UNITED
STATES
ENVIRONMENTAL
PROTECTION
AGENCY
WASHINGTON,
D.
C.
20460
OFFICE
OF
PREVENTION,
PESTICIDES
AND
TOXIC
SUBSTANCES
PC
CODE:
032501
CHEMICAL:
Disulfoton
DP
Barcode:
D280670
MEMORANDUM
February
25,
2002
SUBJECT:
Disulfoton:
Summary
of
Estimated
Drinking
Water
Concentrations
(EDWCs)
use
in
the
Human
Health
Risk
Assessment.

TO:
Betty
Shackleford,
PM
53
Christina
Scheltema,
PM
Team
Reviewer
Michael
Goodis
Reregistration
Branch
3
Special
Review
and
Reregistration
Division
(7508W)

FROM:
James
K.
Wolf,
Ph.
D.,
Soil
Scientist
Environmental
Risk
Branch
3
THRU:
Kevin
J.
Costello,
Acting
Branch
Chief
Environmental
Risk
Branch
3
Environmental
Fate
and
Effects
Division
(7507C)

This
memo
summarizes
the
Tier
II
estimated
drinking
water
concentrations
(EDWC)
for
disulfoton
in
surface
water
for
use
in
human
risk
assessment.
Tier
I
ground
water
concentrations
are
also
presented.
The
drinking
water
assessment
considers
both
parent
disulfoton
and
total
disulfoton
residues
(TDR,
the
sum
of
disulfoton,
D.
sulfoxide,
and
D.
sulfone)
as
these
degradates
were
included
in
the
HED
human
risk
assessment.
The
EDWCs
considered
by
HED
in
the
human
risk
characterization
(i.
e.,
surface
water
peak,
annual
mean,
and
long
term
mean)
are
summarized
by
crop
and
management
practice
in
Table
1.
The
parent
disulfoton
and
TDR
2
concentrations
for
the
cotton
scenario
are
also
given
using
the
default
"for
all
agricultural
"
crops
(0.87)
PCA
factor,
because
all
uses
have
not
been
modeled.
The
cotton
scenario,
with
the
0.87
PCA,
is
given
as
to
represent
an
upper
end
exposure
site.

Table
1.
Tier
II
Estimated
Drinking
Water
Concentrations
(EDWC)
for
parent
disulfoton
and
total
disulfoton
residues
(sum
of
parent
disulfoton
and
D.
sulfone
and
D.
Sulfoxide)
by
crop
in
Index
Reservoir
and
PCA
1
.

Crop
(PCA)
Application
Rate
(lb
ai/
A)
and
Method
Number
of
Applications/
Interval
(days)
Concentration
(
:
g/
L)

Peak
parent/
total
Annual
Average
parent/
total
Long­
term
Average
parent/
total
Barley
(0.87)
1.0
foliar
2
15.51/
34.53
1.61/
7.62
0.95/
4.21
Barley
(0.87)
0.83
granular,
soil
applied
2
12.96/
33.96
1.06/
8.70
0.44/
4.72
Barley
(0.87)
1.0
2
granular,
soil
applied
2
14.88/
39.05
1.22/
10.01
0.51/
5.42
Cotton
(0.20)
1.0
ground
and
soil
1
7.
21/
12.59
0.40/
1.96
0.12/
1.05
Cotton
(0.87)
3
1.0
ground
and
soil
1
31.35/
54.76
1.73/
8.53
0.52/
4.57
Potatoes
(0.87)
Western
states
3.0
ground
and
soil
1
6.
89/
12.53
0.46/
4.77
0.39/
3.71
Potatoes
(0.87)
East
of
Rockies
1.0
foliar
3
13.09/
34.37
1.09/
16.72
0.94/
9.49
Spring
Wheat
(0.56)
0.75
foliar
1
2.
79/
8.02
0.24/
2.39
0.19/
1.82
1
PCA
=
0.87
for
barley
and
potatoes,
0.56
for
spring
wheat,
0.20
for
cotton,
and
0.87
all
agricultural
land.
One
cotton
scenario
the
PCA
is
assumed
to
be
0.87.
2
Barley
with
1.0
lb
ai/
A
application
has
an
EDWC
that
is
(1.0/
0.87)
larger
than
the
0.87
lb
ai/
A
application.
3
The
Agency
default
PCA
for
cotton
is
0.20.
Cotton
without
an
adjustment
for
PCA,
or
0.87,
would
overestimated
the
EDWC.

In
Table
1,
the
peak
surface
water
EDWC
represents
the
upper
1­
in­
10­
year
peak
event
concentration,
the
surface
water
annual
EDWC
represents
the
upper
1­
in­
10
year
mean
annual
concentration,
and
the
long
term
EDWC
is
the
overall
average
for
the
entire
simulation.
The
EDWCs
are
based
on
linked
PRZM
and
EXAMS
models,
with
the
Percent
Crop
Area
(PCA)
and
Index
Reservoir
(IR),
while
using
the
label
maximum
rate
and
number
of
applications,
and
the
3
shortest
re­
application
interval.
Note,
the
cotton
is
shown
with
a
PCA
of
0.87
to
represent
an
upper
end
exposure
site.
SCI­
GROW
and
monitoring
data
was
used
to
estimate
ground
water
concentrations
(Table
2).

Table
2.
Ground­
water
concentrations.

Method
Application
rate/
number
1
Concentration
(
:
g/
L)

Parent
disulfoton
Total
Disulfoton
Residues
SCI­
GROW
(modeled)
3.0
lb
ai/
ac/
1
0.02
1.19
Monitoring
Mostaghimi,
1998
mean
=
0.39
2
(range
0.04
to
2.87)
not
measured
"
NAWQA
0.010
to
0.060
not
measured
"
WI
DNR
4.0
to
100.0
not
measured
1
Cotton
use
rate,
maximum
use
rate
for
major
crops.
2
Overall
mean
of
data.

The
PRZM/
EXAMS
estimated
disulfoton
residue
concentrations
in
surface
water
appear
to
be
strongly
related
to
the
application
rate,
number
of
applications,
application
interval,
and
method
of
application
and
timing
to
application
to
rainfall
events.
Everything
else
being
equal,
several
management
factors
had
an
effect
on
the
estimated
drinking
water
concentrations.
One
factor
was
the
application
rate,
a
second
was
the
application
method
(spray
drift
and
depth
of
incorporation).
A
third
factor
was
the
percent
cropped
area
(PCA).
These
values
were
set
as
presented
to
the
SAP
(Jones
and
Abel,
1997;
Jones
et
al.,
2000).
The
application
rate
and
PCA
result
in
linear
differences
in
estimated
concentrations
(e.
g.,
double
the
rate
the
EDWC
doubles).
Increasing
the
depth
of
incorporation
also
lowered
the
EDWCs.
Method
of
application
and
formulation
also
influenced
the
EDWCs;
the
granular
formulation
essentially
has
no
spray
drift
while
for
aerial
applications
drift
can
be
significant.

Summary
and
Conclusions:

The
models
used
by
EFED
show
that
disulfoton
and
degradates
will
be
found
in
runoff.
Non­
targeted
monitoring
has
found
disulfoton
and
disulfoton
degradates
in
surface
water.
However,
the
fate
of
disulfoton
and
its
degradates
once
in
surface
water
and
sediments,
and
the
likely
concentrations
therein,
cannot
be
modeled
with
a
high
degree
of
certainty
since
data
are
not
available
for
the
aerobic
and
anaerobic
aquatic
degradation
rates.
Surface
water
concentrations
of
disulfoton
and
total
disulfoton
residues
were
estimated
by
using
linked
PRZM3
and
EXAMS
models
using
several
different
scenarios
(barley,
cotton,
potato,
and
spring
wheat).
The
large
degree
of
latitude
available
in
the
disulfoton
labels
also
allows
for
a
wide
range
of
possible
application
rates,
total
amounts,
application
methods,
intervals
between
applications
and
4
application
date(
s).
The
relatively
rapid
rate
of
microbial
degradation
in
the
soil
(<
20
day
aerobic
soil
metabolism
half­
life)
and
direct
aquatic
photolysis,
suggests
that
disulfoton
parent
can
degrade
fairly
rapidly
in
surface
water.
Although
there
is
a
lack
of
some
environmental
fate
data
for
the
degradates,
the
assessment
suggests
that
the
degradates
will
reach
higher
concentrations
than
the
parent
because
they
are
more
persistent
and
probably
more
mobile.

The
estimated
drinking
water
concentrations
(EDWC)
for
parent
disulfoton
and
total
disulfoton
residues
for
different
crops
were
determined
using
the
IR
and
PCA
concepts
(Table
1).
The
peak
concentrations
of
disulfoton
in
the
IR
appear
capable
of
being
quite
high,
with
1­
year­
in
10
peak
surface
water
concentrations
of
2.79
to
15.51
µg/
L
and
annual
mean
concentrations
of
0.24
to
1.61
:
g/
L
for
the
parent
compound.
The
mean
EDWCs
of
the
annual
means
of
disulfoton
ranged
from
0.12
to
0.95
:
g/
L.
Although,
there
is
a
lack
of
some
environmental
fate
data
for
the
degradates,
the
assessment
suggests
that
the
degradates
will
reach
higher
concentrations
than
the
parent
because
they
are
more
persistent
and
probably
more
mobile.
The
estimated
1­
in­
10
year
peak
concentrations
for
the
total
disulfoton
residues
in
the
IR
ranged
from
8.02
to
34.53
:
g/
L
and
annual
mean
ranged
from
1.96
to
16.72
:
g/
L,
and
the
mean
of
the
annual
means
ranged
from
1.05
to
9.49
:
g/
L.
These
estimated
concentrations
were
highly
influenced
by
the
value
PCA
value.
The
PCA
values
have
been
estimated
by
OPP
for
spring
wheat
(0.56)
and
cotton
(0.20).
The
default
for
value
for
all
agricultural
land
of
0.87
was
used
for
the
barley
and
potatoes
scenarios.
Better
estimates
of
the
PCA
for
these
crops
would
reduce
the
uncertainty
associated
with
the
estimated
drinking
water
concentrations.

The
parent
disulfoton
and
TDR
concentrations
for
the
cotton
scenario
are
also
given
with
all
agricultural
land
PCA
factor
being
factored
in
for
one
of
the
cotton
scenarios
(Table
1).
Because
all
uses
have
not
been
modeled,
the
cotton
scenario,
the
all
agricultural
land
PCA,
is
given
as
to
represent
an
upper
end
exposure
site.

The
EDWC
values
for
disulfoton
parent
have
less
uncertainty
than
the
total
residue,
because
there
is
more
certainty
surrounding
the
"estimated"
aerobic
aquatic
metabolism
half­
life
for
the
estimated
aerobic
aquatic
half­
life
for
the
total
disulfoton
residues.
It
is
recommended
that
the
Virginia
data
be
considered
in
the
"quantitative"
drinking
water
assessment
for
ground
water
exposure.
The
Wisconsin
data
should
be
noted
and
addressed
more
qualitatively.
Highly
vulnerable
areas,
such
as
the
Central
Sand
Plain,
do
not
represent
the
entire
use
area
and
can
probably
be
better
mitigated
or
managed
a
local
or
state
level.
Specifically,
it
is
recommended
that
the
1.2
:
g/
L
be
used
for
acute
and
chronic
exposure
from
ground
water
(see
Table
2).
Based
upon
the
fate
properties
of
disulfoton,
the
sulfoxide
and
sulfone
degradates
(more
persistent
and
probably
more
mobile)
have
a
greater
probability
of
being
found
in
ground
water.
The
Agency
has
requested
more
data
on
the
mobility
and
persistence
of
the
disulfoton
sulfone
and
sulfoxide
degradates.
Depending
upon
the
results
of
the
mobility
studies,
a
ground
water
study
(ies)
may
be
required
to
better
assess
the
potential
exposure
from
the
degradates
(and
also
parent).

Monitoring
Data
5
Surface­
water
samples
were
collected
in
a
study
to
evaluate
the
effectiveness
of
Best
Management
Practices
(BMP)
in
a
Virginia
watershed.
Approximately
half
of
the
watershed
is
in
agriculture
and
the
other
half
is
forested.
The
detections
of
parent
disulfoton
in
surface­
water
samples
ranged
from
0.037
to
6.11
:
g/
L
and
fell
within
an
order
of
magnitude
with
the
estimated
environmental
concentrations
(EECs)
obtained
from
the
PRZM/
EXAMS
models.

The
surface­
water
monitoring
in
the
USGS
in
the
NAWQA
(USGS,
1998)
project
found
relatively
few
detections
of
parent
disulfoton
in
surface
water
with
a
maximum
concentration
of
0.060
:
g/
L.
Degradates
were
not
included
in
the
NAWQA
study.
In
a
separate
study,
disulfoton
degradates
were,
however,
reported
in
surface
water,
when
a
rainfall
event
occurred
following
application
to
wheat,
where
fish
kills
occurred;
pesticide
residue
concentrations
ranged
from
29.5
to
48.7
:
g/
L
for
D.
sulfoxide
and
0.02
to
0.214
:
g/
L
for
D.
sulfone
(Incident
Report
No.
I001167­
001).
The
wheat
field
was
located
several
miles
from
the
pond.
The
volume
of
run
off
water
raised
the
level
of
the
pond
fifteen
feet.
The
PRZM/
EXAMS
estimates
of
peak
TDR
correspond
fairly
well
with
the
levels
noted
above
in
the
fish
kill
incident.

EFED
also
made
inquires
to
all
fifty
states
concerning
the
whether
there
had
been
any
monitoring
for
organophosphates
pesticides
in
ground
water
and
surface
water
as
part
of
the
cumulative
assessment.
(http://
www.
epa.
gov/
pesticides/
cumulative/
pra­
op/
iii_
e_
3­
f.
pdf).
The
following
states
conducted
monitoring
which
included
parent
disulfoton:
HI,
KS,
KT,
MD,
MI,
NE,
NC,
WV,
WI,
and
WY.
There
were
no
detections
reported.

Surface­
and
ground­
water
monitoring
data
available
in
STORET
were
evaluated
in
detail,
but
were
generally
not
considered
due
to
limitations
associated
with
high
detection
limits
and
difficulty
in
interpreting
the
data.
About
50
percent
of
the
well
samples
reported
in
STORET
as
<1
:
g/
L
(low
range)
of
disulfoton
residues
and
the
other
50%
were
reported
as
<
250
:
g/
L
(high
range).
Therefore,
the
specific
concentration
of
the
well
is
not
indicated.
The
low
range
wells
could
have
concentrations
from
zero
to
0.99
:
g/
L),
while
the
high
range
could
have
concentrations
from
zero
to
249.99
:
g/
L.
Disulfoton
concentrations
were
simply
given
as
less
than
a
value.
Thus,
considerable
uncertainty
exists
with
respect
to
the
STORET
monitoring
data.
A
pilot
reservoir
monitoring
study
was
initiated
by
USEPA
Office
of
Pesticide
Programs,
Environmental
Fate
and
Effects
Division
(USEPA/
EFED/
OPP),
USEPA
Office
of
Ground
Water
and
Drinking
Water
(USEPA/
ODWGW/
OPP),
and
the
USGS
National
Water
Quality
Assessment
(USGS/
NAWQA)
to
assess
pesticide
concentrations
in
raw
and
finished
drinking
water.
(http://
www.
epa.
gov/
pesticides/
cumulative/
pra­
op/
iii_
e_
3­
f.
pdf).
Disulfoton,
and
disulfoton
sulfone
and
disulfoton
sulfoxide
were
included
in
the
study.
Parent
disulfoton
was
not
detected
(limit
of
detection
=
0.005
:
g/
L).
Degradates
disulfoton
sulfone
were
detected
(0.013
:
g/
L)
in
1
of
316
samples
(LOD
=
0.005
:
g/
L)
and
disulfoton
sulfoxide
(0.006
:
g/
L)
in
1
of
316
samples(
LOD
=
0.016
:
g/
L).
While
this
pilot
study
does
not
allow
for
a
definitive
assessment
of
potential
disulfoton
residues
in
drinking
water,
it
does
show
that
the
degradates
can
be
found
in
drinking
water
sources.
No
disulfoton
residues
were
detected
in
the
finished
water
samples.
More
detail
can
be
obtained
from
the
draft
Cumulative
Risk
Assessment
for
Organophosphates.
6
Water
Resources
Assessment
i.
Summary
and
Conclusions
The
Tier
II
modeling
of
disulfoton
residue
concentrations
in
surface
water
used
the
PRZM3
and
EXAMS
models
as
applied
to
barley,
cotton,
potatoes,
and
spring
wheat,
using
maximum
label
application
rates
and
several
application
methods
(Table
1).
The
Tier
II
EEC
assessment
uses
a
single
site,
or
multiple
single
sites,
over
multiple
years
which
represents
a
high­
end
exposure
scenario
from
pesticide
use
on
a
particular
crop
or
non­
crop
use
site
for
ecological
exposure
assessments.
The
EECs
for
disulfoton
were
generated
for
multiple
crop
scenarios
using
PRZM3.12
(Carsel,
1997;
5/
7/
98)
which
simulates
the
erosion
and
run­
off
from
an
agricultural
field
and
EXAMS
2.97.5
(Burns,
1997;
6/
13/
97)
which
simulates
the
fate
in
a
surface
water
body.
Each
scenario,
or
site,
was
simulated
for
20
to
40
(depending
on
data
availability)
years.

The
sites
selected
generally
were
the
EFED
(standard
scenarios)
to
represent
a
reasonable
"at
risk"
soil
for
the
region
or
regions
being
considered.
The
scenarios
selected
represent
highend
exposure
sites.
The
sites
are
selected
so
that
they
generate
exposures
larger
than
for
most
sites
(about
90
percent)
used
for
growing
the
selected
crops.
An
"at
risk"
soil
is
one
that
has
a
high
potential
for
run­
off
and
soil
erosion.
Thus,
these
scenarios
are
intended
to
produce
conservative
estimates
of
potential
disulfoton
concentrations
in
surface
water.
The
crop,
MLRA,
state,
site,
and
soil
conditions
for
each
scenario
are
given
in
Tables
3
and
4.

The
SCI­
GROW
(Screening
Concentration
in
Ground
Water)
screening
model
developed
in
EFED
(Barrett,
1997)
was
used
to
estimate
potential
ground
water
concentrations
for
disulfoton
parent
and
total
disulfoton
residues
under
"generic"
hydrologically
vulnerable
conditions.
SCI­
GROW
provides
a
screening
concentration,
an
estimate
of
likely
ground
water
concentrations
if
the
pesticide
is
used
at
the
maximum
allowed
label
rate
in
areas
with
ground
water
exceptionally
vulnerable
to
contamination.
In
most
cases,
a
majority
of
the
use
area
will
have
ground
water
that
is
less
vulnerable
to
contamination
than
the
areas
used
to
derive
the
SCIGROW
estimate.

ii.
Application
Rates
Used
in
Modeling
Disulfoton
application
rates
(Table
1)
selected
for
use
in
the
modeling
scenarios
were
based
upon
information
submitted
by
the
registrant,
analysis
conducted
by
BEAD,
and
the
disulfoton
(Di­
Syston)
labels.
Three
factors
were
considered
when
selecting
the
application
rate:
1)
the
labels
range
of
allowable
application
rates;
2)
the
number
of
applications;
and
3)
the
application
interval.
The
maximum
rate
(ounces
or
pounds
a.
i.
per
crop
simulated),
maximum
number
of
applications,
and
the
shortest
application
intervals
were
selected.

iii.
Modeling
Scenarios
7
Surface
Water:
The
disulfoton
scenarios
(Tables
3
and
4)
are
representative
of
high
run­
off
sites
for
barley
in
the
Southern
Piedmont
of
Virginia
(MLRA
136),
cotton
in
the
Southern
Mississippi
Valley
Silty
Uplands
of
Mississippi
(MLRA
134),
potatoes
in
the
New
England
and
Eastern
New
York
Upland
of
Maine
(MLRA
144A),
and
spring
wheat
in
the
Rolling
Till
Prairie
of
South
Dakota
(MLRA
102A).
The
wheat
scenario
was
selected
because
of
high
disulfoton
use
on
wheat
in
South
Dakota
was
high.
Soils
property
data
(Table
4)
and
planting
date
information
were
obtained
from
the
EFED
Standard
Scenarios
or
the
PRZM
Input
Collator
(PIC)
data
bases
(Bird
et
al,
1992).
The
Percent
Crop
Area
(PCA)
values
used
for
the
four
scenarios
for
estimated
drinking
water
concentrations
are
also
given
in
Table
3.
8
Table
3.
Crop,
location,
soil
and
hydrologic
group
for
each
modeling
scenario.

Crop
MLRA
1
State
Soil
Series
Soil
Texture
Hydrologic
Group
Period
(Years)
PCA
2
Barley
136
VA
Gaston
sandy
clay
loam
C
270.
87
Cotton
131
3
MS
Loring
silt
loam
C
20
0.
20
Potatoes
144A
ME
Paxton
sandy
loam
C
36
0.
87
Spr.
Wheat
102A
SD
Peever
clay
loam
C
40
0.
56
1
MLRA
is
major
land
resource
area
(USDA,
1981).
2
PCA
is
the
Percent
Crop
Area.
3
Meteorological
file
met131.
met
is
used
in
the
EFED
standard
cotton
scenario,
since
the
weather
station
is
closer
to
the
simulated
site
then
met134.
met.

Table
4.
Selected
soil
properties
used
modeling.

Soil
Series
(MLRA)
Depth
(in)
Bulk
Density
(g/
cm
3
)
Organic
Carbon
(%)
Field
Capacity
(cm
3
/cm
3
)
Wilting
Point
(cm
3
/cm
3
)

Gaston
(136)
16
1.6
1.
740
0.246
0.126
84
1.6
0.
174
0.321
0.201
50
1.6
0.
116
0.222
0.122
Loring
(131)
10
1.6
1.
160
0.294
0.094
10
1.6
1.
160
0.294
0.094
105
1.8
0.
174
0.147
0.087
Paxton
(144A)
20
1.6
2.
90
0.
166
0.66
46
1.8
0.
174
0.118
0.38
34
1.8
0.
116
0.085
0.035
Peever
(102A)
18
1.35
1.740
0.392
0.202
82
1.60
0.116
0.257
0.177
50
1.60
0.058
0.256
0.176
Ground
Water:
The
SCI­
GROW
(Screening
Concentration
in
Ground
Water)
screening
model
developed
in
EFED
(Barrett,
1997)
was
used
to
estimate
potential
ground
water
concentrations
for
disulfoton
parent
and
total
disulfoton
residues
under
"generic"
hydrologically
vulnerable
9
conditions,
but
not
necessarily
the
most
vulnerable
conditions.
The
SCI­
GROW
model
is
based
on
scaled
ground
water
concentrations
from
ground
water
monitoring
studies,
environmental
fate
properties
(aerobic
soil
half­
lives
and
organic
carbon
partitioning
coefficients­
Koc's)
and
application
rates.

iv.
Modeling
Procedure
Environmental
fate
parameters
used
in
PRZM3
and
EXAMS
runs
are
summarized
in
Table
5.
A
site
specific
Index
Reservoir
was
used
for
each
scenario.
The
PRZM3
simulations
were
run
for
a
period
of
36
years
on
potatoes,
beginning
on
January
1,
1948
and
ending
on
December
31,
1983.
Barley
was
run
for
27
years
(1956­
1983)
and
spring
wheat
was
run
for
40
years
(1944­
1983).
Cotton
was
run
for
20
years
of
data
(January
1,
1964­
December
31,
1983).
Scenario
information
is
summarized
in
Tables
4
and
5.
10
Table
5.
Disulfoton
fate
properties
and
values
used
in
(PRZM3/
EXAMS)
modeling.

Parameter
Value
Source
Molecular
Weight
274.39
MRID
150088
Water
Solubility
15
mg/
l
@20
MRID
150088
Henry's
Law
Coefficient
2.
60
atm­
m3/
mol
EFED
One­
liner
05/
21/
97
Partition
Coefficient
(Koc)
551.5
(mean
of
4
)
MRID
43042500
Vapor
Pressure
1.8E­
04
mmHg
EFED
One­
liner
05/
21/
97
Hydrolysis
Half­
lives
@
pH
4
pH
7
pH
9
1174
days
323
"
231
"
MRID
143405
Hydrolysis
Rate
Constants
(needed
for
EXAMS
derived
from
Hydrolysis
halflives
Kah
=
(negative)
Knh
=
8.88E­
05
Kbh
=
3.58
Aerobic
Soil
Half­
life
(Disulfoton)
6.12
days
(0.113/
d)
Upper
90%
confidence
bound
on
the
mean
of
"half­
lives"
for
the
two
aerobic
soils
tested
in
the
laboratory.
MRIDs
40042201,
41585101,
43800101
Aerobic
Soil
Half­
life
1
(Total
Disulfoton
Residues)
259.63
days
(2.67E­
03/
d)
Upper
90%
confidence
bound
on
the
mean
of
half­
lives
for
the
two
aerobic
soils
tested
in
the
laboratory.
MRIDs
40042201,
41585101,
43800101
Water
Photolysis
3.87
days
(pH
=
5)
(0.179/
d)
MRID
40471102
Aerobic
Aquatic
Half­
life
(Disulfoton)
(Kbaws,
Kbacs)
12.2
days
(0.05682/
day)
Estimated
per
EFED
guidance
Aerobic
Aquatic
Half­
life
(Total
Disulfoton
Residues)
(Kbaws,
Kbacs)
259.63
days
(2.67E­
03/
d)
Did
not
multiple
half­
life
by
2
per
EFED
guidance
to
account
for
uncertainty.
Half­
lives
greater
than
a
year
would
show
residue
accumulation.

Foliar
Dissipation
Rate
3.3
days
(0.21/
d)
MRID
41201801
1
Half­
lives
for
total
residues
were
determined
from
the
total
residues
at
each
sampling
interval.
Total
disulfoton
residues
did
follow
first­
order
kinetic
decay
(The
slope
(decay
rate
constant,
k)
of
the
transformed
(natural
log
or
ln)
(ln
C(
t)
=
ln
Co
­
kt,
where
Co
is
the
initial
concentration,
C
is
concentration,
and
t
is
time)
).
11
A.
Surface
Water
Drinking
Water
Assessment
with
Percent
Crop
Area
and
Index
Reservoir.

The
estimated
drinking
water
concentrations
(EDWCs)
were
evaluated
using
the
methodology
outlined
in
EPA­
OPP
draft
Guidance
for
Use
of
the
Index
Reservoir
and
Percent
Crop
Area
Factor
in
Drinking
Water
Exposure
Assessments
(USEPA,
2000).
This
generally
results
in
the
modification
of
the
scenarios
developed
for
farm
ponds
to
scenarios
for
the
index
reservoirs.

The
purpose
the
Index
Reservoir
(IR)
scenario
and
the
Percent
Crop
Area
(PCA)
for
use
in
estimating
the
exposure
in
drinking
water
derived
from
vulnerable
surface
water
supplies.
Since
the
passage
of
the
Food
Quality
Protection
Act
(FQPA)
in
1997,
the
Agency
has
been
using
the
standard
farm
pond
as
an
interim
scenario
for
drinking
water
exposure
and
has
been
assuming
that
100%
of
this
small
watershed
is
planted
in
a
single
crop.
The
Agency
is
now
implementing
the
index
reservoir
to
represent
a
watershed
prone
to
generating
high
pesticide
concentrations
that
is
capable
of
supporting
a
drinking
water
facility
in
conjunction
with
the
percent
cropped
area
(PCA)
which
accounts
for
the
fact
that
a
watershed
large
enough
to
support
a
drinking
water
facility
will
not
usually
be
planted
completely
to
a
single
crop.
These
two
steps
are
intended
to
improve
the
quality
and
accuracy
of
the
drinking
water
exposure
for
pesticides
obtained
by
models.

Percent
Crop
Area
(PCA):
PCA
is
a
generic
watershed­
based
adjustment
factor
that
will
be
applied
to
pesticide
concentrations
estimated
for
the
surface
water
component
of
the
drinking
water
exposure
assessment
using
PRZM/
EXAMS
with
the
index
reservoir
scenario.
The
output
generated
by
the
linked
PRZM/
EXAMS
models
is
multiplied
by
the
maximum
percent
of
crop
area
(PCA)
in
any
watershed
(expressed
as
a
decimal)
generated
for
the
crop
or
crops
of
interest.
Currently,
OPP
has
PCA
adjustments
for
four
major
crops
–
corn,
cotton,
soybeans,
and
wheat.
Two
are
appropriate
for
disulfoton,
cotton
and
wheat.

The
concept
of
a
factor
to
adjust
the
concentrations
reported
from
modeling
to
account
for
land
use
was
first
proposed
in
a
presentation
to
the
SAP
in
December
1997
(Jones
and
Abel,
1997).
This
guidance
results
from
a
May
1999
presentation
to
the
FIFRA
Scientific
Advisory
Panel
(SAP),
Proposed
Methods
For
Determining
Watershed­
derived
Percent
Crop
Areas
And
Considerations
For
Applying
Crop
Area
Adjustments
to
Surface
Water
Screening
Models,
and
the
response
and
recommendations
from
the
panel.
A
more
thorough
discussion
of
this
method
and
comparisons
of
monitoring
and
modeling
results
for
selected
pesticide/
crop/
site
combinations
is
located
at:
http://
www.
epa.
gov/
scipoly/
sap/
1999/
may/
pca_
sap.
pdf.

The
Agency
will
continue
to
develop
PCAs
for
other
major
crops
in
the
same
manner
as
was
described
in
the
May
1999
SAP
presentation.
However,
the
Agency
expects
that
it
will
use
smaller
watersheds
for
these
calculations
in
the
near
future.
For
minor­
use
crops,
the
SAP
found
that
the
use
of
PCAs
produced
less
than
satisfactory
results
and
advised
OPP
to
further
investigate
possible
sources
of
error.
Thus,
for
the
near
term,
OPP
is
not
be
using
PCAs
in
a
crop­
specific
manner
for
both
major
crops
that
do
not
yet
have
PCAs
and
minor­
use
crops.
12
Instead
it
will
use
a
default
PCA
that
reflects
the
total
agricultural
land
in
an
8­
digit
Hydrologic
Unit
Code
(HUC).
The
PCA
values
used
in
this
assessment
are
listed
in
Appendix
VII.
The
OPP
guidance
document
provides
information
on
when
and
how
to
apply
the
PCA
to
model
estimates,
describes
the
methods
used
to
derive
the
PCA,
discusses
some
of
the
assumptions
and
limitations
with
the
process,
and
spells
out
the
next
steps
in
expanding
the
PCA
implementation
beyond
the
initial
crops.
Instructions
for
using
the
index
reservoir
and
PCA
are
provided
below.
Discussion
on
some
of
the
assumptions
and
limitations
for
both
the
PCA
and
Index
Reservoir
are
included
in
the
Reporting
section.
One
should
note
that
there
is
an
entry
for
`All
Agricultural
Land'
in
Appendix
VII.
This
is
a
default
value
to
use
for
crops
for
which
no
specific
PCA
is
available.
It
represents
the
largest
amount
of
land
in
agricultural
production
in
any
8­
digit
hydrologic
unit
code
(HUC)
watershed
in
the
continental
United
States.

The
unadjusted
EDWC
(PRZM/
EXAMS
output)
is
multiplied
by
the
appropriate
PCA
for
that
crop
to
obtain
the
final
estimated
drinking
water
concentration
(EDWC).
Note
that
if
Tier
2
modeling
is
done
for
an
area
other
than
the
standard
scenario,
the
PCA
would
still
be
applied,
since
it
represents
the
maximum
percent
crop
area
for
that
particular
crop.
(As
regional
modeling
efforts
are
expanded,
regional
PCAs
could
be
developed
in
the
future.)
As
an
example,
for
a
pesticide
used
only
on
cotton,
the
PRZM/
EXAMS
estimated
environmental
concentrations
would
be
multiplied
by
0.20.
This
factor
would
be
applied
to
the
standard
PRZM/
EXAMS
scenario
for
cotton
or
any
non­
standard
cotton
scenario
until
such
time
as
regional
PCAs
are
developed.

When
multiple
crops
occur
in
the
watershed,
the
co­
occurrence
of
these
crops
needs
to
be
considered
(maximum
of
0.87).
The
PCA
approach
assumes
that
the
adjustment
factor
represents
the
maximum
potential
percentage
of
a
watershed
that
could
be
planted
to
a
crop
(0.87).
If,
for
example,
a
pesticide
is
only
used
on
cotton,
then
the
assumption
that
no
more
than
20%
of
the
watershed
(at
the
current
HUC
scale
used)
would
be
planted
to
the
crop
is
likely
to
hold
true.

The
Index
Reservoir
(IR):
Barley,
cotton,
potatoes,
and
spring
were
considered
because
they
represent
significant
uses,
maximum
application
rates,
and
are
grown
in
vulnerable
regions
of
the
United
States.
This
excludes
the
Christmas
tree
use,
for
which
there
is
not
a
adequate
Tier
II
scenario.
For
the
PRZM,
the
input
files
for
each
IR
scenario
are
essentially
the
same
as
its
farm
pond
scenario.
Three
parameters
in
the
PRZM
input
file
require
modification,
AFIELD,
HL,
and
DRF
(http://
www.
epa.
gov/
scipoly/
sap/
1998/
index.
htm).

The
Tier
II
modeling
results
(Table
1)
from
PRZM/
EXAMS
fall
within
the
range
of
parent
disulfoton
concentrations
for
surface
water
reported
in
a
Virginia
monitoring
study
(0.37
to
6.11
:
g/
L)
and
NAWQA
(0.010
to
0.060
:
g/
L).
The
modeled
parent
disulfoton
concentration
estimates
are
generally
greater
than
those
seen
in
the
monitoring
data.
The
modeling
results
of
the
degradates
correspond
reasonably
well
with
those
measured
at
fish
kill
incident
site,
but
were
greater
than
the
detections
in
the
pilot
reservoir
study.
The
monitoring
data
for
the
disulfoton
degradates
is
extremely
limited.
13
Uncertainty
surrounds
these
estimates
because
the
sites
selected
for
modeling
represent
sites
thought
to
be
representative
of
vulnerable
sites.
Additionally,
the
IR
was
generic
(to
each
scenario)
and
data
to
fully
understand
of
the
fate
of
disulfoton
and
disulfoton
residues
is
not
available.
Evidence
suggests
that
the
concentrations
will
not
be
as
high
as
suggest
by
the
modeled
estimates.
The
PCA
values
have
been
estimated
by
OPP
for
spring
wheat
(0.56)
and
cotton
(0.20).
The
default
for
value
for
all
agricultural
land
of
0.87
was
used
for
the
barley,
potatoes,
and
tobacco
scenarios.
Better
estimates
of
the
PCA
for
these
crops
would
reduce
the
uncertainty
associated
with
the
estimated
drinking
water
concentrations.

B.
Ground
Water
Assessment
For
this
assessment,
the
maximum
rate
and
number
of
disulfoton
applications
were
used,
while
assuming
average
environmental
properties
(90
percent
upper
confidence
bound
on
the
mean
aerobic
soil
half­
life
of
6.12
days
and
an
average
Koc
value
of
551
mL/
g).
The
maximum
parent
disulfoton
concentration
predicted
in
ground
water
by
the
SCI­
GROW
model
(using
the
maximum
rate
3
lb.
a.
i./
ac
@
1
applications
­
potatoes
or
1
lb.
ai./
ac
@
3
applications)
was
0.02
:
g/
L.
The
maximum
total
disulfoton
residue
concentration
predicted
in
ground
water
by
the
SCI­
GROW
model
is
1.19
:
g/
L
(90
percent
upper
bound
on
mean
half­
life
of
total
residues
is
259.6
days).

It
should
be
noted
that
all
the
detections
of
disulfoton
residues
in
ground
water
in
Wisconsin
(range
4.0
to
100.0
:
g/
L)
and
some
detections
in
Virginia
(range
0.04
­2.87
:
g/
L)
exceeded
the
concentrations
predicted
by
SCI­
GROW
(0.02
:
g/
L).
Although
SCI­
GROW,
which
is
thought
to
be
conservative
(e.
g.,
a
vulnerable
site),
is
based
on
a
regression
relationship
between
monitoring
data
(detected
concentrations)
and
pesticide
fate
chemistry
at
vulnerable
sites,
SCI­
GROW
does
not
account
for
preferential
flow,
point­
source
contamination,
pesticide
spills,
misuses,
or
pesticide
storage
sites.
Many
unknowns,
data
limitations,
such
as
on­
site
variability,
are
also
present
in
the
prospective
ground­
water
monitoring
studies
which
were
not
included
when
developing
SCI­
GROW.
The
difference
between
monitoring
and
modeling
is
discussed
further
in
the
next
section.

Disulfoton
Monitoring
Data
Surface
Water
Monitoring:

Virginia:
The
previously
discussed
study
to
evaluate
the
effectiveness
of
Best
Management
Practices
(BMP)
in
a
3616­
acre
watershed
in
the
Nomini
Creek
Watershed,
Westmoreland
County,
Virginia
(Mostaghimi,
1989;
Mostaghimi
et
al.
1998)
also
collected
a
limited
number
of
runoff
and
surface
water
samples
at
two
stations.
For
more
discussion
see
Ground­
Water
Section
above,
and
Appendix
1.
The
results
the
surface
water
monitoring
for
disulfoton
parent
is
presented
in
Table
6.
14
Table
6.
Disulfoton
detections
in
surface
water
samples
collected
in
the
Nomini
Creek
Watershed
(Virginia),
during
1986.

Sample
date
Site
number:
Sample
number
Concentration
(
:
g/
L)

8/
18/
86
QN1:
1
(9:
13
am)
6.11
8/
18/
86
QN1:
2
(12:
25
pm)
0.37
9/
28/
86
QN2:
(only
1
sample)
1.62
NAWQA:
Disulfoton
residues
have
been
detected
in
surface
water
at
a
low
frequency
in
the
USGS
NAWQA
study.
The
percentage
of
detections
with
disulfoton
concentrations
>0.01
:
g/
L
for
all
samples,
agricultural
streams,
urban
streams
were
0.27%,
0.20,
and
0.61%,
respectively.
The
corresponding
maximum
concentrations
were
0.060,
0.035,
and
0.037
:
g/
L.
Disulfoton
has
not
been
detected
in
ground
water
in
the
NAWQA
study.
Although
pesticide
usage
data
is
collected
for
the
different
NAWQA
study
units,
the
studies
are
not
targeted,
specifically
for
disulfoton.

STORET:
About
50
percent
of
the
well
samples
reported
in
STORET
had
low
levels
(<
1
:
g/
L)
of
disulfoton
residues.
However,
there
were
indications
of
some
high
concentrations
(the
other
50%
were
reported
as
<250
:
g/
L),
which
may
be
a
reflection
of
how
the
data
were
reported
as
the
disulfoton
concentrations
in
the
monitoring
were
not
always
known.
This
is
because
the
detection
limit
was
extremely
high
or
not
specified,
and/
or
the
limit
of
quantification
was
not
stated
or
extremely
high.
Disulfoton
concentrations
were
simply
given
as
less
than
a
value.
Therefore,
considerable
uncertainty
exists
with
respect
to
the
STORET
monitoring
data.
The
STORET
data
was
considered
only
from
a
"qualitative"
standpoint.
EFED
considered
in
the
assessment
that
while
one
does
not
know
the
concentration
in
the
wells
reported
as
disulfoton
concentrations
<
1
:
g/
L
,
you
know
do
it
is
not
>
than
1
:
g/
L.

Ground
Water
Monitoring:

Monitoring
Studies
With
No
Disulfoton
Residues
Detections
in
Ground
Water:
The
Pesticides
in
Ground
Water
Data
Base
(USEPA,
1992)
summarizes
the
results
of
a
number
of
groundwater
monitoring
studies
conducted
which
included
disulfoton
(and
rarely
the
disulfoton
degradates
D.
sulfone
and
D.
sulfoxide).
Monitoring,
with
no
detections
(limits
of
detections
ranged
from
0.01
to
6.0
:
g/
L),
has
occurred
in
the
following
states
(number
of
wells):
AL
(10),
CA
(974),
GA
(76),
HI
(5),
IN
(161),
ME
(71),
MS
(120),
MN
(754),
OK
(1),
OR
(70),
and
TX
(188).
The
range
of
detection
limits,
especially
the
high
ones
(e.
g.,
6
:
g/
L)
reduce
the
certainty
of
these
data.

One
hundred
twenty
wells
were
analyzed
in
MS
for
disulfoton
degradates
sulfone
and
sulfoxide
and
188
wells
were
analyzed
in
TX
for
sulfone.
Limits
of
detection
were
3.80
and
15
1.90
:
g/
L
for
the
sulfone
and
sulfoxide
degrade,
respectively,
in
MS.
There
were
no
degradates
reported
in
these
samples.

North
Carolina:
The
North
Carolina
Departments
of
Agriculture
(NCDA)
and
Environment,
Health,
and
Natural
Resources
(DEHNR)
conducted
a
cooperative
study
under
the
direction
of
the
North
Carolina
Pesticide
Board
(NCIWP,
1997).
The
purpose
of
the
statewide
study
was
to
determine
if
the
labeled
uses
of
pesticide
products
were
impacting
the
ground
water
resources
in
North
Carolina.

The
study
was
conducted
in
two
phases.
In
phase
one,
55
wells
in
the
DEHNR
Ground
Water
Section's
ambient
monitoring
network
representing
the
major
drinking
water
aquifers
of
the
state
were
sampled
at
least
twice
and
analyzed
for
selected
pesticides.
In
phase
two,
97
cooperator
monitoring
wells
were
installed
and
subsequently
sampled
at
least
twice
in
36
counties
across
the
North
Carolina.
Sites
for
the
cooperator
monitoring
wells
were
chosen
based
on
an
evaluation
of
the
vulnerability
of
ground
water
to
risk
of
contamination
from
the
use
of
pesticides.

Monitoring
wells
were
located
adjacent
to
and
down­
gradient
from
areas
where
pesticides
were
reported
to
have
been
applied
(within
300
feet)
during
the
previous
five
years.
Wells
were
constructed
so
that
the
shallowest
ground
water
could
be
collected
for
analysis.
The
objective
of
these
criteria
was
to
use
a
scientific
method
for
determining
monitoring
well
locations
so
that
the
results
could
be
used
as
an
early
indication
of
the
potential
for
problems
associated
with
pesticides
leaching
to
ground
water.
Disulfoton
residues
were
monitored
for
in
five
North
Carolina
counties,
Allegheny,
Ash,
Beaufort,
Madison,
and
Robeson.
Seven
wells
were
located
in
Christmas
Tree
growing
areas,
one
in
wheat
growing
county,
and
two
in
tobacco
areas.
The
study
authors
make
the
following
statement,
"Results
cannot
be
interpreted
as
representing
the
quality
of
ground
water
near
pesticide
use
areas
statewide
because
the
study
methods
targeted
areas
of
highly
vulnerable
ground
water".

There
were
no
detections
of
disulfoton,
disulfoton
sulfoxide,
and
disulfoton
in
the
ground­
water
monitoring
study
conducted
in
North
Carolina.
Efforts
were
made
to
place
the
wells
in
vulnerable
areas
where
the
pesticide
use
was
known,
so
that
the
pesticide
analyzed
for
would
reflect
the
use
history
around
the
well.
Limitations
of
the
study
include
that
sites
were
sampled
only
twice
and
the
limits
of
detections
were
high
(e.
g.,
>
1.0
:
g/
L)
for
some
of
disulfoton
analytes.
Uncertainties
associated
with
the
study
include
whether
two
samples
from
eight
wells
are
adequate
to
represent
the
ground­
water
concentrations
of
disulfoton
residues,
if
DRASTIC
correctly
identified
a
site's
vulnerability,
and
if
the
wells
were
placed
down­
gradient
of
the
use
areas.

The
study
used
tools
and
information
available
at
the
time
of
the
study
to
identify
vulnerable
locations
for
well
placement.
This
included
statewide
agricultural
data
from
the
N.
C.
Agricultural
Statistics
which
were
used
to
identify
crop
growing
areas,
the
USEPA
DRASTIC
method
(Aller
et
al.,
1987)
was
used
to
locate
the
most
vulnerable
locations
in
the
target
crop
growing
areas,
and
local
county
agents
of
the
USDA
Natural
Resources
Conservation
Service
16
(NRCS)
helped
identify
cooperators­
farmers
for
placement
of
wells.
The
Pesticide
Study
staff
and
county
agents
also
met
with
the
cooperators
to
obtain
pesticide
use
information.
Other
studies
have
shown
that
DRASTIC
is
not
as
good
a
method
to
identify
vulnerable
areas
as
hoped.
The
study
appeared
to
QA/
QC
practices.

Monitoring
Studies
With
Disulfoton
Detections
in
Ground
Water:
Two
of
the
studies
cited
in
the
PGWDB
(USEPA,
1992)
report
the
detection
of
disulfoton
residues
in
ground
water.
The
disulfoton
detections
in
ground
water
in
occurred
studies
conducted
by
Virginia
Polytechnic
Institute
and
State
University
(VPI&
SU,
Mostaghimi,
1989;
Mostaghimi
et
al.
1998)
in
Virginia
where
disulfoton
concentrations
ranged
from
0.04
to
2.87
:
g/
L
and
in
a
Wisconsin
Department
of
Natural
Resources
study
in
Wisconsin
(WDNR,
after
Barton,
1982)
where
concentrations
ranged
from
4.00
to
100.00
:
g/
L.
Of
specific
are
the
disulfoton
concentrations
of
parent
disulfoton
reported
in
these
studies
(VA
and
WI)
that
exceeded
the
estimate
of
0.02
:
g/
L
obtained
from
EFED's
SCI­
GROW
(ground­
water
screening
model)
model.

Virginia:
A
monitoring
study
was
conducted
to
evaluate
the
effectiveness
of
Best
Management
Practices
(BMP)
in
a
3616­
acre
watershed
in
the
Nomini
Creek
Watershed,
Westmoreland
County,
Virginia.
(A
more
detailed
assessment
of
the
final
report
of
Nomini
Creek
Watershed
BMP
study
was
prepared
earlier
(11/
20/
01)
in
Appendix
1).
Approximately
half
of
the
watershed
is
in
agriculture
and
the
other
half
is
forested.
The
major
focus
of
this
study
was
surface­
water
quality
rather
than
ground­
water
quality.
However,
in
addition
to
the
surfacewater
monitoring,
twelve
wells
were
analyzed
for
pesticides,
including
disulfoton.

Samples
were
initially
taken
in
1985
and
1986
from
four
household
wells
in
the
Nomini
Creek
Watershed
(NCW).
Water
samples
from
these
wells
were
analyzed
for
24
pesticides.
Detectable
levels
of
(not
specified)
pesticides
were
found
in
all
four
wells
at
concentrations
below
the
respective
MCL.
One
of
these
four
household
wells
consistently
had
higher
pesticide
levels
than
the
other
wells.
The
study
authors
suggested
that
this
household
well
was
not
"sufficiently
protected
and
was
contaminated
by
surface
runoff
from
adjacent
land".

Based
upon
these
results
of
the
four
household
wells
sampled,
eight
pairs
of
groundwater
monitoring
wells
(39
to
54
feet
deep)
were
installed
at
eight
sites
in
the
NCW
and
sampled
approximately
monthly
from
June
1986
through
June
1997.
Information
concerning
farming
practices
in
the
watershed
was
obtained
from
farmer
interviews
and
questionnaires.
Monitoring
consisted
of
two
(2)
runoff
and
surface
water
monitoring
stations;
seven
rain
gauges;
one
weather
station;
and
eight
(8)
pairs
of
ground­
water
monitoring
wells.
The
ground
water
wells
were
located
primarily
in
agricultural
areas.
These
wells
were
drilled
in
pairs,
100
­
150
meters
apart,
with
one
in
pair
located
hydraulically
down­
gradient
of
the
other.
Approximately
monthly
samples
were
taken
from
each
monitoring
well
and
analyzed
for
a
number
of
analytes
including
22
pesticides.
QA/
QC
procedures
were
followed.
The
mean
of
all
parent
disulfoton
detections
is
0.39
:
g/
L,
the
mean
of
the
pre­
BMP
is
0.52
:
g/
L,
and
post­
BMP
is
0.08
:
g/
L
(Table
7).
The
maximum
detection
was
2.87
:
g/
L
The
final
report
is
discussed
in
Appendix
1.
17
Table
7.
Ground­
water
parent
disulfoton
sampling
results
and
detection
statistics
in
the
Nomini
Creek
Watershed
(Mostaghimi,
1998).

Pesticide
Total
Samples
Detections
1
Detection
2
Frequency
(percent)
Concentration
(
:
g/
L)

Max
Mean
SD
Disulfoton
1010
10
1.0
2.87
0.39
0.32
Pre­
BMP
3
(5/
86
­
10/
88)

229
7
3.1
2.87
0.52
Post
BMP
4
(11/
89
­
9/
96)

693
3
0.4
0.10
0.08
1
Number
of
samples
with
detectable
levels
of
pesticide
2
(samples
with
detectable
levels
of
pesticide
*
100)/
total
number
of
samples
3
Before
agricultural
Best
Management
Practice
(BMP)
implemented
in
watershed.
4
Following
the
implementation
of
BMP
within
the
watershed.

Wisconsin:
Barton,
1982.
In
May
and
June
1982,
the
Wisconsin
Department
of
Natural
Resources
(WDNR)
sent
twenty­
nine
water
samples
from
wells
in
the
Central
Sands
area
of
Wisconsin
to
the
EPA's
Office
of
Pesticide
Programs
for
pesticide
residue
analysis.
Samples
were
taken
from
one
municipal
well,
two
or
three
community
wells,
and
twenty­
five
home
wells;
all
of
which
were
sources
of
drinking
water.
Of
the
29
samples,
15
samples
were
reported
as
no
detects
whereas
14
samples
were
reported
disulfoton
detections.
Disulfoton
detections
ranged
from
4.00
to
100.00
:
g/
L,
with
a
mean
(samples
with
detections)
of
38.43
:
g/
L
and
standard
deviation
of
31.56
:
g/
L.
No
detection
limit
was
specified
for
disulfoton,
although
detections
as
low
as
1
:
g/
L
are
reported
for
other
pesticide
residues
(aldicarb,
and
aldicarb
sulfone,
dinoseb,
sencor,
linuron,
carbofuran,
and
Lasso/
Bravo).

Holden
(1986)
wrote
that
the
WDNR
sampling
program
was
criticized
for
a
number
of
reasons
including
that
the
quality
assurance
and
quality
control
procedures
(QA/
QC)
were
not
always
followed
during
some
stages
of
sampling
and
analysis
(Holden,
1986).
Holden
(1986)
further
indicates
that
"Harkin
et
al.
(1984)
noted
in
their
WIS
WRC
report
Pesticides
in
Groundwater
beneath
the
Central
Sand
Plain
of
Wisconsin
that
some
detections
of
pesticides
in
initial
screening
were
false
positives
and
were
not
supported
by
resampling
and
reanalysis
by
more
sensitive
analytical
methods."

Aldicarb
and
aldicarb
sulfone
were
also
found
in
this
study
and
in
follow
up
studies,
while
disulfoton
was
apparently
not
found
in
follow­
up
sampling.
Aldicarb
is
no
longer
registered
for
use
in
Wisconsin.

The
criticisms
of
the
WDNR
study
must,
however,
be
put
in
some
sort
of
perspective.
First,
a
study
that
did
not
follow
QA/
QC
criteria
does
not
and
should
not
automatically
mean
that
18
the
data
is
bad
or
wrong,
the
detections
may
be
correct
(presence
and
wrong
magnitude).
Frequently
"older"
monitoring
studies
often
had
problems
associated
with
them,
such
as
QA/
QC
problems,
limited
pesticide
usage
information,
and
no
knowledge
about
the
study
area's
hydrology.
Frequently,
studies
with
QA/
QC
programs
are
poorly
designed,
so
that
the
results
may
be
meaningless.
These
data
were
considered
in
the
water
assessment,
but
were
not
included
when
deriving
the
disulfoton
concentrations
(EDWCs)
for
human
health
assessment.

Pesticide
residues
not
being
found
in
follow­
up
sampling
may
be
the
result
of
dissipation
processes
and
should
not
be
used
to
discount
detections
in
earlier
samples.
The
environmental
fate
properties
and
site
hydrology
must
also
be
considered.
Because
ground
water
is
a
dynamic
system,
pesticides
may
be
present
at
one
sampling
event
and
not
at
another.
So
when
the
sample
is
collected,
in
relationship
to
pesticide
use
and
rainfall,
is
important.
All
that
can
be
said
is
that
residues
were
not
found
in
follow­
up
samples.
It
is
unknown
which
samples
were
re­
analyzed
with
more
sensitive
methods.

The
disulfoton
detections
in
the
Central
Sand
Plain
may
have
been
the
result
of
preferential
flow
and
transport
processes.
Literature
documents
preferential
flow
in
the
Central
Sand
Plain.
Thus,
disulfoton
residues
may
have
by­
passed
the
soil
matrix
and
gone
directly
to
ground
water
which
is
possibly
reflected
in
the
"high"
level
of
the
detections.
Although
preferential
flow
is
currently
an
ongoing
area
of
research
and
much
remains
unknown,
it
is
known
that
preferential
flow
is
influenced
by
a
number
of
factors,
including
rainfall
amounts,
intensity,
and
frequency.
Disulfoton
generally
appears
to
be
not
very
persistent
under
aerobic
soil
conditions
and
therefore
may
also
not
be
very
persistent
in
aquifers
that
are
aerobic.
Therefore
it
may
have
also
been
missed
by
utilizing
a
predetermined
sampling
schedule
(e.
g.,
monthly).
Whereas
a
persistent
chemical,
such
as
aldicarb
and
aldicarb
sulfone,
will
be
found
at
greater
frequencies
and
be
less
dependent
upon
timing
of
sampling.
Disulfoton
usage
history
before
the
detections
and
prior
to
the
follow­
up
sampling
is
not
specified.

Additional
Monitoring
Without
Disulfoton
Detections
OPP's
EFED
contacted
individuals
in
nearly
all
the
states
concerning
whether
organophosphate
(OP)
pesticides
had
been
sampled
for
in
their
state
as
part
of
the
cumulative
OP
assessment
(http://
www.
epa.
gov/
pesticides/
cumulative/
pra­
op/
iii_
e_
3­
f.
pdf).
The
following
presents
the
survey
of
states
conducted
by
EFED
for
the
cumulative
OP
assessment
with
respect
to
disulfoton,
disulfoton
sulfone,
and
disulfoton
sulfoxide.
Florida,
did
not
include
disulfoton
or
disulfoton
degradates,
but
is
included
because
the
sulfone
and
sulfoxide
degradates
of
the
OP
fenamiphos
was
included.
EFED
has
not
evaluate
whether
any
of
the
monitoring
studies
noted
below
were
also
included
in
those
reported
in
the
PGWDB
(USEPA,
1992).

Florida
Over
20,000
"determinations"
were
made
for
OPs
in
several
Florida
ground­
water
monitoring
programs.
Only
wells
with
detections
are
reported,
both
for
fenamiphos
sulfoxide.
19
This
is
a
large
dataset,
and
it
will
require
more
conversations
with
Florida
to
understand
the
full
significance
of
these
data.

Continued
monitoring
in
the
Lake
Ridge
monitoring
program
includes
fenamiphos
and
transformation
products
fenamiphos
sulfoxide
and
sulfone.
These
have
not
been
detected
in
quarterly
sampling
of
monitoring
wells
in
11
to
33
wells
over
the
last
two
years.

Hawaii
Robert
Boesch
of
the
Department
of
Agriculture
described
a
drinking­
water
study
conducted
this
past
March.
In
preparation
for
the
OP
risk
assessment,
Hawaii
sampled
36
drinking­
water
wells
in
areas
where
OPs
are
used
on
pineapples,
or
for
urban
use.
These
water
supply
wells,
which
have
shown
contamination
for
other
organic
chemicals,
did
not
have
detections
(LOD
0.5
ppb)
of
the
following
OPs:
acephate,
azinphos
methyl,
chlorpyrifos,
DDVP,
demeton,
diazinon,
dimethoate,
disulfoton,
ethoprop,
fenamiphos,
malathion,
methidation,
methyl
parathion,
mevinphos,
monocrotophos,
naled
and
parathion.

Kansas
Theresa
Hodges
of
the
Kentucky
Department
of
Health
and
Environment
reports
that
of
the
OPs,
only
diazinon
has
been
detected
in
their
routine
ambient
surface
water
quality
sampling
network.
While
diazinon
is
not
on
the
list
of
pesticides
routinely
included,
it
was
added
because
it
had
been
detected.
Since
1995,
44
detections
were
found
at
16
urban
or
golf
course
sites.
The
range
of
detections
was
from
0.19
to
1.5
micrograms/
liter.

Dale
Lambley
of
the
Kansas
Department
of
Agriculture
sent
information
on
their
ground­
water
monitoring
of
chemigation
wells.
The
objective
of
the
study
"is
to
assess
and
monitor
groundwater
quality
by
obtaining
water
samples
at
selected
chemigation
sites
located
at
agricultural
irrigation
wells."
In
sampling
from
1987
to
2000,
chlorpyrifos
was
detected
three
times
at
concentrations
of
1.9,
3.5
and
4.2
ppb
(LOD
=
0.5
:
g/
l).
Dimethoate,
disulfoton
and
methyl
parathion
were
included
in
sampling,
but
were
not
detected
above
detection
levels
of
2.0,
0.5
and
1.0
:
g/
l,
respectively.

The
100
samples
taken
annually
are
apportioned
among
five
Groundwater
Management
Districts
based
on
the
number
of
registered
chemigation
sites
in
each.
Highest
priority
is
given
to
finding
active
chemigation
sites.
Ranking
of
wells
has
also
been
based
on
proximity
to
public
water
supplies
(within
3
miles),
depth
to
water,
soil
type,
and
whether
chemigation
misuse
is
suspected.

Kentucky
Peter
Goodman
reports
that
the
following
OPs
are
included
in
their
ground­
water
monitoring
program:
acephate,
chlorpyrifos,
diazinon,
disulfoton,
ethoprop,
malathion,
methyl
parathion
and
20
terbufos.
Each
was
included
in
more
than
1300
analyses
from
over
300
wells,
but
only
diazinon,
chlorpyrifos
and
malathion
were
detected.

Maryland
Rob
Hofstedter
of
the
Maryland
Department
of
Agriculture
reports
that
their
agency
has
a
current
ground­
water
study
that
includes
diazinon.
Results
of
this
study
are
not
yet
available.
He
referred
me
to
the
Maryland
Geological
Survey
for
information
on
previous
surface­
water
studies
which
included
malathion.

David
Bolton
of
the
Maryland
Geological
Survey
provided
summary
tables
from
the
MGS
Report
of
Investigations
number
66,
"Ground­
Water
Quality
in
the
Piedmont
Region
of
Baltimore
County,
Maryland."
Analysis
in
this
rural
region
included
12
OPs,
10
of
which
are
still
registered.
Disulfoton
was
not
detected
in
ground
water.
Results
of
the
monitoring
are
as
follows,
which
concentrations
in
:
g/
l.

Pesticide
#
samples
MRL
1
>/=
MRL
<MRL
Maximum
Conc.

Disulfoton
112
0.017
0
0
1
MRL
=
Minimum
Reporting
Limit
Michigan
Mark
Breithart
of
the
MDEQ
Drinking
Water
Division
examined
their
database,
and
found
that
analysis
was
done
for
the
following
OPs
in
Michigan
drinking
water:
azinphos
methyl,
chlorpyrifos,
diazinon,
dimethoate,
disulfoton,
fenamiphos,
malathion,
and
methyl
parathion
None
of
these
were
detected
in
49
analyses
of
public
water
supplies.
Of
the
421
analyses
from
private
water
supplies,
only
dimethoate
was
detected.
This
single
detection
of
2
micrograms/
liter
occurred
at
an
aerial
spray
service,
and
therefore
it
is
not
clear
if
it
was
the
result
of
a
point
source.

Nebraska
Nebraska
maintains
the
"Quality­
Assessed
Agricultural
Contaminant
Database
for
Nebraska
Ground
Water,"
which
was
created
from
ground
water
quality
data
submitted
by
many
organizations."
There
were
no
disulfoton
detections
in
185
analyses.
The
following
OPs
are
included
in
the
database:
chlorpyrifos,
diazinon,
ethion,
malathion,
methyl
parathion,
phorate,
and
terbufos.
The
levels
of
detection
are
generally
below
1
ppb.

Mr.
John
Lund,
supervisor
in
the
Surface
Water
Unit
of
the
Nebraska
Department
of
Environmental
Quality,
indicated
that
OPs
have
not
been
included
in
the
State's
surface­
water
monitoring.
21
North
Carolina
Dr.
Henry
Wade
described
the
"Interagency
Study
of
the
Impact
of
Pesticide
Use
on
Ground
Water
in
North
Carolina,"
which
took
place
between
1991
and
1995.
Sampling
of
mostly
shallow
monitoring
wells
was
performed
based
on
information
by
farmers
on
which
pesticides
they
used
within
300
feet
of
the
wells.
By
the
end
of
the
study,
more
than
240
pesticides
were
included
as
analytes.

Sixteen
OPs
were
included
in
the
analysis,
but
none
were
detected.
The
number
of
wells
sampled
for
each
OP
is
shown
below:
acephate
(23
wells),
azinphos­
methyl
(7),
chlorpyrifos
(25),
diazinon
(8),
dimethoate
(5),
disulfoton
(12),
ethoprop
(6),
fenamiphos
(4),
fonofos
(1),
malathion
(9),
mevinphos
(1),
parathion
(5),
phorate
(3),
phosmet
(2),
terbufos
(13)
and
trichlorfon
(2).

Other
pesticides
were
detected
in
these
wells,
especially
herbicides.
The
main
focus
of
the
study
was
herbicides
which
the
EPA
had
identified
as
"potential
leachers."

West
Virginia
Doug
Hudson
of
the
WV
Department
of
Agriculture
says
that
West
Virginia
DoA
does
intermittent
ground
water
sampling,
including
an
OP
screen.
He
could
recall
only
a
single
detection
of
diazinon,
which
they
could
not
confirm.
Other
OP
detections
in
ground
water
were
in
response
to
improper
termiticide
use.

Chad
Board
of
the
DEP
sent
a
spreadsheet
with
analytical
results
which
included
the
following
OPs:
chloropyrifos,
diazinon,
disulfoton,
ethoprop,
malathion,
phorate,
and
terbufos.
Each
were
sampled
in
12
wells,
but
not
detected.
The
detection
limits
ranged
from
0.005
to
0.027
ppb.

Wisconsin
Bill
Phelps,
of
the
Wisconsin
Department
of
Natural
Resources
Bureau
of
Drinking
&
Groundwater
provided
a
summary
of
monitoring
Wisconsin
has
done
in
public
and
private
water
supply
wells
and
information
on
monitoring
from
their
GEMS
database
performed
at
regulated/
investigated
sites.
The
detections
of
disulfoton
occurred
at
a
pesticide
formulation
plant
­
thus
this
would
be
a
point
source
rather
than
non­
point
source
normal
use.

Analyte
#
Water
Supply
Wells
#
Detects
in
Water
Supply
Wells
#GEMS
wells
#
GEMS
wells
with
detections
Maximum
concentration
detected
(ug/
l)

disulfoton
0
190
9
240
Wyoming
22
Jim
Bigelow
of
the
Wyoming
Department
of
Agriculture
described
the
generic
Pesticide
Management
Plan
ground­
water
program,
which
includes
a
network
of
178
wells.
A
total
of
54
active
ingredients
are
included
as
analytes,
including
eight
active
OPs:
azinphos­
methyl,
chlorpyrifos,
diazinon,
disulfoton,
malathion,
methyl
parathion,
phorate
and
terbufos.

Ms.
Miller
indicated
that
there
have
been
detections
of
pesticides
in
117
of
178
wells.
The
Agency
will
investigate
further
details
of
this
monitoring
program.

Limitations
of
Monitoring
Data
The
interpretation
of
the
monitoring
data
is
limited
by
the
lack
of
correlation
between
sampling
dates
and
the
use
patterns
of
the
pesticide
within
the
study's
drainage
basin.
Additionally,
the
sample
locations
were
not
associated
with
actual
drinking
water
intakes
for
surface
water
nor
were
the
monitored
wells
associated
with
known
ground
water
drinking
water
sources.
Also,
due
to
many
different
analytical
detection
limits,
no
specified
detection
limits,
or
extremely
high
detection
limits,
a
detailed
interpretation
of
the
monitoring
data
is
not
always
possible.

Limitations
for
the
monitoring
studies
include
the
use
of
different
limits
of
detection
between
studies,
lack
of
information
concerning
disulfoton
use
around
sampling
sites,
and
lack
of
data
concerning
the
hydro
geology
of
the
study
sites.
The
spatial
and
temporal
relationship
between
disulfoton
use,
rainfall/
runoff
events
and
the
location
and
time
of
sampling
cannot
often
be
adequately
determined.
Thus,
it
is
not
always
possible
to
judge
the
significance
of
the
level
or
the
lack
of
detections.

Although
no
assessment
can
be
made
for
degradates
due
to
lack
of
monitoring
data,
limited
data
suggests
that
the
degradates
are
more
persistent
(>
200
days)
than
disulfoton,
suggesting
their
presence
in
water
for
a
longer
period
of
time
than
the
parent.
The
degradates
also
appear
to
be
more
mobile
than
the
parent
compound.

vii.
Limitations
of
this
Modeling
Analysis
There
are
number
of
factors
which
limit
the
accuracy
and
precision
of
this
modeling
analysis
including
the
selection
of
the
high­
end
exposure
scenarios
and
maximum
number
of
applications
and
rates,
the
quality
of
the
data,
the
ability
of
the
model
to
represent
the
real
world,
and
the
number
of
years
that
were
modeled.
There
are
additional
limitations
on
the
use
of
these
numbers
as
an
estimate
of
drinking
water
exposure.
Individual
degradation/
metabolism
products
were
also
not
considered
due
to
lack
of
data.
Another
major
uncertainty
in
the
current
EXAMS
simulations
is
that
the
aquatic
degradation
rate
used
an
estimated
rate
due
to
lack
of
data.
Direct
aquatic
photolysis
was
also
included.
The
total
disulfoton
residue
decline
rate
was
estimated
from
data,
but
Kocs
and
hydrolysis
rates
for
D.
sulfoxide
and
D.
sulfone
were
not
known
and
assumed
to
be
equal
to
those
of
parent
disulfoton.
These
limitations
influence
the
estimates
of
23
pesticides
transported
off
the
field
(loading
files)
to
the
reservoir,
plus
the
degradation
once
in
the
reservoir.

Spray
drift
is
determined
by
method
of
pesticide
application,
and
is
assumed
to
be
0%
percent
when
applied
as
broadcast
(granular)
or
in­
furrow,
and
6.4%
ground
and
16.4%
aerial
spray
for
the
Index
Reservoir
scenario
(Jones
et
al.,
2000).

The
Tier
II
scenarios
are
also
ones
that
are
likely
to
produce
high
concentrations
in
aquatic
environments.
The
scenarios
were
intended
to
represent
sites
that
actually
exist
and
are
likely
to
be
treated
with
a
pesticide.
These
sites
should
be
vulnerable
enough
to
provide
a
conservative
estimates
of
the
EDWC,
but
not
so
vulnerable
that
the
model
cannot
properly
simulate
the
fate
and
transport
processes
at
the
site.
The
EDWCs
in
this
analysis
are
accurate
only
to
the
extent
that
the
sites
represent
the
hypothetical
high
exposure
sites.

The
quality
of
the
analysis
is
also
directly
related
to
the
quality
of
the
chemical
and
fate
parameters
available
for
disulfoton.
Acceptable
data
are
available,
but
rather
limited
(minimal)
or
not
available
for
the
degradates.
Data
were
not
available
for
degradates
and
the
aquatic
aerobic
metabolism
rate
was
not
known,
but
estimated.
Degradates
with
greater
persistence
and
greater
mobility
would
be
expected
to
have
a
higher
likelihood
of
leaching
to
ground
water,
with
greater
concentrations
in
surface
water.
The
measured
aerobic
soil
metabolism
data
is
limited,
but
has
sufficient
sample
size
to
establish
an
upper
90%
confidence
bound
on
the
mean
of
halflives
for
the
three
aerobic
soils
tested
in
the
laboratory
(and
submitted
to
EFED)
and
reported
in
the
EFED
One­
liner
Database
(MRIDs
40042201,
41585101,
43800101).
The
use
of
the
90%
upper
bound
value
may
be
sufficient
to
capture
the
probable
estimated
environmental
concentration
when
limited
data
are
available.
PRZM
assumes
pesticide
decline
follows
firstorder
kinetics.
As
discussed
in
the
aerobic
soil
metabolism
section
of
the
disulfoton
RED,
disulfoton
doesn't
entirely
follow
first­
order
kinetics.

The
models
themselves
represent
a
limitation
on
the
analysis
quality.
These
models
were
not
specifically
developed
to
estimate
environmental
exposure
in
drinking
water
so
they
may
have
limitations
in
their
ability
to
estimate
drinking
water
concentrations.
Another
limitation
is
the
lack
of
field
data
to
validate
the
predicted
pesticide
run­
off.
Although,
several
of
the
algorithms
(volume
of
run­
off
water,
eroded
sediment
mass)
are
somewhat
validated
and
understood,
the
estimates
of
pesticide
transport
by
PRZM3
has
not
yet
been
fully
validated.
Other
limitations
of
PRZM
are
the
inability
to
handle
within
site
variation
(spatial
variability),
crop
growth,
and
the
overly
simple
water
balance.
Another
limitation
is
that
20
to
40
years
of
weather
data
were
available
for
the
analysis.
Consequently
there
is
a
1
in
20,
27,
36,
or
40
chance
that
the
true
10%
exceedence
EDWCs
are
larger
than
the
maximum
EDWC
in
the
analysis.
If
the
number
of
years
of
weather
data
were
increased,
it
would
increase
the
level
of
confidence
that
the
estimated
value
for
the
10%
exceedence
EDWC
was
close
to
the
true
value.

EXAMS
is
limited
because
it
is
a
steady­
state
model
and
cannot
accurately
characterize
the
dynamic
nature
of
water
flow.
A
model
with
dynamic
hydrology
would
more
accurately
reflect
concentration
changes
due
pond
overflow
and
evaporation.
Thus,
the
estimates
derived
24
from
the
current
model
simulates
a
pond
having
no­
outlets,
flowing
water,
or
turnover.
Another
major
limitation
in
the
current
EXAMS
simulations
is
that
the
aquatic
(microbial)
and
abiotic
degradation
pathways
were
adequately
considered.
Disulfoton
and
the
sulfone
and
sulfoxide
degradates
were
considered
as
total
disulfoton
residues.
The
binding
potentials
of
the
degradates
were
not
known
(they
were
not
considered
individually),
but
were
assumed
to
be
the
same
as
parent
disulfoton.

Citations:

Barrett,
M.
R.
1999.
Updated
Documentation
on
the
SCI­
GROW
Method
to
Determine
Screening
Concentration
Estimates
for
Drinking
Water
Derived
from
Ground
Water
Sources.
Memorandum
From:
M.
R.
Barrett
To:
J.
Merenda.
Environmental
Fate
and
Effects
Division,
Office
of
Pesticide
Programs,
U.
S.
Environmental
Protection
Agency,
Arlington,
VA.

Barton,
A.
1982.
Note
to
Ed
Johnson
dated
12/
10/
82
describing
joint
effort
between
EPA/
OPP
and
Wisconsin
Department
of
Natural
Resources
to
monitor
pesticides
in
ground
water
per
communication
with
the
Wisconsin
Department
of
Natural
Resources.
1982.
Pesticide
Monitoring
in
Wisconsin
Ground
Water
in
the
Central
Sands
Area.
Madison,
WI
Harken,
J.
M.,
F.
A.
Jones,
R.
Fathulla,
E.
K.
Dzanton,
E.
J.
O'Neill,
D.
G.
Kroll,
and
G.
Chesters.
1984.
Pesticides
in
Groundwater
beneath
the
Central
Sand
Plain
of
Wisconsin.
Univ.
of
Wisc.
Resources
Center
Technical
Report
WIS
WRC
84­
01.

Holden,
P.
W.
1986.
Pesticide
and
Groundwater
Quality
Issues
and
Problems
in
Four
States.
National
Academy
Press.
Washington,
D.
C.

Jones,
R.
D.,
J.
Breithaupt,
J.
Carleton,
L.
Labelo,
J.
Lin,
R.
Matzner,
R.
Parker,
W.
Effland,
N.
Thurman,
and
I.
Kennedy.
2000.
Guidance
for
Use
of
the
Index
Reservoir
and
Percent
Crop
Area
Factor
in
Drinking
Water
Assessments.
Draft
3/
21/
2000.
Environmental
Fate
and
Effects
Division,
Office
of
Pesticide
Programs,
U.
S.
Environmental
Protection
Agency,
Arlington,
VA.

Mostaghimi,
S.
et
al.
1989.
Watershed/
Water
quality
monitoring
for
evaluating
BMP
effectiveness
­
Nomini
Creek
Watershed.
Report
N­
P1­
8811.
Agricul.
Engineer.
Dept.
Virginia
Tech.

Mostaghimi,
S.,
S.
Shukla,
and
P.
W.
McClellan.
1998.
BMP
Impacts
on
Nitrate
and
Pesticide
Transport
to
Groundwater
in
the
Nomini
Creek
Watershed.
Final
Report
No.
NC­
0298
Biological
Systems
Engineering
Department,
Virginia
Polytechnic
Institute
and
State
University,
Blacksburg,
VA
NCIWP,
1997.
The
Interagency
Study
of
the
Impact
of
Pesticide
Use
on
Ground
Water
in
North
Carolina.
Prepared
for
North
Carolina
Pesticide
Board
by
The
Interagency
Work
Group.
March
4,
1997.
North
Carolina
Department
of
Agriculture,
Raleigh,
NC.
25
PRZM/
EXAMS
Modeling,
dated
4/
13/
99)
from
Water
Quality
Technology
Team,
Environmental
Fate
and
Effects
Division,
Office
of
Pesticide
Programs,
USEPA.
Arlington,
VA.

USEPA.,
2000.
DP
Barcode
267486
EPA
Review
of
NCIWP,
1997.
The
Interagency
Study
of
the
Impact
of
Pesticide
Use
on
Ground
Water
in
North
Carolina.
Prepared
for
North
Carolina
Pesticide
Board
by
The
Interagency
Work
Group.
March
4,
1997.
North
Carolina
Department
of
Agriculture,
Raleigh,
NC
and
its
relevance
to
the
disulfoton.

References­
Personal
Communications
Kevin
Costello,
EFED
(http://
www.
epa.
gov/
pesticides/
cumulative/
pra­
op/
iii_
e_
3­
f.
pdf).

Note:
All
states
contacted
are
listed
below.
Only
the
states
which
monitored
for
disulfoton
(except
Florida)
are
include
in
this
document.

Monitoring
contacts
Tony
Cofer,
Pesticide
Administrator
of
the
Alabama
Department
of
Agriculture
and
Industry
Groundwater
Protection
Section
Dr.
Enid
Probst
,
Alabama
Department
of
Environmental
Management
Rose
Lombardi
,
Alaska
Department
of
Environmental
Conservation
Pesticide
Program
Charles
Armstrong,
Assistant
Director,
Arkansas
State
Plant
Board
Dr.
Robert
Matzner,
California
Department
of
Pesticide
Regulation
Frank
Spurlock,
California
Department
of
Pesticide
Regulation
Brad
Austin,
Water
Quality
Control
Division,
Colorado
Department
of
Public
Health
and
Environment
Judith
Singer,
Connecticut
Department
of
Environmental
Protection
Pesticide
Management
Division
Scott
Blaier
,
Hydrologist,
Delaware
Department
of
Agriculture
Keith
Parmer,
Florida
Department
of
Agriculture
and
Consumer
Services
Doug
Jones,
Georgia
Department
of
Agriculture
Robert
Boesch,
Hawaii
Department
of
Agriculture
Pesticides
Branch
Gary
Bahr,
Idaho
Dept
of
Agriculture
Division
of
Agricultural
Technology
26
Dave
McMillan,
Illinois
Environmental
Protection
Agency,
Bureau
of
Water,
Ground
Water
Section
Ryan
McDuffee
,
Environmental
Scientist,
Indiana
Department
of
Environmental
Management
Al
Lao,
Indiana
Department
of
Environmental
Management
Mary
Skopec,
Acting
Section
Supervisor,
Iowa
Department
of
Natural
Resources
Water
Monitoring
Section
Theresa
Hodges,
Kansas
Department
of
Health
and
Environment
Dale
Lambley,
Special
Environmental
Assistant
to
the
Secretary,
Kansas
Department
of
Agriculture
Peter
Goodman,
Manager,
Ground
Water
Branch,
Division
of
Water,
Kentucky
Department
of
Environmental
Protection
Karen
Irion,
Louisiana
Julie
Chizmas,
Senior
Water
Quality
Specialist,
Maine
Department
of
Agriculture
Rob
Hofstedter,
Maryland
Department
of
Agriculture
David
Bolton,
Maryland
Geological
Survey
Kenneth
Pelotiere,
Massachusetts
Department
of
Environmental
Protection
Source
Water
Assessment
Program
Dennis
Bush,
Surface
Water
Quality
Division,
Michigan
Department
of
Environmental
Quality
Mark
Breithart,
Drinking
Water
Division,
Michigan
Department
of
Environmental
Quality
Daniel
Helwig,
Minnesota
Pollution
Control
Agency
Mark
Zabel,
Minnesota
Department
of
Agriculture
Rusty
Crowe,
Mississippi
Department
of
Agriculture
and
Commerce
Bureau
of
Plant
Industry
Shedd
Landreth,
Mississippi
Department
of
Environmental
Quality
Paul
Andre,
Program
Coordinator,
Department
of
Agriculture
Plant
Industries
Division
27
Terry
Timmons,
Missouri
Department
of
Natural
Resources
John
Ford,
Missouri
Department
of
Natural
Resources
Donna
Rise,
Montana
Department
of
Agriculture,
Agricultural
Sciences
Division,
Technical
Services
Bureau
John
Lund,
Supervisor,
Surface
Water
Unit,
Nebraska
Department
of
Environmental
Quality
Craig
Romary,
Nebraska
Department
of
Agriculture,
Bureau
of
Plant
Industry
Scott
Cichowlaz,
Nevada
Depart
(of
Ag?)

Pat
Bickford,
New
Hampshire
Department
of
Environmental
Services
Dr.
Roy
Meyer,
Pesticide
Monitoring
and
Evaluation,
New
Jersey
Department
of
Environmental
Protection
Doug
Henson,
New
Mexico
Department
of
Agriculture,
Bureau
of
Pest
Management.

Jeff
Myers,
New
York
Department
of
Environmental
Conservation
Bureau
of
Technical
Support
Dr.
Henry
Wade,
Environmental
Programs
Manager,
North
Carolina
Department
of
Agriculture
and
Consumer
Services
William
Schuh,
North
Dakota
State
Water
Commission
Norene
Bartelson,
North
Dakota
Department
of
Health
Todd
Kelleher,
Ohio
Environmental
Protection
Agency
Department
of
Drinking
and
Ground
Waters
Julie
Letterhos,
Ohio
Environmental
Protection
Agency
Department
of
Drinking
and
Ground
Waters
Gail
Hess,
Ohio
Environmental
Protection
Agency
Don
Molnar,
Oklahoma
Department
of
Agriculture,
Plant
Industry
and
Consumer
Services
Division
John
Pari,
Pennsylvania
Department
of
Agriculture,
Bureau
of
Plant
Industry
Eugene
Pepper,
Rhode
Island
Department
of
Environmental
Management,
Division
of
Agriculture
and
Resource
Marketing
28
Jerry
Moore,
South
Carolina
Pesticide
Regulation
Board,
Clemson
University
Peter
Stone,
South
Carolina
Department
of
Health
and
Environmental
Control
Kathy
Stecker,
South
Carolina
Department
of
Health
and
Environmental
Control
Stan
Pence,
Senior
Hydrologist,
South
Dakota
Geological
Survey
Brad
Berven,
South
Dakota
Department
of
Agriculture
Pesticide
Program
Ken
Nafe,
Tennessee
Department
of
Agriculture
Dr.
Ambrose
Charles,
Texas
Department
of
Agriculture
Mark
Quilter,
Utah
Department
of
Agriculture
and
Food
Arne
Hulquist,
Utah
Department
of
Environmental
Quality
Cary
Giguere,
Vermont
Department
of
Agriculture,
Food
and
Markets
Marvin
Lawson,
Virginia
Department
of
Agriculture
and
Consumer
Services
Daniel
Schweitzer,
Virginia
Department
of
Agriculture
and
Consumer
Services
Doug
Hudson,
West
Virginia
Department
of
Agriculture
Chad
Board,
West
Virginia
Department
of
Environmental
Protection
William
Phelps,
Wisconsin
Department
of
Natural
Resources
Bureau
of
Drinking
&
Groundwater
Jim
Bigelow,
Wyoming
Department
of
Agriculture
Cheryl
Eddy
Miller,
United
States
Geological
Survey,
Wyoming
Robert
Sneed,
United
States
Army
Corps
of
Engineers
29
APPENDIX
1.

UNITED
STATES
ENVIRONMENTAL
PROTECTION
AGENCY
WASHINGTON,
D.
C.
20460
OFFICE
OF
PREVENTION,
PESTICIDES
AND
TOXIC
SUBSTANCES
TO:
Christina
Scheltema
Betty
Shackleford
Michael
Goodis
Special
Review
and
Reregistration
Division
(7508C)

FROM:
James
Wolf
ERB3
Environmental
Fate
and
Effects
Division
(7507C)
30
DATE:
November
20,
2001
RE:
Disulfoton
residues
in
ground
water
found
in
the
Virginia
BMP
Study:

BMP
Impacts
on
Nitrate
and
Pesticide
Transport
to
Groundwater
in
the
Nomini
Creek
Watershed.
Final
Report
No.
NC­
0298
S.
Mostaghimi,
S.
Shukla,
and
P.
W.
McClellan.
1998.
Biological
Systems
Engineering
Department
Virginia
Polytechnic
Institute
and
State
University
Blacksburg,
VA
#
The
ground
water
monitoring
component
was
started
in
1986
and
ended
in
June,
1997.

#
Nomini
Creek
Watershed
is
located
in
Westmoreland
County,
Va.
The
1463
ha
watershed
has
typical
Coastal
Plain
land
use
49%
cropland,
47%
woodland,
and
4%
used
for
homestead
and
roads
(different
reports
has
slightly
different
breakdown,
but
have
the
same
major
uses).
Average
annual
precipitation
is
102
cm,
with
most
of
the
rainfall
occurring
between
April
and
September.
Most
ground
water
recharge
occurs
in
late
Fall
or
early
spring.

#
Nomini
Creek
Watershed
is
located
in
the
Coastal
Plain
Physiographic
providence.
Soils,
geology
and
topography
are
similar
to
the
of
the
unglaciated
Atlantic
Coastal
Plain.
Soils
are
mostly
Ultisols.
The
major
soil
series
are
Suffolk
and
Rumford.
These
soils
cover
91
percent
of
the
area
and
have
similar
physical
properties.

Soil
Taxonomy
Sulfolk
Coarse­
loamy,
siliceous,
thermic
Typic
Hapudults
Rumford
Coarse­
loamy,
siliceous,
thermic
Typic
Hapudults
The
Coastal
Plain
has
been
identified
as
a
vulnerable
area
to
ground
water
contamination.
Other
vulnerable
regions
have
also
been
identified.
The
soils
could
also
be
used
to
identify
possible
problem
areas.
(Can't
be
done
by
tomorrow).
These
are
vulnerable
soil
for
leaching.

#
Agriculture
is
primarily
row
crops.
Major
crops
are
corn,
soybeans,
and
small
grains
(wheat
and
barley).
Typical
rotation
is
conventionally­
tilled
corn,
followed
by
small
grains
with
no­
till
soybeans
planted
in
the
small
grain
residues.
Occasionally,
full
season,
conventionally­
tilled
soybeans
is
also
grown.
USDA
Ag
Statistics
do
not
report
31
tobacco
production
for
Westmoreland
County.
Potatoes
are
reported
to
be
produced,
but
production
appears
to
be
declining.

#
Study
Objective
to
study
the
quality
of
surface
and
ground
water
as
influenced
by
the
agricultural
practices
in
the
watershed.

#
Monitoring
consisted
of
two
(2)
runoff
and
surface
water
monitoring
stations;
seven
rain
gauges;
one
weather
station;
and
eight
(8)
ground­
water
monitoring
wells
(GN1
to
GN8).
The
ground
water
wells
were
located
primarily
in
agricultural
areas.
These
wells
were
drilled
in
pairs,
100
­
150
meters
apart,
with
one
in
pair
located
hydraulically
downgradient
of
the
other.

Characteristic
(m)
Value
Well
GN1
GN2
GN3
GN4
GN5
GN6
GN7
GN8
Well
depth
13.7
12.8
15.2
13.7
16.5
12.0
15.8
11.9
GW
depth
Mean
10.3
9.
6
13.1
9.
4
12.9
8.
2
13.3
8.
6
"
Max.
12.0
10.8
14.0
12.7
13.9
9.
1
14.4
9.
6
"
Min
8.
5
7.1
11.5
7.
0
11.3
7.
0
11.8
7.
4
#
Approximately
monthly
samples
were
taken
from
each
monitoring
well
and
analyzed
for
a
number
of
analytes
including
22
pesticides.
QA/
QC
procedures
were
followed.

#
Herbicide
and
insecticide
application
information
in
the
watershed
were
obtained
from
farmer
surveys.
The
rate
and
time
of
herbicide
application
was
dependent
on
the
crop
rotation
adopted
by
the
farmer.
Corn
is
usually
planted
between
late
April
and
early
May.
Post­
emergence
sprays
applications
occur
in
early
July.
The
timing
and
application
rates
of
insecticides,
applied
individually
or
in
combination,
in
the
watershed
depending
on
the
type
and
extent
of
the
insect
problem
observed.

Note:
the
label
does
allow
for
fall
application
to
wheat.
Perhaps
fall
application
and
greater
fall
recharge
resulted
in
the
observed
concentration
(2.87
:
g/
L).
Possible
mitigation
option?

#
Disulfoton
sampling
results
and
detection
statistics
in
the
Nomini
Creek
Watershed
(Table
15,
after
Mostaghimi,
1998).
These
are
disulfoton
parent.

Pesticide
Total
Samples
Detections
1
Detection
2
Frequency
(percent)
Concentration
(
:
g/
L)

Max
Mean
SD
Disulfoton
1010
10
1.0
2.87
0.39
0.32
32
Pre­
BMP
3
(5/
86
­
10/
88)

229
7
3.1
2.87
0.52
Post
BMP
4
(11/
89
­
9/
96)

693
3
0.4
0.10
0.08
1
Number
of
samples
with
detectable
levels
of
pesticide
2
(samples
with
detectable
levels
of
pesticide
*
100)/
total
number
of
samples
3
Before
agricultural
Best
Management
Practice
(BMP)
implemented
in
watershed.
4
Following
the
implementation
of
BMP
within
the
watershed.

Note:
I
only
had
(raw)
data
through
1990.
Thus,
I
only
had
6
of
the
10
detections,
mean
was
0.57
:
g/
L,
which
is
only
slightly
greater
than
the
mean
with
7
samples
(pre­
BMP).

Discussion
and
recommendation:

The
following
table
was
included
in
Feb.
7,
2000
Additional
Clarification
of
Disulfoton
GroundWater
Monitoring
Data
Assessment.
In
a
recent
discussion
about
a
"chronic"
exposure
for
ground
water
the
following
suggestions
was
put
forth
(mean
=
1.49
:
g/
L
=
(2.87
+0.1)/
2
for
well
site
GN3.
Considering
there
are
many
"monthly
samples,
with
most
being
less
than
the
detection
limit,
a
lower
mean
is
probably
justified
(disulfoton
parent
only).
The
mean
of
all
the
detections
is
0.39
:
g/
L,
the
mean
of
the
pre­
BMP
is
0.52
:
g/
L,
and
post­
BMP
is
0.08
:
g/
L.
Without
specifically
estimating
a
concentration,
I
think
that
as
far
as
parent
disulfoton
goes,
the
average
concentration
would
be
expected
to
be
considerably
less
than
the
DWLOC
of
1.2(?).

Summary
of
Disulfoton
Detections
in
ground
water
from
the
eight
ground­
water
monitoring
wells
in
Nomini
Creek
Watershed
(Virginia),
during
1986
and
1987.

Sampling
Date
Well­
Site
Number
Concentration
(
:
g/
L)

11/
5/
86
GN3
2.87
11/
5/
86
GN6
0.04
3/
13/
87
GN4
0.10
8/
20/
87
GN1
0.13
8/
20/
87
GN2
0.16
8/
20/
87
GN3
0.10
