Page
1
of
27
WASHINGTON,
D.
C.
20460
OFFICE
OF
PREVENTION,
PESTICIDES
AND
TOXIC
SUBSTANCES
UNITED
STATES
ENVIRONMENTAL
PROTECTION
AGENCY
PC
Code:
080803
DATE:
March
26,
2003
SUBJECT:
EPA
Response
to
Comments
from
Syngenta
and
its
Contractors,
the
Triazine
Network,
the
Center
for
Regulatory
Effectiveness,
the
American
Water
Works
Association,
the
State
of
New
York
Office
of
Attorney
General,
the
U.
S.
Department
of
Interior
Fish
and
Wildlife
Service,
the
California
Regional
Water
Control
Board,
the
American
Farm
Bureau
Federation,
and
the
Louisiana
Farm
Bureau
Federation,
about
the
EPA
Reregistration
Eligibility
Science
Chapter
for
Atrazine,
Environmental
Fate
and
Effects
Chapter,
dated
April
22,
2002.

TO:
Kimberly
Lowe,
Chemical
Review
Manager
Special
Review
and
Reregistration
Division
7508C
FROM:
Mary
Frankenberry,
Statistician
James
Lin,
Environmental
Engineer
Timothy
Bargar,
Aquatic
Biologist
Douglas
Urban,
Senior
Scientist
Environmental
Risk
Branch
3
Thomas
Steeger,
Aquatic
Biologist
Environmental
Risk
Branch
4
Environmental
Fate
and
Effects
Division
7507C
THRU:
Stephanie
Irene,
Acting
Chief
Environmental
Risk
Branch
3
Environmental
Fate
and
Effects
Division
7507C
The
U.
S.
Environmental
Protection
Agency
(
EPA)
has
reviewed
the
comment
documents
from
Syngenta
Crop
Protection,
Inc.
and
its
contractors,
the
Triazine
Network,
the
Center
for
Regulatory
Effectiveness,
the
American
Water
Works
Association,
the
State
of
New
York
Office
of
Attorney
General,
the
U.
S.
Department
of
Interior
Fish
and
Wildlife
Service,
the
California
Regional
Water
Control
Board,
the
American
Farm
Bureau
Federation,
and
the
Louisiana
Farm
Bureau
Federation.
Many
of
Syngenta's
comments
referred
to
the
assessment
performed
by
the
2
Atrazine
Ecological
Risk
Assessment
Panel,
"
Aquatic
Ecological
Risk
Assessment
of
Atrazine­­
A
Tiered
Probabilistic
Approach"
(
Giddings
et
al.
2000),
as
well
as
Syngenta's
"
Supplement
to
`
Aquatic
Ecological
Risk
Assessment
of
Atrazine
­
A
Tiered
Probabilistic
Approach'
Including
Responses
to
EPA's
Comments,"
dated
February
26,
2002
(
Giddings
et
al.
2002).
EPA
has
previously
responded
to
the
former
in
Urban
et
al.
2002,
and
is
responding
to
the
latter
in
Frankenberry
et
al.
2002.
EPA
has
thoroughly
reviewed
the
comments
and
made
extensive
responses
on
specific
areas,
such
as
the
choice
of
technical
studies
included
for
the
assessment,
the
validation/
status
of
the
exposure
models
employed,
problem
formulation
and
scope
of
the
study,
and
the
level
or
"
tier"
of
assessment
actually
performed,
especially
in
light
of
refined
risk
assessment
guidelines
under
development
by
the
Environmental
Fate
and
Effects
Division
(
EFED)
in
EPA's
Office
of
Pesticide
Programs
(
OPP).
EPA
has
met
with
Syngenta
and
its
panel
of
scientists
to
discuss
these
comments.
Therefore,
this
response
to
comments
will
address
both
general
and
specific
statements
from
all
the
commenters,
and
will
also
refer
to
EPA's
review
of
the
Syngenta
risk
assessment
and
supplement
where
appropriate.

General
Response
to
Comments
Concerning
Potential
Endocrine
Effects
EPA
is
in
the
process
of
evaluating
data
on
the
potential
endocrine
disrupting
effects
of
atrazine
on
a
range
of
non­
target
animals.
These
data
focus
primarily
on
the
effects
of
atrazine
on
endocrine­
mediated
gonadal
development
in
amphibians,
fish
and
aquatic
organisms.
Some
of
the
studies
under
evaluation
have
been
submitted
by
pesticide
registrants
while
the
remainder
have
been
taken
from
open
literature;
thus,
the
studies
include
both
published
and
unpublished
literature.
Effect
concentrations
vary
widely,
with
effects
on
gonadal
development
reported
at
exposures
from
0.1
part
per
billion
(
ppb)
to
25
ppb.
EFED
is
in
the
process
of
reviewing
the
studies
and
writing
data
evaluation
records
(
DERs).
Where
possible,
raw
data
from
these
studies
are
being
analyzed
and
study
methodologies
are
being
documented
to
determine
whether
procedural
errors
may
have
confounded
the
study's
ability
to
detect
treatment
effects.
Based
on
this
review,
EFED
will
evaluate
whether
the
data
support
the
conclusions
reached
by
the
study's
author.

Pertinent
studies
are
being
performed
both
by
the
pesticide
registrants
and
by
independent
researchers,
which
could
shed
additional
light
on
this
sub­
lethal
endpoint
for
atrazine.
In
accord
with
the
agreement
reached
with
the
Natural
Resource
Defense
Counsel,
OPP
is
planning
to
summarize
all
these
studies
in
preparation
for
a
Federal
Insecticide
Fungicide
and
Rodenticide
Act
(
FIFRA)
Scientific
Advisory
Panel
(
SAP)
meeting
scheduled
for
June
2003.
To
facilitate
the
preparation
of
this
summary
document,
OPP
will
continue
to
accept
studies
on
this
endpoint
up
to
February
28,
2003.
These
studies
can
only
be
accepted
in
final
form
and
must
represent
"
publication
quality";
no
preliminary
data
can
be
accepted
for
this
summary
document.
Data
submitted
prior
to
the
February
cutoff
date
will
be
included
in
the
summary
document
and
discussed
in
the
June
2003
SAP
meeting.
While
EPA
does
not
intend
to
exclude
any
pertinent
studies
from
consideration,
the
February
28th
deadline
is
necessary
for
the
Agency
to
have
sufficient
time
to
review,
evaluate
and
summarize
the
available
studies
and
present
them
for
timely
science
review
by
the
FIFRA
SAP.
In
addition,
the
results
from
this
SAP
meeting
are
3
considered
crucial
input
for
the
amended
interim
reregistration
eligibility
decision
on
atrazine
scheduled
for
October
2003.

Syngenta
Comments
in
Response
to
Reregistration
Eligibility
Science
Chapter
for
Atrazine
Environmental
Fate
and
Effects
Chapter,
April
22,
2002
[
OPP
02­
0026­____]

P.
11
Terrestrial
Birds
and
Mammals
­
Syngenta
states
that
the
current
risk
assessment
inaccurately
depicts
the
chronic
risk
of
atrazine
to
birds
and
mammals.
While
Syngenta
agrees
with
the
EPA
that
the
risk
quotients
are
overestimates
due
to
the
preliminary
screening
methods
used,
they
disagree
with
the
use
of
two
specific
inputs.
First,
they
disagree
with
the
use
of
a
No
Observed
Adverse
Effect
Level
(
NOAEL)
for
small
mammals
of
10
parts
per
million
(
ppm)
and
points
to
the
Atrazine:
Toxicology
Chapter
for
the
Reregistration
Eligibility
Decision,
SECOND
REVISION,
dated
April
11.
2002
which
reports
on
page
67
that
the
offspring
NOAEL
is
3.78
mg/
kg/
day,
which
is
equivalent
to
50
ppm.
Second,
they
disagree
with
the
17­
day
short­
grass
foliar
dissipation
half­
life
value
used
by
EFED
in
the
FATE
model,
and
maintains
that
the
value
should
be
4­
days.
Finally,
Syngenta
maintains
that
when
realistic
exposure
information
is
used,
the
data
show
that
there
are
no
unacceptable
risks
to
terrestrial
vertebrates.

While
there
are
data
in
the
Health
Effects
Division
(
HED)
TOX
ONELINER
database
(
dated
11,
August
1999)
that
reports
a
NOAEL
of
10
ppm
based
on
body
weight
reduction,
EFED
agrees
to
use
the
50
ppm
value
as
reported
by
the
HED
in
their
second
revision
of
the
HED
toxicology
chapter
.
The
change
will
result
in
a
reduction
in
the
LOC
exceedences
for
small
mammals,
but
not
in
the
elimination
of
the
potential
risk.
The
available
foliar
residue
data
show
that
estimates
of
atrazine
half­
lives
vary
considerably
(
from
3.2
to
17­
days).
Since
EFED
was
limited
by
the
available
data
for
performing
a
screening
level
assessment
for
risks
to
birds
and
mammals,
the
choice
of
a
conservative
estimate
of
17­
days
for
the
foliar
half­
life
value
is
appropriate.
EFED
will
verify
the
FATE
model
calculations
including
the
number
of
days
exceeding
the
LOCs.

P.
12
Terrestrial
Plants
­
Syngenta
questions
EPA's
statements
that
widespread
atrazine
exposure
results
in
direct
acute
effects
in
many
terrestrial
plant
species
causing
indirect
effects
on
wildlife
claiming
the
conclusion
is
not
based
on
field
investigations
and/
or
refined
risk
assessments.

EFED
agrees
that
there
are
no
field
studies
showing
the
linkage
between
direct
effects
of
atrazine
and
indirect
effects
on
wildlife.
Also,
EFED
recognizes
that
the
incident
reports
of
direct
effects
of
atrazine
on
terrestrial
plants
is
limited.
However,
the
screening
level
assessments
shows
a
great
potential
for
direct
effects
on
terrestrial
plants.
Therefore,
EFED
will
change
the
statements
to
clearly
indicate
that
there
is
a
potential
risk
for
direct
effects
on
terrestrial
plants,
and
that
potential
indirect
effects
on
wildlife
exist
for
atrazine
as
well
as
other
herbicides
via
habitat
degradation
and
possible
disruption
of
the
food
chain.

P.
12
Toxicity
of
Degradates
Compared
to
Parent
Atrazine
­
Syngenta
disagrees
with
EPA's
description
of
the
atrazine
metabolites
as
"
long
lived".
They
point
to
previously
submitted
metabolite
residue
information
from
HED's
Product
and
Residue
Chemistry
Chapters
(
DP
Barcode
D272006).
In
summary,
they
point
to
the
maximum
reported
residue
after
30­
days
of
4
3.21
ppm
for
the
chloro­
metabolites.
They
state
that
this
residue
level
is
far
below
any
effect
level
for
terrestrial
vertebrates
indicating
that
metabolites
are
not
"
long­
lived."

EFED
believes
the
environmental
fate
profile
of
atrazine
and
monitoring
data
support
the
existence
of
"
long­
lived"
atrazine
metabolites.
The
3.21
ppm
level
for
the
chloro­
metabolites
in
a
residue
monitoring
study
merely
shows
that
these
metabolites
were
transformed
from
the
parent
atrazine.
The
study
does
not
provide
sufficient
data
to
establish
the
kinetics
of
the
formation
and
decline
of
the
chloro­
metabolites.
Further,
the
3.2
ppm
residue
level
does
not
refute
EFED's
conclusion
that
the
atrazine
metabolites
are
"
long­
lived."

P.
13
Incident
Reports
­
Syngenta
continues
to
disagree
with
EPA's
summary
of
incident
data
reported
in
the
reregistration
eligibility
decision
chapter
(
RED).

As
stated
in
the
previous
response
to
comments
(
April
9,
2002,
EFED
Review
of
Comments
from
Syngenta
and
its
Contractors
about
the
EPA
Revised
Environmental
Risk
Assessment
for
Atrazine,
page
15),
EFED
minimally
agrees
with
the
14
incidents
that
Syngenta
attributed
to
atrazine.
These
incidents
in
addition
to
the
screening
level
risk
assessment
for
terrestrial
plants
results
in
a
revised
EPA
conclusion
that
there
is
a
potential
for
direct
effects
on
terrestrial
plants,
and
that
the
potential
for
indirect
effects
(
habitat
degradation
and
disruption
of
food
chain)
on
wildlife
exists
for
atrazine
as
well
as
for
other
herbicides.

P.
14
Uncertainty
of
Atrazine
Co­
Occurrence
With
Other
Chemicals
­
Syngenta
provides
arguments
questioning
EPA's
suggestion
that
synergistic
effects
could
have
caused
some
of
the
non­
plant
related
incidents.

EFED
agrees
that
references
to
synergistic
effects
as
causing
the
non­
plant
related
incidents
are
speculative
and
therefore
will
remove
them
from
the
document.

P.
15
Indirect
Effects
­
Syngenta
states
that
there
no
data
to
support
aquatic
and
terrestrial
indirect
effects
caused
by
atrazine
given
the
reversible
nature
of
its
phytotoxic
effect
on
plants.

EFED's
response
concerning
indirect
effects
has
been
presented
in
the
preceding
responses
on
terrestrial
effects.
However,
concerning
indirect
effects
on
aquatic
systems,
EFED
disagrees
with
Syngenta.
As
stated
in
a
previous
response
(
April
9,
2002,
EFED
Review
of
Comments
from
Syngenta
and
its
Contractors
about
the
EPA
Revised
Environmental
Risk
Assessment
for
Atrazine,
page
20),
although
aquatic
plants
initially
showed
rapid
recovery
of
photosynthetic
capacity,
mean
photosynthesis
rates
for
atrazine­
treated
plants
were
less
than
controls
77
hours
following
exposure.
Additionally,
Jones
et
al.
did
not
report
the
results
for
biomass
for
controls
and
treatments
to
confirm
full
recovery
compared
to
controls.

Syngenta's
reference
to
recovery
may
only
hold
under
short­
term
exposures
to
atrazine
in
lotic
(
flowing)
aquatic
environmental.
Following
a
2­
day
exposure
to
10
µ
g/
L
atrazine,
Johnson
(
1986)
reported
recovery
by
Day
7.
Although
the
rate
of
photosynthesis
may
recover
to
pre­
5
exposure
levels,
plant
biomass
in
atrazine­
treated
plants
do
not
recover
if
the
exposure
duration
was
sufficient
to
kill
some
plant
species,
as
reported
by
Kettle
et
al.
(
1987).

P.
16
to
18
Indirect
Aquatic
Effects
­
Syngenta
questions
EPA's
use
of
three
field
studies
as
primary
support
for
the
conclusion
that
atrazine
may
have
caused
indirect
effects
on
aquatic
communities,
including
reductions
in
populations
in
aquatic
macrophytes,
invertebrates
and
fish
at
from
10
to
20
µ
g/
L,
i.
e.,
Kettle
et
al.
1987,
Lampert
et
al.
1989,
and
Davies
et
al.
1994.
Syngenta
contends
that
these
field
studies
are
questionable
due
to
design
weaknesses
such
as
lack
of
replication,
lack
of
controls,
and
lack
of
consideration
of
possible
confounding
effects.
In
addition,
Syngenta
states
that
the
conclusion
is
inconsistent
with
the
weight
of
evidence
provided
in
numerous,
higher
quality
field
studies
and
the
conclusions
of
Giddings
et
al,
2000
and
2002,
i.
e.,
that
indirect
effects
on
aquatic
animals
are
unlikely
to
occur
at
or
below
the
inferred
noadverse
effect
level
for
plants
of
50
µ
g/
L.

EFED
has
responded
to
these
questions
previously
in
Urban
et
al.
2001,
in
its
responses
to
Syngenta's
earlier
comments
(
EFED
Review
of
Comments
from
Syngenta
and
its
Contractors
about
the
EPA
Revised
Environmental
Risk
Assessment
for
Atrazine,
April
9,
2002),
and
most
recently
in
Frankenberry
et
al.
2002.
All
three
provide
detailed
responses
to
Syngenta's
questions.
EFED
acknowledges
that
there
are
reasons
to
question
each
of
the
studies
that
Syngenta
considers
unreliable
or
questionable;
however,
EFED
continues
to
maintain
that
these
studies
and
others
identified
in
the
EPA
Revised
Risk
Assessment
for
Atrazine
are
useful
in
the
risk
assessment.
They
raise
reasonable
concerns
that
low
atrazine
levels
(
10
to
20
µ
g/
L)
may
indirectly
impact
fish
and
aquatic
invertebrate
populations
as
well
as
macrophyte
communities,
and
thus
support
the
EFED's
concerns
for
indirect
effects
on
aquatic
communities.

Giddings
et
al.
2002
also
identifies
Bester
et
al.
1995,
used
by
EFED
to
support
potential
reduction
in
primary
production
in
estuarine/
marine
aquatic
environments
as
questionable
and
unreliable.
Upon
further
review,
EFED
agrees
that
there
are
major
problems
with
this
study
and
will
remove
it
from
the
assessment
endpoints
of
concern
for
estuarine/
marine
environments.

As
noted
in
the
EPA
Revised
Risk
Assessment
for
Atrazine
(
page
12),
the
Giddings
et
al.
2002
conclusion
is
consistent
with
EPA's
position
when
they
states
that
in
certain
high­
exposure
situations,
atrazine
may
reach
concentrations
that
could
cause
ecologically
significant
effects
on
plant
productivity
and
community
structure.
In
cases
where
the
effects
are
severe
(
and
there
are,
to
our
knowledge,
no
confirmed
effects
on
plant
communities
in
nature),
indirect
effects
on
fish
and
invertebrates
are
possible."

P.
19
Direct
and
Indirect
Terrestrial
Effects
­
Syngenta
questions
EPA's
conclusion
that
in
areas
of
high
atrazine
use,
there
is
widespread
environmental
exposure
that
has
resulted
in
direct
acute
effects
on
many
terrestrial
plant
species
at
both
maximum
and
typical
use
rates,
as
well
as
indirect
terrestrial
wildlife
effects.
Syngenta
questions
the
basis
for
these
conclusions,
i.
e.,
tier
1
risk
quotients,
incident
reports,
and
a
review
paper
by
Freemark
and
Boutin,
1994.
6
As
noted
in
responses
to
previous
comments,
EFED
recognizes
that
risk
quotients
in
screening
level
risk
assessments
provide
a
basis
for
potential
adverse
effects
as
well
as
indicating
a
need
for
additional
data
for
a
refinement
in
the
risk
assessment.
EFED
acknowledges
that
there
are
no
field
studies
showing
the
linkage
between
direct
effects
and
indirect
effects
of
atrazine
on
wildlife.
However,
the
screening
level
assessment
shows
a
great
potential
for
direct
effects
on
terrestrial
plants.

In
a
previous
response
to
Syngenta's
comments
by
Syngenta
(
EFED
Review
of
Comments
from
Syngenta
and
its
Contractors
about
the
EPA
Revised
Environmental
Risk
Assessment
for
Atrazine,
April
9,
2002,
page
7),
EFED
referenced
Freemark
and
Boutin
(
1994)
to
propose
that
herbicide
use
has
had
an
adverse
indirect
effect
on
terrestrial
wildlife
via
habitat
alteration.
Syngenta
contends
that
that
is
not
the
whole
picture,
since
atrazine
has
a
role
in
conservation
practices
which
decrease
detrimental
effects
on
wildlife.
EFED
acknowledges
that
arguments
can
be
made
to
support
both
contentions,
but
definitive
data
remain
to
be
collected.
Therefore,
EFED
will
change
the
statements
to
clearly
indicate
that
there
is
a
potential
risk
for
indirect
effects
on
wildlife
due
to
atrazine
as
well
as
other
herbicides.

P.
21
­
25
Sublethal
Effects
­
Syngenta
claims
that
a
detailed
review
and
assessment
of
its
own
research
and
the
published
literature
on
the
potential
endocrine
effects
of
atrazine
in
fish,
amphibians
and
reptiles,
do
not
indicate
widespread
effects
or
serious
consequences
to
wildlife
from
potential
exposure
to
atrazine.
Syngenta
then
provides
a
review
of
available
atrazine
studies
on
largemouth
bass,
amphibians,
and
atrazine
olfactory
effects
on
salmon
attempting
to
explore
endocrine­
mediated
effects.
Syngenta
concludes
that
it
is
premature
to
consider
any
of
these
results
in
the
atraine
risk
assessment
until
the
studies
are
validated
and
the
significance
of
their
results
is
established.

EPA
is
in
the
process
of
evaluating
data
on
the
potential
endocrine
disrupting
effects
of
atrazine
on
a
range
of
non­
target
animals.
These
data
focus
primarily
on
the
effects
of
atrazine
on
endocrine­
mediated
gonadal
development
in
amphibians,
fish
and
aquatic
organisms.
The
studies
under
evaluation
have
been
submitted
by
pesticide
registrants
and
taken
from
open
literature.
They
include
both
published
and
unpublished
studies.
The
reported
results
vary
widely,
with
effects
on
gonadal
development
reported
at
exposures
from
0.1
parts
per
billion
(
ppb)
to
25
ppb.
EFED
is
in
the
process
of
reviewing
the
studies
at
hand
and
writing
data
evaluation
records
(
DERs).
Where
possible,
raw
data
from
these
studies
are
being
analyzed
and
study
methodologies
are
being
documented
to
determine
whether
procedural
errors
may
have
confounded
the
study's
ability
to
detect
treatment
effects.
Based
on
this
review,
EFED
will
evaluate
whether
the
data
support
the
conclusions
reached
by
the
study's
author.

Additional
pertinent
studies
are
being
performed
both
by
the
pesticide
registrants
and
by
independent
researchers,
which
could
shed
additional
light
on
this
sub­
lethal
endpoint
for
atrazine.
In
accord
with
the
agreement
reached
with
the
Natural
Resource
Defense
Counsel,
OPP
is
planning
to
summarize
all
these
studies
in
preparation
for
a
FIFRA
Science
Advisory
Panel
(
SAP)
meeting
scheduled
for
June
2003.
To
facilitate
the
preparation
of
this
summary
document,
OPP
will
continue
to
accept
studies
on
this
endpoint
up
to
February
28,
2003.
These
studies
can
7
only
be
accepted
in
final
form
and
must
represent
"
publication
quality";
no
preliminary
data
can
be
accepted
for
this
summary
document.
Data
submitted
prior
to
the
February
cutoff
date
will
be
included
in
the
summary
document
and
discussed
in
the
June
2003
SAP
meeting.
While
EPA
does
not
intend
to
exclude
any
pertinent
studies
from
consideration,
this
deadline
is
necessary
for
the
Agency
to
have
sufficient
time
to
review,
evaluate
and
summarize
the
available
studies
and
present
them
for
timely
science
review
by
the
FIFRA
SAP.
In
addition,
the
results
from
this
SAP
meeting
are
considered
crucial
input
for
the
amended
for
the
interim
reregistration
eligibility
decision
on
atrazine
scheduled
for
October
2003.

Concerning
the
Syngenta
sponsored
Wieser
and
Gross
(
2002)
study
on
largemouth
bass
(
LMB)
which
EPA
has
reviewed
and
evaluated,
Syngenta
questions
EPA's
contention
that
the
presence
of
quantitative
levels
of
vitellogenin
(
egg
yolk
precursor
protein)
in
male
bass
indicates
exposure
to
an
estrogenic
compound,
and
EPA's
conclusion
that
high
variability
confounds
the
study.
Syngenta
noted
that
male
fish
can
express
variable
amounts
of
this
protein,
and
that
the
interpretations
and
differences
presented
in
the
report
are
both
conservative
and
within
normal
physiological
expectations.

EPA
believes
that
vitellogenin
is
a
sex­
specific
protein
and
can
only
be
detected
in
males
exposed
to
estrogenic
substances.
While
Syngenta
is
correct
that
male
fish
can
express
variable
amounts
of
this
protein,
LMB
can
only
do
so
in
response
to
variable
levels
of
estrogen.
The
current
test
results
indicate
that
this
particular
vitellogenin
assay
provided
inconsistent
results.
Given
that
mean
vitellogenin
levels
in
male
LMB
exposed
to
100
ug/
L
atrazine
was
7
times
control
values,
the
test
was
unable
detect
a
treatment
effect.
EFED
is
concerned
that
atrazine's
ability
to
impact
vitellogenesis
may
be
a
subtle
effect
that
the
present
study
was
unable
to
differentiate
conclusively.
Furthermore,
recent
evidence
has
suggested
that
atrazine
affects
on
reproduction
may
include
a
suite
of
endpoints
not
typically
captured
by
current
guideline
studies.
It
is
clear
from
the
largemouth
bass
study
that
endogenous
hormones
are
affected.
However,
the
ability
of
atrazine
to
induce
vitellogenesis
in
males
remains
uncertain.

P.
25
Endangered
Species
­
Syngenta
notes
that
they
have
previously
identified
inconsistencies
in
EPA's
risk
assessment
as
it
pertains
to
endangered
species.
In
addition,
Syngenta
states
that
EPA
inaccurately
indicates
that
terrestrial
plant
endangered
species
are
likely
to
be
affected
and
wildlife
species
may
be
indirectly
affected
by
atrazine
use.

EPA
bases
it
risk
concerns
for
endangered
terrestrial
plants
risk
exceedences,
and
these
risk
exceedences
are
based
on
spray
drift
and
runoff
into
the
habitats
for
terrestrial
and
semi­
aquatic
plants.
The
revised
EPA
risk
assessment
clarifies
this
point.
In
addition,
EPA
has
revised
the
section
under
Endangered
Species
Concerns
to
eliminate
statements
indicating
that
wildlife
species
may
be
indirectly
affected
by
atrazine
use.
EPA
determined
that
such
statements
for
atrazine
alone
were
speculative.
Rather,
the
use
of
any
herbicide
could
have
adverse
chronic
effects
on
terrestrial
and
aquatic
plants
in
areas
adjacent
to
treated
fields
that
may
have
indirect
effects
on
these
animals
from
the
loss
of
food
sources
and
the
loss
of
vegetative
habitat
for
cover,
reproduction
and
the
survival
of
offspring
(
Freemark
and
Boutin,
1994).
8
P.
26
Environmental
fate
Syngenta
questions
that
EFED
has
not
used
the
best
available
environmental
fate
data
for
atrazine.
Specifically,
Syngenta
has
suggested
that
EFED
consider
the
submissions
by
Burnett
et
al,
2000
and
2002.

In
reviewing
the
two
submissions
by
Burnett,
et
al.,
EPA
found
that
only
summaries
were
presented.
EFED
needs
to
review
complete
study
reports
in
order
to
determine
a
study's
acceptability.
In
the
atrazine
technical
briefings
in
April,
2002,
EFED
used
Syngenta's
`
suggested'
fate
parameters
as
input
for
the
modeling
runs
to
demonstrate
that
the
resultant
exposure
values
still
pose
potential
aquatic
risk
concerns.
With
regard
to
the
results
of
the
new
soil
photolysis
study,
EFED
will
upgrade
the
DER's
classification
to
acceptable.
However,
regarding
the
use
of
this
parameter
in
the
Pesticide
Root
Zone
Model
(
PRZM)
runs,
currently,
EFED's
guidance
does
not
support
the
use
of
this
input
parameter.
EFED
believes
that
the
suggested
changes
to
PRZM
would
have
a
minimal
impact
on
the
model's
output
and
therefore
would
be
unlikely
to
alter
the
current
risk
assessment.

P.
27
Clarification
of
Estuarine
Analysis
Syngenta
states
that
"
This
conclusion
[
of
estuary]
is
in
error
since
the
atrazine
concentration
data
used
in
the
EPA
comparisons
are
not
representative
of
concentrations
monitored
in
the
Chesapeake
Bay
proper
and
near­
shore
embayments.
These
higher
concentrations
are
representative
of
freshwater
sites
on
streams
and
rivers
within
the
Chesapeake
watershed
and
not
sites
within
the
open
waters
of
the
Bay
or
near­
shore
Bay
waters."
Syngenta
also
questions
the
use
of
data
from
the
Louisiana
Department
of
Agriculture
and
Forestry
(
LDAF).

EFED
questions
how
the
registrant
arrives
at
this
conclusion,
especially
the
claim
that
the
higher
concentrations
are
representative
of
freshwater
sites.
In
the
book
entitled
"
Life
in
the
Chesapeake
Bay:
An
Illustrated
Guide
to
Fishes,
Invertebrates,
and
Plants
of
Bays
and
Inlets
from
Cape
Cod
to
Cape
Hatteras,"
the
authors
(
Alice
and
Robert
Lippson)
describe
the
Chesapeake
Bay,
which,
"
like
all
estuaries,
contains
waters
that
range
from
fresh
to
nearly
as
salty
as
ocean
waters.
An
estuary
is
defined
as
a
somewhat
restricted
embayment
in
which
the
flow
of
freshwater
mixes
with
high
salinity
ocean
water.
At
the
head
of
the
Bay
and
at
the
head
of
each
tributary
stream
(
the
geographical
fall
line),
tidal
influence
is
apparent,
but
little
or
no
ocean­
derived
salt
is
present."
First,
the
registrant
did
not
provide
any
salinity
data
to
show
that
the
monitoring
results
are
for
fresh
water,
and
secondly,
freshwater
can
also
be
part
of
an
estuary.

The
same
statement
can
be
applied
to
the
Louisiana
estuary.
If
the
registrant
can
provide
more
monitoring
data
at
the
locations
which
they
view
as
"
estuarine"
sites,
then
the
results
can
be
used
for
refinement
purposes.
No
monitoring
data
were
presented
to
verify
the
assumption
of
no
detections.

The
registrant
claims
that,
based
upon
the
available
toxicity
data,
there
is
no
indication
that
estuarine
organisms
are
more
sensitive
to
atrazine
than
freshwater
organisms
and
refers
to
9
Giddings
et
al.
2000.
However,
a
closer
review
of
page
168
and
169
of
that
report,
Table
5.3
(
Summary
of
atrazine
acute
toxicity
distributions)
and
Table
5.4
(
Summary
of
atrazine
chronic
toxicity
distributions)
show
that
while
the
sensitivities
of
freshwater
and
saltwater
organisms
to
atrazine
are
similar,
the
10th
percentiles
for
chronic
effects
for
all
saltwater
species
are
almost
three
times
lower
than
those
for
freshwater
species
(
4.9
to
16
µ
g/
L)
.
The
number
of
chronic
studies
upon
which
this
comparison
is
based
is
limited
(
n
=
8
for
saltwater
species;
23
for
freshwater
species);
however,
increased
concern
for
potential
adverse
effects
in
estuarine
systems
appears
to
be
warranted
at
this
time.

Syngenta's
Supplement
to
"
Aquatic
Ecological
Risk
Assessment
of
Atrazine
­
A
Tiered
Probabilistic
Approach"
Including
Responses
to
EPA's
Comments,
February
26,
2002
(
Giddings
et
al.
2002)
[
OPP
02­
0026­____]

Syngenta's
Atrazine
Ecological
Risk
Assessment
Panel
,
EcoRisk,
Inc.
provided
EPA
with
a
supplemental
document
(
Giddings
et
al.
2002)
to
their
original
tiered
probabalistic
risk
assessment
for
atrazine
(
Giddings
et
al.
2000).
The
former
is
a
formal
response
to
Urban
et
al.
2002
which
in
turn
was
EPA's
response
to
Giddings
et
al.
2000.

EPA
has
prepared
its
review
and
response
to
the
supplemental
report
on
atrazine
(
Frankenberry
et
al.
2002).
Following
the
pattern
of
these
reports
from
Syngenta's
Atrazine
Ecological
Risk
Assessment
Panel,
this
response
(
like
the
previous
one)
is
organized
by
tiers.
With
the
technical
assistance
of
it's
contractors,
EPA
provides
responses
to
the
issues
raised
as
well
as
a
critique
of
the
supplemental
information.
In
summary,
the
EPA
concludes
that
the
supplemental
information
is
not
sufficiently
compelling
to
warrant
a
significant
change
in
EPA's
conclusions
presented
in
the
Revised
Environmental
Risk
Assessment
for
Atrazine
dated
April
22,
2002,
i.
e.,
"
Based
on
the
results
of
this
refined
assessment,
the
Agency
finds
that
in
areas
of
high
atrazine
use,
there
is
widespread
environmental
exposure
that
(
1)
may
have
resulted
in
and
may
continue
to
result
in
direct
acute
effects
on
many
terrestrial
plant
species
at
both
maximum
and
typical
use
rates,
(
2)
may
have
caused
and
may
continue
to
cause
direct
effects
on
aquatic
non­
vascular
plants
which
in
turn
could
have
caused
reductions
in
primary
productivity,
(
3)
may
have
caused
and
may
continue
to
cause
reductions
in
populations
of
aquatic
macrophytes,
invertebrates
and
fish,
(
4)
may
have
caused
and
may
continue
to
cause
indirect
effects
on
aquatic
communities
due
to
loss
of
species
sensitive
to
atrazine
and
resulting
in
changes
in
structure
and
functional
characteristics
of
the
affected
communities.
Potential
adverse
effects
on
sensitive
aquatic
plants
and
other
nontarget
aquatic
organisms
as
well
as
their
populations
and
their
communities,
are
likely
to
be
greatest
where
atrazine
concentrations
in
water
equal
or
exceed
approximately
10
to
20
µ
g/
L
on
a
recurrent
basis
or
over
a
prolonged
time
period.
Based
on
monitoring
data,
maximum
concentrations
at
up
to
35%
of
the
sites
exceeded
the
atrazine
concentration
(>
10
µ
g/
L)
at
which
these
adverse
effects
are
found
in
simulated
field
studies.
Up
to
20%
of
the
sites
exceeded
the
atrazine
concentration
(>
20
µ
g/
L)
at
which
adverse
effects
are
found
in
simulated
field
studies
as
well
as
many
of
the
10th
[
per]
centile
values
for
acute
and
chronic
effects
from
analyses
of
laboratory
data.
The
frequency
of
occurrence
and
extent
of
the
potential
impacts
will
vary
depending
upon
the
type
of
water
bodies
and
their
proximity
in
time
and
space
to
atrazine
10
applications.
Recovery
from
the
effects
of
atrazine
and
the
development
of
resistence
to
the
effects
of
atrazine
in
some
vascular
and
non­
vascular
aquatic
plants
is
reported
and
adds
uncertainty
to
these
findings.
Further
research
is
needed
to
quantify
the
impact
that
these
effects
would
have
on
these
risk
conclusions."

Triazine
Network
­
written
comments
(
OPP­
34237C)
[
OPP
02­
0026­
0218]

P.
10
Analysis
Failed
To
Consider
Trend
of
Atrazine
Usage
and
Application­­
The
Triazine
Network
believes
that
the
EFED's
evaluations
are
not
representative
of
current
conditions
and
therefore
overstate
atrazine
exposure.

Table
1
presented
by
the
Triazine
Network
shows
the
patterns
of
decreasing
usage
and
decreasing
rates
of
application
per
unit
area.
EFED
would
like
to
see
the
rate
reduction
reflected
in
the
labels.
If
the
lower
rate
applications
are
feasible
and
accurate
as
the
data
suggest,
then
there
is
no
need
to
maintain
the
higher
maximum
application
rates
as
listed
on
the
current
label,
and
label
reduction
would
be
a
best
mitigation
measure.
Except
for
the
farm
pond
exposure
assessment
which
is
based
on
modeling,
EFED
has
based
its
assessment
on
available
monitoring
data
for
other
types
of
water
bodies.
The
monitoring
data
should
be
reflective
of
the
recent/
current
conditions;
therefore
EFED
is
not
overstating
atrazine
exposure.

P.
11
Trends
in
Surface
and
Groundwater
Water
Concentration­­
The
Triazine
Network
claims
that
in
its
review
of
the
median
atrazine
concentration,
observed
in
the
high­
use
areas,
the
atrazine
levels
have
consistently
decreased
over
time.

Figure
3
presented
by
the
Triazine
Network
shows
the
median
concentrations
from
1989
to
1996.
Upon
closer
examination,
EFED
does
not
observe
a
clear,
decreasing
trend
in
this
figure.
Also,
since
the
Triazine
Network
has
not
provided
any
monitoring
data
at
smaller
scale
sites
such
as
farm
ponds,
which
are
widely
distributed
in
agricultural
ecosystems,
EFED
can
not
determine
whether
or
not
atrazine
concentrations
in
the
surface
water
of
such
ecosystems
have
truly
decreased.

Use
of
data
from
community
water
systems
to
demonstrate
trends
can
be
misleading,
since
the
results
are
for
finished
water.
It
would
be
expected
that
due
to
better
detection
and
treatment
technologies,
the
atrazine
in
finished
water
would
be
lower,
as
the
data
suggest.
For
the
purpose
of
demonstrating
decreasing
atrazine
levels,
the
Triazine
Network
should
analyze
raw
water
samples
and
not
finished
water
samples.
EFED
has
reviewed
the
trend
analysis
reports
and
found
that
for
nearly
63%
(
or
39/
62)
of
the
CWSs,
the
data
indicate
either
no
change
in
the
concentration
or
an
increase
in
the
concentration.
Reevaluating
the
data
with
"
=
0.05
or
0.01
would
likely
result
in
even
fewer
indications
of
a
negative
trend.

P.
13
Occurrence
of
Atrazine
in
Surface
and
Groundwaters
Focused
on
High
Use
Areas
and
Characterized
Worst­
Case
Impacts­­
The
Triazine
Network
criticizes
EFED
for
focusing
on
high
use
areas
based
on
limited
sampling
data
and
only
considering
the
maximum
concentrations.
11
They
point
out
that
the
EFED
report
focused
on
sampling
data
from
receiving
waters
within
21
major
atrazine
use
states.

In
the
original
RED,
EFED
considered
primarily
the
PLEX
monitoring
data,
which
were
available
only
for
21
states
at
that
time.
With
subsequent
submissions
by
Syngenta,
EFED
has
also
reviewed
and
considered
an
additional
11
states.
EFED
is
considering
all
the
submitted
data
from
a
total
of
32
states
from
1993
to
2000.

P.
13
Limited
Sampling­­
The
Triazine
Network
questions
that
in
every
case,
EPA
based
its
evaluation
on
a
consideration
of
limited
data
sets.

EFED
agrees
with
the
limitations
of
sampling
as
identified
by
the
Triazine
Network,
and
is
puzzled
by
its
comments
that
EFED
did
not
consider
all
the
available
data.
EFED
has
pointed
out
that
the
available
monitoring
programs,
including
the
U.
S.
Geological
Survey
(
USGS)
76
mid­
western
reservoirs,
National
Water
Quality
Assessment
(
NAWQA)
and
others,
were
not
designed
to
capture
possible
peak
atrazine
concentrations
which
usually
occur
in
association
with
high
rainfall
events
closely
following
atrazine
applications.
Since
the
monitoring
results
did
not
likely
capture
the
possible
peaks,
EFED
can
not
derive
the
true
90th,
95th,
or
99th
percentile
values
with
confidence.
For
this
reason,
EFED
decided
to
present
the
results
as
shown
in
the
RED.

P.
14,
15
Maximum
Concentrations
(
Lakes,
Streams,
Estuaries)­­
EFED
methods
overestimate
the
nature
of
potential
effects
by
comparing
peak
exposure
values
to
chronic
toxicity
test
endpoints.
The
Triazine
Network
expressed
concerns
that
the
use
of
"
maximum
values"
from
the
abundance
of
monitoring
data
was
inappropriate
and
that
time­
weighted
averages
should
be
calculated
and
used
in
the
risk
assessment.

EFED
has
noted
that
most
of
the
sampling
programs
were
not
designed
to
allow
calculation
of
appropriate
time­
weighted
averages.
In
the
absence
of
such
data,
available
exposure
data
must
be
used.
Since
these
values
likely
under
represent
peak
values,
they
are
more
likely
closer
to
longer
term
averages.
As
such
their
use
in
the
risk
assessment
more
likely
approximates
a
timeweighted
average
exposure
rather
than
a
peak
exposure.

EFED
has
constructed
the
exceedence
curves
using
Weibull
plots
by
ranking
the
available
maximum
concentrations
for
each
of
the
water
bodies
and
assigning
a
probability
associated
with
the
rank
divided
by
the
sum
of
the
number
of
samples
plus
1.
The
use
of
this
methodology
may
be
subject
to
discussion,
but
with
the
limitations
of
available
monitoring
data,
EFED
felt
that
it
is
the
most
reasonable
way
to
characterize
potential
exposure
conditions.

The
Triazine
Network
claimed
that
the
USGS
study
of
9
mid­
western
streams
was
specifically
conducted
to
compare
the
peak
atrazine
concentration
following
runoff
events
and
this
study
obtained
samples
collected
at
close
intervals
readily
amenable
to
time
series
analysis.
If
the
Triazine
Network's
contention
is
correct,
they
are
encouraged
to
perform
the
analyses
and
present
the
results
to
EFED.
12
P.
16
Detailed
Evaluation
of
Magnitude
and
Duration
of
Atrazine
Exposure
Demonstrates
Exposure
Evaluation
Flawed­­
The
Triazine
Network
states
that
without
accurately
defining
the
duration
of
exposure,
it
is
impossible
to
associate
an
assessment
endpoint
with
a
concentration.

EFED
agrees
with
the
assessment
that
"
without
accurately
defining
the
duration
of
exposure,
it
is
impossible
to
associate
an
assessment
endpoint
with
a
concentration."
EFED
has
explained
previously
that
due
to
the
limited
availability
of
time
series
data,
the
exposure
duration
can
not
be
accurately
defined,
and
EFED
has
decided
to
use
the
available
maximum
observations
for
each
available
sampling
site
to
construct
the
exposure
exceedence
probability
curve.
The
Triazine
Network
claimed
that
there
are
peak
atrazine
concentrations
being
captured
in
some
streams
and
lakes,
and
EFED
should
base
its
assessment
on
these
to
characterize
long­
term
average
atrazine
concentration.
It
is
not
clear
to
EFED
what
sets
of
monitoring
data
are
being
referred
to.
The
Triazine
Network
is
encouraged
to
present
their
findings.
Even
if
the
quality
and
quantity
of
monitoring
data
were
good
enough
for
these
streams
and
lakes,
they
could
only
be
used
for
those
specific
monitoring
sites;
due
to
the
limited
number
of
monitoring
sites
available,
the
monitoring
data
could
not
be
considered
representative
of
all
areas
in
which
atrazine
is
used.

P.
16
USGS
Data
on
Small
Streams­­
The
Triazine
Network
presents
its
approach
in
analyzing
this
dataset.

The
Triazine
Network
presented
the
time
series
plots
of
Maple
Creek,
Nebraska,
Robert's
Creek,
Iowa
and
Silver
Creek,
Illinois.
Figure
5
represents
the
results
of
Maple
Creek.
There
are
no
monitoring
results
for
1994,
1995,
and
1996.
The
monitoring
results
in
this
figure
clearly
demonstrate
that
sampling
frequency
is
insufficient
to
accurately
characterize
long­
term
exposure.
Figures
6
and
7
are
time
series
plots
for
Robert's
Creek,
and
Silver
Creek,
respectively.
EFED's
plots
are
based
on
the
monitoring
results
in
USGS
Open­
File
Report
94­
396
for
the
period
from
1990
to
1992.
Since
the
Triazine
Network
has
the
most
recent
results
from
April
2002
to
July/
August
2002,
EFED
would
like
to
request
a
copy
of
these
data.
As
for
Table
3,
which
presents
the
summary
results
of
nine
streams,
the
Triazine
Network
calculated
the
30­
day
and
60­
day
averages
based
on
observations
within
30
or
60
calendar
days.
The
use
of
available
observations
in
calculating
the
average
may
miss
the
true
average
value
since
samples
were
not
collected
daily.
Even
in
this
case,
the
results
of
the
last
sampling
site,
West
Fork
Big
Blue
River,
Nebraska,
clearly
show
exceedence
of
the
draft
chronic
criterion
of
12.0
ug/
L
atrazine.

P.
20
USGS
Data
on
Lakes/
Reservoirs­­
The
Triazine
Network
presents
its
approach
in
analyzing
this
data­
set.

EFED
has
cited
the
observed
peak
concentrations
of
12.4
ug/
L
in
1992
and
11.0
ug/
L
in
1993.
These
were
based
on
the
USGS
Open­
Report
96­
393,
which
consisted
of
3
to
5
grab
samples
collected
in
each
water
body
in
1992
and
1993.
Three
to
five
samples
are
insufficient
to
characterize
long
term
exposure,
especially
since
it
is
quite
possible
the
sampling
scheme
of
only
3
to
5
samples
per
site
will
likely
miss
the
"
true"
peak
concentrations.
More
rigid
and
frequent
sampling
is
needed
to
truly
quantify
the
proper
exposure
duration
results.
13
P.
22
Reassessment
of
Estuarine
Exposure
to
Atrazine­­
The
EFED
assessment
did
not
address
atrazine
in
the
open
waters
of
an
estuary.

The
Triazine
Network
claimed
that
EFED
did
not
address
atrazine
in
the
open
waters
of
an
estuary.
EFED
used
available
data
such
as
that
from
the
Upper
Terrebonne
Watershed
and
the
Chesapeake
Bay.
Although
the
sampling
data
may
not
have
been
taken
in
the
open
waters,
they
are
in
close
proximity
to
the
open
water;
and
in
the
absence
of
more
appropriate
data,
EFED
believes
that
these
data
are
reasonable
surrogates.
EFED
can
not
refine
it's
analysis
without
additional
data.

All
available
monitoring
data
within
the
Chesapeake
Bay
should
be
representative
of
an
estuary.
EFED
based
its
assessment
on
the
monitoring
data,
even
though
they
are
limited
in
their
scope
of
sampling
sites
and
frequencies.
The
Triazine
Network
summarized
the
data
as
shown
in
Table
4.
They
classified
the
data
according
to
receiving
water
type
and
presented
the
maximum,
Spring
mean
and
overall
average
values.
The
Triazine
Network
suggested
that
if
these
are
then
compared
with
EFED's
assessment
endpoint,
there
are
no
expected
environmental
impacts.
Before
these
inferences
can
be
made,
EFED
needs
to
closely
examine
the
data.
Due
to
the
limited
data,
Table
4
did
not
show
any
Spring
means
for
non­
tidal
creeks,
nor
were
some
of
the
overall
averages
available.
The
limited
monitoring
data
can
be
used
as
a
potential
indicator
of
atrazine
residues.
For
the
purpose
of
a
more
precise
risk
assessment,
however,
sampling
frequencies
and
locations
would
have
to
be
markedly
increased.

P.
24
Surface
Water
Exposure
Summary­­
The
Triazine
Network
claims
that
EFED
has
not
properly
evaluated
the
available
data
to
calculate
the
long­
term
exposure
of
atrazine.

The
Triazine
Network
discussed
various
water
bodies:
small
streams,
large
streams
and
rivers,
lakes
and
reservoirs,
and
estuaries.
The
potential
exposure
in
small
streams
can
be
representative
of
possible
worst­
case
exposure.
There
are
many
small
streams
in
the
atrazine
use
area,
in
which
the
impact
can
be
great.
Also,
the
Triazine
Network
did
not
discuss
the
exposure
of
the
"
farm"
pond
environment.
Without
specific
monitoring
data
for
farm
pond
environments,
EFED
has
had
to
rely
on
the
PRZM
coupled
with
the
Exposure
Analysis
Model
System
(
EXAMS)
modeling
results
indicating
that
the
aquatic
exposure
is
likely
to
be
highest
for
this
type
of
environment.
(
On
page
35)
The
Network
claims
that
EFED
did
not
provide
any
data
to
support
the
model
results.
It
questions
EFED's
model
results,
which
suggest
that
elevated
concentrations
persist
for
an
extended
period
of
time.
Furthermore,
it
claims
that
if
a
farm
pond
responds
in
this
manner,
it
should
be
relatively
easy
to
find
high
atrazine
concentrations
in
farm
ponds
long
after
the
spring
application
period.
The
lack
of
any
supporting
data
compromises
the
credibility
of
the
Triazine
Network's
assertions.
EFED
disagrees
with
the
Network's
assessment
and
welcomes
the
Network
to
provide
monitoring
data
from
farm
ponds
in
order
to
substantiate
their
claims
and
refute
EFED's
assessment.
However,
based
on
available
monitoring
data
for
other
types
of
water
bodies,
EFED
expects
atrazine
exposure
to
be
higher
due
to
less
dilution
and
to
the
close
proximity
of
the
treated
agricultural
field
to
the
farm
pond.
14
P30­­
All
studies
indicating
effects
at
low
levels
(<
20ppb)
were
reversible
effects,
or
were
ecologically
insignificant
(
i.
e.
would
not
reduce
populations
significantly).

The
Triazine
Network
believes
that
the
majority
of
effects
which
occur
at
<
20ppb,
are
reversible
or
ecologically
insignificant,
i.
e.,
would
not
significantly
reduce
populations,
and
should
not
be
used
in
the
risk
assessment.
Recovery
is
defined
as
a
return
of
a
measured
parameter
to
the
normal
range
of
the
controls.
Whether
or
not
recovery
occurs
for
effects
observed
at
atrazine
concentrations
<
20ppb
depends
on
the
selected
parameters
and
a
careful
assessment
of
preexposure
status.

EFED
has
selected
direct
effects
on
terrestrial
and
aquatic
plants
and
indirect
effects
on
aquatic
communities
as
assessment
endpoints.
There
is
little
argument
that
these
effects
can
result
from
atrazine
exposure.
However,
EFED
contends
that
the
issue
of
recovery
is
highly
speculative
and
subjective.
Without
complete
characterization
of
pre­
exposure
conditions,
post­
exposure
conditions
can
not
adequately
be
addressed,
and
certainly
can
not
be
classified
as
recovery.
This
issue
is
also
discussed
in
Frankenberry
et
al
2002,
pages
15,
and
21
to
24.
While
EFED
agrees
that
the
extent
of
ecological
effects
resulting
the
loss
of
sensitive
species
is
uncertain,
in
the
absence
of
data,
it
would
be
speculative
and
inconsistent
with
EPA's
mandate
to
be
protective
to
conclude
that
no
adverse
effects
on
aquatic
populations
or
communities
are
likely.

P32­­
Other
factors
could
affect
plant
growth.

EFED
acknowledges
the
role
other
factors
can
play
in
organism
response.
Atrazine
effects
are
dependent
on
available
light
since
it
affects
photosynthesis.
In
the
absence
of
light,
atrazine
has
little
effect.
If
light
penetration
is
diminished,
so
should
the
effect
be
diminished.
However,
EFED
can
not
confidently
state
whether
or
not
any
remaining
effect
upon
plants
is
significant,
nor
the
relative
importance
of
agrochemical
and
non­
chemical
effects.

P32­­
Algal
endpoints
should
not
be
applied
to
small
streams.

The
Triazine
Network
does
not
believe
that
algal
endpoints
should
be
applied
to
streams
where
algal
communities
will
not
be
prevalent.
EFED
responds
that
retention
times
in
small
temporary
and
permanent
streams
are
irrelevant
for
other
photosynthetic
plants
species
such
as
periphytic
algal
communities
upon
which
numerous
macroinvertebrate
organisms
depend
for
survival.

P35­­
Impact
evaluation
in
ponds
should
be
deferred
until
confirmation
of
1)
atrazine
persistence
in
ponds,
2)
ecological
relevance
of
ponds,
and
3)
extension
of
federal
regulations
to
farm
ponds.

The
Triazine
Network
does
not
accept
the
atrazine
concentration
profile
modeled
for
farm
ponds
given
the
pattern
of
atrazine
attenuation
observed
in
stream
and
reservoir
monitoring
data.
EFED
recognizes
the
limitation
of
current
models
to
take
overflow
into
consideration.
However,
many
farm
ponds
in
fact
do
not
overflow
with
each
rain
event,
doing
so
only
when
major
rainfalls
take
place.
Evaporation
does
occur
but,
as
stated
in
the
Triazine
Network
document,
would
not
result
in
complete
drying
of
the
water
body
unless
there
is
a
severe
drought.
Complete
drying
would
15
also
occur
if
the
water
body
was
quite
shallow,
less
so
than
the
standard
water
body
scenario
modeled
by
EFED.

PRZM/
EXAMS
models
are
a
Tier
2
aquatic
exposure
models
that
EFED
and
the
regulated
community
use
to
estimate
pesticide
concentrations
in
surface
waters
following
applications
of
pesticides
to
agricultural
crops.
Typically,
these
concentrations
are
compared
to
ecotoxicological
values
derived
from
laboratory
or
field/
semi­
field
tests,
in
order
to
arrive
at
a
ratio
of
estimated
exposure
to
observed
effects,
i.
e.,
a
risk
quotient.
These
risk
quotients
are
compared
to
established
levels
of
concern
so
that
EFED
can
conclude
that
there
either
is
or
is
not
a
potential
risk
to
aquatic
organisms
from
exposure
to
the
pesticide
in
surface
water,
not
farm
ponds.
We
acknowledge
that
because
the
modeled
system
is
a
farm
pond,
applying
these
risk
estimates
to
other
types
of
water
bodies
increases
the
uncertainty
of
our
risk
estimates.
This
uncertainty
can
be
reduced
by
refining
the
risk
assessment,
i.
e.,
by
gathering
additional
data,
either
ecotoxicological
or
exposure
data.
Concerning
the
exposure
data,
additional
exposure
data,
such
as
monitoring
data
for
other
water
bodies,
such
as
chemical
monitoring
data,
can
reduce
the
uncertainty
in
the
PRZM/
EXAMS
models
due
to
modeled
system
being
a
farm
pond.

The
Triazine
Network
does
not
believe
that
farm
ponds
are
ecologically
relevant.
Many
natural
water
bodies
have
been
eliminated
over
the
past
century
through
developmental
and
agricultural
activities.
These
water
bodies
range
from
small
ephemeral
pools
that
last
a
few
weeks
to
larger,
more
permanent
water
bodies.
Many
organisms
have
depended
on
these
water
bodies
either
throughout
the
year
(
most
obligatory
aquatic
organisms)
as
their
only
site
of
existence
or
for
brief
periods
of
time
as
a
layover
point
during
migration
(
waterfowl),
as
feeding
sites
(
mammals,
birds,
reptiles)
or
as
breeding
sites
(
amphibians,
birds,
reptiles).
In
the
absence
of
natural
water
bodies,
these
organisms
must
find
suitable
replacements.
Farm
ponds
serve
act
as
replacements
for
these
natural
water
bodies,
and
are
therefore
ecologically
relevant
and
essential
for
the
organisms
that
depend
on
small
water
bodies
for
their
existence.

Comments
on
EFED
RED
Chapter
for
Atrazine,
April
22,
2002
­
by
Atrazine
Ecological
Risk
Assessment
Panel
(
June
27,
2002),
submitted
by
Syngenta
[
OPP
02­
0026­____]

P.
4
Exposure
Modeling­­
The
panel
questions
EFED's
exposure
modeling
approach
and
the
use
of
selective
parameters
for
model
inputs.

The
panel
discussed
the
issues
of
scenario,
application
methods,
and
physicochemical
properties.
Relating
to
the
scenario
issues,
neither
the
dynamic
pond
volume
issue
nor
the
approach
used
have
not
yet
been
officially
reviewed
by
EFED.
The
buffer
zone
issue
and
its
relationship
to
runoff
can
not
be
properly
quantified
by
the
model.
With
regard
to
application
methods,
if
aerial
application
is
very
rare
for
corn
and
sugarcane,
then
it
is
desirable
to
remove
this
method
of
application
from
the
label.
As
for
the
physicochemical
properties,
EFED
follows
the
model
input
parameter
guidance
document
to
perform
PRZM/
EXAMS
modeling
simulations.
Based
on
the
results
presented
by
the
panel
with
variable
pond
volume
consideration,
the
exposure
estimate
using
the
model
PONDWQ
for
the
upper
10th
percentile
value
is
about
25
ug/
l
(
ppb),
which
is
in
16
the
range
of
aquatic
risk
concern.
In
addition,
the
farm
pond
environment
is
assumed,
as
a
minimum,
to
have
10
hectares
of
drainage
area
around
the
pond
to
maintain
water
in
the
pond
during
droughts.
This
is
based
on
the
requirement
of
drainage
area
according
to
the
USDA
Ponds
Handbook
(
No.
590)
for
the
Mississippi,
Alabama,
and
Georgia
Region.
Given
the
smaller
annual
rainfall
amount
in
Ohio
area,
the
minimum
drainage
area
should
be
higher;
however,
this
factor
is
not
reflected
in
the
current
modeling
simulations.

Since
there
are
several
aspects
up
for
discussion
regarding
the
proper
use
of
model
simulations,
it
would
be
helpful
for
the
registrant
to
provide
monitoring
results
from
this
type
of
water
body
for
discussion
of
EFED's
modeling
results.

P.
4
Exposure
Model
Sensitivity
Analysis­­
The
panel
brings
up
the
issue
of
using
model
parameters
for
soil
photolysis
and
plant
uptake.

The
panel
raised
the
input
parameter
issue
about
soil
photolysis
and
plant
uptake.
The
use
of
this
information
is
not
yet
approved
according
to
EFED's
model
input
parameter
guidance
document.
Even
if
they
were
to
be
considered,
EFED
believes
that
the
change
would
not
be
large
enough
to
impact
the
modeling
results.
Again,
as
mentioned
previously,
it
will
be
helpful
to
examine
monitoring
results
for
this
type
of
water
environment
for
comparison
with
the
modeling
results.

P.
5
Influence
of
Mixing
on
Exposure­­
The
panel
questions
the
accuracy
in
the
modeling
of
farm
pond
scenarios
regarding
the
influence
of
mixing
for
larger
water
bodies.

EFED
only
uses
PRZM/
EXAMS
to
assess
the
farm
pond
environment.
For
large
water
bodies,
EFED
uses
available
monitoring
data
for
the
exposure
assessment.

P.
5
Use
of
Monitoring
Data
in
Probabilistic
Ecological
Risk
Assessment­­
The
panel
questions
EFED's
descriptive
statement
(
pp.
4,
53)
concerning
the
NAWQA
monitoring
data.

The
panel
questions
EFED's
statement
(
pp.
4,
53)
concerning
the
study
design
and
use
of
NAWQA
monitoring
data.
The
statements
on
pages
4
and
53
are
based
on
descriptions
and
consultation
from
USGS
scientists
who
have
been
involved
in
the
NAWQA
study.
EFED
is
puzzled
by
the
panel's
claim
that
the
distribution
of
concentrations,
including
upper
centiles,
can
be
estimated
from
proper
analysis
of
such
data,
given
the
Panel's
awareness
that
maximum
concentrations
are
likely
missed
due
to
the
infrequent
sampling
design
of
the
NAWQA
study.

The
panel
also
questions
whether
the
pesticide
was
used
as
directed
or
not.
Since
monitoring
data
were
used,
EFED
would
assume
that
the
results
should
be
reflective
of
directed
use,
rather
than
misuse,
unless
the
panel
has
evidence
otherwise.

As
for
the
claim
about
the
unreliability
of
two
high
detections
in
Robert's
Creek,
EFED
relies
on
the
quality
assurance/
quality
control
(
QA/
QC)
performed
by
USGS
on
all
of
the
data
it
collects.
17
P.
6
Concentrations
in
Estuaries­­
The
panel
questions
EFED's
use
of
monitoring
data
for
estuarine
environments.

As
a
common
practice,
"
salinity"
is
used
to
determine
whether
a
water
body
is
brackish
or
fresh
water.
Even
if
the
samples
are
fresh
water
samples,
they
were
taken
from
the
Chesapeake
Bay
and
constitute
that
part
of
its
makeup,
as
noted
above,
where
"
little
[
or
no]
ocean­
derived
salt
is
present."

With
regard
to
the
Upper
Terrabonne
Watershed
in
Louisiana,
EFED
based
its
analysis
on
the
available
data.
Although
the
sampling
data
may
not
have
been
taken
in
the
open
waters,
they
are
in
close
proximity.

P.
6
Persistence
in
Lake
Michigan
and
in
Streams
 
The
panel
questions
EFED's
reference
with
regard
to
estimated
half
life.

The
panel
questions
the
reference
EFED
has
provided
by
Rygwelski,
estimating
a
half
life
in
Lake
Michigan
of
31
years.
It
presents
results
by
Schottler
and
Eisenreich
(
1997),
which
show
the
degradation
half­
life
of
atrazine
in
Lake
Michigan
as
14
years.
The
work
by
Rygwelski
has
been
published
in
J.
Great
Lakes
Res.
25(
1):
94­
106,
Inetrnat.
Assoc.
Great
Lakes
Res.,
1999,
entitled
"
A
Screening­
Level
Model
Evaluation
of
Atrazine
in
the
Lake
Michigan
Basin."
In
this
article,
the
author
also
discusses
the
discrepancy
between
his
results
and
the
work
of
Schottler
and
Eisenreich.
Even
with
a
half­
life
of
14
years,
atrazine
shows
considerable
persistence
in
Lake
Michigan.

P.
6
Deposition
Inputs­­
The
panel
questions
the
importance
of
atrazine
deposition
inputs
and
questions
EFED's
estimates.

Whether
it
is
25%
or
30%
of
the
total
load,
the
numbers
show
the
relative
importance
of
the
deposition
inputs
to
Lake
Michigan.
Even
when
all
surface
runoffs
of
atrazine
would
have
been
eliminated
from
Lake
Michigan,
about
25%
or
more
of
the
current
atrazine
total
load
is
still
calculated
to
have
been
input
to
Lake
Michigan
through
deposition.

Response
to
Comments
[
OPP
02­
0026­
0198]
by
the
Center
for
Regulatory
Effectiveness
(
CRE)
on
the
Atrazine
Environmental
Fate
and
Effects
Risk
Assessment
(
Docket
Control
Number
OPP­
34237C),
November
4,
2002,
and
the
Request
for
Correction
of
Information
Contained
in
the
Atrazine
Environmental
Risk
Assessment
by
the
CRE,
Kansas
Corn
Growers
Association,
and
the
Triazine
Netwark,
November
25,
2002
[
OPP
02­
0026­____].

CRE
Comment
The
CRE
commented
that
the
science
chapter
should
be
revised
to
delete
any
conclusion
that
atrazine
causes
indirect
effects
on
wildlife.
They
based
this
on
excerpts
from
the
EFED
Review
of
Public
Comments
in
Response
to
the
EFED
Revised
Environmental
Risk
Assessment
for
Atrazine,
page
3,
dated
April
10,
2002,
and
from
the
Registration
Eligibility
Science
Chapter
for
18
Atrazine,
Environmental
Fate
and
Effects
Chapter,
pages
63
and
64,
dated
April
22,
2002.
In
these
documents
EPA
admitted
that
there
were
no
methods
or
models
available
which
can
be
used
to
statistically
analyze
indirect
field
effects
data.
Additionally,
EPA
had
identified
uncertainties
in
the
data
used
in
the
assessment,
especially
in
the
field
effects
data
and
the
monitoring
data.

EPA
Response
The
sections
extracted
from
the
risk
assessment
chapter
and
the
responses
to
comments
were
written
to
establish
that
the
major,
but
by
no
means
the
only,
endpoint
of
concern
for
the
current
use
of
atrazine,
was
the
potential
indirect
adverse
effects
on
aquatic
populations
and
communities.
Further,
they
intended
to
explain
that
field
data
showing
these
potential
effects,
as
well
as
the
monitoring
data
that
established
widespread
potential
exposure,
could
not
be
used
in
a
probabilistic
analysis
in
a
manner
similar
to
the
existing
aquatic
laboratory
toxicity
data
and
the
exposure
modeling
data.
Risk
assessments
are
typically
performed
with
data
limitations
and
under
uncertainty.
Neither
prevent
risk
managers
from
arriving
at
risk
conclusions;
rather,
by
clearly
identifying
the
data
limitations
and
the
uncertainties,
and
describing
the
risk
conclusions
as
potential,
EPA
is
being
objective
and
transparent
in
its
assessment.
Finally,
in
spite
of
the
data
limitations
and
uncertainties,
EPA
contends
that
"
the
robust
body
of
surface
water
monitoring
data,
combined
with
extensive
effects
data
for
aquatic
organisms,
enabled
EFED
to
provide
quantitative
conclusions
on
the
frequency
and
extent
of
adverse
effects
of
atrazine
in
a
refined
aquatic
risk
assessment.
The
extensive
databases
as
well
as
the
refined
assessment
increase
the
certainty
of
the
conclusions
beyond
preliminary
risk
assessments
that
are
typical
for
all
other
herbicides."
(
Registration
Eligibility
Science
Chapter
for
Atrazine,
Environmental
Fate
and
Effects
Chapter,
pages
65,
dated
April
22,
2002)

CRE
Comment
The
CRE
maintained
that
since
there
are
no
validated
test
methods
for
assessing
endocrine
effects,
the
atrazine
environmental
risk
assessment
should
state
that
there
is
no
reliable
evidence
that
atrazine
causes
endocrine
effects
in
the
environment.
According
to
the
CRE:

"
EPA's
final
Reregistration
Eligibility
Science
Chapter
for
Atrazine:
Environmental
Fate
and
Effects
Chapter
(
April
22,
2002)("
Environmental
Risk
Assessment")
pages
11,
90
­
94,
states
that
atrazine
causes
endocrine
effects
in
various
organisms
including
frogs.
In
that
there
are
no
validated
test
methods
for
assessing
any
such
effects,
these
pages
of
the
Environmental
Risk
Assessment
should
be
corrected
to
state
that
there
is
no
reliable
evidence
that
atrazine
causes
endocrine
effects
in
the
environment.

EPA
Response
Contrary
to
what
the
CRE
maintains,
the
risk
assessment
does
not
state
that
atrazine
"
causes
endocrine
effects."
Rather,
the
assessment
identifies
potential
effects
of
atrazine
on
endocrine­
mediated
systems
as
an
endpoint
that
warrants
additional
study.
To
assure
that
there
is
no
ambiguity
about
the
Agency's
position,
minor
editorial
changes
have
been
incorporated
into
the
revised
assessment.
The
revised
assessment
does
not
suggest
that
endocrine
disruption,
or
potential
effects
on
endocrine­
mediated
pathways,
be
regarded
as
19
a
legitimate
regulatory
endpoint
at
this
time.
Nor
does
the
Agency
have
evidence
to
state
that
there
is
no
reliable
evidence
that
atrazine
causes
endocrine
effects
in
the
environment.
In
response
to
the
CRE,
we
revised
the
chapter
which
clearly
states
that
based
on
the
existing
data
uncertainties,
the
chemical
should
be
subject
to
more
definitive
testing
once
the
appropriate
testing
protocols
have
been
established.

To
reduce
some
of
the
uncertainties
in
understanding
potential
atrazine
effects
on
amphibian
endocrinology
and
reproductive
and
developmental
responses,
pertinent
studies
are
being
performed
by
external
parties.
In
accordance
with
the
agreement
reached
with
the
Natural
Resources
Defense
Council,
these
studies
in
progress
along
with
the
studies
in
question
will
be
summarized
and
analyzed
for
an
external
scientific
review
by
the
Federal
Insecticide,
Fungicide
and
Rodenticide
Act
(
FIFRA)
Science
Advisory
Panel
(
SAP)
at
a
public
meeting
which
is
scheduled
for
June,
2003.
The
Agency
anticipates
that
the
results
from
this
SAP
meeting
will
provide
significant
input
for
us
to
amend
the
analysis
of
the
potential
effects
of
atrazine
on
amphibian
endocrinology
and
development
in
the
interim
reregistration
eligibility
decision
which
is
scheduled
for
October
2003.

CRE
Comment
The
CRE
also
states
that
the
"
Data
Quality
Act"
requires
proper
test
validation
before
it
is
used
to
generate
information
to
support
regulatory
decisions.

.
.
.
for
Influential
Scientific
Information,
such
as
the
Environmental
Risk
Assessment,
EPA's
Data
Quality
Guidelines
require
that
EPA
"
ensure
reproducibility
for
disseminated
original
and
supporting
data
according
to
commonly
accepted
scientific,
financial,
or
statistical
methods."
EPA
cannot
ensure
reproducibility
of
original
and
supporting
data
from
atrazine
endocrine­
effects
tests
until
and
unless
that
data
is
generated
by
reliable
tests,
and
validation
is
necessary
to
ensure
reliable
tests.

EPA's
Environmental
Risk
Assessment
statements
regarding
atrazine's
endocrine
effects
also
violate
the
Data
Quality
Act's
Utility
Standard.
This
Standard
requires
that
information
disseminated
by
EPA
be
useful
to
its
intended
users,
including
the
public.
Information
from
new
or
revised
tests
is
not
useful
when
it
is
generated
from
unvalidated,
unreliable,
unreproducible
tests."

EPA
Response:
EPA's
use
of
data,
including
the
use
of
studies
that
have
not
been
formally
validated,
is
consistent
with
EPA
and
OMB
Information
Quality
Guidelines
and
applicable
statutory
and
regulatory
requirements.
Rather
than
being
requirements,
EPA's
Information
Quality
Guidelines
contain
EPA's
policy
regarding
information
the
Agency
disseminates
to
the
public.
EPA
implements
these
Guidelines
in
order
to
achieve
the
purposes
of
the
information
quality
law,
section
515,
P.
L.
106­
554.
EPA's
use
of
data
for
pesticide
registration
is
governed
by
40
CFR
Part
158,
which
references
EPA's
Pesticide
Assessment
Guidelines
containing
suggested
protocols
for
developing
data
(
§
158.70).
It
is
important
to
note
that
the
availability
of
a
final
guideline
does
not
in
any
way
affect
the
Agency's
20
authority
to
collect
the
data.
The
use
of
non­
guideline
studies
is
consistent
with
the
process
outlined
in
the
40
CFR
Part
158.
Part
158
states
that
"
data
routinely
required
by
Part
158
may
not
be
sufficient
to
permit
EPA
to
evaluate
every
pesticide
product.
If
the
information
required
under
this
part
is
not
sufficient
to
evaluate
the
potential
of
the
product
to
cause
unreasonable
adverse
effects
on
man
or
the
environment,
additional
data
requirements
will
be
imposed"
(
§
158.75).
In
cases
where
a
final
guideline
is
not
available,
the
Agency
or
a
registrant
may
search
for
available
scientific
literature.

In
determining
whether
data
are
acceptable,
the
Agency
considers
a
number
of
factors,
(
e.
g.,
study
design
including
sample
size,
replication,
use
of
appropriate
controls,
etc.)
including
the
use
of
GLPs,
to
determine
whether
scientifically
sound
methods
were
employed.
"
If
data
are
submitted
prior
to
the
development
of
appropriate
protocols,
the
Agency
will
consider
the
data
provided
they
permit
sound
scientific
judgements
to
be
made."
The
40
CFR
specifically
states
that
such
"
data
will
not
be
rejected
merely
because
they
were
not
developed
in
accordance
with
suggested
protocols"
(
§
158.80).
Additionally
the
CFR
states
that
the
Agency
can
implement
changes
in
the
data
requirements
on
a
case­
by­
case
basis
(
§
158.85).
Whether
a
test
has
been
validated
may
be
taken
into
account
when
the
Agency
considers
what
weight
the
test
should
be
given
in
a
weight
of
the
evidence
approach.
For
atrazine,
EPA
considered
what
weight
should
be
given
to
studies
that
have
not
been
formally
validated,
and
revised
the
assessment
to
clarify
the
discussion
of
the
existing
data
uncertainties,
as
indicated
in
the
response
to
the
previous
comment.
Thus,
the
analysis
is
consistent
with
the
weight
of
the
evidence
approach
and
applicable
regulations.

The
use
of
open
literature
in
the
atrazine
assessment
to
identify
significant
knowledge
gaps
is
consistent
with
EPA
policy
to
establish
data
quality
objectives
that
must
be
addressed
to
help
reduce
uncertainties
regarding
the
potential
effects
of
any
chemical
under
review.
While
the
CRE
is
correct
in
pointing
out
that
specific
guideline
tests
for
studying
the
endocrine
disrupting
potential
of
chemicals
have
not
been
formally
adopted
by
the
Agency,
the
absence
of
such
tests
does
not
preclude
EPA
from
using
information
available
from
peer­
reviewed
open
literature
to
identify
uncertainties
and
areas
for
additional
research.

The
information
generated
from
unvalidated
tests
is
useful
to
intended
users,
including
EPA
and
the
public.
When
a
review
of
the
peer­
reviewed
literature
chemicals
demonstrates
the
potential
of
a
chemical
to
cause
a
broad
range
of
effects,
not
detected
in
current
guideline
studies,
it
is
widely
accepted
in
the
scientific
community
and
the
Agency's
Ecological
Risk
Assessment
Guidelines
(
EPA,
1998)
to
use
this
literature
as
a
means
to
identify
uncertainties.
Consistent
with
the
process
defined
in
the
40
CFR,
the
Agency
can
direct
that
additional
testing
be
conducted
to
examine
non­
guideline
endpoints
and
uncertainties
in
greater
detail.
Consequently,
EPA
used
the
open
literature
to
identify
uncertainties
and
additional
research
that
may
need
to
be
conducted
to
examine
the
endocrine
disrupting
potential
of
atrazine.
This
approach
is
consistent
with
FIFRA
and
with
EPA
guidance
on
data
quality
objectives.
In
identifying
an
uncertainty,
OPP
can
then
work
with
risk
managers
to
better
define
data
quality
objectives
that
will
have
to
be
addressed
through
more
detailed
research.
21
CRE
also
raised
the
issue
of
reproducibility
of
the
studies
in
question.
This
is
one
of
several
science
issues
that
the
Agency
has
identified
and
is
asking
the
SAP
to
address.
EPA
has
not
determined
that
the
studies
are
influential
information.
Until
we
receive
input
from
an
independent
scientific
panel
at
the
SAP
meeting,
we
are
not
prepared
to
address
CRE's
concerns
regarding
reproducibility
of
the
studies.
The
Agency
will
address
issues
regarding
the
transparency
of
information
in
the
amended
registration
eligibility
document
scheduled
to
be
completed
after
the
SAP
meeting
(
October,
2003).

CRE
Comment
The
CRE
commented
that
the
quotient
method
EPA
used
in
the
Registration
Eligibility
Science
Chapter
for
Atrazine,
Environmental
Fate
and
Effects
Chapter
is
arbitrary
and
lacks
utility,
and
EPA
should
defer
any
conclusions
about
the
environmental
effects
of
atrazine
until
EPA
has
developed
a
reliable
probabilistic
risk
assessment
method
for
adverse
environmental
effects.
Further,
CRE
states
that
the
quotient
method
is
not
accurate,
reliable
or
useful
in
assessing
the
indirect
or
other
environmental
effects
of
atrazine.

EPA
Response
According
to
the
EPA
Guidelines
for
Ecological
Risk
Assessment
(
1998;
pages
92,
95­
97),
the
first
step
in
characterizing
ecological
risk
is
estimating
the
risk.
This
is
the
process
of
integrating
exposure
and
effects
data
and
evaluating
any
associated
uncertainties.
Risk
estimates
can
be
developed
using
a
number
of
different
techniques
including
comparisons
of
single­
point
exposure
and
effects
estimates.
The
simplest
approach
for
comparing
the
estimates
is
a
ratio
(
or
quotient)
where
an
exposure
concentration
is
divided
by
an
effects
concentration.
The
guidelines
state
that
"
the
principal
advantages
of
the
quotient
method
are
that
it
is
simple
and
quick
to
use
and
risk
assessors
and
managers
are
familiar
with
its
application.
It
provides
an
efficient,
inexpensive
means
of
identifying
high­
or
low­
risk
situations
that
can
allow
risk
management
decisions
to
be
made
without
the
need
for
further
information."
The
guidelines
also
recognize
that
"
a
number
of
limitations
restrict
application
of
the
quotient
method....
While
a
quotient
can
be
useful
in
answering
whether
risks
are
high
or
low,
it
may
not
be
helpful
to
a
risk
manager
who
needs
to
make
a
decision
requiring
an
incremental
quantification
of
risks...
In
addition,
the
quotient
method
may
not
be
the
most
appropriate
method
for
predicting
secondary
effects
(
although
such
effects
may
be
inferred)...
Finally,
in
most
cases,
the
quotient
method
does
not
explicitly
consider
uncertainty....
Some
uncertainties,
however,
can
be
incorporated
into
single­
point
estimates
to
provide
a
statement
of
likelihood
that
the
effects
point
estimate
exceeds
the
exposure
point
estimate."
Thus,
from
the
standpoint
of
published
and
peer
reviewed
EPA
guidance
for
ecological
risk
assessments,
the
use
of
the
quotient
method
in
the
EFED
science
chapter
for
atrazine
is
neither
arbitrary
nor
does
it
lack
utility.
Consistent
with
this
guidance,
EPA
used
the
risk
quotient
method
to
characterize
atrazine
as
being
problematic
and
then
proceeded
to
utilize
probabilistic
methods
to
further
characterize
the
risks.
EPA
compared
the
estimates
of
effects
to
distributions
of
monitoring
data,
a
probabilistic
approach
which
addresses
some
of
the
uncertainties
inherent
in
the
data.
22
American
Water
Works
Association
(
July
3,
2002)
[
OPP
02­
0026­
217]

Estimate
of
Water
Systems
"
At
Risk"

AWWA
commented
that
EPA
has
likely
underestimated
the
number
of
water
systems
"
at
risk".
EFED
agrees
with
this
comment
and
would
encourage
more
frequent
monitoring
be
conducted
at
more
systems.

Significant
data
gaps
still
remain
for
analytical
methodologies,
treatment
data
and
occurrence
data
Specifically
AWWA
has
concerns
with
respect
to
metabolites
related
to
the
above
issues.
Among
these
issues,
EFED
is
most
interested
in
the
limitation
due
to
lack
of
occurrence
data.
AWWA's
analysis
shows
that
the
sampling
frequency
can
strongly
affect
the
bias
in
90­
day
average
concentrations,
based
on
a
comparison
of
two
log­
normal
distributions
using
daily,
weekly
and
biweekly
sampling
frequencies.
AWWA
also
questioned
OPP's
exposure
assessment
which
was
based
on
weekly
or
biweekly
monitoring
data.
AWWA
is
considering
conducting
an
occurrence
study
of
both
the
metabolites
and
parent
compounds
in
calendar
year
2002
and
invites
EPA's
input
into
the
design
of
this
study.
EFED
applauds
AWWA
for
this
effort
and
is
willing
to
provide
technical
assistance
in
the
design
of
this
study
AWWA
suggested
a
number
of
mitigation
options
OPP
appreciates
the
suggested
mitigation
options
presented
by
AWWA.
The
seven
items
suggested
will
be
considered
for
the
overall
mitigation
measures:
further
label
restrictions
(
application
rate
and
residential
use);
further
restriction
on
use;
the
implementation
of
additional
best
management
practices
(
buffer
strips,
containment
ponds,
etc.);
a
memorandum
of
understanding
with
the
registrant
to
pay
for
the
additional
testing
necessary
to
characterize
the
intermediate
exposure;
the
enactment
of
a
"
water
quality
tax"
(
in
the
range
$
0.01­$
0.25
per
ton)
to
pay
for
additional
testing
and
treatment;
a
partial
ban
for
impacted
watersheds;
and,
a
complete
ban
on
the
use
of
the
chemical.

New
York
State
Office
of
Attorney
General
(
Docket
Control
Number
OPP­
34237C)
July
5,
2002
[
OPP­
02­
0026­
160]

The
NYOAG
commented
to
the
Agency
that
EPA
must
initiate
consultations
with
the
FWS
because
EPA's
issuance
of
a
reregistration
decision
for
atrazine
triggers
the
ESA
consultation
requirement
and
stated
that
the
ESA
requires
that
the
Agency
consider
any
existing
FWS
biological
opinion.

Atrazine
has
been
reviewed
on
several
occasions
by
the
FWS
as
described
in
Section
III
above
under
the
discussion
on
endangered
species.
Currently,
the
the
Agency
is
developing
a
proposal
to
implement
its
Endangered
Species
Protection
Plan
(
ESPP).
The
Agency
is
23
soliciting
public
opinion
on
this
proposal
through
issuance
of
a
Federal
Register
Notice,
Endangered
Species
Protection
Program
Field
Implementation,
December
2,
2002.
The
Agency
obtained
input
on
several
key
aspects
of
the
program
in
a
workshop
held
in
September
2002
that
included
the
pesticide
industry,
pesticide
user
groups,
and
environmental
advocacy
organizations.
An
Advance
Notice
of
Proposal
Rulemaking
(
ANPR),
Endangered
Species
and
Pesticide
Regulation,
was
issued
jointly
by
the
Agency,
the
Department
of
Interior
and
the
Department
of
Commerce
on
January
24,
2003.
The
ANPR
is
soliciting
comments
regarding
methods
to
make
the
consultation
process
more
efficient..

Contamination
of
New
York
Lakes,
Streams
and
Groundwater
Wells
The
State
AG's
Office
submitted
four
reports
of
studies
conducted
by
the
U.
S.
Geological
Survey
containing
monitoring
data
on
many
compounds
for
New
York
rivers,
streams,
lakes,
and
groundwater
wells.
The
documents
included
monitoring
data
from
areas
in
the
Hudson
River
Basin,
in
Hemlock
and
Canadice
Lakes
and
tributaries
in
Western
New
York,
pesticides
and
their
metabolites
in
well
water
in
Suffolk
County,
and
samples
from
three
small
public
water
systems
in
Western
New
York.

The
reports
show
the
prevalence
of
the
compound
at
levels
generally
found
in
the
various
types
of
water
sources
considered
the
EFED
assessment.
Although
the
State
is
correct
in
contending
that
OPP
did
not
assess
the
potential
health
effects
afforded
by
exposure
to
multiple
sources
of
contaminated
drinking
water,
the
EPA
reports
did
attempt
to
assess
exposure
to
and
potential
health
effects
of
the
total
chlorotriazines.

Atmospheric
Deposition
 
The
State
contends
that
the
risk
assessment
"
fails
to
consider
the
additional
exposure
pathway
of
atrazine
in
air
and
its
deposition
in
surface
water
and
drinking
water
supplies."

Although
EFED
qualitatively
considered
the
additional
contribution
to
exposure
of
atrazine
deposition
to
drinking
water
sources,
it
did
not
possess
the
substantial
amount
of
data
necessary
to
properly
assess
exposure
from
this
source
in
a
quantitative
manner.
We
agree
that
it
may
be
significant
in
some
areas,
and
presented
information
on
Lake
Michigan
in
the
chapter.

State
of
Connecticut,
Office
of
Attorney
General
RE:
Docket
Control
Number
OPP­
34237C
Atrazine
(
July
3,
2002)
[
OPP
02­
0026­
0186]

The
State's
Office
of
Attorney
General
stated
that
atrazine
is
ubiquitous
and
warrants
careful
and
thorough
study.
It
reported
that,
according
to
the
US
Geological
Study
between
1992
and
1995,
atrazine
was
the
most
commonly
detected
pesticide
in
the
Connecticut,
Housatonic
and
Thames
River
Basins.
In
addition,
it
was
detected
in
50%
of
the
samples
taken
from
the
Norwalk
River
at
Winnipauk,
Connecticut.
The
office
also
stated
that
the
revised
risk
assessment
does
not
adequately
consider
the
endocrine
effects
of
atrazine,
it's
24
widespread
presence
in
ground
water,
drinking
water
and
the
atmosphere,
and
endangered
species
concerns.
The
office
concludes
that
atrazine
should
be
immediately
cancelled.

EFED
appreciates
the
comments
and
information
that
indicate
that
even
outside
the
Mid­
West,
atrazine
can
be
commonly
detected
in
water
bodies.
This
shows
that
there
is
widespread
environmental
exposure
due
to
the
use
of
atrazine,
even
in
water
bodies
in
Connecticut.
EFED
also
notes
that
the
frequency
of
occurrence
and
extent
of
the
potential
impacts
will
vary
depending
upon
the
type
of
water
bodies
and
their
proximity
in
time
and
space
to
atrazine
applications.
Thus,
additional
information
would
be
needed
to
address
the
potential
risk
to
aquatic
organisms
from
this
exposure.

U.
S.
Department
of
Interior,
Fish
&
Wildlife
Service
(
EPA
Docket
Control
Number
OPP­
34237C)
[
OPP
02­
0026­____]

EFED
has
reviewed
comments
(
Document:
FWS/
AFHC/
DEQ)
submitted
by
the
U.
S.
Fish
and
Wildlife
Service
(
FWS)
regarding
the
recently
completed
environmental
risk
assessment
conducted
to
support
the
reregistration
review
for
atrazine.
In
their
letter,
FWS
expressed
concern
that
because
atrazine
is
located
in
all
environmental
compartments,
e.
g.,
soil,
water,
fog,
rain,
it
is
likely
to
affect
a
broad
range
of
organisms.
FWS
believes
that
EFED's
ecological
risk
assessment
does
not
adequately
characterize
either
direct
or
indirect
impacts
to
endangered
species
since
it
relies
on
toxicity
data
collected
from
a
limited
number
of
surrogate
species,
none
of
which
are
listed.
Additionally,
the
FWS
believes
the
risk
assessment
failed
to
account
for
sublethal
effects
and
the
aggregate
effects
of
atrazine
plus
each
of
the
chemicals
(
inerts)
co­
formulated
with
the
active
ingredient.
FWS
is
particularly
concerned
about
the
potential
impacts
of
atrazine
on
amphibians
given
recent
studies
showing
atrazine
effects
on
African
clawed
frogs
(
Xenopus
laevis)
and
Northern
leopard
frogs
(
Rana
pipiens).
Based
on
their
concerns,
FWS
provided
several
recommendations
for
mitigating
risks
associated
with
the
use
of
atrazine.

EFED
appreciates
FWS'
comments
regarding
the
environmental
fate
and
ecological
effects
assessment
in
support
of
the
reregistration
eligibility
decision
on
atrazine.
EFED
acknowledges
that
there
are
uncertainties
associated
with
any
risk
assessment
which
limit
the
comprehensiveness
of
the
assessments.
In
writing
its
assessment
of
potential
ecological
effects
of
a
pesticide,
EFED
has
to
rely
on
surrogate
species
toxicity
test
data.
Generally
these
data
are
restricted
to
a
relatively
select
group
of
animals
that
survive
well
under
laboratory
conditions
and
may
not
be
representative
of
the
most
sensitive
organisms.
Invariably,
these
data
do
not
include
toxicity
tests
run
on
threatened
and/
or
endangered
species.
FWS
is
encouraged
to
provide
any
data
it
may
have
to
support
its
concerns
that
EFED's
assessment
is
not
likely
representative
of
risks
posed
to
threatened
and/
or
endangered
species.

Risk
assessments
are
typically
performed
with
data
limitations
and
under
uncertainty.
Neither
prevents
risk
assessors
from
arriving
at
risk
conclusions;
rather,
by
clearly
25
identifying
these
data
limitations
and
uncertainties,
and
describing
the
risk
conclusions
as
potential,
EPA
is
being
objective
and
transparent
in
its
assessment.
Finally,
in
spite
of
the
data
limitations
and
uncertainties,
EPA
contends
that
"
the
robust
body
of
surface
water
monitoring
data,
combined
with
extensive
effects
data
for
aquatic
organisms,
enabled
EFED
to
provide
quantitative
conclusions
on
the
frequency
and
extent
of
adverse
effects
of
atrazine
in
a
refined
aquatic
risk
assessment.
The
extensive
databases
as
well
as
the
refined
assessment
increase
the
certainty
of
the
conclusions
beyond
preliminary
risk
assessments
that
are
typical
for
all
other
herbicides."
(
Registration
Eligibility
Science
Chapter
for
Atrazine,
Environmental
Fate
and
Effects
Chapter,
pages
65,
dated
April
22,
2002)

At
present,
EFED
does
not
conduct
aggregate
risk
assessments
of
active
ingredients
and
all
of
the
potential
"
inert"
chemicals
associated
with
the
formulated
products.
In
many
cases,
formulated
product
testing
is
required
if
EFED
has
reason
to
believe
that
the
formulated
product
will
represent
a
substantially
increased
risk
to
nontarget
organisms.
Similar
to
risk
assessments
conducted
on
technical
grade
pesticides,
the
Agency
is
in
the
process
of
evaluating
risks
associated
with
individual
inert
ingredients.
Again,
if
the
FWS
has
data
to
support
its
concern
that
specific
inert
ingredients
in
combination
with
atrazine
represent
an
increased
risk
to
nontarget
organisms,
it
is
encouraged
to
submit
these
data
for
review.

EFED
is
in
the
process
of
evaluating
data
on
the
potential
endocrine
disrupting
effects
of
atrazine
on
a
range
of
non­
target
animals.
These
data
focus
primarily
on
the
effects
of
atrazine
on
endocrine­
mediated
gonadal
development
in
amphibians,
fish
and
aquatic
organisms.
The
studies
under
evaluation
have
been
submitted
by
pesticide
registrants
and
taken
from
open
literature.
They
include
both
published
and
unpublished
studies.
The
reported
results
vary
widely,
with
effects
on
gonadal
development
reported
at
exposures
from
0.1
ppb)
to
25
ppb.
Where
possible,
raw
data
from
these
studies
are
being
analyzed
and
study
methodologies
are
being
documented
to
evaluate
whether
the
data
support
the
conclusions
reached
by
the
study's
author.

Additional
pertinent
studies
are
being
performed
both
by
the
pesticide
registrants
and
by
independent
researchers,
which
could
shed
additional
light
on
this
sub­
lethal
endpoint
for
atrazine.
In
accord
with
the
agreement
reached
with
the
Natural
Resources
Defense
Council,
EPA
is
planning
to
summarize
all
these
studies
in
preparation
for
a
FIFRA
SAP
meeting
scheduled
for
June
2003.
To
facilitate
the
preparation
of
this
summary
document,
OPP
will
continue
to
accept
studies
on
this
endpoint
up
to
February
28,
2003.
These
studies
can
only
be
accepted
in
final
form
and
must
represent
"
publication
quality";
no
preliminary
data
can
be
accepted
for
this
summary
document.
Data
submitted
prior
to
the
February
cutoff
date
will
be
included
in
the
summary
document
and
discussed
in
the
June
2003
SAP
meeting.
While
EPA
does
not
intend
to
exclude
any
pertinent
studies
from
consideration,
this
deadline
is
necessary
for
the
Agency
to
have
sufficient
time
to
review,
evaluate
and
summarize
the
available
studies
and
present
them
for
timely
science
review
by
the
FIFRA
SAP.
In
addition,
the
results
from
this
SAP
meeting
are
considered
crucial
input
for
the
amended
reregistration
eligibility
decision
on
atrazine
scheduled
for
October
2003.
Therefore,
FWS
is
encouraged
to
provide
data
on
atrazine's
effects
on
amphibians
for
review
by
February
28,
2003.
26
EFED
is
encouraged
by
FWS'
willingness
to
participate
in
the
process
of
developing
a
comprehensive
and
meaningful
risk
assessment
for
pesticides
under
review
by
the
Agency
and
appreciates
the
Service's
willingness
to
provide
constructive
comments
and
propose
mitigation
options.
While
it
is
clear
that
the
Agency
will
likely
be
engaged
in
a
consultation
on
atrazine
under
Section
7
of
the
Endangered
Species
Act,
informal
discussions
between
the
two
agencies
can
only
serve
to
improve
EFED's
understanding
of
FWS'
concerns
and
perhaps
provide
a
conduit
for
exchanging
relevant
data
on
chemicals
under
review.

California
Regional
Water
Quality
Control
Board
(
June
19,
2002)
[
OPP
02­
0026­
0223]

The
Board
has
concerns
about
the
discrepancies
between
OPP
and
OW
regarding
the
effect
levels
and
exposure
methodologies.

While
OPP
found
effects
at
a
range
of
10­
20
ug/
L,
Office
of
Water's
(
OW)
has
proposed
an
effect
criterion
at
essentially
the
same
level,
11­
12ug/
L.
However,
EPA
does
not
regulate
pesticides
under
the
Clean
Water
Act
but
rather
under
the
FIFRA.
Although
methodologies
to
derive
aquatic
life
criteria
in
OW
differ
from
the
methods
used
by
OPP
to
derive
risk
quotients
and
compare
them
against
established
levels
of
concern,
both
offices
reviewed
all
of
the
available
data
with
the
result
being
that
the
proposed
OW
standard
for
atrazine
approximates
the
level
proposed
by
OW.
OPP
and
OW
are
currently
exploring
methods
to
better
reconcile
each
office's
approach.

Endangered
Species
Concerns
(
esp.
Salmonids
and
sublethal
effects
on
amphibians)
EPA
has
responded
earlier
in
this
document
to
comments
regarding
endangered
species
and
sublethal
effects.
The
reader
is
encouraged
to
review
the
relevant
preceding
sections
in
this
document.

Urban
uses
The
Board
pointed
out
that
OPP
only
addressed
agricultural
uses
and
not
urban
use.
Since
most
of
the
atrazine
load
has
been
associated
with
agricultural
uses,
EFED
focused
on
this
aspect.
The
NAWQA
monitoring
data
EFED
considered,
however,
also
included
samples
taken
at
some
urban
indicator
sites.
The
detections
of
atrazine
at
these
sites
were
not
as
high
as
some
of
the
detections
in
agricultural
indicator
sites.
If
the
Board
has
any
additional
atrazine
monitoring
data
emphasizing
the
contribution
of
atrazine
from
urban
uses,
EFED
would
like
to
review
them.

American
Farm
Bureau
Federation
(
AFBF)
(
July
3,
2002)
[
OPP
02­
0026­____]

Exposure
through
Drift
 
While
the
AFBF
claims
that
atrazine
is
not
subject
to
volatile
drift,
EFED's
concern
is
the
off­
target
spray
drift
during
the
application.
The
Spray
Drift
Task
Force's
Agdrift
model
output
shows
the
likely
occurrences
of
off­
target
spray
drift
for
both
ground
and
aerial
spray
applications.
27
Persistence
 
AFBF
contends
that
atrazine
"
breaks
down
in
the
soil
in
a
matter
of
weeks
or
a
few
months."
More
detailed
environmental
fate
properties
of
atrazine
have
been
presented
in
the
RED
and
in
the
technical
briefing.
Atrazine
half­
lives
range
from
20
to
146
days
in
the
aerobic
soil
metabolism
studies.
In
aquatic
environments,
however,
depending
on
conditions
in
aquatic
bodies
(
lakes,
mesocosm,
and
experimental
pond),
the
half­
lives
from
six
studies
vary
from
41
to
237
days.
In
addition,
a
large
Agency
project,
the
"
Lake
Michigan
Lake­
Wide
Management
Plan"
estimated
a
half­
life
of
31
years,
due
to
the
cold
temperature,
low
productivity,
high
pH,
low
nitrate
and
low
dissolved
organic
carbon
of
the
lake.

Effect
level
 
AFBF
cites
effect
levels
of
20
and
50
ppb
for
long
and
short
term
exposure,
respectively,
from
a
study
performed
by
Solomon
and
Giddings
(
Ecological
Risk
Assessment
Panel).
OPP's
assessment
found
effects
in
the
field
at
environmental
levels
from
10­
20
ppb.
OPP's
response
to
the
above
study
cited
by
AFBF
is
included
in
this
document
and
in
detailed
responses
in
Urban
et
al
2002
and
Frankenberry
et
al
2002.

Louisiana
Farm
Bureau
Federation
(
July
3,
2002)
[
OPP
02­
0026­
0156]

EFED
has
based
its
assessment
on
the
PRZM/
EXAMS
modeling
results
and
on
the
Louisiana
Department
of
Environmental
Quality's
monitoring
data
of
1998
for
the
risk
assessment
report.
Since
there
are
new
data
from
Lousiana
Department
of
Agriculture
and
Forestry,
the
Farm
Bureau
Federation
is
welcome
to
submit
these
new
data
for
EFED
to
consider.
