White
Paper
on
Potential
Developmental
Effects
of
Atrazine
on
Amphibians
In
Support
of
an
Interim
Reregistration
Eligibility
Decision
on
Atrazine
Submitted
to
the
FIFRA
Scientific
Advisory
Panel
for
Review
and
Comment
June
17
­
20,
2003
Office
of
Prevention,
Pesticides
and
Toxic
Substances
Office
of
Pesticide
Programs
Environmental
Fate
and
Effects
Division
Washington,
D.
C.

May
29,
2003
Page
2
of
95
ACKNOWLEDGMENTS
Authors:

Thomas
Steeger,
Office
of
Prevention,
Pesticides,
and
Toxic
Substances,
Office
of
Pesticide
Programs,
Environmental
Fate
and
Effects
Division
and
Joseph
Tietge,
Office
of
Research
and
Development,
National
Health
and
Environmental
Effects
Research
Laboratory,

Mid­
Continent
Ecology
Division
were
responsible
for
preparation
of
the
white
paper.

Reviewers:

The
authors
acknowledge
the
input
of
a
number
of
reviewers.
Reviewers
included:
Les
Touart
and
Joseph
Merenda
from
the
Office
of
Science
Coordination
and
Policy;
William
Jordan
from
the
Immediate
Office
of
Pesticide
Programs,
Kimberly
Nesci
(
Special
Review
and
Reregistration
Division);
Arthur­
Jean
Williams
(
Field
and
External
Affairs
Division);
and
Mary
Frankenberry,
Stephanie
Irene,
Karen
McCormack,
Edward
Odenkirchen,
Ingrid
Sunzenauer,
and
Douglas
Urban
(
Environmental
Fate
and
Effects
Division)
from
the
Office
of
Pesticide
Programs.

Additionally,
Gary
Ankley,
Sigmund
Degitz
and
Patricia
Schmieder
of
the
Mid­
Continent
Ecology
Division,
Office
of
Research
and
Development
and
Nancy
Beck
of
the
Office
of
Information
and
Regulatory
Affairs,
Office
of
Management
and
Budget
provided
helpful
reviews.
Page
3
of
95
TABLE
OF
CONTENTS
EXECUTIVE
SUMMARY
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
5­

Background
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
5­

Overview
of
Laboratory
and
Field
Studies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
6­

Laboratory
Studies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
6­

Field
Studies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
8­

Conclusions
from
Laboratory
and
Field
Studies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
8­

Conceptual
Model
for
Future
Studies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
9­

CHAPTER
1
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
12­

Introduction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
12­

Regulatory
Background
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
12­

Summary
of
Pesticide
Effects
Characterization
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
13­

Background
on
Atrazine
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
14­

CHAPTER
2
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
17­

Study
Reviews
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
17­

Laboratory
Studies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
18­

Field
Studies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
33­

CHAPTER
3
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
48­

Study
Strengths
and
Limitations
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
48­

Laboratory
Studies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
48­

Field
Studies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
50­

CHAPTER
4
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
52­

Uncertainties
in
Assessing
Potential
Atrazine
Effects
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
52­

Ecological
Relevancy
of
Endpoint
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
53­
Page
4
of
95
Dose­
Response
Relationships
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
54­

Mechanistic
Plausibility
of
Atrazine
Effects
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
54­

Laboratory
to
Field
Extrapolation
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
56­

Application
of
Available
Studies
to
Assess
Potential
Atrazine
Effects
.
.
Page
­
56­

CHAPTER
5
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
58­

Reducing
the
Uncertainties
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
58­

Phase
1.
Test
for
Apical
Gonadal
Effects
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
60­

Species
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
61­

Stage
Sensitivity
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
61­

Test
Conditions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
62­

Dose­
Response
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
64­

Positive
Controls
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
65­

Sampling
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
65­

Endpoints
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
66­

Quality
Indicators
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
66­

Analysis,
Interpretation,
and
Iteration
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
68­

Phase
2:
Sex
Steroid
Measurements
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
69­

Phase
3:
Aromatase
Activity
Measurements
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
69­

Phase
4:
Aromatase
Inhibitor
Study
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
70­

Phase
5:
Ecological
Relevancy
Study
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
71­

CHAPTER
6
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
74­

Conclusions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
74­

CHAPTER
7
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
77­

Charge
to
the
Panel
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
77­

Questions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
78­

References
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Page
­
85­
Page
­
5­
EXECUTIVE
SUMMARY
BACKGROUND
In
response
to
an
amended
consent
decree
entered
to
conclude
litigation
between
EPA
and
the
Natural
Resources
Defense
Council
(
NRDC),
the
Agency
released
on,
January
31,
2003,

an
atrazine
Interim
Reregistration
Eligibility
Decision
(
IRED)
that
was
based,
in
part,
on
an
assessment
of
the
pesticide's
ecological
risk.
However,
several
studies
concerning
the
potential
effects
of
atrazine
on
amphibian
development
were
published
in
a
time
frame
that
prevented
the
Agency
from
completing
a
rigorous
evaluation
of
these
data
for
inclusion
in
the
January
31,
2003
IRED.

Thus,
the
decree
stipulated
that
"
EPA
shall
sign,
on
or
before
October
31,
2003,
a
revised
IRED
for
atrazine
that
addresses
.
.
.
data,
received
by
EPA
prior
to
February
28,
2003,
relating
to
the
potential
effects
of
atrazine
on
amphibian
species."
The
Consent
Decree
also
stipulated
that
"[
a]
t
least
three
months
prior
to
signing
this
revised
IRED,
EPA
shall
develop
a
paper
and
submit
it
to
the
FIFRA
Scientific
Advisory
Panel
(
SAP)
for
review
and
comment."
The
paper
is
required
to
address
"
the
significance
of
the
amphibian
risk
data"
and
"
whether
there
is
a
need
for
additional
data
to
characterize
more
fully
atrazine's
potential
risks
to
amphibian
species,
and,
if
so,
what
data
should
be
developed."

In
accordance
with
the
consent
decree,
EPA
has
developed
this
white
paper
as
the
basis
for
an
SAP
meeting
to
be
held
on
June
17
­
20,
2003.
The
white
paper
provides
an
overview
of
relevant
studies
published
in
the
scientific
literature
and
studies
submitted
by
the
registrant
through
February
28,
2003.
This
white
paper
assesses:

°
the
strengths
and
limitations
of
the
available
studies
to
evaluate
the
extent
to
which
atrazine
may
elicit
developmental
effects
in
amphibians;

°
the
extent
of
concordance
for
the
entire
body
of
information
derived
from
these
laboratory
and
field
studies
to
assess
the
plausibility
that
atrazine
can
cause
developmental
effects;

and
if
so,

°
the
nature
and
strength
of
associated
dose­
response
relationships.
Page
­
6­
The
paper
also
outlines
a
conceptual
model
for
potential
future
studies
that
could
provide
information
to
resolve
inconsistencies
and
address
gaps
in
the
existing
knowledge
base
that
currently
preclude
establishing
a
definitive
characterization
of
atrazine's
effects
on
amphibian
development.

OVERVIEW
OF
LABORATORY
AND
FIELD
STUDIES
In
developing
this
white
paper,
EPA
reviewed
a
total
of
17
studies
(
Table
A)
that
were
published
in
the
scientific
literature
or
submitted
by
the
registrant.
Seven
were
laboratory­
based
investigations,
and
ten
were
field
studies.
Of
the
twelve
registrant­
sponsored
studies,
seven
were
considered
preliminary
or
in
progress
by
the
registrant.
All
studies
were
individually
evaluated
with
regard
to
the
following
parameters:
experimental
design,
protocols
and
data
quality
assurance,
strength
of
cause­
effect
and/
or
dose­
response
relationships,
mechanistic
plausibility,

and
ecological
relevancy
of
measured
endpoints.

Laboratory
Studies
The
laboratory
studies
reviewed
for
this
paper
addressed
a
variety
of
endpoints:
time
to
metamorphosis,
growth
(
body
length
and
weight),
gonadal
abnormalities,
sex
ratios,
laryngeal
dilator
muscle
area,
plasma
steroid
concentration,
and
brain/
gonad
aromatase
activity.
In
evaluating
the
effects
of
atrazine
on
amphibian
development,
researchers
tested
a
variety
of
frog
species,
including
the
African
clawed
frog
(
Xenopus
laevis),
the
leopard
frog
(
Rana
pipiens),
and
the
green
frog
(
R.
clamitans).
Not
all
species
were
used
in
every
study.
All
of
the
laboratory
studies
used
static
renewal
exposures.
Within
individual
studies
the
number
of
atrazine
exposures
ranged
from
only
one
concentration
to
up
to
six
different
concentrations.
In
some
cases,
the
nominal
atrazine
concentrations
were
confirmed
analytically.

The
Agency's
evaluation
of
the
relevant
studies
in
the
open
literature
and
registrantsubmitted
laboratory
investigations
concludes
that
none
of
the
laboratory
studies
fully
accounted
for
environmental
and
animal
husbandry
factors
capable
of
influencing
endpoints
which
the
studies
Page
­
7­
were
attempting
to
measure.
For
example,
in
several
studies
excessive
mortality
and/
or
delayed
development
in
untreated
organisms
or
a
failure
of
organisms
to
respond
to
a
positive
control
(
e.
g.,
estradiol)
confounded
the
interpretation
of
study
results.
In
addition,
in
all
the
studies,

loading
rates
(
i.
e.,
tadpole
biomass
per
liter
of
water)
were
typically
too
high
for
the
solution
changes
associated
with
static
renewal
atrazine
exposures.
Resulting
poor
water
quality
(
e.
g.,
low
dissolved
oxygen,
high
ammonia)
could
have
created
environmental
conditions
unfavorable
to
optimum
survival,
growth
and
development.
In
other
instances,
the
condition
of
the
organisms
may
have
been
compromised
by
poor
quality
feed.

After
reviewing
the
entire
body
of
information
available
from
the
laboratory
investigations,

EPA
has
concluded
that
the
scientific
evidence
does
not
support
many
of
the
conclusions
reached
by
the
various
study
authors.
Some
of
the
major
limitations
of
these
laboratory
studies,
that
make
it
difficult
to
draw
conclusions
with
confidence
about
the
effects
of
atrazine
on
amphibian
species,

include
the
following:

°
For
studies
conducted
at
atrazine
concentrations
in
the
range
of
0.1
to
25
ug/
L,
the
interpretation
of
dose­
response
relationships
for
measured
endpoints
(
e.
g.,
gonadal
abnormalities,
aromatase
activity,
plasma
steroid
concentrations
and
laryngeal
dilator
muscle
diameter)
was
problematic.
Analytical
measurements
of
atrazine
were
incomplete,

or
atrazine
was
detected
in
the
dilution
water
for
the
control
organisms
at
concentrations
comparable
to
low
concentration
treatments.

°
Gonadal
abnormalities
(
ovotestes
and
discontinuous
gonads)
and
laryngeal
dilator
muscle
diameter
effects,
reported
at
atrazine
concentrations
in
the
range
of
0.1
to
200
ug/
L,
have
not
been
reproduced
to
date.
The
extensive
variability
in
the
study
design
protocols
has
made
it
difficult
to
determine
if
this
lack
of
reproducibility
for
demasculinizing
effects
(
decreased
laryngeal
dilator
muscle
area),
and
reports
of
an
inverted
dose­
response
relationship
for
other
gonadal
developmental
endpoints,
are
valid
results
or
artifacts
of
the
design
and
quality
of
the
investigations.

°
The
potential
gonadal
developmental
effects
of
atrazine,
as
well
as
an
explanation
for
the
non­
monotonic
dose­
response
curve
reported
by
some
investigators,
has
been
proposed
to
result
from
induction
of
aromatase
activity.
This
increased
enzyme
activity
would
in
turn
Page
­
8­
lead
to
elevated
estrogen
levels
and
ultimately
the
observed
effects,
i.
e.,
ovotestes
and
reduced
secondary
sex
characteristics,
in
males.
Aromatase
induction
by
atrazine,
though,

has
not
been
demonstrated
in
any
anuran
in
controlled
laboratory
investigations.

Field
Studies
The
field
studies,
which
included
a
microcosm
experiment,
evaluated
growth
(
body
weight
and
length),
gonadal
abnormalities,
sex
ratios,
plasma
steroid
concentration
and
brain/
blood
aromatase
activity
in
X.
laevis,
R.
clamitans,
cricket
frogs
(
Acris
crepitans),
bullfrogs
(
R.

catesbiena),
and
cane
toads
(
Bufo
marinus)
in
Florida,
Indiana,
Iowa,
Michigan,
Nebraska,
Utah,

and
Wyoming,
USA
and
South
Africa.

The
Agency
also
concludes
that
the
currently
available
field
studies
are
of
limited
value
due
to
the
high
variability
in
environmental
conditions
(
e.
g.,
photoperiod,
temperature,
water
quality)
under
which
field­
collected
organisms
lived,
uncertainty
as
to
amphibian
developmental
status
and
condition
at
the
initiation
of
the
studies,
and
an
inability
to
relate
the
co­
occurrence
of
atrazine
with
key
developmental
windows
for
the
organisms
under
investigation.
Consequently,

EPA
cannot
determine
whether
the
failure
of
several
studies
to
show
any
relationship
between
measured
or
predicted
aqueous
atrazine
concentrations
and
developmental
effects
reflects
the
absence
of
a
causal
relationship
or
the
limitations
of
the
study
methodologies.
In
addition,
the
actual
or
possible
co­
occurrence
of
additional
chemical
and/
or
non­
chemical
stressors
confound
attempts
to
attribute
any
observed
responses
to
atrazine
exposure.

Conclusions
from
Laboratory
and
Field
Studies
Overall,
the
weight­
of­
evidence
based
on
currently
available
studies
does
not
show
that
atrazine
produces
consistent,
reproducible
effects
across
the
range
of
exposure
concentrations
and
amphibian
species
tested.
The
current
body
of
knowledge
has
deficiencies
and
uncertainties
that
limit
its
usefulness
in
interpreting
potential
atrazine
effects.
Specifically,
the
demasculinizing
(
decreased
laryngeal
dilator
muscle
area)
effects
were
not
replicated
in
multiple
laboratories.
Page
­
9­
Additionally,
the
feminizing
effects
(
intersex/
hermaphroditism/
ovotestes)
of
atrazine
were
observed
in
three
laboratory
studies
whose
experimental
designs
could
not
be
easily
reconciled
and
that
reported
significant
effects
at
different
concentrations:
one
at
25
ug/
L
atrazine
and
the
other
two
at
0.1
ug/
L.
While
the
feminizing
effects
observed
in
these
different
studies
were
consistent
qualitatively,
there
was
no
consistency
across
the
studies
in
the
reported
dose­
response
relationships.
That
inconsistency,
together
with
the
limitations
in
methodology
in
each
study,

does
not
allow
a
reliable
determination
of
causality
or
the
nature
of
any
dose­
response
relationship.
Although
the
Florida
cane
toads
monitored
in
the
field
exhibited
both
demasculinizing
effects
(
genetic
males
with
female
coloration)
and
feminizing
effects
(
oogenesis
in
male
Bidder's
organ),
there
were
insufficient
data
to
conclusively
link
atrazine
exposure
to
the
phenomena.
Thus,
the
available
data
do
not
establish
a
concordance
of
information
to
indicate
that
atrazine
will
or
will
not
cause
adverse
developmental
effects
in
amphibians.

Conceptual
Model
for
Future
Studies
If
the
Agency's
risk
management
decision
requires
a
greater
degree
of
certainty
in
the
ecological
risk
assessment
for
atrazine
than
possible
from
currently
available
data,
then
additional
information
would
be
necessary
to
evaluate
the
potential
causal
relationships
between
atrazine
exposure
and
gonadal
development
in
amphibians.
If
a
causal
relationship
is
indicated,
then
further
testing
may
be
needed
to
establish
the
nature
of
the
dose­
response
relationship,
the
ecological
relevancy
of
the
effects,
the
plausibility
of
an
underlying
mechanism(
s),
and
the
degree
of
interspecies
variability
for
any
effects.

Laboratory­
based
studies,
planned
and
executed
to
address
critical
design
and
protocol
issues,
can
provide
data
to
adequately
assess
whether
atrazine
exposure
results
in
gonadal
developmental
effects
in
frogs,
and
if
so,
the
strength
of
a
cause­
effect
relationship(
s)
and
the
nature
of
the
underlying
dose­
response
relationship(
s).
Because
field
investigations
may
introduce
considerable
variability
and
make
it
difficult
to
establish
cause­
effect
relationships,
EPA
recommends
that
more
controlled
laboratory
studies
be
conducted
before
field
studies
are
undertaken.
Page
­
10­
EPA
proposes
that
the
studies
be
conducted
in
a
tiered
approach
to
ensure
that
the
potential
effects
of
atrazine
are
evaluated
in
a
systematic
and
efficient
manner,
thus
minimizing
the
level
of
effort
and
resources
required,
while
maximizing
reductions
in
the
extant
uncertainties.
The
first
objective
in
this
approach
would
be
to
determine
the
effects
of
atrazine
on
gonadal
developmental
endpoints
at
the
organism
level
(
ovotestes,
sex
ratios),
using
high
quality
aquatic
toxicology
methods
with
X.
laevis
and,
to
a
lesser
degree,
North
American
Ranids.
These
studies
would
adhere
to
American
Society
for
Testing
Materials
(
ASTM)
recommended
standards
for
loading
rates,
husbandry
and
water
quality
parameters.
The
studies
would
address
inconsistencies
in
previous
research
regarding
the
causal
relationship
between
atrazine
exposure
and
developmental
effects
and
would
provide
sufficiently
robust
data
to
determine
the
nature
of
any
dose­
response
relationship.

If
the
results
from
the
initial
studies
are
negative,
then
there
would
be
no
rationale
to
conduct
further
investigation.
However,
if
the
results
are
positive,
then
EPA
proposes
that
additional
studies
would
be
designed
to
investigate
plausible
mechanisms
involved
in
the
etiology
of
the
response.
Establishing
a
mechanistic
rationale
is
a
critical
basis
for
inter­
species
extrapolation.
Mechanistic
understanding
would
permit
the
development
of
bioindicators
that
can
be
applied
to
future
field
studies,
if
deemed
necessary,
and
it
would
reduce
the
uncertainties
associated
with
developing
a
causal
relationship
between
effects
on
measurement
endpoints
and
atrazine
exposure.
Assuming
that
the
Panel
reaches
conclusions
consistent
with
those
proposed
in
the
current
review
and
evaluation
of
the
existing
data,
EPA
may
request
a
consultation
with
the
SAP
to
review
in
greater
detail
study
designs
and
protocols
for
any
higher
tiered
studies.
Page
­
11­
Table
A.
Summary
of
studies
evaluated
on
the
effects
of
atrazine
on
amphibian
gonadal
development
and
sexual
differentiation.

Study
First
Author
EPA
Study
Number
or
Open
Literature
Citation
Species
Study
Type
Hayes
et
al.
2002a
Proceedings
of
the
National
Academy
of
Sciences,
99:
5476­
5480
Xenopus
laevis
laboratory
Goleman
et
al.
2003
MRID
No.
458677­
07
Xenopus
laevis
laboratory
Tavera­
Mendoza
et
al.
2001a
Environmental
Toxicology
and
Chemistry,
21:
527­
531
Xenopus
laevis
laboratory
Tavera­
Mendoza
et
al.
2001b
Environmental
Toxicology
and
Chemistry,
21:
1264­
1267
Xenopus
laevis
laboratory
Hecker
et
al.
2003
MRID
No.
458677­
04
Xenopus
laevis
laboratory
Hecker
et
al.
2003
MRID
No.
458677­
03
Rana
clamitans
laboratory
Villeneuve
et
al.

2003
MRID
No.
458677­
08
Xenopus
laevis
laboratory
Hayes
et
al.
2002b
http://
dx.
doi.
org/
[
Online
23
October
2002]

Environmental
Health
Perspectives,

111(
4):
568­
575
Rana
pipiens
laboratory
and
field
Du
Preez
et
al.
2003
MRID
No.
458677­
11
Xenopus
laevis
microcosm
Giesy
et
al.
2003
MRID
No.
458675­
01
Xenopus
laevis
field
Smith
et
al.
2003
MRID
No.
458677­
01
Xenopus
laevis
field
Smith
et
al.
2003
MRID
No.
458677­
09
Xenopus
laevis
field
Smith
et
al.
2003
MRID
No.
458677­
10
Xenopus
laevis
field
Crabtree
et
al.
2003
MRID
No.
458677­
05
Rana
catesbiena
field
Jones
et
al.
2003
MRID
No.
458677­
02
Rana
clamitans
field
Sepulveda
et
al.

2003
MRID
No.
458677­
06
Bufo
maninus
Bufo
terrestris
field
Reeder
et
al.
1998
Environmental
Health
Perspectives,

106:
261
­
266
Acris
crepitans
field
Page
­
12­
CHAPTER
1
INTRODUCTION
Regulatory
Background
In
April
2002,
the
Office
of
Pesticide
Programs
(
OPP)
of
the
U.
S.
Environmental
Protection
Agency
(
EPA)
issued
a
document
that
characterized
the
environmental
fate
and
ecological
effects
of
atrazine
in
support
of
the
interim
reregistration
eligibility
decision
(
IRED)
on
atrazine
(
IRED
Science
Chapter
2002).
On
April
16,
2002,
EPA
held
a
technical
briefing
where
Agency
officials
discussed
human
health
and
ecological
risk
assessments
for
atrazine.
At
the
same
time
as
the
public
briefing,
newly
generated
information
regarding
the
potential
effects
of
atrazine
on
amphibian
development
was
published
in
the
open
literature
(
Hayes
et
al.
2002a),
and
concerns
were
raised
that
EPA
had
not
sufficiently
accounted
for
the
potential
effects
of
atrazine
on
amphibian
development
and
possible
effects
on
human
health.

In
an
amended
consent
decree
(
Consent
Decree
2002)
between
EPA
and
the
Natural
Resources
Defense
Council
(
NRDC),
EPA
was
required
to
issue,
by
January
31,
2003,
an
IRED
on
atrazine;
the
consent
decree
acknowledged
that
the
IRED
would
not
take
into
account
recent
literature
on
the
effects
of
atrazine
on
amphibians.
The
decree
further
stipulated
that
EPA
sign,

by
October
31,
2003,
a
revised
IRED
that
incorporated
recommendations
from
a
FIFRA
Scientific
Advisory
Panel
(
SAP)
regarding
studies
on
the
potential
effects
of
atrazine
on
amphibians.
The
decree
indicated
that
EPA
would
review
data
received
by
the
Agency
prior
to
February
28,
2003,
and
that
the
SAP
would
review
a
paper
developed
by
EPA
which
described
and
evaluated
these
studies.
In
accordance
with
this
consent
decree,
EPA
has
developed
a
white
paper
that
critically
evaluates
currently
available
data,
discusses
the
nature
of
remaining
uncertainties
in
evaluating
the
potential
effects
of
atrazine
on
amphibian
development,
and
outlines
the
nature
of
future
studies
that
will
address
these
uncertainties.
Page
­
13­
Summary
of
Pesticide
Effects
Characterization
Although
EPA
routinely
requires
testing
in
multiple
species
(
40
CFR
158),
pesticide
toxicity
data
are
not
likely
to
be
available
for
all
potentially
exposed
non­
target
organisms.
In
the
majority
of
cases,
the
risk
assessment
process
relies
on
a
suite
of
toxicity
studies
performed
on
a
limited
number
of
surrogate
organisms.
For
example,
mallard
duck
(
Anas
platyrhynchos)
and
bobwhite
quail
(
Colinus
virginianus)
are
common
surrogates
for
birds,
while
rat
and
mouse
strains,
typically
used
in
studies
to
support
human
health
risk
assessments,
serve
as
surrogates
for
mammalian
wildlife.
With
regard
to
aquatic
life
associated
with
freshwater
ecosystems,
typical
surrogate
species
include
the
water
flea
(
Daphnia
magna),
bluegill
sunfish
(
Lepomis
macrochirus),
rainbow
trout
(
Oncorhynchus
mykiss),
and
fathead
minnow
(
Pimephales
promelas).

Within
these
broad
taxonomic
groups,
one
acute
and
one
chronic
endpoint
(
usually
mortality
and
frank
measures
of
growth
and
reproduction,
respectively)
are
selected
from
the
available
test
data.
Data
from
the
most
sensitive
species
tested
within
that
taxonomic
group
are
selected.
If
additional
toxicity
data
for
more
species
of
organisms
in
a
particular
group
are
available,
the
selection
need
not
be
limited
to
the
species
mentioned
previously,
but
may
be
expanded
to
include
additional
data
that
meet
the
Agency's
data
quality
requirements.
Regardless
of
the
extent
of
data
available
beyond
the
required
set
of
toxicity
studies,
the
risk
assessment
typically
relies
on
the
selection
of
endpoints
from
the
most
sensitive
species
tested
in
acceptable
studies.

While
the
above
mentioned
surrogates
and
toxicity
endpoints
are
routinely
used
in
Agency
risk
assessments,
they
do
not
represent
a
limitation
on
the
types
of
toxic
endpoints
that
may
be
considered
in
the
risk
assessment.
Through
the
evaluation
of
available
effects
data,
the
EPA
risk
assessment
team
may
encounter
other
effects
information
that
provides
insight
on
endpoints
and
organisms
not
routinely
considered.
Professional
judgment
is
used
by
the
risk
assessment
team
to
determine
whether
and
how
available
data
on
other
toxicological
endpoints
are
included
in
the
risk
assessment.
This
evaluation
includes
reference
to
data
quality
objectives
for
specific
types
of
studies,
the
degree
to
which
adequate
documentation
is
available
to
evaluate
the
technical
merit
of
the
data,
and
whether
the
data
are
applicable
to
the
assessment
endpoints
established
for
the
risk
assessment.
In
deciding
whether
data
are
applicable
to
assessment
endpoints,
the
risk
assessment
Page
­
14­
team
uses
professional
judgment
and
available
lines­
of­
evidence
to
determine
if
the
toxicological
endpoints
can
be
linked
to
assessment
endpoints
in
a
reasonable
and
plausible
manner.

Consistent
with
the
previous
discussion
on
endpoint
selection,
the
selection
of
amphibian
species
as
the
ecological
entity
for
a
risk
assessment
endpoint
may
be
appropriate
based
on
a
variety
of
considerations,
such
as
a
pesticide's
mode
of
action,
its
use
patterns
and
the
species
habitat
requirements.
Because
amphibians
undergo
profound
changes
during
metamorphosis
and
occupy
a
variety
of
ecological
niches
during
their
life­
span,
specific
species,
or
groups
of
species,

may
be
relevant
entities
for
formulating
risk
assessment
endpoints
for
certain
pesticides.

Amphibians
exhibit
a
life
cycle
in
which
conditions
must
be
favorable
for
their
survival
in
aquatic
ecosystems
where
they
breed
and
larvae
develop,
and
in
terrestrial
habitats,
where
many
adult
amphibians
reside.
Living
first
as
gilled­
and
skin­
breathing
aquatic
larvae
and
later
as
completely
or
partly
land­
dwelling
lung­
and
skin­
breathing
adults,
these
animals
rely
on
a
wide
range
of
mechanisms
for
interfacing
with
their
environments.
Unlike
many
organisms
that
have
either
keritinized
and/
or
scaled
outer
surfaces,
the
amphibian
skin
is
thin
and
is
used
as
a
respiratory
membrane
for
gas
exchange
and
water
absorption.
During
the
larval
stage,
many
amphibians
rely
on
filter
feeding
although,
as
a
group,
they
are
considered
the
primary
vertebrate
predator
for
invertebrates
in
many
freshwater
and
moist
terrestrial
environments.

Background
on
Atrazine
Atrazine
is
a
selective,
pre­
and
post­
emergence
herbicide
used
on
a
variety
of
terrestrial
food
crops,
non­
food
crops,
forests,
residential
turf,
golf
course
turf,
recreational
areas,
rights­

ofway
and
rangeland.
Although
used
to
control
broadleaf
and
many
other
weeds
on
a
range
of
agricultural
and
nonagricultural
sites,
the
herbicide's
largest
use
is
on
corn,
sorghum
and
sugarcane.
Atrazine's
primary
mode
of
action
in
plants
is
through
inhibition
of
photosynthesis
by
disruption
of
the
Photosystem
II
pathway.
Atrazine
is
characterized
as
a
persistent,
mobile
compound
that
may
be
transported
to
surface
water
via
runoff,
spray
drift,
and
atmospheric
deposition
(
IRED
2003).
Because
of
the
chemical's
resistance
to
abiotic
routes
of
degradation,

and
its
moderate
susceptibility
to
biotic
degradation,
it
is
likely
to
persist
in
water.
These
characteristics,
combined
with
the
herbicide's
level
of
use,
have
contributed
to
its
widespread
Page
­
15­
detections
in
surface
waters
throughout
the
United
States
and
its
associated
exposure
potential
for
aquatic
organisms.

Although
atrazine
exhibits
varying
degrees
of
acute
and
chronic
toxicity
to
animals,
its
major
identified
impact
on
aquatic
organisms
appears
to
be
an
indirect
effect
resulting
from
a
decrease
in
primary
productivity
in
associated
ecosystems
(
i.
e.,
decreased
plant
biomass).
In
terms
of
direct
effects,
atrazine
is
moderately
toxic
to
fish
(
LC
50
=
5.3
mg/
L)
and
highly
toxic
to
aquatic
invertebrates
(
LC
50
=
0.72
mg/
L)
on
an
acute
exposure
basis.
Although
screening­
level
assessments
did
not
show
an
acute
risk
to
aquatic
animals,
the
level
of
concern
for
chronic
risk
to
fish
and
invertebrates
was
exceeded
for
certain
uses
of
atrazine.
With
respect
to
terrestrial
organisms,
atrazine
was
found
to
be
slightly
toxic
to
birds
(
LD
50
=
940
mg/
kg)
and
mammals
(
LD
50
=
1,869
mg/
kg)
on
an
acute
exposure
basis
(
IRED
Science
Chapter
2002).
Chronic
sublethal
effects
in
mammals
included
effects
on
the
hypothalamic­
pituitary
axis
in
rats
(
HED
Science
Chapter
2002),
where
prolonged
atrazine
treatment
attenuated
pre­
ovulatory
lutenizing
hormone
surge
thereby
inhibiting
ovulation
and
potentially
delaying
the
onset
of
puberty
for
both
males
and
females.

In
April
2002,
Hayes
et
al.
(
2002a)
reported
in
the
Proceedings
of
the
National
Academy
of
Sciences
that
exposure
to
atrazine
concentrations
as
low
as
0.1

g/
L
resulted
in
demasculinization
(
reduced
laryngeal
muscle
growth)
and
feminization
(
testicular
oogenesis/
hermaphroditism)
of
African
clawed
frogs
(
Xenopus
laevis)
in
the
laboratory.
Studies
previous
to
those
of
Hayes
et
al.
(
2002a)
reported
the
incidence
of
hermaphroditism
in
cricket
frogs
(
Acris
crepitans)
correlated
with
atrazine
exposure
(
Reeder
et
al.
1998).
Other
studies
showed
that
exposure
of
Xenopus
laevis
to
a
single
concentration
of
atrazine
(
21

g/
L)
increased
the
frequency
of
secondary
oogonia
and
atresia
of
both
primary
and
secondary
oogonia
in
ovaries
(
Tavera­
Mendoza
et
al.
2001a)
and
reduced
testicular
volume,
spermatogonia,
and
nurse
cells
in
testes
(
Tavera­
Mendoza
et
al.
2001b).

Sanderson
et
al.
(
2002)
reported
that
atrazine
up­
regulated
the
cytochrome
P450
CYP­
19
gene
in
human
cancer
cell
lines
and
increased
synthesis
of
aromatase.
According
to
the
authors,

increased
aromatase
levels
could,
in
turn,
increase
the
baseline
conversion
of
testosterone
to
estrogen.
Additional
studies
published
by
Hayes
et
al.
(
2002b,
c)
in
Environmental
Health
Perspectives
and
Nature
reported
that
similar
effects
on
gonadal
development
were
achieved
using
the
non­
native
X.
laevis
and
the
native
leopard
frog
(
Rana
pipiens).
Finally,
field
studies
Page
­
16­
reported
that
leopard
frogs
collected
in
atrazine­
exposed
sites
exhibited
testicular
oogenesis
and
hermaphroditism
at
rates
as
high
as
92%
(
Hayes
et
al.
2002b,
c).
In
response
to
this
published
literature,
the
registrant
(
Syngenta
Crop
Protection,
Inc.)
sponsored
a
series
of
studies
to
further
examine
the
potential
for
atrazine
to
affect
development
in
both
native
and
non­
native
frog
species.
The
primary
focus
of
these
registrant­
sponsored
studies
was
to
examine
gonadal
development,
laryngeal
muscle
growth,
blood
testosterone
and
estrogen
levels,
and
gonadal
and/
or
brain
aromatase
activity
in
frogs
exposed
to
atrazine.

The
intent
of
this
white
paper
is
to
critically
review
the
available
open
literature
and
registrant­
sponsored
studies
that
examine
the
effects
of
atrazine
on
amphibians
and
to
evaluate
whether
there
is
sufficient
information
to
conclude
that
atrazine
exposure
does
or
does
not
result
in
toxic
effects
on
amphibian
development.
In
this
white
paper,
Chapter
1
provides
an
overview
of
the
regulatory
history
of
atrazine,
the
Agency's
process
for
evaluating
information
used
in
characterizing
potential
risks
associated
with
the
use
of
pesticides,
and
the
exposure/
effects
characterization
of
atrazine.
In
Chapter
2,
actual
study
reviews
are
then
grouped
into
laboratory
and
field
investigations;
study
methods
and
results
are
presented
followed
by
a
discussion
of
specific
issues
associated
with
each
study.
Following
these
reviews,
Chapter
3
discusses
common
strengths
and
limitations
found
in
the
laboratory
and
field
studies.
Based
on
the
study
reviews
and
their
associated
strengths
and
limitations,
major
uncertainties
are
identified
in
Chapter
4
that
could
limit
the
usefulness
of
the
studies
in
charactering
the
potential
developmental
effects
of
atrazine
on
amphibians
for
the
purposes
of
risk
assessment.
Chapter
5
provides
recommendations
on
how
various
uncertainties
could
be
addressed,
and
Chapter
6
presents
the
conclusions
of
the
paper.

Finally
in
Chapter
7,
EPA
poses
a
series
of
questions
that
require
input
from
the
SAP
on
the
Agency's
study
evaluations
and
recommendations
to
reduce
uncertainties
related
to
establishing
a
causal
relationship
between
atrazine
exposure
and
effects
on
amphibian
development.
Page
­
17­
CHAPTER
2
STUDY
REVIEWS
In
response
to
studies
published
in
the
open
literature
regarding
the
potential
effects
of
atrazine
on
amphibian
development,
Syngenta
sponsored
a
series
of
studies
conducted
by
the
Atrazine
Endocrine
Ecological
Risk
Assessment
Panel
administered
by
Ecorisk,
Inc.
of
Ferndale,

WA.
According
to
the
registrant,
the
studies
were
initiated
to
investigate
potential
endocrine
effects
of
atrazine
in
wildlife
species,
including
amphibians.
The
studies
were
submitted
by
February
28,
2003
and
were
provided
specifically
for
consideration
at
the
FIFRA
Scientific
Advisory
Panel's
meeting
in
June
2003.

In
the
following
sections,
summary
reviews,
grouped
into
laboratory
and
field
studies,
are
presented
for
each
individual
study.
Each
study
was
individually
evaluated
with
regard
to
experimental
design,
protocols
and
data
quality
assurance,
strength
of
cause­
effect
and/
or
doseresponse
relationships,
mechanistic
plausibility,
and
ecological
relevancy
of
measured
endpoints.

Throughout
the
reviews,
the
terms
intersex,
hermaphroditism,
and
ovotestes
were
used
interchangeably
and
refer
to
situations
where
ovarian
and
testicular
tissue
were
observed
in
the
same
animal/
gonad.

Consistent
with
EPA's
process
for
evaluating
scientific
studies,
the
Agency
completed
data
evaluation
records
(
DERs)
for
each
of
the
Syngenta­
sponsored
studies
and
conducted
a
critical
review
of
the
pertinent
studies
reported
in
the
open
literature.
In
the
DERs
for
the
registrant­
sponsored
studies,
the
Agency
reviewers
outlined
the
study
methodologies,
re­
analyzed
raw
data,
and
documented
any
uncertainties
or
differences
in
conclusions
from
those
of
the
study
authors.
It
is
important
to
note
that
the
review
of
these
registrant­
sponsored
studies
was
more
detailed
than
for
the
studies
obtained
from
open
literature.
For
these
latter
studies,
the
reviews
were
less
detailed
because
EPA
did
not
have
access
through
the
study
authors
to
the
full
range
of
raw
data
and
quality
control
information
required
for
registrant­
submitted
studies.
Page
­
18­
Laboratory
Studies
One
of
the
advantages
of
conducting
laboratory
studies
is
that
they
allow
researchers
to
control
a
range
of
conditions
that
could
potentially
impact
the
outcome
of
a
study.
Environmental
factors,
water
quality,
loading
rate,
chemical
exposure,
study
animals,
animal
husbandry
and
health
can
all
be
manipulated
more
easily
to
identify
actual
treatment
effects
in
laboratory
studies.

Laboratory
studies
also
allow
greater
flexibility
in
study
design
to
account
for
known
sources
of
variability.
For
example,
sample
size
and
replication
can
be
manipulated
to
reduce
confidence
intervals
around
treatment
means.
Another
advantage
of
laboratory
studies
is
that
researchers
can
use
positive
controls
to
gauge
the
responsiveness
of
the
test
organisms
to
treatment
effects.

Finally,
laboratory
studies
facilitate
sample
processing
since
analyses
can
be
conducted
in
close
proximity
to
where
samples
are
collected.

This
chapter
includes
a
review
of
seven
laboratory
studies,
three
taken
from
open
literature
and
the
remainder
sponsored
by
the
registrant.
A
summary
of
important
attributes
of
these
studies
is
presented
in
Table
1.
Although
Table
1
depicts
nine
studies,
two
of
the
studies
were
field
studies
that
had
laboratory
components.
Page
­
19­

Table
1.
Summary
of
basic
conditions
and
findings
of
various
laboratory
or
controlled
exposure
studies
on
the
effects
of
atrazine
on
amphibian
development.

Study
First
Author
Study
Number
or
Citation
Study
Type
Species
Length
of
Exposure
Developmental
Stage
Nominal
Exposure
Concentrations
(
µ
g/
L)
Primary
Endpoints
Significant
Atrazine
Effects
as
Determined
by
Authors
and
Associated
Atrazine
Concentration
Effects
(
µ
g/
L)
3
Hayes
PNAS,

99:
5476­
5480
static
renewal
X.
laevis
nr
initial:
NF46
final:
NF66
0.01,
0.1,
1.0,
10,
25
0.1,
0.4,
0.8,
1.0,
25,
200
testicular
morphology
laryngeal
muscles

ovotestes

laryngeal
dilator
size
0.10
1.0
46
d
initial:
adult
final:
adult
25
plasma
testosterone

plasma
testosterone
25.0
Goleman
EPA
MRID:

458677­
07
static
renewal
X.
laevis
78
d
initial:
48
h
ph
final:
NF66
1.0,
10,
25
testicular
morphology
ovarian
morphology
laryngeal
muscles

ovotestes
25.0
Du
Preez
4
EPA
MRID:

458677­
11
static
microcosm
X.
laevis
133
d
initial:
120
h
pf
final:
juveniles
1.0,
10,
25
testicular
morphology
ovarian
morphology
none
na
Tavera­
Mendoza
ET&
C,

21:
527­
531
static
X.
laevis
48
h
initial:
NF56
final:
NF56
21
testicular
development

testes
size

spermatogonial
nests

nurse
cells
21.0
21.0
21.0
Tavera­
Mendoza
ET&
C,

21:
1264­
1267
static
X.
laevis
48
h
initial:
NF56
final:
NF56
21
ovarian
development

primary
oogonia

secondary
oogonia

atresia
21.0
21.0
21.0
Page
­
20­

Study
First
Author
Study
Number
or
Citation
Study
Type
Species
Length
of
Exposure
Developmental
Stage
Nominal
Exposure
Concentrations
(
µ
g/
L)
Primary
Endpoints
Significant
Atrazine
Effects
as
Determined
by
Authors
and
Associated
Atrazine
Concentration
Effects
(
µ
g/
L)
3
Hecker
EPA
MRID:

458677­
04
static
renewal
X.
laevis
185
d
initial:
72
h
ph
final:
juvenile
0.1,
1.0,
10,
25
testicular
morphology
ovarian
morphology
gonad
aromatase
brain
aromatase
plasma
estradiol
plasma
testosterone
laryngeal
muscles
none
na
Villeneuve
EPA
MRID:

458677­
08
static
renewal
X.
laevis
26,
43,
47
d
initial:
adult
final:
adult
25
gonad
aromatase
brain
aromatase
none
na
Hayes1
EHP,

111:
568­
575
static
renewal
R.
pipiens
nr
initial:
48
h
ph
final:
juvenile
0.1,
25
testicular
morphology

ovotestes

dysgenesis
0.10
0.10
Hecker
EPA
MRID:

458677­
03
static
renewal
R.
clamitans
273
d
initial:
NF462
final:
NF662
10,
25
testicular
morphology
ovarian
morphology
none
na
1
Also
reviewed
as
field
study
(
information
presented
in
table
represents
laboratory
data
only);
2NF
stage
used
for
convenience,
3
Nominal
concentration,

4Reviewed
as
field
study
Abbreviations:
nr:
not
reported;
na:
not
applicable;
ph:
post
hatch;
pf:
post
fertilization;
NF:
Nieuwkoop
and
Faber
Stage;
1Nieuwkoop
and
Faber
(
1994)
define
Stage
66
of
X.
laevis
development
as
the
point
at
which
the
tail
is
only
represented
by
small
dorsal
swelling
of
loose
connective
tissue,
covered
with
degererating
larval
skin;
the
tail
[
is]
only
a
very
small
triangle,
no
longer
visible
from
ventral
side.

2The
G­
test
(
or
log­
likelihood
ratio
test)
of
goodness­
of­
fit,
may
be
used
to
test
the
fit
of
data
in
treatments
or
classes
against
proportions
expected
from
control
group(
s)
or
from
historical
reference
data.
The
null
hypothesis
is
that
the
observed
frequencies
fit
a
particular
ratio
(
50:
50,
or
3:
1,
etc.);
a
significant
result
indicates
that
the
observed
frequencies
are
different
from
the
expected
ratio.

Page
­
21­
Hayes,
T.
B.,
A.
Collins,
M.
Lee,
M.
Mendoza,
N.
Noriega,
A.
A.
Stuart,
and
A.
Vonk.
2002a.

Hermaphroditic,
demasculinized
frogs
after
exposure
to
the
herbicide
atrazine
at
low
ecologically
relevant
doses.
Proceedings
of
the
National
Academy
of
Sciences
99
(
8):
5476
­
5480.

The
objective
of
this
study
was
to
test
whether
atrazine
interfered
with
metamorphosis
and
sex
differentiation
at
environmentally
relevant,
low
doses
via
endocrine­
disrupting
mechanisms.

Using
three
replicates
with
30
tadpoles
per
replicate,
Xenopus
laevis
were
exposed
under
static
renewal
conditions
(
complete
exposure
water
change
every
72
hrs)
to
nominal
atrazine
concentrations
ranging
from
0.01
to
200

g/
L
from
96­
hr
post­
hatch
through
complete
tail
resorption
(
Nieuwkoop­
Faber
(
NF)
Stage
661)
(
Nieuwkoop
and
Faber
1994).
Animals
were
maintained
in
plastic
containers
typically
used
to
house
laboratory
mice
(
personal
communication:

Tyrone
Hayes,
University
of
California
at
Berkeley,
2002);
containers
were
filled
with
4
L
of
exposure
solution.
At
the
end
of
the
exposure
period,
animal
growth
(
length
and
weight),
time
to
metamorphosis,
gonadal
abnormalities
and
size
(
cross­
sectional
diameter)
of
the
larynx
dilator
muscle
were
recorded.

Exposure
to
atrazine
concentrations

0.1

g/
L
resulted
in
gonadal
abnormalities
in
16
­

20%
of
the
animals.
The
actual
incidence
of
gonadal
abnormalities
at
each
exposure
level
was
not
reported.
For
this
reason,
the
quantitative
relationship
(
dose­
response)
between
atrazine
concentration
and
gonadal
effects
is
uncertain.
Abnormalities
included
multiple
gonads
or
hermaphrodites
(
multiple
testes
and
ovaries
in
the
same
animal);
these
abnormalities
were
not
observed
in
controls.
Although
males
typically
exhibited
larger
laryngeal
muscle
diameters
than
females,
atrazine
exposure

1

g/
L
significantly
decreased
(
G
test2,
p
<
0.05)
the
proportion
of
males
that
were
at
or
above
the
mean
control
for
males.
This
finding
suggested
a
threshold
effect
3The
Agency
did
not
receive
data
on
these
additional
studies
and
was
not
able
to
verify
these
conclusions.

Page
­
22­
at

1

g/
L
in
which
80%
of
the
exposed
males
had
below
average
laryngeal
muscle
diameters.

In
this
study,
Hayes
et
al.
(
2002a)
hypothesized
that
the
co­
occurrence
of
oocytes
and
testicular
tissue
(
hermaphroditism)
and
the
decreased
male
larynx
muscle
size
(
demasculinization)
were
consistent
with
increased
endogenous
estrogen
concentrations.
The
authors
proposed
that
one
possible
mechanism
for
increased
estrogens
would
be
through
increased
aromatase
activity.
In
support
of
this,
Hayes
et
al.
(
2002a)
reported
that
an
average
adult
male
Xenopus
exposed
to
atrazine
at
25

g/
L
had
significantly
reduced
plasma
testosterone.

This
study
states
that
the
results
have
been
repeatedly
verified,
but
additional
data
have
not
yet
been
provided
in
the
open
literature
or
submitted
to
EPA3
The
lack
of
a
dose­
response
relative
to
the
phenomenon
of
hermaphroditism
raises
problems
for
interpreting
cause­
effect
relationships.
Although
there
appears
to
be
a
dose­
dependent
reduction
in
laryngeal
muscle
area
relative
to
atrazine
concentrations,
the
reliance
on
the
proportion
of
animals
falling
below
average
is
an
indirect
measure
of
the
effect.
A
direct
comparison
of
measured
laryngeal
muscle
area
between
controls
and
exposed
animals
would
further
clarify
the
magnitude
and
extent
of
any
developmental
effects.
Information
on
how
diminished
dilator
muscle
area
or
the
gonadal
abnormalities
might
relate
to
the
reproductive
success,
growth
or
survival
of
the
affected
species
in
the
environment
would
also
provide
further
insights
on
the
ecological
relevancy
of
effects.

This
study
was
useful
in
identifying
a
potential
hazard
to
amphibians
and
presented
information
on
measurement
endpoints,
such
as
gonadal
deformities
and
laryngeal
muscle
diameter.
The
study,
however,
did
not
show
a
clear
dose­
response
that
demonstrates
a
causal
relationship
between
atrazine
exposure
and
amphibian
developmental
effects.
Page
­
23­
Tavera­
Mendoza,
L.,
S.
Ruby,
P.
Brousseau,
M.
Fournier,
D.
Cyr
and
D.
Marcogliese.
2001a.
Response
of
the
amphibian
tadpole
Xenopus
laevis
to
atrazine
during
sexual
differentiation
of
the
testes.
Environmental
Toxicology
and
Chemistry
21
(
3):
527
­
531.

In
an
effort
to
examine
the
effects
of
atrazine
on
gonadal
differentiation
and
reproductive
impairments,
male
X.
laevis
(
NF
Stage
56)
were
exposed
to
mean­
measured
concentrations
of
atrazine
at
18

g/
L
(
nominal:
21

g/
L)
under
static
conditions
for
48
hrs.
Animals
(
16
larvae
per
replicate
with
2
replicates
per
treatment)
were
fasted
during
the
48­
hour
test.
Total
testicular
volume
decreased
significantly
(
p
=
0.004)
from
0.026
±
0.003
mm3
in
controls
to
0.01
±
0.001
mm3
in
atrazine­
treated
tadpoles,
representing
a
57%
decrease
after
48
hrs
of
exposure.
The
number
of
spermatogonial
cell
nests
decreased
significantly
(
p
<
0.001)
from
a
mean
of
242.4
±
35.7
in
controls
to
72.9
±
21.8
in
atrazine­
exposed
tadpoles,
representing
a
70%
reduction.
The
number
of
nursing
cells
declined
significantly
(
p
<
0.001)
from
a
mean
of
9.62
±
0.17
in
controls
to
2.35
±
0.36
in
atrazine
treated
tadpoles.
Testicular
resorption
was
observed
in
70%
of
the
male
tadpoles
exposed
to
atrazine
relative
to
controls;
failure
of
full
development
of
the
testis
(
aplasia)
was
observed
in
10%
of
the
testes
examined.
Histological
examination
of
the
pituitary
suggested
that
tissues
were
actively
secreting
hormones
based
on
the
absence
of
chromophores.

This
study
provides
useful
information
on
hazard
identification
and
measurement
endpoints,
such
as
gonadal
abnormalities.
However,
the
study
does
not
provide
sufficient
information
to
establish
a
dose­
response
relationship
because
only
one
concentration
of
atrazine
was
tested.
Other
information
which
is
needed
to
more
fully
interpret
the
significance
of
this
study
include:
documentation
of
the
measurement
endpoints,
elaboration
on
the
ecological
relevancy
of
the
measurement
endpoints,
and
clarification
of
the
number
of
exposed
animals.
Page
­
24­
Tavera­
Mendoza,
L.,
S.
Ruby,
P.
Brousseau,
M.
Fournier,
D.
Cyr
and
D.
Marcogliese.
2001b.
Response
of
the
amphibian
tadpole
Xenopus
laevis
to
atrazine
during
sexual
differentiation
of
the
ovary.
Environmental
Toxicology
and
Chemistry
21
(
6):
1264
­
1267.

To
examine
the
effects
of
atrazine
on
gonadal
differentiation
during
larval
tadpole
development
of
female
X.
laevis,
NF
Stage
56
tadpoles
were
exposed
to
mean­
measured
concentrations
of
atrazine
at
18

g/
L
(
nominal:
21

g/
L)
under
static
conditions
for
48
hrs.

Animals
(
16
larvae
per
replicate
with
2
replicates
per
treatment)
were
fasted
during
the
48­
hour
test.
The
frequency
of
occurrence
of
primary
oogonia
was
significantly
(
p
<
0.05)
lower
in
atrazine­
exposed
(
43.7%)
tadpoles
relative
to
controls
(
74%);
however,
the
frequency
of
occurrence
of
secondary
oogonia
was
significantly
(
p<
0.05)
higher
in
atrazine­
exposed
(
36%)

tadpoles
compared
to
controls
(
23%).
The
incidence
of
atretic
primary
and
secondary
oogonia
was
significantly
higher
(
p
<
0.05)
in
atrazine­
exposed
ovaries
(
20.2%)
relative
to
control
(
2%).

Furthermore,
sections
of
the
pituitary
revealed
no
histological
evidence
that
the
pituitary
was
actively
secreting
hormones.
The
authors
concluded
that
atresia
could
reduce
the
reproductive
capacity
of
the
tadpole
since
primary
germ
cells
provide
oocytes
for
all
the
subsequent
cycles
of
oogenesis
in
the
reproductive
life
of
the
frog.
The
authors
also
speculated
that
atrazine
may
be
affecting
aromatase
activity;
however,
it
was
uncertain
whether
enhanced
conversion
of
androgens
to
estrogen
could
provide
the
effects
observed.

Similar
to
the
evaluation
of
Tavera­
Mendoza
et
al.
(
2001a),
this
study
provides
useful
information
on
hazard
identification
and
measurement
endpoints,
such
as
gonadal
abnormalities.

However,
this
study
also
contained
uncertainties
concerning
use
of
a
single
dose,
the
nature
of
the
measurement
endpoints,
and
the
relevancy
of
the
endpoint
to
higher
order
measures
of
organismal
fitness.
Page
­
25­
Hecker,
M.,
K.
K.
Coady,
D.
L.
Villeneuve,
M.
B.
Murphy,
P.
D.
Jones
and
J.
P.
Giesy.
2003.
A
Pilot
Study
of
Response
of
Larval
Rana
clamitans
to
Atrazine
Exposure:
Assessment
of
Metamorphosis
and
Gonadal
and
Laryngeal
Morphology
and
Selected
Hormones
and
Enzyme
Activities.
Aquatic
Toxicology
Laboratory,

Michigan
State
University,
National
Food
Safety
and
Toxicology
Center,
E.
Lansing,
MI.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
MSU­
03.
(
EPA
MRID
No.
458677­
03)

The
objectives
of
this
study
were
to
develop
and
validate
culture
and
test
methods
for
conducting
laboratory
studies
with
Rana
clamitans
and
to
determine
the
response
of
larval
R.

clamitans
to
atrazine
by
assessing
metamorphosis
and
reproduction
indices
when
animals
were
exposed
to
10
and
25

g/
L
during
larval
development.
In
this
study
the
following
indices
were
evaluated:
1)
percent
of
larvae
initiating
metamorphosis,
2)
percent
completing
metamorphosis,

3)
time
to
metamorphosis,
4)
fresh
post­
mortem
body
weight
and
snout­
vent
length,
5)
incidence
of
gross
gonadal
abnormalities,
and
5)
histology
of
the
gonads.
Beginning
five­
days
post­
hatch,

green
frog
tadpoles
reared
from
field­
collected
eggs
were
exposed
to
atrazine
for
273
days.

Positive
controls,
dihydrotestosterone
and
17­
 
estradiol
(
0.1

g/
mL
in
0.005%
ethanol),
a
dilution
water
control
and
a
solvent
control
(
0.005%
ethanol),
were
also
included.
Replicates
(
9)

consisted
of
30
free­
swimming
tadpoles
each.
From
exposure
days
0
­
67,
animals
were
maintained
under
static
renewal
conditions
in
4
L
of
test
solution,
with
50%
tank
changes
conducted
every
72
hrs.
From
days
68
to
273,
tadpoles
were
maintained
in
tanks
containing
16L
of
test
solution
under
static
renewal.
After
273
days,
exposures
were
terminated
and
tadpoles
were
maintained
in
continuous
flow­
through
10­
L
glass
tanks
housed
in
large
acrylic
tanks
containing
80
L
of
continuously
renewing
freshwater.

At
fore­
limb
emergence,
tadpoles
were
either
housed
individually
or
in
small
groups
in
10­
L
glass
tanks
containing
approximately
500
mL
of
fresh
water.
Over
the
study
period,

mortality
across
all
treatment
groups
averaged
76.5%,
which
was
attributed
to
poor
water
quality
and
overcrowding
during
the
273­
day
static­
renewal
phase
of
the
study.
Mean­
measured
concentrations
of
atrazine
were
relatively
consistent
with
nominal
values,
but
these
measurements
were
made
on
freshly
prepared
stock
solutions,
which
made
it
difficult
to
determine
atrazine
concentration
in
aged
exposure
solutions.
Measurable
levels
of
atrazine
were
detected
in
the
dilution
water
controls.
Although
the
concentrations
of
positive
control
hormones
were
not
Page
­
26­
measured,
the
positive
control
treatments
(
dihydrotestosterone
and
17­
 
estradiol)
suggested
that
green
frogs
only
reacted
to
androgenic
chemicals,
skewing
the
sex
ratio
to
predominantly
male
frogs
(
97.6%).
It
is
uncertain
if
these
results
indicate
that
green
frogs
were
unresponsive
to
estrogenic
chemicals
or
whether
there
was
insufficient
estradiol
in
solution
to
elicit
an
effect.

Based
on
gross
morphology,
no
hermaphroditism
(
testicular
and
ovarian
tissue
in
the
same
animal)
was
observed
in
any
of
the
treatment
groups.
There
were,
however,
difficulties
in
discerning
the
presence
of
gonads
using
this
process.
At
the
lower
dose
of
10

g/
L
atrazine,
time
to
and
age
at
metamorphosis
and
the
size
of
metamorphs
were
reduced
in
frogs,
while
at
the
higher
dose
of
25

g/
L
atrazine,
there
was
no
difference
in
these
same
parameters.

From
a
design
standpoint,
results
from
this
study
were
difficult
to
interpret
because
only
two
concentrations
of
atrazine
were
tested,
comparatively
few
frogs
survived
to
complete
metamorphosis,
gonadal
histology
and
aromatase
levels
were
not
provided,
and
only
one
tank
per
treatment
level
was
used.
The
means
to
establish
a
dose­
response
relationship
was
also
limited
by
the
use
of
two
exposure
concentrations.
Furthermore,
the
high
mortality
noted
in
this
study
may
be
indicative
of
poor
water
quality
and
overcrowding.
The
lack
of
response
to
the
positive
estradiol
control
and
the
presence
of
atrazine
in
the
dilution
water
control
may
have
confounded
interpretation
of
the
data
and
suggests
that
the
study
design
may
not
be
adequate
to
evaluate
the
potential
effects
of
atrazine
on
green
frogs.
Alternatively,
the
species
used
in
this
study
may
not
be
a
sensitive
model
for
examining
the
effects
of
atrazine
on
amphibian
development.

Villeneuve,
D.
L.,
K.
Coady,
M.
Hecker,
M.
B.
Murphy,
P.
D.
Jones
and
J.
P.
Giesy.
2003.
Methods
development
for
the
study
of
mechanism
of
action
of
atrazine
in
adult
and
metamophosing
Xenopus
laevis
and
Rana
clamitans:
aromatase
induction.
Aquatic
Toxicology
Laboratory,
Michigan
State
University,

National
Food
Safety
and
Toxicology
Center,
E.
Lansing,
MI.
Sponsor:
Syngenta
Crop
Protection,
Inc.,

Laboratory
Study
ID
ECORISK
Number
MSU­
01
(
EPA
MRID
No.
458677­
08)

In
this
study,
frogs
were
exposed
to
nominal
atrazine
concentrations
of
25

g/
L
to
test
whether
the
pesticide
could
up­
regulate
aromatase
activity
in
sexually
mature
male
and
female
X.

laevis
and
to
evaluate
whether
atrazine
would
decrease
plasma
testosterone
and
increase
estradiol
(
consistent
with
an
increase
in
aromatase
activity)
in
frogs.
In
three
separate
studies,
two
Page
­
27­
involving
adult
males
and
one
with
adult
females,
frogs
were
exposed
either
to
atrazine
or
to
fresh
water
under
static
renewal
conditions
with
50%
exposure
solution
changes
every
72
hrs
(
single
tank
per
treatment).
In
the
first
study,
15
males
were
exposed
for
26
days.
In
the
second
study,

15
males
were
exposed
for
43
days
(
single
tank
per
treatment),
and
in
the
third
study
13
females
(
six
in
one
replicate
and
seven
in
the
second
replicate)
were
exposed
for
47
days.
Overall
mortality
was
3,
7,
and
19%
in
the
26­,
43­
and
47­
day
exposures,
respectively;
mortality
was
primarily
associated
with
disease
(
fungal/
bacterial)
and
was
positively
correlated
(
r=
0.77)
with
the
number
of
exposure
days.

Homogenates
from
a
single
testes
and
from
brain
were
used
to
measure
aromatase
activity
in
males.
Aromatase
activity
in
the
testes
was
at
or
near
the
level
of
detection
(
LOD
=

0.025
fmol/
h/
mg
protein).
After
26
days
of
exposure,
mean
brain
aromatase
activity
from
atrazine­
exposed
males
(
8.4
±
4.2
fmol/
h/
mg
protein)
was
not
statistically
different
(
p
=
0.678)

from
the
controls
(
7.1
±
4.2
fmol/
h/
mg
protein).
Following
43
days
of
exposure,
mean
brain
aromatase
activity
in
atrazine­
treated
males
(
5.8
±
3.4
fmol/
h/
mg
protein)
again
was
not
statistically
different
(
p
=
0.199)
from
the
controls
(
10.4
±
7.1
fmol/
h/
mg
protein).
In
the
second
study,
atrazine
concentrations
as
high
as
0.25

g/
L
were
measured
in
the
dilution
water
control.

In
the
third
exposure
with
female
frogs,
ovarian
aromatase
activity
averaged
4.5
±
1.7
fmol/
h/
mg
protein
and
did
not
differ
statistically
(
p=
0.447)
from
controls
(
5.4
±
2.1
fmol/
h/
mg
protein).

Mean
aromatase
activity
of
brain
homogenates
from
atrazine­
exposed
females
was
7.3
±
5.0
fmol/
h/
mg
protein,
and
did
not
differ
significantly
(
p=
0.582)
from
control
females
(
8.9
±
4.2
fmol/
h/
mg
protein).

Tank
effects
were
difficult
to
document
in
the
first
two
studies
because
only
one
tank
was
used
per
treatment
level.
There
also
was
considerable
variability
in
aromatase
activity
between
frogs
receiving
the
same
treatment,
with
coefficients
of
variability
at
or
exceeding
100%.
Other
confounding
factors
included
the
use
of
only
one
atrazine
concentration
and
atrazine
contamination
of
the
controls
in
the
second
study.

The
authors
recommended
that
future
testing
use
replication,
higher
sample
sizes
and
a
broader
range
of
atrazine
levels
to
test
for
potentially
subtle
effects.
Although
not
discussed
in
the
report,
water
quality
may
have
also
compromised
the
study.
Evidence
in
support
of
this
Page
­
28­
includes
the
correlation
of
mortality
with
the
number
of
days
the
frogs
were
confined
and
the
references
to
the
diseased
state
of
the
dead
animals.
Apparently,
Xenopus
are
susceptible
to
bacterial
septicemia
(
red­
leg
disease)
if
poor
water
quality
persists
(
Sive
et
al.
1998).

Hecker,
M.,
K.
K.
Coady,
D.
L.
Villeneuve,
M.
B.
Murphy,
P.
D.
Jones
and
J.
P.
Giesy.
2003.
Response
of
Xenopus
laevis
to
Atrazine
Exposure:
Assessment
of
the
Mechanism
of
Action
of
Atrazine.
Aquatic
Toxicology
Laboratory,
Michigan
State
University,
National
Food
Safety
and
Toxicology
Center,
E.
Lansing,

MI.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
MSU­
04.(
EPA
MRID
No.
458677­
04).

The
goal
of
this
study
was
to
determine
the
effects
of
atrazine
on
metamorphosis
and
reproductive
indices
of
larval
X.
laevis
exposed
from
72
hrs
after
hatch
until
completion
of
metamorphosis.
Indices
evaluated
at
metamorphosis
included
percentage
of
animals
initiating
metamorphosis,
percentage
of
animals
completing
metamorphosis,
time
to
metamorphosis,
fresh
body
weight,
snout­
vent
length,
size
of
the
laryngeal
dilator
muscle,
and
gonad
development.

Additionally,
the
study
was
designed
to
determine
concentrations
of
circulating
hormones,

including
testosterone
and
estradiol,
in
control
and
atrazine­
treated
X.
laevis
and
to
measure
aromatase
activity
in
the
gonads
and
brain
tissue
of
control
and
atrazine­
exposed
X.
laevis.

In
this
study,
X.
laevis
larvae
were
exposed
to
atrazine
at
nominal
concentrations
of
0.1,

1.0,
10,
and
25

g/
L
in
frog
embryo
teratogenesis
assay
 
Xenopus
(
FETAX)
media
(
a
moderately
hard
reconstituted
water)
using
a
static
renewal
system
where
50%
of
the
exposure
solutions
were
changed
every
72
hrs.
Exposures
were
also
conducted
using
dilution
water,
positive
(
0.1

g/
L17­

 
estradiol
and
0.1

g/
L
dihydrotestosterone)
and
solvent
(
0.005%
ethanol)
controls.
Larvae
were
exposed
from
72­
hrs
post­
hatch
through
metamorphosis
(
NF
stage
66).
At
metamorphosis,

a
subset
(
number
not
reported)
was
euthanized;
gonads
were
examined
for
gross
morphology,
and
gonads
plus
the
larynges
were
prepared
for
histology.
The
remaining
animals
were
exposed
until
2­
3
months
post­
metamorphosis.
After
the
exposure
period,
half
the
frogs
were
examined
for
gross
morphology
and
the
other
half
(
50
frogs
per
treatment)
examined
for
histology
of
the
gonads.
One
frog
from
each
replicate
tank
(
64
frogs
total)
was
randomly
selected,
and
blood
(
drawn
by
cardiac
puncture),
brain,
and
gonads
were
collected
for
sex
steroid
hormone
and
Page
­
29­
aromatase
activity
assays.
Plasma
concentrations
of
testosterone
and
estradiol
were
measured
by
enzyme­
linked
immunosorbent
assay
(
ELISA),
while
a
tritium­
labeled
androstenedione
water
release
assay
was
used
to
measure
aromatase
in
brain
and
gonad
tissue.

The
authors
concluded
that
atrazine
treatment
did
not
affect
mortality,
time
to
metamorphosis,
sex
ratio,
gonadal
development,
aromatase
activity
or
steroid
hormone
plasma
concentrations
in
a
dose­
dependent
fashion.
They
also
concluded
that
estradiol
(
positive
control)

treatment
only
appeared
to
increase
estradiol
plasma
concentrations.
Dihydrotestosterone
(
positive
control)
increased
larynx
dilator
muscle
area
in
females,
and
neither
positive
control
influenced
sex
ratios.

Although
the
most
frequent
gonadal
abnormality
based
on
gross
morphology
was
discontinuous
gonads,
histology
indicated
that
mixed
sex/
intersex
(
ovarian
and
testicular
tissue
in
the
same
frog)
was
much
more
common
than
indicated
by
gross
morphology.
Since
histology
is
still
being
conducted,
the
authors
could
not
conclude
that
gonadal
abnormalities
were
treatmentrelated

Poor
water
quality
(
elevated
ammonia
and
nitrite
with
low
dissolved
oxygen)
probably
resulting
from
relatively
high
loading
rates
(
30
tadpoles/
4
liters
of
exposure
solution)
under
static
conditions
may
have
compromised
the
growth
and
development
of
the
test
animals.
On
average,

it
took
73
days
for
frogs
to
complete
metamorphosis,
as
opposed
to
58
days
reported
by
Nieuwkoop
and
Faber
(
1994),
and
17
(<
2%)
frogs
in
the
study
never
underwent
metamorphosis.

The
dilution
water
controls
had
measured
atrazine
concentrations
of
0.1

g/
L.
High
variability
(
coefficients
of
variation
ranging
as
high
as
524%)
associated
with
gonadal
aromatase
activity
and
with
plasma
steroid
hormone
concentrations
made
it
difficult
to
differentiate
treatment
effects.

Also,
estradiol
treatment
failed
to
skew
sex
ratios
significantly
in
favor
of
females
as
expected.
In
conclusion,
a
combination
of
tank
effects,
contaminated
controls,
prolonged
development
time,

high
variability,
and
lack
of
responsiveness
to
estradiol
limited
the
usefulness
of
this
study
in
differentiating
treatment
effects.
Page
­
30­
Goleman,
W.
L.
and
J.
A.
Carr.
2003.
Response
of
larval
Xenopus
laevis
to
Atrazine
Exposure:
Assessment
of
Metamorphosis
and
Gonadal
and
Laryngeal
Morphology.
The
Institute
of
Environmental
and
Human
Health,
Texas
Tech
University,
Texas
Tech
University
Health
Sciences
Center,
Lubbock,
Texas.
Sponsor:

Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
TTU­
01
(
EPA
MRID
No.
458677­

07).

Although
this
study
also
appeared
recently
in
Environmental
Toxicology
and
Chemistry
(
Carr
et
al.
2003),
the
raw
data
for
this
study
were
submitted
to
EPA
by
the
registrant
allowing
a
more
detailed
evaluation.
The
goal
of
this
study
was
to
determine
the
response
of
larval
X.
laevis
to
atrazine
by
assessing
metamorphosis
and
reproductive
indices
when
animals
were
exposed
from
48
­
72
hrs
after
hatching
through
the
completion
of
metamorphosis.
Indices
evaluated
included
percent
of
animals
initiating
metamorphosis,
percentage
of
animals
completing
metamorphosis,

time
to
metamorphosis,
percentage
of
intersex
gonads,
fresh
post­
mortem
body
weight,

snoutvent
length,
and
laryngeal
size.

In
a
78­
day
exposure,
48­
to
72­
hr
post­
hatch
X.
laevis
larvae
were
exposed
to
nominal
concentrations
of
1,
10
and
25

g
atrazine/
L
in
FETAX
medium,
FETAX
medium
alone
(
dilution
water
control),
0.1

g/
L
17­
 
estradiol,
0.1

g/
L
dihydrotestosterone,
or
solvent
control
(
0.0025%
ethanol)
in
a
static
renewal
system
where
50%
exposure
solution
water
changes
occurred
every
72
hrs.
For
the
first
seven
days,
60
­
65
larvae
were
maintained
in
100
mL
of
exposure
solution.
At
day
14,
animals
were
transferred
to
1
L
of
exposure
solution,
and
by
Day
21,
animals
were
maintained
in
4
L
of
exposure
solution.
At
NF
Stage
66
animals
were
weighed,

measured
for
snout­
vent
length
and
examined
for
gonadal
gross
morphology.
Larynx
and
gonads
also
underwent
histological
analysis.

Mortality
over
the
study
period
ranged
from
10
­
14%
for
those
animals
that
reached
stage
66
by
80
days
post­
hatch.
Time
to
complete
metamorphosis
did
not
differ
significantly
across
treatments
although
the
specific
time
was
not
reported.
In
all
treatments,
weight
and
snout­
vent
length
were
inversely
proportional
to
the
number
of
days
required
to
complete
metamorphosis,

i.
e.,
animals
completing
metamorphosis
early
tended
to
be
larger
than
animals
that
took
longer
to
complete
metamorphosis.
Sex
ratios
ranged
from
48
­
50%
male
across
all
treatments
except
for
the
estradiol
treatment,
which
significantly
skewed
the
ratio
in
favor
of
females
(
67%;
statistical
Page
­
31­
significance
not
reported).
While
the
incidence
of
intersex
was
significantly
correlated
(
p
=

0.0003)
with
atrazine
exposure
concentrations,
only
the
25

g
atrazine/
L
(
4.7%)
and
estradioltreated
(
7.4%)
males
had
incidence
rates
significantly
different
(
p
=
0.0061
and
p=
0.01,

respectively)
from
dilution
water
(
0.6%
)
and
solvent
(
0.0%)
controls.
Intersex
in
males
treated
with
25

g
atrazine/
L
was
characterized
by
distinguishable
testicular
and
ovarian
tissue,
while
males
treated
with
estradiol
sometimes
had
ambiguous
tissue
structures.

There
was
no
difference
in
the
cross­
sectional
area
of
larynx
dilator
muscle
in
atrazinetreated
males
relative
to
dilution
water
controls.
Dihydrotestosterone­
treated
females
had
significantly
larger
(
p
<
0.0001)
cross­
sectional
dilator
muscle
areas
than
solvent
control
females.

In
this
study,
atrazine
did
not
impact
length,
weight,
time
to
metamorphosis
or
dilator
muscle
area
relative
to
the
controls;
however,
exposure
to
25

g
atrazine/
L
appeared
to
significantly
increase
the
number
of
intersex
males
and
animals
with
discontinuous
gonads.
The
observation
that
both
body
weight
and
length
were
inversely
correlated
with
the
length
of
time
to
complete
metamorphosis
suggested
that
animals
in
all
treatment
groups
were
developmentally
delayed.
In
addition,
17­
 
estradiol
treatment
only
resulted
in
67%
females,
further
suggesting
that
study
animals
may
not
be
totally
responsive
to
the
positive
control.
Reported
dissolved
oxygen
concentrations
did
not
drop
below
3.9
mg/
L;
however,
ammonia
levels
as
high
as
27
mg/
L
were
reported.
These
high
levels
of
ammonia
suggested
that
the
50%
static
renewal
and
loading
rates
(
number
of
tadpoles
per
liter
of
exposure
solution)
may
have
resulted
in
poor
water
quality,

which
could
have
slowed
development
of
the
animals.
In
support
of
this
possibility,
roughly
42%

of
the
animals
did
not
appear
to
have
reached
stage
66
by
Day
78,
suggesting
that
a
large
proportion
of
the
animals
were
not
developing
at
a
normal
rate.
Given
the
declining
condition
of
the
frogs
with
increased
length
of
time
to
maturity,
it
is
unclear
whether
these
animals
would
have
metamphosed
and/
or
survived.
Because
all
the
animals
in
the
study
had
not
undergone
metamorphosis,
the
percent
initiating
metamorphosis,
time
to
metamorphosis
and
percentage
of
gonadal
abnormalities
cannot
be
accurately
calculated
relative
to
the
total
animals
used
in
the
study.

In
the
context
of
hazard
identification,
Goleman
and
Carr
(
2003)
provides
additional
information
that
atrazine
exposure
could
potentially
cause
gonadal
developmental
effects
in
Page
­
32­
amphibians,
albeit
at
a
much
higher
concentration
than
reported
by
Hayes
et
al.
(
2002a;

0.1
ug/
L)
but
comparable
to
the
concentration
reported
by
Tavera­
Mendoza
et
al.
(
2001
a,
b;
18
ug/
L).
However,
Goleman
and
Carr
(
2003)
did
not
report
effects
on
secondary
sexual
characteristics,
i.
e.
laryngeal
muscle
diameter,
as
observed
in
other
amphibian
studies
(
Hayes
et
al.,
2002a).

Hayes,
T.,
K.
Haston,
M.
Tsui,
A.
Hoang,
C.
Haeffele
and
A.
Vonk.
2002b.
Atrazine­
induced
hermaphroditism
at
0.1
ppb
in
American
leopard
frogs
(
Rana
pipiens):
laboratory
and
field
evidence.

Environmental
Health
Perspectives.
http://
dx.
doi.
org/
[
Online
23
October
2002]

Because
components
of
this
study
were
conducted
both
in
the
laboratory
and
in
the
field,

EPA
has
described
its
review
in
both
sections
of
the
document.
The
laboratory
phase
of
the
study
is
discussed
here,
while
the
field
component
of
the
investigation
is
discussed
in
the
next
section
of
this
paper.

The
objective
of
this
study
was
to
investigate
whether
gonadal
effects
observed
in
previous
studies
of
non­
native
X.
laevis
exposed
to
atrazine
were
also
exhibited
in
a
native
species,
i.
e.,
R.
pipiens.
Leopard
frog
larvae
were
exposed
from
48­
hrs
post­
hatch
through
complete
tail
resorption
(
NF­
Stage
66)
to
nominal
atrazine
concentrations
of
0.1
and
25

g/
L
(
0.0036%
ethanol)
in
10%
Holtfreter's
solution.
Following
exposure,
animals
were
sacrificed.

Gross
morphology
and
histological
analysis
of
gonads
revealed
that
36%
and
12%
of
the
males
treated
with
atrazine
at
0.1
and
25

g/
L,
respectively,
suffered
from
gonadal
dysgenesis
(

underdeveloped
testes
with
poorly
structured,
closed
lobules
and
low
to
absent
germ
cells).
Further,

29%
of
the
0.1

g/
L
and
8%
of
the
25

g/
L
animals
displayed
varying
degrees
of
sex
reversal;

testicular
lobules
of
sex­
reversed
males
contained
oocytes,
and
males
that
metamorphosed
later
contained
large
numbers
of
oocytes.
In
a
few
cases,
testicular
oocytes
were
reported
to
be
vitellogenic,
i.
e.,
to
contain
yolk.

The
study
suggested
that
the
enhanced
response
at
lower
doses
may
be
consistent
with
the
low­
dose
effect
reported
for
certain
endocrine­
disrupting
chemicals
although
the
data
did
not
show
a
clear
dose­
response
relationship.
In
a
previous
study
using
Xenopus
(
Hayes
et
al.
2002a),
Page
­
33­
there
appeared
to
be
a
threshold
effect
for
testicular
abnormalities
at
0.1

g/
L
atrazine
(
nominal);

however,
the
response
appeared
to
remain
steady
across
concentrations
of
atrazine
up
to
200

g/
L,
with
16
­
20%
of
the
males
exhibiting
gonadal
abnormalities.
However,
the
incidence
of
these
changes
at
each
dose
was
not
reported.
As
a
result,
the
authors
concluded
that
an
"
inverted­
U"
(
parabolic)
dose­
response
curve
appropriately
described
the
observations.

This
study
essentially
replicates
previous
work
conducted
by
Hayes
et
al.
(
2002a)
that
demonstrated
gonadal
developmental
effects
in
X.
laevis
and
thus
provides
additional
hazard
identification
data
on
potential
effects
in
the
native
leopard
frog.
However,
because
the
study
employed
two
atrazine
concentrations
(
nominal
concentrations
of
0.1
and
25
ug/
L),
there
is
not
sufficient
information
to
establish
a
dose­
response
relationship.

Field
Studies
In
addition
to
the
laboratory
studies,
EPA
evaluated
a
series
of
field
studies
ranging
from
pilot
reconnaissance
surveys
to
full
multi­
year
investigations.
While
field
studies
may
provide
useful
insights
on
"
real­
world"
responses
and
reflect
what
may
actually
occur
in
a
natural
setting,

they
can
be
subject
to
a
wide
range
of
environmental
effects
that
can
influence
a
study's
measurement
endpoints
beyond
treatment
effects.
Unlike
controlled
laboratory
studies
where
environmental
conditions
and
test
animal
homeostasis
are
artificially
maintained
to
some
extent,

field
studies
are
subject
to
a
wide
range
of
conditions
that
can
influence
the
fitness
of
organisms
being
studied.
In
addition,
under
natural
conditions,
organisms
are
potentially
exposed
to
a
wide
range
of
chemical
and
non­
chemical
stressors
simultaneously,
which
makes
the
elucidation
of
cause­
effect
and
dose­
response
relationships
difficult
to
establish.
Studies
measuring
the
effects
of
chemical
exposures
must
allow
for
the
potential
presence
of
a
wide
range
of
chemicals
in
the
field
and
the
possibility
that
these
chemicals
may
have
interactive
effects
that
could
be
antagonistic,
additive
or
synergistic.
In
addition,
the
presence
of
non­
chemical
stressors
(
e.
g.,

suspended
solids)
can
modulate
or
attenuate
potential
chemical
exposures
or
effects.
Thus,
while
field
studies
provide
an
opportunity
to
understand
how
laboratory
effects
may
be
expressed
under
more
natural
conditions,
these
studies
may
introduce
a
range
of
interpretation
challenges
that
Page
­
34­
need
to
be
considered.
Additionally,
the
wide
range
of
conditions
may
increase
the
variability
associated
with
measurement
endpoints,
making
it
difficult
to
detect
treatment
effects.
For
these
reasons,
variability
caused
by
non­
treatment
related
environmental
parameters
should
be
controlled
or
accounted
for
through
study
designs
that
include
a
sufficient
characterization
of
field
conditions
to
permit
statistical
analyses
of
direct,
indirect,
and
interactive
effects
and
typically
employ
large
sample
sizes.

EPA
evaluated
a
total
of
ten
field
studies
consisting
of
two
open
literature
studies
and
eight
Syngenta­
sponsored
studies.
Although
a
South
African
study
of
X.
laevis
was
a
composite
of
four
studies,
it
was
reviewed
collectively
as
a
single
study.
In
two
of
the
field
studies,
part
or
all
of
the
study
was
conducted
under
relatively
controlled
conditions
either
in
the
laboratory
or
in
an
outdoor
microcosm.
The
field
studies
were
conducted
on
a
range
of
species,
i.
e.,
cricket
frogs
(
Acris
crepitans),
X.
laevis,
R.
clamitans
,
bull
frogs
(
Rana
catesbiena),
and
cane
toads
(
Bufo
marinus)
in
a
wide
range
of
locales
(
Florida,
Indiana,
Iowa,
Michigan,
Nebraska,
Utah
and
Wyoming,
USA,
and
South
Africa).
Each
of
the
studies
was
conducted
in
the
native
range
for
the
species
and
in
many
cases
the
studies
represented
interim
reports
which
were
the
first
year
summary
of
a
planned
three­
year
study.

Hayes,
T.,
K.
Haston,
M.
Tsui,
A.
Hoang,
C.
Haeffele
and
A.
Vonk.
2002b.
Atrazine­
induced
hermaphroditism
at
0.1
ppb
in
American
leopard
frogs
(
Rana
pipiens):
laboratory
and
field
evidence.

Environmental
Health
Perspectives
http://
dx.
doi.
org/
[
Online
23
October
2002]

As
discussed
earlier,
this
investigation
included
two
components:
a
laboratory
phase
and
a
field
phase.
Field
studies
were
initiated
along
with
laboratory
studies
in
order
to
address
the
ecological
significance
and
relevance
of
the
initial
laboratory
studies
discussed
previously.
The
objective
of
the
field
component
phase
of
the
study
was
to
determine
if
the
effects
of
atrazine
on
leopard
frogs
observed
under
controlled
laboratory
conditions
could
also
be
observed
in
wild
R.

pipiens
from
a
variety
of
habitats
in
areas
with
reportedly
low­
and
high­
atrazine
use.
In
this
field
study,
water
samples
were
collected
at
each
site
to
determine
atrazine
exposure.
Page
­
35­
In
the
field
reconnaissance
survey
of
leopard
frogs,
frogs
were
sampled
from
a
total
of
eight
sites
with
the
assumption
that
low­
and
high­
atrazine
sales
data
would
be
indicative
of
low­

(
four
sites)
and
high­
(
four
sites)
atrazine
exposures.
Aqueous
atrazine
concentrations
were
quantified
by
gas
chromatography/
mass
spectrophotometry
with
a
level
of
detection
of
0.1

g/
L
in
100­
mL
water
samples
at
each
of
the
eight
sampling
sites.
It
should
be
noted,
though,
that
in
this
study
it
was
difficult
to
determine
if
atrazine
concentrations
reflected
the
true
exposure
conditions
during
larval
development
because
the
study
did
not
include
information
on
when
the
animals
underwent
metamorphosis.
Field
collection
of
frogs
and
water
samples
proceeded
eastward,
starting
in
Utah
on
July
15,
2001
and
ending
in
Iowa
on
July
28,
2001.
At
each
site,

the
researchers
collected
100
leopard
frogs
for
histological
analysis
and
100
mL
of
water
for
chemical
analysis.

The
researchers
identified
testicular
oocytes
in
males
from
seven
of
the
eight
collection
sites.
All
sites
with
atrazine
levels
exceeding
0.2

g/
L
had
males
that
displayed
sex­
reversal
similar
to
those
abnormalities
induced
by
atrazine
in
the
laboratory
study.
The
highest
incidence
(
92%)
and
most
advanced
cases
of
hermaphroditism
were
observed
in
animals
collected
from
the
North
Platte
River
in
Wyoming
where
the
measured
atrazine
concentrations
were
lower
than
at
most
of
the
other
sites.
At
sites
with
similar
atrazine
concentrations,
e.
g.,
0.2

g/
L,
the
incidence
of
gonadal
abnormalities
ranged
from
9
to
92%
suggesting
that
there
was
no
clear
pattern
of
response.
At
sites
where
no
atrazine
was
found,
the
incidence
of
testicular
oogenesis
appeared
to
be
as
high
as
18%.

This
study
was
useful
in
identifying
field
effects
similar
to
those
observed
in
laboratory
studies
of
the
leopard
frog.
Like
the
laboratory
study,
this
study
was
unable
to
establish
a
quantitative
dose­
response
relationship.
Notably,
atrazine
was
often
measured
where
it
was
not
reportedly
used.
Moreover,
it
is
likely
that
other
chemical
contaminants
were
also
present
but
not
measured.
Although
the
researchers
conducted
pesticide
residue
analyses
for
chemicals
that
were
reportedly
used
in
the
watershed,
the
scope
of
the
chemical
analyses
varied
on
a
site­
by­
site
basis.

Information
concerning
the
chemical
profile
and
morphological
characteristics
of
the
sampling
sites
needs
to
be
provided
in
order
to
determine
the
comparability
of
the
associated
aquatic
ecosystems.
Page
­
36­
DuPreez,
L.
H.,
A.
M.
Jooste
and
K.
R.
Solomon.
2003.
Exposure
of
Xenopus
laevis
larvae
to
different
concentrations
of
atrazine
in
semi­
natural
microcosms.
School
of
Environmental
Sciences
and
Development,

Zoology
Department,
Potchefstroom
University
fo
CHE,
Potchefstroom
2520,
South
Africa.
Sponsor:

Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
SA­
01­
D
(
EPA
MRID
No.
458677­
11).

This
report
represents
an
interim
summary
of
a
longer­
term
investigation.
The
objective
of
this
study
was
to
assess
the
effects
of
atrazine
exposure
on
gonadal
abnormalities
in
X.
laevis
metamorphs
and
sub­
adults
under
semi­
natural
conditions
in
microcosms.
Based
on
the
absence
of
atrazine
use
and
corn
production
in
the
watershed
and
the
presence
of
X.
laevis,
adult
frogs
were
collected
from
a
Potchefstroom,
South
African
earthen
pond
C6,
which
was
considered
to
be
reference
site.
Adults
were
induced
to
spawn
in
the
laboratory
and
the
96­
hr
old
progeny
were
divided
among
four
nominal
treatments
(
0,
1,
10
and
25

g/
L
atrazine)
with
three
replicates
per
treatment
(
800
tadpoles
per
replicate).
Mean
measured
concentrations
of
atrazine
ranged
from
0.9
to
1.82

g/
L,
10
to
15.9

g/
L,
and
23.8
to
39.7

g/
L
for
the
1,
10,
and
25

g/
L
nominal
treatments,
respectively.
Atrazine
in
the
reference
tanks
ranged
from
0
to
0.1

g/
L.
Microcosms
consisted
of
1600­
L
cement
tanks,
lined
with
polyethylene
with
a
3­
cm
sand
bottom,
and
filled
with
1,100
L
of
tap
water.
Larvae
were
exposed
until
they
reached
NF
stage
66;
the
study
was
terminated
after
133
days
of
exposure.
A
total
of
150,
NF
stage
66
frogs
were
sampled
per
treatment.
Snout­
vent
length,
weight,
days
to
NF
stage
66
were
recorded
and
animals
were
examined
for
gross
gonadal
morphology.

A
series
of
figures
depicted
water
temperature,
pH,
dissolved
oxygen
(
DO)
and
conductivity
over
the
study
period.
Dissolved
oxygen
concentrations
fluctuated
from
1
to
8.0
mg/
L
over
the
course
of
the
study.
From
day
90
to
100,
conductivity,
temperature
and
pH
all
dropped
markedly.
Conductivity
in
most
of
the
exposure
tanks
tended
to
increase
from
days
0
to
90;
however,
tanks
2,4,
6
and
9
had
erratic
conductivity
with
marked
declines.
In
tanks
4
and
6,

conductivity
dropped
by
roughly
40%
on
day
40,
and
in
tank
2
conductivity
declined
by
a
similar
amount
by
day
60.
Page
­
37­
Although
some
animals
reached
stage
66
by
70
days,
most
did
not
reach
metamorphosis
until
126
­
133
days.
Xenopus
laevis
tadpoles
took
58
days
to
complete
metamorphosis
under
controlled
laboratory
conditions
of
20
­
25oC
(
Nieuwkoop
and
Faber
1994).
Under
natural
conditions,
where
water
temperatures
may
fluctuate
widely,
metamorphosis
may
take
from
56
to
63
days,
which
is
still
a
shorter
period
of
time
than
the
70
days
in
the
current
study.
The
authors
attributed
low
water
temperatures
to
delayed
development
of
the
tadpoles.
Based
on
gross
morphology,
the
incidence
of
gonadal
deformities
in
1,
10
and
25

g/
L
atrazine
groups
was
1.3,

0.7
and
3.3%
of
the
total
frogs
examined
(
150),
respectively;
reference
frogs
exhibited
a
4%

incidence
of
gonadal
deformities.
Discontinuous
testis
was
the
only
gonadal
abnormality
identified;
no
abnormalities
were
observed
in
the
ovaries
except
for
one
ovary
that
was
reduced
in
size.
Males
comprised
48,
39,
and
47%
of
the
1.0,
10,
and
25

g/
L
atrazine­
treated
samples,

respectively,
while
the
reference
samples
were
45%
male.
In
addition,
male
and
female
frogs
not
exposed
to
atrazine
were
larger,
in
terms
of
length
(
p<
0.03)
and
weight
(
p<
0.02),
than
their
atrazine­
treated
counterparts.

This
interim
microcosm
study
represents
a
reasonable
step
forward
in
testing
for
effects
observed
in
the
laboratory
and
may
provide
useful
information
when
all
the
analyses
are
completed.
The
present
study,
though,
showed
that
developmental
rates
across
all
treatments,
in
terms
of
time
to
metamorphosis,
were
delayed.
Fluctuating
water
quality
in
the
microcosm
units
may
have
impacted
the
developmental
rate.
Additionally,
the
study
was
designed
with
the
assumption
that
phytoplankton
growth
would
serve
as
a
source
of
food
for
developing
tadpoles.

Although
the
authors
note
that
phytoplankton
"
flourished",
no
measure
(
e.
g.,
chlorophyl
a
concentration)
of
phytoplankton
growth
was
reported
and
supplemental
food
(
rabbit
pellets)
had
to
be
provided
to
the
developing
tadpoles.
However,
phytoplankton
growth
could
have
been
significantly
limited
by
atrazine
treatment,
thereby
resulting
in
indirect
effects
of
atrazine
on
development.
In
support
of
this
possibility,
atrazine­
treated
male
and
female
frogs
were
significantly
smaller
in
terms
of
length
and
weight
than
their
"
untreated"
counterparts,
further
complicating
interpretation
of
the
study
results.
This
assessment
was
based
on
gross
morphology
alone,
and
the
results
may
change
when
histology
is
completed.
Page
­
38­
Reeder,
A.
L.,
G.
L.
Foley,
D.
K.
Nichols,
L.
G.
Hansen,
B.
Wikoff,
S.
Faeh,
J.
Eisold,
M.
B.
Wheeler,
R.

Warner,
J.
E.
Murphy,
and
V.
R.
Beasley.
1998.
Forms
and
Prevalence
of
Intersexuality
and
Effects
of
Environmental
Contaminants
on
Sexuality
in
Cricket
Frogs
(
Acris
crepitans).
Environmental
Health
Perspectives
106
(
5):
261
­
266
To
assess
the
prevalence
of
gonadal
abnormalities
in
adult
and
juvenile
cricket
frogs,
and
to
determine
whether
sexual
development
is
altered
in
response
to
exposure
to
environmental
contaminants,
cricket
frogs
were
collected
over
a
three­
year
period
(
1993
­
1995)
in
various
locations
throughout
the
state
of
Illinois.
Additionally,
water/
sediment
samples
were
collected
at
sampling
sites
in
1994
and
1995
to
determine
whether
the
prevalence
of
intersex
could
be
related
to
chemical
residues.
In
a
separate
study,
cricket
frogs
were
collected
at
a
site
known
to
be
contaminated
with
PCBs
and
PCDFs,
and
the
prevalence
of
intersex
was
determined
relative
to
control
sites.

Of
the
55
adult
and
juvenile
male
and
female
frogs
collected
in
1993,
two
(
3.6%)
had
both
an
ovary
and
testis.
In
the
testis
of
one,
spermatogenesis
was
normal;
in
the
other,
an
immature
ovary
was
present
as
well
as
a
testis
with
no
active
spermatogenesis.
Of
the
243
frogs
examined
in
1994,
six
(
2.5%)
contained
both
an
ovary
and
a
testis,
and
five
of
the
affected
animals
had
areas
of
normal
spermatogenesis
in
the
testis
interspersed
with
oocytes.
One
animal
had
a
mature
ovary
and
mature
testis
with
normal
spermatogenesis.
Of
the
43
frogs
examined
in
1995,
only
one
(
2.3%)
had
an
ovotestis.
Across
all
three
sampling
years
the
prevalence
of
intersex
was
2.8%.
In
specimens
with
an
ovary
on
one
side
and
a
testis
on
the
other,
ovarian
size
ranged
from
welldeveloped
mature
female
size
to
extremely
small
ovaries
with
a
few
oocytes
present.

Of
the
five
sites
where
intersex
organisms
were
found,
four
had
detectable
atrazine
(
limit
of
detection:
0.5

g/
L).
Of
the
four
sites
where
no
intersex
organisms
were
collected,
only
one
contained
detectable
levels
of
atrazine.
According
to
the
authors,
the
relationship
between
detection
of
atrazine
and
prevalence
of
intersex
approached
significance
(
p
=
0.07).
At
one
site
treated
with
copper
sulfate
in
1994,
one
out
of
33
frogs
collected
had
an
ovotestis.
In
1995,
no
relationship
was
found
between
the
detection
of
atrazine
and
the
prevalence
of
intersex.
No
intersex
was
identified
in
frogs
collected
from
a
pond
treated
with
endothall
in
1995.

Concentrations
of
lead
from
both
years
were
not
associated
with
intersex.
Page
­
39­
Of
the
frogs
collected
from
PCB
and
control
sites,
one
frog
with
an
ovotestis
was
identified
from
the
control
site.
Sex
ratios
differed
significantly
(
probability
not
given)
between
contaminated
and
control
sites.
In
13
juveniles
from
control
and
13
from
contaminated
sites,

gonadal
tissue
was
immature
and
could
not
be
identified
for
histological
preparation.
The
association
between
sex
ratios
of
PCB/
PCDF
contaminated
and
control
groups
revealed
a
significant
difference
(
p
=
0.0007).

While
a
wide
range
of
chemical
residue
analyses
were
conducted,
only
atrazine
data
were
reported.
The
authors
suggested
that
there
may
be
a
trend
between
atrazine
and
the
proportion
of
animals
exhibiting
intersex;
however,
the
only
statistically
significant
relationship
was
between
sex
ratios
in
PCB/
PCDF
contaminated
sites
relative
to
controls.
It
should
be
noted
that
the
sample
size
for
making
this
determination
was
low,
with
three
control
and
three
contaminated
site
animals.

In
this
paper,
Reeder
et
al.
(
1998)
discussed
the
range
of
chemical
residues
in
the
field
collection
sites
and
how
these
chemicals,
combined
with
environmental
conditions,
could
impact
gonadal
development.
These
factors
contributed
to
the
limited
utility
of
this
study
because
the
investigation
did
not
demonstrate
a
significant
effect
of
chemical
residues
on
the
prevalence
of
intersex
in
cricket
frogs.
This
study
underscored
the
need
to
have
focused
study
designs
with
sufficient
power
in
terms
of
sample
size
to
discriminate
effects
if
they
exist.
Also,
the
report
acknowledged
that
little
is
known
about
natural
intersex
rates
in
cricket
frogs.
Without
a
better
understanding
of
the
biology
of
the
cricket
frog
and
the
toxicological
phenomenon
being
examined,
it
is
difficult
to
interpret
the
significance
of
the
reported
observations.
Page
­
40­
South
African
X.
laevis
Field
Studies:

Smith,
E.
E.,
L.
DuPreez,
and
K.
Solomon.
2003.
Field
exposure
of
Xenopus
laevis
to
atrazine
and
other
triazines
in
South
Africa:
feasibility
study
for
site
characterization
and
assessment
of
laryngeal
and
gonadal
responses.
The
Institute
of
Environmental
and
Human
Health,
Texas
Tech
University,
Lubbock,
Texas
(
USA)
and
School
of
Environmental
Sciences
and
Development,
Potchefstroom
University
for
CHE,

Potchefstroom,
South
Africa.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID:
ECORISK
Number
SA­
01A.

Smith,
E.,
L
DuPreez
and
K.
Solomon.
2003.
Field
exposure
of
Xenopus
laevis
to
atrazine
and
other
triazines
in
South
Africa:
exposure
characterization
and
assessment
of
laryngeal
and
gonadal
responses.

The
Institute
of
Environmental
&
Human
Health,
Texas
Tech
University,
Lubbock,
Texas
79490
(
USA)
and
School
of
Environmental
Sciences
and
Development,
Potchefstroom
University
for
CHE,
Private
Bag
X6001,

Potchefstroom
2520
(
South
Africa).
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
SA­
01B
Smith,
E.
E.,
L.
DuPreez,
and
K.
Solomon.
2003.
Gonadal
and
laryngeal
responses
to
field
exposure
of
Xenopus
laevis
to
atrazine
in
areas
of
corn
production
in
South
Africa.
The
Institute
of
Environmental
and
Human
Health,
Texas
Tech
University,
Lubbock,
TX
and
School
of
Environmental
Sciences
and
Development,
Potchefstroom
University
for
CHE,
Potchefstroom,
South
Africa.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID:
ECORISK
Number
SA­
01C.

Giesy,
J.
P.,
M.
Hecker,
and
P.
D.
Jones.
2003.
South
African
Analytical
Support
 
Hormone
and
Aromatase
Analysis
(
SA­
01C).
Aquatic
Toxicology
Laboratory,
Michigan
State
University,
National
Food
Safety
and
Toxicology
Center,
E.
Lansing,
MI.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
MSU­
07
The
objective
of
these
studies
was
to
examine
the
effects
of
atrazine
on
X.
laevis
in
its
native
habitat
(
South
Africa).
Initially
these
studies
were
intended
to
test
whether
morphological
and
biochemical
differences
existed
between
clawed
frogs
in
atrazine­
exposed
(
experimental)

versus
non­
exposed
(
reference)
ponds.
The
criteria
for
differentiating
reference
and
experimental
sites
included
production
of
corn
and
use
of
atrazine
in
the
vicinity,
plus
the
presence
of
X.
laevis
in
a
pond.
Based
on
an
initial
survey
of
the
sampling
area,
five
experimental
(
atrazine
exposure)
Page
­
41­
and
three
reference
(
no
atrazine
exposure)
sites
were
selected
(
458677­
09).
However,

subsequent
sampling
during
later
phases
of
the
study
revealed
that
the
reference
sites
all
contained
measurable
residues
of
atrazine,
its
degradates,
and
terbutyhylazine
(
triazine
herbicide
not
registered
for
use
in
the
USA)
that
were,
in
some
cases,
higher
than
sites
considered
representative
of
atrazine
exposure
(
458677­
01).

The
study
sites
were
subject
to
unusual
conditions,
including
abnormally
high
rainfall,

extremes
in
pH
(
10.2
­
10.8),
and
shifts
in
the
natural
mortality
rate
due
to
predation
by
the
introduction
of
sharptooth
catfish
(
Clarius
gariepinus).
Although
other
pesticides
besides
herbicides
were
used
at
these
sites,
no
data
were
presented
to
characterize
these
residues.

Monthly
terbuthylazine
residues
were
higher
at
some
reference
sites
than
at
experimental
sites,

and
some
reference
ponds
had
unusually
high
amounts
of
chromium
in
the
soil
(
100
mg/
kg)
and
titanium
in
the
water
(
0.6
mg/
L
)
(
458677­
01).
Depending
on
the
form
in
which
these
metals
were
in,
chromium
and
titanium
concentrations
at
these
levels
exceeded
the
reported
LC
50
s
for
these
metals
in
Fowler's
toad,
Bufo
fowleri,
and
the
Eastern
narrow­
mouth
toad,
Gastrophryne
carolinenis
(
Birge
et
al.
2000)
by
several
orders
of
magnitude.

In
this
study,
fewer
frogs
were
targeted
in
high
atrazine
exposure
(
atrazine
concentration
range:
1.46
­
11.6

g/
L)
sites
compared
to
low
atrazine
exposure
(
atrazine
concentration
range:

0.41
­
1.62

g/
L)
sites,
and
there
was
difficulty
in
obtaining
even
the
reduced
number
of
test
animal
at
experimental
sites.
For
this
reason,
the
animals
had
to
be
collected
over
an
extended
period
of
time
(
up
to
6
months).
The
collection
method
(
traps
baited
with
meat
scraps)
may
have
placed
adult
male
and
females
in
very
close
proximity
to
one
another
for
48
hrs.
As
opportunistic
breeders,
X.
laevis
held
in
close
proximity
to
one
another
may
undergo
marked
changes
in
sexual
readiness.
The
collection
method
may
have
also
provided
an
exogenous
source
of
steroid
hormones
via
the
feed.
Studies
have
shown
that
ground
beef
liver
can
produce
growth
inhibition
and
abnormalities
in
X.
laevis
larvae
(
Nieuwkoop
and
Faber
1994).

The
authors
concluded
that
there
were
no
differences
in
the
lengths
or
weights
of
either
males
or
females
collected
from
reference
(
low
atrazine)
and
experimental
(
high
atrazine
exposure)
sites
although,
at
both
sites,
females
were
larger
than
males.
Testes
from
frogs
collected
at
high
atrazine
sites
tended
to
weigh
more
than
testes
collected
from
frogs
at
reference
Page
­
42­
sites,
and
the
gonadosomatic
index
(
ratio
of
gonad
weight
to
total
body
weight)
tended
to
be
higher
for
males
at
experimental
sites
(
458677­
10).

Testicular
oocytes
were
observed
in
3%
of
the
reference
frogs
and
in
2%
of
the
frogs
from
experimental
sites.
No
other
significant
morphological
differences
were
observed
in
the
testes
collected
from
experimental
and
reference
sites.
(
458677­
10)

Males
from
ponds
with
the
highest
atrazine
concentrations
had
significantly
lower
plasma
median
testosterone
concentrations
than
males
from
reference
sites,
and
log
male
testosterone
levels
were
negatively
correlated
(
r=­
0.839;
p=
0.009)
with
the
log
of
the
atrazine
degradate,

diaminochlorotriazine
concentration.
Females
collected
at
high
atrazine
exposure
sites
had
significantly
higher
(
p=
0.018)
testosterone
levels
than
females
collected
at
reference
sites;
female
testosterone
levels
were
negatively
correlated
with
atrazine
surface
water
concentrations
(
458675­
01).
Similarly,
plasma
estradiol
concentrations
were
lower
in
males
and
females
collected
from
high
atrazine
sites.
Male
estradiol
concentrations
were
negatively
correlated
(
r=­
0.779;
p=
0.023)
with
diaminochlorotriazine
(
DACT),
while
female
estradiol
concentrations
were
negatively
correlated
(
r
=
­
0.833;
p
=
0.010)
with
concentrations
of
the
parent
atrazine
(
458675­
01).

In
this
study,
aromatase
activity
in
testes
could
not
be
quantified
at
all
sites;
thus,
no
comparisons
were
made
between
low
and
high
atrazine
exposure
sites.
Ovarian
aromatase
activity
was
not
significantly
different
between
sampling
sites.
The
highest
amount
of
variability
in
aromatase
was
observed
in
animals
collected
over
a
protracted
period
of
time,
which
underscores
how
the
extended
collection
period
influenced
the
results.
The
current
data
suggest
that
atrazine
and/
or
its
degradates
may
impact
plasma
testosterone;
however,
the
data
are
not
conclusive
and
the
mechanism
underlying
this
phenomena
cannot
be
identified.

Since
atrazine
was
present
to
some
degree
in
all
of
the
sampling
sites,
it
was
difficult
to
test
a
null
hypothesis
in
which
atrazine
was
assumed
to
be
absent
in
the
control
group.
The
presence
of
atrazine
in
all
of
the
ponds
suggested
that
a
regression
based
approach
to
data
analysis
may
be
appropriate,
but
this
type
of
analysis
would
require
a
substantially
different
experimental
design
for
optimal
power
to
differentiate
treatment
effects.
Of
course,
EPA
Page
­
43­
recognizes
that
controlling
critical
parameters
and
reducing
confounding
effects
is
one
of
the
major
difficulties
facing
the
conduct
of
all
field
studies.

Jones,
P.
D.,
M.
B.
Murphy,
M.
Hecker,
J.
P.
Giesy.
2003.
Tissue
Pesticide
Residues
and
Histology
of
the
Larynx
and
Gonads
in
Native
Green
Frogs
(
Rana
clamitans)
Collected
from
Agricultural
Areas
in
Michigan:

Hormone
Analysis.
Aquatic
Toxicology
Laboratory,
Michigan
State
University,
218C
National
Food
Safety
and
Toxicology
Center,
E.
Lansing,
MI.
Sponsor:
Syngenta
Crop
Protection,
Inc.
Laboratory
Study
ID:

ECORISK
Number
MSU­
02
(
EPA
MRID
No.
458677­
02).

The
objective
of
this
study
was
to
assess
the
effects
of
atrazine
on
kidney
and
gonad
histology
and
plasma
steroid
hormone
concentrations
and
gonadal
aromatase
activity
of
green
frogs
(
R.
clamitans)
and
other
incidental
ranid
species
collected
from
various
field
sites
within
their
native
Michigan
ranges.
For
Phase
I
of
a
three­
year
study,
R.
clamitans,
R.
pipiens,
and
R.

catesbeiana
were
collected
from
three
reference
ponds
(
atrazine
concentration
range:
0.015
to
0.093

g/
L)
and
six
atrazine­
exposed
ponds
(
atrazine
concentration
range:
0.025
to
250

g/
L)
in
Michigan.
Both
juvenile
(
372)
and
adult
(
340)
green
frogs
were
collected
and
examined
for
gross
gonadal
abnormalities,
and
blood
plasma
estradiol
and
testosterone
levels
were
measured.
A
total
of
four
mixed
or
unknown
sex
animals
were
identified
in
all
of
the
frogs
collected.
Because
hormone
levels
exhibited
considerable
variability
among
locations
and
individuals,
relationships
between
plasma
hormone
levels
and
atrazine
exposure
could
not
be
determined.

While
this
study
represents
an
interim
report,
the
preliminary
results
suggested
that
R.

clamitans
was
not
markedly
impacted
by
atrazine
exposure
in
terms
of
gonadal
deformities.

Because
the
reference
sites
may
have
contained
atrazine,
ecological
effects
could
not
be
accurately
discriminated.
Additionally,
plasma
testosterone
and
estradiol
levels
exhibited
considerable
variability,
and
differences
between
"
exposed"
and
"
reference"
sites
could
not
be
detected.
Plasma
steroid
levels
were
of
questionable
utility,
though,
since
data
were
collected
over
four
months
in
adult
frogs,
and
coefficients
of
variation
ranged
as
high
as
10,628%.
In
this
study,
there
were
no
gender­
specific
differences
in
plasma
estradiol
levels.
Exposed
males
contained
similar
(
roughly
90%)
amounts
of
estradiol
as
females,
while
at
the
reference
sites
males
exhibited
roughly
5.0
times
the
plasma
estradiol
concentration
than
females.
Page
­
44­
Because
of
the
variability
associated
with
the
measured
plasma
steroid
hormone
levels
and
the
presence
of
low
levels
of
atrazine
at
all
the
sites,
this
study
has
limited
usefulness
in
distinguishing
any
treatment­
related
effects
in
amphibians.

Sepulveda,
M.
S.
and
T.
S.
Gross.
2003.
Characterization
of
Atrazine
Exposures
and
Potential
Effects
in
Florida
Ecosystems
Dominated
by
Sugarcane
Agriculture:
A
Reconnaissance
Survey
of
Amphibians
in
South
Florida
for
the
Assessment
of
Potential
Atrazine
Effects.
Department
of
Physiological
Sciences,
University
of
Florida,
Caribbean
Science
Center,
Gainesville,
Florida.
Sponsor:
Syngenta
Crop
Protection,
Inc.
Study
ID:
ECORISK
Number
UFL­
02
(
EPA
MRID
No.
458677­
06)

The
goal
of
this
study
was
to
determine
whether
exposure
of
toads
to
atrazine
in
sugarcane
agricultural
areas
in
south
Florida
could
result
in
a
higher
incidence
of
intersex
and/
or
other
gonadal/
developmental
anomalies.
Secondarily,
the
study
was
conducted
to
examine
amphibian
populations
exposed
to
a
complex
mixture
of
agrochemicals,
e.
g.,
insecticides,

herbicides,
fungicides
and
fertilizers.
This
reconnaissance
survey
for
cane
toads
in
the
southern
Florida
sugarcane­
dominated,
agricultural
sites
(
Belle
Glade
and
Canal
Point)
and
in
nonagricultural
sites
(
University
of
Miami)
indicated
an
increased
incidence
of
intersex
(
ovarian
tissue
in
Bidder's
organ)
in
toads
identified
as
having
testes.
Approximately
29%
of
the
males
collected
from
Belle
Glade
and
39%
of
the
males
collected
from
Canal
Point
were
intersexed,

while
no
intersex
frogs
were
identified
among
the
University
of
Miami
samples.
Bufo
marinus
is
typically
a
sexually
dimorphic
amphibian;
however,
100%
of
the
cane
toads
collected
at
Belle
Glade
and
roughly
55%
of
the
male
cane
toads
collected
at
Canal
Point
exhibited
female
coloration.
Additionally,
males
typically
exhibit
nuptial
pads,
but
in
this
study
71%
and
0%
of
the
intersex
toads
collected
from
Canal
Point
and
Belle
Glade,
respectively,
had
nuptial
pads.

Vitellogenin,
a
female­
specific
protein,
is
not
typically
expressed
in
males;
however,
intersex
toads
had
vitellogenin
levels
(
774
±
29
PO
4/
mg
protein)
similar
to
the
females
(
853
±
34
PO
4/
mg
protein)
and
was
roughly
double
that
of
male
toads
(
375
±
34
PO
4/
mg
protein)
collected
from
the
nonagricultural
site.
Although
plasma
sex
steroids
(
17­
 
estradiol
and
testosterone)
were
relatively
gender­
specific;
testosterone
levels
in
intersex
males
exhibited
roughly
twice
the
amount
of
variability
as
similar
estimates
for
males.
The
data
showed
that
agricultural
sites
had
atrazine
Page
­
45­
concentrations
ranging
from
<
0.01
to
24.45

g/
L
over
the
6­
month
sampling
period,
but
atrazine
levels
were
not
measured
at
the
University
of
Miami
(
nonagricultural)
site.

In
this
same
study,
the
southern
toad
(
B.
terrestris)
was
also
examined
and
found
to
have
an
increased
incidence
of
intersex
(
Bidder's
organ
containing
ovarian
tissue)
in
both
agricultural
(
14%
at
Belle
Glade
and
22%
at
Fisheater
Creek)
and
nonagricultural
sites
(
33%
at
Archibald
Biological
Station).
The
authors
speculated
that
the
presence
of
a
Bidder's
organ
may
have
rendered
the
animals
more
sensitive
to
developmental
effects.
Bidder's
organ
is
characterized
as
a
nonfunctional,
rudimentary
ovary;
however,
no
information
is
available
on
whether
the
organ
has
an
endocrine
function
at
any
time
during
the
development
of
the
animals
(
Duellman
and
Trueb
1994;
Petrini
and
Zaccanti
1998).

This
study
was
useful
in
identifying
the
incidence
of
hermaphroditism
in
field­
collected
toads.
As
with
the
previous
studies,
toads
with
testes
also
appeared
to
have
ovarian
tissue,
but
unlike
previous
studies
the
ovarian
tissue
was
associated
with
the
Bidder's
organ
rather
than
the
testes.
While
toads
collected
in
agricultural
sites
may
have
been
exposed
to
atrazine
during
development,
it
is
unknown
whether
atrazine
was
present
at
the
nonagricultural
sites.
Based
on
the
study
design,
toads
were
also
exposed
to
other
agrochemicals
which
further
makes
it
difficult
to
establish
causality.
Finally,
underlying
environmental
conditions
relevant
to
development
of
toads,
e.
g.,
water
quality
characteristics,
were
not
reported,
which
further
limits
an
interpretation
of
the
findings.
Based
on
these
data,
it
is
difficult
to
conclude
that
atrazine
exposure
was
associated
with
an
increased
incidence
of
intersex.
Also,
the
study
does
not
provide
insights
on
the
ecological
relevance
of
the
data.
If
toads
depend
on
coloration
to
attract
mates,
though,
the
toads
from
agricultural
sites
may
have
impaired
ability
to
attract
mates
due
to
their
distinctly
female
appearance.
Page
­
46­
Crabtree,
C.;
E.
E.
Smith;
J.
A.
Carr.
2003.
Histology
of
the
gonads
and
analysis
of
hormone
levels
in
the
native
bull
frog
(
Rana
catesbiena)
collected
from
agricultural
areas
in
southern
Iowa:
pilot
project.
The
Institute
of
Environmental
and
Human
Health,
Texas
Technical
University,
Lubbock,
Texas.
Sponsor:

Syngenta
Crop
Protection,
Inc.
Laboratory
Identification
Number
ECORISK
Number
TTU­
02
(
EPA
MRID
No.
458677­
05).

This
study
presented
the
results
of
Phase
1
of
a
three­
phase
study
in
which
14
pond
sites
in
southern
Iowa
(
3
reference
and
11
atrazine­
exposed)
were
characterized.
The
overall
study
was
designed
to
select
sites
and
validate
biochemical
and
analytical
methods
and
organism
sampling
techniques
to
assess
the
effects
of
atrazine
on
kidney
and
gonad
histology
of
bullfrogs
(
R.
catesbiena)
and
other
species
collected
from
various
field
sites
in
southern
Iowa.

Experimental
sites
were
located
in
corn
and
soybean­
dominated
agriculture
areas,
while
reference
sites
were
located
in
wooded
or
grassland
areas.
Pond
sizes
ranged
from
0.14
to
2.94
ha
with
average
watershed
areas
ranging
from
2.19
to
84.02
ha.
Measured
atrazine
concentrations
in
reference
ponds
averaged
0.06

g/
L.
Mean
atrazine
concentrations
in
corn­
dominated
sites
over
the
June
to
September
time
period
ranged
from
1.07
to
19.26

g/
L
with
a
maximum
as
high
as
35.07

g/
L.
For
soybean­
dominated
watersheds,
the
highest
residues
ranged
from
3.19
to
3.85

g/
L.
Similarly,
maximum
deisopropyl
atrazine
residue
concentrations
were
highest
in
corndominated
areas
in
June/
July
at
4.17

g/
L.
Maximum
desethyl
atrazine
(
DEA)
residues
were
highest
in
corn­
dominated
ponds
at
16.55

g/
L
in
June/
July
and
16.10

g/
L
in
August.
During
the
sampling
period,
residues
of
DACT
remained
relatively
constant
and
the
highest
average
residues
(
0.65

g/
L)
were
from
corn­
dominated
ponds.

Although
bullfrogs
were
present
at
all
sites
in
sufficient
numbers
for
collection,
not
all
life
stages
were
collected
at
every
site.
No
significant
differences
were
found
for
adult
body
weight
or
snout­
vent
length
(
SVL),
but
mean
weight
and
SVL
for
juvenile
females
were
significantly
lower
in
reference
sites
than
atrazine­
exposure
sites
(
p=
0.001
and
p=
0.0001,
respectively).
Mean
SVL
for
juvenile
males
was
also
significantly
lower
in
reference
sites
than
in
atrazine­
exposure
sites
(
p
=
0.01);
however,
mean
weight
of
juvenile
males
was
not
statistically
different
between
sites.
Gonadosomatic
index
(
GSI
=
weight
of
gonad
÷
body
weight)
was
not
significantly
different
between
sites
for
either
adults
or
juveniles.
No
gross
gonadal
abnormalities
were
Page
­
47­
observed
based
on
visual
examination
(
gross
morphology),
and
the
incidence
of
external
abnormalities
was
less
than
1%
of
the
total
frogs
collected.

This
study
provides
information
showing
that
bullfrogs
did
not
exhibit
a
high
incidence
of
hermaphroditism
when
exposed
to
atrazine
under
field
conditions.
In
this
study,
the
number
of
water
samples
may
not
provide
sufficient
characterization
of
the
exposure
potential
to
atrazine,

particularly
in
reference
sites.
While
an
effort
was
made
to
characterize
a
limited
number
of
herbicides,
other
pesticides
were
not
measured.
Based
on
the
preliminary
results,
none
of
the
indices
measured
(
weight,
length,
GSI
or
the
incidence
of
gross
gonadal
deformities
)
in
the
bullfrog
indicated
that
variable
exposure
levels
to
atrazine
and
other
triazine
degradates
adversely
affected
this
species.
While
the
bullfrog
was
clearly
present,
interpreting
the
field
observations
was
difficult
because
information
on
the
bullfrog
relative
to
some
of
the
indices
of
interest
(
i.
e.,

steroid
hormone
levels,
aromatase
levels,
background
incidence
and
types
of
gonadal
abnormalities)
were
not
available.
Page
­
48­
CHAPTER
3
STUDY
STRENGTHS
AND
LIMITATIONS
Laboratory
Studies
The
laboratory
studies
reviewed
by
the
Agency
were
useful
in
identifying
measurement
endpoints
that
may
represent
potential
hazards
to
amphibians.
They
also
provided
valuable
insights
for
designing
future
studies
that
further
examine
these
measurement
endpoints.

Although
laboratory
studies
are
intended
to
provide
an
opportunity
to
control
potential
sources
of
variability
that
could
affect
the
endpoints
of
interest,
none
of
the
experiments
reviewed
by
the
Agency
fully
accounted
for
environmental
conditions
and
husbandry
factors
capable
of
influencing
endpoints
the
studies
were
attempting
to
measure.
Many
open
literature
studies
did
not
use
standardized
protocols
and
were
therefore
difficult
to
evaluate
in
the
absence
of
confirmatory
investigations.
In
instances
where
laboratory
experiments
used
protocols
and
standard
operating
procedures
that
were
in
existence
prior
to
initiating
the
studies,
the
observed
variability
associated
with
the
various
measurement
endpoints
suggested
that
study
designs
did
not
adequately
account
for
variability,
i.
e.,
sample
sizes
were
insufficient
to
detect
treatment
effects
due
to
large
measured
variances.
Quality
control
assurance
data
indicated
that
many
of
the
studies
were
subject
to
poor
water
quality.
Low
dissolved
oxygen
levels
and
elevated
ammonia
nitrogen
and
nitrite
concentrations
indicate
that
these
factors
alone
could
have
negatively
impacted
growth
and
development
of
the
test
organisms.

In
the
reviewed
studies,
loading
rates
ranged
as
high
as
30
to
65
tadpoles
per
liter
with
incomplete
and
infrequent
exposure
solution
changes.
These
conditions
could
create
unfavorable
environmental
conditions
for
optimum
growth
and
development,
and
in
fact,
reported
growth
and
development
of
unexposed
organisms
were
inconsistent
with
patterns
cited
in
the
open
literature
(
Nieuwkoop
and
Faber
1994).
Laboratory
manuals
on
the
use
of
X.
laevis
recommend
that
stocking
rates
not
exceed
1
tadpole/
L
and
that
complete
water
changes
be
made
at
regular
intervals
(
Sive
et
al.
1998).
Many
of
the
laboratory
studies
reviewed
exchanged
only
50%
of
the
Page
­
49­
exposure
solution
every
72
hrs.
Additionally,
although
laboratory
studies
offer
the
best
opportunity
to
control
exposure
conditions,
in
several
of
the
laboratory
studies,
either
the
exposure
conditions
were
poorly
characterized
or
atrazine
was
detected
in
control
tanks.
In
some
cases,
the
level
of
atrazine
detected
was
comparable
to
those
in
atrazine­
treated
groups.

Feeding
rates
and
the
quality
of
food
used
were
not
described
for
many
of
the
studies.
In
several
studies,
the
authors
indicated
that
feed
analysis
was
"
inconclusive;"
however,
no
data
were
provided
to
independently
interpret
the
analyses.
Because
the
organisms
appeared
to
be
in
relatively
poor
condition
in
many
of
the
laboratory
studies,
it
appears
that
feeding
regimes
were
not
adequately
considered.

Generally,
the
plasma
hormone
concentrations
and
gonad/
brain
aromatase
activity
assays
were
characterized
by
extremely
high
levels
of
variability.
Coefficients
of
variation
ranged
as
high
as
550%
and
indicated
that
the
means
to
detect
a
treatment
effect
was
problematic.
In
some
cases,
hormone
assays
could
not
distinguish
males
from
females.
The
variability
in
these
measurement
endpoints
may
have
been
related
to
handling
stress
and
protracted
sampling
periods.

In
some
cases
sampling
extended
over
several
hours
to
many
months
and
may
have
resulted
in
a
comparison
of
animals
that
were
at
different
stages
of
their
sexual
maturation
cycles.

Additionally,
X.
laevis
is
an
opportunistic
breeder,
and
it
is
unclear
what
environmental
factors
initiate
breeding
responses.
Placing
males
and
females
in
close
proximity
to
one
another
during
collection
could
potentially
impact
plasma
sex
hormone
levels.

Although
most
of
the
studies
employed
positive
controls
using
dihydrotestosterone
and
17­
 
estradiol,
typically
there
was
a
lack
of
responsiveness
to
these
hormones,
implying
that
hormone
concentrations
were
not
sufficiently
high,
or
that
the
test
animals
were
not
responsive
to
the
chemicals,
or
that
there
were
technical
problems
with
the
analytical
methods
used
to
measure
sex
steroid.
According
to
the
open
literature,
X.
laevis
treated
with
estradiol
72­
hrs
post­
hatch
through
metamorphosis
should
produce
100%
females
(
Hayes
1998);
however,
in
the
current
battery
of
studies,
estradiol
treatment
only
yielded
a
maximum
of
67%
females
and
in
some
studies
did
not
have
any
effect
on
sex
ratios.
Furthermore,
measured
estradiol
concentrations
in
some
of
the
studies
were
considerably
lower
than
nominal.
While
it
is
possible
that
some
species
were
not
responsive
to
hormone
treatments,
it
is
uncertain
why
these
species
would
be
selected
as
Page
­
50­
appropriate
measures
of
estrogenic/
androgenic
effects.
The
current
studies
suggested
that
poor
water
quality,
overcrowding,
and
insufficient
hormone
concentrations
may
have
contributed
to
the
lack
of
responsiveness
for
many
of
the
animals
to
respond
to
estradiol
and
dihydrotestosterone.

The
poor
condition
of
the
study
animals
was
a
recurrent
theme
in
the
majority
of
the
laboratory
study
reviews.
With
mortality
rates
as
high
as
75%
across
all
treatments
in
one
study
and
disease
in
another,
the
viability
of
the
test
organisms
was
clearly
in
question.
The
protracted
length
of
time
required
for
the
frogs
to
undergo
metamorphosis
(
80+
days
as
opposed
to
the
more
typical
58­
day
period
for
X.
laevis)
and
the
tendency
for
animals
to
lose
weight
with
increasing
length
of
time
to
metamorphosis
underscored
EPA's
concerns
that
study
conditions
were
not
conducive
to
optimal
growth.
The
decreased
condition
and
growth
could
have
impacted
the
developmental
status
of
the
animals
as
evidenced
by
the
number
of
frogs
that
did
not
complete
metamorphosis.

Field
Studies
As
mentioned
previously,
field
studies
can
help
evaluate
the
relevancy
and
significance
of
toxicological
effects
observed
in
laboratory­
based
investigations.
Unlike
laboratory
studies,

however,
the
more
natural
conditions
of
the
field
can
add
variability
to
the
data.
Consequently,

considerable
effort
is
required
to
assure
that
natural
variability
across
field
study
sites
is
accounted
for
and
that
sites
do
not
have
characteristics
capable
of
interfering
with
the
interpretation
of
the
data.
Most
of
the
studies
reviewed,
however,
did
not
provide
sufficient
information
to
characterize
study
sites
or
provide
sufficient
rationale
for
site
selection
and
variability.
Atrazine
exposure
in
reference/
control
sites,
with
likely
exposure
to
a
much
wider
array
of
chemicals,

reduced
if
not
eliminated
the
option
of
conducting
hypothesis
testing
and
led
some
investigators
to
consider
a
regression­
based
approach,
even
though
the
study
design
was
not
originally
intended
to
support
such
an
analysis.
In
addition,
the
evaluation
of
potential
confounding
effects
of
nonchemical
stressors,
e.
g.,
habitat
condition,
prey
availability,
nutrient
loading,
were
not
described
or
evaluated.
Page
­
51­
While
the
difficulty
in
selecting
field
sites
that
have
similar
morphological
characteristics
is
well
recognized
and
appreciated,
many
of
the
study
sites
had
widely
divergent
conditions.

Although
atrazine
is
frequently
detected
in
monitoring
studies,
the
potential
for
exposure
of
field
study
reference
sites
was
not
adequately
considered
in
the
majority
of
studies.
Quantifiable
levels
of
atrazine
and/
or
its
degradates
were
detected
in
many
of
the
studies.
Although
some
of
the
study
authors
conducted
pilot
studies
to
establish
atrazine
levels
in
experimental
and
reference
sites,
follow­
up
monitoring
suggests
that
the
initial
surveys
were
misleading
,
i.
e.,
atrazine
concentrations
in
reference
sites
were
considerably
higher
than
initial
sampling
indicated.

Additionally,
other
triazine
herbicides
(
e.
g.,
simazine,
terbuthylazine)
were
detected
in
reference
sites
at
comparable
if
not
higher
levels
than
at
atrazine­
exposed
sites.
Although
other
pesticides
were
admittedly
used
in
the
vicinity
of
many
study
locations,
none
of
the
pesticide
concentrations
was
characterized.

Ideally,
field
studies
should
be
designed
based
on
the
variability
associated
with
the
measurement
endpoints,
and
sample
sizes
should
reflect
the
number
of
test
animals
required
to
identify
a
specified
difference
within
a
given
level
of
certainty.
Potential
sources
of
variability
should
also
be
identified
and
controlled
to
the
extent
possible.
The
current
studies
did
not
appear
to
be
designed
based
on
the
variability
associated
within
the
range
of
measurement
endpoints.
In
some
cases,
animals
were
collected
over
extended
periods
of
time
(
up
to
6
months)
which
could
increase
variability
due
to
differences
in
developmental
stage
and
reproductive
status
of
the
organisms
at
time
of
collection.
Page
­
52­
CHAPTER
4
UNCERTAINTIES
IN
ASSESSING
POTENTIAL
ATRAZINE
EFFECTS
At
face
value,
the
weight­
of­
evidence
might
suggest
that
atrazine
did
not
produce
any
consistent
effects
across
the
range
of
frog
species
tested
and
that
the
demasculinizing
(
decreased
laryngeal
dilator
muscle
area)
effects
previously
reported
by
Hayes
et
al.
(
2002a,
b,
c)
could
not
be
replicated.
While
Tavera­
Mendoza
et
al.
(
2001a,
b)
reported
that
exposure
to
higher
atrazine
concentrations
(
21

g/
L)
significantly
increased
the
incidence
of
gonadal
abnormalities
in
X.

laevis,
others
reported
that
exposure
to
similar
concentrations
had
no
effect
in
X.
laevis.
The
feminizing
effect
(
intersex/
hermaphroditism)
of
atrazine
could
only
be
demonstrated
statistically
in
one
study
(
Goleman
and
Carr
2003),
but
at
exposure
concentrations
of
25

g/
L
as
opposed
to
the
0.1

g/
L
effect
level
reported
by
Hayes
et
al.
(
2002a).
Additionally,
the
studies
did
not
demonstrate
whether
gonadal
and/
or
brain
aromatase
activity
was
induced
by
atrazine
treatment
or
whether
plasma
testosterone/
estrogen
levels
were
affected.
Although
the
Florida
cane
toad
study
(
Goleman
and
Carr
2003)
indicated
both
demasculinizing
effects
(
genetic
males
with
female
coloration)
and
feminizing
effects
(
oogenesis
in
male
Bidder's
organ),
there
were
insufficient
data
to
clearly
link
atrazine
exposure
to
this
phenomena.
However,
as
previously
discussed,
each
of
the
studies
has
deficiencies
and
uncertainties
that
limit
their
usefulness
in
differentiating
treatment
effects.

One
of
the
issues
that
has
been
raised
regarding
the
potential
effects
of
atrazine
on
amphibian
development
has
been
the
enhanced
response
at
lower
doses
(
i.
e.,
in
the
range
of
0.1
to
0.3

g/
L).
According
to
Hayes
et
al.(
2002a),
X.
laevis
and
R.
pipiens
exhibited
hermaphroditism
at
0.1

g/
L
in
both
laboratory
and
field
studies,
and
responses
at
higher
doses
were
either
similar
to
or
less
than
those
observed
at
this
level.
Unfortunately,
most
of
these
studies
only
evaluated
atrazine
effects
at
concentrations
ranging
from
1
to
25

g/
L,
and
background
levels
of
atrazine
were
as
high
as
0.1

g/
L.
As
a
result,
these
studies
could
not
be
used
to
differentiate
a
"

lowdose
effect
since
controls
were
often
at
levels
at
which
effects
have
been
observed
in
other
studies.
Page
­
53­
Although
these
studies
do
not
serve
to
reject
or
confirm
studies
claiming
to
demonstrate
that
atrazine
exposure
results
in
demasculinizing
and/
or
feminizing
effects
in
frogs,
they
were
useful
in
identifying
measurement
endpoint
responses
that
may
represent
potential
hazards
to
amphibians.
They
also
provide
important
information
on
how
to
design
laboratory
and
field
experiments
to
assess
these
potential
effects.
As
previously
discussed,
the
ability
of
the
current
battery
of
tests
to
discriminate
treatment
effects
appears
limited
given
the
variability
surrounding
each
of
the
measurement
endpoints.
When
variability
is
coupled
with
confounding
effects,
such
as
impaired
development,
disease,
and
atrazine
contamination
in
control/
reference
groups,
it
is
difficult
to
draw
any
definitive
conclusions
about
the
test
results.
However,
these
observations
provide
important
insights
in
designing
studies
that
can
minimize
the
introduction
of
variability,

which
is
addressed
in
the
next
section
of
the
paper
(
Chapter
5,
Reducing
Uncertainties).

Ecological
Relevancy
of
Endpoint
In
addition
to
the
above
uncertainties
surrounding
these
amphibians
studies,
the
ecological
relevancy
of
the
measurement
endpoints
used
in
the
studies
is
uncertain
and
needs
to
be
examined
further.
Even
if
atrazine
did
result
in
gonadal
abnormalities,
such
as
ovarian
and
testicular
tissue
in
the
same
individual
or
testicular
oogenesis,
there
are
no
data
currently
available
to
suggest
that
reproduction
is
impaired.
Additionally,
if
laryngeal
muscle
growth
were
impaired
by
chemical
exposure,
are
males
less
able
to
attract
females
by
attenuated
mating
calls
and
is
reproduction
subsequently
impaired?
At
present
there
are
no
data
to
relate
the
above
mentioned
measurement
endpoints
to
more
traditional
chronic
assessment
endpoints
of
survival,
growth
and
reproductive
impairment
that
could
impact
animal
populations.
In
Hayes
et
al.
(
2002c),
there
apparently
was
little
difficulty
in
collecting
specimens,
even
though
92%
of
the
northern
leopard
frogs
sampled
at
one
site
were
hermaphrodites.
It
is
uncertain,
therefore,
whether
gonadal
effects,
such
as
hermaphroditism,
significantly
impaired
population
levels.
According
to
the
literature
(
Gray
et
al.

1996;
DePrado
et
al.
2000),
resistance
to
atrazine
has
developed
in
many
species
of
weeds
and
nontarget
plants.
Hayes
et
al.
(
2002a;
c)
has
suggested
that
chemical
exposure
may
result
in
Page
­
54­
resistance
to
the
proposed
feminizing
effects
of
atrazine
by
delayed
metamorphosis.
However,

there
are
no
data
currently
available
to
assess
a
possible
compensatory
mechanism.

Dose­
Response
Relationships
Hayes
et
al.
(
2002a,
c)
has
suggested
that
developmental
effects
resulting
from
atrazine
exposure
are
nonmonotonic
and
do
not
follow
typical
dose­
response
curves
where
higher
concentrations
elicit
greater
effects.
Rather,
the
response
of
both
clawed
frogs
and
leopard
frogs
has
been
either
a
trigger
followed
by
a
consistently
similar
response
or
an
enhanced
response
at
low
concentrations
followed
by
diminished,
although
significant,
effects
at
higher
concentrations
(
low­
dose
effect).
Although
accentuated
responses
at
lower
doses
have
been
associated
with
herbicides
in
an
effect
termed
hormesis
(
Calabrese
and
Baldwin
2001),
the
relevancy
of
this
effect
to
risk
assessment
has
not
been
established.
Because
most
of
the
current
studies
did
not
test
below
1

g/
L
and
many
of
the
reference
sites
were
contaminated
with
atrazine
at
levels
where
other
studies
have
claimed
to
demonstrate
accentuated
responses,
there
is
not
sufficient
information
available
to
ascertain
the
reproducibility
of
a
potential
low­
dose
effect.

Mechanistic
Plausibility
of
Atrazine
Effects
Successful
extrapolation
of
the
effects
of
a
chemical
among
different
species
is
largely
dependent
on
the
degree
to
which
biological
processes
are
conserved.
Conservation
in
biological
processes
are
typically
highest
in
basic
cellular,
early
ontogenetic,
and
endocrine
mechanisms.
For
example,
the
mechanisms
associated
with
the
synthesis,
metabolism,
and
receptor
activity
among
vertebrates
for
estrogen,
androgen,
and
thyroxine
are
considered
to
be
generally
highly
conserved,

albeit
some
differences
exist.
The
"
downstream"
effects
of
these
hormones
which
occur
subsequent
to
receptor
binding,
on
the
other
hand,
tend
to
be
divergent.
Therefore,
the
probability
of
successful
inter­
species
extrapolation
is
highest
when
analyzing
the
"
upstream"
Page
­
55­
processes,
that
is,
those
that
are
typically
associated
with
homeostatic
mechanisms
and
receptor
binding.

In
the
case
of
atrazine,
the
premise
has
been
that
aromatase
induction
results
in
elevated
estrogen
levels
which
leads
to
the
apical
effect,
i.
e.,
ovotestes
and
reduced
secondary
sex
characteristics
in
males
(
Hayes
et
al.
2002a,
c).
Aromatase
induction
by
atrazine,
though,
has
not
been
demonstrated
in
any
anuran
in
the
laboratory,
and
attempts
to
replicate
hermaphroditism
in
other
studies
have
not
been
successful.
While
Hayes
et
al.
(
2002a)
reports
that
adult
males
treated
with
25

g/
L
atrazine
had
significantly
lower
(
p<
0.05)
testosterone
levels
compared
to
control
males,
aromatase
activity
was
not
measured
in
these
animals.
None
of
the
studies
reviewed
demonstrated
a
statistically
significant
increase
in
aromatase
activity
following
atrazine
exposure.
Based
on
the
Agency's
understanding
of
the
variability
associated
with
the
aromatase
assay
used
in
these
analyses,
the
study
designs
were
not
sufficient
to
detect
treatment
effects.

Consequently,
at
this
time
there
do
not
appear
to
be
data
available
that
would
directly
support
a
hypothesis
that
aromatase
induction
in
amphibians
occurs
as
a
result
of
atrazine
exposure.
If
the
inability
to
demonstrate
aromatase
induction
is
related
to
problems
in
replicating
study­
specific
conditions,
then
it
is
important
to
consider
whether
those
conditions
are
environmentally
relevant
and
whether
this
introduction
of
variability
represents
an
essential
design
component
for
future
studies
to
more
fully
evaluate
the
potential
role
of
aromatase
induction.
If
an
endpoint
were
to
be
identified,
then
the
normal
pattern
would
need
to
be
described
in
both
field
and
laboratory
settings
to
discriminate
actual
deviations.
Page
­
56­
Laboratory
to
Field
Extrapolation
A
major
challenge
to
interpreting
the
ecological
significance
and
relevancy
of
effects
observed
in
the
laboratory
is,
in
part,
a
function
of
the
difficulty
in
extrapolating
responses
observed
under
relatively
greater
controlled
and
monitored
treatment
exposures
and
environmental
conditions
to
the
more
variable
and
uncontrolled
nature
of
exposures
and
conditions
in
the
field.
For
example,
typical
laboratory
conditions
for
X.
laevis
utilize
constant
water
temperatures
(
20­
25oC)
and
constant
photoperiod
(
12
hrs
light:
12
hrs
dark).
These
conditions
maintain
X.
laevis
adults
in
a
continual
reproductive
status
in
which
the
animals
can
be
induced
to
spawn.
However,
in
the
field,
organisms
are
subjected
to
fluctuating
(
diurnal
and
seasonal)
temperatures
and
to
continuing
changes
in
photoperiod,
according
to
the
season.
These
two
environmental
factors,
temperature
and
photoperiod,
are
known
to
be
key
determinants
of
reproductive
status
in
most
aquatic
organisms.
As
a
result
of
these
differences
in
environmental
factors,
the
physiological
status
of
organisms
in
the
field
may
differ
significantly
from
those
in
the
laboratory.
Finally,
the
studies
conducted
so
far
have
not
taken
into
account
the
developmental
biology
of
X.
laevis,
nor
have
they
considered
fluctuations
in
atrazine
exposure.
For
example,
the
extended
sampling
periods
could
result
in
comparing
organisms
at
different
stages
of
their
annual
reproductive
cycles
and
months
after
atrazine
application.

Application
of
Available
Studies
to
Assess
Potential
Atrazine
Effects
Based
on
the
currently
available
data,
the
Agency
has
determined
that
there
is
sufficient
certainty
and
consistency
across
several
studies
(
Tavera­
Mendoza
et
al.
2001a,
b;
Hayes
et
al.

2002a,
b,
c;
Goleman
and
Carr
2003;
and
Sepulveda
and
Gross
2003)
to
establish
the
plausibility
of
a
hypothesis
that
atrazine
could
affect
amphibian
development.
However,
the
uncertainties
Page
­
57­
described
previously
preclude
establishing
a
definitive
characterization
of
atrazine's
effects
on
amphibian
development.
Further,
based
on
the
available
information,
conclusions
regarding
the
magnitude
and
likelihood
of
any
potential
adverse
developmental
effects
span
an
approximately
250­
fold
range
of
atrazine
exposure,
due
to
the
existing
uncertainties.
If
the
Agency's
risk
management
decision
requires
a
greater
degree
of
certainty
in
the
ecological
risk
assessment
for
atrazine
than
possible
from
currently
available
data,
then
additional
data
would
be
necessary
to
evaluate
the
potential
causal
relationships
between
atrazine
exposure
and
gonadal
development
in
amphibians
and
if
so,
the
nature
of
the
dose­
response
relationship.
To
the
extent
that
any
mode
of
action
for
any
observed
effect(
s)
can
be
proposed
and
supported,
such
information
would
strengthen
the
plausibility
of
the
relationship.
Although
field
studies
provide
useful
insights
on
"
real­
world"
responses
and
reflect
what
may
actually
occur
in
a
natural
setting,
they
may
be
subject
to
a
wide
range
of
environmental
effects
that
can
influence
a
study's
measurement
endpoints
beyond
treatment
effects.
The
current
studies
show
the
extent
to
which
field
experiments
can
introduce
variability
and
make
it
difficult
to
identify
atrazine­
specific
effects.
As
described
in
Chapter
5
(
Reducing
the
Uncertainties),
EPA
recommends
that
reliable
laboratory
experiments
be
conducted
before
field
studies
are
undertaken.
Additionally,
EPA
welcomes
discussions
with
the
SAP
on
potential
future
study
protocols
to
help
establish
a
perspective
for
both
the
Agency
and
the
scientific
community
to
consider
in
reviewing
experimental
design
options
before
initiating
any
experiments
intended
to
address
the
uncertainties
described
in
the
current
evaluation.
Page
­
58­
CHAPTER
5
REDUCING
THE
UNCERTAINTIES
As
discussed
in
Chapter
4
,
there
are
numerous
uncertainties
associated
with
the
studies
reviewed
for
this
paper.
While
there
is
sufficient
information
to
establish
a
plausible
hypothesis
concerning
potential
atrazine
effects,
the
uncertainties
associated
with
these
studies
are
significant,
making
it
difficult
to
address
the
core
risk
question
as
to
whether
or
not
atrazine
can
adversely
impact
anuran
development
and
reproduction.
Current
studies
show
that
field
experiments
may
introduce
considerable
variability
and
make
it
difficult
to
identify
atrazinespecific
effects.
Therefore,
before
additional
resources
are
devoted
to
field
studies,
laboratory
studies
should
be
designed
and
conducted
regarding
the
working
hypothesis
that
atrazine
exposure
causes
gonadal
developmental
effects
in
amphibians.
Once
this
effect
can
be
established,

with
a
corresponding
dose­
response
relationship,
resources
should
be
devoted
to
examining
induction
in
aromatase
activity,
which
elevates
circulating
estrogen
concentrations
and
in
turn
results
in
the
feminization
of
male
anurans.
EPA
recommends
that
study
protocol
characteristics
of
higher
tier
studies
be
reviewed
in
some
detail
by
the
SAP
in
a
future
consultation,
assuming
a
review
of
the
current
Agency
analysis
is
consistent
with
the
Panel's
evaluation.

The
tiered
approach
for
conducting
these
studies,
which
is
partitioned
into
five
distinct
phases,
follows
a
typical
reductionist
approach
(
Figure
1).
Briefly,
the
objective
of
the
first
phase
is
to
conduct
studies
that
determine
whether
or
not
atrazine
exposure
causes
changes
in
apical
effects
(
e.
g.,
hermaphroditism)
related
to
gonadal
development
and
reproduction.
A
secondary
objective
of
this
phase
is
to
provide
information
on
the
repeatability
of
previous
observations,

develop
a
sound
dose­
response
relationship,
and
determine
the
developmental
sensitivity
of
the
Page
­
59­
test
species,
X.
laevis.
If
apical
effects
are
observed
in
response
to
atrazine
exposure,
then
three
additional
phases
are
proposed
that
address
the
mode
and
mechanism
of
atrazine
action
along
the
lines
of
the
working
hypothesis.
Before
proceeding
through
the
proposed
study
phases,
a
risk
management
decision
would
be
required
to
determine
if
further
reduction
in
uncertainty
is
needed.

In
the
second
phase,
sex
steroid
(
i.
e.,
testosterone
and
estrodiol)
measurements
would
be
conducted,
and
in
the
third
and
fourth
phases
aromatase
activity
would
be
measured.
These
mechanistic
studies
would
be
conducted
for
three
reasons.
First,
understanding
the
mode
of
action
is
essential
for
extrapolating
between
species.
Because
it
is
impossible
to
test
all
species
of
concern
for
apical
effects,
this
overall
approach
uses
X.
laevis
as
a
surrogate
species
where
doseresponse
developmental
sensitivity,
and
mechanism
of
action
data
are
used
as
the
basis
to
investigate
the
effects
of
atrazine
on
native
species.
This
approach
also
allows
for
studies
with
native
anurans
to
be
focused
and
efficient
and
permits
one
to
test
the
hypothesis
that
native
species
respond
similarly
to
atrazine
as
X.
laevis.

Second,
mechanistic
information
is
important
in
that
it
reduces
the
uncertainties
associated
with
developing
a
causal
relationship
between
apical
effects
and
atrazine
exposure
by
demonstrating
a
plausible,
mechanistic
basis
for
any
such
effects.
In
this
context,
such
data
also
assist
with
the
interpretation
of
dose­
response
relationships
and
appropriate
dose
metrics.
Finally,

mechanistic
studies
set
the
stage
for
the
development
of
potential
bioindicators
of
response
that
can
be
applied
if
future
field
work
is
needed.
The
lack
of
such
information
is
a
critical
factor
that
limits
the
utility
of
the
field
studies
conducted
thus
far.
Page
­
60­
Phase
5
of
this
proposal
attempts
to
develop
data
to
establish
whether
or
not
the
effects
of
atrazine
on
gonadal
development,
if
established,
are
ecologically
relevant.
This
phase
would
require
some
novel
approaches
for
evaluating
fertility
of
males
after
atrazine
exposure
in
order
to
develop
data
suitable
to
an
analysis
with
population
models.

There
are
no
universally
accepted
standardized
methods
for
the
studies
outlined
above.

Therefore,
the
methods
proposed
in
the
following
sections
represent
an
ad
hoc
approach,
using
information
and
experience
of
several
different
laboratories.
Although
this
proposal
defines
a
series
of
studies
to
develop
a
comprehensive
understanding
of
atrazine's
potential
effects
at
the
organismal
and
sub­
organismal
levels,
this
approach
does
not
need
to
be
fully
executed
as
currently
outlined.
If
the
studies
proceed
from
the
initial
phase
to
other
phases,
then
there
is
an
opportunity
to
review
the
outcomes
and
revise
future
study
objectives
accordingly
and
to
evaluate
the
utility
of
conducting
the
next
phase
relative
to
the
potential
improvement
of
an
eventual
risk
assessment.
The
advantage
to
this
approach
is
that
it
minimizes
the
generation
of
unnecessary
data,
thus
minimizing
costs
and
delays
in
completing
the
risk
assessment.
This
flexibility,

however,
requires
iterative
review
and
communication
between
the
risk
assessors
and
risk
managers
to
ensure
that
the
Agency's
data
needs
are
met.

Phase
1.
Test
for
Apical
Gonadal
Effects
The
ability
of
only
one
laboratory
to
produce
ovotestes
at
low
atrazine
concentrations
is
problematic
because
this
is
the
endpoint
on
which
much
of
the
concern
hinges.
Therefore,

additional
studies
on
the
effects
of
atrazine
on
gonadal
differentiation
should
be
conducted.

Repeatability
is
the
hallmark
of
sound
science
and
must
be
considered
as
an
important
step
in
determining
the
relevancy
of
previously
conducted
studies.
Positive
results,
besides
supporting
Page
­
61­
the
conclusions
of
previous
studies,
would
provide
the
rationale
for
conducting
further
studies
which
attempt
to
elucidate
the
pathway
of
potential
atrazine
effects.
Negative
results,
on
the
other
hand,
that
do
not
support
the
previous
studies
would
eliminate
the
need
to
proceed
to
additional
phases
of
investigation.

Species
The
species
used
for
this
phase
should
be
both
amenable
to
laboratory
experimentation
and
sensitive
to
the
effects
in
question.
Xenopus
laevis
and
Rana
pipiens
meet
both
of
these
requirements.
X.
laevis,
however,
should
be
used
as
the
primary
species
because
of
its
ease
of
culture,
shorter
developmental
time,
and
the
ability
to
conduct
studies
throughout
the
year.
Rana
pipiens,
while
perhaps
more
environmentally
relevant
to
North
America,
should
only
be
used
in
focused
studies
for
additional
confirmation
of
results
obtained
using
X.
laevis.

Stage
Sensitivity
These
studies
should
first
be
initiated
with
free
swimming
larvae
(
96­
hr
post­
fertilization;

NF
Stage
46)
to
make
a
direct
comparison
to
the
previously
published
studies
and
to
include
the
developmental
stages
known
to
be
sensitive
to
feminization
by
exogenous
estrogens
(
Villalpando
and
Merchant­
Larios
1990;
Figure
2).
Exposures
should
continue
through
metamorphosis
(
NF
Stage
66),
at
which
time
the
study
can
be
terminated.
If
atrazine
is
shown
to
induce
gonadal
effects
in
larvae,
then
additional
studies
should
be
conducted
at
the
effective
atrazine
concentrations.
These
studies
should
initiate
exposure
with
different
developmental
stages
(
e.
g.

NF
stages
52
and
58)
to
determine
if
there
are
stage
sensitivity
differences
and
whether
those
differences
are
consistent
with
previous
studies
(
Villalpando
and
Merchant­
Larios
1990;
Figure
Page
­
62­
2).
By
establishing
stage
sensitivity,
further
experimentation
in
subsequent
phases
could
be
more
focused.

Test
Conditions
The
tests
should
be
conducted
using
flow­
through
conditions
which
favor
growth,

survival,
and
development.
Static
or
static
renewal
methods
should
not
be
used.
The
type
of
water
is
not
specified
in
this
method
because
there
is
no
known
optimal
water
type.
Instead,
the
type
of
water
used
must
be
established
to
promote
normal
survival,
growth,
and
development
under
the
conditions
of
the
proposed
studies.
Using
a
flow­
through
system,
there
is
no
need
to
aerate
the
exposure
water.
Feeding
should
be
done
on
a
daily
basis
using
a
food
type
and
quantity
that
also
has
been
demonstrated
to
promote
normal
survival,
growth,
and
development.

Biological
loading
rates
are
a
major
concern
of
the
previously
conducted
studies.
Loading
rate
is
known
to
be
an
important
factor
in
maintaining
adequate
water
quality
which
is
essential
for
valid
aquatic
toxicity
testing.
Excessive
loading
diminishes
water
quality
through
metabolic
activity
and
respiration
and
can
result
in
inefficacious
exposures.
Standards
have
been
established
by
the
American
Society
for
Testing
and
Materials
(
ASTM
1998)
for
acceptable
loading
rates
of
fishes,
amphibians,
and
invertebrates
in
both
static
(
or
static
renewal)
and
flow­
through
conditions.
ASTM
(
1998)
standards
for
loading
in
tests
conducted
using
static
renewal
conditions
at
22o
C
are
0.5
g/
L
of
test
solution.
All
of
the
studies
reviewed
in
this
report
exceed
this
recommendation
and
suggest
that
the
studies
failed
to
maintain
adequate
conditions
for
a
valid
bioassay
(
loading
rates
were
from
1.1
to
24.4
g/
L;
see
Table
2).
Page
­
63­
Table
2.
Summary
of
loading
rates
used
in
the
various
laboratory
studies
on
the
effects
of
atrazine
on
amphibian
gonadal
development.

Exposure
Type
Species
Number
of
Tadpoles
per
Replicate
Maximum
g/
organism
Total
g
Liters
of
solution
g/
L
Hayes
et
al.
2002a
static
renewal
X.
laevis
30
1.51
45
4
11.3
Goleman
and
Carr
2003
static
renewal
X.
laevis
60
1.51
90
4
22.5
Hayes
et
al
2002c
static
renewal
R.
pipiens
30
2.52
75
4
18.8
Hecker
et
al.
2003.
static
renewal
R.
clamitans
303
205
24
24
60
40
4
16
15.0
2.5
Du
Preez
et
al.
2003.
static
X.
laevis
800
1.51
1,200
1,100
1.1
Goleman
and
Carr
2003.
static
renewal
X.
laevis
65
1.51
97.5
4
24.4
Tavera­
Mendoza
et
al.

2001a;
2001b.
static
X.
laevis
16
1.51
24
15
1.6
Hecker
et
al.
2003
static
renewal
X.
laevis
30
1.51
45
4
11.3
1
not
reported;
estimated
based
on
data
provided
in
Figure
3.

2
not
reported;
estimated
based
on
Ankley
et
al
(
1998).

3
initial
number,
but
high
mortality
confounds
analysis
4
not
reported;
estimated
based
on
metamorph
weights
5
not
reported;
estimated
based
on
approx.
30%
mortality
in
controls
at
transfer
to
larger
tanks
on
day
67
The
ASTM­
recommended
loading
rates
for
studies
conducted
using
flow­
through
conditions
is
1
g/
L
for
every
liter
dispensed
in
24
hrs.
Current
studies
conducted
with
X.
laevis
at
the
EPA,
Office
of
Research
and
Development's
Mid­
Continent
Ecology
Division
(
MED)
employ
flow­
through
conditions
that
were
designed
to
meet
the
ASTM
standard.
Typically,
the
flow
rates
in
each
exposure
tank
are
0.025
L/
min,
which
is
equivalent
to
36
L/
day.
Biological
loading
Page
­
64­
is
usually
20
organisms,
which
reach
about
1.8
g
maximum
body
weight
(
Figure
3),
resulting
in
an
approximate
loading
rate
of
1
g/
L
for
each
liter
dispensed
in
24
h.
Under
these
conditions,

minimal
mortality
has
been
observed,
and
metamorphosis
is
typically
completed
in
about
55
days
(
personal
communication:
J.
Tietge,
U.
S.
EPA,
Mid­
Continent
Ecology
Division,
2003).

Water
concentrations
of
atrazine
and
its
major
degradates,
i.
e.,
diaminochlorotriazine
(
DACT),
desethylated
atrazine
(
DEA),
and
desisopropyl
atrazine
(
DIA),
should
be
measured
in
duplicate
at
specified
frequencies
using
an
appropriate
analytical
method
to
cover
the
concentration
range
of
the
parent
compound
used
in
the
test.
Water
chemistries
(
i.
e.,
dissolved
oxygen,
pH,
ammonia)
should
be
measured
periodically
throughout
the
study
and
reported.

Dose­
Response
The
studies
to
date
provide
inconsistent
information
as
to
the
nature
of
a
dose­
response
relationship
for
potential
atrazine
effects.
Further
studies
suggest
that
lower
concentrations
of
atrazine
are
more
effective
than
higher
concentrations
in
eliciting
gonadal
effects
although
no
biological
basis
has
been
demonstrated
for
this
response.
However,
without
knowing
the
biological
basis,
the
shape
of
a
dose­
response
curve
can
only
be
empirically
established
by
testing
multiple
concentrations
in
the
range
of
interest.
The
highest
test
concentration
should
be
at
or
above
25

g/
L,
while
the
lowest
should
be
below
0.1

g/
L.
If
the
results
of
the
initial
study
using
this
concentration
range
demonstrate
non­
monotonic
dose­
response
patterns,
then
subsequent
tests
that
focus
on
a
specific
concentration
range
may
be
indicated.
If
atypical
dose­
response
patterns
are
observed,
i.
e.,
inverted
U
shape
dose­
response,
then
additional
work
may
be
necessary
to
develop
a
biological
basis
for
such
a
response.
This
may
require
dosimetry
measurements
to
ensure
that
the
organismal
dose,
i.
e.,
whole
organism
residue,
is
correlated
and
Page
­
65­
dependent
upon
the
applied
dose,
i.
e.,
exposure
water
concentration.
Deviations
from
such
a
relationship
could
indicate
that
other
mechanisms,
e.
g.,
metabolism,
active
uptake,
active
elimination,
are
at
work
to
alter
the
expected
steady­
state
relationship,
thereby
potentially
altering
the
shape
of
the
dose­
response
curve.

Positive
Controls
Many
of
the
atrazine
studies
have
used
"
positive"
controls.
Typically
they
have
used
both
an
estrogen
and
an
androgen.
The
use
of
an
estrogen
is
desirable
because
it
demonstrates
the
sensitivity
of
the
organisms
to
a
potential
estrogen
effect
that
is
hypothesized
to
be
in
the
pathway
impacted
by
atrazine.
Estrogen
applied
concentrations
should
be
measured
at
specified
intervals
to
assure
adequate
delivery
to
organisms.
The
use
of
an
androgen
as
a
"
positive"
control
is
not
necessary
because
there
is
no
plausible
relationship,
based
on
the
working
hypothesis,
between
androgen
responsiveness
and
the
effects
of
concern.

Sampling
Sampling
of
all
organisms
under
study
is
essential
to
avoid
potential
bias
related
to
differences
in
developmental
rate.
Sampling
should
occur
at
the
same
developmental
stage
to
make
sure
that
the
inter­
organism
comparisons
are
valid.
For
X.
laevis,
sampling
when
each
organism
reaches
NF
66
is
appropriate.
Replication,
loading,
and
final
sampling
should
be
determined
using
statistical
procedures
to
ensure
that
the
hypothesis
of
the
study
can
be
tested
appropriately.

Endpoints
Page
­
66­
The
primary
endpoints
of
interest
are
those
indicative
of
abnormal
sexual
differentiation.

These
include:
gross
analysis
of
chemically
prepared
(
fixed)
gonads
to
determine
the
morphological
condition
of
the
gonads,
histological
analysis
of
gonads
from
each
organism
to
determine
whether
effects
such
as
testicular
oocytes
and
reduced
spermatogonial
cell
nests
occur
in
males,
and
alterations
in
the
numbers
of
primary
and
secondary
oogonia
occur
in
females.

Additionally,
determination
of
male:
female
sex
ratios
and
other
observations
on
secondary
sexual
characteristics
should
be
made.
Evaluation
of
laryngeal
musculature
is
not
necessary,
as
the
gonads
appear
to
be
more
sensitive
to
the
effects
of
atrazine
exposure.
In
addition,
the
typical
endpoints
of
daily
survival,
growth
(
as
determined
by
wet
weight
at
the
termination
of
the
study),

and
development
(
as
indicated
by
time
to
complete
metamorphosis)
should
be
evaluated.

Quality
Indicators
Because
there
are
currently
no
standardized
methods
available
for
reproductive
and
developmental
assays
with
anurans,
in
general,
nor
with
X.
laevis
specifically,
the
following
guidelines
are
proposed
as
quality
indicators
of
a
valid
study.
Of
utmost
importance,
the
study
should
attempt
to
conform
to
the
ASTM
recommended
standards
for
biological
loading
rates
of
1.0
g/
L
for
flow­
through
studies.
These
loading
rates
should
be
determined
using
the
maximal
mass
of
the
organisms,
not
the
initial
or
final
masses,
as
considerable
weight
loss
occurs
(
up
to
50%)
during
metamorphic
climax.
Minor
deviations
can
be
tolerated,
but
the
ASTM
guidance
should
be
strongly
considered
in
test
design
and
conduct.

Survival
is
obviously
an
important
indicator
of
study
quality.
And,
while
there
is
no
bright
line
between
acceptable
and
unacceptable
survival
rates,
survival
of
90%
or
more
of
the
organisms
indicates
a
quality
study,
particularly
in
the
controls.
Although
the
level
of
Page
­
67­
unacceptable
mortality
is
not
defined,
excessive
mortality
serves
to
bias
the
results
of
the
study
by
limiting
analysis
to
the
survivors.

Like
survival,
there
are
no
set
criteria
for
acceptable
growth
and
development.
However,

taken
in
aggregate,
growth
and
development
should
be
as
optimized
as
practical.
Maximal
larval
weights
of
X.
laevis
should
be
about
1.5
to
1.8
g,
with
weights
of
NF
66
organisms
approximately
half
of
maximal
weights
(
Figure
3).
Development
in
controls
should
proceed
without
great
deviations
from
those
determined
by
Nieuwkoop
and
Faber
(
1994),
with
metamorphic
completion
in
approximately
7
to
9
weeks.
Conditions
that
delay
development
of
unexposed
organisms
beyond
10
weeks
suggest
problems
with
food
or
water
quality
and
introduce
uncertainties
regarding
the
effects
of
overall
delayed
development
on
system
specific
development.
Although
there
are
little
or
no
known
data
regarding
delayed
metamorphosis
on
organ­
specific
development,
there
is
a
large
body
of
literature
on
the
effects
of
accelerated
development
demonstrating
that
accelerated
metamorphosis,
through
exposure
to
thyroid
hormone,
results
in
abnormal
system
development
(
Huang
et
al.
2001).
Clearly,
development
is
a
coordinated
event
with
temporal
constraints.

Besides
the
biological
indicators,
there
are
several
standard
exposure­
related
measurements
that
are
indicators
of
a
quality
study.
The
most
common
are
dissolved
oxygen,

ammonia,
and
pH.
ASTM
guidelines
suggest
that
dissolved
oxygen
should
be
maintained
between
60
and
100%
of
saturation
(
ASTM
1998).
These
guidelines
recommend
that
pH
be
within
an
acceptable
range
for
the
test
organism,
which
in
the
case
of
X.
laevis
is
suggested
to
be
within
6.5
and
9.0
for
earlier
developmental
stages
used
in
the
FETAX
assay
(
ASTM
1998).

While
no
specific
guidance
is
provided
for
ammonia
concentrations,
two
studies
have
evaluated
Page
­
68­
the
toxicity
of
ammonia
to
X.
laevis
larvae
and
found
that
concentrations
of
25­
50
ppm
are
lethal
and
that
concentrations
of
about
15
ppm
significantly
reduced
growth
(
Tietge
et
al.
2000;

Schuytema
and
Nebeker
1999)
in
short­
term
studies.
These
studies
suggest
that
total
ammonia
concentrations
should
be
well
below
10
ppm
because
total
ammonia
(
recognizing
that
the
toxic
form,
unionized
ammonia)
is
dependent
on
the
pH
of
the
test
media.

Analysis,
Interpretation,
and
Iteration
If
the
study
or
studies
indicate
that
atrazine
affects
gonadal
differentiation,
then
the
experimentation
could
move
to
the
next
phase
or
phases
which
deal
with
questions
regarding
the
mechanism
of
action
and
reproductive
fitness.
Although
these
mechanistic
and
higher
order
questions
are
presented
sequentially,
they
could
be
evaluated
simultaneously.

Negative
results,
however,
present
a
much
more
difficult
problem.
At
issue
is
whether
or
not
a
study
design
is,
indeed,
a
valid
attempt
to
repeat
previous
studies
using
the
same
conditions.

Strictly
speaking,
the
conditions
of
any
of
the
atrazine
studies
cannot
be
repeated
because
it
would
be
highly
improbable
to
have
organisms
of
the
same
genetic
makeup,
which
could
be
a
contributing
factor
to
the
outcome.
The
conditions
recommended
here
do
not
attempt
to
actually
replicate
the
conditions
of
any
of
the
atrazine
studies
conducted
thus
far.
Those
conditions
appear
to
have
been
inadequate
to
promote
normal
survival,
growth,
and
development.
The
underlying
assumption
is
that
a
biological
phenomenon,
if
significant,
is
pervasive
under
a
broad
range
of
conditions
that
are
favorable
to
the
viability
of
the
organism
being
tested.
If
effects
are
only
observed
in
the
laboratory
due
to
specific
test
conditions
that
are
not
likely
to
be
found
in
the
field,
then
those
effects
may
not
be
relevant.
If,
however,
the
studies
are
negative,
after
Page
­
69­
reasonable
attempts
are
made
to
use
well­
established
aquatic
toxicity
methods
in
a
systematic
approach,
then
there
is
no
plausible
rationale
to
proceed
with
the
next
phases
of
the
investigation.

Phase
2:
Sex
Steroid
Measurements
If
positive
effects
are
observed
in
Phase
1,
then
it
is
logical
to
hypothesize
that
estrogen
levels
are
increased
and
result
in
feminization.
At
this
point,
with
the
stage
and
dose
dependency
of
the
effects
documented,
it
should
be
possible
to
construct
a
refined
exposure
protocol
which
maximizes
a
presumptive
increase
in
estrogen
concentrations
in
organisms
at
specific
developmental
stages.
Kloas
(
2002)
has
demonstrated
the
ability
to
measure
estrogen
in
organisms
at
early
developmental
stages,
including
those
that
presumably
would
be
present
in
this
hypothetical
study.

If
estrogen
levels
are
determined
to
be
elevated,
then
it
is
reasonable
to
propose
moving
into
Phase
3,
which
proposes
to
measure
aromatase
activity.
If
estrogen
levels
are
not
elevated,

then
an
alternative
path
should
be
followed,
which
evaluates
other
mechanisms
associated
with
gonadal
differentiation.
Since
the
alternative
path
is
not
part
of
the
working
hypothesis,
no
further
discussion
of
it
will
be
presented
here.

Phase
3:
Aromatase
Activity
Measurements
The
test
for
increased
aromatase
activity
as
a
result
of
atrazine
exposure
at
the
appropriate
developmental
stages
assumes
that
estrogen
levels
have
already
been
elevated
by
atrazine.
The
question
is
whether
or
not
elevated
estrogen
levels
result
from
enhanced
aromatase
activity.

Aromatase,
the
enzyme
responsible
for
conversion
of
testosterone
to
estradiol,
is
expressed
in
several
tissues
of
X.
laevis.
The
developmental
expression
of
this
enzyme
has
been
determined
and
detectable
expression
of
aromatase
mRNA
in
the
gonad
occurs
at
NF
Stage
51.
A
small
Page
­
70­
transient
rise
of
expression
occurs
at
Stage
53,
followed
by
a
transient
depression
at
Stage
54,
and
a
large
rise
of
expression
after
Stage
55
(
Miyashita
et
al.
2000;
Figure
3).

Because
exposure
to
exogenous
estrogen
results
in
feminization
of
the
male
phenotype,
it
is
reasonable
to
hypothesize
that
enhanced
aromatase
activity
could
lead
to
abnormally
elevated
endogenous
estrogens
that
can
also
affect
feminization
of
males.
Under
normal
developmental
circumstances,
the
activity
of
aromatase
is
either
absent
or
very
low
during
the
developmental
stages
that
are
most
sensitive
to
feminization
by
estrogens,
suggesting
that
this
phenomenon
does
not
occur
naturally
(
Figure
3).
In
order
for
aromatase
activity
to
effect
feminization
of
the
testes,

the
activity
of
this
enzyme
must
be
sufficiently
elevated
during
the
developmental
period
sensitive
to
estrogen­
induced
feminization.
This
requires
that
atrazine
cause
aromatase
expression
to
occur
prematurely
and
at
high
enough
levels
to
produce
sufficiently
high
endogenous
estrogen
levels
to
affect
feminization.

If,
on
the
other
hand,
atrazine
does
not
increase
aromatase
activity,
then
consideration
should
be
given
to
studying
its
effect
on
other
factors
that
are
involved
in
the
homeostasis
of
estrogen.
A
likely
research
avenue
would
include
studies
on
hypothalamic
and
pituitary
factors,

which
have
already
been
shown
to
be
affected
by
atrazine
in
rodents
(
Cooper
et
al.
2000).

Phase
4:
Aromatase
Inhibitor
Study
Assuming
that
aromatase
activity
is
elevated
by
atrazine,
a
logical
confirmatory
study
would
be
to
expose
organisms
to
atrazine
in
combination
with
an
aromatase
inhibitor.
This
study
could
be
performed
at
all
of
the
preceding
levels
and
incorporate
apical
endpoints
as
well
as
sex
steroid
and
aromatase
measurements.
Page
­
71­
Phase
5:
Ecological
Relevancy
Study
Quantifying
the
likelihood
of
population­
level
responses
due
to
the
potential
effects
of
atrazine
exposure
to
anurans
is
challenging.
The
hypothesis
is
that
atrazine
exposure
adversely
affects
reproductive
fitness
of
a
specific
anuran
species
to
the
extent
that
population
levels
decline
in
response.
To
test
this
hypothesis,
an
anuran
species
that
permits
the
analysis
of
reproduction
is
required.
Currently,
there
are
no
known
laboratory
methods
to
evaluate
this
endpoint
directly
with
native
species.
And,
even
though
reproduction
in
X.
laevis
appears
to
be
a
useful
surrogate
model
to
identify
the
effects
of
atrazine
on
development
of
the
reproduction
system,
it
probably
cannot
be
used
as
a
surrogate
for
determining
ecological
effects
on
native
anuran
species.

The
essential
elements
of
determining
reproductive
fitness
can
be
readily
identified,
but
collecting
data
to
confidently
assess
them
is
relatively
difficult.
Because
the
primary
effect
of
concern
is
related
to
males,
the
primary
endpoint
of
interest
is
the
ability
of
males
to
attract
mates
and
fertilize
eggs
during
mating
(
amplexus).
Fecundity,
per
se,
as
an
indication
of
female
reproductive
capacity
is
of
little
value,
unless
further
experimental
data
support
the
results
reported
by
Tavera­
Mendoza
et
al.
(
2001b),
which
suggest
that
adverse
gonadal
effects
occur
in
females
as
well.

The
essential
elements
affecting
the
male's
ability
to
effect
egg
fertilization
can
be
broken
down
into
three
main
components:
gamete
function,
gamete
release,
and
reproductive
behavior.

The
first,
gamete
function,
can
be
assessed
using
in
vitro
fertilization
methods.
Briefly,
this
method
requires
that
previously
exposed
males
are
either
maintained
under
temperature
and
lighting
conditions
that
induce
reproductive
readiness
or
are
induced
using
pituitary
extracts
or
gonadotropins.
Testes
would
be
removed
and
macerated
in
order
to
release
the
spermatocytes,
Page
­
72­
which
would
be
applied
to
a
fixed
number
of
previously
collected
oocytes.
The
efficacy
of
fertilization
could
then
be
analyzed
by
determining
the
prevalence
of
developing
embryos.
This
type
of
methodology
is
routinely
used
in
X.
laevis
and
X.
tropicalis
and
could
be
adapted
for
use
in
native
species
that
are
amenable
to
the
procedure.
In
the
case
of
most
native
anurans,
this
would
require
exposing
the
organisms
during
the
period
of
atrazine
sensitivity,
followed
by
1.5
to
2
years
of
holding
to
allow
for
sexual
maturation
to
occur.
Once
again,
although
testing
the
ranid
species
could
eventually
be
necessary,
it
may
be
advisable
to
first
use
X.
tropicalis,
which
reaches
sexual
maturity
in
about
six
months,
to
determine
if
there
is
a
fertility
hazard.
If
a
significant
reduction
in
fertilization
is
observed,
then
histological
analysis
of
male
gonads
of
native
anurans
may
be
an
important
consideration
in
extrapolating
potential
effects
on
spermiogenesis
among
native
and
non­
native
species.

If
in
vitro
fertility
is
unimpaired,
then
it
would
be
advisable
to
determine
the
ability
of
the
males
to
successfully
release
the
gametes.
This
concern
is
based
on
the
fact
that
the
presence
of
ovotestes
alters
the
structure
of
the
gonad,
which
may
impede
gamete
release.
This
type
of
study
would
require
an
in
vivo
approach
using
induced
reproduction.

Because
the
reported
effects
on
the
gonads
result
in
ovarian
tissue
occurring
in
males,
it
is
possible
that
males
with
ovotestes
have
elevated
circulating
estrogen
concentrations.
This
could
impact
the
normal
endocrinological
control
of
reproductive
behavior
and
secondary
sexual
characteristics
related
to
reproduction,
as
suggested
by
the
Florida
cane
toad
study
where
males
expressed
female
coloration
patterns
and
lacked
nuptual
pads
(
Sepulveda
and
Gross
2003).
The
only
viable
way
to
test
for
these
effects
is
to
use
natural
reproductive
behavior
in
an
in
vivo
study.

These
are
the
most
difficult
studies
to
successfully
conduct.
Page
­
73­
Finally,
the
data
collected
from
fertility
studies
may
need
to
eventually
be
translated
into
an
estimate
of
reproductive
output.
The
changes
in
reproductive
output
would
in
turn
need
to
be
expressed
in
probability
terms
associated
with
atrazine
exposure.
Finally,
population
models
would
need
to
be
developed
or
applied
to
assimilate
the
reproductive
output
data.
Page
­
74­
CHAPTER
6
CONCLUSIONS
A
significant
body
of
information
is
available
to
evaluate
the
potential
effects
of
atrazine
on
amphibian
development.
Clearly,
the
body
of
studies
reflect
the
commitment
of
an
international
group
of
research
teams
to
independently
examine
whether
atrazine
exposure
results
in
effects
on
gonadal
development,
and
these
investigations
provide
useful
information
to
help
resolve
this
complex
scientific
and
risk
assessment
challenge.
Any
future
research
will
certainly
benefit
from
and
build
upon
this
existing
foundation.

Based
on
a
review
of
available
studies
in
the
open
literature
along
with
recently
submitted
registrant­
sponsored
studies
examining
the
effects
of
atrazine
on
gonadal
and
laryngeal
development
in
frogs,
the
Agency
has
concluded
that
none
of
the
studies
fully
account
for
environmental
and
animal
husbandry
factors
capable
of
influencing
endpoints
that
the
studies
were
attempting
to
measure.
Therefore,
as
discussed
below,
EPA
finds
the
overall
weight­
of­
evidence
so
uncertain
that
it
does
not
support
any
definitive
conclusions
with
regard
to
whether
or
not
atrazine
exposure
adversely
affects
amphibian
development.

The
current
weight­
of­
evidence
does
not
show
that
atrazine
produces
consistent,

reproducible
effects
across
the
range
of
exposure
concentrations
and
amphibian
species
tested.

Only
one
study
(
Hayes
et
al.
2002a)
has
demonstrated
statistically
significant
reductions
in
laryngeal
muscle
area
(
demasculinization)
in
atrazine­
exposed
males.
Additionally,
while
Hayes
et
al.
(
2002a,
b)
have
reproduced
"
feminizing"
gonadal
developmental
effects
(
hemaphroditism/
ovotestes/
intersex)
in
males
at
atrazine
concentrations
as
low
as
0.1

g/
L,

similar
gonadal
effects
have
only
been
demonstrated
in
one
other
laboratory
(
Goleman
et
al.

2003)
at
25

g/
L
for
X.
laevis
at
roughly
the
same
length
of
exposure
and
stage
of
animal
Page
­
75­
development.
The
only
other
laboratory
studies
demonstrating
gonadal
effects
(
Tavera­
Mendoza
et
al.
2001a,
b)
followed
a
considerably
shorter
(
48­
hr)
exposure
to
a
single
atrazine
concentration
(
21

g/
L)
and
resulted
in
dissimilar
endpoints,
i.
e.,
reduced
testicular
volume
and
numbers
of
spermatogonial
cell
nests
in
males
and
decreased
numbers
of
primary
and
secondary
oogonia
in
females.
While
Hayes
et
al.
(
2002b,
c
)
demonstrated
feminizing
effects
in
male
leopard
frogs
collected
in
the
field,
the
incidence
of
hermaphroditism
varied
widely
relative
to
atrazine
concentration.
In
another
field
study,
Sepulveda
and
Gross
(
2003)
demonstrated
increased
incidence
of
hermaphroditism/
intersex
in
cane
toads
and
southern
toads
collected
in
Florida,
but
the
relationship
to
atrazine
exposure
was
uncertain.

Although
the
weight­
of­
evidence
does
not
show
that
atrazine
produces
a
consistent,

reproducible
effect,
both
laboratory
and
field
studies
provide
evidence
that
atrazine
exposure
may
be
associated
with
effects
on
gonadal
development
and
secondary
sexual
characteristics.
The
Agency
believes
that
the
lack
of
reproducibility
across
studies,
which
may,
in
part,
be
due
to
an
inconsistency
in
the
methods
used
by
the
various
research
teams,
and
the
absence
of
a
doseresponse
at
this
point
do
not
refute
the
hypothesis
that
atrazine
exposure
may
result
in
gonadal
developmental
effects
in
amphibians.
Studies
conducted
by
Hayes
et
al.
(
2002a,
b),
Tavera­

Mendoza
et
al.
(
2001a,
b)
and
Goleman
et
al.
(
2003)
suggest
that
atrazine
exposure
at
various
levels
resulted
in
some
degree
of
gonadal
developmental
effects
and
thus
serve
to
identify
a
potential
hazard
to
X.
laevis.
However,
none
of
these
studies
permit
a
clear
understanding
of
how
potential
gonadal
effects
vary
with
exposure.

While
this
paper
evaluated
the
experimental
designs
employed
in
each
of
the
studies
reviewed
in
some
detail,
it
is
important
to
note
that
none
of
the
open
literature
studies
were
conducted
for
regulatory
purposes
nor
were
specific
protocols
available
to
assist
in
the
design
of
Page
­
76­
registrant­
sponsored
studies.
Notwithstanding
the
limitations
on
the
available
data,
collectively,

the
research
has
provided
valuable
and
useful
insights
into
sources
of
variability
that
can
facilitate
future
study
designs
to
reduce
uncertainties.

In
summary,
while
the
current
research
does
not
support
a
definitive
conclusion
regarding
a
quantitative
dose­
response
relationship
between
atrazine
exposure
and
effects
on
gonadal
development,
it
provides
sufficient
information
to
formulate
a
hypothesis
that
atrazine
exposure
may
affect
gonadal
development.
If
the
Agency's
risk
management
decision
requires
a
greater
degree
of
certainty
in
the
ecological
risk
assessment
for
atrazine
than
possible
from
currently
available
data,
then
additional
information
would
be
necessary
to
evaluate
the
potential
causal
relationships
between
atrazine
exposure
and
gonadal
development
in
amphibians.
Any
future
studies
should
take
into
account
lessons
learned
from
the
currently
available
research.
While
the
Agency
does
not
discount
the
utility
of
field­
based
studies
for
examining
effects
under
"

realworld
conditions,
it
believes
that
at
this
time
laboratory­
based
studies
provide
the
best
opportunity
to
control
environmental
and
animal
husbandry
factors
in
order
to
demonstrate
whether
a
quantitative
relationship
exists
between
atrazine
exposure
and
effects
on
gonadal
development.

In
the
next
chapter
(
Charge
to
the
Panel)
the
paper
solicits
input
from
the
FIFRA
Scientific
Advisory
Panel
on
the
Agency's
interpretation
of
the
existing
data
regarding
the
effects
of
atrazine
on
amphibian
gonadal
development.
Additionally,
the
Agency
is
seeking
input
on
its
proposal
to
further
test
the
hypothesis
that
atrazine
exposure
results
in
gonadal
developmental
effects.
Page
­
77­
CHAPTER
7
CHARGE
TO
THE
PANEL
The
Agency
has
reviewed
relevant
studies
in
the
scientific
literature,
as
well
as
studies
submitted
by
the
registrant,
to
evaluate
the
potential
for
atrazine
to
elicit
developmental
effects
in
amphibians.
The
strengths
and
limitations
of
the
individual
studies
were
assessed,
and
the
extent
of
concordance
for
the
entire
body
of
information
derived
from
these
laboratory
and
field
studies
was
considered
to
assess
the
plausibility
that
atrazine
can
cause
developmental
effects,
and
if
so,

the
nature
and
strength
of
associated
dose­
response
relationships.
These
developmental
effects
included
time
to
metamorphosis,
growth,
gonadal
abnormalities,
sex
ratios,
laryngeal
dilator
muscle
area,
plasma
steroid
concentrations
and
aromatase
activity
of
brain
and
gonadal
tissue.

The
Agency
has
identified
a
number
of
uncertainties
in
the
previous
studies'
attempts
to
demonstrate
whether
atrazine
affects
development
in
amphibians.
To
resolve
these
uncertainties,

the
Agency
has
outlined
a
conceptual
model
for
studying
atrazine
action
on
amphibian
development
and
has
proposed
an
approach
using
focused
empirical
studies
to
test
the
hypotheses
embedded
in
the
conceptual
model.

The
Agency
is
seeking
feedback
from
the
SAP
on
a
number
of
questions
surrounding
the
current
body
of
evidence
regarding
the
potential
effects
of
atrazine
on
development
in
amphibians
and
the
relevancy
of
these
potential
effects
to
ecological
risk
assessment.
Additionally,
the
Agency
is
requesting
that
the
SAP
review
and
comment
on
the
conceptual
model
for
potential
future
studies.
Page
­
78­
Questions
1)
In
reviewing
the
available
laboratory
and
field
studies,
the
Agency
used
a
number
of
criteria
to
evaluate
individual
investigations.
Criteria
such
as
experimental
design,
test
protocols,
and
quality
assurance
information
were
used
to
ascertain
the
reliability
of
the
generated
data
in
terms
of
its
ability
to
adequately
assess
a
hypothesis
that
atrazine
elicits
developmental
effects
in
amphibians,
and
if
so,
the
nature
and
strength
of
associated
doseresponse
relationships.

a)
Does
the
SAP
have
any
comments
and
recommendations
regarding
the
EPA's
approach
and
criteria
used
to
evaluate
the
studies?

b)
Given
the
evaluation
criteria
employed
by
the
Agency,
please
comment
on
EPA's
overall
characterization
of
the
currently
available
studies.

c)
Please
comment
on
the
availability,
as
of
February
28,
2003,
of
additional,
relevant
studies
in
the
open
literature
that
were
not
addressed
in
the
white
paper.

Since
February
28,
2003,
is
the
Panel
aware
of
any
studies
that
would
be
relevant?

2)
In
its
evaluation
of
existing
field
studies,
the
Agency
has
concluded
that
these
investigations
are
of
limited
value.
The
reasons
include:
(
1)
the
high
variability
in
environmental
conditions
and
uncertainties
in
the
pre­
existing
status
and
condition
of
fieldcollected
animals,
(
2)
the
spatial
and
temporal
aspects
of
atrazine
exposure
(
i.
e.,
spatial
and
temporal
variability
over
the
course
of
the
studies
and
the
extent
to
which
such
aspects
of
atrazine
exposure
were
empirically
measured
or
otherwise
accounted
for),
and
(
3)
the
possible
co­
occurrence
of
additional
chemical
and/
or
non­
chemical
stressors.
Page
­
79­
c)
To
the
extent
that
the
field
studies
appear
to
indicate
that
atrazine
may
not
adversely
affect
development,
please
comment
on
EPA's
conclusion
that
the
body
of
data
from
field
studies
does
not
provide
the
means
to
ascertain
whether
the
lack
of
a
relationship
between
atrazine
exposure
and
developmental
effects
is
due
to
the
absence
of
a
causal
relationship
or
limitation
in
study
methodologies.

d)
To
the
extent
that
any
field
studies
appear
to
indicate
that
atrazine
may
adversely
affect
development,
please
comment
on
EPA's
conclusion
that
these
field
studies
do
not
provide
sufficient
information
to
resolve
the
potential
role
of
additional
cooccurring
stressors.

3)
In
an
evaluation
of
the
existing
laboratory­
based
studies,
the
Agency
concluded
that
there
was
sufficient
information
to
establish
a
hypothesis
that
atrazine
could
cause
adverse
gonadal
developmental
effects.
However,
due
to
different
experimental
designs
and
variability
in
the
nature
and
extent
of
experimental
conditions
(
e.
g.,
level
of
excessive
mortality,
delayed
development
in
untreated
organisms,
lack
of
response
to
positive
controls)
it
was
not
possible
to
adequately
assess
the
hypothesis
that
atrazine
causes
developmental
effects.
It
was
further
concluded
that
the
current
body
of
information
did
not
provide
the
means
to
characterize
the
nature
of
any
associated
dose­
response
relationships.

d)
Please
comment
on
EPA's
determination
that
the
laboratory
studies
provide
a
plausible
basis
for
the
means
to
establish
a
hypothesis
concerning
the
potential
for
atrazine
to
cause
developmental
effects.
Also,
please
comment
on
whether
the
overall
body
of
available
data
is
adequate
to
demonstrate
whether
or
not
atrazine
causes
developmental
effects
under
the
conditions
described
in
these
studies.
Page
­
80­
b)
Please
comment
on
EPA's
conclusion
that
given
the
variability
in
the
available
dose­
response
data
across
the
studies
(
e.
g.,
an
approximately
250­
fold
difference
in
reported
thresholds
for
observed
developmental
effects
as
well
as
reports
of
monotonic
and
non­
monotonic
dose­
response
curves),
it
is
not
possible
to
ascertain
the
relationship,
if
any,
of
atrazine
exposure
to
developmental
effects
in
amphibians.

4)
Many
of
the
available
studies
proposed
that
aromatase
induction
results
in
elevated
estrogen
levels
that
lead
to
feminization
(
ovotestes/
intersex/
hermaphroditism)
in
genetically
male
amphibians.

e)
Please
comment
on
EPA's
conclusion
that,
to
date,
aromatase
induction
by
atrazine
has
not
been
demonstrated
in
any
anuran
in
controlled
laboratory
investigations.

f)
The
variability
associated
with
plasma
sex
steroid
concentrations
and
aromatase
activities
is
high.
Is
this
variability
normal?
Please
comment
on
any
readily
apparent
or
available
methodological
improvements
(
e.
g.,
changes
in
sampling
design,
analytical
techniques)
that
could
efficiently
address
this
variability
in
future
studies.

c)
Please
comment
on
whether
there
are
additional
data,
other
than
those
summarized
in
the
white
paper,
that
suggest
late
exposure
of
amphibians
(
i.
e.,
juveniles
or
adults)
to
estrogens
or
estrogenic
chemicals
can
induce
ovotestes
formation.
Page
­
81­
d)
Please
comment
on
whether
there
are
additional
data,
other
than
those
summarized
in
the
white
paper,
that
suggest
alternative
mechanisms
that
could
explain
the
apparent
feminization
of
genetically­
male
amphibians.

5)
With
regard
to
specific
endpoints,
the
Agency
does
not
currently
have
sufficient
information
to
quantitatively
relate
gonadal/
laryngeal
effects
to
reproductive
outcomes.
A
major
underlying
uncertainty
is
the
ecological
relevance
of
ovotestes
occurrence
to
the
maintenance
of
anuran
populations.

g)
Can
the
Panel
provide
sources
of
data
on
background
rates
of
ovotestes
occurrence
in
amphibian
species
and
any
associated
considerations
for
interpreting
this
information
in
the
context
of
the
reviewed
studies?

h)
Can
the
Panel
characterize
any
evidence
that
suggests
that
the
presence
of
ovotestes
in
male
anurans
results
in
reproductive
impairment
via
reductions
in
fertility?

i)
The
reduction
of
laryngeal
muscle
area
suggests
diminished
testosterone
in
males.

If
this
is
found
to
be
a
valid
observation
and
if
estrogen
concentrations
do
increase
as
testosterone
concentrations
decrease,
what
other
endpoints
(
e.
g.,
secondary
sexual
characteristics
and
reproductive
behavior)
would
likely
be
affected?

6)
While
some
of
the
available
data
indicate
there
may
be
an
association
between
atrazine
exposure
and
developmental
effects
in
amphibians,
the
Agency's
evaluation
of
the
existing
body
of
laboratory
and
field
studies
has
determined
that
there
is
not
sufficient
scientific
evidence
to
indicate
that
atrazine
consistently
produces
effects
across
the
range
of
amphibian
species
examined.
However,
the
current
body
of
knowledge
has
deficiencies
Page
­
82­
and
uncertainties
that
limit
its
usefulness
in
assessing
potential
developmental
atrazine
effects
and
the
extent
of
any
associated
cause­
effect
and
dose­
response
relationships.

Consequently,
the
Agency
has
determined
that
there
are
not
sufficient
data
to
reject
the
hypothesis
that
atrazine
can
cause
adverse
developmental
effects
in
amphibians.

f)
Does
the
SAP
concur
with
these
conclusions?
If
not,
what
lines­
of­
evidence
would
lead
to
an
alternative
conclusion?

7)
Assuming
the
Agency
determined
an
ecological
risk
assessment
with
a
greater
degree
of
certainty
concerning
developmental
effects
of
atrazine
on
amphibians
were
needed,
please
comment
on
EPA's
conclusion
that
additional
information
is
required
to
evaluate
potential
causal
relationships
between
atrazine
exposure
and
gonadal
development.
Please
also
comment
on
the
added
utility,
if
any,
of
additional
information
to
interpret
the
shape
of
dose­
response
curves
for
potential
developmental
endpoints
and
the
extent
to
which
threshold
or
non­
threshold
response
relationships
can
be
quantified.

8)
The
Agency
has
developed
a
conceptual
model
from
which
to
develop
a
set
of
study
protocols
for
evaluating
the
potential
effects
of
atrazine
on
gonadal
development
in
amphibians.
The
Agency
has
proposed
a
research
approach
using
focused,
empirical,

laboratory
studies
based
on
initial
investigations
with
X.
laevis
followed
by
selective,

confirmatory
studies
with
frog
species
native
to
North
America.

a)
Please
comment
on
the
proposed
sequence
of
study
objectives.

b)
Please
comment
on
whether
the
Agency's
first
set
of
proposed
studies
has
accounted
for
the
major
sources
of
uncertainty
associated
with
the
potential
effects
Page
­
83­
of
atrazine
on
anuran
sexual
differentiation.
In
addition
to
time
to
metamorphosis,

gonadal
abnormalities,
and
sex
ratios
in
the
proposed
Phase
I
assays,
please
comment
on
any
other
endpoints
that
should
be
considered
in
this
initial
phase.

c)
Please
also
comment
on
the
range,
spacing
and
number
of
atrazine
concentrations
that
should
be
employed
in
the
proposed
testing
sequence
to
resolve
uncertainties
in
the
shape
and
nature
of
dose­
response
relationships
for
any
observed
developmental
effects.

d)
Please
comment
on
the
Agency's
recommendation
that
X.
laevis
be
used
as
the
primary
biological
model
in
the
proposed
studies
and
whether
or
not
the
mechanisms
involved
in
sexual
differentiation
of
the
ranid
and
pipid
species
are
sufficiently
similar
to
predict
effects
and
associated
dose­
response
curves
for
Rana
and/
or
to
efficiently
design
Rana
studies.

e)
In
this
regard,
are
there
important
differences
between
the
species
to
conclude
that
any
affected
developmental
processes
observed
in
X.
laevis
would
not
occur
in
Rana?

f)
Alternatively,
are
there
developmental
pathways
in
Rana,
but
not
in
X.
laevis,
that
raise
concerns
about
using
X.
laevis
as
the
primary
biological
model
in
any
future
atrazine
studies?

g)
Assuming
X.
laevis
and
Rana
are
sufficiently
concordant
from
a
toxicodynamic
perspective
with
regard
to
potential
developmental
effects
of
atrazine,
what
critical
toxicokinetic
processes
should
be
considered
for
extrapolating
X.
laevis
doseresponse
relationships
to
Rana
and/
or
for
designing
subsequent
studies
with
Rana?
Page
­
84­
REFERENCES
Ankely,
G.
T.,
J.
E.
Tietge,
D.
L.
DeFoe,
K.
M
.
Jensen,
G.
W.
Holcombe,
E.
J.
Durhan,
S.
A.

Diamond.
1998.
Effects
of
ultraviolet
light
and
methoprene
on
survival
and
development
of
Rana
pipiens.
Environmental
Toxicology
and
Chemistry
17:
2530­
2542.
Page
­
85­
Birge,
W.
J.,
A.
G.
Westerman
and
J.
A.
Spromberg.
2000.
Comparative
toxicology
and
risk
assessment
of
amphibians.
Pages
727
­
791
in
D.
W.
Sparling,
G.
Linder
and
C.
A.
Bishop,

editiors.
Ecotoxicology
of
Amphibians
and
Reptiles.
Society
of
Environmental
Toxicology
and
Chemistry
Press.

Calabrese,
E..
J.
and
L.
A.
Baldwin.
2001.
Hormesis:
a
generalizable
and
unifying
hypothesis.

2001.
Critical
Reviews
in
Toxicology
31
(
4
&
5):
353
­
424.

Carr,
J.
A.,
A.
Gentles,
E.
E.
Smith,
W.
L.
Goleman,
L
J.
Urquidi,
K.
Thuett,
R.
J.
Kendall,
J.
P.

Giesy,
T.
S.
Gross,
K.
R.
Solomon,
G.
Van
Der
Kraak,
2003.
Response
of
larval
Xenopus
laevis
to
atrazine:
assessment
of
growth,
metamorphosis,
and
gonadal
and
laryngeal
morphology.

Environmental
Toxicology
and
Chemistry
22(
2):
396
­
405.

40
CFR.
1997.
Code
of
Federal
Regulations
40.
Parts
158
to
189
(
Protection
of
the
Environment).
Office
of
the
Federal
Register
National
Archives
and
Records
Administration.

Consent
Decree
(
as
amended)
entered
in
Natural
Resources
Defense
Council
v.
Whitman,
Case
Number
C
­
99­
3701
CAL,
N.
D.
California
(
2002).

Crabtree,
C.;
E.
E.
Smith;
J.
A.
Carr.
2003.
Histology
of
the
gonads
and
analysis
of
hormone
levels
in
the
native
bull
frog
(
Rana
catesbiena)
collected
from
agricultural
areas
in
southern
Iowa:

pilot
project.
The
Institute
of
Environmental
and
Human
Health,
Texas
Technical
University,
Page
­
86­
Lubbock,
Texas.
Sponsor:
Syngenta
Crop
Protection,
Inc.
Laboratory
Identification
Number
ECORISK
Number
TTU­
02.

De
Prado,
R.,
N.
Lopez­
Martinez
and
J.
Gonzalez­
Gutierrez.
2000.
Identification
of
two
mechanisms
of
atrazine
resistance
in
Setaria
faberi
and
Setari
viridis
biotypes.
Pesticide
Biochemistry
and
Physiology
67(
2):
114
­
124.

Duellman,
W.
E.
and
L.
Trueb.
1994.
Biology
of
Amphibians.
Johns
Hopkins
University
Press,

Baltimore,
MD.

DuPreez,
L.
H.,
A.
M.
Jooste
and
K.
R.
Solomon.
2003.
Exposure
of
Xenopus
laevis
larvae
to
different
concentrations
of
atrazine
in
semi­
natural
microcosms.
School
of
Environmental
Sciences
and
Development,
Zoology
Department,
Potchefstroom
University
for
CHE,

Potchefstroom
2520,
South
Africa.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
SA­
01­
D.

Giesy,
J.
P.,
M.
Hecker,
and
P.
D.
Jones.
2003.
South
African
analytical
support
 
hormone
and
aromatase
analysis
(
SA­
01C).
Aquatic
Toxicology
Laboratory,
Michigan
State
University,

National
Food
Safety
and
Toxicology
Center,
E.
Lansing,
MI.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
MSU­
07*

Goleman,
W.
L.
and
J.
A.
Carr.
2003.
Response
of
larval
Xenopus
laevis
to
atrazine
exposure:

assessment
of
metamorphosis
and
gonadal
and
laryngeal
morphology.
The
Institute
of
Page
­
87­
Environmental
and
Human
Health,
Texas
Tech
University,
Texas
Tech
University
Health
Sciences
Center,
Lubbock,
Texas.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
TTU­
01.

Gray,
J.
A.,
N.
E.
Balke
and
D.
E.
Stoltenberg.
1996.
Increased
glutathione
conjugation
of
atrazine
confers
resistance
in
a
Wisconsin
velvetleaf
(
Abutilon
theophrasti)
biotype.
Pesticide
Biochemistry
and
Physiology
55(
3):
157
­
171.

Hayes,
T.,
K.
Haston,
M.
Tsui,
A.
Hoang,
C.
Haeffele
and
A.
Vonk.
2002b.
Atrazine­
induced
hermaphroditism
at
0.1
ppb
in
American
leopard
frogs
(
Rana
pipiens):
laboratory
and
field
evidence.
Environmental
Health
Perspectives
http://
dx.
doi.
org/
[
Online
23
October
2002]

Hayes,
T.
B.,
A.
Collins,
M.
Lee,
M.
Mendoza,
N.
Noriega,
A.
A.
Stuart,
and
A.
Vonk.
2002a.

Hermaphroditic,
demasculinized
frogs
after
exposure
to
the
herbicide
atrazine
at
low
ecologically
relevant
doses.
Proceedings
of
the
National
Academy
of
Sciences
99
(
8):
5476
­
5480
Hayes,
T.,
K.
Haston,
M.
Tsui,
A.
Hoang,
C.
Haeffele
and
A.
Vonk.
2002c.
Feminization
of
male
frogs
in
the
wild:
water­
borne
herbicide
threatens
amphibian
populations
in
parts
of
the
United
States.
Nature
419:
895
­
896.
Abridged
Version
of
Hayes
et
al.
2002b
and
not
specifically
reviewed
in
white
paper.

Hayes,
T.
B.
1998.
Sex
determination
and
primary
sex
differentiation
in
amphibians:
genetic
and
developmental
mechanisms.
Journal
of
Experimental
Zoology
281:
373
­
399.
Page
­
88­
Hecker,
M.,
K.
K.
Coady,
D.
L.
Villeneuve,
M.
B.
Murphy,
P.
D.
Jones
and
J.
P.
Giesy.
2003.
A
pilot
study
of
response
of
larval
Rana
clamitans
to
atrazine
exposure:
assessment
of
metamorphosis
and
gonadal
and
laryngeal
morphology
and
selected
hormones
and
enzyme
activities.
Aquatic
Toxicology
Laboratory,
Michigan
State
University,
National
Food
Safety
and
Toxicology
Center,
E.
Lansing,
MI.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
MSU­
03.

Hecker,
M.,
K.
K.
Coady,
D.
L.
Villeneuve,
M.
B.
Murphy,
P.
D.
Jones
and
J.
P.
Giesy.
2003.

Response
of
Xenopus
laevis
to
atrazine
exposure:
assessment
of
the
mechanism
of
action
of
atrazine.
Aquatic
Toxicology
Laboratory,
Michigan
State
University,
National
Food
Safety
and
Toxicology
Center,
E.
Lansing,
MI.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
MSU­
04.

HED
2002.
U.
S.
EPA
Office
of
Pesticide
Programs
Human
Health
Effect
Division
Revised
Human
Health
Risk
Assessement
for
the
Reregiatration
Eligibility
Decision
(
RED).
April
16,

2002.
OPP
Docket
#
2002­
0026.
Docket
Number
0012.

Huang,
H,
L
Cai,
BF
Remo,
and
DD
Brown.
2001.
Timing
of
metamorphosis
and
the
onset
of
the
negative
feedback
loop
between
the
thyroid
gland
and
the
pituitary
is
controlled
by
type
II
iodothyronine
deiodinase
in
Xenopus
laevis.
Proceeding
of
the
National
Academy
of
Sciences,

98,
7348­
7353
Page
­
89­
IRED
Science
Chapter
for
Atrazine:
Environmental
Fate
and
Effects
Chapter,
April
22,
2002
OPP
Docket
Number
2002­
0026
Document
Number
0005.

IRED.
2003.
Interim
Reregistration
Eligibility
Decision
Science
Chapter
for
Atrazine:

Environmental
Fate
and
Effects
Chapter
January
28,
2003.
OPP
Docket
ID
OPP­
2003­
0072;

Document
ID
OPP­
2003­
0072­
0010.

Jones,
P.
D.,
M.
B.
Murphy,
M.
Hecker,
J.
P.
Giesy.
2003.
Tissue
pesticide
residues
and
histology
of
the
larynx
and
gonads
in
native
green
frogs
(
Rana
clamitans)
collected
from
agricultural
areas
in
Michigan:
hormone
analysis.
Aquatic
Toxicology
Laboratory,
Michigan
State
University,
218C
National
Food
Safety
and
Toxicology
Center,
E.
Lansing,
MI.
Sponsor:

Syngenta
Crop
Protection,
Inc.
Laboratory
Study
ID:
ECORISK
Number
MSU­
02.

Miyashita,
K,
N
Shimizu,
S
Osanai,
and
S
Miyata.
2000.
Sequence
analysis
and
expression
of
the
P450
aromatase
and
estrogen
receptor
genes
in
the
Xenopus
ovary.
J
Ster
Bioch
Molec
Biol,
75,

101­
107.

Nieuwkoop,
P.
D.,
and
J.
Faber.
1994.
Normal
table
of
Xenopus
laevis
(
Daudin).
North
Holland
Publishing
Company,
Amsterdam,
Holland.

Petrini,
S.
F.
Zaccanti.
1998.
The
effects
of
aromatase
and
5­
 
reductase
inhibitors,

anitandrogens
and
sex
steroids
on
Bidder's
organs
development
and
gonadal
differentiation
in
Bufo
bufo
tadpoles.
Journal
of
Experimental
Zoology
280:
245
­
259.
Page
­
90­
Reeder,
A.
L.,
G.
L.
Foley,
D.
K.
Nichols,
L.
G.
Hansen,
B.
Wikoff,
S.
Faeh,
J.
Eisold,
M.
B.

Wheeler,
R.
Warner,
J.
E.
Murphy,
and
V.
R.
Beasley.
1998.
Forms
and
prevalence
of
intersexuality
and
effects
of
environmental
contaminants
on
sexuality
in
cricket
frogs
(
Acris
crepitans).
Environmental
Health
Perspectives
106
(
5):
261
­
266
Reregistration
Eligibility
Decision
Science
Chapter
for
Atrazine:
Environmental
Fate
and
Effects
Chapter,
April
22,
2002
OPP
Docket
Number
2002­
0026
Document
Number
0005
Sanderson,
J.
T.,
J.
Boerma,
G.
W.
A.
Lansbergen
and
M.
van
den
Berg.
2002.
Induction
and
inhibition
of
aromatase
(
CYP19)
activity
by
various
classes
of
pesticides
in
H295R
human
adrenocortical
carcinoma
cells.
Toxicology
and
Applied
Pharmacology
182:
44
­
54.

Sepulveda,
M.
S.
and
T.
S.
Gross.
2003.
Characterization
of
atrazine
exposures
and
potential
effects
in
Florida
ecosystems
dominated
by
sugarcane
agriculture:
a
reconnaissance
survey
of
amphibians
in
south
Florida
for
the
assessment
of
potential
atrazine
effects.
Department
of
Physiological
Sciences,
University
of
Florida,
Caribbean
Science
Center,
Gainesville,
Florida.

Sponsor:
Syngenta
Crop
Protection,
Inc.
Study
ID:
ECORISK
Number
UFL­
02.

Sive,
H.
L.,
R.
M.
Granger
and
R.
M.
Harland.
1998.
Early
Development
of
Xenopus
laevis,
a
Laboratory
Manual.
Cold
Springs
Harbor
Laboratory
Press.
4*
Individual
DERs
written
on
study,
but
report
collectively
reviewed
as
one
of
four
component
studies
to
the
South
African
Xenopus
investigations.

Page
­
91­
Smith,
E.
E.,
L.
DuPreez,
and
K.
Solomon.
2003.
Field
exposure
of
Xenopus
laevis
to
atrazine
and
other
triazines
in
South
Africa:
feasibility
study
for
site
characterization
and
assessment
of
laryngeal
and
gonadal
responses.
The
Institute
of
Environmental
and
Human
Health,
Texas
Tech
University,
Lubbock,
Texas
(
USA)
and
School
of
Environmental
Sciences
and
Development,

Potchefstroom
University
for
CHE,
Potchefstroom,
South
Africa.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID:
ECORISK
Number
SA­
01A*
4.

Smith,
E.,
L
DuPreez
and
K.
Solomon.
2003.
Field
exposure
of
Xenopus
laevis
to
atrazine
and
other
triazines
in
South
Africa:
exposure
characterization
and
assessment
of
laryngeal
and
gonadal
responses.
The
Institute
of
Environmental
&
Human
Health,
Texas
Tech
University,

Lubbock,
Texas
79490
(
USA)
and
School
of
Environmental
Sciences
and
Development,

Potchefstroom
University
for
CHE,
Private
Bag
X6001,
Potchefstroom
2520
(
South
Africa).

Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
SA­
01B*

Smith,
E.
E.,
L.
DuPreez,
and
K.
Solomon.
2003.
Gonadal
and
laryngeal
responses
to
field
exposure
of
Xenopus
laevis
to
atrazine
in
areas
of
corn
production
in
South
Africa.
The
Institute
of
Environmental
and
Human
Health,
Texas
Tech
University,
Lubbock,
TX
and
School
of
Environmental
Sciences
and
Development,
Potchefstroom
University
for
CHE,
Potchefstroom,

South
Africa.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID:
ECORISK
Number
SA­
01C*.
Page
­
92­
Tavera­
Mendoza,
L.,
S.
Ruby,
P.
Brousseau,
M.
Fournier,
D.
Cyr
and
D.
Marcogliese.
2001.

Response
of
the
amphibian
tadpole
Xenopus
laevis
to
atrazine
during
sexual
differentiation
of
the
ovary.
Environmental
Toxicology
and
Chemistry
21
(
6):
1264
­
1267.

Tavera­
Mendoza,
L.,
S.
Ruby,
P.
Brousseau,
M.
Fournier,
D.
Cyr
and
D.
Marcogliese.
2001.

Response
of
the
amphibian
tadpole
Xenopus
laevis
to
atrazine
during
sexual
differentiation
of
the
testes.
Environmental
Toxicology
and
Chemistry
21
(
3):
527
­
531.

Villalpando,
I
and
H
Merchant­
Larios.
1990.
Determination
of
the
sensitive
stages
for
gonadal
sex­
reversal
in
Xenopus
laevis
tadpoles.
International
Journal
of
Developmental
Biology,
34,

281­
285.

Villeneuve,
D.
L.,
K.
Coady,
M.
Hecker,
M.
B.
Murphy,
P.
D.
Jones
and
J.
P.
Giesy.
2003.

Methods
development
for
the
study
of
mechanism
of
action
of
atrazine
in
adult
and
metamophosing
Xenopus
laevis
and
Rana
clamitans:
aromatase
induction.
Aquatic
Toxicology
Laboratory,
Michigan
State
University,
National
Food
Safety
and
Toxicology
Center,
E.
Lansing,

MI.
Sponsor:
Syngenta
Crop
Protection,
Inc.,
Laboratory
Study
ID
ECORISK
Number
MSU­

01.
Page
­
93­
Test
for
Apical
Gonadal
Effects
Aromatase
Activity
Measurements
Sex
Steroid
Measurements
Test
for
Alternative
Effects
Relevant
to
Sex
Steroid
Homeostasis
Test
for
Alternative
Effects
Relevant
Gonadal
Differentiation
Stop
YES
YES
NO
NO
NO
Aromatase
Inhibitor
Study
YES
1
2
3
4
Working
Hypothesis
Ecological
Relevancy
Study
5
Figure
1.
Scheme
for
laboratory
studies
to
determine
the
effects
of
atrazine
on
gonadal
differentiation
in
anuran
amphibians.
The
grey
block
contains
the
working
hypothesis
for
atrazine
action
on
gonadal
differentiation,
which
is
partitioned
into
four
logical
phases.
See
text
for
details.
Page
­
94­
Developmental
Stage
42
44
46
48
50
52
54
56
58
Aromatase
mRNA
0
100
200
300
400
500
Sex
Reversal
Intersex
No
effect
Figure
2.
The
effect
of
estradiol
exposure
on
male
gonads
of
different
stages
compared
to
the
developmental
expression
of
aromatase
mRNA.
The
horizontal
bar
indicates
that
there
are
three
periods
of
differential
sensitivity
to
the
feminizing
effects
of
estrogen
exposure
for
male
X.
laevis
larvae
(
Villalpando
and
Merchant­
Larios,
1990).
The
line
indicates
the
relative
expression
of
aromatase
mRNA
at
different
developmental
stages
(
Miyashita
et
al.
2000).
Page
­
95­
Developmental
Stage
46
48
50
52
54
56
58
60
62
64
66
68
Wet
Weight
(
mg)

0
200
400
600
800
1000
1200
1400
1600
1800
Figure
3.
The
relationship
of
weight
to
developmental
stage
of
X.
laevis.
Each
dot
represents
one
organism
(
Tietge
et
al.,
EPA
Midcontinent
Ecology
Division,
unpublished
results.)
