[Federal Register Volume 60, Number 114 (Wednesday, June 14, 1995)]
[Notices]
[Pages 31300-31308]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 95-14683]
-----------------------------------------------------------------------
ENVIRONMENTAL PROTECTION AGENCY
[OPP-260055; FRL-4944-2]
Pesticide Tolerances; Partial Response to Petition to Modify EPA
Policy
AGENCY: Environmental Protection Agency (EPA).
ACTION: Notice; Response to Petition.
-----------------------------------------------------------------------
SUMMARY: This notice responds in part to a petition filed with EPA by
the
[[Page 31301]]
National Food Processors Association and other food and grower trade
associations. That petition sought the repeal or revision of several
EPA policies and interpretations related to how EPA coordinated actions
under its various statutory authorities over pesticide residues in
food. EPA regulates pesticides under the Federal Insecticide,
Fungicide, and Rodenticide Act and sections 408 and 409 of the Federal
Food, Drug, and Cosmetic Act. Although EPA has not resolved all of the
policy questions raised by the NFPA petition, EPA has concluded that
changes are warranted to its policy concerning when FFDCA section 409
is applicable to a pesticide use and several related legal
interpretations.
FOR FURTHER INFORMATION CONTACT: By mail: Niloufar Nazmi, Special
Review and Reregistration Division (7508W) or Jean Frane, Policy and
Special Projects Staff (7501C), Environmental Protection Agency, 401 M
St., SW., Washington, DC 20460. Telephone numbers: 703-308-8028 or 703-
305-5944; e-mail: nazmi.niloufar@epamail.epa.gov. or
frane.jean@epamail.epa.gov.
SUPPLEMENTARY INFORMATION:
Table of Contents
I. Introduction
II. Background
A. Statutory Background
B. EPA Coordination of the Statutory Provisions Governing Pesticides
III. The NFPA Petition
IV. Summary of EPA's Partial Response to NFPA Petition
V. Concentration Policy
A. General Issues
B. Monitoring Data and the Concentration Policy
C. Revisions to the Concentration Policy
1. Introduction and summary.
2. Factors relied upon by EPA in determining whether a pesticide
which concentrates in fact is likely to produce residues in
exceedance of the section 408 tolerance.
3. Other factors potentially relevant to whether residues exceed
the section 408 tolerance.
4. Evaluation of factors.
5. Conclusion.
VI. Ready to Eat
A. NFPA's Argument and Views of Commenters
B. EPA's Response
1. The definitional issue.
2. Enforcement approach.
3. Animal feeds.
4. Future actions.
VII. Are EPA's Policies Rules That Have Not Been Properly
Promulgated?
I. Introduction
In Les v. Reilly, 968 F.2d 985 (9th Cir. 1992), cert. denied, 113
S.Ct. 1361 (1993), the Ninth Circuit U.S. Court of Appeals held that
the Delaney anti-cancer clause in the food additives provision of the
Federal Food, Drug, and Cosmetic Act was not subject to an exception
for pesticide uses which pose a de minimis cancer risk. Prior to the
decision becoming final, food processors and growers filed a petition
with EPA challenging a number of policies and interpretations relating
to how EPA implements its authority under the FFDCA. The petition
proposes policies and interpretations that would reduce the impact of
the Les decision. This notice responds to the petition in part.
II. Background
A. Statutory Background
Pesticide residues in human and animal food in the United States
are regulated under provisions of the Federal Food, Drug and Cosmetic
Act (FFDCA) and the Federal Insecticide, Fungicide and Rodenticide Act
(FIFRA). The interplay between sections 402, 408 and 409 of the FFDCA
and, to a more limited extent, between the FFDCA and FIFRA, have
created a complex, and sometimes contradictory, statutory framework
underlying residue regulation in food.
Before a pesticide may be sold or distributed, it must be
registered under the FIFRA. 7 U.S.C. 136 et seq. To qualify for
registration, a pesticide must, among other things, perform its
intended function without causing ``unreasonable adverse effects on the
environment.'' 7 U.S.C. 136a(c)(5). The term ``unreasonable adverse
affects on the environment'' is defined as ``any unreasonable risk to
man or the environment taking into account the economic, social and
environmental costs and benefits of the use of any pesticide.'' 7
U.S.C. 136(bb).
The FFDCA, 21 U.S.C. 301 et seq., authorizes the establishment by
regulation of maximum permissible levels of pesticides in foods. Such
regulations are commonly referred to as ``tolerances.'' Without such a
tolerance or an exemption from the requirement of a tolerance, a food
containing a pesticide residue is ``adulterated'' under section 402 of
the FFDCA and may not be legally moved in interstate commerce. 21
U.S.C. 331, 342. EPA was authorized to establish pesticide tolerances
under Reorganization Plan No. 3 of 1970. 5 U.S.C. App at 1343 (1988).
Monitoring and enforcement of pesticide tolerances are carried out by
the U.S. Food and Drug Administration (FDA) and the United States
Department of Agriculture (USDA).
The FFDCA has separate provisions for tolerances for pesticide
residues on raw agricultural commodities (RACs) and for residues on
processed food. For pesticide residues in or on RACs, EPA establishes
tolerances, or exemptions from tolerances when appropriate, under
section 408. 21 U.S.C. 346a. EPA regulates pesticide residues in
processed foods under section 409 which pertains to ``food additives.''
21 U.S.C. 348. Maximum residue regulations established under section
409 are commonly referred to as food additive tolerances or food
additive regulations (FARs). Section 409 FARs are needed, however, only
for certain pesticide residues in processed food. Under section
402(a)(2) of the FFDCA, a pesticide residue in processed food generally
will not render the food adulterated if the residue results from
application of the pesticide to a RAC and the residue in the processed
food when ``ready to eat'' is below the RAC tolerance set under section
408. This exemption in section 402(a)(2) is commonly referred to as the
``flow-through'' provision because it allows the section 408 raw food
tolerance to flow through to the processed food form. Thus, a section
409 FAR is only necessary to prevent foods from being deemed
adulterated when the concentration of the pesticide residue in a
processed food when ``ready to eat'' is greater than the tolerance
prescribed for the RAC, or if the processed food itself is treated or
comes in contact with a pesticide.
To establish a tolerance regulation under section 408, EPA must
find that the regulation would ``protect the public health.'' 21 U.S.C.
346a(b). In reaching this determination, EPA is directed to consider,
among other things, the ``necessity for the production of an adequate,
wholesome, and economical food supply.'' Id. Prior to establishing a
food additive tolerance under section 409, EPA must determine that the
``proposed use of the food additive [pesticide], under the conditions
of use to be specified in the regulation, will be safe.'' 21 U.S.C.
348(c)(3). Section 409 specifically addresses the safety of
carcinogenic substances in the so-called Delaney clause which provides
that ``no additive shall be deemed safe if it has been found to induce
cancer when ingested by man or animal or if it is found, after tests
which are appropriate for the evaluation of the safety of food
additives, to induce cancer in man or animal * * *.'' Id. Although EPA
has interpreted the general standard under section 408 to require a
balancing of risks and benefits, where a pesticide which is an animal
or human carcinogen is involved, the section 409
[[Page 31302]]
Delaney clause, in contrast to section 408 and FIFRA, explicitly bars
such balancing no matter how infinitesimal the potential human cancer
risk. Les v. Reilly, 968 F.2d at 989.
B. EPA Coordination of the Statutory Provisions Governing Pesticides
In its administration of FIFRA and FFDCA sections 408 and 409, EPA
has specified that FIFRA registrations for food-use pesticides will not
be approved until all necessary tolerances and food additive tolerances
have been obtained. 40 CFR 152.112(g). As a policy matter, EPA has
taken a similar approach to FFDCA sections 408 and 409, not granting
section 408 tolerances until needed section 409 FARs have been granted.
This linkage of its statutory authorities has been described by EPA
as its coordination policy. Basically, EPA's coordination policy is an
expression of EPA's intent to take into account all of the applicable
provisions governing pesticides in taking action under any one of the
three. EPA's view has been that it should not be approving pesticide
uses under one of the three provisions if an approval needed under one
of the other provisions cannot be obtained.
EPA's concentration policy establishes the criterion as to when
approval is needed for food-use pesticides under FFDCA section 409, and
hence when the Delaney clause applies. Generally, EPA has used a
``concentration in fact'' standard as the test of whether a use needs a
section 409 FAR. The concentration in fact standard focuses on the
level of the pesticide residue in the processed food, measured on a
weight to weight basis, compared to the level of the residue in the
precursor raw agricultural commodity. If a processing study shows that
the level of pesticide residue in the processed food exceeds the level
of residue in the precursor raw agricultural commodity, EPA would
conclude there has been a concentration in fact of the pesticide
residues in the processed food.
EPA believes the concentration in fact test is relevant to the
inquiry of whether a section 409 FAR is needed because residues in the
raw crop may be at or near the section 408 tolerance level. Residues in
the raw crop may be close to the section 408 tolerance level because
section 408 tolerance levels are established based on actual field
trials and designed to be set no higher than necessary given approved
usage directions for the pesticide established in the FIFRA
registration. Under EPA regulations, the section 408 tolerance level
should ``reasonably reflect the amount of residue likely to result when
the pesticide chemical is used in the manner proposed.'' 40 CFR 180.4.
If residue levels in the raw crop are at or near the section 408
tolerance level and concentration in fact occurs during processing, the
residue level in the processed food is likely to exceed the section 408
tolerance. The National Academy of Sciences (NAS) has acknowledged the
logic behind EPA's reliance on a concentration in fact standard:
In determining whether a section 409 food additive tolerance is
required, the EPA focuses on whether residues in any processed
product exceed those found on the unprocessed crop, not whether
residues concentrate above some hypothetical section 408 tolerance.
The logic of the EPA's practice is clear. A section 408
tolerance represents a residue level that may in some cases be
realized. A section 409 tolerance must reflect the possible residue
levels in processed foods derived from that raw commodity.
National Research Council, Regulating Pesticides in Food: Delaney
Paradox 28 (1987).
III. The NFPA Petition
On September 11, 1992, the National Food Processors Association
(NFPA), the United Fresh Fruit and Vegetable Association, the Florida
Fruit and Vegetable Association, the Northwest Horticultural Council,
and the Western Growers Association filed a petition with EPA
challenging the policies followed by EPA in linking its regulatory
activities under the various pesticide provisions of FIFRA and FFDCA.
(Petition to the Environmental Protection Agency, Office of Pesticide
Programs, Concerning EPA's Pesticide Concentration Policy (1992))
(hereinafter cited as ``NFPA petition''). The NFPA petition explicitly
attacks what it calls EPA's ``concentration policy.'' In actuality, the
petition is a challenge to two interrelated policies described by EPA
as its coordination and concentration policies. The NFPA petition
argues that the coordination and concentration policies are both
unlawful and unnecessary. The petition requests that the EPA
coordination policy be repealed so that section 408 tolerances can
remain in effect (or can be established) for pesticide uses even if,
under the Les decision, the associated section 409 FARs have to be
revoked (or cannot be established). The petition asks that the
concentration policy be modified so that it takes into account factors
beyond the concentration in fact test. Additionally, the petition
requests that EPA apply the term ``ready to eat'' in the flow-through
provision according to what NFPA asserts is its plain meaning.
EPA sought public comment on the petition (58 FR 7470, Feb. 5,
1993). Extensive public comment was received, and significant comments
are discussed in this notice. Several more narrowly focused comments
are discussed in a separate document that has been included in the
docket.
IV. Summary of EPA's Partial Response to NFPA Petition
Sections V through VII below set forth EPA's partial response to
the NFPA petition. EPA has not reached a decision on NFPA's challenge
to the coordination policy. EPA, however, has completed evaluation of
NFPA's contentions regarding the concentration policy and EPA's
interpretation of the term ``ready to eat.'' This document responds to
the NFPA petition on these two issues. In brief, EPA agrees with NFPA
and many of the commenters that modifications should be made to its
concentration policy so that it is a better predictor of the likelihood
that residues in processed food may exceed the applicable section 408
tolerance. EPA, however, cannot accept all of NFPA's suggested changes
to the concentration policy. As to interpretation of the phrase ``ready
to eat,'' EPA agrees that such term must be given its common-sense
meaning.
V. Concentration Policy
A. General Issues
EPA's concentration policy is the trigger for when a pesticide use
needs a section 409 FAR. EPA has treated a pesticide use as needing a
section 409 FAR generally whenever a processing study shows that
pesticide residues are greater in the processed food than in the raw
agricultural commodity before processing. In other words, EPA looks to
see if the pesticide ``concentrates in fact.'' EPA has used
concentration in fact as the trigger for when a food additive
regulation is needed because, in theory, RAC tolerances are set at
levels no higher than necessary to cover maximum legal usage under the
FIFRA registration. RAC tolerances are established based on field trial
data showing the range of residues likely to result from maximum legal
application of the pesticide. Generally, the RAC tolerance level is set
just slightly above the maximum residue value found in the field
trials. Thus, if concentration in fact occurs during processing,
overtolerance residues in processed food can result if the RACs used
for processing contain pesticide residues reflecting maximum legal
usage.
[[Page 31303]]
NFPA challenges EPA's concentration policy on two grounds. First,
NFPA claims that all available data support the view that food additive
regulations are unnecessary to avoid adulterated processed food.
Second, NFPA argues that EPA has ignored the ``ready to eat''
requirement in the flow-through provision. EPA's interpretation of the
term ``ready to eat'' will be addressed in the following section.
B. Monitoring Data and the Concentration Policy
NFPA cites various data sources which it claims show residues on
both raw and processed foods generally to be well below the level of
the RAC tolerance. NFPA argues that residues in processed foods
generally fall below RAC tolerances because of the careful attention
paid to the flow-through provision by food processors.
When the flow-through provision was adopted and as it operated
for a number of years, processors clearly understood that it was
their obligation to produce a processed product that stayed within
the raw product tolerance. This obligation could be met through any
number of steps, including supervision of growers' pesticide
practices, careful and informed buying practices, analysis of raw
product, handling, cleaning and treatment of the raw product, and
testing of the finished produce to assure that it would be in
compliance with the Act * * *. [T]hey recognized that if their
process involved some degree of concentration [and the food is
consumed in the concentrated form], they were well advised to use
raw product that at the time of processing was below the prescribed
tolerance levels, and that failure to take such steps could possibly
result in adulteration and a costly enforcement action.
(Comments of NFPA at 37-38).
NFPA asserts that the steps taken by processors to avoid
overtolerance residues show that EPA's reliance on processing studies
to require food additive regulations is unwarranted.
The data relied upon by NFPA do show that pesticide residues in raw
and processed food generally are below section 408 tolerance levels. On
the other hand, EPA is often presented with processing studies by
pesticide manufacturers that demonstrate that particular pesticides
concentrate in processed food to levels 2 times, 10 times, or even 50
times above the level found in the raw crop. EPA has examined carefully
the factors cited by NFPA and commenters as an explanation for the low
levels of residues to determine whether any adjustments to the
concentration policy are appropriate. Although EPA has concluded that
some adjustment to the concentration policy is warranted, EPA believes
that the basic rationale of the concentration policy with its focus on
concentration in fact is sound. As the National Academy of Sciences has
found:
The logic of EPA's practice is clear. A section 408 tolerance
represents a residue level that may in some cases be realized. A
section 409 tolerance must reflect the possible residue levels in
processed foods derived from that commodity.
National Research Council, Regulating Pesticides in Food: Delaney
Paradox 28 (1987)
At the same time, EPA recognizes that reliance solely on processing
studies may not, in some circumstances, accurately ``reflect the
possible residue levels in processed foods.''
In challenging the concentration policy, some commenters argue that
EPA's policy is a theoretical exercise with no basis on actual data and
that this is confirmed by EPA's description of its policy in its
request for comment on the NFPA petition. EPA did not mean to suggest
in that notice that its concentration policy focuses on theoretical
possibilities. EPA's policy has always sought to determine whether
residues greater than the section 408 tolerance can occur in processed
food. EPA makes this determination based on hard data-- actual
processing studies involving, in most cases, the pesticide and crop in
question. EPA's revisions to its policy do not change the basic focus
of the concentration policy. Rather, as explained below, EPA has
expanded the range of data and other information it will consider in
determining whether residues greater than the section 408 tolerance can
occur in processed food.
It is worth noting that the same data relied upon by NFPA to show
that most food, whether raw or processed, is well below section 408
tolerance levels also reinforces EPA's judgment that many section 408
tolerances may currently be set higher than necessary and may need to
be lowered so that they reasonably reflect actual residues. If section
408 tolerances are lowered, the chances of residues over the section
408 tolerance in processed foods where residues concentrate in fact
would be greater.
C. Revisions to the Concentration Policy
1. Introduction and summary. EPA's concentration policy is designed
to evaluate when residues in processed food may exceed the raw food
tolerance due to concentration during processing. Generally, in
implementing its concentration policy, EPA has used a test of
concentration in fact as an indicator that residues over the section
408 tolerance may occur because residue levels in the RAC may exist at
the tolerance level. EPA, however, also has historically considered, to
a limited extent, at least two other factors in evaluating whether a
processing study showing concentration of residues indicates there is a
real possibility of residues over the section 408 tolerance. Below, EPA
discusses those factors and other factors that may prevent the
occurrence of residues over the section 408 tolerance.
EPA concludes that it has too rigidly applied its concentration in
fact test. EPA continues to believe that information from processing
studies is generally the most important single piece of information is
assessing the likelihood that residues in processed food could exceed
the section 408 tolerance. EPA will also continue to consider factors
such as the variability of the analytical method and the degree of
rounding used in establishing the section 408 tolerance. In a departure
from past practice, EPA will, as explained below, take into account,
where appropriate, information pertaining to the averaging of residues
during processing. EPA will also, where appropriate, consider
information obtained from properly designed market basket surveys. EPA,
however, is not convinced at this time by the NFPA suggestion that,
despite data showing residues concentrate during processing, processors
can insure residue levels stay below section 408 tolerance levels.
2. Factors relied upon by EPA in determining whether a pesticide
which concentrates in fact is likely to produce residues in exceedance
of the section 408 tolerance. As noted, EPA follows a concentration in
fact test to determine if section 409 FARs are necessary. For the most
part, EPA's concentration in fact test is applied based on the results
from data from processing studies. Historically, EPA has also
occasionally considered two other factors in determining whether a
processing study which shows concentration in fact does show that
residues in processed food can exceed the appropriate section 408
tolerance.
The first of these factors is the degree of rounding that was used
in setting the RAC tolerance. To a limited extent, EPA has considered
the degree of rounding in past decisions on whether a section 409 FAR
is needed. Generally, the highest value obtained from field trials is
rounded up in selecting the tolerance level. For example, if the
highest value from field trials was 8 parts per million (ppm), that
data point might be rounded to 10 ppm for the tolerance value. Where
rounding increases the observed residue level by 25 percent, the
pesticide would have to concentrate by
[[Page 31304]]
a factor of greater than 25 percent (1.25X) to produce residues over
the section 408 tolerance.
The second factor currently relied upon by EPA is the degree of
variability in the analytical method used to measure residue levels in
the field and processing studies and for enforcement of the tolerance.
If residues do not concentrate to a greater degree than the variability
in the methods, no residues over the section 408 tolerance could be
reliably detected.
3. Other factors potentially relevant to whether residues exceed
the section 408 tolerance. In the past, EPA has generally not taken
into consideration various other factors that may explain why, despite
the fact that a processing study suggests there is a possibility of
residues greater than the RAC tolerance, that event seems to occur
infrequently. One factor that lessens the possibility of residues over
the section 408 tolerance in processed food is that EPA's judgment
concerning whether such residues could occur assumes that the pesticide
will be used at the maximum label rate and applied the maximum number
of times permitted, and that the crop will be harvested at the shortest
preharvest interval allowed. Frequently, however, these maximum
application and harvest practices are not followed resulting in
residues far below tolerance levels in the raw crop, with
correspondingly lower levels in the processed food.
A second factor that serves to result in lower residue levels is
that tolerance values are set to reflect the maximum residue level that
could result from maximum legal application and harvest practices but
field trials generally show a wide range of residue levels even when
maximum legal application and harvest practices used in each trial.
Thus, average residue values from such field trials tend generally to
be significantly below the maximum residue level found in field trials
and, thus, also significantly below the tolerance level.
A third factor that may explain lower observed residues in
processed foods is that the processing of many crops involves mixing or
blending of large amounts of the raw crop. Oftentimes this can result
in significant lowering of residue values as untreated crop is blended
with treated crop. Further, this blending accentuates the above two
factors as lightly treated crops are mixed with crops having received
maximum treatment and high and low level residues from crops receiving
maximum treatment are mixed.
Another reason why residues over the section 408 tolerance may not
occur in processed food is that pesticides often degrade significantly
during the time in which the crop is transported and stored prior to
processing. Thus, even if crops bearing tolerance level residues at
harvest were the only ingredient used in food processing, any
concentration of residues might be offset by normal degradation of
residues.
NFPA suggests additionally that the chance of residues over the
section 408 tolerance is not great because of various steps taken by
food processors. NFPA cites ``supervision of growers' pesticide
practices, careful and informed buying practices, [and] analysis of raw
product'' as actions which serve to reduce residues. Further, various
commenters have contended that residues over the section 408 tolerance
in some processed foods could be avoided by restrictions on pesticide
use to crops grown for the fresh market.
4. Evaluation of factors. Below, EPA evaluates its concentration
policy including EPA's use of processing studies, the factors
considered by EPA in evaluating whether processing studies show the
possibility of residues over the section 408 tolerance, and the
relevance of the various reasons noted above why overtolerance residues
infrequently occur.
Processing studies. EPA guidelines on residue data specify that
processing studies should ``simulate commercial processing as closely
as possible.'' Pesticide Assessment Guidelines, Subdivision O at 21
(1982). Data from such studies, EPA believes, remain the most relevant
information in determining whether residues over the section 408
tolerance may occur. Because section 408 tolerance values represent a
level of residues which field trial studies show can occur, data from a
processing study showing concentration can be a good indicator
regarding the possibility of overtolerance residues in processed food.
EPA has not issued extensive industry-by-industry guidance on what
constitutes ``commercial processing'' but rather has left it to the
pesticide manufacturer to insure that modern commercial processing is
reflected in the processing studies. Thus, EPA disagrees with comments
by NFPA and other commenters which suggest it is EPA which is at fault
for not taking into account practices such as washing and peeling that
routinely occur during processing. If those practices are a part of
commercial processing for certain foods and are not reflected in the
processing studies designed and submitted by pesticide manufacturers,
the pesticide manufacturers need to provide EPA with data that are
truly representative of the industry practice.
Rounding. To a limited extent, EPA has considered the rounding up
that occurs in the selection of the section 408 tolerance value in
making concentration determinations. EPA believes the degree of
rounding remains a legitimate consideration in determining the
likelihood that processing may produce residues in processed food
greater than the section 408 tolerance. Moreover, as noted below, EPA
believes it is appropriate to consider the difference between residue
levels that can occur on crops and the section 408 tolerance level in
evaluating the possibility of residues over the section 408 tolerance
in processed food.
But EPA is concerned that its past practice of rounding up has
resulted in section 408 tolerances being set at a level higher than is
necessary to cover legally treated crops. EPA is currently examining
whether older section 408 tolerances have been set at inappropriately
high levels owing to rounding or for other reasons. EPA is also
exploring whether there might not be statistical techniques for better
assigning section 408 tolerance levels. To the extent EPA alters its
approach to selecting section 408 tolerance levels, these revised
section 408 levels will need to be considered in making determinations
under the concentration policy.
Variability of methods. EPA continues to believe that the
variability of the analytical method should be evaluated in determining
whether residues over the section 408 tolerance are likely to be
reliably detected despite a processing study showing concentration in
fact. The aim of the concentration policy is to identify those uses
which can produce residues over the section 408 tolerance in processed
food. If any possible concentration is so low that it could not be
clearly identified by the relevant analytical method, then, in fact,
instances of residues over the section 408 tolerance in processed food
would not be expected. The degree of variability in analytical methods
must be assessed on a case-by-case basis. Generally, the variability in
analytical methods suggests that residues over the section 408
tolerance are not likely to be reliably detected where processing
studies show concentration factors in the range of 1.1X to 1.5X.
Treatment rates and processor control. EPA believes that it is
appropriate to assume that some growers will treat a portion of their
crop at the maximum treatment rate allowed by the label. EPA's
experience has shown that due to unexpected weather
[[Page 31305]]
and pest pressures it is unrealistic to assume that no grower will
treat his or her crop with a pesticide in the manner that yields the
highest lawful residues.
Moreover, where residues do concentrate during processing, EPA
questions the ability of the processor or grower to manage pesticide
residue levels so as not to produce over-tolerance residues in
processed food. Although processors may know the concentration factor
of residues from processing studies, the concentration factor does not
suggest with any precision how processors could instruct growers to
change their pesticide application procedures so that residues over the
section 408 tolerance will not result in processed food. Levels of
residues in raw crops are dependent not only on how much pesticide is
applied but on when and how the pesticide is applied. Little data exist
that describe the effect of varying any of these procedures on residue
levels. Similarly, EPA believes little information is available
concerning how changes in their manufacturing processes affect residue
levels in processed food. Finally, as discussed below, the comments
received on the NFPA petition reinforce EPA's experience that farmers
often do not know the ultimate destination of their crop. Therefore,
EPA believes it would be very difficult for growers or processors to
manipulate residue levels in processed food.
EPA would be open to considering further industry proposals laying
out a potential policy framework that more specifically delineates how
processor practices could be taken into account in determining the
likelihood that residues in processed food would exceed the applicable
section 408 tolerance. It would be helpful if such policy proposals
contained criteria for evaluating whether specific processor claims
regarding pesticide/commodity combinations are reasonable. Among other
things, these criteria should address (1) what data would be submitted
to EPA to verify residue levels, (2) how the practicality of the
proposed scheme would be evaluated (e.g., degree of concentration of
processing operations and ability to separate raw food streams), and
(3) whether processor control of residue levels for a specific
pesticide/commodity combination could be feasibly enforced. If such
further policy proposals are received, EPA would seek public input
before making any decision on the merits of the proposals and using the
proposed criteria in evaluating specific pesticide uses.
Mixing and blending. EPA believes that in many instances it would
be appropriate to take into account mixing and blending in determining
the likelihood that residues over the section 408 tolerance could
result. This change in practice is warranted, EPA believes, because
EPA's prior assumption, i.e., that all raw food have the potential to
have residues at or near the section 408 tolerance level, does not
adequately take into account the realities of food processing. Because
of the way EPA sets section 408 tolerances, individual raw commodities
do have the potential of having residues at or near the tolerance
level. The data from field residue trials show, however, that residue
values even from a single field can vary significantly. When individual
raw commodities are mixed in processing operations, it is realistic to
expect that there will be an averaging effect on the residues in the
processed food.
Accordingly, if EPA determines that there is a sufficient degree of
mixing or blending during processing such that the normal variation
among individual samples from a field will be substantially evened out,
EPA will consider comparing some ``average'' residue value from field
trials times the concentration factor to the RAC tolerance level in
determining the likelihood of residues over the section 408 tolerance.
EPA generally believes that the most relevant ``average'' residue value
from crop field trials is the highest average residue value from the
series of individual field trials. Using an average of all samples from
all field trials in all regions of the U.S. would tend to suppress the
variability in residue values to a greater extent than can be expected
by mixing or blending. Generally, crops grown in different regions of
the U.S. are not mixed prior to processing. Rather, crops are often
processed field-by-field as they are harvested by the grower.
There are a number of constraints EPA thinks are critical here.
First, considering average field trial residues is only appropriate
where the values being averaged are from field trials involving maximum
treatment rates. In other words, averaging may be used to take into
account the variation in residues which occurs in crops receiving
maximum treatment and minimum preharvest intervals but not residue
variations as result of different levels of treatment. As laid out
above, EPA has no basis on which to make assumptions about whether
crops in specific instances would be treated at rates lower than the
maximum permitted on the pesticide label or what residues those lower
rates would produce. Second, whether considering blending would be
appropriate would depend on the quality of the data base. Consideration
of any ``average value'' would be less appropriate where adequate data
from all representative regions of the country are not available.
Finally, even where it would be appropriate to consider average
residues, EPA believes a simple calculation showing that the average
residue multiplied by the concentration factor from a processing study
is less than the RAC tolerance alone may not conclusively show that
residues over the section 408 tolerance could not result. In
appropriate circumstances, EPA may need to consider a number of other
factors, such as the variability in the field trial data, in
determining the likelihood of residues over the section 408 tolerance.
Degradation of residues. Although EPA recognizes that degradation
of residues frequently occurs, it is not apparent how EPA could take
that phenomenon into account in its concentration policy other than to
the extent the effects of degradation are captured in processing
studies. EPA would need detailed data on the degradation rates of
pesticides as well as on the minimum time between the harvesting of
crops and when such crops are manufactured into ready-to-eat processed
foods. Without such information, it would be difficult to establish a
tolerance level that would assure that legally treated crops did not
result in illegal food.
Some comments filed in response to the NFPA petition suggest that
marketplace survey or FDA monitoring data would be relevant to whether
there is a likelihood of residues over the section 408 tolerance.
Certainly, data from marketplace studies have some degree of relevance
to the question of whether residues in processed food may exceed the
section 408 tolerance. The relevance of marketplace studies, however,
depends on how the marketplace study was performed. For example, the
principal reason marketplace studies have been conducted in the past is
to obtain better data concerning actual residue values close to the
point at which food is consumed. Thus, marketplace studies generally
involve sampling commodities in retail grocery stores. A tolerance for
processed food would not only apply to food in retail stores but at all
prior points at which the food moved in interstate commerce. This fact
would have to be taken into account in assessing the relevance of a
marketplace study in determining the likelihood of residues in
processed food in excess of the section 408 tolerance. Monitoring data
can also be relevant to determining
[[Page 31306]]
the likelihood of residues in processed food exceeding the section 408
tolerance. However, FDA monitoring data, especially monitoring data on
processed foods, generally has been limited and thus may not be a
reliable predictor of the level of residues of a particular pesticide
in a particular processed food.
Market segregation. Several commenters contend that, even where
residues could be expected to concentrate in processed food above the
section 408 tolerance, if EPA were to permit pesticides to be labeled
solely for crops grown for fresh market, no section 409 FAR would be
needed for such pesticide uses. These commenters claim that certain
crops are so specialized that they are grown specifically for the fresh
or processed market, and, in some instances, that even different
pesticides are used on crops depending on whether they are intended for
the fresh or processed market. Thus, these commenters argue that
allowing pesticides to be labeled for crops grown only for the fresh
market where a specialized crop has been developed solely for the fresh
market would not pose an enforcement problem. On the other hand, EPA
received other comments stating that placing such label restrictions on
pesticides would subject growers to a form of ``Russian Roulette.''
EPA's observations indicate that it is difficult to achieve total
market segregation; however, if a party can show that a market for a
specific crop can be segregated and that such segregation can be
feasibly monitored, EPA will not require a section 409 FAR for a
pesticide on that crop.
5. Conclusion. In sum, EPA's concentration policy will continue to
focus on ``possible residues'' in the processed food. EPA will place
primary emphasis on whether processing studies show that the processing
of a commodity results in a level of residues in the processed food
which is greater than the level of residues in the raw food. EPA will
also consider the variability of the analytical method, the degree of
rounding involved in establishing the section 408 tolerance, and, where
circumstances permit, information concerning blending of crops and
average field trial values, and market basket surveys. EPA will
consider information concerning potential market segregation and
pesticide segregation, but such segregation must be established by
clear evidence. But EPA remains unconvinced at this time that it should
give much weight at all to degradation information or the possibility
that farmers are applying pesticides at lower application rates or that
processors will control whether residues over the section 408 tolerance
occur.
VI. Ready To Eat
A. NFPA's Argument and Views of Commenters
The NFPA petition argues that EPA has failed to take into account
language in the flow-through provision of FFDCA section 402 specifying
that processed food is to be evaluated at the ``ready-to-eat'' stage in
determining whether the food exceeds the relevant section 408
tolerance. According to NFPA, the ``ready to eat'' language was added
to the statute to ``take care of any particular problem that might be
raised with respect to a product that was concentrated or dehydrated.''
(NFPA Petition at 34). In its comments, NFPA proposed a definition of
not ready-to-eat food as food ``customarily reconstituted by the
consumer or food manufacturer, or [food] sold for use as an ingredient
in the preparation of finished foods.'' (Comments of NFPA at 12).
Further, NFPA cites several examples from the Code of Federal
Regulations and the Federal Register in which Federal agencies have
used the term ``ready to eat'' to distinguish between various foods.
Except for two comments from State agencies (Florida Department of
Agriculture and North Dakota Department of Agriculture), most of the
commenters on the NFPA petition assert that EPA's approach of treating
any food available for sale as ``ready to eat'' is violative of the
plain words of the statute. Many of these commenters also contend that
EPA overstated the enforcement difficulties of construing the term
``ready to eat'' more narrowly.
As to the definitional issue, numerous commenters contend that the
literal or plain meaning of the term ``ready to eat'' food is food
consumed ``as is.'' One commenter quotes the dictionary definitions of
``ready'' and ``eat'' to derive a definition of ``ready to eat'' food
as ``prepared for immediate taking through the mouth as food.''
(Comments of Catherine Clay at 1). Many commenters mention specific
foods and assert that they were not consumed ``as is.'' In their
comments, fruit growers are particularly adamant that juice
concentrates are not ``ready to eat.'' (See, e.g., comments of Sun-
Diamond Growers at 7 (``People simply do not consume a quart of prune
juice concentrate or even a cup of concentrate.'')). Another commenter
contends that EPA should focus on what the usual practice was as to
foods:
We suggest that for those food items that are never or seldom
consumed in their concentrated forms (e.g., tomato paste, oils,
flour, and juice concentrates), Section 402 should be followed * *
*. Those few situations in which product might be consumed in the
concentrated form do not present an imminent hazard and will not add
significantly to the risk calculation.
(Comments of Del Monte Foods at 1).
As to potential enforcement difficulties with following a consumed
``as is'' approach to ``ready to eat,'' several commenters argue that
EPA could adopt action levels to determine if processed not ready-to-
eat food is adulterated. (Comments of Monsanto; Grocery Manufacturers
Association; NFPA). Such action levels would be established using
dilution factors that take into account the dilution of pesticide
residues as a food is mixed with other foods in processing operations.
The dilution factors, these commenters urge, should be based on the
most concentrated form of ready-to-eat food that the not- ready-to-eat
food was used to produce.
Finally, several commenters claim that commodities such as fruit
pomaces and seed hulls which are commonly used as animal feeds are not
``ready to eat.'' According to these commenters, most animal feeds are
a blend of different ingredients because commodities such as pomaces
and hulls are both nutritionally deficient and unpalatable.
B. EPA's Response
1. The definitional issue. EPA has considered NFPA's arguments and
the comments received and has examined the previous uses of the term
``ready to eat'' by EPA and other Federal agencies. EPA agrees that the
term ``ready to eat'' food has a common-sense meaning of food which is
consumed without further preparation. EPA intends to apply that
interpretation in future actions. Basically, EPA believes that food
should be considered ``ready to eat'' if it is consumed ``as is'' or is
added to other ready-to-eat foods (e.g., condiments). Use of this
interpretation, of course, will not clarify all issues regarding
``ready to eat'' foods. EPA envisions that this definition may be
difficult to apply in many instances.
Some foods will be easier to classify than others. EPA has, in the
past, established section 409 FARs for some foods that clearly do not
meet a common-sense interpretation of ``ready to eat'', and EPA did so
without closely considering what level of residue would occur in
derivative foods which are ``ready to eat.'' Examples would include
dried hops, mint oil, citrus oil, and guar
[[Page 31307]]
gum. These foods are not generally available to consumers in grocery
stores and, even if a consumer could purchase such a food, it would not
be consumed ``as is'' but would be further processed (e.g., dried hops
used in brewing beer) or used as an ingredient in a food product. Other
foods for which EPA has set food additive regulations, such as raisins,
olives, and potato chips, clearly are ``ready to eat.''
EPA generally believes that foods that are mixed prior to
consumption are not ``ready to eat.'' Mixing generally involves the
combining of foods with the intent of creating a different food
product. For example, combining a tea bag with hot water is intended to
create a new food product, the beverage tea. Thus, the dried tea in the
tea bag would not be considered ``ready to eat.'' On the other hand,
EPA does not believe this mixing principle applies to condiments.
Condiments are consumed as a supplement to other ``ready to eat'' food.
A condiment is also consumed ``as is.''
There remain, however, many commodities for which EPA has
traditionally set food additive regulations which are not so easily
characterized under the ``ready to eat'' standard and which will
require a case-by-case inquiry. One of the reasons for the fact-
intensive nature of this inquiry is that foods have many uses and
eating habits vary widely in the United States. Thus, determining
whether a food is ``ready to eat'' involves identifying all significant
uses of a food and then determining if any of those uses meets the
definition of ``ready to eat.'' For example, perhaps the most common
use of vegetable oil is as a cooking medium or as an ingredient in
baked products. However, another use of vegetable oil is as a
``dressing'' for a green salad. When used in this manner, oil is
directly added to the salad as a condiment, and thus oil generally
would qualify as ``ready to eat.'' Additionally, EPA will need to
explore whether some foods which have traditionally not been consumed
without further preparation, are actually being consumed on an ``as
is'' basis. Comments submitted by DuPont Agricultural Products support
this approach:
We appreciate that some concentrated products can be consumed
without mixing. The likelihood of occurrence of this consumption
pattern is a factor which should be considered in determining which
form is best viewed as the ready-to-eat stage. In our view, a
reasonable approach would be to weigh such a consumption pattern
based on the frequency of occurrence. If the consumption of the
concentrate occurs with great infrequency, the appropriate ready-to-
eat food would still be the diluted product.
(Comments of DuPont Agricultural Products at 8).
In circumstances where EPA's revised approach to the term ``ready
to eat'' results in particular food forms of a commodity being dropped
from the category of ``ready to eat,'' EPA will need to explore whether
there is a possibility of concentration of residues above the section
408 tolerance in any other, ready-to-eat forms of that commodity. In
many instances further preparation of a not-ready-to-eat commodity will
so significantly reduce residues that, even if the not-ready-to-eat
precursor processed food contained residues over the section 408
tolerance, the ready-to-eat commodity will not. Use of citrus oil as a
flavoring in ice cream may be an example of this phenomenon. Citrus oil
may be such a small proportion of the total product that any residues
over the section 408 tolerance in the oil would be diluted below the
section 408 tolerance in the ice cream. However, in other instances,
the dilution involved in further preparation of a not-ready-to-eat
processed food is not so dramatic. For example, flour, assuming it is
found to be a not-ready-to-eat food, is prepared into commodities such
as crackers or tortillas in which the dilution factor may be fairly
modest. In situations such as this, EPA will have to determine whether
it should be setting section 409 FARs on different commodities than has
been EPA's traditional practice.
2. Enforcement approach. EPA's revised approach to the term ``ready
to eat'' will make enforcement of the FFDCA more challenging as regards
foods no longer considered ``ready to eat.'' EPA does not view as
satisfactory NFPA's suggestion that for enforcement purposes EPA should
develop dilution tables and from such tables promulgate action levels
to evaluate the legality of not-ready-to-eat processed food. Although
this is a possibility, EPA regards it as cumbersome and lacking the
enforcement ease of binding tolerances. An action level is not binding
on anyone and thus even though use of a dilution table may suggest that
a food is adulterated, FDA could only successfully proceed against the
food if it could prove in court that the level of residue found in the
not-ready-to-eat food would render ready-to-eat food adulterated.
Instead, EPA has decided to use its general rule-writing authority
under FFDCA section 701 to establish maximum residue levels for not-
ready-to-eat processed food. Section 701 grants EPA the authority ``to
promulgate regulations for the efficient enforcement of this Act.'' 21
U.S.C. 371. These maximum residue levels would be set no higher than
the levels which could result in the processed food assuming legal
residues in the raw food and that good manufacturing practices were
followed.
EPA's authority to set such maximum residue levels arises from the
flow-through provision. The flow-through provision does not legalize
residues in ready-to-eat processed food unless three criteria are met:
(1) the residues are at or below the applicable section 408 tolerance;
(2) the precursor raw food had residues within the section 408
tolerance; and (3) good manufacturing practices were followed in
preparing the processed food. The maximum residue levels set under
section 701 would establish binding regulations as to when the two
latter criteria of the flow-through provision are met for a specific
pesticide use. If such a maximum residue level were exceeded in a
processed food, then as a matter of law the flow-through provision
would not apply to the food (whatever the residues in the food when it
is ``ready to eat''), and thus the food would be adulterated as a
matter of law under FFDCA section 402(a)(2)(C).
3. Animal feeds. As noted, a number of commenters claimed that food
processing byproducts such as grape pomace, soybean hulls, etc. are not
``ready to eat'' either because they are unpalatable or nutritionally
deficient or because they are not a significant portion of the diet of
animals. EPA generally intends to apply a similar approach to
processing byproducts used as animal feeds as it will to human foods in
determining whether the byproducts are ``ready to eat'' and will also
use section 701 maximum residue levels, as described above, where
appropriate. Determinations on specific processing byproducts will have
to be made on a case-by-case basis. To the extent it can be shown that
any individual processing byproduct is unpalatable when fed ``as is''
or that for other reasons the processing byproduct is generally not fed
absent further processing or mixing, EPA would not categorize that
particular byproduct as ``ready to eat.'' EPA believes this showing
probably can be made for a substantial number of processing byproducts.
In response to comments stating that EPA required examination of
processing byproducts not currently used as animal feeds (e.g., apple
pomace), EPA would note that it has recently revised its guidelines on
what processing byproducts are used as animal feeds. This revision
followed a comprehensive survey of animal feed practices. EPA has
[[Page 31308]]
also sought public comment on those guideline revisions and will
continue to consider comments on this issue.
4. Future actions. EPA intends to apply its revised approach to the
term ``ready to eat'' in all future tolerance actions. When any action
is taken based on EPA's revised approach, EPA will seek public comment
on designations for specific commodities prior to making any final
determinations.
VII. Are EPA's Policies Rules That Have Not Been Properly
Promulagted?
NFPA contends in its petition that EPA's coordination and
concentration policies are not in compliance with the Administrative
Procedure Act (APA) because they have not been promulgated as a binding
regulation through notice and comment procedures. As to the
concentration policy, EPA has in this notice announced a revised
concentration policy that EPA believes is fully consistent with the
requirements of the APA. This revised policy is not intended to be of
controlling effect either on EPA or regulated parties. Rather, it is
intended as guidance for EPA in administering its authority under
FFDCA. For example, EPA has explained in some detail in its revised
concentration policy what types of data it intends to place primary
reliance upon in determining whether section 409 FARs are needed.
However, EPA has noted its willingness to consider other information
and arguments. Thus, because the revised concentration policy is not
intended as a binding regulation, it need not be promulgated through
notice and comment rulemaking.
List of Subjects
Environmental protection, Agricultural commodities, Food additives,
Feed additives, Pesticides and pests.
Dated: June 9, 1995.
Lynn R. Goldman,
Assistant Administrator for Prevention, Pesticides and Toxic
Substances.
[FR Doc. 95-14683 Filed 6-12-95; 12:20 pm]
BILLING CODE 6560-50-F
