United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S4-89/001 Mar. 1990
x°/EPA Project Summary
Short-Term Methods for
Estimating the Chronic Toxicity
of Effluents and Receiving
Waters to Freshwater
Organisms
Second Edition
Cornelius I. Weber, William H. Peltier, Teresa J. Norberg-King, William B.
Horning, II, Florence A. Kessler, John R. Menkedick, Timothy W. Neiheisel,
Philip A. Lewis, Donald J. Klemm, Quentin H. Pickering, Ernest L. Robinson,
James M. Lazorchak, Larry J. Wymer, and Ronald W. Freyberg
This methods manual is a revision
of EPA/600/4-85/014, and describes
short-term (four- to seven-day)
methods for estimating the chronic
toxicity of effluents and receiving
waters to the fathead minnow
(Pimephales promotes), a cladoceran
(Cerlodaphnia dubia), and a green
alga (Se/enastrum caprlcornutum).
Also included are guidelines on
laboratory safety, quality assurance,
facilities and equipment, dilution
water, effluent sampling and holding,
data analysis, report preparation, and
organism culturing and handling.
Supplementary information on
statistical techniques for test design
and analysis of toxicity test data is
provided in the Appendices.
A supplement to the report was
published In September 1989
(EPA/600/4-89/001 a), to provide an
additional method (Linear Inter-
polation Method) for the analysis of
data from the Fathead Minnow Larval
Survival and Growth Test and the
Ceriodaphnia Survival and Repro-
duction Test. This supplement
consists of 42 pages arranged in four
parts to facilitate insertion In the
appropriate places in the existing
report
This Project Summary was devel-
oped by EPA's Environmental
Monitoring Systems Laboratory,
Cincinnati, OH, to announce key
findings of the research project that Is
fully documented In a separate report
of the same title (see Project Report
ordering Information at back).
Introduction
As a result of the increased awareness
of the value of effluent toxicity test data
for toxics control in the water quality
program and the National Pollutant
Discharge Elimination System (NPDES)
permit program, which emerged from the
extensive effluent toxicity monitoring
activities of the regions and states, and
the availability of short-term chronic
toxicity test methods, the U. S.
Environmental Protection Agency
(USEPA) issued a national policy
statement entitled, "Policy for the
Development of Water Quality-Based
-------
Permit Limitations for Toxic Pollutants,"
in the Federal Register, Vol. 49, No. 48,
p. 9016-9019, Friday, March 9, 1984.
This policy proposed the use of toxicity
data to assess and control the discharge
of toxic substances to the Nation's waters
through the NPDES permits program.
The policy states that "biological testing
of effluents is an important aspect of the
water quality-based approach for
controlling toxic pollutants. Effluent
toxicity data, in conjunction with other
data, can be used to establish control
priorities, assess compliance with State
water quality standards, and set permit
limitations to achieve those standards."
All states have water quality standards
which include narrative statements
prohibiting the discharge of toxic
materials in toxic amounts.
Objective
The four short-term tests described in
this manual are for use in the NPDES
Program to estimate one or more of the
following: (1) the chronic toxicity of
effluents collected at the end of the
discharge pipe and tested with a
standard dilution water; (2) the chronic
toxicity of effluents collected at the end of
the discharge pipe and tested with
dilution water consisting of non-toxic
receiving water collected upstream from
or outside the influence of the outfall, or
with other uncontaminated surface water
or standard dilution water having
approximately the same hardness as the
receiving water; (3) the toxicity of
receiving water downstream from or
within the influence of the outfall; and (4)
the effects of multiple discharges on the
quality of the receiving water. The tests
may also be useful in developing site-
specific water quality criteria
These methods were developed to
provide the most favorable cost-benefit
relationship possible, and are intended
for use in effluent toxicity tests performed
on-site or off-site.
The tests include:
1. A seven-day, sub-chronic, fat-
head minnow (Pimephales
promelas), static renewal, larval
survival and growth test.
2. A three-brood, seven-day,
chronic, cladoceran (Cerio-
daphnia dubia), static renewal,
survival and reproduction test.
3. A seven-day, sub-chronic, fat-
head minnow (Pimephales
promelas), static renewal,
embryo-larval survival and
teratogenicity test.
4. A four-day, chronic, algal,
(Selenastrum capricornutum),
static, growth test.
Short-Term Methods for
Estimating Chronic Toxicity
The objective of chronic aquatic
toxicity tests with effluents and pure
compounds is to estimate the highest
"safe" or "no-effect concentration" of
these substances. For practical reasons,
the parameters observed in these tests
are usually limited to hatchability, gross
morphological abnormalities, survival,
growth, and reproduction, and the results
of the tests are usually expressed in
terms of the highest toxicant concentra-
tion that has no statistically significant
observed effect on these parameters,
when compared to the controls. The
terms currently used to define the
endpoints employed in the rapid, chronic
and sub-chronic toxicity tests have been
derived from the terms previously used
for full life-cycle tests. As shorter chronic
tests were developed, it became
common practice to apply the same
terminology to the endpoints. The
primary terms in current use are as
follows:
Safe Concentration
The highest concentration of toxicant
that will permit normal propagation of fish
and other aquatic life in receiving waters.
The concept of a "safe concentration" is
a biological concept, whereas the "no-
observed-effect concentration" (below) is
a statistically defined concentration.
Ato-Ofcservecf-Effect-
Concentration (NOEC)
The highest concentration of toxicant to
which organisms are exposed in a full
life-cycle or partial life-cycle test, that
causes no observable adverse effects on
the test organisms (i.e., the highest con-
centration of toxicant in which the values
for the observed parameters are not
statistically significantly different from the
controls). This value is used, along with
other factors, to determine toxicity limits
in permits.
Lowest-Observed-Effect-
Concentration (LOEC)
The lowest concentration of toxicant to
which organisms are exposed in a life-
cycle or partial life-cycle test, which
causes adverse effects on the i
organisms (i.e., where the values for
observed parameters are statistic
significantly different from the controls
Effective Concentration (EC)
A point estimate of the toxicant <
centration that would cause
observable adverse effect (such as de
immobilization, serious incapacitat
reduced fecundity, or reduced growth
a given percent of the test organis
calculated by point estimat
techniques. For example, the EC50 f
a Probit Analysis is the estimated (
centration of toxicant that would ca
death, or some other observable quai
"all or nothing," response, in 50% of
test population. If the observable el
is death (mortality), the term LC - Le
Concentration, is used (see below). If
observable effect is a non-qua
biological measurement, the te
Inhibition Concentration (1C), may
used (see below). A certain EC, LC
1C value might be judged fron
biological standpoint to represen
threshold concentration, or lowest c
centration that would cause an advi
effect on the observed parameters.
Lethal Concentration (LC)
Identical to EC when the observ;
adverse effect is death or mortality.
Inhibition Concentration (1C)
A point estimate of the toxicant <
centration that would cause a gi
percent reduction in a non-qua
biological measurement such
fecundity or growth. For example,
IC25 would be the estimated concei
tion of toxicant that would cause a J
reduction in mean young per femali
some other non-quantal biolog
measurement.
If the objective of chronic aqu
toxicity tests with effluents and |
compounds is to estimate the higl
"safe or no-effect concentration" of tt
substances, it is imperative to unders
how the statistical endpoint of these 1
is related to the "safe" or "no-eff
concentration. NOECs and LOECs
determined by hypothesis testing,
LCs, ECs, and ICs are determined
point estimation techniques. There
inherent differences between the us
an NOEC, LOEC, or other estin
derived from hypothesis testinc
estimate a "safe" concentration, and
use of a LC, EC, 1C, or other (
estimate derived from curve fitl
interpolation, etc.
-------
Most point estimates, such as the LC,
"C, or 1C are derived from a
athematical model that assumes a
:ontinuous dose-response relationship.
5y definition, any LC, EC, or 1C value is
in estimate of some amount of adverse
iffect. Thus the assessment of a safe
oncentration must be made from a
ilological standpoint. In this instance,
ie biologist must determine some
mount of adverse effect that is deemed
5 be "safe," in the sense that it will not
•om a practical biological viewpoint,
ffect the normal propagation of fish and
ther aquatic life in receiving waters.
'hus, to use a point estimate such as an
C, EC, 1C to determine a "safe" concen-
ation requires a biological judgment of
'hat constitutes an acceptable level of
dverse effect.
The use of NOECs and LOECs, on the
ther hand, assumes either (1) a
ontinuous dose-response relationship, or
>) a noncontinuous threshold model of
ie dose-response relationship.
In the first case, it is also assumed that
dverse effects that are not "statistically
bservable" are also not significant from
biological standpoint, since they are not
renounced enough to test statistically
ignificant against some measure of the
atural variability of responses.
In the second case, it is assumed that
are exists a true threshold, or concen-
•ation below which there is no adverse
ffect on aquatic life, and above which
lere is an adverse effect. The purpose
if the statistical analysis in this case is to
estimate as closely as possible where
iat threshold lies.
In either case, it is important to realize
iat the amount of the adverse effect that
statistically observable (LOEC) or not
ibservable (NOEC) is highly dependent
in all aspects of the experimental design.
"hese aspects include the choice of
tatistical analysis, the choice of an alpha
9vel, and the amount of variability
>etween responses at a given concentra-
ion. The sensitivity of the test, which is
elated to the magnitude of the adverse
iffect that is statistically observable, can
je controlled by the experimental design
ind by controlling the amount of
variability between responses at the
jiven concentration.
In the first case, where the assumption
3f a continuous dose-response relation-
hip is made, clearly the NOEC estimate
s an estimate of some amount of
adverse effect that is dependent on the
experimental design. In the second case,
the NOEC may be an estimate of a
ife" or "no-effect" concentration but
y if the amount of adverse effect that
appears at the threshold is great enough
to test as statistically significantly differ-
ent from the controls in 'the face of all
aspects of the experimental design
mentioned above. The NOEC in that
case would indeed be an estimate of a
"safe" or "no-effect" concentration. If,
however, the amount of adverse effect
were not great enough to test as
statistically different, then the NOEC
might well be an estimate that again
represents some amount of adverse
effect which is assumed safe because it
did not test as statistically significant. In
any case, the estimate of the NOEC with
hypothesis testing is always dependent
on the aspects of the experimental
design mentioned above. For this
reason, the reporting and examination of
some measure of the sensitivity of the
test (either the minimum significant
difference or the percent change from the
control that this minimum difference
represents) is extremely important.
In summary, the assessment of a
"safe" or "no-effect" concentration
cannot be made from the results of
statistical analysis alone, unless (1) the
assumptions of a strict threshold model
are accepted, and (2) it is assumed that
the amount of adverse effect present at
the threshold is statistically detectable by
hypothesis testing. In this case,
estimates obtained from a statistical
analysis are indeed estimates of a "no-
effect" concentration. If the assumptions
are not deemed tenable, then estimates
from a statistical analysis can only be
used in conjunction with an assessment
from a biological standpoint of what
magnitude of adverse effect constitutes a
"safe" concentration. In this instance, a
"safe" concentration is not necessarily a
"no-effect" concentration, but rather a
concentration at which the effects are
judged to be of no biological significance.
Health and Safety
Collection and use of effluents in
toxicity tests may involve significant risks
to personal safety and health. Personnel
collecting effluent samples and conduct-
ing toxicity tests should take all safety
precautions necessary for the prevention
of bodily injury and illness which might
result from ingestion or invasion of infec-
tious agents, inhalation or absorption of
corrosive or toxic substances through
skin contact, and asphyxiation due to lack
of oxygen or presence of noxious gases.
Prior to sample collection and
laboratory work, personnel will determine
that all necessary safety equipment and
materials have been obtained and are in
good condition.
Quality Assurance
Quality Assurance (QA) practices for
effluent toxicity tests consist of all
aspects of the test that affect data quality,
such as: (1) effluent sampling and
handling; (2) the source and condition of
the test organisms; (3) condition of
equipment; (4) test conditions; (5) instru-
ment calibration; (6) replication; (7) use of
reference toxicants; (8) record keeping;
and (9) data evaluation.
Dilution Water
The source of dilution water used in
effluent toxicity tests will depend largely
on the objectives of the study:
1. If the objective of the test is to
estimate the inherent chronic
toxicity of the effluent, which is
the primary objective of NPDES
permit-related toxicity testing, a
standard dilution water
(moderately hard water) is used.
2. If the objective of the test is to
estimate the chronic toxicity of
the effluent in uncontaminated
receiving water, the test may be
conducted using dilution water
consisting of a single grab
sample of receiving water (if
non-toxic), collected upstream
and outside the influence of the
the outfall, or with other
uncontaminated surface water or
standard dilution water having
approximately the same
characteristics (pH, hardness,
alkalinity, conductivity, and total
suspended solids) as the
receiving water. Seasonal
variations in the quality of
surface waters may affect
effluent toxicity. Therefore, the
pH, alkalinity, hardness, and
conductivity of receiving water
samples should be determined
before each use.
3. If the objective of the test is to
determine the additive effects of
the discharge on already
contaminated receiving water,
the test is performed using
dilution water consisting of
receiving water collected
upstream from the outfall.
Effluent and Receiving Water
Sampling and Handling
The effluent sampling point usually
should be the same as that specified in
-------
the NPDES discharge permit. Conditions
for exception would be: (1) better access
to a sampling point between the final
treatment and the discharge outfall; (2) if
the processed waste is chlorinated prior
to discharge to the receiving waters, it
may also be desirable to take samples
prior to contact with the chlorine to
determine toxicity of the unchlorinated
effluent; or (3) in the event there is a
desire to evaluate the toxicity of the
influent to municipal waste treatment
plants or separate wastewater streams in
industrial facilities prior to their being
combined with other wastewater streams
or non-contact cooling water, additional
sampling points may be chosen.
The decision on whether to collect grab
or composite samples is based on the
objectives of the test and an
understanding of the short and long-term
operations and schedules of the
discharger. If the effluent quality varies
considerably with time, which can occur
where holding times are short, grab
samples may seem preferable because
of the ease of collection and the potential
of observing peaks (spikes) in toxicity.
However, the sampling duration of a grab
sample is so short that full character-
ization of an effluent over a 24-h period
would require a prohibitive number of
separate samples and tests. Collection
of a 24-h composite sample, however,
may dilute toxicity spikes, and average
the quality of the effluent over the
sampling period. Sampling recommend-
ations are provided for grab and
composite samples.
Sample Handling and
Preservation and Shipping
If the data from the samples are to be
acceptable for use in the NPDES
Program, the lapsed time from collection
of a grab or composite sample and its
first use for initiation of a test, or for test
solution renewal, should not exceed 36 h.
Composite samples should be chilled
during collection, where possible, and
maintained at 4°C until used. Samples
collected for on-site tests should be used
within 24 h. Samples collected for off-
site toxicity testing are to be chilled to
4°C when collected, shipped iced to the
central laboratory, and there transferred
to a refrigerator (4°C) until used. Every
effort must be made to initiate the test
with an effluent sample on the day of
arrival in the laboratory.
Samples may be shipped in 4-L (1-gal)
CUBITAINERS" or new plastic "milk"
jugs. All sample containers should be
rinsed with source water before being
filled with sample. After use,
CUBITAINERSR and plastic jugs are
punctured to prevent reuse. Several
sample shipping options are available,
including Express Mail, air express, bus,
and courier service. Express Mail is
delivered seven days a week. Shipping
and receiving schedules of private
carriers on weekends vary with the
carrier.
Sample Preparation
With the Ceriodaphnia and fathead
minnow tests, effluents and surface
waters must be filtered through a 60-u.m
plankton net to remove indigenous
organisms that may attack or be
confused with the test organisms (see
Ceriodaphnia test method for details).
Surface waters used in algal toxicity tests
must be filtered through a 0.45-nm pore
diameter filter before use. It may be
necessary to first coarse-filter the dilution
and/or waste water through a nylon sieve
having 2- to 4-mm holes to remove
debris and/or break up large floating or
suspended solids. Caution: filtration may
remove toxicity.
The DO concentration in the dilution
water should be near saturation prior to
use. Aeration will bring the DO and other
gases into equilibrium with air, minimize
oxygen demand, and stabilize the pH.
If the dilution water and effluent must
be warmed to bring them to the
prescribed test temperature,
supersaturation of the dissolved gases
may become a problem. To prevent this
problem, the effluent and dilution water
are checked for dissolved oxygen (DO)
with a probe after heating to 25°C. If the
DO is greater than 100% saturation or
lower than 40% saturation, the solutions
are aerated moderately with a pipet tip
for a few minutes until the DO is within
the prescribed range.
Data Analysis
The choice of a statistical method to
analyze toxicity test data and the
interpretation of the results of the
analysis of the data from any of the
toxicity tests described in this manual
can become problematic because of the
inherent variability and sometimes
unavoidable anomalies in biological data.
Analysts who are not proficient in
statistics are strongly advised to seek the
assistance of a statistician before
selecting the method of analysis and
using any of the results.
The recommended statistical methods
presented in the manual are not the only
possible methods of statistical analysis.
Many other methods have be
proposed and considered. Amo
alternative hypothesis tests some, I
Williams' Test, require additional i
sumptions, while others, like the bo
strap methods, require computer-
tensive computations. Alternative pc
estimation approaches most probal
would require the services of
statistician to determine the a
propriateness of the model (goodness
fit), higher order linear or nonlim
models, confidence intervals
estimates generated by inver
regression, etc. In addition, po
estimation or regression approach
would require the specification
biologists or toxicologists of some I
level of adverse effect that would
deemed acceptable or safe. Certai
there are other reasonable and defensi!
methods of statistical analysis of this k
of toxicity data. The methods contair
in this manual have been chosen, amc
other reasons, because they are (1) w
tested and well-documented,
applicable to most different toxicity t
data sets for which they a
recommended, but still powerful,
hopefully "easily" understood by n<
statisticians, and (4) amenable to i
without a computer, if necessary.
The data should be plotted, both a:
preliminary step to help detect proble
and unsuspected trends or patterns in i
responses, and as an aid in interpretat
of the results. Further discussion j
plotted sets of data are included in
methods and the Append
Transformations of the data, e.g., arc s
square root and logs, are used wh<
necessary to meet assumptions of 1
proposed analyses, such as t
requirement for normally distributed da
Statistical independence amo
observations is a critical assumption
the statistical analysis of toxicity de
One of the best ways to ensi
independence is to properly foil
rigorous randomization procedun
Randomization techniques should
employed at the start of the te
including the randomization of the pla
ment of test organisms in the test cha
bers and randomization of the test cha
ber location within the array of chambt
A discussion of statistical independen
outliers and randomization, and a sam
randomization scheme, are included
Appendix A.
The number of replicates employed
each toxicant concentration is
important factor in determining 1
sensitivity of chronic toxicity tests. T
sensitivity generally increases as
-------
number of replicates is increased, but the
point of diminishing returns in sensitivity
may be reached rather quickly. The level
of sensitivity required by a hypothesis
test or the confidence interval for a point
estimate will determine the number of
replicates, and should be based on the
objectives for obtaining the toxicity data.
In a statistical analysis of toxicity data,
the choice of a particular analysis and the
ability to detect departures from the
assumptions of the analysis, such as the
normal distribution of the data and
homogeneity of variance, is also
dependent on the number of replicates.
More than the minimum number of
replicates may be required in situations
where it is imperative to obtain optimal
statistical results, such as with tests used
in enforcement cases or when it is not
possible to repeat the tests. For
example, when the data are analyzed by
hypothesis testing, the nonparametric
alternatives cannot be used unless there
are at least four replicates at each
toxicant concentration. If there are only
two replicates, Dunnett's Procedure may
be used, but it is not possible to check
the assumptions of the test.
The recommended statistical analysis
of most data from chronic toxicity tests
with aquatic organisms follows a decision
process illustrated in a flow chart. An
initial decision is made to use point
estimation techniques and/or to use
hypothesis testing. If hypothesis testing
is chosen, subsequent decisions are
made on the appropriate hypothesis
testing procedure for a given set of data,
as illustrated in the flow chart. If point
estimation is chosen, the equivalent of an
NOEC can be calculated. A specific flow
chart is included in the analysis section
for each test.
Since a single chronic toxicity test
might yield information on more than one
parameter (such as survival, growth, and
reproduction), the lowest estimate of a
"no-observed-effect concentration" for
any of the parameters would be used as
the "no-observed-effect concentration"
for each test. It follows logically that in
the statistical analysis of the data, con-
centrations that had a significant toxic
effect on one of the observed parameters
would not be subsequently tested for an
effect on some other parameter. This is
one reason for excluding concentrations
that have shown a statistically significant
reduction in survival from a subsequent
statistical analysis for effects on another
parameter such as reproduction. A
second reason is that the exclusion of
such concentrations usually results in a
more powerful and appropriate statistical
analysis.
Analysis of Growth and
Reproduction Data
Growth data from the fathead minnow
larval survival and growth test are
analyzed using hypothesis testing or
point estimation techniques. (Note that
the nonparametric hypothesis tests can
be used only if at least four replicates
were used at each toxicant concentra-
tion.)
Reproduction data from the
Ceriodaphnia survival and reproduction
test, after eliminating data from concen-
trations with a significant mortality effect
as determined by Fisher's Exact Test,
are analyzed using hypothesis testing or
point estimation techniques. (Note that
the nonparametric hypothesis tests can
be used only if at least four replicates
were used at each toxicant concentra-
tion).
Analysis of Algal Growth
Response Data
The growth response data from the
algal toxicity test, after an appropriate
transformation if necessary to meet the
assumptions of normality and
homogeneity of variance, may be
analyzed by hypothesis testing. Point
estimates, such as the EC1, ECS, EC10,
or EC50, would also be appropriate in
analyzing algal growth data.
Analysis of Mortality Data
Mortality data from the fathead minnow
larval survival and growth test and the
fathead minnow embryo-larval survival
and teratogenicity test are analyzed by
Probit Analysis, if appropriate. The
mortality data can also be analyzed by
hypothesis testing, after an arc sine
transformation.
Mortality data from the Ceriodaphnia
survival and reproduction test are
analyzed by Fisher's Exact Test prior to
the analysis of the reproduction data.
The mortality data may also be analyzed
by Probit Analysis, if appropriate.
Dunnett's Procedure
Dunnett's Procedure consists of an
analysis of variance (ANOVA) to
determine the error term, which is then
used in a multiple comparison method for
comparing each of the treatment means
with the control mean, in a series of
paired tests. Use of Dunnett's Procedure
requires at least two replicates per
treatment and an equal number of data
points (replicates) for each concentration.
However, as stated above, it is not
possible to check the assumptions of the
test. In cases where the number of data
points for each concentration are not
equal, a t test may be performed with
Bonferroni's adjustment for multiple
comparisons, instead of using Dunnett's
Procedure.
The assumptions upon which the use
of Dunnett's Procedure is contingent are
that the observations within treatments
are independent and normally distributed,
with homogeneity of variance. Before
analyzing the data, the assumptions must
be verified using the procedures provided
in Appendix B.
Some indication of the sensitivity of the
analysis should be provided by
calculating: (1) the minimum difference
between means that can be detected as
statistically significant, and (2) the
percent change from the control mean
that this minimum difference represents
for a given test.
The estimate of the safe concentration
derived from this test is reported in terms
of the NOEC. A step-by-step example of
Dunnett's Procedure is provided in the
Appendix.
If, after suitable transformations have
been carried out, the normality as-
sumptions have not been met, Steel's
Many-One Rank Test should be used if
there are four or more data points per
toxicant concentration. If the numbers of
data points (replicates) for each toxicant
concentration are not equal, the Wilcoxon
Rank Sum Test with Bonferroni's
adjustment should be used.
Bonferroni's T-Test
Bonferroni's T-test is used as an
alternative to Dunnett's Procedure when
the number of replicates is not the same
for all concentrations. This test sets an
upper bound of alpha on the overall error
rate, in contrast to Dunnett's Procedure,
for which the overall error rate is fixed at
alpha. Thus Dunnett's Procedure is a
more powerful test.
Steel's Many-One Rank Test
Steel's Many-One Rank Test is a
multiple comparison method for
comparing several treatments with a
control. This method is similar to
Dunnett's Procedure, except that it is not
necessary to meet the assumption for
normality. The data are ranked, and the
analysis is performed on the ranks rather
than on the data themselves. If the data
are normally or nearly normally
distributed, Dunnett's Procedure would
-------
be more sensitive (would detect smaller
differences between the treatments and
control). For data that are not normally
distributed, Steel's Many-One Rank Test
can be much more efficient. It is
necessary to have at least four replicates
per toxicant concentration to use Steel's
test. The sensitivity of this test cannot be
stated in terms of the minimum
difference between treatment means and
the control mean.
The estimate of the safe concentration
is reported as the NOEC. A step-by-step
example of Steel's Many-One Rank Test
is provided in the Appendix.
Wilcoxon Rank Sum Test
The Wilcoxon Rank Sum Test is a
nonparametric test for comparing a
treatment with a control. The data are
ranked and the analysis proceeds exactly
as in Steel's Test except that Bonferroni's
adjustment for multiple comparisons is
used instead of Steel's tables. When
Steel's test can be used (i. e., when there
are equal numbers of data points per
toxicant concentration), it will be more
powerful (able to detect smaller
differences as statistically significant)
than the Wilcoxon Rank Sum Test with
Bonferroni's adjustment.
The estimate of the safe concentration
is reported as the NOEC. A step-by-step
example of the use of the Wilcoxon Rank
Sum Test is provided in the Appendix.
Interpolation Approach
Chronic toxicity test data can be
analyzed by an interpolation approach.
Precision estimates can be calculated
using this approach. The round robin
data show that the endpoints estimated
by this approach are much less variable
than those estimated by hypothesis
testing.
Probit Analysis
Probit Analysis is used to analyze
percentage data from concentration-
response tests. The analysis can provide
an estimate of the concentration of
toxicant affecting a given percent of the
test organisms and provide a confidence
interval for the estimate. Probit Analysis
assumes a normal distribution of log
tolerances and independence of the
individual responses. To use Probit
Analysis, at least two partial mortalities
must be obtained. If a test results in
100% survival and 100% mortality in
adjacent treatments (all or nothing effect),
an LC50 may be estimated using the
graphical method, and the LC50 and
confidence interval may be estimated by
the moving average angle, Spearman-
Karber, or other methods.
It is important to check the results of
Probit Analysis to determine if the
analysis is appropriate. The chi-square
test for heterogeneity provides one good
test of appropriateness of the analysis. In
cases where there is a significant chi-
square statistic, where there appears to
be systematic deviation from the model,
or where there are few data in the
neighborhood of the point to be
estimated, Probit results should be used
with extreme caution.
The natural rate of occurrence of a
measured response, such as mortality in
the test organisms (referred to as the
natural spontaneous response), may be
used to adjust the results of the Probit
Analysis if such a rate is judged to be
different from zero. If a reliable,
consistent estimate of the natural
spontaneous response can be
determined from historical data, the
historical occurrence rate may be used to
make the adjustment. In cases where
historical data are lacking, the
spontaneous occurrence rate should
optimally be estimated from all the data
as part of the maximum likelihood
procedure. However, this can require
sophisticated computer software. An
acceptable alternative is to estimate the
natural occurrence rate from the
occurrence rate in the controls. In this
instance, greater than normal replication
in the controls would be beneficial.
A discussion of Probit Analysis and the
natural occurrence rate, along with a
computer program for performing the
Probit Analysis, are included in Ap-
pendix I.
Summary of Test Methods
1. Fathead minnows, Pimephales
promelas, larvae are exposed in
a static renewal system for
seven days to different concen-
trations of effluent or to receiving
water. Test results are based on
the survival and growth (increase
in weight) of the larvae,
compared to the controls.
2. Fathead minnows, Pimephales
promelas, embryos and larvae
are exposed to different concen-
trations of effluent or to receiving
water in a static renewal system
for seven days, starting shortlv
after fertilization of the eggs
Test results are based on the
total frequency of both mortality
and gross morphologica
deformities (terata), compared tc
the controls.
3. Cladocera, Ceriodaphnia dubia
are exposed in a static renews
system to different concentra
tions of effluent, or to receivinc
water, until 60% of surviving
control organisms have thre<
broods of offspring. Test result!
are based on survival ant
reproduction. If the test i:
conducted as described, th<
control organisms shouh
produce three broods of youn<
during a seven-day period
compared to the controls.
4. The fresh water alga
Selenastrum capricornutum, i
exposed in a static system to
series of concentrations c
effluent, or to receiving water, fc
96 h. The response of th
population is measured in term
of changes in cell density (ce
counts per mL), biomas:
chlorophyll content, o
absorbance, compared to th
controls.
-------
The EPA authors, Cornelius I. Weber (also the EPA Project Officer, see below),
William B. Horning, II, Timothy W. Neiheisel, Philip A. Lewis, Donald J.
Klemm, Quentin H. Pickering, Ernest L Robinson, and James M. Lazorchak
are with Environmental Monitoring Systems Laboratory, Cincinnati, OH 45268;
William H. Peltier is with Region IV, Atlanta Georgia 30308; Teresa J. Norberg-
King is with Environmental Research Laboratory, Duluth, MN 55804; and
Florence A. Kessler, John R. Menkedick, Larry J. Wymer, and Ronald W.
Freyberg are with Computer Sciences Corporation, Cincinnati, OH 45268.
The complete report, entitled "Short-Term Methods for Estimating the Chronic
Toxicity of Effluents and Receiving Waters to Freshwater Organisms—Second
Edition," (Order No. PB 89-207 013/AS; Cost: $31.00, subject to change); and a
supplement to the report (Order No. PB 90-145 7641 AS; Cost $15.00) will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
Copies are also available (at $13.00, Stock #055-00-00288-6) from:
Superintendent of Documents
U.S. Government Printing Office
Washington, DC 20402
Telephone: 202-783-3238
The EPA Project Officer can be contacted at:
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S4-89/001
CHICAGO
------- |