A radioactive carbon technic is described which shows definite promise as a rapid test for pollution of water in time of emergency. Problems that still exist are discussed.
1. RAPID CARBON14 TEST FOR COLIFORM
BACTERIA IN WATER Robert M. Scott, M.S.; Dorothea Seiz, B.S.; and Howard J. Shaughnessy, Ph.D., F.A.P.H.A.
DRESENT methods of testing water for Fbacteriological safety depend upon the detection of the coliform group of bacteria. The more rapid of the present technics utilizes the membrane filter test. This test is dependent upon the growth and multiplication of coliform cells until they reach sufficient numbers to be visible to the unaided eye as a colony. The other test, referred to as the fermentation tube method, is dependent upon sufficient growth and metabolic activity of the coliform bacteria so that visible gas will be formed from lactose in a fermentation tube. This may take from two to four days, while the membrane filter technic requires 24 hours.' It would be most desirable if a much quicker test could be developed for use in a state of emergency. The development of such a test was the objective of the following study. Both of the described tests depend on the conditions prevailing during the logarithmic or growth phase of a bacterial culture. When bacteria are introduced into a new medium, especially a more complex one, they start metabolizing at once.2 During this time, cell division is virtually at a standstill while metabolism is accelerated. Thus, if coliform bacteria are introduced into a medium containing lactose, multiplication may not occur for several hours but MAY, 1964
carbon dioxide should be given off almost immediately. Levin, et al.,3 at Georgetown University conceived the idea of using lactose containing radioactive carbon instead of standard lactose in broth. Their results showed that the small amounts of carbon dioxide given off in the lag phase could be measured with a radioactive isotope detector since a portion of the radioactive carbon was converted to C14 dioxide.4 When radioactive lactose became unavailable, Levin and his group started using sodium formate with radioactive carbon.5 Formate is associated with carbohydrate metabolism in which bacterial cells utilize hydrogen to form metabolic water and liberate carbon dioxide. If the carbon in the formate is radioactive, the evolved carbon dioxide should also be radioactive, and measurable with any sensitive beta activity detector.
Methods The basic medium used was MF (millipore filter) Endo broth. Formic acid C14 sodium salt with a specific activity of 4 mc per millimole was made up with sterile distilled water to a concentration of 0.008 per cent. Equal volumes of this radioactive formate solution and double strength MF lactose broth were mixed just prior to use. This made a final 827
Table 1-Comparison of Coliform Counts and Evolved Radioactivity (All Samples Studied) Bacterial Counts as Determined by the MF Technic 25- 501002004008001,6003,2001-25 50 100 200 400 800 1,600 3,200 6,400 Cells Cells Cells Cells Cells Cells Cells Cells Cells 5 138 19 26 28 56 277 54 13 106 17 13 106 69 32 20 28 5 10 9 0 12
46 60 31
Total Tests
22
Geom. Av.
19
0
.=
a
o
o
Z
721 73 106 35
60 325 628 124
3
4
4
44
120
200
576
1,038
338
964
5,027
6,560
6
3
2
1
3
1
1,200
990
1,500
9,900
7,900
17,000
6,688
202
9,875
11,680
17,284
3,718 1,689
1,357 1,035 516
medium of standard MF lactose broth containing 0.004 per cent radioactive sodium formate. The medium was then agitated gently on a Palo Laboratory shaker to disperse any chemical build-up of radioactivity. The water samples were filtered directly through a one inch standard HA membrane with pore size of 0.47 microns. Each test sample membrane was then placed in the bottom of an individual aluminum planchet (1 inch in diameter, about 1/4 inch deep with a flat lip of about 1/8 inch) and 0.5 ml of the formic acid Endo medium was added directly on top of the membrane in each planchet and immediately covered with a glass cover slip. The planchets were gently 828
23,000 Cells
agitated for approximately one minute to provide adequate mixing of sample and medium, and then placed in a 350 C dry heat incubator for three and one-half hours. For each sample planchet containing medium and membrane, a second planchet was prepared containing an absorbent filter pad impregnated with five drops of a saturated solution of barium hydroxide. At the end of the three and one-half hour incubation period, the sample planchets were removed and allowed to set at room temperature for five minutes. During this five minute period, the cover slip was removed from each incubated sample planchet and immediVOL. 54, NO. 5. A.J.P.H.
RAPID CARBON14 TESTS
ately covered with the second planchet prepared with the barium hydroxide pad. This formed a double planchet, one inverted over the other. These were then re-incubated for another thirty minutes. At the end of the second incubation period, the double planchets were removed and gently shaken to release as much as possible of the dissolved carbon dioxide from the medium. The bottom half of the double planchet containing the membrane and medium was discarded. The upper, or inverted, planchet was dried under an infra red heat lamp for a few minutes, then covered with five drops of a 0.5 per cent collodion solution and dried until caramel brown under the heat lamp. The dried pads were counted in a windowless gas flow proportional counter (Nuclear Measurements of Indianapolis) for five minutes. A sterile medium control was set up with
each series of tests, run exactly as all the rest except no organisms were added. For a coliform numbers control, all samples were counted by making the appropriate dilutions using the standard membrane filter technic with MF Endo broth.
Results Table 1 shows all the results obtained after the described technic was established. Most of the results are in the lowest numerical range since the lower limits of the test were of primary interest. It is recognized that there is insufficient data in the higher cell groupings to make the results valid. In Figure 1, the count per minute (CPM) log averages of each cell grouping were plotted against bacterial counts. It is of interest to note that the first few group-
...-
-4
It"C
--
20D
0 c2;tceiie 4Icells
1.a,--te, 1
cm
nts
......
4C0
cl0 C0113
-.Iterc:Cunt
-y th,e
cells
,:
:
Figure 1 MAY, 1964
829
Table 2-Comparison of Coliform Counts and Evolved Radioactivity (Polluted Water Samples) Bacterial Counts as Determined by MF
Sources of Samples
25- 501001-25 50 100 200 Cells Cells Cells Cells
Sewage Sewage
46 *> 54 13 Sewage Sewage 106 Pure Cult. o 28 River A r 17 Drain Ditch o River B1 River B2 ° Rivers C & D X Rivers E & F
200400 Cells
400800 Cells
1,600
1,6003,200
Cells
Cells
800-
964 721
11,680
6,688 576
60 106 35
3,2006,400 Cells
73 325 628 124
9,875
1,357 5,027 1,035 516 338-D
6,559-E
3,718-C 17,284-F
* Evolved radioactivity In counts per minute.
Table 3-Coliform Counts and Evolved Radioactivity (Routine Samples of Polluted Water) Source
Radioactivity. Net CPM
River A "p "1)
296 104 1
1,357 73 17
Drain Ditch "
740 (est) 193 43
5,027 325" 60
259 185 73 370 (est) 148
1,035 628 106 516 124
1)
PIP
River B1 B1 B1 9 B2 " B2 " B2 River C River D River E River F
Sewage "9 "9 "9 "9
830
Coliform Counts/100 ml
62
35
5,000 1,702 1,330 23,000
3,718 338 6,559 17,285
370
6,688
74 25 16 14
721 46 13 106
Method
Samples filtered through MF "
"
"
" " "
" " " 9
0.1 ml planted direct ' "
"
VOL. 54. NO. 5, A.J.P.H.
RAPID CARBON14 TESTS Table 4-Coliform Counts and Evolved Radioactivity (Public Water Supplies) Type of Water
Rawv (untreated) Finished (NG) Finished (NG) Finished (NG) Raw (untreated) Finished (NG) Finished (HG)* Finished (NG)
Coliform Count/100
Net CPM
Type of Water
Coliform Count/100
Net CPM
5,000 0 0 0 1,702 0 0 0
3,718 13 105 69 166 20 223 28
Raw (untreated) Finished (NG) Finished (NG) Finished (NG) Raw (untreated) Finished (NG) Finished (SG)t
1,330 0 0 0 23,000 0 0
6,559 5 10 9 17,284 1 12
Estimated 5,000 noncoliform colonies. Above counts were made with the membrane filter technic. NG means no growth. SG means slight growth. HG means heavy growth.
*
t 10 noncoliform colonies.
ings show a definite trend. The higher cell groupings may conform better when sufficient data are accumulated. Table 2 shows the actual counts of various dilutions of samples of sewage, rivers, and drainage ditches. Again it should be noted that there is a definite increase in counts as the bacterial populations increase. Table 3 gives some routine counts of coliform bacteria with the corresponding radioactivity in CPM.
Table 4 shows the results of tests of some raw untreated waters and finished treated waters with the corresponding radioactivity. The finished water samples listed below each raw water sample are from the same source. Table 5 shows the results of Table 1 and Figure 1 combined. In Figure 1, the average of the logs of the CPM for the various cell groupings are plotted, and the plotted line for each grouping is converted to the geometric average, so
Table 5-Theoretical and Actual Radioactivity Findings in Relation to Numbers of Bacteria Bacterial Cell Counts
Theoretical Test Results Geo. Mean CPM Geo. Mean CPM from Figure 1 from Table 1
1-25 25-50 50-100 100-200
19 44 120 200
200-400
1,200 990 1,500 9,900 7,900
400-800 800-1,600
1,600-3,200 3,200-6,400
18 44 90 200 510
1,000 2,500 5,600 13,000
Actual No. of Tests 22 3 4 4 6 3 2 1 3
This table is an attempt to convert the line in Figure 1 to theoretical averages with a range of counts which may be expected. Insufficient data in the higher ranges are recognized.
MAY. 1964
831
-_____________
t3#,h.
ttda
~~
........
..
..... ....-
g..
. ... . .
...........
*). ta
,
-
..l.
:,.".
4,
:....
t
..
I
"' A I .... i -.1
::
Figure 2
that column two of Table 5 shows the actual geometric mean for the test results, and in column three there is a theoretical geometric mean from the plotted line on Figure 1. This would be the theoretical results in CPM for each cell group, all things being equal. At this point, it was decided that the use of pure cultures of coliforms would advantageously provide a "calibration" of the instruments for coliform activity only. This also provided a better numbers control. The equipment was also changed, and the counter used was a "Nuclear of Chicago" proportional counter with a one inch detector head with end window. Cultures used were from 24 hour lactose broth tubes with appropriate dilutions so that the cell populations were approximately 50, 500, and 5,000. Numbers controls were made on nutrient agar. However, a series of tests made in 832
accordance with the described technics gave little activity. In seeking the cause, consideration was given to the possibility of the toxic effect exerted by the membrane as reported by Levin.6 Various technics were applied in an effort to overcome this condition. These included filtering the sample through powdered charcoal, boiling the membrane in thioglycollate, and boiling in Sorensen's buffer. None of these methods produced better results. Efforts to induce lag phase by refrigeration, both in media and dilution water, and by allowing the cultures to set overnight, both at refrigeration and room temperatures, were unsuccessful. The use of different propagating media also proved fruitless. Finally by resorting to longer incubation periods, it was found that organisms transplanted a few times in laboraVOL. 54. NO. 5, A.J.P.H.
RAPID CARBON14 TESTS
tory culture media require a longer first incubation time than do those obtained from natural waters. Furthermore, different cultures do not give the same amount of activity under the same condition. The activity of three different cultures incubated three and one-half hours and five and one-half hours is plotted on Figure 2.
Discussion Radioactive decay is a constant phenomenon and, all things being equal, we should have a fairly constant line on Figure 1. The fact that we do not have is an indication of other influencing factors, such as the presence of other bacteria that may utilize formate to some degree, or possibly synergism. The factors of age, attenuation and natural environment of the sample bacterial cells would also influence the rate of metabolism and, correspondingly, the amounts of evolved C14 dioxide. Too, different cultures would react differently in a given situation. Table 4 shows that high numbers of noncoliform bacteria may cause increased counts, although not to the degree that coliforms do. A study of the data on Table 4 also indicates that if there is no bacterial growth (such as would be expected in a sample of an average finished drinking water) there will be little, if any radioactivity. In water with a high mixed bacterial population, some increased radioactivity may be manifest. Whenever coliforms were present, there was radioactivity. The data also indicate
that heavy populations of mixed bacteria may give a rise in radioactivity.
Conclusions 1. This radioactive carbon technic shows definite promise as a quick test for pollution in time of emergency, since it can be completed in about four hours. 2. The results of the tests on Table 4 show that drinking water with no bacterial growths produce little radioactivity, while a raw or contaminated water shows activity proportional to the density of the coliform present. If activity caused by other bacteria is detected, it must come from organisms in the water itself; although this is undesirable, it is probably not dangerous. Any error in interpretation of results would be made on the safe side. 3. Efforts to set up standards using pure cultures as a criteria of coliform activity required longer incubation periods than in the test itself, thus unduly delaying the results obtainable with this technic. REFERENCES 1. American Public Health Association, Standard Methods for the Examination of Water and Waste Water (llth ed.). New York, N. Y.: The Association, 1960. 2. Oginsky, E. L., and Umbreit, W. W. Introduction to Bacterial Physiology (2nd ed.). San Francisco, Calif.: W. H. Freeman, 1959. 3. Levin, G. V.; Harrison, V. R.; and Hess, W. C. Preliminary Report on a One-Hour Presumptive Test for Coliform Organisms. J.A.W.W.A. 48:75, 1956. . Use of Radioactive Culture Media. Ibid. 4. 49:1069, 1957. 5. Levin, G. V.; Harrison, V. R.; Hess, W. C.; Heim, A. H.; and Stauss, V. L. Rapid Radioactive Test for Coliform Organisms. Ibid. 51:101, 1959. 6. Levin, G. V.; Stauss, V. L.; and Hess, W. C. Rapid Coliform Organism Determination with C-14. Presented at the Rudolf's Research Conference at Rtitger's University, Juine 20, 1961.
(Part 11 of this article follows on page 834)
MAY, 1964
833
11. RAPID CARBON14 TEST FOR SEWAGE BACTERIA A REVIEW of the data reported at the April, 1960, meeting of the Middle States Branch of the American Public Health Association held in Chicago' revealed the need for determining the cause of inconsistent results and occasional high counts. Consideration was given to the choice of media and the incubation temperatures. The method used for gas collection had proved to be less than satisfactory because self-absorption,* back scatter,t and geometryi would be somewhat different in each case.2 Media, while of importance, should not present too great a problem since any medium supporting good growth and oxidation of formate should be satisfactory. A medium with an optimum pH for growth that permits a maximum of carbon dioxide liberation without the formation of the bicarbonate ion was desired. Therefore, a few comparative tests were made with various media and as shown in Table 1, the results did not differ greatly. We therefore selected the m-Endo broth MF (membrane filter) media since it was especially designed for water and sewage bacteriology. Geometry, back scatter, and self-absorption were the factors that seemed to * Refers to the "absorption of radiation emitted by radioactive atoms by the matter in which the atoms are located; in particular, the absorption of radiation within a sample being assayed."** t "The deflection of radiation by scattering processes through angles greater than 90 degrees with respect to the original direction of motion."** t "The fraction of the total solid angle about the source of radiation that is subtended by the face of the sensitive volume of a detector." "Not all radiation from a sample is emitted in the direction of the detectors. The geometry factor accounts for the fraction emitted in the proper direction."** ** As defined by radiological handbook of the U. S. Department of Health, Education, and Welfare, Division of Radiological Health.
834
demand most attention. The pads used thick, inviting self-absorption, and their curling and twisting sometimes evident after drying caused considerable back scatter. The counting of these one inch planchets in a two inch detector head would not give the best geometry. To correct the latter, we changed from a two inch detector head to a one inch detector. The instrument now used is a Nuclear of Chicago 192A ultra scaler, with a D47 one inch detector with mylar window. This has a calculated 46 per were
cent
deficiency.
The thick gas collecting pads were discarded and various other materials and methods were tried. The regular bacteriological cellulose acetate membranes did not prove satisfactory because they curled up immediately on drying. Addition of acetone and other solvents to the membrane did not prove satisfactory since it caused wrinkling during the solvent process. Number 51 Whatman paper retained barium hydroxide well, but also curled on drying. Other filter papers which absorbed barium hydroxide well also curled on drying. Addition of Aquadag did not hold the membranes, and plastic sprays, while holding the membranes flat, were not satisfactory because a consistent amount could not be added. Double coated Scotch tape was tried and finally adopted for use. We discarded the aluminum planchets because the concentrated barium hydroxide reacted with the aluminum giving off hydrogen which formed gas pockets under the filter disc. Stainless steel planchets which were one inch in diameter and one-fourth inch deep were adopted for culture media. The collecting planchet was 11/4 inch wide and 1/8 inch deep. The filter paper was cut into round discs, 11/8 inch in diameter, and when saturated with barium hydroxide, VOL. 54. NO. 5. A.J.P.H.
RAPID CARBON14 TESTS
provided a good gas seal over the one inch lower culture planchet. The double coated Scotch tape was used as follows: A disc of Whatman filter paper, 11/8 inch in diameter, was cut and sealed on one side of the double coated Scotch tape. Just prior,to using, the protective coating of the other side of the Scotch tape was removed and the disc was sealed onto the planchet. This provided a constant thickness of filter pad which retained all the barium hydroxide on the surface of the filter paper disc. These were then inverted over the lower culture planchet during the gas collection interval and remained flat during drying, giving a reasonably constant geometry. It became necessary to treat the lower planchets with Desicote to keep the media from "climbing" during the incubation period due to excess moisture.
Procedure Media and Reagents: Barium Hydroxide Solution-Saturated solution of reagent grade Ba(OH)2 in distilled water. Add 1 per cent dry sucrose. Collodion Solution-0.5 per cent solution of collodion in acetone. MF Endo Broth Sodium Formate-C14 specific activity of 4 millicuries per millimole. Equipment: 1. Planchets-Stainless steel, 1 inch in diameter by 1/4 inch high with flat bottom. 2. Planchets-Stainless steel, 11/4 inch in diameter by 1/8 inch deep with flat bottom. 3. Scotch tape, double coated.
4. Membrane filter-1 inch glass filter
holder. 5. Membrane-I inch HA filters 0.47 micron pore size. 6. Glass cover slips of sufficient size to cover the 1 inch planchet. 7. Filter paper-Number 51 Whatman is recommended. Preparation of Media:
Radioactive sodium formate was acquired in a standard package of 0.05 millicuries with a specific activity of 4 millicuries per millimole. Sodium formate has a molecular weight of 68.02 grams and 1 millimole would be 68.02 milligrams. Therefore, there are 4 millicuries of activity in 68 milligrams. 0.05 millicuries 4 millicuries (activity)
x 68 mg (1 millimole) X=0.85 mg or 0.05 millicuries is contained in 0.85 mg of sodium formate.
It had been previously determined' that we should use a medium containing 0.004 per cent of C14 sodium formate. Therefore, we made up the formate in an 0.008 per cent solution. 0.008 grams 0.85 mg x 100 ml X= 10.625 ml of water.
The 0.85 mg of formate was dissolved in 10.625 ml of sterile water and stored in the refrigerator. Just prior to using, equal volumes of the above solution and MF broth were mixed. This makes a final total volume of
Table 1-Counts per Minute on Media Tested Agar Coliform Count
Lactose Broth
74 70 54 Av.
2.0 8.4 7.4
MAY. 1964
5.9
Broth
Sulfate Broth
MF Endo Media
6.2 10.5 19.5 12.0
0.6 2.1 cont.
17.2 22.5 9.0 16.2
Purple Broth
Tryptose Phosphate
Base
14.8 15.7 0.2 11.1
Lauryl
1.3
835
Cl
r-4
u
E
C> 0
06
Cl
o
tCl1co -
Cl c0
'o
0
Co
C.
Cl
t-t-
0
o -
;Ul)C/) -
5.
oC
CY,o
o.
E
U4
0 --
t-
C
C
c
O
-
4
Cl
Cl
C,)
Clo mCl
C
Co
UCu
o -
m
U&uCu
co
L
O
"
0
Cu
0 o 0
u
Cu
LO
U0
t-) 'I) LI)
S. _
*-
t-
E
Lr) oo
oO e Itt:
00% O %0%%
r-0%C Io c
+
C.
Cl)
gt
PCu0
o
co C0% L Co
oo LOf
"C co 0%N
O OO
H
Co
C
L)
O
-
CO %0 Cl oo co Li) (ON %0
NC-
ClN
04 C
LI
_
Co CI
0
C) C o t cC cor 0
._
0%
Cu 5.
0~07
L
0_
co%0lCd
Lf~
%OCCN
L
ClTC6
d L)
4
C
4 C--
"Cu PC Cu
s
5.O
CD
Cl
l Cl
o) o
e C6 e Cl
o
o
U oSrz't-e U:i N
Cl) ow Ca
I0.
o co ---~ LI Co C--
d0
Co CI) Cl Cl
7
Lf~~LI)
8-
LI)
C)I)
co
LO
CO4
C)-
co £
LI-
C)%01 C Cl 0
co
C-
Cl Co CC
cl
8
CD
A
C2. C/)
0% 0% 00I C) 8 C) C>
Co oy
C)
o
PC4
r
It
Id
Cl Co r-
C
C)
Co Co 0 C8 Co Co C)
C)
8- L)o C0
o Co- 8 Co 8 o 8 o o o
co
-
Co
o
CLI)
C-- M4 C) LI) L) O-'DC--LDO
Cl~
a)
Co Co Co
C 000
~0 Cu4
o
000
e
o
Cl
_
C: m 'IC cs s c
%D 'IC C
C
C° No %0 C CN 'I CC C CD eI c
C
C
C-
-t
a
E--
cts Cu
836
Cd
Cd
i
4
cd
cds
eid
Cd
cd
O) O)
C
u)
¢
U
VOL. 54, NO. 5, A.J.P.H.
RAPID CARBON14 TESTS Technic:
21.25 ml of media and at 0.5 ml per test should make 42.5 tests. Since we have a total of 0.05 millicuries (50 microcuries) 50 microcuries we have
42.5
curies of C14
per
one
1. Filter sample through a 1 inch membrane, type HA pore size 0.47 microns, place in the bottom of the 1 inch planchet, immediately add 0.5 ml of the MF formate media, and cover with a cover slip.
or 1.17 micro-
0.5 ml of media
or
test.
Figure 1Averages* for Number of Coliform Bacteria and for Carbon-14 Counts per Minute for 1 Milliliter of Sewage Effluent and Dilutions to 0.1, 0.01, 0.001, and 0.0001 Milliliter and Trend of Averages of Number of Coliform Bacteria 1,
OO,C000 _
* * -
E Y ~~~~~K Average* Number of Coliform Bacteria Average* Carbon-14 Counts per Minute Trend of Number of Coliform Bacteria
43 £ 0
100, 000
co
o04
-I ^
4 )
0
10, 000
/i.~~~~~~
_
OH
1o,000
,O0i C) C)
/~~~~~
1,OOO
S
0
*
-H
Id
A
100
so
a0)
V 1
Milliliters
.0001
.001
.01
.1
1
5 4
35 32
370 400
14,200 4, 200
71,,000 27,000
(6)
(11)
(15)
(17)
(12)
Average*: Coliform Bacteria C-14 Counts per Minute No. of Measurements *
Averages
MAY, 1964
are
based
on
logarithms.
837
Figure 2-Carbon14 Counts per Minute and Number of Coliform Bacteria for 1 Milliliter of Sewage Effluent and Dilutions to 0.1, 0.01, 0.001, and 0.0001 Milliliter 1,000,000 K EY d
1..f v
+
-
Counts per Minute
*
-
Coliform Bacteria
100, OOC
ci
11 e0
5
1. 1. v
td
+
0-4
10,000
v -
v
10-
19C)
I
1.0
I A 9
1,000
4-
~4- :-0-0-0
-;:-
v
k v
4. + 100
0.
9
4. 10
0
~~~~~4-4-
-in~~~~~-
.18.
011
1
.1
.01
.0001
1 Milliliter of Sewage Effluent and Dilutions
2. Incubate three and one-half hours under moist conditions at 350 C. 3. Prepare the gas collection planchet as follows: To one side of the double coated Scotch tape press a filter paper disc, 11/8 inch diameter, and remove the second layer of protective paper from the tape. Press this side of the tape into the bottom of the planchet. Add three drops of the Ba (OH)2 solution and replace the cover slip with the prepared planchet. These can be prepared in advance. This whole procedure should be planned so that the entire operation does not take more than five minutes. 4. Re-incubate 30 minutes under moist conditions at 350 C. 5. At the end of the incubation period the
upper
planchets with the Ba(OH)2
pads are removed and air dried for a few minutes. 6. After partial drying, the pads are covered with three drops of the collodion solution and dried in a warm oven or on a hot plate. 7. The planchets containing the dried 838
pads
are then counted for five minutes in a gas proportional counter.
With each series of test results, a sterile water control is run, and the resulting CPM is subtracted from the test results to give a net CPM for each
sample. Results
Levin, et al.,3 used the British McConkey method which has an incubation temperature of 440 C which will eliminate many of the aerogenes types and permit the E. Coli to grow. In order to compare our results with standard tests for coliform established by the U. S. Public Health Service, we decided to continue our tests at the 350 C temperature using regular MF media. While this may not be as specific for pollution with fecal material as Levin's method, it should compare more favorably with results obtained by the standard drinking water technic. At this temperature, we also ran the risk of interference from VOL. 54. NO.
5.
A.J.P.H.
RAPID CARBON14 TESTS
the presence of other bacteria which utilize formate. Previous work' had demonstrated that under these conditions pure cultures of coliform bacteria were not as active as
water-borne cultures. Since the detection of sewage contamination was the prime objective of this study, it was decided to use sewage as the coliform source rather than pure cultures. Strong medium, and
Figure 3-Number of Coliform Bacteria in 1 Milliliter of Sewage Effluent and Dilutions to 0.1, 0.01, 0.001, and 0.0001 Milliliter-Observed Data, Averages, and 90%o Confidence Limits 1,000,000
100, 000 a)
.-Iq
0
10,000
C)
Cl) C-)
-H
i 0 4c0
1,000 0
$4
sq)
100
10
1
.0001
.001
.01
.1
1
5
35
370
4, 200
71,000
Upper Limit Lower Limit
25
170
2,200
26,C000
240,000
1
7
62
700
21, 000
'Number of Measures
(6)
(11)
(15)
(17)
(12)
Milliliters
-
-
-
-
Coliform Bacteria: Average * 901 Confidence:
*
Based on
MAY. 1964
logarithms. 839
Figure 4-Carbon14 Counts per Minute for 1 Milliliter of Sewage Effluent and Dilutions to 0.1, 0.01, 0.001, and 0.0001 Milliliter-Observed Data, Averages, and 90% Confidence Limits 1,000, 000
1o0, 000
0L)
10,000
:3 a) P4
U)
to
1, 000 '-H
0I
A,
0
100
10
1
Milliliters Carbon-14 c. p. m. Average * 90% Confidence: Upper Limit Lower Limit Number of Measures *
.0001
.001
.01
.1
1
4
32
400
4,200
27,(000
74 0
340
2,000
3
80
16,ooo 1, 100
50,4000 15, 000
(6)
(11)
(15)
(17)
(12)
Based on logarithms.
weak sewages were used, and decimal dilutions were made and run at the same time. To validate the data and also to determine if the coliform counts could be "forecast" by this method, coliform tests were run on each sample. All of the results run by the described 840
procedure are shown in Table 2. When all the net CPM and coliform counts of each sewage dilution are averaged and plotted against the known milliliters of sewage, we obtained a good correlation among the coliform counts, net CPM, and the sewage dilution (Figure 1). A VOL. 54, NO. 5. A.J.P.H.
RAPID CARBON14 TESTS
scatter diagram of all the information on Table 2 and Figure 1 is shown on Figure 2. These data show that, on the average, there is fairly close correlation among the net CPM, the coliform counts, and the sewage dilution. A statistical evaluation was made by the Bureau of Statistics, Illinois Department of Public Health. The 90 per cent confidence limits were calculated for the coliform counts per dilution of sewage (Figure 3). The same statistical evaluation was made for the net CPM and the results are shown on Figure 4. It is
worthy of note that only two results exceeded the 90 per cent limits. Sewage samples vary in strength in that a 1 ml aliquot of a weak sewage may contain approximately the same number of organisms as a 0.1 ml aliquot of a strong sewage. Therefore, it would not be expected that when several different strength sewages are examined that all results in one dilution would be uniform. Furthermore, the results, as an average, would not be representative if only one strength sewage were used. Since we used different strength sewage, the re-
Figure 5-Number of Coliform Bacteria Associated with Carbon-14 Counts per Minute Individual Observations, Averages* for Each Dilution, Regression Line and 90% Confidence Limits 100,000 r
I
r
I
I I
KEY
Individual Observations Average for Specified Dilution Line of Regression - - 90% Confidence Limifs * Note. Avgs. are based on logs. *
10,000
.L.
~4-,0 a
_
-I .2 tt
0
-o E
z
10
I
L I
to
a
I,oa'
55.0l'
Carbon-14 Counts per Minute (Logarithmic Scale) The 61 observations are based on 17 samples of sewage effluent taken at Ottawa, Astoria, Springfield and Lincoln, Ill., from October 16, 1962, to February 6, 1963, measured at 1 milliliter and dilutions to .1, .01, .001, and .0001 milliliter.
MAY. 1964
841
Table 3-Coliform Waters
Counts,
Standard Plate Counts, and CPM of Finished Drinking
Date
City
Amount Filtered in ml
10-26-62 10-29-62 11- 5-62 11- 5-62 11-20-62 11-20-62 11-26-62 11-26-62 11-30-62 11-30-62 12- 3-62 12- 3-62 2- 1-63 2-15-63 2-15-63 2-15-63 2-15-63
Springfield Springfield Springfield Albany Waukegan Springfield Springfield Rock Island Pekin Springfield Sparta Springfield Springfield Carthage Carthage Carthage Springfield
50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50
Coliform Count 50 ml 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Total
Agar Count per 50 ml
Net CPM
* * * 29,000,000 150 200 250 250
- 5.6 - 9.2
29,000 2,500 20,000,000 600 1,500 200 100 350 *
-11.2 56.6 1.6 - 8.0 13.4 16.4 4.5 -- 5.0 3,972.0 - 3.9 - 1.0 -10.8 - 3.0 - 2.6 - 0.6
* No restults obtained. These data show that with only four exceptions (and two of these were slight increases) coliform free water gave net counts of less than 10 CPM and that high counts other than coliform may increase the CPM.
sults of the various dilutions would be expected to overlap one another. For example, the results obtained from 0.01 ml of strong sewage may approximate the results when using 0.1 ml of weak sewage. This then would give a "cloud" effect along the line of averages when the results are plotted individually. This is shown in Figure 5 where the log of
each CPM result is plotted against the associated bacterial log counts. The results, as would be expected, grade from the lowest dilution upward through the highest dilutions. The averages for each sewage dilution and the 90 per cent confidence limits are also shown. In order to test drinking water samples where no coliform or other bacteria is
Table 4-Counts per Minute of Two Concentrations of C-14 Under the Same Test Conditions
Control (No inoculum) Control (No inoculum) 0.001 ml sewage 0.01 ml sewage 0.01 ml sewage 0.1 ml sewage 0.1 ml sewage 1.0 ml sewage
842
C-14 1.17 ,uc CPM
C-14 0.58 Ac CPM
Coliform Counts by MF
Bacterial Count on Agar
83.6 43.4 3.2 137.2 191.0 1,569.4 1,375.4 35,611.8
71.0 25.2 16.0 103.4 193.6 439.2 678.4 17,701.0
34 340 620 3,400 6,200 62,000
1,100 11,000 24,000 110,000 240,000 2,400,000
VOL. 54. NO. 5. A.J.P.H.
RAPID CARBON14 TESTS
expected, we ran a series of waters from various distribution taps from public water supplies (Table 3). The fact that bacteria other than coliform normally found in drinking water,
streams, and soil3 may utilize formate in varying amounts is a complicating factor. This makes the forecast of the probable number of coliform bacteria less reliable. However, for the purposes of this
Figure 6-Carbon-14 Counts per Minute and Number of Coliform Bacteria for 1 Milliliter of Sewage Effluent and Specified Dilutions-Conversion Scale, Gauge for 95% Confidence, and Table of Central Values and Ranges C O N V E
S I O iN
S C At L E 40, 000
100,000-_ 80,000 6o, ooo -
-AUGP:
20, )
40,000
1 ml.
-axirum
20, 000 - -10,000
8,o0o0 6,o000
10,00C-
8, ooo
6,000 4,000
.1 m.
Carbon-14
FOR
Counts per Minute
95";
24,000
-
LIMITS
-
M iniyminm
?I, 000 2,000
-
L8000
:8: 6oo 2
1,
600