PESTICIDE REREGISTRATION
REJECTION RATE ANALYSIS
ECOLOGICAL EFFECTS
-------
-------
Table of Contents
Introduction , 1
Scope of analysis 1
Process 2
Description of the Discipline . . 3
Description of Studies 5
Industry Comments on "Description of Discipline" Section ................ 26
References 34
Rejection Rate Analysis - General Issues Raised by Industry 34
Current Rejection Rate 36
Summary 37
Rejection Rate Graphs 38
Rejection Factors 44
Summary Table of Rejection Factors 162
Unresolved Issues 165
Conclusions 166
Recommendations 167
Appendix A: Additional Supporting Documents . 168
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance ......... 169
Attachment 1, Memorandum: "Reevaluation of Previously Rejected Mollusk Shell
Deposition Studies, 72-3(b)," Oct. 29, 1992, from Paul Shuda to Dan Barolo . . 184
Rapid Feedback SOP 187
Appendix B: List of EPA Guidance Documents 188
-------
-------
List of Figures
Figure 1 List A - Rejection Rates for Ecological Effects Guideline Requirements
Figure 2 List A - Current (Post) 1988 Rejection Rates by Guideline - Avian Acute and
Reproduction Toxicity
Figure 3 List A - Current (Post 1988) Rejection Rates by Guideline - Fish Acute and
Reproduction Toxicity
Figure 4 List A - Current (Post 1988) Rejection Rates by Guideline - Other
Figure 5 List A - Ecological Effects Guidelines with Lower Rejection Rates Over Time
Figure 6 List A - Ecological Effects Guideline Rejection Rates Without Consistent
Improvement Over Time
-------
-------
Introduction
This rejection rate analysis has been undertaken by the Special Review and Reregistration
Division (SRRD) and the Environmental Fate and Effects Division (EFED) in the Office of
Pesticide Programs (OPP) of the Environmental Protection Agency (EPA or "the Agency").
The purpose of this guideline-by-guideline analysis is to identify those factors that most
frequently cause rejection of ecological effects studies required for reregistration. This
information will enable OPP to (a) provide registrants with information on rejection factors
to minimize their reoccurrence in future studies, (b) reassess the adequacy of its guidance,
(c) determine the appropriate regulatory response to a future rejected study, and (d) make
any internal changes in process, procedures, or criteria deemed appropriate.
The decision to analyze these factors was made after a FIFRA Reregistration recosting
analysis, conducted in the spring of 1991, indicated that rejected studies posed the most
significant potential for delays in the production of Reregistration Eligibility Documents
(REDs). Reregistration eligibility decisions require that reasonable risk assessments be
performed for all relevant human health and ecological end points for each chemical.
Performing such risk assessments requires a substantially complete data base. A substantially
complete data base requires that registrants submit studies of acceptable quality. Also, a
significant reduction in rejection rates for most disciplines is required for OPP to be able to
meet its production schedule for REDs.
Scope of analysis
The scope of this analysis is limited almost entirely to an examination of rejected studies.
While a scientist's review could result in a finding of acceptable, upgradable, unacceptable,
or supplemental, rejected (i.e., unacceptable) studies are the focus here because a rejected
study will more than double the amount of time and resources required to satisfy that
guideline. Upgrading usually does not require as much time to accomplish as repeating the
study. A rating of supplemental by a scientist might require substantial new work and add
additional time delays to the process.
The scope of this analysis is also limited primarily to List A studies, although for some
guidelines rejected studies from List B were used to augment the sample size for the rejection
factor identification process. The analysis was confined primarily to List A chemicals
because List A chemicals: (1) represent those chemicals with the longest reregistration
history—each chemical case had a Registration Standard published between 1980-1988;
(2) are high-volume, food-use chemicals that could pose the greatest potential risk for human
health and the environment, and therefore have the highest priority in reregistration; and
(3) generate the most extensive data requirements.
To what extent are List A rejection factors representative of lists B, C, and D?
Unfortunately, it is not possible at this time to make such a determination, since a random
-------
sample of List A, B, C, and D studies was not chosen as the basis for this analysis. Such a
sample was not feasible since List B chemicals completed Phase 4 (in fiscal year (FY) 91),
List C chemicals completed Phase 4 in FY 92, and List D chemicals completed Phase 4 in
FY 93. Consequently, at the time when the rejection rate project began in 1991, there was
not an adequate pool of reviewed studies across lists for each guideline to support a randomly
drawn data base. Furthermore, many List B and C study reviews conducted in Phase 4 were
based on examination of the summaries -only.
The rejection factors identified in this assessment of List A rejected studies could
plausibly either overstate or understate the number of rejection factors likely to be found in
any future assessment of List B, C9 and D rejected studies. On the one hand, many List A
studies were initiated in response to the Registration Standards prior to both the 1982
guidelines and development of acceptance criteria in Phase 3 (1989), and consequently may
have been rejected by criteria that were not in place at the time the study was conducted. In
this case, the corresponding rejection factors are not likely to be repeated in List B, C, and
D studies, since the data call-ins have all been issued subsequent to OPP's publication of its
guidelines and acceptance criteria. On the other hand, many of the studies judged to be
acceptable now may be repeat studies. Consequently, the rejection factors identified here may
omit factors that were responsible for previous submissions being rejected.
Process
The Agency first reviewed the data evaluation records (study reviews) on a guideline-by-
guideline basis in order to:
(1) identify those factors that most frequently caused each guideline study to be rejected;
(2) determine the rejection rates and trends (where the sample size was adequate) for each
guideline requirement;
(3) assess the adequacy of EPA's guidance documents with respect to each rejection
factor; and
(4) determine if each rejection factor is avoidable.
Second, a draft was provided to an industry work group of scientists for review and
comment, in order to (1) obtain from a user's perspective the adequacy of EPA's guidance
documents corresponding to each rejection factor, and (2) better understand why the rejection
factors occur. The industry work group included: Richard Balcomb, CD3A; Richard Brown,
Zeneca; Peg Cherney, Rhone Poulenc; Joseph Dulka, DuPont; Reinhard Fischer, AgrEvo;
Robert Graney, Miles; Richard Holt, DuPont; Catherine Holmes, BASF; Michael McKee,
Monsanto; Ellen Mihaich, Rhone Poulenc; and John Schupner, AgrEvo. Industry and EPA
scientists met on April 7 and .8, 1994, to discuss the problem areas in order to develop a
-------
better understanding of them. The revised ecological effects chapter explicitly includes
industry comments on each rejection factor, and EPA's response to them.
Description of the Discipline
Ecological effects data are used by the Agency to determine the toxicological hazards of
pesticides to various terrestrial and aquatic nontarget organisms. These effects data are
compared to environmental fate and exposure data when the Agency performs a risk
characterization for a pesticide use.
The eco-effects tests include acute, subacute, chronic, and field studies, which are part of
a testing scheme that moves from the basic acute laboratory studies to chronic and applied
field studies and, essentially, from the least difficult studies to perform to the most difficult.
The testing scheme is a tiered one; i.e., results from one tier are evaluated to determine
potential toxicological hazards and if further testing is required in the higher tiers. In this
system, surrogate organisms are used. ' Special testing (guidelines 70-1) may be required, but
generally the tests move from the least expensive to the most expensive, and the adverse
effects examined include the following: mortality, reduction in growth, reproductive
impairment, changes in numbers of species, bioaccumulation of residues in nontarget
organisms, and in the highest tier studies, structure and function changes in the ecosystem.
The Ecological Effects Branch (EEB), in conjunction with the Environmental Fate and
Ground Water Branch (EFGWB), performs ecological risk characterizations based upon eco-
effects and environmental fate data submitted to the Agency. For pesticide registration or
reregistration, ecological risk characterizations generally consist of the following major
activities:
(1) Review and evaluation of eco-toxicity data submitted to support the registration or
reregistration of pesticides and their uses.
(2) Establishment of the endpoints of concern and the overall toxicity of the pesticide,
and its use(s) to non-target, non-human organisms based upon the data submitted and
evaluated.
(3) Calculation of risk quotients based upon the eco-toxicity data and the pesticide use
data, fate and transport data, and estimates of exposure.
RISK QUOTIENT = EXPOSURE / TOXICITY
(4) Comparison of the risk quotients to established regulatory Levels of Concern (LOCs).
An LOG is a measure of risk to nontarget organisms that may lead to regulatory
action. Several ecological LOCs are used in OPP regulatory decision-making.
-------
Specific levels of concern are defined in the table below. EEC is defined as the
estimated environmental concentration.
Endpoint & Scenario
Mammalian acute (granular)
Mammalian acute (spray)
Mammalian chronic (granular)*
Mammalian chronic (spray)
Avian acute (granular)
Avian acute (spray)
Avian chronic (granular)*
Avian chronic (spray)
Aquatic acute
Aquatic chronic
Non-target insects
Non-target plants
Risk Quotient
LDso/FT2
EEC/LC50
EEC/NOEL
EEC/NOEL
LDso/FT2
EEC/LC50
EEC/NOEL
EEC/NOEL
EEC/LC50
EEC/NOEL**
NOT
QUANTIFIED
NOT
QUANTIFIED
LOC - Non-
Endangered
0.5
0.5
1.0
1.0
0.5
0.5
1.0
1.0
0.5
1.0
LOC-
Endangered
0.1
04
1.0
1.0
0.1
0.1
1.0
1.0
0.05
1.0
* There are no standard procedures for estimating chronic exposure levels for granular
pesticides.
** It is the goal of the Agency to regulate on the MATC where warranted. However,
depending upon study results, the NOEL may be deemed appropriate for use in the risk
assessment.
-------
Description of Studies
The following is a description of the ecological effects guideline studies required by the
Agency to support registration.
Wildlife and Aquatic Organisms
Avian Oral LD50 (guidelines 71-1): The avian oral LD50, using either an upland game
bird (e.g., bobwhite quail) or a waterfowl species (e.g., mallard duck), is an acute, single-
dose laboratory study designed to determine the quantity of toxicant required to cause fifty
percent mortality in a test population of birds. Technical grade active ingredient (TGAI) is
administered by oral intubation to adult birds, and the results (expressed as LD50 milligrams
(mg) active ingredient (a.i.)/kilogram (kg)) are used to make the following determinations:
Value of Information (questions answered):
(1) Determine the category of toxicity to birds (mg/kg):
< 10 very highly toxic
10-50 highly toxic
51-500 moderately toxic
501-2000 slightly toxic
> 2000 practically non-toxic
(2) Determine the need for precautionary label statements. (If LD50< 100mg/kgs then
"This pesticide is toxic to birds" is required.)
(3) Determine acute dose quotients (Q = quotient) (EEC (mg active ingredient available
on the surface)/LD50/bird = LD50/ft2).
(4) Establish the level of minimal acute risk to birds. (If Q < 0.2, then the pesticide has
a minimum acute risk to birds, and no additional data are required.)
(5) Determine the need for Restricted Use classification. (If Q > 0.2, but < 0.5, then
risk may be mitigated by Restricted Use classification.)
(6) Establish the level of acute risk to endangered birds (If Q > 0.1, then consultation
with U.S. Fish and Wildlife Service is required.)
(7) Determine the need for further regulatory action (If Q > 0.5, then further regulatory
action may be required, which could include mitigation, field test(s), screening study
71-5.)
-------
(8), Establish the level of high acute risk to birds (If Q > 0.5, then the pesticide has a
potential high acute risk to birds.)
Guideline Uses (Her Progression):
(1) If Q < 0.2, then there are no additional data requirements.
(2) If Q > 0.5, further regulatory action may be required, which may include field
test(s), screening study 71-5.
Regulatory Use(s):
(1) If Q < 0.2, then there is minimal acute risk to birds, and there are no additional data
requirements; recommend registration/reregistration proceed.
(2) If Q > 0.2, but < 0.5, potential acute risk to birds may be mitigated by Restricted
Use classification.
(3) If Q > 0.5, there is potential high acute risk to birds; further regulatory action may
be required, which may include mitigation, field test(s), screening study 71-5.
(4) If Q > 0.5, and bird incident reports or field tests with bird kills are available, the
recommend pesticide for special review or other risk-benefit balancing.
Avian Dietary LC50 (guidelines 71-2):
The avian dietary LC50, using both an upland game bird (bobwhite quail) and a waterfowl
species (mallard duck), is an acute, eight-day dietary laboratory study designed to determine
the quantity of toxicant required to cause fifty percent mortality in a test population of birds.
TGAI is administered by mixture into juvenile birds diets for five days (followed by three
days "clean" diet), and the results (expressed as LC50 parts per million (ppm) active
ingredient (a.i.)) are used to make the following determinations:
Value of Information (questions answered):
(1) Determine the category of toxicity to birds (ppm):
<50
50-500
501-1000
1001-5000
>5000
very highly toxic
highly toxic
moderately toxic
slightly toxic
practically non-toxic
-------
(2) Determine the need for precautionary label statements. (If LC50 < SOOppm, then
"This pesticide is toxic to birds" is required.)
(3) Determine acute dietary quotients (Q = quotient) for non-granular formulations (EEC
(residue levels in avian food items from available field studies or Kenaga,
1973)/LC50).
(4) Establish the level of minimal acute risk to birds. (If Q < 0.2, then the pesticide has
a minimum acute risk to birds, and no additional data are required.)
(5) Determine the need for Restricted Use classification. (If Q > 0.2, but < 0.5, then
acute risk may be mitigated by Restricted Use classification.)
(6) Establish the level of acute risk to endangered birds. (If Q > 0.1, then consultation
with U.S. Fish and Wildlife Service is required.)
(7) Determine the need for further regulatory action. (If Q ^ 0.5, then further
regulatory action may be required which may include mitigation, field test(s),
screening study 71-5.)
(8) Establish the level of high acute risk to birds. (If Q > 0.5, then the pesticide has a
potential high acute risk to birds.)
Guideline Uses (Tier Progression):
(1) If Q < 0.2, then there are no additional data requirements.
(2) If Q > 0.5, further regulatory action may be required, which may include field
test(s), screening study 71-5.
Regulatory Use(s):
(1) If Q < 0.2, then there is minimal acute risk to birds and no additional data
requirements; recommend registration/reregistration proceed.
(2) If Q > 0.2, but < 0.5, potential acute risk to birds may be mitigated by Restricted
Use classification.
(3) If Q > 0.5, then there is potential high acute risk to birds and further regulatory
action may be required, which may include mitigation, field test(s), screening study
71-5.
(4) If Q > 0.5, and bird incident reports or field tests with bird kills are available,
recommend pesticide for special review or other risk-benefit balancing.
-------
Freshwater Fish LC50 (guidelines 72-1):
The freshwater fish LC50, using both a cold water (rainbow trout) and a warm water
(bluegill) species, is an acute, ninety-six-hour laboratory study designed to determine the
concentration in water required to cause fifty percent mortality in a test population of fish.
TGAI1 is administered into water containing fish, providing exposure for ninety-six hours,
and the results (expressed as ppm a.i.) are used to make the following determinations:
Endpoint:
Mortality: LC50 value in ppm with 95% confidence limits, lethal concentration in water
likely to kill 50% of the cold water and warm water fish exposed
Value of Information (questions answered):
(1) Determine the category of toxicity to fish (ppm):
<0.1 very highly toxic
0.1-1 highly toxic
> 1-10 moderately toxic
> 10-100 slightly toxic
> 100 practically non-toxic
(2) Determine the need for precautionary label statements. (If LC50< 1 ppm, then "This
pesticide is toxic to fish" is required.)
(3) Determine acute quotients (Q = quotient) for all formulations (EEC (residue levels in
water from back-of-the-envelope calculations or fate models)/LC50)
(4) Establish the level of minimal acute risk to fish. (If Q < 0.1, then the pesticide has a
minimum acute risk to fish and no additional data are required.)
(5) Determine the need for Restricted Use classification. (If Q > 0.1, but < 0.5, then
acute risk may be mitigated by Restricted Use classification.)
(6) Establish the level of acute risk to endangered fish, (If Q > 0.05, then consultation
with U.S. Fish and Wildlife Service is required.)
(7) Determine the need for chronic testing, fish early life-stage test 72-4. (If Q > 0.01,
then fish early life-stage test 72-4 is triggered.)
1 Depending on certain criteria, testing with typical end-use product (TEP) may be done.
8
-------
(8) Determine the need for further regulatory action. (If Q > 0.5, then further
regulatory action may be required, which may include mitigation, field test(s),
mesocosm, or actual aquatic field study 72-7.)
(9) Establish the level of high acute risk to fish. (If Q > 0.5, then the pesticide has a
potential high acute risk to fish.)
Guideline Uses (Tier Progression):
(1) If Q < 0.1, then there are no additional data requirements.
(2) If Q > 0.5, further regulatory action may be required, which may include field
test(s), mesocosm, or actual aquatic field study 72-7.
Regulatory Use(s):
(1) If Q < 0.1, then there is minimal risk and no additional data requirements;
recommend registration/reregistration proceed.
(2) If Q > 0.1, but < 0.5, potential risk may be mitigated by Restricted Use
classification.
(3) If Q > 0.5, there is potential high acute risk and further regulatory action may be
required, which may include mitigation, field test(s), mesocosm, or actual aquatic
field study 72-7.
(4) If Q > 0.5, and incident reports or field tests with fish kills are available,
recommend pesticide for special review or other risk-benefit balancing.
Acute LC50 Freshwater Invertebrates (guidelines 72-2):
The freshwater invertebrate vLC5o/EC5o, using a freshwater invertebrate (Daphnia sp.), is
an acute, forty-eight-hour laboratory study designed to determine the concentration in water
required to cause fifty percent mortality, or immobilization, in a test population of
invertebrates. TGAI (see footnote 1) is administered into water containing invertebrates,
providing exposure for forty-eight hours, and the results (expressed as ppm a.i.) are used to
make the following determinations:
Endpoints:
Mortality: LC50 value in ppm with 95 % confidence limits; lethal concentration in water
likely to kill 50% of the aquatic invertebrates exposed; or,
-------
Immobilization: EC50 in ppm with 95% confidence limits; concentration in water likely to
immobilize 50% of the aquatic invertebrates exposed.
Value of Information (questions answered):
(1) Determine the category of toxicity to aquatic invertebrates (ppm):
<0.1 very highly toxic
0.1-1 highly toxic
> 1-10 moderately toxic
> 10-100 slightly toxic
> 100 practically non-toxic
(2) Determine the need for precautionary label statements. (If LC50< 1 ppm, then "This
pesticide is toxic to aquatic invertebrates" is required.)
(3) Determine acute quotients (Q = quotient) for all formulations (EEC (residue levels in
water from back-of-the-envelope calculations or fate models)/LC50).
(4) Establish the level of minimal acute risk to aquatic invertebrates. (If Q < 0.1, then
the pesticide has a minimum acute risk and no additional data are required.)
(5) Determine the need for Restricted Use classification. (If Q > 0.1, but < 0.5, then
acute risk may be mitigated by Restricted Use classification.)
(6) Establish the level of acute risk to endangered aquatic invertebrates. (If Q > 0.05,
then consultation with U.S. Fish and Wildlife Service is required.)
(7) Determine the need for chronic testing, invertebrate life cycle test 72-4. (If Q >
0.01, then invertebrate life cycle test 72-4 is triggered.)
(8) Determine the need for further regulatory action. (If Q > 0.5, then further
regulatory action may be required, which may include mitigation, field test(s),
mesocosm, or actual aquatic field study 72-7.)
(9) Establish the level of high acute risk to aquatic invertebrates. (If Q > 0.5, then the
pesticide has a potential high acute risk to aquatic invertebrates.)
Guideline Uses (Tier Progression):
(1) If Q < 0.1, then there are no additional data requirements.
(2) If Q > 0.5, further regulatory action may be required, which may include field
test(s), mesocosm, or actual aquatic field study 72-7.
10
-------
Regulatory Use(s):
(1) If Q < 0.1, there is minimal risk and no additional data requirements; recommend
that registration/reregistration proceed.
(2) If Q > 0.1, but < 0.5, potential risk may be mitigated by Restricted Use
classification.
(3) If Q > 0.5, there is potential high acute risk and further regulatory action may be
required, which may include mitigation, field test(s), mesocosm, or actual aquatic
field study 72-7.
(4) If Q > 0.5, and incident reports or field tests with kills are available, recommend the
pesticide for special review or other risk-benefit balancing.
Wild Mammal Toxidty (guidelines 71-3):
The wild mammal toxicity study, using a representative wild mammal, is typically a
laboratory study designed to determine at least one of several eco-effects endpoints: an acute
LD50/LC5o, a no observable effect level (NOEL), or a maximum acceptable toxicant
concentration (MATC). Typically, TGAI (see footnote 1) is administered either by oral
intubation or via mixing in the diet, and the results, expressed either as mg a.i./kg or ppm
a.L, are used to: provide additional support for precautionary labeling; establish eco-effects
endpoints; compare with measured or estimated environmental concentrations (i.e., mg
a.i./sq. ft. or ppm a.i.); and indicate if further wild mammal testing is required.
Avian Reproduction (guidelines 71-4):
The avian reproduction studies, using both an upland game bird (bobwhite quail) and a
waterfowl species (mallard duck), are laboratory tests designed to determine the quantity of
toxicant required to adversely affect the reproductive capabilities of a test population of
birds. TGAI is administered by mixture into breeding birds diets throughout their breeding
cycle, and the results (expressed as NOEL and various observable effect levels, such as
Lowest Observed Effect Level (LOEL) in ppm a.i.) are used to assess effects on
reproduction. Impaired reproduction is measured in terms of difference between treatment
and controls for:
# eggs laid per hen in eight weeks;
eggs cracked of eggs laid (%);
# viable embryos of eggs set;
live three-week embryos of viable embryos (%);
normal hatchlings of live three-week embryos (%);
# 14-day old survivors of normal hatchlings; and
# 14-day old survivors per hen.
11
-------
Results of this test are used to make the following determinations:
Value of Information (questions answered):
(1) Determine chronic/reproductive quotients (Q = quotient) (EEC (residue levels in
avian food items, from available field studies or Kenaga, 1973)/NOEL).
(2) Establish the level of minimal chronic risk to birds. (If Q(EEC/NOEL) < 1, then the
pesticide has a minimum chronic risk to birds, and no additional data are required.)
(3) Establish the level of chronic risk to endangered birds. (If Q(EEC/NOEL) > 1, then
consultation with U.S. Fish and Wildlife Service is required.)
(4) Determine the need for further regulatory action. (If Q(EEC/NOEL) > 1, then
further regulatory action may be required, which may include mitigation, field test(s),
screening study 71-5.)
(5) Establish the level of high chronic risk to birds. (If Q(EEC/NOEL) > 1, then the
pesticide has a potential high chronic risk to birds.)
Guideline Uses (Tier Progression):
(1) If Q(EEC/NOEL) < 1, there are no additional data requirements.
(2) If Q(EEC/NOEL) > 1, further regulatory action may be required, which may include
field test(s), screening study 71-5.
Regulatory Use(s):
(1) If Q(EEC/NOEL) < 1, there is minimal chronic risk to birds and no additional data
requirements; recommend registration/reregistration proceed.
(2) If Q(EEC/NOEL) > 1, there is potential high chronic risk to birds and further
regulatory action may be required, which may include mitigation, field test(s),
screening study 71-5.
(3) If Q(EEC/NOEL) > 1, and bird incident reports or field tests with bird kills are
available, recommend pesticide for special review or other risk-benefit balancing.
12
-------
Acute LC50 Estuarine and Marine Organisms (guidelines 72-3):
The estuarine/marine organisms LC5o/EC50 studies, using marine/estuarine fish, shrimp,
and mollusc species, are acute, forty-eight or ninety-six hour laboratory tests designed to
determine the concentration in water required to cause fifty percent mortality, incomplete
shell growth (oyster larvae), or reduced shell growth (oyster). TGAI (see footnote 1) is
administered into water containing test organisms, providing exposure for the appropriate
time period, and the results (expressed as ppm a.i.) are used to make the following
determinations:
Endpoint:
Mortality: LC50 value in ppm with 95 % confidence limits; lethal concentration in water
likely to kill 50% of the organisms exposed. Applies to acute toxicity tests for estuarine and
marine fish and shrimp.
Inhibition of Shell Growth: EC50 in ppm with 95% confidence limits. The concentration in
water inhibiting shell deposition by 50%. Applies to oyster shell deposition study.
Reduction in Larval Development: EC50 value in ppm with 95 % confidence limits. The
concentration in water likely to result in a 50% reduction in successful development of free-
swimming, fully-shelled veliger larvae from fertilized eggs. Applies to mollusc embryo
larvae study.
Value of Information (questions answered):
(1) Determine the category of toxicity to fish and aquatic invertebrates (ppm):
very highly toxic
highly toxic
moderately toxic
slightly toxic
practically non-toxic
0.1-1
>1-10
>10-100
>100
(2) Determine the need for precautionary label statements. (If LC50< 1 ppm, then "This
pesticide is toxic to fish (and aquatic invertebrates)" is required.)
(3) Determine the acute quotients (Q = quotient) for all formulations (EEC (residue
levels in water from back-of-the-envelope calculations or fate models)/LC50).
(4) Establish the level of minimal acute risk to fish and aquatic invertebrates. (If Q <
0.1, then the pesticide has a minimum acute risk and no additional data are required.)
13
-------
(5) Determine the need for Restricted Use classification. (If Q > 0.1, but < 0.5, then
acute risk may be mitigated by Restricted Use classification.
(6) Establish the level of acute risk to endangered estuarine fish and aquatic invertebrates.
(If Q > 0.05, then consultation with U.S. Fish and Wildlife Service is required.)
(7) Determine the need for chronic testing, fish early life-stage and invertebrate life-cycle
tests 72-4. (If Q > 0.01, then fish early life-stage and/or invertebrate life-cycle tests,
72-4 are triggered.) The determination to require either both the estuarine fish early
life-stage study and the estuarine invertebrate life-cycle study or only one of the two
studies will be made on a case-by-case basis to allow for flexibility. If the results
from both freshwater chronic tests and other data raise concerns for chronic effects,
then the Agency will ask for both estuarine chronic studies.
(8) Determine the need for further regulatory action. (If Q > 0.5, then further
regulatory action may be required, which may include mitigation, field test(s),
mesocosm, or actual aquatic field study 72-7.)
(9) Establish the level of high acute risk to estuarine fish and aquatic invertebrates. (If Q
> 0.5, then the pesticide has a potential high acute risk to fish and aquatic
invertebrates.)
Guideline Uses (Tier Progression):
(1) If Q < 0.1, then there are no additional data requirements.
(2) If Q > 0.5, then further regulatory action may be required, which may include field
test(s), mesocosm, or actual aquatic field study 72-7.
Regulatory Use(s):
(1) If Q < 0.1, then there is minimal risk and no additional data requirements;
recommend registration/reregistration proceed.
(2) If Q > 0.1, but < 0.5, potential risk may be mitigated by Restricted Use
classification.
(3)
(4)
If Q > 0.5, there is potential high acute risk and further regulatory action may be
required, which may include mitigation, field test(s), mesocosm, or actual aquatic
field study 72-7.
If Q > 0.5, and incident reports or field tests with kills are available, recommend the
pesticide for special review or other risk/benefit balancing.
14
-------
Fish Early Life Stage and Aquatic Invertebrate Life-Cycle (guidelines 72-4):
The freshwater and estuarine/marine fish early life-stage and invertebrate life-cycle
studies, using representative species, are laboratory tests designed to determine the quantity
of toxicant required to adversely affect the reproductive capabilities of a test population of
fish or invertebrates. TGAI is administered into water containing test organisms, providing
exposure throughout a critical life-stage (fish) or a life-cycle (invertebrates), and the results
(expressed as a MATC and a NOEL, in ppm a.i.) are used to make the following
determinations:
Endpoint(s):
Impaired Reproduction and Development: Effects on reproduction and development are
described as LOEL in ppm and NOEL in ppm in the aquatic environment of fish and aquatic
invertebrates likely to be exposed. Identification of the lowest effect level is dependent upon
study results and may be represented by either the NOEL or MATC. Impaired reproduction
and development are measured in terms of difference between treatment and controls for:
Fish
• number of embryos hatched
• time to hatch
• mortality of embryos, larvae, and juveniles
• time to swim-up (if appropriate)
• growth-weight and length
Aquatic Invertebrates
• survival of first generation
• production of young by first generation
• length of first generation at end of test
NOEL and LOEL are determined for each endpoint.
Value of Information (questions answered):
(1) Determine the chronic/reproductive quotients for all formulations (EEC (residue levels
in aquatic environment, from back-of-the-envelope calculations or fate
models)/NOEL).2
2 It is the goal of the Agency to regulate on the MATC where warranted. However,
depending upon study results, the NOEL may be deemed appropriate for use in the risk
assessment.
15
-------
(2) Establish the level of minimal chronic risk to fish and aquatic invertebrates. (If Q
(EEC/NOEL) < 1 (see footnote 2), then the pesticide has a minimum chronic risk to
fish and aquatic invertebrates, and no additional data are required.)
(3) Establish the level of chronic risk to endangered fish and aquatic invertebrates. (If
Q(EEC/NOEL) > 1, then consultation with U.S. Fish and Wildlife Service is
required.)
(4) Determine the need for additional chronic testing for fish, fish full life-cycle test 72-5.
(If Q (EEC/NOEL) > 0.1 (see footnote 2), and there is information that other parts
of the reproduction cycle (not tested in the fish early life-stage test) may be affected
by the pesticide, then fish full life-cycle test 72-5 is triggered.)
(5) Determine the need for further regulatory action. (If Q > 1, then further regulatory
action may be requked, which may include mitigation, field test(s), mesocosm, or
actual aquatic field study 72-7.)
(6) Establish level of high chronic risk to fish and aquatic invertebrates. (If Q > 1, then
the pesticide has a potential high chronic risk to fish and aquatic invertebrates.)
Note: Item (2) above is a regulatory decision element. Item (4) above is a tier progression
element. These two items are not contradictory.
Guideline Uses (Tier Progression):
(1) If Q (EEC/NOEL) < 0.1, then fish full life-cycle test 72-5 is not requked.
(2) If Q (EEC/NOEL) > 0.1, and there is information that other parts of the
reproduction cycle (not tested in the fish early life-stage test) may be affected by the
pesticide, then fish full life-cycle test 72-5 is requked.
Regulatory Use(s):
(1) If Q < 1, then there is minimal risk and no additional data requkements; recommend
registration/reregistration proceed.
(2) If Q > 1, there is potential high chronic risk and further regulatory action may be
requked, which may include mitigation, field test(s), mesocosm, or actual aquatic
field study 72-7.
(3) If Q > 1, and incident reports or field tests with kills are available, recommend the
pesticide for special review or other risk-benefit balancing.
16
-------
Fish Life-Cycle (guidelines 72-5):
The fish (freshwater and estuarine/marine) life-cycle study, using representative species,
is a laboratory test designed to determine the quantity of toxicant required to adversely affect
the reproductive capabilities and various life stages of a test population of fish. TGAI is
administered into water containing test organisms, providing exposure throughout a full life-
cycle (e.g., eggs from Fl adults through F2 eggs, embryos, larvae, and juveniles), and the
results (expressed as a MATC and no observable effect concentration (NOEC) in ppm a.i.)
are used to make the following determinations:
Endpoint:
Impaired Reproduction and Development: Effects on reproduction and development are
described as Lowest Observed Effect Level (LOEL) in ppm and No Observed Effect Level
(NOEL) in ppm in the aquatic environment of fish and aquatic invertebrates likely to be
exposed. Identification of the lowest effect level is dependent upon study results and may be
represented by either the NOEL or MATC. Impaired reproduction and development are
measured in terms of difference between treatment and controls for:
(1) Effects on 1st and 2nd generation:
• number of days to complete hatching
• number of embryos hatched
• number of surviving larvae hatched
(2) Effects on 1st and 2nd generation juveniles:
• number of abnormal fish
• length of survivors
• weight of survivors
(3) Effects on Ist-generation juveniles and adults, and 2nd-generation juveniles:
• mean length
•• mean weight
• number of survivors
• number of embryos
(4) NOEL and LOEL are determined for each endpoint.
Value of Information (questions answered):
(1) Determine chronic/reproductive quotients for all formulations (EEC (residue levels i
aquatic environment, from fate models)/NOEL) (see footnote 2).
in
17
-------
(2) Establish the level of minimal chrome risk to fish. (If Q (EEC/NOEL) < 1 (see
footnote 2), then the pesticide has a minimum chronic risk to fish, and no additional
data are required.)
(3) Establish the level of chronic risk to endangered fish. (If Q (EEC/NOEL) > 1 (see
footnote 2), then consultation with U.S. Fish and Wildlife Service is required.)
(4) Determine the need for further regulatory action. (If Q > 1, then further regulatory
action may be required, which may include mitigation, field test(s), mesocosm, or
actual aquatic field study 72-7.)
(5) Establish the level of high chronic risk to fish. (If Q > 1, then the pesticide has a
potential high chronic risk to fish.)
Guideline Uses (Tier Progression):
(1) If Q < 1, then there are no additional data requirements.
(2) If Q > 1, field test(s), mesocosm, or actual aquatic field study 72-7 is triggered.
( „„ , „ , , ,, , ,
Regulatory Use(s):
(1) If Q < 1, then there is minimal risk and no additional data requirements; recommend
registration/reregistration proceed.
(2) If Q > 1, then there is potential high chronic risk and further regulatory action may
be required, which may include mitigation, field test(s), mesocosm, or actual aquatic
field study 72-7.
(3) If Q > 1, and incident reports or field tests with kills are available, recommend
pesticide for special review or other risk-benefit balancing.
Simulated and Actual Field Testing - Mammals and Birds (guidelines 71-5(a) and (b)):
Simulated and actual field testing with mammals and birds consists of field tests designed:
(1) to either refute the assumption (based on laboratory tests) that risks to wildlife will occur
under conditions of actual pesticide use (Level One Study, 71-5a); or (2) to provide some
quantification of what risks may occur during actual use (Level Two Study, 71-5b). Under
most conditions TEP is applied: at maximum use rate(s), during actual use situations, under
situations representing "worst case" risk, and in areas containing adequate numbers of
wildlife. Results (expressed as the quantity of toxicant required to produce an adverse effect:
18
-------
e.g., ppm a.i., mg a.i./kg, pounds (Ibs) a.i./acre (A)) are used to: usually address field
mortality of birds and mammals, but may also address sublethal toxic effects, altered
behavior, reduced food resources, or lowered reproductive capabilities as well as effects to
amphibians and reptiles; compare with measured or estimated environmental concentrations
(i.e., ppm a.L, mg a.i./sf); possibly determine susceptibility differences between different
species; provide support for additional precautionary labeling; indicate if further field or
laboratory testing is required; and provide field data relative to the registrability of the
pesticide use.
Specifically, the following determinations are made:
Level One Study, 71-5(a)
Endpoint:
Percent Mortality occurring less than a specified percent of the time (20% in most cases):
Effects in the field are measured in terms of number of dead birds per test plot,
acetylcholinesterase inhibition, abnormal behavior, density differences, diversity differences,
residues in bird tissues, residues in bird food items.
Value of Information (questions answered):
(1) Determine if potential acute or chronic risk indicated by quotients (acute Q > 0.5;
chronic Q > 1) are confirmed.
(2) Determine the need for precautionary label statements. (If field studies demonstrate
that the use of the pesticide may result in fatality to birds, then the statement "This
pesticide is extremely toxic to wildlife" is required.)
(3) Establish the level at which potential acute or chronic risk to birds is negated by
results of field testing. (If no mortality is indicated in the screening study at not >
20%, no additional data are required.)
(4) Determine if further regulatory action is required. (If mortality is indicated in the
screening study at > 20%, then further regulatory action may be required, which
may include mitigation, field test(s), quantitative study 71-5(b>.)
Guideline Uses (Tier Progression):
(1) If mortality or impaired reproduction at minimal levels, then there are no additional
data requirements.
19
-------
(2) If mortality is not shown to be < 20 %, further regulatory action may be required,
which may include field test(s), quantitative study 71-5(b).
Regulatory Use(s):
(1) If mortality is not < 20 %, then the statement "This pesticide is extremely toxic to
•wildlife" is required.
(2) If mortality < 20 %, potential acute/chronic risk to birds is negated and there are no
additional data requirements; recommend registration/reregistration proceed.
(3) If mortality is not shown to be < 20 %, then further regulatory action may be
required, which may include mitigation, field test(s), quantitative study 71-5(b), or
recommend pesticide for special review or other risk-benefit balancing.
Level Two Study, 71-5(b)
Endpoint:
Percent Mortality, percent Impaired Reproduction (as % of vulnerable local populations):
Effects in the field may be measured in terms of mortality or survival rates, rates of
reproductive impairment, survival of dependent young, acetylcholinesterase inhibition,
abnormal behavior, age structure of the populations, residues in bird tissues, residues in bird
food items.
Value of Information (questions answered):
(1)
(2)
(3)
(4)
Determine if acute or chronic risk indicated by results from Tier 3 Testing (as
birds that use the field) are confirmed and quantify impacts.
%of
Determine the need for precautionary label statements. (If field studies demonstrate
that the use of the pesticide may result in fatality to birds, then the statement "This
pesticide is extremely toxic to wildlife" is required.)
Establish the level at which acute or chronic risk to birds is mitigated or limited in
magnitude and/or duration. (If mortality or impaired reproduction on vulnerable local
populations of birds are indicated in the quantitative study at minimal levels,
indicating that no additional data are required.)
Establish the level at which acute or chronic risk to birds is confirmed and quantified
as to the magnitude and/or duration. (If mortality or impaired reproduction on
20
-------
vulnerable local populations of birds is indicated in the quantitative study to be at
levels potentially detrimental, further regulatory action may be required, which may
include placing the pesticide in special review, or imposing other risk reduction
measures.)
Guideline Uses (Tier Progression): If mortality or unpaired reproduction is at minimal
levels, then there are no additional data requirements.
Regulatory Use(s):
(1) If mortality or impaired reproduction is significant, then the statement "This pesticide
is extremely toxic to wildlife" is required.
(2) If mortality or impaired reproduction is minimal, acute/chronic risk to birds is limited
in magnitude and/or duration. There are no additional data requirements.
(3) If mortality or impaired reproduction is biologically significant (e.g., ^ 20%),
further regulatory action may be required, which may include mitigation,
recommending the pesticide for special review or other risk-benefit balancing.
Simulated or Actual Field Testing - Aquatic Organisms (guidelines 72-7):
Simulated and actual field testing with aquatic organisms consists of field tests designed:
(1) to either refute the assumption (based on laboratory tests) that risks to aquatic organisms
will occur under conditions of actual pesticide use; and (2) to provide some quantification of
what risks may be during actual use. Typically, the study design used is a mesocosm one, in
which TEP is applied in a manner that simulates aquatic EECs likely to occur under actual
pesticide use conditions. Results (expressed as the quantity of toxicant required to produce
an adverse effect: e.g., ppm a.i.) are used to: establish what communities (e.g., benthic,
zooplankton, phytoplankton, finfish) are adversely affected and the magnitude of such effects;
compare with measured or estimated environmental concentrations (i.e., ppm a.i.); provide
support for additional precautionary labeling; indicate if further field or laboratory testing is
required; and provide field data relative to the registrability of the pesticide use.
Specifically, the following determinations are made:
Endpoint:
15-20% adverse effect (depending on measured parameter) hi local populations of aquatic
biota (e.g., plankton, macrophytes, macro-invertebrates, and finfish). Effects in the field are
21
-------
measured in terms of adverse effects on survival, biomass, growth, reproduction, and
population size in populations of aquatic organisms.
Value of Information (questions answered):
(1) Determine if potential acute or chronic risk indicated by quotients (acute Q > 0.5;
chronic Q > 1) are confirmed.
(2) Determine the need for precautionary label statements. (If field studies demonstrate
that the use of the pesticide may result in fatality to aquatic organisms, then the
statement "This pesticide is extremely toxic to fish (and/or aquatic invertebrates)" is
required.)
(3) Establish the levels at which acute or chronic risk to aquatic organism populations is
negated, mitigated, or limited in magnitude or duration. (If adverse effects on
exposed populations of aquatic organisms are indicated at < 15% (or 20%,
depending on parameter measured), then no additional data are required.)
(4) Establish the level at which acute or chronic risk to aquatic organism populations is
confirmed and quantified as to the magnitude and/or duration. (If adverse effects on
exposed populations of aquatic organisms are indicated at > 15% (or 20%,
depending on parameter measured), further regulatory action may be required, which
may include mitigation, placing the pesticide in special review, or requiring other use
restrictions.)
Guideline Uses (Tier Progression): If adverse effect is indicated at < 15% (or 20%,
depending on parameter measured), then there are no additional data requirements.
Regulatory Use(s):
(1) If adverse effect is indicated at > 15% (or 20%, depending on parameter measured),
then the statement, "This pesticide is extremely toxic to fish (and/or aquatic
invertebrates)" is required.
(2) If adverse effect < 15% (or 20%), acute/chronic risk to aquatic organism populations
is mitigated or limited in magnitude and/or duration, and no additional data are
required; recommend registration/reregistration proceed.
(3) If adverse effect > 15% (or 20%), further regulatory action may be required, which
may include mitigation, recommending pesticide for special review, or other risk-
benefit balancing.
22
-------
Special Tests (guidelines 70-1):
Special tests (typically using birds, mammals, or fish) are required on a case-by-case
basis to address particular toxicological hazards of concern. Such tests are usually performed
in the laboratory and may relate to metabolites, routes of exposure, or certain chemical
properties. Examples of these tests include (but are not limited to):
1. fish, mammalian, or avian cholinesterase studies designed to address levels of
cholinesterase inhibition;
2. avian, mammalian, or fish metabolism studies designed to address metabolites formed
and their toxicity;
3. secondary toxicity studies with birds and mammals designed to address potential
secondary, or tertiary, toxicity of pesticides (e.g., rodenticides and predacides); or
4. aquatic microcosm studies designed to address potential effects in aquatic communities
may be accepted as supplemental data when appropriate.
Results from these studies are used to: provide support for additional precautionary labeling;
establish toxicity or effect levels; compare, when possible, with measured or estimated
environmental concentrations (i.e., ppm a.i., mg a.i./sf, Ibs a.i./A); and indicate if further
laboratory or higher tier testing is required.
Nontarget Plants
Tier I Nontarget Area Terrestrial Plant Phytotoxicity (guideline 122-1):
The terrestrial nontarget plant phytotoxicity tests are greenhouse or growth chamber tests
that consist of three parts, a test for seed germination, a test for seedling emergence, and a
test for vegetative vigor. In all three tests the test organisms are: the required three species
corn, soybeans, and a root crop; plus 7 other species, usually tomato, cucumber, lettuce,
cabbage, oat, ryegrass, and onion (6 species of at least 4 families of dicots and four species
of at least 2 families of monocots). The soil or plant surface is treated with technical grade
active ingredient (TGAI) at a concentration comparable to the maximum label application rate
or at a concentration 3 times the estimated environmental concentration. Results are reported
in grams or pounds a.i. per acre and are expressed as the percent of detrimental effect on
germination or growth compared to the control after at least 5 days for the seed germination
test and 14 days for the seedling emergence and vegetative vigor tests. Parameters measured
include radicle length, % germination, % emergence, plant height, plant dry weight, and %
phytotoxicity. The results are used to: establish acute toxicity levels, compare with measured
23
-------
or estimated environmental concentrations, and to indicate if further testing at a higher tier is
necessary.
Tier I Nontarget Area Aquatic Plant Phytotoxicity (guideline 122-2):
The aquatic nontarget plant phytotoxicity tests are laboratory tests that evaluate the acute
toxicity of pesticides to the two aquatic species: Selenastrum capricornutum, (a freshwater
green alga) and Lemna gibba (an aquatic macrophyte). A one dose concentration comparable
to the maximum label application rate (or a concentration 3 times the estimated
environmental concentration) is used in these static tests. The Selenastrum test is 5 days long
and the Lemna test is 14 days long. Technical grade active ingredient is used, and the results
are reported as the percent of detrimental effect on growth compared to the control.
Parameters measured include number of cells or fronds and observed phytotoxicity. The
results are used to: establish acute toxicity levels, compare with measured or estimated
environmental concentrations, and to indicate if further testing at a higher tier is necessary.
Tier II Nontarget Terrestrial Plant Phytotoxicity (guideline 123-1):
The terrestrial nontarget plant phytotoxicity tests are greenhouse, growth chamber, or
small field plot tests that consist of three parts: a seed germination test, a seedling emergence
test, and a vegetative vigor test. These studies evaluate the effects of multiple dosage levels
on plant growth, using less than the maximum label rate with dosages in a geometrical
progression of no more than two-fold, and with subtoxic (
-------
using less than the maximum label rate with dosages in a geometric progression of no more
than two-fold, and with subtoxic (< EC50) and non-toxic (no observable effect level, NOEL)
concentrations. Knowing that herbicides kill plants, many registrants start their herbicide
testing at the Tier n level. The technical grade active ingredient is used, and the 5 day
results are reported for Selenastnon, Anabaena, Skeletonema, and the unspecified diatom.
Fourteen-day results are reported for Lemna. Results are reported as EC50 and NOEL values
in mg a.i/liter. Parameters measured include number of cells or fronds, and observed
phytotoxicity. The results are used to establish acute tpxicity levels, compare with measured
or estimated environmental concentrations, and to indicate if further testing at a higher tier is
necessary.
Tier III Nontarget Plant Phytotoxicity Field Studies (guideline 124-1 (Terrestrial Plants), 124-
2 (Aquatic Plants):
The Tier El terrestrial and aquatic nontarget plant phytotoxicity field studies are only
required if greater than 25% adverse effects on plant growth for terrestrial plants and 50%
adverse effects on pliant growth for aquatic plants are expected to occur when the product is
used as directed by the label (the estimated environmental concentration exceeds the £€25
terrestrial or EC50 aquatic value(s)). The Tier El tests are expected to provide information
on detrimental effects to plants during critical stages of development. The typical end-use
product (TEP) is used to assess effects on a broader range of nontarget plant species in a
number of geographic areas. Due to the absence of test protocols, the registrants have not
been required to perform Tier m tests to date.
Target Area Phytotoxicity Testing (guideline 121-1):
Data concerning the phytotoxic effects of a pesticide on desirable area plants are
generally waived by 40 CFR Part 158. Under certain circumstances, the Agency may
request these studies.
Nontarget Insects
Honey Bee Acute Contact LD50 (guidelines 141-1) :
The honey bee acute contact LD50, using the honey bee, (Apis melliferd), is an acute,
single-dose laboratory study designed to determine the quantity of toxicant required to cause
fifty percent mortality in a test population of bees. Technical grade active ingredient (TGAI)
is administered by one of two methods: whole body exposure to technical pesticide in a
nontoxic dust diluent; or, topical exposure to technical pesticide via microapplicator. Results
are expressed as LD50 in 'micrograms (jjig) of a.i. per bee, and are used to make the following
determinations:
25
-------
(1) Determine Category of Toxicity to Bees (/*g/bee):
< 2 highly toxic
2-10.99 moderately toxic
> 11 practically nontoxic
(2) Determine need for precautionary labeling
(3) Determine need for additional testing: if LD50 < 11
from foliar residue testing.
per bee, data are required
Honey Bee - Toxicity of Residues on Foliage (guidelines 141-2):
The honey bee foliar residue study is a laboratory test designed to determine the length of
time over which field-weathered foliar residues remain toxic to honey bees. The
representative end-use product is applied to crop foliage, the foliage is harvested at
predetermined intervals after application, and test bees are caged on the treated foliage.
Results are expressed in terms of the length of time, following application, during which
residues-continue to cause significant mortality in honey bee test populations. Results are
used to: provide support for precautionary labeling; establish extent of residual hazard of the
pesticide to honey bees; and, indicate if additional testing is required.
Industry Comments on "Description of Discipline" Section
Avian Acute
(1) The use of estimated LC50 values calculated from LD50 values concerns NACA.
Additional discussion of the validity and use of this approach is warranted prior to its
use as a standard procedure.
EPA Response
The use of estimated LC503 values calculated from LD504 values is discussed in detail in
Urban and Cook (1986), pp. 34-38. This Standard Evaluation Procedure has undergone peer
review by the EEFRA Scientific Advisory Panel. This method is used primarily to convert
rat LD50 values to estimated rat LC50 values. If actual rat LC50 values exist, these values
would be used (i.e., McCann, et al, 1981). A discussion of the variation in results using
this method is also found in this reference. The Agency typically uses this method in
3 Median lethal concentration in the diet necessary to effect (kill) 50% of the test population.
4 Median lethal dose necessary to effect (kill) 50% of the test population.
26
-------
preliminary risk assessments. Refined risk assessments consider additional exposure and
toxicity data.
(2) For the Restricted Use classification, Urban and Cook (1986) list the quotients as 0.2
to 1.0. In the Rejection Rate Analysis (RRA), the Agency has reduced the upper
quotient limit from 1.0 to 0.5. Without specific justification and consideration of
appropriate public comment, NACA believes the quotient should remain as outlined in
Urban and Cook (1986).
EPA Response:
In Urban and Cook (1986), Table 1, it states that presumption of risk that may be
mitigated by Restricted Use for birds occurs at 1/5 or 0.2 the LC50 ^ EEC < LC50. Thus,
the upper quotient5 limit is equal to the LC50 or 1.0 for Restricted Use to birds. Beyond the
quotient of 1.0 was presumption of unacceptable risk to birds. The reference for these limits
dates to Agency regulations published in 19756. Current regulations reaffirm the lower limit
of 1/5 or 0.2 for birds7, but do not specify a numerical upper limit8. Recent publicly
released documents, the decisions from the Ecological, Fate, and Effects Task Force9 and
the Implementation Paper for the New Paradigm10 establishes a numerical upper limit at 0.5
for birds. This 0.5 level of concern is consistent with that for aquatic organisms which was
stated in the 1975 regulations. Support for the 0.5 limit for birds comes from the results of a
preliminary retrospective analysis of 20 terrestrial field studies. The results of the analysis
support the conclusion that bird kills in the field may occur at quotients at or near 0.5.
6 Quotient is the ratio of chemical exposure value(es) to toxicity value(es), e.g., 7 to 25 parts per
million (ppm) estimated or measured in the diet of birds divided by LCBO values of 10 to 25 ppm
yielding quotient values of 0.3 to 1.
6 Part 162 - Regulations for the Enforcement of FIFRA [FR 40 (129): 28260-28265; 28281-28284;
Thursday, July 3, 1975].
7 1992, 40 CFR Protection of the Environment. Part 152.170. Criteria for restriction to use by
certified applicators, (c)(i){A) and (B), criteria for hazard to non-target species.
81992, 40 CFR Protection of the Environment. Part 154.7 Criteria for Initiation of Special Review,
(a) (3).
9 U.S. Environmental Protection Agency. 1992. Memorandum from Linda J. Fisher, Assistant
Administrator, to Douglas Campt, Director, Office of Pesticide Programs, dated October 29, 1992.
Subject: Decisions on the Ecological, Fate and Effects Task Force.
10 U. S. Environmental Protection Agency. 1993. Memorandum from Douglas. D. Campt, Office
Director, OPP to Victor J. Kimm, Acting Assistant Administrator, OPPTS, dated August 25, 1993.
Subject: Implementation Paper for the New Paradigm.
27
-------
As previously noted, the Levels of Concern (LOCs)11 for ecological effects have been
established based on Agency regulations and guidance12. In 1975, a Rebuttable Presumption
Against Registration (RPAR, the predecessor to Special Review), could be initiated when the
estimated environmental exposure was equal to or greater than one-half the acute toxicity
value for aquatic organisms. The "trigger" for birds at that time was equal to the acute avian
toxicity value with no "safety factor." Internally, OPP always believed that some safety
factor was needed, but the majority of pesticides for which avian field studies were required
had estimated exposures much greater than the acute toxicity values for birds. Thus, OPP
never established a number that captured the notion of "approaching" the bird toxicity value.
In the recent proposed revision to Part 158, OPP stated that the avian field study would
be required when the estimated environmental exposure is equal to or greater than one-half
the acute avian toxicity value. This is identical to the trigger for the aquatic mesocosm.
The establishment of regulatory LOCs depends on both science and policy, and they are
designed to serve policy based decisions. EPA believes that the current LOCs have a strong
regulatory and historical context and are supported by available data.
Industry Comment:
NACA continues to contend that there is minimal scientific justification for reducing the
upper bound limit form 1.0 to 0.5. NACA requests that the Agency obtain external peer
review of such a retrospective analysis before it is used as justification to change the
guidelines.
(3) The concept of LD50/ft2 is included in the triggering process and regulatory uses. This
trigger was not in the original guidelines and has not been thoroughly evaluated. In
addition, the Agency states that this "trigger" can be used to conclude the "pesticide
has a potential high acute risk to birds." NACA does not agree with this statement
and requests that it be removed.
[Note: EEB believes that the following argument by NACA is outside the scope of this
BRA, and that this section written by NACA enmeshes scientific issues leading to
11 The Level of Concern or LOG is a measure of risk to nontarget organisms or to groundwater
which may lead to regulatory action.
12 Examples include the regulations for Registration,.Reregistration and Classification (40 FR 28242);
the current regulations {40 CFR 154.7); the criteria for initiation of Special Review (40 CFR 154);
Criteria for Restriction for Use by Certified Applicators [40 CFR 152.170(c)]; 1988 Guidance Document
for Conducting Terrestrial Field Studies.
28
-------
regulation (risk assessment) and the use of data in regulatory decision-mating (risk
management). During the April meeting, EEB stated that many of the issues discussed
in this section are unrelated to the scope of the RRA. The RRA is focused on the
science needed to make sound risk assessments, not on how the data are used to
develop risk assessments. EEB believes that the issues addressed below should be
addressed in a separate venue.]
NACA believes that the LD50/sq-ft calculation, and the associated LOCs (e.g., significant
hazard for values >0.5), have a number of weaknesses and in their current form should not
be used for risk assessment. These weaknesses include;
1. The historical origins of the calculation are weak and not relevant to current chemistry.
The original speculation that significant adverse effects on birds could be associated with
applications equal to or greater than 1 LD50/sq-ft are based on small-scale pen studies
(DeWitt 1963) with persistent, bioaccumulative compounds (DDT, dieldrin, and
heptachlor).
2. Correlation of LD50/sq-ft calculations with the results of new field studies is incomplete
and not publicly available.
The agency has, at various times, made reference to a preliminary analysis of recent
avian field studies which reportedly supports the use of the current LD50/sq-ft model and
associated LOCs. Until such an analysis is complete and available for public review it is
inappropriate to reference this work.
3. The calculation has inherent flaws that lead to clearly untenable conclusions.
Exploratory calculations for broadcast applications indicate that the LD50/sq-ft model
predicts that a chemical applied at 1 Ib a.i./acre, having an LD50 of 400 mg/kg
(moderately toxic rating), would exceed the LOG of highest concern (0.5) for a 50 gram
songbird. Using the same example for a 20 gram songbird and a chemical with an LD50
of 1000 mg/kg (slightly toxic rating) produces a result also exceeding an LOG of 0.5.
Using these same calculation procedures, it can be shown that a compound that is
practically non-toxic (LD50 > 2000 mg/kg) would exceed the LOCs for restricted use
(0.2) and for endangered species concern (0.1). NACA feels these examples demonstrate
that the LD50/sq-ft model is flawed and will not discriminate between safe and potentially
toxic chemicals.
4. The calculation assumes no upper limit on granule consumption.
Consumption of large numbers of granules is limited, among other things, by the
repellent properties of most formulations, the relativity low attractiveness of most
29
-------
commercial granules as grit and simply the dilution-effect of naturally occurring grit.
Recent field studies (Fischer and Best 1994) with untreated granules indicate that
consumption of greater than 20 granules would be unusual. With no upper limit, the
current model results in products with relatively low toxicity triggering regulatory
concern.
5. The LD5o/sq-ft model has not been adequately peer-reviewed.
The model received cursory review by the SAP as part of the carbofuran Special Review.
NACA strongly feels that this review can not be construed as a blanket endorsement for
all granular compounds. Moreover, given the importance of this model as the primary
assessment tool for this important class of formulations, it is essential that the model be
the focus of an avian workshop or special scientific forum.
EPA Response:
The LD50/ft2 risk index concept is not a new one. It was first presented by Felthousen
(1977) and has been used by OPP/EEB ever since. More recently it was presented in the
Carbofuran Technical Support Document13 for the carbofuran Position Document 2/3, which
was issued for public comment through the Federal Register (54 FR 3744), January 25,
1989. As part of this process, the Agency presented the risk index to the FIFRA Scientific
Advisory Panel for peer review14. The Panel supported the risk index as an acceptable
initial approach to characterize avian risk and as a useful indication of potential risks of
granular insecticides to birds.
In addition, in March 1992, the Agency released the "Comparative Analysis of Acute
Avian Risk from Granular Pesticides"15. This document provided additional discussion,
explanation, and analysis using LD50/ft2.
(4) Requirement for avian field studies conflicts with New Paradigm. Also need
definition of "unusual circumstances."
13 U.S. Environmental Protection Agency. 1989 (January). CARBOFURAN: Special Review Technical
Support Document. Office of Pesticides and Toxic Substances.
14 FIFRA Scientific Advisory Panel. 1989. A set of scientific issues being considered by the Agency
in connection with the Special Review of carbofuran. Report dated 2/23/89 in response to public review
2/15/89, Arlington, VA.
16 U.S. Environmental Protection Agency. 1992 (March). Comparative Analysis of Acute Avian Risk
from Granular Pesticides. Office of Pesticide Programs, Washington, D.C.
30
-------
EPA Response:
Under the New Paradigm, requirements for avian field studies will not be based solely
upon the criteria in the footnotes to Part 158, but will be the result of risk management
decisions. The field study data requirement in Part 158 has not changed; rather, the criteria
describing when these studies will be required has changed. Thus, the when-required criteria
in the footnotes to this Part will be modified to reflect the policy changes.
The overriding emphasis of the policy is that risk managers should only ask for a study
when the information from such a study will improve the Agency's ability to make decisions.
Thus, if OPP has significant uncertainty about the risk of a chemical in the field which
cannot be resolved without a field study, then a field study should be considered. However,
before requiring such a study (or any other study) OPP should know how to use the results
of the study.
The Task Force conclusion that the avian field study provides very limited new
information to regulatory decision-making and thus would only be required under unusual
circumstances was based to a large degree upon two factors. First, a consensus of Task
Force members indicated that the results of the avian field studies reviewed by OPP always
confirmed predictions of adverse effects, primarily bird mortality, based upon lower tiered
studies. Second, a preliminary analysis of "in-house" avian field studies supported this
conclusion. The Task Force recognized that the conclusion was based upon limited data on
organophosphate and carbamate pesticides. Thus, the "unusual circumstances" where avian
field studies may be required in the future could include new pesticides which exceed the
LOCs for birds and which have chemistry and modes of action significantly different from
organophosphate and carbamate pesticides. However, other cases may arise which also may
warrant the requirement of a field study.
Avian Chronic
(1) Requirement for screening field study conflicts with New Paradigm.
EPA Response: See previous response concerning when field studies would be required.
(2) In the Rejection Rate Analysis (RRA), EPA proposes that a chronic risk quotient for
evaluating and categorizing reproductive effects be calculated using the Kenaga EEC
and the experimentally derived NOEL or LOEL. This conflicts with the New
Paradigm, which states that the Kenaga data should only be used in the absence of
"actual residue information."
Furthermore, the SEP for Ecological Risk Assessment clearly acknowledges the
general inadequacy of the Kenaga data for chronic exposure estimation.
31
-------
NAG A believes that these recent documents (New Paradigm and RRA), in their
simplistic approach to chronic exposure data, may promote a review process that will
produce an unwarranted number of chronic toxicity triggers for pesticides.
EPA Response:
Recently, there has been considerable discussion concerning the conduct of the avian
reproduction test, the interpretation of its results, and the use of the results in determining
LOC's. At the request of OPP, the Office of Research and Development Environmental
Research Laboratory at Corvallis, Oregon presented an overview of this test for evaluating
effects of pesticides on reproduction hi birds16. A number of changes in the conduct and
design of the test were discussed. One important point of discussion in the report was that
pesticide effects on avian reproduction are not simply a function of chronic exposure.
Reproductive effects have been shown within days (e.g., 8 days or less) after treatment.
Thus, short-term exposures should be considered in evaluating the potential for avian
reproductive effects. In 1994, the Agency will be participating with OECD and SETAC in a
review of this test and its use in ecological risk assessment. Until completion of that review,
and based upon the ORD overview, the Agency will use short-term exposure values for
determining avian chronic LOCs.
Residue data in Agency files usually lack residue values occurring immediately or shortly
after application. As specified in the Decision on the Ecological, Fate and Effects Task
Force, when actual residue data are lacking, the estimated residue values from Kenaga will
be used in calculating avian chronic LOC's.
Simulated and Actual Field Testing
(1) The RRA uses 20% adverse effect as a benchmark for determining regulatory actions
and requiring additional data. This appears to conflict with New Paradigm, which
indicates that widespread and repeated avian mortality will not be tolerated in the face
of low benefits (mortality of 20% could still be considered widespread).
EPA Response:
Quantification of effects as not less than 20% on exposed individuals is used as an
important measurement endpoint for field studies. It is used to indicate that further
regulatory action may be required. Whether this endpoint alone would satisfy the
"widespread and repeated" risk/benefit standard for decision-making would be determined
18 Bennett, R. S. and L. M. Ganio. 1991 (July). Overview of methods for evaluating effects of
pesticides on reproduction in birds. EPA 600/3-917048. U.S. EPA, Environmental Research Laboratory,
Corvallis, Oregon, pp.106.
32
-------
after it was determined that further regulatory action is required. Changes were made in the
document reflecting this response. As discussed earlier, under the New Paradigm policy,
there is a reduced emphasis on using field studies to make a regulatory decision.
Aquatic Chronic
(1) The sections on value of information, guideline uses, and regulatory uses for the fish
early life stage, fish full life cycle, and aquatic invertebrate (72-4) tests appear to
contain a number of discrepancies. The risk criteria should compare the EEC directly
to the NOEL or LOEL values, but instead they compare the EEC to a portion of the
NOEL (0.10 or 0.5). The correct interpretation of the triggers is summarized by the
Agency in the discussion of the flow charts presented in this document.
EPA Response:
We agree. The aquatic chronic LOG is the ratio of the EEC to the LEL, the lowest
effect level. Identification of the LEL is dependent upon study results and may be
represented by either the NOEL or MATC. It is the goal of the Agency to regulate on the
MATC where warranted. However, depending upon study results, the NOEL may be
deemed appropriate for use in the risk assessment. Additionally, the use of the data for
either tier study progression decisions or regulatory decisions were clarified in the document.
Appropriate changes were made in the document to reflect this.
(2) In the RRA, EPA states:
"A bioaccumulation study (guideline 165-4) is required if a significant concentration
of the pesticide and/or its principal degradation product(s) are expected to occur in
aquatic environments and may accumulate in aquatic organisms. If the potential for
aquatic food chain effects is established, an aquatic organism accumulation study
(guideline 72-6) is also required."
The requirement for also testing bioaccumulation according to guideline 72-6 is new.
The new requirement is unnecessary and should be deleted.
EPA Response:
This requirement (guideline 72-6) is not new. It has been part of the data requirements
since the establishment of Part 158 in 1984. This explanation was included to indicate the
tiered requirements for potential aquatic food chain effects. The fish bioaccumulation test
(guideline 165-4) in Subdivision N, is required first. Then if the results from this test
indicate a potential for adverse effects on the aquatic food chain, the bioaccumulation test
using other aquatic organisms (guideline 72-6) in Subdivision E is required.
33
-------
REFERENCES
Felthousen, R. 1977. Memorandum to J. Akerman, dated September 9, 1977.
Subject: Classification of granulated formulations.
Fite, E.C., L.W. Turner, N.J. Cook and C. Stunkard. 1988. Guidance Document for
Conducting Terrestrial Field Studies. EPA 540/09-88-109. U.S. Environmental Protection
Agency, Office of Pesticide Programs, Washington, D. C. 20460.
McCann, J.A., W. Teeters, D. J. Urban, and N. Cook. 1981.
A Short-Term Dietary Toxicity Test on Small Mammals. Avian and Mammalian Wildlife
Toxicology: Second Conference. ASTM STP 757. D.W. Lamb and E.E. Kenaga, Eds.,
American Society for Testing and Materials, pp. 132-142.
Urban, DJ. and N. Cook. 1986.
Hazard Evaluation Division Standard Evaluation Procedure - Ecological Risk Assessment.
EPA-540/9-86-167. U.S. Environmental Protection Agency, Office of Pesticide
Programs, Washington, D. C. 20460.
Rejection Rate Analysis - General Issues Raised by Industry
In the introductory material of the "NACA Response to EPA's Ecological Effects
Rejection Rate Analysis," Industry provided a number of general comments. These
comments addressed protocol and methodology problems, procedural and policy concerns,
and a variety of other generic issues. Since these comments were general in nature, and did
not relate to any specific rejection factor or specific study, the Agency's responses will be
provided here in general terms.
Industry submitted several comments relating to the need for resolution of protocol and
methodology problems. Concerns were raised for the following tests: fish full life-cycle,
avian reproduction, nontarget plant studies, oyster studies, and Daphnia chronic studies. The
issues raised are outside the scope of the RRA, and will be addressed by other venues (e.g.,
the guideline harmonization process and revision of SEPs).
One comment requested closure on analytical issues associated with aquatic tests. As
noted in Industry's comment, these issues are discussed in the 1992 EPA/NACA guidance
document, Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance. This
guidance document is considered closed and is an attachment to this Rejection Rate Analysis.
(See Appendix A).
34
-------
One comment questioned the need for NOELs in aquatic and avian acute studies. The
Agency agrees that automatic rejection of acute studies which lack an NOEL is not justified.
Currently, the NOEL is not required in acute studies.
A number of issues were raised relating to policy and procedure, the need for
clarification of guidance, and the need for more flexibility on the part of the Agency. Two
suggestions discussed in the April meeting are:
• Streamline the process for industry feedback to reviewers.
For rapid feedback to questions from both sides, all contact should go through the
Product Managers. NACA will help establish a procedure, similar to the SOP set up for
rapid feedback from the EPA, for EPA scientists to use when a technical question arises
in the review of a study.
• Integrate risk assessment into the study review process.
If data is not an important component of the risk assessment, then repeating the study is
not necessary. The problem is that the individual scientist will not have all of the
necessary information available to him/her in order to make that determination. HRR
suggested that a one page summary of all of the compound's uses, use rates, application
methods, timing of use, maximum uses, etc., can be added as a cover letter to go with
each study. This would help the scientist decide how important a study is to the risk
assessment. It is important that EEB coordinate with BEAD and HED to evaluate
comparable use rates, provided this is possible within program time constraints.
EPA Response:
The hazard assessment (ecotoxicology data) and the fate assessment (exposure data) are
separate components of the risk process and must be adequately characterized independently
prior to integration into the risk assessment. It is possible that a data requirement can be
waived or a rejected study not be repeated. Coordination with BEAD and HED is only
possible during special reviews and maybe RED's-not registration actions (sections 3, 5, 18,
and 24(c)).
Note: EEB believes that the following issues raised by NACA are outside the scope of
the KRA:
• provide clearer guidelines for the use of EEC values;
• consolidate all guidance into a single document;
• integrate the new paradigm into KRA;
• standardize criteria which define an acceptable study; and
• implement more flexible policies in areas of accepting test conditions outside
recommended range, accepting chronic studies which lack NOEL's, reviewing non-
standard studies, and reviewing studies conducted with very low toxicant concentrations.
35
-------
Finally, although it was not presented in the introductory material, the need for a
definition of "data" was expressed several times in the Industry's response document. The
Agency intends this to mean the actual raw data sheets, or a summary of the raw data, with
sufficient information to allow for independent statistical analysis of the data.
EPA Response/Comments:
NACA wants to include alternate test species not currently approved by the Agency.
This issue is being addressed through internal and external (e.g., OECD) evaluation via the
guideline harmonization process.
Under the Rejection Factors section of this document, the term "tabulated and unanalyzed
data" is substituted for "raw data" to clarify what the Agency needs in order to evaluate the
studies.
Current Rejection Rate
The following graphs demonstrate the current and historical rejection rates for each of the
ecological effects guidelines. The historical rate includes all List A studies reviewed as of
December 1, 1992, but does not include studies that were reviewed prior to the publication
of the Registration Standards. Due to the limited number of cases, the results reported in
this section have not been tested for statistical significance, and therefore caution should be
exercised in their interpretation. The purpose here is not to develop an empirically
defensible rejection rate value. Rather, the intent is to use rejection rates as the best
indicator available of where additional Agency/registrant attention and efforts are warranted
to improve the quality of the studies.
Figure 1 illustrates the overall rejection rate for ecological effects, which is now
estimated at approximately 20%. This is down from the overall rate of 32% prior to 1986,
and up slightly from the 1986-1988 rate of 17%. While there has been a significant
reduction in rejection rates (38%) between the pre-1986 and the post-1988 time periods, the
current (post-1988) rate still remains quite high.
Figures 2 through 4 illustrate what the Agency believes are the current rejection rates for
each guideline. The avian acute toxicity studies rejection rates are low, but the avian chronic
reproduction studies are very high (Figure 2). While five of the six fish acute toxicity
studies have current rejection rates greater than 10%, they are significantly lower than the
fish chronic reproduction studies (Figure 3). For both the avian and aquatic reproduction
studies, the high rejection rates indicate that significant problems exist.
36
-------
Figure 5 illustrates those guidelines whose rejection rates appear to have dropped
significantly over time. All four of the guideline studies that have shown such improvement
are acute toxicity studies.
Figure 6 illustrates those guidelines whose rejection rates have not demonstrated
consistent and significant improvement over time.
Studies not included in this analysis due to insufficient data include 71-3 wild mammal
toxicity, 71-5 avian field studies, 72-5 fish life cycle, 72-6 aquatic organism accumulation,
and 72-7 mesocosm and pond field studies.
Summary
Key implications that might be drawn from these graphs include:
(1) overall rejection rates in ecological effects appear to have gone down significantly.,
but current rates remain high;
(2) for both avian and aquatic reproduction studies, the high rejection rates indicate
significant problems;
(3) five of the six fish acute toxicity studies have current (post-1988) rejection rates
greater than 10%;
(4) four acute toxicity guidelines—71-1, acute avian oral; 72-1C, acute toxicity trout;
72-2, acute toxicity Daphnia; and 72-3B, acute toxicity mollusk—have shown
encouraging and consistent reduction in their rejection rates over time; and
(5) none of the implications discussed above can be tested for statistical significance,
and caution should therefore be exercised in interpreting them.
37
-------
o1 2 *
5 S *
Q. 0> -,
® *£•
3? 8
Rejection Rate (percent)
10
o
CO
o
i.
5*
M
%
III
I* s
1. F 3
S** *** 4£
S* %^ s-
CO
00
o>
fit
Cft
-
I-
a. .A
09 -^-
IS
Q) _*
"• co
en 03
-< CO
CO 03
C CO
O1
I
Q.
CO
00
00
I
I
(l)
K)
m >
:
»
-------
I*
Rejection Rate (percent)
10
O
CO
O
sf r
• 9; J.A
SL w
M
« 5'
I
(K
O
rb
•••
I
-------
Rejection Rate (percent)
-------
* 3
8 3
5, *
II -
Q>
s> *
s
1 1
to
*
N>
0)
K)
to
ro
to
Q
C
Q.'
2.
CD
10
Rejection Rate (percent)
to
o
03
O
T
T
o
K>
t
M V
(0
O
c
cr 5
a
O ^
a
CD (O
i 00
oS
? 30
CD £
"* CD"
o
o"
TI
(5*
-------
Rejection Rate (percent)
-o
3
CO
00
O)
CD
00
O)
I
_»
CO
00
CO
3"
V)
CO
00
00
-------
Rejection Rate (percent)
I
CD
00
O)
CO
03
0>
CO
00
00
H
CO
CO
00
00
O o m
*
=* a ff
3
-------
Rejection Factors
The following ecological effects guidelines were analyzed to determine the most common
reasons that studies submitted to meet these guidelines were rejected. After each rejection
factor, specific references to EPA guidance addressing that factor are given. The EPA
guidance is analyzed and is referred to in this report to determine if the guidance documents
available to registrants adequately cover the areas where problems have occurred. A list of
all guidance documents available for ecological effects studies is provided in Appendix B.
After each rejection factor and the corresponding references to EPA guidance, an
Industry Comment section has been provided with the industry scientists' assessment of the
adequacy of the EPA guidance; an explanation of technical difficulties, if any, associated
with the rejection factor; and recommendations. Following each industry comment is EPA's
response to that comment.
After each guideline, the rejection factors are assessed in terms of the registrants' ability
to avoid the factor in the future. This is presented in a list that represents what EPA would
consider to be rejection factors that could or should be avoidable by the registrants. It is the
intention of the Agency in the future to take the appropriate regulatory action should such
avoidable factors cause a future study submission to be rejected. This standard will not be
applied retrospectively to the studies analyzed in this assessment.
AVIAN ORAL LDSO (Guideline 71-1)
1. Rejection Factor: Study failed to establish a valid LD50 value with corresponding 95%
confidence limits, or an LD50 greater than 2000 mg/kg.
Agency Guidance: Pesticide Assessment Guidelines - Subdivision E (EPA-540/9-82-024;
October, 1982):'7 p. 33, no. 71- 1 (b)(4) and p. 36, 71-1 (d),
Dosage-mortality data section. FIFRA Accelerated Reregistration -
Phase 3 Technical Guidance (EPA-540/09-90-078; December 24,
1989),18 p. C-14, 71-1, no. 9. and C-16, 71-1, no. 9., provide
guidance on this subject.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes.
17 For the rest of this paper this document will be referred to as Subd-E.
18 For the rest of this paper this document will be referred to as FIFRA-TG.
44
-------
Industry Comment:
As all of the 21 studies on list A chemicals since 1988 have been
accepted, there is clearly not a rejection rate problem with this
guideline.
EPA Response: Industry and the Agency agree.
2. Rejection Factor: A "split-dosing" procedure, which is inconsistent with other comparable
acute toxicity testing methods, was used.19
Agency Guidance: Subd-E: p. 33, 71-1 Heading and no. (b) (4) and p. 34, 71-1, no. (d)
Acceptable protocol.
Assessment of
Guidance:
Avoidable:
Industry Comment:
Guidance on this subject is adequate.
Yes. Possibly not in some cases with Mallard Duck at the doses
required. The guidance is not adequate if following the guidance with
a recommended species at a recommended rate leads to a study
rejection. Additional guidance should recommend the use of Bobwhites
where regurgitation is a problem with Mallards.
The quoted guideline does not actually specify that split dosing is not
allowed, and split dosing should not be rejected out of hand if the doses
are administered closely together. Regurgitation is a toxic response of
Mallards to some chemicals and not necessarily related to the size of
the dose, so split dosing does not always resolve the issue. Mallards
also regurgitate if the volume exceeds 5 % of the body weight. ,
Bobwhite Quail are less prone to regurgitation and should be the
species of choice.
EPA Response:
The Agency agrees with Industry's comment that regurgitation may be
an inevitable consequence of toxicity of some test compounds.
Fortunately, this reaction has been relatively limited, and therefore
should be addressed on a case-by-case basis. The result of the avian
single-dose oral LD50 is an integral component of the Agency risk
assessment. The Agency believes that, when the problem of
regurgitation is encountered, a reasonable effort to resolve the problem
19 Another, possibly more important, rejection factor is that regurgitation occurred in all test
levels. It appears that this may be why the study authors chose a "split-dosing" technique: to
counteract the regurgitation.
45
-------
should be made, so that satisfactory data are available from the study.
As suggested in Industry comments, bobwhites are less prone to
regurgitation and would be one way of avoiding the problem. However,
the Agency believes that guidance on number of doses for the "Avian
Single-dose Oral LD50" is adequate.
3. Rejection Factor: The length of time birds were fasted prior to dosing was not specified.
Agency Guidance: Subd-E: p. 35, 71-1, no: (d) Diet and fasting and FIFRA-TG: p. C-
13, 71-1, no. 4 and C-15, 71-1, no. 4.
Assessment of
Guidance: The guidance documents which address this factor appear adequate.
Avoidable:
Yes.
Industry Comment: The length pf tune that the birds were fasted should have been
reported.
EPA Response: Industry and the Agency agree.
AVIAN DIETARY LC50 (Guideline 71-2)
1. Rejection Factor: The reliability of the data was questionable because mortalities occurred
that were attributed to interactions between animals (rather than to the
test chemical), but no such interactions were observed in the controls.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
The study author can draw his/her own conclusions concerning any
mortalities that occur during a test, particularly when there is an
inadequate dose-response pattern. P. 38 of Subd E 71-2(b)(4) indicates
that satisfactory data should establish the 8-day dietary LC50.
Not clear, in that it is subject to interpretation as to whether data are
adequate to calculate an LC50. Statistical means are available that will
generate a "number for an LC50," but the confidence in that number
may be low if the dose response is poor.
No.
Industry Comment: Subd-E does not define "satisfactory data." The SEP says only that
"Mortality that cannot be fully attributed to the effects of the test
46
-------
EPA Response:
Resolution:
material may be better understood if assessed in the light of behavioral
observations." Other quail species, such as the Japanese Quail, are less
aggressive than bobwhites, but they are not currently acceptable. The
study should be allowed to be conducted with Japanese Quail, as is
acceptable under European guidelines.
The SEP says only that "Mortality that cannot be fully attributed to the
effects of the test material may be better understood if assessed in the
light of behavioral observations." The SEP does mention the use of ten
bobwhite quail to a pen (p 2. 2. Pen Facilities) and this species is
known sometimes to express aggression under these circumstances.
Other quail species, such as Japanese quail, are less aggressive than
bobwhites.
NACA recommends the use of Japanese quail in cases where aggressive
behavior cannot be avoided. EPA maintained that Japanese quail are
unacceptable, although the use of Japanese quail is undergoing
evaluation via internal and external (OECD) guideline harmonization.
D. Barolo indicated that if one cannot eliminate aggressive behavior for
a specific study, then results from Japanese quail studies will be
considered as supplemental information and used in the "weight of
evidence" evaluation of the need for additional data.
The Agency agrees with Industry's comment that chemical induced
interaction between animals in initial tests is not avoidable. However,
the Agency does not agree with the suggested solution to accept data on
Japanese quail. The acceptability of this species is being addressed in
the guideline harmonization process.
NACA and EPA agreed that aggressive behavior can be avoided by
caging the bobwhite quail individually, even though this is a protocol
deviation.
2. Rejection Factor: The reliability of the test results were questionable because of high
variability in the measured test concentrations in the test diet. The test
material was known to be unstable at room temperature, so the treated
test diet (bird food) was stored at 4 degrees C throughout the study. At
the test termination, samples of feed were taken from the stored batch
to verify concentrations through chemical analysis. These samples
were then stored at -20 degrees C for about one-half year before
chemical analysis. At that tune, the chemical analysis results were
highly variable.
47
-------
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
Standard Practice for Conducting Subacute Dietary Toxicity Tests with
Avian Species, Section 9.3 page 372.
Standard Evaluation Procedure; Avian Dietary LC50 Test, EPA-540/9-
85-008, page 5, Section m D.
Subdivision E Guidelines for Hazard Evaluation for Wildlife and
Aquatic Organisms, Section 70-3 (b)(l) page 26.
The guidance clearly requires that steps be taken to ensure that test
concentrations are the same throughout the study. However, there is
no specific guidance which specifies how soon after a sample is
collected that chemical analysis must be performed. This was
considered to be more a problem of scientific judgement than a
disregarding of any specific guidance.
The Good Laboratory Practice Standards Subpart F of 40 CFR Section
160.105 (which is currently published, but was not in effect when these
studies were performed) provide substantially more guidance on the
need to verify concentrations, and to ensure that stored test material
(and, implicitly, stored media that contains the test material) is tested to
determine stability of test material under storage conditions. Even at
that, currently there is nothing in the Standards that requires a specific
maximum storage time such that this problem could be avoided in the
future.
Unclear. On the one hand, it seems that a laboratory would not wait
half a year to do chemical analysis. On the other hand, if the
variability of the resulting measurements had been acceptably low, the
procedure would have been accepted because it would have been
confirmation that the test material was in the test diet at a given
concentration. It is not clear that this type of problem would be
avoidable, given the guidance provided.
EPA has focused on the length of sample storage as the potential
problem here, but Industry contends that there may be a number of
other issues—such as over sampling, homogeneity, and recovery—that
affect dietary analysis. It is well known that the accuracy of analytical
measurement is affected by the size of the nominal concentration.
Although concentrations within the range of hundreds of ppm can be
measured very accurately, at the lower end of a dose-response range, if
concentrations are in ppb or less, error will be proportionally greater
due to a great excess of binding sites.
48
-------
EPA Response:
Resolution:
Industry Comment:
Although EPA does not directly refer to the magnitude of the problem,
if it is within the 70 to 130% range, if the dilution and dosing
procedures are good and well documented, and if there is a consistent
dose-related response in the biological endpoints, the study should not
be rejected purely on the basis of variation in the analysis of the diet.
By setting a criterion of 70 to 130% for the measured doses compared
with nominal, the requirements for avian dietary tests are then
generally in line with those jointly established by EPA and NACA for
aquatic laboratory testing (EPA/NACA (1992), Conducting Acceptable
Aquatic Laboratory Studies: Proposed Guidance,, pp. 10-12).
The permissible variability of measured concentrations in the diet
should be viewed in context of how it affects the ability to characterize
toxicity. Generally, the greater the variability, the less certainty that
can be placed in statistically calculated endpoints such as LC50s, etc.
Variability, in and of itself, should only be used as a study rejection
factor if it is so great that it precludes distinction of test levels (i.e., the
variability is so great that one test level is not significantly different
than the next higher or next lower). The 70% and 130% may be a
good rule, of thumb for initial screening, but would not preclude under
some conditions still requesting additional testing to better characterize
risk, even if the variability was between 70% and 130%.
EPA believes that the percent of nominal can be established to be used
as a benchmark for acceptable analytical recoveries. A review of
appropriate studies is needed to establish acceptable analytical limits.
The limits of 70 to 130% established in the EPA/NACA guidance
document apply to aquatic studies and may not be applicable to avian
and mammalian studies.
Designation of 70-130% as a "rule of thumb" means that a registrant
will not know a priori if the study might be rejected on this parameter.
Unless 70-130% is designated as an acceptable range, there is not
agreement between EPA and Industry, and the issue remains
unresolved.
3. Rejection Factor: Test material was not technical grade. The test material was 42.2%
a.i., a formulation presumably, rather than the technical grade product.
Agency Guidance:
Subdivision E Guidelines for Hazard Evaluation for Wildlife and
Aquatic Organisms, Section 71-2 (b)(l) page 38.
49
-------
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Standard Evaluation Procedure; Avian Dietary LC50 Test, EPA-540/9-
85-008, page 1, Section I C 1.
The guidance for this requirement is clearly stated and well recognized
throughout the pesticide industry. Guidance states that avian subacute
dietary testing must be done with the technical grade active ingredient
unless the Agency specifically indicates a need for formulated product
testing. No new guidance is needed to clarify this requirement.
Yes, this deficiency is clearly avoidable. Note that it would be possible
for a test done with a formulation to provide useful supplemental
information, but it does not fulfill requirements for registration unless
specifically indicated by the Agency.
From the information given, one cannot be certain whether the test
material was a formulation or not. Some materials can, for good
reasons, be described as being of low purity, or it may be difficult to
define exactly what the active ingredient is.
The Agency agrees with Industry. In the rare event that the technical
grade is relatively low in purity (e.g., 42%), the registrant should
explain that fact in each study where the low purity technical grade test
material is used. Industry should be sure that the percent active
ingredient is included in all reports. This is because cover letters and
other supporting documentation frequently get separated from tests,
thus making the information found in them unavailable to the science
reviewer. A study should not be rejected just because the technical
grade is of relatively low purity.
4. Rejection Factor: The top test level was only 2110 ppm a.L, and no LC50 was
established. The only condition where it is acceptable not to establish
an LC50 is when testing is conducted up to 5000 ppm and it can be
comfortably stated that the LC50 is likely to be greater than 5000 ppm.
Agency Guidance:
Assessment of
Guidance:
Subdivision E Guidelines for Hazard Evaluation for Wildlife and
Aquatic Organisms, Section 71-2 (b)(4) page 38.Standard Evaluation
Procedure; Avian Dietary LC50 Test, EPA-540/9-85-008, page 3,
Section H C 1.
The guidance for this requirement is clearly stated. Note that in terms
of test material as a formulated product, the maximum test
50
-------
Avoidable:
Industry Comment:
concentrations were 5000, and the LC50 would be >5000 ppm for the
42.2% test material. No new guidance is needed to clarify this
requirement,,
Yes.
This is clearly a problem associated with the study already discussed in
Rejection Factor 3, and would inevitably follow from the original
misunderstanding.
EPA Response:
The Agency agrees with Industry. If the test material is a formulation
of relatively low percentage a.i., it cannot fulfill the requirement for
testing with the technical grade active ingredient. Tests using the
formulation may, if there are no other deficiencies, provide useful
supplemental information.
5. Rejection Factor: There was a 50% drop from nominal concentration by day 5.
Inadequate justification and explanation were provided. Note that the
primary problem was a lack of information explaining the graphs and
results of the study to determine if binding caused the apparent decline
in measured test concentrations. It appears that this study is capable of
being upgraded if the problem is adequately addressed.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
Standard Practice for Conducting Subacute Dietary Toxicity Tests with
Avian Species, section 9.3, page 372.
Subdivision E Guidelines for Hazard Evaluation for Wildlife and
Aquatic Organisms, Section 70-3 (b)(l), page 26.
The guidance primarily focuses on ensuring that test concentrations
remain constant and that reliable comparisons can be made concerning
relative toxicity and relative sensitivity. They do not specifically
indicate the exact wording and nature of explanations required for
variation in measured test concentrations compared to nominal test
levels.
Yes.
This rejection factor is avoidable, but should not necessarily undermine
the study. The testing facility should understand and adapt to the
51
-------
EPA Response:
stability of the compound under test conditions and keep variability in
the range of 70 - 130% of nominal.
The Agency agrees with Industry. Knowing the stability of the
compound in avian test diet under test conditions is essential. A study
rejected because of declining concentrations may be upgradeable if
adequate explanation is provided and the decline curve is well
characterized. Refer to the EPA Response and Resolution sections for
Rejection Factor 2 above for comments regarding limitations of the 70
130% variability range.
WILD MAMMAL TOXICITY (Guideline 71-3)
1. Rejection Factor: The method used to prepare the treatment diets, first forming a stock
feed and then mixing with clean feed, could lead to erroneous results
because animals might choose to eat untreated rather than treated feed.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
Subd-E: pp. 40-41, 71-2, (d)(5) Test diets20 and FIFRA-TG: p. C-
17, 71-2, no. 2, give guidance on diet preparation, and, initially, it
appears to indicate this method of diet preparation is acceptable to use.
However, diet mixing practices can still be open to question, and in this
case the scientist who reviewed this study indicated that he questioned
the homogeneity of the treatment diets and the mixing methods; e.g.,
how soon after the stock diet was prepared were the treatment diets
prepared.
Agency guidance on this may need to be reexamined, since issues
concerning diet preparation are recurring,,
Probably not with current guidance.
Guidance appears adequate, although the FIFRA-TG appears to be
more relaxed than are the reviewers about acceptance criteria. The
rejection factor quoted does not seem inconsistent with the guidelines.
Industry welcomes the opportunity to establish exactly what problems
FJRR has with diet preparation. It is possible that contact with the
registrant before issuing the DER could have resolved the issue.
20 This section is under 71-2, Avian Dietary LC50, because 71-3, Wild Mammal Toxicity,
refers to 71-2 for performing dietary tests.
52
-------
EPA Response:
Resolution:
While other issues concerning diet preparation, as previously suggested,
may need to be reexamined, the Agency agrees with Industry's
comments that guidance on mixing appears adequate. The question
raised in this factor does not appear to be a generic issue, but a
problem specific to this individual study. The issue relates to the
description of preparation of the diets, which did not appear to meet the
guidance given to ensure a uniform concentration in the feed.
Use of a premix (i.e., prepare stock feed and mix with clean feed) for
diet preparation is a commonly used, acceptable approach in toxicology
and is not cause for rejection of a study. Methods must include
adequate procedures to ensure that prepared feed concentration is
uniform.
2. Rejection Factor: Extensive cannibalism occurred during the study, which may indicate
that improper caging of animals was used.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
Subd-E: p. 40, 71-2, (d)(l) Pen facilities., nos. (1) and (4), and
FIFRA-TG: p. C-18, 71-2, no. 6. provide guidance on pen facilities.
The guidance given is for pen facilities for birds. However, issues
such as cannibalism are handled on a case-by-case basis. This is a
chemical-specific decision that is important for the overall assessment
of the hazards of the chemical.
No. It is not possible to prevent animals placed in cages from
exhibiting cannibalism, but the registrant could have placed the animals
in separate pens. Guidance is not specific in this case.
Although the results are likely to be obscured by the cannibalism, they
may still include sufficient information to conduct a risk assessment.
On the grounds of animal welfare, repetition of the study should not be
requested unless the results are completely unusable. The issue arises
through lack of specific guidelines, though it appears that the animals
may not have been cared for in the best possible way.
It should be routine that the protocol for this experiment be agreed
upon by EPA and the registrant before the study is conducted. Expert
advice should be sought where the behavior of the test species in
captivity is not known by either party. The study should not have to be
repeated unless the results of the current study add no value to the risk
assessment.
53
-------
EPA Response:
Resolution:
EPA Additional
Response:
This test is not normally required and a variety of species may be used.
The behavior of the species in captivity may not be well known by
either the registrant or the Agency. The SEP shows flexibility and
Industry supports this. The rejection factor could have been avoided by
a discussion between the Agency and the registrant, and by seeking
specialist advice by both parties before initiation of the test.
The Agency agrees with Industry that extensive consultations with
experts are needed before conducting these less-than-routine tests. The
Agency disagrees with Industry that rejection, and subsequent
repeating, only be imposed if the results are of no value. A study can
be repeated if additional testing will increase the understanding of the
risk to nontarget organisms. In some cases, the results may add some
value, but. still not provide sufficient information to allow for a
mammal risk assessment in which high confidence can be placed. It is,
however, Agency policy that the "value of information" should be taken
into account before any study has to be repeated.
EPA and NACA agree that these studies are seldom conducted and thus
standard methods do not exist. Consultation with experts is
recommended before conducting such studies, and inclusion of protocol
review and consultation prior to test initiation is recommended.
Problems such as cannibalism is certainly undesirable; however, they
may be unavoidable given the lack of experience in conducting such
tests.
Unavoidable problems such as cannibalism may not result in automatic
rejection of the study. However, if the extent of the problem is such
that it precludes deriving useful lexicological information from the
study that may be used to assess risk, then a repeat of the study may be
requested. For example, if only one animal in ten, and in only one test
level, succumbs to cannibalism, that cannibalism would probably not,
by itself, result in study rejection. If, however, substantial cannibalism
occurred at most of the test levels, the study would probably not
provide sufficient information to characterize toxicity. The Agency
would likely conclude that there would be high value of information for
requesting thai the study be conducted again.
It is recommended that when conducting tests which are less than
routine, sufficient preliminary effort be expended to discover many of
these problems and devise ways to avoid them. For example, groups
of test animals similar in size and composition to those proposed in the
study should be held for a period equivalent to the full duration of the
54
-------
study. If cannibalism occurs, the test groups may have to be modified
or reduced, or it is even possible that each individual would have to be
housed separately.
AVIAN REPRODUCTION (BOBWHITE QUAIL) (Guideline 71-4)
1. Rejection Factor: The value of the control birds as a valid biological control was
diminished because the percent of cracked eggs in the control group
was higher than two of the treatment groups, and the highest incidence
of mortality and excoriation was found in the control group.21
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Subd-E: pp. 48 - 57, 71-4 (all sections) and FIFRA-TG: pp. C-27 -
C-32, 71-4 (all sections) provide general guidance on conducting these
types of studies. Specific guidance concerning acceptable levels for
reproductive parameters is not available, since such levels are handled
on a case-by-case basis.
Due to the complexities involved in performing these types of studies,
these rejection issues are handled on a case-by-case basis.
No. Current testing guidance does not specify the maximum numbers
or percent of cracked eggs in the control that are acceptable.
Eggshell cracking may often be influenced by bird behavior, cage
conditions, or handling procedures. Consequentially, rejection of
studies based on eggshell cracking in controls is unlikely to improve
assessment of pesticide uses that truly will affect bird reproduction.
Before rejecting studies for eggshell cracking, the following questions
should be considered: (1) is cracking in treatment groups significantly
higher than in controls; (2) without reference to controls, does a dose-
related trend in cracking exist; (3) is cracking elevated in other species
tested; and (4) are other egg-related reproductive variables (fertility,
viability, hatchability, shell thickness) affected in a dose-related
manner?
The Agency does not agree with the Industry comments. The Agency
has been consistent in rejecting studies in which there is more than
10% cracked eggs in the controls. A background rate of 10% as the
limit of acceptability is reasonable as the Agency sees an average of
21 There were four groups: a control, 20 ppm treatment level, 100 ppm, and 500 ppm.
55
-------
Industry Comment:
Resolution:
5 % from most research laboratories. The EPA-ERL at Corvallis was
able to achieve background rates of < 1 % for mallard duck and <5%
for bobwhite quail. A high rate of cracking leads to a loss of
sensitivity of the study and a decrease in the statistical power of the
test. The ability of the study to detect differences between control and
treated birds for other avian reproduction parameters is diminished.
The avian reproduction study does not achieve high statistical power for
its parameters as is, and, therefore, further reduction in power is not
desirable.
1. EPA states that the "EPA-ERL Corvallis was able to achieve
background rates of < 1 % for mallard duck and < 5 % for bobwhite
quail." It should be noted that the Corvallis lab does not conduct
studies according to EPA or ASTM guidelines relative to light
intensity. They have significantly reduced light intensity in order to
reduce aggressive behavior and egg cracking. Given this, it is
inappropriate to use this lab as a comparative example of what is
attainable.
2. The Agency is incorrect in stating that a "high rate" of egg cracking
leads to loss of statistical sensitivity or power. Statistical power
depends upon the number of cages and not the number of eggs, per se.
This concern has been rebutted by registrants in the past using actual
statistical power analysis based on real data.
3. The Agency states that "the ability of the study to detect differences
between control and treated birds for other avian reproduction
parameters is diminished." NACA disagrees with this point.
EPA asserts that the following rejection criterion shall be applied:
(1) < 10% cracked eggs in the control birds is acceptable (i.e., no
rejection);
(2) ^ 10% and < 13% will result in a review of eggshell cracking in
relation to other study endpoints. However, this is not cause—by
itself—for rejection; and
(3) ^ 13% is grounds for rejection.
2. Rejection Factor: Data discrepancies dimmish the integrity of the study; e.g., (1) the
photoperiod used in the study appeared inappropriate, (2) postmortem
examinations of chicks could not be performed because chicks were so
56
-------
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
cadaverous, (3) it was not reported why medication was administered in
the feed to adults and chicks, or (4) the statistical evaluation did not
appear to include the total number of data points.
Subd-E: pp. 48 - 57, 71-4 (all sections) and HFRA-TG: pp. C-27 -
C-33, 71-4 (all sections) provide general guidance on conducting these
types of studies. Specific guidance concerning how to handle data
discrepancies or acceptable levels for reproductive parameters is not
available, since such levels are handled on a case-by-case basis.
Due to the complexities involved in performing these types of studies,
these rejection issues are handled on a case-by-case basis.
Possibly.
The rejection factors are a collection of "data discrepancies." It is not
stated whether each one of these issues is considered by itself to be a
"fatal flaw," or whether it takes multiple issues of this type to result in
study rejection (for other study types, EPA has generally listed such
factors separately). Because three of the four issues are stated with
qualifications ("appeared inappropriate"), it would be useful to know if
the reviewers were essentially requesting additional information (and
therefore deferring judgement on validity) or were classifying a study
as scientifically unsound for these reasons.
Photoperiod: EPA guidelines provide recommendations on photoperiod
that should be followed (not "must" be followed). Furthermore, the
SEP states that "EEB does not endorse any one protocol.... This
referenced protocol [Subd-E] provides flexible guidance to help
researchers design scientific protocols ..." (p. 9). Deviations from
the recommendation should be considered grounds for rejection only if
there is evidence that the photoperiod used did not stimulate the birds
to reproductive activity. Under the current protocol, EPA recommends
16 to 17 hours of light during the stimulation period. This represents a
significant daily photoperiod (equivalent to sunrise at 6:00 a.m. and
sunset at 11:00 p.m.). Birds are stimulated to reproduction on shorter
or longer daily photoperiod.
Postmortem examinations: EPA guidelines only require postmortem
necropsy for adults (all must be necropsied). Necropsies are not
routinely performed on offspring (0 to 14 days old) because their small
size makes examination of internal organs difficult. Chicks are housed
at high temperatures, which exacerbates the problem by promoting
57
-------
rapid autolysis. If postmortem examination requirements were applied
to progeny, the number of required analyses (> 2,000) would be
excessive.
Medication use: Any medication given during the test must be
documented and the reason for its use explained. However, EPA
guidelines do not require an explanation for why such treatments are
given. Future guideline revisions should clarify this point. In many
cases, issues such as this can be most efficiently handled by telephone
contacts with registrants, instead of going through the entire process of
study rejection and registrant rebuttal.
Statistical evaluation: Any statistical analysis should include all the
appropriate data points. Given the number of checks that such data
currently receive prior to submission, this should not be a general
problem. Before rejecting studies for such reasons, registrants could be
contacted by telephone during the review process to resolve such
matters.
Industry concurs that these are complex studies requiring case-by-case
review. Given the high rejection rate reported for reproduction studies,
it seems possible that studies are being rejected for data discrepancies
that are not necessarily "fatal flaws." As a general rule, the more
complex and lengthy a study, the greater the probability that some
deviations from the ideal will exist in execution or biological
performance. The need to repeat studies should be viewed in terms of
the overall exposure and risk to birds, with clear documentation that an
adequate safety margin can not be inferred from existing information.
Finally, strong feelings exist within and outside EPA that the current
avian reproduction test may be of limited usefulness for risk assessment
for most currently registered chemicals. This factor should also be
considered when the decision is made to request replacement studies.
Industry encourages EPA to proceed expeditiously to review the
existing test protocol.
EPA Response: The Agency agrees with Industry on most points:
photoperiod: [partially agree] The photoperiod should not be
referenced as reason for rejecting study unless it is considered likely,
scientifically, that the photoperiod adversely impacted the results,
making them unreliable for risk assessment.
post mortem examination: [agree] The study should not be rejected
just because the chicks were not necropsied.
58
-------
Resolution:
medication: [agree tentatively] The issue is not so much whether the
use of medication in the diet of test animals is acceptable; rather, it is
whether it is explained, and whether it impacts the validity of the study.
Medication of animals may be acceptable, but its use should be justified
and explained in detail. The Agency must determine whether its use
may have impacted the results. Further, the Agency must consider
whether the fact that medication was needed suggests other problems
with the study, such as unhealthy organisms or poor animal husbandry.
These deficiencies are usually evidenced in other ways during the
study, but may also be manifested by the need for medication. Use of
unhealthy organisms or poor husbandry is likely to affect the
acceptability of the study.
statistical analysis: [agree] Studies should never be rejected solely
because of the statistical analysis the registrant did or did not do. It is
the Agency's responsibility to conduct statistics and evaluate the results
of the study using the data. The inclusion of data, and possibly the
format of that data, may be an issue in acceptability of the "report" of
the study. If a study report is deemed unacceptable because of missing
data or data submitted in an inappropriate format, the study would be
considered for upgrading when the needed information was provided in
the proper format.
Industry and the Agency were generally in agreement that:
(1) photoperiod should not be a rejection criteria unless it impacts on
the scientific validity of the study,
(2) generally, the lack of a postmortem examination of chicks should
not be a rejection criterion unless study-specific factors warrant it,
(3) an explanation of how and why animals were medicated should be
included in the report, and
(4) all appropriate data should be provided in the report.
AVIAN REPRODUCTION (DUCK) (Guideline 71-4)
1. Rejection Factor: Data discrepancies diminish the integrity of the study; e.g., (1) the
photoperiod used in the study appeared inappropriate, (2) egg collection
procedures appeared inappropriate, (3) it was not reported why
medication was administered in the feed to birds, or (4) the overall
fertility of control birds appeared to be too low.
59
-------
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
Subd-E: pp. 48 - 57, 71-4 (all sections) and FIFRA-TG: pp. C-27 -C-
33, 71-4 (all sections) provide general guidance on conducting these
types of studies. Specific guidance concerning how to handle data
discrepancies or acceptable levels for reproductive parameters is not
available, since such levels are handled on a case-by-case basis.
The guidance in Subd E for lighting period is specific (see page 53, 3rd
and 4th para). The other issues are less clearly defined. Due to the
complexities involved in performing these types of studies, these
rejection issues are handled on a case-by-case basis. These types of
decisions are important for the overall assessment of the hazards of the
chemical. '
The lighting deficiencies are avoidable, but it is not clear that the other
deficiencies are specifically cautioned against in written guidance.
The rejection factors are a collection of "data discrepancies." It is not
stated whether each one of these issues is considered by itself to be a
"fatal flaw," or whether it takes multiple issues of this type to result in
study rejection (for other study types, the Agency has generally listed
such factors separately). Because three of the four issues are stated
with qualifications ("appeared inappropriate"), it would be useful to
know if the reviewers were essentially requesting additional information
(and therefore deferring judgement on validity) or were classifying a
study as scientifically unsound for these reasons.
Photoperiod: EPA guidelines provide recommendations on photoperiod
that should be followed (not "must" be followed). Furthermore, the
SEP states that "EEB does not endorse any one protocol .... This
referenced protocol [Subd-E] provides flexible guidance to help
researchers design scientific protocols . . ." (p. 9). Deviations from
the recommendation should be considered grounds for rejection only if
there is evidence that the photoperiod used did not stimulate the birds
to reproductive activity. Under the current protocol, EPA recommends
16 to 17 hours of light during the stimulation period. This represents a
significant daily photoperiod (equivalent to sunrise at 6:00 a.m. and
sunset at 11:00 p.m.). Birds are stimulated to reproduction on shorter
or longer daily photoperiod.
Egg collection: This is unlikely to be a significant general factor in
study rejection. Gross errors (e.g., no system existed to separate eggs
collected from different treatment groups) should have been described
in the EPA analysis.
60
-------
EPA Response:
Medication use: Any medication given during the test must be
documented and the reason for its use explained. However, EPA
guidelines do not require an explanation for why such treatments are
given. Future guideline revisions should clarify this point. In many .
cases, issues such as this can be most efficiently handled by telephone
contacts with registrants, instead of going through the entire process of
study rejection and registrant rebuttal.
Egg Fertility: Critical parameters such as egg fertility should not be
judged on so vague a basis as "appeared to be too low." Industry 4
postulates two points in responding to this: (a) "normal" values have
not been published by EPA nor do any consensus values exist in the
scientific community at large; and (b) if the chemical is not exerting
any deleterious effect, then any treatment group, including the controls,
has an equal chance of being the lowest performing group. Given the
cost and length of reproduction studies, EPA should allow use of
laboratory historical data, where appropriate, as an additional indicator
of reproductive performance. EPA should work with NACA and other
groups to determine if normal or acceptable values can be established.
NACA concurs that these are complex studies requiring case-by-case
review. Given the high rejection rate reported for reproduction studies,
it seems possible that studies are possibly being rejected for data
discrepancies that are not necessarily "fatal flaws." As a general rale,
the more complex and lengthy a study, the greater the probability that
some deviations from the ideal will exist in execution or biological
performance. The need to repeat studies should be viewed in terms of
the overall exposure and risk to birds, with clear documentation that an
adequate safety margin can not be inferred from existing information.
Finally, strong feelings exist within and outside EPA that the current
avian reproduction test may be of limited usefulness for risk assessment
for most currently registered chemicals. This factor should also be
considered when the decision is made to request replacement studies.
Industry encourages the Agency to proceed expeditiously to review the
existing test protocol.
See Rejection Factor 2 above for comments regarding photoperiod and
use of medication. It is unlikely that egg collection methods and/or
frequency would be sole rejection criteria, unless they appeared to be
rather haphazard and affected hatchability.
The Agency agrees with Industry's comment that critical parameters
such as egg fertility should not be judged on so vague a basis as
"appears to be too low," and this was not the case. This statement is a
61
-------
summary of information discussed in the review of this study which
addressed those points, such as historical data, which Industry indicated
should be addressed. The range of egg fertility found in this type of
study is stated and compared to the range reported for this study.
FRESHWATER FISH LCSO (BLUEGILL) (Guideline 72-1)
1. Rejection Factor: The concentration level selected was less than 100 mg/L and was not
high enough to produce an LC50.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Subd-E: p. 67, 72-1, no. (b)(3)(i)(A)(B) and (ii); pp. 67- 68, 72-1, no.
(c)(l)(i)(A)CB). FIFRA-TG: p. C-38, 72-1, no. 8. and p. C-40, 72-1,
no. 8. Stephan, Charles E. (EPA-660/3-75-009; April, 1975):22 pp.
6 - 7 and pp. 45 - 49, nos. 12. and 13. plus Table 6 and Section F.,
no. 8.
The guidance documents which address this factor appear adequate.
Yes, unless the solubility of the test material is below the level at
which any mortality occurs. In this case, the registrant is expected to
submit a request for an exception to this requirement because of
insolubility. The guidance for this type of request has been developed
jointly by EPA and NACA and is contained in Appendix A, Conducting
Acceptable Aquatic Laboratory Studies: Proposed Guidance (1992).
None of the guidance documents provides true guidance in the case of
solubility problems. As the purpose of the study is to determine the
acute toxicity expressed as an LC50 value, Industry basically agrees
with EPA. Solubility problems frequently occur below 100 mg/L with
pesticides, making it sometimes difficult to test up to this level.
However, application of the theoretical rate of 10 Ib/A directly into 0.5
ft. of water results in only 7.4 mg/L (Urban and Cook, 1986).
Depending on the concentration achieved, the data may be used for a
risk assessment. The expectation to request a waiver in these
circumstances was not known to many registrants.
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue. However, the Agency
22 This reference provides examples of acceptable methods for conducting acute aquatic
toxicity tests. For the rest of this paper this document will be referred to as Stephan.
62
-------
stresses that resolution of solubility problems in aquatic studies is
independent of risk assessment issues. Toxicity (hazard) must be
characterized independently of an exposure estimate.
2. Rejection Factor: The test chambers were aerated.
Agency Guidance: FIFRA-TG: p. C-38, 72-1, no. 7 and Stephan: p. 16 and p. 31, no.
2.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes. Either the test chambers should not be aerated, or if aeration is
necessary, chemical analysis of the test solution periodically throughout
the test should be performed to verify exposure.
Industry Comment: Industry agrees with EPA's assessment. However, a rejection is only
warranted if there is aeration but no analysis.
EPA Response:
EPA and NACA agree.
3. Rejection Factor: The biological loading of test vessels was twice the recommended
amount.
Agency Guidance: FIFRA-TG: p. C-37, 72-1, no. 7. and Stephan: pp. 33 - 34, no. 4.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes.
Industry Comment: Industry recognizes that the guidance documents clearly recommend
levels for biological loading. However, if the dissolved oxygen and
ammonia levels are within acceptable limits, if there is a distinct
concentration-effect relationship, and if the concentrations are stable
over time, the biological loading alone should not be reason for
rejection. A flexible, case-by-case evaluation of an otherwise
scientifically valid study should be considered.
EPA Response:
The Agency agrees that this factor alone may not be sole reason for
rejection of a study. However, a loading factor of twice the
recommended amount is a significant deviation and can warrant
63
-------
rejection of a study. Furthermore, in studies where loading exceeds
recommended levels ammonia levels should be measured daily. This
factor will be handled on a case-by-case basis.
4. Rejection Factor: Test substance purity was not identified.
Agency Guidance: Subd-E: p. 30, 70-4 (c)(2)(i). Stephan: p. 47, no. F.2.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
The guidance documents which address this factor appear adequate.
Yes.
Industry agrees that the guidance documents provide appropriate
guidance, and that knowledge of the purity of the test substance is
mandatory for a risk assessment. However, not all impurities may
have to be known, unless there is reason to suspect the influence of
impurities. In order to avoid unnecessary delays in time and effort for
EPA and the registrant, industry recommends that the reviewer contact
the registrant prior to finalizing and issuing the DER.
Both the Agency and Industry agree that this factor is avoidable. This
information needs to be included with the study upon its first
submission to the Agency.
5. Rejection Factor: Inappropriate test species were used and test species were not clearly
identified.
Agency Guidance: Subd-E: p. 30, 70-4, (c)(3)(i) - (viii); p. 67, 72-1 (b)(2)(i) -(iii).
FIFRA-TG: p. C-37, 72-1, no. 3. Stephan: pp. 20 - 22, no. D.I.
and Table 3.
Assessment of
Guidance:
Avoidable:
Industry Comment:
The guidance documents which address this factor appear adequate.
Yes.
Industry concedes that a clear identification of the test species is
essential. However, it is not quite clear which species are appropriate.
While the FEFRA-TG and the SEP renders acceptable rainbow trout,
brook trout, and Coho salmon as cold-water species, and bluegill,
channel catfish, and fathead minnow as warm-water fish, Stephan and
64
-------
EPA Response:
Resolution:
ASTM also recommend the goldfish as a warm-water species. Subd-E
requests one warm- and one cold-water fish, preferably bluegill and
trout. Several other warm-water species are recommended by OECD.
Although EPA has good reason for preferring certain species, use of
other internationally accepted species in a scientifically valid test should
not lead to automatic rejection, especially in view of international
harmonization.
The preferred test species are rainbow trout and bluegill sunfish, which
are listed in Subdivision E, FEFRA-TG and the SEP. EPA is in the
process of harmonizing its testing guidelines with other groups,
including OECD. A result of this process may be the acceptance of
test species that are currently not on the preferred species list.
However, accepting a study does not necessarily mean that an
additional species that is more appropriate for the risk assessment will
not be required. It is likely that EPA will always require trout and
bluegill for comparative risk assessments.
In the discussion, HHB indicated that it would be likely that EPA will
harmonize with other groups and accept species that are not on the
current preferred test species list. However, accepting a study does not
necessarily mean that an additional species will not be required that is
more appropriate for the risk assessment. It is likely that EPA will
always require trout and bluegill for comparative risk assessments.
6. Rejection Factor: Fish were fed during the 96-hour exposure period of the study.
Agency Guidance: SEP Acute Toxicity Test for Freshwater Fish: p. 4, H.B.3. Stephan:
p. 30, IV.D.7 p. 37, IV.E.7. ASTM E - 729: p. 388, 10.8.3 and p.
391, 11.6.
Assessment of
Guidance: The guidance documents which address this factor appear adequate.
Avoidable:
Yes.
Industry Comment: Industry agrees that the fish should not be fed during the 96-hour
exposure period.
EPA Response: EPA and NACA are in agreement.
65
-------
7. Rejection Factor: Minimum limit of detectability, or the minimum quantifiable limit, was
not defined quantitatively.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Subd-E: p. 29, no. 70-4 (b)(l)(2), p. 32, no. 70-4 (c)(6)(ii)(F) and p.
68 no. 72-1 (c)(4). SEP Acute Toxicity Test for Freshwater Fish: p.
9, m.H. Stephan: p. 39 - 40, IV. E.lO.b.l and p. 49, IV.F.9.
FIFRA-TG: 72-1, p. C-40, no. 8. ASTM E - 729: p. 394, 15.1.9.
Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
(1992). The guidance provided in this document is adequate to address
this issue.
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
Guidance to avoid this problem in the future is offered by EPA/NACA,
Conducting Acceptable Aquatic Laboratory Studies: Proposed
Guidance (1992).
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
8. Rejection Factor: The variability limit for test concentrations was greater than 1.5.
Agency Guidance: ASTM E-729: p. 392, 11.9.3.4(2).
Assessment of
Guidance: Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
(1992). The guidance provided in this document is adequate to address
this issue.
Avoidable:
Industry Comment:
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
Variability of measurements may be encountered in spite of all efforts
to limit the variation, especially if test concentrations are in the ppb or
ppt range, and if the compound readily adsorbs to surfaces, has very
limited solubility, or has limited stability. Coefficients,of variation at
66
-------
EPA Response:
these concentrations are high, as noted in the literature, and variation is
more likely to be related to analysis than to malfunction of a diluter (if
used). In these or other circumstances where the analytical method is
acceptable, as determined by spiked control samples, and serious efforts
to produce less-variable data have not been successful, EPA should be
flexible—a rejection does not seem warranted. On a case-by-case
basis, EPA should accept studies with variable data, as long as an
adequate defense has been made.
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
9. Rejection Factor: The dilution water contained higher levels than recommended of lead,
iron, and aluminum.
Agency Guidance: Stephan: p. 15, IV. C. 2. ASTM E-729: 384, 8.2.2.1 and p. 385,
8.4.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Generally, the guidance given in this area is adequate.
Yes.
While Stephan on page 15 recommends specific levels of lead, iron,
and aluminum for reconstituted water, which is basically recommended,
on page 14 acceptable water for acute toxicity tests is defined as
"healthy test organisms will survive in it for the duration of acclimation
and testing without showing signs of stress. . . . Because daphnids are
more sensitive to many toxicants ... a more realistic criterion for an
acceptable dilution water is that first instar daphnids will survive in it
for 48 hours without food." Similarly, ASTM gives as a minimal
requirement for acceptable dilution water "that at least one test species
will survive, grow, and reproduce satisfactorily in it." Taking criteria
of biological responses to the test water into consideration, studies do
not need to be rejected on this basis.
The Agency agrees with Industry that biological responses of organisms
to the dilution water need to be considered. However, if these and
other metals are present in the dilution water at concentrations which
exceed acceptable criteria, the Agency should be notified prior to test
initiation.
67
-------
Resolution:
The study will not be rejected solely on the basis that metals exceeded
the criteria. The biological performance of the organisms will also be
considered in evaluating the study.
10. Rejection Factor There was no solvent control.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Resolution:
Subd. E: p. 26, no. 70-3 (b)(l)(2)(3). ASTM E-729: p. 385 - 386,
9.2.5 - 9.2.7, p. 389, 11.1.1.1 and p. 393, 13.1.3. SEP Acute
Toxicity Test for Freshwater Fish: p. 6, H.D.3. Stephan: p. 31,
IV.E.1. FIFRA - TG: 72-1, p. C-36, no. 2 and p. C-39, no. 7.
The guidance documents which address this issue appear adequate.
Yes.
Only Stephan and ASTM require solvent controls. The Subd-E
guideline states that the solvent control "should be used." (ASTM
defines "should" to mean that the specified condition is recommended
and ought to be met, if possible. Although violation of one "should" is
rarely a serious matter, violation of several will often render the results
questionable.)
Industry agrees that the use of a solvent control is desirable and
scientifically appropriate. However, if the solvent control would have
displayed signs of intoxication, but was omitted, these signs of
intoxication would be attributed to the test compound. Although this
might not reflect the true toxicity, for risk assessment purposes it would
reflect a "worse case" scenario, and safety for the assessment. It
therefore may not be necessary to reject a study for this parameter
alone.
The Agency guidance on this factor is adequate. Industry agrees with
the Agency that using a solvent control is scientifically valid and
correct.
NACA suggested that the Agency indicate in guidance documents that a
solvent control "must" (instead of "should") be included. The Agency
agrees.
11. Rejection Factor The results for two of the test concentrations were obtained from a
separate test conducted a few weeks after the definitive study.
68
-------
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Resolution:
Subd. E: p. 27, 70-3, (b)(4).
p. 380, 4.1.
Stephan: p. 31, IV.E.l. ASTM E-729:
The guidance documents which address this issue appear adequate.
Yes.
Industry basically agrees with EPA's assessment. Adding
concentrations at a later time may occur, however, if in cases of very
flat concentration-effects relations, space in the diluter system or in the
facility is insufficient. This may arise especially if the determination of
a NOEL is required (studies of NACA members were rejected for this
reason). Add-on studies must establish a continuation of the biological
responses (e.g., by overlapping of tested concentrations). Study
rejection therefore should be based on a case-by-case consideration.,
Adding concentrations from a separate study and combining them with
the ones used in the definitive study is inappropriate. The arguments
provided by Industry are unpersuasive. If the definitive study does not
generate an acceptable dose-response curve, a new study needs to be
conducted. Note, however, that determination of a NOEL in acute
toxicity studies is not required. It is the responsibility of the registrant
to ensure that the design of the test system results in the study being
conducted according to guideline requirements.
NACA assumed that the results of the study were combined to produce
a mortality/effect profile which would allow an UCSO/ECSO or a NOEL
to be estimated. NACA agrees that combining results of tests
conducted at different times is not desirable, especially to complete
dose-response curves.
12. Rejection Factor Not all test solutions were measured at 96 hours as well as at zero
hours. The concentration of the test material greatly decreased for at
least one test level after 96 hours of exposure.
Agency Guidance:
Subd, E: p. 29, 70-4 (b)(l)(2), p. 32, 70-4 (c)(6)(ii)(F). Stephan: p.
41 - 42, IV.E.10.B.(3). SEP Acute Toxicity Test for Freshwater Fish:
p. 7, H.D.6 p. 9, m.H. and p. 11, IV.C.l. ASTM E - 729: p. 392,
11.9.3.3 and 11.9.3.4(1) and (3).
69
-------
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
(1992). The guidance provided in this document is adequate to address
this issue.
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
Industry agrees that rejections in this area were not avoidable due to
lack of clarity. Studies were accepted in the past by EPA with and
sometimes without analytical measurements.
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
13. Rejection Factor A control group of fish for the inert/carrier ingredients present in the
formulation was lacking. (The test material was 42% a.i. and 58%
inert/carrier ingredients.)
Agency Guidance: Stephan: p. 31, IV.E.I.
Assessment of
Guidance:
Avoidable:
Industry Comment:
Normally, this guidance is adequate. However, in this case the
scientist concluded that, without an inert/carrier control (i.e., blank
formulation) it was not possible to discern whether the toxic effects
were due to the a.i., the inerts/carrier, or to both. This is a chemical-
specific decision that is important for the overall assessment of the
hazards of the chemical.
No.
While Stephan requires an additive control, it also states that "none of
the ingredients of the mixture or formulation is considered an additive."
Therefore, one cannot conclude that testing of blank formulations is
needed. Since an a.i. will be applied to the environment only as a
formulation., knowledge of the contribution of the formulation additives
may not change the overall risk assessment, but may serve only
scientific or comparative purposes. If the formulation is being tested
for a reason other than low solubility of the a.i., then it is likely that
data already exist for the TGAI. Comparison of the toxicities of the
a.i. and the formulation should suffice to determine if the inert
70
-------
EPA Response:
Resolution:
ingredients cause greater toxicity. The inert/carrier control would
therefore not be necessary.
The Agency does not routinely require testing of inert/carrier
ingredients of pesticides. If this testing is necessary for the overall
assessment of the hazards of a specific chemical, it will be stipulated in
the Data Call-in Notice or in other correspondence. The Agency
agrees with Industry that the guidance provided in the EPA/NACA
document may be sufficiently adequate, on a chemical-specific basis, to
address this issue without the need for further testing.
NACA requested clarification as to what specifically was to be tested.
The Agency indicated that the terms "inert/carrier control" and "blank
formulation" are used interchangeably. If testing of the blank
formulation is needed, this will be stipulated in the DCI or other
correspondence.
14. Rejection Factor The acclimation period for the fish prior to initiation of the test was
less than half the recommended length of time.
Agency Guidance: ASTM E - 729: p. 388, 10.7 - 8. SEP Acute Toxicity Test for
Freshwater Fish: p. 4, H.B.3. Stephan: p. 28 - 29, IV.D.7.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this issue appear adequate.
Yes.
Industry Comment: The guidance documents appear contradictory:
ASTM requires maintaining the test organisms in the dilution water at
the test temperature for at least 48 hours before they are placed in the
test chambers;
SEP recommends _>. 2-week acclimation to study conditions; and
Stephan says to acclimate to changing water/temperature for 2 days,
and 2 days in that water/temperature.
Because this parameter may not be critical for the validity of a test, if
the test organisms display normal behavior and reactions, rejection
because of this parameter does not seem necessary.
71
-------
EPA Response:
There can be confusion regarding the terms "holding" and
"acclimation." The SEP inadvertently used the wrong term. When the
test organisms are brought into the laboratory, they are to held for at
least 2 weeks in holding tanks under stable water quality and
temperature conditions. After transfer to the acclimation tank from the
holding tank, the water needs to be gradually changed from 100%
holding water to 100% dilution water and test temperature over a
period of 2 or more days. They must be kept in 100% dilution water
and temperatures for at least 2 additional days before use in the toxicity
study. The Agency guidance cited above, plus the procedures
described in parts 801 and 810 of Standard Methods for the
Examination of Water and Wastewater, provide adequate details. The
Agency agrees that this factor alone may not be sole reason for
rejection of a study. This factor will be handled on a case-by-case
basis.
FRESHWATER FISH LCSO (RAINBOW) (Guideline 72-1)
1. Rejection Factor: The concentration level selected was less than 100 mg/L and was not
high enough to produce an LC50.
Agency Guidance Subd-E: p. 67, 72-1, (b)(3)(i)(A)(B) and (ii); pp. 67- 68, 72-1,
(c)(l)(i)(A)(B). FEFRA-TG: p. C-38, 72-1, no. 8. and p. C-40, 72-1,
no. 8. Stephan: pp. 6 - 7 and pp. 45 - 47, nos. 12. and 13. plus
Table 6 and section F., no. 8.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes, unless the solubility of the test material is below the level at
which any mortality occurs. In this case, the registrant is expected to
submit a request for an exception to this requirement because of
insolubility. The guidance for this type of request has been developed
jointly by EPA and NACA and is contained in the guidance document
found in Appendix A.
Industry Comment: Although the guidance documents cited clearly require the provision of
an LC50, none of the documents provides true guidance for solubility
problems. Solubility problems frequently occur below 100 mg/L,
making it sometimes difficult to test up to this level. However,
application of the theoretical rate of 10 Ib/A directly to 0.5 ft. of water
results in only 7.4 mg/L (Urban and Cook, 1986). Depending on the
concentration achieved, the data may be used for a risk assessment.
72
-------
EPA Response:
The expectation to request a waiver in these circumstances was not
known to many registrants.
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
2. Rejection Factor: The test chambers were aerated.
Agency Guidance: FIFRA-TG: p. C-38, 72-1, no. 7 and Stephan: p. 16 and p. 31, no.
2.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
The guidance documents which address this factor appear adequate.
Yes. Either the test chambers should not be aerated, or if aeration is
necessary, chemical analysis of the test solution periodically throughout
the test should be performed to verify exposure.
The guidance documents do suggest that the water is not to be aerated
unless the concentrations are measured; but, a rejection is only
warranted if there is aeration but no analysis.
EPA and NACA agree.
3. Rejection Factor: The biological loading of test vessels was twice the recommended
amount.
Agency Guidance: FIFRA-TG: p. C-37, 72-1, no. 7 and Stephan: pp. 33 - 34, no. 4.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes.
Industry Comment: Industry recognizes that the guidance documents clearly recommend
levels for biological loadings. However, if the dissolved oxygen and
ammonia levels are within acceptable limits, if there is a distinct
concentration-effect relationship, and if the concentrations are stable
over time, the biological loading alone should not be reason for
rejection. A flexible, case-by-case evaluation of an otherwise
scientifically valid study should be considered.
73
-------
EPA Response: The Agency agrees that this factor alone may not be sole reason for
rejection of a study. However, a loading factor of twice the
recommended amount is a significant deviation and can warrant
rejection of a study. Furthermore, in studies where loading exceeds
recommended levels, ammonia levels should be measured daily. This
factor will be handled on a case-by-case basis.
4. Rejection Factor: Test substance purity was not identified.
Agency Guidance: Subd-E: p. 30, 70-4, (c)(2)(i). Stephan: p. 47, no. F.2.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
The guidance documents which address this factor appear adequate.
Yes.
Industry agrees that the guidance documents provide appropriate
guidance, and that knowledge of the purity of the test substance is
mandatory for a risk assessment. However, not all impurities may
have to be known, unless there is reason to suspect the influence of
impurities. In order to avoid unnecessary delays in time and effort for
EPA and the registrant, Industry recommends that the reviewer contact
the registrant prior to finalizing and issuing the DER.
Both the Agency and Industry agree that this factor is avoidable.
information needs to be included with the study upon its first
submission to the Agency.
This
5. Rejection Factor: Inappropriate test species were used and test species were not clearly
identified.
Agency Guidance: Subd-E: p. 30, 70-4, (c)(3)(i) - (viii); p. 67, 72-1, (b)(2)(i) - (iii).
FIFRA-TG: p. C-37, 72-1, no. 3. Stephan: pp. 20 - 22, no. D.I.
and Table 3.
Assessment of
Guidance: The guidance documents which address this factor appear adequate.
Avoidable:
Yes.
Industry Comment: Industry concedes that a clear identification of the test species is
essential. However, it is not quite clear which species are appropriate.
While the FIFRA-TG and the SEP renders acceptable rainbow trout,
74
-------
EPA Response:
brook trout, and Coho salmon as cold-water species, and bluegill,
channel catfish, and fathead minnow as warm-water fish, Stephan and
ASTM also recommend the goldfish as a warm-water species. Subd-E
requests one warm- and one cold-water fish, preferably bluegill and
trout. Several other warm-water species are recommended by OECD.
Although EPA has good reason for preferring certain species, use of
other internationally accepted species in a scientifically valid test should
not lead to automatic rejection, especially in view of international
harmonization.
The preferred test species are rainbow trout and bluegill sunfish, which
are listed in Subdivision E, FIFRA-TG and the SEP. EPA is in the
process of harmonizing its testing guidelines with other groups,
including OECD. A result of this process may be the acceptance of
test species that are currently not on the preferred species list.
However, accepting a study does not necessarily mean that an
additional species that is more appropriate for the risk assessment will
not be required. It is likely that EPA will always require trout and
bluegill for comparative risk assessments.
Resolution: In the discussion, EER indicated that it would be likely that EPA will
harmonize with other groups and accept species that are not on the
current preferred test species list. However, accepting a study does not
necessarily mean that an additional species will not be required that is
more appropriate for the risk assessment. It is likely that EPA will
always require trout and bluegill for comparative risk assessments.
6. Rejection Factor: Fish acclimation records indicate the test fish were acclimated during
the time the definitive study was conducted.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
SEP Acute Toxicity Test for Freshwater Fish: p. 4, H.B.3. Stephan:
pp. 28 - 29, IV.D.6 and 7. ASTM E - 729: p. 388, 10.8.
The guidance documents which address this factor appear adequate.
Yes. The contract laboratories are required to keep records regarding
the acclimation conditions of the fish. Evidently, if the acclimation
period overlapped the period for the definitive test, the fish were not
properly acclimated to the test conditions and/or the laboratory was
careless in conducting the study.
Industry Comment: The guidance documents appear contradictory:
75
-------
EPA Response:
ASTM requires maintaining the test organisms in the dilution water at
the test temperature for at least 48 hours before they are placed in the
test chambers;
SEP recommends > 2-week acclimation to study conditions; and
Stephan says to acclimate to changing water/temperature for 2 days,
and 2 days in that water/temperature.
Industry agrees that fish should be properly acclimated prior to testing,
and that appropriate records must document the acclimation period.
But, the guidance documents should be clarified. In order to avoid
unnecessary delays in tune and effort for EPA and the registrant,
Industry recommends that the reviewer contact the registrant prior to
finalizing and issuing the DER.
This factor appears to be a Quality Assurance/Quality Control problem.
The Agency is not stating that the fish were improperly acclimated to
the test conditions prior to beginning the definitive study, but that the
dates for the acclimation period and the testing period overlapped. The
Agency's definition of holding and acclimation were described earlier.
This factor is correctable by providing records that are accurately
maintained and reported.
7. Rejection Factor: Fish were fed during the 96-hour exposure period of the study.
Agency Guidance: SEP Acute Toxicity Test for Freshwater Fish: p. 4, H.B.3. Stephan:
p. 30, IV.D.7, p. 37, IV.E.7. ASTM E - 729: p. 388, 10.8.3 and p.
391, 11.6.
Assessment of '
Guidance: The guidance documents which address this factor appear adequate.
Avoidable:
Yes.
Industry Comment: Industry agrees that the fish should not be fed during the 96-hour
exposure period.
EPA Response: EPA and NACA are in agreement.
8. Rejection Factor: Minimum limit of detectability, or the minimum quantifiable limit, was
not defined quantitatively.
76
-------
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Subd-E: p. 29, no. 70-4 (b)(l)(2), p. 32, no. 70-4 (c)(6)(ii)(F) and p.
68 no. 72-1 (c)(4). SEP Acute Toxicity Test for Freshwater Fish: p.
9, m.H. Stephan: p. 39 - 40, IV.E.lO.b.l and p. 49, IV.F.9.
FIFRA-TG: 72-1, p. C-40, no. 8. ASTM E - 729: p. 394, 15.1.9.
Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
(1992). The guidance provided in this document is adequate to address
this issue.
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
Guidance to avoid this problem in the future is offered in EPA/NACA
(1992), Conducting Acceptable Aquatic Laboratory Studies: Proposed
Guidance.
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
9. Rejection Factor: Chemical was recovered in the dilution water control at a level
exceeding those in the two lowest test concentrations.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
Subd. E: p. 26, no. 70-3 (b)(l)(2)(3). ASTM E - 729: p. 380, 4.1,
p. 389, 11.1.1.1 and p. 393, 13.1.3. SEP Acute Toxicity Test for
Freshwater Fish: p. 6, H.D.3. Stephan: p. 6, m.B and p. 45,
IV.E.12.
The guidance documents which address this factor appear adequate.
Yes. Contamination of the controls with the test chemical is
unacceptable.
Industry agrees that the control water should not contain any test
material. However, when working at very low concentrations (ppb and
ppt), sporadic "findings" in the control water may occur, due to the
fact that the method of chemical analysis is operated at its lowest level.
False positive results may be as possible as false negative results.
Depending on the level at which the test is performed and the nature of
77
-------
EPA Response:
the finding, the occurrence of positive findings should not be over-
judged.
Contamination of the control water with the test material is
unacceptable. If the chemical is being tested at very low concentrations
that bracket the limit of quantification, and there is a possibility of
obtaining false positive results, this concern must be clearly detailed
being in the report.
10. Rejection Factor There was very low recovery of the test chemical from the stock
solutions at zero and 96 hours. Measurements for the two lowest test
concentrations were only given for 96 hours, as the chemical was not
detected in the zero-hour samples.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Subd. E: p. 32, no. 70-4 (c)(6)(ii)(F). SEP Acute Toxicity Test for
Freshwater Fish: p. 6, H.D.l, p. 7, H.D.6, p. 9, m.H. and p. 11,
rV.C.l. Stephan: p. 45, IV.E.12. ASTM E-729: p. 392,
11.9.3.4(1) and (3).
Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
(1992). The guidance provided in this document is adequate to address
this issue.
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
Industry agrees that low recovery from stock solution and at 0 hours
may cause the rejection of a study.
The Agency and Industry agree that this factor may affect the
acceptance of a study. The Agency believes that the guidance provided
in the EPA/NACA document is adequate to address this issue.
11. Rejection Factor The variability limit for test concentrations was greater than 1.5.
Agency Guidance: ASTM E -729: p. 392, 11.9.3.4(2).
78
-------
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
(1992). The guidance provided in this document is adequate to address
this issue.
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
Variability of measurements may be encountered in spite of all efforts
to limit the variation, especially if test concentrations are in the ppb or
ppt range, and if the compound readily adsorbs to surfaces^ has very
limited solubility, or has limited stability. Coefficients of variation at
these concentrations are high, as noted in the literature, and variation is
more likely to be related to analysis than to malfunction of a diluter (if
used). In these or other circumstances where the analytical method is
acceptable, as determined by spiked control samples, and serious efforts
to produce less-variable data have not been successful, EPA should be
flexible—a rejection does not seem warranted. On a case-by-case
basis, EPA should accept studies with variable data as long as an
adequate defense has been made.
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
12. Rejection Factor The dilution water contained higher levels of lead, iron, and aluminum
than recommended.
Agency Guidance: Stephan: p. 15, IV.C.2. ASTM E - 729: 384, 8.2.2.1 and p. 385,
8.4.
Assessment of
Guidance:
Avoidable:
Generally, the guidance given in this area is adequate.
Yes.
Industry Comment: While Stephan on page 15 recommends specific levels of lead, iron,
and aluminum for reconstituted water, which is basically recommended,
on page 14 acceptable water for acute toxicity tests is defined as
"healthy test organisms will survive in it for the duration of acclimation
and testing without showing signs of stress. . . . Because daphnids are
more sensitive to many toxicants ... a more realistic criterion for an
79
-------
EPA Response:
Resolution:
acceptable dilution water is that first instar daphnids will survive in it
for 48 hours without food." Similarly, ASTM gives as a minimal
requirement for acceptable dilution water "that at least one test species
will survive, grow, and reproduce satisfactorily in it." Taking criteria
of biological responses to the test water into consideration, studies do
not need to be rejected on this basis.
The Agency agrees with Industry that biological responses of organisms
to the dilution water need to be considered. However,, if these and
other metals are present in the dilution water at concentrations which
exceed acceptable criteria, the Agency should be notified prior to test
initiation.
The study will not be rejected solely on the basis that metals exceeded
the criteria. The biological performance of the organisms will also be
considered in evaluating the study.
13. Rejection Factor There was no solvent control.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
Subd. E: p. 26, no. 70-3 (b)(l)(2)(3). ASTM E - 729: p. 385 - 386,
9.2.5 - 9.2.7, p. 389, 11.1.1.1 and p. 393, 13.1.3. SEP Acute
Toxicity Test for Freshwater Fish: p. 6, H.D. 3. Stephan: p. 31,
IV.E.1. FIFRA - TG: 72-1, p. C-36, no. 2 and p. C-39, no. 7.
The guidance documents which address this factor appear adequate.
Yes.
Only Stephan and ASTM require solvent controls. The Subd-E
guideline states that the solvent control "should be used." (ASTM
defines "should" to mean that the specified condition is recommended
and ought to be met, if possible. Although violation of one "should" is
rarely a serious matter, violation of several will often render the results
questionable.)
Industry agrees that the use of a solvent control is desirable and
scientifically appropriate. However, if the solvent control would have
displayed signs of intoxication, but was omitted, these signs of
intoxication would be attributed to the test compound. Although this
might not reflect the true toxicity, for risk assessment purposes it would
reflect a "worst-case" scenario, and safety for the assessment. It
80
-------
EPA Response:
Resolution:
therefore may not be necessary to reject a study for this parameter
alone.
The Agency guidance on this factor is adequate. Industry agrees with
the Agency that using a solvent control is scientifically valid and
correct.
NACA suggested the Agency indicate in guidance documents that a
solvent control "must" (instead of "should") be included. The Agency
agrees.
14. Rejection Factor The results for several of the test concentrations were obtained from
separate tests conducted a few weeks after the definitive study.
Agency Guidance: Subd. E: p. 27, 70-3, (b)(4). Stephan: p. 31, IV.E.l. ASTM E -
729: p. 380, 4.1.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Resolution:
The guidance documents which address this factor appear adequate.
Yes.
Industry basically agrees with EPA's assessment. Adding
concentrations at a later time may occur, however, if in cases of very
flat concentration-effects relations, space in the diluter system or in the
facility is insufficient. This may arise especially if the determination of
a NOEL is required (studies of NACA members were rejected for this
reason). Add-on studies must establish a continuation of the biological
responses (e.g., by overlapping of tested concentrations). Study
rejection therefore should be based on a case-by-case consideration.
Adding concentrations from a separate study and combining them with
the ones used in the definitive study is inappropriate. The arguments
provided by Industry are unpersuasive. If the definitive study does not
generate an acceptable dose-response curve, a new study needs to be
conducted. It is the responsibility of the registrant to ensure that the
design of the test system can adequately conduct the study according to
guideline requirements.
NACA assumed that the results of the study were combined to produce
a mortality/effect profile which would allow an LC50/EC5o or a NOEL
to be estimated. NACA agrees that combining results of tests
conducted at different tunes is not desirable, especially to complete
81
-------
dose-response curves. For clarification, NACA requested that the
Agency reiterate in its response that NOELs are not required for acute
studies.
15. Rejection Factor Not all test solutions were measured at 96 hours, as well as at zero
hours.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Subd. E: p. 29, 70-4 (b)(l)(2), p. 32, 70-4 (c)(6)(ii)(F). Stephan: p.
41 - 42, IV.E.10.B.(3). SEP Acute Toxicity Test for Freshwater Fish:
p. 7, H.D.6 p. 9, m.H and p. 11, IV.C.l. ASTM E - 729: p. 392,
11.9.3.3 and 11.9.3.4(1) and (3).
Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
(1992). The guidance provided in this document is adequate to address
this issue.
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
Industry agrees that rejections in this area were not avoidable due to
lack of clarity. Studies were accepted in the past by EPA with and
sometimes without analytical measurements.
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
16. Rejection Factor A control group of fish for the inert/carrier ingredients present in the
formulation was lacking. (The test.material was 42% a.L and 58%
inert/carrier ingredients.)
Agency Guidance: Stephan: p. 31, IV.E.l.
Assessment of
Guidance:
Normally, this guidance is adequate. However, in this case the
scientist concluded that, without an inert/carrier control (i.e., blank
formulation) it was not possible to discern whether the toxic effects
were due to the a.l, the inerts/carrier, or to both. This guidance may
need strengthening.
82
-------
Avoidable:
No.
Industry Comment: While Stephan requires an additive control, it also states that "none of
the ingredients of the mixture or formulation is considered an additive."
Therefore, one cannot conclude that testing of blank formulations is
needed. Since an a.i. will be applied to the environment only as a
formulation, knowledge of the contribution of the formulation additives
may not change the overall risk assessment, but may serve only
scientific or comparative purposes. If the formulation is being tested
for a reason other than low solubility of the a.i., then it is likely that
data already exist for the TGAI. Comparison of the toxicities of the
a.i. and the formulation should suffice to determine if the inert
ingredients cause greater toxicity. The inert/carrier control would
therefore not be necessary.
EPA Response:
Resolution:
The Agency does not routinely require testing of inert/carrier
ingredients of pesticides. If this testing is necessary for the overall
assessment of the hazards of a specific chemical, it will be stipulated in
the Data Call-In Notice or in other correspondence. The Agency
agrees with Industry that the guidance provided in the EPA/NACA
document may be sufficiently adequate, on a chemical-specific basis, to
address this issue without the need for further testing.
NACA requested clarification as to what specifically was to be tested.
The Agency indicated that the terms "inert/carrier control" and "blank
formulation" are used interchangeably. If testing of the blank
formulation is needed, this will be stipulated in the DCI or other
correspondence.
17. Rejection Factor The weights of the test fish exceeded the recommended range.
Agency Guidance: Subd. E: p. 28, 70-3 (d)(2) and p. 67, 72-1 (b)(2)(ii) and (iii). SEP
Acute Toxicity Test for Freshwater Fish: p. 3, H.B.2. Stephan: p. 23
- 24, IV.D.3. ASTM E - 729: p. 386 - 387, 10.2. FIFRA -TG: 72-
1, C-37, no. 3.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes.
Industry Comment: While the guidance for this factor is appropriate, the size of the test
fish may become a problem due to the availability of seasonally
83
-------
EPA Response:
spawning fish, especially if a test has to be performed to meet a
specific deadline. Rejection may not be necessary, if the deviation is
not excessive and the results of the study are comparable to others.
The range of preferred weights for test fish is rather wide, thereby
giving the individual registrants flexibility in obtaining appropriate sized
fish for the test. The Agency requires consistency in size of test fish to
allow for comparison of test results among chemicals.
18. Rejection Factor The test temperature exceeded that recommended for rainbow trout.
Agency Guidance: FEFRA -TG: 72-1, C-37, no. 7. SEP Acute Toxicity Test for
Freshwater Fish: p. 5, H.C.2 and p. 10, IV.C.l. ASTM E - 729:
386, Table 4. Stephan: p. 21, Table 3.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes.
Industry Comment: Industry agrees that a rejection due to this factor is avoidable.
However, depending on the extent, it may not be necessary to reject a
study, especially when considering harmonization issues. OECD
allows testing at 13-17°C, Germany requires 15°C. Testing in this
range, instead of 12 + 1°C, may not lead to such different results as to
justify a repetition of a test.
EPA Response:
The issues of preferred test temperature for selected fish species as
covered in the guidelines harmonization process is being addressed in
other venues. No further comment on this factor is necessary.
19. Rejection Factor The biological loading of the system was greater than recommended.
Agency Guidance: SEP Acute Toxicity Test for Freshwater Fish: p. 5, H.C.5. FEFRA -
TG: 72-1, C-37, no. 7. Stephan: p. 33 - 34, IV.E.4 and 5. ASTM
E - 729: p. 390, 11.4.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes.
84
-------
Industry Comment: Industry recognizes that the guidance documents clearly recommend the
level for biological loading. However, if the dissolved oxygen level is
within acceptable limits, and if there is a distinct concentration-effect
relationship, the biological loading alone should not be reason for
rejection. A flexible, case-by-case evaluation of otherwise scientifically
valid studies should be considered.
EPA Response:
The Agency agrees that this factor alone may not be sole reason for
rejection of a study. However, a loading factor greater than the
recommended amount is a significant deviation and can warrant
rejection of a study. Furthermore, in studies where loading exceeds
recommended levels, ammonia levels should be measured daily. This
factor will be handled on a case-by-case basis.
20. Rejection Factor The acclimation period for the fish prior to initiation of the test was
less than half the recommended length of time.
Agency Guidance: ASTM E - 729: p. 388, 10.7 - 8. SEP Acute Toxicity Test for
Freshwater Fish: p. 4, H.B.3. Stephan: p. 28 - 29, IV.D.7.
Assessment of
Guidance:
Avoidable:
Generally, the guidance in this area is adequate.
Yes.
Industry Comment: The guidance documents appear contradictory:
• ASTM requires maintaining the test organisms in the dilution water at
the test temperature for at least 48 hours before they are placed in the
test chambers;
• SEP recommends > 2-week acclimation to study conditions; and
• Stephan says to acclimate to changing water/temperature for 2 days,
and 2 days in that water/temperature.
This parameter may not be critical to the validity of the test, if the test
organisms display normal behavior and reactions. In order to avoid
unnecessary delays in time and effort for EPA and the registrant,
industry recommends that the reviewer contact the registrant prior to
finalizing and issuing the DER.
85
-------
EPA Response:
There can be confusion regarding the terms "holding" and
"acclimation." The SEP inadvertently used the wrong term. When the
test organisms are brought into the laboratory, they are to held for at
least 2 weeks in holding tanks under stable water quality and
temperature conditions. After transfer to the acclimation tank from the
holding tank the water needs to be gradually changed from 100%
holding water to 100% dilution water and test temperature over a
period of 2 or more days. They must be kept in 100% dilution water
and temperatures for at least 2 additional days before use in the toxicity
study. The Agency guidance cited above, plus the procedures
described in parts 801 and 810 of Standard Methods for the
Examination of Water and Wastewater, provide adequate details. The
Agency agrees that this factor alone may not be sole reason for
rejection of a study. This factor will be handled on a case-by-case
basis.
21. Rejection Factor Fish mortality during the acclimation period was higher than
recommended.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
ASTM E - 729: p. 388 - 389, 10.8 - 10.9. SEP Acute Toxicity Test
for Freshwater Fish: p. 4, H.B.3. Stephan: p. 28 - 29, IV.D.7.
Generally, the guidance hi this area is adequate.
Yes.
Industry agrees that mortality above the levels specified in the guidance
documents may indicate problems with the test organisms, and
therefore may be reason to invalidate a study.
EPA and NACA are in agreement.
22. Rejection Factor The dissolved oxygen during the test was supersaturated (over 100%
saturation).
Agency Guidance: SEP Acute Toxicity Test for Freshwater Fish: p. 7, n.D.5. Stephan:
p. 31, IV.E.2. ASTM E - 729: p. 390, 11.4.2 and p. 393, 13.1.9.
Assessment of
Guidance:
Supersaturation of oxygen is not directly addressed in our guidance
documents.
86
-------
Avoidable:
Industry Comment:
EPA Response:
Resolution:
Not at this time.
Oversaturation can occur after intensive aeration of test aquaria prior to
testing at lower concentrations. Normally this should not be persistent
over the entire testing period, but would be seen only at the beginning
of the test. If the observation is only made then, if the control fish
behave and look normal, and if chemical analysis verifies the
concentration level, there should be no reason to reject the study. This
factor may be solved on a case-by-case consideration, and not by a
check-box approach.
The Agency agrees with Industry's comments. This factor will be
handled on a case-by-case basis.
NACA encouraged the Agency to indicate that if the animals in the
control group perform normally, then the test is acceptable. The
Agency indicated that it is unlikely that a study would be rejected for
this factor alone.
ACUTE LCSO FRESHWATER INVERTEBRATES (Guideline 72-2)
1. Rejection Factor: Organisms were not randomly distributed to test vessels.
Agency Guidance: Subd-E: p. 28, 70-3, no. (d)(3). Stephan: p. 30, E.I and p. 31, E.I.
. . .cont.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
The guidance documents which address this factor appear adequate.
Yes.
The guidance documents clearly state that organisms must be randomly
distributed to test vessels and that this is a valid scientific precaution
against skewing results. As long as some unbiased selection is used,
however, it is doubtful that this factor would significantly impact the
results of this study. In fact, it is not clear that a true random
distribution is needed. Unless the dose-response curve shows erratic,
highly variable data points, this should not be reason in itself to reject a
study. Studies should not be rejected for this factor if the mortality
data are not irregular.
The guidance is clear.
87
-------
Resolution:
EPA/NACA agreed to add a statement as follows: "Impartially
distributing organisms among randomly placed test chambers is an
acceptable procedure for randomizing organisms."
2. Rejection Factor: Temperature of the water was not monitored during the tests.
Agency Guidance: FIFRA-TG: p. C-41, 72-2, no. 2., and p. C-42, 72-2, no. 7., and p.
C-44, 72-2, no. 7. Stephan: pp. 31 - 33, no. 3.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes.
Industry Comment: Unless temperature data can be supplied in some other way, this is a
clear violation of the regulations. The study in all probability will have
to be re-done, and this study will become supplemental data.
EPA Response:
EPA and NACA are in agreement.
3. Rejection Factor: Chemical analyses (measurements of test concentration levels) were not
performed on the test solutions.
Agency Guidance: Subd-E: p. 72, 72-2,(c)(4). Stephan: pp. 39 - 42, no.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate. (In
this particular situation, the registrant had submitted previous data
which showed that measured concentrations were substantially lower
than nominal concentrations. Based on this information, the Agency
concluded that the present (i.e., "rejected") study should have utilized
chemical analyses.)
Yes.
Industry Comment: There is an overall lack of clarity in the guidance documents
concerning analytical determinations (e.g., how many samples to take,
and when to take them).
Since previous tests showed that nominal concentrations were difficult
to obtain/maintain, analysis should have been done (assuming the
88
-------
EPA Response:
variations between nominal and actual test concentrations in the
previous tests were greater than 30%, and that physical data—i.e.,
stability and solubility—suggested that maintaining a constant exposure
may be a problem). Test solutions should be maintained within 70% to
130% of nominal concentration, unless all efforts to do so fail and
these efforts are carefully documented (EPA/NACA (1992), Conducting
Acceptable Aquatic Laboratory Studies: Proposed Guidance).
It is evident that in this test situation chemical analyses were necessary.
More guidance on the need for measuring test concentrations is
provided in the EPA/NACA document.
4. Rejection Factor: A control group of animals for the inert/carrier ingredients present in
the formulation was lacking.23
Agency Guidance: Stephan: p. 31, E.I. . . . cont., 2nd paragraph.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Normally, this guidance is adequate. However, in this case the
scientist found that, without an inert/carrier control (i.e., blank
formulation) it was not possible to discern whether the toxic effects
were due to the a.i., the inerts/carrier, or to both. This guidance may
need strengthening.
Yes.
While Stephan requires an additive control, it also states that "none of
the ingredients of the mixture or formulation is considered an additive."
Therefore, one cannot conclude that the testing of blank formulations is
needed. And, as an a.i. will be applied to the environment as a
formulation, knowledge of the contribution of the formulation additives
may not change the overall risk assessment, but may serve only
scientific or comparative purposes.
The Agency does not routinely require testing of inert/carrier
ingredients (i.e., blank formulation) of pesticides. If this testing is
necessary for the overall assessment of the hazards of a specific
chemical, it will be stipulated in the Data Call-in Notice or in other
correspondence. The Agency agrees with Industry that the guidance
provided in the EPA/NACA document may be sufficiently adequate, on
23 The test material was 30% a.i. and 70% carrier/inert ingredients.
89
-------
Resolution:
a chemical-specific basis, to address this issue without the need for
further testing.
NACA requested clarification as to what specifically was to be tested.
The Agency indicated that the terms "inert/carrier control" and "blank
formulation" are used interchangeably. If testing of the blank
formulation is needed, this will be stipulated in the DCI or other
correspondence.
5. Rejection Factor: Minimum limit of detectability, or the minimum quantifiable limit, was
not defined quantitatively.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Subd-E: p. 29, no. 70-4 (b)(l)(2), p. 32, no. 70-4 (c)(6)(ii)(F) and p.
68 no. 72-1 (c)(4). SEP Acute Toxicity Test for Freshwater
Invertebrates: p. 9, m.H. Stephan: p. 39 - 40, IV. E.lO.b.l and p.
49, IV.F.9. FIFRA-TG: 72-2, p. C-42, no. 11. ASTM E - 729: p.
394, 15.1.9.
Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
(1992). The guidance provided in this document is adequate to address
this issue.
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
Industry Comment: Guidance to avoid this problem is offered in EPA/NACA (1992),
Conducting Acceptable Aquatic Laboratory Studies: Proposed
Guidance.
EPA Response:
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
6. Rejection Factor: The variability limit for test concentrations was greater than 1.5.
Agency Guidance: ASTM E-729: p. 392, 11.9.3.4(2).
Assessment of
Guidance: Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
90
-------
Avoidable:
Industry Comment:
EPA Response:
(1992). The guidance provided in this document is adequate to address
this issue. ,
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
Variability of measurements may be encountered in spite of all efforts
to limit the variation, especially if test concentrations are in the ppb or
ppt range, and if the compound readily adsorbs to surfaces, has very
limited solubility, or has limited stability. Coefficients of variation at
these concentrations are high, as noted in the literature, and variation is
more likely to be related to analysis than to malfunction of a diluter (if
used). In these or other circumstances where the analytical method is
acceptable, as determined by spiked control samples, and serious efforts
to produce less-variable data have not been successful, EPA should be
flexible—a rejection does not seem warranted. On a case-by-case
basis, EPA should accept studies with variable data, as long as an
adequate defense has been made.
The Agency believes that the guidance provided hi the EPA/NACA
document is adequate to address this issue.
7. Rejection Factor: Not all test concentrations were measured at both zero hours and 48
hours.
Agency Guidance:
Assessment of
Guidance:
Subd. E: p. 29, 70-4 (b)(l)(2), p. 32, 70-4 (c)(6)(ii)(F). Stephan: p.
41 - 42, IV.E.10.B.(3). SEP Acute Toxicity Test for Freshwater
Invertebrates: p. 7, H.D.6 p. 9, m.H, and p. 11, IV.C.L ASTM E -
729: p. 392, 11.9.3.3 and 11.9.3.4(1) and (3).
Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
(1992). The guidance provided in this document is adequate to address
this issue.
Avoidable:
Industry Comment:
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
EPA should not require that all test concentrations be analyzed, unless
there is reason to suspect that there could be a problem with
91
-------
EPA Response:
Resolution:
maintaining exposure. If the data provided showed that the compound
was stable over the 48-hour test period, and the assayed concentrations
are within 70% to 130% of nominal, the study should be acceptable.
The concern could be further mitigated by the physical characteristics
of the compound.
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
NACA agreed with the response, and EPA agreed that studies which
analyzed low, medium, and high test concentrations would be
"grandfathered" so they would not be rejected. Studies initiated after
publication of the RRA need to follow the guidance document.
8. Rejection Factor: The dilution water contained higher levels of lead, iron, and aluminum
than recommended.
Agency Guidance: Stephan: p. 15, IV. C. 2. ASTM E - 729: 384, 8.2.2.1 and p. 385,
8.4.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Resolution:
Generally, the guidance given in this area is adequate.
Yes.
If EPA has data to show that the levels of these metals in the dilution
water used in the test will have a detrimental effect on Daphnia, and if
the controls show more than 10% mortality in a static system or 5% in
a flow-through system, then the study should be rejected. Industry
supports the Stephan concept. Taking criteria of biological responses to
the test water into consideration, however, studies do not need to be
rejected on this basis.
The Agency agrees with Industry that biological responses of organisms
to the dilution water need to be considered. However, if these and
other metals are present in the dilution water at concentrations which
exceed acceptable criteria, the Agency should be notified prior to test
initiation.
The Agency responded that the study would not be rejected solely on
the basis that metals exceeded the criteria. The biological performance
of the organisms would also be considered in evaluating the study.
92
-------
9. Rejection Factor: Percent a.i. of the tested formulation was not given. The test material
was not identified by lot or batch numbers. There was no indication if
the concentrations used in the study were based on the percent a.i. or
total formulated product.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Subd. E: p. 30, 70-4 (c)(2)(i) and (ii). Stephan: p. 47, IV.F.2. SEP
Acute Toxicity Test for Freshwater Invertebrates: p. 8, m.A. ASTM
E-729: p. 394, 15.1.2.
The guidance documents which address this issue appear adequate.
Yes.
EPA is reasonable in requiring a description of the test material. In
order to avoid unnecessary delays in tune and effort for EPA and the
registrant, however, Industry recommends that the reviewer contact the
registrant prior to finalizing and issuing the DER.
EPA and NACA are in agreement.
10. Rejection Factor The photoperiod was not 16-hour light/8-hour dark as recommended,
but total darkness.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
SEP Acute Toxicity Test for Freshwater Invertebrates: p. 5, H.C.4.
Stephan: p. 10, IV.B.l. ASTM E-729: p. 381, 7.1.
A photoperiod of total darkness is not directly addressed in our
guidance documents.
Not at this time.
Unless there is data to show that Daphnia are adversely affected by
changing the photoperiod over 48 hours, the requirement should only
be a recommendation. In cases where a compound undergoes rapid
aqueous photolysis, total darkness may be the photoperiod of choice.
Furthermore, complete darkness is acceptable for the OECD 202 24-
hour EC50 Daphnia acute mobilization test. In the interest of
international harmonization of pesticide testing guidelines, this should
not be a reason to reject a study if adequate explanations have been
made.
93
-------
EPA Response:
A photoperiod is useful because it simulates natural conditions and
Daphnia generally do not exist under total darkness.
It would be less stressful to the test organisms to experience standard
periods of light and darkness, than to be shocked suddenly with light at
observations times. The Agency agrees that photolytically sensitive
chemicals should be tested in darkness or under flow-through
conditions.
11. Rejection Factor No temperature data were provided to indicate if the chambers were
monitored at least every 6 hours.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Subd. E: p. 29, 70-4 (b)(l), p. 31, 70-4 (c)(6)(ii)(B), p. 71, 72-2
(c)(3)(vii). Stephan: p. 49, IV.F.10. SEP Acute Toxicity Test for
Freshwater Invertebrates: p. 7, H.D.5 and p. 9, m.G. FIFRA-TG:
72-2, p. C-44, no. 7. ASTM E-729: p. 392, 11.9.2.2 and p. 393,
13.1 and 13.1.10 to 13.1.12.
This guidance may need clarification. In general, the guidance states
that the temperature range and average temperature should be reported.
However, the study can be rejected if temperature is not measured
enough or if the deviations are too great. Reporting only the range and
mean will not provide sufficient data to determine the study's
acceptability.
No.
EPA needs to clarify what data should be submitted. Temperature data
needs to be collected, but submission of tabled data should be
acceptable. Studies that have been rejected due to the absence of raw
data should be upgraded if the appropriate data is provided to EPA. If
the reviewer or designee contacted the registrant for the needed
information prior to finalizing and issuing the DER, significant time
and rework could be avoided.
Submission of the daily range and daily average temperatures measured
during the course of the study is sufficient. Any large deviations in
temperature must be explained in the report for the study.
12. Rejection Factor Use of dechlorinated water as a portion of the dilution water is not
recommended.
94
-------
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
SEP Acute Toxicity Test for Freshwater Invertebrates: p. 4, H.C.l.
ASTM E-729: p. 384, 8.2.3. Stephan: p. 15, IV.C.2.
The guidance documents which address this factor appear adequate.
Yes.
The guidance documents are somewhat contradictory. The SEP states
(p. 4), "Dechlorinated water should not be used because removal of
chlorine is rarely complete and residual chlorine can be quite toxic to
aquatic organisms. The dilution water must be able to support the test
animals without stress. Organisms should be able to survive, grow,
and reproduce in acceptable diluent." If a laboratory establishes that
reproduction is acceptable using in-house dechlorinated water, then the
water is satisfactory and the test should be acceptable.
A key word in the above rejection factor is "recommended."
Recommended guidance helps to establish "generally recognized as
acceptable" boundaries. However, it may not be prudent to use them
in all cases to reject a study. Taking criteria of biological responses to
the test water into consideration, it may not have been appropriate to
reject the study.
The Agency strongly recommends against the use of dechlorinated
water for the reasons stated in the SEPs for freshwater organism
testing. If the use of dechlorinated water cannot be avoided, the
biological responses of the control organisms and chemical analyses
must meet acceptable criteria.
ACUTE LCSO ESTUARINE AND MARINE ORGANISMS (FISH) (Guidelines
72-3A & D)
1. Rejection Factor: Unexplained variability between the 0-hour and 96-hour measured
concentrations at the two highest test levels (i.e., mean measured
concentrations substantially increased).
Agency Guidance:
Assessment of
Subd-E: pp. 30 - 31, 70-4 (c)(4)(i) - (viii); p. 75, 72-3 (c)(3)(xi) and
(c)(4). FIFRA-TG: pp. C-48, 72-3 no. 13. and C-50, 72-3, no. 9.
Stephan: pp. 39 - 42, no. 10. a.b.(l)(2)(3).
95
-------
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Generally, the guidance given in this area is adequate, and unexplained
variations such as this are handled on a case-by-case basis.
Yes, if the variation was due to deficient chemical analysis techniques
or inappropriate test conditions. However, if the variation was due
primarily to chemical characteristics, it is unlikely that it could be
avoided.
Not avoidable. Measured concentrations may increase or decrease
when testing close to the solubility of the compound in flow-through
situations, if the materials is readily adsorbed to walls.
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
2. Rejection Factor: The concentration levels selected were less than 100 mg/L and because
of solubility problems were not high enough to produce an LC50.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
LC50 Determinations: Subd-E: p. 73, 72-3 (b)(3)(i)(A)(B). FIFRA-
TG: p. C-48, 72-3, no. 8. and 9. and p. C-50, 72-3, no. 8. Stephen:
pp. 6-7 and pp. 45 - 49, nos. 12. and 13. plus Table 6 and section
P., no. 8.
Solubility Issues: Stephen: p. 35, no. 5., 1st paragraph.
The guidance documents which address the LC50 or 100 mg/L factor
appear adequate. The guidance addressing solubility issues may need
improvement.
The problem may not be avoidable, but this eventuality should be
accommodated by study evaluation policy,, If solubility of the test
material is below the level at which any mortality occurs, the registrant
is expected to submit a request for an exception to this requirement
because of insolubility. The guidance for this type of request is being
developed and will be disseminated to the registrants.
Although the guidance documents cited clearly require the provision of
an LC50, none of the documents provides true guidance for solubility
problems. Solubility problems frequently occur below 100 mg/L,
making it sometimes difficult to test up to this level. In these cases,
the significance of the 100 mg/L criterion needs to be reconsidered.
For example, assuming that a theoretical application rate of 10 Ib/A is
96
-------
EPA Response:
applied directly to 0.5 ft. of water, the resulting concentration would be
only 7.4 mg/L, well below the 100 mg/L criterion. Hence, depending
on the concentration achieved, the data may be used for a risk
assessment. Many registrants were not aware that they would need to
request a waiver in these circumstances.
The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
3. Rejection Factor: Dissolved oxygen levels fell below 60% saturation at 48 hours and
below 40% saturation at 96 hours in two test levels.
Agency Guidance: SEP Estuarine Fish Acute Toxicity: P. 7, H.E.5. Stephan: p. 31,
IV.E.2. ASTM E - 729: p. 390, 11.4.2 and p. 393, 13.1.9.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Resolution:
The guidance documents which address this factor appear adequate.
Yes.
While boundaries help to establish if a study is acceptable, it may not
be necessary to use them in all cases to reject a study. Performing a
test with solvent at higher temperatures and larger fish can lead to
oxygen depletion. If the fish behave normally, results of the study may
be acceptable. Also, fish stressed by lower oxygen levels may be a
worse case than under normal circumstances. Therefore, flexible use
of this rejection factor may be indicated.
The purpose of a toxicity test is to test organisms at optimal conditions
to toxic agents. Lower D.O. generally results in altered metabolic
rates, which alters chemical uptake. Therefore, it is possible to still get
a good dose-response curve, but the slope of the curve could be greatly
affected. A study is unacceptable if the dissolved oxygen levels fall
below acceptable limits in the control and treatment chambers. Slight
aeration is permissible to counter oxygen depletion below the
acceptable limits provided chemical concentrations are adequately
measured. Refer to ASTM E-729, 11.2.2 for guidance on how to
properly aerate test chambers.
The Agency agreed that rejection is not necessary if effective corrective
action is taken immediately (i.e., the corrective action raises the D.O.
to recommended levels by the next measurement tune or within 24
hours, whichever is shorter, and there is no adverse effect noted in
97
-------
control organisms). However, the study will be rejected if corrective
measures fail to bring the D.O. to acceptable levels.
4. Rejection Factor: Although a static test system was used, the concentrations in the test
vessels should have been measured, as other studies with the same test
chemical showed that the concentrations varied as much as 30% during
a 96-hour exposure period. Therefore, the nominal concentrations did
not accurately reflect the true concentrations to which the fish were
exposed.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
SEP Estuarine Fish Acute Toxicity: p. 7 H.E.6 p. 10, m.C.L ASTM
E-729: p. 392, 11.9.3.3.
Sufficient guidance did not exist until EPA/NACA produced
Conducting Acceptable Aquatic Laboratory Studies: Proposed Guidance
(1992). The guidance provided in this document is adequate to address
this issue. (In this particular situation, the registrant had submitted
previous data showing that there was a substantial amount of variation
between measured and nominal concentrations. Based on this
information, the Agency concluded that the present (i.e., "rejected")
study should have utilized chemical analyses.)
No, due to lack of clarity in issues concerning analytical
determinations. These issues are now resolved in the EPA/NACA
document. Future submissions containing this factor will be rejected.
Industry agrees that, due to lack of clarity., rejections in this area were
not avoidable. Studies were accepted in the past by EPA with and
sometimes without analytical measurements. According to the new
EPA/NACA guidance document, a variation of 30% is within
acceptable ranges for demonstrating stability, and does not necessitate
analytical verification of concentrations.
Variability of measurements may be encountered in spite of all efforts
to limit the variation, especially if test concentrations are in the ppb or
ppt range, and if the compound readily adsorbs to surfaces, has very
limited solubility, or has limited stability. In these or other
circumstances where the analytical method is acceptable, as determined
by spiked control samples, and serious efforts to produce less-variable
data have not been successful, EPA should be flexible and rejection
does not seem warranted. On a case-by-case basis, EPA should accept
98
-------
studies with variable data, as long as an adequate defense has been
made.
EPA Response: The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
ACUTE LC50 ESTUAREVE AND MARINE ORGANISMS (MOLLUSKS) (Guidelines
72-3B and E)24
1. Rejection Factor: There was insufficient new shell growth in the control oysters to
adequately determine the effect of the pesticide on shell deposition.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Subd-E: p. 26, 70-3 (b)(2) and (3). Stephan: pp. 30 - 31, no. E.I.
and p. 39, no. 9., last paragraph.
Guidance in this area could be improved, since published guidance does
not specifically address shell growth of oysters.
Possibly.
Industry Comment: It would not be possible to avoid rejection of the studies, given the
current guidance, since there are no accepted EPA guidelines which
.specify minimum growth criteria for control oysters. Even if EPA
adopted the current draft proposal of 2 mm of growth for control
oysters, many of the studies would be rejected and would need to be
repeated, resulting in delays, the wasting of resources, and the often
needless use of additional test organisms.
It is difficult from the description of the rejection factor to determine
how "insufficient" the growth of control oysters was in the rejected
study. However, industry is aware of a draft document issued on 7
August 1990 by the EPA, in which a minimum growth criterion for
control oysters of 2 mm in 96 hours is proposed. Industry will assume
that the insufficient growth is based on the draft criteria of less than
2 mm.
Literature cited by EPA in justifying the 2 mm control shell growth
(Epifanio et al. 1975; Epifanio and Mootz, 1976; Epifanio, 1979) are
24 Note that there are two types of mollusc studies: a 96-hour shell deposition test and a 48-
hour embryo larvae study. The former measures sheE growth; the latter measures mortality to
embryos/larvae.
t
99
-------
not supportive of the EPA position that it is normal to obtain up to 1.0
mm of peripheral shell growth per day. Under conditions similar to
those used at many contract laboratories, Epifanio (1979) reported that
over a 42-day test period employing 15 different diets, oysters grew
between 1.22 mm and 8.09 mm. Ignoring the two lowest values as
outliers, the remaining 13 diets produced a mean shell growth of 0.14
mm per day, or 0.56 mm in 96 hours. Epifanio and Mootz (1976)
reported that the fastest growing oysters in their experiments in
recirculating mariculture systems only deposited an average of 0.58 mm
per week, or 0.33 mm in 96 hours. No reference to shell deposition
rates up to 1.0 mm per day exist in the three Epifanio articles.
Examination of 16 groups of control oyster growth data obtained
between August, 1990, and November, 1991, by Toxikon
Environmental Sciences, showed a mean oyster growth rate range .of
0.7 to 3.3 mm per group in 96-hour tests. Growth of individual
oysters ranged from 0 to 5.6 mm. Toxikon also provided data from 7
random groups of control oysters tested in 1983 which showed that
although new shell growth averaged 2.1 and 2.7 mm per group, 31%
of these oysters produced less than 2 mm of growth over 96 hours. A
summary of control oyster growth from Springborn Laboratories, Inc.,
consisting of 5 years of data and 68 studies (approximately 4,000
individuals tested) showed a mean oyster growth rate range of 0.9 to
3.8 mm hi 96-hour studies.
A study was conducted by Springborn Laboratories, Inc., to determine
the sensitivity of 2 different groups of oysters to a test material. The
study compared the response of a group of Crassostrea virginica to a
group of the "fast-growing" Wilde strain of C. virginica. This study
was designed to determine if differences in rate of growth would affect
the establishment of an EC50. Using potassium chromate as the test
substance, the EC50 for typical C. virginica was 5.6 mg/L and the EC50
for the Wilde strain of C. virginica was calculated to be 4.7 mg/L.
The typical C. virginica deposited an average of 0.8 mm over the
96-hour test, and the Wilde strain deposited an average of 3.1 mm.
The conclusion was that the 0.9 mg/L difference in EC50 values could
easily be attributed to test biological variability in oyster response, and
that the sensitivity to toxicants is not related to their inherent growth
rate.
Discussions with oyster growers has identified important information
concerning the growth endpoint. Growth rate in oysters varies greatly
depending on season and strain. Even when a fast-growing strain such
as the Wilde is used, the variability within a spawn can be large.
100
-------
EPA Response:
"Even under the best possible conditions healthy oysters may not on
average grow the minimum 2 mm required during the 96-hour testing
period" (M. Cummins, P. Cummins Oyster Company., personal
communication., letter on file). Gary Wickfors of the National Marine
Fisheries Service (Milford, CT) stated that 2 mm of growth in 4 days
would occur only under the most optimal nutritional and temperature
conditions (personal communication, letter on file). He also states that
a 10-day acclimation period to a new food source or water temperature
is insufficient to maximize optimal growth. Dr. Wickfors suggests that
the seasonal variability in growth can be quite large and if the assay is
needed in all seasons, then some changes in the protocol should be
considered. He suggested that a modification of the acclimation
schedule or relaxation of the 2 mm control growth requirement would
be appropriate.
Oyster shell growth is variable, depending on size, age, source of the
oyster, diet, or season of the year. Large variability hi oyster growth
is often exhibited in oysters from the same source and maintained under
the same test conditions. The validity of a study should not be based
on the highest growth rate obtained in reported studies or even on the
average growth rate from many studies. Inherent biological variability
in growth restricts the utility of a stringent growth performance
standard. Since effects on oyster growth are calculated relative to the
control growth for every study, the calculated EC50 and NOEL
endpoints, as well as no-effect limit tests, would not be expected to be
different if the shell growth was 0.8 mm (the lower end of the range of
historical control data) or 2.0 mm over 96 hours. The range of the
means, not the overall mean, is what should be considered in
establishing a growth guideline. Reviewers must examine the data
closely and not rely on a strict growth criteria that might invalidate
reasonable and technically sound testing results.
In a memorandum from Paul Schuda, Deputy Director of EFED, to
Daniel Barolo, Director of SRRD, and Lawrence Culleen, Acting
Director of RD, dated Oct. 29, 1992, shell growth in mollusk studies
was addressed. (See attachment 2). To summarize, the memo stated
that EFED scientists, after consulting with experts in the field of
estuarine organism testing (Roger Mann, Virginia Institute of Marine
Science, and Jerry Zaroogian, USEPA - ERL, Narragnasett, RI),
determined that 2 mm of new shell growth in clean water controls in a
96-hour period is easily obtainable and is the minimum growth to allow
for valid comparisons.
101
-------
Resolution:
EFED proposes to present this determination of minimum shell growth
to the Science Advisory Panel (SAP) prior to formally amending the
appropriate SEP. In the interim, if the sole deficiency in the study is
insufficient shell growth (i.e., less than 2 mm in the controls) the study
is supplemental (i.e., scientifically sound but does not fulfill guideline
requirements). If the mollusks do not appear to be more sensitive than
other tested organisms to that pesticide, the study does not need to be
.repeated. If mollusks do appear to be more sensitive, the potential for
the pesticide to affect estuarine organisms will be considered before
determining the need to repeat the study. Refer to attachment 2 for
specific details. This memo provides the necessary guidance until the
SEP is amended to address this issue.
The Epifanio studies were not cited by EPA (EEB) to justify the 2 mm
control shell growth.
NACA consulted experts in the field (P. Cummins, Cummins Oyster
Company and Gary Wickfors, National Marine Fisheries Service,
Milford, CT) and found that 2 mm of growth in 96 hours was not
"easily obtainable" and studies are often not submitted and are redone
in order to obtain 2 mm of growth in controls, reflecting a much higher
failure rate of studies than acknowledged by the rejection rate analysis.
This was due to seasonal variation and differences between strains.
NACA held that currently there was just only one supplier who could
produce material to meet this criterion, but that they could not supply
the whole industry and the source was not secure as it was based on a
particular location in the Chesapeake prone to disturbance. RRB
named Sammy Ray (Texas A&M) and Liam Anderson as experts
advising that 2 mm of growth in Oysters was "easily obtainable" and
added that at present they were not rejecting studies obtaining < 2 mm
but > 1 mm growth in controls.
Listed is additional guidance regarding shell growth in control oysters:
Butler, P. A., and J. I. Lowe. 1978. Flowing sea water toxicity test
using oysters (Crassostrea virginicd). In: Bioassay Procedures for the
Ocean Disposal Permit Program, EPA-600/9-78-010. Environmental
Research Laboratory, U. S. Environmental Protection Agency, Gulf
Breeze, FL. pp. 25-27.
American Public Health Association, American Water Works
Association, and Water Pollution Control Federation. 1985. Toxicity
test procedures using Mollusks (Tentative). In: Standard Methods for
102
-------
the Examination of Water and Wastewater. APHA, Washington, DC.
pp. 796-797.
EEB plans to pursue a refinement of the minimum growth criteria.
However, this will entail looking at historical data and round-robin
testing before this criteria can be refined.
EPA and NACA were agreed that decisions should be made on a case-
by-case basis as stated in the letter from Paul Schuda to Daniel Barolo
and that due to apparent contradiction between experts, a resolution
meeting of experts was proposed. Prior to any meeting, however, EEB
would like to review the data from Toxikon Environmental Sciences
and Springborn Laboratories, Inc, cited above, including the
Springborn study comparing the sensitivity of two strains of
Crassostrea virginica. The data should be presented in a tabulated,
unanalyzed form to allow EER to verify the conclusions stated above.
2. Rejection Factor: There were insufficient concentration levels to result in percent
mortality to mollusc embryos/larvae of greater than 65 %, thus resulting
in a statistically less-reliable LC50.
Agency Guidance: Stephan: pp. 43, 45, and 46, nos. 11. and 12.
Assessment of
Guidance:
Avoidable:
Industry Comment:
Guidance in this area is generally adequate. However, as issues
concerning concentration/dosage levels are recurring, the Agency
handles these on a case-by-case basis.
Possibly. The registrant can choose to avoid submitting studies with
results such as this, but it is possible that even with substantial
preparation and care in study design, results such as this cannot be
avoided.
The criteria for needing greater than a 65 % effect is not specified in the
Subd-E, and as such should only be considered a recommendation, not
a required criterion. Each study should be evaluated on the merit of
the data produced, whether or not the dose-response is within some
specified range. Solubility problems with the test compound may
preclude testing at a high enough concentration to get the recommended
response range. In addition, due to the typically flat dose-response
curve in oyster studies, it is difficult to obtain both an EC50 and a
NOEL. If there is a consistent slope for the effects data, then a valid
103
-------
EPA Response:
EC50 can be calculated even when a concentration does not produce
more than 65 % effect.
It is difficult to assess the above rejection factor, since it refers to 65 %
mortality and an LC50 level. The oyster embryo/larval study is an
effect test, not a mortality test, and as such there is no requirement to
obtain "greater than 65% mortality" or determine an "LC50 level." The
SEP (1985) states that "the measured response in this test should be the
percentage of larvae dying or failing to develop complete shells."
Since the only required endpoints outlined in Subd-E are an EC50 and a
NOEL, the above statement must refer to the identification of one
endpoint that includes both dead and partially shelled organisms.
Another possible assumption regarding the above rejection criteria is
that the reference to LC50 is a typographical error, and the issue is one
of not obtaining an EC50, due to solubility problems with the test
compound. Current EPA policy allows chemicals that are poorly
soluble to be tested up to the maximum water solubility obtainable for
the given test conditions, provided that all measures to maximize water
solubility are employed. While Industry agrees that the solubility of a
compound should be maximized in order to determine an effect
concentration, there are many instances where an LC50 or EC50 is not
obtainable due to lack of solubility. Studies are often rejected due to
the inability to obtain this 50% effect level, yet it is impossible to
p'erform the study so that an effect level will be determined. EPA
states in the "Avoidability" section that indeed it may not be possible to
avoid this rejection factor.
If the first assumption of mortality versus effect is what caused the
reviewed study to be rejected, then Industry recommends that clear
guidance as to the actual assessment endpoint is needed. Since larvae
that fail to develop a complete shell in 48 hours are considered
ecologically non-viable (ASTM 1987), the only appropriate endpoint
for this study combines both organisms that are dead and those that do
not develop a full shell by the end of the test.
If the second assumption of solubility is what resulted in study
rejection, Industry recommends that as long as the solubility was
maximized in the test, the study should be accepted and the results
reported with a complete description of measures taken to maximize
solubility.
The Agency agrees with Industry that development is the valid endpoint
in .the mollusk embryo/larval study as calculated by an EC50 value.
104
-------
Resolution:
This study is not a mortality study, although dead organisms are
counted with those that do not fully develop a full shell. However,
whether or not mortality or poor development were the endpoint, the
Agency still requires that there must be at least one test concentration
producing an effect of at least 65% to be statistically valid, provided
there are no problems with chemical solubility.
EPA could not accept the deletion of the study, though they classed it
as "techniquey" but were unwilling to relax the criteria. As the NACA
proposals were unacceptable to the EPA, this issue remains unresolved
and so it is proposed that this test is re-assessed by a workgroup,
possibly in conjunction with the shell deposition issue., following a
review of historical control survival data.
3. Rejection Factor: Raw data on shell deposition were not provided.
Agency Guidance: Subd-E: p. 29, 70-4 (b)(l)(2). FIFRA-TG: p. C-57, 72-3 (c), no. 8.;
p. C-58, 72-3 (c), no. 12.; p. C-59, 72-3 (c), no. 3.; p. C-60, 72-3 (c)
nos. 8. and 9. Stephan: pp. 47 - 49, F.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
The guidance documents which address this factor appear adequate.
Yes.
Industry agrees that the raw data, defined as "enough data for the
Agency to verify calculated statistical values ..." (Subd-E: p. 29, §
70-4), should be included with the report in order for the reviewer to
be able to verify the stated EC50 and NOEL based on shell deposition.
Studies that have been rejected due to the absence of raw data should
be upgraded, however, if the appropriate data is provided to EPA. If
the reviewer or designee contacted the registrant for the needed
information prior to finalizing and issuing the DER, significant time
and rework could be avoided.
Growth measurements of individual oysters for each test concentration
need to be provided so that the reported statistical values can be
verified. Mean growth per test replicate or per test concentration is not
sufficient.
4. Rejection Factor: The test was aerated without chemical analyses of test solutions.
105
-------
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
FIFRA-TG: p. C-57, 72-3 (c), no. 7. and Stephan: p. 16 and p. 31,
no. 2. and pp. 39 - 42, no. 10. a.b.(l)(2)(3).
The guidance documents which address this factor appear adequate.
Yes. Either the test chambers should not be aerated, or if aeration is
necessary, chemical analysis of the test solution periodically throughout
the test should be performed to verify exposure.
Industry disagrees that the referenced guidance is adequate. The
FIFRA-TG (C-57, No. 7) only requires chemical analysis for
embryo/larval studies when aeration is required. However, there is a
requirement that all flow-through assays (FIFRA-TG C-58, No. 12)
have concentration analysis. Stephan says that there should never be
aeration, but that the dissolved oxygen should be between 60% and
100% for flow-through studies. No reference is made to chemical
analysis if aeration is necessary in the Stephan document, although
there is a requirement in this document to measure test concentrations
during all studies.
Industry agrees that chemical analysis should be performed in cases
where aeration of test solutions is necessary. Chemical analysis
requirements have been proposed in EPA/NACA (1992), Conducting
Acceptable Aquatic Laboratory Studies: Proposed Guidance. Current
guidelines recommend that test solutions be analyzed during flow-
through studies; this would include the situation where aeration is
needed. There does not appear to be a change needed in any guidance,
and this should not be a problem in the future.
EPA Response: EPA and NACA are in agreement.
5. Rejection Factor: There were deviations from recommended test solution characteristics.
For example, the dissolved oxygen concentration was below
recommended values of 60% to 100% saturation.
Agency Guidance: Stephan: p. 31, no. 2.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes. Steps can be taken to maintain acceptable dissolved oxygen
levels.
106
-------
Industry Comment: A key word in the above rejection factor is "recommended."
Recommended guidance helps to establish "generally recognized as
acceptable" boundaries. However, it may not be prudent to use them
in all cases to reject a study. Specific compound characteristics,
solvents, and temperature can affect the dissolved oxygen concentration
during a study. If the compound is volatile, then aeration may not be
an option, and a study may have to be run at less than optimal
dissolved oxygen concentrations.
Whether or not the rejection was avoidable depends on the extent of
reduced dissolved oxygen, both in concentration and occurrence. It is
unclear from the description in the rejection factor whether the
dissolved oxygen levels were low in all concentrations, or just at one or
a few test levels. Also, there is no clarification of what "below 60%"
means. For example, there should be no reason to reject a study based
on lower-than-recommended dissolved oxygen in the lowest dose of a
test where the NOEL is the dose above the lowest.
If there is an adequate dose-response curve obtained at the end of the
study, then it can be assumed that the deviation from recommended
dissolved oxygen concentrations did not affect the outcome of the test.
The study must be evaluated on the data produced, and on what effect
the reduced dissolved oxygen might have had, not just on a
recommended guideline.
EPA Response:
Resolution:
The purpose of a toxicity test is to test organisms at optimal conditions
to toxic agents. Lower D.O. generally results in altered metabolic
rates, which alters chemical uptake. Therefore it is possible to still get
a good dose-response curve, but the slope of the curve could be greatly
affected. A study is unacceptable if the dissolved oxygen levels fall
below acceptable limits in the control and treatment chambers. Slight
aeration is permissible to counter oxygen depletion below the
acceptable limits, provided chemical concentrations are adequately
measured. Refer to ASTM E-729, 11.2.2 for guidance on how to
properly aerate test chambers.
The Agency agreed that no rejection is necessary if corrective measures
were taken immediately. If corrective action increased the dissolved
oxygen to recommended levels by the next measurement time period or
within 24 hours, whichever is shorter, and no adverse effects on control
organisms were observed, then the study should not be rejected solely
on low dissolved oxygen. However, the study will be rejected if
corrective measures fail to bring the D.O. to acceptable levels.
107
-------
6. Rejection Factor: Dissolved oxygen levels were below 60% saturation in the last 48 hours
of exposure in a flow-through test system.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Resolution:
SEP Estuarine Mollusc Acute Shell Deposition Study: p. 6, n.E.5.
Stephan: 31, IV.E.2. ASTM E - 729: p. 390, 11 A3 and p. 393,
13.1.9.
The guidance documents which address this factor appear adequate.
Yes.
This factor is not addressed in Subd-E. If the above listed references
are required, rather than recommended, then that fact should be made
clear.
Whether or not the rejection was avoidable depends on the extent of
reduced dissolved oxygen, both in concentration and occurrence. It is
unclear from the description hi the rejection factor whether the
dissolved oxygen levels were low in all concentrations, or just at one or
a few test levels. Also, there is no clarification of what "below 60%"
means.
If there is an adequate dose-response curve obtained at the end of the
study, then it can be assumed that the deviation from recommended
dissolved oxygen concentrations did not affect the outcome of the test.
The study must be evaluated on the data produced, and on what effect
the reduced dissolved oxygen might have had, not just on a
recommended guideline.
The purpose of a toxicity test is to test organisms at optimal conditions
to toxic agents. Lower D.O. generally results in altered metabolic
rates, which alters chemical uptake. Therefore it is possible to still get
a good dose-response curve, but the slope of the curve could be greatly
affected. A study is unacceptable if the dissolved oxygen levels fall
below acceptable limits in the control and treatment chambers. Slight
aeration is permissible to counter oxygen depletion below the
acceptable limits provided chemical concentrations are adequately
measured. Refer to ASTM E-729, 11.2.2 for guidance on how to
properly aerate test chambers.
The Agency agreed that no rejection is necessary if corrective measures
were taken immediately. If corrective action increased the dissolved
oxygen to recommended levels by the next measurement tune period or
108
-------
within 24 hours, whichever is shorter, and no adverse effects on control
organisms were observed, then the study should not be rejected solely
on low dissolved oxygen. However, the study will be rejected if
corrective measures fail to bring the D.O. to acceptable levels.
Other Industry Comments on Guidelines 72-3B and E:
Embryo/Larval percent control mortality by end of test:
According to the SEP for mollusc embryo/larval tests, the organisms should be placed into
the exposure vessels within 1 hour of fertilization. This requirement has led to problems
with test interpretation and acceptability, because only one cell division occurs during this
one-hour period, making it very difficult to verify successful fertilization. Any non-viable
embryos counted in the initial inoculum are therefore apparent mortalities at the end of the
exposure. This increases the apparent mortality rate in the control vessels. Alternatively,
the ASTM guidelines for this test require a count of viable embryos 3 hours after
fertilization, after approximately 8 cell divisions. The number of organisms added to each
test vessel can then be adjusted to reflect a more reliable count of viable embryos/ml. The
SEP should be revised to account for the introduction of non-viable embryos into the test,
which can result in biased survival data.
EPA Response: Industry makes a valid point. The Agency agrees that the additional 2-hour
waiting period should not affect test results and would improve reliability of viable embryo
counts.
Requirement of 70% control survival in the embryo/larval study:
The 70% survival/development performance criteria for control embryos has resulted in the
failure of many embryo/larval studies, and the refusal by at least one contract laboratory to
perform the study. Conversations with aquaculturists indicate that successful development of
mollusc larvae 48 hours post-fertilization ranges from 10-90%, based on the inherent
variability of the performance of the adult population.
One contract laboratory recommends reducing the requirement for survival/development in
the control to 50% fully-shelled and an additional 10% underdeveloped. In order to have the
minimum number of larvae to assess at the end of the test, the study could be started with
40,000 embryos/L. With 50% developing normally, the final number of embryos assessed
would adhere to the minimum of 20,000/L indicated in the SEP. The whole issue of oyster
growth and development studies should be reexamined. Amending the guidance may not
prove beneficial if the study design is flawed.
EPA Response: The 70% survival requirement should be retained in order to assure that
quality organisms and suitable culture conditions are maintained.
109
-------
The conditions and specific methods used routinely by all aquaculturists surveyed should be
analyzed closely. Consultations should be initiated with governmental (i.e., EPA-ERL at
Gulf Breeze, Florida, or the Virginia Institute of Marine Science) and private laboratories to
establish historical survival success. The Agency should do an internal analysis to determine
what percentage of submitted studies have been identified as having survival less than 70
percent. From this data a statistically valid correction factor for control mortality will be
developed, provided that control mortality does not exceed some predetermined criteria.
Resolution: Survival of <70% in the controls can be acceptable if the overall quality of the
study (coefficients of variation (CV), etc.) are high; <70% survival alone is not a fatal flaw.
EPA defines "high test quality" as including CV's under 20; CV's greater than 40 are
unacceptable in most cases. A further evaluation of historical data and literature plus
research at an EPA-ERL is required to determine the lower limit for acceptable control
survival. EPA will address this issue on a case-by-case basis.
ACUTE LCSO ESTUARINE AND MARINE ORGANISMS (SHRIMP) (Guidelines
72-3C & F)
1. Rejection Factor: Test substance purity was not identified.
Agency Guidance: Subd-E: p. 30, 70-4 (c)(2)(i). Stephan: p. 47, no. F. 2.
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes.
Industry Comment: If the registrant cannot provide the test material characterization, the
study should be rejected. In the case that the omission of purity
information was an oversight, significant time and rework could be
avoided if the reviewer contacts the registrant for the needed
information prior to finalizing and issuing the DER.
EPA Response:
Both the Agency and Industry agree that this factor is avoidable. This
information needs to be included with the study upon its first
submission to the Agency.
2. Rejection Factor: Chemical analyses (measurements of test concentration levels) were not
performed on the test solutions.
Agency Guidance: Subd-E: p. 75, 72-3 (c)(4). Stephan: pp. 39 - 42, no.
110
-------
Assessment of
Guidance:
Avoidable:
The guidance documents which address this factor appear adequate.
Yes.
Industry Comment: The listed guidance is not adequate. Subd-E states that if chemical
analysis is conducted, the methods must be described. Stephan does
state that if the toxicant is known, the concentration must be measured;
however, Stephan is only recommended—not required—guidance. If
there is additional information on the stability of the test compound,
there may be no need to measure test concentrations.
EPA Response:
More guidance on the need for measuring test concentrations is
provided in the EPA/NACA document.
3. Rejection Factor:
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
The type of solvent used in the study, and the amount used in the
solvent control and each treatment level, were not provided..
Subd-E: p. 28, 70-3 (c)(6)(i) - (iv) and p. 30, 70-4 (c)(4)(v). FIFRA-
TG: p. C-51, 72-3 (b), no. 2.; p. C-52, 72-3 (b), no. 7.; and C-54,
72-3 (b), no. 2. Stephan: p. 31, 2nd paragraph and p. 35, no. 5.,
1st paragraph.
The guidance documents which address this factor appear appropriate.
Yes.
Industry agrees that the solvent used in a study should be identified,
and that the concentration of solvent in the solvent control, in relation
to the treatments, should be described.
EPA Response: EPA and NACA are in agreement.
4. Rejection Factor: Solubility needed to achieve LC50 was not obtained (and was
questionable, since previously submitted data appeared to indicate that
high levels of solubility could be achieved).
Agency Guidance: Stephan: p. 35, no. 5, 1st paragraph.
Ill
-------
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Resolution:
Issues of data discrepancies, relative to previously submitted data, are
handled by the Agency on a case-by-case basis. However, guidance
addressing solubility issues may need improvement.
Probably, but it may not have been avoidable even if higher solubility
was achieved in other tests, since solubility can vary significantly under
different test conditions. When the maximum solubility level is lower
than the level which will allow calculation of an LC50s the registrant
must submit an explanation justifying an exception to that requirement.
Guidance for this is being developed.
Stephan does not address solubility issues and thus is inadequate
guidance for this rejection factor. EPA should consider the guidance
outlined in EPA/NACA (1992), Conducting Acceptable Aquatic
Laboratory Studies: Proposed Guidance. Also, it is unclear from the
description of the rejection factor to what "high levels of solubility"
refer. If that statement refers to the solubility reported under "perfect"
conditions, then it is inappropriate to expect to achieve that level of
solubility under actual test conditions. In addition, large differences in
solubility have been noted between freshwater and saltwater systems.
The solubility that should be used as a measure is that which is
achievable given the test conditions.
•The Agency believes that the guidance provided in the EPA/NACA
document is adequate to address this issue.
It is important to distinguish between "perfect" versus "real" solubility.
The 1992 EPA/NACA (1992) document "Conducting Acceptable
Aquatic Laboratory Studies : Proposed Guidance" outlines measures to
take in order to maximize solubility. Solubility is defined as the
amount of chemical retained in the supernatant of a conventionally
centrifuged sample of test media. A measure is only considered to
have increased solubility if the increase is two tunes or greater.
Other Industry Comments on Guideline 72-3(c):
Requirement that the age of mysids be less than 24 hours at the beginning of a study:
Neither Subd-E nor the SEP indicates an age requirement for the mysid shrimp acute study,
though FEFRA-TG lists an age requirement of < 24 hours. The age requirement is based on
the recommendation that exposure to the compound encompass an entire molt cycle in order
to expose the organism during the most vulnerable tune period. Industry agrees that a
112
-------
uniform age is necessary to prevent cannibalism and to decrease the variability in response.
However, it should be acceptable to begin a mysid study at 96 hours, in order to allow
laboratories time to receive newly hatched mysids and acclimate them to the test conditions.
The concern surrounding the age requirement for mysids in acute toxicity tests appears to
center around a possible increase in sensitivity to toxicants immediately post-molt. S.
Lussier (EPA, Narragansett) has reported that mysids undergo 9 molts at 25 °C in the first 12
days of life with a mean intermolt period of 12.3 hours (Mysidopsis sp.: Life History and
Culture. Proceedings from a Workshop held in Gulf Breeze, Florida, 15-16 Oct 1986). This
information suggests that a study performed from 96 hours to 192 hours (8 days) would
include a full molt cycle.
Goodman et al. (1988) tested juvenile mysids from 3 age ranges (< 1, 5, and 10 days) to
determine their sensitivity to 3 toxicants at the various ages. In these tests, the 96-hour LC50
values were within a factor of two for the 3 compounds tested in the 3 age groups. The
authors concluded that age was not a significant factor in the acute toxicity of the test
compound to Mysidopsis bahia.
In a comparative study by Nor-Am Chemical company, a compound was tested with < 24-
hours-old and 4-day-old mysids. The LC50 and 95 % confidence interval for the < 24-hours-
old mysids, calculated by the moving average method, was 33.8 mg/L (26.4-46.4 mg/L).
The LC50 and 95% confidence interval for the 4-day old mysids, again calculated by the
moving average method, was 30.6 mg/L (23.9-40.9 mg/L). Again, age was not a factor in
sensitivity to the toxicant.
Studies should not be rejected only on the basis of mysid shrimp being older than 24 hours,
as long as the age of the mysids in the test is uniform within a specified range. As long as
the mysids are juveniles during the study (up to approximately 9 days of age), data generated
at the EPA Laboratory in Narragansett suggest that the mysids will be exposed to the test
compound during a full molt cycle, and thus be exposed at potentially the most sensitive
stage of their development.
EPA Response: Industry presents a valid point and the Agency suggests that all available
data, including that of S. Lussier (EPA, Narragansett, and Nor-Am Chemical Company)
cited above be analyzed to establish a reasonable age range (> 24 hours) adaptable to the
use of uncultured test organisms.
Resolution: The EPA reconfirmed the <24 hour-old requirement until further data can be
assessed to show that 24-, 48- or 96-hour-old shrimps are equally as susceptible. This data
evaluation should be conducted as a peer review and must contain values for individual
organisms. The data should be presented in a tabulated, unanalyzed form to allow REB to
verify the conclusions. This requirement is unchanged unless it is overturned by a
subsequent review of data and further NACA/EPA discussions, according to joint priorities
ofEPAandNACA.
113
-------
FISH EARLY LIFE STAGE (Guideline 72-4)
1. Rejection Factor: Significant mortality occurred at all test concentrations, making it
impossible to determine an MATC.
Agency Guidance: ASTM25: p. 862, no. 9.3; p. 864, nos. 11.1.1.1 and 11.1.1.2.
FIFRA-TG: p. C-63, 72-4, no. 15. and C-65, 72-4, no. 10.
Assessment of
Guidance:
Avoidable:
Industry Comment:
Generally, guidance on this factor is adequate because of the experience
of the laboratories doing such studies. Issues such as this are handled
on a case-by-case basis, but improvement in the guidance may be
needed.
Yes.
The FIFRA-TG document is inadequate. It does not address the issue
of "significant" mortality at all tested concentrations. A footnote in the
document indicates that a NOEL is "supplemental and may not be
required for every study."
The ASTM guidance appears to be adequate in that it suggests that
regression analysis may be used as an alternative to the hypothesis
testing approach. It also discusses the issue of biological versus
statistical significance. However, this document does not address what
level of effect is of ecological significance.
Fish Early Life Stage studies are designed to evaluate the effects of a
test substance on the survival and growth of embryos, larvae, and
juvenile fish. Industry agrees that significant mortality at all test
concentrations may prevent the determination of an MATC. A
distinction must be made between statistical significance within the test
and the normal or historical control variability (e.g., 20%). Based on
the study design and the power of the statistical tests employed, a small
difference in survival may be statistically significant but still within the
acceptable range of mortality typical for control organisms. It is
important to consider when and under what circumstances the mortality
occurred. A careful review of the magnitude of the effect should be
undertaken, considering the dose response and the other endpoints
25 Various versions are available. In this case, Vol. 11.04, 1991, which was readily
available, was used and will be used for this study and the Aquatic Invertebrate Life-Cycle test.
114
-------
EPA Response:
Resolution:
evaluated in the study. The results need to be evaluated in relation to
the risk assessment.
Studies such as this should be handled on a case-by-case basis. EPA
should be flexible in their study reviews. Industry recommends that the
magnitude of the effect on the fish in the lowest concentration group be
reviewed carefully in respect to (1) when the mortality occurred,
(2) statistical significance, (3) the overall concentration-effect
relationship, and (4) relevance of the NOEL to the estimated
environmental exposure level. Definition of a NOEL from regression
analysis (e.g., LC20) should be possible from studies with significant
mortality in the lowest test concentration.
NOEL's are not required for acute toxicity studies, but are required in
chronic toxicity studies for regulatory purposes. In order for a NOEL
to be valid, there must be at least one concentration level with effects
not significantly greater than the appropriate control, and a second
concentration delineating an acceptable LOEL. The FIFRA-TG
guidance states that an MATC value must be reported (C-62) and a
NOEL determined (C-65). It was an inadvertent error that a footnote
indicated that the NOEL is a supplemental data requirement.
Studies will not automatically be rejected if a NOEL is not
experimentally achieved. Registrants can submit a calculated NOEL
for consideration. In this discussion, EPA also emphasized the general
.distinction between studies "not fully acceptable" versus "needing to be
repeated." Studies that are not core (fully acceptable) may still provide
enough information for a risk assessment and therefore need not be
repeated. The need for a new study will be considered on a case-by-
case basis.
2. Rejection Factor: Raw data were not submitted to allow the Agency to verify the MATC
and NOEL.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Subd-E: p. 29, no. 70-4 (b)(l)(2) and p. 78, no. 72-4 (c)(l) - (11).
FIFRA-TG: pp. C-61 - C-65, 72-4 (all sections). ASTM: pp. 867
868, no. 15.
The guidance documents which address this factor appear adequate.
Yes.
115
-------
Industry Comment:
EPA Response:
Resolution:
Industry agrees that raw data, defined as "enough data for the Agency
to verify calculated statistical values," should be included with the
report in order for the reviewer to be able to verify the stated MATC
and NOEL values. Studies that have been rejected due to the absence
of raw data should be upgraded if the appropriate data are provided to
EPA.
For the fish early life stage study, data in the following manner need to
be submitted in the final report: number of embryos hatched, time to
hatch, number of dead embryos and dead larvae are reported daily for
each incubation cup; time to swim-up, number of dead juveniles are
reported daily for each test chamber; and growth measurements are
reported for each live individual fish for each test concentration and
control at the end of the study. This will provide sufficient data for the
reviewer to verify the reported MATC, NOEL, and LOEL values for
each biological endpoint.
Agreement was reached that data needed to calculate statistics must be
submitted. EPA prefers to call this necessary data "tabulated and
unanalyzed data."
3. Rejection Factor: The survival rate in the control group was substantially lower than
recommended (70%).
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
Subd-E: p. 26, 70-3 (b)(2) and (3). (For background information:
Stephan: pp. 30 - 31, no. E.I. and p. 39, no. 9., last paragraph).
ASTM: p. 867, no. 13.1.8.
Generally, guidance on this factor is adequate because of the experience
of the laboratories doing such studies. However, the guidance may
need improvement.
Yes.
Industry finds the guidance inadequate. Subd-E: p. 26, §70.3(b)(2) and
(3) does not address control survival. The Stephen reference does not
address fish early life stage tests. The ASTM document has variable
criteria depending on species, and it differs from the SEP. The Agency
should provide clear guidance as to how to calculate control survival.
Industry needs to know how many studies were rejected for poor
control survival and what the survival was in each case. Industry
116
-------
EPA Response:
Resolution:
agrees that control survival should not be "substantially lower than
recommended (70%)." The term "substantially lower" is open to
interpretation. The calculation of control survival is handled differently
among and within various documents. For example, ASTM
recommends that control survival for fathead minnows be > 70%, and
be calculated for the period of 48 hours to 32 days. In the same
•document, survival must be S: 70% for rainbow trout, as calculated
from the time of thinning to test termination. The SEP states that
control survival at the end of the study must be 80% overall,, and not
less than 70% in any replicate.
If control mortality occurs later in a study (e.g., due to the malfunction
of test equipment, outbreak of disease, etc.) resulting in overall control
survival that is < 70%, the study should be acceptable. Repeating the
study will not provide additional useful information, is wasteful of
resources, and will impede the reregistration process.
Control survival below 60% should be considered to be "substantially
lower" than recommended. Where survival is questionable, the issue
should be handled on a case-by-case basis, with consideration of the
time point within the study where survival dropped below the
recommended criterion.
The Agency acknowledges differences among criteria as prescribed by
the Agency SEP and various protocols. The Agency should consolidate
appropriate criteria from the best guidance documents available into one
SEP. Mortality within any replicate exceeding 30% would justify
rejection of a study. The Agency agrees with Industry that the term
"substantially lower" is vague. Stating that mortality was less than the
criteria of 70% is adequate. The Agency will reconsider its study
rejection levels to align them with the more conservative vis-a-vis more
protective criteria prescribed by ASTM and Stephen (pp. 30-31, No.
E.I and p. 39, No. 9, last paragraph).
EPA and NACA agreed that a flexible approach be used in the analysis
of control survival. The difference between EPA recommendations
(80% overall, 70% per replicate) and ASTM (70% overall) was noted.
4. Rejection Factor: Measured test concentrations showed highly erratic results throughout
the test, raising questions about whether test equipment (i.e., diluter
system) was functioning properly.
117
-------
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Resolution:
Subd-E: pp. 30-31, 70-4 (c)(4)(i) - (viii); p. 78, 72-4 (b)(4).
FEFRA-TG: p. C-62, 72-4, no. 12.; p. C-63, 72-4, no. 15.; and C-65,
72-4, no. 10. ASTM: p. 867, nos. 13.1.14 and 13.1.15.
Generally, guidance on this factor is adequate because of the experience
of the laboratories doing such studies. Further, unexplained variations
such as these are handled on a case-by-case.
Yes.
Industry finds the guidance inadequate. Subd-E: p. 30-31,
§70.4(c)(4)(i) to (viii); and p. 78, §72-4(b)(4) does not address erratic
analytical results. The pages and sections of the FIFRA-TG document
do not give clear guidance on this issue. ASTM offers the clearest
guidance, but it differs from EPA/NACA (1992) Conducting
Acceptable Aquatic Laboratory Studies: Proposed Guidance.
Industry agrees that highly erratic results in measured test
concentrations may indicate problems with the dosing apparatus.
However, if there is no apparent problem with the dosing apparatus,
and the test substance is highly unstable in aqueous solution, is poorly
soluble in water, adsorbs to test apparatus, or the tested concentrations
are near the detection limit of the analytical method, then the erratic
results are probably unavoidable.
The Agency agrees that erratic results need to be handled on a case-by-
case basis. The guidance provided in the EPA/NACA document should
be adequate to address this issue provided the erratic results are caused
by inherent properties of the test chemical. However, if inconsistent
analytical results are due to malfunction of the test (diluter) system, the
study will likely be rejected.
It was agreed that long-term studies should not be automatically
rejected for equipment failures. The registrant must be able to
demonstrate the duration of the exposure fluctuation and that
appropriate corrective actions were taken (e.g., repairs and sampling
within 24 hours).
5. Rejection Factor: The raw data for the acclimation period and embryo hatchability,
survival, and growth of the larvae were not provided. Therefore, the
reported MATC and NOEL values could not be verified.
118
-------
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
Subd-E: pp. 29 - 32, no. 70-4 (b) (1) and (2) and (c) (3), (5) - (7) and
p. 78, no. 72-4 (c)(l) - (11). FIFRA-TG: 72-4, pp. C-61 - C-65 (all
sections). ASTM E -1241: pp. 865 - 866, no. 11.8 and pp. 867 - 868,
no. 15. SEP (Fish ELS): p. 6, H.A.7 and p. 9, H.D.
The guidance documents which address this factor appear adequate.
Yes.
Although industry agrees that the guidance appears adequate, a clearer
distinction should be made between the definition of raw data as used
herein and that used in the Good Laboratory Practice Standards.
Raw data on acclimation is not pertinent to calculation of an MATC or
NOEL. Industry agrees that raw data, defined as "enough data for the
Agency to verify calculated statistical values," should be included with
the report in order for the reviewer to be able to verify the stated
MATC and NOEL values. Studies that have been rejected due to the
absence of raw data should be upgraded if the appropriate data is
provided to the Agency.
The Agency agrees that data for the acclimation period are not
necessary for calculation of the MATC value. However, the data on
the conditions during the acclimation period are needed for verification
of maintenance of the test organisms. This data can take the form of
ranges of data points confirming the condition and maintenance of the
organisms. For the fish early life stage study, data in the following
manner need to be submitted in the final report: number of embryos
hatched, time to hatch, number of dead embryos and dead larvae are
reported daily for each incubation cup; time to swim-up, number of
dead juveniles are reported daily for each test chamber; and growth
measurements are reported for each live individual fish for each test
concentration and control. This will provide sufficient data for the
reviewer to verify the reported MATC, NOEL, and LOEL values for
each biological endpoint.
Other Industry Comments on Guideline 72-4
Individual Dry Weights Were Not Measured:
Dry weight is not a requirement in the guidance documents cited; measurement of wet
weight is acceptable. In cases where the fish appear edematous, dry weight is more
119
-------
appropriate. In the study design for the fish ELS test, the experimental unit is the replicate
chamber. Industry believes that the use of individual fish as independent experimental units
in the evaluation of growth parameters constitutes pseudo-replication and is not appropriate.
Assuming the fish were not edematous, measurement of wet weight rather than dry weight
should not be grounds for study rejection. Use of mean growth values of all fish within a
replicate in the statistical analysis is appropriate if the replicate test chambers are the
experimental units. Studies should not be rejected when the data is analyzed in this manner.
EPA Response: The Agency agrees that wet weight measurements are acceptable. In tests
where the fish are edematous, dry weight measurements are valid. Individual measurements
are important to determine "within" variances and therefore, are necessary. EEB would like
to consult with statisticians on the use of individual versus mean replicate values for growth.
Only Two True Replicates Were Used for Embryo Exposure:
It is clear that Agency guidance documents recommend the use of four replicates in the
fish ELS test. However, ASTM allows the use of two replicates. It is Industry's
understanding that there has been some inconsistency on the part of EPA in respect to
rejection of studies using two replicates. Some studies using two replicates have been
accepted by EPA. The use of two replicates should not be the sole cause for rejection of a
study.
EPA Response: The Agency requires four replicates for embryo exposure. However if
Industry can provide scientifically valid reasons for using two rather than four replicates, the
Agency will, on a case-by-case basis, judge whether a study should or should not be
rejected.
Statistical Evaluation of Growth Data (Length and Weight) Used Mean Values Rather
Than Individual Measurements:
In the study design for the fish ELS test, the experimental unit is the replicate chamber.
Industry believes that the use of individual fish in the evaluation of growth parameters
constitutes pseudo-replication and is not appropriate. Standardization of statistical procedures
is required. Use of mean growth values of all fish within a replicate hi the statistical analysis
is appropriate if the replicate test chambers are the experimental units. Studies should not be
rejected when the data is analyzed in this manner.
EPA Response: The Agency will accept mean growth data (Replicate = Experimental Unit)
for statistical analysis with the following conditions: 1) individual length and weight data be
provided to the Agency electronically and 2) time-weighted chemical averages be used for the
analyses.
120
-------
Resolution: Individual measurements are important to determine "within" variances and,
therefore, are necessary. F,RR would like to consult with statisticians on the use of
individual versus mean replicate values for growth. Industry can do their calculations using
mean data, but the Agency will continue to validate the results using the individual
measurements until or unless the statistical consultation indicates otherwise.
Industry Comment: NACA requests the opportunity to be party to such consultation. NACA
member companies have evaluated the issue statistically and have statistical expertise that
would be useful to the discussion. The evaluation should result in a written document
outlining the rationale and justification for the statistical procedures to be used.
Dissolved Oxygen Levels Were Below Recommended Levels:
Industry agrees that dissolved oxygen should not be substantially lower than
recommended; however, the term "substantially lower" is open to interpretation, and
recommended dissolved oxygen levels differ among the various guidance documents. ASTM
recommends maintaining dissolved oxygen at > 60% of saturation, based on a time-weighted
average. ASTM also permits aeration of the test chambers. The SEP infers that dissolved
oxygen be maintained at > 75% of saturation and does not allow aeration of test chambers.
Studies such as this should be handled on a case-by-case basis. EPA should be flexible in
their study reviews. Rejection should only occur if the dissolved oxygen concentrations are
below a time weighted average of 60%. Industry recommends that low dissolved oxygen
occurrences be reviewed carefully in respect to when and why they occurred, and the overall
effect on the outcome of the study.
EPA Response: The purpose of a toxicity test is to test organisms at optimal conditions to
toxic agents. A study is unacceptable if the dissolved oxygen levels fall below acceptable
limits in the control and treatment chambers. Slight aeration is permissible to counter
oxygen depletion below the acceptable limits, provided chemical concentrations are
adequately measured. Refer to ASTM E-1241, 11.2 for guidance on how to properly aerate
test chambers.
Resolution: The Agency agreed that rejection is not necessary if effective corrective action is
taken immediately (i.e., the corrective action raises the D.O. to recommended levels by the
next measurement time or within 24 hours, whichever is shorter, and there is no adverse
effect noted in control organisms). However, the study will be rejected if corrective
measures fail to bring the D.O. to acceptable levels.
Solvent Appeared to Affect Growth;
Use of a solvent is routine in aquatic toxicity tests of poorly soluble compounds. EPA's
SEP and the ASTM guidelines contain recommendations for the solvents that can be used in
121
-------
a test and the maximum concentration permitted. These solvents are generally regarded as
non-toxic at the recommended concentrations. If there was an apparent effect on growth
with the solvent, then the solvent control is used to compare with the other treatments.
Studies where a solvent is used to solubilize the test compound should not be rejected if there
appears to be an effect due to the solvent. In these cases, data evaluation should be based on
comparisons with the solvent control.
EPA Response: The Agency will judge these issues on a case-by-case basis. If an
acceptable solvent causes unacceptable control mortality or affects other biological endpoints,
it should be determined whether solvent levels (< 10%) are adequate or if a replacement
solvent is needed. If the first two alternatives are not feasible, then the appropriateness of
comparing the data to the solvent control will be evaluated.
Solvent Concentrations Not Equal Across Treatments:
There is no absolute requirement that solvent concentrations be equal across all
treatments. EPA's SEP allows differing concentrations of solvent, as long as the solvent
control contains at least as much solvent as in any of the other test chambers. Standard
Mount and Brungs diluter systems that are used at many laboratories preclude the use of
equivalent solvent concentrations. Studies performed using a cosolvent should not be
rejected if solvent concentration is not equal across test chambers. As long as a solvent
control is tested that contains solvent at a concentration equal to the highest level in any test
chamber, the study should be acceptable.
EPA Response: The Agency agrees with Industry. The only additional comment is that the
concentration of the solvent should not exceed 0.1 ml/L in any test concentration in a flow-
through system.
Length Measurements Were Not Provided:
Neither EPA's SEP nor the ASTM guidance states that length must be measured. The
HFRA-TG document does list length as a growth parameter. Guidance documents are
ambiguous on the issue of length measurements, and the need for measurement of length
should be clarified.
EPA Response: Length measurements are a vital component of growth indices. Therefore,
individual standard length measurements are required on the individual fish at the end of the
study. The SEP, section H.D., states that length is one of the continuous measurements that
needs to be statistically analyzed.
AQUATIC INVERTEBRATE LIFE CYCLE (Guideline 72-4, Daphnia)
122
-------
1. Rejection Factor: Raw data were not submitted to allow the Agency to verify the MATC.
Agency Guidance: Subd-E: p. 29, 70-4 (b)(l)(2) and p. 78, no. 72-4 (c)(l) - (11).
FIFRA-TG: p. C-67, no. 9. ASTM: p. 792, 13.1.9 and pp. 792 -
793, no. 15! SEP Daphnia Life Cycle: p.6, D.
Assessment of
Guidance:
Avoidable:
Industry Comment:
The guidance documents which address this factor appear adequate.
Yes.
Industry disagrees that guidance on this factor is adequate. Below is a
list of the referenced EPA guidance with relevant text. Industry feels
that guidance documents with asterisks clearly establish that some form
of raw data is required; however, since "raw" data can be defined in a
variety of ways, the guidance can cause confusion.
• Subd-E: p. 29, §70-4(b)(l)(2). Both sections relevant, especially (2).
Each test report should include "... enough data for the Agency to
verify calculated statistical values."
• Subd-E: p. 78, §72-4(c)(l) to (11). All sections are not pertinent, only
#4 ". . . include mortality data."
ASTM: pp 792-793, No. 15. All sections are not pertinent, only
15.1.11 "... data in sufficient detail to allow independent statistical
analyses ..."
SEP, Daphnia Life Cycle (not included in original analysis): p. 6, D.
Requires that "results (of the study) must be accompanied by copies of
the original (raw) data ..."
A second source of confusion is the inconsistency with which EPA
invokes various documents as guidance. For this rejection factor, for
example, EPA has listed ASTM as guidance, but not EPA's own SEP.
In general, industry agrees with EPA that this was avoidable.
However, clear delineation of appropriate guidance might help
eliminate future problems. Requiring submission of raw data is not
acceptable guidance, without a definition of "raw."
123
-------
EPA Response:
Resolution:
In addition, industry believes that:
1. Subd-E is the most appropriate guidance for this subject, since it
establishes that only enough data needs to be submitted to allow
independent verification of endpoints. i
2. EPA should contact the registrant and request data needed, rather
than rejecting the study because the data were not submitted. This
would prevent needless rejection of studies.
Failure to initially include the SEP as guidance was an oversight. The
citation is now included in the guidance for this rejection factor. For
the Daphnia life-cycle study, data in the following manner need to be
submitted in the final report: survival, mortality, and reproduction data
for each replicate chamber for each day observed; lengths and or dry
weights for each live individual daphnid for each test concentration and
control at the end of the study. This will provide sufficient data for the
reviewer to verify the reported MATC, NOEL, and LOEL values for
each biological endpoint.
Industry and EPA are in agreement with what needs to be submitted.
2. Rejection Factor: The study was unable to produce NOEL values for reproduction and
growth; thus, correct MATC values cannot be established for
reproduction and growth.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
ASTM: p. 784, no. 4. p. 792, no. 13.1.9 and 15.1.11. FIFRA-TG:
P. 67, 72-4(b), no. 9.; p. C-68, 72-4(b), no. 13. and p. C-70, 72-4(b),
no. 9. SEP Daphnia Life Cycle: p.7, DDL A. Subd-E: p. 29, no. 70-
4 (b)(l) and (2) and p. 78 no. 72-4 (c)(l) to (11).
Generally, guidance on this factor is adequate because of the experience
of the laboratories doing such studies. Issues such as this are handled
on a case-by-case basis, but improvement in the guidance may be
needed.
Yes.
Industry agrees that guidance on this factor can be improved. Below is
a list of the referenced EPA guidance with relevant text. Asterisks
indicate guidance documents that are particularly clear in establishing
the NOEL requirement. EPA's selection of guidance documents to
124
-------
include in this report produced some confusion. For example, ASTM
is listed, but not EPA's Subd-E Guidelines or SEP.
• ASTM, p. 784, no 4. Section does not require or mention MATC or
NOEL (general description of study).
• ASTM, p. 792, No. 13.1.9. Section does not require or mention
MATC or NOEL (discusses measurement endpoints for the study).
* • FIFRA-TG, §72-4(b), p. 67, No. 9. Requires ".
reported in ppm."
* • FIFRA-TG, p. C-68, §72-4(b), No. 13. Requires ".
effect level and reproductive effects defined."
MATC value
no observed-
• FIFRA-TG, p. C-70, §72-4(b), No. 9. Requires MATC and NOEL.
• Subd-E: p. 32, §70-4(c)(7) (not included in original analysis). General
requirements for data analysis. §70-4(c)(7)(iii) specifically requires
determination of NOEL. Problems: (1) These are general guidelines
that apply to acute as well as chronic studies. Consequently, this
implies that a NOEL is required for acute studies. (2) These
requirements do not require an adverse effect level.
• Subd-E (not included in original analysis): p. 78, §72-4(c)(l) to (11).
General requirements for reporting; (c)(3) requires no-effect level.
• SEP, Daphnia Life Cycle (not included in original analysis): p. 7,
HI. A. Requires verification of proper calculation of MATC.
In principle, Industry agrees that a NOEL and an adverse effect level
are reasonable endpoints for this test. However, Industry feels that
because of the complexity of these tests, it is important to remain
flexible in interpreting results. The following three factors can
influence NOEL determination:
First, selection of test concentrations that will yield a NOEL and a
reproductive effect level (Required in FEFRA-TG, §72-4(b), p. C-68,
No. 13) can be difficult. The selection problem is exacerbated by the
routine spacing of test concentrations by 2X of the lower concentration.
It would be helpful if EPA provided clear guidance that it is acceptable
to increase the size of the interval.
125
-------
EPA Response:
Second, uncertainty exists regarding appropriate control groups to be
used in statistical comparisons in determination of the NOEL. ASTM-
1193, No. 9.2.4.3 suggests that if a solvent effect exists, then all
comparisons should be made to the solvent control. In the absence of a
solvent effect, both controls should be used in meeting requirements
(Industry assumes this means pooling controls). It is important that
guidance is clear on this issue, since these factors can influence NOEL
identification.
Third, absence of a NOEL for a particular study should result in
rejection only after considering the type of variable affected, the
magnitude of the. effect, the existence of a dose response relationship,
the importance to EPA's risk assessment, and the possibility of
estimating a theoretical NOEL.
Industry therefore suggests that EPA:
1. Cite only relevant guidance documents, as indicated with asterisks
above.
2. Allow flexibility in establishing test concentrations for chronic tests.
Do not require an adverse effect level unless it is necessary to
complete the risk assessment. Allow test concentrations to be
separated by levels greater than 2X the lower concentration.
3. Clearly indicate the type of statistical comparisons to be made (e.g.,
compare treatment levels to control, solvent control, or pooled
control), since the NOEL is statistically derived by comparing to a
control group, and since most aquatic studies have both a negative
and a solvent control.
4. Not categorically reject studies if a NOEL is not identified. Weight
of evidence based on such factors as magnitude of effect and
importance to the risk assessment should be considered.
Failure to initially include the SEP and Subd-E as guidance was an
oversight. The citations are now included in the guidance for this
rejection factor. The Agency agrees to provide better guidance
document citations, advice for test concentration selection, and test
concentration intervals. The Agency can require data beyond the
ASTM protocol. See the Rapid Feedback SOP.
When solvents are used, the solvent control should be used for data
comparison. The negative control is designed to discern potential
126
-------
effects of the solvent. In cases where there is no control mortality and
no significant difference between the negative and positive controls, use
of a pooled control may be permissible.
Resolution:
NOELs are not required for acute toxicity studies, but are required in
chronic toxicity studies for regulatory purposes. In order for a NOEL
to be valid, there must be at least one concentration level with effects
not significantly greater than the appropriate control, and a second
concentration delineating an acceptable LOEL.
Studies will not automatically be rejected if a NOEL is not
experimentally achieved. Registrants can submit a calculated NOEL
for consideration.
There needs to be some flexibility in dose selection and spacing. The
Agency needs to develop a scientifically sound method for spacing
concentrations in addition to good general guidance for this chronic
study. The Agency agreed to provide guidance (advice) through the
rapid feedback process. The SOP developed for obtaining rapid
feedback explains the documentation needed for inclusion in the final
report for EPA review.
3. Rejection Factor: The survival rate in the control group (at one test level) was
substantially lower than recommended.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
Subd-E: p. 26, 70-3 (b)(2) and (3). (For background information:
Stephan: pp. 30 - 31, no. E.I. and p. 39, no. 9., last paragraph).
ASTM: p. 792, no. 13.1.10. SEP Daphnia Life Cycle: p.4, H.A.7.
FIFRA-TG: 72-4(b), p. C-67, #12.
Generally, guidance on this factor is adequate because of the experience
of the laboratories doing such studies. Issues such as this are handled
on a case-by-case basis, but improvement in the guidance may be
needed.
Yes.
Industry agrees that improvement in guidance is needed. Below is a
list of the referenced EPA guidance with relevant text. The guidance
that Industry felt was particularly relevant (indicated below with
asterisks) was not included as guidance in EPA's original rejection rate
127
-------
analysis. These guidance documents were cited in other parts of this
document, so it is unclear why they were not included here. Because
of the inconsistencies in designating guidelines, it is not clear what the
acceptability criteria is. For example, the ASTM guidance says
> 20% mortality is unacceptable, whereas EPA's own guidance
indicates > 30% mortality is unacceptable. None of the guidance
indicates whether it is per replicate or the average of all replicates.
• Subd-E: p. 26, §70-3(b)(2) and (3). Passage addresses inclusion of
solvent control (2) and maintenance of control groups similar to other
test levels. Stephan: pp. 30-31, No. E.I and p. 39, No. 9., last
paragraph. This reference specifies 10% mortality in static and 5% in
flow-through as being unacceptable.
• ASTM: p. 792, No. 13.1.10. Test rejected if controls have > 20%
mortality, < 60 young per adult (13.1.11), or ephippia were produced
(13.1.12).
• FIFRA-TG (not included in analysis): 72-4(b), p. C-67, #12. Success
of controls—the study is rejected if >: 30% total mortality, ^ 40
young were produced, or if ephippia are produced.
• SEP Daphnia (not included in analysis). Provides identical guidance
as in FIFRA-TG
Also confusing is EPA's assessment that "guidance on this factor is
adequate because of the experience of the laboratories doing such
studies."
The rejection factor statement referred to "one test level in the control
group." Industry assumed that this refers to the solvent and negative
levels of the control groups.
Industry suggests that EPA:
1. Clarify why the study was rejected (e.g., What is one test level of
the control group? What does "substantially" mean?).
2. Resolve conflicting acceptability criteria in guidance documents.
Industry suggests following the 30% control mortality criteria
established in guidance documents with asterisks. The acceptability
criteria is based on the average of replicates where there are more
than one animal per replicate.
128
-------
EPA Response:
Resolution:
3. Clarify the statement "this factor is adequate because of the
experience of the laboratories doing such studies."
Failure to initially include the SEP and FIFRA-TG as guidance was an
oversight. The citations are now included in the guidance for this
rejection factor. The Agency acknowledges differences among criteria
as prescribed by the Agency SEP and various protocols. The Agency
also acknowledges that the wording, "one test level" is inappropriate.
It should have read "one control group." Nevertheless, mortality
within any replicate exceeding 30% would justify rejection of a study.
The Agency will reconsider its study rejection levels to align them with
the more conservative vis-a-vis more protective criteria prescribed by
ASTM and Stephen (pp. 30-31, Nol E.I and p. 39, No. 9, last
paragraph).
EPA and NACA agreed that a flexible approach be used in the analysis
of control survival. The difference between EPA recommendations
(80% overall, 70% per replicate) and ASTM (70% overall) was noted.
4. Rejection Factor: Growth parameters (length and weight) were not measured
quantitatively on adults at the end of the study.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Subd-E: p. 29, no. 70-4 (b)(l)(2) and p. 78, no. 72-4 (c)(l) - (11).
FIFRA-TG: 72-4(b) p. C-70, no. 9. ASTM E - 1193: p. 790 - 791,
no. 11.10, pp. 792 - 793, no. 13, 14 and 15. SEP Daphnia Life
Cycle: p. 5, H.A.8 and p. 6, H.D.
Guidance in this area needs to be improved. The EPA publications
indicate that length measurements are adequate to analyze effects on
growth. However, in 1987 ASTM determined that weight is a better
growth indicator than length. An amendment to the SEP was drafted in
September 1990 and indicated that dry weight and length measurements
are needed. However, these changes need to be incorporated formally
into EPA documents, particularly the SEP.
No.
Industry Comment:
We agree with EPA that guidance in this area needs to be improved.
Below is a list of the referenced EPA guidance with relevant text. We
feel that the guidance indicated with the asterisks are most appropriate.
Industry could not check the accuracy of the statement that "ASTM, in
1987, determined that (dry) weight is a better growth indicator than
129
-------
length," since no reference was included in the ASTM document.
Industry does not believe that EPA should reject studies solely on the
basis that dry weight measurements were not made, until it is clearly
mandated in the guidance documents.
Subd-E: p. 29, §70-4(b)(l) and (2); p. 78, §72-4(c)(l) to (11). General
data requirements; no reference to Daphnia or growth.
FBFRA-TG: 72-4(b) p. G-70, No. 9. Requires length of first
generation organisms.
ASTM E-1193: p. 790-791, No. 11.10. Requires dry weight
determination of individual Daphnia or length of individual Daphnia, if
it has been determined length and dry weight are correlated.
ASTM E-1193: pp. 792-793, Nos. 13, 14, and 15. No. 13.1.9
requires data on growth. No. 14.1 requires dry weight or length. No.
15.1.11 requires data sufficient to allow recalculation of growth
MATC.
i ...
SEP, Daphnia Life Cycle: p. 5, n.A.8 requires length of first
generation. P. 6, n.D requires statistical analysis of length.
Amendment to Ecological Effects SEP (not included in original
analysis): Section (8). Changes SEP to say "include data on ... the
dry weight of first generation daphnids at the end of the test."
The issue relative to this rejection factor is principally related to the
additional requirement of measuring dry weight, as well as length, as
an estimate of growth. Industry does not feel that addition of this
endpoint is warranted, for the following reasons:
First, Industry does not agree with EPA's statement on assessment of
guidance that ASTM determined that dry weight was a better indicator
of growth rate. ASTM guidance states that "dry weight of each
individual first-generation daphnid . . . must be determined to the
nearest microgram, except that length may be determined in place of
dry weight if a correlation has been shown between length and dry
weight." Kooijman and Metz (1984) noted that length of daphnids is
proportional (correlated) to the third root of wet weight. Gurney, et al.
(1990), point out that even though length and weight are correlated,
carapace length will tend to fluctuate less than total body mass. This is
because carapace length remains constant during the molting cycle but
weight can fluctuate. Accordingly, body mass would tend to be more
130
-------
EPA Response:
variable in a changing environment compared to length. A less
variable parameter such as length could be a "better" indicator of
growth.
Second, growth and reproduction of individual daphnids is closely
related (Kooijman and Metz, 1984). Since the principal endpoint of
this test is reproduction, it is important that balance be maintained in
the power of statistical comparisons for growth and reproduction.
Industry agrees that a measure of growth is necessary, but maintains
that reproduction should be the focus of the study and that length serves
as an adequate estimate of growth.
Gurney, W.S.C., E. McCauley, R.M. Nisbet, and W.W. Murdoch
(1990) The physiological ecology ofDaphnia: A dynamic model of
growth and reproduction. Ecology 71:716-732.
Kooijman, S.A.L.M. and J.A.J. Metz (1984) On the dynamics of
chemically stressed populations: The deduction of population
consequences from effects on individuals. Ecotox. Environ. Safety
8:254-274.
Industry suggests that EPA:
1. Clearly indicate whether dry weight or length were not measured,.
2. Do a rigorous statistical power analysis on any new endpoints added
to this test. Since this is principally a test for determining
reproductive effects, it is important that the statistical power for
growth measurements do not deviate greatly from the power
inherent to the reproduction test. Moreover, Industry believes that
one measure of growth should be acceptable.
3. Compile new guidance specific for this test system, since many
Daphnia Life-Cycle studies are flow-through systems. Specific
attention should be focused on the types of endpoints that should be
included and statistical power.
The Agency recommends that both length and dry weight be
determined, however no study reporting length measurements only will
be rejected. The Agency concurs that appropriate guidance should be
provided for the flow-through test design, and specific guidance
provided on preferred end-points.
131
-------
Resolution: Industry and EPA agree that, while both length and weight are
preferable, only length is required until the Agency provides specific
guidance indicating otherwise.
There was agreement that guidance is needed for the flow-through
Daphnia Life-Cycle study. EEB is working toward developing a
protocol and SEP for this study. ASTM has a draft protocol for the
flow-through design which is still in the process of comment and
revision. It cannot be recommended as a general guidance document
by the Agency until it is finalized.
5. Rejection Factor: Failure to obtain a NOEL. Thus, correct MATC values for
reproduction and growth cannot be established.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Industry Comment:
ASTM: p. 784, no. 4. p. 792, no. 13.1.9 and 15.1.11. FIFRA-TG:
P. 67, 72-4(b), no. 9.; p. C-68, 72-4(b), no. 13. and p. C-70, 72-4(b),
no. 9. SEP Daphnia Life Cycle: p.7, m.A. Subd-E: p. 29, no. 70-
4 (b)(l) and (2) and p. 78 no. 72-4 (c)(l) to (11).
Generally, the guidance in this area is adequate.
Yes.
Industry suggests that guidance on this factor can be improved. Below
is a list of the referenced EPA guidance with relevant text. Guidance
documents indicated with asterisks are particularly clear in establishing
the NOEL requirement. EPA's selection of guidance documents to
include for this factor caused some confusion. For example, the
guidance documents listed for this factor are not the same as those
listed for the very similar factor no. 2 above.
Subd-E: p. 29, §70-4(b)(l) and (2). General data requirements; no
mention of NOEL or MATC.
* • Subd-E: p. 78, §72-4(c)(l) to (11). General requirements for
reporting; (c)(3) requires no effect level.
* • FIFRA-TG: p. C-67, 72-4(b), No. 9. Requires MATC value to be
reported in ppm.
* • Daphnia Life Cycle: p. 7, ffl. A. Requires verification of proper
calculation of MATC.
132
-------
EPA Response:
In principle, Industry agrees that a NOEL and an adverse effect level
are reasonable endpoints for this test. However, Industry feels that
because of the complexity of these tests, it is important to remain
flexible in interpreting results. Three of the factors that can affect
NOEL determination are described below:
First, selection of test concentrations that will yield a NOEL and a
reproductive effect level (Required in FIFRA-TG, §72-4(b), p. C-68,
No. 13) can be difficult. The selection problem is exacerbated by the
routine spacing of test concentrations by 2X of the lower concentration.
It would be helpful if the Agency provided clear guidance that it is
acceptable to increase the size of the interval.
Second, uncertainty exists regarding appropriate control groups to be
used in statistical comparisons in determination of the NOEL. ASTM-
1193, No. 9.2.4.3 suggests that if a solvent effect exists, then all
comparisons should be made to the solvent control. In the absence of a
solvent effect, both controls should be used in meeting requirements
(industry assumes this means pooling controls). It is important that
guidance is clear on this issue, since these factors can influence NOEL
identification.
Third, absence of a NOEL for a particular study should result in
rejection only after considering the type of variable affected, the
magnitude of the effect, the existence of a dose response relationship,
the importance to EPA's risk assessment, and the possibility of
estimating a theoretical NOEL.
The guidance for this rejection factor and factor No. 2 for this
guideline study have been corrected and are now identical. The
Agency agrees to provide better guidance document citations, guidance
for test concentration selection, and test concentration intervals. The
Agency can require data beyond the ASTM protocol.
When solvents are used, the solvent control should be used for data
comparison. The negative control is designed to discern potential
effects of the solvent. In cases where there is no control mortality and
no significant difference between the negative and positive controls, use
of a pooled control may be permissible.
NOELs are not required for acute toxicity studies, but are required in
chronic toxicity studies for regulatory purposes. In order for a NOEL
to be valid, there must be at least one concentration level with effects
133
-------
Resolution:
not significantly greater than the appropriate control, and a second
concentration delineating an acceptable LOEL.
Studies will not automatically be rejected if a NOEL is not
experimentally achieved. Registrants can submit a calculated NOEL
for consideration.
There needs to be some flexibility in dose selection and spacing. The
Agency needs to develop a scientifically sound method for spacing
concentrations in addition to good general guidance for this chronic
study. The Agency agreed to provide guidance (advice) through the
rapid feedback process. The SOP developed for obtaining rapid
feedback explains the documentation needed for inclusion in the final
report for EPA review.
6. Rejection Factor: The raw data for production of offspring, adult survival, weight, and
length were not provided. Therefore, the reported MATC and NOEL
values could not be verified.
Agency Guidance: Subd-E: p. 29, no. 70-4 (b)(l) and (2) and p. 78, no. 72-4 (c)(l) -
(11). ASTME- 1193: p. 792, 13.1.9 and p. 792, 15.1.11. FIFRA-
TG: 72-4(b), p. C-67, no. 9.
Assessment of
Guidance:
Avoidable:
Generally, the guidance hi this area is adequate.
Yes.
Industry Comment: Industry disagrees that guidance on this factor is adequate. Below is a
list of the referenced EPA guidance with relevant text. Industry feels
that guidance documents with asterisks above clearly establish that
some form of "raw" data is required. However, since raw data can be
defined in a variety of ways, the guidance can cause confusion.
* • Subd-E: p. 29, §70-4(b)(l) and (2). Both sections are relevant,
especially (2). Each test report should include "... enough data for the
Agency to verify calculated statistical values."
• Subd-E: p. 78, §72-4(c)(l) to (11). All sections are not pertinent, only
#4 ". . . include mortality data."
• ASTM E-1193: p. 792, 13.1.9. Requires data on survival, growth, and
reproduction.
134
-------
ASTM E-1193: p. 792, 15.1.11.
recalculation of growth MATC.
Requires data sufficient to allow
EPA Response:
Resolution:
• FIFRA-TG: p. C-67, 72-4(b), No. 9. MATC value reported in ppm.
Raw data provided.
• SEP Daphnia Life Cycle: p. 6, D. Requires that "results (of the
study) must be accompanied by copies of the original (raw) data ..."
A second source of confusion is the inconsistency with which EPA
invokes various documents as guidance. For example, the guidance
documents listed for this factor are not the same as those listed for the
very similar rejection factor no. 1 above.
In general, Industry agrees that this is avoidable. However, clear
delineation of appropriate guidance might help eliminate future
problems. Raw data must be clearly defined.
In addition, Industry believes that:
1. Subd-E provides the best guidance for this subject, since it
establishes that only enough data needs to be submitted to allow
independent verification of endpoints.
2. EPA should contact registrant and request data needed prior to
finalizing and issuing the DER, rather than rejecting the study
because the data were not submitted. This would prevent needless
rejection of studies.
For the Daphnia life-cycle study, data in the following manner need to
be submitted in the final report: survival, mortality, and reproduction
data for each replicate chamber for each day observed; lengths and or
dry weights for each individual live daphnid for each test concentration
and control at the end of the study. This will provide sufficient data
for the reviewer to verify the reported MATC, NOEL, and LOEL
values for each biological endpoint.
Industry and EPA are in agreement with what needs to be submitted.
Other Industry Comments on Guideline 72-4
A Daphnia chronic study was initially rejected by T3HR because of three factors:
1) variable test concentrations, 2) no raw data, and 3) experimental design did not follow
135
-------
SEP (static renewal protocol with 7 beakers containing individual adults and 3 beakers
containing 5 individuals per beaker). All factors were ultimately resolved and the study was
accepted as core.
Industry included this case study to illustrate the lack of sufficient guidance in identifying
an appropriate protocol for the Daphnia chronic study. EPA's SEP is written for a static
renewal system, whereas many Daphnia chronic studies are performed under flow-through
conditions. Consequently, the experimental design in the SEP may be difficult to implement
under conditions imposed by splitter devices or beaker turnover rates.
EPA Response: The Agency suspects that the study was rejected primarily on the basis of
the absence of tabulated and unanalyzed data coupled with erratic test concentrations. It has
been established historically through numerous written and verbal communications that data
(all unsummarized and unanalyzed data) is needed by the Agency to facilitate its evaluation.
The Agency concurs that the SEP in particular, and the study design generally, need
complete revisions to reflect flow-through conditions.
INVERTEBRATE LIFE CYCLE (ESTUARINE SPECIES (Mysid))
(Guideline 72-4)
1. Rejection Factor: Reproduction rate of the mysids hi the controls and in the test
concentrations was very low and could not be statistically analyzed.
Agency Guidance: ASTM E - 1191: p. 764, 13.1.10. FIFRA-TG: 72-4(c), p. C-72, no.
9 C-75, no. 9.
Assessment of
Guidance: The guidance documents which address this factor appear adequate.
Avoidable:
Industry Comment:
EPA Response:
Yes.
Some confusion may arise if laboratories performing the study do not
have explicit instructions to follow the ASTM protocol. This should be
clearly identified in official guidance. Subd-E Guidelines reference the
protocol by Nimmo et al. (1978) as the only example of an acceptable
protocol for the mysid chronic reproduction test. This reference was
not listed as guidance in the present analysis. Perhaps it will be
necessary to update the Guidelines to acknowledge this change, so that
laboratories will know which protocol to follow. EPA should provide
an official list of protocols that will yield acceptable studies if followed.
The Agency concurs that acceptable protocols should be listed by the
Agency. Two suggested protocols are listed below „
136
-------
American Society for Testing Materials. 1990. Standard Guide for
Conducting Life-Cycle Toxicity Tests With Saltwater Mysids. E 1191-
90. Published by ASTM Committee E-47 on Biological Effects and
Environmental Fate, Philadelphia, PA, 19103.
Nimmo, D.E., L. H. Banner, R.A. Rigby, J.M, Sheppards and A.J.
Wilson, Jr. 1977. Mysidopsis bahia: An estuarine species suitable for
life cycle bioassays in determining sublethal effects of a pollutant. In
Aquatic Toxicology and Hazard Evaluation (F.L. Mayer and J. L.
Hamelink, eds.)5 American Society For Testing Materials, Philadelphia,
PA, STP 634, pp. 109-116.
Nimmo, D.E., T.L. Hamaker, and C.A. Sommers. 1978. Entire life-
cycle toxicity test using mysids (Mysidopsis bahia) in flowing water.
Pp. 64-68 in Bioassay Procedures for the Ocean Disposal Permit
Program. U.S. Environmental Protection Agency, Office of Research
and Development. EPA-600/9-78-010.
2. Rejection Factor: Adult body lengths were not measured at the end of study.
Agency Guidance: ASTM E - 1191: p. 762, 11.9.4 and p. 764, 13.1.8. FIFRA-TG: 72-
4(c), p. C-72, no. 9 and C-75, no. 9.
Assessment of
Guidance:
Avoidable:
Industry Comment:
EPA Response:
The guidance documents which address this factor appear adequate.
Yes.
Industry questions the need to have two estimates of mysid growth in a
study principally aimed at determining reproductive effects in
chemicals. EPA should (1) provide an official list of protocols that will
yield acceptable studies if followed, and (2) require only one estimation
of growth. Industry recommends length as the principal measure of
growth.
The Agency concurs that acceptable protocols should be listed by the
Agency. Suggested protocols are listed below.
American Society for Testing Materials. 1990. Standard Guide for
Conducting Life-Cycle Toxicity Tests With Saltwater Mysids. E 1191-
90. Published by ASTM Committee E-47 on Biological Effects and
Environmental Fate, Philadelphia, PA, 19103.
137
-------
Nimmo, D.E., L. H. Bahner, R.A. Rigby, J.M. Sheppard, and AJ.
Wilson, Jr. 1977. Mysidopsis bahia: An estuarine species suitable for
life cycle bioassays in determining sublethal effects of a pollutant. In
Aquatic Toxicology and Hazard Evaluation (F.L. Mayer and J. L.
Hamelink, eds.), American Society For Testing Materials, Philadelphia,
PA, STP 634, pp. 109-116.
Nimmo, D.E., T.L. Hamaker, and C.A. Sommers. 1978. Entire life-
cycle toxicity test using mysids (Mysidopsis Bahia) in flowing water.
Pp. 64-68 in Bioassay Procedures for the Ocean Disposal Permit
Program. U.S. Environmental Protection Agency, Office of Research
and Development. EPA-600/9-78-010.
The Agency concurs that this test method was designed primarily for
determining reproductive effects. However, a complete environmental
assessment includes data on growth, survival, and reproduction.
Consequently, the Agency believes that growth data is important for the
overall risk assessment. The Agency believes that growth is two-
dimensional, longential as well as transversal. Therefore, the Agency
would prefer to see both length and weight measurements for mysids.
Resolution: NACA raised the question of whether one measure of growth would be
sufficient for the mysid, since EPA has agreed that only one measure of
growth (length) is necessary in the Daphnia Life-Cycle test.
EPA did not agree during the meeting that only one measure of growth
is acceptable. EPA wants to work with the ERL at Gulf Breeze to
resolve this issue. As data becomes available comparing the two
measures of growth, NACA may like to follow-up with EPA on this
point.
3. Rejection Factor: The feeding rate of the mysids was below the recommended daily
ration.
Agency Guidance: ASTM E - 1191: p. 760, 10.5 and p. 761-762, 11.5.
Assessment of
Guidance:
Avoidable:
Generally, the guidance given in this area is adequate.
Yes.
Industry Comment: Industry recognizes that ASTM provides guidance on the appropriate
feeding rate for mysid shrimp in the chronic study. However, Industry
138
-------
EPA Response:
Resolution:
feels that deviations from recommended feeding rates is not sufficient
cause for rejection of the study. Rejection of the study would not be
appropriate if the control group was meeting all acceptability criteria.
Also, Industry feels that some confusion may arise if laboratories
performing the study do not have explicit instructions to follow the
ASTM protocol. This should be clearly identified in guidelines or the
SEP. EPA should provide an official list of protocols that will yield
acceptable studies if followed.
Guidance is adequate for recommended mysid feeding rates. While
study guidance recommends acceptable feeding rates, a study will not
be rejected solely because feeding deviates from recommended levels.
Biological response (i.e., survival, growth, reproduction, lack of
cannibalistic behavior) will also be evaluated in determining if optimal
feeding was maintained during a study.
NACA does not believe that feeding rate should be a reason for
rejection if the controls are growing and surviving. Mysids are
cannibalistic if there is insufficient food. The biological response is
what needs to be optimal, not the feeding rate. HEP disagreed and said
that the feeding should be optimal.
NACA pointed out that the feeding rate was only mentioned in one
recommended protocol and is probably not the only feeding rate which
is acceptable. EEB agreed that biological response is important, and
agreed that there would be more flexibility within reason. FF;R
continues to emphasize that in order for toxicity studies to be truly
comparative, absolute standardized, optimal conditions must be used.
4. Rejection Factor: The raw data for adult mortality and growth, production, and survival
of offspring were not provided. Therefore, the reported MATC and
NOEL values could not be verified.
Agency Guidance:
Assessment of
Guidance:
Avoidable:
Subd-E: p. 29, no. 70-4 (b)(l) and (2) and p. 32, no. 70-4 (c)(7) and
p. 78, no. 72-4 (c)(l) - (11). ASTM E - 1191: p. 764, 13.1.8 and p.
764-765, 15.1.11. FIFRA-TG: 72-4(c), p. C-72, no. 9 and 10.
Generally, the guidance given in this area is adequate.
Yes.
139
-------
Industry Comment:
Industry disagrees that guidance on this factor is adequate. Below is a
list of the referenced EPA guidance with relevant text. Industry feels
that guidance documents with asterisks clearly establish that some form
of raw data is required; however, since raw data can be defined in a
variety of ways, the guidance can cause confusion.
A second source of confusion is the inconsistency with which EPA
invokes various documents as guidance. Guidance documents cited for
this rejection are conflicting. For example, one document requires
sufficient data to calculate statistical endpoints, whereas another
requires submission pf raw data.
• Subd-E: p. 29, §70-4(b)(l) and (2). Both sections relevant, especially
(2). Each test report should include "... enough data for the Agency
to verify calculated statistical values."
• Subd-E: p. 32, §70-4(c)(7). Not relevant; specifies types of data
analysis.
• Subd-E: p. 78, §72-4(c)(l) to (11). All sections are not pertinent,, only
#4 ", . . include mortality data."
• ASTM E-1192: p. 764, 13.1.8. Requires data on survival, growth and
reproduction.
• ASTM E-1192: p. 764-765, 15.1.11. Provide a table of data on
survival, growth, and reproduction of mysids ". . .in sufficient detail
to allow independent statistical analysis."
• FEFRA-TG: p. C-72, 72-4(c), No. 9 and 10. No. 10 specifically
requires raw data.
In general, Industry agrees that this was avoidable. However, clear
delineation of appropriate guidance might help eliminate future
problems. Requiring submission of raw data is not acceptable
guidance, without an definition of "raw."
In addition, Industry believes that:
1. Guidance provided by Subd-E is most adequate for this subject,
since it establishes that only enough data needs to be submitted to
allow independent verification of endpoints.
140
-------
EPA Response:
2. EPA should contact registrant and request data needed, rather than
rejecting the study because the data were not submitted. This
would prevent needless rejection of studies.
For the mysid life cycle study, data in the following manner need to be
submitted in the final report: daily survival, mortality, and
reproduction data for each replicate chamber; lengths and dry weights
for each individual live mysid for each test concentration and control
alive at the end of the study. This will provide sufficient data for the
reviewer to verify the reported MATC, NOEL, and LOEL values for
each biological endpoint
Other Industry Comments on Guideline 72-4
Mysid spawn can be affected by individual or group matings. Guidance is not clear, but
EPA rejected this study because spawns were individual rather than groups. Principally, this
is a problem of inadequate guidance. Official guidance should be published that addresses
appropriate experimental designs for mysid chronic studies.
EPA Response: The Agency agrees with Industry that it needs to provide better guidance for
all aspects of this study.
Resolution: EPA agreed that the guidance is not clear. ASTM standard guide E-l 191
requires that mature mysids be paired prior for reproduction, whereas the Nimmo protocols
require group spawning. Until the Agency determines that one method is preferable to the
other, it will be flexible in reviewing the studies, provided that the study clearly states which
protocol it is following.
Additional Industry Comment:
Concerns Regarding the Study:
• Rejection rate is high.
• Studies are expensive.
• The utility of the data is unclear. Of the 20 or more Life-cycle studies performed, it is
difficult to determine how they were used in the risk assessment.
• The restart rate for the studies is high, more of a problem since variability is often seen
late in a study.
• Difficult to conduct the study.
There was a NACA/EPA Work Group which began to address the issues in 1992.
NACA submitted a protocol to the EPA for review. Due to EPA resource constraints, the
141
-------
review of the protocol was suspended. Concerning the use of the data, EPA said there is a
difference hi how the data is used for regulatory purposes and how OPP uses the data to
understand the compound. It is still unclear how useful the study actually is and since it is
so problematic, it should be reviewed.
NACA Recommendation: A small, focused work group consisting of 2-3 NACA
representatives and 2-3 EPA personnel should be formed, with an established completion date
(6-12 months) to evaluate the Fish Life-cycle study.
[EPA priorities must weigh strongly in determining when this can be accomplished.]
Additional EPA Comment:
The mysid test is relevant. This study determines whether a pesticide present in an
estuarine environment will affect reproduction in exposed estuarine invertebrates. This study
is the true counterpart to the freshwater Daphnia invertebrate toxicity test.
Resolution:
1) NACA stated that due to difficulties hi study design/conduct, the sensitivity of the study
to identify reproductive effects is not optimal. EPA does not agree that the sensitivity of
the study is not optimal. EPA wants to consult the literature and the ERL at Gulf Breeze
in addition to its own data records prior to determining if it agrees with this statement.
2) The reproductive NOELs from the Early Life-Stage are often similar to the NOEL from
the Life-cycle study.
3) Time extensions will be granted (assuming adequate justification is provided) when
requested as long as it is not the last study controlling the finalization of the RED.
4) NACA agreed to resubmit the protocol developed in the original work group by the end
of April and EPA agreed to review.
5) Consideration should be given to the development of a test that addresses reproductive
endpoints more effectively.
6) Rejection rate analysis for the submitted Fish Life-Cycle studies would be useful,
although resource constraints must be considered.
142
-------
SIMULATED AND ACTUAL FIELD TESTING - MAMMALS AND BIRDS
(SIMULATED STUDY) (Guideline 71-5A) and (ACTUAL STUDY) (Guideline 71-5B)26
A discussion of the rejection factors, Agency guidance, or assessment of guidance is not
presented because, essentially, this experimental design (i.e., simulated (pen) study) is no
longer acceptable for addressing potential terrestrial hazards. With publication of the
Guidance Document for Conducting Terrestrial Field Studies (EPA-540/09-88-109)27 in
September 1988, simulated (or pen) studies are no longer done or considered appropriate by
the Agency.
In addition, with the publication of the New Paradigm, terrestrial field studies, under
most conditions, are no longer required. Therefore, there are no comments for Guideline
71-5B.
Additional Industry Comment:
Industry commented on the following guidelines:
GUIDELINE §§122-l(a) and 123-l(a)—SEED GERMINATION
EPA did not include non-target plant testing in the rejection rate analysis of ecological
effects studies. Industry has encountered the following issues in meeting EPA acceptance
criteria for non-target plant-testing requirements:
1. Industry Concern: Seedling germination test not included in registration package.
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1, 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
• FIFRA-TG: §122-2
Industry Assessment of Guidance: Industry questions the need for a separate set of tests to
study germination, as germination is part of the seedling emergence test. If a shoot does not
emerge above the soil line, it is clear that the plant has been affected. Whether the product
is embryocidal or impacts emergence of the shoot is immaterial.
26 This section concerns studies done using the GDTFS.
27 For the rest of this paper this document will be called GDTFS.
143
-------
Industry Assessment ofAvoidability: Avoidable. The germination test can continue to be
part of the registration submittal. However, Industry questions whether this study
substantially adds to the risk assessment process, as no company has observed a test
substance passing the emergence test but failing the germination test.
Industry Recommendations: Industry requests that EPA evaluate the need for this test.
Industry believes that elimination of these germination tests would not severely impede
EPA's ability to assess the potential impact of products to non-target terrestrial plants, while
it would relieve the number and cost of studies to the registrant, and reduce EPA cost and
manpower requirements to review these studies. EPA personnel could focus on review of
the seedling emergence and vegetative vigor/shoot length test results.
EPA Response: The Agency will consider a waiver of seed germination studies (122-1,
123-1).
Resolution: NACA requested that the Agency consider a waiver of seed germination studies.
The testing scheme proposed by the Agency suggests that all seed germination studies (as
currently required under Subdivision J of the guidelines, 122-1 and 123-1) be waived.
Percent germination standards are discussed in the new proposed testing scheme.
2. Industry Concern: Insufficient germination in the control group.
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 15 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
• FIFRA-TG: §122-2
Industry Assessment of Guidance: Rejection has been noted if the percent germination is
below 70%. There is no set number given in any of the guidance documents stating or
suggesting that seeds must demonstrate 70% germination. Rather, the guidelines do state
that the seed batch should be "healthy and germinable." Not all plant species will provide
70% germination. Additionally, the test conditions recommended by EPA are not optimal
for all species. Few seed companies use this type of test for determining seed germination
rates. Industry questions the need for this test, as germination is part of the seedling
emergence test (see comments in next section).
Industry Assessment ofAvoidability: Avoidable. Rather than basing acceptance on a set
percent germination number, focus should be given to whether the study answers/refutes the
question of whether the test material produces a 25 % effect.
144
-------
Industry Recommendations: Some companies have designed this test so that there is
sufficient replication based on percent germination for each specific species. The number of
seeds per replicate is set to statistically refute a 25 % effect, even though the germination may
be less than 70%. The approach advances any species to Tier n if a presumption of 25 %
inhibition cannot be refuted. EPA has also stated that the germination rate for each seed lot
should be reported. We believe that controls and historic control data produced from
germination trials are more appropriate and should serve this purpose.
EPA Response: The Agency has never invalidated a plant study solely because the seed
germination fell below 70%. We would, however, consider a study invalid if a sufficient
number of control plants died or did not germinate so that a statistical analysis was not
possible. Minimum percent germination is not currently addressed by Subdivision J
Guidelines, however, in the interests of Good Laboratory Practices, the minimum acceptable
percent seed germination must be determined for each seed lot tested if the study is to have
statistical validity. The Agency suggests that Industry supply percent germination
information from the seed packages for Agronomic crops. For vegetable crops, the Agency
suggests that the Association of Official Seed Analysts (AOSA) and Federal Seed Act test
standards be used. Minimum acceptable seed germination standards are set by the USDA
under Federal Seed Act Regulations (7 CFR, Section 201.31) The values for the agronomic
crops are based on the experience of the agronomists in EEB and will be used for study
rejection purposes. The Agency does not believe that historic control data from previous
tests are more appropriate or serve this purpose because certified germination will vary from
year to year and from packet to packet.
VEGETABLE MIN. % GERM AGRON. CROPS MIN % GERM
artichoke 60
asparagus 70
asparagus bean 75
bean, garden 70
bean, lima 70
bean, runner 75
beet 65
broadbean 75
broccoli 75
brussels sprout 70
burdock, great 60
cabbage 75
cabbage,
tronchuda 75
cantaloupe 75
alfalfa 70
small grain28 80
bean, field 70
beet, field 65
sugar beet 55
bentgrass 65
bluegrass 65
bermudagrass 75
bluestem 50
brome 65
buckwheat 60
buffalograss 50
canarygrass 60
corn, field 85
28Small grains include sorghum, barley, wheat, oats, rye, millet, and rice.
145
-------
cardoon
carrot
cauliflower
celeriac
celery
chard, Swiss
chicory
60
55
75
55
55
65
65
Chinese cabbage 75
chives 50
citron 65
collards 80
corn, sweet 75
cornsalad 70
cowpea 75
cress, garden 75
cress, upland 60
cress, water 40
cucumber 80
dandelion 60
eggplant 60
endive 70
kale, all types 75
kolhrabi 75
leek 60
lettuce 80
mustard,
all types 75
okra 50
onion 70
pak-choi 75
parsley 60
parsnip 60
pea 80
pepper 55
pumpkin 75
radish 75
rhubarb 60
rutabaga 75
satisfy 75
soybean 75
spinach 60
spinach,
New Zealand 40
squash 75
corn, pop 75
vetch 75
fescue, tall 75
fescue, other 60
grama 50
lespedeza 60
lupine 70
panicgrass 50
orchardgrass 70
rape, all types 75
redtop 55
ryegrass 75
safflower 75
soybean 75
sudangrass 70
tobacco 55
wheatgrass 50
clover 70
146
-------
tomato 75
tomato, husk 50
turnip 80
watermelon 70
The susceptibility or tolerance of plant species to pesticides can vary significantly. Data
on plant sensitivity generated in pre-registration efficacy screening trials should be considered
when choosing species that are most appropriate for Subdivision J testing. The separation of
cold hearty and warm season plants for optimum seed germination, seedling emergence, and
vegetative vigor under different temperatures is suggested.
3. Industry Concern: Presence of a seed treatment.
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
Industry Assessment of Guidance: Species selection is generally based on species that are
known to be affected by the product. Laboratories select a suite of species that will produce
a spectrum of response for the compound. Therefore, for crop protection products that are
known to be selective (grass control versus broadleaf control), the same suite of species may
not be used for every compound tested. This is not always an issue for non-selective
compounds. For crop selective products it is also known, a priori, that a grass-control agent
is likely to be more selective toward monocotyledons and therefore little-to-no effect might
be observed to dicotyledon species, and visa versa. Where possible, laboratories conducting
these tests have used treatment-free seeds. However, this is not always possible for all
species. To select species on the basis of availability of treatment-free seed would be
inappropriate. There are also problems arising from fungal contamination with some species,
making testing extremely difficult. The seed treatment, when compared to appropriate
controls, should not interfere with developing adequate response data. Additionally, EPA
has given verbal approval to the use of treated seed if no source of untreated seed was
available.
Industry Assessment ofAvoidability: Avoidable. Laboratories conducting these tests could
limit species selection to only those available from suppliers as seed-treatment-free.
However, this would limit testing and might not allow the spectrum of compound sensitivities
to be established for risk assessment purposes.
Industry Recommendations: Industry believes that seed treatments do not impact results on
the control group, and therefore EPA should accept studies performed with treated seeds.
147
-------
EPA Response: The Subdivision J Guidelines currently specify that seeds may be steam
sterilized; however, no other seed treatments are allowed. Verbal approval to use treated
seeds has been given by Agency staff as long as the registrant proves that there are no
synergistic or antagonistic interactions with the test chemical, or direct phytotoxicity to each
of the tested plants (all 10). This would, of course, require the submission of phytotoxicity
data for each seed treatment and combinations of them if used, as compared to a negative
control. Industry may provide research data that supports their position that the use of other
chemicals in the test will not adversely impact study results. This data must be presented in
a form that will allow Agency reviewers to apply it to all reviews. The Agency suggests a
published, peer-reviewed document.
4. Industry Concern: Lack of negative controls.
EPA Guidance: None
Industry Assessment of Guidance: EPA has rejected some studies that do not include a
solvent-free control when a solvent is used to apply the test substance and only solvent
controls are used. This includes tests where the test substance has been applied in a solvent,
the solvent is allowed to evaporate, and water is added for germination. Such studies have
been rejected due to concerns that residual acetone may have resulted in reduced germination
in the controls. Additionally, when differences exist between solvent controls and solvent-
free controls (e.g., aquatic non-target plant testing), the guidelines state that the EC50 must be
calculated based on the solvent control. Industry questions the need for a solvent-free
control, as it is not used in the data analysis.
Industry Assessment ofAvoidability: Avoidable. Tests conducted to evaluate the effects of
the solvent should be performed to verify that the solvent does not reduce germination.
Industry Recommendations: If it can be shown from a control study that the prescribed
solvent used in a test on a specific species does not result in effects, routine inclusion of a
negative control (solvent free control) should not be necessary. If a difference is apparent,
then the percent inhibition data should be developed for the solvent control and the solvent
control study should be referenced whenever it is used in subsequent studies. The percent
inhibition due to the test substance, however, should still be determined from the solvent
control data.
EPA Response: The Subdivision J Guidelines do specify that a negative control is required at
aU times, even when a solvent control is used. Without the negative control, there is no way
to determine if a reduction in growth in the solvent control (if and when it occurs) is a
significant reduction if no comparative data are available. Once again, the Agency has been
flexible with Industry regarding this point. The Agency encourages that Industry provide
research data to support the position that the solvents routinely used in the plant studies are
not directly phytotoxic to plants, and that they are not interactive with the test compound. If
148
-------
a significant adverse effect to plant growth occurs in the solvent control, then the study
should be repeated using a less toxic solvent. If there is no significant difference between
the two sets of control data as determined by the Williams test, then the data may be pooled
for use in the statistical analysis.
GUIDELINE §§122-1 and 123-1—SEEDLING EMERGENCE and VEGETATIVE
VIGOR
1. Industry Concern: Presence of a seed treatment.
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1 30 Julv
1985 J
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
Industry Assessment of Guidance: All species where seed treatments have been used have
been rejected.
Industry Assessment of Avoidability: Avoidable.
Industry Recommendations: See Industry Concern above under "Germination."
EPA Response: Refer to Agency response to NACA Concern 3, above.
Resolution: Untreated seeds must be used wherever possible for test and control plants.
Studies with treated seeds will be categorized as supplemental (that is unless other rejection
factors render the study invalid) until or unless industry shows with testing that the seed
treatment did not influence the toxicity of the test material or adversely affect seed
germination and emergence. These supplemental studies will not have to be repeated unless
it is deemed necessary, by the Agency, for risk assessment or risk reduction evaluation.
2. Industry Concern: Lack of negative controls.
EPA Guidance: None
Industry Assessment of Guidance: For low-solubility compounds, the test substance is
sometimes applied with a solvent in the carrier liquid. Some tests conducted where only a
control that has been sprayed with a water and solvent have been rejected or deemed
supplemental. EPA insists that a water-only control for comparison to evaluate solvent
effects is necessary. As described above, tests comparing a solvent control against a water-
149
-------
only control for both emergence and vigor testing could be used to determine whether the
solvent concentration produces an effect. However, if it does, the solvent control—not the
water control—is appropriate for use in calculation of the EC2S.
Industry Assessment ofAvoidability: Avoidable.
Industry Recommendations: See Industry Concern above under "Germination."
EPA Response: Refer to Agency response to NACA Concern 4, above.
3. Industry Concern: Test duration of 14 days is not adequate.
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1, 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
• HFRA-TG: §122-2
Industry Assessment of Guidance: Guidelines indicate that the duration of the germination
and seedling growth test is 14 days. It can be extended for a longer period to allow for
vigor measurements at the registrants discretion.
Industry Assessment ofAvoidability: Not avoidable. This is not avoidable as 14 days is
given as the study duration in the guidelines.
Industry Recommendations: Industry believes that 14 days is generally sufficient for test
duration, as stated in the guidelines.
EPA Response: The Agency finds the 14-day rating period unacceptable for many slow
acting, foliar-applied herbicides (translocate slowly to the active growing points in the plant
and result in a slow phytotoxic effect). Some recent labels clearly inform growers of a delay
before phytotoxic symptoms appear. These statements are placed on labels to inform
growers that the herbicide is slower acting than previously used herbicides. The pesticide
registrants often know the mode-of-action, weed-control spectrum, and longevity of efficacy
of their product before Subdivision J tests are conducted. The Agency recommends a
minimum test duration of 4 weeks for seedling emergence and vegetative vigor studies.
Resolution: Generally, the test period should remain 14 days. However, if the compound is
a slow-acting foliar herbicide, the test period should be extended from 14 days to 4 weeks.
Registrants should make best effort up front to meet this requirement.
150
-------
4. Industry Concern: Use of pest control agents, fertilizer, or other materials to maintain
healthy plants is unacceptable.
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1, 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
• FIFRA-TG: §122-2
Industry Assessment of Guidance: This issue is not addressed in the guidelines. While it is
preferable to not have pests and diseases in the greenhouse during these tests, it is not
uncommon that plants reared in greenhouses may sometimes be subjected to outbreaks of
such pests as aphids, white flies, mites, and fungal pathogens. While some greenhouses
incorporate the use of natural predators to control some of these pests, this approach is not
always effective for every pest. To prevent the plants from becoming impacted by these
pests, appropriate pest control products may be used occasionally to prevent the loss of a
test. It is Industry's belief that as long as the products used and reasons for use are clearly
documented, this should not be a reason for rejection. This is consistent with animal testing,
where animal health treatments are permitted under well-documented circumstances.
Fertilization is necessary to establish healthy control plants, as soils—especially those used in
these tests—may lack sufficient nutrients to grow plants successfully.
Industry Assessment ofAvoidability: Not avoidable. Treatment of greenhouses with pest
control agents is critical for greenhouse culturing of plants. For the longer-term studies
(emergence and vigor studies), an outbreak may occur which will impact the validity of the
study if not controlled. These treatments will not impact the outcome of the study.
Conversely, the studies could be discarded and restarted. However, it is unlikely that this
effort will result in more reliable results.
Industry Recommendations: Even under the best controlled conditions, and particularly
during certain times of the year, pest outbreaks will occur under greenhouse conditions and
pest control must be exercised to maintain control in the greenhouse and for the study.
Under these circumstances, the curative products used, the period during the test when these
conditions existed, and "historical" phytotoxicity data on the compounds used as curative
treatments, should be presented in the report. For example, if a company is required to use
a particular compound during a test to control whiteflies and has generated data to indicate
that this compound is not phytotoxic, then referencing such data should be adequate, and
additional "controls" for such treatments should not be required. This is critical, because it
is not possible logistically or practically to have extra controls available to test any product
that might be needed to control diseases or infestations during conduct of a test, nor would
any such controls be a representative control for the test. NACA requests that the following
statement be added to the "Resolution" section: "In lieu of treating additional controls during
151
-------
the test with any curative compounds, potential phytotoxicity of curative compounds can be
established in separate trials, and those results could be referenced to support use of such
compounds during the test."
EPA Response: The logic for not using pesticide treated seed was previously presented.
This same logic holds for other pesticide treatments that may be applied during the course of
the study. The introduction of additional pesticides during the study adds variability to an
already highly variable study. The Agency is aware of no plant studies that were rejected
because fertilizer was used in the study. Excessive use or lack of fertilizer are factors that
are considered during the study review.
Resolution: If the use of pest controls is absolutely necessary, Industry should apply
compound to all treatments and additional controls. One completely untreated control must
be maintained for the duration of the study. Industry should provide information on the
antagonistic/synergistic effects of the pest control agent and the test substance. It was agreed
that a Workshop/Working Group should be established to further evaluate this issue, as the
Agency and NACA have not reached a resolution.
5. Industry Concern: Guidelines state that the test substance must be the technical grade
of the active ingredient. The test substance was a formulated product rather than the
technical grade active ingredient.
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1, 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
• FIFRA-TG: §122-2
Industry Assessment of Guidance: Currently, the guidelines require testing to be done on
TGAI. However, to apply TGAI to plants it is necessary to use a solvent and/or make a
simple formulation. Though it is not considered necessary to test all formulations or to
remove the option of testing TGAI, it is proposed that the guideline should be modified to
permit the testing of either TGAI or a typical end-use single formulation.
Industry Assessment ofAvoidability: Industry is unaware of studies being rejected on the
basis of testing appropriate formulations, and so believes the guidelines should reflect this
practice.
Industry Recommendations: Industry believes that the guidelines should be modified to state
that the testing of a formulation and/or the end-use product should be allowable.
152
-------
EPA Response: The Agency agrees with this statement, and encourages the use of TEP in
future plant testing. If the pesticide registrant has many different formulations registered, the
one with the highest percentage of a.i. and /or the one most widely used should be tested
Additional tests might be required upon the introduction of new formulations.
Resolution: It was agreed that either TGAI or TEP would be acceptable until the time new
direction is given by the Agency via 40 CFR part 158 or the guidelines, Subdivision J.
6. Industry Concern: Watering of pots in the greenhouse from the top for emergence
testing. fe
EPA Guidance: None
Industry Assessment of Guidance: Some studies have been rejected because the plants have
been watered from the top. This is a usual testing practice, even in efficacy testing and is
not excluded in the current guideline. Therefore, it should not be a reason for rejection if
the registrant can provide sufficient evidence that the test chemical was retained within the
test system and would be in contact with the seeds, roots, etc., during the study. Bottom
watering could be done, but can lead to a higher incidence of plant failure through root
disease.
Industry Assessment of Avoidability: Avoidable. However, because of the points mentioned
above, Industry does not believe that watering of the plant from the top should be cause for
rejection.
Industry Recommendations: The most important point is that the chemical is contained within
the system. This cannot be done easily with bottom watering. With top watering, water
will quickly penetrate the dry soil and the net movement of water will be upwards due to
evapotranspiration. Therefore, watering from the top should be an acceptable practice.
EPA Response: One particular study was rejected because it was believed that a highly
soluble pesticide may have been washed through the soil because of excessive watering
events. In this case, bottom watering was recommended so that the chemical would move
upward through the soil with evaporation, thus coming in contact with the germinating seeds.
There are pluses and minuses to both methods of watering. The Subdivision J guidelines are
not specific with regard to method of watering or frequency of watering, for good reason.
Each greenhouse or growth chamber may have different environmental conditions that would
require a different watering frequency and/or methodology. The Agency does not routinely
reject plant studies solely on the basis of watering method or frequency. The Agency
requests that Industry utilize data regarding the solubility, volatility, and Kd value of the test
chemical to determine its affinity to the growth media prior to conducting the study so that
the best method and frequency of plant watering is used. The Agency would prefer that
153
-------
seedling emergence studies utilize bottom watering. For vegetative vigor studies, top
watering under the foliage or bottom watering can be used.
The Agency recommends that non-porous plant containers be used. Peat and clay containers
may adsorb the test chemical. With regard to propagation media, soil mixes containing
>2% organic matter should be avoided. At the other extreme, 100% acid washed sand,
glass beads, or rockwool are not recommended either.
Resolution: Watering from the bottom is preferred. However, when watering from the top,
care should be taken not to wash the compound off of the plant, and chemical physical
properties of the test substance should be considered (solubility, Kow, K^ etc.)
7. Industry Concern: Placing more than one plant of the same species in a pot per
replicate.
EPA Guidance: None
Industry Assessment of Guidance: Rejection of a study due to the presence of more than one
plant per pot is inappropriate. It is usual greenhouse practice to grow more than one plant in
a pot when appropriate for that species. Pots for each plant species are selected to minimize
interplant competition and allow treatment of all plants by the application. Not to place more
than one plant per pot for this test would be impractical.
Industry Assessment ofAvoidability: Not avoidable. It is impractical to grow plants hi
individual containers for all species.
Industry Recommendations: Industry recommends that as long as the density of plants does
not impact the treatment of the plants, plants may be grown with more than one plant to a
pot. This practice should not be a reason for rejection.
EPA Response: The Subdivision J guidelines do not currently specify the number of
individual plants that are allowed per pot. Different types and sizes of pots are used in
different greenhouses in different parts of the country. Occasionally, we review studies in
which the plants are obviously crowded for the size and type of container. If we consider
this a significant detriment to plant growth, the study will be rejected on this basis. In the
absence of specific guidance in Subdivision J, the Agency suggests the following: 15 cm
diameter X 15 cm deep pot, use a maximum of 1 plant per pot for corn, soybean, tomato,
and cucumber; use a maximum of 3 plants per pot for sugarbeet, rape, and pea; use a
maximum of 6 plants per pot for onion, wheat, and sorghum. All others, use a maximum of
1 plant per pot if uncertain.
Resolution: The Agency agrees with the NACA proposal. However, if care is not taken to
prevent overcrowding, the study may be rejected on that basis.
154
-------
GUIDELINE §§123-1 and 123-2—PLANT PROTECTION TIER H
1. Industry Concern: Selection of rates for Tier n.
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1, 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
• HFRA-TG: §122-2
Industry Assessment of Guidance: The guidelines suggest five rates in geometric progression
in steps of 2X. At times, test rates in increments of 4X or grouping of test rates in
appropriate increments to better define the EC25 are appropriate, depending on the slope of
the dose-response curve. If steps of 2X were used, it would require a much greater number
of rates to be tested, and would require much greater greenhouse space and resources to
complete the test. These additional rates are not required, provided that the rates selected
and the magnitude of the steps between rates adequately define the response curve. Industry
believes that both EC50 and EC25 can be routinely determined.
Industry Assessment ofAvoidability: Avoidable. Studies conducted in only 2X geometric
increments can be conducted, but it is unlikely that they will improve the quality of data if
the appropriate end-points can be calculated in a reputable fashion.
Industry Recommendations: Industry suggests that studies be accepted or rejected on the
reliability of the study end-points. If concentrations other than 2X geometric increments
appropriately define the study end-points, then the study should be deemed acceptable.
EPA Response: The Subdivision J guidelines clearly state that dosages in a 2X geometric
progression be used. The Agency agrees with Industry with regard to this request. A
number of waivers of the 2X requirement have already been given, however, the Agency
must have a sufficient number of test dosages at 2X, 3X, or 4X to perform a proper
statistical analysis. A preliminary range finding study is suggested prior to final rate
selection. Studies will be rejected if there is not at least one dose greater than the EC50
value and one dose lower than the EC25 value.
2. Industry Concern: Development of appropriate dose-response curves for each
parameter tested.
EPA Guidance:
• Subd-J: §122-2
155
-------
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1, 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
• EIFRA-TG: §122-2
Industry Assessment of Guidance: Study end-points are based on the most sensitive end-point
measures. These are generally selected by observations made in a preliminary range-finding
test. Quantitatively, the effect on shoot weight has typically been observed as the most
sensitive measure, and provides the most useful and accurate dose-response curve. Due to
variability introduced with root washing, this measure is believed to be less reliable. Shoot
height and visual injury symptoms can be used, but these tend to be less quantitative. It is
impossible to provide a dose-response curve on root inhibition that covers the EC50 and EC25
in most instances, since rates of compound producing these effects would have killed the
plant through shoot effects. If a valid response curve is provided for the most sensitive of
the parameters (e.g., shoot weight), it should not be necessary to do additional work to
establish a complete response curve for all other possible measures.
Industry Assessment ofAvoidability: Avoidable. As stated in the SEP, it is not always
possible, nor is it useful, to evaluate an EC25, EC50, and a NOEL for all growth measures.
A valid dose-response curve for the most sensitive measure should suffice to perform a risk
assessment analysis.
Industry Recommendations: Industry recommends that the study end-points be based on the
most sensitive end-point measure (shoot weight, if appropriate, based on observations in the
Tier 1 tests) and if possible provide end-points for other plant growth measures when
available. However, if other less sensitive end-point growth measures cannot be measured
reliably, this should be stated in the report.
EPA Response: The Agency is hi general agreement with the Industry recommendation.
Study end-points were briefly discussed in the 1990 Corvallis Non-target Plant Testing
Workshop. Another workshop was recommended by the Corvallis workgroup so that further
discussion on this and other concerns could occur. The Agency recommends that a second
workshop be conducted.
GUIDELINE §§122-2 and 123-2—AQUATIC PLANT GROWTH
1. Industry concern: Registration and reregistration of several products, principally
herbicides and fungicides, have been hampered when a complete battery of aquatic non-
target plant studies [Selenastrum capricornutum, Lemna gibba, Anabaena flos-aqua, a
freshwater diatom (e.g., Naviculla pelliculosa) and a marine diatom (e.g., Skeletonema
costatum)] have not been submitted as part of the registration/reregistration package.
EPA Guidance: Subd-J: §122-2
156
-------
Industry Assessment of Guidance: More-specific guidance is needed to establish when non-
target aquatic plant studies are required, and which species and tests should be used. Subd-J
states: "Data on the toxic effects of a pesticide on growth and reproduction of aquatic plants
are required by 40 CFR Part 158 on a case-by-case basis to support the registration of each
end-use product intended for outdoor pesticide application." However, these tests now seem
to be required routinely for all pesticides or at least for all herbicides and fungicides.
Industry Assessment ofAvoidability: Avoidable
Industry Comments: Aquatic non-target plant testing should initially be limited to S.
capricornutum, as suggested in the guidelines, or to another comparable internationally
acceptable green algae (e.g., Scenedesmus subspicatus, Chlorella vulgaris). If the EC50 for
S. capricornutum is greater than the EEC, then further testing on other species should not be
required. Additionally, data generated by F.L. Mayer, EPA (Presentation to Aquatic Effects
Dialogue Group, 1991) support the concept that testing S. capricornutum, along with other
the aquatic toxicity tests (Daphnia and fish acute toxicity), is a reasonable approach to screen
products for potential adverse effects to the aquatic environment.
Industry Recommendations: A tiered approach developed jointly by EPA and NACA should
be used when requiring aquatic non-target plant testing, to determine if and when an entire
suite of aquatic non-target plant tests is necessary.
EPA Response: Because of the potential for phytotoxicants to reach aquatic habitats, an
analysis of the potential for adverse effects to nontarget and endangered aquatic plants is
conducted. In order to perform such a risk assessment, the 5 above-referenced aquatic plant
dose-response studies are required.
A number of issues are driving the need to improve our plant toxicity data base, and to
clarify when and which studies are required for Section 3 and reregistration.
Required plant studies for new chemicals (Section 3 actions) should be submitted along with
the required basic 6 ecotoxicology studies. The Agency data base is generally devoid of
phytotoxicity data for insecticides, miticides, nematicides, lampricides, molluscicides, or
piscicides. In the future, plant toxicity ecological risk assessments will be conducted for all
pesticide registration and reregistration actions.
2. Industry Concern: Study is rejected because it was performed according to OECD,
TSCA, or Canadian guidelines, rather than EPA guidelines.
Rejections are most commonly due to differences in cell inoculum levels (1 X 104 rather
than 3 X 103 cells/ml), light intensity, study duration (96 hours rather than 120 hours for
algae, and 120 hours rather than 14 days for Lemna), and/or growth medium.
157
-------
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1, 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
Industry Assessment of Guidance: Studies conducted according to Canadian, OECD, TSCA,
or FIFRA Subd-J guidelines all reach the same end-point: a 50% reduction in algal growth
orLemna frond development within a prescribed period of tune. Differences in cell
inoculum, light intensity, medium, and test duration were selected to reach a predetermined
growth level within a specified time, and should have little effect on the end-point relative to
the growth observed for the control group.
From a policy perspective, EPA has signed agreements with the European Community (EC)
to mutually accept studies conducted according to mutually accepted protocols. This
agreement should be acknowledged hi order for FEFRA- conducted studies to be mutually
accepted by the EC.
Industry Assessment ofAvoidability: Avoidable
Industry Recommendations: To maximize the use of available data, reduce costs, and reduce
duplication of effort, Industry proposes that studies performed under EC, OECD, Canadian,
FIFRA, and other scientifically acceptable testing procedures be mutually accepted by EPA
and other regulatory agencies around the world. Furthermore, if a study has already been
accepted by one country, review tune in another country might be reduced or eliminated.
EPA Response: The harmonization of Agency guidelines with Canadian and OECD
guidelines is currently in progress at the Agency. In the meantime, the Agency will continue
to evaluate each study on its individual scientific merits. With regard to inoculum levels, the
Agency is currently allowing up to 10,000 cells at the beginning of any of our algae studies.
The Data Evaluation Record (DER) for a given study contains a section called "Test
Procedure," in which deviations from test procedures in Subdivision J guidelines or the SEPs
are listed. The study is rarely rejected based on improper lighting, temperature, humidity,
and other environmental factors alone, unless the control organisms are significantly affected.
Subdivision J guidelines clearly specify that algae studies must be of 120-hour duration.
At the tune that the initial guidelines were written, Industry provided comments on the draft
guidelines. Industry, academia, and the Agency agreed to the 120-hour test interval for algae
studies. Generally, it was believed at that tune that the additional day was necessary to
demonstrate recovery from initial toxic effects (if recovery was to occur). A 96-hour test is
currently required by TSCA. The Agency would welcome a scientific analysis of studies in
which 96- and 120-hour test results are compared for a number of chemical classes.
158
-------
The American Society for Testing and Materials (ASTM) has published a protocol for a
7-day Lemna gibba study. Before adopting this 7-day standard, the Agency would again
welcome a scientific analysis of studies that compare 7- versus 14-day test results for a
number of chemical classes. Data in OPP files suggests that the 7- and 14-day results are
not similar for most studies.
Additional scientific input is requested with regard to 96- versus 120-hour test intervals
for algae and 7 days versus 14 days for Lemna gibba. These issues are topics for workshop
discussion.
Resolution: Some effort is being made to harmonize with other guidelines. It was agreed
that while the 120-hour test is preferred, if a valid 4 day study is submitted it will not lead to
an automatic rejection. If the study does not serve the purpose for risk assessment, the
Agency can ask for more information. NACA indicated it has no objection to the 14 day
Lemna test.
3. Industry Concern: Studies have been rejected or considered supplemental if a NOEL
is not obtained.
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1, 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
• FIFRA-TG §122-2
Industry Assessment of Guidance: Subd-J and the SEPs do not require that a NOEL be
determined in the Tier 1 and 2 non-target plant toxicity studies. However, FIFRA-TG
requires that a NOEL be reported in Tier 2 studies.
Industry Assessment ofAvoidability: Not avoidable
Industry Comments: A NOEL cannot be routinely calculated from an EC50 study. The main
purpose of an EC50 study is to determine a mean effect concentration for comparison to a
control group. Typically, five treatment concentrations are used that will bracket the EC50.
It is often not possible to run a study where sufficient effects are observed to determine an
EC50, and also have a low enough concentration to determine a NOEL.
Industry Recommendations: The dose-response relationship generated in an EC50 study is
generally adequate for determining an EC^, as discussed in the guidelines, and this can be
used for Ecological Risk Assessment and for protection of endangered species.
159
-------
EPA Response: Subdivision J does require that the NOEL be reported for Tier n studies;
however, in many cases the NOEL estimate is higher than the EC25 value. Because plant
studies are subject to a high degree of biological variation due to differences in soil type,
watering regimes, nutrients used, lighting, temperatures, greenhouses and growth chambers,
seed sources, seed germination, and varietal or hybrid differences, the number of replicates
must be increased from 3 to 5 to reduce this inherent variability. Further, all studies should
have at least one dosage above the EC50 value and at least one dosage below the EC25
value. A no-effect concentration will become increasingly important in our risk assessments
and in our "may affect" determinations for endangered/threatened species. When the
expected down-wind concentration is matched with the plant no-effect values, the Agency
should be able to derive a reasonable buffer distance for drift. The Agency sees merit in
striving for the no-effect values, especially for low-dose herbicides, but will not invalidate
studies when it cannot be obtained. Instead, we will now require that the EC50 value be
determined for endangered species purposes. In the future, we expect to set bounds
regarding acceptable and unacceptable confidence intervals.
The Agency is currently evaluating a recent publication of a statistical method for
continuous biological data: Bruce, R.D. and D.J. Versteeg. 1992. A Statistical Procedure
For Modeling Continuous Toxicity Data. Environmental Toxicology and Chemistry. Vol.
11, pp. 1485-1494. This statistical method is considered to be a more appropriate method
than the currently used probit method.
4. Industry Concern: Degradation of test substance during test period due to the presence
of the algae or Lemna sp.
EPA Guidance:
• Subd-J: §122-2
• SEP: Non-Target Plants, Growth and Reproduction of Aquatic Plants - Tier 1, 30 July
1985
• SEP: Growth and Reproduction of Aquatic Plants - Tier 2, Draft
• FDFRA-TG: §122-2
Industry Assessment ofAvoidability: Not avoidable. Industry continues to develop
compounds that complete their designated function and then degrade. Rejection of a study
because the test substance is not persistent seems to contradict the goal of both EPA and
Industry—that of protection of non-target organisms.
Industry Recommendations: Industry and EPA discussed this problem under. NACA's
ERAWG, and agreed that the test solutions should be analyzed both before and after
completion of a study, and that the EC50 should be determined based on the measured initial
concentrations, not the concentration after the study. Industry strongly supports this previous
conclusion and requests that EPA incorporate this into its guidance.
160
-------
EPA Response: The Agency will abide by the mutually agreed-upon conclusion of the
NACA/EPA guidance document.
Additional Discussion and Resolution:
There were several discussions concerning the draft Non-Target Plant testing scheme
proposed by the Agency under 40 CFR Part 158. Following is a summary of the resolutions
of those discussions. It was suggested that several of these issues be further addressed by a
work group.
1. Discussion and Resolution: Minimum USDA standards for seed germination will be
used to determine seed viability. Separate testing of viability rates is not necessary. It
was agreed that the final document will contain the standard USDA numbers for all ten
selected terrestrial species. NACA will check to see if USDA standards can be met with
non-treated seeds.
2. Discussion and Resolution: In the discussion of testing with TEP plus adjuvants, it was
decided that only a single representative adjuvant (designed to increase efficacy)
recommended on the label requires testing. No testing of possible interaction between the
adjuvant and the TEP is required. Testing requirements when more than one class of
adjuvant is recommended (anionic, ionic, non-ionic, etc.) were not resolved and should
be further discussed in a work group setting.
3. Discussion and Resolution: It was clarified that this document is NOT requiring native
species testing be initiated. This issue should be referred to a work group/workshop.
4. Discussion and Resolution: During the discussion of phytotoxicants and non-target
plants, it was decided that all phytotoxicants (not just herbicides) should be subjected to
Tier n terrestrial testing.
5. Discussion and Resolution: It was clarified that all aquatic non-target plant testing could
be completed with TGAI.
6. Discussion and Resolution: The possible use of efficacy data to determine terrestrial
EC50 phytotoxicity data for compounds non-phytotoxic substances was discussed. It was
decided that because this data is generally not collected under GLP, efficacy data would
be evaluated as support information but would be considered supplemental for risk
assessment purposes. This issue should be further discussed in a work group setting.
161
-------
Summary Table of Rejection Factors
GUIDELINE
REJECTION FACTORS
Avian Oral LD50
71-1
-Failure to establish a valid LD^ value with corresponding 95 %
confidence limits or an LD^ greater than 2000 mg/kg.
-Use of a "split-dosing" procedure.
-Fasting period prior to dosing not specified.
Avian Dietary LC50
71-2
-Mortalities attributed to interactions between animals (rather than to
test chemical), when such interactions were not observed in the
controls.
-High variability in the measured test concentrations in the test diet.
-Test material was not technical grade.
-LCjo not established when testing at dose levels < 5000 ppm a.i.
-Variation in test concentration and/or failure to adequately justify the
variations.
Freshwater Fish LCSO
(Bluegill)
72-1
-Concentration level < 100 mg/1, not high enough to produce an
LC50.
-Aeration of test chambers.
-Biological loading of test vessels twice the recommended amount.
-Test substance purity not identified.
-Inappropriate test species and/or test species not clearly identified.
-Fish fed during the exposure period.
-Minimum limit of detectability, or the minimum quantifiable limit,
not defined quantitatively.
-Test concentrations variability limit > 1.5.
-Levels of lead, iron and aluminum present in dilution water higher
than recommended.
-No solvent control.
-Results for some test concentrations obtained from tests conducted
after the definite study.
-Not all test solutions measured at 96 and 0 hours.
-No control group for the inert/carrier ingredient component of the
formulation.
-Acclimation period half that recommended.
162
-------
GUIDELINE
REJECTION FACTORS
Freshwater Fish LC
(Rainbow)
72-1
so
-All of the rejection factors listed above for guideline 72-1, Bluegill.
-Fish acclimation period overlapping with the definite study period.
-Contamination of the controls with test chemical.
-Low recovery of test chemical from the stock solutions.
-Fish weights exceeding the recommended range.
-Test temperature exceeding the recommended for the test species.
-Biological loading of the system greater than recommended.
-Fish mortality during the acclimation period higher than
recommended.
-Supersaturation of oxygen.
Acute LC50 Freshwater
Invertebrates
72-2
-Organisms not randomly distributed to test vessels.
-Water temperature not monitored.
-Chemical analyses (concentration levels) not performed on test
solutions.
-No control group for the inert/carrier ingredient component of the
formulation.
-Minimum limit of detectability, or the minimum quantifiable limit,
not defined quantitatively.
-Test concentrations variability limit > 1.5.
-Not all test concentrations measured at 0 and 48 hours.
-Levels of lead, iron and aluminum present in dilution water higher
than recommended.
-Percent of a.i. of the test formulation not identified.
-Photoperiod not as recommended.
-Raw temperature data not provided.
-Use of dechlorinated water as a portion of the dilution water.
Wild Mammal Toxicity
71-3
-Diet preparation method not adequate.
-Improper animal caging, as indicated by extensive cannibalism.
Avian Reproduction Quail
71-4
-Percent of cracked eggs in the control higher than in treatment
groups.
-Data discrepancies:
=inappropriate photoperiod;
=reasons for administration of medication not provided;
=total number of data points not included in statistical evaluation.
Avian Reproduction
Duck
71-4
-Data discrepancies:
=inappropriate photoperiod;
=inappropriate egg collection procedures;
=low overall fertility of control birds.
Acute LC50 Estuarine and
Marine Organisms Fish
72-3D
-Unexplained variations in concentrations.
-Concentration level < 100 mg/1, not high enough to produce an
LCjo.
-Dissolved oxygen levels lower than recommended.
-Analytical determination of the concentration in the test vessels not
provided.
163
-------
GUIDELINE
REJECTION FACTORS
Acute LCSO Estuarine and
Marine Organisms Mollusc
72-3B and E
-Insufficient new shell growth in control oysters.
-Insufficient dosage levels to produce a reliable LC^.
-Raw data on shell deposition not provided.
-Aeration of test chambers without chemical analyses of test
solutions.
-Dissolved oxygen levels lower than recommended.
Acute LC50 Estuarine and
Marine Organisms Shrimp
72-3C
-Test substance purity not identified.
-Chemical analysis of test solutions concentration not performed.
-Type and quantity of solvent used not provided.
-Solubility needed to achieve LC^ not obtained.
Fish Early Life Stage
72-4
-Mortality too high at all concentrations.
-Raw data not submitted.
-Survival rate in the control group lower than recommended.
-Erratic results in measured test concentrations.
Aquatic Invertebrate Life
Cycle
72-4
-Raw data not submitted.
-NOEL values for reproduction and growth cannot be established
from study results.
-Survival rate in the control group lower than recommended.
-Adults' growth (length and weight) not measured quantitatively.
Invertebrate Life Cycle
Estuarine Species
72-4
-Reproduction rates too low to be statistically analyzed.
-Adult body lengths not measured at the end of the study.
-Feeding rate below the recommended daily ration.
-Raw data not provided.
164
-------
Unresolved Issues
In Order of NACA's Priority for Resolution
Study
1. Terrestrial and
Aquatic Non-Target
Plants
2. Avian Risk
Assessment
3. Fish Full Life Cycle
4. Oyster deposition
5. Oyster Larvae
6. Shrimp Acute
Issue
General testing
methodology and risk
assessment procedures
Exposure assessment
issues
Methodology for study
conduct
Shell deposition of 2 mm
Acceptable Control
Mortality
Age of test organisms
Action/Resolution
Some issues resolved,
others remain open; work
group may be necessary
to resolve remaining
issues,
NACA proposes work
group formation to
address avian issues.
NACA proposes work
group to address issue.
Temporary solution in
place. Follow-up needed
for review of historical
growth information.
Evaluation of historical
survival data.
Evaluation of data.
165
-------
Conclusions
The overall rejection rate for ecological effects has shown improvement, having declined
from a pre-1986 rate of 36 percent to the current 20 percent. Four acute toxicity guidelines
have shown consistent improvement: avian oral (71-1), trout (72-1C), Daphnia (72-2), and
mollusk (72-3B). Despite this encouraging individual progress, however, the aggregate
rejection rate for ecological effects remains unacceptably high.
Avian and aquatic reproduction studies have current rejection rates near or above 40
percent—a clear indication that serious problems exist. Two other studies (both of which are
combined with 123-1A, Tier n nontarget terrestrial plant phytotoxicity) with rejection rates
over 40 percent are seed germination/seed emergence (122-1A) and vegetative vigor (122-
1B). Plant toxicity testing requirements are relatively new compared to animal testing
requirements. The number of rejected studies is expected to sharply diminish as testing
laboratories become more familiar with OPP data requirements and guidelines. Further, a
number of laboratories have only recently (within the past two years) begun to conduct plant
studies.The following studies have rejection rates of substantially less than 40 percent, but
over time their rejection rates have not improved significantly: bluegill acute toxicity
(72-1A), marine fish acute toxicity (72-3A), shrimp acute toxicity (72-3C), invertebrate life
cycle (72-4B), and avian dietary (71-2B).
Progress on rejection rates for Tier I avian and aquatic studies is therefore mixed. The
avian Tier I studies have lower rejection rates than their aquatic counterparts, but the avian
dietary guideline has not shown significant improvement. Several aquatic Tier I guidelines
have shown improvement, but for the most part their rejection rates remain high.
With rejection rates greater than 30 percent, the Tier n reproduction guidelines for both
aquatic and avian species have higher rejection rates than the Tier I guidelines. Rejection
factors common to the aquatic and avian reproduction guidelines include data discrepancies
and lack of tabulated and unanalyzed data submission.
In 1982, the Environmental Fate and Effects Division (then called the Hazard Evaluation
Division) issued three Pesticide Assessment Guidelines on ecological effects. The Division
also issued six Standard Evaluation Procedures during the period 1985-1986. Appendix B
gives a complete list of the ecological effects guidance documents.
The Agency's dedication to reducing ecological effects rejection rates goes beyond
improving its guidance. When future studies are rejected for factors where the Agency
believes its guidance to be adequate, and where it deems the factor avoidable, regulatory
action may be appropriate.
EPA is confident that this assessment will accelerate the downward trend in the ecological
effects rejection rate. This cooperative effort between EPA and Industry scientists furthers
OPP's goal of meeting its production schedule for REDs.
166
-------
Recommendations
As a result of the rejection rate analysis and the ensuing discussions with Industry, both
EPA and Industry realize the need for more in-depth follow-up on various ecological effects
data requirements. In order of priority for resolution are:
(1) Terrestrial and Aquatic Non-target Plants
(2) Fish Full Life Cycle
(3) Oyster Deposition
(4) Oyster Larvae
(5) Shrimp Acute Toxicity
Industry/EPA work groups have been proposed to resolve these critical issues, and EPA
intends to address them in priority order as resources permit.
Finally, SRRD intends to continue tracking rejection rates for ecological effects guideline
studies. If significant reductions in the rejection rates for these studies are not realized,
further regulatory action may be required.
167
-------
-------
Appendix A: Additional Supporting Documents
-------
-------
CONDUCTING ACCEPTABLE AQUATIC 1AB STUDIES:
FBOF08XD GUIDANCE
INTRODUCTION
The primary goal of this proposed guidance is to provide
registrants with more detailed information on how to design and
conduct aquatic laboratory studies for pesticide risk assessment.
It should be noted that this document is a "working** proposed
guidance which may be subject to further scientific review within
the Agency, and that no new policy is being introduced at this
time. This guidance emphasizes the importance of adequate
characterization of the test material and proper understanding of
how the material reacts under test conditions. This document
also attempts to:
1. Provide guidance to help pesticide registrants develop the
best toxicity information possible, taking into consideration any
constraints in the experimental design and test material:
information that is scientifically defensible and protective of
human health and the environment.
2. State EFED's position on specific areas of concern. In this
guidance, provisions are made for exceptions, while at the same
time maintaining a high level of scientific integrity.
3. State EFED's position and reasonable approach for enforcing
testing criteria, limits and standards. Standards are set with
the recognition that certain problems will arise and provisions
must be made to accommodate unavoidable problems.
4. Interpret those areas that need to be defined and set .limits
for designing and conducting laboratory studies.
DEFINITIONS AMD SITUATIONS
Analysis. 0-Hour
In static or static renewal tests, when 0-hour measurement
is indicated, it is assumed that the method of mixing the test
material in the test.solution allows adequate time for complete
dissolution equilibrium such that these initial measurements
accurately reflect exposure throughout the test container. The
registrant may need, to justify an exception from the 30-minute
requirement for adding test organisms if the characteristics of
the test material and test system require a longer equilibrium
time. If preliminary trials have been done, this delay should be
predictable.
-------
-2-
In flow-through tests, the study should be conducted with
knowledge of the tine it will take for the test material to reach
equilibrium or steady-state in the test container. Initiation of
the test and scheduling the sampling times must be based on this
information. In some cases, a flow-through system may have to be
run for an extended-time "pre-test" to try to achieve equilibrium
or steady-state. If equilibrium or steady-state cannot be
achieved, and/or it appears that the measured concentrations will
be «ubstantially below <<70%) nominal, t&e study report should
reflect that the laboratory was aware of this problem. The study
report should clearly identify the problem, indicate the steps
taken to mitigate it and justify the study design and dosing
levels. However, if sufficient analytical methods are available
and acceptable toxicity data are produced, additional testing
and evaluation with the sole objective of obtaining initial
measured concentrations greater than 70% of nominal, will not be
required.
.ASTM.. .Re.commen.dat i ons
In many cases, ASTM guidance makes recommendations based on
"when needed" or "when necessary." When referring to testing
performed to support pesticide risk assessment, EFED is
"responsible for defining "when needed" or "when necessary".
Chemistry Method
When a chemistry method is used in preliminary trials, in
range-f inding*-tests, in establishing percent purity'of batches of
test material, or in measuring concentrations in test containers,
it must be submitted with the study or referenced by MRID #. The
documentation must include a complete description of the method
so that a bench chemist can determine the necessary equipment and
perform the analysis. It must also include the raw data/
standards and chromatograro from a representative analysis using
the method. This representative analysis must be conducted with
the specific media for which it will be used during the test;
i.e., under test conditions. The actual minimum detection level
and level of quantification must be identified.
SLP Standards
According to 6LP regulations, tests not conducted according
to those regulations should not be used to support permits such
at> the registration of pesticides. Among the specific 6LP
requirements are that the test material's solubility and
stability be known -and that chemical analysis of the batch test
material be performed. Determining the solubility and stability
of the test material in the mixture or test solution is an
important part of these studies.
-------
-3-
Nominal Concentration
For aquatic tests, the nominal test level is the
concentration that would occur if all test material added to the
test solution were completely dissolved and did not dissipate in
any way •
1SFED Recommended
Recommended means that the procedure or test is preferred in
order to avoid problems, but it is not required. If the
recommended procedure or test is not performed, the study will
not necessarily be rejected.
Renewal Cvcle
Static renewal is one method to ensure relatively continuous
concentrations when the test material is not stable under test
conditions. At a minimum, the renewal cycle* should be based on
the stability of the test material under test conditions. The
time to renewal (renewal cycle) should be shorter than the time
<70% concentration of the test material to decline to
Replicates and Concentrating
When replicate test containers and measurement .of test
concentration are required, each replicate in each test
concentration must be analyzed separately. This is because the
responses in each replicate are viewed as independent and it is
necessary to know what the concentrations were so variation can
be determined. Exceptions to this occur when:
1. Treatment replicate containers under static tests or
static renewal conditions are filled from a bulk preparation. In
this case, only samples from the bulk supply for each test level
must be analyzed; or
2. A "splitter" is used in a flow-through test to feed more
than one replicate. In this case, only samples from one
replicate per treatment level require analysis. It is
recommended that samples be collected from all replicates and be
stored, in r^se anomalous concentrations are measured in the one
that is analyzed. Analyzing the other replicates may shed liqht
on the cause and extent of the anomalous measurements.
1 The renewal cycle may be shorter than required by
stability characteristics of the test material because of other
factors, such as dissolved oxygen, feeding, etc.
-------
-4-
Replicates receiving flow from a splitter should be sampled and
analyzed alternately. In other words, if there are two
replicates (A and B), replicate A should be analyzed in the first
week and replicate B in the second week, etc.
Average concentrations of replicates are used in regression
analysis.
Stability
A test material is considered to be stable uafl«r test
conditions if, under those conditions, it does not degrade,
volatilize, dissipate, precipitate, sorb to test container walls,
or otherwise decline to concentrations less than 70% of the day-0
measured concentration during the study period. If it is
expected to decline to less than 70% of the day-0 measured
concentration during the study period, either static renewal or
flow-through design is needed to try to ensure that the test
concentration is maintained at levels greater than or equal to
70%. The only exception is testing with algae and diatoms, which
cannot be tested in static renewal or flow-through systems (see
discussion on testing with algae and diatoms).
Sample Stora_ge
If samples of growth medium, stock solutions, or test
solutions collected for chemical analysis cannot be analyzed
immediately, they should be handled and stored appropriately to
minimize loss-of the test material. Loss could be caused by such
.processes as microbial degradation, hydrolysis, oxidation,
photolysis, reduction, sorption, volatilization, etc. Stability
determination under storage conditions, whether it refers to
storing the test material before testing or storing samples
awaiting analysis, is required by GLP regulation.
Under Test Conditions
The behavior of a test material should not be based on
experiments which are conducted under conditions different from
those occurring during the test. When there is a problem with
chemical solubility and/or stability, the behavior of a test
material should be determined under test conditions or those
conditions that are likely to affect the behavior of the test
material. These include but are not limited to:
test solution characteristics (salt or freshwater)
temperature, pH, conductivity, lighting
with test organisms in place
use of the same test containers
use of the same flow-through systems where appropriate
-------
-5-
Varlability in Measured Concentrations
The goal for limiting variability of measurements between
replicates of the same concentration, and over time in the same
concentration, is maintaining the ratio of the highest
concentration to the lowest concentration at 1.5 or less. A test
may be rejected if variability exceeds this amount.
An important factor in considering the limits of variability
is the avoidance of overlapping mean test concentrations between
test levels. High variability puts into question the reliability
2 ??? Jnvlronmental chemistry method and/or the concentrations
on which to base statistical analysis and toxicological
conclusions, if variability beyond the 1.5 ratio occurs, an
exception to it should be justified.
This justification should clearly state the problem, explain
why it occurred, provide scientific justification, and identify
all measures taken to mitigate the problem. The justification
also should include the fully developed chemistry method,
including the documentation necessary for a bench chemist to
review and evaluate it.
For cases in which variability problems are suspected,
registrants wishing to avoid possible rejection of a study are
strongly advised to conduct extensive preliminary trials. If it
becomes clear that high variability cannot be avoided, an
exception should be justified.
It is recommended that any justification be provided in
advance. EFED scientists will decide on the validity of the
rationale for the exception, and possibly recommend other methods
to reduce potential variability. netnoas
PRELIMINARY TRIXLS
EFED recommends that more effort be expended in designing
acceptable studies before they are conducted, especially for
problem chemicals. Subpart F (S 160.105 of 40 CFR), Good
Laboratory Practice Standards, requires that certain information
be known before beginning toxicity tests used to support permits
by EPA. The information indicated as necessary in the GLP
guidance should be developed under test conditions. This
information can be gained while doing the currently required
range-finding studies. Below is a list of recommended
preliminary tests:
-------
-6-
A. Stability trials should be conducted under test
conditions. These trials must be documented and submitted to
EFED for review with the study to which they apply.
B. Solubility trials should be conducted under test
conditions. These trials must be documented and submitted with
the study to EFED for review.
C. If solubility is a problem (<100 ppm), trials should be
conducted under test conditions using various solvents that are
most likely to be effective and that are widely recognized as
being nontoxic and other means to ensure that the appropriate
methods are used, during the laboratory tests to enhance
solubility. Once a solvent is chosen based upon more simplistic,
comparative evaluations, the decision should be confirmed in the
range-finder preliminary trials with only that solvent.
D. Chemical Analysis Methods used during laboratory studies
must be provided (or referenced by MRID f) in the laboratory
study. The submitted documentation should specify the LOD and
..LOQ, and the precision,- based on trials conducted using the
Method with samples representing test conditions. The
description of the chemical method must be in sufficient detail
that a chemist could evaluate the method.
E. Stability of the test material in the samples to be
collected for-chemical analyses should be -determined during the
laboratory studies. This includes determining whether and how
samples can be stored for future analysis.
Laboratory studies must be designed taking into account this
preliminary information. This means the above trials are to be
conducted before the definitive laboratory studies are initiated.
Publications, such as ASTM, rarely specify exactly how a
study should be designed in the detail necessary to be used as
"cookbook protocols." Suggestions are provided for study designs
with qualifying phrases such as "as needed" or "if necessary."
This implies that the behavior of the test material under test
conditions is known and assumes that the best methods and
procedures will be chosen by the laboratory doing the study.
CONFIRMING EXPOSURE WITH ANALYSTS
Acute Testincr
Acute Static Tests: Except for acute aquatic algae and diatom
studies (122-2 and 123-2, which can only be conducted as static
tests), acute static tests may be conducted only if, among other
things, the test material has been shown to be stable under the
-------
-7-
test conditions, as defined above.2 In an acute static test with
a test material that is stable and readily soluble at the
treatment levels, measurements of each test concentration are not
absolutely required. However:
1. ASTM guidance indicates that for static tests, if
possible, the concentration of toxicant should be measured at the
beginning and end of the test in all test chambers. Further,
measurement of the toxicant*s degradation products is desirable,
but not required.
2. The study may be rejected if the following occurs: 1)
the test material was not stable under test conditions, 2)
precipitates formed, or 3) solubility was likely to have been a
problem at the levels tested. However, if the recommended
chemical measurements were made to verify exposure levels, then
the study may not be rejected. Whether the study design was
modified in a scientifically defensible attempt to accommodate
these chemical characteristics will also be considered.
If variability is expected to be a problem, it is
recommended that measurements of test concentrations be made at
each test level at 0-hour, 48 hours and, for tests longer than 48
hours, at test termination. Replicate test containers should be
measured separately, except as explained above.
Acute Static Renewal: See the general discussion of replicates
above.
If a static renewal test is conducted, each test chamber'must be
sampled for chemical analysis at the 0-hour, at the end of the
first (or longest) cycle, and at test termination. It is
recommended that measurements be made at the end of each renewal
cycle.
Other factors not addressed in this guidance may preclude
conducting a static test even if the test material is stable
under test conditions conditions. These include, but are not
limited to, problems in maintaining dissolved oxygen levels,
feeding requirements, and concern for bacterial/microbial
contaminants. "' "
EFED certainly considers it "possible" to measure
concentrations of toxicants. Increased expense does not make it
.impossible.
-------
Acute Flow-Through:
above.
-8-
See 'the general discussion of replicates
If a flow-through test is conducted , each test concentration
must be measured at the 0-hour and at test termination. It is
recommended that for 96-hour tests, an intermediate measurement
be made at 48 hours to verify mid-test exposure if variability is
expected to be a problem.
Chronic Testing
Chronic static Renewals
above.
See the general discussion of replicates
Concentrations must be measured at each test level at 0-hour, at
the end of the last renewal cycle (at test termination), and at
the beginning and end of an intervening cycle at least once per
week. The longest cycle in a sequence should be used! if
variable-cycle periods are employed.
Chronic Flow-Through:
above.
See the general discussion of replicates
In each concentration, measure at 0-hour, every 7 days, and at
test termination. At the beginning of a study, the exact flow
of the system, and water output at each splitter must be
documented. In addition, system flow must be metered and
monitored visually or mechanically on a daily basis (every 24
hours), and it is recommended that the system flow be metered and
monitored twice a day (approximately every 12 hours).
Measurement of test concentration is required each time metering
fluctuation or malfunction is detected or observed. A record of
the regular inspections must be maintained and provided with the
study report.
All acute aquatic algae and diatom tests must be conducted as
static. Flow-through and static renewal systems are not
recommended for these test, since they are conducted with
microscopic organisms that cannot be protected from loss when
renewing or draining water from the test containers. Static
tests for Iiemna aibba can be conducted, regardless of stability.
-------
-9-
TBBTB WITH POORLY 8QMTBLB MATERIALS
Existing OPP guidelines for aquatic toxicity tests require
that chemicals be tested up to a maximum dissolved concentration
of 100 ppm (mg/1) in an effort to obtain an LCw, or EC.,,. Current
w, .,,.
policy allows chemicals that are poorly soluble (solubility < 100
ppm) to be tested up to the maximum water solubility obtainable
for the given test conditions employed, provided that certain
prerequisites apply:
1) The technique used to maximize chemical dissolution in
the test media under standard conditions for the test
is justified. Consideration of the optimum technique
should include use of non-toxic solvents, saturation
(solubility) columns, sonication, minor adjustments to
environmental conditions (i.e., temperature, pH, etc.),
etc., as appropriate. Minor adjustments should not
extend outside the recommended range of conditions for
the specific test organism.
2) Testing with a more soluble formulation (e.g.,
emulsifiable concentrate), if one exists, is provided
in addition to testing with the technical-grade
material. Testing with a more soluble formulation will
not be required if it does not increase the solubility
by 2X.
3) Measured concentrations of test media at appropriate
intervals and from appropriate test chambers of, all
test levels are determined from centrifuged
supernatant.
Studies that involve radical changes in environmental test
conditions outside the recommended range of values for
temperature, salinity, pH, etc. will be considered on a
case-by-case basis.
Solubility is defined as the amount of chemical .retained in
•the supernatant of a conventionally centrifuged sample of test
anedia. This amount of test material is considered to represent a
conservative measure of the most bioavailable fraction which may
include some colloidal material not removed by centrifugation in
addition to the truly dissolved fraction. A condition or
mechanism is only considered to have increased solubility if the
increase is two times or greater.
-------
-10-
OTHER S88UES OP SPECIAL CONCERN
position on Solubili"ty Enhancement
Saturator columnss The use of saturation columns as an aid in
the dissolution of test material and in confirming maximum
solubility is recommended but not required for nonvolatile test
chemicals with test media solubilities of 10 ppm or less.
Methods for using these columns in aquatic toxicity tests can be
adapted from the methods established for their use in determining
water solubility under OECD's Column Elution Method (Ref. No.
105). Saturator columns will be considered to generate test
solutions for static studies and not for flow-through studies.
Furthermore, saturator columns for static studies need only be
considered if ..conventional techniques for dosing the water do not
result in water concentrations within 2X of the stated solubility
of the compound.
Emulsifiers and Formulation Testing: Testing with a more soluble
.formulation, if one exists and which may contain emulsifiers,
.dispersants, solubilizing agents, etc., is required for all
active ingredients subject to aquatic organism testing and having
a water solubility less than 100 ppm and less than an EC/LC50. A
defined EC/LC50 provides a greatly improved basis for risk
assessment.
Effect of-Temperature: Solubility is a function-of temperature
and is especially sensitive at the limits of solubility. It is
recognized that increases of as much as 10 degrees may affect the
solubility no more than a factor of two. However if test
solutions are close to saturation, small changes in temperature
may result in supersaturated solutions. In addition, control of
temperature is important because of its well-known effects on the
actual toxicity of the compound.
Centrifugation: Conventional centrifugation is required for all
test media where undissolved test material, precipitate,
flocculant, colloidal suspension, etc., is/are observed in the
test chambers or. where-the solubility and hence bioavailability
are in question. Filtration may be used instead of
centrifugation if the analytical method is validated over a range
of acceptable concentrations. For aquatic toxicity tests, EFED
dsfinss solubility as the extractable chemical in the centrifuged
or filtered supernatant.
Position on Nominal Versus Measured Concentrations
When a laboratory test design has been specifically modified
to accomodate the instability of test material or other factors
likely to cause variability in test concentrations, and the
design is judged adequate based on sufficient preliminary
-------
-11-
information*, then the study will not be rejected solely on the
grounds that measured concentrations varied by sore than 30% of
the nominal concentration. An increase in measured test
concentration of more than 30% from the nominal concentration
during the test will generally not result in rejection, provided
that the following conditions are met:
1. a reasonable and scientific •xplanation is given, and
the variability of results produced by the chemical
analysis method is adequately characterized.
2. all test containers exhibit a similar (but not
necessarily identical) shift ,
3. the variability of the measured concentrations is
acceptable,
4. a statistically valid endpoint can be derived from the
measured concentrations , and
5. the preliminary stability information is provided with
complete documentation and description of methods used
to derive such information.
When High Variability Cannot Be Avoided
In some cases,, high variability cannot be avoided because
the test concentrations are approaching the limit of detection or
because of unavoidable binding of the test material to the
chemical analysis apparatus. When the ratio of the highest
This assumes that the preliminary stability tests were
conducted under test conditions essentially identical to the
actual test conditions.
6 If concentrations in some containers go up substantially
(>30%) and test concentrations in other containers go down
substantially (>30%), they will not be considered to have
exhibited a similar shift. The most important criterion is that
test levels aust not experience a shift in "order." That is, the
highest test level should remain highest, the next should remain
second, etc. If orders are shifted, the test may be rejected,
since regression analysis would not yield statistically sound
median lethal concentrations and confidence limits.
7 Either an LC50, £C50, or that the LC50 or ECSO is greater than
100 ppm.
-------
-12-
concentration to the lowest measured concentration is expected to
vary by more than 1.5, the registrant is strongly advised to
justify an exception to this requirement in advance of conducting
the aquatic laboratory studies. This exception justification
should consist of:
1. documentation of the preliminary trials indicating this
problem;
2. "the specific steps that will be taken to reduce the
variation;
3. the fully developed chemical analysis method; and
4. the raw data, standards and chromatogram from a
representative analysis using the method. For each
chemistry method, the actual minimum detection level
and level of quantification must be identified.
EFED will decide on each exception justification on a
case-by-case basis. However, if a series of aquatic tests are to
be conducted with one chemical and it is anticipated that these
limits will be exceeded, one exception justification may cover
more than one study. EFED will then exercise judgment in
evaluating studies with test materials that are difficult to
measure.
Test Material--Decline in Algae and Diatom Tests
Conducting flow-through or static renewal tests with aquatic
algae and diatoms is not feasible with the current state of the
•practice. Therefore, the following is recommended for a test
material that, based on preliminary stability testing, is
expected to degrade to less than 70% of the nominal
concentration. The study should be conducted normally, with
concentrations measured at 0-hour and at test termination.
Although it is undesirable to allow the concentrations to decline
throughout the study, the problem may be unavoidable. In this
case, the LC50 regression analysis is based on 0-hour test levels.
For purposes of consistency, the aquatic test with a
vascular plant (Lemna aibba) need not be done using a flow-
through or static renewal system with the sole purpose of
maintaining test concentrations. There may be other reasons for
conducting a static renewal study.
-------
Appendix
Why
Measured Concentrations
of Nominal
concentrations?
Where measured concentrations are indicated, they are
considered necessary because:
1. There are concerns that the actual concentrations to
which the test organisms are exposed may be different than
"nominal." This variation may be due to chemical characteristics,
test conditions or mechanical apparatus.
2. Measured concentrations confirm that the test system was
designed appropriately and is operating acceptably.
Characteristics that make testing difficult (low solubility,
short half-life, high binding potential, etc.) must be accounted
for in the exposure estimates (see below). They are not a reason
for developing misleading toxicity values from laboratory tests.
Measurement of test concentrations is not done -lust to
determine if the technician knows how to mix the test solution
once. Among other things, it also ensures that the test solution
was mixed correctly each time: It corroborates the precision of
the technician or mechanics of the test system.
If test levels are not measured, the nominal values are used
to calculate the LC50, EC50f NOEL and 1OEL. If the test material
has degraded or has become unavailable because of insolubility or
sorption, the pesticide may be characterized as less toxic than
it really is. For example, if based on nominal test levels, the
LC50 is 5 ppm, the pesticide would be considered moderately
toxic. No higher tier testing would be required and that value (5
ppm) would be the basis for developing concern levels with which
to compare EECs. But if, in reality, the concentrations to which
the organism was actually exposed were only between 0.1 and 1
ppm, the LC50 may well be closer 0.5 ppm. This would result in
labeling, and could trigger higher tier tests. More importantly,
it would yield substantially lower concern levels with which to
compare exposure levels.
WhY Vse Measured Concentrations When They Are Available?
Measured concentrations are used when they are available
because they indicate what the exposure was in the test chambers.
Why Pe Concerned with Pesticides That Are Difficult to Test?
Presumably, a safer pesticide is one that that may degrade
rapidly, has low solubility, and is used at low rates. While
these characteristics may result in lower exposure levels in the
field, the risk they represent can only be determined if the
-------
-2-
actual toxicity of the pesticide is known8. When potentially
.low, realistic exposure levels are calculated and used for risk
assessments, it is imperative that the actual toxicity of the
pesticide at those levels be determined. If the test were
conducted using nominal concentrations, the results could reflect
a higher apparent effect concentration (e.g. LC$0, ECj0 or NOEL,
as discussed above). As a result, potential risk may be missed
because the comparison would be between a low "realistic"
exposure and a high nominal test level that was not the true
toxicity level. A risk assessment based on such a comparison and
data would be faulty and could not be scientifically defended.
Pesticides that are used at very low levels tend to have
high biological activity. For this reason, it is imperative that
the toxicity data developed for these pesticides be accurate and
scientifically defensible.
Why Analyze at Test Termination?
Where indicated, measurement at test termination is
considered necessary to determine if the test organisms were
exposed to the test material throughout the entire study and at
what levels. A significant change in test concentration during
the last part of the study may substantially alter the results.
For example, if the test concentration dropped dramatically
during the last few days of a study, the effects that may have
been caused by such exposure may not occur. The EC50.or LCe0
developed from that study would be midleading if it is called a
"96-hour LC50".
Test endpoints are used as if the organisms were exposed to
the test material at the statistically developed value (LC50 or
EC50) for the entire test duration. Granted, one aspect of the
risk assessment is to compare concern levels based on LC50,s to
initial immediate concentrations. However, it is assumed that
field conditions exist in which concentrations that may be of
acute concern may last longer or occur frequently enough to be
comparable to the 48-hour, 96-hour or 120-hour test duration.
Even though a pesticide may degrade rapidly under one condition
(in water, for example), the possibility of repeated exposure
needs to be considered. Repeated loadings from reservoirs of the
active ingredient, occurring in compartments where persistence is
greater, say occur. Without taking into account these
eventualities, EFED would not be able to generate risk
assessments that adequately protect the aquatic ecosystem.
8 Or at least know with high certainty the level below which the
pesticide is not likely to result in 50% mortality (i.e., an
> X concentration" situation.
-------
-3-
tP Achieve Current Test Standards
The costs of developing chemical analysis methods,
S*?^?1 analvsi? *? Bost toxicity tests, conducting
3N J*? »»d PFcliffiinary trials, and measuring the
x in^r?dien* in bfttch supplies of the test material should
not be considered an extra cost. These activities are part of
the overall test procedure, and have been included in our cost
estimates •
n w?h?-.GLP ^Wires that batch mixes be analyzed and that
solubility and stability be known. Both require methods
tSFEFSSt^. *h?r+f°ff ' **• «?•* of ffiet«ods development should
not be factored into the additional chemical analysis that may be
recommended to do scientifically sound and acceptable studies for
pesticide risk assessment.
nM,* *Sti?a?5* ProTided by NACA *i* not include cases in
which the pesticide active ingredient does not cause mortality at
maximum practical solubility (see solubility discussion)! In
these cases, only one treatment level is required with 30 test
organisms, if it can be shown that no mortality occurred
throughout the test period, the LC50 is assumed to be greater
Th^Sj +£°JCJ^ration and no nore testin9 would be required.
The fact that this may occur should be known from range-finding
tests. In such cases, the cost for chemical analysis would be
reduced dramatically because there would be fewer samples to
analyze. Adequate preliminary trials to determine the behavior
of the test material under test conditions could avoid
unnecessary sampling and analyses.
-------
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
OCT 29 19S7.
MEMORANDUM
OFFICE OF
PESTICIDES AND TOXIC
SUBSTANCES
SUBJECTS Reevaluation of Previously Rejected
Mollusk Shell Deposition Studies 72-i[b]
FROM:
TO:
Paul P. Schuda, Deputy Director
Environmental Fate and Effects
Vision H7507C
Daniel Barolo, Director
Special Review and Reregistration Division H7508W
Lawrence E. Culleen, Acting Director
Registration Division H7505C
SUMMARY
Mollusk shell deposition studies (72-3[b]) that were
rejected because of inadequate amount of shell growth in control
organisms will be re-examined to determine if the study has to be
repeated. The need-for a certain minimum amount of shell growth
in the control mollusks is a relatively new requirement, of which
the registrants have not been officially notified. This
memorandum presents the criteria for when the EFED will, or will
not recommend that the rejected mollusk studies be repeated. It
also explains what the new criteria are and how they will be
promulgated.
BACKGROUND
One of the 3 studies used by EFED to characterize potential
toxicity of a chemical to estuarine and marine species is the 96-
hour mollusk shell deposition study. This study is required for
a pesticide when significant exposure in estuaries or marine
habitats is expected. This study depends on a reduction in the
shell-growth of mollusks in response to exposure to various
concentrations of a pesticide compared to the shell-growth of
control aollusks in clean water. In the 72-3 guidelines and in
the Standard Evaluation Procedure (SEP) for this study, no
minimum amount of shell growth was established for the control
mollusks (those exposed to clean water). It was stated that 1 mm
per day could be expected from healthy mollusks, but that was not
established as a criteria for rejection.
Printed on Recycled Piper
-------
Since the SEP's were published, the EFED scientists have
received information from experts in the field of estuarine
organism testing indicating that there should be a minimum amount
of shell-growth that occurs in the control to make the study
acceptable. With inadequate shell-growth" of the control
organisms, these experts believe there is a likelihood that
reductions in the shell-growth of the mollusks exposed to the
pesticide may not be detectable and the resulting EC50 developed
from such a study may not reliably indicate toxicity. In other
words, the poor performance of the control mollusks could mask
affects on shell growth caused by the pesticide. The consensus
of the experts is that 2 mm of shell growth in a 96-hour period
is easily obtainable and is the minimum they would allow for a
valid comparison with responses from mollusks exposed to
pesticides.
PROPOSAL
The EFED is proposing to implement the requirement that, for
a mollusk shell deposition study to be acceptable, the organisms
in the clean water control must produce a minimum average of 2 mm
of shell growth during the 96 hours. The EFED will present this
proposal to the Science Advisory Panel (SAP). With the comments
from SAP, the EFED will formally amend the SEP and notify the
registrants of this change.
IMPLEMENTATION
The EEB has begun incorporating this change into its
evaluation criteria and has, in the past few years, rejected
several studies solely because the average shell growth in the
clean water control was less than 2 mm during the 96-hour test.
However, in light of the fact that this change was not formally
established and communicated to RD and SRRD, and the registrants,
the EEB will reconsider the category of these rejected studies
and evaluate them under the following interim procedures.
Old studies, and future studies begun before the proposed
formal notification takes effect, will be evaluated in the
following way. If they were rejected because shell growth in the
clean water control is less than 2 mm, and that is the primary
problem with the study (i.e. no other deficiencies were noted
that, by themselves, would cause the study to be unacceptable ),
the study will be called supplemental (i.e. not fulfilling
guideline requirements). Then a value of data assessment will be
made. The results of the mollusk study will be considered with
other available freshwater and estuarine organism test data to
see if mollusks seem more or less sensitive to that pesticide.
If mollusks do not appear to be more sensitive, EFED will not
recommend that the study be repeated. The toxicity values for
If another deficiency caused the study to be unacceptable
this interim procedure will not apply.
-------
the more sensitive species will be used for characterizing risk
to freshwater or estuarine organisms.
If, for a given pesticide, it is determined that the mollusk
may be more sensitive than any other freshwater*or aquatic
organism, the EFED will then appraise the hazard represented by
the pesticide-and determine the level of concern it represents.
The study may still not have to be repeated for a pesticide that
has a low potential for adverse effects, even if the mollusk was
the most sensitive.
If the mollusk is the most sensitive, and if there appear to
be risks to aquatic or estuarine organisms, the EFED will
maintain the position that the mollusk study should be repeated.
This is because of the possibility that risk would be based on
test results that do not adequately characterize hazard.
Mollusk shell deposition studies initiated after the
modification takes effect and is distributed to the registrants
will be evaluated based on the new criteria. At that time, the
EFED will recommend those not meeting the criteria be repeated.
BENEFIT OF THIS INTERIM SOLUTION
This proposed interim procedure will likely permit OPP to
reduce substantially the number of mollusk studies that need to
be repeated. In a review of a random group of pesticides
where data from mollusks and other estuarine and freshwater
organisms were available, the mollusk was usually not the most
sensitive.
IDENTIFYING REJECTED MOLLUSK STUDIES
t The EFED is requesting that the product managers or chemical
review managers identify mollusk studies that have been rejected
solely because of problems with shell growth. The EEB scientists
will determine which of those studies do not need to be repeated.
An alternative is to submit this proposal to the registrants that
have done mollusk studies and allow them to identify the studies
they feel were rejected based on this criterion.
If you have questions, please contact Douglas Urban, Acting
Chief, EEB or Daniel Rieder (305-5314).
cc: Stephanie Irene
Peter Caulkins
-------
-------
Rapid Feedback SOP
187
-------
Appendix B: List of EPA Guidance Documents
The following is a list of guidance documents that outline procedures for conducting
ecological effects studies. Specific references to these materials are made under each of the
rejection factors listed.
Hazard Evaluation Division, Standard Evaluation Procedures:
Acute Toxicity Test for Estuarine/Marine Organisms (1985).
Acute Toxicity Test for Freshwater Fish (1985).
Acute Toxicity Test for Freshwater Invertebrates (1985).
Avian Dietary LC50 Test.
Daphnia Life Cycle Chronic Toxicity Test (1986).
Ecological Risk Assessment (1986).
Pesticide Assessment Guidelines:
Subdivision E: Hazard Evaluation, Wildlife and Aquatic Organisms (1982).
Subdivision J: Hazard Evaluation, Nontarget Plants (1982).
Subdivision L: Hazard Evaluation, Nontarget Insects (1982).
Guidance Document for Conducting Terrestrial Field Studies (1988).
EEFRA Accelerated Reregistration Phase 3 Technical Guidance (1989).
Stephan, Charles E., "Committee on Methods for Toxicity Tests with Aquatic
Organisms." 1975. Methods for Acute Toxicity Tests with Fish, Macroinvertebrates, and
Amphibians. US EPA, Ecol. Res. Series.
ASTM Standard E, American Society for Testing and Materials, 1916 Race Street,
Philadelphia, PA.
188
-------
Standard Operating Procedure for Obtaining Rapid Feedback from EPA
on Technical Issues Associated with Ecological and Environmental Fate Studies
Background:
Rejection rates for Ecological Effects and Environmental Fate studies may be reduced if
feedback from the Agency on specific technical issues was obtained prior to initialing the
studies. Given the strict time-lines associated with reregistration, it can be difficult to obtain
such definitive feedback in a short time period. This limitation was identified by both the
Agency and Registrants and it was determined that a specific mechanism should be
established for communication.
Procedure:
When registrants have specific questions concerning the technical conduct of studies, the
following procedure should be used to obtain guidance from the Agency:
1. For minor problems requiring rapid response from EPA, the registrant should call
the Branch Chief or Deputy Branch Chief of EEB or the Branch Chief of EFGWB as
appropriate for the problem.
2. The Branch Chief will determine if the problem is minor and urgent, and if so, will
assign it to the appropriate Section Head or form an ad hoc team. Minor problems will
be defined as those which can be solved within one week, with less than approximately
one staff hour of work.
3. To facilitate discussion, it may be best to send data or information to the Branch
Chief or appropriate Section Head by FAX or mail before calling. When appropriate, it
is assumed that the registrants will perform adequate preliminary research for the
compound and test system of concern and, if necessary, have this data available to
support any position they are defending.
-------
The current Section Heads (SH) and Branch Chiefs (BC) phone numbers are:
Ecological
Effects
Phone No.
703-305-
Environmental
Fate
Phone No.
703-305-
N. Cook (SH) 5322
H. Craven (SH) 5320
D Rieder (SH) 5314
A. Stavola (SH) 5354
L. Touart (SH) 6134
D. Urban (Deputy BC) 5746
T. Maciorowski (BC) 7347
P. Mastradone (SH) 5335
M. Shamim (SH) 5025
A. Abramovitch (SH) 5975
H. Behl (SH) 6128
H. Nelson (SH) 7356
H. Jacoby (BC) 5734
4. EPA's decision on the problem raised by the registrant will be communicated back
to the registrant by telephone.
5. A memorandum to the Branch chemical file will be prepared and the registrant will
immediately (within one week) follow up the telephone communication with a submission
through the appropriate registration division (RD or SRRD) documenting the request and
decision. A written response will be made to that submission.
6. When the final study report is submitted to the EPA, it should reiterate any agreed-
upon decisions and document when they were requested and granted.
If the request is not considered minor and cannot be decided within one week with less than
approximately one staff hour of work, the registrant will be informed as soon as practical
(within one week). If the registrants wish to pursue the matter, they must submit their
request with necessary supporting studies and explanation through the appropriate division
(RD or SRRD). They may, if necessary, submit a request for a time extension.
Relative to field or monitoring studies, every effort will be made by the Agency to
accommodate questions and concerns of the registrants.
When contacting EPA for advice, questions should only relate to technical issues. Questions
regarding the status of reviews, etc., should be addressed through the product managers in
RD or chemical review managers in SRRD. Contacting reviewers directly is discouraged.
-------
/5«*#- 'V^
>
aiXlyzrS
s§.|ll§
£a*j = i
i
^
4
-7
/
r-
D>
E
3
0
^
fi>
OQ
m
o
n>
i
O
coO
xo
>0
z>
or-
m
-------
CO
LL1
DC
CD
IELLJ
OH
ZDC
CO
I-
O
o
-------
tlia
52
S.J8
-
'
r-' oi
1<
•
l
-------
i
-------
Sl-iS
CO
-------
5
Is
it
o
I
I
-------
A
s- §
Q_ 5"
V
if H-ef
p.! Ill
S^NCO* '
" " 1
I
g
^
in
5.2-g
SE
•Si
-------
m
t
?
I?
------- |