EPA/600/R-92/091
June 1992
PB92-169655
STATISTICAL APPROACH TO PREDICTING CHRONIC TOXICITY OF
CHEMICALS TO FISHES FROM ACUTE TOXICITY TEST DATA
by
Foster L. Mayer
U.S. Environmental Protection Agency
Environmental Research Laboratory
Sabine Island
Gulf Breeze, FL 32561
Gary F. Krause
Mark R. Ellersieck
Gunhee Lee
University of Missouri
Agricultural Experimental Station
105 Math Sciences Building
Columbia, MO 65211
U.S. ENVIRONMENTAL PROTECTION AGENCY
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH & DEVELOPMENT
GULF BREEZE, FL 32561
-------
DISCLAIMER
This document is intended for internal Agency use only.
Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
ii
-------
PREFACE
The Office of Research and Development initiated a multi-
laboratory Ecological Risk Assessment Research Program in 1986 to
develop scientifically defensible methods for use by the Office of
Pesticides and Toxic Substances (OPTS) in assessing ecological
risks. The Ecological Risk Assessment Research Program provides
the technical basis to improve Agency risk assessments for
chemicals in view of the Agency's interest in protecting ecological
resources and the OPTS state of the practice in conducting
ecological risk assessments. Many research needs remain, and the
demands on OPTS to consider risks to ecological resources in
chemical regulation will continue to grow.
The area of ecological risk assessment described in this
report involves a major advancement in predictive toxicology. For
the last 20 years, we have continued to use and refine various
acute-chronic ratios and correlation analyses of acute (LC50s) and
chronic data (maximum acceptable toxicant concentrations) to
estimate chronic toxicity from acute data. Until this research was
conducted, no accurate method for truly predicting, and not
estimating, chronic toxicity existed.
A technically defensible concept and methodology are described
wherein simultaneous consideration is given to exposure, degree of
response, and time course of effect, all of which are usually
included in the results of an acute test, but seldom used. The
predictive technique may reduce chronic testing requirements and
will be highly beneficial in initial chronic assessments of
chemicals and effluents and in predicting chronic toxicity for
species difficult to culture, including those that are rare and
endangered.
iii
-------
ABSTRACT
A comprehensive approach to predicting chronic toxicity from
acute toxicity data was developed in which simultaneous
consideration is given to concentration, degree of response, and
time course of effect. A consistent endpoint (lethality) and
degree of response (0%) were used to compare acute and chronic
tests. Predicted no-effect concentrations were highly accurate 92%
of the time (within a factor of 2.0 of the limits of the maximum
acceptable toxicant concentrations for lethality) and did not vary
by more than a factor of 4.8 when the technique was applied to a
data base of 18 chemicals and 7 fish species. Growth effects can
be predicted from chronic lethality, but reproductive effects
should not be.
iv
-------
TABLE OF CONTENTS
Page
Preface iii
Abstract iv
Introduction 1
Methods 3
Degree of Response 3
Time Course of Effect 3
Technique 4
Data Base 6
Results and Discussion 9
Relation to Other Endpoints 12
Acknowledgements 14
References 15
Appendixes
A. Multifactor Probit Analysis Program 27
B. Acute and Chronic Toxicity Data Base 63
v
-------
TABLES
Number Page
1 Comparisons of observed maximum acceptable
toxicant concentrations (MATC) and predicted
no effect concentrations (PNEC) for lethality
based on flow-through acute tests 20
2 Correlations of survival and growth 2 2
FIGURES
1 Dose-response curves used to derive the LCO (0.01%)
for various observation times in acute toxicity
tests (1.281 = a+bX). Probit % mortality:
1.281 = 0.01%, 5.000 = 50%, and 8.719 = 9.99 23
2 Prediction of the chronic no-effect value for
lethality from acute toxicity test data with Kepone
and fathead minnows (LC50 = a+b [l/t]). The
intercept (a) represents the LCO (2.9 ~/jg/L) over an
indefinite exposure time, and the maximum acceptable
toxicant concentration (MATC) for chronic lethality
was between 1.2 and 3.1 nq/L 24
3 Acute flow-through toxicity test results with
carbon tetrachloride and sheepshead minnows
demonstrating dose-response data in time
(96-h LC50 = 19 mg/L) 25
vi
-------
INTRODUCTION
Using acute lethality data to estimate chronic toxicity to
fishes customarily involves deriving an application factor (Mount
and Stephan 1967) or an acute-chronic ratio (Kenaga 1982), both of
which require acute-and chronic toxicity testing. Kenaga (1979)
reviewed the principal measurements of the acute LC50, the maximum
*
acceptable toxicant concentration (MATC), and the application
factor (AF) used in determining chronic no-effect concentrations
for many chemicals. The application factor is derived by dividing
the MATC for a compound, as determined in a chronic toxicity test
with a given fish species, by the acute flow-through LC50 for the
same compound tested with the same species. The acute-chronic
ratio (ACR) is the inverse of AF. The AF or ACR is then used to
estimate chronic no-effect concentrations for other species for
which only acute toxicity data exist (Buikema et al. 1982). Both
approaches have limitations in using these ratios to estimate
chronic toxicity.
One limitation is that the biological endpoints and degrees of
response are often not comparable between acute and chronic
toxicity data. When one uses either the AF or ACR, the acute
median lethal concentration (LC50) is compared with the MATC, often
derived from an endpoint other than lethality. Even though the
mode of action for lethality is often assumed to be the same under
acute and chronic exposures, the mode of action may not be the same
for different endpoints (e.g., growth or reproduction compared with
lethality). Although different degrees of response (acute 50%
1
-------
versus chronic no-effect or 0%) could be used when response slopes
are similar, the slopes may be different. Additionally, the use of
the AF or ACR method does not take into consideration the
progression of lethality through time that is observed in acute
toxicity tests. The concentration-time-response interaction has
been addressed by Shirazi and Lowrie (1988), but they directed
their efforts toward better defining the Lt50. The acute toxicity
value represents only one point in time (96-h LC50) , and the
progression of degree of response with duration of exposure should
be essential when one predicts chronic toxicity from acute toxicity
data.
A more comprehensive, alternative approach is proposed here in
which simultaneous consideration is given to concentration, degree
of response, and time course of effect, all of which are usually
included in the results of an acute test, but seldom used. A
consistent endpoint (lethality) and degree of response (0%) are
used to predict chronic lethality from acute toxicity tests. Two
assumptions may be required: (1) concentration-response is a
continuum in time, and (2) the mode of action for lethality is
similar under acute and chronic exposures.
2
-------
METHODS
Simple linear regression (Y = a+bX) was used to derive lethal
concentrations of no effect (LCO = 0.01%) for each observation time
in an acute toxicity test and to predict the chronic no-effect
concentration for lethality from those LCO's.
Degree of Response
In chronic toxicity tests, we are most often interested in the
no-effect concentration (e.g., that concentration causing 0%
effect), whereas in acute tests, the degree of response usually
used is 50%. Although a probit value (Finney 1971) does not exist
for 0% or 100%, an approximate value can be derived. In the use of
probit analysis of acute toxicity data (Finney 1971, Litchfield and
Wilcoxon 1949) , the probit value used for 100% mortality is
actually the probit value for 99.99%. An approximate value for LCO
can thus be derived by subtracting the probit value for 99.99%
(8.7190) from 10 to provide a probit value of 1.2810 for 0.01%
mortality.
Time Course of Effect
Predicting chronic toxicity from acute toxicity data requires
a means of estimating the LCO for an indefinite period of time
(chronic) from an acute toxicity test conducted over a finite
period of time (96-h LC50). Green (1965) and Sprague (1969)
provided approaches to the problem of estimating tolerance over an
indefinite time period, although it was with the LD50 or LC50.
They noted that as the time of exposure becomes sufficiently long,
the LD50 or LC50 approaches an asymptotic value. Green (1965)
3
-------
suggested using a hyperbola to describe the relationship. A
hyperbola can be expressed as a straight line by using the
reciprocal of time (t) as the independent variable. The equation
becomes LD50 = a+b(l/t). Since 1/t approaches zero as t approaches
infinity, the intercept (a) represents the LD50 over an indefinite
time of exposure. The proposed method for estimating LCO makes use
of Green's approach to predict chronic toxicity from acute toxicity
data.
Technique
The acute toxicity test must be conducted with strict
adherence to standard test methods (Committee on Methods for
Toxicity and Tests with Aquatic Organisms 1975, American Society
for Testing and Materials 1980) to obtain estimates of LCO over
time. The times of 24, 48, 72, and 96 hours were selected because
observations in standard acute toxicity tests are usually made at
these time periods. Less than 24-h observations were used when
available. Inclusion of these observations is very important when
most toxicity occurs during the early part of a 96-h test. The
greatest concentration that causes no mortality and the least
concentration that causes complete mortality were the
concentrations used for 0% and 100% responses. All concentrations
causing mortality (0%
-------
if a concentration-response is evident in time.
Linear regression analysis (Snedecor and Cochran 1980) was
used to calculate the estimated LCO at all observation times from
acute flow-through tests (Fig. 1) as probit % mortality = a+b(log
concentration). The LCO' s at each time period were then regressed
against the reciprocal of time (Fig. 2) where LCO = a+b(l/t). The
intercept (a) of this regression is the predicted no-effect
concentration for chronic lethality. Log transformations, log LCO
= a+b (1/t) or log LCO = a+b log (1/t) , were required for ten tests
because of negative intercepts and/or curvilinear nature of the
data.
When test data permits, response-surface models (multiple
regression) for analyzing all data from an acute toxicity test
simultaneously (Fig. 3) are preferable to the two-step simple
linear regression approach described above. We therefore developed
a probit surface methodology and a user-friendly software program
(simple linear and multiple regression models) to predict chronic
toxicity based on acute time-exposure-effect data (Appendix A) .
This method is called Multifactor Probit Analysis (MPA) and uses
the iterative reweighed least squares method to estimate the
parameters of the probit surface. The independent variables
consist of time of exposure and concentration of the toxicant. The
dependent variable is the probit of the proportion responding to an
exposure concentration. MPA allows the user to predict the
concentration of a toxicant at any time and percent mortality as
well as calculate a measure of variability (95% confidence limits).
The calculation of LCOs is dependent upon slope and time
5
-------
course of effect, both of which are influenced by sample size
(number of fish per concentration) and dose separation (dilution
factor among concentrations). In this study, sample sizes ranged
from 10 to 30 organisms and dilution factors ranged from 50 to 75%.
The minimum acceptable sample size and maximum dilution factors
were not determined in this study, but could be from the data set
(Appendix B).
Data Base
The acute and chron'ic tests (Appendix B) selected for analyses
were taken from those conducted at the Columbia National Fisheries
Contaminant Research Center (U.S. Fish and Wildlife Service,
Columbia, MO) and the Gulf Breeze Environmental Research Laboratory
(U.S. Environmental Protection Agency, Gulf Breeze, FL) on seven
fish species: rainbow trout, Oncorhvnchus mvkiss: cutthroat trout,
0. clarki; brook trout, Salvelinus fontinalis: lake trout, S.
namavcush; fathead minnow, Pimephales promelas; channel catfish,
Ictalurus punctatus: and sheepshead minnow, Cvprinodon varieaatus.
With the exception of a few static acute tests used, acute and
chronic tests were conducted in flow-through diluter systems
modeled after that described by Mount and Brungs (1967) . Each
diluter delivered four to seven concentrations of toxicant and a
control. Water temperature was maintained within ± 1° C of the
desired temperature, and day length was regulated by the methods of
Drummond and Dawson (1970). Acute and chronic tests were conducted
in accordance with standard procedures (Committee on Methods for
Toxicity Tests with Aquatic Organisms 1975, American Society for
Testing and Materials 1980, Clesceri et al. 1989). The specific
6
-------
methods and materials, experimental design, and measurements are
included among the published articles cited as footnotes at the end
Table l. Concentrations of all chemicals were measured.
The concentration-response data in our historical data base
for chlordane, endrin, EPN, heptachlor, methoxychlor, and toxaphene
with sheepshead minnows were inadequate for observations prior to
the 9 6 h point in time; for that reason, several acute tests with
sheepshead minnows were repeated. Acute and chronic flow-through
tests were also conducted with carbon tetrachloride and sheepshead
minnows as another test of the LCO method, because carbon
tetrachloride is considered to have different modes of action
between acute and chronic exposures with mammals (Haley 1987,
Hardin 1954, Recknagel et al. 1989).
Two additional types of data sets' having available and
acceptable acute toxicity tests were analyzed — a pond study with
bluegills, Lepomis macrochirus. and fluorene, a component of
petroleum (Boyle et al. 1985, Finger et al. 1985), and the other
was with coturnix quail, Coturnix iaponica. and mercuric or methyl
mercuric chloride (Hill and Soares 1984). Fourteen 0.08 ha ponds
were treated with various concentrations of fluorene (July 26,
1982) . The ponds were drained approximately 70 days after exposure
(early October, 1982), and the fish were counted, measured, and
weighed to determine survival, growth, and production of recruits.
Fluorene exposures in the ponds were based on Average measured
concentrations following treatment on days 1, 3, and 7. Acute
toxicity tests with bluegills were conducted in the laboratory
under static conditions to simulate pond exposures. Five-d acute
7
-------
dietary tests were conducted with coturnix quail by presenting the
chemicals at various concentrations in turkey starter mash for 5
days. Daily observations for evidence of toxicity were made until
clinical signs were no longer detectable (10 days). Chronic
toxicity was determined by feeding the mercurials at various
concentrations in ad libitum diets from hatching to adulthood (9
weeks).
8
-------
RESULTS AND DISCUSSION
When the predicted values were compared with the observed
values of chronic tests (early life-stage arid partial and full life
cycle toxicity tests), they proved highly accurate for a variety of
chemicals and fish species (Table 1; Appendix C, model 5) . The
predicted no-effect concentrations (PNEC) were very close to or
within the limits (highest concentration without effect on survival
and the next higher concentration with a significant effect) of the
.maximum acceptable toxicant concentrations (MATC) for lethality and
varied by less than a factor of two from an MATC 92% of the time.
The other 18% of the predictions (two observations) consisted of
factors of 2.5 and 4.8 of the observed concentrations.
The technique worked very well in predicting chronic lethality
of carbon tetrachloride to sheepshead minnows (PNEC = 10.4 mg/L,
observed = 4.5
-------
The predictive technique was also highly accurate among
various single chemicals and mixtures; it seemed representative of
a wide range of octanol-water partition coefficients (log Kow).
Results of acute static tests might be used when flow-through tests
results are not available and the log Kow is low (e.g., fluridone) .
Chemicals that are highly water soluble will not adsorb to the test
container or be taken up by the test organisms as much as with
chemicals of low water solubility, and exposure will more closely
resemble that for flow-through tests. However, additional research
is needed to determine the log Kow below which static test data can
be used to predict chronic toxicity.
Although the other studies (pond and quail) analyzed represent
a very small data set, it is notable that the PNECs were accurate.
The ponds were dosed in a static acute manner (MATC for lethality'
= 0.032 ug/g for mercuric chloride and 2.0
-------
an order of magnitude. The cause is being investigated, and when
determined, the MPA program will be modified to default to model 5
under those conditions. PNECs can be derived with model 5 when the
other models do not work, but confidence limits are not provided.
The technique for deriving PNECs uses some aspects of concepts
developed previously. Sprague (1969) recommended that acute tests
be conducted until the toxicity curve becomes parallel to the time
axis, indicating a threshold concentration. An incipient LC50 is
then estimated by selecting an exposure time from the asymptotic
part of the toxicity curve. The reciprocal of mean survival times
within concentrations was used as early as 1917 by Powers (Jones
1964). Regressing the reciprocal of mean survival time on
concentration to derive theoretical thresholds of toxicity was
further developed by Abram (1964, 1967) and Alderdice and Brett
(1957). Although observing survival times in acute tests has
merits, it is laborious and is only infrequently done.
The approach of incorporating all data in an acute test
(concentration, degree of response, and time course of effect) to
predict chronic lethality has a technical basis. During the last
2 0 years, fish chronic toxicity tests have been shortened from full
life cycle tests to 30-90 d early life stage or partial life cycle
tests (Macek and Sleight 1977; McKim 1977, 1985) and-then to 7-d
subchronic tests (Norberg and Mount 1985). Reviews of subchronic,
early life stage, partial life cycle, and full life cycle toxicity
tests with several fish species demonstrated that the shorter tests
are good estimators of chronic toxicity and MATCs observed in the
longer life cycle tests (Macek and Sleight 1977; McKim 1977, 1985;
-------
Norberg-King 1989; Woltering 1984). Although the success of
developing briefer tests to estimate chronic toxicity is
empirically based, it does support the toxicological concept of
time course of effect in using acute data to predict chronic
lethality.
Another use of acute toxicity data to estimate chronic
toxicity is the toxicity threshold value or LCI (Birge et al. 1985,
Birge et al. 1989), which is calculated for 1.0% mortality and at
one point in time. This application of acute tests should work
well for those chemicals, effluents, and so on that differ little
in toxicity between acute and chronic effects or where the LCI is
derived at a duration approaching or within chronic exposure
conditions. However, the LCI does not take into account time
course of effect, and its use for predictive purposes is limited
for a wide range of chemicals; particularly those that
bioconcentrate or have cumulative effects.
Relation to Other Endpoints
Chronic toxicity tests commonly include the measurement of
long-term effects of a contaminant on the survival, growth, and
reproduction of a test organism. Assessments of sensitivity in
relation to chronic endpoints in fishes have been conducted
(Woltering 1984, Mayer et al. 1986, Suter et al. 1987, Ward and
Parrish 1980). Survival and growth are often equally sensitive,
and growth may not be of critical importance in establishing no-
effect concentrations in most tests. In tests for which growth is
the single most sensitive endpoint, survival could be used to
estimate the no-effect concentration within a factor of 3.
12
-------
Mayer et al. (1986) found growth-related endpoints to be
highly predictable from survival effects with freshwater fishes
(Table 2). Length was less variable than weight, and although all
of the coefficients of determination (r2) exceeded 0.9, they were
slightly higher for length (0.941 to 0.949) than for weight (0.901
to 0.931). Also, no alteration was noted in the intercepts (a) for
length versus survival between 30 and 90 days of exposure; the
intercepts of weight versus survival varied, without trends, over
time. Using these equations (Table 2) , estimated no-effect
concentrations for growth may be derived from the predicted values
for chronic lethality.
No-effect concentrations were always less for reproduction
endpoints than for survival (Mayer et al. 1986, Suter et al. 1987).
Attempts to relate acute lethality to chronic reproductive effects
by regression analysis have not been successful (Suter et al.
1987) . Because of the likelihood of different modes of action
between lethal and reproductive effects, we do not recommend that
reproductive effects be predicted using the proposed method.
However, the proposed technique is highly beneficial in the
preliminary assessment of chronic toxicity of effluents and other
chemicals and in predicting chronic no-effect concentrations for
survival and growth with fish species that are difficult to culture
under chronic testing conditions.
13
-------
ACKNOWLEDGEMENTS
The suggestions and encouragement of Drs. M.R. Ellersieck,
J.W. Gillett, and J.J. Lech contributed significantly to this
research. R.S. Stanley conducted the acute tests for sheepshead
minnows, J.C. Moore provided partition coefficients, J.A. Folse and
T.L. Simon assisted in data management and analysis, Valerie A.
Coseo and Maureen Stubbs typed the report. The project was
sponsored by the U.S. Environmental Protection Agency's Office of
Research and Development Ecological Risk Assessment Research
Program.
14
-------
REFERENCES
Abram, F.S.H. 1964. An application of harmonics to fish
toxicology. Int. J. Air Wat. Pollut. 8:325-338.
Abram, F.S.H. 1967. The definition and measurement of fish
toxicity thresholds. Advances in Water Pollution Research.
Proc. 3rd. Int. Conf. 1:75-95.
Alderdice, D.F. and J.R. Brett. 1957? Some effects of kraft mill
effluent on young Pacific salmon. J. Fish. Res. Board Can.
14:783-795.
American Society for Testing and Materials. 1980. Standard
practice for conducting acute toxicity tests with fishes,
macroinvertebrates, and amphibians. In Annual Book of ASTM
Standards, E729-80. American Society for Testing and
Materials, Philadelphia, PA.
Birge, W.J., J.A. Black, T.M. Short, and A.G. Westerman. 1989.
A comparative ecological and toxicological investigation of a
secondary wastewater treatment plant effluent and its
receiving stream. Environ. Toxicol. Chem. 8:437-450.
Birge, W.J., J.A. Black, and A.G. Westerman. 1985. Short-term
fish and amphibian embryo-larval- tests for determining the
effects of toxicant stress on early life stages and estimating
chronic values for single compounds and complex effluents.
Environ. Toxicol. Chem. 4:007-821.
Boyle, T.P., S.E. Finger, R.L. Paulson, and C.F. Rabeni. 1985.
Comparison of laboratory and field assessment of fluorene -
Part II: Effects on the ecological structure and function of
experimental pond ecosystems. Pages 134-151 in T.P. Boyle,
ed. Validation and Predictability of Laboratory Methods for
Assessing the Fate and Effects of Contaminants in Aquatic
Ecosystems. American Society for Testing and Materials STP
865, Philadelphia, PA.
Buckler, D.R., A. Witt, Jr., F.L. Mayer, and J.N. Huckins. 1981.
Acute and chronic effects of Kepone and mirex on the fathead
minnow. Trans. Am. Fish. Soc. 110:270-280.
Buikema, A.L., Jr., B.R. Niederlehner, and J. Cairns, Jr. 1982.
Biological monitoring. Part IV-toxicity testing. Water Res.
16:239-262.
Clesceri, L.S., A.E. Greenberg, and R.R. Trussell, eds. 1989.
Standard Methods for the Examination of Water and Wastewater.
17th ed. American Public Health Assoc., Washington, D.C.
15
-------
Cleveland, L. , D.R. Buckler, F.L. Mayer, and D.R. Branson. 1982.
Toxicity of three preparations of pentachlorophenol to fathead
minnows - A comparative study. Environ. Toxicol. Chem. 1:205-
212.
Committee on Methods for Toxicity Tests with Aquatic Organisms.
1975. Methods for acute toxicity tests with fish,
macroinvertebrates, and amphibians. EPA-660/3-75-009. U.S.
Environmental Protection Agency, Duluth MN.
Cripe, G.M., L.R. Goodman, and D.J. Hansen. 1984. Effect of
chronic exposure to EPN and to Guthion on the critical
swimming speed and brain acetylcholinesterase activity of
Cvprinodon varieaatus. Aquatic Toxicol. 5:255-266.
Drummond, R.A. and W.F. Dawson. 1970. An inexpensive method for
simulating diel patterns of lighting in the laboratory.
Trans. Am. Fish. Soc. 99:434-435.
Dwyer, W.P., F.L. Mayer, J.L. Allen, and D.R. Buckler. 1978.
Chronic and simulated use-pattern exposures of brook trout
(Salvelinus fontinalis^ to 3-trifluromethyl-4-nitrophenol
(TFM). Investigations in Fish Control No. 84. U.S. Fish and
Wildlife Service, Washington, DC.
Finger, S.E., E.F. Little, M.G. Henry, J.F. Fairchild, and T.P.
Boyle. 1985. Comparison of laboratory and field assessment
of fluorene - Part I: Effect of fluorene on the survival,
growth, reproduction, and behavior of aquatic organisms in
laboratory tests. Pages 120-133 in T.P. Boyle, ed.
Validation and Predictability of Laboratory Methods for
Assessing the Fate and Effects of Contaminants in Aquatic
Ecosystems. American Society for Testing and Materials STP
865, Philadelphia, PA.
Finney, D.J. 1971. Probit Analysis, 3rd ed. Cambridge University
Press, London.
Goodman, L.R., D.J. Hansen, J.A. Couch, and J. Forester. 1976.
Effects of heptachlor and toxaphene on laboratory-reared
embryos and fry of the sheepshead minnow. Southeast Assoc.
Game and Fish Comm. 30:192-202.
Green, R.H. 1965. Estimation of tolerance over an indefinite time
period. Ecology 46:887.
Haley, T.J. 1987. Solvents and chemical intermediates. Pages
504-554 in T.J. Haley and W.0. Berndt, eds. Handbook of
Toxicology. Hemisphere Publ. Corp., Washington, DC.
Hamelink, J.L., D.R. Buckler, F.L. Mayer, D.U. Palawski, and H.O.
Sanders. 1986. Toxicity of fluridone to aquatic
invertebrates and fish. Environ. Toxicol. Chem. 5:87-94.
16
-------
Hansen, D.J. and P.R. Parrish. 1977. Suitability of sheepshead
minnows fCyprinodon varieaatus^ for life-cycle toxicity tests.
Pages 117-126 in F.L. Mayer and J.L. Hamelink, eds. Aquatic
Toxicology and Hazard Evaluation. American Society for
Testing and Materials STP 634, Philadelphia, PA.
Hansen, D.J., S.C. Schimmel, and J. Forester. 1977. Endrin:
Effects on the entire life cycle of a saltwater fish,
Cyprinodon varieaatus. J. Toxicol. Environ. Hlth. 3:721-733.
Hansen, D.J. and G.M. Cripe. 1984. Interlaboratory comparison of
the early life-stage toxicity test using the sheepshead minnow
f Cyprinodon varieqatus). EPA-600/X-84-081. U.S.
Environmental Protection Agency, Gulf Breeze, FL.
Hardin, B.L. 1954. Carbon tetrachloride poisoning - A review.
Ind. Med. Surg. 23:93-105.
Hill, E.F. and J.H. Soares, Jr. 1984. Subchronic mercury exposure
in coturnix and a method of hazard evaluation. Environ.
Toxicol. Chem. 3:489-502.
Jones, J.R.E. 1964. Fish and River Pollution. Butterworth,
London.
Kenaga, E.E. 1979. Aquatic test organisms and methods useful for
assessment of chronic toxicity of chemicals. Pages 101-111 in
K.L. Dickson, A.W. Maki, and J. Cairns, Jr., eds. Analyzing
the Hazard Evaluation Process. American Fisheries Society,
Washington, DC.
Kenaga, E.E. 1982. Predictability of chronic toxicity from acute
toxicity of chemicals in fish and aquatic invertebrates.
Environ. Toxicol. Chem. 1:347-358.
Litchfield, J.T., Jr. and F. Wilcoxon. 1949. A simplified method
of evaluating dose-effect experiments. J. Pharmacol. Exp.
Ther. 96:99-113.
Macek, K.J. and B.N. Sleight III. 1977. Utility of toxicity tests
with embryos and fry of fish in evaluating hazards associated
with the chronic toxicity of chemicals to fishes. Pages 137-
146 in F.L. Mayer and J.L. Hamelink, eds. Aquatic Toxicology
and Hazard Evaluation. American Society for Testing and
Materials STP 634, Philadelphia, PA.
Mayer, F.L., P.M. Mehrle, and W.P. Dwyer. 1975. Toxaphene effects
on reproduction, growth, and mortality of brook trout. EPA-
600/3-75-013. U.S. Environmental Protection Agency, Duluth,
MN.
17
-------
Mayer, F.L., P.M. Mehrle, and W.P. Dwyer. 1977. Toxaphene:
Chronic toxicity to fathead minnows and channel catfish. EPA-
600/3-77-069. U.S. Environmental Protection Agency, Duluth,
MN.
Mayer, F.L., W.J. Adams, M.T. Finley, P.R. Michael, P.M. Mehrle,
and V.W. Saeger. 1981. Phosphate ester hydraulic fluids: An
aquatic environmental assessment of Pydrauls 50E and 115E.
Pages 103-123 in D.R. Branson and K.L. Dickson, eds. Aquatic
Toxicology and Hazard Assessment. American Society for
Testing and Materials STP 737, Philadelphia, PA.
Mayer, F.L., K.S. Mayer, and M.R. Ellersieck. 1986. Relationship
of survival to other endpoints in chronic toxicity tests with
fish. Environ. Toxicol. Chem. 5:737-748.
Mayer, F.L., C.H. Deans, and A.G. Smith. 1987. Inter-taxa
correlations for toxicity to aquatic organisms. EPA-600/8-
87/017. U.S. Environmental Protection Agency, Gulf Breeze,
FL.
McKim, J.M. 1977. Evaluation of tests with early life stages of
fish for predicting long-term toxicity. J. Fish. Res. Board
Can. 34:1148-1154.
McKim, J.M. 1985. Early life stage toxicity tests. Pages 58-95
in G.M. Rand and S.R. Petrocelli, eds. Fundamentals of
Aquatic Toxicology. Hemisphere Publishing Corp., Washington,
DC.
Mount, D.I. and W.A. Brungs. 1967. A simplified dosing apparatus
for toxicology studies. Water Res. 1:21-29.
Mount, D.I. and C.E. Stephan. 1967. A method for establishing
acceptable limits for fish - Malathion and the butoxyethanol
ester of 2,4-D. Trans. Am. Fish. Soc. 96:185-193.
Norberg, T.J. and D.I. Mount. 1985. An new fathead minnow
(Pimephales promelas^ subchronic toxicity test. Environ.
Toxicol. Chem. 4:711-718.
Norberg-King, T.J. 1989. An evaluation of the fathead minnow
seven-day subchronic test for estimating chronic toxicity.
Environ. Toxicol. Chem. 8:1075-1089.
Parrish, P.R., S.C. Schimmel, D.J. Hansen, J.M. Patrick, and J.
Forester. 1976. Chlordane: Effects on several estuarine
organisms. J. Toxicol. Environ. Hlth'. 1:485-494.
Recknagel, R.O., E.A. Glende, Jr., J.A. Dolak, and R.L. Waller.
19891 Mechanisms of carbon tetrachloride toxicity.
Pharmacol. Ther. 43:139-154.
18
-------
Shirazi, M.A. and L. Lowrie. 1988. Comparative toxicity based on
similar asymptotic endpoints. Arch. Environ. Contain. Toxicol.
17:273-280.
Snedecor, G.W. and W.G. Cochran. 1980. Statistical Methods, 7th
ed. Iowa State University Press, Ames, IA.
Sprague, J.B. 1969. Measurement of pollutant toxicity to fish.
I. Bioassay methods for acute toxicity. Water Res. 3:793-
821.
Suter, G.W. II, A.E. Rosen, E. Linder, and D.F. Parkhurst. 1987.
Endpoints for responses of fish to chronic toxic exposures.
Environ. Toxicol. Chem. 6:793-809.
U.S. Environmental Protection Agency. 1981. Acephate, aldicarb,
carbophenothion, DEF, EPN, ethoprop, methyl parathion, and
phorate: Their acute and chronic toxicity, bioconcentration
potential, and persistence as related to marine environments.
EPA-600/4-81-023. U.S. Environmental Protection Agency, Gulf
Breeze, FL.
Ward, G.S. and P.R. Parrish. 1980. Evaluation of early life-
stage toxicity tests with embryos and juveniles of sheepshead
minnows (Cvprinodon varieaatus^. Pages 243-247 in J.G. Eaton,
P.R. Parrish, and A.C. Hendricks, eds. Aquatic Toxicology.
American Society for Testing and Materials STP 707,
Philadelphia, PA.
Woltering, D.M. 1984. The growth response in fish chronic and
early life stage toxicity tests: A critical review. Aquat.
Toxicol. 5:1-21.
Woodward, D.F. and F.L. Mayer. 1978. Toxicity of three herbicides
(butyl, isooctyl, and propylene glycol butyl ether esters of
2,4-D) to cutthroat trout and lake trout. Technical Paper No.
97. U.S. Fish and Wildlife Service, Washington, DC.
19
-------
Table 1. Comparison of observed maximum acceptable toxicant
concentrations (MATC) and predicted no-effect concentrations (PNEC) for
lethality based on flow-through acute tests.
Chemical and
Species
Log
Kow
MATCa
(Mg/L)
PNEC
(Mg/L)
Butyl benzyl phthalate
Fathead minnows
Carbon tetrachloride
Sheepshead minnows
Chlordane
Sheepshead minnows
Complex effluent
Fathead minnows
2,4-D Butyl ester
Cutthroat trout
Lake trout
2,4-D PGBEE
Cutthroat trout
Lake trout
Endosulfan
Sheepshead minnows
Endrin
Sheepshead minnows
EPN
Sheepshead minnows
Fluridone
Channel catfish
Heptachlor
Sheepshead minnows
Kepone
Fathead minnows
Methoxychlor
Rainbow trout
Sheepshead minnows
4.44
2 . 64
5.80
2 .81
4 . 88
4.90—6.00c
4.56-5.30
4 . 80
1.87
5 .44
6.08
4.20
>360
4,500
-------
Pentachlorophenol
Fathead minnows
Phorate
Sheepshead minnows
Pydraul 50E
Fathead minnows
5.01
3.50
4.62-6.08f
>142
0 . 24
-------
Table 2. Regression correlations3 of survival and growth*5 no observed effect
concentrations among all fish species and chemicals at various time periods.
Coefficient of
Analysis and days Intercept Slopec Determination
of exposure n (a) (b) (r2) y±95%C.I.
Weight vs. survival
30
6
0. 395
0.920
0. 931
4.35+0.41
60
10
0. 682
0.901
0.901
4.63+0.31
90
15
0.194
0.993
0.916
4.3 3+0.22
:ngth vs. survival
30
16
0. 284
0.968
0.945
4.64+0.18
60
17
0. 263
0.965
0. 949
4.60+0.15
90
18
0.275
0.971
0.941
4.41+0.17
aLog y = a+b(log x) , where y is no-effect concentration for survival and x is
no-effect concentration (ng/L) for growth (length or weight).
bMayer et al. (1986).
CA11 slopes were significantly different from 0 (p<0.01).
22
-------
Fig. l. Dose-response curves used to derive the LCO (0.01%) for
various observation times in acute toxicity tests (1.281
= a+bX). Probit % mortality: 1.281 = 0.01%, 5.000 = 50%,
and 8.719 = 99.99%.
23
-------
Log Concentration
-------
Fig. 2. Prediction of the chronic no-effect value for lethality
from acute toxicity test data with Kepone and fathead
minnows (LCO * a+b [1/t]). The intercept (a) represents
the LCO (2.9 ng/L) over an indefinite exposure time, and
the maximum acceptable toxicant concentration (MATC) for
chronic lethality was between 1.2 and 3.1 iig/L.
24
-------
10
8.0
6.0
4.0
.a = 2.9 jig/L, NOEC = 1.2 - 3.1 ng/L
2.0
0.010
0.021
1/time (h)
0.042
-------
Fig. 3. Acute flow-through toxicity test results with carbon
tetrachloride and sheepshead minnows demonstrating dose-
response data in time (96-h LC50 = 19 mg/L)•
25
-------
100
Ctt
£
\
75
c
o
o
©
50
25
20
AO
60
10
-------
26
-------
APPENDIX A
Multifactor Probit Analysis Program
Preceding page blank
27
-------
28
-------
EPA / 600/X-.91/101
July 1991
MULTIFACTOR PROBIT ANALYSIS
by
Gunhee Lee, M.S., Mark Ellersieck, Ph.D. and Gary Krause, Ph.D.
University of Missouri-Columbia
105 Math Sciences Building
Columbia, Missouri 65211
Cooperative Agreement No. CR816166
Project Manager
Gary Krause
Professer of Statistics
University of Missouri-Columbia
Project Officer
Foster L. Mayer, Jr.
U.S. Environmental Protection Agency
Environmental Research Laboratory
Gulf Breeze, Florida 32561
Preceding page blank 2s
-------
DISCLAIMER
The research described in this article has been
subjected to Agency review for internal Agency distribution.
Mention of trade names does not constitute endorsement or
recommendation for use.
30
-------
TABLE OF CONTENTS
Acknowledgement 3 2
Abstract 33
Introduction 35
Conclusions 35
Recommendations and Important Considerations 3 6
Approach 41
Multifactor Probit Analysis Software 43
Description of Multifactor Probit Analysis Main Menu . . . 44
Main Program Menu Item 1 44
Main Program Menu Item 2 47
Main Program Menu Item 3 47
Main Program Menu Item 4 49
Main Program Menu Item 5 4 9
Main Program Menu Item 6 _ 49
Main Program Menu Item 7 50
Main Program Menu Item 8 51
Output Control 58
Description of Output 58
References 61
31
-------
ACKNOWLEDGEMENT
The project was sponsored by the U.S. Environmental Protection
Agency's Office of Research and Development Ecological Risk
Assessment Research Program, through Gulf Breeze Environmental
Research Laboratory.
32
-------
ABSTRACT
Environmental toxicologists are interested in the long-time-
exposure effect of a low concentration of a toxic substance.
Long-time-exposure toxicity testing is time consuming and
expensive; consequently, accurate methods for estimating long-
time-exposure effects which eliminate this time and expense, are
desirable. In the past long-time-exposure toxicity was
determined by calculating an application factor or an acute-
chronic ratio for a limited number of species and then applying
these multiplicative factors to other species. This method may
not give accurate estimates and does not give any measure of the
sampling variance of the estimate.
A methodology has been developed that will predict long-
time-exposure effect toxicity based on acute data. This method
is called Multifactor Probit Analysis (MPA) and uses the
iterative reweighed least squares method to estimate the
parameters of the probit surface. The independent variables are
time of exposure and concentration of the toxicant. The
dependent variable is the probit of the proportion responding to
a concentration. MPA allows the user to predict the
concentration of a toxicant at any time and percent mortality,
LCt>p. The Multifactor Probit Analysis calculates a point
estimate and a measure of dispersion (95% approximate confidence
limits).
The Multifactor Probit Analysis software is versatile and
the user can choose from seven different probit models and seven
different transformation combinations of the independent
variables. This software is entirely menu driven.
MPA predicts long-time-exposure mortality from acute data.
This prediction represents the amount of a toxic substance that
can exist in a laboratory environment for an extended exposure
time that will produce 0.01 percent mortality.
33
-------
34
-------
SECTION 1
INTRODUCTION
Environmental toxicologists are interested in determining
low concentration mortality of a chemical to an organism when
exposed for extended periods of time. More specifically: What
concentration of a chemical can exist in a laboratory environment
with only small effect on biological life? In the past,
estimates for these chronic no-effect concentrations have been
estimated using a combination of chronic and acute data for a
particular species.
A methodology and a computer program have been developed
cooperatively by the Environmental Protection Agency and the
University of Missouri-Columbia that predicts the long-time-
exposure lethality of chemicals from acute toxicity test data.
The software is called Multifactor Probit Analysis (MPA). This
software calculates the lethal concentration of a chemical for
expected effect, P (probability of response), for extended
periods of exposure time.
SECTION 2
CONCLUSIONS
Statistical models are programmed that utilize information
from several acute bioassay data sets to quantify the
relationship between exposure time, and dose of a chemical and
mortality. This user friendly software provides maximum
likelihood estimates of relevant parameters. Output from this
software includes a predicted concentration LCp, that has
expected effect P, (P is the proportion of subjects responding
which may be very small), under extended exposure time.
Approximate confidence limits are provided on true LCp.
Computer software, called MULTIFACTOR PROBIT ANALYSIS, may
be used to: a) assist selection of a model to relate exposure
time and concentration to probit mortality, b) estimate the
functional relationship among parameters in the best model and
exposure time to predict long-time-exposure LCp. P may be very
small. Comparison of the predicted LCP to long-time-exposure LCp
estimates from long-time-exposure trials may be done. Evaluation
of the appropriateness of this scheme would be dependent upon
this comparison.
35
-------
SECTION 3
RECOMMENDATIONS AND IMPORTANT CONSIDERATIONS
1. Short term exposure tests should be independent. The
methodology used assumes this condition. If observed
mortality at time t and a concentration is cumulative, then
bias may result.
2. Choice of model depends on mode of action of the toxicant.
Parallel concentrations-mortality lines assumes mode of
action is constant as time varies. An interaction between
concentration and time allows for. changes in mode of action.
Several models may be evaluated using the same data. This
permits a scheme, based on heterogeneity chi-square to
select the best model of those tried.
3. If control mortality is observed at zero dose (control),
Abbott's adjustment can be used to adjust non-zero dose
mortality. However, if control mortality is not constant
over the entire range of exposure time, Abbott's adjustment
is no longer valid. In this situation analysis can be done
by ignoring control mortality. A rule of thumb is, control
mortality should not be greater than 10%. (If control
mortality is greater than 10% the entire test should be
redone).
4. If"exposure tests are done at different times, a separate
probit analysis at each time will give an indication of
parallel or non parallel slopes. If the slopes are similar,
a multifactor probit model using parallel lines should be
used. If the slopes are different, a model with non-
parallel slopes should be used.
ESTIMATING NO OBSERVABLE EFFECT CONCENTRATIONS
We set the tolerable long-time-exposure effect at 0.01%.
The concentration that causes this effect (probit value of 1.281)
will be called the No Observable Effect Concentration (NOEC).
First Approach:
This approach is based on a simple probit analysis or a
least square linear regression for each exposure time (Type 5).
The procedure is as follows:
For given time t, NOEC will be estimated from the estimated
simple probit line. Then one estimates the regression line
between NOEC estimates and reciprocal of exposure using the
-36
-------
following model:
NOEC = a + /9/time.
If the model is correct, NOEC converges to a as time become
large.
Mote 1:
Several dose and time transformations should be evaluated
for each different chemical and species. Estimated NOEC will
be the one which has maximum R-square for the regression
model.
NOTE 2:
Suppose there are not enough data points to estimate a
probit regression line. In this situation, we recommend
that the maximum no-mortality concentration for a specific
toxicant be included for analysis. In this case, the
experimenter should identify the NOEC for different exposure
times and check the monotonicity assumption. For example,
NOEC of 4 8 hours cannot be less than 72 hours but should be
greater than 24 hours.. Generally, we recommend the analysis
include the maximum no-mortality concentration for a large
number of exposure times in the experiment. This will give
more information about NOEC at time infinity.
NOTE 3:
The use of the least-squares-method to estimate the probit
regression line when responses are the same at different
exposure times, may have a danger of over estimation, i.e.
NOEC estimate may be higher than the true NOEC. The reason
is, that data at 96 hours gives more information for time
infinity predictions than earlier exposure times. This
method treats them equally. If slopes between any two
(NOEC,time) points changes after some period of time, for
example mode of action changes, this method is very
insensitive to change and provides a compromised slope which
will be smaller.
Note 4:
MPA program uses MLE (Maximum Likelihood Estimation) of
simple probit analysis and also calculates a simple LS
(least square) analysis for each time. Theoretically, the ML
estimator is usually superior to LS estimator. ML
estimation is sensitive to changes in observed data.
However, there is a greater possibility of an estimate
against the monotonicity assumption. For example, ML
estimate of NOEC increases as time increases or p% lethal
concentration decreases as p increases. With MLE, we highly
recommend the option of screen plotting (which is provided
by MPA program) be chosen and checked for the monotonicity
37
-------
assumption, i.e. 72 hours simple probit line estimate should
lie above the 48 hours simple probit line est-imate. If MLE
provides an unreasonable estimate, LS should be selected.
The results from LS has the serious overestimation problem
and is quite sensitive to the choice of concentration in the
experiment. LS, however, guarantees the monotonicity
assumption as long as the observed data holds the assumption
and always produces an estimate of NOEC when MLE may not.
Second Approach:
This approach is based on a multiple probit model, Dose-
Time-Response surface. Unlike the first approach, which
estimates the NOEC probit for each time, this approach solves the
Dose-Time-Response equation simultaneously. The MPA program can
compute four cases. The cases are as follows:
Case 1*.
Specific long time exposure is specified and assumes equal
slope for every time (Type 2).
Dose - Time - Response relationship is defined as
Probit (p) — q +• 0*(Dose) + Y*(Time).
Let us denote specific long exposure time as T. NOEC at T hours
can be found as follows:
NOECt = 1'2.8Ip
where, 1.281 is the probit value of 0.01%.
Case 2:
Specific long time exposure is unknown and equal slope is
expected for every time (Type 6).
Dose - Time - Response relationship is defined as
Probit (p) = a + £*(Dose) + -//(Time).
NOEC at long exposure time can be found as follows:
NOEC = 1'2^1"a
P
38
-------
where, 1.281 is the probit value of 0.01%.
Case 3:
Specific long time exposure is specified and one assumes
slope changes with constant rate as time increases (Type 3).
Dose - Time - Response relationship is defined as
Probit (p) = a + 0*(Dose) + v*(Time) + S* (Dose) * (Time) .
Let us denote specific long exposure time as T. NOEC at T hours
can be found as follows:
where, 1.281 is probit value of 0.01%.
Case 4:
Specific long time exposure is unknown and one assumes slope
changes with constant rate as time increase (Type 7).
Dose - Time - Response relationship is defined as
Probit (p) = a + /3*(Dose) + Y/(Time) + S* (Dose) / (Time) .
NOEC at infinity hours can be found as follows:
where, 1.281 is probit value of 0.01%.
Note 5:
Type 6 and Type 7 can be utilized with specified long
exposure time. NOEC at T hours can be found as:
NOECt
1.281-g-y»r
noect
1. 281-a-y/T
P
if TypeS is appiled
NOECt
= 1 if Type? is aPPUed.
39
-------
Note 6:
Several dose and time transformations should be evaluated
for each different chemical and species. The best estimated
NOEC will be the one which has the minimum computed
heterogeneity factor. The heterogeneity factor equals the
computed chi-square divided by degrees of freedom.
Note 7:
Since long exposure time is quite dependent on different
species (or average life of species), in some cases, an
experimenter may want to set a specific time (for example,
1440 hours). If there is knowledge of a life cycle of a
species, estimation of NOEC should be based on the average
life time to avoid underestimation. NOEC at time infinity is
always less than NOEC at a specified time.
Note 8:
When a cross product term is used, i.e. 5*(Dose)*(Time) term
in Type 3 and S*(Dose)/(Time) term in Type 7, there is still
a small chance to get an estimate against the monotonicity
assumption. If this happens, both Type 3 and Type 7 of any
dose time transformation should not be considered to
estimate NOEC. In this situation, the assumption of
monotonicity is not met even though it has a small computed
chi-square.
Note 9:
If no candidate model has reasonable small computed chi-
square, i.e. every -candidate model has large chi-square
which is greater than 10 times the degrees of freedom, care
should be taken when estimating the NOEC. Multiple probit
model is not appropriate with a large chi-square.
40
-------
SECTION 4
APPROACH
MPA has two primary functions. The first function is for
entering and editing datum not only for the MPA subroutine, but
also for other software. Data files already prepared in ASCII
format can be retrieved using this function. The data entry and
editing function is described in detail in a later section.
»
The second function is statistical analysis. Once a data
set has been entered, a selected MPA subroutine executes an
analysis. The output which is produced depends on the analysis
option chosen.
The analysis options are:
A:
Simple probit analysis using one independent variable, which
is concentration. A single exposure time is assumed.
B:
Multifactor probit analysis which includes two independent
variables, (concentration and exposure time). This option
assumes parallel probit regression lines at each exposure
time.
C:
This is also"a multifactor probit analysis using •
concentration, time and interaction as independent
variables. This option is different than B in that non-
parallel probit lines over time are assumed.
D:
This option computes a multifactor probit analysis with
three independent variables (concentration of toxicant 1,
exposure time and a third variable which could be a second
toxicant).
E.
This is a simple probit analysis using concentration as the
independent variable and is computed at each level of
exposure time.
F:
This option is the same as option B except that the
independent variable representing exposure time is the
reciprocal of time, (1/time).
G:
This option is the same as C except that the independent
variable representing exposure time is l/Time.
41
-------
Since the MPA uses time as one of the independent variables,
the mortality or other quantal response needs to be observed at
different times. Mortality must be observed at least two times.
However, when only two time tests are available relationships
will be poorly determined resulting in wide confidence intervals.
Therefore, it is preferable to observe mortality more than two
times.
If the model chosen includes the independent variable
1/Time then the long time exposure small effect concentration is
estimated conditional on 1/Time or l/(log time) being 0 and a
choice of mortality (perhaps .01 percent). The estimate is the
y-intercept of the regression of an LC value on x (x being the
time factor), the predicted concentration of a toxicant that will
essentially produce small effect (perhaps .01 percent) under
long-time-exposure.
42
-------
SECTION 5
MULTIFACTOR PROBIT ANALYSIS SOFTWARE
The anticipated computer is an IBM PS/2 or an earlier PC.
High resolution graphics are preferred. The Multifactor Probit
Analysis (MPA) Software is initiated by placing the program disk
into disk drive A and typing A:MPA. A logo should appear on the
screen. This will remain until you press the key.
After pressing the key, the following MAIN PROGRAM
MENU, (above the dashed line below) and Current Program Status,
(below the dashed line below) will appear on the monitor.
MULTIFACTOR PROBIT ANALYSIS
MAIN PROGRAM MENU
1 - CHOOSE TYPE OF PROBIT MODEL AND LOG TRANSFORMATION
2 CHOOSE EXPOSURE TIMES
3 ENTER NEW DATA
4 EDIT DATA IN MEMORY
5 GET DATA FROM DISK
6 SAVE DATA ON DISK
7 DEFINE A TITLE, CHANGE GRAPHICS MODE
8 STATISTICAL ANALYSIS
9 QUIT
CHOOSE 1-9 (enter a single number, you do not press )
CURRENT MODEL STATUS
CURRENT MODEL
..ONE INDEPENDENT VARIABLE (DOSE)
CURRENT EXPOSURE TIMES 24 48 72 96
CURRENT TRANSFORMATION NATURAL LOG
LAST DISK FILE READ
LAST DISK FILE WRITTEN ON ..
TITLE
Above the dashed line is a menu of the possible operations
that the MPA has available. Below the dashed line is the Current
Model Status showing the current statistical model, current
exposure times, current transformation, last disk file read, last
disk file written on, and title.
43
-------
DESCRIPTION OF MULTIFACTOR PROBIT ANALYSIS MAIN MENU
MAIN PROGRAM MENU ITEM 1:
1)CHOOSE TYPE OF PROBIT MODEL AND LOG TRANSFORMATION.
Number 1 in the MAIN PROGRAM MENU serves three functions. The
first function permits the choice of model. If number 1 is
selected from the MAIN PROGRAM MENU a second menu appears on the
monitor:
PROBIT MODEL
MENU OF STATISTICAL MODELS
1
ONE INDEPENDENT VARIABLE (DOSE)
2
TWO INDEPENDENT VARIABLES (DOSE AND TIME) WITH
PARALLEL SLOPE
3
INTERACTION BETWEEN DOSE AND TIME WITH NON-PARALLEL SLOPE
4
THREE INDEPENDENT VARIABLES (DOSE, TIME AND A THIRD
VARIABLE, eg. second dose)
5
ONE INDEPENDENT VARIABLE (DOSE) GROUPED BY TIME
6
TWO INDEPENDENT VARIABLES WITH PARALLEL SLOPE USING
RECIPROCAL-OF TIME
7
TWO INDEPENDENT VARIABLES WITH NON-PARALLEL SLOPE USING
RECIPROCAL OF TIME
8
QUIT
CHOOSE 1-8 (enter a single number, you do not need to press
)
THE STATISTICAL MODELS;
NOTATION:
Suppose n subjects are tested at k different dose levels.
r : the response frequency from n subjects given dose
level z.
x : the transformed value of z (natural log or loglO
transformation).
P : the proportion of subjects responding at dose level
z, (P = r/n).
tx : the representation of exposure time for these
values of r, n and z.
ft : the third transformed interval scale factor which may be
a dose level of a second chemical.
The basic Probit function of P is:
* 1 2
/I * x
——e 2 -p
,%/S¥
44
-------
Type 1 : Probit model with one independent variable.
Probit(p) = a+bx.
where a is the intercept and
b is the coefficient on the metric of
dose concentration x.
Type 2 : Probit model with two independent variables.
Probit(p) = a+bx+ctx.
where a is the intercept,
b is the coefficient on the metric of
dose concentration x and
c is the coefficient on the metric of
time (hours).
NOTE:
In Type 2, parallel probit lines are assumed for each time.
Type 3 : Probit model with two independent variables and
interaction between dose and time.
Probit(p) = a+bx+ctx+dxtx.
where a is the intercept,
b is the coefficient on the metric of
dose concentration x,
c is the coefficient on the metric of
time (hours) and
d is the coefficient on cross product term.
NOTE:
In Type 3, the slopes are changing at rate d so the slope
will be b+dtx. For example, if d equals -0.2, then the
slope will decrease -0.2 as time increases 1 unit.
Type 4 : Probit model with three independent variables for
exposure time i.
Probit(p) «= a+bx+ctx+dft.
where a is the intercept,
b is the coefficient on the metric of
dose concentration x,
c is the coefficient on the metric of
time (hours) and
d is the coefficient on the metric of
third interval scale factor ft.
Type 5 : Probit model with one independent variable.
Probit (pi) = a^+bj^x
NOTE:
May be used for a regression at each level of time.
45
-------
Type 6 : Probit model with two
is the inverse of tx.
Probit(p) = a+bx+c/tx.
independent variables one of which
Type 7 : Probit model with two independent variables and
interaction which uses inverse of tx.
Probit(p) = a+bx+c/tx+dx/tx.
NOTE:
Type 6 and Type 7, assume parallel probit lines and non-
parallel probit lines, respectively. Type 2 and Type 3 has
time as one of the independent variables. This differs from
Type 6 and Type 7 which involves the reciprocal of time
(l/tx) as one of the independent variables.
2)CHOOSE TYPE OF TRANSFORMATION. After a probit model has
been chosen by selecting one of the seven models, a Data
Transformation Menu will appear. The transformation selected
applies to independent variables. This is the second function of
MAIN PROGRAM MENU Item 1.
DATA TRANSFORMATION MENU
1 NATURAL LOG OF DOSE AND NATURAL LOG OF TIME
2 LOG 10 OF DOSE AND LOG 10 OF TIME
3 INPUT VALUE OF DOSE AND INPUT VALUE OF TIME
4 NATURAL LOG OF DOSE AND LOG 10 OF TIME
5 NATURAL LOG OF DOSE AND INPUT VALUE OF TIME
6 LOG 10 OF DOSE AND NATURAL LOG TIME
7 LOG 10 OF DOSE AND INPUT VALUE OF TIME
8 QUIT
CHOOSE 1-8 (enter a single number, you do not need to press
< ENTER >)
When a transformation has been selected by entering a
number, the MAIN PROGRAM MENU will appear.
3) The third function identifies the order of variables in
the input record for the model selected. This is a data
requirement. You will not see this until you utilize MAIN
PROGRAM MENU Item 3.
Three variable orders are possible.
The Type l model requires the order be z, n and r.
46
-------
The Type 2, 3, 5, 6 and 7 models requires the order be z, t,
n and r.
The Type 4 model requires the ordering z, t, ft, n and r.
MAIN PROGRAM MENU ITEM 2:
CHOOSE EXPOSURE TIMES. These only influence output. The
default times are 24, 48, 72 and 96 for model types 2, 3 and 4.
For model types 6 and 7, the default times are 24, 48, 72, 96 and
time infinity. If MAIN PROGRAM MENU Item 2 is used, times other
than the default times can be chosen. These times (in hours)
dictate 'the times for which LC values will be calculated and
presented in PRINTED OUTPUT. If times other than default times
are needed they can be entered in response to the cue from MAIN
PROGRAM MENU Item 2. Ex:
ENTER EXPOSURE TIMES SEPARATED BY BLANKS
(Example: 24 48 72 96)
Model types 1 and 5 calculate simple probit analysis, so
exposure times are not present. Exposure times should be entered
within one line separated by blanks.
MAIN PROGRAM MENU ITEM 3;
ENTER NEW DATA. If data has not been entered previously, it
can now be entered by selecting item 3. The following
instructions and data input prompt will now appear on the
monitor, (this is an illustration assuming a Type 1 model was
chosen, see third function description of MAIN PROGRAM MENU Item
1) •
47
-------
ENTER INFORMATION USING THE FOLLOWING FORMAT. LEAVE AT LEAST
ONE SPACE BETWEEN NUMBERS. USE ARROWS, and KEYS
TO MOVE THE CURSOR. YOU CAN USE THE OR
KEYS. WHEN DONE PRESS THE KEY.
ENTER DOSE, NUMBER TESTED AND NUMBER RESPONDING ON EACH LINE
o
o
o
o
o
o
o
LINE 1 COL 1 12:00:00 01/01/91
The cursor should be at the first line and column as
indicated at the bottom of the monitor. The order of entry will
come from a prompt. At least one space is needed between values
for successive variables.
There are a number of commands that are useful for data
entry: "
,
Key or the Key is depressed to send the cursor
to the top of a file (Home) or bottom of a file (End).
,
If the or keys are depressed it moves the file
up or down one editor screen.
The key puts the user in and out of insert mode. If
one is in the insert mode it is indicated in the first
window. This key is only used for inserting data or
characters within a line.
The key deletes data or characters within a line.
The key allows data to be typed in specific columns.
(Default is every 5 spaces.)
The can be used to position the cursor while in the
48
-------
work window.
This allows the user to insert a line previous to the
current line.
Deletes the current line.
To exit MAIN PROGRAM MENU Item 3, press .
MAIN PROGRAM MENU ITEM 4;
EDIT DATA IN MEMORY. The same commands used to enter new
data also work for editing existing data sets. Once the data are
in memory, use number 4 for editing. The same screen presented
under the discussion on MAIN PROGRAM MENU. Item 3 will appear. If
data are in memory and one utilizes MAIN PROGRAM MENU Item 3, the
data in memory and on the screen will be erased, waiting for new
data to be entered. The data must be in memory for the program
to run. To exit MAIN PROGRAM MENU Item 4, press .
MAIN PROGRAM MENU ITEM 5:
GET DATA FROM DISK. If a data set already resides on a
disk, choose MAIN PROGRAM MENU Item 5. The screen will now
present the statement.
ENTER THE ASCII FILE NAME (Example. B:PROBIT.DAT)
If one enters the ASCII data set name and presses enter,
this data set will be in memory and can be edited by MAIN PROGRAM
MENU Item 4.
MAIN PROGRAM MENU ITEM 6:
SAVE DATA ON DISK. After data has been entered and/or
edited it should be saved. Enter 6 and the following message
will appear on the screen.
ENTER A FILE NAME (Example. B:PROBIT.DAT)
¦p
^ssssassasssBi^SBsssssasssssssssssssaBsasEaaas&sBnsaHBSBSBKSssas&ssassaiasssBSsssssssssss^s:
After entering the ASCII data set name and pressing the
key, the following WARNING will appear on the screen.
49
-------
THIS PROCESS WRITES OVER AN OLD DATA SET WITH THE SAME
NAME IF ONE EXISTS. ARE YOU SURE YOU WANT TO DO THAT ?
(Y/N) ?
If Y is entered the data set will be saved or replaced. If
N is entered the data will not be saved under the name specified,
however, it still resides in memory. After the Y or N is entered
the MAIN PROGRAM MENU will appear. One may enter MAIN PROGRAM
MENU Item 6 again and save the ASCII file under a different name.
MAIN PROGRAM MENU ITEM 7:
1)DEFINE A TITLE. CHANGE GRAPHICS MODE. Item 7 in the MAIN
PROGRAM MENU serves two main functions. The first is to define a
title. If a number of data sets are to be analyzed it is
important to title each output. If MAIN PROGRAM MENU Item 7 is
selected a Miscellaneous Menu will appear:
MISCELLANEOUS MENU
1DEFINE A TITLE
2CHANGE THE GRAPHICS MODE
3QUIT
CHOOSE 1-3 (enter it single number, you do not need to
press )
If 1 is selected from the Miscellaneous Menu, the response
will be:
ENTER A TITLE:
After a title has been entered the program will return to
the MAIN PROGRAM MENU.
2)GRAPHICS HARDWARE CONTROL. The second function of MAIN
PROGRAM MENU Item 7 serves as graphic hardware control. When
selected the Miscellaneous Menu will appear. If 2 is selected
from Miscellaneous Menu (CHANGE THE GRAPHICS MODE), the following
response will appear:
ENTER THE GRAPHICS MODE (eg. EGAMONO, HIRES, EGAHIRES, VGA)
•>
50
-------
Four responses are possible and are defined as follows:
EGAMONO:
Monochrome graphics with 640 pixels horizontally by 350
pixels vertically.
HIRES:
CGA High Resolution Graphics.
EGAHIRES:
EGA High Resolution Graphics. This can be used if the
computer is equipped with an EGA card. This is default.
VGA:
VGA High Resolution Graphics.
After entering one of the 4 options the program will return
to the MAIN PROGRAM MENU.
If 3 is entered from the Miscellaneous Menu, the MAIN
PROGRAM MENU will appear.
MAIN PROGRAM MENU ITEM 8:
1)STATISTICAL ANALYSIS. After the data has been typed or
retrieved from disk, a statistical model chosen and a title
(optional) entered, the data will be analyzed when 8 is selected.
If this is done, the following response will appear if models
chosen were 2, 3, 4, 5, 6 or 7. The plotting option is not
available if model 1 was chosen.
DO YOU WANT TO DO SCREEN PLOTTING? (Y/N)?
If Y or N is entered (which is the command for Y (yes) or N
(no) screen plotting of probit line), another menu will appear if
mortality occurs at any 0 dose level. If no mortality exists at
dose level 0, screen plotting will begin. If N was entered (no
plotting), all screen graphics are suppressed and the program
proceeds to the output control menu after control mortality is
checked.
51
-------
2)CONTROL MORTALITY OPTIONS.
NON-ZERO RESPONSE IS PRESENT AT DOSE LEVEL 0.
1 STOP PROCESSING.
2 IGNORE RESPONSE AT DOSE LEVEL 0.
3 ADJUST RESPONSE USING ABBOTT'S FORMULA.
CHOOSE 1-3 (enter a single number, you do not need
to press )
This menu will only appear if there is non-zero response at
dose level 0 (control mortality). If there is no control
mortality, this menu will not appear. If number 1 is entered,
the analysis will not be calculated and the MAIN PROGRAM MENU
will appear on the screen.
If number 2 is entered, all 0 dose records will be deleted
and the probit analysis will be computed.
If number 3 is entered, all mortalities are adjusted for
control mortalities using Abbott's formula.
To facilitate computation the data file is sorted by time .
and dose, then stored. Further the software truncates off all
records, except that for highest dose, with zero response and all
records, except that for lowest dose, with response = n.
If you have more than one independent variable, (i.e., Type 2, 3,
4, 6, 7), and control mortality varies for each exposure time, we
recommend one choose option 2, (IGNORE RESPONSE AT DOSE LEVEL 0).
Abbott's adjustment is only applied when control mortality is
constant for every exposure time assay.
52
-------
Once control mortality method (solution 2 or 3) has been
selected, Y (yes for plotting on the monitor) entered, and if the
type of Model chosen was 2, 3, 4, 6 or 7 a plot will appear on
the monitor.
The plot (for an example, use the figure above) has Dose
Concentration on the X axis and the Probit on the Y axis. Each
line represents time starting at 10 hours exposure with 10 hour
increments ending at 100 hours. The lines are in order with time
from bottom to top if mortality increases with time, if
decreasing they are in order from top to bottom.
; v y ;
53
-------
To proceed to the next plot, depress the key, the
following plot will appear. This graph will NOT appear if Probit
Model Type 5 was chosen.
This plot (figure above), has either time or 1/Time as the X
axis (this determined by the model chosen) and dose concentration
as the Y axis. The lines represent LCP values for P: .01%, 5%,
15%, 25%, 35%, 45%, 55%, 65%, 75%, 85% and 95%, graphed over time
from bottom to top.
54
-------
If the probit model chosen is Type 5, which is a separate
probit line for each exposure time, a different plot and an
additional option are available. As with other probit models the
option after entering Item 8 from the MAIN PROGRAM MENU
(Statistical Analysis), is the plot command. Once a decision on
plotting is made a second prompt will appear if control mortality
occurs. If control mortality is present a prompt will appear
allowing the user to use Abbott's formula or delete control
mortality. This is the same command for all other probit models
discussed earlier. Model type 5 is calculated by two different
methods. The first method is the maximum likelihood for each
time. The second method is a simple least square regression of
probit mortality on log dose for each time. Once the decision on
plotting and control mortality is checked, another prompt will
appear. If the calculation for least square estimates have less
than three points at a specific time the following response will
appear on the monitor.
ESTIMATION OF LEAST SQUARE REGRESSION HAD LESS THAN 3 OBS. AT
HOURS.
DO YOU WISH TO INCLUDE MAXIMUM CONCENTRATION WITH NO MORTALITY
FOR FURTHER REGRESSION ANALYSIS ? (Y/N)
-p
NOTE: MAXIMUM CONCENTRATION WITH NO MORTALITY IS
This response will appear for each time when less than three
points are available.
After the least square regression method has been checked for
each time, a plot will be produced similar to the plot on page
20, if plotting was requested. The maximum likelihood prompt
will appear if probit analysis cannot be computed at a specific
time. This situation will occur if the responses at a specific
time are all zeros, there are no partial mortalities or
convergence is not obtained. The response will be,
ESTIMATION OF PROBIT REGRESSION HAS FAILED AT HOURS.
DO YOU WISH TO INCLUDE MAXIMUM CONCENTRATION WITH NO MORTALITY
FOR FURTHER REGRESSION ANALYSIS ? (Y/N)
NOTE: MAXIMUM CONCENTRATION WITH NO MORTALITY IS
55
-------
This allows the user to either keep or delete the highest
concentration with zero mortality. This prompt will appear for
each time-exposure were the probit analysis fails. If none of
the probit analysis fails, this prompt will NOT appear. Those
concentrations with zero mortality are used in predicting long
time exposure at .01% mortality. After this prompt, the
following plot will appear if it was requested.
56
-------
1,27 2,25 3,24 4,22
NATURAL LOG (DOSE CONCENIBAIION)
The plot (for example, use the figure above) has Dose
Concentration on the X axis and the Probit on the Y axis. In
this example, probit lines are estimated for each 24, 48, 72 and
96 hours.
The equations for the lines at each time are shown in the output.
24 hours : Y = -0.799385 + 1.881835 X
48 hours : Y = 0.367254 + 1.831574 X
72 hours : Y = 0.446714 + 1.973819 X
96 hours : Y = 0.087199 + 2.354164 X
This plot will remain until the user presses the key.
57
-------
OUTPUT CONTROL
After the second plot appears, press , the monitor
will present the following menu. If one previously entered N for
plotting, this menu would have appeared without showing the
graphs.
OUTPUT MENU
ION THE SCREEN
2ON A PRINTER
3ON A DISK
4QUIT
CHOOSE 1-4 (enter a single number, you do not need to
press )
If number 4, is selected, the MAIN PROGRAM MENU will
appear. If 3 is selected, a prompt will appear asking for disk
drive identification and a data set name to identify the output
that will be stored. Number 1 and 2 direct the output to the
monitor or printer, respectively.
DESCRIPTION OF OUTPUT
The analysis procedures use the iterative reweighed least
squares method to .estimate the parameters of a probit plane.
During the process, MPA uses the convergence criteria of 10~5 or
100 maximum iterations to determine the completion of analysis.
The 10"5 criteria is based on the regression equation intercept
and regression coefficients. If the difference from one
iteration to the next is less than 10~5 for the intercept and
regression coefficient or partial regression coefficients,
depending on the model chosen, the convergence criteria is met.
If 10"5 is not met for any of the parameters in the model, a
further iteration is performed. If the 10"5 criteria is not met
after 100 iterations the analysis is terminated. After one of
these criteria has been met, a goodness-of-fit chi-square
statistic is computed. The output includes:
A.
The data and the data points that have been deleted as a
result of multiple 0% lethality or 100% lethality.
B.
Title if one is specified using MAIN PROGRAM MENU Item 7 and
the description of the probit model and transformation.
58
-------
c.
A listing of the values of the estimator is given for each
iteration. Each iteration produces an estimate of the
intercept and regression coefficient or partial regression
coefficients of concentration and time.
D.
Chi-square value is given for the goodness-of-fit test.
E.
The chi-square critical value a = .05, and the DF for chi-
square.
F*
The variance covariance matrix of estimators are given.
G.
Statement of applying a heterogeneity factor if the
calculated chi-square is greater than or equal to the
critical value of chi-square. The heterogeneity factor is
computed by dividing the goodness of fit chi-square
statistic by the degrees of freedom. This application of
heterogeneity factor is discussed by Finny (1971).
H.
The adjusted variance covariance matrix and heterogeneity
factor are printed. (Output items G and H are not printed if
the calculated chi-square is less than tabulated .chi-
square) .
I.
All analysis (except type 4) includes a listing of
mortality, lower and upper approximate confidence limits
(95%) and the point estimate of LCp.
If type l was chosen, a probit analysis is computed for a
single exposure time.
If type 5 was chosen, three different analysis are
performed.
The regression equation, (Dose = Intercept + Slope/Time), is
calculated along with the ANOVA table for percent probit
probabilities of .01, .1, 1, 5, 10, 50 and 90 percent. The
transformation of dose and time found in the regression equation
is controlled by the Data Transformation Menu described earlier.
This additional output is described below.
J.
The description of the regression equation is printed.
59
-------
K.
Probit probability, the intercept and slope for the equation
present in step J.
L.
The estimate of the lethal dose concentration at time
infinity is given. This estimate is calculated according to
the regression equation described in step J.
M.
Analysis of variance table and R-square value for the probit
analysis are printed for each 'probit.
If one of type 2, 3, 6 or 7 was chosen, the statistical
analysis is based on a multiple regression equation in which the
default times are 24, 48, 72 and 96 hour.
If type 6 or 7 was chosen, an additional time of infinity is
present.
60
-------
REFERENCES
Literature
1. Finney, D. J. (1971). Probit analysis. Cambridge
University Press, London.
2. Finney, D. J. (1978). Statistical method in biological
assay. Griffin, London.
3. Hastings, C. (1959). Approximations for digital computers.
Princeton Univ. Press, Princeton, NJ.
4. Mayer, F. L. (1990). Predicting chronic lethality to
fishes from acute toxicity data. Proc. Soc. Environ.
Toxicol. Chem. 11:93.
5. Mayer, F. L. (1990). Predicting chronic lethality of
chemicals to fishes from acute toxicity test data. U.S.
Environmental Protection Agency, Report No. EPA/600/X-
90/147, Gulf Breeze, FL. 15p.
Software
6. True BASIC Reference Manual. (1985). Addison-Wesley,
Reading, MA.
61
-------
62
-------
APPENDIX B
Acute and Chronic Toxicity Data Base
Preceding page blank
63
-------
64
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: Butyl benzyl phthalate
Species: Fathead minnows
Number/Concentration: 30
Age (days): Mean*
Length (mm): Mean= 47
Weight (g): Mean= 0.35
Temperature (C): Mean= 21
pH: Mean* 8.1
Hardness (mg/L): Mean= 297
Measured
[cone]
(ug/L)
Number dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
740
0
0
0
0
1060
0
1
1
1
2100
2
7
8
8
2770
22
27
28
28
3230
28
29
29
29
CHRONIC TOXICITY
Test type: ELS " Test duration (days): 30
Observed no-effect concentrations (ug/L):
Lethality: >360 Growth: 140-360 Reproduction:
Reference:
Unpublished data, Monsanto Company.
Preceding page blanK
65
-------
MARINE FISH TESTS
ACUTE TOXICITY
Chemical: Carbon tetrachloride
Species: Sheepshead minnows
Number/Concentration: 20
Age (days): Mean*
Length (mm): Mean* 16
Weight (g): Mean* 0.11
Temperature (C): Hean= 26
Salinity (o/oo): Mean* 21
Measured
[cone]
(ug/L)
Number dead at selected observation time (hours)
3
6
12
24
48
72
96
0
0
0
0
0
0
0
0
11700
0
0
0
0
0
0
0
16100
0
1
3
4
4
4
4
25400
17
17
17
18
18
18
19
38500
20
20
20
20
20
20
20
CHRONIC TOXICITY
Test type: ELS Test duration (days): 28
Observed no-effect concentrations (ug/L):
Lethality: 4500-11200 Growth: Reproduction:
Reference:
This study
66
-------
MARINE FISH TESTS
ACUTE TOXICITY
Chemical: Chlordane
Species: Sheepshead minnows
Number/Concentration: 20
Age (days): Mean> 27
Length (mm): Mean* 9
Weight (g): Mean» 0.02
Temperature (C): Mean* 25
Salinity (o/oo): Mean= 22
Measured
[cone]
(ug/L)
Nuntoer dead at selected observation time (hours)
24
46
77
96
0 1 0
0
0
0
5.1
0
0
0
0
9.3
0
0
3
6
K
0
0
0
0
19
0
4
10
15
31
0
11
19
20
Reference:
This study
CHRONIC TOXICITY
Test type: ELS Test duration (days): 28
Observed no-effect concentrations (ug/L):
Lethality: 7.1-17 Growth: Reproduction:
Reference:
Parrish, P.R., S.C. Schimmel, D.J. Hansen, J.M. Patrick, and J. Forester.
1976. Chlordane: Effects on several estuarine organisms. J.
Toxicol. Environ. Hlth. 1:485-494.
67
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: Complex effluent
Species: Fathead minnows
Number/Concentration: 40
Age (days): Mean-
Length (mm): Mean®
Weight (g): Mean=
Temperature (C): Mean=
pH: Mean- 7.3
Hardness (mg/L): Mean=
Measured
[cone]
X
Muflber dead at selected observation time (hours)
24
46
72
96
0
0
0
0
0
5.6
0
0
0
0
10
2
8
10
13
18
4
15
39
40
32
40
40
40
40
56
40
—
40
40
40
CHRONIC TOXICITY
Test type: ELS Test duration (days): 14
Observed no-effect concentrations (ug/L-):
Lethality: 2.0-3.5% Growth: 1.1-2.0% Reproduction:
0.6-1.1%
Reference:
Unpublished data, Monsanto Company.
68
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: 2,4-D BE
Species: Cutthrout trout
Number/Concentration: 30
Age days): Mean- 210
Length (mm): Mean= 79
Weight (g): Mean= 4.2
Temperature (C): Mean=10
pH: Mean« 7.4
Hardness (mg/L): Mean=162
Measured
[cone]
(ug/L)
Nunber dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
48
0
0
0
0
100
0
0
0
0
191
0
0
0
0
386
0
1
6
10
785
23
30
30
30
CHRONIC TOXICITY
Test type: ELS Test' duration (days): 60
Observed no-effect concentrations (ug/L):
Lethality: 24-44 Growth:
Reproduction:
Reference:
Woodward, D.F. and F.L. Mayer. 1978. Toxicity of three herbicides
(butyl, isooctyl, and propylene glycol butyl ether esters of
2,4-D) to cutthrout trout and lake trout. Technical Paper No.
97. U.S. Fish and Wildlife Service, Washington, DC.
69
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: 2,4-D BE
Species: Lake trout
Number/Concentration: 30
Age (days): Mean- 120
Length (mm): Mean* 60
Weight (g): Mean= 1.5
Temperature (C): Mean= 10
pH: Mean= 7.4
Hardness (mg/L): Mean= 162
Measured
Cconcl
(ug/L)
Number dead at selected observation time (hours)
24
68
72
96
0
0
0
0
0
48
0
0
0
0
100
0
0
0
0
191
0
1
1
1
386
0
1
4
K
785
29
30
30
30
CHRONIC TOXICITY
Test type: ELS Test duration (days): 60
Observed no-effect concentrations (ug/L):
Lethality: 33-60 Growth: 15-33 Reproduction:
Reference:
Woodward, D.F. and F.L. Mayer. 1978. Toxicity of three herbicides
(butyl, isooctyl, and propylene glycol butyl ether esters of
2,4-D) to cutthrout trout and lake trout. Technical Paper No.
97. U.S. Fish and Wildlife Service, Washington, DC.
70
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: 2,4-D PGBEE
Species: Cutthroat trout
Number/Concentration: 30
Age (days): Mean"
Length (mm): Mean' 79
Weight (g): Mean= 4.2
Temperature (C): Mean= 10
pH: Mean- 7.4
Hardness (mg/L): Mean= 162
Measured
[cone]
(ug/L)
Nunber dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
40
0
0
0
0
75
0
0
0
0
153
0
0
0
0
308
0
23
26
29
617
30
30
30
30
CHRONIC TOXICITY
Test type:.ELS Test duration (days): 60
Observed no-effect concentrations (ug/L):
Lethality: 31-60 Growth: Reproduction:
Reference:
Woodward, D.F. and F.L. Mayer. 1978. Toxicity of three herbicides
(butyl, isooctyl, and propylene glycol butyl ether esters of
2,4-D) to cutthrout trout and lake trout. Technical Paper No.
97. U.S. Fish and Wildlife Service, Washington, DC.
71
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: 2,4-D PGBEE
Species: Lake trout
Number/Concentration: 30
Age (days): Mean*
Length (mm): Mean* 60
Weight (g): Mean= 1.5
Temperature (C): Mean= 10
pH: Mean= 7.4
Hardness (mg/L): Mean= 162
Measured
[cone]
(ug/L)
Number dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
40
0
0
0
0
75
0
0
0
0
153
0
0
0
0
30S
0
5
19
23
617
30
30
30
30
CHRONIC TOXICITY
Test type: ELS Test duration•(days): 60
Observed no-effect concentrations (ug/L):
Lethality: 52-100 Growth: 52-100 Reproduction:
Reference:
Woodward, D.F. and F.L. Mayer. 1978. Toxicity of three herbicides
(butyl, isooctyl, and propylene glycol butyl ether esters of
2,4-D) to cutthrout trout and lake trout. Technical Paper No.
97. U.S. Fish and Wildlife Service, Washington, DC.
72
-------
MARINE FISH TESTS
ACUTE TOXICITY
Chemical: Endosulfan
Species: Sheepshead minnows
Number/Concentration: 20
Age (days): Mean-
Length (mm): Mean =
Weight
-------
MARINE FISH TESTS
ACUTE toxicity
Chemi ca1: Endrin
Species: Sheepshead minnows
Number/concentration: 20
Age (days): Mean* 23
Length (mm) : Mean * 9
Weight (g)s Mean=
Temperature (C): Mean= 26
Salinity (o/oo): Mean= 18
Measured
tconc]
(ug/L)
Nunber dead at selected observation time (hours) ,
24
48
72
96
0
0
0
0
0
0.16
0
0
0
0
0.26
0
0
11
15
0.52
3
17
20
20
0.76
16
20
20
20
1.4
20
20
20
20
Reference:
This study
CHRONIC TOXICITY
Test type: LC Test duration (days): 140
Observed no-effect concentrations (ug/L):
Lethality: 0.12-0.31 Growth: 0.12-0.31 Reproduction: 0.12-0.31
Reference:
Hansen, D.J., S.C. Schimmel, and J. Forester. 1977. Endrin:
Effects on the entire life cycle of a saltwater fish,
Cyprinodon varieqatus. J. Toxicol. Environ. Hlth. 3:721-733.
74
-------
MARINE FISH TESTS
ACUTE TOXICITY
Chemical: EPN
Species: Sheepshead minnows
Number/Concentration: 20
Age (days): Mean- 22
Length (mm): Mean * 12
Weight (g): Mean=
Temperature (C): Mean- 25
Salinity (o/oo): Mean= 20
Measured
[cone]
(ug/L)
Number dead at selected observatiorwtime (hours)
24
48
72
96
0
0
0
0
1
46
0
0
1
2
63
1
2
4
5
150
6
9
13
19
213
14
19
19
20
381
20
20
20
20
Reference:
This study
CHRONIC TOXICITY
Test type: PLC Test duration (days): 229
Observed no-effect concentrations (ug/L):
Lethality: 4.1-7.9 Growth: 4.1-7.9 Reproduction: >7.9
Reference:
Cripe, G.M., L.R. Goodman, and D.J. Hansen. 1984. Effect of chronic
exposure to EPN and to Guthion on the critical swimming speed
and brain acetylcholinesterase activity of Cvprinodon
varieaatus. Aquatic Toxicol. 5:255-266.
75
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY (Static)
Chemical: Fluorene
Species: Bluegills
Number/Concentration: 10
Age (days): Mean*
Length (mm): Mean*
Weight (g): Mean= 0.8
Temperature (C): Mean= 22
pH: Kean* 7.5
Hardness (mg/L)Kean= 280
Measured
[cone]
(ug/L)
Number dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
133
0
0
0
0
237
0
0
0
0
414
0
0
0
0
740
0
5
10
10
1332
0
8
10
10
2368
1
10
10
10
4144
4
10
10
10
7400
9
10
10
10
Reference:
Finger, S.E., E.F. Little, M.G. Henry, J.F. Fairchild, and T.P.
Boyle. 1985. Comparison of laboratory and field assessment of
fluorene- Part I: Effects of fluorene on the survival, growth,
reproduction, and behavior of aquatic organisms in laboratory
tests. Pages 120-133 in T.P. Boyle, ed. Validation and
Predictability of Laboratory Methods for Assessing the Fate and
Effects of Contaminants in Aquatic Ecosystems.. American
Society for Testing and Materials STP 865, Philadelphia, PA.
CHRONIC TOXICITY
Test type: Pond study Test duration (days): 70
Observed no-effect concentrations (ug/L) :
Lethality: 0-67 Growth: >433 Reproduction: 0-67
* Measured concentration based on average of day 1,3, and 7 analyses.
Reference:
Boyle, T.P., S.E. Finger, R.L. Paulson, and C.F. Rabeni.
1985.Comparison of laboratory and field assessment of fluorene-
Part II: Effects on the ecological structure and function of
experimental pond ecosystems. Pages 134-151 in T.P. Boyle, ed.
Validation and Predictability of Laboratory Methods for
Assessing the Fate and Effects of Contaminants in Aquatic
Ecosystems. American Society for Testing and Materials STP 865,
Philadelphia, PA.
76
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY (Static)
Chemical: Fluridona
Species: Channel catfish
Number/Concentration: 10
Age (days): Mean>=
Length (mm): Mean=
Weight (g): Mean- 0.70
Temperature (C): Mean= 22
pH: Mean- 7.1
Hardness (mg/L): Mean= 40
Measured
[cone]
(ug/L)
Number dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
1800
0
0
0
0
3200
0
0
0
0
5600
0
0
0
1
10000
0
2
6
9
18000
1
1
5
10
32000
10
10
10
10
56000
10
10
10
10
100000
10
10
10
10
CHRONIC TOXICITY
Test type: ELS Test duration (days): 60
Observed no-effect concentrations (ug/L):
Lethality: 1000-2000 Growth: 1000-2000 Reproduction:
Reference:
Hamelink, J.L.,
SanderB.
and fish.
D.R. Buckler, F.L. Mayer, D.U. Palawski, and H.O.
1986. Toxicity of fluridone to aquatic invertebrates
Environ. Toxicol. Chem. 5:87-94.
77
-------
MARINE FISH TESTS
ACUTE TOXICITY
Chemical: Heptachlor
Species: Sheepshead minnows
Number/Concentration: 20
Age (days): Mean" 37
Length (mm): Mean ¦ 10
Weight (g): Mean* 0.02
Temperature (C): Mean* 25
Salinity (o/oo): Mean= 22
Measured
Ccooc]
(ug/L)
Nkjnber dead at selected observation time (hours)
24
48
72
96
0
0
0
0
1
4.2
0
0
0
0
6.8
0
0
0
1
11
0
6
9
14
15
2
14
18
20
31
20
20
20
20
Reference:
This study
CHRONIC TOXICITY
Test type: PLC, ELS Test duration (days): 96,28
Observed no-effect concentrations (ug/L):
Lethality: 1.9-2.8 Growth: Reproduction: 0.97-1.9
2.2-3.5 2.2-3.5
Reference:
Goodman, L.R., D.J. Hansen, J.A. Couch, and J. Forester. 1976.
Effects of heptachlor and toxaphene on laboratory-reared
embryos and fry of the sheepshead minnow. Southeast Assoc.
Game and Fish Comm. 30:192-202. Hansen, D.J. and P.R. Parrish.
1977. Suitability of sheepshead minnows (Cvrarinodon
varieoatus) for lifecycle toxicity tests. Pages 117-126 in F.L.
Mayer and J.L. Hamelink, eds. Aquatic Toxicology and Hazard
Evaluation. American Society for Testing and Materials STP
634, Philadelphia, PA.
78
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: Kepone
Species: Fathead minnows
Number/Concentration: 20
Age (days): Mean- 30
Length (mm): Mean* 15
Weight (g): Mean= 0.03
Temperature (C): Mean= 25
pH: Mean* 7.8
Hardness (mg/L): Mean= 290
Measured
[cone]
(ug/L)
Number dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
10
0
-1
5
5
16
3
7
12
12
22
5
12
13
13
27
8
20
20
20
40
20
20
20
20
56
20
20
20
20
73
20
20
20
20
CHRONIC TOXICITY
Test type: ELS Test duration (days): 60
Observed no-effect concentrations (ug/L):
Lethality: 1.2-3.1 Growth: 1.2-3.1 Reproduction:
Reference:
Buckler, D.R., A. Witt, Jr., F.L. Mayer, and J.N. Huckins.
1981. Acute and chronic effects of Kepone and mirex on
the fathead minnow. Trans. Am. Fish. Soc. 110:270-280.
79
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: Methoxychlor
Species: Brook trout
Number/Concentration: 20
Age (days): Mean-
Length (mm): Mean*
Weight (g): Mean- 0.97
Temperature (C): Mean* 12
pH: Mean-
Hardness (mg/L): Mean-
Measured
[cone]
(ug/L)
Ninter dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
4.6
0
0
0
0
9.4
0
0
3
14
24
3
11
20
20
CHRONIC TOXICITY (Rainbow trout)
Test type: ELS Test duration (days): 90
Observed no-effect concentrations (ug/L):
Lethality: 1.1-3.1 Growth: 1.1-3.1 Reproduction:
Reference:
Unpublished data, Columbia National Fisheries Contaminant Research
Center.
80
-------
MARINE FISH TESTS
ACUTE TOXICITY
Chemical: Mothoxychlor
Species: Sheepshead minnows 1
Number/Concentration: 20
Age (days): Mean* 47
Length (mm): Mean • 15
Weight (g): Mean* 0.04
Temperature (C): Mean* 25
Salinity (o/oo): Mean* 22
Measured
[cone]
(ug/L)
Nunber dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
13
0
0
0
0
20
2
2
3
3
43
0
8
18
20
60
5
20
20
20
98
20
20
20
20
Reference:
This study
CHRONIC TOXICITY
Test type: PLC, ELS Test duration (days): 112,28
Observed no-effect concentrations (ug/L):
Lethality: 23-48 Growth: Reproduction: 12-23
12-23 >12
Reference:
Hansen, D.J. and P.R. Parrish. 1977. Suitability of sheepshead
minnows (Cyprinodon varieaatus) for lifecycle toxicity tests.
Pages 117-126 in F.L. Mayer and J.L. Hamelink, eds. Aquatic
Toxicology and Hazard Evaluation. American Society for Testing
and Materials STP 634, Philadelphia, PA.
81
-------
MARINE FISH TESTS
ACUTE TOXICITY
Chemical: Methoxychlor
Species: Sheepshead minnows 2
Number/Concentration: 20
Age (days): Mean" 21
Length (mm): Mean = 9
Weight (g): Mean= 0.007
Temperature (C): Mean= 25
Salinity (o/oo): Mean= 21
Measured
[cone]
(ug/L)
Number dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
13
0
0
0
0
20
0
0
0
0
35
0
6
19
20
61
9
20
20
20
120
20
20
20
20
Reference:
This Btudy
CHRONIC TOXICITY
Test type: PLC,ELS Test duration (days): 112,28
Observed .no-effect concentrations (ug/L):
Lethality: 23-48 Growth: Reproduction: 12-23
12-23 >12
Reference:
Hansen, D.J. and P.R. Parrish. 1977. Suitability of sheepshead
minnows (Cyprinodon varieaatus) for lifecycle toxicity tests.
Pages 117-126 in F.L. Mayer and J.L. Hamelink, eds. Aquatic
Toxicology and Hazard Evaluation. American Society for Testing
and Materials STP 634, Philadelphia, PA.
82
-------
MARINE FISH TESTS
ACUTE TOXICITY
Chemical: Methoxychlor
Species: Sheepshead minnows 3
Number/Concentration: 20
Age (days): Mean* 27
Length (mm): Mean = 9
Weight (g): Mean= 0.007
Temperature (C): Mean= 25
Salinity (o/oo): Mean= 20
Measured
[cone]
(ug/L)
Nuntoer dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
15
0
0
0
0
31
0
0
1
1
49
0
14
15
15
57
4
17
20
20
66
19
20
20
20
Reference
This study
CHRONIC TOXICITY
Test type: PLC, ELS Test duration (days): 112,28
Observed no-effect concentrations (ug/L):
Lethality: 23-48 Growth: Reproduction: 12-23
12-23 >12
Reference:
Hansen, D.J. and P.R. Parrish. 1977. Suitability of sheepshead
minnows /Cvprinodon varieoatua) for lifecycle toxicity tests.
Pages 117-126 in F.L. Mayer and J.L. Hamelink, eds. Aquatic
Toxicology and Hazard Evaluation. American Society for Testing
and Materials STP 634, Philadelphia, PA.
83
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: Pentachlorophenol-P
Species: Fathead minnows
Number/Concentration: 20
Age (days): Mean= 40
Length (mm): Means 26
Weight (g): Mean= 0.02
Temperature (C): Mean= 22
pH: Mean* 7.4
Hardness (mg/L): Mean= 272
Measured
[cone]
(ug/L)
Umber dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
237
0
0
0
0
311
2
4
4
4
414
20
20
20
20
CHRONIC TOXICITY
Test type: PLC Test duration (days): 90
Observed no-effect concentrations (ug/L):
Lethality: >142 Growth: 36-85 Reproduction:
Reference:
Cleveland, L. , D.R. Buckler, F.L. Mayer, and D.R. Branson. 1982.
Toxicity of three preparations of pentachlorophenol to fathead
minnows-A comparative study. Environ. Toxicol. Chem. 1:205-
212.
84
-------
MARINE FISH TESTS
ACUTE TOXICITY
Chemical: Phorate Length (mm): Mean « 7
Species: Sheepshead minnows Weight (g): Mean*
Number/Concentration: 20 Temperature (C): Mean= 25
Age (days): Mean- Salinity (o/oo) : Mean«= 27
Measured
[cone]
(ug/L)
Member dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
0.12
0
0
0
0
0.22
0
0
0
0
0.50
0
0
0
1
0.83
2
2
2
2
1.1
2
4
4
5
1.5
4
4
5
14
4.2
20
20
20
20
6.3
20
20
20
20
10
20
20
20 -
20
CHRONIC TOXICITY
Test type: ELS Test duration (days): 28
Observed no-effect concentrations (ug/L):
Lethality: 0.24-0.41 Growth: 0.24-0.41 Reproduction:
Reference:
U.S. Environmental Protection Agency. 1981. Acephate
aldicarb, carbophenothion, DEF, EPN, ethoprop, methyl
parathion, and phorate: Their acute and chronic toxicity,
bioconcentration potential, and persistence as related to
marine environments. EPA-600/4-81-023. U.S. Environmental
Protection Agency, Gulf Breeze, FL.
85
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: Pydraul 50E
Species: Fathead minnows
Number/Concentration: 20
Age (days): Mean'
Length (mm): Mean*
Weight (g): Mean= 1.6
Temperature (C): Mean= 16
pH: Mean* 7.7
Hardness (mg/L): Mean= 272
Measured
[cone]
(ug/L)
Nunber dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
754
0
0
0
0
1034
0
0
1
3
1301
0
1
6
12
1S30
13
19
20
20
CHRONIC TOXICITY
Test type: PLC Teat duration (days): 90
Observed no-effect concentrations (ug/L):
Lethality: 317-752 Growth: 317-752 Reproduction:
Reference:
Mayer, F.L., W.J. Adams, M.T. Finley, P.R. Michael, P.M. Mehrle,
and V.W. Saeger. 1981. Phosphate ester hydraulic fluids:
An aquatic environmental assessment of Pydrauls 50E and
115E. Pages 103-123 in D.R. Branson and K.L. Dickson, eds.
Aquatic Toxicology and Hazard Assessment. American
Society for Testing and Materials STP 737, Philadelphia, PA.
86
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: TFM
Species: Brook trout
Number/Concentration: 10
Age (dayB): Mean* Adult
Length (mm): Mean= 261
Weight (g): Mean= 212
Temperature (C): Mean= 10
pH: Mean* 7.4
Hardness (mg/L): Mean= 272
Measured
[cone]
(ug/L)
Nunber dead at selected observation time (hours)
1
3
6
24
48
72
96
0
0
0
0
0
o ¦
0
0
4500
0
0
0
0
0
0
0
6500
0
0
1
1
1
5
7
8600
0
0
5
7
7
7
8
11700
0
5
8
10
10
10
10
14400
1
8
10
10
10
10
10
CHRONIC TOXICITY
Test type: PLC (adult) Test duration (days): 120
Observed no-effect concentrations (ug/L):
Lethality: 4000-8800 Growth: 4000-8800 Reproduction: 1600-4000
Reference:
Dwyer, VI.P., F.L. Mayer, J.L. Allen, and D.R. Buckler. 1978.
Chronic and simulated use-pattern exposures of brook trout
(Salvelinus fontinalis) to 3-trifluromethyl-4-nitrophenol
(TFM). Investigations in Fish Control No. 84. U.S. Fish and
Wildlife Service, Washington, DC.
87
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: Toxaphene
Species: Brook trout 1
Number/Concentration: 20
Age (days): Mean= 480
Length (mm): Mean= 231
Weight (g): Mean= 133
Temperature (C): Mean= 10
pH: Mean* 7.4
Hardness (mg/L): Mean= 272
Measured
teonej
(ug/L)
Munber dead at selected observation time (hours)
72
96
120
1U
0
0
0
0
0
3.8
0
0
0
0
5.1
0
0
0
3
6.2
0
1
4
15
8.8
0
i.
15
20
12
16
20
20
20
CHRONIC TOXICITY
Test type: PLC (adult) Test duration (days): 180
Observed no-effect concentrations (ug/L):
Lethality: 0.14-0.29 Growthj 0.14-0.29 Reproduction: 0.039-0.068
Reference:
Mayer, F.L., P.M. Mehrle, and W.P. Dwyer. 1975. Toxaphene effects on
reproduction, growth, and mortality of brook trout. EPA-
600/3-75-013. U.S. Environmental Protection Agency, Duluth,
MN.
88
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: loxapbene
Species: Brook trout 2
Number/Concentration: 26
Age (days): Hean«
Length (mm): Mean-
Weight (g): Mean' 9.2
Temperature (C): Mean= 12
pH: Mean2 7.4
Hardness (mg/L): Mean= 272
Measured
[cone]
(ug/L)
Nunber dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
2.0
0
0
0
0
2.9
0
0
0
4
4.2
0
0
9
22
6.2
0
10
26
26
8.2
0
23
26
26
11
2
26
26
26
16
25
26
26
26
CHRONIC TOXICITY
Test type: ELS Test duration (days): 90
Observed no-effect concentrations (ug/L):
Lethality: 0.068-0.14 Growth: 0.068-0.14 Reproduction:
Reference:
Mayer, F.L., P.M. Mehrle, and W.P. Dwyer. 1975. Toxaphene effects
on reproduction, growth, and mortality of brook trout. EPA-
600/3-75-013. U.S. Environmental Protection Agency, Duluth,
MN.
89
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: Toxapben*
Species: Fathead minnows
Number/Concentration: 10
Age (days): Mean*
Length (mm): Mean*
Weight (g): Mean- 0.3
Temperature (C): Mean= 25
pH: Mean* 7.4
Hardness (mg/L): Mean= 272
Measured
[cone]
(ug/L)
Number dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
2.8
0
0
0
0
4.2
0
0
0
0
6.0
0
1
2
4
7.8
0
2
3
4
11
0
6
9
10
15
2
10
10
10
20
7
10
10
10
CHRONIC TOXICITY
Test type: PLC,ELS Test duration (days): 98,30
Observed no-effect concentrations (ug/L):
Lethality: >0.17 Growth: 0.054-0.097 Reproduction: >0.17
0.097-0.17
Reference:
Mayer, F.L., P.M. Mehrle, and W.P. Dwyer. 1975. Toxaphene: Chronic
toxicity to fathead minnows and channel catfish. EPA-600/3-77-
069. U.S. Environmental Protection Agency, Duluth, MN.
90
-------
FRESHWATER FISH TESTS
ACUTE TOXICITY
Chemical: Xoxaphene
Species: Channel catfish
Number/Concentration: 20
Age (days): Mean-
Length (mm): Mean=
Weight (g): Mean=
Temperature (C): Mean= 20
pH: Mean= 7.4
Hardness (mg/L): Mean= 272
Measured
[cone]
(ug/L)
Nunber dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
0.56
0
0
0
0
1.0
0
0
4
10
1.8
0
16
18
20
3.2
20
20
20
20
CHRONIC TOXICITY
Test type: ELS of PLC Test duration (days): 90
Observed no-effect concentrations (ug/L):
Lethality: 0.070-0.13 Growth: 0.070—0.13 Reproduction:
0.13-0.30
Reference:
Mayer, F.L., P.M. Mehrle, and W.P. Dwyer. 1975. Toxaphene: Chroni
toxicity to fathead minnows and channel catfish. EPA-600/3-
069. U.S. Environmental Protection Agency, Duluth, MN.
91
-------
MARINE FISH TESTS
ACUTE TOXICITY
Chemical: loxaphene
Species: Sheepshead minnows
Number/Concentration: 20
Age (days): Mean" 23
Length (mm): Mean = 7
Weight (g): Mean' 0.004
Temperature (C): Mean= 27
Salinity (o/oo): Mean= 22
Measured
[cone]
(ug/L)
Nimber dead at selected observation time (hours)
24
48
72
96
0
0
0
0
0
1.7
0
0
0
1
2.4
0
1
2
6
4.4
1
6
18
20
6.4
0
18
20
20
9.7
18
20
20
20
Reference:
This study
CHRONIC TOXICITY
Test type: ELS Test duration (days): 28
Observed no-effect concentrations (ug/L):
Lethality: 1.1-2.5 Growth: >2.5 Reproduction:
Reference:
Goodman, L.R., D.J. Hansen, J.A. Couch, and J.Forester. 1976.
Effects of heptachlor and toxaphene on laboratory-reared
embryos and fry of the sheepshead minnow. Southeast Assoc.
Game and Fish Comm. 30:192-202.
92
-------
BIRD TESTS
ACUTE TOXICITY (Subacute)
Chemical: Mercury (HgCl,)
Species: Coturnix quail
Number/Concentration: 15
Age (days): Mean* 14
Measured
[cone]
(ug/g)
Nuntoer dead at selected observation time (hours)
24
48
72
96
120
144
168
192
216
240
0
0
0
0
0
0
0
0
0
0
0
2500
0
0
1
2
2
2
'2
2
2
2
3535
0
0
0
0
1
4
5
5
5
5
5000
0
0
0
1
5
7
7
7
7
7
7070
0
1
1
5
8
11
11
11
11
11
10000
0
2
4
5
8
11
11
12
12
12
Chemical was presented at various concentrations in turkey starter mash for
5 days. Daily observations for evidence of toxicity were made from first
presentation of treated feed until clinical signs were no longe.r
detectable.
CHRONIC TOXICITY
Test type: sublethal Test duration (days): 63(hatch-9wks)
Observed no-effect concentrations (ug/g):
Lethality: >32 Growth: >32 Reproduction: >32
Reference:
Hill, E.F. and J.H. Soares, Jr. 1984. Subchronic mercury exposure in
coturnix and a method of hazard evaluation. Environ. Toxicol.
Chem. 3:489-502.
93
-------
BIRD TESTS
ACUTE TOXICITY (Subacute)
Chemical: Methyl mercury (CH^HgCl)
Speciea: Coturnix quail
Number/Concentration: 15 (exceptions: 0=10,30 ug/g=16)
Age (days): Mean- 14
Measured
[cone]
(ug/g)
Nunber dead at selected observation time (hours)
24
48
72
96
120
144
168
192
216
240
0
0
0
0
0
0
0
0
0
0
0
15
0
0
0
0
0
0
0
0
0
0
21
0
0
0
0
0
0
0
0
0
1
30
0
0
0
0
0
1
1
2
2
2
42
0
0
0
0
1
1
2
3
3
6
60
0
0
0
0
0
0
1
4
7
11
Chemical was presented at various concentrations in turkey starter mash for
5 days. Daily observations for evidence of toxicity were made from first
presentation of treated feed until clinical signs were no longer
detectable.
CHRONIC TOXICITY
Test type: sublethal Test duration (days): 63(hatch-9wks)
Observed no-effect concentrations (ug/g):
Lethality: 2-8 Growth: >32 Reproduction: 2-8
Reference:
Hill, E.F. and J.H. Soares, Jr. 1984. Subchronic mercury exposure in
coturnix and a method of hazard evaluation. Environ. Toxicol.
Chem. 3:489-502.
94
-------
Computer Product Information Sheet
NTIS Federal Computer Products Center Software
UNITED STATES DEPARTMENT OF COMMERCE
National Technical Information Service
5285 Port Royal Road, Springfield, VA 22161
TO ORDER: Phone: (703) 487-4650 FAX: 703-321-8547 Telex: 64617
(Also available on a Rush basis for an added fee)
Title:
Statistical Approach to Predicting Chronic Toxicity of Chemicals to Fishes from Acute Toxlcity
Test Data (for microcomputers)
Source: Environmental Protection Agency
NTIS Order Number: PB92-503119 Product Type: Software-Diskette
Date: as of June 1992
Price Code: D02 U.S.. Canada, & Mexico. $90.00, all other addresses: $180.00
(Price includes documentation, add $3 to each order (or handling)
Summary:
A methodology was developed to predict chronic toxicity based on acute data. This method is called Multifactor
Probit Analysis (MPA) and uses the iteractive reweighed least squares method to estimate parameters of the
probit surface. The independent variables are time of exposure and probit of the proportion responding to a
concentration. MPA allows the user to predict the concentration of toxicant at any time and percent mortality, L(
t,p. MPA calculates a point estimate and a measure of dispersion (95% Confidence limits). The MPA is versatile,
entirely menu-driven and offers 7 probit models and transformation combinations of the independent variables
The software is on one 51/4 inch diskette, 360K double density. File format: ASCII. Documentation included; may
be ordered separately as PB92-169655.
System: IBM PS2 Model 50; DOS 4.0 operating system, 640K. Language: True BASIC.
Refund Policy: NTIS does not permit return of items for credit or refund. A replacement will be provided if an error is made in
filling your order, if the item was received in damaged condition, or if the item is defective.
-------
PB92-503119
COMPUTER DISKETTE FILE PROPERTIES
7/13/92
(FCPC#507S)
01.
Completion Date
Year
Month
Day
02. Long Title
Statistical Approach to
Predicting Chronic
Toxicity of Chamicals to
Fishas from Acuta
03. Short Title
Multifactor Probit
Analysis (MPA)
MT
Copying Date
Year
Month
Day
05. Subscription
No
06.
ca New Product
~ Replacement
07. Number of
Diskettes
08. Submitting Organization and Address
Environmental Protection Agency
Office of Research and Development
Environmental Research Laooratory
G«if Breeze, FL 32561
09. For information about the content, contact:
Dr. F.L. Mayar () (904)-934-9380
Dr. G.F. Krause () (314)-882-6663
10. Host Computer/Model
I3K PS2 Mcael 50
11. Memory Requirement
640 K
12. Language/Format
True BASIC
ASCII
13. Diskette Size
; '.it, inch.
14. Diskette Capacity
360K
15. Operating System/Version
DOS 4.0
16. Number of Files
17. Number of Records
18. Record Length
19. Documentation
P392-169655
20. Supplemental Information
System: IBK PS2 Mode. SO; 2CS 4.C operating system, 640K. Language: True BASIC,
Thr software is or, on* i I/< inch disxette, 360K double density, file format: ASCI
Documentation includes; may be oraered separately as PB92-169655.
21. For Submitting Organization Use
Form NTIS-FCPC-01(1'B9)
-------
TECHNICAL REPORT DATA
fPleate reed Instruction on the rtvent bt/on completing)
VpV60T0/R-ff^!"/I1K-92/0'^
3. RECIPIENT'S ACCESSION NO
PB92-169655
«. title and subtitle
STATISTICAL APPROACH TO PREDICTING CHRONIC
¥8H6iT¥ 9fsfH6S;FSALS"T0 FISHES FR0M ACUTE
S. REPORT OATE
JUNE 1992
S. PERFORMING ORGANIZATION CODE
EPA/ORH
7 AUTHOAISI
E.L. M § y e r ^ , G.F. Krause' , M.R. Ellersieck^,
G . Lee
s. performing organization REPORT NO
9. PERFORMING ORGANIZATION NAME ANO ADDRESS
7
"University of Missouri Agricultural
Station, Columbia, Missouri 65211
Experiment
io. program element no.
1 1 CONTRACT/GRANT NO
If. SPONSORING AGENCY NAME ANO ADDRESS
1J.S. ENVIRONMENTAL PROTECTION AGENCY
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
GULF BREEZE, FLORIDA 32561
13. TYPE OP REPORT AND PERIOO COVERED
14. SPONSORING AGENCY CODE
is. supplementary notes
FOR DISKETTE SEL: I'.R9 2-50 31 1 9
is. abstract
A comprehensive approach to predicting chronic toxicity from
acute toxicity data was developed in which simultaneous
consideration is given to concentration, degree of response, and
time course of effect. A consistent endpoint (lethality) and
degree of response (0%) were used to compare acute and chronic
tests. Predicted no-effect concentrations were highly accurate 92%
of the time (within a factor of 2.0 of the limits of the maximum
acceptable toxicant concentrations for lethality) and did not vary
by more than a factor of 4.8 when the technicjue was applied to a
data base of 18 chemicals and 7 fish species. Growth effects can
be predicted from chronic lethality, but reproductive effects
should not be.
17. KBV WOROB ANO OOCUMBNT ANALYSIS
1 descriptors
b.lOCNTlFltRS/OPEN INOEO TERMS
c. cosati Fitld,Group
1B. DISTRIBUTION STATSMINT
It. SECURITY CLASS (Phi Report)
31. NO. OP PAGES
mi
20. SICURiTy Class i Thit petti
» PR'CE
EPA '•"» 2230-1 (Ha*. 4.771 •acviout boit.on >* oaseucrt
i
------- |