1
UNITED
STATES
ENVIRONMENTAL
PROTECTION
AGENCY
WASHINGTON,
D.
C.
20460
Date:
September
15,
2005
Chemical:
Thidiazuron
PC
Code:
120301
DP
Barcode:
D320941
Subject:
Thidiazuron
B
Response
to
Bayer's
60­
day
Response
on
the
EFED
Science
Chapter
for
the
Reregistration
Eligibility
Decision
Document
To:
John
Pates,
Chemical
Review
Manager,
Reregistration
Branch
II
Special
Review
and
Reregistration
Division
(
SRRD)
(
7508C)

From:
William
Evans,
Biologist,
Environmental
Risk
Branch
I
James
Hetrick,
Ph.
D.,
Environmental
Scientist,
Environmental
Risk
Branch
I
Environmental
Fate
and
Effects
Division
(
EFED)
(
7507C)

Approved
By:
Kevin
Costello,
Acting
Branch
Chief,
Environmental
Risk
Branch
I
Environmental
Fate
and
Effects
Division
(
EFED)
(
7507C)

We
have
attached
our
responses
to
Bayer
CropScience's
June
11,
2005,
"
60­
day
Response
to
Environmental
Fate
and
Effects
Risk
Assessment
for
the
Reregistration
Eligibility
Decision"
comments
on
EFED's
April
5,
2004,
Science
Chapter
for
the
Reregistration
Eligibility
Decision
(
RED)
Document
for
thidiazuron.

Our
responses
follow
the
same
RED­
page
sequential
order
that
Bayer
used.
We
use
essentially
the
same
format.
Note
that
the
page
numbers
to
which
Bayer
refers
are
those
from
April
5,
2005
RED
document,
and
would
not
necessarily
correspond
to
those
in
the
revised
RED.

Please
inform
Bayer
that
EFED
appreciates
the
comments
expressed
and
has
corrected
and
revised
the
document
as
appropriate.
Although
some
risk
quotients
have
changed,
we
are
pleased
to
report
that
the
errors
Bayer
detected
essentially
do
not
affect
the
risk
characterization
and
conclusions
we
reached
previously.
2
Cover
Memorandum
for
EFED
Phase
2
Risk
Assessment
Use
of
Thidiazuron
Page:
2
Paragraph:
1
Line:
na
EPA
statement:

EPA
provides
an
overview
of
the
use
pattern
for
thidiazuron.

Bayer's
comment:

It
may
be
useful
to
include
in
the
use
pattern
summary,
the
average
application
rates
provided
by
BEAD.
As
the
EFED
risk
assessment
notes
on
page
23
the
average
application
rates
range
from
0.05
to
0.07
lb
ai/
A;
significantly
lower
than
the
maximum
allowed
rates.

EPA
response:
A
sentence
has
been
added
which
describes
the
average
application
rates.

Ecological
Risk
Summary
Page:
3
Paragraph:
1
Lines:
3­
5
EPA
statement:

A
fourth
reason
that
risk
may
be
underestimated
is
that
product
label
statements
about
citrus
and
cantaloupe,
for
which
we
have
no
test
data,
clearly
imply
that
these
or
other
plant
species
may
be
more
sensitive
than
species
for
which
we
have
test
data.

Bayer's
comment:

The
label
statements
regarding
citrus
and
cantaloupe
do
not
necessarily
indicate
that
these
species
are
more
sensitive
than
the
species
tested,
but
indicate
that
some
grower
concerns
were
raised
for
these
species.
The
buffer
language
that
was
developed
was
not
based
on
a
quantitative
analysis
for
buffer
distances.
It
is
highly
likely
that
a
smaller
buffer
would
be
equally
protective.

EPA
response:
EPA
will
remove
the
language
which
infers
that
citrus
and
cantaloupe
are
more
sensitive
than
surrogate
species
tested.
Instead,
more
generic
language
will
be
included
which
states
that
non­
target
plants
can
be
more
or
less
sensitive
than
those
tested
in
the
laboratory.

Page:
3
Paragraph:
2
Bullet
2
for
mammals
EPA
statement:

Mammals
are
at
potential
chronic
risk.

Bayer's
comment:

While
a
screening
level
assessment
may
indicate
the
potential
for
concern
for
mammals,
a
more
refined
assessment
demonstrates
that
this
risk
is
minimal.

°
The
study
on
which
the
screening
level
assessment
was
based
was
old
and
did
not
meet
all
guideline
requirements.
For
example,
food
consumption
was
not
assessed.

There
was
a
poor
dose­
response
relationship
in
the
endpoint
in
question.
None
of
the
3
current
studies
included
in
the
HED
review
(
rat
reproduction
study
meeting
guidelines,
rat
and
rabbit
developmental
toxicity
tests)
demonstrate
the
same
type
of
effect
observed
in
the
older
study.

°
The
screening
level
risk
assessment
using
the
results
from
the
newer,
more
reliable
reproduction
study,
as
well
as
further
refinements
in
the
assessment,
demonstrate
that
the
potential
risk
is
low.

°
All
RQ's
using
the
predicted
mean
residue
values
were
below
the
Level
of
Concern
(
LOC)
.
All
RQ's
using
the
dietary
exposure
estimates
are
below
the
LOC.
It
should
be
noted
this
was
also
true
for
the
screening
level
assessment
using
the
3­
generation
study.

°
Potential
food
sources
off­
site
of
the
target
application
would
have
significantly
less
exposure,
and
result
in
a
minimal
risk
for
mammals.

°
Mammals
are
unlikely
to
be
consuming
100%
of
their
diet
in
fully
treated
thidiazuron
treated
food
items.

°
Cotton
plants,
particularly
late
in
the
season
are
not
a
primary
food
source
for
wild
mammals.

°
According
to
Linders
et
al.
(
2002),
the
crop
interception
factor
for
cotton
during
stem
elongation/
crop
cover/
flowering
is
70%,
and
during
senescence
is
90%.
If
we
use
the
70%
as
a
conservative
estimate,
this
would
indicate
that
the
cotton
intercepts
at
least
70%
of
the
applied
thidiazuron.
Given
that
the
mammals
do
not
consume
the
cotton,

the
potential
exposure
on
other
food
items
within
the
field
will
be
significantly
lower
(
70%)
than
that
used
in
the
models
for
the
screening
level
risk
assessment.
If
we
reduce
the
exposure
estimates
by
70%,
none
of
the
scenarios
in
Table
1
would
exceed
the
chronic
LOC.

EPA
response:

The
risk
assessment
has
been
revised
to
include
the
2­
generation
rat
study
and
to
use
the
results
from
that
study
for
risk
quotient
calculation.
The
screening
assessment
does
not
assume
that
mammals
are
consuming
cotton
plants,
but
other
plant
feed
items
in
and
at
the
margins
of
treated
cotton
fields.

Page:
3
Paragraph:
2
Bullet
3
for
birds
EPA
statement:

Potential
chronic
risk
to
birds
cannot
be
determined
because
of
the
absence
of
chronic
data.

Bayer's
comment:

Chronic
or
reproductive
toxicity
tests
with
birds
are
not
necessary
as
there
is
minimal
to
no
exposure
during
the
breeding
season.
The
normal
use
pattern
is
for
one
application,
a
second
application
may
occur
but
within
a
short
time
frame.
The
application
occurs
in
the
autumn
4
(
when
birds
are
not
breeding)
on
leaves
which
then
rapidly
drop
and
degrade
on
the
soil.

Moreover,
cotton
leaves
are
not
an
attractive
food
source
for
birds
and
thidiazuron
is
not
expected
to
bioaccumulate
in
animal
tissues.
Therefore
the
chronic
risk
to
birds
is
low.

EPA
response:
See
EPA
response
below.

Page:
3
Paragraph:
2
Bullet
3
for
birds
EPA
statement:

Potential
chronic
risk
to
birds
cannot
be
determined
because
of
the
absence
of
chronic
data.

Bayer's
comment:

Chronic
or
reproductive
toxicity
tests
with
birds
are
not
necessary
as
there
is
minimal
to
no
exposure
during
the
breeding
season.
The
normal
use
pattern
is
for
one
application,
a
second
application
may
occur
but
within
a
short
time
frame.
The
application
occurs
in
the
autumn
(
when
birds
are
not
breeding)
on
leaves
which
then
rapidly
drop
and
degrade
on
the
soil.

Moreover,
cotton
leaves
are
not
an
attractive
food
source
for
birds
and
thidiazuron
is
not
expected
to
bioaccumulate
in
animal
tissues.
Therefore
the
chronic
risk
to
birds
is
low.

EPA
response:
In
the
absence
of
data,
EFED
assumes
that
chronic
effects
from
exposure
to
thiadiazuron
may
occur
whenever
birds
may
come
into
contact
with
the
chemical.
Reproductive
effects
could
occur
as
a
result
of
a
one
time
exposure
at
any
time
without
regard
to
the
breeding
season.
With
regard
to
cotton
leaves
not
being
an
attractive
food
source
for
birds,
hedgerows
near
cotton
fields
which
have
been
contaminated
with
thiadiazuron
may
provide
more
of
a
food
source
for
birds
foraging
in
or
near
cotton
fields.
Further,
chronic
avian
reproduction
studies
are
required
for
a
screening
level
assessment.

Data
Gaps
 
Ecological
Effects
­
Nontarget
Terrestrial
Plant
Page:
5
Paragraph:
3
Lines:
1­
4
EPA
statement:

Definitive
tier
II
EC25'
s
and
NOAECs
are
lacking
for
the
vegetative
vigor
of
all
monocot
tested
and
for
the
seedling
emergence
of
two
monocots
tested;
this
is
because
the
test
concentrations
were
not
low
enough
to
establish
these
endpoints.

Bayer's
comment:

This
statement
is
inaccurate.
For
the
vegetative
vigor
tests
a
NOAEC
was
obtained
for
all
monocots
species
tested,
it
was
equal
to
the
highest
test
concentration
(
see
EFED
risk
assessment,
page
104).
The
EC25
was
greater
than
this
concentration
(
0.1783
lbs
ai/
A).
5
EPA
response:
EFED
agrees
with
the
statement
regarding
the
vegetative
vigor
test.
However,
the
statement
remains
accurate
for
the
seedling
emergence
testing.
Tier
2
testing
of
monocots
is
not
required.
The
sentence
with
regard
to
vegetative
vigor
will
be
modified.

Data
Gaps
 
Ecological
Effects
­
Birds
Page:
6
Paragraph:
7
Lines:
1­
4
EPA
statement:

Chronic
risks
to
birds
can
not
be
quantitatively
evaluated
at
this
time
because
chronic
data
were
not
submitted.
In
order
to
assess
this
potential
risk,
we
recommend
chronic
testing
with
an
upland
game­
bird
species
(
bobwhite
quail)
and
a
waterfowl
(
mallard
duck).

Bayer's
comment:

Chronic
or
reproductive
toxicity
tests
with
birds
are
not
necessary
as
there
is
minimal
to
no
exposure
during
the
breeding
season.
The
normal
use
pattern
is
for
one
application,
a
second
application
may
occur
but
within
a
short
time
frame.
The
application
occurs
in
the
autumn
(
when
birds
are
not
breeding)
on
leaves
which
then
rapidly
drop
and
degrade
on
the
soil.

Moreover,
cotton
leaves
are
not
an
attractive
food
source
for
birds
and
thidiazuron
is
not
expected
to
bioaccumulate
in
animal
tissues.
Therefore
chronic
testing
risk
for
birds
is
low
and
additional
testing
would
bring
little
value
to
the
risk
assessment.

EPA
response:
Please
see
response
to
"
Page:
3
Paragraph:
2
Bullet
3
for
birds"
above.

PHASE
2:
EFED
RISK
ASSESSMENT
I.
Environmental
Risk
Conclusions
Terrestrial
Plant
Risks
of
Concern
Page:
4
Paragraph:
5
Lines:
4­
7
EPA
statement:

EPA
provides
a
summary
of
RQ's
for
the
terrestrial
plant
risk
assessment.

Bayer's
comment:

As
discussed
for
comments
on
Appendix
G
(
Tables
11­
15,
pages
121­
127),
the
aerial
drift
estimates
did
not
account
for
the
application
efficiency.
Therefore,
the
RQ's
should
be
adjusted
to
reflect
the
change
in
the
estimated
environmental
exposure
(
EEC).

EPA
response:
The
aerial
drift
estimates
were
based
on
simulations
using
the
AGDRIFT
model,
which
was
developed
using
Spray
Drift
Task
Force
data.
These
data
measured
spray
drift
at
varying
distances
from
the
treated
field
directly.
The
application
efficiency
to
which
Bayer
refers
is
the
efficiency
of
aerial
application
to
the
treated
field
assumed
in
the
TERR­
PLANT
model.
That
efficiency
factor
does
not
pertain
to
spray
drift.
6
Page:
5
Paragraph:
5
Lines:
1­
3
EPA
statement:

To
explore
the
apparent
risk
posed
by
spraydrift,
we
used
EFED's
current
AgDrift
computer
model
and
thidiazuron
phytotoxicity
data
to
estimate
plant
exposure
as
a
function
of
downwind
distances
from
application
areas.

Bayer's
comment:

EPA
generated
a
series
of
bar
graphs
showing
a
percent
effect
vs.
a
no­
spray
zone.
These
graphs
are
not
appropriately
utilized
and
it
is
not
clear
what
the
graphs
are
depicting.
There
is
no
description
as
to
how
these
bar
graphs
were
generated.
Also,
each
graph
should
have
a
footnote
itemizing
which
effect
was
used
to
generate
for
the
regression
analysis.
Not
all
effects
to
nontarget
plants
are
equal
in
severity,
and
therefore
it
would
be
useful
for
risk
management
decisions,
to
have
a
better
understanding
of
the
graphs.

It
is
common
in
nontarget
plant
tests
to
have
a
weak
dose
response
relationship
between
exposure
and
effect.
Indication
that
the
dose­
response
relationship
is
not
optimized
for
the
full
range
of
the
ECx
values
being
attempted
to
be
generated
for
the
graphs.
For
example,

the
lettuce
vegetative
vigor
response
(%
inhibition
relative
to
control)
for
shoot
weight
was
12,
27,
12,
22,
and
40%
for
lowest
to
highest
doses
in
the
study.
With
such
a
poor
doseresponse
relationship,
the
confidence
intervals
around
the
ECx
values
that
were
generated
would
be
very
large.
However,
no
confidence
intervals
were
provided
to
assist
in
the
interpretation
of
the
bar
graphs.
For
the
soybeans
and
tomato
vegetative
vigor
studies,
no
effect
greater
than
25%
was
observed
on
shoot
length
or
weight,
and
yet
a
regression
to
generate
EC10
to
EC90
values
was
attempted.
It
is
clear
that
these
regressions
were
extrapolating
well
beyond
the
range
of
the
concentrations
tested,
and
there
would
be
high
uncertainty
(
high
variability)
in
the
results
obtained.
As
demonstrated
by
these
examples,
the
use
of
the
bar
graphs
should
be
limited
and
should
be
interpreted
with
extreme
caution.

The
application
scenario
for
the
nontarget
plant
studies
is
to
spray
a
high
volume
of
a
dilute
solution,
mimicking
a
direct
spray
scenario.
However,
in
a
spray
drift
scenario
(
which
is
relevant
for
nontarget
plants)
the
droplet
spectrum
consists
of
a
few
drops
of
a
more
concentrated
solution.
The
potential
impact
on
a
plant
may
be
reduced
under
this
different
exposure
scenario,
which
should
also
be
considered
in
assessing
the
potential
risk
The
endangered
species
terrestrial
plant
concern
occurs
when
the
environmental
concentration
is
greater
than
the
NOEC.
The
EC05
can
be
utilized
in
cases
where
an
NOEC
was
not
achieved
in
the
nontarget
plant
tests.
Since
for
thidiazuron,
a
NOEC
was
achieved
for
the
most
sensitive
species,
and
therefore
the
NOEC
was
(
and
should
be)
utilized
for
the
risk
assessment.
As
stated
above,
the
bar
graphs
utilizing
regression
analysis
outside
the
boundaries
of
the
concentrations
tested
(
i.
e.,
below
the
NOEC)
may
be
inappropriate
for
risk
management
decisions.
7
EFED
response:

The
RED
chapter
informs
the
reader
that
a
lognormal
toxicity
distribution
is
assumed,
and
the
calculation
by
which
different
effect
levels
on
the
distribution
are
estimated
is
provided.
However,
the
document
has
been
revised
to
acknowledge
that
there
is
uncertainty
in
this
calculation
of
dose
response
curves.
The
uncertainty
in
the
vegetative
vigor
dose­
response
curves
for
soybean
and
tomato
does
not
lead
to
a
greater
finding
of
risk;
the
bar
graphs
do
not
indicate
deposition
at
loadings
greater
than
that
equivalent
to
the
EC25
at
any
distance
from
the
treated
field.

II.
Introduction
Use
and
Use
Characterization
Page:
7
Paragraph:
3
Lines:
4­
5
EPA
statement:

All
labels
indicate
that
citrus
trees
are
particularly
sensitive
to
thidiazuron.

Bayer's
comment:

The
label
statements
regarding
citrus
and
cantaloupe
do
not
necessarily
indicate
that
these
species
are
more
sensitive
than
the
species
tested,
but
indicate
that
some
grower
concerns
were
raised
for
these
species.
The
buffer
language
that
was
developed
was
not
based
on
a
quantitative
analysis
for
buffers.
It
is
highly
likely
that
a
smaller
buffer
would
be
equally
protective.

EPA
response:
EPA
will
remove
the
language
which
infers
that
citrus
and
cantaloupe
are
more
sensitive
than
surrogate
species
tested.
Instead,
more
generic
language
will
be
included
which
states
that
non­
target
plants
can
be
more
or
less
sensitive
than
those
tested
in
the
laboratory.

Risk
Assessment
Approach
Page:
9
Paragraph:
3
Lines:
1­
3
EPA
statement:

For
the
aquatic
risk
assessment,
in
addition
to
parent
thidiazuron,
we
include
the
compounds
AE
F132147
and
AE
C421200
(
photodegradates
of
thidiazuron
produced
in
water)
and
the
thidiazuron
metabolite
AE
F132145.

Bayer's
comment:

While
BCS
acknowledges
the
approach
taken
with
this
screening
level
risk
assessment,
it
should
be
made
clear
that
this
approach
is
highly
conservative.
The
risk
assessment
is
based
on
the
estimated
environmental
concentration
(
EEC)
of
the
parent
compound;
the
EEC
for
the
photodegradate
would
be
significantly
less
than
that
of
the
parent.
This
approach
may
not
be
appropriate
for
all
risk
assessments.

EPA
response:
EFED
agrees
that
this
approach
is
conservative,
and
this
is
stated
in
the
aquatic
exposure
portion
of
the
risk
assessment.
Additional
language
indicating
this
has
been
added
to
the
"
Risk
Assessment
Approach"
section.
8
Measures
of
effects
Page:
9
Paragraph:
5
Line:
na
EPA
statement:

EPA
provides
a
list
of
studies
for
the
potential
measure
of
effects
for
mammals.
For
chronic
studies,
the
list
only
lists
a
3
generation
rat
study
and
note
that
a
rabbit
developmental
study
was
not
available.

Bayer's
comment:

Numerous
chronic
studies
are
available,
including
a
core
reproduction
study
and
rat
and
rabbit
developmental
studies
(
rat
developmental
toxicity
MRID
00077127,
rabbit
developmental
toxicity
MRID
46121507,
2­
generation
reproduction
MRID
46209601).
The
US
EPA
HED
chapter
lists
and
reviews
all
of
these
studies.

EPA
Response:
Language
indicating
that
a
rabbit
developmental
study
was
not
available
has
been
removed.
The
chronic
risk
assessment
has
been
revised
to
be
based
on
the
2­
generation
rat
study.

Measures
of
Exposure
 
Terrestrial
Plants
Page:
12
Paragraph:
7
Line:
na
EPA
statement:

EPA
provides
a
description
of
the
exposure
estimates
for
nontarget
plants,
which
includes
potential
screening
level
exposure
assessments
for
runoff.

Bayer's
comment:

The
runoff
estimates
used
for
the
nontarget
plant
risk
assessment
are
highly
conservative.
A
more
refined
estimate
could
be
derived
using
the
loading
estimates
from
the
PRZM
model.

EPA
response:
The
Agency
is
developing
a
terrestrial
plant
exposure
model
which
is
intended
to
improve
the
ability
to
estimate
potential
pesticide
exposure
from
runoff
and
spray
drift.
Until
such
a
model
has
undergone
the
necessary
QA/
QC,
TERR­
PLANT
will
remain
the
approved
model
for
terrestrial
plant
risk
assessments.

III.
Integrated
Environmental
Risk
Characterization
Risk
to
Aquatic
Organisms
Page:
17
Paragraph:
2
Line:
1
EPA
statement:

EPA
provides
the
risk
quotients
for
aquatic
assessment.

Bayer's
comment:

It
should
be
noted
that
the
acute
aquatic
risk
assessment
for
fish
is
utilizing
the
toxicity
values
of
one
of
the
degradates,
but
the
estimated
environmental
exposure
values
of
the
parent
compound.
For
a
screening
level
assessment,
this
is
a
highly
conservative
approach.

A
discussion
about
the
conservative
nature
of
the
assessment
should
be
included.
9
The
endangered
species
RQ's
for
aquatic
plants
were
calculated
using
an
EC05
value
instead
of
the
available
NOEC
values.
The
RQ's
for
aquatic
plants
should
be
corrected
(
see
comments
Appendix
G,
page
110).

EPA
response:
Language
has
been
added
to
the
"
Risk
Assessment
Approach"
discussion
which
notes
that
use
of
the
degradate
toxicity
data
is
conservative.
However,
since
no
direct
risk
to
freshwater
fish
from
exposure
to
thidiazuron
is
indicated,
no
further
characterization
of
the
approach
is
necessary
in
the
Risk
Characterization.

The
endangered
species
RQ
values
for
aquatic
plants
were
correct,
and
did
not
exceed
the
level
of
concern.
The
EC
05
values
that
were
listed
in
the
table
have
been
replaced
with
the
NOEC
values
on
which
the
RQs
were
based.

Risk
to
Birds
Chronic
Risk
­
Birds
Page:
18
Paragraph:
4
Line:
1­
4
EPA
statement:

Chronic
risk
to
birds
cannot
be
quantitatively
evaluated
at
this
time
because
chronic
data
were
not
submitted.
However,
available
data
showing
that
chronic
levels
of
concern
for
mammals
are
exceeded
and
that
thidiazuron
is
persistent
in
the
field.
Suggest
that
there
is
also
potential
for
chronic
risk
to
birds
should
they
be
of
equal
or
greater
sensitivity
than
mammals.

Bayer's
comment:

Chronic
or
reproductive
toxicity
tests
with
birds
are
not
necessary
as
there
is
minimal
to
no
exposure
during
the
breeding
season.
The
normal
use
pattern
is
for
one
application,
a
second
application
may
occur
but
within
a
short
time
frame.
The
application
occurs
in
the
autumn
(
when
birds
are
not
breeding)
on
leaves
which
then
rapidly
drop
and
degrade
on
the
soil.

Moreover,
cotton
leaves
are
not
an
attractive
food
source
for
birds
and
thidiazuron
is
not
expected
to
bioaccumulate
in
animal
tissues.
Therefore
the
chronic
risk
to
birds
is
low.

EPA
response:
Please
see
response
to
"
Page:
3
Paragraph:
2
Bullet
3
for
birds"
above.

Risk
to
Mammals
Chronic
Risk
­
Mammals
Page
18
(
Paragraph
6)
to
Page
20
EPA
statement:

A
summary
of
the
rat
3­
generation
reproduction
study
is
provided,
followed
by
the
chronic
10
screening
level
risk
assessment.

Bayer's
comment:

The
rat
3­
generation
study
used
for
the
wild
mammal
risk
assessment
is
an
older
study
that
did
not
meet
all
current
guideline
requirements,
and
should
not
be
relied
upon
for
determining
potential
effects
to
wild
mammals.

°
The
study
did
not
have
food
consumption
measurements
meaning
that
there
is
uncertainty
in
the
actual
exposure
concentrations
expressed
in
mg/
kg
bw/
d.
Risk
quotients
based
on
the
dietary
concentration
do
not
exceed
the
level
of
concern.

°
The
potential
effect
observed
in
the
3
generation
study
was
an
effect
on
litter
size.

There
was
no
dose
response
relationship
noted
in
the
study
and
no
consistent
trend
for
an
effect
in
the
other
generations.

°
Within
the
3
generation
study
the
mean
litter
size
for
the
control
group
varied
from
8.7
to
12.8.
The
mean
litter
size
for
all
treatment
groups
and
all
generations,
except
one,
completely
falls
within
the
same
range
as
the
control.
For
the
one
exception
(
generation
F2a),
there
was
no
significant
difference
between
the
controls
and
the
treatment
groups.

°
None
of
the
current
studies
used
in
the
HED
review,
including
the
rat
2
generation
study,
the
rat
developmental
study
or
the
rabbit
developmental
study,
noted
any
statistically
or
biologically
significant
effect
of
thidiazuron
on
litter
size.

According
the
HED
chapter
the
NOAEL
for
the
rat
2
generation
study
was
400
ppm
(
35
mg/
kg
bw/
day),
with
the
LOAEL
being
1200
ppm
(
108.5
mg/
kg
bw/
d)
based
on
a
low­
level
effect
on
body
weight
gain.
These
low
level
alterations
in
body
weight
gain
may
have
little
ecological
relevance
The
screening
level
risk
assessment
should
be
refined
utilizing
the
results
from
the
newer,

more
reliable
reproduction
study.
Table
1
provides
the
risk
quotients
for
a
screening
level
analysis
utilizing
the
NOEL
of
400
ppm
or
35.4
mg/
kg
bw/
d.
While
there
are
some
low
exceedences
for
the
level
of
concern
(
LOC),
the
majority
of
the
exposure
scenarios
do
not
exceed
the
LOC,
and
all
RQ's
are
less
than
2.
However
further
refinements
in
the
assessment
demonstrate
that
the
potential
risk
is
low
and
that
under
more
realistic
environmental
conditions,
many
factors
are
present
that
would
mitigate
any
potential
risk.

°
All
RQ's
using
the
predicted
mean
residue
values
were
below
the
LOC
(
Table
1).

°
All
RQ's
using
the
dietary
exposure
estimates
(
Table
2)
are
below
the
LOC.
It
should
be
noted
this
was
also
true
for
the
screening
level
assessment
using
the
3­

generation
study.

°
Potential
food
sources
off­
site
of
the
target
application
would
have
significantly
less
exposure,
and
result
in
a
minimal
risk
for
mammals.

°
Using
the
maximum
exposure
values,
the
assumption
is
the
animal
consumed
100%

of
its
diet
from
fully
treated
food
items.
This
assumption
is
highly
conservative
11
under
normal
circumstances,
but
is
particularly
inappropriate
for
the
cotton
defoliant
application
scenario.
Cotton
plants,
particularly
late
in
the
season
are
not
a
primary
food
source
for
wild
mammals
(
or
birds
for
that
matter).
At
the
time
of
application,

the
cotton
plant
is
fully
grown.
The
spray
application
results
in
a
high
percentage
of
the
thidiazuron
being
applied
to
the
cotton,
the
leaves
will
then
desiccate
and
fall
to
the
ground,
which
would
not
be
a
desirable
food
source.
And
as
mentioned
above
the
RQ's
using
the
mean
residue
values
do
not
exceed
the
LOC.

°
According
to
Linders
et
al.
(
2002),
the
crop
interception
factor
for
cotton
during
stem
elongation/
crop
cover/
flowering
is
70%,
and
during
senescence
is
90%.
If
we
use
the
70%
as
a
conservative
estimate,
this
would
indicate
that
the
cotton
intercepts
at
least
70%
of
the
applied
thidiazuron.
Given
that
the
mammals
do
not
consume
the
cotton,

the
potential
exposure
on
other
food
items
within
the
field
will
be
significantly
lower
(
70%)
than
that
used
in
the
models
for
the
screening
level
risk
assessment.
If
we
reduce
the
exposure
estimates
by
70%,
none
of
the
scenarios
in
Table
1
would
exceed
the
chronic
LOC.

In
conclusion,
while
the
screening
level
risk
assessment
may
result
in
some
LOC
exceedences,
a
refined
risk
assessment
indicates
that
there
are
mitigating
factors
in
the
environment
that
significantly
reduce
the
risk
and
therefore
there
is
low
chronic
risk
for
mammals.

EPA
response:

The
risk
assessment
has
been
revised
to
base
the
mammalian
chronic
risk
assessment
on
the
results
of
the
2­
generation
rat
reproduction
study.
While
there
were
no
frank
reproductive
effects
observed
in
this
study,
there
was
a
delay
of
female
sexual
maturation
and
disruption
of
the
estrous
cycle
observed
at
the
1200
ppm
dose.
Therefore,
the
study
indicates
the
potential
for
effects
which
could
affect
reproduction
and
survival
in
wild
mammals.

Risk
to
Terrestrial
Plants
Pages:
21
and
22
Tables
summarizing
the
risk
quotients
for
single
and
multiple
applications
EPA
statement:

Terrestrial
plant
risk
quotients
for
single
and
multiple
applications
for
plants
in
adjacent
to
treated
sites
and
plants
in
semi­
aquatic
areas.

Bayer's
comment:

The
terrestrial
plant
risk
quotients
will
need
to
adjusted
to
account
for
the
inclusion
of
the
application
efficiency
in
the
drift
exposure
estimates
for
aerial
applications
(
see
comments
for
Appendix
G,
Table
10)
12
EPA
response:
See
response
above
on
the
inapplicability
of
application
efficiency
factors
to
spray
drift.

Page:
21
­
23
Text
Discussing
Risks
for
Nontarget
Plants
Also
relevant
for
pages
3­
4
of
the
cover
memo,
and
pages
4­
6
of
the
risk
assessment
EPA
statement:

EPA
provides
their
risk
characterization
for
nontarget
terrestrial
plants.

Bayer's
comment:

It
is
important
to
note
that
not
all
nontarget
plant
species
were
equally
sensitive
to
thidiazuron.
In
the
seedling
emergence
test
7
of
the
10
species
tested
had
minimal
to
no
effect
from
exposure
to
product.
Similarly,
for
the
vegetative
vigor
test,
6
of
the
10
species
had
minimal
to
no
effects
observed.
In
terms
of
the
risk
assessment,
the
majority
of
the
plant
species
tested
triggered
no
concern
(
no
LOC
exceedence)
for
nontarget
species.
For
all
of
the
possible
scenarios:
vegetative
vigor
or
seedling
emergence,
single
or
multiple
applications,
adjacent
species
or
semi­
aquatic
species,
and
non­
endangered
and
endangered
species,
no
more
than
2
species
trigger
potential
concerns
for
nontarget
plants.

EFED
response:

The
comment
that
not
all
test
species
exhibited
sensitivity
to
thidiazuron
in
laboratory
tests
is
correct.
The
Risk
Characterization
has
been
amended
to
reflect
this
fact.

Bayer's
comment:

°
Plants
in
close
proximity
to
agriculture
are
generally
highly
resilient
and
used
to
disturbance,
which
may
lessen
the
impact
of
any
potential
thidiazuron
exposure.

EFED
response:
It
is
beyond
the
scope
of
this
screening
assessment
to
catalog
which
non­
listed
plants
are
found
in
close
proximity
to
agriculture,
and
whether
they
are
resistant
to
exposure.
For
this
reason,
the
screening
assessment
is
based
on
surrogate
plant
species
tested
in
seedling
emergence
and
vegetative
vigor
laboratory
tests.

Bayer's
comment:

Thidiazuron
at
the
specified
application
rates
would
not
have
an
effect
on
many
plant
species,
and
the
probability
of
such
an
effect
is
reduced
even
more
at
the
average
application
rates.

EPA
response:
The
Agency
calculated
terrestrial
plant
RQs
using
average
application
rates,
and
included
them
in
the
risk
assessment.
Even
at
average
application
rates,
RQs
for
dicots
exceeded
some
levels­
of­
concern.
13
Bayer's
comment:

°
Thidiazuron
is
most
efficacious
when
the
target
plant
is
vegetatively
dormant
and
through
the
reproductive
cycle
(
Cathey,
1986).
Therefore
when
considering
potential
effects
on
plant
reproduction,
must
consider
the
efficacy
at
the
different
life
stages.
(
Cathey,
1986.
Physiology
of
defoliation
in
cotton
production.
In:
Cotton
Physiology,
Mauney
and
Stewart
(
eds),
The
Cotton
Foundation,
Memphis,
TN)

EPA
response:
Seedling
emergence
and
vegetative
vigor
tests
are
acute
toxicity
tests
which
do
not
test
plants
when
they
are
vegetatively
dormant
nor
through
the
reproductive
cycle.
Therefore,
the
test
plants
in
the
laboratory
may
less
sensitive
than
non­
target
plants
in
the
field,
and
the
results
of
the
tests
may
underestimate
thidiazuron
toxicity.

Bayer's
comment:

°
There
are
many
environmental
factors
that
influence
the
potential
impact
of
thidiazuron.
As
noted
by
Cathey,
1986,
the
potential
for
defoliation
is
dependent
upon
the
condition
of
the
plant,
the
concentration
of
the
defoliant,
and
environmental
factors.
Defoliants
work
best
with
high
temperatures
and
high
humidity,
night
temperature
>
16C,
and
low
nitrogen
supplies.
Therefore
potential
impacts
of
thidiazuron
on
nontarget
plants
must
also
consider
these
environmental
factors.

EPA
response:

1)
The
consideration
of
the
condition
of
non­
target
plants
near
treated
fields
would
represent
an
advanced
refinement
to
the
risk
assessment
which
may
not
be
possible
given
currently
available
data
and
assessment
techniques.

2)
Non­
target
plants
would
be
exposed
to
thidiazuron
under
the
same
climatic
conditions
which
make
it
an
effective
cotton
defoliant
when
it
is
applied.
Since
laboratory
tests
are
not
conducted
under
these
conditions,
test
plants
may
be
exposed
under
conditions
which
make
them
less
sensitive
to
thidiazuron
than
they
are
in
the
field.

Bayer's
comment:

°
The
application
scenario
for
the
nontarget
plant
studies
is
to
spray
a
high
volume
of
a
dilute
solution,
mimicking
a
direct
spray
scenario.
However,
in
a
spray
drift
scenario
(
which
is
relevant
for
nontarget
plants)
the
droplet
spectrum
consists
of
a
few
drops
of
a
more
concentrated
solution.
The
potential
impact
on
a
plant
may
be
reduced
under
this
different
exposure
scenario,
which
should
also
be
considered
in
assessing
the
potential
risk.

EPA
response:
The
number
of
droplets
to
which
non­
target
plants
would
be
exposed
would
presumably
be
dependent
on
the
distance
from
the
treated
field.
A
study
that
reports
the
effects
of
thidiazuron
drift
at
varying
distances
from
the
treated
field
would
be
useful
for
a
refined
terrestrial
plant
risk
assessment.
The
guideline
laboratory
data
submitted
for
the
screening
assessment
does
14
not
allow
an
assessment
of
whether
actual
spray
drift
in
the
field
would
lead
to
greater
or
lesser
effects
to
non­
target
plants.

Bar
graphs
and
following
discussion
on
page
43
EPA
statement:

EPA
generated
bar
graphs
depicting
percent
effect
on
plants
and
the
no­
spray
zone
and
the
follow
discussion.

Bayer's
comment:

The
bar
graphs
showing
a
percent
effect
vs.
a
no­
spray
zone
are
not
appropriately
utilized
and
it
is
not
clear
what
the
graphs
are
depicting.
There
is
no
description
as
to
how
these
bar
graphs
were
generated.
Also,
each
graph
should
have
a
footnote
itemizing
which
effect
was
used
to
generate
for
the
regression
analysis.
Not
all
effects
to
nontarget
plants
are
equal
in
severity,
and
therefore
it
would
be
useful
for
risk
management
decisions,
to
have
a
better
understanding
of
the
graphs.

EPA
response:
As
described
in
the
RED
chapter
under
"
Phytotoxicity
and
Downwind
Distance",
the
bar
graphs
indicate
the
distance
to
which
AGDRIFT
calculated
thidiazuron
drift
at
loadings
at
which
a
particular
toxic
effect
level
would
be
experienced
by
a
particular
tested
plant
species.
This
section
describes
how
the
graphs
were
generated,
and
describes
how
the
varying
effects
levels
were
calculated.
The
effects
used
to
generate
the
regression
analysis
are
those
used
for
RQ
calculations;
this
has
been
clarified
in
the
document.

Bayer's
comment:

It
is
common
in
nontarget
plant
tests
to
have
a
weak
dose
response
relationship
between
exposure
and
effect.
Indication
that
the
dose­
response
relationship
is
not
optimized
for
the
full
range
of
the
ECx
values
being
attempted
to
be
generated
for
the
graphs.
For
example,

the
lettuce
vegetative
vigor
response
(%
inhibition
relative
to
control)
for
shoot
weight
was
12,
27,
12,
22,
and
40%
for
lowest
to
highest
doses
in
the
study.
With
such
a
poor
dose
response
relationship,
the
confidence
intervals
around
the
ECx
values
that
were
generated
would
be
very
large.
However,
no
confidence
intervals
were
provided
to
assist
in
the
interpretation
of
the
bar
graphs.
For
the
soybeans
and
tomato
vegetative
vigor
studies,
no
effect
greater
than
25%
was
observed
on
shoot
length
or
weight,
and
yet
a
regression
to
generate
EC10
to
EC90
values
was
attempted.
It
is
clear
that
these
regressions
were
extrapolating
well
beyond
the
range
of
the
concentrations
tested,
and
there
would
be
high
uncertainty
(
high
variability)
in
the
results
obtained.
As
demonstrated
by
these
examples,
the
use
of
the
bar
graphs
should
be
limited
and
be
interpreted
with
extreme
caution.
15
EPA
response:
The
RED
chapter
informs
the
reader
that
a
lognormal
toxicity
distribution
is
assumed,
and
the
calculation
by
which
different
effect
levels
on
the
distribution
are
estimated
is
provided.
However,
the
document
has
been
revised
to
acknowledge
that
there
is
uncertainty
in
this
calculation
of
dose
response
curves.
The
uncertainty
in
the
vegetative
vigor
dose­
response
curves
for
soybean
and
tomato
does
not
lead
to
a
greater
finding
of
risk;
the
bar
graphs
do
not
indicate
deposition
at
loadings
greater
than
that
equivalent
to
the
EC25
at
any
distance
from
the
treated
field.

Bayer's
comment:

EPA
notes
on
page
43,
second
paragraph,
the
endangered
species
terrestrial
plant
LOC
is
exceeded
when
the
environmental
concentration
is
greater
than
the
EC
05.
This
statement
is
inaccurate.
The
endangered
species
terrestrial
plant
concern
occurs
when
the
environmental
concentration
is
greater
than
the
NOEC.
The
EC
05
can
be
utilized
in
cases
where
an
NOEC
was
not
achieved
in
the
non­
target
plant
tests.
However,
for
thidiazuron,
a
NOEC
was
achieved
for
the
most
sensitive
species,
and
therefore
the
NOEC
was
utilized
for
the
risk
assessment.
As
stated
above,
the
bar
graphs
utilizing
regression
analysis
outside
the
boundaries
of
the
concentrations
tested
(
i.
e.,
below
the
NOEC)
may
be
inappropriate
for
risk
management
decisions.

EPA
response:
The
bar
graphs
do
not
include
the
EC
05,
but
the
estimated
EC
10
through
EC
90.
The
bar
graphs
are
meant
to
show
the
distance
to
which
various
effect
levels
might
occur,
not
to
establish
an
endpoint
for
endangered
species
RQ
calculation.

Endangered
Species
Assessment
Species
Affected
Page:
44
Paragraph:
5
First
bullet
EPA
statement:

Chronic
LOC's
are
exceeded
for
mammals
foraging
on
short
grass
and
broadleaf
forage
and
small
insects.

Bayer's
comment:

As
noted
elsewhere
in
the
response
to
the
risk
assessment,
the
presumed
chronic
risk
for
mammals
that
was
derived
in
the
screening
level
risk
assessment
is
mitigated
through
a
refined
toxicity
threshold
and
through
environmental
factors
that
significantly
reduce
the
potential
exposure
for
mammals.
Therefore,
there
is
minimal
risk
for
endangered
mammals.

Additionally,
the
majority
of
mammals
that
would
coincide
at
the
county
level
with
thidiazuron
applications
would
not
be
consuming
the
potential
food
items
and/
or
would
not
be
close
proximity
to
cotton
fields.
16
EPA
response:
See
comments
above
on
the
inclusion
of
the
2­
generation
rat
reproductive
study
in
the
risk
assessment.

APPENDIX
E:
Ecological
Hazard
Data
Page:
97
Table
9
 
Acute
Toxicity
of
Thidiazuron
to
Birds
(
Dietary
administration)

EPA
statement:

An
MRID
of
00066168
was
reported
for
the
acute
dietary
toxicity
study
for
mallard
duck.

Bayer's
comment:

The
correct
MRID
No.
for
this
study
is
00066169.

EPA
response:
This
has
been
changed.

Page:
99
Table:
11
Line:
na
EPA
statement:

Table
11
only
provides
information
on
an
older
3
generation
reproduction
study
with
rats.

Bayer's
comment:

As
noted
earlier,
other
data
for
mammals
is
available
and
should
be
included
in
this
table.

EPA
response:

The
table
has
been
revised
to
include
the
2­
generation
rat
reproduction
study.

Page:
102
Table:
16
Line:
na
EPA
statement:

Carrot
%
response
was
given
as
>
25%,
however
no
endpoint
is
indicated
as
affected.

Bayer's
comment:

The
carrot
%
response
should
be
<
25%
since
no
effects
were
observed
in
the
study,
as
noted
in
the
table.

EPA
response:
This
has
been
changed.

APPENDIX
G:
Detailed
Risk
Quotients
Page:
109
Table:
1
Line:
Freshwater
fish
EPA
statement:

The
data
for
the
acute
freshwater
fish
endpoint
is
the
LC50
for
the
degradate.

Bayer's
comment:

While
it
is
noted
in
the
footnote
that
the
acute
toxicity
data
is
for
the
photodegradate,
it
should
also
be
noted
that
this
approach
is
highly
conservative.
The
exposure
estimated
for
17
the
photodegradate
would
be
significantly
less
than
that
of
the
parent.

EPA
response:
This
has
been
noted
in
the
body
of
the
document
under
"
Risk
Assessment
Approach".

Page:
110
Table:
2
Line:
Aquatic
vascular
plant
EPA
statement:

The
duckweed
(
Lemna
gibba)
EC05
value
is
used
for
the
endangered
species
toxicity
threshold.

Bayer's
comment:

The
thidiazuron
duckweed
test
resulted
in
an
NOEC
of
8600
(
as
noted
in
Appendix
E,
Table
7).
It
is
not
appropriate
to
calculate
a
EC05
in
such
cases.
Therefore
the
NOEC
should
be
used
for
the
aquatic
vascular
plant
toxicity
threshold.
The
resulting
endangered
species
RQ
would
be
0.001.

EPA
response:
This
has
been
changed.

Page:
110
Table:
2
Line:
Blue­
green
algae
EPA
statement:

The
blue­
green
algae
EC05
value
is
used
for
the
endangered
species
toxicity
threshold.

Bayer's
comment:

The
thidiazuron
blue­
green
algae
test
resulted
in
an
NOEC
of
2800
(
as
noted
in
Appendix
E,

Table
7).
It
is
not
appropriate
to
calculate
a
EC05
in
such
cases.
Therefore
the
NOEC
should
be
used
for
the
aquatic
vascular
plant
toxicity
threshold.
The
resulting
endangered
species
RQ
would
be
0.004.

EPA
response:
This
has
been
changed.

Page:
110
Table:
2
Line:
Marine
diatom
EPA
statement:

The
marine
diatom
EC05
value
is
used
for
the
endangered
species
toxicity
threshold.

Bayer's
comment:

The
thidiazuron
marine
diatom
(
Skeletonema)
test
resulted
in
an
NOEC
of
110
(
as
noted
in
Appendix
E,
Table
7).
It
is
not
appropriate
to
calculate
a
EC05
in
such
cases.
Therefore
the
NOEC
should
be
used
for
the
aquatic
vascular
plant
toxicity
threshold.
The
resulting
endangered
species
RQ
would
be
0.10.

EPA
response:
This
has
been
changed.
18
Page:
120
and
124
Table:
10
and
13
Line:
Aerial
application
scenarios
EPA
statement:

The
estimated
environmental
exposure
(
EEC)
for
drift
from
aerial
applications
is
given
as
0.01
lb
ai/
A.

Bayer's
comment:

As
noted
on
page
13,
the
spray
drift
for
aerial
applications
is
assumed
to
be
5%
with
an
application
efficiency
of
60%.
However
the
application
efficiency
was
not
included
in
the
drift
calculation.

In
Table
10,
the
aerial
application
spray
drift
EEC
should
be
0.006
lb
ai/
A,
the
total
loading
to
adjacent
areas
should
be
0.008
lb
ai/
A
and
the
total
loading
to
semi­
aquatic
areas
should
be
0.026
lb
ai/
A.

In
Table
13,
the
aerial
application
spray
drift
EEC
should
be
0.006
lb
ai/
A,
the
total
loading
to
adjacent
areas
should
be
0.009
lb
ai/
A
and
the
total
loading
to
semi­
aquatic
areas
should
be
0.036
lb
ai/
A.

These
corrections
would
impact
all
succeeding
aerial
application
risk
quotient
tables
for
nontarget
plants.

Pages:
121­
127
Tables
:
11,
12,
14,
15
Line:
na
EPA
statement:
The
aerial
application
drift
EEC's,
total
loadings
and
resulting
RQ's
are
based
on
the
calculations
in
Tables
10
and
13.

Bayer's
comment:

As
noted
above,
the
aerial
application
drift
EEC
did
not
account
for
application
efficiency.

Therefore
the
EEC's
and
risk
quotients
for
aerial
applications
should
be
corrected.
A
significant
reduction
in
the
RQ's
is
observed
with
the
corrected
exposure
values.

Table
11,
for
plant
adjacent
to
treated
sites
and
for
plants
in
semi­
aquatic
areas,
the
aerial
application
RQ's
ranged
from
0.03
to
5.45.

Table
12,
for
plant
adjacent
to
treated
sites
and
for
plants
in
semi­
aquatic
areas,
the
aerial
application
RQ's
ranged
from
0.03
to
31.58.

Table
14,
for
plant
adjacent
to
treated
sites
and
for
plants
in
semi­
aquatic
areas,
the
aerial
application
RQ's
ranged
from
0.03
to
5.45.

Table
15,
for
plant
adjacent
to
treated
sites
and
for
plants
in
semi­
aquatic
areas,
the
aerial
application
RQ's
ranged
from
0.03
to
31.58.

EPA
Response:
As
noted
above,
the
application
efficiency
does
not
apply
to
spray
drift
from
the
treated
field,
but
to
the
amount
of
applied
pesticide
that
is
successfully
applied
to
the
field.
It
would
not
be
appropriate
to
apply
that
efficiency
factor
to
the
default
spray
drift
assumption
of
5%
in
the
TERR­
PLANT
model,
nor
to
spray
drift
estimates
from
AGDRIFT.
19
Page:
124
Table:
13
Line:
footnote
1
EPA
statement:

Footnote
1
states
that
the
EEC
assumed
the
grass
peak
residue
value
(
day
6)
from
the
first
application
at
0.2
lb
ai/
A
and
added
initial
residue
value
at
0.1
lb
ai/
A
for
a
second
application.

Bayer's
comment:

It
is
unclear
why
the
grass
peak
residue
value
would
be
used
for
a
nontarget
plant
risk
assessment.
The
footnote
should
be
corrected
to
reflect
that
the
EEC
values
were
based
on
the
same
assumptions
as
for
a
single
applications
(
runoff
and
drift),
but
that
a
second
application
was
modeled
using
the
ELLFATE
program
and
its
assumptions
for
half­
life.

EPA
response:

The
multiple
application
risk
quotients
have
been
removed
from
the
risk
assessment,
and
a
qualitative
discussion
of
potential
risks
from
multiple
applications
has
been
added.
