oEPA
United States
Environmental Protection
Agency
Municipal Environmental Research EPA-600 2-80-097
Laboratory August 1980
Cincinnati OH 45268
Research and Development
Hindrance of
Coliform
Recovery by
Turbidity and
Non-Coliforms
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution-sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-80-097
August 1980
HINDRANCE OF COLIFORM RECOVERY BY TURBIDITY AND NON-COLIFORMS
by
Diane S. Herson
University of Delaware
Newark, Delaware 19711
and
Hugo T. Victoreen
Wilmington Water Department
Wilmington, Delaware 19801
Grant No. R805102
Project Officer
Edwin E. Geldreich
Drinking Water Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, 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.
ii
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled
land are tragic testiony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its
components require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem
solution and it involves defining the problem, measuring its impact, and
searching for solutions. The Municipal Environmental Research Laboratory
develops new and improved technology and systems for the prevention,
treatment, and management of wastewater and solid and hazardous waste
pollutant discharges from municipal and community sources, for the
preservation and treatment of public drinking water supplies, and to
minimize the adverse economic, social, health, and aesthetic effects of
pollution. This publication is one of the products of that research; a
most vital communications link between the researcher and the user
community.
This paper describes studies done on factors which affect
interactions between coliforms and non-col iforms isolated from the
distribution system. Additional studies were done on the effect of
non-bacterial turbidity on coliform recovery. The data obtained provides
information to all investigators who are concerned with water quality and
who are responsible for interpretation of microbiological data.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
iii
-------
ABSTRACT
The principal objectives of this project were to evaluate the
recoverability of coliforms from waters which have a) high populations of
non-coliform organisms and b) high levels of turbidity due to natural
mineral turbidity, hydrated oxides and organic debris.
The approach to the first objective was to isolate and identify
coliform and non-coliform organisms from Brandywine River source water and
from the Wilmington Water Department distribution system. Sampling was
done at various times of the year to get as wide a range of organisms as
possible.
After initial isolation and identification of coliforms and
non-col iforms, our studies concentrated on the interactions between these
two groups of organisms. When interactions between E^ coli,
Flavobacterium sp., Acinetobacter sp., Arthrobacter sp. and a CDC group II
K-I sp. were studied ail organisms except Tor the latter one were found to
be capable of inhibiting the coliform. The greater the numbers of
non-col iforms the more pronounced the inhibition. The medium in which the
interaction was studied also was important. The standard growth medium
used was a low nutrient bacterial extract produced from organisms
originally isolated from dead end water. Supplementation of this extract
with a salts mix and phosphate buffer appeared to protect the coliforms
from inhibition by the non-col iforms. Supplementation with washed
tubercle solids appeared to have no effect.
Another important factor in determining the outcome of an interaction
was the physiological status of both members. Coliform inhibition was not
observed unless the non-coliform was able to grow and increase in numbers
in the growth medium. This was also true for the coliform member of the
interaction. In those studies where the coliform control numbers did not
increase, but were only maintained in the eight day course of the
experiment, inhibition of the coliforms by the non-col iforms was no longer
observed in the coliform: non-coliform mix. Therefore coliforms may be
less vulnerable to inhibition or stress by non-coliform when either member
of the interaction is not actively growing. However, when inhibition is
occurring, preliminary data indicates that the coliforms are being
stressed by the non-coliforms. Under these conditions recovlry on
selective media is less than recovery on non-selective media.
A series of experiments, some with raw water and some with
distribution system water, were set up to distinguish turbidity inhibition
of coliforms from the inhibition caused by other bacteria. It was found
that within those turbidity limits acceptable to the consumer, turbidity
per se was not an impediment to coliform growth, but it did make it more
iv
-------
difficult to recognize coliforms on membrane filters. The strikingly
persistent tendency for larger sample volumes on membrane filters to give
proportionately smaller coliform yields is not attributable to turbidity.
The more serious inhibitions to coliform detection seemed to be caused by
surprisingly large populations of non-coliforms exceeding the resident
coliforms in water mains by factors of 102 to 105.
This report was submitted in fulfillment of Grant #R805102 by the
University of Delaware under the sponsorship of the Drinking Water
Research Division, MERL, U. S. Environmental Protection Agency. This
report covers the period March 13, 1977 to August 15, 1979, when work was
completed.
-------
CONTENTS
Foreword iii
Abstract iv
Figures v11i
Tables x
Abbreviations and Symbols xi
1. Introduction -. 1
2. Conclusions 3
3. Recommendations 5
4. Experimental Procedures 6
Isolation of Organisms 6
Interaction Experiments 9
Organisms to be Used 9
Determination of Numbers 9
Growth Medium 9
Standard Conditions 14
Turbidity Experiments 49
Studies Using Raw Water 56
Studies Using Distribution Water 58
Studies Using Recirculating Distribution
Water from an Excised Main 60
References * . . . . 64
vi i
-------
FIGURES
Number Page
1 E_. coli and Flavobacterium sp. (1:1) in bacterial extract .... 17
2 E,. coli and Flavobacterium sp. (1:1) in bacterial extract .... 18
3 £. coli and Flavobacterium sp. (7:1) in bacterial extract .... 19
4 E. coli and Flavobacterium sp. (10:1) in bacterial extract ... 20
5 E. coli and Flavobacterium sp. (1:14) in bacterial extract ... 21
6 £. coli and Flavobacterium sp. (1:16) in bacterial extract ... 22
i
7 £. coli and Flavobacterium sp. (1:2) in bacterial extract
~ supplemented with salts and phosphate buffer 24
8 E. coli and Flavobacterium sp. (1:1) in bacterial extract
~ supplemented witn salts and phosphate buffer 25
9 £. coli and Flavobacterium sp. (1:8) in bacterial extract
~ supplemented with salts and phosphate buffer 26
10 E. coli and Flavobacterium sp. (1:16) in bacterial extract
~" sTippTemented witn salts and phosphate buffer 27
11 £. coli and Flavobacterium sp. (6:1) in bacterial extract
~ supplemented witn salts and phosphate buffer 28
12 £. coli and Flavobacterium sp. (9:1) in bacterial extract
"" supplemented with salts and phosphate buffer 29
13 E. coli and Flavobacterium sp. (1:1) in bacterial extract
"" supplemented witn pnospnate buffer 30
14 E. coli and Flavobacterium sp. (1:2) in bacterial
"~ extract supplemented with phosphate buffer . 31
15 E. coli and Flavobacterium sp. (1:14) in bacterial
" extract supplemented with phosphate buffer 32
16 E. coli and Flavobacterium sp. (1:17) in bacterial
~" extract supplemented with phoshate buffer 33
(Continued)
vi ii
-------
FIGURES (Continued)
Number Paqe
17 £. coll and Flavobacterium sp. (8:1) in bacterial
~" extract supplemented witn phosphate buffer 34
18 E. coli and Flavobacterium sp. (5:1) in bacterial
" extract supplemented with phosphate buffer 35
19 E. coli and Acinetobacter sp. (1:10) in bacterial
" extract 36
20 E. coli and Acinetobacter sp. (1:10) in bacterial
~ extract supplemented with salts and phosphate buffer 37
21 Growth of E. coli and Arthrobacter sp. in bacterial extract ... 39
22 E. coli and Arthrobacter sp. (1:1) in bacterial extract 40
23 £. coli and Arthrobacter sp. (1:10) in bacterial extract .... 41
24 Growth of E. coli and Arthrobacter sp. in bacterial extract
supplemented with salts and phosphate buffer 42
25 E. coli and Arthrobacter sp. (1:1) in bacterial extract
"" supplemented witn salts and phosphate buffer 43
26 E. coli and Arthrobacter sp. (1:10) in bacterial extract
~ supplemented with salts and phosphate buffer 44
27 Growth of Arthrobacter sp. in various media 45
28 Growth of E. coli in various media 46
29 £. coli and Arthrobacter sp. (1:10) in bacterial extract 47
30 £. coli and Arthrobacter sp. (1:10) in bacterial extract
~ supplemented with crushed tubercle solids 48
31 £. coli and Flavobacterium sp. (1:1) in bacterial extract .... 50
32 £. coli and Flavobacterium sp. (1:10) in bacterial extract .... 51
33 Growth of E. coli and Arthrobacter sp. on bacterial extract
at 6.60C". TT. 52
34 E. coli and Arthrobacter sp. (1:1) on bacterial extract
" an>760C .TTTTTT 53
IX
-------
TABLES
Number Page
1 Organisms Isolated From Raw Water 6
2 Organisms Isolated From Finished Water . . . .^ 8
3 Interaction Experiments 14
4 A Comparison of Bacterial Recovery on Plate Count Agar
and M-5 Agar 56
5 Coliform Recovery From Good Quality Raw Water 57
6 Coliform Recovery From Intermediate Quality Raw Water ... 58
7 Coliform Recovery From Distribution Water 59
8 Coliform Recovery From Distribution Main Water 60
9 Coliform Recovery From Recirculating Main Water 60
10 Coliform Recovery From a Combination of Distribution Water
and Recirculating Main Water 61
11 Coliform Recovery From a Combination of Distribution Water
and Recirculating Main Water 62
-------
LIST OF ABBREVIATIONS AND SYMBOLS
BGB — Brilliant Green Bile Broth
LES-Endo ~ An Endo agar formulated for membrane filter work
LS — Lauryl sulphate (Lauryl Tryptose) broth
M-Endo Broth ~ A broth designed for membrane filter work
MF — Membrane filter or membrane filter method
M-5 — A low nutrient, high iron, plate count agar
MPN — Most Probable Number technique, usually used with LS
broth in coliform enumeration
PCA ~ Plate Count Agar
VRBA — Violet Red Bile Agar
xi
-------
SECTION 1
INTRODUCTION
One coliform per 100 ml of water is the standard currently in use for
drinking water. Situations have been known to exist in the water
distribution system of the Wilmington Water Department where 1-2 coliforms
per 100 ml had to be isolated from water containing a total of 100 to
100,000 total organisms per ml. Attempts were made to approximate these
as well as other conditions in the distribution system in our study of
coliform:non-coliform interactions. Standard procedures for these studies
were established using organisms isolated from the distribution system
growing in a bacterial extract prepared from dead end water isolates. The
experiments reported attempted to determine the effects of:
1. the specific non-col iform
2. the ratio of coliform:non-coliform
3. the media used
4. the physiological status of each member on the coliform:
non-col iform interaction.
In addition, coliforms must be detected in waters containing high
levels of turbidity due not only to high bacterial numbers but also to
natural mineral turbidity, hydrated oxides, and organic debris. This
situation, which is not unique to the Wilmington Water Department, serves
to complicate detection of coliforms by routine analysis. Natural mineral
turbidity, hydrated iron oxides, organic debris, and the live cells of
other bacteria can all form mats on the MF surface from as little as 100
ml of a water sample. Their presence may inhibit coliforms, render the
Endo components less inhibitory toward non-coliforms or promote confluent
growth. In addition, if the turbidity is due to particles with embedded
bacteria, these bacteria may be protected from the effects of
chlorination. Inhibition of coliforms by non-coliforms can potentially
pose a serious health hazard if any of the non-coliforms are pathogenic.(1)
In the turbidity studies the MF and MPN techniques were employed for
coliform detection. Both methods were used side by side wherever
possible. This was because it was assumed that the MF technique would be
vulnerable to turbidimetric interference and that the MPN technique would
be almost immune to such interference, but extremely vulnerable to
non-col iform bacterial interference.
Most of the turbidity experiments were carried out in water from the
distribution system frequently blended from several districts, and
some-times containing water from a 2 meter cross section of 150 mm main
cut from a neighborhood experiencing coliform regrowth problems. This
1
-------
isolated main was kept active for many months by circulating small amounts
of dead-end water through it with a peristaltic pump.
The experiments reported may be divided into 2 groups:
(1) Turbidity augmentation experiments where coliforms and the
general bacterial population were measured before and after the
addition of a chlorine-free main deposit. This adds a turbidity
which covers a broad spectrum, as it was derived from water
mains by way of hydrants on 150 mm branches.
(2) Turbidity reduction experiments where coliforms and total
bacteria were measured before and after a low speed
centrifugation designed to remove most of the turbidity while
leaving most of the bacteria in the supernatant. The
centrifuging procedure subtracted the largest and heaviest
particles so that the residual turbidity was probably more
organic, smaller in particle size, and lower in specific gravity
than the original turbidity. Use of both of these techniques
subjected standard coliform detection procedures to a wide
variety of turbidimetric interference.
A determination of the effects of the above factors on coliform
recovery would provide important information to all investigators who are
concerned with water quality and who are responsible for interpretation of
microbiological data. Ultimately this information may be used to provide
further insights into alternate methods of water treatment which would
diminish or eliminate the high numbers of microbes and high levels of
turbidity. It may also be used to develop techniques to enable the
detection of coliforms in water containing high numbers of non-coliforms
and/or high turbidities so that this water is not erroneously thought to
be safe.
-------
SECTION 2
CONCLUSIONS
A wide variety of coliform and non-coliform organisms were isolated
from Brandywine River water and from the Wilmington Water Department
distribution system. Four of these isolates (Flavobacterium sp.,
Acinetobacter sp., Arthrobacter sp. and a CDC tiroup 11 K.-I sp.) were
tested in interaction experiments for their ability to inhibit E. colI i.
The coliform was inhibited by all the non-coliforms except for TheTDT
Group II K-l isolate. Inhibition of the coliform was observed to be
dependent upon a number of factors. When the coliform: non-coliform ratio
was decreased from 1:1 to 1:10 the inhibition was greater as determined by
the recoverability of coliform colonies on plate count agar. The medium
in which the interaction was studied was also important. Dead end water
from the distribution system was found to be unsatisfactory as a growth
medium because of the inability to have a coliform control without
altering the water in the process of ridding it of the non-coliform
indigenous population. The growth medium used was a low nutrient
bacterial extract produced from organisms originally isolated from dead
end water. Supplementation of this extract with a salts mix and phosphate
buffer appeared to protect the coliforms from inhibition by the
non-coliforms. Supplementation with washed tubercle solids appeared to
have no effect on the coliform:non-coliform interaction studied.
Another important factor in determining the outcome of a coliform:
non-coliform interaction is the physiological status of both members.
This is demonstrated by comparing the growth of the coliform and
non-coliform controls (each species growing alone in the test medium) with
the growth of each of the two species when they are mixed together in the
same medium. Coliform inhibition is not observed unless the non-coliform
control is able to grow and increase in numbers in the growth medium.
This is also true for the coliform member of the interaction. In the
studies where control coliform numbers are not observed to increase, but
are only maintained in the eight day course of the experiment, inhibition
of the coliform by the non-coliform is no longer observed in the coliform:
non-coliform mix. Therefore it appears that coliforms may be less
vulnerable to inhibition or stress by non-coliforms when either member of
the interaction is not actively growing. However, when inhibition is
occurring preliminary data indicates that the coliforms are being stressed
by the non-coliforms. Under these conditions recovery on selective media
is less than recovery on non-selective media.
Turbidity per se is not the major hindrance to the growth of
coliforms on memBrane filters. On the other hand it does produce colonies
more difficult to recognize as coliforms. With most of the waters
-------
studied, proportionately larger quantities of water filtered through
membrane filters do not give proportionately higher yields of coliform
colonies. This phenomenon is still observed when the turbidity of a water
sample is reduced by low speed centrifugation which does not remove a high
proportion of the bacterial density. It appears then that the inhibition
of coliforms is due more to the presence of large numbers of non-col iform
bacteria rather than to the suspended matter present. The intercolonial
spaces on incubated MF Endo membrane filters contain coliforms which have
been inhibited and have not produced discernible coliform colonies. The
inhibitory influence of non-coliforms upon coliform detection has probably
escaped notice because the total bacterial population has not been
measured. The M-5 agar medium used in one turbidity study gives very high
colony yields, but other low nutrient media may serve as well.
-------
SECTION 3
RECOMMENDATIONS
Interactions between coliforms and non-col iforms can be expected to
occur in the distribution system. The outcome of the interaction is
dependent upon numerous factors including the specific non-coliform
present, the numbers of each type of organism, the nutrient environment in
which the interaction is occurring, the physiological status of the
interacting organisms and the type of media used to recover the
coliforms. It is therefore important that investigators testing for the
presence of the indicator coliform organisms be aware of these factors and
their influence on coliform recovery.
Additional studies on the coliform: non-coliform interaction should
be done. Our preliminary data indicates that when grown in the presence
of non-col iforms, the coliforms may be stressed. If this is so, the
ability to recover them using some standard techniques will be
diminished. In addition, waters containing high numbers of non-col iforms
in the absence of coliforms may not be considered safe until further
studies are done to eliminate the possibility of the presence of stressed
coliforms.
Purveyors of drinking water should concentrate their bacterial
monitoring (efforts) in the chlorine-free or low chlorine outskirts of
distribution systems rather than in the fresh output of wells and water
treatment plants.
Wherever turbidity is such that the filtration of 100 ml leaves a
conspicuous deposit on a membrane filter, the MPN method should be used.
If particular sampling points consistently yield heavily turbid
cultures in the MPN test, transfers should be made to B6B even though gas
is not present at the presumptive stage.
Where chlorine is low or absent a long term low nutrient plate count
should be employed to determine the size of the background bacterial
population within which coliform enumeration is being attempted.
-------
SECTION 4
EXPERIMENTAL PROCEDURES
ISOLATION OF ORGANISMS
Organisms were isolated by the Wilmington Water Department and
identified at the University laboratory. Primary media for coliforms
included Lauryl Tryptose, Mac Conkey, Eosin Methylene Blue and
Desoxycholate Agar. Plate count agar and M-5 agar were used for
non-coliforms. All of the organisms were routinely Gram stained and
tested in the API 20E system. The definition of coliform used here is the
same as that in Standard Methods, 14th ed. (2).
In those cases where the API did not yield a number usable for
identification of an isolate, Sergey's Manual (3) was used as a guide to
determine the genus and if possible, species. Supplementary biochemical
tests were done using standard procedures from the Manual for Clinical
Microbiology (4) and Diagnostic Microbiology (5). In addition, American
Type Culture Collection samples were used to further confirm our
identifications. Tables 1 and 2 list the identity of the isolated
organisms according to Geldreich et al.'s classification of total
coliforms, coliform antagonists, opportunistic pathogens and category not
established (6) and according to von Graevenitz's classification of
opportunistic pathogens (7).
TABLE 1. ORGANISMS ISOLATED FROM RAW WATER
Category
1. Total coliforms
Organism name
Enterobacter cloacae
Citrobacter freundii
Enterobacter agglomerans
Escherichia coli
Klebsiella pneumoniae
API
1305573
3305723
3305573
3305773
3305763
1305173
3704553
1604532
1205573
1004353
1244573
1297573 '
5144572
1144573
1144572
714555.2:.
1205773:!
(Continued)
-------
TABLE 1. (Continued)
Category
Organism name
API
2. Coliform antagonists
3. Opportunistic pathogens
4. Category not established
Klebsiella pneumoniae
Pseudomonas maltophilia
Pseudomonas aeruginosa
Kseuaomonas" sp. 2
I- lavobacterium sp.
Bacillus sp.
*AGinetobacter sp.
*Arthrobacter sp.
Actinomycetes
Pseudomonas aeruginosa
Pseudomonas" mal tophi-TTa
Klebsiella pneumoniae
Acinetobacter calcoaceticus
Aeromonas nyarophina
**Baci11us sp.
**Corynebacterium sp.
r i avooac ter 1 um "sp.
Streptomyces sp.
5255773
5215773
5255763
020000040
020200040
2202000
220000451
000000040
000200010
30071234
22020044
02030000
0202000
0201000
100717357
0 API
2204000
0002000
1203000
040200400
0 API
2202000
020000040
020200040
1205773
5255773
5215773
5255763
0201000
70471245
33471775
30461045
30071234
22020044
020300000
0202000
0201000
0002004
000000040
000200010
1002000
1002000
1003000
0203000
(Continued)
-------
*
**
TABLE 1. (Continued)
Interaction studies with these organisms described in this report
indicate they are also coliform antagonists.
These organisms are considered opportunistic pathogens by von
Graevenitz (7), but were placed in category not established by
Geldreich et al. (6).
TABLE 2. ORGANISMS ISOLATED FROM FINISHED WATER
gory
urqanism name
ftrl
1. Total coliforms
2. Coliform antagonists
3. Opportunistic pathogens
Enterobacter cloacae
Citrobacter freundii
Enterobacter agglomerans
Escherichia coli
MeDsiefTa'"pneumoniae
Pseudomonas maltophilia
rseuaomonas" sp. z
Flavobacterium sp.
Baciilus sp.
*Aclnetobacter sp.
*Arthrobacter sp.
Pseudomonas maltophilia
KI ebs 1 e i TT"pneumon i ae
Serratia liquifasciens
noraxeila sp.
Aeromonas hydrophilia
**Bacil1us sp.
**uoryneDacterium sp.
(Continued)
3305533
3305773
3305573
3205533
3704553
1344573
1205573
1005573
1144572
5255773
5215773
100000040
220000451
000200451
0 API
2202000
11051735
0204040
33051531
0 API
4200000
1000000
1003000
0001000
0002000
1002000
100000040
5255773
5215773
53077635
000000440
000100400
2000000
70071075
3047526
2202000
00000004
004300004
8
-------
TABLE 2. ORGANISMS ISOLATED FROM FINISHED WATER
category
4.
uppor turn stic pathogens
Category not established
Organism name
Mavobacterium sp.
uuu Group TTT-i
Yersinia pseudo tuberculosis
API
0 API
000000442
1014153
* Interaction studies with these organisms described in this report
indicate they are also coliform antagonists.
** These organisms are considered opportunistic pathogens by von
Graevenitz (7), but were placed in category not established by
Geldreich et al. (6).
Analysis of the isolated organisms indicates that the same species
are found in finished and raw waters. Numerous organisms categorized as
opportunistic pathogens are found in both waters. As noted on the tables
our interaction studies indicate that Acinetobacter sp. and Arthrobacter
sp. can be added to the list of co1iform antagon 1ists. In agreement witfr
Geldreich et al. (6) we found Flavobacterium sp. also to be a potential
coliform antagonist.
INTERACTION EXPERIMENTS
In our study of interactions between coliforms and non-coliforms
standard conditions had to be established. Each of the following was
cons i dered:
Organisms to be Used
The organisms used in the study were isolates from the Wilmington
Water Department distribution system. Their identification is indicated
in the previous section.
Determination of Numbers of Coliforms and Non-Coliforms
For those studies where coliform and non-col iform numbers were equal
or relatively close, Plate Count Agar (PCA), was used to determine total
numbers of bacteria and Endo Agar was used to enumerate coliforms. When
coliforms were grown alone, their numbers on PCA correlated well with
their numbers on Endo Agar.
Growth Medium
An organism's growth characteristics vary depending on the type and
concentration of medium used. It is expected that their interaction with
other organisms will vary accordingly. Since the distribution system is a
low nutrient environment we attempted to duplicate this condition in our
experiments.
-------
Dilute Nutrient Broth-
Concentrations of 2 x 10'3 and 1 x 10"3 nutrient broth were used
in preliminary interaction screening experiments. Use of this medium was
discontinued because it did not qualitatively or quantitatively duplicate
the water environment.
Dead end Water From the Wilmington Water Department Distribution System—
This water was chosen because:
1. It is from the distribution system.
2. It supports the growth of microbes.
3. It may be that in this water coliform recovery is hindered by the
presence of non-col iforms.
The major problem with using this water untreated is that it is
impossible to have a proper control for coliform growth alone. The
untreated (i.e., not filtered or autoclaved) water has microbes present
which may affect coliform growth or recovery. Procedures used to remove
these indigenous organisms would also alter the water chemically or
physically. The growth of the coliforms alone in dead end water in the
absence of the other organisms but with everything else present can
therefore not be accurately assayed. We felt that under these conditions
a way to estimate coliform control numbers would be to determine the
growth of these organisms in untreated dead end water and in autoclaved
dead end water.
As a preliminary experiment we found these studies to be very useful,
but decided against continuing the use of dead end water because of the
problem noted above with the coliform control. In all cases coliforms
were undetectable by MPN assays by the sixth day of the experiment. We
did not know if this was due to the inability of the water to support the
growth of these organisms or if it was due to the inhibitory action of
other organisms. The autoclaved water control was also not able to
support the growth of coliforms. However, this may have been because of
the effect of autoclaving which may have destroyed some organic material
and/or produced toxic materials.
Bacterial Extract—
Our studies using dead end water as a growth medium led us to
conclude that problems with setting up a proper coliform control far
outweighed the advantages of using dead end water. We felt that in
addition to the above noted problem, the dead end water would most likely
vary chemically from time to time during the course of a year. We
therefore decided to prepare a medium in which to grow the organisms.
'»
Since lysed bacterial cells may themselves be providing nutrients for
other organisms in the distribution system, (8) we decided to prepare
bacterial extracts and test the ability of E^ coli (C 29) and a
non-coliform mix made up of organisms originally isolated from the
distribution system to grow in these extracts.
10
-------
Preparation of bacterial extracts—
M. tariy (\og pnasej non-coiiTorm extract
(1) Seven non-col iform organisms were isolated from the dead
end water tested in the previous experiment. They were
identified as:
1 Moraxella sp., 1 Flavobacterium sp., 3 Arthrobacter sp.,
1 Staphyloc'occus saprophytlcus, and 1 Strep tomyces sp.
(2) Ali tne organisms were grown on PCA plates, colonies were
scraped and used to inoculate 1 liter of nutrient broth in
2 liter flasks. The flasks were allowed to shake at room
temperature for 9 hours at which time the cultures were in
logarithmic phase.
(3) Cells were harvested in 150 ml buckets spun at 10,000 RPM
for 15 min. The cell preparations were kept on ice for the
remainder of the proceaure.
(4) The cell pellet was resuspended and washed 3 times in
distilled water. The pellets were drained, dried and the
wet weights determined. A 1:5 ratio (cell weight:distilled
water) was prepared for the next step in which the cell
suspension was put through a Carver Laboratory press at
15-20 K Ibs/square inch and then spun at 5 K for 5 minutes.
(5) The supernatant obtained was saved and sonicated with a
microprobe for 8-15 second bursts (Sonifier Cell Disruptor,
Heat Systems Co.).
(6) The supernate was then filtered through Whatman #1,
Millipore 8 urn 0.8 urn and 0.45 urn filters. The final
0.45 m filtration was to sterilize the extract and the
resulting filtrate was aseptically transferred to small
sterile tubes in 1 ml aliquots and frozen at -5°C.
(7) The optical density of an aliquot was read at 260 and 280
nm and compared with that of a 10~3 concentration of
nutrient broth. This was the basis for determining
concentration levels to be used in future interaction
studies. Prior to addition to samples, aliquots of the
extact were re-filtered through a sterile Swinney filter
unit fitted with a 0.45 urn filter.
B. Late (stationary phase) non-coliform extract:
The above extract is designated "early," the cells still being
in logarithmic growth. A second extract termed "late" was
prepared as described above, except in step 2, the cultures were
grown for 60 hours before harvesting.
C. E. coli extract:
An E. coli extract was also prepared using a frozen paste of E.
colT~B mid-log phase cells obtained from Grain Processing Co.,
nuscatine, Iowa. It was prepared as described above.
The bacterial extracts prepared as described above were capable of
11
-------
supporting the growth of the coliform and non-coliform isolates. The
concentration of extract necessary to support the growth of the control
organisms was determined for each preparation.
Chemical analysis of bacterial extract—As noted earlier, the amount
of bacterial extract aaaea was Dased on the 260/280 ratio of the extract
compared to 10~3 nutrient broth. To gain more information on the
composition of the extract, the following chemical and biochemical
determinations were made.
Total Carbohydrate Determination:
Using a modification of Hassid and Abraham's starch and glycogen
analysis(9), it was possible to determine the amount of total carbohydrate
(ug/g wet weight) and the amount of glycogen (ug/g) in the bacterial
extact.
1. A 2.0 ml aliquot was digested with 1 ml of 6036 KOH.
2. 0.1 ml of this solution was brought up to 5 ml with distilled
water.
3. One ml and 0.5 ml aliquots were then tested by adding 10 ml
Anthrone reagent.
4. Tubes were swirled gently and boiled for 10 min, then immersed
in ice and cooled to below room temperature.
5. Readings were done at 620 nm and compared to a standard curve
for glucose.
We found 34.4 ug total carbohydrate/g wet weight of extract and 30.87 ug
glycogen/g.
(10).
Protein Determination:
The amount of protein (ug/ml) was determined using the Bio-Rad method
1. A stock of BSA (60 mg/ml) was appropriately diluted for the
standard curve.
2. An aliquot of extract was diluted to fall within a straight line
portion of the standard curve.
3. Five ml of Bio-Rad reagent was added to each 0.1 ml of test and
standard sample.
4. Samples were read at 595 nm.
One extract prepared 1-4-79 had 41.4 ug protein/ml while another prepared
4-20-79 contained 347.5 ug protein/ml. The 4-79 extract was, 6 times more
concentrated on a g wet weight/ml extract basis.
DNA Determination:
This method allows the direct measurement of DNA (ug/ml) by
fluorescent detection (11). It only works on double stranded DNA. A
drawback to this method is that it should be used on freshly broken cells
12
-------
so there is a minimum of nicking of the DNA by nucleases. Since nuclease
inhibitors were not added to the bacterial extracts and since the length
of time from the preparation of the extract to the DNA determination
varied from days to months, the value of this test is mainly to show the
presence of DNA in the extract.
1. The sample should have a 280 nm O.D. of 0.05. This is in a
final volume of 1.8 ml.
2. 100 ul of 300mM MgCl2 is added, the tube is shaken, 100 ul of
mithramycin is added and the tube is again shaken.
3. Incubate at room temperature 30 min. to 18 hr. We found the
greatest stabilization of readings after 18 hrs.
4. Calf thymus DNA at a concentration of 12 ug/ml was used as the
standard. In our extracts we had 560 ug DNA/ml. This is about
10% of that expected in freshly sonicated cells.
Supplementation of the bacterial extract—in some experiments the
bacterial extract is supplemented, me rationale for these additions is
discussed below.
1. Salts + phosphate buffer:
The salt mix + phosphate buffer solution used is the one described in
Standard Methods page 888 (2) for testing the purity of distilled water.
Since this mix is used when coliform growth is being tested, we thought it
might enhance the growth of the coliforms when they were growing in a low
nutrient medium.
2. Phosphate buffer:
This serves as a control for the addition of salts + phosphate. If
the presence of salts + phosphate did have any effect on the coliform:
non-coliform interaction adding back phosphate alone would indicate if the
difference was due to the presence of the phosphate buffer.
3. FeS04:
i
As is discussed later, it was observed that the salts + phosphate mix
did affect the coliform:non-coliform interaction. FeS04 was added to
the bacterial extract to determine if the observed effect was due to the
iron containing salt in the mix.
4. Tubercle solids:
Tubercles are present in most water mains. The material from the
tubercles may leach into the distribution water thereby becoming available
to organisms in the system. Tubercle solids were added to determine if
they would in any way affect the coliform:non-coliform interaction.
The effect of various factors on the coliform:non-coliform interaction
The following factors were considered:
13
-------
a. the specific non-coliform
b. the ratio of coliforms to non-col iforms
c. the growth medium used
d. the presence or introduction of other factors believed to be
potentially stressing such as the isolation media.
The specific experiments which contributed to our information on
col iform:non-coliform interactions are listed in Table 3.
TABLE 3. INTERACTION EXPERIMENTS
igures
isms
or uonaitions
1. E_. coli^ and Flavobacterium sp.
2. £_._ coli and Flavobacterium sp.
3. E^ coli and Flavobacterium sp.
4. £_._ coli and Acinetobacter sp.
5. £. coli and Acinetobacter sp.
6. E_^ coli and Arthrobacter sp.
7. E^_ coli and Arthrobacter sp.
8. £._ coli and Arthrobacter sp.
9. £_._ coli and Flavobacterium sp.
10. E. coli and Arthrobacter sp.
Bacterial extract 1-6
Bacterial extract +salts 7-12
mix + phosphate buffer
Bacterial extract + 13-18
phosphate buffer
Bacterial extract 19
Bacterial extact + salts 20
mix + phosphate buffer
Bacterial extract 21 - 23
Bacterial extract + 24-26
salts mix + phosphate
buffer
Bacterial extract, 27 - 30
Bacterial extract +
tubercle solids
Bacterial extract, 31 - 32
physiologically altered
Flavobacterium sp.
Bacterial extract, 33 - 34
grown at 6.6°C.
Standard Conditions for Interaction Experiments
Numerous experiments were done in the course of our study to
determine the conditions and procedures for running the interaction
experiments. The procedures adopted and used throughout the remainder of
our work are detailed below.
1. 250 ml flasks were filled with 100 ml of double distilled deionizefl
water and then autoclaved.
14
-------
2. Bacterial extract was added at a concentration previously determined
to support growth of the controls.
3. The coliforms and non-coliforms were streaked on PCA plates and
incubated overnight. Colonies were scraped into sterile centrifuge
tubes with buffered water. Each organism was washed three times.
Absorbance was read at 650 nm to determine the number of cells/ml.
In those experiments where different ratios of coliforms to
non-coliforms were tested, dilutions were made in an attempt to have
the following ratios of col iforms:non-col iforms: 1:1, 1:10, 1:0,
0:1, and in some cases 10:1, in the test flasks. We also attempted
in all cases to have a total of 1.1 x 10^ organisms inoculated per
flask.
4. Ratios were run in duplicate. Aliquots from each flask were plated
in triplicate.
5. Plates were incubated and counted after 24 - 48 hours.
6. Controls: the flasks containing each organism growing alone (1:0 and
0:1 ratios) served as controls.
In addition to the establishment of standard procedures we had to run
experiments every time a new extract was prepared to determine the proper
levels of extract to use. When in the course of our experiments we found
that the control organisms were not growing as well as they had in
previous experiments we had to determine if the extract had been altered
with time and was therefore not capable of supporting growth as
effectively or if the growth characteristics of the isolate had changed
with time. Both situations were observed. These control experiments are
not included in this report.
The Effect of Specific Non-Col if orms on Coliforms—
The non-coliforms tested were: Flavobacterium sp.
Arthrobacter sp.
Acinetobacter sp.
cuu group 11 K-l
Under specific conditions which will be expanded upon below we were
able to demonstrate inhibition of the coliforms by all of the
non-coliforms except for the CDC group II K-l organism. Our operational
definition of inhibition is low levels of growth or recovery of coliforms
when they are growing in combination with non-coliforms as compared to
recovery or numbers obtained when the coliforms are growing alone in the
same media.
The Effect of the Ratio of Coliforms:Non-Coliforms and the Medium Used on
Coliform Recovery—
E. coli and F-lavobacterium sp. interactions were tested in 1:1, 1:10
and 15:1 ratios in non-coiiTorm extract, and the same extract supplemented
with an E. coli salts mix plus phosphate buffer, and with phosphate buffer
alone. The results are graphed in Figures 1 - 18.
15
-------
The media used and the ratio of co1 iforms:non-col iforms initially
inoculated were both found to have a profound effect on the interactions
between the two groups of organisms tested. The day 8 data has been
statistically analyzed by comparing the growth rates of the col iforms when
they were grown in 1:0, 1:1, 1:10 and 10:1 ratios with non-col iforms.
This was done by running an analysis of variance followed, when necessary,
by the Student-Newman-Keuls test (12).
When just the coliform control growth rates are compared it is found
that they grow best in the bacterial extract supplemented with salts +
phosphate, followed by the bacterial extract alone and least well in the
extract supplemented with phosphate.
When the coliforms were inoculated into early non-coliform extract
alone (1:0 ratio = control) or in 1:1, 10:1, and 1:10 ratios with
non-col iforms the profiles in Figures 1-6 were observed. The statistical
analysis which compared the growth rates of the coliforms growing in the
presence of different numbers of non-coliforms indicated the following:
1:0 (control) 1:1 10;1 1:10
This means that the growth rate of the coliform control was significantly
different (at the .05 level) than the growth rates observed when the
coliforms were growing in the presence of non-coliforms. When growing in
the presence of non-coliforms, the growth rates of the coliforms in 1:1
and 10:1 coliform:non-coliform ratios were not significantly different
from each other. In addition, the growth rate of the coliforms in a 1:10
coliform: non-coliform ratio was significantly different from the growth
rate of the coliforms when they were in any of the other ratios tested.
In these experiments it was necessary to use numbers of organisms
that are higher than those observed in the distribution system in order to
more fully characterize the coliform/non-coliform interaction. For
example, observation of figures 1-4 indicate that in non-coliform extract
when the coliforms are added in equal numbers and even when they are added
at approximately a ten fold advantage in numbers over the non-coliforms
they are inhibited. Coliform inhibition is more pronounced when the
non-coliforms are initially given an approximate ten fold advantage in
numbers (Figures 5 and 6). Under these conditions coliforms could not be
recovered from the mix by day 4. By using higher densities of
non-coliforms it was possible to follow the course of inhibition and
determine the difference in coliform recovery when the non-coliform
numbers were increased. This data also points up the importance of the
determination of total bacterial numbers in water testing as is done by
the Standard Plate Count since increased non-coliform levels appear to be
a factor in coliform inhibition.
For the early non-coliform extract + salts + phosphate medium
(Figures 7 - 12) our statistical analysis indicated the following:
I:0(contro1) 1:1 1:10 10:1
16
-------
en
LJJ
o
5xl05-
O £. coli alone
• £. coli in mix
D Flavobacterium alone
Flavobacterium in mix
5 xI03-
IxlO
01 4 6 8
DAYS
Figure 1. £. coli and Flavobacterium sp. (1:1) in bacterial extract.
17
-------
co
UJ
5xl06<
I x I06J
5xl05
I x I05-
5xl04-
0 5xlO°J
O £. coli alone
• £. coli in mix
D Flavobacterium alone
• Flavobacterium in mix
5x10
I x 10
DAYS
Figure 2. E. coli and Flavobacterium sp. (1:1) in bacterial extract.
18
-------
5xl06
O £. coli alone
• £. coli in mix
D Flavobacterium alone
Flavobacterium in mix
Ix 10
SxlO1-
Ix 10
8
4 6
DAYS
Figure 3. £. coli and Flavobacterium sp.» (7:1) in bacterial extract.
19
-------
CO
I x I07-
5 x I06«
Ix I06H
5xl05
5x10-
txlO
O £. coli alone
• E. coli in mix
D Flavobacterium alone
• Flavobacterium in mix
0 I
8
Figure 4. £. coli and Flavobacten'um sp. (10:1) in bacterial extract.
20
-------
I x 10
I x I0-
5 x I0
O £. coli alone
• £. coli in mix
D Flavobacterium alone
• Flavobacterium in mix
Ix 10
8
0 I 4
DAYS
Figure 5. £. coli and Flavobacterium sp. (1:14) in bacterial extract.
21
-------
CO
UJ
o
IxlO7
5xl06 -
1x10°
. coli in mix
DFTavobacterium alone
Flavobacterium in mix
4 6
DAYS
8
Figure 6. £. coli and Flavobacterium SP. (1:16) in bacterial extract,
22
-------
This means there was no significant difference in the coliform growth
rates when they were grown alone or in combination with non-col iforms in a
1:1 or 1:10 ratio.
When the col iforms were grown in a 10:1 ratio of
coliforms:non-coliforms their growth rates were significantly different
from the growth rates observed for the col iforms in the other ratios
tested.
Observation of figures 7, 8, 9 and 10 indicates that when
non-coliform extract is supplemented with salts + phosphate the col iforms
are not inhibited by the non-coliforms when the two are grown together in
a 1:1 or 1:10 coliform: non-coliform ratio. When the coliforms are given
an initial ten fold advantage in numbers under these conditions they are
not inhibited and their growth rate in the mix exceeds that of the control
(Figures 11 and 12).
For the early non-coliform extract + phosphate medium (Figures 13 -
18) our statistical analysis indicated the following:
1:0 (control) 1:1 1:10 10:1
This means there is no significant difference in the growth rates of
the coliforms in the different ratios tested.
This experiment was a control for the addition of salts + phosphate
to see if the protective effect on the coliforms of the salts + phosphate
supplementation was due to the phosphate buffer. Observation of Figures
13 -18 and comparing them to Figures 7-12 indicates this is not the case.
The growth of E, coli and Acinetobacter sp. was tested on bacterial
extract and this extract supplemented with salts + phosphate (Figures 19
and 20). The unsupplemented extract supported high levels of coliform
growth but was unable to support the Acinetobacter which was undetectable
by day four (Figure 19). Growth of each organism in the mix (in a 1:10
ratio of coliforms: non-coliforms) was comparable to that of its
respective control. When the medium was supplemented with salts +
phosphate (Figure 20) the Acinetobacter control growth equalled the E.
coli control growth. In a i:iu conform: non-coliform mix the ~
ftcinetobacter growth equalled control levels but the coliforms were
inniDited. An interesting aspect of this interaction is that the
inhibition of the coliform under the conditions of this experiment appear
to decrease with time. On day four there is a 10^ difference in the
col iform levels reached in the control and test flasks and by day eight
there is only a fifteen fold difference in the levels. It appears that
this medium may be conferring a protective effect on both members of the
interaction.
Another study on interactions between a coliform and a non-coliform
using bacterial extract alone and supplemented with salts + phosphate was
done with E. coli and an Ar throb acter sp. (Figures 21-26). In this
experiment"^/iolet Red Bile Agar ^VKBA), a selective medium, was used to
23
-------
CO
UJ
o
IxlO
5x10'
IxlO
5x10
IxlO
1
O £. coli alone
• £. coli in mix
D Flavobacterium alone
• Flavobacterium in mix
DAYS
1
6
8
Figure 7. E_. coli and Flavobacterium sp. (1:2) in bacterial extract
supplemented with salts and phosphate buffer.
24
-------
CO
_l
-I
UJ
o
IXI07-|
5xl06
IxlO
5x10
IxlO'
O E_. coli alone
• £. coli in mix
D Flavobacterium alone
• Flavobacterium in mix
8
DAYS
Figure 8.
E. ejali and Flaypbacten'um sp. (1:1) in bacterial extract
supplemented with salts and phosphate buffer.
25
-------
CO
LJ
O
5x10
IxlO
5x
Ix
I01-
I01-
• E. coli in mix
G Flavobacterium;
• Flavobacterium
0 1 4
alone
in mix
8
DAYS
Figure 9. £_• coli and Flavobacterium sp. (1:8) in bacterial extract
supplemented with salts and phosphate buffer.
26
-------
IxlO
5x10
IxlO1
O £. coli alone
• £. coli in mix
D Flavobacterium alone
• Flavobacterium in mix
8
DAYS
Figure 10. £. coli and Flavobacterium sp. (1:16) in bacterial extract
supplemented with salts and phosphate buffer.
27
-------
I x I06-
5xl05
5x10
1 x I02-
5x I01-
O E. coli alone
• E. coli in mix
CJ Fl avobacteri urn
• Fl avobacteri urn
0 1 4
alone
in mix
A.
€
Figure 11. E.
DAYS
_E. col 1 and Flavpbacterium sp. (6:1) in bacterial extract
supplemented with salts and phosphate buffer.
28
-------
Lul
O
IxlO
5x10
5x10
I x 10-
O £. coli alone
• £. coli in mix
D Flavobacterium alone
• Flavobacterium in mix
i
4
8
DAYS
Figure 12. E_. coli and Flavobacterium sp. (9:1) in bacterial extract
supplemented with salts and phosphate buffer.
29
-------
CO
UJ
o
IxlO2
5x10'
IxlO1 -
5x10° -
1x10°
I
*> E. coli alone
• E. coli in mix
O Flavobacterium alone
• Flavobacterium in mix
r
4
DAYS
8
wr\ i *J
Figure 13. !E. cgli and Flavpbacterium sp. (1:1) in bacterial extract
supplemented with phosphate buffer.
30
-------
CO
_J
LJ
O
O E_. coli alone
. coli in mix
D Flavobacterium alone
Flavobacterium in mix
IxlO1
DAYS
. (1:2
supplemented with phosphate buffer.
8
Figure 14. £, coli and Flavobacterium sp. (1:2) in bacterial extract
ppleme
31
-------
CO
LJ
O
O £. coli alone
• E. coli in mix
U Flavobacterium alone
Flavobacterium in mix
IxlO1
01 4 6 8
DAYS
Figure 15. E. coli and Flavobacterium sp. (1:14) in bacterial extract
supplemented with phosphate buffer.
32
-------
CO
_J
IxlO7 -
5xl06 -
IxlO6 •
5xl05
IxlO5
5xl04
IxlO4
5xl03
IxlO2
5x10'
IxlO1
O £. coli alone
• £. coli in mix
g Flavobacterium alone
Flavobacterium in mix
DAYS
Figure 16. g. coli and Flavobacterium sp. (1:17) in bacterial extract
supplemented with phosphate buffer.
33
-------
CO
_l
_J
LJ
O
IxlO6
5xl05
O . coli alone
. coll in mix
D FTavobacterium alone
• Flavobacterium in mix
IxlO2
5x10'
IxlO1 | . i
01 4 68
DAYS
Figure 17. £. coli.and Flavobacterium sp. (8:1) in bacterial extract
supplemented with phosphate buffer
34
-------
3
Id
O
O £• coli alone
. coli in mix
D Flavobacterium alone
Flavobacterium in mix
IxlO5
5xl04
IxlO2 -
5x10' •
\,
) 1 4 6 8
DAYS
Figure 18. E_. coli and Flavobacterium sp. (5:1) in bacterial extract
supplemented with phosphate buffer.
35
-------
3
iLl
o
IxlO6
5xl05
T
IxlO5 -
5xl04 -
O £. coli alone
• £. coli in mix
D Acinetobacter alone
• Acinetobacter in mix
0 .
IxlO
Figure 19. E> coli and Acinetobacter sp. (1:10) in bacterial extract.
36
-------
8
IxlO
5xio7 • 2£- ^H-?!°"?
_ coli in mix
DAcinetobacter alone
• Acinetobacter in mix
IxlO7 -
5xl06 •
IxlO6
5xl05
3
UJ
o
IxlO1
DAYS
Figure 20. E_. coli and Acinetobacter sp. (1:10) in bacterial extract
supplemented with salts and phosphate buffer
37
-------
enumerate coliforms and to compare coliform recovery with recovery on
Plate Count Agar (PCA), a non-selective medium. If the coliforms were
stressed, their recovery on VRBA wou.ld be less than that on PCA. Using
this criteria of stress, the coliform control (Figure 21) was unstressed
based on the level of growth attained at 8 days since recovery on VRBA and
PCA was similar. When grown in combination in a 1:1 ratio (Figure 22) in
non-coliform extract, after an initial lag of four days coliform numbers
increase and recovery again is identical on the two media. However, when
the ratio of coliform to non-coliform is changed to 1:10 after a similar
four day lag when numbers increase on PCA a definite decrease on VRBA is
observed (Figure 23). This suggests that the coliforms are stressed.
Addition of salts + phosphate to the non-coliform extract again
appears to confer protection on the coliforms (Figures 24-26). Control
growth reached higher levels in a shorter period of time on this
supplemented medium (Figure 24). As expected recovery on selective and
non-selective medium was identical. In a 1:1 ratio with Arthrobacter,
coliform numbers approximate control numbers with no indication or stress
(Figure 25). In a 1:10 coliform: non-coliform ratio (Figure 26) the
protective effect of the salts + phosphate is still evident. Growth
levels in the mix again approximate the control and the difference in
recovery on the selective and non-selective media is much less pronounced
than that observed in the unsupplemented medium (Figure 23).
The above series of experiments demonstrates the protective effect of
the salts + phosphate mix on the coliforms in their interaction with
non-coliforms. In the distribution system in the presence of unlined
mains, iron salts may be introduced from pipe corrosion into distribution
water. Tubercles with large numbers of associated bacteria can be
isolated from these unlined mains. In the next experiment with E. coli
and Arthrobacter we attempted to see if tubercle solids or the iron salt
in the salts mix used above (FeS04) would enhance the growth of E^
coli. We also tested the interaction of E. coli and Arthrobacter in the
presence of bacterial extract supplementecTwitn tubercle solids.
Neither tubercle solids alone nor FeS04 alone supported the growth
of either isolate (Figures 27 and 28). Finding that the non-coliform
extract also did not support the growth of Arthrobacter was surprising and
most likely a technical error since a later rerun figure 29) of this
interaction indicated that the same concentration of extract did support
the growth of Arthrobacter. Controls here reached levels of approximately
10? organisms/ml. In a 1:10 ratio with coliforms in this extract the
coliforms were observed to be inhibited. By day eight there was a 103
difference in the levels obtained by the coliforms in the mix when
compared to control levels. When the bacterial extract is supplemented
with tubercle solids the Arthrobacter control grows well but active growth
of the coliform contra! i§ h"6l ObSfiVVed. When the two are combined in
this medium in a 1:10 coliform: non-coliform ratio (Figure 30),
Arthrobacter growth resembles its control growth. Coliform numbers in the
mix, as was seen for the control in this medium are barely maintained.
The tubercle solids therefore did not appear to enhance the growth or
confer protection on the coliforms.
38
-------
_J
2
\
CO
UJ
o
coli on PCA
coli on VRBA
AArthrobacter on PCA
IxlO2
4 6
DAYS
8
Figure 21. Growth of E. coli and Arthrobacter sp. in bacterial extract.
39
-------
LJ
O
• E. coli on PCA
£. coli on VRBA
AArthrobacter on PCA
IxlO3
5xl02
IxlO2
0 I
4 6
DAYS
Figure 22. E. coli and Apthrobacter sp. (1:1) in bacterial extract.
40
-------
IxlO8
5xl07 -I
IxK)5
5xl04
LU
O
E. coli on PCA
E. coli on VRBA
Arthrobacter on PCA
IxlO4
5xl03
IxlO3
5xl02
IxlO2
5x10'
IxlO1
Figure 23. £. coli and Arthrobacter so. (1:10).In bacterial extract.
41
-------
CO
_J
LJ
O
O E. coli on PCA
DE. coli on VRBA
A Arthrobacter on PCA
IxlO3 -
5xl02
IxlO2
DAYS
8
Figure. 24. Growth of £. coli and Arthrobacter sp. in bacterial extract
supplemented with salts and phosphate buffer.
42
-------
UIO8 •
5xl07 •
IxlO7 -
5xl06
£. coli on PCA
i. coli on VRBA
Arthrobacter on PCA
3
O
IxlO6
5xl05
IxlO5
5xl04
IxlO4
5xl03
IxlO3
5xl02
IxlO2
DAYS
Figure 25. E.. coli and Arthrobacter sp. (1:1) in bacterial extract
supplemented with salts and phosphate buffer.
43
-------
IxlO8 .
5xl07 -
CO
Ld
O
IxlO6 -
5xl05 -
coli on PCA
coli on VRBA
Arthrobacter on PCA
IxlO1
0
I
DAYS
8
Figure 26. E.. coli and Arthrobacter sp. (1:10) in bacterial extract
supplemented with salts and phosphate buffer.
-------
CO
_J
_1
UJ
o
O extract
8 tubercle
FeS04
• extract + tubercle
A extract + FeSO
1x10°
0 I
4 6
DAYS
Figure 27. Growth of Arthrobacter sp. in various media.
45
-------
IxlO6
5xl05 -
IxlO6 -
5xl04 •
O extract
• tubercle
DFeSO
extract + tubercle
extract + FeSO
IxlO
0 --"I
Figure 28. Growth of £. coli in various media.
46
8
-------
IxlO8
5xl07 -I
CO
UJ
o
IxlO7 -
IxlO8 -
IxlO5
5xl04
O E_. coli alone
• E. coli in mix
D Arthrobacter alone
Arthrobacter in mix
IxlO2 -
IxlO1
4 6
DAYS
Figure 29. E_. coli and Arthrobacter sp. (1:10) in bacterial extract.
47
-------
IxlO9 •
5xl08 -
IxlO8 -
5xl07 -
UJ
o
U Arthrobacter alone
• Arthrobacter in mix
IxlO
DAYS
8
Figure 30. E_. coli and Arthrobacter sp. (1:10) in bacterial extract
supplemented with crushed tubercle solids.
48
-------
The Effect of the Physiological Status of Either Member of an Interaction—
In previous experiments where the interaction of E. coli and
Flavobacterium was studied, we observed inhibition of the coliform by the
non-coliform when the two were inoculated together into non-col if orm
extract (Figures 1-6). In these experiments the numbers of the
Flavobacterium controls increased with time reaching levels of 105 -
1U". However, when the non-coliform is physiologically changed as is
indicated by the control not growing and increasing in numbers in the
extract, the ability of it to inhibit the coliform is altered. In figures
31 and 32 it can be seen that when the Flavobacterium control is only able
to maintain itself in the medium and the conform control is increasing,
the interaction of the two organisms is affected. When the two organisms
are grown together in the same medium the coliform is not inhibited. The
Flavobacterium is unable to inhibit the coliform and is unaffected by the
presence or tne coliform.
The physiological state of the coliform member of the interaction is
also important in determining the outcome of an interaction. When E. coli
and Arthrobacter were grown on non-coliform extract at 6.6° C, the
coliform control (Figures 33-34) maintained itself over the eight day
course of the experiment. An additional assay at eighteen days indicated
only a slight decrease in viable numbers. The Arthrobacter control only
increased approximately twenty fold over the course of the eight days
(Figure 33). When the two were grown together in a 1:1 ratio in
non-coliform extract at the low temperature (Figure 34) Arthrobacter
numbers increased over the control reaching a final value of lU^
organisms/ml whereas the coliform numbers remained unchanged. Under these
conditions the coliforms were not inhibited. Recovery of the coliforms
was done using PCA and Endo pour and spread plates. The use of the
selective Endo medium and the pour plate which subjects the organisms to a
high temperature stress might be expected to recover less organisms if the
coliforms in the mix were stressed by the presence of the non-col iforms.
This does not appear to be the case. Coliform numbers were similar to
those observed for the control. Recovery on PCA and Endo pour and spread
plates was similar. It appears then that the coliforms may be less
vulnerable to inhibition or stress by non-coliforms when either member of
the interaction is not actively growing.
TURBIDITY EXPERIMENTS
The purpose of this series of experiments was to determine if
non-bacterial turbidity such as that due to natural minerals, hydrated
oxides and organic debris was responsible for inhibition of coliform
recovery. The study was carried out at the Wilmington Water Department,
Wilmington, Delaware using various types of water and turbidity. The
turbidity was either added to the water, in turbidity augmentation
studies, or removed from the water, in turbidity reduction studies. Since
this series of experiments was concerned with the effect of non-bacterial
turbidity on coliform recovery an attempt was made to alter the turbidity
of the water samples without altering the total bacterial count. This was
done in the turbidity augmentation studies by adding turbidity which was
49
-------
LJ
O
IxlO7.
5xlOe .
IxlO6 -
5xl05 -
IxlO5
5xl04
IxlO2
5x10'
IxlO1
O £_. coli alone
• £. coli in mix
D F1avobacteriurn alone
• Flavobacterium in mix
I
4 6
DAYS
8
Figure 31. E. coli and F1avobacteriurn sp. (1:1) in bacterial extract.
50
-------
IxlO7 •
5xl06 -
IxlO6 -
5xl05 •
IxlO5
5xl04
3
UJ
o
OL coll alone
• £. coli in mix
D Flavobacterium alone
B Flavobacterium in mix
IxlO1
DAYS
8
Figure 32. £. coli and Flavobacterium sp. (1:10) in bacterial extract.
51
-------
IxlO8 •
5xl07 -
IxlO7 -
5xl06 '
O E_. coll on PCA spread plate
• E_. coli on PCA pour plate
D E. coli on Endo spread plate
•E' co1i on Endo pour plate
AArthrobacter on PCA spread plate
IxlO6
5xl05
Ld
O
IxlO2
0 I
4 6
DAYS
Figure 33. Growth of E_. coli and Arthrobacter sp. in bacterial extract at
6.6 C.
52
-------
IxlO8 .
5xl07
IxlO7 -
5xl06 '
CO
UJ
o
O £. coli on PCA spread plate
• E_. coli on PCA pour plate
D E_. coli on Endo spread plate
1 L-. c°1i on Endo pour plate
A Arthrobacter on PCA spread plate
IxlO2 "I
4 6
DAYS
8
Figure 34. E. coli and Arthrobacter sp. (1:1) in bacterial extract
at 6.6°C.
53
-------
assayed and known to be bacterial free and in the turbidity reduction
studies by using very low speed centrifugation which would not pellet
bacteria. As a result of these manipulations the resulting two water
samples would vary only in the level of non-bacterial turbidity. If this
non-bacterial turbidity was the major factor in inhibition of coliform
recovery then the water containing the higher level of non-bacterial
turbidity would be expected to yield fewer coliforms than the low
turbidity water.
Since this was not observed, additional studies were done with some
of the water samples to further distinguish the role of bacterial and
non-bacterial turbidity in inhibition of coliform recovery. As in the
first series the two samples differed in levels of non-bacterial
turbidity, but had similar numbers of bacteria. From these samples
increasing volumes of water were removed and membrane -filtered. In the
filtration series for both of the two waters inhibition of coliform
recovery was observed to be a function of the volume filtered with the
greatest inhibition observed for the highest volume filtered. Since the
two original water samples varied only in their levels of non-bacterial
turbidity, filtering the same volume from each original water sample
resulted in the deposition on the filter pad of the same number of
bacteria but different levels of non-bacterial turbidity. As was noted,
the high volume sample from each of the two waters showed the greatest
inhibition of coliform recovery. If the coliform recovery of these two
samples was comparable, it would appear that the inhibition is due
primarily to high levels of bacteria. If however coliform recovery was
inhibited to a greater extent in the high volume sample from the original
high turbidity water, then inhibition would appear to be due primarily to
the presence of high levels of non-bacterial turbidity.
Materials and Methods
Water Samples—
The waters used in the turbidity experiments were samples of raw
water, distribution water and recirculating water from a main excised from
an area of the distribution system which had regrowth problems. The
excised main was two meters long and had a 150 mm diameter. Feed water
for the main came from areas of the distribution system which had lost
their chlorine residual.
In some early experiments raw water was used because it guaranteed
the presence of coliforms particularly during the winter months. Water
for later experiments was usually taken from remote areas of the
Wilmington distribution system carrying little or no chlorine. These
locations also had a history of coliform occurrence ranging from
occasional to frequent. Because the actual coliform level was difficult
to predict, an attempt was made to maximize coliforms by blending samples
from two or more "trouble spots," and by occasionally adding small amounts
of raw water or water from the isolated main. Despite these efforts many
membrane filters had low colony counts, a situation familiar to those who
routinely search for coliforms in distribution systems.
54
-------
Turbidity Sources--
In mains known to be carrying chlorine, but where loose deposits were
also suspected, a surging of the main frequently yielded a moderate
chlorine, high turbidity mixture at hydrants. When such suspensions were
collected in sterile containers without thiosulphate and stored at 2 -
4°C for several weeks many of them were then found to be chlorine-free
and bacteria-free.
Confluence situations also yielded suspensions suitable for turbidity
studies. When water traveling by one route degrades to a turbid,
chlorineless condition, and then converges with water which has traveled
by a superior route, retaining its chlorine, this mixture can often be
refrigerated for a few weeks to yield an apparently sterile chlorine-free
turbidity.
Turbidity Measurement and Centrifugation—
Centrifugal separation of bacteria from other turbidity was
accomplished in an International UV Centrifuge using 50 ml sterile capped
tubes. G forces used were from 160 to 450 and were maintained for 6 or 8
minutes. The decanted supernatants were pooled before use.
Turbidity was measured using a DRT-100 nephelometer type instrument.
Coliform Recovery--
Col if orm recovery was assayed by both MPN and MF in case one of these
techniques was more sensitive to inhibition by turbidity. It is difficult
to compare the absolute numbers obtained using these two techniques
because the MPN number actually represents a range. It was therefore
necessary to determine if the MF coliform number fell within the MPN range.
MPN coliform measurements used Lauryl Sulphate Broth (sometimes
called Lauryl Tryptose Broth), 20 ml of 1 1/2 strength broth for 10 ml
portions, and single strength for 1 ml and smaller portions. 3 or 4
groups of 3 decimal ranks of 5 tubes each were employed and presumptive
positives were transferred to Brilliant Green Bile Broth for confirmation.
Membrane filter measurements of coliforms were made with both LES
Endo agar and MF Endo broth usually without enrichment. The broth was
used more often on the assumption that more routine analyses in the water
industry are being conducted on this medium. Incubation was for 22 to 24
hours at 35°C. Unless indicated in the text, MF coliforms were not
confirmed. Therefore the numbers reported as MF coliforms represent the
highest possible number of these organisms recovered. Even with this bias
it was observed that in many experiments increasing the volume of the
water sample filtered resulted in decreased recovery of coliforms on a per
ml or 100 ml basis. Therefore of even greater interest than confirmation
of colonies which looked like coliforms is the analysis of colonies which
do not show up as coliforms. This is in principle technically possible,
but in practice not feasible. However, to determine if perhaps there were
coliforms present on the MF filters which were not growing into colonies,
circular 50 mm^ discs were cut from inoculated membrane filters with a
55
-------
sterile cork borer and transferred to Lauryl Sulphate Broth and confirmed
in Brilliant Green Bile broth.
Total Bacterial Recovery--
Measurement of the total bacterial population was done using M-5 agar
(3 grams Plate Count Agar, 12 grams plain agar, 200 mg ferrous gluconate
in 1000 ml distilled water) formulated by H. T. Victoreen. The
suitability of this agar for this purpose and its superiority to plate
count agar is evident from the data in Table 4 where pour plate results
with Plate Count Agar (Standard Methods Agar) are compared with M-5 agar
using a variety of water samples over a 7 month period. Both inoculated
agars were incubated for 7 days at 28°C.
TABLE 4. A COMPARISON OF BACTERIAL RECOVERY ON
PLATE COUNT AGAR AND M-5 AGAR
Source
Date
PC
M-5
M-5/PC
Grove. Pr.
K. Butler
ch Lore
Wynd.
Br. Ns.
Pr. Ess.
708 Pr.
Wynd
N.P.
Br. Ns.
Ch. Lore
Ch. Lore
Woodcr .
Gov. Pr.
Br. Ns
Br. Ns
1509
K. Butler
Eastm.
1212
Han. Wash
Lower Gar!
Maria
3-21-77
3-21-77
3-21-77
3-23-77
3-23-77
3-25-77
3-25-77
4- 3-77
4-13-77
4-13-77
4-29-77
5- 6-77
5- 6-77
5- 9-77
6-22-77
7- 6-77
9- 7-77
9-26-77
9-26-77
9-28-77
10-26-77
10-26-77
11- 7-77
4
13
340
4
320
1300
0
2
140
30
7000
1600
6
28
450
740
46
180
76
40
90
54
100
48
180
6000
22
2700
3200
150
100
2000
1500
13000
5000
36
240
5600
14000
6500
2200
2500
2700
6200
4800
8000
12
14
18
5.5
8.4
2.5
50
14
50
1.9
3.1
6
8.6
12
19
140
12
33
68
69
89
0.80
Studies Using Raw Water
A group of preliminary experiments were done on membrane filters to
determine whether in a turbid sample coliform yields would be obtained
which were proportionate to sample size. In these experiments raw water
of good quality was used. The water samples had phosphate levels of less
than 0.4 and ranged in turbidity from 2.0 to 10.0 FTU. The results appear
in Table 5.
56
-------
TABLE 5. COLIFORM RECOVERY FROM GOOD QUALITY RAW WATER
Date
9/29/77
10/10/77
10/16/77
Sample
Source Size (ml)
Raw Porter* 0.5
2
Raw Brandywine* 0.5
2
Raw Brandywine 0.1
0.2
0.5
Raw Brandywine* 1
2
4
4
8
50
100
150
Colonies
Counted
19
25
28
51
36
55
90
2
7
6
7
16
71
80
75
Calculated
Coliforms/lOOml
3.8X10"*
1.3X103
5.6X103
2.6X103
3.6X104
2.8X104
1.8X104
200
350
150
175
200
140
80
50
*LES Endo agar with enrichment. On remaining dates the same medium was
used without enrichment.
It is apparent that despite variations in source, turbidity and coliform
density, smaller volumes filtered gave better calculated yields on the
same sample. When a smaller sample volume is filtered, less turbidity and
bacteria are being spread over the effective filtering area. Increasing
the sample size filtered may result in inhibition of coliform recovery by
turbidity, non-col iforms, coliforms or an undefined crowding factor.
Microbial inhibition may be a result of toxic by-products of other
organisms or competition for space and nutrients.
Another experiment was conducted with an intermediate quality raw
57
-------
water centrifuged at 200g for 6 minutes to reduce the turbidity from 18 to
3.0. Coliform recovery was determined using Endo broth without
enrichment. Results appear in Table 6.
TABLE 6. COLIFORM RECOVERY FROM INTERMEDIATE QUALITY RAW WATER
Pre-Centrifugation Post-Centrifugation
(Turbidity = 18) (Turbidity = 3)
Quantity
Filtered Coliforms/100 ml
2-0.05 ml
portions
4-0.20 ml
portions
3-0.50 ml
portions
6
2
1
.1 x 104
.88 x 104
.73 x 104
Coliforms/100 ml
6.4
2.85
1.83
x 104
x 104
x 104
Total bacterial numbers were determined using M-5 agar incubated for 7
days at 28° C. Reduction of turbidity from 18 to 3 FTU did not reduce
total bacterial numbers as evidenced by the same bacterial counts (1.25 x
1Q5 organisms/ml) obtained on M-5 agar before and after centrifugation.
In addition, the pattern of decreasing yield with increasing sample size
was again observed for the pre- and post-centrifuged samples. Since
comparable numbers of coliforms were recovered from a given volume of each
of the two water samples, it would appear that turbidity per se is not
responsible for the decreased coliform yield obtained whenTarger volumes
of water are filtered. The observed pattern may therefore be due to the
continued crowding of colonies on the filter surface with a resultant
competition for space and nutrients or from the by-products of the
competing organisms.
Studies Using Distribution Water
Distribution water was used to study the effect of turbidity
augmentation on coliform recovery. The water used was a composite made of
water from five areas, each with a history of coliform regrowth. 2% raw
water was then added to this mix to further improve the chances of
detecting coliforms during the experiment. All of the five finished water
components had lost their chlorine in the distribution system.
The mixture was divided in half and one half was augmented with
additional discolored water deposits to raise the turbidity from 1.2 to
6.6. These deposits were obtained by collecting from an area containing
chlorine in a sterile thiosulphate free container and refrigerating for
several weeks until the deposit was chlorine free.
Coliform determinations were made on the water before and after
turbidity augmentation by the MPN and membrane filter techniques. Results
are reported in Table 7.
58
-------
TABLE 7. COLIFORM RECOVERY FROM DISTRIBUTION WATER
Pre-Augmentation Kost-Augmentation
(Turbidity =1.2) (Turbidity - 6.6)
uontirmea w/iuu mi w/iuu mi
MPN*
MF Counts 15/100 ml 15/100 ml
6 - 10 ml
portions
6 - 50 ml 21/100 ml 19/100 ml
portions
*reported figures represent an average of 4 sets of 15 tubes MPNS.
Parallel assessments of the overall bacterial population were made by 7
day plate counts at 28° C on M-5 agar. The M-5 counts were 2.5 x 103
organisms/ml for the pre-augmentation water and 2.6 x 103 organisms/ml
for the post-augmentation water indicating that the added turbidity
concentrate was bacteria-free and that the only change it brought about in
the coliforms1 environment was to raise the amount of suspended matter.
Under such conditions, there is no impairment of coliform recovery by
either method, attributable to the turbidity augmentation. In addition,
filtering a five fold greater portion through the millipore filter (50 ml
as opposed to 10 ml portions) did not result in a proportionate loss of
coliform recovery. However, when 4 50 mm2 discs were removed from 2
sets of pre- and post-augmentation filters in each case one of the 4
filters was positive in L.S. broth. This may indicate that in both cases
there were inhibited or stressed coliforms on the filter which would not
have been considered as such if special efforts for their recovery had not
been undertaken. In this composite water coliforms are outnumbered by the
non-col iforms by a factor of 5 x 103.
In another experiment with distribution water the water temperatures
were so low that there was no assured source of coliform bearing water in
the distribution system. It was therefore necessary to take a stock
coliform culture previously isolated from a coliform regrowth area, dilute
it heavily, and adapt it to a medium consisting of membrane filtered water
from this same location fortified with pulverized, sterilized main
tubercles. This produced a suspension which 48 hours later had a coliform
MPN of 490. When this suspension was riled and allowed to settle for 30
minutes, the supernatant had an MPN of 130. This supernatant was employed
in the experiment discussed below.
A freshly drawn 50 - 50 mix from two areas showing coliforms during
warmer weather was divided into two sterile flasks, brought up to room
temperature and inoculated with a 1Q% addition of the supernatant
described above. This gave a turbidity of 5.6. One of these flasks
received a further addition, an amount of bacteria-free, chlorine-free,
main deposits sufficient to further raise the turbidity to 11.4. After 30
59
-------
minutes contact coliform and M-5 plate counts were determined on both
flasks. Coliform recovery is reported in Table 8.
TABLE 8. COLIFORM RECOVERY FROM DISTRIBUTION WATER
're-AugmentationKost-Augmentation
(Turbidity = 5.6) (Turbidity - 11.4)
Confirmed MPN* 7.7/100 ml
MF Counts 1.6/100 ml
^Reported figures represent an average or 4 sets or i
4.0/100 ml
2.8/100 ml
u tune MKiNi.
M-5 agar counts were 1.6 x 10^ organisms/ml for the pre-augmentation
water and 1.7 x 10^ organisms/ml for the post-augmentation water, again
indicating that the added turbidity did not contain viable bacteria. When
the MF and MPN results are compared the MF results do fall into the 9595
confidence limits of the MPN. The original water, however, had an MPN of
130 (range = 30 - 3100 coliforms/100 ml) before being diluted 1:10. The "
data is suggestive here, then, that recovery by either method may be at
the lower end of the expected range. This may mean that factors such as
the high numbers of non-col iforms which outnumber the coliforms by a
factor of 106 or the turbidity present in the 5.6 FTU water is high
enough to inhibit coliforms or coliform recovery and increasing the
turbidity to 11.4 does not further contribute to inhibition of coliform
recovery.
Studies Using Recirculating Distribution Water from an Excised Main
The main used, as was previously noted, was excised from an area of
the distribution system which had regrowth problems. A study was done
using distribution water which had recirculated through this main by
blending unfiltered water from the main with some which had been fiber
glass filtered. This resulted in a discolored water sample with a
turbidity of 8.0. As a precaution against an unexpected drop in the
coliform population this mixture was then augmented with low levels of raw
water. Coliform determinations were made on this water by both the MPN
and MF methods and the total bacterial population was evaluated using M-5
agar with 7 days incubation at 28°C. The water was then centrifuged at
460 x g for 7 minutes and the entire evaluation repeated. The final
turbidity was 1.3. The coliform data is reported in Table 9.
TABLE 9. COLIFORM RECOVERY FROM RECIRCULATING MAIN WATER
Hre-leritr1tugati6n rost-uentrirugation
(Turbidity -8.0) (Turbidity•- 1.3)
MHN*
1st set
2nd set
(Continued)
230/100 ml
330/100 ml
330/100 ml
330/100 ml
60
-------
TABLE 9. (Continued)
Kre-uentriTugation
(Turbidity = 8.0)
'ost-uentriTugain
(Turbidity ~ 1.3)
MI- tounts
1 ml portion
2 ml portion
4 ml portion
6 ml portion
8 ml portion
1400/100 ml
950/100 ml
180/100 ml
800/100 ml
500/100 ml
180/100 ml
75/100 ml
*Reported figures represent an average of 4 sets of 15 tube MPNS.
M-5 counts indicated a decline in bacterial numbers from an average of 6.9
x 105 organisms/ml in the original 8 FTU water to 5.7 x 105
organisms/ml in the 1.3 FTU water. This reduction in total bacterial
numbers is also reflected in a reduction of coliform recovery by the MF
technique. However, the MF results are observed to vary dramatically as a
function of the volume filtered. The phenomenon is observed with both the
pre- and post-centrifuged waters. The pre-centrifuged water has an
approximately seven fold higher level of turbidity. If turbidity were
primarily responsible for inhibition of coliform recovery, the recovery of
organisms from the low turbidity water would have been greater than that
which was observed. Centrifugation removed 18% of the total organisms,
and therefore presumably also of coliforms. The coliform numbers
recovered in the post-centrifuged water reflect at least this 18%
decrease. Therefore it appears that removal of turbidity did not enhance
coliform recovery.
A later coliform recovery experiment was done using 80% of the water
from the recirculating main and 20% from a high rust distribution area.
This mixture had a turbidity of 15.5. After 7 minutes centrifugation at
1600 RPM the turbidity was reduced to 2.6. Coliform recovery before and
after centrifugation using the MF and MPN procedures is presented in Table
10.
TABLE 10. COLIFORM RECOVERY FROM A COMBINATION OF DISTRIBUTION WATER AND
RECIRCULATING MAIN WATER
rre-uentrifugation
(Turbidity - 15.5)
Tit
Post-UentriTugation
(Turbidity ^2.6)
tonrirmea
MF Counts
5 - 1.0 ml
volumes
/zu /iuu mi
180/lOOml
/iuu mi
220/lOOml
*Reported figures represent an average of 3 sets of 15 tube MPNS.
61
-------
The rusty deposits which constituted most of the turbidity present were
easy to remove and the centrifugation procedure removed approximately 83%
of the turbidity without effectively altering the bacterial levels
present. Under these conditions coliform recovery did not appear to be
effectively changed as a function of turbidity removal. M-5 counts (7 day
incubation at 28°C) were an average of 1.4 x 106 organisms/ml for the
pre-centrifuged water and 1.2 x 106 organisms/ml for the
post-centrifugaed water. Centrifugation therefore removed 83% of the
turbidity and 14% of the total bacterial numbers. If turbidity £er se
were inhibiting coliform recovery both the MF and MPN techniques migWT be
expected to indicate improved recovery of coliforms. This does not appear
to be the case.
Another coliform recovery experiment was done using 23% of the water
from the recirculating main, 73% of the water from the distribution system
and 2% rusty water from the distribution system. The mixture had a
turbidity of 5.1 and was centrifuged for 7 minutes at 1650 RPM to give a
turbidity of 1.1. Coliform recovery before and after centrifugation using
the MF and MPN procedures is presented in Table 11.
TABLE 11. COLIFORM RECOVERY FROM A COMBINATION OF DISTRIBUTION WATER AND
RECIRCULATING MAIN WATER
(Turbidity = 5.5) (Turbidity ~-2.6)
uonnrmed MKNX wu/iuu mi zou/iuu mi
MF Counts
4 - 5 ml portions 25/100 ml 30/100 ml
4 - 25 ml portions 37/100 ml 26/100 ml
*Reported figures represent an average of 3 sets of 15 tube MPNs.
M-5 counts (7 day incubation at 28^ C) were an average of 5.0 x
organisms/ml for the pre-centrifuged water and 3.3 x 105 for the post-
centrifuged water. Centrifugation therefore removed 78% of the turbidity
and 33% of total organisms.
In this experiment the MF counts are lower than the range obtained
using the MPN. In addition the MPN numbers obtained after centrifugation
of the water reflect a loss of organisms, whereas the numbers obtained
using the MF technique do not. Using the MF technique, filtering 5 or 25
ml portions yielded a similar number of colonies on the filter. One ml
portions had also been run, but the numbers were so low aS to be
statistically unreliable. It may be for the levels of turbidity and
organisms present in the water that larger volumes would have had to be
used to see inhibition of coliform recovery as a function of the volume
filtered.
Two additional types of manipulations were done in this experiment.
62
-------
Discs of 50 mm2 size were aseptically removed from the apparently barren
intercolonial spaces on certain membrane filters after 24 hours incubation
and transferred to single strength Lauryl Sulphate Broth. In the case of
the pre-centrifuge filters 2 discs were cut from each of 2 - 5.0 ml
filters and 4 discs from a 25 ml filter. In the case of the
post-centrifuge filters 2 discs from each of 2 - 5.0 ml filters and 2
discs were cut from each of 2 - 25 ml filters. All 16 of these discs
produced gas in Lauryl Sulphate Broth in 18 hours and confirmed in
Brilliant Green Bile. To say that turbidity was not restricting recovery
is not to say that all the coliforms were detected. The fact that all of
the 50 mm2 discs from apparently barren areas of both pre- and
post-centrifuge filters were positive in Lauryl Sulphate Broth and
confirmed in Brilliant Green Bile suggests heavy microbial interference.
Apparently, coliforms which had not grown into recognizable colonies were
trapped under the surface mat on the filters and removal of 78% of the
turbidity left the mat just as restrictive as before. There was also
imperfect coliform detection among the MPNs. When one 0.1 ml set from the
pre-centrifugation group and one 0.1 ml set from the post centri'fugation
were all transferred to Brilliant Green Bile, one more tube showed
positive in each set. The sets chosen were among those which had
displayed a mixture of positive and negative tubes. If the effort to find
coliforms among the negative, but turbid Lauryl Sulphate tubes had been
extended to all sets, it is possible that more would have been found.
Such studies have previously been done at the Wilmington Water Department
(13). In these studies transfers were made from negative but turbid
Lauryl Sulphate Broth tubes into Brilliant Green Bile broth tubes. The
Brilliant Green Bile broth tubes gave positive results, and further exam-
ination of the organisms in these tubes indicated that coliforms were
present. This study along with the ones presented in this report indicate
that numerous factors influence coliform recovery. This should be taken
into consideration in the interpretation of microbiological data obtained
from assaying distribution system water.
63
-------
REFERENCES
1. Geldreich, E. E., Nash, H. D., Reasoner, D. J., and R. H. Taylor.
The Necessity of Controlling Bacterial Propulations in Potable
Waters: Community Water Supply. J. Amer. Water Works Assn.,
64:596-602, 1972.
2. Standard Methods for the Examination of Water and Wastewater, 14th
ed. American Public Health Association, New York.-, 1975. 1193 pp.
3. Buchanan, R. E. and N. E. Gibbons, Eds. Bergeys manual of
Determinative Bacteriology, 8th ed. The Williams and Wilkins Co.,
Baltimore, Maryland, 1974. 1246 pp.
4. Lennette, E. H., Spaulding, E. H., and J. P. Truant, eds. Manual of
Clinical Microbiology. 2 nd ed. American Society for Microbiology,
Washington, D. C., 1974. 414 pp.
5. Bailey, W. R. and E. G. Scott. Diagnostic Microbiology. C. V.
Mosby, St. Louis, Missouri, 1974. 414 pp.
6. Geldreich, E. E., Nash, H. D. and D. Spino. Characterizing Bacterial
Populations in Treated Water Supplies: A Progress Replrt. In:
Proceedings of the American Water Works Association Water Quality
Technology Conference, Kansas City, Missouri, 1977. Section 2B-5,
pp. 1-13.
7. von Graevenitz, A. The Role of Opportunistic Bacteria in Human
Disease. Ann. Rev. Microbiol., 31:447-471, 1977.
8. Postgate, J. R. and J. R. Hunter. The Survival of Starved Bacteria.
J. Gen. Microbiol., 29:233 -263, 1962.
9. Hassid, W. Z. and S. Abraham. Chemical Procedures for Analysis of
Polysaccharides. In: Methods in Enzymology, S. P. Colowick and N.
D. Kaplan, eds. Academic Press, New York, New York, 1957. pp. 34-50.
10. Bradford, M. M. A Rapid and Sensitive Method for the Quantitation of
Microgram Quantities of Protein Utilizing the Principle of
Protein-dye Binding. Anal. Biochem., 72:248-254, 1976.
11. Hill, B. T. and S. Whatley. A Simple, Rapid Microassay for DNA.
FEBS Letters, 56:20-23, 1975.
(continued)
64
-------
REFERENCES
12. Sokal, R. R. and F. J. Rohlf. Biometry; the Principles and Practice
of Statistics in Biological Research. W. H. Freeman, San Francisco,
California, 1969, 776 pp.
13, Victoreen, H.T. Water Quality Deterioration in Pipelines. In:
Proceedings of the American Water Works Association Water Quality
Technology Conference, Kansas City, Missouri, 1977. Section 3B-5,
pp. 1-9.
65
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-80-097
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
HINDRANCE OF COLIFORM RECOVERY BY TURBIDITY AND
NON-COLIFORMS
5. REPORT DATE
August 1980 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Diane S. Herson and Hugo T. Victoreen
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
School of Life and Health Sciences, Univ. of Delaware
Newark, Delaware 19711
Wilmington Water Dept., 16th & Market Streets,
Wilmington, Delaware 19801
10. PROGRAM ELEMENT NO.
C61C1C
11. CONTRACT/GRANT NO.
R-805102
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory - Cin., OH
Office of Research and Development
U.S. Environment Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final 3/13/77-8/15/79
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: Edwin E. Geldreich (513) 684-7232
16'ABSTRT?ie objectives of this project were to evaluate the recoverability of coliforms
from waters which have: a) high populations of non-coliform organisms, and b) high
levels of turbidity due to natural mineral turbidity, hydrated oxides and organic
debris. After initial isolation and identification of coliforms and non-coliforms
from raw and distribution water interactions between these two groups of organisms
were studied. The outcome of the interaction was found to be dependent upon numerous
factors. These included: the specific non-coliform, the numbers of each type of
organism, the nutrient environment in which the interaction occurred, the
physiological status of the interacting organisms and the type of media used to
recover the coliforms.
Turbidity augmentation and reduction experiments were done to distinguish
non-bacterial turbidity inhibition of coliforms from the inhibition caused by other
bacteria. The more serious inhibition to coliform detection seemed to be caused by
the large populations of non-coliforms which exceeded the resident coliforms in water
mains by factors of 10^ to 10*. Turbidity per se was not an impediment to
coliform growth, but it did make it more difTTclfTT to recognize coliforms on membrane
filters. These results will be of interest to individuals concerned with water
quality and interpretation of microbiological data.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Coliform bacteria
Potable water
Turbidity
Microorganism control (water)
Eschen'chia coli
Flavobacterium
Coliforms
Coliform inhibition
13B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
78
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (Rev. 4-77)
66
U.S. GOVERNMENT PRINTING OFFICE: 1980--657-165/0106
------- |