	

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

WASHINGTON, D.C.  20460

OFFICE OF PREVENTION, PESTICIDES AND TOXIC SUBSTANCES

December 1, 2006

Via Fax and Email

Mr. Ian S. Chart

Vice President, Director of Regulatory Affairs

4695 MacArthur Court, Suite 1250

Newport Beach, CA 92660

Dear Mr. Chart:

	In your letter of November 6, 2006, you noted that the Agency had not
made available the Agency’s responses to the Phase 5 comments for
PCNB.  In scrutinizing the public docket for the PCNB reregistration
decision, we found that the two documents responding to the Phase 5
comments had been omitted inadvertently from the public docket.  We have
since emailed Amvac copies of the two documents (one developed by our
Environmental Fate and Effects Division and one by our Health Effects
Division) and submitted both documents to the docket for posting.  We
apologize for any inconvenience you may have experienced as a result of
our oversight, and have extended the public comment period on the PCNB
RED for approximately 30 days to allow review of these documents by your
company and other members of the interested public.  The notice
announcing the extension, this letter, and your letter of November 6
have been or will be added to the docket.

	

	We will take this opportunity to reiterate our views with regard to
other information you assert is necessary for formulating your comments
on the PCNB RED.  You have requested additional guidance and information
on the food web modeling and the use of an ecological risk endpoint from
the Chemtura two-generation reproduction study in the rat.  Reference
was made in the PCNB RED and/or documentation supporting the RED to both
the model and this reproduction study.

 

Role of food web modeling and an alternate endpoint in ecological risk
conclusions

	As we noted in our letter to you dated October 20, 2006, the food web
model and the endpoint from a different reproduction study play a
limited role in the Agency’s characterization of the ecological risks
associated with the use of PCNB. 

	To emphasize our earlier statements, the food web modeling is just one
of many considerations that support the Agency’s conclusion that the
environmental fate characteristics of PCNB add qualitatively to the
risks estimated via risk quotients.  The Agency has not used the results
of the modeling to quantify risks.

	The Chemtura reproduction study (hereafter referred to as the alternate
study or the alternate reproduction study) was used to illustrate other
possible outcomes for calculating risk quotients based on an endpoint
more applicable to wildlife effects of interest.  The Agency has not
used calculations based on the alternate study endpoint to quantify
risks, but qualitatively, to suggest how the risk quotients would differ
if based on other adverse effects.

	Even when used in a qualitative sense, the food web model and alternate
endpoint are just a subset of information the Agency reviewed when
formulating its conclusion that the risk quotients reported in the PCNB
RED underestimate actual risks to wildlife.  Most importantly, as we
have written in past correspondence, our conclusions about the
ecological risks associated with PCNB and its metabolites would not
differ had the model and the alternate endpoint not been considered.

References to food web modeling and the rodent reproduction studies in
the RED and supporting documentation

	A reference to the food web model and the outputs of this modeling for
PCNB is made in the Environmental Risk section of the RED:  “…
application of the Gobas food web bioaccumulation model developed by
EPA’s Office of Water … suggests that the aquatic food chain can
serve as a mechanism for significant bioaccumulation …”

	Additional references to the model appear in the April 26, 2006 memo
entitled “Synopsis of Pentachloronitrobenzene Environmental Loading
and Ecological Risk” developed by our Environmental Fate and Effects
Division (EFED) and posted to the PCNB docket in support of the RED. 
The memo refers to the model and concludes that examination of available
information (of which the outputs for the model are but one piece) has
led the Agency to the conclusion that “PCNB is…expected to
bioaccumulate in aquatic food chains,” and cites potential
bioaccumulation factors that have been derived from the model for
species at different levels in the food web. 

An alternate study endpoint that is more sensitive than the endpoint
used to calculate chronic mammalian risk quotients for PCNB is also
discussed in the RED.  The RED notes that the endpoint used to calculate
chronic mammalian risk quotients is the same endpoint used in the
estimation of chronic human health risks.  The adverse effects
associated with that endpoint are cellular level effects to the thyroid
and liver.  The RED states that the relevance of these effects and their
relationship to more typical ecological assessment endpoints (e.g.,
impaired reproduction, growth, and survival) are not clear.  Other
effects, such as the reduced pup weights observed in the alternate
study, may be more relevant.  The RED goes on to say that the alternate
endpoint is not part of the risk assessments, but that risk quotients
based on the alternate endpoint would be much greater than the risk
quotients cited in the RED.  The RED does not provide risk quotients
based on the alternate endpoint, and reference is not made in any of the
supporting documents for the RED to risk quotients derived from an
alternate chronic endpoint for mammalian wildlife.

Amvac’s request for additional information on the food web model 

	In the November 6 letter, Amvac has asked for a description of the
modifications made to the Arnot and Gobas food web model.  In previous
letters requesting extensions to the comment period for the PCNB RED,
Amvac has requested additional information on the food web model, and
the Agency has in each reply attempted to provide relevant information
Amvac can use to advance its understanding of the role the model plays
in the reregistration decision.

	Throughout our post-RED correspondence with Amvac, the Agency has
repeatedly written that the food web model plays only a minor role in
conclusions about the ecological risks associated with the use of PCNB. 
We now reiterate that position.

Amvac’s request for additional information on the alternate
reproduction study

	Also in the November 6 letter, Amvac has requested documents from the
Health Effects Division (HED) or Environmental Fate and Effects Division
concerning the endpoint from the alternate study.  Amvac had previously
asked for the data evaluation report (DER) for this study, and all
relevant HED reviews (e.g., a HIARC review).  These same requests were
made in previous correspondence.

	The Agency reiterates that the HIARC review and the HED toxicology
chapter, both of which discuss the reproduction studies, have been
available on the PCNB docket for some time.  As we have advised you
before, these documents do not discuss the relevance of study endpoints
to effects in wildlife, but rather to the assessment of human risks.

	As noted several times previously, the Agency did not in fact rely on
the alternate study for quantifying risks to wildlife.  The RED mentions
it incidentally as a way to show that selection of an endpoint that
might be considered as more relevant would result in higher risk
quotients.  

Potential impact of these two elements on the Agency’s conclusions
about risk

	The Agency has stated in responses to earlier Amvac correspondence that
its conclusions about the ecological risks associated with PCNB and its
metabolites would not differ had the food web model and the alternate
mammalian endpoint not been considered.  Instead, the Agency concluded
that a combination of factors, including most particularly the
environmental persistence of PCNB and its metabolites, serve to indicate
that the calculated risk quotients for PCNB tend to understate the risks
to wildlife.

	Because the limited discussion of the alternate mammalian endpoint in
the RED adds so little to the characterization of wildlife risks
associated with the use of PCNB, the Agency will amend the RED to note
that the Agency has removed the alternate endpoint from consideration in
its risk assessment.

	The Agency’s conclusions about the risks associated with PCNB use
would be the same without the results of the food web model discussed in
the PCNB RED and related documentation.  The model is just one of many
kinds of information that support our conclusion that the environmental
fate characteristics of PCNB add to the risks estimated via the Risk
Quotient approach to ecological risk assessment.  Discussion of the
model in the RED was intended to provide, qualitatively, additional
support for one of those characteristics, i.e., PCNB’s potential to
bioaccumulate.  

	The food web model has not been validated for use on compounds such as
PCNB, nor are empirical data available to corroborate the model outputs
cited in the RED documents.  For this reason, the Agency has determined
that we cannot at this time rely on the model, either qualitatively or
quantitatively, to provide confirmatory data for the characterization of
the bioaccumulative potential of PCNB. The Agency’s overall
conclusions about bioaccumulation and its impact on the ecological risks
posed by PCNB use remain unchanged.  If Amvac believes that the Agency
has not drawn appropriate conclusions about how this bioaccumulation
contributes to ecological risk, the company is free to develop empirical
data to address its concerns.  The Agency has developed guidelines for
bioaccumulation studies which can serve as the basis for such testing.  

	Based on our removal of the alternate mammalian endpoint and the food
web model from consideration in the risk assessment, the Agency believes
that detailed comments from Amvac on either element will have no impact
on the Agency’s conclusions about the risks associated with PCNB use. 
Comments submitted by Amvac on these two subjects will be addressed in
the same way they have been addressed since the release of the RED, and
reiterated in this correspondence—by noting that the conclusions of
the RED would be the same even if we had not considered them.

Depuration and the presence of PCNB and metabolites in the aquatic
environment

	Your letter of November 6 requests that the Agency provide support for
the statement in our October 20 letter that “depuration plays little
or no role in bioaccumulation of PCNB and its metabolites at all trophic
levels in the aquatic environment.”  You further ask that we provide
support for the assumption of continuous steady-state concentrations of
PCNB in the water column and sediment.

	The Agency and Amvac have discussed the potential for depuration of
PCNB residues by aquatic animals before.  The Agency’s notes from a
January 26, 2006 meeting with representatives of Chemtura and Amvac
(including yourself) attribute to Agency personnel the statement that
“[t]he opportunity for depuration is limited because exposures to PCNB
and its PCA degradate would likely be continuous.”  The RED states,
“…[T]he impact of depuration is limited by the persistence of PCNB
and its metabolites in the aquatic environment, which create a regular
source of exposure.”  Since PCNB and its metabolites are very
persistent in the aquatic environment, residues will remain in the water
column, and potentially in food sources for aquatic animals, long after
fresh inputs are made to the environment.  It should be expected that
persistent environmental residues of PCNB will continue to be taken up
by aquatic animals, even as they are eliminating PCNB residues from
their bodies.

	The Agency did not claim that PCNB residues maintain a steady state in
the environment, but that their persistence provides a regular source of
exposure.  Repeated applications during the growing season and
subsequent seasons are likely to result in the continued presence of
residues in the aquatic environment.

Third extension of the comment period

	Although the Agency will remove both the food web model and the
alternate reproduction study from consideration at this time, the public
did not have access to the Phase 5 comments developed by the Agency and
inadvertently omitted from the docket until recently.  To allow
interested parties to review those documents, the Agency has extended
the comment period on the PCNB for a third time, to January 8, 2007.  A
notice was published in the Federal Register on November 29 announcing
the extension.

					Sincerely,

/s/

					Debra Edwards, Ph.D.,

					Director, Special Review and 

					  Reregistration Division (7508P)

 The notes Amvac provided for the same meeting differ somewhat, amending
the language here to add:  “EPA provided no data to support this
hypothesis…”

 PAGE   

 PAGE   5 

