PB-258 099
STANDARD TEST FISH DEVELOPMENT PART I
Minnesota Univ, St Paul
Prepared for:
Environmental Protection Lab, Duluth, Minn
July 1976
DISTRIBUTED BY:
National Technical Information Service
U. S. DEPARTMENT 0F
5285 Port Royal Road, Springfitid Va. 22151
TMs has ton approved for release and sale.
-------
282013
EPA-600/3-76-QI1a
July 1976
PB 258 099
Ecological Research Series
REPRODUCED BT
NATIONAL TECHNICAL
INFORMATION SERVICE
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmente! technology. Elimination of traditional grouping was consciously
planned to (otter technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal
species, and materials. Problems are assessed for their long- and short-term
influences. Investigations include formation, transport, and pathway studies to
determine the fate of pollutantsand their effects. This work provides the technical
basis for setting standards to minimize undesirable changes in living organisms
in the aquatic, terrestrial, .and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Entomology, Fisheries, and Wildlife
University of Minnesota
St. Paul, Minnesota 55108
TECHNICAL REPORT DATA
ftfcase read Inunftiuti* on the rcrtrif before coniplctingf
1 REPORT NO.
EPA-600/3-76-061a
[. RECIPIENT'S ACCESSION>NO.
STANDARD TEST FISH DEVELOPMENT, PART I
6. PERFORMING ORGANIZATION CODE
7. AUTHCRiS)
Ira R. Adelman and Lloyd L. Smith, Jr.
10. PROGRAM ELEMENT NO.
1BA608
12. SPONSORING AGENCY NAME-ANO ADDRESS
U.S. Environmental Protection Agency
Office of Research and Development
Environmental Research Laboratory
Duluth, Minnesota 55804
. RPPnRT DATE
July 1976 (Issuing Date)
. PERFORMING ORGANIZATION REPORT NO.
Grant R800940
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA-ORD
See Part It, EPA-600/3-76-061b
Fathead minnows and goldfish were compared for their suitability as a standard
bioassay fish. Acute bioassays of four potential reference toxicants, sodium
chloride, pentachlorophenol, hexavalent chromium, and Guthion", were conducted
with both species, and results were reported as toxicity curves as well as LC50's
at various times. Both species showed the same variability of bioassay results.
Since goldfish could not complete a life cycle in 1 year under laboratory conditions,
fathead minnows were recommended as a standard species on the basis of their smaller
size and their utility in complete life cycle tests.
Bioassays of .pentachlorophenol were conducted with both species to determine
the effect of testing different sized fish of the same age or testing different
aged fish. Size selection of fish within the ranges tested is unnecessary since
differences in LCSO's were small. Since age of fathead minnows did not affect the
LCSO's after 24 hours, use of younger fish would allow smaller bioassay chambers
or more fish per chamber.
On the basis of seven criteria, sodium chloride was superior for use as a
reference toxicant with pentachlorophenol a close second choice.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Age
Bioassay
Fishes
Chromium
Sodium chloride
Size determination
Toxicity
5 :.1STR!Ev* C'J STATEMENT
RELEASE TO PUBLIC
b-IDENTIFIERS'OPEN ENOEDTERMS
Goldfish, Guthion,
Acute toxi.city, fathepd
minnows, hexavalent
chromium, pentachloro-
phenol, reference toxi-
cant, goldfish culture
• 19. SECURITY C.4SS • Hin
\ UNCLASSIFIED
COSATi 1 ictil<'<.;niii[>
06C
06F
06S
UNCLASSIFIED
I.
-------
EPA-600/3-76-06U
July 1976
STANDARD TEST FISH DEVELOPMENT
PART I
Fathead Mitmovs (Piaephales promelas) and Goldfish (Carassius
auratus) as Standard Fish In Bloassays and Their Reaction
To Potential Reference Toxicants
by-
Ira R. Adelman
and
Lloyd L. Smith, Jr.
Department of Entomology, Fisheries, and Wildlife
University of Minnesota
St. Paul, Minnesota 55108
Grant No. R800940
Project Officer
Robert W. Andrew
Environmental Research Laboratory
Duluth, Minnesota 55804
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL RESEARCH LABORATORY
DULUTH, MINNESOTA 55804
I0u
-------
DISCLAIMER
This report has been reviewed by the Environmental Research Laboratory -
Duluth, 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 recommen-
dation for use.
ABSTRACT
Fathead minnows and goldfish were compared for variability of LCSO's
with four toxicants as a means of determining their suitability as a
standard bioassay fish. Acute bioassays with sodium chloride, penta-
chlorophenol, hexavalent chromium and Guthiorr* were conducted with
both species; and results were reported as toxicity curves as well as
LCSO's at various times. Both species showed similar variability of
bdoassay results. ^_
Bioassays of pentachlorophenol were conducted with both species to
determine the effect of testing different, sized fish of the same age or
testing different aged fish. Size selection of fish within the ranges-
tested is unnecessary since differences in LCSO's were small. Since age
of fathead minnows did not affect the LCSO's after 24 hours, use of
younger fish would allow smaller bioassay chambers or more fish per
chamber.
Since goldfish could not be induced to complete a life cycle in the
laboratory in less than one year, fathead minnows were superior as a
standard species because of their smaller size and their utility in
complete life cycle tests. -
On the basis.of seven criteria, sodium chloride was superior for use as
a reference toxicant with pentachlorophenol a close second choice.
This report was submitted in fulfillment of Project 18050 HOH 'and
Grant R800940 by the Department of Entomology, Fisheries, and Wildlife,
University of Minnesota, under the sponsorship of the Environmental
Protection Agency. Work was completed as of April, 1975.
iii
-------
CONTENTS
Page
Abstract Hi
List of Figures vi
List of Tables vii
Acknowledgements x
Sections
I Conculusions 1
II Reconmendations 2
III Introduction 3
Standard Fish Studies 3
Reference Toxicant Studies 5
Determination'of Suitable Size and Age 6
IV Materials and Methods 8
Acute Tests 8
Goldfish Culture Study 27
V Results 33
Acute Tests 33
Goldfish Culture Study 56
VI Discussion 67
Selection of a Standard Species 67
Selection of a Reference Toxicant 69
VII References 71
VIII Publications 74
IX Glossary 75
Preceding page blank
-------
FIGURES
No.
1 Spatial Arrangement of Test Chambers Drawn to Scale in.
Goldfish Culture Study.
2 Toxicity Curves of Sodium Chloride, Pentachlorophenpl,
GuthionTand Hexavalent Chromium for Goldfish and Fathead
.Minnows.
3 Toxicity Curves of Hexavalent Chromium for Goldfish and
Fathead Minnows.
It Relationship of Reciprocal of LC50 Versus Reciprocal of
Time for Goldfish of Different Sizes and Fathead Minnows
of Different Sizes and Different Ages.
vi
Page
28
34
55
TABLES
No. Page
1 Analysis of Well Water 9
2 Acute Toxicity and Mean Test Conditions in Bioassays to 12
Determine Variability In Response of Fathead Minnows and
Goldfish to Sodium Chloride
3 Acute Toxicity and Mean Test Conditions in Bioassays to 14
Determine Variability in Response of Fathead Minnows and
Goldfish to Pentachlorophenol
4 Acute Toxicity and Mean Test Conditions in Bioassays to 16
Determine Variability in Response of Fathead Minnows and
Goldfish to Hexavalent Chromium
5 Acute Toxicity and Mean Test Conditions in Bioassays to 18
Determine Variability in Response of Fathead Minnows and
Goldfish to Guthioii*
6 Test Conditions in Bioassays with Large and Small Fathead 20
Minnows of Constant Age
7 Test conditions in Bioassays with Fathead Minnows of 21
Various Ages
8 Test Conditions in Bioassays with Large and Small Goldfish 22
from a Constant Stock
9 Diet of Two Groups of Goldfish During Different Periods 32
of the Experiment
vii
-------
No. Page
10 Means and 95Z Confidence Intervals of LCSO's of Four 36
Toxicants and Two Fish Species at Various Times
11 Means and 95% Confidence Intervals of LCSO's of Hexavalent 42
Chromium from Results of the Second Experimental Procedure
12 Mean LCSO's with Coefficient of Variability of Four 43
Toxicants at Various Times
13 Occurrence of Significant Differences in Multiway Analyses 46
of Variance with Four Toxicants
No.
21 Mean Weights of Fish from All Individual Tanks at
Approximately Monthly Intervals
22 Standard Lengths of Individual Fish in All Tanks
at Termination after 393 Days
23 Sex and State of Gonadal Maturity of Individual Fish
in All Tanks at Termination after 393 Days
62
64
66
14 Comparison of LCSO's from Unhealthy Goldfish Stock with
Mean LCSO's from All Tests with "Normal" Goldfish
49
IS Comparison of Four Toxicants with Regard to Their Use
as a Reference Toxicant
51
16 Mean and Standard Deviation of Weight and Length of Fish 52
in All Bioassays
17 Regression Equations and Coefficient of Determination (r ) S4
for the Relationship Between LCSO and Time in Three
Groups of Experiments
18 Means and Standard Deviations of LCSO's with Pentachloro- 56
phenol for Different Groups of Fish at Various Times
19 Means and Standard Deviations of Conditions in Goldfish
Culture Study
58
20 Weight of Individual Fish In All Tanks at Termination
after 393 Days
60
viil
\
"i
tx
-------
ACKNOWLEDGEMENTS
The authors wish to thank personnel at the Newtowii, Ohio field station
of the Environmental Research Laboratory - Duluth for their suggestions
on selection of species of fish and toxicants. .We also wish to thank
Robert G. Ruesink, Richard W. Frenzel, and particularly Gary D. Siesennop
for assistance in conducting the experiments.
SECTION I
CONCLUSIONS
The results described in the following report permit certain conclusions
concerning the selection of a standard bloassay fish and choice of a
reference toxicant.
1. Fathead minnows and goldfish are similar in sensitivity to sodium
®
chloride", pentachlorophenol, hexavalent chromium, and Guthiotr^ in
acute bioassays.
2. Variability of LC50's with the four toxicants was similar for both
species and was more dependent on the toxicant than the species.
3. Since fathead minnows can complete a life cycle in less than one
year and the culture study with goldfish was unsuccessful in achiev-
ing this result, fathead minnows will presently meet this criterion
for a standard species.
A. Neither size of goldfish nor size and age of fathead minnows affected
LCSO's of pentachlorophenol at times longer than 24 hours.
5. Sodium chloride best met some of the desired characteristics of a
reference toxicant with pentachlorophenol a close second.
-------
SECTION II
RECOMMENDATIONS
1. Fathead minnows are recomiended for use as a standard bioassay fish.
2. It is further recomended that the appropriate federal agency pro-
mote the development of a private or public, hatchery for this
species to provide uniform stock.
3. It is recommended that approximately 4 to 7 week-old fathead minnows
be used in standard bioassays.
4. Goldfish from a single gene pool are recommended as a standard
fish until a source of fathead minnows becomes available.
5. Sodium chloride or pentachlorophenol are recommended for use
reference toxicant.
6. Further studies are recommended to determine which reference
toxicant is more suitable for detection of abnormal fish.
SECTION III
INTRODUCTION
STANDARD FISH STUDIES
Many authors have mentioned the desirability of a standard fish species
for use in all types of fisheries research (Marking, 1966; Lennon, 1967;
Cairns, 1969; Sprague, 1970). In an extensive discussion on the use of
selected strains of-fish as bioassay animals, Lennon (1967) stated that
the peed for standard strains of fish was "immediate and extremely im-
portant in view of the expansion of fish bioassays in pharmacology,
pollution control, pesticide evaluation, fish culture, and fish con-
trol." This need was based on the requirement for reproducibility of
test results. Since resistance to a toxicant may vary with species,
strain, previous history, age, size, health, and handling procedures,
the first step in achieving reproducibility of results within or between
laboratories would be development of standard bioassay procedures in-
cluding standard water and a standard bioassay fish.
Recently various government agencies have been attempting to define
standard species and tesing procedures for use by industries to monitor
their effluents. The purpose of developing these procedures is to Iden-
tify and isolate potentially detrimental effluents through bioassay and
to determine the relative toxicity of effluents. Since many industrial
effluents are complex mixtures whose toxic components are unknown or
difficult to Isolate, the biological monitor serves as a more rapid and
probably more realistic identification tool for potential hazards than
a chemical analysis. Industrial personnel who would conduct these tests
-------
are usually not trained fisheries biologists, therefore standard testing
procedures and fish strains are necessary to insure valid comparisons of
the potential hazard of effluents.
The Ohio River Valley Water Sanitation Commission has developed a stan-
dard 24-hour static bioassay using a prescribed strain of goldfish for
monitoring industrial effluents (ORSANCO, 1974). MacLeod (1972) has
proposed a similar program using rainbow trout to establish uniform
enforcement regulations throughout Canada. The Water Resources Control
Board of California is also attempting to designate strains of one or
more fish species for use in a statewide industrial monitoring program.
The more important criteria for selection of any strain used as a stan-
dard fish for bioassay are: 1) relatively constant response to a broad
range of toxicants when tested under similar conditions; 2) available
in large quantities with close quality control; 3) easily handled for
bioassay purposes; 4) easily transported; 5) continuous availability of
the desired size; and 6) capable of successful completion of a life
cycle in i year or less under laboratory conditions.
One objective of the present study was primarily related to the first
criterion. If a standard strain is used to make comparisons of the
toxicity of industrial effluents, there must be some indication that
the strain does not show high variability in response when tested under
standardized procedures at different times or locations. To assess
variability, acute bioassays were conducted over a 2-year period with
two fish species frequently suggested as potential standards and four
toxicants with different modes of action, sodium chloride, pentachloro-
phenol, hexavalent chromium, and Guthion-r
Goldfish, Carasslus auratus (L.), were chosen because of their availa-
bility in large quantities from a single commercial source, Ozark
Fisheries, Inc., Stoutland, Missouri, and because of their proposed use
in a standard bioassay procedure for Industries in the Ohio River Valley
(ORSANCO, 1974). The fathead minnow, Pimephales promelas Rafinesque,
was selected because of successful use in many bioassays (Martin, 1973).
A strain of fathead minnows from the National Water Quality Laboratory
at Duluth, Minnesota has been available for research purposes and a
small-scale hatchery design was developed by personnel at that labora-
tory (U.S. Environmental Protection Agency, 1971). If the present
study indicated that the fathead minnow was more suitable, the technology
for development of a commercial source would be available.
A second objective of the present study related to the last criterion
listed. Was the goldfish capable of completing a life cycle within 1
year under laboratory conditions? Previous attempts to real goldfish
from egg through spawning adult under chronic bioassay conditions in
our laboratory were not successful, although fish that had spent at
least the first summer of their life in outdoor ponds readily spawned
in our laboratory when mature. Hervey and Hem (1968) reported that
field-reared fish spawned as early as 9 months, and pond-reared fish
spawned in our laboratory at approximately that age. An experiment was
therefore conducted to examine the effect of temperature, food, and
crowding on the ability of goldfish to mature in the laboratory. These
three variables were suspected as causes of previous failures.
REFERENCE TOXICANT STUDIES
Marking (1966) and Davis and Boos (1975) have discussed the need for a
reference toxicant to determine the condition of fish at the time of
testing. In theory a reference toxicant bioassay under standardized
test conditions would be conducted just prior to or simultaneously with
a bioassay of the toxicant or effluent under consideration. If results
from the reference toxicant bioassay deviated considerably from pre-
viously accumulated data on this toxicant, the fish stock would be
considered abnormal and results of any bioassays with it would be ques-
-------
tionable. This procedure would be useful both to the industry or labor-
atory conducting the test and to regulatory agencies for determining the
validity of test results. Davis and Hoos (197S) conducted a field test
of reference toxicants including pentachlorophenol where the materials
were bioassayed by different laboratories. This study indicated dif-
ferences between laboratories but not within-laboratory variability, the
basis upon which results would in part be judged. Furthermore, tests
were conducted on rainbow trout from a variety of sources rather than a
single stock.
As with the standard fish strain, a reference toxicant would have to
meet certain criteria to be useful. The more important criteria are:
1) minimum variability in response of normal fish; 2) rapid detection of
abnormal fish by a deviant response; 3) rapid lethal action; 4) simple
analytical technique; 5) usable in static and flow-through bioassays;
and 6) general ease of laboratory handling. Three of the four toxicants
used for determination of species variability were selected for their
potential as reference toxicants as well as their different modes of
toxic action. These potential reference toxicants were sodium chloride,
pentachlorophenol, and hexavalent chromium. A third objective of the
present study was to determine how well the tested materials met the
criteria for a reference toxicant.
DETEKMINATION OF SUITABLE SIZE AND AGE
Different shipments of goldfish from Ozark Fisheries varied considerably
in age and weight and since weight of these fish is generally not corre-
lated with age, conditions in the rearing ponds are probably responsible
for size differentials. With fathead minnows age of fish for testing
can be precisely controlled to 1-week age groups, but conditions during
rearing result in considerable variability in size of these fish. Since
the age of the goldfish cannot be controlled, it is desirable to know
what effect size would have when fish are tested with one of the potential
reference toxicants. Variation in a reference toxicant bioassay
attributable to size might require rejection of large or small stocks of
fish. In using fathead minnows of constant age, different sized fish
night also contribute to variability. Since various aged fathead minnows
could be selected by use of the culture system (U.S. Environmental Pro-
tection Agency, 1971), variation in test results might require use of a
constant age in all tests or lack of variation due to age might permit
testing fish of a variety of ages or very young fish. Either of these
latter alternatives would allow for greater culture unit production or
smaller culture units if younger fish are used.
A final objective of the present study was to determine the difference
in sensitivity to one of the potential reference toxicants, pentachloro-
phenol, of different sized fathead minnows and goldfish of a constant
age or stock and to determine the difference in sensitivity of different
aged fathead minnows.
-------
SECTION IV
MATERIALS AND METHODS
ACUTE TESTS
Apparatus and Water
All bioassays of hexavalent chromium and pentachlorophenol were con-
ducted in the same two f low-rthrough proportional diluters constructed
of glass, G.E. Silicone Seal, and Koroseal plastic tubing (Mount and
Erungs, 1967). Each diluter dispensed seven toxicant concentrations
with a dilution factor of 60Z and a control. A series of flow splitters
permitted duplication of each toxicant concentration and control so
that four separate tests could be conducted simultaneously. The test
chambers were constructed of glass and G.E. Silicone Seal, measured
50 x 25 x 20 cm, and contained 20 liters. With a flow rate of 105
ml/min, 90% replacement of water occurred in approximately 7 hours
(Sprague 1969) and 7.6 test chamber volumes were added each 24 hours.
The pH was controlled by dispensing sulfuric acid into the head tank
with a chemical metering apparatus (Mount and Brungs, 1967), and the
temperature was controlled by a hot water heat-exchange system in the
head tank. Hexavalent chromium was reagent grade potassium dichromate
and pentachlorophenol was technical grade sodium pentachlorophenate.
The pentachlorophenol was analyzed for dioxins, a frequent contaminant.
and none were present.
All flow-through tests were made with water from a deep well, pumped
to the apparatus through polyvinyl choride pipe after iron removal
(Table 1).
Table 1. ANALYSIS OF WELL WATER
(milligrams/liter)
Item
Total hardness as CaCOj
Calcium as CaCO,
Iron
Chloride
Sulfate
• Sulfide
j . Fluoride
Total phosphates
!
( Sodium
Potassium
Copper
Manganese
Zinc
Cobalt, nickel
j Cadmium, mercury
Ammonia, nitrogen
Organic nitrogen
Concentration
220
140
0.02
<1
<5
0.0
0.22
0.03
6
2
0.0004
0.0287
0.0044
<0.0005
<. 0.0001
0.20
0.20
Water taken from well head before aeration and heating; pH 7.5.
-------
Inability to dissolve adequate .concentrations of sodium chloride and
Guthlort^ in the proportional dlluters required the use of a renewal
bioassay with these materials. The same number and size test chambers
were used as for the toxicants tested in flow-through apparatus. Each
test chamber contained a 50-vatt aquarium heater and was aerated con-
tinuously .
Preliminary tests indicated that high pH levels in the aerated labora-
tory well water remulted in sublethal accumulations of unionized
ammonia in static tests. In order to maintain a lower pH and keep the
ammonia in a non-toxic ionized form, a soft water was used in these
tests. This water was prepared by diluting laboratory well water with
80% deionized water and buffering with 272 mg/liter mono-basic potassium
phosphate and 25 mg/liter sodium hydroxide. A fresh preparation of this
water was added to each chamber at the start of the acclimation period
and on Introduction of toxicant, and 88% of the water and toxicant was
changed every 2 or 3 days during the test. Chemical analysis Indicated
that there was no reduction of either sodium chloride or Guthioir* under
test conditions for up to 4 days. Reagent grade sodium chloride crys-
tals were dissolved directly into each test chamber, and 93% technical
©
grade Guthlon^ln an acetone carrier was pipetted directly. One control
chamber for each species received a volume of acetone equivalent to that
of the highest Guthiorr-^concentration (1750 ppm) and no more than 10%
mortality was ever observed in this chamber. This 10% loss occurred in
the acetone control In two bioassays and also in the control without
acetone in two tests.
Fish
Fathead minnows were reared in a laboratory hatchery similar to that
designed at the National Water Quality Laboratory (U.S. Environmental
Protection Agency, 1971). The original stock of juvenile fish was
obtained from the National Water Quality Laboratory at Duluth, and at
approximately 8- to 10-month intervals additional spawning stock from
Duluth was mixed with the hatchery-reared spawners to prevent develop-
10
ment of a genetically divergent strain. Stocks of fish could be se-
lected for testing from within 1-week age groups. The fathead minnows
tested for the determination of bioassay variability were 11 weeks old
at the start of each test, and mean weights ranged from 0.12 to 0..13 g
(Tables 2, 3, 4, 5). In studies on the effect of size and age, various
ages and weights were used depending on the objective of each test
(Tables 6 and 7).
Goldfish were obtained from Ozark Fisheries, Inc., Stoutland, Missouri.
Growth of these fish is sometimes intentionally retarded by the supplier
to maintain a stock of the desired size. Upon arrival all stocks were
treated twice on successive days with 1 mg/liter potassium permanganate
for 1/2 hour to remove Gyrodactylus. Chv the third day they were
treated with 25 mg/liter tetracycline. After-treatment-fish were held
for at least 10 days at 21 C in flowing water prior to testing. In
tests of bioassay variability goldfish varied in age from 5 months to
1-1/2 years, and mean weights ranged from 1.37 to 2.70 g (Tables 2, 3,
4, 5). In tests for determination of the effect of size, fish were
approximately 6 months old and size varied depending on the objective
of the test (Table 8).
Bioassay Procedure
Determination of Variability—Eight series of tests were conducted in
duplicate for a total of 16 bioassays per species per toxicant (Tables
2, 3, 4, 5). From 68 to 72 hours prior to the introduction of toxicant
fish were sorted in a stratified random manner into the two sets of test
chambers, with 10 fish in each. After acclimation to the test chamber,
test water, and temperature, the toxicant was immediately brought to
the desired concentration by pipetting from a stock solution for
Guthionr sodium pentachlorophenate, and potassium dichromate, and by
direct dissolution of sodium chloride. In a series of eight tests an
error in calculation resulted in about 35% of the desired hexavalent
chromium concentration initially with desired concentrations not
attained for 3 to 4 hours. In three additional tests with each species
11
\
\
-------
Table 2 (continued). ACUTE TOXICITY AND MEAN TEST CONDITIONS IN BIOASSAYS TO DETERMINE VARIABILITY
IN RESPONSE OF FATHEAD MINNOWS AND GOLDFISH TO SODIUM CHLORIDE
Temper-
ature,
Test C
G1U
GIL
G2U
G2L
G3U
G3L
G4U
G4L
G5U
G5L
G6U
G6L
G7U
G7L
G8U
G8L
25.1
25.3
24.7
24.7
24.8
24.9
25.0
24.4
24.7
24.9
25.2
24.9
24.9
25.2
25.0
24.7
-'Fatheads -
pH,
meter
reading
7.13
7.11
6.95
7.05
.7.20
7.18
6.98
7.01
7.04
7.03
7.16
7.19
7.24
7.19
7.13
7.10
or
mg/1
6.93
6.69
. 6.80
6.38
7.22
7.06
6.38
6.68
7.02
7.08
6.18
6.55
7.31
7.17
7.18
7.27
Weight, I
g
2.53
1 2-44
1.79
1.92
1.57
1 1.63
2.39
2.34
1.53
1.56
1.69
1.65
2.17
2.18
2'. 77
2.83
.ength.-'
Goldfish
35.2
34.4
30.0
31.5
33.0
33.2
33.1.
32.8
32.1
31.9
33.8
33.7
37.8
37.8
38.0
38.7
24 hr
9900
9750
9800
8350
9850
10000
11050
11050
10100
10200
10450
10400
9980
9600
8800
9950
48 hr
7550
7350
7200
6800
7950
7400
8200
8400
7900
7400
8800
8500
7900
7850
7150
6950
LCSO, at
96 hr 1
7000
6950
7050
6800
7600
7200
7900
8050
7200
7050
7950
7950
7650
7300
7000
6800
1/1
threshold
7000
6950
7050
6800
7600
7200
7900
8050
7200
7050
7750
7950
7550
7300
7000
6800
(days)
(6)
(6)
(5)
(4)
(5)
(5)
(6)
(6)
(5)
(5)
(7)
(6)
(7)
(6)
(5)
(6)
Table 2. ACUTE TOXICITY'AND MEAN TEST CONDITIONS IN BIOASSAYS TO DETERMINE VARIABILITY IN RESPONSE
OF FATHEAD MINNOWS AND GOLDFISH TO SODIUM CHLORIDE
Test
F1U
F1L
F2U
F2L
F3U
F3L
F4U
F4L
F5U
F5L
F6U
•F6L
F7U
F7L .
F8U
F8L
Temper-
ature,
C
25.0
24.7
25.2
25.0
24.7
25.0
25.1
25.2
25,0
24.8
25.0
24.8
24.8
25.0
25.2
25.0
pH,
meter
reading
7.16 .
7.12
7.24
7.26'
7.20
7.17
7.05
7.06
7.33
7.28
7.20
7.19
'7.33
7.31
7.41
7.41
or
mg/1
• _
_
7.22
7.48
7.13
7.37
6.09
6.14
6.98
6.85
7.50
7.53
7.11
7.08
7.12
7.05
Weight
g
0.24
0.26
0.21
0.26
0.20
0.19
0.25
0.26
0.26
0.26
0.24
0.24
0.35
0.32
0,30
0.32
, Length ,-
./
24 hr
48 hr
LC50,
96 hr
mg/1
Threshold
(days)
Fathead minnows
27.2
27.6
24.2
25.7
23.7
22.8
27.2
27.0
27.1
27.7
26.4
27.0
26.9
26.1
27.0
27.4
7100
7200
7400
7700
7750
7500
9000
8300
8150
—
7750
8100
8800
8800
7500
7600
7050
7100
7400
7650
7400
7300
8300
8200
7800
7650
7550
7950
8700
8300
7200
7500
7050
7100
7400
7650
7400
7200
8300
8200
7800
7650
7450
7950
8400
8150
7200
7500
7050
7100
7400
7650
7400
7200
8300
8200
7800
7650
7450
7950
8400
8150
7200
7500
(4)
(4)
(4)
(4)
(4)
(5)
(4)
(4)
(4)
(4)
(5)
(4)
(5)
<5>.
(4)
(4)
-------
Table 3 (continued). ACUTE TOXICITY AND MEAN TEST CONDITIONS IN BIOASSAYS TO DETERMINE VARIABILITY
IN RESPONSE OF FATHEAD MINNOWS AND GOLDFISH TO PENTACHLOROPHENOL
G1U
GIL
G2U
G2L
G3U
G3L
G4U
G4L
G5U
G5L
G6U
G6L
G7U
G7L
G8U
G8L
a/
Temper-
ature,
24.2
23.0
25.1
24.3
25.2
24.4
26.2
25.1
25.8
24.6
25.1
24.2
25.0
24.5
24.5
23.8
Fatheads -
pH
reading
7.81
7.78
7.77
7.75
7.62
7.68
7.54
7.59
7.58
7.59
7.58
7.60
7.83
7.84
7.73
7.76
total length;
0 , Weight,
mg/1 ' ._ g
5.51
5.66
6.17
6.17
5.86
6.31
5.33
5.71
5.64
5.84
5.39
5.72
6.02
5.72
5.84
6.32
goldfish
2.36
2.57
1.46
1.50
1.55
1.40
2.46
2.70
1.66
1.74
1.69
1.65
2.31
2.31
1.76
1.54
Length ,-'
Goldfish
35.o'
36.4
30.2
29.5
29.4
28.7
39.5
39.2
33.5
33.8
33.8
33.7
37.0
37.8
34.8
33.8
LC50. rng/1
24 hr
.21
.24
.23
.26
.21
.18
.32
.30
.32
.31
.23
.22
.37
.36
.24
.36
48 hr
.21
.22
.23
.21
.19
.17
.29
.23
.30
.24
.22
• 19
.35
.34
.21
.30
96 hr
.21
.22
.23
.21
.17
.17
.22
.23
.24
.24
.20
.19
.29
.30
.20
.25
Threshold
.21
.22
.23
.21
.17
.17
.18
.23
.21
.24
.20
.19
.25
.28
.15
.17
(days)
(4)
(4)
(4)
(4)
(5)
(4)
(10)
(4)
(10)
(4)
(5)
(4)
(8)
(8)
(11)
(ID
- standard length.
Table 3. ACUTE TOXICITY AND MEAN TEST CONDITIONS IN BIOASSAYS TO DETERMINE VARIABILITY IN RESPONSE
OF FATHEAD MINNOWS AND GOLDFISH TO PENTACHLOROPHENOL
Test
Flu
F1L
F2U
F2L
F3U
F3L
F4U
F4L
F5U
F5L
F6U
F6L
F7U
F7L
F8U
F8L
Temper-
ature,
C
24.6
23.7
24.9
24.3
25.9
24.7
25.8
25.2
25.8
25.1
25.3
24.2
24.6
23.7
24.9
24.4
PH
meter
reading
7.82
7.83
7.72
7.72
7.69
7.68
7.86
7.78
7.59
.7.62
7,65
7. 65
7.63
7.58
7.83
7.82
o2,
mg/1
6.52
6.67
6.37
6.61
7.02
7.18
6.44
6.61
6.27
6.39
6.22
6.48
6.10
6.20
6.78
6.80
Weight, Length ,-'
K mm
0.31
0.33
0.19
0.20
0.21
0.18
0.30
0.30
0.22
0.23
0.22
0.24
0.29
0.35
0.26
0,24
Fathead
28
28
23
23
24
23
28
28
25
25
26
27
27
28
25
25
Minnows
.7
.8
.4
.9
.1
.5
.6
.4
.5
.5
.0
.0
.2
.5
.2
.0
LC50, mg/1
24 hr
.20
.18
.22
.18
.20
.21
.24
.19
.19
.19
.28
.28
.20
.20
.32
.26
48 hr
.20
.18
.22
.18
.20
.21
.24
.19
.19
.19
.27
.22
.20
.19
.27
.24
96 hr
.20
.18
.22
.18
.19
.21
.22
.18
.19
.19
.24
.20
.20
.19
.27
.23
Threshold
.20
.18
.22
.18
.19
.21
.22
..18
.19
.19
.24
.20
.20
.19
.27
.23
(days)
(4)
(4)
(4)
(4)
(5)
(5)
(5)
(5)
(4)
(4)
(6)
(5)
(5)
(4)
(5)
(6)
-------
Table 4 (continued). ACUTE TOXICITY AND MEAN TEST CONDITIONS IN B10ASSAYS TO DETERMINE VARIABILITY
IN RESPONSE OF FATHEAD MINNOWS AND GOLDFISH TO HEXAVALENT CHROMIUM
Tairmoi-o*- iifo nH tt^t" AT"
Test
G1U
GIL
G2U
G2L
G3U
,- G3L
G4U
G4L
G5U
G5L
G6U
G6L
G7U
G7L
G8U
G8L
C reading
24.7
24.0
25.8
24.2
25.5
24.6
25.9
25.1
26.1
25.2
25.1
24.6
24.7
23.9
24.6
24.1
-' Fatheads - total
Table
Test
7.64
7.65
7.67
7.66
7.65
7.65
7.67
7.66
7.49
7.47
7.60
7.62
7.76
7.77
.7.67
7.68
length;
0,, Weight,
mg/1 .v g
5.75
6.11
6.63
6.46
5.96
6.11
5.93
6.06
5.82
5.96
6.19
6.40
6.24
6.21
5.86
6.55
goldfish -
Length ,-
24 hr
LC50,
48 hr .
nat/1
96 hr
11 days
Goldfish
2.36
2.17
1.64
1.67
1.61
1.49
1.67
1.66
V.85
1.87
1.79
1.61
2.13
2.23
2.30
2.20
standard
37.8
36.3
31.3
30.4
30.3
29.2
29.7
29.7
33.7
34.9
34.4
33.5
36.3
37.5
37.1
36.8
length.
-
-
259
256
280
-
-
-
256
-
250
-
244
296
245
266
238
202
158
231
236
224
204
-"
206
243
208
202
182
205
230
212
123
123
90
125
109
135
110
129
98
133
102
. 133
126
126
133
126
4. ACUTE TOXICITY AND MEAN TEST CONDITIONS IN BIOASSAYS TO DETERMINE VARIABILITY
OF FATHEAD -MINNOWS AND GOLDFISH TO HEXAVALENT CHROMIUM
Temperature ,
C
pH meter
reading
°2'
SIR/I
Weight ,
R
Length ,-'
mm
24 hr
LC50,
48 hr
mg/1
96 hr
29
24
24
28
36
59
22
19
45
43
15
18
38
47
43
32
IN RESPONSE
11 days
Fathead Minnows
F1U
F1L
F2U
F2L
F3U
H F3L
o<
F4U
F4L
F5U
F5L
F6U
F6L
F7U
F7L
F8U .
F8L
24.5
23.6
24.7
23.9
26.0
24.9
24.8
24.1
24.8
24.6
25.8
25.1
25.1
24.8
24.7
23.9
7.65
7.66
7.65
7.67
7.66
7.65
7.58
7.57
7.71
7.70
7.50
7.52
7.62
7.63
7.79
7.78
6.26
6.73
6.62
6.82
6.05
6.51
6.48
6.99
6.31
6.24
6.27
6.49
6.68
6.79
6.63
6.56
0.24
0.24
0.20
0.20
0.16
0.16
0.30
0.31
0.29
0.29
0.18
0.18
0.21
0.21
0.37
0.37
26.9
27.4
24.4
23.9
22.8
22.7
26.9
27.9
27.6
27.3
24.0
24.4
23.1
23.9
29.6
29.6
112
134
160
130
160
180
163
111
108
113
164
182
98
162
110
154
83
95
82
70
92
89
84
79
63
83
144
114
52
61
57
56
56
51
53
49
48
60
50
53
49
37
66
55
38
34
29
34
24
18
17
13
14
12
24
18
21
16
20
22
15
16
19
14
-------
Table 5 (continued). ACUTE TOXICITY'AND MEAN TEST CONDITIONS IN BIOASSAYS TO DETERMINE VARIABILITY
IN RESPONSE OF FATHEAD MINNOWS AND GOLDFISH TO GUTHION*
Temperature
Test C
G1U
GIL
G2U
G2L
G3U
G3L
G4U
G4L
G5U
G5L
G6U
G6L .
G7U
G7L
G8U
G8L
24.8
24.4
25.1
25.0
24.7
24.8
24.8
24.4
25.0
24.7
24.9
24.7
24.8
24.6
24.6
24.4
, pH meter
reading
7.20
7.18
7.31
7.31
7.34
7.3?
7.12
7.09
7.12
7.11
7.21
7.20
7.07
7.13
. 7.18
7.16
02, Weight,
mg/l 8
6.42
6.11
6.15
6.25
6.60
5.00
5.89
5.12
6.99
7.27
6.14
. 5.62
5.65 '
6.62
5.66
6.19 .
2.37
2.51
1.75
1.69
1.87
1.87
1.78
1.75
1.37
1.64
1.72
1.66
1.81
1.80
3.50
3.53
Length ,-'
mm '
Goldfish
33. ,9
34.1
30.3
29.6
32.3
32.3
30.7
30.3
32.6
33.6
35.0
35.3
37.9
36.0
42.4
41.5
24 hr
6.87
8.50
6.98
8.87
5.34
5.35
7.63
11.20
6.81
6.05
-
-•
10.60
7.18
9.70
-
LC50,
48 hr
5.18
4.61
4.96
6.17
4.20
2.80
5.01
4.03
6.81
4.26
6.89
8.75
5.58 .
3.69
8.36
5.37
mg/1
96 hr
2.23
2.18
2.68
2.45
2.48
1.71
2.07
2.05
2.08
2.13
3.86
1.88
3.02
2.05
1.35
3.76
11 days
1.22
1.06
1.13
1.05
1;05
0.47
0.71
0.94
0.73
0.71
0.83
0.75
0.19
0.43
0.90
0.71
— Fatheads - total length; goldfish - standard length.
Table 5. ACUTE TOXICITY AND MEAN TEST CONDITIONS IN BIOASSAYS TO DETERMINE VARIABILITY IN RESPONSE
OF FATHEAD MINNOWS AND GOLDFISH TO GUTHION®
Temperature, pH meter
Test C reading
-------
Table 7. TEST CONDITIONS IS BIOASSAYS WITH FATHEAD MINNOWS OF VARIOUS AGES
Test
Age, Fish weight, g
wk
Fish length,
Mean
SO
Mean
SD
Temperature, C
Mean
pH meter reading
Mean
Oxygen, ag/1
Mean Range
F4A
F5A
F6A
F7A
F4B
F5B
F6B
F7B
F4C
F5C
F6C
F7C
F4D
F5D
F6D
F7D
4
4
4
4
7
7
7
7
11
11
11
11
14
14
14
14
.02
.03
.01
.02
.04
.10
.05
.03
.15
.24
.16
.21
.21
.29
.20
.32
.007
.014
.011
.008
.025 ,
.036
.034
.011
.065
.099
.072
.090
.047
.123
.093
.117
13.2
14.3
12.1
12.8
17.6
20.8
17. 1'
15,0
24.7
27.4
25.1
25.8
27.8
29.6
26.6
30.9
1.59
!1.92
2.55
1.94
1.99
2.55
2.88
1.48
'3.38
3.49
4.01
3.63
2.16
4.13
4.15
3.61
25.3
24.6
24.2
25.2
23.7
24.7
25.2
24.6
24.9
23.7
25.5
25.2
25.0
24.5
25.3
25.2
25.0-25.7
24.3-25.1
23.5-24.6
24.5-26.0
23.3-24.3
24.5-25.0
24.5-25.7
23.8-25.3
24.6-25.2
23.1-24.3
25.0-25.7
24.3-25.5
24.3-25.4
23.9-25.1
24.5-25.6
24.5-26.0
7.72
7.81
7.84
7.72
7.71
7.86
7.88
7.77
7.80
7.83
7.84
7.76
7.70
7.82
7.82
7.72
7.67-7.86
7.71-7.96
7.84-8.03
7.68-7.74
7.65-7.87
7.80-7.98
7.82-8.02
7.69-7.97
7.74-7.88
7.72-8.03
7.67-8.02
7.70-7.93
7.63-7.83
7.72-8.01
7.65-8.07
7.68-7.74
6.6
6.7
6.7
6.7
6.7
6.9
6.7'
6.8
6.8
6.8
6.5
6.6
6.5
6.7
6.5
6.7
6.4-7.0
6.5-7.0
6.6-6.9
6.6-7.0
6.5-7.1
6.8-7.1
6.1-7.0
6.6-7.1
6.6-7.0
6.7-7.0
6.3-6.7
6.4-6.9
6.3-7.0
6.4-7.1
6.0-7.0
6.6-7.0
a/
-Total length.
Table 6. TEST CONDITIONS IN BIOASSAYS WITH LARGE AND SMALL FATHEAD MINNOWS OF CONSTANT AGE
Fish weight, g
Test
F1A
FIB
F2A
F2B
F3A
FSB
X
.09
.27
.11
.28
.05
.26
s
.037
.094
.038
.094
.019
.090
Fish length, mm-
X
16.6
23.8
It s
23.5
16.8
28.8
s
2.37
2.61
i IV-
2.j,
1.71
2.97
Temperature, C
X
24.8
24,8
25.0
24.2
24.9
24.6
Range
23.5-25.6
24.0-25.9
23.9-25.5
23.6-24.5
24.0-25.5
24.0-25.6
pH meter reading
X
7.83
7.81
7.70
7.68
7.74
7.75
Range
7.67-7.99
7.64-7.98
7.65-7.85
7.64-7.71
7.64-7.92
7.69-7.91
Oxygen,
X
6.7
6.8
7.0
7.1
6.7
6.7
mg/1
Range
6.
6.
6.
6.
6.
6.
5-7.1
6-7.1
8-7.3
8-7.3
4-7.2
5-7.0
- Total length.
-------
'£.
en
rt
3
D.
(D
ri
a.
S1
9
00
rt
O
W
In
*°
OD
LO
GO
^
W
u>
w
t-1
0
in
to
vo
1
to
ro
U1
?
^j
\o
ON
O
ON
ON
2
H*
-j
O
^
ro
0*
*•
4>
to
in
0
ro
in
ro
00
O">
to
•vj
O
to
1
-J
OO
.p.
00
0
o^
b
o
to
60
r-
*j
in
^
*-
to
U)
^
w
In
ro
I
ro
in
o^
— j
o
1
00
1/1
<*
ro
0
Oh
O
ro
o\
ro
^
in
in
W
in
OD
10
O
ro
in
to
1
ro
in
"^
S
?
Oo
O
in
^i
1
Oh
P
CD
ro
o»
00
ON
in
to
_^
lo
00
w
o
ro
in
to
1
ro
in
00
§
1
00
§
0>
0
T
Oh
o
£»
I--
u>
10
00
U)
to
u>
to
in
to
l
ro
a*
r-1
51
?
•^1
OO
in
in
^
T
in
S
a
•"*
X 1
ta
XI
00
X 1
0
(D
XI
1
n
§
JO
*«J
H-
0)
D-
c
H-
g.
rt
R
H-
00
8
OQ
rt
v
1
H
n
rt
rt
C
0
0
PC
g
rt
1C
U
a.
9
R
o
1
§
B
a
tt>
C
M
E
!
the desired concentration was immediately attained and results compared
to the previous series. Goldfish were acclimated for at least 10 days
at 21 C and for 3 days at the 25 C test temperature. Fathead minnows
were reared at a constant 25 C. Photoperiod during the test consised
of 12 hr of light and 12 hr of dark. Fish in all tests were fed a
small amount of Glencoe pelleted trout food each day, and once per day
debris was siphoned from the aquaria.
Deaths were recorded as each fish died during the first 8 to 12 hours
and at 24-hour intervals thereafter. Tests were conducted for 11 days
or until no additional mortality occurred during a 48-hour period,
whichever came first. Lethal concentration to 50% (LC50) was deter-
mined graphically on semi-logarithmic paper (APHA, 1971), and the LC50
computed at the start of a 2-day period without mortality was used as
the criterion for achieving a threshold LC50. With hexavalent chromium
(R)
and Guthioir-^a threshold LC50 was not usually attained by 11 days, but
technical limitations made it necessary to terminate tests at that time.
Effect of Size and Age—Three series of experiments were conducted:
1) effect of size with a constant stock of goldfish (Table 8); 2) effect
of size with constant aged fathead minnows (Table 6) ; and 3) effect of
different aged fathead minnows (Table 7). A single stock of fish was
used for all goldfish experiments. For each individual test 160 fish
were randomly selected from the stock, and these were divided visually
into groups of 80 large and 80 small fish. The two groups were then
assigned to bioassay aquaria in a stratified random manner with 10 fish
per aquarium. In the series of experiments for determination of size
effect with fathead minnows, constant aged (11-week) fish were selected
from the culture unit on three occasions. Each group was separated into
large and small fish and then tested in the same manner as the goldfish.
In the four experiments for the determination of the effect of age, fat-
head minnows of 4, 7, 11, and 14 weeks were removed from the culture
unit; very small, large or malformed individuals were discarded, and
each group was sorted into the four sets of test chambers in a stratified
23
-------
random manner with 10 fish per chamber. The bioassay procedures and
computations of LCSO's were the same as described above except that
logarithmic-probability graph paper was used for graphical determination
of LC50's.
Chemical Analysis
In the series of tests for determination of bioassay variability, con-
centrations of toxicant in each test chamber were analyzed at 2- or 3-
day intervals. In addition, composite samples of pentachlorophenol were
accumulated twice a day for the last five series of tests and analyzed
at the end. A comparative analysis of fresh and stored samples indi-
cated no deterioration of toxicant. In the series of tests for deter-
mination of effect of size or age, only composited samples of penta-
chlorophenol were anlayzed at the end of each test.
Haxavalent chromium was analyzed colorimetrically [method 211(II)D]
without filtration since a filtered and unfiltered analysis gave the
same result. Chloride was analyzed by mercuric nitrate titration
(method 112B) and results were reported as sodium chloride (AFHA, 1971).
In the first three series of tests for determination of bioassay varia-
bility, pentachlorophenol was analyzed colorimetrically after addition
of Safranin-O and extraction into chloroform (Haskins, 1951). A stan-
dard curve was prepared by analysis of known quantities of reagent grade
pentachlorophenol dissolved in water and extracted and analyzed in the
same manner as the unknown samples. In the last five series of bio-
assays for determination of variability and in all tests for determina-
tion of effect of size and age, pentachlorophenol was extracted into
benzene, acetylated and analyzed on a Beckman GC 72-5 gas chromatograph
with a 122 cm glass column packed with 5% DC 200 on 80/100 mesh Gas
Chrom-Q and a nonradioactive electron capture detector (Rudling, 1970).
Comparison of the colorimetric and gas chromatographic methods indicated
no difference. Standard curves were prepared by solution of reagent
grade pentachlorophenol into benzene, acetylation and measurement of
24
peak height on the gas chromatograph. Due to non-linearity of the
standaiu curve, -ar.kr.cwtis vere diluted so that all samples fell in the
narrow range of 0.05 to 0.15 mg/liter pentachlorophenol. A control
water sample spiked with a known quantity of reagent grade pentachloro-
phenol was analyzed with each group of unknowns. The mean recovery
efficiency of the spike for the series of tests determining variability
was 93.7% (standard deviation 4.73), and for the tests of size or age
effect, mean recovery efficiency was 86.9% (standard deviation 7.13).
Unknowns were corrected by the mean recovery efficiency for each bio-
assay.
dD
After extraction into benzene, Guthioir^was analyzed with the same gas
chromatograph parameters as pentachierophenol. Again, due to non-
linearity of the standard curve"; samples from bioassay chambers were
diluted so thait the finar~concentraTlon In ben2ene was 1 To 3 mg/liter.
(R)
Standards were prepared by solution of the 93% technical grade GuthioriEy
into benzene in that range of concentrations. Since control water was
contaminated, a delonized water sample was spiked with a known concen-
-------
experimental design permitted a determination of the coefficient of
variation (CV) for the 24-hour, 96-hour, and threshold or 11-day LCSO's
for each toxicant with each species. Since each individual bioassay
was conducted in duplicate, multiway analyses of variance of the LCSO's
were also performed to determine significant differences between dif-
ferent stocks of fish and between the two sets of bioassay chambers in
which the duplicates were conducted. To estimate the probability of an
uncontrolled variable affecting the LCSO's with different fish stocks, a
modification of the standard multiway analysis of variance was used
where error due to the slope of individual treatments against their '
deviation from the mean response was partitioned (Handel, 1961).
Since eight analyses were performed at each time interval, a rigid per
comparison error rate of 0.0064 was computed by the formula:
a - 1 - (1-P)
1/n
where a = the probability of making a Type I error
P = probability of occurrence of an error
n • number of analyses
This reduced the possibility of making at least one Type I error by
chance in rejection of the null hypothesis (no significant differences
between stocks, duplicate tests, or slopes). The per comparison error
rate of 0.0064 was equivalent to a per experiment error rate of 0.05
with the experiment consisting of eight analyses of variance at each
time interval (two species, four toxicants).
Determination of Effect of Size and Age—Toxicity curves (LC50 versus
time) were drawn for comparison of the sensitivity of small versus large
goldfish and fathead minnows and for various aged fathead minnows. The
26
curves were drawn from the combined data of all bioassays with each
species within a size or age group. These hyperbolic curves were trans-
formed to straight lines by plotting the reciprocal of the LC50 versus
the reciprocal of time to permit statistical comparison. Analysis of
covariance or t-tests were then calculated to examine differences in
slope and elevation of the straight lines.
GOLDFISH CULTURE STUDY
Apparatus
During most of the experimental period the 16 test chambers used were
constructed of 19 mm thick exterior plywood coated on the inside with
Sears polyester finishing resin. Eight of the tanks measured 61 x 46
x 46 cm and eight others measured 122 x 46 x 46 cm. A polyvinyl chloride
standpipe at one end of each tank maintained~the water depth at 34 cm
so that the small tanks contained 95 and the large tanks 190 liters.
Each tank contained an air diffusion stone, use of which was initiated
2 months after the start of the experiment. The tanks were spacially
arranged in a stratified random manner to accomodate the 23 factorial
design of the experiment (Figure 1).
In the initial stages of the experiment eggs were hatched in a two-
chanbered glass container. The incoming water flowed into a 20 x 20 x
10 cm deep chamber and then through a perforated glass tube into a
10 x 10 x 5 cm deep compartment. The eggs were not moved by the flow
of water and movement of dye indicated that water flowed evenly over
the eggs. Hatched fry were caught in a 7.5 x 7.5 x 10 cm deep basket
as they came through the overflow tube. This basket had three glass
sides and bottom and one Nitex (nylon) screened side and was immersed
in a 22-liter glass aquarium measuring 50 x 25 x 26 cm deep. The same
well water as used in flow-through acute bioassays (Table 1) was sup-
plied to the to the test chambers via two modified proportional diluters
constructed of glass, G.E. Silicone Seal, and Koroseal tubing (Mount and
Brungs, 1967). Each diluter consisted of only the upper distribution
system for delivery of fresh water. The eight chambers in this section
27
-------
3.66 M
alternated in size with four dispensing 500 ml to the smaller tanks and
four delivering 1000 ml to the large tanks. Water from a ninth chamber,
normally used for delivery of control water in toxicant dispensing
systems, ran to the drain since maintenance of a precise volume was
difficult. With a cycle time of 2 minutes 10 seconds, 90% of the water
in both small and large tanks was replaced in approximately 15 hours;
Water was supplied to each diluter from a separate head tank containing
a stainless steel hot water heat exchange system controlled by a
thermostat. The pH was effectively controlled by a chemical metering
device (Mount and Brungs, 1967) dispensing sulfuric acid into the head
tank along with the incoming water. ' .
Six fluorescent light fixtures, each holding two 40-watt cooj-white
fluorescent tubes, were suspended 55 cm above the water, surface and
were arranged so that all tanks received approximately equal illumina-
tion. Photoperiod throughout the experiment was 16 hours light and 8
hours dark.
Experimental Design
The experiment was arranged in a 7? factorial design with two replica-
tions (Figure 1). The three treatments were temperature, nominally 20
and 25 C; fish density, same number in large and small tanks; and food,
Oregon moist pelleted trout food and Oregon moist supplemented with a
variety of other foods. One-way analysis of variance was used for de-
termining significant differences due to each factor as well as the two
and three factor interactions. Since one tank of fish was lost due to
an accident, the average for the replicate tank was used to supply the
missing data, and one degree of freedom was subtracted from the error
term.
Figure 1. Spatial arrangement of test chambers drawn to scale in gold-
fish culture study. Conditions in each chamber are detailed
in Table 19.
23
The effects of the three factors on survival, growth, and maturation
were measured. If maturation occurred, then observations would have
been made on the presence or absence of spawning, fecundity, and time
till initiation of spawning.
29
\
-------
Procedure
A group of 4-year-old goldfish from Ozark Fisheries, Inc. that had been
held in our laboratory at approximately 11 C for about 18 months was used
as brood stock for obtaining eggs. These fish had been induced to spawn
many times by raising the water temperature to 20 C over a period of 2
days. On the third day a portion of the stock would spawn. With this
procedure eggs from three ripe females were hand stripped into three
Petri dishes; milt froa excised testes of three ripe males was squeezed
through cheesecloth onto the eggs; and water at 22 C was added and the
mixture swirled for 2 minutes to complete fertilization. From each
spawning pair eggs were randomly sorted into two groups of 200 eggs.
One group from a spawning pair was placed in an-egg hatching chamber at
20 C and one at 25 C after the Petri dish was floated on the surface of
the appropriate water until temperature equilibrated.
Six days after the start of the experiment all healthy fry from each
temperature were consolidated into two 22-liter aquaria at the respec-
tive temperature. Thirteen days later fish within each temperature
group were assigned in a stratified random manner to each of the wooden
tanks, 18 per tank, after removal of the largest, smallest, and obviously
deformed fish. After 42 days of the experiment fish were thinned to 12
per tank and after 122 days to 10 per tank with prior removal of largest,
smallest, and unhealthy or deformed fish.
After 2-1/2 months and at monthly intervals thereafter fish were not fed
for 24 hours, and on the next day fish from each tank were weighed as a
group in water on a triple beam balance. At termination after 393 days
weights and standard lengths of individual fish were taken, fish were
sexed, and a judgment was made on the state of maturation of the gonads
using a Roman numeral classification system.
Food
The quantity and quality of the food was changed at various times de-
pending on the size of the fish and the availability of certain foods.
30
At all times fish were fed in excess with equal amounts given to each
tank. For the first 43 days all fish received the same diet, and for
the remainder of the experiment one group received only Oregon moist
pelleted trout food and the other group received proportionally less
Oregon moist supplemented by other foods indicated in Table 9_ Other
foods used were Glencoe dry trout pellets, ground for consumption by
fry; pulverized hard-boiled chicken egg yolk; live brine shrimp nauplii
and adults; a mixture of fresh lettuce and fathead minnows ground in a
blender; chopped frozen spinach; live earthworms; and fresh duckweed
(Lemna minor) and various incidental organisms associated with it.
Monitoring Test Conditions
Temperature, pH, and dissolved oxygen concentrations were recoreded once
per week for each tank, and total alkalinity once per week for each
temperature group. Cycle time for each diluter was measured twice each
week with adjustments made as necessary to attempt to maintain a 2-
minute interval.
Temperature in each tank was measured with a YSI Telethermometer and
each diluter was also monitored by a continuously recording Honeywell
24-channel recorder. Total alkalinity, pH, and dissolved oxygen were
analyzed as described in the acute bioassay section.
31
-------
Table 9. DIET
OF TWO GROUPS
OF GOLDFISH DURING
THE EXPERIMENT-'
DIFFERENT PERIODS OF
!
Duration,
days
1-19
20-43
44-161
162-205
206-327
328-391
Daily
schedule
AM-1
AM-2
PM-1
PM-2
AM-1
AM-2
AM-3
PM-1
PM-2
AM-1
PM-1
PM-2
AM-1
AM-2
PM-1
PM-2
AM-1
AM-2
PM-1
PM-2
AM-1
AM-2
PM-1
PM-2
Straight diet
Egg yolk
Glencoe
Glencoe
Brine shrimp
Glencoe
Fish - lettuce
Egg yolk
Fish - lettuce
Brine shrimp-
Oregon moist
Oregon moist
Oregon moist
Oregon moist
—
Oregon moist
Oregon moist
Oregon moist
'
Oregon moist
Oregon moist
Oregon moist
—
Oregon moist
Oregon moist
Mixed diet
Egg yolk
Glencoe
Glencoe
Brine shrimp
Glencoe
Fish - lettuce
Egg yolk
Fish - lettuce
Brine shrimp
Fish - lettuce
Oregon moist
Oregon moist
Spinach
Oregon moist
Oregon moist
Earthworms
Duck weed
Oregon moist
Oregon moist
Duck weed
Spinach
Oregon moist
Oregon moist
Spinach
— Consult text for description of the food indicated.
32
SECTION V
RESULTS
ACUTE TESTS
Acute Toxtcity
Sodium chloride—Mean dissolved oxygen concentrations for all tests were
7.05 and 6.87 mg/liter, mean pH meter readings 7.23 and 7.11, mean tem-
peratures 25.0 and 24.9 C (Table 2), and mean total alkalinities 87 and
83 mg/liter £aCO, (rang.es 80-100 and 73-91) for fathead minnows and
goldfish, respectively.
Most mortality from sodium chloride occurred within the first 48 hr
after which the toxicity curves paralleled the abscissa (Figure 2).
Goldfish were initially more resistant than fathead minnows but after
48 hr became significantly less resistant with a threshold LC50 of
7322 mg/liter for goldfish and 7650 mg/liter for fathead minnows (t -
2.159, p - ,02-.05) (Table 10).
Immediately upon introduction of sodium chloride both species were
affected even at concentrations that caused no mortality during the tests.
For about the first 5 min of exposure there was increased swimming
activity particularly toward the water surface. After this the fish
settled into a pattern of normal activity and food searching near the
aquarium bottom. Within 1 to 2 hr prior to death fish began increased
respiratory movements at the surface accompanied by occasional bursts
of frenzied swimming. Just 'prior to death swimming ceased and opercular
movements slowed considerably. Death was presumed to be due to massive
33
-------
IOOOO
7OOO
0.8
0.9
0.3
O.2
10
o
*>
o
300
200
too
SO
30
a « IP 20 ap
_gOO_
B
• 1111
"S «
To-
"To SO"55 R55 2oo
HOURS
Figure 2. Toxicity curves of (A) sodium chloride, (B) pentachlorophenol,
(C) Guthion, and (D) hexavalent chromium for goldfish (dashed
lines) and fathead minnows (solid lines). At times of 24 hr
or longer curves are visually fitted through mean LCSO's from
at least 14 tests and at earlier times through mean LCSO's
from at least B tests. Hypothesized extensions of the Guthion
curves based on limited data are represented by dots.
34
osmoregulatory failure.
Pentachlorophenol—Mean dissolved oxygens were 6.54 and 5.82 nig/liter,
mean pH meter readings 7.72 and 7.69, mean temperatures 24.8 and 24.7 C
(Table 3), and mean total alkalinities 211 and 211 mg/liter CaC03 (ranges
204-216 and 206-217) for fathead minnows and goldfish, respectively.
Toxicity curves for pentachlorophenol were similar to those for sodium
chloride in that initial mortality was rapid and the curves became
parallel to the abscissa by 48 hr for fathead minnows and by about 96 hr
for goldfish (Figure 2). Goldfish were initially more resistant but
there was no significant difference in the threshold LC50 of 0.21 mg/
liter for both species (t = .117, p >.5) (Table 10).
Fish exposed to pentachlorophenol seemed unaffected until a few hours
before death. They then exhibited increased swimming activity near the
surface, followed by quiescence, then death at the bottom of the test.
chamber. Webb and Brett (1973) list toxic effects of pentachlorophenol
and classify it as a general metabolic stressor, uncoupling oxidation
phosphorylation.
Guthion^-—Mean dissolved oxygen concentrations were 6.55 and 6.09 mg/
liter, mean pH meter readings 7.30 and 7.20, mean temperatures 24.8 and
24.7 C {Table 5), and mean total alkalinities 89 and 87 mg/liter as CaCOj
(ranges 82-93 and 81-92) for fathead minnows 'and goldfish, respectively.
The toxicity curves (Figure 2) for Guthion^'are similar in shape for
both species but differ from the curves for sodium chloride and penta-
chlorophenol as well as typical curves for many other toxicants as re-
ported by Sprague (1969). The curves are nearly parallel to the abscissa
at 6 to 12 hr for fathead minnows and at 18 to 26 hr for goldfish. This
result is presumed to be an artifact resulting from concentrations too
low to-cause measurable mortality in less than 6 hr for fathead minnows
and 18 hr for goldfish. Individual points used to construct these curves
35
-------
Table 10. HEADS AND 95% CONFIDENCE INTERVALS OF LCSO'S (MG/LITER)
OF FOUR TOXICANTS AND TWO FISH SPECIES AT VARIOUS TIMES
NaCl
Hours FH^ GF^-
3
4
5
8
12 8530
+663
24 7910
+330
48 7691
+257
72 7650
+234
96 7650
+234
120 7650
+234
144 —
168
—
—
—
•
9952
+365
7706
+308
7388
+236
7341
+234
7322
+224
7322
+224
__
PCP
FH
.70
+.07
.55
+ .04
.38
+.03
.25
+.03
.22
+.02
.21
+.02
.21
+.02
.21
+ .01
.21
+.01
— —
GF
—
—
.69
+.08
.46
+.05
.33
+.03
.27
+.03
.24
+.03
.24
+.03
.22
+.02
.21
+.02
.21
+.02
_—
Guthion"*
FH
—
—
—
7.43
+2.94
6.82
+1.96
5.94
+1.87
3.63
+ .90
2.31
+.50
1.90
+ .40
1.51
+.33
1.25
+.30
1.06
+.25
GF
—
—
—
—
—
7.78
+1.14
5.42
+.87
3.65
+.42
2.37
+.36
1.99
+.32
1.44
+.23
1.25
+.23
Cr"
FH
—
—
—
257
+40
195
+14
140
+15
82
+13
56
+8
48
+6
43
+5
38
+5
32
+3
K>
GF
—
—
—
—
261
+13
212
+13
' 167
+15
120
+8
83
+10
65
+12
53
+9
36
Table 10 (continued). MEANS AND 95% CONFIDENCE INTERVALS OF LCSO'S
(MG/LITER) OF FOUR TOXICANTS AND TWO FISH SPECIES AT VARIOUS TIMES
NaCl PCP Guthion®.
Hours FH GF FH
192 — —
216
240 — —
264 — - — —
•
GF FH
.95
+.23
.81
+.21
.78
+.22
.76
+.22
GF
1.15
+ .24
1.02
+.18
.92
+.17
.80
+.15
Cr*6
FH
27
+2
23
+2
20
+2
18
+2
GF
45
+9
39
+7
34
+7
33
+7
^Fathead minnow.
-'Goldfish.
were the means of at least 14 LC50 values at times of 24 hr or longer
and means of at least 8 LC50 values at times earlier than 24 hr. In two
tests with goldfish where test concentrations were set somewhat higher
and early mortality occurred, the mean 8-hr LC50 was 17 ng/liter and the
mean 4-hr LC50 from two fathead minnow tests under similar conditions
was 8.5 mg/liter. As indicated by the dotted line (Figure 2), it is
hypothesized that these curves bend steeply upward at shorter time inter-
vals. The resultant change in slope throughout the curves may indicate
different mechanisms of toxic action or at least a period of physiologi-
cal adjustment before further mortality occurs. The toxic effect of
(B
Guthion** and its oxidized metabolite is inhibition of acetylcholines-
terase (Chemagro Division Research Staff, 1974).
As with sodium chloride and pentachlorophenol, goldfish were initially
more resistant, but at 11 days the LC50's were not significantly dif-
37
-------
ferent with 0.76 nig/liter for fathead minnows and 0.80 mg/liter for
goldfish (t - .381, p>.5) (Table 10). At 11 days threshold LC50 values
had not been attained for goldfish, but the toxicity curve for fathead
minnows was just starting to parallel the abscissa (Figure 2).
Early mortality of fathead minnows was unusual in that deaths were
poorly related to toxicant concentration. In typical tests the follow-
ing percentage mortalities had occurred after 1 hr:
Concentration (og/l): 0 0.5 0.9 1.6 2.5 4.8 6.8 10.0
Percentage mortality: 0 20 0 0 20 40 90 40
Upon introduction of the Guthion*', fish in all concentrations would
appear stressed within 5 min, many lost equilibrium, and some died as
indicated by the results above. After the initial mortality the death
rate decreased for up to 24 hr and then again increased (Figure 2).
Fish that lost equilibrium would either die almost immedaltely or would
lie on the bottom of the aquarium in this condition for a few days be-
fore death occurred. This loss of equilibrium in the environment would
be essentially equivalent to death since the individual is immediately
vulnerable to predators and is at the mercy of water currents. There-
fore, LC50 is not a good predictor of environmental damage. ECSO's
based on loss of equilibrium would be lower than LCSO's at the same
time intervals.
Hexavalent Chromium—Mean dissolved oxygen concentrations from all tests
were 6.53 and 6.14 ing/liter, mean pH meter readings 7.65 and 7.64,
mean temperature 24.7 and 24.9 C (Table 4), and mean total alkalinities
214 and 214 mg/liter CaCOj (ranges 200-230 and 200-222) for fathead
minnows and goldfish, respectively.
The toxicity curves for hexavalent chromium are somewhat unusual when
compared to typical hyperbolic curves of other toxicants (Figure 2).
The shape of the curve for goldfish is similar to the Guthion curves
and limited data obtained for earlier times suggest that it bends
sharply upward at the shorter time intervals. The curve for fathead
38
minnows indicates an initial period of rapid mortality, followed by a
period of reduced mortality from 72 to 120 hr, and then again increased
mortality. This curve suggests the possibility of two modes of toxic
action, one at high and one at lower chromium concentrations. Grindley
(1946) reported a similar curve showing the effect of potassium dichro-
raate on rainbow trout. However, this test was conducted at considerably
higher concentrations of chromium. In a series of studies on the toxic
effect of hexavalent chromium it was found that the primary path of
entry was by diffusion through the gills of rainbow trout (Knoll and
Fronm, 1960), that no gill pathology occurred but the epithelial cell
lining of the Intestinal tract of largeoouth bass was sloughed off
(Fromm and Schiffman, 1958), and that glucose uptake by epithelial cells
of the intestinal tract was inhibited (Stokes and Fromm, 1965). These
authors did .not suggest more than one mode of toxic action. Since glu-
cose inhibition was found in live fish exposed to 2.5 mg/liter hexa-
valent chromium for 7 days, this may be the mode of toxic action at
times longer than 72 hr. The cause of death at earlier times is unknown*.
if another mode of toxic action exists.
With the other three toxicants the goldfish were initially more resis-
tant, but by the end of the test, LCSO's were the same for both species
(K]
(pentachlorophenol and Guthion—9 or less for goldfish (sodium chloride).
With hexavalent chromium the goldfish were more resistant throughout
the entire test. The 11-day LC50 of 33 mg/liter for goldfish was sig-
nificantly greater than that of 18 mg/liter for fathead minnows (t =
4.601, p - .01-.02) (Table 10). Unless these species differ considerably
in their resistance to chromium as compared to the other toxicants, it
would be expected that after 11 days the chromium toxicity curve for
goldfish would bend downward resulting in a curve similar to that for
fathead minnows.
• Since initial chromium concentrations were gradually increased to the
desired concentrations in most tests, three tests on each species were
conducted to compare the difference in LC50's when test concentrations
were attained immediately. The toxicity curves for both species were
39
-------
similar in shape to the curves obtained by the first experimental pro-
cedure (Figures 2 and 3). The LCSO's for fathead minnows were lower at
all times, but the LCSO's for goldfish were higher after 24 hr (Tables
10 and 11). The difference in procedure affected-the overall acute mor-
tality of fathead minnows, and the 11-day LC50 of 18 mg/liter by the
first procedure was significantly higher than the 11-day LC50 of 12
mg/liter (t = 2.482, p = .02-.05) by the second procedure. There was
no significant difference in overall toxicity to goldfish using the two
procedures with 11-day LCSO's of 0.33 and 0.38 mg/liter (t - .722, p
The onset of detrimental effects was usually observable from 24 to 48 hr
before death. Respiratory movements became more rapid and pronounced
and fish surfaced frequently. Prior to death respiratory movements
slowed, fish lost equilbrium, and they died shortly thereafter. Within
24 hr after the introduction of even the lowest chromium concentrations
tested, the bottom of the goldfish chambers became littered with tape-
worms .
Variability of Test Species
In the previous section mortality of the fish due to the four toxicants
was discussed primarily in terms of toxicity curves depicting death rates
during the entire test. Standard bioassays for comparative purposes by
industrial or research laboratories will probably continue to report
results as LCSO's at fixed times, most likely 24 hours as with the
ORSANCO (1974) bioassay or 96 hours as with the proposed Canadian pro-
cedure (MacLeod, 1972). Reporting of threshold LCSO's has become more
common where these values are attained in a relatively short time.
Therefore, the comparison of variability described here is in terms of
24- and 96-hr LCSO's for all toxicants as well as threshold LCSO's for
sodium chloride and pentachlorophenol and 11-day LCSO's for hexavalent
chromium and Guthion with which threshold values were not attained by
termination of the bioassay. Table 12 lists the mean LC50 values for
200
100
O 50
in
O
20
30 60 100
HOURS
200
Figure 3. Toxicity curves of hexavalent chromium for goldfish (dashed
line) and fathead minnow (solid line). Curves are visually
fitted through mean LCSO's from three tests using the second
experimental procedure.
40
41
\
-------
Table 11. MEANS AND 95Z CONFIDENCE INTERVALS OF LCSO'S OF
HEXAVALENT CHROMIUM FROM RESULTS OF THE SECOND EXPERIMENTAL PROCEDURE
Hours
LC50. mg/1 Cr
Fathead
Goldfish
24
48
72
96
120
144
168
192
216
240
264
66 +18
50 +12
33 +21
26 +11
26 +11
24 +12
20 +12
18 +12
14 +13
13 +12
12 + 8
245 +62
213 +49
188 +57
124 +33
101 +35
76 +69
60 +18
53 +17
45 + 5
40 +18
38 +17
42
Table 12. MEAN LCSO'S (MG/LITER) WITH COEFFICIENT OF VARIABILITY
(PERCENTAGE) OF FOUR TOXICANTS AT VARIOUS TIMES
Toxicant
NaCl
PCP
Cr*6
Guthion
NaCl
PCP
Cr*6
Guthion
24
LC50
7910
0.22
140
5.94
9952
0.27
261+
7.78
hr
CV
8
19
20
58
7
23
24
Time
96 h
LC50
Fathead Minnow
7650
0.22
48
1.90
Goldfish
- 7341
0.22
120
2.37
r
CV
6
12
22
39
*
17
12
29
Terminal-'
LC50
7650
0.21
18
0.76
-
-7322
0.21
33
0.80
CV
6
12
21
52
6"
16
38
35
^Threshold LC50 for NaCl and PCP, 11-day LC50 for Cr*6 and Guthion.
— 24-hr LC50 not obtained in all tests due to slow mortality.
43
-------
the four tested materials at these times along with the coefficient of
variation. Since the LCSO's of the toxicants differ by up to four orders
of magnitude, easy comparison of variability by standard deviation is
impossible without normalization to the coefficient of variation.
The LC50 values of sodium chloride were the least variable of the four
toxicants at all times, with essentially no difference between fathead
minnows and goldfish. Fentachlorophenol ranked second although with
this toxicant goldfish were slightly more variable. Hexavalent chromium
and Guthion** ranked third and fourth, respectively, although missing
data and some variation at different times make this ranking less defini-
tive. With sodium chloride and pentachlorophenol there was little
difference in variability between 96-hr and threshold LCSO's, but varia-
bility at 24 hours was greater than at the other times. With hexavalent
chromium there was essentially no difference in variability between 24-,
96-hr, and 11-day LCSO's with fathead minnows but with goldfish the
variability of the 96-hr LC50 was considerably less than at 11 days, and
because initial mortality was slow too few 24-hr LCSO's were determined
for computation of a reliable coefficient. The fathead minnows had mini-
mum variability to Guthioir* at 96 hours and similar variability at 1
and 11 days, whereas the goldfish showed increasing variability at 1, 4,
and 11 days. In general, the 96-hr measurement was the least variable,
but where threshold values were attained, they showed similar variability.
Neither fish species showed consistent superiority in minimal variability.
Both were the same for sodium chloride. The fathead minnows were slightly
less variable with pentachlorophenol and considerably more variable with
Guthioii? With chromium the goldfish were less variable at 96 hours, more
variable at 11 days, and no comparison was available at 24 hours.
The analyses of vari^'ce indicated only one instance of a significant
difference between the i iplicate bioassay apparatus used for testing
each stock. That differeri •*. occurred with goldfish at 96 hours testing
hexavalent chromium (Table 1- . Since there was also a significant
difference in slope, it appeared that some independent variable was
responsible for the difference between results from the Lwo types of
apparatus. Correlation coefficients were computed for the LCSO's and
all quantifiable independent variables, which included weight, length,
condition factor, age of fish, time of pre-bioassay holding, dissolved
oxygen concentration, pH, and temperature. Only temperature correlated
(r " -.69). An anlysis of covariance with temperature as the covariate
was then computed and the results from the two bioassay apparatus were
still significantly different, indicating that temperature was not the
cause of the difference. Since slope was not significantly different
in any other comparison, there was no inidication of an independent
variable affecting the LCSO's.
Reference Toxicant
The same data used to evaluate the variability of the two fish species
also .indicate the variability of the toxicants. Sodium chloride bio-
assays were considerably less variable at all times (Table 12). There-
fore this compound best meets the criterion of minimum variability in
response of the fish to a reference toxicant. Pentachlorophenol ranked
second in variability, and although the coefficient of variation was
about two or three times greater than for sodium chloride, it is still
acceptable as a reference toxicant. The analytical precision for penta-
chlorophenol was twice as great as for sodium,chloride and may have con-
tributed somewhat to the difference in variability. The coefficients of
(R)
variation for chromium and Guthioir* were considerably higher and varied
so much between species that neither of these is suitable as a reference
toxicant on the basis of this criterion.
The results of the bioassays can be used only indirectly to assess whether
the toxicant would detect abnormal fish by their having a deviant
response. If stocks of fish, particularly goldfish, differed during the
2-year period of testing, the experimental design would permit detection
of these differences in response to the toxicants by the multiway analysis
of variance. Only the sodium chloride bioassays consistently revealed a
45
-------
Table 13. OCCURRENCE OF SIGNIFICANT DIFFERENCES IN MULTIWAY ANALYSES
OF VARIANCE WITH FOUR TOXICANTS^
Toxicant
NaCl
PCP
+6
Cr°
Guthion
NaCl
PCP
Cr*
Guthion
NaCl
PCP
Species
Fathead
Goldfish
Fathead
Goldfish
Fathead
Goldfish
Fathead
Goldfish
Fathead
Goldfish
Fathead
Goldfish
Fathead
Goldfish
Fathead
Goldfish
Fathead
Goldfish
Fathead
Goldfish
Stocks
of fish
24-hr LC50
ns
ns
ns
sig
ns
—
sig
—
96-hr LC50
sig
sig
ns
ns
sig
ns
ns
ns
Terminal LC50
sig
sig
ns
ns
Source of variance
Bioassay
apparatus
ns
ns
ns
ns
ns
—
ns
—
ns
ns
ns
ns
ns
sig
ns
ns
ns
ns
ns
ns
Slope
ns
ns
ns
ns
ns
—
ns
—
ns
ns
ns
ns
ns
sig !
ns
ns
ns
ns
ns
ns
46
Table 13 (continued). OCCURRENCE OF SIGNIFICANT DIFFERENCES
IN MULTIWAY ANALYSES OF VARIANCE WITH FOUR TOXICANTS^
Source of variance
Toxicant
Cr"**
Guthion
Species
Fathead
Goldfish
Fathead
Goldfish
Stocks
of fish
Terminal LC50
ns
ns
ns
ns
Bioassay
apparatus
ns
ns
ns
ns
Slope
ns
ns
ns
ns
- "Sig" indicates a significant difference; "ns" indicates no significant
difference; per experiment p<.05.
significant difference between the eight stocks of fish tested with both
the 96-hr and threshold LCSO's different for both species (Table 12).
Three other indications of significant differences between fish stocks,
goldfish with pentachlorophenol at 24 hours and fathead minnows with
chromium at 96 hours and Guthion^at 24 hours, were not consistent be-
tween species or other test durations. Since the same stock of goldfish
was tested with each toxicant, significant differences might be expected
with all toxicants. However, bioassay results with sodium chloride were
considerably less variable (Table 12) and thus significant differences
between stocks tested with the other toxicants were either masked by the
greater overall variability or did not occur. Therefore, sodium chloride
appears to be the most sensitive indicator of differences between fish
stocks. Whether or not these differences are due to previous history of
the stock, fish condition at the time of testing, or some uncontrolled
variable cannot be evaluated from these data. The mortality of all gold-
fish stocks during the pre-test holding period was similar (1-10%) so
that differences cannot be attributed to this effect.
One attempt was made to determine if a previously unhealthy and stressed
47
-------
stock of fish could be detected by deviant LCSO's when exposed to any of
the four toxicants. A stock of goldfish arrived in extremely poor con-
dition, heavily infested with skin flukes and probably a bacterial
disease. Daily mortality was high for over a month, and during that
period numerous treatments with potassium permanganate, tetracyeline, or
neomycin were administered. After about 50% of the fish were lost, mor-
tality ceased, and the fish were tested in duplicate with each toxicant.
It was hypothesized that these fish would either be more resistant to the
toxicants if the weaker individuals had been eliminated from the popula-
tion, less resistant if the total stress from the diseases and treatments
had increased their sensitivity, or equally resistant if the two factors
balanced out.
With the unhealthy fish the equivalent 96-hr and threshold LCSO's of
sodium chloride were considerably lower than the means from all bio-
assays with the presumed healthy fish and well outside the range of all
of these tests. These results correspond to the analyses of variance
of the 16 regular bioassays where there were significant differences in
96-hr and threshold LC50's but no difference at 24 hours (Table 13).
The reverse trend occurred with pentachlorophenol. Threshold and 96-hr
LCSO's of the unhealthy fish were within the range of the "normal" fish,
but the 24-hr LCSO's were greater and outside the range for normal fish
(Table 14). These results correspond with the analyses of variance for
pentachlorophenol where differences in stocks of goldfish were detected
at 24 hours but not at the other two times (Table 13). The only other
instances where the LCSO's for the unhealthy fish were outside the_
VR}
range for normal fish were the 96-hr and 11-day values for Guthion
(Table 14). The analyses of variance did not detect stock differences
with Guthion^'at these times, but the great overall variability of
Guthion® probably masked any possible differences.
Sodium chloride is most capable of detecting abnormal fish by deviant
LCSO's but bioassays must be conducted for longer than 24 hours. Penta-
chlorophenol is very rapid in detecting abnormal fish, but after 24 hours
48
Table 14. COMPARISON OF LCSO'S FROM UNHEALTHY GOLDFISH STOCK
WITH MEAN LCSO'S FROM ALL TESTS WITH "NORMAL" GOLDFISH
(milligrams/liter)
LC50 .
Stock
"Normal"
Unhealthy
"Normal"
Unhealthy
Mean.
Range
Test 1
Test 2
Mean
Range
Test 1
Test 2
NaCl
24-hr LC50
9952
8350-11050 0.
9270
10270
- 96-hr LC50
7341-
6800-8050 0.
6170
6180
PCP
0.27
18-0.37
0.38
0.44
0.22
17-0.30
0.24
0.30
Cr*
-
-
-
120
90-135
93
106
Guthion*
7.78
5.34-11.20
7.35
6.25
2.37-
1.35-3.86
<0.18
0.17
Terminal LCSO^'
"Normal"
Unhealthy
Mean
Range
Test 1
Test 2
7322
6800-8050 0.
6170
6180
0.21
15-0.28
0.19
0.25
33
15-59.
37
56
0.80
0.19-1.22
<0.18
-•0.17
Threshold LC50 for NaCl and PCP; 11-day LC50 for Cr*6 and Guthion?
the differences disappear. This difference may give an advantage to
pentachlorophenol if reference bioassays are conducted for just 24 hours
JCT\
prior to testing an unknown. Both hexavalent chromium and Guthion6'are
relatively incapable of discovering abnormal fish.
49
\
-------
A reference toxicant should be suitable for use in either flow-through
or static systems depending on the needs of the laboratory, but static
tests will probably be used more frequently by industries. With the
flow-through systems currently in use it is impracitcal to use sodium
chloride as a reference toxicant because the quantity of the salt
required is too great. Pentachlorophenol can be used for tests in
either system. When the same stock of goldfish was tested simultaneously
in duplicate static bioassays and duplicate flow-through systems, the
threshold LCSO's were 1-1/2 to 2 times greater in the static tests.
The cause of this difference was not determined. It was not due to
deterioration or loss of toxicant in the static system since concen-
trations were monitored daily and no change occurred. No static bio-
assays were attempted with hexavalent chromium but others have conducted
apparently successful static tests (Trama and Benoit, 1960). Although
static Guthion*' tests were performed, flow-through tests are possible as
was demonstrated in a chronic test conducted in our laboratory. The
(6)
major difficulty with a flow-through acute bioassay of Guthioir was
solution of the high concentrations needed for acute tests.
Table 15 summarizes the ability of the four toxicants to meet the seven
desired characteristics of a reference toxicant. The first five cri-
teria (top to bottom) have been discussed previously in depth. Ease of
chemical analysis is an important consideration for research labora-
tories, but industrial laboratories will probably use calculated concen-
trations without chemical analysis. The subjective judgment on ease of
handling is based on use of the chemical in a laboratory including such
factors as equipnent cleaning and administration of the material and on
considerations of degradation of waste material in the environment and
toxicity to humans (including carcinogenicity).
Effect of Size and Age
Goldfish—Constant stock—Mean lengths and weights of the three groups of
small fish ranged from 34-36 cm and 1.41-1.70 g and for the large fish
Table 15. COMPARISON OF FOUR TOXICANTS WITH REGARD
TO THEIR USE AS A REFERENCE TOXICANT-'
Consistency of fish response
Detection of abnormal fish
Rapid attainment of
threshold LC50
Use in static bioassay
Use in continuous-flow
bioassay
Ease of chemical analysis
General ease in handling
NaCl
1
2
1
2
4
1
2
PCP
2
2
1
2
1
3
2
Cr4*
3
4
4
2
1
1
3
Guthion^'
4
3
4
2
3
4
4
— 1 - very good; 2 - good; 3 - fair; 4 - poor.
from 43-44 cm and 2/68-3.59 g (Table 16). The regression equations and
the percentage of variability in the reciprocal of LC50 attributable to
2
reciprocal of time of the measurement (coefficient of determination - r )
are presented in Table 17 for the combined data for groups of small and
large fish. An analysis of covariance indicated no significant difference
in slope (F
, ,
J., £
1.63, p>.05) or elevation (F, ,, - 3.68, p>.05),
J., LI
thus both large and small goldfish from the same stock responded simi-
larly to the pentachlorophenol (Figure 4). If the 24-hr, 96-hr, and
threshold LCSO's are compared between the large and small fish (Table
18), the 24- and 96-hr LCSO's are not significantly different (t = 2.49,
p>.05 and t - .390, p > .05) but the threshold LCSO's are different
(t » 5.00, p « .025-. 05). This apparent contradiction of the analysis
of covariance results because it compares variability over the entire
test period whereas the t-test compares the variability at one instant.
Thus, the rate of mortality throughout the entire test was similar for
both groups, but the concentration at which mortality ceased (threshold
LC50) was greater for the small fish indicating that these were ultimately
50
51
\
-------
Table 16. MEAN AND STANDARD DEVIATION OF WEIGHT AND LENGTH
OP FISH IN ALL BIOASSAYS
Table 16 (continued). MEAN AND STANDARD DEVIATION OF WEIGHT
ASD LENGTH OF FISB TM ALL BIOASSAYS
Teat
G1A
GIB
G2A
G2B
G3&
G3B
fflA.
FIB
F2A
E2B
F3A
F3B
F4A
F4B
F4C
F4D
F5A
FSB
F5C
F5D
F6A
F6B
F6C
F6D
Age,
wks
_
_
.
_
_
". - '
11
11
11
11
U
n
4
7
11
14
4
7
11
14
4
7
11
14
Noninal
size
Goldfish -
small
large
small
large
. snail
large
Fathead
small
large
small
large
snail
large
Fathead
-
-
-
-
-
. -
-
-
-
-
-
- .
.Weight
Mean
Constant
1.41
2.68
1.62
3 as
1.70
3.59
r ft
SD
Stock
.336
.652
.455
.751
.426
.838
Length,
Mean
35
43
36
44
34
43
,c^'
SD
3.2
3.0
3-°
3.4
2.5
3.1
- Constant Age
.09
-27
.11
.28
.05
.26
.037
.094
.038
.094
.019
.090
17
24
16
24
17
29
2.4
2.6
2.0
2.6
1.7
3.0
- Various Ages
.02
.04
.15
.21
.03
.10
.24
.29
.01
.OS
.16
.20
.007
.025
.065
.047 .
.014
.036
.099
.123
.011
.034
.072
.093
13
18
25
28
1*
21
27
30
12
17
25
27
1.6
2.0
3.4
2.2
1.9
2.6
3.5
4.1
2.6
2.9
4.0
4.2
Test
F7A
F7B
F7C
F7D
Age,
wks
4
7
11
14
Nominal Weight
size Mean
Fathead - Various
.02
.03
.21
.32
, R
SD
Ages
.008
.011
.090
.117
Length
Mean.
13
15
26
31
^
SD
1.9
1.5
3.6
3.6
a/Goldfish - standard length; fathead minnow - total length.
slightly less sensitive. — - - —.
Fathead Minnows—rConstant age—Mean lengths and weights of the three
groups of small fathead minnows ranged from 16-17 cm and .05-.09' g and-
for the large fish from 24-29 cm and .26-. 28 g (Table 16). The regression
equations for the reciprocal toxicity curves with these two groups re-
sulted In the best fit of any groups tested with coefficients of deter-
mination of .83 and .88 (Table 17). The slopes of these regression lines
were significantly different (F.^ J3 - 7.59, p<.01) indicating a different
rate of mortality over the entire test period (Figure 4). Difference in
elevation was not tested since this is meaningless with different slopes.
There was no significant difference in 24-hr, 96-hr, or threshold LCSO's
(Table 18) between the two size groups (t - 2.00, p ? .05, t - 1.61, p
7-05 and t "= 1.61, p .05, respectively). Although the regression lines
were separating at the time of the 24-hr, 96-hr and threshold measurements
(Figure 4), differences between the two groups were not sufficient to
reveal statistical significance. In summary, the small fathead minnows
died sooner at a given concentration but the rate of mortality once death
began to occur was greater for the large fish. By the time 24-hr and
threshold LCSO's occurred there was no difference between sizes.
52
53
-------
Table 17. REGRESSION EQUATIONS AND COEFFICIENT OF DETERMINATION
(r2) FOR THE RELATIONSHIP BETWEEN LC50 AND TIME
IN THREE GROUPS OF EXPERIMENTS
Experiment
Regression
Equation
r2
Goldfish
Small
Large
1/Y
1/Y
= 4
- 4
.09 -
.82 -
Fathead -
Small
Large
1/Y
1/Y
« 4
* 5
.35 -
.05 -
Fathead -
4 weeks
7 weeks
11 weeks
14 weeks
1/Y
1/Y
1/Y
1/Y
- 4
» 4
- 4
* 4
.77 -
.09 -
.03 -
.87 -
11
15
.57
.05
(1/x)
(1/x)
.82
.78
Constant Age
8.
13
74
.28
(1/x)
(1/x)
.83
.88
Various Ages
11
10
10
14
.44
.15
.30
.80
(1/x)
(1/x)
(1/x)
(1/x)
.80
.76
.71
.74
Different ages—The mean lengths and weights of fathead minnows increased
with age in each experimental group although sizes between different
test groups overlapped (Table 16). Density of fish in the culture unit
aquaria was probably responsible for this effect. The four regression
equations based on concentration against time did not fit the data as
well as those related to size of fathead minnows and goldfish, but the
coefficients of determination were still adequate (Table 17). Analysis
of covariance indicated a significant difference in slopes (F „- =
3.16, p m .025-.05) so each line was compared with all others by t-tests
(Figure 4) . The slope of the line for 14-week-old fish was significantly
different than for 11-, 7-, and 4-week-old fish (t - 2.52, p< .05, t =
2.76, p<.05, and t - 1.98, p " .05, respectively). None of the other
comparisons approached significance.
54
02
ID
o
O.I
HOURS
0.2
-I
0.3
Figure 4. Relationship of reciprocal of LC50 of pentachlorophenol versus
reciprocal of time. A: goldfish, 1) small, 2) large;
B: fathead minnow - constant age, 1) small, 2) large-
C: fathead minnow - different ages, 1) 4 weeks; 2) 7 weeks
3) 11 weeks, 4) 14 weeks.
55
-------
Table 18. MEANS AND STANDARD DEVIATIONS OF LC50*S WITH
PENTACHLOROPHENOL FOR DIFFERENT GROUPS OF FISH AT VARIOUS TIMES
(mg/liter pentachlorophenol)
Experiment
Small
Large
Small
Large
It weeks
7 weeks
11 weeks
14 weeks
24-hr
Mean
.267
.250
.240
.213
.222
.245
.232
.200
LC50
SD '
Goldfish -
.042
.053
Fathead
.036
.025
Fathead
.021
.039
.052
.016
96-hr LC50
Mean SD
Constant Stock
.247 .025
.190 .020
- Constant Age
.227 .029
.203 .012
- Various Ages
.198 .017
.230 .036
.222 .039
.190 .012
Threshold
Mean
.240
.190
.227
.203
.198
.230
.222
..190
LC50
SD
.026
.020
.029
.012
,017
.036
.039
.012
The 24-hr, 96-hr, and threshold LCSO's for the four age groups are pre-
sented in Table 18. Analyses of variance indicated no significant dif-
ferences between different groups of fish. The 4-, 7-, and 11-week-old
fish died sooner at a given high concentration but the rate of mortality
after death began was greater for the 14-week-old fish than the other
three so that at the time of the 24-hr, 96-hr and threshold measurements
there was no difference between ages.
GOLDFISH CULTURE STUDY
Test Conditions
The temperature in each tank was maintained close to the desired tempera-
tures with a range in means of 19.85 to 20.02 C for the nominal 20 C tanks
56
and 24.58 to 25.10 C for the nominal 25 C tanks (Table 19). The ranges
in mean pH meter reading overlapped for the 20 and 25 C tanks with 7.48-
7.58 for the former and 7.55-7.62 to the latter. Mean total alkalinity
was 195.6 (standard deviation 7.2) and 191.-3 mg/liter CaCO, (standard
deviation 5.4) for the 20 and 25 C tanks, respectively. Dissolved
oxygen was more difficult to maintain in the higher temperature tanks
even with the use of air dispersion stones. Mean dissolved oxygen concen-
trations ranged from 5.29 to 6.36 and from 4.92 to'5.70 mg/liter in the
20 and 25 C tanks, respectively (Table 19). After 115 days an accident
occurred which resulted in an increase of the temperature in tank #1 to
lethal levels, thereby eliminating this tank from the experiment.
Eggs, Fry and Mortality
The three groups of eggs held-at 20 C had survival-percentages through -
hatching of 68.1, 30.2, and 34.1 and the three groups at 25 C survived
at percentages of 56.6, 21.0, and 39.0. Total survival was 40% at 20
C and 38% at 25 C indicating no effect of temperature. The relatively-
low survival was probably due to infertility of the artificially fer-
tilized eggs.
After fry were consolidated within a temperature group no attempt was
made to assess mortality until the fish were distributed to individual
tanks. During the 29 days from the assignment of 18 fish tb each tank~
until the first thinning to 12 fish per tank, 81% of the try survived
at 20 C and 89% at 25 C. The fish were ultimately thinned to 1O per
tank and thereafter occasional fish were lost from accidents or unknown
causes. This mortality is reflected in the number of individual fish
specified for each tank in Table 20.
Coloration
After 90 days the proportion of fish that had changed from olive-brown
to a gold or white coloration was recorded. Analysis of variance after
an arcsin transformation of the proportions indicated that temperature
57
-------
IT ^J>
«i I
3-
O
01
?•
f
»
t
vOU>H>£~N>l/iOOsK*ViUJao-C*
sssassss
§M
UJ
I
alone was the only factor showing a significant effect on the rate of
color change (F - 54.12, p<.005). At 20 C 71 of the fish had com-
pleted the color change and 79Z had done so at 25 C. Size of the fish
versus coloration was not recorded, but direct observation indicated
that size was probably not a significant covariate since smaller fish
had changed color at 25 C and most larger fish at 20 C had not. Once
all fish had turned white or gold, the intensity of the color was much
greater in tanks of fish fed the mixed diet, particularly after the
fresh duckweed was included.
Growth
Fish in all tanks gained weight throughout the entire test period (Table
21). Analysis of variance indicated significant differences in final
weights (Table 20) due to all main factors and one two-factor inter-
relation. There was a significant difference in weight due to tempera-
ture (Fx ? - 48.97, p<.005), due to food (F - 29.15, p<.005), due to
tank size (F, -8.29, p - .01-.025), and due to the temperature x
J., / •
food interaction (F » 7.35, p * .025-.05). The temperature x size,
food x size, and temperature x food x size interactions were not signif-
icantly different (p>.05). Weight of the fish was greater at the higher
temperature, with the straight Oregon moist diet, and in the larger
tanks. The interaction between the 25 C temperature and the straight
Oregon moist diet also resulted in greater weight gain.
Analysis of variance of final lengths (Table 22) also indicated signif-
icant differences due to temperature (F - 63.56, p<.005), due to food
1,'
(Fj^ 7 - 6.01, p - .025-.05), and due to tank size (.f^ ? - 14.52, p »
.005-.01). None of the two-factor interactions resulted in significant
differences in final lengths (p>.05), but the three-factor interaction
of temperature, food, and tank size did result in a significantly
greater length (F
^ ?
7.51, p - .025-. 05) with higher temperature,
straight Oregon moist diet, and larger tank size.
59
-------
Table 20 (continued), WEIGHT OF INDIVIDUAL FISH IN ALL TANKS AT TERMINATION AFTER 393 DAYS
(grams)
' • . I
Nominal 20 C
Replir
cate
tank
2
Fish
num-
ber^'
7
8
. 9
10
Mean
Straight
Large
tank
102.5
104.0
141.0
151.0
93.05
diet .
Small
tank
85.0
92.5
119.0
141.0
84.40
Mixed
Large
tank
110.5
113.5
122.0
-
92.00
diet
Small
tank
107.5
154.5
-
-
71.75
Straight
Large
tank
155.5
156.5
186.5
-
128.27
Nominal 25 C
diet
Small
tank
133.5
171.5
202.5
-
112.83
Mixed
Large
tank
81.0
120.0
161.0
162.0
89.05
diet
Small
tank
128.5
130.5
152.0
-
91.70
a/
— Number corresponds to fish listed in Tables 22 and 23.
— Fish lost after 115 days due to accident.
Table 20. WEIGHT OF INDIVIDUAL FISH IN ALL TANKS AT TERMINATION AFTER 393 DAYS
(grama)
Nominal 20 C
Repli-
cate
tank
1
2
Fish
num-
1
2
3
4
5
6
7
8
9
10
Mean
1
2
3
4
5
Straight
Large
tank
22.0
55.0
74.0
75.5
79.0
83.0:
• 86.0
92.0
105.5
1«5.0
85.70
24.5
58.0
65.5
87.5
97.5
diet
Small
tank
38.5
41.0
53.5
55.5
71.0
83 .'s
90.5
103.0
121.5
147.5
80.55
55.5
60.0
63.0.
71.0
78.5
Mixed
Large
tank
25.0
42.0
57.0
73.5
. 77.5
82.0
82.5
94.5
97.0
104.5
. 73.55
61.0
66.0
73.5
80.0
100.5
diet
Small
tank
-]>/
-
-
-
-
-
-
\ -
-
-
14.0
31.5
55.0
55.0
75.0
Straight
Large
tank
49.0
50.0
50.5
57.0
80.5
110.0
111.5
197.0
237.0
274,5
121.70
75.5
89.5
94.5
124.0
136.0
Nominal 25 C
diet
Small
tank
35.5
47.0
59.5
78.5
124.0
135.5
146.5
148.0
179.5
-'
106.00
23.0
26.5
98.5
119.5
120.0
Mixed
Large
tank
18.5
59.0
75.0
91.0
95.5
98.0
102.5
135.0
143.0
149.5
96.70
25.0
55.5
64.5
69.0
71.5
diet
Small
tank
20.0
45.5
59.5
66.0
67.5
78.0
88.5
92.5
151.5
199.5
80.85
34.5
43.5
49.5
56.0
57.0
-------
Table 21 (continued)
MEAN WEIGHTS OF FISH FROM ALL INDIVIDUAL TANKS AT APPROXIMATELY MONTHLY
INTERVALS
, (grams)
Nominal 20 C
Days
after
start
310
340
393
Repli-
cate
tank
1
2
1
2
1
2
Straight
Large
tank
74.67
81.00
82.51
90.07
85.70
93.05
diet
Small
tank
71.89
74.35
78.01
79.09
80.55
84.40
Mixed
Large
tank
62.87
78.72
72.14
87.56
73.55
92.00
diet
Small
tank
-
58.11
-
66.27
-
71.75
Straight
Large
tank
98.23
100.32
112.67
114.67
121.70
128.27
Nominal 25 C
diet
Small
tank
86.21
88.67
93.72
97.96
106.00
112.83
Mixed
Large
tank
79.70
71.81
89.62
81.48
96.70
89.05
diet
Small
tank
70.15
65.46
80.62
82.46
86.85
91.70
— Fish lost after 115 days due to accident.
Table 21. MEAN WEIGHTS OF FISH FROM ALL INDIVIDUAL TANKS AT APPROXIMATELY MONTHLY INTERVALS
(grams)
Nominal 20 C
Days
after
start
70
99
127
158
188
219
250
280
Repli-
cate
tank
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
Straight
Large
tank
1.92
2.47
6.46
8.19
12.73
15.33
21.15
25.16
32.01
34.57
43.99
49.00
56.64
63.08
67.46
75.35
diet
Small
tank
1.44
1.67
6.14
7.40
13.59
14.78
22.98
23.48
30.92
31.75
42.41
43.02
53.70
56.58
63.63
67.53
Mixed
Large
tank
2.10
2.56
6.37
7.48
12.29
13.34
19.05
20.75
26.85
28.43
35.43
38.69
46.77
57.57
57.36
70.57
diet
Small
tank
0.98
1.33
4.63
5.46
10.50
10.45
a/
17.13
-
24.31
-
32.21
-
43.16
-
52.84
Straight
Large
tank
4.67
4.04
10.28
10.08
16.12
15.72
25.67
24.69
36.38
35.54
55.19
48.49
71.76
75.46
88.03
89.43
Nominal 25 C
diet
Small
tank
4.09
3.59
10.74
10.22
17.13
16.67
25.31
25.02
34.64
33.19
48.49
47.55
63.10
62.46
74.22
79.02
Mixed
Large
tank
3.17
3.21
8.43
7.14
13.69
11.72
20.37
19.60
29.00
25.26
38.26
35.77
52.19
50.42
70.44
61.83
diet
Small
tank
2.91
3.21
7.92
7.59
12.66
13.26
19.23
20.41
27.11
28.87
37.00
40.18
51.00
55.62
63.02
58.42
-------
Table 22. STANDARD LESGTHS OF' INDIVIDUAL FI3K IN ALL TANKS
AT TERMINATION AFTER 393 DAYS
(millimeters)
Nominal
Repli- Fish
cate num-
tank ber—
1 1
2
3
.4
5
6
7
8
9
10
Mean
2 1
2
3
4
5
6
7
8
9
10
Mean
•Straight
Large
tank
76
96
114
124
117
121
119
117
128
160
117.2
74
102
92
112
125
122
•125
127
136
146
116.1
diet
Small
tank
83
92
96
104
107
112
113
117
132
142 '
109.8
99
99
112
110
100
118
123
128
133
141
116.3
20 C
Mixed
arge
tank
75
. 94
99
119
120
121
123
128
128
134
114.1
102
107
120
120
134
132
132
135
140
-
124.7
Nominal 25 C
diet
Small
tank
- —
-
-
-
-
-
-
-
-
-
64
84
84
94
112
116
130
145
-
-
103
Straight
Large
tank
' 99
105
94
108
125
128
.139
166
177
-189
133.0
111
119
124
135
142
143
153
146
151
-
.6 136.0
diet
Small
tank
89
99
101
121
143
139
150
139
160
-
126.8
77
85
125
137
136
125
144
147
159
-
126.1
Mixed
Large
tank
74
110
111
128
126
132
139
150
150
153
127.3
87
106
113
113
115
115
125
142
156
156
122.8
diet
Snail
tank
80
100
107
115
120
I"
136
137
160
177
124.7
79
?7
105
107
139
118
.139
148
153
163
124.8
Maturation
At termination of the experiment the gonads were examined and a subjec-
tive judgment on the degree of maturity was made using the following
classification (Table 23):
I. Immature - gonads extremely small and difficult to locate with
the naked eye
II. Resting stage - sexual products have not yet begun to develop;
gonads very small but easily located
III. Maturation - gonads of moderate size; eggs not distinguishable
to the naked eye and milt not running when pressure is applied
to abdomen
IV. Maturity - gonads have nearly achieved maximum size; eggs
distinguishable to naked eye, but not extruded with pressure;
milt is extruded under light* pressure and some secondary
sexual characteristics are visible.
Although f ish were large enough and old enough to spawn based on our
previous experience with field-reared fish, only one female and two
males had nature gonads. These three fish all were in a large, 25-C
tank and fed the mixed diet, but the replicate of this treatment was
similar to all other tanks (Table 23). The presence of the three most
mature fish in the same tank may indicate an uncontrolled and unknown
variable influencing those fish.
To determine if any experimental condition resulted in more mature fish,
an analysis of variance was calculated on the percentage of III and IV
group fish in each tank after an arcsin transformation of the percentages.
No significant differences were detected between treatments, factors, or
factor interactions. This experiment was unsuccessful in providing con-
ditions suitable for complete maturation within a desired 1-yr period.
corresponds to fish listed in Tables 20 and 23.
- Fish lost after 115 days due to accident.
65
\
-------
Table 23. SEX AHD STATE OP GONADAL MATURITY4' OF INDIVIDUAL FISH
IN ALL TAJIKS AT TERMINATION AFTER 393 DAYS
Nominal
Repli- Fish
cate num-
tank ber^'
1
Female
1
2
3
4
5
6
7
8
9
10
(Z)
Straight
Large
I
tank
F-II
F-II
F-II
F-II
M-III
F-I
M-II
F-II
F-II
F-II
80
III & IV (Z) .10
2
Female
III &
1
2
3
4
5
6
7
8
9
10
(Z
IV
M-I
M-II
M-II
.F-II
F-II
F-II
F-II
F-II
M-II
M-II
) 50
(Z) 0
diet
Small
tank
F-II
F-II
M-I
F-II
M-III
M-III
M-III
M-II
F-II
F-II
50
30
M-I
F-II
M-III
M-II
M-II
M-III
F-II
F-II
F-II
F-II
50
20
20 C
Mixed
Large
tank
F-II
M-II
F-II
M-III
F-II
M-II
M-II
M-II
F-II
M-II
40
10
M-I
F-II
M-I
F-II
F-II
F-II
F-II
F-II
F-II
-
77.8
0
Nominal 25 C
diet
Small
tank
_c/
-
- '
-
-
. -
-'
-
-
-
-
-
M-I
F-II
F-II
F-II
M-II
M-II
F-II
F-II
-
-
62.5
0
Straight
Large
tank
M-II
F-II
M-II
M-III
M-II
F-II
F-II
M-III
F-II
M-II
40
20
M-II
F-II
F-II
F-II
M-III
M-H
M-II
F-II
M-II
-
44.4
11.1
diet
Small
tank
M-II
M-II
F-II
F-II
F-II
M-II
M-II
F-II
F-II
-
55.6
0
F-II
M-II
M-II
M-II
M-II
M-II
M-II
M-II
M-III
-
11.1
11.1
Mixed
Large
tank
F-II
M-III
M-III
F-II
M-IV
M-IV
F-II
F-II
M-III
F-IV
50
60
F-II
F-III
M-II
F-II
F-II
F-I
M-II
M-II
F-II
M-III
60
20
' — I
diet
Small
tank
M-II
1
F-II 1
M-I i
i
M-I '
F-II ;
i
M-II i
F-II
F-II
F-II
M-II
50
I
M-II |
M-II
M-II
F-II
1
M-III
M-II
M-II
F-II
F-II
M-II
30
10
—.Roman numeral maturity rating discussed in text.
—'Number corresponds to fish listed in Tables 20 and 22.
—All fish lost after 115 days due to accident.
66
SECTION VI
DISCUSSION
SELECTION OF A STANDARD SPECIES
Acute Toxicity
A desirable attribute of a standard fish species would be moderate
sensitivity to many toxicants. Fish with extreme sensitivity would
require tremendous dilution of effluents for proper concentrations in
an acute, test and chemical analysis of low concentrations might" be
less accurate or precise. Extreme resistance to many toxicants might
require such high concentrations of industrial effluent that no kill
could be attained with full strength effluents. Either situation would
add complexities to standard bioassays meant to be conducted with only
semi-experienced technicians. Since little data in the literature on
the toxicity of sodium chloride, pentachlorophendl, hexavalent chromium,
(ft
or Guthion*- is comparable to the present study because different bio-
assay conditions were maintained, the sensitivity of the fathead minnow
and goldfish cannot be compared to other species, only to each other.
The toxicity curves indicate that although mortality of fathead minnows
from three of the toxicants (pentachlorophenol, sodium chloride, and
Guthioir) is Initially more rapid than mortality of goldfish, by ter-
mination of the tests goldfish were at least as sensitive, if not more
so, than the fathead minnows. The toxicity curves for hexavalent
chromium are incomplete but goldfish may ultimately also be more sensi-
tive to this toxicant (Figure 2). The initial resistance of the gold-
fish make that species less desirable as a standard since some industries
67
-------
may use 24-hour tests.
Variability of Bioassay Results
On the basis of minimum variability in results of acute bioassays,
neither the goldfish nor the fathead minnow was superior, and varia-
bility was more dependent on the toxicant than on the species. With
the toxicants for which a threshold LC50 was attained rapidly, sodium
chloride and pentachlorophenol, variability was relatively small, but
with GuthioiM and hexavalent chromium, where a threshold LC50 was not
attained by 11 days, variability was considerably greater (Table 12).
Thus the decision on selection of a standard fish between these two
species must be based on other criteria.
Goldfish Culture and Other Criteria
Both species are easily handled for bloassay purposes; however smaller
.test containers can be used for fathead minnows since even 14-week-old
fathead minnows are considerably smaller than the size of goldfish
available from Ozark Fisheries. Both species should be easily trans-
portable since both are tolerant' of low oxygen concentration. The
fathead minnow has been used successfully in many chronic studies since
sexual maturity can be achieved in about 3 to 5 months. The goldfish
culture study was unsuccessful in the attempt to raise fish from an egg
to spawning adult in one year, and we are not aware of anyone having
done so under laboratory conditions. Further studies are necessary to
achieve success. On the basis of these biological considerations, the
fathead minnow is recommended as the better of the two species for a
standard bioassay fish. Since the goldfish is available from a large
commercial source it could be used at present, although stocks obtained
during the sunnier are in poor condition and generally cannot be used for
bioassay. Regulating agencies could probably promote the development of
a source for fathead minnows. Fathead minnows from a variety of sources
may be undesirable if they vary considerably in resistance to a toxicant.
68
Size and Age of Standard Fish
Since the threshold LCSO's for the large and small goldfish exposed to
pentachlorophenol were significantly different (Table 18), it may be
desirable to restrict the size selected for bioassay when that species
is used as a standard. From a practical point of view, however, the
actual difference in LCSO's was so small that this size restriction is
probably unnecessary. Variability resulting from factors other than
size was just as large (Table 3).
Although different rates of mortality occurred between large and small
fathead minnows, this difference affected LCSO's only at very early
time intervals (Table 18). Most standard bioassays require a minimum •
of 24 hours and LCSO's computed at or .subsequent.to that time were not
significantly different. Therefore, size selection of constant age
fathead minnows would be unnecessary for pentachlorophenol.
A similar result occurred with fathead minnows of different ages. Al—
though rates of mortality differed, there was no significant difference
between LCSO's computed at 24 hours or longer with 4-, 7-, 11-, and
14-week-old fish (Table 18). If culture units similar to that described
by the U.S. EPA (1971) are used for rearing fathead minnows, various
age groups could be mixed to increase numbers available for bioassay.
However, since fish of various ages were only tested with pentachloro-
phenol we cannot, be sure if differences due to age might occur with
other toxicants. It is therefore recommended that only the 4- or 7-
week-old fish be tested. By using younger fish, culture unit production
could be Increased. These smaller fish would also allow use of smaller
bioassay chambers or more fish per chamber. The main disadvantage of
using the small fish is difficulty of observation in a large chamber,
but this criticism is minor provided observations are made carefully.
SELECTION OF REFERENCE TOXICANT
On the basis of minimum variability and limited information on the
69
-------
ability to detect abnormal fish, of the four toxicants tested sodium
chloride best met the requirements for a reference toxicant (Table 15).
Although pentachlorophenol can be used and may have the advantage of
more rapid detection of abnormal fish, careful consideration must be
given to pH of the test water since toxicity is greatly influenced by
pH (Crandall and Goodnight, 1959). This necessity might complicate
industrial bioassays. Additional studies are needed to determine
clearly the ability of both sodium chloride and pentachlorophenol to
detect abnormal fish.
Use of any reference toxicant will require previous standardization by
many bioassays to establish a baseline with which comparisons can be
made. In establishment of this baseline, complete information on the
fish stocks will be required to.eliminate stressed fish from considera-
tion. Since even presumed healthy fish vary, rejection of a bioassay
stock should be considered only if it falls 1 1/2 to 2 standard devia-
tions outside of the mean of baseline tests. Anything more restrictive
would result in rejection of many stocks of relatively, normal fish.
Stocks falling outside of this criterion would be expected to be
suffering from some known or unknown abnormality.
70
SECTION VII
REFERENCES
American Public Health Association, American Water Works Association,
and Water Pollution Control Federation. 1971. Standard Methods for
the Examination of Water and Wastewater. 13th ed. New York, American
Public Health Association, Inc. 874 p.
Cairns, J., Jr. 1969. Fish Bioassays - Reproducibilify and Rating.
Revista de Biologia (Rio de Janeiro). £(1-2):7-12.
Chemagro Division Research Staff. 1974. Guthion (Azinphosmethyl):
Organophosphorus Insecticide. Residue Rev. 51:123-180.
Crandall, C.A., and C.J. Goodnight. 1959. The Effect of Various
Factors on the Toxicity of Sodium Pentachlorophenate to Fish. Limnol.
Oceanog. 4^53-56.
Davis, J.C., and R.A.W. Hoos. 1975. Use of Sodium Pentachlorophenate
and Dehydroabietic Acid as Reference Toxicants for Salmonid Bioassays.
J. Fish. Res. Board Can. 32:411-416.
Fromm, P.O., and R.H. Schiffman. 1958. Toxic Action of Haxavalent
Chromium on Largemouth Bass. J. Wildl. Mgt. 22:40-44.
Grindley, J. 1946. Toxicity to Rainbow Trout and Minnows of Some Sub-
stances Known to be Present in Wast Waters Discharged to Rivers. Ann.
Appl. Biol. 33:103-112.
71
-------
Haskins, W.T. 1951. Colorimetric Determination of MicroRram Quan-
tities of Sodium and Copper ?entachlorophenates.. Anal.. Chera. 23:1672-
1674.
Heryey, G.F., and J. Hem. 1968. The Goldfish. London, Faber and Faber.
271 p.
Knoll, J., and P.O. Fromm. 1960. Accumulation and Elimination of Hexar
valent Chromium in Rainbow Trout. Physiol. Zool. 33:1-8.
Lennon, R.E. 1967. Selected Strains of Fish as Bioassay Animals. Frog.
Fish-Cult. 29_( 3): 129-132.
MacLeod, J.C. 1972. A Standard Rainbow Trout Unit.(RTU) for Acute
Toxicity Determinations in Industrial Effluents. Fish. Res. Board Can.,
Freshwater Inst. Winnipeg. 14 p.
Handel, J. 1961. Non-additivity in Two-way Analysis of Variance. J.
Am. Stat. Assoc. 56:878-888.
Marking, L.L. 1966. Evaluation of p,p'-DDT as a Reference Toxicant in
Bioassays. U.S. Bur. Sport Fish. 6, Wildl., Wash., Resource Pub. 14.
10 p.
Martin, D,M. 1973. Freshwater Laboratory Bloassays - A Tool in Environ-
mental Decisions. Dept. Limnol., Ac ad. Nat. Sci., Philadelphia. Contrib.
No. 3. 51 p.
Mount, D.I., and W.A. Brungs. 1967. A Simplified Dosing Apparatus for
Fish Toxicology Studies. Water Res. 1^21-29.
ORSAHCO Biological Water Quality Committee. 1974. ORSANCO 24-hour Bio-
assay. Ohio River Valley Water Sanitation Commission, Cincinnati, Ohio.
21 p. .
72
Rudling, L. 1970. Determination of Fentachlorophenol in Organic
Tissues and Water. Water Res. 4_: 533-537.
Sprague, J.B. 1969. Measurement of Pollutant Toxicity to Fish. I.
Bioassay Methods for Acute Toxicity. Water Res. 3^793~821-
Sprague, J.B. 1970. Measurement of Pollutant Toxicity to Fish. II.
Utilizing and Applying Bioassay Results. Water Res. 4^:3-32.
Stokes, R.M., and P.O. Fromm. 1965. Effects of Chromate on Glucose
Transport by the Gut of Rainbow Trout. Fhysiol. Zool. 38:202-205.
Trama, F.B., and R.J. Benolt. 1960. Toxicity of Hexavalent Chromium to
Bluegills. J. Water Pollut. Contr. Fed. ^2:868-877. ' _
U.S. Environmental Protection Agency. 1971. Tentative Plans for the
Design and Operation of a Fathead Minnow Stock Culture Unit. Natl.
Water Quality Lab., Duluth, Minn. 8 p.
Webb, P.W., and J.R. Brett. 1973. Effects of Sublethal Concentrations
of Sodium Pentachlorophenate on Growth Rate, Food Conversion Efficiency,
and Swimming Performance in Underyearling Sockeye Salmon (Oncorhynchus
nerka). J. Fish. Res. Board Can. 30:499-507.
73
-------
SECTION VIII
PUBLICATIONS
Adelman, I. R., L. t. Smith, Jr., and G. D. Siesennop. Acute Toxicity
of Sodium Chloride, Pentachlorophenol, GuthionT and Hexavalent Chromium
to Fathead Minnows (Pimephales promelas) and Goldfish (Carassius
auratus). J. Fish. Res. Board Can. ^3:000, 1976.
Adelman, I. R., and L. L.-Smith, Jr. Fathead Minnows (Pimephales
promelas) and Goldfish (Carassius auratus) as Standard Fish in Bio-
assays and Their Reaction to Potential Reference Toxicants. J. Fish.
Res. Board Can. 33:000, 1976.
Adelman, I. R., L. L. Smith, Jr., and G. D. Siesennop. Effect of Size
or Age of Goldfish and Fathead Minnows on Use of Pentachlorophenol as
a Reference Toxicant. Water Res. (In press)
SECTION IX
GLOSSARY
Bioassay - A toxicity test: the estimation of the strength of a poison
by its effect on a living organism.
Bioassay, acute - A toxicity test of short duration, usually less than
3 weeks.
Bioassay, chronic - A toxicity test of long duration, usually 1 month
or longer.
Bioassay, flow-through - A toxicity test where the toxicant and diluent
water are continuously replaced by fresh material.
Bioassay, renewal - A toxicity test where the entire toxicant and
diluent water are replaced at period intervals; in this study every 2
or 3 days.
Coefficient of Determination (r2) - The proportion of the sum of squares
of the dependent variable that can be attributed to the independent
variable.
Coefficient of Variation (CV) - The sample standard deviation expressed
as a percentage of the mean.
Error rate, per comparison - The probability of occurrence of one or
more false significant differences among comparisons of different treat-
cents.
75
-------
Error race, per experiment - The probability of occurrence of one or
more false significant differences among comparisons from all experi-
ments.
Fry - Postembryonic fish up to about 10 days old.
Guthion (a2inphps-methyl) - An organic phosphorus insecticide - chemical
name: o,o-Dimethyl S-[4-oxo-l,2,3-benzotriazin-3(4H)-ylmethyl-]-phosphoro-
dithioate. •
Homeostasis - A state of physiological equilibrium; in this case where
processes of excretion or detoxification just balance the toxic action
of the poison.
Juvenile - Fish older than 10 days but not sexually mature.
LC50 - The concentration of poison that will kill 50%. of the test
organisms at a specified time.
LC50, threshold - The concentration of poison at which the organism
presumably reaches homeostasis with the toxicant; arbitrarily indi-
cated by 2 days of no deaths in the bioassay.
Type I Error - A statistical error where the experimenter rejects the
null hypothesis and it is true.
Toxicity Curve - A mathematical description of the effect of a toxicant
where the LC50 is the independent variable, plotted on the ordinate,
and time the dependent variable., plotted on the abscissa.
SU-S. GOVERNMENT PttlNTmG OTFItt. 137S-65T f
76
------- |