m "3 1979
ENDRIN
POSITION DOCUMENT 4
Special Pesticide Review Division
Office of Pesticide Programs
Office of Toxic Substances
U.S. Environmental Protection Agency
- Xo/3
-------
REPORT DOCUMENTATION \ '¦ RE™:" K0-
PAGE -	! EPA/SPRD -80/39
4. Title and Subtitle
Endrin: Position Document 4
/. Author',s)
! 9. r.'c i f otti in^ 0-'£j i-1 ion Niuno ind Awdra^'.
Special Pesticide Review Division
Environmental Protection Agency
Crystal Mall #2
Arlington, VA 22202
12. Sponsoring Or^.»nujfon .Name and Aci(^css
Environmental Protection Agnecy
401 M St. S.W.
Washington, D.C. 20460
15. Sur>pl< 'CHvM.iiy N.oU-4
. 2.
I
A..
3. Recipient's Accejilon No
_ _ ?BSI	1_0?A_8-L
5. Report Do?c
5/3/79
3. PorJo/ijjina Organ!.? ;;{:cn Rc;U. Nn
10. Project/ Tasy/Work Unit No.
i 11. Coniract(C) cr Grant'G) f\o.
I (O
I
i (G)
"1
13. Typ« of R-5 pert & PcriotJ Cove-rccl
j i/..
16. Ary.Uuct O.irmt" ?OQ v^ordi.i
j	Agency Regulatory Decision on Registration of Pesticide: Also contains
j	comments from USDA and Scientific Advisory Panel and EPA's response to these
:	concerns on economic, agricultural, environmental and health impact of
j	the regulatory decision.
f	- - - -
• x7.	A--.:;.'/vs a r> „c i iptor
0504,0606,0703
t b '.f.ontilioi^/Opcn-CriOeC Verms
j. CO.SATI Kieto'/fifoup
;n. Availability Stfitemon'.
Unlimited
i
I	
(5oe ANCS1-Z30.1 35
i 19. Security C'css (This Report)
! Unclassified
^•20. Socurlty Clout Cfhla Page)
Unclassified
21- No. of
i 22. Price
J
Sac /n*f.*uct.'cn& on i?ovcr%o
O?T50iSAL TORM 772 (4-7/)
(For.-f.orly NT1S-2S)
Cnpnrtmon! of Coinrr.erce
-------
Acknowledgements
Writing Staff
Kyle R. Barbehenn, Project Manager, SPRD, OPP
EPA Project Support Team
Mitchell Bernstein, Attorney, OGC
William Burnam, Pharmacologist, HED, OPP
David Coppage, Aquatic Biologist, HED, OPP
George Ludvik, Entomologist, BFSD, OPP
David Severn, Chemist, HED, OPP
EPA Pesticide Chemical Review Committee (PCRC)
Dr.	William A. Wells, Chairperson, SPRD, OPP
Dr.	Elizabeth L. Anderson, CAG, ORD
Dr.	Richard N. Hill, OTS
Dr.	Allen L. Jennings, SRD, OPM
Dr.	Donna R. Kuroda, OHEE, ORD
Mr.	John J. Neylan, PED, OE
Mr.	Michael S. Winer, OGC
-------
TABLE OF CONTENTS
I.	Introduction
II.	Analysis of Comments
A. Garments Relating to Risk
1.	Teratogenicity	4
2.	Acute Toxicity to Wildlife	14
3.	Population Reductions of Aquatic Organisms	15
4.	E&tality to Endangered Species	16
5.	Oncogenicity	20
B.	Ccnanents Relating to Benefits	23
1.	Cotton	23
2.	Small Grains	26
3.	Apple Orchards	30
C.	Comments Relating to Regulatory Options	31
1.	Designation of Target Species in Apple Orchards	31
2.	Equipment	33
3.	Distance Restrictions from Aquatic Habitats	34
4.	Distance Restrictions frcan Hunan Habitation	50
5.	Posting of Contaminated Ponds	51
6.	Teratogenic Warning	53
7.	Protective Clothing for workers	56
8.	Warnings on Prophylactic Use	57
9.	Enforcement	60
10. Grasshopper Control
D.	Comments Relating to Procedural Matters	61
1.	Availability of the Agency's Rebuttal Analysis	62
2.	Initiation of New Teratology Studies	63
3.	Use of Relevant Information on Risk Assessment	66
4.	Development of State Programs for Use on Cotton	67
III. Conclusions	68
Bibliography	73
Appendices
A.	Comments by The Scientific Advisory Panel
B.	Comments by the U.S. Department of Agriculture
C.	Comments by the U.S. Fish and Wildlife Service
-------
I. Introduction
The Federal Insecticide, Fungicide, and Rodenticide Act,
as amended (FIFRA) [7 U.S.C. Section 136 et seq.] regulates
all pesticide products. Section 6(b) of FIFRA authorizes
the Administrator of the Environmental Protection Agency
("EPA" or the "Agency") to issue a notice of intent to
cancel the registration of a pesticide or to change its
classification if it appears to him that the pesticide or
its labeling "does not comply with the provisions of
[FIFRA] or, when used in accordance with widespread and
commonly recognized practice, generally causes unreason-
able adverse effects on the environment."
The Agency designed the Rebuttable Presumption Against
Registration (RPAR) process to gather risk and benefit
information about problem pesticides and to make balanced
decisions concerning them in a manner which allows all
interested groups to participate. This process is set
forth in 40 CFR 162.11.
On July 27, 1976, the Agency issued an RPAR notice
for pesticide products containing endrm (41 FR 31316).
The endrin RPAR was one of the first issued by the Agency.
At the time it was issued, Agency RPAR procedures were
still in a formative stage, and a detailed Position Document
1 did not accompany the endrin RPAR notice. Copies of this
Document, however, were provided to all registrants and
other concerned parties.
-1-
-------
On October 20, 1978, the Agency issued Endrin: Position
Document 2/3 (EPA, 1978; hereinafter called PD 2/3), and
published a Notice of Determination and Availability of the
Position Document in the Federal Register on November 2,
1978 (43 FR 51132). In PD 2/3 the Agency analyzed the
rebuttals it received in response to the original RPAR
notice, presented its analysis of both the risks and benefits
associated with the uses of endrin, and proposed a decision
to conclude the RPAR process.
FIFRA requires the Agency to submit notices issued
pursuant to Section 6 to the Secretary of Agriculture
("Secretary" or "USDA") for comment on the impact of the
proposed action on the agricultural economy [Section 6(b)]
and to a Scientific Advisory Panel (SAP) for comment on the
impact of the proposed action on health and the environment
[Section 25(d)]. The Agency is required to submit these
documents to the Secretary and the SAP at least 60 days
before making the final notice effective by sending it to
registrants and making it public. The Secretary and the
SAP may comment in writing within 30 days of receiving
the notice; the Agency is required to publish any of their
comments and the Administrator's responses with publication
of the final notice.
-2-
-------
Additionally, since the RPAR notice indicated that
endrin had caused fatality to an endangered species, the
Agency was required by Section 7 of the Endangered Species
Act of 1973 (16 O.S.C. 1531; see also 50 CFR 402, 43 FR 870)
to initiate formal consultation with the U.S. Fish and
Wildlife Service, U.S. Department of the Interior ("FWS").
The biological opinion submitted to the Agency by the FWS on
June 8, 1978 directed the Agency to take appropriate action
to reduce risks to endangered species from the use of endrin
and to reinitiate formal consultation on the proposed
actions (Greenwalt, 1978a). The comments of the FWS to the
actions proposed in PD 2/3 were made on December 14, 1978
and in a supplementary revision on March 1, 1979 (Greenwalt,
1978b; 1979).
The Agency is not required under the statute to afford
registrants and other interested persons an opportunity to
comment on the bases for the proposed action while it
is under review by the USDA and the SAP. However, the
Agency decided that it was consistent with the purpose of
the RPAR process and the Agency's overall policy of open
decision-making to do so. Accordingly, PD 2/3 solicited
such public comments.
The Agency has received a number of public comments in
response to the November 2, 1978 Notice of Determination and
the Endrin PD 2/3. Responses from the SAP, the USDA, the
-3-
-------
FWS, Velsicol Chemical Corporation ("Velsicol"), which is
the sole manufacturer of endrin in the United States, the
Environmental Defense Fund (EDP) , and other interested
parties have been analyzed and are addressed in Section II
of this document. The entire responses from the SAP, the
USDA and the FWS are contained in the Appendices to this PD 4.
II. Analysis of Comments
A. Comments Relating to Risk
1. Teratogenicity
Comments have been received, regarding the
validity of the tests on which the Agency relied in con-
cluding that endrin has a teratogenic potential, regarding
levels of exposure that can reasonably be anticipated, and
regarding a margin of safety (MOS) that can be considered as
"ample". These comments and the Agency's response are:
a. Validity of the Tests
The Agency's risk analysis (PD 2/3) noted
that a single exposure of 5 rag/kg endrin on the eighth day
of pregnancy caused significant numbers of meningoencephalo-
celes in hamsters. A no-observed-effect-level (NOEL) of 1.5
rag/kg was established by this study (Chernoff et al., 1978a).
Although the teratogenic studies were discussed at length in
the FIFRA-SAP Meeting of October 26, 1978 (Transcript of
Proceedings, hereinafter referred to as SAP, date, page) the
-4-
-------
SAP did not make a formal comment on this issue. A consultant
for Velsicol seemingly challenged the validity of the
Chernoff study (Velsicol, Exh.31) but the points raised
were rebutted by Chernoff et al. (1978b). Velsicol's second
consultant accepted the validity of the NOEL of 1.5 mg/kg
(Velsicol, Exh. 30, p. 6). Since Velsicol and its first
consultant now apparently accept the validity of the estab-
lished NOEL for purposes of risk assessment (Velsicol, p.38
and Exh. 61) the details of the related comments and rebuttal
do not require further discussion.
b. Levels of Exposure that can be Anticipated
The Agency's exposure analysis focused on
dermal exposure to bystanders and persons associated with
the process of applying endrin and on the ingestion of
contaminated fish. The Agency believes that dermal exposure
to applicators and bystanders can be reduced adequately by
requiring protective clothing, prohibiting application
within specified distances from human habitation and similar
measures less stringent than cancellation. Ingestion
exposure, however, is of particular concern to the Agency
because the contamination of fish-bearing waters by runoff
is difficult to control, especially where endrin is used on
cotton in areas which receive substantial rainfall.
-5-
-------
Velsicol has challenged the validity of the
Agency's estimate of potential exposure from the consumption
of contaminated fish. The Agency based its risk assessment
on the consumption of 25Og of fish containing 1.0 ppm
endrin. This level of exposure was conceived, not as a
"worst case" estimate, but as a reasonable one (SAP, October
26, 1978, p.14). Velsicol did not challenge the use of 250g
but contends that 0.4 ppm is the highest concentration of
endrin that can be expected to occur in fish (Velsicol, p.39
and Exh. 5) and relies on the National Pesticide Monitoring
Program (NPMP, Seabolt, 1978) results to support its contention.
It is true that the highest concentration
of endrin in fish reported for 1977 in the NPMP was 0.4 ppm.
However, the NPMP samples fish from major rivers throughout
the nation without regard to sources of potential contamina-
tion. Moreover, the sampling program is not designed to
determine maximum residues that might occur m fish in crop-
land areas. NPMP samples from Alabama, Arkansas, Louisiana,
Mississippi and Tennessee were taken in major rivers where
cotton is grown on only a small fraction of the drainage
area; where only a small fraction of the cotton that is
grown is actually treated with endrin (EPA, 1977); and
without regard to actual or potential runoff episodes.
-6-
-------
Thus, it is somewhat surprising to find endrin present at
any concentrations m the fish sampled. The widespread and
regular occurrence of endrin in these fish is strong evidence
that endrin is likely to be present in much higher concentra-
tions in fish more closely associated with cotton culture.
Levels of endrin in the edible portion of
catfish killed by endrin may in fact exceed 4 ppm ( Mount
and Putnicki, 1966). Since the record establishes that many
fish kills have been associated with the use of endrin,
especially on cotton, it can reasonably be expected that
fish-bearing waters have been contaminated with sub-lethal
doses of endrin with a much greater frequency. Velsicol's
consultant (Velsicol, Exh.5) has estimated that fish exposed
to sub-lethal doses of endrin could accumulate as much as
2.0 ppm. Thus, the Agency's use of 1.0 ppm endrin as a
concentration that could reasonably be anticipated in fish
consumed by humans is on firm ground^.
1/ Velsicol also relies on estimates of runoff concentrations
which would allegedly occur if a quarter-mile barrier strip
were to be imposed to show that the resulting concentrations
would be "safe" for fish and apparently for residues m the
fish (Velsicol p.26; Exh. 24). As discussed more fully in
Section II C 3, the Agency cannot rely upon their calculations
and assumptions involved in the barrier strip model and
cannot reasonably conclude that residue levels would be
acceptable under that proposal.
-7-
-------
What must be anticipated m the field is a
wide range of endrm residue levels in fish that vary in the
probability of occurrence — from the infrequent but very
high levels associated with dead and dying fish through all
degrees of sub-lethally exposed fish. The amounts of such
fish that may be consumed in a day may range widely from
small (125 g) through large (250 g) to exceptional portions
(500 g). Ranges in the margins of safety associated with
these variables will be presented below.
c. Adequacy of the Margin of Safety (MPS)
The Agency has no rule of general appli-
cability for determining the ampleness of the teratogenic
margin of safety associated particular compounds — each
chemical is evaluated individually. In evaluating endrin
the Agency noted that humans might be 50 times more sensitive
to the convulsive effects of endrin than are hamsters and
concluded that such a difference in sensitivity might also
be true for teratogenic effects (PD 2/3, p.51). Thus, the
Agency concluded that exposure levels that would give rise
to an MOS of 500 or lower would be cause for concern. Both
the SAP (SAP, October 26, 1979, pp.30-32) and Velsicol
(Velsicol, p.40 and Exh. 32) objected to the derivation of
this MOS. Informally, the SAP members indicated that an
-8-
-------
ample MOS should be somewhere between 100 and 1000 but could
arrive at no scientific method for establishing an appropriate
value (SAP, October 26, 1979, pp. 28-34, 124-126). No
formal recommendation was made.
Velsicol has attempted to make several
points bearing on the assessment of teratogenic risk that
require a clarifying response:
1)	Velsicol claims that the "actual" NOEL
lies somewhere between the lowest observed effect level (5
mg/kg) and the observed NOEL (1.5 mg/kg) (Velsicol, pp.41-2,
Footnote (£N) 7). This contention is merely speculative. The
Agency must rely on established values in estimating the
MOS.
2)	Velsicol argues that an MOS of 100 is
appropriate for endrin. Velsicol states "As Or. Wilson
notes (Exhibit 32), and as the Agency acknowledges (Position
Document 2/3, p. 51), however, a margin of safety of 100 is
normally ample for low potential environmental teratogens
such as pesticides..." (Velsicol, p.40). The Agency has
neither characterized pesticides in general nor endrin
specifically as "low potential environmental teratogens".
Elirther, what the Agency did say concerning the adequacy of
margins of safety was:
-9-
-------
While the Agency has not esta-
blished official guidelines for
determining the adequacy of the MOS
for teratogens in general, Agency
toxicologists believe that an MOS
below 100 would be a matter of
serious concern. Interpreting these
values, however, requires a judgement
based on other factors associated with
characteristics of the chemical,
routes of exposure, and the probabi-
lity of various levels of exposure.
Thus, the above value should not
be construed as an established Agency
policy but only as a toxicological
guideline for risk assessment against
which benefits must be balanced and
additional safety requirements imposed
(PD 2/3, p.50-1).
3) Velsicol's consultant, Or. Wilson,
has taken exception to the informal comments by the SAP
suggesting that an MOS of 1000 might be appropriate.
According to Dr. Wilson, "The only reasonable justification
for a margin of safety of 1000 would be in the event that
endrin were an environmental pollutant of no or negative
economic importance and totally without benefit to man. To
the contrary, it makes a significant contribution toward
providing food and fiber to meet human needs" (Velsicol,
Exh. 35). Clearly, Dr. Wilson's concept of the adequacy of an
MOS is not cast solely in terms of assessing risk, per se,
but is predicated on assumptions concerning environmental
pollution and benefits of use. In admitting, however, that
a MOS of 1000 may be justified in some circumstances,
-10-
-------
Dr. Wilson apparently concedes that the teratogenic risk at
2/
margins of safety greater than 100 may be cause for concern—'.
As discussed above, the Agency does take benefits xnto
account in reaching a final regulatory decision concerning a
use of a pesticide. Here, the facts that meningoencephalocele
is a very serious defect and that the benefits from the use of
endrin on cotton are very low lead the Agency to conclude
that an MOS much greater than 100 is appropriate for this
use of endrin.
4) Another of Velsicol's arguments implicitly
objects to the Agency's use of any MOS at all. Velsicol
contends that a pregnant woman would have to consume "ludi-
crously massive amounts of endrin-contaminated fish to
incur a teratogenic hazard" (Velsicol, p.38). Velsicol then
17 In fact, Dr. Wilson's reasoning would indicate that an
MOS of between 100 and 1000 is appropriate for endrin.
Endrin was found in the vast majority of the fish inhabiting
all major rivers sampled by the NPMP in Alabama, Arkansas,
Louisiana, Mississippi and Tennessee in 1977 (Seabolt, 1978)
and it has occurred at lethal or near lethal levels in the
brains of brown pelicans, white pelicans and bald eagles (PD
2/3, pp. 37-9). While endrin may not share the apparent
ubiquity of certain other organochlorines m the environment,
it certainly qualifies as an "environmental pollutant".
Further, the value of endrin in protecting the nation's
cotton crop, rather than being "significant", is marginal at
best.
-11-
-------
goes on to calculate the amount of "maximally contaminated"
fish—^that a pregnant woman would have to consume "in order
to incur the threshold teratogenic dose" (Velsicol, p.39).
Presentation of the data in this fashion completely ignores
the concept of providing an adequate margin of safety to
prevent susceptible persons from ever receiving a "threshold
teratogenic dose". The Agency must reject any approach to
risk assessment which is premised on the expression of risk
in terms of exposure with no margin of safety associated
with it. Rather, the Agency must exercise its judgment
based on the margins of safety which are afforded by the
levels of exposure that can reasonably be anticipated.
The following table indicates the terato-
genic margins of safety associated with various levels of
consumption of contaminated fish by a 50 kg woman:
Endrm
concentrat ion
(ppm)
Level
of consumption
(qrams)
125
250
500
0.1
7500
3750
1870
0.5
1500
750
375
1.0
750
375
187
2.0
375
187
93
4.0
187
93
46
3/ Velsicol assumes this to be at levels of 0.4 ppm rather
than 1 ppm even though the same consultant elsewhere esti-
mates that sublethal doses may result in bioaccumulation
as high as 2 ppm.
-12-
-------
The Agency's illustration of a MOS of 375
associated with the consumption of 250 g of fish containing
1.0 ppm endrin (PD 2/3, p.58) should be put in the context of
the total array of possible risk situations rather than
isolated as a single point of contention. As indicated m
the above table, the lowest MOS that can reasonably be
anticipated (46) would result from an opportunistic harvest
of fish in the final throes of endrin toxicity that are
consumed in very large quantity, perhaps because refrigeration
is lacking. Such a scenario can reasonably be anticipated
but may not be a very common event-^/. On the other hand,
judging by NPMP data, women consuming fish caught in the
major rivers of the Delta region would commonly be exposed
to endrin residues but seldom at levels providing an MOS of
less than 1000. Between these two extremes lies an area of
intermediate teratogenic risk that is associated with the
consumption of fish from many ponds and streams that are
contaminated by sublethal levels of endrin because of their
proximity to cotton culture. The risks from such exposure
must be considered as unreasonable in light of the low
benefits associated with the use of endrin on cotton.
£/ While pesticides are deliberately used to harvest fish
in some parts of the world, it is more reasonable to hypothe-
size that adults or children may encounter fish in distress
from endrin toxicity, may harvest these fish before they are
dead, and that pregnant women may consume these fish.
-13-
-------
2. Acute Toxicity to Wildlife
Velsicol (pp. 46-51) has commented at some
length on the issue of acute toxicity to wildlife, emphasiz-
ing theoretical reasons why the Agency erred in presuming
the existence of this risk and noting an absence of confir-
matory evidence. Their major argument is that many forms
of wildlife will develop an avoidance response from consuming
sub-lethal levels of endrin. The Agency agrees, in principle,
that, because of behavioral characteristics, certain
individuals or certain species may not be susceptible to
poisoning by endrin. On the other hand, the record indicates
that wildlife kills have been observed from the use of
endrin on wheat in Colorado (Hinkle, 1979); on cotton fields
in California and Alabama, and on alfalfa in California
(Bushong, 1978). It can reasonably be inferred from these
incidents that the foliar application of endrin at any
registered dosage has a potential for killing wildlife,
despite the theorizing of Velsicol's consultant (Velsicol,
Exh. 46). This conclusion is not ameliorated by self-serving
allegations of lack of observed effects which are not
supported by an appropriate investigation or analysis^/.
5/ For instance, Warren Smith (Velsicol, fN 11) has reported
that deer, woodchucks and rabbits continued to thrive a
year after orchards in New York were treated with endrin.
Deer, however, are browsers unlikely to be feeding extensive-
ly on the ground vegetation of orchards, woodchucks should
all be in hibernation at the time endrin is applied, and
rabbits have a high reproductive potential to compensate for
(Footnote Continued)
-14-
-------
The Agency is not aware of adequate surveys conducted by
wildlife biologists that demonstrate the absence of adverse
effects on wildlife from the use of endrin.
3. Population Reduction of Aquatic Organisms
In PD 2/3 the Agency set forth the circumstances
surrounding many events that led the Agency to conclude that
runoff of endrin has been a major cause of the reported fish
kills. Velsicol does not dispute that endrin may have caused
fish kills in the past but persists in maintaining that the
"reported problems arose from misapplication or misuse of
endrin" (Velsicol, p.45). Velsicol's claims of misuse are
purely conjectural and are insufficient to overcome the
presumption of risk—'
Perhaps the most persuasive evidence that the
lethal endrin concentrations associated with many fish kills
arose from normal application practices rather than from
misuse stems from the association of those incidents with
5/ (Con't) any excessive mortality. Mr. Smith's training is
not in the area of wildlife biology and his argument reflects
this lack of expertise. Moreover, the report of the monitor-
ing study of these orchards by wildlife biologists is not
yet available.
6/ Moreover, Velsicol contends that "the empirical data of
sporadic fish kills relied upon by the Agency actually tend
to confirm Dr. Freed's theoretical kinetics" concerning
runoff potential (discussed more fully in Section II C 3,
below) (Velsicol, p.23). Whether or not this is true, this
argument appears to be inconsistent with Velsicol's misuse
argument since Dr. Freed's theoretical kinetics are not
based on misuse.
-15-
-------
toxaphene as well. Velsicol has asserted that "Toxaphene
was identified by the Agency as the causal agent in many of
the very same PERS incidents which earlier had been attri-
buted to endrin" (Velsicol, FN 12). As noted in PD 2/3
(p.22), such a coincidence does not serve to exonerate
endrin. The frequent coincidence of high concentrations of
both endrin and toxaphene, however, is strong circumstantial
evidence that the source of both is runoff since it is
highly improbable that both endrin and toxaphene were
misapplied or misused independently of each other on numerous
occasions at the same time and general area. The Agency
continues to believe that most fish kills that have resulted
from either endrin, toxaphene or both were the result of use
consistent with the label. As discussed more fully below,
the Agency also remains unconvinced that a 1/4 mile barrier
strip will adequately reduce runoff of these pesticides.
4. Fatality to Endangered Species
At the request of the Director of the U.S.
Fish and Wildlife Service (Greenwalt, 1978a) the Agency
reinitiated Section 7 Consultation prior to determining what
final actions to take with respect to endrin. In response,
the FWS expressed the opinion, in essence, that most
of the Agency's proposed regulatory actions and use restric-
tions alleviated the FWS' concern for adverse impacts of
-16-
-------
endrin to threatened and endangered specxes (Greenwalt,
1978b). An exception made by the FWS was that the use of
endrin to control orchard mice was likely to jeopardize the
continued existence o£ the Arctic and American peregrine
falcons and the bald eagle when used in the normal ranges o£
these birds.
At the FIFRA-SAP meeting (SAP, October 26, 1978
pp.102-3) the Agency's attention was directed to an unpubli-
shed manuscript by Stickel et al. (Undated) in which endrm
was identified as the cause of death for two bald eagles.
This manuscript was not previously available to the endrin
RPAR record. The SAP formally recommended that the Agency
address the concerns of the FWS by imposing geographical
restrictions for the use of endrin in orchards (Fowler,
1978). The EDF, relying on the FWS position/ proposed
additional cancellation of endrin uses (Hinkle, 1979).
Following additional investigations among its
staff and consultation with the Agency, however, the FWS
revised its position on the use of endrin to control orchard
mice, concluding in essence that a case for risk to the
Arctic and American peregine falcons and to bald eagles
could not be substantiated by currently available information
(Greenwalt, 1979). Accordingly, the FWS concluded that
the use of endrin to control voles in orchards is not
-17-
-------
likely to jeopardize the continued existence of those
species. Any new evidence, however, could require a
reappraisal of the FWS' opinion.
Velsicol has made several comments on the risks
to endangered species (Velsicol, pp. 54-8) that do not
affect the Agency's position but which require response.
1) Velsicol states, "In view of the evidence
on the Louisiana brown pelican presented by Velsicol and in
view of the conclusion drawn by Director Greenwalt, the
Agency has conceded that the endangered species trigger had
been rebutted successfully (Position Document 2/3, pp.
33-40)". While the Position Document does in fact conclude
that "the risk to endangered species has been rebutted for
the brown pelican", the Agency specifically rejected all
of Velsicol's arguments as the basis for such a conclusion
(PD 2/3, pp.35-8). Further, the probable fatality to two
bald eagles noted in PD 2/3 and the confirmed fatality of
two additional bald eagles introduced to the record by the
SAP are sufficient to maintain the Agency's concern for that
species. However, the Agency has addressed this concern by
requiring that fish killed by endrin be collected and
buried, thus substantially reducing the most likely source
-18-
-------
of exposure to bald eagles. With this new requirement, the
Agency agrees that the continued use of endrin "is not
likely to jeopardize the continued existence of the bald
eagle" (Greenwalt, 1979).
2) Velsicol contests the conclusion of the FWS
regarding the potential for secondary poisoning of raptors
from birds and rabbits that may be exposed to endrin-treated
orchards by citing reports submitted by Warren Smith (Velsicol,
p.57, FN 11). As noted above (FN 5) however, Mr. Smith's
observations on wildlife do not address the issue. The
relevant document on this issue will be the report of the
monitoring program conducted in New York in 1977 and this
report, as indicated in PD 2/3 (pp.61-2), is not yet available
because chemical analyses are incomplete.
*
3) Velsicol cites the opinion of Dr. Howard
regarding hazards of secondary poisoning potential to
falcons and eagles. However, Velsicol specifically addresses
lack of hazard associated with the consumption of orchard
mice (Velsicol, p.58) rather than the non-target birds
and rabbits that are the concern of the FWS.
In summary, the Agency agrees with the revised
position of the FWS in concluding that the available evidence
does not support the conclusion that the use of endrin is
-19-
-------
likely to jeopardize the continued existence of peregrine
falcons and bald eagles and also concurs that the situation
should be reevaluated as new information becomes available.
Thus, the available evidence does not require changes.in the
Agency's position on endangered species.
5. Oncogenicity
The Agency's Carcinogen Assessment Group
(CAG) analyzed the available evidence on the oncogenicity of
endrin and concluded that endrin was unlikely to be a human
carcinogen. Accordingly, in PD 2/3, the Agency took the
position that the oncogenicity "trigger" had been rebutted.
Dr. Melvin Reuber criticized the Agency's conclusion. Dr.
Reuber made an oral presentation to the SAP on October 26,
1978 and submitted written comments to the SAP and to the
Agency (Reuber, 1978a; b). The CAG has submitted written
responses (Anderson, 1978a; b). At the December 14, 1978
SAP meeting, Dr. Reuber and Dr. Roy Albert, Chairman of the
CAG, both discussed the issue of endrin's oncogenicity and
responded to Panel members' questions.
Dr. Reuber's major points are (1) that most of
the oncogenic studies conducted with endrin are invalid for
various reasons and the negative findings reported from such
studies should receive no weight in assessing the oncogenicity
of endrin and (2) that certain of the studies resulted
-20-
-------
in statistically significant increases in tumors associated
with exposure to endrin. The CAG responded that some of the
allegedly positive results involved differences of opinion
among pathologists. For instance, Dr. Reuber's diagnosis of
certain liver tumors in the FDA rat study could not be
confirmed by two other consultant pathologists. Since
Dr. Reuber declined an invitation to participate in a joint
examination of the slides, the CAG accepted the opinion of
the other pathologists. Additionally, Dr. Reuber differed
with another consultant, Dr. Frith, on the relative number
of malignant vs. benign mammary tumors m rats and the CAG's
position was that the total tumor yield, whether or not they
were classified as benign or malignant, was not related to
endrin dosage (SAP, Dec.14, 1978, p.23). While the CAG
acknowledged that most of the available studies had some
deficiences, it also indicated that all contributed some
information and it determined, on balance, that the slight
indication of positive endrin effect found in the FDA rat
study and the Kettering mouse study was insufficient to
indicate that endrin was likely to be a human carcinogen.
(For detailed discussion, see SAP, December 14, 1978).
The EDF also commented on the oncogenicity
issue at the SAP meeting and in a submission to the Agency
(Hinkle, 1979). The EDF argues that respectable scientific
authority is sufficient evidence upon which the Agency may
-21-
-------
rely in a determination and that such authority does not
have to reflect a majority opinion. The EDP alleges that
critical questions regarding the endrin studies remain
unanswered and, especially in light of Dr. Reuber's diag-
nosis, the issue of endrin1s carcinogenicity remains an open
question. Dr. Albert, in response to similar questioning
by the SAP, indicated that Dr. Reuber's opinions had been
given a great deal of attention by the CAG (SAP, Dec.14,
1978, p. 35).
Velsicol has also submitted comments on the
question of oncogenicity and the Agency must respond to
certain of their statements. By incomplete quotation of the
NCI Technical Report Series No. 12 (Velsicol, Exh. 48),
Velsicol has incorrectly implied that the NCI conclusions
were unqualified. Velsicol reported that NCI had concluded
that endrin "was not carcinogenic for...rats or for...mice"
(Velsicol, p.53). The full statement reads, "It is concluded
that under the condition of this bioassay, endrin was not
carcinogenic for Osborne-Mendel rats or for B6C3P1 mice." By
removing the limitations on the conclusion, Velsicol has
improperly generalized its applicability. The NCI clearly
did not speak to the conclusions of other existing tests, to
what might be the case in other strains of rodents, or
to how differences in conditions might have affected the
outcome of the NCI tests.
-22-
-------
Velsicol's comments also incorrectly characteri-
zed both the SAP's and the Agency's positions with respect
to the oncogenic status of endrin. Velsicol states, "More-
over, the Agency's Scientific Advisory Panel reviewed the
evidence of record and concurred in the Agency's conclusion
that endrin is not carcinogenic." (Velsicol, p.53). In
fact, the SAP has made no formal comment on the issue and
thus has neither concurred in nor dissented from the Agency's
conclusion. Moreover, it is the Agency's position only
that the weight of the evidence is that endrin is "unlikely
to be a human carcinogen" (PD 2/3, p. 44); the Agency has
never unqualifiedly concluded that "endrin is not carcinogenic."
The Agency recognized that there was some evidence suggestive
of oncogenic effects but that the evidence as a whole
supported the conclusion that endrin is unlikely to be a
human carcinogen. That there was insufficient evidence for
the Agency to consider endrin a probable human carcinogen
does not mean	as -Velsicol suggests—that the Agency has
made an unqualified negative finding.
B. Comments Relating to Benefits
1. Cotton
The USDA (Bergland, 1978) comments:
"We believe the cancellation of
uses on cotton in the Southeast and
Delta will have limited economic
impact at the present time. However,
a number of entomologists in the
-23-
-------
cotton producing States have pointed
out that the use of endrin on cotton
could become critical if current
alternative pesticides are lost
through, the regulatory process,
reduced market availability, or
become less effective".
In response, the Agency is pleased that the
USDA concurs with the Agency's analysis of the present
benefits of endrin on cotton which provides the basis for
its regulatory decision. If future events indicate a
substantial change in the risk/benefit picture, at that time
the Agency can reconsider its regulatory decision.
Velsicol has commented extensively on the
essentiality of endrin in Integrated Pest Management Programs
(IPMP), especially for control of the bollworm complex, and
on the economic impact of an endrin cancellation (Velsicol,
pp. 60—74). Velsicol states that experts in Alabama,
Arkansas, and Mississippi have presented data which indicate
that bollworm resistance to endrin in those states is not
currently a widespread or major problem (Velsicol, pp. 62
and 66) but that it remains effective on light to moderate
infestations. Judicious use of endrin, allegedly, can retard
the development of resistance to newer pesticides. Velsicol
offers some recent data on treatment costs indicating that a
replacement of endrin with the most likely non-RPAR pesticides
would result in a total increase in costs of ?1,436,116
rather than the Agency's estimate of $717,850.
-24-
-------
The Agency's perception of the benefits from
the use of endrin on cotton differs only in detail from
Velsicol's. The Agency's understanding is that bollworm
resistance to endrin per se is virtually complete (Lincoln,
1979; EPA, 1977) and therefore, as Velsicol indicates,
endrin is currently formulated for use on cotton mainly in
combination with methyl parathion. Moreover, Velsicol
admits that the combination is generally efficaceous only
for light to moderate infestations. While the Agency
concedes that some benefits could be derived from the use of
endrin on cotton, Velsicol's claim of essentiality is hardly
supported by the dramatic decline in use in Mississippi,
"from approximately 275,000 pounds in 1973 to approximately
75,000 pounds in 1977, and even less in 1978" (Velsicol,
Exh. 18). The Agency agrees in principle that the use of a
variety of pesticides may be desirable to retard the develop-
ment of resistance and that endrin could theoretically
continue to play some role in IPM programs. Velsicol's
claims that endrin is essential for this purpose, however,
are not supported by the record.
Velsicol's conclusion that a cancellation of
endrin would increase costs by $1.4 million rather than the
Agency's estimate of $0.72 million may well reflect current
costs of pesticides more accurately than did the Agency's
-25-
-------
analysis which was based on 1976 values. Even so, the
impact projected by Velsicol is probably exaggerated for
several reasons. The RPAR compounds EPN and toxaphene
were not considered as viable substitutes by Velsicol.
While some regulatory action may be taken on toxaphene, the
RPAR on EPN has not been issued yet and, while it is too
early to predict, EPN may well remain a viable substitute
for endrin on cotton beyond the immediate future, thus
lessening the long term economic impact of an endrin cancel*
lation. Additionally, Velsicol's analysis fails to account
for the reduction in benefits that would result from the
regulatory action it would prefer, imposition of a 1/4 mile
distance restriction from water (and human habitation) in
the states east of Interstate Highway 35 (1-35). Even under
Velsicol's analysis, however, its conclusion with regard to
endrin's economic importance to the cotton growing industry
is remarkably similar to the Agency's and the Agency continues
to maintain that endrin is only of minor significance to the
cotton industry.
2. Small Grains
The USDA (Bergland, 1978) has criticized the
Agency for an alleged failure to give sufficient consideration
to the possible benefits of relatively insignificant uses of
endrin for which economic data may be lacking, such as the
use of endrin to control chinch bugs. With respect to the
-26-
-------
use of endrin on chinch bugs, however, public response to
the proposed decision set forth in PD 2/3 indicates that
the Agency was correct in attributing little or no economic
value to endrin. Dr. Leroy Brooks (Kansas), the only
proponent of the use, did not renew his appeal for retaining
that use after the issuance of PD 2/3, although he submitted
other comments to the USDA (Brooks, 1978). No other comments
(including those from Velsicol) were received on chinch bugs
and the OSDA did not specify other uses with respect to
which USDA believed the Agency incorrectly assessed the
benefits of endrin.
Velsicol's comments with regard to the use of
endrin on small grains contain substantial errors of fact.
Regarding Kansas, Velsicol claims that 1,200,000 acres are
treated with endrin (presumably for control of pale western
cutworm) and, for Oklahoma, 2,000,000 acres are said to be
infested and treated for army cutworm (Velsicol, pp. 76-7).
These values are inconsistent with the cited references,
with the Agency's estimate of usage, and with Velsicol's
own production and sales figures. Regarding pale western
cutworms in Kansas, Dr. Brooks stated that, "Some localized
infestations requiring treatment of 10 to 20 thousand
acres occur every two to three years. Large scale outbreaks
... that would necessitate treatments over a much larger
-27-
-------
area (possibly up to a million acres) could occur..."
(Velsicol, Exh. 58). Regarding Oklahoma, Dr. Coppock
reported that, "Between five and six million acres of winter
wheat were sprayed for the greenbug [emphasis added] and
army cutworms during that time [1976] "(Coppock, 1976) and
Velsicol's estimate can not be derived from that reference.
The Agency's estimate for all states combined was an annual
average of 416,000 acres for pale western and 691,000 acres
for army cutworms. Velsicol's own production estimates
indicate a range in usage on small grains from 201,000
pounds in 1976 to 25,000 pounds in 1977 (PD 2/3, p.6). At
3-4 oz• active ingredient of endrm per acre, these amounts
would have treated from a maximum of one million acres in
1976 to a minimum of one hundred thousand acres in 1977.
Thus, the Agency's analysis may have over-estimated the
benefits of endrin's use on small grains somewhat but
Velsicol's claims of treated acreage can not be remotely
supported by their own production and sales figures, which
were provided by Velsicol to provide a more accurate
picture of recent usage.
Velsicol also notes the need for endrin to
control grasshoppers on small grains but stated, "This use
is not discussed herein because presently endrin is registered
for use on grasshoppers only in Montana" (Velsicol, p. 75,
-28-
-------
ra 23) . The Agency noted that it had no data which could be
used to evaluate the benefits from the use of endrin to
control grasshoppers (PD 2/3, p. 143) and the EDF has
protested that there are many available substitutes registered
for grasshopper control in small grains and rangeland (Hinkle,
1979). However, endrin is not registered for use in
rangeland and several of EDF's alternative pesticides are
not registered for use on small grains. PD 2/3 (p. 39)
incorrectly listed phorate as an alternative since it is
recommended as a border treatment in Montana. To clarify
the record, the only federally registered alternatives for
the use of endrin on wheat are malathion, parathion, methyl
parathion and toxaphene. Nosema locustae is also registered
but its usefulness in the field is not. well established.
Velsicol has submitted new data on the compara-
tive efficacy of insecticides to control grasshoppers
in Oklahoma (Coppock and Pitts, undated). The results of
these tests indicate that malathion, parathion and toxaphene
gave excellent control under the test conditions.
Velsicol has also stated, "Endrin is applied at
the exceedingly low rate of 3-4 oz. a.i./acre, and this is an
environmental benefit of usage" (Velsicol, p. 77). The
Agency does not believe that the circumstances of dosage
rates constitute an "environmental benefit" in the normal
sense of that term and Velsicol"s unsupported claim to that
effect is rejected.
-29-
-------
3. Apple Orchards
Velsicol's comments on the benefits of endrin's
use in apple orchards contains some erroneous points critical
of the Agency's Benefit Analysis (EPA, 1977). Velsicol
states that the Agency inappropriately focused upon total
apple acres when assessing endrin, and infers that the
Agency was naive in not recognizing the "economically .
disastrous" effects of voles in affected orchards, Velsicol
also states that the Agency's national estimates of the
effect of an endrin cancellation upon apple supplies is
"misleading". Further, Velsicol charges "that the Agency
has attempted to minimize the vole problem". Velsicol
concludes by asserting that the Agency should have limited
its analysis to vole-infested orchards and should have
extended the analysis beyond the three-year time horizon
presented in the Benefit Analysis (Velsicol, pp.78-9).
These assertions center around two factors:
1) Velsicol is apparently unaware of the Agency's policy
to analyze the effects of its intended regulations at all
levels, and 2) has misunderstood the Agency's Benefit
Analysis. The Benefit Analysis includes an assessment
of market (domestic apple supplies and prices), consumer
(retail price effects), and producer (production, farm level
price, revenue, and cost changes) impacts.
-30-
-------
The inclusion of the effect of an endrin
cancellation upon national apple supplies and prices is
neither "misleading" nor an "attempt to minimize the vole
problem" but an Agency requirement. The Endrin Benefit
Analysis for Orchards also includes an extensive discussion
of producer-level impacts, including analysis of apple
losses, production cost changes, farm level prices, and farm
revenues. The cost-effectiveness of endrin and various
alternative control strategies used by apple growers
are also discussed at length. Since the effects of an
endrin cancellation may extend beyond current users, the
Agency was prudent in not limiting the analysis to owners of
vole-infested orchards.
With regard to the three-year time horizon
*
used in the report, the Agency is fully aware that impacts
could extend beyond this period. However, available data do
not permit an accurate assessment of the economic effects
likely to occur in the long term (EPA, 1977 pp. 74-5, 8Q-1,
84-5).
C. Comments Relating to Regulatory Options
1. Designation of Target Species in Apple Orchards
The USDA (Bergland, 1978) notes, "The use of
the word 'only' in identifying the vole species to be
controlled in apple orchards may cause unnecessary enforce-
ment problems when more than one species is established in
-31-
-------
an orchard. We assume that the destruction of eastern
meadow voles incidental to the control of pine voles would
not be considered inconsistent with the labeling." Addi-
tionally, Dr. Don Hayne (personal communication, Nov. 14,
1978) has noted that the Agency's use of the terms "eastern"
and "western" meadow voles has no basis in scientific
nomenclature. Dr. Ross Byers (1978) has indicated that the
prairie vole (Microtus ochroqaster) has behavioral charac-
teristics similar to that of the pine vole and that the need
for endrin to control this species in the mid-West should
be investigated.
Having received no additional information,
the Agency is unable to address Dr. Byers' concern. The
Agency agrees with Dr. Hayne's point. Accordingly, labels
for use in apple orchards should designate the pest species
as follows:
Eastern United States: Pine Voles (Microtus pinetorum)
Western United States: Meadow Voles (Microtus species)
The distribution of commercial apple growing areas is such
that the broad geographical limits do not pose problems of
interpretation.
The OSDA is correct in its assumption that
the destruction of meadow voles incidental to the control of
pine voles from the use of endrin in eastern orchards
containing both pine voles and meadow voles would not be
-32-
-------
inconsistent with the label. The use of the word "only" on
the label is necessary, however, to preclude the use of
endrin where it is intended to control meadow voles rather
than pxne voles in the East. The presence of pine voles
in an orchard in the East may not be used as a pretext for
the use of endrin intended only to control meadow voles.
2. Equipment
The USDA commented that, for aerial application,
wind velocities should be stated as the maximum allowed (i.e.
10 mph) rather than as a range (i.e. 2-10 mph) and that the
flying heights should be the same for all crops. Additionally,
/
the State of North Carolina recommended that the label
specify "apply only with ground equipment" for use in apple
orchards (Blaylock, 1978).
In specifying a minimal wind velocity for
aerial application, the Agency is following the recommenda-
tion of Velsicol's Expert Panel (Akesson, 1977) and believes
that this represents sound advice for controlling drift
because it is supported by empirical data and rational
analyses. The variable height of application referred to by
the USDA stems from a typographical error. The maximum
height for aerial application should be 10 feet above all
crops. While it is unlikely that anyone would attempt to
control voles by treating apple orchards with endrin by air,
such an attempt would be extremely hazardous. Accordingly,
such application will be prohibited by a label restriction.
-------
3. Distance Restriction from Aquatic Habitats
a. Cotton Osage Bast of Interstate Highway 35
The Agency concluded in the Sndrin PD 2/3
that the hazard o£ endrin to fish arises from transport
to water by both drift and runoff. It concluded that a
distance restriction can substantially reduce endrin
contamination of water resulting from drift, but that no
information was available to assess the impact of a distance
restriction on the reduction of contamination from runoff.
In response to the recommended regulatory option in PD 2/3
to cancel endrin use on cotton east of Interstate Highway
35, Velsicol submitted an extensive discussion defending
the efficacy of a 1/4 mile restriction from water bodies in
diminishing endrin runoff to water in the southeastern
United States. The Agency has already engaged in several
exchanges of comments on the runoff question with Velsicol
and its consultant, both prior to and during the SAP proceed-
ings (Velsicol Exh. 21,22,23,25; Severn, 1978; SAP, October
26, 1978). The Agency will now respond to Velsicol's
summary presentation of the issues which it made in its
comments on Position Document 2/3 (Velsicol, pp.17-30). In
Velsicol's summary, the following points were made:
(a) endrin is strongly adsorbed to soil
particles at the application site;
-34-
-------
(b)	endrin has a "comparatively short
residual life" in the environment;
(c)	not more than 1% of applied endrin
would be carried to the edge of a treated plot by an intense
rainfall;
(d)	an intervening 1/4 mile of bare
cultivated soil would reduce runoff concentrations of endrin
to 1% of this 1%, or 10.01% of the amount applied;
(e)	vegetation in the barrier strip would
further reduce the runoff by another factor of 10;
(f)	maximum concentration in a pond
containing two acre-feet of water resulting from application
of 1.25 lbs of endrin to one acre separated from the pond by
a barrier strip covered with vegetation would be approxi-
mately 2 ppt;
(g)	the efficacy of the distance restric-
tion has been demonstrated by the reduction in fish kills
observed in Arkansas after the imposition of a distance
restriction;
(h)	the acceptability of the distance
restriction is also demonstrated by its imposition m
Mississippi as part of an emergency exemption for synthetic
pyrethroid application to cotton;
-35-
-------
(i) monitoring data from Alabama confirm
the efficacy of intervening land in decreasing endrin
residues in ponds;
(j) distance restrictions imposed by
the United States Forest Service for pesticide applications
in forest areas also support the efficacy of distance
restrictions in reducing contamination of adjacent waters?
and
(3c) in summary, the evidence that a
quarter-mile barrier would render endrin runoff from
southeastern cotton fields innocuous is overwhelming, and
warrants revision of the Agency's preliminary recommendation
to cancel the cotton use in the Southeast.
The Agency's response will discuss each of
these points in order.
(a) Adsorption of Endrin to Soil
The Agency accepts the view put forward by
Velsicol that endrin may be strongly bound to soil or
suspended sediment. It is generally agreed (Fionke and
Chesters, 1973; USDA/EPA, 1976) that compounds which are
strongly adsorbed will move mostly on sediment particles.
Thus the major mode of runoff transport of endrin is probably
through erosion processes.
-36-
-------
Since endrin is applied as a foliar spray,
rather than directly to the ground, a potential problem
appears to be washoff from the foliage soon after applica-
tion. The record o£ endrin-related fish kills appears to be
correlated with rainfall incidents. Estimates of the amount
of pesticide deposited on foliage from aerial application
vary, but 50% on foliage and 30% on the ground appears to be
a reasonable estimate. Sparr et al. (1966) observed a
concentration of endrin in runoff water during a rainfall
event seven days after application which was higher than
that found during irrigation prior to the rain and stated:
We believe that this higher endrin
concentration resulted from washing the
endrin off the foliage.
While there are major flaws in this study (the particulate
fraction of the runoff was apparently not analyzed, although
this fraction would be expected to contain most of the
endrin, as noted above}, the study at least suggests that
foliar washoff during a rainfall event is an additional
consideration in evaluating the overall extent of endrin
transport by runoff.
Another study indicating that foliar washoff of
pesticides from cotton makes an important contribution to
runoff was recently reported by Willis et al. (1976). These
-37-
-------
workers applied toxaphene and other pesticides to cotton in
a nearly flat watershed equipped with instrumentation to
measure surface runoff and sediment and chemical yields at
the point where runoff entered a four-acre pond. They found
a total of 0.038 lbs/acre of toxaphene in runoff during the
period from August to February (a period of low sediment
yield); a total of 9 lbs/acre of toxaphene had been applied
in August and September. They concluded that the freshly
sprayed leaves were an important source of toxaphene m
runoff in August and September. These workers also observed
that:
Current cultural practices in the
Mississippi Delta may be intensifying
sediment and chemical transport from
agricultural fields. After harvest, many
farmers shred plant residues, till the
soil, and form rows. The fields are
left with little or no vegetative
cover throughout the winter and early
spring, and are subject to the erosive
forces of rainfall and runoff until
adequate cover develops.
Since, as noted above, endrin is bound to
soil particles, this study suggests that substantial runoff
transport of endrin may occur under current cotton cultural
practices.
(b) Environmental Persistence of Endrin
Velsicol concluded that endrin has a
comparatively short residual life in the environment.
Persistence on foliage or soil is an important issue, since
-38-
-------
the longer a chemical resides at the site of application,
the more opportunity there will be for runoff events to
occur. The Agency realizes that persistence is not an
important factor for runoff events which occur immediately
following application, but it is, of course, concerned with
all runoff events occurring subsequent to application.
Velsicol cited studies on endrin photodegrada-
tion (Baker and Applegate, 1974) and soil metabolism
(Castro and Yoshida, 1971; Matsumura et al., 1971) m
support of its conclusions with regard to endrin persistence.
The study by Baker and Applegate used blacklight lamps to
irradiate thin films of endrin and other pesticides on glass
in the laboratory; they reported a 10-30% photodecomposition
of endrin in 20 hours, compared to dark controls. This study
has little utility for evaluating the environmental photode-
gradation of endrin, since it presents no data on the
photochemistry of soil-bound endrin. It is likely that
bound endrin would be much less accessible to sunlight and
in addition might be inherently less photoreactive. The
artificial light source employed also makes this study less
valid. The claim that mirex was photodegraded suggests
that the emission spectrum of the lamps used extended to
well below 290 nm (the lower limit of natural sunlight),
since mirex has virtually no absorption afiove 250 nm and no
photoreaction could be detected using natural sunlight
-39-
-------
(Alley et al./ 1974). In summary, the information presented
by Baker and Applegate may not be used as a reliable indicator
of the environmental photodegradation of endrin.
The soil metabolism study of Castro and Yoshida
(1971) was performed in flooded and upland soils in the
Philippines. Endrin was found to degrade rapidly in a
flooded soil but was in fact quite persistent (88% re-
covered after two months) when the same soil was maintained
at 80% of the maximum water-holding capacity. In any event,
this study is of dubious utility in evaluating persistence
of endrin in the soils of the southeastern United States. A
monitoring study performed in 1966 in Greenville, Mississippi
(QSDA, 1968) found high residues of endrin in soil more than
one year after treatment. Soybeans planted in these soils
had endrin residues resulting from translocation. While the
studies of Matsumura et al. (1971) showed that 25 of 150 soil
cultures had the capacity to degrade endrin in laboratory
culture, it is clear that endrin can be sufficiently persis-
tent in the southeastern United States to survive a winter
season.
(c) Estimates of Extent of Runoff From
Treated Fields
Velsicol concluded that, as a worst case,
not more than 1% of the endrin applied would be carried to
the edge of a treated plot by soil erosion. In support of
-40-
-------
this conclusion, a limited number of controlled runoff
studies were cited, in which the total amount of pesticide
leaving the field was measured to be less than 1%. However,
two recent reviews (Pionke and Chesters, 1973; Leonard et
al., 1976) have compiled a much larger number of such runoff
studies? the overall range of extent of loss varied from
0.007% to approximately 40%, with 11 studies reporting losses
in excess of 1%. These studies encompassed a wide range of
conditions (type of pesticide and application conditions,
rainfall characteristics, type of soil, crop, slope, etc.),
all of which strongly influence runoff, as noted by Velsicol.
In addition, a very recent study (Smith et al., 1978) used
paraquat as a tracer compound for estimation of sediment
transport in a Southern Piedmont watershed. When applied to
the soil surface, runoff losses of paraquat commonly exceeded
5%. Although no cover crop was present in this case, it
appears that sediment transport of bound pesticides can be a
reasonably effective process. Precise predictions of the
behavior of endrin when applied to cotton in the Southeast
may not be made based on data currently available. However,
based on the studies which are available, the Agency concludes
that a 1% runoff yield, while reasonable some of the time, is
certainly not a "worst case". If it were necessary to
establish such a "worst case", 10% would probably represent
an upper limit of runoff transport from treated fields for
most situations.
-41-
-------
(d) Efficiency of a Barrier Strip in Reducing
Runoff
The Agency in PD 2/3 concluded that no informa-
tion was available on which to base a quantitative estimate
of the efficacy of intervening land in reducing the runoff
potential of endrin. This conclusion derived in part from
the observation that quantitative runoff studies (as discussed
in (c) above) commonly measured runoff immediately adjacent
to the treated field. This point was also made by Velsicol.
However, the summary document submitted by Velsicol also
states that:
...on the basis of this worst-case
runoff model, Dr. Freed calculated that
the quarter-mile barrier (assuming it
was bare-cultivated) would reduce runoff
concentrations of endrin to 1% of what they
would be under similar worst-case circum-
stances with no_ barrier strip... (Velsicol, p.25)
The calculation referred to above is the use of
the Universal Soil Loss Equation and a sediment delivery
ratio equation to calculate the amount of chemical in
overland runoff; the value computed was 0.0127 pounds. The
values of the input parameters for the equations are not
presented, nor is the manner of carrying out the calculation.
In any event, these calculations are not based on any field
experiments with endrin, despite Velsicol's contention that
detailed data and other information have been provided to
the Agency to evaluate endrin runoff.
-42-
-------
There is no question but that intervening land
areas can have the effect of reducing sediment runoff and
thus sediment-bound pesticide transport. However, erosion
continues as a major problem; for example, an annual sediment
yield of 11.6 tons per acre was measured on nearly flat land
in the Mississippi Delta (Willis et al., 1976). A general
equation for the amount of sediment transported overland is
apparently not available. Values of the ratio of sediment
transported from a specific area by erosion to the amount
received by a body of water range from about 0.1 to 0.3
(USDA/EPA, 1975). This report also concluded that:
The sediment discharged to large rivers
is usually less than one-fourth of that eroded
from the land surface.
Obviously, this amount will vary with rainfall intensity
and previous surface conditions, as well as the distance
over which it is transported. A major runoff event may
also pick up sediment deposited during prior runoff events
(USDA/EPA, 1976).
In conclusion, the Agency's perception is
that it is not possible to predict the extent of overland
transport of endrin by erosive processes because of the
variable nature of these processes and thus, the efficacy of
a barrier strip in reducing endrin runoff cannot be
predicted. The Agency concludes that Velsicol's contention
-43-
-------
that a "worst case" of endrin transport across a 1/4 mile
barrier strip is 1% of that leaving the treated field is not
justified since it is not clearly supported by any available
information.
(e) Effect of Vegetation in Attenuating
Runoff
Velsicol concluded that vegetative cover on the
proposed barrier strip would further reduce endrin transport
by a factor of ten. An exploratory survey (Moubry et al.,
1967) of endrin runoff through heavy turf in a Wisconsin
orchard was cited, in which no endrin was detectable in
runoff water; the water was observed to be devoid of silt.
The relevance of this study to cotton runoff is questionable,
since it does not appear that cotton fields generally are
surrounded by heavy turf. The observations of Willis et al.
(1976), quoted above, suggest that very little vegetative
cover may be available throughout much of the year in cotton
culture. For the cover and management factor appearing in
the Universal Soil Loss Equation (OSDA/EPA, 1975), a value
of 0.34-0.4, corresponding to about 60% reduction in
sediment yield, is a generalized value for cotton. Loss
Equation (USDA/EPA, 1975). Thus, a factor of perhaps
two or three, rather than ten, appears to be a reasonable
estimate of the effect of vegetative cover on a barrier
strip in reducing endrin runoff.
-44-
-------
(£) Calculation of Maximum Sndrin
Concentration
Based on its estimates of endrin runoff from a
treated plot, across a barrier strip, and through vegetation,
Velsicol calculated that the maximum concentration in a two
acre-foot pond located 1/4 mile away from a single treated
acre would be 2 ppt. As discussed above, the Agency
does not accept these three estimated runoff percentages,
or, therefore, the calculated pond concentration based on
them. Moreover, the use of a single acre as a plot size is
particularly unreasonable; clearly, many acres of cotton
could be treated in a single watershed. As noted by Leonard
et al. (1976):
The pesticide load in runoff and on
sediment times the areal extent of usage
is the pesticide dosage entering the
receiving water.
The Agency believes that integrated sampling of
a watershed area, in which all of the runoff is channeled
through a flume and sampled continuously, is the only
reliable way to quantitate pesticide losses m runoff. The
studies by Willis et al. (1976) and Smith et al. (1978) are
examples of such studies. In the absence of adequate data
of this nature, the Agency can not reasonably conclude
that Velsicol's calculation of maximum endrin concentration
in receiving waters is supported.
-45-
-------
(g)	The Arkansas Distance Restriction
Velsicol stated that a reduction in fish kills
in Arkansas following imposition of a 1/4 mile aerial
application distance restriction from commercial fish ponds
and hatcheries demonstrates the efficacy of such restrictions
in diminishing runoff transport. Arkansas and Mississippi
are areas of intensive commercial catfish farming. Crockett
et al. (1975) sampled catfish from SO farms in 1970 and
reported that 76% of the fish samples contained endrin.
They concluded that aerial transport of endrin from nearby
cotton areas was the most probable route of contamina-
tion. They also observed that commercial fish ponds are
generally constructed to prevent the entry of surface
runoff. Thus it appears likely that the reason that the
imposed distance restriction resulted in a decrease in fish
kills, to the extent that those data are accurate and
completer was because drift was the main source of contami-
nation of the commercial fish ponds. The Agency accordingly
concludes that the alleged success of the distance restric-
tion in Arkansas does not answer the question of reduction
in runoff transport.
(h)	The Emergency Exemption in Mississippi
Velsicol claimed that the imposition of a
1/4 mile distance restriction for an emergency exemption
involving synthetic pyrethroids in Mississippi and the
-46-
-------
resulting lack of fish kills demonstrates the efficacy of
the restriction in diminishing runoff. A 1/4 mile restriction
is commonly imposed as a condition of an emergency exemption
use as a precautionary measure to reduce aquatic contamination
while adequate data are being developed for registration
purposes. However, the restriction is not imposed on the
basis of any particular data or information regarding
environmental transport or the effectiveness of a barrier
strip in reducing transport. The fact that a distance
restriction was invoked for an emergency exemption does not
necessarily establish its efficacy as a condition of permanent
registration. Additionally, Velsicol has not submitted any
monitoring data to support their allegation of lack of fish
kills.
(i) Alabama monitoring data
The Alabama monitoring data (Elliott, undated)
reported a wide range of endrin residues in pond water,
sediment, fish, soil, forage, rats and birds. Endrin
treatment history was not reported, so that correlation
between endrin use and resulting environmental residues is
not possible. This study was not designed to evaluate the
efficacy of the distance restriction in diminishing endrin
residues in water, and no conclusions regarding the efficacy
can be drawn from the study.
-47-
-------
(j) Distance Restrictions in Forest Areas
The United States Forest Service employs
distance restrictions for pesticide applications in forest
areas. Since vegetative cover and soil surface conditions
in forest areas are entirely different from those expected
adjacent to southeastern cotton fields, the Agency concludes
that distance restrictions used m the forest have no
relevance to cotton agriculture.
(k) Summary
Velsicol stated that there is overwhelming
evidence that a 1/4 mile distance restriction would
render innocuous any endrin runoff from southeastern cotton
fields. The Agency has reviewed all the information submitted
by Velsicol concerning this issue, as well as additional
information cited above. The Agency concludes that endrin
transport to water by runoff would still be a substantial
possibility if the distance restriction were to be imposed,
and that no reliable information is available to insure that
the attenuation of this transport by a barrier strip would
consistently be of the order of magnitude suggested by
Velsicol. Accordingly, the conclusions submitted by
Velsicol on the issue of runoff cannot be considered adequate
to support its proposal of allowing endrin use on cotton m
the southeast subject to a 1/4 mile distance restriction
from water.
-48-
-------
Additionally, Velsicol has stated that its
proposed distance restrictions from bodies of water "would
reduce runoff to innocuous levels even under worst-case
circumstances"; that "the Agency has acknowledged the
validity of this point with respect to small grains regions
and western cotton regions where heavy rainfall is infrequent
and that "The Agency also agrees with Velsicol that a
similar distance restriction of 50 feet is appropriate for
the apple orchard use" (Velsicol, p.24). These statements
totally distort the Agency's position on the effectiveness
of barrier strips in reducing runoff. In all cases where
the Agency has proposed that a distance restriction be
imposed, the purpose was to reduce drift to acceptable
levels relative to the perceived benefits of usage.
In the case-of orchards, at the very place cited by Velsicol
(PD 2/3, p.157), the Agency stated that:
Major risks to fish and wildlife
would remain because of the high applica-
tion rate to the terrestrial habitat and
because the potential for runoff would be
little affected by a distance restriction
of 50 feet.
The Agency has never accepted Velsicol's
contention with regard to barrier strips as an effective
means of reducing runoff and objects to Velsicol"s inaccu-
rate representation of the record.
-49-
-------
b. Small Grains
In response to the Agency's proposal to
permit applications of endrin adjacent to ponds owned by the
user, Velsicol has repeated its proposal to prohibit applications
within 1/4 mile of all lakes, ponds, and streams (Velsicol,
p.90). In PD 2/3 the Agency presented its rationale for
excepting ponds owned by the user: as a matter of policy,
the farmer should have the right to choose between risking
his fish and protecting his wheat (PD 2/3, p.145). Velsicol
has given no reason for denying the farmers that option, and
Agency sees no reason to change its position.
Dr. Leroy Brooks (1978) has recommended that
the distance restriction be reduced to 1/8 mile if endrin is
applied by ground equipment. Dr. Brooks' recommendation is
consistent with the intent of the regulations." Drift from a
boom ground sprayer two feet above the wheat will travel
less than half the distance than will the drift from an
airplane at an elevation of 10 feet if both have similar
nozzles and pressures. Therefore a 1/8 mile distance restric-
tion is appropriate for such ground equipment and will be so
indicated on the label.
4. Distance Restriction from Human Habitation
The SAP questioned the basis for the Agency's
proposal to prohibit application of endrin within 150 yards
of human habitation (SAP, Dec. 15, 1978 pp.5-6) and recom-
mended that the distance be extended to 1/4 mile from
-50-
-------
human habitation (Fowler, 1978). Velsicol supported the
imposition of a quarter-mile restriction. The basis for the
Agency's proposal, set forth in PD 2/3 (p. 128), was that
the MOS for teratogenic risk estimated for a distance of 150
yards is ample (5500). Neither the SAP nor Velsicol demon-
strated any deficiency in the Agency's assessment. If the
MOS is ample, it would be unreasonable to restrict further
the economic benefits to the user.
The Agency concedes that its risk estimate
assumed compliance by the applicator regarding equipment,
wind speed and other restrictions, in the absence of full
compliance, the calculation of the MOS would be in error by
an unknown amount. Therefore, in consideration of the
recommendations of both the SAP and Velsicol and consistency
in the specification of distances on labels, the Agency will
compromise its position and direct that this restriction be
modified to read "1/8 mile" (220 yards) instead of 150
yards.
5. Posting of Contaminated Ponds
In the event of a fish kill, the Agency pro-
posed that the pond be posted "Contaminated: No Fishing"
for a period of one year. Velsicol characterizes this
warning as "inadequate" and indicates that a more appropriate
warning would be as follows: "Contaminated: Use of this
-51-
-------
Water For Drinking, Pishing, Swimming or Other Recreational
Purposes Is Prohibited" (Velsicol, p.90). Velsicol *s
position, however, is unsupported by any analysis. A direct
overspray was estimated to produce a concentration of
0.009 mg/1 (9 ppb) in a pond 2 feet deep (Velsicol, Exh. 5).
Were a woman to drink as much as a gallon of water containing
10 ppb endrin, the MOS would be 2000. The Agency does not
consider Velsicol's concern for this risk to be well founded
and will not consent to the addition of misleading warnings
on labels.
Other issues concerning posting have also been
raised. One is whether the duration of posting should be
less than a year. Another is whether posting is necessary in
situations where contamination is likely but the level is
below that which kills fish. The latter issue was raised at
the SAP meeting (SAP, October 26, 1978, p.121) but was left
unresolved. Unfortunately, there is no ample body of field
data to provide a basis for setting safety standards. It
seems reasonable and prudent, therefore, to require that, if
treatment is made at distances closer than 1/4 mile by air
or 1/8 mile by ground from ponds owned by the user, that
such ponds be posted for a period of 6 months if no fish are
killed and 12 months if a fish kill occurs. In any case,
fishing may be permitted if laboratory analysis indicates
-5 2-
-------
that endrin concentration in the edible portion of fish
do not equal or exceed 0.3 ppra (which is the current FDA
Action Level for endrin residues in fish), since the MOS for
a pregnant woman consuming 315 g (11 oz.) of fish contaminated
at this level is 1000. These restrictions may be revised
when a body of data regarding residue reduction in the field
becomes available.
6. Teratogenicity Warning
The Preliminary Determination proposed that
appropriate endrin labels bear a "Warning to Female workers"
that "Excessive Exposure to Endrin May Cause Birth Defects".
Velsicol opposes the inclusion of such a warning on endrin
labels. (Velsicol, p.88). The bases for Velsicol's oppo-
sition and the Agency's responses are:
a. The margin of safety for applicators is 300
and this is three times the acceptable level. Barring
accidents and assuming that they follow label instructions,
applicators and other workers are at little risk from the
teratogenic effects of endrin. One of the purposes of the
warning, however, is to insure that vulnerable female
workers are aware of the potential risks so that they may
exercise the appropriate precautions and respond properly to
accidental exposure. One drop of a 19.7% EC formulation
contains 10 mg of endrin and 10% absorption of that drop
-53-
-------
provides an MOS of only 75 for teratogenic effects. Certainly
such potential exposure should be of substantial concern.
Further, as discussed above, the Agency does not conclude
that an MOS of 100 is "acceptable" for all teratogenic
risks.
b. Also, the phrase "excessive exposure to
endrin may cause birth defects" is factually inaccurate
because exposure to threshold teratogenic levels of endrin
would cause acute toxicity or death to humans before such
exposure could cause birth defects (see p. 41 of this
response and Exhibit 43). The references cited do not
elaborate on the above issue but only indicate that endrin
may cause single convulsions in humans at dosages of 0.20 to
0.25 mg/kg and multiple convulsions at 1 mg/kg. It is the
Agency's position that, since teratogenic effects in the
hamster were observed at doses which did not produce convul-.
sions or other overt signs of toxicity, the same relative
relationship may exist for humans. That is, a teratogenic
hazard in humans may occur before any toxic warning signs
are observed. Velsicol's argument would have some validity
only if terata in test animals were associated only with
severe toxic effects in the dams, which is not the case
with endrin. In addition, even if endrin did cause acute
toxicity in humans at doses below the teratogenic threshold,
the teratogenic concern would not thereby be eliminated.
-54-
-------
c.	Furthermore, any such theoretical hazard
would only apply to pregnant pesticide applicators ... and
only during the early months of pregnancy. The Agency
agrees that only pregnant women are at risk of birth defects
but the period of vulnerability has not been established.
The warning should be modified to read ""Excessive exposure
to endrin during pregnancy may cause birth defects".
d.	If a teratogenicity warning is warranted for
a weak teratogen such as endrin with only a remote likelihood
of exposure to pregnant women, strong teratogens to which
women are commonly exposed ... should contain teratogeni-
city warnings as well. The Agency does not agree that
endrin should be characterized as a "weak" teratogen^/. It
agrees, in principle, that many compounds should bear
teratogenicity warnings and intends to enforce that prini-
ciple when appropriate.
7/ Velsicol's characterization of endrin as a "weak" teratogen
apparently derives from statements made by one of its
consultants regarding the Chernoff study (Velsicol, Exh.30,
p. 8 and SAP, October 26, 1978 p.51 and Velsicol, Exh.32).
At the SAP meeting, Velsicol's consultant stated, "....
there is some teratogenic potential, albeit a low level for
this compound [endrin], but that potential occurs
only at maternal toxic levels or very near to maternal
toxic levels" (SAP, p. 51). The basis for this position has
never been made clear. Previously, the consultant stated,
"A single dose as high as 10 mg/kg produced no maternal
toxicity and had no effect on intrauterine mortality or
growth of the offspring. Two types of malformations ... were
(Footnote Continued)
-55-
-------
7. Protective Clothing for Workers
The tJSDA commented, "We question the advisabi-
lity of requiring protective clothing for all female workers.
The teratogenicity risk, as defined, should apply only to
female workers capable of bearing children." On the other
hand, Velsicol states, "Protective clothing should be worn
by men as well as women. This is because any hazards
to applicators or field workers would be from acute exposure,
not from a teratogenic hazard" (Velsicol, p.91).
Since the risk criterion for acute dermal
toxicity for endrin had been rebutted by Velsicol, the
Agency determined to only impose additional protective
clothing requirements for female workers since they were
imposed on the basis of a teratogenic risk. On that point,
although it is true that only women who are capable of
bearing children are at risk, the Agency believes that it is
prudent to impose protective clothing requirements for all
2/ (Con't) significantly increased at the three highest
single doses, 5.0,7.5, and 10.0 mg/kg" (Velsicol, Exh.30
pp.5-6). It appeared, on further discussion at the SAP
meeting (SAP, pp.51-4), that the consultant's misperceptions
of the data should have been rectified. The record is clear
that frank terata were produced by single doses of endrin
that were substantially below that which caused observable
maternal toxicity to female hamsters. The argument that
near-toxic levels were required to produce terata from
single-dose exposure is not based on fact and any conclusion
regarding endrin's "low potential" for teratogenicity based
on that argument is spurious.
-56-
-------
women involved in application of endrin since the vast
majority of such female workers are likely to be of child-
bearing age. Second, as an independent matter (and before
the teratogenic risk came into focus), Velsicol proposed
general improvements in labeling to protect against acute
risks to all workers (Akesson, 1978). The Agency agrees
that all workers could be protected more effectively against
acute risks by improvements in protective clothing and
equipment and registrants are encouraged to voluntarily
apply such label improvements to men as well as women.
8- Warnings on Prophylactic Use
In its Notice of Determination, the Agency
proposed the following language, "Prophylactic Use.
Unnecessary use of this product can lead to resistance
in pest populations and subsequent lack of efficacy." The
Agency received several comments on this proposal.
The USDA commented, "We do not believe the
statement on prophylactic use adds anything to the usefulness
of the label information and should be deleted" (Bergland,
1978). The SAP report indicated that, "The Panel is concerned
with the risks inherent with the prophylactic use of endrin
and urges that the Agency reexamine the label statements
regarding such use" (Fowler, 1978). Velsicol's comment was,
"Velsicol proposed to prohibit prophylactic use of endrin.
-57-
-------
The Agency's proposed label language, however, merely
is in the form of a warning and is not emphatic enough
deterrent against prophylactic use" (Velsicol, p.90).
Finally, Dr. Ross Byers (Byers, 1978) wrote:
The statement, page 33 concerning
"Prophylactic Use" is not based on fact.
Resistance in vole populations is not the
result o£ using Endrin when not needed!
Where resistance develops is when partial
control is achieved through low dosage
applications and/or poor application
technique. Partial control allows suffi-
cient animals within the area to continue
the reproduction of survivors in the
presence of the toxicant. Pine vole
populations were first found resistant to
Endrin in the areas most seriously infested
and where growers were using reduced rates
per acre and/or using rather poor applica-
tion techniques.
The comments of USDA and the SAP on prophylactic
use are diametrically opposed and in neither case is the
basis for the position fully articulated. The Agency can
only respond by a fuller explanation of its position.
The Agency considered imposing a prohibition
against prophylactic use (that is, use when economic infesta-
tions are not present), such as that encouraged by Velsicol,
rather than a warning. The Agency decided against the
prohibition because it believed such a restriction would be
generally unenforceable. Unless substantial damage is
visible, it is usually not possible to determine, after
control measures have been applied, whether or not the
-58-
-------
controlled populations had been at economic levels. In any
event, the Agency believes that the educational aspect of
the proposed label language accomplishes the Agency's
primary objective in this respect, so that a prohibition
per se is not necessary.
Dr. Byers' account is not necessarily at
variance with the principles on which the Agency relies
in its concerns about resistance.. Dr. Byers indicates that
repeated usage was necessary because of poor control and
implicates the poor control as a critical factor in the
development of resistance. While, historically, this may
have been the case in Virginia, the reason for making
frequent applications is not relevant to the principles of
natural selection that lead to genetic resistance; selection
should be even more rapid if repetitive control is highly
effective. Dr. Byers' comment does highlight the impor-
tance of proper application methods and the proposed
label changes regarding rates and equipment should help to
prevent situations such as those described by Dr. Byers.
9. Enforcement
The EDF notes that pests other than those for
which the Agency proposes to maintain registration may occur
in small grains and orchards and asks, "How does the 1978
amendment (Section 2 (ee)), which allows use on a site
-5 9-
-------
against pests not named on the label, affect these 'cancelled'
uses?" (Hinkle, 1979). The Agency was cognizant of this
problem and addressed it in accordance with Section 2
(ee) of FIFRA by requiring that the labeling specifically
state that endrin may be used "only" for the pests specified
on the label, after it was determined that the use of
endrin against other pests would cause an unreasonable
adverse effect on the environment. The Agency is aware that
strict enforcement of label restrictions may be impossible
but believes that, where its regulatory actions have been
reasonable, an adequate level of compliance can be anticipated.
Any substantial evidence that misuse has become a common
practice would provide a basis for further regulatory
action.
10. Grasshopper Control
The EDF strenously opposes the use of endrin
to control grasshoppers, citing the existence of risks to
fish, wildlife and livestock from the use of endrin on wheat
and the availability of safer alternatives (Hinkle, 1979).
Since this use of" endrin was not fully analyzed in PD 2/3
because available information on both risks and benefits
appeared to be inadequate, the Agency will continue to
review any new evidence as it becomes available. If that
review indicates that additional regulatory action is
-60-
-------
desirable, the Agency will issue a supplement to PD 2/3 for
review by the SAP, the USDA, the FWS, and the public.
Concerned parties are requested to submit any additional
information regarding the risks and benefits from the use of
endrin to control grasshoppers as soon as possible.
When the Agency began its risk/benefit analysis,
the only registration for endrin to control grasshoppers was
for small grains in Montana. This old state registration is
now pending as an application for federal registration for
use in Montana. While PD 2/3 was in preparation, the Agency
also received endrin registrations for special local
needs in the states of Nebraska and Oklahoma, pursuant to
Section 24(c) of FIFRA, to control grasshoppers both in
winter wheat and as perimeter treatments in non-cropland
(but not on rangeland). This latter site for the use of
endrin to control grasshoppers will also be re-evaluated in
the event that new data on the risks and benefits of that
use become available. Currently endrin use for grasshopper
control is restricted to the above three states.
D. Comments Relating to Procedural Matters
The Agency has received several comments with
regard to the RPAR process and how, in the case of endrin,
the Agency has administered the process. Since some
of these comments reflect misunderstanding, misconstrue the
record, or otherwise influence the public perception of
Agency activities, the issues raised by these comments
require some discussion and clarification.
-61-
-------
1. Availability of the Agency's Rebuttal Analysis
Velsicol has stated that "the [RPAR] regulations
require the Administrator to issue prior to initiation of a
risk/ benefit analysis a notice of determination as to
whether the cited risk presumptions have been rebutted. See
40 CFR 162.11(a)(5). In the case of the endrin RPAR, however,
the Agency's rebuttal analysis was not made available to
Velsicol until after the Agency's risk/benefit analysis had
been completed." (Velsicol, pp.5-6, FN 1, emphasis in
original).
Velsicol has misinterpreted the relevant
provisions of the RPAR regulations. It is true that Section
162.11 (a)(5)(ii) states that "... if after review of the
evidence submitted in rebuttal the Administrator determines
%
that the applicant or registrant has not rebutted the
presumption..., then he shall issue a notice in accordance
with sections 3(c)(6), or 6(b)(1) of the Act..., as appro-
priate, for the use(s) of the pesticide subject to the
presumption and not rebutted." However, Section 162.11
(a)(5)(iii) specifically provides that "in determining
whether to issue a notice pursuant to section 3 (c)(6) or
section 6 (b)(1) ... in accordance with paragraph (a)(5)(ii)
of this section 162.11, the Administrator may, in his
discretion, take into account staff recommendations resulting
from preliminary analysis, if any, concerning the balancing
-62-
-------
of risks against benefits." In other words, the regulations
clearly contemplate that the Administrator may evaluate
benefits, and the balancing o£ those benefits against risks,
in determining whether or not to issue a notice of intent to
cancel or deny registration in cases where the risk presump-
tions have not been rebutted. Contrary to Velsicol's
assertions, nothing in those regulations or otherwise
requires the Administrator to issue a separate document as
to whether the risk presumptions have been rebutted, prior
to initiating the risk/benefit analysis^.
2. Initiation and Evaluation of New Studies
on Teratogenicity
Velsicol has portrayed a situation which does
not correctly represent either the Agency's timing of, or
motive for, initiation of additional teratology studies on
endrin by its own scientists. First, Velsicol claims that
because the Agency was "[alpparently dissatified with IRDC's
findings, in January, 1978, the Agency requested Dr. Chernoff
of its Health Effects Research Laboratory to confirm the
8/ Velsicol also argues that the Agency's alleged refusal to
disclose its rebuttal analysis prior to completion of its
risk/benefit analysis "unnecessarily delayed Velsicol from
developing further information on the Agency's remaining
risk concerns." (Ibid). Even assuming that the regulations
contemplate repeated opportunities for registrants to
rebut presumptions of risk (by "developing further information"
after it is determined that the presumption was not rebutted
by the initial submission), the Agency does not believe that
Velsicol was prejudiced in the circumstances of this case.
In any event, the Agency accepted Velsicol's comments on
PD 2/3 on January 5, 19 79 - over two months after Velsicol
received a copy of PD 2/3.
-63-
-------
results of the Ottolenghi and IRDC studies." (Velsicol,
p.35). As a matter of fact, however, HERL was requested to
do a comparative study of single vs. multiple dosing of
hamsters on August 15, 1977 — well before even the draft
results of the IRDC studies were forwarded to the Agency—
More significantly, it is surprising that
Velsicol would continue in its January 1979 submission
to make the unfounded allegations concerning the timing
of the internal request, particularly since in December 1978
it specifically acknowledged that the Agency's internal
request was made in August 1977—'The Agency does not
believe that any valid purpose is served by what is at best
careless presentation of the facts.
9/ However, Dr. Chernoff did not initiate the study itself
until January 1978 because the project had not been assigned
due to an internal misunderstanding at BERL.
10/ See Velsicol Exhibit 28, which is a letter from Louis
G. Nickell, Vice-President of Velsicol, to H. Wade Fowler,
Director of SAP, dated December 7, 1978. On page 1, Mr.
Nickell states that the "Chernoff study 5/ was requested by
Dr. H/le Barbehenn 6/...". Reference 6 to the letter reads:
"6. Barbehenn, K.R., SPRD, EPA. August 15, 1977. Request
for Assistance: Teratogenicity of Enarin. Memorandum to
John ftoelson, Environmental Research Center, Health Effects
Research Laboratory, EPA, Research Triangle Park, North
Carolina".
64-
-------
Second, the Agency's inquiry into the validity
of the IRDC studies did not occur until Dr. Chernoff attempted
to replicate the IRDC studies and was unsuccessful in
solubilizing endrin in methocel, the test vehicle used by
IRDC in alleged conformity with the protocols which had
previously been- submitted to the Agency. As Velsicol is
well aware, those protocols specifically stated (Velsicol,
Exh. 37): "Vehicle will be selected that will solubilize
the Endrin, and will not potentiate teratogenic or fetotoxic
effects" (Emphasis Added).
When Dr.. Chernoff could not solubilize endrin
in methocel, it was discovered that the final IRDC reports
(Velsicol, Exh. 38,39) indicated that "endrin was suspended
in a 0.5% aqueous Methocel solution at varying concentrations
... " (emphasis added). When the Agency demanded an explana-
tion of this apparent deviation from the protocols, IRDC
responded (Velsicol, Exh. 40): "With regard to the protocols,
we have interpreted the word 'solubilize' to mean to make
more soluble rather than to prepare a true chemical solution.
By the use of Methocel as a vehicle, some Endrin was solubi-
lized and the remainder suspended, resulting in fulfillment
of the protocol requirement of solubilizing Endrin."
Whether or not, as a scientific matter, this
unilateral interpretation in any way affected the validity
of the results of the IRDC studies, it seriously undermines
-65-
-------
Velsicol's implications that the Agency capriciously raised
after-the-fact questions about the IRDC protocols "despite
the efforts of Velsicol in advance of the IRDC studies
to ensure that the protocols were acceptable to the Agency
and despite Velsicol's belief that the Agency had approved
the protocols in every respect." (Velsicol, p.36).
Finally, Velsicol asserts that it did not
receive Chernoff's final report until September 6, 1978,
whereupon it arranged to have two outside teratologists
(Drs. Steffek and Wilson) review the studies and data and
prepare expert analyses (Velsicol, p.7, FU2). However,
Velsicol elsewhere acknowledges that it received a draft
copy of the Chernoff study in July, 1978 (Velsicol, p.36);
and, in fact, Dr. Steffek visited Dr. Chernoff shortly
thereafter (on Velsicol's behalf) to discuss his draft
paper.
3. Use of Relevant Information on Risk Assessment
Velsicol has alleged that the Agency "apparently
was unable to take into account ... in Position Document
2/3 significant risk information on teratogenicity and other
matters which had been developed at the Agency's request"
(Velsicol, pp.7 and 17). Although the Agency extended the
opportunity to Velsicol to comment upon the teratogenicity
issue prior to the issuance of PD 2/3, the matters referred
to were certainly not developed "at the Agency's request."
-66-
-------
Moreover, while the Agency indicated a willingness to
consider any new information for its potential impact on the
pending decision, preliminary reviews by the Agency indicated
that none of Velsicol's last minute submissions contained
any information that required any change in the Agency's
position. And, as indicated in this PD 4, the Agency has
reviewed and commented on all relevant information supplied
by Velsicol before making this final decision, so that the
Agency's review process has not resulted in any prejudice to
Velsicol.
4. Development of State Programs for Use on
Cotton
Representatives of the states of Alabama,
Arkansas and Mississippi requested the Agency to defer
the final decision on the use of endrin on cotton until the
States can develop programs that would substantially alter
the risk/benefit picture (Lane, 1979; Lincoln, 1979a; Coley,
1979). The Agency responded by indicating that it wishes to
encourage the development of such programs in general but,
in the absence of new information, the Agency had no basis
for deferring a decision already overdue (Johnson, 1979a).
It also indicated that should new information on risk/benefit
relationships be developed, including the institution of
state programs which would establish appropriate controls to
enhance the risk/ benefit ratio for the use of endrin on
-67-
-------
cotton, it would then be appropriate for the Agency to
reconsider the registration for the use of endrin on cotton
in areas east of 1-35.
The state of Arkansas then proposed that the
State would establish a new category for certain restricted
use pesticides such as endrin, in effect making them available
for use only under emergency conditions to be identified by
extension personnel (Martin, 1979). The Agency responded
that many specific details of such a program would have to
be developed for further consideration, that a revised
risk/benefit analysis would be necessary, and that any new
decisions proposed by the Agency would require reconsulta-
tion with the FWS and public review (Jbhnson, 1979b). Thus,
the Agency still has no basis for deferring its decision but
will reconsider it whenever it is justified by the availabi—
lity of new information.
III. Conclusions
After considering the comments received from the USDA,
the SAP, the USDI, Velsicol and other concerned parties, the
Agency has decided to make the following revisions to the
Notice of Determination:
A. Registration for Use on Cotton
1. Warning to Female Workers. "Excessive exposure
to endrin may cause birth defects"-will be amended to
read, "Excessive exposure to endrin during pregnancy
may cause birth defects."
-68-
-------
2.	Aerial Application. "Do not release this
material at greater than 19 feet height above the crop"
will be amended to read "...10 feet height above the crop."
3.	Application Restrictions. "Do not use this
product within 150 yards of human habitation "will be
amended to read "Do not apply this product within 1/8 mile
of human habitation."
"Do not use this product within 1/4 mile of
streams, lakes, or ponds. Application may be made within
1/4 mile of ponds owned by the user, but application within
200 yards of such ponds may result in fish kill" will
be amended to read, "Do not apply this product by air within
1/4 mile or by ground within 1/8 mile of lakes, ponds, or
streams. Application may be made at distances closer
to ponds owned by the user but such application may result
in excessive contamination and fish kills."
4.	"Procedures to be Followed if Fish Kills
Occur. In case of fish kills, fish must be collected
promptly and disposed of by burial. At ponds, post signs
stating: Contaminated: No Fishing. Signs must remain for
one year after fish kill has occurred" will be amended to
read, "Procedures to be Followed if Fish Kills Occur or if
Ponds are Contaminated. In case of fish kills, fish must be
-69-
-------
collected promptly and disposed of by burial. Ponds in
which fish kills have occurred, and user-owned ponds exposed
to endrin by application at distances closer than otherwise
prohibited, must be posted with signs stating: 'Contaminated:
No Fishing.' Signs must remain for one year after a fish
kill has occurred or for six months after lesser contamina-
tion unless laboratory analysis shows endrin residues in the
edible portion of fish to be less than 0.3 parts per million
(ppm)."
5. Add: "For use in areas west of Interstate
Highway #35 only".
B.	Registration for Use on Small Grains
Amendments 1, 2, 3, and 4 for cotton (A, above)
are applicable for small grains.
C.	Registration for Use in Apple Orchards
1.	Amend the "Warning to Female Workers" as
above.
2.	Pests for Which this Product May be Applied,
"This product may be applied to control the following pest
only: pine vole? western meadow vole" will be amended to
read,. "This product may be applied to control the following
pests only: Eastern United States-Pine Voles (Microtus
pinetorum); Western United States-Meadow Voles (Microtus
species)".
-70-
-------
3.	Equipment. Add, "Apply fay ground equipment
only."
4.	Procedures to be Followed If Fish Kills
Occur
"In case o£ fish kills, fish must be collected
promptly and disposed of by burial. At ponds, post signs
stating: 'Contaminated: No Fishing'. Signs must remain for
one year after fish kill has occurred." will be amended to
read "In case of fish kills, fish must be collected promptly
and disposed of by burial. Ponds in which fish kills have
occurred must be posted with signs stating: 'Contaminated:
No Fishing'. Signs must remain for one year after a fish
kill has occurred unless laboratory analysis shows endrin
residues in the edible portion of fish to be less than 0.3
parts per million (ppm)."
D.	Registrations for Use on Sugarcane
Amend the "Warning to Female Workers" as above.
E.	Registration for Treatment of Conifer Seed
Application Restrictions. "Do not apply when
large numbers of migratory birds are expected " will be
amended to read: "Do not sow treated seed when large numbers
of migratory birds are expected."
F.	Registrations for CJse as Tree Paint
Amend the "Warning to Female Workers" as above.
-71-
-------
G.	Registration of Use	on Alfalfa and Clover Seed
Crops
Amendments 1, 2, 3,	and 4 for cotton (A, above)
are applicable to alfalfa and	clover seed crops.
H.	Registration for Ose in Enclosed Bird Perch
Treatments
Amend the "Warning to Female Workers" as above.
Except for the above amendments, all provisions of
the Notice of Determination will be adopted as the final
decision on the registration and continued registration of
pesticide products containing endrin.
-72-
-------
BIBLIOGRAPHY
Akesson, Norman B. 1977. The Technical Advisory Committee
on Label Restrictions for Endrin-Methyl Parathion 1.6 EC and
Endrin 1.6 EC. Report from the committee's meeting in
Denver, Colorado, December 8-9, 1977.
Akesson, Norman B. 1978. Letter from Agricultural Engineer-
ing Department, University of California, Davis, California to
Lawrence Ebner, Sellers, Conner and Cuneo, Attorneys and
Counselors, Washington, D.C., defining water bodies and
streams for endrin use and limitations. August 25, 1978.
Alley, E. G., B. R. Layton, and J. P. Minyard, Jr. 1974.
Photoreduction of Mirex in aliphatic amines. J. Ag. Food
Chem. 22 (4), 727-9.
Anderson, Elizabeth L. 1978a. Memo from Executive Director,
CAG, EPA, to H. Wade Fowler, Executive Secretary, FIFRA
Science Advisory Panel (SAP), transmitting a response to
Melvin Reuber's critique of the endrin carcinogenicity
review. December 7, 1978.
Anderson, Elizabeth L. 1978b. Memo from Executive Director,
CAG, EPA, to H. Wade Fowler, Executive Secretary, FIFRA
SAP, transmitting additional comments to Melvin Reuber's
critique of CAG's endrin review. December 13, 1978.
Baker, R. D., and ff. G. Applegate. 1974. Effect of ultra-
violet radiation on the persistence of pesticides. Texas J»
Sci. 25 (1-4), 53-9.
Bergland, Bob. 1978. Letter from Secretary, U.S. Department
of Agriculture, to Douglas Costle, Administrator, U.S.
Environmental Protection Agency, responding to EPA's notice
of determination concluding the endrin RPAR. November 22,
1978.
Blalock, T. C. 1978. Letter from Associate Dean, North
Carolina State University School of Agriculture and Life
Sciences, to Technical Services Division, OPP, EPA, commenting
on proposed actions on endrin. November 14, 1978.
Brooks, Leroy. 1978. Letter from Extension Specialist,
Kansas Cooperative Extension Service, to Dr. Paul W. Bergman,
USDA, concerning endrin use on small grains and endrin
use restrictions. November 20, 1978.
-73-
-------
Bushong, Clayton. 1978. Memo to I^le Barbehenn, Project
Manager, SPRD, OPP, EPA, concerning "Opinions on Endrin"
Letter with Four Separate Risk Analyses Attached. June 28,
1978.
Byers, Ross E. 1978. Letter from Assoicate Professor
of Horticulture, Virginia Polytechnic Institute and State
University, to Dr. James P. Barnett, Endrin Impact Assessment
Team Leader, USDA, concerning endrin for vole control.
November 14, 1978.
Castro, T. F., and T. Yoshida. 1971. Degradation of organo-
chlorine insecticides in flooded soils in the Philippines.
J. Ag. Food. Chem. 19 (6), 1168-70.
Chernoff, Neil, Robert J. fevlock, and L. Earl Gray, Jr.
1978a. Endrin-induced perinatal toxicity studies in the
hamster with a review of the pertinent literature. Health
Effects Research Laboratory, Experimental Biology Division,
EPA. Research Triangle Park, North Carolina (Unpublished)
August 25, 1978.
Chernoff, Neil, Robert J. tevlock and L. Earl Gray, Jr.
1978b. Memo to tyle Barbehenn, SPRD, OPP, EPA, concerning
review of statements by Drs. A. J. Steffek and J. G. Wilson
about endrin teratology studies. November 13, 1978.
Coley, Jack. 1979. Letter from Mississippi State Entomologist,
to Douglas Costle, Administrator, EPA, concerning possible
state restrictions to allow some uses of endrin. January
24, 1979.
Coppage, David. 1978. Memo from Aquatic Biologist, EEB,
HED, OPP, EPA, to Kyle Barbehenn, Project Manager, SPRD,
OPP, EPA, concerning theoretical movement of endrin from
soil into water and the effect of this movement on non-target
organisms. December 13, 1978.
Coppock, Stanley. 1976. Letter from Extension Entomologist,
Cooperative Extension Service, Oklahoma State University,
Stillwater, Oklahoma, to Edwin L. Johnson, OPP, EPA.
December 29, 1976.
Coppock, Stanley, and Terry Pitts. Undated. Grasshoppers
and insecticides for their control. Oklahoma State University
Extension Service. (Unpublished).
-74-
-------
Crockett, A. B., 6. B. Wiersma, H. Tai, and W. Mitchell.
1975. Pesticide and mercury residues in commercially grown
catfish. Pest. Monit. J. J3 (4), 235-40.
EPA. 1977. Preliminary Benefit Analysis of Endrin.
Economic Analysis Branch, (EAB), CED, OPP,
Washington, O.C.
EPA. 1978. Endrin Position Document 2/3. SPRD, OPP,
OSEPA, October 20, 1978.
Fowler, H. Wade, Jr. 1978. Memorandum from Executive
Secretary, FIFRA SAP, to Edwin Johnson, Deputy Assistant
Administrator, OPP, EPA, transmitting the SAP's findings on
endrin. December 21, 1978.
Greenwalt, Lynn A. 1978a. Director, Fish and Wildlife
Service, O.S. Department of the Interior, to Douglas
Costle, Administrator, EPA, concerning effects of endrin on
populations of endangered species. JUne 8, 1978.
Greenwalt, Lynn A. 1978b. Letter from Director, Fish
and Wildlife Service, U.S. Department of the Interior,
to Douglas H. Costle, Administrator, EPA, amending biological
opinion of JUne 8, 1978. December 14, 1978.
Greenwalt, Lynn A. 1979. Letter from Director, Fish
and Wildlife Service, O.S. Department of the Interior,
to Douglas Costle, Administrator, EPA, revising biological
opinion of December 14, 1978. March 1, 1979.
Hinkle, Maureen. 1979. Letter from Pesticides Monitor,
Environmental Defense Fund, to U.S. Environmental Protection
Agency, concerning the preliminary determination of the
endrin RPAR. February 15, 1979.
Johnson, Edwin L. 1979a. Letter from Deputy Assistant
Administrator, OPP, EPA, to McMillan Lane, Commissioner,
Alabama State Department of Agriculture and Industries,
concerning state programs to affect endrin risk/benefit
balancing. March 6, 1979.
Johnson, Edwin L. 1979b. Letter from Deputy Assistant
Administrator, OPP, EPA, to Mr. Wallace Martin, Arkansas
State Plant Board, concerning a proposed state program
for special restricted-use pesticides. April 16, 1979.
Lane, McMillan. 1979. Letter from Commissioner, Alabama
State Department of Agriculture and Industries, to Douglas
Costle, Administrator, EPA, concerning the need for endrin
for use on Alabama cotton. January 23, 1979.
-75-
-------
Lincoln, Charles. 1978. Letter from member, Arkansas
State Plant Board, to Edwin L. Johnson, Deputy Assistant
Administrator, OPP, EPA, concerning presumptions supposedly
rebutted at the SAP meeting. March 15, 1979.
Lincoln, Charles. 1979. Letter from member, Arkansas
State Plant Board, to Douglas Costie, Administrator, EPA,
concerning possible state restrictions to preserve some
endrin uses. January 19, 1979.
Leonard, R. A., G. W. Bailey, and R.R. Swank, Jr. 1976.
Transport, detoxification, fate, and effects of pesticides
in soil and water environments, in Land Applications of
Waste Materials, published by Soil Conservation Society
of America, Ankeny, Iowa.
Martin, Wallace. 1979. Memo from Chairman, Arkansas
State Plant Board, to EPA, concerning state restriction of
pesticide uses. March 14, 1979.
Matsumura, P., V. G. fbanzilkar, K. C. Patil, and G. M.
Boush. 1971. Metabolism of endrin by certain soil micro-
organisms. J. Agric. Pood Chem. 19:27-31.
Moubry, R. J., J» M. Helm, and G. R. Myrdal. 1968. Chlori-
nated pesticide residues in an aquatic environment located
adjacent to a commercial orchard. Pest. Monit. J. 1_ (4) ,
27-9.
Mount, D.I. and G.J. Putnicki. 1966. Summary report of
1963 Mississippi fish kill. In: Trans. 31st N. Amer.
Wildlife Conf. 177-184. Wildl. Manag. Inst., Washington,
D.C.
Pionke, H. B., and G. Chesters. 1973. Pesticide-sediment-
water interactions. J. Environ. CCual. 2 (1) , 29-45.
Reuber, Melvin D. 1978a. Critique of the carcinogenicity
studies on endrin in animals. November 2, 19 78.
Reuber, Melvin D. 1978b. Carcinogenicity of endrin.
Frederick Cancer Research Center, Frederick, M.D.
(Unpublished) November 28, 1978.
Science Advisory Panel. 1978a. Transcript of Proceedings
of FIFRA SAP Meeting, October 26, 1978, in Arlington,
Virginia. Reported by Acme Reporting Company, Washington,
D.C.
-76-
-------
Science Advisory Panel. 1978b. Transcript of Proceedings
of PI ERA SAP Meeting, December 14-15, 1978, in Arlington,
Virginia. Reported by Acme Reporting Company, Washington,
D.C.
Seabolt, Buford. 1978. Memo to Dr. Ren Macek concerning
endrin residues from NPMP (fish). September 18, 1978.
Severn, David J. 1978. Memo from Chemist, EFB, HED,
OPP, EPA, to lyle Barbehenn, Project Manager, SPRD, OPP,
EPA, concerning Dr. V. H. Freed*s statement to the SAP on
endrin runoff. November 9, 1978.
Smith, C. N., R. A. Leonard, G. W. Langdale, and G. W.
Bailey. 1978. Transport of agricultural chemicals from
small upland piedmont watersheds.
Sparr, B. I., W. G. Appleby, D. M. DeVries, J. V. Osmun,
J. M. McBride, and G. L. Foster. 1966. Insecticide residue
in waterways from agricultural use, in Organic Pesticides in
the Environment, American Chemical Society Washington,
D.C.
Stickel, William H., William L. Reichel; Donald L. Hughes.
Undated. Endrin in birds: lethal residues and secondary
poisoning. Patuxent Wildlife Research Center, U.S. Fish
and Wildlife Service, Laurel, M.D. (Unpublished).
USDA. 1968. Monitoring for chlorinated hydrocarbon insecti-
cide residues in soybeans	1966. Pesticide Monit. J. 2
(1), 58-67.
USDA/EPA. 1975. Control of Water Pollution from Cropland.
Vol. I. Report No. ARS-H-5-1.
USDA/EPA. 1976. Control of Water Pollution from Cropland,
Vol II. Report No. ARS-H-5—2.
USDA/State Assessment Team on Endrin. Pesticide Impact
Assessment: Endrin. November 4, 1976.
Velsicol Chemical Corporation. 1979. Response of registrant
Velsicol Chemical Corporation to the preliminary determination
concerning the rebuttable presumption against registration
and continued registration of pesticide products containing
endrin. 91 pp + 61 Exhibits. January 5, 1979.
Willis, G. H., L. L. McDowell, J. F. Parr, and C. E. Murphree.
1976. Pesticide concentrations and yields in runoff and
sediment from a Mississippi Delta watershed. Proc. Fed.
Inter-Agency Sediment. Conf., 3rd.
-77-
-------