823R93006
GREAT LAKES WATER QUALITY GUIDANCE
ATTACHED ARE MATERIALS DISTRIBUTED ON SEPTEMBER 6 1991 IN
PREPARATION FOR THE SEPTEMBER 25-26,1991 GREAT LAKES STEERING COMMITTEE
MEETING
AQUATIC LIFE MATERIALS 1 - 242
o Procedure for Deriving Aquatic Life Criteria 3-45
o Aquatic life criteria for cadmium 46 - 56
o Aquatic life criteria for copper 57 - 64
o Aquatic life criteria for free cyanide 65 - 67
o Aquatic life criteria for chromium (III) 68 - 72
o Aquatic life criteria for chromium (IV) 73 - 78
o Aquatic life criteria for Dieldrin 79 - 85
o Aquatic life criteria for Endrin 86 - 93
o Aquatic life criteria for Lindane 94 - 100
o Aquatic life criteria for mercury (II) 101 - 110
o Aquatic life criteria for nickel 111 - 115
o Aquatic life criteria for parathion 116 - 122
o Aquatic life criteria for pentachlorophenol 123 - 127
o Aquatic life criteria for phenol 128 - 135
o Aquatic life criteria for selenium 136 - 141
o Aquatic life criteria for silver 142 - 149
o Methodology for deriving Tier II values 150 - 242
BIOACCUMULATION FACTOR MATERIALS 243 - 262
WILDLIFE MATERIALS 263 - 305
o Procedure for Deriving Criteria for Protection of
Wildlife 264 - 270
-------
&EPA
United States Office of Water
Environmental Protection WH-585 May 1993
Agency Washington, DC 20460
ERRATA SHEET FOR
THE GREAT LAKES WATER QUALITY GUIDANCE
As of May 4, 1993
On April 16, 1993, EPA published the proposed Water Quality Guidance for the Great
Lakes System and a correction notice identifying missing text and other changes
inadvertently omitted during editing of the proposed rule (58 FR 20802 and 21046).
Additional typographical or editorial errors identified since publication are listed
below If you identify any suspected errors, please notify Wendy Schumacher, Water
Quality Branch (WQS-16J), U.S. EPA, Region V, 77 West Jackson Blvd., Chicago, Illinois
60604, telephone: 312-886-0142.
The first sentence of the second last paragraph of section II.G of the preamble
should read, "SPA invites comment on the proposed BAF level of 1000 and any
alternative BAF levels for use in defining BCCs " The vord "defining" replaces the
word "defending," which was a typographical error This affects p. 50 of 3/31/93
prepublication copy; PRE-1.GLI, p. 200, of 3/30/93 prepublication disk copy; and
58 FR 20845.
The ambient water quality criteria for aquatic life for pentachlorophenol and phenol
should be corrected in Tables III-l and III-2 of the preamble and Tables 1 and 2 of
part 132. The correct values were used, however, in the support document, "Great
Lakes water-Quality initiative Criteria Documents for the Protection of Aquatic Life
in Ambient Water." The correct values are shown below (the published values are in
parentheses). This affects pp. 65-66 and 256-257 of 3/31/93 prepublication copy;
PRE-2.GLI, pp. 240 and 244, and RUL-1 GLI, pp. 1048-1049; and 58 FR 20853 and 58 FR
21014.
Acute Chronic
ug/L ug/L
Pentachlorophenol
should be 53 4.1
(was published as) (5.3—OK) (3 3—wrong)
Phenol
should be 3600 110
(was published as) (3700—wrong) (120—wrong)
The headings "Percentile" and "Sample Size" for Table III-3 of the preamble were
transposed in the Federal Register publication. The headings were correct in the
prepublication copy and disk. This affects SB FR 20856.
In Table IX-1 of the preamble, the entries for Major Direct Dischargers—Municipal
should read 348.9 and 353.5 for Scenarios 3 and 4 respectively. The initial "3"s
were omitted through typographical error. The totals in the Federal Register
publication were also in error for the same reason. They should read 473.9 and
505.5 for Scenarios 3 and 4 respectively. This affects p. 221 of 3/31/93
prepublication copy; PRE-6.GLI, p. 924, 58 FR 20987. Please note that these errors
occur only in the table. EPA used the correct figures in its analysis of the costs
of implementing the Guidance, and in the preamble text discussing the analysis. The
figures are also stated correctly in the support document, "Assessment of Compliance
Costs Resulting from Implementation of the Proposed Great Lakes Water Quality
Guidance."
-------
o Technical Support Document 271 - 305
DDT and Metabolites 283 - 287
Mercury 288 - 292
Polychlorinated Biphenyl (PCB) 293 - 295
2,3,7,8-TCDD 296 - 299
Summary of parameters 300
Reference used for the Great Lakes Wildlife
Criteria 301 - 305
HUMAN HEALTH MATERIALS 306 - 514
o Procedure for deriving human health criteria 307 ~ 323
o Technical support document for
human heall h criteria 324 - 429
o Human health criteria for p,p'-
dichlorodiphenyltrichloroethane (DDT) 430 - 437
o Human health criteria for dieldrin 438 - 445
o Human health criteria for chlordane 445 - 451
o Human health criteria for hexachlorobenzene 452 - 461
o Human health criteria for heptachlor 462 - 467
o Human health criteria for mercury 468 ~ 475
o Human health criteria for lindane 483 - 491
o Human health criteria for polychlorinated biphenyl 492 - 501
o Human health criteria for 2,3,7,8-
tetrachlorodibenzo-p-dioxin 503 - 514
GLWQI IMPLEMENTATION PROCEDURES 515 - 548
o Purpose 517
o Applicability 517
o Definitions 517 - 521
o Site-specific 521 - 522
o Total maximum daily loads/wasteload allocation
and mixing zones for point sources 522 - 533
o Compliance schedules 533
o WQBELS below the level of detection 533 - 535
H
-------
o Additivity 535 - 539
o Loading limits for point sources 539
o Whole effluent toxicity requirements for
point sources 539 - 542
o Background concentrations of pollutants 543 - 543A
o Reasonable potential to exceed WQBELS 543A - 545
o Variance from water quality standards for
point sources 545 - 547
o Environmental fate 547 - 548
GLWQI ANIDEGRADATION MODEL REGULATION 549 - 556
in
-------
AQUATIC LIFE MATERIALS
-------
GLI Aquatic Criteria
Methodology 9/5/91
EDITOR'S NOTE:
Text that is included in the U.S. EPA 1985 National Aquatic Guidelines, but has
been deleted from the GLI Aquatic Criteria Proposal, has been bracketed.
Text that is new, and is not found in the 1985 National Aquatic Guidelines, has
been underlined.
Text that was included in previous versions of this document, but since deleted,
are double braketed.
-------
GLWQI Aquatic Criteria
Methodology 9/5/91
Great lakes Water Quality Initiative
Procedure for Deriving Aquatic Life Criteria
A two tiered approach shall be used to derive acute and chronic values for the
protection of aquatic organisms.
Tier 1 acute and chronic numeric criteria shall be derived using a modification
of procedures described in the U.S. Environmental Protection Agency (U.S.EPA)
"Guidelines for Deriving Numerical National Water Quality Criteria for the
Protection of Aquatic Organisms and Their Uses", January, 1985 (hereafter
referred to as National Guidelines). Criteria derived using Tier l procedures
shall be adopted as numeric criteria by the Great Lake States and shall be based
upon data available in U.S.EPA's AQUIRE Data Base, Michigan's CESARS Data Base,
or any other appropriate data meeting the requirements set forth in Section II
of the Tier 1 procedure.
When insufficient data are available to derive aquatic criteria using the Tier
1 approach, Tier 2 acute and/or chronic criteria shall be derived using the
process described in Tier 2, The Tier 2 process shall be adopted by the Great
Lakes States as a narrative procedure to provide more flexibility for toxicity
data generation by dischargers. The procedure shall be vised to derive values for
interpreting concentrations of a chemical in an effluent or in ambient waters of
the Great Takes Basin. They would represent an agency's best professional
judgement and serve as the basis for a water quality based effluent limitation.
However, when used in a permit, a discharger should be given the flexibility to
-------
GLWQI Aquatic Criteria
Methodology 9/5/91
provide the minimum data base necessary for a Tier l calculation. If data were
provided prior to the effective date of the limit, a permit modification may be
appropriate.
Site-specific modifications of the Tier 1 and 2 procedures may be justified
following the guidance in the GLWQI Implementation Procedure (Procedure 4).
In addition to the criteria derived by procedures set forth in Tier l and
Tier 2, biological techniques including whole effluent toxicity requirements
shall be used as necessary, following the guidance contained in the GLWQI
Inplementation Procedure (Procedure 10) and U.S. EPA's "Technical Support
Document for Water Quality Based Toxics Control", 1985, to assure that toxic
conditions do not exist for aquatic life in surface waters of the Great
System.
-------
GLI Aquatic Criteria
Methodology 9/5/91
Great takes Water Quality Initiative
Procedure for Deriving Aquatic Life Criteria
Tier I
I. DEFINITION OF MATERIAL OF CONCERN
A. Each separate chemical that does not ionize substantially in most
natural bodies of water should usually be considered a separate
material, except possibly for structurally similar organic compounds
that only exist in large quantities as commercial mixtures of the
various compounds and apparently have similar biological, chemical,
physical, and toxicological properties.
B. For chemicals that do not ionize substantially in most natural
bodies of water (e.g. some phenols and organic acids, some salts of
phenols and organic acids, and most inorganic salts and coordination
complexes of metals and metalloids 1, all forms that would be m
chemical equilibrium should usually be considered one material. Each
different oxidation state of a metal and each different nonionizable
covalently bonded organometallic compound should usually be
considered a separate material.
C. The definition of the material should include an operational
analytical component. Identification of a material simply, for
example, as '"sodium" obviously implies "total sodium", but leaves
room for doubt. If "total" is meant, it should be explicitly
*_^
J
-------
GLI Aquatic Criteria
Methodology 9/5/91
stated. Even "total" has different operational definitions, some of
which do not necessarily measure "all that is there" in all samples.
Thus, it is also necessary to reference or describe the analytical
method that is intended. The operational analytical component
should take into account the analytical and environmental chemistry
of the material, the desirability of using the same analytical
method on samples from laboratory tests, ambient water, and aqueous
effluents, and various practical considerations, such as labor and
equipment requirements, and whether the method would require
measurement in the field or would allow measurement after samples
are transported to a laboratory.
The primary requirements of the operational analytical component are
that it be appropriate for use on samples of receiving water, that
it be compatible with the available toxicity and bioaccumulation
data without making extrapolations that are too hypothetical, and
that it rarely result in underprotection or overprotection of
aquatic organisms and their uses. Because an ideal analytical
measurement will rarely be available, a compromise measurement will
usually have to be used. This compromise measurement must fit with
the general approach that if an ambient concentration is lower than
the criterion, unacceptable effects will probably not occur, i.e.,
the compromise measure must not err on the side of underprotection
when measurements are made on a surface water. Because the chemical
and physical properties of an effluent are usually quite different
from those of the receiving water, an analytical method that is
2
-------
GLI Aquatic Criteria
Methodology 9/5/91
acceptable tar analyzing an effluent might not be appropriate for
analyzing a receiving water, and vice versa. If the ambient
concentration calculated front a measured concentration m an
effluent is higher then the criterion, an additional option is to
owasur* the concentration after dilution of the effluent with the
receiving water to determine if the measured concentration is
lowered by such phenomens as oomplexation or sorption. A further
option, of course, is to derive a site-specific criterion. Thus,
the criter.Lon should be based on an appropriate analytical
measurement, but the criterion is not rendered useless if an ideal
measurement either is not available or is not feasible.
Note; The analytical chemistry of the material might have to be
taken into account when defining the material or when judging the
acceptability of some toxicity tests, but a criterion should not be
based on the sensitivity of an analytical method. When aquatic
organisms are more sensitive than routine analytical methods, the
proper solution is to develop better analytical methods, not to
underprotect aquatic life.
II. COLLECTION OF MIA
A. Collect all data available on the material concerning toxicity to
aquatic am ma Is and plants, [[giving particular attention to
resident species of the Great Lakes basin] ]. [ (b) FDA action levels
and(c) chronic field studies and long term field studies with
wildlife species that regularly consume aquatic organisms]
3
7
-------
GLI Aquatic Criteria
Methodology 9/5/91
B. All data that are used should be available in typed, dated, and
signed hard copy (publication, manuscript, letter, memorandum, etc.)
with enough supporting information to indicate that acceptable test
procedures were used and that the results are probably reliable. In
some cases, it may be appropriate to obtain written information front
the investigator, if possible. Information that is confidential or
privileged or otherwise not available for distribution should not be
used.
C. Questionable data, whether published or unpublished, should not be
used. For exanple, data should usually be rejected if they are from
tests that did not contain a control treatment, tests in which too
many organisms in the control treatment died or showed signs of
stress or disease, and tests in which distilled or deionized water
was used as the dilution water without the addition of appropriate
salts.
D. Data on technical grade materials may be used if appropriate, but
data on formulated mixtures and emulsifiable concentrates of the
material should not be used.
E. For some highly volatile, hydrolyzable, or degradable materials, it
is probably appropriate to use only results of flow-through tests in
which the concentrations of test material in test solutions were
measured using acceptable analytical methods.
-------
GLI Aquatic Criteria
Methodology 9/5/91
F. Data should be rejected if they were obtained using:
1. Brine shrinp, because they usually only occur naturally in
water with salinity greater than 35g/kg.
2. Species that do not have reproducing wild populations in North
America.
3. Organisms that were previously exposed to substantial
concentrations of the test material or other contaminants.
G. Questionable data, data on formulated mixtures and emulsifiable
titrates, and data obtained with non-resident species or
previously exposed organisms may be used to provide auxiliary
information but should not be used in the derivation of criteria.
III. REQUIRED DMA.
A. Certain data should be available to help ensure that each of the
four ma^or kinds of possible adverse effects receives adequate
consideration. Results of acute and chronic toxicity tests with
representative species of aquatic animals are necessary so that data
available for tested species can be considered a useful indication
of the sensitivities of appropriate untested species. Fewer data
rung toxicity to aquatic plants are required because
procedures for conducting tests with plants and interpreting the
5
-------
GLI Aquatic Criteria
Methodology 9/5/91
results of such tests are not as well developed.
B. To derive a Great Takes Tier 1 Criterion for freshwater aquatic
organisms and their uses, the following should be available:
1. Results of acceptable acute tests (see Section IV) with at
least one species of freshwater animal in at least eight
different families such that all of the following are
included:
a. the family Salmonidae in the class Osteichthyes
b. one other family (preferably a coramercially, or
recreationally important warmwater species) in the class
Osteichthyes (e.g., bluegill, channel catfish, etc.)
c. a third family in the phylum Chordata (e.g., fish,
amphibian, etc.)
d. a planktonic crustacean (e.g., a cladoceran,
copepod, etc.)
e. a benthic crustacean (e.g., ostracod, isopod, amphipod,
crayfish, etc.)
f. an insect (e.g., mayfly, dragonfly, damselfly, stonefly,
caddisfly, mosquito, midge, etc.)
W
-------
GLI Aquatic Criteria
Methodology 9/5/91
g. a family in a phylum other than Arthropoda or Chordata
(e.g., Rot if era, Annelida, Mollusca, etc.)
h. a family in any order of insect or any phylum not
already represented.
2. Acute-chronic ratios (see Section VI) with a species of
aquatic animals in at least three different families provided
that of the three species:
at least one is a fish
at least one is an invertebrate
at least one species is an acutely sensitive freshwater
species (the other two may be saltwater species).
3. Results of at least one acceptable test with a freshwater
algae or vascular plant is desirable but not required for
criterion derivation (see Section VIII). If plants are among
the aquatic organisms most sensitive to the material, results
of a test with a plant in another phylum (division) should
also be available.
4. At least one acceptable bioconcentration factor determined
with an appropriate freshwater species, if a maximum
permissible tissue concentration is available.]
C. If all required data are available, a numerical criterion can
usually be derived, except in some cases. For exanple, derivation
7
-------
GLI Aquatic Criteria
Methodology 9/5/91
of a criterion might not be possible if the available aojte-chronic
ratios vary by more than a factor of ten with no apparent pattern.
Also, if a criterion is to be related to a water quality
characteristic (see Sections V and VII) more data will be required.
Similarly, if all required data are not available, a numeric
criterion should not be derived except in special cases. [For
example, even if enough acute and chronic data are available, it
might not be possible to derive a criterion if the available data
clearly indicate that the Final Residue Value should be much lower
than either the Final Chronic Value or the Final Plant Value.]
D. Confidence m a criterion usually increases as the amount of
available pertinent information increases. Thus, additional data
are usually desirable.
IV. FINAL ACUTE DATA
A. Appropriate measures of the acute (short term) toxicity of the
material to a variety of species of aquatic animals are used to
calculate the Final Acute Value. The Final Acute Value is an
estimate of the concentration of the material corresponding to a
cumulative probability of 0.05 in the acute tenacity values for the
genera with which acceptable acute tests have been conducted on the
material. However, in sane cases, if the Species Mean Acute Value
of a ocranercially, or recreationally important species of the Great
8
-------
GLI Aquatic Criteria
Methodology 9/5/91
System rrsasin]] is lower than the calculated Final Acute
Value, then the Species Mean Acute Value replaces the calculated
Final Acute Value in order to provide protection for that ojiportant
species.
B. Acute tenacity tests should have been conducted using acceptable
procedures.
C. Except for results with saltwater annelids and inysids, results of
acute tests during which the test organisms were fed should not be
used, unless data indicate that the food did not affect the toxicity
of the test material.
D. Results of acute tests conducted in unusual dilution water, e.g. ,
dilution water in which total organic carbon or particulate matter
exceeded 5mg/l, should not be used, unless a relationship is
developed between acute toxicity and organic carbon or particulate
natter, or unless data show that organic carbon or particulate
matter, etc., do not affect toxicity.
E. Acute value; should be based upon endpoints which reflect the total
severe adverse impact of the test material on the organisms used in
the test. Therefore, only the following kinds of data on acute
toxicity to aquatic animals should be used:
1. Tests with daphnids and other cladocerans should be started
9
-------
GLI Aquatic Criteria
Methodology 9/5/91
with organisms less than 24 hours old and tests with nudges
should be started with second or third instar larvae. Die
results should be the 48 hour EC50 based on percentage of
organisms killed. If such an EC50 is not available for a
test, the 48 hour LC50 should be used in place of the desired
48 hour EC50. An BC50 or I£50 of longer than 48 hours can be
used as long as the animals were not fed and the control
animals were acceptable at the end of the test.
2. The results of a test with embryos and larvae of barnacles,
bivalve molluscs (clans, mussles, oysters and scallops), sea
urchins, lobsters, crabs, shrimp and abalones should be the 96
hour EC50 based on percentage of organisms with inccnpletely
developed shells plus percentage of organisms killed. If such
an EC50 is not available from a test, of the values that are
available fron the test the lowest of the following should be
used in place of the desired 96 hour EC50: 48 to 96 hour BCSOs
based on percentage of organisms with inccnpletely developed
shells plus percentage of organisms killed, 48 to 96 hour
BCSOs based upon percentage of organisms with inccnpletely
developed shells, and 48 to 96 hour LCSOs.
3. The result of tests with all other aquatic animal species and
older life stages of barnacles, bivalve molluscs (clams,
mussles, oysters and scallops), sea urchins, lobsters, crabs,
shrimp and abalones should be the 96 hour EC50 based on
10
-------
GLI Aquatic Criteria
Methodology 9/5/91
percentage of organisms exhibiting loss of equilibrium plus
percentage of organisms immobilized plus percentage of
organisms killed. If such an EC50 is not available from a
test, of the values that are available from a test, the lower
of the following should be used m place of the desired 96
hour EC50: the 96 hour BCSO based on percentage of organisms
exhibiting loss of equilibrium plus percentage of organisms
immobilized and the 96 hour L£50.
4. Tests whose results take into account the number of young
produced, such as most tests with protozoans, are not
considered acute tests, even if the duration was 96 hours or
less.
5. If the tests were conducted properly, acute values reported as
'•greater than" values and those which are above the solubility
of the test material should be used, because rejection of such
acute values would bias the Final Acute Value by eliminating
acute values for resistant species.
F. If the acute toxicity of the material to aquatic animals apparently
has been shown to be related to a water quality characteristic such
as hardness or particulate matter for freshwater animals [or
salinity or particulate matter for saltwater animals] a Final Acute
Equation should be derived based on that water quality
characteristic. Go to Section V.
11
JS
-------
GLI Aquatic criteria
Methodology 9/5/91
G. The agreement of the data within and between species should be
considered. Acute values that appear to be questionable in
comparison with other acute and chronic data for the same species
and for other species in the same genus probably should not be used.
For example, if the acute values available for a species or genus
differ by more than a factor of 10, rejection of some or all of the
values is probably appropriate.
H. If the available data indicate that one or more life stages are at
least a factor of two or more resistant than one or more other life
stages of the same species, the data for the more resistant life
stages should not be used in the calculation of the Species Mean
Acute Value because a species cannot be considered protected from
acute toxicity if all the life stages are not protected.
I. For each species for which at least one acute value is available,
the Species Mean Acute Value (SMAV) should be calculated as the
geometric mean of the results of all flow-through tests in which the
concentrations of test material were measured. For a species for
which no such result is available, the SMAV should be calculated as
the geometric of all available acute values, i.e., results of flow-
through tests in which the concentrations were not measured and
results of static and renewal tests based on initial concentrations
(nominal concentrations are acceptable for most test materials if
measured concentrations are not available) of test material.
12
A,
-------
GLI Aquatic Criteria
Methodology 9/5/91
Note; Data reported by original investigators should not be rounded
off. Results of all intermediate calculations should be rounded off
to four significant digits.
Note: Ihe geometric mean of N numbers is the Nth root of the
product of the N numbers. Alternatively, the geometric mean can be
calculated by adding the logarithms of the N numbers, dividing the
sum by N, and taking the antilog of the quotient. Ihe geometric
mean of two numbers is the square root of the product of the two
numbers, and the geometric mean of one number is that number.
Either natural (base e) or common (base 10) logarithms can be used
to calculate geometric means as long as they are used consistently
within each set of data, i.e. the antilog used must match the
logarithms used.
Note; Geometric means, rather than arithmetic means are used here
because the distributions of sensitivities of individual organisms
in toxicity tests on most materials and the distributions of
sensitivities, of species within a genus are more likely to be
lognormal than normal. Similarly, geometric means are used for
acute-chronic ratios [and bioooncentration factors] because
quotients are likely to be closer to lognormal than normal
distributions. In addition, division of the geometric mean of a set
of numerators by the geometric mean of the set of denominators will
result in the geometric mean of the set of corresponding quotients.
13
-------
GLI Aquatic Criteria
Methodology 9/5/91
J. For each genus for which one or more SMAVs are available, the Genus
Mean Acute Value (GMAV) should be calculated as the geometric mean
of the SMAVs available for the genus.
K. Order the GMAVs from high to low.
L. Assign ranks, R, to the GMAVs fron "1" for the lowest to "N" for the
highest. If two or more GMAVs are identical, arbitrarily assign
them successive ranks.
M. Calculate the cumulative probability, P, for each GMAV as R/(N + 1).
N. Select the four GMAVs which have cumulative probabilities closest to
0.05 (if there are less than 59 GMAVs, these will always be the four
lowest GMAVs).
O. Using the selected GMAVs, and Ps, calculate
S2 = £ ((In GMAV)2) - ((2 (In GMAV))2/4)
L =
A = SCfojQS) + L
EAV = e*
Note: Natural logarithms (logarithms to base e, denoted as In) are
used herein merely because they are easier to use on some hand
14
-------
GLI Aquatic Criteria
Methodology 9/5/91
calculators and conputers than common (base 10) logarithms.
Consistent use of either will produce the same result.
P. If for a oonraercially, or recreationally important species of the
Great lake** System [ FBasinT 1 the geometric mean of the acute values
from flow through tests in which the concentrations of test material
were measured is lower than the calculated Final Acute Value, then
that geometric mean should be used as the Final Acute Value instead
of the calculated Final Acute Value.
Q. Go to Section VI.
V. FINAL ACUTE
A. When enough data are available to show that acute toxicity to two or
more species is similarly related to a water quality characteristic,
the relationship should be taken into account as described in
Sections B - G below or using analysis of covariance. The two
methods are equivalent and produce identical results. The manual
method described below provides an understanding of this application
of covarianoe analysis, but computerized versions of covariance
analysis are much more convenient for analyzing large data sets. If
two or more factors affect toxicity, multiple regression analysis
should be used.
B. For each species for which comparable acute toxicity values are
15
-------
GLI Aquatic Criteria
Methodology 9/5/91
available at two or more different values of the water quality
characteristic, perform a least squares regression of the acute
toxicity values on the corresponding values of the water quality
characteristic to obtain the slope and its 95% confidence limits for
each species.
Note: Because the best documented relationship is that between
hardness and acute toxicity of metals in fresh water and a log-log
relationship fits these data, geometric means and natural logarithms
of both toxicity and water quality are used in the rest of this
section. For relationships based on other water quality
characteristics, such as pH, temperature, or salinity, no
transformation or a different transformation might fit the data
better, and appropriate changes will be necessary throughout this
section.
c. Decide whether the data for each species is useful, taking into
account the range and number of the tested values of the water
quality characteristic and the degree of agreement within and
between species. For example, a slope based on six data points
might be of limited value if it is based only data for a very narrow
range of values of the water quality characteristic. A slope based
on only two data points, however, might be useful if it is
consistent with other information and if the two points cover a
broad enough range of the water quality characteristic.
In addition, acute values that appear to be questionable in
16
-------
GLI Aquatic Criteria
Methodology 9/5/91
comparison with other acute and chronic data available for the same
species and for other species in the same genus probably should not
be used. For example, if after adjustment for the water quality
characteristic, the acute values available for a species or genus
differ by more than a factor of 10, rejection of sane or all of the
values is probably appropriate. If useful slopes are not available
for at least one fish and one invertebrate or if the available
slopes are too dissimilar or if too few data are available to
adequately define the relationship between acute toxicity and the
water quality characteristic, return to Section IV.G., using the
results of tests conducted under conditions and in waters similar to
those commonly used for toxicity tests with the species.
D. Individually for each species calculate the geometric mean of the
available acute values and then divide each of the acute values for
a species by the mean for the species. This normalizes the acute
values so that the geometric mean of the normalized values for each
species individually and for any combination of species is 1.0.
£. Similarly normalize the values of the water quality characteristic
for each species individually.
F. Individually for each species perform a least squares regression of
the normalized acute values of the water quality characteristic.
The resulting slopes and 95% confidence limits will be identical to
those obtained in Section B above. Now, however, if the data are
17
-------
GLI Aquatic Criteria
Methodology 9/5/91
actually plotted, the line of best fit for each individual species
will go through the point 1,1 in the center of the graph.
G. Treat all the normalized data as if they were all for the same
species and perform a least squares regression of all the normalized
acute values on the corresponding normalized values of the water
quality characteristic to obtain the pooled acute slope, V, and its
95% confidence limits. If all the normalized data are actually
plotted, the line of best fit will go through the point 1,1 in the
center of the graph.
H. For each species calculate the geonetric mean, W, of the acute
toxicity values and the geometric mean, X, of the values of the
water quality characteristic. (These were calculated in steps D and
£ above).
I. For each species calculate the logarithm, Y, of the SMAV at a
selected value, Z, of the water quality characteristic using the
equation:
Y - In W - V(ln X - In Z)
J. For each species calculate the SMAV at X using the equation:
SMAV = eY
Note; Alternatively, the SMAVs at Z can be obtained by skipping
step H above, using the equations in steps I and J to adjust each
18
-------
GLI Aquatic Criteria
Methodology 9/5/91
acute value individually to Z, and then calculating the geometric
mean of the adjusted values for each species individually. This
alternative procedure allows an examination of the range of the
adjusted acute values for each species.
K. Obtain the Final Acute Value at Z by using the procedure described
in Section IV.J-O.
L. [If the SMAV at Z of a ccntnercially, or recreationally important
species of the Great" Takes Basin is lower than the calculated Final
Acute Value at Z, then the 3RV should be used as the Final Acute
Value at Z instead of the calculated Final Acute Value. ]
If f
or a
ionally uipurtant species of the
svsteni the qeciuetz''^ T^^^n of the acute values at Z frcni
were
"8s lower than the
5tions of the test
Final Acute Value at Z.
instead of the calculated Final Acute Value.
M. The Final Acute Equation is written as: Final Acute Value =
e(V[ln(water quality characteristic) ] + A - V[ln Z]) / ^Sf& y = p^i^ acute
slope and A = In (Final Acute Value at Z). Because V, A, and Z are
known, the Final Acute Value can be calculated for any selected
value of the water quality characteristic.
19
-------
GLI Aquatic Criteria
Methodology 9/5/91
VI. FINAL CHRCNIC VALUE
A. Depending on the data that are available concerning chronic toxicity
to aquatic animals, the Final Chronic Value might be calculated in
the same manner as the Final Acute Value or by dividing the Final
Acute Value by the Final Acute to Chronic Ratio. In some cases it
may not be possible to calculate a Final Chronic Value for Tier 1.
Note: As the name implies, the acute-chronic ratio (ACR) is a way
of relating acute and chronic toxicities. The acute-chronic ratio
is basically the inverse of the application factor, but this new
name is better because it is more descriptive and should help
prevent confusion between "application factors" and "safety
factors". Acute-chronic ratios and application factors are ways of
relating the acute and chronic toxicities of a material to aquatic
organisms. Safety factors are used to provide an extra margin of
safety beyond the known or estimated sensitivities of aquatic
organisms. Another advantage of the acute-chronic ratio is that it
will usually be greater than one; this should avoid the confusion as
to whether a large application factor is one that is close to unity
or one that has a denominator that is much greater than the
numerator.
B. Chronic values should be based on results of flow-through (except
renewal is acceptable for daphnids) chronic tests in which the
concentration of test material in the test solutions were properly
20
-------
GLI Aquatic Criteria
Methodology 9/5/91
measured at appropriate tines during the test.
C. Results of chronic tests in which survival, growth, or reproduction
in the control treatment was unacceptably low should not be used.
The limits of acceptability will depend on the species.
D. Results of chronic tests conducted in unusual dilution water, e.g.,
dilution water in which total organic carbon or particulate matter
exceeded 5 mg/1, should not be used, unless a relationship is
developed between chronic toxicity and organic carbon or particulate
matter, or unless data show that organic carbon, particulate matter,
etc., do not affect toxicity.
E. Chronic values should be based on endpoints and lengths of exposure
appropriate to the species. Therefore, only results of the
following kirds of chronic toxicity tests should be used:
1. Life-cycle toxicity tests consisting of exposures of each of
two or more groups of individuals of a species to a different
concentration of the test material throughout a life cycle.
To ensure that all life stages and life processes are exposed,
tests with fish should begin with embryos or newly hatched
young less than 48 hours old, continue through maturation and
reproduction, should end not less than 24 days (90 days for
salmonids) after the hatching of the next generation. Tests
with daphnids should begin with young less than 24 hours old
21
-------
GLI Aquatic Criteria
Methodology 9/5/91
and last for not less than 21 days, for Ceriodaphnia not less
than 7 days. Tests with mysids should begin with young less
than 24 hours old and continue until 7 days past the median
tune of first brood release in the controls. For fish, data
should be obtained and analyzed on survival and growth of
adults and young, maturation of males and females, eggs
spawned per female, embryo viability (salmonids only), and
hatchability. For daphnids, data should be obtained and
analyzed on survival and young per female. For mysids, data
should be obtained and analyzed on survival, growth, and young
per female.
2. Partial life-cycle tenacity tests consist of exposures of each
of two more groups of individuals of a species of fish to a
different concentration of the test material through most
portions of a life cycle. Partial life-cycle tests are
allowed with fish species that require more than a year to
reach sexual maturity, so that all major life stages can be
exposed to the test material in less than 15 months. Exposure
to the test material should be begin with immature juveniles
at least 2 months prior to active gonad development, continue
through maturation and reproduction, and end not less than 24
days (90 days for salmonids} after the hatching of the next
generation. Data should be obtained and analyzed on survival
and growth of adults and young, maturation of males and
females, eggs spawned per female, embryo viability (salmonids
22
-------
GLI Aquatic Criteria
Methodology 9/5/91
only), and hatchability.
3. Early life-stage tenacity tests consisting of 28 to 32 day (60
days post hatch for salmonids) exposures of the early life
stages, of a species of fish from shortly after fertilization
through embryonic, larval, and early juvenile development.
Data should be obtained and analyzed on survival and growth.
tjpte; Results of an early life-stage test are used as
predictions of results of life-cycle and partial life-cycle
tests with the sane species. Therefore, when results of a
life-cycle or partial life-cycle test are available, results
of an early life-stage test with the sane species should not
be used. Also, results of early life-stage tests m which the
incidence of mortalities or abnormalities increased
substantially near the end of the test should not be used
because the results of such tests are possibly not good
predictions of comparable life-cycle or partial life-cycle
tests.
F. A chronic value nay be obtained by calculating the geometric mean of
the lower and upper chronic limits from a chronic test or by
analyzing chronic data using regression analysis. A lower chronic
limit is the* highest tested concentration (a) in an acceptable
chronic test, (b) which did not cause an unacceptable amount of
adverse effect on any of the specified biological measurements, and
23
-------
GLI Aquatic Criteria
Methodology 9/5/91
(c) below which no tested concentration caused an unacceptable
effect. An upper chronic limit is the lowest tested concentration
(a) in an acceptable chronic test, (b) which did not cause an
unacceptable amount of adverse effect on one or more of the
specified biological measurements, and (c) above which all testpri
concentrations also caused such an effect.
Note; Because various authors have used a variety of terms and
definitions to interpret and report results of chronic tests,
reported results should be reviewed carefully. The amount of effect
that is considered unacceptable is often based on a statistical
hypothesis test, but might also be defined in terms of a specified
percent reduction from the controls. A small percent reduction
(e.g., 3%) might be considered acceptable even if it is
statistically significantly different from the control, whereas a
large percent reduction (e.g. , 30%) might be considered unacceptable
even if it is not statistically significant.
G. If the chronic toxicity of the material to aquatic animals
apparently has been shown to be related to a water quality
characteristic such as hardness or particulate matter for freshwater
animals [or salinity or particulate matter for saltwater animals] a
Final Chronic Equation should be derived based on that water quality
characteristic. Go to Section VII.
H. If chronic values are available for species in eight families as
24
-------
GLI Aquatic Criteria
Methodology 9/5/91
described in Section III.B.I. a Species Mean Chronic Value (SMCV)
stolid be ca]culated for each species for which at least one chronic
value is available by calculating the geometric mean of all chronic
values availcible for the species, and appropriate Genus Mean Chronic
Values should be calculated. The Final Chronic Value should be
obtained using the procedure described in Section IV.J-O. Then go
to Section VI .M.
I. For each chronic value for which at least one corresponding
appropriate acute value is available, calculate an acute-chronic
ratio, using for the numerator the geometric mean of the results of
all acceptable flow-through (except static is acceptable for
daphnids and midoesl acute tests in the same dilution water and
which the conaentrations were measured. For fish, the acute test(s)
should have been conducted with juveniles. The acute test(s) should
have been part of the sane study as the chronic test. If acute
tests were not conducted as part of the same study, acute tests
conducted in the same laboratory and dilution water, but in a
different study, may be used. If no such acute tests are available,
results of acute tests conducted in the same dilution water in a
different laboratory may be used. If no such acute tests are
available, an acute-chronic ratio should not be calculated.
J. For each species, calculate the species mean acute-chronic ratio as
the geometric mean of all acute-chronic ratios available for that
species.
25
-------
GLI Aquatic Criteria
Methodology 9/5/91
K. For some materials the acute-chronic ratio seems to be the same for
all species, but for other materials the ratio seems to increase or
decrease as the Species Mean Acute Value (SMAV) increases. Thus the
Final Acute-Chronic Ratio can be obtained in four ways, depending on
the data available:
1. If the species mean acute-chronic ratio seems to increase or
decrease as the SMAVs increases, the Final Acute-Chronic Ratio
should be calculated as the geometric mean of the acute-
chronic ratios for species whose SMAVs are close to the Final
Acute Value.
2. If no major trend is apparent and the acute-chronic ratios for
[a number of species] all species are within a factor of ten,
the Final Acute-Chronic Ratio should be calculated as the
geometric mean of all the species mean acute-chronic ratios
for both freshwater and saltwater species.
3. For acute tests conducted on metals and possibly other
substances with embryos and larvae of barnacles, bivalve
molluscs, sea urchins, lobsters, crabs, shrimp, and abalones
(see Section IV.E.2), it is probably appropriate to assume
that the acute to chronic ratio is 2. Chronic tests are very
difficult to conduct with most such species, but it is likely
that the sensitivities of embryos and larvae would determine
the results of life cycle tests. Thus, if the lowest
26
-------
GLI Aquatic Criteria
Methodology 9/5/91
available SMAVs were determined with embryos and larvae of
such species, the Final Chronic Value is equal to the
Criterion Maximum Concentration (see Section IX. B) .
4. If the most appropriate species mean acute-chronic ratios are
less than 2.0, and especially if they are less than 1.0,
acclimation has probably occurred during the chronic test.
Because continuous exposure and acclimation cannot be assured
to provide adequate protection in field situations, the Final
Acute-Chronic Ratio should be assumed to be 2, so that the
Final Chronic Value is equal to the Criterion Maximum
Concentration (see Section X.B) .
If the available species mean acute-chronic ratios do not fit one of
these cases, a Final Acute-Chronic Ratio probably cannot be obtained
and a Final Chronic Value probably cannot be calculated for Tier 1.
Pref eroTV"^ ^ypMl^ be given to fr^shwat^r (vgrsng saltwater)
ratios whgn calc^ila'tincr a Final Acute—Chronic Ratio.
' - c ratio data
In TTT.B.2.} are not met with freshwater data alone, saltwater data
L. Calculate the Final Chronic Value by dividing the Final Acute Value
by the Final Acute-Chronic Ratio. If there was a Final Acute
27
-------
GLI Aquatic Criteria
Methodology 9/5/91
Equation rather than a Final Acute Value, see also Section VILA.
M. If the Species Mean Chronic Value of a commercially or
recreationally important species of the Great Lakes r TBasinl 1 System
is lower than the calculated Final Chronic Value, then that Species
Mean Chronic Value should be used as the Final Chronic Value instead
of the calculated Final Chronic Value.
N. Go to Section VIII.
VII. FINAL CHRONIC EQUATION
A. A Final Chronic Equation can be derived in two ways. The procedure
described here in Section A will result in the chronic slope being
the same as the acute slope. The procedure described in Sections B-
N will usually result in the chronic slope being different from the
acute slope.
1. If acute-chronic ratios are available for enough species at
enough values of the water quality characteristic to indicate
that the acute-chronic ratio is probably the same for all
species and is probably independent of the water quality
characteristic, calculate the Final Acute-Chronic Ratio as the
geometric mean of the available species mean acute-chronic
ratios.
28
-------
GLI Aquatic Criteria
Methodology 9/5/91
2. Calculate the Final Chronic Value at the selected value Z of
the water quality characteristic by dividing the Final Acute
Value at Z. (see Section V.M.) by the Final Acute-Oironic
Ratio.
3. Use V == pooled acute slope (see Section V.M.) as
L = pooled chronic slope.
4. Go to Section VTI.M.
B. When enough data are available to show that chronic toxicity to at
least one species is related to a water quality characteristic, the
relationship should be taken into account as described in Sections
B-G below or using analysis of covariance. The two methods are
equivalent and produce identical results. The manual method
described below provides an understanding of this application of
covariance analysis, but computerized versions of covariance
analysis are Touch more convenient for analyzing large data sets. If
two or more factors affect toxicity, multiple regression analysis
should be used.
C. For each species for which comparable chronic toxicity values are
available at two or more different values of the water quality
characteristic, perform a least squares regression of the chronic
toxicity values on the corresponding values of the water quality
characteristic to obtain the slope and its 95% confidence limits for
29
-------
GLI Aquatic Criteria
Methodology 9/5/91
each species.
*: Because the best documented relationship is that between
hardness and acute toxicity of metals in fresh water and a log-log
relationship fits these data, geometric means and natural logarithms
of both toxicity and water quality are used in the rest of this
section. For relationships based on other water quality
characteristics, such as pH, temperature, or salinity, no
transformation or a different transformation might fit the data
better, and appropriate changes will be necessary throughout this
section. It is probably preferable, but not necessary, to use the
same transformation that was used with the acute values in
Section V.
D. Decide whether the data for each species is useful, taking into
account the range and number of the tested values of the water
quality characteristic and the degree of agreement within and
between species. For example, a slope based on six data points
might be of limited value if it based only on data for a very narrow
range of values of the water quality characteristic. A slope based
on only two data points, however, might be more useful if it is
consistent with other information and if the two points cover a
broad range of the water quality characteristic. In addition,
chronic values that appear to be questionable in comparison with
other acute and chronic data available for the same species and for
other species in the same genus probably should not be used. For
30
-------
GLI Aquatic Criteria
Methodology 9/5/91
example, if after adjustment for the water quality characteristic,
the chronic values available for a species or genus differ by more
than a factcar of 10, rejection of sane or all of the values is
probably appropriate. If a useful chronic slope is not available
for at least one species or if the available slopes are too
dissimilar or if too few data are available to adequately define the
relationship between chronic toxicity and the water quality
characteristic, it might be appropriate to assume that the chronic
slope is this same as the acute slope, which is equivalent to
assuming that the acute-chronic ratio is independent of the water
quality characteristic. Alternatively, return to Section VI.H.,
using the results of tests conductpfl under conditions and in waters
similar to those commonly used for toxicity tests with the species.
E. Individually for each species calculate the geometric mean of the
available chronic values and then divide each chronic value for a
species by the mean for the species. This normalizes the chronic
values so that the geometric mean of the normalized values for each
species individually and for any combination of species is 1.0.
F. Similarly normalize the values of the water quality characteristic
for each species individually.
G. Individually for each species perform a least squares regression of
the normalized chronic toxicity values on the corresponding
normalized values of the water quality characteristic. The
31
35"
-------
GLI Aquatic Criteria
Methodology 9/5/91
resulting slopes and the 95% confidence limits will be identical to
those obtained in Section B above. Now, however, if the data are
actually plotted, the line of best fit for each individual species
will go through the point 1,1 in the center of the graph.
H. Treat all the normalized data as if they were all the same
species and perform a least squares regression of all the
normalized chronic values on the corresponding normalized
values of the water quality characteristic to obtain the
pooled chronic slope, L, and its 95% confidence limits.
If all normalized data are actually plotted, the line of best fit
will go through the point 1,1 in the center of the graph.
L. For each species calculate the geometric mean, M, of the toxicity
values and the geometric mean, P, of the values of the water quality
characteristic. (These were calculated in steps E and F above.)
J. For each species calculate the logarithm, Q, of the Species Mean
Chronic Value at a selected value, Z, of the water quality
characteristic using the equation:
Q - In M - L{ln P - In 2)
: Although it is not necessary, it will usually be best to use
the same value of the water quality characteristic here as was used
in Section V.I.
32
-------
GLI Aquatic Criteria
Methodology 9/5/91
K. For each species calculate a Species Mean Chronic Value at Z using
the equation;
SMCV » e°
Note; Alternatively, the Species Mean Acute Chronic Value at Z can
be obtained by skipping step J above, using the equations in steps
J and K to adjust each acute value individually to Z and then
calculating the geometric means of the adjusted values for each
species individually. This alternative procedure allows an
examination of the range of the adjusted chronic values for each
species.
L. Obtain the Final Chronic Value at Z by using the procedure described
in Section IV.J-O.
M. If the Species Mean Value at Z of a ccranercially or recreationally
uqportant species of the Great T.?!?*>**? [ [Paginl T sys^*8*** is lower than
the calculated Final Chronic Value at Z, then that Species Mean
Chronic Value should be used as the Final Chronic Value at Z instead
of the calculated Final Chronic Value.
N. The Final Chronic Equation is written as:
Final Chronic Value =
e(L( In (water quality characteristic)] + InS- LflnA])
where L = pooled chronic slope and S = Final Chronic Value at Z.
Because L, S,, and Z are known, the Final Chronic Value can be
33
-------
GLI Aquatic Criteria
Methodology 9/5/91
calculated for any selected value of the water quality
characteristic.
VIII. FINAL PLAOT VALUE
A. Appropriate measures of the toxicity of the material to aquatic
plants are used to compare the relative sensitivities of aquatic
plants and animals. Although procedures for conducting and
interpreting the results of toxicity tests with plants are not well
developed, results of tests with plants usually indicate that
criteria which adequately protect aquatic animals and there uses
will probably also protect aquatic plants and their uses.
B, A plant value is the result of a 96 hour test conducted with an alga
or a chronic test conducted with an aquatic vascular plant.
Note: A test of the toxicity of a metal to a plant usually should
not be used if the medium contained an excessive amount of a
completing agent, such as EDIA, that might affect the toxicity of
the metal. Concentrations of EDEA above 200 ug/1 should probably be
considered excessive.
C. The Final Plant Value should be obtained by selecting the lowest
result from a test with an important aquatic plant species in which
the concentrations of test material were measured and the endpoint
was biologically important.
34
-------
GLI Aquatic Criteria
Methodology 9/5/91
TOE EWTIBE SECTION DEMJMS WTffl THE CaiTULATION OF THE FINAL RESIDUE VALUE HAS
_PJ
IX. OltifcK DATA
Pertinent information that could not be used in earlier sections might be
available concerning adverse effects on aquatic organisms and their uses.
The most important of these are data on cumulative and delayed toxicity,
reduction in survival, growth, or reproduction, or any other adverse
effect that has been shown to be biologically mportant. Especially
important are data for species for which no other data are available.
Data from behavioral, biochemical, physiological, microcosm, and field
studies might also be available. Data might be available from tests
conducted in unusual dilution water (see IV.D and VI.D), from chronic
tests in which the concentrations were not measured (see VI.B), from tests
with previously exposed organisms (see II.F), and from tests on formulated
mixtures or amlsifiable concentrates (see II. D). Such data might affect
a criterion if the data were obtained with an important species, the test
concentrations were measured, and the endpoint was biologically important.
X. CRITERION
A. A criterion consists of two concentrations: the Criterion Maximum
Concentration and the Criterion continuous Concentration.
35
-------
GLI Aquatic Criteria
Methodology 9/5/91
B. The Criterion Maximum Concentration (CMC) is equal to one-half the
Final Acute Value.
C. The Criterion Continuous Concentration (CCC) is equal to the lowest
of the Final Chronic Value or the Final Plant Value [or the Final
Residue Value] (if available) unless other data (see Section DC)
show that a lower value should be used. If toxicity is related to
a water quality characteristic, the CCC is obtained from the Final
Chronic Equation or Final Plant Value (if available) that results in
the lowest concentrations in the usual range of the water quality
characteristic, unless other data (see Section DC) show that a lower
value should be used.
D. Round both the CMC and the CCC to two significant digits.
£. The criterion is stated as:
The procedures described in the Tier 1 methodology indicate that,
except possibly where a locally important species is very sensitive,
aquatic organisms and their uses should not be affected unacceptably
if the four-day average concentration of (l) does not exceed (2)
ug/1 more than once every three years on the average and if the one-
hour average concentration does not exceed
(3) ug/1 more than once every three years on the average.
where:
(l) = insert name of material
(2) = insert the Criterion Continuous Concentration
36
-------
GLI Aquatic Criteria
Methodology 9/5/91
(3) = insert the Criterion Maximum Concentration
F. r (Tables 2 and 3 contain the acute and chronic criteria.
XI. FINAL REVIEW
A. The derivation of the criterion should be carefully reviewed by
rechecking each step of the Guidelines. Items that should be
especially checked are:
1. If unpublished data are used, are they well documented?
2. Are all required data available?
3. Is the range of acute values for any species greater than a
factor of 10?
4. Is the range of Species Mean Acute Values for any genus
greater than a factor of 10?
5. Is there more than a factor of 10 difference between the four
lowest Genus Mean Acute Values?
6. Are any of the lowest Genus Mean Acute Values questionable?
7. Is the Final Acute Value reasonable in comparison with the
Species Mean Acute Values and Genus Mean Acute Values?
8. For any ccnmercially or recreationally important species gf.
the Great- fr*frfg Pft?in is the geometric mean of the acute
values from flow-through tests in which the concentrations of
test material were measured lower than the Final Acute Value?
9. Are any of the chronic values questionable?
37
-------
GLI Aquatic Criteria
Methodology 9/5/91
10. Are any dhronic values available for acutely sensitive
species?
11. Is the range of acute-chronic ratios greater than a factor of
10?
12. Is the Final Chronic Value reasonable in comparison with the
available acute and chronic data?
13. Is the measured or predicted chronic value for any
ccnmercially or recreationally important species of the Great
Calces Basm below the Final Chronic Value?
14. Are any of the other data important7
15. Do any data look lite they might be outliers?
16. Are there any deviations from the Guidelines? Are they
acceptable0
B. On the basis of all available pertinent laboratory and field
information, determine if the criterion is consistent with sound
scientific evidence. If it is not, another criterion, either higher
or lower, should be derived using appropriate modifications of these
Guidelines.
38
-------
GLI Aquatic Criteria
Methodology 9/5/91
Great Lakes Water Quality initiative
Procedure for Deriving Aquatic Life Criteria
Tier 2
I. If all eight minimum data requirements for calculating an FAV using Tier
1 are not net, a secondary acute value (SAV) for the waters of the Great
Basin shall be calculated for a chemical as follows:
To calculate a SAV, the most sensitive GMRV in the data base is
divided by the Final Acute Value Factor (FAVF) (Table 1)
corresponding to the number of satisfied imnintim data requirements
listed in the Tier I methodology (Section III (B) (1) ) . If all eight
requirements are satisfied, a tier 1 criterion
calculation may be possible. In order to calculate a SAV, the data
base mist contain, at a wininmn, a genus mean acute value (GMAV) for
one of the following three genera in the family Daphnidae -
Cerodaphnia sp. , Daphnia sp. , or Simocephalus sp. .
Table 1. Final Acute Value Factors
Number of Satisfied Mini mm
FAVF
1 20
2 13
3 8.6
39
-------
GLI Aquatic Criteria
Methodology 9/5/91
4 6.5
5 5.0
6 4.0
7 3.6
If appropriate, the SAV will be made a function of a water quality
characteristic in a manner similar to that described in Tier 1.
II. If three or more experimentally determined acute chronic ratios (ACRs)
which are acceptable based on Tier 1 are available for the chemical,
determine the Final Acute-Chronic Ratio (FACK) using the procedure
described in Tier 1. If fewer than three acceptable experimentally
determined ACRs are available, use enough assumed ACRs of 18 so that the
total number of ACRs equals three. Calculate the Secondary Acute-Chronic
Ratio (SACK) as the geometric mean of the three ACRs. Thus, if no
experimentally determined acute-chronic ratios are available, the
SACK is 18.
III. Calculate the Secondary Chronic Value (SCV) by dividing the FAV (fron
Tier 1) by the SACK, or the SAV by the PACK or the SACK.
IV. Secondary Criteria shall consist of two concentrations: the Secondary
Maximum Concentration (SMC) and the Secondary Continuous Concentration
(SOC); and shall be derived in the same manner as described in Tier 1
(Section X).
40
-------
GLI Aquatic Criteria
Methodology 9/5/91
V. if for a commercially or recreationally important species of the Great
Takes Basin the geometric mean of the acute values fron flow-through tests
in which the concentrations of the test materials were measured is lower
than the calculated SAV then that geometric mean should be used as the SAV
instead of the calculated SAV.
VI. If for a ocranercially or recreationally important species of the Great
T»r«*g Basin the geometric mean of the chronic values from flow-through
tests in which the concentrations of the test materials were measured is
lower than the calculated SCV then that geometric mean should be used as
the SCV instead of the calculated SCV.
VII. On the basis of all available pertinent laboratory and field information,
determine if the criterion is consistent with sound scientific evidence.
If it is not, another criterion, either higher or lower, should be derived
using appropriate modifications of these procedures.
[ [VIII. Tables 4 and 5 contain acute and chronic criteria, respectively, that
are calculated using these Tier 2 procedures]]
VIII. the most lemait secondary criteria shall be complied on an annual basis by
U.S.EPA Region V Vfoter Division and be available for distribution to the
public.
41
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier l Aquatic Life Criteria for Cadmium
Class 013
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for cadmium (EPA, 1986)
led to the addition of new acute and chronic toxicity data to
the EPA data base. These data are found in Tables l and 2,
respectively. The combined data were used as the basis for the
Tier 1 criteria calculations. The toxicity of cadmium is
hardness dependent. Therefore, all data used in the criteria
calculations were normalized to a hardness of 50 mg/1.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (Table 3). This analysis resulted in
an FAV of 4.591 ug/1 at a hardness of 50 mg/1. The CMC was
calculated by dividing the FAV by 2, resulting in a CMC of 2.3
ug/1.
For waterbodies protected for salmonid fisheries, the FAV was
lowered to 4.25 ug/1 to protect the commercially and
recreationally important Chinook salmon. The CMC is 2.1 ug/1.
The final acute equation for nonsalmonid fisheries is:
1.128 (InH) - 2.889
e
The final acute equation for salmonid fisheries is:
1.128 (InH) - 2.966
e
Criterion Continuous Concentration (CCC)
The Final Chronic Value (FCV) was developed following the
approach given in the EPA criteria document. Due to the large
variety of species with which chronic tests have been
conducted, the FCV was calculated in the same way as the FAV by
using the ranked approach. The ranked chronic data are shown in
Table 4.
The resulting FCV is 0.3166 ug/1 at a hardness of 50 mg/1. The
CCC is 0.32 ug/1. There was no need to lower this value to
protect commercially or recreationally important species of the
Great Lakes basin. The final chronic equation is:
0.7852 (InH) - 4.222
e
Summary
The cadmium criterion are:
-------
GLWQI
For nonsalmonid fisheries:
FAV = 4.591 (at hardness = 50 mg/1)
CMC = 2.3 ug (at hardness - 50 mg/1)
FCV = 0.3166 ug/1 (at hardness * 50 mg/1)
CCC = 0.32 ug/1 (at hardness = 50 mg/1)
1.128 (In H) - 2.889
Final Acute Equation = e
0.7852(ln H) - 4.222
Final Chronic Equation = e
For salmonid fisheries:
FAV - 4.25 ug/1 (at hardness - 50 mg/1)
CMC =2.1 ug/1 (at hardness • 50 mg/1)
FCV = 0.3166 ug/1 (at hardness = 50 mg/1)
CCC =0.32 ug/1 (at hardness = 50 mg/1)
1.128 (In H) - 2.966
Final Acute Equation « e
0.7852 (In H) - 4.222
Final Chronic Equation = e
EPA
FAV = 3.589 ug/1 (at hardness = 50 mg/1)
CMC = 1.7945 ug/1 (at hardness = 50 mg/1)
FCV = 0.6582 ug/1 (at hardness = 50 mg/1)
1.121 (In H) - 3.13
Final Acute Equation = e
0.7852 (In H) - 3.490
Final Chronic Equation = e
References
EPA, 1985. Ambient Aquatic Life Water Quality Criteria for
Cadmium. EPA 440/5-84-032. January 1985.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure
for Deriving Aquatic Life Criteria. April, 1991
Biesinger, K.E. and G.M. Christensen. 1972. Effects of various
metals on survival, growth, reproduction, and metabolism of
Daphnia magna. Jour. Fish. Res. Board Can. 29: 1691
Cusimano, R.F., D.F. Brakke and G.A. Chapman. 1986. Effects of
pH on the toxicities of cadmium, copper and zinc to steelhead
trout (Salmo gardnezi). Can J. Fish. Aquati. Sci. 43: 1497-1503
Elnabarawy, M.R., A.N. Wetter and R.R. Rabidea. 1986. Relative
sensitivity of three daphnid species to selected organic and
inorganic chemicals. Env. Tox. Chem. 5:393-398
Hall, W.S., R.L. Paulson, L.W. Hall Jr., and D.T. Burton. 1986.
Acute toxicity of cadmium and sodium pentachlorphenate to
Daphnids and fish. Bull. Environ. Contam. Toxicol. 37:308-316
Ingersoll, C.G. and R.W. Winner. 1982. Effect of Daphnia pulex
(DeGeer) to daily pulse exposures to copper or cadmium. , t ^\
Environ. Toxicol. Chem. 1:321 Li (
-------
Khangarot, B.S., and P.K. Ray. 1987. Correlation between heavy
metal acute toxicity values in Daphnia magna and fish. Bull.
Env. Contam. Toxicol. 38:722-726
Martin, S.R. and D.M. Holdich. 1986. The acute lethal toxicity
of heavy metals to peracarid crustaceans (with particular
reference to freshwater asellids and gammarids). Wat. Res.,
20f9): 1137-1147.
Mirenda. R.J. 1986. Toxicity and accumulation of cadmium in the
crayfish. Orconected virilis (hagen). Arch. Env. Contam. Tox.,
15(4):401-407
Niederlehner,B. 1984. Cadmium toxicity to a cladoceran. In: A
comparison of techniques for estimating the hazard of chemicals
in the aquatic environment. M.S. Thesis. Virginia Polytechnic
Institute and State University.
Palawski, D., J.B. Hunn and F.J. Dwyer. 1985. Sensitivity of
young striped bass to organic and inorganic contaminants in
fresh and saline waters. Trans. Am. Fish. Soc. 114:748-753
VanLeewen, C.J, et al.. 1985. Differences in susceptibility of
early life stages of rainbow trout (Salmo gairdneri)
environmental pollutants. Aq. Tox. 7:59-78.
-------
Table 1. Acute Data Generated for Cadmium Subsequent to the 1984 EPA Ambient Hater Quality Criteria Document
(EPA, 1985)
Species
Hardness Adjusted
(mg/1 as LC50/EC50 Acute Value
Method* Chemical CaCO3) (ug/1) (ug/1)**** Reference
Cladoceran, S,\J
Ceriodaphnia reticulata
Cladoceran, S,U
Daphnia pulex
Cladoceran, S,U
Daphnia pulex
Cladoceran, S,U
Daphnia pulex
Cladoceran, S,U
Daphnia pulex
Cladoceran, S,tf
Daphnia magna
Amphipod, S,U
Crangonyx pseudgracilis
Crayfish, S,U
Orconectes virilis
Rainbow trout, PT,M
Oncorhynchus mykiss
Rainbow trout, FT,M
Oncorhynchus mykiss
Rainbow trout, FT,N
Oncorhynchus mykiss
Rainbow trout (28-day egg), FT,M
Oncorhynchus mykiss***
Rainbow trout (14-day egg), FT,M
Oncor us mykiss***
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
Cadmium
chloride
240
120
200
200
240
240
50
26
9.2
50
50
50
50
184
70
50
100
319
178
1700
6100
<0.5
30
10
9200
00
31.36
26.07
10.47
20.94
54.37
30.3
1700
12755
3.37
30
10
9200
75OO
Elnabarawy et al . ,
Hall et al., 1986
Hall et al., 1986
Hall et al., 1986
Elnabarawy et al..
Elnabarawy et al,
Martin and Holdich
Mirenda, 1986
Cusimano and Brakk
Vanleeuwen et al.,
Vanleeuwen et al.,
Van Leeuwen et al,
Van Leeuwen et al,
1986
1986
1986
, 1986
e, 1986
1985
1985
1985
1985
-------
Table 1. cont. Acute Data Generated for Cadmium Subsequent to the 1984 EPA Ambient Hater Quality Criteria
Document (EPA, 1985)
Species
Hardness Adjusted
(mg/1 as LC50/EC50 Acute Value
Method* Chemical CaCO3) (ug/1) (ug/1)**** Reference
Rainbow trout (24-hr, egg), FT,M
Oncorhynchus mykiss***
Rainbow trout (0-hr, egg), FT,M
Oncorhynchus mykiss***
Striped bass, S,U
Morone saxatilis
Striped bass, S,U
Morone saxatilis
Cadmium 50 13000
chloride
Cadmium 50 13000
chloride
Cadmium 40 4
Chloride
Cadmium 285 10
Chloride
13000 Van Leeuwen et al, 1985
13000 Van Leeuwen et al, 1985
5 14 Palawski, et al. 1985
1.4 Palawaki, et al. 1985
* FT = Flow-through, M = measured, S = static, U = unmeasured
*** These data were not used in development of the SMAV since data were
available for a more sensitive life stage
**** Adjusted to a hardness of 50 mg/1.
6,
-------
Table 2. Chronic Data Compiled for Cadmium Subsequent to the 1984 EPA Ambient Water Quality Criteria Document
(EPA, 1985)
Species
Hardness Adjusted
mg/1 as MATC MATC
Method* Chemical CaCO3 (ug/1) (ug/1)** Reference
Cladoceran, Ceriodaphnia reticulata PLC cadmium
chloride
Cladoceran, Daphnia magna***
Cladoceran, Daphnia pulex
H
N
oligocheate, Aeolosoma headleyi
PLC
L-C
LC
PLC
ELS
240
240
106
65
240
65
0.
4.
7.
7
13
25
4
3
07
49
.7
.19
0.
1.
1.
6.
4
20.
12
25
8
1
5
Elnabarawy et
Elnabarawy et
al.
al.
Ingersoll & Winner
Niederlehner, 1984
Elnabarawy et al.
Niederlehner,
1984
1986
1986
, 1982
1986
* PLC - partial life cycle, LC = life cycle, ELS = early life cycle
** Adjusted to 50 mg/1 hardness
*** Not used in SMAV calculation since a life cycle test data is available (EPA, 1985)
°\
-------
Table 3. Ranked Genus Mean Acute Values with SpecIBB Mean
Acute-Chronic Ratios for Cadmium.
Genus Mean
Acute Value
Rank* (ug/1)**
Species
Species Mean
Acute Value
-------
Tabla 3. Continued.
Genus Mean
Acute Value
Rank* (ug/1)**
Species
Species Mean
Acute Value
(ug/1)**
30
29
28
27
26
25
24
23
22
21
20
19
18
3018
2888
2400
2395
2310
2137
1700
1700
1200
736
401
322.8
221.9
Tubificid worm,
Branchiura sower by i
Flagfish,
Jordanella floridae
Northern squawfish,
Ptychocheilus oregonensis
Green sunfish,
Lepomia cyanellus
Pumpkinseed ,
Lepomis gibbosus
Bluegill,
Lepomis macrochirus
Mayfly,
Ephemerella grandis
Tubificid worn,
Luuiodrilus hoffmeisteri
Worn,
Nais sp.
Amphipod,
Crangonyx pseudogracilis
Mtdge,
Chironomus sp.
American eel,
Anguilla roatrata
leopod,
Asa llus bicrenata
Mayfly,
Paraleptophlebia praepedita
Bryozoan,
3018
2888
2400
2399
1347
4249
2310
2317
1700
1700
1200
736
401
322.8
221.9
Plunatella enarginata
17 215.5 Common carp,
Cyprinus carpio
16 204.9 Amphipod,
Hyalella azteca
215.5
204.9
-------
Table 3. Continued.
Genua Mean
Acute value
Rank* (ug/1)**
Species
SpecIBB Mean
Acute Value
(ug/1)**
15
14
13
12
11
10
9
8
7
6
5
156.9 Snail,
Phyaa gyrina
142.5 Bryozoan,
Pectinatella magnifica
104 Snail,
Aplexa hypnorum
98.79 Banded killifiah,
Fundului diaphanuB
62.66 Amphipod,
GaxnmaruB paeudolunnaeus
Amphipod ,
Gammarus sp.
60.3 cladoceran,
Ceriodaphnia reticulata
42.8 leaped,
Lirceua alabamae
40.78 Cladoceran,
Moina macrocopa
37.19 Cladoceran,
SimocephaluB oerrulatua
Cladoceran,
Simochephalua vetulua
30.54 Bryozoan,
Lophopodella carter!
30.50 Fathead minnow.
156.9
142.5
104
98.79
55.9
70.00
60.3
42.8
40.78
33.2
41.65
30.54
30.50
Pimephalea promelaa
19.3 Cladoceran,
Daphnia magna
Cladoceran,
Daphnia pulex
5.42 Coho Balmon,
Oncorhynchua kiautch
Chinook Balmon,
OncorhynchuB tahawytsha
14.2
26.3
6.48
4.25
-------
Table 3. Continued.
Rank*
Genus Mean
Acute Value
(ug/1)**
Specie*
Specie* Mean
Acute Value
(ug/1)**
Rainbow trout,
OncorhynchuB mykisa
2.7*** White perch,
Morone americana
Str Lped bass ,
Morone saxatilia
1.63 Brown trout,
Salrao trutta
5.78
7544
2.7
1.63
Ranked from most resistant to most sensitive baaed on Genue Mean Acute
Value.
Freshwater Genus Mean Acute Values and species Mean Acute Values are
at a hardness of 50 mg/1.
The striped bass SMAV was used &• the GMAV due to the wide variation in
in the SMAVs within the genus Morone
-------
Table 4-. Ranked Freshwater Genus and Specie* Mean Chronic Valuea
for Cadmium
Genus Mean Species Mean
Acute Value Acute Value
Rank* (ug/1) Species (ug/1)
14
13
12
11
10
9
8
7
6
5
4
3
2
20.5
16.32
15.22
8.17
8.138
7.771
7.849
5.336
4. 841
4.383
3.399
0.1918
0.1354
0.12
Oligochaete,
Aeoloeoma headleyi
Bluegill,
Lepomis macrochirus
Fathead minnow,
Punephales promelas
Smallmouth bass,
Micropterus dolomieui
Northern pike,
Esox lucius
Atlantic salmon,
Salnto salar
Brown trout,
S a lino trutta
White sucker,
Catostomus commerconi
Flagf ish,
Jordanella floridae
Snail,
Aplexa hypnorum
Brook trout,
Salvelinus fontinalis
Lake trout,
Salvelinus namaycush
Coho salmon,
Oncorhynchus kiautch
Chinook salmon,
Oncorhynchus tshawytscha
Cladoceran,
Moina macrocopa
Cladoceran,
Daphnia magna
Cladoceran,
Daphnia pulex
Cladoceran,
20.5
16.32
15.22
8.17
8.138
8.192
7.372
7.849
5.336
4.841
2.362
8.134
4.289
2.694
0.1918
0.1354
3.5
0.12
1 Cerodaphnia reticulata
* Ranked from most resistant to most sensitive based on Genus Mean
Acute Value.
Sit,
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Copper
Class 017
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for copper (EPA, 1985)
led to the addition of new acute and chronic toxicity data to
the EPA data base. These data are found in Tables 1 and 2,
respectively. The combined EPA and added data were used as the
basis for the Tier 1 criteria calculations. The toxicity of
copper is hardness dependent. Therefore, all data used in the
criteria calculations were normalized to a hardness of 50 mg/1.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (GMAV)(Table 3). This analysis resulted
in an FAV of 14.57 ug/1 (at hardness = 50 mg/1). No adjustment
of this value was necessary to protect for commercially or
recreationally important species of the Great Lakes basin.
Using the EPA derived pooled slope of 0.9422, the resulting
Final Acute Equation -
0.9422 (In hardness) - 1.007
e
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 7.3 ug/1 (at a hardness of 50 mg/1).
Criterion Continuous concentration (CCC)
Insufficient chronic toxicity data were available to calculate a
Final Chronic Value (FCV) using the eight family approach of the
procedure. Therefore, the FCV was calculated by dividing the FAV
by the EPA derived final acute to chronic ratio of 2.823,
resulting in a FCV of 5.16 ug/1 (at a hardness of 50 mg/1). No
adjustment of this value was necessary to protect commercially
or recreationally important species of the Great Lakes basin.
The CCC is 5.2 ug/1. Using the EPA chronic slope of 0.8545, the
Final Chronic Equation «
0.8545 (In hardness) - 1.702
e
Summary
The copper criteria are:
GLWQI
FAV - 14.57 ug/1 (at hardness = 50 mg/1)
-------
CMC =7.3 ug/1 (at hardness = 50 mg/1)
FCV = 5.16 ug/1 (at hardness - 50 mg/1)
CCC * 5.2 ugl (at hardness = 50 mg/1)
0.9422 (In hardness) - 1.007
Final Acute Equation = e
0.8545 (In hardness) - 1.702
Final Chronic Equation - e
EPA
FAV - 18.45 ug/1 (at hardness - 50 mg/1)
CMC =9.23 ug/1 (at hardness - 50 mg/1)
FCV = 6.539 ug/1 (at hardness - 50 mg/1)
0.9422 (In hardness) - 0.770
Final Acute Equation « e
0.8545 (In hardness) - 1.465
Final Chronic Equation = e
References
Bailey, H.C., D.H.W. Liu and H.A Javitz. 1985. Time/toxicity
relationships in short-term static, dynamic, and plug-flow
bioassays. Aquatic Toxicology and Hazard Assessment: Eighth
Symposium, ASTM STP 981.
Cusimano, R.F. and D.F. Brakke. 1986. Effects of pH on the
toxicities of cadmium, copper, and zinc to steelhead trout
(Salroo gairdneri). Can. J. Fish. Aquat. Sci., 43:1497-1503.
Elnabarawy, M.T., A.N. Welter and R.R.Robideau. 1986. Relative
sensitivity of three daphnid species to selected organic and
inorganic chemicals. Environ. Toxicol. and Chem. 5:393-398.
EPA, 1985 Ambient Aquatic Life Water Quality Criteria for
Copper. EPA 440/5-84-031. January 1985.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for
Deriving Aquatic Life Criteria. May, 1991
Harrison, F.L., J.P. Knezovich, and D.W. Rice Jr. 1984. The
toxicity of copper to adult and early life stages of the
freshwater clam, Corbicula manilensis. Arch. Environ. Centam.
Toxicol. 13:85-92.
Kosalwat, P. and A.W. Knight. 1987. Acute toxicity of aqueous
and substrate-bound copper to the midge, Chironomus decorus.
Arch. Environ. Contain. Toxicol. 16:275-282.
Martin T.R. and D.M Holdich, 1986. The acute lethal toxicity of
heavy metals to peracarids crustaceous (with particular
reference to fresh-water Asellids and Gammarids). Wat. Res.
20(9): 1137-1147.
Palawski, D., J.B. Hummand, and F.J. Dwyer. 1985 Sensitivity of
young striped bass to organic and inorganic contaminants in fresh
and saline waters. Trans. Am. Fish. Soc. 114:748-753.
Spehar, R.L. and J.T. Fiandt. 1986. Acute and chronic effects of
water quality criteria-based metal mixtures on three aquatic
species. Environ. Toxicol. Chem. 5:917-931.
-------
Table 1. Aquatic Acute Toxicity Data Added Subsequent to Publication of the Ambient Hater Quality
criteria for Copper (EPA, 1985).
Species
Hardness
(mg/1 as
Method* CaCO3) Chemical
Adjusted
LC/EC50 LC/EC50
(ug/1) (ug/1)** Reference
Cladoceran,
Ceriodaphnia reticulata
Cladoceran,
Daphnia raagna
Cladoceran,
Daphnia pulex
Amphipod,
Crangonyx pseudogracilus
Asiatic clam,
Corbicula manilensia
Midge,
Chironomus decorus
Fathead minnow,
Pimephalea promelas
Bluegill,
Lepomia macrochirus
Bluegill,
Lepomis macrochirus
Rainbow trout,
Oncorhynchus mykiss
Striped bass,
Morone saxatilia
S,U
s,u
S,U
S,U
FT,M
S,M
FT,M
S,M
FT,M
FT,M
S,U
240
240
240
SO
17
44
43.9
31.2
31.2
9.2
285
Cupric
chloride
CupL ic
chloride
Cupric
chloride
-
-
-
Cupric
nitrate
Copper
sulfate
Copper
sulfate
Copper
sulfate
23
41
31
1290
>2600
739
96
340
550
2.8
270
5.2
9.4
7.1
1290
>7184
834
109
*** 530
858
14
*** 52
Elnabarawy
Elrabara"y
Elnabarawy
Martin and
et al., 1986
et al - 1986
et al., 1986
Holdich, 1986
Harrison et al., 1984
Kosalwat &
Knight, 1987
Spehar & Fiandt, 1986
Bailey et
Bailey et
Cueimano &
al., 1985
al., 1985
Brakke, 1986
Palawski et al, 1985
* s » static, FT - flow through, 0 - unmeasured, M - measured
** Adjusted to a hardness of 50 mg/1
•** Data not used due to preferential use of FT, tudies
*** Value not used because acute values are too rgent for this species
-------
Table 2. Aquatic Chronic Toxicity Data Added Subsequent to Publication of the Ambient Hater Quality
Criteria for Copper (EPA, 1985).
Hardness Adjusted
(tng/1 as HATC MATC
Species Method* CaCO3) Chemical (ug/1) (ug/1)** Reference
Fathead minnow, ELS 44 Cupric 6.2 6.7 Spehar & Fiandt, 1986
Pimephales promelas nitrate
* ELS - early life stage
** Adjusted to a hardness of 50 mg/1
-------
Table 3. Ranked Genus Meun Acute Values with Specj.ee Mean Acute-Chronic
Ratio* for Copper.
Rank*
43
42
41
40
39
38
3?
36
35
34
33
Genus Mean
Acute Value
(ug/1)**
10240
7184
6200
5860
4600
4305
1990
1877
1397
1290
1057
Species
Stonefly,
Acroneuria lycorias
Astatic clam,
Corbicula manilensis
Caddis fly,
Unidentified sp.
White perch.
Mo rone amencanus
Dainaelfly,
Unidentified sp.
American eel,
Anguilla rest rat a
Crayfish,
Procambarus clarku
Snail,
Campeloma decisum
Crayfish,
Orconectes rusticus
Aftphipod,
Crangonyx pseudogracilus
Pumkinaeed,
Species Mean Species Mean
Acute Value Acute-Chronic
(ug/1)** Ratio
10240
> 7184
6200
5860
4600
4305
1990
1877 156.2***
1397
1290
640.9
L«'pomia gibbosus
32
31
30
29
28
900
790.6
684.3
331.8
289
Bluegill,
Letpomia macrochirue
Snail,
Amnicola sp.
Banded killifish,
Fundulus diaphanus
Mozambique tilapia
T.tlapia mossambica
Striped shiner,
Notropis chrysocephalus
Goldfish,
1742
900
790.6
684.3
331.8
289
Carassius auratus
-------
Table 3. Continued.
Rank*
27
26
25
24
23
22
21
20
19
18
17
16
Genus Mean
Acute Value
(ug/l)«*
242.7
196.1
193.36
166.2
156.8
141.2
135
133
110.4
97.9
90
86.67
Species
Worn,
Lumbriculus variegatus
Moaquitofiah,
Gambusia af finis
Midge ,
Chironomua tentans
Midge,
Chironomus decoruB
Midge,
Chironomus sp.
Snarl,
Goniobaais livescens
Common carp,
CyprinuB carpio
Rainbow darter
Etheoatoma caeruleum
Orangethroat darter,
Etheoetoma spectabile
Bryozoan,
Pectinatella magnifica
Chroelmouth,
Acrocheilus alutaceua
Brook trout.
SalvelinuB fontinalia
Bluntnose minnow,
Pimephalee notatue
Fathead minnow.
Pimephalea promelas
Worm,
Nais ap.
Blacknoee dace,
SpecieB Mean
Acute Value
(ug/1)**
242.7
196.1
298
834
30
166.2
156.8
86.67
230.2
135
133
110.4
72.16
132.9
90
86.67
Species Mean
Acute-Chronic
Ratio
—
-
-
-
-
-
-
-
-
-
7.776***
26.36***
10.33***
-
-
Rhinichthys atratulus
-------
Table 3. Continued.
Rank*
15
14
13
12
11
10
9
8
7
6
5
4
Genus Mean
Acute Value
(ug/1)**
85.42
83.97
83
78.55
76.07
69.81
56.21
53.08
39.33
37.05
37.05
22.09
Specie*
Cutthroat trout,
Salmo clarkii
Atlantic salmon,
Salmo aalar
Creek chub,
Semotilus atromaculatus
Guppy,
Poecilia reticulata
Central • toner oiler,
Campostoma anomalum
Coho salmon,
Oncorhynchue kisutch
Sockeye salmon,
Oncorhynchua nerka
Chinook Balmon,
Oncorhynchue tshawytacha
Rainbow trout,
Oncorhynchua mykiBS
Brawn bullhead,
Ictalurus nebulosus
Snail,
Gyraulus circumstriatus
Horn,
Limnodrilus hoffmeiateri
Snail,
Physa heteroBtropha
Snail,
Physa integra
Bxyozoan,
Lophopodella carter!
Bzyozoan,
P lunate la emarginata
Aoiphipod,
Species Mean
Acute Value
(ug/1)**
66.26
110.12
83.97
83
78.55
87.1
233.8
42.26
38.89
69.81
56.21
53.08
35.91
43.07
37.05
37.05
22.09
Species Mean
Acute-Chronic
Ratio
-
-
-
-
-
-
-
> 4.473***
—
—
-
-
-
3.585***
—
—
3.297
Giunoarus pseudolimnaeuB
-------
Table 3. Continued.
Rank*
3
Genus Mean
Acute Value
(ug/1)** Species
Species Mean Species Mean
Acute Value Acute-Chronic
(ug/1)** Ratio
16.74 Northern Squawfish,
Ptyehocheilus oregonensit
14.48 Cladoceran,
Daphnia magna
Cladoceran,
Dapnia pulex
Cladoceran,
Daphnia pulicaria
9.92 Cladoceran,
Cenodaphnia reticulata
16.74
19.88
16.5
9.263
9.92
2.418
* Hanked from moat resistant to most sensitive based on Genus Mean Acute Valui
** Genus Mean Acute Values are at hardness = 50 mg/1.
*** Value not used in calculating the final acute-chronic ratio
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier l Aquatic Life Criteria for Free Cyanide
Class 018
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier l approach of the Great Lakes Water Quality
Inititative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for cyanide (EPA, 1985}
produced no new data to supplement the EPA data base. Therefore,
the EPA data base was used as the basis for the Tier 1 criteria
calculations. A modification of the EPA data base was made
during this review due to the recent reclassification of the
rainbow trout as Oncorhynchus mykiss. This reclassification
resulted in no change to the 1985 EPA calculated acute criterion
for free cyanide.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (GMAV)(Table 1). This analysis resulted
in an FAV of 45.77 uq/1. The CMC was calculated by dividing
the FAV by 2, resulting in a CMC of 23 ug/1.
For waterbodies protected for salmonid fisheries, the FAV was
lowered to the rainbow trout SMAV of 44.73 ug/1 in order to
provide adequate protection for this commercially and
recreationally important species of the Great Lakes basin. The
CMC is 22 ug/1.
Criterion Continuous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate a
Final Chronic Value (FCV) using the eight family approach of the
procedure. Therefore, the FCV was calculated by dividing the FAV
(45.77 ug/1) by the EPA derived final acute to chronic ratio of
8.568 (Table 1), resulting in a FCV of 5.342 ug/1. The CCC is
5.3 ug/1. For waterbodies protected for salmonid fisheries, the
FCV is 5.22 ug/1 and CCC is 5.2 ug/1. No adjustment of either
FCV was necessary to protect commercially or recreationally
important species of the Great Lakes basin.
Summary
The Free cyanide criteria are:
GLWQI EPA
Non-salmonid Salmonid
FAV: 45.77 ug/1 44.73 ug/1 44.73 ug/1
CMC: 23 ug/1 22 ug/1 22.36 ug/1
FCV: 5.342 ug/1 5.221 ug/1 5.221 ug/1
CCC: 5.3 ug/1 5.2 ug/1
-------
References
EPA, 1985. Ambient Water Quality Criteria for Cyanide - 1984.
EPA 440/5-84-028. January 1985.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for
Deriving Aquatic Life Criteria. April, 1991
-------
Table 1. Ranked Genus Mean Acute Values with Species Mean Acute-Chronic Ratios
for Free Cyanide
Rank*
16
15
14
13
12
11
10
9
8
7
6
S
4
3
2
1
Genus Mean
Acute value
(ug/1)
2490
2326
432
426
318
167
147
125.1
123.6
102
102
99.28
92.64
90
85.80
44.73
Species
Midge,
Tanytarsus dissimilis
Isopod,
Asellus communis
Snail ,
Physa heterostropha
Stone fly,
Pteronarcys dorsata
Goldfish,
Carassius Auratus
Amphipod,
Gammarua pseudolunnaeus
Guppy,
Poecilia reticulata
Fathead minnow,
Pimephales promelas
Cladoceran,
Daphnia magna
Cladcceran,
Daphnia pulex
Largemouth bass,
Micropterus salmoides
Black crappie,
Porno* is nigromaculatus
Blueciill,
Lepomis macrochirus
Yellow perch,
Ferca flavescens
Atlantic salmon,
Salmo salar
Brook trout,
Salve linus fontinalis
Rainbow trout
Species Mean
Acute Value
(ug/D
2490
2326
432
426
318
167
147
125.1
160
95.55
102
102
99.28
92.64
90
85.8
44.73
Species Mean
Acute-Chr on ic
Ratio
-
68.29**
-
-
-
9.111
-
7.633
-
—
—
—
7.316
-
-
10.59
—
Oncorhynchus mykiss
* Ranked from most resistant to most sensitive based on Genus Mean Acute Value.
** Value not used in calculation of the Final Acute-Chronic Ratio
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Chromium(III)
Class # 015
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for Chromium(III)
(EPA, 1984) produced additional acute toxicity data (Table 1).
The combined data were used as the basis for the Tier 1
criteria calculations. The toxicity of chromium(III) is
hardness dependent. Therefore, all data used in the criteria
calculations were normalized to a hardness of 50 mg/1.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (Table 1). This analysis resulted in
an FAV of 2044 ug/1 at a hardness of 50 mg/1. No adjustment of
this value was necessary to protect for commercially or
recreationally important species of the Great Lakes basin.
Using the EPA derived pooled slope of 0.819, the Final Acute
Equation =
0.819 (In hardness) + 4.419
e
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 1022 ug/1.
Criterion Continuous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate
a Final Chronic Value (FCV) using the eight family approach of
the procedure. Therefore, the FCV was calculated by dividing
the FAV by the final acute to chronic ratio of 41.8 (geometric
mean of the fathead minnow and rainbow trout ACRs) (Table 2).
The resulting FCV is 48.86 ug/1. The CCC is 49 ug/1. The Final
Chronic Equation =
0.819 (In hardness) + 0.685
e
No adjustment of the FCV was necessary to protect commercially
or recreationally important species of the Great Lakes basin.
Summary
The Chromium(III) criterion are:
GLWQI
FAV = 2044 ug/1 (at hardness = 50 mg/1)
CMC - 1022 ug/1 (at hardness = 50 mg/1)
FCV = 48.86 ug/1 (at hardness = 50 mg/1) .
CCC = 49 ug/1 (at hardness = 50 mg/1) /C7
-------
0.819 (In hardness) + 4.419
Final Acute Equation » e
0.819 (In hardness) + 0.685
Final Chronic Equation = e
EPA
FAV = 1968 ug/1 (at hardness =50 mg/1)
CMC = 984 ug/1 (at hardness - 50 mg/1)
FCV = 117 ug/1 (at hardness • 50 mg/1)
CCC = 120 ug/1 (at hardness - 50 mg/1)
0.819 (In hardness) +4.38
Final Acute Equation - e
0.819 (In hardness) + 1.561
Final Chronic Equation = e
References
EPA, 1985. Ambient Aquatic Life Water Quality Criteria for
Chromium(III) - 1984. EPA 440/5-84-029. January 1985.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure
for Deriving Aquatic Life Criteria. May, 1991
Martin, T.R. and D.M. Holdich. 1986. The acute lethal toxicity
of heavy metals to peracarid crustaceans (with particular
reference to fresh-water asellids and gammarids). Wat. Res.
20(9):1137-1147.
-------
Table 1. Acute Data Generated for Chromium Subsequent to the 1984 EPA Ambient Water Quality Criteria
Document (EPA, 1985)
Species
Hardness Adjusted
(mg/1 as LC50/EC50 Acute Value
Method* Chemical CaCO3) (ug/1) (ug/1)** Reference
Amphipod,
Crangonyx paeudogracilis
S,U
Chromium
chloride
50
291
291
Martin & Holdich, 1986
* S - static, U = unmeasured
** Adjusted to a hardness of 50 mg/1
-------
-------
Table 2. Ranked Genus Mean Acute Values with Speciea Mean Acute-Chronic Ratios
for Chromium(III).
Genus Mean
Acute Value
Rank* (ug/l)«»
Species
Species Mean Species Mean
Acute Value Acute-Chronic
(ug/1)** Ratio
19
18
17
16
15
14
13
12
11
10
9
S
7
6
5
291,000
71060
50000
43100
16010
15630
15370
14770
13230
12860
11000
10320
10210
9669
9300
Amphipod ,
Crangonyx pseudogracilis
Caddis fly,
Hydropsyche betteni
Caddisfly,
Unidentified sp.
Damself ly,
Unidentified sp.
cladoceran,
Daphnia magna
Banded killifiah,
Fundulus diaphanus
Pumpkin seed,
Lepomis gibboaua
Bluegill,
Lepomis macro chi.ru e
White perch,
Morone americana
Striped bass,
Morone aaxatilie
Common carp,
Cyprinus carpio
American eel,
Anguilla rostrata
Midge,
Chironomua sp.
Fathead minnow,
Pimephales promelas
Snail,
Amnicola sp.
Rainbow trout,
Oncorhynchus mykisa
Worm,
Nais sp.
291,000
71060
50000
43100
16010 >356.4**
15630
15720
15020
13320
16370
13230
12860
11000
10320 27.30
10210
9669 64.11
9300
7/
-------
Table 2. Continued.
Rank*
4
3
2
1
Genus Mean
Acute Value
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier l Aquatic Life Criteria for Chromium (VI)
CAS # 18540-29-9
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Inititative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for chromium (EPA, 1985)
led to the addition of new acute toxicity data to the EPA data
base (Table l). The combined data bases were used as the basis
for the Tier 1 criteria calculations.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (GMAV)fTable 2). This analysis resulted
in an FAV of 32.04 ug/1. No adjustment of this value was
necessary to protect for commercially or recreationally
important species of the Great Lakes basin.
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 16 ug/1.
Criterion Continuous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate a
Final Chronic Value (FCV) using the eight family approach of the
Erocedure. Therefore, the FCV was calculated by dividing the FAV
y the EPA derived final acute to chronic ratio of 2.917 (Table
1), resulting in a FCV of 10.98 ug/1, and a CCC, of 11 ug/1. No
adjustment of this value was necessary to protect commercially
or recreationally important species for the Great Lakes basin.
Summary
The Hexavalent chromium criteria are:
GLWQI EPA
FAV = 32.04 ug/1 FAV = 31.49 ug/1
CMC = 16 ug/1 CMC = 15.74 ug/1
FCV = 10.98 ug/1 FCV = 10.80 ug/1
CCC = 11 ug/1
References
EPA, 1985. Ambient Water Quality Criteria for Chromium - 1984.
EPA 440/5-84-029. January 1985.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for
Deriving Aquatic Life Criteria. May, 1991.
Berglind, R. and G. Dave, 1984. Acute toxicity of chromate,
DDT, PCP, TPBS, and zinc to Daphnia magna cultured in hard and
soft water. Bull. Environ. Contam. Toxicol. 33:63-68
73
-------
Dorn, P.B., J.H. Rodgers, Jr., K.M. Jop, J.C. Raia, and K.L.
Dickson. 1987. Hexavalent chromium as a reference toxicant in
effluent toxicity test. Environ. Toxicol. and Chem. 6:435-444.
Elnabarawy, M.T., A.N. Welter and R.R.Robideau, 1986. Relative
sensitivity of three daphnia species to selected organic and
inorganic chemicals. Environ. Toxicol. and Chem. 5:393-398.
Jop, K.M, T.F. Parkerton, J.H. Rodgers and K.L. Dickson.
1987. Comparative toxicity and speciation of two hexavalent
chromium salts in acute toxicity tests. Environ. Toxicol. Chem,
6:697-703
Martin, J.R. and D.M. Holdich, 1986. The acute lethal toxicity
of heavy metals to peracarid crustaceans (with particular
reference to fresh-water asellids and gammarids). Wat.Res.
20(9):1137-1147.
-------
able 1. Aquatic Acute Toxicity Data Added Subsequent to publication of
the Ambient Water Quality Criteria for Hexavalent Chromium (EPA, 1985)
Specie*
Method* Chemical
LC/ECSO
(ug/1) Reference
Cladoceran,
Daphnia magna
Cladoceran,
Daphnia magna
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulex
Amphipod ,
Carngonyx pseudogracilis
Amphipod ,
Carngonyx peeudogracilia
Bluegill,
Leporine macrochirus
Bluegill,
Lepomis macrochirus
Bluegill,
S,U
S,U
S,H
s,u
S,M
s,u
S,M
s,u
s,u
S,M
S,M
R,U
R,U
S,M
S,M
S,M
K-dichr ornate
Na-dichromate
K-dichromate
K-dichromate
K-dichromate
K-dichromate
K-d ic hr ornate
K-dichr ornate
Na -d i c hr omat e
K-dichromate
K-dichromate
K-dichromate
K-dichromate
K-dichromate
K-dichromate
K-dichromate
900**
112**
170**
190**
20**
20**
40**
40**
122**
ISO**
180**
420
810
182,000**
154,000**
201,240**
Berglind & Dave, 1984
Elnabaraway et al.
Dorn, et al., 1987
Dorn, et al., 1987
Dorn, et al., 1987
Dorn, et al., 1987
Dorn, et al., 1987
Dorn, et al., 1987
Elnabaraway et al.
Jop et al, 1987
Jop et al, 1987
Martin & Holdich,
Martin & Holdich,
Jop et al, 1987
Jop et al, 1987
Dorn et al. , 1987
, 1986
, 1986
1986
1986
Lepomis macrochirus
-------
Table 1. cont.
Species
Method Chemical
LC/EC50
(ug/1) Reference
Bluegill,
Lepomis macrochiruB
Bluegill,
Lepomis macrochiruB
Bluegill,
Lepomis macrochiruB
Bluegill,
LepomiB macrochirus
Bluegill,
LepomiB macrochiruB
Fathead minnow,
Pimephales promelas
Fathead minnow,
Punephales promelas
Fathead minnow,
Pimephales promelas
Fathead minnow,
S,U
S,M
S,U
S,M
s,,u
S,M
£,M
S,U
S,M
K-dichr ornate
K-dichr ornate
K-dichromate
K-dichr ornate
K-dichromate
K-dichromate
K-dichromate
K-dichr ornate
K-dichr ornate
164,730** Dorn et al.
199,200** Dorn et al.
158,360** Dorn et al.
148,310** Dorn et al.
146,530** Dorn et al.
46,000** Jop et al.
34,000** Jop et al,
26,130** Dorn et al.
26,410** Dorn et al.
, 1987
, 1987
, 1987
, 1987
, 1987
1987
1987
, 1987
, 1987
Pimephales promelaa
* S « static, FT = flow through, M * measured, U «
** These data were not used in criteria development
of FT,M tests for the Cladocerans and fish.
unmeasured
due to the availability
-------
Table 2 Ranked Freshwater Genus Mean Acute Values with Species Mean Acute-Chronic
Ratios for Hexavalent Chromium.
Genus Mean
Acute Value
Species Mean Species Mean
Acute Value Acute-Chronic
Rank*
28
27
26
25
24
23
22
21
20
19
18
17
16
15
(ug/i)
1,870,000
176,000
140,000
123,500
119,500
72,600
69,000
67,610
61,000
59,000
57,300
51,250
49,600
47,180
Species
Stonefly,
Neophasganophora capitata
Crayfish,
Orconectes rusticus
Damsel fly,
Enallagma aspersum
Green sunfish,
Lepomis cyanellus
Bluegill
Lepomis macrochirus
Goldfish
Car as BLUB auratus
White crappie,
Pomoxia annular is
Rainbow trout.
Oncorhynchue mykiss
Emerald shiner.
Notropis atherinoides
Striped shiner.
Notropis chrysocephalus
Sand shiner,
Notropis stramineus
Midge,
Chironomus tentans
Brook trout,
Salvelinus fontinalis
Midge ,
Tanytarsus dissxmilis
Central stoneroller.
Campos toma anomalum
Silver •) aw minnow,
Ericymba buccata
Bluntnose minnow,
Pimephales notatus
Fathead minnow,
-------
Table 2. cont.
»
Genus Mean
Acute Value
Rank* (ug/l)
Species
Species Mean Species Mean
Acute Value Acute-Chronic
(ug/l) Ratio
14
13
12
11
10
9
8
7
6
5
4
3
2
46,000
36,300
30,450
30,000
23,010
1,560
1,440
650
630
583
67.1
45.1
36.35
Johnny darter,
Etheoiitoma nigrum
Yellow perch,
Perca flavescena
Stripitd bass,
Moronn saxatilis
Guppy,
PoeciJ la reticulata
snail,
Physa heterostropha
Bryozoan,
Lophopodella carteri
Bryozoan,
Pectinatella magnifica
Bryozoan,
Plumatella eroarginata
Amphipod,
Hyale] la azteca
Amphipod ,
Crangonyx paeudogracilis
Amphipod,
Ganunarus pseudolimnaeus
Cladoceran,
Ceriodaphnia reticulata
Cladoceran,
Simocophalua serrulatus
Cladoceran,
46,000
36,300
30,450
30,000
23,010
1,560
1,440
650
630
583
67.1
45.1
40.9
32.3
-
-
-
-
-
—
—
—
-
—
1.13
2.055
5.267
Simocephalus vetulus
28.94 Cladoceran,
Daphnia cnagna
Cladoceran,
Daphnta pulex
23.07 > 6.957**
36.3 5.92
* Ranked from most resistant to most sensitive based on Genus Mean Acute Value.
** Value not used in calculation of the Final Acute-Chronic Ratio
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Dieldrin
CAS #60-57-1
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991). A search of aquatic toxicity literature generated
subsequent to publication of the EPA criteria document for
dieldrin (EPA, 1980) led to the addition of new acute toxicity
data to the EPA data base (Table 1).
Some acute toxicity data from the EPA criteria were not used as
the test protocol did not meet current acceptable toxicity
testing procedures. These included tests conducted by Santharam
et al. (1976), Gaufin (1965) and Jensen and Gaufin (1964).
The combined EPA and added data were used as the basis for the
Tier l criteria.
Criterion Maximum Concentration (CMC)
Sufficient data were available to calculate a Tier l acute
criterion (Final Acute Value or FAV) for dieldrin. The FAV was
calculated using the lowest 4 Genus Mean Acute Values (GMAV)
(Table 2). This analysis resulted in a FAV of 0.4781 ug/1. No
adjustment of this value was necessary to protect commercially
or recreationally important species of the Great Lakes basin.
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 0.24 ug/1.
Criterion Continous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate a
Final Chronic Value (FCV) using the eight family approach of the
procedure. Therefore, the FCV was calculated by dividing the FAV
by the EPA derived final acute to chronic ratio (FACR) of 8.53.
The resulting FCV is 0.0561 ug/1. The CCC is 0.056 ug/1. No
adjustment of this value was necessary to protect for
commercially or recreationally important species of the Great
Lakes basin.
Summary
The dieldrin criteria are:
GLWQI EPA
FAV - 0.4781 ug/1 FAV - 2.5 ug/1
CMC =0.24 ug/1
FCV - 0.0561 ug/1 FCV = 0.29 ug/1
CCC = 0.056 ug/1
-------
Table 1. Aquatic Acute ToxicJ.ty Data Added Subsequent to Publication of
the Ambient Hater Quality Criteria for Dxeldrin (EPA, 1980)
Species
Teat
Duration LC/EC50
Method* (hra) (ug/1)
Reference
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulex
Stonefly,
Claaaaenia aabuloaa
Stonefly,
Pteronarcya californica
Stonefly,
Pteronarcella badia
Damaelf ly,
lechnura verticalia
Annelid,
Lumbriculua variegatus
Rainbow trout,
Oncorhynchua mykiea
Rainbow trout,
Oncorhynchua mykia*
Rainbow trout,
Oncorhynchua mykiaa
Goldfish,
Caraaaiua auratue
Fathead minnow,
Pinephalea promelae
Bluegill,
Lepomia macrochirus
Bluegill ,
Lepomia macrochirua
Pumpkinaeed,
Lepomia gibboaua
Cutthroat trout,
Salmo clarki
S,M
S,U
s,u
s,u
s,u
s,u
FI,M
S,U
FT,M
S,U
S,U
S,U
s,u
s,u
£.,U
S,U
48
48
96
96
96
96
96
96
96
96
96
96
96
96
96
96
251
190
0.6
0.5
0.5
12
21.8
1.2**
0.62
3**
1.8
3.8
3.1
7
6.7
6
Daniels & Allan, 1981
Mayer 6 Ellersieck, 1986
Mayer & Elleraieck, 1986
Mayer fi Elleraieck, 1986
Mayer fi Elleraieck, 1986
Mayer & Elleraieck, 1986
EPA, 1991
Mayer fi Elleraieck, 1986
Shubat & Curtis, 1986
Vanleeuwen et al., 1985
Mayer 6 Elleraieck, 1986
Mayer & Elleraieck, 1986
Mayer fi Elleraieck, 1986
Sanders, 1972
Cairna fi Scheier, 1964
Mayer fi Elleraieck, 1986
-------
Table 1. continued
Species
Test
Duration LC/EC50
Method* (hrs) (ug/1)
Reference
Channel catfish,
Xctalurus punctatus S,U
Largemouth bass,
Micropterus salmoides S,U
96
96
4.5 Mayer & Ellersieck, 1986
3.5 Mayer & Ellereieck, 1986
* S = static, FT * flow through, U = unmeasured, M » measured
** Value was not used in criteria development due to the preferential use
of a FT,M test.
-------
Table 2. Ranked Genus Mean Acute Values for Dieldrin
Rank*
19
18
17
16
15
14
13
12
11
10
9
8
7
GMAV
ug/1
740
534
228
214
21.8
20
17.7
12
8.6
8.5
6
5
4.5
Species
Crayfish,
Orconectes naxs
Amphipoda ,
Gamiiarus lacustris
Amphipoda ,
Gamnarus fasciatus
Cladoceran,
Daphnia pulex
Cladoceran,
Simocephalus serrulatus
Annelid,
Lumbriculus variegatus
Glass shrimp
Palaemonetes kadiakensis
Fathead minnow,
Pimephales promelas
Damsel fly,
Ischnura verticalzs
Goldfish,
carassius auratus
Pumpkinseed ,
Lepomis gibbosus
Bluegill,
Lepomis macrochirus
Green sunfish,
Lepomis cyanellus
Cutthroat trout,
Sal no clarki
Isopoda,
Asellus brevicaudus
Channel catfish,
Ictalurus punctatus
SMAV
ug/1
740
460
620
228
214
21.8
20
17.7
12
8.6
6.7
11.5
8
6
5
4.5
-------
Table 2. cont. Ranked Genus Mean Acute Values for Dieldrin
Rank*
6
5
4
3
2
1
GMAV
ug/1
4.5
3.5
0.62
0.6
0.5
0.5
Species
Guppy,
Poecilia reticulata
Largemouth bass,
Micropterus salmoides
Chinook salmon,
Oncorhynchus tshawytscha
Coho salmon,
Oncorhynchus kisutch
Rainbow trout,
Oncorhynchus mykiss
Stonefly,
Claassenia sabulosa
Stonefly,
Pteronarcys californica
Stonefly,
Pteronarcella badia
SMAV
ug/1
4.5
3.5
6.1
10.8
0.62
0.6
0.5
0.5
* Ranked from most resistant to most sensitive based on GMAV.
** The rainbow trout SMAV was used as the GMAV because it was
the result of a FT,M test, whereas the coho and chinook data
were from S,U tests.
-------
Table 3. Acute to Chronic Ratios for Dieldrin
Species
Guppy,
Poecilia reticulata
Rainbow trout,
Oncorhynchus mykiss
Mysid shrimp,
Hysidopsis bahia
Acute
value
(ug/1)
4.1
2.5
4.5
Chronic
Value
(ug/1)
0.45
0.22
0.73
Ratio
9.1
11
5.6
Refei
EPA,
EPA,
EPA,
rence
1980
1980
1980
-------
REFERENCES
Cairns, J.Jr.& A. Scheier. 1964. The effect on the sunfish, Lepomis gibbosus,
of chronic exposure to lethal and sublethal concentrations of dieldrin
Notulae Naturae. 310:10
Daniels, R.E. and J.D. Allan. 1981. Life table evaluation of chronic
exposure to a pesticide. Can. J. Fish. Agu. Sci. Vol 38(5):485-494
EPA, 1980. Ambient Water Quality Criteria for Aldrin/Dieldrin. EPA
440/5-80-019
EPA, 1991. Acute 96-hr flow-through freshwater exposures with endrin and
dieldrin using an annelid (Lumbriculus variegatus). Unpublished results.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for Developing
Aquatic Life Criteria. April, 1991
Mayer, F.L. and M.R. Ellersieck. 1986. Manual of Acute Toxicity:
Interpretation and Data Base for 410 Chemicals and 66 Species of Fresh-
water Animals. USDI Publication # 160.
Sanders, H.O., 1972. Toxicity of some insecticides to four species of
malocostracan crustaceans. U.S. Bureau Sports Fishery and Wildlife Tech.
Papers. Tech. Paper 66
Shubat, P.J. and L.R. Curtis. 1986. Ration and toxicant preexposure
influence on dieldrin accumulation by rainbow trout (Salmo gairdneri).
Environ. Toxicol. and Chem. Bol. 5, pp. 69-77
VanLeeuwen, C.J. et al. Differences in susceptibility of early life stages
of rainbow trout (Salmo gairdneri) to environmental pollutants. Aquatic
ToxiCOl. Vol. 7, pp 59-78
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Endrin
CAS #72-20-8
July, 1991
Introduction
The criteria presented in this document were developed pusuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991). A search of aquatic toxicity literature generated
subsequent to publication of the EPA criteria document for
endrin (EPA, 1980} led to the addition of new acute toxicity
data to the EPA data base (Table 1).
Some acute toxicity data from the EPA criteria document were not
used in development of the GLI criteria because the test
protocol did not meet current acceptable toxicity testing
procedures. These included tests conducted by Naqui and Ferguson
(1968), Nebeker and Gaufin (1964), Gaufin et al (1965), Jensen
and Gaufin (1966), Post and Schroeder (1971), Mount (1962), and
Solon (1969).
The combined EPA and added data (Table 1) were used as the basis
for the Tier 1 criterion.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
genus mean acute values (Table 2). This analysis resulted in a
FAV of 0.1792 ug/1. No adjustment of this value was necessary to
protect commercially or recreationally important species of the
Great Lakes basin.
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 0.09 ug/1.
Criterion Continuous Concentration (CCC)
The EPA criteria document for endrin lists four acute to chronic
ratios. Three of these (Table 3) were considered appropriate for
developing a final acute to chronic ratio (FACR). The resulting
FACR equals 4.8.
The Final Chronic Value (FCV) is equal to the FAV divided by the
FACR. The resulting FCV is 0.0373 ug/1. The CCC is 0.037 ug/1.
No adjustment of this value was necessary to protect for
commercially or recreationally important species of the Great
Lakes Basin.
Summary
The endrin criterion are:
GLWQI EPA
FAV m 0.1792 ug/1 FAV - 0.18 ug/1
CMC - 0.09 ug/1
FCV - 0.0373 ug/1 FCV - 0.045 ug/1
CCC = 0.037 ug/1
-------
able. 1 Aquatic Acute Toxicity Data Added Subsequent to Publication
of the Ambient Water Quality Criteria for Endrin (EPA, 1980)
Species
Cladoceran ,
Cerodaphnia reticulata
Cladoceran,
Daphnia magna
Cladoceran ,
Daphnia magna
Cladoceran,
Daphnia magna
Cladoceran,
Daphnia magna
Cladoceran,
Daphnia magna
Cladoceran,
aphnia pulex
Cladoceran,
Daphnia pulex
Annelid,
Lumbnculus vanegatus
Snipe fly,
Atherix variegatua
Midge,
Tanytarsus dissimilis
Stonefly,
Acroneuria pacifica.
Crayfish,
Orconectes immunis
Damselfly,
lachnura verticalus
Danselfly,
lachnura verticalus
Yellow perch,
Perca flavescens
Method*
S,U
S,U
S,0
s,u
s,u
S,M
S,U
s,u
FT,M
S,U
S,M
s,u
FT,M
S,U
S,U
FT,U
Test
Duration
(hre)
48
48
48
48
48
48
48
48
96
96
48
96
96
96
96
96
LC/ECSO
(mg/1)
24
4.2
59
41
74
160
20
30
42.6
4.6
0.84
> 0.18**
89
2.4
2.1
0.15
Reference
Elnabarawy, et al 1986
Mayer and Ellersieck,
Elnabarawy, et al 1986
Mayer and Elleraieck,
Mayer and Ellersieck,
Thurston et al., 1985
Mayer and Elleraieck,
Elnabarawy et al, 1986
EPA, 1991
Mayer and Elleraieck,
Thuraton et al., 1985
Mayer and Ellersieck,
Thurston et al., 1985
Mayer and Ellersieck,
Mayer and Ellersieck,
Mayer and Ellersieck,
1985
1985
1985
1985
1985
1985
1986
1986
1986
-------
Table. 1 cont. Aquatic Acute Toxicity Data Added Subsequent to Publication
of the Ambient Water Quality Criteria for Endrin (EPA, 1980)
Specie*
Test
Duration LC/EC50
Method* (hrs) (mg/1) Reference
Largemouth baas,
Micropterus salmoides
Black bullhead,
Zctalurus melas
Channel catfish.
Ictalurus puntatus
Channel catfish,
Ictalurus puntatus
Channel catfish,
Ictalurus puntatus
Rainbow trout.
Oncohynchus my kiss
Rainbow trout,
Oncohynchus mykiss
Goldfish,
Carassius auratus
Goldfish,
Carassius auratus
Fathead minnow,
Pimephales promelas
Fathead minnow,
Pimephales promelas
Mosquitofish,
Gambusia affinis
Mosquitofish,
Gambusia affinis
Carp,
Cyprinus carpio
Bluegill,
Lepomis macrochirus
Bullfrog tadpole,
Rana cateabiana
S,U
s,u
s,u
s,u
FT,M
s,u
FT,M
FT,U
FT,M
S,U
FT,M
S,U
FT,M
FT,0
FT,M
FT,«
96
96
96
96
96
96
96
96
96
96
96
96
96
96
96
96
0.31
1.1
0.32**
1.1**
0.42
0.75**
0.3
0.44**
0.95
1.8**
0.65
1.1**
0.69
0.32
0.21
2.5
Mayer and Ellersieck,
Mayer and Ellersieck,
Mayer and Ellersieck,
Mayer and Ellersieck,
Thurston et al. 1985
Mayer and Ellersieck,
Thurston et al. 1985
Mayer and Ellersieck,
Thurston et al. 1985
Mayer and Ellersieck,
Thurston et al. 198S
Mayer and Ellersieck,
Thurston et al. 1985
Mayer and Ellersieck,
Thurston et al. 1985
Thurston et al. 1985
1986
1986
1986
1986
1986
1986
1986
1986
1986
* FT « flow through, S * static, U * unmeasured, M - measured
** not used in criteria development due to preferential use of FT,M tests
-------
Table 2. Species Mean Acute Values and Genus Mean Acute Values
for Endrin
Rank
28
27
26
25
24
23
22
21
20
19
18
17
16
GMAV
ug/1
64
53
43
38
34
24
4.6
3.0
2.5
2.1
1.6
1.5
1.3
Species
Mayfly,
Hexagenia bilineata
Crayfish,
Orconectes nais
Crayfish,
Orconectes immunis
Annelid,
Lumbnculus variegatus
Cladoceran,
Daphnia magna
Cladoceran,
Daphnia pulex
Cladoceran,
Simocephalus serrulatus
Cladoceran,
Ceridaphnia reticulata
Snipe fly,
Atherix variegatus
Anphipod,
Gamnarus fasciatus
Amphipod ,
Garamarus lacustris
Bullfrog tadpole
Rana catesbiana
Damselfly,
Ischnura verticalus
Guppy,
Poecilia reticulata
Sowbug ,
Asellus brevicaudus
Glass shrimp,
SMAV
ug/1
64
32
89
43
59
25
34
24
4.6
3.1
3.0
2.5
2.1
1.6
1.5
1.3
Palaemonetes kadiakensis
-------
Table 2. cont. Species Mean Acute Values and Genus Mean Acute
values for Enclrin
Rank
15
14
13
12
11
10
9
8
7
6
5
4
3
2
GMAV
ug/l
0.95
0.85
0.84
0.78
0.76
0.69
0.68
0.57
0.54
0.49
0.46
0.32
0.31
0.21
Species
Goldfish,
Carassius auratus
Flagfish,
Jordanella floridae
Midge,
Tanytarsus dissimilis
stonefly,
Pteronarcys californica
Stonefly,
Claassenia sabulosa
Mosquitofish,
Gambusia affinis
Black bullhead,
Ictalurus melas
Channel catfish,
Ictalurus punctatus
Coho salmon,
Oncorhynchus kisutch
Chinook salmon,
Oncorhynchus tshawytscha
Rainbow trout,
Oncorhynchus mykiss
Stonefly,
Pteronarcella badia
Fathead minnow,
Pimephales promelas
Brook trout,
Salvelinus fontinalis
Carpt
Cyprinus carpio
Larqemouth bass,
Micropterus salmoides
Bluegill,
SMAV
ug/l
0.95
0.85
0.84
0.78
0.76
0.69
1.1
0.42
0.51
1.2
0.3
0.54
0.49
0.46
0.32
0.31
0.21
Lepomis macrochirus
-------
Table 2. cont. Species Mean Acute Values and Genus Mean Acute
values for Endrin
Rank
1
GMAV
ug/l
0.15
Species
Yellow perch,
SMAV
ug/l
0.15
Perca flavescens
-------
Table 3. Acute to Chronic Ratios for Endrin
Species
Grass shrimp,
Palaemonetes pugio
Flagfish,
Jordanella floridae
Sheepshead minnow,
Cyprinodon variegatus
Acute
Value
0.72
0.85
0.36
Chronic
Value
(ug/1)
0.039
0.26
0.19
Ratio
18
3.3
1.9
Reference
EPA, 1980
-------
REFERENCES
EPA. 1980. Ambient Water Quality Criteria for Endrin. EPA 440/5-80-047
EPA, 1991. Acute 96-hr flow-through freshwater exposures with endrin and
dieldrin using an annelid (Lumbriculus variegatus). Unpublished data.
Elnabarawy, M.T., A.N. Welter and R.R.Robideau. 1986. Relative sensitivity
of three daphnid species to selected organic and inorganic chemicals.
Environ. Toxicol. Chem. 5:393-398
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for Developing
Aquatic Life Criteria. April, 1991
Mayer, F. and M.R. Ellersieck. 1986. Manual of Acute Toxicity: Interpretation
and Data Base for 410 Chemicals and 66 Species of Freshwater Animals.
USDI Publication #160.
Thurston, E.V, T.A Gilfoil, E.L Meyn, R.K. Zajdel, T.I. Aoki and G.D.
Veith. 1985. Comparative toxicity of ten organic chemicals to ten common
aquatic species. Water Res. 19(9):1145-1155.
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 and 2 Aquatic Life Criteria for Lindane
CAS # 58-89-9
July, 1991
Introduction
The criteria presented in this document were developed pusuant to
both the Tier 1 and Tier 2 approaches of the Great Lakes Water
Quality Initiative Procedure for Deriving Aquatic Life Criteria
(GLWQI, 1991). A search of aquatic toxicity literature generated
subsequent to publication of the EPA criteria document for
lindane (EPA, 1980) led to the addition of new acute toxicity
data to the EPA data base (Table 1). The combined data bases were
used as the basis for the Tier 1 and 2 criteria.
Criterion Maximum Concentration (CMC)
Sufficient data were available to calculate a Tier 1 acute
criterion for lindane (Table 2). The Final Acute Value (FAV) was
calculated using the lowest 4 Genus Mean Acute Values (GMAV)
(Table 2). This analysis resulted in an FAV of 1.902 ug/1. No
adjustment of this value was necessary to protect for
commercially or recreationally important species of the Great
Lakes Basin.
The Criterion Maximum Concentration (CMC) was calculated by
dividing the FAV by 2, resulting in a CMC of 1.0 ug/1.
Secondary Continuous Concentration (SCC)
The EPA criteria document lists two acceptable acute to chronic
ratios (Table 3). A secondary acute to chronic ratio (SACK) was
calculated as the geometric mean of the cladoceran and midge ACRs
and a default value of 25 for the lacking ACR. The resulting SACR
is 37.3.
The Secondary Chronic Value (SCV) was calculated as the FAV
divided by the SACR. The resulting SCV is 0.0509 ug/1. The SCC is
0.05 ug/1. There was no need to lower this value to protect for
commercially or recreationally important species of the Great
Lakes basin.
Summary
The lindane criterion are:
GLWQI EPA
FAV = 1.902 ug/1 FAV - 2 ug/1
CMC =1.0 ug/1
SCV - 0.0509 ug/1 FCV =0.08 ug/1
SCC - 0.05 ug/1
-------
Table 1. Aquatic Acute Values for Lindane Added Subsequent to Publication of the
Ambient Water Quality Criteria for Lindane (EPA, 1980)
Test
Duration LC/EC50
Species
Cladoceran,
Daphnia magna
Cladoceran,
Daphnia magna
Amphipoda,
Gammarus lacustris
Snail,
Lymneae stagnalis
Stonef ly,
Pteronarcys californicus
Stonef ly,
Pteronarcys californicua
Damselfly,
Lestes congener
Backswimmer ,
Notonecta undulata
Crawling water beetle,
Peltodytea sp.
Coho salmon,
Oncorhynchus kisutch
Lake trout,
Salvelinus namaycush
Lake trout,
Salvelinus namaycush
Brown trout,
Salmo trutta
Brown trout,
Salmo trutta
Brown trout,
Salmo trutta
Rainbow trout,
Oncorhynchus my kiss
Rainbow trout,
Oncorhynchus mykiss**
Method*
S,U
S,M
s,u
s,u
s,u
s,u
s,u
s,u
s,u
s,u
s,u
s,u
s,u
s,u
FT, U
FT,M
S,U
(hrs)
48
48
96
96
96
96
96
96
96
96
96
96
96
96
96
96
96
-------
Table l.(cont). Aquatic ftcute Valuea for Lindane Added Subsequent to
Publication of the Ambient Water Quality Criteria for Lindane
{EPA, 1980)
Specie*
Test
Duration LC/ECSO
Method* (hrB) (ug/1)
Reference
Rainbow trout,
Oncorhynchui mykiaa**
Rainbow trout.
Oncorhynchue mykiaa**
Rainbow trout,
Oncorhynchua mykiaa**
Rainbow trout.
Oncorhynchua mykiaa
Bluegill,
Lepomia macrochirus
Bluegill,
Lepomia macrochirus
Green sunfieh.
LepomiB cyanellua
Green sunfish,
Lepomis cyanellua
Yellow perch,
Perca flaveacens
Fathead minnow,
Pimephalea promelaa
Fathead minnow,
Pimephalea promelaa
Fathead minnow,
Pimephalea promelaa
GoldfiBh,
Caraaaiua auratua
GoldfiBh,
Caraaaiua auratua
Channel catfiah,
Ictalurua punctatua
Fowler a toad.
Bufo woodhouaei fowleri
Heetern chorua frog,
Paeudacria triaeriata
s,u
s,u
s,u
FT,M
s,u
s,u
s,u
s,u
FT,U
FT,U
S,U
S,U
S,D
S,U
s,o
s,u
s,o
96
96
96
96
96
96
96
96
96
96
96
96
96
96
96
96
96
24
31
41
30
57
56
70
63
23
77
67
86
90
105
49
3200
2650
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
Tooby and Ourbin, 197S
Randall, et al. 1979
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
Macek and McAllister, 1970
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
Mayer & Elleraieck, 1986
« S - atatic, SR - static renewal, FT * flow through, U - unmeasured,
M - measured
** Value not uaed due to preferential use of FT,M tests.
-------
Table 2. SMAVs and GMAVs for Lindane*
Rank
23
22
21
20
19
18
17
16
15
14
13
12
11
Specj.es
Fowlers toad,
Bufo woodhouBi fowleri
Western chorus frog,
Pseudacris trxaeriata
Cladoceran,
Simocephalus eerrulatue
Cladoceran,
Daphnia magna
Cladoceran,
Daphnia pulex
Midge,
Chironomus tentans
Guppy,
Poecilia reticulata
Goldfish,
Carassius auratus
Carp,
Cyprinus carpxo
Fathead minnow,
Pimephalea promelas
Bluegill,
Lepomis macrochirus
Redear sunfish,
Lepomia microlophus
Green Bunfish,
Lepomis cyanellus
Catfish,
Ictalurue punctatus
Black bullhead,
Ictalurus tnelaa
Yellow perch,
Perca flavescens
Brook trout
Salvelinus fontinalis
Lake trout,
Salvelinus namaycush
SMAV
ug/1
3200
2650
676
630
460
207
138
121
90
72
59
83
76
46
64
40
44
28
GMAV
ug/1
3200
26SO
676
-
538
207
138
121
90
72
-
-
72
-
55
40
-
35
-------
Table 2. (Con't.) SMAVB and GHAVs for Lindane
Rank
10
9
8
7
6
5
4
3
2
1
Specie*
Rainbow trout.
Oncorhynchus mykiss
Coho salmon,
Oncorhynchus kisutch
Chinook ealmon,
Oncorhynchus tachawytacha
Largemouth bass,
MicropteruB salmoxdes
Amphipod,
Ga/nmaru* favciatus
Amphipod ,
Gammaruc lacuatris
Damsel f ly,
Leetee congener
Crawlxng water beetle,
Peltodytes Bp.
Brown trout.
Salmo trutta
iBOpod,
Aaallus brevrcaudua
Snai.1,
Lymnaea ntagnalia
Backswiinaer,
Notonecta undulata
Stonefly,
Fteronarcyc californicua
SMAV
ug/i
26
36
40
32
11
65
20
20
13
10
3.3
3
2.1
GMAV
ug/1
-
—
33
32
-
27
20
20
13
10
3.3
3
2.1
* includes data published in EPA, 1980 and data added subsequent to
publication of EPA, I960.
-------
Table 3. Lindane acute to chronic ratios
Acute Chronic
Value Value
Species ug/1 ug/1 Ratio Reference
Cladoceran,
Daphnia magna 485 14.5 33 EPA, 1980
Midge,
Chironomus tentans 207 3.3 63 EPA, 1980
-------
REFERENCES
Bluzat, R. and J. Senge. 1979. Effets de Trois insecticides (lindane,
Fenthion and carbaryl): Toxicite' aigue sur gnatre espace d'invertebres
limneques: Toxicite' chronique chez le mollusque pulmone' Lymnea. Environ,
Pollut. 18:51-70
EPA, 1980. Ambient Hater Quality Criteria for Hexachlorocyclohexane.
EPA 440/5-80-054
Federle, P.F, and W.J. Collins. 1976. Insecticide toxicity to three insects
from Ohio ponds. Ohio J. Sci. 76(l):19-24
GLHQI, 1990. Great Lakes Hater Quality Initiative Procedure for Developing
Aquatic Life Criteria.
Hermens, J., H. Canton, M. Stezger, and R. Hagman. 1984. Joint effects of
a mixture of 14 chemicals on mortality and inhibition of reporduction
of Daphnia magna. Aquatic Toxicol. 5:315-322
Macek, K.J. and H.A. McAllister. 1970. Insecticide susceptibility of some
common fish family representatives. Trans. Am. Fish. Soc. 99.20-27\
Mayer, F.L., and M.R. Ellersieck, 1986. Manual of acute toxicity: Interpretation
and database for 410 chemicals and 66 species of freshwater animals.
U.S. Dept. Interior, Fish and Wildlife Service, Resource Pub 160
Randall, H.F., W.H. Dennis and M.C. Harner. 1979. Acute Toxicity of
dechlorinated DDT, chlordane and lindate to Bluegill (Lepomis macrochirus)
and Daphnia magna. Bull. Environ. Contain. Toxicol. 31 (4):459-466
Tooby, T.E. and F.J. Durbin. 1975. Lindane residue accumulation and elimination
in rainbow trout (Salmo gairneri Richardson) and roach (Rutilua rutilus
linneaus). Environ. PoLlut. 8(2):79-89.
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Mercury (II)
Class 021
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Inititative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for mercury (EPA, 1985)
led to the addition of new acute toxicity data to the EPA data
base (Table 1). The combined data bases were used as the basis
for the Tier 1 criteria calculations.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (GMAV)(Table 2). This analysis resulted
in an FAV of 1.652 ug/1. No adjustment of this value was
necessary to protect for commercially or recreationally
important species of the Great Lakes basin.
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 0.83 ug/1.
Criterion Continuous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate a
Final Chronic Value (FCV) using the eight family approach of the
procedure. Therefore, the FCV was calculated by dividing the FAV
by the EPA derived final acute to chronic ratio of 3.731 (Table
1), resulting in a FCV of 0.442 ug/1 and CCC of 0.44 ug/1. No
adjustment of this value was necessary to protect commercially
or recreationally important species of the Great Lakes basin.
Summary
The mercury II criteria are:
GLWQI EPA
FAV = 1.652 ug/1 FAV = 4.857 ug/1
CMC =0.83 ug/1 CMC = 2.428 ug/1
FCV = 0.442 ug/1 FCV = 1.302 ug/1
CCC =0.44 ug/1
References
EPA, 1985. Ambient Water Quality Criteria for Mercury - 1984.
EPA 440/5-84-026. January 1985.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for
Deriving Aquatic Life Criteria. April, 1991
Elnabarawy, M.T., A.N. Welter and R.R. Robideau. 1986. Relative
sensitivity of three daphnia species to selected organic and
inorganic chemicals. Environ. Toxico. Chem. 5:393-398.
Khangarot, B.X. and P.K. Ray. 1987 B. Studies on the acute
toxicity of copper and mercury alone and in combination to the
-------
common guppy, Poecilia reticulata (PETERS). Arch. Hydrobiol.
110(2): 303-314
Kirubagaran, R. and K.P. Joy, 1988. Toxic effects of threee
mercurial compounds on survival and histology of the kidney of
the catfish Clarias batrachus (L.). Ecotoxicol. and Environ.
Saf. 15:171-179.
Martin, T.R. and D.M. Holdich. 1986. The acute lethality of
heavy metals to peracarid crustaceans (with particular
reference to fresh-water asellids and gammarids). Wat. Res.
20(9):1137-1147.
Paulose, P.V. 1988. Comparative study of inorganic and organic
mercury poisoning on selected freshwater organisms. J. Environ.
Biol. 9(2): 203-206.
Rossaro, B., G.F. Gaggino and R. Marchetti. 1986. Accumulation
of mercury in larvae and adults, Chironomus riparius (Meigen).
Bull. Environ. Contain. Toxicol. 37:402-406.
Spehar, R.L. and J.T. Fiandt. 1986. Acute and chronic effects of
water quality criteria-based metal mixtures on three agjuatic
species. Environ. Toxicol. and Chem. 5:917-931.
f
\&
-------
Table 1. Aquatic Acute Toxicity Data Added Subsequent to Publication of
the Ambient Water Quality Criteria for Mercury (EPA, 1985)
Species
LC/ECSO
Method* Chemical (ug/1) Reference
Cladoceran, Mercuric
Ceriodaphnia reticulata S,U chloride
Cladoceran,
Daphnia magna, S,U
Cladoceran,
Daphnia pulex S,U "
Amphipod,
Crangonyx pseudogracilis S,U
Midge,
Chironomus riparius S,M "
Mosquitof ish,
Gambusia af finis S,U "
Walking catfish,
Clarias batrachus S,U
Fathead minnow, Mercuric
Punephales promelas FT,M nitrate
Guppy, Mercuric
Poecilia reticulata R,U chloride
2.9
9.6
3.8
1.0
750
230
507
172
26
Elnabarawy et al.
Elnabarawy et al.
Elnabarawy et al.
Martin & Holdich,
Rossaro et al., 1
Paulose, 1988
Kirubagaran & Joy
Spehar & Fiandt,
Khangarot & Ray,
, 1986
, 1986
, 1986
1986
988
, 1988
1986
1987b
* S = static, FT = flow through, U » unmeasured, M = measured
-------
Table 2. Ranked Genus Mean Acute Values with Species Mean Acute-chronic
Ratios for inorganic Mercury.
Rank*
Genus Mean
Acute Value
(ug/1) species
Species Mean species Mean
Acute Value Acute-Chronic
(ug/1)** Ratio
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
2000
2000
2000
1200
1200
1000
1000
406.2
370
257
250
240
230
203
180
Stone fly,
Acroneuria lycorias
Mayfly,
Ephemerelia subvaria
Caddisfly,
Hydropayche betteni
Caddis fly,
(Unidentified)
Damsel fly,
(Unidentified)
Worm,
Nais sp.
Mozambique tilapia
Tilapia mossambica
Tubificid worm,
Spi rosperma ferox
Tubificid worm,
Spiroaperma nikolskyi
Snail,
Ap.exia hypnorum
Coho salmon,
Oncorhynchus kxsutch
Rainbow trout,
Oncorhynchus mykiss
Tu&ificid worm,
Quistardrilua multisetosus
Tubificid worm.
Rhyacodrilus montana
Walking catfish,
Clarias batrachus
Mcsquitofish,
Gambusia affinia
Tubifxcid worm,
2000
2000
2000
1200
1200
1000
1000
330
500
370
240
275
250
240
230
203
180
-
—
_
_
_
_
_
_
—
_
-
-
-
-
-
-
Limnodnlus hoffmeisten
-------
Table 2. Continued.
Rank*
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
Genus Mean
Acut* Value
649.2***
160
140
140
20
750
100
80
80
50
28
20
10
3.7 4.498
2.9
2.9
1.0
* Ranked from most resistant to most sensitive based on Genus Mean Acute Value.
** Calculation of the final acute-chronic ratio (ACR) also included a
mysid (Mysidopsis bahia) ACR of 3 095
*** not used in development of the final ACR
-------
common guppy, Poecilia reticulata (PETERS). Arch. Hydrobiol.
110(2): 303-314
Kirubagaran, R. and K.P. Joy, 1988. Toxic effects of threee
mercurial compounds on survival and histology of the kidney of
the catfish Clarias batrachus (L.). Ecotoxicol. and Environ.
Saf. 15:171-179.
Martin, T.R. and D.M. Holdich. 1986. The acute lethality of
heavy metals to peracarid crustaceans (with particular
reference to fresh-water asellids and gammarids). Wat. Res.
20(9):1137-1147.
Paulose, P.V. 1988. Comparative study of inorganic and organic
mercury poisoning on selected freshwater organisms. J. Environ.
Biol. 9(2): 203-206.
Rossaro, B., G.F. Gaggino and R. Marchetti. 1986. Accumulation
of mercury in larvae and adults, Chironomus riparius (Heigen).
Bull. Environ. Contain. Toxicol. 37:402-406.
Spehar, R.L. and J.T. Fiandt. 1986. Acute and chronic effects of
water quality criteria-based metal mixtures on three aquatic
species. Environ. Toxicol. and Chem. 5:917-931.
col
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Mercury (II)
Class 021
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Inititative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for mercury (EPA, 1985)
led to the addition of new acute toxicity data to the EPA data
base (Table 1). The combined data bases were used as the basis
for the Tier 1 criteria calculations.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (GMAV)(Table 2). This analysis resulted
in an FAV of 1.652 ug/1. No adjustment of this value was
necessary to protect for commercially or recreationally
important species of the Great Lakes basin.
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 0.83 ug/1.
Criterion Continuous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate a
Final Chronic Value (FCV) using the eight family approach of the
grocedure. Therefore, the FCV was calculated by dividing the FAV
y the EPA derived final acute to chronic ratio of 3.731 (Table
1), resulting in a FCV of 0.442 ug/1 and CCC of 0.44 ug/1. No
adjustment of this value was necessary to protect commercially
or recreationally important species of the Great Lakes basin.
Summary
The mercury II criteria are:
GLWQI EPA
FAV = 1.652 ug/1 FAV = 4.857 ug/1
CMC =0.83 ug/1 CMC = 2.428 ug/1
FCV = 0.442 ug/1 FCV = 1.302 ug/1
CCC =0.44 ug/1
References
EPA, 1985. Ambient Water Quality Criteria for Mercury - 1984.
EPA 440/5-84-026. January 1985.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for
Deriving Aquatic Life Criteria. April, 1991
Elnabarawy, M.T., A.N. Welter and R.R. Robideau. 1986. Relative
sensitivity of three daphnia species to selected organic and
inorganic chemicals. Environ. Toxico. Chem. 5:393-398.
Khangarot, B.X. and P.K. Ray. 1987 B. Studies on the acute
J__ J _» .__? ____ __ _ J •_»_. _•* m_B _—.» * —_••« # ** T _—k——. _*>h —» *—__r4 ^ W4 _~*_~*~Vl V* ~i V* ^ ^ "1 _^'l^ ^ -j , , j ,_
JO
_nangaro^, O.A. ana r.j\. nay. xi»o/ D. <_>v-fluxes w.i i_ii«- aww.w«
toxicity of copper and mercury alone and in combination to the /
-------
Table 1. Aquatic Acute Toxicity Data Added Subsequent to Publication of
the Ambient Water Quality Criteria for Mercury (EPA, 1985)
Species
LC/EC50
Method* Chemical (ug/1) Reference
Cladoceran, Mercuric
Ceriodaphnia reticulata &,U chloride
Cladoceran,
Daphnia magna, S,U "
Cladoceran,
Daphnia pulex S,U "
Amphipod,
Crangonyx pseudogracilis S,U "
Midge,
Chironoenus riparius S,M "
Mosquitof ish,
Gambusia af finis S,U "
Walking catfish,
Clarias batrachus S,U
Fathead minnow. Mercuric
Pimephales promelas FT,M nitrate
Guppy, Mercuric
Poecilia reticulata R,U chloride
2.9
9.6
3.8
1.0
750
230
507
172
26
Elnabarawy et al.
Elnabarawy et al.
Elnabarawy et al.
Martin S Holdich,
, 1986
, 1986
, 1986
1986
Rossaro et al., 1988
Paulose, 1988
Kirubagaran 6 Joy
Spehar & Fiandt,
Khangarot & Ray,
, 1988
1986
1987b
* S = static, FT = flow through, U * unmeasured, M = measured
-------
Table 2 Ranked Genus Mean Acute Values with Species Mean Acute-Chronic
Ratios for Inorganic Mercury.
Rank*
30
29
28
27
26
25
24
23
22
21
20
19
IS
17
16
Genus Mean
Acute Value
2000
2000
2000
1200
1200
1000
1000
406 2
370
257
250
240
230
203
180
Species
Stonef ly,
Acroneuria lycorias
Mayfly,
Ephemerelia subvaria
Caddisfly,
Hydropsyche betteni
Caddisfly,
(Unidentified)
Damsel fly,
(Unidentified)
Worm,
Nais sp.
Mozambique tilapia
Tilapia mossambica
Tubificid worm,
Spirosperma ferox
Tubificid worm,
Spirosperma nikolskyi
Snail,
Aplexia hypnorum
Coho salmon,
Oncorhynchus kisutch
Rainbow trout,
Oncorhynchus mykiss
Tubificid worm,
Quistardrilua multisetosus
Tubificid worm,
Rhyacodrilus montana
Walking catfish,
Clarias batrachus
Mosquitof ish,
Gambusia affinis
Tubificid worm,
Species Mean
Acute Value
(ug/1)**
2000
2000
2000
1200
1200
1000
1000
330
500
370
240
275
250
240
230
203
ISO
Species Mean
Acute-Chronic
Ratio
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Limnodrilue hoffmeisten
-------
Table 2. Continued.
Rank*
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
GenuB Mean
Acute Value
(ug/1)
163
160
140
140
122
100
80
80
SO
28
20
10
3.3
2.9
1.0
Species
Fathead minnow,
Punephales promelaa
Bluegill,
Lepomia macrochiruB
Tufcificid worm,
Tub if ex tub if ex
Tub if ic id worm,
St> lodrilua heringianua
Mictge,
Cturonomua ap.
Midge,
Chj ronomua ripariua
Tubificid worm,
Vanchaeta pacifica
Tubificid worm,
Branchiura aowerbyi
Snail,
Amnicola ap.
Crayfish,
Orconectea lumoaua
Guppy,
Poecilia reticuLata
Crayfish,
Faxonella clypeatus
Amphipod ,
Gammurua ap.
Cladoceran,
Daphnia magna
Cladoceran,
Daphnia pulex
Cladoceran,
Cerodaphnia reticulata
Amphipod ,
Crangonyx pseudogracilia
speciea Mean
Acute Value
(ug/i)
163
160
140
140
20
750
100
80
80
SO
28
20
10
3.7
2.9
2.9
1.0
Specie a Mean
Acute-Chronic
Ratio
> 649.2***
-
-
-
-
-
-
-
-
-
-
-
—
4.498
-
-
—
* Ranked from moat remstant to most sensitive based on Genua Mean Acute Value.
** Calculation of the final acute-chronic ratio (ACR) alao included a
myBid (Mysidopaia bahia) ACR of 3.095
*** not used in development of the final ACR
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Nickel
Class 022
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for nickel (EPA, 1986)
led to the addition of new acute data to the EPA data base. The
new data are found in Table 1. The combined data bases were used
as the basis for the Tier 1 criteria calculations. The toxicity
of nickel is hardness dependent. Therefore, all data used in the
criteria calculations were normalized to a hardness of 50 mg/1.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (Table 2). This analysis resulted in an
FAV of 615 ug/1 at a hardness of 50 mg/1. No adjustment of
this value was necessary to protect for commercially or
recreationally important species of the Great Lakes basin.
The Final Acute Equation =
0.846(lnH) + 3.11
e
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 308 ug/1 (at a hardness of 50 mg/1).
Criterion Continuous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate a
Final Chronic Value (FCV) using the eight family approach of the
procedure. Therefore, the FCV at a hardness of 50 mg/1 was
calculated by dividing the calculated FAV by the EPA derived
final acute to chronic ratio of 17.99 (Table 2), resulting in a
FCV of 34.21 ug/1 at a hardness of 50 mg/1. No adjustment of
this value was necessary to protect commercially or
recreationally important species of the Great Lakes basin. The
CCC is 34 ug/1 at a hardness of 50 mg/1.
The acute slope of 0.846 was used to develop the final chronic
equation, as was done in the EPA document. The Final Chronic
Equation =
0.846 (InH) + 0.223
e
Summary
The nickel criterion are:
GLWQI
FAV = 615 ug/1 ( at hardness = 50 mg/1)
CMC = 308 ug/1 (at hardness = 50 mg/1)
FCV = 34.21 ug/1 (at hardness = 50 mg/1)
CCC = 34 ug/1 (at hardness = 50 mg/1)
-------
0.846(lnH) + 3.11
Final Acute Equation = e
0.846(lnH) + 0.223
Final Chronic Equation - e
EPA
FAV = 1578 ug/1 (at hardness = 50 mg/1)
CMC » 789 ug/1 (at hardness « 50 mg/1)
FCV » 87.72 ug/1 (at hardness - 50 mg/1)
0.846 (In H) + 4.054
Final Acute Equation = e
0.846 (In H ) + 1.1645
Final Chronic Equation = e
References
EPA, 1987. Ambient Aquatic Life Water Quality Criteria for
Nickel. EPA 440/5-86-004. September, 1986.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for
Deriving Aquatic Life Criteria. April, 1991
Martin, T.R., and D.M. Holdich. 1986. The acute lethal toxicity
of heavy metals to peracarid crustaceous (with particular
reference to freshwater asellids and gammands). Wat. Res.
20(9):1137-1147.
Nebeker, A.V., et al. 1986. Effects of copper, nickel, and zinc
on three species of Oregon freshwater snails. Environ. Tox.
Chemi. 5:807-811.
Powlesland, C. and J. George. 1986. Acute and chronic toxicity
of nickel to larvae of Chironomus riparis (Meigan). Env. Poll.,
Ser. A. 42:47-64
-------
Table 1. Aquatic Acute Toxicity Data Added Subsequent to Publication of the Ambient Water Quality Criteria
for Nickel (EPA, 1985)
Species
Adjusted
Hardness LC/EC50 LC/EC50
Method* Chemical (mg/1) (ug/1)** (ug/1)*** Reference
Snail, F,U
Physa gyrina
Amphipod, S,U
Crangonyx pseudogracilis
Midge, (1st instar) S,U
Chironomus riparis
Midge, (let instar) S,U
Chironomus riparis
Midge, (1st instar) S,U
Chironomus rxpans
Midge, (2nd instar) S,U
Chironomus riparis****
Midge, (2nd instar) S,U
Chironomus riparis****
Midge, (2nd instar) S,U
Chironomus riparis****
Nickel 26 298 518 Nebecker, et al., 1986
chloride
50 66,100 66,100 Martin & Holdich, 1986
55 72,400 66,635 Powlesland & George, 1986
55 81,300 74,832 Powlesland & George, 1986
55 84,900 78,146 Powleeland & George, 1986
55 184,000 169,362 Powlesland & George, 1986
55 150,000 138,066 Powlesland & George, 1986
55 174,000 160,157 Powlesland & George, 1986
* S = static; F = flow-through, U = unmeasured
** Results are expressed as nickel, not as the chemical.
*** LCSOa and ECSOs were adjusted to hardness - 50 mg/1 using the pooled slope of 0.8460
****Not used in criteria development due to availability of data for a
more sensitive life stage
-------
Table 2. Ranked Genus Mea.n Acute Values with Species Mean Acute-Chronic
Ratios for Nickel.
Genus Mean
Acute Value
Species Mean Species Mean
Acute Value Acute-Chronic
Rank*
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
(ug/1)**
73041
66100
43250
40460
30200
21320
21200
14100
13380
13000
12770
12180
9839
9661
8697
9530
Species
Midge,
Ch ironomus riparis
Amphipod ,
Cr«ngonyx pseudogracilis
Banded kill i fish,
Fundulua diaphanis
stonef ly,
Acironeuria lycorias
Caddisfly
Unidentified ap.
Goldfish,
Carasaius auratus
Damsel fly,
Unidentified sp.
Worm,
Nais sp.
Rainbow trout,
Oncorhynchua mykiss
Amphipod,
Gammarus flp.
Snail,
Airmicola sp.
American eel,
Anguilla rostrata
Common carp,
Cyprinus carpio
Guppy,
Poecilia reticulata
White perch,
Morone americana
Striped baas,
Morone saxatilis
Pumpkinseed,
(ug/1)**
73041
66100
43250
40460
30200
21320
21200
14100
13380
13000
12770
12180
9839
9661
12790
5914
7544
Ratio***
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Lepomxs gibbosus
-------
Table 2. Continued.
Rank*
Genus Mean
Acute Value
(ug/1)** Species
518
Daphnia pulicaria
Claodoceran,
Daphnia magna
Snail,
Physa gyrina
Species Mean Species Mean
Acute Value Acute-Chronic
(ug/1)** Ratio***
5
4
3
2
6707
4636
4312
1SOO
Bluegill,
Leporoia macrochirus
Fathead minnow,
Pimephales promelaa
Mayfly,
Ephemerella subvaria
Rock base,
Ambloplitea rupeatns
Claodoceran,
12040
6707 35.58
4636
4312
2042
1102
518
29.86
* Ranked from most resistant to most sensitive based on Genus Mean Acute Value.
** Genus Mean Acute Values are at hardness = 50 mg/1.
*** Calculation of the final Acute-chronic ratio {ACR) also included a
mysid (Mysidosis bahia) ACR of 5.478
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Parathion
CAS #56-38-2
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative (GLWQI, 1991). No new data were available to
supplement the parathion data base in EPA, 1986. Therefore,
these data (Table 1) were used as the basis for the Tier 1
criteria.
Criterion Maximum Concentration
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (Table 2). This analysis resulted in an
FAV of 0.1298 ug/1. No adjustment of this value was necessary to
protect commercially or recreationally important species of the
Great Lakes basin.
The CMC was calcualted by dividing the FAV by 2, resulting
in a CMC of 0.065 ug/1.
Criterion Continuous Concentration (CCC)
The acute to chronic ratio (ACR) selected for developmemt of the
CCC is the Daphnia magna ACR of 10.1. This value was selected
based on the guidance in GLWQI, 1991, which indicates that if
the species mean ACR seems to increase as the Species Mean Acute
Values increase (as occurs with the parathion ACR data set), the
ACR should be calculated as the geometric mean of the ACRs for
species whose SMAVs are close to the FAV.
The Final Chronic Value (FCV) was calculated by dividing the FAV
(0.1298 ug/1) by the ACR (10.1). The resulting Final Chronic
Value is 0.0128 ug/1. The CCC is 0.013 ug/1. No adjustment of
this value was necessary to protect commercially or
recreationally important species of the Great Lakes basin.
Summary
The parathion criteria are:
GLWQI EPA
FAV = 0.1298 ug/1 FAV = 0.1298 ug/1
CMC = 0.065 ug/1 CMC = 0.0649 ug/1
FCV = 0.0128 ug/1 FCV = 0.0128 ug/1
CCC = 0.013 ug/1
-------
Taala I. Ac** totclty «f rtnrtftlo* «• A*a»tl«
IC90 Saacta*
•r GCW Ac«ta Valw
<»oA>*«» U«A>
Hat
FRgSHMATER SPECI |S
Tub 1 field vora.
Tubltlcld warm,
Tublfex sp.
Cladocaran,
Daphnla naqna
Cladocaran,
Oaphn la tugna
Cladocaran (<24 hr) ,
Oaphn la aaana
Cladocaran, (<24 hr) ,
Oaphn la **<]na
Cladocaran «24 hr) ,
Daphnla Mgna
Cladoceran (1st Instar).
Daphnla pule*
Cladocaran (1st Instar),
SlMocaphalus sarrulatus
Iso pod,
Asallus bravlcaudus
Isopoda (*ature) ,
A sail us bravlcaudus
Amphlpod (natural,
GaMMrus lasclatus
Am phi pod (itatura),
GaMMMrus lasclatus
Amphtpod (*atura) ,
GaMMrus fasclatus
S,
s,
s,
s.
s.
s.
F,
s.
s.
S.
S.
s,
F.
S.
U
U
U
U
M
U
H
U
U
U
U
U
U
U
Analytical
(99.61)
Analytical
Raagant
(99|)
Analytical
(991)
Raagant
(99«)
Technical
(98.71)
Technical
Tachntcal
(9B.7f)
Tachn leal
(98.7J)
Tachn leal
(98.71)
Tachn leal
(98.71)
Tachntcal
(98.71)
9,250""»
9,230«"
0.8
1.8
1.27
1.3
t.o
0.60
0.47
600
2,130
2.1*
4.5*
I.J*
9.230 Mnlttan and Goodnight
1966
9,230 Mhlttan and Goodnight
1966
Boyd 1997
Brtngaann and Kuhn I960
Spacla 1976, Spacla
at al. 1981
Portland 1980
1.0 Spacla 1976, Spacla
at al. 1981
0.60 Johnson and F Inlay
1980
0.47 Johnson and F Inlay
1980
Sandars 1972
1.130 Johnson and F Inlay
1980
Sandars 1972
Sandars 1972
Johnson and F Inlay
1980; Sandars 1972
T.M* I.
-------
Ta»U I.
Saacla*
Mayfly.
Cloaoji dlgtariM
Mayfly,
Cloao* dlptariai
Mayfly.
Cloaon dlptaru*
Mayfly tjuv«n!«*r ,
Haxaqanla blUnaata
Oansalfly (Juvanlla).
Ischnura van tl calls
OaMsalfly.
Lastas conaanar
Stonafly,
Ptaronarcalla bad la
Stonafly 1 naiad),
Ptaronarcys callfornlca
Stonafly (2nd yaar class),
Ptaronarcys callfornlca
Stonafly (naiad),
Acronaurla pact flea
Stonafly (2nd yaar class),
Claassanla sabiilosa
s, u
s. u
R. U
S, U
s. u
s, u
s, u
s. u
s, u
s. u
s. u
Analytical
Analytical
(991)
Analytical
(99X)
Technical
(98.71)
Tachn leal
(98.7*)
Tachn leal
Tachn leal
(98.7*)
Tachn leal
(95|)
Tachn leal
(98.7*)
Tachn leaf
Tachn leal
(98.7*)
or GCM
(.a/I)"*
2.5
2.6
1.7
15
0.64
3.0
4.2
52*
5.4
2.9
1.5
Acvta V«la»
(••A)
"
2.227
15
0.64
3.0
4.2
5.4
2.9
1.5
Nafaraaca
Oortland 1980
Dortland I960
Oortland 1980
Johnson and F Inlay
1980
Johnson and F Inlay
1980
Fadarla and Collins
1976
Johnson and F Inlay 1
Sandars and Copa I9C
Jansan and Gaufln I!
Johnson and Fin lay 1
Sandars and Copa 1*
Jansan and Gaufln H
Johnson and F Inlay 1
Sandars and Copa 196
-------
Table I.
Specie*
CCS* Specie*
or K90 Acvt* Vel
l.aA>M« UttA)
00
Rainbow trout
(try, 7? .
Sal no galrdnerl
Brown trout (16-19 cm),
Sal*o trutta
Brook trout (Juvenile),
Salvellnut fontlnalU
Lake trout (0.7 g) ,
Salvellnus naawycush
Goldtlsh (Juvenile),
Carasslus auratus
GoldHsh (0.9 g),
Carasslus auratus
Fathead winnow (1-1.9 g>,
Plaephales prowelas
Fathead Minnow (1-1.5 g),
Pl*ephale> prone las
Fathead nlnnow (1-1.5 g),
Pluephales proaelas
fathead Minnow (1-1.) g) ,
PlMphales proMelas
Fathead Minnow (Juvenile),
PlMephales promelas
Fathead Minnow (adult),
PNephales prone las
R.
F,
F.
s.
s.
S.
s,
s,
s,
s,
s.
s.
u
M
M
U
U
U
U
U
U
U
U
M
(99f)
t
Reagent
<99f)
Reagent
I99O
Technical
(98.71)
Techn leal
(99* )
Technical
(98.71)
Technical
(96.)()
Technical
< 96.51)
Technical
(96.)|)
Technical
(96.)|)
Technical
(99%}
Reagent
(99O
1,400
1,510
1.760
1,920
2,700
1,830
1,400
1.600
2,800
1,700
1,300
1.600
1,419
1.510
1,760
1,920
2,223
-
-
-
Ret«
Yen Leeuwen et al . 196)
Specie 1976, Spade
•t al. 1981
Spacle 1976, SpocU
•t al. 1981
Johnson and F Inlay I960
Pickering «t al . 1962
Johnson and FTnUy I960
Henderson and Pickering
19)8
Henderson and Pickering
1956
Henderson and Pickering
19)8
Henderson and Pickering
19)8
Pickering et al. 1962
Spacle 1976, Spacle
et al. 1961
Ta*l*
-------
Tafela I. f«MWtl«M«I
Koraan shrlep (adult),
Palaa»on •acrodactylus
Koraan stir lap (adult).
PalaaaoB •acrodactvlus
Strlpad bass I Juvanlla),
Morona saxatllls
F, U
S, U
f, 0
UCM Saaelas
or KM teat* Val
«a«AI*»» UaAl
5AITMATER SPECIES
(991) 17.8
11.5
17.8
(M<)
14.51
17.8
Hat
Earnast 1970
Earn«»t 1970
Korn and Earnast 1974
" S • static, R • ranawal; F • flow-through. U • uiMaasurad. H » aaasiirad.
** Parcant purity Is glvan In paranthasas «tian avallabla.
*•• if tha eoncantrattons vara not naasurad and tha publlshad rasults Mara not raportad to ba adjusted
for purity, tha publlshad rasults vara MI)tip)lad by tha purity It It MS raportad to ba lass than
971.
•*•• Llanodrllls sp. and TubI fax sp. «ara tastad togathar, but appaarad to ba aqually raslstant.
tt
Not tisad In calculation of Spaclas Maan Acuta Valua bacausa data ara avallabla for a aora
sansltlva Ufa staga.
4*C, not usad In calculations.
22 *C.
T«fclty
-------
Ae«t« aliwa «ltfc SfNwtai MM* Ac«tarQmMf« Mtlo*
tajik*
Jl
30
29
28
27
26
29
24
23
22
(a«A)
9,230
9,230
2,690
2,223
1.836
..«.
1.130
1.000
839.6
688.7
1l~lff
FRESHWATER SPECIES
Tub! He Id Nora.
Tublfax sp.
Tub 1 flc Id wore.
Llmodrllu* tp.
Channal cattish,
Ictalurus punctatus
Goldfish,
Carasslus auratus
Brook trout.
Salvallnus fontlnaMs
Laka trout,
Salvallnus namaycush
Cutthroat trout.
Salax> clark)
Brown trout,
Salao trutta
Rainbow trout.
Se>ax> aalrdiwl
Isopod,
AS«|)MS bravlcaudus
Mestarn chorus frog,
Psaudacrls trtsartata
Fathaad nlnnow,
Plmaphalas proaalas
Gra«n sun fish,
Lapoaili cyanal lus
Hluaglll.
$a«cla« NMM
Acvt* Valu*
l.aAl"
9,230
9,230
2,690
2,223
1.760
1,920
1.960
1.910
1.419
1,130
1,000
839.6
930
310
^SSS?*
-
-
-
-
-
79.49
-
2,121
Lepoails nacrochlrus
-------
ro
*ea*» Vain* tevt* Vila* Acvta~Chraa)lc
Rank" faaAI SMtlw (affAIM Mtla**«
8
7
6
5
4
3
2
1
1.697 Midge,
Chiroaoawa rlaarlus
1.3 Stonafly,
Claatsenla sabulosa
1.127 Aaphlpod,
Gaamarus fasclatus
Gaaawus lacustrls
0.8944 Phantoai *ldqa,
ChaobOTiit sp.
0.7746 Cladoceran,
Daphnla aiagna
Cladocaran,
Daphnla gulax
0.64 OaMselfly,
1 schnur a vent 1 ca 1 1 a
0.47 Cladocaran,
Slaocaphalua serrulatus
0.04 0- ay fish.
Orconactas nal«
1.697
1.5
0.3628
3.5
0.8944
1.0
0.60
0.64
0.47
0.04
10.10
• Ranked fro* Most raslctant to «o*t sensitive bused on Genus Mean Acute Value.
Inclusion of "greater than* values does not necessarily l«ply a true ranking, but
does allow us* of all genera for which data are available so that the Final Acute
Value Is not unnecessarily looered.
•
•» From Table t.
»•• Frow Tab la 2.
Tatl* S.
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Pentachlorophenol
CAS # 87-86-5
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for pentachlorophenol
(EPA, 1986) led to no acceptable new acute or chronic toxicity
data to supplement the EPA data base. Therefore, the EPA data
base was used as the basis for the Tier 1 criteria calculations.
A modification of the EPA data base was made during this review
due to the recent reclassification of the rainbow trout from
Salmo gairdneri to Oncorhynchus mykiss.
The toxicity of pentachlorophenol is pH dependent. Therefore,
all data used in the criteria calculations were normalized to a
pH of 6.5 S.U..
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (Table 1). This analysis resulted in an
FAV of 10.57 ug/1 (at pH = 6.5). No adjustment of this value was
necessary to protect for commercially or recreationally
important species of the Great Lakes basin.
Using the EPA derived slope of 1.005, the Final Acute Equation ~
1.005(pH) - 4.174
e
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 5.3 ug/1 at a pH of 6.5.
Criterion Continuous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate a
Final Chronic Value (FCV) using the eight family approach of the
procedure. Therefore, the FCV was calculated by dividing the
calculated FAV by the EPA derived final acute to chronic ratio
of 3.166, resulting in a FCV of 3.338 ug/1 (at pH = 6.5). The
CCC is 3.3 ug/1. No adjustment of this value was necessary to
protect commercially or recreationally important species of the
Great Lakes basin. The Final Chronic Equation =
1.005(pH) - 5.327
e
Summary
The pentachlorophenol criteria are:
GLWQI
FAV - 10.57 ug/1 (at pH - 6.5)
CMC =5.3 ug/1 (at pH - 6.5)
FCV = 3.338 ug/1 (at pH - 6.5)
-------
CCC =3.3 ug/1 (at pH =6.5)
1.005 (pH) - 4.174
Final Acute Equation = e
1.005 (pH) - 5.327
Final Chronic Equation - e
EPA
FAV = 10.97 ug/1 (at pH = 6.5)
CMC = 5.485 ug/1 (ar pH = 6.5)
FCV =3.46 ug/1 (at pH = 6.5)
1.005(pH)-4.137
Final Acute Equation = e
1.005 (pH) - 5.290
Final Chronic Equation = e
References
EPA, 1986. Ambient Aquatic Life Water Quality Criteria for
Pentachlorophenol. EPA 440/5-88-009. September 1986.
GLWQI, 1991. Great Lakes Water Quality Inititive Procedure for
Deriving Aquatic Life Criteria. April, 1991
-------
Table 1. Ranked Genus Mean Acute Values with specie* Mean Acute-Chronic
Ratios for Pentachlorophenol.
Genus Mean
Acute Value
Rank* (ug/1)**** Species
Species Mean Species Mean
Acute Value Acute-Chronic
(ug/1)**** Ratio***
32
31
30
>43920
11260
10610
Crayfish,
Orconotes immunia
Midge,
Tanytarsus diasimilis
Sciomyzid,
>43920
11260
10610
29
28
27
26
25
24
23
22
21
20
19
18
17
Sepedon fuscipennie
417.7 Tubificid worm, 417.7
Rhyacodrilus montana
408.2 Tubificid worm, 408.2
Stylodrilua heringianua
403.2 Snail, 403.2
Gilila altilis
361.6 Tubificid worm, 239.5
Spirosperma ferox
Tubificid worm, 545.8
Spirosperma nikoiakyl
317.5 Tubificid worm, 317.5
Quistradrilus multisetoaus
291.6 Flagfiah, 291.6
Jordaneila floridae
224.2 Tubificid worm, 224.2
Tubifex tubifex
195.4 Guppy, 195.4
Lebistes reticulata
182.5 Tubificid worm, 182.5
Limnodnlus hoffmeisteri
172.1 Amphipod, 172.1
Crangonyx pseudogracilia
155.9 Tubificid worm, 155.9
Branchiura sowerbyl
132.1 Snail, 132.1
Physa gyrina
122.1 Amphipod, 122.1
GammaruB pseudolimnaeuB
>10.27**
-------
Table 1. Continued.
Genus Mean
Acute Value
Rank* (ug/1)**** species
Species Mean Species Mean
Acute Value Acute-Chronic
(ug/1)**** Ratio***
16
15
14
13
12
11
10
9
8
7
6
5
4
3
105
87.48
78.1
67.13
65.53
63.11
60.50
60.43
58.47
57.72
56.41
44.48
34.13
31.26
Laigemouth baas.
MicropteruB salmoidea
Amphipod,
Hy a lei la azteca
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia magna
Cladoceran,
Cex lodaphnia reticulata
Goldfish,
Caiassius auratus
Fathead minnow,
Punephales promelas
Moiiquitof ish,
Gambusia af finis
Snail,
ApJ exa hypnorum
Tuhi ficid worm.
Varichaeta pacifica
Cladoceran,
SuaocephaluB vetulus
Bluegill,
Lepomis macrochirus
Bu Llf rog.
Rano catesbiana
Brook trout,
SaLvelinuB fontinalis
Rainbow trout,
Oncorhynchus tnykiBB
Coho salmon,
OncorhynchuB kisutch
Sockeye salmon,
105
87.48
90.83
67.15
67.13
65.53
63.11
60.50
60.43
58.47
57.72
56.41
44.48
34.13
35.34
31.82
32.85
_
_
2.5
>15.79«*
—
4.535
—
_
-
0.8945
-
-
-
4.564
-
-
Oncorhynchus nerka
-------
Table 1. Continued.
Genus Mean
Acute Value
Rank* (ug/1)**** Species
species Mean Species Mean
Acute Value Acute-Chronic
(ug/1)**** Ratio***
3 (con't.) Chinook salmon,
Oncorhynchus tshawytscha
2 26.54 Channel catfish,
Ictalurus punctatus
1 4.355 Common carp,
CyprinuB carpio
25.85
26.54
4.355
* Ranked from most resistant to most sensitive based on Genus Mean Acute
Value.
** Value not used in calculation of the final acute-chronic ratio
*** Final acute-chronic ratio (ACR) includes a sheepshead minnow
(Cyprinodon variegatue) ACR of 6.873
**** values are adjusted to a pH of 6.5
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aqucitic Life Criteria for Phenol
CAS # 108-95-2
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative Procedure lor Deriving Aquatic Life Criteria (GLWQI,
1991}.
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for phenol (EPA, 1980)
led to the addition of new acute and chronic toxicity data to
the EPA data base. The combined acute and chronic data bases are
found in Tables 1 and 3, respectively. These data were used as
the basis for the Tier 1 criteria calculations.
Criterion Maximum Concentration (CMC)
Sufficient acute toxicity data were available to calculate a
tier 1 acute criterion for phenol. The Final Acute Value (FAV)
was calculated using the lowest 4 Genus Mean Acute Values
(Table 2). This analysis resulted in an FAV of 7031 ug/1. No
adjustment of this va Lue was necessary to protect for
commercially or recreationally important species of the Great
Lakes basin.
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 3500 ug/1.
Criterion Continuous Concentration (CCC)
Three acute to chronic ratios (ACR) were available for phenol
(Table 2). Because there is a trend in the data (the ACRs
increase with decreasing SMAVs), the ACR for the species whose
SMAV is closest to the FAV was used as the Final ACR. The FACR
is the rainbow trout ACR of 63.6.
The Final Chronic Value (FCV) was calculated by dividing the FAV
by the FACR of 63.6, resulting in a final chronic value of 110
ug/1. The CCC is also 110 ug/1. No adjustment of the FAV was
necessary to protect commercially or recreationally important
species of the Great Lakes basin.
Summary
The phenol criteria are:
GLWQI EPA
FAV = 7031 ug/1 not determined
CMC - 3500 ug/1
FCV = no ug/1 not determined
CCC - 110 ug/1
-------
Table 1. Acute Toxicity of Phenol to Aquatic Organisms.
Species
Cladoceran,
Daphnia magna
n
n
tt
ft
tt
ti
H
tt
ft
N
H
n
n
H
H
"
»
It
Cladoceran,
Daphnia pulex
n
ft
ft
ft
Method*
S,U
II
M
ft
n
tt
ft
ft
M
ft
S,M
n
n
n
S,M
M
It
It
n
s,u
S,M
M
tt
ft
tt
EC50/LC50
mg/1
14.5
13.3
11.2
21.3
8.6
11.5
6.6
30.1
12
19.8
23
7.7
7
21
100
92
91
88
91.2
28
93
87.8
85
81
79
Refer
Cowgill et al,
tt
n
n
tt
n
Keen fi Baillod,
Parkhurat et al
LeBlanc, 1980
Milleman et al,
Herroene et al,
Lewis, 1983
EPA, 1980
"
ft
n
ti
ti
it
n
n
it
tt
«t
tt
ence
1985
1985
. , 1979
1984
1984
Cladoceran,
Daphnia pulicaria
Cladoceran,
Cenodaphnia dubia/affinia
Midge,
Chironomua ripanus
Midge,
Chironomua tentana
Midge,
chironomus pinguis
Midge ,
Endochironomua natchitochaea
F,M
S,U
If
N
H
H
n
F,M
S,U
H
> 109
13 2
7.8
17.2
4.47
3.81
4.65
500
187.1
105
80.5
69.8
DeGraeve et al., 1980
Cowgill et al. 1985
Green at al, 1985
Franco et al. 1984
Milleman et al, 1984
Franco et al, 1984
Midge,
Tanytarsus neopunctipennis
70
-------
Table 1. cont. Acute Toxi.cj.ty of Phenol to Aquatic Organisms.
specxes
Snail,
Physa heterostropha
Snail,
Aplexa hypnorum
Mayfly,
Baetis rhodani
Planar ia,
Polycelis tenius
Caddisfly,
Hydropsyche angustipennia
Tubificid worm,
Lunnodrilus hoffmeisten
Guppy,
Poecilia reticulata
Bluegill,
Lepomia macrochirus
if
n
H
n
H
•*
Fathead minnow,
Pimephales promelaa
n
ft
n
n
it
Goldfish,
Carassius auratue
Mosquito fish,
Garabusia affinis
n
Channel catfish,
Ictalurus punctatus
Rainbow trout,
Oncorhynchus mykiss
ft
«t
Ml
Method*
S,U
P,M
F,M
It
H
M
s,u
S,M
s,u
tt
1*
u
M
s,u
R,M
s,u
s,u
P,M
fl
N
N
S,U
S,M
S,U
s,u
s,u
P,M
F,M
R,U
EC50/LC50
mg/1
94
>51.1
15.5
88
260
780
39.2
31
13.5
13.5
20
23.8
24
11.5
19.3
32
34.27
67.5
32.4
49.7
25
44.5
26
56
16.7
9.7
8.9
11.6
5
Reference
EPA, 1980
Holcombe et al. 1987
Green et al., 1985
N
M
n
EPA, 1980
It
M
H
n
n
Pickering & Henderson, 1966
EPA, 1980
«f
EPA, 1980
EPA, 1980
n
Geiger et al., 1985
Geiger et al., 1985
Degraeve et al., 1980
EPA, 1980
tf
Hann et al., 1977
EPA, 1980
Hodson, 1984
EPA, 1980
It
1i
* S • static, F = flow through, U * unmeasured, M * measured
/30
-------
Table 2.
Rank*
Ranked Genu* Mean Acute Value* with Species Mean Acute-Chronic Ratio*
for Phenol.
Genu* Mean
Acute Value
(mg/1) Specie*
Specie* Mean Specie* Mean
Acute Value Acute-Chronic
(mg/1) Ratio
18
17
16
15
14
13
12
11
10
9
8
7
6
780
260
177.9
94
88
70
69.8
55.47
51.1
44.5
40.6
38.2
34.9
Tubificid worm
Lunnodrilu* hoffmei*ten
Caddis fly
Hydropayche angustipennis
Midge
Chironomu* tentan*
Midge,
Chironomu* ripariu*
Midge,
Chironomu* pingui*
Snail,
Phyea heterostropha
Flatworm,
Polyceli* teniu*
Midge,
Tanytarsu* neopunctipennis
Midge,
Endochironomue natchitochaea
Cladoceran,
Daphnia magna
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulicaria
snail ,
Aplexa hypnorum
Goldfish,
Caraaaiu* auratu*
Fathead minnow,
Pimephale* promela*
Mosquito fish,
Gambuaia affini*
Guppy,
780
260
140.2
500
80.5
94
88
70
69.8
22.18
70.6
109
51.1
44.5
40.6
38.2
34.9
—
-
-
-
-
-
-
-
-
3.9**
-
-
-
-
49.3**
-
—
Poecilia reticulata
-------
Table 2. cont.
Rank*
Genus Mean
Acute Value
(mg/1) Species
Species Mean Specie* Mean
Acute Value Acute-Chronic
(mg/1) Ratio
5 17.3 BJuegill, 17.3
Lctpomis macrochirus
4 16.7 Channel catfish, 16.7
Ictalurus punctatus
3 IS. 5 Mayfly, 15.5
Baetis rhodani
2 10.16 Rainbow trout, 10.16
Oncorhynchua mykiss
1 7.2 CLadoceran, 7.2
Ceriodaphnia dubia/affinia
* Ranked from most resistant to least resistant based on GMAVs
** value not used in development of the acute to chronic ratio
63.6
-------
Table 3. Chronic Toxicity of Phenol to Aquatic Organisms.
Speclet
Cladoceran,
Daphnia magna
Fathead minnow,
Funephales promelas
Fathead minnow,
Pimephale* promelae
Test
Type
L-C
E-L
E-L
Chronic Value
mg/1
Reference
3.07 LeBlanc & Suprenant, 1980
1.37 DeGraeve et al., 1980
2.56 Hoicombe et al., 1982
-------
REFERENCES
Cowgill, P.W., I.T. Takahashi and S.L. Applegath. 1985. A comparison
of the effect of 4 benchmark chemicals on Daphnia magna and Ceriodaphni
dubia/affinis tested at two different temperatures. Env. Tox. Chem.
4:415-422
Degraeve, B.M., R.L. Overcast and H.c. Bergman. 1980. A comparison of
underground coal gasification condenser water and selected constituents
to aqutic biota. Arch. Environ. Contam. Toxicol. 9:543-555
EPA, 1980. Ambient Water Quality Criteria for Phenol. EPA 440/5-80-066.
October 1980.
Franco, P.J., K.L. Daniels, and R.M. Cushman et al. 1984. Acute toxicity
of synthetic oil, aniline and phenol to laboratory and natural populati
of chironomid (Diptera) larvae. Env. Poll. (A) 34:321-331.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for Deriving
Aquatic Life Criteria. April, 1991
Geiger. D.L., L.E. Northcott and D.T. Call, et al. 1985. Acute toxicities
of organic chemicals to fathead minnows (Pimephales promelas). Vol. Ill
Center for Lake Superior Environmental Studies, Univ. Wis.-Supl pp326
Green, D.W.J., K.A. Williams and D. Pascoe. 1985. Studies on the acute
toxicity of pollutants to freshwater macroinvertebrates. II. Phenol.
Arch. Hydrobiol. 103(1):75-82
Hann, R.W., Jr., et al. 1977. Water quality characteristics of hazardous
materials. PB285946. Texas A. and M. Univ., Colletge Station, Environme
Engineering Div.
Hermens, J., H. Camton and N. Steygor et al. 1984. Joint effect of a mixt
of 14 chemicals on mortality and inhibition of reproduction of Daphnia
magna. Aq. Tox. 5:315-322
Hodson, P.V., D.G. Dickson and K.L.E. Kaiser. 1984. Measurement of median
lethal dose as a rapid indication of contaminant toxicity to fish. Env.
Chem. 3:243-254.
Holcombe. G.W., G.L. Phipps, A.H. Sulaimen and A.D. Hoffman. 1987.
Simultaneous multiple species testing: Acute toxicity of 12 diverse
freshwater amphibians, fish and invertebrate families. Arch. Environ. C
Toxicol. 16:697-710
Holcombe, G.W., G.L. Phipps and J.T. Fiandt. 1982. Effects of phenol,
2,4-dimethylphenol, 2.4-dichlorophenol, and pentachlorophenol on embryo
larval, and early-juvenile fathead minnows (Pimephales promelas). Arch.
Environ. Contam. Toxicol. 11:73-78
Keen, R. and C.R. Baillod. 1985. Toxicity to Daphnia of the end products
of wet oxidation of phenol and substituted phenols. Wat. Res. 19(6):76
LeBlanc, G.E. and D.C. Suprenant. 1980. The chronic toxicity of 8 of the
priority pollutants to the water flea (Daphnia magna) Eg. and G. Bionom
draft manuscript.
-------
anc, G.E. 1980. Acute toxicity of priority pollutants to water flea
aphnia magna) Bull. Environ. Contain. Toxicol. 24:684-691.
Milleman, R.E., W.J. Birge and J.A. Black, et al. 1984. Trans. Am. Fish.
Soc. 115:74-85
Pickering, A.H. and C. Henderson. 1966. Acute toxicity of some important
petrochemicals to fish. J. Wat. Poll. Cont. Fed. 38(9):1419-1429.
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Selenium
Class 023
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for Selenium (EPA,
1987) led to the addition of new acute toxicity data for
selenium IV and VI to the EPA data base. The additional data
are found in Table 1. The combined data were used as the basis
for the Tier l criteria calculations.
Individual criteria are presented in this document for selenium
IV and VI, and for total selenium. The approach used by EPA
(EPA 1987), which is to regulate selenium as total selenium, is
the preferred approach for selenium regulation.
Selenium VI:
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (GMAV) (Table 2). This analysis
resulted in an FAV of 25.06 ug/1. No adjustment of this value
was necessary to protect for commercially or recreationally
important species of the Great Lakes basin.
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 13 ug/1.
Criterion Continuous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate
a Final Chronic Value (FCV) using the eight family approach of
the procedure. Therefore, the FCV was calculated by dividing
the calculated FAV by the EPA derived final acute to chronic
ratio of 2.651 (Table 2), resulting in a FCV of 9.453 ug/1. No
adjustment of this value was necessary to protect commercially
or recreationally important species of the Great Lakes basin.
The Selenium(VI) criterion are:
GLWQI EPA
FAV = 25.06 ug/1 FAV = 25.65 ug/1
CMC = 13 ug/1 CMC - 12.82 ug/1
FCV - 9.453 ug/1 FCV - 9.676 ug/1
CCC =9.5 ug/1
Selenium IV:
Criterion Maximum Concentration
The FAV was calculated using the lowest 4 Genus Mean Acute
Values (GMAV) (Table 3). This analysis resulted in an FAV of
371.8 ug/1. No adjustment of this value was necessary to
-------
protect for commercially or recreationally important species of
the Great Lakes basin.
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 190 ug/1.
Criterion Continuous Concentration
Insufficient chronic toxicity data were available to calculate
a FCV using the eight family approach of the procedure.
Therefore, the FCV was calculated by dividing the calculated
FAV by the EPA derived final acute to chronic ratio of 8.314
(Table 3), resulting in a FCV of 44.72 ug/1. This value was
lowered to 27.6 ug/1 to protect the commercially and
recreationally important rainbow trout (justification for this
value is given in EPA, 1987).
The selenium IV criteria are:
GLWQI EPA
FAV = 371.8 ug/1 FAV = 371.8 ug/1
CMC = 190 ug/1 CMC = 190 ug/1
FCV =27.6 ug/1 FCV =27.6 ug/1
CCC = 28 ug/1
Total selenium:
Field studies conducted on Belews Lake in North Carolina
(described in EPA, 1987) suggested that selenium might be more
toxic to certain species of freshwater fish than has been
observed in chronic toxicity tests. Based upon these studies,
the freshwater CCC for total selenium was determined to be 5
ug/1. Assuming the final acute-chronic ratio to be 7.993
(geometric mean of all acute to chronic ratios between 2.651
and 16.26), the FAV is 39.96 ug/1. The CMC is 20 ug/1.
Summary
The total selenium criteria are:
GLWQI EPA
FAV = 39.96 ug/1 FAV = 39.96 ug/1
CMC = 20 ug/1 CMC = 19.98 ug/1
CCC = 5 ug/1 CCC = 5 ug/1
References
EPA, 1987. Ambient Aquatic Life Water Quality Criteria for
Selenium. EPA 440/5-87-006. September 1987.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure
for Deriving Aquatic Life Criteria. April, 1991
Johnston, P.A. 1987. Acute toxicity of inorganic selenium to
Daphnia magna (Straus) and the effect of sub-acute exposure
upon growth and reproduction. Aquatic Toxicol. 10:335-352.
-------
Table 1. Aquatic Acute Toxicity Data Added Subsequent to Publication of
the Ambient Water Quality Criteria for Selenium IV and VI (EPA, 1987)
LC/EC50
Species Method Chemical (ug/1) Reference
Cladoceran, S,U Na-selenite 680 Johnston, 1987
Oaphnia magna (Selenium IV)
Cladoceran, S,U Na-aelenate 750 Johnston, 1987
Oaphnia magna, (Selenium VI)
-------
able 2. Ranked Genus Mean Acute Values with Species Mean Acute-Chronic Ratios
for Selenium(VI).
Genus Mean
Acute Value
Rank* (ug/1)
Species
Species Mean
Acute Value
(ug/1)
Species Mean
Acute-Chronic
Ratio
11
10
442000 Leech, 442000
Hephelopsis obscura
193000 Snail, 193000
Aplexa hypnorum
66000 Channel catfish, 66000
Ictalurus punctatus
63000 Bluegill, 63000
Lepomis macrochirus
47000 Rainbow trout, 47000
Oncorhynchus mykiss
20000 Midge, 20000
Paratanytarsus parthenogeneticus
7300 Hydra, 7300
Hydra ap.
5500 Fathead minnow, 5500
Pimephales promelas
760 Amphipod, 760
Hyalella azteca
549.8 Cladoceran, 1229
Daphnia magna
Cladoceran, 246
Daphnia pulicaria
65.38 Amphipod, 65.38
Gaznmurus peeudolimnaeus
16 26
9.726
2.651**
* Ranked from most resistant to most sensitive based on Genus Mean Acute Value.
** Since the species mean acute-chronic ratio increases as the SMAVs
increase, the acute-chronic ratio for the Daphnia magna (whose SMAV is
closest to the FAV) was used as the FACR.
-------
Ill
Table 3. Ranked Genua Mean Acute Values with Species Mean Acute-Chronic
Ratios for Selenium IV.
Rank*
Genua Mean
Acute Value
(ug/1) Species
Species Mean species Mean
Acute Value Acute-Chronic
(ug/1) Ratio**
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
203000
42500
35000
34910
30176
28500
26100
25934
24100
13600
12600
11700
10490
10200
6500
2704
1783
Let-ch,
Nephelopsis obscure
Midge,
Tanytarsus diasunilis
Common carp,
Cyprinua carpio
Snail,
Apiexa hypnorum
White sucker.
Catostomus commersoni
Bluegill,
Lepomia macrochirus
Goldfish,
Carassius auratus
Midge,
Chtronomus plumosus
Snail,
Phfsa sp.
Channel catfish,
IctaluruB punctatus
Mosguitof ish,
Gajibuaia affinia
Yellow Perch,
Perca flavescens
Rainbow Trout,
Oncorhynchus mykiss
Brook trout,
Salvelinus fontinalis
Flagfish,
Jordanella floridae
Airiphipod,
Gammarus pseudolimnaeus
striped bass,
203000
42500
35000
34910
30176
28500
26100
25934
24100
13600
12600
11700
10490 141.5***
10200
6500
2704
1783
Morone saxatxlis
-------
Table 3. Continued.
Genus Mean
Acute Value
Rank* (ug/1) Species
Species Mean Species Mean
Acute Value Acute-Chronic
(ug/1) Ratio
1796 Cladoceran,
Daphnia magna
Cladoceran,
Daphnia pulex
1700 Hydra,
Hydra ap.
1601 Fathead minnow,
Pimephales promelas
<603.6 Cladoceran,
Ceriodaphnia affinis
340 Amphipod,
Hyalella azteca
834
3870
1700
1601
<603.6
340
13.31
5.586
6.881
* Ranked from most resistant to most sensitive based on Genus Mean Acute Value.
** Calculation of the final acute-chronic ratio included, in addition
to the above data, a mysid (Mysidopsis bahia) ACR of 7.085 and
a sheepshead minnow (Cyprinidon variegatus) ACR of 10.96
*** Not use in calculation of the acute-chronic ratio
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier l Aquatic Life Criteria for Silver
Class 024
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Water Quality
Initiative Procedure for Deriving Aquatic Life Criteria (GLWQI,
1990).
A search of aquatic toxicity literature generated subsequent to
the draft publication of the EPA criteria document for silver
(EPA, 1987) produced no new data to supplement the EPA data
base. Therefore, the EPA data base was used as the basis for
the Tier 1 criteria calculations and the resulting criteria are
the same as in the 1987 criteria document.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute Values (GMAV) in the EPA data base (Table 1).
This analysis resulted in an FAV of 1.833 ug/1. No adjustment
of this value was necessary to protect for commercially or
recreationally important species of the Great Lakes basin.
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 0.9 ug/1.
Criterion continuous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate
a Final Chronic Value (FCV) using the eight family approach of
the procedure. Therefore, the FCV was calculated by dividing
the calculated FAV by the EPA derived final acute to chronic
ratio of 15.7 (Table 1), resulting in a FCV of 0.1168 ug/1. No
adjustment of this value was necessary to protect commercially
or recreationally important species of the Great Lakes basin.
Summary
The silver criteria are:
GLWQI EPA
FAV - 1.833 ug/1 FAV - 1.833 ug/1
CMC - 0.9 ug/1 CMC = 0.9165 ug/1
FCV - 0.1168 ug/1 FCV = 0.1168 ug/1
CCC =0.12 ug/1
References
EPA, 1987. Ambient Aquatic Life Water Quality Criteria for
Silver. DRAFT. September 24, 1987.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for
Deriving Aquatic Life Criteria. April, 1991.
Af
-------
able 1. Ranked Freshwater Genus Mean Acute Values with Species Mean Acute-Chronic
ratios for Silver.
Genus Mean
Acute Value
Rank* (ug/1)
Species
Species Mean Species Mean
Acute Value Acute-Chronic
Ratio***
18
17
16
IS
14
13
12
11
10
9
8
7
6
5
4
3
2
1
560
420
S3
29
26
17.3
15
13 38
13
11.34
9.2
8.49
8.163
5
4.5
3.924
2 2
2.155
Crayfish,
Orconectes uraaunis
Midge,
Tanytarsus dissunilis
Snail,
Aplexa hypnorum
Leach ,
Nephelopsis obscura
Hydra,
Hydra sp.
Channel catfish,
Ictalurua punctatus
Cladoceran,
Sunocephalus vetulus
Rainbow trout,
Oncorhynchus roykiss
Bluegill,
LepomiB raacrochirus
Fathead minnow,
Punephales promelas
Flagfish,
Jordanella floridae
Mottled sculpin,
Cottus bairdi
Speckled dace,
Rhinichthys osculus
Amphipod,
Crangonyx pseudogracilis
Amphipod ,
Gammarus pseudolimnaeus
Cladoceran,
Ceriodaphnia reticulata
Mayfly,
Leptophlebia sp.
Cladoceran,
Daphnia pulex
Cladoceran,
560
420
83
29
26
17.3
15
13.38
13
11.34
9.2
8.49
8.163
5
4 5
3.924
2.2
5.158
0.9
«•>
_
_
_
_
_
_
33.29
_
13.66
_
—
_
-
-
_
-
_
0.4994**
Daphnia magna
* Ranked from most resistant to most sensitive based on Genus Mean Acute Value.
* Value not used in calculation of the final Acute-Chronic Ratio
** Calculation of the Final Acute-Chronic Ratio included, in addition to the above
data, a Mysid (Mysidopais bahia) ACR of 8.512
-------
GREAT LAKES WATER QUALITY INITIATIVE
Tier 1 Aquatic Life Criteria for Zinc
Class 027
July, 1991
Introduction
The criteria presented in this document were developed pursuant
to the Tier 1 approach of the Great Lakes Hater Quality
Initiative Procedure of Deriving Aquatic Life Criteria (GLWQI,
1991).
A search of aquatic toxicity literature generated subsequent to
publication of the EPA criteria document for zinc (EPA, 1987)
led to the addition of new acute toxicity data to the EPA data
base. These data are found in Table 1. The combined data were
used as the basis for the Tier 1 criteria calculations. The
toxicity of zinc is hardness dependent. Therefore, all data
used in the criteria calculations were normalized to a hardness
of 50 mg/1.
Criterion Maximum Concentration (CMC)
The Final Acute Value (FAV) was calculated using the lowest 4
Genus Mean Acute values (GMAV) (Table 2). This analysis
resulted in an FAV of 152.7 ug/1 at a hardness of 50 mg/1. The
EPA pooled slope of 0.8473 did not change by the addition of
new data. No adjustment of the FAV was necessary to protect
for commercially or recreationally important species of the
Great Lakes basin.
The Final Acute Equation =
0.8473(In hardness) + 1.577
e
The CMC was calculated by dividing the FAV by 2, resulting in a
CMC of 67 ug/1.
Criterion Continuous Concentration (CCC)
Insufficient chronic toxicity data were available to calculate
a Final Chronic Value (FCV) using the eight family approach of
the procedure. Therefore, the FCV was calculated by dividing
the calculated FAV by the EPA derived final acute to chronic
ratio of 2.208 (Table 2). This results in a FCV of 60.33 ug/1
at a hardness of 50 mg/1. The CCC equals 60 ug/1. No adjustment
of this value was necessary to protect commercially or
recreationally important species of the Great Lakes basin. The
Final Chronic Equation =
0.8473 (In hardness) + 0.785
e
Summary
The zinc criterion are:
GLWQI
FAV = 119.4 ug/1 (at a hardness of 50 mg/1).
CHC = 67 ug/1 ( at a hardness of 50 mg/1)
FCV = 60.33 ug/1 (at a hardness of 50 mg/1)
CCC = 60 ug/1 ( at a hardness of 50 mg/1)
-------
0.8473(In hardness) + 1.577
Final Acute Equation - e
0.8473 (In hardness) + 0.785
Final Chronic Equation « e
EPA
FAV = 130.1 ug/1 (at a hardness of 50 mg/1)
CMC - 65.05 ugl (at a hardness of 50 mg/1)
FCV =58.9 ug/1 (at a hardness of 50 mg/1)
0.8473 (In hardness) +1.5537
Final Acute Equation = e
0.8473 (In hardness) + 0.7614
Final Chronic Equation » e
References
Berglind, R. and G. Dave, 1984. Acute toxicity of chromate,
DDT, PCP, TPBS, and zinc to Daphnia magna cultured in hard and
soft water. Bull. Environ. Contain. Toxicol. 33:63-68.
Dawson, D.A., E.F. Stebler, S.L. Burks and J.A. Bautle.
1988. Evaluation of the Developmental Toxicity of Metal-
Contaminated Sediments Using Short-Term Fathead Minnow and
Frog Embryo-Larval Assays. Environ. Toxicol. Chem. 7:27-34.
EPA, 1987. Ambient Aquatic Life Water Quality Criteria for
Zinc. EPA 440/5-87-003. February 1987.
GLWQI, 1991. Great Lakes Water Quality Initiative Procedure for
Deriving Aquatic Life Criteria. April, 1991.
-------
Table 1. Aquatic Acute Toxicity Data Added Subsequent to Publication of the Ambient Water Quality
Criteria for Zinc (EPA, 1967).
Hardness Adjusted
(mg/1 as LC/ECSO LC/ECSO
Species Method CaCO3) chemical (ug/1) (ug/1)* Reference
Frog, S,M 100 Zinc sulfate 34500 19149 Daweon et al., 1968
Xenopus laevis
Cladoceran, S,U 300 Zinc eulfate 1100 241 Berglind & Dave, 1984
Daphnia magna
* Adjusted to a hardness of 50 mg/1.
-------
Table 2. Ranked Genus Mean Acute Values with Species Mean Acute-Chronic
Ratios for Zinc.
Genus Mean
Acute Value
Rank* (ug/1)** Species
Species Mean Species Mean
Acute Value Acute-Chronic
(ug/1)** Ratio****
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
88960
19800
19149
18400
17940
16820
13630
10560
10250
9712
8157
8100
7233
6580
6053
Damsel fly,
Argia sp.
Amphipod ,
Crangonyx pseudogracilis
Frog,
Xenopus laevis
Worm,
Nais up.
Banded killifish,
Fundulus diaphanus
Snail,
Amnicola sp.
American eel.
Anguilla rostrata
Pumpkinseed,
Lepomis gibbosus
Bluegill,
Lepomis macrochirus
Goldfish,
Carassius auratus
Worm,
Lumbriculus variegatus
Isopod,
Asellus bicrenata
Isopod,
Asellus communis
Amphipod,
Gamroarus sp.
Common carp,
Cyprinus carpio
Northern squawfish,
Ptychocheliua oregonesis
ouppy,
88960
19800
19149
18400
17940
16820
13630
18790
5937
10250
9712
5731
11610
8100
7233
6580
6053
_
_
_
_
_
—
—
_
ta
_
_
_
_
-
-
Poecilia reticulata
-------
Table 2. Continued.
Hank*
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
Genue Mean
Acute Value
(ug/1)**
6000
5228
4900
4341
3830
3265
2176
2100
1707
1672
1607
1578
1353
1307
>1264
Specie*
Golden ahiner,
NotemigonuB crycoleucua
White sucker,
Catoetomus commersoni
Asiatic clam,
Corbicula fluminea
Southern platyfiah,
XiphophoruB maculatus
Fathead minnow,
PimephaleB promelaa
iBopod,
Lirceua alabamao
Atlantic salmon,
Salnto salar
Brook trout,
Salvelinus fontinalia
Bryozoan,
Lcphopodella carteri
Flagfiah,
Jordanella floridae
Bryozoan,
Plumateila rostrata
Snail,
Heliaoma campanulatum
Snail,
Physa gyrina
Snail ,
Physa heterostropha
Bzyozoan,
Pectinetella magnifica
Tubificid worm,
Species Mean
Acute Value
(ug/1)**
6000
5228
4900
4341
3830
3265
2176
2100
1707
1672
1607
1578
1683
1088
1307
>1264
Species Mean
Acute-Chronic
Ratio****
^
_
_
_
5.644***
_
_
2.335***
_
41.2***
_
_
_
_
—
-
Limnodrilua hoffmeisteri
931 Rainbow trout,
Oncorhynchua mykias
689.3
1.S54
-------
Table 2. Continued.
Rank*
5
4
3
2
Genus Mean
Acute Value
(ug/1)** Specie*
Coho salmon,
Oncorhynchua kieutch
Sockeye ealraon,
Oncorhynchua nerka
Chinook salmon,
OncorhynchuB tshawytscha
790 Mozambique tilapia
Tilapia mossanbica
299.8 Cladoceran,
Daphnia magna
Cladoceran,
Daphnia pulex
227. 8 Longfin dace,
Agosia chryaogaster
119.4 striped bass,
Species Mean
Acute Value
-------
PLEASE NOTE THE PROPOSED METHODOLOGY FOR DERIVING TIER II AQUATIC LIFE VALUES
WAS SUBSTANTIALLY REVISED BY THE TECHNICAL WORKGROUP AT THEIR AUGUST 25 - 26, 1991
MEETING THEREFORE, REVISED TIER II CRITERIA CALCULATIONS ARE NOT AVAILABLE AT THIS
POINT THE ENCLOSED DESCRIBES THE BASIS FOR THE NEW TIER II APPROACH IN GENERAL TERMS,
IT IS PROPOSED THAT THE LOWEST AVAILABLE GMAV WILL BE DIVIDED BY A FACTOR WHICH IS
A FUNCTION OF THE NUMBER OF AVAILABLE ACUTE VALUES SATISFYING THE MINIMUM DATA REQUIRE-
MENTS OF THE TIER I METHODOLOGY. THESE FACTORS ARE LISTED BELOW
SAMPLE SIZE FACTOR
1 (must be Daphmd) 20 5
2 13.2
3 86
4 6.5
5 50
6 4.0
7 3.6
UNDER THIS NEW METHODOLOGY, A DEFAULT ACUTE TO CHRONIC RATIO OF 18 IS ALSO PROPOSED
750
-------
MICHIGAN DEPARTMENT OF NATURAL RESOURCES
INTEROFFICE COMMUNICATION
September 5, 1991
To: Joan Karnauskus, EPA Region V
From: Brenda Sayles, MDNR
Subject: Aquatic Life Tier 2 Technical Support Document
As decided during the August 22-23, 1991 GLI Technical
Workgroup meeting, the EPA draft document (1/4/91) "Analyses
of Acute and Chronic Data for Aquatic Life", with some
modifications, will serve as the basis for the aquatic life
Tier 2 Technical Support Document.
These modifications include:
1) Using only the Final Acute Value Factors which included
in their derivation a Daphnid
2) Using the 80% Final Acute Value Factors (provided in an
August 23, 1991 letter from Charles Stephan to Jim Grant
- Attachment A), and;
3) Using the 80% freshwater estimate acute-chronic ratio of
18 as the default acute to chronic ratio (provided in a
August 21, 1991 letter from Charles Stephan to Jim Grant
- Attachment B).
cc. Jim Grant/MDNR
at
1030 *....
* I
-------
flUG-23-1991 13144 FROM EPA-fiRL-DULUTH TO £r * &35B P.02/07
f UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
ENVIRONMENTAL Ri$£AflCM LABORATORY - DYUTTH
«Z01 CONOOON BOULEVARD
OULUTH. MINNESOTA 5MQ4
August 23, 1991
Mr. Jin Grant
Michigan Department of Natural Resources
Xnapps center
P. O. Box 30028
Lansing, MI 48909
Dear Jin:
I asked George Host to generate additional Final Acute Value
Factors (FAVFs) to supplement those given in Table 18 on page 62
of the 1-4-91 draft of the report titled ANALYSES OF ACUTE AND
gin^9flTC nATA yos *OPATIC LIFE. The 95th percentile FAVFs that I
have been using are in the group titled "Overall 95th Percentile"
at the bottom of Table 18. The 50th percentile FAVFs that I have
been using are in the group titled "Overall Median" that is
second from the bottom in the table. In each group I used the
FAVFs for the cases labelled "Daphnid required" and "Daphnid
excluded".
George reanalyzed the data to determine FAVFs that
correspond to selected percentilea between the 50th and the 95th
for all five cases; the results are presented in the attached
three pages. From these results I extracted the FAVFs for the
two cases labelled "Daphnid required" and "Oaphnid excluded" and
created the attached Tables A and B. (Note that the order of the
cases within a group is different in George's new data than in
Table 18.) The "sample size" also is the number of minimum data
requirements that are satisfied because each acute value in the
sample was selected to satisfy a different minimum data
requirement. The FAVFs given in Tables A and B for the 50th and
95th percentiles agree with those given in Table 18 of the
report.
As expected, the FAVF increases slowly with an increase in
the low percentiles, but increases rapidly with an increase in
the high percentiles. This is not the result of a statistical
assumption concerning the distribution of the data; this is a
direct result of the data that were in the data sets that were
used in this analysis. These FAVFs were determined by randomly
sampling actual data sets, calculating the ratio of the FAV to
the lowest value in the sample, and then empirically determining
the percentiles. These are not hypothetical examples that might
never occur in the real world; each of these results corresponds
to real data. At the bottoms of Tables A and B I noted the /i
IS'
-------
13-44 FROM EPfi-ERL-DULUTH TO 85173739958 P 33/07
SablE ft «S^^iS«.J°Vhi?h the »«•*•<»•• ™*» given in
rSi^L?,,! n«ii of th« fraft r«Port were greater than the
re»p«ctiv« 95th percentile FAVTs. Although the 95th percentile
8 9 **** lar**' ^•y are consistent with the available
data
uJ sanP1« »iz« • « would n«v«r be us«d
the 8 minimum data requir«n«ntB are satimfied tht
LSSH^ ^^^^ u»^g the procedure specified in^he Tier
procedure. These values do indicate, however, that about 75
ar* Within a factor of 3 of thTlow^S acute
t *llow'd 8ampU 8i" of 8 f and about 5
*re *°" ^^ 7 tlMS lower «"» ^^ lowest
aCUt* V5iUGS: Thft Differences between the factor, for
d required" and "Daphnid excluded" for sample site - 8 are
. samPlln^ variation because there are no substantial
differences between these two cases when all eight minimum data
requirements are satisfied.
^ a major consideration in selecting the
5° ** U8e
-------
«JG-23-1991 13=45 FROM EPft-£RL-~_ . TO 85173739958
50
55
60
65
70
75
80
85
90
95
TABLE A: FAVF6 WITH DAPHNID REQUIRED
SAMPLE ST2K
1
4.9
6.4
7.8
10.1
13.8
15.7
20.5
31.6
51.0
93.5
2
2
3.2
3.7
4.5
5.7
7.2
9.0
13.2
18.6
30.1
57.8
2
3
2.6
2.9
3.4
3*9
5.0
6.5
8.6
13.1
22.1
50.5
2
4
2.4
2.6
2.8
3.3
3.9
4.9
6.5
9.1
16.3
41.8
3
5
2.2
2.4
2.6
2.8
3.3
3.9
5.0
7.0
12.4
31.0
5
6
2.0
2.2
2.4
2.6 <
2.9
3.4
4.0
5.6
8.6
22.0
7
7
1.9
2.1
2.3
2.4
2.7
3.0
3.6
4.3
6.2
13.1
10
8
1.9
2.0
2.2
2.3
2.5
2.7
3.2
3.7
4.9
7.3
12
TABLE Bs FAVFs KITH DAPHNID EXCLUDED*
1
18.0
22.6
34.8
52.2
80.0
134.
242.
739.
262.
'000.
2
7
9
12
17
24
41
64
116
239
1680
.7
.8
.6
.5
.8
.4
.8
«
•
•
3
4.
6.
7.
10.
13.
20.
36.
59.
116.
424.
8
1
8
1
3
1
2
2
4
3.
4.
5.
7.
9.
12.
20.
37.
68.
176.
7
5
7
1.
3'
8
1
5
8
5
3.
3.
4.
5.
6.
8.
12.
21.
41.
111.
1
7
4
5
8
9
9
4
0
6
2.6
3.1
3.7
4.4
5.6
6.9
9.2
14.2
25.9
64*2
7
2.4
2.6
3.1
3.7
4.6
5.8
7.2
10.1
17.4
39.9
8
1.9
2.0
2.1
2.3
2.5
2.7
3.2
3.7
4.9
7.5
50
55
60
65
70
75
80
85
90
95
Me 5 8 9 8 10 9 8 12
K * number of chemicals (out of 31) for which the worst-case
FAVF in Table 11 in the draft report is greater than the 95th
percentile FAVF.
Daphnid not excluded at sample sice - 8.
M = number of chemicals (out of 31) for which the worst-case
FAVF in Table 8 in the draft report is greater than the 95th
percentile FAVF.
-------
•*=««J(r-23»1991 13-45 FROM EPfl-ERL-DULUTH
TO
-05/07
^^^^MBMMMM^^MH
^ m^^^ ^ «J
•••^^•••••••^••^•H
^ -^— ^__
50th Pareantite
AnyFamfly
Daphrod Raqwrad
Both Daphnid ft Salmonld
Daphnid Exdudad
Any Family
Daphnid Raquvad
Both Daphnid & Salmomd
Daphnd Exokidad
Salmon* Raquirad
60th Pf rc»nta«
Any Family
Daphnid Raqurad
Both Daphnid & Salmonid
Daphnid Exdudad
Salmon* Raquirad
65th Parcanbla
Any Family
Daphnid Raquirad
Both Daphnid i Salmon*
Daphnid Exdudad
Salmon* Raquirad
mm^mm^m
10th through
PM^HHMH
^""""^^H
1
BaaaaiHai^iHNk^iH
1466
4.89
—
17.98
1083
19.18
6.37
_
22.60
13.12
26.85
777
_
3477
1798
4092
1011
~~
5215
2967
^^MMHHi
IStti Paraar
l^M^^^^
••H^^M
2
620
3.23
2.83
767
00
787
374
335
979
717
10^9
4.51
4.11
1261
9J7
1304
574
5.43
1748
13.04
mm^^^m
itHMtarP!
^•MMM^H
•^^H^H)
3
3.73
2.60
2.47
4.77
365
4.45
291
2.89
6.07
4.45
567
3.40
334
778
581
733
3.93
393
1007
7.27
•^^••^
IMM4-R
fta^fl^A^fe 4
4
2.81
2.36
2.28
388
270
3.34
2.57
2.53
4.45
320
387
283
292
5.65
389
4.68
331
338
708
495
MBMM^M
^^J^^—* — —
wwnwwttw
Mxa
5
2.40
2.18
2.16
3.05
2.37
^65
2,36
2JJ3
3,65
2.60
3.06
2.56
2-57
437
301
3.58
2.81
2.92
554
3J9
HBH>^MM
2.16
102
2.02
2.61
2.14
2J4
2.20
2.19
3.06
2.30
2.58
2.37
2.36
365
2.54
2.84
2.5*
2,59
4.38
2.83
•^MM
•iHHi^^B
7
196
192
1.91
2.36
1.94
2.14
2.08
2.09
2.63
2.13
2.25
Z25
2^2
309
2^5
2JO
2.42
2.40
3.68
246
•HMIW
••••••
8
186
186
185
186
197
1.97
197
1.96
1J7
215
2.15
216
2.13
214
2.25
2.26
225
2.25
2.25
-------
FROM EPPHERL-DULUTH
TO
B5173739958
SOth ht5thPwoantflaaforP1iaaa4-ftaatNMMr
SamptoSfea
Any Family
Daphmd Raqurad
Beth Daphnid I SrimonU
Daphnid Exdudad
Salmonid Raqwad
75th Parcantia
Any Family
Daphmd Raqu*ad
Both Daphrtd ft. Satanomd
Daphnid Exdudad
Salmonid Raqwrad
80th ParcanMa
Any Family
Oaphnid Raqurad
Both Daphnid & Salmond
Daphnid Exdudad
Safcnonid Raqurod
86th Pafoantla
Family
Daphnid Raqtwad
Both Daphnid 4 Salmonid
Daphnid Exdudad
Salmonid Raqunad
59.72
13.75
_
79.65
50.55
107.88
1668
—
13432
67.80
17429
2050
—
24235
136.12
442.60
31.81
—
73852
29137
1739
7.15
6.80
24.75
1734
2469
903
7.86
4137
2823
44.00
13.12
13.12
6432
5039
7526
1858
16.16
11572
84.76
9.96
497
5.15
1329
9.90
1329
6.47
664
20.09
13.12
2029
8.59
8.75
36.19
2050
38.97
13.12
13.12
5919
4055
6.18
3.87
397
932
659
854
4.90
5.09
12.77
885
1225
653
700
20.09
1145
1978
908
1069
37.48
2052
425
3.31
336
662
438
559
3.87
3.93'
8.90
598
762
5.04
524
12.86
812
^.
11.95
700
7.48
2137
1235
338
196
2.98
557
335
3.93
3.42
3.43
6.93
394
5.21
4.00
4.05
916
535
7.15
564
553
14.16
7.69
2.68
2.65
262
4.56
2.66
320
3.04
3.05
5.80
3.18
3.63
3.61
3.63
721
3.77
4.93
430
4.35
10.08
4.99
247
2.47
2.47
2.47
244
2.6S
266
2.65
2.
2.
3.17
3.17
3.20
317
3.13
376
374
374
370
3.69
-------
HJG-23-1991 13 47 FROM EPfl-ERL-DU_UTH
ro
85173739958 P 07/07
50th through Mth PwotntilM for Phut « • tm*tmur
90w1 PWOtAlM
Any Family
Daphnd fequirad
Both DapHnd 4 Satonomd
Dapftnd Exduted
Salmon* Rtqidrtd
95ttP«o«flia«
Any Family
Dtphnid Roquirtd
B«h Daphntd & Salmorad
Daphnid Exdud«d
Salmomd Rtqurtd
8«mp*tSd«
171254
S101
_
2262.17
2015 46
836656
93.52
_
10024.15
20210.61
160^2
30.11
2SJ2
236J6
175,68
100688
5778
5191
168004
185740
69.60
22.06
20JO
11572
7526
19000
5034
48.20
424,11
21620
39.44
1629
1823
68.79
42.49
9014
41.83
38.83
17561
9789
21.19
1226
12.61
41.02
21.59
53-59
31.02
29.00
111.17
5966
13.14
8.60
6.75
28.85
1306
3317
22.00
1990
6429
33.73
7.10
624
647
17.35
744
16.94
1312
1322
3990
1718
4,94
485
490
486
497
7.99
7.26
942
7.51
806
-------
/
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
iNVIHONMENTAl RESEARCH LABORATORY - OUU/TM
«01 CONOOON BOULEVARD
DULUTH, MWNISOTA $1904
August 21, 1991
Mr. Jin Grant
Michigan Department of Natural Resources
Knapps center
P. o. Box 30028
Lansing, MX 48909
Dear Jim:
I discussed Section 7 (pages 22 and 23) of the 1-4-91 draft
AQUATIC LIFE with Dr. Ron Regal since he performed the analyses
of the Final Acute-Chronic Ratios. The medians given on page 23
are 7.8 for fresh water and 4.0 for salt water. These were
determined by using the classic definition of sample medians the
middle value when sample size is odd and the average of the two
middle values when sample size is even.
The estimates of the 95th percentiles were obtained by using
a parametric method because of the "greater than" value for fresh
water that was near the sample 95th percentile. Also, parametric
methods are useful with small data sets because they "smooth11 the
data. The selected parametric method was used because it allowed
for both right and left censoring; it could handle the "greater
than" value and could be tailored to the estimation of a
percentile in the right tail. For fresh water the estimated mean
and standard deviation were 1.96 and 1.10., respectively, on the
natural log scale. For salt water the estimated mean and
standard deviation were 1.48 and 1.33, respectively, also on the
natural log scale. From these values the following estimates of
the percentiles are obtained:
»
Pere«nt:ila Freshwater Saltwater
Estimate Estimate
50 7.1 4.4
55 8.2 5.2
60 9*4 6.2
65 10.8 7.3
70 12.6 8.8
75 14.9 10.8
80 17.9 13.5
85 22.2 17.4
90 29.1 24.2
95 43.4 39.2
-------
96 48.7 45.1
97 56.2 53.6
97.5 61.3 59.5
98 68.0 67.5
These estimates of the 95th percentiles agree with the values
given on page 23 of the draft report. In addition, the value of
25 is between the 85th and 90th percentiles for fresh vater and
is slightly above the 90th percentile for salt vater. The
freshwater sample median of 7.8 is between the 50th and 55th
pereentiles for fresh water, whereas the saltwater sample median
of 4.0 is below the 50th percentile for salt water. The
percentile estimates for fresh water are similar to those for
salt water.
Sincerely,
Charles Stephen
cc: Joan KamausKus, U.S. EPA, Region V, Chicago
Ron Regal, University of Minnesota - Duluth
-------
FOR EXTERNAL REVIEH
DRAFT
1-4-91
ANALYSES OF ACUTE AND CHRONIC DATA
FOR AQUATIC LIFE
by
tnst
George E. Host
Natural Resources Research Institute
Duluth, MN 55811
Ronald R. Regal
University of Minnesota
Duluth, MN 55812
and
Charles E. Stephan
Environmental Research Laboratory - Duluth
6201 Congdon-Blvd.
Duluth, MN 55804
OFFICE OF ENVIRONMENTAL PROCESSES AND EFFECTS RESEARCH
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, DC 20460
-------
NOTICE
Th* information in this document has bean funded wholly or in
part by th« United States Environmental Protection Agency under
cooperative Agreement No. CR-814147-02-0 with the University of
Wisconsin - Superior and through Contract No. 68-03-3544 with AScl
Corporation. The extramural portion of this worJc was funded by the
criteria and Standards Division of the Office of Water and by
Region III in connection with a RARE (Regional Applied Research
Effort) project concerning pesticides. This document has been
subjected to the Agency's peer and administrative review, and it
has been approved for publication as an EPA document. Mention of
trade names or commercial products does not constitute endorsement
or recommendation for use.
11
-------
FOREWORD
The U.S. EP* is responsible for the protection of the
environment in which we live and work. An essential part of
exercising this responsibility is the research that allows the
Agency to formulate and implement actions that lead to a compatible
balance between human activities and preservation of the
environment.
The Environmental Research Laboratory - Duluth is the Agency' s
center of expertise in freshwater aquatic toxicology. To help the
Aaency fulfill its responsibilities as set forth in the Clean water
Act. ERL-Duluth is examining ways that might be used to draw valid
and useful conclusions on the basis of a small amount of data
concerning toxicity to aquatic organisms without requiring the
collection of large amounts of additional toxicity data in all
situations. The work reported here deals with analyses of acute
values and final acute-chronic ratios that were derived from
aquatic life criteria documents.
111
-------
ACKNOWLEDGEMENT
The authors thank R.J. EricJcson for his help in designing
these analyses and interpreting the results.
-------
ABSTRACT
Data concerning acute toxicity and acute-chronic ratios were
obtained from draft and final aquatic life criteria documents so
that several kinds of analyses could be performed. An analysis of
the dependence of the Final Acute Value (FAV) on the number of
Genus Mean Acute Values (GMAVs) used in its calculation
demonstrated that, although for many data sets the FAV increased as
the number of GMAVs increased above the minimum of eight, the FAV
decreased for a few data sets. Final Acute Value Factors (FAVFs),
which are intended to relate the results of one or a fev acute
toxicity tests to a FAV, were derived by empirical and theoretical
methods. The empirical derivation of FAVFs was accomplished using
specially selected samples (i.e., subsets) of acute values obtained
from data sets contained in the criteria documents. FAVFs for
samples containing 1 to S acute values, where each acute value
satisfied a different minimum data requirement, were determined by
two methods. Whether calculated on a worst case basis or as 95*
percentiles from random sampling, the FAVFs decreased substantially
as sample size increased from 1 to 8 for both fresh and salt water.
Also for both waters, when the sample was required to contain an
acute value for one or more species that are sensitive to many
chemicals, the FAVF was much smaller, especially at small sample
sizes. The theoretical derivation of FAVFs was based on the log-
triangular distribution. Final Chronic Value Factors (FCVFs) wer
calculated by applying the log-normal distribution to Final Acute
Chronic Ratios given in criteria documents; the 95th percent lie FCVF
was 43 for fresh water and 39 for salt water.
-------
CONTENTS
Pace
Notice 11
Foreword ni
Acknowledgement iv
Abstract v
Figures vii
Tables viii
1. Introduction l
2. Database Management ........ 4
3. Dependence of the FAV on the Number of GMAVs 6
4. FAVFs Based on Highest Acute Values 8
5. FAVFs Based on Random Samples of Acute Values 11
6. FAVFs Based on the Log-triangular Distribution 16
7. Final Chronic Value Factors 22
8. Discussion 24
References 29
Appendix 66
VI
-------
FIGURES
page
1 probability Plot for Chlordane in Salt Water 30
2 Log-normal Probability Plot of Freshwater FACRs .... 31
3 Log-normal Probability Plot of Saltwater FACRs ..... 32
VII
(W
-------
TABLES
Paoe
1 Sources of Data Sets 33
2 Fields Used in the Database COMBO 35
3 Data Sets Not Satisfying the Minimum Data Requirements . 37
4 Values of FAV and S for Data Sets in Criteria Documents 38
5 Dependence of the FAV and S on the Number of GMAVs in
Fresh Water 40
6 Dependence of the FAV and S on the Number of GMAVs in
Salt Water 46
7 Worst Case Acute Values 51
8 Worst Case FAVFs 52
9 Minimum Data Requirements Ranked by Acute Value .... 53
10 Worst Case Acute Values When Daphnids Were Required . . 54
11 Worst Case FAVFs When Daphnids Were Required 55
12 Worst case Acute values When Salmonids Were Required . . 56
13 Worst Case FAVFs When Salmonids Were Required 57
14 Worst Case Acute Values When Both Daphnids and
Salmonids Were Required 58
15 Worst Case FAVFs When Both Daphnids and Salmonids
Were Required 59
16 Summary of Geometric Mean Worst Case FAVFs .60
17 Freshwater Summary FAVFs (Version 1) 61
18 Freshwater Summary FAVFs (Version 2) 62
19 Saltwater Summary FAVFs (Version 2) 63
20 FAVFs Based on Log-triangular Distribution 64
21 FACRs from Criteria Documents .... 65
Vlll
-------
SECTION 1
INTRODUCTION
The most recent version of the "Guidelines for Deriving
Numerical National Water Quality Criteria for the Protection of
Aquatic Organisms and Their Uses" (Stephan, et al., 1985), commonly
known as the "National Guidelines1*, describes a set of procedures
that can be used to derive water quality criteria for aquatic life.
The National Guidelines specify minimum data requirements that
should be satisfied if aquatic life criteria are to be derived.
Each minimum data requirement specifies both a type of test and a
type of species for which an acceptable test result should be
available. If both a freshwater criterion and a saltwater
criterion are to be derived for a chemical, results of sixteen
acute toxicity tests and three chronic toxicity tests are required.
If a criterion is to be derived for a chemical in }ust one of the
waters, results of only eight acute tests and three chronic tests
are required.
The data required to derive either a freshwater or a saltwater
criterion are available for only a very small percentage of the
commercially important chemicals and the cost of generating a
complete set of the required data for a chemical in either wate
will usually exceed $100,000. In many situations, howeve
although a criterion might be desirable, it might not be necessar
For example, in order to determine whether a measured or predicted
concentration of a chemical in a body of water is cause for concern
because of toxicity to aquatic organisms, it might be sufficient to
know that the, probability is very high that that concentration is
lower than the criterion would be for that chemical. If the
measured, or jyredic ted concentration of the chemical was high enough
to warrant further*con«ideratIron, a dtcivion might be» "Matte to
generate additional toxicity data, but it still might be possible
to adequately evaluate the situation without generating all the
data necessary to derive an aquatic life criterion. If, however,
the chemical was found in enough bodies of water at concentrations
that are cause for concern, it might be decided that a criterion
should be derived for the chemical. Thus if acceptable ways can be
developed to evaluate the acceptability of measured and predicted
concentrations of chemicals in ambient water using results of one
or a few acute toxicity tests, it might be possible to avoid the
generation of additional toxicity data, which might help ensure
that resources are applied to the chemicals for which data are most
needed.
An aquatic life criterion derived according to the National
Guidelines consists of two numbers, the lower of which is the
Criterion Continuous Concentration (CCC). For many chemicals the
CCC is equal to a Final Chronic Value (FCV), which is usual
derived by dividing a Final Acute Value (FAV) by a Final Acu
-------
Chronic Ratio (FACR), i.e., it is often true that: CCC * FCV •
FAV/FACR. The FACR is used to extrapolate from acute toxxcity to
chronic toxicity. Although the CCC is often equal to a FCV, it
might also be based on field data and data concerning effects of
residues or effects on plants. The higher of the two numbers in an
aquatic life criterion is the Criterion Maximum Concentration
(CMC), which is derived by dividing the FAV by a factor of 2.
Because the FAV is an estimate of an LC50 or EC50, this factor of
2 is used to reduce the concentration from one that kills or
affects 50 percent of the individuals of the species of concern to
a concentration that will kill or affect a much lower percentage of
the individuals.
If it is desirable to design a procedure that can be used to
draw a conclusion concerning a criterion for a chemical on the
basis of results of one or a few acute toxicity tests on that
chemical, it would seem appropriate to mimic the procedure
described in the National Guidelines for calculating a water
quality criterion. A simple way of imitating this procedure would
be to divide the lowest Genus Mean Acute Value (GMAV) available for
a chemical by a Final Acute Value Factor (FAVF) in order to
calculate a Lower Bound FAV (LBFAV) . The FAVF would be derived so
that the probability is high (e.g., 95%) that the FAV for the
chemical would be higher than the LBFAV. Then the LBFAV could be
divided by a Final Chronic Value Factor (FCVF) to calculate a Lower
Bound FCV (LBFCV), on the assumption that the LBFCV can be used as
a lower bound on the ccc. This procedure can be summarized as:
(Lowest GMAV)/FAVF - LBFAV afcd LBFAV/FCVF - LBFCV.
Because the FAVF and FCVF are denominators, if they are derived as
upper bounds, then the quotients can be considered lover bounds.
If the LBFCV can be considered a lower bound on the water quality
criterion, then a measured or predicted concentration that is below
the LBFCV can probably be considered of little concern. This
approach was used in Michigan's "Rule 57" (Michigan DNR, 1987) and
in the "Guidelines for Deriving Ambient Aquatic Life Advisory
Concentrations" (U.S. EPA, 1987).
An alternative approach would be to develop factors that would
be applied to the mean of the available GMAVs rather than to the
(lowest of the available GMAVs. This approach was not tested here
for two reasons. First, the occurrence of "greater than" values
will be more of a problem if factors are applied to the mean,
rather than the lowest, of the available GMAVs. Second, adding
values to a data set can raise or lower the mean, whereas
additional values can only lower the lowest value. Thus it might
be easier to understand the dependence of the factors on sample
size if the factors are applied to the lowest, rather than the
mean, of the available GMAVs. When only one GMAV is available, the
choice between the mean and the lowest GMAV is moot because the
mean and the lowest value are the same.
-------
Th* purpose of the work described here v«s to derive FAVFs a
FCVFfi, specifically taking into account the fact that for chemicals
for which few data are available, most available GMAVs will be an
acute value from one acute toxicity test and will rarely be the
mean of a number of acute values. In addition, it vas desired that
the FAVFs be dependent on the size of the data set, so that the
FAVF decreases as the uncertainty decreases. This vas implemented
as a dependence on the number of minimum data requirements that
were satisfied by the data set. Because this work concerned FAVs
and FCVs, it was natural to base the derivation of FAVFs and FCVFs
on data contained in draft and final aquatic life criteria
documents (Table 1).
Section 2 of this report describes the formation of the
database on which the analyses described in Sections 3, 4, and 5
were performed. Section 3 describes the analysis of the dependence
of the FAV on the number of GMAVs in data sets that satisfy the
eight minimum data requirements concerning acute toxicity.
sections 4 and 5 describe two methods that were used for empirical
calculation of FAVFs. The method described in Section 4 calculates
worst case FAVFs. In Section 5, two versions of a different method
are used to calculate FAVFs on the basis of random samples.
Section 6 of this report describes the theoretical calculation of
FAVFs. calculation of FCVFs is described in Section 7.
I TO
-------
SECTION 2
DATABASE MANAGEMENT "^
Empirical analyses concerning FAVs and FAVFs were bas<*3 on the
toxicological database WQUAL and the taxonomic database TAXON. All
toxicity values in WQUAL concern acute toxicity to aquatic animals
and are derived from the draft and final criteria documents listed
in Table 1. Both databases were originally compiled **. the U.S.
EPA's Environmental Research Laboratory in Narragansett, RI, by
David Hansen and Walter Berry with the help of Charles stephan at
the U.S. EPA's Environmental Research Laboratory in Duluth, MN.
After the databases were transferred to Duluth, they were checked
against the source criteria documents and some data were also
checked against the original references; all errors found were
corrected. To facilitate analyses, pertinent fields from WQUAL and
TAXON were used to create a third database named COMBO. Data
fields included in COMBO are described in Table 2.
A computer program was written to determine whether each data
set in COMBO satisfied the minimum data requirements concerning
acute toxicity specified in the National Guidelines. Eleven
freshwater and two saltwater data sets did not satisfy all eight
minimum data requirements (Table 3). Most of these data set's were
from criteria documents that were based on the 1980 version of the
National Guidelines, because the minimum data requirements were not
aj. strict in the 1980 version of the National Guidelines as they
are in the 1985 version.
After the incomplete data sets were removed from
consideration, Species Mean Acute Values (SMAVs) were calculated
for the remaining 29 freshwater and 28 saltwater data sets
according to the following rules:
1. Acute values with B, E, H, or M in the Remark field (Table 2}
were not used.
2. If, after application of rule 1, a data set contained one or
more "flow-through measured" acute values (indicated by an "F"
in the Technique field and an "M" in the Measurement field) for
a species, only those values were used to calculate the SMAV for
that species, as specified in the National Guidelines. If the
data set did not contain a "flow-through measured** value for the
species, then all acute values, except as per rule l, in the
data set for that species were used to calculate the SMAV.
3. The freshwater data sets for ammonia, cadmium, chromium(III),
copper, lead, nickel, pentachlorophenol, and zinc were nqt
subject to rule 2. In the criteria documents for these
chemicals acute toxicity in fresh water is related to one or
more water quality characteristics such as hardness, pH, or
-------
temperature. For each of these data sets all acute values
except as per rule 1, were first adjusted to the same value o
the character!:§tic(s) using the relationship given in the
criteria document and then all adjusted values were used to
calculate SMAVs.
For each of the 57 data sets, GMAVs and FAVs were then calculated
using the procedures described in the National Guidelines, except
that the FAV was never lowered to equal the SNAV for a sensitive
important species, as was done, for example, in the criteria
document for ammonia in fresh water. For the criteria documents
that were based on the 1985 version of the National Guidelines, the
SMAVs, GMAVs, and FAVs calculated by the computer were compared
with those given in the criteria documents, and all discrepancies
were examined. (Cn the criteria documents prepared according to
the 1980 version of the National Guidelines, the FAV was calculated
from SMAVs, not SMAVs, and so only SMAVs could be compared.)
WQUAL, TAXON, and COMBO were corrected so that all remaining
discrepancies were due to errors in the criteria documents.
The scale factor 5, which is the slope of the probability plot
of the four lowest GMAVs, was also calculated for each data set.
The scale factor .is an estimate of o>/I?, where o is the standard
deviation of the population from which the GMAVs are assumed to
have been drawn, which population is assumed to have a log-
triangular distribution. The scale factor was calculated becaus
it might be useful in interpreting results presented in Section
and so that it could be used in the theoretical derivation of Fina
Acute Value Factors (FAVFs) in Section 6. The scale factor is
based on the four lowest GMAVs because these determine the slope
used in the calculation of the FAV.
The values of the FAV and 5 calculated from each data set are
presented in Table 4.
-------
SECTION 3
DEPENDENCE OP THE FAV ON THE NUMBER OF GMAVs
The procedure described in the National Guidelines for
calculating an FAV requires at least eight GMAVs such that the
eight minimum data requirements concerning acute toxicity are
satisfied. This procedure vas designed to be conservative in the
sense that, on the average, greater uncertainty (i.e., fewer data)
should result in more protection (i.e., a lover FAV). The intent
was to establish this relation between uncertainty and conservatism
by (a) specifically requiring data for one or more generally
sensitive species in the minimum data requirements in addition to
requiring data for a variety of species, and (b) calculating the
PAV as a 95th percentile so that the amount of extrapolation below
the lowest GMAV depends on the number of GMAVs that can be
calculated from the data set. (When there are more than twenty
GMAVs, the FAV is usually calculated by interpolation between the
two lowest GMAVs rather than by extrapolation below the lowest
GMAV). Thus the procedure was intended to produce, on the average,
higher FAVs as the number of GMAVs increases above 8. The intended
relation between uncertainty and conservatism will be accomplished,
however, only if the minimum data requirements result in bias
toward an overabundance of sensitive species for each chemical.
The actual relationship of the PAV to n * the number of GMAVs
was determined for each data set by calculating FAVs from samples
(i.e., subsets) of randomly selected acute values. For each sample
the first eight acute values were randomly selected, subject to the
restriction that each acute value satisfied a different minimum
data requirement. Additional acute values, for n>8, were selected
so that each new value was for a genus which was not already
represented in the sample. Because the mean FAV obtained here for
n-8 was to be used in Sections 4 and 5, this sampling was designed
to be compatible with the sampling used in Sections 4 and 5. Thus
acute values, not SMAVs or GMAVs, were randomly sampled and the
presence of a B or M in the Remarks field and the "flow-through
measured" rule were ignored, as explained in Section 4.
For each data* set the sample sizes used began with n«8 and
progressively increased by 2 (i.e., 8, 10, 12,...) up to, but not
including, the total number of GMAVs available from the data set.
One hundred samples were randomly selected for each of the sample
sizes used with any one data set and then 100 FAVs were calculated.
The geometric mean and the minimum and maximum values of the FAV
are presented in Tables 5 and 6 for each sample size used with each
data set. Tables 5 and 6 also contain similar data concerning the
scale factor S. The values of FAV and 5 given in Table 4 were
calculated using a different set of rules and so are not comparable
to the values given in Tables 5 and 6. No values are in Table 5
-------
for phenanthrene, thallium, and tributyltin or in Table 6 fo
cyanide because only eight GMAV» could be calculated from thes
data sets.
Many data sets show*d the trend of higher mean FAVs with
higher values of n, although for some data sets the mean FAV
increased very little as n increased. For a few data sets,
however, the geometric mean FAV decreased as n increased.
Especially for pentachlorophenol in fresh water and copper in salt
water, the mean FAV decreased as n increased over a wide range of
values of n. The difference between increasing and decreasing
trends did not seem to be related to the scale factor. For most
data sets, the standard deviations of both the FAV and S decreased
as n increased.
One way to summarize the data presented in Tables 5 and 6 is
to examine the ratio of the mean FAV for n-8 to the mean FAV for
n-18 for those data sets for which at least 18 GMAVs could be
calculated. (The sample size of 18 is used merely because it is a
reasonably large size for which many ratios can be calculated.)
Twelve freshwater data sets and ten saltwater data sets contain
acute values for more than 18 genera and the 22 ratios are:
RATIO NUMBER
6.8 1
4.4 1
3-4 2
2-3 4
1-2 11
0.6-1 3
The range of the ratios is about a factor of ten. For most data
sets there was a small to moderate increase in the FAV when n
increased from 8 to 18, but for a few data sets the increase was
more than a factor of 3. Theee results*- show that, the intended
conservatism was achieved by a factor of two or more for only 8 of
the 22 data sets. For 3 of the 22 data sets the relationship
between FAV and the number of GMAVs was the opposite of that
intended.
-------
SECTION 4
FAVFs BASED ON HIGHEST ACUTE VALUES
A Final Acute Value Factor (FAVF) is intended to be used in
the calculation of a LBFAV from the lowest GMAV available from a
data set. The FAVF is intended to be conservative in the sense
that, on the average, greater uncertainly (i.e., fewer data) should
result in more protection. In other words, a lack of data should
rarely result in underprotaction, which can only be achieved by
having overprotection most of the time. Both empirical and
theoretical methods were used to derive FAVFs and if both kinds of
methods are used appropriately, the results should be comparable.
Two kinds of empirical calculations were performed. The worst case
approach is described here in Section 4, whereas a statistical
approach is described in section 5. After these two kinds of
empirical approaches are discussed in Sections 4 and 5, a
theoretical approach is discussed in Section 6.
The objective of the analyses presented in this section was to
determine the worst case (i.e., largest) FAVF that could be
obtained from each data set by using one through eight acute values
that are actually in the data set, where each selected acute value
had to satisfy a different minimum data requirement. To accomplish
this, the eight highest acute values that collectively satisfied
the eight minimum data requirements were selected from each data
set. This selection of acute values ignored the presence of a B or
M in the Remarks field and also ignored the "flow-through measured"
rule. These portions of the rules discussed in Section 2 were
ignored here because FAVFs are intended to be used with chemicals
for which few toxicity data are available. Some types of decisions
that might be possible with large data set are unlikely to be
possible with small data sets. For example, when few data are
available, the B and H Remarks and the "flow-through measured1* rule
will rarely be applicable. (Acute values ad3usted for water
quality were used for ammonia, pentachlorophenol, and some metals
even though such adjustments are not likely to be possible when
dealing with small data sets.) FAVFs are intended, however, to be
applied only to acute values that have been subjected to screening
procedures similar to those described in the National Guidelines.
The GMAVs and FAVs discussed in Section 2 and presented in Table 4
are intended to be calculated from data sets that satisfy the
minimum data requirements and therefore will generally be based on
more acute values.
The eight highest acute values selected from one data set,
with the restriction that each acute value had to satisfy a
different minimum data requirement, were ranked in descending
order. These ranked acute values were the worst case samples
(i.e., subsets) of sizes N«l to N»8, where the highest of the eight
acute values was considered the sample of 1, the two highest values
8
ITS
-------
w*r* con«id*r*d th* sample of 2, and so on. The acute valu* tha
was th* saapl* size of 1 was th« highest acut* valu* in th* data
s*t becaus* each acut* valu* will satisfy a minimum data
requirement. For *ach data set, th* lowest valu* in *ach of the
eight samples was then divided by th* geometric mean FAV for the
sample size of 8 given in Table 5 for the sarn* data s*t. (Because
Table 5 does not contain values for phenanthrene, thallium, and
tributyltin, values for these chemicals were obtained from Table
4.) These eight 'quotients ar*, therefore, the worst case FAVFs
that can be generated from the data set using acute values that are
actually in the data set. This analysis was conducted on
freshwater data sets only.
For each of the 29 freshwater data sets, the ranked eight
highest acute values that satisfied the eight minimum data
requirements are presented in Table 7. Table 8 presents the worst
case FAVFs, which were calculated by dividing the values for a data
set in Table 7 by the mean FAV given in Table 5 (or Table 4 when
necessary) for the same data set. Table 8 also shows the
arithmetic and geometric means across data sets. The values were
not tested for normality on either an arithmetic or logarithmic
scale, but quotients are usually more like a log-normal than a
normal distribution. Th* geometric mean FAVFs decreased from 541
for only on* (i.e., the highest) acut* value to 6 when all eight
minimum data requirements were satisfied. Interestingly, in Tables
7 and 8, the values for antimony(III), chromium(VI), coppe
parathion, and thallium for the sample size of 7 are more than 1
times greater than the values for the sample size of 8.
Analyses were also conducted on the effects of requiring that
each sample, regardless of sample size, contain an acute value for
one or more sensitive species. Daphnids and salmonids were tested
as likely sensitive sp*cies. In th* case of the daphnid, for
example, when th*> high**t acut* value** w*r* selected fro* a- data
set, th* highest acut* valu* for a daphnid was selected as the
sample of 1 and thus was contained in each successively larger
sample in th* sequence of eight samples from that data set. Table
9 shows th* minimum data requirement that was satisfied by each of
the acut* values that ar* in Tabl* 7. For example, the positions
of th* daphnid valu*s (cod* Dl; s*« Tabl* 2) within rows shows the
ranking of daphnids relative to the other types of species. In
this worst case analysis, the daphnid was the most resistant type
of species to acenaphthene and phenanthrene, but the most sensitive
to cadmium, chlorine, chromium(VI), copper, 2,4-dimethylphenol,
mercury, nickel, parathion, selenium(IV), and zinc. (This
indication of relative sensitivity might not agree with that seen
in the criteria documents because a 8 and/or M in the Remark field
and the "flow-through measured" rule were ignored in this worst
case analysis.) Table 10 presents the highest acute values with ,
the restriction that each sample had to contain the highest acute
value available for a daphnid, and Table 11 presents th
corresponding woist case FAVFs. For most data sets, requiring t
-------
highest value for a daphnid in each sample reduced the FAVFs
substantially. The geometric mean FAVF averaged across data sets
for N-l, for example, was reduced from 541 to 22 (Tables 8 and 11).
The impact of requiring the highest acute value for a daphnid
decreased as sample size increased.
A similar analysis was conducted to examine the effects of
requiring that each sample, regardless of size, contain the highest
acute value for a salmonid. Table 12 presents the highest acute
values with the restriction that each sample, regardless of sample
size, had to contain the highest acute value for a salmonid in the
data set, and Table 13 presents the corresponding worst case FAVFs.
As with the daphnid, the mean FAVF was substantially reduced by the
presence of the highest acute value for a salmonid, although the
reduction was not as great as that produced by daphnids. For N-l,
the geometric mean FAVF across data sets was reduced from 541 to 51
by the inclusion of the value for a salmonid.
The last analysis of worst case FAVFs was to examine the
effects of requiring both the highest acute value for a salmonid
and the highest acute value for a daphnid in each sample of size
N-2 through N»8. Table 14 presents the highest acute values,
subject to the requirement that each sample in the sequence of
samples from a data set must contain both the highest acute value
for a daphnid and the highest acute value for a salmonid; Table 15
presents the corresponding worst case FAVFs. This dual requirement
resulted in the lowest mean FAVFs, because salmonids could be
sensitive when daphnids were resistant and vice versa. Requiring
both a value for a daphnid and a value for a salmonid yielded only
a slight reduction in the mean FAVF compared to the mean FAVF
derived when a value for only a daphnid was required (Table 16).
The maximum effect occurred at N-2, where the FAVF was reduced from
19 (only daphnid required) to 12 (both daphnid and salmonid
required). Adding the acute value for a salmonid to a sample that
already contained the acute value for a daphnid had little effect
on the worst case FAVF. Even when a value for a salmonid was
required (Tables 12 and 13), the more than a factor of ten decrease
when sample size increased from 7 to 8 occurred for the same five
chemicals noted above for Tables 7 and 8; when a value for a
daphnid was required, the large decrease only occurred for
antimony(III) and thallium (Tables 10, 11, 14, and 15).
All four mean worst case FAVFs at N»8 were € (Table 19). The
mean FAVFs are the same for all four cases because at a sample size
of 8, the four separate worst case samples obtained from any one
data set will contain the same eight highest acute values
regardless of what, if any, of the conditions tested here are
imposed concerning sensitive species.
10
111
-------
SECTION 5
FAVFS BASED ON RANDOM SAMPLES OF ACUTE VALUES
Whereas the previous section concerned FAVFs that were
deliberately calculated to be worst case FAVFs, the FAVFs discussed
in this section were based on random sampling. The empirical FAVFs
discussed in this section were calculated by dividing the lowest of
N randomly selected acute values by a FAV. Two versions of this
empirical methodology were used. The major difference between the
two versions concerned how the FAV used as the divisor was
calculated:
1. In the first version the divisor was the geometric mean FAV
presented for N-8 in Tables 5 and 6, except that the FAV from
Table 4 was used for phenanthrene, thallium, and tributyltin
in fresh water and for cyanide in salt water.
2. In the second version the divisor used for each of the eight
samples in a sequence was the FAV that was calculated from the
sample of size N«8 in that sequence.
Thus, in the first version the same mean FAV was used with each
sample obtained from a particular data set, whereas in the second
version a FAV was separately calculated for use with each sequence
of 8 samples.
The first version of this random selection method was applie
to freshwater data sets only, whereas the second version w-
applied to both freshwater and saltwater data sets. In bo
versions of this method, samples (i.e., subsets) of eight acute
values were randomly selected from each data set, with the
restriction that each selected acute value within the sample of
eight had to satisfy a different minimum data requirement. Each
time a sample of eight acute values was created, samples of sizes
1 through 7 were also created, with each nev acute value increasing
the sample size by 1. Thus when a sample of size eight was
selected, a sequence of eight samples of sizes 1 through 8 was
being selected.
To avoid the possibility that the minimum data requirements
might confound the effect of sample size (e.g., by always selecting
an acute value for a salmonid first), the minimum data requirements
themselves were randomized before each random selection of acute
values; one acute value was then selected at random to satisfy each
of the randomized minimum data requirements. The B and M Remarks
and the "flow-through measured" rule were ignored here for the same
reason they were ignored in Section 4, and adjusted values were
used here just as they were in Section 4.
The lowest acute value present in each sample of size 1 to 8
was used as the numerator when calculating the FAVF. In the first
version of this method the random sampling procedure was repeated
99 times to generate 99 sequences of eight samples from each da
11
-------
set; then the lowest acute value in each sample was divided by the
geometric mmmn FAV presented for N-8 in Tables 5 and 6, except that
the FAV given in Table 4 was used for the four data sets named
above. In the second version, 199 sequences of eight samples were
generated from each data set and the lowest acute value in each
sample was divided by the FAV calculated from the sample of size
N-8 in that sequence. The sampling was repeated 199 times in the
second version to account for the probable increase in variability
caused by using an individual FAV, rather than a mean FAV, as the
divisor.
The design and results of these analyses can be summarized as
follows:
version 1 Version 2 Versiop 2
Freshwater Freshwater Saltwater
8 sample sizes 8 sample sizes 8 sample sizes
29 data sets 29 data sets 24 data sets
99 samples per 199 samples per 199 samples per
sample size sample size sample size
For each group of 99 FAVFs calculated in version 1, the median (50th
percentile) FAVF and the 95th percentil* FAVF were calculated,
resulting in twenty-nine medians and twenty-nine 95th percentiles
for each sample size. Then the median and 95th percentile of the
twenty-nine medians and the median and 95th percentile of the
twenty-nine 95th percentiles were calculated. Thus these
calculations resulted in four summary FAVFs:
a. median of 29 medians.
b. 95th percentile of 29 medians.
c. median of twenty-nine 95th percentiles.
d. 95th percentile of twenty-nine 95th percentiles.
In addition to these two medians (a and c) and two 95th percentiles
(b and d) that were calculated for each sample size, one more
median and one more 95th percentile were calculated using all 29*99
- 2871 FAVFs that had been calculated for each sample size. These
last two summary FAVFs are:
e. the overall median - the median of the 2871 FAVFs.
f. the overall 95th percentile - the 95th percentile of the 2871
FAVFs.
In the second version of the method, these same six summary FAVFs
were calculated from 199 FAVFs, rather than from 99 FAVFs, for each
sample size for each data set. These six summary FAVFs were
calculated for each of the eight sample sizes for version 1 in
fresh water (Table 17), version 2 in fresh water (Table 18), and
version 2 in salt water (Table 19).
The overall 95th percentile FAVF is probably the best of the
six summary FAVFs for two reasons:
(1) The two overall FAVFs are directly based on 2871 values. The ,
other four summary FAVFs are directly based on only 29 values
and are thus more dependent on the magnitudes of the mean
12
-------
FAVF» for a few data sets. This will have much more impact on
a 9S1" percentile FAVF than on a median. Indeed, the median
of the 29 medians is very similar to the overall median in
Table 17; the same in true for Table 18 and for Table 19.
(2) The two sources of variability of concern here are variability
within each chemical and variability between chemicals. The
only two summary FAVFs that deal with both sources of
variability are the 95th percentile of the 95th percentiles and
the overall 95th percentile. The 95th percentile of the 95th
percentiles is probably an extreme summary FAVF; it is always
the largest of the six summary FAVFs, and often by a large
margin. The overall 95th percentile FAVF has the most
straightforward conceptual basis.
For these reasons most of the discussion of the results will focus
on the overall 95th percentile FAVF.
In the course of these analyses, a concern arose as to whether
there was a "weighting1* effect of taking 99 random samples from
data sets for which the total number of possible combinations was
fewer than 99, because some combinations would necessarily be
obtained more than once. To evaluate this., results obtained using
99 random samples were compared with results obtained using each
possible combination once and only once. For example, if a data
set contained 16 acute values and sampling was done for the sample
size of N«l, this data set would contribute only 16 samples if eac
possible combination was used only once, whereas each combinati
would be used about 6.5 times on the average if 99 samples were
taken. Calculations of medians and 95th percentiles showed that the
two sampling strategies did not produce substantially different
results; for example, FAVFs calculated as overall 95th percentiles
using the first version at a sample size of N-l were 10,500 and
13,600 for the two sampling strategies.
The effect of requiring that each sample, regardless of sample
size, contain an acute value for one or more sensitive species was
also examined using this empirical method. For freshwater data
sets daphnids and salmonids were used as sensitive species, whereas
for saltwater data sets crustacean species in the families Mysidae
and Penaeidae and fish species in the genus Menidia were used as
sensitive species. The saltwater minimum data requirements require
an acute value for either a mysid or a penaeid or both, partly
because these are considered sensitive species. Most of the
saltwater data sets also contain a value for a species in the genus
Menidia. An examination of the results obtained with salmonids and
daphnids in fresh water and a review of the saltwater data sets
indicated that the three kinds of saltwater species named above
probably would have a major impact on the saltwater FAVF. In order
to have the best comparison it was desirable to use only data sets
that contained acute values for all the kinds of species that were
to be tested as sensitive species. Many saltwater data sets
contain a value for a mysid or a penaeid but not for both, so the
two were considered together as one kind of sensitive species, ju
13
-------
as they are considered together in the saltwater minimum data
requirements. All but four of the saltwater data sets contained a
value for a species in the genus Menidia. so these four data sets
were not used in this analysis. Therefore the FAVFs in salt water
were examined using 24 data sets.
Similar to the procedure used in Section 4 to assess the
impact of sensitive species, additional 99 or 199 sequences of 8
samples were randomly selected with the restriction that an acute
value for the sensitive species was chosen first (e.g., at N-i the
acute value chosen was a randomly selected value for a daphnid).
Each sample of 8 was completed with the remaining minimum data
requirements being satisfied in random order. Then, in order to
obtain a better indication of the effect of the sensitive species
with the second version of the method, in another group of 199
sequences, each sample of 8 was drawn by excluding the sensitive
species until the eighth (i.e., the last) acute value was drawn
(e.g., the requirement for a daphnid was satisfied last). This
allowed comparisons of samples of sizes 1 to 7 that did and did not
contain an acute value for the sensitive species.
Table 17 summarizes the results obtained with the first
version of this method, which was used only with freshwater acute
values. For the four different sampling strategies, which differed
in their consideration of sensitive species, the six summary FAVFs
are presented for each of eight sample sizes. Sample size had a
strong effect on the FAVF. In several cases, FAVFs showed an
approximately exponential decrease as sample size increased, when
no sensitive species was required, the overall 95th percentile FAVFs
decreased from 10,500 to 15 as sample size increased from 1 to 8.
Requiring a randomly selected acute value for a daphnid in every
sample reduced the overall 95th percentile FAVF at N-l from 10,500
to 75. The reductions that occurred at sample sizes of 2 through
8 were less dramatic. In contrast, requiring an acute value for a
salmonid increased the overall 95th percentile FAVF from 10,500 to
16,398. Also, adding a value for a salmonid to a sample that
already contained a value for a daphnid usually raised, and never
lowered, the FAVF.
Results obtained using the second version of this method with
freshwater acute values (Table 18) were quite similar to those
obtained using the first version of the method (Table 17). The
overall 95th percentile FAVFs generated using the two versions were
generally within a factor of two. The same exponential
relationship of decreasing FAVF with increasing sample size was
evident with the second version. When no sensitive species was
required or excluded, the overall 95th percentile FAVF at N=l was
10,500 for version 1 and 8367 for version 2 (Tables 17 and 18).
Requiring a randomly selected value for a daphnid resulted in FAVFs
of 75 and 94 for the first and second versions, respectively. f
FAVFs generated when daphnids were required were consistently lower
than FAVFs generated without concern for sensitive species, and
14
-------
FAVF* generated when daphnids were excluded were consistentl
higher than FAVFs generated without concern for sensitive species.
Ac with the* first version of th* method, requiring an acute value
for a salaonid in the sample did not reduce the overall 95th
percentile FAVF at any sample size.
Saltwater acute values were studied using only the second
version of the method and generally produced FAVFs (Table 19) that
were somewhat lower than the comparable freshwater FAVFs (Table
18). The saltwater overall 95th percentile FAVF for N-l was 4903,
compared with 8400 for fresh water. Requiring sensitive species
substantially reduced the saltwater FAVFs. Requiring a randomly
selected acute value for either a mysid or a penaeid, for example,
reduced the FAVF at N-l from 4903 to 74. Requiring an acute value
for a species in the genus Menidia reduced the FAVF at N»l from
4903 to 314. At sample sizes of N-2 to N-7, when mysid or penaeid
was required, Menidia was excluded and vice versa. This is why at
N-3, for example, requiring Manidia raised the FAVF from 101 to
138. A better comparison is to compare Manidia or mysid-penaeid
required to both Menidia and mysid-penaeid excluded. In this case,
requiring Menidia and mysid-penaeid, respectively, even with the
other excluded reduced the FAVF at N-3 from 174 to 22 and 38 (Table
19}. The FAVF was affected very little by adding a value for a
species in the genus Menidia to samples that already contained a
value for either a mysid or a penaeid.
with both versions of the method, samples of size 8 a
complete data sets for calculating FAVs. At N-8 with the secon
version, the overall 95th percentile FAVF is 8 in fresh water (Table
18) and 4 in salt water (Table 19). For the first version of the
method, the overall 95th percentile FAVF was IS, probably because
the mean FAV was less appropriate to individual samples than was
the individually calculated FAV. Also, because the sample of 8 is
a complete data set, the fact that the range of the four overall
95* percentile FAVFs in Table 17 is very smell indicates that the
four separate sets of 99 samples of size 8 gave nearly identical
results. The comparable overall 95th percentile FAVFs in Tables 18
and 19 indicate that the same conclusions apply to the separate
sets of 199 samples. This low degree of variation at N-8 indicates
that the 99 and 199 random samples used with the first and second
versions of this method, respectively, were sufficiently large to
produce robust results.
Version 2 of the method is probably the better version because
the FAV used with each sequence of samples is derived from the
sample of size N-8 that is in the sequence itself. Within sampling
error, the two versions should give the same mean FAV for each
chemical and therefore the same median FAVF for a chemical. Thus
the median of the medians and the overall median should be the
same, within sampling error, for the two versions. Indeed, the
medians of the medians are in close agreement between Tables 17 and
18, and the overall medians are also similar in the two tables.
15
-------
SECTION 6
FAVTs BASED ON THE LOG-TRIANGULAR DISTRIBUTION
The empirical simulations discussed in th« previous section
gave FAVFs by which the lowest GMAV in a sample should be divided
in order to have a high probability that the resulting LBFAV is
below the FAV. The FAV itself is an estimate of the 5th percent lie
assuming a log-triangular distribution in the lower tail (Erickson
and Stephan, 1988). This section describes a method for
determining exact theoretical 95 percent confidence limits on the
5th percentile assuming a log-triangular distribution. This exact
theoretical method gives FAVFs that can be compared with the
empirical FAVFs.
The Triangular Distribution
The procedure used to calculate FAVs applies a log-triangular
distribution to the values in the lower tail (Erickson and Stephan,
1988). The probability density function of the standard triangular
distribution at any value z is given by
f(z)
-±-Z -/6 4 Z 4 0
6
Z
Integrating this density function and simplifying gives the
cumulative distribution function
o
F(z) - { 12 ,
1 - -~(z - v/5")2 0 s z s ,/S" •
The inverse of the cumulative distribution function for any
probability p is then
vfl2p - y/g 0 s p < 1/2
- p) * v/5" 1/2 s p s
I
\ -
Parameter and Percentile Estimation
If a random variable X has a triangular distribution with mean
ti and variance o2, then X can be viewed as a linear rescaling of
z, such that X * \L * oz . Given data X^, . . . .X,,, a probability plot
can be formed by plotting the ordered values X(c) versus
standardized triangular quantiles Q(f) » ^"l-rpr) (Rice, 1988, p. ,
292). For points below the median, this corresponds to plotting
16
-------
ordered values of X versus vT3p - v/S" - Thus by plotting X(r, versu
fi>fr) • J 12(-]^) " & • tfc*- intercept is (1 , which is an estimate of
ji, and ths slop* is d, which is an estimate of o.
Whan an FAV is calculated using the procedure described by
Erickson and Stephen (1988), the FAV is usually calculated from the
four lowest ordered GKAVs ^(1)iJC(2)iX(3)s^U) and their standardized
quantiles, i.e., their cumulative probabilities. The FAV is then
J?p, which is the estimate of the p" percentile; thus the FAV is
the estimate of the value that would be greater than p percent of
the values of X. If X is the average of Xa), Jf(2> ,^O),X<4) and 0
is the average of the corresponding standardized quantiles, a_line
fitted to a plot of Jf(r) versus Q(t] must pass through (Q,X] .
Therefore
x - a
and
£ - X - do .
The lower limit L of the resulting triangular distribution is
estimated as
- Jf - d(Q
To estimate £p, the pM percentile of a triangularly distributed X,
for p $ 0.5,
£ - fl +
If instead of plotting X(r) versu* J 12J— £-\ - ^5", we plot X([} versus
J — ^- , the result ing slope is S * d/I? . Hence
The relationship between percent lies of a triangular
distribution can be used to derive the confidence bound on the 5tn
percentile. Suppose X^ is the p" percentile estimate from a
triangular distribution with scale factor 5 - dfTS . For pl < 0.5
X • A * d
If Apj is the corresponding estimator of another percentile with p2 t
< 0.5,
17
-------
p, * & *
Then
+ a
For example, if £g,40 estimates the 40th percentile, then the 5th
percent lie is estimated by
#o as « *o 40 * Sty^rre
» #0 «o • 0.415 ,
If, however, pl > 0.5, then
» I * d ( -i2(l-p) *
Therefore, for p± > 0.5 and p2 < 0.5,
x - x * a
If, for example, £0 60 estimates the 60th percentile, then
os • o 60 *
Xa 6a - 0.565 .
Calculating a LBFAV from th^ Lowest Acute Value When N<8
Let X,...,XW be a random sample from a triangular
distribution, and let X^^ be the lowest X. In the notation used
above, jfm was the lowest value, but Xnia will be used here for
clarity. An upper 95% confidence limit for the percentile
corresponding to A^ is 1 - 0.051'". If X^ corresponds at most to
percentile V (for Upper limit), then Xmin provides a lower bound
for the U"1 percentile. A lower bound for the 5th percentile, is
then
Lax,,.,,,"-
where U » 1 - 0.051'*.
The acute values in the lower tails of the data sets are
assumed to be log-triangular. Because these values are
concentrations C^,C2> ...,€„, then Xl * ln(Ca) , where "In" signifies
the natural logarithm, i.e., to base e. In particular
X.,,, - ln(CL,J . Therefore
-------
£7*0.5.
For example, assume a data sat that consists of the eight
GMAVs available for chlordane in salt water: 120, 17.5, 16.9, 11.8,
6.4, 6.2, 4.8, 0.4 (Erickson and Stephan, 1988). Then imagine, for
illustration, that the only data available happened to be the
highest four values: 120, 17.5, 16.9, and 11.8. To calculate a
LBFAV, one vould need to select a value of 5 large enough to give
a conservative estimate of the 5th percentile. A large 5 means that
the 5th percentile potentially could be far below the lowest
observed value. Based on the values of S in Table 4, a
conservative value of 5 - 10 (see below) would achieve 95%
confidence in bounding the 5th percentile. Using this value of
S « 10 for illustration,
lnUl.8)
2.47
U
LBFAV
1 - 0.051/4
0.527.
_ _
2.47 * 5(v^T7oT * LBFAV
below the FAV. En this example this occurs partly because X,5) *
11.8 is less than one would predict on the basis of the probability
plot (Figure 1) of the lowest 4 values, and partly because S > S.
As another example, suppose that the only value available was
120, which is the highest of the saltwater chlordane values. ' Then
U * 0.95, ln(120Ji - 4.79, and
In(LSFAV) - 4.79 + 3(,/0""TJ? * ^1 - 0.95 -
- 4.79 + 10(0.223 * 0.223 -
* -4.88.
LBFAV * e'4 ••
* 0.008.
Equivalently
19
-------
LBFAV * 120/(e° »«7)10
» 0.008.
Again th« LBFAV is below the FAV. This is also partly because
X(l) • 120 is less than would be predicted from the probability plot
of the lowest four values and partly because 3 > S. In this case,
however, the LBFAV is less than the FAV also partly because one of
eight values is at most the 89th percentile of the sample, whereas
the upper limit for the percentile of one random sample is
U - 0.95.
Table 20 shows FAVFs by which the lowest value in the data set
must be divided in order to have the specified percent confidence
that the LBFAV is less than the true 5™ percentile. The FAVFs for
a given value of 5 >1 are the FAVFs for 5*1 raised to the 5 power.
For example, the FAVF for 95% confidence with 5*2 5 and N-l is
(2.63)25 » 11.2. The FAVFs for 50% confidence correspond to
estimating the 5th percentile, rather than estimating an upper limit
on the 5th percentile. Thus if S»10 and we have N*l value, then we
would have to divide that one value by 126 in order to have a 50-50
chance of ending up below the true 5th percentile. If 5-10 and
tf-8, an estimate for the 5th percentile could be found by dividing
the lowest acute value by 1.9. In order to be 95% confident that
the resulting value is below the 5th percentile, the lowest acute
value at tf*8 would have to divided by 28.6.
Using a FAVF based on 95% confidence of being below the 5th
percentile is conservative compared to calculating an FAV. when
eight minimum data requirements are satisfied, the FAV is an
estimate of the 5th percentile. When fewer than eight requirements
are satisfied, the LBFAV will usually be below the 5th percentile
if 5 is chosen to be sufficiently large.
Choosing a Vqlue of S from a Range of Values
If a number of different chemicals have different values of o,
we are interested in the situation where a chemical is chosen at
random. The probability that a LBFAV computed using a fixed 5 is
below the true 5th percentile is a mixture of the probabilities for
the chemicals. When d is used to compute the LBFAV and a is the
true standard deviation and U
-------
Th«n the probability can be rearranged to
P
The corresponding probability for 17 > 0.5 is
<>/0~7oT
which can be rearranged to
p
J *»ia " ** { -RffiStt - v/5" * RJIJTZ + (!-.R)yT2Vo~7oT > .
The probabilities P j -^2 — - < z\ are computed by the standard log-
triangular cumulative distribution function F(z) above.
For example, if R - d/o - 1.5, then the assumed scale factor
3 is 50% too large, and the computed LBFAV will usually be even
lower than desired, so the probability that the LBFAV is less than
the true 5th percentile will be above the targeted 95%. If we have
tf*4 acute values, then £7*0.527 . Using the formula given above for
J7>0 5 and R-1.5, the actual probability that the computed LBFAV
is less than the true 5th percentile is 0.999. If we have tf-7
values, then £7*0.35. Using the formula given above for C/<0.5 and
R-1.5, the actual probability that the computed LBFAV is less tha
the true 5th percentile is 0.998. If R-0.5, the assumed sea
factor is too small and the probability that the LBFAV is less th
the true 5th percentile will be below the targeted 95%. with tf*4
and R-0.5, the actual percentile is 0.64. With tf-7 and R-0.5, the
actual probability that the computed LBFAV is less than the true 5th
percentile is 0.718.
For a given, value of S the^probabilities for a LBFAV computed
from a randomly chosen set of chemicals can be compared by
averaging the probabilities for the given chemicals. Such combined
probabilities can be used to solve* for 3 such that overall 95% of
the computed LBFAVs are less than the true 5th percentile. For
example, suppose the true values of 5 for six chemicals are' 1, 2,
4, 8, 16, and 32. For these six chemicals an iterative numerical
solution gave 18.84 as the value of 3 that would result in an
overall 95% of the computed LBFAVs being less than the true 5th
percentile.
21
-------
SECTION 7
FINAL CHRONIC VALUE FACTORS
Because a Final Chronic Value Factor (FCVF) is intended to be
used when a Final Acute-Chronic Ratio (FACR) cannot be derived, the
FACRs that are in the draft and final aquatic life water quality
criteria documents (Table 1) are the most appropriate data for use
in determining a FCVF. The available FACRs are listed in Table 21.
This list contains only the FACRs, not all the acute-chronic ratios
(ACRs) given in the documents. Only the FACRs are listed because
ACRs are sometimes higher or lower for acutely resistant species
than for acutely sensitive species and the FCVF is intended to be
appropriate for use with sensitive species. In the footnoted
cases, however, the value given in Table 21 is not the same as the
FACR published in the criteria document. For example, the value
given in Table 21 for aluminum is larger than the published FACR
because the FCV in the criteria document was lowered to protect an
important sensitive species. For aluminum, therefore, the value
given in Table 21 was obtained by dividing the published FAV by the
published lowered FCV. No such changes were made because a Final
Residue Value was lower than a FCV. All values in Table 21 are
rounded to two significant figures.
Table 21 gives a "greater than" value for the FACR for
toxaphene in fresh water. This" "greater than1* value is a "right-
censored11 value because the only information available is that the
FACR is somewhere to the right of 38 on the number line. Due to
this right-censored value, nonparametric methods cannot be used to
estimate high percentiles by counting up to a certain ordered
value. To estimate high percentiles in fresh water, the FACRs were
modeled with a parametric distribution and the parameters of that
distribution estimated. The distributions of both the freshwater
and the saltwater FACRs are modeled well by a log-normal
distribution, particularly in the upper end of the distribution.
Figures 2 and 3 show log-normal probability plots of the FACRs for
fresh and salt water, respectively. To estimate percentiles, the
FACRs are modeled as a log-normal distribution for values above 10.
Values below 10 are treated as left-censored so that they
contribute to the analysis but do not affect the shape of the curve
for the high percentiles that are of interest here. In fresh water
62% of the FACRs are below 10, whereas 74% of the saltwater FACRs
are below 10.
Due to the right-censored toxaphene value and the left-
censored values below 10, a computer program that can handle both
types of censoring was needed. In order to perform the analyses
desired here, maximum likelihood routines given in chapter 3 of
Kalbfleisch and Prentice (1980) were revised to handle both types ,
of censoring simultaneously. Originally these routines were for
right censoring but a FORTRAN program was written to handle right
22
-------
and left censored data for both normal and log-norma
distributions. For fresh water this program estimated the mean and
standard deviation to be 1.96 and 1.10, respectively, on the
natural log scale. Using a normal probability function and
transforming back to the original scale, the 95th percentile was
calculated to be 43. For salt water the mean and standard
deviation were estimated to be 1.48 and 1.33, respectively,
resulting in an estimated 95th percentile of 39. The median was
estimated to be 7.8 in fresh water and 4.0 in salt water, both of
which are below 10. The value of 25, which was recommended by
Kenaga (1982) and Call, et al. (1985), was the 87th percentile for
fresh water and the 90th percentile for salt water.
The 95th percentiles are quite similar, partly because of
fortuitous canceling effects. For most chemicals the freshwater
and saltwater FACRs are similar. The major exceptions are
toxaphene with a much larger FACR in fresh water and chromium(VI)
with a much larger FACR in salt water. Thus in botlt-waters there
is one large value that is not in the other water, and the
resulting 95th percentiles are quite similar.
23
tfO
-------
SECTION 8
DISCUSSION
The acute values in some of the data seta used in these
analyses could not be reviewed as ouch as desired. Additional
reviev was desirable because the screening procedures used in the
preparation of the 1980 criteria documents were not as stringent as
those used for the later criteria documents. It is hoped that
additional review would not have substantially impacted the results
of these analyses.
The results presented in this report are based on all the data
sets that were available in draft and final aquatic life criteria
documents at the time this work was begun. Many of the these
pollutants are metals and pesticides, but some of these pollutants
are industrial organic chemicals and others are such pollutants as
ammonia, chloride, chlorine, and cyanide. Thus these results are
based on data for a wide variety of pollutants in both fresh and
salt water and are likely to be applicable to a wide variety of
pollutants.
The results concerning the relationship of the FAV to the
number of GMAVs from which it is calculated indicate that ,-the
minimum data requirements do not provide the intended conservatism
for all data sets. Additional analyses could probably indicate
desirable changes in the minimum data requirements.
Another major result of these analyses is that the range of
acute values for any one chemical is likely to be much larger than
the range of acute-chronic ratios among chemicals. The forty-three
FACRs, 24 of which were in fresh water and 19 of which were in salt
water, ranged from 2 to 51, although one value is only known to be
greater than 38. In contrast, the overall 95th percentile FAVFs
ranged from 4 to 20,000. Thus there is much more uncertainty in
the FAVFs than in the FCVFs. This can be considered fortunate
because acute values are certainly cheaper to determine than acute-
chronic ratios.
The very large range in the individual FAVFs also means that
the magnitude of a summary FAVF depends greatly on the percentile
chosen for its basis. For example, in fresh water when N»l and a
sensitive species is not required, the overall median (i.e., the
50th percentile) was 14.66 and the overall 95th percentile was 8367
(Table 18). Lowering the percentile will simultaneously reduce the
FAVF and reduce the amount of conservatism. As the conservatism is
reduced, there will be less incentive to generate additional data
because additional data will have less effect on the LBFCV. Too
much conservatism, however, will unnecessarily cause the generation *
of additional data.
24
-------
If both the FAVF and the FCVF are calculated as 95
percentilee, the probability that the LBFCV is lower than the CCC
will be greater than 95%. Because the relative range of the FAVFs
is much greater than that of the FCVFs, it is likely that the
probability of the LBFCV being below the CCC will depend almost
entirely on the larger range and so the probability might not be
too much greater than 95%. Additional analyses that combine the
FAVF and FCVF into one factor might be possible. Having two
separate factors i» desirable, however, because there will probably
be chemicals for which an FAV can be calculated but a FACR cannot.
The possibility that a FCVF might be used with a FAV should be
taken into account in the selection of the percentile that is to be
the basis of the FCVF.
The range in the FAVF is due in part to variation among
replicate tests on a chemical using the same species. In a few
cases the agreement among replicate results is very poor. When
many data are available, the chances of detecting "outliers" is
much greater. For chemicals for which few data are available,
SMAVs and GMAVs will usually be the result of one test and will
have greater variability and uncertainty than SMAVs and GMAVs based
on results of many tests. This should be taken into account in the
determination of FAVFs.
The studies of sensitive.species produced some unexpected
results, especially for salmonids. The overall 95th percent!
FAVFs (as well as other summary FAVFs) were so much lower wh
daphnids were required than when salmonids were required that it
was felt necessary to understand why these results did not agree
with the conventional wisdom that both daphnids and salmonids are
sensitive species. Examination of the ranked GMAVs for the 29
freshwater data sets revealed that both daphnids and salmonids were
among the most sensitive species in about the same percentage of
the data sets. It was discovered, however, that even when a
daphnid species appeared to be resistant, the species was rarely
more than 200 times more resistant than the most sensitive species
in the data set. In contrast, salmonids were sometimes up to
44,000 times more resistant than the most sensitive species in the
data set. Thus whether a species (or group of species) will have
a major impact on the overall 95th percentile FAVF depends mostly
on the percent of data sets in which the species is very resistant.
Judgments concerning the relative sensitivities of species
will depend on the criteria used to define "sensitive". For
example, salmonids would probably be considered freshwater
sensitive species based on the worst case FAVFs given in Table 16,
but they would not be considered sensitive species based on the
overall 95th percentile FAVFs given in Tables 17 and 18. Another
consideration is the comparison that is made. Probably the best
way to judge the impact of a sensitive species is to compare
results obtained when the species is required to be present and
when the species is required to be absent. Such comparisons w
25
-------
made for daphnids in Table 18 and for sensitive saltwater species
in Table 19.
Studies of sensitive saltwater species indicated that two
kinds of species can cause a large reduction in the FAVF. Although
salmonids did not reduce the freshwater FAVFs, fish species in the
genus Menidia did reduce saltwater FAVFs. In addition, crustacean
species in the families Mysidae and Penaeidae also caused large
reductions. When two kinds of species deserve special
consideration, there are more possible combinations of inclusion
and exclusion that can be tested (Table 19).
Results of computerized studies concerning the kinds of
species that can cause large reductions in FAVFs need to be
examined closely to determine exactly which species are being
studied. For example, in the freshwater family Oaphnidae the only
genera for which many acute values are available are Ceriodaohnia.
Daohnia. and sj.mocephalus. Similarly for saltwater crustaceans,
Mvaidonaia and Penaeua are the only genera in the families Mysidae
and Penaeidae, respectively, for which many data are available.
Species within a genus probably have sufficiently similar
toxicological characteristics that generalization within a genus is
possible, but generalizations within a family might be more risky.
The theoretical bounds on the 5th percentile in Table 20
correspond most directly with the cases in Tables 18 and 19 labeled
"any family" where no consideration was given to sensitive species.
Application of the iterative numerical calculation procedure
discussed at the end of Section 6 to the values of S given in Table
4 indicated that value of S-8.2 would give an overall 95%
confidence level. The overall 95th percentile FAVFs for "any
family1* in Tables 18 and 19 are similar in magnitude to the FAVFs
in Table 20 for S-7.5 and 95 percent. The FAVFs are not expected
to be exactly the same for three reasons. First, data for
freshwater and saltwater species are considered separately in
Tables 18 and 19, but are considered together in Table 20. Second,
the FAVFs in Tables 18 and 19 answer slightly different questions
than the FAVFs in Table 20. The FAVFs in Tables 18 and 19 give
divisors such that one is 95% confident that the resulting LBFAV is
less than the FAV that would be obtained from the sample when the
8 minimum data requirements are satisfied. The FAVFs in Table 20
give bounds on the true population 5th percentile. Third, the acute
values chosen to satisfy different of the eight minimum data
requirements are not truly independent as assumed by the derivation
using the log-triangular distribution. Thus, the simulations will
not match any theoretical results exactly. Despite these three
reasons why the values in Table 20 will not agree with those in
Tables 18 and 19, the overall 95th percentile FAVFs for "any family"
are similar in magnitude to the FAVFs in Table 20 for s=7.5 and 95
percent. The overall median FAVFs for "any family" in Tables 18
and 19 are also similar in magnitude to the FAVFs in Table 20 for
S-7.5 and 50 percent.
26
-------
Similarly, th« geometric mean worst case FAVFs given in Tabl
16 are quit* close to the medians of the 95th percentiles given in
Tables 17 and 18. This is in spits of th« fact that this is a
comparison of geometric means in Table 16 with medians in Tables 17
and 18; in addition, this is a comparison of "worst case" values in
Table 16 with 95th percentiles in Tables 17 and 18. The general
agreement between the values in Tables 16, 17, 18, 19, and 20
confirms that all of these different approaches are reasonably
appropriate and that none of them give results that are "outliers11.
For example, the very large FAVFs given in Tables 18, 19, and 20
for small sample sizes are not an artifact of the analyses. These
very large FAVFs are due to a very large range of acute values in
more than 5 percent of the data sets.
The overall 95th percentile FAVFs obtained at N-8 using the
second version of the method for calculating empirical FAVFs range
from 4 to 8 (Tables 18 and 19). Similarly, the mean worst case
FAVFs at N-8 are 6 (Table 16). These values are in agreement with
the fact that in criteria documents the FAV is never more than a
factor of 3 below the lowest SMAV. (The lowest SMAV is usually
equal to the lowest GKAV.) These factors are small considering
that the extrapolations are from n-8 to the 95th percentile. The
lowest value is an estimate of the p » 100/(n+l) * 100/9 - 11th
percentile when n-8, whereas 19 values would be needed in order for
the lowest value to be an estimate of the p - 100/20 »~ 5th
percentile. These small factors might be due in part to the use o
the log-triangular distribution because other distributions such a
the log-normal and log-logistic would have longer tails and given
lower values when extrapolating from n-8 to the 95th percentile.
The ratios of the FAV for n-8 and n«18 given on page 7, however,
indicate that large factors are not necessary when n-8. This is
probably due in part to the conservatism built into the minimum
data requirements, but it does indicate that the small factors and
the log-triangular distribution are appropriate. Also, these
considerations give no support for increasing the minimum data
requirements to more than 8.
Additional work could be done to provide more definitive
information concerning the items examined in this report. such
additional work could involve the following:
1. Additional data sets could probably be added to COMBO, and
literature searches could be performed for data that could be
added to the existing data sets. All data in COMBO could be
screened using updated procedures for reviewing the
acceptability of results of acute toxicity tests. Toxicity
tests could be conducted to fill gaps in data sets that
satisfy only six or seven minimum data requirements.
2. An alternative approach could be used to study the dependence
of the FAV on the number of GMAVs by randomly selecting
samples (n>8) of GMAVs, with the restriction that the GMAVs
27
-------
each sample must satisfy the minimum data requirements. The
geometric mean FAVs for the various sample sizes from n-8 to
the maximum number of GMAVs in the data set could then be
compared. This approach would result in an FAV for the
maximum number of GMAVs that would be the same as the FAV
given in Table 4. Preliminary results obtained using this
approach with the freshwater data sets were similar to the
results reported in Section 3. The relationship between the
FAV and the number of GMAVs could be used to study the impact
of specific changes in the minimum data requirements on the
conservatism of the FAV.
3 Analyses could be performed on an updated version of COMBO
using version 2 of the empirical method to calculate overall
95th percentile FAVFs. Additional percentiles could be
calculated if desired.
4. Variance could be compared within species, genera, and
families, and among families.
5. Analyses could be performed for specific fcinds of chemicals,
such as organics.
The results presented in this report should be useful for
interpreting data concerning the toxicity of chemicals to aquatic
life, especially when not enough data are available to allow
calculation of water quality criteria using the procedures
described in the National Guidelines. An example of how the
results of these analyses might be used is presented in the
Appendix.
28
-------
REFERENCES
Call, D.J., L.T. Brooke, M.L. Knuth, S.H. Poirier, and M.D.
Hogland. 1985. Fish Subchronic Toxicity Prediction Model for
Industrial Organic Chemicals that Produce Narcosis. Environ.
Toxicol. Chen. 4:335-341.
Erickson, R.J., and C.E. Stephan. 1988. Calculation of the Final
Acute Value for Water Quality Criteria for Aquatic Organisms.
PB88-214994. National Technical Information Service, Springfield,
VA.
Kalbfleisch, J.D., and R.L. Prentice. 1980. The statistical
Analysis of Failure Time Data. John Wiley, New York. 321 pp.
Kenaga, E.E. 198.2. Predictability of Chronic Toxicity from Acute
Toxicity of Chemicals in Fish and Invertebrates. Environ. Toxicol.
Chem. 1:347-358.
Michigan Department of Natural Resources. 1987. Updated Support
Document for the Aquatic Chronic Value of the Rule 57(2)
Guidelines. Surface water Quality Division, Lansing, MI.
Rice, J.A. 1983. Mathematical Statistics and Data Analysis.
Wadsworth, Pacific Grove, CA.
Stephan, C.E., D.I. Mount, D.J. Hansen, J.H. Gentile, G.A. Chapma ,
and W.A. Brungs. 1985. Guidelines for Deriving Numerical National
Water Quality Criteria for the Protection of Aquatic Organisms and
Their Uses. PB85-227049. National Technical Information Service,
Springfield, VA.
U.S. EPA. 1987. Guidelines for Deriving Ambient Aquatic Life
Advisory Concentrations. Office of Water Regulations and Standards,
Washington, DC, and Office of Research and Development, Duluth, MN.
29
/ft
-------
I
s
> 2
Plotted Line is Fit to First 4 Points
J I I I
•1.5 -1.0 -0.5 0.0 0.5 10 15
Figure 1. Probability Plot for Chlordane in Salt Water
30
-------
i
s
GC
O o
-3 -2 -1
Figure 2. Log-normal Probability Plot of Freshwater FACRs
31
-------
s
o
in
or
O o
•2
-1
Figure 3. Log-normal Probability Plot of Saltwater FACRs
32
IT*
-------
Table 1. Sources of Data Sets
Chemical
Water*
Source Document"
Acenaphthene
Acrolein
Aldrin
Aluminum
Ammonia
Ammonia
Antimony (III)
Arsenic (I II)
Cadmium
Chlordane
Chloride
Chlorine
Chlorpyrifos
Chromium (II I)
Chromium (VI)
Copper
Cyanide
DDT
Dieldnn
2 , 4-Dimethylphenol
Endosulfan
Endrin
Heptachlor
Lead
Lindane
Mercury
Methyl parathion
Nickel
Parathion
Pentachloropheno L
Phenanthrene
Phenol
Selenium (IV)
Selenium (VI)
Silver
Thallium
Toxaphene
Tributyltin
1,2, 4-Trichlorobenzene
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
' F
F
F
F
F
F
F
F
F
F
F
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
Draft: 10-11-88
Draft: 9-21-87
EPA 440/5-80-019
EPA 440/5/86/008
EPA 440/5-85-001
EPA 440/5-88-004
Draft: 8-30-88
EPA 440/5-84-033
EPA 440/5-84-032
EPA 440/5-80-027
EPA 440/5-88-001
EPA 440/5-84-030
EPA 440/5-86-005
EPA 440/5-84-029
EPA 440/5-84-029
EPA 440/5-84-031
EPA 440/5-84-028
EPA 440/5-80-038
EPA 440/5-80-019
Draft: 9-1-88
EPA 440/5-80-046
EPA 440/5-80-047
EPA 440/5-80-052
EPA 440/5-84-027
EPA 440/5-80-054
EPA 440/5-84-026
Draft: 6-7-88
EPA 440/5-86-004
EPA 440/5-86-007
EPA 440/5-86-009
Draft: 8-16-88
Draft: 9-13-88
EPA 440/5-87-006
EPA 440/5-87-006
Draft: 9-24-87
Draft: 3-4-88
EPA 440/5-86-006
Draft: 6-16-88
Draft: undated
33
-------
2,4.5-Trichlorophenol F S Draft: 9-25-87
Zinc F S EPA 440/5-87-003
F - freshwater; S - saltwater.
Published documents are titled "Ambient Water Quality Criteria
for " and are available from the National Technical
Information Service, Springfield, VA. The drafts used were the
newest ones available as of 10-12-88.
34
-------
Table 2. Fields used in the Database COMBO
Record Wumber
Chemical
Water
F - Freshwater
S - Saltwater
MDR Codes' for freshwater species:
A -A salmonid.
Bi - B13 * Fanilies other than Salaonidae in the class
osteichthyes.
C >A species that is in the phylum Chordata but is
not in the class Osteichthyes.
Dl » A daphnid, i.e., in the family Oaphnidae.
D2 * Other planktonic crustaceans.
E -A species that is a benthic crustacean.
Fl - F7 * Orders of insects.
Gl - G6 * Families in phyla other than Arthropoda or
Chordata, e.g., hydra, planana, rotifers, worms,
snails, clams, mussels, bryozoans, etc.
X - A species that is nonresident and therefore should
not have been used in the criteria document.
MDR Codes' for saltwater species:
Families in the phylum Chordata.
The family Mysidae.
* The family Penaeidae.
• Other families in the phylum Arthropoda.
Families in the phylum Annelida.
* Families in the phylum Molluscs.
Families in the phylum Echinodermata.
A species that is nonresident and therefore should
not have been used in the criteria document.
Genus
Species
Life Stage
Remark
A » Mo remark (default).
B * This acute value is in the source criteria document but
was not used in the calculation of the FAV in the criteria
document or here for the reason given in the criteria i
document.
E » This acute value should not be used because it should not
have been included in the criteria document. (This re
35
Ml
Ml
N2
N3
PI
51
Rl
X
- M22
- N26
- P8
- S13
• R4
*
—
m
—
»
at
m
m
-------
was usually because the species is not a resident
species.)
G - This acute value is a "greater than*1 value.
H - The hardness, or another pertinent water quality
characteristic, of the dilution water is unknown, so this
acute value cannot be adjusted and therefore was not used
in calculations in the criteria document or here. (This
remark is used only with those chemicals for which the FAV
is calculated from adjusted acute values.)
L - This acute value is a "less than" value.
M - This acute value was not used in the calculations in the
source document or here because an acute value for this
chemical is also available for a more sensitive life stage
of this species.
R » This acute value has been changed from that in the source
criteria document, based on either (a) the reference used
in the source document or (b) a newer reference for the
sane value.
Technique
F » Flow-through
S - static
Measurement
M - Measured
U * Unmeasured
Acute Value
' MDR Codes were designed so that an algorithm could be written
to determine which of the minimum data requirements concerning
acute toxicity were satisfied by any group of species.
36
-------
Table 3. Data Sets Not Satisfying the Hininua Data Requirements
Chemical
Hater Requirement(s) not Satisfied1
Aldrin
Chlordane
Chlorpyrifos
DDT
DieXdrin
Endosulfan
Endrin
Heptachlor
Lead
Lindane
Toxaphene
Aldrin
Chlordane
Fresh
Fresh
Fresh
Frash
Fresh
Fresh
Fresh
Fresh
Fresh
Fresh
Fresh
Salt
Salt
a
a,b
c
a
a
a
a
a,b
b
a,b
a
d
e
1 The minimum data requirements not satisfied are:
a. A species in any order of insect or any phylum not already
represented.
b. A species in a phylum other than Arthropoda or Chordata.
c. A benthic crustacean.
d. A species in the family Mysidae or the family Penaeidae.
e. Several requirements were not satisfied.
37
-------
TABU 4. VALUfS OF PAV AND S FOR DATA SETS IN C»TE«A DOCUMENTS
Crtwtml
ActnipMMiw
Aootan
Aluminum
Ammontt
AmfflOflK
Animony (III)
Anpmony (III)
AfMrae (III)
Arune (111)
Cadmium
Cadmium
Chtondt
Chtonrw
Chtonn*
Chlorpynto*
Chromium (III)
Chromium (VI)
Chromium (VI)
Coppw
Coppar
Cyand*
Cyandt
DOT
Dtfdrm
2 4-0«»tf)ylph»nol
2 4-Oimwhylpfwnol
Endosultan
Endnn
H«ptacrtor
WaMr
F
F
F
F
S
F
S
F
S
F
S
F
F
S
S
F
F
S
F
S
F
S
S
S
F
S
S
S
S
Numb*
of QMAVt
9
12
14
34
17
9
n
14
11
44
33
12
28
21
12
18
27
21
41
20
IS
a
16
19
12
9
12
19
'8
S
482
383
877
389
237
12.80
228
409
774
1938
714
229
571
329
571
653
440
299
190
829
183
954
558
562
359
653
450
4 11
1398
FAV
8002
615
1496
070
051
176
2959
718
137
893
8509
1720
3832
2524
002
1968
31 49
2158
1809
533
€268
203
0 14
062
2663
54S
r-i ^T
QC3
r, /•-
* *
38
-------
TABU 4. (CONTWUIO)
Lwd
Undm
Mtfcury
Mweury
Mtfcyi Pvifcion
Nidwl
NicMJ
Paratmn
PcnCKMoraplwnol
Olh*lt!8M4lbW4Wli^BhnAl
r*VlllflUHUIU|JIIVtlUI
PtenaftttWf*
PnSrttt^ftfBfW
B8*AMMi
rTiwnoi
Satanium(IV)
S*toraum(IV)
S^fwm(VI)
SiKw
Sdvw
Hn«um
ToKflpnClW
TnbwtKlbn
TribuiyMn
124-TricMarabMMm
l24>TricMorotan»M
2.4 s-TriehtaraphMoi
2 4 S-Tnchtoraphwiol
Zinc
Zinc
S
S
F
S
F
F
S
F
F
F
F
F
S
F
F
S
F
S
F
S
f
3
F
S
F
S
11
17
20
29
22
18
21
30
33
17
8
10
28
22
IS
11
t8
19
•
IS
8
18
14
IS
10
10
35
28
220
1707
1193
357
1509
711
951
1781
1313
391
276
7 19
519
905
330
1459
389
202
982
348
674
340
251
323
284
067
632
304
267
021
486
413
063
1577
157
012
1095
2506
5962
1546
7445
370
588
2565
1 83
1460
4665
042
030
053
689
280
199
473
127
'93
FAV • Final Acu» Vtiut
S « Stop* fccttr
GMAVi • G«m* M«an Acutt ValuM
39
-------
TABLE «. UtPWOIMCI Of THf «V AND 3 ON THE NUMBER Of GMAV» IN PRESH WATER
Ctwmcrt
J. —^^t^^^^t^^t^
ACiWtBipnywvv
Acretan
Acreton
Aluminum
AJummum
Aluminum
AmmoniA
Affifnontt
Ammontt
AfflvnoniB
AmmonA
A^Rn^OAlB
AfAfltOAtt
Amman*
AmmoAtt
AflimflfiB
AlfWnOfHB
Ammonia
Ammonal
Numtar
ert
GMAVi
a
a
10
•
10
12
a
10
12
U
16
18
20
22
24
26
2t
30
32
Mn
189
257
32S
ai9
990
11*2
031
036
040
048
049
053
056
OS*
060
062
0<4
068
068
—find Ac
MOT
189
664
523
2644
1887
1881
067
071
074
070
068
070
067
067
066
068
088
068
070
1M V*kM_
Max
189
2367
2153
14132
14571
10709
146
157
120
130
1 16
1 17
121
121
094
1 14
098
084
085
Std.
SSL
000
684
337
4678
2894
1847
029
027
024
022
019
018
015
012
009
011
008
006
005
Mm
284
274
299
130
1 43
244
033
022
043
047
044
048
034
033
140
058
151
236
244
1
MtM
284
537
621
473
665
734
197
191
173
209
235
230
275
298
329
323
340
355
372
Max.
284
802
• 40
804
889
962
481
434
502
435
446
466
436
454
463
445
457
431
465
Std
&CL
000
184
129
253
175
125
120
104
1 14
102
096
1 12
096
062
076
080
065
055
047
Antmony (IB)
142
168
192
25
1253 1273 121
017
ArMne(IK)
ArMne(IN)
Amnc (III)
Cadmium
Cadnurn
Cadmium
Cadmium
Cadmium
a
10
12
a
to
12
14
16
152
213
260
052
080
149
i 86
336
212
281
441
362
530
657
680
841
286
630
674
3611
4296
7061
4469
6296
45
97
178
822
1043
H98
945
1196
956
350
381
352
296
190
1 79
215
1046
944
673
927
844
823
891
829
11 49
1223
1277
1713
1869
1529
18S3
1542
054
1 86
332
342
343
323
3 11
298
40
-------
TAMJl (COMTMUtJD)
CatiMum
Catttuii
Caawuiii
Cadmium
Cadmium
Cadmum
Cadmium
Cadmium
Cadmwn
Cattmum
Cadmium
Cadmium
Cadmium
Cadmium
Chtond*
Chtondf
Chtond*
CMOTW
Chtonm
CNMta*
Chtorirw
CNMfcw
CMonw
CMorim
ChMM
CMom*
CMonw
Chromium (III)
Chremum (ill)
Chromium (III)
Chromium (III)
Chromium (HI)
11
i>0
;a
.24
26
26
30
32
34
36
31
40
42
44
a
10
12
•
to
12
14
It
It
20
22
24
26
«
to
12
14
16
351
410
471
sst
642
721
7W
It2
939
1020
not
1197
12.93
1394
231
270
330
1762
21 tt
23.4S
25.73
28.08
2987
31 «
33.33
34 9t
3664
1064
1232
1383
ISM
1744
066
932
911
1062
1048
1101
1136
1133
not
1172
12.70
1310
1357
1412
346
369
365
4019
40 tt
42.31
39 2t
3726
3a\5t
36.3T
3173
3706
3758
1669
2013
1951
2020
1934
57 Tt
4633
3674
6771
4167
4509
4002
3923
1910
3S33
22.46
2307
2450
2399
563
635
617
6914
7144
7111
7149
7300
72,07
72,50
7017
65 55
6546
2431
266S
2175
3095
3316
1147
937
911
12.06
696
t77
723
510
326
462
30
299
247
t 19
140
12t
104
1727
1901
1t.9t
It 10
1712
1644
14 tt
11 It
662
566
563
647
663
671
473
233
157
084
067
091
094
191
197
961
209
1015
1040
1068
1090
366
405
440
031
02t
031
033
029
022
032
034
079
063
342
378
4 11
441
470
855
931
939
924
1028
1024
1063
1108
1302
12.84
1363
1415
1451
1498
529
541
549
176
173
183
190
247
260
304
391
451
513
416
456
512
550
612
1540
1702
1693
1975
1589
1582
1642
1477
1871
1552
1594
1634
1580
1616
668
739
677
500
488
530
522
491
522
549
546
569
590
540
S97
649
673
699
307
293
318
359
279
320
299
208
143
224
148
1 15
090
048
107
0
1 41
155
168
183
160
190
188
155
1 30
092
061
073
079 »
063
41
-------
TABLES. (CONTWUED)
Chromium (VI)
Chromium (VI)
Chremunt (VI)
Chramum (VI)
Chramum (VI)
Chromium (VI)
Chramum (VI)
Chromium (VI)
Chromium (Vt)
Chromium (VI)
Copper
Copper
Copper
Copper
Copper
Copper
Copper
Copper
Copper
Copper
Copper
COM^V
^TWJ^^BW
Copper
Copper
Copper
Copper
Copper
Cyano*
Cyanide
Cyanide
Cyar«d»
2 4-Oirrtfihylphenol
2 4 Oirwhytohenol
•
10
12
14
16
11
20
22
24
26
B
10
12
14
16
18
20
22
24
26
28
30
32
34
36
3*
40
8
10
12
14
a
10
079
069
140
301
639
829
1674
25 OS
28 75
32.96
279
619
652
870
840
1079
11 3t
1208
1287
1315
1482
1582
1«2«
1708
1768
1827
1887
3784
5581
6242
6674
1594
1345
418
764
1156
1761
2980
2837
3158
3436
3750
41 14
888
1147
1180
1316
1415
1433
1555
1601
1673
1743
1734
1788
1822
18.72
1891
1683
1938
S736
7021
S98S
7094
1S44
1924
2136
5889
7088
83.04
9719
11462
12640
72.27
52.67
4639
tZ78
1S4S
)747
1793
1916
2164
2332
2205
2348
2806
2573
2588
2655
3008
2906
2654
2340
9059
9721
9951
3813
2088
2366
42
400
699
1123
1359
1794
1S4S
1180
725
SIS
357
231
200
261
257
281
274
305
279
273
317
248
258
201
287
231
149
138
1303
1379
961
502
175
295
1582
344
374
402
428
452
475
497
519
539
314
339
297
354
377
359
377
394
457
427
492
458
525
487
500
571
585
1 44
1 09
1 19
128
454
434
2015
1817
1584
1502
1307
1136
1000
948
620
687
621
610
628
603
624
603
595
564
587
576
575
562
581
578
585
584
595
291
1 91
1 87
1 75
517
506
2612
2743
3217
2945
2971
3201
2959
2032
1814
1536
1350
807
925
862
886
924
980
940
954
1081
10 11
762
891
849
861
668
706
481
297
274
236
609
633
237
538
628
669
557
539
493
458
465
407
1 72
096
1 11
1 13
1 17
140
1 40
1 26
1 14
126
1 04
070
081
064
057
032
029
080
055
049
036
045
c?o
-------
TAIL! I (CONIMUB}
MMUy
Mamiy
Mwawy
MMwy
Manwy
Mwouy
Mwony
Mwoiry
Mwoury
Mwewy
MiMhjn PWVUOA
Mit>y( pwfMon
Mvhtiyl pjrutttOft
t^K^hhJ J**»^a4k«*«4k
MMnyi pwntvn
Mttiyf parafwn
Mttyt paratfuon
Mttyi oarattion
Niekal
NfeM
U-^£~Ai
NMMI
Ui^^^^
HHCMI
NMM
ft, [•fr.l i
pnnon
M^^^^^^M
i" ••NUH
«h^^^^^^^^.
rtnnoii
Pannon
Paretwn
Parstven
PantMn
A^^^^k^B^
rtranon
Pantmn
Pantvon
DAAM^MM^M
faranon
•
!0
12
14
1»
11
20
22
24
26
•
10
1£
t«
1«
111
20
H
10
12
14
If
•
10
12
14
10
18
20
22
24
26
2S
041
017
097
12S
164
202
212
343
404
4M
010
014
oia
023
030
040
050
707
Ml
too
1290
13M
000
ooa
016
021
021
032
034
040
047
049
053
077
115
111
2.M
253
303
344
3M
434
472
034
043
049
OS4
080 _
055
057
832
1002
11ft
1311
1450
"Off
026
031
040
043
046
051
051
053
054
OSS
150
181
2.20
2.53
301
3 69
433
455
522
547
1 18
195
216
241
264
289
262
060
1100
1256
1420
1599
060
069
075
081
069
067
093
095
096
065
067
025
030
034
031
032
029
0.27
0.28
029
020
049
064
079
064
085
069
053
57
60
51
39
32
017
016
014
013
012
013
Oil
007
007
003
003
763
665
766
622
675
926
973
1016
1062
1103
713
451
490
526
560
593
623
469
S41
559
632
673
361
200
216
234
249
263
277
269
302
314
325
12.32
1Z33
12.22
1164
1155
1167
1162
1174
1162
1152
969
1005
1013
1051
1123
12.43
1329
663
690
699
697
70S
1022
731
644
S72
S5S
570
487
466
422
413
374
1530
1586
1664
1787
1783
1689
16.92
1739
1716
1604
13.67
1111
1642
1764
1829
1711
17M
767
790
8.27
921
847
2523
2171
1353
1303
1472
995
939
923
954
859
617
166
201
232
235
229
212
202
172
166
1 11
296
316
331
359
355
2
22
047
070
060
063
054
599
401
213
< 90
217
1 75
1 77
1 84 »
1 59
1
43
210
-------
TAOJS. (CQOTlNUfD)
P««U*ju*h*nul
P«,eu*«phenoi
Per*******
Pen^-or^o.
PenttMonenenol
HaniM i lil i li 1
Pvftttcntoropne9ooi
PeneKMorophenol
W«KtUWO***al
ptfiAbnttro^rwuH
_
*"
Phenol
Phenol
Phenol
Phenol
Phenol
Phenol
Phenol
Phenol
Phenol
Phenol
Selenwm(IV)
SetonunKIV)
SotorwHHIV)
So*M*MIV)
SetommdV)
Selenium (IV)
Selenium (IV)
Setonum (VI)
Selenium (VI)
6
10
18
22
24
**
26
9B
e.0
30
32
8
10
12
14
16
18
20
22
24
26
8
10
12
14
16
18
20
8
10
172
239
296
* ft*
J 96
42S
500
570
/w
646
7 14
f »•*
790
• At>
9 9V
952
1040
2899
3294
3834
4322
4789
5291
5508
6107
6447
6909
50
99
136
164
196
260
320
583
18 12
1759
1354
12.75
14 CM
1* 99
12.86
1199
11 M
1 1 «•?
1149
1207
tm-Wf
1079
1ft &4
IB *•
1032
1075
5191
5752
5979
6170
6398
6466
6721
7066
7233
7314
272
309
316
336
338
378
346
3703
31 19
5639
5584
3468
mm Am
9645
5747
4308
37 as
3815
10 rm
JC.Uel
3081
1 4 *¥t
31 23
3163
2819
6500
7424
7901
8472
9067
9549
10070
10832
10672
11133
127«
1322
1368
1407
627
1485
713
13652
17378
11 07
1191
1026
Irt B%9
1Q O*
1189
1143
fl_
1000
Q Aft
» VIP
850
7 K9
' 52
484
305
1101
1307
1450
1510
1443
1301
1226
1222
1101
891
357
358
329
232
135
226
43
5380
6435
025
027
1 14
n 94
U 21
046
136
I «A
I ^v
152
1 U
1 OV
164
4 ffk
i 7W
176
186
278
307
334
359
382
404
424
444
483
461
ISO
185
180
193
515
217
573
961
1062
304
361
412
3^M
96
469
537
• id
IV
629
796
• •*»
• BO
1096
1119
433
442
449
467
471
486
502
497
487
503
646
650
666
713
757
769
865
1336
1407
1046
1040
1094
11 80
12.45
1244
13 M
1369
1483
i *• id
13 34
1357
1346
S67
615
658
695
690
744
711
686
715
666
1318
1285
1164
11 70
1247
11 10
10 13
2066
1749
220
303
325
334
388
413
e> 2Q
429
4 14
A Aft
4 OB
276
1 90
059
073
078
083
077
079
080
075
059
056
. 290
271
209
1 58
121
i 39
035
329
211
44
511
-------
TA8U1 (OOM1MUB)
SOW
SiMr
saw
SOW
saw
Tffeuyttn
1 2 4-TrichlorBe*ttBf»
1 2.4-Trichtarebenatne
1 2 4-TrfcMarQeenatn*
2,4 S-Tnehlorophtnol
Zinc
Zine
Zinc
Zkic
Znc
Zinc
Zine
Zine
Zinc
Zne
Zine
Zine ~
Zine
Zine
I
10
ia
u
id
ii
«
10
12
1
8
10
12
14
16
ii
30
22
.M
26
26
30
32
34
on
104
144
168
160
026
666
721
791
1724
1674
3138
4019
4908
5693
6739
7692
•6.77
9701
10679
12171
13420
14646
T6142
140
170
167
208
216
029
1156
1017
930
1813
52.70
•028
•617
10387
11394
11353
12383
128,44
13339
14)84
14738
132 30
16166
16388
1092
666
382
320
327
030
2014
2037
1724
187 5
14668
16993
19269
21722
24311
27049
30043
32$ 89
38743
39108
30093
422.28
28882
30824
132
076
037
028
023
001
424
364
282
74
4008
5242
5874
«731
6757
7010
68.75
7047
88.11
6888
52.28
4472
3636
1689
150
130
142
152
182
674
048
053
108
257
599
631
671
717
763
607
648
888
928
962
997
1038
1142
1178
499
419
371
347
325
700
170
220
267
281
987
901
957
989
1041
11 10
1142
1198
1177
12.21
1242
1241
1228
1220
717
718
703
601
527
757
325
359
384
321
1596
1697
1506
1554
1369
1413
1408
1470
1517
1570
1619
1664
1577
1602
1 25
131
1 13
097
070
039
089
069
070
031
2,39
210
t
2
185
1 86
200
185
190
197
182
169
1 44
080
PAV . fin* Aeu» Valu*.
SBStoeetacfcr
QMAVt . Genus Mean Acu» Value*
45
-------
TABU 6, DIWWOfNC! Of THE FAV AND S ON THE NUMBER OF QMAVt IN SALT WATER
Ammonia
Antmony (III)
Antmony (III)
Araarae (III)
Anarae (lit)
Cadmium
Cadmium
Cadmium
Cadmium
Cadmium
Cadmium
Cadmium
Cadmium
Cadmium
Cadmium
Cadmium
Cadmium
Cadmium
Cnloflna
Chionna
Chionna
Chionna
Chionna
Chionna
Chionna
Numbar
of
GMAVi
•
10
12
14
16
»
10
8
10
8
10
12
14
16
18
20
22
24
26
28
30
32
8
10
12
14
16
18
20
— Final AnM Viki*.
Mbi Maan Max.
026
033
036
040
044
1980
2945
50
104
a
12
14
20
24
34
34
SO
56
63
72
71
84
969
1141
1606
1806
2003
2163
2333
050
051
048
048
047
2719
3133
15S
153
63
55
55
62
56
66
76
75
77
80
81
82
89
1530
1628
1984
2139
2262
2375
2489
091
075
079
076
076
3407
3S47
1041
360
2507
1246
624
1174
233
314
275
329
179
203
140
116
123
2023
2154
2350
2517
2551
2697
2766
SVL
fiBL
015
013
010
008
006
430
121
213
68
390
180
123
142
34
39
43
40
22
29
14
10
8
230
231
235
1 89
1 61
1 36
1 10
Mm
029
028
031
033
039
159
175
276
574
090
247
241
317
387
410
371
357
469
487
763
769
814
2 11
233
253
272
299
306
322
M*an
1 60
1 73
187
199
216
222
163
682
727
796
887
947
904
922
934
871
863
851
835
835
824
819
437
396
391
364
356
355
347
e
Max
490
404
340
318
262
417
242
1085
895
1463
1616
1584
16 13
1546
1750
1636
1465
1424
1379
11 60
1200
955
606
696
589
541
535
542
456
SUL
5SL
064
076
063
047
333
096
020
1 51
085
342
316
295
291
237
233
247
215
169
1 48
108
078
022
769
1 01
088
072
066
069
050
46
-------
TABLE t (CQN1MUB)
Chtorpimtoi
(Mo**.
Chromium (VI)
Chromium (VI)
Chromium (VI)
Chromium (VI)
Chromium (VI)
Chromum (VI)
Chrarmim(VI)
Coppor
Coppor
Coppw
CfifllMf
^"^^^•*
Coppor
Coppor
DOT
DOT
DOT
DOT
OMttm
DMttvt
OMOin
DM*M
Owttto
°-1*"
24-Owwtnylphtnflt
EndoMMn
EndosuHon
II
10
II
19
12
14
16
18
;to
6
10
12
14
16
18
8
10
12
14
8
10
12
14
16
16
6
8
10
002
009
515
1139
1490
1890
1904
1994
2118
209
267
380
467
573
660
008
007
009
019
017
029
092
099
049
053
416
001
002
002
009
1999
2011
2290
2199
2291
2229
2169
1017
789
901
815
633
778
009
010
012
013
042
049
047
051
059
080
532
002
003
002
003
2699
2770
2999
2949
2979
2999
2943
5889
2911
3448
3856
2797
3025
010
011
012
013
339
279
207
133
1 41
095
940
009
006
000
000
597
449
439
429
409
392
179
1099
981
»90
893
829
614
002
001
001
000
099
039
022
011
012
005
161
002
001
413
457
109
071
077
099
099
093
099
189
139
150
161
172
182
406
449
499
524
162
216
339
402
428
547
567
374
4 14
413
457
303
236
192
200
201
2.08
245
510
583
535
592
631
770
574
564
545
550
628
659
653
628
635
589
639
631
521
413
457
854
698
401
450
373
334
289
11 15
1151
1131
11 46
1052
986
878
955
903
661
948
962
919
911
876
776
'58
926
813
000
000
192
1 14
076
072
072
068
038
226
262
290
277
274
2 '
1 54
1 25
083
037
163
139
121
1 15
1 IS
081
051
i 72
1 32
47
-------
TABLE*. (CONTINUE)
Endrtn
End*
EmMn
Endrin
Endm
Efldm
*p*e*or
H«piK**v
H«KM**r
HwpOENor
H^Nor
LMd
LMd
Lndan*
Undtnt
Undm
Lndm
Undw»
Mtrcury
Mtreury
Mweury
Mweury
M»o*y
Mwaury
M*reury
Mtrcury
Mercury
NK**
Nttfcrt
Nicfcrt
Ntcfc*
6
10
12
14
16
18
6
10
12
14
16
8
10
8
10
12
14
16
6
10
12
14
16
18
20
22
24
a
10
12
14
001
001
001
002
002
003
001
002
003
004
008
236
260
004
006
011
015
019
090
OM
OM
163
I7f
261
300
320
341
755
1242
2362
3682
001
002
002
003
003
003
003
004
004
OOS
006
306
297
016
016
019
016
021
206
2.54
269
301
341
369
375
364
401
5761
6663
7622
10167
002
003
003
003
003
003
2.30
209
229
049
064
461
434
452
464
475
4M
459
1134
1317
8M
637
6M
947
572
545
671
24005
31144
40405
41061
000
001
001
001
000
000
052
042
025
006
010
67
56
149
152
146
097
075
260
254
1 16
104
090
126
052
043
050
7212
8569
7650
8956
629
305
332
35*
379
401
306
2M
267
670
655
097
108
105
1 16
126
136
144
250
216
156
187
1 78
186
196
207
216
906
673
731
786
939
720
663
542
459
4 11
12.75
1328
1491
1564
1606
216
201
831
964
1076
1391
1544
665
617
523
544
479
439
425
400
376
1401
1374
1342
1233
12.24
1159
1137
1023
1071
500
1891
1901
1927
1834
1815
333
253
1624
1683
1660
1693
1706
1287
1233
1127
1352
1262
817
820
792
795
2026
2221
21 29
21 30
1 47
197
233
239
185
029
377
347
250
190
184
074
046
456
451
453
315
260
235
246
219
213
195
146
1 49
1 20
121
269
293
358
373
-------
TABLE*. ('
..•*.«
rMM
1*1^^1
PHOWI
NIcM
D^^^^^ri^h^M^M^^^flM^
Pvn^BnwpnQTVfl
PwittcMoropnVAQi
P^niKraorapfWioi
ff«mlmrtUnmnluuwti
*•_»-.
SMMMm(IV)
S*tonun(IV)
SrtMum(IV)
Satenwm (IV)
Sifcw
Sriwir
S*«r
Silver
S*«r
Sfr-r
To*apt»n»
TM*HW»
TO*«*OT
Touptant
TribuiyMn
TribuiyiM
T*«*H»
Tnbulyltn
Tnbutytm
14
16
20
8
10
12
M
141
16
8
8
10
12
14
8
10
12
U
16
16
6
10
12
14
8
10
12
14
16
5952
7899
11223
623
1537
1754
2S3B
4K9 04J
3494
1162
224
283
504
560
339
407
582
544
714
1344
012
023
034
044
OOS
012
019
023
026
12053
13831
15016
25.27
2416
33-96
37 an
*pf W
3437
12.72
523
602
634
592
720
917
1153
1344
1401
1444
044
050
050
054
024
027
031
024
028
42536
41288
26775
7809
5817
5951
A.1 14
&9.9B
4414
1324
903
928
952
887
2101
1684
1748
1809
1602
1780
170
139
102
081
128
064
135
059
052
7167
5110
2Z24
1134
744
714
A 9S
• •*
199
057
187
167
152
60
436
356
291
228
139
094
024
017
015
012
027
015
021
009
006
836
884
930
179
198
215
231
» *r •
244
650
10S
1 16
124
135
139
154
167
180
191
202
183
202
220
236
120
323
1 44
445
473
1138
1098
979
504
440
383
3 12
w ••
279
680
261
221
213
260
507
404
327
177
244
222
447
418
376
342
530
538
561
607
631
2064
1946
1483
1047
749
786
787
' 9'
409
752
752
650
353
320
965
969
1073
1033
983
346
1127
772
628
453
11 44
930
784
784
834
3U
310
157
167
184
1 48
1 11
064
034
174
124
068
037
2
263
253
1 97
1 19
053
1 59
120
102
063
'95
1 44
1 42
097 »
072
49
-------
TABLf «. (CONTMUED)
1
1,2.4-TneNorob«uwi»
1 2,4-TricMorataruww
I
10
12
14
140
183
230
210
238
271
292
300
358
300
tot
431
57
S6
47
30
290
329
3S0
3S5
424
422
420
441
626
850
56«
500
093
0«3
036
022
379
4«0 30
001
134
203
04S
Zne
Zinc
Zinc
Zinc
Zinc
Zinc
Zinc
Znc
Znc
Zinc
8
10
12
14
19
18
20
22
24
26
1003
3030
7806
9980
11867
12400
13279
14101
14935
16472
12050
14036
15707
18089
161 11
18890
17055
18911
17109
17186
43794
40000
44589
51541
42512
40075
38900
39720
27007
20079
90,32
9098
7331
7020
*•
5770
5407
40M
3410
2410
1240
120
101
157
160
1 70
190
200
209
420
465
502
409
456
450
470
405
490
482
492
404
12.12
1107
837
765
604
724
816
796
784
594
200
193
133
1 IS
1 12
104
102
085
075
037
FAV . Rrwl AcuM Vakw
S» Stop* tear
GMAVt • OWMM MMR Acuto V*MM
50
-------
TABLE 7 WOW CASE ACUTC VALUES
Chemical
Acenapnthene
Acrolein
Aluminum
Ammonia
Antimony (III)
Arwrae (HI)
Cadmium
Chloric*
Chlorine
Chromium (III)
Chromium (VI)
Copper
Cyanide
2 4 Dunethylphenoi
Mercury
Methyl Parattnon
Nickel
Paratruon
PentacNoropnenot
Phenanthrene
Phenol
Selenium (IV)
Selenium (VI)
Silver
Thaftum
Tnbutyrtn
i 2 4-TnchtaroDenxene
2 4 5-Tnchtorophenol
Zinc
Metn Sample Size
PAV»
^^^•••i
1(0
6M
2644
OIJ7
1M
212
362
3*6
4019
1880
4ia
888
5736
1(144
077
034
«32
C 11
1759
SSI 62
5191
272
ara
140
4486
330
1156
1813
5270
\
41000
5920
79900
22.80
25100
97000
14850
1308S
960
71064
1870000
10242
2490
67600
2100
15200
43978
10000
43921
1190
2250000
203000
442000
3160
1460000
22740
50000
3060
144928
2
2090
510
55600
1141
25700
41760
12220
9466
710
43100
176000
9300
2326
62500
2000
14800
40460
5230
18150
1150
122000
§3000
193000
580
229000
6630
23800
2660
88964
3
2040
366
50000
800
25700
28130
8400
6743
400
17776
170000
6200
1800
36300
2000
5300
30200
3700
1486
490
100000
42500
66000
280
120000
1020
21200
1340
27678
4
1720
180
49800
799
25700
24500
8100
6222
390
16371
140000
6056
507
33400
1000
37
21565
2650
1141
419
100000
35000
63000
270
2700
550
4333
1268
20200
5
1700
93
40000
494
25700
22040
7921
4039
258
15600
135000
5869
432
19300
420
33
15734
2130
663
375
67500
31400
47000
241
2300
540
3900
672
19037
6
670
90
38200
413
21830
14960
5395
3563
179
12436
69000
1990
426
18100
180
IS
14300
48
SOS
234
51100
12500
20000
60
2200
390
3400
611
18400
_7m.
460
74
23000
276
16460
5278
400
2960
102
12400
40600
636
364
9200
SO
9
13000
32
308
126
25000
4300
5300
55
1800
370
3010
336
2694
3
240
61
22000
262
500
874
133
2540
34
3200
3490
60
350
4600
-
2259
2
30?
96
11600
3870
'60
53
00
3 SO
: AD
•• — l*t
»-*•*
5-i"
'Value* tor phenanftrene thallium and sibutyrtn are from Table 4, all other values are from Table 5 for N-8
FAV « Final Acute Value
51
-------
TABLE* WOfVT CAK FAVFt
ChamM
Maan
Samp*
Ac******
Acnaw
Alumnum
Anwianai
Antmenytlll)
Ananc(ltl)
Cadmium
Chtonda
Chtoma
Chramum (lit)
CtwmwMVI)
Coppar
Cyarada
24-0«naftylpha«ol
Manwy
Maftyl paraffMon
PamMon
Phanantrana
Pfwnol
Satonum(IV)
SatonumfVl)
Sfear
Thatwrn
Trfbu^tti
1 l+TricMarafeanana
245-TncWorephanrt
Tine
AT* maan
Qao maan
FAV*
189
694
2944
067
189
212
362
346
4019
1889
419
888
5736
1844
077
0 34
832
011
1759
5992
5191
272
3703
140
4996
029
1159
1813
5270
1
217
891
30
34
153
459
4044
39
24
38
447159
1154
43
37
2732
53
89909
2497
19
433
749
11939
2262
31192
779
43
17
2750
22184
541
JL
11
77
21
17
153
197
3373
27
18
23
42089
1048
41
34
2602
49
46446
1032
19
24
196
5212
'399.
4899
227
21
15
1699
5295
235
3
WHMMMP
11
55
19
12
153
133
2319
19
10
9
40651
699
31
20
2602
1 ** 4 A
15616
36
32859
84
6
19
159
1792
200
2591
35
16
7
525
3471
134
4
9
27
19
12
153
116
2239
18
10
9
33477
683
9
18
1301
109
26
23633
65
7
19
129
1701
. 192
59
19
4
7
383
2219
77
5
9
14
15
7
153
104
2197
12
6
8
32282
661
a
10
546
A9
97
19
18916
38
6
13
115
1269
172
49
18
3
4
361
1969
60
8
4
14
14
6
130
71
1489
10
4
7
16499
224
7
10
234
44
17
426
29
4
10
46
540
43
47
13
3
3
349
700
36
7
2
11
9
4
110
25
111
9
3
7
9708
72
6
5
65
27
16
264
18
2
5
16
143
39
38
13
3
2
Si
372
'9
|
1
9
8
4
3
4
37
7
2
2
836
7
6
3
11
20
3
16
17
2
2
14
21
36
2
2
2
1
10
38 ,
6
V«UM art tram Tabta S tor N » 8
FAVFi . final Aeuta Valua Facto*
FAV . final Aeu» VakM
52
-------
TAMJi.
DATA flKUMEMEMTa RANKED BY ACUTE VALUE
*****
fca^phfla.^
J^^^
Alunwni
AfWitw
AMknanyp)
Afwnc(HI)
Cadiwim
CMendt
CNonrw
Chromium (lit)
Chromium (VI)
Coppar
Cyan*
2,4-0«iattiy
-------
TABLE 10 WORST CASE ACUTE VALUES WHEN OAPHNIOS WERE REQUIRED
Chemical
Sample Size
Ac*nai»*em
Acrolem
Aluminum
Ammonia
Anomony (III)
Arsenic fill)
Cadmium
Chloride
Chlonne
Chromium (III)
Chromium (VI)
Copper
Cyanide
2 4-Dimethylphenol
Mercury
Methyl parathion
Nickel
Paratoon
Pentachtorophenol
Phenanthrene
Phenol
Selenium (IV)
Selenium (VI)
Silver
ThaSum
Tnbutyiin
1 2 4-Tnenlorober«ene
4 x ncnnrepnenoi
Zinc
FAV*
189
664
2644
067
168
212
362
346
4019
1889
418
888
5736
1844
077
034
832
0 11
1759
5962
5191
272
3703
140
46 as
030
1156
1813
5270
1
41000
93
38200
494
18460
5271
133
3583
45
17778
3490
5964
1800
4800
877
9 1
2259
1 8
306
1150
109000
3870
5300
55
2200
663
50000
2660
547
2
2000
93
38200
494
18460
5271
133
3563
4$
17776
3490
5964
1800
4800
877
91
2299
18
308
1150
109000
3870
5300
55
2200
663
23800
2660
547
3
2040
93
38200
494
18460
5270
133
3583
43
17776
3490
5964
1800
4800
. 877
91
2259
18
308
490
100000
3870
5300
55
2200
102
21200
1340
547
4
1720
93
38200
494
18460
527*
133
3583
45
16371
3490
5964
507
4800
877
91
2259
18
308
419
100000
3870
5300
55
. 2200
55
4333
1268
547
5
1700
93
38200
494
18460
5270
133
3563
45
1S800
3490
5964
432
4800
877
91
2259
18
308
375
67500
3870
5300
55
2200
54
3900
672
547
6
670
90
38200
413
18460
5276
133
3583
45
12436
3490
5964
426
4800
877
91
2259
18
308
234
51100
3870
5300
55
2200
39
3400
611
547
7
460
74
23000
276
16460
5278
133
2950
45
12400
3490
5964
364
4800
877
9 1
2259
18
308
126
25000
3870
5300
55
1600
37
3010
336
547
B
240
61
22000
262
500
874
133
2540
45
3200
3490
5954
350
4800
877
69
22S9
18
307
96
11600
3870
760
S3
100
05
2<40
260
547
'Values tor phenanthrene thallium and tnbutylan are from Table 4 all other values are from Table 5 tor N«8
PAV » Fmal Acute Value
54
-------
TABLE 11 WOflBJTCAaf FAVFs WHEN OAPHNIOS WERE
ChemtCBJ
Aeenapntten*
Acroiein
Aluminum
Ammotw
Anttnony (III)
Arsenic (IN)
Cadmium
Chtonde
Chlorine
Chromium (Hi)
Chromium (VI)
Copper
Cyanrfe
2 4-Dunethylpnenol
Mercury
MeBiyl parathwn
Nicfcet
Parathwn
PenaeMorophenoJ
Phenol
Setonum(IV)
Sefenum(VI)
Silver
ThaJbum
TrfbutyMn
1 2 4-Trtcftlorobenane
2 4 5-TncMerephsnot
Zinc
Arith mean
Geo mean
•Vali i*Mt btr nhAnsnttMmMA 1
189
664
2«44
067
1«i
21?
362
34«
4019
1889
411
868
5736
1844
077
034
832
0 11
1799
59*2
5191
272
3703
140
4888
030
11 58
1813
5270
itkiBilti ••* «h..&d
1
217
1400
1448
737
110
2491
3868
1038
1 12
941
838
672
3138
260
1141
2682
272
1598
1753
1929
2100
1423
143
3919
4898
227
4324
1467
1037
6808
2201
I ^^ah. .*•!«••• Mtf
REOUIREO
Sample Size
2 3 4 S
1108
1400
1446
737
110
2491
3668
1035
1 12
941
835
672
3138
260
1141
2682
272
1598
1753
1929
2100
1423
143
3919
4896
227
2058
1467
1037
60.18
1936
«. CM. T«*»
••MMMMM*
1082
1400
1445
737
110
2491
38.68
1038
1 12
941
835
672
3138
260
1141
2682
272
1598
1753
822
1928
1423
V43
3919
4698
3492
1833
739
1037
5278
1708
JA A all M*
912
1400
1445
737
110
2491
3868
1035
1 12
667
835
672
884
260
11 41
2662
272
1598
1753
703
1928
1423
143
3919
4898
1683
375
700
1037
5080
1491
tA» «> mtt i4Ui m
^^MM*
902
1400
1445
737
110
2491
3668
1035
1 12
837
835
672
753
260
11 41
2682
272
1598
1753
629
1300
1423
143
3919
4696
1849
337
371
1037
5037
1417
>*tfb (•*»«•» T^J
JL.
355
1355
1445
616
110
2491
3668
1035
1 12
658
835
672
743
260
1141
2682
272
1598
1753
392
984
1423
143
3919
4896
1335
294
337
1037
4966
1291
Nfcijk £ fjmr *J.
7
244
11 14
870
412
110
2491
3668
852
1 12
657
835
672
635
260
11 41
2682
272
1598
1753
211
482
1423
143
3919
3842
t267
260
185
1037
48 55
11 19
_a
1 27
918
332
391
297
412
3668
733
' 12
1 69
S3S
672
610
ISO
11 41
2033
2 '2
!59S
161
223
1423
2052
3^*6
2 '3
' '1
35
J3
: 3-
• "" :3
630
FAVFs . Fin* Aeutt Value Factors
FAV - Final Acutt Value
55
-------
TABLE 12. WORST CASE ACUTE VALUES WHEN SALMONIOS WERE REQUIRED
Chemical
Acenaphtftene
Acrolen
Aluminum
Ammonia
Antimony (ill)
Anwmc(IU)
Cadmium
Chloride
Chlonne
Chromium (III)
Chromium (VI)
Copper
Cyanide
2 4-Omethylphenol
Mercury
Methyl pawhion
Nickel
Parafton
Pentachtorophenot
PhenarMhren*
Phenol
Selenium (IV)
Selenium (VI)
Silver
Thaftum
Tnbutylsn
i < •- 1 ncruoroMncefie
2 4 5-Tnchloropheno»
Zinc
Mean
FAV
189
664
2644
067
16S
212
362
346
4019
1869
418
888
5736
1844
077
034
832
011
1759
5962
5191
272
3703
140
4486
030
1156
1813
5270
1
670
74
40000
262
25700
14960
5399
6743
179
12436
69000
636
507
9200
420
5300
15734
10000
1485
375
11600
12500
47000
280
2300
39
4333
260
2694
670
74
40000
262
25700
14960
5395
6743
179
12436
69000
636
507
9200
420
5300
15734
5230
1465
375
11600
12500
47000
280
2300
39
4333
260
2694
670
74
40000
262
25700
14960
5395
6743
179
12436
69000
636
507
9200
420
5300
15734
3700
1485
375
11600
12500
47000
280
2300
39
4333
260
2694
Sami
4
670
74
40000
262
25700
14960
5395
6222
179
12436
69000
63*
507
9200
420
371
15734
2650
1141
375
11600
12500
47000
270
2300
39
4333
260
2694
XeStza
_5_
670
74
40000
262
25700
14960
5395
4039
179
12436
69000
636
432
9200
420
33
15734
2130
663
375
11600
12500
47000
241
2300
39
3900
260
2694
-i_
670
74
38200
262
21830
14960
5395
3583
179
12436
69000
636
426
9200
180
151
14300
48
505
234
11600
12500
20000
" 60
2200
39
3400
260
2694
7
460
74
23000
262
18460
5278
400
2950
102
12400
40600
636
364
9200
50
9 1
13000
32
308
126
11600
4300
5300
55
1800
37
30'0
260
2694
9. _
240
61
22000
262
500
874
133
2540
84
3200
3490
60
350
4800
877
69
2259
i 8
307
96
11600
3870
-60
53
'00
35
2 40
250
£4-
'Values tor phenantivww ttallium and tnbutytsn are from Table 4 all other values art from Table 5 tor N.8
FAV > Final Acute Value
56
-------
j"l
TABLE 13. WORST CASE FAVft WHEN SALMONIDS WERE REQUIRED
ChaflMCai
AcanaptMham
AcrolMn
Aluminum
Ammooia
Antmony (IK)
ArMrae(W)
Cadmium
CMonda
Chlonw
Chromium (ill)
Chromium (VI)
Coppar
Cyand*
2 4-0
-------
TABLE 14 WORST CASE ACUTE VALUES WHEN BOTH OAPHNIDS AND SALMONIOS WERE REQUIRED
Chemical
Sample Size
Aceoaphthene
Acroltn
Atuminum
Ammom
Antimony (III)
Ar»ene (III)
Cadmium
Chlonde
Chlorine
Chromium (III)
Chromium (VI)
Copper
Cyanide
2 4-Oimethylphenol
Mercury
Methyl parathion
Nickel
Panutvon
Pentaehtorophenol
Phenanftrene
Phenol
Selenium (IV)
Selenium (VI)
Silver
Th«ium
Tnbutyltn
1 2 4-Trtchloroberuene
2 4 5»Tncftlofopne~noi
Zmc
FA\T
188
664
2644
067
168
212
362
346
4019
1889
418
888
5736
1844
077
034
832
011
1759
5962
5101
272
3703
140
4686
030
1156
1813
5270
f _j_
670
74
38200
262
18460
5278
133
3583
45
12436
3490
5964
507
4800
877
91
2259
18
308
375
11600
3870
5300
55
2200
39
4333
260
547
3
670
74
38200
262
18480
5271
133
3583
45
12436
3490
5964
507
4800
877
91
2259
18
308
375
11600
3870
5300
55
2200
39
4333
260
547
4
670
74
38200
262
18480
5278
133
3583
45
12438
3490
5964
507
4800
877
91
2259
18
308
37S
11600
3870
5300
55
2200
39
4333
260
547
•k
5
670
74
38200
262
18480
5278
133
3583
45
12436
3490
5964
432
4800
877
9 1
2259
18
306
375
11600
3870
5300
55
2200
39
3900
260
547
6
670
74
38200
262
18460
5278
133
3583
45
12436
3490
5964
426
4800
877
91
2259
1 a
308
234
11600
3870
5300
55
2200
39
3400
250
547
7
460
74
23000
262
18460
5278
133
2950
45
12400
3490
5964
364
4800
877
9 1
2259
1 8
308
126
11600
3870
5300
55
1800
37
3010
260
547
8
*•«••••
240
€1
22000
262
500
874
133
2540
45
3200
3490
5964
350
4800
877
69
2259
1 8
307
96
11600
3870
760
53
100
;s
: jo
250
54'
•Valuts lor phtnanthrtn* thallium and tnbutylon art from Tabl* 4 all ofttr valutt ar» from Table 5 tor N=8
'Not applicable two values required
FAV « Rnal Acute Value
53
-------
TABLE 19 WORST CASE FAVFt WHEN BOTH OAPHHIOS AND SALMONIOS WERE REQUIRED
ChtflHCtf
Acer.****.
Aeraien
Aluminum
Ammoraft
Anfimony (III)
Arsene (III)
Cadmium
Chtonde
Chlonne
Chromium (III)
Chromium (VI)
Copper
Cyarude
2 4-Oimethylphenol
Mercury
Mediyl parathcn
Nickel
Paratfuon
Pentacnlorapnenoi
Phenol
Selenium (IV)
Selenium (VI)
Stiver
Thafcum
Tnbutyfon
i 2 4-Tnchtorabenxene
2 4 S-TncMorephenol
Zinc
Arrth mean
Geo mean
Mean
FAV"
18»
664
2644
067
161
2t?
36?
340
4010
1880
4111
88)1
5736
1844
077
034
832
011
59(12
5191
2*
3703
1 (0
4686
030
1156
1813
5270
Sample SIM
1* 2 3 4 S fi
355
1114
1445
391
110
2491
3668
1036
1 12
658
835
672
884
260
1141
2682
272
1598
1753
629
223
1423
143
3919
4696
1335
375
143
1037
4933
1201
355
11 14
1445
391
110
2491
3668
1035
1 12
658
835
672
884
260
1141
2682
272
1596
1753
629
223
1423
143
3919
4696
1335
375
143
1037
4933
1201
355
11 14
1445
391
110
2491
3668
1035
1 12
656
839
672
884
260
1141
2682
272
1598
1753
629
223
1423
143
3919
4696
1339
375
143
1037
4933
1201
^^••WMMt
355
11 14
1446
391
110
2491
3668
1035
1 12
658
835
672
753
260
11 41
2682
272
1598
1753
629
223
1423
143
3919
4696
1335
337
143
1037
4927
1190
••MMa^M
355
11 14
1445
391
110
2491
3666
1035
1 12
65t
835
672
743
260
1141
2682
272
1598
1753
392
223
1423
143
3919
4696
1335
294
143
1037
4917
1165
7
244
11 14
870
391
110
2491
3668
952
1 12
657
635
672
635
260
11 41
2662
272
1596
1753
211
223
1423
143
3919
3842
1267
260
1 43
•037
4844
'078
-J-
'27
918
832
391
297
412
3666
'33
1 12
'69
835
672
610
260
11 41
2033
: 72
598
744
161
223
1423
2052
3776
2 13
"1
35
-13
C 3"*
":3
530
'Valun for phtnanmrvn* tfiallium and tnbuiyltm are from Table 4 all other value* are from Table 5 lex N=3
'Not applicable two value* required
PAVFs • final Acute Value Factors
FAV > final Acute Value
59
-------
TABU 1C SUMMMIV OF OEOMETraC MEAN WORST CASE FAVft
SmpfeSin
J. A i
Oaphndfltqund
S41
22
51
236
10
50
12
134
17
90
12
77 80
19 14
41 31
12 12
36 19
13 It
2» It
12 11
8
6
8
FAVFt • Final Acutt VakM Faem
60
-------
TAMJ17 fflCTIWATCT 8UMIMRY FAVPt (VERSHON 1)
3 4 _5_.T §_ jr_ _g_
Any tam*
Dapfcradmqund
Saknorad wound
Bet) Oaptwd and Safcnamd
9Qw) rWCSWlkw Of IflQfBM
Anyf»«iy
Daehnd wqurad
Salmon* raqund
toft Daphnrt and Salmon**
Mttftft o* 9otf) pftfcwi&lot
Any*!*
Daphnd nqurcd
BQv) DflptwuQ MO Sg^fnoftid
OS* Buc^Atf* a* OSMi naiujji.
Any tan*
Daphrad i*4umd
Saknorad mqurad
OvaraflUadtai
Any tan*
Oatfmd wound
SMmondra^imd
8ott Ovptmd and Satmonid
Any tan*
Oaphmd raqwrad
Salmond rtQMVd
Botft OapAfMd and S*mc«* '99
11 10
10000
202
15100
•TJl/»
202
420
303
425
299
17500
3874
19633
3845
291
257
343
260
2914
2100
2170
2109
1373
979
19 6A
l4fcWJ
979
449
112
4000
202
333
257
303
257
8718
2614
6716
32.63
290
197
259
2.29
2170
1593
15.93
1399
12.63
979
n7l
/ 1
914
209
112
279
tj\rt
202
260
229
257
216
3645
21 70
3945
2381
218
190
2.20
192
1593
1399
15.93
1099
906
922
1O lal
*W «•
914
112
118
202
221
194
221
190
2614
2100
2893
2170
219
182
201
192
1593
1058
1059
1059
703
590
7(11
/ UJ
590
6716
»4«7
6716
n.A AV
9467
193
182
187
182
1917
19 17
19 17
19 17
1 74
162
182
1 77
1058
1583
1058
1058
401
401
1 AM
4 **
401
2893
2893
2893
MB AA
2893
182
1 77
132
180
1500
1583
1593
1389
FAVF* m find AoM VakM
61
-------
TAOLE It. mnNWMTER 3LHMMKY PAVPt (VERSION 2)
M^atmadta.
Any tarty
Oaphnd raqund
Oaphnd aadudad
Satmond wound
Bo* Oaphnid and Sabnomd
***«**•«*««*
Anytamiy
Daphndraqund
Oaphradaxdudad
SaWltOnKI fVQUaTMl
Bo* Daphnid and Salmon*
Anytanrty
Oaphnid raquvad
Oaphnid aictudad
SaJmond nomad
Soft Oaphnid and Safmonid
95tfi PMtwtte at Mfti jLinujinLu.
Any tarty
Daphradraqund
*^Pl^ii9f^BP Vtat^HMp^PW
Stirnoflrtfa^avvd
NDflh QApftniQ flflO SVfUQtlttl
OvandMadlan
Any tarty
Oaphnd raqund
OvanJ 99t) Paraanota
Any tarty
DqtpnfWI IVQUaTQQ
Oaphnd aidudad
Safenond raqund
Soft Oaphmd and Salmon*
i
1793
399
19 17
784
—
2298
49
4101
22451
—
42649
2169
42321
5427
—
166272
339
174114
133099
—
1499
499
1093
8367
94
10024
20211
—
-*-
703
319
1137
564
319
83
33
292
1149
47
6449
1195
8480
3918
789
39614
168
39990
100021
101
620
323
600
9 at
i Via
1007
58
1680
1857
52
J-
394
240
709
400
272
5911
22.97
7529
9749
3929
4*73
799
51 14
2230
714
9999
132
17930
29929
99
373
290
4**
ff
399
19000
5054
42411
21620
4320
Sami
4
292
229
4 14
304
249
1909
22.42
2940
5994
2952
2799
699
4912
2017
714
1297
107
1901
2130
99
291
239
4 MA
J 99
270
5 9»
< 2a
9014
4193
17561
9799
3983
triaSua
-^
209
213
341
239
223
722
949
2073
979
979
2007
909
3941
1999
909
90392
9799
99133
97709
9789
240
218
4 rtat
3 09
237
24*
TO
5359
3102
111 17
5966
2900
JL
199
207
299
203
219
646
603
1763
660
639
1*53
520
2419
1314
499
433 30
9034
99099
39912
9719
216
202
9 A1
261
214
9 rt4
i we
3317
2200
$423
3373
1990
7
190
196
289
199
202
607
576
1561
598
622
749
430
765
792
430
8915
6999
19361
7529
7528
196
192
M 4£
2 3o
194
1694
1312
3990
1718
1322
•*-
199
193
199
193
191
532
509
509
499
519
304
300
310
303
302
4071
3997
3402
4979
4893
86
86
86
799
726
751
806
842
FAVFt * final ACM* Vaiua factors
62
-------
TABLf If lAttWTWlUMMMIYFAVRi (VERSION 2)
Any ami*
MP* i*qund,
M8r*qum*MP«B*id»d
MP ano M6
MM* UA
no* MB
MPmqumd MBaiduotd
Ml mound MP«dud«d
Sett MP and M6
flOf MV
Anylamdr
MBrtqwad MPtxdudw)
Bott) MP and Ml
MP nor M8
95ttt
Anylmriy
MP rvqunri. Ml «xdudKt
NMhtr MP nor Ml
Any
NMrMrMPnorMI
MP raqurao, MB •xtiudrt
MSiwqumd M
Bo* MP and M8
nor M6
2080
382
2061
—
3041
34448
3168
19013
—
42502
394 66-
12.50
52.76
—
21084
428709
271
2915
—
82424
16.53
388
1481
__
2408
4903
74
314
—
5419
594
288
747
398
716
7888
1091
6864
2088
11374
12052
11 61
4323_
11 44~
6326
2557
132
1738
129
3193
- 5-40
284
628
326
841
28772
5071
191 70
4677
43678
350
284
511
268
407
1798
563
3Z58
990
6433
4143
952
2717
987
3482
1425
95
433
108
1337
328
251
416
266
435
101 11
2182
13802
3415
17426
288
233
290
297
309
814
377
3011
593
6808
1658
901
1660
776
2081
54762
7728
432.18
8879
61594
267
232
317
252
328
6334
1364
9058
22.20
10550
241
211
253
240
293
401
351
2656
429
4053
1181
664
1183
769
1222
19812
7188
43137
7308
38961
244
217
268
235
271
2261
882
7661
1501
7823
226
200
239
210
262
366
351
2477
331
3320
819
545
856
710
626
8627
12.92
38671
7048
27688
229
206
258
220
252
1202
606
6273
936
6688
207
1 98
226
206
*••»
342
351
2214
331
^"**
539
287
484
47S
"*"
1990
847
19401
12.38
'"'~LJ "
211
204
243
209
~
650
462
5720
607
-
206
197
201
199
***"
331
351
380
33!
*•"""
335
2J?
285
286
*
S23
8*7
a 49
836
~
204
204
'97
205
"
442
e~
450
, 47
-
' MP indudM al
* M8 indudM all
FAVFt. Final Acuta
m tht tarnily MytidM and all spcews m m« lamriy Pcnwda*
in m« genus M»nipia
63
-------
Table 20. FAVFs Based on Log-triangular Distribution.
s :
1.0
1.0
1.0
2.5
2.5
2.5
5.0
5.0
5.0
7.5
7.5
7.5
10
10
10
•»*•«% ^VIM4dl
PERCZN
95
80
50
95
80
50
95
80
50
95
80
50
95
80
50
1
2.63
2.10
1.62
11.2
6.4
3.3
125
41
11
1411
264
38
15835
1692
126
2
2.05
1.69
1.37
6.0
3.7
2.2
36.2
13.6
4.9
217
50
11
1309
185
24
3
1.79
1.52
1.26
4.3
2.9
1.8
18.5
8.2
3.2
79.6
23.5
5.6
342.6
67.2
10.0
O1VVDT V
4
1.65
1.42
1.19
3.5
2.4
1.5
12.4
5.8
2.4
43.5
14.0
3.7
152.9
33.8
5.8
aiiib
5
1.56
1.35
1.15
3.1
2.1
1.4
9.4
4.5
2.0
28.7
9.6
2.8
88.0
20.3
3.9
6
1.50
1.30
1.11
2.7
1.9
1.3
7.5
3.7
1.7
20.6
7.1
2.2
56.4
13.7
2.9
7
1.44
1.26
1.09
2.5
1.8
1.2
6.2
3.2
1.5
15.6
5.6
1.9
39.0
9.9
2.3
8
1.40
1.23
1.07
2.3
1.7
1.2
5.3
2.8
1.4
12.4
4.6
1.6
28.6
7.6
1.9
FAVFs - Final Acute Value Factors,
S * Slope factor.
64
-------
Table 21. FACRs from Criteria Documents.
Chemical
Aluminum
Ammonia
Ammonia
Arsenic (III)
Cadmium
Chlordane
Chloride
Chlorine
Chlorpyrifos
Chromium (II I)
Chromium (VI )
Copper
Cyanide
Dieldrin
Endosulfan
Endrin
Lead
Linda ne
Mercury
Nickel
Parathion
PentachlorophenoL
Selenium
Toxaphene
Zinc
Freshwater
17*
18e
3.8
6.9d
14
7.6
3.3
4.1
16d
2.9
3.0"
8.6
8.5
3.9
4.0
51
25
3.7
18
10
3.2
8.0
>38*
2.2
Saltwater
NAb
13
3.8
9.1
14
NA
3.3
4.1
NA
43
2.0
2.0
8.5
3.9
4.0
51
NA
3.7
18
NA
3.2
8.0
2.0
2.2
Reference
EPA 440/5-86-008
EPA 440/5-85-001
EPA 440/5-88-004
EPA 440/5-84-033
EPA 440/5-84-032
EPA 440/5-80-027
EPA 440/5-88-001
EPA 440/5-84-030
EPA 440/5-86-005
EPA 440/5-84-029
EPA 440/5-84-029
EPA 440/5-84-031
EPA 440/5-84-028
EPA 440/5-80-019
EPA 440/5-80-046
EPA 440/5-80-047
EPA 440/5-84-02
EPA 440/5-80-0
EPA 440/5-84-0
EPA 440/5-86-004
EPA 440/5-86-007
EPA 440/5-86-009
EPA 440/5-87-006
EPA 440/5-86-006
EPA 440/5-87-003
' This value was obtained by dividing the FAV by the lowered FCV
given in Table 3 in the criteria document.
6 NA - Not Available.
0 Calculated as a weighted average of the ratios for the range of
pH from 6.5 to 9.0. (See page 95 in the criteria document.)
d Calculated as two times the geometric mean of the quotients of
the CMC divided by the CCC at hardnesses of 50, 100, and 200
mg/L given in the National Criteria in the criteria document.
The purpose of the factor of two is to account for the fact
that the numerator in this ratio is the CMC, which is one-half
the FAV.
* This value was obtained by dividing the FAV by the value for
brook trout given in Table 3 in the criteria document.
FACRs * Final Acute-Chronic Ratios.
65
-------
APPENDIX
GUIDELINES FOR DERIVING AN
AQUATIC LIFE-PESTICIDE CONCENTRATION
by
Charles £. Stephan
and
Russell J. Erickson
ENVIRONMENTAL RESEARCH LABORATORY - DULUTH
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
DULUTH, MN 55804
66
-------
I. introduction
Region III of the U.S. EPA intends to develop a Pesticide
Hazard index (PHI) that can be used to rank the toxicologxcal
hazards of chemicals used as active ingredients in registered
pesticides. The PHI will contain three toxicological components in
order to separately take into account effects on aquatic organisms,
humans, and wildlife. These Aquatic Life-Pesticide Concentration
(ALPC) Guidelines were developed to be considered for use as the
aquatic life component of the PHI and are based in part on the
"Guidelines for Deriving Numerical National Water Quality criteria
for the Protection of Aquatic organisms and Their Uses" (stephan,
et al., 1985), commonly known as the "National Guidelines'*. Use of
these ALPC Guidelines will usually require much professional
judgment; all necessary decisions should be based on a thorough
knowledge of aquatic toxicology and an understanding of the
National Guidelines.
II. Rationale
The most desirable input into the PHI concerning the toxicity
of a pesticide to aquatic organisms is the freshwater and/or
saltwater Criterion Continuous Concentration (CCC), derived as
specified in the National Guidelines. In most cases, a CCC is
equal to a Final Chronic Value (FCV), which is usually derived
dividing a Final Acute Value (FAV) by a Final Acute-Chronic Rat
(FACR). A FAV is derived only if at least eight Genus Mean Acute
Values (GMAVs) are available, such that the minimum data
requirements concerning acute toxicity are satisfied. Similarly,
a FACR is derived only if three or more acceptable experimentally
determined acute-chronic ratios (ACRs) are available.
For a pesticide; for which-too few data are available-to allow
derivation of a CCC, the procedure described in Part III below
specifies that a freshwater and/or saltwater Aquatic Life-Pesticide
Concentration (ALPC) be calculated and used as the aquatic life
input into the PHI. The primary rationale of this procedure is
that the greater uncertainty associated with fewer data should
rarely result in an underestimate of hazard. Therefore the
procedure was designed so that at least 95% of the ALPCs are
expected to be lower than the comparable FCVs would be if enough
data were available to allow derivation of the FCVs. ALPCs are
thus biased estimators of hazard, relative to FCVs, but the average
amount of bias decreases as the amount of pertinent data concerning
the pesticide increases and uncertainty decreases. A pesticide
ranked on the basis of an ALPC will almost always appear to be more
hazardous than a comparable pesticide ranked on the basis of a CCC.
A second rationale is that the procedure should use, as much
as possible, and otherwise be consistent with, the procedures used
in the National Guidelines to derive a FCV. To this end, an A
67
-------
is intended to reflect chronic toxicity and is calculated by
dividing an acute value by an ACR. The acute value is a FAV if one
can be calculated according to the National Guidelines; if not, a
Pesticide Acute Value (PAV) is calculated by dividing the lowest
available GMAV by a Pesticide Acute Factor (PAF). An ALPC is then
obtained by dividing the acute value (FAV or PAV) by a FACR if one
can be derived according to the National Guidelines; otherwise the
acute value is divided by a Pesticide Acute-Chronic Ratio (PACK) to
obtain the ALPC. If at least one acceptable GMAV is available for
fresh water and at least one for salt water, both a freshwater and
a saltwater ALPC can be derived for the pesticide of concern.
The intent of having nearly all ALPCs be lower than the
comparable FCVs would be was implemented in three ways - two
concerning the PAF and one concerning the PACR:
1. The PAFs given in Tables 1 and 2 were derived so that 95% of the
PAVs are expected to be lower than the FAVs that would be
obtained if all eight minimum data requirements were satisfied.
Calculation of the PAFs was based on a computerized sampling of
the results of individual acute toxicity tests given in 29
freshwater and 24 saltwater final and draft aquatic life
criteria documents (Host, et al., 1990). To obtain the PAFs
given in Table 1, samples of size eight were randomly selected
from the individual freshwater acute test results in each
criteria document, except that each sample was required to
satisfy the minimum data requirements concerning acute toxicity
so that a FAV could be calculated for each sample. Then
subsamples of sizes one through seven were randomly selected
from each sample, except that (a) in one case each of the seven
subsamples obtained from each sample of eight was required to
contain an acute value for a daphnid, and (b) in the second case
none of the subsamples was allowed to contain a value for a
daphnid. For each subsample a factor was then calculated by
dividing the lowest value in the subsample by the FAV for the
sample of eight from which the subsample was obtained. The
factors so calculated for any one of the fourteen kinds of
subsamples were pooled across chemicals and the 95th percentile
was calculated and used as the PAF given in Table 1. Similar
analyses were performed using the results of individual
saltwater acute toxicity tests given in the criteria documents
to obtain the PAFs presented in Table 2.
a. Fresh water (Table 1):
For data sets that do not contain a GMAV for at least one of
three specified genera in the family Oaphnidae fCeriodaphnia.
Daphnia. and Simocephalus), the PAFs given in Table 1 decrease
from 10,000 to 40 as N (» the number of minimum data
requirements that are satisfied) increases from l to 7. Thus if
a data set contains only one GMAV, and it is not for one of the
specified daphnid genera but it happens to be for the
68
-------
sensitive genun, addition of more GMAVs will not change th
lowest GMAV, but the PAF can decrease from 10,000 to 40. In
contrast to these 95th percentiles of 10,000 and 40, the
corresponding medians are 18 and 2.4. Therefore the PAF is more
than 550 times higher than half of the factors obtained when
N-l, and more than 16 times higher than half of the factors
obtained when N«7.
For data sets that contain a GMAV for at least one of the
specified daphnid genera, the PAFs decrease from 94 to 13 as N
increases from 1 to 7. The PAFs are much lower when a GMAV for
at least one of the specified daphnid genera is available
because, in the data sets analyzed, these genera were rarely
more than 100 times more resistant than the most sensitive genus
whereas npn-daphnid genera were sometimes more than 10,000 times
more resistant than the most sensitive genus. (Such other
daphnid genera as Moina are not included in the list of
specified genera because little is known about the relative
sensitivities of these other genera.) Although salmonids are
often considered sensitive, the PAFs obtained for subsamples
that were required to contain a GMAV for a genus in the family
Salmonidae were similar to the PAFs obtained for subsamples that
were not required to contain a GMAV for a salmonid.
when N«l for fresh water, the magnitude of the PAF depends on
(a) whether a GMAV is available for one or more of the specifi
daphnid genera and (b) the highest test results that could
randomly selected for use in the calculation of the PAF, because
the PAFs are 95th percentiles and therefore depend mostly on the
extreme values. When N is greater than 1, the PAF also depends
on the number of minimum data requirements that are satisfied,
which can be understood as a reduction in the PAF as the number
of available pertinent data increases.
b. salt water (Table 2):
Whereas freshwater PAFs depend greatly on whether the data set
contains a GMAV for any of the three specified genera in the
family Daphniclae, saltwater PAFs depend greatly on whether the
data set contains a GMAV for either of the following two
crustacean genera: Mvsidopsis and Penaeua. If the data set does
not contain a GMAV for either Myaidopsis or Penaeus and if the
data set satisfies only a few minimum data requirements, the PAF
depends greatly on whether the data set contains a GMAV for the
fish genus Menidia. The situation is more complex for salt
water than for fresh water because sufficient data are available
to indicate that two types of saltwater genera deserve special
consideration.
Because the saltwater situation is more complex, it becomes
important that the amount of reduction in the PAF as mo
minimum data requirements are satisfied depends not only on
69
-------
number of requirements that are satisfied, but also on the
percentage of chemicals for which the special genera are the
most sensitive. The special crustacean genera fMvsidopgia and
Penaeuel not only have a low PAF when N-l but one or the other
is the most sensitive saltwater genus for 40% of the chemicals
for which data were used in the derivation of the saltwater
PAFs. In contrast, when N-l the PAF is about 5 times higher
when the GMAV is for Menidia than when the GMAV is for
Mvsidoosis or Penaeus; also, Menidia is the most sensitive genus
for only 4% of the chemicals. The fact that Menidia is never
extremely resistant is very important when N»l but is
counterbalanced at larger N by the fact that Manidia is the most
sensitive saltwater genus for very few chemicals. Thus whether
a data set contains a GMAV for Menidia is only an important
consideration when (a) the data set satisfies only a few minimum
data requirements and (b) the data set does not contain a GMAV
for either Mysidopsis or Penaeus.
2, The PAFs given in Tables 1 and 2 are based on FAVs calculated
from samples that contained eight acute values and, therefore,
eight GMAVs, but many FAVs are based on more than eight GMAVs.
Therefore an analysis was conducted concerning the dependence of
the FAV on the number of GMAVs used in the calculation of the
FAV (Host, et al.r 1990). For each chemical, sets of eight or
more GMAVs were randomly selected, except that each set of GMAVs
was required to satisfy the minimum data requirements concerning
acute toxicity. A FAV was calculated for each set of GMAVs and
a mean FAV was then calculated for each selected number of
GMAVs. For most chemicals the mean FAV increased slightly as
the number of GMAVs in the set increased from 8 to 40. Thus the
PAFs in Tables 1 and 2 are slightly larger on the average than
they would be if they had been based on FAVs calculated from
samples containing more than eight GMAVs.
3. The PACR was given a default value of 40 based on an analysis of
FACRs available from final and draft aquatic life criteria
documents (Host, et al., 1990). For fresh water, 24 FACRs were
available, whereas 19 were available for salt water. The 95th
percentile FACR was calculated to be 43 for fresh water and 39
for salt water. In fresh water the estimated median FACR is
7.8, so a value of 40 is more than 5 times larger than the FACR
for half of the chemicals. In salt water the estimated median
FACR is 4.0, so a value of 40 is more than 10 times larger than
the FACR for half of the chemicals.
For these three reasons the ALPC is expected to be lower than the
FCV would be for nearly all pesticides and substantially lower for
most.
70
337
-------
III. Procedure
A. All available data concerning the acute and chronic toxicity of
the pesticide of concern to freshwater and saltwater aquatic
animals should be collected and screened for acceptability as
described in the National Guidelines.
B. For each species for which the result of at least one acceptable
acute toxicity test is available, a Species Mean Acute Value
(SMAV) should be determined as described in the National
Guidelines and then GMAVs should be calculated.
c. If the eight minimum data requirements concerning acute toxicity
are satisfied for fresh water or for salt water, the FAV(s)
should be derived as described in the National Guidelines. When
a desired FAV cannot be derived because the eight minimum data
requirements aze not satisfied, the lowest available GMAV should
be divided by the appropriate PAF specified in Tables 1 and 2 in
order to calculate the PAV.
0. If three or more experimentally determined ACRs, which are
acceptable according to the National Guidelines, are available
for the pesticide of concern, the FACR should be derived as
described in the National Guidelines. If fewer than three
acceptable experimentally determined ACRs are available, enough
ACRs of 40 should be assumed so that the total number o
experimentally determined and assumed ACRs equals three; a PA
should then be calculated as the geometric mean of the thre
ACRs. Thus if no acceptable experimentally determined ACR is
available, the PACR is 40.
E. If both a FAV and a FACR are available for either fresh water or
salt water or both, the FCV(s) and CCC(s) should be derived as
described, in the National Guidelines and the ccC(s) used, as the
aquatic life input into the PHI. When a desired CCC cannot be
derived, the corresponding ALPC should be calculated by dividing
the PAV (or preferably the FAV if it can be derived) by the PACR
(or preferably the FACR if it can be derived) and used as the
aquatic life input into the PHI.
IV. Discussion
This procedure obviously contains a large amount of
uncertainty, and therefore a substantial bias for overprotection
relative to the FCV, when few pertinent data are available. If a
freshwater ALPC must be derived for a pesticide for which the only
toxicity test result available is a GMAV for a non-daphnid
freshwater genus, the PACR would be calculated using a PAF of t
10,000 and a PACR of 40. If this genus happens to be the one that
is the most acutely sensitive to this pesticide and if this genus
also has a very low ACR with this pesticide, the ALPC would
71
-------
almost 400,000 tin** lower than what th« FCV would b* if sufficient
data were available to allow its derivation. The procedure
presented for the calculation of the ALPC shows how additional
pertinent data could affect the PAF and PACR and the National
Guidelines indicate what data would be needed to allow derivation
of a FCV rather than an ALPC. The amount of bias in the ALPC would
be reduced by basing the PAFs and PACK on a percentile such as the
80th rather than the 95th; this would also result, however, in more
ALPCs being underprotective in relation to FCVs. Because this
procedure contains three "margins of safety", the percent of ALPCs
that would be lower than the comparable FCVs might still be about
95 if the PAFs and the PACR were both based on 80th percentiles.
If an ALPC is used to evaluate the importance of a measured or
predicted concentration of a pesticide in a body of water, the
evaluation should take into account the fact that ALPCs are
designed to be lower than FCVs would be. Thus a concentration that
is slightly below an ALPC should be of even less concern than would
be a concentration that is slightly below a FCV. Similarly, a
concentration that is slightly above an ALPC should be of less
concern than would be a concentration that is slightly above a FCV.
Although a concentration that is slightly above an ALPC should
usually be of little concern, such a concentration probably does
indicate a need for more data concerning the toxicity of the
pesticide to aquatic species. If a measured or predicted ambient
concentration is above an ALPC or a CCC, more information
concerning the ambient concentration is probably desirable.
Because the chemicals that are active ingredients in
registered pesticides are very diverse, the PAFs and PACR used in
the procedure presented above were based on data for all chemicals
for which final or draft national water quality criteria for
aquatic life were available, not all of which are or were
ingredients in registered pesticides. Although it would be
desirable for these ALPC Guidelines to depend at least in part on
the mode of action of the individual pesticide, no way of
accomplishing this is known at this time.
72
-------
TABLE 1
Freshwater Pesticide Acute Factors (PAFs)1
Number of
freshwater
minimum data
requirements
that are
satisfied.
If the data set
does contain a
CMAV2 for a
specified daphnid3,
select the PAF
from this column.
If the data set
does not contain
a GMAV2 for a
specified daphnid3,
select the PAF
from this column.
1
2
3
4
5
6
7
94
58
51
42
31
22
13
10,000
1,700
420
180
110
64
40
1 All PAFs have been rounded to two significant digits.
2 Genus Mean Acute Value.
3 A GMAV for at least one of the following genera in the
family Daphnidae: Ceriodaphpia. Daphnia. or Simoeephalus.
73
-------
TABLE 2
Saltwater Pesticide Acute Factors (PAFs)1
Number of
saltwater
minimum
data
require-
ments
that are
satisfied.
1
2
3
4
5
6
7
If the data
set contains
a GMAV2 for
MVSldOPSl,3
or Penaeus
and a GMAV
for Menidia.
NA3
47
34
22
15
9.4
6.1
If the data
set contains
a GMAV2 for
Mvaidopsis
or Penaeus .
but not for
Menidia.
74
51
22
14
8.8
6.1
4.6
If the data
set contains
a GMAV2 for
flenidia. but
not for
Mysidopsis
or genaeus .
310
190
140
90
77
63
57
If the data
set does
not contain
a GMAV2 for
Mvsidopsis
or Penaeus
or Menidia.
5,400
440
170
110
78
69
MA3
1 All PAFs have been rounded to two significant digits.
2 Genus Mean Acute value.
3 Not Applicable.
74
-------
REFERENCES
Host, G.E., R.R. R«gal, and C.E. Stephan. 1990. Analyses of Acute
and Chronic Data for Aquatic Life, internal Report. U.S. EPA,
Duluth, MN.
Stephan, C.E., D.I. Mount, D.J. Hansen, J.H. Gentile, G.A. Chapman,
and w.A. Brung.. 1985. Guidelines for Deriving Numerical National
Water Quality Criteria for the Protection of Aquatic Organisms and
Their Uses. PB85-227049. National Technical Information Service,
Springfield, VA.
75
-------
BIOACCUMULATION FACTOR
MATERIALS
-------
DRAFTi September 6, 1991
GRBAT LAKES INITIATIVE
PROCEDURE TO DETERMINE A BIOACCUMULATION FACTOR
I. INTRODUCTION
The purpose of this procedure is to determine bioaccumulation factors to be used
in the calculation of Great Lakes Initiative human health and wildlife criteria
and levels of concern. Also, BAFs vill be used to identify the Bioaccumulativ«
Substances to be considered under the Great Lakes Initiative programs.
Bioaccumulation reflects uptake of substances by aquatic organisms exposed to
the substance through all routes, as would occur in netuie. Bioconcentration
reflects uptake of substances by aquatic organisms exposed to the substance only
from the surrounding vater medium. Both bioaccumulation and bloconcentration
factors (BAF and BCF) are proportionality constants, relating the concentration
of a substance in aquatic organisms to its concentration in the surrounding
vater. Measured or predicted BAFs, rather than BCFs, vill be used to calculate
criteria and levels of concern because BAFs represent the bioaccumulation that
occurs in natural aquatic systems. Predicted BAFs will be calculated from a BCF
by multiplying the BCF by a food chain multiplier (PCM).
II. DEFINITIONS
Bioaccumulation. Uptake and retention of substances by an aquatic organism
from its surrounding media and food.
Bioaccumulation Factor (BAF). The ratio of a substance's concentration in
tissue versus its concentration in ambient water, in situations vhere th«
organism and the food chain are exposed.
Bioconcentration. Uptake and retention of substances by an aquatic
organism from the surrounding vater through gill membranes or other external
body surfaces.
Bioconcentration Factor (BCF). The ratio of a substance's concentration in
tissue versus its concentration in ambient waiet, in situations vhere the
organism is exposed through the vater only.
Food Chain Multiplier (FCH). A factor that varies vith the log K of an
organic chemical by vhich a BCF is multiplied to account for uptake of the
chemical from sources other than the vater, particularly uptake through the
organism's diet.
Steady-State BAF/ECF, A BAF or BCF that does not change substantially ov
time; that is, the BAF or BCF that exists vhen uptake and depuration are equa
Depuration. The loss of a substance flora an aquatic organism.
-------
Octanol/water paitition coefficient (Kov>- Tn« tatio, at equilibrium, of
the concentration of a substance in the octanol phase to its concentration in
the aqueous phase in a tvo-phase octanol/vater system.
Uptake. The soiption of a substance into
-------
S. Unpublished data.
B. Data Review
Measured BCFs and, if applicable, measured BAFs should meet the procedural and
quality assurance requirements specified in the ASTM (1990) "Standard Practice
for Conducting Bioconcentration Tests with Fishes and Saltvater Bivalve
Molluscs", ASTM designation E 1022-84; and in the USEPA guidance contained in
Stephen et al. (1985) "Guidelines for Deriving Numerical National Water Quality
Criteria for the Protection of Aquatic Organisms and Their Uses". In
particular, the follovirg should be met:
1. The BAF/BCF is Steady-State, or steady-state conditions can b«
estimated.
2. The concentration of the substance did not have an adverse effect
on the organisms,
3. The concentration of the substance in the water was measured and
relatively constant over the exposure period. The concentration
should be averaged over the period steady-state conditions vcre
achieved.
4. The organisms vere exposed to the substance using a flow-throug
or renewal procedure.
5, For organic chemicals, the percent lipid has been measured in, or
can be reliably determined for, the teat organism.
C, Data Screening and Selection
These procedures provide overall guidance for the determination of BAFs, but
they can't cover all the decisions that must be made in the review and selection
of acceptable data. Professional judgement is required throughout the process.
A degree of uncertainty is associated vith the determination of any BAP. The
amount of uncertainty depends on the quality of data available and the method
used to derive the BAF.
Field measured BAPs should be based on fish species, preferably living in the
Great Lakes at or near the top of the food chain (trophic level four). This is
particularly true for organic chemicals vith log K values greater than four.
The conditions of the field study should not be so unique that the BAF is not
applicable to other locations.
Laboratory measured BCJ's also Should be based on fish species, but BCFs for
molluscs and other invertebrates may be used vith caution. For example,
invertebrates metabolize some chemicals less efficiently than vertebrates vhich
inflates the BCF.
The percent lipid content of the test fish should be reported as part of a
BAF/BCF study on organic chemicals.
All average (mean) values are geometric means unless specified otherwise.
-------
Plant BAFs/BCFs should not be used for human health criteria. If used for
wildlife criteria, they should be used in proportion to the weight, relative to
the total diet, of plant material consumed by the wildlife species to be
protected.
BAFs/BCFs are expressed on a wet weight basis. BAFs/BCFs reported on a dry
weight basis must be converted to wet weight. If no conversion factor is
reported with the BAF/BCF, multiply the dry weight value by 0.1 for plant and
animal plankton, or by 0.2 for fish and invertebrate*.
V. DETERMINATION Of BAFs FOR INORGANIC CHEMICALS
BAPs are assumed to be equal to BCFs for inorganic substances.
Concentrations of an inorganic substance in a BAF/BCF study should be greatei
than natural background levels and greater than levels required for normal
nutrition of the test organism if the substance is a micronutrient, while still
below levels which adversely affeti the organism. Bioaccumulation of inorganic
substances may be overestimated if concennations are at or below normal
background levels.
A. BAF for Human Health Criteria
1. BAFt/BCPs for human health criteria should be based on edible
tissue of fish unless it can be demonstrated that whole body
BAFs/BCFs are not inflated due to preferential accumulation in
nonedible tissues, such as liver and kidney.
2. The BAF for an inorganic chemical equals the geometric mean of
the acceptable BAFs, or, in the absence of BAF data, the
geometric mean of the acceptable BCFs.
B. BAF for Wildlife Criteria
1. BAFs/BCFs for wildlife criteria should be based on whole body
fish data.
2. The BAF for an inorganic chemical equals the geometric mean of
the acceptable BAFs, or, in the absence of BAF data, the
geometric mean of the acceptable BCFs*
3. BAFs for invertebrates and aquatit. plants may be consideied on a
case-by-case basis if the diet of the wildlife species which the
criterion will protect includes nonflsh organisms.
VI, DETERMINATION OF BAFs FOR ORGANIC CHEMICALS
A. Lipid Normalisation
For lipophilic organic chemicals, the BAF/BCF is assumed to be directly
proportional to the percent lipid from one tissue to another and from one
-------
aquatic species to another. Percent lipid data are used to convert the reported
BAFs/BCPs to a BAP appropriate for the fisheries of the Great Lakes and their
tributary waters. Percent lipid data are also used to determine BAFs for
calculating human health and wildlife criteria from the same date.
Determination of the appropriate percent lipid for a fish species is often
complicated by the variability of lipid content due to changes in environmental
conditions, changes over time, differences due to fish size, age, sex, diet, and
other factors.
1, The average percent lipid content of the test organism (whole
body or edible tissue) is used to normalize the measured
BAFs/BCPs. Percent lipid of the tissue sample is determined
gravimetrically following extraction and evaporation using a
solvent such as hexane, petroleum ether or dichlorobenzene.
2. Percent lipid should be measured in the teat organism* used In
the BAF or BCP study. If percent lipid is not reported for the
test organisms in the original BAF/BCP publication it may be
obtained from the author, or, in the case of a laboratory study,
lipid data for the same laboratory population of test organisms
that were used in the original study may be used.
3. Reported BAPs and BCPs are normalized to one percent lipid by
dividing the BAFs and BCFs by the average percent lipid for the
test organisms. Both whole body and edible portion BAF/BCFs are
normalised using the respective whole body and edible portion
percent lipid values.
B. Pood Chain Multiplier (PCM)
In the absence of measured BAP data, a PCM is used to predict the BAP. The
appropriate PCM in selected from Table 1 under trophic level four, unless
case-specific information supports the use of a PCM for a lower trophic level. A
PCM greater than one Is applicable to most lipophillc organic chemicals with log
K values of four or wore.
The FCMs listed in Table 1 were determined by staff at the SPA Environmental
Research lab-Duluth, based on a model of increasing bioaccumulation in
successive trophic levels (U.S.BPA 1991, Thomann 19B9). The application of an
PCM is a reasonable means to predict a BAP from a BCP, with order of magnitude
confidence, for most chemicals with log K values in the range of about 3.5 to
6.5 (Thomann 1989). However, predicted BAPs for chemicals with log K values
in this range are likely to have greater uncertainty than measured BAPs. Also,
FCMs are most applicable to chemicals that are not readily metabolised by
aquatic organisms. The more the chemical is metabolized, the more the specified
PCM is likely to be too large*
C. Predicted BCPs Based on Octanol/Vater Partition Coefficient
In the absence of acceptable measured BAPs and/or BCPs for llpophilic organic
chemicals, a BAP is calculated using the relationship between bioconcentration
-------
and the log,ft of the octanol/water partition coefficient. BCFs bused on log K
values will Be multiplied by the appropriate FCM to reflect bioaccuraulation due*
to exposure through the food chain.
Measured log K values are usually preferred, but if measured values aie not
available, the? are estimated using quantitative structure activity
relationships (QSAR). If more than one measured value is available an average
value can be used, or professional judgment used to select the best value.
A BCF is calculated from the chemical's log K using equation 1 front Veith and
Kosian (1983). °
Iog10 BCF - 0.79 log1Q K0tf -0.40. (1)
Where log,A K « the log of the octanol-vater partition coefficient.
10 0V
Individual BCPs calculated from equation 1 have 95 percent confidence limits of
about one order of magnitude (Veith and Kosian 1963). Thus, a predicted BCF of
100 would have 95 percent confidence limits ranging from 10 to 1000. This
level of confidence, however, does not extend to very highly lipophilic
chemicals (log K greater than 6.5). Equation 1 is more likely to over
estimate the true BCP as log K values increase above 6.5. Veith and Kosian
(1983) suggest not using equation 1 vith chemicals having molecular weights
greater than 600. The BCF model in QSAR, which is based on equation 1, places a
cap on predicted BCFs of 100,000. Equation 1 equates a BCF of 100,000 to a log
KQW of about 6.8.
0. BAF Calculation
1. A species mean BAF or BCP is calculated if more than one
acceptable (measured) normalized BAF or BCF is available for a given species.
For each chemical, the mean of the normalized species mean BAFa or BCPs is
calculated.
2. The BAF for a chemical for which field measured BAFs are
available is calculated as follows:
a. Human Health BAF * (normalised mean BAF)(6)
b. Wildlife BAF » (normalized mean BAF)(9)
Wherei 6 is the mean percent lipid for the edible portions
of game fish in the Great lakes.
9 is the mean percent lipid for whole fiah (ffame and
nongame) in the Great Lakes
3. The BAF for a chemical for which laboratory measured BCFs are
available is calculated as follows:
-------
a. Humin Health BAF - (normalized mean BCF)(6)(FCM)
b. Wildlife BAF * (normalized mean BCF)(9)(FCM)
Wherei 6 and 9 are as described above
FCM is the appropriate food chain multiplier from
ttble 1.
4. A BAF for a chemical for vhlch no measured BAF or BCF is
available is calculated aa follovs:
a. Hunuin Health BAF . (predicted BCF)(6/7.6)(FCM)
b. Vilcilift BAF • (predicted BCP)(9/7.6)(FCM)
Vherei predicted BCF is from equation 1, not to exceed
100,000
6 and 9 are as described above
7.6 is the aveiage percent lipid of the organisms
used to establish the BCF/log R relationship.
FCH is the appropriate food chain multiplier from
table 1.
VII. LITERATURE CITED
Thomann, R.V. 1989. Bioaccumuletion nodel of organic chemical distribution in
aquatic food chains. Environ. Sci. technol. 23i 699-707.
U.S. Environmental Protection Agency. 1991. Assessment and control of
bioconcentratable contaminants in surface waters. Draft. U.S.EPA, Office of
Vater, Washington, D.C.
Veith, G.D. and P. Kosian. 1985* Estimating bioconcentration potential from
octanol/vater partition coefficients. Chapter 15 In FCBs in the Great Lakes.
Hacfcey, D., R. Patterson, S. EUenreich, and H. Simmons (eds.) Ann Arbor
Science.
-------
/
8
•ow
Table 1
Food Chain Multipliers
Trophic Levels*
2
1.0
1.1
1.1
1.1
1.1
1.2
1.2
1.2
1.3
1,4
1.5
1,6
1.7
1.9
2.2
2.4
2.8
3.3
3.9
4.6
5.6
6.8
8.2
10
13
15
19
**
3
1.0
1.0
1.1
1.1
1.1
1.1
1.2
1,3
1.4
1.5
1.8
2.1
2,5
3.0
3.7
4.6
5.9
7,5
9.8
13
17
21
25
29
34
39
45
**
4
1.0
1.0
1.1
1.1
1.1
1.1
1.2
1.3
1.4
1.6
2.0
2.6
3.2
4.3
5.8
8.0
11
16
23
33
47
67
75
84
92
98
100
**
5 3.9
4.0
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
5.0
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5,8
5.9
6.0
6.1
6.2
6.3
6.4
6,5
>6.5
*Trophic level» 2 is cooplankton
3 is small fish
4 is piscivorous fish including top predator*
**For chemicals with log K values greater than 6.5 a FCH could range from O.l
to 100. Such chemicals should be evaluated individually to determine the
appropriate PCM.
-------
DRAFT
8-27-91
The following are proposed human health bioaccumulation factors
(HHBAFs) that are being suggested for use in deriving human health
tier 1 values (criteria) and tier 2 values (levels of concern) for
the GLI The chemicals are the ones that Rick Balla and I feel
are currently on the "Universe of Pollutants" of the GLI based on
the last meeting of the TWG The procedures used to derive these
HHBAFs have been discussed with Dave Maschwitz, but most of the
proposed numbers have not been reviewed by Dave. Many of these
proposed HHBAFs have been derived using information at hand
because it was not possible to conduct thorough literature
searches in the time available. Most of these contain an
excessive number of digits to prevent roundoff error in subsequent
calculations.
Charles Stephan
R£F<
CAS#O
HHBAFC
NAME"
001
077
002
003
089GNRX
ALN
078
114
1 1 5AGNR
116G
072N
83-32-9
208-96-8
107-02-8
107-13-1
309-00-2
7429-90-5
120-12-7
7440-36-0
7440-38-2
1332-21-4
56-55-3
483 61
227 4'
430 21
60. 01
1,157,358'
1,3543
0.5"
1.0'
396, 1981
Acenaphthene
Acenaphthylene
Acrolein; 2-propenal
Acrylonitrile
Aldrin
Aluminum
Anthracene
Antimony
Arsenic
Asbestos
1,2-Benzanthracene;
benz[a]anthracene
004Y
71-43-2
15.43 Benzene
-------
005
074N
075N
079
073N
117R
043
018
042
047
041
C67R
118AGNRX
006Y
091GLNR
007
022
051
016
019
023Y
020
024
92-87-5
205-99-2
207-08-9
191-24-2
50-32-8
7440-41-7
111-91-1
111-44-4
108-60-1
75-25-2
101-55-3
85-68-7
7440-43-9
56-23-5
57-74-9
108-90-7
59-50-7
124-48-1
75-00-3
110-75-8
67-66-3
91-58-7
95-57-8
4.46'
1,611,4503
1,611,450'
7,960 to
7,960,200>'k
1,199,970'
19'
1 23'
13 73
5 883
23 5j
18,638'
1, 160'
11'
37 ^
263,250'
59 2'
90. 01
18'
4.32j
1.92'
7.51
497. 4'
16'
Benzidine
3,4-Benzofluoranthene,
benzo[b]fluoranthene
11,12-Benzofluoranthene,
benzo[k]fluoranthene
1,12-Benzoperylene;
benzo[ghi]perylene
Benzo[a]pyrene;
3,4-benzopyrene
Beryllium
Bis (2-chloroethoxy)methane
Bis (2-chloroethyl) ether
Bis (2-chloroisopropyl) ether
Bromoform; tribromomethane
4-Bromophenyl phenyl ether
Butyl benzyl phthalate
Cadmium
Carbon tetrachloride,
tetrachloromethane
Chlordane
(also CAS# 12789-03-6)
Chlorobenzene
p-Chloro-m-cresol,
4-chloro-3-methylphenol
Chlorodibromomethane
Chloroethane
2-Chloroethyl vinyl ether
Chloroform; tnchloromethane
2-Chloronaphthalene
2-Chlorophenol
-------
040
AC
1 1 9AGRXY
076N
120AGLNR
121AR
CR
094GLN
093GLN
092GLNR
066GRY
G
082
068GR
025XY
026XY
027XY
028
048Y
013
010
029
030
7005-72-3
2921-88-2
7440-47-3
218-01-9
7440-50-3
57-12-5
94-75-7
72-54-8
72-55-9
50-29-3
117-81--'
333-41-5
53-70-1.
84-74-2
95-50-1
541-73-1
106-46-7
91-94-1
75-27-4
75-34-3
107-06-2
75-35-4
156-60-5
1,712'
6,8971
0 5'
243,156'
0.0*
1.0*
54 I1
486,804'
9, 128,274'
2,296,650'
14 25 to
14,250i'"
173 O1
7,960 to
7,960,200''"
3,115j
175. 21
1711
1871
2083
14. 2>
8.2'
2 51
14. 83
8 3>
4-Chlorophenyl phenyl ether
Chlorpyrifos
Chromium
Chrysene
Copper
Cyanide
2,4-D; 2,4-Dichloro-
phenoxyacetic acid
4,4'-DDD; p,p'-DDD; 4,4'-TDE,
p,p'-TDE
4,4'-DDE, p,p'-DDE
4,4'-DDT; p,p'-DDT
DEHP; di (2-ethylhexyl)
phthalate
Diazinon
1,2:5,6-Dibenzanthracene;
dibenz[a,h]anthracene
Dibutyl phthalate;
di-n-butyl phthalate
1,2-Dichlorobenzene
1,3-Dichlorobenzene
1,4-Dichlorobenzene
3,3'-Dichlorobenzidine
Dichlorobromomethane;
bromodichloromethane
1,1-Dichloroethane
1,2-Dichloroethane;
vinylidene chloride
1,1-Dichloroethylene
1,2-trans-Dichloroethylene
-------
031
032
033
090GLNR
070R
034
071R
060
059
035
036
069RY
037
NRX
095
096
097
098GRXY
099
038
039Y
080
G
GR
120-83-2
78-87-5
542-75-6
60-57-1
84-66-2
105-67-9
131-11-3
534-52-1
51-28-5
121-14-2
606-20-2
117-84-0
122-66-7
115-29-7
959-98-8
33213-65-9
1031-07-8
72-20-8
7421-93-4
100-41-4
206-44-0
86-73-7
16984-48-8
86-50-0
73 83
11 8'
5 76'
222, 9249
1461
187 51
71 41
33 1'
5 77'
12 2l
16 41
14 25 to
14,250:
67 8'
333 6m
333 6<
333 6'
83.4'
7,4581-n
23 8'
110'
10,950'
l,929j
41.5'
2,4-Dichlorophenol
1,2-Dichloropropane
1,3-Dichloropropylene;
1,3-dichloropropene
Dieldrin
Diethyl phthalate
2,4-Dimethylphenol;
2,4-xylenol
Dimethyl phthalate
4,6-Dinitro-o-cresol,
2-methyl-4,6-dinitrophenol
2,4-Dinitrophenol
2,4-Dinitrotoluene
2,6-Dinitrotoluene
Dioctyl phthalate;
di-n-octyl phthalate
1,2-Diphenylhydrazine
Endosulfan, thiodan
alpha-Endosulfan
beta-Endosulfan
Endosulfan sulfate
Endrin
Endrin aldehyde
Ethylbenzene
Fluoranthene
Fluorene; 9H-fluorene
Fluoride
Guthion
-------
100GNRX
101GNX
009LN
052Y
NX
102
103
105
053
012
083
GLNR
054
122AGNRX
104GR
CR
123AGLNR
GNR
046
045
044Y
GLNR
055
76-44-8
1024-57-3
118-74-L
87-68-J
608-73-L
319-84-6
319-85-7
319-86-8
77-47-1
67-72-L
193-39-5
7439-89-6
78-59-1
7439-92-1
58-89-9
121-75-5
7439-97-6
72-43-5
74-83-9
74-87-3
75-09-2
2385-85-5
91-20-3
28,798*
12,507i
239,154"
3,038*
l,954h
l,954h
l,954h
l,954h
98. 61
1,139*
4,997 to
4,996,800'
8 761
3.8e
1,954*
26 41
17,03s1
2.45>
1.683
3.053
5,969,130*
99. 01
Heptachlor
Heptachlor epoxide
Hexachlorobenzene
Hexachloro-1,3-butadiene;
hexachlorobutadiene
Hexachlorocyclohexane; BHC,n
alpha-Hexachlorocyclohexane;
alpha-BHC
beta-Hexachlorocyclohexane;
beta-BHC
delta-Hexachlorocyclohexane,
delta-BHC
Hexachlorocyclopentadiene
Hexachloroethane
Indeno[1,2,3-cd]pyrene;
2,3-o-phenylene pyrene
Iron
Isophorone
Lead
Lindane, gamma-BHC,
ganuna-hexachlorocyclohexane;
Malathion
Mercury
Methoxychlor
Methyl bromide; bromomethane
Methyl chloride, chloromethane
Methylene chloride;
dichloromethane
Mirex; dechlorane
Naphthalene
-------
124AGRXY
056
057
058
061
062
063
LN
AGR
1068GLNR
NYZ
064AY
081
065RY
N
084Y
125AGRY
126R
129LN
NX
N
015
085N
127
086Y
7440-02-0
98-95-3
88-75-5
100-02-7
62-75-9
86-30-6
621-64-7
29082-74-4
56-38-2
608-93-5
87-86-5
85-01-8
108-95-2
39801-14-4
129-00-0
7782-49-2
7440-22-4
1746-01-6
634-66-2
95-94-3
79-34-5
127-18-4
7440-28-0
108-88-3
3.5'
2 55l
8 64-
9.201
0 093'
271 21
3,693
2,315,538'
315 63
2,132,232"
22, 116'
755 4*
4,80s1
2,0731'°
2,791,446'
10,669'
4 8f
0 5'
43,7149
10. O1
61. 21
130"
47. 63
Nickel
Nitrobenzene
2-Nitrophenol
4-Nitrophenol
N-Nitrosodimethylamine
N-Nitrosodiphenylamine
N-Nitrosodipropylamine;
N-nitrosodi-n-propylamine
Octachlorostyrene
Parathion
PCBs, polychlorinated
biphenyls
Pentachlorobenzene
Pentachlorophenol
Phenanthrene
Phenol
Photomirex
Pyrene
Selenium
Silver
2,3,7,8-TCDD
1,2,3,4-Tetrachlorobenzene
1,2,4,5-Tetrachlorobenzene
1,1,2,2-Tetrachloroethane
Tetrachloroethylene
Thallium
Toluene; methylbenzene
-------
113AGNRX
008XY
Oil
014
087
021Y
088
8001-35-2
120-82-1
71-55-6
79-00-5
79-01-6
88-06-2
75-01-4
128AGRXY 7440-66-6
11.31
14 3'
21.21
42 6l
3 73'
3 5'
Toxaphene
1,2, 4-Tnchlorobenzene
1,1,1-Tnchloroethane
1,1, 2-Tnchloroethane
Trichloroethylene;
trichloroethene
2,4,6-Trichlorophenol
Vinyl chloride;
chloroethylene;
chloroethene
Zinc
This column identifies references concerning the reason the
substance is on the list Numbers are from the list of 126
given on pages 1Z<5 and 736 of Appendix A of Part 423 of the 7—1-
90 edition of the Code of Federal Regulations Even though 129
is the highest number on the list/ numbers 17, 49, and 50 were
delisted by the U S. EPA; thus there are only 126 items on the
list. The three substances that were delisted were
017
049
050
5^2-88-1 Bis(chloromethyl) ether
75-69-4 Tnchlorofluoromethane;
fluorotrichloromethane
75-71-8 Dichlorodifluoromethane
8,
The letters refer to the following'
A. An aquatic life criterion based on the 1985 National
Guidelines was published by the U.S. EPA.
PCBs were on the list of 126 as Aroclors 1016, 1221, 1232,
1242, 1248, 1254, and 1260, which were numbers 106 to 112
Rather than possibly imply that the best way to deal with
PCBs is as Aroclors, they are listed here under the
generic term "PCBs".
This substance was added on a case-by-case basis.
A Specific Objective is given for this item in Annex 1 of
the Great Lakes Water Quality Agreement.
This is on list IA or IB in the report "Categorization of
Toxics in Lake Ontario" (July 18, 1988).
This is on list IA or IB in the report "Categorization of
Toxic Substances in the Niagara River" (June 1990).
C.
G,
N,
-------
R. This is listed in the "Red Book" (1976).
X. This is on list 1C, ID, or IE in the report
"Categorization of Toxics in LaKe Ontario" (July 18,
1988)
Y. This is on list 1C, ID, or IE in the report
"Categorization of Toxic Substances in the Niagara River"
(June 1990)
Z. This is a substance for which a human health criterion was
published in a 1980 criteria document, but which is not on
the list of 126
3 This column gives a CAS number for the chemical Chlordane is
the only chemical on the list for which two CAS numbers are
commonly used by chemists
" This column gives the value that is being proposed as the human
health bioaccumulation factor for use in the GLI.
d This column gives one or more names for the chemical. If more
than one name is given, the first name was selected as the
preferred name; other names given are synonyms that might be
useful when comparing lists or checking CAS numbers Synonyms
are separated by semicolons because parentheses, brackets,
commas, and colons are used in some rules for naming chemicals.
The preferred names are not claimed to be definitive, they were
selected because they are used in the list of 126, are used in
the CAS Registry, and/or should allow chemists to identify the
specific chemical intended Common names are used for
pesticides; most other names follow one or more generally
accepted sets of rules for naming chemicals. In CAS
nomenclature, "butyl" means "n-butyl", "octyl" means "n-octyl",
"propyl" means "n-propyl", etc , thus when it is not a straight
chain, the chemical name will explicitly say so. ""
" This value is recommended on the basis that it was the value
used for the edible portion of freshwater fish in the 1980
criteria document
£ Revised from the value of 16 that is in the 1980 criteria
document based on a reinterpretation of the data reported by
Adams (1976) .
g This HHBAF is based on field data from Lake Ontario for fish at
trophic level 4.
" The HHBAF for this chemical was set equal to the HHBAF that was
derived for lindane, which is an isomer
1 This HHBAF is based on results of a bioconcentration test. If
this chemical is metabolizable and the FM used in the derivation
is greater than 1.0, the BAF is probably too high.
-------
' This HHBAF is based on a calculated value for Log P. If the
chemical is metabolizable, the BAF is probably too high,
especially if the FM used is greater than 1.0
* Because the "typical Log P" is greater than 6.5, the FM is
assumed to be between 0 1 and 100. Although a predicted BAF
cannot be calculated, lower and upper limits on the BAF can be
calculated
1 The HHBAF for this chemical was set equal to the HHBAF that was
derived for DEHP
m The HHBAF for this chemical was set equal to the HHBAF that was
derived for the alpha and beta isomers of this chemical.
" This HHBAF is prooably too low because the measured (field) BAF
for dieldrin was ,nuch higher than the predicted BAF
0 This HHBAF seems Tiuch too high
-------
HUMAN HEALTH
BIOACCUMULATION FACTORS
HHBAF MAKE
9,128,274 4,4'-DDE; p,p'-DDE
5,969,130 Mirex; dechlorane
2,791,446 Photomirex
2,315,538 Octachlorostyrene
2,296,650 4,4'-DDT; p,p'-DDT
2,132,232 PCBs; polychlorinated biphenyls
1,611,450 3,4-Benzofluoranthene;benzo[b]fluoranthene
1,611,450 11,12-Benzofluoranthene;benzo[k]fluoranthene
1,199,970 Benzo[a]pyrene; 3,4-benzopyrene
1,157,358 Aldrin
486,804 4,4'-DDD; p,p'-DDD; 4,4'-TDE; p,p'-TDE
396,198 l,2-Benzanthracene;benz[a]anthracene
263,250 Chlordane (also CAS# 12789-03-6)
243,156 Chrysene
239,154 Hexachlorobenzene
222,924 Dieldrin
43,714 2,3,7,8-TCDD
28,798 Heptachlor
22,116 Pentachlorobenzene
18,638 4-Bromophenyl phenyl ether
17,035 Methoxychlor
12,507 Heptachlor epoxide
10,950 Fluoranthene
10,669 Pyrene
-------
HUMAN HEALTH PAGE 2
BIOACCUMULATION FACTORS
NAME
7,960 to l,12-Benzoperylene;benzo[ghi]perylene
7,960,200
7,960 to 1,2:5,6-Dibenzanthracene;
7,960,200 dibenz[a,h]anthracene
7,458 Endrin
6,897 Chlorpyrifos
4,997 to Indeno(l,2,3-cd]pyrene;2,3-o-phenylene pyrene
4,996,800
4,805 Phenanthrene
3,115 Dibutyl phthalate;di-n-butyl phthalate
3,038 Hexachloro-l,3-butadiene;hexachlorobutadiene
2,073 Phenol
1,954 Hexachlorocyclohexane;BHC,n
1,954 aIpha-Hexachlorocyclohexane;aIpha-BHC
1,954 beta-Hexachlorocyclohexane;beta-BHC
1,954 delta-Hexachlorocyclohexane;delta-BHC
1,954 Lindane; gamma-BHC;
gamroa-hexachlorocyclohexane
1,929 Fluorene; 9H-fluorene
1,712 4-Chlorophenyl phenyl ether
1,354 Anthracene
1,160 Butyl benzyl phthalate
1,139 Hexachloroethane
14.25 to Dioctyl phthalate;di-n-octyl phthalate
14,250
14.25 to DEPH; di(2-ethylhexyl) phthalate
14,250
-------
WILDLIFE MATERIALS
-------
F - e - S> 1 FR J __ 1..-4 ,:_0 S W_j_j=. C o 1-1 S._i_ri DM_P
- O IL
September 6, 1991
GREAT LAKES WATER QUALITY INITIATIVE
PROCEDURE FOR DERIVING CRITERIA FOR PROTECTION OF WILDLIFE
The Great Lakes Wildlife Criterion is the concentration of a substance which if not exceeded protects
wildlife populations inhabiting the Great Lakes Basin from adverse effects resulting from mgcstion of
surface waters and from tngesiion of aquatic life taken from surface waters of the Great Lakes Basin
A tiered approach shall be used to derive criteria for the protection of wildlife populations in the Great
Lakes Basin Tier 1 criteria stall be derived for any chemical of concern for which the minimum
database requirements can be met. When insufficient data are available to derive Tier 1 criteria. Tier 2
criteria shall be domed for those chemicals of concern which satisfy the Tier 2 minimum ternary
database requirements.
Tier 1 Minimum Tenacity Database For Wildlife Criterion Development To derive a
Tier 1 criteria for wildlife, the minimum toxictiy database required shall provide enough data to generate
a subchronic or chronit dose-response curve for any given chemical for both mammalians and avians.
From this data, a no observed adverse effect level (NOAEL) or lowest observed advene effect level
(LOAEL} value shall be obtained for criteria calculation. Peer reviewed field studies of wildlife species
shall be considered when such studies are available. In the absence of acceptable wildlife field studies, a
Tier 1 criterion sbal! he developed from mammalian cod avian laboratory data as follows.
Lab Mammals
At least one well conducted subchronic study in one mammalian species consisting of repeated
oral exposure over a '90 day or greater period, or
At least one well conducted reproductive or developmental effects study tn one mammalian
species via the oral route
-------
•SEP— 6. — 3 1 F" P I 1 <* : O 9 U I s o o n s. i r-. DMF
P . C1 3
Lab Birds
At least one well continued study of 28 days or greater in one avian species designed to observe
iubchromc a& well as reproductive or developmental effects,
Laboratory data shall be used in tombinauon with field studies whenever possible Preference shall be
given to laboratory studies with wildlife species over traditional laboratory animals to reduce
uncertainties in making mtervpecics extrapolation*
Tier 2 Minimum Toriaty Database For Wildlife Value Development All data that do not
meet the Tier I minimum database requirements shall be considered in the development of Tcr 2
values Subchronic or chronic toxicity data
-------
P- •£-•=•! FRI 1. 4 : e 9 W i s c o i-. =. i r, DMP
<=• . O
Where. WV « Wildlife value in milligram*; per liter (mg,L)
NOAEL * No observed adverse effect level m milligrams of substance per kilogram of
bod)T weight per day (mg/kg-U) as derived from mammalian or avum studies or
as specified below.
WtA m Average weight m kilograms (kg) of the test animals
WA • Average daily volume of water consumed by the test animals in liters per
day (L/d).
SSF - Species icrniuvlty factor An extrapolation factor to account for differences in
loxicity between species (0 01 to 1 for Tier 1 and 0 001 to 1 for Tier 2)
FA - Average daily amount of fond consumed by the test animal in kilograms per
day (kg/d).
BAF » Aquatic life bioaecumulation factor in liters per kilogram (L/kg). Chosen
using guidelines for wildlife in the Procedure to Determine a Bioaccumulation
Factor, Great Lakes Water Quality Initiative.
A species WV ii calculated as the geometric mean of the WVs if more than one WV u available for a
species.
In those eases in which a no observed advene effect level (NOAEL) a available in units other than
mg/kg-d, the following procedures shall be used to express the NOAEL prior to calculating the wildlife
value:
-------
P - •= - =• 1 F R I 14:10 W i =. c o r-, = . r-, D M F R
If the NOAEL a given in milligrams of toncam per liter of water consumed (mj/L), the NOAEL
shall be multiplied by the daily average volume of water consumed by the test animals m liters per
day (L/d) and divided bv the average weight of the test animals in kilograms (kg)
If the NOAEL n given in milligrams of toxicant per kilogram of food consumed (mg/kg), the
NOAEL shall be multiplied by the average amount of food m kilograms consumed daily by the
test animals (lig/d) and divided ty the average weight of the test animals in kilograms (kg).
In those cases m which a NOAEL is unavailable and a lowest observed adverse effect level (LOAEL) is
available, the LOAEL may be substituted with proper adjustment to estimate the NOAEL. An
uncertainty factor of between one and 10 may be applied to the LOAEL, depending on the
dose-response of the adverse effect, to reduce ihc LOAEL into the range of a NOAEL If the LOAEL
is available m units other than mg/kg-d, the LOAEL shall be expressed in the same manner as that
specified for the NOAEL
In certain instances where only subchronic data arc available, a 10-fold uncertainty factor may be applied
to the NOAEL to extrapolate from subchronic to chronic levels This factor may be used when
assessing highly bioaccumuUtive chemicals where toxicokmetic considerations suggest that a bioassay of
limited length may underestimate hazard
The selection of the species sensitivity factor (SSF) shall be based on the available lexicological data and
on available physicochcmical and tojdcokJnetic properties of the substance in question This value is an
extrapolation from (he toxicity data to account for differences between species, when evidence implies
that the indicated research animal may not be the most sensitive. Outdance for choosing the SSF is
provided in the Technical Support Document.
-------
FRI 1 <4 10 Wisconsin D M P "=.01=
If drinking or feeding rates are not given in the study or studies from which a WV is being calculated,
drinking (WA) and feeding rates (FA) shall be determined from appropriate data tables for the particular
study species For studio, done with domestic laboratory animals, the following reference shall be
tonsuJtcd National Institute for Occupational Safety and Health, 1989, Registry of Tone Effects of
Chemical Substance*. When insufficient data exist for wildlife and other mammalian or avian species,
the aliomctnc equations given below shall be applied.
For mammalian species the aliomctnc equations are as follows
1 FA « 00687 x(WtA)M!
Where FA = Feeding rate of mammalian species in kilograms per day (kg/d).
WiA • Average weight in kilograms (kg) of the test animals
2. WA «• 0099x(WtA)"°
Where WA - Drtnkug rate of mammaUaa species in liters per day (L/d).
WtA - Average weight in kilograms (kg) of the test animals.
For avian specie* the allometric equation* are a» follows
i. FA - 0.0582
Where: FA • Feeding rate of avian speacs in kilograms per day (kg/d).
WtA - Average weight in kilograms (kg) of the test animals.
2. WA . 0059x(WtA)"'
Where. WA « Drinking rate of avian species In liters per day (L/d)
-------
F— •£ — ^ 1 F"RI 1 «*• : 1 1 Ul i s G o rt * i r-i D M F c . t3
WtA • Average weight in kilograms (kg) of the test animals
Toxic Equivalency Factors. Where data are available, the use of toxic equivalency factor (TEF)
values are justified and appropriate. For purposes of implementing the criteria in Table 1, th< TEF
values provided in the Implementation Procedure are used for convening concentrations of chlorinated
dibcnzodiowns, chlorinated dibenzofurans, and chlorinated biphenyls, into "equivalent" amounts of
23,7,8-TCDD The toxic equivalences are to be used when developing a total toncity for a discharge
containing a mixture of these compounds
Table 1
Great Lakes Wildlife Criteria - Tier 1
QHCTJa
DOT A Metabolites 0.044 pg/L
Mercury 0.27 ng/L
PCBs (total) 200 pg/L
2,3,7,S.TCDD 1 5 pg/L
Table 2
Great Lakes Wildlife Values - Tier 2
Substance Value
To be developed .
v \perm\wr9gliwc.bbg
. 6 -
"2-7D
-------
p _ ^ _ =» j_ fr fs i 1 >* : 1 1 U I .5 c z i-i .5 I r-i D M P - . O =
September 6, 1991
GREAT LAKES WATER QUALITY INITIATIVE
TECHNICAL SUPPORT DOCUMENT
PROCEDURE FOR DERIVING CRITERIA FOR PROTECTION OF WT1JDLIFE
OVERVIEW
Th* purpose of establishing water quality criteria for wildlife is eo
datormina surface water concentrations of toxicants that will remain
protective of wildlife populations that utilize waters of the Great Lakes
Basin as a drinking or foraging source For the purpose of these regulations,
this concentration is called the Great Lakes Wildlife Criterion (GLWC)
The GLWC is the highest aqueous concentration of a toxicant which causes no
significant reduction ir the growth, reproduction, viability or usefulness (in
a commercial or recreational sense) of a population of exposed animals which
use receiving waters for drinking or as a food source, over several
generations
To derive a CLWC, scientific literature shall be reviewed for mammalian and
avian toxicity studies vhich can be utilized to determine an appropriate GLWC
for any toxicant of concern. A tiered approach shall be used in the
derivation of these criteria.
-------
P — «£ — =•! F F I 1 •* : 1 2 Ul i A •= o r-i 5 i ri D M F
Numerical criteria (Tier 1) have been calculated according to the specified
procedures for four substances (DDT and metabolites, mercury, polychlcrinated
oipnenyls and 2,3,7,8-tetrachlorodibenzo-p-dioxin) Numerical criteria for
the first three substances have been derived, in part, from wildlife toxicity
data that was reviewed and summarized by EPA, in establishing National Water
Quality Criteria
EXPOSURE EQUATIONS
The equation used to calculate a Wildlife Value (WV) and ultimately the GLWC
is an exposure equation It's derivation is given in the following
paragraphs
The starting point of the derivation is the total intake of a toxicant The
aquation must model the intake of the toxicant in the field via food and
water. The equation includes food and water consumption rates, as well as
body weight, for the species from which the toxicity data is derived Then
the intake is set equal to a no observable adverse effect level (NOAEl.)
Therefore, the WV equation Includes a food and water intake component as
follows.
Toxicant intake through drinking water - (WV x WA)/WtA (equation 1)
Toxicant intake through food - (WV x FA x BAF)/WtA (equation 2)
- 2
-------
F — •= — ="! R F I I -* 1 7 W ( .5 r Zi r-i = i r-i D 14 F i= . 1 O
Where- WV - Wildlife value in milligrams p«r liter (mg/L)
UA - A/erag* daily volume of water consumed by the test:
animal in liters per day (L/d).
FA - Average daily amount of food consumed by the te:>t
animal in kilograms per day (kg/d)
BAF - Aquatic life bioaccumulation factor ulth units of liter
p«r kilogram (LAg)
WtA " Average weight of the test animal in kilograms (kg)
The total toxicant intake of water and food is set equal to the NOAEL in
mg/kg-d to ensur* protection of wildlife. Then, equation one and two are
combined to yield equatisn three
NOAEL > (WV x WA)/WtA + (WV x PA x BAF)/WtA (equation 3)
Factoring and rearranging produces:
WV < NOAEL x .WtA (equation 4)
WA * [FA x BAF]
To account for differences in toxicity between species, the NOAEL is
multiplied by « species sensitivity factor (5SF) ranging from 0.01 to 1 for
Tier 1 and 0 001 to 1 for Tier 2. This gives the final equation for the WV,
WV - JNQAEL x SSF1 x Wyfl (equation 5)
WA + [FA x BAF]
- 3
-------
F— «= — =•! FPI 1 4 : 1 3 LU i =. - o I-, = i n D M R
In establishing the wildlife criterion for a given substance when NOAF1 values
are available for more than one species, the lowest species WV calculated
becomes the GLWC If more chan one WV can be calculated for the same species,
the geometric mean of the WV's becomes the final criterion Prior to
calculating a WV, a variety of considerations for each variable must be
understood. These considerations are discussed in the following paragraphs
NOAEL. LOAEL
If a NOAEL is available from the scientific literature, already in proper
units, it may be directly substituted into the equation. In many instances,
however, a NOAEL is unavailable and a LOAEL is available for a particular
animal In these instances the LOAEL must be adjusted and then with proper
units can be substituted directly into the equation.
The LCAEL is adjusted by a factor of from one to ten to drop the LOAEL into
the range of the NOAEL. Experimental support of this concept is provided by
Weil and KcCollister (1963) A discussion and endorsement of this concept can
be found in Stokinger (1972) and Dourson and Start (1983) In addition, this
concept is endorsed by U S EPA in the Federal Register for Water Quality
Criteria Documents. (Federal Reglster/Vol. 45, No, 23I/Friday
November 28, 1980).
Tabla 1 demonstrates the vide variety of toxic effects that may be encountered
in toxicity studies It may provide guidance on how to assess the severity of
toxic effects that may be associated with a LOAEL or a NOAEL evaluation This
guidance has been adapted from DeRosa et al. (1985)
• 4
-------
-------
F- •=-=•! FFI 1-4:1^ W i i. c. o n s i 1-1 D M F <- . 1
S'JBCHMKIC TO CHRONIC EXTRAPOLATIONS
In certain Instances where only subchronlc data on the test species are
available, a 10-fold subchtonic uncertainty factor may be applied This
factor, when applied, would account for uncertainty in extrapolating front a
subchronic NOAEL co a chronic NOAEL This factor nay be used when assessing
highly bioaccuffiulative chemicals where toxicokinetic considerations sugg*st
that a bioassay of limited length may underestimate hazard
- 5 -
-------
F -
p I
1 •+ : 1 -*
U
s, o o n s. i r-i D M F
1 S
TASIE 1 RATING VALUES TO SANK TOXICITY
RATIISia
EFFECT
Enzyme Induction or other biochemical change with no pathological changes and no change In
organ wotgrrt*
Enzyme induction and aubcollular proliferation or ethe' change* In orgenollo* but no other
apparent effort*
Hyporptasla, hypertrophy or atrophy but no change In organ weight*
HyporplMla, hypertrophy, or atrophy with change* Ir ergon wolghti
floverelble oolluiar change* cloudy twclllng, hydropic change, or fatty change*
Necroilr or met«pl«»la with no apparent decrement or organ function Any neuropath/ without
apparent behavioral, loneory, or phyetologioal change*
Noeroan atrophy hypertrophy or motapla»la with a detectable decrement of organ function*
Any neuroprthy with a meaauroble change In behavioral, eoneory or phytlologlcal activity
Necroel* atrophy hypertrophy, or metaplaela with definitive organ dyefunetion any neuropathy
with grate ohongoe in b«h«vlor, lonsory, or motor performance, Any decrease In reproductive
capacity Any evidence of fototoxlcrty
Pronounced sathologleal change* with ievora organ dytfunctton Any neuropathy with loti of
behavioral or motor control or losi of sensory ability Reproductive tfytfunctlon Any
taratogenle effort wtth maternal to»c!ty
Death or pronounced tHo ehortonlng Any t«ratogen!o effect without dgn* ef maternal loxielty
Source. Adapted from OcRota et al. (1*85)
WEIGHT OF THE TEST ANIMAL
Th* weight of che test arlm»l IB also n*td«4 input for the model equation. If
this information La not given in the chosen study, the average weight of the
test species shall be determined from available literature , For instance, in
the derivation of the DDT criterion the average weight of a brown pelican had
to be estimated. Based on work by J. B. Dunning, 1984, an average weight of
3500 grams was chosen,
DRISKISC
FEEPING^HATES
-------
E.EF- «= — =•! FRI 1-4:15 Uisoor-isiri DNP f ,
Drinking and feeding rates of the test sp«cits are necessary Co accurately
predict exposure through chose routes, Based on the general uae of tzophtc
level 4 bioaccumulation factors (BAFs) in calculating wildlife values (see
BIOACCVXl'IATION FACTOR .se&filfln.), it is assumed in most cases that the
sensitive species identified for protection are piscivorous animals When
toxicit> data used to derive a wildlife criterion is not available fot a
piscivorous species, it shall be assumed that there is no quantifiable
difference in either body weight or food and water consumption rates between
the test species and a piscivorous species. When rates are given in the
study of choice, they may be directly substituted into the equation If this
information is not available from the chosen coxicity studies, it shall be
obtained from other appropriate literature concerning the species In some
instances, however, this information is not directly available and needs to be
estimated. The following reference may be consulted for studies done with
domestic laboratory animals. National Institute for Occupational Safety and
Health and Registry of Toxic Effects of Chemical Substances (Latest Edition)
Allometric equations shall be used to approximate feeding or drinking rates
for other mammalian and avian species These equations were adopted as
recommended by the Wisconsin TAC from the following references
Calder, William A , III and Eldon J. Braun. 1983. Scaling of Osmotic
Regulation in Mammals and Birds. American Journal of Physiology.
244-601-606.
Nagy, Kenneth A 1987 Field Metabolic Rat* and Food Requirement Scaling
in Mammals and Birds. Ecological Monographs. 57(2)'111-128
-------
lRFIl-*:15Wis.c.Or-i.ZinDMF
BIOACCUMULATION FACTOR
A bioaceumulation factor (BAT) is necessary Co properly estimate the
concentration of the chemical in the food source, based on its concentration
in the water source, The procedure to derive the BAF is specified in the
separate guidance on bio&ccumulation, entitled, Procedure to Determine a
Bioaccumulation Factor This guidance document specifies that, in general,
trophic level 4 BAFs be used In the derivation of wildlife values, although
options to use plant and lower trophic level BAFs are permitted based on the
identification of sensitive species that are not obligate piscivores.
SPECIES SENSITIVITY FACTOR
The KOAEL may be adjusted to accommodate differences in interspecies toxicity
with the use of an extrapolation factor. This adjustment may be necessary
since the information upon which a criterion is developed will not necessarily
be based on the most sensitive wildlife species In order to provide
protection for the most sensitive species, an extrapolation factor called the
species sensitivity factor (SSF) shall be used, based on the available
toxicological data and the physlcochemlcal and toxicoklnetic properties of the
substance in question The SSF Is not Intended to adjust for potential
differences between the test species and the sensitive species with regard to
body weight (VtA) and food (FA) and water consumption (Wa) The factor
selected shall reflect the uncertainty in which the available toxicity data
represent the most sensitive species to be protected. For instance, a
database containing both chronic and reproductive/developmental data for a
diversity of species may require a SSF of between 0 1 and I If these data
are from numerous species and represent the most sensitive mammalian and avian
8
-------
P— 1= — =" 1 F F I 1 «* : 1 i=. Ut I A. = O rl £ i r-i D N F (=.!•=
species, the SSF may be equal to 1 Also, when a NOAEL for a species that is
used to derive a WV is greater than the lD5o value of another species, then the
difference shall be compensated for by the SSF.
To determine the proper range for the species sensitivity factor, LD.. data
was reviewed for approximately SO chemicals and chronic toxicity data was
reviewed where available Table 2 contains LD.. data for nine pesticides, FOB
and 2,3,7,8-tetrachlorodibenzo-p-dioxin This table demonstrates how coxicity
from certain chemicals differs between species Table 3 contains chronic
toxicity data for organomercury compounds in mammalian species These data
support both the use and the range of a species sensitivity factor
For Tier 1 wildlife values, the SSF shall be used for extrapolating toxicity
data across species within a class The Tier 1 SSF is generally not intended
for interclass extrapolations because of the poorly defined uncertainties in
comparing toxicokinetic and toxicodynamic parameters in mammals and avians
An interclass extrapolation employing an SSF may be used for a given chemical
if it can be supported a validated biologically-based dose-response model or
by an analysis of interclass toxicological data, Incorporating the endpoints
in question, for a chemical analog that acts under the same mode of toxic
action
For Tier 2 wildlife values, interclass extrapolation may be permitted.
Because of the uncertainties in performing Interclass extrapolations, an SSF
as low as 0.001 may be appropriate for calculating Tier 2 values.
- 9 -
-------
= E P -
— •=> 1
R I
1 -4 : 1 T
U
* c- o t~i s.
D H K
Dieldna
Zndnn
Hejuchlorobaniane
(Confd)
TABU 2 SENSITIVITY OF SPECIES BASED ON LD_0 DATA
Chemical $pee>8»
Aldnn Fulvous whittling duck
Mallard
Bobwhite
Pheasant
Mule Dear
Chlordane Mallard
California Quail
Pheasant
DDT Bulllrof
Mallard
California Quail
Japanese Quail
Pheasant
Sandhill Crane
RockDovt
Canada Goo*e
Fulvous WhmLny Duck
Mallard
California Quail
Japanese Quafl
Pheasant
Chukar
Gray Pamidfc
Rock Dove
House Sparrow
Mule Dear
Domestic Goat
Mallard
Sharp Tailed Gieute
California Quail
Pheasant
Rock Dove
Mult Deer
Domestic Goat
Mallard
Pheasant
Confidence IntevaU
29 2 [22 2-38 4]
520 '.229-1,210]
6 59 {5.00-8 66]
16,8 £14.1-20.0]
1S.8-37.S
1.200 [954-1,5101
14 1 [9.14-21 7}
24.0-72.0
72,000
72,240
S9S (430-825]
841 £607 1.170]
1,334 1894-1,990]
71,200
74,000
< 141
100-200
381
878 [6.47-11 9]
69.7 [40.0-121]
790 [216-289]
2S.3 [15.2-42.21
8 S4[ 124-62.8]
26.6 [19 2-36 9]
47 6 [34.3*66.0]
75-150
100-200
564 C2.7M17]
1 06 [0.552-2 04]
1.19 [0.857 1.65}
1 78 [1,12 2 83]
2.0-SO
605 12.S
25.0.50 0
71,414
118 [93 6-148]
10
-------
p —
240
24.0 [16.8-34J]
16 0 [4 00-64 0]
2.52 [1.82-3.50]
3.36 [2 43-4 66]
22.0-44.0
28.0-56 0
2,098
2,078
1.0-8.6
>20
0.8-11.0
87
> 2,000
79.4 [383-163]
18 9 [15 0-23 8]
84 1 [60.6-116}
35.4 [25.5-49 0]
240 [110-521]
SO 1 [16 7-150]
35.4 [8.85-141]
0.6-2 ug/Xl
2245 Uf/kg
<70 ug/kg
100-200 uff/kg
114-284 ug/kg
115 uf/kg
1,157-5,051 Uf/kg
99 0 [37.2-264]
30.8 [23.3-40.6]
70.7 [37.6-133]
19.9 [14.1-282]
85 5 [59 3-123]
23.7 [11.9-47.4)
40.0 [20.0-80.0]
23.7 [20.0-28.3]
100-316
581 [425-794]
139-240
>160
Adopted from llsltr ft. 1986», IJiltr ft. 19Mb, and Hudson tt *L. 1984,
-------
•= e p -
S. c. O n
D M R
TASLE 3 TOXICITY OF ORGANOMERCURY COMPOUNDS TO SELECTED MAMMALIAN SPECIES
Species DOM Effect
Dog
Cat
Pi*
Rhesu Monkey
Mmk
Riv«r Otr«r
0.2S
mg/kf
O.Smj/kg
O.S mj/kj
1 0 mg/kg
Sdllbirth*
Death
Stillbirths
Maternal Toxicey
Death
Death
Tible 5 f» idoptcd *rom Eitler
- 12 -
-------
F _ ,__=,! F F I 1 -4 : 2. O W i .= c o r-i =- i r- , D M
CALCVLA1J.UN UK GKKAT LAKES WILDLIFE CRITERIA
In this section the Great Lakes Wildlife Criteria for four substances are
discussed, These substances are DDT and metabolites, mercury, polychlorinated
biphenyls (PCBs), and 2 , 3,7, 8-tetrachlorodibenzo-p-dioxin (TCDD) For each
substance a brief discussion is presented of the following.
(1) The studies chosen during the review process.
(2) Input parameters to the criteria equation These parameters are
summarized in Table 4 of the Technical Support Document.
(3) A brief discussion of the LOAEL to NOAEL adjustment and the chosen species
sensitivity factor.
OPT and Metabolites
The Great Lakes Wildlife Criterion is based on the study by.
Anderson, D, W , J, R Jehl , R W, Ris through, L A. Woods, L R. Deweese
and W G Edgeconbe. 197S. Brown Pelicans Improved Reproduction Off
The Southern California Coast Science. 190:806-808
This work is based on a five year field study of brown pelicans. Reproductive
success of these birds was monitored and the concentration of DDT in their
diet and eggs was measured From the results, a LOAEL of 0 15 tug/kg w«t
weight in food is inferred. At this dietary level there was successful
reproduction but at levels too low for population stability Therefore, a
- 13
-------
_ ,g — =» I F" F I 1 4 : 2 1 Ul I .B. <=. o r-i s. i ri D M R ^ <- . O -*
LOAEL to NOAJEL adjustment factor ot iU is used for criteria development. The
species sensitivity factsr is 0 1 This was chosen based on the fact chat
numerous accounts exist of DDT toxicity In birds, of which piscivorous birds
are among ehe most severely affected. It is impractical, based on th«
toxicity database, to decide that the pelican is the most sensitive species
The Wildlife Value equation and input parameters are as follows
WV • fKOAEL x SSF1 x Wt. (equation 5)
WA 4- [FA x BAT]
LOAEL from study 0 15 mg/Vg vet weight in food
LOAEL to NOAEL Adjustsent Factor- 10
NOASL: 2 82 x lO^gAg-day
Weight of animal, 3.5 kg
Feeding rate 0.1314 kg/day-dry weight*
Drinking rate 0 137 U/day
Bioaecunulation factor- 3,446,975 L/kg wet weight
Species sensitivity factor1 0 1
*Note: When feeding rates are reported in dry weight values, other tissue-
related values reported in wee weight must be converted to dry weight
values by multiplying the wet tissue weight value by five.
Conversion of LOAEL to NOAEL with proper units results in the following.
LOAEL - 0 15 mgAg in food wet weight
Adjusting the LOAEL to a NOAEL yields: 0 15/10 - 0.015 mg/kg
- 14 -
-------
p _ ^ _ =1 i FRI 14:21 Wisconsin BMP i-.O
Converting the NOAEL from a wet to a dry weight produces
0,015 x 5 - 0.075 mg/kg dry weight
NOAEL in rag/kg-day - 0 075 me/ke x 0 1314 kg/dav
3.5 kg
- 2 82 x 10*3 ng/kg.d*y
Adjusting cht BAF from a wee weight to a dry weight yields:
3.444,975 LAg x 5 - 17,224,875 L/kg
Final Equation
WV in mg/L - 2 82 x 104 mt/ka-dav x 0 I lc» x 3 S
0.137 L/day + [0,1314 kg/day x 17.224,875 L/kg]
- 4.36 x 10" mg/L
- 4 4 x 10 '• mg/L
Converting to pieogra&s/liter yields 4 4 x 10" mg/L x 10*ng/mg • 0.044 pg/L
This Wildlife Value (0.044 pg/L) is the chosen Great Lakts Wildlife Criterion
for DDT and metabolites.
- 15
-------
,£ — 9! (=• p> i 1 «4 : 2 .2 W I s o o r-i = i i-i DMR /=•
The Tier 1 minimum toxic tty database requirements for DDT criterion
development become satisfied by further noting the subchronic feeding study in
rats, by laug et al. (1950). From this study, a NOAEL of 1 ppm is darlved
Weanling rates were fed commercial -grade DDT For 15-27 weeks. The critical
toxic effect was liver Coxicity. The species sensitivity factor is 0 1, given
that there are too few good mammalian toxic I ty studies to decide that a lab
rate is the most sensitive mammal
The Wildlife Value equat'.on and input parameters are as follows:
W - r NOAEL x SSF1 x We. (equation 5}
WA + [FA X
NOAEL from study 0 05 mg/kg-day
Weight of animal- 0 2 kg
Feeding rate 0.01 kg/day
Drinking rate 0 02 L/day
Bioac cumulation factor. 3,444,975 L/kg
Species sensitivity factor* 0 1
WV - Q.OS mt/ke-dav * 0 1 x 0.2 kg
0.02 L/day + [0.01 kg/day x 3,444,975 L/kg]
- 2 9 x 104 ng/L
Converting to picograae/Liter yields 2,9 x 104 mg/L x 10* pg/mg - 29.0 pg/L
16
-------
— <= — •=< 1 FFI 14:2:2 UJ i 5 o o r-. =. I r-i DMP
*Note The wildlife value derived using ch« Laug ec al. data support the
Tier 1 dacab»»« requiremencs buc, Is not the wildlife value chosen for
deriving th» final Great Lakes Wildlife Criterion for DDT
17 -
-------
F — «=.— =>! FRI 1 ** : Z "S W I * c C ri * i r-i DHP _ r- . Ct 3
Mercury
Tht Great Lakes tfildlifo Criterion 1$ based on the following study
Heinz, G. H, 1979, Methylmercury Reproductive and behavioral effects on
three generations of mallard ducks. J. Wildl. Manage. 43.394*401
Heinz (1979) reported decreased reproductive success and altered behavior of
offspring over three generations of mallard ducks fed a dietary concentration
of 0 5 ppm (0 5 »g/kg) nethyl mercury. In determining the avian wildlife
value, a species sensitivity factor of 0 1 is chosen, based on the limited
number of avian species used in dose-response toxicity studies, and on reports
that vild waterfowl hav« been found with higher tissue levels of mercury than
had the mallards in thin study.
The Wildlife Value equation and input parameters are as follows:
WV - fNOAEL x SSF1 x Wt.
WA + [FA X BAF)
LOAEL from study: 0.5 mg/kg
LOAEL to NOAEL Adjujitnant Factor. 10
NOAEL: 8.7 x 10*' ng/kg-day
Weight of animal; 1.1 kg
Feeding rate- 0 172 kg/day
Drinking rate- 0 063 L/day
Bioaccumulation facror: 16,799 LAg wet waight
Species sensitivity factor: 0.1
-------
p _ «= — ? 1 F F I 1 -4- I .2 3 U I 3. c- o r-i = I r-i BMP
Conversion of LOAEL to NOAEL with proper units.
LOAEL - 0 5 ing/kg in food wet weight
Adjust to NOAEL - 0.5/10 - 0 05 mg/kg
NOAEL in mg/kg-d*y - 0.05 mg/kg x 0 172 kg/day
1.1 kg
- 7 8 x 10J nigAg-day
Final equation*
WV - 7 8 x 10' mg/kg-dav x 0 1 x 1 1 kg
0.063 L/d«y + [0 192 kg/day x 16,799 L/kg]
- 2.66 x 107 ng/L
Converting to nanograras/Liter yields 2 66 x 107 mg/L x 10* ng/mg -0,27 ng/L
This Wildlife Value (0.27 ng/L) !• the chosen Great Lakes Wildlife Criterion
for mercury
-------
_ t — =• 1 F R I 1 ** : 2 •* Ul I * c o r-i s i r-i D M F* t= . 1 O
The .ier 1 minimum toxic icy database requirements for criterion development
become satisfied by further noting the following studies
Wobeser, G,, N D Nielsen and B Schiefer 1976 Mercury and Mink I The
Use of Mercury Contaminated Fish as a Food for Ranch Mink Canadian
Journal of Comparative Medicint 40,30-33
Wobeser, G., N D Nielsen, and B Schiefer 1976. Mercury and Kink II:
Experimental Methyl Mercury Intoxication. Canadian Journal of
Comparative Medicine 40 34-45.
This research consisted of feeding studies in mink that covered a range of
dietary mercury levels Pathological, behavioral and developmental effects
were examined in adult female and juvenile mink. The authors reported
clinical findings of ataxia, anorexia, weight loss, convulsions and death, all
related to mercury poisoning They monitored accumulation in tissue and
reported pathological findings indicative of mercury poisoning. From these
studies a NOAEL. of 0.44 eg/kg wet weight from food was determined. At this
dose level no significant indications of mercury poisoning were reported The
dose level was based on a 731 diet ration Therefore, the NOAEL was adjusted
to 0.33 mg/kg to meet the assumptions of a 100% contaminated diet, The
species sensitivity factor is 0.1 This factor was based on a review of
available toxicological research and the lack of data for other mammals
The Wildlife Value equation and input parameters are as follows:
WV - fNQAEL x SSF1 x Wt. (equation 5)
WA + [FA x BAF]
- 20 -
-------
F — «= — =• 1 FFI 1 4 : Z. «* Wis-^oriain D M P
NOAEL from itudy: 0.44 mg/kg In food (75% of ditt)
Adjusted NOAEL 0.33 mgAg wet wiight (100% of diet)
NOAEL- 0 05 mg/kg-day
Weight of animal 1 0 kg
Feeding rat*' 150 g/day wet weight
Drinking rate. 0.099 L/day
Bioaecumulatlon factor- 16,799 L/kg wet weight
Sptciti tensitivicy factor 0 1
Conv«riion of NOAEL to proper units results In the following Inputs
NOAEL in rag/kg - 0 44 mg/kg wet weight In food based on 75% diet ration
NOAEL In mg/kg - 0.33 mg/kg wet weight baaed on 100% diet.
NOAEL in mg/kg-day - 0.33 at/kg x 0 15 kg/dav
1.0 kg
- 0 05 mg/kg-day
final Equation*
WV in mg/L • 0.05 ma/kg-day x 0 1 x 1 0 k«
0.099 L/day + [0 13 kg/day x 16,799 L/kg]
- 2.0 x 10-* ng/L
Converting to nanograms/Liter yleldi 2.0 x 10* ag/L x 10* ng/mg - 2 0 ng/L
-------
F — €. — SI F" R I 1 4 : ;i 5 W i * c. o ri .£ i t"i DMR «= . 1
*Note. The wildlife value derivtd using the Wobeser «c *1. d*c* support ch*
Tier I database requirements but, is nej: cha wildlif* valu* chosen for
deriving the final Great Lakes Wildlife Criecrion for mercury.
22
-------
F— to - =• 1 F" F I 1 -+ I ^ 5 U I A c O n S. I r-i D N P r-.l
Polvchlorlnated Biphenyla f?CB)
Trie Great Lakes Wildlife Criterion is based on th* following study
Flacnow, N S and L K Karstad 1973 Dietary Effects of Polychlorinaeed
Biphenyls on Mink Canadian Journal of Comparative Medicine 37:391-
400
This paper describes a nine monch study of dietary effects of PCS on mink It
was previously chosen by EPA to assess the reproductive toxieity of PCBs to
mink, The LOAEL indicated from this study is representative of extreme
reproductive failure, Clinical signs and pathological lesions were also
manifested. The species sensitivity factor of 1 was based on the available
toxicological database for PCB, A wide range of species have been tested and
the mink is the most sensitive of all wildlife species tested to date
The Wildlife Value equation and input parameters are as follows
WV - rNOA£L x SSF1 x WgA (equation 5)
WA * [FA x BAT]
from «t-uHy 0 (& »g/fcg in food
LOAEL to NOAEL Adjustment Factor. 10
NOAEL- 9 6 x 10J mg/kg-day
Weight of animal: 1.0 kg
Feeding rate. 150 g/day wet weight
Drinking rate: 0 099 L/day
-------
p — & — •=•! FFI 1 *» : .2 «=. W i 3- c o n =. i r-i DMF ^ i=- . 1 •*
Sioaccumulatior. facccr: 3,198,348 L/kg wet weight
Species sensitivity factor 1
References for feeding rate are taken from the following sources
Bleavins, H R. and Aulerich, R J, 1931 Feed Consumption and Food
Passage Time in Mink and European Ferrets, Lab Animal Sci, 31(3)
268-269
Schaible, P,J 1970. Nutrition and Feeding Sect III In. Blue Book
of Fur Farming Milwaukee, VI; Editorial Service Co
Conversion of LOAEL to NOAEL with proper units,
LOAEL - 0 64 mg/kg in food wet weight
Adjust to NOAEL - 0.64/10 - 0 064 fflg/kg
NOAEL in mg/kg -day - 0.064 mg/kt x 0.13
1.0 kg
• 9 6 x 10J Bg/kg-day
- 24
-------
F— e — 91 FRI I -4 : :i €. Wisconsin D M R
Final Equation1
WV in mg/L - 9 6 x 10* mz/kg-dav x 1 x 1 0 lee
0 099 L/day + [0.15kg/day x 3,198,348 L/kg]
- 2 0 x 104 mg/L
Converting to picograms/Liter yield* 2 0 x 10* mg/L x 10' pg/mg - 20 0 pg/L
This Wildlife Value (20.0 pg/L) la the chosen Great Lakes Wildlife Criterion
for total PCS*
The Tier 1 minimum toxicity database requirements for PCB criterion
development become satisfied by further noting the subchronic feeding study In
chickens by Britton and Huston (1973). White Leghorn hens were fed diets
containing 0, S, 10, 20, 40 or 80 ppm PCB as Aroclor 1242 for 6 weeks From
this study, a LOAEL of 5 ppm was noted The critical toxic effect was
hatchability.
25
-------
F— •£—•?! F R I 1 •* I — 7 W I » C O n a i n DNR r= . 1 >=
2.3.7.8-TCDD
The Great Lakes Wildlife Criterion is based on the following studies.
Nosek, J.A.. J.R. Sullivan, T.E. Amundson, S R Craven, Ltt Miller,
A G Fitzpat-ick, M.E. Cook and R.E Peterson. 1991.
Embryotoxicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Ring-
Necked Pheasants
Nosek, J A , J R Sullivan, S.S, Hurley, J R Olson, S.R. Graven, and
R E Peterson 1991 Metabolism and Disposition of 2,3,7,8-
Tetrachlorodibenzo-p-dioxin in Ring-Necked Pheasant Hens, Chicks,
and Eggs,
Nosek, J.A., J.R Sullivan, S S Hurley, S.R Craven, and R E Peterson
1991 2,3,7,8-Tetrachlorodibenzo-p-dioxin Toxicity in Ring-Necked
Pheasants Hens
These papers have been (submitted to journals for publication and are chapters
in The Effects of Exposvire to 2,3,7,8-TetrachlordlbeiUo-p-dioxln in
Ring*necked Pheasants, J. A Nosek, 1991. Ph.D. dissertation, University of
Wisconsin, Madison.
The work done by Nosek nt. al represents a comprehensive laboratory toxicity
Investigation using a wildlife species and therefore shall be given preference
over other animal dioxin toxicity studies. Based upon the LOAEL for adult
pheasants, reproductive performance was impaired and hatchability decreased
The species sensitivity factor of 0 1 Is based on the available toxicologlcal
. ae .
-------
F— «£ — =•! FFI 1 •+ : .2 T Uis.zoi-ijs.ir-. r>MF , «- . 1
database for TCDD, There is evidence that ch« pheasant is not the most
sensitive of all wildlife species chat have been evaluated, Current research
on the wood due* indicates that it may have a lower threshold of toxicity by
at least one order of magnitude (Donald H White, pers, com )
The Wildlife Value equation and input parameters are as follows
WV - fNOA£L x SSFi x Wt» (equation 5)
WA + [PA X BAF]
LOAEL from study 0 01 mg/kg vet weight
LOAEL to NOAEL Adjustment Factor 10
NOAEL 0 29 x 10"3 mg/kg-day
Weight of Animal: 1 0 kg
Feeding Rate 0.0582 kg/day dry weight*
Drinking Rate: 0 059 L/day
Bioaccumulation Factor: 62,571 L/kg wet weight
Species Sensitivity Factor. 0.1
*Noce When feeding rates are reported in dry weight values, other tissue-
related values reported in wet weight must be converted to dry
weight values by multiplying the wet tissue weight value by five
Conversion of LOAEL to NOAEL with proper units results in the following
LOAEL - 0 01 mg/kg vet weight
Adjusting the LOAEL to a NOAEL yields- 0 01/10 - 0.001 mg/kg
- 27 -
-------
p _ £ _ a i p- p I 1 ** S ^ *? W i s. C O 1-1 -S i r-i D M P __ <=.!=•
Converting the NOAEL from a wet to dry weight produce!
0.001 x 3 - 0 005 mg/kg dry weight
>30AEL in ntg/kg-day - 0 005 ma/ki x 0 0582 kg /day
I 0 kg
- 0.29 x 10J mg/kg-day
Adjusting the BAF from a wet weight to a dry weight yields
65,571 LAg x 5 - 327,855
Final Equation
WV in mg/L - 0 29 > 10 * me/kg-dav x 0 1 x 1 0 kg
0 059 L/day + [0.0582 kg/day x 327,855 LAg]
- 1.51 x 10-* mg/L
Converting to nanograini/litar : 1.5 x 10*' mg/L x 10* pg/ng -15 pg/L
This Wildlife Value (1,3 pg/L) if tha chosen Great Lakts Wildlife Criterion
for 2,3,7, 8. TCDD,
28 -
-------
R — >~ — =• i R F I 1 -4 : .2 £• W I =. c o r-i =. i r. DMF F.I
The Tier 1 minimum toxicity database requirements for 2,3,7,8-TCDD criterion
development become satisfied by further noting the following mammalian studies
in rhesus monkeys. Studies by Schantz et al. (1979) and Allen et al, (1979)
suggest that rhesus monkeys are more sensitive to 2,3,7,8-TCDD than rats. In
a continuation of these rhesus monkey studies, Bowman et al. (1989a, 1989b)
evaluated the effects of 5 and 25 ppt 2,3,7,8-TCDD in feed on reproduction and
on behavior, respectively Breeding of the animals after 7 and 24 months of
exposure resulted in impalrtd reproductive success with 25 ppt (LOAEi.) but not
with 5 ppt (NOAEL) in feed (approximately 0 67 and 0 13 ng/kg-day,
respectively) In deriving a wildlife value from these data, a species
sensitivity factor of 1 0 is used.
The wildlife value derived from rhesus monkeys is not chosen as the Great
Lakes Wildlife Criterion.
Z97
29 -
-------
p — t — =• 1 F R I 1 4 : 2 •?> Ul i * c o n » i r-i DM
TAUll 4 WMVIAllY Of INPUT WUWMgTWS TOR OEIWAT1CV 0> 1>tf OMAT UKM WllflU^, CIUTIWA
NOA1L
OR LOASt AOJUSTCS IOOY D» MCISC
U>ACl AOJUF'VIENT VOAtl WDOIIT RATt
suM'Asff SsSJSuSi PACTDH ytJn.<»y kj ^jnj
j
5Df * 0:j 10 JM«tD IS 0117
M£TAi!P<.T'S «>(/k|i
we* wtljhl
J
MuriiKv 0^4 NOT io»io it OOM
n^kl APPUCAtU
3
KB OW 10 »»«10 10 0099
££«,
z.i78Rnn 0,0: in OMtio 10 oxue
r525.VC
HA*^
let/day
0,1314
4fy wtlghf
Oil
W*l W«|tlt
O.IS
»•« wetglM
oou
WOACCUMUIATION
PACTOH
I/*!
3444,97S
*f< welghi
16 TM
W*.«i|hi
3 I9t^iS
**W<|ht
61,5?)
SfECIHC
KNSHVITY
FACTOR
a i
a*
,0
01
• 30 -
-------
F_ ,£_=,! FRI 1 4 : r: •=! Ul i s. c o r-> .=. i r-i IJNR r= . ^
REFERENCES USED FOR THE GREAT LAKES WILDLIFE CRITERIA
Allan, J R et al. 1979 Reproductive effects of halogenated aromatic
hydrocarbons on nonhuaan primates. Ann NY Acad Scl. 320.419-425
Anderson, D W., J. R Jehl, R. W. Risebrough, L. A. Woods, L. R Deweese
and W G Edgecomba. 1975 Brown Pelicans: Improved Reproduction Off The
Southern California Coast. Science. 190:806-808,
Bleavlns, M. R and R J. Aulerich, 1981, Feed Consumption and Food
Passage Time in Mink and European Ferrets Lab Animal Sci 31(3).268-269
Bowman, RE , et al. 1989 Chronic dietary intake of 2,3,7,8-tetrachlordibenzo-
p-dioxin (TCDD) ac 5 or 25 parts per trillion in the monkey TCDD kinetics
and dose-effect estimate of reproductive toxicity. Chemosphere
18(l-6):243-252
Bowman, RE , et al. 1989. Behavioral effects in monkeys exposed to 2,3,7.8-
TCDD transmitted maternally during gestation and for four months of
nursing Chemosphere 18(1-6):235-242.
Britton, W M and T.M. Huston> 1973 Influence of Polychlorinated Biphenyls is
the Laying Hen, Poultry Science 52 1620>1624
Calder III, W. A., and E J Braun. 1983 Scaling of Osmotic
Regulation in Mammals and Birds. American Journal of Physiology
244.601-606.
DaRosa, C. T , J, F. Stara, and P. R. Durkin. 1985. Ranking Chemicals
-------
P — € — =» 1 FPI 1 «* : ^ Q 14 i s o o n * i ri D M P
Basad on Chronic Foxicicy Data Toxicology and Industrial Health
Dour son, M L and J F Scar* 1983. Regulatory History and Experimental
Support of Uncertainty (Safety) Factori. Regulatory Toxicology and
Pharmacology 3 -224- 2 38
Dunning. J 5 1964 Body Weight! of 686 North American Birds,
Monograph #1, Western Bird Banding Association.
Eisler, R 1986a Dioxin Hazards to Fish, Wildlife, and Invertebrates A
Synoptic Review, U.S. Fish and Wildlife Service Biological Report.
85(1 8) 37pp
EUler, R. 1986b Polychlorinated Biphenyl Hazards to Fish. Wildlife, and
Invertebrates' A Synoptic Review. U.S. Fish and Wildlife Service
Biological Report. 8S(1.7) 72 pp.
Eisler, R. 1987. Mercury Hazard to Fish, Wildlife and Invertebrates- A
Synoptic Review U S Fish and Wildlife Service Biological Report.
85(1.10): 90pp.
Heinz, G. H. 1979. Methylmercury : Reproductive and behavioral effects on
three generations of mallard ducks. J. Wildl Manage, 43.394-401.
Hudson, R. H , R X Tucker, and M. A. Haegele. 1984 Handbook of
Toxicity of Pesticides to Wildlife, U.S. Fish and Wildlife Service,
Resource Publication *153, 90 pp.
-------
F - •= — =• 1 F" F I 14:31 W I s. o z. f-i = i r-i D M P ,= . .
Laug, E., A, Nelson, 0 Fitzhugh and F Kunze 1950 Liver cell alteration and
DDT storage In che fat of the rat induced by dietary levels of 1-50 ppra
DDT. J. Pharmacol Exp Therap 98'268-273
Nagy, K. A 1987 Field Metabolic Rate and Food Requirement Scaling
in Mammals and Bird* Ecological Monograph* 57(2).111-128
Mosek, J A, 1991. The Effects of Exposure to 2,3,7,8-Tetrachlorodibenzo-
p-dioxin in Ring-necked Pheasants Fh D Dissertation, University of
Wisconsin, Madison,
Nosek, J A , J R Sullivan, T,£. Aoundson, S R Craven, L.M, Miller,
A. G. fitzpatrick, M E Cook and R.E, Peterson 1991 Embryotoxicity of
2,3,7,8-T«trachlorodibenzo-p-dioxin in Ring-Necked Pheasants (In
preparation)
Nosek, J A , J R Sullivan, S S. Hurley, J.R Olson, S R, Craven, and R E
Peterson. 1991. Metabolism and Disposition of 2,3,7,8-
Tetrachlorodibenzo-p-dioxin in Ring-Necked Pheasant Kens, Chicks, and Eggs
(In preparation)
Nosek, J A., J.R. Sullivan, S.S. Hurley, S,R, craven, and R E, Peterson 1991
2,3,7,8-Tetrachlorodibanzo-p-dloxin Toxicity in Ring-Necked Pheasants Hens
(In preparation)
Platnow, N.S and L. H. Karstad 1973. Dietary Effect! of Polychlorinated
Biphenyls on Mink Canadian Journal of Comparative Medicine 37 391-400
-------
p— e — •=• 1 F F I 1 4 : 3: 1, U4i.scor-i.sin D M P* . !=•._•*
Registry of Toxic Effects of Chenical Substances, 1978 Nation*! Institute fo
Occupational Safety and Health
Safe, S 1990 Folychlorinactd Biphenyls (PCBs) , Dibenzo-p-Dioxins (PCDDs) ,
Dibenzofurans (PCCFs), and Related Compounds Environmental and
Mechanistic Considerations which Support the Development of Toxic
Equivalency Factor* (TEFs), CRC Press. Critical Reviews in Toxicology
21
-------
F — t — i* 1 FRI 1 -* : 3 r: Ul I s. C o ri * * r-, DMP
Weil, C,S , and D D, McCollister, 1963. Relationship Between Short and
Long-Term Studies in Designing an Effective Toxicity Test Agric. Food
Cham 11 486-491,
White, Donald H 1991. Personal Communication U.S. Fish and Wildlife Service
School of Forest Resources. University of Georgia-Athens
Wobeser, G,, N D Nielsen and 3 Schiefer, 1976 Mercury and Mink I The Usa
of Mercury Contaminated Fish aa a Food For Ranch Mink Canadian Journal of
Comparative Medicine, 40 30-33.
Wob«s«r, G , N D Nielsen, and B, Schiefer 1976 Mercury and Mink II
Experimental Methyl Mercury Intoxication. Canadian Journal of Comparative
Medicine 40 34-45
v,\perra\wr9glit«.bbg
-------
HUMAN HEALTH MATERIALS
-------
Rerevised Draft
September 5, 1991
GREAT LAKES INITIATIVE
PROCEDURE FOR DERIVING HUMAN HEALTH CRITERIA
I. INTRODUCTION
A. Goal
The goal of the human health criteria for the Great Lakes and their
tributaries is the protection of humans from unacceptable exposure to
toxicants via consumption of contaminated fish, drinking water and water
oriented recreational activities.
B. Level of Protection
The criteria developed shall protect against unacceptable exposure to
chemicals which may produce carcinogenic and/or nonearcinogenic effects.
The adequacy of this protection is a function of the level of design
risk or no adverse effect estimation along with the other elements of
the procedure such as selection of data, exposure assumptions and
implementation, etc. and their interactions. Ambient criteria for
single carcinogens shall not exceed a lifetime incremental risk of 1 In
100,000 of developing cancer. For noncarcinogenic chemicals, criteria
-------
shall be established below levels expected to produce adverse health
effects, i.e. acute, subchronic and chronic toxicity including
reproductive and developmental effects.
C. Two-tiered Classification
Chemical exposure levels in surface water protective of human health
shall be derived based on either a Tier 1 or Tier 2 classification. The
distinctions between Tier 1 and Tier 2 are as follows:
1. Application
Tier 1. Numeric water quality criteria shall be derived via an
adopted narrative procedure. These numeric criteria shall be adopted
as enforceable ambient water quality standards throughout the Great
Lakes System (see Table 1).
Tier 2: The adopted narrative procedure will also be used to
calculate Tier 2 levels of protection for chemicals that do not meet
Tier 1 minimum data requirements. The Tier 2 levels are not adopted
as numeric criteria in order to allow for greater flexibility in
considering new toxicity data. Tier 2 levels developed via the
narrative procedure may be used as enforceable limits in discharge
permits and environmental cleanup actions. A listing of current Tier
2 levels developed via the procedure shall be available for review
through EPA Region V Water Program Office.
-------
2. Data Gaps:
Tier 1: In general, Tier 1 numeric criteria will be derived using a
comprehensive and well established chemical effects database. The
need for uncertainty adjustments are generally lower than for Tier 2.
The probability of significant future revisions to these numeric
criteria because of additional information is also lower than for
Tier 2.
Tier 2- Derivation of a level of protection under the Tier 2 process
allows the use of less data and therefore, carries less confidence
than a Tier 1 criterion. For noncarcinogens, the use of a database
with greater gaps of information generally requires higher
uncertainty adjustments. This is necessary to assure that a
sufficient margin of safety exists when establishing an
environmentally protective chemical exposure concentration. For
carcinogens, less confidence relates to a lower weight-of-evidence
of predicting actual human cancer risk either as a result of lesser
quality data or weaker evidence of carcinogenicity. The database
utilized by the Tier 2 procedure may generally be expected to change
or improve with further research. The goal is to encourage further
data development for Tier 2 chemicals in order to support the future
derivation of Tier 1 criteria, where possible.
II. MINIMUM DATA REQUIREMENTS
The best available toxicity data on the adverse health effects of a
chemical shall be used when developing criteria or levels of
3
-------
protection. Such information should be obtained from the USEPA
Integrated Risk Information System (IRIS) database, the scientific
literature, and other informational databases, studies and/or reports
containing adverse health effects data of adequate quality for use
in this procedure. To encourage consistency with other efforts
within the USEPA to control chemical contaminants, strong
consideration should be given Co the most currently available
guidance provided by IRIS for chemicals undergoing criterion or
level of protection development using this procedure.
A. Carcinogens
Adequate weight of evidence of potential human carcinogenic effects for
a chemical sufficient to calculate Tier 1 human cancer criteria or
Tier 2 levels of protection shall be defined according to the USEPA
classification system for chemical carcinogens (USEPA, 1986).
1. Tier 1
Weight of evidence of potential human carcinogenic effects
sufficient to derive a Tier 1 human cancer criterion shall
generally Include Group A - human carcinogens, and Group B -
probable human carcinogens. Certain Group C - possible human
carcinogens, may be suitable for Tier 1 criterion development
where studies have been well conducted yet are limited because
they involve only a single species, strain or experiment which
does not demonstrate a high incidence, unusual site or type of
tumor, or early onset.
-------
DRAFT
2. Tier 2
tfeight-of-evidence of possible human carcinogenic effects sufficient
to derive a Tier 2 human cancer level of protection shall include
those Group C - possible human carcinogens, with data sufficient for
quantitative risk assessment, however, inadequate for Tier 1
criterion development.
B. Noncarcinogens
All available toxicity data should be evaluated considering the full
range of possible health effects of a chemical, i.e. acute /subacute,
chronic/subchronic and reproductive /developmental effects, in order to
calculate human noncancer criteria and levels of protection. Although it
is desirable to have an extensive database which considers a wide range
of possible adverse effects, this type of data exists for a very limited
number of chemicals. For many others, there is a range in quality and
quantity of data available. To provide defensibility and maintain
objectivity in decision-making, it is necessary to establish a minimum
database below which development of criteria or levels of protection
cannot proceed. The following, although not ideal, represent the minimum
data sets necessary for this procedure.
1. Tier 1
The minimum data sufficient to derive a Tier 1 human criterion
shall include at least one well conducted epidemiologic study or
animal study. A well conducted epidemiologic study for a Tier 1
human noncancer criterion must quantify exposure levels and
demonstrate a positive association between exposure to a chemical
-------
and adverse effects in humans. A well conducted study in animals
must demonstrate a dose-response relationship involving critical
effects relevant to humans based on a defensible biological rationale.
Ideally, the duration of a study should span multiple generations of
exposed test species or at least a major portion of the lifespan of
one generation. This type of data is currently very limited. By
the use of uncertainty adjustments, shorter term studies with
evaluation of more limited effects may be used to extrapolate to
longer exposures or to account for a variety of adverse effects.
For this procedure such a limited study must be conducted for at
least 90 days in rodents or 10Z of the lifespan of other appropriate
test species and demonstrate a no-observable-adverse-effeet-level
(NOAEL). In some cases, chronic studies of one year or longer in
rodents or 50% of the lifespan or greater in other appropriate test
species that demonstrate a lowest-observable-adverse-effect-level
(LOAEL) involving relatively mild and reversible effects, may be
sufficient for use in Tier 1 criterion derivation.
2. Tier 2
When sufficient data are not available to meet the Tier 1 data
requirements, a more limited database may be considered for Tier 2
level of protection development. As with Tier 1, all available data
shall be considered and ideally should address a range of adverse
health effects with exposure over a substantial portion of the
lifespan of the test species. When such data are lacking it may be
necessary to rely on less than ideal data in order to establish a
level of protection for exposure.
-------
With the use of appropriate uncertainty factors to account for such
limited data, the minimum data sufficient to derive a Tier 2 level of
protection shall include a NOAEL from at least one well conducted
short-term repeated dose study. This study shall be of at least 28
days duration, in animals demonstrating a dose-response, and
involving critical effects relevant to humans based on a defensible
biological rationale. An additional uncertainty factor may be
applied to the standard uncertainty factor approach in order to
further accommodate the extrapolation of this very short study
duration to lifetime exposure. Structure-activity relationships
(SAR) may be used along with all other data available on a chemical
to determine the appropriate additional uncertainty factor to be
used with such limited data.
III. PRINCIPLES FOR CRITERIA DEVELOPMENT
The procedure to calculate Tier 1 criteria or Tier 2 levels of protection
is the same. However, the amount of data, the weight-of-evidence or
level of uncertainty In the data available for calculating criteria
varies significantly between the two Tiers.
A. Carcinogens
1. A non-threshold mechanism of carcinogenesis shall be assumed unless
biological data adequately demonstrate the existence of a threshold
on a chemical specific basis.
-------
2. All appropriate human epidemiologic data and animal cancer bioassay
data shall be considered. Data specific to an environmentally
appropriate route of exposure shall be used. Oral exposure should
be used preferentially over dermal and inhalation.
a. If acceptable human epidemiologic data are available for a
chemical, a risk associated dose shall be set equal to the
lifetime average exposure which would produce an Incremental
cancer risk of 1 in 100,000.
b. If acceptable human epidemiologic data are not available, the
risk associated dose shall be derived from available animal
bioassay data.
3. When animal bioassay data are used and a nonthreshold mechanism of
carcinogenicity is assumed, the data are fitted to a linearized
multistage computer model. The upperbound 95% confidence limit on
risk (or, the lower 95% confidence limit on dose) at the 1 in
100,000 risk level shall be used to calculate acceptable exposure.
Other models, modifications or variations of the multistage model
which consider the data more appropriately may be used on a
case-by-case basis.
4. If the duration of experiment is significantly less than the natural
lifespan of the test animal, the slope will be adjusted to
compensate for latent tumors which were not expressed (USEPA, 1980).
-------
5, A species scaling factor shall be used to account for differer'-es
between test species and humans. It shall be assumed that
milligrams per surface area per day is an equivalent dose between
species (USEPA, 1986). However, if adequate pharmacokinetic and
metabolism studies are available, these data may be factored into
the adjustment for species differences on a case-by-case basis.
6. Additional data selection and adjustment decisions must also be made
in the process of quantifying risk. Consideration must be given to
tumor selection for modeling, e.g. pooling estimates for multiple
tumor types and identifying and combining benign and malignant
tumors. All doses shall be adjusted to give an average daily dose
over the study duration. Adjustments in the rate of tumor response
must be made for early mortality in test species. The
goodness-of-fit of the model to the data must also be assessed.
7. When a linear, nonthreshold dose-response relationship is assumed,
the risk associated dose shall be calculated using the following
equation:
RAD - 0.00001
V
Where:
RAD - risk associated dose in milligrams of toxicant per kilogram
body weight per day (mg/kg/day).
-------
0.00001 (1 x 10" ) - incremental risk of developing cancer equal to
1 in 100,000.
q * = slope factor (mg/kg/day)
8. Whenever appropriate human epidemiologic data are not available, and
the biological data indicate that a chemical causes cancer via
a threshold mechanism, the risk associated dose may be calculated
via a method other than the linearized multistage model on a
case-by-case basis.
B. Noncarcinogens
1. Noncarcinogens shall generally be assumed to have a threshold dose
or concentration below which no adverse effects should be observed.
Therefore, the Tier 1 human safe criterion or Tier 2 level of
protection is i he maximum water concentration of a substance at or
below which a Lifetime exposure from drinking water, consumption of
contaminated fish, and water related recreation activities would be
expected to have no adverse effects.
Exceptions to this case may exist. Chemicals acting as genotoxic
teratogens and germline mutagens are thought to produce reproductive
and/or developmental effects via a genetically linked mechanism which
may have no threshold. Other chemicals also may not demonstrate a
threshold. Therefore, since there is no well established methodology
for calculating criteria protective of human health from the effects
of such agents, criteria will be established on a case-by-case basis.
10
-------
2. All appropriate human and animal toxicologic data shall be reviewed
and evaluated. Exposure should be via a route most relevant to
environmental exposure. When human data are not available, animal
data from species most relevant to humans shall be used. In the
absence of data to distinguish the most relevant species, data from
the most sensitive animal species tested, i.e. the species showing a
toxic effect at the lowest administered dose (given a relevant route
of exposure) shall generally be used.
3. The experimental exposure level representing the highest level
tested at which no adverse effects were demonstrated (NOAEL) shall be
used for criteria calculations. In the absence of a NOAEL, the
lowest observable adverse effect level (LOAEL) may be used in
certain cases.
4. Uncertainty factors shall be used to account for the uncertainties in
predicting acceptable exposure levels for the general human
population based upon experimental animal data or limited human data.
a. An uncertainty factor of 10 shall generally be used when
extrapolating from valid experimental results from studies on
prolonged exposure to average healthy humans. This 10-fold
factor Is used to protect sensitive members of the human
population.
b. An uncertainty factor of 100 shall generally be used when
extrapolating from valid results of long-term studies on
experimental animals when results of studies of human exposure
11
-------
are not available or are Inadequate. In addition to (a) above,
this represents an additional 10-fold uncertainty factor in
extrapolating data from the average animal to the average human.
c. An uncertaiity factor of up to 1000 is generally used when
extrapolating from animal studies for which the exposure duration
is less than chronic or when other significant deficiencies in
study quality are present, and when useful long-term human data
are not available. In addition to (a and b) above, this
represents an additional uncertainty factor of up to 10-fold.
The level of additional uncertainty applied for subchronic
exposure depends on the duration of the study used relative to
the lifetime of the experimental animal.
d. An additional uncertainty factor of between one and ten may be
used when deriving a criterion from a lowest observable adverse
effect levc'l (LOAEL) . This uncertainty factor accounts for the
lack of an identifiable no observable adverse effect level
(NOAEL) . "lie level of additional uncertainty applied depends
upon the severity of the observed adverse effect.
e. An additional factor of between one and ten may be
applied when there are limited effects data or incomplete
subacute or chronic toxicity data. The level of quality and
quantity of the experimental data available as well as
structure-activity relationships may be used to determine the
factor selected.
12
-------
5. All study results shall be converted, as necessary, to the standard
unit of milligrams of toxicant per kilogram of body weight per day
(mg/kg/day). Doses will be adjusted for continuous exposure, i.e.,
7 days/week, 24 hrs/day, etc.
C. Criteria Derivation
I. Standard Exposure Assumptions
The following represent the standard exposure assumptions used to
calculate criteria for carcinogens and nonearcinogens. There may
be situations where site specific data on fish consumption may
indicate higher levels of exposure are more appropriate in order to
address particular subpopulations substantially different from the
average exposure.
Wh - weight of an average human (Wh * 70kg).
WC » average per capita water consumption (both drinking and
incidental exposure) for surface waters classified as public
water supplies (WCd - 2 liters/day)
-or-
average per capita incidental dally water exposure for
surface waters used only for recreational activities
(WC - 0.01 liters/day)
3/7
13
-------
FC « average per capita daily consumption of sportcaught fish
0.015 kg/d,iy
BAF * bioaccumuLation factor.
2. Carcinogens
The Tier 1 human cancer criteria or Tier 2 level of protection shall
be calculated as follows*
HCV - RAD x Wn
WC + (FC x BAF)
Where-
HCV - Human Cancer Value In milligrams per liter (mg/L).
RAD - Risk associated dose in milligrams toxicant per kilogram body
weight per day (mg/kg/day) that Is associated with a lifetime
incremental cancer risk equal to 1 In 100,000.
3. Noncarcinogens
The Tier 1 human noncancer criteria or Tier 2 level of protection
shall be calculated as follows:
<32o
-------
HNV - APE x Wh x RSC
WC + (FC x BAF)
Where
HNV » Human noncancer value in milligrams per liter (mg/L) .
ABE « Acceptable daily exposure in milligrams toxicant per kilogram
body weight per day (mg/kg/day) .
RSC - Relative source contribution factor of 0.8 for persistent
bioaccumulative compounds. This is generally used for
bioaccumulative organic compounds to allow for potential
exposure via sources other than consumption of contaminated
water and fish and recreational exposure.
15
-------
Revised 9/6/91
Table 1. Great Lakes Initiative Human Health Criteria (Tier 1) and Levels of Protection (Tier 2)
Substance BAF*
CAS No. Common Name
50-29-3 DDT 2,296,650
57-74-9 Chlordane 263,250
58-89-9 Lindane 1,954
60-57-1 Dieldrin 222,924
76-44-8 Heptachlor 28,798
87-86-5 Pentachlorophenol 755
H8-74-1 Hexachlorobenzene 239,154
1336-36-3 PCBs (class) 2,132,232
1746-01-6 2,3,7,8-TCDD 43,714
7439-97-6 Mercury 470,000**
(includes
Methylmercury, 22967-92-6)
HNV (ng/1)
drinking
0.8 (Tier
8 (Tier
600 (Tier
0.8 (Tier
200 (Tier
2E5 (Tier
10 (Tier
0.01 (Tier
1E-4 (Tier
0.5 (Tier
nondrinking
i)
1)
1)
1)
1)
1)
1)
2)
1)
1)
0.8 (Tier
8 (Tier
600 (Tier
0.8 (Tier
200 (Tier
2E5 (Tier
10 (Tier
0.01 (Tier
1E-4 (Tier
0.5 (Tier
i \
i )
1)
1)
1)
1)
1)
1)
2)
1)
1)
HCV (ng/1)
drinking nondrinking
0.0& (Tier 1) 0.06 (Tier 1)
0.1 (Tier 1) O.I (Tier 1)
20 (Tier 2) 20 (Tier 2)
0.01 (Tier 1) 0.01 (Tier 1)
0.4 (Tier 1) 0.4 (Tier 1)
400 (Tier 1) 500 (Tier 1)
0 1 (Tier 1) 0.1 (Tier 1)
3E-3 (Tier 1) 3E-3 (Tier 1)
1E-5 (Tier 1) 1E-5 (Tier 1)
ft
*~ _^* _«_ _* * v. it. • * «•* w »«% •"!..«- l_ d._bJI \M .1 ___.__.__.*. ^ f"l/"» A _ _ _.__.»._ J
**Interlm value estimated by Michigan DNR.
-------
BIBLIOGRAPHY
U.S. Environmental Protection Agency (EPA), Water Quality Criteria
Availability, Appendix C Guideline and Methodology Used in the
Preparation of Health Effects Assessment Chapters of the Consent
Decree Water Quality Criteria Documents, Federal Register, Vol. 45,
November 28, 1980, 79347-79357.
U S. Environmental Protection Agency (EPA), 1986, Guidelines for
Carcinogen Risk Assessment. Federal Register Vol. 51, No. 185.
September 24, 1986, 33992-34002.
17
-------
DRAFT
Technical Support Document
Human Health Criteria
Great Lakes Initiative
September 5, 1991
-------
Human Health Criteria
Technical Support Document
I. INTRODUCTION
Goal
Level of Protection
Two Tiered Aoproach
Technical Background
II. MINIMUM DATA REQUIREMENTS
Carcinogens
Weight of Evidence
Humans
Animals
Appropriate Studv Design and Data Development
JJoncarcinogens
Appropriate Studv Design and Data Development
Acute Toxicity
14 Day and 28 Day Repeated Dose Toxicity
Subchronic and Chronic Toxicity
Reproductive and Developmental Toxicitv
Tier Designation
Carcinogens
Noncarcinogens
III PRINCIPLES FOR CRITERIA DEVELOPMENT
General
Carcinogens
Mechanism
Data Review
Model
Nonthreshold Approach
Threshold Approach
Lifespan Adjustment
Species Scaling
Noncarcinogens
Mechanism
Data Review
Uncertainty Factors
Exposure Assumptions
Body Weight
Duration of Exposure
Incidental Exposure
Drinking Water
Fish Consumption
Relative Source Contribution
IV. CRITERIA CALCULATIONS
Standard Exposure Assumptions
Carcinogens
Noncarcinogens
References
Tables, Appendices
^bles
Appendix A
-------
HUMAN HEALTH CRITERIA
Technical Support Document
I. INTRODUCTION
A. Goal
The goal of the human health criteria for the Great Lakes and their
tributaries is the protection of humans from unacceptable exposure to
toxicants from consumption of contaminted fish, drinking water and water
oriented recreational activities. Emphasis is on protection of the
individual in evaluating toxicity information and its application in the
derivation of criteria and levels of protection. Exposure assumptions
follow trends for the general population as a region. Based on extreme
differences in behavior, there may be some individuals who receive a
greater level of protection or a lesser level of protection via these
procedures.
B. Level of Protection
Criteria or levels of protection developed for human health generally
restrict chemical carcinogen exposure to levels estimated to result in a
lifetime incremental risk of no greater than 1 in 100,000 of developing
cancer. The procedure generally used to estimate the risk level leads to
the development of a plausible upper limit of the risk. When the design
risk is combined with the other aspects of the initiative procedures,
i.e., selection of data, exposure assumptions, implementation, etc., the
1
-------
true cancer risk, particularly from actual exposure to the resulting low
level environmental concentrations, is unknown and may even be as low as
zero.
The selection of an "acceptable" target risk level is both difficult and
controversial. Differences in perception of risk, opinions as to benefit
versus risk reduction costs, as well as distinctions between risks that
are considered voluntary or involuntary, all play a meaningful role in
the debate of acceptability For this initiative, the approach to
defining acceptable is to consider levels of existing, essentially
involuntary and nonbeneficial environmental risks routinely experienced
by Great Lakes Basin residents. Such risks may be considered as
background risks associated with living within this region Table i
evaluates the everyday risks of death from several of these naturally
occurring incidents such as tornados, floods, lightning and animal bites
or stings. When extrapolated to lifetime risks, we see that these risks
range from 1.4 in 100,000 for animal bites or stings to 4 in 100,000 for
floods and tornadoes. To establish acceptable upper limit lifetime risks
at 1 in 100,000 provides a level of protection within the range estimated
for these naturally occurring incidents.
For noncarcinogens, protection of human health is generally centered on
providing a margin of safety against chronic adverse effects from long
term, low level contaminant exposure. However, protection also includes
consideration of potential reproductive and developmental impairment as
well as other acute or subchronic adverse effects. The concept of
acceptable exposure incorporates the potential for long term exposure of
a broad spectrum of the population to an environmental contaminant
~7
without any anticipated adverse health effects.
-------
TABLE 1
EVERYDAY RISKS
*'• u AccuauUta i On*
ffl 100.000 Msk of Ooatft
9«r CiPtU
ExtrioolitM U'
Ui
ftjttp teiilelo accident
falls)
9nwiin
Ifl-4
lO'5
«••
10'S
10*S
u-4
Ifl-7
Ifl-7
lO'7
io-7
io-5
io-»
10-4
io"4
u)-4
io-4
io-3
10 ~*
Ifl-4
Ifl-4
Ifl-4
i 4 i io'2
4 Z x IO*3
2 a i 10°
Z i 10°
7. IO-1
3 S • lO-4
4 i 10"S
4 t 10"S
3.J x 10'S
1 4 i lfl'S
S.I i Ifl'3
3 S i 10°
7 x 10°
3 • 10**
4 i 10'*
4 l 10'2
7 x 10*2
3 i Ifl*2
1.4 • W2
1.4 . lO*2
5 4 i ;o"2
Soit One o«u
ft aowMt of OMWt otittar <«fi«taxin)
IN ointt of «11V (tflattim)
200 9aMom of drinking ««ur frv
90 pewit of brail* ttttk (eanetr
OP No» Or!««
ealyf
2 eisiPtttei
Crqucft 4M Mllson (1912)
70
-------
C. Two Tiered Approach
A two tiered approach is used to derive exposure levels protective of
human health. Both tiers relv generally on the same standard procedure
for data review and criteria derivation The difference between the two
focuses heavily on the certainty with which one can predict a level of
risk or a level of safety for humans from the data available. The more
adequate the database to estimate actual human risk or to establish no
adverse effect levels, the greater the certainty in the appropriateness
of the criterion or level of protection. This level of certainty depends
heavily on the weight of experimental evidence which includes factors
such as* the quantity of studies or size of the experimental database
available for review; the quality of study design, its conduct and range
of effects evaluated; the potency or range and type of adverse effects
observed and, the appropriateness of this data in predicting human
effects, i.e. evaluation of effects in humans or in animal species
biologically similar to human.
The greater the level of certainty in the database for noncancer effects,
generally the lower the need for adjustment of the research findings to
assure a necessary margin of safety. The greater the weight of evidence
for carcinogenicity, the greater the strength in predicting cancer risk
to humans. Chemicals with databases providing a high level of certainty
in predicting a level of risk or safety for humans from adverse health
effects are suitable for Tier 1 numeric criteria derivation. Tier 1
criteria are conceptionally those criteria where the probability of
change is low. Tier 1 criteria will be applicable across the Great Lakes
Basin. Great Lakes states will be expected to adopt such criteria by
4
-------
regulation for use in their water pollution control programs. The
narrative procedure to develop Tier 1 criteria will also be adopted by
regulation, thereby allowing for development of Tier 1 criteria and their
implementation for chemicals with data sufficient to meet the Tier 1 data
requirements vet not formallv adopted yet as numeric criteria.
Chemicals with less extensive data or where the weight of evidence toward
predicting human health effects is less certain, are subject to Tier 2
levels of protection development. Under Tier 2, the probability of
future change is greater than for Tier 1 as demonstrated by the extent,
level of quality and/or weight of evidence or conclusiveness of effects
demonstrated by the database. The levels of protection derived via Tier 2
need to remain subject to change pending new data and/or reinterpretatlon
of effect or potencv. It is critical that state chemical regulatory
programs maintain the opportunity to easily react to these changes and
therefore depend on the narrative procedure for recalculation. The
process is the emphasis here versus "fixed" numeric criteria. States
will be expected to adopt, bv regulation, the procedure for developing
Tier 2 levels of protection, rather than the numeric value the procedure
generates.
D. Technical Background
The process used to evaluate effects and in development of criteria shall
be based on currently acceptable scientific methods and consider guidance
offered by the various USEFA methods. Particular attention must be paid
to RfD and cancer risk estimation development contained within the
Integrated Risk Information System (IRIS),
-------
To achieve consistency with other USEPA guidance for chemical management,
it is important to review and strongly consider the IRIS values for
chemicals undergoing criteria or level of protection development whenever
available. Although consistency is important, it is also important that
the most current and complete data should be used when generating
criteria or levels of concern whether IRIS has considered this data or
not. Further, since IRIS values are developed under guidance and through
the judgement of workgroups, the final values may not always be arrived
at corsistently, i.e., duration of studies, selected, uncertainty
factors, applied basis for derivation of potency slopes, etc., may differ
between decisions. This may present difficulties, from a procedural
perspective, where consistency with a narrative procedure is important.
Therefore, the most current and complete data will be used for criteria
and level of protection development. If the data have not been
considered by the IRIS RfD or CRAVE workgroups, the appropriate workgroup
should be advised of the data. In cases where IRIS RfDs or potency
slopes have not been developed consistent with these procedures, the
rationale for RfD or potency slope development should be evaluated and
determination made whether 1) Justification is sufficient to support
deviating from these procedures, or 2) justification exists to deviate
from IRIS guidance.
Specific references which should be reviewed and evaluated for assistance
in these criteria and subsequent policy development are as follows:
National Cancer Institute (NCI). 1976. Guidelines for Carcinogen
Bioassay in Small Rodents, Technical Report Series No. 1, U.S.
Department of Health, Education and Welfare, NCI-CG-TR-1. -
33 f
o
-------
Office of Science and Technology Policy (OSTP). 1985. Chemical
Carcinogens; A Review of the Science and Its Associated Principles,
Federal Register, Vol. 50, No. 50 March 14, 1985, 10371-10442.
Organization for Economic Cooperation and Development (OECD). 1987.
Guidelines for Testlng of Chemicals, Paris, France.
U.S. Environmental 'rotection Agency (EPA). 1989. Risk Assessment
Guidance for Superfund, Volume 1, Human Health Evaluation Manual
(Part A) - Interim Final, Office of Emergency and Remedial
Response, Washington, D.C., EPA/540/1-89/002.
U.S. Environmental Protection Agency (EPA). 1980. Water Quality Criteria
Availability, Appendix C Guidelines and Methodology Used in the
Preparation of Health Effects Assessment Chapters of the Consent
Decree Water Quality Criteria Documents, Federal Register, Vol. 45,
November 28, 1980, 79347-79357.
U.S. Environmental Protection Agency (EPA). 1985. Toxic Substances
Control Act Test Guidelines; Final Rules, Federal Register,
Vol. 50, NO. 188. September 27, 1985, 39421-39425.
U.S. Environmental Protection Agency (EPA). 1986. Guidelines for
Carcinogen Rink Assessment. Federal Register, Vol. 51, No. 185.
September 24, 1986, 33992-34002.
U.S. Environmental Protection Agency (EPA). 1986. Guidelines for the
Health Assessment of Suspect Developmental Toxicants, Federal
Register, No. 51, No. 185. September 24, 1986 34028-34040.
-------
This is by no means a complete list. Other multiple sources of
information and guidance may also be considered as appropriate.
II. MINIMUM DATA REQUIREMENTS
A. Carcinogens
1. Weight of Evidence
Evidence of a chemical's possible carcinogenic effects in humans
shall be categorized according to the EPA weight of evidence
classification system, which is adapted from the International Agency
for Research on Cancer (IARC). The five categories or groups are as
follows*
Group A - Human Carcinogen
"sufficient" evidence from epidemiologic studies to support a
causal association between exposure to the chemical and cancer;
Group B - Probable Human Carcinogen
"limited" evidence from epidemiologic studies with or without
supporting animal data (Group Bl); or. "sufficient" evidence of
carcinogenlcity based on animal studies, but for which there may
be "Inadequate evidence" or "no data" from epidemiologic studies
(Group B2);
333
-------
Group C - Possible Human Carcinogen
"limited" evidence of carcinogenicitv in animals and the absence of
data for hum.ins;
Group D - Not Classifiable as to Human Carcinogenicity
"inadequate" evidence of carcinogenicity in humans and animals, or,
for which no data are available; and
Group E - Evidence of Noncarcinogenicity for Humans
"no evidence" for carcinogenicity in at least two adequate animal
tests in different species or in both adequate epidemiologic and
animal studies.
The definitions of the EPA weight of evidence classifications are as
follows:
1. Humans
a. Sufficient evidence - a causal association can be inferred
between exposure to the chemical and human cancer.
b. Limited evidence - a causal Interpretation is credible, but
that alternative explanations, such as chance, bias or
confounding could not adequately be excluded.
c. Inadequate evidence - there were few pertinent data, or, a
causal interpretation is not credible from available studies
since they did not exclude change, bias or confounding.
d. No evidence - no association was found between exposure and
an increased risk of cancer In well-designed and well-
conduci ed Independent analytical epidemiologic studies.
9
-------
2. Animals
a. Sufficient evidence - an increased incidence of malignant or
combined malignant and benign tumors: 1) in multiple species
or strains, 2) in multiple experiments using different dosage
levels and possible different routes of exposure; or 3) in a
single experiment with a high incidence, unusual site or type
of tumor, or early onset.
b. Limited evidence - data suggest a carcinogenic effect but are
limited because' 1) the studies involve a single species,
strain or experiment which does not demonstrate a high
incidence, unusual site or type of tumor, or earlv onset; 2)
the experiments used inadequate dosage levels, inadequate
exposure duration, inadequate follow-up periods, poor
survival, too few animals, or inadequate reporting; 3) an
increase in benign tumor incidence only and no response in a
variety of short-term tests for mutagenicity; or 4) tumor
responses of marginal statistical significance due to
inadequate study design or reporting, or, in tissue known to
have a high or variable background rate.
c. Inadequate evidence - because of major qualitative or
quantitative limitations, the studies cannot be interpreted
as shoving either the presence or absence of a carcinogenic
effect.
d. No evidence - no increased tumor incidence in at least two
well-designed and well conducted animal studies in different
species.
10
-------
Further detail regarding this classification system for
categorizing weight of evidence for carcinogenicity may be found
in the EPA, Guidelines for Carcinogen Risk Assessment (E?A, 1986).
2. Appropriate Study Design and Data Development
The following discussion summarizes the process for evaluating
evidence of carcinogenicitv and outlines an approach study design by
which one may measure the quality and adequacy of data development.
When available, human epidemiologic data with quantifiable exposure
levels are preferred for evaluating a chemical's carcinogenic
potential ovei use of animal data alone. Epidemiological studies can
provide direct evidence of a chemical's carcinogenicity in humans
(05TF, 1985). The type of epidemiologic study conducted indicates
whether the study may be useful in assessing carcinogenic risk to
exposed humane; (analytical studies) or if it is merely hypothesis-
generating and inherently incapable of proving a causal association.
Case reports, descriptive studies and ecological (correlational)
studies generally cannot establish whether risks are associated with
particular exposures. Analytical studies can assess carcinogenic
risks to exposed humans, and can infer a casual association (Mausner
and Kramer, 1985; OSTP, 1985). The two general types of analytical
studies are case-control and cohort. In case-control studies, a
group of diseased "case" individuals is initially identified and
matched with nondiseased "controls". Information on past exposure to
reputed risk factors or causative agents is then collected for both
groups. If the proportion of cases with a certain exposure is
significantly different than that of controls, an association between
11
-------
exposure and disease mav be indicated. A cohort study starts by
identifying a group of individuals with a particular exposure and a
similar group of unexposed persons and follows both groups over time
tc determine subsequent health outcomes. The rates of disease in the
exposed and unexposed groups are then compared. Cohort studies may
be based on current exposure and future health outcomes (prospective
cohort study), or on past exposure information and disease occurrence
(historical cohort study). As with case-control studies, cohort
studies that are well-designed, well-conducted, and well-evaluated
can test hypotheses and provide the basis for causal inferences
(OSTP, 1985; EPA, 1986). Factors such as proper selection and
characterization of exposed and control groups, adequacy of duration
and quality of follow-up, proper identification and characterization
of confounding factors and bias, appropriate consideration of latency
effects, valid ascertainment of causes of morbidity and mortality,
and the ability to detect specific effects are all elements for
determining the adequacy of epidemiologlc studies (EPA, 1986).
In interpreting a reported causal association, reference may be made
to the following criteria, as described by IARC (1985), EPA (1986),
and the Tripartite Working Group (1985):
- There Is no identifiable positive bias which could explain the
association.
- The possibility of positive confounding factors has been considered
and ruled out as explaining the association.
- The association is unlikely to be due to chance alone.
337
12
-------
Although the weight of evidence increases with the number of adequate
studies, in some instances, a single epidemiologic study mav be
strongly indicative of a cause-effect relationship (IARC, 1985; EPA,
1986). Confidence to infer a causal association is increased bv
any of the following- when several independent studies are concordant
in showing the association; when the association is strong; when
there is a dose-response relationship when a reduction in exposure is
followed by a reduction in the incidence of cancer; when the effect
is biologically plausible; or when the effect is specific for a
particular chemical. When epidemiological evidence based on
analvtical studies appears to be significantly flawed, the evidence
mav then be downgraded to being suggestive of an association based on
scientific judgment. This may still provide evidence that a causal
interpretation is credible, but that alternative explanations, such
as chance, bias or confounding factors, could not adequately be
excluded.
Epidemiological studies are inherently capable of detecting only
comparatively large increases in the relative risk of cancer (EPA,
1986). Other limitations of epidemiological studies include the long
latency of cancer, and the difficult task of exposure assessment,
including multiple exposures. Therefore, negative results from such
studies do not verify that a particular agent is noncarcinogenic in
humans {IARC, 1985; EPA, 1986; OSTP, 1985).
Although epidemiologic studies are preferable for assessing
carcinogenic potential for humans, the relative paucity of such data
necessitates the use of animal data as a surrogate for humans in
13
43 p
-------
most situations. In the aDsence of adequate data on humans, it is
biologically plausible and prudent to regard agents for which there
is sufficient evidence of carcinogenicity in experimental animals as
if thev present a carcinogenic risk to humans (IARC, 1987). The
weight of evidence that an agent is potentially carcinogenic in
humans increases with: a) the increase in tissue sites affected; b)
the increase in number of animal species, strains, sexes, doses and
experiments showing a carcinogenic response; c) the occurrence of
clear-cut dose-response relationships as well as a high level of
statistical significance of the increased tumor incidence in treated
groups as compared to controls; d) a dose related shortening of the
time-to-tumor occurrence or time to death with tumor; and e) a
dose-related increase in the proportion of tumors that are malignant
(EPA, 1986).
The guidelines detailed by EPA (1985), OSTP (1985) and NCI (1976) for
evaluating long-term carcinogenicity bioassays will be utilized to
determine the adequacy of design and the strength of evidence
provided by the study. Specific study design elements of these
guidelines are synopsized as follows:
Species used: The most videlv used and accepted test species is
the rat. NCI/NTT bioassays routinely use the Fischer inbred
(F344) strain of rat and the B6C3F1 hybrid mouse. Hamsters have
also been frequently used. Other animal species and strains may
also be appropriate surrogates to demonstrate a chemical's
carcinogenic potential.
H
-33 °r
-------
Number of animals: At least 100 rodents (50 of each sex) should
be used at each dose level and concurrent control.
Age at scare: Dosing of roaents should begin as soon as possible
after weaning to allow for the long latency of cancer. For rats,
dosing ideally begins before the age of 6 weeks and should not
begin after 8 weeks of age.
Survival: All groups should have at least 50Z survival at the
time of termination.
Concurrent control groups: These should be untreated, sham
treated, or, if a vehicle is used in administering the test
substance, vehicle control groups. The use of historical control
data is desirable for assessing the significance of changes
observed in exposed animals, but only if the strain of animals and
laboratory conditions have not changed. For the evaluation of
rare tumors, even small tumor responses may be significant
compared to historical data. The review of tumor data at sites
with high spontaneous background requires special consideration
(OSTP, 1985). For instance, a response that is significant with
respect to the experimental control group may become questionable
if the historical control data indicate that the experimental
control group had an unusually low background incidence (KIP, 1984)
Dose levels.: At least 3 dose levels are recommended in addition
to the concurrent control group, for the purpose of risk
assessment (OSTP, 1985; EPA, 1985). For the purpose of hazard
15
-------
assessment, detection of a carcinogenic response is possible with
one dose level, although 2 dose levels are preferred and are
necessary to demonstrate a dose-response relationship. The
highest dose level should target the maximum tolerated dose
(MTD). The MTD is the dose which, when given for the duration of
the chronic study, elicits signs of minimal toxicity (e.g., less
than or equal to 10Z weight gain decrement) without substantially
altering the normal life span due to effects other than
carcinogenicitv The MTD is intended to provide an adequate
statistical power for the detection of carcinogenic activity.
While not an ideal solution to the problem of low bioassay
sensitivity, use of the MTD is appropriate if it is properly
determined (OSTP, 1985; EPA, 1986).
Dosing route. The test substance should be administered via the
oral, dermal or inhalation route.
Dosing schedule: The animals should ideally be dosed on a 7 day
per week basis. Hovever, based primarily on practical
considerations, dosing on a 5 day per week basis is acceptable.
Treatment preferably should be continued for the major portion of
the animal's lifespan. This Is at least 18 months for mice and
hamsters, and 24 months for rats.
Data collection: During the study, animals should be monitored
for body weight and food intake, as well as for the onset and
progression of all toxic effects. Clinical examinations,
including hematology, biochemistry of blood, urinalysis, and
16
-------
ophthalmological examination, should be made. Gross necropsy and
hlstopathology should be performed on all animals. Specific
requirements are too numerous to list here, but mav be reviewed
via the EPA (1985) and NCI (1976) guidelines.
All observed results should be evaluated by an appropriate and
generally accepted statistical method. Evidence for carcinogenic
action should be based on the observation of statistically
significant tumor responses in specific organs or tissues.
Appropriate statistical analysis should be performed on data from
long-term studies to help determine whether the effects are
treatment-related or possiblv due to chance. These should at
least include a statistical test for trend, including appropriate
correction for differences in survival. The weight to be given
to the level of statistical significance (the p-value) and to
other available pieces of information is a matter of overall
scientific judgment. In a review of 25 NTP feeding studies as
discussed by OSTF (1985), a simple statistical rule was derived
by Haseman which appeared to mimic the scientific judgment
process used in those experiments. "Regard as carcinogenic any
chemical that produces a high dose increase in a common tumor
that is statistically significant at the 0.01 level or a
high-dose Increase in an uncommon tumor that is statistically
significant at the 0.05 level. The overall false positive rate
associated with this rule was estimated to be no more than 7-8Z
for the NT13 two-sex, two-species protocol". A statistically
significant excess of tumors of all types in the aggregate, in
the absence of a statistically significant increase of any
1 7
-------
individual tumor type, should be regarded as minimal evidence of
carcinogenic action unless there are persuasive reasons to the
contrary (OSTP, 1985).
These guidelines represent ideal parameters. Studies will not be
expected to meet all of these desirable conditions in order to be
further considered for use in the process. The adequacy and
appropriateness of all animal carcinogenicity bioassays will be
carefully considered. It is crucial that judgment of adequate
testing be based on sound scientific principles. In general, it can
be expected that most substances tested for carcinogenicity have been
reviewed by NCI/NTP, IARC, and/or EPA. Historically the evaluations
by these agencies have been sufficient for decision-making. A
thorough assessment of the data should be performed regardless of the
findings of those independent agencies since these reviews might be
dated in that research data available subsequent to the date of
review were not considered by the reviewing group. The overall
assessment of a chemical's carcinogenic potential will depend on
welght-of-evidence based upon full consideration of all the evidence.
A variety of studies may be encountered which may be considered
flawed or lacking in adequate design or reporting. Such studies may
only be able to be utilized anecdotally and only considered
suggestive evidence of carcinogenicity. Examples of conditions
meeting such a criteria are:
1. Borderline conditions of:
a. Statistical significance. A general example would be a study
in which the MTD was administered and the test for positive
-------
dose-related trend (e.g., Cochran-Armitage Test) determined
that the slope of the dose-response curve was different from
zero; however, comparisons of the tumor incidences in treated
groups vith that in the control group (e.g., Fisher-Irwin
exact test) were not significant at p - 0.05.
b. Study dosign.
c. Study reporting. A general example would be a study reporting
a tumorLgenic response, but lacking statistical analyses to
verify that an apparent increase in incidence was statistically
significant.
d. A tumor response in a tissue known to have a high and variable
background rate.
2. Tumor responses or lack of response which are more than likely
attributable to excessive doses that compromise major organ
systems. Positive studies at levels above the MTD should be
carefully reviewed to ensure that the responses are not due to
factors which do not operate at exposure levels at or below the MTD.
Evidence indicating that high exposures alter tumor responses by
indirect mechanisms that may be unrelated to effects at lower
exposures should be dealt with on an individual basis. As noted
by the OSTF (1985), "Normal metabolic activation of carcinogens
may possibly also be altered and carcinogenic potential reduced
as a consequence [of high-dose testing].'1 Negative long-term
animal studies at exposure levels above the MTD may not be
acceptable if animal survival is so impaired that the sensitivity
of the study is significantly reduced below that of a conventional
chronic animal study at the MTD.
19
-------
"The carcinogenic effects of an agent may be influenced by non-
physiological responses (such as extensive organ damage, radical
disruption of hormonal function, saturation of metabolic
pathways, formation of stones in the urinary tract, saturation of
DNA repair with a functional loss of the system) induced in the
model systems. Testing regimes inducing these responses should
be- evaluated for their relevance to the human response to an
agent and evidence from such a study, whether positive or
negative, must be carefully reviewed." (OSTP, 1985).
3. Tumors at the site of oral, dermal or inhalation administration
attributable to irritation or frank tissue damage.
4. Tumor responses following administration by a route other than
oral, dermal or inhalation. Such tumors may be at the site of
administration or removed from it. Some general examples are
tumors induced following intraperitoneal, intravenous or
subcutaneous Injection, or bladder implantation.
Solid-state carcinogenisis is the occurrence of tumors around an
inserted inert object. It is a phenomenon that is dependent
primarily on the size and shape of the object, rather than the
chemical composition of the implanted material (Williams and
Weisburger, 1986). Therefore, induction of solid-state tumors
generally will not be considered in the weight-of -evidence
approach .
20
-------
Data from all long-term animal studies snould be considered in
evaluating carcinogenicitv. However, carcinogenic responses
should be evaluated as to their relevance of predicting cancer
risks to humans. Therefore, data from species that respond most
like humans should be used preferentially when such information
exists. Data on tumors in organs or as a result of effects on
metabolic or biochemical pathways that don't exist in humans
should be evaluated very carefully as to their inference of human
cancer risk. Further, a positive carcinogenic response in one
species/strain or sex is not generally negated by negative
results in other species. Replicate negative studies, however,
that are essentially identical in all other respects to a
positive study may cast doubt on the validity or reproducibility
/
of a positive study. A variety of other weight of evidence
issues may make it difficult to interpret the significance of
tumor data and therefore result in a lower classification of
carcinogenLcity. Examples of such issues include: increased
incidence of tumors in the highest dose group only and/or only at
the end of the study; no substantial dose-related increase in the
proportion of tumors that are malignant; the occurrence of tumors
that are predominantly benign; no close-related shortening of the
time to th« appearance of tumors; negative or inconclusive
results from a spectrum of short-term tests for mutagenic
activity; or, the occurrence of excess tumors only In a single
sex (EPA, 1985).
21
-------
B. Noncarcinogens
The full range of possible adverse health effects shall be evaluated when
establishing an acceptable exposure to noncarcinogens. Acute/subacute,
subchronic/chronic and reproductive/developmental effects shall be
considered. The principles of data selection are similar to those for
carcinogenic effects, a well-conducted epidemiologic study which
demonstrates a positive association between a quantifiable exposure to a
chemical and human disease is generally preferred for evaluating adverse
health effects. At present, however, human data adequate to serve as a
basis for quantitative risk assessment are available for only a few
chemicals. Frequently, inference of adverse health effects to humans
must be drawn from toxicity information gained through animal experiments
with human data serving qualitatively as supporting evidence. Under this
condition, health effects data must be available from well conducted
studies in animals relevant to humans based on a defensible biological
rationale, i.e. similar metabolic pathways, etc.
The following provides guidance on appropriate study design for a
variety of types of toxicity studies against which one may evaluate the
quality and adequacy of data development. This evaluation of adequacy
of data coupled with effects information forms the basis for selection
of uncertainty factors and subsequent acceptable exposure levels.
1 . Appropriate Study Design
a. Acute Toxicity
Determination of an LDSO or LC50 Is often an initial step in
experimental assessment and evaluation of a chemical's toxic
-------
characteristics. Such studies are used in establishing a dosage
regimen in subchronic and other studies and tnav provide initial
information on the mode of toxic action of a substance. Because
LD50 or LC59 studies are of short duration, inexpensive and easy
to conduct, thev are commonly used in hazard classification
systems. Acute lethality studies are of limited use in this
process. However, the data from such studies do provide
information on health hazards likely to arise from individual
short-term exposures. Although this process should never allow
exposures which approach such acute levels, such studies provide
high dose effects data from which to evaluate potential effects
from exposures which may temporarily exceed the acceptable
chronic exposure level. An evaluation of the data should
include the incidence and severity of all abnormalities, the
reversibility of abnormalities observed other than lethality,
gross lesions, body weight changes, effects on mortality, and
any other toxic effects.
In recent years guidelines have been established to improve
quality and provide uniformity in test conditions. Unfortunately,
many published LD50 or LCSO tests were not conducted in
accordance with current EPA or OECD guidelines since they were
conducted prior to establishment of guidelines. For this reason,
it becomes necessary to examine each test or study to determine
if the study was conducted in an adequate manner.
The following is a list of ideal conditions compiled from various
testing guidelines which may be used for determination of
23
-------
adequacy. Unfortunately, many published studies do not report
details of test conditions making such determinations difficult.
However, test conditions guidelines that might be considered
ideal may include:
animal age and species identified;
minimum of 5 animals per sex per dose group (Both sexes
should be used.};
14 day or longer observation period following dosing;
minimum of 3 dose levels appropriately spaced. (Most
statistical methods require at least 3 dose levels.);
identification of purity or grade of test material used
(particularly important in older studies);
if a vehicle used, the selected vehicle is known to be non-toxic;
gross necropsy results for test animals; or
acclimation period for test animals before initiating study.
Specific conditions for oral LD50-
dosing by gavage or capsule;
total volume of vehicle plus test material remain constant
for all dose levels; and
animals were fasted before dosing.
Specific conditions for dermal LD50:
exposure on intact, clipped skin and Involve approximately 102
of body surface; and
24
-------
animals prevented from oral access to test material by
restraining or covering test site.
Specific conditions for inhalation LC50*
duration of exposure at least 4 hours; and
if an aerosol (mist or particulate) the particle size (median
diameter and deviation) should be reported.
Although the above listed conditions would be included in an
ideally conducted study, not all of these conditions need to be
included in an adequately conducted study. Therefore, some
discretion is required on the part of the individual reviewing
these studies (EPA, 1985, OECD, 1987).
b. 14 Day or 28 Day Repeated Dose Toxicity
The following guidelines were derived using the OECD Guideline
for Testing of Chemicals (1987), for determining the design and
quality of a repeated dose short-term toxicity study. The
similarity between the conduct of a 14-day and 28-day study is
sufficient to consider them under the same guideline. The main
difference is the time period over which the dosing takes
place. These guidelines represent ideal conditions and studies
will not be expected to meet all standards in order to be
considered. For example, the National Toxicology Program's
cancer bioassay program has generated a substantial database of
short-term repeated dose studies. The study periods for these
range from 14 days to 20 days with 12 to 15 doses administered
25
-------
generally for 5 dose levels and a control. Since the quality of
this data is good, it is desirable to consider these study
results even Chough thev do not alwavs identically follow the
protocol.
The purpose of short-term repeated dose studies is to promote
information on possible adverse health effects from repeated
exposures over a limited time period. Where subchronic or
chronic data are lacking, short-term repeated dose studies of 28
davs or longer, with the application of appropriate uncertainty
factors, may be used bv this initiative to estimate acceptable
long-term exposure levels.
According to OECD Guidelines, short-term repeated dose studies
should include the following:
minimum of 3 dose levels administered and an adequate
control group used;
minimum of 10 animals per sex, per dose group (both sexes
should be used);
- the highest dose level should ideally elicit some signs of
toxicity without inducing excessive lethality and the lowest
dose should ideally produce no signs of toxicity;
ideal dosing regimes include 7 days per week for a period
of 14 days or 28 days;
all animals should be dosed by the same method during the
entire experiment period;
animals should be observed dally for signs of toxicity
a t-he ri-eatn«»nt ««riod (1 «. 14 or 28 HavO Animals
-------
which die during the study are necropsied and all survivors
in the treatment groups are sacrificed and necropsied at the
end of the study period;
all observed results, quantitative and incidental, should be
evaluated bv an appropriate statistical method;
clinical examinations should include hematology and clinical
biochemistry, urinalysis may be required when expected to
provide an indication of toxlcity. Pathological examination
should include gross necropsy and histopathology.
The findings of short-term repeated dose toxicity studies should
be considered in terms of the observed toxic effects and the
necropsy and histopathological findings. The evaluation will
include the incidence and severity of abnormalities, gross
lesions, identified target organs, body weight changes, effects
on mortality and other general or specific toxic effects (OECD,
1987).
c. Subchronic and Chronic Toxicity
The following guidelines were derived using the EPA Health
Effects Testing Guidelines (1985), for determining the quality of
a flubchronjc or chronic (long term) study. Additional detailed
guidance may be found in that document. These guidelines
represent ideal conditions and studies will not be expected to
meet all standards in order to be considered. The subchronic and
chronic studies have been designed to permit determination of
no-observed-effect levels (NOEL) and toxic effects associated
with continuous or repeated exposure to a chemical. Subchronic -?r~ «•>
27
-------
studies provide information on health hazards likely to arise
from repeated exposure over a limited period of time. They
provide information on target organs, the possibilities of
accumulation, and, with the appropriate uncertainty factors, may
be used in establishing safetv criteria for human exposure.
Chronic studies provide information on potential effects
following prolonged and repeated exposure. Such effects might
require a long latency period or are cumulative in nature before
manifesting disease. The design and conduct of such tests should
allow for detection of general toxic effects including
neurological, physiological, biochemical and hematological
effects and exposure-related pathological effects.
According to the EPA Guidelines, high quality subchronic/chronic
studies include the following:
minimum of 3 dose levels administered and an adequate control
group used;
minimum of 10 animals for subchronic, 20 animals for chronic
studies per sex, per dose group (both sexes should be used);
the highest dose level should ideally elicit some signs of
toxicity without Inducing excessive lethality and the lowest
dose should ideally produce no signs of toxicity;
ideal dosing regimes include dosing for 5-7 days per week
for 13 weeks or greater (90 days or greater) for subchronic
and at least 12 months or greater for chronic studies in
rodents. For other species, repeated dosing should ideally
28
-------
occur over 107, or greater of animals lifespan for subchronic
studies and 50% or greater of the animal's lifespan for
chronic studies;
all animals should be dosed by the same method during the
entire experimental period;
animals should be observed daily during the treatment period
(i.e., 90 days or greater);
animals which die during the study are necropsied and, at the
conclusion of the studv, surviving animals are sacrificed
and necropsied and appropriate histopathological
examinations carried out;
results should be evaluated by an appropriate statistical
method selected during experimental design; and
such toxicity tests should evaluate the relationship between
the do«.e of the test substance and the presence, incidence
and severity of abnormalities (including behavioral and
clinical abnormalities), gross lesions, identified target
organs,, body weight changes, effects on mortality and any
other toxic effects noted (EPA, 1985).
d. Reproductive and Developmental Toxicity
Studies considered here can be evaluated for quality by
comparing the study protocol or methods section with accepted
testing guidelines prepared by EPA, OECD or Interagency
Regulatory Liaison Group (IRLG). The EPA Health Effects Testing
Guidelines (1985) include guidelines for both reproduction and
fertility studies and developmental studies. These EPA o?ZT~~(f
29
-------
guidelines can serve as the ideal experimental situation with
which to compare study quality. Studies being evaluated do not
need to match precisely but rather should be similar enough that
one can be assured that the chemical was adequately tested and
that the results closely reflect the true reproductive or
developmental toxicity of the chemical.
Developmental toxicity can be evaluated via a relatively
short-term study in which the compound is administered during
the period of organogenesis. Some of the specific guidelines
for developmental studies are cited below.
- minimum of 20 young, adult, pregnant rats, mice or hamsters or
12 young, adult, pregnant rabbits recommended per dose group;
minimum of 3 dose levels with an adequate control group used;
the highest dose should induce some slight maternal toxicity
but no more than 10% mortality. The lowest dose should not
produce grossly observable effects in dams or fetuses. The
middle dose level, in an ideal situation, will produce minimal
observable toxic effects;
dose period should cover the major period of organogenesis (days
6 to 15 gestation for rat and mouse, 6 to 14 for hamster, and 6
to 18 for rabbit);
dams should be observed dally; weekly food consumption and
body weight measurements should be taken;
necropsy should include both gross and microscopic examination f the dams; the uterus should be examined so that the number of
ibryonic or fetal deaths and the number of viable fetuses can
counted; fetuses should be weighted; and
-------
- one-third to one-half of each litter should be prepared and
examined for skeletal anomalies and the remaining animals
prepared and examined for soft tissue anomalies.
The EPA Health Effects Testing Guidelines (1985) recommend a
two-generation reproduction study to provide information on the
ability of a chemical to impact gonadal function, conception,
parturition and the growth and development of the offspring,
Additional information concerning the effects of a test compound
on neonatal morbidity, mortality and developmental toxicity may
also be provided. The recommendations for reproductive testing
are lengthy and quite detailed and may be reviewed further in
the Health Effects Testing Guidelines. In general, the test
compound Js administered to the parental (F) animals (at least
20 males and enough females to yield 20 pregnant females) at
least 10 weeks before mating, through the resulting pregnancies
and through weaning of their Fl offspring. The compound is then
administered to the Fl generation similarly through the
production of their F2 offspring until weaning. Recommendations
for numbers of dose groups and dose levels are similar to those
reported for developmental studies, Details are also provided
on mating procedures, standardization of litter sizes (if
possible, 4 males and 4 females from each litter are randomly
selected), observation, gross necropsy and histopathology. Full
histopathology is recommended on the following organs of all
high dose and control P and Fl animals used In mating: vagina,
uterus, testes, epididymldes, seminal vesicles, prostate,
pituitary gland and target organs. Organs of animals from other
31
-------
dose groups should be examined when pathology has been
demonstrated in high dose animals (EPA, 1985).
As with any other tvpe of study, the appropriate statistical
analyses must be performed on the data for a study to qualify as
a good quality study. In addition, developmental studies are
unique in the sense that they yield two potential experimental
units for statistical analysis, the litter and the individual
fetus. The EPA testing guidelines do not provide any
recommendation on which unit to use, but the Guidelines for the
Health Assessment of Suspect Development Toxicants (EPA, 1986)
states that "since the litter is generally considered the
experimental unit in most developmental toxicity studies, the
statistical analyses should be designed to analyze the relevant
data based on incidence per litter or on the number of litters
with a particular end point". Others (Palmer, 1981 and Madson
et al., 1982) identify the litter as the preferred experimental
unit as well.
Information on maternal toxicity is very important when
evaluating developmental effects because it helps determine if
differential susceptibility exists for the offspring and
mothers. Since the conceptus relies on Its mother for certain
physiological processes, interruption of maternal homeostasis
could result in abnormal prenatal development. Substances which
affect prenatal development without compromising the dam are
considered to be a greater developmental hazard than chemicals
which cause developmental effects at maternally toxic doses.
32
-------
Unfortunately, maternal toxicitv information has not been
routinely presented in earlier studies and has become a routine
consideration in studies only recently. In an attempt to use
whatever cata are available, maternal toxicity information may
not be required if developmental effects are serious enough to
warrant consideration regardless of the presence of maternal
toxicity.
C. Tier Designation
1 . Carcinogens
Adequate weight-of-evidence of potential human carcinogenic effects
sufficient to calculate a Tier 1 Human Cancer Criterion (HCC)
generally consists of data sufficient to meet the categorical
definition of a Group A - Human Carcinogen and Group B - Probable
Human Carcinogen. Certain Group C - Possible Human Carcinogens may
also be suitable for Tier 1 criterion development. Where cancer
bioassays have been veil conducted, yet are limited because they only
involve a single animal species, strain or experiment and do not
demonstrate a high incidence, unusual site or type of tumor, or early
onset of tumor igenes is, such data may be suitable for Tier 1
criterion development.
As discussed earlier, data used for developing Tier 1 criteria are
expected to carry a high degree of certainty in their ability to
predict an effect. In this case, the quality of data and the
weight-of-evidence needs to be sufficient to ascertain that the
-------
chemical holds at least a good potential of producing carcinogenic
effects in humans.
For chemicals where the weight-of-evidence and quality of data is not
sufficient for Tier 1 numeric criteria the database may be adequate
to develop Tier 2 levels of protection. In this case, the data needs
to be sufficient to ascertain that the chemical is at least a
possible human carcinogen, i.e. Group C. As discussed previously
under Weight-of-Evidence and Appropriate Study Design, data on
chemicals in this Group suggest only limited evidence of
carcinogenicity. Studies may be flawed or lacking adequate design
or reporting yet show strong enough evidence of carcinogenicity or
the potential for carcinogenic effects such that the data should not
be ignored. Examples of such data may be studies where statistical
analysis may be lacking or tumor incidence may be only marginally
significant; tumor responses or lack of response may be attributable
to excessive dosing, or there may be high mortality in the exposed
groups also due to excessive dosing; increases exist for benign
tumors only with no evidence of mutagenicity, etc. Further
discussion as to how these data are treated in criteria derivation
and what potential differences may exist in such treatment will be
discussed further in the section on criteria development.
2. Noncarcinogens
All available toxicity data should be evaluated considering the full
range of possible effects of a chemical. Unfortunately, expansive
data exists for a limited number of chemicals. Although all data
34
-------
are evaluated, a line must be drawn below which data are not
sufficient for criteria development Adequate data necessary to
develop a Humai Noncancer Criterion (HNC) for noncancer effects should
ideally incorporate at least one well conducted epidemiologic study
which demonstrates a positive association between a quantifiable
exposure to a chemical and human disease. Such data exist for only a
few chemicals, therefore, reliance on animal data in establishing an
adequate margin of safety for humans is necessary. Although a more
extensive effects database is desirable, for this initiative, the
minimum database for a Tier 1 criterion must contain at least a well
conducted subchronic mammalian study. The duration of the study
must be at least 90 days in rodents or 10Z of the lifespan of other
appropriate species with exposure preferably via the oral route.
Such a study must establish a frank-effect-level (FED, a lowest-
observed-adver-se-effeet-level (LOAEL) and a no-observed-adverse-
effect-level (HOAEL). The study must be conducted in an animal
species relevant to humans based on a defensible biological rationale
and generally follow the studv protocol previously discussed. To
further reduce uncertainty, data from longer studies approaching the
lifetime of the test animal are preferrable. In some cases, chronic
studies of one year or longer in rodents or 50% of the lifespan or
greater in other appropriate test species may be sufficient. Dose
response must be demonstrated in these longer term studies, however a
LOAEL involving relatively mild and reversible effects may be
considered an acceptable data point for decision making. Reproductive/
developmental effects data as well as supportive epidemiologic
evidence of similar effects seen in test animals are also highly
desirable in order to evaluate the full range of potential adverse
effects to humans.
-------
When data are not sufficient to meet the minimum requirements for
deriving Tier I numeric criteria, such data mav be considered for
development of Tier 2 levels of protection. As with Tier I, all
available data should be considered, however, a minimum database
suitable for Tier 2 must contain at least a well conducted subacute
mammalian studv with an exposure period of at least 28 davs,
preferably via the oral route of exposure. The study should
establish a dose-response relationship including a frank-effect-level
(FEL), a lowest-observed-adverse-effect-level (LOAEL), and a
no-observed-adverse-effect-level (NOAEL). Acceptable protocol for
conducting such 28 day studies may be found in the OECD
Testing Guidelines (OECD, 1987) as discussed previously. Although
the effects observed from short duration studies are usually fewer
than normally evaluated in longer duration studies, such effects
should at least include mortality, clinical observations, body weight
changes and necropsy of major organs with whatever histopathology
that mav be available. The minimum data point for decision making on
such short term exposure data must be a NOAEL. Structure-activity
relationship (SAR) review should also accompany the minimum data
evaluation. SAR compares a chemical with substances that have
structural similarities in order to predict whether the chemical
might cause similar toxic effects. Such information may then be
used in deciding what uncertainty factors may be appropriate to
apply to such limited data in order to protect against potential
similar effects.
Studies of longer duration than 28 days and with greater evaluation
of effects are more desirable for use in Tier 2 and may allow the use
36
-------
of a LOAEL for decision making, depending on the quality of the
study. As with Tier 1, reproductive/developmental effects data as
well as anv supportive epidemiologic evidence is highly desirable in
order to evaluate the full range of potential adverse effects of the
chemical. As with carcinogens, further discussion as to how these
data will be applied in the derivation of acceptable exposure levels
and what adjustments must be made to account for uncertainty will be
discussed in further detail in the section on criteria development.
III. PRINCIPLES FOR CRITERIA DEVELOPMENT
A. General
The process to derive Tier 1 criteria or Tier 2 levels of protection is
generally the same. The weight of evidence and level of certainty in the
data available for calculating acceptable exposure levels establishes the
major difference between the two. For risk assessment of noncarcinogenic
effects, the minimum data requirements differ between tiers. Therefore,
differences in adjustments to the data (i.e., uncertainty factors) may
also occur between tiers. These differences reflect differing levels of
certainty for establishing adequate margins of safety. In the case of
carcinogens, the same quantitative risk assessment approach generally
followed for Tier 1 is used as veil for Tier 2 when the data allow.
When the bioassay data for Tier 2 carcinogens are not suitable for
quantitative risk assessment yet the weight-of-evidence supports concern
for possible carcinogenic effects, an additional uncertainty factor may
be applied to the LOAEL or NOAEL for the chemical in order to establish
an extra marzin of safetv reflective of this concern. - -5&3
-------
All available appropriate human epidetniologic data and animal toxicologic
data shall be considered. Data specific to an environmentally
appropriate route of exposure shall be used for criteria development,
i.e. oral, dermal or inhalation versus injection, implantation, etc.
Findings from studies using less than appropriate routes of exposure may
be considered supportive of data obtained through more appropriate
routes. Although local effects are important, for the purposes of this
initiative oral exposure should be considered preferential to dermal and
inhalation data since ingestion is the primary route of exposure, i.e.
water and fish consumption. Caution must be exercised in the use of
dermal and inhalation data. Strong consideration must be given for
pharmacokinetlc information on absorption, distribution and metabolism in
establishing equivalent doses with oral exposure. Effects produced
through exposure via a non-oral route generally should be as a result of
svstemic distribution of a toxicant rather than as local effects to the
skin or the respiratory tract.
In general, study results shall be converted, as necessary, to the
standard unit of milligrams of toxicant per kilogram of body weight
per day (mg/kg/day).
If a study does not specify water or food consumption rates, or body
weight of the test animals, standard values may be used for the test
species, such as may be obtained from the National Institute of
Occupational Safety and Health, Registry of Toxic Effects of Chemical
Substances (RTECS) or similar appropriate references.
Study results from multiple exposures shall be adjusted, as
necessarv. to a dailv dose exoosure as if received dailv for the
-------
duration of the exposure period. The exposure period shall be
defined as the interval beginning with administration of the first
dose through the last dose, inclusively.
8. Carcinogens
1. Mechanism
The mechanism by which chemicals cause cancer is not completely
known, and mav involve a varietv of mechanisms occurring at various
stages in the carcinogenic process. A chemical may act at a single
stage or more :han one stage. Currently, the dominant theory
regarding the process bv which a chemical causes cancer is based on
two stages' initiation and promotion (Borzsonyi, 1984; OSTP, 1985;
Trosko, 1983; Williams, 1986), The concept of two-stage
carcinogenesis has been supported by Investigations involving skin
and liver systems (Argyris, 1985; Pitot and Sirica, 1980). This
operational theory allows the classification of carcinogens according
to their apparent biological activity. Some chemicals are capable,
by a variety of genotoxlc mechanisms, of triggering the carcinogenic
process (initiation). Other chemicals may only alter the expression
of the initiated genome and enhance tumor development by a variety of
non-genetic mechanisms (promotion). Complete carcinogens operate by
both processes. Initiators are capable of directly altering in an
irreversible manner the native structure of the DNA. Promotion may
be reversible in the early stages, appears to be highly dose-
dependent, and apparently requires prolonged or repeated exposure
(Pitot, 1981; Slaga, 1984; Thomas, 1986).
-------
Calling an agent a promoter does not eliminate the carcinogenic
hazard potential of a chemical. Indeed, data indicate that promoting
phenomena are largely responsible for the expression of many human
cancer tvpes (Williams, 1986) However, it is verv difficult both in
principle and in practice to confirm the assertion that a given
chemical acts bv promotion alone (OSTP, 1985).
Currently, for most if not all chemicals, data are not available to
determine the exact mechanism by which they cause cancer. As a
result, significant controversy exists regarding the existence of
thresholds for carcinogens. Therefore, for the purpose of routine
cancer risk assessment, agents that are positive in long-term animal
experiments should be considered as complete carcinogens unless
there is evidence to the contrary because, at present, it is
difficult to determine whether an agent is acting only as a
promoting or cocarcinogenic agent (EPA, 1986).
For the purpose of this initiative, unless adequate data
demonstrate otherwise, a nonthreshold mechanism will be assumed for
those chemicals classified as Group A, B and C carcinogens.
2. Data Review
If acceptable human epidemiologic data are available, a risk
associated dose shall be set equal to the lifetime exposure which
would produce an Incremental increased cancer risk of 1 in 100,000.
If more than one study is judged acceptable, the study resulting in
the most protective risk associated dose is generally used to ^/ <^"
calculate the human cancer criterion.
-------
In the absence of appropriate human studies, data from a species
that responds most like huaans should be used, if information to
this effect ecists. Where several studies are available, which may
involve different animal species, strains, and sexes at several
doses and bv different routes of exposure, the following approach to
selecting the data sees is used*
The tumor incidence data are separated according to organ site
and tumor type.
All biologically and statistically acceptable data sets are
presented.
The range of the risk estimates is presented with due regard to
biological relevance (particularly in the case of animal
studies) and appropriateness of route of exposure.
Because is it possible that human sensitivity is as high as the
most sensitive responding animal species, in the absence of
evidence to define the most relevant species to humans, the
biologically acceptable data set from long-term animal studies
showing the greatest sensitivity should generally be given the
greatest emphasis, again with due regard to biological and
statistical consideration (EPA, 1986).
Exceptions tc the above may exist as follows:
If two or more studies exist which are identical with respect to
species, strain, sex and tumor type and are of equal quality, the
geometric mean of the potency from these studies may be used (E?A,
1980). In certain instances where there are several studies in
-------
various strains and even several species and where there Is no
indication of a single studv or species judged most appropriate, the
geometric mean estimates from all studies may be used to determine
the potency. This ensures that all relevant data are included in the
derivation (EPA, 1989d)
Where two or more significantly elevated tumor sites or types are
observed in the same study, extrapolations may be conducted on
selected sites or types. These selections will be made on
biological grounds. To obtain a total estimate of carcinogenic
risk, animals with one or more tumor sites or types showing
significantly elevated tumor incidence should be pooled and used for
extrapolation so long as double-counting of tumor-bearing animals is
prevented. The pooled estimates will generally be used in
preference to risk estimates based on single sites or types.
Quantitative risk extrapolations will generally not be done on the
basis of totals that include tumor sites without statistically
significant elevations (EPA, 1986).
Benign tumors should generally be combined with malignant tumors for
risk estimates unless the benign tumors are not considered to have
the potential to progress to the associated malignancies of the same
histogenic origin. The contribution of the benign tumors, however,
to the total risk should be indicated (EPA, 1986).
42
-------
3. Model
a. Nonthreshcld Approach
When acceptable human epidemiologlc data are not available and a
nonthreshold mechanism is assumed, carcinogenesis bioassay data,
as appropiiate, are fitted to a linearized multistage computer
model. (Mote: Other models, such as time-to-tumor,
modifications or variations of the multistage model may be used
which consider the data more appropriately under case-by-case
circumstances.)
Since risks at low exposure levels cannot be measured directly
either by animal experiments or by epidemiologic studies, a
number of mathematical models have been developed to extrapolate
from high to low dose. Different extrapolation models, however,
may fit the observed data reasonably well but may lead to large
differences in the projected risk at low doses (EPA, 1986).
"No single mathematical procedure is recognized as the most
appropriate for low-dose extrapolation in carcinogenesis. When
relevant biological evidence on mechanism of action exists
(e.g., pharmacokineticsi target organ dose), the models or
procedures employed should be consistent with the evidence.
When data and information are limited, however, and when much
uncertainty exists regarding the mechanism of carcinogenic
action* models or procedures which incorporate low-dose
linearity are preferred when compatible with the limited ^.
06?
information." (OST?, 1985)
-------
In an attempt to characterize the underlying dose-response
relationship, models which use the nonthreshold assumption of
carcinogenicity are commonly used. The linearized multistage
(LMS) model calculates an upper bound based on the theory that a
developing tumor goes through several different stages which can
be affected by a chemical carcinogen. The LMS model is forced to
be linear in the low-dose region, regardless of the shape of the
dose response curve, and therefore LMS-based risk estimates may
be regarded as relatively conservative when used for public
health protection.
In calculating upper bounds on potency from the LMS model, the
bloassay data are fitted to the LMS model, e.g. Global 82 developed
by Howe and Crump (1982). The 95 percent upper bound estimate on
the linear term, q.*, is used to calculate the upper confidence
bound on risk for a given dose, or the lower confidence bound on
dose for a given risk. The slope factor (q.*) is taken as an
upper bound of the potency of the chemical in inducing cancer at
low doses.
When pharmacokinetic or metabolism data are available, or when
other substantial evidence on the mechanistic aspects of the
carcinogenesis process exists, a low-dose extrapolation model
other than the linearized multistage procedure might be
considered more appropriate on biological grounds. When a
different model is chosen, the risk assessment should clearly
discuss the nature and welght-of-evidence Chat led to the
choice. Considerable uncertainty will remain concerning
44
-------
response at low doses, therefore, in most cases an upper-limit
risk estimate using the linearized multistage procedure should
also be presented for comparison (EPA, 1986).
fa. Threshold ^pproacn
Whenever appropriate human epidemiological data are not
available, and the preponderance of data suggest that a
chemical causes cancer via a threshold mechanism, the risk
associated dose may be calculated via a method other than the
linearized multistage model on a case-by-case basis.
As a default, a safety factor approach may be pursued after
thorough evaluation of the toxicologic and pharmacologic data on
the compound including, but not limited to* mechanism of
carcinogenesis, number and type of tumors induced, the
spontaneous incidence of tumors, the number of animal species
tested and affected, metabolic considerations, epideoiologic
data, extent of data supporting a non-genotoxic mechanism of
tumor induction, i.e. mutagenicity assay data, initiation/
promotion assay data, etc.
4. Lifespan Adjustment
If the duration of the study (Le) is significantly less than the
natural lifespan for the species (L) , the slope factor, q.* is
adjusted to account for unobserved tumors due to the short study
45
-------
duration. The assumption is that if the duration of the study was
increased, tumor incidence would continue to increase as a constant
function of the background rate. It is assumed that the cumulative
tumor rate would increase at least bv the 3rd power of age since age
specific rates for humans increase at least by the 2nd power of
age and often considerably higher (EPA, 1980).
For mice and rats, the natural lifespan (L) is defined as 90 weeks
and 104 weeks, respectively, the slope factor adjustment will be
conducted for mice and rat data if the study duration (Le) is less
than 78 weeks for mice or 90 weeks for rats, by multiplying the slope
factor by the factor (L/Le) . For other species, this adjustment
factor should also be used whenever appropriate, using species-
specific values for L and the Le trigger level. The latter may be
determined using the trigger levels for mice and rats as a guideline.
5. Species Scaling
Low-dose risk estimates derived from laboratory animal data
extrapolated to humans are complicated by a variety of factors that
differ among species and potentially affect the response to
carcinogens. Included among these factors are differences between
humans and experimental test animals with respect to life span, body
size, genetic variability, population, homogeneity, existence of
concurrent disease, pharmacokinetic effects such as metabolism and
excretion patterns, and the exposure regimen.
Lf*
-------
The usual approach for making interspecies comparisons has been to
use standardized scaling factors. Commonly employed standardized
dosage scales include ng per kg bodv weight per day, ppm in the diet
or water, mg per m body surface area per day, and mg per kg body
weight per lifetime. In the absence of comparative toxicological,
physiological, metabolic and pharmacokinetic data for a given
chemical, extrapolation shall be made on the basis of surface area
(EPA, 1986).
The species sensitivity factor is calculated by dividing the average
weight of a human (Wh) by the weight of the test species (Wa) and
taking the cube root of the resultant value. This is based on the
premise that a close approximation of the surface area is 2/3 the
power of weight, and that the exposure in og-2/3 the power of body
weight/day is suniarlly considered to be an equivalent exposure (EPA,
1980). The animal slope factor is multipled by this factor to obtain
the human slope factor.
q * (human) • q.* (animal) x
\/
Wa kg
The weight (Wa) of the test species should be the average adult
weight from tbe particular bioassay if possible, or derived from
available data tables or standard assumed weights.
47
-------
C. Noncarcinogens
1 Mechanism
Noncarcinogens generally are assumed to have a threshold dose or
level below which no adverse effects should be observed (NOAEL).
For many noncarcinogenic effects, protective mechanisms are believed
to exist that must be overcome before an adverse effect is
manifested. For example, where a large number of cells perform the
same or similar function, the cell population may have to be
significantly depleted before the effect is seen. As a result, a
range of exposures exists from zero to some finite value that can be
tolerated by the organism with essentially no expression of adverse
effects. In the development of a margin of safety from exposure to a
chemical, the effort exists to find the upper bound of this tolerance
range (i.e., the maximum subthreshold level). Because variability
exists in the human population, attempts are made to assure a
subthreshold level which is protective of sensitive individuals in the
population. For most chemicals, this level can only be estimated and
incorporates the use of uncertainty factors indicating the degree of
extrapolation used to derive the estimated value (EPA, I989d).
Exceptions to this principle exist. Noncarcinogenic chemicals may
exist with no Identifiable threshold. Examples of this
exception may include genotoxic teratogens and germline mutagens.
These agents have been specifically identified to differentiate
between chemicals thought to produce reproductive and/or
developmental effects via a genetically linked effect from those /3/J5*
68
-------
chemicals more routinelv considered to act via a nongenetic
mechanism. There are few chemicals, if any, which currently have
sufficient mechanistic information about their mode of action to
link teratogeric or developmental effects to mutational events
during organogenesis, histogenesis or other stages of development.
These chemicals mav also interact with germ cells to produce
mutations which may be transmitted to the zygote and also be
expressed during one or more of these stages of development.
EPA has recognized this potential and discussed this issue in their
1989 Proposed Amendments to Agency Guidelines for Health Assessments
of Suspect Developmental Toxicants (EPA, 1989c) and in their 1986
Guidelines foi Mutagenicity Risk Assessment (EPA, 1986). Various
statements within these guidelines should raise concern for the
potential for future generations inheriting chemically induced
germline mutations or suffering from mutational events occurlng in
utero:
"It is estimated that at least 10% of all human disease is
related to specific genetic abnormalities..."
"Life in our technological society results in exposure to many
natural and synthetic chemicals. Some have been shown to have
mutagenic activity in mammalian and sub-mammalian test systems.
and these may have the potential to Increase genetic damage in
the human population... The extent to which exposure to natural
and synthetic environmental agents may have increased the
frequency of genetic disorders In the present human population
And contributed to the mutational "load" that will be
-------
transmitted to future generations is unknown at this time.
However, for the reasons cited above, it seems prudent to limit
exposure to potential mutagens."
"Approximately 3Z of newborn children are found to have one or
more significant congential malformation at birth, and by the
end of the first postnatal year, about 3t more are found
to have serious developmental defects. Of these, it is estimated
that 20Z are of known genetic transmission, 10Z are attributable
to known environmental factors..."
An awareness of the potential for such teratogenic/mutagenic effects
should be established in order to deal with such data should it
occur in the future. However, without adequate data to support a
genetic or mutational basis for developmental or reproductive
effects, the default becomes an uncertainty factor approach. This
approach follows the procedure identified for noncarcinogens assumed
to have a threshold. Genotoxic teratogens and germline mutagens
should be considered an exception while the traditional uncertainty
factor approach is the general rule for calculating criteria or
levels of protection for chemicals demonstrating developmental/
reproductive effects.
A nonthreshold mechanism shall be assumed for genotoxic teratogens
and germline mutagens. Since there is no well established mechanism
for calculating criteria protective of human health from the effects
of these agents, criteria will be established on a case-by-case
37S
50
-------
2. Data Review
All toxicity data on a chemical should be evaluated for criterion or
level of concern development. Those studies representing the best
quality and most appropriate data as discussed previously under
appropriate study design should be selected for defining adverse
effects and their level of occurance. As previously discussed,
adequate human epidemiologic data should be used in evaluating the
adverse health effects of a chemical whenever available. When
adequate human data are not available, animal data from species
most relevant to humans should be used. In the absence of data on
the "most relevant" species or the inability to identify the most
relevant species, data from the most sensitive animal species tested,
i.e., the species demonstrating an adverse health effect at the
lowest administered dose via a relevant route of exposure, shall
generally be used.
For guidance, adverse health effects are those deleterious effects
which are or may become debilitating, harmful or toxic to the normal
functions of an organism including reproductive and developmental
effects. These do not include such effects as
tissue dlscoloxation without other noted effects, or the induction
of enzymes involved in the metabolism of the substance. Guidelines
for defining the severity of adverse effects have been suggested by
Hartung and Dutkin (1985) which proposes a ranking from slight to
severe effects. Distinguishing slight effects such as reversible
enzyme induction and reversible subcellular change from more severe
effects is critical in distinguishing between a no observed adverse ~~-"£J~Jf
effect level (NOAEL) and a low-observed-adverse-effect (LOAEL).
-------
The experimental exposure level representing the highest dosage
level tested at which no-adverse-effects were demonstrated (NOAEL)
shall be used in the formula for criteria development. In the
absence of such data, the dosage level at which the lowest-observed
adverse-effect-level was demonstrated may be used in some
circumstances for criteria development.
Preference should be given to studies Involving exposure over a
significant portion of the animal's lifespan since this is
anticipated to reflect the most relevant environmental exposure. An
exception to this is where reproductive and/or developmental effects
mav be demonstrated to have a lower NOAEL over a shorter exposure
period. When two or more studies of equal quality and relevance
exist, the geometric means of the NOAEL or LOAEL may be used.
3. Uncertainty Factors
Since the adverse effects data for many chemicals are quite limited,
varying degrees of uncertainty surround the margin of safety that
would exist by direct use of this data for criteria development.
Consequently, adjustments or uncertainty factors are used to assure
*
an appropriate margin of safety for protection of human health
exists when it is necessary to use data of limited certainty in
criteria development.
The following are examples of where uncertainty exists as a result
of weakness either in the data base or the process which needs
accommodation:
52
-------
using dose-response information from effects observed at high
doses to pi edict the adverse health effects that may occur
following exposure to the low levels expected from human contact
with the agent in the environment;
using dose- response information from short-term exposure studies
to predict the effects of long-term exposures, and vice-versa;
using dose-response information from animal studies to predict
effects in humans; and
- using dose-response information from homogeneous animal
populations or healthy human populations to predict the effects
likely to be observed in the general population consisting of
individuals with a wide range of sensitivities. (EPA, 1989d)
For this initiative, accommodation for these uncertainties will
be handled in the following process. For further detail in the
selection of these uncertainty factors, please see Appendix A.
a. An uncertainty factor of 10 shall generally be used when
extraplating from valid experimental results from studies on
prolonged exposure to average healthy humans. This 10-fold
factor is used to protect sensitive members of the human
population.
b. An uncertainty factor of 100 shall generally be used when
extrapolating a NOAEL from valid results of long-term studies on
-------
experimental animals when results of studies of human exposure
are not available or are madeauate. In addition to (a) above,
this represents an additional 10-fold uncertainty factor in
extrapolating data from the average animal to the average human.
c. An uncertainty factor of up to 1000 may be used when
extrapolating a NOAEL from animal studies for which the exposure
duration is less than chronic or when other significant
deficiencies in study quality are present and when useful
long-term human data are not available. In addition to (a) and
(b) above, this represents an additional uncertainty factor of up
to 10-fold. The level of additional uncertainty applied depends
on the duration of the study used relative to the lifetime of the
experimental animal.
d. An additional uncertainty factor of between one and ten may be
used when deriving a criterion from a lowest-observed-adverse-
effect level (LOAEL) or when less than adequate quality data
must be used. This uncertainty factor reduces the LOAEL into
the range of a no-observed-adverse-effect level (NOAEL). The
level of additional uncertainty applied depends upon the
severity of the observed adverse effect and/or the quality of
the experimental data available.
e. An additional uncertainty factor of between one and ten may be
used when there are limited effects data or incomplete subacute
or chronic toxicity data, such as when deriving a level of
protection based on subacute 28 day exposure data. This
54
-------
additional subacute to chronic application factor is assumed to
better adjust the 10-fold uncertainty factor described in (C) for
such short term study data to a chronic study no-observed-
adverse-effect level (NOAEL) . The level of quality or extent
of the experimental data available as well as structure-activity
relationship information determines the factor selected.
D. Exposure Assumptions
When dealing with site specific and individual specific exposure, it is
more accurate to use actual available exposure information to estimate
an individual's specific risk or margin of safety. Individual behaviors
can be assessed anc specific activity information compiled to address
quantity, frequenc\ and duration of exposure. When dealing with such
diverse populations of individuals covering as large an area as the
Great Lakes Basin, extreme ranges of behaviors and activities are likely
Therefore, deriving, default assumptions that can estimate reasonable
exposures which address the vast majority of the Basin population becomes
necessary.
1. Body Weight
National body weight data has been compiled by the National Center
for Health Statistics from a survey conducted from 1976 through 1980
entitled the socond National Health and Nutrition Examiantion Survey
(NHANES II). Approximately 28,000 people aged 6 months to 74 years
were surveyed with other 20,000 individuals actually interviewed and
examined. Weighted mean body weights have been determined from this
data. Since body weights change so rapidly during childhood, it is
-------
reasonable to use mean adult body weight to reflect population body
weights when assuming a long exposure duration. From national survey
data, the mean adult body weight appears to be approximately 72kg.
There is some evidence the mean may even be higher for the Great
Lakes Basin population. However, as a matter of convention, 70 kg
has been used for many years in chemical regulatory programs and
still appears appropriate for this initative.
Body Weights of Adults (kilograms)
Men
Women
Std. Error
Age
18 < 25
25* 35
35* 45
45*55
55*65
65<75
18*75
Mean
73.7
78.7
80.8
81.0
78.8
74.8
78.1
of Mean
0.0035
0.0034
0.0040
0.0041
0.0041
0.0051
0.0016
Mean
60.6
64.2
67.1
67.9
67.9
66.6
65.4
Men and Women
Std. Error Std. Error
of Mean
0.0032
0.0037
0.0043
0.0044
0.0045
0.0048
0.0017
Mean of Mean
67.2
71.5
74.0
74.5
73.4
70.7
71.8
56
3SY
-------
Body Weights of Children (kilograms)
Age
Mean
Boys
Girls
Std. Error
of Mean Mean
Bovs and Girls
Std. Error
of Mean Mean
Std. Error
of Mean
< 3
3< 6
6< 9
9 < 12
12< 15
15< 18
11.9
17.6
25.3
35.7
50.5
64.9
0.0016
0.0014
0.0023
0.0038
0.0051
0.0047
11.2
17.1
24.6
36.1
50.7
57.4
0.0011
0.0015
0.0024
0.0043
0.0049
0.0042
11.6
17.4
25.0
36.0
50.6
61.2
2. Duration of Exposure
(USEPA, 1989a)
a. Tine Spent at Home
Although even the most homebound individual may periodically get
away from home for some period of time, the standard default
assumption is to assume an individual is at home 100Z of the
time. Several time use studies have been conducted to identify
human activity patterns by which duration of exposure to
contaminants in a particular scenario can be analyzed. One "^C~)—:>
57
-------
particular study by Robinson, 1977 was a nationally representative
study conducted from fall 1975 to fall 1976 by the University of
Michigan's Institute for Social Research. The sampling base
included the entire population 18 years and older, regardless of
occupation, excluding the population in Institutional settings,
i.e. nursing homes, college dormitories, etc. Diaries were kept
as to time spent in various activities by the respondents and
several interviews of the respondents were conducted as well.
The value of such data may be demonstrated in estimating the
importance of exposure to contaminants received at their
residence (i.e. drinking water) relative to time spent away from
home where drinking water, etc. may come from a different source.
Evaluation of this data by the USEPA resulted in the following
mean of exposure duration per week.
Men
Women
Average Adult
Time spent at home 97.80 hours 115.98 hours 107.59 hours
Time spent away 70.27 hours 52.10 hours 60.49 hours
from home
b. Population Mobility
The default assumptions for mobility is to consider that an
individual remains in the same residence for a "lifetime". >*^c~> -p
Movement of individuals from individual residences, communities,
58
-------
or even regions of the country mav influence exposure duration
to contaminants from sources such as drinking water and sport
caught fish dramatically. If movement occurs within the same
community, the influence by drinking water may not change. If
movement is still within the region, the influence of
contaminated sport fish may not change. Be that as it may,
mobility may lower exposure duration.
Based on a survey conducted by the Oxford Development Corporation,
a property management company, the average residence time for an
apartment dweller is estimated to range from 18 to 24 months. A
survey conducted by the Bureau of the Census in 1983, determined
that 93Z of householders moved into their present home between
1950 and 1983. Using this information, the following time of
residence ranges have been determined:
YearsinCurrent Home Total % of Householders
0-1 7.5
1 - 3 16.9
3-13 40.2
13 - 18 H.O
18 - 23 7.9
23 - 33 9.5
> 33 7.0
Based on these statistics, the 50th percentile of householders
living in their current residence is 9.4 years and the 90th <~2& /
percentile is 29.8 years.
-------
c.
Life Expectancy
Statistical data on life expectancy is gathered annually by the
U.S. Department of Commerce. Data presented by the Bureau of
Census for 1985 show that life expectancy for the total U.S.
population is 74.7 years. The breakdown of this average is as
follows:
Male
Female
Total
white
71.8
78.7
75 3
black and other
67.2
75.2
71.2
black
65.3
73.7
69.5
total average
71.2
78.2
74.7
(USEPA, 1989a)
Although the average life expectancy now is approximately 75 years,
given the mobility of the population, as previously demonstrated,
i.e., time spent away from home and lower number of years spent in
the same residence, the traditional default value for "lifetime"
exposure of 70 years appears adequate for considering chronic
"lifetime" exposure.
60
-------
3. Incidental Exposure
Although the adjustment for recreational exposure appears
insignificant, it provides a clear indication that such a source of
exposure has been considered by this process. The suggested 0.01
liter/day is based on the assumption that a swimmer may consume a
mouthful (30 ml) of water per swimming event. A worse case
assumption might be an average of one swimming event per day during
the four month warm weather period starting in mid May and ending in
mid September. Exposure potential, when averaged over a year,
equals 0 01 liters/day for a lifetime. No estimate of duration of
swimming event is made in the swimming assumption. However, if one
would assume each event occurs for a one hour duration, total annual
exposure duration could be estimated to be 123 hours/year.
EPA has recently estimated a national average frequency of swimming
to be 7 days/year with a 2.6 hour duration (EPA, 1989d). An earlier
EPA publication estimated an average annual frequency of 9 days/year
with a 2 hour duration of exposure. Other total body contact
recreation such as water skiing was also identified as having
approximately 20 million participants with a total exposure of 260
million hours per year or an average of 14 hours exposure per
participant. Partial body contact was identified as 20% body
exposure for fishing and 40Z body exposure for boating. This earlier
reference listed 68 million people involved nationally in boating
with an average duration of 1600 million person hours per year and 5
million people involved nationally in fishing with 6600 million
person hours duration per year. The resulting individual average
61
-------
exposure duration equals approximately 24 hours and 122 hours of
participation, respectively (EPA, 1979). If each hour of total body
contact equivalent for bathing, water skiing, boating and fishing
were calculated (18, 14, 10 and 24 hours, respectively), the total
equals 66 hours of average body contact exposure.
Various recreational surveys have been conducted in Michigan and may
serve as a typical example of Great Lakes Basin activity.
Estimations similar to EPA's for activities per participant and
hours per participation may be calculated from this information. If
we were to assume an individual were to participate in all
activities for the number of days listed from the 1981 Michigan
Travel and Recreation Survey and for the duration of hours per
participation as identified in the 1976 Recreation survey, and the
percentage adjustment made for total body contact exposure from the
older EPA reference, the following calculations might be made:
Activity Days Hours per Body Contact Hours of
per Participant Participation Adjustment Exposure
Swimming
Fishing
Power Boating
Water Skiing
Sailing
Canoeing
13.3
14.3
24.5 (total)
9.6
10.4 (total)
4.8
2.1 (ave.)
3.7 (ave.)
3.2
1.5
3.2
3.9
1.0
.2
.4
1.0
.4
.4
27.9
10.6
31.4
14.4
13.3
7.5
TOTAL
105.1
(Wells, 1990)
-------
TABLE 2
WATER-BASED RECREATION ACTIVITY
3Y MICHIGAN RESIDENTS, 1976 & 1981
- Recreation by Michigan Residents (Michigan 1976 Recreation Survey)
Participations (b) Hours Per Hours Per
Activity Per Rasufcpt Resident Participation
Swimming - Inland Lakes or
Streams
Swimming - Great Lakes
Fishing (a) - Inland Lakes
Fishing - Streams • .«. j.* ~> «.
Fishing - Great Lakes 07 29 42
Fishing - Ice Fishing 05 20 40
Power Boating 14 44 32
Water Skiing 07 10 15
3.9 8.6 2 2
08 16 20
2.6 8.9 3 4
1.2 3.7 3 2
r w*w * w w *• 11 ?«i
Water Skiing
Sailing 04 13 3.2
Canoeing 04 15 39
Rowing 03 05 16
Other Boating 01 01 2.0
1981 • Recreation by Michigan Residents (age 2 and over)
(1981 Michigan Travel & Recreation Survey)
No of Percentage of
Participants Population Activity Days (c) Activity Days
Activity (Hillionsl Participating Per MI Resident Per Participant
Lake/Stream
Swimming/Sun-
Bathing 31 36 48 13.3
Fishing 3 1 36 5.1 14 3
Power Boating -
Inland Lakes 15 17 21 11.8
Great Lakes 0.9 10 1 3 12.7
Water Skiing 0.9 11 11 96
Sailing -
Inland Lakts 0.4 5 0.4 7.1
Great Lakes 0 ? 3 0.1 3.3
Canoeing 1 3 15 0.7 4.8
(a) Total for all types of fishing is 5.0, not 10.0 as stated in 19
Recreation Plan.
(b) Person could participate more than once IQ a given activity in one day
(c) "On how many different days did (you} do the activity'"
Dwells, 1990)
63
-------
Given these comparisons of water recreation activities, the suggested
incidental exposure level appears appropriate, given
the variability in individual behavior.
Drinking Water
Two liters of water has been the nationwide conventional estimate of
adult human's daily water consumption. According to EPA, the 2
liters of water per day is a historical figure set by the U S. Army
in determining the amount of water needed for each person in the
field. The National Academy of Sciences (NAS) estimates that daily
water consumption may vary with physical exercise and fluctuations
in temperature and humidity. It is reasonable to assume those
living in a more humid, hot climage will consume higher levels of
water. NAS has calculated the average per capita water consumption
to be 1.64 liters per day. The National Cancer Institute (NCI) also
has looked at this issue with an overall tap water consumption rate
of 1.39 liters of water per day as their study average The NCI
study is of particular interest since data was compiled from Detroit,
Iowa, New Jersey and Connecticut giving a database of over 3500
respondents with similar weather conditions to the Great Lakes Basin.
The consumption rate of less than or equal to 1.96 liters per day is
equated to the 100Z cumulative frequency level as seen in the
following table*
64
-------
Frequency Distribution of Tap Water Consumption Rates*
Consumption Rate (L/day) Cumulative Frequency (%)
£ 0.80 19.2
0.81 - 1.12 39.6
1.13 - 1.33 59.7
1.45 - 1.95 79.9
< 1.96 100.0
*Represents consumption in a "typical" week.
Other researchers have discovered average levels both higher and
lower than NCI. The Food and Drug Administration's (FDA) Total
Diet Study estimated rates for water and water-based foods for two
groups of adults to be 1.07 and 1.3 liters per day with an average
of 1.2 liters per day. The U.S. Department of Agriculture (USDA) in
the 1977-78 Nationwide Food Consumption Survey identified daily
beverage intakes of from 1.24 to 1.73 liters per day. After review
of all these studies, EPA has judged the average adult drinking
water consumption rate to be 1.4 liters per day with a reasonably
convervative assumption of 2 liters per day as being the 90th
percentile value (USEPA, 1989a).
65
-------
5. Fish Consumption
Much debate has occured over the years as to the appropriate
regionally caught fish consumption rate for the Great Lakes
Basin. This is one area where extreme differences exist in the
region's consumption behavior. A large segment of the population
consumes little or no fish caught from the region, while a small
segment of the population consumes a significant quantity of
regionally caught fish. For this initiative, the sport angler
and his/her family are the focus for determining a mean
regionally caught fish consumption value.
Several studies of fish consumption and sport angler behavior have
been evaluated to estimate an appropriate fish consumption value
for the region. The details of this evaluation will be discussed
further in an appendix. Three regional surveys; Michigan (West,
1989), Wisconsin (Fiore, 1989) and New York (Connelly, 1990);
have been used to form the basis of this initiative's assumption.
In summary, the results of the Michigan survey suggest that
approximately 65% of the licensed anglers consume less than one
meal per week of all fish. This is consistent with Wisconsin
data which estimates the mean annual total number of all fish
meals consumed by anglers to be 41. This is also consistent with
New York anglers who consume 45.2 meals statewide and
approximately 41.6 total meals in the regions with the greatest
number of sport anglers and greatest sport fishing effort. Based
on the Michigan and Wisconsin surveys, approximately 43% of the
fish meals consumed are sport caught, or approximately 18-19
66
-------
meals per >ear. Estimates of meal sizes range up to 8 ounces
(0.5 pounds) or an approximate total of 9-9.5 pounds per vear.
This equates to a daily fish consumption rate of 11-12 gms/day.
The Michigan survey data indicate a mean annual total fish
consumption rate of 17 gms/day or (at 43%) approximately 7 gms of
sport caught fish. There is poor data on the proportion of the
nonsport caught (commercial) fish consumed within the region
which is a< tually caught within the region. Using the Michigan
survey data, at least 22% of the fish consumed are species from
outside the region. Thereby, the maximum proportion of
regionally caught and consumed fish in Michigan may be estimated
to be only 78% or 13 gms per day. All those contacted familiar
with commercial fishing within the region estimated the major
amount of -egionally caught commercial fish are sold outside of
the region and therefore, generally not available to regional
anglers. If one assumes a conservative mean total of sport
caught and regional commercial caught meals to equal 24 meals per
vear at 8 ounces per meal or up to 48 meals per year at 4 ounces
per meal, the mean dailv consumption rate is 15 gm/day.
For this initiative, the assumption of 15 gm/day of regionally
caught fish should adequately estimate the consumption rate of
the mean angler population and their families for all sport
caught fish. A much larger segment of the sport angler
population is included if this consumption is attributed totally
to species of fish more susceptible to persistent and
bioaccumulative contaminants, i.e., the salmonids. This
67
-------
represents at least the 95% exposure level for regionally caught
fish for the regional population as a whole, i.e , fisherpersons
as well as nonfisherpersons.
6. Relative Source Contribution
For many chemicals, surface water may represent a very small
portion of an individual's exposure potential. This is
particularly true for those chemicals with a high vapor pressure,
e.g. solvents where volatilization would quickly remove them from
the water column. Using toluene as an example, it is estimated
that less than 0.2Z of an individual's exposure comes via surface
water Diet, as well, is not considered to be a significant source
(ATSDR, 1989). Therefore, indoor and ambient air, as well as
occupational exposure may provide the majority of exposure to
volatile chemicals for individuals.
On the other hand, diet, including fish, may provide the most
significant exposure for persistent lipophilic chemicals such as
PCBs, DDT, dioxins, etc. For these chemicals, surface water
exposure through fish consumption may be the most significant
source of contamination. It Is important to recognize these
differences in source contributions when calculating risk or
developing safe exposure levels in order not to significantly
underestimate (or overestimate) the Importance of surface water's
contribution to the total exposure of certain compounds. For
chemicals such as PCS, the FDA Total Diet Study estimates that
contaminated fish represent the most significant exposure. T
68
-------
average adult's dailv intake of PCBs via diet is estimated to be
560 ng, versus estimated inhalation levels of 100 ng per dav.
Based on these estimates, diet contributes approximately 85% of
exposure (ATSDR, 1987). It appears likely that, for other highly
bioaccumulatlve chemicals, a similar esimate may be made as well.
Following the concept proposed by EPA for the National Primary
Drinking Water Regulations (EPA, 1989) where data Indicate
exposure via water is between 80 and 100 percent of the total
exposure to a chemical, an 80% relative source contribution is
used. This concept for bioaccumulatlve materials only is adopted
by the initiative. The 80Z ceiling ensures that exposure should
be low enough to provide adequate protection for individuals whose
total exposure to a contaminant is higher than current data
indicate, i.e., through diet or other exposure such as occupation.
This approach provides for responsible management of those
chemicals where surface water contamination may lead to the most
significant avenue for exposures and may also have a significant
impact on controlling a chemicals total exposure. Where
environmental media other than surface water represents the most
significant exposure, this initiative encourages efforts to
reduce exposure via those media as being most effective in
reducing total exposure.
69
-------
IV. CRITERIA CALCULATIONS
A. Standard Exposure Assumptions
Wh - weight of an average human (Wh - 70kg).
WC = average per capita water consumption for surface waters
classified as public water supplies (WC, « 2 liters/day)
-or-
average per capita incidental daily water exposure for
surface waters used only for recreational activities
(WC - 0.01 liters/day)
FC - average per capita daily consumption of sportcaught fish '
0.015 kg/day
BAF = bioaccumulation factor.
B. Carcinogens
When a linear, nonthreshold dose-response relationship is assumed, the
risk associated dose shall be calculated using the following equation:
RAD - 1 x 0.00001
V
70
-------
Where•
RAD = risk associated dose in milligrams of toxicant per kilogram
body weight per day (mg/kg/day)
0.00001 (1 x 10 ) - incremental risk of contracting cancer equal to
1 in 100,000.
q.* - slope factor for humans in (mg/kg/day)
The human cancer value shall be calculated as follows:
HCV - RAD x rfh
WC + (FC x BAF)
Where:
HCV «• Human Cancer Value in milligrams per liter (mg/L).
RAD • Risk associated dose in milligrams toxicant per kilogram body
weight per day (mg/kg/day) that is associated with a lifetime
incremental cancer risk equal to 1 In 100,000.
3. Noncarclnogens
a. The human noncancer value shall be calculated as follows:
HNV - AIIE x Wh x RSC
WC -f (FC x BAF)
-------
Where:
HNV - Human noncancer value in milligrams per liter (mg/L).
ADE » Acceptable daily exposure in milligrams toxicant per kilogram
body weight per day (mg/kg/day).
RSC - Relative source contribution factor of 0.8 is used to account
for routes of exposure other than consumption of contaminated
water and fish and recreational exposure. Use of the RSC
is limited to bioaccumulative toxicants where the main source
of exposure may be assumed to be fish consumption.
b. An ADE may be derived directly from the following example methods
depending on the type and quality of the toxicity database:
1. a scientifically valid reference dose (RfD) as identified
through best available information sources, such as IRIS; and
2. a scientifically valid acceptable daily intake (ADI) as
identified from the U.S. Food and Drug Administration.
3. a chronic or subchronic NOAEL for humans exposed to the
toxicant via contaminated drinking water as follows:
ADE - NOAEL (mg/1) x wr
d
U x Wh
72
-------
Where.
U « Uncertainty factor of 10-100 depending on the quality
of the data.
A. a chronic or subacute NOAEL from a mammalian test species
exposed to toxicant contaminated drinking water as follows;
ADE « NOAEL (og/1) X Vw
_ Wa
U
Where:
Vw • Volume of water consumed per day by test animals (L/day)
Wa « Vieight of test animal (kg) .
U • Uncertainty factor of 100-1000 depending on quality of
data. An additional uncertainty factor of up to 10 may
be used to account for studies of very short term,
e.g., 14 days.
5 . a chronic or subacute NOAEL from a mammalian test species
exposed to toxicant -contaminated food as follows:
ADE - NOAEL (mg/kg food) x fc
U
-------
Where:
fc - Daily food consumption by test animal (kg).
Wa = Weight of test animal (kg).
U - Uncertainty factor of 100-1000 depending on quality of
data. An additional uncertainty factor of up to 10
may be used to account for studies of very short term,
i.e., 28 days.
6. a chronic or subacute NOAEL from a mammalian test species
exposed to a toxicant by gavage as follows:
ADE = NOAEL (mg/kg) x Fv
U
Where-
Fw - Fraction of week dose.
U - Uncertainty factor of 100-1000 depending on quality of
data. An additional uncertainty factor of up to 10 may
be used to account for studies of very short term,
i.e., 28 days.
74
-------
7. A chronic or subacute NOAEL from a mammalian test species
exposed to a toxicant by inhalation:
= NOAEL (mg/m3) x I x fw x fd x r
U x Wa
Where-
I = Inhalation rate for test species (m /day)
fw = fraction of week exposed.
fd = Fraction of day exposed.
r * Absorption coefficient.
Wa = height of test animal (kg).
U * Uncertainty factor of 100-1000 depending on quality of data.
8. Similar approaches shall be followed when data is limited to a
LOAEL with an appropriate increase in uncertainty factor.
For example, a subacute LOAEL from a mammalian test species
exposed to toxicant contaminated drinking water would be
calculated as follows:
75
-------
ADE » LOAEL (mg/1) x Vw
Wa
U
Where-
Vw «• Volume of water consumed per day by test animal (L/day).
Wa - Weight of the test animal (kg).
U = Uncertainty factor of 1000-10,000 depending on quality of
data and severity of effect.
76
-------
REFERENCES
Agency for Toxic Substances and Disease Registry (ATSDR), 1987,
Toxicology Pro-file for Selected PCBs (Aroclor-1260, -1254, -1248,
-1242, -1232, -1221, and -1016), Syracuse Research Corp., Contract
No. 68-03-3228, Oak Ridge National Laboratory, Oak Ridge, TN.
Agency for Toxic Substances and Disease Registry (ATSDR), 1989,
Toxicological Profile for Toluene, Clement Assoc., Inc., Contract No.
205-88-0608, Centers for Disease Control, Atlanta, GA.
Argyris, T.S., 1985, Regeneration and the Mechanism of Epidermal Tumor
Promotion, CRC Reviews in Toxicology, 14.211-258.
Borzsonyi, M. et s.1., (Ed.), 1984, Models, Mechanisms and Etiology of
Tumor Promotion, International Agency for Research on Cancer, World
Health Organisation, IARC Scientific Publications No. 56, Lyon,
France.
Connelly, N.A., T.L. Brown and B. A. Knuth, 1990, New York Statewide
Angler Survey, 1988, New York State Department of Environmental
Conservation, Albany, NY.
Crouch, A.C. and R. Wilson, 1982, Risk/Benefit Analysis, Ballinger
Publishing Co., Cambridge, MA.
Fiore, B.J. et al., 1989, Sport Fish Consumption and Body Burden Levels
of Chlorinated Hydrocarbons: A Study of Wisconsin Anglers, Archives
of Environmental Health, 44:82-88.
-------
Howe, R.B. and K.S. Crump, 1982, A Computer Program to Extrapolate
Quantal Animal Toxicity Data at Low Doses, Prepared for Office of
Carcinogen Standards, Occupational Safety and Health Administration,
U.S. Department of Labor, Contract 41 USC 252C3, Washington, D.C.
International Agency for Research On Cancer (IARC), 1985, IARC
Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to
Humans, Volume 36, Preamble, World Health Organization, Lyon, France.
International Agencv for Research on Cancer (IARC), 1987, IARC
Monographs on the Evaluation of Carcinogenic Risks to Humans,
Preamble, Final Draft, January, 1987, World Health Organization,
Lyon, France.
Hanson, J.M. et al., 1982, Teratology Test Methods for Laboratory
Animals, In- Principles and Methods of Toxicology, Hayes, A.W. (Ed),
Raven Press, New York, NY.
Mausner, J. and S. Krammer, 1985, Mausner and Bahn Epidemiology; An
Introductory Text, W.B. Saunders Company, Philadelphia, PA.
National Cancer Institute (NCI), 1976, Guidelines for Carcinogen
Bioassay in Small Rodents, Technical Report Series No. 1, U.S.
Department of Health, Education and Welfare, NCI-CG-TR-1.
National Toxicology Program (NTP), 1984, Report of the Ad Hoc Panels on
Chemical Carcinogenesis Testing and Evaluation of the National
Toxicology Program, Board of Scientific Counselors, U.S. Government
Printing Office, Washington, D.C., 1984-421-1324726.
-------
Office of Science and Technology Policy (OSTP), 1985, Chemical
Carcinogens; A Review of the Science and Its Associated Principles,
Federal Register, Vol. 50, No. 50. March 14, 1985, 10371-10442.
Organization for Economic Cooperation and Development (OECD), 1987,
Guidelines for Testing of Chemicals, Paris, France.
Palmer, A.K., 1981, Regulatory Requirements for Reproductive
Toxicology: Theory and Practice, In: Developmental Toxicology,
Kinnnel, C.A. and J. Buelke-Sam (Eds), Raven Press, New York, NY.
Pitot, H., T. Goldworthy, and S. Moran, 1981, The Natural History of
Carcinogenesis: Implications of Experimental Carcinogenesis in the
Genesis of Human Cancer, Journal of Supramolecular Structure and
Cellular Biochemistry, 17:133-146.
Pitot, B.C. and A.E. Sirica, 1980, The Stages of Initiation and
Promotion in Hepatocarcinogenesis, Biochimica et Biophysica Acta,
605:191-215.
Slaga, T., 1984, Can Tumor Promotion be Effectively Inhibited7 In:
Borzsonyi, M. et al. (Eds), 1984, Models, Mechanisms and Etiology of
Tumor Promotion, International Agency for Research on Cancer, World
Health Organization, IARC Scientific Publications No. 56, Lyon,
France.
Thomas, R. (Ed), Safe Drinking Water Committee, 1986, Drinking Water and
Health, National Research Council, National Academy of Sciences,
National Academy Press, Washington, D.C., p. 141, 157.
-------
Tripartite Working Partv, 1985, Criteria for Identifying ana Classifying
Carcinogens, Mutagens and Teratogens, Developed jointly by the Safety
of Chemicals Committee of CEFIC, the International Affairs Group of
CMA/SOCMA, and the Canadian Chemical Producers Association, Obtained
from Chemical Manufacturers Association, Washington, D.C.
Trosko, J., C. Jones and C. Chang, 1983, The Role of Tumor Promoters on
Phenotypic Alterations Affecting Intercellular Communication and
Tumorigenesis, Annals N Y. Academy of Science, 407:316-327.
U.S Department of Health and Human Services, 1979, Registry of Toxic
Effects of Chemicals Substances (RTECS), National Institute for
Occupational Safety and Health, Cincinnati, OH.
U.S. Environmental Protection Agency (EPA), 1979, Identification and
Evaluation of Waterborne Routes of Exposure from Other Than Food and
Drinking Water, Washington, D.C. Office of Water Planning and
Standards. EPA 440/4-79-016.
U.S. Environmental Protection Agency (EPA), 1980, Water Quality Criteria
Availability, Appendix C Guidelines and Methodology Used in the
Preparation of Health Effects Assessment Chapters of the Consent
Decree Water Quality Criteria Documents, Federal Register, Vol. 45,
November 28, 1980, 79347-79357.
U.S. Environmental Protection Agency (EPA), 1985, Toxic Substances
Control Act Test Guidelines; Final Rules, Federal Register, Vol. 50,
No. 188, September 27, 1985, 39421-39425.
80
-------
U.S. Environmental Protection Agencv (EPA), 1986, Guidelines for
Carcinogen Risi Assessment, Federal Register, Vol. 51, No. 185,
September 24, 1986, 33992-34002.
U.S. Environmental Protection Agency (EPA), 1986a, Guidelines for the
Health Assessment of Suspect Developmental Toxicants, Federal
Register, Vol. 51, No. 185, September 24, 1986, 34028-34040.
U.S. Environmental Protection Agency (EPA), 1986b, Guidelines for
Mutageniclty Assessment, Federal Register, Vol. 51, No. 185,
September 24, 1986, 34006-34012.
U.S. Environmental Protection Agency (EPA), 1989a, Exposure Factors
Handbook. Washington, D.C. Office of Health and Environmental
Assessment, Exposure Assessment Group. EPA/600/8-89/043.
U.S. Environmental Protection Agency (EPA), 1989b, National Primary and
Secondary Drinking Water Regulations, Proposed Rule, Federal Register
Vol. 54, No. 97, May 22, 1989, p. 22069.
U.S. Environmental Protection Agency (EPA), 1989c, Proposed Amendments
to Agency Guidelines for Health Assessment of Suspect Developmental
Toxicants, Federal Register, Vol. 54, No. 9386, March 6, 1989.
U.S. Environmental Protection Agency (EPA), 1989d, Risk Assessment
Guidance for Superfund, Volume 1, Human Health Evaluation Manual
(Part A) - Interim Final, Office of Emergency and Remedial Response,
Washington, D.C., EPA/540/1-89/002.
81
-------
Wells, Philip, 1990, Michigan Department of Natural Resources (MDNR),
Recreation Division, Personal Communication to Gary Hurlburt, Surface
Water Quality Division, MDNR.
West, P.C. et al., 1989, Michigan Sport Anglers Fish Consumption
Survev A Report to the Michigan Toxic Substance Control Committee,
University of Michigan Natural Resource Sociology Research Lab.
Technical Report #1, Ann Arbor, Michigan, MDMB Contract #87-20141.
Williams, G and J. Weisburger, 1986, Chemical Carcinogens, In*
Klaassen, C , M Amdur and J Derell (Eds), 1986, Casarett and
Doull's Toxicology; The Basic Science of Poisons, Third Edition,
MacMillan Publishing Co., New York, NY.
82
-------
APPENDIX A
UNCERTAINTY FACTORS
A. INTRODUCTION
Uncertainty factors (also called safety factors) are intended for
use in extrapo Lating toxic responses thought to have a threshold
(i.e., noncarcLnogenic effects). "Uncertainty factor" is defined as
a number that reflects the degree or amount of uncertainty that must
be considered when experimental data in animals are extrapolated to
man (EPA, 1980). In addition, uncertainty factors are used when
extrapolating from small populations of humans to the entire
heterogeneous human population and when extrapolating from a single
animal species to wildlife communities. The use of uncertainty
factors in extrapolating animal toxicity data to acceptable exposure
levels for humans has been the cornerstone of regulatory toxicology
(National Academy of Sciences, 1980). This appendix will provide
the risk assessor with additional guidelines, rationale and
information concerning the selection of uncertainty factors.
Because of the high degree of judgment involved in the selection of
uncertainty factors, the risk assessment justification should
include a detailed discussion of the selection of the uncertainty
factors along with the data to which they are applied.
This report is organized with the recommended uncertainty factors
listed in Part B for quick reference, and a discussion of those
factors and their support in Part C. Also included in Part C is a f-fi ft
83
-------
discussion of the exposure duration terms "subacute", "subchronic",
and "chronic"
B. RECOMMENDED UNCERTAINTY FACTORS
1 . A 10-fold factor is recommended when extrapolating from valid
experimental results from human studies of prolonged exposure.
2. A 100-fold factor is recommended when extrapolating from valid
results of long-term studies on experimental animals with results
of studies of human exposure not available or scantv (e.g., acute
exposure only)
3. A factor of up to 1000 is recommended when extrapolating from
animal studies for which the exposure duration is less than
chronic (i.e., less than 50% of the lifespan) or when other
significant deficiencies in study quality are present, with no
useful long-term or acute human data.
4. An additional uncertainty factor or between 1 and 10 is recommended
depending on the severity and sensitivity of the adverse effect
when extrapolating from a LOAEL rather than a NOAEL.
5. An additional uncertainty factor of up to 10 may be applied when
there are limited or incomplete subacute or chronic toxicity
data, such as with short-term repeated dose animal studies where
the exposure regime involves a limited period that is markedly
short-term relative to the lifespan of the test species (e.g.,
28-day rodent NOAEL) .
-------
C. DISCUSSION
Dourson and Stara (1983) reviewed available literature on
uncertainty faccors which are used to estimate acceptable daily
intakes (ADIs) for toxicants. They found that the use and choice of
these factors is supported by reasonable Qualitative biological
premises and specific biological data. Therefore, in the absence of
adequate chemical-specific data, uncertainty factors for criteria
derivation may be selected according to reasonable assumptions and
approximations rather than total arbitrariness. They presented a
set of guidelines for the use of uncertainty factors based on those
utilized by the FDA, WHO, NAS, and EPA, indicating consistency and
widespread acceptance among the scientific community. Those
guidelines have been adapted herein for use in risk assessment under
the Great Lakes Initiative. Their rationale and experimental
support are discussed below. The guidelines should not be
misconstrued as, being unalterable and inflexible. They are intended
to help ensure appropriateness and consistency of risk assessments.
They should be regarded as general recommendations, with the
realization that the data for a particular chemical may be such that
a different uncertainty factor would be more appropriate.
1. A 10-fold factor is recommended when extrapolating from valid
experimental results from human studies of prolonged exposure.
People of all ages, states of health, and genetic
predispositions may be exposed to environmental contaminants.
The 10-fold factor is intended to offer protection for the
85
-------
sensitive subpopulations (the very young, the aged, medically
indigent, genetically predisposed, etc.)» since the observed
no-effect level is generally based on average healthy
individuals. Experimental support for this 10-fold factor is
provided by log-probit analysis and the study of composite human
sensitivity (Dourson and Stara, 1983).
However, Calabrese (1985) has presented data on human
variability in several physiological parameters and in
susceptibility to several diseases, and concluded that human
variation may range up to two or three orders of magnitude.
While human variation in the metabolism of various xenobiotics
may have a 1000-fold range, Calabrese (1985) noted that the vast
majority of the responses addressed fell clearly within a factor
of 10. Another studv on key human pharmacokinetic parameters
indicates that the 10-fold factor to encompass human variability
may only capture the variability among normal healthy adult
humans. That report recommends further study to determine the
degree of additional susceptibility among sensitive
subpopulations (EPA, 1986).
Given the heterogeneous and highly outbred state of the human
population, and the multifactorial nature of disease
susceptibility, reliance on the adequacy of the 10-fold factor
for extrapolation to "safe" levels appears somewhat precarious.
But because of its history of use and current widespread
acceptance, this factor may continue to be used until the
availability of new data indicating quantitatively a more
acceptable factor.
-------
2. A100-fold factor is recommended when extrapolating from valid
results of long-term studies on experimental animals with
results of stud Lesof human exposure not available or scanty
(e.g., acute exposure only).
This represents the 10-fold factor for intraspecies extrapolation
(see C.I) and an additional 10-fold uncertainty factor for
extrapolating data from the average animal to the average man.
The 100-fold uncertainty factor has been justified for use with the
risk extrapolation for food additives. That justification has been
based on differences in body size, differences in food requirements
varying with age, sex, muscular expenditure, and environmental
conditions within a species, differences in water balance of
exchange between the body and its environment among species, and
differences among species in susceptibility to the toxic effect of a
given contaminant (Blgwood, 1973). The use of the 100-fold
uncertainty factor has also been substantiated by citing differences
in susceptibility between animals and humans to toxicants,
variations in sensitivities in the human population, the fact that
the number of animals tested is small compared with the size of the
human population that may be exposed, the difficulty in estimating
human intake, and the possibility of synergistic action among
chemicals (Vettorazzi, 1976).
On a dose per unit of body weight basis, large animals (e.g., man)
are generally more sensitive to toxic effects than small animals
(e.g., rats, mice). This principle is attributed to the
relationship between animal size and pharmacokinetics, whereby the £f f •-?
87
-------
tissues of a large animal are exposed to a substance (mg/kg dose)
for a much longer time than the tissues of a small animal. This
principle has been demonstrated experimentally. The pharmacokinetic
processes underlying this phenomenon include: in general, large
animals metabolize compounds more slowly than do small animals;
large animals have many more susceptible cells; in large animals,
substances are distributed more slowly and tend to persist longer;
the blood volume circulates much more rapidly in small animals.
Thus, for the same mg/kg dose, human tissues are exposed to a
substance for a much longer time than rodent tissues (National
Academy of Sciences, 1977}
Experimental support for the additional 10-fold uncertainty factor
when extrapolating from animal data to humans is provided by studies
on body-surface area dose equivalence and toxicity comparisons
between humans and different animal species (Dourson and Stara,
1983). On a dose per unit of body-surface area basis, the effects
seen in man are generally in the same range as those seen in
experimental animals. An interspecies adjustment factor accounts
for differences in mg per kg body weight doses due to different
body-surface areas between experimental animals and man. The factor
may be calculated by dividing the average weight of a human (70 kg)
by the weight of the test species (in kg) and taking the cube root
of this value. Thus on a body weight basis, man is assumed to be
more sensitive than the experimental animals by factors of
approximately 5 and 13 for rats and mice, respectively. For most
experimental animal species (I.e., all species larger than mice),
the 10-fold decrease in dose therefore appears to Incorporate a LCt ^
88
-------
margin of safety. For mice, the interspecies adjustment factor
suggests that the additional 10-fold uncertainty factor for
interspecies extrapolation to humans is not large enough (Dourson
and Stara, 1983). Nevertheless, the additional 10-fold factor is
considered adequate to adjust from mice to humans when
chemical-specif \c data are not available.
3. A factor of up co 1000 is recommended when extrapolating from animal
studies for which the exposure duration is less than chronic
(i.e.. less than 50% of the lifespan) or when other significant
deficiencies in study quality are present, with no useful long-term
or acute human data.
This represents the 10-fold factors for intraspecies and
interspecies extrapolation (see C.2), and an additional uncertainty
factor of up to 10-fold for extrapolating from less than chronic to
chronic animal exposures (or when the data are significantly flaved
in some other way) . Injury from chronic exposure may occur in at
least three ways: by accumulation of the chemical to a critical
concentration at sites of action sufficient to induce detectable
injury; by accumulation of injury until physiological reserves can
no longer compensate (i.e., repair is never complete); or after a
long, latent period beginning with an exposure that has an
unrecognized biological effect and precipitates the eventual
appearance of Injury (National Academy of Sciences, 1977).
Obviously, sui ficient duration of exposure is necessary in order for
the effects seen in chronic toxicity to become manifest. Subchronic
toxicology studies may not offer reliable means for assessment of
89
-------
long-term toxic effects in animals, let along extrapolation to
chronic effects in man (National Academy of Sciences, 1977).
Hovever, it is often the case that a good quality, chronic exposure
study for a particular chemical is unavailable. The intention of
this additional uncertainty factor is to enable the use of subchronic
or flawed studies to protect against the risk of adverse effects which
might only appear with chronic dosing.
Experimental support for the additional uncertainty factor is given
bv literature reviews which compare subchronic NOAELs and chronic
NOAELs for many compounds (McNamara, 1971; Weil and McCollister,
1963). The studies reviewed by those investigators employed a
variety of rodent and non-rodent species. The duration of the
subchronic exposures was usually 90 days, but ranged from 30 to
210 davs. Wide variations in endpoints and criteria for adverse
effects were encountered in these literature reviews. However,
their findings do give a rough indication of the general subchronic
and chronic NOAELs for other than carcinogenic or reproductive
effects. For over SOZ of the compounds tested, the chronic NOAEL was
less than the 90-day NOAEL by a factor of 2 or less. There was some
indication that chronic dosing may result in the development of
tolerance toward certain chemicals, as the chronic NOAEL was larger
than the 90-day NOAEL in a few cases. However, it was also found
that the chronic NOAEL may be less than the 90-day NOAEL by a factor
of 10 or more. The latter situation appeared to be uncommon.
Therefore, these reviews report that the additional 10-fold
uncertainty factor appears to be adequate or incorporate a margin of
safety in the majority of cases. / /S
90
-------
As the literature reviews bv McNamara (1971) and Weil and
McCollister (1963) are limited and the studies reviewed utilized a
variety of toxicologic endpoints with questionable sensitivities,
one must be cautious in interpreting their conclusions. But for
lack of data to the contrary, it appears that application of the
additional 10-fold uncertainty factor is appropriate and justified
when extrapolating a NOAEL from a 90-day study to a chronic NOAEL
estimate. This practice may underestimate the true chronic NOAEL
far more often than overestimating it, thus adding a margin of safety
to the risk calculations.
One remaining question regarding exposure duration is* At what
point is the duration considered adequate, such that the additional
uncertainty factor of up to 10 is unnecessary9 In other words, how
is "chronic" defined for the sake of this guideline?
At this point, further discussion of the terms "chronic"
"subchronic", and "subacute", is necessary. The term "subacute" has
been used to describe a duration less than subchronic, while it has
also been used as a term analogous to subchronic. EPA (1980)
describes "subacute" exposures (in this case, analogously to
"subchronic") as often exceeding 102 of the lifespan, e.g., 90 days
for the rat with an average lifespan of 30 months. However, as
pointed out by the Organization for Economic Cooperation and
Development (OECD, 1981), the term "subacute" is semantically
incorrect. The OECD prefers to use the phrase "short-term repeated
dose studies", referring to 14, 21 and 28 day studies, to
distinguish from "subchronic" studies of greater duration.
91
-------
"Subchronic" is generally defined as part of the lifespan of the
test species, although opinions differ on the precise definition.
Klaassen (1986) defines "subacute" as repeated exposure to a
chemical for one month or less, and "subchronic" as repeated
exposure for 1-3 months. Chan et al. (1982) describe "subchronic"
exposure durations as generally ranging from 1 to 3 months in
rodents and one year in longer-lived animals (dogs, monkeys), or for
part (not exceeding 10%) of the lifespan. Stevens and Gallo (1982)
define "long-term toxicity tests" (encompassing subchronic and
chronic toxicity studies) as studies of longer than 3 months
duration, i.e., greater than 10% of the lifespan in the laboratory
rat. EPA (1985) describes "subchronic" toxicity testing as
involving continuous or repeated exposure for a period of 90 days,
or approximately 10% of the lifespan for rats.
The various definitions offered for "chronic" are generally
inconsistent. Klaassen (1986) defines "chronic" as repeated
exposure for more than 3 months. According to the National Academy
of Sciences (1977), chronic exposure in animals is generally
considered to be at least half the life span. In estimating chronic
SNARLs, the National Academy of Sciences (1980) in most cases
utilized data from studies lasting a "major portion of the lifetime
of the experimental animal". According to the EPA's Health Effects
Testing Guidelines (EPA, 1985), chronic toxicity tests should involve
dosing over a period of at least 12 months. The application of
their guidelines, they add, should generate data on which to
identify the majority of chronic effects and shall serve to define
long-term dose-response relationships. The OECD (1981) states that
92
-------
the division between subchronic and chronic dosing regimes is
sometimes taken as 10Z of the test animal's life span. They also
state that the duration of the exposure period for chronic toxicity
studies should be at least 12 months. They describe "chronic" as
prolonged and repeated exposure capable of identifying the majority
of chronic effects and to determine dose-response relationships.
Others have insestigated the delayed appearance of toxic effects
which might be missed under shorter dosing regimes. Frederick
(1986) conducted a pilot survey of new drug evaluators for
incidences of delayed (greater than 12 month) drug-induced
pathology It was concluded that new toxic effects "not
infrequently" arise after one year of dosing in rodents. It was
further stated that those findings formed the basis for the
conclusion of the Bureau of Human Prescription Drugs* the duration
of the long-term toxicity tests of drugs that are likely to be used
in man for moie than a few davs should be at least 18 months.
Glocklin (1986) reviewed the issues regarding testing requirements
for new drugs, and concluded that 12 month chronic toxicity studies
seemed to be .in appropriate requirement for characterization of the
dose-response.
It is evident that there are discrepancies in the qualitative and
quantitative characterization of "chronic" animals studies. An
appropriate and reasonable working definition for "chronic" would
appear to be at least half the life span (therefore, at least 52
weeks for rats and at least 45 weeks for mice). Qualitatively,
"chronic" means that the exposure duration was sufficient to
-------
represent a full lifetime exposure, in terms of dose-response
relationships. For example, a study providing an experimental NOAEL
which approximates a lifetime NOAEL is considered a chronic study.
It is recognized that the above quantitative definition (at least
half the life span) does not demonstrate the flexibility inherent in
the above qualitative description. That flexibility reflects the
vast differences in the toxicology of various chemicals:
demonstration of a lifetime NOAEL for some chemicals may require
dosing for half the life span, while the toxicology of most
chemicals may allow demonstration of a lifetime NOAEL under a much
shorter dosing regime. It may be argued that the lifetime NOAEL for
noncarcinogenic effects of many chemicals can be demonstrated in
rodent studies of much less than one year. While the
previouslv-discussed works of McNamara (1971) and Weil and
McCollister (1963) support that view, they also demonstrate that the
chronic NOAEL may be less than the 90-day NOAEL by a factor of 10 or
more, for some chemicals.
This discussion is necessary in order to properly interpret the
uncertainty factor guideline, which recommends that the additional
uncertainty factor of up to 10 be applied when the exposure duration
is less than "chronic". The intent of the uncertainty factor is to
adjust the experimental NOAEL to a lifetime NOAEL in those cases
where the lifetime NOAEL was presumably not adequately demonstrated.
The key issues are summarized in the following points and
recommendations:
-------
a. An acceptable quantitative definition of "chronic" is elusive.
Due to differing toxicological properties, the necessary minimum
exposure duration to demonstrate a lifetime NOAEL differs widely
among chemicals. A qualitative, philosophical definition of
chronic is. "Chronic" is when the exposure duration is
sufficient for the identification of the majority of long-term
effects and their dose-response relationships. Therefore, a
"chronic" studv reporting a NOAEL is one which can be reasonably
presumed to predict the lifetime NOAEL.
b. The use of scientific judgment is predominant in the decision of
when chronic exposure conditions exist, and hence, when the
additional uncertainty factor is no longer appropriate.
c. That scientific judgment should be guided by a review of all
available pertinent data, e.g., metabolism, pharmacoklnetics,
bioaccumulation, mechanism of action, target organ
characteristics, potential for latent effects, etc.
d. Available reviews of rodent studies indicate that, for many
chemicals, studies of much less than one year duration can
provide reasonable estimates of lifetime NOAELs. However, it is
also recognized that the toxicological characteristics of some
chemicals, will prevent the qualitative and quantitative
demonstration of latent adverse effects and a lifetime NOAEL if
the duration is less than one year. If the lack of additional
data prevents scientific judgment in these cases, 50% of the
lifespan (52 weeks for rats; 45 weeks for mice) may be
95
-------
considered the minimum necessary duration for a "chronic"
exposure. Application of the additional uncertainty factor for
these apparently "subchronic" studies may later provide to be
excessively conservative in some cases. But, if the toxicologic
database is inadequate, the additional uncertainty factor should
be included, both as a matter of prudent public policy and as an
incentive to others to generate the appropriate data.
e. Ordinarily, the additional 10-fold factor mav be applied for all
rodent studies of 90 days duration, unless there is
chemical-specific data indicating that would be unnecessary and
overly conservative.
f. For rodent studies of between 90 days and 12 months duration,
the use of the additional 10-fold uncertainty factor is best
determined by professional judgment. As described above, if
data are not available to sufficiently guide professional
judgment, then such studies may be subject to part or all of the
additional 10-fold factor. A "sliding scale" or between 1 and
10 is a reasonable means of selecting a lesser factor when 10
appears excessive. Under this concept, the additional
uncertainty factor applied may vary on a scale of one to ten
according to how closely the dosing duration approached 50Z of
the lifespan. Of course, consideration must be given of the
study quality and the other pertinent data mentioned in 3.c
above. A 90-day rodent study would be subject to a 10-fold
additional factor, if study quality Is otherwise nominal and
other chemical-specific data are lacking. A nominal-quality tf, -> /
'
96 '
-------
study, with exposure over 50Z of the lifespan, would be subject
to a "1", i.e., no additional adjustment. Situations where the
exposure duration is between 90 days and 50% of the lifespan,
and/or study quality is flawed, must be handled on a
case-by-case basis. This "sliding scale" concept may offer
guidance t3 the scientific judgment that will be necessary.
Dosing duration is but one parameter upon which to assess the
adequacy of a study Other deficiencies in the study design may
cause increased concern about the validity of the reported NOAEL or
LOAEL. Therefore, risk assessors may utilize part or all of this
additional 10-fold uncertainty factor to compensate for data which
appears less-than-adequate. Factors which may affect the degree of
confidence in the data include the number of animals per dose group,
the sensitivity and appropriateness of the endpoints, the quality of
the control group, the exposure route, the dosing schedule, the age
and sex of the exposed animals, and the appropriateness of the
surrogate species tested, among others. EPA's Health Effects Testing
Guidelines (EFA, 1985) provide specific information on the desirable
qualities of subchronic and chronic toxicity tests.
4. An additional uncertainty factor of between 1 and 10 is recommended
depending on the severity and sensitivity of the adverse effect
when extrapolating from a LOAEL rather than a NOAEL. This uncertainty
factor reduces the LOAEL into the range of a NOAEL, according to
comparisons of LOAELs and NOAELs for specific chemicals. There is
evidence available which Indicates, for a select set of chemicals,
97
-------
96% have LOAEL/NOAEL ratios of 5 or less, and that all are 10 or
less (Dourson and Stara, 1983) In practice the value for this
variable uncertainty factor has been chosen by the U.S. EPA from
values among 1 through 10 based on the severity and sensitivity of
the adverse effect of the LOAEL. For example, if the LOAEL
represents liver cell necrosis, a higher value is suggested for
this uncertainty factor (perhaps 10). If the LOAEL is fatty
infiltration of the liver (less severe than liver cell necrosis),
then a lower value is suggested (perhaps 3; see the following
discussion). The hypothesized NOAEL should be closer to the LOAEL
showing less severe effects (Dourson and Stara, 1983).
In some cases the data do not completely fulfill the conditions for
one category of the above guidelines, and appear to be intermediate
between two categories. Although one order of magnitude is
generally the smallest unit of accuracy that is reasonable for
uncertainty factors, an intermediate value may be used if felt
necessary (Dourson, 1987). According to EPA (1980), such an
intermediate uncertainty factor may be developed based on a
logarithmic scale rather than a linear scale. Calculating the mean
logarithmically may be the more appropriate option, because the
precision of all uncertainty factor estimates is poor, and a
logarithmic scale is the best way to estimate the mean of two
imprecise estimates (Dourson, 1987). Halfway between 1 and 10 is
approximately 3.16 on a logarithmic scale. However, so as not to
imply excessive accuracy in the estimate, that mean value should be
rounded-off to 3 (Dourson, 1987).
98
-------
5. An additional uncertainty factor of up to 10 may be applied when
there are limited or incomplete subacute or chronic toxicitv data,
such as with short-tenn repeated dose animal studies where the
exposure regime involves a limited period that is markedly short-term
relative to the lifespan of the test species (e.g., 28-day rodent NOAEL
As previously noted (see C 3) the OECD (1981) distinguishes between
14, 21 or 28 day studies and "subchronic" studies of greater
duration, by referring to the former as "short-term repeated dose
studies11. The short-term studies are commonly conducted by the NT?
to enable appropriate dose selection in subchronic studies (NCI,
1976). When a limited database exists, short-term animal studies of
28 days or longer may be of sufficient quality to support risk
assessment of potential chronic exposure. Because the duration of
exposure is suostantially less than the 90-day period discussed
under C.3, the risk assessment may require an additional uncertainty
factor in conjunction with the 1000-fold factor recommended under
C 3. As guidance, an additional factor of up to 10 is recommended
when extrapolating from a short-term NOAEL (e.g., 28 days) to
subchronic duration (e.g., 90 days).
Although the extrapolation from oral LD5Qs to chronic oral NOAELs
has been reported by several investigators (Venman and Flaga, 1985;
Layton et al., 1987; McNamara, 1971), there has been relatively
little investigation of the extrapolation from short-term NOAELs
(much less than 90 days in rodents) to chronic NOAELs. EPA (1989)
states that when experimental data are available only for shorter
durations thar desired for subchronic RfD derivation an additional
uncertainty factor is applied. However, further details on the
99
-------
selection of an adequate and appropriate uncertainty factor for those
"shorter durations" are not provided. Weil et al. (1969) evaluated
the relationship between 7-day, 90-day and 2-year minimum effect
levels (MiE) for 20 materials via feed exposure. They found that the
median value for a 90-day MiE was obtained by dividing the 7-day MiE
by a factor of 3 The 95th percentile for the 90-day MiE was
obtainea by dividing the 7-day MiE by 6.2. Also noteworthy is the
finding that the 95th percentile for the 2-year MiE was obtained by
dividing the 7-dav MiE by a factor of 35.3.
These data, albeit limited, support the general principle that as
exposure duration decreases, the ability of the data to demonstrate
chronic dose-response relationships also decreases. While an
additional 10-fold uncertainty factor may reasonably and
appropriately convert a 90-day NOAEL to a surrogate chronic NOAEL,
an additional uncertainty factor may be necessary when extrapolating
from short-term exposures. Applying an additional uncertainty
factor of up to 10 will help ensure that the risk assessment for
potential chronic exposures is adequately conservative, i.e., the
true chronic NOAEL will generally not be overestimated.
100
-------
REFERENCES:
Bigwood, E. 1973. The Acceptable Daily Intake of Food Additivies. CRC
Crit. Rev. Toxicol. June 41-93. As cited in: Dourson, M. and J.
Stara. 1983. Regulatory History and Experimental Support of
Uncertainty (Safety) Factors. Regulatory Toxicology and Pharmacology.
3:224-238.
Calabrese, E. 1985. Uncertainty Factors and Interindividual
Variation. Regulatory Toxicology and Pharmacology. 5-190-196.
Chan, P., O'Hara, G and A. Hayes. 1982. Principles and Methods for
Acute and Subchronic Toxicity In: Hayes, A. (Ed.) 1982.
Principles and Methods of Toxicology. 1982. Raven Press. New York,
NY.
Dourson, M. 1987. U.S. EPA. Environmental Criteria and Assessment
Office. Personal communication with Robert Sills, Michigan
Department of Natural Resources.
Dourson, M. and J. Stara. 1983. Regulatory Hlstorv and Experimental
Support of Uncertainty (Safety) Factors. Regulatory Toxicology and
Pharmacology. 3:224-238.
Frederick, G. 1986. The Evidence Supporting 18 Month Animal Studies.
In: Walker, S. and A. Dayan. 1986. Long-Term Animal Studies; Their
Predictive Value for Man: Proceedings of the Centre for Medicines
Research Workshop held at the Ciba Foundation, London, 2 October
1984.
-------
Glocklin, V. 1986. Justification for 12 Month Animal Studies. In-
Walker, S. and A. Dayan. 1986. Long-Term Animal Studies; Their
Predictive Value for Man: Proceedings of the Centre for Medicines
Research Workshop held at the Ciba Foundation, London, 2 October
1984.
Klaassen, C. 1986. Principles of Toxicology. In Doull, J., Klaassen,
C. and M. Amdur (Eds.). 1986. Casarett and Doull's Toxicology - The
Basin Science of Poisons. 3rd Edition. Macmillan Publishing Co.,
Inc. New York, NY.
Layton, D W., et al. 1987. Deriving allowable daily intakes for
systemic toxicants lacking chronic toxicity data. Regulatory
Toxicology and Pharmacology. 7 96-112.
McNamara, B., 1971. Concepts in Health Evaluation of Commercial and
Industrial Chemicals. In: Mehlman, M., Shapiro, R. and H. Blumenthal
(Eds.). 1971. Advances in Modern Toxicology. Volume 1, Part 1:
New Concepts in Safetv Evaluation. Chapter 4. John Wiley and Sons.
New York, NY.
National Academy of Sciences. 1980. Drinking Water and Health.
Vol. 2. National Academy Press. Washington, D.C.
National Academy of Sciences. 1977. Drinking Water and Health.
Vol. 1. National Academy Press. Washington, D.C.
102
-------
National Cancer Institute. 1976. Guidelines for Carcinogen Bioassay in
Small Rodents. U.S. DREW. Technical Report Series No. 1
Organization for Economic Cooperation and Development (OECD). 1981.
OECD Guidelines for Testing of Chemicals. Section 4 Health
Effects. Paris, France.
Stevens, K. and M. Gallo 1982. Practical Considerations in the
Conduct of Chronic Toxicity Studies. In- Hayes, A. (Ed.) 1982.
Principles and Methods of Toxicology. 1982. Raven Press. New York,
NY.
U.S Environmental Protection Agencv (EPA). 1986. Human Variability in
Susceptibility to Toxic Chemicals - I. Noncarcinogens.
EPA/600/8-86/033.
U.S. Environmental Protection Agency (EPA). 1985. Health Effects
Testing Guidelines. 40 CFR Part 798. Federal Register, v. 50,
n. 188, September 27, 1985.
U.S. Environmental Protection Agency (EPA). 1980. Water Quality
Criteria Documents: Availability. Appendix C - Guidelines and
Methodology U&ed in the Preparation of Health Effect Assessment
Chapters of the Consent Decree Water Criteria Documents. Federal
Register, v. 45, n. 231, November 38, 1980. 79347-79357.
103
-------
Venman, B. and C. Flaga. 1985. Development of an Acceptable Factor to
Estimate Chronic End Points from Acute Toxicitv Data. Toxicology and
Industrial Health. 1(4). 261-269
Vettorazzi, G. 1976. Safety Factors and their Application in the
Toxicological Evaluation. In' The Evaluation of Toxicological Data
for the Protection of Public health. Pennagon, Oxford. As cited in:
Douson, M. and J. Stara. 1983. Regulatory History and Experimental
Support of Uncertainty (Safetv) Factors. Regulatory Toxicology and
Pharmacology. 3- 224-238.
Weil, C. and D. McCollister. 1963. Relationship Between Short- and
Long-Term Feeding Studies in Designing an Effective Toxicitv Test.
Agricultural and Food Chemistry. 11(6): 486-491.
104
-------
DRAFT
September 6, 1991
Great Lakes Initiative
Human Health Criteria for
p,pf-Dichlorodiphenyltrichloroethane (DDT)
CAS No. 50-29-3
1. Tier 1 Human Noncancer Criterion
A review of the available literature indicates that the most
appropriate basis for HNC derivation for DDT is the NOAEL from the
subchronic rat feeding study of Laug et al. (1950). Weanling rats
(15/sex/group) were fed commercial-grade DDT (81Z p,p'-DDT,
19* o.p'-DDT) At levels of 0, 1, 5, 10 or 50 ppm for 15-27 weeks.
The critical toxic effect was liver toxicity, demonstrated as
relatively mild dose-dependent histopathologic changes in
hepatocytes at doses of 5 ppm and higher. These included
hepatocellular hypertrophy, increased cytoplasmic oxyphilia, and
peripheral basophilic cytoplasmic granules. The NOEL was 1 ppm, or
0.05 mg/kg bw/day assuming a food consumption rate of 5Z body weight
per day. The LOAEL was 5 ppm (0.25 mg/kg bw/day).
The database is judged to be sufficient for Tier 1 HNC derivation.
The key study (Laug et al., 1950) provides a subchronic (greater
than 90 day) NOEL which is supported and supplemented by other data.
In a 2-year rat dietary exposure study (Fitzhugh, 1948) rats were
exposed to 10-800 ppm DDT in feed, resulting in liver lesions at all
dose levels with a LOAEL of 10 ppm (0.5 mg/kg bw/day). The
-------
available mammalian reproduction and developmental studies of DDT
indicate that an HNC derived from the critical effect of liver
toxicity will be protective of potential human reproductive/
developmental effects (EPA, 1985). The HNC is based on the
subchronic rat NOEL of 0.05 mg/kg bw/day, with a total uncertainty
factor of 100. An uncertainty factor for subchronic to chronic
conversion is not Included because of the corroborating chronic
study in the database. This approach is consistent with the oral
RfD development by EPA (1985).
ADE - 0.05 mg/kg/d - 0.0005 mg/kg/d - 0.5 ug/kg/d
100
Where: Uncertainty Factor « 100, composed of:
lOx for intraspecies variability
lOx for interspecies extrapolation
Drinking Water Sources*
HNV - ADE x Wh x RSC - 0.5 ug/kg/d x 70 kg x 0.8
WC + CFC x BAF) - 2 1/d + (0.015 kg/d x 2,296,650 I/kg*)
- 0.00081 ug/1 (rounded off to 0.8 ng/1 (Tier 1))
-------
Nondrinking Water Sources'
HNV - APE x Vh x RSC - 0.5 ug/kg/d x 70 kg x 0.8
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 2,296,650 I/kg*)
« 0.00081 ug/1 (rounded off to 0.8 ng/1 (Tier 1))
Where: RSC - 0.8; the substance Is persistent and bioaccumulative
*BAF - 2,296,650, provided by EPA-Duluth and Minnesota PCA.
References:
Fitzhugh, 0. 1948. Use of DDT insecticides on food products.
Industrial and Engineering Chemistry. 40(4):704-705.
Laug, E., A. Nelson, 0. Fitzhugh and F. Kunze. 1950. Liver cell
alteration and DDT storage in the fat of the rat induced by dietary
levels of 1-50 ppm DDT. J. Pharmacol. Exp. Therap. 98:268-273.
U.S. Environmental Protection Agency (EPA). 1985. Integrated Risk
Information System (IRIS). Chemical file for DDT (50-29-3).
Verification Date 12/18/85. Last Revised 9/30/87.
2. Tier 1 Human Cancer Criterion
A review of the available literature for DDT carcinogenicity
reveals a lack of adequate epidemiological data and an extensive
-------
database of chronic oral rodent bioassays. These studies indicate
that the induction of liver tumors is the most consistent and
significant tumorigenic response to DDT in rodents. EPA (1987) has
classified the weight of evidence of DDT carcinogeniclty as B2
based on multiple positive studies in two species (mice and rats),
with ancillary evidence including promoting activity, genotoxicity,
and structural relation to other rodent liver carcinogens.
Therefore, the data are sufficient for Tier 1 HCC derivation.
The animal bioassay providing the highest slope factor estimation
is the multigeneration mouse feeding study of Tarjan and Kemeny
(1969). The predominant tumor types were leukemias and lung
tumors; a significant liver response was not seen. EPA (1980)
derived ambient water quality criteria from the slope factor of
8.422 (mg/kg/day)~ from this study.
EPA (1986a) evaluated the carcinogenicity of DDT and other related
compounds and determined that the Tarjan and Kemeny (1969) study
was not the most appropriate basis for quantitative risk
assessment. The study's findings were not consistent with the
numerous other positive bioassays In terms of the organ site
(lung/leukemia versus liver) and the slope factor (about an order
of magnitude greater). This slope factor was judged to be a
statistical outlier in relation to the liver tumor induction data
from six key studies, and the quality and validity of the study was
also questionable. EPA (1986a) derived a slope factor from the
consistent finding of liver tumor induction In rats and mice, for
which the six key studies provided slope factors within a 13-fold
-------
range. The recommended slope factor of 3.4 E-l (mg/kg/day)"
was derived as the geometric mean of ten slope factors from those
six studies (Turusov et al., 1973; Terracini et al., 1973; Thorpe
and Walker, 1973; Tomatls and Turusov, 1975; Cabral et al., 1982;
Rossi et al., 1977). The averaging procedure was followed because
no further database refinement or rejection could be logically made,
and the geometric average of the values was viewed as the best
rational estimate of the slope factor (EPA, 1986a). The EPA's CRAVE
workgroup has reviewed and accepted this approach to slope factor
estimation as a method to include all relevant data (EPA, 1987) .
This averaging approach to slope factor estimation utilizing
multiple studies, species, strains and sexes has not generally
been recommended in earlier EPA guidelines (EPA, 1980; I986b).
However, more recently, EPA (1989) has stated: "Occasionally, in
situations where no single study is judged most appropriate, yet
several studies collectively support the estimate, the geometric
mean of estimates from all studies may be adopted as the slope.
This practice Insures the inclusion of all relevant data" (EPA,
1989). In the specific case of DDT, the averaging process as
applied to the best available studies may be the most reasonable
means of quantitatively characterizing the carcinogenicity of DDT
(Schoeny, 1991; Holder, 1991; Bayard, 1991).
The Tier 1 Human Cancer Criteria for DDT are derived from the slope
factor of 3.4 E-l (mg/kg/d)~ based on rodent liver tumor induction
in the six key studies.
-------
RAD - 1 x Id"5 - 2.94 x 10~5 mg/kg/d
3.A x 10"1 (mg/kg/d)"1
29.4 ng/kg/d
Drinking Water Sources*
HCV - RAD x Wh - 29.4 ng/kg/d x 70 kg
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 2,296,650 I/kg*)
- 0.0597 ng/1 (rounded off to 0.06 ng/1 (Tier 1))
Nondrinking Water Sources:
HCV - RAD x Wh - 29.4 ng/kg/d x 70 kg
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 2,296,650 I/kg*)
- 0.0597 ng/1 (rounded off to 0.06 ng/1 (Tier 1))
Where: *BAF - 2,296,650, provided by EPA-Duluth and Minnesota PCA.
References:
Bayard, S. 1991. lexicologist/Statistician with the U.S. EPA Office of
Research and Development, Human Health Assessment Group. Personal
communication with R. Sills, Michigan Department of Natural
Resources.
-------
Cabral, J. et al. 1982. Effects of long-term intake of DDT on rats.
Tumori 68:11-17.
Holder, J. 1991. lexicologist with the U.S. EPA Office of Research and
Development, Hjman Health Assessment Group. Personal communication
with R. Sills, Michigan Department of Natural Resources.
Rossi, L. et al. 1977. Long-term administration of DDT or
phenobarbital-Na in Wistar rats. Int. J. Cancer. 19:179-185.
Schoeny, R. 1991. U.S. EFA Environmental Criteria Assessment Office,
Chair of the Cancer Risk Assessment Verification Endeavor (CRAVE)
workgroup. Personal communication with R. Sills, Michigan
Department of Natural Resources.
Tarjan, R. and T. Keneny. 1969. Multigeneration studies on DDT in
mice. Food Cosmet. Toxicol. 7:215-222.
Terracini, B. et al. 1973. The effects of long-term feeding of DDT to
BALB/c mice. Int. J. Cancer. 11:747-764.
Thorpe, E. and A. Walker. 1973. The toxicology of dieldrln. II.
Comparative long-term oral toxicity studies in mice with dieldrin,
DDT, phenobarbital, beta-BHC and gamma-BBC. Food Cosmet. Toxicol.
11:433-442.
Totnatis, L. and V. Turusov. 1975. Studies on the carcinogenicity of
DDT. Gann Monograph on Cancer Research. 17-219-241.
-------
Turusov, V. et al. 1973. Tumors In CF-1 mice exposed for six
consecutive generations to DDT. J. Natl. Cancer Inst. 51:983-998.
U.S. Environmental Protection Agency (EPA). 1980. 45 Federal Register
No. 231, pp. 79347-79356. Appendix C - Guidelines and Methodology
Used in the Preparation of the Consent Decree Water Criteria
Documents.
U.S. Environmental Protection Agency (EPA). 1986a. The Assessment of
the Carcinogenicity of Dicofol (Kelthane), DDT, DDE, and DDD (TDE).
OHEA/ORD. EPA/600/6-86/001. PB 87-110904.
U.S. Environmental Protection Agency (EPA). 1986b. 51 Federal Register
No. 185, pp. 33992-34003. Guidelines for Carcinogen Risk Assessment,
U.S. Environmental Protection Agency (EPA). 1987. Intergrated Risk
Information System (IRIS database). Chemical file for DDT (59-29-3).
Verification Date 6/24/87. Last Revised 5/1/91.
U.S. Environmental Protection Agency (EPA). 1989. Risk Assessment
Guidance for Superfund. Volume 1. Human Health Evaluation Manual
(Part A). Interim Final. OERR. EPA/540/1-89/002.
-------
DRAFT
September 6, 199!
Great Lakes Initiative
Human Health Criteria for
Dieldrin
CAS No. 60-57-1
1. Tier 1 Human Noncancer Criterion
A review of the available literature indicates that the most
appropriate studv for HNC derivation for dieldrin is a two year
study conducted by Walker et al. (1969). In this study, 25 Carworth
Farm "E" rats of each sex were administered 0.1, 1.0 or 10.0 ppm
dieldrin in their diet and 45 rats of each sex were used as
controls. At the end of two years, the females exposed to 1.0 and
10.0 ppm had increased liver weights and liver-to-faody weight
ratios. Histopathological examination of these animals found
changes in p«»renchymal cells which included focal proliferation and
focal hyperp.asia. A NOAEL of 0.1 ppm (estimated to be 0.005
mg/kg/day) was determined from the study In support of this value
a systemic NOEL of 0.005 mg/kg/day was calculated for dogs in the
same study.
Studies examining the reproductive effects of dieldrin are lacking
(EPA, 1987). A review of studies which examine the developmental
effects of dieldrin in mice (Chernoff et al. , 1975; Dix et al.,
-------
1977) and rats (Harr et al., 1970; Chernoff et al., 1975) suggest
that exposure levels which may result in adverse developmental
effects are higher than the NOAEL determined in the Walker et al.
(1969) study.
The quality of the Walker et al. (1969) study was deemed sufficient
to derive a Tier 1 HNC. This study was also used by EPA (1987) to
derive the oral RfD for dieldrin. The HNC was derived from the
NOAEL (0.005 mg/kg/day) using an uncertainty factor of 100 to
account for intraspecies variability and interspecies extrapolation.
ADE = Q 005 mg/kg/d - 0.00005 mg/kg/d = 50 ng/kg/d
100
Where Uncertainty Factor - 100, composed of
lOx for intraspecies variability
lOx for interspecies extrapolation
Drinking Water Sources:
HNV - ADE x Wh x RSC - 50 ng/kg/d x 70 kg x 0 8
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 222,924 I/kg*)
- 0.84 ng/1 (rounded off to 0.8 ng/1 (Tier 1))
-------
Nondrinking Water Sources
HNV = APE \ Wh x RSC = 50 ng/kg/d x 70 kg x 0.8
WC +
-------
U.S. Environmental Protection Agencv (EPA). 1987. Integrated Risk
Information System (IRIS database). Chemical file for dieldrin
(60-57-1). Verification Date A/16/87. Last Revised 9/1/90.
Walker, A.I.T., D.E. Stevenson, J. Robinson, E. Thorpe and M. Roberts.
1969. The toxicology and phannacodynamics of dieldrin (HEOD): Two
year oral exposures of rats and dogs. Toxicol. Appl. Pharmacol.
15:345-373.
Tier 1 Human Cancer Criterion
According to EPA (1987a), there are inadequate data available to
ascertain whether dieldrin is a human carcinogen. However, chronic
studies have shown that dieldrin induces the formation of liver
tumors in seven strains of mice when administered orally.
Additional support for dieldrin's carcinogenicity is provided bv
its structural similarity to other compounds (i.e. heptachlor and
chlordane) which have been found to induce tumors in rodents.
Dieldrin has also produced a positive response in several
mutagenicity studies. The weight-of-evidence for dieldrin
carcinogenicity is sufficient for B2 (probable human carcinogen)
classification (EPA, 1987a). The data are sufficient to derive a
Tier 1 HCC.
Six key studies (Davis, 1965 as reevaluated by Reuber and cited in
Epstein, 1975; Walker et al., 1972; Thorpe and Walker, 1973; NCI,
-------
1978, Tennekes et al , 1981; Meierhenrv et al. , 1983) have reported
liver tumor induction in mice exposed orally to dieldrin. EPA
(1987a; 1987b) calculated 13 different slope factors using data from
these studies. The calculated slope factors were within an
eight-fold range. EPA (1987a; 1987b) calculated a single oral slope
factor of 1.6 x 10 (mg/kg/day) by taking the geometric mean of
the 13 slope factors computed from the key studies. This method of
computing a slope factor is used "in situations where no single
study is judged most appropriate, yet several studies collectively
support the estimate. ." (EPA, 1989). According to EPA (1989), the
advantage of this method of determining the slope factor is that all
relevant data are used in the computation.
RAD - 1 x L0~ - 6 x 10~ mg/kg/d - 0.6 ng/ke/d
1.6 *. 101 (mg/kg/d)'1
Drinking Water Sources
HCV - RAD x Wh - 0 6 ng/kg/d x 70 kg
WC + (FC x BAF) 2 1/d + (0 015 kg/d x 222,924 I/kg*)
- 0.013 ng/1 (rounded off to 0.01 ng/1 (Tier 1))
-------
Nondrinking Water Sources
HCV - RAD x Wh - 0.6 ng/kg/d x 70 kg
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 222,924 I/kg*)
= 0 013 ng/1 (rounded off to 0.01 ng/1 (Tier 1))
*BAF - 222,924, provided by EPA-Duluth and Minnesota PCA.
REFERENCES
Davis, K J 1965. Pathology report on mice fed aldrin, dieldrin,
heptachlor or heptachlor epoxide for two years Internal FDA
memorandum to Dr A. J. Lehman. July 19. As cited in: Epstein,
1975; EPA, 1987a.
Epstein, S.S., 1975. The carcinogenicity of dieldrin. Part 1. Sci.
Total Environ. 4 1-52.
Meierhenrv, E.F., B.H. Reuber, M.E. Gershwin, L.S. Hsieh and S.W.
French. 1983. DieIdrin-induced mallory bodies in hepatic tumors of
mice of different strains. Hepatology. 3-90-95.
National Cancer Institute. 1978. Bioassays of aldrin and dieldrin for
possible carcinogenicity. DHEW Publication No. (NIH) 78-822
National Cancer Institute Carcinogenesis Technical Report Series,
No. 22. NCI-CG-TR-22.
-------
Tennekes, H.A., A.S. Wright, K M. Dlx and J.H. Koeman. 1981. Effects
of dieldrin, diet and bedding on enzvme function and tumor incidence
in livers of male CF-1 mice. Cancer Res. 41-3615-3620.
Thorpe, E. and A.I.T. Walker. 1973. The toxicology of dieldrin
(HEOD). Part II. Comparative long-term oral toxicology studies in
mice with dieldrin, DDT, phenobarbitone, beta-BHC and gamma-BHC.
Food Cosmet Toxicol. 11 433-441
U.S. Environmental Protection Agencv (EPA) 1987a. Integrated Risk
Information System (IRIS database). Chemical file for chlordane
(57-74-9), Verification Date 3/5/87. Last Revised 1/1/91.
U.S. Environmental Protection Agency (EPA). 1987b. Carcinogenicity
Assessment of Aldrin and Dieldrin. Prepared by Carcinogen
Assessment Group, Office of Health and Environmental Assessment,
Washington, DC for Hazard Evaluation Division, Office of Pesticide
Programs, Office of Pesticides and Toxic Substances. OHEA-C-205.
U.S. Environmental Protection Agency (EPA). 1989. Risk Assessment
Guidance for Superfund. Volume 1. Human Health Evaluation Manual
(Part A). Interim Final. OERR. EPA/540/1-89/002.
Walker, A.I.T., E. Thorpe and D.E. Stevenson. 1972. The toxicology of
dieldrin (HEOD). I. Long-term oral toxicity studies in mice. Food
Cosmet. Toxicol. 11:415-432.
-------
DRAFT
September 6, 1991
Great Lakes Initiative
Human Health Criteria for
Chlordane
CAS No 57-74-9
Tier 1 Human Noncancer Criterion
A review of the available literature indicates that the most
appropriate studv for HNC derivation for chlordane is a study
conducted by Velsicol Chemical Corporation (1983a). In this study,
80 Fischer 344 rats of each sex were administered 0, 1, 5 or 25 ppm
chlordane for 130 weeks Hematological, biochemical, urinarv and
pathological measurements were made on eight animals/sex/group at
weeks 26 and 52. The same measurements were made on animals which
survived to week 130. Liver hypertrophy occurred in females at 5
ppm (0.273 mg/kg/d) and a NOEL of 1 ppm (0.055 mg/kg/d) was
determined. No liver lesions were found in male rats and a NOEL of
25 ppm (0.1175 mg/kg/d) was determined.
The NOEL for female rats is slightly lower than the NOELs determined
for mice and dogs. A 24-month chronic study in ICR mice found
NOELs of 0.123 rag/kg and 0.138 mg/kg for male and female mice,
respectively (Velsicol, 1983b). A study using Beagle dogs found
NOELs of 0.06 mg/kg and 0.09 mg/kg for male and female dogs,
respectively (Wazeter, 1967).
-------
Data on Che reproductive and developmental effects of chlordane are
limited. ATSDR (1989) cited a study bv Usami et al. (1986) which
showed no malformations in pups whose dams were administered 20, «0
or 80 mg/kg/d chlordane from day 7 to 17 of gestation. The finding
of this study suggests that exposure levels which may cause adverse
effects on development are higher than the NOEL cited above for the
Velsicol study (1983a). Other studies reported by Chernoff and Kavlock
(1982) and Ingle (1952 as cited by EPA, 1990) also indicate that
criteria derived from the chronic NOAEL of 0.055 mg/kg/d (Velsicol,
1983a) should be protective of potential developmental effects.
The quality of the Velsicol (1983a) rat study was deemed sufficient
to derive a Tier 1 HNC. This study was also used by EPA (1989; 1990)
to derive the oral RfD for chlordane. The HNC was derived from the
female rat NOIL using an uncertainty factor of 100 to account for
intraspecies variability and interspecies extrapolation.
ADE - 0.055 mg/kg/d - 0.00055 mg/kg/d - 550 ng/kg/d
100
Where: Uncertainty factor - 100. composed of*
lOx for intraspecies variability
lOx for interspecies extrapolation
-------
Drinking Water Sources.
HNV - APE x Wh x ESC - 550 ng/kg/d x 70 kg x 0.8
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 263,250 I/kg*)
- 7.80 ng/1 (rounded off to 8 ng/1 (Tier D)
Nondrinking Water Sources*
HNV - APE x Wh x RSC - 550 ng/kg/d x 70 kg x 0.8
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 263,250 I/kg*)
- 7 80 ng/1 (rounded off to 8 ng/1 (Tier 1))
Where: RSC - 0.8; the substance is persistent and bioaccumulative,
*BAF - 263,250, provided by EPA-Duluth and Minnesota PCA.
REFERENCES
Agency for Toxic Substances and Disease Registry (ATSDR). 1989.
Toxicologlcal Profile for Chlordane. Department of Health and
Human Services. U.S. Public Health Service.
Chernoff, N. and R. J. Kavlock. 1982. An in vivo teratology screen
utilizing pregnant mice. J. Toxicol. Environ. Hlth. 10:541-550
-------
Ingle, L. 1952. Chronic oral toxicitv of chlordane to rats. Arch.
Ind. Hyg Occup. Med. 6.357-367
U.S. Environmental Protection Agencv (EPA). 1989. Integrated Risk
Information Svsteo (IRIS database). Chemical file for chlordane
(57-74-9). V«rification Date 3/22/89. Last Revised 7/1/89.
U.S. Environmental Protection Agency (EPA). 1990. Drinking Water
Criteria Docunent for Heptachlor, Heptachlor Epoxide and Chlordane.
Revised November, 1990. ECAO-CIN-406.
Usami, M., K. Kawa;>hima, S. Nakaura, et al. 1986. Effects of chlordane
on prenatal development of rats. (Abstract). Eisei Shikenso Hokoku.
104:68-73.
Velsicol Chemical Corporation. 1983a. Yonemura, T., F. Takamura and
Y. Takahashi. Thirty-month chronic toxicity and tumorigenicity test
in rats by chLordane technical. (Unpublished study by Research
Institute for Animal Science in Biochemistry and Toxicology, Japan).
Velsicol Chemical Corporation. 1983b. Inui, S., K. Yamazaki,
T. Yonemura, «t al. Twenty-four month chronic toxicity and
tuaorigenicitv test in mice by chlordane technical. (Unpublished
study by Research Institute for Animal Science in Biochemistry and
Toxicology, Japan).
Wazeter, F.X. 196?. Two-Year Chronic Feeding Study in the Beagle Dog.
Sponsored by Velsicol Chemical Corporation (Unpublished).
-------
Tier 1 Human Cancer Criterion
There are inadequate data available to determine whether chlordane
is a human carcinogen (EPA, 1987, 1990). Chronic studies using four
strains of mice (CD-I, B6C3FI, C5781/6N, ICR) of both sexes have
shown an increase in the occurrence of liver tumors. Additional
weight-of-evidence for chlordane's carcinogenicity is provided by
its structural similarity to other compounds (i.e., dieldrin and
heptachlor) which have been found to induce hepatocellular
carcinomas in mice. The results of various mutagenicity studies
are inconclusive as to this chemical's ability to cause mutagenic
effects. The weight-of-evidence for chlordane carcinogenicity is
sufficient for B2 (probable human carcinogen) classification (EPA,
1986; 1987; 1990). The data are sufficient to derive a Tier 1 HCC.
Two key studies (NCI, 1977; Velsicol, 1973 as cited in EPA, 1986)
found a significant increase in hepatocellular carcinomas in
treatment groups when compared to controls. Both studies also
shoved a dose-response relationship between exposure of mice to
chlordane and the occurrence of liver tumors. EPA (1986; 1987; 1990)
calculated four separate slope factors from these key studies,
and derived a recommended slope factor of 1.3 (mg/kg/d) from the
geometric mean of these slope factors. This method of computing a
slope factor is used "in situations where no single study is Judged
most appropriate, yet several studies collectively support the
estimate ..." (EPA, 1989). According to EPA (1989), the advantage
of this method of determining the slope factor is that all relevant
data are used in the computations. /T"!
-------
RAD - 1 x 10'5 - 7 7 x 10~6 mg/kg/d - 7.7 ng/kg/d
1.3 (mg/kg/d)-i
Drinking Water Sources'
HCV - RAD x Wh - 7.7 ng/kg/d x 70 kg
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 263,250 I/kg*)
- 0.14 ng/1 (rounded off to 0.1 ng/1 (Tier 1))
Nondrinking Water Sources:
HCV - RAD x Wh - 7 7 ng/kg/d x 70 k«
WC + ''FC x BAF) 0.01 1/d + (0.015 kg/d x 263,250 I/kg*)
- 0.14 ng/1 (rounded off to 0.1 ng/1 (Tier 1))
*BAF - 263,250, provided by EPA-Duluth and Minnesota PCA.
REFERENCES
National Cancer Institute (NCI). 1977. Bioassay of Chlordane for
possible carclnogenicity. NCI Carcinogenesis Tech. Rep. Ser. No. 8,
U.S. DHEW Publ. No. (NIH) 77-808. Bethesda, MD.
-------
U.S Environmental Protection Agency (EPA). 1986. Carcinogenicity
Assessment of Chlordane and Heptachlor/Heptachlor Epoxide
Carcinogen Assessment Group. Office of Health and Environmental
Assessment, Washington, DC.
U.S. Environmental Protection Agency (EPA). 1987. Integrated Risk
Information System (IRIS database). Chemical file for chlordane
(57-74-9). Verification Date 4/1/87. Last Revised 1/1/91.
U.S. Environmental Protection Agency (EPA). 1989 Risk Assessment
Guidance for Superfund. Volume 1. Human Health Evaluation Manual
(Part A). Interim Final. OERR. EPA/540/1-89/002.
U.S. Environmental Protection Agency (EPA). 1990. Drinking Water
Criteria Document for Heptachlor, Heptachlor Epoxide and Chlordane.
Revised November, 1990. ECAO-CIN-406.
-------
DRAF1
September 6, 1991
Great Lakes Initiative
Human Health Criteria for
Hexachlorobenzene
CAS No. 118-74-1
1. Tier 1 Human Noncancer Criterion
A review of the available information on hexachlorobenzene (HCB)
toxicity, including reviews by EPA (1980; 1985a; 1985b; 1988),
indicates that the database is sufficient for Tier 1 HNC derivation.
The best available data consist of laboratory animal studies.
The principle human data on HCB toxicity consist of widespread toxic
effects among several thousand Turkish citizens exposed to HCB via
consumption of fungicide-treated grain during 1955-1959. The
resulting effects included porphyria cutanea tarda (PCT),
neurotoxicity, liver damage, and increased infant mortality. The
exposure has been estimated at 50-200 mg/day over an extended
period, without further description of the dosage estimation method
(Cam and Nlgogosyan, 1963). The human data cannot be used for
quantitative risk assessment because accurate exposure data are not
available (EPA, 1988).
The available literature indicates that HCB is a potent developmental
toxicant in several animal species. Kitchin et al. (1982) exposed
-------
rats to 0, 60, 80, 100, 120 and 140 ppm HCB In feed (doses were
approximately 4.5-10 mg/kg/day). Thev reported a dose-dependent
increase in mortality of pups in the Fla and Fib litters. Grant et
al. (1977) conducted a 4-generation reproduction study with rats at
food HCB levels in the diet of 0, 10, 20, 40, 80, 160, 320 and 640
ppm. They concluded that 20 ppm (about 1.5 mg/kg/day} was a NOAEL,
while 40 ppm (about 3 mg/kg/day) resulted in increased liver weights
and aniline hydroxylase activities in weanlings.
Rush et al. (1983) exposed mink to 0, 1 or 5 ppm in feed (about 0.16
or 0.78 mg/kg/day), resulting in profound effects on kit
survivabilitv to weaning at the high dose. Mortality was 8.2
percent, 4 1 percent and 77.4 percent among controls, low dose, and
high dose groups, respectively. Bleavins et al. (1984a, 1984b) also
reported that mink, as well as ferrets, are highlv sensitive to the
developmental effects of HCB. Mink were found to be more sensitive
than ferrets, while both appeared more sensitive than rats according
to published data. The most profound effects reported were
decreased mink birth weights at adult dietary levels as low as 1 ppm
and a dose-related increase in kit mortality at three weeks of age
among both mink and ferrets at levels as low as 1 ppm (about 0.14
and 0.11 mg/kg/day for mink and ferrets, respectively).
Additionally, effects were seen on the levels of hypothalamic
dopamine of mink kits and on hypothalamic serotonin in adult mink.
These changes were statistically significant at levels as low as
1 ppm in feed. A NOAEL was not reported.
-------
Arnold et al. (1985) exposed male and female rats to dietary HCB
levels of 0, 0.32, 1.6, 8.0 or 40 ppm for 90 days prior to mating
and until 21 days after parturition (at weaning). The offspring
were exposed in utero, from maternal nursing, and from their diets
for the remainder of their lifetime. The total study period was 130
weeks. A NOAEL was reported at 1.6 ppm (about 0.08 mg/kg/day). At
8 ppm (0.29 mg/kg/day) the parental (Fo) males demonstrated
increased heart and liver weights and the Fl generation had an
increased inc tdence of hepatic centrilobular basophilic
chromogenesis The 40 ppm Fl groups showed increases in pup
mortalitv, hematic centrilobular basophilic chromogenesis, and
severe chronic nephritis (males only).
The effects of HCB on adult animals has been further demonstrated
in many other studies, a few of which report NOAELs. Kuiper-Goodman
et al. (1977) exposed rats via the diet to 0, 0.5, 2, 8 and 32 mg/kg
bw/day for up to 15 weeks. A NOAEL was reported at 0.5 mg/kg/day,
while at the higher dose levels, increased tissue porphyrin,
increased organ weights and increased severity of centrilobular
liver lesions, [>«lliu-lufeJLii1 Likmi(j,i u were noted. Grant et al. (1974)
exposed rats to HCB at dietary levels of 10, 20, 40, 80 and 160 ppm
for 9-10 months. Porphyria was induced at levels as low as 20 ppm
(about 1 mg/kg/day).
The data selected for HNC determination are from the Arnold et al.
(1985) study. This study involved the exposure of adult
Sprague-Dawley rats and subsequent exposure of the offspring in
utero, via lactation, and via the diet. Cross-fostering studies
-------
have demonstrated that the neonate is particularly sensitive to the
toxic effects of HCB. The transfer of HCB to neonates via the milk
of exposed adults has also been shown to be significant (Bailey et
al. 1980; Bleavins et al. 1982). The Arnold et al. (1985) study
demonstrated NOAELs of 0.32 and 1.6 ppm in feed (estimated to be
0.016 and 0.08 mg/kg/day). Therefore, the NOAEL of 0.08 mg/kg/day,
with an uncertainty factor of 100 (lOx for each inter- and
intraspecies extrapolation) is used for HNC derivation.
Consideration was also given to the mink and ferret data (Rush et
al., 1983; Bleavins et al., 1984a, 1984b) which demonstrate that
these species are highly sensitive to the developmental toxic
effects of HCB. Adverse effects on the development of mink and
ferrets have been reported at doses only slightly higher than the
rat NOAEL of 0.08 mg/kg/day. However, the rat NOAEL is utilized
preferentially because the Sprague-Dawley rat, unlike the mink or
ferret, is a human-surrogate species which has been extensively
studied as an animal model for toxicity testing, and because the
Arnold et al. (1985) study is of very good quality. The selection
of this key study is consistent with the development of the RfD for
HCB by EPA (1988).
ADE - 0.06 mg/kg/d = 0.0008 mg/kg/d - 0.8 ug/kg/d
100
Where: Uncertainty factor » 100, composed of:
lOx for intraspecies variability
lOx for interspecies extrapolation
-------
Drinking Water Sources.
HNV = APE * Wh x RSC - 0.8 ug/kg/d x 70 kg x 0.8
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 239,154 I/kg*)
- 0.012 ug/1 (rounded off to 10 ng/1 (Tier 1))
Nondrinking Water Sources.
HNV - APE > Wh x RSC - 0.8 ug/kg/d x 70 kg x 0.8
WC + (FC x BAF) 0.01 1/d -1- (0.015 kg/d x 239,154 I/kg*)
- 0.01? ug/1 (rounded off to 10 ng/1 (Tier 1))
Where- RSC == 0.8; the substance is persistent and bioaccumulative.
*BAF • 239,1 '54, provided by EPA-Duluth and Minnesota PCA.
REFERENCES
Arnold, D.L. et al. 1985. Long-term toxicity of hexachlorobenzene in
the rat and the effect of dietary vitamin A. Fd. Chem. Toxic.
23(9):779-793.
Bailey, J., V. Knauf, W. Mueller and W. Hobson. 1980. Transfer of
hexachlorobenzene and polychlorinated biphenyls to nursing infant
rhesus monkeys: Enhanced toxicity. Environ. Res. 21(1): 190-196
-------
Bleavins, M.R., W.J. Breslin, R.J. Aulerich and R.K. Ringer. 1982.
Excretion and placental and mammary transfer of hexachlorobenzene in
the European ferret (Mustela putorius furo). J. Toxicol. Environ.
Health. 10:929-940
Bleavins, M., R. Aulerich and R. Ringer. 1984a. Effects of chronic
dietary hexachlorobenzene exposure on the reproductive performance
and survivability of mink and European ferrets. Arch. Environ.
Contain. Toxicol. 13 357-365.
Bleavins, M. et al. 1984b. Effects of dietary hexachlorobenzene
exposure on regional brain biogenic amine concentrations in mink
and European ferrets Toxicol Environ. Hlth 14-363-377.
Cam, C and G. Nigogosvan 1963 Acquired toxic porphyria cutanea
tarda due to hexachlorobenzene. Report of 348 cases caused by this
fungicide J. Am. Med. Assoc. 183:88-91.
Grant, D. et al. 1974. Effects of hexachlorobenzene on liver porphyrin
levels and microsomal enzymes in the rat. Environ. Physiol.
Biochem. 4:159-165.
Grant, D., W. Phillips and G. Hatina. 1977. Effect of
hexachlorobenzene on reproduction in the rat. Arch. Environ.
Contain. Toxicol. 5(2): 207-216.
Kitchin, K. et al. 1982. Offspring mortality and maternal lung
pathology in female rats fed hexachlorobenzene. Toxicol.
23-33-39.
-------
Kuiper-Goodman, T. et al. 1977. Subacute toxicity of hexachlorobenzene
in the rat. Toxicol. and Appl. Phannacol. 40:529-549.
Rush, G. et al. 1983. Perinatal hexachlorobenzene toxicity in the
mink. Environ. Res. 31:116-124.
U.S. Environmental Protection Agency (EPA). 1980. Ambient Water
Quality Criteria for Chlorinated Benzenes. EPA 440/5-80-028.
U.S. Environmental Protection Agency (EPA). 1985a. Drinking Water
Criteria Document for Hexachlorobenzene (Final Draft).
EPA - 600/X-84-179-1. PB-86-117777.
U S Environmental Protection Agency (EPA) 1985b. Health Assessment
Document for Chlorinated Benzenes. EPA/600/8-84/015F.
U.S. Environmental Protection Agency (EPA). 1988. Integrated Risk
Information System (IRIS database). Chemical file for
hexachlorobenzene (118-74-1). Verification Date 5/26/88. Last
Revised 4/1/91.
2. Tier 1 Human Cancer Criterion
A review of the available literature indicates that there are
Inadequate epidemiological studies and sufficient animal
carcinogenicity data, supporting a B2 weight-of-evidence
-------
classification (EPA, 1989). The animal bioassays, which have been
comprehensively reviewed and summarized by EPA (1980; 1985a; 1985b;
1989), indicate that HCB induces tumors of the liver predominantly,
with neoplasm induction of the thyroid and kidney also reported.
The data are judged sufficient for Tier 1 HCC derivation.
EPA (1991) derived an oral slope factor of 1.6 per (mg/kg)/day
from a chronic rat bioassay demonstrating hepatocellular carcinoma
induction (Erturk et al., 1986). This slope factor is among the
highest of those derived for HCB from 14 different datasets, which
fell within a range of 8.3 E-2 to 1.7 E+0 (EPA, 1989). This dataset
was also selected for slope factor estimation because the study was
well-conducted and the tumors were malignancies of the primary
target organ (liver cancers).
In the key study, Erturk et al. (1986; abstracts previously
published as Lambrecht et al., 1983a; 1983b) exposed groups of 94
Sprague-Dawley rats/sex/dose to HCB via feed at 0, 75 or 150 ppm in
the diet for up to two years. Treated animals of both sexes
surviving past 12 months showed significant increases in liver and
renal tumors. Females were far more susceptible to hepato-
carcinogenicity while males were generally more sensitive to renal
carcinogenicity. The slope factor of 1.6 per (mg/kg)/day is derived
from the induction of hepatocellular carcinomas in female rats.
This is consistent with EPA (1989).
RAD - 1 x 10"5 " 6 2 x 10~6 mg/kg/d = 6.2 ng/kg/d
1.6 (mg/kg/d)"1
-------
Drinking Watei Sources
HCV = RAD x Wh - 6.2 ng/kg/d x 70 kg
WC + IFC x BAF) 2 1/d + (0.015 kg/d x 239,154 I/kg*)
- 0.12 ng/1 (rounded off to O.| ng/1 (Tier 1))
Nondrinking Water Sources'
HCV = RAD x Wh - 6.2 ng/kg/d x 70 kg
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 239,154 I/kg*)
0.12 ng/1 (rounded off to 0.1 ng/1 (Tier 1))
Where: *BAF = 239,154, provided by EPA-Duluth and Minnesota PCA,
REFERENCES
Erturk, E. et al. 1986. Oncogenicity of hexachlorobenzene. In:
Hexachlorobenzene: Proc. Int. Symp., C.R. Morris and J.R.P. Cabral,
Eds. IARC Scientific Publ. No. 77, Oxford University Press, Oxford.
pp. 417-423.
Lambrecht, R., et al. 1983a. Renal tumors in rats chronically exposed
to hexachlorobenzene (HCB). Proceedings of the American Association
for Cancer Research 24:59.
-------
Lambrecht, R., et al. 1983b. Hepatocarcinogenicity of chronically
administered hexachlorobenzene in rats. Federation Proceedings.
42(4)-786.
L.S. Environmental Protection Agency (EPA). 1980. Ambient Water
Quality Criteria for Chlorinated Benzenes. EPA 440/5-80-028.
U.S. Environmental Protection Agency (EPA). 1985a. Drinking Water
Criteria Document for Hexachlorobenzene (Final Draft).
EPA-600/X-84-179-1. NTIS PB 86-117777.
U S Environmental Protection Agency (EPA). 1985b. Health Assessment
Document for Chlorinated Benzenes. EPA/600/8-84/015F.
U.S. Environmental Protection Agency (EPA). 1989. Integrated Risk
Information System (IRIS database). Chemical file for
hexachlorobenzene (118-74-1). Verification Date 3/1/89.
Last Revised 3/1/91.
-------
DRAFT
September 6, 1991
Great Lakes Initiative
Human Health Criteria for
Heptachlor
CAS No 76-U-8
1. Tier 1 Human Noncancer Criterion
A review of (he available literature indicates that the most
appropriate s»tudy for HNC derivation for heptachlor is a two-year
study conducted by Velsicol Chemical Corporation (1955) as cited by
EPA (1987). In this study, CF strain rats (20/sex/group) were fed
diets containing 0, 1.5, 3, 5, 7, or 10 ppm heptachlor. The NOEL
for male rat-s was 3 ppm (0.15 mg/kg/d) with the liver-to-body weight
ratio reported as the sensitive endpoint. The LEL for this critical
effect was 5 ppm (0.25 mg/kg/d).
Two studies suggest that exposure levels which may cause adverse
effects on reproduction are higher than the NOEL determined in the
study cited above. EPA (1987) cited a NOEL of 0.25 mg/kg/d in a
one-generation rat reproduction test (Velsicol, 1955) and a NOEL of
0.5 mg/kg/d in a three-generation rat reproduction test (Velsicol,
1967). However, Green (1970) found adverse effects on reproduction
in rats which were administered heptachlor in feed at 5 ppm
(approximately 0.25 mg/kg/d). This value was considered a LOAEL
and no NOAEL was established in this study. With respect to
-------
developmental toxicity, Yamaguchi et al. (1987) found no teratogenic
effects in the offspring of female rats dosed orally with 5, 10 or
20 mg/kg heptachlor from days 7-17 of gestation.
The quality of the Velsicol (1955) studv and supporting data was
deemed sufficient to derive a Tier 1 HNC. This study was also used
by EPA (1987) to derive the oral RfD for heptachlor.
ADE - 0.15 mg/kg/d - 0.0015 mg/kg/d - 1.5 ug/kg/d
100
Where Uncertainty Factor « 100, composed of.
lOx for intraspecies variability
lOx for Interspecies extrapolation
Drinking Water Sources-
HNV • ADE x Wh x RSC - 1.5 ug/kg/d x 70 kg x 0.8
WC -f (FC x BAF) 2 1/d + (0.015 kg/d x 28,798 I/kg*)
• 0.19 ug/1 (rounded off to 0.2 ug/1 (Tier 1))
Nondrinking Water Sources:
HNV " ADE x Wh x RSC - 1.5 ug/kg/d x 70 kg x 0.8
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 28,798 I/kg*)
- 0.19 ug/1 (rounded off to 0.2 ug/1 (Tier 1))
-------
Where: RSC • 0.8; the substance is persistent and bioaccumulative,
*BAF - 28,798, provided by EPA-Duluth and Minnesota PCA.
REFERENCES
Green, V. 1970. Effects of Pesticides on Rat and Chick Embryo. In-
Hemphill, D. (Ed.). 1970. Trace Substances in Environmental
Health-Ill. Proceedings of University of Missouri's 3rd Annual
Conference on Trace Substances in Environmental Health, June 24-26,
1969.
U.S. Environmental Protection Agency (EPA). 1987. Integrated Risk
Information System (IRIS database). Chemical file for heptachlor
(76-44-8). Verification Date 4/16/87. Last Revised 3/1/91.
Velsicol Chemical Corporation. 1955. MRID No. 0062599. Available from
EPA. Write to FOI, EPA, Washington, DC 20460.
Velsicol Chemical Corporation. 1967. MRID No. 00147058. Available
from EPA. Write to FOI, EPA, Washington, DC 20460.
Yamaguchi, M., S. Tanaka, K. Kawashima, S. Nakaura and A. Takanaka.
1987. Effects of heptachlor on fetal development of rats. Natl.
Inst. Hyg. Sci. 105:33-36.
-------
2. Tier 1 Human Cancer Criterion
There are inadequate data available to ascertain whether heptachlor
is a human carcinogen. However, heptachlor exposure has caused a
significant increase in hepatocellular carcinomas in two strains
of mice. In addition, heptachlor is structurally related to other
compounds (i.e., dieldrin and chlordane) which have been found to
induce the formation of liver tumors in mice. Heptachlor has
produced negative results in gene mutation assays, mouse dominant
lethal assays and in vitro DNA repair assays (EPA, 1980; 1986; 1987).
According to the weight of evidence approach employed by EPA (1986;
1987; 1990), there is sufficient evidence to classify heptachlor as
a B2 carcinogen (probable human carcinogen). There is also
sufficient evidence to derive a Tier 1 HCC.
Two key studies (Davis, 1965 as cited by Epstein, 1976; NCI, 1977)
have reported hepatocellular tumors in mice exposed orally to
heptachlor. EPA (1986; 1987; 1990) calculated four slope factors
using data from these studies. The slope factors ranged from 0.83
to 14.9 (mg/kg/d)~ . EPA (1987) calculated a single slope factor of
4.5 (mg/kg/day)~ by taking the geometric mean of the four slope
factors from the key studies. This method of computing a slope factor
Is used "in situations where no single study is judged most
appropriate, vet several studies collectively support the estimate
..." (EPA, 1989). The advantage of this method of determining the
slope factor is that all relevant data are used in the computations
(EPA, 1989).
-------
RAD - I x 10'5 - 0.0000022 mg/kg/d - 2.2 ng/kg/d
4.5 (ag/kg/d)"1
Drinking Water Sources*
HCV - RAD x Wh - 2.2 ng/kg/d x 70 kg
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 28,798 I/kg*)
- 0.35 rg/1 (rounded off to 0.4 ng/1 (Tier 1))
Nondrinking Water Sources:
HCV - RAD x Wh - 2 2 ng/kg/d x 70 kg
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 28,798 I/kg*)
- 0.36 ng/1 (rounded off to 0.4 ng/1 (Tier 1))
*BAF - 28,798, provided by EPA-Duluth and Minnesota PCA.
REFERENCES
Davis, K. 1965. Pathology Report on Mice Fed Aldrin, Dieldrin,
Heptachlor and Heptachlor Epoxide for Two Years. Internal FDA
memorandum to Dr. A. J. Lehman, July 19.
Epstein, S.S. 1976. Carcinogenicity of heptachlor and chlordane. Sci.
Total Environ. 6:103-154.
-------
National Cancer Institute (NCI). 1977 Bioassay of Heptachlor for
Possible Carcinogenicitv. SCI Carcinogenesis Tech. Rep. Ser. No. 9,
U.S. Environmental Protection Agency (EPA). 1980. Ambient Water
Quality Criteria for Heptachlor. Criteria and Standards Office.
Washington, DC. EPA 440/5-80-052.
U.S. Environmental Protection Agency (EPA). 1986. Carcinogenicity
Assessment of Chlordane and Heptachlor/Heptachlor Epoxide.
Carcinogen Assessment Group. Office of Health and Environmental
Assessment, Washington, DC.
U.S. Environmental Protection Agency (EPA). 1987. Integrated Risk
Information System (IRIS database) Chemical file for heptachlor
(76-44-8). Verification Date 4/1/87. Last Revised 1/1/91.
U.S. Environmental Protection Agency (EPA). 1989. Risk Assessment
Guidance for Superfund. Volume 1. Human Health Evaluation Manual
(Part A). Interim Final. OERR. EPA/540/1-89/002.
U.S. Environmental Protection Agency (EPA). 1990. Drinking Water
Criteria Document for Heptachlor, Heptachlor Epoxide and Chlordane
Revised November, 1990. ECAO-CIN-406.
-------
DRAFT
September 6, 1991
Great Lakes Initiative
Tier 1 Human Health Criteria for
Mercury
CAS No. 7439-97-6
(including methylmercury, Cas No. 22967-92-6)
1. Tier 1 Human Noncancer Criterion
A review of the available literature on the environmental cycling,
fate, and toxlcity of raercurv and mercury compounds indicates that
HNC derivation is most appropriately based upon the human
dose-response to methylmercury. Numerous reviews on mercury
toxicity (e.g , WHO, 1976, 1990; EPA, 1980; 1984a; 1984b; 1985a)
describe the human dose-response relationship resulting from
food-borne exposure to methvlmercury in Iraq (1971-72), Japan
(1940s thru l%0s) , and elsewhere. These data are judged to be
sufficient foi Tier 1 criterion derivation.
Studies of widespread human food-borne exposure to methylmercury in
fish (Minamata and Niigata, Japan) and in seed grain (Iraq) have
shown that neurological symptoms of mercury toxicity in adults
appear with b.ood levels of mercury in the range of 200 to 500 ng/ml
(Nordberg and Strangert, 1976; Clarkson et al., 1976; WHO, 1976;
1990; EPA, 1980; 1984a; 1984b; 1985a). However, there are a few
studies of workers exposed occupationally to mercury via inhalation /. 7)
-------
which suggest that blood mercury levels as low as 10-20 ng/tnl may
result in the development of signs of renal dysfunction (Increased
proteinurea and albumlnurea) and abnormal psychomotor performance
(Roels et al., 1982; Piikivi et al., 1984; Buchet et al., 1980).
The adult LOAEL of 200 ng/ml in blood has been associated with an
intake level of 200-500 ug/d (EPA, 1980; WHO, 1990), although the
human adult population's variability in mercury elimination rate is
significantly bimodal (Clarkson et al, 1976; Nordberg and Strangert,
1976). The human LOAEL of 200 ug/d, or 3 ug/kg/d, for the
development of neurological effects forms the basis for the RfD
derived by EPA (1985b) and the fish consumption criteria derived by
EPA (1980). It has been estimated that less than 5% of the adult
population will experience neurological effects at these levels
(WHO, 1990).
The risk assessments by EPA (1980) and EPA (1985b) utilized a total
uncertainty factor of 10 in conjunction with the LOAEL dose, and
both stated that the LOAEL and the risk assessment addressed the
sensitivity and the adequate protection of both pre- and postnatal
exposures. EPA (1980) justified the 10-fold uncertainty factor as
an accounting for "individual differences in habits of fish
consumption and in susceptibility to the toxic effects of
methylmercury, including prenatal exposures". EPA (1985b) justified
the 10-fold uncertainty factor "to adjust the LOAEL to what is
expected to be a NOAEL. Since the effects are seen in sensitive
individuals for chronic exposure, no additional factors are deemed
necessary".
-------
For the derivation of the Tier 1 Human Noncancer Criterion, a total
uncertainty factor of 50 will be utilized. This is composed of
a 10-fold factor to adjust the adult LOAEL to a presumed adult NOAEL
and an additional 5-fold factor to protect CNS development during
the sensitive fetal life stages. The use of a 10-fold factor for
LOAEL-to-NOAEL conversion is Justified by consideration of the
severity and Lrreverslbillty of the effects at the LOAEL, the long
latency of mercury effects, and the occupational studies which
suggest that the threshold may be considerably lower than 200 ng
Hg/ml blood.
An uncertainty factor of 5 is utilized to ensure that the criterion
will be protective of the fetal effects of mercury exposure via
maternal ingestion of mercury-contaminated fish. The particular
sensitivity of the fetus has been recognized in reviews of mercury
toxicity (WHO, 1976; 1990, D'ltri, 1978; EPA, 1980; 1984a; 1984b;
1985a). The earliest of these assessments (WHO, 1976) developed a
dose-response relationship for the adult which was not presented as
being accurate for the more sensitive fetal effects. It was noted
that many infant victims reported from Minamata had severe cerebral
involvement (palsy and retardation) whereas their mothers had mild
or no manifestations of poisoning. Although these observations
were qualitatively confirmed by animal studies, quantification of
the difference in the degree of sensitivity between human fetuses
and adults has been elusive. EPA (1980; 198Sb) utilized a total
uncertainty factor of 10 and assumed that the resulting risk
assessments were adequately protective of fetal effects. However,
-------
WHO (1990) reviewed the database on oral methylmercury ingestion,
Including more recent studies, and made significant advances in
delineating quantitatively the greater sensitivity of prenatal
exposure relative to adult exposure. Although WHO (1990) did not
recommend a particular numeric sensitivity factor for the fetus,
their assessment sufficiently demonstrates that an additional
uncertainty factor is reasonable and prudent to help ensure adequate
protection. They concluded that adult effects occur at a LOAEL
(for 57, increased occurrence rate) of 200 ng/ml blood, or at 50
ug/g in hair Fetal effects on CNS development occur at a LOAEL (5%
increased occurrence rate) of 10-20 ug/g as a peak level in maternal
hair. Since the level of mercury in maternal blood correlates to
the simultaneous level in new hair growth, the hair serves as a
fairlv reliable indicator of maternal blood mercury levels during
pregnancy. The data suggest that the fetal effects LOAEL may be
2.5 to 5 times lower than the adult effects LOAEL.
The HNC is derived from the adult LOAEL dose of 3 ug/kg/d which is
associated with the LOAEL in blood of 200 ng/ml, and an uncertainty
factor of 50. The methylmercury fonn is the most significant of the
mercury compounds from the standpoint of ambient environmental
mercury and human exposures and health impacts. Aqueous
concentrations of mercury, and especially methylmercury, may be very
low in ambient waters. Other forms of mercury, such as elemental
mercury or mercury (I), may be reasonably anticipated to be
transformed predominantly to methylmercury in the aquatic
environment via oxidation to mercury (II) and biomethylation.
The biomethylation of inorganic mercurv and the very high propensity
477
-------
for methylmerc-iry to bioaccumulate in aquatic organisms result in a
high and significant human exposure potential (EPA, 1980; D'ltri,
1990; Annett et al.f 1975). The various forms of mercury released
to and found in the ambient aquatic environment may be assumed to be
converted primarily to methylmercury. Therefore, the HNC is
expressed as the total recoverable mercury concentration.
ADE - 3 ug/kg/d - 0.06 ug/kg/d
50
Where* Uncertainty Factor - 50, composed of:
lOx for LOAEL-to-NOAEL conversion
5x for intraspecies variability (protection of fetal CNS
development)
Drinking Water Sources.
HNV - ADE x Wh x RSC - 0.06 ug/kg/d x 70 kg x 0 8
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 470,000 I/kg*)
- 0.00048 ug/1 (rounded off to 0.5 ng/1 (Tier 1))
Nondrinking Water Sources:
HNV - ADE x Wh x RSC - 0.06 ug/kg/d x 70 kg x 0.8
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 470,000 I/kg*)
- 0.00048 ug/1 (rounded off to 0.5 ng/1 (Tier 1))
-------
Where* RSC » 0.8; the substance is persistent and
bioaccutnulative.
*Interim BAF - 470,000, calculated by Michigan DNR.
REFERENCES
Annett, C.S. et al. 1975. Mercury in fish and waterfowl from Ball
Lake, Ontario J. Environ Qual. 4(2) 219-222.
Buchet, J.P., H. Roels, A Bernard and R. Lauwerys, 1980. Assessment of
renal function of workers exposed to inorganic lead, cadmium or
mercury vapor. J. Occup. Med. 22:741-750.
Clarkson, T.W., L. Amin-Zaki and S. K. Al-Tikriti. 1976. An outbreak of
methylmercury poisoning due to consumption of contaminated grain.
Federation Proceedings. 35(12)-2395-2399.
D'ltri, P.A. and F.M. D'ltri. 1978. Mercury contamination: a human
tragedy. Environmental Management. 2(1):3-16.
D'ltri, F.M. 1990. Mercury contamination - what we have learned since
Minamata. Environmental Monitoring and Assessment, v. 16.
473
-------
Nordberg, G.F. and P. Strangert. 1976. Estimations of a dose-response
curve for long-term exposure to methylmercuric compounds in human
beings taking Into account variability of critical organ
concentration and biological half-time: a preliminary communication.
In: Effects aid Dose-Response Relationships of Toxic Metals 1976.
Elsevier Scientific Publishing Company. Amsterdam, The
Netherlands, p. 273-282.
Piikivl, L., H. Hanninien, T. Martelin et al. 1984. Pyschological
performance and long term exposure to mercury vapors. Scand. J.
Work. Environ. Health 10:35-41.
Roels, J., R. Lauwerys, J.P. Buchet et al. 1982. Comparison of renal
function and psychomoter performance in workers exposed to elemental
mercury. Int. Arch. Occup. Environ. Health 50-77-93.
U.S. Environmental Protection Agency (EPA). 1980. Aabient Water
Quality Criteria Document for Mercury. EPA 440/5-80-058.
U.S. Environmental Protection Agency (EPA). 1984a. Mercury Health
Effects Update: Health Issue Assessment. OHEA. EPA-600/8-84-019F.
U.S. Environmental Protection Agency (EPA). 1984b. Health Effects
Assessment for Mercury. EPA/540/1-86/042. NTIS- PB86-134533.
-------
U.S. Environmental Protection Agency (EPA). 1985a. Drinking Water
Criteria Document for Mercury. Prepared for Office of Drinking
Water, by Environmental Criteria and Assessment Office.
EPA-600/X-84-178-1. Final Draft. PB86-117827.
U.S. Environmental Protection Agency (EPA). 1985b. Integrated Risk
Information System (IRIS database). Chemical file for methvlmercury
(22967-97-6). Verification Date 12/2/85. Last Revised 2/1/89.
World Health Organization (WHO). 1976. Environmental Health Criteria 1:
Mercury. WHO, Geneva.
World Health Organization (WHO). 1990. Environmental Health Criteria 101:
Methylmercury WHO, Geneva.
-------
DRAFT
September 6, 1991
Great Lakes Initiative
Tier 1 Human Health Criteria for
Pentachlorophenol
CAS No. 87-86-5
1. Tier 1 Human Noncancer Criterion
A review of the available literature indicates that HNC derivation for
pentachlorophenol (PeCP) is most appropriately based on the chronic
oral rat study by Schwetz et al. (1978). Twenty-five rats/sex
were administered PeCP in the diet for two years at levels resulting
in doses of 1, 3, 10 or 30 mg PeCP/kg bw/day. The test substance
was representative of Dowicide EC-7, a commercially available and
purified grade of PeCP. An accumulation of pigment in the liver and
kidneys was observed in females receiving 10 or 30 mg/kg/day and in
males receiving 30 mg/kg/day. The NOAEL was 3 mg/kg/day.
The chronic rat NOAEL at 3 mg/kg/day Is supported by several
subchronic anc reproduction/development studies (EPA, 1985a; 1985b).
Johnson et al. (1973) administered 3, 10 or 30 mg/kg bw/day purified
PeCP to rats lor 90 days via feed. Increased liver weights were
observed at 10 or 30 mg/kg/d, and increased kidney weights occurred
at 30 mg/kg/d The NOAEL was 3 mg/kg/day. Kimbrough and Linder
(1978) administered purified PeCP to rats via dietary levels of 20,
100 or 500 ppin for eight months. Only the highest exposure of 500
-------
ppm (approximately 25 mg/kg bw/day) resulted in hepatocellular
changes. Kidney weights were elevated over controls at all dose
levels including 20 ppm (approximately 1 mg/kg bw/day) , but without
a dose-related increase. Goldstein et al. (1977) reported a NOAEL
at 5 mg/kg bw/day (100 ppm in feed) to female rats over eight
months, with hepatic and body weight effects at 25 mg/kg bw/day (500
ppm in feed). Reproductive/developmental studies have reported a
very low tendency for placental transfer (Larsen et al., 1975), a
lack of teratogenic effects, and fetotoxicity at 15 or 30 mg/kg
bw/day (Schwetz et al., 1974; 1978). A statistically significant
increase in delayed skull ossification has been reported at doses as
low as 5 mg/kg bw/day (Schwetz et al. , 1974), but no effects on
reproduction, neonatal growth, survival, or development occurred at
3 mg/kg bw/day (Schwetz et al., 1978). The EPA (1985b) concluded
that 3 mg/kg bw/day could be considered a NOEL for PeCP's
fetotoxicity
The quality of the key study and supporting database are judged to
be sufficient for Tier 1 HNC development. The derivation of the HNC
from the chronic rat NOAEL at 3 mg/kg/day, with 10x uncertainty factors
for inter- and intraspecies extrapolation, is consistent with the RfD
development by EPA (1985a).
ADE - 3 mg/kg/d - 0.03 mg/kg/d
100
Where. Uncertainty Factor » 100, composed of:
lOx for intraspecies variability
10x for interspecies extrapolation
-------
Drinking Water Sources*
HNV » APE x Wh x RSC - 0.03 mg/kg/d x 70 kg
WC +
-------
Johnson, R.L., et al. 1973. Chlorinated dibenzodioxins and
pentachlorophenol. Environ. Health Perspect. 5-171-175.
Kimbrough, R.D. and R.E Linder. 1978. The effect of technical and
purified pentachlorophenol on the rat liver. Toxicol. Appl.
Pharmacol. 46-151-162.
Larsen, R.V et al. 1975. Placental transfer and teratology of
pentachlorophenol in rats Environmental Letters. 10(2)' 121-128.
Schwetz, B.A. et al. 1974 The effect of purified and commercial grade
pentachlorophenol on rat embryonal and fetal development Toxicol.
Appl. Pharmacol. 28(1)- 151-161.
Schwetz, B A. et al. 1978. Results of two-year toxicity and
reproduction studies on pentachlorophenol in rats. In: Rao, K.R.
(Ed.). 1978. Pentachlorophenol Chemistry, Pharmacology, and
Environmental Toxicology. Plenum Press, New York, NY. pp. 301-309.
U.S. Environmental Protection Agency (EPA). 1985a Integrated Risk
Information System (IRIS database). Chemical file for
Pentachlorophenol (87-86-5) Verification Date 5/20/85. Last
Revised 5/7/91.
U.S. Environmental Protection Agency (EPA). 1985b. Drinking Water
Criteria Document for Pentachlorophenol. Prepared for the Office of
Drinking Water, by the Environmental Criteria and Assessment Office.
EPA-600/X-84-177-1. PB-86-118015.
-------
2. Tier 1 Human Cancer Criterion
A review of the available literature on the carcinogenicity of
pentachlorphenol (PeCP) indicates a lack of evidence of human or
animal carcinogenicity prior to NTP (1989). As pointed out in NTP
(1989), prior studies were generally limited by study design flaws.
EPA has concluded that the evidence provided by NTP (1989) would
support classification of PeCP into Group B2; probable human
carcinogen (EPA, 1989; 1990). The conclusions of NTP (1989) are that
under the conditions of the 2-year feeding studies, there was clear
evidence of carcinogenic activity for male and female B6C3F1 mice
exposed to PeCP as Dowicide EC-7, and for male mice exposed to PeCP
as a technical grade composite. Additionally, there was some
evidence of carcinogenic activity for female mice exposed to the PeCP
technical-grade composite. The database is judged to be sufficient
for Tier 1 HCC derivation.
The study des: gn of NTP (1989) involved exposure via feed for 103
weeks to groups of 50 mice/sex, with groups of 35/sex serving as
controls for each grade of PeCP studied. The grades of PeCP selected
for the chronic bioassays were commercial samples (Dowicide EC-7 and
a technical-grade composite), considered to be representative of PeCP
forms which humans are exposed to. Exposure levels in feed were 0,
100 or 200 ppm technical-grade PeCP or 0, 100, 200 or 600 ppm
Dowicide EC-7. There was clear evidence that administration of the
technical-grade composite resulted in increased incidences of
hepatocellular and adrenal medullary neoplasms in male mice, and
some evidence of induction of hepatocellular tumors and
-------
hemangiosarcomas in female mice. There was clear evidence that the
administration of Dowicide EC-7 resulted in hepatocellular tumors
and adrenal medullary pheochromocytomas in both sexes and
hemangiosarcomas in female mice.
Among the three types of tumors Induced by pentachlorophenol, the
EPA Science Advisory Board considered the hemangiosarcomas to be the
tumor of greatest concern (EPA, 1990). To give preference to the
hemang10sarcoma data and because some male groups experienced
significant early loss, only the female mice are used in the
quantitative risk assessment. Pooled tumor incidence was utilized,
with exclusion of animals which died prior to the first tumor
observation The slope factor of 1.2 E-l per (mg/kg)/day is
calculated as the geometric mean of the slope factors for each
pentachlorophenol preparation. This approach is consistent with EPA
(1990).
RAD = 1 x 10~5 - 8.33 x 10"5 mg/kg/d
1.2 E-l (mg/kg/d)"1
83.3 ng/kg/d
Drinking Water Sources:
HCV = RAD x Wh - 83.3 ng/kg/d x 70 kg
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 755 I/kg*)
- 438 ng/1 (rounded off to 0.4 ug/1 (Tier 1))
-------
Nondrinking Water Sources*
HCV = RAD x Wh « 83.3 ng/kg/d x 70 kg
WC + (RC x BAF) 0.01 1/d + (0.015 kg/d x 755 I/kg*)
= 514 ng/1 (rounded off to 0.5 ug/1 (Tier 1))
Where: *BAF - 755, provided by EPA-Duluth and Minnesota PCA.
REFERENCES'
NTP. 1989. Toxicology and Carcinogenesis Studies of Two
Pentachlorophenol Technical-Grade Mixtures in B6C3F1 Mice (Feed
Studies). U S. DHHS. Technical Report Series No. 349.
U.S. Environmental Protection Agencv (EPA) 1989 54 FR 22062-22160.
National Primary and Secondarv Drinking Water Regulations; Proposed
Rule. May 22, 1989.
U.S. Environmental Protection Agency (EPA) 1990. Integrated Risk
Information System (IRIS database). Chemical file for
Pentachlorophenol (87-86-5). Verification Date 8/2/90. Last
Revised 5/7/9'.
-------
DRAFT
September 6, L991
Great Lakes Initiative
Human Health Criteria for
Lindane
(gamma-Hexachlorocyclohexane)
CAS No. 58-89-9
Tier 1 Human Noncancer Criterion
A review of the available literature Indicates that the most
appropriate study for the derivation of the HNV for lindane is a
subchronic study conducted by Zoecon Corporation (1983) as evaluated
by EPA (1991) and summarized by EPA (1986). In this study, Wistar
KFM-Ham (outbred) SPF rats (20/sex/dose) were administered 0, 0.2,
0.8, 4, 20 or 100 ppm lindane in the feed. Fifteen animals/sex/group
were sacrificed after 12 weeks. The remaining rats were fed the
control diet for an additional six weeks before sacrifice. Rats
exposed to 20 and 100 ppm lindane had a greater incidence of liver
hypertrophy, kidney tubular degeneration, hyaline droplets, tubular
distension, interstitial nephritis and basophilic tubules than did
the controls. The NOAEL for this study was 4 ppm. This dose was
estimated to be equivalent to 0.29 mg/kg/d for the male and 0.33
mg/kg/d for the female rats.
Two chronic studies which examined the effects of lindane on rats
and dogs were cited by EPA (1986). A two-year study by Fitzhugh
-------
(1950) reported a NOAEL of 2.5 mg/kg/d in Wistar rats with liver
weights and liver damage evaluated as the endpoints. In a two-vear
study in beagle dogs, Rivett et al. (1978) reported a NOAEL
of 1.6 mg/kg/d for liver toxicity.
A review of the database on developmental and reproductive effects
of lindane suggests that these effects may occur at levels higher
than the NOAE1 calculated in the study conducted by Zoecon
Corporation (1983). Palmer et al. (1978a) found no adverse effects
on reproductive function and development following exposure of
female rats to lindane in the feed at levels of 1.25, 2.5 and 5
mg/kg/d for three generations. Khera et al. (1979) found no
reproductive effects in Wistar rats exposed to lindane at levels
ranging from 6.25 to 25 mg/kg from the 6th to the 15th day of
gestation. No adverse effects were found in a teratogenicity study
on pregnant rabbits fed lindane on gestation days 6-18 at levels of
5, 10 and 15 mg/kg (Palmer et al., 1978b). However, Sircar and
Lahiri (1989) reported that even the lowest exposure group (3.75
mg/kg/d) of Srfiss mice receiving lindane during gestation
experienced reproductive failure.
The quality of the study conducted by Zoecon Corporation (1983) was
deemed sufficient to derive a Tier 1 HNC. The results of studies
which examine the reproductive or developmental effects of lindane
are either negative or indicative of possible effects at doses
substantially higher than the NOAEL reported by Zoecon Corporation
(1983). Although subchronic in duration (12 weeks), the key study
is supported by chronic studies in the database This study was
-------
also used by EPA (1986) to derive the oral RfD for lindane. The
HNC was derived from the female rat NOAEL (0.33 mg/kg/d) using an
uncertainty factor of 1000 to account for intraspecies variability,
interspecies extrapolation and the extrapolation from a subchronic
to chronic study. The magnitude of this uncertainty factor is
expected to result in adequate protection from any potential
reproductive or developmental effects as well as chronic noncancer
effects.
ADE - Q 33 mg/kg/d - 0.00033 mg/kg/d - 0.33 ug/kg/d
1000
Where: Uncertainty Factor - 1000, composed of*
lOx for subchronic to chronic extrapolation
lOx for intraspecies variability
lOx for interspecies extrapolation
Drinking Water Sources:
HNV - ADE x Wh x RSC - 0.33 ug/kg/d x 70 kg x 0.8
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 1,954 I/kg*)
- 0.59 ug/1 (rounded off to 0.6 ug/1 (Tier 1))
-------
Nondrinking Water Sources'
HNV - APE x Wh x RSC * 0.33 ug/kg/d x 70 kg x 0.8
WC -I- (FC x BAF) 0.01 1/d + (0.015 kg/d 1,954 I/kg*)
- 0.63 ug/1 (rounded off to 0.6 ug/1 (Tier 1))
Where: RSC - 0.8; the substance is bioaccuaulative.
*BAF - 1,954, provided by EPA-Duluth and Minnesota PCA.
REFERENCES
Fitzhugh, O.G., A.A. Nelson and J. P, Frawley. 1950. The chronic
toxicities of technical benzene hexachloride and its alpha, beta and
gamma isomers. J. Phann. Exp. Ther. 100:59-66.
Khera, K.S., C. Whalen, G. Trivett and G. Angers. 1979. Teratogenicity
studies on pesticidal formulations of dimethoate, diuron and lindane
in rats. Bull. Environ. Contain. Toxicol. 22(4-5):522-529.
Palmer, A.K., D.D. Cozens, E.J.F. Spicer and A.N. Worden. 1978a.
Effects of lindane upon reproductive function in a 3-generation
study in rats. lO(l):45-54.
Palmer, A.K., A.M. Bottomley, A.N. Worden, H. Frohberg and A. Bauer.
1978b. Effect of lindane on pregnancy in the rabbit and rat.
Toxicol. 9(3):239-247.
-------
Rivett, K.F., H Chesterman, D.N. Kellett, A J Newman and A N. Worden
1978. Effects of feeding llndane to dogs for periods of up to two
years. Toxicol 9-273-289.
Sircar, S. and P. Lahiri. 1989. Lindane (gamma-HCH) causes
reproductive failure and fetotoxicity in mice. Toxicol. 59-171-177.
U.S. Environmental Protection Agency (EPA). 1986. Integrated Risk
Information System (IRIS database). Chemical file for lindane
(58-89-9), Verification Date 1/22/86. Last Revised 3/1/88
U.S. Environmental Protection Agency (EPA) 1991 Data Evaluation
Record (DER) for lindane Office of Pesticide Programs.
Zoecon Corporation. 1983 Unpublished report. MRID No. 00128356.
Available from EPA. Write to FOI, EPA, Washington, B.C. 20460.
2. Tier 2 Human Cancer level of protection
There are inadequate data available to ascertain whether lindane
is a human carcinogen (IARC, 1982; EPA, 1985; ATSDR, 1989). The
preponderance of evidence indicates that lindane is carcinogenic to
mice (EPA, 1985). Animal bioassays conducted by Thorpe and Walker
(1973), NCI (1977), Goto et al. (1972) and Hanada et al. (1973)
provide evidence that lindane induces liver tumors. Possibl
-------
relevance to humans is Indicated by the occurrence of a carcinogenic
metabolite (2,4,6-trichlorophenol) in humans and other species
following exposure to lindane (EPA, 1980; 1985; ATSDR, 1989). The
weight-of-evidence for lindane carcinogenicity is reportedly
sufficient for "B2-C" classification (EPA, 1985; 1991). The
database was net judged sufficient for B2 (probable human
carcinogen) classification because of limitations in the quality of
the bioassay data and also because the results of most mutagenicity
tests have been negative (EPA, 1985). According to EPA (1985), "the
weight-of-evidence appears to be closer to a Category C carcinogen
than to Categoiy B2 carcinogen". The cancer risk assessment for
lindane is under review by EPA (EPA, 1990). For this initiative,
the data are sufficient to derive a Tier 2 level of protection
because liver tumor induction has been found in more than one
bioassay, in both sexes and in multiple strains of mice. A Tier 1
criterion is not Indicated, due to the limited quality of the
bioassay data, the lack of mutagenicity data, and the pending
assessment by EPA.
The Thorpe and Walker (1973) study was used to determine the Tier 2
level of protection because it provides the best quality data for
quantitative zisk assessment. In this study, 30 CF1 mice of each
sex were exposed to 400 ppm lindane in their diet for up to 110
weeks. The pooled control group consisted of 45 animals of each
sex. Besides having only one exposure level, the quality of the
study was compromised because a low percentage of treated mice
survived (3% of females and 17Z of males) to the end of the study
(EPA 1980; 1935). There was a significant increase in the incidence
-------
of hepatic neoplasms in treated male and female mice. Liver
neoplasms were found in 27/28 (96%) treated male mice and 20/21
(95%) treated female mice as compared to 11/45 (24%) and 10/44 (23%)
liver neoplasms in male and female controls, respectively The male
mouse data were used to calculate a slope factor of 1.3 (mg/kg/d)~
via the linearized multistage model Global 82.
RAD - 1 x 10~5 - 7.6 x 10~6 mg/kg/d - 7.6 ng/kg/d
1.3 (mg/kg/d)"1
Drinking Water Sources:
HCV - RAD x Wh - 7.6 ng/kg/d x 70 kg
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 1,954 I/kg*)
- 17 ng/1 (rounded off to 20 ng/1 (Tier 2))
Nondrinklng Water Sources:
HCV - RAD x Wh - 7.6 ng/kg/d x 70 kg
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 1,954 I/kg*)
- 18.1 ng/1 (rounded off to 20 ng/1 (Tier 2))
*BAF - 1,954, provided by EPA-Duluth and Minnesota PCA.
-------
REFERENCES
Agency for Toxic Substances and Disease Registry (ATSDR). 1989.
Toxicological Profile for Alpha-, Beta-, Gamma- and
Delta-Hexachlorocyclohexane. U.S. Public Health Service.
Goto, M., M. Hattori and T. Mizagawa. 1972. Contributions to ecology. II,
Hepatoma development in mice after administration of HCH isomers in
high dosages. Chemosphere. 1*279-282.
Hanada, M., E. Kawano, S. Kawamura and M. Shiro. 1981. Radiation and
photo-induced degradation of five isomers of 1,2,3,4,5,6-hexachloro-
cyclohexane. Agric. Biol. Chem. 45(3):659-665.
International Agency for Research on Cancer (IARC). 1982. IARC
Monographs on the Evaluation of the Carcinogenic Risk of Chemicals
to Humans. Suppl. 4:133-135.
National Cancer Institute (NCI). 1977. Bioassay of Lindane for
Possible Carcinogeniclty. NCI Carcinogenesls Tech. Rep.
S«r. No. 14. 99 p. NTIS PB-273-480.
Thorpe, E. and A.I. Walker. 1973. The toxicology of dieldrin (HEOD). II.
Comparative long-tern and toxicity studies in nice with dieldrin,
DDT, phenobaibitone, Betz-BHC and gamma-BHC. Food Cosmet. Toxicol.
11:433-442.
-------
U.S. Environmental Protection Agency (EPA) 1980. Ambient Water
Quality Criteria for Hexachlorocyclohexane. Criteria and Standards
Office. Washington, DC. EPA 444/5-80-054.
U.S. Environmental Protection Agency (EPA). 1985. Drinking Water
Criteria Document for Lindane. Prepared by the Office of Health and
Environmental Assessment. Environmental Criteria and Assessment
Office, Cincinnati, OH for the Office of Drinking Water, Washington,
D.C.
U.S. Environmental Protection Agency (EPA). 1990. Integrated Risk
Information System (IRIS database). Chemical file for lindane
(58-89-9). Last Revised 8/1/90.
U.S Environmental Protection Agency (EPA). 1991. Health Effects
Assessment Summary Tables.
-------
DRAFT
September 6, 1991
Great Lakes Initiative
Human Health Criteria for
Polychlorinated Biphenyls (PCBs)
CAS No. 1336-36-3
Tier 2 Human ^oncancer Level of Protection
Studies of low-level oral PCB exposure in several species have
demonstrated effects on serum chemistry, liver toxicity,
reproductive capability and other endpoints at doses of less than 5
mg/kg bw/day. The most appropriate data for use in HNV development
are the rhesus monkev data due to the high sensitivity of the
species and the relative wealth of the database including studies on
reproduction and development. Also, as a nonhuman primate the
rhesus monkey mav serve as the most appropriate model species for
potential human effects.
Adult male and female rhesus monkeys were administered Aroclor 1248
in the diet at levels of 2.5 and 5.0 ppm for up to 18 months (Allen,
1975; Allen and Barsotti, 1976; Barsotti et al., 1976; Allen et al.,
1980). Assuming that rhesus monkeys consume daily an amount of
food equivalent to 4* of their body weight, these exposure levels of
2.5 and 5.0 ppm are equivalent to 0.1 and 0.2 mg/kg bw/day,
respectively (EPA, 1985). After six months exposure, the females
were bred with unexposed males. Conception occurred in 12/12, 8/8
-------
and 6/8 of the adult females receiving 0, 2.5 and 5.0 ppm,
respectively. The number of live infants born at 0, 2.5 and 5.0 ppm
was 12, 5 and 1, respectively. Exposure of the females continued
until three months after parturition. Infants were allowed to
remain with their mothers and nurse for a minimum of four months,
resulting in transmammary as well as transplacental exposure. At
both dose levels (dose-specific responses not distinguished), adult
females developed acne, alopecia (hair loss), erythema, swelling of
the evelids, abnormal menstrual cycles, and abnormal serum
chemistries Two treated females that died after 173 days (2.5 ppm
group) or 310 days (5 ppm group) were found at necropsy to have
signs of liver toxicity including focal areas of necrosis. Treated
animals were noted to appear more susceptible to the opportunistic
intestinal pathogen Shigella flexneri type IV. Offspring of the
exposed groups (combined) demonstrated decreased birth weights (399
± 22 g, vs. 507 t 59 g. in controls) and hyperpigmentation of the
skin. Three of the 6 infants died, which was attributed to PCB
toxicity. These included the only infant from the 5 ppm group and
two of the five infants from the 2.5 ppm group. Necropsies of these
infants showed rudimentary thymuses, small spleens, underdeveloped
splenic lymph nodes, hypocellularity of the bone marrow and fatty
infiltration of liver cells, among other effects. In a continuation
of these studies, Barsotti (1980) and Allen et al. (1980) reported
that in subsequent breeding trials during the recovery period for
these same adult female monkeys, effects on reproduction and
offspring development were still apparent for greater than one year.
-------
Barsotti (1980) and Barsotti and VanMiller (1984) administered
diets containing 0, 0.25 or 1.0 ppm Aroclor 1016 to groups of 8
adult female ihesus monkeys for seven months prior to breeding,
through gestation and a A-month nursing period. The total exposure
period was 87 ± 9 weeks. Exposure did not result in signs of overt
toxicity in adult females. All treated females conceived, carried
their fetuses to term and delivered viable offspring. The birth
weights from the control, 0.25 and 1.0 ppm groups were 512 ± 64,
491 ± 24 and 422 ± 29 g., respectively. The newborn weights in the
1.0 ppm group were significantly less than the controls (p less
than 0.01). Experimental groups of Infants gained weight
consistently, and the infant weights among the 1.0 ppm group were
not significantly lower than the control group at weaning (864 ± 97g
vs. 896 ± 90g). The authors determined that during the
pre-breeding exposure period, the 0.25 and 1 0 ppc treated females
consumed 1.7 i 0.3 mg/kg bw/7 months and 6.1 ± 0.9 mg/kg bw/7
months, or approximately 0.008 and 0.03 mg/kg bw/dav, respectively.
Bowman et al (1981) reported that offspring of female rhesus
monkeys fed diets containing 0.5 or 1.0 ppm Aroclor 1248 three days
per week or 1.5 ppm daily (estimated average doses were 0.006, 0.013
and 0.085 mg/kg bw/day, respectively) displayed greater locomotor
activity than controls. However, group sizes were small (n « 3-7),
the quantitative differences in activity were not dose-related, and
the variability within each group was noted to be substantial.
Becker et al. (1979) found that groups of 1-2 rhesus monkeys fed
diets containing 3, 10, 30 or 100 ppm of PCBs as Aroclor 1242 for
several months had dose-dependent findings of gastric lesions, lack
41 *f
-------
of body weight gain, reduced hemoglobin, persistent leukocytosis,
and early mortality. These effects occurred even in the animal
receiving the lowest dose of 3 ppm (0.12 mg/kg bw/day), which expired
after 245 days of dosing.
The substantial studies of Aroclor 1248 in rhesus monkeys establish
a subchronic LOAEL for marked systemic toxicity, reproductive and
developmental effects at 2.5 ppm (0.1 mg/kg bw/day). One study
utilizing Aroclor 1016 indicates a LOAEL for neonatal weight
depression at 1 0 ppm (0.03 mg/kg bw/day) with a NOAEL for this
effect at 0.25 ppm (0 008 mg/kg bw/day) (Barsotti, 1980, Barsotti
and \anMiller, 1984). None of these studies were chronic in
duration, generally spanning less than 10% of the expected lifespan
of about 20 years (Gold et al., 1984). The composition of Aroclor
1016 is primarily di-, tri-, and tetrachloro isomers of biphenyl,
with an average chlorine percentage (41%) that is very similar to
Aroclor 1242 (EPA, 1980) Other Aroclor mixtures composed of more
highly chlorinated congeners have inadequate data to identify the
approximate threshold level for the sensitive systemic and
reproductive/developmental effects. EPA (1985) considered the poor
metabolism of PCBs and their bioaccumulation tendency, and the
severity of effects seen at the subchronic LOAEL of 0.1 mg Aroclor
1248/kg bw/day, and declined recommendation of an Acceptable Daily
Intake (ADI).
The database is judged insufficient for Tier 1 Human Noncancer
Criterion development. A Tier 2 level of protection is derived from
the Aroclor 1016 NOAEL in rhesus monkeys at 0.25 ppm (approximately
-------
0.008 mg/kg bv/day). The Tier 2 level of protection is intended to
be applicable to all PCS isomers and Aroclor mixtures (i.e., total
PCBs) until a more appropriate methodology may be developed. In
particular, the dose-response for chronic and reproductive/
developmental effects of the more highly chlorinated PCBs needs to
be investigated and characterized. A total uncertainty factor of
1,000 is used in the calculation:
0.008 mg/kg/d <* 8 x 10~6 mg/kg/d - 8 ng/kg/d
1,000
Where: Uncertainty factor • 1,000 composed of:
lOx for intraspecies variability
lOx for interspecies extrapolation
lOx for subchronic exposure duration
Drinking Water Sources
HNV = APE x Wh x RSC - 8 ng/kg/d x 70 kg x 0 8
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 2,132,232 I/kg*)
- 0.014 ng/1 (rounded off to 10 pg/1 (Tier 2))
Nondrinking Water Sources.
HNV - APE x Wh x RSC - 8 ng/kg/d x 70 kg x 0.8
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 2,132,232 I/kg*)
- 0.014 ng/1 (rounded off to 10 pg/1 (Tier 2))
-------
Where: RSC - 0.8; the substance is persistent and bioaccumulative,
*BAF - 2,132,232, provided by EPA-Duluth and Minnesota PCA.
References*
Allen, J.R. 1975. Response of the non-human primate to polychlorinated
biphenyl exposure. Fed. Proc. 34: 1675-1679
Allen, J R. and D A. Barsotti. 1976. The effects of transplacental and
mammary movement of PCBs on infant rhesus monkeys. Toxicology.
6 331-340.
Allen, J.R., D.A. Barsotti and L.A. Carstens. 1980. Residual effects
of polychlorinated biphenyls on adult nonhuman primates and their
offspring. J. Toxicol. Environ. Health. 6(1)'55-66.
Barsotti, D.A., R J. Marlar and J.R. Allen. 1976 Reproductive
dysfunction in rhesus monkeys exposed to low levels of polychlorinated
biphenyls (Aroclor 1248). Fd. Cosmet. Toxicol. 14:99-103.
Barsotti, D.A. 1980. Gross, clinical and reproductive effects of
polvchlorinated biphenyls (PCBs) in the rhesus monkey. Diss. Abstr.
Int. 41(10):3744-5.
ill
-------
Barsotti, D.A. and J.P. VanMillcr. 1984, Accumulation of a commercial
polychlorinated biphenyl mixture (Aroclor 1016) in adult rhesus
monkeys and their nursing infants. Toxicology. 30(l):31-44.
Becker, G.M., W.P. McNulty and M. Bell. 1979. Polychlorinated biphenyl
induced morphologic changes in the gastric mucosa of the rhesus
monkey. Lab. Invest. 40(3):373-383.
Bowman, R.E., M.P. Heironimus and D.A. Barsotti. 1981. Locomotor
hyperactivity in PCB-exposed rhesus monkeys. Neurotoxicology.
2(2)-251-68.
Gold, L.S , et al. 1984. A carcinogenic potency database of the
standardized lesults of animal bioassavs. Environmental Health
Perspectives. 58*9-319.
U.S. Environmental Protection Agency (EPA). 1985. Drinking Water
Criteria Document for Polvchlorinated Biphenyls (PCBs).
Environmental Criteria and Assessment Office. EPA-600/X-84-198-1.
PB-86-118312.
2. Tier 1 Human Cancer Criterion
PCBs (as a class) have sufficient carcinogeniclty weight-of-evidence
for a B2 classification (probable human carcinogen) based on the
induction of hepatocellular carcinomas in three strains of rats and
two strains of mice and inadequate yet suggestive evidence of excess
-------
risk of liver cancer in humans (EPA, 1987). The data are judged
sufficient for Tier 1 HCC derivation. Although animal feeding
studies demonstrate the carcinogenicity of commercial PCB
preparations, it is not known which of the PCB congeners in such
mixtures are responsible for these effects. EPA (1987) developed a
carcinogenicitv risk assessment for PCBs with a slope factor derived
from Aroclor 1260 data, clearly stating the intent that the
assessment be considered representative for all PCB mixtures. The
application of this approach to regulatory programs is a prudent
approach to ensure adequate protection of public health.
A review of the available carcinogenicity data indicates that the
most appropriate studies for quantitative cancer risk assessment are
the bioassays of Kimbrough et al (1975) and Norback and Weltman
(1985). These studies utilized different rat strains — Sherman
rats in the Kimbrough et al (1975) study, Sprague-Dawley rats in the
Norback and Weltman (1985) study — but otherwise had several
similarities. Both utilized large numbers of animals in chronic
Aroclor 1260 feeding studies with only one exposure group. Dosed
groups received 100 ppm for 630 days in the bioassay by Kimbrough
et al. (1975), while Norback and Weltman (1985) administered 100
ppm for 16 months followed by a 50 ppm diet for an additional 8
months, then a basal diet for 5 months. The predominant neoplastic
effect in each study was the increased incidence of hepatocellular
neoplasms in female rats.
-------
Using the linearized multistage procedure, EPA (1987) estimated
slope factors of 7.7 (mg/kg/d)" and 3 9 (mg/kg/d)~ from the data
of Norfaack and Weltman (1985) and Kimbrough et al. (1975),
respectively. The larger of these slope factors, 7.7 (mg/kg/d)~ ,
was selected by EPA (1987) as the preferred slope factor estimate.
Although the Norback and Weltman (1985) study included a test
protocol of partially hepatectomizlng some of the animals, EPA
(1987) noted that the study had favorable qualities. The rat strain
used (Sprague-Dawley) is known to have a low incidence of
spontaneous hepatocellular neoplasms, the study duration spanned the
natural life of the animal, and concurrent morphologic liver studies
showed the sequential progression of liver lesions to hepatocellular
carcinomas. Extrapolation modeling utilized a female rat liver
tumor incidence rate of 45/47 in the dosed group. This includes 7
animals which had earlier undergone partial hepatectomy, and the
liver tumor incidence for this subgroup was unreported. Exclusion
of this group would have verv little impact on the resulting slope
factor, and the tumor promoting effect of the partial hepatectomization
should be minimal (Riremath, 1991). The Tier 1 Human Cancer
Criterion for PCBs is based on the slope factor of 7.7 (mg/kg/d)~
derived from the rat bloassay of Norback and Weltman (1985).
RAD - 1 x LO"3 - 1.3 x 10"6 mg/kg/d
7.7 (mg/kg/d)'1
1.3 n,g/kg/d
-------
Drinking Water Sources'
HCV - RAD x Wh - 1.3 ng/kg/d x 70 kg
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 2,132,232 I/kg*)
2.8 x 10"3 ng/1 (rounded off to 3 pg/1 (Tier 1))
Nondrinking Water Sources
HCV - RAD x Wh - 1.3 ng/kg/d x 70 kg
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 2,132,232 I/kg*)
2 8 x 10~3 ng/1 (rounded off to 3 pg/1 (Tier 1))
Where- *BAF - 2,132,232, provided by EPA-Duluth and Minnesota FCA.
References'
Hiremath, C. 1991. lexicologist, U.S. EPA Office of Research and
Development. Personal communication with R. Sills, Michigan
Department of Natural Resources.
Kimbrough, R.D. et al. 1975. Induction of liver tumors in Sherman
strain female rats by Aroclor 1260. J. National Cancer Institute,
55(6):1453.
-------
Norback. D. and R.N. Weltman. 1985. Polychlorinated biphenyl induction
of hepatocellular carcinomas in the Sprague-Dawley rat. Env. Health
Persp. 60:97-105.
U.S. Environmental Protection Agency (EPA). 1987. Integrated Risk
Information System (IRIS database). Chemical file for
polychlorinatcd biphenyls (PCBs) (1336-36-3). Verification
Date 4/22/87. Last Revised 1/1/90.
-------
DRAFT
September 6, 1991
Great Lakes Initiative
Tier 1 Human Health Criteria for
2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD).
CAS No. 1746-01-6
1. Tier 1 Human Noncancer Criterion
Of the many subacute and chronic studies available for 2,3,7,8-TCDD,
a few stand out as supporting Tier 1 criterion derivation. In a
two-year toxicity and oncogenicity study, rats were administered
doses of 0, 0.001, 0.01 and 0.1 ug/kg bw/day of 2,3,7,8-TCDD via
diet (Kociba et al., 1978). Animals given the high dose exhibited
increased mortality, decreased weight gain, slight depression of
erythroid parameters, increased urinary excretion of porphvrins and
delta-aminolevulinic acid and increased serum levels of certain
enzymes. Histopathologic or gross effects were seen in liver,
lymphoid, lung and vascular tissues. An increased tumor incidence
was also seen. Similar effects, but to a lesser degree, were seen
in mid-dose animals. A NOAEL of 0.001 ug/kg/day (1 ng/kg/day) was
reported in this study.
A NOAEL of 0 001 ug/kg bw/day via feed exposure was also reported in
a three-generation rat reproduction study (Murray et al., 1979). At
0.1 ug/kg/day, decreases in F. generation fertility and F. generation
-------
litter size ware reported. At 0.01 ug/kg/day, significant decreases
in fertility were seen in the F and F_ generations; other effects
included decreased litter size at birth, decreased gestational
survival and decreased neonatal growth and survival. The
reproductive capacity of the low dose rats did not appear to be
significantly affected in any generation. However, a reevaluation
of these data using different statistical methods indicated that
both lower dose levels resulted in significant reductions in
offspring survival indices, increases in liver and kidney weight of
pups, decreased thvmus weight of pups, decreased neonatal weights
and increased incidence of dilated renal pelvis (Nisbet and Paxton,
1982). Nisbet and Paxton (1982) concluded that 0 001 ug/kg/day (1
ng/kg/day) was not a NOEL in the Murray et al. (1979) study. Kimmel
(1988) consicered the data of Murray et al. (1979) to be suggestive
of a pattern of decreased offspring survival and increased offspring
renal pathology even at 0.001 ug/kg/day, although the pooling of
data from dilferent generations by Nisbet and Paxton (1982) was
considered biologically inappropriate.
Studies by Schantz et al. (1979) and Allen et al. (1979) suggest
that rhesus monkeys are more sensitive to 2,3,7,8-TCDD than rats.
When monkeys were administered 50 ppt 2,3,7,8-TCDD in feed for 7 to
20 months, decreases in fertility, increases in abortions and other
toxic effects (alopecia, hyperkeratosis, weight loss, decreased
hematocrit and white blood cell count and increased serum levels of
SGPT) were noted. The 50 ppt dietary residue level corresponds to a
daily dose cf 1.5 ng/kg bw/day (EPA, 1984) Therefore, 1.5 ng/kg/day
can be considered a LOAEL for rhesus monkevs from these studies.
-------
In a continuation of the rhesus monkey studies by Schantz et al.
(1979) and Allen et al. (1979), Bowman et al. (1989a, 1989b) have
evaluated the effects of 5 and 25 ppt 2,3,7,8-TCDD in feed on
reproduction and on behavior, respectively. Breeding of the animals
after 7 and 24 months of exposure resulted in impaired reproductive
success at 25 ppt but not at 5 ppt (approximately 0,67 and 0.13
ng/kg bw/day, respectively). The exposures were discontinued after
4 years, and a third breeding ten months post-exposure did not
indicate reproductive impairment (Bowman et al., 1989a). The
offspring from these breeding experiments were evaluated for
development and behavioral effects utilizing several testing methods
(Bowman et al., 1989b) Although there were no significant effects
of TCDD exposure on birth weight, growth, or physical appearance of
the offspring, some behavioral test results were interpreted to be
indicative of TCDD effects. These included alterations in the
social behavior between the mothers and their infacts and of peer
groups of the offspring after weaning. However, the study groups
were verv limited in size and the statistical and biological
significance of the findings are unclear. This study may be
interpreted to provide only suggestive evidence of possible
behavioral effects. The reproduction study of Bowman et al. (1989a)
provides much clearer evidence of a LOAEL at 25 ppt (0.67 ng/kg/day)
and a NOAEL at 5 ppt (0.13 ng/kg/day).
The EPA has used the equivocal evidence for a rat LOAEL at 1 ng/kg/day,
supported by an unequivocal rhesus monkey LOAEL at 1.5 ng/kg/day, in
the development of an Acceptable Daily Intake (ADI) (EPA, 1984;
1985a) and Drinking Water Equivalent Level (DWEL) (EPA, 1985b;
-------
1990). In light of the more recent rhesus monkey study of Bowman et
al. (1989a), there is improved resolution of the threshold for the
sensitive effect of reproductive impairment in this species. The
Human Noncancer Criterion is based on the NOAEL of 0.13 ng/kg/day
for reproductive effects from this study. The entirety of the
rhesus monkey studies, supported by the evidence in rats cited
above, is judged sufficient for Tier 1 criterion development.
ADE = 0.13 ng/kg/d - 1 3 x 10"3 ng/kg/d - 1.3 pg/kg/d
LOO
Where. Uncertainty Factor • 100, composed of'
lOx for intraspecies variability
lOx for interspecies extrapolation
Drinking Water Sources
HNV = ADE x Wh x RSC - 1.3 pg/kg/d x 70 kg x 0.8
WC + (FC x BAF) 2 1/d -I- (0 015 kg/d x 43,714 I/kg*)
- 0.11 pg/1 (rounded off to 0.1 pg/1 (Tier 1))
Non-drinking Water Sources
HNV - ADE x Wh x RSC - 1.3 pg/kg/d x 70 kg x 0.8
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 43,714 I/kg*)
- 0.1J pg/1 (rounded off to 0.1 pg/1 (Tier 1))
-------
Where: RSC » 0.8; the substance is persistent and bioaccumulative,
*BAF - A3,714, provided by EPA-Duluth and Minnesota PCA.
REFERENCES:
Allen, J.R. et al. 1979. Reproductive effects of halogenated aromatic
hydrocarbons on nonhuman primates. Ann. NY Acad. Sci. 320.419-425.
Bowman, RE., et al. 1989a. Chronic dietary intake of
2,3,7,8-tetrachlorodibenzo-p~dioxin (TCDD) at 5 or 25 parts per
trillion in the monkey. TCDD kinetics and dose-effect estimate of
reproductive toxicity. Chemosphere. 18(1-6). 243-252.
Bowman, R.E., et al. 1989b, Behavioral effects in monkeys exposed to
2,3,7,8-TCDD transmitted maternally during gestation and for four
months of nursing. Chemosphere. 18(1-6)•235-242.
Kimmel, G.L. 1988 Appendix C. Reproductive and Developmental
Toxicity of 2,3,7,8-TCDD. Reproductive Effects Assessment Group,
OHEA/ORD, EPA. In: EPA. 1988. A Cancer Risk-Specific Dose
Estimate for 2,3,7,8-TCDD. Appendicef A-F. Review Draft.
EPA/600/6-88/007Ab.
Kociba, R. J. et al. 1978 Results of a two-year chronic toxicity and
oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats.
Toxicol. Applied Pharmacol. 46:279-303.
-------
Hurray, F. J. et al. 1979. Three-generation reproduction study of rats
given 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in the diet.
Toxicol. Applied Phannacol. 50:241-252.
Nisbet, I.C.T. and M.B. Paxton. 1982. Statistical aspects of
three-generation studies of the reproductive toxicity of TCDD and
2,4,5-T. The American Statistician. 36(3)-290-298.
Schantz, S. L. et al. 1979. Toxicological effects produced in nonhuman
primates chronically exposed to 50 ppt TCDD. Toxicol Applied
Phannacol. 48:A180. (Abstract No. 360).
U.S. Environmental Protection Agency (EPA). 1984. Ambient Water
Quality Criteria for 2,3,7,8-Tetrachlorodibenzo-p-dioxin. Office of
Water Regulations and Standards. EPA 440/5-84-007.
U.S. Environmental Protection Agency (EPA). 1985a. Health Assessment
Document for Polvchlorinated Dibenzo-p-dioxins. Office of Health and
Environmental Assessment. EPA/600/8-84/014F.
U.S. Environmental Protection Agency (EPA). 1985b. Drinking Water
Criteria Document for 2,3,7,8-Tetrachlorodibenzo-p-dioxin. ECAO/ODW.
EPA-600/X-84-194-1. PB 86-117983
U.S. Environmental Protection Agency (EPA). 1990. 55 Federal Register
No. 143. Wednesday, July 25, 1990 National Primary and Secondary
Drinking Water Regulations; Synthetic Organic Chemicals and Inorganic
Chemicals Proposed rule.
-------
2. Tier 1 Human Cancer Criterion
The EPA (1984) evaluated the available epidemiological and animal
bioassay data on the potential carcinogenicity of 2,3,7,8-TCDD They
determined that some case-control studies provide limited evidence
for the human carcinogenicity of phenoxy acids and/or chlorophenols,
which contain impurities including 2,3,7,8-TCDD. They concluded
that the evidence for the human carcinogenicity of 2,3,7,8-TCDD
based on the epidemiologic studies is only suggestive due to the
difficulty of evaluating the risk of 2,3,7,8-TCDD exposure in the
presence of the confounding effects of phenoxy acids and/or
chlorophenol. Recently published epidemiology studies may be
interpreted to provide suggestive evidence of carcinogenicity (Zober
et al., 1990; Fingerhut et al., 1991) The potential use of these
new studies for quantitative risk assessment has not yet been fully
explored. With regard to animal bioassavs, the EPA (1984)
concluded that several rodent studies establish that 2,3,7,8-TCDD is
an animal carcinogen in multiple species and organs and is probably
carcinogenic in humans. The weight of evidence of carcinogenicity
is sufficient for Group B2 classification (probable human carcinogen)
and satisfies the database requirements for Tier 1 criterion
derivation.
Among the carcinogenicity bioassays, NTP conducted bioassays with
both Osborne-Mendel rats and B6C3F1 mice (NTP, 1982a). Groups of 50
mice and 50 rats of each sex were given 2,3,7,8-TCDD in corn
oil-acetone by gavage twice per week for 104 weeks. Doses of 0,
0.01, 0.05 or 0.5 ug/kg/week were administered to rats and male mice
-------
while female mice received 0, 0.04, 0.2 or 2.0 ug/kg/week. Controls
consisted of /5 rats and 75 mice of each sex. Animals were killed
at weeks 105-107. 2,3,7,8-TCDD caused an increased, dose-related
incidence of i ollicular-cell adenomas or carcinomas of the thyroid
in male rats. A significant increase in subcutaneous tissue
fibromas was also seen in high-dose males. High-dose female rats
exhibited increased incidences of hepatocellular carcinomas and
neoplastic nodules, subcutaneous tissue f ibrosarcomas and adrenal
cortical adenomas. In male and female mice, 2,3,7,8-TCDD induced an
increased dose-related incidence of hepatocellular carcinomas.
High-dose female mice also exhibited increased incidences of thyroid
follicular-cell adenomas.
In a dermal study also conducted under contract for NTP (NTP,
1982b), 30 male and 30 female Swiss Webster mice were treated with
2,3,7,8-TCDD in acetone for 3 days/week for 104 weeks. Doses of
0.005 ug and 0 001 ug 2,3,7,8-TCDD were administered to the clipped
backs of males and females, respectively. A similar group was
pretreated with one application of 50 ug dimethylbenzanthracene
(DMBA) one week before 2,3,7,8-TCDD administration. 2,3,7,8-TCDD
induced a statistically significant increase of f ibrosarcomas in the
integumentary system of females given both 2,3,7,8-TCDD alone and
following a single application of DMBA.
Van Miller et al. (1977) administered diets containing 0, 0.001,
0.005, 0.05, 1, 50, 500 and 1000 ppb 2,3,7,8-TCDD to groups of 10
male Sprague-Dawley rats Animals received the diets for 78 weeks
and were then placed on control feed until they were killed at week
95. All rats fed the higher concentrations (1-1,000 ppb) died
-------
early. A variety of tumors were produced and the total number of
animals with tumors generally increased, but the small number of
animals limits the value of the data.
Kociba et al. (1978) administered 2,3,7,8-TCDD via the diet to
groups of 50 male and 50 female Sprague-Dawley rats for 2 years.
Control groups consisted of 86 animals of each sex. The doses
administered were 0, 0.001, 0.01 and 0 1 ug/kg/day. 2,3,7,8-TCDD
induced an increased incidence of hepatocellular carcinomas and
hepatocellular hyperplastic (neoplastic) nodules in female rats at
the two highest dose levels. The highest dose of 2,3,7,8-TCDD also
induced an increase in the incidence of stratified squamous cell
carcinomas of the hard palate and/or nasal turbinates in both males
and females, squamous cell carcinomas of the tongue in males and
squamous cell carcinomas of the lungs in females.
Kociba et al. (1978) is chosen as the basis for quantitative cancer
risk assessment. The Kociba study found that the principal target
organ for 2,3,7,8-TCDD-induced tumors was the liver in female rats,
demonstrating a dose-related statistically significant increase of
hepatocellular carcinomas and hyperplastic (neoplastic) nodules. For
quantitative risk assessment, the data were adjusted for early
mortality by eliminating those animals that died during the first
year of the study. Also, in the mid-dose group, two of the reported
20 females with tumors had both nodules and carcinomas; 18 affected
animals were used as the input for the dose group. Using the
linearized multistage model, the resulting slope factor for
2,3,7,8-TCDD is 1.51 x 10 (mg/kg/dav)~ . However, an independent
-------
pathologist (Squire) was engaged by EPA to reevaluate the
histopathologlc slides from the Kociba study (EPA, 1984). Squire
reported higher tumor incidences than Kociba, generating a slightly
higher slope factor of 1.61 x 10 (mg/kg/day) . EPA (1984) used an
average of the two slope factors, 1.56 x 10 (mg/kg/day)~ , to
generate surface water criteria.
In March 1990 a panel of seven independent pathologists referred to
as the Pathology Working Group (PWG) blindly reevaluated the female
rat liver slides from Kociba et al. (1978). Liver lesions were
classified according to the National Toxicology Program's 1986 liver
tumor classification scheme (Sauer, 1990). Using the linearized
multistage model, the liver tumor Incidence rates reported by the
4 -1
PWG result in a slope factor of 5.1 x 10 (mg/kg/day) for liver
4 _i
tumors onlv, and a slope factor of 7.5 x 10 (mg/kg/day) for pooled
significantly increased tumors of the liver, lung or nasal
turbinates/hard palate. The latter method avoids double-counting of
tumor-bearing animals (Bayard, 1990).
The Human Cancer Criterion is based on the pooled significant tumors
in female rats of Kociba et al. (1978) with the liver tumor
reevaluation of the Pathology Working Group (Sauer, 1990). The
4
linearized multistage model generates a slope factor of 7.5 x 10
(mg/kg/day) from these data.
RAD - I x IP"5 _ - 1.33 x 10~10 mg/kg/d
7.5 x 104 (mg/kg/d)"1
0.133 pg/kg/d
-------
Drinking Water Sources:
HCV - RAD x Wh - 0.133 pg/kg/d x 70 kg
WC + (FC x BAF) 2 1/d + (0.015 kg/d x 43,714 I/kg*)
1.4 x 10"2 pg/1 (rounded off to 0.01 pg/1 (Tier 1))
Nondrinking Water Sources'
HCV = RAD x Wh - 0.133 pg/kg/d x 70 kg
WC + (FC x BAF) 0.01 1/d + (0.015 kg/d x 43,714 I/kg*)
1.4 x 10~2 pg/1 (rounded off to 0.01 pg/1 (Tier 1))
Where: *BAF • 43,714, provided by EPA-Duluth and Minnesota PCA.
REFERENCES:
Bayard, S. 1990. lexicologist/Statistician with the U.S. EPA Office of
Research and Development, Human Health Assessment Group. Personal
communication with R. Sills, Michigan Department of Natural
Resources.
Fingerhut, M. et al. 1991 Cancer mortality in workers exposed to
2,3,7,8-tetrachlorodibenzo-p-dioxin. The New England Journal of
Medicine. 234(4).212-218
-------
Kociba, R.J. et al. 1978. Results of a two-year chronic toxicity and
oncogenicity study of 2,3,7,8-tetrachlorodlbenzo-p-dioxin in rats.
Toxicol Applied Pharmacol. 46 279-303.
National Toxicology Program (NTP). 1982a. Bioassay of
2,3,7,8-tetrachlorodibenzo-p-dioxin in Osborne-Mendel Rats and
B6C3F1 Mice (Gavage Study). NTP-TR-209. National Toxicology
Program, U.S. DHHS, Research Triangle Park, NC.
National Toxicologv Program (NTP) 1982b Carcinogenesis Bioassay of
2,3,7,8-tetrachlorodibenzo-p-dioxin in Swiss-Webster Mice (Dermal
Study). NTP-rR-201. National Toxicology Program, U.S. DHHS,
Research Triangle Park, NC.
Sauer, R.M. 1990. Pathology Working Group: 2,3,7,8-Tetrachlorodibenzo-
p-dioxin in Sorague-Dawley Rats. Pathco, Inc. Submitted to the
Maine Scientific Advisory Panel.
U.S. Environmental Protection Agency (EPA). 1984. Ambient Water Quality
Criteria for 2,3,7,8-Tetrachlorodibenzo-p-dioxin EPA 440/5-84-007.
Van Miller, J.P. et al. 1977. Increased incidence of neoplasms in rats
exposed to low levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin.
Chemosphere 6(10)-625-632.
Zober, A., P. Messerer and P. Huber 1990. Thirty-four-year mortality
follow-up of BASF employees exposed to 2,3,7,8-TCDD after the 1953
accident. Int. Arch. Occup. Environ. Health. 62(2):139-157.
-------
Revised 9/06/91
6LWQI IMPLEMENTATION PROCEDURES
INTRODUCTION
This guidance package includes the procedures with which to
translate the water quality criteria developed under the Great
Lakes Water Quality Initiative into specific controls for sources
of pollution.
Key procedures have been developed to replace or supplement
portions of existing regulations. These procedures used in
conduction with the remaining State and Federal requirements will
enhance the programs that are already in place in order to meet
the mandate of the Great Lakes Critical Programs Act.
It is the intent of the GLWQI guidance to distinguish between
Tier I criteria and Tier II levels of concern. In this document,
the terms, water quality criteria or criteria, refers to both
Tier I criteria and Tier II levels of concern, unless otherwise
specified.
In addition to the main proposals which are described in this
guidance, comments will be solicited regarding alternative
proposals for Procedures 4, 7, and 8. For the purposes of
fostering Steering Committee discussion, the alternatives are
described at the end of the applicable Procedures. It is the
intention to include the alternative proposals in the preamble to
the proposed GLWQI guidance at the time of publication.
The preamble will also include the technical justification for
many of the conditions of these procedures.
-------
GLWQ] IMPLEMENTATION PROCEDURES
Table of Contents Page
1. Purpose 3
2. Applicability 3
3. Definitions 3
4. Site-Specific 7
5. Total Maximum Dally Load
Wasteload Allocations Procedures/Mixing Zones for Point
Sources 8
6. Compliance Schedules 19
7. WQBELs Below the Levels of Detection 19
8. Additivity 21
9. Loading Limits 25
10. Whole Effluent Toxicity Requirements for Point Sources 25
11. Background Concentrations of Pollutants 29
12. Reasonable Potential to Exceed WQBELs 30
13. Variances of Water Quality Standards 32
14. Environmental Fate 34
15. General 35
-------
1. PURPOSE. The purpose of this guidance in conjunction with
the antidegradation guidance is to provide procedures
whereby GLI pollutant criteria for Aquatic Life, Wildlife
and Human Health will be translated into specific controls
on sources of pollutants subject to those criteria. These
procedures shall apply to all continuous point and non point
source discharges, and nay be applied to wet weather point
source discharges on a case by case basis or at the
discretion of the State. These procedures shall apply to
both Tier I criteria and Tier II levels of concern for GLI
pollutants and may be applied to criteria for pollutants
that are not GLI pollutants. In addition, this guidance
contains definitions of terms used in Great Lakes guidance
for aquatic life criteria, human health criteria and
wildlife criteria.
2. APPLICABILITY. The criteria developed under the Great Lakes
Water Quality Initiative apply to decision-making involving
contributions of pollutants from both point and nonpoint
sources to the waters of the Great Lakes System. The
aquatic life criteria and the wildlife criteria shall apply
to all waters of the Great Lakes System. Human health -
public water supply criteria shall apply to the Open Waters
of the Great Lakes, all connecting waters of the Great
Lakes, and at all other approved points of withdrawal for
public water supplies within the Great Lakes System. Human
health - non-public water supply criteria shall apply to all
other waters. It should be noted that fish consumption is
considered for each category of human health criteria.
The criteria shall provide a basis for decision making
concerning contributions of pollutants to the Great Lakes
System, including but not limited to decisions made under
the National Pollutant Discharge Elimination System (NPDES)
program, State Implementation Plan (SIP) revisions, RCRA
corrective action activities, Remedial Action Plans (RAP)
and Lakewide Area Management Plans (LAMP) development and
implementation, and shall be considered applicable or
relevant and appropriate requirements (ARAR'S) under
proposed CERCLA activities.
3. DEFINITIONS
Acceptable Daily Exposure (APE)t an estimate of the WMiytiwm daily dose of a
substance which is not expected to result in adverse effects to the general human
population, including sensitive subgroups.
Acute-Chronic Ratio fACRl; is the ratio of the acute toxicity of a material to its
chronic toxicity. Zt is used as a factor for estimating chronic toxicity on the
basis of acute toxicity data.
SI 7
-------
Acute Toxic Unit (TU.l; is the reciprocal of the effluent concentration that cau
50 percent of the organisms to die by the end of the acute exposure period (i.e.,
100 LC*).
Adverse Effect 8 any observed deleterious effect to organisms due to exposure to a
substance. This includes effects which are or may become debilitating, harmful or
toxic to the normal functions of the organism, but does include effects such as
tissue discoloration alone or the induction of enzymes involved in the metabolism of
the substance.
Allowable Dilution Flow fQ^>: is the portion of the stream design flow used in
developing the W1A for sources regulated by the NPOBS program. The Q^ is calculated
in Procedure 5.1.8(4).
Bioaccumulatien; is the uptake and retention of substances by an aquatic organism
from its surrounding media and food.
Bioaccumulation Factor fBAFVt is the ratio of the substance's concentration in
tissue versus its concentration in ambient water, in situations where the organism
and the food chain are exposed.
Bioaccumulative Chemical of Concern (SCO; is any chemical which, upon entering the
surface waters, by itself or as its toxic transformation product, bioaccumulates in
aquatic organisms by a factor greater than 1000. (Formerly
Bioaccumulative/Persistent Toxic Substance (B/PTS) and Persistent Bioaccumulative
Toxic Substance (PBTS))
Bioconcentration; is the uptake and retention of substances by an aquatic organism
from the surrounding water through gill membranes or other external body surfaces.
Bioconcentration Factor iBCFlt is the ratio of the substance's concentration in
tissue versus its concentration in water, in situations where the organism is
exposed through the water only.
Carcinogen; a substance which causes an increased incidence of benign or malignant
neoplasms, or substantially decreases the time to development of neoplasms, in
animals or humans.
Chronic Toxic Unit (TIL); is the reciprocal of the effluent concentration that
causes no observable effect on the test organisms by the end of the chronic exposure
period (i.e., 100/NOEC).
Chronic Toxicitv; is the causation of an adverse effect in an organism as a result
of exposure to a major portion of the organism's natural lifespan.
Criteria Continuous Concentration fCCCl; is equal to the lower of the Final Plant
Value or the Final Chronic Value. It is the highest instream concentration of a
material to which organisms can be exposed indefinitely without causing unacceptable
effect.
Criterion Maximum Concentration tCMCl; is equal to one half the FAV. It is the
highest instream concentration of a toxicant to which organisms can be exposed for a
brief period of time without causing an acute effect.
Connecting Channels of the Great Lakes: means any of the following: the Detroit
River; Lake St. clair; the Keweenaw waterway; the St. Marys River; the St. Clair
River; the Niagara River; and the St. Lawrence River upstream from the point at
which it becomes the international boundary between Canada and the United States.
Depuration; is the loss of a substance from an aquatic organism.
-------
Detection Level (PL): detection level, for the purposes of Procedure 7 of this
guidance, is defined as the minimum level, the level at which the entire analytical
system gives recognizable spectra and acceptable calibration points. The minimum
level is developed baaed on inter-laboratory analyses of the analyte in the matrix
of concern.
Dilution fraction; is a number used in determining the allowable dilution flow (Q^)
to be used in the wasteload allocation (HLA)• The dilution fraction is calculated
as detailed in Procedure 5.1 Part B(l)(4)(b).
Discharge induced mixing; is the area where the velocity and momentum associated
with an effluent being discharged from the end of a pipe is dissipated and the
dilution or mixing which then occurs is associated only with the natural dispersion
1CSO; is the concentration of a test material at which SO percent of the exposed
organisms exhibit a specified response during a specified time period.
Existing Sources: any source which does not meet the definition of a new source.
Existing Uses; are those uses actually attained in the water body on or after
November 28, 1975, whether or not they are included in the water quality standards.
Final Acute Value fFAVl; an estimate of the concentration of a material
corresponding to a cumulative probability of 0.05 in the acute toxicity values for
the genera with which acceptable acute tests have been conducted on the material.
The calculated FAV may be lowered to be equal to the Species Mean Acute Value of a
commercially or recreationally important species of the Great Lakes basin.
Final Chronic Value (FCV1; an estimate of the concentration of a material
corresponding to a cumulative probability of 0.05 in the chronic toxicity values for
the genera with which acceptable chronic tests have been conducted on the material.
Alternatively, the FCV may be determined by dividing the FAV by an acute-chronic
ratio. The calculated FCV may be lowered to be equal to the species Mean Chronic
Value of a commercially or recreationally important species of the Great Lake basin.
Final Plant Value; is the result of a 96-hour alga or chronic test conducted with
an aquatic vascular plant.
GLI Pollutants* are pollutants for which criteria are developed using the aquatic
life, wildlife, or human health criteria development methodologies.
Great Lakes states; means the States of Illinois, Indiana, Michigan, Minnesota, Mew
York, Ohio, Pennsylvania, and Wisconsin.
Great Lakes System; means all of the streams, rivers, lakes and other bodies of
water that are within the drainage basin of the St. Lawrence River at or upstream
from the point at which this river becomes the international boundary between Canada
and the United States. This also includes the waters of Lake Erie, Lake Huron, Lake
Michigan, Lake Ontario, and Lake Superior.
Human Cancer Criterion (HCC1: a Human Cancer Value (HCV) that is adopted as a
numeric ambient water quality criterion under this guidance.
Human Cancer Value IHCV)i the maximum ambient water concentration of a substance at
or below which a lifetime of exposure from drinking the water, consuming fish from
the water, and water-related recreation activities will protect humans from an
unreasonable risk of contracting cancer at a level of 1 in 100,000.
Human Noncancer Criterion (HNC\t a Human Noncancer Value (HMV) that is adopted as a
numeric ambient water quality criterion under this guidance.
-------
Human Noncancer Value (HNV|: the maximum ambient water concentration of * subst
at or below which no adverse noncancer effects are anticipated to occur from
lifetime exposure via drinking the water, consuming fish from the water, and water-
related recreation activities.
Load Allocation fLAl; is the total load of a substance from nonpoint sources to a
water body. Nonpoint sources include: in-place contaminants, direct wet and dry
deposition, groundwater inflow, and overland runoff.
LCSO; is the concentration of a test material at which SO percent of the exposed
organisms die during a specified time period.
Lowest Observed Adverse Effect Level; the lowest tested dose or concentration of a
substance which resulted in an observed adverse effect in exposed test organisms
when all higher levels resulted in the same effect.
No Observed Adverse Effect Level; the highest tested dose or concentration of a
substance which resulted In no observed adverse effect in exposed test organisms.
Honcarinoaen; a substance which is not classified as a known, probable, or possible
human carcinogen according to the weight-of-evidence assessment.
Non-public water supply; is a water that is not designated as a public water
supply.
Octanol/water partition coefficient (K_>; is the ratio, at equilibrium, of the
concentration of a substance in the octanol phase to its concentration in the
aqueous phase in a two-phase octanol/water system.
Open Waters of the Great Lakes: means all of the waters within Lake Erie, Lake
Huron, Lake Michigan, Lake Ontario, and Lake Superior lakeward from a line drawn
across the mouth of tributaries to the Lakes, including all waters enclosed by
constructed breakwaters, but not including the connecting channels.
Public Water Supply; is a designation by a State for a water that, with
conventional treatment, will be suitable for human intake and meets federal
regulations for drinking water.
Reasonable Potential; is where an effluent is projected or calculated to cause an
excursion above a water quality standard based on existing controls, effluent
variability, species sensitivity, and dilution allowance, as minimum considerations.
Relative Source Contribution fRSC): a factor used in the calculation of an HNC to
account for anticipated exposure other than those directly addressed by the HNC
derivation.
Risk Associated Dose fRADl; a dose of a known or presumed carcinogenic substance in
milligrams/kilogram/day which is calculated to be associated with a lifetime
incremental cancer risk equal to 1 in 100,000.
Species Mean Acute Value; is the geometric mean of the results of all flow-through
tests in which the concentrations of test material were measured. For a species for
which no such result is available, the SMAV should be calculated as the geometric
mean of all available acute values.
Steady-state BAF/BCFt in a BAF or BCF that does not change substantially over time;
that ie, the BAF or BCF that exists when uptake and depuration are equal.
Stream Peaion Flow; is the stream flow that represents critical conditions,
upstream of the source, for protection of each classification of criteria.
-------
Threatened or endangered species; those species that are listed as threatened or
endangered under state or Federal law.
Threshold Effect; an effect of a substance that theoretically has a minimally
effective dose or concentration, below which the effect does not occur.
Total Maximum Daily Load fTMDLlt is the maximum amount of a pollutant which may be
discharged to a water body and still ensure attainment or maintenance of water
quality standards as defined through application of the procedures established in
this guidance.
Toxic Substancei a substance which is recognized as being capable of eliciting
adverse effects in exposed organisms due to anticipated or assumed exposure.
Tributaries of the Great Lakes System; means all waters which are upstream of the
point where the tributary enters the Open Waters of the Great Lakes.
Uptake; is the sorption of a substance into or onto an aquatic organism.
Wasteload Allocation fWLAlt is the portion of a receiving water's loading capacity
that is allocated to one of its existing or future point sources of pollution.
The definitions provided in 40 CFR Parts 122, 130 and 131 also apply to this
guidance and to the Great Lakes guidance prepared for aquatic life, wildlife and
human health criteria and antidegradation.
4. SITE-SPECIFIC
A. Requirements for Site-Specific modifications of criteria.
Criteria may be modified to reflect site-specific conditions only
in accordance with the following guidance:
(1) Aquatic Life. Site-specific modifications to the aquatic
life criteria or levels of concern to reflect local
environmental conditions may be developed in accordance with
the guidance provided in Chapter 4 of the U.S. EPA Water
Quality Standards Handbook, 1983, or subsequent revisions.
These site-specific modifications may be utilized by the
Great Lakes States as a basis for establishing water quality
standards to protect the designated uses of a specific water
body, and as such, will reflect the requirement at 40 CFR
131.11(a)(i) for sound scientific and technological
rationale.
(2) Wildlife. Wildlife criteria or levels of concern maybe
designed to protect the individuals of a threatened or
endangered species population. This can be accomplished
through the use of an additional uncertainty or other
documented factor in the equation for the Wildlife Value.
(3) pioaccumulatjon. Bioaccumulation factors may be modified on
a site-specific basis where reliable data show that local
bioaccumulation is greater than the basin-wide average.
-------
(4) ffW*n *Tf «"»**- Human health criteria or levels of concern
may be designed to protect the individuals of a specific
population which consume fish at a rate higher than the
basin-wide average. This can be accomplished by using a
fish consumption rate of greater than 15 g/day in the human
health criteria equations, as long as the lipid content of
the consumed fish is equal to or greater than 6%.
B. notification of other Great Lakes States. When a State
determines that information warrants a site-specific modification
to a criterion or level of concern as allowed in Part A, the
State shall notify the other Great Lakes States of such a
determination and supply any justification as necessary.
*****ALTERNAT1VE PROPOSAL FOR PREAMBLE: This alternative approach
proposes to restrict site-specific modifications to the criteria for
bioaccumulative chemicals of concern (BCCs) to only more restrictive
changes, and to allow site-specific modifications to criteria for
pollutants that are not BCCs to be increased or decreased as the
situation warrants.
5. TOTAL MAXIMUM DAILY LOADS
WASTELOAD ALLOCATIONS AND MIXING ZONES FOR POINT SOURCES
5.0 Total MaH*'fiffiMffl PflilY Loads
It is the goal that total maximum daily loads (TMDL) be
determined when required under Section 303(d) of the Clean Water
Act for all affected waters of the Great Lakes System including
the Lakes and all waters tributary to the Lakes.
The TMDL process is a determination of the maximum amount of a
pollutant that may be discharged to a water body and still ensure
conformance with water quality standards. When a TMDL has been
completed, it facilitates the allocation of pollutant loads to
point, nonpoint and natural background sources, including a
margin of safety where appropriate. Any loading above the TMDL
risks violating water quality standards.
TMDL's may be developed for an entire water body system involving
many point and nonpoint sources, a water body segment with a
cluster of sources, or a single source. TMDL's, including a
margin of safety (MOS), are calculated to meet the most stringent
water quality criteria for the protection of human health,
wildlife, or aquatic life under all design requirements provided
in this procedure.
A. General.
(1) TMDLs. A total maximum daily load (TMDL) is the maximum
amount of a pollutant which may be discharged to a water
8
-------
body and still ensure attainment, or maintenance of water
quality standards as defined through application of the
procedures established in this guidance. The process for
determining TMDLs when required by Section 303(d) of the
Clean Water Act, for a parameter and waterbody is defined as
follows:
TMDL = WLA + LA 4- MOS
where: WLA = wasteload allocation; that portion of
the TMDL which may be allocated to point
sources to the water body;
LA = load allocation; that portion of the
TMDL which may be allocated to non-point
source contributions including
background in the water body; and
MOS = margin of safety; that portion of the
TMDL reserved to account for any lack of
knowledge concerning the establishment
of the TMDL.
(2) Conditions of TMDLs. A TMDL may be determined for all or
any portion of the Great Lakes system. The TMDL, including
a specifically identified MOS, for each tributary as it
enters a Lake or connecting channel shall be determined such
that all water quality criteria are met in the tributary at
the point of entry at the appropriate stream design flow as
provided in Section 5.1 Part 6(4)(a). It is not necessary
to establish a TMDL for all waters which may be affected by
a point source or nonpoint source prior to establishing a
WLA or LA for the respective sources. The WLA or LA shall
be the most stringent values which have been determined
through the establishment of a TMDL for any waterbody which
may be affected by those sources.
(3) Load Allocations. It is not necessary to establish a LA for
nonpoint sources prior to the establishment of a WLA for
point sources. A load allocation may be established as
necessary by the regulatory agency or as determined through
the procedures described below. If a load allocation is not
specifically established, then, under the procedures of this
guidance, the load allocation is accounted for through the
establishment of the margin of safety (MOS) as specified in
this section.
(4) Other Conditions. If the sum of point source discharges and
the nonpoint source discharges exceeds the TMDL minus the
MOS for that substance, then either the point source
discharges or the nonpoint sources discharges or both shall
be reduced such that the TMDL is not exceeded.
-------
B. WLAa for BCCs. A primary policy of the GLWQG is to protect th
Great Lakes from the effects of bioacummulative chemicals of
concern (BCC) and to assure progress toward zero discharge and
virtual elimination of the discharge of BCCs. The following
requirements, therefore, apply:
(1) Existing sources. The WLA portion of the TMDL for a point
source which exists on the date of publication of this
Initiative shall not exceed:
(a) Beginning January 1, 2004, a value equivalent to the
most stringent water quality criterion in effect on
that date or the value determined by Section ll.B of
this guidance, for the substance or a fraction of the
WLA which has been established for an open water of the
Great Lakes (OWGLs) or a portion thereof, whichever is
more stringent;
(b) Until January 1, 2004, for direct discharges to OWGL's,
a value determined in accordance with Section 5.1 Part
A(l)(b) or Section ll.B, or a fraction of a WLA for a
Lake or a portion thereof, which ever is more
stringent;
(c) Until January 1, 2004, for dischargers to tributaries
of the Great Lakes System (TGLS), a value determined '
accordance with Section 5.1 Part B(l)(b) or Section
ll.B, or a fraction of a WLA for a Lake or a portion
thereof, whichever is more stringent.
(2) New sources. Subject to the provisions of the
antidegradation procedures, and beginning on the effective
date of this Intitiative, the WLA portion of the TMDL for
new point sources shall not exceed a value which is
equivalent to the most stringent water quality criterion in
effect on that date or a value determined by Section ll.B,
for the substance or a fraction of the WLA which has been
established for a Lake or a portion thereof.
C. TMDLs for TQL8 for pollutants that are pot BCCs.
(1) TMDLs for tributary basins. Whenever a TMDL is established
for a basin which is tributary to the Great Lakes or
connecting channels, the tributary shall meet the conditions
specified in Section 5.0 Part A.(2). In addition, all
points within the basin shall meet the conditions specified
in subsections (2) and (3) of this Section, including all
MOSs for segments or individual TMDLs. A MOS shall be
incorporated into the basin TMDL analysis to account for the
lack of knowledge of nonpoint source, background and point
source loadings. The MOS may vary from tributary basin to
tributary basin dependent on the confidence in the source
10
-------
loading information available, but shall not be less than
25% of the TMDL.
(2) Segment or individual TMDLs. Whenever a TMDL is established
for a portion of a tributary to the Great Lakes, the margin
of safety and the WLA shall be calculated subject to the
conditions which follow:
(a) Margin ^f Safety.
(i) If a LA and a mixing zone are not specifically
determined, the MOS shall be equal to not less
than 75% of the dilution or assimilative capacity
at the stream design flow as determined in Section
5.1 Part B.(4)(a).
(ii) In TGLSs, the MOS for pollutants that are not BCCs
for which both a mixing zone and LA have been
determined is a variable value which is not less
than 25% of the dilution or assimilative capacity
at the stream design flow as determined in Section
5.1 Part B(4)(a).
Wasteload Allocations.
(i) If a mixing zone demonstration under Section 5.2
has not been completed by the permittee, approved
by the permitting authority and/or LA has not been
determined by the permittee then, the WLA shall be
the difference between the MOS determined in
Section 5.0 Part C(l)(a) and the total dilution
or assimilative capacity at the stream design flow
and shall not be greater than 25% of the dilution
or assimilative capacity at the stream design flow
determined in Section 5.1 Part B.(4)(a).
(ii) If a mixing zone demonstration under Section 5.2
has been completed by the permittee, approved by
the permitting authority and a LA has been
specifically determined as a portion of the TMDL,
the WLA shall be that portion of the TMDL not
assigned to the LA and the MOS (as established by
Section 5.0 Part C.(l)(b)) but shall not be
greater than 75% of the dilution or assimilative
capacity at the stream design flow determined in
Section 5.1 Part B.(4)(a).
D. TKDLa for Direct Dischargers to OWQLs for pollutants that are
BCCS.
(1) Lakewide TMDLs.
11
-------
(a) If a TMDL has been determined for a Lake, or portio
thereof, receiving the direct discharge of the point
source(s), the magnitude of the MOS shall be
established at the discretion of the individual
permitting agencies.
(b) If a TMDL has been determined for a Lake, or portion
thereof, receiving the direct discharge of the point
source(s), then the WLA for each point source including
direct discharges to the Lake shall equal a fraction of
the WLA for all point sources which discharge to the
Lake, but no greater than the WLA established by the
methods in Section 5.1 or Procedure 11. The portion of
the WLA and LA which is allocated or assigned to the
individual point or nonpoint sources shall be
determined on a case-by-case basis by the affected
permitting authorities.
(2) Individual TMDLs
(a) If a TMDL has pofr been determined for a Lake, or
portion thereof, receiving the direct discharge of the
point source(s), the MOS is assumed accounted for
through establishment of the allowable dilution in
Section 5.1 Part A.
(b) If a TMDL has not been determined for a Lake, or
portion thereof, receiving the direct discharge of t
point source(s), then the TMDL for each point source
location shall equal the WLA as determined in Section
5.1 Part A.
5.1 procedures for Developing Source Specific TMDLs/wasteload
Allocations. This section describes how to calculate wasteload
allocations (WLA) for point source discharges and shall be used
in conjuction with the determination of TMDLs under Section 5.0.
The applicable WLA shall be the more stringent of that determined
under Section 5.0 or this section. These procedures are
applicable to the determination of wasteload allocations for
point source discharges where the load allocation (LA) for
nonpoint sources is determined by the regulating agency to be
zero or negligible or where the nonpoint source component of the
total maximum load is accounted for in the determination of the
background concentration used in the equations. The wasteload
allocations which are determined by the equations in this
procedure may be expressed as concentrations of substances by
deleting the conversion factor which is included in the right
hand side of each equation.
When the LA is determined not to be zero or not to be negligible,
the WLA for point sources will be determined in accordance with
Section 5.0 of this guidance.
12
-------
A. Discharges to Lakes. Procedure for developing wasteload
allocations (WLAs) for sources to the Open Wat«rs of the Great
Lakes (OWGL1. inland lakes and other waters of the Great Lakes
System which do not exhibit unidirectional flow.
(1) Existing Sources. WLAs based upon chronic aquatic life,
wildlife and human health criteria for existing sources
shall be developed in accordance with the following
requirements.
(a) Non-bioaccumulative. WLAs for pollutants that are not
bioaccumulative chemicals of concern shall not exceed:
(b)
WLA < L! 1 (criterion) - 10 (background concentration)]
(X)
where:
criterion =* a chronic aquatic life criterion,
wildlife criterion, or human health
criterion specified in units of mass
per unit of volume, and
background - background of the substance
specified in units of mass per unit
of volume.
X - a conversion factor which converts
units of mass per unit volume to
units of mass per unit time.
(ii) A demonstration for a larger mixing zone may be
provided, approved and implemented in accordance with
Section 5.2 Part A of this Procedure, but in no case
shall a demonstration result in a mixing zone which
exceeds the area where discharge induced mixing occurs.
(iii) If the background concentration in the OWGL is
equal to or exceeds the criterion, the WLA shall not
exceed the levels specified in Procedure 11 -
Background Concentrations of Pollutants of this
guidance.
pjifrfrecpmulative. Beginning January 1, 2004, WLAs for
bioaccumulative chemicals of concern (BCCs) shall not
exceed the criteria effective at that time. Until
January 1, 2004, WLAs for BCCs shall be set according
to the following formula:
(i)
NLA < L!1(criterion) - 10(background concentration)J (I)
where:
criterion
a chronic aquatic life criterion,
wildlife criterion, or human health
criterion specified in units of mass
per unit of volume, and
SStf
-------
background * background of the substance
specified in units of mass per u
of volume.
X - a conversion factor which converts
units of mass per unit volume to
units of mass per unit time.
(ii) Compliance with criteria expressed in terms of
concentration shall not be achieved through flow
augmentation.
(iii) If the background concentration in the OWGL is
equal to or exceeds the criterion, the WLA shall not
exceed the levels specified in Procedure 11 -
Background of this guidance.
(2) New Sources. WLAs based upon chronic aquatic life, wildlife
and human health criteria for new sources shall be developed
in accordance with the following requirements.
(a) Non~bioaccumulative. WLAs for pollutants that are not
BCCs shall equal the criteria, unless a mixing zone
demonstration is provided, approved and implemented in
accordance with Section 5.2 Part A of this Procedure,
but in no case shall a demonstration result in a mixing
zone which exceeds the dilution provided by the
equation in Section 5.1 Part A.(1)(a)(i).
(b) Bioaccumulative. WLAs for BCCs shall equal the
criteria. Compliance with criteria expressed in terms
of concentration shall not be achieved through flow
augme ntat ion.
(3) Existing and New Sources. WLAs based on acute aquatic life
criteria for both existing and new sources shall not exceed
the Final Acute Value (FAV).
(4) Other Conditions. For waters that are not OWGLs, a
permitting authority may establish more restrictive
requirements than those specified elsewhere in this
procedure.
B. Discharges to Tributaries, Procedure for developing WLAs for
sources to tributaries of the Great Lakes System (TQL8) and the
connecting channels of the Great Lakes except those that do not
exhibit unidirectional flow. Whenever the effects of two or more
point sources of the same pollutant overlap, or a portion
thereof, the WLA calculation by this procedure shall be the total
WLA for all point sources. This total WLA shall be further
allocated to all point source discharges to the water body at the
discretion of the individual state permitting agency.
14
-------
(1) Existing Sources. WLAs based upon chronic aquatic life,
wildlife and human health criteria for existing sources
shall be developed in accordance with the following
requirements.
(a) Non-bioaccumulative. WLAs for pollutants that are not
BCCs shall be calculated using the following equation:
WLA - [(critenonUQad + fl-flsource flow) - Pad I background l] (X)
source flow
where: criterion - a chronic aquatic life criterion,
wildlife criterion, or human health
criterion specified in units of mass
per unit of volume,
Q.J - allowable dilution as calculated in
Section 5.1 Part B.(4),
f » fraction of the source flow that is
withdrawn from the receiving water,
source flow - flow rate of the source specified in
units of volume per time,
background * background of the substance
specified in units of mass per unit
of volume.
X * a conversion factor which converts
units of mass per unit volume to
1 units of mass per unit time.
(i) Q«i may be greater than that specified in the
conditions in Section 5.1 Part B.(4)(b) but not
1 greater than 0.75 where a mixing zone demonstration, as
set forth in Section 5.2. has been provided by the
permittee and approved by the permit issuing authority
i which shows that an alternative value should be used.
i
I (11) If the background concentration in the TGLS is
equal to or exceeds the criterion, the WLA shall not
exceed the levels specified in Procedure 11 -
Background Concentrations of Pollutants of this
guidance.
(b) Bioaccumulative. Beginning January 1, 2004, WLAs for
BCCs shall equal the criteria. Until January l, 2004,
WLAs shall be calculated by the following equation:
WLA -[(criterionUOad + J (X)
source flow
where: criterion * a chronic aquatic life criterion,
wildlife criterion, or human health
15
-------
criterion specified in units of
per unit of volume,
Q.4 - allowable dilution as calculated in
Sections 5.1 Part B.(4),
f « fraction of the source flow that is
withdrawn from the receiving water,
source flow « flow rate of the source specified in
units of volume per time,
background - background of the substance
specified in units of mass per unit
of volume.
X -a conversion factor which converts
units of mass per unit volume to
units of mass per unit time.
Compliance with criteria expressed in terms of
concentration shall not be achieved through flow
augmentation.
(i) If the background concentration in the TGLS is
equal to or exceeds the criterion, the WLA shall not
exceed the levels specified in Procedure 11 -
Background Concentrations of Pollutants of this
guidance .
(2) New Sources. WLAs based upon chronic aquatic life, vildl
and human health criteria for new sources shall be developed
in accordance with the following requirements.
(a) Non-b Loaccumulative . WLAs for pollutants that are not
BCCs shall equal the criteria, unless a mixing study is
provided, approved, and implemented in accordance with
Section 5.2 Part A of this Procedure, but in no case
shall a demonstration result in a mixing zone which
exceeds the dilution provided by the equation in
Section 5.1 Part B. (1) (a) using the appropriate values
from Section 5.1 Part B(4).
(b) Bjpaee^roulative,. WLAs for BCCs shall equal the
criteria. Compliance with criteria expressed in terms
of concentration shall not be achieved through flow
augmentation.
(3) Existing and New Sources. WLAs based on acute aquatic life
criteria for both existing and new sources shall not exceed
the Final Acute Value (FAV) .
(4) Allowable Dilution Flow fO..l . The allowable dilution flow
will be determined by the following equation:
16
,530
-------
= stream design flow * dilution fraction
(a) Stream Design Flow. The stream design flow shall be:
(i) either the 7 day, 10 year low flow (7Q10) or the
biologically based design flow for chronic aquatic life
criteria;
(ii) the Harmonic mean flow for human health criteria;
and
(iii) the 30 day, 5 year low flow (30Q5) for wildlife
criteria.
(b) Dilution Fraction. The dilution fraction for all
design flows and for all criteria is based upon the
ratio of the 7Q10 to the source flow. It is determined
by the following conditions when the ratio is:
(i) less than or equal to 250, the dilution fraction
shall be no greater than 0.25.
(11) greater than 250 but less than 300, the dilution
fraction shall be no greater that the results
calculated by the following equation:
Dilution Fraction * 103 - 0.3I(7Q10/Sourc« Flow!
100
(111) is equal to or exceeds 300, the dilution
fraction shall be no greater than 0.1.
5.2 Mixing Zone Demonstration
A. Mixing Zone Demonstration Requirements.
(l) Requirements for demonstration. The mixing zone
demonstration must:
(a) describe the size, shape, and location of the area of
mixing, including the manner in which diffusion and
dispersion occur;
(b) for sources to the open waters of the Great Lakes,
define the location at which discharge induced mixing
ceases;
(c) document the substrate character and geomorphology
within the mixing zone;
17
-------
(d) show that the mixing zone does not interfere with or
block passage of fish or aquatic life at the conflu
of tributary streams.
(e) show that the mixing zone does not extend to drinking
water intakes; and
(f) show that the mixing zone would not otherwise interfere
with the designated uses of the receiving water or
downstream waters.
(2) Other factors. In addition, the demonstration shall address
the following factors:
(a) that the zone of passage for mobile aquatic life, which
meets all applicable criteria is provided, if the
demonstrated mixing is granted;
(b) whether or not contiguous mixing zones overlap;
(c) whether organisms would be attracted to the area of
mixing, as a result of the effluent character;
(d) whether the habitat supports endemic or naturally
occurring species;
(e) that the mixing does not promote undesirable aquatic
life or result in a dominance of nuisance species; a
L.
(f) that by allowing additional mixing/dilution:
(i) substances will not settle to form objectionable
deposits;
(ii) floating debris, oil, scum, and other matter in
concentrations that form nuisances will not be
produced;
(iii) objectionable color, odor, taste or turbidity
will riot be produced.
B. Other Conditions. For situations where a mixing zone
demonstration, as set forth in Section 5.2 Part A, has been
provided by the permittee:
(1) the permittee must also supply information with which to
establish the LA to the same receiving water.
18
£3*-
-------
(2) if the permitting authority approves the demonstration made
pursuant to the conditions of this Part, an adjustment to
the dilution fraction specified in Section 5.1 Part B(4)(b)
may be made. This adjustment shall not increase the
dilution fraction to greater than 0.75.
6. COMPLIANCE SCHEDULES
A. Hew or more restrictive limitations due to existing conditions.
Any existing permit which is reissued or modified to contain a
new or more restrictive effluent limitation based upon Tier I
criteria or Tier II levels of concern or whole effluent toxicity
criteria, may allow a reasonable period of time, not to exceed
the term of the permit or three years whichever is less, for the
permittee to comply with that limit, provided that limitation is
not necessary due to a new or increased discharge.
When a permit is issued, reissued or modified to contain an
effluent limitation derived from a Tier I criterion or Tier II
level of concern or whole effluent toxicity criterion, that
effluent limitation, having become necessary due to a new or
increased discharge, shall be effective upon the commencement of
the new or increased discharge.
C. Delayed effectiveness ofTier II limitations. Whenever a limit
based upon a Tier II level of concern is proposed for
incorporation into a permit, the permittee shall be allowed a
period of up to three years in which to provide additional
studies necessary to develop a Tier I criterion or to modify the
existing criterion. In such cases, the permit shall contain a
reopener clause indicating that the Tier II based limitation
becomes effective at the end of the study period, unless the
necessary studies have been provided by the permittee, during
that study period. When the permittee demonstrates through
additional studies that a revised limit is appropriate, that
limit shall be incorporated through permit modification and shall
become effective within the permit term. The limit revised based
upon additional studies is not affected by antibacksliding.
7. WQBELS BELOW THE LEVEL OF DETECTION. When an effluent limitation
based upon a GLI criterion or value of concern (WQBEL) for a
chemical is determined to be less than the detection level (DL)
of the most sensitive, approvable analytical technique available,
the permitting authority shall use the following strategy to
regulate the source of that chemical in the NPDES permit.
19
533
-------
A. Permit Limit. Include the WQBEL in the NPDES permit with an
effective date prior to the expiration date, specify an
analytical method and measurement frequency, and identify the DL
for the chemical that is not to be exceeded. The permittee shall
be given the opportunity to demonstrate that a higher level is
appropriate because of matrix interference.
B. Narrative Statement. Include permit language explaining that the
WQBEL for the chemical is less than the DL of the specified
analytical method.
C. Status of Samples Above Detection. Include permit language
clarifying that any discharge of the chemical at or above the DL
is an exceedance requiring further action to resolve the status.
D. Program Requirement. Include a condition in the permit which
establishes a program for proceeding towards compliance with the
WQBEL and towards the goal of eliminating all detectable levels
of the chemical in all internal (industrial) or indirect
(municipal) wastewater streams contributing to the permittee's
wastewater collection system. The chemical minimization program
should include, but not be limited to, the requirements described
in Part A(8).
E. Compliance Language. Include permit language specifying that the
permittee will be considered in compliance by the permitting
authority with the WQBEL if all the samples in any monthly
reporting period are less than the DL, and full compliance with
the chemical minimization program requirements is being achieved.
F* BCCs. If the chemical is a BCC, include a condition in the
permit which requires the permittee to determine if the chemical
is bioconcentrating or bioaccumulating in fish exposed to the
effluent. Resident fish monitoring, caged fish monitoring,
effluent chemical bioconcentration studies, and/or application of
other procedures in conjunction with the EPA procedure described
in "Assessment and Control of Bioconcentratable Contaminants in
Surface Water" (EPA 600/ ) should be required as part of the
permit condition. To the extent that these studies reveal
unacceptable accumulation in fish tissue as a result of the
discharge, the control strategy to minimize the chemical shall be
reviewed and modified as appropriate.
G. Other Requirements. The permit may also require the development
and implementation of other innovative monitoring programs.
These programs would be determined on a case-by-case basis and
may include:
(1) new analytical equipment and methods more sensitive than the
analytical method specified in the permit;
20
-------
(2) internal waste stream monitoring and mass balance modeling
techniques; and
(3) other innovative monitoring techniques capable of adequately
determining the compliance status of the effluent.
H. Chemical Minimization Program. The goal of the chemical
minimization program should be to reduce and maintain all
potential sources of the chemical below the DL. The minimization
program should, as a minimum, include the following:
(1) an annual review and semi-annual monitoring of potential
sources of the chemical;
(2) quarterly monitoring for the chemical in the influent to the
wastewater treatment system;
(3) submittal of an approvable control strategy designed to
proceed toward the goal of maintaining all sources of the
chemical to the wastewater collection system below the DL;
(4) if the sources of the chemical are discovered, appropriate
control measures should be implemented, consistent with
approved control strategies; and
(5) an annual status report should be sent to the permitting
authority including:
(a) all minimization program monitoring results for the
previous year;
(b) a list of potential sources of the chemical; and
(c) all action taken to determine and eliminate the
chemical.
*****ALTERHATIVE PROPOSAL FOR PREAMBLE: This alternative approach
proposes to assign a value to the nondetected samples. The average of
the assigned values and the detected values for a particular period of
time is then calculated. The average concentration is compared to the
effluent limit. A value which assigns an equal amount of risk to the
environment and to the discharger is 50% of the DL or the WQBEL,
whichever is less.
Comments will be solicited on other types of detection levels which
could be used inplace of the minimum level.
8. ADDITIVITY. The assumption of additivity will apply to human
health cancer or wildlife criteria when other information
regarding the effects of mixtures is not available. This
provision applies to discharges from point sources, only.
21
-------
A. Aquatic Life Criteria. When establishing requirements under t
provisions of these implementation procedures, for aquatic li
criteria, the effects of chemical mixtures are assumed to not be
additive. Whole effluent toxicity requirements established under
Procedure 10 are designed to account for additive effects to
aquatic organisms.
B. Wildlife Criteria. When establishing requirements under the
provisions of these implementation procedures, for wildlife
criteria, the effects of chemical mixtures are assumed to pot be
additive, except where specified in the criteria through the
application of toxicity equivalency factors as provided in Part E
of this Procedure.
requirements under the provisions of these implementation
procedures, for human health criteria for noncarcinogens
(threshold), the effects of chemical mixtures are assumed to not
be additive, except where specified through the application of
toxicity equivalency factors.
D. Human Health - Cancer criteria. When establishing requirements
under the provisions of these implementation procedures, for
human health criteria for carcinogens, the effects of chemical
mixtures are »^gtnn«>4 to be additive and the following apply:
(1) If an effluent for a discharger contains detected levels
more than one substance for which a Tier I criterion or T
II level of concern, exists at levels which warrant water
quality-based limitations as determined for the individual
substances permitting authority, the incremental risk of
each carcinogen is assumed to be additive. Except as
provided in (2) below, the limitation for each carcinogen
shall be established in a permit to protect against additive
effects possibly associated with simultaneous multiple
chemical human exposure such that the following condition is
met:
c, •*•
Limit, Limit} Limit.
Where: C,..n - the monthly average concentration of each
•eparate carcinogen in the effluent.
Limiti ..n - the effluent limitation concentration calculated
for each aubatance independent of other
carcinogena which may be preaent in the
background baaed on the human cancer criterion
for each respective carcinogen.
(2) If it can be shown that the carcinogenic risk for a mixture
of substances is not additive, the limitations for each
carcinogen will be determined based on that information.
22
-------
TEFa applied to Wildlife Criteria. Toxicity equivalency factors
shall apply when establishing requirements under the provisions
of these Implementation procedures for wildlife criteria for
chlorinated dibenzodioxins (CDDs), chlorinated dibenzofurans
(CDFs) and chlorinated biphenyls (CBs). Concentrations of CDDs,
CDFs, and CBs in a discharge shall be converted into "equivalent"
amounts of 2,3,7,8-TCDD by multiplying the concentration of the
substance by the TEF value shown in Table l. All resultant
concentrations are added to produce an "equivalent11 value for
2,3,7,8-TCDD which is used to calculated the WLA for the
discharger.
equivalency factors shall apply when establishing requirements
under the provisions of these Implementation procedures for human
health criteria for CDDs and CDFs. Concentrations of CDDs and
CDFs in a discharge shall be converted into "equivalent" amounts
of 2,3,7,8-TCDD by multiplying the concentration of the substance
by the TEF value shown in Table 1. All resultant concentrations
are added to produce an "equivalent" value for 2,3,7,8-TCDD which
is used to calculated the WLA for the discharger. If multiple
carcinogens are present in a discharge, the value of CB in the
equation in part D{1) is the equivalent value for 2,3,7,8-TCDD.
23
53?
-------
Table 1
Toxic Equivalency Factor Values
compound TEF Value
1. Dioxina
Mono-, Di-, and TriCDDa 0
2,3,7,8-TCDD 1
other TCDDs 0
2,3,7,8,-PeCDD 0.5
other PeCDDa 0.0
2,3,7,8-HxCDDfi 0.1
other HxCDDe 0.0
2,3,7,8-HpCDD 0.01
other HpCDDs 0.0
OCDD 0.001
2. Furans
Mono-, Di-, and TriCPDe 0
2,3,7,8-TCDF 0.1
other TCOFs 0.0
2,3,4,7,8-PeCDF O.S
1,2,3,7,8-PeCDF 0.05
other PeCDFs 0.0
2,3,7,8-HxCDFn 0.1
other HxCDFa 0.0
2,3,7,8-HpCDFi! 0.01
other HpCDFe 0.0
OCDF 0.001
3(a) Coplanar PCBs IUPAC*
3,3',4,4',5-PeCB 126 0.1
3,3',4,4',5,5'-HxCB 169 O.OS
3,3',4,4'-TCB 77 0.01
3{b) Monoortho Coplanar PCBi
2,3,3',4,4'-P«CB 105 0.001
2,3,4,4'5-PeCB 114 0.001
2',3,4,4'5-PeCB 123 0.001
2,3',4,4',5-P«CB 118 0.001
2,3,3',4,4',5-HxCB 1S6 0.001
2,3,3',4,4',5-HxCB 157 0.001
2,3',4,4',5,5'-HxCB 167 0.001
2,3,3'4,4',5,5',-HpCB 189 0.001
TEF References:
1. Kubiak, T.J., et al. 1989. Microcontaminants and Reproductive
Impairment of the Forstar's Tern on Green Bay, Lake Michigan - 1983.
Archives of Environmental Contamination and Toxicology. Vol. 18, pp.
706-727.
2. Personal communication. 1990. Robert K. Ringer, Michigan State
University - Center for Environmental Toxicology, East Lansing,
Michigan.
24
-------
3. Smith, L.M., et al. 1990. Determination and Occurance of AHH-
Active Polychlorinated Biphenyls, 2,3,7,8-tetrachlorodibenzofuran in
Lake Michigan Sediment and Biota - The Question of Their Relative
lexicological Significance. Chemosphere. Volume 21, No. 9, pp. 1063-
1077.
4. Tillitt, D.E., 1989. Planar Chlorinated Hydrocarbons (PCHs) in
Colonial Fish Eating Bird Eggs From the Great Lakes. Marine
Environmental Research. Vol. 28 (1-4), pp. 505-508.
5. Kannan, N. 1989. Critical Evaluation of Polychlorinated Biphenyl
Toxicity in Terrestrial and Marine Mammals: Increasing Impact of Non-
ortho and Mono-ortho Coplanar Polychlorinated Biphenyls From Land to
Ocean. Archives of Environmental Contamination and Toxicology. Vol.
18, pp. 850-857.
• ""ALTERNATIVE PROPOSAL FOR PREAMBLE! This alternative approach
proposes to address the interactions among noncarcinogens. If the
target organ for two or more water quality limited chemicals is the
same, the interaction among those chemicals should be accounted for in
one of the following ways: 1) actual data on this interaction should
be used, if available, or 2) additivity should be assumed in the
absence of actual data.
9. LOADING LIMITS FOR POINT SOURCES. Whenever a WLA is determined
necessary and has been calculated under the provisions of
Procedure 5, the WLA shall be established as both a concentration
value and a mass value. Both values shall be consistent in terms
of daily, weekly or monthly averages or in other appropriate
time-related terms. The mass loading values shall be calculated
using the average design flow of the treatment facility in the
case of a publicly-owned treatment works or the average annual
flow in the case of an industrial source. If determined
necessary, wet weather WLAs for non BCCs for point sources may be
established at the discretion of the individual state permitting
agencies, but such loadings shall result in attainment of the
water quality criteria.
10. WHOLE EFFLUENT TOXICITY REQUIREMENTS FOR POINT SOURCES
A. WET Requirements. Permitting authorities shall require the
aquatic toxicities of industrial and municipal effluents to meet
the following conditions:
(1) At the point of discharge, the effluent shall not exceed 1.0
acute toxic unit (TUa), and
(2) All points of the receiving water outside the allocated
mixing zone shall not exceed l.o chronic toxic unit (TUc).
25
-------
B. Representative conditions. Determination of the need for WET
permit requirements shall be made considering data characteris
of existing discharge conditions and design flows as appropriate
for each receiving water body.
C. Conditions requiring WET limit. Where adequate toxicity data are
available to demonstrate that an effluent has the reasonable
potential to exceed the aquatic toxicity requirements set forth
in Part A., above, appropriate WET limitations should be included
in the NPDES permit. The basis for establishing WET limitations
is discussed in Part F, below. When WET is regulated by a
narrative WQS, the requirements for a WET limit may be waived
when the WQBELs on individual chemicals in the effluent are shown
to attain and maintain all applicable numeric and narrative water
quality standards. An appropriate schedule of compliance should
be incorporated in the NPDES permit for an effluent which
consistently exceeds the aquatic toxicity requirements set forth
in Part A., above. The schedule of compliance shall allow the
permittee sufficient time to complete a toxicity
identification/reduction evaluation (TI/RE) on the effluent.
D. Monitoring Requirements for WET. Where the effluent is suspected
to contain toxic substances but inadequate toxicity data are
available to determine if the effluent has the reasonable
potential to exceed the aquatic toxicity requirements set forth
in Part A., above, short-term or periodic long-term WET testing
requirements should be used to generate the data needed to
adequately characterize the aquatic toxicity of the effluent.
general, one year of quarterly acute or chronic toxicity testing
using at least one fish and macroinvertebrate test species should
be considered the minimum database needed to adequately
characterize the aquatic toxicity of the effluent. Appropriate
language should also be included in the NPDES permit which
requires the initiation and completion of a TI/RE by the
permittee if the toxicity testing data indicate that the effluent
exceeds the aquatic toxicity requirements set forth in Part A,
above.
E. Other retirements. Where adequate toxicity data are available
to demonstrate that an effluent is toxic to aquatic life but does
not have the reasonable potential to exceed the aquatic toxicity
requirements set forth in Part A., above, periodic WET testing
may be required in the NPDES permit.
F. Authority to require monitoring. Regardless of the results of
the analysis conducted under this procedure, the permitting
authority may, whenever determined necessary, require whole
effluent toxicity limits for a discharge. Considerations
include:
(1) available mix determined in accordance with the requirements
in Procedure 5;
26
-------
(2) species sensitivity;
(3) discharge category and predicted effluent quality;
(4) proximity to other discharges; and
(5) any relevant information that indicates a reasonable
potential to exceed the aquatic toxicity requirements set
forth in Part A.
G. Requirements for new sources. New sources shall be required to
meet any applicable whole effluent toxicity limits upon issuance
of the NPDES permit.
H. Basis for establishing WET limitations in MPDEfl permits.
(1) Permitting authorities shall consider a discharge to have a
reasonable potential to exceed the aquatic toxicity
requirements established for the water body when sufficient
site-specific information is available to demonstrate that:
(a) for acute toxicity, the following condition is met:
50% < B
% effect in 100% effluent
where: Effect - immobilization or mortality
B = Multiply factor taken from Table 3.2 in
U.S.EPA Technical Support Document. For
a data set where n < 10, the CV is set
equal to 0.6. For a data set where n >
10, the CV is calculated as standard
deviation/mean.
(b) for chronic toxicity, the following condition is met:
[chronic toxicity (TUe) of the effluent] > 1/(B x RWC)
where: B = Multiplying factor taken from Table 3.2 in
U.S.EPA Technical Support Document. For a
data set where n < 10, the CV is estimated to
be equal to 0.6. For a data set where n >
10, the CV is calculated as standard
deviation/mean.
RWC - Receiving water concentration of the effluent
in decimal form. For discharges to the TGLS,
RWC is determined by dividing the source flow
by the Qrf, where the facility's water supply
27
-------
is the receiving water or by dividing the
source flow by the quantity of the source
flow plus the Q.4, where the facility's water
supply is not the receiving water. For
dischargers to OWGLs, RWC is the source flow
divided by 11.
(c) When B \s greater than 1/RWC, the above formula is not
applicable due to the fact that chronic toxicity levels
below 1.0 TUC cannot be measured with adequate
statistical confidence. In these situations, the
permitting authority shall review the raw toxicity test
data available to determine whether the discharge has
the reasonable potential to exceed the chronic toxicity
requirements of the Part A.
(2) Toxicity testing is performed in accordance with methods
described an the following documents or subsequent
revisions:
(a) "Methods for Measuring the Acute Toxicity of Effluent
to Freshwater and Marine Organisms", EPA/600/4-85/013.
(b) "Short-Term Methods for Estimating the Chronic Toxicity
of Efiluents and Receiving Waters to Freshwater
OrganJsms", EPA/600/4-89/001 (except Method #1001 and
#1003).
(c) Other acute or chronic toxicity testing methods
determined to be acceptable by the permitting
authority.
(3) The permitting authority shall use the following guidance
when evaluating aquatic toxicity data available for an
effluent:
(a) chron:ic toxicity values for each species generated
within the same calendar month are averaged to evaluate
whether an effluent has the reasonable potential to
exceed the chronic toxicity requirements of Part A.
(b) acute toxicity values generated on samples collected
during the same day are averaged by species to evaluate
whether an effluent has the reasonable potential to
exceed the acute toxicity requirements of Part A.
(c) Effluent*specific acute:chronic toxicity ratios receive
preference when predicting chronic toxicity values from
existing acute toxicity data. However, when effluent-
specific acute-chronic toxicity ratios are not
available an acute-chronic ratio of 10 is assumed.
28
-------
11. BACKGROUND CONCENTRATIONS OF POLLUTANTS
A. Representative Background Concentration. The representative
background concentration of a toxic substance shall be used in
deriving the WLAs under Procedure 5.
(1) Except as provided elsewhere in this procedure, the
representative background concentration shall equal the
geometric mean of the acceptable available data for a
substance. Values below the level of detection shall be set
equal to one-half the detection level for the purposes of
calculating the geometric mean. Other statistical methods
may be used to estimate the levels of the non-detected
values.
(2) Background concentrations may not be measured at a location
within the direct influence of a point source discharge.
Representative background concentrations shall be determined
on a case-by-case basis using available data on the
receiving water or similar water bodies in the state and
best professional judgment. Background concentrations of
substances may be determined to include nonpoint source
contributions in the calculation of WLAs for point sources.
B- Background concentrations greater than the water quality standard
or criteria. This Section included provisions for determining
effluent limitations when the background concentration of a
pollutant in a receiving water exceeds an applicable water
quality standard or criterion. In applying these provisions,
however, all effluent limitations derived by this provision must
not cause any applicable TMDL to be exceeded. In such cases, the
effluent limitations shall be adjusted so that the TMDL is not
exceeded.
(1) Point source using water froma source other than the water
body to which the effluent is discharged.
(a) Whenever the representative background concentration
for a toxic substance in the receiving water is
determined to be greater than any applicable water
quality standard or criterion for that substance and
the source of at least 90% of the wastewater is from
groundwater or a public drinking water supply system,
the concentration value of the WLA for that substance,
shall be equal to the lowest applicable water quality
standard or criterion except as provided by Part
B.(1)(b). POTWs which discharge to the same surface
water from which the water supply is withdrawn shall be
subject to Part B(2). of this procedure.
29
-------
(b) The concentration value of the WLA may be establishe
at a concentration greater than the water quality
standard or criterion for the substance in the
receiving water as required by Part B.(l)(a) in a range
up to, but not greater than the representative
concentration of the substance in the receiving water.
The WLA shall only be increased above the standard or
criterion if it is demonstrated to the permitting
agency that the concentration of the substance in the
groundwater or public drinking water supply system at
the point of intake exceeds the applicable standard or
criterion for that substance ajod that reasonable,
practical or otherwise required methods are implemented
to minimize the addition of the toxic substance to the
wastewater. This part shall not apply where
groundwater is withdrawn from a location of
contain mated groundwater.
(2) point sources usingyatgrfrom the same water body to which
the effluent is discharged.
(a) Whenever the representative background concentration of
a toxic substance in the receiving water is determined
to be greater than any applicable water quality
standard or criterion for that substance and the source
of more than 10% of the wastewater for any discharger
is from the same receiving water, the concentration
value of the WLA for that substance shall equal the
representative background of that substance in the
receiving water. In addition, or as an alternative,
the mass value of the WLA may be established at a value
which requires that there be no net addition of the
toxic substance in the wastewater as compared to the
intake or source water.
C. Ma^im^iB background concentration values. The determination of
representative background concentrations for toxic substances in
Parts B.(l)(b) and B.(2)(a) shall be statistically (P<0.01) or
otherwise appropriately determined as the reasonably expected
maximum background concentration for that substance.
12. REASONABLE POTENTIAL TO EXCEED WQBELS
A. Determining Reaiionable Potential with large data sefrp. If 10 or
more representative effluent data samples are available for a
substance discharged from a point source, projected effluent
quality (PEQ) should be calculated utilizing valid statistical
procedures. The PEQ will then be compared with various
limitations based on the WLA. Limits will be specified in an
NPDES permit if any one of the following conditions is met:
30
5 43 A
-------
(1) The maximum reported discharge concentration exceeds the
level of detection and exceeds the limitation based on acute
aquatic life criteria.
(2) If the 99th percentile of the daily values exceeds the level
of detection and exceeds the limitation based on the acute
aquatic life criteria.
(3) If the 99th percentile of the monthly average exceeds the
level of detection and exceeds the limitation based on the
chronic aquatic life, human health, and/or wildlife
criteria.
(4) If the loads calculated from the PEQ data exceed 25% of the
WIA based on chronic aquatic life, human health, and/or
wildlife criteria (for free flowing streams) and the maximum
reported discharge concentration exceeds limits of detection
and exceeds 50% of the limitation based on acute aquatic
life criteria.
B. Determining Reasonable Potential with snail data sets. If less
than 10 representative effluent data samples are available for a
substance discharged from a point source, PEQ should be
calculated utilizing the following procedures:
(1) Determine the number of total observations ("n") for a
particular set of effluent data, and determine the highest
value from the data set.
(2) Using a coefficient of variation (CV) of 0.6, determine the
appropriate multiplying factor from Table 2.
(3) Multiply the highest value from a data set by the value from
Table 2 to calculate PEQ.
(4) Compare the PEQ to the WQBEL. If the PEQ is greater than
the WQBEL, an effluent limitation is necessary.
c. other <|ppli?flfrl« conditions. In addition to the above
procedures, permit limits may be required to comply with other
state and federal rules and regulations, such as, but not limited
to, pretreatment, categorical standards, etc. Also, when
determining whether permit limits are necessary, information from
chemical-specific, whole effluent toxicity and biological
assessments should be considered independently.
31
-------
Table 2
Reasonable Potential Multiply Factor
99% Confidence Level and 95% Probability Basis
Number of Samples Multiplying Factor
CV - 0.6
1 6.2
;; 3.8
-J 3.0
4 2.6
!> 2.3
6 2.1
7 2.0
8 1.9
9 1.8
13. VARIANCES FROM WATER QUALITY STANDARDS FOR POINTS SOURCES
A. Maximum fcJTf*frft!"? f9f Variances. The permitting authority may
grant a variance to a water quality standard (WQS) which was th
basis of a water quality-based effluent limitation included in
surface water discharge permit. The variance shall not exceed
years. This provision shall not apply to new sources.
B. conditions to grant a variance. A variance shall be granted if
the permittee demonstrates to the permitting authority that the
attaining the WQS is not feasible because:
(1) Naturally occurring pollutant concentrations prevent the
attainment of the WQS; or
(2) Natural, ephemeral, intermittent or low flow conditions or
water levels prevent the attainment of the WQS, unless these
conditions may be compensated for by the discharge of
sufficient volume of effluent discharges without violating
State water conservation requirements to enable WQSs to be
st; or
(3) Human caused conditions or sources of pollution prevent the
attainment of the WQS and cannot be remedied or would cause
more environmental damage to correct than to leave in place;
or
(4) Dams, diversions or other types of hydrologic modifications
preclude the attainment of the WQS, and it is not feasible
to restore the water body to its original condition or to
32
-------
operate such modification in a way that would result in the
attainment of the WQS; or
(5) Physical conditions related to the natural features of the
water body, such as the lack of a proper substrate cover,
flow, depth, pools, riffles, and the like, unrelated to WQ,
preclude attainment of WQSs; or
(6) Controls more stringent than those required by Sections
301(b) and 306 of the Clean Water Act would result in
substantial and widespread economic and social impact; and
the extent of any increase risk to human health and the
environment associated with the variance is determined and
is consistent with the protection of the public health,
safety and welfare.
C. Tlf^F*1"* to submit application. The permittee shall submit an
application for a variance within 60 days after the permitting
authority reissues or modifies the permit. The application shall
include all relevant information necessary to make a
demonstration based on one or more of the conditions in Part B,
above.
D. Public Notice of Tentative Decision. Upon receipt of a complete
application for a variance, the permitting authority shall public
notice the request and tentative decision regarding the variance
and shall notify the other Great Lakes States of the tentative
decision.
E. ripalDecision on VarianceRequest. The permitting authority
shall issue a final decision on the variance request within 90
days of the expiration of the public comment period as required
in Part D, above. If all or part of the variance is approved,
the decision shall include all permit conditions needed to
implement the variance. The permitting authority shall deny a
requested variance if the permittee fails to make the
demonstration required under Part B of this Procedure.
F. Applicability. A variance applies only to the permittee
requesting the variance and only to the substance specified in
the variance. A variance does not affect or require the
permitting authority to modify the corresponding water quality
standard.
G. Incorporating Approved Variance into Permit. The permitting
authority shall initiate a permit modification to establish all
conditions needed to implement the variance. A permit
modification at a minimum shall require:
(1) compliance with an initial effluent limitation which at the
time the variance is granted represents the level currently
achievable by the permittee; and
33
-------
(2) that reasonable progress be made toward attaining the vat
quality standards through
(a) a mechanism such as the establishment of a capital
improvements fund, where appropriate, and
(b) continued investigation of treatment technologies,
process changes, pollution prevention, wastewater reuse or
other techniques that will reduce the level of the substance
or result in compliance by the permittee with the WQS and
submission of reports on the investigations at such time
specified by the permitting authority.
H. Renewal of Variance. A variance may be renewed, subject to the
requirements of Parts B. through 6., above. A variance may not
be renewed if the permittee did not substantially comply with the
conditions of the variance.
I. U.S. EPA Approval. All variances must be approved by the U.S.
EPA.
J. State wpe Revisions. All variances must be appended to the State
WQS rules.
14. ENVIRONMENTAL PATE
A. Mixing gone Degradation. WQBELs may be modified to account for
degradation or attenuation of toxicity of a substance inside the
mixing zone, if the following conditions are met:
(1) The following information is provided to the permitting
authority:
(a) The rate of degradation inside the mixing zone is
documented by field studies supplied by the permittee;
(b) The f-ield studies demonstrate rapid and significant
loss of the substance inside the mixing zone under the
full range of critical conditions expected to be
encountered; and
(c) The study includes other factors affecting the water
column such as but not limited to resuspension of
sediments, speciation, and transformation.
(2) The field studies are reviewed and approved by the
permitting authority.
(3) The degradation of the pollutant is not achieved by the
transfer of the pollutant to other media.
34
-------
B. Bediments. WQBELs may be reduced to prevent contamination of
sediment with toxic substances or to prevent accumulation of the
substance in sediments if determined necessary to protect water
quality. This applies whether the deposition occurs in sediments
located inside or outside the mixing zone of the source.
15. GENERAL. The permitting authority may require that additional
data be provided to meet the minimum data requirements, when
otherwise not available, for derivation of Tier II levels of
concern developed pursuant to the aquatic life, wildlife, and
noncarcinogenic human health procedures of the Great Lakes
Initiative. However, the permitting authority may determine that
this minimum data gathering provision is not appropriate if the
available environmental fate and/or toxicity data (including
qualitative structure activity relationship information) indicate
that a substance has no reasonable potential for exposure in
water resulting in adverse impacts to aquatic life, wildlife and
human health.
35
-------
GL WOIANTIDEGRADA TION MODEL REGULA TION - DRAFT 9/4/91
The model regulation is constructed in the following sequence
(a) Antidegradation Policy is a statement of the general policy with regard to maintenance
and protection of water quality in the Great Lakes System It is generally the same as the
national regulation, excupt that it more broadly defines High Quality Waters which are
subject to protection under antidegradation, and it prohibits the lowering of water quality m
impaired waters
(b) Antidegradation Implementation Procedures define the procedures to be used by the
Great Lakes States to implement the general Policy. In particular, this section identifies
priorities of the Great Lakes States with regard to bioaccumulative chemicals of concern
and situations m which the lowering of water quality will be considered significant and
subject to a detailed antidegradation review Subparagraphs (b)(4) through (b)(6) establish
procedures specific to the situation covered in the Policy Subparagraph (b)(6) specifically
addresses maintenance of high quality waters with respect to d) bioaccumulative chemicals
of concern, and (n) other pollutants
(c) Antidegradation Demonstration defines the information that an entity that is seeking to
lower water quality must provide in support of that request It promotes pollution
prevention and requires, that entities develop information regarding the costs associated
with eliminating or reducing the extent of the lowering of water quality
(d) Antidegradation Decision identifies the process that the State will follow in evaluating
the information provided pursuant to (c) and in reaching a decision on the lowering of water
quality The procedure establishes policy on minimum expenditures that will be required of
an entity to eliminate or reduce the extent of the lowering of water quality It further
requires that the State considers the social and economic benefits in light of the
environmental effects associated with the lowering of water quality in order to reach the
decision The State may either conduct a full review of the technical merit of the
demonstration and make its tentative decision accordingly, or alternatively, the State may
choose to determine that the administrative requirements of
-------
GL WQIANTIDEGRADA TION MODEL REGULA TION - DRAFT 9/4/91
{a) Ant/degradation Policy This antidegradation policy shall be applicable to any source, point or
nonpomt, of pollutants to surface waters of the Great Lakes System Pursuant to this policy
(1) Existing instream water uses, as defined pursuant to 40 CFR 131, and the level of
water quality necessary to protect existing uses shall be maintained and protected Where
designated uses of the water body are impaired, there shall be no lowering of the water
quality with respect to the pollutant or pollutants which are causing the impairment Where
uses specified in Section 101 (a I of the Clean Water Act are impaired, there shall be no
lowering of water quality with respect to the pollutant or pollutants causing the impairment,
(2) Where, for any parameter, the water quality exceeds that level necessary to meet the
applicable water quality standards, that water shall be considered high quality for that
parameter and that quality shall be maintained and protected unless the State finds, after
full satisfaction of intergovernmental coordination and public participation provisions of the
State's continuing planning process, that allowing lower water quality is necessary to
accommodate important economic or social development in the area in which the waters
are located In allowing such degradation, the State shall assure water quality adequate to
protect existing uses fully Further, the State shall assure that there shall be achieved the
highest statutory and regulatory requirements for all new and existing point sources and all
cost effective and reasonable best management practices for nonpomt source controls,
(3} Where high quality waters constitute an outstanding National resource, such as waters
of National and State parks and wildlife refuges and waters of exceptional recreational or
ecological significance, that water quality shall be maintained and protected, and
(4) In those cases where the potential lowering of water quality is associated with a
thermal discharge, the decision to allow such degradation shall be consistent with Section
316 of the Act
(b) Antidegradation Implementation Procedures
(1) Applicability These procedures are applicable to actions that result in, or have the
potential to result in, a significant lowering of water quality relative to the water quality
conditions that exist prior to the implementation of such action The decision-making
procedures described in this section shall be applied on a pollutant-specific basis to any
situation to which they are applicable
(2) Exemptions. Except as the Director may determine on a case-by-case basis that the
application of these procedures is required to adequately protect water quality, or as the
affected water body is an outstanding National resource water, the procedures in this part
do not apply to
(0 actions that do not result in a significant lowering of water quality,
(n) actions, such as remedial actions pursuant to the Comprehensive Environmental
Response, Compensation and Liability Act, as amended, or corrective actions
pursuant to the Resource Conservation and Recovery Act, as amended, which have
the effect of improving water quality in a surface water body even where the action
may result in the short-term, temporary (less than one year) significant lowering of
water quality in the same surface water body,
(in) bypasses as defined at 40 CFR 122 41 (m); and
-------
GL WQIANTIDEGRADA TION MODEL REGULA TION - DRAFT 9/4/91
(iv) response actions pursuant to the Comprehensive Environmental Response,
Compensation and Liability Act, as amended, undertaken to alleviate a release into
the environment of hazardous substances, pollutants or contaminants which may
pose an imminent and substantial danger to public health or welfare
(3) Definitions
(i) Pollutant The term pollutant is as defined in Section 502 of the Clean Water
Act and includes toxic, conventional and nonconventional pollutants, and
bioaccumulative chemicals of concern as they are defined in this section
(n) Significant lowering of water quality A significant lowering of water quality
occurs when any of the following conditions exist
(A) there is an increase in the mass loading, in excess of the loading
associated with the effluent quality existing prior to the action, of any
bioaccumulative chemical of concern to the surface water from an action by
the permittee at an existing, expanding or new point source,
(B) thei e is an increase in the mass loading, in excess of that existing prior
to the action, of any bioaccumulative chemical of concern to the surface
water ftom an action by the regulated entity at an existing, expanding or
new nonpomt source,
(C) there is an increase, other than a de miminis increase, in the permitted
loading of any pollutant that is not a bioaccumulative chemical of concern to
the surface water from an action by the permittee at an existing, expanding
or new point source,
(D) there is an increase in the loading of any pollutant that is not a
bioaccumulative chemical of concern from a nonpomt source, which is in
excess of the loading incorporating all applicable requirements of the
governing nonpomt source program,
(E) there is any lowering of water quality for any pollutant in an
Outstanding National Resource Water, or
IF) for any other action, where such action is determined by the Director,
on a case-by-case basis, to be significant
(in) Bioaccumutath/e chemical of concern A bioaccumulative chemical of concern
is: any chemical which, upon entenng the surface waters, by itself or as its
transformation product, bioaccumulates in aquatic organisms by a factor greater
than 1000 The bioaccumulation factors used in identifying bioaccumulative
chemicals of concern will be determined according to paragraph (cite BAF
procedurelof this part For the purposes of this part, bioaccumulative chemicals of
concern will include, at a minimum, the substances in Table of Appendix [list
of GLWQI chemicals of concern with BAFs > 10001, but may include additional
substances identified by the Director using the criteria identified above
(iv) Outstanding National Resource Waters Outstanding National Resource Waters
shall be those designated as such by the states Categories of waters which are
-------
GL WO1ANTIDEGRADA TION MODEL REGULA TION • DRAFT 9/4/91
eligible for designation include but are not limited to the following four categories,
which are waters recognized as
(A) important because of protection through official action, such as Federal
or State law, Presidential or Secretarial action, international treaty, or
interstate compact,
(B) having exceptional recreational significance,
(C) having exceptional ecological significance,
(D) having other special environmental, recreational, or ecological
attributes, or
(E) waters whose designation as Outstanding National Resource Waters is
reasonably necessary for the protection of waters identified in
subparagraphs (A) through (D) above
(v) High Quality Waters High quality waters are those that satisfy the criteria
specified in paragraph (a) (2) above regarding the quality of the water
(vi) De mintmis The lowering of water quality by a pollutant will be considered de
mimmis if it satisfies all of the following criteria for the pollutant under
consideration, and such a determination is consistent with applicable
implementation requirements and limitations pursuant to the Great Lakes Critical
Programs Act
(A) the lowering of water quality does not involve a bioaccumulative
chemical of concern,
(B) the lowering of water quality uses less than 10 percent of the
unallocated wasteload allocation or load allocation, as appropriate, which
remained as of the effective date of this regulation;
(C) a minimum of 50 percent of the unallocated wasteload allocation or
load allocation, as appropriate, that existed as of the effective date of this
regulation remains after accommodating the lowering of water quality, and
(D) no more than 90% of the total wasteload allocation or load allocation,
as appropriate, has been allocated, where
(E) the wasteload allocation or load allocation evaluated in the decision is
developed pursuant to [cite Implementation Procedure #51 for the pollutant
under consideration in the water body in the area where the water quality is
proposed to be lowered
(4) For all waters, the Director shall ensure that the level of water quality necessary to
protect existing uses is maintained In order to achieve this requirement, and consistent
with 40 CFR 131 10, water quality standards use designations must include alt existing
uses Controls shall be established as necessary on point and nonpomt sources of
pollutants to ensure that the criteria applicable to the designated use are achieved in the
water
-------
GLWQJ ANTIDEGRADAT1ONMODEL REGULATION• DRAFT9/4/97
(5) (0 For Outstanding National Resource Waters, the Director shall ensure, through
the application of appropriate controls on pollutant sources, that water quality is not
lowered, significantly or otherwise, as the result of any action
(n) Exception A short term, temporary (less than 1 year) lowering of water quality
may be permitted by the Director
(6) For high quality waters, the Director shall ensure that no significant lowering of water
quality occurs except as the action resulting in the significant lowering of water quality
satisfies the conditions of paragraph (c) of this part regarding completion of an
antidegradation demonstration and the information thus provided is determined by the
Director to adequately support the significant lowenng of water quality
d) To prevent the significant lowenng of water quality that would result from any
increased rate of loading of a bioaccumulative chemical of concern from any source,
the Director sh
-------
GL WQJ ANTIDEGRADA TION MODEL REGULA TION - DRAFT 9/4/9 1
0 ) Pollution Prevention and Alternatives Analysis. Identify such alternatives and
techniques that are available to the entity that would eliminate or reduce the extent of the
significant lowering of water quality
d) Alternatives that must be evaluated include, but are not limited to
(A) Substitution of bioaccumulative chemicals of concern with non-
bioaccumulative and/or non-toxic substances
(B) Application of water conservation methods
(C) Waste source reductions within process streams
(D) Recycle/reuse of waste by-products, either liquid, solid, or gaseous
(E) Alternative best available treatment scientifically available applied to
waste water streams
(F) Manufacturing process operational changes
(11) The evaluation of alternatives and treatment options shall identify the most cost
effective means of eliminating the significant lowering of water quality In addition,
the evaluation shall define the incremental costs of the removal of each pollutant
associated with the significant lowering of water quality, up to 1 00 percent
removal, where achievable
(2} Mandatory Pollution Control Expenditures Identify such alternatives as are available at
a cost less than or equal to that at which the ratio of the annualized cost of the waste water
pollution control to the cost of the expansion or development is 1 percent, and the extent
to which the significant lowering of water quality will be lessened by each alternative
identified Should the mandatory implementation of alternatives identified in this paragraph
eliminate the significant lowering of water quality, the entity shall not be required to provide
information specified in paragraphs (c)(3) through (5)
(3) Additional Affordable Pollution Control Expenditures Identify such affordable
alternatives in addition to those identified in paragraph (cH2) of this part that are available
and the extent to which the significant lowenng of water quality will be lessened by each
Affordable alternatives are those alternatives that are available at a cost less than or equal
to that where the ratio of annualized pollution control costs to the value of the entire annual
production or operations is 1 percent Should the implementation of alternatives identified
m this paragraph eliminate the significant lowenng of water quality, the entity shall not be
required to provide information specified in paragraphs
-------
GL WQIANTIDEGRADA TION MODEL REGULA TION - DRAFT 9/4/91
(HI) reduction in the unemployment rate or other social service expenses,
(iv) increase in tax revenues
(5) Environmental Effet ts Identify the effect of the significant lowering of water quality
after incorporating all affordable pollution control alternatives Such evaluation shall
consider the pollutant or pollutants involved, the applicable standards for the pollutant,
existing ambient levels of the pollutant or pollutants in water, sediments and biota, and any
other factors as required by the Director
(d) Antidegradatton Decision
(1) Once the Director determines that the information provided by the entity pursuant to
paragraphs (c)(1) through (5) of this part is administratively complete he shall use the
information, as follows, to determine the extent to which water quality may be lowered by
the entity in no event may the decision reached under this part allow the water quality to
be lowered below the minimum level required to support existing uses fully
(i) At a minimum, the entity requesting to significantly lower water quality shall be
required to implement the alternative identified pursuant to paragraph (c)(2) of this
part that most effectively eliminates or reduces the need to significantly lower
water quality Should no alternatives have been identified pursuant to paragraph
(c)(2) of this part that will eliminate the significant lowering of water quality, the
entity may still be required to implement controls identified in paragraph (c)(3) of
this part
(H) If the information provided pursuant to paragraph (c)(3) of this part indicates
that there exist pollution control alternatives available to the entity and which lessen
the significant lowering of water quality more than the mandatory alternative
identified pursuant to paragraph (c)<2), the Director shall require the implementation
of the most effective alternative or combination of alternatives, which the Director
determines are affordable Such alternatives may incorporate the alternative
required pursuant to (0 of this paragraph, or may substitute for it, as the Director
determines would be most appropriate to prevent or lessen the significant lowering
of water quality
(HI) Should the evaluation in In) of this paragraph indicate that no significant
lowering of water quality is necessary to accommodate the action, the Director shall
establish control requirements applicable to the action that prohibit the significant
lowering of water quality
(iv) Should th« evaluation in dO of this paragraph indicate that some significant
lowering of water quality is still necessary to accommodate the proposed action,
the Director shall consider the social or economic developments associated with the
action identified pursuant to paragraph (c)(4| of this part and the environmental
effects of the ugnificant lowenng of water quality identified pursuant to paragraph
(cH5) of this part. Based on this analysis, the Director shall determine if the
significant lowering of water quality should be proposed to be allowed
(v) The Director may choose to defer the review in dHiv) of this paragraph until
after the public is provided the opportunity to comment, subject to the conditions of
(2)00 of this paragraph.
6 5-5--T
-------
GL WQIANTIDEGRADA TION MODEL REGULA TION - DRAFT 9/4/91
(2) The tentative decision of the Director regarding the extent to which water quality may
be significantly towered shall be subject to the public participation requirements of 40 CFR
25 To the extent that the tentative decision is embodied in the conditions of a National
Pollutant Discharge Elimination System permit, the public participation requirements may be
satisfied by the public notice of the draft permit and fact sheet which discusses that
antidegradation demonstration and decision regarding the significant lowering of water
quality
d) If the Director's decision is based on the analysis in (IHni) or (iv) of this
paragraph, then the public notice of the decision shall define the extent of
significant lowering of water quality tentatively determined by the Director to be
allowable, and the factors considered in reaching that decision.
(11) If the Director chooses to defer the review as provided in <1}
-------
|