Appendix
A,
Page
1
of
13
Appendix
A.
Executive
Summary
from:
A
Probabilistic
Model
and
Process
to
Assess
Acute
Lethal
Risk
to
Birds.

Full
document
is
available
at:
http://
www.
epa.
gov/
oscpmont/
sap/
2001/
index.
htm
Appendix
A,
Page
2
of
13
A
Probabilistic
Model
and
Process
to
Assess
Acute
Lethal
Risks
to
Birds
Prepared
for
the
March
13­
16,
2001
Scientific
Advisory
Panel
Meeting
February
16,
2001
Prepared
by
Edward
Fite,
M.
S.*
Edward
Odenkirchen,
Ph.
D.*
Timothy
Barry,
Sc.
D.**
*
Environmental
Fate
and
Effects
Division/
Office
of
Pesticide
Programs
**
Office
of
the
Administrator/
Office
of
Policy,
Economics,
and
Innovation
United
States
Environmental
Protection
Agency
Appendix
A,
Page
3
of
13
EXECUTIVE
SUMMARY
Background
EPA
is
implementing
a
new
tiered
process
to
conducting
ecological
risk
assessments,
which
will
be
used
under
the
FIFRA
regulatory
framework.
This
approach
will
include
probabilistic
tools
and
methods
at
the
more
refined,
higher
levels
(
tiers)
to
provide
information
regarding
the
probability
or
likelihood
of
the
impact
as
well
as
the
magnitude
or
severity
of
the
effect.
This
new
process
will
be
used
to
better
characterize
the
impact
of
pesticides
on
non­
target
terrestrial
and
aquatic
organisms.
The
results
of
these
analyses
will
in
turn
be
used
by
risk
management
divisions
in
decision­
making.

Under
this
new
tiered
approach,
screening
level
(
deterministic)
assessments
are
first
conducted
and
are
generally
based
on
conservative
assumptions
and
generic
data.
More
complex
probabilistic
assessments,
representing
increasingly
realistic
biological
and
exposure
scenarios,
are
performed
for
those
pesticides
judged
at
the
screening
level
to
potentially
pose
the
most
serious
risk
and
that
are
believed
to
require
further
characterization
to
determine
appropriate
regulatory
action.

The
Environmental
Fate
and
Effects
Division
(
EFED)
in
EPA's
Office
of
Pesticide
Programs
conducted
a
screening
level
assessment
for
agricultural
uses
of
ChemX
on
a
variety
of
crops
using
existing
EFED
deterministic
methods.
This
deterministic
assessment,
which
was
based
on
risk
quotients,
indicated
high
avian
acute
risk.
An
evaluation
of
the
field
studies,
data
from
monitoring
programs,
open
literature,
and
well
documented
incident
reports
provided
confirmatory
evidence
of
the
occurrence
of
acute
impacts
to
birds
following
ChemX
application
under
current,
approved
label
rates
and
methods.
Thus,
the
weight­
of­
the
evidence
for
the
effects
on
birds
was
compelling,
leading
to
a
high
level
of
certainty
in
concluding
acute
high
risks
to
birds.
In
addition,
significant
concerns
were
raised
about
potential
chronic
exposure
to
birds
and
for
possible
impacts
to
other
non­
avian
wildlife
as
well.

As
a
result
of
the
potentially
significant
risk
to
non­
target
terrestrial
species,
ChemX
was
identified
as
a
candidate
for
a
more
refined,
higher
level
assessment
using
probabilistic
techniques.
The
refined
assessment
approach
presents
the
risk
in
more
quantitative
terms,
providing
an
estimate
of
the
probability
and
magnitude
of
adverse
effects.
Owing
to
the
existence
of
a
comparatively
more
robust
avian
toxicological
data
set,
the
scope
of
the
refined
risk
assessment
is
confined
to
direct
acute
effects
on
birds.

Overview
of
the
Refined
Assessment
Use
Scenarios
The
use
scenarios
for
the
refined
assessment
were
selected
based
on
a
use
profile
of
ChemX
and
the
results
of
the
screening
level
assessment.
Scenarios
were
selected
based
on
(
1)
the
large
Appendix
A,
Page
4
of
13
volume
of
ChemX
used,
(
2)
total
acres
potentially
treated,
and
(
3)
the
high
risk
quotients
calculated
along
with
the
supporting
field
evidence.

This
resulted
in
the
selection
of
the
following
use
scenarios:
at­
plant
and
foliar
application
of
flowable
ChemX
to
corn
and
foliar
application
to
alfalfa.
Both
scenarios
take
place
in
the
spring
and
in
the
midwestern
region
of
the
United
States.
It
should
be
noted
that
other
regions
of
the
country,
notably
western
states,
may
also
be
important
when
considering
alfalfa.
However,
the
results
of
an
analysis
of
midwestern
alfalfa
are
believed
to
be
representative
of
risks
to
avian
species
in
western
states
as
well.

For
both
corn
and
alfalfa,
the
scenarios
were
based
on
single
applications
of
ChemX.
The
application
rates
used
ranged
from
the
lowest
to
the
highest
on
the
label
along
with
typical
rates,
which
are
those
that
fall
within
the
range
of
application
rates
reported
to
encompass
the
largest
percentage
of
treated
acres.
In
some
cases,
only
a
single
application
rate
was
identified
on
the
label.
The
following
table
presents
a
complete
picture
of
the
use
scenarios
considered
in
the
refined
assessment.
The
application
rate
is
presented
in
ounces
(
oz)/
1000
feet
(
ft)
and
pounds
(
lb)
active
ingredient
(
ai)/
acre.
It
should
be
noted
that
2.5
oz/
1000
ft
is
equivalent
to
1
lb
ai/
acre.

Crop
Application
Method
Application
Rate
Rationale
for
Selection
of
Application
Rate
corn
at
plant
in­
furrow
2.5
oz/
1000
ft
only
rate
on
label
corn
at
plant
banded
2.5
oz/
1000
ft,
7
inch
band
only
rate
on
label
corn
foliar
1
lb
ai/
acre
maximum
rate
corn
foliar
0.75
lb
ai/
acre
typical
rate
corn
foliar
0.25
lb
ai/
acre
24­
c
minimum
rate
alfalfa
foliar
1
lb
ai/
acre
maximum
rate
alfalfa
foliar
0.5
lb
ai/
acre
typical
rate
alfalfa
foliar
0.125
lb
ai/
acre
minimum
rate
Selection
of
Focal
Species
In
screening
level
assessments,
the
species
tested
under
the
standard
guideline
studies
(
40
CFR
158)
are
assumed
to
be
representative
of
avian
species
in
general,
and
therefore
results
of
these
tests
along
with
conservative
exposure
assumptions
are
used
to
make
predictions
for
all
species.
In
contrast,
focal
species
are
selected
for
use
in
more
refined
assessments
to
provide
more
realistic
and
appropriate
biological
scenarios.
The
focal
species
provide
information
specific
to
that
species
and
are
also
considered
representative
of
other
species
with
similar
biological/
behavioral
characteristics
sharing
a
treated
area.
Selection
criteria
for
focal
species
included,
but
were
not
Appendix
A,
Page
5
of
13
limited
to,
species
likely
to
be
found
in
fields
treated
by
the
pesticide,
and
their
natural
history.

In
selecting
focal
species
for
the
refined
assessment,
EFED
considered
the
results
of
field
studies.
Avian
species
in
and
around
corn
and
alfalfa
fields
where
ChemX
was
being
applied
and
which
were
reported
dead
were
considered.
In
addition,
open
literature
on
birds
inhabiting
corn
and
alfalfa
fields
and
their
surrounding
habitat
as
well
as
information
on
the
feeding
and
nesting
habits
of
birds
associated
with
these
crops
were
evaluated.
Based
on
these
criteria,
the
following
species
were
selected
for
midwestern
corn
and
alfalfa:

Corn
Species
Diet
Preferences
Feeding
Sites
Nesting
Sites
Meadow
Lark
(
eastern
and
western)
insectivore
ground
ground
Horned
Lark
omnivore
ground
ground
Vesper
Sparrow
omnivore
ground
ground
Red­
winged
Blackbird
omnivore
ground
shrubs
Mourning
Dove
granivore
ground
tree
Killdeer
insectivore
ground
ground
Alfalfa
Species
Diet
Preferences
Feeding
Sites
Nesting
Sites
Dickcissel
omnivore
ground
ground
Western
Meadowlark
insectivore
ground
ground
Grasshopper
Sparrow
omnivore
ground
ground
Mourning
Dove
granivore
ground
ground
tree
Vesper
Sparrow
omnivore
ground
ground
Mallard
Duck
herbivore
ground
ground
Appendix
A,
Page
6
of
13
Construct
of
the
Refined
Terrestrial
Risk
Assessment
Model
EFED
has
developed
a
refined
assessment
model
to
estimate
the
magnitude
and
probability
of
acute
effects
to
non­
target
avian
species
from
pesticides.
The
basic
structure
of
the
model
can
be
expressed
by
the
general
equation:

Risk
=
f
(
exposure,
toxicity).

Since
risk
is
a
function
of
exposure
and
toxicity,
the
model
is
based
on
the
characterization
of
exposure
and
effects.
Distributions
were
developed
for
the
major
exposure
and
effects
variables,
which
were
combined
using
a
Monte
Carlo
analysis
to
estimate
the
probability
and
magnitude
of
effects.

The
major
parameters
addressed
in
the
model
are:
°
Food
habits
of
selected
focal
species,
which
are
proportioned
for
each
food
type
consumed;
°
Daily
ingestion
rate
of
food
and
water
as
a
function
of
body
weight;
°
Frequency
of
feeding
and
drinking
on
the
sprayed
field;
°
Water
sources
(
dew,
puddles,
and
ponds);
°
Distribution
of
residues
on
the
food
and
water
sources
as
a
function
of
application
rate;
°
Degradation
rates
of
food
residue
estimates,
and
°
Inter­
and
intra­
species
sensitivity
distributions.

A
flow
diagram
illustrating
the
overall
relationship
of
the
model
construct
is
provided
in
Figure
E­
1.
Appendix
A,
Page
7
of
13
.
Appendix
A,
Page
8
of
13
Model
Input
Data
The
conceptual
model
presented
may
be
used
to
conduct
a
refined
assessment.
The
model
uses
probabilistic
techniques
to
quantify
acute
mortality
for
a
variety
of
pesticides
based
on
oral
exposure
from
food
and
water
taken
in
treated
fields.
The
input
data,
however,
will
vary.
For
the
refined
assessment
of
ChemX,
the
following
input
data
were
used
in
the
model
simulations:

°
The
use
scenarios
for
midwestern
corn
and
alfalfa
described
previously.
Aerial
applications
were
run
with
existing
on­
field
puddles,
rain
events
forming
puddles
on
the
day
following
application,
and
no
rain
events;
°
The
focal
species
identified
earlier;
°
The
three
assumptions
of
bird
sensitivity
(
low,
medium,
and
high)
to
ChemX
when
species­
specific
toxicity
data
were
not
available;
and
°
A
seven
day
exposure
window
based
on
the
half­
life
of
ChemX.

Refined
Assessment
Results
for
ChemX
The
results
of
the
refined
assessment
for
ChemX
are
summarized
in
Tables
E­
1
through
E­
4.
Table
E­
1
provides
a
quantitative
summary
of
acute
bird
mortality
for
the
exposed
complex
of
species.
Since
the
sensitivity
of
most
of
the
focal
species
to
ChemX
is
unknown,
model
simulations
were
run
based
on
low,
medium,
and
high
sensitivity
to
ChemX
in
order
to
address
the
uncertainty
this
introduces
into
the
refined
assessment.

In
contrast,
Tables
E­
2,
E­
3
and
E­
4
provide
a
quantitative
summary
of
acute
bird
mortality
for
specific
species,
the
red­
winged
blackbird
and
the
mallard
duck.
Toxicity
data
were
available
for
both
species
and
were
thus
used
in
model
simulations.

Mean
Avian
Mortality
Levels
Table
E­
1
provides
a
summary
of
the
mean
mortality
levels
of
avian
species
that
occur
in
and
around
fields
treated
with
ChemX.
They
include
the
range
of
mortality,
which
is
based
on
the
mean
percent
mortality,
and
the
percentage
of
species
with
predicted
mortality
to
be
greater
than
0,
10,
and
70%.

For
example,
under
the
scenario
for
aerial
application
to
corn
at
the
lowest
application
rate
of
0.25
lb
ai/
acre,
the
range
of
mortality
for
the
complex
of
avian
species
exposed,
on
average,
is
between
0
and
88%.
That
is,
over
the
long
term,
some
species
are
experiencing
no
mortality
and
other
species
may
be
experiencing
an
average
mortality
up
to
88%.
Further,
on
average,
(
1)
70%
of
the
exposed
complex
of
species
will
have
some
mortality
due
to
ChemX
exposure,
with
30%
of
the
species
experiencing
no
mortality;
(
2)
35%
of
the
species
are
expected
to
experience
10%
mortality
or
greater;
and
(
3)
10%
are
experiencing
70%
mortality
or
greater,
which
may
range
up
to
88%.
Appendix
A,
Page
9
of
13
Table
E­
1.
Range
of
Predicted
Mean
Mortality
Results
(
Across
Species,
Application
Scenarios,
and
Exposure
Scenarios)
and
Percent
of
Species
Above
Selected
Levels
of
Mortality
Use
Scenario
Range
of
Mean
Percent
Mortality
Percentage
of
species
with
mean
mortality
X%
or
greater
than:

>
0
%
10%
70%

corn
aerial
0.25
lb
ai/
acre
(
minimum
rate)
0
to
88
70
35
10
corn
aerial
0.75
lb
ai/
acre
(
typical
rate)
0
to
98
85
65
15
corn
aerial
1
lb
ai/
acre
(
maximum
rate)
0
to
99
95
90
20
corn
banded
1
lb
ai/
acre
(
only
rate)
0
to
86
60
33
<
5
corn
in­
furrow
1
lb
ai/
acre
(
only
rate)
0
to
86
70
37
<
5
alfalfa
aerial
0.125
lb
ai/
acre
(
minimum
rate)
0
to
50
55
27
0
(
max.
50%
mortality)

alfalfa
aerial
0.5
lb
ai/
acre
(
typical
rate)
0
to
89
70
57
17
alfalfa
aerial
1
lb
ai/
acre
(
maximum
rate)
0
to
92
95
62
23
An
overview
of
Table
E­
1
suggests
that
the
majority
of
avian
species
using
ChemX­
treated
fields
of
corn
and
alfalfa,
regardless
of
application
rate
and
method,
will
experience,
on
average,
some
mortality.
Approximately
one­
third
or
more
of
the
exposed
species
expected
to
experience
10%
mortality
or
greater.
At
typical
aerial
application
rates
for
both
corn
and
alfalfa
crops,
approximately
60%
or
more
of
the
species
are
expected
to
experience
mean
mortality
of
10%
or
Appendix
A,
Page
10
of
13
greater
and
approximately
15%
are
expected
to
experience
70%
or
greater.
Except
for
the
lowest
aerial
application
rate
to
alfalfa,
very
high
average
levels
of
mortality
(
70%
or
greater)
will
occur
in
some
species.
Even
at
the
lowest
application
rate
for
alfalfa,
some
species
are
still
predicted
to
experience
nearly
50
percent
mortality.

Predicted
Mortality
for
Red­
Winged
Blackbirds
and
Mallard
Ducks
Tables
E­
2
through
E­
4
provide
summaries
of
predicted
mortality
for
the
red­
winged
blackbird
and
mallard
duck.
Both
species
have
toxicity
data
for
ChemX,
which
indicate
they
are
fairly
sensitive
to
this
pesticide.
Model
runs
were
based
on
the
toxicity
of
these
species
to
ChemX,
and
therefore
more
quantitative
conclusions
could
be
made
for
these
birds.

Table
E­
2
provides
a
range
of
predicted
mortality
for
red­
winged
blackbirds.
Even
at
the
lowest
rate
of
aerial
application
to
corn,
24%
mortality
of
the
exposed
red­
winged
blackbirds
is
expected
to
result,
on
average.
Mortality
is
expected
to
be
10%
or
greater
in
the
majority
of
cases
(
95%
probability).
However,
in
a
few
cases
(
5%
probability),
the
mortality
is
expected
to
be
between
45
and
50%
at
the
lowest
application
rate.
At
higher
application
rates,
the
mortality
is
predicted
to
be
even
greater.

Table
E­
2.
Range
of
Predicted
Mortality
for
Red­
Winged
Blackbirds
Use
Scenario
Mean
Mortality
Level
95%
Probability
of
Mortality
5%
Probability
of
Mortality
corn
aerial
0.25
lb
ai/
acre
(
minimum
rate)
24%

10%
45
­
50%

corn
aerial
0.75
lb
ai/
acre
(
typical
rate)
57%

40%
75
­
85%

corn
aerial
1
lb
ai/
acre
(
maximum
rate)
64%

45%
85
­
95%

corn
banded
1
lb
ai/
acre
(
only
rate)
30%

15%
45
­
64%

corn
in­
furrow
1
lb
ai/
acre
(
only
rate)
30%

15%
50
­
60%
Appendix
A,
Page
11
of
13
Predicted
mortalities
for
mallard
ducks
exposed
to
ChemX
are
summarized
in
Table
E­
3
for
zero
days
after
application
and
in
Table
E­
4
for
12
days
after
application.
High
mortality
rates
are
predicted
in
alfalfa
fields
for
both
temporal
application
cases,
with
the
exception
of
exposure
at
the
lowest
application
rate
approximately
two
weeks
after
pesticide
application.
If
mallard
ducks
feed
in
a
recently
treated
field
at
the
lowest
application
rate
zero
days
after
application,
the
mean
mortality
level
expected
is
82%.
There
is
a
95%
probability
that
mortality
will
be
65%
or
greater,
and
in
a
very
few
cases
(
5%
probability)
mortality
of
95%
or
greater
is
expected.
At
this
same
application
rate,
mortality
is
still
expected
12
days
later,
but
ChemX
residues
appear
to
decline
to
a
level
that
does
not
induce
high
mortality.

Table
E­
3.
Range
of
Predicted
Mortality
for
Mallard
Ducks
Zero
Days
After
Application
Use
Scenario
Mean
Mortality
Level
95%
Probability
of
Mortality
5%
Probability
of
Mortality
alfalfa
aerial
0.125
lb
ai/
acre
(
minimum
rate)
82%

65%

95%

alfalfa
aerial
0.5
lb
ai/
acre
(
typical
rate)
99%

95%

100%

alfalfa
aerial
1
lb
ai/
acre
(
maximum
rate)
100%

100%

100%

Table
E­
4.
Range
of
Predicted
Mortality
for
Mallard
Ducks
12
Days
After
Application
Use
Scenario
Mean
Mortality
Level
95%
Probability
of
Mortality
5%
Probability
of
Mortality
alfalfa
aerial
0.125
lb
ai/
acre
(
minimum
rate)
3%

0%

10%

alfalfa
aerial
0.5
lb
ai/
acre
(
typical
rate)
28%

10%

45%

alfalfa
aerial
1
lb
ai/
acre
(
maximum
rate)
54%

35%

70%
Appendix
A,
Page
12
of
13
Tables
E­
2
through
E­
4
provide
species­
specific
predictions
of
mortality
to
the
red­
winged
blackbird
and
mallard
duck
from
ChemX
application.
Results
for
both
species
indicated
significant
mortality
events
were
likely,
depending
upon
application
rate
and
method.
Analysis
of
all
mean
mortality
predictions
across
species
and
sensitivity
assumptions
suggests
that
significant
mortality
events
are
likely
for
a
large
number
of
species.

Conclusions
The
results
of
the
refined
assessment
quantify
the
conclusions
reached
in
the
lower
level,
screening
assessment.
Based
on
the
construct
of
this
probabilistic
model,
high
mortality
in
at
least
some
avian
species
will
occur,
regardless
of
the
application
rate
and
method,
following
the
application
of
ChemX
to
corn
and
alfalfa
using
the
application
methods
and
rates
examined
in
the
refined
assessment.

Specifically,
results
show
that
from
55%
to
95%
of
the
bird
species
using
midwestern
corn
and
alfalfa
fields
treated
with
ChemX
will
experience
some
mortality,
on
average.
Twenty­
seven
to
90%
of
the
species
are
likely
to
experience
at
least
10%
or
greater
mortality,
on
average,
while
up
to
23%
of
the
species
are
likely
to
experience
at
least
70%
mortality,
again
on
average.

The
relative
risks
of
some
application
methods
and
application
rates
can
be
differentiated.
Banded
and
in­
furrow
applications
of
ChemX
to
corn
are
predicted
to
result
in
lower
levels
of
mortality
to
more
species
than
aerial
application
of
ChemX
to
corn
at
comparable
and
even
lower
rates.
For
both
corn
and
alfalfa
aerial
applications,
lower
rates
of
application
result
in
lower
predicted
levels
of
mortality
for
more
species.
However,
even
for
those
application
rates
and
methods
showing
lower
impacts,
several
species
are
still
predicted
to
experience
relatively
high
mortality
levels.

In
reference
to
red­
winged
blackbirds,
mean
mortality
in
corn
is
predicted
to
be
from
24
to
64%.
But
for
some
groups
of
birds,
mortality
could
go
as
high
as
95%.
The
results
for
mallard
ducks
in
alfalfa
are
even
more
striking.
Depending
on
the
application
rate,
the
mean
mortality
is
predicted
to
range
from
57
to
99%
3
days
after
application,
39
to
95%
6
days
after
application,
and
from
3
to
54%
12
days
post
application.
Appendix
A,
Page
13
of
13
