FIFRA
SCIENTIFIC
ADVISORY
PANEL
(
SAP)

OPEN
MEETING
POTENTIAL
DEVELOPMENTAL
EFFECTS
OF
ATRAZINE
ON
AMPHIBIANS
June
17,
2003
[
8:
30
a.
m.]

Crowne
Plaza
Hotel
Washington
National
Airport
1489
Jefferson
Davis
Highway
Arlington,
Virginia
22202
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
2
PARTICIPANTS
FIFRA
SAP
Session
Chair
Stephen
M.
Roberts,
Ph.
D.

Designated
Federal
Official
Paul
Lewis
FIFRA
Scientific
Advisory
Panel
Members
Steven
Herringa,
Ph.
D.

Gary
E.
Isom,
Ph.
D.

Fumio
Matsumura,
Ph.
D.

Mary
Anna
Thrall,
DVM,
MS
FQPA
Science
Review
Board
Members
Joel
Coats,
Ph.
D.

Peter
Delorme,
Ph.
D.

Robert
J.
Denver,
Ph.
D.

James
Gibbs,
Ph.
D.

Sherril
L.
Green,
DVM,
Ph.
D.

Darcey
B.
Kelley,
Ph.
D.

Werner
Kloas,
Ph.
D.

Gerald
A.
LeBlanc,
Ph.
D.

Carl
Richards,
Ph.
D.

David
Skelly,
Ph.
D.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
3
DR.
ROBERTS:
The
topic
for
this
session
is
going
to
be
the
potential
developmental
effects
of
atrazine
on
amphibians.
I
would
like
to
begin
the
meeting
by
introducing
our
designated
federal
official,
Mr.
Paul
Lewis,
and
ask
if
he's
got
any
announcements
for
us.

MR.
LEWIS:
I
thank
you
Dr.
Roberts.
And
I
want
to
first
thank
Dr.
Roberts
for
serving
as
our
incoming
chair
for
the
FIFRA
SAP,
looking
forward
to
working
with
him.
And
also
to
acknowledge
our
permanent
panel
members,
Dr.
Gary
Isom
and
Dr.
Steven
Herringa.
Dr.
Handworker,
also
another
permanent
panel
member,
is
unavailable
at
this
meeting
today
but
will
be
here
in
July.

I
am
Paul
Lewis
and
I
will
be
serving
as
a
designated
federal
official
to
the
FIFRA
SAP
for
this
meeting
over
the
next
four
days.
I
want
to
thank
both
the
members
of
the
panel
and
the
ad
hoc
members
for
agreeing
to
serve
the
next
four
days
for
what
I
think
we'll
find
a
very
challenging
and
interesting
discussion
that
we'll
be
having.

Reviewing
the
potential
developmental
effects
of
atrazine
on
amphibians,
we
appreciate
the
time
and
the
effort
of
the
panel
members
in
reviewing
the
materials
and
preparing
their
remarks,

taking
into
account
their
busy
schedules.

By
way
of
background,
the
FIFRA
SAP
is
a
federal
advisory
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
4
committee
that
provides
independent
scientific
peer
review
and
advice
to
the
Agency
on
pesticides
and
pesticide­
related
issues
regarding
the
impact
of
proposed
regulatory
actions
on
human
health
and
the
environment.
The
FIFRA
SAP
only
provides
advice
and
recommendations
to
the
Agency,
that
is,
making
implementation
authority
remains
with
the
Agency.

FIFRA
established
what
is
called
a
permanent
panel
which
consists
of
seven
members.
The
expertise
of
the
panel
is
also
augmented
through
a
science
review
board
and
science
review
board
members
serve
as
ad
hoc
temporary
members
of
the
SAP
providing
additional
scientific
expertise
to
assist
in
reviews
conducted
by
the
panel.

As
the
designated
federal
official
for
this
meeting,
I
serve
as
liaison
between
the
panel
and
the
Agency.
And
I'm
also
responsible
for
insuring
provisions
of
the
Federal
Advisory
Committee
Act
are
met.

The
Federal
Advisory
Committee
Act
establishes
a
system
for
governing
the
creation,
operation,
termination
of
executive
branch
advisory
committees.
It
highlights
the
consideration
of
federal
advisory
committees
on
the
FACA
as
follows:
The
committees
are
chartered.
They
are
governed
by
uniform
procedures.
They
provide
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
5
only
advice,
are
open
to
public
scrutiny.

In
addition,
FIFRA
SAP
meetings
are
also
subject
to
FACA
requirements.
And
these
include
public
meetings,
timely
public
notice
of
the
meetings,
and
document
availability.
In
that
respect,
we
have
documents
for
this
meeting.
The
background
paper,
public
comments,
and
the
final
meeting
will
be
available
through
the
Office
of
Pesticide
Programs
docket.

In
terms
of
financial
conflicts
of
interest,
as
the
designated
federal
official
of
this
meeting,
a
critical
responsibility
is
to
work
with
appropriate
Agency
officials
to
ensure
all
appropriate
ethics
regulations
are
satisfied.
In
that
capacity,
panel
members
are
briefed
with
provisions
of
the
federal
conflict
of
interest
laws.
Each
participant
has
filed
a
standard
government
financial
disclosure
report.

I
along
with
our
deputy
ethics
officials
for
the
Office
of
Prevention
of
Pesticide
and
Toxic
Substances,
in
a
consultation
with
the
office
general
counsel,
have
reviewed
the
report
to
ensure
all
ethics
requirements
are
met.
And
a
sample
copy
of
the
form
is
available
on
our
FIFRA
SAP
web
site.

The
panel
will
review
challenging
science
issues
over
the
next
several
days.
We
have
a
full
agenda
and
meeting
times
are
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
6
approximate.
This
may
not
keep
to
the
exact
times
as
noted
due
to
panel
discussions
and
public
comments.
We
strive
to
ensure
adequate
time
for
Agency's
presentations,
public
comments
being
presented,

and
panel
deliberations.

For
panel
members
and
public
commentors,
please
identify
yourselves
and
speak
into
the
microphones
provided
since
the
meeting
is
being
recorded.
Copies
of
all
presentation
materials,
public
comments,
will
be
available
in
the
Office
of
Pesticide
Program
docket,
as
I
mentioned
previously,
within
the
next
few
days.

For
members
of
the
public
requesting
time
to
make
a
public
comment,
please
limit
your
comments
to
five
minutes
unless
prior
arrangements
have
been
made.
For
those
that
have
not
preregistered,

please
notify
myself
or
a
member
of
the
FIFRA
SAP
staff
who
are
sitting
just
to
the
right
of
me
here.

As
I
mentioned
previously,
there
is
a
public
docket
for
this
meeting.
All
background
materials,
questions
posed
to
the
panel
by
the
Agency,
and
other
documents
related
to
this
SAP
meeting
are
available
in
the
docket.
Overheads
will
be
available
in
the
next
few
days.

Finally,
background
documents
are
also
available
on
the
SAP
web
site.
And
the
agenda
for
this
meeting
lists
contact
information
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
7
for
that
type
of
material.

For
members
of
the
press,
we
have
a
press
contact
from
our
office,
Mr.
David
Deacon,
of
the
Office
of
Media
Relations,
is
available
to
answer
your
questions.
Mr.
Deacon,
please
stand.
Thank
you.
Any
interested
people
who
have
questions
from
the
press,
please
refer
them
to
Mr.
Deacon.

At
the
conclusion
of
the
meeting,
the
FIFRA
SAP
will
prepare
a
report
as
response
to
questions
posed
by
the
Agency,
background
materials,
presentation,
and
public
comments.
The
report
serves
as
meeting
minutes,
and
we
anticipate
the
minutes
to
be
available
in
approximately
two
to
four
weeks
and
will
be
posted
on
our
web
site
and
in
the
OPP
docket.

Finally,
due
to
unforeseen
circumstances,
this
Friday's
meeting
will
move
from
this
location
to
the
Holiday
Inn
National
Airport,

2650
Jefferson
Davis
Highway
in
Arlington,
Virginia.
The
meeting
address
for
this
Friday's
meeting
is
noted
on
the
meeting
agenda
outside
this
room,
and
you'll
notice
some
placards
in
the
hallway
when
you
first
enter
the
room.
The
Holiday
Inn
is
approximately
8
to
10
blocks
from
our
present
location
and
parking
is
available.
This
meeting
change
will
only
be
for
this
Friday,
June
20.
All
other
days,

our
meeting
will
be
occurring
here.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
8
We
apologize
for
any
inconvenience
that
may
occur
and
are
making
special
arrangements
for
this
change
of
meeting
location
this
Friday.
If
you
require
any
assistance
to
attend
the
Friday
meeting,

including
maps
or
shuttle
service
to
the
Holiday
Inn
hotel,
please
visit
a
member
of
the
FIFRA
SAP
meeting.

Thank
you.
Dr.
Roberts.

DR.
ROBERTS:
Thank
you,
Paul.
The
SAP
staff
has
assembled
an
outstanding
panel
of
experts
to
deal
with
this
topic,
and
I
would
like
to
introduce
that
panel
now
and
do
so
by
asking
each
member
of
the
panel
to
state
their
name,
their
affiliation,
and
their
area
of
expertise.
And
I
think
we'll
just
go
around
the
table
counter­
clockwise
starting
with
Dr.
LeBlanc
on
my
immediate
right
and
then
for
each
member
of
the
panel
around
the
table
to
introduce
themselves.
Dr.
LeBlanc.

DR.
LEBLANC:
Thank
you.
My
name
is
Gerry
LeBlanc.
And
I'm
a
professor
in
the
Department
of
Environmental
and
Molecular
Toxicology
at
North
Carolina
State
University.
My
area
of
expertise
is
awake
toxicology.

DR.
KELLEY:
I'm
Darcy
Kelley.
I'm
professor
of
Biological
Sciences
and
a
member
of
the
Center
for
Environmental
Research
and
Conservation
at
Columbia
University.
And
my
area
of
expertise
is
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
9
sexual
differential
of
the
amphibian
Xenopus
laevis.

DR.
KLOAS:
My
name
is
Werner
Kloas.
I'm
professor
for
endocrinology
at
University
of
Berlin.
And
I'm
also
heading
the
Inland
Fisheries
Department
of
Leibniz­
Institute
of
Freshwater
Ecology
and
Inland
Fisheries.
My
working
group
is
the
research
focused
on
awake
disruption
in
amphibians,
especially
addressing
sexual
differentiation
and
also
the
thyroid
system.

DR.
GREEN:
My
name
is
Sherril
Green,
and
I'm
an
associate
professor
in
the
Department
of
Comparative
Medicine
at
Stanford
University.
My
area
of
interest
and
expertise
is
in
laboratory
Xenopus
laevis,
Xenopus
laevis
and
Rana
pipiens
specifically
as
a
laboratory
animal
model.

DR.
COATS:
My
name
is
Joel
Coats.
I'm
professor
of
entomology
and
toxicology
at
Iowa
State
University.
My
areas
of
expertise
are
in
pesticides,
especially
environmental
toxicology
and
environmental
chemistry.

DR.
DENVER:
My
name
is
Robert
Denver.
I'm
associate
professor
of
molecular,
cellular,
and
development
biology
at
the
University
of
Michigan.
And
my
area
of
expertise
is
developmental
neuroendocrinology
of
amphibians.

DR.
ROBERTS:
Let's
jump
over
to
Dr.
Gibbs.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
10
DR.
GIBBS:
My
name
is
James
Gibbs.
I'm
an
associate
professor
of
conservation
biology
at
the
State
University
of
New
York's
College
of
Environment
Science
and
Forestry
in
Syracuse,

New
York.
And
my
area
of
expertise
is
amphibian
demography
and
population
dynamics.

DR.
RICHARDS:
My
name
is
Carl
Richards.
I'm
a
professor
of
biology
at
University
of
Minnesota
Duluth
and
director
of
the
Minnesota
Sea
Grant
College
Program.
My
expertise
is
that
of
an
aquatic
ecologist
and
landscape
ecologist.

DR.
DELORME:
My
name
is
Peter
Delorme.
I'm
a
senior
risk
assessor
with
the
Canadian
Government
working
on
risk
assessments
of
pesticides.
My
area
of
expertise
is
aquatic
ecology.

DR.
SKELLY:
My
name
is
David
Skelly.
I'm
an
associate
professor
of
ecology
at
Yale
University.
And
my
area
of
expertise
is
population
and
community
ecology
of
amphibians.

DR.
MATSUMURA:
My
name
is
Fumio
Matsumura.
I'm
at
the
Department
of
Environmental
Toxicology.
My
area
of
expertise
is
molecular
toxicology.
I'm
interested
in
the
frogs,
too.

DR.
THRALL:
I'm
Mary
Anna
Thrall.
I'm
a
professor
of
veterinary
pathology
in
the
College
of
Veterinary
Medicine
at
Colorado
State
University.
And
my
area
of
expertise
is
veterinary
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
11
clinical
pathology.

DR.
ISOM:
I'm
Gary
Isom,
professor
of
toxicology
at
Purdue
University.
And
my
area
is
neural
toxicology
and
neural
degenerative
diseases.

DR.
HEERINGA:
I'm
Steve
Heeringa,
biostatistician
and
director
of
the
statistics
design
group
at
the
Institute
for
Social
Research
at
the
University
of
Michigan.
My
specialty
is
population­
based
studies
and
design
of
population­
based
studies.

DR.
ROBERTS:
And
I'm
Steve
Roberts.
I'm
a
professor
with
joint
appointments
in
the
College
of
Veterinary
Medicine
and
College
of
Medicine
at
the
University
of
Florida.
I
also
serve
as
director
of
the
Center
for
Environmental
and
Human
Toxicology
there.
My
areas
of
expertise
are
mechanisms
of
toxicity,
particularly
involving
the
liver
and
immune
systems
and
also
methods
of
risk
assessment.

I
would
now
like
to
welcome
Mr.
Merenda
who
is
Director
of
the
Office
of
Science
Coordination
and
Policy.
Good
morning,
Mr.

Merenda.

DR.
MERENDA:
Good
morning,
Steve,
and
welcome
to
all
of
the
panelists
as
well
as
all
of
those
who
are
participating
in
attending
this
session
in
the
audience.

Within
EPA,
the
concept
of
independent,
external
scientific
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
12
peer
review
plays
a
very
important
role
in
our
evaluation
of
decision
processes.
And
so
this
kind
of
event,
while
complex
to
organize,
is
a
very
important
part
of
our
job.
In
fact,
it's
an
important
part
of
the
job
of
my
office,
the
Office
of
Science
Coordination
and
Policy
in
EPA's
Office
of
Prevention
Pesticides
and
Toxic
Substances.

We
find
that
these
kinds
of
meetings
with
both
the
permanent
panel
members
and
a
number
of
expert
ad
hoc
members
who
are
selected
specifically
for
your
expertise
on
the
subject
atrazine
and
is
extremely
valuable
to
the
Environment
Protection
Agency
in
helping
us
to
better
understand
where
we've
done
things
well
and
where
may
have
missed
some
points
or
where
we
need
to
look
further
and
dig
deeper
as
we
evaluate
the
data
available
to
us
and
make
risk
management
and
regulatory
decisions.

This
is
going
to
be,
as
Paul
said,
a
very
full
program.
Over
the
next
four
days,
there
are
a
number
of
complex
issues
for
us
to
deal
with.
And
we
are
quite
pleased
to
have
the
expertise
that
is
being
brought
to
us
by
all
of
you
and
we're
quite
thankful
for
your
willingness
to
take
time
from
busy
schedules
and
other
commitments
to
spend
these
four
days
with
us
in
helping
us
better
understand
these
problems.

So
welcome,
and
just
to
reemphasize
what
Paul
said,
that
if
you
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
13
have
any
needs
for
logistics
or
assistance
with
making
this
meeting
work
better
or
making
your
own
arrangements
work
out
while
you're
here
at
the
meeting,
please,
do
not
hesitate
to
contact
a
member
of
our
staff
on
the
FIFRA
Scientific
Advisory
Panel.

DR.
ROBERTS:
Thank
you
very
much.
I
would
also
like
to
extend
the
panel's
welcome
this
morning
to
Ms.
Anne
Lindsay,
who
is
the
Acting
Deputy
Director
of
the
Office
of
Pesticide
Programs.
Good
morning.

MS.
LINDSAY:
Good
morning.
You've
taken
my
first
line
away.
I
was
going
to
say
I
was
Anne
Lindsay,
Acting
Deputy
Director
for
Programs
in
the
Office
of
Pesticide
Programs.

I'm
actually
here
on
behalf
of
Jim
Jones
who
is
our
relatively
new
office
director,
though
not
new
to
the
Pesticide
programs.
He
is
actually
dealing
with
family
responsibilities
this
week
and
asked
me
to
welcome
you
to
Washington,
and
in
particular,
to
thank
you
for
agreeing
to
serve
on
this
scientific
advisory
panel
and
the
sort
of
extensive
meeting
that
we've
got
set
up
for
the
week.
I'd
actually
hoped
I
was
going
to
welcome
you
to
a
sunny
Washington.
But
that
doesn't
look
like
that's
going
to
happen.

The
topic
of
this
meeting,
the
Potential
Developmental
Effects
of
Atrazine
on
amphibians,
is
one
that
has
generated
an
extraordinary
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
14
amount
of
interest
since
the
first
data
appeared
suggesting
a
link
between
atrazine
exposure
and
development
effects
in
frogs.
It's
not
only
a
topic
that's
drawn
a
lot
of
interest,
I
think
it's
fair
to
say,
that
it's
actually
been
extremely
controversial.

Now
for
those
of
us
who
work
at
EPA
on
Pesticide
regulation,

such
controversy
is
actually
often
or
frequently
part
and
parcel
of
doing
our
job.
It's
not
that
we
want
it
to
be
that
way.
But
that
is
often
how
it
is.
And
I
think
that's
the
case
because
the
issues
we
deal
with
are
usually
very
complicated
and
often,
as
I
think
the
subject
of
this
meeting,
really
falls
at
the
cutting
edge
of
science
so
the
answers
may
not
be
clear
cut.

We
find
ourselves
dealing
frequently
with
situations
in
which
the
answers
are
not
obvious.
And
different
groups
also
have
strongly
held
views
and
strongly
held
contrasting
views.
Over
the
years,
we've
developed
a
deep
appreciation,
therefore,
for
the
value
of
science
as
the
basis
of
our
work
and
to
have
that
science
guide
our
decisions.

And
that's
really,
frankly,
where
all
of
you
as
a
panel
and
as
experts
in
your
various
areas
come
in.
Your
job
is
to
help
us
figure
out
what
the
scientific
information
actually
does
tell
us,
where
we
can
trust
information,
where
there
are
questions
about
it.
We
want
your
advice
to
guide
us
as
we
move
forward
making
our
regulatory
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
15
decisions.

I've
spent
probably
not
only
the
largest
part
of
my
career
in
public
service
maybe
the
largest
part
of
my
life
at
this
point
in
public
service.
I
think
when
you
do
that,
you
have
to
be
able
to
have
a
great
deal
of
pride
and
draw
a
great
deal
of
satisfaction
in
contributing
back
to
the
community
and
the
country
that
you
come
from.
And
that
is
certainly,
I
think,
how
I
and
the
other
EPA
folks
who
will
be
presenters
here
today
feel
about
our
work.

We
hope
that
you,
too,
since
at
this
point
as
members
of
the
panel,
you're
part
of
the
public
service
in
effect,
will
take
that
sense
of
satisfaction
for
making
a
very
real
contribution
to
the
civic
life
of
our
country.
There's
no
doubt
in
my
mind
that
the
work
that
you
will
do
as
panel
members
has
enormous
value
to
us
at
EPA.
And
because
of
that,
I
believe
it
will
have
enormous
value
to
the
citizens
of
this
country.
So
I
want
to
thank
you
very
much
for
making
that
contribution
of
your
time
and
expertise.

I'd
like
in
particular
to
acknowledge
EPA's
thanks
to
Dr.

Roberts,
the
chairman
of
the
SAP.
This
is
actually,
I
believe,
your
first
meeting
as
chairman
of
the
SAP.
But
you
are
well­
known
and
a
highly
regarded
scientist
who's
served
for
a
number
of
years
on
our
panel
and
has
presided
as
session
chair
at
some
of
our
most
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
16
controversial
meetings
with
firmness
and
grace.
So
we're
looking
forward
to
a
continuation
of
that
firmness
and
grace.
We
thank
you
for
contributing
on
the
panel
with
these
new
responsibilities.

We'd
also
like
to
extend
our
appreciation
to
the
other
new
permanent
members
of
the
panel
who
I
think
Paul
has
introduced
but
Dr.
Steward
Handworker,
Steve
Heeringa,
and
Gary
Isom.
It's
a
delight
to
have
Dr.
Heeringa
back
with
us
after
serving
as
an
ad
hoc
panel
member
on
multiple
occasions
and
also
we
welcome
Dr.
Isom
as
a
permanent
member
of
the
SAP.

In
addition,
my
thanks
to
all
of
the
scientists
who
have
agreed
to
serve
as
expert
advisors
on
the
atrazine
issues.
Your
willingness
to
contribute
your
knowledge
and
expertise
to
sorting
out
pressing
scientific
issues
is
invaluable.

And
then
finally,
Paul,
I
want
to
thank
you
and
the
other
members
of
the
SAP
staff.
You
do
a
great
job
taking
care
of
the
panel
members
and
a
great
job
running
the
meeting.
And
for
that,
we're
deeply
appreciative.

So,
finally,
let
me
wish
you
the
best
for
the
upcoming
meetings
and
we
look
forward
to
receiving
your
report.

DR.
ROBERTS:
Well,
thank
you
for
your
kind
remarks.

They're
very
much
appreciated.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
17
Our
discussion
of
the
scientific
issues
is
going
to
begin
with
presentations
by
the
Agency.
And
the
first
presenter
is
Dr.
Steven
Bradbury
of
the
Office
of
Pesticide
Programs.
Good
morning,
Dr.

Bradbury.

DR.
BRADBURY:
Good
morning,
thank
you.
I'd
also
like
to
extend
my
thanks
to
the
SAP
staff
for
helping
to
organize
the
meeting
and
to
the
panel
for
the
discussions
we'll
be
having
over
the
next
several
days.
Your
input
and
advice
will
be
greatly
appreciated
and
very
important
part
of
the
scientific
analysis
that
we've
embarked
upon
with
the
white
paper.

What
I'd
like
to
do
in
my
presentation
is
go
over
a
few
of
the
issues
you'll
be
hearing
in
more
detail
from
Tom
Steeger
and
Joe
Tietge
later
this
morning.
And
to
review
a
bit
why
we're
here
and
what
we
hope
to
accomplish
over
the
next
several
days,
as
I've
mentioned
in
my
opening
remarks,
we're
looking
forward
to
obtaining
your
recommendations
on
our
analysis
to
date
regarding
the
potential
developmental
effects
of
atrazine
on
amphibians.
And
as
we
go
through
the
morning's
discussions,
we'll
be
walking
through
several
topics
of
particular
note.

One
will
be
the
integration
of
the
available
information
as
we've
summarized
in
the
white
paper.
We'll
also
be
discussing
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
18
aspects
of
that
information
in
the
context
of
how
it
allows
one
to
formulate
risk
hypotheses
about
the
potential
effects
of
atrazine
on
amphibian
development.
And
from
those
risk
hypotheses,

establishing
a
conceptual
model
for
potential
effects.
And
ultimately,

then
taking
a
look
at
an
analysis
plan
as
to
how
to
move
forward
in
the
context
of
that
available
information.

To
provide
a
little
background
as
to
where
we've
been
and
how
we've
got
here,
let's
go
back
to
January
31,
2003.
That's
when
EPA
released
an
Interim
Reregistration
Eligibility
Decision,
an
IRED.

And
that
document
included
an
assessment
of
human
health
risk
assessment
issues
as
well
as
taking
a
look
at
ecological
risk
assessment
issues.

In
the
January
'
03
ecological
risk
characterization,
we
followed
sort
of
the
basic
tenets
of
the
Agency's
guidelines
for
ecological
risk
assessment
and
took
a
look
at
exposure
issues
and
took
a
look
at
the
physical
chemical
properties
of
atrazine
and
compared
both
modeling
and
monitoring
and
find
general
consistency
between
monitoring
studies
and
the
classes
of
watersheds
and
uses
that
atrazine
is
associated
with.

The
effects
analysis
or
effects
characterization
in
that
document
focused
on
the
integrity
of
aquatic
communities
or
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
19
stability
of
aquatic
communities
structure
and
function
as
a
risk
assessment
endpoint
and
came
to
the
conclusion
the
likelihood
of
adverse
effects
would
occur
at
approximately
10
to
20
micrograms
per
liter
over
recurrent
or
prolonged
periods
of
exposure.
And
this
analysis
was
similar
to
a
somewhat
independent
analysis
going
on
in
the
Office
of
Water
as
they
develop
their
draft
water
quality
criteria
for
atrazine.

In
the
document,
we
also
discuss
some
of
the
uncertainties
that
are
associated
with
this
risk
assessment
and
all
risk
assessments
have
different
levels
in
context
of
uncertainty.
Then
we
discuss
some
of
the
exposure
characterizations
uncertainties,
including
some
data
gaps
in
terms
of
being
able
to
predict
the
atrazine
concentration
patterns
and
attributes
across
a
full
population
of
water
body
types
in
the
United
States.
And
we
also
discussed
some
of
the
uncertainties
in
terms
of
the
spatial
and
temporal
variability
of
atrazine
and
the
ability
to
predict
or
model
those
patterns.

In
terms
of
effects
characterization,
we
discussed
the
challenge
of
converting
steady
state
atrazine
exposures
to
fluctuating
or
transient
atrazine
exposures
and
how
to
work
through
the
dosimetry
of
that
kind
of
an
exposure
scenario.
We
talked
about
the
issues
in
terms
of
quantifying
aquatic
community
recover
or
resistance
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
20
repeated
exposures
to
the
compound.
And,
finally,
we
also
discussed
the
uncertainties
concerning
the
potential
developmental
effects
of
atrazine
on
amphibians.

At
the
time
that
the
IRED
was
being
prepared,
there
were
several
studies
that
were
addressing
the
potential
effects
of
atrazine
on
amphibian
development
and
were
being
published
at
about
the
same
time
or
during
the
period
of
time
that
we
were
preparing
that
January
'
03
document.

As
a
consequence,
we
really
didn't
have
time
to
perform
a
rigorous
evaluation
of
these
data
for
inclusion
in
that
January
IRED.

And
consequently
we
agreed
with
NRDC
that
we
would
proceed
with
our
analysis
of
this
issue
of
the
potential
effect
of
atrazine
on
amphibian
development
during
the
time
period
that
we're
all
in
now.

And
we
set
course
on
a
path
to
review
the
available
information
through
February
28
of
2003
and
then
convene
a
panel
as
we're
doing
today
to
discuss
the
information
at
hand
and
to
gain
your
insights
and
comments
on
the
conclusions
we've
reached
to
date
concerning
that
information.

More
specifically,
what
we
wanted
to
set
out
in
terms
of
the
agreement
between
January
and
October,
was
to
take
a
look
at
the
significance
of
the
amphibian
risk
data
and
determine
whether
there
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
21
was
a
need
for
additional
data
to
characterized
more
fully
atrazine's
potential
risks
to
amphibian
species,
and
if
so,
what
data
should
be
developed
to
further
reduce
the
uncertainties
associated
with
this
question.

What
I'd
like
to
do
over
the
next
few
minutes
is
just
provide
a
bit
of
a
roadmap
and
some
of
the
milestones
for
the
rest
of
the
morning's
presentations
by
the
EPA
folks.
And
the
white
paper
actually
sort
of
sets
up
the
problem.
It
helps
formulate
the
problem
before
us.
And
I've
used
the
words
"
formulate
the
problem"
in
the
context
of
the
Agency's
1998
ecological
risk
assessment
guidelines.

And
the
roadmap
that
we're
going
to
use
today
to
summarize
the
highlights
of
the
White
Paper
to
in
fact
use
the
Agency's
guidelines
as
a
framework
for
the
presentations
and
the
logic
train
that
we
went
through
in
interpreting
the
information
that's
currently
available.

I
think
you're
all
aware
of
Agency's
guidelines,
but
I'd
like
to
spend
just
a
few
minutes
touching
upon
what
I
feel
are
some
of
the
important
aspects
of
the
Agency's
guidelines
and
the
role
of
science
in
risk
management
and
regulatory
decision
making.

Obviously,
I
believe,
ecological
risk
assessments
are
the
science
part
of
the
overall
decision
making
process.
And
in
our
risk
assessments,
which
is
the
box
in
that
figure
on
the
screen,
is
where
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
the
science
of
ecological
risk
assessment
occurs.
And
through
the
development
of
an
ecological
risk
assessment,
clarifying
the
uncertainties,
establishing
what
we
know
and
what
we
don't
know
the
assumptions
behind
the
analyses
helps
inform
the
overall
decision­
making
process
which
blends
in,
of
course,
to
many
other
considerations
other
than
the
science.

So
the
key
over
the
next
few
days
is
to
take
a
look
at
the
science
associated
with
this
issue
and
determine
where
our
certainties
are,
where
are
uncertainties
are,
and
how
that
can
be
helpful
in
helping
to
inform
the
process
of
overall
decision
making
with
regard
to
atrazine
and
its
potential
effects
on
amphibians.
So
the
focus
is
on
the
science.

Over
the
course
of
the
next
several
presentations,
or
my
presentation,
Tom's
and
Joe's,
we're
going
to
focus
on
especially
the
problem
formulation
phase
of
an
ecological
risk
assessment
because,

in
fact,
that's
sort
of
where
we
are
right
now.
We're
in
the
stage
of
formulating
the
problem,
starting
to
define
what
we
know,
what
we
don't
know,
what
types
of
risk
hypotheses
can
we
formulate
with
the
existing
information,
and
to
develop
a
pathway
for
moving
forward.

So
we're
in
the
process
of
generating
and
evaluating
preliminary
hypotheses
about
why
effects
may
occur
or
could
be
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
23
occurring
and
to
articulate
the
uncertainties
that
would
be
associated
with
moving
forward
in
a
risk
assessment.
I
think
it's
also
important
to
remember
that
the
guidelines
for
ecological
risk
assessments
acknowledge
an
iterative
process.
And
that
as
the
science
evolves
and
risk
management
issues
evolve,
the
risk
assessments
can
evolve
too.
And
so
in
a
sense,
we're
in
an
iterative
process
of
the
overall
risk
assessment
for
atrazine.

As
I
mentioned
earlier,
in
the
January
document
we
articulated
one
uncertainty
that
was
associated
with
the
potential
effects
of
atrazine
on
the
development
of
amphibians.
And
in
a
sense,
we're
going
through
an
iteration
of
the
risk
assessment
for
atrazine;
and
we're
at
the
stage
of
looking
at
problem
formulation.

So
what
does
that
mean?
What
I've
done
in
the
next
slide
is
expanded
on
the
concepts
of
problem
formulation
and
what
some
of
the
key
aspects
of
this
phase
of
a
risk
assessment
entail.
And
this
will
be
the
focus,
this
is
really
the
focus
of
the
White
Paper
and
it
is
the
focus
of
the
presentations
we
will
be
making
today.

Through
the
problem
formulation,
one
sets
the
stage
for
moving
forward
in
the
overall
risk
assessment.
Outcomes
of
a
problem
formulation
could
include
a
decision
that
there's
no
need
to
go
forward,
that,
in
fact,
there
isn't
an
issue
that
requires
a
risk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
24
assessment.
Another
option,
talking
about
broad
pathways,
one
can
move
forward
after
problem
formulation
would
be
that,
in
fact,
there
is
sufficient
data
to
move
ahead
with
the
risk
assessment
with
specified
levels
of
certainty.
And
another
option
would
be
that,
while
it's
possible
to
form
risk
hypotheses,
there
are
certain
uncertainties
that
are
associated
with
the
risk
assessment;
and
the
analysis
plan,

therefore,
may
call
for
or
suggest
additional
data
that
could
be
needed.

The
key
in
problem
formulation
is
associated
with
box
the
on
my
left
and
the
arrows
that
are
going
both
ways.
And
that's
the
dialogue
between
the
risk
managers
and
the
risk
assessment
team.

The
risk
assessment
is
science,
but
it's
not
science
in
a
vacuum.
It's
science
to
inform
regulatory
decision
assessment
and
risk
management.
And
so
the
decisions
with
regard
to
certainty
and
uncertainty
certainly
have
a
scientific
basis,
but
they
also
have
a
context.
And
the
context
is
in
the
context
of
how
much
certainty
is
required
to
make
a
regulatory
decision
with
a
specified
level
of
confidence.

The
role
of
the
risk
assessor
is
to
provide
the
risk
manager
insights
into
the
uncertainties,
the
certainties,
the
risk
hypotheses,

and
to
engage
in
a
dialogue
as
to
the
pathway
and
moving
forward.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
25
The
White
Paper
is
a
problem
formulation.
The
White
Paper
is
designed
to
help
inform
all
of
us
and
you
and
gain
your
insights
as
well
to
inform
the
broader
community
as
to
the
potential
pathways
that
make
sense
given
the
information
that
we
currently
have.

The
key
products
in
a
problem
formulation
are
the
items
in
the
circles
on
the
figure.
The
clarification
of
the
risk
assessment
endpoints
and
the
measures
of
effects
is
an
important
outcome
of
problem
formulation
as
is
the
conceptual
models
which
is
essentially
as
series
of
one
or
more
risk
hypotheses
as
to
how
one
may
envision
or
hypothesize
that
atrazine,
in
this
case,
could
have
potential
developmental
effects
on
amphibians.

And
then
an
analysis
plan,
given
the
current
body
of
information,
a
plan
in
moving
forward.
How
are
we
going
to
use
the
available
information,
if
additional
information
could
be
gained,
what
kind
of
information
would
contribute
to
closing
what
knowledge
gaps.

And
again,
before
moving
forward
into
actual
risk
characterization
dialogue
with
the
risk
management
community
to
ensure
that
the
regulator
decision
makers
in
the
agencies
have
a
clear
understanding
of
the
uncertainties,
and
the
issues
associated
with
moving
forward
with
different
levels
or
amounts
of
data.
The
risk
assessment
is
designed
to
inform
regulatory
decisions;
it
doesn't
make
regulatory
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
26
decisions.

Let
me
just
spend
a
few
minutes,
then,
with
this
roadmap
as
background,
hit
a
few
of
the
milestones
on
the
journey
we'll
take
over
the
next
couple
of
hours
in
reviewing
the
White
Paper.

As
I
indicated
in
that
previous
slide,
one
of
the
most
critical
steps
actually
in
problem
formulation
is
integrating
the
available
information.
What
do
we
know,
what
don't
we
know,
and
how
do
we
use
that
information
to
draw
conclusions
concerning
risk
assessment
endpoints,
measures
of
effects;
how
do
we
use
that
information
to
help
establish
risk
hypotheses,
and
ultimately
establish
the
analysis
plan
for
moving
forward.

Tom
Steeger
will
be
providing
an
overview
of
the
key
studies
that
were
discussed
in
the
White
Paper
in
some
detail.
And
as
you'll
recall,
there
were
17
studies
available
for
analysis
by
the
Agency
as
of
February
28,
2003.
Seven
of
these
studies
were
laboratory­
based
studies;
and
10
of
the
studies
were
field
experiments.

The
White
Paper
describes
how
we
looked
at
the
study
attributes,
experimental
designs,
the
various
protocols
that
were
used,

and
how
we
looked
across
those
studies
to
take
a
look
at
the
body
of
knowledge,
looking
at
the
consistency
across
the
studies,
how
the
studies
as
a
whole
provide
insights
into
the
strength
of
cause­
effect
or
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
27
dose
response
relationships,
the
extent
to
which
the
body
of
knowledge
provides
insights
on
mechanistic
plausibility
concerning
the
potential
effects,
and
issues
regarding
ecological
relevancy.

And,
of
course,
throughout
the
White
Paper
and
throughout
the
process
of
developing
a
problem
formulation,
one
keeps
track
of
the
certainties
and
uncertainties
as
one
integrates
and
evaluates
the
existing
information.

Tom,
in
his
talk,
will
go
over
the
body
of
the
information
and
sort
of
a
synthesis
mode.
Tom's
not
going
to
go
through
each
individual
study
because
the
White
Paper
provides
that
level
of
analysis.
And
rather
Tom's
going
to
summarize
the
synthesis
and
the
integration
of
the
information.
And
so
as
he
talks
about
study
protocols
and
design,
it's
designed
to
be
reflective
of
the
entire
body
of
studies
and
not
necessarily
a
specific
comment
for
a
specific
study.

During
Thursday
or
Friday,
if
you'd
like
to
talk
in
more
detail
on
specific
studies,
of
course,
we'd
be
happy
to
do
so.

One
of
the
outcomes
of
problem
formulation,
which
I
mentioned
previously,
is
establishment
of
the
risk
assessment
endpoints;
and
they
draw
in
part
from
the
available
information,
but
they'd
also
have
to
be
connected
to
the
Agency,
in
this
case,
EPA's
mission.
What
is
it
all
about
in
terms
of
protecting
the
environment
and
human
health.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
28
And
so
you
have
a
train
of
logic
that
needs
to
connect
environmental
management
goals
to
a
risk
assessment
endpoint
and
then
to
the
measures
of
effects
which
will
be
used
to
estimate
how
those
risk
assessment
endpoints
may
change
based
on
different
exposure
scenarios.

So
in
this
context,
we're
talking
about
environmental
management
goal
which
is
the
viability
of
anuran
populations.
So
the
analysis
of
the
studies
that
Tom
and
Joe
will
talk
about,
we're
focusing
on
a
risk
assessment
endpoint
which
involves
the
reproduction
and
recruitment
of
native
anurans.

Again,
a
risk
assessment
endpoint
needs
to
be
a
ecological
entity
and
the
attributes
of
the
entity.
So
the
entity
is
anuran,
native
anurans
of
North
America,
native
anurans,
and
the
attribute
reproduction
and
recruitment.

Through
the
analysis
of
the
existing
information
that
Tom
will
describe
as
further
highlighted
in
the
discussion
that
Joe
will
provide,

there's
a
whole
family
of
measures
of
effects
that
have
been
reported
in
the
literature
and
it
can
be
useful
in
terms
of
estimating
how
the
risk
assessment
endpoint
may
change
based
on
different
atrazine
exposures.

And
on
the
slide,
I've
listed
many
of
the
measures
of
effects
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
29
that
were
included
in
those
studies;
and
Tom
will
be
talking
about
the
interpretation
of
that
information
in
the
context
of
estimating
the
risk
assessment
endpoint
behavior.

A
second
major
output
of
the
problem
formulation
stage
and
of
the
White
Paper
is
the
conceptual
model.
The
conceptual
model
is
a
way
of
pulling
together
the
lines
of
evidence,
the
information
that's
available,
to
formulate
risk
hypotheses
and
to
provide
insights
into
how
we
need
to
go
forward.
As
I
said,
I'm
going
to
provide
some
of
the
milestones
on
the
path;
and
Tom
and
Joe
will
provide
more
of
the
details
of
how
we
got
to
some
of
the
milestones
that
are
shown
on
this
slide.

We
went
through
the
available
information
and
as
we
discussed
in
the
White
Paper
we
concluded
that
the
lines
of
evidence
did
not
show
a
consistent,
reproducible
effect
of
atrazine
across
the
exposure
concentrations
in
the
amphibian
species
tested.
But
we
also
noted
that
there
were
issues
concerning
the
study
protocols,
the
experimental
designs,
and
inherent
uncertainties
in
the
issue
at
hand
to
make
it
difficult
to
fully
interpret
the
information.
And
consequently
and
as
a
result
of
the
strength
of
the
studies,
we
did
come
to
the
conclusion
that
the
available
data
is
of
sufficient
quality
to
establish
a
risk
hypothesis
that
atrazine
could
cause
developmental
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
30
effects
in
amphibians.

Between
Tom's
talk
and
Joe's
talk,
we'll
review
the
White
Paper
in
the
context
of
how
we
reach
that
risk
hypothesis
and
establish
in
more
detail
the
conceptual
mode
of
atrazine's
mechanism
of
proposed
mechanism
of
action
and
how
that
could
lead
to
developmental
effects
or
reproductive
effects
in
the
context
of
the
risk
assessment
endpoint
that
I
mentioned
previously.

And,
finally,
the
last
major
product
from
a
problem
formulation
and
as
discussed
in
the
White
Paper,
is
the
analysis
plan.
The
analysis
plan,
again,
is
a
series
of
options
to
have
a
dialogue
with
the
risk
managers
in
the
Agency
in
terms
of
where
to
move
forward.
And
here,
again,
I
think
it's
important
to
come
back
to
the
interface
between
risk
assessment
and
risk
management.

In
an
analysis
plan,
options
are
created
based
on
the
scientific
certainties
and
uncertainties
available
in
the
information.
The
decision
as
to
how
much
uncertainty
or
how
much
uncertainty
one
can
make
a
decision
is
part
of
that
interface
between
the
risk
assessor
and
the
risk
manager.
And
so
the
analysis
plan
lays
out
concepts
for
future
studies,
lays
out
a
roadmap
for
future
studies.
But
they're
in
the
context
of
whether
or
not
greater
certainty
is
required
to
make
the
regulator
decision.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
31
There's
some
science
and
then
there's
policy
in
FDM
that
goes
on.
And
the
White
Paper
is
discussing
the
scientific
certainties
and
uncertainties
that
EPA
feels
exists
in
the
current
body
of
information
and
provides
some
thoughts
and
some
concepts
and
plans
as
to
how
those
uncertainties
could
be
closed
if
they
need
to
be
closed
to
make
a
regulatory
decision.

So
the
analysis
plan
then,
based
on
our
risk
hypotheses
and
the
conceptual
model,
are
designed
to
enhance
or
improve
the
clarity
of
potential
causality
in
terms
of
atrazine's
potential
effects
on
amphibians
as
well
as
to
further
characterize
the
potential
dose
response
relationship
between
atrazine
exposure
and
developmental
effects.

In
further
phases
of
the
analysis
plan,
it
talks
about
making
connections
to
mechanisms
of
action
as
well
as
ecological
relevancy.

I
think
that's
another
important
point
to
bring
out
is
that,
as
I
mentioned
earlier,
risk
assessments
can
be
iterative,
they
can
be
phased,
they
can
be
tiered,
and
in
fact,
an
analysis
plan
in
an
ecological
risk
assessment
can
lay
out
a
phased
or
tiered
approach
to
reducing
uncertainties,
incremental
gains
in
knowledge
as
a
basis
of
needs
for
informing
the
risk
managers
in
the
decisions
they
need
to
make.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
32
And
as
you'll
note
in
the
White
Paper
and
as
we'll
summarize
this
morning,
the
analysis
plan
lays
out
a
phased
or
tiered
approach
to
looking
at
specific
uncertainties
in
sort
of
cascade
approach.

Again,
the
decisions
to
move
through
the
phases
in
the
analysis
plan
would
be
tied
to
risk
management
decision
criteria.
But
how
is
the
science
in
terms
­­
what's
the
state
of
the
science
in
informing
the
risk
managers
for
different
decisions
they
may
need
to
make.

So
in
conclusion
from
my
talk,
I
want
to
stress
that
the
White
Paper
reflects
our
conclusions
to
date
based
on
our
analysis
of
the
information
and
our
interpretation
of
this
information
in
the
context
of
the
Agency's
risk
assessment
guidelines.
Now
we're
at
the
important
stage
of
gaining
insights
and
advice
and
counsel
from
our
scientific
peers
in
terms
how
we've
taken
a
look
at
the
data,
how
we've
integrated
the
available
information,
gain
your
insights
and
advice
and
counsel
on
how
we
evaluated
the
studies,
how
we
characterize
the
available
studies,
and
the
conclusions
that
we
drew
from
the
available
body
of
information.

Of
course,
we're
also
looking
forward
to
your
thoughts
and
opinions
in
terms
of
the
risk
assessment
endpoints
and
the
measures
of
effects.
And
then
ultimately
the
conceptual
model
and
the
analysis
plan.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
33
So
with
that,
I'll
close.
And
I'll
be
happy
to
answer
any
clarifying
questions
at
this
point.

DR.
ROBERTS:
Thank
you,
Dr.
Bradbury,
for
laying
out
the
task
in
front
of
us.
Let
me
ask
the
panel
if
they
have
any
questions
or
clarifications
for
you.

I
see
no
questions,
so
thank
you
very
much.

Let's
go
on
then
next
to
Dr.
Steeger's
presentation
on
an
overview
of
the
atrazine
studies.
Good
morning,
Dr.
Steeger.

DR.
STEEGER:
Good
morning.
Thank
you
for
the
opportunity
to
discuss
some
of
the
recent
literature
that
has
become
available
regarding
the
potential
effects
of
atrazine
on
amphibian
development.

As
indicated,
the
Agency
has
developed
a
White
Paper
intended
to
review
recent
studies
conducted
on
the
effects
of
atrazine
on
amphibian
development.
This
presentation
will
provide
an
overview
of
the
study
reviews
and
attempt
to
integrate
the
information
to
answer
whether
there
is
sufficient
information
to
substantiate
claims
that
atrazine
exposure
results
in
development
effects
in
amphibians.

Panel
members
have
had
an
opportunity
to
review
each
of
the
studies
for
themselves.
This
presentation
will
look
at
the
studies
collectively
rather
than
focus
on
individual
studies.

As
early
as
1998
and
continuing
through
this
year,
a
series
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
34
studies
have
been
published
indicating
the
variable
lengths
of
exposure
to
atrazine
is
associated
with
gonadal
effects
in
amphibians
based
on
two
studies
published
by
Tevera­
Mendoza
in
2001,
three
studies
published
by
Hayes
in
2002,
research
described
in
a
poster
presented
by
McCoy,
et
al.,
in
2002
at
the
meeting
of
the
Society
of
Toxicology
and
Enviornmental
Chemistry;
and
a
study
by
Carr,
et
al.,

in
2003;
there
is
sufficient
information
to
formulate
a
plausible
hypotheses
that
atrazine
exposure
may
result
in
development
effects
in
amphibian
gonads
and
that
these
effects
may
impact
secondary
sexual
characteristics
in
these
animals.

However,
also
based
on
these
studies
reported
in
the
open
literature,
there
is
a
lack
of
consistency
in
the
type
of
effect
produced
and
the
concentration
of
atrazine
required
to
produce
that
effect.

As
part
of
a
consent
decree
between
the
Agency
and
the
Natural
Resource
Defense
Council,
the
Agency
agreed
to
conduct
and
review
the
available
literature
regarding
the
effects
of
atrazine
on
amphibian
development.
The
Agency
reviewed
a
total
of
17
studies
that
were
submitted
as
of
February
28,
2003.
As
Steve
indicated,
12
of
the
studies
were
sponsored
by
the
registrant
where
5
were
drawn
from
the
open
literature.

Registrant­
submitted
studies
received
more
scrutiny
since
more
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
35
data
were
available.
Although
none
of
the
studies
were
conducted
under
good
laboratory
practice
conditions,
many
of
the
studies
had
standard
operating
procedures
and
some
level
of
quality
assurance
in
place.
Additionally,
on
studies
where
raw
data
were
available,
the
data
were
re­
subjected
to
statistical
analyses.

Since
most
of
the
published
studies
did
not
have
standard
operating
procedures
nor
were
raw
data
available
for
review
on
most
of
the
studies,
the
open
literature
studies
were
evaluated
at
face
value
with
the
understanding
that
all
these
published
studies
would
have
been
subject
to
some
degree
of
scrutiny
already
through
the
journal's
peer
review
process.

No
formal
guidelines
existed
for
specifically
examining
the
effects
of
atrazine
on
gonadal
development
in
amphibians.
Currently,

there
are
no
guideline
studies
for
amphibians
and
the
Agency
relies
on
other
aquatic
and
terrestrial
test
species
to
serve
as
surrogates
for
estimating
risks
to
amphibians.

Additionally,
many
of
the
measurement
endpoints
examined
in
the
recent
studies
differ
from
those
regularly
utilized
by
the
Agency
to
estimate
acute
and
or
chronic
risk.
However,
the
Agency
is
not
confined
to
using
guideline
studies
to
identify
potential
hazards.
The
Agency
routinely
relies
on
open
literature
to
provide
additional
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
36
insights
on
the
potential
effects
of
pesticides
and
may
use
this
information
to
request
additional
studies
to
address
uncertainties.

The
registrants
Ingenta
voluntarily
undertook
all
of
the
studies
submitted
for
the
Agency
review.
The
studies
were
prompted
by
concerns
that
atrazine
exposure
could
potentially
result
in
developmental
effects
in
amphibians.
Although
over
many
years
the
registrant
has
completed
both
acute
and
chronic
ecological
effect
testing
on
a
range
of
species
in
both
the
laboratory
and
the
field,
we
are
focused
today
on
the
recently
completed
studies
completed
on
amphibians.

As
Steve
indicated,
a
total
of
17
studies
were
submitted
by
the
agreed
upon
February
28,
2003,
deadline.
The
deadline
was
imposed
to
allow
sufficient
time
to
review
the
studies
and
write
a
White
Paper
regarding
the
review
for
submission
to
this
SAP.
Seven
of
the
studies
were
conducted
exclusively
in
the
laboratory,
while
10
of
the
studies
were
conducted
in
the
field.
Field
studies
included
Florida,
Illinois,

Indiana,
Iowa,
Michigan,
Nebraska,
Utah,
Wyoming,
and
South
Africa.

When
studies
are
submitted
to
the
Agency,
data
evaluation
records
are
completed
on
each
of
the
studies.
Typically,
data
evaluation
records
detail
how
and
why
the
study
was
conducted,
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
37
results,
and
what
the
study's
author
concluded
from
the
data.
The
Agency
then
analyzes
the
study's
raw
data
and
attempts
to
draw
its
own
conclusions
from
the
data.
Reviewers
identify
any
inconsistencies
in
study
methods
and
results
and
then
summarize
their
interpretation
of
the
study
results.

As
noted
earlier,
most
of
the
open
literature
did
not
have
sufficient
detail
of
the
complete
in­
depth
data
evaluation
records.

And,
in
fact,
data
evaluation
records
are
not
typically
completed
on
open
literature.
However,
evaluation
records
were
completed
for
the
five
open
literature
studies
to
capture
as
much
of
the
methodology,

data,
and
results
that
were
available
in
the
published
study.

All
data
evaluation
records
completed
by
the
Agency
undergo
secondary
review
to
verify
the
primary
reviewers
interpretation
of
the
study.
For
each
of
the
17
studies
reviewed
in
the
White
Paper,
data
evaluation
records
were
reviewed
by
three
secondary
reviewers.

Copies
of
the
data
evaluation
records
for
the
amphibian
effects
studies
have
been
provided
to
the
panel
members.

Reviewed
were
the
studies'
protocols
and
quality
assurance,
the
strength
of
cause­
effect
relationship,
whether
there
was
a
dose
response,
whether
the
observed
effects
have
a
plausible
mechanism
of
action
that
is
consistent
with
what
is
known
about
the
chemical,
and
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
38
finally,
whether
the
measured
effects
are
ecologically
relevant.

A
range
of
amphibian
species
were
tested
in
the
studies.

Although
the
laboratory
studies
may
have
relied
on
non­
native
species,
each
of
the
field
studies
examined
species
within
their
native
range;
thus
cane
toads
were
studied
in
Florida,
bull
frogs
were
studied
in
Iowa,
northern
leopard
frogs
were
studied
in
Wyoming,
Utah,

Nebraska,
and
Indiana,
green
frogs
were
studied
in
Michigan,
cricket
frogs
were
studied
in
Illinois,
and
the
African
clawed
frogs
were
studied
in
South
Africa.
Although
most
of
the
studies
relied
on
tadpoles,
field
studies
examined
both
larval
and
adult
forms.

Endpoints
measured
in
the
laboratory
and
field
studies
included
time
to
metamorphosis,
growth
in
terms
of
length
and
weight,

presence
of
gonadal
abnormalities,
laryngeal
muscle
area,
sex
rations,

plasma
steroid
concentrations,
and
brain
and
gonad
aromatase
activity
levels.

Gonadal
abnormalities
include
misshapen
gonads,
for
example,

discontinuous
testes
or
multi­
lobe
testes.
However
hermaphroditism
was
also
observed.
For
the
purposes
of
this
presentation,
the
terms
hermaphroditism,
intersex,
and
ovotestes
are
used
interchangeably
to
represent
the
co­
occurrence
of
testicular
and
ovarian
tissue
either
in
the
same
gonad
or
individual.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
39
Effects
on
the
amphibian
laryngeal
dilator
muscle
were
also
described.
Although
a
variety
of
methods
were
used
to
document
this
effect,
generally
the
cross­
sectional
area
through
the
laryngeal
dilator
muscle
was
measured.
Typically,
male
frogs
have
a
larger
dilator
muscle
than
females.

No
effort
was
made
in
his
presentation
to
single
out
a
particular
study.
Rather
the
focus
is
on
issues
that
were
identified
in
the
studies
collectively.
This
is
not
to
say
that
all
the
studies
exhibited
similar
difficulties.
Some
studies
contained
relatively
few
issues,
while
others
may
have
contained
several.
However,
no
study
was
devoid
of
uncertainties
and
or
inconsistencies.

Since
each
of
the
studies
contained
sufficient
uncertainties
and
consistencies
or
inconsistencies
that
rendered
the
data
of
questionable
utility,
data
evaluation
records
focus
primarily
on
methodological
issues
rather
than
on
a
statistical
analysis
of
the
data.

As
mentioned
previously,
there
were
7
laboratory
studies
and
10
field
studies.
Most
of
the
field
studies
had
some
laboratory
analyses.
Collectively,
the
following
issues
were
identified
in
the
laboratory
studies:
Atrazine
contamination
of
the
controls,
poor
water
quality,
poor
growth
and
development
and
or
survival,
high
variability
in
endpoint
measures,
lack
of
reproducibility,
and
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
40
unresponsiveness
to
positive
controls.

Atrazine
contamination
in
the
controls
seemed
to
be
a
recurrent
theme
in
several
of
the
studies.
Measured
concentrations
of
atrazine
exceeded
the
levels
of
detection
by
a
factor
of
two
and
were
at
concentrations
reported
to
cause
effects
in
other
studies.
For
example,
in
several
of
the
studies,
atrazine
concentrations
in
control
tanks
was
higher
than
0.1
micrograms
per
liter.
The
concentration
of
atrazine
reported
by
Hayes
to
cause
developmental
effects
in
frog
testes.

Additionally,
several
studies
suggested
that
animal
feed
used
in
the
studies
may
have
contained
atrazine
residues.
However,

separation
techniques
are
not
sufficiently
developed
to
allow
the
researchers
to
verify
and
or
quantify
the
concentration
of
atrazine
in
the
feed.

Laboratory
studies
ranged
from
testing
a
single
concentration
of
atrazine
to
testing
a
broad
spread
of
concentrations.
Although
gonadal
effects
have
been
observed
between
0.1
and
25
micrograms
per
liter,
most
of
the
studies
did
not
sufficiently
bracket
these
concentrations
to
verify
whether
atrazine
at
these
concentrations
can
result
in
a
consistent
developmental
effect.

Poor
water
quality
was
one
of
the
most
frequent
issues
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
41
surrounding
the
laboratory
studies.
Although
the
Agency
does
not
receive
many
amphibian
studies
and
does
not
have
specific
guidelines
to
conduct
these
studies,
several
sources,
for
example,
the
ASTM,

exist
that
do
provide
guidance
for
conducting
aquatic
toxicity
testing
using
amphibians.
Unfortunately,
high
loading
rates
and
frequent
and
incomplete
exposure
water
changes
resulted
in
diminished
water
quality
as
evidenced
by
high
ammonia
and
nitrite
levels
coupled
with
low
dissolved
oxygen.

As
a
result
of
poor
water
quality,
many
of
the
study
animals
exhibited
poor
growth,
low
developmental
rates,
disease
and
high
mortality
rates
that
contributed
to
the
tests'
inability
to
differentiate
treatment
effects.
In
some
cases,
growth
was
negatively
correlated
with
length
of
time
to
metamorphosis.
Where
Xenopus
laevis
typically
requires
58
days
to
complete
metamorphosis,
in
some
studies
larvae
had
not
undergone
metamorphosis
by
as
late
as
100
days.
High
mortality
rates
confounded
some
of
the
studies;
and
in
some
cases,
required
a
proposed
study
methodologies
be
abandoned.

Several
of
the
studies
elected
to
measure
plasma
testosterone
and
estradiol
concentrations
and
aromatase
activity
in
the
brain
and
gonad.
Variability
in
measured
steroid
concentrations
were
so
high
that
in
some
cases
the
study
was
unable
to
differentiate
males
from
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
42
females.
Aromatase
activity
in
the
gonads
ranged
from
being
below
the
level
of
detection
to
sporadic
peaks
in
activity
rendering
within
group
variability
so
high
that
it
would
be
difficult
to
differentiate
any
treatment
effect.

With
coefficients
of
variation
as
high
as
roughly
500
percent,

study
designs
were
insufficient
to
account
for
this
level
of
variability
and
still
be
able
to
detect
treatment
effects.
In
some
studies,

measurement
endpoints
would
have
had
to
differ
by
roughly
80
percent
before
this
study
would
have
been
able
to
detect
the
difference.

Many
of
the
studies
did
not
run
positive
controls.
However,
on
some
of
the
studies
which
did
utilize
dihydrotestosterone
and
17­
beta
estradiol,
low
percentages
of
animals
responded
to
the
treatment.

This
response
differed
from
other
studies
that
indicated
that
the
treatment
of
frogs
with
steroids
would
markedly
impact
sex
ratios
and
the
rate
of
hermaphroditism.

It
is
uncertain
whether
the
lack
of
responsiveness
to
positive
controls
was
due
to
animals
genuine
insensitivity
to
the
steroid
hormones
or
whether
there
was
insufficient
chemical
present
to
elicit
a
response.

The
Agency
recognizes
that
field
studies
can
be
difficult
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
43
conduct
since
researchers
are
not
able
to
control
environmental
conditions.
Also
the
Agency
recognizes
the
difficulty
in
identifying
sampling
sites
that
can
be
considered
true
replicates
of
one
another.

However,
of
the
field
studies
submitted,
there
tended
to
be
considerable
variability
between
sampling
sites.
Similar
to
some
of
the
laboratory
studies,
atrazine,
both
the
parent
and
its
degradates,

was
present
in
reference
groups.
Additionally,
other
trizine
herbicides
and
chemicals
were
present
but
not
always
well­
characterized.

Where
pesticides
were
characterized,
their
concentrations
were
in
some
case
relatively
high.
And
it
is
unclear
what
impact
they
might
have
had
on
the
outcome
of
the
study.
In
some
studies,
there
were
unusual
environmental
conditions
that
may
have
impacted
the
study.
Unusually
high
rainfall
and
increased
predation
due
to
introduced
species
were
problematic.

Similar
to
laboratory
studies,
variable
hormone
concentrations
in
aromatase
activities
were
problematic.
The
variable
plasma
hormone
levels
may
have
been
a
result
of
collecting
animals
over
a
protracted
period
of
time.
In
one
study,
animals
were
collected
over
roughly
a
six­
month
period
where
study
animals
were
likely
to
be
at
different
stages
of
their
sexual
cycles.
Additionally,
it's
unclear
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
44
whether
housing
Xenopus
laevis
in
close
proximity
to
one
another
following
their
collection
influenced
their
hormonal
concentrations
and
or
aromatase
activity
of
these
opportunistic
breeders.

In
spite
of
all
the
issues
identified
in
the
available
studies,
the
Agency
believes
that
the
laboratory
and
field
studies
have
provided
useful
information.
The
studies
provide
sufficient
information
with
which
to
formulate
a
hypothesis.
They
provide
insight
on
the
potential
sources
of
variability
and
they
provide
insight
on
future
test
species
and
study
conditions.

Although
many
of
the
studies
did
not
demonstrate
any
effect
of
atrazine
on
amphibian
development,
there
are
sufficient
data
to
suggest
that
atrazine
may
be
affecting
gonadal
development.
In
six
of
the
studies,
atrazine
exposure
was
associated
with
a
range
of
gonadal
effects
across
three
species
of
amphibians.
There
are
sufficient
data
to
minimally
formulate
the
hypothesis
that
atrazine
exposure
may
impact
gonadal
development.
However,
there
are
insufficient
data
to
refute
or
confirm
whether
atrazine
is
actually
causing
gonadal
effects.

The
Agency
believes
that
there
are
insufficient
data
to
refute
or
confirm
the
hypotheses
that
atrazine
exposure
may
impact
gonadal
development
because
of
the
collective
uncertainties
associated
with
the
existing
studies.
Uncertainties
include
whether
the
cause­
effect
is
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
45
real
and
can
be
readily
repeated
in
different
laboratories;
what
is
the
dose
response
relationship,
the
mechanistic
plausibility
that
atrazine
exposure
is
causing
a
given
effect;
the
inability
to
readily
extrapolate
laboratory
effects
to
the
field;
and
the
uncertain
ecological
relevancy
of
the
measurement
endpoints.

Without
addressing
these
uncertainties,
the
Agency
has
no
way
to
determine
whether
a
particular
effect
can
consistently
be
expected
to
occur
at
a
particular
level,
whether
the
effect
if
real
can
be
expected
to
occur
in
other
animals,
and
whether
the
effect
is
likely
to
adversely
effect
an
animal's
reproductive
fitness.

While
gonadal
development
appears
to
be
the
primary
effect
associated
with
atrazine
exposure
in
amphibians,
a
consistent
measurement
endpoint
for
the
effect
has
differed.
Atrazine
exposure
has
been
demonstrated
to
result
in
hermaphroditism
in
several
studies
and
laryngeal
effects
in
a
single
study.
However,
other
studies
have
not
been
able
to
demonstrate
similar
effects.
While
males
have
been
primarily
affected,
there
are
conflicting
data
on
whether
females
are
also
impacted.

Obtaining
a
clear
dose
response
relationship
has
been
problematic
for
most
of
the
researchers
engaged
in
studying
the
effects
of
atrazine
on
amphibians.
In
some
studies,
atrazine
exposure
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
46
resulted
in
no
effects;
while
in
others,
concentrations
as
low
as
a
tenth
of
a
microgram
per
liter
resulted
in
hermaphroditism
in
the
laboratory.
Efforts
by
some
researchers
to
substantiate
these
laboratory
results
were
only
successful
at
atrazine
concentrations
250
times
higher.

Additionally,
data
from
some
studies
have
suggested
that
following
a
threshold
effect
concentration,
there
is
either
a
leveling
off
of
a
response
or
a
diminished
response
at
higher
doses.
Therefore,

the
existing
data
have
not
demonstrated
a
traditional
monotonic
dose
response
curve.

Several
of
the
current
studies
have
proposed
that
atrazine
exposure
results
in
up­
regulation
of
aromatase
activity
and
a
subsequent
decline
in
testosterone
concentrations
and
an
increase
in
estrogen
that
in
turn
lead
to
feminizing,
that
is,
hermaphroditism,
and
demasculinizing,
that
is,
decreased
laryngeal
muscle
effects
in
atrazine­
exposed
males.

However,
no
study
thus
far
has
directly
demonstrated
that
aromatase
activity
has
indeed
been
up­
regulated.
And
only
one
study
has
demonstrated
that
plasma
testosterone
has
decreased
in
atrazine­
treated
males.

Although
many
of
the
studies
thus
far
have
examined
plasma
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
47
steroid
levels
and
brain
aromatase
activity
levels,
it
is
uncertain
whether
the
proposed
mechanism
of
action
is
likely
to
be
observed
on
the
basis
of
the
whole
animal.
Rather
aromatase
activity
is
proposed
to
increase
in
the
testes
where
androgenous
testosterone
is
converted
to
estrogen.
It
is
unclear
whether
these
localized
increases
in
steroid
conversions
could
be
detected
in
blood
plasma
at
all.

Out
of
the
17
studies,
one
demonstrated
gonadal
effects
in
both
the
laboratory
and
the
field.
However,
in
this
single
study,
there
was
a
clear
lack
of
a
dose
response.
Coupled
with
the
variable
effects
that
have
been
noted,
even
within
the
same
species,
extrapolating
atrazine
to
potential
field
effects
is
difficult.

While
intuitively
it
may
seem
that
the
presence
of
ovotestes
and
reduced
numbers
of
spermatogonial
cell
mass
in
males
and
reduced
numbers
of
primary
and
secondary
oogonia
in
females,
may
impair
the
reproductive
fitness
of
frogs
and
that
reduced
laryngeal
muscle
mass
and
secondary
sexual
characteristics
may
impair
an
animals
ability
to
attract
mates.
There
are
not
data
currently
available
to
the
Agency
with
which
to
gage
impaired
reproductive
function,
recruitment,
or
survival.

Additionally,
the
current
ecological
risk
assessment
of
atrazine
identifies
that
some
plants
have
exhibited
resistance
to
atrazine.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
48
Some
researchers
have
speculated
that
amphibians
may
also
develop
resistance
to
the
potential
effects
of
atrazine
on
amphibian
development.
The
Agency
is
uncertain
regarding
the
role
of
resistance
and
recovery
from
the
potential
developmental
effects.

The
primary
criteria
for
conducting
ecological
risk
characterizations
in
the
Agency
are
that
they
be
transparent,
clear,

consistent,
and
reasonable.
Of
these
criteria,
transparency
is
viewed
as
the
principal
value
from
among
the
four
since
it
leads
to
clarity,

consistency,
and
reasonability.

Consistent
with
the
EPA's
process
for
conducting
ecological
risk
assessments,
it
has
evaluated
the
available
data
following
specific
evaluation
criteria.
That
included
experimental
design,
the
strength
of
the
cause­
effect
relationship,
the
dose
response
relationship,
the
mechanistic
plausibility,
and
the
ecological
relevancy.
The
Agency
has
provided
these
reviews
to
panel
members.

Based
on
its
review
of
the
available
literature,
the
Agency
believes
that
there
is
sufficient
information
to
formalize
a
hypothesis
regarding
the
potential
effects
of
atrazine
on
amphibian
development.

But
because
of
the
uncertainties
surrounding
each
of
the
studies
conducted
thus
far,
the
Agency
is
recommending
that
additional
studies
be
conducted.
The
next
presentation
by
Joe
Tietge
will
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
49
discuss
the
Agency's
recommendations
for
additional
testing.

In
conclusion,
as
of
February
28,
2003,
the
Agency
has
reviewed
a
total
of
17
studies
examining
the
effects
of
atrazine
on
amphibian
gonadal
development.
These
studies
have
involved
both
laboratory
and
field
work
and
have
looked
at
six
species
of
anurans.

In
each
of
the
studies,
the
Agency
has
identified
concerns
regarding
the
study
methodologies
and
or
results
that
potentially
limit
the
utility
of
the
studies.

Based
on
all
17
studies,
atrazine
exposure
did
not
produce
consistent,
reproducible
effects
across
all
species
tested;
therefore,

the
weight
of
evidence
suggests
that
atrazine
exposure
does
not
impact
gonadal
development.
However,
there
are
lines
of
evidence
from
both
laboratory
and
field
studies
that
support
the
formulation
of
a
plausible
hypothesis
that
atrazine
exposure
may
result
in
developmental
effects
in
amphibians.

Although
the
current
studies
cannot
be
used
to
refute
or
confirm
the
hypothesis
that
atrazine
exposure
may
result
in
gonadal
development
effects,
the
studies
do
provide
useful
information
of
the
sources
of
variability.
This
information
will
be
critical
to
the
design
of
future
studies.

There
are
insufficient
data,
as
I
indicated,
to
refute
or
confirm
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
50
the
effects
of
atrazine
on
amphibians.
If
risk
managers,
however,

wish
to
reduce
the
current
uncertainties
regarding
the
potential
effects
of
atrazine
on
amphibians,
the
Agency
recommends
that
additional
studies
be
initiated.
These
studies
should
build
on
the
current
body
of
information.

If
additional
testing
is
required,
the
Agency
is
proposing
that
a
phased
approach
be
used
to
examine
the
cause­
effect
dose
response
mechanistic
plausibility,
and
the
ecological
relevancy
of
any
effects
observed
following
the
exposure
of
amphibians
to
atrazine.
Joe
Tietge
who
will
follow
me
will
present
what
the
Agency
is
proposing
as
follow­
up
studies.

Are
there
any
questions?

DR.
ROBERTS:
Great.
Thank
you,
Dr.
Steeger.
Before
we
move
on
to
the
next
presentation,
I
think
this
is
good
opportunity
for
the
panel
to
ask
you
any
questions
they
might
have
on
the
Agency
review
of
the
17
studies.
Are
there
any
questions
among
panel
members
regarding
the
Agency's
review.

DR.
KELLEY:
I
have
a
question.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
So
for
instance
you
said
that
in
some
of
the
studies
you
failed
to
document
sex
differences
in
steroid
levels.
How
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
51
did
you
know
that
you
would
expect
to
see
those
sex
differences?

DR.
STEEGER:
We
didn't.

DR.
KELLEY:
So
that
is,
in
fact,
then
not
such
a
useful
criteria.

DR.
STEEGER:
We
­­

DR.
KELLEY:
If
you
didn't
know
you
expected
to
see
them
and
then
you
didn't
see
them,
how
did
you
know
they
were
there
anyway?

DR.
STEEGER:
You're
talking
about
sex
differences
in
the
steroid
level
hormone
concentrations.

DR.
KELLEY:
In
steroid
levels.
Right.

DR.
STEEGER:
Well,
we
didn't
know
what
to
expect.
Because
as
I
indicated,
this
was
a
new
area
for
the
­­
these
measurement
endpoints
were
new
for
the
Agency
to
consider.

DR.
KELLEY:
All
right.

DR.
ROBERTS:
Yes,
Dr.
Skelly.

DR.
SKELLY:
I
just
had
a
general
question
about
whether
in
your
review
you
considered
any
sort
of
a
line
below
which
the
quality
of
data
issues
meant
that
you
wouldn't
consider
the
evidence
from
that
study.

DR.
STEEGER:
When
the
Agency
receives
guideline
studies,

we
have
what's
called
the
"
rejection
rate
analysis"
where
there
are
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
52
certain
characteristics
of
the
study
which
will
eliminate
it
from
consideration.
Some
of
the
studies,
one
potential
factor
that
can
eliminate
a
study
for
consideration
is
the
presence
of
the
test
chemical
in
control
sites.
Because
these
studies
weren't
conducted
following
guidelines,
we
didn't
really
have
a
criteria
that
would
really
eliminate
it
from
consideration.
But
that
would
constitute
a
reason
to
reject
a
study.
Does
that...

DR.
SKELLY:
So
you
considered
rejecting
studies
and
decided
not
to.

DR.
STEEGER:
We
considered
that
we
would
just
review
the
studies
as
they
existed
without
any
consideration
for
what
would
constitute
a
fatal
flaw
in
the
study.

DR.
KELLEY:
Can
I
ask
then
­­

DR.
ROBERTS:
Excuse
me.

DR.
KELLEY:
­­
just
follow
up
­­

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
Yeah.

DR.
ROBERTS:
You've
got
to
wait
for
me
to
call
on
you.
Dr.

Kelley.

DR.
KELLEY:
Thank
you.
Just
to
follow
up
on
that
question.

So
do
you
use
a
criterion
that
might
take
into
account
the
weight
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
53
the
evidence,
in
that
case
just
simply
the
number
of
studies?
So
if
you
had
22
studies
that
got
an
effect
and
5
that
didn't
you
would
consider
not
having
rejected
any
of
them
as
weighing
in
on
the
weight
of
evidence
in
favor
of
the
first
result
rather
than
the
second?

DR.
STEEGER:
In
this
case,
we
are
not
using
a
­­
because
of
the
inconsistencies
in
all
the
studies,
we
couldn't
really
use
a
weight
of
evidence
approach.
As
I
indicated
throughout
my
presentation,

what
seems
to
be
a
recurrent
theme,
a
line
of
evidence,
that
there
seemed
to
be
some
effects
that
recur
over
the
studies.
But
the
weight
of
evidence
approach
does
not
work
for
us
in
this
case
because
there
were
such,
in
our
view,
glaring
problems
with
each
of
the
studies
that
it
was
not
difficult
to
weight
them
per
se.

DR.
KELLEY:
One
last
question.
You
had,
whatever
it
was,
19
studies
­­
is
that
right?
­­
12
and
7.

DR.
STEEGER:
Seventeen
studies.

DR.
KELLEY:
Seventeen
studies.
After
you
completed
the
White
Paper,
did
you
become
aware
of
any
studies
that
were
not
included
in
the
White
Paper
that
you
had
missed
for
one
reason
or
another?

DR.
STEEGER:
Our
contractor
provided
us
with
­­
are
we
talking
about
amphibian
studies?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
54
DR.
KELLEY:
Amphibian
studies.

DR.
STEEGER:
Not
that
we
were
aware
of.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
I'm
paraphrasing
here.
But
you
said
something
to
the
effect
that
the
hypothesis
wasn't
accepted
because
no
consistent
reproducible
effects
across
all
species
were
observed.
And
I
was
wondering
if
that
was
actually
a
requirement
for
accepting
the
hypothesis,
that
consistency
among
all
species
be
observed.

DR.
STEEGER:
No,
it's
not.

DR.
LEBLANC:
And
a
second
question
is
were
other
species,

vertebrate
species,
aquatic
vertebrate
species
considered
in
the
literature
review?

DR.
STEEGER:
No,
they
were
not.
We
do
have
information
on
other
species.
But
our
review
looked
at
amphibians
only.

DR.
ROBERTS:
Dr.
Gibbs.

DR.
GIBBS:
Just
quickly,
in
your
characterization
of
available
studies,
you
mentioned
that
data
evaluation
records
focused
primarily
on
methodological
problems
rather
than
statistical
analyses.
Could
you
elaborate
on
that?

DR.
STEEGER:
Because
of
the
problems
with
atrazine
contamination
in
the
controls,
the
unresponsiveness
of
animals
to
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
55
positive
controls,
the
slow
development
of
the
animals,
the
high
mortality
rates
that
were
exhibited
on
many
of
the
studies,
those
were
what
we
considered
to
be
problematic
that
would
have
rendered
the
data
of
questionable
utility.
So
we
didn't
really
focus
on
analyzing
the
data
per
se.

Now
I
did
mention,
though,
that
on
the
aromatase
and
plasma
steroid
concentrations,
there
was
high
variability.
We
do
analyses
to
verify
that
there
was
indeed
high
levels
of
variability
in
the
measurement
endpoints.
But,
again,
because
of
the
way
that
the
information
was
collected,
it
would
have
been
problematic
for
us
to
move
forward
with
the
study
independent
of
what
the
analyses
told
us.

DR.
GIBBS:
Was
the
assumption
that
that
peer­
review
process
would
have
caught
any
problems
or
issues
with
the
statistical
analyses
as
reported?

DR.
STEEGER:
Well,
the
peer
review
process,
are
we
talking
about
for
open
literature?

DR.
GIBBS:
Yeah.

DR.
STEEGER:
For
the
open
literature,
it's
rare
for
journals
to
have
access
to
the
author's
raw
data.
So
it's
unlikely
they
would
have
caught
that.

DR.
ROBERTS:
Okay.
Any
other
questions?
Okay.
I
see
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
56
none.
Then
we
have
scheduled
a
break
now.
We're
a
little
ahead
of
schedule.
I
guess
whether
or
not
we
go
to
break
kind
of
depends
on
the
length
of
the
next
presentation.
So
let
me
just
ask
you,
Mr.

Tietge,
we
have
allotted
an
hour
for
that
talk.
What's
your
best
guess?

MR.
TIETGE:
I
think
it
will
take
about
30
minutes.

DR.
ROBERTS:
Then
let's
go
ahead
and
move
on
to
your
presentation.

MR.
TIETGE:
Thank
you.
I'm
just
going
to
get
started
here.

The
basis
for
this
talk
today
is
that
the
evaluation
of
the
currently
available
data
as
previously
reviewed
by
Dr.
Steeger
suggests
that
anuran
reproductive
fitness
may
be
adversely
affected
by
exposure
to
atrazine.
However,
the
data
are
insufficient
to
conclude
that
atrazine
adversely
affects
anuran
reproduction.

Therefore,
further
studies
are
proposed
following
the
Guidelines
for
Ecological
Risk
Assessment
to
reduce
the
uncertainties
and
permit
an
eventual
risk
characterization
if
warranted.

These
conclusions
are
based
on
the
fact
that
there
are
a
number
of
remaining
uncertainties
including
the
following:
The
number
of
affirmative
studies,
that
is,
those
that
seem
to
demonstrate
an
effect
on
gonadal
development
and
secondary
sexual
characteristics,
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
57
sample;
there's
limited
evidence
of
repeatability
between
laboratories;
the
dose­
response
relationship
remains
undefined
due
to
the
lack
of
sufficient
dose­
response
data;
the
mechanistic
plausibility
of
the
hypothetical
mode
of
action
is
currently
unsupported
by
the
available
data
on
amphibians;
and,
finally,
the
ecological
relevancy
of
the
potential
effects
of
atrazine
exposure
on
amphibians
remains
undetermined.

Based
on
these
observations,
it
is
EPA's
recommendation
that,

if
the
risk
management
process
requires
further
reductions
in
these
uncertainties,
then
additional
laboratory
studies
need
to
be
conducted
before
any
additional
risk
assessment
activities
regarding
the
effects
of
atrazine
on
amphibian
reproduction
are
undertaken.

The
objectives
of
this
presentation
are,
first,
to
review
the
concept
of
problem
formulation
as
used
in
the
Agency's
ecological
risk
assessment
process;
second,
to
restate
the
environmental
goals
and
assessment
endpoints
necessary
to
make
the
risk
management
decisions;
third,
propose
a
conceptual
model
for
atrazine
action
on
anuran
reproduction
by
defining
a
risk
hypothesis;
fourth,
propose
an
analysis
plan
which
identifies
measures
of
effect
relevant
to
the
assessment
endpoints
and
risk
hypothesis
and
includes
a
phased­
study
approach
to
test
central
components
of
the
risk
hypothesis;
fifth,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
58
identify
critical
decision
points
in
the
phased
approach;
and,
finally,

to
provide
some
conclusions.

It's
not
my
intent
here
to
give
a
detailed
discussion
of
the
ecological
risk
assessment
paradigm,
but
I
want
to
remind
you
or
familiarize
you
with
the
basic
components
of
an
ERA.
Because
as
already
mentioned
by
Dr.
Bradbury
this
morning,
we
are
using
the
ERA
paradigm
to
guide
our
approach
on
this
issue.

Briefly,
this
process
can
be
represented
as
three
distinct
phases:
Problem
formulation,
analysis,
and
risk
characterization.

Problem
formulation
is
the
foundation
of
the
ecological
risk
assessment
process
as
it
lays
out
the
goals
and
approaches
necessary
for
the
successful
completion
of
an
assessment.
Much
of
what
I'm
presenting
to
you
today
is
indeed
part
of
the
problem
formulation
phase.

The
analysis
phase
is
the
phase
that
implements
the
approach
developed
in
the
problem
formulation
and
generates
the
data
required
to
complete
the
final
phase,
which
is
risk
characterization.
I'll
only
touch
on
the
analysis
phase
today
as
it
relates
to
the
approach
developed
and
problem
formulation.
I
will
not
risk
characterization
except
to
say
that
this
is
the
phase
that
takes
into
account
the
probabilities
associated
with
exposures
and
effects
and
results
in
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
59
some
conclusion
regarding
risk.

In
this
slide,
I
have
expanded
the
problem
formulation
box
here
in
order
to
demonstrate
that
there
are
four
main
components
to
consider.
The
first
is
the
integration
of
the
available
data
as
presented
previously
by
Dr.
Steeger.
This
is
a
critical
step
because
it
identifies
the
potential
problems
associated
with
a
particular
stressor,

in
this
case
atrazine,
and
serves
to
refine
and
focus
the
risk
assessment
questions
and,
therefore,
all
subsequent
activities.

The
remaining
three
components
of
problem
formulation:

selecting
the
assessment
endpoints,
developing
a
conceptual,
and
developing
an
analysis
plan,
is
the
focus
of
this
talk
today.
Before
I
launch
into
these
areas,
I'd
like
to
point
out
the
critical
connection
of
problem
formulation
to
analysis.

In
this
slide,
I've
expanded
the
analysis
phase
to
show
its
components
as
well.
In
general,
in
the
case
of
a
chemical,
this
phase
typically
evaluates
exposure
as
depicted
in
the
left
side
of
the
large
gray
box
and
effects
as
shown
on
the
right.
What
I'd
like
emphasize
here
is
that
the
analysis
plan
that
I
will
present
today,
which
is
part
of
the
problem
formulation
phase,
provides
guidance
for
how
to
conduct
the
studies
on
effects
and
is
thereby
a
prerequisite
to
measuring
effects.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
60
I'd
like
to
return
to
the
problem
formulation
phase
and
discuss
it
in
terms
of
atrazine.
First,
the
overall
environmental
goal
is
to
ensure
anuran
populations
are
viable
and
self­
sustaining.
This
goal
is
rather
generic.
But
it
serves
to
orient
the
assessment
process
by
focusing
on
a
specific
objective.
And
although
it
is
somewhat
simplistic,
it
can
be
difficult
to
actually
assess
the
impacts
of
atrazine,
or
any
other
chemical
for
that
matter,
on
anuran
populations
directly
in
the
field.

But
the
assessment
endpoint,
successful
reproduction
and
recruitment
of
native
anurans,
is
directly
related
to
population
status.

And
as
indicated
by
some
of
the
atrazine
studies
evaluated
earlier
today,
some
aspects
of
reproduction,
including
gonadal
development,

are
measurable
endpoints
relevant
to
the
concern
surrounding
atrazine
exposure.

Knowing
the
environmental
goal
and
the
assessment
endpoints,

and
given
the
fact
that
some
studies
indicate
that
reproductive
system
development
and
secondary
sexual
characteristics
may
be
affected
by
atrazine,
we
need
to
construct
a
conceptual
model.
The
conceptual
model
in
the
form
of
a
risk
hypothesis
is
an
attempt
to
develop
a
model
that
uses
existing
information
to
form
a
plausible
explanation
of
the
potential
effects
of
atrazine
on
the
assessment
endpoints.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
61
That
model
is
depicted
diagrammatically
in
this
slide.
The
hypothesized
effects
of
atrazine
are
presumed
to
be
initiated
by
a
molecular
interaction.
This
interaction
results
in
increased
aromatase
activity,
the
enzyme
responsible
for
the
conversion
of
testosterone
to
estradiol.
The
increased
activity
of
aromatase
results
in
an
evaluation
of
endogenous
estradiol
which
affects
feminization,
for
example,
in
the
male
gonad.

If
the
effects
in
the
male
gonad
are
severe
enough,
then
reduction
is
fertility
and
reproductive
success
could
be
realized.

Which
leads
to
a
hypothetical
reductions
in
recruitment
thereby
impairing
population
maintenance
which
is
in
fact
the
assessment
endpoint.

This
risk
hypothesis,
which
is
based
on
the
information
on
the
literature
and
from
submitted
studies,
may
or
may
not
be
correct.
But
it
forms
the
basis
of
the
proposed
studies
and
can
also
be
thought
of
as
a
working
hypothesis.
Because
of
the
uncertainties
associated
with
the
risk
hypothesis
are
relatively
high,
it
is
likely
that
it
will
be
modified
when
data
become
available.

As
with
any
hypothesis,
some
elements
are
easier
and
or
are
more
important
to
test
than
others.
So
the
questions
is:
At
what
point
in
the
risk
hypothesis
should
hypothesis
testing
be
introduced
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
62
evaluate
the
specific
sub­
questions?
Or
alternatively,
what
is
the
best
strategy
to
evaluate
the
train
of
events
in
the
risk
hypothesis
to
test
its
validity.

In
the
case
of
this
specific
risk
hypothesis,
it
is
our
view
that
the
most
appropriate
entry
point
is
at
the
level
of
determining
the
effects
of
atrazine
on
gonadal
development,
the
apical
organismal
level
endpoint.
The
reasons
for
this
are,
first
of
all,
this
is
the
endpoint
on
which
much
of
the
concern
hinges.
But,
secondly,
readily
available
methods
exist
to
test
the
sub­
hypothesis
with
relatively
inexpensive
methods
that
permit
the
analysis
of
large
sample
sizes.

And
perhaps
most
importantly,
this
endpoint
is
the
linchpin
in
the
entire
train
of
events.
That
is,
if
atrazine
is
found
to
affect
gonadal
development
with
a
greater
degree
of
certainly
than
currently
exists,
then
this
result
provides
strong
rationale
to
conduct
studies
on
the
proceeding
and
subsequent
elements
of
the
risk
hypothesis.

If
on
the
other
hand,
atrazine
does
not
affect
gonadal
development
following
a
systematic
effort
to
study
this
potential
phenomenon,
then
the
logic
train
of
this
risk
hypothesis
is
broken
and
there
may
be
no
impetus
to
follow
up
by
testing
the
upstream
and
downstream
elements.

So
now
that
we
have
an
assessment
endpoint
selected
and
a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
63
conceptual
model
in
place,
we
can
now
develop
an
analysis
plan.

There
are
four
major
elements
to
the
analysis
plan:
A
strategy
to
evaluate
the
risk
hypothesis
which
I've
already
touched
on
in
the
last
slide;
selection
of
endpoints
to
evaluate
also
referred
to
as
measures
of
effects;
determinization
of
appropriate
methods;
and
a
sequence
of
analysis
that
follows
the
most
efficient
path
to
accept
or
refute
the
risk
hypothesis
in
a
systematic
and
organized
manner.

I've
excerpted
the
first
section
of
the
risk
hypothesis
in
the
first
panel
that
I
just
put
on
the
screen.
In
the
proposed
analysis
of
a
risk
hypothesis
as
shown
as
hypothesis
testing
which
is
in
the
second
panel.
As
I
mentioned
earlier,
the
entry
point
for
testing
the
risk
hypothesis
proposed
to
be
the
effects
on
the
gonads
at
the
organismal
level.
So
beginning
at
the
organismal
level,
the
effects
of
atrazine
on
gonadal
development,
particularly
in
the
males,
is
the
primary
endpoint.

Developing
data
at
this
level
is
critical
in
that
it
may
provide
the
rationale
and
justification
for
conducting
relevance
and
or
mechanistic
studies.
If
these
organismal
level
tests
are
affirmative,

then
measurements
of
sex
steroids
should
be
conducted.
And
if
estrogen
levels
are
shown
to
be
elevated
in
the
atrazine
treatments,

then
measurements
of
aromatase
activity
could
be
indicated
as
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
64
previous
studies
have
attempted.
If
positive,
the
data
from
these
studies
will
be
useful
to
establish
a
mechanistic
basis
for
inter­
species
extrapolation,
further
develop
the
plausibility
of
the
mechanisms
involved,
and
develop
appropriate
biomarkers
that
could
be
used
in
future
field
studies.

Although
sex
steroid
and
aromatase
measurements
are
necessary
to
test
the
mechanistic
aspects
of
the
risk
hypothesis,
they
do
not
provide
meaningful
information
on
the
ecological
relevancy
of
a
potential
gonadal
effect.
Therefore,
if
gonadal
effects
are
observed
at
the
organismal
level,
it
is
possible
to
proceed
directly
to
studies
which
evaluate
fertility
endpoints
that
are
relevant
to
the
maintenance
of
populations.
Furthermore,
if
the
working
hypothesis
is
supported
by
organismal
and
suborganismal
studies,
then
it
may
be
possible
to
confirm
the
mode
of
action
by
conducting
confirmatory
studies
which
utilize
no
aromatase
inhibitors.
Rescue
of
normal
morphology
of
the
male
gonad
by
an
aromatase
inhibitor
co­
administered
with
atrazine
would
provide
substantial
support
of
the
risk
hypothesis
in
general
and
more
specifically
the
mode
of
action
involved.

However,
if
any
of
the
studies
conducted
as
part
of
the
hypothesis­
testing
phase
are
negative,
then
alternatives
should
be
considered.
If
no
consistent
and
reproducible
effects
are
observed
at
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
65
the
organismal
level,
then
there
may
be
no
need
to
continue
any
further
testing.
If
on
the
other
hand,
the
organismal­
level
tests
are
affirmative
and
either
the
steroid
or
aromatase
studies
are
negative,

then
an
alternative
hypothesis
may
need
to
be
evaluated.

Since
this
is
purely
hypothetical
at
this
time,
and
it
is
outside
the
scope
of
the
current
risk
hypothesis,
no
further
discussion
of
the
alternative
testing
will
be
presented.

So
now
there's
an
analysis
plan.
I'd
like
to
discuss
some
of
the
details
of
the
proposed
studies.
These
are
labeled
as
phases
here
as
they
were
also
labeled
in
the
White
Paper.

Phase
1,
the
Test
for
Apical
Gonadal
Effects.
The
first
and
most
important
phase
of
hypothesis
testing
in
this
phase
is
to
determine
if
atrazine
exposure
results
in
consistent
and
reproducible
gonadal
effects
in
males
and
females
and
determine
the
shape
of
the
dose­
response
curve,
if
any.

This
slide
lays
out
the
key
experimental
process
for
consideration
in
the
Phase
1
studies.
The
primary
species
recommended
for
this
work
is
Xenopus
laevis.
The
species
is
recommended
because
it
is
amenable
to
laboratory
testing
and
has
been
shown
by
four
studies
to
be
potentially
responsive
to
the
effects
of
atrazine
on
gonadal
development
and
differentiation.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
66
Xenopus
laevis,
however,
is
not
a
native
anuran
at
least
from
a
North
American
perspective.
It
may
or
may
not
be
representative
of
native
anurans
when
it
comes
to
this
issue.
Therefore,
a
secondary
species
is
suggested
such
as
Rana
pipiens,
the
northern
leopard
frog,

which
can
be
used
in
corroborative
studies.

There
is
one
study
that
suggests
that
Rana
pipiens
is
sensitive
to
the
effects
of
atrazine
on
gonadal
development.
However,
this
species
is
more
difficult
to
work
with
in
the
laboratory
and
most
labs
do
not
have
culture
methods
that
permit
continuous
breeding
and
are,

therefore,
unable
to
conduct
studies
throughout
the
year
with
Rana
pipiens
as
can
be
done
with
Xenopus
laevis.
Despite
the
limitations
associated
with
Rana
pipiens
culture,
comparative
studies
may
be
useful
to
develop
data
to
determine
the
ecological
relevancy
of
potential
atrazine
effects
on
a
native
species.

The
developmental
stages
used
in
these
studies
need
to
include
those
that
are
sensitive
to
the
effects
of
estrogens.
In
a
study
conducted
by
Villalando,
1990,
with
Xenopus,
exposure
to
an
estrogen
elicited
effects
on
gonadal
differentiation
during
the
pre­

metamorphic
period.
However,
after
entering
metamorphosis,
the
gonads
were
less
sensitive
to
estrogen
exposure.
The
proposed
studies
should
include
the
pre­
metamorphic
period
and
continue
until
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
67
the
organism
completes
metamorphosis
at
which
time
they
should
be
evaluated.

The
test
condition
in
the
studies
reviewed
earlier
all
use
static­
renewal
conditions.
These
methods
did
not
conform
to
the
generally
accepted
biological
loading
regs
recommended
by
ASTM
and
resulted
in
delayed
development
and
in
some
cases
excessive
mortality.
It
is
likely
that
these
problems
resulted
form
static­
renewal
conditions
themselves
which
resulted
in
the
accumulation
of
nitrogenous
wastes
and
other
metabolic
products
and
generally
poor
water
quality.
It
is
our
recommendation
that
flow­
through
conditions
be
used
that
adhere
to
ASTM
standards
and
thereby
promote
survival,
growth,
and
development.

The
concentrations
of
atrazine
to
be
used
in
these
proposed
studies
should
bracket
those
found
to
be
effective
in
perturbing
gonadal
differentiation
in
previous
studies,
that
is,
at
or
below
.1
micrograms
per
liter
for
the
low
and
at
or
above
25
micrograms
per
liter
for
the
high.
And,
of
course,
these
concentrations
need
to
be
verified
analytically.

The
use
of
estradiol
as
a
positive
control
is
recommended
since
the
potential
effects
on
the
gonad
are
proposed
to
be
mediated
through
this
pathway.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
68
The
sample
size
and
replication
are
not
detailed
here,
but
they
should
be
determined
a
priori
to
be
sufficient
to
test
the
stated
hypothesis
using
appropriate
statistical
assumptions.
Sampling
should
include
all
organisms
on
test
to
avoid
potential
biases.

And,
finally,
the
endpoints
should
include
survival,
growth,

development,
gross
gonadal
morphology,
gonadal
histopathology.

From
these
data,
male­
to­
female
sex
ratios
can
be
derived
in
the
shape
of
the
dose­
response
curve
determined
for
each
endpoint.

Because
there
are
issues
with
the
existing
studies
that
limit
the
usefulness
of
the
data,
we
propose
that
quality
indicators
be
established
as
a
guide
to
evaluate
validity
of
the
proposed
studies.

First
and
foremost,
the
tests
need
to
be
conducted
in
accordance
with
ASTM
standards
for
biological
loading
and
basic
water
quality
parameters
of
pH,
ammonia,
dissolved
oxygen
and
need
to
be
contained
within
acceptable
limits
and
verified
regularly
throughout
the
conduct
of
the
study.

With
regard
to
the
biological
endpoints,
while
there
is
no
bright
line
between
acceptable
and
unacceptable
survival
percentages,

survival
of
90
percent
or
more
is
indicative
of
a
quality
study.
This
is
a
reasonable
standard
to
adopt
particularly
for
Xenopus
laevis
studies
as
the
species
is
particularly
hardy
in
the
laboratory.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
69
Similarly,
growth
of
Xenopus
laevis
should
result
in
organisms
of
about
one­
and­
a­
half
grams;
and
this
will
vary
between
laboratories.
Metamorphic
development
should
be
completed
within
10
weeks.

But
there
are
no
standardized
methods.
And
some
methods
that
are
proposed
here
are
the
acceptance
­­
there
are
no
acceptance
criteria
because
of
the
lack
of
standardized
methods.
So
these
issues
should
be
evaluated
in
aggregate
using
some
professional
judgement.

Measurements
of
sex
steroids.
I'll
now
discuss
the
remaining
phases
of
the
analysis
plan
very
briefly.
Since
the
conduct
of
each
of
these
phases
is
dependent
on
the
outcome
of
the
previous
phases,
it
is
premature
to
discuss
them
in
much
detail.
I
will,
however,
lay
out
the
objective
and
potential
approaches
for
each
phase,
recognizing
that
these
may
change
as
more
information
becomes
available.

The
second
phase
of
the
study
should
be
conducted
if
the
Phase
1
studies
are
positive.
The
aim
of
the
Phase
2
studies
is
to
determine
if
concentrations
of
estradiol
and
testosterone
are
altered
by
exposure
to
atrazine.

The
approach
to
this
phase
is
based
on
the
fact
that
the
developmental
sensitivity
toward
the
feminizing
effects
of
estrogen
in
male
Xenopus
laevis
has
been
experimentally
determined
as
depicted
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
70
in
the
panel
in
the
right.
In
this
panel,
developmental
stages
indicated
is
indicated
on
the
X
axis.
The
developmental
period
between
stage
44
and
50
represent
a
period
in
which
estrogen
is
capable
of
completely
overriding
testicular
differentiation
resulting
in
100
percent
of
the
test
population
with
ovaries.

As
natural
gonadal
differentiation
proceeds,
their
sensitivity
to
this
effect
diminishes
as
is
indicated
by
a
period
of
incomplete
feminization
from
androgenous
estrogen
in
a
period
of
apparent
insensitivity
coincident
with
the
onset
of
metamorphosis.

This
suggests
that
if
ovotestes
formation
in
the
male
is
dependent
on
estrogen,
that
the
elevated
estrogen
levels
need
to
be
present
during
the
sensitive
developmental
stages.
Any
associated
studies
of
this
phenomenon
should
focus
on
the
effects
of
atrazine
on
sex
steroids
during
these
sensitive
periods.

However,
it
remains
uncertain
as
to
whether
more
developmentally
advanced
organisms
are
sensitive
to
the
feminizing
effects
of
estrogen
on
the
gonad.
This
is
an
area
of
uncertainty
that
requires
more
investigation
as
well.

Moving
on
to
Phase
3,
the
measure
of
aromatase
activity,
the
objective
of
the
Phase
3
studies
is
to
determine
if
aromatase
activity
is
increased
by
exposure
to
atrazine
during
sensitive
developmental
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
71
stages.
Whether
or
not
this
phase
is
conducted,
is
dependent
on
whether
the
proceeding
studies
on
sex
steroids
suggest
that
modulation
of
aromatase
activity
may
be
responsible
for
elevated
estrogen
levels
for
example.

Similar
to
the
approach
for
measurement
of
the
sex
steroids,

there's
a
developmental
component
to
the
approach.
The
developmental
expression
of
aromatase
mRNA
has
been
determined
for
Xenopus
laevis
and
is
presented
in
the
panel
on
the
right.

Expression
of
aromatase
mRNA
is
apparent
at
approximately
stage
50
and
generally
increases
with
development.
This
expression
pattern
is
overlaid
on
the
previous
graphic
depicting
the
developmental
sensitivity
toward
estrogen­
induced
feminization
which
decreases
with
development.

Taken
together,
these
studies
suggest
that
aromatase
activity
must
be
elevated
prematurely
and
at
sufficiently
high
levels
during
the
estrogen­
sensitive
stages
to
result
in
feminization
in
males.

Therefore,
at
this
point
in
time,
the
most
appropriate
approach
may
be
to
examine
this
phenomenon
prior
to
the
onset
of
metamorphosis
which
is
generally
considered
to
begin
at
about
Stage
54.

Phase
4,
Aromatase
Inhibitor
Study.
If
it
is
demonstrated
that
the
previous
phase
of
aromatase
activity
is
increase
by
atrazine,
then
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
72
it
may
be
desirable
to
determine
if
co­
administration
of
an
aromatase
inhibitor
with
atrazine
rescues
the
male
gonad
from
feminization.

This
approach
would
require
that,
first,
the
effective
concentration
of
an
aromatase
inhibitor
be
empirically
determined.
Then
based
on
the
dose
response
data
in
the
first
three
phases,
the
organisms
would
be
exposed
to
the
inhibitor
simultaneously
will
affect
atrazine
concentrations.
Then
the
effects
could
be
analyzed
similar
to
the
Phase
1
studies.

On
Phase
5,
Evaluating
Ecological
Relevancy,
the
objective
here
in
this
final
phase
is
to
determine
if
the
potential
effects
of
atrazine
on
gonadal
differentiation
results
in
reduced
fertility.

There
may
be
several
approaches
to
this,
but
based
on
the
premiss
that
feminization
of
males
occurred
in
Phase
1,
the
approach
that
I've
outlined
here
is
to
determine
if
feminization
alters
fertility
using
either
in
vitro
or
in
vivo
fertilization
methods.
Although
such
methods
are
used
routinely
for
reproductive
purposes
in
numerous
laboratories,
they
are
not
currently
used
to
quantify
fertility
as
an
important
parameter
to
estimating
reproductive
output.
If
such
studies
are
warranted,
then
additional
research
would
have
to
be
conducted
to
establish
quantifiable
methods.

So
in
conclusion,
it
is
possible
to
reduce
the
major
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
73
uncertainties
associated
with
the
potential
risk
of
atrazine
to
amphibians
by
following
a
phased
sequence
of
laboratory
studies
focused
on
the
critical
components
of
the
risk
hypothesis,
using
currently
available
high
quality
methods
that
are
standard
for
aquatic
toxicology,
and
establishing
and
adhering
to
study
quality
indicators.

Although
the
analysis
plan
that
I
have
presented
lays
out
a
relatively
comprehensive
set
of
studies
to
evaluate
the
risk
hypothesis,
the
extent
to
which
the
proposed
studies
are
actually
conducted
depend
on
two
important
factors.
The
first
is
whether
or
not
risk
management
decisions
require
reduction
in
the
current
level
of
uncertainties
to
proceed.
And,
second,
is
whether
or
not
the
outcomes
of
the
initial
phases
indicate
that
additional
studies
are
logical
and
valuable
in
terms
of
testing
the
components
of
the
risk
hypothesis.
These
issues
will
have
to
be
evaluated
as
more
data
become
available.

Thank
you
for
your
attention.
I'd
be
happy
to
entertain
questions.

DR.
ROBERTS:
Thank
you
for
your
presentation.
I'd
like
to
give
the
panel
the
opportunity
now
to
ask
you
any
questions
they
might
have
about
the
proposed
Agency
approach
that
you've
described.
Let's
start
with
Dr.
LeBlanc.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
74
DR.
LEBLANC:
Joe,
do
you
anticipate
a
temporal
sequence
to
the
performance
of
the
phases
or
might
some
of
the
phases
be
conducted
at
the
same
time?

DR.
TIETGE:
I
think
it
would
be
up
to
the
laboratory
who's
proposing
to
do
the
studies.
They
certainly
could
be
conducted.
Or
one
could,
for
example
conduct
an
organismal­
level
study
and
then
archive
samples
for
further
analysis
that
are
in
that
tier.
I
wouldn't
want
to
propose
the
tier
as
being
too
linear.
So
I
think
some
of
them
could
be
done
at
the
same
time.

DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
I
would
like
to
know
is
the
hypothesis
is
focused
on
aromatase
production.
So
what
we
found
and
up
to
now
and
what
is
more
or
less
a
verified
is
feminization
or
demasculinization.
So
this
could
be
also,
I
think,
obtained
by
anti­
androgenic
effects.
I
think
for
conceptional
frame
work,
we
should
include
all
as
a
possibility
so
that
we
have
an
alternative
pathway
to
receive
feminization
or
demasculinization
via
the
anti­
androgenic
pathways.

This
should
be
maybe
from
a
serial
point
of
view
at
least
included.

DR.
TIETGE:
I
would
agree.
And
I
tried
to
leave
the
door
open
in
the
alternative
path.
I
think
that
once
you
get
away
from
the
organismal
level
effects
in
the
risk
hypothesis,
you
have
more
and
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
75
more
uncertainty
especially
as
you
go
toward
the
mechanistic
side.

So
I
agree
with
you
totally.

DR.
ROBERTS:
Dr.
Kelley
then
Dr.
LeBlanc.

DR.
KELLEY:
With
regard
to
the
flow­
through
proposal,
as
you
know,
most
Xenopus
colonies
aren't
raised
in
flow­
through.
And
in
the
wild,
of
course,
you
don't
find
very
many
tadpoles
in
streams
with
any
motion
what
so
ever.
So
are
you
aware
of
data
that
indicate
that
a
flow­
through
system
as
opposed
to
a
static
renewal
system
with
large
volumes
of
water
would
have
differential
effects
on
mortality?

DR.
TIETGE:
Well,
the
issue
of
the
static
renewal
versus
flow­
through
isn't
­­
let's
see,
how
am
I
going
to
answer
this.
If
one
goes
back
to
the
ASTM
guidelines,
which
I
think
are
fairly
valuable
in
terms
of
establishing
guidance
for
biological
loading,
there
are
guidelines
for
static
tests.
However,
for
a
typical
Xenopus
individual
it
would
require
probably
three
to
four
liters
of
solution
per
individual
to
meet
those
standards.
So
if
you
want
to
have
tests
that
have
high
enough
end
value
in
order
to
test
your
hypothesis,
it
would
require
very
large
exposure
chambers.
And
I
think
that
would
be
very
limiting.

In
fact,
in
our
laboratory,
we
use
flow­
through
conditions
routinely
and
achieve
metamorphic
completion
in
about
seven
weeks
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
76
typically
post
fertilization
and
usually
with
99
percent
or
higher
survival
and
apparently
good
growth
rates.
Often
we
have
organisms
in
the
1.8­
to
2­
gram
range
when
they're
right
around
stage
60
prior
to
the
weight
loss
that
occurs
through
metamorphosis.

I'm
not
sure
if
I
answered
your
question.
Yes,
there
is
some
basis
for
it.
Also,
I
understand
that
even
with
the
native
Ranas,
they
don't
necessarily
live
in
flow­
through
conditions
in
the
field;
but
they
also
don't
live
in
a
static
aquarium
in
the
field
because
the
system
is
more
complex.

DR.
ROBERTS:
If
you
have
a
follow­
up
questions
that
would
be
fine.

DR.
KELLEY:
So
this
does
bear
to
the
issue
of
ecological
relevance,
however.
So
it's
not
entirely
clear
how
a
continuous
flow­
through
system
would
bear
either
on
Xenopus
that
avoid
a
flow
system
or
on
Rana
even
if
it's
not
totally
static
since
much
of
the
data,

since
some
of
the
concern,
at
least,
comes
from
things
like
drainage
ditches
and
ponds
accumulating
in
runoff
from
fertilized
fields.
So
one
should,
I
think,
think
about
whether
flow­
through
data,
although
well­
controlled
from
the
point
of
view
of
water
quality,
actually
would
mimic
the
conditions
under
which
exposure
might
occur.

DR.
TIETGE:
I
understand
your
point.
It's
a
point
well­
taken.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
77
However,
in
aquatic
toxicology,
I
would
suggest
that
the
more
control
you
have
over
the
experiment
in
terms
of
water
quality,
the
more
confidence
you
have
that
the
effect
that
you're
observing
is
related
to
the
concentration
of
the
chemical.
I
think
it
is
a
much
different
question
to
ask
whether
or
not
a
laboratory
study
is
directly
applicable
or
representative
of
field
conditions.
I
think
that's
the
state
of
the
science
actually.

DR.
ROBERTS:
Next
Dr.
LeBlanc,
followed
by
Dr.
Isom,
Dr.

Kloas,
and
then
Dr.
Skelly.

DR.
LEBLANC:
In
formulating
your
hypothesis
on
the
mechanism
by
which
atrazine
might
elicit
effects
on
developing
gonads,
did
you
consider
the
work
of
Ralph
Cooper
showing
in
rats
the
effects
on
gonadotropins?

MR.
TIETGE:
The
hypothalamus,
hypothalamic
vectors?

DR.
LEBLANC:
Yeah,
in
suppressing,
glueinizing
hormone.

MR.
TIETGE:
I'm
certainly
aware
of
it.
Of
course,
Ralph's
in
the
Agency
so...
But,
no,
I
think
what
we
tried
to
follow
was
the
information
that
we
thought
was
more
specific
or
germane
to
the
amphibian
issue.
So,
no,
we
didn't
really
take
it
into
consideration.

However,
as
I
also
answered
Dr.
Kloas,
we
did
leave
the
door
open
that
as
you
find,
as
the
data
indicates,
you
can
go
to
an
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
78
alternative,
take
alternative
paths.

DR.
LEBLANC:
I
think
induction
of
aromatase
may
be
consistent
with
his
observations.
But
certainly
in
the
young
male,
the
profound
effect
that
he
says
in
terms
of
steroid
levels
was
a
suppression
of
testosterone.
So
I
would
agree
with
Dr.
Kloas
that
certainly
anti­
androgens
is
something
you
might
want
to
consider
as
a
positive
control.

DR.
ROBERTS:
Dr.
Isom,
Dr.
Kloas,
and
then
Dr.
Skelly.

DR.
ISOM:
The
objective
of
your
proposed
study
is
obviously
to
determine
the
reproductive
fitness
of
the
species.
And
I
noted
in
a
number
of
the
studies
that
have
been
published
that
not
only
do
we
have
observed
or
postulated
effects
upon
gonotropic
development,
but
also
secondary
sex
characteristics
that
are
important
for
reproduction
like
laryngeal
muscle.

I
was
wondering
why
you
aren't
proposing
to
at
least
measure
that
in
the
species
in
your
exposure
studies.
And
then
a
second
question
is
have
you
thought
about
positive
controls
for
aromatase
that
is
inducers
exposure.

MR.
TIETGE:
Okay.
The
first
question
was
­­

DR.
ISOM:
The
laryngeal
muscle.

MR.
TIETGE:
Right.
Laryngeal
muscle.
If
I
recall
the
data
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
79
correctly,
laryngeal
muscle
effects
were
observed
at
higher
concentrations.
And
so
the
gonadal
effects
were
more
sensitive.

They
certainly,
if
you
affected
laryngeal
muscle
to
a
certain
level
undetermined
at
this
point
in
time,
you
certainly
could
expect
to
have
some
potential
effects
on
reproductive
activity.
And
I
think
Dr.

Steeger
mentioned
that
in
his
talk
as
well.

I
mean
the
door
could
be
open
to
that.

DR.
ISOM:
It
seems
to
me
that
if
you're
doing
the
study,
you
have
the
animal
there.
It
wouldn't
be
that
difficult
to
do
that.

MR.
TIETGE:
Actually,
we
have
­­
I
have
no
experience
with
that
endpoint.
I'm
not
sure
how
to
deal
with
it.
Would
anybody.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
I
do
have
experience
with
that
endpoint
since
that's
what
I've
studied
for
a
good
long
while.
And
one
of
the
things
that
I
think
would
be
required
in
a
study
of
this
kind
is
to
enable
the
animals
to
grow
until
they
reached
reproductive
maturity.
If
you
wanted
to
study
the
endpoint
of
sexual
differentiation
functionally,

both
in
terms
of
active
spermatogonia
in
testing
situations,
either
removing
the
testes
or
doing
natural
matings
which
would
be
more
variable,
and
if
you
also
wanted
to
study
the
endpoint
of
laryngeal
function,
which
is
to
produce
the
male
advertisement
call
among
other
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
80
calls,
then
you
would
need
to
actually
have
your
animals
go
probably
for
about
a
year.
Although,
if
your
animals
are
growing
fast,
you
can
probably
get
them
to
call
in
six
months.

So
those
seem
like
natural
endpoints
that
relate
very
closely
to
the
issue
of
reproductive
success.

DR.
ROBERTS:
Okay.

MR.
TIETGE:
A
positive
aromatase
inducer,
that
certainly
could
be
done.
We
didn't
include
it
because
we
were
trying
to
stay
as
directly
on
the
track
as
we
could.
But
certainly
it's
a
fine
idea.

DR.
ROBERTS:
Next
question
or
clarification
from
Dr.
Kloas
followed
by
Dr.
Skelly,
Dr.
Denver,
and
then
Dr.
Green.

DR.
KLOAS:
Of
course,
I
would
like
to
come
back
to
flow­
through
versus
static
renewal
system.
I
think
up
to
now
we
have
no
real
indication
of
if,
at
least
some
unofficial
indications,
that
maybe
a
flow­
through
would
reduce
positive
control
effects.
For
instance,
for
estradiol,
for
feminization
there's
one
study
I'm
aware
of
they
have
a
reduced
feminization
effect
because
it's
flow­
through.

I'm
not
sure.
So
I
think
before
just
fixing
everything,
we
should
also
maybe
be
aware
that
we
need
a
comparative
study,
a
comprehensive
study,
between
flow­
through
and
static
renewal.

I
know
also
from
this
flow­
through
experiment,
the
tadpole's
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
81
grow
well
and
also
the
developmental
stages
were
reached.
But
I'm
a
little
bit
concerned
about
sexual
differentiation.
I'm
not
sure
that
you
can
say
it
doesn't
matter.

MR.
TIETGE:
I'm
familiar
with
the
study,
I
think,
that
you're
referring
to.
And
I
find
the
results
to
be
somewhat
enigmatic.

However,
in
using
the
flow­
through
conditions
in
our
laboratory,
I
don't
think
there's
any
effect
on
sexual
differentiation
based
on
the
method
itself.
I
think
that
from
an
efficacy
of
exposure
point
of
view,

the
flow­
through
system
probably
ought
to
be
more
effect
than
in
a
static
system
because
there's
a
lot
of
evidence
that
exists
with
the
more
hydrophobic
chemicals
will
be
depleted
under
static
conditions.

That's
very
well
established
in
the
aquatic
toxicology
literature.

DR.
KLOAS:
I
agree
from
a
theoretical
point
of
view.
But
did
you
do
a
positive
control
using
estradiol
for
inducing
feminization
in
parallel
in
this
system
and
it
works?

MR.
TIETGE:
It
works,
yes.

DR.
KLOAS:
I
would
like
to
see
it.

DR.
ROBERTS:
Let's
go
to
Dr.
Skelly.
But
before
we
continue
with
the
questions,
let
me
just
remind
the
panel
that
you
will
have
the
opportunity
to
provide
feedback
on
this
approach
as
we
address
the
questions.
The
purpose
now
is
really
just
to
get
clarification
on
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
82
Agency's
approach.
Dr.
Skelly.

DR.
SKELLY:
I
had
a
question
about
the
designation
of
Xenopus
laevis
as
a
primary
species
and
Rana
pipiens
as
a
secondary
species.
And
I
guess
I'll
leave
it
a
little
bit
open­
ended.
But
I'm
interested
in
why
you
made
that
distinction
and
what
it's
going
to
mean
in
terms
of
the
timing
in
your
conceptual
model
of
when
things
would
happen
and
what
that
means
in
terms
of
how
the
studies
will
be
used
in
terms
of
weight
of
evidence
or
how
they'll
be
prioritized
in
your
thinking.

MR.
TIETGE:
Well,
on
the
first
point,
the
speed
and
utility
of
Xenopus
laevis
over
Rana
pipiens,
I
think,
is
fairly
universally
accepted.
I
mean
you
can
do
­­
I
realize
that
some
laboratories
can
produce
Rana
pipiens
throughout
the
year
for
studying.
However,

most
cannot.
And
most
are
limited
to
collections
that
occur
early
in
the
spring
to
conduct
Rana
studies.

By
contrast,
one
could
have
multiple
studies
within
one
year
with
Xenopus
laevis
and
make
some
headway
without
having
to
wait
until
the
natural
breeding
season
for
the
Rana
species.

Also,
I
think
if,
based
on
the
studies
that
were
submitted,
I
think
the
question
is:
Can
Xenopus
be
a
reliable
surrogate
for
Rana?

With
regard
to
this
endpoint,
taking
the
studies
that
have
been
done
at
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
83
face
value,
I
guess
you'd
have
to
say
yes.
If
you
have
a
surrogate
that's
more
efficient,
it's
probably
the
way
to
go.

However,
as
with
any
of
the
issues
of
species
extrapolation,

there's
no
substitute
for
a
corroboration
in
terms
of
some
comparative
studies.
And
so
the
intent
of
my
point
that
I
made
was
that
you
might
do
the
hard
work,
the
voluminous
work,
with
Xenopus
and
try
to
get
as
far
into
understanding
the
phenomenon
as
you
can;
and
then,
when
you
have
a
good
handle
on
what's
going
on
and
you
have
a
level
of
confidence
that
allows
you
to
go
forward,
that
you
would
then
go
back
to
Rana
and
do
some
confirmatory
studies
because
I
think
you
are
limited
in
terms
of
the
methods
that
are
available
with
Rana.
So
that
was
my
­­

DR.
SKELLY:
I'll
just
follow
up
quickly.
So
does
this
mean
with
your
flow
chart
that
you
might
do
the
work
on
Xenopus
and
come
up
with
a
negative
result
and
stop
there
because
you
don't
need
to
corroborate
a
negative
result
versus
a
positive
result?

MR.
TIETGE:
Well,
that's
a
great
question
because
you
never
know
what
to
do
with
negative
results.
It's
hard
to
make
decisions
based
on
negative
results.
But
you
might
still
have
enough
concern,

based
on
the
existing
information,
that
you'd
go
back
and
do
a
Rana
study
at
the
organismal
level
even
if
it
was
negative
in
Xenopus.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
84
I
think
where
Xenopus
really
has
the
advantage
is
when
you
get
into
the
iterative
studies,
especially
at
the
mechanistic
level,
where
you're
trying
to
define
what's
going
on.
Because
then
that
would
if
you,
for
example,
if
the
working
hypothesis
were
demonstrated
using
that
approach
that
I
laid
out,
you
might
be
able
to
very
quickly
then
go
onto
Rana
pipiens
and
verify
that
the
same
thing
is
going
on.
And
that's
what
I
was
referring
to
as
establishing
a
basis
for
inter­
species
extrapolation
when
it
came
to
the
value
of
the
mechanistic
data.

DR.
ROBERTS:
Dr.
Denver
followed
by
Dr.
Green
and
then
Dr.

Gibbs.

DR.
DENVER:
Joe,
some
of
the
main
concerns
of
the
studies
that
were
reviewed
by
the
Agency
were
the
variability
within
studies
and
also
the
variability
among
studies.
And
I'm
curious
if
the
Agency
has
considered
ways
to
control
for
this
variability
in
terms
of
the
assays
that
have
been
chosen
and
the
way
to
validate
these
assays
among
different
labs.

I'm
thinking
about
the
two
types
of
assays
that
you're
proposing.
One
is
the
morphological
assay
where
you're
looking
at
gonadal
morphology
and
the
presence
or
absence
of
intersex
individuals.
And
these
types
of
scoring
of
intersex
individuals
may
be
influenced
by
subjective
measures.
Are
there
any
other
types
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
85
assays
that
may
be
applied
to
this,
for
example,
by
a
chemical
assay
that
may
assay
for
testicular
or
ovarian
antigens
that
may
be
more
objective
and
may
be
even
more
sensitive?
Has
there
been
consideration
given
to
developing
standardized
pools,
say,
of
plasma
that
can
be
distributed
to
the
laboratories
for
estimates
of
estrogens
and
androgens
in
blood
plasma
to
validate
those
assays
among
laboratories,
things
like
that?

MR.
TIETGE:
Well,
with
regard
to
the
more
mechanistic
things
you
brought
up
in
your
last
two
points,
no,
we
don't
really
have
anything
going
on
there.

But
going
back
to
your
first
comment
regarding
the
subjectivity,
I
think
our
approach
is
that
the
histopathology
has
to
be
included.
And
histopathology
or
pathology
in
general
is
a
subjective
science.
But
I
think
it's
often
­­
I
should
say
it's
often
a
subjective
science.
However,
it's
a
science
that
has
a
lot
of
confidence
based
on
the
experience
and
the
review
process
that's
a
typical,
modern
pathology
reviews
and
such.

So
I
think
that
you
can
use
a
subjective
endpoint
effectively,
I
think,
if
you
include
the
histopathology.
But
with
regard
to
the
more
mechanistic­
based
things,
I
don't
know
of
anything
that's
going
on
in
that
regard.
And
we
certainly
don't
have
anything
right
now.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
86
We
are
interested
in
developing
those
ideas.
And
I
think
there's
movement
in
the
Agency
to
develop
and
validate
amphibian
methods.

But
there
are
no
validated
methods
currently.

DR.
ROBERTS:
Dr.
Green
then
Dr.
Gibbs
then
Dr.
Richards.

DR.
GREEN:
Regarding
some
of
the
secondary
characteristics
in
the
male
that
could
be
studied,
I
wondered
if
you
could
comment
on
why
in
the
literature
and
anywhere
else
that
I've
seen
so
far
there
hasn't
been
an
evaluation
of
the
nuptial
pads
in
Xenopus
laevis.

They're
very
easy
to
see,
quite
prominent,
in
post­
metamorphic
young
juveniles.
And
certainly
if
they
were
feminized
as
a
result
of
exposure
to
chemicals
in
the
wild,
I
would
think
that
those
would
diminish
and
that
could
be
followed
up
in
field
studies
as
well
as
in
the
laboratory.

MR.
TIETGE:
So
your
question
is
­­
are
you
making
a
point,
or
are
you
asking
a
question?

DR.
GREEN:
No.
I'm
asking:
Do
you
have
plans
to
look
at
that?
And
it
would
follow
up
on
Dr.
Darcey's
comment
that
you
might
have
to
extend
the
studies
a
little
longer
in
order
to
be
able
to
see
them.
But
they
can
be
detected
grossly
and
histologically
quite
early,

I
believe.

MR.
TIETGE:
No.
I
mean
our
plan
is
just
a
plan,
and
we're
87
1
looking
forward
to
your
input.
And
I
think
we'd
accept
that
and
move
2
forward.

3
DR.
GREEN:
Okay.
In
addition
to
the
nuptial
pads,
another
4
thing
would
be
the
ventral
folds
around
the
cloaca
in
females
are
5
quite
prominent.
And
you
would
expect,
perhaps,
if
males
were
6
feminized,
that
they
might
become
quite
prominent
in
the
laboratory
7
as
in
the
wild
as
well.

8
DR.
ROBERTS:
Dr.
Gibbs.

9
DR.
GIBBS:
A
quick
question
about
endpoints.
There
seems
to
10
be
an
inconsistency
insofar
as
in
the
Phase
1
studies
insofar
as
the
11
survival
is
listed
as
an
independently
varying
endpoint
to
be
12
measured.
And
yet
in
the
recommended
study
protocol,
survival
was
13
something
to
be
constrained
to
remain
above
90
percent.

14
MR.
TIETGE:
90
percent
survival
in
the
controls
is
what
I
was
15
referring
to
with
regard
to
be
an
indicator
the
methods
used
in
the
16
tests
were
sufficient
to
promote
survival.

17
DR.
GIBBS:
Okay.
Under
the
various
treatments.

18
MR.
TIETGE:
Under
the
treatments,
of
course,
that
would
be
19
given.

20
DR.
ROBERTS:
Dr.
Richards
then
Dr.
Coats.

21
DR.
RICHARDS:
Just
a
very
general
question
relating
to
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
88
two
words
"
ecological
effects."
We've
talked
about
reproduction,

some
aspects
of
that.
I
just
wonder,
in
a
very
general
sense,
what
has
the
Agency
thought
about
that
term.
Is
what
we're
talking
about
here
so
far
enough
to
cover
the
realm
of
ecological
effects?
In
a
broader
sense
of
­­

MR.
TIETGE:
Broader
than
the
ecological
relevancy
of,
I
think,
Steve,
you
might
want
to
jump
in
on
this.

DR.
BRADBURY:
I
think
when
you're
doing
a
chemical
risk
assessment
and
we're
starting
with
some
observations
from
the
field,

but
certainly
starting
from
sort
of
the
building
blocks
of
building
of
a
hypothesis,
a
working
hypothesis,
it's
sort
of
building
up
as
opposed
to
top
down,
bottom
up
kind
of
thing.

Certainly
most
of
the
ecological
risk
assessments
for
chemical
stressors
are
across
the
Agency
not
just
in
pesticides,
working
through
what
are
the
relevant
organismal
responses
that
give
you
insights
into
population
or
community
responses.
And
I
think
it's
fair
to
say
that
we're
all
at
sort
of
the
edge
of
moving
into
how
do
you
take
a
look
at
population
models,
how
do
you
start
thinking
about
meta­
population
models,
how
do
you
start
interfacing
chemical
effects
with
habitat
quality
to
understand
the
relative
roles
of
those
different
stressors
on
a
population
or
community
structure.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
89
I
think
today,
which
is
probably
reasonably
representative
of
other
types
of
chemical
risk
assessments,
we're
still
at
that
interface
of
how
does
the
toxicology
start
to
merge
into
population
biology
or
landscape
ecology.
What
are
the
insights
from
population
biology
information
or
landscape
ecology
information
that
give
insights
into
those
organismal
responses
that
are
most
critical
for
population
viability
and
what's
a
plausible
toxicological
mechanism
to
influence
those
endpoints.

On
the
surface,
at
least
in
this
specific
example,
the
discussions
of
measures
of
effects
and
risk
assessment
endpoints
are
at
least
qualitatively
associated
with
reproductive
fitness.
But
resolution
spatial
and
temporal
in
defining
that
obviously
would
take
increasing
levels
of
information
at
all
sorts
of
levels
of
biological
organization.

DR.
ROBERTS:
Dr.
Coats.

DR.
COATS:
Yes,
when
you're
evaluating
the
reports,
did
you
consider
the
analytical
methodology
used
for
chemistry
especially
in
terms
of
the
quality
of
the
data
or
the
selection
of
the
methodology
and
how
sensitive
it
might
be
or
how
specific
it
might
be?

MR.
TIETGE:
We
were
aware
of
the
level
of
detection
that
was
associated
with
the
assays
that
were
used.
We
didn't
have
any
input
on
the
assay
that
were
­­
again,
there
were
no
guidelines
for
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
90
registrant
to
conduct
these
studies
by.
So
we
accepted
the
studies
at
face
value
in
terms
of
what
the
level
of
detection
was.
But
it
was
clear
that
the
levels
of
variability
that
were
associated
with
the
measurements
themselves
were
high
and
in
many
cases
­­
in
some
cases,
when
backgrounds
were
subtracted
from,
when
the
assay
background
was
subtracted
from
the
treatment
samples,
they
were
actually
negative
values
afterwards.

In
our
opinion,
other
methods
from
looking
at
those
endpoints
may
be
necessary.
Rather
than
relying
on
ELISA
assays,
some
other
method
may
be
more
appropriate.
Again,
we're
looking
for
input
from
the
Panel
to
address
those
concerns.

DR.
ROBERTS:
Any
other
questions
from
the
Panel?

If
not,
I
think
this
would
be
a
good
time
for
a
break.
Let's
take
a
break
and
reconvene
at
10
minutes
before
the
hour
when
Dr.
Steeger
will
present
Agency
conclusions.
So
let's
break
for
15
minutes.

[
Break
at
10:
35
a.
m.;
session
resumed
at
10:
55
a.
m.]

DR.
ROBERTS:
We're
starting
now.
So
if
folks
in
the
audience,
if
you
could
make
your
way
to
your
seats
promptly,
please.

DR.
STEEGER:
The
computer
is
in
the
process
of
re­
booting.

DR.
ROBERTS:
All
right.
We'll
await
the
computer.

DR.
STEEGER:
As
I
indicated
earlier
today,
the
criteria
for
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
91
doing
a
ecological
risk
characterization
is
that
the
characterization
has
to
be
transparent,
it
has
to
be
clear,
it
has
to
be
consistent,
and
it
has
to
be
reasonable.
The
Agency
has
reviewed
a
number
of
studies.

And
the
process
that
we've
used
in
making
those
reviews
has
been
captured
by
Steve
Bradbury,
and
it's
discussed
throughout
my
presentation.
And
critical
in
that
process
is
the
problem
formulation.

And
the
iterative
series
of
processes
that
occur
between
the
risk
manager
and
the
risk
assessor.

Based
on
the
studies
that
the
Environmental
Fate
and
Effects
Division
has
reviewed
these
studies
that
were
submitted
as
of
February
28,
it
has
concluded
that
there
are
lines
of
evidence
that
suggest
that
the
exposure
to
atrazine
may
result
in
developmental
effects
in
amphibians.
However,
there
were
basic
inconsistencies
and
variability
in
the
studies
that
we
reviewed
that
prevent
us
from
either
refuting
or
confirming
those
effects.

The
Agency
has
recommended,
based
on
its
review,
that
a
phased
process
be
undertaken
to
examine
specific
measurement
endpoints.
And
in
conducting
this
phased
process
that
Joe
Tietge
outlined,
the
Agency
would
be
working
its
risk
managers
to
determine
the
level
of
uncertainty
that
would
be
necessary
to
resolve
in
order
for
some
decision
to
be
reached.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
92
One
of
the
questions
that
was
posed
to
Joe
was
what
was
the
temporal
sequence
in
doing
the
phased
studies.
Joe
indicated
studies
or
different
phases
could
be
done
concurrently.
But
the
critical
input
comes
from
the
risk
manager.
The
risk
manager
has
to
be
able
to
define
the
level
of
uncertainty
that
they're
willing
to
accept
in
order
to
make
a
risk
management
decision.
The
risk
assessors
collect
the
information
and
help
the
risk
manager
understand
how
much
of
the
uncertainty
is
associated
with
the
data
that
are
available.

We
believe
that
the
current
studies
contain
sufficient
uncertainty
that
we're
unable
to,
as
I
said,
refute
or
confirm
whether
atrazine
is
indeed
having
effects
on
amphibian
development.
But
the
bottom
line
is,
and
throughout
my
presentation,
I
stressed
that,
given
the
fact
that
over
several
studies
and
environmental
conditions
and
species,
atrazine
exposure
did
appear
to
be
having
some
impact
on
gonadal
development.

But
because
of
the
lack
of
consistency
and
the
type
of
effect
elicited,
the
lack
of
a
dose
response,
the
inability
of
the
current
studies
to
demonstrate
a
plausible
mechanistic
action,
and
our
inability
or
the
Agency's
inability
to
link
the
measurement
endpoints
that
have
been
reported
with
our
traditional
assessment
endpoints
of
reproduction,
survival,
and
growth
currently,
we're
unable
to
make
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
93
any
statements
regarding
the
utility
of
these
data.

As
I
indicated
in
my
presentation,
though,
the
current
studies
do
provide
beyond
the
line
of
evidence,
information
on
sources
of
variability
and
how
future
studies
might
be
designed
to
better
account
for
those
sources
of
variability
and
provide
reliable
means
of
measuring
the
effects
of
atrazine
on
gonadal
development.

Steve,
would
you
care
to
add
anything?

DR.
BRADBURY:
I
think
that
Tom
summed
it
up
quite
well.
I
guess
as
the
Panel
deliberates
and
we
have
some
discussions,
just
emphasizing
again
using
the
Agency's
risk
assessment
guidelines
as
a
way
to
organize
our
thoughts
and
to
think
about
the
science
at
hand
and
the
science
in
the
context
of
making
a
regulatory
decision,
and
sometimes
that
creates
some
different
choices
that
one
makes
in
terms
of
phrasing
questions
and
articulating
and
understanding
what
the
uncertainties
are
associated
with
different
aspects
of
the
science
when
the
aspects
of
this
science
are
in
the
context
of
making
a
regulatory
decision.

And
certainly
we're
looking
very
forward
to
your
comments
and
input
in
terms
of
helping
to
define
and
understand
the
certainty
that
exists
in
the
information
today
and
your
thoughts
on
those
if
you
feel
there
are
uncertainties
that
remain,
the
nature
of
those
uncertainties
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
94
in
the
context
of
how
we've
taken
a
look
at
it,
and
then
some
approaches
to
refine
the
knowledge
base.

The
ultimate
decision
on
how
much
information
is
enough
to
make
a
decision
starts
to
leave
the
realm
of
science,
but
science
is
important
to
inform
that
process
so
that
the
decisions
that
are
made
are
reasonable
and
transparent
and
clear
as
Tom
indicated.

So
I
think
at
that
we'd
be
happy
to
turn
it
over
to
Steve
and
carry
forward.

DR.
ROBERTS:
Thank
you.
This
concludes
the
Agency's
presentation.
I'd
like
to
ask
the
Panel,
again,
if
you
have
any
questions
regarding
the
presentations
this
morning
before
we
move
on
in
the
agenda.
Yes,
Dr.
Kelley.

DR.
KELLEY:
Given
that
you've
decided
that
Xenopus
laevis
is
going
to
be
your
primary
experimental
target
in
this,
I
wondered
to
what
extent
you
would
also
require
before
changing
the
current
regulatory
environment
that
effects
be
demonstrated
in
native
North
American
species.
Is
that
a
fair
question?

DR.
BRADBURY:
Run
that
by
me
one
more
time.
I
don't
think
I
understand
the
question.

DR.
KELLEY:
Well,
look,
so
you
decide
to
use
an
experimental
animal
that's
found
in
South
Africa.
There
are
feral
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
95
populations
here
in
America.
But
I
assumed
that
we
as
Americans
and
EPA
aren't
too
much
worried
about
the
survival
of
feral
populations
of
Xenopus
in,
you
know,
Arizona
golf
courses,
right,
which
is
where
they
tend
to
live.
So
suppose
a
whole
scenario
was
developed
around
Xenopus
and
effects
of
atrazine,
to
what
extent
could
we
then
­­
to
what
extent
could
you
use
that
information
to
apply
to
our
own
native
species
here
in
America.

DR.
BRADBURY:
Now
I
understand
your
question,
sorry.
It's
a
challenging
question,
and
I
think
it
transcends
much
of
ecotoxicology
and
ecological
risk
assessment
where
you
may
be
able
to
test,
get
information
on
a
handful
of
species
and
potentially
have
to
extrapolate
to
hundreds
of
species
or
tens
of
species
at
least
in
different
landscape
scenarios.

I'd
like
to
back
up
to
the
first
way
you
phrased
your
question
to
indicate
that
part
of
the
lines
of
evidence
to
determine
where
we
are
and
where
we
may
need
to
go,
in
fact,
took
advantage
of
some
field
studies
as
well
as
some
laboratory
studies
to
develop
the
lines
of
evidence
that
it's
plausible
to
formulate
a
hypothesis,
although
we
may
not
have
the
confidence
right
now
to
quantify
the
probabilities
of
risk
based
on
the
information
at
hand.
But
the
epidemiology­
type
investigations
as
well
as
reductionist
studies
combined
together
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
96
increase
your
understanding
and
your
ability
to
quantify
what
you
know
and
what
you
don't
know
and
the
gaps
that
may
remain.

In
the
context
of
starting
with
a
biological
model
to
start
to
get
some
clarity
into
the
issue
at
hand
for
the
risk
assessment
is
sort
of
a
fundamental
question.
In
aquatic
toxicology,
typically
the
Agency
is
using
fathead
minnows
and
rainbow
trout
and
bluegill
as
species
to
try
to
represent
what
could
happen
to
the
thousands
of
fish
species
in
the
country.

Now,
through
problem
formulation
and
thinking
about
which
landscapes
we're
talking
about,
one
starts
to
narrow
down
the
type
of
species
that
one
needs
to
focus
on.
But
you're
still
dealing
with
the
fundamental
species
extrapolation
challenge.
What
are
the
toxicokinetic
differences
between
species?
What
are
the
toxicodynamic
differences
or
similarities
across
species
to
help
put
some
bounds
on
the
potential
variability
across
the
species.
And
you
have
to
blend
that
with
some
of
the
practicality
of
generating
toxicological
information
that
provides
the
ability
to
control
some
of
the
natural
variability
of
the
world
so
that
one
can
tease
out
the
signal
the
chemical
may
be
sending
in
terms
of
a
dose­
response
study
and
try
to
understand
what
that
chemical
is
doing
in
light
of
all
the
other
variabilities.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
97
So
one
trades
off
between
a
biological
model
that's
well
understood.
What
do
we
know
about
the
developmental
biology
of
a
given
species
or
the
reproductive
biology
of
a
given
species
that
helps
us
get
insights
into
the
potential
effects
a
chemical
may
have.
And
it
also
then
gives
us
insights
into
the
issues
we
should
be
thinking
about
in
terms
of
how
do
we
extrapolate
an
effect
seen
in
one
species
to
other
species.
Do
these
species
also
contain
the
same
receptor?
How
well­
conserved
is
the
receptor?
How
well­
conserved
are
other
biochemical
pathways
in
terms
of
mechanism
of
action
or
detoxification
or
activation?
What
are
the
issues
in
terms
of
toxicokinetics
uptake
distribution?
So
all
those
things
sort
of
come
into
play.

In
the
context
of
getting
started
on
this
challenge
and
our
proposal,
and,
again,
it's
a
proposal.
It's
a
plan
put
before
our
scientific
peers
to
gain
your
wisdom
and
insights
as
well.
In
terms
of
amphibian
toxicology
and
in
the
context
of
developmental
biology
issues
associated
with
a
ecotoxicological
risk
assessment,
Xenopus
offers
one
way
to
get
started
efficiently
to
start
to
get
some
clarity
in
terms
of
the
ability
of
atrazine
to
cause
a
reproducible
response
in
terms
of
a
developmental
endpoint.

If
one
sees
that,
it
gives
some
insights
into
aspects
of
species
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
98
extrapolation.
If
that
happens,
what
would
be
the
observed
mechanisms
of
endocrinology
or
developmental
biology
that
would
have
to
be
present
in
other
amphibians
for
an
observation
in
Xenopus
to
be
relevant
to
other
amphibian
species?
To
the
extent
we
can
extrapolate
or
determine
a
dose,
be
it
an
aqueous
dose
or
a
dose
inside
the
organism,
that
gives
a
sense
of
sensitivity?
What
are
the
attributes
of
atrazine's
physical
chemical
properties
and
the
toxicokinetic
properties
that
go
on,
processes
that
go
on
that
help
us
extrapolate.

But
our
analysis
plan
doesn't
rely
just
on
­­
a
proposed
analysis
plan
doesn't
rely
just
on
Xenopus
and
then
modeled
into
other
species.

We
are
proposing
to
use
at
least
the
Northern
Leopard
Frog
as
a
North
American
species
to
take
a
look
to
see
if
we
get
some
consistency
in
an
atrazine
signal
in
a
North
American
species.

But
that's
where
sort
of
the
juggling
match
between
efficiencies
and
ability
to
get
on
with
the
question
at
hand.
Xenopus
we
can
use,

or
laboratories
can
use
all
the
time.
Most
labs
would
have
to
wait
for
spring
cycles
before
they
could
investigate
this
issue
in
the
Northern
Leopard
Frog
at
least.
So
it's
a
balance
in
terms
of
clarity
of
the
model
we're
using,
what
you
understand
about
you
biological
model,

what
you
understand
about
fundamental
toxicological
processes;
and
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
99
then
sort
of
your
ability
to
get
information
efficiently
and
effectively.

And
you
will
have
extrapolation
uncertainties
for
sure.

It
also
doesn't
preclude,
as
I
indicated
at
the
beginning
and
as
some
of
the
questions
at
the
end
of
our
last
session,
is
blending
what
you
know
from
a
controlled
world
of
toxicology,
the
reductionist
approach
to
sorting
these
things
out
and
how
does
that
get
blended
in
with
the
landscape
ecology
or
meta­
population
perspective
on
what
this
all
means.
And
I
think
the
ultimate
sort
of
risk
assessment,
and,

again,
it
depends
on
how
much
certainty
one
needs
to
make
a
decision
starts
to
blend
those
sciences
together
to
get
the
context
of
the
landscape
as
well
as
the
cross­
species
vulnerabilities.

DR.
ROBERTS:
Yes.
A
follow­
up
question
from
Dr.
Kelley.

DR.
KELLEY:
So
speaking
of
spring,
this
is
a
question
for
Joe.

What
Xenopus
are
you
planning
to
use?
Which
laevis
subpopulations?
You
know,
in
South
Africa
there
are
a
number
of
different
sub­
populations,
not
subspecies
just
sub­
populations,
that
have
different
breeding
seasons.

MR.
TIETGE:
I
haven't
given
much
thought
to
that.
I
think
there
is,
among
the
laboratories
in
the
United
States
anyway,
what
would
be
a
strain,
I
suppose,
that
is
commonly
used.
But
I
have
to
­­
I
don't
­­
I
haven't
given
much
thought
to
that.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
100
If
you're
proposing
especially
going
to
South
Africa
to
obtain
specific
strains?

DR.
KELLEY:
No.
I
think
I'd
be
against
that.
But
I'm
just
going
to
tell
you
that
the
groups
that
we
have
here
in
the
States
are
representative
of
a
sub­
population
of
laevis.
They
were
bred
originally
from
a
sub­
population;
and
they
retain
to
a
degree,

unsuspected
by
the
unwary,
an
androgenous
annual
circannual
rhythm,

for
instance,
in
hormone
production.
So
you
have
to
know
what
population
you're
dealing
with,
whether
you're
in
their
winter
when
they,
you
know,
when
one
population
will
breed
and
another
one
won't
or
there
summer
vice
versa.

So
I
think
it's
just
worth
bearing
in
mind
where
get
your
animals
from.

DR.
ROBERTS:
Any
other
questions
from
the
Panel?

If
not,
I
would
like
to
thank
Dr.
Bradbury,
Dr.
Steeger,
and
Dr.

Tietge
for
your
excellent
presentations
this
morning.
I
think
you've
given
us
a
clear
picture
of
the
Agency's
analysis
of
the
information
that's
available,
the
dilemma
that
lies
in
that
analysis,
and
your
thoughts
on
where
to
go
from
here.
So
your
presentations
were
very
helpful
for
the
Panel
and
your
answering
our
questions
was
very
useful
for
our
Panel
so
that
we
get
a
clear
understanding
of
what
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
101
Agency
proposes
to
do.

I'd
like
to
move
on
now
in
the
agenda.
The
next
item
is
public
comments.
We've
had
several
people
request
substantial
blocks
of
time
for
public
comments.
And
that
includes
the
first
public
commentor.
Rather
than
break
up
that
public
comment,
I
think
it
would
be
best,
my
preference
would
be,
to
begin
public
comments
after
lunch,
to
go
to
lunch
early,
come
back
early,
get
started,
and
begin
the
public
comments
then.

If
we
break
now
for
lunch
and
were
to
come
back
at
12:
30,
that
would
give
us
more
than
an
hour
for
lunchtime.
But
let
me
ask
our
first
scheduled
public
commentor,
which
is
the
Eco
Risk
Group,
if
they
could
be
ready
to
go
at
12:
30.

Is
Dr.
Kendall
here?
Not
to
put
you
on
the
spot.
Okay.
Great.

Let's
go
ahead
an
adjourn
now
for
lunch,
reconvene
promptly
at
12:
30.
We
will
begin
the
public
comments
then.

A
quick
announcement
from
the
DFO.

MR.
LEWIS:
Just
for
the
members
of
the
Panel,
this
room
will
be
open
during
lunch.
So
if
any
valuables,
please
take
them
with
you,

your
lap
tops
and
other
personal
belongings.
Thank
you.
See
you
at
12:
30.

[
Lunch
recess
taken
at
11:
25
a.
m.;
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
102
session
resumed
at
12:
30
p.
m.]

DR.
ROBERTS:
Let's
reconvene
the
meeting.
At
this
point
in
the
agenda,
the
Panel
would
like
to
listen
to
public
comments
on
these
issues.
And
we
have
several
individuals
or
groups
that
have
requested
the
opportunity
to
address
the
Panel
and
present
information.

Before
we
begin
the
public
comments,
I
would
like
to
remind
the
public
commentors
that
the
issues
that
we
are
focused
on
here
are
scientific
issues.
They
relate
to
a
very
specific
set
of
data
and
problems
and
issues
that
are
of
a
scientific
nature.
There
are,
of
course,
broader
issues
of
policy
and
so
forth.
But
those
are
really
outside
the
deliberations
of
this
Panel.
So
I
would
like
to
request
from
all
of
the
public
commentors
that,
when
they
address
the
Panel,

they
really
confine
their
comments
to
the
scientific
issues.

There
are
some
legitimate
policy
issues
and
points
to
be
made,

but
this
is
really
not
the
venue
to
make
those.
There
are
other
avenues
to
get
that
information,
those
viewpoints,
to
the
EPA.
So
if
you
could
in
fact
confine
your
comments
to
the
scientific
issues
that
the
Panel
is
trying
to
wrestle
with,
that
would
be
very
helpful
for
us.

With
that
being
said,
I
would
like
to
say
that
the
Panel
welcomes
public
comments
and
different
viewpoints
and
opinions
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
103
regarding
the
scientific
issues
that
we
face.
This
is
very
helpful
to
us
and
we
look
forward
to
hearing
from
you.

The
first
group
that
has
requested
the
opportunity
to
address
the
Panel
is
Eco
Risk.
And
they
are
represented
by
Dr.
Ron
Kendall.

Welcome,
Dr.
Kendall.

DR.
KENDALL:
Thank
you
very
much,
Dr.
Roberts.
First
of
all,
I
wanted
to
say
thank
you
for
the
opportunity
for
our
team
to
address
the
Panel
today
and
we
look
forward
to
providing
to
the
distinguished
members
of
the
SAP
and
you,
Mr.
Chairman,
some
perspectives
we
developed
over
a
number
of
years
now
on
the
response
of
amphibians
to
atrazine.

We're
here
as
an
eco
risk
panel,
but
we
really
are
university
scientists,
faculty
members
at
universities
across
the
nation
and
internationally.
I'm
going
to
introduce
my
colleagues
in
just
a
minute.

The
Eco
Risk
organization
has
led
a
facilitative
effort
to
bring
multiple
universities
together
and
multiple
members
under
an
opportunity
to
coordinate,
focus
efforts,
and
to
move
forward
in
what
we
have
felt
was
a
very
exciting
opportunity
to
engage
these
cutting­
edge
scientific
questions.

The
sponsor
has
been
Syngenta,
the
registrant
in
the
case
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
104
atrazine.
We
have
appreciated
their
support
and
their
willingness
for
the
Eco
Risk
panel
to
move
forward
in
a
open,
forthright
way
and
to
communicate
our
science
to
the
open
literature
as
well
as
here
at
the
SAP.
So
we
appreciate
that
support.

In
terms
of
the
this
afternoon,
we
appreciate
the
patience
of
the
SAP
in
giving
us
the
time
to
address
you,
we've
been
engaged
in
this
process
for
a
number
of
years
now.
And
we've
developed
a
core
presentation
which
is
before
you.
And
it
summarizes
our
efforts
and
tries
to
give
you
some
perspective
on
our
opinions
on
the
subject.

And,
of
course,
we
welcome
your
opinions
as
well.

Feel
free
to
ask
questions
at
any
time.
But
we
will
go
through
the
core
presentation
which
will
provide
you
a
summary
process
as
to
what
we've
been
through.
And
then
with
my
colleagues,
each
one
of
them
will
spend
a
few
minutes
summarizing
some
of
the
highlights
going
on
in
their
laboratories
with
their
graduate
students
and
post­
docs
and
so
on.
So
the
SAP
will
have
a
chance
to
discuss
with
each
faculty
member,
scientist,
at
the
table
what
has
been
their
contribution
to
some
of
the
current,
emerging
knowledge
that
we
have
on
this
particular
subject.

To
my
right
as
we
would
proceed,
Dr.
Glen
Van
Der
Kraak
will
give
the
core
presentation.
So
I'd
like
to
proceed,
Glen,
to
introduce
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
105
you
so
if
we
can
go
to
the
next.

The
panel
members
represent
a
variety
of
expertise
from
across
the
nation.
We've
drawn
as
well
on
consultants.
But
as
I
would
go
around
to
my
right,
Dr.
Glen
Van
Der
Kraak
is
professor
and
chairman
of
zoology
at
the
University
of
Guelph
and
the
co­
editor
of
the
document
that's
recently
been
published
called
"
Global
Assessment
of
the
State
of
the
Science
of
Endocrine
Disruptors,"
funded
through
WHO.
So
we've
asked
Glen
to
provide
the
presentation
and
the
core
presentation
at
least.

Next
to
my
right
is
Dr.
John
Giesy
from
Michigan
State
University.
Dr.
Giesy
is
the
university
distinguished
professor
with
various
appointments
there
and
well­
known
in
the
field
of
environmental
toxicology.

Dr.
Jim
Carr
is
in
biological
sciences
at
Texas
Tech
University.

He
has
published
on
the
subject
of
atrazine
and
amphibians
and
has
contributed
heavily
to
our
process.

Dr.
Ernest
Smith
is
engaged
with
the
Institute
of
Environmental
&
Human
Health
at
Texas
Tech
University.
He
is
a
reproductive
biologist
and
has
engaged
our
subject
area
in
environmental
toxicology
of
atrazine
from
the
reproductive
endpoint
perspective.

Dr.
Louis
Du
Preez
is
from
Potchefstroom
University
in
South
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
106
Africa,
a
member
of
the
School
of
Environmental
Sciences
and
is
an
expert
on
Xenopus
in
their
native
habitat.
And
we
have
been
engaged
in
field
studies
in
South
Africa.
Dr.
Du
Preez
will
report
on
the
results
of
that
work
that
has
engaged
the
panel
directive
in
working
with
him.

And
then
Dr.
Tim
Gross
from
the
University
of
Florida,
the
Caribbean
Science
Research
Center,
and
is
heavily
involved
in
amphibian
ecotoxicological
work
in
Florida
on
multiple
species
and
has
reported
for
the
panel
various
projects
over
the
years
involving
not
only
amphibians
but
fish
and
reptiles.

And
Dr.
Keith
Solomon,
last
but
not
least,
professor
at
the
University
of
Guelph,
well­
known
in
field
of
environmental
toxicology
and
risk
analysis.
And
this
is
our
team.

We've
also
had
others
that
we
have
participated
with
over
the
years.
Dr.
Tyrone
Hayes,
University
of
California
Berkeley,

participated
with
the
panel
the
first
three
years.
Resigned
in
November
of
2000
to
pursue
his
own
research.
Dr.
Bob
Silken
from
Silken
&
Associates
has
served
as
a
consultant
to
us
and
we
have
valued
his
contribution
as
we
have
engaged
the
statistical
interpretation
of
a
lot
of
these
questions,
both
from
a
field
standpoint
as
well
as
a
laboratory
standpoint.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
107
Again,
I
want
to
emphasize
that
the
purpose
of
this
scientific
panel
here
is
that
we
work
together.
We
have
moved
forward
in
a
way
in
which
we
design
projects
together.
We
don't
send
an
individual
faculty
member
out
with
some
graduate
students
and
they
go
do
the
work.
We
design
work
in
consultation.
We
meet
regularly.
We
have
conference
calls.
We
get
together.

And
we've
designed
through
our
standard
operating
procedures.

Which
I
might
add,
for
most
of
our
projects
they
may
not
be
totally
GLP,
but
they're
close
to
it.
Particularly
for
emerging
science
as
we
are
doing,
no
validated
protocols
are
in
place,
it's
kind
of
tough
to
put
a
GLP
study
together.
But
with
the
encouragement
with
our
sponsor
and
the
Eco
Risk
organization,
and
Ms.
Katherine
Vins
that
heads
up
the
Eco
Risk
QA
unit,
we've
been
able
to
move
in
that
direction.
So
all
of
our
procedures
do
have
standard
operating
procedures.
These
can
be
checked.
All
the
data
that
we've
developed
to
date
has
been
turned
into
the
Agency
for
full
and
complete
scrutiny.

So
never
the
less,
we
have
worked
as
a
team
to
combine
our
efforts
to
focus
on
research
as
we
envision
it
to
be
needed;
and
we
have
designed
and
implemented
these
projects
with
full
opportunity
to
freely
pursue,
publish,
discuss,
and
engage
our
graduate
students
as
necessary.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
108
So
from
my
scientific
perspective,
it's
been
a
great
opportunity.

And
I
think
every
one
of
our
panel
members,
whether
they
be
from
North
America
or
not,
would
agree
with
me.

So
we
would
first
like
to
proceed,
Mr.
Chairman,
with
the
presentation
of
our
core
presentation.
It
might
be
best
to
let
Dr.
Van
Der
Kraak
at
least
get
through
the
core
presentation
before
questions
because
he's
worked
very
hard
with
our
team
to
put
all
this
together.

And
then
let's
have
a
question
and
answer
period.
I'm
sure,
based
on
the
questions
this
morning,
there
will
be
lots
of
questions
from
this
panel.
Then
we
will
proceed
to
the
individual
investigators
if
that
will
meet
your
wishes.

DR.
ROBERTS:
Great.
That's
fine.
Let's
go
ahead
and
proceed
with
the
core
presentation
from
Dr.
Van
Der
Kraak.

DR.
KENDALL:
Thank
you.
Dr.
Van
Der
Kraak.

DR.
VAN
DER
KRAAK:
Thank
you.
Before
I
get
into
the
meat
of
the
presentation,
thank
you
very
much
for
the
opportunity
to
speak
on
behalf
of
the
panel
to
this
issue.

In
terms
of
chronology,
this
tends
to
set
out
some
of
the
activities
associated
with
the
panel.
In
1996,
EDSTAC
was
formed.

The
atrazine
endocrine
panel
began
its
work.
And
in
'
97
was
the
first
report
that
came
through
the
panel.
'
98
the
studies
began
in
earnest
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
109
on
fish,
reptiles,
and
frogs.
And
in
2001,
the
panel
began
further
studies
using
Xenopus
as
a
model
both
in
the
laboratory
and
in
the
field.

Associated
with
this,
there
were
other
activities
going
on
with
the
United
States
Environmental
Protection
Agency,
formation
of
the
Endocrine
Disruptor
Methods
Validation
Subcommittee,
to
try
to
bring
together
some
standardized
testing
protocols
for
looking
at
endocrine­
active
substances
across
the
spectrum
of
animals
from
humans
through
to
invertebrates.

Our
work
occurred
during
a
period
of
time
when,
as
Ron
had
mentioned,
there
were
no
standard
protocols
that
were
available.
We
then
shifted
our
focus
in
2002
to
begin
some
very
detailed
mechanistic
and
field
studies
in
North
America
along
with
activity
that
was
going
on
in
South
Africa.

We
produced
the
second
panel
report
that
was
available
to
the
United
States
Environmental
Protection
Agency.
We
followed
up
with
a
third
panel
report
in
2003.
And
this
is
all
again
occurring
in
the
backdrop
of
the
next
generation
of
EDSTAC
who
is
trying
Tier
2
endocrine
distruptor
tests
that
we
hope
we
will
contribute
at
least
in
some
small
way
to
providing
some
of
these
validated
methods
that
will
be
applicable
for
studies
with
amphibians.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
110
In
terms
of
the
panel
activities,
the
panel
activities
were
to
establish
and
direct
research
programs
in
multiple
laboratories
to
test
hypotheses
and
to
understand
mechanisms.
And
I'll
go
into
that
in
a
little
bit
more
detail.
In
terms
of
the
activity
that
we
do,
we
review
science,
we
integrate
and
evaluate
data
from
other
laboratories,
and
design
our
own
studies.
And
this
is
facilitated
in
part
through
the
preparation
of
reports
as
I
mentioned
were
made
available
to
the
U.
S.

Environmental
Protection
Agency.

The
overall
objective
of
our
program
is
to
assess
the
effects
of
environmentally
relevant
levels
of
atrazine
on
amphibians.
We
have
a
multi­
pronged
approach
that
you
can
put
under
two
broad
umbrellas.

One
being
studies
involving
what
we
might
call
field
studies.
The
other
in
laboratory
studies.
And
as
listed
there,
these
studies
encompass
a
number
of
different
species
some
of
which
are
native
to
North
America
and
are
environmentally
relevant
in
our
environment;

others
are
laboratory
surrogates
like
Xenopus,
but
we've
gone
and
done
the
unique
thing
of
studying
this
in
its
native
habitat.

We
also
study
some
introduced
species,
in
this
case
the
cane
toad
in
Southern
Florida,
because
it's
found
in
areas
where
there
is
overlap
with
potential
exposures
to
the
chemical
in
question.

This
is
just
a
brief
summary
of
some
of
the
reports
that
have
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
111
been
prepared
by
the
panel.
And
as
you
can
see,
our
focus
started
with
a
risk­
based
assessment
of
the
endocrine
system,
moved
into
a
more
global
evaluation
of
the
endocrine
system
in
non­
mammalian
vertebrates,
and
then
specifically
now
has
focused
some
of
our
attention
looking
at
the
questions
associated
with
amphibians.

If
you
look
at
what
were
some
of
the
highlights
of
these
reports,

these
focused
initially
on
some
of
the
traditional
endpoints
associated
with
the
ecotoxicological
potential,
the
ecotoxicological
effects
that
could
be
potentially
mediated
by
atrazine.
We
looked
at
and
identified
that
endocrine
and
reproductive
effects
had
not
been
specifically
addressed.
Where
they
had
been
studied,
they
had
been
looked
at
in
microcosm
and
in
full
life­
cycle
tests
and
these
tended
to
focus
on
responses
in
fish.

In
terms
of
reproductive
and
endocrine
effects,
test
guidelines
were
still
under
development.
We
needed
to
refine
and
optimize
assays,
and
we
needed
to
establish
a
framework
for
assessing
these
responses
given
that
standardized
protocols
were
not
available.

So
by
way
of
introduction,
I'll
get
into
the
meat
of
the
core
presentation.
And
the
core
presentation
has
an
overarching
question
of
trying
to
understand
the
ecological
effects
of
atrazine
on
amphibians.
In
order
that
we
could
accomplish
that
goal
of
increasing
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
112
our
understanding,
we
needed
to
have
an
approach
to
identify
causality.
And
part
of
that
came
through
the
development
and
implementation
of
the
weight
of
evidence
approach
to
the
question.

The
other
part
was
that
we
wanted
to
embrace
the
scientific
method.
And
so
we
developed
and
tested
a
suite
of
different
hypotheses
that
were
based
at
different
levels
of
biological
complexity.
Some
of
these
were
associated
with
effects
on
the
endocrine
system
through
modulation
of
various
endocrine
endpoints.

We
quickly
moved
to
try
to
look
at
very
specific
activities
at
the
tissue
level
and
looked
at
tissue
toxicity
through
studies
on
the
effects
on
the
gonad
and
the
larynx.
And
then
we
also
attempted
in
a
general
sense,
and
I
will
come
to
back
to
this
in
a
few
minutes,
of
trying
to
get
to
the
tough
question
of
what
might
be
some
of
the
population
level
impacts.

So
while
this
approach
may
look
like
we're
going
from
the,
I
guess
you
could
say,
bottom
up
or
the
top
down
depending
on
where
you
put
these,
we
were
trying
to
look
at
different
scales
of
biological
complexity.

In
terms
of
the
first
question
that
came
to
the
panel
was
how
do
we
evaluate
causality.
And
we
could
certainly
go
back
into
the
literature
and
we
could
identify
that
this
question
has
been
around
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
113
since
the
1800s.
There
have
been
a
very
strong
focus
on
this
from
a
human
health
epidemiological
perspective.
But
there
have
been
very
significant
developments
that
have
been
made
over
time
in
terms
of
looking
at
this
from
an
eco
perspective
and
looking
at
it
in
the
real
world
with
wildlife
species.

Some
of
the
pivotal
work
by
Glen
Fox
published
in
1991.
Other
work
by
Gary
Ankley
who
involved
some
of
the
members
of
our
panel.

But
it's
important
to
note
that
Gary
Ankley
was,
in
fact,
one
of
the
reviewers
of
the
White
Paper
that
is
before
the
SAP
today.

Now,
as
Dr.
Kendall
mentioned
a
few
minutes
ago,
I
was
very
fortunate
to
be
involved
for
a
period
of
about
three
years
in
work
sponsored
by
the
International
Programme
on
Chemical
Safety
that
ended
up
with
the
publication
of
a
book
that
was
the
Global
Assessment
of
the
State
of
the
Science
Associated
with
Endocrine
Disruptors.

Why
do
I
bring
this
up
here?
Because
the
causal
criteria
that
was
developed
through
this
document
is
the
very
criteria
that
we
have
tried
to
use
in
trying
to
evaluate
the
potential
effects
of
atrazine
on
amphibian
populations.

Now,
I
put
this
up
as
if
it's
my
own
work.
In
fact,
it's
not
my
own
work.
I
was
a
member
of
a
very
strong
team
of
people
that
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
114
included
representatives
from
over
32
­­
there
was
32
different
international
experts
that
were
associated
with
the
construction
of
this
document.
Some
of
those
members
are
in
fact
members
of
the
Eco
Risk
atrazine
panel.
Others
are
members
of
the
U.
S.

Environmental
Protection
Agency.
But
I
think
the
key
point
that
I
wish
to
make
here
is
that
this
was
a
world­
wide
perspective
on
trying
to
develop
a
criteria
document
for
evaluating
the
potential
effects
of
endocrine
disrupting
chemicals.

This
slide
talks
to
the
fact
that
there
are
a
number
of
mechanisms
in
which
chemicals
could
be
having
effects
on
development
and
endocrine
processes
in
amphibians
or
in
other
vertebrates
for
that
matter.
There
were
direct
effects
where
compounds
could
act
as
hormone
mimics
or
antagonists.
Indirect
effects
associated
with
changes
in
the
hormone
titer,
effects
directly
on
tissue
development
such
as
effects
on
gonadal
development.

And
when
you
look
back
at
where
we
were
as
a
panel
about
three
years
ago,
we
were
left
with
the
starting
point
that
there
was
very
little
in
the
way
of
responses
that
were
evident
in
amphibians.

The
one
response
that
was
before
us
was
a
potential
effect
of
laryngeal
development
in
amphibians
associated
with
exposure
to
atrazine.
And
there
were
discussion
at
the
time
that
this
possible
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
115
mechanism
may
involve
interference
with
the
androgen
and
estrogen
titer,
and
it
was
the
focus
of
some
specific
hypothesis
testing
that
we
could
develop
that
was
focused
not
only
on
just
androgen
estrogen
titers
but
responses
associated
with
changes
in
aromatase
activity.

As
a
panel,
we
developed
a
main
hypothesis.
And
the
main
hypothesis
is
probably
a
little
longer
than
what's
written
there.
But
what
we
were
interested
in
was
the
question
of
whether
exposures
to
environmentally
relevant
concentrations
of
atrazine
caused
adverse
effects
on
endocrine
function
in
amphibians.
And
by
endocrine
function,
we
mean
that
in
the
broadest
sense
­­
changes
in
endocrine
function,
growth,
reproduction,
and
development
­­
all
components
that
are
under
the
control
of
the
endocrine
system.

This
enabled
us
to
develop
a
series
of
sub
hypotheses
that
I'm
going
to
go
through
in
some
degree
of
detail
in
the
next
series
of
slides.
But
we're
going
to
look
at
whether
these
effects
could
be
mediated
through
estrogen­
dependent
mechanisms,

androgen­
dependent
mechanisms,
effect
on
the
thyroid
systems,
direct
effects
on
the
gonad,
and
then
the
potential
that
there
may
be
affects
on
the
population
level
in
exposed
amphibians.

Now,
I
put
this
slide
up
to
try
to
remind
everyone
that
the
endocrine
system
in
amphibians
is
designed
in
the
same
manner
and
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
116
the
same
fashion
as
it
is
in
all
other
vertebrates.
There
is
a
hypothalamic
pituitary
target
organ
regulation
of
endocrine­
dependent
processes.
And
this
slide
helps
to
illustrate
some
of
the
potential
targets
for
which
there
may
be
effects
that
we
could
develop
hypotheses
around.

One
of
the
initial
hypothesis
that
we
were
interested
in
was
whether
or
not
there
were
effects
on
the
titer
estrogens
and
effects
mediated
through
changes
in
aromatase
activity.
We
were
very
interested
in
whether
there
were
effects
on
androgen
levels
and
effects
mediated
through
the
androgen
receptor.
We
were,
of
course,

interested
in
whether
there
were
effects
on
the
thyroid
system
because
of
their
important
developmental
role
in
amphibians.

We
were
also
interested,
of
course,
on
some
of
the
apical
endpoints
and
whether
there
were
responses
associated
with
secondary
sex
characteristics,
laryngeal
growth,
gonadal
growth
as
an
example.
And
then
we'll
try
to
translate
this
to
higher
levels
of
biological
complexity
by
looking
ultimately
at
population­
level
responses.

So
if
we
go
into
the
hypotheses
that
we've
considered,
the
first
hypothesis
we
considered
was
whether
atrazine
caused
adverse
effects
in
amphibians
through
estrogenic
or
anti­
estrogen­
mediated
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
117
mechanisms.

Now
the
next
slide
illustrates
a
number
of
the
endpoints
that
we
considered
in
doing
these
analyses.
And
for
each
of
the
hypotheses,

I'm
going
to
have
a
similar
format.
And
in
this
case,
the
kinds
of
endpoints
that
we
considered
were
binding
to
the
estrogen
receptor
changes
in
the
amount
of
circulating
estrogens,
inductions
of
aromatase,
and
whether
or
not
the
responses
that
were
induced
in
studies,
both
in
the
lab
and
in
the
field,
could
be
correlated
with
the
responses
that
we
saw
to
estrogen
exposure.

I
won't
go
through
all
of
the
details
of
these
responses
or
the
conclusions
to
the
studies
as
these
are
going
to
be
the
focus
of
the
remaining
slides.
But
to
cut
to
the
chase,
the
conclusion
for
this
section
was
that
it's
highly
unlikely
that
atrazine
could
be
exerting
effects
through
mechanisms
that
are
involved
with
estrogen.

An
obvious
question
being
does
atrazine
exert
its
effects
through
binding
to
the
estrogen
receptor.
To
our
knowledge,
this
is
not
been
explicitly
tested
in
amphibians;
but
there
is
an
extensive
literature
available
that
suggests
that
atrazine
does
not
bind
with
any
significant
affinity
to
the
mammalian
estrogen
receptor.
Given
the
high
degree
of
homology
between
these
receptors
across
classes,
we
don't
expect
that
this
is
an
issue
we
need
be
concerned
about.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
118
In
terms
of
changes
associated
with
atrazine
and
affecting
plasma
estrogen
titer,
some
of
the
initial
studies
looked
at
both
Xenopus
and
the
green
frog
exposed
throughout
development
in
lab
conditions,
and
there
were
no
effects.
Similarly,
we
saw
no
effects
on
Xenopus
laevis
adults
exposed
in
the
laboratory
for
periods
of
up
to
47
days.
There
were
some
indications
from
field
studies
that
there
was
a
negative
correlation
between
estrogen
titers
and
triazines
under
field
conditions.
And
in
other
studies
looking
at
the
cane
toad,
adults
exposed
to
atrazine
under
field
conditions,
again,
we
saw
no
significant
effects.

Just
to
give
you
an
example
of
the
kinds
of
data
that
we
saw
in
these
types
of
experiments,
this
is
the
result
of
an
experiment
that
was
conducted
by
a
post­
doc
in
John
Giesy's
lab
at
Michigan
State
University.
And
Dr.
Hecker
showed
that
exposure
to
atrazine
caused
no
specific
concentration­
related
response.
Of
the
various
doses
that
were
tested,
only
one
dose
caused
a
reduction
in
estradiol
concentration
in
the
plasma.

And
this
finding,
coupled
with
the
other
responses
that
we
had
failed
to
show
a
response
in
terms
in
changes
in
estradiol
titer,

suggested
to
us
that
this
was
not
a
particularly
robust
response
and
certainly
one
that
was
difficult
to
envisage
from
a
mechanistic
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
119
standpoint.

In
this
slide
here
on
the
left­
hand
side
we
slow
the
kinds
of
responses
that
you
see
when
you
expose
animals
to
estradiol
in
the
water.
And
not
surprisingly,
if
you
put
estradiol
in
the
water,
the
amounts
of
estrogen
in
the
blood
go
up.
So
that's
very
much
what
one
would
expect.

Now
we
took
these
studies
and
we
conducted
some
of
these
actually
in
the
field
situation
in
South
Africa.
The
triangle
on
this
slide
illustrates
the
main
corn­
growing
area
of
South
Africa.
The
red
dot
here
indicates
the
study
site,
and
that
is
in
the
vicinity
of
Potchefstroom
where
Dr.
Louis
Du
Preez,
a
member
of
our
team,
is
a
faculty
member.

This
is
not
Kansas.
This
is
South
Africa.
And
this
is
a
picture
of
a
corn
field
in
the
corn­
growing
area.
And
if
you
look
at
the
nature
of
the
soil
type
in
the
area,
this
soil
type
is
particularly
sandy.
As
a
result
of
that,
there
is
the
rapid
movement
of
any
chemical
that's
put
on
fields
in
this
area.
And
this
raises
the
potential
may
well
get
into
receiving
environment
into
ponds
that
would
be
the
likely
home
of
native
amphibians
in
South
Africa.

And
what
we
were
able
to
do
was
initiate
a
series
of
studies
in
which
we
looked
specifically
at
amphibian
populations
living
in
these
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
120
ponds.
These
are
two
of
the
experimental
sites
in
the
corn­
growing
area
E1
and
E8.
And
this
green
here
represents
corn
fields
that
grow
essentially
right
up
to
the
edge
of
the
farm
ponds.

When
we've
done
these
studies
and
we've
gone
out
and
measured
circulating
levels
of
hormones,
in
this
case
these
are
estradiol
levels
in
males
and
females.
And
in
this
case,
we
found
a
significant
reduction
in
circulating
estradiol
levels
in
corn­
growing
areas
in
both
the
male
and
the
females.

I'd
like
to
leave
you,
though,
with
a
couple
of
points
associated
with
this
slide.
The
first
slide
or
the
first
thought
is:
Is
that
if
indeed
this
response
of
atrazine
was
associated
with
an
induction
of
aromatase
activity,
this
would
be
contrary
to
what
one
might
predict.

The
second
issue
is,
is
that
in
sampling
populations
of
frogs
in
this
area,
you
have
frogs
at
various
stages
of
sexual
maturity.
And
it's
quite
clear
that
if
you
look
at
the
range
and
the
variance
of
the
data,

there
is
considerable
overlap
and
certainly
it's
very
difficult
to
partition
out
responses
that
one
might
immediately
attribute
to
exposure
to
atrazine.

Now,
a
considerable
amount
of
attention
has
been
paid
to
the
question
of
whether
or
not
atrazine
has
effects
on
estrogen
titer
through
induction
of
aromatase
activity.
The
first
discussion
of
this
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
121
point
came
in
a
publication
by
Hayes
where
he
hypothesized
that
atrazine
would
be
induced
in
frogs
exposed
to
atrazine.
In
this
situation,
there
was
no
data
that
was
provided.
But
in
subsequent
studies
conducted
by
the
panel,
we
found
no
effect
on
atrazine
activity
levels
in
Xenopus
exposed
through
development
in
the
laboratory,
both
in
juveniles
and
in
adults.
And
in
field
studies,
we
showed
no
correlation
with
triazine
levels
in
adults
that
were
collected
again
under
field
conditions.

This
illustrates
some
of
the
data
that
has
been
associated
with
our
evaluation
of
aromatase
activity
in
the
gonad
of
Xenopus.
If
you
look
in
the
first
instance
on
the
right­
hand
side
of
the
panel
here,
we
see
a
marked
sexual
dimorphism
in
the
total
amounts
of
aromatase
activity
in
the
gonad.
That's
not
unexpected.
But
we
see
no
concentration­
dependent
effect
of
atrazine
on
aromatase
activity
levels.

The
contrast
to
this,
if
you
look
at
the
data
on
the
left­
hand
side
of
the
panel,
if
we
expose
the
animals
to
estradiol,
at
least
in
the
females,
we
induce
a
significant
reduction
in
ovarian
aromatase
activity
in
the
females.

Again,
back
to
the
original
hypothesis
is
atrazine
acting
like
an
estrogen.
In
this
case,
we're
seeing
no
evidence
that
atrazine
is
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
122
having
that
type
of
a
response.

We
also
tried
to
compare
the
effects
of
atrazine
to
those
associated
with
estradiol.
Now
the
first
of
the
points
that
we
have
up
there
was
the
summary
of
some
work
that
was
done
a
number
of
years
ago
by
Dr.
Tyrone
Hayes
and
his
associate,
looking
at
the
possibility
that
atrazine
would
affect
sexually
dimorphic
characters,
that
would
be
coloration
in
a
frog
species
called
Hyperolius.
And
in
this
case,

it's
my
understanding
that
they
found
no
response.
As
well
as
I
mentioned
the
earlier
slide,
atrazine
also
did
not
mimic
the
effects
of
estradiol
on
sex
ratio
in
Xenopus.

So
I
guess
the
question
is
where
are
we
in
terms
of
this
overall
hypothesis
and
using
this
weight
of
evidence
criteria.
In
terms
of
temporality,
we
have
little
indication
of
data
that
we
can
apply
in
that
context.
But
when
we
look
in
terms
of
the
other
key
components
of
the
weight
of
hypothesis
testing
framework
in
terms
of
strength
of
association,
consistency,
or
biological
plausibility,
there's
little
evidence
to
support
that
type
of
a
mechanism.
And
we're
left
with
the
overall
summary
that
there's
little
evidence
to
support
the
concept
that
atrazine
has
affects
in
amphibians
through
either
estrogenic
or
anti­
estrogen
mediated
processes.

The
second
hypothesis
that
we
considered,
and
this
was
one
that
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
123
was
brought
up
in
question
earlier
today
to
the
Science
Advisory
Panel
to
the
U.
S.
EPA,
was
whether
or
not
atrazine
may
be
exerting
it's
effects
in
amphibians
through
androgenic
receptors
through
acting
as
either
an
androgen
agonist
or
an
androgen
antagonist.

So
we've
considered
this
by
both
empirical
studies
and
by
looking
at
the
literature.
And
we
looked
at
a
variety
of
endpoints
including
binding
to
the
androgen
receptor,
changes
in
androgen
receptor
type,
androgen
titer,
comparing
the
responses
to
DHT,
and
looking
at
whether
or
not
atrazine
mimics
the
effects
of
androgens
on
androgen­
dependent
processes.

So
to
summarize
some
of
these
data,
in
terms
of
does
atrazine
bind
to
the
androgen
receptor,
to
our
knowledge
this
is
not
been
tested
in
amphibians.
But
if
we
look
at
the
extensive
literature
that's
available
for
mammals,
there
is
little
evidence
to
suggest
that
atrazine
binds
to
the
mammalian
androgen
receptor.

In
terms
of
androgen­
dependent
gene
activation,
this
has
not
been
specifically
tested
in
amphibians.
But
again,
the
results
are
negative
in
mammals.
And
given
the
homology
of
receptors
across
those
species,
we
don't
anticipate
that
there
is
an
issue
here
that
we
have
to
be
concerned
with
in
the
immediate
term.

In
terms
of
effects
of
plasma
androgen
titers,
in
terms
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
124
looking
at
laboratory
studies,
we
see
very
different
types
of
responses.
In
work
that
has
been
conducted
by
the
panel,
there
were
no
effects
in
Xenopus
or
in
the
green
frog
exposed
during
development
in
the
laboratory.
We
also
saw
no
effects
in
adults
following
exposure
in
the
laboratory
to
atrazine.
This
is
in
contrast
with
a
study
that
was
produced
and
published
by
Dr.
Tyrone
Hayes,
in
which
he
showed
a
significant
reduction
in
plasma
testosterone
levels
in
adults
Xenopus
laevis.

We've
also
looked
at
this
under
field
conditions.
And
here
we
see
a
variety
of
different
types
of
responses
that
are
not
consistent.

In
terms
of
studies
with
the
cane
toad,
we
saw
no
effects
of
plasma
androgen
levels
following
collection
in
reference
and
atrazine­
exposed
locations.
In
terms
of
studies
that
were
conducted
in
South
Africa,
there
was
a
correlation
with
lower
T
levels
associated
with
the
exposure
to
one
of
the
metabolites,
DACT,
but
not
to
atrazine
or
trebuthylazine
under
field
conditions.
And
in
the
female,
we
did
see
a
negative
correlation
between
concentrations
of
atrazine,
trebuthylazine,
and
their
metabolites.

To
illustrate
some
of
these
data,
these
are
the
responses
that
were
observed
in
Xenopus
laevis
collected
in
South
Africa
from
both
reference
and
corn­
growing
locations
in
males.
We
did
not
see
a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
125
significant
change
in
the
median
androgen
titers
in
the
blood.
In
contrast,
we
did
see
a
significant
reduction
in
females.
But,
again,

the
levels
of
androgens
in
the
plasma
of
these
frogs
are
highly
variable
and
it
makes
interpretation
of
these
data
somewhat
difficult.

Now,
there
was
an
initial
report
that
suggested
that
one
of
the
androgen­
dependent
processes
that
occurs
in
frogs,
that
being
the
size
of
the
laryngeal
dilator
muscle,
may
well
be
affected
by
atrazine.

And
this
was
a
study
that
was
produced
Dr.
Tyrone
Hayes.
And
he
reported
a
decrease
in
laryngeal
dilator
muscle
using
cross­
sectional
area
as
the
indicator.

In
three
other
studies
involving
Xenopus,
members
of
the
panel
have
failed
to
show
an
effect
of
atrazine
on
laryngeal
dilator
muscle
size.
And
if
we
test
the
original
hypothesis
that
atrazine
may
well
be
functioning
as
an
androgen­
receptor
agonist,
we
would
anticipate
that
atrazine
would
mimic
the
effects
of
DHT.
And,
in
fact,
in
our
studies,

we
have
consistently
failed
to
show
that
atrazine
mimics
the
effects
of
DHT
on
the
size
of
the
laryngeal
dilator
muscle.

Now,
here
I
have
redrawn
some
work
that
came
from
Dr.
Tyrone
Hayes.
And
he
reported
in
the
Proceedings
of
the
National
Academy
of
Science
a
response
in
males
such
that
exposure
to
concentrations
of
atrazine
at
the
highest
doses
caused
a
significant
depression
of
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
126
size
of
the
laryngeal
muscle
in
terms
of
cross­
sectional
area.

Interestingly,
he
showed
a
similar
kind
of
trend,
although
not
a
statistically
significant
response
in
females.
And
this
was
an
interesting
observation
and
one
which
the
panel
was
very
interested
in
trying
to
see
whether
it
would
hold
up
under
other
studies.
But
this
reduction
is
not
synonymous
in
our
minds
with
an
effect
that
would
likely
be
mediated
by
an
induction
of
aromatase.

In
the
kinds
of
studies
that
we've
conducted
as
a
panel,
these
are
some
work
from
Jim
Carr,
published
in
2003
in
Environmental
Toxicology
and
Chemistry,
we
see
in
the
course
of
our
study
at
Stage
66
that
the
sexual
dimorphic
response
is
quite
evident,
suggesting
that
they
are
responding
to
androgenous
hormones.
But
we
see
no
dose­
related
affects
of
atrazine
in
males
or
in
females.
By
comparison,
if
we
do
treat
these
animals
with
DHT,
we
see
the
anticipated
and
expected
rise
in
the
size
of
the
laryngeal
dilator
muscle.

In
other
studies
conducted
at
Michigan
State
University
looking
at
this
endpoint
in
terms
of
responses
looking
at
atrazine,
we
saw
no
significant
differences
associated
with
the
size
of
the
laryngeal
muscle
in
males
or
in
females.
But
once
again,
the
positive
control
of
DHT
had
a
clear
stimulatory
effect
on
the
size
of
the
muscle.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
127
When
we
looked
at
Xenopus
from
the
field
situation
in
South
Africa,
we
did
these
data
a
little
bit
differently.
We
tried
to
co­
vary
this
with
the
weight
of
the
frog.
And
so
we
calculated
a
larynx­
somatic
index,
the
weight
of
the
dilator
muscle
and
the
associated
cartilage
versus
the
body
weight
of
the
frog.
Again,
we
saw
a
clear
sex­
related
difference
in
both
the
references
areas
and
the
corn­
growing
areas.
But
there
was
no
association
with
whether
the
animals
were
collected
in
reference
or
corn­
growing
locations.

So
in
terms
of
this
second
hypothesis,
if
we
looked
at
whether
or
not
there
was
evidence
to
support
the
conclusion
that
atrazine
effects
or
exerts
effects
through
androgen­
mediated
processes,
the
evidence
was
either
there
was
no
evidence
available;
or
where
there
was,
the
evidence
was
scant
and
certainly
not
indicative
of
a
robust
type
of
response.

The
third
hypothesis
that
we
considered
was
one
of
whether
or
not
atrazine
would
exert
it's
effects
through
influences
on
the
thyroid
hormone
system.
And
this
was
an
obvious
hypothesis
to
us
given
the
importance
of
the
thyroid
in
mediating
both
metamorphosis
and
what
is
known
across
vertebrates
in
terms
of
the
permissive
effects
of
thyroid
hormones
on
other
aspects
of
development
such
as
gonadal
development.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
128
So
we
were
interested
in
whether
or
not
there
were
changes
in
thyroid
hormone­
mediated
responses.
And
the
bottom
line
here
is
that
atrazine
does
not
appear
to
affect
metamorphosis.
In
terms
of
an
obvious
other
place
to
look,
would
be
whether
or
not
atrazine
had
effects
on
plasma
thyroid
hormone
titer.
There
is
no
information
available
at
this
time.

So
binding
to
the
thyroid
hormone
receptor
has
not
been
tested
in
amphibians.
And
in
terms
of
effects
of
thyroid
hormone­
dependent
gene
activation,
there
was
no
effect
on
metamorphosis
in
a
suite
of
different
studies
using
a
range
of
species
including
Xenopus,
the
green
frog,
and
the
Leopard
Frog.

So
in
terms
of
the
bottom­
line
conclusion
for
thyroid­
mediated
response,
we
see
no
evidence
that
atrazine
affects
thyroid­
mediated
processes
in
amphibians.
And
this
conclusion
falls
well
in
line
with
the
conclusions
that
are
coming
out
in
terms
of
the
mammalian
literature.
Again,
in
mammals,
there
is
no
indication
that
atrazine
is
affecting
thyroid­
dependent
processes.

The
fourth
hypothesis
and
one
that
has
been
the
focus
of
much
of
the
attention
of
the
panel,
but
also
of
the
discussions
today,
was
whether
or
not
atrazine
causes
adverse
effects
on
gonadal
development
in
amphibians.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
129
And
in
this
regard,
we
focused
our
evaluations
on
both
testicular
morphology
and
ovarian
morphology
and
development.
The
bottom
line
in
these
studies
is,
that
if
you
look
across
the
literature,
if
there
is
a
sex
that
has
the
potential
to
be
affected,
it
is
more
likely
the
males.
But
in
this
situation,
it
is
clearly
a
variable
type
response.

And
I'll
try
to
highlight
some
of
those
differences
in
various
studies
that
have
been
evaluated.

In
terms
of
the
effects
on
testicular
development,
the
kinds
of
endpoints
have
included
­­
well,
first
of
all,
there's
been
a
variety
of
endpoints
that
have
been
evaluated.
In
terms
of
the
ones
that
I'm
going
to
highlight
on
this
slide
was
in
terms
of
effects
both
hermaphrodism
and
on
the
presence
of
discontinuous
testes
or
breaks
of
the
structure
of
the
testes.

Dr.
Hayes
reported
in
the
PNAS
paper
that
there
was
an
induction
of
both
of
these
events
at
doses
greater
than
or
equal
to
0.1
microgram
per
liter.
The
work
done
by
the
panel
showed
aspects
of
similar
responses,
but
at
doses
that
were
about
250­
fold
higher
in
concentration.
And
in
other
studies
using
Xenopus,
there
were
no
effects
in
field
and
microcosm­
exposed
populations
in
South
Africa.

And
there
was
no
effects
in
the
laboratory
study
conducted
at
Michigan
State
University.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
130
There
was
one
other
study.
And
that
was
the
study
by
Tavera­
Mendoza
and
colleagues
who
reported
a
decrease
in
testicular
volume
in
Xenopus.
They
showed
this
response
at
21
micrograms
per
liter.
But
we
have
serious
questions
and
reservations
about
that
study
as
there
is
inconsistency
between
the
published
work
and
the
replicate
experiments
that
are
reported
in
the
thesis
describing
the
entirety
of
the
work
conducted
in
that
laboratory.

In
other
studies,
there
was
no
effect
at
doses
less
than
or
equal
to
25
micrograms
per
liter
in
the
green
frog
in
work
by
Hecker.
And
then
there
was
study
by
Hayes
showing
that
there
was
in
increase
in
hermaphrodism
in
Rana
pipiens,
the
Leopard
Frogs;
but
it
was
an
inverse
concentration
response
that
was
somewhat
difficult
to
interpret.

There
were
other
studies
that
have
looked
at
hermaphrodism
in
frogs.
And
this
turns
out
that
this
has
been
a
response
that
has
been
observed
for
decades.
There
was
response
indicating
that
hermaphrodism
does
occur
in
other
frogs
well
prior
to
the
use
of
atrazine.
And
in
the
cricket
frogs,
there
was
a
clear
indication
of
intersex
in
a
retrospective
study
that
evaluated
museum
specimens.

In
terms
of
the
types
of
responses
that
have
been
seen,
these
are
some
work
from
Jim
Carr's
studies
that
were
published
in
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
131
Environmental
Toxicology
and
Chemistry.
In
terms
of
discontinuous
testes,
there
was
an
increase
associated
with
atrazine
exposure
but
only
at
the
highest
dose
both
in
terms
of
discontinuous
testes
and
intersex.
If
we
treated
with
DHT
or
estradiol,
we
saw
no
effect
in
terms
of
discontinuous
testes
but
an
induction
by
estradiol
of
an
increase
in
the
proportion
of
intersex.

In
other
studies
that
were
conducted
by
Hecker
and
associates
at
Michigan
State
University,
we
saw
no
significant
differences
associated
with
exposure
to
various
doses
of
atrazine
in
terms
of
looking
at
discontinuous
gonads,
mixed­
sex
gonads,
size
irregularities,
intersex,
or
other
anomalies.
So,
clearly,
there
is
a
discordance
between
different
laboratories
in
terms
of
types
of
responses
that
are
seen
in
terms
of
testicular
development.

In
a
field
study,
this
was
work
conducted
in
Iowa
in
which
there
was
an
evaluation
of
various
­­
pardon
me.
This
is
work
from
South
Africa.
And
this
was,
again,
looking
at
Xenopus
from
areas
which
were
references
sites
and
corn­
growing
sites.
And
here
a
serological
evaluation
was
done
to
look
at
the
distributional
volume
of
different
cell
types
within
a
microscopic
field,
looking
at
spermatogonia,

spermatocytes,
sperm,
blood
vessels,
and
other
cell
types.
And
we
could
differentiate
no
difference
in
this
distribution
of
cell
types
from
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
132
both
corn­
growing
and
reference
locations.

If
we
switch
gears
a
little
bit
and
we
look
at
what
happens
in
terms
of
some
of
the
types
of
responses
that
we
see
in
amphibians,

this
is
an
interesting
observation
that
we
and
now
others
have
clearly
made.
And
that's
the
presence
of
an
oocyte
that
is
found
growing,
not
necessarily
growing,
found
present
in
the
vicinity
of
what
appears
to
be
normal
testicular
tissue.
In
this
case,
we
have
an
oocyte
with
multiple
nucleoli.
We
have
a
development
of
epithelial
cell
layer.

And
this
testicular
oocytes
are
turning
out
to
be
almost
a
ubiquitous
feature
of
the
development
of
amphibians.

If
you
look
across
a
range
of
studies,
various
authors,
Hayes,

Smith,
Du
Preez,
Hecker,
and
others
going
back
to
Witschi
in
the
1920s,
have
identified
that
there
are
testicular
oocytes
that
are
present
in
amphibians.

In
terms
of
whether
this
response
is
associated
with
atrazine,

there's
one
paper
suggesting
that
these
occur
in
association
with
exposure
to
atrazine.
That
being
the
Hayes
work.
But
the
other
studies
show
that
these
are
present
at
all
doses
associated
with
both
reference
and
doses
lower
than
30
or
25
micrograms
per
liter.

Just
to
give
you
an
example
of
the
kinds
of
responses
that
people
have
seen,
this
is
some
work
on
testicular
oocytes
in
Xenopus
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
133
that
were
exposed
as
juveniles
until
Stage
66
and
then
allowed
to
grow
up
or
grow
out
from
there.
When
we
looked
at
these
data
in
terms
of
the
incidence
of
testicular
oocytes
or
of
intersex,
there
was
no
significant
difference
in
that
distribution
associated
with
exposure
to
atrazine.

Now,
I
mentioned
earlier
that
if
we
switch
and
look
at
the
opposite
sex
and
we
look
at
responses
in
females,
the
general
scheme
of
things
is
that
there's
no
evidence
that
there
are
affects
on
ovarian
morphology
in
Xenopus
associated
with
exposure
to
atrazine.
Hayes
in
his
work
showed
no
effect
at
doses
up
to
200
micrograms
per
liter.

Others
at
the
highest
doses
that
they
looked
at
25,
approximately
30,

in
the
field
studies,
or
25
in
the
Michigan
State
study
by
Hecker,

showed
no
effect.

The
one
study
that
seems
to
be
contrary
to
this
is
the
Tavera­
Mendoza
in
a
second
paper.
And
this
group
reported
that
associated
with
exposure
to
atrazine,
there
was
a
reduction
in
the
number
of
primary
oocytes
but
actually
an
increase
in
the
number
of
secondary
oocytes.
An
interpretation
of
this
would
be
that
atrazine
is
actually
promoting
ovarian
development.

Again,
we
have
some
concerns
about
this,
that
the
replication
between
the
published
study
and
what's
reported
in
the
thesis
is
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
134
certainly
not
there
and
it's
not
consistent.

In
other
studies
where
investigators
have
looked
in
a
number
of
ranids,
again,
there
was
no
effect
in
terms
ovarian
development.

So
in
terms
of
evaluating
whether
or
not
atrazine
exposure
is
associated
with
effects
on
gonadal
development,
in
terms
of
the
temporality,
there
seems
to
be
a
very
serious
question
associated
with
causality
in
that
these
responses
were
present
well
before
the
introduction
of
atrazine
to
the
marketplace.
In
terms
of
the
strength
of
association,
there
is
some
evidence
of
responses.
But
it's
an
inconsistent
concentration
response.

In
terms
of
the
consistency
where
there
are
concentration
responses,
these
are
typically
not
dose­
related.
There's
clearly
some
indication
that
there
may
be
some
effects
that
are
occurring.
There's
little
evidence
to
indicate
that
those
are
severe
effects.
But
at
this
point
in
time,
we
have
little
or
no
evidence
in
terms
of
the
mechanism
that
may
be
contributing
to
these
types
of
responses.

So
our
overall
assessment
here
is
that
there's
little
evidence
the
atrazine
affect
gonadal
development
in
male
frogs.

The
last
hypothesis
is
one
that
we
very
much
wished
to
get
to.

And
that
is
to
address
the
question
­­
some
of
you
may
call
it
the
select
question.
And
that
was
whether
or
not
atrazine
causes
adverse
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
135
effects
at
the
population
level
in
exposed
amphibians.
And
so
the
kinds
of
endpoints
can
be
and
have
been
considered
in
this
regard
are
looking
at
the
abundance
of
species,
age­
size
class
distributions
as
examples.

And
the
conclusion
to
these
studies
is
relative
to
causality,

there's
little
evidence
of
effects
linked
to
atrazine
exposure.

Now,
I'll
go
through
this
slide,
but
I'd
to
remind
those
ecologists
in
the
group
that
there
is
a
caveat
coming
on
the
next
slide.

When
we've
looked
at
population
responses,
the
kinds
of
responses
that
we've
measured
are
there
are
robust
populations
and
there
are
no
differences
in
age­
size
class
distributions
of
Xenopus
in
corn­
growing
and
reference
sites
in
South
Africa.
When
others,

Hayes,
has
looked
at
the
Leopard
Frog
across
a
range
of
atrazine
exposures,
he
found
robust
populations.
When
we
looked
in
South
Florida,
we
found
much
higher
populations
of
the
cane
toad
in
areas
that
were
associated
with
sugar
cane
production
which
would
have
higher
exposure
to
atrazine
than
our
reference
locations.

And
when
we
looked
at
the
bullfrog
across
a
range
of
atrazine
exposures
in
Southern
Iowa,
there
were
numerous
individuals
in
what
appeared
to
be
robust
populations.
But
my
caveat
for
the
ecologists
in
the
group
was
that
few
studies
have
been
undertaken
to
explicitly
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
136
address
the
question
of
whether
there
were
adverse
population
level
impacts
associated
with
atrazine
exposure
in
amphibians,
at
least
in
native
amphibians.

Now,
just
a
couple
of
slides
here
so
we
don't
think
all
of
the
work
occurs
in
the
lab.
This
is
Louis
Du
Preez
taking
an
oxygen
meter
sample
from
a
study
site
in
South
Africa.
These
are
the
traps
that
we
used
to
collect
Xenopus.
These
are
weighed
down
and
put
underwater
because
Xenopus
is
obviously
an
aquatic
species.
We
then
collect
the
frogs
in
the
traps.
We
then
can
do
mark,
recapture,

and
release
studies.

And
when
we
do
these
kinds
of
things,
these
are
types
of
data
that
we
have
seen.
And
these
are
looking
at
both
reference
and
corn­
growing
locations
in
South
Africa.
And
the
various
colors
on
there
represent
various
age
classes.
In
terms
of
statistical
evaluation,

there
was
no
difference
in
the
proportion
of
different
age
classes
across
the
reference
and
corn­
growing
sites.
And
if
you
look
at
those
population
structures,
you've
got
a
blend
of
young
and
old
frogs
in
all
of
the
locations.

So
in
terms
of
our
overall
evaluation
here
in
terms
of
whether
or
not
there
are
responses
that
are
manifest
at
the
populations,
to
our
knowledge,
there's
no
evidence
at
this
point
linking
atrazine
exposure
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
137
and
adverse
effects
at
the
population
level.

Now,
where
are
we
in
terms
of
the
summary?
In
terms
of
the
overall
strength
of
association,
if
we
look
at
the
global
characteristics
of
temporality,
we
see
no
evidence
of
a
correlation
between
the
occurrence
of
gonadal
effects
and
the
introduction
and
use
of
atrazine.

In
terms
of
strength
of
association
and
the
various
kinds
of
parameters
that
we've
looked
at,
on
general
or
in
general,
there's
little
evidence
to
point
to
a
concentration­
dependent
response
with
atrazine
and
the
various
endpoints
that
we've
looked
at.
No
one
has
evoked
cautious
postulates
to
remove
the
stressor
to
try
to
establish
causality.
That's
something
that
with
a
robust
responses,
we
certainly
would
be
willing
to
and
would
like
to
consider.

In
terms
of
incidence
rates
in
the
population,
these,
for
the
variety
of
parameters
that
we've
looked
at,
are
clearly
inconsistent.

And
more
often
than
not,
the
various
types
of
confounders
that
could
have
influenced
the
types
of
responses
that
we've
seen,
particularly
in
the
field
situation,
have
not
been
specifically
addressed.

In
terms
of
consistency,
generally,
there
is
not
particularly
good
consistency
where
there
have
been
responses
measured.
In
terms
of
biological
plausibility,
in
terms
of
the
kinds
of
mechanisms
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
138
that
are
evoked
in
the
literature,
in
terms
of
effects
through
estrogenic,
androgen,
or
thyroid­
mediated
mechanisms,
we
see
little
evidence
to
suggest
that
atrazine
is
exerting
effects
in
that
way.

And
we
can't
specifically
address
the
question
of
recovery
given
that
there's
a
lack
of
consistent
and
robust
response
in
the
types
of
endpoints
that
we've
looked
at.

Now,
one
of
the
things
that
you
may
have
picked
up
is
that
the
group
that
we
work
with
is
an
atrazine
ecological
risk
analysis
panel.

So
what
we
should
be
doing
is
conducting
a
risk
analysis.
Well,
we're
hamstrung
and
we're
unable
to
do
a
risk
analysis
in
the
sense
that
we're
not
seeing
consistent
effects,
there's
no
consistent
concentration­
dependent
responses.
And
at
this
point
in
time,
a
risk
analysis
is
not
feasible
or
possible.

Thank
you
very
much
for
your
time
and
attention.

DR.
ROBERTS:
Thank
you
for
your
presentation.
I
would
like
now
to
ask
the
Panel
if
they
had
any
questions
for
you.
Dr.
Kelley.

DR.
KELLEY:
Yeah,
I
have
some
questions
about
the
Carr
study
that
was
published
in
2003.
Is
that
okay
to
ask
you
about?
One
of
the
puzzling
aspects
of
the
Carr
study
were
the
results
with
the
positive
control
which
was
raising
the
tadpoles
in
estradiol.
And
as
I
understand
it,
they
were
raised
beginning
48
hours
after
fertilization
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
139
all
the
way
up
until
they
got
to
Stage
66,
and
you
only
looked
at
Stage
66
animals.
And
there
are
a
number
of
studies
in
the
literature
that
indicated
that
these
animals
in
other
studies
would
have
a
hundred
percent
female
at
that
dose
and
nobody
had
ever
reported
intersex
at
that
dose;
although
it
is
reported
within
a
smaller
time
window.

And
I
wondered
if
you
had
a
feeling
for
why
you
had
relative
insensitivity
in
this
paradigm
to
the
positive
control?

DR.
CARR:
I
don't
know
if
it
was
a
lack
of
sensitivity.
If
you
actually
look
at
the
estradiol
levels
in
the
tanks,
which
is
in
the
technical
report,
they
are
a
lot
lower
than
they
should
be.
So
it
may
have
been
a
dose
response
effect.

DR.
KELLEY:
Oh,
so
you
think
it
was
actually
sticking
to
the
glass.

DR.
CARR:
No.
I
think
it
might
have
been
a
fact
of
the
tank
change
paradigm.
We
didn't
do
complete
tank
changes.
We
didn't
think
that
would
affect
atrazine.
In
fact,
it
didn't
affect
atrazine
levels.
And
that
was
the
purpose
of
the
experiment.

We've
done
other
studies
to
show
that
that
concentration,
if
you
maintain
target
concentrations
at
100
parts
per
billion
estradiol,
you
will
get
100
percent
females.

DR.
KELLEY:
Well,
so
what
I'm
disturbed
about
is
in
Figure
2.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
140
So
the
summary
statement
from
Figure
2
is
that,
in
fact,
the
effect
of
atrazine
did
not
resemble
the
effect
of
estradiol.
But
if
you
actually
look
at
Figure
2,
in
fact,
they
look
quite
similar;
although
clearly
the
effect
of
atrazine
is
not
as
significant
as
the
effect
of
estradiol.
So
I'm
just
worrying
that
the
paradigm
itself
diluted
the
delivery
to
such
an
extent
that
an
effect
present
could
not
have
been
picked
up.

DR.
CARR:
Diluted
the
delivery
of
­­

DR.
KELLEY:
Well,
clearly,
you've
just
told
me
it
diluted
the
delivery
of
estradiol
so
that
you
didn't
have
an
effective
concentration.

DR.
ROBERTS:
Dr.
Kelley,
I'm
sorry
to
interrupt.
But
can
you
make
it
clear
which
figure
you're
looking
so
the
rest
of
the
panel
can
see
what
you're
referring
to.

DR.
KELLEY:
Yes.
This
is
a
figure
­­
there
is
a
published
paper
this
year
on
whom
the
first
author
is
Dr.
Carr
who's
down
at
the
end
of
the
table.

DR.
CARR:
So
this
is
Figure
2
in
our
paper.
Okay.

DR.
KELLEY:
Yeah,
this
is
Figure
2
in
the
paper.
I've
also
read
the
technical
report
which
I'm
looking
at
here.
But,
you
know,

the
paper
is
a
lot
shorter.
It's
easier
to
get
through.

Anyway,
so
I
guess
I'm
concerned
about
the
fact
that,
although,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
141
of
course,
applying
statistics
to
it
­­
in
one
case,
you
have
a
statistical
effect
and
in
the
other
you
don't.
But,
in
fact,
the
graphs
are
actually
rather
similar.

DR.
CARR:
You're
right.
There
was
a
reduction
in
the
percentage
of
males
in
the
highest
atrazine
concentration.
It
was
not
statistically
significant.
There
was
no
alteration
in
the
percentage
of
females.

DR.
KELLEY:
Okay.

DR.
CARR:
And
that
led
us
to
our
other
conclusion
in
that
paper
that
atrazine
was
principally
affecting
male
gonadal
differentiation.

DR.
KELLEY:
Okay.
So
you're
attributing
your
lack
of
the
positive
control
to
the
fact
that
you
didn't
have
an
effective
enough
dose
of
estradiol.

DR.
CARR:
Correct.

DR.
KELLEY:
But
you
don't
think
that
that
applied
to
the
atrazine
in
the
study
­­

DR.
CARR:
Well,
we
know
it
didn't
because
­­

DR.
KELLEY:
­­
because
you
measured
it.

DR.
CARR:
­­
we
measured
it.

DR.
KELLEY:
That's
my
first
question.
I
will
cede
the
stage
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
142
somebody
else.

DR.
ROBERTS:
All
right.
Dr.
Skelly
and
then
Dr.
Green.

DR.
SKELLY:
During
your
presentation,
you
suggested
that
the
work
of
one
group,
Tavera­
Mendoza.
You
suggested
that
this
study
design
was
flawed
and
that
you
sort
of
implied
that
that
should
influence
how
we
think
about
the
evidence
that
came
out
of
that.
I
wanted
to
ask
you
a
general
question
and
a
specific
question.
And
that
is,
in
general,
do
you
think
study
design
flaws
should
influence
how
this
panel
views
the
evidence
that
we're
being
asked
to
look
at?

And
specifically,
if
atrazine
is
being
detected
at
control
sites,
should
that
influence
how
we
think
about
study
outcome?

DR.
VAN
DER
KRAAK:
In
terms
of
the
Tavera­
Mendoza
paper,
we
were
taken
back
by
the
lack
of
reproducibility
of
the
data
across
what
was
reported
in
the
thesis
and
what
was
reported
in
the
published
literature.
So
in
terms
of
our
evaluation,
we
felt
that
it
was
appropriate
that
we
identify
that
there
was
that
inconsistency.
And
so
rather
than
in
our
weight
of
evidence
providing
a
very
resounding
positive
response,
as
an
example,
we
felt
that
it
was
inappropriate
to
do
that;
given
that
within
their
own
hands,
that
wasn't
a
reproducible
effect.
And
so
we've
tended
to
diminish
the
value
of
that
in
our
scheme.
We
didn't
exclude
it
to
completion
in
that
we
reported
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
143
data
and
said
what
they
found.
So
50
percent
of
the
time,
they
get
a
response.

In
terms
of
the
specifics
of
their
experimental
design,
I
think
there's
a
number
of
issues
associated
with
the
type
of
experiments
they
did
in
terms
of
a
very
short
duration
response,
a
short
duration
exposure
paradigm,
looking
at
a
response
without
subsequent
follow
up
to
find
out
whether
this
was
a
long­
term
advancement
in
ovarian
development
as
the
case
was
in
females
or
a
significant
change
in
testicular
development
as
they
seem
to
report
in
males.
So
it
was
difficult
to
try
to
address
that
from
the
robustness
perspective.

Should
you
exclude
that
in
your
evaluations?
No.
I
think
you
should
include
it
in
your
evaluations.
But
you
should
look
at
all
of
the
available
data
in
arriving
at
your
individual
conclusions
as
to
how
you
placed
weight
on
individual
studies.

The
second
questions
I'll
give
my
response
to
it,
and
then
I'll
ask
others
on
the
panel
if
they
wish
to
add
something
additional.

Your
questions
was
whether
or
not
the
presence
of
atrazine
in
some
of
the
experiments
in
the
controls
would
be
something
that
would
cause
me
to
throw
out
that
data.
And
the
answer
to
that
is
that
in
my
mind,
that's
not
the
issue
that
would
throw
it
out
in
my
perspective
would
be
because
we
would
be
looking
at
that
in
a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
144
dose­
response­
related
paradigm
and
seeing
no
evidence
of
dose
response.
I
wouldn't
be
­­
I
would
feel
that
that
negated
the
nature
of
the
experiment.

Does
anybody
wish
to
add
to
that?

DR.
GIESY:
I
think
that's
an
excellent
question.
And
the
way
I
would
answer
it
is
I
think
that
would
preclude
being
able
to
ask
certain
kinds
of
questions.
But
it
wouldn't
negate
the
ability
to
ask
other
kinds
of
questions.

The
way
we've
approached
it
in
the
field
where
it
is
difficult
to
find
situations
where
there
is
no
atrazine
but
very
low
concentrations,

is
to
take
a
Type
2
statistical
approach
or
regression­
type
approach
to
look
at
that
data
because
it
is
difficult
to
ask
the
question
completely
without
and
with
atrazine.

So
I
think
you
have
to
look
at
each
study
specifically.
And
I
would
reiterate
what
Dr.
Van
Der
Kraak
said,
I
don't
think
you
throw
all
the
data
out.
But
I
think
it
does
preclude
the
ability
to
ask
certain
questions.

DR.
ROBERTS:
Moving
on
then.
Dr.
Green.

DR.
GREEN:
This
question
is
along
the
same
lines.
There
were
two
studies
that
you
referred
to
quite
frequently,
the
one
by
Dr.

Carr
in
2003
and
the
one
published
Dr.
Giesy
in
2003,
in
which
you
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
145
said
that
data
was
interpreted
as
atrazine
had
no
effect
on
the
various
parameters
you
were
looking
at.
A
very
nagging
concern
I
have
about
those
studies
has
to
do
with
stocking
densities,
loading
densities,
and
the
water
quality
in
those
tanks.

Ammonia
levels
as
high
as
27
milligrams
per
liter
are
quite
toxic.
And
that
alone
could
affect
the
outcome
of
that
study.
It
could
inhibit
the
growth
of
the
animals,
make
them
susceptible
to
infectious
diseases.
And
given
that
you
have
such
variability
in
stocking
density,
the
tadpoles
were
stocked
quite
heavily,
as
well
as
variability
in
water
quality,
how
can
you
support
the
conclusion
that
atrazine
had
no
effect
in
the
face
of
such
background
levels
of
other
toxic
substances.

And
I
have
a
follow­
up
question
to
that,
too,
if
that's
okay.

DR.
ROBERTS:
That's
fine.

DR.
GIESY:
Yeah,
that's
a
good
question.
The
studies
with
my
name
on
them
were
field
studies
from
South
Africa.
They
weren't
the
lab
studies.
I
think
the
ones
you
refer
to
are
the
ones
by
Hecker,

Environmental
Toxicology.

DR.
GREEN:
Yes.

DR.
GIESY:
Yeah,
and
you're
right.
All
those
issues
are
ones
we
identified
in
our
report
that
are
limitations
of
the
studies.
When
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
146
we
went
into
the
study,
we
tried
to
do
a
power
analysis
to
look
at
what
we
needed
in
the
way
of
sample
sizes
to
be
able
to
make
some
conclusions.
And
that
included
stocking
numbers
and
numbers
of
tanks
to
look
for
tank
effects
and
number
of
doses.

So
when
we
did
all
that,
in
the
end,
everything
was
a
compromise.
And
in
hindsight,
certainly,
if
we
had
the
space
and
ability
to
do
it,
we
would
have
chosen
to
use
lower
density
stocking
for
sure.
So
all
those
criticisms
that
the
EPA
has
pointed
out,
and
I'm
sure
the
Panel
will
pick
up
on,
are
valid
and
we
certainly
would
recognize
those.

Whether
it
completely
negates
the
utility
of
the
data,
I
personally
don't
think
so.
I
think
it
would
be
nice
to
be
able
to
do
it
again.
That's
why
personally
I
think
the
EPA's
conclusions
are
sound.

And
their
proposal
to
move
forward
is
a
good
one,
to
try
to
remove
of
those
uncertainties
that
we
readily
admit
are
there.

DR.
ROBERTS:
Follow­
up
by
Dr.
Green.

DR.
GREEN:
Yes.
I'd
like
to
know
just
in
general
by
members
of
the
Panel
who
have
labs
where
they
are
conducting
these
experiments.
What
test
kits
do
you
use,
and
how
frequently
do
you
monitor
water
quality
analysis
in
these
studies?
Are
they
color
metrics,
that
sort
of
thing?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
147
DR.
CARR:
The
standard
operating
procedures,
I
think,
were
made
available
as
part
of
the
GLP
conditions
for
the
study
and
there
should
be
water
quality
operating
procedures
in
there.
We
measure
temperature,
dissolved
oxygen,
conductivity,
ammonia,
using
the
Hawk
photometric
method
on
a
weekly
basis.
In
many
cases,

dissolved
oxygen
on
an
every­
other­
day
basis
or
every
three
days
when
we
do
tank
changes.

DR.
ROBERTS:
Dr.
LeBlanc
then
Dr.
Kloas
and
then
Dr.

Thrall.

DR.
KENDALL:
Mr.
Chairman,
I
think
Dr.
Carr
wants
to
add
to
the
first
question.

DR.
ROBERTS:
Okay.
That's
fine.

DR.
CARR:
Right.
Dr.
Green
was
asking
about
water
quality.

There
are
concerns
about
water
quality
in
a
static
exchange
design.

The
27
ppm
levels
that
came
up
­­
towards
the
end
of
the
study
when
the
animals
are
larger,
completing
metamorphosis,
the
unionized
ammonia
levels
were
about
.2
ppm.
We
didn't
see
high
mortality.

And
we
don't
think
that
ammonia
was
toxic
to
the
animals.

Did
it
effect
growth?
Well,
the
animals
did
develop
slowly.

They
were
at
a
lower
temperature.
But
we
also
saw
99
percent
of
the
animals
sexually
differentiated.
So
we
don't
think
it
impacted
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
148
degree
of
sexual
differentiation
and
that
the
critical
aim
of
the
study
was
to
examine
atrazine
effects
on
gonadal
development.

DR.
ROBERTS:
Okay.
Were
there
other
members
of
your
group
that
wanted
to
respond
to
Dr.
Green's
question?
Dr.
LeBlanc.

DR.
LEBLANC:
Thank
you.
When
I
think
about
androgens
and
estrogens,
I
tend
to
think
about
them
having
different
roles
in
adults
versus
juveniles.
That
is,
in
the
adult,
I
think
about
them
having
roles
in
reproduction.
And
in
the
juvenile,
the
larvae,
I
think
about
them
having
roles
in
development.
And
I
think
what
we're
concerned
about
today
is
a
role
that
atrazine
might
have
in
perturbing
development
of
these
larvae.

But
it
seems
like
a
lot
of
the
negative
data
that
was
just
presented
discounting
or
at
least
not
being
able
to
demonstrate
any
effect
of
atrazine
on
androgens
or
estrogens,
were
in
the
adult.

Correct
me
if
I'm
wrong
if
that's
not
the
case.
But
in
terms
of
steroid
hormone
levels,
aromatase
activity,
I
just
got
the
feeling
like
you
were
aiming
at
the
wrong
target
when
generating
this
information.

Could
you
comment
on
that
anyone?

DR.
ROBERTS:
Dr.
Giesy.

DR.
GIESY:
I
wanted
to
make
sure
you
introduced
me.

Yeah,
that's
a
good
question.
Some
of
the
studies
were
with
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
149
adults.
The
Xenopus
studies,
there
was
one
that
was
conducted
through
Stage
66
where
there
were
samples
taken
for
analysis
at
that
point.
And
there
was
a
subpopulation
that
was
grown
out
for
about
two­
and­
a­
half
months
beyond
that.

A
similar
study
was
done
with
the
ramaclamatns.
So
those
were
exposed
throughout
their
entire
development.
The
Carr
study
was
terminated
at
Stage
66.
And
those
were
exposed
throughout
the
entire
developmental
period.
There
were
other
studies
that
were
the
field
studies
where
those
were
adults.
So
they
were
collected
as
adults,
but
presumably,
they
were
exposed
to
the
environmental
concentrations
of
atrazine
in
those
situations
throughout
development.

And
then
there
were
some
studies
that
were
done
only
as
adults
to
look
at
potential
mechanisms
of
action
at
a
fairly
crude
high
level
to
see
if
we
could
get
induction
in
the
gonad
because
that
had
been
reported
in
the
literature
previously.
So
in
the
adults,
we
did
want
to
see
if
we
could
reproduce
that.

So
it
was
a
combination
of
adults.
But
mostly
it
was
throughout
development.

DR.
LEBLANC:
Can
I
follow
up?

DR.
ROBERTS:
Yes,
please.

DR.
LEBLANC:
As
related
to
androgenic
or
anti­
androgenic
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
150
effects
of
atrazine,
one
of
the
conclusions
that
was
reached
by
the
group
was
that
atrazine
did
not
mimic
THT.
And
I
think
the
data
demonstrated
that
rather
clearly,
though
I
suspect
that
never
really
was
a
hypothesis.
So
I
don't
think
any
of
the
data
that's
been
reported
suggests
that
it
is
acting
as
an
androgen.
And
if
anything,
perhaps
it's
acting
as
an
anti­
androgen.

And
I
was
wondering
if
the
Eco
Risk
group
has
ever
evaluated
it,
an
anti­
androgen,
to
see
if
the
effects
are
consistent
with
atrazine.

DR.
VAN
DER
KRAAK:
The
short
answer
is
no.
Have
we
considered
it?
Yes.
But
it
has
been
considered
in
relation
to
a
whole
host
of
various
hypotheses
that
in
the
goodness
of
time
will
get
tested.

DR.
ROBERTS:
Dr.
LeBlanc,
I
thought
your
first
question
might
have
encompassed
not
only
the
time
of
exposure
but
the
time
of
assessment,
developmental
stage
at
assessment.
And
I
wasn't
sure
whether
the
response
­­
Dr.
Giesy,
I
think,
focused
on
the
duration
of
exposure
and
the
developmental
stages
of
exposure
but
not
necessarily
at
the
times
of
assessment.

So
if
I
might
jump
in
and
follow
up.
Dr.
Giesy,
can
you
touch
on
that
in
terms
of
stages
of
development
at
which
assessment
was
conducted
and
how
that
might
factor
into
the
interpretation.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
151
DR.
GIESY:
Yeah.
We,
as
a
panel,
have
talked
about
that
a
lot
and
think
that's
really
critical
to
doing
experiments
and
also
interpreting
the
data.
In
our
studies,
we
developed
protocol.
If
we
were
to
do
additional
studies,
certainly,
we
would
want
to
design,
I
think
like
the
EPA
is
proposing
to
design,
a
system
where
we
could
look
at
some
of
the
critical
windows.
We
think
that's
very
important.

And
I'm
going
to
let
Jim
Carr
mention
things
in
a
minute.

But
we
think,
also,
that
it
may
lead
to
some
of
the
difficulties
in
interpretation
and
comparison
among
data
sets,
among
laboratories,

how
animals
are
exposed
and
when
they
are
collected,
and
whether
or
not
they're
grown
out.
We
agree
with
the
EPA
that
to
do
that
grow­
out
study
is
important.
And
I
think
Dr.
Kelley
mentioned
that
this
morning.
I
couldn't
agree
more.
It's
very,
very
appropriate
to
do
that.

So
interpreting
the
data,
the
timing
of
exposure,
and
in
a
minute,
I'll
make
some
comments
relative
to
aromatase
when
the
time
is
appropriate
that
also
would
impinge
on
when
you
collected
it
in
the
developmental
cycle.

DR.
CARR:
One
of
the
technical
issues
with
looking
at
hormone
levels
in
the
tadpoles,
of
course,
is
that
you're
restricted
by
the
amount
of
blood
that's
available
to
look
at
blood
hormone
levels.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
152
And
what's
often
done
is
to
look
at
whole­
body
hormone
levels.
So
that's
a
technical
issue
that
would
need
to
be
addressed
in
the
interpretation
of
whole­
body
hormone
levels
relative
to
the
onset
of
gonadal
steroid
secretion.

The
other
issue
is
the
transfer
of
maternal
steroids
into
the
egg
and
the
contribution
of
those
steroids
and
separating
those
contributions
out
from
the
steroids
that
are
produced
andogenously
from
the
animal's
gonads.

So
I
think
there
are
some
technical
issues
that
would
need
to
be
addressed,
too.
And
we
have
discussed
those
several
times.
And
it
would
be
important
to
look
at
those,
I
think.

DR.
ROBERTS:
Dr.
Kloas
and
then
Dr.
Thrall.

DR.
KLOAS:
I
would
like
to
continue
in
this
field.
I
would
like
to
know
something
about
why
did
you
use
this
experimental
design
for
measuring
aromatase
steroid
levels.
So
as
you
are
aware,

the
endocrine
system
you
have
some
counter­
regulation.
So
you
assessed
estradiol
and
testosterone
after
at
least
26
days.
And
why
didn't
you
use
the
short­
term
exposure.
For
instance,
let's
say
half
a
day,
one
day,
three
days,
seven
days?
And
then
if
there
is
any
change
in
aromatase
activity
and
also
in
estradiol
and
androgen
levels,

because
after
26
days
or
43
days,
there
might
be
a
readjustment
by
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
153
endocrine
counter­
regulation
by
hypothalamic
pituitary
levels.
So
I
think
you
cannot
really
exclude
any
aromatase
effect.

And,
furthermore,
my
second
question
is
methodology
by
Miyita,
environmental
toxicology
al.,
would
also
allow
to
assess
aromatase
activity
in
tadpoles.
Just
a
couple
of
seconds
ago,
you
were
talking
about
sensitive
windows.
Why
not
to
do
short­
term
exposures
in
tadpoles
and
measuring
aromatase
activity
by
using
semi­
quantitative
auto
PCR?

DR.
GIESY:
All
great
suggestions.
I'd
love
to
do
it
all.

DR.
ROBERTS:
Dr.
Giesy's
responding.

DR.
GIESY:
Those
are
all
great
suggestions.
In
fact,
in
our
laboratory
now,
we've
developed
molecular
beacons
for
CYP19.
We
can
do
that.
So
these
were
initial
studies.
We
wanted
to
start
at
a
sort
of
a
high
level,
gross
look
and
see
if
we
could
reproduce
what
was
reported
in
the
literature.
But
I
would
agree
it
does
not
allow
us
to
preclude
the
potential
effects
through
an
aromatase
mechanism
of
action
in
specific
localized
tissues.
So
I
think
timing
is
important
to
do
that
and
look
at
it.
To
do
that
in
small
tissue
amounts
we
would
have
to
use
PCR.
And
like
I
said,
later,
whenever
it's
appropriate,
I'll
talk
more
about
the
aromatase
hypothesis
and
what
I
think
about
it
and
its
future.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
154
But
I
think
all
of
those
are
great
suggestions.
I
would
not
disagree
with
any
of
them.

DR.
ROBERTS:
Dr.
Thrall
then
Dr.
Kelley.

DR.
THRALL:
Maybe
I
missed
it.
But
I
was
still
just
a
little
confused
about
what
the
inconsistency
was
on
the
Tavera­
Mendoza
study.
You
said
there
was
an
inconsistency
between
the
thesis
and
the
published
paper.
And
I
wondered
if
you
could
be
more
specific
about
that.
This
was
in
relation
to
testicular
volume.

DR.
SOLOMON:
We
originally
saw
these
papers
only
in
publication,
and,
subsequently,
obtained
a
copy
of
the
thesis.
I
don't
know.
Has
the
Panel
seen
the
thesis?

DR.
ROBERTS:
I
do
not
believe
that
that's
been
entered
into
the
docket.

DR.
SOLOMON:
One
of
the
issues
in
the
published
paper
was
that
they
exposed
the
animals
for
a
relatively
short
period
of
time.

And
then
they
reported
a
decrease
in
the
size
of
the
testes,
in
the
volume
of
the
testis,
up
to
70
percent
as
I
recall.
However,
they
didn't
actually
measure
the
size
of
the
testes
in
the
animals
when
they
started.
It
was
just
comparing
controls
to
the
treated
or
the
exposed
animals
which
raised
concerns.

Some
other
concerns
were
differences
in
the
numbers
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
155
animals
reported
between
the
figures
and
the
text
of
the
papers
that
were
not
consistent.
We,
subsequently,
obtained
a
copy
of
the
thesis
and
found
that
a
second
experiment
had
been
conducted
with
greater
range
of
concentrations;
not
just
21,
which
was
actually
measured
at
18,
three
different
concentrations.
And
they
had
not
seen
a
concentration
response
and
no
statistically
significant
differences.

So
on
that
basis,
we
felt
that
there
was
obviously
some
design
flaws
in
addition
to
the
small
numbers
of
animals
used,
the
small
number
of
tanks.
There
were
only
two
tanks
used.
So
they
couldn't
look
at
inter­
tank
variation.
So
that
to
our
mind,
diminished
the
value
of
that
data
in
interpreting
these
responses.

DR.
ROBERTS:
Dr.
Kelley
and
then
Dr.
Skelly.

DR.
KELLEY:
So
I
have
two
sets
of
questions.
The
first
really
to
the
field
data
in
South
Africa
since
I
have
to
report
on
that.

So
the
animals
were
sampled
in
April
and
May
which
is
just
at
the
end
of
the
rains.
Could
I
have
some
information
on
the
relation
between
the
data
sampling
of
the
adults
and
the
onset
of
the
breeding
season?

DR.
DU
PREEZ:
In
the
Potchefstroom
area,
Xenopus
breed
from
September,
September,
October,
November.
That's
the
onset
of
the
rainy
season.
But
Xenopus
has
got
a
prolonged
breeding
season
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
156
from
September
right
until
April,
end
of
April.
So
­­

DR.
KELLEY:
So
you
were
sampling
at
the
end
of
the
breeding
season
then.

DR.
DU
PREEZ:
Yes.
We
sampled
the
great
majority
of
the
specimens
at
the
end
of
the
breeding
season.
In
a
few
of
the
ponds,

we
had
difficulty
collecting
the
targeted
number
of
specimens;
and
we
collected
those
during
subsequent
months
after
that.

DR.
KELLEY:
Did
you
see
any
difference
in
your
measurements
of
plasma
steroid
levels
depending
on
the
time
of
year?

So
you
had
most
of
your
animals
in
April
and
May.
But
you
had
this
one
group
where
you
collected
at
four
different
times.
Were
they
pooled?
Or
were
you
able
to
look
at
those
data
separately?

DR.
GIESY:
That's
an
excellent
point
one
we've
discussed
at
length
within
the
panel.
Let
me
cut
to
my
conclusion.
Then
I'll
go
back
and
try
to
backfill
with
some
details.

From
where
we
are
now,
I
would
have
two
conclusions.
One,
I
don't
think
it's
very
useful
unless
we
understand
the
seasonal
cycles
and
are
able
to
stratify
our
sampling
to
use
measurements
of
estradiol
and
testosterone
as
functional
endpoints.
EPA
has
come
to
that
conclusion
in
their
White
Paper,
and
I
agree.

Now,
the
reason
for
that
is
the
sample
sizes
required
to
have
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
157
any
power
to
show
effects
would
be
pretty
large.
So
we
recorded
when
we
collected
them.
We
also
determined
gonadal
stage
development
with
a
scoring
system,
and
we
also
determined
age.
But
when
we
went
to
go
to
try
to
stratify
the
organisms
by
time
of
collection
in
the
season,
gonadal
stage,
and
age,
things
got
pretty
thin
in
the
sample
sizes.
So
that's
a
trade­
off.
So
I
think
any
field
work
that
people
want
to
do
is
going
to
be
very
limited
because
of
that
problem.

Now,
in
South
Africa,
as
you
well
know,
the
Xenopus
are
not
synchronous
spawners.
They
spawn
continuously
throughout
the
season.
Some
may
not
spawn
at
all.
Some
may
spawn
once,
and
some
may
spawn
several
times.
So
I
think
that's
what
leads
to
the
great
amount
of
variation.

So
at
the
same
time,
I
then
think,
well,
with
the
effects
we
saw,

we
do
see
effects
in
the
corn­
growing
region.
They
are
fairly
small
relative
to
the
variation
that
we
see.
And
one
question
is
what
ecological
relevance
does
that
have.

But
to
answer
your
question,
I
think
it's
absolutely
critical
that
we
be
able
to
stratify
our
sampling
by
season,
by
age,
and
by
reproductive
condition
to
be
able
to
interpret
any
potential
effects
of
compounds
like
atrazine
on
hormones.
It's
a
difficult
problem
as
you
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
158
well
know.

DR.
KELLEY:
So
if
I
could
just
bring
the
attention
of
your
group
to
some
available
data
on
hormone
levels.
So
we
measured
hormone
levels
both
in
serum
and
in
mid­
sections
of
bodies
at
various
stages
and
development
which
is
in
the
Kang,
environmental
toxicology
al.,
1995
paper
from
General
and
Comparative
Endocrinology.
And
where
we
could
compare
the
mid­
section
level
to
the
serum
level
because
we
had
enough
tissue,
they
were
very
close.

So
it
may
be,
in
fact,
that
that's
an
adequate
way
to
do
that
study.

And
this
is
also
the
way
that
a
more
recent
study
by
Bogge,

environmental
toxicology
al.,
in
Comparative
Biochemistry
and
Physiology,
Part
B
in
2002,
measured
both
17­
beta
estradiol
and
androgen,
both
T
and
DHT,
throughout
development,
were
able
to
document
the
contribution
of
maternal
hormones
very
early
in
development,
and
then
the
later
contribution
of
hormones.

And
in
their
paper,
although
we
did
not
see
this
in
ours,
the
levels
are
comparable.
But
it
looks
like
their
variability
is
lower.

They
actually
have
a
sex
difference
in
androgen
level
and
also
in
estrogen
level
at
the
time
of
sexual
differentiation.

So
there
clearly
are
some
data
available
now
that
would
enable
or
approaches
that
would
enable
you
to
look
at
that.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
159
That
was
a
comment.

DR.
GIESY:
One
I
didn't
fully
appreciate.
Can
you
tell
me
about
the
method?
Could
you
resample
the
same
individual?
Is
that
what?

DR.
KELLEY:
No.
What
we
did
was,
you
know,
you
can't
get
enough
blood
out
of
a
tadpole
to
run
a
reliable
radioimmunoassay.

They're
pretty
small.
And
we
actually
did
these
assays
down
to
Stage
56,
which
is
just
towards
the
beginning
of
metamorphosis.
But
at
Stage
66,
the
end
of
metamorphosis,
we
were
able
both
to
get
tissue
samples,
not
from
the
­­
actually,
did
we
do
them?
I
think
actually
we
did
do
them
from
the
same
individual
come
to
think
of
it.
But
it
didn't
make
any
difference.
The
variability
was
quite
low.
And
in
that
case,

the
serum
levels
agreed
quite
well
with
­­
this
is
a
mid­
body
segment
that
includes
both
the
liver,
which
would
be
the
major
clearance
organ,
and
the
gonads.
So
I
think
it
is
possible
to
do.

And
the
comment
which
we
got
from
the
reviewers,
which
I
will
forward
to
you,
was
that
they
were
worried
about
contamination
from
lipids.
But
we
were
able
to
extract
lipids
and
come
up
with
exactly
the
same
numbers.

DR.
ROBERTS:
Dr.
Skelly
followed
by
Dr.
Green
and
then
Dr.

LeBlanc.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
160
DR.
SKELLY:
In
summing
up
what
you
had
concluded
from
concerning
Hypothesis
V,
which
is
atrazine
causes
adverse
effects
at
the
population
level
in
exposed
amphibians,
you
mention
that
a
number
of
studies,
which
you
had
done
and
which
are
part
of
the
open
literature,
had
sampled
robust
populations.
And
wondered
if
you
could
tell
me
what
your
group
defines
as
a
robust
population
and
what
sort
of
demography,
breeding
behavior
and
breeding
success
sorts
of
parameters
you've
measured
and
you
plan
to
measure.

DR.
DU
PREEZ:
As
part
of
this
study,
we
did
a
mark
and
recapture
study
to
determine
populations
in
both
corn­
growing
and
cattle­
farming
areas.
And
in
all
of
these
sites,
we
found
large
numbers
of
Xenopus.
Male
female
ratio
were
the
same.
No
statistical
difference.
Xenopus
populations
do
fluctuate
sometimes
due
to
introduction
of
catfish.
And
as
we've
seen
this
past
year,
catfish
can
wipe
out
a
Xenopus
in
one
specific
pond
in
a
relatively
short
time.
So
you
have
this
constant
fluctuation.

But
if
you
set
the
traps,
you
get
a
feeling
for
the
population.
If
you
have
difficulty
getting
the
specimens
during
a
certain
part
of
the
year,
it's
easier
to
trap
Xenopus
in
spring.
You
get
larger
numbers
in
the
traps.
But
in
all,
those
populations
appear
to
do
very
well.

DR.
SKELLY:
I
spend
a
lot
of
my
time
wearing
rubber
pants
as
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
161
I'm
sure
some
of
you
do.
I
guess
what
I
was
asking
specifically
is,

based
on
what
you
have
done
which
is
look
at
literature,
published
literature
studies,
and
your
own
studies
which
seem
to
be,
maybe
with
some
exceptions,
going
out
and
sampling
either
over
a
short
interval
or
just
looking
at
a
study
that
might
have
gone
in
just
once,
you're
declaring
something
to
be
a
robust
population.

I'm
a
population
demographer.
That
raises
antennae.
So
what
is
a
robust
population.
And
if
that's
just
sort
of
a
vague
descripter,

I'd
like
to
know
that.

DR.
KENDALL:
Mr.
Chairman?

DR.
ROBERTS:
Dr.
Kendall.

DR.
KENDALL:
I'd
like
for
you
to
tell
us
what
you
think
is
a
robust
population.
No
disrespect.
I
would
like
for
you
to
address
that.
And
then
we
will
respond.

DR.
KELLEY:
Okay.
Well,
I
mean,
I
guess
going
back
to
the
conceptual
model
that's
been
forwarded
here,
we're
ultimately
trying
to
get
at
viable
populations.
And
I
guess
you
could
also
define
it
comparatively.
You've
gone
out
and
measured
atrazine­
exposed
sites
and
control
sites
and
you
could
do
comparisons
as
well.

What
I
was
asking,
none
of
that
information
was
presented
when
you
mentioned
robust
populations.
I
didn't
know
whether
you
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
162
had
collected
information
that
you
weren't
showing
us.

DR.
KENDALL:
I
think
that
point's
well­
taken.
Dr.
Ernest
Smith
would
like
to
respond.

DR.
SMITH:
We
are
responsible
for
the
Iowa
study.
We
have
not
done
mark
and
recapture.
But
based
on
the
profile
over
our
first
year,
which
was
really
a
pilot
study,
we
observed
a
significant
change
in
the
number
of
juveniles
as
we
sampled
during
the
late
spring,

early­
late
summer,
and
early
fall.
And
as
a
result
of
that,
I
think
from
that
standpoint,
I
would
say
there
is
an
evidence
of
reproduction,

evidence,
indication
of
juvenile
metamorphosis,
differences
in
that
increases
as
you
sample.

So
we're
back
into
the
same
site
for
a
second
year.
And
I
think
we'll
have
a
better
feel
for
what
is
a
robust
population
relative
to
last
year's.
But
at
this
point,
I
think
there
is
enough
indication
there
from
the
numbers
that
we
have
captured
and
released
back
into
those
sites.

DR.
KENDALL:
Dr.
Gross.

DR.
GROSS:
We've
been
coordinating
and
looking
at
the
cane
toad
which
was
summarized
in
the
previous
presentation
in
South
Florida.
And
at
least
in
our
studies
on
the
sugar
cane
sites,
we
see
populations
we
consider
robust,
to
answer
you
question
in
part,
due
to
the
fact
that
we
see
all
age
classes
present
within
the
group
that
we're
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
163
looking
at.
We're
able
to
collect
several
hundred
animals,
actually,

on
those
sites
within
a
couple
of
hours
time
frame
representing
all
these
age
classes
as
compared
to
control
or
reference
sites
where
it
would
usually
take
us
weeks
to
collect
similar
numbers
if
even
possible
to
collect
those
numbers.

We've
also
been
able
in
data
you
haven't
seen
demonstrate
that
there
are
tadpoles
on
those
particular
sites,
egg
masses,
and
so
on.
So
reproduction
is
obviously
occurring
on
those
particular
sites.
And
we
consider
them
to
be
fairly
robust
for
those
reasons.

DR.
ROBERTS:
Did
you
have
a
follow­
up,
Dr.
Skelly?

DR.
SKELLY:
I'm
done.

DR.
SOLOMON:
Just
an
additional
comment
on
the
South
Africa
studies.
We
obtained
estimates
of
the
total
population
size
based
on
the
mark
and
recapture.
And
these
were
not
inconsistent
with
the
sizes
of
the
sites.
The
smaller
sites
had
smaller
populations.

The
larger
sites
had...
So
if
one
thinks
that
a
site
may
have
a
certain
carrying
capacity,
it
was
consistent
with
what
we
saw
there
except
for
the
cases
of
introduced
predators,
which
would
obviously
affect
numbers
for
different
reasons.

DR.
ROBERTS:
Dr.
Green
then
Dr.
LeBlanc,
Dr.
Richards,
Dr.

Gibbs,
Dr.
Kloas,
and
Dr.
Kelley.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
164
DR.
GREEN:
I
think
you
just
answered
part
of
my
question
about
the
viability
of
the
eggs
produced
by
these
females
on
sites
where
atrazine
contamination
is
known
and
the
viability
of
the
sperm
from
the
male,
and
are
the
eggs
able
to
be
fertilized.
Apparently
so
if
you
say
there
are
healthy
populations.

It
came
to
my
attention
here
when
you
were
presenting
some
of
your
data
in
the
core
presentation,
the
Hayes
and
Hecker
studies,

where
you
cited
ovarian
morphology
in
frogs
and
other
species
was
normal.
And
these
were
laboratory
frogs;
correct?
So
I
was
wondering
if
anyone
from
the
panel
had
extended
those
studies
to
actually
if
those
eggs
were
fertilizable
because,
in
my
experience
in
the
laboratory,
a
good
healthy
looking
egg
may
not
yield
the
kind
of
data
you're
looking
for.
It's
not
viable
even
though
it
appears
to
be
so
by
physical
characteristics.

DR.
GIESY:
I
knew
you
had
assembled
a
super
panel,
Steve;

but
these
questions
are
great.

Once
again,
we
couldn't
agree
more.
We've
thought
about
that.

We've
even
gotten
to
the
point
of
designing
some
studies,
both
ex
vivo
type
studies
and
I
think
that
it's
very
appropriate
to
do
that
kind
of
a
grow­
out
study.
So
both
ways
that
I
heard
suggested
earlier
today
in
discussion,
I
think,
have
some
merit.
I
think
the
EPA
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
165
suggested,
and
I
think
Professor
Kelley
might
have
mentioned,
maybe
ex
vivo
approaches
to
try
to
get
at
that
in
a
staged
way.
I
think
that's
a
good
thing
to
do.
But
the
definitive
study
is
that
grow­
out
study.

Absolutely.
And
I
think
the
way
the
EPA
has
proposed
to
stage
that
and
work
through
it
has
some
merit.

But
I
would
add
the
caveat
that
I'm
not
completely
comfortable
with
the
decision
tree
of
saying
if
we
don't
see
testicular
oocytes,
for
instance,
or
some
other
histological
response,
that
the
decision
would
be
to
not
do
that
study.
I
personally
am
not
particularly
comfortable
with
that.
I
think
it
is
a
very
important
thing
to
do.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
When
discussing
testicular
oocytes,
Dr.
Van
Der
Kraak
commented
that
the
phenomenon
seems
to
be
relatively
common
at
least
in
the
experiments
unrelated
to
atrazine.
None
the
less,
if
we
look
at
the
figure,
and
I'm
referring
to
the
figure
titled,

Testicular
Oocytes
in
Grow­
Out
Xenopus
by
Hecker,
it
certainly
at
face
value
it
appears
that
testicular
oocytes
are
regulated
by
hormones
that
negatively
regulated
by
DHT
and
positively
regulated
by
estradiol
perhaps.

What
I
was
questioning
is
as
related
to
atrazine
we
see
what
appears
to
be
a
greater
than
two­
fold
increase
in
testicular
oocytes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
166
with
.1
microgram
per
liter
atrazine.
There's
no
error
there
to
indicate
the
level
of
variability
and
there's
an
indication
that
it's
not
a
significant
increase.

But
my
question
is,
and
I
recognize
that
we
as
toxicologists
would
be
very
uncomfortable
looking
at
dose­
response
curves
that
don't
conform
to
a
monatonic
response.
But
was
any
consideration
given
to
the
fact
that
this
might
be
real
increase
in
that
the
inability
to
detect
a
significant
increase
reflects
the
statistical
design
or
design
of
the
experiment?

DR.
ROBERTS:
Before
you
respond,
can
we
get
the
particular
figure?
Is
this
the
one?
Okay.

DR.
GIESY:
I'm
not
quite
with
you.
Can
you
repeat
it
again,

please?
I've
got
the
picture
now.

DR.
LEBLANC:
Okay.
So
now
we're
looking
at
­­
well,
first
if
we
look
at
the
bottom,
would
you
concur
that
testicular
oocytes
are
regulated
by
hormones?
It's
hard
to
make
judgements
here
because
we
don't
see
what
the
error
is
associated
with
these
values.
But
­­

DR.
GIESY:
This
is
one
of
our
studies.
In
the
report
that
you
have,
the
means,
medians,
ranges,
and
all
of
the
statistics
for
this
data.
So
that's
maybe
what
you
want
to
look
at,
Gerry,
to
get
the
specifics.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
167
The
one
problem
is
sample
sizes
are
fairly
small
relative
to
the
incidences
that
we're
seeing.
So
I'm
not
sure
we
can
draw
much
conclusion.
What
I
can
say
is
based
on
this,
at
least
with
atrazine,
it
doesn't
look
like
a
very
robust
response.
It's
not
a
huge
response.

Whether
estradiol
affects
this,
I'm
still
open
on
that.
I
think
it
potentially
can.
That's
why
I'm
concerned
about
some
of
the
circulating
plasma
concentrations
and
some
of
the
regression
relationships
we
did
see
from
the
South
Africa
study.
So
that's
why
I
wouldn't
necessarily
focus
only
on
the
testicular
oocytes.

But
the
problem
here
is
I
think
the
sample
sizes
are
fairly
low
and
the
incidences
are
low.
And
that's
what
leads
to
not
being
able
to
show
a
statistical
difference.
So
in
planning,
if
the
EPA
moves
ahead
and
has
this
study
repeated,
they
can
look
at
this
data
as
a
way
to
do
their
power
analysis
to
figure
out
exactly
what
size
they
need.
But
they're
going
to
be
very,
very
substantial
sample
sizes
you're
going
to
need
to
show
a
difference.

DR.
ROBERTS:
Dr.
Richards,
then
Drs.
Gibbs,
Kloas,
and
Kelley.

DR.
RICHARDS:
I'm
interested
in
the
South
Africa
field
studies.
And
it
probably
relates
to
other
field
studies
also.

DR.
ROBERTS:
Dr.
Richards,
can
you
get
the
mike?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
168
DR.
RICHARDS:
I'm
sorry.
I'm
interested
in
the
South
Africa
field
studies.
But
this
relates
to
other
field
studies
in
general.

With
essentially
eight
data
points,
I
wonder,
it's
certainly
critical
to
have
some
confidence
in
the
degree
of
exposure
that
we
think
the
animals
were
exposed
to.
And
my
question
is
sort
of
general
in
that
how
well
do
you
think
you've
characterized
it
in
the
way
you
portray
the
data,
four­
week­
mean
concentration.
I
can't
remember
exactly
your
procedure
or
how
frequently
you
measured.
Certainly,

with
hydrologic
events
very
dramatically
influence
concentrations
of
atrazine.
I
just
wonder
how
well
do
you
think
this
portrays
what
they
were
exposed
to?

DR.
DU
PREEZ:
What
you
have
to
take
in
mind
is
over
one
field
use
season
we
measured
atrazine
and
other
triazine
concentrations.
But
that
is
not
the
concentration
that
that
specific
animal
was
exposed
to
during
its
development.
That
specimen
might
be
four,
five,
six
years
old.
And
we
don't
know
what
those
atrazine
or
triazine
concentrations
were
five,
six
years
ago.

But
from
what
we've
seen
in
the
specific
season,
it
was
a
season
with
a
very
high
rainfall.
And
we
would
hypothesize
that
those
animals
were
exposed
to
probably
much
higher
atrazine
concentrations
than
was
recorded
in
the
specific
season.
The
answer
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
169
is
we
don't
know
what
those
animals
were
exposed
to.

DR.
SOLOMON:
Could
I
just
add
a
comment
to
that?

In
designing
the
experiment,
we
realized
we
couldn't
go
back
in
time
and
measure
prior
atrazine
concentrations.
We
did
have
historical
data
from
theses
and
other
sources
that
showed
the
presence
of
atrazine
at
relatively
high
concentrations
in
surface
waters
in
that
area.
So
we
realized
that
they
could
be
quite
large.

The
season,
as
I
said
to
somebody
earlier
today,
if
I
could
have
predicted
that
rainfall,
I
would
have
sold
my
shares
in
Enron.
But
I'm
not
that
good.

It
was
interesting,
though,
that
in
high
rainfall,
it's
dilute.
But
we
did
measure
concentrations
every
week
during
the
application
season
because
not
all
the
fields
are
treated
on
exactly
the
same
day.

And
then
every
two
weeks
after
that.
So
we
have
a
fairly
good
temporal
exposure
regimen.
And
for
the
purposes
of
the
assessing
the
effects
on
hormones
and
aromatase,
we
decided
to
use
the
concentrations
in
the
four­
week
period
prior
to
the
capture
and
sampling
of
the
animals
because
we
suspected
­­
or
expected
rather
­­

that
these
kinds
of
responses
would
be
related
more
to
recent
exposure
than
previous
exposure.

And,
of
course,
the
metamorphose
that
we
collected
in
the
study
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
170
year,
were
exposed
to
the
concentrations
that
we
measured.
So
we
know
what
they
were
exposed
to.
And
we
also
know
what
their
responses
were
in
terms
of
gonadal
responses,
environmental
toxicology
cetera.
So
I
think
we
have
a
good
handle
on
both
temporal
exposures
and
for
the
site
exposures
as
well.

But
then
as
Dr.
Du
Preez
pointed
out,
we
know
that
they
were
probably
exposed
earlier;
but
we
don't
know
to
what
concentration
or
that
we
theorize
that
it
may
be
greater.

DR.
DU
PREEZ:
If
I
might
add
another
comment.
What
we've
observed
is
a
definite
peak
in
atrazine
directly
after
the
application.

That
would
be
from
December,
January,
February
we
observed
a
peak.

But
the
Xenopus
started
breeding
end
of
September.
So
those
first
couple
of
months,
Xenopus
would
breed
in
fairly
low
atrazine
concentrations.
And
then
those
that
spawned
after
December,
would
be
exposed
to
higher
concentrations.
So
that
is
making
this
whole
interpretation
of
the
field
use
of
the
field
data
even
more
complicated.

DR.
KENDALL:
Dr.
Bob
Silken.

DR.
SILKEN:
Bob
Silken,
statistician­
consultant
to
the
Panel.

Given
that
you
didn't
have
exposure
concentration
information
over
the
entire
profile
of
the
animals
life,
although
you
did
have
a
pretty
good
handle
on
what
it
was
in
the
recent
past,
and
also
in
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
171
keeping
with
Dr.
Giesy's
comments
earlier
that
even
though
you
may
have
some
concentration
of
atrazine
in
the
control
sites
or
the
reference
sites,
that
doesn't
eliminate
making
some
use
of
the
data.

We
not
only
did
comparisons
between
non­
corn­
growing
sites
and
corn­
growing
sites,
but
we
also
took
what
concentration
information
was
there
and
compared
the
sites
with
the
four
lowest
atrazine
concentrations
with
the
four
highest.
And
they
were
fairly
divided.

We
also
did
the
separation
for
anything
below
one
and
anything
above
two
and
compared
the
five
sites
that
were
below
one
with
the
three
sites
that
were
above
two.
So
that
even
though
you
didn't
have
exact,
precise
exposure
concentration,
you
could
still
do
comparisons
at
different
levels
of
exposures.
And
the
comparisons
were
reasonably
consistent
across.
No
matter
how
you
grouped
them,

reference,
low,
low
three,
low
five,
low
four,
the
analyses
came
out
pretty
much
the
same.

DR.
RICHARDS:
One
follow­
up?

DR.
ROBERTS:
Sure.

DR.
RICHARDS:
I'm
just
curious
about
the
biology
of
this
creature.
I
know
in
the
mid­
west,
sometimes
when
I'm
sampling
Rana
species,
when
I
get
a
heavy
rainfall,
there's
water
everywhere
and
there
are
Rana
everywhere,
moving
around
between
ponds.
That's
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
172
probably
part
of
their
metapopulation
dynamics
and
so
forth.

Does
this
occur
during
high
flows
with
this
species
in
South
Africa?

DR.
DU
PREEZ:
Yes.
Xenopus
has
been
well­
documented
to
migrate
and
sometimes
en
mass.
Sometimes
you
find
mass
migrations.
What
we
did
in
this
study
is
we
tried
to
determine,
was
there
any
migrations.
And
what
I
did
was
to
brand
specimens
from
a
specific
pond
with
a
digital
number
corresponding
to
the
site
number.

And
we
did
not
observe
any
migrations
during
our
study.
But,
yes,

they
do
migrate.

DR.
ROBERTS:
Dr.
Gibbs.

DR.
GIBBS:
Yes.
In
your
presentations,
there
were
repeated
presentations
of
negative
results.
I
had
anticipated
more
of
a
consideration
of
a
power
of
tests.
And
I'm
just
curious
particularly
with
the
field
studies.
And
I'm
just
curious
how
widespread
power
analyses
were
in
your
analyses
both
perhaps
post
hoc
or
a
prior.

DR.
KENDALL:
Good
question.
Bob
Silken.

DR.
SILKEN:
I
guess
I'm
not
going
to
get
to
sit
over
there.

Yes,
we
did
try
and
do
power
analyses
before
we
went
into
the
tests.

And,
of
course,
the
power
analyses
varied
in
requirements
depending
upon
whether
it
was
a
lab
study,
a
field
study,
or
what
endpoint
for
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
173
the
obvious
reasons
about
the
difference
in
variability.

For
example,
in
Dr.
Carr's
lab
study,
we
found
that
we
definitely
at
least
eight
tanks
because
we
felt
the
tank
effects
were
going
to
be
a
strong
factor.
We
went
with
11
tanks
in
order
to
maximize
the
power,
given
that
11
tanks
times
the
number
of
treatments
was
a
many
tanks
as
they
could
work
with,
that
they
had
room.
So
power
was
in
there.

We
also
did
power
calculations
as
far
as
the
numbers
per
tank
which
may
have
led
to
some
overcrowding.
Blame
it
on
the
statistician
who
wanted
bigger
numbers.
Although
we
found
the
number
of
tanks
were
much
more
important
than
number
of
animals
per
tank.

When
we
did
the
power
analyses,
for
example,
for
Dr.
Carr's
study,
we
did
look
at
the
high
degree
of
correlation
within
a
tank.

And
we
looked
at
what
the
power
would
be
if
we
effectively
had
one
animal
per
tank.
Even
though
we
put
30
in
there,
if
they
effectively
all
behaved
the
same,
what
would
the
power
be
if
we
had
one
animal
per
tank.
And
it
ranged
in
his
study
from
something
like
70
power
to
detect
the
types
of
departures
that
were
being
talked
about.
A
power
in
the
range
of
70
to
98
percent.

If
we
had
four
animals
per
tank,
for
all
the
types
of
changes,
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
174
powers
went
up
to
99
percent.
So
we
did
look
at
power.

We
looked
at
powers
also
in
the
field
study.
And,
again,
the
variability
there
makes
it
such
that
the
powers
would
be
less.
We
would
strongly
encourage
that
if
you're
going
to
do
field
studies,
that
the
site­
to­
site
variability
probably
dominates
everything
else.
And
that
that
means
that
you
need
a
large
number
of
sites,
both
control
sites
and
treatment
sites
because
the
variability
in
the
controls
is
huge
as
well.
And
two
or
three
control
sites
is
not
enough.

So
we
were
looking
heavily
at
power.
What
else
do
you
want
to
know?

DR.
GIBBS:
You've
considered
it.
Did
you
do
any
post
hoc
analyses?
Or
perhaps
you're
opposed
to
those
of
what
sorts
of
effects
you
could
have
detected
given
your
final
sampling
design.

DR.
SILKEN:
They
ranged
from
kind
of
a
post
hoc
analysis
gives
you
better
estimates
of
variation.
You
can
go
back
and
ask
the
question
of,
given
those
variations,
what
could
you
have
detected.

And
we
found,
for
example,
for
laryngeal
muscle
that
we
had
plenty
of
power.
That
was
not
an
issue.

We
found
that
for
aromatase
and
some
of
the
hormones,
you're
going
to
need
an
awfully
big
study
unless
you
really
want
­­
unless
you
only
want
to
detect
really
big
differences.
Kind
of
looking
at
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
175
slide
that
you
all
were
looking
at
up
there
earlier
about
the
different
effects
of
atrazine
at
.1,
1,
10
and
25,
you
had
power
to
detect
big
differences
relatively
easily.
And
if
big
differences
were
all
that
were
ecologically
relevant,
then
you
have
plenty
of
power.
If
you
wanted
to
get
down
and
say
I
really
wanted
to
fine
tune
this
and
be
able
to
differentiate
between
a
3
percent
response
and
4
percent
response,
we
didn't
have
that
much
power.

But
you
sort
of
want
a
trade
off
between
what's
ecologically
relevant
and
how­
big­
can­
you­
make­
it
type
thing.

DR.
ROBERTS:
Dr.
Kloas
and
then
Dr.
Kelley.

DR.
SOLOMON:
Could
I
just
add
something
to
the
comment?

In
designing
the
South
Africa
studies,
we
started
those
in
2001.

And
at
that
time,
our
initial
interest
was
laryngeal
dilator
muscle.

And
there
was
an
ongoing
study
in
the
lab
at
the
time,
and
that
was
the
effect
we
were
interested
in.
With
that
in
mind,
we
chose
the
best
possible
situations.
We
wanted
to
find
reference
ponds
where
we
could
detect
no
atrazine
at
the
time
of
the
study
which
was
the
season
before
we
started.
And
we
also
wanted
to
have
a
reasonable
number
of
ponds
within
reasonable
operating
distance
of
the
university
as
we
didn't
want
frogs
dying
in
the
field
because
we
couldn't
collect
and
empty
the
traps
fast
enough.
So
a
lot
of
logistical
problems
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
176
associated
with
getting
really
large
sample
sizes.
We
couldn't
go
very
far
a
field
for
practical,
logistical
reasons.

DR.
SILKEN:
Let
me
just
add
one
other
comment.
One
thing
that
we
discovered
in
looking
at
this
data
relatively
extensively
is
that
the
tank
effects,
both
in
the
field
studies
and
in
the
lab
studies,

were
such
that
if
you
were
to
pool
the
data
and
ignore
tank
effects,

you
get
yourself
in
some
very
unexpected
trouble.

And,
in
fact,
and
I
know
experimenters
do
this
all
the
time.

They'll
do
a
test
to
see
whether
there's
homogeneity
among
the
tanks.

And
if
it
passes
an
F
test
or
another
test
for
homogeneity,
then
it
passes
the
test
that
apparently
there
are
no
tank
effects.
So
you
pool
all
the
data
together,
and
then
you
do
a
test
for,
say,
treatment
differences.

We
found
that
when
you
do
that,
instead
of
having
a
5
percent
error
rate,
you
have
between
a
30
and
90
percent
error
rate
when
you
follow
an
F
test
for
homogeneity
with
then
pooling
the
data
and
checking.

So
as
far
as
your
power
is
concerned,
there's
another
aspect,

too.
And
that
is
the
tank
effects
and
whether
or
not
you
pool
animals
within
tanks.
A
very
dangerous
thing
to
do.
So
we
encourage
the
Panel
not
to
do
that
and
to
do
almost
all
of
their
analyses
on
a
tank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
177
basis.

DR.
ROBERTS:
Dr.
Kloas,
then
Dr.
Kelley,
Dr.
Denver.

DR.
KLOAS:
I
appreciate
very
much
that
you
mentioned
the
hypothesis
of
possibility
of
anti­
androgenic
effects
of
atrazine.
I
think
why
didn't
you
go
forward
to
test
this
hypothesis?
You
stated
just
something
about
androgen
effects
but
not
about
anti­
androgenic
effects.

So
for
instance,
I
would
have
liked
to
know
something
about
effects
on
5­
alpha
reductase
or
the
relationship
between
testosterone
to
dehydrotestosterone
because
I
think
we
all
are
aware
that
dehydrotestosterone
is
much
more
power
full,
it's
an
androgen,
which
is
leading
to
masculinization.
And
we
have
a
demasculinization
effect.
I
think
this
could
be
also
another
key
enzyme
or
another
key
target
for
having
adverse
effects
on
demasculinization.

DR.
ROBERTS:
Did
you
want
to
respond?

DR.
VAN
DER
KRAAK:
I
mean
I
think
the
answer
is
simple.
I
agree.
I
mean
it's
just
a
question
of
­­
we've
talked
about
a
number
of
potential
experiments.
We've
not
got
there
yet.
And
we'll
certainly
continue
to
consider
that.
And
if
that
wishes
to
be
recommendation
that
goes
forward
by
the
SAP,
we
and
I'm
sure
others
would
consider
that
very
seriously.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
178
DR.
GIESY:
If
I
could
respond.
Once
again,
I
have
to
completely
agree.
We've
thought
about
it.
We've
talked
about
it.
I
have
a
proposal
written
to
do
it.
So
I,
obviously,
agree
with
you
that
it's
a
thing
to
do.
But
it's
just
a
matter
of
time,
how
much
time
there
was.
And
also
it
was
driven
by
trying
to
look
at
what
was
being
published
in
the
literature
and
get
a
handle
on,
could
we
reproduce
that,
were
those
mechanisms
of
action
that
were
being
proposed
plausible.

So
in
part,
if
we
could
just
design
experiments
ourselves
and
not
be
looking
at
the
literature,
certainly,
we
would
have
agreed
with
you
and
gone
straight
ahead
to
do
that.

DR.
ROBERTS:
Dr.
Kelley
then
Dr.
Denver
and
Dr.
Herringa.

DR.
KELLEY:
So
these
are,
once
again,
questions
about
the
field
data.
From
the
mark
and
recapture
data
and
from,
of
course,
you
know
the
size
of
the
ponds,
can
you
give
me
an
estimate
of
how
big
the
ponds
were
and
how
many
frogs
there
were
in
each
pond?
Were
these
high­
density
ponds?
Low
density
ponds?

DR.
DU
PREEZ:
I
would
say
medium
to
high
density.
In
the
report
that's
been
submitted,
we
gave
the
estimate
surface
area
of
the
ponds.
And
they
varied
from
small
to
really
big
ponds.
And
the
estimated
populations
that
we've
calculated
corresponds
very
well
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
179
with
the
size
of
the
pond.

From
my
personal
experience,
what
I
would
regard
as
a
good
Xenopus
pond
would
be
a
pond
with
muddy
water
because
predation
is
that
much
lower.
So
some
of
our
ponds
were
muddy,
muddy
water.

Some
of
the
other
ponds
were
clear.

DR.
KELLEY:
So
in
your
smallest
pond,
how
many
frogs
do
you
think
you
had?

DR.
DU
PREEZ:
The
population
estimates
that
we
did
were
very
conservative.
But
they
were
in
excess
of
300
in
the
smallest
pond.
And
then
a
couple
of
thousands
would
be
in
the
largest.

DR.
KELLEY:
Okay.
In
the
document
here,
the
laboratory
number
SA01A,
Table
3,
you
give
the
ages
of
male
and
female
frogs
collected.
And
yet
having
only
started
in
2001,
this
can't
be
actual
yearly
observations
because
it
goes
up
to
an
age
of
eight
unless
that's
eight
months.

DR.
DU
PREEZ:
Those
age
determinations
were
done
through
scoliotic
chronology.
So
we
did
histology,
sectioning
through
a
toe
and
then
by
counting
the
growth
rings
to
determine
the
age
of
the
frogs.

DR.
KELLEY:
So
the
eight
on
this
is
years?

DR.
DU
PREEZ:
Eight
years.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
180
DR.
KELLEY:
Is
eight
years.
Okay.

Now
my
final
question
has
to
do
with
your
measurement
of
the
estradiol
levels
in
the
two
sets
of
ponds.
There
are
published
estradiol
levels
for
laboratory
reared
animals.
And
I
guess
my
question
is
as
follows:
The
estradiol
levels
that
you
have
for
males
in
this
study
­­
I'm
now
looking
at
Figure
5,
page
29
of
138,
on
the
Eco
Risk
No.
MSU07
­­
in
which
your
measuring
plasma
estradiol
levels
picagrams
per
mil
in
male
and
female
from
both
your
control
ponds
and
your
atrazine­
sampling
ponds.
My
concern
here
is
I
don't
think
I've
actually
seen
such
high
estradiol
levels
in
males.
Females,
yes.

But
males,
no.
And
I'm
wondering
if
there
might
not
be
some
other
contaminant
in
the
ponds
that
don't
have
atrazine
that
could
be
accounting
for
this.

Both
my
studies
and
Shapiro's
and
almost
anybody
who
looks
at
vitellogenin
induction
fails
to
see
very
much,
if
any,
estradiol
in
adult
males.
And
yet
you
have
levels
that
are
1,200
picagrams
per
mil
which
is
way
in
excess
of
the
published
values.

DR.
GIESY:
We
did
those
measurements
in
our
laboratory
using
an
ELISA
technique.
They
are
what
they
are.
We've
provided
to
the
panel
our
QAQC.
But
we
would
be
glad
to
have
people
look
at
it.
And
if
there's
a
problem,
let
us
know.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
181
Relative
to
other
contaminants,
I
don't
know
other
than
what's
been
measured.
And
I
would
defer
to
Dr.
Solomon
to
talk
about
the
other
measurements
that
were
made.
But
I
don't
know
of
anything
that
might
cause
that
interference.

DR.
KELLEY:
But
the
literature,
people
looked
at
this
quite
carefully
because
they
were
interested
in
using
induction
of
the
vitellogenin
gene
as
an
assay
for
steroid
hormone
control
of
development.
And
Shapiro,
a
number
of
years
ago,
pointed
out
that
there's
this
very
interesting
memory
phenomenon.
And
he
used
male
Xenopus
because
they
had
no
induction
of
vitellogenin
gene.
And
he
showed
that
if
they'd
even
once
been
exposed
to
estrogen,
the
second
time
they
saw
estrogen
he
got
a
very
rapid,
very
large
response.
And
he
went
back
and
measured
estrogen
levels
during
development,
and
we've
measured
them
as
well.
Both
of
us
felt
that
the
data
were
consistent
with
males
having
almost
having
no
estrogen
available.

And
so
to
see
an
animal
with
circulating
levels
that
aren't
any
different
from
females,
they're
identical,
even
at
the
end
of
the
breeding
season
and
that
are
so
high
for
estradiol,
I
mean,
this
is
a
huge
level
for
estradiol,
makes
one
wonder
what's
going
on
in
both
sets
of
ponds.

DR.
GIESY:
That's
a
very
good
point.
We'd
love
to
be
able
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
182
split
samples
with
people
and
make
sure
the
results
are
accurate.

DR.
ROBERTS:
Dr.
Denver.
Dr.
Solomon.

DR.
SOLOMON:
Could
I
just
address
the
other
contaminants?

The
reference
ponds
were
located
in
a
totally
different
geological
area
where
corn
production
was
not
possible
just
because
of
the
type
of
soils.
There
were,
however,
cattle
present
in
the
system.
And
of
course,
cattle
do
come
and
drink
at
the
ponds.
So
there
is
a
possibility
of
contamination
with
both
urine
and
feces
in
cattle
which
may
result
in
presence
of
animal
estrogens
in
the
system.

But
whether
this
was
causing
any
response,
we
don't
know.

The
populations
in
those
ponds
in
terms
of
age,
size,
class,
sex
ratios,
et
cetera,
were
what
we
would
expect
them
to
be.
So
whatever
those
numbers
are,
they
at
least
appeared
to
us
to
not
be
affecting
the
populations.

DR.
ROBERTS:
Dr.
Denver.

DR.
DENVER:
Yes.
My
question
has
to
do
with
the
laboratory
studies.
And
it
goes
to
potential
vehicle
effects.
And
I
was
wondering
if
the
group
could
comment
on
that.
I've
noticed
that
there
are
a
number
of
instances
of
ethanol
having
an
effect.
And
I
think
tht
I
recall
from
reading
the
literature
that
you
published,
the
Carr
study,

that
the
atrazine
was
actually
not
dissolved
in
ethanol;
whereas
in
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
183
Hayes
study,
it
was.
And
I
wonder
if
you've
considered
possibly
there
being
an
interaction
effect
between
atrazine
and
ethanol
that
might
be
responsible
for
the
differences
between
the
laboratory
studies
that
we've
seen
today.

DR.
CARR:
We
chose
not
to
use
ethanol
as
solvent
because
atrazine
soluble
in
water
up
to
30
milligrams
per
liter.
And
there
was
no
reason
to
use
ethanol
as
a
solvent
for
our
study
because
the
concentrations
were
low
enough
that
they
could
be
dissolved
in
an
aqueous
medium.

In
terms
of
addressing
the
differences
in
the
study
and
hindsight,
which
we
didn't
know
before
going
into
the
study,
it
might
be
important
to
look
at
that.
And
I
think
there
is
in
some
of
the
data
evaluation
records
concerns
about
using
ethanol
as
a
potential
solvent
when
it's
not
necessary.

And
that's
something
that
would
need
to
be
done.
We
haven't
done
that.

DR.
DENVER:
On
the
graph
that's
there,
there's
clearly
the
potential
for
effects
of
ethanol
on
some
of
the
parameters
that
you're
looking
at.

DR.
GIESY:
I
would
agree.
In
the
report
that
we
provided
to
you,
when
we
did
our
statistical
analyses,
we
compared
to
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
184
appropriate
control.
So
for
atrazine,
we
controlled
to
water­
only
control.
And
for
the
positive
controls,
the
DHT
and
estradiol,
we
compared
to
the
ethanol
control.
We
can't
get
enough
estradiol
or
DHT
into
solution
readily
to
do
the
studies
without
using
the
carrier
solvent.
So
that's
a
real
limitation
of
the
study.

So
we
chose
not
to
put
estradiol
across
everything
for
the
reasons
that
Dr.
Carr
pointed
out.
So
could
that
lead
to
a
difference
in
results
in
the
designs?
It
certainly
could.
And
I
think
that's
the
kind
of
thing
that
might
led
to
some
of
the
differences
that
you
see
among
studies.

DR.
SOLOMON:
If
I
could
just
add,
I
have
a
comment.
We've
done
some
studies
with
fish
exposed
to
estradiol
and
effulents
containing
substances
that
are
supposed
to
mimic
estradiol.
And
we've
had
a
lot
of
problems
with
ethanol
as
a
carrier
solvent
in
terms
of
mortality
and
lack
of
growth
in
the
controls
in
full
life­
cycle
studies.
So
we
try
to
avoid
it
when
we
can.
But,
obviously,
in
this
situation,
it's
the
only
way
to
get
it
in
so
you
minimize
it.

But
I
did
at
a
meeting
a
couple
of
weeks
ago
find
out
that
ethanol
is
a
good
inducer
of
mixed­
function
oxidizes
in
some
organisms.
So
it
may
be
something
worth
following
up
there.

DR.
ROBERTS:
Dr.
Herringa
then
Dr.
Green.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
185
DR.
HERRINGA:
Bob
Silken's
comments
on
these
inter­
tank
variability
in
some
of
the
lab
experiments,
it
prompts
another
thought
that
I
had
as
I
was
reading
these
papers.
And
that
relates
to
potential
genetic
isolation,
genetic
effects.
I
presume
that
in
the
laboratory
setting,
it's
not
possibly
clearly
in
the
pond­
field
studies,
that
genetic
isolation
is
probably
a
very
real
phenomenon
in
fact.
But
in
the
lab
studies,
is
it
standard
practice
to
split
­­
I
want
to
say
litters.
It's
not
litters
here
­­
but
whatever
the
egg
pool
is.
And
sometimes
it's
not
even
just
sort
of
a
biological
strain
you're
using,
but
you're
actually
field­
capturing
animals.
Do
you
have
fertilized
and
then
randomize
them
between
the
control
and
the
dose
levels?

DR.
CARR:
In
our
studies,
yes,
we
mixed
eggs
between
breeding
pairs.
In
our
studies,
we
used
five
or
seven
breeding
pairs.

The
eggs
were
randomly
selected
and
assigned
to
treatment
tanks.

DR.
HERRINGA:
If
I
may
follow
up.
Dr.
Carr,
is
it
your
impression
that
in
these
studies
that
that's
standard
practice
or
would
be
standard
practice
for
these
types
of
studies?

DR.
CARR:
It
is
in
our
standard
operating
procedure.
Again,

there
are
no
standardized
and
validated
tests
for
these
chronic
exposures.
I'm
trying
to
recall
the
ASTM
guidelines
for
FTEC
studies.
I
don't
know
if
anybody
remembers
off
the
top
of
their
head
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
186
the
numbers
of
animals
to
be
used
in
those.

But
it's
our
sense
that
for
statistical
purposes,
it's
better
to
use
multiple
breeding
pairs
than
one
breeding
or
two
breeding
pairs.

DR.
ROBERTS:
Dr.
Silken.

DR.
SILKEN:
To
follow
up,
Steve,
the
eggs
were
randomly
split
up
among
treatment
groups
in
the
Michigan's
MSUO3
study
as
well
as
well
as
they
were
in
Dr.
Carr's
study.

DR.
ROBERTS:
Dr.
Green.

DR.
GREEN:
I
just
want
to
get
a
feel
for
the
kind
of
atrazine
exposure
the
frogs
in
the
wild
might
have
had.
And
one
way,
I
think
it
would
help
me
to
do
that
would
be
if
you
could
tell
me,
over
the
time
period
preceding
your
collection
of
the
new
metamorphs
and
the
young
juveniles,
could
you
tell
me
things
like
what
was
the
average
daily
temperature,
what
was
the
average
daily
rainfall.

And,
of
course,
I'd
be
interesting
in
knowing
the
water
quality
parameters
of
the
pond
water,
for
example,
do
you
know
what
the
ammonia
levels
got
to
at
the
worst
and
the
best?
Then
I
could
gauge,

you
know,
did
they
experience
a
period
of
drought
where
atrazine
might
be
at
its
highest
versus
rainfall
where
you're
going
to
have
a
dilute
run­
off?

DR.
DU
PREEZ:
In
this
study,
we
measured
those
couple
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
187
parameters
on
a
weekly
basis
and
later
on
bi­
weekly.
As
we've
collected
water
samples,
we
measure
temperature.
We
measure
dissolved
oxygen.
We
measured
conductivity,
pH.
And
that's
all
been
documented
in
our
reports.

DR.
SOLOMON:
There
was
some
additional
water
quality
analyses
conducted
and,
actually,
sediment
as
well
just
for
routine
water
chemistry
parameters.
This
was
done
less
frequently
than
the
sampling
measurements.
But
we've
characterized
and
other
components
of
the
system.
And
that's
also
in
the
reports.

There
were
differences
between
the
reference
and
the
control
­­

I'm
sorry
­­
the
reference
and
the
exposed
sites
in
terms
of
some
of
the
major
ions,
calcium,
sodium
and
some
of
those.
But
in
general,

there
didn't
seem
to
be
any
difficulties
except
some
of
the
elements
were
relatively
high
in
concentration.
But
the
analysis
was
conducted
on
unfiltered
water
samples.
So
we
don't
know
the
speciation
of
some
of
the
metals
in
the
water
because
it
was
total
element
analysis
rather
than
soluble.
Doing
it
again,
one
might
look
at
filtered
water
or
look
for
soluble
form
rather
than
suspended
forms
of
some
of
the
elements.

DR.
GREEN:
If
I
could
follow
up.
I
haven't
gone
back
to
pull
this
information
out
quickly.
If
you
could
summarize
for
me,
was
there
a
period
of
heavy
rainfall
at
some
point
that
might
explain
low
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
188
levels
of
atrazine
and
no
effect.
That's
really
what
I'm
looking
for.

Versus
a
period
of
drought
which...

DR.
DU
PREEZ:
During
this
study,
as
I've
mentioned
early
on,

we've
had
more
than
double
then
annual
rainfall.
So
it
was
a
very
wet
season.
That's
one
point.

Average,
minimum,
maximum
temperatures
was
spot­
on
with
a
10­
year
mean.
So
there
was
no
really
very
cold
or
very
hot
periods
during
this
study.
But
the
rainfall
was
double.

DR.
ROBERTS:
Okay.
I
have
a
very
quick
question.
Dr.

Silken,
I
saw
him
leap
to
the
table.
Did
you
have
something
you
wanted
to
throw
in
on
this.

DR.
SILKEN:
I
just
wanted
to
follow
up
with
Louis'
response
that
the
rainfall,
did
that
increase
the
concentrations
or
decrease
them?

DR.
DU
PREEZ:
Increased
the
concentration?

DR.
SILKEN:
Of
atrazine.

DR.
DU
PREEZ:
No.
We
had
a
definite
dilution
of
atrazine
with
this
excessive
rainfalls.
But
all
these
crafts
are
in
the
reports,

the
temperature,
the
rainfall,
everything
has
been
reported.

DR.
SOLOMON:
I
have
a
comment
to
that.
Work
done
in
relation
to
another
component
of
the
atrazine
risk
assessment
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
189
extensive
modeling
done
by
Marty
Williams
has
looked
at
this
issue
of
rainfall
and
dilution
in
ponds,
specifically
to
try
and
estimate
concentrations
that
might
occur
in
ponds.
And
he
sees
in
his
models
basically
the
same
thing
as
if
you
have
an
out­
flow
to
the
pond,
which
many
models
interestingly
enough
don't
have,
with
a
lot
of
rainfall,

you
actually
end
up
with
lower
concentrations
than
if
you
have
moderate
rainfall.
If
you
have
no
rainfall,
you
get
not
runoffs
and
no
contamination.
So
highest
concentrations
would
be
in
moderate
rainfall
years,
which
preceded
our
year
and
have
followed
our
year
of
study
in
South
Africa.

DR.
ROBERTS:
My
quick
question,
it
was
for
Dr.
Van
Der
Kraak.
It's
just
a
clarification.
When
you
discussed
your
causation
criteria,
one
of
them
was
temporality.
And
you
made
the
statement
that
you
didn't
think
temporality
was
met
because
of
the
prior
existence
of
some
of
these
phenomenon
like
testicular
oocytes.
Were
you
really
viewing
that
in
a
qualitative
sense
or
in
a
quantitative
sense?
In
other
words...

DR.
VAN
DER
KRAAK:
That
was
more
in
a
qualitative
sense.

DR.
ROBERTS:
That's
what
I
thought.
I
just
wanted
to
be
clear
on
that
point.

DR.
VAN
DER
KRAAK:
Yes.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
190
DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
So
let
me
follow
up
on
the
testicular
oocytes.

So
in
other
published
work,
my
understanding
is
that
you
can
actually
detect
the
testicular
oocytes
only
when
you
serially
section
through
the
testes
and
then
you
see
the
occasional
oocyte;
right?
So
maybe
that's
just
normal.

But
in
the
25
picagrams
per
liter
treated
atrazine
group
in
the
Carr,
et
al.,
study,
were
there
frank
­­
were
there
testes
that
upon
visual
inspection
had
testicular
parts
and
ovarian
parts?
We're
not
talking
about
the
stray
oocyte
but
were
frankly
hermaphroditic
comparable
to
other
reports
in
the
literature.

How
did
those
testes
look
really?

DR.
CARR:
The
animals
that
we
looked
at
in
our
study,
were
Stage
66.
99
percent
of
the
controls
were
sexually
differentiated.

And
in
our
study,
it
was
very
clear
to
see
animals
that
shared
both
male
and
female
gonadal
characteristics.

The
most
common
finding
when
we
actually
did
the
histology
on
the
ones
that
were
intersex
at
the
gross
morph
level
was
that
we
could
detect
an
ovarian
cavity,
for
example,
in
the
females
and
call
it
a
female­
like
gonad.
But
most
cases
in
the
males,
the
gonads
were
smaller.
And
so
it
was
more
a
difference
in
the
shape
and
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
191
organization
at
the
gross
level
that
we
were
reporting.

Now,
one
of
the
features
we
did
see
even
in
the
estradiol
exposures
were
rostral
deformities
that
resemble
some
of
the
things
that
Chang
and
Witschi
had
reported
back
in
the
'
50s.
We
saw
that
in
both
the
estradiol
and
25
part
per
billion
atrazine
group.

Those
weren't
entirely
that
common.
In
fact,
the
intersex
­­
and
this
was
another
point
I
was
going
to
bring
up
­­
was
fairly
rare.

When
we
looked
at
300
animals
in
the
23
part
per
billion
atrazine
group,
we
saw
it
in
12.
So
in
our
study,
it
was
a
relative
low
incidence.
But
they
stood
out
fairly
clearly.
It
was
fairly
easy
to
detect.

DR.
KELLEY:
So
if
I
could
follow
up.
I
think
the
point
I
want
to
make
is
that
the
testicular,
having
a
couple
oocytes
in
your
testes,

you
know
­­
I
mean,
this
just
may
be
a
normal
thing,
nothing
to
get
excited
about.
But
having
a
gonad
that
has
frank
ovarian
parts,
I
think
might
be
a
qualitatively
different
phenomenon.

So
the
fact
that
there
might
have
been
oocytes
throughout
all
of
the
literature
and
the
occasional
testes
wouldn't,
actually,
I
don't
think,
bear
on
the
question
of
intersex.
And
the
Panel
has
had
informal
discussions
about
how
many
Xenopus
they've
opened
up
and
how
many
intersex
they've
ever
seen
in
their
entire
life.
And
I
have
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
192
to
tell
you,
I
have
never
seen
an
intersex.
And
I'm
sorry
to
report,

I've
opened
up
thousands
upon
thousands
of
Xenopus.

Yes,
testicular
oddities.
Sometimes
only
one
testis.
But
a
gonad
that's
hermaphroditic?
Un­
huh.
At
least
not
in
normal
adult
lab
populations.

DR.
CARR:
We've
never
seen
anything
that
was
grossly
hermaphroditic
in
our
atrazine
animals.
The
intersex
we
used
as,
in
terms
of
the
terminology,
was
to
mean
that
we
couldn't
identify
it
as
male
or
female
at
the
gross
level.
We
did
not
find
testicular
oocytes
in
our
intersex
animals.

DR.
KELLEY:
Again,
I
guess
I
suggest
we're
going
to
have
to
go
forward
and
grow
these
animals
up.

DR.
CARR:
Absolutely.

DR.
KELLEY:
Because
it
may
be
a
phenomenon
that
becomes
more
obvious
as
the
animals
get
older.

DR.
ROBERTS:
Dr.
Matsumura
and
then
Dr.
Coats.

DR.
MATSUMURA:
I
was
just
wondering
what
the
mechanisms
which
can
create
these
kinds
of
effects
if
there
is
effect.
I
think
you
went
pretty
describing
the
major
hypothesis.
And
you
didn't
mention
anything
about
the
LH
or
prolactin
or
gondaltropins.
Did
you
do
any
experiment,
or
were
you
just
guessing?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
193
DR.
VAN
DER
KRAAK:
I
apologize.
We
were
talking
between
us
at
the
beginning
of
your
question.

DR.
MATSUMURA:
Well,
of
course,
we
must
get
some
mechanistic
information.
And
you
mentioned
about
the
hypothalamus,
pituitary,
LH,
FSH;
and
you
didn't
say
anything
about
data.

DR.
VAN
DER
KRAAK:
Correct.
And
I
think
there
are
a
couple
of
responses
to
your
question.
The
first
one
is,
in
formulating
the
questions
that
were
raised
in
the
White
Paper,
we
did
go
into
a
discussion
about
the
possibility
of
looking
at
aspects
of
various
hypothalamic
hormones,
LH,
FSH,
and
the
like.
The
question
there
becomes
one
of
what
hypothesis
is
one
expecting
to
be
testing.
And
then
we've
got
some
specific
questions
that
we
need
to
do
some
additional
biological
detective
work
in
terms
of
if
the
question
is
related
sexual
differentiation,
is
there,
in
fact,
significant
expression
of
LH
and
FSH
genes
at
that
time
in
development
and
whether
it's
a
viable
hypothesis
to
test.

We've
identified,
again,
as
a
priority
­­
pardon
me.
We've
identified
it
as
a
potential
question,
but
we've
certainly
not
gone
there.
I
know
Dr.
Giesy
has
talked
about
it
extensively
in
our
panel
meetings,
and
we
all
agree.
It's
a
question,
again,
of
time
and
effort.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
194
In
terms
of
are
there
other
hypotheses
that
are
out
there
that
one
could
test?
Sure.
I
think
the
number
of
hypotheses
that
one
could
generate
are
not
endless,
but
there
are
large
numbers
of
those
that
it
could
be.
I
mean
there
a
number
of
genes
that
are
turned
on
during
sex
differentiation.
We
could
go
systematically
and
look
at
the
expression
of
every
one
of
those
genes.
Or
we
could
take
a
molecular
approach
to
try
to
evaluate
those.
But
I
think
the
approach
that
we've
attempted
to
adopt
was,
on
the
one
hand,
let's
look
at
some
of
the
obvious
candidates.
And
then
number
two,
if
we
have
evidence
of
a
frank
effect
that's
reproducible
and
robust,
then
let's
go
back
in
and
do
those
directed,
mechanistic
studies
at
that
point
in
time.

Otherwise,
it
tends
to
be
a
little
bit
of
a
fishing
expedition.

DR.
MATSUMURA:
I
understand.
It's
priority.
This
is
one
mammalian
people
have
really
found
an
effect.
Sos
there's
no
question
that
the
Long­
Evans
rats,
this
is
a
major
finding.
So
that's
why.

DR.
ROBERTS:
Dr.
Giesy.

DR.
GIESY:
Well,
as
usual,
Professor
Matsumura,
you're
very
perceptive.
We've
certainly
thought
about
that.
We've
looked
at
the
mammalian
literature
and
actually
think
that
is
an
area
we
need
to
look
at.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
195
Now,
what
I
have
to
do
is
put
things
into
perspective.
We
have
really
two
ends.
One,
if
we
know
a
mechanism
of
action,
then
we
can
look
at
the
response
that
we
would
expect
to
find
and
use
that
as
our
measurement
endpoint
and
we
can
put
that
into
perspective
relative
to
environmental
risk
assessment.
Or
we
can
choose
the
endpoint
that
we
think
is
important
and
try
to
work
back
and
see
if
there's
a
plausible
mechanism.

Where
I
see
us
now
is
we're
in
the
middle.
We
don't
know
what
endpoint
we
need
to
look
at
because
we
don't
know
the
critical
mechanism
of
action.
And
if
we
did,
then
that
would
be
better.
But
what
EPA
is
saying
in
their
White
Paper
is
that
we're
not
going
to
look
into
those
mechanisms
of
action
until
we
have
looked
at
a
couple
particular
endpoints.

I
think
I
speak
for
the
panel
when
we
would
say,
well,
there
may
be
other
critical
mechanisms
of
action.
We've
looked
at
aromatase
a
lot.
A
lot
of
that
proposed
mechanism
of
action
is
based
on
work
that
was
done
in
my
laboratory.
We've
looked
at
that.
And
at
this
point
in
time,
we
could
give
you
some
suggestions
on
how
you
could
further
test
that
specific
hypothesis.
But
at
this
point
in
time,

my
feeling
is
it's
more
efficient
to
look
other
places.
And
so
we've
identified
the
hypothalamus
and
pituitary.
Hypothalamic
pituitary
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
196
axis
is
a
key
place
to
look.

Now
how
do
you
look
at
it?
Well,
you
have
two
choices.
One,

you
could
look
for
specific
responses.
And
we've
developed
a
preliminary
proposal
to
do
that.
Or
as
Professor
Van
Der
Kraak
indicated,
you
could
look
for
more
general
responses.
And
some
work
he's
done
in
his
laboratory
using
things
like
differential
display,
are
useful
techniques
that
we
could
ask
the
question,
is
anything
change
in
that
axis
and
work
from
that
position.
So
that's
another
idea
you
might
think
about.

But
we
couldn't
agree
more.
I
don't
think
at
this
point
in
time
we
can
key
in
on
any
one
specific
mode
of
action.
Absolutely.

DR.
ROBERTS:
Dr.
Coats.
Oh,
Dr.
Solomon,
did
you
want
to
respond
also?

DR.
SOLOMON:
Just
to
follow
up.
And,
Fumio,
thank
you.

The
effects
in
the
Sprague­
Dawley
rat
were
reproducible
and
consistent
at
different
times,
different
labs,
et
cetera.
Once
the
mechanism
was
understood
in
the
fact
that
it
was
specific
to
Sprague­
Dawleys
that,
I
think,
was
very
helpful.
I
was
at
the
SAP
here
in
Crystal
City
a
few
years
ago
to
listen
to
that
discussion,
so
it
was
very
interesting
to
see
it
being
used
in
a
risk
assessment
context.

But
we
haven't
yet,
at
least
in
our
own
hands,
been
able
to
get
a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
197
reproducible
robust
response
that
we
can
use
to
track
back
to
a
potential
mechanism.
Given
that,
obviously,
we
would
love
to
do
it.

DR.
ROBERTS:
Dr.
Coats
and
then
Dr.
Isom.

DR.
COATS:
In
the
field
studies,
I
guess
it
was
particularly
in
the
South
African
one,
there
are
quite
a
few
other
triazine
or
metabolites
note
there.
I
was
curious
about
the
water
levels
that
are
reported.
Are
they
filtered
samples
or
unfiltered?
And
are
there
sediment
values,
and
do
you
think
those
would
be
important
or
not?

DR.
DU
PREEZ:
If
you're
referring
to
if
they're
filtered
or
not,

you're
referring
to
the
weekly
sampling
of
the
water?

DR.
COATS:
The
water,
yes.

DR.
DU
PREEZ:
No.
Well,
I'm
not
­­
I
wasn't
involved
in
those
analyses.
But
to
the
best
of
my
knowledge,
they
were
not
filtered.

DR.
SOLOMON:
The
analyses
were
done
by
Piet
Johnson
from
Rensburg
at
the
Potchefstroom
University
using
GC
mass
spec.
And
he
used
SPE
cartridges
and
liquid­
liquid
extraction
using
both
methods
to
confirm.
And
then
he
used
unfiltered
water.
But
it
was
filtered
through
the
SPE
cartridges
which
were
then
eluted.
So
anything
that
was
trapped
on
a
solid,
would
have
been
extracted.
And
then
the
liquid­
liquid
were
extracted
from
the
direct
water
samples,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
198
no
filtration.

Sediments
samples
were
looked
at
and
no
pesticides
were
detected
in
the
sediment
samples.
Given
the
KD­
binding
constance
of
atrazine
which
is
relatively
water
soluble,
that's
consistent
with
what
we
would
expect,
that
nothing
was
present
in
the
sediments.

DR.
COATS:
Thanks.

DR.
ROBERTS:
Dr.
Isom.

DR.
ISOM:
Dr.
Giesy,
I'd
like
to
follow
up
on
your
comment
you
just
made
a
few
moments
ago
about
aromatase.
The
EPA
is
proposing
that
that
will
be
the
second
tier,
that
is,
if
we
can
pass
the
first
tier,
to
start
to
look
at
that.
And
you
made
the
comment
that
perhaps
­­
I'll
paraphrase
you
that
perhaps
we're
looking
in
the
wrong
direction.

Yet
if
you
look
at
the
literature
Sanderson's
group
has
shown
in
human
cells,
tumor
lines,
that
it
atrazine
induces
aromatase.
And
I
think
at
higher
concentrations
perhaps
even
inhibits
so
that
you
can
get
that
inverted
U­
shaped
dose
response
curve.

I'd
like
to
have
you
follow
up
on
that
comment
then
and
any
other
studies
that
you
have
done,
could
you
describe
those
with
that
enzyme?

DR.
GIESY:
Absolutely.
Well,
yeah,
I'm
an
author
on
those
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
199
papers;
so
I
do
know
about
them.
I
think
it's
important
maybe
to
explain
a
couple
things
about
how
those
studies
were
done
and
why
they
were
done
and
I'll
do
that.

But
let
me
answer
your
question
very
specifically.
I
think
what
I
meant
was,
if
we
look
at
aromatase,
if
there
are
open
issues
there,
I
think
we
need
to
take
a
fundamentally
different
approach
than
we've
taken
previously.
Not
to
dismiss
it
entirely.
But
if
we
do
go
ahead
with
that,
and
I
do
agree
with
EPA
that
it
shouldn't
be
first
tier.
But
if
we
go
ahead
with
it,
I
think
we
have
to
do
it
in
a
way
that
we
look
in
specific
tissues.
And
we
can
do
that
by
QRTPCR.

In
our
laboratory,
we've
now
developed
a
molecular
beacons
to
do
that.
So
we
can
do
the
quantitative
PCR.
I
think
it
would
be
important
if
we
follow
that
mechanism
up
that
we
do
it
through
the
developmental
stages
where
we
know
there
are
changes
in
aromatase
expression
and
we
do
it
in
a
tissue­
specific
way.

First
of
all,
my
comments
were
not
to
just
look
in
a
gross
way
because
I
don't
think
we'll
see.
So
I
think
it
is
stage­
dependent
and
it's
tissue
specific.
So
if
we
do
go
that
way,
that's
what
we
need
to
do.

But
at
the
same
time,
based
on
the
literature
we
do
know,
it
would
be
appropriate
at
the
same
time
to
look
at
other
plausible
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
200
mechanisms.
And
one
of
the
other
plausible
mechanisms
is
through
the
hypothalamic
pituitary
gonadal
axis.
I
think
we've
got
enough
evidence
as
Professor
Matsumura
point
out
to
not
completely
dismiss
that.

Why
do
I
say
that?
Because,
in
fact,
I
think
some
of
the
effects
that
we
do
observe
are
more
consistent
relative
to
what
can
be
caused
through
that
mechanism
of
action
relative
to
testosterone
depression
for
instance.

Now,
beyond
that,
the
studies
that
Thomas
did
when
he
was
in
my
laboratory
were
designed
specifically
to
try
to
understand
why
we
observing
what
we
considered
to
be
anomalous
results.
And
that
is
we
knew
that
atrazine
didn't
bind,
at
least
in
our
hands,
to
the
estrogen
receptor.
But
in
some
cell
lines,
we
did
see
what
looked
like
estrogenic
responses.
The
question
was
why.

Now
the
first
experiment
I
had
Thomas
do
was
just
to
dose
them
with
perpronobutoxide
because
I
though
maybe
what
we
were
looking
at
were
metabolites
that
were
being
formed,
hydroxy
metabolites
which,
in
fact,
might
be
estrogenic.
Subsequently,
we
tested
all
those.
They
basically
weren't
estrogenic.
But
in
doing
that,
we
thought
we
were
knocking
out
the
MFO
activity
that
might
form
the
metabolites;
when,
in
fact,
what
we
were
doing
most
likely
was
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
201
knocking
out
the
aromatase.

So
we
did
those
studies
in
part
to
understand
mechanistically
why
we
were
observing
this
effect
in
vitro.
They
were
done
at
pretty
high
concentrations.
They
were
done
at
30
micromolar,
which
is
about
6
parts
per
million
in
the
tissue.
And
they
were
done
­­
and
the
result,
then,
of
that,
too,
was
­­
and
we
looked
at
message.
We
looked
at
expression,
and
we
look
at
the
activity.
Depending
on
which
one
you
looked
at,
the
range
of
response
that
we
got
was
somewhere
between
two­
to
four­
fold.
Not
a
huge
response.

So
I
don't
dismiss
it
entirely
because
I
think
we
haven't
investigated
at
these
time­
specific
and
tissue­
specific
responses.
So
if
we
do
go
ahead,
that's
where
we
need
to
look.
But
I
really
think
based
on
the
literature
that's
out
there,
we
shouldn't
dismiss
the
other
potential
pathway
at
the
same
time.
That
was
my
point.
Not
that
you
shouldn't
consider
it
at
all.
But
how
you
do
it
if
you
do
it
and
not
forget
about
this
other
pathway
which
I
think
is
also
consistent
with
the
results
that
were
observed.

DR.
ROBERTS:
Dr.
Kendall,
I
think
that
before
your
panel
continues
with
the
next
aspect
of
your
public
comments,
my
panel
needs
a
break.

DR.
KENDALL:
Mr.
Chairman,
our
panel
yields
to
your
panel.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
202
Let's
take
a
break.

DR.
ROBERTS:
Let's
take
a
15­
minute
break.

(
Break;
session
resumed
at
3:
45
p.
m.)

DR.
ROBERTS:
Before
the
next
round
of
presentations,
Dr.

Green
had
a
question
as
a
follow
up
from
our
discussion
right
before
the
break.

DR.
KENDALL:
Very
good.

DR.
ROBERTS:
Dr.
Green.

DR.
GREEN:
Yes.
I
just
have
a
quick
question
for
clarification
to
Dr.
Du
Preez.
You
mentioned
that
some
of
the
answers
to
the
questions
I
posed
earlier
were
available
in
a
document
that
was
submitted.
Can
you
clarify
which
document
that
was
so
I
can
go
back
and
make
sure
we
all
have
the
same
details?

DR.
DU
PREEZ:
The
documents
that
we've
submitted
to
the
Panel,
that
would
be
SA01B
and
SA01C.

DR.
GREEN:
And
in
those
documents,
there
are
the
details
regarding
water
quality
analysis
on
the
ponds,
frequency
­­

DR.
DU
PREEZ:
Everything
is
in
there.

DR.
GREEN:
­­
changing
and
rainfall
and
that
sort
of
thing.

Okay.

DR.
DU
PREEZ:
Including
the
raw
data.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
203
DR.
GREEN:
Thank
you.

DR.
ROBERTS:
Let's
then
proceed
with
the
next
round
of
presentations.
The
Panel
will
have
opportunity
to
ask
more
questions
after
those.

DR.
KENDALL:
Thank
you,
Mr.
Chairman.
As
we
earlier
indicated,
we
would
lead
off
with
the
core
presentation
by
Dr.
Van
Der
Kraak.
And
we,
as
a
panel,
are
very
impressed
with
your
panel.

You
came
very
well­
prepared.
And
we
do
accept
the
criticism
and
welcome
the
comments
and
have
gained
a
great
deal
of
future
insight
related
to
how
we
would
proceed
with
our
research.

I
think
the
panel
can
relate
to
particularly
the
last
couple
of
years,
spent
a
lot
of
effort
and
we've
amassed
now
a
great
deal
of
information
and
a
lot
of
manuscripts
are
stacking
up
that
are
going
out
for
review
as
we
speak.
So
this
discussion,
interaction
is
welcomed.

And
from
our
core
presentation,
our
panel
is
continuing,
as
I
speak,
to
evolve
data.
And
I
wanted
each
of
them
to
have
a
brief
opportunity
to
­­
I
didn't
realize
we
would
have
as
many
questions
as
we
did.
But
we
welcome
them.
But
I
wanted
each
panel
member
to
have
a
chance
to
briefly
address
this
SAP
to
give
you
an
opportunity
to
further
understand
what
their
contribution
was
to
our
overall
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
204
effort.
And
I
emphasize
our
efforts
have
been
collective.
And
our
publication,
we
have
discussions
on
our
publications
and
they
are
collective.

And
I
also
wanted
to
make
a
comment
about
Dr.
Bob
Silken.

We
probably
ought
to
make
him
an
honorary
panel
member.
He's
made
himself
so
available.
He
sits
in
on
all
panel
meetings
now,

conference
calls
and
so
on.
And
he's
very
gracious
with
his
contribution,
although
he
drives
us
crazy
sometimes
related
to
all
of
his
questions.
But
I
think
it's
made
our
science
better.

But
anyway,
I'd
like
to
begin.
Dr.
Van
Der
Kraak
gave
our
core
presentation.
Dr.
Giesy,
I'd
like
for
him
to
comment
as
he
feels
appropriate.
And
the
Panel
is
welcome
to
ask
any
questions
of
our
scientists
as
we
proceed.
And
then
we
will
close
our
comments
today
after
this
period
by
offering
you
some
responses
to
the
White
Paper
questions
from
our
panel
that
can
be
shared
with
you.
So
Dr.
Giesy.

DR.
GIESY:
First
of
all,
I'd
just
like
to
say
I
think
this
is
an
excellent
panel.
I'm
very
impressed
by
how
well­
prepared
everyone
is,
but
you're
all
experts.
And,
two,
to
reinforce
that
I
think
the
EPA
White
Paper
was
an
excellent
document.
They
have
a
really
tough
task
to
do,
to
balance
all
this
and
try
to
find
a
way
forward.
So
I
thought
they
did
an
excellent
job.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
205
We've
been
involved
in
this
for
a
couple
of
years
now.
I
guess
what
I've
learned
from
that
is
that
we
started
from
not
much
and
have
made
a
lot
of
progress.
But
it's
certainly
a
work
in
progress.

Certainly
things
aren't
perfect
and
can
be
developed
much
more
in
the
future.

My
main
interest
in
really
in
the
mechanism
of
action,
what
is
the
plausible
mechanism
for
how
atrazine
may
cause
effects
so
we
can
put
into
context
what
are
the
right
endpoints
to
measure,
what
we
refer
to
in
risk
assessment
as
measurement
endpoints,
what
would
be
the
most
sensitive
and
relevant
endpoint
to
measure
that
then
could
go
into
an
assessment
endpoint;
which
is
ultimately
what
the
EPA
has
to
deal
with.
And
so
I
feel
strongly
that
we
do
need
to
know
what
the
critical
mode
of
action
is.
And
that's
my
interest.
So
anything
relative
to
where
we
are,
where
we
go
relative
to
that
is
something
I'm
interested
in.

So
we've
set
up
some
studies.
When
we
started,
we
wanted
to
work
with
some
native
species.
The
protocols
aren't
all
completely
worked
out
for
that.
One
of
the
grad
students
in
my
lab,
Katie
Cody,

did
a
lot
of
work
just
to
be
able
to
figure
out
how
to
do
a
study
with
green
frogs.
Having
done
that,
if
the
EPA
asks
me
should
we
use
green
frogs,
I'd
say
no
because
there
are
some
real
issues
with
time
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
206
metamorphosis
for
green
frogs.

But
I
think
we
have
learned
a
lot
as
a
panel.
And
I
think
as
a
scientific
community,
things
are
unfolding
in
this
whole
endocrine
disruptor
area.

With
that,
you
know,
I
think
this
panel
can
have
a
huge
impact
on
the
future,
where
we
go,
where
EPA
goes,
where
the
science
goes.

And
I
think
that
would
just
be
great.
But
I
don't
have
any
other
specific
comments.

We
do
have
one
ongoing
field
study.
What
Dr.
Kendall
wanted
us
to
do
was
get
you
up
to
speed
on
anything
that's
been
done
since
the
reports
that
we've
provided
to
you.
From
my
laboratory,
there
isn't
anything
else
really.
But
we
do
have
an
ongoing
field
study.
We
have
students
in
the
field
right
now
that
will
continue
with
all
the
warts
and
imperfections
of
trying
to
do
field
work.

But
other
than
that,
you
have
everything
that
we've
done
at
this
time.
And
if
you
have
any
further
questions,
I'd
be
glad
to
try
to
field
them.

DR.
ROBERTS:
Any
question?
Dr.
Kelley.

DR.
KELLEY:
So
I'd
like
to
comment
and
I
think
this
also
applies
to
Dr.
Hayes
about
the
use
of
this
laryngeal
dilator
muscle
cross­
sectional
areas
and
assay.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
207
So
we
studies
the
development
of
the
larynx
in
Xenopus.
And
there
are
two
features
that
I
think
are
important
for
your
use
of
it
as
an
assay
for
masculinization.
First
of
all,
let
me
point
out,
it's
a
very
good
assay
for
masculinization
as
you
show
yourselves.
It's
a
very
androgen­
sensitive
organ.
And
it's
certainly
extremely
sexually
dimorphic
in
adulthood.

So
the
cross­
sectional
area
that
you
guys
measure
represents
two
things,
the
size
of
muscle
fibers
and
the
number
of
muscle
fibers.

And
we
actually
never
measured
cross­
sectional
area
because
it
confounds
the
two.
We
looked
at
number
of
muscle
fibers.
And
we
also
showed
that
the
ability
of
androgen
to
cause
growth
of
the
larynx,
both
hypertrophy
and
hyperplasia,
was
dependent
on
prior
exposure
to
thyroid
hormone.
So
in
an
assay
system
where
your
animals
are
taking
a
long
time
to
metamorphose
­­
all
right.
So
let's
say
normal
is
two
weeks.
These
guys
are
taking
a
month
and
a
half
or
longer,
we've
also
shown
that
the
animals
are
exposed
to
androgen
during
that
period.
So
it's
possible
that
variability
in
time
to
metamorphosis
can
contributed
substantively
to
variability
in
whether
you
see
a
sexually
differentiated
laryngeal
cross­
sectional
area.

Now,
that
all
washes
out
by
the
time
the
animal
is
three
months
old.
So
you
don't
have
to
worry
about
that
any
more.
So
I
make
a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
208
plea
for
you
guys
to
stop
looking
at
laryngeal
cross­
sectional
area
at
Stage
66.
I
don't
think
it's
very
appropriate
time.
And
just
give
it
a
couple
months
more.

We
divided
post­
metamorphic
development
into
six
stages
that
are
very
well­
characterized
by
laryngeal
weight.
You
can
standardize
all
of
your
animals
no
matter
how
long
it's
taking
them
by
body
weight
and
laryngeal
weight
to
those
stages.
And
it
should
be
possible
to
come
up
with
some
well­
characterized,
low
variability
assay
for
whether
atrazine
has
an
effect
on
masculinization
by
using
that
assay
at
a
slightly
later
time
period.

Anyway,
that's
my
input
on
the
laryngeal
bioassay
which
has,
in
fact,
been
variable.
But
I
think
it's
been
variable
for
reasons
of
rearing
variability
in
terms
of
length
to
metamorphosis
in
the
studies
that
pretty
much
every
has
done.
And
that
makes
concordance
between
the
studies
very
difficult.

DR.
KENDALL:
Good
points.

DR.
GIESY:
Excellent
point.

DR.
ROBERTS:
Any
other
questions
for
Dr.
Giesy?
Okay.

Let's
move
on.

DR.
KENDALL:
I
did
want
to
make
sure
of
one
thing.
Dr.

Kelley,
you
mentioned
two
references,
two
papers
on
the
hormonal
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
209
measurements.
And
we
need
to
make
sure
we
get
those
for
Dr.
Giesy.

DR.
KELLEY:
I
have
the
papers.
I'd
be
happy
to
give
them
to
you.

DR.
KENDALL:
Very
good.

DR.
CARR:
I
really
don't
have
any
new
data
to
present.
And
I
don't
have
any
additional
specific
comments
that
weren't
addressed
in
the
previous
session.

I
did
have
a
general
comment.
I
think
there
was
discussion
earlier
this
morning
regarding
the
subjectivity
of
some
of
the
gonadal
assessment.
And
you
can
only
come
to
that
conclusion
when
you
look
at
all
the
different
terminology
that's
used
to
assess
intersex,

hermaphrodite,
et
cetera.

But
one
thing
to
remember
in
our
study
and
most,
if
not
all,
of
the
other
studies,
is
that
these
are
analyses
and
the
treatments
are
conducted
with
no
knowledge
of
what
the
treatments
actually
are.

These
are
blind
tests
in
essence.
And
so
I
feel
very
confident
about
our
data
in
terms
of
when
we
see
something
that
is
unusual
and
this
is
an
intersex,
it's
very
different
from
what
we
would
expect
to
see
in
a
normal
male
or
female.
So
I
think
that
reduces
some
of
the
subjectivity
at
least
in
Xenopus.

We
have
finished
an
experiment
in
Rana
pipiens
that
the
things
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
210
were
a
little
bit
different
there.
We
don't
have
data
to
share
on
that
yet.

But
that's
really
all
I
had
to
say.
And
I'll
be
happy
to
answer
any
additional
questions.

DR.
ROBERTS:
Any
questions
for
Dr.
Carr?
Great.
Thanks.

Dr.
Smith,
did
you
want
to
proffer
something?

DR.
SMITH:
We
have
continued
the
work
in
Iowa,
the
field
study
in
Iowa,
primarily,
the
laboratory
component
of
it.
From
the
histology
standpoint,
we
have
found
so
far
one
testicular,
one
animal
with
testicular
oocyte
which
turns
out
to
be
about
.6
percent
of
the
total
number
of
animals
that
we
have
analyzed.

We
have
also
taken
the
plasma
for
testosterone
analysis
and
gone
over
the
period
of
collection
because,
as
I
said
earlier,
the
representative
time
representing
Period
1,
2,
and
3
which
would
be
late
spring,
early
summer,
later
summer
and
early
fall.
And
the
indication
there
is
that
there
is
time­
dependent
change
in
plasma
testosterone
concentration.
So
time
becomes
a
variable
and
a
factor
that
has
to
be
taken
into
consideration.
However,
over
the
period,

there
is
no
significant
difference
for
plasma
testosterone
when
you
compare
the
adult
animals
during
the
specific
period.

The
other
observation
is
that
during
that
period,
there
is
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
211
significant
difference
between
the
adult
and
the
juvenile
plasma
testosterone
concentration
using
a
60
gram
cut­
off
point
as
the
difference
between
the
adult
versus
the
juvenile
for
that
comparison.

We
are
presently
in
the
process
of
utilizing
the
steriological
approach
that
we
use
for
the
South
African
Xenopus
laevis
adult
testicular
histological
analysis
to
evaluate
the
fractional
volume
sperm,
spermatocytes,
spermatogonia
and
what
we
consider
as
other
cells,
which
includes
blood
vessels,
connective
tissue.
And
that
data
will
be
made
available
pretty
soon
as
to
the
contribution
from
that
standpoint.

And
that's
where
we
are
presently.
I'll
entertain
any
questions
from
you.

DR.
ROBERTS:
Questions
for
Dr.
Smith?
Great.
Thanks
for
the
update.
Dr.
Du
Preez.

DR.
DU
PREEZ:
Last
night
I
prepared
a
quick
PowerPoint
presentation,
so
I'm
going
to
walk
you
through
a
couple
of
slides.

This
a
picture
that
I've
quickly
inserted
here
to
give
you
an
image.
This
is
one
of
the
larger
ponds
that
we've
used.
This
specific
pond
was
referred
to
as
Site
E6.

We
did
four
studies
in
South
Africa.
SA01A
was
the
evaluation
of
the
sites
where
we
characterized
the
different
sites,
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
212
populations,
the
mark
and
recapture
studies.
And
based
on
this
study,

we
identified
five
experimental
sites
and
three
reference
sites.

Phase
B,
SA01B,
was
the
study
of
a
one­
field
use
season.
We
on
a
weekly
basis
at
first
and
later
bi­
weekly
collected
water
samples
that
were
analyzed.
Elemental
scans
were
performed
on
water
samples
as
well
as
on
sediment
samples.
And
then
SA01C,
where
we've
collected
the
specimens.
The
blood
samples
were
collected,

shipped
over
to
Dr.
Giesy's
lab
where
they've
conducted
the
hormone
aromatase
analyses.
Gonads
were
shipped
over
to
Dr.
Smith's
lab.

And
he
just
mentioned
the
histology.

Then
SA01D
is
the
study
on
the
microcosm
that
I'm
going
to
expand
a
little
bit
more
on.
And
then
just
for
interest,
I've
been
busy
with
the
first
study
at
this
stage
where
we're
going
to
look
at
the
reproductive
cycle
of
Xenopus.
Our
period
of
months,
we're
going
to
quantify
the
nuptial
pads,
the
cloacal
folds,
the
hormone
levels,
and
so
forth.

If
we
now
focus
on
the
microcosm
study,
this
formed
part
of
a
thesis
of
Alaric
Uester.
We
had
a
microcosm.
We
had
12
ponds,
and
this
was
the
layout.
Three
reference
ponds,
three
ponds
with
one
microgram
per
liter,
three
with
10,
and
three
with
25.

These
were
initially
earthworm
pits
that
we
referred
to.
We
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
213
emptied
them,
lined
them
with
a
membrane,
filled
them
with
water,

add
some
microphytes,
leave
them
for
six
months
to
stabilize,
and
then
started
the
experiment.

Individual
microcosms
were
covered
with
that
frame
with
hail
netting.
And
that
would
be
to
keep
predators
out,
primarily
dragonfly,
because
the
dragonfly
larvae
is
the
one
animal
that
you
don't
want
in
a
microcosm,
and
also
birds.

Just
a
couple
of
tadpoles
there
to
show
that
the
tadpoles
schooled
in
my
opinion
in
a
natural
fashion.
And
at
Stage
66
animal,

they
are
in
the
water.
The
water
quality
from
measured
concentrations
and
from
a
visual
inspection
was
very
good.

And
there
is
a
set
of
microcosms.
What
I
want
to
point
out
in
this
slide
is,
from
a
logistical
point,
we
had
a
problem
that
not
all
the
ponds
were
exposed
to
the
same
amount
of
sunlight.
You
can
see
that
these
first
four
ponds
are
shaded
here
by
a
tree.
Then
there
are
a
couple
of
ponds
in
the
middle
that
received
more
sunlight.
And
then
on
the
other
side,
again,
two
ponds
that
were
semi­
shaded.

The
water
temperature
was
much
lower
than
you
would
expect
in
a
natural
pond.
And
this
had
an
effect
on
the
development.
And
we,
indeed,
experienced
a
delayed
development.
But
that's,
in
my
opinion,
not
a
concern
in
this
study.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
214
This
is
the
actual
recorded
atrazine
levels.
At
one
instance,
in
one
of
the
reference
ponds,
we
did
detect
0.1
microgram
per
liter
atrazine.
Again,
I
was
not
worried
about
this
because
two
weeks
later
it
was
gone.
I
think
it
was
either
an
error
or
contamination.
But
to
make
sure,
I
went
back
and
I
split
the
reference
pond
samples
in
two.

I
compared
the
specimens
collected
from
that
specific
Pond
No.
3
with
the
other
reference
samples,
and
there
was
no
difference.

Then
on
cross
morphology,
that
would
be
a
normal
male,
the
female.
And
this
is
the
type
of
deformities
that
we've
observed.
And
the
only
deformity
that
we
found
was
discontinued
testes.

Now,
I'm
actually
a
parasitologist
working
on
parasites
of
amphibians.
And
for
over
15
years
or
more,
I've
literally
opened
up
thousands
of
frogs.
And
this
is
what
you
see
from
time
to
time,
a
single
testis.
Usually,
when
there's
a
single
testis,
it
would
be
larger.

You
do
find
discontinued
testis,
but
I've
never
come
across
a
true
hermaphrodite
in
frogs
that
I've
collected.
And
I've
worked
in
both
corn­
growing
areas
where
atrazine
would
be
applied
and
areas
more
pristine
reserves
and
so
forth.

Based
on
the
gross
morphology,
we've
observed
a
4
percent
in
the
reference,
1.3
in
the
1
microgram
per
liter,
.6
in
the
10,
and
3.7
in
the
25
microgram
per
liter.
So,
again,
no
statistical
difference
here.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
215
And
this
is
up
to
the
point
where
we've
reported
in
the
study
SA01D.
In
the
recent
weeks,
we've
conducted
the
histology.
And
I'm
quickly
going
to
report
on
our
findings
there.

This
is
my
slave,
Alaric
Uester.
We
selected
54
specimens
per
concentration,
randomly
selected.
That
adds
up
to
216
of
the
600
specimens
selected
for
the
gross
morphology.
Six
specimens
were
lost
in
the
preparation
for
the
histology.
120
to
200
sections
per
specimen,
four
to
six
slides
per
specimen.
That
all
adds
up
to
more
than
a
thousand
slides.
And
31,000
histological
sections
that
were
examined
individually.
So
that's
why
I'm
not
very
popular
with
my
students
at
this
stage.

What
we've
observed
and
what
I
want
to
point,
testicular
oocyte
would
appear
and
disappear
in
seven
slides
sectioned
at
6
micrometer.

So
what
I
wanted
to
stress
with
this
is
it's
not
good
to
check
every
20th
section.
You
have
to
check
every
one
if
you
want
to
detect
the
testicular
oocytes.

A
testis
with
a
testicular
oocyte
there,
a
cross­
section,

longitudinally
section,
and
the
ovary.
And
you
can
clearly
the
lumen
in
the
middle.

This
is
what
the
histology
revealed.
No
less
than
56
percent
of
the
animals
from
the
reference
ponds
had
testicular
oocytes.
Again,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
216
56
in
the
1
microgram,
58
in
10,
and
38
in
the
25.

Just
this
past
week,
we
did
another
study
where
we
looked
at
grow­
outs.
Now
the
rate
columns
represent
Stage
66
and
the
blue,

10­
month
grow­
outs.
And
there
is
a
reduction
in
the
number
of
oocytes.
Now
I
must
point
out
that
this
blue
set,
the
10­
month
grow­
outs
is
only
10
specimens
per
concentration.

Then
we
calculated
the
number
of
specimens
with
testicular
oocytes
where
we
found
the
oocytes
in
the
single
testis
or
both
testes
and
no
statistical
significant
difference
there.

If
we
now
look
at
the
mean
number
of
oocytes
per
specimen
with
testicular
oocytes,
we
find
that
there
is
an
average
of
about
10
oocytes
per
specimen.
But
there
is
a
significant
reduction
if
we
look
at
the
10­
month­
old
grow­
outs.
And
then
if
we
split
up
the
sample
and
divide
the
number
of
oocytes
in
categories,
zero,
1,
2
to
10,
11
to
12
and
so
on,
51
to
60,
we
find
the
following.
And,
again,
not
the
significant
reduction
in
the
number
of
oocytes.

The
maximum
number
of
oocytes
in
the
Stage
66
samples
was
in
this
25
microgram
per
liter.
And
that
was
58
oocytes
in
a
Stage
66
Xenopus.
The
maximum
for
the
10­
month
grow­
out
was
5.
Only
one
specimen
had
5,
three
had
2,
and
the
rest
had
a
single
oocyte.

So
from
this,
tadpoles
developed
slow
as
a
result
of
cold
water.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
217
I've
already
addressed
that.
Ponds
were
partially
shaded,
covered
with
hail
netting
and
did
not
have
a
shallower
part
as
you
would
find
in
a
natural
pond.
Often
these
Xenopus
in
a
natural
pond
would
school
in
the
shallower
water.

I've
already
addressed
this
point
of
the
atrazine
that
was
detected
in
this
one
reference
pond.
Again,
no
reason
for
concern
there.

No
gonadal
abnormalities
were
observed
in
the
females.
No
hermaphrodites
were
observed.
Males
showed
a
low
percentage
of
deformities
at
a
cross­
morphology
level.
A
high
percentage
of
frogs
had
one
or
more
testicular
oocytes
at
all
concentrations,
but
there
was
no
dose
response.
And
then
a
significant
reduction
in
the
number
of
testicular
oocytes
as
metamorphs
grow
older.

And
from
the
South
African
studies,
for
me
the
take­
home
message
would
that
atrazine
does
not
appear
to
affect
the
gonadal
development
of
Xenopus
laevis
at
environmentally
relevant
concentrations.
The
second
point
would
be
that,
in
my
opinion,
one
would
expect
that
if
atrazine
had
a
negative
or
adverse
effect
on
Xenopus
laevis,
it
would
reflect
in
the
population
dynamics
after
four
decades
of
intensive
use
of
atrazine.
And
in
previous
decades,
it's
been
documented
that
in
South
Africa
they
used
atrazine
by
the
tons.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
218
I
mean
they
really
applied
it.

Now
days,
they
use
more
conservative,
recommended
amounts.

But
surely,
if
there
was
something
to
be
worried
about,
you
would
see
it
in
the
Xenopus
populations
out
there.

And
then
in
my
opinion,
field
studies
and
microcosm
studies
does
have
its
limitations
as
greatly
pointed
out
by
the
EPA's
White
Paper.
But
I
don't
think
we
should
underestimate
the
value
of
field
or
microcosm
studies.
Because
in
the
end,
the
question
is
what's
happening
out
there.

Thank
you.

DR.
ROBERTS:
All
right.
Any
questions
from
panel
members.

Dr.
Green.

DR.
GREEN:
I
think
you
could
probably
anticipate
this
question
from
me.
How
cold
was
the
water?
And
you
didn't
mention
anywhere
in
this
document
the
conditions
of
the
water
quality,
pH
conductivity.
If
someone
were
to
try
and
reproduce
this
experiment,

it
would
be
very
important
to
have
that
information
if
you're
going
to
reproduce
this
microcosm
in
a
lab.

DR.
DU
PREEZ:
That's
correct.
That
data
is
available.
What
we've
handed
out
is
a
handout
to
summarize
this
presentation.
That's
not
the
full
report.
In
the
full
report,
we
have
all
the
other
data.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
219
DR.
GREEN:
But
in
my
opinion,
I
think
the
water
quality
and
the
parameters
that
define
this
microcosm
are
very
important,

especially
to
a
laboratory
Xenopus
person.
So
what
was
the
temperature
on
this?

DR.
DU
PREEZ:
Too
cold
to
swim
in.
We
did
not
measure
water
temperature
on
an
hourly
basis
for
example.
We
measured
in
this
study
at
10
o'clock
in
the
morning
when
we
took
the
water
sample
to
check
for
atrazine.
And
that
was
in
the
low
teens,
around
10
to
14,

which
is
lower
than
you
find
in
the
natural
population.

DR.
ROBERTS:
Any
other
questions.
Yes,
Dr.
Kelley.

DR.
KELLEY:
If
you
wouldn't
mind
my
asking
you
a
question
about
the
previous
field
study.
I'm
looking
at
the
mass
of
the
frogs
on
page
59
in
the
study
where
you
collected
them
from
the
three
reference
ponds
and
the
five
experimental
sites.
And
I
just
want
to
make
sure
I'm
reading
this
right.

So
I
think
the
goal
was
to
get
20
of
each
sex
from
each
pond.

But
it
looks
to
me
like
that
was
difficult
to
achieve
at
the
experimental
sites.
Was
there
another
reason
for
that
that
you
could
think
of?

DR.
DU
PREEZ:
I
was
really
upset
when
I
wanted
to
collect
the
specimens,
especially
from
Site
E1.
Because
with
this
excessive
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
220
rains,
some
catfish
were
washed
into
that
pond
and
literally
wiped
out
the
Xenopus
population
in
that
pond
to
such
an
extent
that
we
had
difficulty
to
collect
the
target
number
of
specimens.

DR.
KELLEY:
So
your
feeling
that
the
difficulty
in
collecting
at
this
sites
where
you
have
full
representation
in
your
reference
sites
from
the
cow
pastures,
but
a
sparser
representation
for
the
experimental
site
was
due
to
predation.

DR.
DU
PREEZ:
That's
my
opinion.

DR.
KELLEY:
Thank
you.

DR.
ROBERTS:
Any
other
questions?
If
not,
let's
move
on
to
Dr.
Gross.

DR.
GROSS:
My
comments
will
also
be
relatively
brief.
We
have
been
coordinating
now
for
the
past
year
and
a
half
the
studies
looking
at
another
one
of
the
mini­
field
studies,
looking
at
the
cane
toad
in
sugar
cane
agricultural
areas
in
South
Florida.
Those
studies
are
currently
continuing
and
ongoing,
so
we
have
little
to
add
at
this
point
that's
new
that
the
Panel
has
not
already
seen
the
documents
provided.

I
think
several
things
can
be
noted,
though,
from
the
materials
that
have
been
provided.
First,
I
think
as
indicated
in
EPA's
review
which
we
fully
agree
with,
these
experiments
were
simply
preliminary
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
221
examinations,
mostly
fact­
finding,
range
finding,
site
identification,

species
identifications
kind
of
studies.
And,
therefore,
it's
very
difficult
from
these
studies
if
not
impossible
to
draw
any
conclusions
relative
to
causation
of
single
chemicals
or
even
mixtures
of
chemicals
or
other
factors.

None
the
less,
the
studies
that
we've
seen
do
demonstrated
some
effects.
We
think
that
these
effects,
though,
are
vastly
different
in
this
species
compared
to
what
else
has
been
reported
for
Xenopus
or
the
various
ranid
species
that
had
been
examined
because
we're
looking
at
a
toad
species,
the
cane
toad,
which
has
a
bitters
organ
and
rudimentary
ovarian
structure,
which
basically,
I
think,
allows
a
differential
kind
of
response.
We
believe
these
responses
currently
to
be
mostly
in
the
adult
or
sub­
adult
rather
than
during
the
metamorphic
phases
that
have
been
described
previously
for
the
other
species.

None
the
less
as
I
indicated,
we
are
continuing
with
these
studies.
Actually,
we've
been
continuing
since
February,
have
little
data
yet
to
add
because
these
studies
are
ongoing.
I
think
the
comments
of
the
panel
today
have
been
mostly
addressed
as
we
designed
this
second
phase
of
study,
the
second
tier;
and,
hopefully,

we
will
have
better
answers
in
the
next
year.

But
I'll
be
happy
to
entertain
any
questions
you
might
have
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
222
relative
to
the
studies
we've
conducted
on
cane
toads.

DR.
ROBERTS:
Any
questions
for
Dr.
Gross?
All
right.
Dr.

Solomon.

DR.
SOLOMON:
I'm
sort
of
wrapping
up
this
with
a
few
brief
comments.
And
my
role
in
the
panel,
apart
from
helping
formulate
the
reports,
was
to
bring
the
concept
of
risk
assessment
to
the
data
that
we
were
developing.
And
as
we
heard
earlier
from
Dr.
Bradbury
and
others,
risk
assessment
involves
a
component
of
integration
of
exposure
and
effects
data.
And
as
you
probably
very
well
know,
there
is
an
excellent
data
base
for
atrazine
concentrations
from
the
ecosystem
through
the
efforts
of
the
U.
S.
Geological
Survey
and
other
labs.
We
have
very
large
data
sets
in
the
hundreds
of
thousands
of
data
points.
Maybe
not
hundreds
of
thousands,
but
certainly
tens
of
thousands
of
data
points.
So
there
is
a
good
data
base
to
go
out
there
to
work
on
a
measured
concentrations
to
compare
to
effect
concentrations.

We
also
have
good
modeling
data
that
allows
us
to
estimate
exposures
in
areas
where
we
don't
have
good
measured
values
and
helps
to
get
a
tighter
definition
of
temporal
variation
that
we
don't
get
from
sampling
once
every
week
or
two
weeks.

I've
been
somewhat
frustrated
in
this
because
the
other
side
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
223
the
equation,
the
effect
side,
has
not
materialized
to
the
point
where
I
can
use
this
in
the
risk
assessment
process.
And
I
think
Dr.
Van
Der
Kraak
pointed
out
in
his
slide
that
we
need
robust
and
consistent
concentration
responses
that
we
can
feed
into
the
risk
assessment
process
to
determine
probablistically
or
deterministically
what
types
of
responses
we
might
expect
to
see
in
the
environment.
So
that's
been
my
frustration.
And,
hopefully,
it's
a
frustration
on
the
Panel
as
well.
I
certainly
heard
that
when
I
read
the
White
Paper
as
well,
that
there
was
insufficient
data
at
this
point
to
do
a
risk
assessment.

Thank
you
very
much.

DR.
ROBERTS:
Any
questions
for
Dr.
Solomon?
Yes,
Dr.
Van
Der
Kraak.

DR.
VAN
DER
KRAAK:
They
overlooked
me.

Just
to
comment
in
terms
of
some
ongoing
studies
that
are
occurring
in
my
laboratory.
Some
of
the
experiments
are
being
done
in
relation
to
some
of
the
questions
that
Dr.
Kloas
addressed
in
that
we've
been
looking
at
more
of
the
responses
of
Xenopus
to
atrazine
in
terms
of
effects
on
steroidogenesis.
And
so
we're
looking
and
doing
some
mechanistic
studies,
trying
to
tease
apart
places
in
the
pathway
that
are
responsive
and
develop
the
methodology
in
order
to
do
in
vivo
exposures
and
to
take
gonadal
tissue
outside
of
the
animal
and
do
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
224
ex
vivo
incubations
to
test
for
specific
lesions
in
the
steroidogenic
pathway.
So
those
studies
are
just
ongoing.

DR.
ROBERTS:
Any
questions
for
Dr.
Van
Der
Kraak.
Dr.

Kendall,
I
think
there's
another
phase
of
your
public
comments.

DR.
KENDALL:
Yes.
Thank
you,
Mr.
Chairman.
We
were
encouraged
to
as
a
panel,
and
we
did,
respond
to
the
White
Paper
in
terms
of
the
questions
being
posed
to
the
SAP
and
would
offer,
Mr.

Chairman,
that
we
discuss
that
or
summarize
it
or
provide
this
for
the
record.
But
each
member
of
the
SAP
has
been
given
our
panel's
response
and
literature
backup
as
to
some
of
the
scientific
issues
we
see.
I
would
yield
to
how
you
recommend
that
we
proceed.

DR.
ROBERTS:
I
think
that
if
you
could
summarize
it
for
us,

your
opinions
and
responses
on
these,
that
would
probably
be
the
most
useful.
Let
me
be
sure
that
I've
got
it.

DR.
KENDALL:
This
document
is
entitled,
"
The
atrazine
Endocrine
Ecological
Risk
Assessment
Panel's
Response
to
The
United
States
Environmental
Protection
Agency's
Questions
in
the
`
White
Paper
on
Potential
Developmental
Effects
of
atrazine
on
amphibians.'"

DR.
ROBERTS:
All
right.

DR.
KENDALL:
The
date
is
June
17,
2003.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
225
ATTY2:
Got
it.
Thanks.
If
you
can
maybe
just
proceed
and
summarize
it
for
each
question,
that
would
be
very
useful.

DR.
KENDALL:
The
members
of
the
SAP,
of
course,
will
receive
your
charge
and
the
questions
probably
Thursday
morning.

Mr.
Chairman,
I
really
don't
want
to
read
each
question,
not
unless
you
want
me
to.
But
we
thought
we'd
at
least
project
them
on
the
board.
And
I
wanted
to
just
very
quickly
touch
on
some
of
the
responses
that
we
had
as
a
panel.

Initially,
we
as
a
panel
would
like
to
compliment
the
Agency
on
their
effort
in
bringing
this
White
Paper
together.
Again,
this
is
an
emerging
and
one
in
which
there's
not
a
lot
of
standardization
both
in
the
science
to
do
the
research
as
well
as
the
terminology.
And
would
encourage
the
SAP
to
engage,
not
just
the
science,
but
the
terminology
so
we
call
all
begin
to
speak
the
same
language.

Never
the
less,
in
terms
of
Question
1,
in
particular
1.
a.,
"
Does
the
SAP
have
any
comments
and
recommendations..."
In
general,
our
panel
supports
the
Agency's
evaluation
of
the
existing
body
of
data
and
generally
agrees
with
the
conclusions;
although
there
were
two
point
from
our
perspective
that
needed
to
be
considered.

We
as
a
panel
felt
that
the
Agency
focused
more
on
the
limitations
of
the
data
versus
what
the
data
could
offer
if
it
were
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
226
looked
at
in
a
robust,
comprehensive
manner.
For
instance,
the
limitations
of
low
concentrations
of
atrazine
in
the
reference
site,

relatively
great
inherent
variability
in
hormone
concentrations,
and
other
issues
such
as
the
time
to
metamorphosis.

Again,
we
accept
the
criticisms.
We
welcome
the
opportunity
to
address
these
criticism
in
the
future.
But
let
us
not
lose
sight
of
the
forrest
for
looking
too
hard
at
the
individual
trees.

The
second
point
that
we
think
merits
attention
is
the
Agency's
statistical
analysis
of
the
data.
Whenever
the
Agency
reanalyzed
our
data,
I
want
to
emphasize
for
any
Panel
member,
we
provided
all
raw
data
for
all
studies.
So
from
that
perspective,
one
could
take
this
and
analyze
this
as
one
deemed
appropriate.
Never
the
less,
the
Agency's
analysis
of
our
data,
if
they
found
a
statistically
significant
difference
that
our
panel
did
not
find,
it
was
not
always
because
EPA's
analysis
in
terms
of
how
they
approached
it
in
many
cases
EPA's
analysis
pooled
animals
together
over
sites,
tanks,
replicates
rather
than
approach
it
as
we
did
in
preserving
the
structure
in
the
data
and
including
the
animal
site,
tank,
and
replicate
in
the
analysis.

And
Dr.
Silken
was
adamant
as
to
how
we
approached
these
analyses
in
the
context
of
statistical
background.
And
we
would
off
that
it
just
be
considered
as
one
looks
at
all
of
these
data
together.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
227
In
terms
the
b.
part
of
Question
No.
1,
from
a
statistical
perspective,
the
Agency's
overall
characterizations
of
the
currently
available
studies
tend
to
treat
differences
that
are
found
to
be
statistically
significant
by
one
method
analysis
as
true
differences
that
are
biologically
significant
and
beyond
question
but
tends
to
treat
results,
in
terms
of
no
statistically
differences,
as
questionable.

So
as
we
approached
it,
and
I
heard
the
comments
today
as
my
colleagues
did.
We
approached
it
with
the
best
science
and
the
best
statistics
that
we
could
apply.
We
are
not
seeing
robust,
repeatable
statistical
differences.
Never
the
less,
we
would
welcome
the
SAP
to
look
carefully
at
this
as
well
as
assist
the
Agency
in
looking
at
their
statistical
approaches.

In
1.
c.,
a
number
of
studies
have
been
done
that
address
the
effects
of
atrazine
on
development
of
anurans
and
also
on
the
occurrence
of
testicular
oocytes
have
been
published
in
the
literature.

Others
that
should
have
been
included
include
Allran
and
Karasov
2000
and
2001,
and
Brown­
Sullivan
and
Spence
of
2003.
And
we
go
into
more
detail
on
that
particular
area.

In
Question
No.
2,
we
agree
that
field
studies
have
limitations
as
discussed
by
EPA.
While
these
limitations
are
acknowledged,
field
studies
we
believe
are
extremely
useful
in
a
weight
of
evidence
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
228
approach
and
the
results
are
still
relevant.

This
is
particularly
important
for
emerging
areas
of
science
where
we
need
to
make
decisions.
Data
are
coming
in.
And
we
believe
that
a
comprehensive
field
and
laboratory
integration
are
critical.

The
studies
conducted
in
the
field
we
believe
were
designed
to
test
the
presumption
of
risk.
That
is,
they
were
designed
such
that,
if
there
were
no
differences
locations
or
correlations
between
exposure
to
atrazine
and
responses,
it
could
be
concluded
that
atrazine
did
not
cause
these
effects
under
relevant
environmental
exposures.
And
we
would
welcome
further
deliberation
by
the
SAP
on
this
particular
approach.

Never
the
less,
Dr.
Du
Preez
comments
and
I
think
his
elegant
summary
of
some
of
the
works
going
on
in
South
Africa,
we
challenge
the
concept
that
if
atrazine
had
caused
robust,
sustained,

comprehensive
population­
level
effects
on
native
Xenopus
laevis
in
South
Africa
as
one
may
have
suspected
from
preliminary
studies,

then
it
would
have
been
reflected
in
disturbances
of
population
structure,
particularly
after
40
years
of
application
in
that
area.

Moving
on
to
Question
3,
in
terms
of
3.
a.,
we
believe
that
laboratory
studies
provide
a
plausible
basis
to
establish
a
hypothesis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
229
concerning
the
potential
for
atrazine
to
cause
developmental
effects
provided
that
sample
size
is
adequate
the
small
incidence
of
abnormalities
that
have
been
reported
in
some
studies,
and
these
have
been
relatively
small
in
some
studies,
and
provided
that
the
exposures
overlap
with
critical
windows
of
gonadal
differentiation,
and
provided
that
sampling
design
takes
into
account
the
variability
of
timing
in
gonadal
differentiation
in
different
anuran
species.

Many
of
these
issues
were
raised
today.
I
think
this
SAP
panel
is
highly
alert
to
these
concerns
and
I
think
will
address
this
nicely
and
assist
the
Agency
with
their
recommendations.
We
believe
that
the
overall
body
of
data
clearly
indicates
that
the
response
of
Xenopus
laevis
to
atrazine
varies
under
the
conditions
described
in
the
available
studies.
And
there
are
a
lot
of
reasons
for
this.
And,
of
course,
this
SAP
will
address
many
of
those
reasons.

Although
the
degree
to
which
differences
in
experimental
design
and
husbandry
influence
the
contradictory
findings
remains
a
matter
of
debate,
the
fact
that
a
relationship
between
atrazine
exposure
and
development
of
gonadal
abnormalities
is
not
consistently
found
raises
the
question
of
the
ecological
significant
and
the
relevance
of
observed
effects
of
atrazine
on
gonadal
differentiation.
We
think
that's
at
the
essence
of
the
questions
that
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
230
you
will
be
addressing,
Mr.
Chairman.

In
Subsection
b.
of
that
question,
to
date,
the
laboratory
and
microcosm
studies
have
followed
logical
assumptions
inherent
in
the
scientific
method.
And
we
believe
that
given
the
inconsistent
responses
observed,
it
is
not
possible
at
this
time
to
predict
how
Xenopus
laevis
will
respond
to
atrazine
under
yet
a
different
set
of
environmental
conditions.

Although
it
is
relatively
clear
that
atrazine
does
not
dramatically
affect
thyroid
function
and
does
not
influence
estrogen
receptor
activity
at
environmentally
relevant
concentrations,
the
ability
of
atrazine
to
influence
gonadal
differentiation
is
an
as
of
yet
unelucidated
pathway
cannot
be
predicted
for
available
studies.
And
we
are
still
searching,
as
Dr.
Giesy
and
other
members
of
our
panel
related
to
you,
for
these
underlying
mechanisms
of
action
for
potential
effects.

Give
the
lack
of
a
repeatable
effect
and
the
absolute
lack
of
evidence
for
a
cellular
mechanism
underlying
the
reported
effects
of
atrazine
on
gonadal
differentiation,
it
is
not
possible
at
this
time
to
predict
the
dose
response
relationship
or
the
rank
order
potency
of
atrazine
metabolites
relative
to
the
parent
compound.

In
Question
No.
4,
we
concur
with
the
conclusions
reached
by
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
231
the
EPA
that
there
is
no
evidence
to
conclude
that
atrazine
causes
induction
of
aromatase
in
frogs,
at
least
at
this
time.

In
Subsection
b.
of
that
question,
there
is
not
evidence
to
suggest
that
analytical
issues
are
a
major
factor
contributing
to
variability
in
plasma
sex
steroid
hormone
levels
or
aromatase
activity
measures
in
frogs.
I
might
add
that,
again,
I
emphasize,
our
projects
were
implemented
on
GLP­
like
performance
standards
under
SOPs
where
standards
of
performance
were
measured.
The
Eco
Risk
QA
unit
went
to
every
laboratory,
check
and
recheck
process.
So
we
did
take
into
account
analytical
and
measurement
performance.

At
least
for
the
research
associated
with
the
Eco
Risk
panel
here,
our
quality
measures
are
in
place.
And
we
believe
they
are
within
acceptable
bounds.
And
we
do
believe
that
perhaps
variations
in
analytical
results,
particularly
in
the
area
of
sex
steroids,
may
have
some
biological
underpinning
moreso
than
analytical
underpinning.

In
Subsection
c.,
we
believe
that
in
anurans,
sex
differentiation
is
sensitive
to
sex
steroids
during
critical
periods
of
development.

And
I
think
this
SAP
will
help
elucidate
for
the
Agency
how
one
might
approach
a
better
measure
of
that
for
repeatable
type
data
acquisition.
There
is
little
evidence
to
suggest
that
anurans
are
particularly
sensitive
to
estrogens
in
terms
of
the
induction
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
232
testicular
oocytes
following
sex
differentiation.
And
our
panel
has
discussed
this
issue.
So
this
area,
we
would
welcome
the
SAP
to
fully
explore
and
offer
some
perspectives
to
the
Agency
on
the
relevance
in
either
the
reproductive
or
ecological
relevance
of
these
testicular
oocytes.

Compounds
acting
as
estrogens
or
anti­
androgens
would
be
predicted
to
cause
feminization
of
males.
There
are
no
data
available
to
suggest
that
atrazine
functions
as
an
estrogen
receptor
agonist
through
binding
to
the
estrogen
receptor
or
induction
of
estrogen­
dependent
responses.
Similarly,
there
is
no
evidence
that
atrazine
binds
to
the
androgen
receptor
and
functions
as
an
anti­
androgen.

There
is
some
real
experts
on
this
SAP
related
to
this
particular
subject
area.
This
is
what
we
believe
to
date
as
the
state
of
the
science.
We
welcome
your
critical
review.

Moving
to
the
next
question,
just
a
few
more
minutes,
Mr.

Chairman.
I'm
trying
to
roll
through
this.

Question
5,
Subpart
a.,
our
understanding
of
spermatogenesis
in
anurans
lags
far
behind
that
for
mammals,
especially
for
amphibians.

It
is
not
known
if
accelerated
growth
precedes
histological
differentiation
of
gonads
which
would
influence
the
developmental
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
233
rate
at
early
stages
as
well
as
gonadal
morphology.
It
appears
that
sex
determination
exhibits
considerable
plasticity
and
the
ability
of
gonadal
differentiation
to
respond
to
environmental
factors
may
be
highly
adaptive.
Therefore,
it
may
be
resulting
in
some
of
the
differences
in
data
that
we
are
seeing
currently.

Gonads
in
Rana
curtipes
initially
differentiate
into
ovaries,
and
later,
in
the
prospective
males,
the
ovaries
degenerate
and
transform
into
testes.
This
represents
a
semi­
differentiating
type
of
gonad.

Thus,
interpretation
of
background
rates
of
ovotestes
and/
or
testicular
oocytes
occurrences
in
amphibian
species,
as
well
as
interpretation
in
the
context
of
environmental
exposure
and
risk
assessment,
requires
significant
experimental
evaluation
under
controlled
laboratory
conditions
in
addition
to
the
evaluation
of
populations
in
the
natural
habitats.

And
so
we
believe
this
SAP
can
contribute
much
thought
and
idea
in
the
proposed
future
laboratory
work
necessary
with
atrazine
and
frogs.
But
let
us
not
discount
the
importance,
relevance,
and
contribution
our
field
work
can
contribute
to
this
whole
subject
area.

In
subpart
b.
of
this
question,
there
is
no
information
currently
available
that
explicitly
tests
whether
the
presence
of
a
few
testicular
oocytes
would
result
in
any
impairment
of
reproduction.
Therefore,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
234
one
would
wonder
is
this
an
aberration
or
not.

The
presence
of
robust
populations
of
frogs,
and
this
is
debatable,
Dr.
Skelly,
and
we
respect
your
opinion
and
welcome
your
seat
on
the
SAP
in
contributing
to
the
population
understanding
of
this
subject
area.
But
there
were
presence
of
robust
populations
of
frogs
where
there
is
a
relatively
great
incidence
of
testicular
oocytes
further
argues
against
such
an
adverse
effect
of
this
condition.
This
is
point
we
want
to
make.

Recently,
the
research
accomplished
by
our
research
team
in
male
amphibians
does
not
suggest
that
the
presence
of
testicular
oocytes
and
ovotestes
results
in
the
reproductive
impairment
via
reduced
fertility.

In
Subsection
c.
of
this
question,
the
major
point
here
­­
we
addressed
other
points.
But
the
laryngeal
development
in
frogs,
as
Dr.
Kelley
has
mentioned,
is
a
sexually
dimorphic
process;
and
the
formation
of
a
larynx
capable
of
a
male
calling
behavior
is
androgen­
dependent.
Under
normal
conditions
the
laryngeal
dilator
muscle
of
the
male
Xenopus
laevis
is
larger
than
that
of
females.
It's
been
hypothesized
that
atrazine
could
decrease
plasma
concentrations
of
testosterone
in
Xenopus
laevis
by
up­
regulating
the
expression
of
aromatase,
the
enzyme
that
converts
testosterone
to
estradiol.
Yet
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
235
this
remains
to
be
proven.

And,
again,
this
whole
area,
I
think,
would
welcome
further
input
from
the
SAP.
It
is
probably
inappropriate,
we
believe
at
least
at
this
point
with
the
laryngeal
dilator
muscle,
to
be
using
this
effect
as
an
assessment
endpoint,
perhaps
a
measurement
endpoint.
But
this
point
may
be
moot,
since
at
least
at
this
point
in
time,
we
do
not
seem
to
have
reproducible,
sensitive
measures
of
potential
estrogenic
or
anti­
androgenic
effects
of
atrazine.

In
Question
No.
6,
our
Eco
Risk
panel
supports
EPA's
conclusion
that
the
data
currently
available
from
both
laboratory
and
field
studies
involving
a
wide
range
of
amphibian
species
does
not
support
the
hypothesis
that
atrazine
causes
development
effects
in
amphibians.
That
is
yet
to
be
proven.
Given
the
low
background
incidence
of
gonadal
effects
reported
in
most
studies,
particular
attention
should
be
devoted
to
sufficient
sample
size
so
that
the
statistical
power
of
the
study
is
sufficient
to
properly
test
the
null
hypothesis.
This
has
already
been
described
by
Dr.
Gibbs.

The
use
of
concentrations
greater
than
those
expected
in
the
environment
would
allow
evaluation
of
a
threshold
for
gonad­
specific
effects
to
be
determined.
Therefore,
levels
of
exposure
greater
than
25
ppb
would
be
necessary.
The
study
should
also
consider
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
236
possibility
that
effects
could
be
caused
by
metabolites
and/
or
environmental
break­
down
products
of
atrazine
and
related
triazines.

Hazard
and
risk
assessments
for
pesticides
and
other
substances
have
always
been
based
on
the
principle
of
dose
response
and/
or
concentration
response.
We
believe
there
is
neither
a
framework
nor
a
general
precedent
for
risk
assessment
based
on
U­
shaped,
dose,
or
concentration
responses.
We
will
welcome
SAP
deliberation
on
this
subject.

In
Question
8,
while
some
aspects
of
the
study
plan
proposed
by
the
U.
S.
EPA
are
reasonable;
again,
we
compliment
highly
EPA's
White
Paper
and
their
effort
to
engage
the
best
science
possible
on
this
whole
area
of
atrazine
exposure
in
amphibians.
It
may
be
unwise
to
base
the
decision
tree
solely
on
histological
effects
in
the
gonad
when
there
is
currently
no
reason
to
believe
that
this
the
most
sensitive
endocrine
response.
And
there
has
been
no
plausible
mechanism
of
action
that
has
been
suggested
by
studies
done
to
date.

Dr.
Kloas's
eloquently
argued
on
other
enzyme
approaches
and
other
endpoints
that
may
be
more
elegantly
sensitive
to
such
effects
is
being
considered
by
the
SAP.

In
the
b.
part
of
that
question,
it
appears
that
the
major
set
of
endpoints
is
covered
under
the
present
approach
that
EPA
has
offered
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
237
except
some
of
the
ones
I
just
mentioned.
Work
so
far
has
revealed
that
the
degree
of
gonadal
differentiation
at
completion
of
metamorphosis
is
highly
variable
within
species
and
especially
between
species.
In
fact,
there
is
no
a
priori
reason
to
suspect
that
gonadal
differentiation
would
be
timed
with
completion
of
metamorphosis.
So
it
is
unclear
why
the
Agency
is
using
this
same
sampling
time,
end
of
metamorphosis,
to
gauge
the
degree
of
sexual
differentiation.

And
I'm
sure
the
SAP
will
deliberate
and
offer
the
approaches
in
particularly
the
lab
component
of
these
future
proposed
studies
that
will
address
these
concerns
that
we
have.

In
Subsection
c.,
it
is
extremely
important
to
properly
design
for
possible
tank
and
site
effects.
It
is
also
important
to
include
possible
tank
or
site
effects
in
the
analyses
of
the
study
data.

Not
only
is
it
important
to
design
studies
with
multiple
tanks
or
sites
per
treatment
level,
but
it
also
important
to
include
multiple
tanks
or
sites
at
the
control
level.
Even
two
or
three
tanks
or
sites
at
the
control
level
may
be
insufficient
to
capture
the
true
tank­
to­
tank
variability.
And
this
has
already
been
discussed
by
Dr.
Silken
previously
at
the
table
today.

Based
on
initial
power
analysis
performed
by
our
Eco
Risk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
238
panel
prior
to
conducting
its
laboratory
studies,
a
minimum
of
eight
tanks
per
treatment
are
necessary
to
account
for
inter­
tank
variation.

In
terms
of
the
last
few
comments
in
Subsection
e.,
within
ranid
species
there
are
differences
in
the
expression
of
androgen
hormone­
dependent
secondary
sexual
characteristics.
This
will
have
to
be
take
into
account.
There
are
also
differences
at
the
hormonal
and
potential
differences
at
the
developmental
level
that
are
not
clearly
understood
in
ranids
which
would
point
to
caution
in
assuming
that
the
developmental
processes
and
the
mechanism
of
gonadal
differentiation
can
be
appropriately
tested
in
Xenopus
laevis
alone
which
argues
that
Xenopus
laevis
may
in
fact
be
an
initial
good
model
species
to
test.
But
we
should
not
count
other
native
frog
species
that
may
be
differing
in
their
sensitivity,
gonadal
differentiation,
et
cetera.

Many
ranids,
for
example
Rana
catesbeiana,
are
difficult
if
not
impossible
to
breed
under
laboratory
conditions,
meaning
that
eggs
would
have
to
be
collected
from
natural
ponds
with
unknown
exposure
histories.
Rana
pipiens
can
easily
be
breed
in
the
lab.
So
we
have
differences
of
how
these
various
species
will
reproduce
in
the
lab
and/
or
have
to
be
collected
from
the
environment.

In
Subsection
g.
of
the
last
question,
we
believe
and
we
hope
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
239
that
you
will
talk
about
this.
But
a
toxicokinetic
study
of
the
uptake,

organ
distribution,
and
depuration
of
atrazine
in
Xenopus
laevis,
rana
species,
as
well
as
other
frogs
will
be
useful
in
determining
whether
extrapolation
is
possible
between
frogs.
In
other
words,
we
offer
caution
that
to
take
a
non­
native
laboratory
model
and
try
to
extrapolate
this
to
robustly
to
our
native
frog
populations.

Thank
you,
Mr.
Chairman.
That
concludes
all
we
have
to
say
in
a
summary
nature
of
our
document.

DR.
ROBERTS:
Thank
you,
Dr.
Kendall.
Reminding
the
panel
that
our
opportunity
to
respond
to
these
questions
will
come
later,
are
there
any
questions
to
Dr.
Kendall
or
other
members
of
this
panel
regarding
their
rationale
or
basis
for
their
responses
to
these
questions.
Dr.
Kelley.

DR.
KELLEY:
I
wanted
to
make
sure
I
understand
your
summary.
As
I
understand
it,
you
felt
that
there
was
no
robust
and
reproducible
effect
of
atrazine
in
terms
of
development
of
intersex.

And
yet
in
the
Carr,
et
al.,
study,
you
say,
and
I
quote,
"
Exposure
to
either
estradiol
or
25
micrograms
atrazine
per
liter
increased
the
incidence
of
intersex
animal
based
on
an
assessment
of
gonadal
morphology."

So
do
you
mean
that
that
was
not
a
robust
effect?
I
mean,
there
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
240
were
11
replicates
of
that
particular
treatment
group.

DR.
CARR:
Again,
we
found
4
percent
intersex
in
that
particular
treatment.
In
terms
of
the
percentage
of
animals
that
were
affected,
I
wouldn't
consider
that
a
necessarily
robust
response.

Neither
was
the
qualitative
assessment
of
the
intersex.
Again,
we
had
initially
used
the
term
hermaphrodites
to
refer
to
these
gonadal
abnormalities.
But
they
weren't
true
hermaphrodites
in
the
sense
that
there
were
testes
and
eggs
sticking
out.
In
that
sense,
they
were
fairly
subtle.

And,
again,
when
we
looked
at
the
histological
level,
they
were
identifiable
as
male
or
female
at
the
histo
level.
But
there
were
differences
in
the
shape
and
structure
of
the
gonad
that
popped
out
at
the
gross
level.

DR.
KELLEY:
But
it
was
statistically
significant.

DR.
CARR:
Absolutely.

DR.
KELLEY:
It
was
reliable
within
your
groups.

DR.
CARR:
Right.

DR.
KELLEY:
So
you
would
conclude
that
this
particular
dose
of
atrazine
did
have
an
effect
even
if
you
felt
that
it
was
a
high
dose
and
it
was
a
small
effect.

DR.
CARR:
Absolutely.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
241
DR.
KELLEY:
But
still
you
would
stand
behind
the
effect.

DR.
CARR:
Yes.

DR.
KELLEY:
Okay.
Thank
you.

DR.
KENDALL:
That's
a
good
question.
And,
again,
in
the
presentation
what
we
emphasized,
we
went
through
enormous
planning
on
the
statistical
power
of
that
experiment.
And
had
not
that
statistical
power
been
so
strong,
we
probably
would
not
have
detected
that
effect.
But
we
stand
by
it,
although
it
was
small
and
we
do
not
believe
robust.

DR.
VAN
DER
KRAAK:
I
think
there's
an
additional
point
to
add
to
that.
There's
another
component
to
your
question,
though.

And
the
question
was:
Whether
the
response
was
reproducible
across
multiple
studies.
And
in
fact
­­

DR.
KELLEY:
Oh,
yeah,
within
your
study.
That's
all
I
asked.

You
did
11
replicates.
I
wanted
to
make
sure
it
was
reproducible
within
your
study.

DR.
VAN
DER
KRAAK:
Correct.

DR.
KELLEY:
Right.
Good.

DR.
VAN
DER
KRAAK:
The
question,
the
additional
point
was
whether
it
was
reproducible
across
studies.

DR.
KELLEY:
25
micrograms
at
that
dose.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
242
DR.
CARR:
Right.
And
­­

DR.
KELLEY:
I
really
want
to
know
about
your
study,
whether
you
stood
behind
that
conclusion.

DR.
CARR:
Right.
I
do.
And
the
thing
that
might
add
some
clarity
is
that
those
abnormalities
were
distributed
across
the
replicates.
They
weren't
from
one
tank.
That's
the
point
I
wanted
to
make.

DR.
KELLEY:
Great.
Thank
you.

DR.
KENDALL:
That's
a
fair
question.

DR.
ROBERTS:
Any
other
questions?
Dr.
Green.

DR.
GREEN:
One
last
question.
I
was
wondering
how
the
panel
felt
about
the
possibility
that
in
order
to
get
a
handle
on
the
significance
or
the
number
of
frogs
in
the
wild
population
in
different
species
that
live
with
ovotestes.
How
does
the
panel
feel
about
the
possibility
that
field
studies
would
involve
going
out
and
capturing
healthy
frogs
from
healthy
frog
populations
and
literally
having
to
kill
hundreds
of
them,
maybe
thousands
depending
what
the
statistician
tells
us,
just
to
verify
that
it
is
or
isn't
a
problem
in
the
absence
or
presence
of
atrazine?

DR.
KENDALL:
We
may
ought
to
wait
to
let
the
SAP
answer
that
question.
I
think
that's
a
great
question.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
243
DR.
KELLEY:
They're
not
our
frogs.
They're
SA's
frogs.

DR.
GREEN:
I
assume
there
would
be
some
in
the
United
States.
You
can
see
I
find
it's
a
bit
troubling.
Do
you
think
the
data
right
now
supports,
for
the
edification
of
the
panel,
that
we
consider
having
to
go
out
and
capture
healthy
frogs
to
determine
how
significant
this
is
in
a
wild
population?

DR.
KENDALL:
I
think
initially
what
we
believe
is
going
to
take
a
lot
of
frogs
to
have
the
power
to
detect
the
effects
because
we
don't
believe
the
effects
are
that
robust.
That's
a
very
good
point.

Never
the
less,
if
one
is
going
to
get
the
correct
kinds
of
data
for
future
purposes
to
regulate
this
chemical
relative
to
frogs,
we
will
need
to
take
the
lab
work
to
the
field
to
see
if,
in
fact,
these
effects
are
occurring
and
that
they
have
any
reproductive
and/
or
ecologic
consequence.

DR.
GIESY:
That's
an
excellent
question
and
one,
certainly,

within
my
group
we
consider
all
the
time.
We
have
ongoing
field
studies
that
we're
doing
where
we
are
collecting
a
lot
of
frogs.
And,

of
course,
we
do
that
under
our
animal
use
permits.
But
it's
something
we
don't
do
lightly.
And
I
don't
think
we
should
do
lightly.

And
I
would
encourage
the
Panel
to
consider
that
in
whatever
design
you
come
up
with
for
experiments.
We
would
not
want
to
do
that
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
244
wantonly.

DR.
DU
PREEZ:
One
can
also
look
at
museum
specimens.

There
are
literally
thousands
of
Xenopus
sitting
in
bottles
throughout
the
worlds.
But
it's
difficult
to
persuade
the
curators
of
those
collections
that
you
want
to
open
up
some
specimens.
They
are
sometimes
very
difficult.
But
that
is
a
possible
option.
And
now
we
know
why
frogs
are
declining.

DR.
ROBERTS:
Other
questions
from
the
Panel?
Sorry.
Dr.

Skelly.

DR.
SKELLY:
Perhaps
I
should
have
asked
this
before.
It
dawned
on
me,
thinking
about
this,
that
one
of
the
questions
I
wanted
to
ask
your
group
was
whether
you
have
looked
or
whether
you
plan
to
look
for
any
evolved
tolerance
to
atrazine.

DR.
SOLOMON:
Perhaps
I
can
answer
on
behalf
of
the
group.

We
certainly
thought
about
this.
Two
aspects.
One
is
has
there
been
a
genetic
bottleneck
that
these
frogs
have
gone
through.
And,
in
fact,

in
frozen
at
minus
80
degrees
celsius
we
have
blood
waiting
for
DNA
analysis
if
anybody's
interested
in
it.
We'd
welcome
contributions
from
the
Panel
and
the
audience
as
well.
Not
contributions
financially,
but
suggestions
as
to
how
we
do
that
in
case
you
get
me
wrong.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
245
The
other
issue
is
­­
I
mean
there
have
certainly
been
some
suggestions
of
resistance.
And
I
think
the
White
Paper
mentioned
and
we
know
that
there's
resistance
in
plants.
Plants
have
a
mechanism
of
action,
receptor
mechanism,
that
is
susceptible
to
selection,
change
in
the
protein
sequence.
Of
course,
plants
are
under
strong
selective
pressure,
so
they
have
evolved
being
selected
for
this
resistance.

For
this
to
have
happened
in
frogs,
I
suspect
we'd
have
to
have
a
fairly
strong
response
that
either
causes
death
or
change
in
reproduction.
And
for
my
part,
I
haven't
seen
that
yet.
In
my
own
experience
with
resistance,
I
think
you
need
one
of
those
or
both
of
them
to
actually
have
selection
take
place.

DR.
SKELLY:
Just
as
a
follow­
up.
I
guess
that's
my
point
is
that
there
are
two
ways
to
document
a
demographic
effect.
One
of
them
is
to
watch
a
population
for
a
really
long
time.
The
other
ways
is,
if
you
find
an
evolved
response,
that
implies
a
past
demographic
effect.

DR.
SOLOMON:
Yes.

DR.
KENDALL:
That's
a
good
point.

DR.
ROBERTS:
Any
other
questions?
Or
more
response
from
Dr.
Giesy.

DR.
GIESY:
The
issue
of
resistance,
I
think,
is
a
very
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
246
interesting
one.
And
one
that
I'd
like
to
hear
some
responses
from
the
Panel.
I
personally
have
some
ideas
from
a
molecular
biological
techniques
how
you
could
approach
that.
I
think
it
is
a
question
you
can
pose.
I
think
we
do
have
tools
that
we
can
look
at
that.
And
I
think
it's
probably
worth
looking
at.

DR.
ROBERTS:
Any
other
respondents?
Dr.
Kendall.

DR.
KENDALL:
I
just
want
to
say,
Mr.
Chairman
and
distinguished
members
of
the
SAP,
thank
you
for
the
opportunity.

We're
very
grateful
to
have
had
this
time
frame
to
discuss
these
issues
with
you.

I
also
appreciate
EPA
for
setting
up
the
opportunity
to
have
such
a
scientific
discussion
and
to
give
us
the
opportunity
to
look
towards
the
future.
We
also
appreciate
Eco
Risk
for
facilitating
our
efforts
between
our
universities.
And
we
appreciate
our
sponsor
for
giving
us
the
support
necessary
to
engage
what
we
believe
will
be
some
important
discussion
in
the
next
few
days.
And
we
welcome
hearing
from
you.
And
thank
you
for
your
criticism,
you
input,
and
particularly
your
patience
this
afternoon.

DR.
ROBERTS:
On
behalf
of
the
Panel,
I'd
like
to
express
our
appreciation
for
you
and
your
colleagues
willingness
to
come
here
and
discuss
your
research
in
very
open
fashion
with
the
Panel
and
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
247
answer
all
their
questions
regarding
it.
I
think
that's
been
very
helpful
in
us
understanding
these
issues.

DR.
KENDALL:
Thank
you.

DR.
ROBERTS:
All
right.
Well,
that
concludes
this
public
comment.
There
are
plenty
more
to
come.
But
at
this
point,
I
want
to
assess,
do
a
stamina
check
on
the
Panel.
The
next
public
commentor,

we
have
two
public
comments
from
the
registrant.
They
will
take
a
total
of
two
hours,
but
they
could
be
divided.
Potentially,
we
could
do
one
of
them.
Then
next
one
will
take
about
an
hour.

Okay.
I
can
see
by
the
looks
on
your
faces.
That
answers
my
question.
I
just
thought
I
would
bring
it
up,
but
I
got
a
pretty
clear
non­
verbal
response
on
where
we
stand
today.

In
view
of
that,
let
us
go
ahead
and
adjourn
the
meeting
today.

We
will
reconvene
at
8:
30
tomorrow
morning.
We
will
continue
with
public
comments,
the
first
of
which
will
be
two
public
commentors
from
the
registrant,
Sygenta.
And
then
we
will
proceed
with
other
public
commentors
beyond
that.

Paul,
do
you
have
any
announcements
before
we
close
for
today?

MR.
LEWIS:
Nothing
to
add,
Dr.
Roberts.
I'm
looking
forward
to
continuing
our
discussion
tomorrow
morning.
Thank
you.
248
1
DR.
ROBERTS:
Today's
session
is
closed.
We'll
reconvene
2
tomorrow
morning
at
8:
30.

3
[
Meeting
convened
at
4:
35
p.
m.]

4
­
oo0oo­
1
2
3
4
5
6
7
8
9
10
249
CERTIFICATE
OF
STENOTYPE
REPORTER
I,
Jane
F.
Hoffman,
Stenotype
Reporter,
do
hereby
certify
that
the
foregoing
proceedings
were
reported
by
me
in
stenotypy,

transcribed
under
my
direction
and
are
a
verbatim
record
of
the
proceedings
had.

_______________________________

JANE
F.
HOFFMAN
1
2
3
4
5
6
7
8
­­
9
10
­­
250
**
I­
N­
V­
O­
I­
C­
E****
****
I­
N­
V­
O­
I­
C­
E****

JANE
F.
HOFFMAN
TODAY'S
DATE:
6/
26/
03
DATE
TAKEN:
6/
17/
03
CASE
NAME:
FIFRA
SAP/
Potential
Developmental
Effects
of
Atrazine
on
Amphibians
TOTAL:
PAGES:
363
LOCATION
OF
DEPO:
Crystal
City,
VA
DELIVERY:
7­
day
SPECIAL
INSTRUCTIONS:
Tapes
to
be
included
