PB84-129493
Interlaboratory Comparison of Continuous
Plow, Early Life Stage Testing wJth
Fathead Minnows
(U.S.) Environmental Research Lab., Duluth, MN
Dec 83
r
US. Depslmgflt @$
ffetisas! TeetufeaJ l^mat^t
-------
TECHNICAL REPORT DATA
REPORT NO.
EPA-600/3-84-005
a naeiPtiN-rs Accaasio* MO.
PBS k 12949 5
«. T1TUG AND SU8TITU2
Interlaboratory Comparison of Continuous Flow, Early
Life Stage Testing with Fathead Minnows.
1. REPORT OATH
December 1983
S.POMFORMJWO ORGANIZATION COD3
AUTMORUtt
A. E. Lenke
3. PERFORMING. OROAMIZATIOM RSPORTNO.
PERFORMING ORGANIZATION MAM8 AMO ADDRE6S
U.S. Environmental Protection Agency
Environmental Reoearch Laboratory-Dul^th
6201 Congdon Boulevard
Duluth, Mlnnbsota 55804
1O. PROORAM 8LHMSNT NO.
11. COMTrACr/OMANT NO.
2. SPONSORING AO9NCY NAME AND ADDRESS
Same as above
IX TVPt OP RCPORT AND PRRIOO COViRiO
14. 8PONSORINO AOENCV COOI
EPA/600/03
&. SUPPLBM8NTARV NOTES
1O. ABSTRACT
Six laboratories conducted toxicity experiments according to a supplied protocol. Also
supplied were the chemicals to be tested (acenaphthene and iscphoronc). Test organisms
were fathead minnow (Plmepnales proaelas) embryos which vere raised until 28 days post
hatch. All fish were weighed and compared with the controls, by each investigator and
all analytical work and other necessities were provided by the various participants.
Results ranged between 0.049 mg/1 and 0.42 mg/1 for the low solubility acenaphthene and
between 1.35 mg/1' and 45.4 mg/1 for a more soluble isophorone. The isophorone results
were strongly correlated inversely to growth of the controls which varied between 0.969
gr and 0.018 gr for a high and low, respectively, ft—;——
7.
KBY WORDS AND OOCUM8NT ANALYSIS
DESCRIPTORS
b.lD6NTIFISHS/OPfiN§NOEO TERMS C. CO3ATI FieM/CfOOp
16. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
It. SECURITY CLASS (TJitl Rtport)
UNCLASSIFIED
ai. NO. of PAOSS
29
2O. SECURITY CLASS
UNCLASSIFIED
22. PRICK
2220-1 (8»». 4-77)
BOITIOM i» oesoklTS
-------
EPA-600/3-84-005
December 1983
Intei-laboratory Comparison of Continuous Flow, Early Life Stage Testing
with Fathead Minnows
A. E. temke
U.S. Environmental Protection Agency
Environmental Research Laboratory-Duluth
6201 Congdon Boulevard
Duluth. Minnesota 55804
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. M-ntion of trade names
or coBsnercial products does r.ot constitute endorse-
ment or recommendation for ure.
11
-------
Interlaboratory Comparison of Continuous Flow, Eatly Life Stage Testing
f
with Fathead Minnows
v
A. E. Lemke
Toxicity testing with various organisms and various life stages is an
important part of environmental protection strategy. The fathead minnow
(Pimephalea promelas) is a small cyprinid, well suited to laboratory-
conditions. Tefita have been conducted on various materials using acute (96
hr), sub-chror.ic (embryo, larval, juvenile, 30 day), and full life history
techniques (6-8 months).
Toxicity testing data summarized by MeKim (1977) indicates that the
early life stages, particularly at the time the ycung fish begin to feed
after absorbing their yolk sac, arc the most sensitive.
A set of guidelines (attached) for the fathead minnow embryo larval test
was prepared by the research group at the Environmental Research Laboratory-
Duluth of the U.S. ETA. Seven laboratories participated in the evaluation of
these guidelines. The participating laboratories will be designated by the
following acronynms: (ANSP, ERCO, MBL, JEI, ABI, WF1S, ERL-D). Laboratory
names, addresses and principal investigator are listed in Appendix 1.
General Procedures
A request for proposals was initiated through the EPA contracting
office. This proposal contained the guidelines, a rating scheme for
evaluation, and a scope of work. The acope of work requested that two
chemicals, acenaphthene and isophorone, be tested twice each in tests
beginning with less than 24 hr old embryos and terminating 28 days after
hatching. The guidelines called for the principal reporting parameter to be
the nonlethal effect of growth different from the controls with 96 hr LC50
-------
and 30 day LC50 also to be reported if developed as part of the growth
.differential testing. •,
The highest concentration not different from the control and belov which
all concentrations were not different from the control was designated as the
lower chronic endpoint. Conversely, the lowest concentration which was
different from the control and above which all concentrations were different
from the control was called the upper chronic endpoint. These were compared.
Other parameters compared were number of test concentrations, toxicant
preparation solvents, number of embryos initially, chamber size, embryo cup
construction, water supply, photoperiod, temperature, dissolved oxygen, flow
rate, culture techniques of stock culture, acquisition and handling of
embryos initially, control conditions, feeding (timing, kind, amount), tank
cleaning, termination techniques, and chemistry of both compounds. All
comparisons were made in relation to the guidelines as written. If no
specific suggestions were made, the individual labs compared with each other
and KRL-D experience.
Finally, suggestions for improving the guidelines were requested.
Procedure Adherence by Participating Laboratories
The guidelines recommended that the test water should allow rearing of
fish and other aquatic organisms. Water quality parameters called for
included, temperature 25 +_ I* C, pH 7.1-8.5, dissolved oxygen between 75 and
100 of saturation, and flow rate sufficient to prevent loss of toxicant
concentration in contact with the test organisms. These parameters have been
found to provide a suitable base for good survival and growth of fatnead
minnow embryos and larva. Other recommendations are found in the guidelines
(attached). All laboratories followed water quality procedures well within
2
-------
the guidelines. Hardness varied from a high of 390-310 mg/1 CaC(>3 (ERCO)
to a constant low of 47 mg/1 +_ 1.5 mg (ERL-D) and a variable low between
27-70 (WFTS). All of the laboratories used the test water for culture
purposes also.
. pR levels reported by all but one laboratory were well within the
satisfactory range of 7.1 - 8.5. MBL did not report pH.
Temperatures in the tests were well controlled with all laboratories
reporting a mean of 25°C +^ l*C as recommended by the guidelines.
Four of the six laboratories used well water as dilution water (ERCO,
MBL, JEI, API, and WFTS). One laboratory reported using unfiltered water
from a pristine stream, the watershed of which is controlled by the
laboratory (ANSP). The other laboratory used sand filtered Lake Superior
water obtained by pipeline offshore (ERL-U).
The other water parameter asked for in the guidelines was dissolved
•
oxygen. All laboratories reported dissolved oxygen levels of +70%. WFTS
reported concentration 6.2 - 8.4 rag/D.O./l which converts to 74-- 98%
saturation at the reported test temperature. One laboratory (MBL) did not
follow the guidelines. Instead, they maintained the dissolved oxygen by
aeration of the test tanks.
Guideline test conditions concerned with operational procedures included
starting numbers of embryos, sorting, water turnovers, test chamber size
(Table 1), use of solvents to disperse the test materials, and number of test
concentrat ions.
Embryo Handling
The guidelines recommended 60 embryos or more per concentration
initially. . BI used 40 per concentration in each test and ERL-D used 40 in
one teat and 60 or more in the other three. The sorting or reduction in
3
-------
numbers of organisms before transfer into the rearing tanks was not
recommended. ANSP and WFTS reduced the final number of larva to 40 for
rearing, all others transferred all hatched animals. All participants used
duplicate chambers and five concentrations and a control. They replicated by
space and/or time, giving 4 separate tests as requested in the scope of work.
Exact turnover rates were not specified in the guidelines other than that
they should be sufficient to keep the toxicant concentration stable and
dissolved oxygen at +752 saturation. Turnover rates from 8 to 88 times in 24
hrs. were reported, MRL did not report turnover rate.
All laboratories used diethylformamide as a solvent for the acenapthene
.which is only sparingly soluble in water. MBL and JEI reported difficulty
with che solvent. Isophorone was added directly to the test water by all
reporting units except JEI which used a solvent.
jV
Chemical Methods
Chemical analyses of the test concentrations were accomplished by the
following me*hods.
Analytical Chemistry Methods
Laboratory Acenapthene
ANSP Gas liquid chromatograph
ERCO Fluorometric
MBL Flourometric
JEI Gas chromatograph
ABI Gas chromatograph
WFTS High pressure liquid chromatograph
ERL-D Flourotnetric
Isophorone
UV spectroscopy
Gas chromatograph
Gas chromatograph
Gas chromatograph
Gas chromatograp'u
Gas chromatograph
Gas chromatograph
-------
Analytical problems were reported by MBL and JEI. They both reported'
difficulty with flocculent material associated with the dimethyl foramide
used as the solvent for acenapthene. MBL examined some of their flocculent
material and found large numbers of ciliates associated with the floe.
t* '
i -'
Table 1
. Chamber Sice (in inches) and Egg Cup Condition
L W H Agitation of Embryo Cups
No
Yes
Yes
Stirred water
Yes (siphon)
Yes (screen tray)
Yes
,. Table 1 shows the chamber sizes utilized by the participants for fry
rearing. This range of chamber sizes is as large or larger than suggested in
the guidelines. The guidelines called for making sure that the eggs
(embryos) were exposed to the toxicant and sufficient .oxygenated test water.
Gentle agitation in a reciprocating system was suggested but not mandatory.
Only ANSP did not provide any special consideration for insuring water flow
around the embryos.
Feeding of Fry
All laboratories fed brine shrimp (Artemia salina) nauplii at least
twice a day throughout the test except on weekends when they were fed once by
most participants (ANSP fed twice daily on weekends). ERL-D supplemented the
ANSP
ERCO
KBL
JEI
ABI
WFTS
ERL-D
40
18
34.5
25
24
9
6
20
6
15
16
19
9
16
26
8
14.5
17
19.9
10.5
18
-------
brine shrimp with fine particle trout chow and ANSP reported i large amount
of natural food, i.e., copepods, etc., in their unfiltered stream water.
No mention was made by any laboratories which day in relation to the
weekend the tests were started with the exception of ERL-D. Their report
expressed a belief that a shortened early feeding schedule (larva emerged on
weekend) on one test accounted for leas, but still satisfactory growth than
the other three tests.
Test Results
The effect/no effect levels as reported by the participating labora-
tories fire presented in Table 2. The acenaphthene, no effect means range
from a low of .049 mg/1 to a high of .42 mg/1 or approximately a 10-fold
difference. Isophorone no effect means have a greater range 1.35 mg/1 to
45.4 mg/1, a 40-fold difference.
Growth variations as measured by reported values of control fish
weights (Table 2) vary from a high of .969 gr for ANSP to a low of .0018 gr
for MBL, nearly 3 orders of magnitude (>50 times). That indicates problems
with the feeding regime.
An anomaly in the data set is the difference in ranges between weight,
acenaphthene endpoints, and isophorone endpoints. To resolve the problem,
regression analyses were conducted with each of the sets regressed against
the others. No significant correlation was found with the acenapthene data
versus either of the other sets. The weight versus the isophorone data
resulted in an R* of .87 in the exponential least squares regression of the
form Y = AeBX where A = 24.66 and B « -3.087. Thus as the weight goes
up, differences from the control are revealed at lower concentrations. This
by itself is logical but the acenaphthene data does not show this effect;.
-------
Table 2
Interlaboratory Comparison Test Results
Participating
Organization
ANSP
ERCO
MBL
JEI
ABI
WITS
ERL-D
Ace 1
.953
.948s
.0056
.0030
,0052s
.095
.083
.085s
.031
.037s
.186
,214s
Control
Ace 2
*•
.969
.965s
.0057
NR
.Oils
.012
.091
..095s
.023
.122
.180
Growth
Iso 1 Iso 2
.957 .969
.0061 .0055
.0095 .0018
.045 .088
.026
.141 .202
Ace 1
.075-. 18
(.16)
.064-. 098
(.079)
.034-. 071
(.049)
-
.11-. 24
(.16)
.33-. 50
(.41)
.133-. 265
(.188)
Test Results
Ace 2
.18-. 37
(.26)
.091-. 139
(.112)
.033-. 088
(.054)
"
.12-. 27
(.18)
.35-. 51
(.42)
.146-. 285
( .204)
No Effect Level
Iso 1
.65-2.8
(1.35)
14.8-?4.5
(19.0)
31.5-65.4
(45.4)
-
9.2-20.0
(13.6)
11-19
(14.5)
14.5-27.6
(20.0)
Iso 2
.65-2.8
(1.35)
11.8-53.9
(25.2)
29.3-56.1
(40.5)
-
19.9-40.6
(28.4)
-
4.15-8.78
(6.0)
All weights are in mg and are mean weight of surviving control fish.
All participants met the criteria for control survival of 80* or better.
Result numbers in mg/1 analyzed.
s is solvent control weights
(n) are geometric means
-------
Discussion and Resolution of Procedural Problems
the difficulties encountered by participants io conducting the bioassays
can be categorized in three ways: handling problems with the solvents and
the chemicals, difficulty with the frequency of determinations to be taken,
and-test organism problems, especially lack of growth.
•. _ f
. these problems are universal in bioassay work. Twenty-five yeat3 of
experience by the author have shown all of these to be problems encountered
by experimenters who are trying to follow a slightly different system than
their usual routine. Experience with similar procedures reduces the problems
but they are not eliminated. Discussion with two ERL-D testing crews
(personal communication, Duane Benoit, Gary Phipps) indicated that chemical
handling is a problem for even the most experienced testing units.
Solubility, volatility, and cross-contamination between tests are always a
problem. Each requires different resolution and, although they are not an
actual part of the bioassay procedure, they do contribute to about one-third
of the failures, especially as the types of compounds tested change.
Some work at our laboratory in this area of structure activity has shown
that low water solubilities can cause unexpected results (Veith, personal
communication). Three of the laboratories reported some solubility problems
with acenapthene, and two used saturated solutions as the highest
/
concentration. Isophorone did not cause this problem,' The solubilities in
water are 3.42 mg/1 and 12,000 mj»/l as acenapthene and iiiophorone,
respectively (Callahan et al., 1979). The low solubi l.ity of acenaphthene
supports the contention that the results of the acenaphthene tests are not
affected by the differences in growth as were the isophorone tests.
The second problem area was reported by all contract laboratories. It
concerned the actual number of analyses, type of analyses, and statistical
8
-------
evaluations to be made during the chemical and biological data gathering.
All of the laboratories preferred sets of standard forms for daily use and
for compilation at the end of the test set. the contract laboratories also
reported that they could do more accurate pricing if such a set of forms were
included. Contract laboratories said a further benefit of a form set is that
it would serve as a check list. '
The other reported problem is the difficulty found in actually rearing
and growing the test animals. Only three of the seven participating
laboratories had satisfactory growth compared to that obtained by sone of the
.•surde'.ine writers (Benoit et al., 1982; Jarvinen and Tanner, 1982). It is
the opinion of the participants that the feeding regitre suggested in~ the
guideline is not specific enough, and Chat perhaps other foods should be
tried. Suggestions include the use of protozoa cultures, prepared aquarium
food, natural food from a clean water source, and special kinds of brine
shrimp nauplii since apparently they are not all the same size depending on
the source. Also considered to be a problem was the timing of the feeding
during the first several days after the young fatheads absorb their yolk sac
and begin to feed. ERL-D reported 44% more growth when the test animals
hatched during the week rather than during the weekend. (The guidelines
allow a reduced feeding schedule on the weekend.)
Resolution of the request for a set of data gathering forms is easy.
Experience with four interlaboratory comparison test sets indicate that such
forms should be added to the guidelines.
The other two problems, i.e., chemical handling and organism handling
a,id feeding, are much less amenable to resolution.
Chemical handling revolves around system changes required for various
chemicals. Metals and water soluble organics have entirely different
-------
problems than either volatile organic!, or slightly water soluble materials.
Experience with the type of chemical group in question is the best solution.
t.. -
Unfortunately, experience cannot be written into a set of protocols. The
best generalization to alleviate the problem is to suggest thac groups of
chemicals be tested which have similar physical and toxicity characteristics.
Materials like acetone and ethyl alcohol require different chemical handling
than endrin and PCBs. . . -
Liquids, solids, and biodegradable test materials require different
systems. A contract group conducting a large number of flow through 96 hr
acute bioassays with juvenile fathead minnows has developed four types of
test systems (personal communication, Michael Knuth). Chemicals are assigned
to the various systems on the basis of the parameters previously enumerated.
The guidelines cover this area with the statement "suitable distribution
devices to maintain test concentrations". Perhaps separate guidelines
sections should be written for operation of the test systems based on the
physical and chemical parameters of the test materials.
The wide variation in growth of the fish during the 30 day test period
indicates that the food supply and/or food presentation is a problem in1
completing a satisfactory test based on the guidelines. Several reports,
Beck et al. (1980), Johns et al. (1980), indicate that various brine shrimp
are not alike in suitability as a food source for young fish. The experience
of the author and various test groups at ERL-D also indicate that the age of
the nauplii used, i.e., newly hatched versus late first instar also can
affect the suitability as food for first feeding fathead minnows. This is
perhaps caused by size or mobility differences as newly feeding fish are very
selective (Siefert, 1972). Availability, and tank distribution also are
factors based on observation. This is emphasized by the report of one
10
-------
participant (ANSP) which had fish growth four times above the next two
laboratories and ten to seven hundred timea that reported by certain other
participants. The natural food in that laboratory water supply, available
'continuously, was noted as the reason for this growth. Unfortunately, few
laboratories have water *•«.th a good natural food supply and those thac do
have food that varies with the seasonal progression. The ERL-D report
indicates that the food availability in the first few days after feeding
begins id also a factor. The report of ANSP suggests that sose other food
(hay infusion is suggested) be used during this critical period in future
testing.
Each of the above suggestions have merit. However, the nonetandard
nature of each of the suggested remedies makes recommending such additions to
«*v
the guidelines difficult. In this author's opinion, a test with the desired
species with all factors except the toxicant initially comparing weights of
the usual 5 duplicates as an operations check should be conducted. The
necessary changes to reach a minimum mean weight of no less than 0.1 g per
animal at 28 day post hatch co-ild then be made. This would allow comparison
of the expected results throughout the system and a thorough familiarity with
the system before any testing is begun. Any effect of variations in flow,
lighting, tank position, etc., could also be controlled. A similar pretest
check on the toxicant additions and chemical analysis system should also be
conducted but for less time (a week is suggested). These additions to the
guidelines would save time anJ funds. If this is not done, at least two
tests on each chemical should be conducted, a minimum weight for the controls
should be set, and provision should be mede in the initial planning for
retesting of any material in which this minimum is not met. Another check
parameter that could be included in the guidelines is to discard any test
11
-------
where surviving animals in partial kill concentrations are larger Chan the
controls, indicating that food and/or space stress was a problem. (The three
laboratories adjudged satisfactory met this criteria as there was no
difference in fish mean weight in chambers with fewer fish.)
The condition that mu-jt be worked toward in any continuous and
continuing testing procedure is uniformity in all parameters except that
being tested.
These guidelines in general meet this need. The only part of the
guidelines that appear to be in need of modification is the type and/or
quality of food used.
12
-------
Summary of recommendation for use of these guidelines in toxicity
evaluation.
1. Conduct a training session for the principal biologist prior to stare up
of any system following the guidelines.
2. Require all testing laboratories to conduct a full scale test without
toxicant to evaluate their laboratory techniques.
3. Produce a set of forms for daily recording and post test compilation
including statistics (should include number of chemistries and other
information user needs).
4. Assign chemicals with similar solubilities, volatilities, and other
physical properties to each test program as much as possible to increase
productivity of test systems.
5. In lieu of requirement on #2 above, set a minimum control mean weight of
0.1 gr per fish. (This requirement is not as useful as $1 because
chemicals will be lost and repeats will be required while requirement
two will preclude any laboratory doing the work if they have
unsatisfactory techniques.)
13
-------
References
Beck, A. 0., D. A. Bergston, and W. H. Ho well. 1980. Nutritional value of
five geographical strains of Artenia; Effects on survival and growth of
larvel Atlantic silveraide Menidia menidia. pp. 249-259; Volume 3,
Ecology, The Brine Shrimp Artemia, CulCuring, Use in Aqucculture.
Benoit, D. A., F. A. Puglisl, and D. L. Olson. 1982. A fathead minnow
(Pimephalca promelas) early life stage toxicity test method evaluation
and exposure to four organic chemicalo.
Callahan, M. A,, M. W. Slimak, N. W. Gabel, I. °. May, C. F. Fowler, J, R,
Freed, P. Jennings, R. L. Durfee, F. C. Whit more, B. Maestri, W. R,
Mabey, B, R. Holt, and C. Gould. 1979. Water-related environmental
fate of 129 priority pollutants. U.S. Environmental Protection Agency
Report EPA-440/4-79-029b, Vol. 2.
Jarvinen, A. W., and D. K. Tanner. 1982. Toxicity of selected controlled
i
' release and corresponding unformulated technical grade pesticides to the
fathead minnow (Pimcphales promelas) . Environ. Pollut. (Series A) 27:
179-195.
Johns, D. M., M. E. Peters, and A. 0. Beck. 1980. Nutritional value of
geographical and temporal strains of Artemia; Effects on survival and
growth of two species of Brachycerus larvae, pp. 291-304; Volume 3,
Ecology, The Brine Shrimp Artemia, Culturing, Use in Aquaculture.
McKim, J. M. 1977. Evaluation of tests with early life stages of fish for
predicting long term toxicity. J. Fieh. Res. Board Can. 34(8):
1148-1159.
Siefert, R. E. 1972. First food of larval yellow perch, white sucker,
bluegill, emerald shiner and rainbow smelt. Trans, Am. Fish. Soc. 101:
219-225.
14
-------
Appendix 1
Participating Organisations
ABI Applied Biology, lac.
AHSP Academy of Natural Sciences of
Philadelphia
ERCO Energy Resources Co.
ERL-D Environmental Research Laboratory-Ouluth
JEI Jones Edmunds, Inc.
MBL Marine Bioasaay Laboratory
WPTS Western Kish Toxicity Station
Hark Smith
Dr. Arthur Scheier
Timothy Ward
Armond Lemke
Mary Leslie
John Hansen
Dr. Alan Nebeker
15
-------
Guidelines for Conducting Flow-Through Early Life Stage Toxic ity
Tests with Fathead Minnows
for use in the USEPA, OTS-ORD Round Robin Test
1. In an Early Life Stage Toxicity Test with fathead minnows, organisms are
exposed to toxicant during part of the embryonic stage, all of the
larval stage and part of the juvenile stage. The organisms are examined
for statistically significant reductions in survival and weight in order
to determine lower and upper chronic endpoints.
A lower clironic endpoint is the highest tested concentration (a) in an
acceptable chronic test, (b) which did not cause the occurrence (which
was statistically significantly different from the control at the 952
level) of any specified adverse effect, and (c) below which no tested
concentration caused such an occurrence.
An upper chronic endpoint is the lowest tested concentration (a) in an
acceptable chronic test, (b) which caused the occurrence (which was
statistically significantly different from the control at the 95% level)
of any specified adverse effect and (c) above which all tested
concentrations caused such an occurrence.
2. Not enough information is currently available concerning early life
stage tests with fathead minnows to allow precise specification of
details for all aspects of the test. Enough such tests have been
conducted and enough aspects have been studied, however, to indicate
that these Guidelines are appropriate. A prudent course of action for
anyone planning to conduct such tests would be to initially conduct a
test with no toxicant to gain experience and to determine if the
requirements of sections 10, 11, 19, 20, 26 and 27 can be met using the
16
-------
planned water, food, procedures, etc. If a solvent may be used in the
preparation ol a stock solution, it would also be prudent to test one or
more concentrations of one or more solvents at the same time (see
Section 4). Gener il information on such things as apparatus, dilution
water, toxicant, randomization of test chambers and organisms, and
methods for chemical analyses, can be found in Draft #10 of the proposed
ASTM Standard Practice for Conducting Acute Toxicity Tests with Fishes,
Hacroinvertebrates, and Amphibians.
3. Tests should be conducted with at least five toxicant concentrations in
a geometric series and at least one control treatment. The
concentraiton of toxicant in each treatment, except for highest
concentration and the control treatment, should usually be 50 percent of
that in the next higher one.
4. If 'a solvent other than water is-used to prepare test solutions, o
solvent control (at the highest solvent concentration present in any
other treatment) using twice as many test organisms and test chambers as
the other treatments is required in addition to the regular control,
unless such a control has already been tested in the same water with the
same species of fish, food, and test procedure and the water quality has
not changed significantly. A concentration of solvent is acceptable
only if it is (or has been) shown that that concentration or a higher
one does not cause an increase or decrease in survival or weight at the
end of the test tiiat is statistically significantly different from the
control at the 95% level using a two-tailed test.
5. For each treatment (toxicant concentration and control) there must be at
least two replicate test chambers each containing one or more embryo
17
-------
•• . cups with at least 60 embryos divided equally between the embryo cups at
the beginning of the test.
6. Two test chambers have been used routinely:
a. Twenty fish have been tested in a chambei which is 16 cm x 44 cm x
Ib cm high with a 16 cm x 18 cm 40-mesh stainless steel screen 6 cm
from one end, with a water depth of 12.8 cm and with a flow rate of
190 ml/minute.
b. Fifteen fish have been tested in a chamber which is 6.5 cm x 18.0 cm
x 9.0 xm high with a 6.5 cm x 9.0 cm 40-mesh stainless steel screen
2.5 cm from one end, with a water depth of 4.5 cm and with a flow
rate of 15 ml/minute.
All of the above are inside dimensions. In both test chambers the voter
depth is controlled by a standpipe located in the smaller screened
compartment with the test solution enterin,; at the other end of the test
chamber. ; '
7. Embryo cups should be glass cylinders about 4,5 cm inside diameter and
about 7 cm high with 40-raesh nylon or stainless steel screen glued to
the bottom. Tha embryo cups must be suspended in the test chamber in
such a way as to insure that the organisms are always submerged and that
test solution regularly flows into and out of the cup without agitating
the organisms too vigorously. A rocker arm apparatus driven by a 2
r.p.m. motor and having a vertical-travel distance of 2.5 - 4.0 cm has
been successfully used, as have self-starting siphons that cause the
level of solution in the test chamber to rise and fall.
13
-------
8. Any water in which fathead minnows will survive, grow, and reproduce
satisfactorily should be an acceptable dilution water for early life
stage toxicity tests with fathead minnows. ' •
9. A 16-hr light and 8-hr dark photoperiod should be provided. A 15- to
30-rainute transition period at "lights on" and "lights off" may be
desirable. Light intensities from 10 to 100 lumens at the water surface
have been used successfully, but the intensity should be about the same
for all test chambers. Light should be provided by wide-spectrum (color
Rendering Index > 90) fluorescent lamps.
10. Tests should be conducted at 25*C. The temperature in each test chamber
should be between 24 and 26*C at all times and must be between 20 and
28*C at all times. If the water is heated, precautions should be taken
to assure that supersaturation of dissolved gjises is avoided and total
dissolved gases should be measured at least once during the test in the
water entering the control treatment.
11. The dissolved oxygen concentration should be between 75 percent and 100
percent saturation at all times in all test chambers. At no time during
the test should one test chamber have a dissolved oxygen concentration
that is more tha*« 1.1 times the dissolved oxygen concentration occurring
in another chamber at the same time.
12. The flow rate of test solution through the test chambers must be great
enough to maintain the dissolved oxygen concentration (see sections 11
and 22) and to insure that the toxicant concentrations are not decreased
significantly due to uptake by test organisms and material on the sides
and bottoms of the chambers.
19
-------
13. A test begins when embryos in embryo cups are placed in test solution
and ends 32 days later.
14. Embryos and fish should not be treated to cure or prevent disease or
fungus before or during a test. • , . •
15. Embryos should be obtained from a fathead minnow stock culture
maintained at 25*C and a dissolved oxygen concentration between 75X and
• 100X saturation with a 16-hr light and 8-hr dark photoperiod. Frozen
adult brine shrimp has been successfully used as a food for adult
fathead minnows. The maximua production of embryos by fathead minnows
has been obtained in a 30 cm x 60 cm x 30 cm deep chamber with a water
depth of IS cm when 15 cm x. 30 cm quadrants are formed with stainless
steel screen and one male, one female and one or two substrates are
placed in each quadrant. Half-round spawning substrates (Eenoit and
Carlson, 1977) with an inside diameter of 7.5 cm and a length of 7.5 cm
have been used successfully.
16. The afternoon before a test is to begin, all of the substrates should be
removed from an appropriate number of tanks in the stock culture unit
and should be replaced about the time the lights are turned on the next
morning. Enough (at least three) substrates with embryos on them should
be removed six hours later and soaked in dilution water for two hours.
For each individual substrate the embryos should be gently separated
(Cast and Brungs, 1973) and removed and visually examined using a
dissecting scope or a magnifying viewer. Empty shells and undeveloped
and opaque embryos should be discarded. If less than 50 percent of the
embryos from a substrate appear to be healthy and fertile, all the
20
-------
embryos from Chat substrate should be discarded. Single embryos with no
fungus or partial shells attached are preferable, although embryos with
partial shells attached and clumps of two or three embryos (with or
without separation) have been used successfully. An approximately equal
number of acceptable embryos from one substrate should be impartially
distributed to each embryo cup and the process repeated for at least two
more substrates until the proper number of embryos have been placed in
etch cup to give at least 60 embryos per treatment. The embryo cups
should be standing in dilution water when the embryos are being
distributed and then the cups should be randomly placed in the test
chambers.
17. Twenty to 24 hours after they are placed in the embryo cups, the embryos
should be visually examined under a dissecting scope or magnifying
viewer and all dead embryos should be counted and discarded. -Embryos
that are alive but heavily fungused should also be counted and
discarded. Forty to 48 hours after the start of the exposure all dead
and heavily fungused embryos should be -ounted and removed. The
remaining healthy, fertile embryos should be impartially reduced to the
desired number ot test organisms (at least 30 per treatment). If more
than about 35 percent of the embryos in the control treatnent are
discarded within the first 48 hours of the test because they are dead or
heavily fungused, it will probably be cost-effective to restart the
test. In eddition, if toxicant related effects are seen at 48 hours, it
will probably be cost effective to restart the test since all of the
toxicant concentrations will probably cause adverse effects. Each day
thereafter dead embryos should be counted and discarded.
21
-------
18. la each treatment, when hatching is about 90% completed or 48 hr after
first hatch in that treatment, the live young fish should be counted and
the live fish that are visibly (without the use of a dissecting scope or
magnifying viewer) lethargic or grossly abnormal in either swimming
behavior or physical appearance should be counted. All of the normal
and abnormal live fish should be released into the test chambers.
Unhatched embryos should be left in Che cups and released into the test
chamber when they hatch. The range of time-to-hatch (to the nearest
day) in each cup should be recorded.
19. A test should be terminated if the average percent of embryos (based on
the number of embryo) after thinning) that produce Jive fry for release
into test chambers ir. any control treatment is less than 50 percent or
if the percent hatch in any control embryo cup is more than 1.6 times
that in another control embryo cup,
20. The flow rate, size of the test chamber and the amount of food added
should be such that the average weight of the control fish at the end of
the test would not be significantly greater if only half as many fish
were tested per test chamber.
21. Each test chamber containing live fish over two days old must be fed
live newly hatched brine shrimp at least t«ro times a day at least 6 hrs
apart (or three times a ds; about four hours apart) on days 2-5 after
hatch and at least five days a week thereafter. They must be fed at
least once a day on all other days. Other food may also be provided in
addition to the above. The amount of food provided to each chamber may
be proportional to the nuaber and size of fish in the chamber, but each
22
-------
chamber nivef be treated in a comparable manner. Quantifying the amount
. -»•*-
' of live newly hatched brine shrimp to be fed is difficult, but the fish
should not be excessively overfed or underfed. A large buildup of food
on the bottom of the chamber is a sign of excessive overfeeding. A sign
of not feeding enough of the right kind of food is that in the sideview
the abdomen does not protrude.
•22. Test chambers should be cleaned often enough to maintain the dissolved
i
oxygen concentration (see sections 11 and 12) and to insure that the
toxicant concentrations are not decreased significantly due to sorption
by matter on the bottom and sides. In most tests if the organisms are
rot overfed too much and the flow rate is not too low, removing debris
from the bottom once or twice a week should be adequate. With some
toxicants that promote growth of bacteria the sides^and bottoms-should
be cleaned more often. Debris can be removed with a. large pipette and
rubber bulb or by siphoning into a white bucket. A dark tip on the
pipette or siphor should-help fish avoid being sucked up, but the
pipette or bucket should be examined to insure that no live fish is
discarded,
23. Temperatures should be recorded in all test chambers once at the
beginning of '.he test and once near the middle of the test. In
addition, temperature should be recorded at least hourly in one test
chamber throughout the test. The dissolved oxygen concentration should
be measured in each treatment at least once a week during the test.
Hardness, pH, alkalinity, and acidity should be measured once a week in
the control treatment and once in the.highest toxicant concentration.
23
-------
The concentration of toxicant should be measured at least twice a week
in each treatment.
24. Dead fish should be removed and recorded when observed. At a minimum
11, 18, 25 and 32 days after the beginning of the test, the live fish
should be counted and the fish that are visibly (without the use of a
dissecting scope or magnifying viewer) lethargic and grossly abnormal in'
either swimming behavior or physical appearance should be counted,
4
25. The fish should not be fed for the last 24 hours prior to termination on
day 32. At termination the weight (wet, blotted dry) of each fish that
was alive at the end of the test should be determined. If the fish
exposed to toxicant appear to b<> edematous compared to control fish,
determination of dry, rather than wet, weight is desirable.
26. A test is not acceptable if the average survival of the controls at the
end of the test is less than 80 percent or if survival in any control
chamber is less than 70 percent.
27. A test is net acceptable if the relative standard deviation (RSD ° 100
times the standard deviation divided by the mean) of the weights of the
fish that were alive at the end of the test in .any control test chamber
is greater than 40 percent.
28. Data to be statistically analyzed are:
(A) percent of healthy, fertile embryos at AO-48 hcurs
(B) percent of embryos that produce live fry for release into test
chambers
24
-------
(C) percent of embryos that produce live, normal fry for release into
tesc chambers '
(D) percent of embryos that produce live fish at end of test
(B) percent of embryos that produce live, normal fish at end of test
(P) weights of individual fish that were alive at end of test.
Although item A is based on the number of eabryos initially placed in
embryo cups, items B, C, D, aid £ are based on the number of embryos
after thinning. - ;,
29. Pichotomous data (live-dead, normal-abnormal) should be analyzed using
contingency tables (Sokei and Rohlf, Biometry, 1969, p. 587) or log
linear techniques. For weight data the individual fish are used an the
replicates unless 4 two-tailed f test indicates that differences between
replicate test chambers are not negligible. Weight data may be analyzed
using Bartlett's test and one-way analysis of variance, but Dunnott's
>
procedure (Steel and Torrie, Principles and Procedures of Statistics,
i960, p. Ill) should be used to identify treatments producing weights
that are statistically significantly lower than those of the controls at
the 95Z level.
30. Although the results of the analyses of all six types of data in section
23 should be reported, the lower and upper chronic endpoints are only
based on statistically significant reductions in survival and weight at
the end of the test (items D and F). Item A is apparently relatively
insensitive and item B is included in item D. In addition, abnormal
fish seem to weigh less than normal fish and so will be covered in item
F. Also, since the distinction between normal and abnormel is
subjective, this kind of data is expected to be less reproducible from
25
-------
one investigator to another than the other kinds of data. Although
items D and F are considered primary, the other items are included
. because they may provide useful information.
References
Benoit, D. A., and R. W. Csrleon. 1977. Spawning success of fathead minnows
on selected artificial substrates. Prog. Fish-Cult. 39: 67-69.
Flickinger, S. A. 1969. Determination of sexes in the fathead minnow.
Trans. Am. Fish. Soc. 9S: 526-527.
Cast, H. H., and W. A. Brungs. 1973, A procedure for separating eggs of the
fathead minnow. Prog. Fish-Cult. 35: 54,
May, R. C. 1970. Feeding larval marine fishes in the laboratory: A review.
Calif. Mar. Res, Comra., California Cooperative Oceanic Fisheries
Investigations Report 14: 76-83.
This tentative procedure was written by Charles Stephan with the help of many
members of the staff of the Environmental Research Laboratory in Duluth,
Minnesota.
26
-------
------- |