FIFRA
SCIENTIFIC
ADVISORY
PANEL
(
SAP)

OPEN
MEETING
JUNE
19,
2003
DAY
3
OF
3
Located
at:
Crown
Plaza
Hotel
1489
Jefferson
Davis
Highway
Arlington,
VA
22202
Reported
by:
Frances
M.
Freeman
2
1
C
O
N
T
E
N
T
S
2
3
Proceedings...........................
Page
3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
3
DR.
ROBERTS:
Good
morning.
I
want
to
take
the
opportunity
now
to
briefly
reintroduce
the
panel.
And
let
me
as
we
have
done
in
days
previously
begin
with
Dr.
LeBlanc
and
ask
each
member
of
the
panel
going
around
the
table
to
state
their
name,
their
affiliation
and
their
expertise.

Dr.
LeBlanc.

DR.
LEBLANC:
Good
morning.
My
name
is
Gerry
LeBlanc.

And
I'm
a
professor
in
the
department
of
environmental
and
molecular
toxicology
at
North
Carolina
State
University.
And
my
area
of
expertise
is
endocrine
toxicology.

DR.
KELLEY:
I'm
Darcy
Kelley.
I'm
a
professor
of
biological
sciences
on
the
faculty
for
the
center
for
environmental
research
and
conversation
and
a
member
of
the
Earth
Institute
at
Columbia
University.

And
my
area
of
expertise
is
sexual
differentiation
of
the
amphibian
xenopus
laevis.

DR.
KLOAS:
My
name
is
Werner
Kloas.
I'm
professor
for
endocrinology
at
University
of
Berlin.

I'm
also
heading
the
department
of
inland
fisheries
at
the
Leibniz
Institute
of
Freshwater
Ecology
and
Inland
Fisheries.
And
my
expertise
is
endocrine
disruptors
acting
on
sexual
differentiation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
4
and
thyroid
system
in
amphibians.

DR.
GREEN:
My
name
is
Sherril
Green.
I'm
an
associate
professor
in
the
department
of
comparative
medicine
at
Stanford
University.
My
special
interest
and
expertise
is
in
the
care
and
husbandry
of
laboratory
amphibians,
xenopus
laevis
and
other
species.

DR.
COATS:
My
name
is
Joel
Coats.
I'm
in
the
department
of
entomology,
chair
of
the
department,
and
professor
of
entomology
and
toxicology
at.

DR.
ISOM:
State
University.

My
expertise
is
environmental
toxicology,
environmental
chemistry
of
pesticides.

DR.
DENVER:
I'm
Robert
Denver.
I'm
from
the
department
of
molecular
cellular
developmental
biology
of
the
University
of
Michigan
at
Ann
Arbor.
And
my
expertise
is
developmental
neuroendocrinology
of
amphibians.

DR.
GIBBS:
My
name
is
James
Gibbs.
I'm
an
associate
professor
of
conversation
biology
at
the
State
University
of
New
York's
College
of
Environmental
Science
and
Forestry.
And
my
area
of
expertise
is
amphibian
demography.

DR.
RICHARDS:
My
name
is
Carl
Richards.
I'm
a
professor
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
5
biology
at
the
University
of
Minnesota
Duluth,
and
I'm
director
of
the
Minnesota
Sea
Grant
College
Program.
My
expertise
is
aquatic
ecologist
and
landscape
ecology.

DR.
DELORME:
My
name
is
Peter
Delorme.
I'm
a
senior
environmental
risk
assessor
with
the
Canadian
Government,
Pest
Management
Regulatory
Agency.
My
area
of
expertise
is
aquatic
ecology
and
risk
assessment
methods.

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
a
professor
of
environmental
toxicology.
I
also
serve
as
the
director
of
our
Center
for
Environmental
Health
Sciences.
My
area
of
expertise
are
molecular
toxicology.
And
I
also
study
some
neural
effect
of
the
pesticides.

DR.
THRALL:
I'm
Mary
Anna
Thrall.
I
am
a
professor
of
veterinary
pathology
at
Colorado
State
University.
And
my
area
of
expertise
is
veterinary
clinical
pathology.

DR.
ISOM:
I'm
Gary
Isom,
professor
of
toxicology
at
Purdue
University.
And
my
area
of
expertise
is
neurotoxicology.

DR.
HEERINGA:
I'm
Steve
Heeringa,
senior
research
scientist
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
6
with
the
Institute
for
Social
Research
at
the
University
of
Michigan.

I'm
a
statistician.
My
specialization
is
in
the
design
of
research
and
in
specialist
research
relating
to
human
and
animal
populations.

DR.
ROBERTS:
And
I'm
Steve
Roberts.
I'm
a
professor
at
University
of
Florida
with
joint
appointments
in
the
College
of
veterinary
medicine
in
the
College
of
Medicine,
and
serve
also
there
as
director
for
the
Center
for
Environmental
and
Human
Toxicology.

My
areas
of
expertise
are
toxicology
and
risk
assessment.

Before
we
proceed
with
the
public
comments
this
morning,
at
the
close
of
yesterday,
there
was
a
little
bit
of
confusion
about
the
status
of
some
research
reports
that
were
­­
from
research
performed
by
Dr.
Hayes
while
at
Ecorisk.

Dr.
Sielken
has
offered
to
perhaps
clarify
the
situation
on
that.

I
would
like
Dr.
Sielken
to
do
that
at
this
point
for
the
panel.

DR.
SIELKEN:
Thank
you,
Mr.
Chairman,
and
good
morning
to
the
panel.

Just
to
be
as
clear
as
possible,
I
did
print
out
a
copy
of
the
clarification
statement
that
I
would
like
to
make.
And
this
is
a
clarification
of
the
materials
that
I
submitted
yesterday
to
the
record
and
to
you
all.
That's
my
Texan
coming
through,
you
all.

On
the
morning
of
Wednesday,
yesterday,
Dr.
Hayes
made
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
7
reference
to
his
study
number
99XLATZ2.
At
the
beginning
of
the
afternoon
session,
I
submitted
four
items
to
the
EPA
or
SAP
related
to
my
review
of
the
statistics
in
Dr.
Hayes'
report
for
that
study.

These
four
items
were
as
follows:
And
they
are
numbered
here.

I
won't
read
the
titles.
But
the
first
one
was
a
document
that
was
prepared
for
a
meeting
with
EPA
in
June
of
2002.
And
that
document
was
briefly
summarizing
the
comments
that
I
had
made
in
my
presentation
to
Dr.
Hayes
on
September
19th
of
2000
at
Berkeley.

There
is
no
change
in
that.
I'm
just
clarifying
that
that
was
the
first
thing
that
I
gave
you
yesterday.

DR.
HAYES:
We
didn't
have
a
meeting
in
September
­­

DR.
SIELKEN:
That
is
the
date
that
was
on
the
front
page
of
the
presentation
that
I
gave.
It
was
at
the
meeting
in
Berkeley,

whatever
date
it
was.

DR.
HAYES:
It
was
January
or
February.
It
wasn't
September.

DR.
SIELKEN:
I
have
been
corrected.
It
was
January.
Boy.

This
is
what
happens
when
you
go
to
bed
at
quarter
to
4
in
the
morning.

It
was
at
the
one
and
only
meeting
that
I
attended
in
Berkeley
with
Dr.
Hayes.

The
second
item
was
the
overheads
that
I
presented
at
that
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
8
meeting,
whenever
it
was.
It
expressed
my
professional
concerns
as
a
statistician
about
characteristics
of
the
study,
including
what
Dr.

Hayes
described
as
"
haphazardly
selecting
a
sample
to
analyze
for
laryngeal
size"
instead
of
taking
a
random
sample,
the
need
to
explicitly
provide
more
of
the
numerical
data
and
errors
I
found
in
a
spot­
check
of
the
numerical
accuracy
of
the
report.

These
overheads
also
contain
the
results
of
what
I
believe
to
be
more
appropriate
statistical
analyses
of
the
data
on
laryngeal
cross­
sectional
area.
And
the
details
of
my
findings
are
written
out
there
for
you.

Basically,
what
I
found
was
that
there
was
not
a
substantial
reduction
in
laryngeal
size
due
to
atrazine.
This
was
the
same
conclusion
that
Dr.
Hayes
had
noted
in
the
body
of
his
report,
but
did
not
note
in
his
summary.

The
third
thing
that
I
handed
out,
which
was
probably
pretty
obscure
to
you,
was
the
copies
of
the
24
transparencies
that
I
prepared
to
show
examples
of
some
of
the
errors
and
inconsistencies
I
found
in
Dr.
Hayes'
Excel
spreadsheets,
worksheets,
and
which
I
had
presented
to
Dr.
Hayes
at
that
meeting
in
Berkeley.
Correct
the
date.

I
just
listed
there
some
idea
of
what
those
errors
and
inconsistencies
were.
I
had
copied
the
transparencies
that
I
used
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
9
identify
those
things
and
had
provided
those
to
you
yesterday.

Now,
where
the
confusion
really
probably
arises
is
with
respect
to
Number
4.
And
the
version
of
the
final
report
signed
by
Nigel
Noriega
on
June
23rd,
'
00,
was
the
copy
that
I
submitted
to
you.

The
first
three
of
the
items
that
I
briefly
discussed
above
were
reproduced
from
original
copies
that
I
had
prepared
and
brought
with
me
to
this
meeting.

I
did
not
think
that
I
needed
to
bring
my
copy
of
the
final
report
that
I
reviewed.
I
didn't
think
I
needed
to
bring
those
here
since
I
assumed
that
there
would
be
plenty
of
copies
here
and
I
didn't
need
to
carry
it
in
my
luggage.

Therefore,
I
asked
Ecorisk
staff
to
provide
a
copy
of
the
final
report
for
me
to
submit.
The
Ecorisk
staff
provided
the
only
final
report
that
was
signed.
And
I
assumed
that
that
was
the
same
report
that
I
had
been
looking
at.

Unfortunately,
the
only
version
of
the
­­
and
unfortunately,
the
only
version
that
the
final
report
that
was
on
the
CD
that
was
provided
to
the
SAP
members.

In
other
words,
Syngenta
provided
to
you
the
only
signed
final
report.

When
I
asked
the
Ecorisk
staff
to
reproduce
the
final
report
for
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
10
me,
then
that's
what
they
did.
They
reproduced
that
signed
report.

And
that's
the
same
report
that
is
on
your
CD.

That's
the
version
of
the
report
that
was
readily
available
yesterday
since
Dr.
Hayes'
statement
therein,
that
there
were
no
abnormal,
undifferentiated
or
intersexual
gonads
were
observed
in
any
of
the
treatments
or
controls
had
been
discussed
over
lunch.

So
that
was
why
that
copy
was
just
out
and
ready
and
ready
to
go.

During
Dr.
Hayes'
presentation,
he
had
shown
a
figure
showing
the
distributions
of
laryngeal
sizes
in
male
frogs.
I
understood
him
to
say
that
this
figure
referred
to
later
studies,
that
is,
later
than
this
99XLATZ2.

I
remember
that
figure
since
I
had
referred
to
it
extensively
in
my
presentation
at
Berkeley.

Since
I
had
given
all
the
copies
of
the
final
report
to
the
SAP
staff,
I
left
the
SAP
meeting
during
the
last
public
commentor's
comments
to
check
that
this
figure
was
indeed
in
the
copies
of
the
final
report
I
had
submitted
to
SAP.

To
my
chagrin
(
embarrassment),
I
realized
that
the
copy
of
the
final
report
I
had
submitted
could
not
have
been
the
copy
that
I
had
reviewed
for
that
Berkeley
meeting
since
the
figure
of
interest
was
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
11
not
in
it
and
I
knew
it
was
in
the
copy
that
I
had
seen.

I
understand
that
after
I
left
the
meeting,
Dr.
Hayes
stated
that
the
copy
of
the
final
report
I
had
submitted
was
not
the
final
version.

It
seems
that
the
signed
final
report
that
I
copied
and
gave
to
the
panel
yesterday
was
not
the
final
version.
In
fact,
there
is
no
final
version.
And
there
is
only
one
signed
and
dated
version.
And
that's
the
one
I
gave
you.

Furthermore,
all
other
versions
of
the
final
report
I
could
find
said
that
they
were
final
reports.
All
the
cover
pages
are
the
same.

And
none
were
dated
except
for
that
one
signed
copy
that
I
gave
you.

The
only
version
of
the
final
report
that
contained
the
figure
of
interest
was
an
unsigned
version
of
the
final
report
with
a
handwritten
note
on
the
cover
page
stating
that,
"
Cathy,
here
is
a
copy
of
the
report
that
was
sent
to
Ecorisk
several
months
ago.

And
it's
signed
Kathy
and
then
the
phone
number.

Here
the
first
Cathy
is
Cathy
Benz,
who
is
the
Ecorisk
quality
assurance
person.
And
the
second
Kathy
with
a
K
is
Katherine
Kim
who
is
CEO
of
Sokoke,
Incorporated,
and
also
Dr.
Hayes'
wife.

This
is
the
copy
that
one
Kathy
sent
to
the
other
Cathy.

I
checked
this
version
of
the
final
report.
And
it
did
contain
the
figure
as
I
recalled
it.
Furthermore,
this
version
contained
on
Page
41
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
12
the
erroneous
correlation
coefficient
of
minus
2.61
that
I
had
remembered
commenting
on,
since
a
minus
2
is
impossible.

Therefore,
I
went
to
Kinkos
last
night,
and
skipped
dinner,
and
had
26
copies
of
this
version
of
the
final
report
made
from
the
CD
that
had
been
provided
to
EPA.

So
there
was
a
copy
of
that
signed
version
in
the
miscellaneous
materials
that
Syngenta
had
submitted
to
EPA.

Although
this
version
contains
the
statement
that
there
are
no
abnormal,
undifferentiated
or
intersexual
gonads
were
observed
in
any
of
the
treatments
animals
and
the
figure
as
I
recalled
it
and
the
erroneous
correlation
coefficient
that
I
remembered,
a
closer
examination
of
the
figure,
about
11
o'clock
last
night,
revealed
that
the
figure
labels
were
slightly
different.

The
dots
and
the
figure
were
still
the
same.
But
the
axis
label
had
changed.
And
there
was
a
line
in
there
for
where
the
female
mean
was.

I
said,
well,
this
can't
be
the
report
that
I
reviewed
either.

So
of
the
reports
that
were
available
and
that
Syngenta
had
provided
to
EPA
and
they
had
provided
only
the
final
or
what
they
thought
to
be
the
final
to
the
panel
members,
it
meant
that
there
had
to
be
another
version
somewhere.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
13
I
recall
that
at
the
time
that
I
did
my
review,
Keith
Solomon,

Dr.
Solomon,
had
received
an
electronic
copy
of
the
version
of
the
final
report
and
of
the
Excel
spreadsheet
that
Dr.
Hayes
had
provided
and
that
I
had
reviewed.

Therefore,
this
morning,
instead
of
waking
at
my
office
in
the
middle
of
the
night,
I
asked
Dr.
Solomon
to
print
a
copy
of
that
version
of
the
final
report.

And
I
will
be
distributing
copies
of
that
to
the
SAP
members
as
well
as
to
Dr.
Hayes
as
soon
as
they
come
off
the
printer
downstairs.

This
version
of
the
report
also
contains
the
statement
about
no
abnormal,
undifferentiated
or
intersexual
gonads
were
observed
in
any
of
the
treatments
or
controls.

Let
me
make
a
personal
apology
for
my
unintentional
mistake
of
submitting
the
only
signed
version
of
the
final
report
­­
or
the
only
version
of
the
final
report
that
was
signed
by
the
study
director
to
SAP,
since,
apparently,
from
Dr.
Hayes'
statement,
it
is
not
the
final
report.

I
do
apologize
for
any
confusion.
I
have
made
in
that
box
26
copies
of
the
version
that
Kathy
sent
to
Cathy.
And
as
soon
as
they
come
off
the
printer
downstairs,
I
will
give
you
the
version
that
I
received
prior
to
the
Berkeley
meeting
and
that
I
reviewed
at
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
14
Berkeley
meeting.

Just
for
your
own
edification,
there
is
a
not
a
great
deal
of
difference
in
these
reports,
but
there
are
some
differences.
You
are
welcome
to
review
them
at
your
pleasure
and
make
your
own
decisions.

DR.
ROBERTS:
Thank
you,
Dr.
Sielken,
and
thank
you
for
your
efforts
to
clarify
that.

Dr.
Hayes,
I
see
you
raising
your
hand.

DR.
HAYES:
I'll
be
brief.

DR.
ROBERTS:
That's
fine.
Please
approach
the
microphone.

Let
me
just
explain
that
what
the
panel
is
trying
to
do
is
obtain
copies
of
reports
of
your
earlier
work
with
Ecorisk.
And
we're
trying
to
get
what
represents
the
best
­­
a
best
representation
of
that
research.
And
apparently,
there
are
multiple
versions
going
around.

And
if
you
could
assist
us
by
identifying
which
version
you
consider
to
be
the
most
accurate
representation
of
the
research,
that
would
be
very
helpful.

DR.
HAYES:
Yes.
In
terms
of
the
gonadal
abnormalities,
I've
already
addressed
that.
I
started
out
showing
how
we
discovered
the
abnormalities.
It
was
during
that
experiment
where
we
were
trying
to
measure
steroids
in
animals.
And
we
were
sexing
the
animals
without
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
15
buens
(
ph).
And
that's
when
it
was
discovered.
And
I
reported
that
to
Ecorisk,
not
to
Syngenta,
to
Ecorisk
in
November
of
2000.

And
that's
when
we
reanalyzed
the
data.
So
that
is
correct.
I
already
addressed
that.

The
other
thing
I
wanted
to
say,
I
think
Bob
Sielken
has
already
said
better
than
I
can,
you
see
the
sort
of
level
of
confusion
and
the
round
and
round
they
go
about,
there
is
never
a
final
report
because
there
is
always
we
need
to
do
this,
you
do
that.
And
things
never
get
signed.

That's
exactly
why
I
left
the
panel.
We
started
reporting
adverse
effects
as
early
as
1999.
And
because
there
was
never
a
final
report
and
things
never
got
signed
on
and
they
would
ask
to
change
this
and
that
and
this
version
and
that
­­
I
think
you
see
exactly
why
I
left
the
panel.

So
I
won't
comment
anymore.
It
was
exactly
what
was
just
described
to
you.

DR.
ROBERTS:
Okay.
But
we
will
now
have
several
versions
of
that
report.
And
if
to
the
extent
to
which
­­
before
we
close
public
comment,
if
you
could
look
at
those
and
identify
which
one
you
think
is
the
most
accurate
­­

DR.
HAYES:
I
can
identify
the
last
final
report.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
16
DR.
ROBERTS:
That
would
be
very
useful
for
us.

DR.
HAYES:
Also,
let
me
say,
by
the
way,
the
data
that
he,
Dr.

Sielken
is
talking
about
that
he
analyzed
was
not
sent
by
me.
I
don't
know
where
he
obtained
it.
I
did
send
data
to
Dr.
Solomon
and
Giesy,
but
that
didn't
come
from
me.

DR.
ROBERTS:
Dr.
Thrall
has
a
question.

DR.
THRALL:
Dr.
Hayes,
would
it
be
possible
for
us
to
have
a
copy
of
what
you
presented
to
us
yesterday?
Because
that's
probably
what
you're
considering
to
be
your
final
at
this
stage
of
the
game.
I
think
that
would
help
us
as
we
deliberate.

DR.
HAYES:
Yes.
I
don't
think
I
have
enough
recordable
CDs,

but
I
can
burn
it
on
to
a
CD
and
make
it
available.
Yes.

DR.
ROBERTS:
If
a
CD
can
be
made
available
to
Mr.
Paul
Lewis,
our
designated
federal
official,
he
can
make
copies
of
the
CD
­­
and
entered
into
the
public
docket.

DR.
HAYES:
Okay.
And
I'll
look
at
this
more
carefully,
but
it
does
look
like
the
final
version.
And
it
does
look
like
it
also
has
the
raw
data
and
all
the
numbers.
I'll
look
through
more
carefully,

including
the
numbers
that
Dr.
Sielken
just
said
weren't
available
to
him.

DR.
ROBERTS:
Dr.
Hayes
and
Dr.
Sielken,
both,
we
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
17
appreciate
your
efforts
to
clarify
this
for
the
panel.

Let
us
proceed,
then,
with
public
comments.
And
as
I
have
done
at
the
opening
of
each
of
the
public
comment
sessions
that
we
have
had,
let
me
remind
the
public
commentors
and
the
panel
that
we
are
here
focused
on
some
specific
issues
related
to
potential
developmental
effects
of
atrazine
on
amphibians.

So
we're
focused
on
scientific
issues
because
we
are
the
Scientific
Advisory
Panel.
We're
not
here
to
consider
policy
issues,

legal
issues
and
so
forth.

So
we
could
confine
comments
and
our
discussion
to
the
scientific
issues,
I
think
that
will
be
best.
Let
me
see.
Our
first
commentor
that
we
have
signed
up
for
this
morning
is
Dr.
Jennifer
Sass
on
behalf
of
Natural
Resource
Defense
Council.

And
she
will
be
followed
by
Dr.
Diana
Post,
just
as
a
heads
up,

Dr.
Post.
She
may
not
be
here.

DR.
SASS:
Should
I
start?
I
was
just
going
to
introduce
myself
while
this
is
­­

DR.
ROBERTS:
Good
morning,
welcome.

DR.
SASS:
Thank
you.
I'm
Jennifer
Sass.
I'm
a
Ph.
D.
with
the
Public
Health
Program
of
the
Natural
Resources
Defense
Council.
It
is
an
environmental
nonprofit
group.
We're
an
advocacy
group
for
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
18
public
health
here
in
the
U.
S.

And
we
have
been
following
atrazine
very
closely.
In
fact,
it
is
because
of
a
consent
decree
with
NRDC
that
EPA
is
reviewing
atrazine.

I
want
to
first
thank
you
all
for
giving
your
time
to
this
extremely
important
issue.
I
know
that
you
are
all
extremely
busy.

And
I
know
that
this
is
a
tremendous
amount
of
time,
both
in
attending
the
meeting
and
in
previewing
prior
to
coming
to
the
meeting.
And
I
want
you
to
know
that
we
really
appreciate
your
attention
to
this
important
matter.

I
also
want
to
thank
the
EPA
for
preparing
the
white
paper.
As
you
know,
most
of
the
studies,
in
fact,
all
of
the
Syngenta
submitted
studies
were
not
published
and,
therefore,
not
available
for
peer
review
or
public
scrutiny.
So
we
rely
on
the
EPA
scientists
to
review
boxes
and
boxes
of
data
that
we
would
never
even
want
to
look
at
were
it
publicly
available
because
of
the
time
and
the
energy
required.

So
I
do
really
appreciate
the
excellent
review
that
the
EPA
has
done
in
the
white
paper.
Thank
you.

I'm
going
to
make
my
comments
very
brief.
First
of
all,
I
want
to
point
out
that
the
charge
questions
to
the
EPA
in
the
white
paper
I
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
19
thought
were
unusually,
unprecedently
vague
and
ambiguous.

And
that's
possibly
because
the
charge
questions
were
in
an
unprecedented
interest
of
the
White
House
­­
actually
passed
through
the
White
House
for
review
prior
to
getting
put
into
the
white
paper.

And
I
don't
think
they
reflect
the
white
paper
in
terms
of
the
scientific
assessment.

So
I
would
like
to
start
out
by
redirecting
a
little
bit.
What
I
think
that
the
key
questions
are
that
the
EPA
really
needs
to
address
in
its
legal
obligations
with
the
NRDC
in
reviewing
atrazine
­­
in
fact,

the
reason
that
this
SAP
is
here
today
is
because
NRDC
negotiated
with
EPA
to
send
this
data
to
a
scientific
advisory
panel
for
review
and
since
at
the
time
the
EPA
was
not
going
to
look
at
the
data.

And
the
key
questions
in
that
agreement
were,
does
atrazine
threaten
wildlife
amphibian
populations.
And
does
atrazine
act
as
an
endocrine
disrupter.

Those
questions
are
not
in
the
charge
questions
anywhere.
But
those
are
the
key
questions
that
the
EPA
needs
to
answer
to
fulfill
its
legal
obligations
and
to
complete
its
assessments.

And
those
really
are
the
questions
that
this
scientific
advisory
panel
can
provide
input
and
advice
on
because
of
the
expertise
that
is
here
around
this
table.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
20
I
also
want
to
point
out
that
those
are
the
key
questions
that
are
data
driven.
In
other
words,
those
are
the
questions
that
are
going
to
be
answered
by
a
review
of
the
science
of
the
data
that
is
available
and
the
full
body
of
literature.
And
from
the
answers
to
those
questions,
the
next
question
will
be,
can
atrazine
be
used
safely.

At
that
point,
the
question
that
follows,
which
is
not
a
question
for
the
SAP,
is
at
what
concentration
or
dose
or
under
what
conditions
can
atrazine
be
considered
to
be
safe.
That's
presuming
that
it
can
be
used
safely.
Presuming
there
is
some
answer
to
that
middle
question.

At
what
concentrations
or
dose
can
it
be
used
safe.
In
other
words,
from
a
policy
point
of
view,
we
talk
about
a
baseline
or
a
no
effect
level
or
NOEL
or
reference
dose
and
for
human
consumption.

Those
are
policy
driven
questions.
Those
are
not
appropriate
questions
for
a
scientific
advisory
panel.
And
in
that
light,
I
don't
think
that
the
white
paper
should
be
criticizing
or
critiquing
or
undermining
the
published
literature
because
it
doesn't
provide
that
kind
of
an
answer.

Those
aren't
answers
that
can
be
derived
directly
from
the
kinds
of
studies
that
the
published
literature
were
designed
to
answer,
to
tackle
all
kinds
of
questions.

So
I
would
like
to
focus
on
the
data
driven
questions
and
point
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
21
out
that
it
is
not
a
fair
critique
to
say
that
a
study
doesn't
come
up
with
a
dose
response,
therefore,
the
study
somehow
is
insufficient
to
include
in
the
review.

The
NRDC
asks
that
the
scientific
advisory
panel
provide
a
fair
and
complete
review
of
the
available
literature
with
greatest
consideration
given
to
those
data
from
robust
and
well
designed
studies
published
in
the
peer
reviewed
and
scientific
literature.

We
ask
that
the
amphibian
data
be
evaluated
as
to
its
consistency
with
the
whole
body
of
available
data,
including
mammalian,
aquatic
and
mechanistic
studies.

We
ask
that
the
scientific
advisory
panel
provide
its
expert
scientific
opinion
as
to
the
effect
of
atrazine
on
amphibian
health
and
to
make
recommendations
as
to
whether
or
not
it
can
be
used
in
a
manner
that
will
not
harm
amphibian
populations.

We
believe
based
on
the
published
literature
and
reviews
therein
that
there
is
compelling
evidence
that
atrazine
is
a
multi­
site,

multi­
species
endocrine
disrupter.

Despite
variability
between
study
designs,
there
are
data
for
mutually
consistent
studies
with
sufficient
statistical
power
published
in
the
peer
reviewed
literature
demonstrating
that
atrazine
acts
in
mammals,
amphibians
and
aquatic
organisms
through
at
least
two
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
mechanisms
of
action
to
disrupt
hormonal
pathways
critical
to
reproductive
development
and
function.

These
findings
compel
a
determination
that
atrazine
is
an
endocrine
disrupter
on
wildlife
and
that
its
use
should
be
banned
or
severely
restricted.
That's
the
position
of
the
NRDC.

I've
handed
out
my
comments
yesterday
or
the
day
before,
how
ever
long
we
have
been
sitting
here,
on
paper
looking
something
like
this.
I
handed
out
25
copies.
I
hope
you
have
them.

It
provides
more
of
a
comprehensive
literature
review.
In
the
back
all
references
are
there
for
everything
that
I'm
talking
about
here.
What
I'm
going
to
do
is
very
quickly
breeze
through
that.
I'm
not
going
to
go
into
any
detail
on
any
study.
They
are
not
my
studies.

I
can't
defend
or
answer
to
them
in
any
detail.

I
just
want
to
put
the
question
of
amphibian
risk
in
the
larger
context
of
the
published
literature
on
atrazine.

What
I
want
to
point
out
in
this
graph,
which,
again,
you
have
in
the
hand
out
is
that
there
are
a
number
of
studies
in
rats,
and
there
are
a
number
of
different
studies
that
show
tumor
formation.
But
there
are
strained
specific
responses
or
rather
differences
between
how
strains
of
rats
respond
to
atrazine.

So
for
instance,
in
one
single
study,
a
full
litter
resorption
was
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
23
seen
at
low
doses
in
the
F344
strain,
but
not
in
the
Long
Evans
and
the
Sprague
Dawley
strain,
for
example.
And
that
was
seen
at
50
milligrams
per
kilogram
when
treated
through
gestational
day
6
to
10.

There
has
been
tumors
seen
in
some
strains
of
rats
and
not
in
other
strains
of
rats
and
at
different
doses.

And
those
tumors
have
been
reproductive
organ
associated
tumors
such
as
mammary
and
different
reproductive
organs.

There
has
been
a
suppression
of
luteinizing
hormone
and
prolactin
seen
in
Long
Evans
rats,
but
Sprague
Dawley
rats
did
not
respond
the
same
way.

Wistar
rats,
both
males
and
females,
showed
delayed
puberty.

Females
were
less
sensitive.
They
showed
effects
at
50
milligrams
per
kilogram
when
treated
during
the
period
of
critical
development
of
those
organs,
whereas
males
showed
effects
at
12.5.
Much
less.

I
want
to
point
out
that
I
think
that
the
published
studies
demonstrate
taken
together
that
atrazine
acts
as
an
endocrine
disrupter
in
rats,
but
the
different
strains
respond
differently.

Strains
differ
in
their
response
to
different
concentrations
and
also
with
different
measured
endpoints.
This
does
not
represent
disagreement
in
the
published
literature,
but,
rather,
demonstrates
the
complex
action
of
atrazine
like
all
endocrine
disruptors
on
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
24
complex
web
of
hormonal
regulation.
Don't
think
of
these
pathways
as
linear.

All
results
summarized
below
in
this
table
are
of
statistical
significance.
I
only
picked
what
the
authors
themselves
picked
as
their
conclusions
or
results.

I
think
that
in
addition
there
are
some
multiple
mechanisms
that
have
been
demonstrated.
I
don't
think
any
of
these
have
been
demonstrated
in
many
studies,
but
they
have
been
demonstrated
at
least
to
be
indicative
or
suggestive
of
wider
implications
in
the
luteinizing
hormone
and
prolactin
levels.

There
has
been
some
studies
in
whole
animals.
These
have
been
done
by
EPA
scientists.
And
in
fact,
it's
some
of
these
studies
that
the
EPA
is
now
using
to
set
a
no
effect
level
for
atrazine.

As
well
there
has
been
some
demonstration
of
aromatase
activity.
You
have
heard
about
that
ad
nauseam
in
this
meeting.

I
want
to
point
out,
though,
that
the
Sanderson
study
that
was
referred
to,
the
first
publication
actually
had
one
of
the
Ecorisk
people
on
that
as
authorship,
John
Giesy.
And
then
later
in
the
subsequent
follow­
up
publication,
his
name
was
removed
­­
or
he
did
not
participate
in
the
second
follow­
up
study.

But
he
seemed
to
be
in
agreement
with
those
findings,
at
least
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
25
the
first
time
around.
And
there
have
been
follow­
up
studies.
So
again,
I
don't
think
that
that
data
is
tight,
but
I
think
it
is
indicative.

Atrazine
disrupts
hormonal
pathways
in
multiple
species.
In
rats,
there
is
evidence
that
prostatitis
has
resulted
in
the
suckling
rat
pups
when
the
mothers
were
treated
with
atrazine.

These
are
interesting
studies
because
the
atrazine
does
not
seem
to
have
come
through
the
milk
to
the
pup.
It
actually
affected
the
mother,
the
dam,
and
then
the
pups
were
subsequently
affected
by
alterations
in
the
dam
hormonal
responses
to
atrazine.
And
this
was
by
EPA
scientists.

There
has
also
been
demonstration
of
reduced
testosterone,

reduced
sperm
motility
and
the
delayed
puberty
in
males
and
females
in
the
Wistar
rats.
Again,
this
was
in
the
EPA
study.

In
pigs,
there
is
one.
Delayed
estrus
after
oral
feeding
of
atrazine­
laced
feed.

And
in
alligators,
you
have
already
heard,
there
is
some
induced
aromatase
activity.

There
is
one
paper
I
found
in
tiger
salamanders
that
I
thought
was
interesting
because
it
showed
that
the
salamanders
were
responding
differently
at
different
concentrations,
that
at
lower
concentrations
development
was
delayed,
but
size
and
weight
weren't
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
26
affected,
whereas
at
the
higher
concentrations
development
progressed
normally,
but
size
and
weight
were
reduced.

Again,
I
don't
think
that
this
represents
disagreement.
I
think
it
accurately
represents
the
complex
reactions
of
biological
systems
to
hormonal
disruptors.

I
think
that
the
published
literature
is
consistent
in
demonstrating
that
atrazine
may
disrupt
hormonal
pathways
resulting
in
disruption
of
reproductive
hormones
and
reproductive
cycles.

While
the
Ecorisk
people
were
digging
up
data
last
night,
I
was
actually
on
the
phone
late
last
night
with
someone
named
Shana
Swan
(
ph)
who
has
just
published
yesterday
in
EHP
on
line
a
new
study
that
where
she
collected,
her
group
collected
semen
and
urine
samples
from
fertile
men,
about
200
each
from
Minnesota
and
Missouri,
and
shown
that
the
risk
of
poor
semen
quality
was
elevated
with
several
different
pesticides.
But
with
atrazine
it
was
elevated
12
fold.
And
those
are
very
statistically
significant.

The
populations
are
relatively
small,
but
then
again,
finding
a
needle
in
a
haystack.
If
this
is
able
to
be
seen
in
such
a
small
population,
I
think
it
is
worth
a
follow­
up.
And
so
does
she.
This
data
will
be
followed
up.
But
as
of
today,
this
is,
I
think,
consistent
with
the
body
of
literature
that
we're
reviewing.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
27
In
frogs,
there
has
been
demonstrations
in
multiple
species
both
in
the
wild
and
under
controlled
laboratory
conditions.

The
xenopus
laevis,
of
course,
is
what
the
lab
is
addressing
in
large
part.
Also,
in
rana
pipiens,
lab
and
field
work.
Also,
in
bufo
marinus,
in
field
work.

The
interesting
thing
about
the
bufo
marinus
work,
and
again,
it
is
not
publicly
available,
so
I
have
not
been
able
to
review
that,
any
published
studies
in
any
way,
but
my
understanding
from
discussions
with
the
author
is
that
there
is
a
built­
in
concentration
gradient
because
of
the
frogs
closer
to
the
cane
fields
had
more
effects.

In
other
words,
more
females,
skin
coloration,
than
the
frogs
living
farther
away
from
the
field.
But
obviously,
I'm
not
the
one
to
discuss
that
work.

I
think
that
the
ecological
risks
from
atrazine
are
unacceptably
high
and
that
there
is
no
relief
in
site
as
the
EPA
assessment
now
stands.

And
I
will
quote
from
EPA
environmental
scientists,
the
risk
quotients
exceeded
the
levels
of
concern
for
chronic
effects
on
mammals,
birds,
fish,
aquatic
invertebrates
and
non­
target
plants.

The
risks
are
possible
at
maximum
and
in
some
cases
typical
use
rates.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
28
So
the
EPA
ecological
assessors
are
extremely
aware
that
the
atrazine
at
current
use
rates
under
current
use
patterns
is
posing
an
unacceptably
high
threat
to
wildlife
populations.
And
some
of
these
wild
populations
include
endangered
species.

There
are
concerns,
again,
I'm
quoting
from
the
EPA
scientists,

for
adverse
toxicological
effects
on
freshwater
and
estuarine
plants
and
their
communities
as
well
as
indirect
adverse
effects
on
aquatic
invertebrates
and
fish
populations
at
monitored
atrazine
levels
in
surface
waters.

So
based
on
real
readings
of
real
atrazine
in
the
real
world,

there
is
real
cause
for
concern
according
to
the
EPA
ecological
scientists.

The
ongoing
use
of
atrazine
jeopardizes
endangered
species
and
their
critical
habitats.
The
exposure
of
aquatic
communities
to
atrazine
levels
at
10
to
20
part
per
billion,
this
is
based
on
an
EPA
assessment,
can
result
in
community
level
and
population
level
effects.

This
is
significantly
below
the
EPA's
current
proposal
in
its
current
assessment,
which
is
to
allow
a
seasonal
average
or
90
day
average
up
to
37.5
part
per
billion
before
any
action
is
triggered.

Up
until
37.5
part
per
billion,
the
registrant,
Syngenta,
will
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
29
voluntarily
do
more
intensive
monitoring.

So
other
than
monitoring,
there
will
be
no
change
in
the
use
or
use
rate
or
use
patterns
of
atrazine
unless
the
37.5
part
per
billion
trigger
level
is
exceeded
for
a
90
day
average.

The
U.
S.
Fish
and
Wildlife
submitted
comments
on
this.
They
claim
that
EPA's
assessment
underestimates
the
ecological
impacts
of
atrazine
in
part
because
it
does
not
consider
sublethal
effects
on
reproductive
ability.

So
to
end
and
to
thank
you
for
your
time,
I
want
to
point
out
that
there
are
some
very
serious
and
more
scientific
questions
that
the
EPA
needs
to
address.
And
those
questions
are,
does
atrazine
threaten
wild
amphibian
populations
and
does
atrazine
act
as
an
endocrine
disrupter
in
wildlife
populations.

These
are
data
driven
questions.
And
if
the
answer
is
yes,
then
the
onus
falls
on
the
EPA
to
answer
the
policy
questions
of
what
doses
and
under
what
conditions
atrazine
can
or
cannot
be
used
safely.

Thank
you
for
your
time.

DR.
ROBERTS:
Thank
you,
Dr.
Sass.

Let
me
ask
the
panel
members
if
they
have
any
questions
for
you.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
30
Dr.
Matsumura.

DR.
MATSUMURA:
Regarding
this
new
report
on
the
human
sperms,
did
those
people
really
measure
the
level
of
the
atrazine
in
sperms?

DR.
SASS:
In
the
fluid,
is
my
understanding.

I
spoke
with
the
author
last
night
and
have
read
some
of
the
various
different
reports
of
it.
And
my
understanding
is
they
measured
in
the
fluid
levels.

Atrazine
was
one
of
a
small
handful
of
chemicals
that
came
to
light
as
associated
with
this.
Alachlor
and
diazonal
metabolites
were
the
others.

DR.
MATSUMURA:
Thank
you.

DR.
ROBERTS:
Any
others?
If
not,
thank
you
very
much,
Dr.

Sass.

It
sounds
like
lots
of
folks
were
busy
last
night
working.

Dr.
Post
has
requested
time
to
speak.
I
don't
know
whether
Dr.

Post
is
here
this
morning
yet.
If
not,
we
can
move
to
the
next
person
and
then
come
back
to
her.

I
have
Dr.
Stephen
Sheffield
listed
as
requesting
the
opportunity
for
public
comment.
Is
Dr.
Sheffield
here?

Come
forth,
please,
and
identify
yourself.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
31
DR.
SHEFFIELD:
Thank
you,
Mr.
Chair.
My
name
is
Steve
Sheffield.
And
first
of
all,
I
greatly
appreciate
the
opportunity
to
address
the
SAP
today.

In
my
mind,
this
is
a
very
important
issue.
I
would
echo
the
comments
of
the
previous
speaker
in
acknowledging
the
amount
of
time
and
your
willingness
to
serve
on
this
panel.
It
is
a
very
important
issue.
And
I
greatly
appreciate
all
the
efforts
you
are
going
to
put
into
this
over
the
next
couple
of
days
and
the
efforts
you
already
put
into
it.

I
should
also
indicate
that
I
am
an
affiliate
professor
in
the
department
of
environmental
science
and
policy
at
George
Mason
University,
and
that
I'm
actually
providing
these
comments
to
the
SAP
as
a
professional
wildlife
toxicologist
and
private
citizen,
and
that
these
comments
shouldn't
be
construed
as
an
official
position
of
the
university.

I
should
probably
spend
just
a
second
giving
you
information
on
my
pertinent
background
and
experience.
I
have
a
Ph.
D.
in
environmental
toxicology
from
Oklahoma
State
University.
I
was
on
the
faculty
of
environmental
toxicology
at
Clemson
University
for
four
years.
And
I
have
been
at
George
Mason
for
four
years.

I
have
experience
with
both
laboratory
and
field
experiments
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
32
examining
exposure
and
possible
effects
of
pesticides
on
amphibians,

including
both
frogs,
rana
species,
hyla
species
and
xenopus
as
well
as
salamanders,
ambystoma
species,
and
have
coauthored
a
book
chapter
on
multiple
chemical
stressors
and
amphibians
that
will
be
published
by
SETAC
press
this
summer.

I
have
been
following
with
great
interest
the
atrazine
amphibian
issue
as
it
has
unfolded
over
the
last
several
years.
As
a
result,
I
was
wanting
to
share
with
you
my
perspective
on
the
subject.
I
will
include
some
thoughts
on
atrazine
in
general,
some
comments
on
the
studies
highlighted
in
the
white
paper,
some
comments
based
on
the
peer
reviewed
literature
and
some
recommendations
and
thoughts
on
the
subject.

As
a
wildlife
toxicologist,
one
of
the
things
that
catches
my
attention
is
the
high
volume
pesticides.
That's
not
to
say
that
all
of
them
are
bad.
I'm
just
saying
that
when
I
go
about
looking
at
things,

looking
at
exposure
and
effects
of
pesticides
on
wildlife
species,
it
the
high
volume
ones
that
generally
catch
my
attention.

In
the
United
States,
from
the
estimate
of
the
literature
and
other
sources
I
found,
are
roughly
75
to
150
million
pounds
of
active
ingredient
annually
over
40
years
on
up
to
100
million
acres
of
atrazine
is
applied
or
has
been
applied.
It
is
ubiquitous
in
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
33
environment.
It's
found
in
all
environmental
media,
including
surface
water,
ground
water,
soil,
sediment,
air,
including
fog,
and
biota
and
it's
atmospherically
transported.

I
think
that
this
is
an
incredible
volume
of
atrazine
use
and
it's
very
difficult
to
comprehend
that
sheer
volume
of
amount
of
use
and
the
amount
in
the
environment
that
I
kind
of
liken
it
to
the
idea
of
trying
to
comprehend
how
much
100
million
dollars
is.
I
know
I
certainly
can't
do
that.
I
have
a
hard
time
comprehending
that
number.
That's
why
I'm
saying
that
this
is
a
very
hard
number
to
comprehend
because
it
is
so
high.

Therefore,
as
Dr.
Hayes
made
mention
yesterday,
control
sites
that
are
free
of
atrazine
contamination
are
most
difficult
to
find.

Further,
peak
application
coincides
closely
with
peak
amphibian
reproductive
seasons.
I
see
this
as
a
scenario
for
potential
trouble
for
amphibians,
not
just
from
atrazine,
but
from
the
various
chemical
mixture
in
which
they
are
exposed.

Dr.
Hayes'
presentation
yesterday,
he
had
several
different
slides
showing
atrazine
levels
found
in
the
environment.
I
don't
need
to
repeat
that.
The
levels
that
he
has
looked
at
and
that
some
others
have
looked
at
are
environmentally
relevant,
which
is
important.

I
like
to
tend
to
discount
some
of
the
ones
that
look
at
really
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
34
high
unrealistic
levels
and
look
at
the
ones
that
are
just
­­
focus
on
the
ones
that
are
just
environmentally
relevant.

One
of
the
things
that
I
like
to
focus
on
a
lot
is
exposure
routes.

And
as
far
as
exposure
routes,
I
see
amphibians
getting
exposed
through
oral,
dermal
and
inhalation
routes
for
atrazine.
I
don't
know
if
there
is
any
data
on
maternal
deposition
into
eggs
or
not.
But
I
imagine
that
that
is
possible.
It
does
happen
for
other
compounds.

It
also
needs
to
be
considered
the
persistence
of
atrazine
in
the
environment.
The
estimates
range
very
widely.
But
it's
generally
relatively
persistent.
In
aquatic
systems,
it
can
vary
quite
a
bit.

In
some
of
the
experimental
microcosms
and
mesocosms
it
does
not
have
a
very
long
persistence.
But
in
some
natural
farm
ponds
and
natural
lakes
it
can
persist
almost
a
year.
That's
significant.

In
terrestrial
systems,
one
citation
I
found
recently,
the
Talbert
and
Fletchall
paper,
1964,
atrazine
persisted
for
17
months
in
the
soil
at
a
two­
pound
active
ingredient
per
acre
application
rate.

Now
I'm
going
to
move
to
a
couple
specific
areas
that
I'm
concerned
with.
One
is
bioaccumulation.
At
least
one
person
so
far
in
this
panel
­­
in
this
proceedings
has
mentioned
bioaccumulation.

The
study
by
Allran
and
Karasov
of
2000
found
levels
of
atrazine
in
tadpoles
that
were
six
times
higher
than
the
concentration
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
35
in
the
test
water.

That
may
or
may
not
be
a
significant
finding,
but
it
raises
a
flag
with
me.
It
could
possibly
result
in
more
continuous
exposure
of
the
target
organs.

And
amphibians
are
the
only
vertebrate
taxa
that
I'm
aware
of
that
also
accumulate
organophosphate
insecticides.

Another
finding
that
could
have
a
bearing
on
amphibians,
given
the
suggestion
that
atrazine
effects
may
potentiate
or
even
synergize
organophosphate
effects.

As
far
as
possible
fitness
effects
of
atrazine
on
amphibians,

there
is
a
couple
different
studies
that
have
indicated
possible
fitness
effects.
And
I'm
very
concerned
about
that
as
well.

The
Allran
and
Karasov
2001
paper
used
exposure
rates
of
0
to
20
micrograms
per
letter.
And
frogs
at
the
highest
level
stopped
eating
immediately
upon
introduction
of
atrazine
and
did
not
eat
during
the
entire
14
day
experiment,
an
anorexic
like
effect
that
could
have
implications
for
fitness
in
amphibians.

The
Brown­
Sullivan
and
Spence
2003
paper,
which
was
a
study
that
looked
at
atrazine
and
nitrate
in
combination,
this
particular
study
found
that
at
40
micrograms
per
letter
of
atrazine
and
37
milligrams
per
litter
of
nitrate
using
an
additive
model,
the
snout
vent
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
36
length
at
metamorphosis
significantly
decreased.
They
saw
reduced
growth
and
delayed
metamorphosis.

This
possible
selective
disadvantage
for
these
frogs
could
result
in
fitness
effects.

And
finally,
the
Diana,
et
al.,
2000
paper
that
examined
effects
of
atrazine
on
amphibian
growth
and
survival
in
artificial
aquatic
communities,
they
looked
at
hyla
versicolor,
which
are
the
gray
tree
frogs,
and
atrazine
concentrations
of
0,
20,
200
and
2000
micrograms
per
liter.

They
found
frogs
from
two
high
dose
groups
were
five
percent
shorter
and
10
percent
lower
body
mass
at
metamorphosis
than
those
of
the
control
and
low
atrazine
groups.

Decrease
in
amphibian
length
and
weight
at
metamorphosis
can
indicate
a
reduction
in
fitness
in
wild
populations
of
anurans
exposed
to
atrazine
at
these
levels,
although,
I
do
admit
that
these
levels,
the
200
to
2000
micrograms
per
liter
are
not
common,
usually,
except
during
the
peak
times
when
the
runoff
accumulates.

Now,
moving
to
the
studies
that
have
been
performed
and
published
by
Hayes,
et
al.,
first,
I
have
to
say
that
I
was
greatly
impressed
with
the
presentation
given
by
Dr.
Hayes
yesterday.
He
was
very
thorough
in
scope
and
covered
the
contentious
areas
to
my
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
37
satisfaction.

Overall,
I
would
characterize
these
studies
as
very
carefully
conducted
and
analyzed,
statistical
sound
and
believe
that
the
conclusions
reached
were
reasonable
given
the
data.

Do
I
find
fault
with
some
of
his
methodology?
Yes.
But
not
enough
to
discount
the
findings
of
the
study.

Regarding
the
methodology,
there
is
a
few
things
that
I
would
do
differently.
As
an
example,
I
would
use
glass
containers
instead
of
plastic.
That's
what
I
have
used
in
any
previous
studies.

I
would
also
use
an
atrazine
formulation
instead
of
technical.

That's
an
personal
choice
of
mine.
Just
as
examples.

I
don't
understand
the
criticisms
of
their
work
because
they
suggested
a
mode
of
action
for
the
effects
seen
or
for
failing
to
find
an
acceptable
dose
response
curve.

Although
they
speculate
on
a
mode
of
action
of
a
particular
fact,
that's
a
normal
part
of
a
discussion
section
of
any
peer
reviewed
scientific
publication.

The
high
incidence
of
males
with
gonadal
abnormalities,

whether
it
be
testicular
oocytes,
intersex
or
hermaphroditic,
how
ever
you
want
to
say
it,
is
of
great
concern
to
me.
Particularly,
if
it's
shown
that
these
individuals
are
not
fertile
or
otherwise
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
38
reproductively
impaired,
which
has
not
yet
been
looked
at.

From
what
I
heard
from
the
group
on
Tuesday,
the
intersex
individuals
do
not
get
that
way
by
themselves.
It
apparently
takes
exposure
to
an
endocrine
disrupting
compound
during
gonadal
development
to
make
this
happen.

In
regard
to
the
registrant
studies,
there
was
a
relatively
large
number
of
studies
that
were
funded
by
the
registrant
over
the
past
year
or
two.
Just
as
a
personal
comment
or
a
personal
observation,
it
was
apparent
to
me
that
the
studies
were
highly
reactionary
in
nature.

And
reactionary
not
only
to
the
results
of
Dr.
Hayes,
but
also
for
inclusion
in
the
atrazine
IRED
document.

Therefore,
I
see
that
they
are
put
together
and
conducted
rapidly.
They
contain
many
design
and
other
flaws
that
limited
their
use
and
rendered
them
largely
uninformative.

Further,
only
one
of
these
studies
is
peer
reviewed,
the
Carr,
et
al.,
paper.
And
I
really
hope
that
I
get
the
ultra
rapid
turnaround
this
paper
was
afforded
when
I
submit
my
next
paper
to
Environmental
Toxicology
and
Chemistry.

As
far
as
field
studies,
Dr.
Hayes
beat
me
to
the
punch
on
this
one,
but
I'm
going
to
expand
on
his
comments
from
yesterday.

The
study
by
Reeder,
et
al.,
1998
appears
to
have
been
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
39
completely
written
off
as
a
study
with
nonsignificant
data.
However,

a
point
that
seems
to
be
overlooked
on
their
work
is
the
fact
that
although
they
found
significant
correlation
between
PCB
and
PCDFs
and
sex
ratio
reversal,
they
state
that
there
was
an
association
approaching
significance,
the
P
value
of
0.07
between
the
detection
of
atrazine
and
intersex
individuals.

The
failure
of
this
study
to
find
a
significant
correlation
for
this
was
pointed
out
in
subsequent
studies
on
this
topic,
including
in
the
Carr,
et
al.,
2003
paper.
However,
I
have
to
point
out
that
I'm
very
comfortable
in
assigning
a
significance
level
of
P
equal
of
0.1
for
field
studies
such
as
this
due
to
the
large
inherent
variability
and
my
willingness
to
accept
a
slightly
higher
error
in
these
cases.

In
fact,
this
is
a
common
statistical
practice
for
field
studies,

including
the
set
of
field
studies
conducted
at
the
EPA
Corvallis
mesocosm
facility
over
the
past
10
years,
of
which
I
have
peer
reviewed
many
of
them
for
journals.

Therefore,
I
would
ask
the
panel
to
consider
the
use
of
P
equal
0.1
significance
level
in
field
studies
when
appropriate.

I'm
not
saying
it's
always
appropriate.
I'm
saying
that
it's
very
commonly
used
in
field
studies
because
the
variability
is
so
high
that
you
are
willing
to
accept
a
little
bit
more
error
to
try
to
tease
out
what
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
40
is
going
on.
That's
what
I'm
saying.

As
far
as
endocrine
disrupting
abilities
of
atrazine,
I
strongly
agree
with
Dr.
Matsumura's
assertion
the
other
day
that
we
should
be
looking
at
the
hypothalamic
pituitary
interrenal
axis
in
amphibians,
a
system
that
is
involved
in
the
stress
response,
metamorphosis,
feeding
and
mating.

This
is
where
the
endocrine
disruption
action
is
in
mammals,

and
I
have
a
hard
time
figuring
out
why
we
are
not
looking
at
this
in
amphibians.

As
we
heard
on
Tuesday
from
the
registrants,
these
hormonal
systems
are
evolutionary
highly
conserved
through
the
vertebrate
taxa.
So
we
can
reasonably
expect
that
the
amphibian
HPI
axis
system
would
closely
resemble
that
of
birds
and
mammals.

Therefore,
in
addition
to
the
examination
of
the
hypothesized
mode
of
action
involving
aromatase,
we
should
be
looking
at
possible
impacts
on
ACTH,
gonadotropin
releasing
hormone,
FSH,
LH
and
prolactin
as
well.

The
HPI
axis
has
been
looked
at
in
amphibians
by
Gendron,
et
al.,
1997,
who
examined
the
functional
integrity
of
the
HPI
axis
in
mudpuppies
exposed
to
organochlorines
in
the
field.
They
found
contaminant­
induced
disruptions
within
the
HPI
axis
in
mudpuppies
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
41
collected
at
most
of
their
contaminated
study
sites.

There
has
been
one
mention
so
far
in
proceedings
about
immunotoxicology.
I'm
very
interested
in
that
area.
The
Christin,
et
al,
paper,
2003,
that
was
published
in
the
most
recent
issue
of
ET&
C,

Environmental
Toxicology
and
Chemistry,
exposed
juvenile
rana
pipiens
for
21
days
to
a
mixture
of
six
pesticides,
including
atrazine,

and
then
challenged
with
a
parasitic
nematode.

They
found
that
pesticide
mixtures
caused
diminished
phagocytosis
and
splenocyte
numbers,
thereby
suggesting
a
compromised
immune
system
in
these
frogs.

I
have
three
other
considerations
that
I
want
to
mention.
One
is
an
interaction
consideration.
Both
potentiation
and
synergy.
I
alluded
to
this
study
earlier,
the
Belden
and
Lydy
study,
2000.

Exposure
to
nontoxic
concentrations
of
atrazine
cause
potentiated
toxicity
of
OPs,
of
organophosphate
insecticides
in
amphibians.

In
a
previous
study
by
Lydy
in
1997
showed
the
toxicity
of
atrazine
and
organophosphates
was
synergistic
to
invertebrates.
The
ubiquity
of
atrazine
in
the
environment
makes
me
wonder
how
much
additional
damage
of
the
OPs
are
causing
amphibians
due
to
an
interaction
such
as
this.

And
another
paper
by
Howe,
et
al.,
1998,
there
was
an
atrazine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
42
alachlor
mixture
that
was
more
toxic
than
either
of
the
two
compounds
separately.

As
far
as
formulation,
this
is
a
personal
bias
of
mine,
but
I
would
submit
that
there
is
no
such
thing
as
technical
grade
atrazine
in
the
real
word.

The
real
world,
which
is
the
crops
and
other
plants
in
which
atrazine
is
applied,
receives
various
formulations
of
atrazine.
I
believe
that
it
should
be
formulations
of
atrazine
that
we
should
be
examining
and
not
technical
grade.

I'm
fully
aware
of
the
arguments
for
using
technical
grade
pesticides
in
experiments,
but
in
my
mind,
that
has
little
practical
use
if
it's
not
the
exact
chemical
in
which
the
organisms
are
being
exposed
in
the
field.

And
my
other
area
of
consideration
is
atrazine
metabolites
and
degradates.
These
compounds
such
as
hydroxyatrazine
and
de­
ethylatrazine
and
others
should
not
be
ignored
in
any
amphibian
studies.
These
compounds
are
also
ubiquitous
in
the
environment
and
may
be
exerting
some
effect.

So
the
way
I
see
it,
the
major
studies
that
need
to
be
done,
are
intersex
frogs
fertile?
To
me,
that's
the
Number
1
thing
from
the
two
days
and
part
of
a
third
day
that
I
have
seen
so
far.
That's
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
43
question
I
think
needs
to
be
addressed
immediately.
Are
these
intersexed
frogs
fertile.

If
yes,
is
there
a
problem.
If
no,
you
have
identified
a
big
problem.
And
then
where
do
we
go
from
there.

Number
2,
as
I
alluded
to
already,
the
HPI
axis
in
amphibians
is
something
that
is
in
bad
need
of
being
examined.

Number
3,
I
strongly
suggest
including
a
salamander
species
in
the
testing
that
will
be
done.

A
study
of
mine
that
will
be
published
later
this
year
found
that
salamanders
are
on
the
order
of
two
to
four
times
more
sensitive
to
an
organophosphate
insecticide
than
three
species
of
anuran
frogs
including
xenopus.

In
this
study,
I
used
the
marbled
salamander,
ambystoma
opacum,
and
found
it
to
be
an
ideal
test
species
for
studying
possible
effects
of
contaminants.
It
is
an
autumn
breeder
with
relatively
large
clutch
sizes,
larvae
that
are
relatively
easy
to
raise
in
the
lab
and
eggs
whose
development
can
be
started
on
demand.

Number
4,
the
use
of
xenopus
as
a
primary
test
species.
I
would
point
out
to
the
panel
that
there
are
inherent
strengths
and
weaknesses
to
going
this
route.

The
strengths
are
well­
known.
But
in
addition
to
being
a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
44
non­
native
species
and
only
related
to
our
native
species
at
the
order
level,
there
is
an
important
life
history
difference
between
xenopus
and
native
anurans.
The
adult
xenopus
are
100
percent
aquatic,

whereas
native
anurans
can
spend
significant
time
on
land.
Thus,

differential
exposures
to
atrazine
can
occur.

Native
species
leave
the
water
and
feed
on
terrestrial
pray
items,
move
through
sprayed
vegetation
and
can
pick
up
both
dermal
and
inhalation
exposure
that
could
be
more
significant
or
at
least
different
than
in
water.

For
example,
a
frog
moving
through
a
freshly
sprayed
field
could
be
exposed
to
atrazine
through
oral,
dermal
and
inhalation
routes
where
the
atrazine
would
not
be
diluted
as
it
would
be
in
water.

This
is
the
major
point.
These
two
differences
can
make
extrapolation
between
xenopus
and
native
anurans
less
reliable.

Number
5
is
the
tolerance
resistance
question.
I
noticed
with
interest
that
this
has
been
touched
on
already
in
these
deliberations.

It
is
important
to
address
this
issue
because
this
might
be
contributing
what
is
apparently
being
seen
in
the
field,
the
populations
of
frogs
in
heavy
atrazine
use
areas
are
still
there.

In
areas
where
heavy
atrazine
use
has
gone
on
for
40
years,

such
as
some
of
the
areas
we
have
heard
about
already,
that
is
more
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
45
than
enough
time
to
have
had
enough
generations
in
which
to
evolve
genetic
resistance.

Other
vertebrates
have
done
it.
Mainly,
muroid
rodents
and
gambusia,
mosquito
fish.
But
I
submit
that
the
tolerance
resistance
question
is
very
important
to
this.
Particularly,
in
looking
at
why
all
these
frogs
are
still
there
after
40
years.

And
Number
6,
population
level
effects.
Certainly,
this
could
be
argued
to
be
the
ultimate
question
to
examine.
However,
I
firmly
believe
it
may
take
many
years
to
possibly
see
an
effect
at
this
level.

If
some
males
are
losing
reproductive
function,
other
males
may
hop
right
in
and
be
more
than
happy
to
increase
their
contribution
to
the
gene
pool.
If
fertility
is
negatively
impacted
in
intersex
individuals,
intersex
males
die
and
are
lost
to
the
gene
pool
without
contributing.

As
the
frequency
of
intersex
individuals
increases
in
the
population,
the
effective
population
size
should
decrease
as
fewer
individuals
are
contributing
to
the
subsequent
generations.

Also,
as
frogs
die
out
of
a
particular
area,
other
frogs
may
come
in
and
colonize,
recolonize
the
area,
making
it
likely
very
difficult
to
detect
a
population
crash.

So
if
this
scenario
could
be
maintained
for
a
relatively
long
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
46
period
of
time,
a
population
level
effect
may
take
many,
many
years
to
be
detected.

My
final
thoughts.
I
have
a
couple
final
thoughts
here
and
I'll
end.
The
static
renewal
verse
continuous
flow­
through
tests.
I
have
modified
this
­­
I've
added
this
since
the
deliberations
have
started.

I
strongly
agree
with
Dr.
Hayes
and
others
on
this.
I
believe
for
these
tests
it
is
imperative
to
mimic
as
closely
as
possible
the
natural
conditions
of
the
test
organisms.

Native
anurans
and
xenopus
tend
to
be
found
in
pools,
ponds,

vernal
pools
and
ephemeral
standing
water
such
as
puddles
and
ditches,
not
flowing
waters
like
streams
and
rivers.
And
if
they
are
found
in
streams,
which
I
have
found
some
in
streams,
they
are
likely
found
mainly
in
the
intermittent
pools
within
streams
and
not
in
the
moving
water.

Also,
cost
is
a
major
factor
in
this.
As
being
a
researcher
at
a
university,
static
renewal
is
relatively
inexpensive
and
continuous
flow­
through
is
so
expensive
as
to
be
prohibitive
to
most
researchers.

I
do
agree
that
the
flow­
through
systems
successfully
deal
with
the
water
quality
issue,
but
it's
probably
not
worth
the
trade­
off.

Second,
I'm
not
going
to
spend
any
time
talking
about
this
other
than
just
mentioning
it
because
it
is
getting
away
from
things
a
bit.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
47
But
it
is
proposing
that
the
EPA
consider
some
sort
of
action
the
auspices
of
the
Clean
Water
Act
such
as
the
National
Water
Quality
Criteria
for
Amphibians.

Something
like
this
was
proposed
previously
in
the
1990s,
but
no
water
quality
criteria
exists
for
amphibians
in
the
United
States.

And
it's
automatically
assumed
that
criteria
for
fish
and
human
health
are
adequate
to
protect
all
aquatic
species,
when,
in
fact,
this
assumption
has
yet
to
be
tested.

Finally,
regarding
the
white
paper
and
the
charge
to
the
panel.

I
was
impressed
with
all
the
work
done
by
Dr.
Tom
Steeger
and
his
colleagues
for
this
white
paper.
And
I
commend
Tom
and
his
colleagues
for
their
effort.

However,
when
I
got
to
the
eight
questions
at
the
back
of
the
white
paper,
that
was
a
different
story.
To
say
that
I
was
sorely
disappointed
with
them
is
an
understatement.

In
my
opinion,
they
were
poorly
written
and
off
base
from
where
I
think
this
SAP
should
be
headed.
I
say
this
in
part
because
they
failed
to
address
one
of
the
most
fundamental
questions
one
could
ask
regarding
atrazine
and
amphibians.
Is
it
an
endocrine
disrupter
in
amphibians.

Why
is
there
no
question
that
states,
does
the
panel
think,
given
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
48
the
data
available
from
the
17
studies,
atrazine
is
causing
some
level
of
endocrine
disruption
in
amphibian
populations.
And
does
this
effect
on
the
endocrine
system
have
the
potential
to
translate
into
population
level
effects.

That
concludes
my
comments.
Thank
you
very
much.
And
I
wish
you
the
best
of
luck
in
the
next
two
days
in
your
deliberations.

DR.
ROBERTS:
Thank
you,
Dr.
Sheffield.

Are
there
questions
from
the
panel?
Dr.
Green,
then
Dr.
Thrall.

DR.
GREEN:
I
have
two
short
questions.
You
refer
to
the
paper
by
Belden
and
Lydy
of
2000,
exposure
to
nontoxic
concentrations
of
atrazine
cause
potentiated
toxicity
of
OPs
in
amphibians.

Do
you
recall
just
ballpark
what
these
nontoxic
concentrations
of
atrazine
were?

DR.
SHEFFIELD:
I
don't,
but
I
believe
I
have
­­
no,
I
don't
have
the
paper
with
me.
I
don't
know
what
the
compounds
were.
But
I
know
that
the
two
OPs
tested
were
chlorpyrifos
and
diazinon.
Those
effects
were
potentiated
with
the
atrazine
exposure.

DR.
GREEN:
The
second
question
I
have
is
could
you
clarify
the
difference
between
technical
grade
atrazine
and
other
formulations
of
atrazine?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
49
DR.
SHEFFIELD:
Yes,
absolutely.

As
I
started
into
this
field,
it
just
made
sense
to
me
that
if
you
are
going
to
study
something
in
the
lab,
why
not
study
something
that
has
application.

As
a
toxicologist,
the
tendency
is
to
use
something
that
is
pure
or
as
close
to
pure
as
you
can
get
so
that
you
can
actually
say
this
is
causing
the
effect.
You
don't
want
these
other
things
in
there
mixing
up
with
it.

But
in
pesticides,
the
stuff
that
is
used
in
the
field
are
called
formulations.
So
they
take
the
technical
grade
and
they
make
formulations
out
of
that
so
that
you
get
a
various
percentage
of
the
active
ingredient
in
the
formulation.

For
example,
diazinon,
a
compound
that
I
have
used
quite
a
bit.

50
percent,
an
emulsifiable
concentrate
of
50
percent
is
used.
And
the
other
50
percent
is
inert
ingredients.
So
they
will
put
other
compounds
in
there
to
dissolve
the
diazinon
and
other
things
needed
to
make
that
formulation
go
on
that
particular
crop
or
whatever
application
is
being
used.

So
the
technical
grade
is
the
pure
or
as
pure
as
it
gets.

Sometimes
technical
grade
is
only
95
percent,
sometimes
it's
99
percent.
It
usually
ranges
in
that
area.
Formulations
are
what
is
used
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
50
in
the
real
world.
And
For
atrazine,
there
is
many
different
kinds
of
formulations.

I
don't
know
exactly
what
the
active
ingredient
percent
is
in
these
formulations.
Offhand,
I
don't
know
what
it
is.
But
it
is
some
percentage.
So
what
there
is,
there's
other
compounds,
inert
compounds
in
with
the
atrazine.

And
they
may
have
an
effect
on
their
own.
Some
of
the
compounds
in
with
diazinon
that's
mixed
in
as
an
emulsifiable
concentrate
are
known
teratogens
and
other
things
like
that.

So
you
are
moving
away
from
getting
a
look
at
a
pure
compound,
but
you're
looking
at
something
that
is
environmentally
realistic
because
that's
what
the
amphibians
are
getting
exposed
to.

Does
that
help?

DR.
ROBERTS:
Dr.
Thrall,
then
Dr.
LeBlanc.

DR.
THRALL:
I
was
particularly
interested
in
the
immunosuppression
aspect
of
atrazine
that
we
haven't
touched
on
very
much
here.
And
I
was
wondering
if
you
­­
I'm
not
familiar
with
the
Christin
2003
paper
you
alluded
to.
I
wondered
if
you
could
give
me
a
little
bit
more
detail
on
that,
the
pathophysiology
of
the
nematode
and
how
the
decreased
spleen
function
interacted
with
that
pathophysiology.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
51
DR.
SHEFFIELD:
That's
a
really
good
question.
I
have
the
paper
with
me.
I
don't
remember
all
the
details
of
it
because
it
just
came
out.
But
I
will
tell
you
I
noticed,
because
I
have
done
studies
like
this
before,
what
they
did,
they
used
the
pesticides
as
a
­­
they
exposed
the
organism
to
the
pesticides.
And
they
used
a
parasite
as
a
challenge
to
the
immune
system
and
tested
it
that
way.

And
they
used
a
mixture
which
included
atrazine.
Like
I
said,
I
have
the
paper
with
me.
I
can
provide
you
a
copy
of
that
paper
if
you
want
to
look
at
it.
I
don't
remember
all
the
details
offhand
because,

like
I
said,
it
just
came
out.

DR.
ROBERTS:
Dr.
LeBlanc
and
then
Dr.
­­
Dr.
Green,
do
you
want
to
follow
up
to
that
question?

DR.
GREEN:
I
want
to
follow
up.

I
also
have
that
paper,
Dr.
Thrall.
I
think
relevant
to
this
discussion
about
atrazine,
it
was
used
in
the
mixture
to
treat
these
rana
pipiens
or
various
rana
species.
And
then
these
species
were
inoculated
with
lung
worms.

The
conclusion
by
the
authors,
and
there
were
two
papers,
the
first
paper
just
stated
that
it
appeared
that
in
the
presence
of
atrazine
in
particular,
the
virulence
of
the
pathogen
was
enhanced,
and
that
it
accelerated
their
life
cycle
such
that
they
matured
and
were
present
in
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
52
greater
numbers
in
the
atrazine
treated
frogs.
And
they
speculated
that
it
might
have
something
to
do
with
immune
suppression.

The
follow­
up
paper
by
the
second
group
came
back,
repeated
the
same
experiments
and
did
T
lymphocyte
function
tests.
They
show
lymphocyte
suppression,
and
they
propose
­­
I
don't
believe
they
followed
up
with
this,
that
there
might
be
B
cell
suppression
and
immunosuppression
of
immunoglobulins
that
would
protect
against
parasites.

But
neither
of
those
two
papers
had
any
data
that
documented
that
the
frogs
were
compromised
in
growth
or
that
they
died
earlier
because
of
it.

So
they
didn't
extend
to
make
any
speculation
about
what
this
might
do
to
an
amphibian,
a
wild
population
that
was
exposed
to
atrazine.

But
I
think
we
have
both
of
those
papers
here
if
you
would
like
to
look
at
them.

DR.
SHEFFIELD:
Thank
you,
Dr.
Green.

The
one
other
comment
I
make
on
that
is
that
I
think
another
thing
that
would
be
useful
instead
of
using
the
parasite,
and
it's
something
that
I've
done
in
the
past,
is
use
a
pathogenic
challenge.

That
way
your
endpoint
is
mortality.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
53
So
you
get
to
see
immediately
what
the
effect
is.
So
you
don't
have
to
drag
it
out
and
look
if
the
parasite
is
going
to
kill
the
animal
down
the
line
or
not.

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

DR.
LEBLANC:
You
noted
that
atrazine
has
a
bioconcentration
factor
of
six
in,
I
think,
frog
tadpoles.

In
looking
at
that
number,
my
interpretation
would
be,
well,
at
least
one
thing
we
don't
have
to
worry
about
is
atrazine's
ability
of
propensity
to
bioaccumulate
in
these
animals.

But
it
raises
a
flag
in
your
mind.
And
I
just
wondered
if
you
could
expand
on
that
a
little
bit.
Is
there
something
unique
to
amphibians
and
bioconcentration
that
we
should
be
aware
of?
Or
is
six
really
a
significant
number?

DR.
SHEFFIELD:
I
don't
know.
It
just
struck
me
as
something
that
I
thought
was
­­
I
guess
maybe
it
was
because
of
the
fact
that
I
had
the
other
paper
in
mind
and
I
was
thinking
about
the
Lydy
work
that
showed
the
potentiation
and
synergy
of
atrazine
with
the
OPs.

The
organophosphate
insecticides
are
also
very
ubiquitous
in
the
environment.
That's
getting
away
from
the
purview
of
this
panel,

but
for
me,
anyway,
it
has
implications.

If
there
is
potentiation
and
synergy
effects
from
atrazine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
54
exposure
dealing
with
the
OPs,
that,
in
my
mind,
is
significant.

That's
what
it
was
­­
I
don't
think
the
six
is
alarmingly
high
or
anything.
It
indicates
that
they
do
accumulate
it
and
because
I
know
that
they
are
the
only
taxa
that
accumulates
OPs
also.

So
there
is
something
unique
about
amphibians
that
the
other
vertebrate
taxa
don't
have.

DR.
LEBLANC:
Atrazine
is
pretty
water
soluble,
and
six
is
pretty
nonsignificant.
So
I
think
in
that
respect,
at
least,
the
amphibians
are
probably
responding
like
everybody
else
and
they
are
not
accumulating
the
material.

DR.
ROBERTS:
Dr.
Delorme,
then
Dr.
Skelly.

DR.
DELORME:
With
respect
to
your
concerns
about
formulation,
are
you
aware
of
­­
versus
technical
product,
are
you
aware
of
any
studies
that
show
concentrations
of
the
formulants
that
are
mixed
with
atrazine
in
the
end­
use
products
and
their
concentrations
in
water?

Because,
obviously,
atrazine
isn't
applied
directly
to
water.
It
is
applied
on
fields.
There
is
going
to
be
differential
degradations
and
the
ratios
may
change.

Are
you
aware
of
any
data
that
would
indicate
what
of
the
formulants
get
into
aquatic
systems?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
55
DR.
SHEFFIELD:
Not
offhand
I'm
not.
I'm
not
sure
exactly
what
they
use
to
make
the
atrazine
formulations
either.
Sometimes
­­

I
don't
know
with
atrazine,
but
sometimes
a
lot
of
chemical
companies
are
very
sensitive
as
to
what
they
put
in
there.
And
that's
not
public
information.
I
don't
know
if
that's
the
case
with
atrazine
or
not.

So
I
don't
know
what
that
is.
And
I
don't
know
how
it
differentially
divides.
Once
it
is
applied
and
it
starts
its
fate
in
the
environment,
I
don't
know
how
it
divides
up
or
not.
All
I'm
maintaining
is
that
I
have
seen
many
studies
that
have
shown,
that
have
looked
at
technical
grade
pesticides
and
other
studies
that
have
looked
at
the
same
pesticide
with
formulations
and
found
very
different
results.

So
I'm
saying
as
a
real
world
type
situation,
I
think
the
formulations
are
important
to
look
at
regardless
of
how
they
divide
out.
The
chemical
pool
is
still
going
to
be
there
in
the
amphibian's
habitat
regardless
of
how
they
divide
out.
The
pool
is
still
going
to
be
there
of
all
the
different
chemicals
that
are
in
the
formulations.

Whether
some
of
those
chemicals
disappear
faster
than
others
­­
they
probably
do.
They
have
different
fates
because
they
are
different
chemicals.

I
don't
know.
I
don't
really
know
what
else
to
say
on
that
one.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
56
DR.
ROBERTS:
Dr.
Skelly?

DR.
SKELLY:
Dr.
Sheffield,
several
other
people
that
have
presented
to
the
panel
have
given
the
impression
that
field
studies
should
be
of
kind
of
secondary
importance
or
they
are
messy
or
something.
And
you
are
one
of
the
first
people
to
say
that
field
studies
may
be
of
primary
importance.

And
I
wondered
if
you
could
comment
further
on
that
and
specifically
talk
about
the
distinction
between
observational
field
studies
and
field
experiments
in
this
sort
of
work.

DR.
SHEFFIELD:
Well,
what
I
would
say
to
that
is
that
I'm
of
the
opinion
that
this
is
­­
this
and
many
other
pesticide
questions
are
very
complex.
It
is
not
going
to
be
answered
by
any
one
set
of
studies
necessarily.

You
have
to
use
an
integrated
lab
and
field
approach
to
this
thing.
And
you
may
have
to
do
something
in
the
lab,
then
go
to
the
field
and
at
the
same
time
go
back
in
the
lab
and
do
something
further.

And
then
maybe
you
have
to
go
back
in
the
field
again.

It
is
not
anywhere
near
­­
there
is
no
road
map
written
for
this.

It
is
kind
of
like
what
you
find
by
trial
and
error,
by
experimentation
in
the
lab
you
can
use
to
translate
to
what
you
might
want
to
look
at
in
the
field.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
57
Plus,
you
can
do
some
things
that
Dr.
Haye's
group
and
others
have
done
with
just
looking
at
­­
going
out
to
field
sites
that
are
contaminated.

They
didn't
­­
they
picked
their
sites
based
on
atrazine
sales
and
then
went
there
and
looked
at
various
sites.
You
can
do
a
little
bit
more
as
far
as
preselecting
sites
or
actually
setting
up
experiments,
experimental
ecosystems,
for
example.

There
has
been
some
work
on
that
done.
I
would
like
to
see
some
more
on
that
with
this
particular
question,
setting
up
experimental
ecosystems.
At
that
point,
you
have
a
lot
more
control
over
your
system
and
it
is
in
the
field.

I
would
say
that
the
field
studies
are
definitely
harder
to
deal
with.
They
are
harder
­­
there
just
is
a
lot
more
things
that
can
go
wrong.
But
they
are
very
useful.

And
I
my
emphasis
on
that
may
have
been
simply
because
of
the
fact
that
I
have
a
slight
bias
towards
­­
and
a
lot
of
enthusiasm,

towards
field
studies.

They
are
a
group
package.
They
have
to
go
together.
And
just
doing
one
at
a
time
is
not
necessarily
anything.
I
mean,
they
can
overlap.
There
is
no
written
rules
for
this.

Basically,
it's
adaptive
management.
You
are
going
along.
You
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
58
find
something.
You
adapt.
You
figure
out
what
to
look
at
next.
If
that
question
could
be
best
asked
in
the
field,
then
that's
where
you
go.
And
if
not,
then
you
do
more
lab
work.

So
it
is,
like
I
said
­­
in
some
of
these
areas
in
wildlife
toxicology,
you
are
going
down
a
road
that
has
not
been
paved
yet.
It
is
adaptive
management.

But
the
field
studies
are
important.
But
like
I
said,
there
is
a
lot
more
inherent
variability.
You
can
possibly
accept
a
little
bit
more
error
because
of
that
inherent
variability.
They
are
harder
to
do,

so
they
haven't
been
done
as
much.
But
they
are
valuable.

DR.
SKELLY:
Thank
you.

DR.
ROBERTS:
Thank
you,
Dr.
Sheffield.
We
appreciate
you
coming
in
and
sharing
your
comments
and
thoughts
with
us
this
morning.

DR.
SHEFFIELD:
Thank
you
very
much.
Good
look
to
all
of
you
in
the
next
two
days.

DR.
ROBERTS:
Next
public
commenter
is
Mr.
James
Tozi
(
ph)

on
behalf
of
­­
I'm
not
sure
what
the
note
says.
But
I'm
sure
he
will
tell
us.

And
he
will
be
followed
by
Mr.
John
Hall.

MR.
TOZI:
Thank
you,
Mr.
Chairman.
Good
morning.
I'm
Jim
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
59
Tozi
with
the
National
Business
Services.
And
with
a
title
like
that,

given
the
chairman's
remarks
that
you
want
scientific
remarks,
you
may
say
there
is
somewhat
of
a
divergence
there,
but
I
assure
you
there
won't
be.

One
of
our
main
roles
is
to
ventilate
to
the
public
the
scientific
basis
or
bases
for
regulatory
decisions.

And
the
question
is
how
do
we
do
that.
We
do
that
by
the
use
of
a
web
site
called
Cyberactivist.
US.
You
might
not
have
heard
of
that
site,
but
if
you
go
to
your
web
master,
it
is
probably
by
traffic
gradings
the
highest,
if
not
one
of
the
most
highly
trafficked,
web
sites
in
the
world
on
the
very
specialized
area
of
regulatory
policy.

And
it
is
used
throughout
the
world.

Now,
questions
given
that,
why
am
I
here.
I
think
this
panel,

and
it
just
happens
to
be
that
you
drew
this
card,
that
subsequent
to
the
passage
of
a
very
important
act,
the
Data
Quality
Act,
and
we
look
throughout
the
entire
government
for
proceedings,
this
proceeding
is
the
first
major
proceeding
that's
going
to
be
subject
to
the
Data
Quality
Act
when
the
agency
goes
to
use
it.

And
I
haven't
been
here
all
the
day,
so
you
might
have
heard
some
of
this
before.

Now,
keeping
with
the
chairman's
charge,
I'm
not
going
to
get
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
60
into
law
because
I
don't
think
the
law
is
any
different
than
which
most
scientists
do
anyway.

The
basic
two
important
portions
of
the
law
says
the
data
used
by
the
government
and
regulatory
decisions
first
must
be
reproducible
and
second
it
must
be
transparent.

And
many
scientists
when
a
law
passes,
big
deal,
we
do
that
all
the
time.
Well,
some
of
the
regulators
thought
it
was
a
big
deal.
The
scientists
did
not.

So
I
just
want
to
say
that
those
two
standards,
reproducible
and
transparency,
are
going
to
be
judged
on
anything
that
comes
out
of
this
whole
proceeding.

What
does
that
mean?
That
means
any
third
party
such
as
I,
on
National
Business
Services,
or
anyone
that
testifies,
including
this
committee,
you
can
write
and
say
anything
you
want.
I
can
write
and
say
anything
I
want.
It
need
not
be
reproducible
and
it
need
not
be
transparent.

However,
if
the
agency
is
going
to
use
anything
that
I
give
in
this
proceeding,
anything
any
other
third
party
gave
or
whether
they
are
going
to
use
your
report,
it
is
going
to
have
to
meet
those
two
standards.

If
it
does
not,
any
third
party
can
move
to
strike
that
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
61
information
from
decisionmaking
in
the
government.

Now,
a
point
that
I
wanted
to
emphasize
here
in
terms
of
what
we're
going
to
do
to
advise
this
committee,
we
are
very
interested
in
this
proceeding
because
of
the
data
quality
aspects.

But
I
think
you
can't
look
at
data
quality
unless
you
have
data.

And
so,
we
are
going
to
put
on
this
cyber
activist
web
site
Monday
evening
a
request
that
petitioners
have
made
from
EPA
on
the
data
on
this
proceeding.

I
haven't
looked
at
all
the
data,
so
don't
ask
me
about
all
the
details,
but
I'm
advised
that
­­
and
now
would
I
want
to
compete
with
you
on
what
the
data
means.
But
I'm
advised
that
there's
petitioners
that
have
asked
EPA
for
the
data
and
it
comes
out
in
these
kind
of
categories.

There
were
seven
data
sets,
I
understand,
but
this
will
all
be
on
the
web
site,
five
of
which
were
given
to
people,
some
of
which
were
encryptive,
some
of
which
was
withheld.

We're
not
going
to
take
a
position
on
these
matters
on
the
site.

That
complete
data
set
will
be
put
up
on
Cyberactivist.
US,
and
it
will
be
out
for
public
comment.

And
the
form,
you
are
familiar
with
forms,
it
is
not
a
web
site.

It
is
a
form
and
it
is
live
so
any
person
in
the
world,
anyone
in
this
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
62
panel,
anyone
in
this
room
can
look
at
that
data
set
and
make
a
comment
on
it
in
realtime,
which
we'll
be
expecting
a
lot
of
comments.

And
I
will
tell
you
when
we
put
it
up,
we're
going
to
urge
people
to
ask
three
questions
when
they
look
at
this
data
set.
First,
on
the
data
received,
was
it
reproducible.
Second,
how
does
use
of
encrypted
data
fit
into
regulatory
proceedings.
And
third,
when
the
federal
regulatory
agencies
withhold
data,
how
does
FICA
(
ph)

committees
and
how
do
people
handle
this.

We
think
by
putting
it
on
Cyberactivist.
US
and
the
charge
of
this
committee
as
well
as
the
composition
of
this
committee
on
the
web
site,
we
think
we'll
be
able
to
ventilate
way
beyond
this
proceeding
an
importance
how
this
type
of
information
is
used
by
influential
bodies
such
as
yours.

We
urge
you
to
look
at
that
site
and
the
comments
we'll
receive.

Some
of
them,
we've
received
thousands
of
comments.
Thank
you
very
much.

DR.
ROBERTS:
Thank
you.
I
think
there
is
a
question
from
Dr.
Heeringa
and
then
Dr.
Green.

DR.
HEERINGA:
Mr.
Tozi,
as
a
scientist
operating
in
a
university
environment,
I
believe
in
full
and
open
disclosure
on
this
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
63
data,
obviously,
there
are
proprietary
interests,
so
and
I
commend
you
on
that.

Just
a
comment
too
that
data
alone
aren't
sufficient.
You
need
documentation
to
accompany
it.
I
presume
that
part
of
your
posting
of
these
data
will
be
to
post
the
relevant
documents
that
describe
the
study
designs
and
protocols
used
to
collect
the
data
too.
Because
as
you
open
that
up
to
analysts,
it
is
going
to
be
very,
very
critical.
And
I
think
particularly
in
the
case
of
the
data
sets
we're
looking
at
here,

that
those
documentations
of
procedures
and
assessments
be
available
simultaneously
too.

MR.
TOZI:
Let
me
ask
you
one
question
on
this.
I
think
that's
very
good.
What
I
would
encourage
when
we
post
these
that
anyone
that
has
something
that
they
want,
and
it
is
very
easy,
you
just
press,

click
on
a
button
and
you
put
an
attachment
and
it
goes
right
up,
we
will
put
some
of
that
up.
The
initial
things
are
going
to
be
what
was
response
to
the
FOIA.

But
I
agree
with
you.
There
are
going
to
have
to
be
additional
postings
to
interpret
that.
I
have
looked
at
some
of
the
data.
It
is
not
my
technical
expertise
to
say
what
is
there,
but
I
agree
with
you.

We
would
encourage
anyone
that
is
involved
in
this
proceeding,

whether
around
this
table
or
in
the
room,
that
have
relevant
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
64
information,
because
I
will
tell
you
that
a
lot
of
other
agencies
way
beyond
EPA
­­
which,
by
the
way,
EPA
on
third
party
data
and
the
Data
Quality
Act
is
probably
one
of
the
leading
agencies.

You
can
criticize
some
stuff
they
have
done.
And
I
know
the
agency
from
when
they
were
born,
they
are
doing
a
very
good
job.
I
agree
with
you.
And
to
the
extent
I
can
get
that
kind
of
information
and
people
want
to
put
it
on,
we
welcome
them.

DR.
HEERINGA:
One
more
additional
comment.
Just
a
technical
question
for
people
who
would
access
this.
What
data
storage
format
will
it
be
in.
Will
it
be
in
Excel
spreadsheets?
SPSS
Sass?
Raw
data?

MR.
TOZI:
I
can
tell
you
how
you
do
it
now.
You
register
on
the
site.
You
type
something
in.
You
press
a
button.
I
think
it
comes
all
out
on
HTML.
There
is
attachment.
The
question
is
how
do
you
get
that
attachment
on
there.

Generally,
the
people
that
put
the
attachments
on
now
either
have
it
on
their
web
site
and
they
cut
and
paste
it
over
or
they
send
it
to
us
and
we'll
scan
it
in.

We
prefer
you
have
it
on
your
web
site
because
of
the
amount
of
traffic.

DR.
HEERINGA:
Thank
you.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
65
DR.
ROBERTS:
Dr.
Green
had
a
question.

DR.
GREEN:
Yesterday
and
today
several
presenters
have
alluded
to
encrypted
data,
which
I'm
not
clear
what
data
people
are
talking
about
that
has
been
encrypted.

Is
it
available
to
the
SAP
or
could
you
clarify
that?

MR.
TOZI:
Thank
you.
One
of
the
people
told
me
I
was
supposed
to
say
that.
Having
worked
in
OMB
for
years,
some
things
I
don't
address,
maybe
because
I
don't
want
to.

They
asked
me
this,
yes.
From
the
FOIA
requests
that
I
understand
that
people
filed
out
of
these
seven
­­
I
just
say
I
haven't
gone
through
all
the
data,
five
data
sets
came
back,
four
of
which
are
data.

When
I
say
encrypted,
I
understood,
but
others
in
the
room
would
know
more
about
this
than
I,
the
term
encrypted
means
it
was
in
some
machine
code
that
you
couldn't
interpret.

So
when
they
put
that
up,
I
want
to
ask
them
­­
they
will
identify
which
one
was
encrypted.
What
it
meant.
But
I
got
from
the
conversations
that
unless
you
had
some
transition
or
translation
codes,
you
could
not
interpret
the
data.
You
needed
some
other
kind
of
information
to
turn
it
into
meaningful
data.

DR.
ROBERTS:
Are
there
any
other
questions?
66
1
If
not,
thanks
very
much
for
coming
in
and
your
comments
and
2
letting
us
know
about
that.

3
Mr.
Hall
has
requested
the
opportunity
to
speak
to
the
panel.

4
As
a
heads
up,
he
will
be
followed
by
Mr.
Botts.

5
MR.
HALL:
Good
morning,
Mr.
Chairman,
members
of
the
6
panel.

7
My
name
is
John
Hall.
I'm
with
Hall
and
Associates.
I'm
here
8
today
representing
the
Kentucky
Corn
Growers
Association.

9
By
way
of
background,
I'm
an
environmental
engineer
and
an
10
attorney.
I
suppose
maybe
from
this
panel's
perspective
that
puts
two
11
strikes
against
me
since
I'm
not
a
scientist
per
se.

12
In
general,
I
have
spent
the
last
20
years
of
my
career
dealing
13
with
water
quality
related
issues
under
the
Clean
Water
Quality
act,

14
development
of
water
quality
standards
and
the
like.

15
The
information
that
has
been
presented
over
the
last
two
days,

16
as
you
are
aware,
not
only
has
an
effect
on
the
reregistration
of
17
atrazine,
but
also
affects
what
EPA
is
doing
in
terms
of
its
water
18
quality
standards
development
for
atrazine.

19
I
come
from
that
perspective
in
terms
of
my
review
as
to
the
20
information
presented.
In
the
water
quality
standards
issue
area,
they
21
generally
look
at
impacts
in
terms
of
what
I
will
call
the
big
picture.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
67
They
don't
typically
parse
out
mechanistically
why
an
impact
occurs,
what
the
exact
cause
of
the
effect
is,
but
they
look
at
whether
or
not
there
is
a
significant
effect
on
growth,
reproduction
or
survival,
the
three
main
endpoints
that
are
often
the,
if
you
will,
the
end
result
of
most
research
to
decide
what
the
real
world
impacts
are.

I
have
just
a
couple
comments
with
regard
to
the
information
presented
and
some
comments
regarding
an
earlier
presentation
to
this
panel.

Dr.
Hayes'
presentation
in
my
mind
raised
at
least
one
issue.
I
guess
others
have
thought
it
raised
several.
His
analysis
focused
on
abnormalities
in
gonadal
reproduction.

And
without
looking
at
the
environmental
significance
of
that,

the
question
I
had
was
that
these
abnormalities
were
classified
as
such
because
apparently
they
weren't
seen
in
certain
controls
during
the
study.
But
that
in
possibly
later
studies
that
were
conducted,
did
find
some
of
these
abnormalities
in
the
controls.

And
therefore,
I
guess
there
is
at
least
a
question
in
my
mind
as
to
if
the
later
studies
exhibited
some
of
these
abnormalities,
should
the
results
of
the
earlier
studies
that
classified
certain
things
as
abnormalities
kind
of
gone
back
and
corrected
those
charts.
If
that,
in
fact,
was
the
case.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
68
The
other
thing
I
would
note
is
that
in
terms
of
at
least
statistically,
by
classifying
abnormalities
as
something
you
don't
see
in
controls,
you
automatically
get
a
zero
abnormality
in
your
control
group.

And
then
when
you
do
your
statistics
and
comparison
to
that,

very
often
you
can
get
anything
being
mathematically
significant
in
terms
of
a
difference
compared
to
a
zero
effect
on
the
control.

I
would
note,
though,
that
at
least
out
of
all
the
studies
that
I
have
heard
about
today
and
in
the
prior
days,
that
the
environmental
significance
of
any
of
these
effects
that
have
been
noted
is
not
demonstrated.

I
have
not
seen
any
information
presented
that
these
effects
actually
cause
impacts
on
frog
populations,
which
in
the
end
is
I
think
what
we're
really
concerned
about.

Let
me
just
move
on
to
a
couple
other
points.
In
the
NRDC
presentation,
they
indicated
that
the
charge
to
this
panel
was
misplaced.
I
could
not
disagree
more.

The
charge
to
the
panel
is
not
misplaced.
What
we
have
is
different
agendas
occurring
on
from
different
parties
that
I
believe
are
a
little
misplaced.

NRDC's
view
is
that
endocrine
disruption
equals
ban.
In
other
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
69
words,
any
level
of
impact
you
find
regardless
of
its
significance
to
the
real
world
triggers
a
ban.

And
that
trigger
on
endocrine
disruption
shouldn't
be
ecological
significance.
It
is
simply
statistical
significance.

That
is
not
the
standard.
That
is
applicable
under
any
federal
law
I
am
aware
of.

The
Clean
Water
Act
requires
you
to
avoid
significant
adverse
impacts
on
aquatic
life.
And
the
endpoints
that
EPA
has
used
for
three
decades
are
growth,
reproduction
and
survival.

In
other
words,
does
this
have
a
real
world
effect
in
the
real
world.
You
may
find
an
effect,
a
histopathological
effect
in
the
organism.
But
the
question
is
does
that
cause
any
significant
adverse
effect
later
on
such
as
populations
are
affected
­­
unless
you
don't
classify
something
as
a
substantial
adverse
effect
and
regulate
it.

OPP,
the
pesticides
office,
uses
a
no
unreasonable
adverse
environment
effect
endpoint.
Again,
it
is
a
similar
type
of
legal
standard.

All
these
programs
in
summary
look
at
population­
based
impacts.
They
require
documentation
of
those
impacts
and
not
merely
speculation
that
the
impacts
occur.

And
that,
of
course,
is
the
charge
to
this
panel.
And
I
think
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
70
you
will
serve
us
all
greatly
by
answering
the
question
of,
whatever
impacts
these
studies
have
shown,
whatever
they
are,
is
there
any
information
that
shows
these
impacts
are
demonstrated
to
impact
these
organisms
at
a
population
level?

If
the
answer
is
no,
then
you
don't
regulate
the
pollutant
­­
the
constituent
more
stringently
at
this
time.
You
may
research
it
some
more
if
there
are
questions
that
you
have,
but
you
don't
decide
that
you
are
going
to
cancel
a
registration
over
something
that
is
not
connected
to
a
real
impact.

I
will
tell
you
one
thing
that
was
said
in
the
earlier
presentation
that
is
clearly
misleading
and
incorrect.

NRD
stated
that
EPA's
report
confirmed
that
there
were
real
readings
and
real
world
concerns
showing
significant
risks.
It
is
simply
not
true.

What
EPA
does
when
they
do
these
risk
assessments,
I
call
it
triage,
I
guess
is
the
best
way
to
look
at
it,
you
go
through
doing
a
very
gross
level
of
review
and
decide,
my
god,
if
you
can
pass
that
test,
you
are
clearly
not
an
effect.

That's
what
they
call
a
Tier
1
assessment.
And
they
make
multiple,
multiple
worse
case
assumptions.
And
if
you
can
pass
all
those
worst
case
assumptions,
you
are
out
the
door.
We
leave
you
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
71
alone.

If
you
can't
pass
the
series
of
worst
case
assumptions,
they
go
to
a
higher
level
review.
They
don't
declare
that
they
have
determined
something
that
is
causing
significant
real
world
effects.

You
move
on
forward.

One
example,
and
NRDC
relied
on
it
in
notifying
the
EPA,
they
intended
to
sue
them
under
the
Endangered
Species
Act
on
this
issue.

EPA
in
the
report
took
data
from
what
they
called
estuarine
areas.
The
entire
Chesapeake
Bay.
They
took
data
from
areas
that
were
even
30,
50
miles
upstream
in
non­
tidal
waters,
clearly
freshwater
systems.
And
in
their
report,
they
classified
that
as
estuarine
data.

You
would
have
thought
that
the
Chesapeake
Bay
had
30
parts
per
billion
of
atrazine
occurring
in
it.
In
fact,
it
does
not.

Well,
if
you
read
the
EPA
report,
and
not
very
carefully,
what
you
think
is
the
Chesapeake
Bay
has
these
enormous
high
concentrations
of
atrazine.

When
you
look
at
the
underlying
database,
you
find
the
concentrations
of
atrazine
in
the
Chesapeake
Bay
proper,
around
the
order
of
10ths
and
100ths
of
a
PPB
unmeasurable
levels.

So
in
other
words,
you
have
to
look
a
little
more
carefully
at
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
72
the
data
before
you
decide
that
what
is
in
that
EPA
report
has
demonstrated
there
is
real
readings,
real
world
concerns
before
you
make
claims
that
leather
back
turtles
are
adversely
affected
in
Chesapeake
Bay.

I
guess
I
would
like
to
end
with
I
applaud
what
this
panel
is
doing.
I
think
you
will
do
us
all
a
great
service
by
clarifying
what
the
real
effects
are,
what
the
information
really
does
show
and
whether
and
how
we
should
move
on
to
do
further
evaluations.

I
think
it
would
be
extremely
helpful
if
this
panel
confirms
what
I
believe
is
a
correct
assessment
on
EPA's
part
that
no
one
has
demonstrated
a
real
world
connection
to
population­
based
impacts
from
these
endocrine
disrupter
endpoints.

That
is
not
to
say
no
one
ever
will.
That
is
just
to
say
that
it
has
certainly
not
been
demonstrated
to
date.

Thank
you
very
much.
If
there's
any
questions,
I'll
be
happy
to
take
them.

DR.
ROBERTS:
Thank
you,
Mr.
Hall.
Let
me
ask
the
panel
if
they
have
questions.

I
don't
see
any.
Thanks
very
much
for
coming
in
and
sharing
your
comments
and
thoughts
with
us.

MR.
HALL:
You
are
welcome.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
73
DR.
ROBERTS:
Mr.
Botts
is
up
next.
Welcome,
Mr.
Botts.

MR.
BOTTS:
Thank
you,
sir.
My
name
is
Dan
Botts.
I'm
the
Director
of
the
Environmental
and
Pest
Management
Division
of
Florida
Food
and
Vegetable
Association.

And
as
such,
we're
a
voluntary
trade
association
that
represents
the
producers
of
about
60
different
commodities
in
the
state
of
Florida,
including
sugar
cane,
sweet
corn
and
sod,
all
of
which
use
atrazine.

But
the
reason
I'm
here
today
is
because
I
have
been
following
this
issue
for
at
least
10
years,
even
before
the
special
review
was
issued
in
1994.
And
have
looked
forward
to
some
resolution
at
some
point
down
the
road
in
determining
if,
in
fact,
atrazine
is
safe,
but
more
importantly,
that
triazines
as
a
class
of
compounds.
Larger
acreage
use
of
triazines
in
the
state
of
Florida
is
simazine
use
on
citrus.

And
because
of
that,
your
deliberations
here
today
pave
a
pathway
of
how
this
whole
class
of
compounds
will
be
dealt
with.

And
not
only
this
class
of
compounds,
but
others
as
we
move
into
the
new
process
of
determining
endocrine
disruption
impacts
and
how
to
regulate
those
impacts
for
the
office
of
pesticides
programs.

I
look
forward
to
your
deliberations,
and
looking
extremely
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
74
forward
to
your
final
report,
having
sat
through
many
SAPs
over
the
past
since
1997
after
FQPA
passed
on
all
kind
of
issues,
everything
from
modeling
for
OP,
cumulative
exposure
to
OPs
and
dose
response
curves
for
other
issues.

I'm
extremely
encouraged
by
the
discourse
that
has
already
occurred
among
the
panel
members
and
between
the
panel
members
and
the
people
presenting.
It
is
a
much
more
engaged
panel
than
a
lot
of
SAP
panels
have
been
in
the
past.

And
I
will
keep
this
extremely
brief
for
that
respect.
I
know
you
are
already
a
half
a
day
behind
schedule.
It's
going
to
be
extremely
time
consuming
to
go
through
the
questions
that
have
been
presented.
I
would
first
applaud
the
agency
for
taking
a
very
controversial
set
of
data
and
the
information
and
compiling
it
and
looking
at
it
in
a
manner
to
put
forward
the
review
that
they
did
in
their
white
paper
and
framing
the
issue
to
go
forward.

And
I'll
be
honest
with
you.
I
didn't
read
the
charge
of
the
questions
until
the
plane
ride
up
here,
so
I
don't
have
as
great
a
familiarity
with
the
intent
or
tone
of
the
questions
as
some
of
the
other
previous
speakers
have.

But
after
listening
to
the
comments
and
conversations,
I
am
sure
that
any
issue
that
might
be
floated
out
in
the
vagueness
or
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
75
generalities
associated
with
those
questions
will
come
forward
as
this
committee
deliberates.

I
would
draw
one
thing
to
your
attention,
though.
In
the
presentation
yesterday
afternoon,
there
was
one
slide
that
was
put
up
by
Dr.
Hayes
relative
to
several
epidemiology
studies
relating
to
the
human
health
effects
and
potential
carcigenocity
of
atrazine.

This
committee
can
be
informed
by
the
deliberations
of
the
SAP
panel
which
was
held
in
June
of
2000.
I'm
sure
that
report
is
up
on
their
web
site,
and
would
be
available
for
you.

Most
of
the
studies
that
were
referenced
in
his
presentation
yesterday
were
discussed
at
length
in
that
document.
Whether
they
are
relevant
or
not
to
deliberations
of
this
panel
I
would
suggest
are
probably
not
the
case.

Some
of
the
mechanistic
issues
that
might
have
come
up
in
the
previous
part
on
determining
the
cancer
risk
associated
with
atrazine
probably
are.

There
will
also
be
an
SAP
panel
later
this
summer
to
deal
with
epidemiological
information
relative
to
cancer,
and
those
discussions
would
probably
be
more
appropriate
at
that
point.

With
that,
I
encourage
you
to
move
forward
since
you
don't
have
to
listen
to
us
in
the
public
anymore
and
can
deliberate
among
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
76
yourselves.
And
I
look
forward
to
your
discussions
over
the
next
day
and
a
half.

DR.
ROBERTS:
Thank
you,
Mr.
Botts.
We
appreciate
you
coming
in
and
sharing
your
thoughts
and
comments.
Let
me
ask
the
panel
very
quickly
if
they
have
any
questions
for
you.

I
don't
see
any.
Again,
thanks
very
much
for
your
thoughts.

Dr.
Post
had
requested
the
opportunity
to
speak.
I
asked
if
she
was
here
earlier.
Is
she
here
now?

This
concludes
the
list
of
people
who
have
asked
previously
to
speak.
Let
me
ask
now
if
there
is
anyone
in
the
audience
who
has
not
had
an
opportunity
as
yet
to
address
the
panel
and
would
like
to
do
so.

This
will
be
your
last
opportunity
to
make
a
public
comment.

Because
when
the
public
comment
session
closes,
the
next
item
of
business
will
be
for
the
panel
to
begin
their
deliberations.

So
final
call
to
the
audience.
Is
there
anyone
who
has
not
yet
had
the
opportunity
to
address
the
panel
that
would
like
to
speak?

Seeing
none,
this
closes,
then,
the
public
comment
portion
of
the
meeting.
Let's
take
a
break
for
about
15
minutes
and
reconvene,

and
the
panel
will
then
begin
their
deliberation
of
the
questions.

(
Thereupon,
a
brief
recess
was
taken.)

DR.
ROBERTS:
Before
we
begin
deliberation
of
the
questions
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
77
posed
by
the
agency,
our
designated
federal
official,
Mr.
Paul
Lewis,

has
a
few
announcements
about
keeping
the
docket
squared
away,

some
announcements
about
submission
of
documents.

MR.
LEWIS:
Thank
you,
Dr.
Roberts.
Just
briefly,
during
the
break
we
distributed
to
the
panel
two
pieces
of
information.
One
is
a
CD
that
was
provided
by
the
EPA
Office
of
Pesticides
Programs
of
additional
data
for
the
panel
to
review,
to
consider.
In
addition,

additional
data,
additional
clarification
provided
by
Syngenta
based
on
comments
they
made
yesterday.

The
material
is
available
to
the
panel
now.
In
addition,
we
will
make
it
available
in
the
public
docket.

DR.
KELLEY:
Do
you
know
where
this
came
from?

MR.
LEWIS:
Let
me
give
this
back
to
the
chair.

DR.
ROBERTS:
The
question
came
from
Dr.
Kelley
regarding
a
CD
that
we
were
distributed,
and
she
wanted
to
know
the
source.
And
it
is
the
one
that
says
Hayes
data
sets.

DR.
STEEGER:
The
data
sets
on
that
CD
are
all
of
the
data
that
Dr.
Hayes
has
provided
to
support
his
report
to
Syngenta,
the
report
that
was
distributed
earlier
today.

It
also
contains
data
to
support
the
standard
operating
procedures
that
he
developed
in
his
lab
to
determine
feeding
rates.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
78
There
is
an
additional
file
on
there
that
provides
the
password
for
the
one
password
protected
data
set
that
is
contained
among
the
seven.

DR.
ROBERTS:
Thank
you,
Dr.
Steeger.

We
have
a
follow­
up
question
by
Dr.
Kelley.

DR.
KELLEY:
For
clarification,
so
the
feeding
data
which
were
alluded
to
yesterday,
does
this
just
include
the
methodology
for
gathering
them
and
not
the
results?
Or
does
this
include
the
results
of
different
feeding
regimens?

DR.
STEEGER:
My
understanding
is
they
are
the
results
of
different
feeding
regimens
to
determine
what
would
be
the
ideal
feeding
rate
for
xenopus
in
Dr.
Hayes'
lab.

DR.
KELLEY:
Thank
you.

DR.
ROBERTS:
And
just
as
a
general
comment,
the
panel
very
much
appreciates
the
data
and
the
reports
that
were
submitted
during
the
course
of
the
meeting
and
the
comments.

We
will
try
and
consider
and
utilize
those
in
our
deliberations
as
best
we
can.
But
time
is
very
short
for
us
to
be
able
to
consider
those.
So
we'll
do
the
best
we
can
with
it.

Yes,
Dr.
Steeger.

DR.
STEEGER:
I
want
to
make
one
more
comment
about
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
79
encryption
of
the
data.

As
Dr.
Hayes
indicated
in
his
presentation,
his
treatments
are
color
coded.
One
of
the
data
sets
is
codes
dot
X
L
S.
It's
an
Excel
spreadsheet.
It
provides
the
way
of
associating
the
colors
with
the
actual
treatment
levels.

DR.
ROBERTS:
Thank
you,
Dr.
Steeger.

Dr.
Bradbury,
the
panel
has
received
a
great
deal
of
public
comment
over
the
last
couple
of
days,
a
lot
of
information.

We
appreciated
the
opportunity
to
hear
from
the
investigators
and
asked
them
questions.
We
now
turn
to
deliberation
of
the
questions
posed
by
the
agency.
And
I
wondered
if
you
had
some
introductory
remarks
or
comments
you
wanted
to
make
to
help
us
make
sure
we
have
the
right
focus
as
we
begin
to
that.

DR.
BRADBURY:
Thank
you,
Mr.
Chairman.
I
appreciate
the
opportunity.

What
I
would
like
to
do
is
just
spend
a
few
minutes
to
recapping
where
we
have
been
over
the
last
few
days
and
perhaps
help
set
the
stage
for
moving
through
the
questions.

What
I
would
like
to
do
is
sort
of
again
get
us
back
into
the
context
of
the
science
and
the
risk
assessment
issue
that
we're
facing
and
to
put
this
question
in
the
context
of
the
agency's
ecological
risk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
80
assessment
guidelines,
the
process
whereby
the
science
is
evaluated
to
help
inform
the
regulatory
decisions
that
the
agency
needs
to
make.

We
talked
about
this
on
Tuesday
morning,
that
one
of
the
important
phases
of
an
ecological
risk
assessment
is
the
problem
formulation
stage.
It
provides
the
context,
it
provides
the
foundation
for
proceeding
further
in
the
actual
risk
assessment.

The
risk
assessment,
of
course,
is
designed
to
help
inform
the
risk
management
decision,
science
for
a
purpose.
The
science
for
this
purpose
is
to
help
inform
decisions
about
the
potential
risk
of
atrazine
to
amphibians.

As
we
discussed
on
Tuesday
morning,
ecological
risk
assessments
tend
to
be
an
iterative
process.
And
some
of
the
speakers
over
the
last
couple
days
noted
the
risk
assessment
that
has
already
been
completed
in
terms
of
the
potential
effects
of
atrazine
on
aquatic
community
structure
and
function.

As
part
of
that
it
iterative
process,
questions
came
up
from
the
public,
from
NRDC
and
from
others.
And
in
the
context
of
interacting
with
our
agency's
risk
managers,
it
formulates
another
question.

The
question
being
whether
or
not
atrazine
can
cause
developmental
effects
on
amphibians,
and,
if
so,
what
could
be
the
consequences
if
that
effect
occurs.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
81
As
one
moves
in
theory
from
problem
formulation
into
the
analysis
phase,
that's
where
one
blends
the
exposure
information
with
the
effects
information
and
takes
that
into
risk
characterization
where,
in
fact,
we
attempt
to
provide
an
estimate
of
the
magnitude
and
likelihood
of
potential
adverse
effects.
If
you
will,
try
to
establish
the
probabilities
that
certain
events
can
happen
at
different
atrazine
exposure
concentrations.

The
goal
of
the
risk
assessment
is
to
provide
that
exposure
response
profile
and
articulate
what
the
ecological
significance
of
that
exposure
response
profile
is.
And
then
communicate
that
to
the
risk
manager.

And
as
we
have
heard
over
the
last
few
days,
there
is
a
number
of
issues
that
come
into
play
in
making
a
decision
about
the
registration
and
reregistration
of
a
pesticide.
Science
is
only
one
aspect
of
the
overall
decision.
There
are
many
other
factors
that
go
into
making
a
regulatory
decision.

The
challenge
for
the
scientific
community,
not
just
the
scientists
in
the
office
of
pesticide
programs
and
the
office
of
research
and
development,
but
the
scientific
community
that
spans
academia
industry
and
public
groups,
is
to
ensure
that
the
science
that
goes
into
making
these
decisions
is
clear
and
transparent
and
provides
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
82
the
risk
managers
an
objective
understanding
of
what
we
know
and
what
we
don't
know
so
that
certainties
in
their
decisionmaking
is
clear
to
all
concerned.

So
that's
the
context.
We're
inside
the
box.
We're
inside
the
box
of
science,
but
it
is
science
for
a
purpose
and
it
is
science
that
has
urgency.
Decisions
have
to
be
made.
Making
no
decision
is,
in
fact,

a
decision.
And
it
is
a
decision
that
is
made
in
the
context
of
whatever
scientific
uncertainty
or
certainties
we
have
at
the
time.

So
with
that
in
mind,
let
us
just
walk
through
again
a
little
bit
of
context
on
the
road
map
that
we're
working
through.
Let's
take
a
look
at
problem
formulation
again.
And
that's
really
where
we
are
right
now.

The
white
paper
is,
in
fact,
a
problem
formulation.
Problem
formulation
is
where
we
integrate
available
information
to
try
to
establish
some
sense
of
risk
assessment
endpoints
in
the
context
of
environmental
management
goals
and
start
to
articulate
what
those
measures
of
effects
could
be
to
make
estimates
about
those
risk
assessment
endpoints.

Based
on
the
integration
of
available
information,
we
focus
on
the
formation
of
risk
hypotheses
and
try
to
develop
a
conceptual
model
that
could
be
used,
a
working
hypothesis,
if
you
will,
to
relate
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
83
the
information
that
we
do
have
in
terms
of
exposures
and
effects
to
then
set
up
the
analysis
plan
for
actually
undertaking
the
risk
assessment.

It
is
very
critical
in
a
risk
assessment
that
has
a
lot
of
attention
and
has
a
lot
of
implications
to
use
this
problem
formulation
step
to
its
utmost,
to
have
it
be
rigorous,
to
gain
the
input
from
the
scientific
peers,
to
ensure
that
we've
thought
this
thing
through
the
way
it
needs
to
be
thought
through.

At
the
end
of
the
analysis
phase,
there
are
sort
of
three
broad
paths
one
could
imagine
going
down.
One
possibility
is
that
through
the
analysis
of
the
available
information
and
dialogue
with
the
risk
managers
it
could
be
concluded
that
there
is
no
need
to
do
a
risk
assessment,
that
there
is
sufficient
certainty
in
the
information,
in
the
context
of
the
certainty
that
is
required
to
make
a
regulatory
decision
that
there
is
no
need
to
do
a
risk
assessment.
That's
one
possible
outcome
of
problem
formulation.

Another
outcome,
possible
outcome
of
problem
formulation
would
be
that,
in
fact,
there
is
sufficient
information
to
formulate
the
working
hypotheses
in
the
conceptual
model.
The
analysis
plan
lays
out
how
to
use
the
available
information
to
proceed
with
the
risk
assessment.
Acknowledgment
of
potential
uncertainties
are
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
84
recognized
in
the
analyses
plan,
but
the
decision
is
that
one
could
go
ahead
and
actually
start
doing
the
risk
assessment
to
move
into
the
analysis
phase
and
begin
ultimately
to
characterize
risk.

A
third
outcome
of
problem
formulation
could
be
that
there
is
sufficient
information
to
formulate
a
reasonable
conceptual
model
to
formulate
reasonable
risk
hypotheses,
to
formulate
a
reasonable
working
hypothesis
and
to
develop
an
analysis
plan
that
outlines
the
amount
of
information
that
is
available
and
outline
the
data
gaps
that
are
facing
the
ability
to
proceed
with
the
risk
assessment
with
varying
levels
of
certainty.

That's
sort
of
where
we're
at
right
now
in
taking
a
look
at
the
white
paper,
the
problem
formulation
and
gaining
your
insight
and
input
and
wisdom
on
the
paths,
the
possible
paths
to
go
forward
after
problem
formulation.

Is
there
sufficient
information
to
proceed
with
the
risk
assessment
and
quantify
to
varying
degrees
the
probability
of
adverse
effects
to
amphibians
in
terms
of
development
based
on
varying
atrazine
exposures?

Is
there
sufficient
information
to
say
there
is
no
need
to
go
forward;
there
is
no
plausible
risk
hypothesis
for
the
potential
of
atrazine
to
cause
adverse
effects
on
amphibian
development
or
are
we
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
85
somewhere
in
between.
That's
the
ultimate
question
before
us.
And
a
series
of
questions
to
help
us
get
to
where
we
need
to
get.

One
thing
I
want
to
point
out
is
that
in
the
process
of
going
through
problem
formulation
and
ultimately
getting
to
risk
characterization,
we're
talking
about
uncertainties,
we're
talking
about
what
we
know,
we're
talking
about
what
we
don't
know
and
we're
pulling
that
information
together.

There
has
been
some
discussion
that
it
is
a
weight
of
evidence
approach.
In
fact,
it
is
not
a
weight
of
evidence
approach.
In
fact,

the
agency's
ecological
risk
assessment
guidelines
are
very
clear
that
it
is
not
a
weight
of
evidence
approach.

In
fact,
it
is
termed
the
lines
of
evidence
approach.
And
to
quote
from
the
agency
guidelines,
the
phrase,
lines
of
evidence,
is
used
to
deemphasize
the
balance
of
opposing
factors
based
on
assignment
of
quantitative
values
to
reach
a
conclusion
about
a
"
weight."
in
favor
of
a
more
inclusive
approach,
which
evaluates
all
available
information
even
evidence
that
may
be
qualitative
in
nature.

So
the
point
is
we're
not
balancing
pounds
of
information.

We're
looking
at
lines
of
evidence.
If
I
could
be
so
bold
as
to
change
the
analogy
a
bit
from
the
risk
assessment
guidelines,
you
may
want
to
think
about
it
as
pieces
of
evidence,
pieces
of
information
and
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
86
think
about
how
we
might
try
to
build
a
puzzle.
Think
about
building
a
jigsaw
puzzle
with
lots
of
pieces
on
the
table,
all
different
shapes,

all
different
sizes,
all
different
contours.
Some
with
many
edges,

some
with
a
few
edges.

Are
all
those
pieces
of
information
on
the
table
even
part
of
the
same
jigsaw
puzzle?
If
they
are,
how
many
of
those
pieces
on
the
table
can
one
start
to
put
together
to
start
to
build
the
picture?
Is
there
enough
pieces
on
the
table
that
are
starting
to
connect
to
each
other
to
actually
see
what
the
picture
is
and
to
talk
about
the
picture
and
describe
the
picture
in
great
detail?

Or
is
it
apparent
that
the
pieces
on
the
table
don't
even
allow
one
to
start
to
put
the
pieces
together?
Is
it
even
possible
that
the
pieces
don't
even
belong
in
the
same
puzzle?
Or
are
there
enough
pieces
starting
to
come
together
that
one
can
start
to
imagine
what
that
puzzle
could
look
like
if
one
could
get
more
pieces
and
can
start
to
see
what
kind
of
pieces
would
be
the
most
critical
to
start
building
the
picture,
to
start
painting
the
picture.

I
think
the
other
idea
in
terms
of
not
using
the
term
weight
but
using
the
term
lines
or
pieces
of
evidence
is
that
some
pieces
in
this
puzzle
may
be
very,
very
small,
but
have
lots
of
edges
and
they
connect
lots
of
pieces
that
are
on
the
table.
They
may
be
very
small,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
87
but
they
may
be
very
critical.

I
think
another
aspect
of
thinking
about
this
in
terms
of
painting
a
picture
or
putting
together
a
puzzle
is
that
as
we
have
certain
pieces
come
together,
there
may
be
some
gaps
between
some
of
the
pieces.
But
depending
upon
the
shape
of
the
pieces
and
how
those
pieces
are
coming
together,
it
may
be
very
relatively
easy
to
imagine
what
that
missing
piece
would
look
like,
i.
e.,
we
might
be
able
to
extrapolate
to
what
that
missing
piece
would
look
like,
a
reasonable
level
of
confidence.

In
some
cases
the
edges
of
the
pieces
may
lead
us
out
into
parts
of
the
universe
we
haven't
been
before,
and
that
may
require
actually
getting
some
pieces
to
help
put
the
picture
together.

So
it
is
an
inclusive
process.
There
is
no
right
or
wrong.
There
are
no
winners
or
losers
in
this
process.
The
only
winners
are
the
people
of
the
United
States
getting
the
kind
of
information
it
takes
so
that
the
risk
management
decisions
can
be
informed.

The
only
winners
in
this
operation,
in
this
endeavor
is
a
science,
and
the
science
being
blended
together
to
maximize
all
the
information
possible
to
make
the
most
informed
decision
that
we
can
make.

In
going
through
problem
formulation,
one
sort
of
gets
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
88
impression
it's
a
linear
process.
It
is
not.
You
sort
of
take
different
information
and
you
work
it
through.
Let's
just
recap
real
briefly
sort
of
where
we
are
coming
from
in
the
white
paper.

We
initially
laid
out
on
Tuesday
morning
the
statement
of
the
risk
management
goal,
the
environmental
management
goal
as
well
as
the
risk
assessment
endpoints,
and
those
being
the
reproductive
and
recruitment
capability
of
native
anurans.
I
don't
think
I
need
to
go
into
great
detail
on
that.

But
what
I
would
like
to
do
is
sit
back
and
think
about
what
that
means
in
terms
of
being
able
to
make
some
estimates
about
reproduction
and
recruitment
of
native
anurans.
That's
a
big
question.

That's
a
question
that
brings
in
all
sorts
of
fields
of
biology,

landscape
ecology.

Problem
formulation
helps
us
get
started
on
the
process
of
putting
this
information
together
to
answer
that
question.

When
the
agency
has
to
take
on
chemical
risk
assessments
in
the
ecological
realm,
we
take
on
a
challenge
that,
with
all
due
respect
to
the
human
health
risk
assessors
of
the
agency,
I
think
we
have
a
bigger
challenge
and
a
more
exciting
challenge,
because
we
have
to
work
across
many
layers
of
biological
organization.

And
the
examples
of
the
levels
of
biological
organization
on
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
89
that
slide
don't
even
capture
all
the
levels
of
biological
organization
as
Carl
Richards
and
others
on
the
panel
know,
but
to
at
least
get
it
all
to
fit
on
one
slide,
I
think
I
can
get
the
point
across.

This
concept
on
this
slide
has
been
published
by
many
folks
in
many
different
venues.
But
I
think
it
illustrates
very
nicely
some
of
the
challenges
in
the
integration
of
information
that
is
required
as
one
goes
forward
in
a
risk
assessment.

One
could
enter
our
examination
of
a
question
at
any
level
of
biological
organization,
but
the
concept
is
that
in
general
as
one
goes
down
levels
of
biological
organization,
one
creates
greater
understanding
of
the
potential
or
actual
interaction
of
a
chemical
with
a
biological
system.

But
certainly
as
one
goes
down
in
levels
of
biological
organization,
our
understanding
of
the
relevancy
of
those
events
at
the
population
or
community
or
landscape
level
become
less
and
less
certain.

So,
in
fact,
doing
an
ecological
risk
assessment
is
a
blending
of
many
levels
of
biological
organizations
to
inform,
to
help
us
understand
what
is
going
on
and
to
help
us
understand
the
relevancy
of
the
ecological
risk
assessment.

I
think
this
panel
is
a
nice
example
of
the
kinds
of
skills
and
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
90
disciplines
and
professions
that
need
to
come
together
to
understand
in
the
problem
formulation
context
of
where
we
are
and
where
we
need
to
head.

So
if
I
could
try
one
more
analogy.
We
talked
about
pieces
of
a
puzzle
in
terms
of
creating
that
puzzle
and
creating
the
picture
to
understand
where
we
are
and
what
the
world
looks
like.
We
also
could
think
about
it
in
terms
of
threads
and
how
threads
get
woven
together
to
create
a
tapestry,
a
tapestry
that
has
texture,
a
tapestry
that
has
a
lot
of
vibrant
aspects
to
it.

And
it
means
we're
blending
in
the
skills
and
talents
of
folks
that
are
experts
in
molecular
biology
and
interactions
of
chemicals
with
receptors
or
enzymes
as
well
as
interacting
with
folks
that
are
experts
in
landscape
ecology
and
understanding
how
spatial
and
explicit
descriptions
of
habitat
are
critical
for
understanding
the
population
demographics
of
amphibians
or
other
species.

So
our
challenge
is
how
to
weave
these
threads
together
to
create
a
coherent
picture
that
blends
understanding
and
relevancy.

So
that
as
a
context,
I
would
like
just
a
couple
more
minutes
if
I
can,
Mr.
Chairman,
for
Tom,
Tom
Steeger,
and
I
to
just
touch
base
again
on
the
integration
of
available
information
for
this
specific
risk
assessment,
problem
formulation
for
atrazine,
touch
on
the
conceptual
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
91
model
again
just
to
touch
base
and
kind
of
get
us
all
reequilibrated
and
then
to
just
touch
base
on
the
first
aspect
of
our
analysis
blend
for
the
problem
formulation.

DR.
STEEGER:
Over
the
past
two
days,
panel
members
have
received
range
of
input
from
researchers
engaged
in
studying
the
effects
of
atrazine
on
amphibian
development.

Panel
members
have
been
provided
copies
of
each
of
the
studies
discussed
and
EPA's
assessment
of
the
studies.

To
a
large
extent,
the
presentations
have
focused
on
the
results
generated
from
research
efforts.
As
has
been
pointed
out
by
a
member
of
the
SAP,
it
isn't
sufficient,
though,
to
look
exclusively
at
the
data.

But
rather,
you
have
to
also
consider
the
study
design
and
study
conditions
in
which
the
data
were
collected.

As
Steve
Bradbury,
Joe
Tietge
and
I
have
pointed
out,
the
agency
follows
a
process
for
conducting
ecological
risk
assessment.

The
initial
stage
of
ecological
risk
assessment
is
problem
formulation
where
risk
assessors
work
with
risk
managers
to
integrate
available
information
on
a
chemical
to
define
potential
hazards,
their
impact
on
assessment
endpoints
and
the
uncertainties
associated
with
the
measurement
endpoints
used
to
identify
potential
hazards.

Typically,
the
agency
relies
on
guideline
studies
to
assess
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
92
risks.
However,
open
literature
and
nonguideline
studies
are
also
considered.

Consistent
with
the
agency's
process
for
evaluating
studies,

studies
were
evaluated
using
the
following
criteria:
Experimental
design,
study
protocols,
protocols
and
quality
assurance,
the
strength
of
the
cause
effect
relationship,
whether
there
was
a
dose
response,

whether
the
observed
effects
have
plausible
mechanism
of
action
that
is
consistent
with
what
is
known
about
the
chemical,
and,
finally,

whether
the
measured
effects
are
ecologically
relevant.

Did
we
know
what
to
expect?
No.
While
the
agency
routinely
receives
studies
that
have
a
very
standardized
protocol
and
established
databases
on
which
to
gauge
the
conduct
and
outcome
of
the
studies,
the
studies
under
current
review
on
amphibian
effects
represented
a
new
area
of
information.

Did
the
agency
have
expectations
on
whether
the
data
were
realistic.
As
Dr.
Giesy
testified,
the
data
are
what
they
are.
Agency
reviewers,
though,
examine
how
the
data
were
collected
and
analyzed
to
determine
whether
they
would
have
come
to
the
same
conclusions
as
the
study
authors.

The
agency,
however,
is
looking
for
input
from
this
panel
on
whether
the
data
provided
from
the
current
suite
of
studies
are
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
93
reasonable
for
gauging
the
effects
of
atrazine
on
amphibian
development.

Also,
there
was
some
discussion
about
the
agency's
unrealistic
expectation
of
a
monotonic
dose
response
curve.
The
agency
is
not
focussed
on
the
shape
of
the
curve
as
much
as
on
the
consistency
of
the
dose
response.

We
do
not
insist
that
a
chemical
exhibit
a
monotonic
dose
response,
but
simply
that
we
are
able
to
understand
and
project
what
type
of
response
might
be
expected
in
a
consistent
fashion
from
exposure
to
particular
levels
of
atrazine.

By
now,
the
panel
is
painfully
aware
that
the
agency
has
reviewed
seventeen
studies
that
were
received
as
a
February
28th
consent
decree
deadline.
12
of
the
studies
were
submitted
by
the
registrant.
Five
were
drawn
from
open
literature.
Seven
studies
were
conducted
in
the
laboratory
and
ten
were
conducted
in
the
field.

While
the
presentations
by
researchers
over
the
past
two
days
have
focused
on
study
results,
the
agency
has
a
number
of
evaluation
criteria
that
extend
beyond
the
data
and
examine
the
conditions
under
which
the
data
were
collected.

As
I
indicated,
the
data
evaluation
records
for
the
17
studies
focused
primarily
on
methodological
inconsistencies
that
were
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
94
considered
to
have
potentially
critical
impacts
on
the
study.

While
Dr.
Sielken
was
correct
in
pointing
out
that
collapsing
replicates
can
potentially
increase
the
likelihood
of
a
type
one
error,

that
is,
drawing
a
conclusion
that
there
is
a
statistical
difference
when
one
does
not
actually
exist,
agency
reviewers
conducted
statistical
analyses
based
on
both
the
original
study
design
and
on
collapsing
the
data
sets
in
order
to
explore
whether
statistical
relevancy
could
be
extracted
from
the
highly
variable
data
sets.

However,
it
is
important
to
keep
in
mind
that
regardless
of
what
approach
was
used
to
view
the
data,
confounding
effects
across
all
of
the
studies
limited,
if
not
precluded,
the
utility
of
the
data
regardless
of
how
they
were
analyzed.

Thus,
atrazine
contamination
in
the
controls,
poor
water
quality,
poor
growth
and
development
and
survival,
high
variability
in
endpoint
measurements,
lack
of
reproduceability
and
unresponsive
positive
controls
were
recurrent
themes
that
were
considered
critical.

One
question
that
was
posed
on
Tuesday
was
whether
panel
members
should
use
any
of
these
studies
in
determining
whether
there
are
sufficient
data
to
gauge
whether
atrazine
is
impacting
gonadal
development.

However,
we
look
to
the
panel's
collective
expertise
on
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
95
amphibians'
endocrinology,
plus
extensive
laboratory
and
field
research
coupled
with
the
understanding
of
the
science
of
risk
assessment
to
provide
feedback
to
the
agency
on
that
very
question.

The
agency
views
that
the
studies
do
have
utility
in
helping
to
identify
potential
effects
on
amphibian
development.
Additionally,

the
studies
provide
insight
on
the
sources
of
variability
and
they
provide
insight
on
the
appropriate
test
species
and
study
conditions
that
may
be
utilized
in
future
studies.

If
the
risk
managers
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.

I
suspect
that
it
is
often
times
frustrating
for
the
public
to
recognize
that
the
agency
is
not
omniscient
in
its
understanding
of
the
available
research
on
a
particular
subject.

While
certain
background
information
would
be
particular
relevant
to
the
measurement
endpoints
under
consideration,
the
agency
may
simply
not
be
aware
of
its
existence.

Additionally,
because
of
research
limitations,
the
agency
has
come
to
rely
on
the
testing
of
surrogate
species
to
be
representative
of
the
effects
of
a
very
broad
range
of
organisms.
Thus,
while
it
might
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
96
be
ideal
to
test
actual
species
that
may
be
likely
to
be
exposed
to
pesticides,
these
data
are
not
typically
available.

Additionally,
surrogate
test
species
are
not
selected
based
on
their
sensitivities
to
chemicals,
but
rather
on
their
ability
to
survive
under
laboratory
conditions.

Also,
while
laboratory
test
conditions
may
not
be
deemed
as
ecologically
relevant,
they
are
intended
to
provide
sufficient
control
over
environmental
conditions
to
permit
better
elucidation
of
potential
treatment
effects.

Steve
will
now
recap
on
the
approach
that
the
agency
is
going
to
recommend
to
address
the
current
uncertainties.

DR.
BRADBURY:
Following
up
on
what
Tom
just
indicated,
a
written
description
of
the
risk
hypothesis
and,
as
the
guidelines
recommend
to
the
extent
possible,
try
to
visualize
what
the
risk
hypothesis
is
in
a
conceptual
model.
And
the
white
paper
provided
an
image
similar
to
this.

Before
we
get
started
on
talking
about
this
conceptual
model,

I'm
reminded
of
a
quote,
I
think
it
is
by
George
Fox
that
says,
all
models
are
wrong,
but
some
are
useful.
And
I'm
really
used
to
that
phrase.
My
research
is
more
on
quantitative
structure
activity
relationships.
And
we're
always
dealing
with
the
fact
that
you're
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
97
trying
to
predict
what
you
don't
know.

So
all
models
don't
capture
reality.
But
if
used
in
the
proper
context,
they
provide
insight
into
how
to
move
forward.

I
think
another
way
to
view
it
and
perhaps
a
less
harsh
way
would
be
to
say
that
all
models
have
limitations,
but
all
can
be
useful
if
placed
in
the
proper
context.

I
think
a
lot
of
what
we
heard
over
the
last
few
days
and
we'll
be
discussing
over
the
next
day
and
a
half
will
be
the
context
of
models,
be
they
biological
models,
be
they
experimental
models,
be
they
representations
of
the
field
to
try
to
capture
what
all
of
the
ecosystems
in
the
country
do,
they
are
models.
And
they
are
models
used
to
try
to
help
us
find
a
path
to
move
forward.

So
in
that
spirit,
a
conceptual
model
is
a
working
hypothesis,

but
you
have
to
start
somewhere.
You
have
to
start
somewhere.

So
in
our
conceptual
model,
we're
focusing
on
that
top,
for
me,

the
top
left­
hand
corner
of
our
conceptual
model.
That's
the
focus
and
we're
going
to
move
forward.
We
have
talked
about
the
idea
of
linking
the
potential
molecular
effects
to
the
effects
in
the
working
hypothesis
in
elevated
E
2
(
ph)
to
the
concentration
on
or
the
focus
initially
on
gonadal
effects
in
males
which
would
then
lead
to
the
issue
of
impaired
fertility,
reproductive
and
success
and
then
to
our
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
98
risk
assessment
endpoint.

As
we
discussed
in
the
white
paper,
the
analysis
plan
of
moving
forward
is
a
phased
approach.
Again,
the
idea
that
one
doesn't
have
to
know
everything
to
know
something
to
then
moved
forward
in
a
risk
management
criteria.

This
then
lays
out
the
first
part
of
our
analysis
plan,
which
gets
at
taking
a
look
at
the
apical
effects
of
gonadal
development
in
amphibians,
provide
the
logical
point
to
break
through
that
logic
train
that's
in
the
conceptual
model.
Again,
to
go
back
to
the
earlier
slide,

understanding
combined
with
ecological
relevancy.

This
working
hypothesis
starts
off
with
sort
of
posing
the
question
on
ourselves,
can
we
establish
with
greater
confidence
the
potential
of
atrazine
to
cause
developmental
effects,
start
to
get
a
handle
on
what
the
stressor
response
profile
is
for
that
effect
as
a
launching
pad
to
either
simultaneously
look
for
ecological
relevancy
as
well
as
mechanistic.

That's
what
we'll
be
looking
forward
to
hear,
the
panel's
discussion
on
the
pathway
of
moving
forward.

The
phase
one
aspect
of
the
analysis
plan
we
feel
is
the
most
important
part
or
at
least
would
hope
that
the
panel
would
spend
a
fair
amount
of
time
on
what
we're
proposing
as
phase
one.
Or
if
you
feel
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
99
there
is
a
different
phase
that
we
should
start
on,
of
course
we'll
be
very
happy
to
entertain
that.

But
at
least
from
our
perspective
as
a
working
hypothesis
for
you
all
to
test,
our
phase
one
component
then
is
focusing
on
whether
or
not
atrazine
exposure
results
in
gonadal
effects
in
males
and
perhaps
females,
and
again,
to
try
to
determine
what
the
dose
response
relationship
is.

We
have
no
preconceived
notion
of
what
a
"
right"
dose
response
relationship
is.
That's
not
the
issue.
The
issue
is
what
is
the
dose
response
relationship
and
what
kind
of
confidence
do
we
have
in
quantifying
that
dose
response
relationship.

In
the
context
of
the
phase
one
studies,
we
indicated
on
Tuesday
morning
a
number
of
issues
in
terms
of
getting
started.
We
started
a
little
bit
of
discussion
on
Tuesday
morning
about
the
choice
of
the
biological
model
to
get
started
on.
There
probably
will
be
a
number
of
biological
models
we
can
consider.

Again,
all
models
have
limitations,
but
all
can
be
useful
if
placed
in
the
proper
context,
the
context
of
understanding
chemical
toxicity,
the
context
of
studying
ecological
relevancy.

There
is
probably
not
one
model
that
can
answer
all
our
questions.
But
what
are
the
types
of
models
we
should
use
for
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
100
different
types
of
questions
at
hand.

We
spent
a
lot
of
time
as
Tom
summarized
in
looking
at
the
available
information
in
terms
of
some
standards
that
exist
in
the
scientific
literature
as
a
function
of
scientific
bodies,
in
terms
of
some
of
the
basic
conditions
that
are
required
to
do
aquatic
toxicology
testing.

The
history
of
aquatic
toxicology
is
about
35
year
old
now.

Over
35
years
there
has
been
a
lot
of
advancements
in
the
technology
of
how
to
do
aquatic
toxicology
testing.
There
has
been
a
lot
of
work
to
describe
what
the
conditions
of
organisms
should
be
expected
and
the
terms
of
doing
those
tests.

And
in
this
context,
the
American
Society
for
Testing
Material
has
established
standards,
expectations
of
quality
in
terms
of
undertaking
an
aquatic
toxicology
test,
including
test
with
amphibians.
We
feel
that's
an
important
criteria
to
take
into
account
in
terms
of
doing
aquatic
toxicology.

This
will
be
one
of
the
exciting
parts
of
the
discussion,
because
the
methodologies
used
in
aquatic
toxicology
may
not
be
the
same
methodologies
that
are
used
in
developmental
biology
and
they
may
not
be
the
same
methodologies
that
are
used
in
field
biology
protocols.
We're
going
to
have
to
weave
that
tapestry,
weave
those
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
101
threads
together
to
try
to
create
a
vibrant
textured
tapestry.

Obviously,
you
know
where
we're
coming
from
in
terms
of
how
we
think
there
is
a
way
to
approach
maintaining
appropriate
quality
in
the
bioassays
and
combined
with
keeping
track
at
the
end
that's
going
to
be
required
to
test
some
of
these
hypotheses.

We
obviously
have
a
proposal
on
the
table
in
terms
of
using
a
flow­
through
technology
to
meet
those
ASTM
standards.
That
doesn't
mean
that
all
tests
that
could
be
done
to
answer
these
questions
have
to
be
done
with
flow
through.
In
our
opinion,
the
most
important
criteria
is
going
to
be
whatever
method
is
used,
it
has
to
meet
ASTM
standards.

Whatever
is
needed
then
to
answer
the
statistical
power
issues,

let
it
be
what
it
will
be,
but
ASTM
standards
we
feel
are
important
because
they
are
based
on
35
years
of
aquatic
technology
and
we're
somewhat
nervous
in
backing
away
from
what
people
have
been
developing
over
the
years
of
aquatic
toxicology,
to
not
belabor
the
point,
some
of
the
data
quality
indicators
that
we
discussed
previously
and
which
I'm
highlighting
just
in
the
last
few
minutes.

So
with
that,
we'll
wrap
it
up.
Again,
just
to
bring
it
back
home
to
the
risk
assessment
issue
at
hand,
is
there
sufficient
information
available
to
describe
the
certainties?
What
are
the
certainties?
What
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
102
are
the
uncertainties?

Risk
managers
can
use
that
information
to
decide
if
we
should
go
ahead
and
do
a
quantitative
risk
assessment
or
if
no
risk
assessment
needs
to
be
done.
Under
the
assumption
there
may
be
a
request
for
greater
certainty
in
performing
a
risk
assessment
for
atrazine,
we
laid
out
an
analysis
plan
to
describe
what
we
considered
would
be
important
information
to
reduce
the
uncertainty.

We
look
forward
to
the
dialogue
with
the
panel.

DR.
ROBERTS:
Thank
you,
Dr.
Bradbury.
I
think
your
presentation
and
your
comments
by
Dr.
Steeger
are
very
important
to
help
us
refocus
as
we
begin
our
deliberation
of
the
questions
posed
by
the
agency.

A
considerable
amount
of
effort
has
been
taken
to
put
the
available
information
regarding
potential
effects
of
atrazine
on
amphibians
before
the
panel.
It
is
now
time
for
the
panel
to
go
to
work
and
offer
our
best
scientific
evaluation
and
recommendations.

I
would
like
to
go
ahead
and
take
the
first
question
and
ask
either
Dr.
Bradbury
or
Dr.
Steeger
if
they
would
pose
the
first
question
to
the
panel.

DR.
BRADBURY:
I'll
turn
it
over
to
Tom
to
read
the
questions.

DR.
STEEGER:
First
question.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
103
In
reviewing
the
available
laboratory
and
field
studies,
the
agency
used
a
number
of
criteria
to
evaluate
individual
investigations.
Criteria
such
as
experimental
design,
test
protocols
and
quality
assurance
information
were
used
to
ascertain
the
reliability
of
the
generated
data
in
terms
of
its
ability
to
adequately
assess
the
hypothesis
that
atrazine
elicits
developmental
effects
in
amphibians,
and,
if
so,
the
nature
and
strength
of
associated
dose
response
relationships.

Then
part
one
of
a
two­
part
question.

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

And
secondly,
given
the
evaluation
criteria
employed
by
the
agency,
please
comment
on
EPA's
overall
characterization
of
the
currently
available
studies.

DR.
ROBERTS:
Dr.
Kelley,
as
lead
discussant,
I
will
ask
you
to
start
out
on
this.
I
will
leave
it
to
you.
Do
you
want
to
take
both
parts
together
or
do
you
want
to
have
discussion
on
Part
A
and
then
discuss
Part
B?

DR.
KELLEY:
What
I
thought
I
would
do
is
to
read
a
response
to
the
question
that
I
have
written
after
discussion
with
several
but
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
104
not
all
panel
members,
and
it
includes
both
parts.
And
then
ask
the
panel
for
comments
on
what
I
have
written
so
that
it
can
be
revised
to
more
accurately
reflect
the
viewpoint
of
the
panel.

Let
me
ask
the
panel
to
begin
with
whether
that's
an
acceptable
procedure
to
them.

DR.
ROBERTS:
Sure.

DR.
KELLEY:
And
let
me
just
tell
you.
Feel
free
to
disagree
violently.
Not
a
problem.

DR.
ROBERTS:
I
don't
want
them
to
disagree
violently.

DR.
KELLEY:
Violently
in
the
academic
sense.
That
is
to
say
how
could
you
possibly
say
that.
That
would
be
violent
for
an
academic.

DR.
ROBERTS:
Proceed
then.

DR.
KELLEY:
Bear
in
mind
we
may
amend
this.
All
right?

We
felt
that
the
review
was
thorough
and
the
conclusions
were
appropriate,
given
the
data
reviewed
by
the
EPA.
And
here
I'm
referring
to
the
seventeen
studies.

We
agreed
that
additional
studies
are
warranted.
There
are
several
studies
supported
by
the
registrant
or
by
other
agencies
that
indicate
that
atrazine
can
cause
developmental
abnormalities.
On
the
other
hand,
a
range
of
abnormalities
are
reported
and
they
are
not
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
105
consistent
from
study
to
study
(
bearing
in
mind
that
the
literature
includes
several
anuran
species.)

Further,
though
not
considered
in
the
white
paper,
the
findings
are
consistent
with
studies
in
other
vertebrate,
both
aquatic
(
e.
g.
fish)

and
terrestrial
(
some
rodents.)
Given
the
conservation
of
many
basic
pathways
for
endocrine
regulation,
these
studies
are
relevant
to
this
white
paper
focused
on
anurans.

Comments
from
the
panel?

DR.
ROBERTS:
I
have
one
and
then
I'll
open
it
up
to
other
commenters.
A
asks
regarding
the
approach
in
the
criteria
used
by
the
EPA
to
evaluate
those
studies.

And
maybe
I
missed
it.
But
did
you
basically
agree
with
their
criteria?

DR.
KELLEY:
I
thought
­­
yes.
The
review
was
thorough.
I
think
that's
implicit.
So
let
me
add
a
phrase
saying
that
the
criteria
were
appropriate.

DR.
ROBERTS:
Yes.
If
you
believe
so.
I
think
we
should
make
that
explicit
in
the
response.

DR.
KELLEY:
I'll
add
that.

DR.
ROBERTS:
And
then
Dr.
Coats.
You
had
a
comment.

DR.
COATS:
Yes.
I'm
a
secondary
discussant
on
this.
I
would
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
106
like
to
read
my
opinion
which
is
certainly
consistent
with
the
opinions
expressed
already
by
Dr.
Kelley.
The
agency's
approach
on
criteria
are
valid
and
address
a
number
of
very
important
points
from
the
exposure
side
of
the
risk
assessment,
which
is
where
some
of
my
concerns
are.

Number
1,
the
data
evaluation
reports
could
include
a
description
of
the
analytical
methods,
that
be
ELISA
or
chromatographic
methods
to
enhance
the
completeness
of
the
reports
and
their
interpretation.

Measured
concentrations
which
are
extremely
important
of
the
chemical
in
the
water
need
to
be
obtained
and
any
potentially
bioactive
metabolite
should
also
be
quantified
as
mentioned
in
the
white
paper.

In
some
research
papers,
the
exposure
concentrations
are
nominal
with
no
measured
concentrations
provided
or
recovery
confirmed
given,
but
no
values.

The
importance
of
having
measured
concentrations
could
be
addressed
more
directly
in
weighing
the
validity
of
the
work.

Thirdly,
the
significance
of
the
exposure
method
was
pointed
out,
including
some
of
the
shortcomings
on
the
static
renewal
systems.
The
ASTM
standards
for
flow­
through
systems
should
be
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
107
followed
and
the
method
strongly
encouraged.

Fourth,
body
burdens
must
be
measured
in
the
organisms
to
reflect
the
degree
of
exposure
and
help
explain
the
mode
of
action.

Once
again,
bioactive
metabolites
also
should
be
included.

Determination
of
residues
in
specific
tissues
would
also
be
very
valuable,
but
at
the
very
least
the
whole
body
residues
could
confirm
the
exposure.

On
the
hazard
side
of
the
risk
assessment
equation,
it
is
critical
that
the
dose
response
relationships
be
shown
as
pointed
out
by
the
agency.
Even
if
it
is
an
atypical
inverted
U
shape,
nonmonotonic
response,
a
dose
response
curve
can
be
generated
if
appropriate
concentrations
are
tested.

This
of
course
would
help
delineate
safe
versus
unsafe
concentrations,
but
also
could
assist
in
elucidating
mechanism
of
action.

DR.
ROBERTS:
Dr.
Coats,
I
think
I
heard
three
different
kinds
of
comments
there.
One
was
perhaps
some
deficiencies
in
reporting
among
the
existing
studies.

Some
were
aspects,
experimental
aspects
that
perhaps
should
be
addressed
in
other
studies.
But
the
question
really
asks
about
the
EPA's
characterization
of
currently
available
studies.
And
I
think
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
108
some
of
your
comments
touched
on
that.

But
I
want
to
be
sure
that
when
we
respond
to
this
question
that
we
sort
of
don't
mix
other
things,
other
points
that
we
want
to
make
in
with
it.
So
I
think
it
is
going
to
be
important
that
we
give
it
as
clear
a
response
as
we
can
to
be
about
things
that
you
feel
and
other
panel
members
feel
that
any
deficiencies
in
their
characterization
where
there
are
things
that
they
didn't
describe
or
that
they
inaccurately
described
that
perhaps
could
be
tweaked
in
the
document.

Dr.
Kelley?

DR.
KELLEY:
I
actually
did
in
my
written
comments
separate
out
A
and
B,
I
see
here.
And
so
I
did
have
an
additional
comment,

which
is
that
the
design
of
the
experiments
reviewed
by
the
EPA
and
the
analysis
of
the
data
were
flawed
in
many
instances,
which
was
the
EPA's
conclusion.

And
then
this
relates
to
the
dose
response
­­
if
there
is
a
threshold
effect
for
atrazine,
that
threshold
is
not
firmly
established.

With
respect
to
dose
response,
any
requirement
that
functions
be
monotonic
are
clearly
inappropriate.
And
as
we've
heard,
I
think
the
EPA
would
agree
with
that.

So
I
think
these
were
in
discussions
of
the
panel
members
among
the
most
important
issues
that
came
to
mind,
the
threshold
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
109
dose
for
observing
an
effect
if,
in
fact,
an
effect
can
be
observed
strikes
us
as
a
very
important
piece
of
information
to
be
gathered.

DR.
ROBERTS:
So
you
think
that
perhaps
the
report
as
the
EPA
said
should
clarify
that
in
their
criteria
to
evaluate
the
studies
there
was
not
an
insistence
on
a
monotonic
dose
response
relationship.
Is
that
correct?

DR.
KELLEY:
And
I
think
actually
that
it
is
just
a
wording
matter.
I
think
use
of
the
word
threshold
is
useful
here,
unless
it
has
some
technical
meaning
that
escapes
me.
So
threshold
is
the
question
here.

And
then
of
course
it
will
have
a
shape,
and
we
can't
predict
in
advance
what
shape
that
will
be.

DR.
ROBERTS:
Dr.
Green.

DR.
GREEN:
I'm
the
third
discussant
on
that
question.
And
following
up
on
what
Dr.
Coats
just
read
to
you,
I
think
what
we
were
trying
to
convey
is
that
we
would
like
for
the
EPA
to
include
in
their
approach
and
criteria
for
evaluating
the
study
the
fact
that
so
few
studies,
if
any,
evaluated
water
levels
and
tissue
levels
in
the
animals.

And
that
should
be
emphasized
in
an
EPA
report
describing
their
approach
and
criteria
and
noting
that
so
few
studies
did
that
with
any.
That
would
be
quite
helpful.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
110
DR.
ROBERTS:
Thank
you.

Comments
from
other
panel
members?

Dr.
Matsumura?

DR.
MATSUMURA:
I
really
appreciate
what
Dr.
Kelley
said
about
the
comparative
aspect
in
comparing
the
action
of
atrazine
to
other
organisms,
particularly
in
the
vertebrates,
so
that
we
formulate
some
idea
what
happens
in
other
organisms.
I
think
this
document
could
have
more
information
on
the
other
organisms.
Some
people
have
pointed
out
too.

And
regarding
the
question
that
Joel
Coats
raised,
you
are
using
the
water
concentration
of
atrazine.
In
some
setting,
it
is
just
crazy.

As
Dr.
Hayes
pointed
out,
the
next
day
when
you
come
a
concentration
may
be
quite
different
if
it
has
rained
the
next
day
or
dried
up.

So
criterion
that
one
should
use
really
is
the
residue
levels
in
the
frogs
themselves,
including
metabolites.

So
I
know
our
chairman
don't
want
to
mix
up
with
later
questions,
but
still
other
criterion
we
should
include
whether
residues
can
be
found.

In
the
Reed
study
from
Illinois,
they
were
measuring
the
residues.
And
really
using
the
water
concentration
is
not
a
really
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
111
good
method.
We
have
to
identify
the
residues,
including
the
metabolites.
Even
if
it
is
sort
of
ephemeral,
still
we
have
to
do
that.

DR.
ROBERTS:
Dr.
Richards?

DR.
RICHARDS:
I
believe
that
in
a
general
way
that
the
agency's
review
of
these
is
pretty
much
in
line,
particularly
with
respect
to
the
field
studies.
I
think
that
they
have
done
a
fairly
reasonably
qualitative
look
at
availability
of
the
studies
and
the
data
and
information
that
was
provided.

DR.
ROBERTS:
Are
there
other
criteria
that
you
would
like
to
see
added
or
articulated?

DR.
RICHARDS:
I
think
that
we
will
address
that
in
other
questions,
specifically.

DR.
ROBERTS:
Okay.
Dr.
DeLorme?

DR.
DELORME:
I
just
wanted
to
concur
that
for
the
most
part
the
criteria
they
used
were
reasonable
and
that
they
are
reflecting
the
need
for
sound
science
and
consistency
in
the
science
used
to
conduct
ecological
risk
assessments.
That's
their
job.
And
they
want
to
make
sure
that
the
science
that
they
are
using
and
the
studies
that
they
are
using
are
sound.

With
respect
to
the
overall
characterization
of
the
studies,
I
found
it
to
be
reasonable.
Certainly,
there
were
some
minor
points
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
112
that
we
might
disagree
with
on
specific
studies,
but
the
overall
characterization,
I
think
they
did
a
good
job.
They
have
identified
their
concerns
with
the
studies
from
a
risk
assessment
and
from
a
scientific
perspective.

They
have
also
identified
any
conclusions
or
information
or
contributions
that
they
felt
the
study
made.
And
they
have
also
identified
uncertainties
that
result
from
the
studies.
And
I
think
they
have
done
a
reasonable
job
in
doing
that.

DR.
ROBERTS:
Dr.
Skelly.

DR.
SKELLY:
Overall,
I
think
the
EPA
did
a
good
job
of
characterizing
the
studies.
And
I
just
want
to
make
a
comment
about
how
studies
were
divided
into
categories.

The
distinction
between
laboratory
and
field
studies
was
made.

And
in
a
couple
places
in
the
report,
I
don't
know
whether
it
is
referring
to
all
of
the
studies
that
were
submitted,
but
they
were
characterized
as
experiments.

I
just
would
like
to
point
out
that
the
field
studies
that
are
included
here
include
something
called
a
microcosm
experiment,

which,
in
my
experience,
is
actually
a
mesocosm
experiment.
That's
probably
just
technical.

But
mesocosm
experiments
are
not
field
experiments.
I
think
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
113
that's
an
important
distinction
when
we're
thinking
about
this.
And
most
of
the
field
studies
were
observational,
which
isn't
a
criticism,

but
I
think
the
criteria
for
evaluating
a
field
observational
study,
a
field
experiment,
of
which
from
what
I
can
tell
there
are
none
here,

and
a
mesocosm
experiment
might
be
distinct.

DR.
ROBERTS:
So
some
of
the
terminology
may
need
to
be
tweaked
a
little
bit
in
terms
of
how
they
are
described
or
classified,
is
that
your
recommendation?

DR.
DELORME:
So
that
the
terminology
might
be
tweaked
and
also
the
criteria
for
evaluating
things
like
the
distinction
between
a
field
observational
study
and
a
field
experiment
in
terms
of
what
sort
of
expectations
you
have
for
variation
and
environmental
conditions
could
be
different.

And
I
think
we're
going
to
talk
about
that
a
bit
more
when
we
get
to
the
next
question.

DR.
ROBERTS:
Thank
you.
Just
a
heads
up
for
Dr.
Kelley
­­

DR.
KELLEY:
Would
you
like
me
to
put
that
wording
into
this
first
thing?
I'll
put
it
in.
You
can
check
it
later.

DR.
ROBERTS:
Just
as
a
heads
up
for
Dr.
Kelley
as
the
lead
discussant,
when
we
finish
our
discussion,
I'm
going
to
ask
you
to
sort
of
give
me
your
sense
of
the
capsule
summary
of
this.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
114
DR.
KELLEY:
Thank
you.

DR.
ROBERTS:
Dr.
Denver.

DR.
DENVER:
I
just
want
to
support
a
point
that
Dr.
Coats
made.
And
that
goes
to
the
criterion
used
to
evaluate
especially
the
endocrine
data
and
the
need
to
really
validate
the
assays
that
are
used,

especially
the
ELISA
assays.
And
that
will
come
up
in
another
point
later
and
we'll
discuss
it
further.
But
I
think
that
that
can
be
raised
at
this
point
also.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
First
of
all,
I
felt
that
the
criteria
certainly
was
appropriate,
and
I
was
very
thankful
of
the
EPA
for
making
our
life
a
lot
easier,
I
think,
in
reviewing
this
document
and
assisting
us
in
that
manner.

I
would
like
to
comment
on
a
somewhat
minor
but
I
think
important
terminology
consideration.
Particularly,
in
light
of
the
fact
that
we
have
had
discussions
relating
to
aqueous,
measuring
aqueous
concentrations
versus
measuring
tissue
residues.

I
certainly
think
tissue
residues
are
important,
certainly
from
a
mechanistic
standpoint,
to
understand
what
the
burdens
are
and
the
effect
that
organs
might
be
carrying
in
terms
of
atrazine
loads.

Concentrations,
aqueous
concentrations
are
important
because
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
115
the
bottom
line
is
that's
probably
how
the
regulatory
agency
is
going
to
make
decisions,
based
upon
exposure
concentrations
and
not
true
doses
that
the
animals
get.

Throughout
the
document,
there
is
discussion
of
dose
response
relationships,
and
I
would
just
advise
that
consideration
be
given
where
appropriate
to
use
the
term
concentration
response
relationship
and
use
dose
response
relationship
where
appropriate,
but
don't
interuse
the
two.

DR.
ROBERTS:
I
agree.
Any
other
comments?
Dr.
Delorme.

DR.
DELORME:
I
just
wanted
to
add
something.
Dr.

Matsumura
and
I
believe
Dr.
Kelley
indicated
that
one
of
the
criteria
that
they
felt
might
have
been
addressed
was
inclusion
of
data
on
other
vertebrates.

I
think
that
was
­­
especially
as
goes
to
ecological
relevance,
it
might
be
­­
had
some
consideration.

DR.
ROBERTS:
Other
comments?

I'm
thinking
about,
trying
to
think
about
whether
or
not
I
agree.

I
think
their
intent
was
to
at
least,
I
think
what
this
statement
is,
to
summarize
studies
on
a
particular
topic.

I
think
at
some
point
you
have
to
try
and
perhaps
put
that
in
context,
in
broader
context.
I
think
at
that
point
you
would
look
for
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
116
analogy
from
other
species.

But
I'm
not
sure
I
would
criticize
this
summary
for
not
going
­­

making
that
broader
look.
I
think
it
had
a
very
specific
purpose,
not
to
say
those
other
comparisons
might
not
be
important
in
terms
of
understanding
potential
effects,
but
I
thought
it
was
very
well
focused.

Personally
speaking,
I
was
satisfied
with
it.

Any
other
comments?
Dr.
Gibbs.

DR.
GIBBS:
Just
one
quick
question
in
terms
of
the
case­
by­
case
treatment
of
all
the
studies.
I
felt
that
on
the
possibility
for
type
2
statistical
errors
to
be
committed
in
all
of
the
studies
wasn't
very
frequently
considered.

I
think
the
focus
was
on
the
statistical
significance
reported,

but
I
didn't
­­
and
clearly
EPA
is
aware
of
the
problems
with
sample
sizes.
But
I
felt
many
of
the
studies
had
serious
issues
with
not
being
able
to
detect
effects
should
they
have
occurred.
And
I
just
felt
across
studies
that
that
consideration
of
type
2
errors
wasn't
particularly
prevalent.

DR.
ROBERTS:
Dr.
Heeringa.

DR.
HEERINGA:
I
just
add
my
support
to
that
comment.
I
think
­­
as
I
reviewed
these
studies
and
the
EPA's
assessment
and
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
117
review
of
them,
I
generally
agree
with
the
conclusions,
but
I
also
support
the
last
comment,
too,
that
I
think
if
we
look
at
these
and
the
other
contributing
sources
of
error
we
have
talked
about,
that
many
of
these
studies
are
underpowered
to
detect
the
type
of
effects
that
we
are
measuring.

In
spite
of
some
of
the
identified
measurement,
contamination
or
quality
problems,
all
of
these
contribute
potentially
to
biases
in
results,
but
there
are
other
sources
of
variable
errors
that
I
think
really
would
lead
us
to
assume
that
even
given
nominal
sample
sizes
that
a
lot
of
these
studies
are
underpowered.

And
a
recommendation
I'll
make
later
on
is
that,
if
anything,

when
we
start
into
new
studies,
if
they
are
conducted,
that
they
be
overpowered
to
start
with
simply
so
we
don't
find
ourself
in
this
quandary
of
being
right
on
the
edge
of
type
1
versus
type
2
error
problems.

DR.
ROBERTS:
Any
other
comments
or
responses
to
1
A
and
1
B?

Dr.
Kelley,
I
know
you
have
been
taking
notes.
To
the
extent
possible,
can
you
try
and
summarize
your
discussion?

DR.
KELLEY:
Yes.
I'm
not
totally
up
to
speed
on
Dr.
Skelly's
comments.
But
I
will
try
to
summarize
them.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
118
So
in
general,
the
panel
felt
that
both
the
evaluation
criteria
and
the
thoroughness
of
the
EPA's
white
paper
were
appropriate.

The
panel
will
get
to
the
hypothesis
in
a
moment.
But
the
panel
did
support
the
hypothesis
that
there
was
enough
data
out
in
the
literature
given
this
evaluation
to
proceed
with
evaluating
the
hypothesis
that
atrazine
may
contribute
to
developmental
abnormalities
in
amphibia.

Some
concerns
were
raised
by
the
panel
with
regard
to
the
evaluation,
areas
of
emphasis
that
the
panel
felt
should
have
been
weighed
in
more
heavily
than
other
areas.
That
includes
possibility
of
type
2
errors,
which
is,
for
the
public,
also
my
stat
students
always
get
this
confused,
but
the
inability
to
detect
effects
that
are
really
there
for
whatever
reason.
And
some
of
the
possible
reasons
were
outlined
here.
So
that
was
one
area
of
concern.

Another
area
of
concern
was
the
ability
to
compare
across
studies
between
nominal
concentrations
of
application
of
atrazine
and
actual
tissue
concentrations
of
atrazine.

And
the
third
was
that
the
characterization
of
the
field
data
in
terms
of
what
was
actually
obtained
was
felt
by
panel
members
to
be,

how
can
I
say,
somewhat
superficial,
I
think
would
probably
be
the
best.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
­­
119
DR.
GREEN:
Observational.

DR.
KELLEY:
Observational.
Thank
you.

Although,
there
were
experiments
in
those
field
data,
right,
so
it
wasn't
completely
observational.
The
microcosm
experiments
did
have
experimental
and
control
groups.

DR.
ROBERTS:
Dr.
Skelly
can,
I
think,
fill
in
on
that.

DR.
SKELLY:
We
can
talk
about
it
afterwards
DR.
KELLEY:
Anyway,
he's
going
to
fix
my
terminology
here.

DR.
SKELLY:
The
point
was
that
a
mesocosm
experiment
is
not
really
a
field
experiment.
There
is
an
important
distinction.

DR.
KELLEY:
Yes.
So
that
the
use
of
the
technical
terms.

In
a
field
experiment,
you
don't
take
the
animals
out
of
the
field
and
throw
them
into
a
tank.
You
manipulate
them
in
their
environment
in
situ.
You
get
rid
of
the
red
on
the
red
wing
black
bird's
wing,

right,
or
something
of
that
sort.
That
would
be
a
field
experiment,
as
opposed
to
bringing
them
into
a
mini
lab.
So
there
was
some
concern
about
that.

But
these
concerns
were
felt
by
the
panel
to
be
relatively
minor,

and
to
not
abrogate
the
conclusion
of
the
panel
that
the
initial
analysis
by
the
EPA
was
thorough,
that
the
criteria
were
appropriate
and
that
the
evaluation
was
a
complete
one.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
120
DR.
ROBERTS:
Would
anyone
like
to
edit
or
make
suggestions
regarding
that?

Dr.
Bradbury,
is
our
response
reasonably
clear?

DR.
BRADBURY:
One
question
of
clarification.

In
interpreting
or
evaluating
future
studies,
am
I
to
understand
the
consensus
of
the
panel
that
future
studies
should
have
measured
aqueous
concentrations
of
atrazine
at
a
minimum
and,
depending
upon
the
nature
of
the
hypothesis
being
tested
in
a
given
study,
perhaps
tissue
concentrations
as
well
and
perhaps
activated
metabolites?

DR.
KELLEY:
I
think
actually
the
panel
might
feel
a
bit
more
strongly
than
that.
It
might
feel
that
measured
tissue
concentrations
of
atrazine
is
not
a
perhaps
but
should
be
a
requirement
going
forward.

But
I
don't
want
to
speak
for
the
panel
as
a
whole.

DR.
ROBERTS:
Dr.
Green.

DR.
GREEN:
I
still
think
in
the
laboratory
environment
that
water
concentrations
of
atrazine
are
important
to
know
through
the
experiment.

DR.
ROBERTS:
Others
want
to
weigh
in
on
this?
Dr.
Coats
and
then
Dr.
LeBlanc.

DR.
COATS:
I
want
to
reemphasize.
I
believe
that
the
tissue
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
121
concentrations
would
be
very
helpful
and
water
absolutely
needs
to
be
measured.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
It
is
my
opinion
that
measuring
aqueous
concentrations
is
mandatory
and
measuring
tissue
levels
is
desirable
based
upon
the
questions
being
asked
in
a
given
experiment.

DR.
ROBERTS:
And
speaking
for
myself,
I
would
concur
with
that
as
well.

Anyone
else
want
to
weigh
in?

Are
there
any
other
follow­
ups
or
clarifications?

Dr.
Steeger.

DR.
STEEGER:
I
would
like
to
make
one
additional
comment
for
the
benefit
of
many
of
the
researchers.

Dr.
Heeringa
is
correct
that
type
2
errors
are
a
critical
consideration
in
the
design
of
studies.
Many
of
the
researchers
did
attempt
to
design
their
studies
with
that
in
mind.
The
data
evaluation
records
that
were
prepared
didn't
fully
capture
those
efforts,
but
I
think
that
a
genuine
effort
was
made.

Unfortunately,
as
things
progressed
in
the
labs
and
things
didn't
turn
out
as
well
as
they
had
hoped,
the
study
designs
didn't
support
the
data
in
terms
of
controlling
some
of
the
variability
that
really
got
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
122
out
of
hand.

DR.
ROBERTS:
I
believe
Dr.
Heeringa
wanted
to
respond
to
that.

DR.
HEERINGA:
Thank
you
very
much.
I
agree
completely.
I
didn't
intend
to
say
that,
you
know,
forehand
that
there
was
inappropriate
planning
in
terms
­­
I
think
across
these
studies
that
adequate
planning
went
in.

But
as
you
say,
during
the
course
of
the
studies
as
we
learn
more
about
some
of
the
measurement
problems
and
some
of
the
other
issues
that
came
in,
that
clearly
this
caused
those
sort
of
a
priori
expectations
to
be
modified
and
in
a
way
that
my
recommendation
is
really
that
for
the
future
that
we
can
anticipate
these
things
will
reoccur
to
varying
degrees,
and
we
may
have
even
other
unmeasured
sources
of
error
in
the
experimental
process.

And
instead
of
sort
of
delaying
the
result
of
this,
let's
sort
of
design
with
a
margin
of
error
on
these
type
2
error
problems
so
that
we
can
accommodate
them
as
they
arise
in
laboratory
and
testing
situations.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
I
would
just
like
to
revisit
aqueous
concentrations
again
for
a
moment.
Certainly,
in
many
of
the
studies
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
123
that
I
have
read,
it
appeared
to
me
that
concentrations
of
atrazine
were
measured
only
in
solutions
that
were
freshly
prepared.

And
certainly
depending
upon
the
delivery
system
that's
selected
and
the
design
of
the
experiments,
ultimately,
I
think
consideration
should
be
given
to
make
sure
that
concentrations
are
measured
at
times
at
which
you
would
expect
high
levels
of
atrazine,

but
also
the
lower
levels
that
might
be
resulting
in
these
treatments.

DR.
ROBERTS:
Any
other
comments
or
points
to
add
on
1
A
and
1
B?
We
also
had
a
1
C.

DR.
BRADBURY:
We
have
a
1
C
and
a
1
D.

DR.
ROBERTS:
I
don't
have
a
D
on
mine.

DR.
STEEGER:
Actually,
it's
pretty
much
like
1
C.

1
C
is,
Please
comment
on
the
availability,
as
of
February
28th,

2003,
of
additional
relevant
studies
in
the
open
literature
that
were
not
addressed
in
the
white
paper.

And
1
D
is,
Since
February
28th,
2003,
is
the
panel
aware
of
any
studies
that
would
be
relevant?

DR.
ROBERTS:
We
have
both
of
those.
The
second
one,
at
least
on
mine,
was
not
separately
marked
as
D.

Let
me
then
go
back
to
Dr.
Kelley.

DR.
KELLEY:
Yes.
So
the
review
of
the
panel,
while
very
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
124
thorough
and
complete,
was
narrow
in
its
focus.
So
the
question
is
what
kind
of
guidance
would
the
EPA
like
from
us
about
what
we
consider
to
be
the
relevant
literature.

Let
me
tell
you
what
I
have
in
mind.
So
omitted
from
this
were
several
studies
that
I
found
in
the
open
literature
on
effects
of
atrazine,
usually
at
very
high
doses
on
early
mortality,
say,
in
xenopus,
tadpoles
and
some
more
recent
effects
on
slightly
later
animals.
So
those
were
omitted.

The
EPA
analysis
raised
issues
of
consistency
in
the
data
and
consistency
of
things
like
the
stages
of
sexual
differentiation.
Yet,

of
course,
there
were
none
of
the
background
papers
on
sexual
differentiation
in
that
panel.
So
those
are
two
areas.

And
the
third
is
of
course
the
issue
that
the
panel
has
raised
that
are
relevant
to
things
like
the
mechanism
of
action
of
atrazine
if
it,
in
fact,
has
an
effect
that
might
be
addressed
productively
by
references
that
included
work
on
other
than
amphibians.

So
I'm
in
the
midst
of
collecting
a
series
of
papers
of
that
sort.

But
what
I
would
like
to
know
­­
but
I
have
to
say
I
didn't
find
a
paper
that
you
guys
had
left
out
on
effect
on
gonadal
development
in
amphibia
due
to
atrazine.
I
haven't
found
such
a
thing
yet.

So
I
do
believe
that
in
terms
of
the
focus
of
review
that
you
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
125
didn't
omit
any
obvious
and
glaring
paper.
But
in
view
of
the
panels's
desire
to
provide
scientific
guidance
to
the
process
going
forward,
I
think
that
there
are
probably
a
number
of
studies
that
we
would
like
to
bring
to
your
attention.

DR.
ROBERTS:
Other
panel
members?

Dr.
Matsumura.

DR.
MATSUMURA:
I
would
like
to
ask
the
panel
to
consider
to
add
the
paper
by
Bevan
appeared
in
the
EHS
regarding
action
of
alkylphenols.
Of
course
it
is
not
the
atrazine,
but
this
particular
paper
really
shows
what
the
estrogenic
effect
of
the
alkylphenol
really
look
like.

It
is
summarizing
their
work
on
some
of
the
early
stages
of
development,
and
the
curvature
that
develop
in
the
tadpoles
are
very
clear
on
the
melanocyte
site
formation.
You
can
see
what
real
estrogenic
compounds
could
look
like.

This
is
the
one
endpoint
studied
in
the
xenopus.
And
it
is
a
very
clear
cut
effect
that
you
expect.

And
knowing
that
alkylphenols,
particularly
in
nonylphenol
can
be
found
in
many
detergents,
including
the
triatonix
100
and
all
those
agents
that
you
use
in
the
lab
washing
your
cages
as
well
as
in
the
formulation
in
some
of
the
pesticides.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
126
We
should
really
show
that
this
is
a
kind
of
thing,
if
it
is
a
direct
effect
on
the
estrogen
receptor,
this
is
what
you
expect.
So
I
would
like
to
suggest
that.

I
also
would
like
to
suggest
that
the
Christin's
papers
that
Dr.

Green
mentioned
too
and,
yes,
we
are
looking
very
narrow
way
and
there
are
other
effects
of
the
atrazine
on
the
immuno
competency.
So
that
I
would
like
to
suggest.

DR.
ROBERTS:
Dr.
Isom.

DR.
ISOM:
I
would
just
like
to
point
out
for
the
record
that
the
Hayes
paper
in
EHP
came
out
in
April
2003.
It
has
been
referred
to
several
times
throughout
the
proceedings.
And
that
is
actually
a
detailed
publication
of
the
nature
paper
that
appeared
this
past
year.

DR.
ROBERTS:
Dr.
Steeger.

DR.
STEEGER:
The
requirement
for
the
review
is
that
studies
had
to
be
submitted
by
February
28th.
You
are
correct
that
the
published
hard
copy
of
that
report
came
out
in
EHP
in
April.

However,
the
online
version
was
indeed
published
or
available
in
October.
And
that
was
the
version
that
was
reviewed
for
the
white
paper.

DR.
ROBERTS:
I
think
Dr.
Isom's
response
is
of
1
D.

DR.
STEEGER:
That
April
publication
is
identical
to
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
127
October
one.

DR.
KELLEY:
Could
I
ask
my
question
again
of
the
EPA?

What
sorts
of
guidance
would
you
like
to
have?
Clearly,
we
can
give
guidance
going
forward.
I
still
see
no
­­
within
the
very
narrow
focus
of
what
you
reviewed,
I
don't
think
you
missed
anything,
but
we
have
compiled
a
bibliography
that
we
thought
would
be
helpful.

Would
you
like
that?
Or
do
you
want
to
sift
through
it?

DR.
BRADBURY:
This
is
sort
of
an
awkward
question
to
answer
because
it
presumes
the
path
the
panel
may
or
may
not
go
down.

Hypothetical.
If
the
panel
is
thinking
about
a
pathway
about
gathering
additional
information,
my
sense
is
that
some
of
the
information
you
are
talking
about
would
be
very
instructive
in
thinking
about
the
design
and
nature
and
aspects
of
other
information
that
would
provide
greater
insight
either
to
do
a
risk
assessment
today
or
insights
into
what
kind
of
information
and
pathways
to
gather
new
information
would
be
helpful.

So
my
feeling
is
that,
yes,
it
would
be
very
helpful
is
more
in
the
context
of
the
proceedings
­­

DR.
KELLEY:
But
what
about
my
question
about
the
omission
of
acute
toxic
effects
of
atrazine
on
very
early
tadpoles
from
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
128
bibliography?

DR.
BRADBURY:
I'll
turn
it
over
to
Tom
some,
but
in
looking
at
gonadal
developmental
effects
and
I
think
a
sense
of
what
those
exposure
concentrations
were
sort
of
kept
us
focused
in
that
context.

Putting
it
all
into
the
context
of
full
dose
response
and
full
effects
endpoints,
that
could
be
helpful
in
the
white
paper
now
because
it
provides
some
founding
in
terms
of
effects
of
exposures.

DR.
STEEGER:
From
my
perspective
as
a
scientist,
I
would
very
much
appreciate
to
have
access
to
that
information
to
improve
our
understanding
and
representation
of
the
effects
of
atrazine.

DR.
BRADBURY:
In
the
context
of
getting
through
­­

DR.
ROBERTS:
Right.

DR.
KELLEY:
I
bring
this
up
because,
of
course,
mortality
was
an
endpoint
that
was
studied
in
all
these
studies.

And
even
though
the
doses
in
these
acute
studies
were
much
higher
and
the
mortality
was
much
faster,
still,
in
all,
that's
the
compound
that
you
are
interested
in
and
the
species
that
you
are
interested
in.

So
their
exclusion
seems
to
me
inappropriate
even
if
they
are
not
directly
relevant
to
gonadal
development
because
they
die
before
they
had
any
gonad.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
129
DR.
ROBERTS:
Perhaps
the
panel
could
recommend
that
they
be
included
if
for
no
other
reason
to
help
provide
a
perspective
for
the
doses
at
which
developmental
effects
may
or
may
not
occur.

For
suggestions
on
other
papers,
I
think
we
should
clearly
identify
that
if
the
agency
is
interested
in
studies
that
do
not
relate
specifically
to
gonadal
development
but
may
in
fact
inform
­­
by
analogy
help
understand
whether
or
not
a
potential
phenomenon
occurred,
here
are
some
examples
that
we
can
point
out
so
that
it's
very
clear
that
we
don't
necessarily
consider
them
omissions
for
the
focused
purpose
of
this
paper,
but
they
may
be
useful
for
the
ecological
risk
assessment.

DR.
KELLEY:
Let
me
tell
you
how
I
would
like
to
proceed.

What
I
would
like
to
do
is
to
complete
my
list
of
additional
sources
of
information
that
would
be
helpful
categorizing
the
papers
as
we
have
just
discussed
under
different
categories,
studies
and
other
species,

acute
effects
on
toxicity,
studies
that
might
be
useful
if
a
risk
assessment
goes
forward
and
why
they
might
be
useful.

And
I
will
compile
that
list
over
lunch,
and
I
will
get
it
printed
up
and
I
will
distribute
it,
and
then
I
will
allow
­­
of
course
I
will
allow
­­
the
panel
will
make
me
make
sure
that
we
haven't
missed
studies
that
individual
panel
members
have
discovered.
And
then
at
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
130
that
point
I
think
I
would
be
able
to
sign
off
on
both
1
C
and
1
D.

At
this
point
I
would
just
like
to
revisit
them
afterwards.

DR.
ROBERTS:
Thank
you.
Other
comments?

We're
coming
right
up
on
12
o'clock.
I'm
suspecting
that
­­
I'm
looking
at
Dr.
Richards
as
the
lead
discussant
on
Number
2.
I
suspect
we're
probably
going
to
want
a
little
time
to
discuss
responses
for
Number
2.

So
let's
go
ahead
and
break
now
for
lunch.
And
we
can
come
back
replenished
and
restored
at,
say,
1
o'clock
and
take
up
Question
Number
2.

Let's
adjourn
now,
and
we'll
see
everyone
at
1
o'clock.

(
Thereupon,
a
luncheon
recess
was
taken.)

DR.
ROBERTS:
Before
we
take
up
Question
Number
2,
our
designated
federal
official,
Mr.
Paul
Lewis,
has
an
announcement.

MR.
LEWIS:
Thank
you,
Dr.
Roberts.

The
panel
were
distributed
additional
comment
from
Syngenta
in
reference
to
clarifying
some
of
their
remarks
at
yesterday's
meeting.
It
is
available
to
the
panel
and
will
also
be
available
in
the
public
docket
for
public
review
and
inspection.
Thank
you.

DR.
ROBERTS:
Thank
you,
Paul.

Dr.
Bradbury,
let's
go
ahead
and
proceed
to
Question
Number
2.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
131
DR.
STEEGER:
In
its
evaluation
of
the
existing
field
studies,

the
agency
has
concluded
that
these
investigations
are
of
limited
value.

The
reasons
include,
one,
the
high
variability
in
environmental
conditions
and
uncertainties
in
the
preexisting
status
and
condition
of
field
collected
animals.

Two,
the
spatial
and
temporal
aspects
of
atrazine
exposure,
i.
e.,

spatial
and
temporal
variability
over
the
course
of
the
studies
and
the
extent
to
which
such
aspects
of
atrazine
exposure
were
empirically
measured
or
otherwise
accounted
for.

And
three,
the
possible
cooccurence
of
additional
chemicals
and/
or
non
chemical
stressors.

Question
2
A
is,
to
the
extent
that
the
field
studies
appear
to
indicate
that
atrazine
may
not
adversely
affect
development,
please
comment
on
EPA's
conclusion
that
the
body
of
data
from
the
field
studies
does
not
provide
the
means
to
ascertain
whether
the
lack
of
a
relationship
between
atrazine
exposure
and
developmental
effects
is
due
to
the
absence
of
a
causal
relationship
or
limitation
in
study
methodologies.

DR.
ROBERTS:
Dr.
Richards
is
our
lead
discussant
on
this
question.
I
will
ask
Dr.
Richards
to
begin
our
response.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
132
DR.
RICHARDS:
Thank
you.
I'm
going
to
make
some
general
comments
on
a
number
of
different
aspects
of
this
question.
I
think
I
will
tend
to
treat
A
and
B
together,
although,
will
certainly
address
both
parts
of
that
question.

And
I
invite
the
second
and
third
readers
on
this
question
to
jump
in
at
any
point.

In
essence,
the
field
studies
that
have
been
presented
in
the
literature
and
the
presentations
over
the
last
couple
days
are
very
good.

In
one
respect,
they
provide
a
great
deal
of
information.
I
think
one
of
the
most
important
things
that
they
seem
to
indicate
is
that
the
abnormalities,
gonadal
abnormalities
that
appeared
in
some
cases
in
laboratory
experiments
are
seen
in
the
field.

These
abnormalities
sometimes
are
seen
in
very
different
geographic
regions
to
some
degree
among
different
species.

And
from
that
standpoint,
I
think
that's
critical,
because
I
think
all
these
questions
ultimately
relate
to
an
ecological
role
and
a
population
role.
And
without
looking
at
sort
of
natural
field
type
of
situations,
we
can't
really
see
whether
some
of
the
things
that
we're
looking
at
are
relevant.

Now,
however,
the
studies
do
not
seem
to
provide
substance
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
133
the
nature
of
why
these
gonadal
abnormalities
or
other
endpoints
are
related
to
atrazine.

And
that's
where
it
breaks
down
to
problems
in
experimental
design.
Often
times,
the
experiments
were
never
really
designed
to
ask
such
questions.
Typically,
they
have
very
poor
statistical
power
to
address
such
of
a
question.

Most
of
them
are
essentially
descriptional
studies.

Descriptional
studies
or
observational
studies
can
be
very
good.
They
can
be
powerful.
They
can
give
you
relationships.
But
they
require
that
a
number
of
parameters
be
accounted
for
in
order
to
derive
any
use
from
them.

In
the
studies
that
have
been
set
up
to
use
a
control
or
a
reference
situation,
those
control
and
reference
situations
have
usually
not
been
adequately
described
or
they
have
some
fundamental
flaws.
And
the
sample
sizes
used,
the
end
size
used
is
typically
totally
inadequate
given
the
variability
and
some
of
the
measures
that
we're
looking
at.

The
studies
that
have
been
descriptional
in
attempting
to
use
more
of
aggression
based
approach
have
not
adequately
dealt
with
the
problems
of
scale.

The
studies
we
have
looked
at
have
included
scales
including
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
134
looking
at
continental
scales,
looking
at
the
size
of
a
county,

comparing
a
few
ponds
within
a
small
area.

All
of
those
have
very
significant
things
that
have
to
be
accounted
for
in
order
to
account
for
the
variance
associated
with
individual
assessment
endpoints.

And
I
will
just
begin
to
list
a
few
of
those
things.
But
the
Number
1
thing
that
we're
dealing
with
here
is
aquatic
organisms,

organisms
that
have
a
very
significant
part,
at
least
one
part
of
their
last
cycle
tied
to
the
aquatic
medium.

In
order
to
do
a
descriptional
study
or
to
set
up
a
good
field
experiment,
you
have
to
adequately
account
for
the
movement
of
water
in
and
out
of
those
systems.

A
small
ditch
is
very
different
than
a
river,
is
very
different
than
a
wetland.
It's
very
different
than
a
backwater
on
a
river.

So
it
comes
back
to
both
describing
the
environment
that
the
animal
lives
in,
but
very
much
describing
in
a
general
way
or
a
specific
way
the
exposure
that
these
animals
actually
encounter
in
the
field.

And
for
small
experimental
studies,
that
needs
to
be
accounted
for
and
in
much
detail.
But
on
the
larger,
more
descriptive
studies,

that
can
be
accounted
for
in
a
much
better
way
than
has
even
been
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
135
attempted
in
any
of
these
studies.

There
are
many
tools
available
that
allow
us
to
look
at
even
relatively
large
geographic
scale
studies
that
can
quantify
and
allow
you
to
partition
some
of
that
variance
in
a
stratified
experimental
design.

And
basically,
that's
relating
to
how
does
surface
water
interact
with
the
bodies
of
water
that
these
experiments
are
being
conducted
in
or
observations
are
being
made,
what's
a
relative
contribution
of
groundwater,
what
are
flow
events
during
a
period
of
exposure,
what
are
the
flow
events
that
actually
occur
amongst
seasons,
in
cases
where
experiments
have
been
run,
have
there
been
flooding
events
that
may
influence
exposure
or
may
influence
the
movement
of
organisms
in
or
out
of
those
experimental
situations.

And
also
related
to
hydrology
is
­­
to
have
an
adequate
depiction
of
hydrology,
you
have
to
know
what
the
watershed
is.
You
have
to
know
what
is
upstream
of
the
event
that
you
are
observing.

And
there
are
some
fairly
easy,
relatively
powerful
ways
to
describe
watersheds
both
in
a
strictly
hydrologic
sense,
but
also
in
terms
of
what
the
watershed
is
composed
of.

Things
like
soil
characteristics
and
hydrology
connectivity
to
other
water
bodies,
these
are
all
things
that
are
fundamental
both
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
136
the
biology
and
movement
of
these
creatures,
as
well
as
the
chemical
and
physical
environments
that
the
animals
live
in.

Basically,
some
of
the
studies
have
provided
that
type
of
information
perhaps
in
an
appendix
or
listing
of
things
they
may
or
may
not
have
measured.
Some
of
them
did
not
provide
any
of
that
information.

But
essentially,
none
of
the
studies
attempted
to
use
that
as
a
co­
correlate
in
any
way
to
try
to
partition
variance
amongst
characters
that
again
provide
variation
in
these
populations
that
have
been
observed.

So
in
essence,
I
think
what
we're
looking
at
is
that
statistical
designs
have
not
been
sufficient
to
actually
address
questions
specifically
related
to
atrazine.

And
I
think
that
that
falls
into
the
B
part
of
this
question,
also,

is
that
there
are
co­
stressors,
potential
co­
stressors.
Many
of
those
we
can
guess,
many
of
those
that
we
could
potentially
stratify
out
in
an
appropriate
statistical
design
or
to
partition
into
in
the
analysis,

but
that
takes
a
great
deal
of
forethought
in
terms
of
development
of
a
design.

In
many
cases,
had
we
seen
as
much
attention
given
to
sound
experimental
design
as
we
have
at
least
to
date
with
some
of
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
137
laboratory
measures,
I
think
we
could
have
been
a
little
bit
further
forward.

I
think
I'll
ask
my
co­
readers
to
jump
in
after
that.

DR.
ROBERTS:
Dr.
Gibbs,
you
are
an
associate
discussant.
Is
there
anything
you
want
to
add?
Or
do
you
want
me
to
take
Dr.
Skelly
first?

DR.
SKELLY:
First
of
all,
I
would
like
to
point
out
that
that's
a
single
sentence
up
there,
and
so
it
took
me
a
while
to
figure
out
what
it
means,
but
I
think
I
got
a
handle
on
it.

I
guess
the
end
of
the
sentence
is
the
most
critical
part.
So
do
we
think
that
the
absence
of
causal
relationship
or
limitation
and
study
methodologies
is
the
reason
why
the
EPA's
conclusion
that
field
studies
don't
provide
the
means
to
ascertain
whether
there
is
something
going
on
here.

I
guess
in
my
opinion,
I'm
going
to
agree
with
Dr.
Richards
and
say
that
the
limitation
and
study
methodologies
is
probably
most
important
here,
and
comment
that
the
absence
of
causal
relationship
has
never
slowed
down
any
ecologist
I
have
ever
met.

We
often
like
to
pretend
that
we
know
why
things
happen
when
we're
working
in
the
field,
but
that
shouldn't
stop
us
from
moving
forward
and
trying
to
understand
what
the
associations
are.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
138
Because
after
all,
everything
that
we're
doing
is
keeping
our
fingers
crossed
that
we
get
the
mechanism
right.
Even
in
the
finest
laboratory,
using
the
best
laboratory
practices,
we
can
never
know
for
sure
that
we
have
the
mechanism
right.

So
specifically,
I
think
the
limitations
and
study
methodology
that
were
most
important
concur
largely
with
what
Dr.
Richards
said.

I
think
the
lack
of
power
is
an
issue,
but
it
is
very
easy
to
be
critical
of
that
in
hindsight.

These
studies
that
have
been
done
can
now
be
used
moving
forward
to
come
up
with
better
ideas
with
what
needs
to
happen.

Having
said
that,
the
scale
of
the
field
studies
and
I
say
this
from
the
standpoint
of
having
quite
a
bit
of
experience
in
field
sampling
amphibian
population,
these
were
not
large
studies
in
terms
of
the
number
of
sites
that
were
examined.

There
are
many
more
ambitious
studies
and
I
think
we
can
look
to
a
lesson
from
how
the
UV
light
on
amphibians
controversy
is
being
resolved
now.

It
has
taken
very
large
scale
studies
looking
at
hundreds
of
wetlands
over
very
large
areas
dealing
with
many
of
the
same
issues
that
we
have
been
talking
about
here
today.

Some
of
the
people
in
this
room
have
been
involved
in
this
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
139
work,
so
they
know
what
it
takes
to
pull
inference
out
of
these
sorts
of
big
observational
studies.

That
leads
me
to
the
second
point,
which
is
the
issue
of
site
selection.
I
think
it
is
hard
to
go
out
and
use
a
map
ahead
of
time
or
just
go
out
and
look
at
a
place
and
say,
there
is
a
cornfield
here
or
there
isn't
a
cornfield
here,
to
say
whether
this
is
the
right
site
to
be
an
exposure
site
or
a
control
site.

And
some
of
the
studies
that
were
not
considered
here
and
are
not
necessarily
particularly
relevant
to
this
issue
but
where
people
have
taken
an
alternative
approach,
they
might
take
a
year
or
two
years
just
to
pick
their
sites.

I'm
going
to
come
back
to
this
point
when
we
talk
about
some
of
the
questions
later
on.
I'll
leave
that
for
now.

The
final
point,
the
limitation
in
the
study
methodologies
may
be
this
is
not
a
limitation
study
methodology,
but
a
significant
limitation
in
the
studies
that
I
saw
was
to
me
looking
at
what
the
EPA
is
going
for
here,
the
assessment
endpoints
that
are
talked
about
are
fertility
reproduction
recruitment.

Ultimately,
what
we
are
trying
to
think
about
­­
something
about
viability
of
populations
and
those
people
in
this
room
who
know
about
measuring
population
viability
know
what
a
giant
task
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
140
that
is.

Even
if
we
go
back
to
the
assessment
endpoints
here
the
motivation
for
doing
field
studies
should
be
to
get
at
those
things.
I
don't
think
most
of
the
field
studies
that
have
been
done
to
date
address
that
issue.
In
moving
forward
that's
something
that
should
be
tackled
very
concretely.

In
response
to
B,
to
the
extent
that
we
do
see
something
going
on
here,
is
it
true
that
these
studies
don't
provide
sufficient
information
to
resolve
the
potential
role
of
additional
cooccuring
stressors?

I
guess
that
is
always,
always
going
to
be
true
to
some
extent
with
the
field
study,
but
for
reasons
that
we'll
probably
talk
about
later
on
I
still
think
it
is
extremely
important
for
much
of
this
work
to
be
done
in
a
context,
in
a
field
context.

To
that
end,
as
I
have
harped
on
before
and
I
will
say
now,

virtually
all
the
field
work
that
is
talked
about
here
and
if
we
exclude
the
mesacosm
experiment
all
of
it
has
been
observational.

It
is
possible
even
in
this
difficult
kind
of
context
of
working
with
pesticides
to
do
field
experiments.
Especially
­­
we've
got
a
registered
pesticide
here,
we
can
use
it
in
the
world
out
there
and
see
what
it
does
at
population
scales.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
141
That
can
help
us
significantly
get
at
the
issue
of
co­
occurring
factors
because
we
can
choose
sites
that
are
sort
of
tabularasas
to
begin
with
and
add
our
stressor
that
we're
interested
in.

I'm
going
to
echo
­­
in
finishing,
I
will
echo
what
Dr.
Richards
said.
That
is
that,
I
think,
the
critical
thing
here
that
we
can't
ignore
is
we
can
talk
about
limitations
of
any
study
we
want
all
day
long,

multiple
groups
have
gone
out
and
looked
for
gonadal
abnormalities
in
nature
and
they
have
found
them
and
these
match
the
morphological
characters
people
are
seeing
in
the
lab.

At
this
point,
I
think
it
is
fair
to
say
we
don't
know
why,
but
the
fact
that
those
abnormalities
exist
I
think
is
significant.
I
think
if
a
bunch
of
groups
had
gone
out
there
and
had
not
found
them
we
would
be
thinking
about
this
differently.
I
will
end
there
and
pass
to
Dr.

Gibbs.

DR.
ROBERTS:
Dr.
Gibbs.

DR.
GIBBS:
We
have
spoken
extensively
as
a
group,
so
I
don't
want
to
reiterate
too
many
points.
I
do
personally
agree
with
the
statement
that
essentially
study
designs
have
precluded
really
determining
whether
there
is
or
there
is
not
an
association
between
the
occurrence
of
gonadal
abnormalities
in
the
field
populations
of
amphibians
in
the
presence
or
absence
of
levels
of
atrazine.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
142
We
don't
want
to
imply
that
observational
studies
cannot
address
these
issues.
I
think
they
are
actually
very
important
for
the
reason
that
Dr.
Skelly
just
mentioned.

The
risk
assessment
goal
here
really
does
focus
on
population
viability.
One
needs
to
go
to
the
field
essentially
to
measure
many
of
these
parameters.

I
would
­­
perhaps
this
is
getting
away
from
the
question,
but
there
are
other
modes
of
studying
and
tackling
these
problems
which
essentially
fall
in
the
realm
of
field
experimentation.

We
have
been
discussing
some
other
situations
that
occur
in
the
field
with
temporary
breeding
­­
temporary
pool
breeding
amphibians
such
as
the
amphibians
that
use
vernal
pools.

Vernal
pools
occur
in
isolated,
small,
populations
in
great
numbers
in
homogenous
landscapes
that
are
really
quite
amenable
to
experimental
manipulations
that
could
get
at
both
­­
would
yield
information
both
from
histological
perspective
but
also
from
demographic
perspective
quite
relevant
to
many
of
the
issues
that
the
risk
managers
­­
risk
assessment
folks
have
to
deal
with.
I
think
that's
worth
exploring
at
some
point.
But
to
get
back
to
the
point
at
hand,
I
agree
broadly
with
many
of
the
statements
that
Dr.
Richards
raised,
issues
with
inappropriate
or
not
outlining
sampling
frames,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
143
low
levels
of
replication
in
terms
of
using
wetlands
as
actually
replicates
and
the
amphibians
therein.

And
really
a
difficulty
in
establishing
whether
there
is
or
is
not
causal
relationship
here
between
what
are
some
fairly
high
levels
of
abnormalities
being
seen
in
the
field,
but
whether
or
not
there
is
a
relationship
with
atrazine.
In
my
mind
that
can't
be
attributed
based
on
the
designs
of
many
of
these
studies.

DR.
ROBERTS:
Thank
you,
Dr.
Gibbs.
Let
me
ask
other
members
of
the
panel
if
they
agree
or
disagree?

DR.
MATSUMURA:
I
agree
with
Dr.
Skelly
and
Dr.
Richards.

I
certainly
emphasize
the
importance
of
the
field
studies.

I
was
trying
to
think
what
the
precedence
I
could
think
of,
and,

of
course,
like
Gillette's
work
on
the
alligators
and
how
long
did
it
take,
and
what
started
that
whole
project,
and
like
finding
a
really
contaminated
site
like
Lake
Apopka
was
very
important
for
that
study.

And
comes
from
field
studies
like
eggshell
thinning,
it
took
a
long,

long
time
to
say
it
was
DDT.

But
again,
field
observation
that
all
those
eggs
are
crashing
and
that's
a
problem
and
the
bald
eagles
not
just
thinning
but
by
the
weight
of
their
mothers
just
sitting
on
those
eggs
are
crashed.
Those
observations
really
helped.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
144
And
finally,
I
would
like
to
make
sure
that
one
aspect
to
emphasize,
that
means
analysis
of
burden
of
other
contaminants.
I
was
reading
this
Reeder's
paper
reporting
this
intersex
confined
to
the
­­
correlated
to
atrazine.
But
at
the
same
time
they
had
the
PCDD
and
PCDF.

And
so
you
have
to
leave
some
questions
always.
Knowing
what
the
other
stressors
and
other
chemicals
are
there,
that
could
really
help.

DR.
ROBERTS:
Dr.
Green.

DR.
GREEN:
I
agree
with
your
comments.
For
my
own
edification,
I
would
like
to
know
from
your
experience
and
perspective
what
denotes
a
healthy
amphibian
population
observationally
in
a
field
study?
What
implies
that
the
population
is
sickly
and
on
the
decline,
absolute
head
counts
and
how
long
does
it
take
to
get
that
kind
of
information?

DR.
GIBBS:
Insofar
as
you
linked
health
to
decline,
you
are
talking
about
a
temporal
phenomenon.
That
may
indicate
the
need
for
repeated
sampling
to
actually
detect
a
change
in
population
size,

population
structure.

With
these
particular
organisms,
some
of
them
have
­­
with
frogs,
incredibly
short
life
spans
and
then
generation
times
and
hence
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
145
population
turnovers.

So
one
can
fairly
quickly,
unlike
working
with
many
other
vertebrates,
see
if
there
actually
are
population
declines
or
increases
occurring
because
you
need
much
less
time
with
generation
times
and
sometimes
one
to
two
to
three
years
to
see
these
kinds
of
effects.

Sheer
abundance
is
one
good
measurement.
Dr.
Skelly
may
have
other
ideas.

DR.
SKELLY:
I
would
agree
with
what
Dr.
Gibbs
just
said.
I
would
just
add
that
we
have
settled
on
a
group
here
that
has
notoriously
volatile
population
dynamics
and
going
out
in
a
given
year
and
seeing
recruitment
failure
for
some
species
is
actually
the
norm.

Having
total
cohort
failure
is
­­
if
we
were
talking
about
grizzly
bears
in
Yellowstone
people
might
get
quite
upset
about
that,
but
most
of
the
wood
frog
populations
I
survey
every
year
in
most
years
crash
out.

This
year
they
are
floating
across
the
roadways
and
they
are
doing
quite
well
with
all
the
rain.
In
most
years
they
dry
up
with
their
pond.
It
definitely
is
something
you
have
to
look
at
over
time.

We're
fortunate
many
of
the
species
that
we
would
be
interested
in
in
terms
of
native
amphibians
do
have
short
life
spans.
It
does
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
146
have
to
be
looked
at
over
time
if
we're
interested
in
any
sort
of
understanding
of
population
robustness.

DR.
ROBERTS:
There's
that
word
again.
Dr.
Delorme,
then
Dr.
Richards.

DR.
DELORME:
Just
following
up
on
Dr.
Green's
question.

Could
you
give
us
any
indication
of
some
measurement
endpoints
or
what
variables
you
might
be
looking
for
to
look
for
effects
in
populations
if
you
are
out
there
with
your
rubber
boots?

DR.
SKELLY:
Population
­­

DR.
GREEN:
I
think
that
will
be
addressed
in
question
five.

We
have
some
estrogenic
biomarkers
we
put
together
that
­­

DR.
DELORME:
I
was
asking
from
a
populations
perspective,

like
what
population
variables
might
you
go
out
and
measure.

DR.
ROBERTS:
Were
you
planning
on
bringing
it
up
later
in
the
context
of
five
or
do
you
want
­­

DR.
SKELLY:
I
actually
think
we
will
address
that
later.

DR.
ROBERTS:
Dr.
Richards.

DR.
RICHARDS:
Another
context
sort
of
thing
is
amphibian
population
is
often
very
variable
that
is
the
way
they
are.
That
is
the
nature
of
the
waterbodies
they
live
in,
to
get
back
to
the
hydrology
of
these
situations.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
147
One
of
the
things
that
we
know
about
amphibians
is
that
they
interact
over
a
relatively
large
geographic
scale.
You
just
can't
look
at
a
pond
the
size
of
the
space
between
our
tables
here,
you
have
to
be
cognizant
of
the
larger
network
of
waterbodies
that
they
are
connected
to.

Because
when
one
local
population
disappears
they
are
colonized
by
another
population
that
may
be
half
a
kilometer
away.

Looking
at
a
series
of
small
ditches
that
are
by
nature
connected
is
very
important
as
opposed
to
an
isolated
pool.

Are
you
looking
at
a
species
like
the
South
African
animal
that
lives
entirely,
its
whole
life
cycle
in
water
versus
something
like
the
rana
species
that
are
out
moving
around?

These
are
very
critical
questions.
Both
designing
the
experiment
and
asking
questions
about
is
a
population
healthy
in
part
it
is
related
to
a
little
bit
larger
scale
than
we
frequently
focus
on.

DR.
ROBERTS:
Dr.
Skelly
has
a
follow
up,
then
Dr.
LeBlanc.

DR.
SKELLY:
I
guess
as
a
follow
up
to
that,
we're
painting
a
picture
that
can
seem
to
get
more
and
more
difficult.

Actually,
the
genetic
techniques
that
are
being
developed
now
are
getting
us
to
the
point
were
we
can
actually
distinguish
the
boundaries
of
populations.
Up
until
recently,
people
just
kind
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
148
crossed
their
fingers
and
assumed
a
pond
was
a
population.
I
think
we're
moving
well
beyond
that.
So
we
are
actually
starting
to
understand
in
a
much
more
detailed
way
how
much
movement
there
is
and
over
what
scales
we
should
be
assessing
population
level
responses
in
order
to
say
whether
a
population
is
going
to
remain
there
or
is
it
likely
to
go
extinct.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
This
question
is
posed
to
the
discussants.

Based
on
what
I
just
heard,
my
question
is
would
it
be
­­
due
to
the
volatile
nature
of
the
population
dynamics
of
frogs
would
it
be
difficult
to
discern
population
level
effects
of
environmental
contaminants?

DR.
RICHARDS:
Yes
and
no.
I
think
sometimes
with
some
of
these
measures
dropping
down
out
of
the
sky
on
one
occasion
and
listening
for
how
many
frog
calls
there
are
you
are
not
going
get
a
real
good
answer.
I
think
because
of
the
fluctuating
dynamics
of
these
populations
in
their
actual
environment
some
manner
of
looking
at
more
of
a
prolonged
view
of
the
population
is
needed.

There
are
some
things
as
genetics
that
give
us
a
backward
view
of
what
is
going
on
there,
and
that's
great,
but
there
is
much
room
for
innovation
on
this.
The
book
is
not
completely
written.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
149
DR.
ROBERTS:
I
think
Dr.
Skelly
wants
to
respond
as
well.

DR.
SKELLY:
I
guess,
I
would
say
in
order
to
make
the
picture
a
little
less
bleak
here,
there
are
very
volatile
dynamics.
But
if
you
follow
a
set
of
populations
in
a
defined
area
over
a
long
period
of
time,
you
see
this
volatility
from
year
to
year.

You
also
see
very
striking
patterns.
I
will
give
you
an
example
that
is
directly
relevant
to
the
sort
of
thing
we're
considering
here.

For
many,
many
years
people
have
been
doing
experiments
in
mesacosms
in
the
laboratory
that
were
showing
that
amphibians
undergo
interspecific
competition
and
that
interspecific
competition
could
even
be
possible
for
population
extinction.

That
was
the
way
­­
that
was
the
inference
that
was
coming
out
of
this
work.
There
is
excellent
laboratory
and
mesacosm
based
evidence
that
there
was
competition
going
on,
and
if
we
change
the
words,
that's
exactly
what
we
have
been
talking
about
here,
the
same
sort
of
thing
where
people
were
doing
these
very
focused
studies,
they
were
getting
mechanistic,
they
were
looking
for
interference
compounds,
they
were
looking
for
resource
based
competition,
and
so
forth.

We
have
been
able
to
show
more
recently
that
from
everything
we
can
tell
even
though
this
is
an
actor
in
these
experimental
context,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
150
it
does
not
seem
to
be
important
for
the
kinds
of
patterns
that
the
EPA
is
interested
in
for
assessment
endpoints
and
for
their
overall
criteria.

Instead
we
have
been
able
to
discover
starting
with
field
observations,
then
moving
to
field
experiments,
and
then
into
laboratory
experiments,
that
other
factors
such
as
hydroperiod
and
such
as
the
light
level
in
these
ponds
that
influences
temperature
and
food
gradients
is
much
more
important.

That's
been
a
real
lesson
for
me
in
how
we
discover
what
is
really
going
on,
how
to
parse
this
out
from
the
perspective
of
a
field
scientist
so
that
the
lab
scientists
are
working
on
the
right
stuff
so
they
are
not
just
going
down
the
garden
path.

DR.
ROBERTS:
Dr.
Isom.

DR.
ISOM:
Since
atrazine
has
been
applied
to
the
environment
for
over
40
years
in
North
America,
is
there
historical
population
data
that
perhaps
could
be
looked
at
to
give
us
some
answers
to
these
longer
term
population
issues?

DR.
SKELLY:
There
has
been
­­
because
of
the
overall
interest
in
amphibian
declines,
everything
that
existed
and
much
work
over
the
last
12
years
or
so
has
been
done
along
those
lines.

We
probably
for
the
purposes
of
this
panel
can
parasitize
a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
151
much
larger
database
than
you
might
believe.
There
is
much
known
on
that.
The
problem
of
course
is
going
to
be
attributing
cause
to
these
historical
studies
where
the
other
measurements
that
we
are
interested
in
were
not
made.

I
can
give
you
as
an
example,
there
were
surveys
done
in
the
early
part
of
this
century
in
Iowa
that
describe
high
abundances
of
amphibians
over
many,
many
counties.
When
the
resurvey
was
done
in
1980'
s,
abundances
and
species
composition
over
large
areas
had
changed.

Now
that's
over
the
same
time
span
as
the
green
revolution.

Does
it
have
anything
to
do
with
that,
does
it
have
anything
to
do
with
pesticides?
We
have
absolutely
no
idea.

In
general,
the
consensus
seems
to
be
now
that
many,
many
amphibian
populations
are
declining.
So
that's
the
state
of
our
knowledge,
the
state
of
our
sciences
that
many
populations
have
declined.

When
people
go
out
in
their
back
yards
­­
and
I
give
lots
of
lectures
to
the
public
and
the
first
question
I
get
from
the
public,

which
is
an
excellent
one,
is
I
can
be
deafened
sitting
on
the
back
porch
by
the
spring
peepers.
So
are
amphibians
declining?

What
I
say
to
them
is
because
this
is
a
temporal
phenomenon
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
152
maybe
you
would
have
gone
deaf
in
one
week
instead
of
two,
100
years
ago.
We
don't
know
what
it
was
­­
we
don't
have
enough
data
to
be
able
to
say
over
very
broad
scales
what
amphibian
abundances
were
like.

But
Dr.
Gibbs
maybe
can
answer
this
a
little
bit
better
than
I
can.
One
of
the
things
that
has
happened
over
that
same
time
period
is
that
wetlands
have
disappeared
or
been
modified,
often
in
an
agricultural
context,
in
ways
that
change
their
suitability
for
different
amphibian
species
over
the
same
time
span.

There
has
been
talk
about
multiple
co­
occurring
factors.
It
is
there.
I
think
the
most
interesting
historical
data
that
we
could
get
are
in
the
animals.
So
this
issue
of
evolved
resistance
could
be
really
important
in
showing
or
in
evaluating
whether
there
have
been
broad
scale
evolutionary
responses
to
pesticides.

DR.
ROBERTS:
Dr.
Gibbs,
did
you
want
to
respond
also?

DR.
GIBBS:
Just
quickly,
amphibian
monitoring
in
a
systematic
and
large
scale
fashion
only
has
begun
since
1990
or
so
with
various
initiatives
such
as
the
USGS
and
other
groups
have
put
together.
There
really
is
not
a
lot
of
data,
even
that
effort
is
limping
along.

Unfortunately,
we
are
constrained
and
we
will
remain
so
for
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
153
quite
a
few
more
decades
before
we
can
get
something
like
the
US
­­

the
breeding
bird
survey
on
line
for
amphibians,
that's
the
goal.
We
really
don't
have
any
good
solid
baselines
over
large
areas.
And
a
decade
of
data
on
a
few
spots
is
about
as
good
as
we
can
do.

This
underlies
a
lot
of
the
controversies
and
debates
over
amphibian
decline
this
lack
of
monitoring
data.
Unfortunately,
that
is
the
state
of
the
science
on
amphibian
monitoring.

DR.
ROBERTS:
Dr.
Richards.

DR.
RICHARDS:
I
would
like
to
hopefully
dispel
a
little
bit
of
the
gloom
here
that
I
myself
have
been
raising.

At
one
level
amphibians
are
no
different
than
any
other
aquatic
organisms
that
are
dealt
with
by
EPA
and
aquatic
ecologists
for
a
long
time.

They
still
reproduce,
they
still
have
fecundity,
they
still
recruit
to
a
population,
there
is
predation,
there
is
density
and
nondensity
dependent
aspects
to
their
population
sizes,
and
that
makes
them
particularly
amenable
to
experimentation.
That
doesn't
mean
we
can't
experiment
with
these
animals
and
can't
derive
some
creative
ways
of
examining
population
parameters.

They
do
things
just
like
when
you
are
chasing
little
fish
along
the
margins
of
a
stream.
If
you
have
ever
done
that
you
know
you
can
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
154
go
there
one
time
and
they
will
be
there,
next
time
they
won't.

That
doesn't
keep
us
from
having
a
great
body
of
regulatory
approaches
to
dealing
with
toxic
antieffects
on
fish.

DR.
ROBERTS:
Dr.
Delorme,
did
you
want
to
respond
to
that
as
well?
I
have
a
couple
of
other
folks
who
have
questions.

DR.
DELORME:
I
wanted
to
make
a
comment
on
an
observation.
I
can
wait.

DR.
ROBERTS:
Let
me
go
to
Dr.
Denver,
then
Dr.
Heeringa,

and
then
you.

DR.
DENVER:
I
just
wanted
to
comment
on
something
that
Dr.

Skelly
said.
It
has
to
do
with
single
species
experiments
versus
competitive
interactions,
and
at
the
risk
of
making
this
even
more
complicated
I
think
it
is
important
to
realize
the
importance
of
the
competitive
interactions.
Most
of
the
studies
that
we
have
reviewed
have
been
studies
of
single
species.
We
published
a
study
actually
last
year
that
showed
that
exposure
to
PCBs
can
actually
change
the
competitive
interactions
among
species.

These
were
studying
the
northern
leopard
frog
and
also
wood
frogs.
That's
another
layer
of
complexity
that
needs
to
be
considered.

DR.
ROBERTS:
Dr.
Heeringa.

DR.
HEERINGA:
With
much
more
expertise,
Dr.
Denver
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
155
anticipated
my
question.
Mine
is
much
more
direct
and
I
deal
a
little
bit
with
waterfowl
populations.

Is
there
evidence,
either
from
zoological
samples
in
museums,

et
cetera,
or
other
studies,
on
sort
of
either
a
lack
of
increased
hybridization
or
competition
in
these
natural
populations?

Are
there
species
that
are
invading
historically,
turf
for
other
species
that
have
been
a
concern
to
ecologists,
is
that
practice,
it
could
happen
naturally,
but
do
you
see
that
happening
and
are
these
populations
still
relatively
isolated
and
nonhybridized?

DR.
GIBBS:
Certainly,
and
the
bullfrog
is
a
classic
example
of
invasions,
but
very
much
associated
with
the
activities
of
people
and
moving
them
around.

I'm
trying
to
think
of
an
analogous
situation
with
a
hybrid
zone
shifting
or
one
native
species,
a
sympatric
species,
one
moving
into
the
other's
range.
I'm
failing
to
­­

DR.
SKELLY:
The
example
that
Dr.
Hayes
talked
about
that
was
the
first
one
that
I
had
seen
evidence
for.

DR.
ROBERTS:
We've
got
Dr.
Delorme,
then
Dr.
Matsumura.

DR.
DELORME:
I
was
sitting
here
listening
to
Dr.
Skelly
talk
in
response
to
Gary
Isom's
question
and
it
occurred
to
me
that
one
of
the
things
that
EPA
has
asked
us
to
look
at
is
the
assessment
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
156
endpoints
and
they
put
forth
an
assessment
endpoint
of
the
population.

You
guys
can
shoot
me
down
if
you
want
but
some
of
the
measures
that
you
are
talking
about,
species
diversity
and
relative
species
abundance,
perhaps
maybe
an
alternate
assessment
endpoint
is
also
communities,
in
terms
of
their
composition.

It
gets
into
sensitivity
of
different
species
and
whether
it
is
different
or
not.
Just
want
to
know
if
you
guys
think
this
idea
has
any
merit
as
alternate
assessment
endpoint?

DR.
ROBERTS:
You
three
are
on
the
spot
again.
Who
wants
to
take
it?

DR.
RICHARDS:
I'll
touch
a
little
bit
of
it.

There
are
some
studies
out
there
that
have
indicated
over
relatively
large
geographic
areas
that
anuran
communities
can
be
a
fairly
useful
way
of
looking
at
relationship
to
landscape
parameters.

Whether
they'll
be
useful
for
looking
at
toxic
related
questions,
I
don't
think
has
been
delineated.

Certainly,
communities
can
be
manipulated
­­
the
data
from
communities
can
be
manipulated
into
a
series
of
metrics
much
like
we
use
fish
or
invertebrates
or
other
things
in
a
way
to
look
at
patterns
over
relatively
large
geographic
scales.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
157
DR.
ROBERTS:
Dr.
Skelly,
did
you
want
to
add
anything?

DR.
SKELLY:
I
guess
I
would
just
insert
a
note
of
caution
that
it
can
be
­­
if
we
are
trying
to
link
these
studies,
eventually,
back
to
a
mechanism
and
link
them
to
laboratory
studies,
logistically
and
pragmatically,
it
may
be
challenging
enough
to
do
on
a
single
species
level.

I'm
very
interested
in
communities.
I
think
that
they
can
sometimes
show
responses
that
may
give
us
more
information
than
single
species.

But
on
the
idea
that
we
need
to
walk
before
we
run,
I
guess
I
would
say
that
we
should
try
to
get
out
into
the
field
more
working
with
the
single
species
before
we
look
at
multiple
species
simultaneously.

DR.
ROBERTS:
Dr.
Matsumura.
DR.

MATSUMURA:
I'm
not
doing
that,
but
I'm
collaborating
with
the
good
herbatologist
in
the
pit
studies,
on
the
study
on
the
yellow
legged
mountain
frogs,
rana
muscosa.

As
far
as
that
particular
species
is
concerned,
it
is
declining,

everybody
agrees,
at
least,
in
California.
It
has
a
long
larval
life
span
that
it
spends
in
such
a
cold
place
for
three
to
almost
four
years
in
the
water,
and
they
are
disappearing.
158
1
When
we
made
a
survey,
we
found
that
there
are
still
some
2
ponds
and
lakes
teamed
with
the
yellow
legged
frog
but
they
all
on
3
the
eastern
slope.
On
the
western
slope,
there
are
only
about
three
4
sites
that
we
found.

5
There
is
no
question
that
you
can
talk
to
those
rangers.
They
6
say,
well,
I
used
to
know
that
particular
pond
had
the
muscosa,
lots
of
7
them
and
they
are
disappearing
in
the
last
year
10
years.

8
They
are
species
that
you
can
really
count
on
not
just
by
virtue
of
9
long
life
in
that
particular
case.
You
can
really
see
that's
how
they
10
are
disappearing.

11
So,
it
is
not
related
to
atrazine.

12
But
by
studying
those,
we
start
finding
also,
hyla
regilla,

13
pacific
tree
frogs,
are
also
disappearing
from
the
same
western
slope.

14
That's
the
reason
why
we
are
suspecting
air
pollution
and
particularly
15
PCBs.

16
In
the
lowland
area
also
the
red
legged
frogs
also
disappearing,

17
but
they
are
being
displaced
by
bullfrogs.
So
that
is
the
same
18
example
that
you
are
citing.

19
I
get
the
feeling
that
just
looking
at
those
mountain
frogs
20
endangered
species
which
we
cannot
study
well,
that
there
could
be
21
some
combination
of
right
kind
of
species,
very
sensitive
ones
that
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
159
you
could
pick
on
and
try
to
see
whether
those
are
the
ones
who
are
effected
the
most
by
this
kind
of
pollution.

So
some
field
biologists
can
really
look
for,
they
may
find
a
good
combination.
I
just
wanted
to
add.

DR.
ROBERTS:
Thank
you.

I
haven't
heard
much
in
the
way
of
disagreement
in
this.

Let
me
ask
the
panel,
is
there
anything
that
they
disagree
with
that
has
been
stated
or
are
there
any
other
important
points
that
have
been
left
out
of
the
discussion
that
perhaps
should
be
included
in
the
response?

Dr.
Kelley.

DR.
KELLEY:
I
don't
disagree
with
the
conclusion
as
stated
based
on
the
available
data.
I
just
simply
wish
to
make
a
comment
about
the
field
studies
in
South
Africa,
which
is
that
if
the
EPA
decides
that
it
wants
to
adopt
xenopus
as
some
sort
of
model
system
for
studying
a
sex
environmental
effects,
including
atrazine,
but
there
are
others
as
well
for
which
it's
being
the
considered,
inclusion
of
field
studies
in
gathering
field
data
in
xenopus
will
be
extremely
useful.

It
will
enable
you
to
take
the
best
studied
laboratory
anuran
and
correlate
those
results
with
the
field.
I
don't
think
it
is
possible
for
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
160
any
other
animal.
We
have
neglected
in
our
literature
survey
the
data
on
xenopus
which
exists
in
the
South
African
literature.

I
would,
if
that
route
is
taken,
I
think
­­
going
forward
more
carefully
designed
and
better
powered
field
studies
in
xenopus
will
be
useful
in
trying
to
relate
laboratory
results
to
potential
effects
in
the
field.

Now
they
won't
tell
you
anything
about
our
North
American
anurans,
necessarily,
but
they
will
tell
you
something
about
the
limitations
of
the
transform.

DR.
ROBERTS:
Thank
you.
Dr.
Bradbury,
it
looks
like
there
is
strong
endorsement
for
the
value
of
field
studies
and
the
panel
has
a
number
of
comments
about
what
is
required
to
really
get
good
information
from
those
studies.

Are
there
any
follow
up
questions
that
you
have
for
the
panel
on
this
topic?

DR.
BRADBURY:
Just
for
the
record,
we
have
question
2
B.

DR.
ROBERTS:
I
sort
of
sense
that
our
discussion
has
covered
both
2
A
and
2
B.
I
will
defer
to
the
discussants
on
that.

DR.
BRADBURY:
I
feel
comfortable
that
we
got
the
picture
between
the
two
aspects
but
if
there
is
something
left
in
your
laptop.

DR.
ROBERTS:
Dr.
Richards,
is
there
anything
left
in
your
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
161
laptop?

DR.
RICHARDS:
My
fingers
are
fingered
out,
but
I
think
in
my
own
comments
I
was
implying
that
it
was
very
difficult
to
examine
the
effects
of
costressers
given
the
studies
we
have
seen
so
far.

DR.
BRADBURY:
I
think
we're
in
good
shape
on
our
end.

DR.
ROBERTS:
Are
there
any
other
points
that
need
to
be
made
or
anything
we
need
to
resolve
before
we
move
onto
the
next
question?
If
not,
let's
move
on
to
question
three.

DR.
STEEGER:
In
evaluation
of
the
existing
laboratory
based
studies
the
agency
concluded
that
there
was
sufficient
information
to
establish
a
hypothesis
that
atrazine
could
cause
adverse
gonadal
developmental
effects.

However,
due
to
the
different
experimental
designs
and
variability
in
the
nature
and
extent
of
experimental
conditions,
i.
e.,

level
of
excessive
mortality,
delayed
developmental
and
untreated
organisms,
lack
of
response
to
positive
controls,
it
was
not
possible
to
adequately
assess
the
hypothesis
that
atrazine
causes
developmental
effects.

It
was
further
concluded
that
the
current
body
of
information
did
not
prove
the
means
to
characterize
the
nature
of
any
associated
dose
response
relationships.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
162
Please
comment
on
EPA's
determination
that
the
laboratory
studies
provide
a
plausible
basis
for
the
means
to
establish
hypothesis
concerning
the
potential
for
atrazine
to
cause
development
effects.

Also,
please
comment
on
whether
the
overall
body
of
available
data
are
adequate
to
demonstrate
whether
or
not
atrazine
causes
developmental
effects
under
the
conditions
described
in
these
studies.

DR.
ROBERTS:
An
important
question
Dr.
LeBlanc,
what
do
you
think?

DR.
LEBLANC:
I
concur
with
the
EPA,
their
determination
that
laboratory
studies
on
the
affects
of
atrazine
on
anuran
gonadal
development
are
sufficient
to
establish
the
hypothesis
that
atrazine
interferes
with
normal
gonad
development.

Clearly,
the
available
data
are
limited.
However,
the
existing
data
as
related
specifically
to
the
laboratory
investigation,
I
believe,

support
the
hypothesis
of
the
eight
laboratory
studies
that
we
were
provided
with
that
evaluating
effects
of
atrazine
on
gonad
development.

Five
of
these
studies
detected
such
abnormalities
from
three
different
laboratories.
Certainly,
in
my
laboratory,
if
a
student
comes
to
me
with
an
observation,
be
it
induction
of
enzyme,
suppression
of
protein,
or
a
message
RNA
level,
the
first
thing
I
tell
them
is
go
do
it
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
163
again.

They
come
back
and
I
tell
them
to
go
do
it
again
and
perhaps
after
six,
seven,
or
eight
times
I
will
accept
that
data,
which
doesn't
mean
that
perhaps
two
of
those
times
they
weren't
able
to
reproduce
the
effect.

But
if
the
preponderance
of
the
evidence
consistently
shows
the
effect
that
they
initially
observed
and
that
we
were
pursuing,
I
wouldn't
discount
the
data.

There
certainly
were
differences
observed
among
the
various
studies
that
we
examined
as
related
to
the
effects
that
were
observed
and
many
attributes
were
ascribed
that
could
have
contributed
to
these
differences,
including
species
differences,
differences
in
exposure
design,
and
specific
endpoints
that
were
selected
in
the
individual
evaluations.

Significant
data
gaps
exist
in
our
understanding
of
the
affects
of
atrazine
on
anuran
development.
These
include
a
lack
of
understanding
of
the
mechanism
by
which
atrazine
elicits
its
developmental
toxicity,
the
nature
of
the
concentration
response
relationship,
definition
of
susceptible
windows
of
exposure,
and
identification
of
a
threshold
concentration.

However,
it
is
certainly
my
view,
that
the
existing
data
does
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
164
support
acceptance
of
the
hypothesis
that
atrazine
interferes
with
normal
anuran
gonadal
development.

DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
I
agree
a
great
deal
with
what
Gerald
already
told
you.
I
think
one
of
the
points
why
we
find
so
many
differences,
and
the
big
difference
should
be
that
we
have
no
standardized
protocols
in
various
studies.

I
think
one
of
the
goals
for
the
future
in
future
studies
is,
first,

to
design
more
or
less
a
standardized
protocol
to
work
on
amphibians,

especially
on
xenopus
laevis,
but
also
if
you
use
ranids.

So,
we
are
comparing
apples
with
pears
if
you
don't
go
ahead
with
a
really
standardized
protocol.
That's
one
comment
on
it.

DR.
ROBERTS:
Dr.
Denver,
do
you
have
something
to
add?

DR.
DENVER:
I
just
agree
that
we
have
sufficient
data
to
generate
a
hypothesis
but
insufficient
data
to
test
the
hypothesis.

DR.
ROBERTS:
Other
members
of
the
panel?
Dr.
Matsumura.

DR.
MATSUMURA:
I
agree
with
Dr.
LeBlanc's
statement
so
long
as
it
is
limited
to
gonadal
abnormalities.
I'm
not
convinced
about
the
whole
scale
developmental
effect
even
including
laryngeal
effect.

But
regarding
the
abnormalities,
I
feel
that
the
having
Dr.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
165
Carr's
independent
experiment
as
well
as
John
Giesy's,
I
feel
that
qualitative
there
is
some
clear
cut
independent
verification
of
the
phenomenon.
But
others,
I
would
not
stick
my
neck.
So
that's
just
my
personal
feeling.

That's
why
I
asked
Dr.
Carr
to
just
tell
me
that
he
can
really
stand
behind
the
data.
He
told
me
that.
So,
I'm
satisfied
by
independent
verification.
But
other
parameters
I
would
not
agree
that
there
is
sufficient
base.

DR.
ROBERTS:
Dr.
Matsumura
raised
the
point
­­
you
know
the
question
asked
about
developmental
effects.
I
guess,
we
need
to
be
careful
how
we
define
those
in
terms
of
our
conclusions.

I
would
also
like
to
point
out
that
I
want
to
be
clear,
when
the
question
asks
whether
or
not
the
overall
body
of
available
data
is
adequate
to
demonstrate
whether
or
not
atrazine
causes
developmental
effects.

I
thought
I
heard
two
different
answers
to
that
question
among
panel
members.
I
want
to
sort
of
see
what
people
think.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
I
think
the
question
is
whether
the
data
is
sufficient
to
entertain
the
hypothesis
that
atrazine
produces
developmental
effects.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
166
The
answer
was
a
narrowly
focused
one.
It
said
that
the
data
on
gonadal
development
coming
from
two
very
different
frog
species,

one
totally
aquatic
and
one
at
least
partially
terrestrial,
terrestrial
during
most
of
the
lifespan,
did
indicate
an
effect
on
gonadal
development.

So
for
that
reason,
entertaining
the
hypothesis
was
regarded
as
an
appropriate
thing
to
do.

DR.
ROBERTS:
So,
not
to
put
words
in
your
mouth,
but
your
opinion
is
that
the
data
are
sufficient
to
support
the
hypothesis
or
to
justify
the
hypothesis,
but
not
necessarily
to
demonstrate
which
I
would
interpret
as
prove
the
hypothesis.

DR.
KELLEY:
It
was
very
well
put
by
the
EPA.
You
could
look
at
the
data
and
say
the
data
is
such
that
we
could
reject
this
hypothesis.
I
do
not
believe
the
panel
would
support
that
conclusion
although
we
haven't
heard
from
everybody.
I
do
not
believe
that
we
could
reject
the
hypothesis.
So,
what
that
means
is
the
hypothesis
is
still
viable.
Is
the
hypothesis
more
than
a
hypothesis,

do
we
actually
believe
that
atrazine
will,
at
some
dose
that
we
don't
know
now
and
some
range
of
species
that
we
don't
know
now,
reliably
result
in
gonadal
abnormalities?

I
don't
think
nor
do
I
believe
the
panel
thinks
that
the
data
are
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
167
sufficient
to
accept
that
hypothesis.
But
the
data
are
certainly
sufficient
not
to
rule
it
out
and
to
continue
to
entertain
it
as
a
model
for
looking
further.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
I'm
going
to
push
you
a
little
bit
more
though.

I'm
just
going
to
reread
the
statement
just
to
make
sure
we
answer
the
question
as
asked.
The
statement
­­
the
question
ends
with
essentially
available
data
­­
EPA
is
questioning
whether
available
data
is
adequate
to
demonstrate
whether
or
not
atrazine
causes
developmental
affects.
So,
you're
saying
no.

DR.
KELLEY:
Read
the
last
phrase.

DR.
LEBLANC:
Under
conditions
described
in
these
studies.

DR.
KELLEY:
So
we
have
data.
We
have
data
from
a
reasonably
large
number
of
studies
that
indicate
that
at
least
one
dose
there
was
a
gonadal
defect
in
a
varying
proportion
of
animals.

I
have
a
where
there
is
smoke
there
is
fire
reaction
to
that.
I
think
there
is
something
going
on,
that
is
my
bottom
line,
but
I
don't
know
there
is
something
going
on.

I
believe
it
is
worthwhile
investigating
the
hypothesis
further,

but
I
wouldn't
believe
that
it
was
proven
at
this
point.
It
certainly
raises
concerns.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
168
DR.
ROBERTS:
Whenever
we're
not
sure
exactly
what
the
question
is
asking,
we
sometimes
ask
the
agency
to
clarify
for
us.

Is
this
question
asking
whether
or
not
the
data
provided
by
the
existing
studies
is
sufficient
to
demonstrate
that
affect
or
is
it
more
along
the
lines
of
what
Dr.
Kelley
is
talking
about?

Is
it
justify
pursuit
of
the
hypothesis
but
not
necessarily
demonstrate
the
phenomenon
which
is
what
I,
sort
of,
read
into
the
white
paper.
If
you
can
clarify
that
for
us.

DR.
BRADBURY:
It's
amazing
how
one
can
wordsmith,
and
work
and
work,
and
think
it's
clear,
and
then
you
need
to
work
on
it.

The
second
phrase
is
commenting
on
whether
the
overall
body
of
available
data
is
adequate
to
demonstrate
in
the
describe
studies.

I
think
that's
an
important
phrase.
Based
on
the
body
of
information
is
there
some
sense,
given
the
quality
of
the
studies
and
the
characteristics
of
the
studies,
is
there
something
going
on,
is
there
some
smoke?

That
is
sort
of
in
the
context
in
is
there
enough
smoke
to
say,
I
think
it's
reasonable
to
formulate
a
hypothesis
to
look
to
see
if
there
is
fire
associated
with
the
smoke.
That's
the
context
of
the
question,

intended
meaning
of
the
question.

I
apologize
if
you're
confused.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
169
DR.
ROBERTS:
With
that
clarification,
then,
both
of
you
have
answered
in
the
affirmative
then?

DR.
LEBLANC:
I
have.

DR.
ROBERTS:
Dr.
Kelley?

DR.
KELLEY:
I
have
as
well.

DR.
ROBERTS:
Dr.
Denver?

DR.
DENVER:
Yes.

DR.
ROBERTS:
Is
there
anyone
else
that
would
like
to
weigh
in
on
this?

Dr.
Kloas
you
were
a
discussant
did
you
want
to
indicate
one
way
or
the
other
on
this?
Dr.
Coats?

DR.
COATS:
Yes,
I
will
weigh
in.
I
think
there
is
enough
data
to
pursue
forward
movement
to
test
the
hypothesis.

DR.
DELORME:
I
would
concur
as
well.
I
think
one
of
the
things
we
saw
is
there
are
a
number
of
factors
with
the
husbandry
and
whatnot
that
might
have
resulted
in
something
going
on.
So,
it
is
not
really,
really,
clear
that
atrazine
is
the
root
cause.
So,
we
need
to
pursue
the
hypothesis.

DR.
ROBERTS:
Would
anyone
else
like
to
weigh
in
on
this
before
we
move
on
to
B?

DR.
SKELLY:
I
will
concur
broadly.
I
think
the
way
that
Dr.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
170
LeBlanc
and
the
others
responded
reflects
my
feelings
as
well.

DR.
ROBERTS:
Dr.
Green?

DR.
GREEN:
I
concur
as
well.

DR.
ROBERTS:
Good.
Dr.
Richards?

DR.
RICHARDSON:
I'll
join
the
party.

DR.
ROBERTS:
Then
let's
go
ahead
and
take
part
B?

DR.
STEEGER:
Please
comment
on
EPA's
conclusion
that
given
the
variability
in
the
available
dose
response
data
across
the
studies,
e.
g.,
approximately
250­
fold
difference
in
reported
thresholds
for
observed
developmental
effects
as
well
as
reports
of
monotonic
and
nonmonotonic
dose
response
curves,
it
is
not
possible
to
ascertain
the
relationship,
if
any,
of
atrazine
exposure
to
developmental
effects
in
amphibians.

DR.
ROBERTS:
Dr.
LeBlanc,
what
do
you
think?

DR.
LEBLANC:
The
short
answer
is
I
agree
wholeheartedly,

and
I
think
I
only
need
to
expand
on
that
a
little
bit.

There
is
clearly
major
deficiencies
that
exist
in
the
data
as
related
to
describing
the
relationship
between
atrazine
exposure
and
gonadal
toxicities
that
have
been
reported.

We
can
extract,
I
think,
a
small
amount
of
data
as
related
to
that
relationship
that
presumed
relationship,
at
this
point
in
time,
although
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
171
the
conclusions
are
very,
very
limited.

For
example,
I
think,
I
hope,
we're
all
in
agreement
that
atrazine
at
a
concentration
of
0.01
microgram
per
liter
appears
to
have
no
effect
on
anuran
gonadal
development.

The
data
seems
to
suggest
that
exposure
to
frog
larvae
to
atrazine
concentrations
greater
than
or
equal
to
.1
microgram
per
liter
can
elicit
developmental
abnormalities,
not
does
but
can.

There
is
limited
data
available
to
us
to
make
any
judgments
as
to
whether
or
not
concentrations
in
the
range
of
.1
to
10
micrograms
per
liter
truly
elicit
adverse
effects.

However,
when
we
consider
concentrations
in
the
vicinity
of
25
micrograms
per
liter,
there
seems
to
be
some
reasonably
good
concordance
that
this
concentration
does
elicit
gonadal
effects.

Clearly,
more
data
is
required
to
define
the
concentration
response
relationship
between
atrazine
and
gonadal
development
of
anuran
larvae.

At
this
time,
I
would
have
confidence
only
in
concluding
that
the
threshold
concentration
for
this
material
exists
between
.01
and
25
micrograms
per
liter.

DR.
ROBERTS:
Thoughts
from
other
panel
members?
I
will
start
with
the
lead
discussants,
Dr.
Kloas,
did
you
want
to
add
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
172
anything
or
comment
on
this?

DR.
KLOAS:
Concur.

DR.
ROBERTS:
Dr.
Denver?

DR.
DENVER:
I
would
just
concur
with
the
statement
that
we
put
together
last
night.

DR.
ROBERTS:
Other
comments
from
other
members
of
the
panel?

Does
anyone
feel
that
the
data
are
adequate
to
describe
the
dose
potential
dose
response
relationships?

Dr.
Matsumura?

DR.
MATSUMURA:
I
don't
think
it
is
adequately
describing,

all
we're
saying
is
that
the
threshold
concept
can
be
applied
and
that
some
experiments
indicate
that
there
is
such
a
thing.

Real
threshold
experiments
must
be
done
rather
carefully.
It
is
not
that
easy.
How
many
years
it
took
to
have
any
agreement
on
the
cancer
dose
and
the
effect
relationships,
particularly
with
the
hormones,
there
are
feedbacks
and
­­
you
know,
Jere.

So,
I
would
vote
for
the
statement
that
the
threshold
is
a
good
way
to
go
about
it
and
there
are
some
indications,
incipient
indications.

DR.
ROBERTS:
In
the
spirit
of
Dr.
LeBlanc's
earlier
comment
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
173
which
I
concurred
that
these
were
actually
concentration
response
relationships
and
not
dose
response
relationships.
So,
I'm
going
to
correct
myself
in
agreement
with
that.

DR.
LEBLANC:
I
think
we
need
to
recognize
that
the
great
majority
of
the
studies
that
we
have
reviewed,
I
can
think
of
one
exception,
were
not
designed
to
evaluate
the
concentration
response
relationship.
Certainly,
future
studies
need
to
be
designed
appropriately
so
that
the
response
relationship
can
be
appropriately
evaluated.
I
made
some
attempts
to
drive
some
general
understanding
of
what
this
relationship
might
be
based
upon
all
the
information
that
was
provided
to
us.

The
best
that
I
could
come
up
with,
and
I
don't
have
a
lot
of
confidence
in
it,
is
that
.01
micrograms
per
liter
has
no
effect
and
all
concentrations
evaluated
greater
than
that
had
an
effect
on
average.

And
on
average,
that
effect
regardless
of
the
concentration
seemed
to
have
been
around
20
percent
incident
of
effect
whatever
the
investigators
may
have
been
monitoring.

Now,
I
don't
know
what
that
means.
It
could
mean
that
the
maximum
response
that
we
should
anticipate
is
20
percent.
And
that
the
concentration
response
is
going
to
occur
between
.01
and
.1,
the
threshold
will
be
between
there.
I
don't
know
that
that's
the
case.
It
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
174
is
just
a
suggestion.

And
there
are
precedents
for
observing
partial
effects
where
perhaps
we
would
be
anticipating
greater
effects?
What
comes
to
mind
are
my
own
studies
looking
at
intersex
in
snails
as
related
to
tributal
10.
In
laboratory
exposures
where
we
try
and
replicate
what
we
see
in
the
environment
in
10
contaminated
environments
where
the
incidence
is
100
percent
intersex,
the
best
we
can
generate
is
about
30
percent.
We're
not
alone,
other
labs
have
had
similar
success
in
generating
intersex
in
only
about
30
percent
of
the
animals.
We
really
don't
know
why.
We
assume
it
is
a
deficiency
in
our
experimental
design.

We
know
the
animals
are
capable
of
­­
the
population
is
capable
of
totally
responding,
but
in
the
lab
they
simply
don't.
I
don't
think
we
need
to
go
into
these
experiments
with
the
anticipation
that
we're
going
to
see
100
percent
response
in
a
population
if
20
percent
is
truly
the
maximum
that
we
can
anticipate,
we
should
accept
that
and
live
with
it.

But
we
should
still
be
able
to
define
that
relationship
between
a
threshold
concentration
and
that
maximum
effect
whatever
it
might
be.

DR.
ROBERTS:
Okay.
Are
there
other
comments
you
want
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
175
express
regarding
this?

I'll
say
for
myself
personally,
I
agree
that
I
don't
think
the
data
are
adequate
to
describe
the
dose
response
relationship.

I
have
not
gone
through
the
analysis
you
have.
I
have
no
idea
where
an
apparent
threshold
might
be.
Speaking
personally,
I
have
no
idea
what
the
shape
of
the
dose
response
curve
might
be.
But
I
do
agree
there
is
not
enough
data
to
establish
that.

Anyone
else?
Let's
go
ahead
and
move
on
to
the
next.

DR.
STEEGER:
Many
of
the
available
studies
proposed
that
aromatase
induction
results
in
elevated
estrogen
levels
that
lead
to
feminization
as
characterized
by
ovotestes,
intersex,
and
hermaphroditism
in
genetically
male
amphibians.

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

DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
First
of
all,
I
would
like
to
say
that
it
is
not
the
only
hypothesis
which
could
raise
such
phenomenons,
there
are
feminizational
demasculization
phenomenon.

We
would
have
two
ways
how
to
obtain
them
estrogenic
and
anti­
androgenic
ones,
that's
the
first
remark
I
want
to
make.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
176
Answering
question
4
A,
of
course
it
is
correct.
There
are
no
data
showing
up
any
aromatase
induction
or
stimulation
caused
by
atrazine.
But
however,
I
think
we
have
already
discussed
there
is
no
approach
done
which
would
be
appropriate
to
demonstrate
it.

So,
I
think
that
experimental
designs
done
up
to
now,
they
don't
support
this
hypothesis
but
they
cannot
support
such
a
hypothesis
for
aromatase
stimulatory
effects.

Would
you
like
to
add
something?

DR.
ROBERTS:
Thank
you.
So
in
your
opinion
it
is
still
an
open
question?

DR.
KLOAS:
Yes.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
I
would
agree.
I
would
like
to
point
out
that
the
aromatase
gene
or
at
least
one
of
them,
they
probably
have
two,
the
spene
laevis
(
ph)
has
been
cloned
in
xenopus
laevis.
It
will
be
possible
to
study
the
expression
of
the
aromatase
gene
in
the
developing
gonad
in
the
presence
and
the
absence
of
agents
that
are
thought
to
affect
it.

While
the
MRNA
expression
is
not
definitive
with
respect
to
protein
expression
or
the
activity,
the
enzymatic
activity
of
that
protein,
on
the
other
hand,
you
do
have
to
have
the
gene
expressed
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
177
have
it
there.
So
we
now
have
available
the
tools
to
test
this
hypothesis
in
a
rigorous
fashion.
It
should
be
possible
to
test
that.

I
do
concur
absolutely
that
although
this
is
a
popular
hypothesis,
it
is
not
the
only
hypothesis
that
could
provide
a
mechanism
of
action
for
effects
that
have
been
reported.

And
so
to
focus
on
this
hypothesis
to
the
exclusion
of
other
kinds
of
ideas
like
changes
in
hormones
within
the
animal's
body
and
steroid
hormones
changes
in
the
hypothalamic,
pituitary,
gonadal
axis
is
almost
certainly
a
mistake.
There
are
bound
to
be
other
mechanisms
that
are
involved.
DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
I
would
also
like
to
add
one
comment.
It
could
be
also
due
that
there
is
a
direct
interaction
with
the
enzyme
aromatase.
So
I
think
if
you
could
invitro
assay,
you
should
add
atrazine
at
different
concentrations
to
show
up
if
there
is
any
stimulatory
effect
which
could
be
done
very
easily
because
already
aromatase
was
assessed
biochemically
in
several
labs.
And
it
is
a
routine
assay.

I
think
that's
second
possibility.
Not
only
looking
on
gene
expression,
but
also
on
direct
interference
with
the
enzyme
as
a
second
possibility
to
rule
out
if
there
is
any
interference
with
aromatase
and
the
output
for
estrogens.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
178
DR.
ROBERTS:
Thank
you.

Dr.
Denver.

DR.
DENVER:
The
data
to
at
least
generate
the
hypothesis
are
based
on
studies
in
human
cell
lines
and
a
limited
analysis
in
alligators.
The
hypothesis
has
not
been
directly
tested
in
amphibia.

Just
one
point
I
would
add,
to
test
it,
not
only
do
we
need
to
take
advantage
of
the
molecular
tools
that
are
available,
but
also
do
it
at
the
appropriate
developmental
stage
at
least
in
one
that
is
hypothesized
to
be
sensitivity
to
atrazine.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
First,
I
would
like
to
just
agree
with
everything
that
has
been
said.
But
I
would
like
to
make
the
point
strongly
that
in
agreeing
that
aromatase
induction
has
not
been
demonstrated,
it
should
not
be
construed
to
mean
that
it's
not
induced
from
my
perspective.

It
simply
means
that
the
appropriate
experiments
haven't
been
done
to
demonstrate
whether
or
not
induction
occurs.

And
certainly,
a
point
that
I
feel
extremely
strongly
about
in
future
studies
looking
at
aromatase
induction
is
that
the
experiments
need
to
be
conducted
in
the
right
life
stage,
not
in
the
adult,
but
in
the
larvae
that
we're
interested
in.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
179
DR.
ROBERTS:
Dr.
Kelley,
you
are
nodding.
Agreed?

DR.
KELLEY:
Yes.
To
demonstrate
that
you
have
an
effect
of
an
agent
in
an
adult,
say,
for
example,
hormone
levels,
and
then
to
infer
that
that
same
effect
with
an
unknown
mechanism
should
have
happened
at
developmental
stage
is
taking
the
hypothesis
quite
far
down
the
line
without
evidence.

So
really,
you
have
to
go
back
to
the
stage
when
the
effects
are
thought
to
occur
in
order
for
the
mechanism
to
be
established
or
not
established.

DR.
ROBERTS:
So
the
comments
so
far
I
think
have
expressed
the
opinion
that
it
has
not
been
demonstrated
that
aromatase
is
induced.
However,
the
experiments
have
not
necessarily
been
appropriate
design
to
adequately
test
that.

There
have
been
some
suggestions
about
how
the
best
to
approach
that
test
as
well
as
a
caution
from
the
panel
that
about
focusing
specifically
on
aromatase
as
a
possible
mode
of
action
that
there
are
other
endocrine
mechanisms
that
need
to
be
considered
as
well.

Does
everybody
sort
of
agree
with
that
summarization?
Are
there
any
other
points
or
is
there
disagreement?

Dr.
Matsumura.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
180
DR.
MATSUMURA:
Just
I
want
to
make
sure
that
­­
yes,
our
chairman
really
summarized
right.
We
should
not
really
lock
into
that.
Dr.
Kloas
also
suggested
some
different
approaches
such
as
the
antagonist
of
the
testosterone.

This
is
not
the
typical
estrogenic
response.
It
doesn't
happen
in
the
females.
It
doesn't
look
like
any
effects.
So
should
not
be
locked
in.
I
agree
with
Dr.
Kelley's
statement
that
there
are
other
possibilities
that
we
have
to
keep
our
minds
open.

DR.
ROBERTS:
Any
other
thoughts
or
comments
about
A?
Is
the
feedback
clear?
Great.

Let's
go
to
B.

DR.
STEEGER:
The
variability
associated
with
plasma
sex
steroid
concentrations
and
aromatase
activities
is
high.
Is
this
variability
normal.

Please
comment
on
any
readily
apparent
or
available
methodological
improvements,
for
example,
changes
in
sampling
design,
analytical
techniques
that
could
efficiently
address
this
variability
in
future
studies.

DR.
ROBERTS:
Dr.
Kloas
and
Dr.
Kelley
I
believe
during
some
earlier
discussions
you
made
some
points
about
the
sex
steroid
measurements.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
181
What
is
your
response
to
this
question?

DR.
KLOAS:
First
of
all,
I
think
all
of
us
being
aware
of
this
kind
of
determinations
agree
that
there
is
always
in
lower
vertebrates
a
high
variability
and
steroid
levels
in
general
and
also
activities
in
aromatase
may
vary
widely.

I
think
there
should
be
agreement
here.
So
the
variability
shown
up,
until
yet
measured,
seems
to
be
normal,
what
you
will
normally
get.

Anyhow,
we
have
to
commend
or
we
have
some
remarks
about
ELISA
data,
especially
for
estradiol
measurements
presented.

I
think
there
are
at
least
one
order
or
two
orders
of
magnitude
higher
than
all
the
old
data
measured
also
in
some
of
our
labs
that
would
suggest
to
be
in
this
level.
So
they
are
much
higher.
So
there
should
be
any
proof
that
this
really
is
the
right
way
to
assess
estradiol
in
xenopus.

For
anything
else,
I
think
changes
in
sampling
design
­­
I
would
like
to
show
up
set
series
(
ph)
involvement
in
steroid
or
in
sexual
steroid
genesis.
I
think
first
of
all
there
should
be
short
term
exposures
and
short
term
measurements
for
estradiol
as
well
as
for
androgens.

I
would
also
refer
that
there
could
be
also
some
interference
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
182
with
ratio
of
testosterone,
dehydro
(
ph)
testosterone.
And
I
would
also
refer
that
there
should
be
also
some
improvements
to
include
five
alpha
reductase
measurements,
because
this
could
be
also
another
hypothesis
how
it
works
and
how
it
could
work
in
anti­
androgenic
way
(
ph).

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
I'm
going
to
read
my
remarks.

The
variability
associated,
the
largest
variability,
two
orders
of
magnitude,
in
other
words,
1,000
times,
in
the
studies
reviewed
by
the
EPA
reflects
differences
in
results
obtained
in
studies
sponsored
by
the
registrant
using
ELISA
assays
and
studies
in
the
open
literature
using
radioimmunoassay.

The
most
likely
explanation
for
this
large
discrepancy
is
the
method
used
to
measure
hormone
levels
­­
so
that's
the
analytical
comment
up
here.
Resolving
these
discrepancies
should
be
straightforward.

Within
the
open
literature,
on
the
other
hand,
variability
is
more
typical
typically
in
the
two
times
range,
a
range
well
within
diurnal
and
seasonal
variation.

It
should
be
recognized
that
species
studied
extensively
in
the
laboratory,
xenopus
laevis,
may
have
originated
from
different
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
183
populations
in
their
native
habitat.

Populations
can
differ
in
seasonality
even
within
the
same
country.
For
example,
in
South
Africa,
the
Cape
population
breeds
in
the
African
winter,
and
the
Johannesburg
population
in
the
African
summer.

And
we
would
thus
expect
differences
between
the
two
populations
in
the
field
and
in
the
laboratory
as
this
species
maintains
circannual
rhythms
even
after
many,
many
generations
in
the
laboratory.

When
wild
caught
animals
are
brought
into
the
laboratory
for
study,
they
should
be
characterized
genetically
to
identify
the
population
of
origin,
this
is
possible
now
based
on
literature,
and
also
to
verify
species
because
I'm
sorry
to
tell
you
every
xenopus
looks
like
every
other
xenopus
pretty
much,
except
some
are
big
and
some
are
small.
Although,
if
you
listen
to
their
songs,
you
can
tell
the
difference.

And
the
same
approach
must
be
employed
in
characterizing
groups
of
animals
that
are
used
for
laboratory
studies.
So
this
is
by
way
of
saying
that
if
we're
going
to
standardize
the
animal
as
a
test
species,
we
have
to
develop
standards
that
don't
just
involve
the
laboratory
assays
that
we
do,
but
that
involve
the
biology
of
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
184
animal
that
we're
looking
at.

DR.
ROBERTS:
Dr.
Denver.

DR.
DENVER:
I
just
want
to
reiterate
the
importance
of
validation
of
immunoassays,
especially
assays
for
estradiol.

Estradiol
is
known
to
be
present
at
very
low
concentrations
in
plasma
and
is
routinely
measured
at
least
in
the
laboratories
that
are
represented
here
by
radioimmunoassays.

And
historically,
radioimmunoassays
have
been
used
to
measure
plaza
sex
steroids
and
tend
to
be
more
sensitive
than
ELISA's,
although
some
of
the
current
generation
of
ELISA's
may
be
just
as
sensitive.

But
I
think
it
is
important
to
compare
the
two
methods
to
make
sure
that
they
are
in
concordance
and
to
validate
the
assay
methods
for
each
species
under
study.

There
is
a
tendency
these
days
to
purchase
a
kit
and
put
things
into
it.
I'm
not
saying
that
this
is
what
was
necessarily
done
by
the
investigators
here,
but
it
is
really
important
to
evaluate
that
kit
for
the
species
under
study
and
show
that
you're
actually
measuring
what
you
think
you
are
measuring
and
that
you're
actually
recovering
from
your
sample.

In
experiments
where
you
add
known
amounts
of
cold
steroid,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
185
you
can
actually
recover
that
in
the
assay.

So
those
are
important
things.
Because
the
data
that
you
generate
are
only
as
good
as
the
methods
that
you
are
using
to
generate
them.
There
has
been
a
lot
of
discussion
about
animal
husbandry,
which
is
important,
but
I
think
there
should
be
similar
discussion
of
the
methods
used
to
analyze
steroids,
because
perhaps
much
or
perhaps
some
of
this
variability,
at
least,
could
be
due
to
analytical
methods.

The
fact
that
one
lab
can
use
a
kit
and
get
the
same
result
doesn't
necessarily
mean
that
result
is
correct.
You
can
repeat
an
artifact
until
you
are
blue
in
the
face
and
it
doesn't
necessarily
mean
that
that
is
a
correct
number.

DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
I
think
also
another
possibility
to
overcome
such
a
problem
for
measuring
estradiol
could
be
to
measure
an
estrogenic
biomarker.
A
lot
of
estrogenic
biomarkers
are
now
already
available.

The
methodology
is
well­
established.
For
instance,

vitellogenin,
you
can
measure
it
in
the
plasma,
but
also
as
vitellogenin
MRNA
by
RTPCR
techniques.
That's
routine
measurement.

You
can
also
maybe
in
an
indirect
way
demonstrate
that
there
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
186
could
be
any
estrogenic
interactions
or
interference
caused
by
atrazine.

So
This
is
another
possibility
to
have
another
methodological
approach
to
demonstrate
estrogenic
pathways.

DR.
ROBERTS:
Yes.
Dr.
Kelley
raised
that
issue
earlier
in
discussion
with
one
of
the
presenters.

If
she
wants,
we
could
include
that
as
part
of
our
response
to
this
question.

DR.
KELLEY:
I
also
want
to
make
the
comment
that
snapshot
measurements
of
hormone
levels
at
a
particular
point
in
time
do
not
give
a
valid
picture
of
the
history
of
exposure
of
animal
to
the
hormone
in
question.

What
you
really
want,
what
I
want,
are
biomarkers
that
show
you
in
living
realtime
color
what
the
animal
is
seeing.
We
are
fortunate
with
xenopus
that
those
are
available
now,
actually.
We
will
be
able
to
look
at
flashing
green
frogs
and
know
what
tissue
is
being
exposed
to
what
level
of
hormone,
when
and
how,
which
is
an
unusual
situation.

But
even
without
this
fancy
molecular
method,
the
biomarkers
such
as
have
just
been
described
are
really
very
well
characterized.

And
you
know
that
if
you
see
a
male
that
has
any
vitellogenin
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
187
or
further
that
shows
rapid
response
to
estrogen
challenge,
you
know
that
male
has
seen
a
level
of
vitellogenin
that
is
not
present
in
the
literature.

So
these
are
available
assays
and
should
be
used
and
added
to
the
armamentarium.

DR.
ROBERTS:
Dr.
Kelley,
can
we
provide
some
specifics
and
citations
in
our
minutes?

DR.
KELLEY:
I
have
them
in
my
giganto
reference
list.

DR.
ROBERTS:
Somehow
I
knew
you
would.

Dr.
Tietge
I
believe
has
a
question.

DR.
TIETGE:
I'm
a
little
confused
on
one
point
here.
Maybe
I
didn't
catch
it
quite
correctly.

Dr.
Kloas
said
that
there
is
high
variability
in
these
measurements
and
that's
somewhat
normal.

Dr.
Kelley
said
that
the
typical
range
of
variability
is
twofold.

Are
you
in
agreement,
the
two
of
you?
Did
I
get
that
right?

DR.
KELLEY:
The
typical
range
­­
would
you
agree
that
the
typical
range
of
variation
is
not
two
orders
of
magnitude?
1,000
times?

DR.
ROBERTS:
That's
100.

DR.
KELLEY:
Sorry.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
188
Well,
you
know
what
I
mean.
It's
not
1,000
times?

DR.
ROBERTS:
I
think
she
meant
two
orders
of
magnitude
perhaps.

DR.
KLOAS:
Not
twofold.

DR.
KELLEY:
I'm
sorry.
I
meant
two
orders
of
magnitude.

DR.
TIETGE:
That
clarifies
it.
Thank
you.

DR.
KLOAS:
We
agree
with
each
other.

DR.
KELLEY:
We
agree
that
twofold
is
normal,
and
1,000
times
is
not
normal.
I
actually
wrote
that
down.

DR.
ROBERTS:
Two
orders
of
magnitude,
yes.
Three
orders
of
magnitude,
no.

DR.
KELLEY:
Is
twofold
two
orders
of
magnitude?
No.

Twofold
is
within
an
order
of
magnitude.
It's
within
10
times.

DR.
MATSUMURA:
I
would
like
to
support
the
Werler's
(
ph)

comment.
Really,
those
steroid
levels
can
change
really
by
the
hour.

So
I
would
like
to
really
recommend
that
the
endpoint
of
those
steroid
action
­­
actually,
I
have
plasmid
for
the
vitellogenin
on
the
control
by
ERE.
It
comes
from
xenopus.
I
got
it
from
Shapiro.
And
I
also
have
a
PS2,
which
is
very
sensitive.

It
is
well
constructed.
It
works.
I
offer
to
Joe
Giesy,
if
he
wants,
and
anybody
who
wants
I
have
it.
Of
course,
I
have
to
write
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
189
Shapiro
saying
that
I
will
transfer
that.

But
really,
those
are
sensitive.
I
was
so
amazed.
I
compared
that
to
mammalian
construct
and
the
xenopus
is
better.
Much
purer
ERE.
You
can
detect
the
PS2
very
easily
or
PGR,
progesterone
receptor.

Those
are
well,
well
accepted.
You
can
run
that
easily.
I
support
that.
Because
that
is
more
stable
a
way
of
measuring.

Some
time
ago
hormone
was
up,
but
its
effect
is
still
here.

DR.
ROBERTS:
Any
other
comments
on
this
before
we
move
on
to
the
next
one?

Clarification?

DR.
BRADBURY:
I'm
sure
you
all
do
it.
If
you
can
get
all
this
written
down,
some
of
the
dialogue.
That
will
be
very
helpful.

Based
on
the
discussions
over
the
last
few
days,
still
within
the
realm
of
this
question,
but
one
of
the
hypotheses
was
it
was
­­
the
increase
in
estradiol,
that
the
estradiol
doesn't
get
into
the
plasma,
it
stays
within
the
tissue.

Are
there
techniques
available
to
measure
changes
in
estradiol
concentration
within
the
tissues?

Could
you
comment
on
that?

DR.
ROBERTS:
Dr.
Kloas.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
190
DR.
KLOAS:
I
think
it
is
easy
to
measure
also
from
tissue's
cystorrheic
(
ph)
levels.
You
can
extract
them
easily.
But,
however,
I
think
it
is
a
completely
lipophilic
compound,
how
to
keep
it
­­
it
has
to
be
really
completely
resorped
by
receptors
at
the
same
time.

Normally,
there
should
be
any
leakage
to
the
circulation.
I
wouldn't
expect
that
could
be
stored,
of
course.
So
should
be
a
change.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
As
Dr.
Kelley
said
also,
it
is
really
a
snapshot
that
we're
taking
be
it
in
the
serum
or
be
it
in
a
given
tissue.

And
that
snapshot
tends
to
reflect
very
well
the
level
of
synthesis
that's
taking
place
in
that
tissue
at
that
point
in
time.

So
another
alternative
would
be
rather
than
looking
at
tissue
levels
at
a
given
hormone,
if
we're
working
with
gonads
would
be
perhaps
to
remove
the
gonads
and
in
tissue
culture
look
at
the
production
steroid
synthesis.

And
another
strength
to
that
approach
too
is
you
could
be
at
the
same
time
looking
at
production
of
multiple
steroids,
androgens
and
estrogens.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
Meashita
(
ph)
and
coworkers
in
Japan
have
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
191
routinely
cultured
the
Stage
51
gonad.
In
the
presence
of
aromatase
inhibitors
and
also
in
the
presence
of
estrogen,
are
able
to
observe
sexual
differentiation
invitro
and
are
able
to
observe
effects
of
the
agents
that
they
add.

That
assay
is
very
available.

Also,
I
want
to
point
out,
I
know
I
pointed
this
out
before,
but
the
liver
is
not
that
close
to
the
gonad.
Typically,
if
you
have
evidence
of
vitellogenin
­­
let
me
explain
for
the
public
what
vitellogenin
is.
It
is
the
yolk
proteins.
Birds
have
it.
We
don't
have
so
much
yolk.
But
the
frogs
have
a
lot
of
it.
A
mature
egg
has
to
yolk
up
in
order
for
it
to
be
oviposited,
ovulated
and
oviposited.

If
you
see
vitellogenin
in
the
liver,
it
had
to
have
arrived
there,

I
believe
there
is
agreement,
via
the
circulation.
So
it
had
to
have
circulated
at
some
point.
That
would
mean
that
it
got
out
of
the
organ.

It
is
worth
pointing
out,
it
has
been
pointed
out,
for
example,

females
use
testosterone
as
a
precursor
for
estrogen.
Dr.
Hayes
was
right,
that
he
didn't
get
the
yolk
up
except
in
rana.

But
xenopus
uses
testosterone
as
a
precursor
for
estrogen.
You
can
measure
really
high
levels
of
testosterone
in
females
when
their
ovaries
are
activated.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
192
So
circulating
levels
can
be
useful
and
shouldn't
be
denied,
but
there
are
other
tools
available
to
us.

DR.
ROBERTS:
Let
me
ask
Dr.
LeBlanc
and
Dr.
Kelley
a
question
of
clarification
for
my
edification.
Not
for
the
agency's.

The
experiments
in
gonadal
culture
that
you
described,
are
those
­­
are
you
envisioning
an
exvivo
experiment
where
they
are
exposed,
you
are
removing
gonads
from
exposed
animals
and
culturing
them
and
then
measuring
synthesis
rate?

DR.
LEBLANC:
Yes.

DR.
ROBERTS:
So
the
exposure
would
occur
in
the
animal,
but
not
while
in
consult
(
ph)
culture.

DR.
LEBLANC:
That's
correct.

DR.
ROBERTS:
And
you
would
still
expect
to
see
a
continuing
effect
of
the
atrazine
even
though
the
gonads
in
culture
would
not
be
exposed
to
atrazine?

DR.
LEBLANC:
It
certainly
depends
on
the
mechanism
of
the
effect.
But
if
it
were
induction
of
the
enzyme,
then
in
the
short
term,

yes,
you
would
expect
to
see
­­

DR.
ROBERTS:
Short
term
until
it
declines,
okay.

DR.
KELLEY:
Just
to
echo
a
point
made
here
about
short
term
exposure,
it
is
a
tremendous
advantage
that
you
can
get
sexual
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
193
differentiation
over
a
48
hour
period
in
xenopus.

Because
of
that,
you
can
do
it
in
culture.
You
don't
have
to
worry
about
the
rundown
of
the
cells
so
much.
You
don't
have
to
worry
about
the
compensatory
mechanisms
within
the
entire
animal.

And
that's
a
big
experimental
advantage.

DR.
ROBERTS:
Are
there
any
follow­
up
questions
from
the
agency
on
this?
Any
other
comments?

Let's
then
move
on
to
the
next
question.

DR.
STEEGER:
Please
comment
on
whether
there
are
additional
data,
other
than
those
summarized
in
the
white
paper,
that
suggest
late
exposure
of
amphibians,
i.
e.,
juveniles
or
adults
to
estrogens
or
estrogenic
chemicals
can
induce
ovotestes
formation.

DR.
KLOAS:
As
far
as
we
are
aware,
not
in
xenopus,
at
least.

I
cannot
give
you
the
citation,
but
there
is
some
Japanese
people
talking
about
ranids
where
you
could
reverse
probably
in
late
stages
of
rana
rugosa.
You
can
induce
sex
reversal.
But
not
for
xenopus.
It
is
not
reported,
at
least.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
In
ranids,
it
has
long
been
known
that
aging
animals
will
show
gonadal
changes
spontaneously.
Whether
those
are
associated
with
environmental
agents
has
never
known.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
194
I
remember
reading
an
old
report
where
old
female
ranids
will
shed
sperm.
In
xenopus,
this
has
never
been
reported,
to
my
knowledge.

So
I
don't
know
of
any
data
on
this,
but
there
may
be
people
who
know
better
than
we
do.

DR.
ROBERTS:
Dr.
Denver.

DR.
DENVER:
No,
I
don't
have
any
additional
data
to
add.

DR.
ROBERTS:
Anybody
else
on
the
panel?

DR.
ROBERTS:
Let's
go
ahead
on
to
the
next
one.

DR.
STEEGER:
Please
comment
on
whether
there
are
additional
data
other
than
those
summarized
in
the
white
paper
that
suggest
alternative
mechanisms
that
could
explain
the
apparent
feminization
of
genetically
male
amphibians.

DR.
ROBERTS:
I
suspect
Dr.
Kloas
has
an
idea
or
two
on
this.

Go
ahead.

DR.
KLOAS:
I
think
we
don't
have
really
completely
new
data
and
not
been
included
in
the
white
paper.
But
from
the
last
days
and
from
some
presentations
we
saw,
we
feel
that
you
can
create
quite
a
lot
of
hypothesis
more
or
less
which
could
account
for
these
findings
presented
by
different
groups.

I
just
would
like
to
summarize
a
little
bit
which
possibilities
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
195
could
happen.

First,
aromatase
induction
or
effect
­­
change
in
activity
of
aromatase
could
be
one.
Still
one
of
the
hypotheses
which
could
work.
The
next
one,
we
could
also
have
anti­
androgenic
effects.
A
third
one
would
be
as
already
mentioned
yesterday,
influence
by
the
hypothalamus
pituitary
gonad
axis.
So
more
than
(
inaudible)

endocrine
pathway.

First
possibility
would
be
inhibition
of
steroidogenesis,

especially
on
sex
reverse
steroids.

We
have
shown
very
easily
that
if
you
expose
animals
to
atrazine,
at
least
testosterone
levels
went
down
and
really
pronounced.

And
I
think
furthermore
there
is
not
completely
evidence
for
me
that
you
can
exclude
any
interference
with
the
thyroid
system.
There
is
still
some
data
available
also
from
the
public
comments.
Dr.

Sheffield
also
he
reminds
me
again
to
one
paper
on
larval
salamanders
that
shows
the
inhibition
or
delay
of
metamorphosis.
I
think
still
from
the
data
we
got
and
because
of
conduction
arose
(
ph)
from
the
husbandry,
it's
not
completely
clear
that
there
couldn't
be
any
interference
with
the
thyroid
system.

Next
week
we're
going
to
present
some
data
that
sometimes
you
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
196
will
not
­­
you
cannot
see
it
really
for
developmental
stages,
but
you
will
still
have
counter­
regulation
if
you
have
inhibitory
effects
on
the
thyroid
system.

This
could
be
counter­
regulated
by
more
pronounced
TSH
production.
This
way,
with
this
pathway
we
couldn't
exclude
also.

I'm
making
up
everything.
I'm
sorry
for
that.
But
the
data
presented
here
still
suggests
that
we
have
so
many
facilities
and
possibilities
which
could
be
one
of
the
pathways
or
several
pathways
how
it
could
work.

DR.
ROBERTS:
Dr.
Kelley,
do
you
have
anything
to
add?

DR.
KELLEY:
Let
me
comment
a
little
bit
on
the
thyroid
hormone
question.
Metamorphosis
in
xenopus
is
completely
dependent
upon
thyroid
hormone
and
requires
as
has
been
shown
in
a
number
of
studies
the
expression
of
thyroid
hormone
receptor,
which
there
are
two
which
exist
in
a
variety
of
isoforms.

I
was
not
able
to
detect
thyroid
hormone
receptor
alpha
in
developing
gonads.
So
I
do
not
know
if
the
developing
gonad
needs
expression
for
formation,
for
what
I
call
sex
determination.

In
fact,
in
that
study
in
which
thyroid
hormone
was
blocked,
the
gonad
proceeded
to
develop
to
the
point
of
having
spermatids.

Although,
of
course,
frank
spermatid
gonia
did
not
develop
because
­­
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
197
this
is
the
other
side
of
the
equation,
you
cannot
get
effects
of
steroid
hormones
in
xenopus
unless
the
animal
has
first
been
exposed
to
thyroid
hormone.

And
some
effects
require
thyroid
hormone
induction
of
prolactin.

So
it
is
clear
in
xenopus
that
there
is
this
strong
interaction
in
terms
of
steroid
hormone
effects
for
many
tissues
between
pituitary
hormones,
thyroid
hormones,
this
is
extending
now
to
the
axis,
the
pituitary
hypothalamic
gonadal
axis
now
to
the
thyroid
and
the
steroid
hormones.

While
this
may
sound
like
a
nightmare
like
the
ecological
studies
in
point
effect,
it's
an
exquisitely
regulated
developmental
system
which
provides
then
a
very
good
assay
system
for
a
variety
of
endocrine
perturbations.

So
I
agree
that
we
can't
at
this
point
completely
exclude
thyroid
hormone
effects.
I
don't
know
about
the
gonads.
But
if
there
are
any
other
effects
of
the
agents
that
are
being
studied,
we
should
take
this
into
account.

DR.
ROBERTS:
Dr.
Denver.

DR.
DENVER:
I
don't
know
of
any
additional
data
that
would
support
any
alternative
mechanisms,
but
I
think
it
is
important
to
at
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
198
least
entertain
the
hypothesis
that
if
there
are
gonadal
effects,
that
they
could
be
mediated
by
nonendocrine
mechanisms,
that
is,
that
there
could
be
direct
effects
on
genes
that
are
important
for
gonadal
determination
that
may
have
nothing
to
do
with
hormones.

I
don't
think
that
should
necessarily
be
ruled
out.

DR.
ROBERTS:
It
seems
we
didn't
have
a
lot
of
specific
data
to
offer,
but
the
panel
members
did
have
suggestions
for
a
number
of
alternative
mechanisms
and
seemed
to
strongly
think
that
those
ought
to
be
considered
as
well.

Are
there
any
other
comments?

Dr.
Matsumura.

DR.
MATSUMURA:
I
was
intrigued
at
least
by
the
fact
that
the
HCG
reversed
some
of
those
actions,
but
not
all
the
way.

I
would
like
to
suggest
at
least
that
area
should
be
followed
up
to
see
what
it
is.

It
remind
me
of
the
Precosin
(
ph)
studies,
insect
studies.
This
really
is
part
of
capsulatum
(
ph).

In
that
case,
you
can
reverse
it
except
that
it
doesn't
come
back
to
the
same
level
because
part
of
the
function
of
the
capsulatum
is
gone.
It
looks
like
that,
but
again,
just
one
experiment.
I'm
quite
sure
Dr.
Hayes
is
intrigued
by
that
too.
So
we'll
see
what
happens
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
199
then.

DR.
ROBERTS:
Any
other
comments?

Dr.
LeBlanc.

DR.
LEBLANC:
I
just
want
to
make
a
point
that
in
relation
to
all
the
possible
effects
that
Dr.
Kloas
talked
about
with
terminal
hormones,
the
androgens,
the
estrogens,
thyroid,
hormone,
that
the
effects
need
not
be
self
exclusive,
that
there
could
be
effects
at
the
level
of
the
hypothalamus,
the
pituitary
that
results
in
several
effects
on
these
terminal
hormones
resulting
in
decreased
testosterone,

increased
estradiol
and
some
effect
on
thyroid
hormone
perhaps.

So
they
don't
have
to
be
self
exclusive.
There
could
be
a
common
target
upstream
that
is
affecting
many
hormones.

DR.
ROBERTS:
Anything
else?

Any
follow­
up
questions
or
clarifications
needed?

DR.
ROBERTS:
I
would
like
to
go
ahead
and
take
question
5
and
then
do
a
break.

DR.
STEEGER:
With
regard
to
specific
endpoints,
the
agency
does
not
have
currently
have
sufficient
information
to
quantitatively
relate
gonadal
laryngeal
effects
to
reproductive
outcomes.

A
major
underlying
uncertainty
is
the
ecological
relevance
of
ovotestes
occurrence
to
the
maintenance
of
anuran
populations.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
200
Can
the
panel
provide
sources
of
data
on
background
rates
of
ovotestes
occurrence
in
amphibian
species
and
any
associated
considerations
for
interpreting
this
information
in
the
context
of
the
reviewed
studies?

DR.
ROBERTS:
Our
lead
discussant
on
this
is
Dr.
Green.

DR.
GREEN:
Witchie
reported
sporadic
cases
of
hermaphroditism
early
­­
in
the
late
1950s.

And
there
is
at
least
one
study
in
which
the
prevalence
of
ovotestes
in
a
controlled
population
of
laboratory
frogs
has
been
recently
described.
And
this
is
a
paper
by
Dr.
Kloas
which
is
in
press
right
now.

To
our
knowledge,
however,
the
background
rates
of
ovotestes
in
any
amphibian
population
have
not
been
reported.

The
panel
members
believe
the
frequency
of
occurrence
of
ovotestes
in
normal
healthy
populations
of
amphibians
is
probably
very
low.

This
is
of
course
based
on
the
rare
occurrence
and
our
observations
of
ovotestes
in
our
own
laboratory
animal
populations.

So
without
information
on
the
background
rates
of
ovotestes,
it
is
not
possible
to
assess
the
impact,
if
any,
of
the
presence
of
ovotestes
on
anuran
populations.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
201
DR.
ROBERTS:
Comments
by
other
discussants?

Dr.
LeBlanc.

DR.
LEBLANC:
No
comments.
I
agree.

DR.
KLOAS:
I
agree.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
Yes,
I
think
this
falls
into
the
category
of
data
that
we
had
described
by
the
EPA
panel
before,
which
was
that
in
some
cases
you
will
only
have
anecdotal
data.

None
of
us
ever
thought
it
was
important
to
document
how
many
ovotestes
you
saw
in
opening
up
thousands
of
frogs.
There
are
no
published
data
on
those.

But
if
you
will
take
the
collective
wisdom
of
people
who
have
opened
up
thousands
of
frogs
will
tell
you
that
Witschi
reports
only
one
case,
actually,
in
the
50s
by
somebody
else.
And
I
actually
have
never
seen
one
in
the
laboratory.

DR.
ROBERTS:
Is
this
an
issue
for
which
we
need
­­
is
there
reasonably
clear
understanding
about
what
you
mean
by
ovotestes?

DR.
KELLEY:
I
have
made
a
PDF
file
of
the
Witschi
paper.

It's
actually
a
chapter
from
a
book.
It
is
has
a
very
nice
picture
of
a
mature
gonad
with
clear
testicular
tissue
and
very
well
yolked
up
eggs.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
202
I
submit
that
that
was
what
I
would
call
an
ovotestis.
And
everything
else
I
regard
as
kind
of
subpar
and
up
for
grabs.
But
when
you
have
an
ovary
that
has
a
testis
and
eggs,
which
are
pretty
unmistakable,
that's
what
I
would
call
it.
And
this
is
an
important
issue,
the
terminology
issue
is
very
important.

The
question
of
whether
the
occasional
oocyte
in
a
testis
is
normal
or
not,
Witschi
himself
in
this
­­
for
you
guys
who
don't
know
who
Witschi
is
­­
was
he
German
or
Swiss?
I
believe
he
was
Swiss.

He
was
Swiss,
and
he
started
working
on
xenopus.
He
brought
them
in
a
tea
kettle
to
Bozzle
(
ph).
He
performed
almost
all
the
early
experiments
with
Chang
and
McComo
(
ph)
on
sexual
differentiation
in
xenopus
and
was
a
very
gifted
biologist,
I
believe.

He
has
a
very
nice
picture
of
this
in
this
chapter
summarizing
many,
many
years
of
work.

Anyway,
I
have
given
this
to
the
EPA.
I
have
the
paper
with
me.
And
the
book
should
be
available
in
the
libraries.
It
will
be
in
the
bibliography.

DR.
ROBERTS:
Perhaps
it
would
be
useful
since
I
believe
we're
probably
going
to
comment
subsequently
on
the
problems
created
by
lack
of
terminology
that
we
clarify
in
our
response
what
we
mean
by
ovotestes.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
203
DR.
KELLEY:
That's
a
very
good
suggestion.

DR.
ROBERTS:
Any
other
comments
to
add
on
this
one?

Let's
go
to
B
then.

DR.
STEEGER:
Can
the
panel
characterize
any
evidence
that
suggests
that
the
presence
of
ovotestes
in
male
anurans
results
in
reproductive
impairment
via
reductions
in
fertility?

DR.
GREEN:
To
our
knowledge,
there
are
no
studies
that
show
that
the
presence
of
ovotestes
in
male
anurans
results
in
reproductive
impairment.
However,
the
panel
recommends
that
feminized
males
be
included
in
grow­
out
studies
for
the
purposes
of
using
them
in
breeding
experiments
to
test
this
hypothesis.

DR.
ROBERTS:
Other
responses?
Other
panel
members?

And
by
no
evidence,
you
mean
that
there
is
no
evidence,
but
is
there
any
evidence
that
they
do
not?
There
is
just
no
evidence?

DR.
GREEN:
There
is
no
evidence
either
way.

DR.
ROBERTS:
Either
way.
We
should
make
that
clear
in
our
response.

Dr.
Skelly.

DR.
SKELLY:
This
may
be
the
most
appropriate
point
to
make
this
comment.
I
was
glad
to
hear
that
Dr.
Green
mentioned
reproductive
behavior.
Because
I
think
fertility
is
only
part
of
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
204
story.

I
think
it
is
going
to
be
very
important
if
we're
interested
in
ecological
relevancy
to
see
what
these
genetically
male
and
phenotypically,
whatever
they
are,
how
they
behave
when
they
are
given
the
opportunity
to
mate.

And
I
see
three
broad
categories
of
possibilities
at
the
population
level.
One
is
that
there
may
very
well
be
no
effect.
And
I
think
someone
else
mentioned
this
earlier.
If
you
remove
a
few
males
from
a
population
of
frogs
out
in
nature
­­
I
mean,
ecologists
make
jokes
about
how
males
are
superficial
all
the
time.

DR.
KELLEY:
We
make
jokes
about
that
too.

DR.
GREEN:
I
left
those
out.

DR.
SKELLY:
Thank
you.
It
may
not
take
very
many
many
males
to
keep
a
population
going.

This
is
a
group
of
organisms
for
which
minority
of
males
in
many
populations
may
routinely
do
most
of
the
matings.

However,
it
may
be
possible
if
these
animals
that
have
developmental
abnormalities,
if
they
actually
behave
as
males
but
are
not
fertile
and
they
convince
females
that
they
are
fertilizing
their
eggs
and
leave
a
bunch
of
rotting
unfertilized
eggs
around,
that
could
be
a
serious
population
level
effect.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
205
On
the
other
hand,
if
these
developmentally
abnormal
genotypic
males
actually
function
and
act
as
females,
you
could
increase
the
population
size.
You
could
increase
viability
of
populations.

I
don't
know
how
that's
going
to
work
genetically.
Maybe
Dr.

Kelley
can
comment
on
that.

But
I
see
those
three
broad
categories.
And
I
don't
see
any
way
of
figuring
out
how
to
get
towards
the
broader
goals
of
ecological
relevancy
that
the
EPA
set
out
without
looking
at
this
in
a
field
context
on
native
species.

DR.
ROBERTS:
Perhaps
our
response
could
include
that
caution.

Anything
else,
Dr.
Gibbs
and
then
Dr.
Matsumura.

DR.
GIBBS:
Just
another
body
of
evidence
that
pertains
to
this
particular
question
is
that
interspecies
comparisons
have
shown
that
testes
size
and
sperm
production
are
positively
correlated.

So
it
stands
to
reason
that
within
the
individual,
any
structure
such
as
­­
over
that
reduce
the
size
of
testes
would
reduce
the
sperm
production.

I
don't
know
if
that's
too
much
of
an
extrapolation
to
make.
But
there
is
that
correlation
there.

DR.
ROBERTS:
Perhaps
we
should
mention
that,
then,
with
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
206
some
appropriate
citation
in
our
response.

DR.
MATSUMURA:
I
think
it
is
pretty
important
to
at
least
test
in
the
lab
that
those
individuals
with
the
abnormality
would
indeed
be
reproductively
successful
or
not.

But
I
was
just
thinking
how
would
you
do
that.
You
can't
do
any
invasive
method
to
say
which
one
was,
really
had
discontinuous
ovotestes.

How
would
you
design,
Dr.
Green
or
Dr.
Kelley?

DR.
ROBERTS:
Dr.
Green.

DR.
GREEN:
I
have
not
actually
tested
this
myself,
but
I
have
been
to
amphibian
workshops
where
I
have
seen
this
done.

Some
amphibians
are
small
enough
that
as
you
know
you
can
see
through
them.
If
you
hold
them
up
to
the
light,
you
can
see
internal
organs.

One
method
would
be
to
ultrasound
the
animal's
abdomen
because
there
are
ultrasound
probes
now
that
have
been
miniaturized
for
the
purpose
of
use
in
mice.
So
that's
a
possibility
that,
if
you
are
good
enough,
you
might
be
able
to
detect
gonadal
structures
with
an
ultrasound
probe.

Another
possibility
would
be
minimally
invasive
endoscopic
techniques.
There
are
endoscopes
now
that
have
been
miniaturized
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
207
for
mice,
again,
that
would
be
a
small
puncture.
You
could
go
in
and
look
and
suture
the
animal
up.
Very
similar
to
harvesting
oocytes.

They
would
probably
recover
just
fine
and
go
on.
So
those
are
two
methods.

And
then,
of
course,
fancier
methods
would
be
MRI
and
that
sort
of
thing,
and
which
they
now
make
coils
that
are
small
enough
for
mice,
so
I'm
sure
they
would
easily
accommodate
xenopus
laevis,

which
would
not
require
sacrificing
the
animal,
any
of
those
three
test
methods.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
I'm
going
to
make
my
pitch,
although
not
entirely
appropriate
now,
for
using
known
ZZ
individuals.

If
you
ran
this
experiment
in
animals
where
the
offspring
of
sex
converted
females
that
had
been
phenotypically
female
because
they
have
been
grown
up
in
estrogen,
and
males,
all
of
their
offspring
are
male,
so
then
you
know
that
you
have
nice
sibs
to
compare
to,
and
then
you
can
actually
run
a
very
well­
controlled
experiment
under
those
conditions.

And
you
can
study
not
only
testicular
development
invasively
and
noninvasively,
but
you
can
also
study
clasping,
which
is
the
major
reproductive
behavior
related
to
fertilization,
and
also
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
208
courtship
song,
whose
importance
in
xenopus
for
successful
reproduction
is
not
clear,
but
is
a
very
good
marker
for
sexual
differentiation.

DR.
ROBERTS:
I
think
what
I
have
heard
is
our
panel's
response
that
there
is
no
definitive
information
one
way
or
another
regarding
ovotestes'
effects
as
an
impairment,
perhaps
some
basis
to
speculate
that
it
might,
and
some
discussion
of
challenges
associated
with
running
those
kinds
of
tests
to
provide
the
answer
to
the
question
and
some
suggestions
as
to
approaches
and
some
discussion
about
difficulties
in
setting
up
that
kind
of
test
and
interpreting
it.

Any
follow­
up
questions
or
clarifications
needed?
Let's
go
to
the
next
one.

DR.
STEEGER:
Reduction
in
laryngeal
muscle
area
suggests
diminished
testosterone
in
males.
If
this
is
found
to
be
a
valid
observation
and
if
estrogen
concentrations
do
increase
as
testosterone
concentrations
decrease,
what
other
endpoints,
for
example,

secondary
sexual
characteristics
and
reproductive
behavior
would
likely
be
affected?

DR.
ROBERTS:
Dr.
Green
again.

DR.
GREEN:
The
panel
came
up
with
a
list
of
ten
additional
estrogenic
biomarkers.
The
first
five
that
I
have
here
are
not
invasive
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
209
and
would
not
require
sacrificing
the
animal
and
are
easy
to
identify
phenotypically.

Obviously,
the
measurement
of
snout
to
vent
length
and
body
weight
in
feminized
males
should
be
bigger
than
the
control
males.

As
Dr.
Hayes
pointed
out
yesterday,
nuptial
pads,
the
presence
or
absence
or
diminishment
of
nuptial
pads
and
enlargement
of
the
ventral
folds
of
the
cloaca
strengthen
the
pattern
of
the
male
calling
signal.

Dr.
Kelley
has
some
really
nice
recordings
if
anybody
maybe
wants
to
hear
them
during
the
break.
We
were
impressed
with
how
reproducible
and
how
subjectively
these
might
be
measured.

We
also
were
impressed
by
the
fact
that
if
you
want
to
correlate
loss
of
function,
potentially
the
shrinking
of
the
larynx
muscles,
with
the
morphological
findings,
then
this
would
be
a
good
way
to
show
that
not
only
is
the
larynx
small,
it
is
not
functioning.
You
could
do
that
by
recording
their
calling.

Last,
of
course,
like
Dr.
Kelley
just
mentioned,
clasping.

There
were
some
additional
biogenic
or
biomarkers
that
people
suggested.
One
of
them
we
have
talked
about
at
length
already.
I
won't
go
into
it.
But
it
is
a
time
course
examining
the
synthesis
or
the
presence
of
vitellogenin
in
response
to
an
estrogen
challenge.
And
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
210
then,
of
course,
checking
oviduct
development.

There
are
proteins
expressed
in
xenopus
laevis
in
the
harderian
glands
around
the
eyes.
Three
different
proteins
are
expressed
by
those
animals
that
are
uniquely
female
and
just
one
protein
by
the
males.

Also,
the
number
and
the
size
of
muscle
fibers
in
the
larynx
and
myosin
expression
in
the
larynx
muscle.

And
last,
we
have
had
some
additional
discussions
on
this
one,

but
it
seems
like
seminal,
I'm
not
sure
seminal
fluid
analysis
is
the
appropriate
term
in
amphibian,
but
analysis
of
the
sperm
in
some
of
these
feminized
males
would
be
important
to
look
at.

For
example,
are
they
morphologically
normal
when
they
mature.
Do
they
have
normal
motility.
Is
the
fluid
they
are
found
in
normal.

So
there
would
be
an
additional
assessment
of
the
fertility
in
some
of
these
feminized
males.

DR.
ROBERTS:
Comments
by
other
panels
members?

Dr.
Heeringa.

DR.
HEERINGA:
I
would
like
to
make
a
statistician's
comment
on
the
use
of
the
laryngeal
muscle
or
some
other
sort
of
continuous
measure
of
masculinity
as
distinct
from
other
endpoints
such
as
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
211
gonadal
abnormalities.

Looking
at
the
data,
I
think
­­
and
secondarily
to
these
other
measures,
I
think
having
a
continuous
endpoint
or
continuous
measure
on
a
characteristic
that
appears
to
be
and
demonstrated
to
be
related
to
testosterone
levels
in
the
male
frog,
I
think
would
be
important
as
sort
of
a
secondary
confirmation.

It
may
actually
get
a
little
further
ahead
in
looking
at
mode
of
action
than
just
at
whether
there
is
an
effect.
But
I
think
to
the
extent
that
this
was
not
costly
or
disruptive
to
measure
when
these
animals
are
sacrificed,
I
think
that
it
would
add
value
to
the
experimental
data.

DR.
ROBERTS:
Other
comments.
Dr.
Kelley.

DR.
KELLEY:
As
I
have
pointed
out
before,
the
measurement
of
cross­
sectional
area
of
laryngeal
muscle
is
a
function
of
two
properties,
the
size
of
the
muscle
fibers
and
the
number
of
the
muscle
fibers.

Now,
both
reflect
the
history
of
exposure
to
androgen.
So
the
two
of
them
together
are
some
kind
of
indication
of
the
history
of
exposure
to
androgen.

However,
let
me
just
point
out
that
while
the
size
of
the
muscle
cross­
sectional
area
is
not
very
well­
documented
developmentally
in
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
212
the
literature,
the
number
of
muscle
fibers
is
very
well­
documented.

And
becomes
statistically
significantly
different
between
males
and
females
at
the
stage
which
we
call
PM
1,
which
we
have
also
characterized.

I
want
to
point
out
that
one
of
the
advantages
of
xenopus
are
the
very
well­
characterized
metamorphic
stages,
which
allow
standardization
of
experiments,
so
that
when
we
talk
about
stage
56
animal,
we
know
what
we're
talking
about.

We
carried
out
a
similar
set
of
study
for
post
metamorphic
development.
Because
the
development
of
the
larynx
is
largely
post
metamorphic.
The
brain
is
premetamorphic.

So
we
have
those
standardized
data
for
numbers
of
muscle
fibers.
You
can
do
it
in
the
paraffin
section
and
you
can
replicate
previous
studies.

So
that
will
provide
­­
it
is
a
continuous
variable
because
a
number
of
fibers
is
distributed.
But
the
data
are
very
clear
on
when
those
are
expected
to
become
different
and
at
what
time.
And
it
should
provide
a
sensitive
marker.

DR.
ROBERTS:
Dr.
Heeringa
and
then
Dr.
Richards.

DR.
HEERINGA:
Thank
you
very
much,
Dr.
Kelley.
I
yield
to
the
experts
on
the
exact
nature
of
the
continuous
measurement.
I
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
213
think
that's
an
excellent
contribution.

My
point
is
I
think
we
would
like
to
add
this
continuous
measurement,
something
that
is
differentiated
in
terms
of
its
outcome
potentially
at
least
in
terms
of
observation
from
the
abnormal
gonadal
development.

DR.
KELLEY:
The
point
I
forgot
to
make
was
that
in
the
model
that
EPA
wants
to
develop
for
xenopus,
you
are
going
to
terminate
your
experiments
at
stage
66.
And
there
is
variability
in
the
results
that
we
have
been
presented
here
as
to
whether
there
is
a
sex
difference
in
laryngeal
cross­
sectional
area
at
stage
66.

We
don't
get
a
difference
in
weight.
I
never
measured
a
cross­
sectional
area.
Dr.
Hayes
does
get
a
difference
in
cross­
sectional
area.
Other
people
don't.

If
you
are
going
to
do
a
grow
out,
the
shortest
period
you
should
grow
out
is
three
months
when
the
animals
are
PM
1.
I
refer
to
Tobias,
et
al.,
1991
A,
for
a
description
of
those
experiments
and
these
stages.

DR.
ROBERTS:
'
91
A?

DR.
KELLEY:
The
first
was
the
stages.
The
second
one
was
when
during
those
stages
the
processes
are
hormone
sensitive.

Developmental
biology.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
214
DR.
ROBERTS:
Great.
Dr.
Richards.

DR.
RICHARDS:
I
wanted
to
echo
that
the
development
of
some
of
these
measures,
particularly
Dr.
Kelley
mentioning
the
vocalization,
the
sounds
related
to
the
laryngeal
muscles
and
clasping
behavior,
these
would
be
really
strong
links
to
then
begin
looking
at
real
population
level
studies.

If
there
is
some
concrete
analyses
and
relationships
that
can
be
developed
there,
that
would
be
particularly
strong
for
the
next
step
of
studies.

DR.
ROBERTS:
Dr.
Gibbs.

DR.
GIBBS:
One
quick
comment.
I'm
a
little
bothered
with
what
seems
to
be
an
implicit
assumption
that
bigger
is
better
when
it
comes
to
laryngeal
muscle
area.
Because
females
are
cuing
in
all
sorts
of
qualities
to
male
calls.
Not
simply
volume
and
repetition.

I
just
don't
think
it
is
necessarily
that
simpler
a
relationship
that
larger
laryngeal
muscle
area
corresponds
to
greater
mating
success
via
the
effects
on
the
vocalizations
of
these
males.

DR.
KELLEY:
That's
certainly
true
in
rana,
which
are
highly
discriminating
animals.

In
xenopus,
how
can
I
put
it,
in
xenopus,
unless
you
have
a
male
number
of
muscle
fibers,
you
don't
call
at
all.
So
if
you
make
a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
215
genetic
female
have
a
male
number
of
muscle
fibers
and
have
a
male
size
of
muscle
fibers
and
she
has
circulating
androgen,
this
is
a
testicular
transplant
study,
she
will
call.

There
is
very
good
­­
I
can
tell
you
how
much
circulating
androgen
you
need
to
have
in
a
male
for
him
to
call.
It
is
a
central
effect.
Once
his
larynx
is
masculinized,
it's
masculinized
forever.

But
what
I
can't
tell
you
and
what
nobody
knows
in
xenopus
is
whether
the
calling
male
gets
the
females.

This
is
well­
known
in
rana
where
you
can
do
phono
taxas
experiments
in
the
ponds
in
South
Africa.
Finding
out
whether
the
calling
male
gets
the
female
has
not
yet
been
established.

So
unfortunately,
the
tractable
experimental
prep
is
not
the
prep
for
which
we
have
those
same
kinds
of
data
that
you
have
in
rana,
which
is
­­
I'm
not
going
to
call
them
a
more
sensitive
system,

but
in
some
ways
a
more
subtle
system.

DR.
GIBBS:
But
laryngeal
area
may
not
well
pertain
to
being
a
success
in
rana.

DR.
KELLEY:
That's
perfectly
possible.
And
the
other
thing
is
of
course
in
rana,
you
guys
don't
talk
about
female
calls,
but
in
xenopus,
females
have
two
calls.
They
have
an
acoustic
aphrodisiac
call
that
drives
a
male
nuts,
and
they
have
a
turn­
off
call.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
216
So
they
have
a
very
highly
developed
vocal
system
which
they
use
in
their
social
behaviors
beyond
sex
­­
centered
around
sex,
but
getting
a
little
bit
further
out.

So
these
are
actually
rather
different
vocal
systems.
The
xenopus
one
is
much
more
complex
in
terms
of
call
number
than
the
rana
one.
And
you
are
absolutely
right
about
ranae.
We
don't
know
the
relationship.

DR.
ROBERTS:
If
you
think
it
is
important,
Dr.
Gibbs,
we
could
perhaps
include
the
caution
about
extrapolation
of
the
findings
to
reproductive
success.

Dr.
Skelly,
did
you
want
to
add?

DR.
SKELLY:
No.
That's
what
I
­­

DR.
ROBERTS:
We
can
certainly
put
that
in
our
report.

Anything
else
to
add
on
this
one?
If
not,
let's
go
ahead
and
take
a
break.
It's
3
o'clock.
Let's
reconvene
at
about
3:
15.
We
are
miraculously
ahead
of
schedule.

(
Thereupon,
a
brief
recess
was
taken.)

DR.
ROBERTS:
Let's
go
ahead
and
get
started.
Before
we
take
the
next
question,
just
before
lunch
Dr.
Kelley
promised
to
assemble
a
bibliography.
This
was
in
response
to
Question
1,
I
believe,
and
some
literature
that
we
could
perhaps
recommend
to
the
agency.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
217
She
has
put
that
together.
That
has
been
distributed
to
the
panel.
I
would
just
ask
each
panel
member
to
go
over
the
list
to
see
whether
or
not
you
agree
with
it.
If
you
have
papers
that
you
might
want
to
add
to
it
and
get
back
with
Dr.
Kelley
on
that.

DR.
KELLEY:
I
apologize.
They
are
slightly
out
of
order
and
jumbled,
but
I
didn't
have
very
much
time.
If
there
are
papers
that
aren't
on
it,
just
put
the
papers
on
it
and
I'll
add
them.
If
there
are
papers
you
want
me
to
take
off,
I'll
take
them
off,
except
if
they
are
my
papers,
in
which
case
that's
nonnegotiable.
They
stay
on.

DR.
ROBERTS:
Let's
go
ahead
to
Question
6.

DR.
STEEGER:
While
some
of
the
available
data
suggests
that
there
may
be
an
association
between
atrazine
exposure
and
developmental
effects
in
amphibians,
the
agency's
evaluation
of
the
existing
body
of
laboratory
and
field
studies
has
determined
that
there
is
not
sufficient
scientific
evidence
to
indicate
that
atrazine
consistently
produces
effects
across
the
range
of
amphibian
species
examined.

However,
the
current
body
of
knowledge
has
deficiencies
and
uncertainties
that
limit
its
usefulness
in
assessing
potential
developmental
atrazine
effects
and
the
extent
of
any
associated
cause
effect
in
dose
response
relationships.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
218
Consequently,
the
agency
has
determined
that
there
are
not
sufficient
data
to
reject
the
hypothesis
that
atrazine
can
cause
adverse
developmental
effects
in
amphibians.

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

DR.
ROBERTS:
Dr.
Delorme,
do
you
concur
with
that
conclusion?

DR.
DELORME:
What
I'm
going
to
do
is
I'm
going
to
actually
go
through
the
conclusions
and
break
it
down.
I
tried
to
put
it
through
chromatic,
and
my
computer
crashed.

The
first
conclusion
that
I
pull
out
is
that
there
is
not
sufficient
scientific
evidence
to
indicate
that
atrazine
consistently
produces
effects
across
the
range
of
amphibian
species
examined.
I
had
to
agree
with
the
conclusion.

We
as
a
panel,
I
think,
already
concluded
that
atrazine
could
produce
or
might
produce
effects
on
gonadal
development.
However,

the
consistency
of
the
response
across
a
species
studied
was
difficult
to
asses
because
of
the
problems
identified
with
respect
to
the
design
and
conduct
of
both
the
laboratory
and
the
field
study.
That
confounds
their
interpretation.

The
second
conclusion
is
the
current
body
of
knowledge
has
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
219
deficiencies
and
uncertainties
that
limit
its
usefulness
in
assessing
potential
developmental
atrazine
effects.
I
interpreted
this
as
a
risk
assessor
as
meaning
limits
its
usefulness
in
a
risk
assessment
contact.

Certainly,
I
don't
think
I
would
want
to
conduct
a
risk
assessment
with
the
data
that's
been
presented.
So
I
agree
with
that
conclusion.

Further,
EPA
needs
to
have
results
from
studies
done
where
other
factors
can
be
ruled
out
as
a
cause
in
either
the
presence
or
the
absence
of
effects.
I
think
that's
one
of
the
key
things
that
you
guys
have
brought
out.

You
need
studies
that
are
done
where
it's
unequivocal
that
atrazine
is
the
route
cause
of
the
effect.
For
example,
you
need
good
husbandry
in
the
lab
studies,
good
design
for
both
the
field
and
lab
studies
and
some
of
the
other
factors
that
have
been
discussed
in
the
other
questions.
I
think
that's
key
in
gaining
the
data
you
need
to
do
the
risk
assessment.

Another
conclusion,
it
was
stated
as,
and
the
extent
of
any
associated
cause
effect
and
concentration
response
relationship.
I
think
we
had
already
agreed
that
from
the
data
that
has
been
presented,
we
can't
say
anything
about
the
exact
nature
of
the
response,
either
the
shape
of
the
dose
response
function
or
thresholds
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
220
or
whatnot.

We
just
can't
characterize
it
at
this
point
from
the
data
that
has
been
presented.
It's
recognized
­­
this
is
a
necessary
element
to
conduct
a
risk
assessment.
You're
going
to
need
to
have
some
sort
of
idea,
whether
it's
a
threshold
response,
if
it's
a
dose
response.
What
is
the
nature
of
the
function
of
the
dose
response
in
order
to
conduct
a
risk
assessment.

Consequently,
the
agency
­­
this
is
the
next
conclusion.

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

We
agree
that
with
the
available
­­
we
agree
with
EPA
that
the
available
data
does
suggest
that
atrazine
can
affect
amphibian
gonadal
development.
However,
the
available
data
does
not
allow
for
a
proper
characterization
of
the
nature
and
magnitude
of
the
response,
nor
does
it
offer
sufficient
support
for
the
identification
of
a
plausible
mechanism.

I
guess
in
the
end,
if
you
add
it
all
up,
we
agree
with
the
conclusions,
or
at
least
I
do
and
Joel.
We
wrote
this
together.

Joel,
did
you
have
anything
to
add?

DR.
ROBERTS:
That
sort
of
answers
my
next
question,
but
I'll
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
221
let
Dr.
Coats
respond.

DR.
COATS:
I
concur
with
the
opinion
there
since
it
includes
some
of
my
ideas.

One
other
comment.
We
have
spent
over
this
time
period
and
through
17
or
so
studies
an
awful
large
amount
of
time
picking
them
apart
and
looking
at
every
detail.

On
the
other
hand,
they
do
really
constitute
the
body
of
what
we
do
know
about
this.
There
is
a
lot
we
don't
know
yet.
And
I
wanted
to
make
the
point
they
all
have
made
contributions
toward
the
progress
here
and
the
understanding
of,
is
it
a
problem,
is
it
not,
or
how
big
a
problem
is
it
or
not.

That's
one
thing
we
should
acknowledge.

DR.
ROBERTS:
Dr.
Richards,
did
you
have
anything
to
add?

DR.
RICHARDS:
That
I
concur
with
the
statements
that
the
two
previous
speakers
have
made.
I
have
nothing
more
to
add.

DR.
ROBERTS:
I
think
Dr.
Isom
had
a
question.

DR.
ISOM:
I
have
a
question
for
clarification.
What
do
you
mean
by
adverse
developmental
effects
as
opposed
to
just
developmental
effects,
which
we
have
been
talking
about
previously?

DR.
BRADBURY:
Good
question.
And
it,
I
think,
ties
back
to
the
some
of
the
previous
questions
in
terms
of
the
responses
that
have
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
222
been
described
in
the
literature
thus
far
in
the
context
of
the
risk
assessment
endpoint,
which
is
getting
at
issues
of
reproductive
fitness.

So
again,
had
we
probably
polished
the
question
a
little
better,

it
would
have
been
clearer,
but
in
the
context
of
these
endpoints,

measures
of
effects
in
the
context
of
the
risk
assessment
endpoint
that
we
laid
out
in
the
problem
formulation.

So
to
the
extent
these,
for
example,
gonadal
abnormalities
could
be
related
to
male
fertility
or
reproductive
fitness
measures.

DR.
ISOM:
With
that
definition,
would
that
change
then
the
discussion
we
just
had?

DR.
ROBERTS:
Does
anyone
feel
that
that
would
change
the
response
based
on
­­
change
the
panel's
answer
based
on
the
response
by
Dr.
Bradbury?

Dr.
Green.

DR.
GREEN:
Would
you
repeat
that
one
more
time?

DR.
BRADBURY:
I
should
pull
out
the
problem
formulation.
I
think
we're
all
tracking,
but
we
should
make
sure.

That
abnormalities
that
have
been
described
in
the
literature
for
an
ecological
risk
assessment
then
need
to
be
connected
to
the
measures
of
effects,
the
risk
assessment
endpoint
in
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
223
environmental
management
goal.
The
environmental
management
goal
was
maintain
or
viability
of
anuran
populations,
the
risk
assessment
endpoint
being
connected
to
reproduction
and
fitness
of
populations,
and
the
measures
of
effects
being
connected
to
those
developmental
processes
that
go
on
in
amphibians
that
are
related
to
their
ability
to
reproduce
successfully
and
then
hence
maintain
populations.

So
it
is
trying
to
maintain
that
causal
chain.
Does
this
toxicological
effect
move
up
the
levels
of
biological
organization?
If
you
go
back
to
that
slide
I
had
at
the
beginning
of
today,
this
midmorning,
sort
of
the
connection
between
the
effects
at
different
levels
of
biological
organization
connecting
to
the
risk
assessment
endpoint.

DR.
GREEN:
I
guess
in
my
mind
the
question
too
implies
the
difference
between
whether
or
not
the
effect
that
we
see
is
going
to
have
an
adverse
effect.

For
most
of
these
parameters
we
have
been
looking
at,
we
don't
know
yet
because
they
haven't
been
carried
out
in
grow­
out
studies
far
enough.

So
at
this
point,
any
developmental
effect
would
be
just
a
developmental
effect.
And
maybe
we
should
strike
the
word
adverse
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
224
because
we
don't
know
if
it
is
adverse
yet
or
not
unless
it
is
associated
with
mortality.

DR.
BRADBURY:
Right.

DR.
GREEN:
Is
that
fair?

DR.
ROBERTS:
We
can
clarify
that
in
our
response
that
we
interpreted
this
as
being
developmental
effects
and
as
we
have
stated
earlier
at
least
with
respect
to
some
of
these
effects
that
we
don't
know
yet
the
consequences
of
these
observations.

DR.
BRADBURY:
Right.

If
we
go
back
to
some
of
the
earlier
questions,
you
all
were
discussing
sort
of
the
connections
between
changes
in
the
larynx
to
potential
calling
or
other
kind
of
secondary
characteristics,
which
is
part
of
that
discussion
of
what
is
the
causal
link,
what
is
the
toxicological,
the
ecological
pathway
that
we're
addressing.

DR.
ROBERTS:
We
can
put
that,
draft
that,
put
the
caveat
in
there.

DR.
DELORME:
The
way
it's
written
now
is
actually
kind
of
broad.
Because
what
I
said
is
the
available
data
does
not
allow
proper
characterization
of
the
nature
and
magnitude
of
the
response.

What
we're
saying
is
we
don't
know
how
far
it
is
going,
but
I
can
amend
that
to
say
at
the
organism
or
population
level,
if
that's
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
225
agreeable.

DR.
ROBERTS:
Sure.
Does
anyone
disagree
with
the
statements
that
have
been
made
so
far?

I
don't
see
any.
Any
clarifications
or
follow­
up
questions
from
the
agency?

Let
me
clarify
one
thing
for
the
audience.
From
time
to
time
now
as
the
panel
have
given
their
responses,
they
have
indicated
that
they
have
worked
with
somebody
else
on
the
panel
in
terms
of
writing
something
up.

Under
the
Federal
Advisory
Committee
Act,
individual
panel
members
can
talk
to
other
panel
member
is
they
have
responsibility
on
a
same
topic,
and
sort
of
discuss
the
issues.

But
I
want
to
make
very
clear
that
the
panel
has
not
met
and
undergone
any
deliberations
other
than
in
this
room
in
open
session.

There
have
been
discussions
among
panel
members
during
breaks
and
in
the
evening,
that
sort
of
thing.
As
individuals,
that's
allowed,
but
there
has
not
been
any
closed
session
of
this
panel
to
deliberate
any
of
these
issues.

Let's
go
Question
7.

DR.
STEEGER:
Assuming
the
agency
determined
an
ecological
risk
assessment
with
a
greater
degree
of
certainty
concerning
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
226
developmental
effects
of
atrazine
on
amphibians
were
needed,
please
comment
on
EPA's
conclusion
that
additional
information
is
required
to
evaluate
potential
causal
relationships
between
atrazine
exposure
and
gonadal
development.

Please
also
comment
on
the
added
utility,
if
any,
of
additional
information
to
interpret
the
shape
of
dose
response
curves
for
potential
developmental
endpoints
and
the
extent
to
which
threshold
or
nonthreshhold
response
relationships
can
be
quantified.

DR.
ROBERTS:
Dr.
Delorme.

DR.
DELORME:
I'm
going
to
handle
the
first
part
of
this
question
and
let
Dr.
Coats
handle
the
part
about
the
utility
of
interpreting
the
shape
of
the
dose
response
curve.

I
think
we
agree
with
that,
the
statement
that
­­
or
EPA's
conclusion
that
additional
information
is
required
to
evaluate
the
potential
causal
relationship
between
atrazine
exposure
and
gonadal
development.

The
relationship
is
there,
we
think.
There
is
tantalizing
evidence
that
something
is
going
on.
But
it
needs
to
be
confirmed.
Or
what
is
being
suggested
needs
to
be
confirmed.
And
we
need
to
characterize
the
nature
of
the
dose
response
function.

One
of
the
tenets
of
the
scientific
methods
is
the
repeatability
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
227
of
experiments.
Certainly,
we
have
had
attempts
made
to
repeat
this
that
have
been
submitted
to
EPA
and
we
have
looked
at.

But
definitely,
we
need
to
firm
up
the
existence
of
the
causal
relationship,
and,
if
it
is
there,
if
it
does
exist,
we
need
to
characterize
the
nature
of
the
dose
response
function.

There
is
also
a
need
to
identify
a
plausible
mechanism.
I
think
one
thing
we
have
to
recognize
is
the
identification
of
the
mechanism
is
important
because
it
can
in
part
aid
in
the
extrapolation
of
the
results
from
the
surrogate
test
species,
the
species
of
concern
in
the
environment.

DR.
ROBERTS:
Dr.
Coats.

DR.
COATS:
I
have
a
few
things
to
add.

The
plausible
mechanism
explanation
needs
to
be
put
forth
with
some
data.

Secondly,
there
needs
to
be
some
similarity
of
data
or
patterns
or
trends
from
several
research
groups
to
show
repeatability
of
the
experiments.

The
dose
response
curves
are
extremely
important
to
the
question
of
any
detrimental
effect
of
any
toxicant
on
an
organism,

regardless
of
whether
the
relationships
demonstrate
a
typical
or
atypical
concentration
response
curve
for
a
given
endpoint.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
228
It
should
be
possible
to
ascertain
the
shapes
of
the
curve
given
enough
concentrations,
enough
replications
and
controlled
conditions.

Repeatability
in
other
labs
should
also
be
feasible
if
the
same
species,
stage,
water
concentrations
and
timing
are
utilized.

Another
point
is
that
studies
on
quantitative
structure
activity
relationships
can
often
provide
information
about
the
mechanism
of
action
as
well
or
provide
rationale
for
the
data
is
generated
from
comparative
testing.

Experiments
that
use
a
series
of
closely
related
compounds,

atrazine,
cyanazine,
propazine,
simazine,
terbutyl
azine
(
ph)
et
cetera,
could
elucidate
patterns
that
can
help
explain
the
interaction
between
the
molecule
and
the
putative
receptor
addressing
the
causal
relationship.

This
approach
seems
to
be
lacking
so
far
and
could
be
valuable
in
the
invivo
test
for
gonadal
development
as
well
as
enzyme
induction
or
MRNA
expression.

DR.
ROBERTS:
Thank
you,
Dr.
Coats.

Dr.
Kelley,
did
you
have
anything
to
add?

DR.
KELLEY:
No.
I
completely
concur.

DR.
ROBERTS:
Anyone
else
have
any
comments?

DR.
GREEN:
I
also
concur.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
229
DR.
ROBERTS:
Thank
you,
Dr.
Green.
Anyone
else
like
to
weigh
in
on
this
one?

Does
anyone
disagree,
I
guess
I
should
ask
that.
I
don't
hear
any
disagreement.

Any
follow­
ups
or
any
clarification
needed?
Let's
then
go
on
to
Question
8,
which
I
will
point
out
is
a
seven­
part
question.

Dr.
Richards
has
got
that
one.

DR.
RICHARDS:
Being
the
nonexpert
that
I
am
on
almost
every
part
of
this
­­

DR.
ROBERTS:
Let's
let
the
agency
go
ahead
and
pose
the
question
to
us.

DR.
STEEGER:
The
agency
has
developed
a
conceptual
model
from
which
to
develop
a
set
of
study
of
protocols
for
evaluating
the
potential
effects
of
atrazine
on
gonadal
development
on
amphibians.

The
agency
has
proposed
a
research
approach
using
focused
empirical
laboratory
studies
based
on
initial
investigations
with
xenopus
laevis
followed
by
selective
confirmatory
studies
with
frog
species
native
to
North
America.

This
is
a
proposal.
It
is
not
set
in
stone.
We
have
­­
as
has
been
indicated
in
a
number
of
follow­
up
questions.
Please
comment
on
the
proposed
sequence
of
the
study
objectives.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
230
DR.
ROBERTS:
Dr.
Richards,
we'll
go
through
these
I
guess
one
at
a
time.
Do
you
want
to
take
the
first
one?

DR.
RICHARDS:
Yes.
I'm
going
to
invite
the
persons
listed
and
not
listed
on
this
question
to
jump
in
on
an
open
discussion
here.

But
on
the
please
comment
on
the
proposed
study
of
sequence
objectives,
I
think
Dr.
Kelley
had
already
and
several
others
have
mentioned
about
the
potential
and
importance
of
some
types
of
field
oriented
studies.

That
was
the
one
comment
that
I
had
on
this
also,
and
not
to
preclude
them
in
a
parallel
track
with
some
of
the
other
laboratory
based
studies.

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

DR.
SKELLY:
Just
to
support
what
Dr.
Richards
just
said,
I
think
it
is
going
to
be
important
to
get
the
field
component
of
any
evaluation
undergoing
as
soon
as
any
of
this
happens
in
part
because
field
studies
for
lots
of
reasons
take
a
long
time
to
set
up
and
get
going.

They
are
not
necessarily
as
money
intensive
as
some
of
the
laboratory
studies,
but
observational
studies
and
experimental
studies
take
time
to
find
locations
as
we
have
seen
in
evaluating
some
of
the
past
data.
That
can
make
a
lot
of
difference
in
how
useful
these
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
231
studies
will
be
when
it
comes
to
interpreting
them.

I
guess
I
just
can't
emphasize
enough
that
I
think
it
is
going
to
be
critical
to
get
that
side
of
things
moving
in
order
to
meet
the
objectives
laid
out
in
the
conceptual
model
by
the
EPA.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
Could
you
remind
me
again,
what
is
the
first
thing
that
you
plan
to
do?

DR.
BRADBURY:
Phase
one
in
the
white
paper.

DR.
KELLEY:
Oh,
yes.

DR.
DELORME:
Test
for
apical
gonadal
effects
of
­­

DR.
KELLEY:
So
I
concur
with
that.
I
would
argue
that
we
should
attempt
to
strictly
replicate
some
of
the
studies
that
show
an
effect,
especially
a
low
dose
effect.

I
accept
the
fact
that
in
a
replication
one
will
often
want
to
add
groups
and
so
on.
But
we
have
a
real
discrepancy
in
the
threshold
for
an
effect
if
in
fact
one
exists.

I
would
argue
for
a
strict
replication,
perhaps
a
replication
of
the
high
dose
and
a
replication
of
the
study
that
did
the
low
dose.

I
think
it
is
worthwhile
knowing
how
reliable
and
repeatable
the
initial
observations
are
before
we
go
forward.

Now,
there
has
been
­­
anyway,
I
could
go
through
it.
But
I
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
232
think
it
is
very
important
to
track
down
the
sources
of
variability
and
the
results
before
going
forward.
Because
suppose
you
were
never
able
to
repeat
any
of
them
or
you
always
repeated
them
and
in
a
much
lower
dose.

It
would
affect
so
strongly
the
next
steps
in
the
sequence
that
I
would
argue
very
strongly
for
starting
with
that.

DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
I
concur
with
the
comments
already
made.

I
would
also
point
out
maybe
it
is
time
saving
and
saving
money
if
you
could
start
again
already
in
parallel
with
doing
some
mechanistic
studies.

For
instance,
there
was
some
claim
there
is
no
interference
with
estrogen
receptors,
with
androgen
receptors.

There
are
some
experience
which
could
be
very
easily
maybe
to
say
there
is
nothing
going
on.
Especially,
I
have
some
concern
maybe
if
there
is
only
a
small
effect
on
thyroid
system,
you
will
not
see
it
on
morphological
­­
looking
just
on
morphological
stages.

There
could
also
be
one
biomarker,
for
instance,
TSH,
to
show
up
any
inhibitory
effect
in
addition
which
could
be
done
in
parallel
for
doing
such
a
developmental
study.

DR.
ROBERTS:
I'm
going
to
jump
in
and
make
a
comment
too.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
233
I
think
that
if
we
do
additional
mechanistic
studies,
if
there
are
variable
aspects,
I
think
we
­­
as
Dr.
Kelley
pointed
out,
I
think
we
have
to
get
a
handle
on
those
first
because
we
have
to
know
when
animals
respond
and
when
they
don't
respond.

Because
we
want
to
be
sure
we're
looking
in
animals
that
are
responding
to
do,
to
see
whether
or
not
the
mechanisms
we
think
might
be
operating
are,
in
fact,
in
place.

I
don't
disagree
that
mechanistic
studies
are
important,
but
I
think
getting
a
real
solid
handle
on
the
phenomenon
and
being
able
to
reproducebly
observe
that,
I
agree
with
Dr.
Kelley,
is
a
first
priority.

Dr.
DeLorme
and
Dr.
LeBlanc
and
Dr.
Denver.

DR.
DELORME:
I
just
wanted
to
concur
with
Dr.
Kelley.
I
think
that
what
has
been
proposed
is
quite
a
logical
sequence.
The
first
thing
you
really
need
to
do
is
confirm
whether
or
not
there
is
any
relationship
between
atrazine
and
the
effect.

And
certainly
what
Dr.
Kelley
had
proposed
was
repeating
the
experiments
to
see
if
you
can
find
that,
but
do
it
under
controlled
conditions
where
the
husbandry
and
whatnot
is
­­
any
factors
that
might
affect
it
are
controlled
for.

Followed
by
looking
in
possibly
native
anurans
from
North
America
to
see
whether
or
not
you
are
going
to
get
the
same
kinds
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
234
effect.
I
think
that's
another
good
logical
step
to
take.

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

DR.
LEBLANC:
From
my
perspective,
the
further
you
go
up
in
that
cascade
of
complexity
that
Dr.
Bradbury
had
up
there
earlier
on,

you
are
reducing
your
chances
of
seeing
the
effect
that
you
are
trying
to
confirm
upfront
and
sort
of
give
you
some
evidence
that
you
should
proceed.

Certainly,
I
wouldn't
do
preliminary
studies
or
initial
studies
at
the
population
level.
I
think
gonad
development
is
the
critical
endpoint
of
interest,
but
I'm
not
sure
I
would
do
it
at
that
level.

I
think
that
perhaps
the
first
experiments
that
I
would
do
would
be
looking
at
some
cellular
response
that
I'm
more
likely
to
see.
And
once
I
have
identified
doses
at
which
that
response
occurs
and
I'm
comfortable
that
that
response
is
occurring,
I
think
I
would
build
upon
that.

And
then
I
would
look
at
tissue
level
effects
in
terms
of
gonadal
development
and
ultimately
effects
of
the
individual
and
the
populations,
is
my
perspective.

DR.
ROBERTS:
Dr.
Denver.

DR.
DENVER:
What
we
really
want
to
know
here,
we
want
to
know
whether
there
are
population
level
effects
ultimately.
Is
that
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
235
correct?

And
that's
something
that
is
going
to
take
presumably
a
while
to
sort
out.

The
next
level,
I
assume,
is
to
understand
what
might
be
fitness
effects
at
the
individual
level.

And
I
wonder
­­
this
actually
is
a
more
general
question
that
goes
beyond
just
simply
atrazine.
But
I
wonder
if
we
can
begin
to
get
a
handle
on
those
fitness,
those
individual
fitness
effects
by
looking
at
individuals
or
individuals
that
are
presumably
intersex
or
have
gonadal
abnormalities
in
the
field,
which
I
think
we
agree
has
been
documented,
and
find
a
way
to
identify
those
individuals
which
we
discussed
a
bit
earlier
perhaps
using
some
MRI
approaches
or
something
of
that
sort.
And
do
the
grow­
out
experiments
and
determine
if
there
are
really
fitness
consequences
of
having
these
intersex
gonads.

Is
that
something
that
we
should
be
doing
now?
As
I
said,
it
goes
to
a
larger
issue,
larger
than
just
simply
atrazine.

And
these
intersex
individuals
may
result
from
contaminants
other
than
atrazine.
But
is
that
something
that
we
want
to
consider
sooner
rather
than
later?

DR.
ROBERTS:
We're
getting
some
interesting
differing
views
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
236
on
sequence
of
or
highest
priorities.

Any
other
panel
members
want
to
weigh
in
on
this?

Dr.
Delorme.

DR.
DELORME:
That
is
actually
an
interesting
thought.
I
never
thought
of
that
before.

But
if
there
was
some
chemical
that
you
could
come
up
with
to
induce
intersex
that
you
knew
for
search
was
going
to
the
happen,
that
might
answer
some
of
the
questions
with
respect
to
fertility
and
possibility
of
effects
at
higher
levels
of
biological
organization.

Instead
of
waiting
to
find
out
what
is
actually
going
to
go
on
and
identify
­­

You
could
do
it
in
parallel
with
atrazine.
Because
certainly
as
we
go
through
­­
as
risk
assessor,
I'm
talking,
as
we
go
through
other
chemicals,
there
certainly
are
going
to
be
cases
where
these
kinds
of
effects
arise.

It
is
always
a
question.
What
is
the
ecological
relevance
of
ovotestes.
I
know
in
the
fish
community
it
is
something
they
are
grappling
with
now.
Because
in
a
lot
of
the
rivers
in
Britain
and
in
the
states,
they
have
identified
fish
populations
where
ovotestes
type
effects
do
occur.
The
so
what
question
is
still
there.
Well,
do
they
contribute
to
the
population
or
does
it
have
a
population
level
effect.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
237
So
that's
actually
an
interesting
take
on
things.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
So
there
is
good
news.
There
is
a
chemical
that
does
it.
In
xenopus,
estrogen
does
it.
It
is
very
well
worked
out.
The
critical
period
is
worked
out.
How
much
ovotestes.
I'm
using
that
word
in
the
sense
in
which
I
used
it
before,
and
I
refer
you
to
the
figure
in
the
Witschi
paper.
It
is
very
well
worked
out.

The
animals
that
were
in
that
paper
were
adult.
They
persisted
until
adulthood.
It
looked
like
it
wasn't
resorbed.
I
would
suggest
that
there
is
no
necessity
to
run
this
as
a
serial
experiment.
I'm
totally
opposed
to
that.

You
run
it
in
parallel.
You
run
it
in
parallel
also
with
beginning
your
ecological
assessments,
because
I'm
in
total
agreement.
Field
work
takes
forever.
Getting
it
published
takes
forever.
Anyway,
just
doing
the
work
takes
forever.

So
I
would
start
the
field
studies
and
I
would
start
a
replication
of
the
previous
studies
on
the
atrazine
effects.
I
would
do
it
in
ZZ
animals.
You
know
what
genotype
you
were
dealing
with.

I
would
do
it
in
a
well­
characterized
population
of
animals
and
I
would
run
a
group
along
with
it
with
estrogen
at
various
time
periods
during
the
defined
critical
window.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
238
And
then
you
have
to
grow
those
animals
out.
It's
a
male.

That's
the
good
news.
It's
only
six
months.
And
then
you
run
those
animals
in
behavioral
and
fertility
tests.
And
then
you
know
if
animals
with
66
percent
ovaries
and
33
percent
testes
or
something
produce
fewer
offspring
than
animals
with
a
different
ratio.

Those
studies
are
actually
fairly
easy
to
do
in
a
well­
defined
controlled
way.
You
can
either
use
natural
behaviors
or
you
can
kill
the
animals
and
mush
up
their
testes
and
fertilize
eggs
with
them.

You
don't
have
to
kill
the
animals.
You
can
just
take
their
testes
out.

DR.
ROBERTS:
Other
thoughts?

Lots
of
different
ideas,
Dr.
Richards,
unfortunately
for
you
to
capture.

DR.
RICHARDS:
I'm
sure
many
of
those
words
will
be
on
to
a
text
file
and
given
to
me.

DR.
KELLEY:
Yes,
after
dinner.

DR.
ROBERTS:
Before
we
move
on
to
B,
let
me
make
sure.
Is
there
any
follow­
up
question
or
is
it
as
clear
as
it
can
be
given
the
fact
that
there
are
some
differences
of
opinion
about
sequences?

DR.
BRADBURY:
I
think
it
would
be
helpful
for
the
dialogue
to
continue
a
bit.
I'm
approaching
this
carefully.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
239
And
with
all
due
respect
to
all
members
of
the
panel,
in
the
context
of
science,
for
a
purpose,
which
is
trying
to
incrementally
improve
our
ability
to
reduce
the
uncertainties
on
the
risk
assessment
and
how
that
marries
up
with
and
is
complimentary
to
advancing
knowledge,
because
it
is
important
to
advance
knowledge,
and
to
the
extent
the
panel
has
any
feel
of,
for
lack
of
a
better
word,
where
you
could
get
the
biggest
bang
for
your
buck
to
start
either
sequence
how
you
would
stage
information
gathering
or
a
sense
of
what
information
we
start
to
clarify
or
reduce
some
of
the
uncertainties
in
the
current
ability
to
go
through
the
risk
assessment
process.

It
is
sort
of
­­
the
dialogue
has
been
interesting.
I
think
Jere
in
thinking
about
the
endocrine
disrupter
screening
assays
and
some
of
the
other
descriptions
about
using
QSAR
and
then
going
from
the
field
down
sort
of
to
illustrate
very
legitimate
differences
of
opinion
in
terms
of
how
you
blend
mechanistic
understanding
with
ecological
relevancy
and
where
are
you
in
that
continuum,
it
seems
like
we're
getting
advice
to
do
everything
all
at
once.

And
while
maybe
that's
the
only
way
we
can
solve
this
problem,

if
that's
the
conclusion,
then
that's
the
conclusion.

I
think
it
would
be
helpful
to
hear
about
what
some
of
the
trade­
offs
would
be
at
least
in
picking
different
places
to
start
in
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
240
phased
approach
that
we
provided.

It
may
mean
you
may
all
feel
that
you
should
start
all
five
phases
all
at
once,
and
that's
cool.
But
it
would
be
helpful
to
at
least
get
input
from
the
panel
in
terms
of
the
trade­
offs
if
one
couldn't
do
all
five
phases
at
the
same
time.

DR.
ROBERTS:
Let's
ask
the
panel.
What
if
you
couldn't
do
them
all
at
once,
where
would
you
start?
Dr.
Kelley
and
then
Dr.

Green.

DR.
KELLEY:
One
in
five.

DR.
ROBERTS:
Beg
your
pardon?

DR.
KELLEY:
That
was
it.

DR.
GREEN:
Testing
the
working
hypothesis
phase
experiment
Number
1,
the
test
for
apical
gonad
effects
and
Number
5,
ecological
relevancy
of
the
study.
I
concur
with
Dr.
Kelley's
comment.

But
since
you
have
solicited
additional
dialogue,
I
have
a
few
more
things
that
I
think
are
imperative
to
standardizing
the
conditions.
And
this
should
be
done
in
Phase
1.
So
right
from
the
get­
go
we
can
do
the
best
we
can.

One
of
the
things
that
struck
me
yesterday
about
the
housing
and
the
husbandry
conditions
was
the
extreme
variability
and
the
fact
that
one
of
the
things
we
were
measuring
was
growth
rate
and
trying
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
241
to
make
a
connection
between
the
effect
of
atrazine
on
growth
rate
and
then
on
the
growth
rate
in
the
control
tank
when
between
labs
and
within
the
lab
there
were
specific
issues
with
water
quality
and
feeding
and
other
things
that
could
account
for
variability
in
growth
rate
alone
whether
atrazine
was
there
or
not.

So
I'm
going
to
take
the
opportunity
to
make
a
plea.
And
part
of
this
bleeds
into
Question
Number
4,
which
I
was
actually
waiting
to
answer,
about
two
issues.

And
the
first
one
is
with
water
quality.
This
is
not
pertaining
to
measuring
levels
of
atrazine.
But
I
think
we
should
make
some
attempt
to
define
the
stocking
density
first
off
for
both
embryos
and
adults
and
juveniles
for
xenopus
laevis.

Unfortunately,
that
hasn't
been
standardized
in
Laboratory
Animal
Medicine
even
yet,
but
there
are
some
recommendations.

The
proceedings
for
the
National
Academy
of
Sciences
has
recommended
one
to
two
liters
for
adults.
And
there
are
various
textbooks
and
original
papers
for
tadpole
stocking
and
density.
I
think
we
talked
about
that
a
little
bit
already.
So
that
would
be
one
thing.

One
thing
that
was
lacking
in
the
descriptions
of
many
of
the
papers
was
the
kind
of
water
that
was
used.
I
couldn't
tell.
Was
this
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
242
deionized
or
R
O
treated
or
reconstituted
salt
water.
Was
it
well
water
from
a
source.
Was
it
chlorinated
or
chloraminated,
potable
tap
water
that
had
been
filtered.

And
I
think
all
those
water
sources
are
perfectly
suitable.
But
it
needs
to
be
stated
and,
if
possible,
standardized
so
that
everybody
who
runs
the
test
on
atrazine
uses
the
same
kind
of
water.

The
other
thing
would
be
water
quality
testing.
For
this
purpose,
because
there
is
evidence
in
the
literature
that
atrazine
and
nitrites
interact
and
that
they
have
some
by­
product
of
ammonium
metabolism,
when
the
ammonia
levels
get
high
in
tanks,
it
may
falsely
elevate
the
detrimental
effects,
if
any,
that
any
other
chemical
has
in
there.
It
could
be
chlorine.
It
could
be
atrazine.

So
water
quality
should
be
measured
on
a
regular
basis.

Depending
on
the
duration
of
the
experiment
and
the
stocking
density
and
the
water
turnover,
I
would
say
once
a
day.
It
could
be
done
with
a
quick
dip
stick
that
is
a
relatively
inexpensive
test.
It
doesn't
have
to
be
the
Hawk
(
ph)
analytical
test
every
day,
but
some
notation
of
water
quality
parameters
on
a
regular
basis.

That
would
include
pH,
conductivity,
water
temperature,

ammonia,
nitrate,
nitrite.
And
because
we
want
to
know
what
kind
of
water
is
used,
we
should
be
measuring
for
chlorine
and
chloramine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
243
and
the
heavy
metals
that
we
know
interfere
with
tadpole
development
like
copper
and
some
of
the
other
things.

So
I
would
like
to
see
that
in
most
of
these
reports
at
least
have
been
part
of
the
quality
control.

So
water
quality
is
one
issue
in
Phase
1
that
I
think
if
we
could
standardize.
And
you
can't
go
to
the
literature
and
find
this.

There
is
a
paper
that
was
published
in
the
Laboratory
Animal
Medicine
Journal
that
reported
a
survey
across
the
nation
of
what
most
xenopus
users
are
doing
right
now
that
you
can
use
as
a
guide.

I
have
the
reference
here
that
I'll
add
to
the
list
for
Dr.
Kelley
on
all
these
parameters.

The
second
issue
that
I
wanted
to
bring
up
was
feeding.
I'm
sure
Dr.
Kelley
will
bring
this
up
as
well.

But
Dr.
Hayes
pointed
out
that
he
felt
that
maybe
some
of
the
animals
in
the
Syngenta
studies
had
been
underfed.
And
I
looked
at
that
­­
I
went
back
and
looked
at
that
paper
and
looked
at
some
of
his
publications
on
what
he
was
feeding.

It
struck
me
that
he
is
feeding
rabbit
chow.
I
find
this
a
little
bit
disconcerting
because
adult
xenopus
are
strict
carnivores.
And
that
is
an
herbivore
diet.

DR.
HAYES:
The
adult
xenopus
were
fed
trout
chow,
not
rabbit
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
244
chow.

DR.
GREEN:
Dr.
Hayes
in
the
back
said
his
adult
xenopus
were
fed
trout
chow.
But
if
we
continue
these
studies
into
grow­
out
experiments
where
juveniles
and
adults
are
used,
then
we
should
probably
feed
them
a
diet
that's
at
least
14
percent
protein,
which
is
more
appropriate
for
a
carnivore.

There
are
references.
Kevin
Wright
has
published
in
another
textbook,
Amphibian
Medicine
and
Husbandry,
a
chapter.
In
that,
he
recommends
that
we
not
feed
adult
and
juvenile
xenopus
laevis,

anyway,
diets
for
omnivorous
fish
and
turtles
or
for
herbivores.

So
you
can
go
to
a
feed
company,
there
are
several
reputable
ones,
that
will
make
xenopus
chow
for
you
or
you
can
buy
it
from
the
big
distributors.
Nasco
and
Xenopus
1
have
their
own
that
they
sell.

I
think
in
an
attempt
to
standardize
across
labs,
that's
probably
the
appropriate
food
for
that
age.

Now,
tadpole
feeding
­­
those
companies
also
sell
food,
if
I'm
correct,
for
younger
metamorphs
and
juveniles
as
well.
And
they
eat
phytoplankton.

I'm
not
so
certain
how
important
it
is
that
they
have
a
high
protein
diet
to
maximize
growth
and
health
under
these
conditions.

But
if
you
are
going
to
make
those
comparisons
in
growth
rates
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
245
and
fecundity,
then
probably
standardizing
to
a
carnivorous
amphibian
diet
would
be
the
best
thing
to
do.

The
companies
also
have
based
those
diets
and
the
amount
of
feed
on
growth
rate
curves
that
they
have
developed.
I
believe
they
will
share
them
with
you.
I
have
asked
for
them
and
they
have
been
very
helpful
in
advising
me.

They
also
make
recommendations
on
how
much
to
feed
per
animal.
That's
based
on
known
anuran
kilocaloric
requirements
at
different
temperatures.

So
if
we
hold
tanks
for
some
of
the
juvenile
and
adults
at
19
degrees
versus
25,
we
should
feed
them
accordingly.
There
are
ways
to
do
that.

If
you
look
to
the
fish
literature,
as
young
fish
grow,
they
are
regularly
weighed
once
a
month
on
mass.
The
tank
feed
is
adjusted
accordingly.

I
know
that
it
is
not
standard
practice
right
now
to
do
that
for
xenopus.
Usually,
you
have
about,
whatever,
5,
50
or
100
in
a
tank
and
you
throw
so
much
food
and
make
sure
they
eat
it
all.

But
in
terms
of
water
quality
level,
what
is
not
eaten
and
the
amount
they
excrete
will
affect
the
amount
of
ammonia,
which
in
turn
affects
the
amount
of
atrazine
that
is
or
isn't
available
or
potentiated.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
246
Those
are
two
conditions,
at
least,
the
water
quality
and
the
feeding
that
I
think
this
Phase
1
should
be
spelled
out
as
best
we
can.

And
the
temperature
that
we
all
agree
would
be
the
right
place
to
start
with
in
conducting
these
experiments.

DR.
ROBERTS:
I
think
it
will
be
important
for
us
to
convey
Dr.
Green's
recommendations
in
our
report.
I
don't
think
it
necessarily
belongs
as
part
of
8
A,
but
we'll
get
it
in
there.
I
think
it
is
important
to
pass
that
information
along
to
the
agency.

With
regard
to
the
question
posed
by
Dr.
Bradbury,
if
you
couldn't
do
them
all
at
once,
what
would
you
do?

Dr.
Delorme.

DR.
DELORME:
I
would
concur
with
Dr.
Kelley
that
probably
the
first
thing
you
want
to
do
is
­­
if
your
goal
is
to
move
the
atrazine
risk
assessment
forward,
then
I
think
you
have
to
confirm
the
causal
relationship.
That's
the
first
thing
you
need
to
do.
That
would
be
the
first
priority
to
do.

DR.
ROBERTS:
Dr.
Coats.

DR.
COATS:
I
think
that
the
laboratory
approach
as
described
in
Phase
1
is
probably
the
most
obvious
and
least
risky
from
a
benefit
­­
from
the
point
of
time
invested
and
money
invested.

And
so
that's
where
I
think
we
should
start.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
247
DR.
ROBERTS:
Dr.
Skelly.

DR.
SKELLY:
At
the
risk
of
boring
everyone,
I'm
going
to
reiterate
that
I
think
pushing
ecological
relevancy
off
to
Phase
5
could
be
a
mistake.

We're
testing
for
apical
gonadal
effects
in
Phase
1.
Then
Phase
2
is
sex
steroid
measurements.
Phase
3
is
aromatase
activity
measurements.
Phase
4
is
aromatase
inhibitor
study.
Phase
5
is
ecological
relevance.

I
think
there
could
be
a
giant
woops
there
if
we
get
to
the
end
and
find
out
that
the
population
level
effects
aren't
what
we're
all
being
concerned
about
here.

DR.
ROBERTS:
Dr.
Delorme.

DR.
DELORME:
Just
as
a
point.
If
you
look
at
figure
one
in
the
white
paper,
it
actually
shows
after
the
test
for
apical
gonadal
effects
and
a
yes
being
found,
there
is
a
line
off
to
the
side
with
Number
5.

Can
EPA
comment
on
what
was
intended
with
that?
It
wasn't
clear
whether
or
not
you
intended
the
ecological
relevance
test
to
start
at
that
point.

DR.
SKELLY:
Is
ecology
fifth
level
priority
or
not?

DR.
TIETGE:
It
just
so
happens
that
we
used
numbers.
We
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
248
could
have
used
colors
I
guess
to
indicate
what
order
to
do
them
in.

Clearly,
as
Dr.
Delorme
pointed
out,
we
had
that
split
in
the
sequence
to
suggest
that
if
you
get
effects
at
the
individual
level,
then
it
may
be
more
prudent
from
a
risk
assessment
point
of
view
to
conduct
some
field
work.

I
wouldn't
read
into
the
ordinal
numbers
here
very
much.

DR.
ROBERTS:
Does
that
help,
Dr.
Skelly?

DR.
SKELLY:
Yes.
If
I
can
follow
up.

I'm
fine
with
that
with
one
important
caveat.
That
is,
we're
going
to
get
to
this
a
little
bit
farther
down.
But
if
the
panel
feels
that
there
are
significant
concerns
about
context
dependence
between
species,
and
at
least
in
my
case
I
think
we
have
seen
some
evidence
that
that
could
be
the
case,
then
I
don't
see
that
a
no
on
apical
gonadal
effects
in
xenopus
necessarily
means
that
we
shouldn't
be
thinking
about
an
ecological
relevance
study
for
native
species.

DR.
ROBERTS:
Dr.
Bradbury.

DR.
BRADBURY:
Just
to
get
a
clarification,
because
the
white
paper
talks
about
using
xenopus
to
get
started
as
a
biological
model
bla,
bla,
bla,
but
also
talks
about
other
species
to
also
be
looked
at
in
the
context
of
phase
one
style
experiments.

I
guess
the
clarification
that
if
xenopus
didn't
reproduce
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
249
previous
results,
would
the
panel
feel
it
would
be
useful
to
try
it
out
in
rana
in
the
Phase
1
style
experiment
or
to
immediately
go
to
a
long
term
reproductive
study
or
go
out
to
the
field
to
look
at
rana
or
other
native
species?

DR.
ROBERTS:
Expression
of
preference
on
that?

Dr.
Kelley.

DR.
KELLEY:
Well,
you
know,
I'm
for
parallel
processing.

Let's
face
it.
We
don't
care
here
in
America
about
the
survival
of
our
xenopus
because
they
are
well
taken
care
of
in
the
lab.

But
we
care
about
the
survival
of
our
native
frog
species.
You
are
just
going
to
have
to
start
the
field
studies
right
away.
But
I
do
agree
that
it
would
be
useful
to
see
if
we
could
run
some
rana
experiments
in
the
lab.

Of
course
the
problem
is
rana
is
a
lot
harder
to
run
in
the
lab.

And
rana
is
about
a
zillion
times
more
difficult
to
work
on
than
xenopus
and
so
forth.

But
that
isn't
to
say
that
it
shouldn't
be
attempted,
because
it
will
be
very
informative.
There
are
folks
that
are
good
at
rana.
I'm
not
one
of
them.
But
I
know
there
are
people
that
are.

It
would
be
worthwhile
to
try
in
a
limited
way
to
see
if
we
could
get
that
to
go.
But
it
is
going
to
be
a
lot
harder.
The
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
250
background
literature
that
is
available
for
xenopus
is
just
not
available
for
rana.

It
will
be
somewhat
more
difficult
to
interpret
the
results.
But
those
are
the
results
we
care
about.
Because
these
are
our
frogs,
our
North
American
frogs.
We
care
a
lot
about
them.

We
should
study
them
along
with
the
indicator
species,
which
has
many,
many
advantages,
but
is
not
native
to
our
country.

DR.
ROBERTS:
Other
thoughts?

Dr.
Richards.

DR.
RICHARDS:
I
concur
with
what
Dr.
Kelley
has
brought
up
here.
I
think
that
ultimately
the
agency
is
going
to
have
to
go
there
no
matter
what
answer
we
get
from
xenopus.
And
it
could
lead
us
down
some
very
good
trails.

Ultimately,
the
question
is
going
to
come
to
is
what
does
that
mean
to
ranid
species.
And
if
we
don't
start
now
in
trying
to
further
hone
or
develop
methods
that
are
appropriate
for
them,
it
is
just
going
to
prolong
the
whole
process.

DR.
ROBERTS:
Dr.
Tietge.

DR.
TIETGE:
I
think
what
we
have
to
keep
in
mind
is
that
using
the
rana
species
or
the
intent
of
using
the
rana
species
is
to
do
the
confirmatory
work.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
251
But
you
do
the
hard
work
I
think
in
xenopus
because
it
is
a
more
useful
model
at
this
point
in
time.
So
I
don't
think
we
meant
one
or
the
other.

I
think
­­
I
agree
with
what
I'm
hearing
here
that
you
probably
will
end
up
doing
it
in
both
species
at
least
in
that
Phase
1
study.

But
the
intent
of
the
xenopus
is
to
go
faster
and
farther
to
understand,
for
example,
mechanistic
pathways
that
you
would
then
go
back
and
in
a
more
focused
approach
look
at
with
the
rana
species.

DR.
BRADBURY:
Just
to
help
clarify
because
you've
been
getting
some
understandably
different
kinds
of
views.

At
one
point
we
heard
we
should
try
taking
account
of
variability
associated
with
some
of
the
current
studies
to
try
to
see
if
we
could
replicate
what
had
happened
in
the
past.

And
all
there
is
a
ranid
study.
There
are
several
xenopus
studies.
That
would
imply
get
started
with
xenopus
to
see
if
you
can
get
it
to
happen
in
xenopus
again.
Start
with
ranid.

Again,
I
know
you
are
probably
thinking
this
through
as
we
go.

But
as
the
panel
deliberates,
it
will
be
helpful
to
get,
if
not
definitive
answers,
at
least
some
thoughts
of
the
cost
benefit,
that's
not
the
right
word,
but
the
ups
and
downs,
the
trade­
offs
associated
with
some
of
the
different
choices
in
the
pathway.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
252
DR.
ROBERTS:
Dr.
Denver
and
then
Dr.
Green.

DR.
DENVER:
Let
me
sort
of
echo
the
comments
of
Dr.

Richards
and
play
devil's
advocate.

What
if
you
are
unable
able
to
replicate
the
xenopus
results
and
you
are
unable
to
show
in
fact
any
effects
of
atrazine?
What
decision
would
you
then
make?

Would
you
then
decide
to
go
on
to
ask
the
question
in
North
American
ranid
species
or
would
you
conclude
that,
in
fact,
there
is
no
basis
for
concern?

DR.
BRADBURY:
I
think
some
of
that
dialogue
­­
some
of
the
issues
that
one
would
face
in
making
that
decision
are
some
of
the
latter
questions
we
have
in
the
series
of
question
eight
which
gets
at
some
of
the
issues
about
toxico
dynamic
and
toxico
kinetic
differences
among
frogs.

And
to
not
dodge
the
question,
I
think
I
would
benefit
greatly
from
the
panel's
dialogue
on
those
issues
that
could
drive
interspecies
extrapolation.

DR.
ROBERTS:
Fair
enough.

We've
had
a
lot
of
dialogue.
I
hope
Dr.
Richards
has
been
keeping
up.

Let
me
see
if
we
can
decide
whether
or
not
the
­­
is
there
some
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
253
sort
of
consensus
in
terms
of
priority
or
are
there
still
differences
of
opinion?

Dr.
Richards,
what
is
your
sense?

DR.
RICHARDS:
I'm
hearing
that
most
of
the
panel
seems
to
feel
comfortable
that
some
of
the
­­
the
Phase
1
experiments
need
to
be
replicated,
techniques
cleaned
up.
That
work
needs
to
go
forward.

I'm
hearing
that
from
most
people.

I
have
heard
from
a
little
bit
smaller
number
that
some
aspects
of
Number
5
need
to
go
forward
in
terms
of
grow
out
or
some
of
the
other
basic
ecological
studies.
Not
necessarily
full
ecological
studies.

DR.
ROBERTS:
I
sort
of
heard
unanimity
in
the
desirability
for
the
Phase
1
studies
to
proceed.
And
some
support
for,
if
possible,

beginning
the
Number
5
ecological
studies,
because
recognizing
that
that
­­
demonstrating
that
relevance
would
be
very
important
for
the
risk
assessment.

Does
somebody
else
have
a
different
take?

Dr.
Gibbs.

DR.
GIBBS:
I
would
concur.
I
thought
Dr.
Kelley
had
outlined
a
sequence
that
is
really
quite
doable
in
a
fairly
short
time
frame
in
terms
of
looking
at
the
fitness
consequences
with
intersex
in
normal
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
254
males
and
then
you
could
actually
put
those
in
a
natural
environment
and
look
at
the
consequences
and
mating
with
maybe
even
molecular
genetic
markers
getting
the
tadpoles
or
metamorphs.

But
I
don't
think
until
there
is
­­
rather
than
a
full­
blown
ecological
study,
100
wetlands,
et
cetera,
we're
not
talking
about
that
necessarily
for
Phase
5.
I
think
just
a
shorter
term
study
like
Dr.

Kelley
outlined
I
think
is
quite
doable.

DR.
ROBERTS:
Does
that
sound
reasonable
to
the
panel?
All
right.
Good.
Then
let's
try
and
have
the
minutes
the
reflect
that
recommendation.

Any
follow­
up
questions
from
the
agency?
Dr.
Green
and
then
Dr.
Isom.

DR.
GREEN:
I
wasn't
clear
now.
Was
the
decision
that
ranA
and
xenopus
experiment
should
be
run
in
parallel,
if
possible?

DR.
ROBERTS:
I
think
we're
going
to
take
up
the
species
­­

DR.
GREEN:
Later
on?

DR.
ROBERTS:
Yes.
As
we
go
a
little
bit
further
down.

Hopefully,
after
we
have
that
discussion,
then
we'll
be
able
to
clarify
our
recommendation
for
that
aspect.

Dr.
Isom.

DR.
ISOM:
I
wasn't
clear
also
with
regards
to
the
first
study.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
255
A
lot
of
data,
mechanistic
data,
can
be
obtained,
related
mechanistic
data
can
be
obtained
from
that
study.
Are
we
just
advocating
looking
at,
say,
anatomical
gross
abnormalities
in
these
animals,
the
survivability,
or
are
we
also
saying
we
should
be
looking
at
blood
levels
of
hormones
and
other
associated
effects
in
those
animals?

If
you
have
those
tissues
of
those
animals,
why
not
get
the
mileage
out
of
them?

DR.
BRADBURY:
8
B
starts
to
get
at
some
of
the
very
issues
you
are
bringing
up.

DR.
ROBERTS:
I
guess
my
initial
response,
Gary,
would
be
to
encourage
doing
superimposed
mechanistic
studies
while
they
are
doing
these.

DR.
ISOM:
It
wasn't
clear.

DR.
BRADBURY:
In
fact,
Question
Eight
B
charged
the
panel,

very
explicitly
asks
the
panel
for
advice
and
counsel
on
the
attributes
of
these
studies.

DR.
ROBERTS:
Then
let's
go
to
Eight
B.

DR.
STEEGER:
Please
comment
on
whether
the
agency's
first
set
of
proposed
studies
has
accounted
for
the
major
sources
of
uncertainty
associated
with
the
potential
effects
of
atrazine
on
anuran
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
256
sexual
differentiation.

In
addition
to
the
time
to
metamorphosis,
gonadal
abnormalities
and
sex
ratios
in
the
proposed
Phase
1
assays,
please
comment
on
any
other
endpoints
that
should
be
considered
in
this
initial
phase.

DR.
ROBERTS:
Dr.
Richards.

DR.
RICHARDS:
No.
The
first
part
I
think
we
pretty
much
elaborated
on,
it
seems
to
me,
in
several
of
the
other
questions.
No,

we
haven't
accounted
for
the
major
sources
of
uncertainty
associated
with
potential
effects.

And
I'll
let
the
discussion
open
up
in
terms
of
the
other
measures
as
Dr.
Isom
and
others
have
initially
brought
up.

DR.
ROBERTS:
Dr.
Green
would
like
to
comment
on
that.

DR.
GREEN:
I
think
I
already
commented
basically
on
the
husbandry
issues.
But
in
terms
of
in
addition
to
time
to
metamorphosis
gonadal
abnormalities
and
sex
ratios,
please
comment
on
any
other
endpoints,
and
we
addressed
that
with
the
10
different
estrogenic
biomarkers.

I'm
not
clear
if
the
goal
here
is
to
get
these
studies
done
quickly
and
in
what
level
of
detail
we
want
to
look
at
these
things.
Because
all
10
of
them
would
be
a
lot
of
work.
The
first
four
or
five,
which
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
257
are
observational
and
easy
to
see,
I
wouldn't
suggest
would
be
the
place
to
start.

DR.
BRADBURY:
I
think
that
kind
of
a
discussion
would
be
helpful.
What
would
be
the
experimental
investments,
time,
effort
associated
with
adding
the
different
endpoints
in,
what
is
the
kind
of
information
that
is
gained
as
one
gets
that
information.

Some
of
the
logic
in
the
different
phases,
probably
isn't
the
right
word,
but
the
components
or
the
trying
to
sequence
things
is
sort
of
an
approach
to
if
we
get
the
frank
apical
effects,
then
­­
part
of
this
question
is
getting
at
are
there
ways
to
maybe
blend
some
of
the
concepts
that
are
in
the
analysis
plan
and
try
to
link
some
things
up
more
quickly.
That's
the
question.

Getting
some
insights
on
what
some
of
the
efforts
would
be
and
some
thought
about
what
would
this
experimental
design
start
to
look
like
I
think
is
important.

DR.
ROBERTS:
Go
ahead
and
follow
up,
Dr.
Green.

DR.
GREEN:
We're
trying
to
decide
here
­­
in
Phase
1
let's
say
we're
going
to
start
with
xenopus
laevis.
And
in
Phase
1
we're
going
to
grow
them
out
to
three
months.
That
was
something
Dr.
Kelley
suggested.
And
I
concur
that
would
be
the
minimum
because
then
we
could
see
some
of
the
secondary
sexual
characteristics
start
to
emerge
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
258
or
not.

And
we
would
look
at
different
stages
and
ages
along
the
way
out
to
three
months.
Is
that
correct?

DR.
KELLEY:
Let
me
just
talk
about
a
moment
the
sex
ratio.

You
are
not
going
to
be
able
to
interpret
the
sex
ratio
unless
you
have
a
uniform
genotype.
Here
is
what
I
would
do.
I
would
get
Nasco
or
Kelley
Evans
at
Xenopus
1
to
establish
for
me
a
stock
colony
for
researchers
to
draw
their
animals
from.

They
will
be
willing
to
do
this.
I
would
like
to
point
out
to
the
audience
that
these
places
are
agricultural
facilities
in
their
home
states.
They
support
the
greater
agricultural
goods.
Xenopus
is
a
farm
product.

Regulate
is
a
farm
product
in
many
states.
We're
growing
frogs
here.
It's
not
corn,
but
frogs.

Anyway,
so
you
have
to
start
with
ZZ
animals,
otherwise
the
sex
ratio
becomes
very
difficult
to
interpret
without
long
breeding
experiments.
It
takes
two
years
to
get
a
female
to
breed.
You
don't
want
that.
It
will
really
slow
you
down.

How
long
will
it
take
to
get
ZZ
animals.
If
you
get
donated
ZZ
animals
that
people
have,
you
would
have
to
confirm
they
really
are
ZZ
by
mating
them
to
normal
males
and
having
all
male
offspring.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
259
That
would
take
three
months.

If
you
have
to
make
them,
that
will
take
six
months.
Three
months
minimum,
six
months
maximum.

Then
you
would
run
your
experiments
in
the
presence
and
absence
of
atrazine
and
various
other
things,
and
you
should
be
able
to
get
data
on
that
in,
I
would
say,
six
months.
That's
a
year.

Nine
months
to
a
year
minimum.
With
things
going
wrong,

that's
12
months
basically
for
sure
and
maybe
18
months.

So
18
months
to
replicate
with
a
known
population
under
defined
growing
conditions.
That's
what
that
would
take
in
a
lab
that
was
up
and
functional.

In
a
lab
that
had
never
dealt
with
frogs
before,
though,
I
would
like
to
point
out
that
it
will
take
longer.
If
you
contract
this
out
or
something
to
people
who
have
never
raised
xenopus,
it
will
take
longer
because
there
is
a
learning
curve,
learning
how
to
keep
the
animals
and
keep
them
happy
and
so
forth.
Just
catching
them.

Learning
to
put
lids
on
so
they
don't
hop.
These
are
major
things.

Is
that
the
sort
of
thing
you
want?
That's
how
long
it
is
going
to
take
just
with
xenopus.

DR.
BRADBURY:
In
the
white
paper
we
laid
out
attributes
that
may
or
may
not
be
consistent.
Some
may
be,
some
probably
not
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
260
consistent
with
what
you
just
described.

On
the
screen,
we
have
laid
out
some
of
the
details
of
that.

DR.
KELLEY:
I
disagree
deeply
with
some
of
the
details.

DR.
BRADBURY:
Again,
it
was
a
plan
to
get
this
kind
of
dialogue
going.
I'm
trying
to
figure
a
way
see
if
I
can
get
the
chairman
to
help
us
sort
of
systematically
explore
some
of
these.

DR.
ROBERTS:
I
think
there
is
lots
of
issues.
In
CS
for
range
and
spacing
and
number
of
concentration,
some
of
those
things
are
up
here
­­
you
are
looking
for
feedback
on
what
should
be
done
in
these
experiments.
And
I
think
we
can
go
sort
of
go
through
those
aspects.

I
think
your
first
question
was
in
B
is
what
endpoints
should
we
look
for.
And
so
let's
answer
that
question.
And
then
we
can
go
on
to
provide
you
with
feedback
on
other
aspects
of
the
experimental
design.

You
asked
for
a
little
bit
of
dialogue
for
us
in
terms
of
if
you
did
this,
it
adds
this,
but
it
costs
this
kind
of
thing.
Not
only
in
terms
of
dollars,
but
just
to
give
you
some
sort
of
feedback
in
terms
of
what
you
get
for
what
extra
effort
and
resources
are
invested
in
terms
of
adding
different
endpoints.

If
I
understood
you
correctly,
Dr.
Bradbury,
that's
the
kind
of
advice
they
are
seeking
from
the
panel
right
now
as
part
of
this
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
261
particular
question.

So
if
we
have
some
suggestions
for
them
in
addition
to
the
endpoints
that
are
identified
here,
some
information
about
what
else,

pros
and
cons.

DR.
KELLEY:
Since
you
are
concerned
about
fertility,
you
have
to
add
a
rate
of
fertilization
of
eggs
as
an
endpoint.
And
there
are
a
number
of
different
ways
to
do
that.

Some
less
variable
than
others.
But
that's
an
important
endpoint.
And
that
will
require
growing
animals
out
longer.
Six
months.

So
that
will
add
to
the
cost
of
it.
But
it
is
an
important
endpoint
for
your
goal.
I
would
argue
the
most
important
endpoint
for
your
goal.

Do
you
need
more
detail
on
how
to
do
that?
There
are
42
different
ways.

DR.
ROBERTS:
I
don't
think
at
this
point.
But
I
think
that's
exactly
what
­­
provide
you
with
this
information
which
is
important,

but
it
means
the
experiment
has
to
go
longer.

DR.
KELLEY:
Right.
But
it
is
not
very
fancy.
You
don't
have
to
learn
to
do
a
vitellogenin
induction
assay.

You
get
the
testes
and
mush
it
up
and
fertilize
the
eggs,
you
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
262
know.

DR.
ROBERTS:
Other
suggestions?
Dr.
Isom
and
then
Dr.

Skelly.

DR.
ISOM:
We
have
talked
around
this
and
some
of
our
speakers
have
mentioned
this.
I
think
perhaps
this
accounts
for
some
of
the
variability
we
have
seen
among
studies.

That
is,
clear
definition
of
the
endpoints.
We
really
­­
even
to
define
sex
ratios,
how
are
we
doing
that?
There
is
all
kinds
of
variability
in
the
terminology,
and
I
would
really
encourage
somebody
to
sit
down
and
come
up
with
some
standardized
terminology
and
endpoints
and
how
you
are
going
to
evaluate
those
before
these
studies
are
conducted,
at
least
this
phase
one.

DR.
ROBERTS:
I
think
that's
something
that
everyone
on
the
panel
will
concur
with,
is
that
the
agency,
as
you
begin
this
effort,
is
going
to
have
to
standardize
the
terminology
for
the
endpoints
by
convening
a
workshop
or
whatever
mechanism
is
the
best
way
to
do
it.

But
there
has
to
be
some
terminology
that
everybody
is
using
and
agrees
upon.
We
clearly
saw
the
issues
as
the
agency
has.
This
is
not
something
new.
You
are
not
surprised
by
this
recommendation.

But
I
think
that's
something
we'll
make
in
our
minutes.

Other
endpoints?
Dr.
Skelly.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
263
DR.
SKELLY:
Just
to
build
on
what
Dr.
Kelley
suggested,
if
you
are
going
to
grow
animals
out
and
add
that
extra
investment,
I
think
it
is
going
to
be
important
to
beyond
castrating
the
males
and
mashing
up
their
testes
to
figure
out,
and
I
don't
know
enough
about
xenopus
to
know
how
this
could
work,
but
in
the
species
I
work
with,

it
would
be
possible
to
measure
reproductive
behavior
and
reproductive
endpoints
of
whole
living
animals.

That
would
measure
their
reproductive
function.
Because
what
Dr.
Kelley
suggested
is
taking
a
look
at
some
index
of
male
fertility.

But
we
don't
know
what
is
going
on
in
this
animal's
brain
and
whether
it
is
going
to
act
like
a
male
or
female
or
what
it
is
going
to
do.
I
think
that's
critical
to
getting
beyond
the
kind
of
physiology
to
the
behavior
in
the
ecology.

DR.
ROBERTS:
So
you
are
suggesting
another
possibility
is
to
add
reproductive
tests
that
involve
behavioral
component,
but
the
downside
is
that
there
is
­­
is
the
methodology
relatively
standard
or
this
something
they
would
have
to
work
through?

Dr.
Kelley
is
­­
I'll
let
you
respond
to
this.

DR.
SKELLY:
Before
Dr.
Kelley
builds
on
something
that
I
say,
just
let
me
say
that
I
don't
know
that
the
methods
are
standardized.
They
probably
vary
quite
a
bit
from
species
to
species,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
264
because
you
are
turning
on
frogs.
They
all
have
their
own
little
things
they
like.

But
it's
certainly
possible
to
do.
And
I
think
it
adds
significantly
too,
and
then
you
can
mash
their
testes
up.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
So
that's
what
we
study,
the
reproductive
behavior
of
male
frogs
and
female
frogs.
You
can
do
it.
And
we
have
standardized
assays
for
doing
it.

There
are
uncertainties
associated
with
this.
I
find
in
my
lab
that
the
more
things
you
try
to
get
out
of
a
single
animal,
the
greater
the
possibility
that
you
will
compromise
one
of
the
measurements.

I
don't
like
killing
animals.
I
always
try
to
get
everything
possible
from
the
animal
if
it
is
going
to
give
its
life
in
the
name
of
science.
But
there
are
problems
with
running
that
kind
of
experiment
where
you
end
up
confounding
it.

If
you
have
a
protocol
for
stimulating
the
testes
that
is
very
reliable
and
then
you
take
that
same
animal
and
test
it
behaviorally,
I
think
you
could
do
that.
And
you
could
test
clasping
for
which
there
is
normative
data.
The
papers
are
on
the
list.
You
could
test
the
vocal
behavior
quite
easily.

So
you
could
do
that.
It
would
add
to
the
study.
It
will
be,
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
265
be
honest
with
you,
behavior
is
hard
and
more
variable
than
numbers
of
fertilized
eggs.
It
will
add
a
level
of
uncertainty
and
a
level
of
interlab
variability
that
will
not
be
trivial.

But
it
is
doable.
I
agree
with
you
that
it
is
important.

On
the
other
hand,
you
have
to
say
to
yourself,
hey,
suppose
you
could
never
get
these
male
xenopus
to
clasp.
Is
that
going
to
be
something
that
you
are
going
to
regard
as
a
valid
endpoint
or
not.

I'm
a
little
bit
worried
about
that,
particularly
since
it
won't
be
very
easy
to
transpose
that
same
paradigm
to
rana
where
the
hormonal
requirements
for
male
reproductive
behavior
are
rather
different.

Very
different.

DR.
ROBERTS:
Dr.
Green,
did
I
see
your
hand
up?
And
then
Dr.
Denver.

DR.
GREEN:
I
was
just
thinking
along
the
lines
you
brought
up,
Dr.
Skelly,
that
an
endpoint
might
be
to
just
see
if
these
animals
can
naturally
mate
in
a
laboratory.
It
is
a
simple
easy
way
to
do.
It
doesn't
require
castration
or
anything.

And
the
evaluation
of
that
would
be
the
number
of
fertilized
eggs
at
the
end
of
the
time
they
have
been
in
the
bucket
together
or
something
simple
like
that.
There
is
a
lot
of
variability
in
that.

I
know.
I
have
seen
that.
And
the
females
will
eat
the
eggs
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
266
sometimes
and
that
kind
of
thing.
But
one
way
to
assess
their
behavior,
and
I'm
sure
you
can
elaborate
on
this,
Darcy,
would
be
if
you
put
them
together
after
hormonal
priming,
can
they
­­
do
they
mate
like
controlled
animals
do.

DR.
KELLEY:
So
again,
I
stress
that
you
need
a
very
well­
defined
stock
to
do
that
experiment
with.
You
need
to
make
sure
that
your
controls
and
your
experimentally
treated
animals
are
equivalent
in
establishing
that
group
that
everybody
agrees
are
going
to
be
the
­­
whatever
those
mice
are,
C
J
B
57
or
whatever,
you
know,

of
the
xenopus
world
is
going
to
take
a
little
bit
of
work.

DR.
ROBERTS:
I
think
as
we
make
some
comments,
it
is
important
to
give
the
agency
some
advice
about
how
straightforward
these
assays
are.
Let's
be
honest,
some
things
may
be
possible.
But
if
there
is
one
laboratory
in
the
world
that
can
do
it
because
they
have
enough
experience,
that's
an
important
practical
consideration
for
the
agency
as
opposed
to
other
techniques
that
are
more
straightforward
that
depending
upon
who
happens
to
do
it,
they
are
likely
to
get
decent
results
as
opposed
to
something
that
is
trickier.

I
think
it's
maybe
important
to
make
it
clear
to
the
agency
as
we
discuss
these
things,
give
them
some
sort
of
sense
about
that.

DR.
KELLEY:
I
can
give
you
a
protocol
for
watching
mating.
I
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
267
have
this
long
boring
paper
if
you
want
to
go
read
it
from
my
dissertation
on
hormone
levels,
amount
of
time
spent
in
amplexus,

number
of
clasp
attempts
and
controlling
for
females
and
so
forth.

It
is
spelled
out
very
clearly.
I
believe
it
could
be
replicated,

although
I
noticed
nobody
has
ever
wanted
to
do
it.
I
think
it
is
possible
to
do.

In
my
experience,
killing
the
animal
and
mushing
up
the
testes
is
far
and
away
the
easiest
way
to
get
a
first,
easy
measurement.

I
think
you
would
have
to
take
eggs
from
a
variety
of
females
and
you
will
have
to
have
a
good
control
and
you
will
have
to
make
sure
it
is
done
double
blind
bla,
bla,
bla.

But
you
will
end
up
with
if
you
fertilize
X
number
of
eggs,
you
get
X
number
of
offspring.
You
will
get
a
quantifiable
measure
that
can
be
subjected
to
robust
statistical
procedures
as
opposed
to
nonparametrics
requirements,
which
are
required
for
these
noncontinously
distributed
variables
like
percentage
of
males
clasping,
which
is
what
I
used
in
my
original
studies.

I'm
happy
to
teach
anybody
how
to
watch
frogs
clasp
each
other
and
score
it.
I
think
it's
pretty
easy
to
do.
I'm
just
telling
you
that
it
would
be
about
a
zillion
times
easier,
even
for
me,
to
do
a
fertility
assay
first.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
268
I'm
not
against
studying
reproductive
behavior.
It
is
what
I
do.

I'm
just
telling
you
it
is
a
more
variable
endpoint
in
a
field
in
which
the
endpoints
are
already
very
variable.

DR.
ROBERTS:
My
point
in
raising
this
is
I'm
not
trying
to
discourage
any
particular
kind
of
assay.
I
just
want
to
disclose
as
best
we
can
the
advice
­­
take
advantage
of
the
experience
of
you
individuals
and
experts
who
have
done
these
kinds
of
things
to
let
them
know
how
easy
or
how
hard
they
are
to
do.

Dr.
Richards,
Dr.
Green
and
then
Dr.
Delorme.

DR.
DENVER:
I
want
to
say
that
getting
an
estimate
of
reproductive
output
when
the
putative
species
is
the
male
is
going
to
be
really
difficult.

On
the
face
of
it,
it
sounds
fairly
straightforward
to
do
invitro
fertilization
with
oocytes,
mash
up
the
testes.
But
you
don't
need
a
lot
of
sperm
to
get
a
reasonably
efficient
rate
of
fertilization.

Also,
it
depends
on
the
quality
of
the
oocytes,
as
you
know.
So
standardizing
something
like
that
I
think
would
be
a
nightmare.
I
think
­­
if
the
target
species
were
a
female,
it
would
be
a
lot
easier
to
quantify,
say,
yolk
deposition
or
other
aspects
of
female
fecundity.

But
I
think
that
given
it
is
a
male,
we
may
need
to
think
about
other
ways
to
do
that.
Maybe
the
behavioral
tests
might
be
more
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
269
informative
or
more
capable
of
standardizing
the
­­

DR.
KELLEY:
I
have
the
behavior
paper
with
me.
You
guys
can
read
it
and
see
what
you
think.

DR.
DENVER:
But
what
do
you
think
about
that,
Darcy?
You
mentioned
the
invitro
fertilization
as
a
way
to
measure
­­

DR.
KELLEY:
You
are
absolutely
right.
It
is
a
problem.
The
other
way
to
do
it
is
simply
to
do
histology
on
the
gonads
and
look
at
the
spermatozoa.
You
can
stage
spermatozoa.
It
is
not
as
well
worked
at
as
mice,
but
you
could
stage
it.

If
you
get
a
big
effect,
it
is
never
a
problem.
If
you
have
a
marginal
problem,
it
is
always
a
problem.
Right?
What
I
would
like
to
do
is,
if
you
wouldn't
mind,
to
go
back
and
reread
my
dissertation
paper
from
1975
and
see
what
the
error
bars
are
like,
basically.
See
how
variable
the
measure
is
for
looking
at
clasping.

Calling
is
not
going
to
do
you
any
good.
Either
they
call
or
they
don't.
That's
it.
You
could
look
at
amount
of
time
calling
at
various
levels,
but
that
reflects
internal
androgen.
And
calling
is
really
difficult.
To
get
robust
callers
at
certain
times
of
year
is
really
hard.
But
clasping
is
much
less
difficult,
the
amplectant
(
ph)

position.

Let
me
go
back
and
look
at
the
clasping
data
and
get
back
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
270
you.

DR.
ROBERTS:
Dr.
Richards,
Dr.
Green,
Dr.
Delorme.

DR.
RICHARDS:
I
just
want
to
point
out
I
have
never
worked
with
behavior
in
amphibians,
but
I
have
done
a
little
bit
with
fish
and
invertebrates.

Doing
behavioral
experiments
in
the
lab
are
problematic.
They
are
very
complex.
They
are
difficult
to
anticipate
outcomes
often
times.

And
I
think
behavior
is
best
viewed
in
the
field.
I
think
that
some
of
the
mashing
and
counting
things
might
be
better
measurements
at
this
point
in
the
game.

Plus,
I'm
not
sure
that
federal
employees
are
allowed
to
watch
clasping
behaviors.

DR.
ROBERTS:
Good
point.

Dr.
Green.

DR.
GREEN:
You
had
asked
to
indicate
how
feasible,
how
easy
some
of
these
proposed
experiments
might
be
to
do.

Just
to
communicate
to
you,
in
our
animal
facility,
natural
matings
are
quite
common
just
for
the
purpose
of
collecting
the
fertilized
eggs
and
studying
the
eggs.

And
usually
in
the
fall,
we
have
the
arrival
of
150
or
so
new
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
271
graduate
students
and
post
docs
who
have
never
dealt
with
frogs
before.
I'll
go
through
the
frog
rooms
and
there
may
be
as
many
20
to
50
buckets
lined
up
on
the
floor
where
animals
are
paired.
They
have
been
primed
hormonally.
They
are
put
together
and
essentially
left
overnight
undisturbed.
Some
labs
have
video
cameras
to
watch.
I
don't
know
why.

So
there
is
a
way
to
record
the
behavioral
aspects
of
the
mating
without
disturbing
the
frogs.
And
then
the
students
come
back
the
next
day
hopefully
early
enough
that
the
females
haven't
eaten
the
frogs,
but
there
are
ways
to
prevent
that
from
happening
with
a
mesh
grid
at
the
bottom.

And
then
they
take
the
eggs
on
upstairs.
And
as
I
mentioned,
I
think
that
seems
relatively
straightforward.
I
know
all
about
the
variability
in
that.
You
go
to
the
dissecting
scope
and
you
look
for
the
ones
that
are
fertilized,
and
you
could
carry
that
one
step
further
to
see
how
many
actually
become
viable
tadpoles.

The
endpoint
could
go
on
and
on
and
on.
But
that
basic
part
is
fairly
simple
to
do.

DR.
ROBERTS:
Dr.
Bradbury.

DR.
BRADBURY:
I
just
wanted
to
do
a
check­
in
to
see
if
I'm
synthesizing
the
dialogue
properly.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
272
At
one
point
there
was
a
statement
that
seemed
to
be
accepted
by
several,
at
least,
that
there
should
be
an
attempt
to
try
to
replicate
the
studies
that
have
been
discussed
over
the
last
few
days,
the
lab
studies.
And
that
that
was
an
important
task
at
hand.

And
that
xenopus
and/
or
rana
simultaneously
or
­­
we're
going
to
think
about
that
a
little
bit
later,
the
Phase
1,
and
again,
don't
get
it
too
sequenced,
was
a
proposal
to
get
out
to
the
panel
that
wasn't
intended
to
be
a
detail
by
detail
replication
of
the
previously
published
studies
in
part
because
we
had
concerns
about
growth
and
water
quality
and
those
kind
of
things.

But
in
the
spirit
of
many
of
the
discussions
over
the
last
few
days,
that
if
a
response
seems
to
be
consistent
and
is
occurring
in
a
reproducible,
logical
pattern,
that
study
designs
don't
have
to
be
identical,
but
there
are
certain
principles
they
are
holding
to
to
see
if,

in
fact,
they
are
getting
concordance
of
the
information
development.

Phase
1
is
designed
in
that
spirit.
And
seemed
­­
was
sort
of
our
proposal
for
your
consideration
in
terms
of
"
repeating
what
had
been
done
in
the
previous
studies."

Am
I
to
interpret
that
that
is
one
aspect
of
an
experiment
or
a
study
that
would
go
on,
and
some
of
the
other
discussion
we
have
been
hearing
would
be
in
the
context
of
aspects
of
studies
we
have
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
273
been
using
as
phase
five?
Or
am
I
to
interpret
that
not
only
do
you
want
us
to
come
up
with
an
experimental
design
that
can
capture
"
what
has
been
previously
studied,"
but
also
include
all
these
additional
endpoints?

I
could
imagine
an
experimental
design
that
would
do
that.
But
I
would
be
curious
if
that's
what
you
mean.
Or
are
we
talking
about
a
series
of
separate
studies?

DR.
ROBERTS:
Responses?
Dr.
Delorme,
actually
you
were
next
up
anyway.
DR.
DELORME:
Actually,
I
was
going
to
bring
up
that
point,
because
I
was
getting
confused
as
to
what
is
going
on.

And
I
flipped
back
to
the
slide
on
phase
one
which
is,
test
for
apical
gonadal
effects,
and
the
objective
was
to
determine
if
atrazine
exposure
results
in
gonadal
effects
in
males
and
females
in
brackets.

When
I
had
put
in
my
consensus
that
repeating
the
experiments,

basically,
that's
what
this
one
is
about,
and
then
the
side
bar
with
the
ecological
relevance
is
sort
of
the
next
thing,
I
mean,
bringing
it
back
and
wearing
my
risk
assessor
hat,
because
that's
what
I'm
doing,
I
think
we
have
to
recognize
that
EPA
wants
to
move
this
risk
assessment
forward.

And
the
first
step
in
doing
that
is
determining
whether
or
not
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
274
that
cause
effect
between
atrazine
and
gonadal
effects
is
there.

So
perhaps
maybe
we
want
to
consider
an
experiment
which
would
try
and
replicate,
not
replicate,
but
at
least
firm
up
that
cause
effect.

DR.
ROBERTS:
Dr.
Green.

DR.
GREEN:
I
think
the
issue
was
whether
or
not
those
gonadal
effects
that
you
see
morphologically
have
any
physiological
consequence.
It
is
hard
to
answer
that
question
if
all
you
are
looking
for
is
gonadal
abnormalities.

That
would
be
the
endpoint,
I
guess,
and
a
place
to
start
in
phase
one.
If
they
do,
then
maybe
go
on
with
other
experiments
to
then
test
the
hypothesis
as
to
whether
or
not
that
has
any
effect
on
fertility
and
fecundity.

DR.
ROBERTS:
Dr.
Bradbury.

DR.
BRADBURY:
The
way
that
discussion
was
going
is
good.
I
think
it
would
be
helpful
for
us
to
hear
you
explore
is
it
sort
of
a
sequential
­­
get
this
information,
check
in
what
to
do
next
or
do
you
try
to
do
it
all
at
once.

And
I
think
you
opened
up
dialogue
that
would
be
good
to
hear
some
more
about.

DR.
ROBERTS:
Dr.
Delorme,
then
Dr.
Skelly
and
then
Dr.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
275
Coats.

DR.
DELORME:
You
may
be
able
to
design
a
study
that
you
could
actually
phase
the
results
where
you
get
the
results
of
the
initial
study
but
you
have
enough
animals
left
over
that,
if
you
need
to
take
it
to
looking
at
the
physiological
consequences
of
having
disrupted
or
abnormal
gonads
whether
or
not
that
has
effect.

I
guess
that
would
be
possible.
But
again,
I
come
back
to
what
Dr.
Green
said
earlier.
The
more
things
you
try
and
get
out
of
a
study,
then
you
risk
having
uncertainty
creep
in
or
losing
animals
or
whatever.

So
you
are
going
to
have
to
evaluate
the
relative
merits
of
what
you
are
going
to
get
out
based
on
the
design.

If
you
want
to
design
a
study
where
you
could
initially
try
and
replicate
or
look
at
the
causal
effect
with
the
idea
that
you
are
going
to
have
enough
animals
built
into
it
that
you
can
take
it
beyond
that
should
you
find
out
that
there
is
effect,
certainly
you
could
do
that.

DR.
ROBERTS:
Dr.
Skelly.

DR.
SKELLY:
Dr.
Delorme
said
much
of
what
I
wanted
to
say.

I
guess
one
way
to
phrase
the
recommendation
back
to
EPA
would
be
in
the
context
of
doing
a
power
analysis
and
thinking
about
how
many
animals
need
to
be
available
at
each
stage.
How
many
of
them
are
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
276
going
to
have
to
be
sacrificed
at
each
stage.

Some
things
that
we're
not
doing
when
we're
sitting
here
talking
about
it
may
become
obvious
one
way
or
another
that
it
makes
sense
to
economize
and
break
those
up
or
it
makes
sense
to
economize
and
put
them
together.

DR.
ROBERTS:
Good
point.
Dr.
Coats.

DR.
COATS:
My
thinking
is
that
phase
one
is
presented
up
there
and
as
described
in
the
book
is
pretty
much
where
we
need
to
start.
But
that
there
has
been
discussion
of
maybe
adding
one
more
endpoint
on
to
the
end
of
that
or
two
if
they
are
not
too
complicated.

DR.
ROBERTS:
Others?
Dr.
Gibbs.

DR.
GIBBS:
If
I
can
just
add
one
point.
Molecular
genetics
gets
right
to
the
heart
of
the
matter,
which
would
be
paternity.
There
are
simple
ways
without
getting
into
behavior.
For
male
frogs,

paternity
is
the
bottom
line.
However,
they
get
there.
And
I
think
you
could
devise
some
fairly
simple
experiments,
toe
clips
from
all
of
your
adults
and
all
of
the
metamorphs
and
you've
­­
genotyping
them
and
assigning
paternity.
And
you
have
got
some
good
data.
It
gets
right
at
the
issue.

DR.
ROBERTS:
Any
other
comments?
What
do
you
think,
Dr.

Bradbury,
is
that
­­
we
have
sort
of
honed
in
on
it
a
little
more?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
277
DR.
BRADBURY:
Yes.
I
know
you
are
going
to
be
writing.
I
now
have
a
feeling
that
I
can
imagine
some
of
the
dialogue
that
will
be
on
paper
that
will
be
helpful
based
on
this
last
round
of
discussions.

DR.
ROBERTS:
I
hope
Dr.
Richards
can
also
imagine
that
dialogue.
Dr.
Coats.

DR.
COATS:
One
point
had
not
been
discussed
very
much
was
the
dose
response
possibility
as
described
in
the
tier
one,
phase
one
there.

And
I
think
it
is
an
extremely
important
part
of
it,
and
would
be
very
informative
to
elaborate
on
the
dosing
scheme.
That
may
be
discussed
at
a
later
point.

DR.
ROBERTS:
I
think
it
comes
up
in
C
as
I
would
interpret
C.

Have
we
finished
then
with
B,
Dr.
Richards?
Do
you
want
to
poll
the
panel?
Do
you
have
a
pretty
good
feel
for
where
we
are
on
this?

DR.
RICHARDS:
I
don't
know
if
the
point
about
conditions
and
this
thing
about
the
ASTM
standards
and
flow
through
­­

DR.
ROBERTS:
I
think
we're
going
to
get
to
that.

DR.
RICHARDS:
­­
is
that
going
to
come
later?

DR.
ROBERTS:
I
think
so.
We're
still
sort
of
what
components
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
278
need
to
be
in
that
phase
one
and
that
sort
of
thing.

Dr.
Delorme.

DR.
DELORME:
I
just
want
to
ask
for
a
clarification.
You
have
replication
up
there
as
two.
Does
that
mean
two
tanks
or
two
species
or
two
xenopus?

DR.
TIETGE:
It
was
meant
to
represent
two
tanks.
I
think
we
have
since
­­
this
is
an
older
version
of
the
slide
for
some
reason.

DR.
DELORME:
Do
you
want
to
comment
on
whether
or
not
you
think
that's
appropriate
given
the
context
of
the
discussion?

DR.
TIETGE:
No.
In
the
context
of
the
power
analysis
and
the
discussion
that's
been
going
on,
it
is
probably
not
sufficient.

It
is
a
commonly
used
approach
in
aquatic
toxicology,
at
least
two.

DR.
ROBERTS:
Should
we
go
on
to
8
C
then?

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

DR.
ROBERTS:
I
think
for
this
particular
question
­­
I'm
sort
of
looking
ahead.
We
have
some
questions
­­
most
of
the
rest
of
them
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
279
deal
rather
specifically
with
the
species
involved
and
differences
and
those
kinds
of
things.

I
think
this
is
probably
the
best
question
to
raise
not
only
issues
regarding
to
spacing
and
dosing
of
atrazine
concentrations,
but
perhaps
some
of
these
other
things
about
experimental
design
that
we
have
been
sort
of
eager
to
bring
up.

With
that
said,
let's
go
to
Dr.
Richards.

DR.
RICHARDS:
I'm
going
to
open
it
up
to
those
first
statements
on
range
and
spacing.
I
know
some
of
you
had
some
stronger
feelings
about
that
if
you
want
to
just
jump
in
here.

DR.
ROBERTS:
Dr.
Heeringa
had
to
leave
for
the
airport.
But
he
did
give
me
typed
up
comments.
With
your
forbearance,
let
me
read
them,
because
I
think
that's
probably
the
best
way
to
get
them
into
the
record.
And
I
would
not
dare
try
and
paraphrase
his
points
for
fear
of
not
getting
them
right.

In
regard
to
8
C,
he
says,
this
question
must
be
answered
in
the
context
of
a
presumed
constraint
on
the
cost
and
effort
that
can
be
devoted
to
a
single
replication
of
the
study
to
determine
if
aqueous
concentrations
of
atrazine
bears
a
relationship
to
gonadal
irregularities
and
any
associated
mechanicism
for
endocrine
disruption
in
frogs.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
280
The
scope
of
the
study
is
determined
by
the
number
of
concentrations
and
controls
tested,
the
number
of
intralab
replications,
for
example,
tanks
for
each
control
concentration
level,

and
the
number
of
test
animals
per
experimental
replication.

Range
in
spacing
of
the
experimental
concentration
levels
is
obviously
related
to
the
number
of
feasible
experimental
points.

Range
in
spacing
should
also
be
governed
by
the
specific
test
hypothesis
concerning
the
potential
shape
of
any
underlying
concentration
response
relationship.

Consider
the
components
of
this
design
in
the
following
order.

One,
selection
of
controls.
Two,
range
of
observations
for
experimental
concentrations.
Three,
number
of
independent
replicates
per
treatment.
Four,
number
of
test
animals
per
experimental
replicate.
And
five,
number
in
spacing
of
experimental
concentrations.

Number
one,
selection
of
controls.
The
experiment
should
include
untreated
control
replicates
and
a
positive
control
under
which
test
animals
are
exposed
to
a
concentration
of
estrogen.
I
support
the
EPA
statement
that
a
positive
androgen
control
group
is
not
needed,
although
would
be
beneficial
if
laryngeal
muscle
measurement
is
included
as
an
endpoint.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
281
As
a
protective
factor
in
the
test
of
hypothesis
concerning
the
dichotomous
or
polyotomous,
(
ph)
Dr.
Hayes'
classification,
gonadal
deformity
endpoint,
which
is
based
on
existing
data,
suggests
a
one­

sided
test.
The
sample
size
for
the
untreated
control
should
be
increased
beyond
the
levels
that
would
be
assigned
to
concentration
points
in
the
nonzero
domain
of
the
test
range.

Number
two.
Range
of
observations.
Setting
aside
the
positive
controls,
the
range
of
experimental
concentration
should
span
zero
or
untreated
atrazine
concentration
across
ecologically
relevant
concentrations
and
extend
to
concentrations
that
meet
and
at
least
one
point
exceeds
the
upper
percentile
bounds
that
have
been
measured
in
natural
aquatic
environments.

Number
three,
number
of
replications
for
each
experimental
treatment.
For
the
xenopus,
and
that's
Hayes
and
Carr
studies,
and
the
rana
studies,
Hayes,
estimates
of
the
empirical
intrareplication
or
intratank
correlation
should
be
obtainable.

Based
on
concordance
of
the
values
or
maximum
of
estimates
if
highly
variable,
the
estimated
intraclass
correlation
should
be
used
to
determine
the
number
of
replicates
per
treatment
arm
and
the
allocation
of
total
sample
size
to
replications
and
test
animals
per
replication.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
282
This
should
be
based
on
the
best
empirical
data
from
the
existing
studies.
Underestimating
the
intrareplicate
correlation
and
planning
the
sample
size
allocation
may
seriously
attenuate
the
true
power
of
the
test
of
the
hypothesis
concerning
the
chosen
endpoints
and
in
particular
test
of
hypotheses
concerning
a
dichotomous
that
is
deformity
outcome
or
the
differentiated
polytomous
classification
of
deformities
proposed
by
Dr.
Hayes
in
his
presentation
to
the
panel.

Number
4,
number
of
animals.
Subjects
per
replication
having
established
a
working
value
for
the
intrareplicate
correlation
for
the
class
of
outcomes
of
interest
and
a
desired
level
of
statistical
power
for
a
specific
hypothesis
test.

The
determination
of
the
optimal
number
of
animal
subjects
can
be
determined
jointly
with
the
determination
of
number
of
treatment
replicates.

This
allocation
is
obviously
constrained
by
bio­
loading
and
water
quality
considerations
that
are
discussed
in
the
EPA
white
paper.

Finally,
Number
5,
optimal
determination
of
the
number
in
spacing
of
treatments
as
governed
by
the
hypothesized
shape
of
any
underlying
concentration
response
relationship.

The
panel
has
determined
that
data
from
existing
studies
lend
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
283
support
to
further
testing
the
hypothesis
of
a
relationship
between
atrazine
concentration
and
gonadal
abnormalities
in
frogs.

Dr.
Hayes
provided
data
and
arguments
that
the
relationship
is
not
monotonic
potentially
in
inverted
response.
But
this
has
not
been
replicated
in
other
studies.

At
this
stage,
there
is
insufficient
basis
to
set
the
spacing
of
treatments
to
optimize
the
experimental
design
for
a
functional
form
for
a
potential
concentration
response
curve.

A
robust
design
would
use
multiple
concentration
points
to
accommodate
the
possibility
that
any
effect
is
monotonic
or
alternately
that
there
is
a
simple
nonmonotonic
or
convex
relationship.

There
was
also
an
advantage
to
retaining
concentration
points
that
have
been
used
in
the
prior
research
by
Hayes,
Carr,
Hecker
and
others,
0,
0.01,
1,
10
and
25
micrograms
per
liter,
and
adding
an
upper
concentration
level
that
exceeds
the
25
microgram
per
liter
value,
at
which
Dr.
Hayes
and
Carr
studies
have
detected
increases
in
the
number
of
abnormalities,
such
as
basing
of
concentration
treatments
should
be
sufficient
to
test
the
hypothesis
of
an
effect
and
to
secondarily
test
whether
any
real
effect
is
monotonic
or
nonmonotonic.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
284
Obviously,
these
concentration
levels
would
not
permit
testing
threshold
responses
below
the
.01
microgram
per
liter
level.

I
will
give
these
to
Dr.
Richards
to
work
into
our
comments.

Dr.
Green,
did
you
want
to
­­

DR.
GREEN:
I
was
just
going
to
say
I
support
his
recommendations,
and
I
particularly
like
the
dose
range
that
he
has
proposed
in
that.

DR.
ROBERTS:
Other
comments.
Dr.
Coats.

DR.
COATS:
Yes.
I
liked
some
of
the
comments.
I
don't
think
that
dose
range
is
going
to
depict
the
curve
any
more
clearly
than
­­

concentration
response
any
better
than
what
we
have
already
seen
other
than
one
laboratory
doing
all
those
concentrations,
which
Dr.

Hayes
has
mostly
done.

I
think
if
there
is
an
unusual
response
at
which
point
.1
part
per
billion
is
a
significant
concentration,
that
points
near
that,
above
and
below
ought
to
help
delineate
the
response
curve
if
it
is
a
curve.

And
that
if
you
are
going
by
orders
of
10,
orders
of
magnitude,

you
would
miss
that.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
I
agree
completely
with
Dr.
Coats.
I
think
we
need
to
recognize
here
or
at
least
acknowledge
what
the
intent
of
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
285
experiment
is
and
design
it
appropriately.

We
can
design
the
experiment
to
replicate
earlier
experiments,

that
is,
determine
whether
or
not
atrazine
is
eliciting
the
effect,
or
we
can
try
and
characterize
the
shape
of
the
concentration
response
curve.

But
I
don't
know
that
we
can
do
both
unless
we
get
lucky.
And
I
would
suggest
that
probably
the
first
time
around
we
follow
the
recommendations
of
Dr.
Heeringa
recognizing
that
we're
probably
not
going
to
characterize
the
concentration
response
curve,
but
will
have
a
good
design
to
replicate
previous
observations
and,
if
we
get
lucky,

perhaps
we'll
gain
information
on
the
concentration
response
curve
as
well.

DR.
ROBERTS:
Dr.
Green.
Then
Dr.
Delorme.

DR.
GREEN:
The
aspect
I
liked
about
the
dose
responses
or
the
doses
that
he
has
proposed
is
that
they
do
include
the
ones
that
have
been
looked
at
previously.

And
I
agree.
They
could
be
added
to.
And
I
like
the
fact
that
he
exceeded
the
maximum
dose
that
was
looked
by
both
labs.
I
can't
recall
how
many
fold,
how
many
times
he
said
higher
than
the
25.

What
was
the
next
value?
Did
he
give
one?

DR.
ROBERTS:
Based
on
some
percentile,
I
believe,
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
286
observations
from
the
field.

DR.
GREEN:
That
seems
important
to
do.

DR.
ROBERTS:
Dr.
Delorme.

DR.
DELORME:
I
just
wanted
to
ask
a
clarification
of
EPA.

Were
you
intending
this
as
an
inova
(
ph)
design
study
which
is
a
hypothesis
testing
or
one
which
­­
a
linear
design
­­
or
a
regression
type
of
design
where
you
actually
want
to
get
a
dose
response?

Because
that
actually
makes
a
difference
on
how
you
space
your
doses
and
how
many
animals
you
have
in
your
replicates.

This
actually
goes
to
some
of
the
other
comments
that
have
already
been
made.

DR.
ROBERTS:
Dr.
Tietge.

DR.
TIETGE:
To
the
extent
possible,
I
think
dose
response,
but
I
think
recognizing
what
Dr.
LeBlanc
said
is
you
may
not
be
able
to
hit
the
proper
range
of
concentrations
to
achieve
that.

DR.
ROBERTS:
As
Dr.
LeBlanc
said,
often
times
when
you
do
these
kinds
of
studies,
especially
if
there
is
a
sharp
inflexion
in
the
curve,
then
you
have
to
go
back
and
start
loading
in
in
that
critical
range.
So
it
wouldn't
be
surprising.

But
I
think
at
least
the
recommendations
by
Dr.
Heeringa
span
the
right
range
of
doses.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
287
Any
other
recommendations
or
comments
about
spacing,
et
cetera?
Dr.
Delorme.

DR.
DELORME:
Just
out
of
curiosity,
are
you
worried
about
the
low
end
of
the
dose
response
curve
or
the
upper
end?
From
a
regulatory
perspective,
often
times
what
we're
looking
for
is
near
to
no
effects
or
what
we
would
putatively
call
acceptable
effects.

DR.
BRADBURY:
I
think
we're
sort
of
getting
into
a
phase
within
a
phase.
I
think
the
first
question
is
can
you
reproduce
the
effect
and
can
we
use
that
in
dose
basings
that
were
recommended
I
think
are
reasonable
to
get
some
sense
of
consistency
across
studies.

Give
me
a
concentration
response
curve
to
start
work
even
if
it's
crude,
then
we
can
start
talking
about
those
kinds
of
things.
I
think
it's
premature
until
we
get
a
response
relationship.

DR.
ROBERTS:
Using
the
chairman's
prerogative,
I'm
going
to
change
my
mind
about
talking
about
other
experimental
design
aspects
at
this
time.
Let's
go
through
the
rest
of
the
questions,
and
then
at
the
end
if
there
are
recommendations
regarding
other
aspects
of
the
experimental
design
or
other
points,
let's
go
ahead
and
bring
them
in
at
that
time.

Are
there
any
other
comments,
then,
on
the
specific
issues
raised
in
this
particular
question
on
the
range,
spacing
and
number
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
288
atrazine
concentrations?

Let's
move
on,
then,
to
D.

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

DR.
ROBERTS:
Dr.
Richards,
do
you
want
to
lead
off
or
throw
it
open
for
discussion?

DR.
RICHARDS:
Let
me
make
a
comment.
I
think
that
the
agency
and
others
have
indicated
there's
lots
of
reasons
you
use
xenopus,
for
a
variety
of
laboratory
tests
and
quick
techniques
and
so
forth.

I
think
we
have
also
previously
identified
some
need
to
start
initiating
a
tighter
rana
procedure
in
the
laboratory.

But
I
would
like
to
throw
it
over.
I
hope
Dr.
Kelley
will
respond
on
the
differences
in
the
differentiation.

DR.
KELLEY:
I
don't
think
we
know
enough
to
be
able
to
answer
this
question.
We
know
a
huge
amount
about
xenopus
laevis,

and
we
know
so
much
less
about
rana
that
I
couldn't
tell
you.
I
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
289
couldn't
answer
this
question.

DR.
ROBERTS:
Dr.
Skelly.

DR.
SKELLY:
I
guess
Dr.
Kelley's
point
suggests
to
me
that
­­

just
to
support
what
EPA
has
proposed,
that
rana
is
used
as
a
corroborating
species
for
these
early
experiments.

DR.
ROBERTS:
Any
other
comments
or
any
comparison
of
the
species
or
the
suitability
of
one
to
serve
as
a
model
for
the
other?
Dr.

Denver,
and
then
Dr.
Green.

DR.
DENVER:
I
just
want
to
concur
with
Dr.
Skelly
that
the
ranid
species
do
need
to
be
considered
early
in
the
game.

DR.
ROBERTS:
Dr.
Green.

DR.
GREEN:
I
think
it
is
probably
obvious,
but
we
could
probably
all
predict
that
there
will
be
differences
between
the
two
species.

DR.
ROBERTS:
Any
other
comments
on
this
one?
Dr.
Gibbs.

DR.
GIBBS:
Only
that
the
rana
maybe
should
be
clarified.
It
should
be
a
North
American
rana.
There's
rana
on
different
continents.

DR.
ROBERTS:
Then
let's
let
our
minutes
reflect
that
clarification.

Let's
go
ahead
and
go
to
E
which
is
sort
of
the
flipside
of
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
290
question.

DR.
STEEGER:
In
this
regard,
are
there
important
differences
between
the
species
to
conclude
that
any
affected
developmental
processes
observed
in
xenopus
laevis
would
not
occur
in
rana?

DR.
ROBERTS:
Dr.
Richards.

DR.
RICHARDS:
I
will
throw
that
one
open.

DR.
ROBERTS:
Dr.
Kelley.

DR.
KELLEY:
Every
species
develops
in
its
own
way,
but
they
all
develop
their
gonads,
and
they
differentiated,
you
know,
not
too
roughly
inappropriate
times.

There
actually
are,
although
I'm
not
sure
how
valid,
there
are
tables
that
relate,
the
tables
of
normal
development
in
xenopus
to
roughly
equivalent
stages
in
rana.
So
it
is
possible
to
normalize
a
little
bit
in
that
way.

The
developmental
biologists
believe,
although
they
now
mostly
study
xenopus,
in
the
old
days
they
studied
rana,
and
the
developmental
biologists
believe
that
the
fundamental
processes
are
extremely
similar,
shpay
mons
(
ph)
organizer,
induction
and
so
forth.

I
don't
know
of
any
strong
species
difference
that
would
lead
me
to
believe
that
there
would
be
some
fundamental
reason
­­
some
fundamental
difference
at
the
developmental
level
that
would
lead
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
291
anything
affected
in
xenopus
to
not
necessarily
be
affected
in
rana.

So
that's
a
very
couched
statement,
but
I
believe
that
the
developmental
biology
community
would
agree
with
me.

DR.
ROBERTS:
Dr.
Skelly.

DR.
SKELLY:
I
will
support
everything
Dr.
Kelley
just
said
with
just
an
added
comment.
It
seems
like
timing
is
important
here.

And
one
aspect
that
xenopus
differs
from
almost
all
other
frogs
is
in
how,
say,
the
onset
of
reproduction
and
the
development
of
reproductive
morphology
differs
with
respect
to
the
onset
of
other
sort
of
metamorphic
characters
in
almost
all
other
frogs.

So
there
are
things
that
just
never
happen
in
xenopus.

Transitions
that
either
don't
take
place
or
take
place
differently
in
xenopus
than
they
would
in
rana
and
almost
all
other
frogs
that
undergo
metamorphosis.

That
suggests
that
either
the
genes
that
control
development
and
some
of
the
developmental
pathways,
and
I'm
getting
way
out
on
a
limb,
I'm
going
back
to
grad
school
to
remember
this
stuff,
but
suggest
to
me
that
developmental
pathways
and
timing
of
things
are
somewhat
different.

And
since
timing
of
the
exposure
to
atrazine
or
whatever
can
happen
at
different
times
relative
to
developmental
sequence,
even
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
292
though
there
are
these
tables
that
relate
xenopus
development
to
rana
development,
what
is
going
on
inside
and
when
genes
are
turning
on
and
those
sorts
of
things,
I
don't
know
that
we
know
that
much
about
that
right
now.

I
guess
that's
just
a
note
of
caution
in
assuming
that
these
things
are
going
to
happen
similarly.
I
will
just
finish
by
saying
that
I
think
we
have
seen
pictures
anyway
that
suggest
that
some
of
the
outcomes
in
gonadal
abnormalities
seem
to
manifest
themselves
somewhat
differently
in
the
two
species.

DR.
ROBERTS:
Other
points?

Should
we
go
to
F
which
is
again
related
to
the
same
­­

different
way
of
sort
of
tackling
the
same
kind
of
issue?

DR.
STEEGER:
Alternatively,
are
there
developmental
pathways
in
rana
but
not
in
xenopus
laevis
that
raise
concerns
about
using
xenopus
laevis
as
the
primary
biological
model
in
any
future
atrazine
studies?

DR.
ROBERTS:
Dr.
Richards.
You
were
going
to
defer
to
the
person
who
just
walked
out
the
door.

DR.
RICHARDS:
Like
a
flipside
of
what
we
just
spoke
of
here.

DR.
SKELLY:
I
guess
I'm
getting
tired,
but
I
was
hoping
to
make
the
comment
I
made
for
E
for
F.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
293
You
can
move
that.

(
Thereupon,
the
time
was
5
o'clock
p.
m.)

DR.
ROBERTS:
I
think
we
could
sort
of
take
these
up
kind
of
collectively.
Because
if
I
understand
it
correctly,
I
think
the
agency
is
trying
to
ask
in
different
ways
to
what
extent
can
we
extrapolate
data
from
xenopus
to
rana,
and
does
it
serve
as
an
effective
surrogate
or
is
there
any
reason
to
be
concerned
if
we
used
it
as
our
primary
model
that
it
would
mislead
us
about
what
was
going
on
in
rana.

Dr.
Denver.

DR.
DENVER:
I
don't
have
any
evidence
that
would
suggest
that
the
basic
developmental
pathways
would
differ,
although,
there
are
distinct
differences,
obviously,
in
life
history
and
physiology
and
we
have
documented
differences
in
the
development
of
the
stress
axis
when
the
stress
axis
becomes
responsive,
the
production
of
stress
hormones
in
the
two
species
throughout
development,
things
like
that.

So
the
timing
of
things
are
different
in
the
two
species
as
Dr.

Skelly
pointed
out.
But
I
don't
have
any
evidence
to
suggest
that
the
basic
mechanisms
are
that
different.
So
I
think
the
main
point
there
is
to
realize
that
the
critical
periods
or
the
sensitive
periods
may
differ
between
the
two
species.
So
that
needs
to
be
considered
in
any
experiments
that
are
designed.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
294
DR.
ROBERTS:
Other
points?
Dr.
Delorme.

DR.
DELORME:
Just
stepping
back
and
taking
the
broad
picture,
I
guess
what
I'm
hearing
is
that
you
can
make
the
assumption
that
there
may
be
not
too
many
differences,
but
you
are
making
an
assumption.

And
at
some
point,
you
are
going
to
have
to
test
that
assumption.
It
may
not
have
to
be
right
away.
But
certainly,

sometime
in
the
future
you
are
going
to
have
to
do
some
work
to
work
on
it.

DR.
BRADBURY:
I
think
I'm
picking
up
the
message
that
xenopus
is
a
reasonable
biological
model
to
further
confirm
a
toxicological
signal
related
to
gonadal
development
and
test
that
with
atrazine.
That
that's
a
reasonable
biological
model
to
get
started.

Does
that
mean
that
all
amphibians
will
respond
exactly
the
same
way
as
xenopus
does?
I
realize
we're
not
saying
that.
But
just
as
we
do
our
aquatic
toxicology
testing,
we
don't
have
the
luxury
to
test
all
the
thousands
of
species
of
fish
in
North
America.
We
have
to
settle
on
a
few
surrogate
species
to
give
us
a
sense
of
the
toxicological
potential
of
the
chemical
and
then
through
other
analyses
one
deals
with
species
extrapolation
and
other
aspects
of
life
history
to
refine
our
risk
assessments
as
needed.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
295
So
for
that
matter,
we
have
a
handful
of
mice
and
rats
that
we
use
to
go
to
humans.
But
at
least
we
have
several
species
to
go
to
one
species.
Ecological
toxicology,
we're
going
from
a
handful
of
species
out
to
potentially
several
or
many.

So
what
I
think
I'm
hearing
is
that
xenopus
is
a
reasonable
biological
model
to
get
a
handle
on
the
consistency
of
previous
studies
in
terms
of
the
potential
for
atrazine
to
initiate
these
developmental
effects.

The
proposal,
then,
is
to
use
a
North
American
species
to
get
some
sense
of
species
variability
and
to
be
using
a
species
of
North
America.
Does
that
represent
all
the
North
American
species?
No.

But
at
this
point,
I
would
say
that
we're
sort
of
in
the
venue
of
all
of
the
challenges
we
have
in
ecological
risk
assessment
across
the
board
in
how
to
extrapolate
across
many
species
when
you
only
have
data
for
a
few,
unless
the
panel
wants
to
probe
that
a
bit.

DR.
ROBERTS:
I
think
I
heard
not
only
just
now
but
discussions
earlier
during
the
last
couple
of
days
that
there
are
some
­­
since
a
primary
objective
as
we
just
discussed
at
least
initially
is
to
get
a
handle
and
do
some
well
controlled
studies,
that
there
are
a
number
of
advantages
in
using
xenopus
to
do
that
because
it's
well­
characterized.
There
is
a
lot
of
experiments
and
it
lends
itself
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
296
doing
those
kinds
of
experiments
that
really
need
to
be
done
initially.

But
as
you
say,
we'll
also
have
to
do
at
least
some
experiments
on
rana
because
of
uncertainty
about
the
extent
to
which
xenopus
would
be
representative.

Does
anybody
else
want
to
add
anything?

DR.
DENVER:
I
just
wanted
to
add
a
related
point.
That
came
up
earlier
with
regard
to
measurements
of
uptake
of
atrazine
and
body
burdens
and
that
sort
of
thing.

These
two
species
have
very
different
feeding
ecology
as
tadpoles.
And
that
could
translate
into
differential
rates
of
uptake
and
different
exposure
to
the
compound.

And
so
I
think
that
at
the
outset
it
is
important
to
address
those
issues.
I
think
as
already
been
­­
that
point
has
been
made,
I
don't
know
that
has
been
made
today,
but
it
has
been
made
previously
that
that
be
measured
and
perhaps
compared
between
species
and
the
studies.

DR.
ROBERTS:
Dr.
Denver,
can
you
provide
a
reference
that
we
can
include
in
our
minutes
so
we
can
be
sure
and
pass
that
message?

DR.
DENVER:
A
reference
to
the
previous
discussion
­­

DR.
ROBERTS:
No,
to
the
different
feeding
behaviors.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
297
DR.
DENVER:
Oh,
yes,
sorry.

Xenopus
tadpoles
are
filter
feeders
whereas
ranids
tend
to
feed
on
­­
try
to
sit
at
the
bottom
of
the
pond.
So
they
have
quite
different
modes
of
feeding.
And
so
their
rates
of
uptake
of
compounds
from
the
environment
may
be
different.

DR.
ROBERTS:
If
you
can
draft
a
couple
of
lines
or
two
and
throw
in
a
few
citations
for
a
report,
that
would
be
very
helpful.

Any
other
comments
on
this?

DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
Maybe
if
you
want
to
generalize
the
effects
on
sexual
differentiation
on
amphibians,
I
think
I
would
like
to
mention
that
we
have
also
another
order
of
amphibians,
oradales
(
ph),
which
have
different
androgens.

If
you
would
like
to
have
some
comprehensive
studies,
of
course
I
agree
fully
that
you
should
go
ahead
to
start
out
with
xenopus.
But
if
you
would
expect
something
different,
then
you
should
more
go
to
oradales
to
look
in
this
­­
all
of
amphibians.

Because
they
have
probably
also
a
little
bit
differences
or
more
pronounced
differences
concerning
sexual
differentiations
in
comparison
to
xenopus
and
ranids.

DR.
ROBERTS:
Dr.
Delorme.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
298
DR.
DELORME:
Just
a
comment
on
Dr.
Denver's
little
discourse
there
on
uptake
from
feeding.

I
don't
know
if
there
is
any
data,
but
it
might
be
worthwhile
looking
at
the
relative
contribution
from
food
and
bioconcentration
from
the
water.

I
don't
know
if
one
would
outweigh
the
other,
but
certainly
it
is
a
thing
to
consider,
something
to
consider
in
the
exposure,
because
we're
really
treating
the
water
here,
not
the
food.

DR.
ROBERTS:
Any
other
points
or
comments
on
this,
on
the
appropriateness
of
the
models?

Let's
go
ahead
and
take
G,
then.

DR.
STEEGER:
Assuming
xenopus
laevis
and
rana
are
sufficiently
concordant
from
a
toxico
dynamic
perspective
with
regard
to
potential
developmental
effects
of
atrazine,
what
critical
toxico
kinetic
processes
should
be
considered
for
extrapolating
xenopus
laevis
dose
response
relationships
to
rana
and/
or
for
designing
subsequent
studies
with
rana.

DR.
ROBERTS:
Who
wants
to
tackle
this
one?
Dr.
Delorme,

then
Dr.
Green,
then
Dr.
Coats.

DR.
DELORME:
I
think
we
already
touched
on
one
of
them.

That's
uptake.
The
second
one
would
be
looking
at
whether
or
not
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
299
degradation
or
the
depuration
from
the
animal
is
similar.
I
can't
tell
you.

DR.
ROBERTS:
Dr.
Green.

DR.
GREEN:
I
can't
answer
this
directly,
but
I
know
from
looking
at
pharmacokinetic
studies
again
for
the
purpose
of
trying
to
apply
veterinary
drugs
to
treat
sick
animals
that
there
are
differences
between
rana
pipiens
and
mammals.
There
have
been
some
comparisons
in
small
rodents.

Differences
between
rana
pipiens
and
X.
laevis
in
pharmacokinetic
studies,
the
comparisons
aren't
usually
directly
made.
But
the
margin
of
safety
for
many
drugs
has
been
proposed
to
be
much
lower
for
X.
laevis
because
it's
a
fully
aquatic
species,
which
means
that
when
they
get
sick
or
weak,
they
can't
get
up
to
the
surface
to
gulp
air.

So
they
die
from
drowning,
which
I
might
envision
that
if
at
the
higher
ends
of
this
dose
range
that
we
propose
for
atrazine,
the
mortality
perhaps,
maybe
not,
could
be
higher
in
the
juvenile
and
older
animals
that
are
fully
aquatic
at
that
point,
because
they
will
get
sick
and
weak,
and
not
being
semiterrestrial
like
rana
pipiens,
you
might
see
more
of
them
die
earlier
on
in
the
study
than
you
would
for
rana
pipiens.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
300
That's
speculative
on
my
part.
But
there
is
some
suggestion
in
the
literature
that
when
it
comes
to
some
veterinary
drugs,
that's
what
happens
to
xenopus
laevis.

So
you
can't
apply
a
drug
dose
for
a
rana
to
a
xenopus
to
treat
a
particular
condition
without
being
aware
you
could
kill
the
xenopus.

DR.
ROBERTS:
Dr.
Coats.

DR.
COATS:
I
agree
there
is
basis
to
probably
expect
different
pathways
of
degradation.
Most
of
them
probably
are
detoxifications.

There
is
possibility
of
some
of
them
not
being
at
detoxification
and
still
resulting
in
a
molecule
that
would
be
bioactive.
From
the
uptake
perspective,
I
agree
that
the
water
and
the
food
are
both
important
probably
and
need
to
be
studied.

And
certainly
absorption
through
the
skin
is
something
relatively
unique
or
at
least
feasible
that
would
have
to
be
looked
at
I
think
in
the
aquatic
forms.

DR.
ROBERTS:
Just
to
comment
on
my
part,
I
think
you
could
do
a
lot
of
toxico
kinetic
studies
in
terms
of
absorption
and
depuration.
But
an
easy
thing
to
do
to
sort
of
see
what
the
summation
of
those
processes
are
is
to
look
at
the
tissue
levels
in
animals
at
different
concentrations,
which
really
gives
you
sort
of
the
integration
of
the
intake
and
outflow.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
301
And
we
have
already
talked
about
the
desirability,
if
possible,

of
putting
that,
getting
that
information
as
part
of
the
study
so
you
can
see
whether
or
not
at
a
particular
concentration
of
atrazine
and
water
you
get
the
same
concentrations
in
xenopus
as
you
do
in
rana.

On
a
very
simple
level,
that
might
be
very
useful
for
extrapolations.

Dr.
Delorme
and
then
Dr.
Kloas.

DR.
DELORME:
I
guess
my
only
concern
with
this
would
be
that
the
development
of
the
different
organs,
especially
the
liver,

which
is
probably
going
to
be
the
major
detoxification
organ,
you
have
to
be
aware
of
any
differences
between
the
two
species
and
how
that
can
affect
what
is
going
on.

I
don't
know
how
you
are
going
to
do
that.
I'm
not
a
developmental
biologist.
It
theoretically
I
guess
could
have
an
effect.

DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
For
toxico
dynamics
I
think
during
the
larval
developments
there
shouldn't
be
a
big
difference
between
ranid
and
xenopus.

But
after
metamorphosis,
the
skin
of
xenopus
is
relatively
impermeable.
So
there
should
be
a
big
difference
in
toxico
dynamics,

especially
if
some
experiments
for
the
modes
of
actions
might
become
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
302
designed
for
juvenile
or
adults.
This
should
be
taken
in
account
that
there
is
severe
differences
between
ranid
species
versus
skin
still
that
keeps
on
being
quite
permeable
in
xenopus.

DR.
ROBERTS:
Good.
Any
other
comments
or
suggestions
on
this
point?

Dr.
Bradbury,
is
the
input
reasonably
clear?

DR.
BRADBURY:
Yes.

DR.
ROBERTS:
This
was
the
last
part
of
the
last
question.
But
during
the
course
of
some
discussions,
I
think
there
have
been
some
aspects
of
potential
experiments
that
the
panel
has
been
anxious
to
recommend
or
offer
their
advice
in
terms
of
how
some
of
these
studies
should
be
performed.
We
have
already
heard
some
of
them.
And
I
think
Dr.
Green
has
provided
some
excellent
suggestions
on
issues
in
terms
of
practicalities
of
doing
studies
and
things
that
need
to
be
considered.

Maybe
if
we
could
go
back
to
the
slide
which
sort
of
laid
out
the
things
in
some
of
the
initial
phase
one
studies
to
sort
of,
to
serve
as
prompts.

I
will
ask
the
panel
now
as
we
sort
of
move
beyond
the
questions
posed
to
use
if
there
are
any
specific
suggestions
that
they
might
have
that
haven't
been
mentioned
so
far.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
303
Dr.
Green
and
LeBlanc.

DR.
GREEN:
I
think
we
talked
about
water
quality
issue
and
nutrition
and
feeding
a
little
bit.
But
the
issue
of
what
kind
of
tanks
these
studies
would
be
performed
in,
whether
they
were
flow
through
or
static
renewal,
I
don't
know
if
that
was
resolved.

I
was
just
wondering
if
the
EPA
were
to
conduct
these
studies
in
their
lab,
and
I'm
not
saying
that
they
would,
what
kind
of
tanks
are
routinely
used
for
such
studies?

DR.
TIETGE:
We
typically
use
glass
tanks
using
flow
through
conditions.

DR.
GREEN:
What
is
the
water
turnover
rate
in
the
flow
through?

DR.
TIETGE:
In
terms
of
flow
rate,
it
is
25
mils
per
minute,

which
is
about
36
liters
per
day.

DR.
GREEN:
And
the
total
volume
of
the
tank
is
how
much?

DR.
TIETGE:
In
that
particular
system,
the
standing
tank
volume
I
believe
is
four
liters
and
our
stocking
density
is
from
20
to
25
organisms.

And
if
you
calculate
the
maximal,
the
approximate
maximal
weight,
the
total
loading
approaches
the
ASTM
standard
of
one
gram
per
liter
per
day.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
304
DR.
GREEN:
I
don't
have
a
feel
for
how
difficult
this
would
be
for
labs
that
might
be
asked
to
perform
these
studies
to
reproduce.

DR.
TIETGE:
Well,
I
kind
of
got
prepared
for
this.
I
was
anticipating
these
questions
last
night
while
everybody
else
was
awake.

I
have
a
brief
little
presentation
here.

Perceived
problems.
We
hear
a
lot
of
different
problems
brought
up.
And
one
is
system
costs.

Actually,
system
costs
are
not
necessarily
high
because
there
are
proportional
diluter
devices
which
work
on
hydrolic
principles
that
really
cost
probably
a
few
hundred
dollars
to
build.

These
have
been
published.
There
is
published
designs
for
these
that
date
back
to
the
70s,
I
believe.
One
of
them
is
called
the
Mount
Brungs
(
ph)
Deluter.
There
is
a
Banoit
(
ph)
Deluter.
These
kinds
of
technologies
are
very
old.
We
have
many
of
them
in
our
laboratory.

I
think
the
notion
of
cost
is
often
considered
an
impediment
because
there
are
fancier
systems
that
can
be
computer
controlled
and
can
do
very
complicated
exposure
or
can
be
used
to
achieve
very
complicated
exposure
designs.

But
the
basic
systems
can
be
very
inexpensive
and
very
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
305
reproducible
because
you
are
relying
on
gravity
to
make
them
work.

They
are
very
low
cost
systems.

There
are
concerns
about
operational
costs.
For
example,
waste,

chemical
costs
in
terms
of
the
test
chemical,
which
I
don't
think
in
this
case
would
be
a
great
impediment,
and
then
costs
for
water.
And
I'll
get
into
a
few
of
those
a
little
bit
more
in
a
moment.

We
often
hear
that
it
is
suboptimal
for
anurans.
And
I
guess
I'm
wondering
what
the
biological
basis
is
for
that.
I
have
read
anecdotally
that
it
is
a
suboptimal.
All
I
know
is
empirically
it
seems
to
work
pretty
well.
So
if
someone
has
a
biological
basis
for
that
­­

Dr.
Kelley
mentioned
yesterday
that
stimulation
of
lateral
line
could
relate
to
­­
could
relate
to
some
stress
response,
and
I
have
seen
that
in
the
literature,
but
I
have
never
seen
it
actually
documented.
It
seems
to
be
kind
of
an
informal
opinion.

Just
a
quick
history
about
these
methods.
As
I
mentioned,
some
of
the
technology
for
doing
the
flow
through
methods
were
developed
back
in
the
1970s
at
the
end
of
the
period
of
aquatic
toxicology
where
it
was
called
the
kill
them
and
count
them
period
where
acute
lethality
in
short
term
static
tests
were
kind
of
the
norm.

As
water
quality
criteria,
for
example,
became
more
complicated
or
more,
what
do
I
want
to
say
­­
well,
we
started
to
move
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
306
toward
more
sophisticated
approach
to
water
quality
criteria.
And
we
began
using
studies
that
were
more
capable
of
dealing
with
chronic
and
subchronic
sublethal
endpoints.

That's
where
flow
through
systems
really
became
developed,

was
to
achieve
those
ends.
And
when
you
get
into
more
subtle
effects,
such
as
endocrine
disruption,
these
types
of
systems
provide
highly
reproducible
exposures,
which
minimizes
the
variance
in
the
system.

Currently,
EPA
requires
flow
through
studies
in
different
offices
of
at
least
12
species.
We
just
kind
of
threw
these
numbers
together
last
night.
And
typically,
there
is
300
studies
submitted
annually.
These
are
primarily
fish
studies.

But
the
point
is
that
there
is
adequate
facilities
in
the
research
community
to
do
this.
There
is
the
toxicological
expertise
and,
in
fact,
guidance
such
as
in
the
ASTM
guidelines
on
how
to
conduct
these
study
was
amphibians
as
well
as
fishes.

In
general,
at
our
laboratory,
we
also
maintain
a
very
large
database
for
aquatic
toxicity.
And
I
can
tell
you
that
there
is
thousands
of
studies
that
have
been
conducted
over
the
last
20
years
using
flow
through
methods
with
at
least
30
aquatic
species.
This
is
not
anything
new.
That's
my
point
here.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
307
So
there
is
two
main
lines
of
reasoning
to
use
flow
through
methods.
The
first
is
the
biological
rationale.
Maintenance
of
water
quality.
These
issues
have
been
brought
up.
Temperature,
pH,
et
cetera.

Reduced
stress
caused
by
repeated
exchanges
or
handling
of
organisms
during
static
renewal
studies.
In
a
flow
through
study,
you
don't
need
to
do
that.

And
with
most
species,
at
least
the
species
that
we
have
worked
with,
these
methods
have
been
demonstrated
to
promote
survival,

growth
and
development.

Now,
the
toxicological
rationale
for
the
flow
through
methods,

first
of
all,
is
the
maintenance
of
a
stable
chemical
concentration.

Remembering
that
in
this
particular
case
we're
not
trying
to
achieve
an
ecological
approach,
it
is
a
toxicological
approach,
to
reduce
the
variables
involved
so
we
can,
you
know,
assess
the
chemical
in
a
tightly
controlled
system.

It
is
particularly
good
for
highly
hydrophobic
chemicals
because
it
eliminates
the
mass
limitations
that
occur
when
you
are
dealing
with
chemicals,
let's
say,
that
are
in
the
6
to
7
KOW
range
with
labile
chemicals
which
may
be
­­
will
degrade
due
to
metabolism,
hydrolysis,
photolysis,
or
they
might
be
volatile
in
a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
308
system.

It
also
has
the
added
benefit
that
it
reduces
or
eliminates
the
accumulation
of
chemical
metabolites.
That's
of
the
parent,
which
was
brought
up
as
an
issue
yesterday.
And
we
get
improved
dose
response
or
concentration
response
data
from
that.

So
at
our
laboratory
in
Duluth,
we
have
flow
through
methods
that
we
have
used
for
xenopus
laevis.
And
we
have
gone
through
some
of
these
baseline
studies
that
have
been
mentioned
previously.

We
have
done
loading
studies
to
evaluate
whether
or
not
we
needed
to
approach
the
ASTM
or
utilize
the
ASTM
recommendations.

And
indeed,
when
we
pushed
the
loading
organism
performance
based
on
mostly
growth
and
development,
it
became
problematic.
So
we
have
stuck
with
the
ASTM
guidelines.

We
have
done
some
feeding
comparisons.
I
could
get
into
the
detail
if
anybody
is
interested.
And
we
have
done
some
baseline
developmental
studies
looking
at
developmental
rate.

One
of
the
added
improvements
that
we
see
in
the
flow
through
is
we
have
better
developmental
synchrony,
which
makes
the
tests
a
little
bit
easier
to
conduct
and
reduces
some
variation.

If
you
use
the
ASTM
standards,
you
can
conduct
static
tests
or
you
can
conduct
flow­
through
tests.
And
I
have
listed
the
biological
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
309
loading
rates
for
the
two
there.

And
the
reason
I
was
able
to
pull
that
number
off
the
top
of
my
head
a
little
bit
earlier
is
because
I
went
through
the
calculations
last
night.
In
our
studies,
we
typically
use
25
organisms.
Their
maximal
weights
are
about
one
and
a
half
to
1.8
grams.
Occasionally,
we'll
get
a
2
gram
organism.
That
usually
occurs
at
about
stage
61,
62,
I
think.

So
we
base
our
loading
on
our
maximum
weights.

And
as
you
can
see,
if
you
go
through
the
math,
we're
at
about
one
gram
per
liter
per
day.

One
of
the
things
to
note
here
is
if
you
choose
the
static
route
and
you
are
going
to
adhere
to
this
guideline,
because
we're
working
in
the
absence
of
other
validated
protocols,
then
you
will
see
that
if
you
want
to
run
25
organisms,
I
did
the
calculation
just
for
comparison,
you
would
need
75
liters
per
day
for
those
25
organisms
as
opposed
to
in
a
flow
through
condition
where
you
would
only
need
38
liters
per
day.

So
I
think
if
you
make
the
decision
that
you
are
going
to
stick
to
the
guidance
that
was
developed
and
first
published
in
1980
and
it
had
­­
it
has
taken
a
lot
of
­­
taken
advantage
of
a
lot
of
information
that
was
developed
up
to
that
point,
then
I
think
it
is
more
efficient
to
go
with
the
flow
through.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
310
At
our
laboratory,
we
have
conducted
approximately
35
toxicological
studies.
If
you
are
interested,
12
to
15
of
those
studies
are
published,
and
there
are
several
in
press.
But
the
thing
I
want
to
point
out
here
is
that
we
have
worked
with
xenopus
laevis,
rana
pipiens,
clamitans,
sylvatica,
septentionalis.
And
all
those
species
are
amenable
to
these
methods.

Some
are
better
than
others.
I
know
someone
was
complaining
about
a
doing
green
frog
work.
I
kind
of
had
to
smile
because
it
just
takes
them
forever
to
develop.
So
I
wouldn't
recommend
doing
green
frog
or
mink
frog
work
with
these
methods.
But
they
can
be
done
if
you
are
studying
larval
period.

But
if
you
want
to
go
through
metamorphosis,
I
think
the
xenopus
laevis,
rana
pipiens
and
rana
sylvatica
would
be
the
organisms
of
choice.

We
also
hold
in
culture
­­
all
of
our
cultures
flow
through.
And
we
have
currently
xenopus
laevis
and
xenopus
tropicalis
in
culture.

That's
all
in
flow
through
conditions.

I
would
be
happy
to
take
any
questions
on
that.

DR.
ROBERTS:
Are
there
any
questions
or
comments
on
the
flow
through
versus
static
issue?

Dr.
Green.
I
believe
Dr.
LeBlanc
had
indicated
previously
you
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
311
wanted
to
­­
no?
Then
Dr.
Kloas.

Dr.
Green.

DR.
GREEN:
I'm
in
full
support
of
using
a
flow­
through
system,
I
think
­­
for
the
reasons
you
just
pointed
out
there.

Although,
I
had
many
people
complain
about
having
to
do
that
seems
like
once
they
switched
they
are
overall
more
satisfied.

And
I
can't
find
anything
in
the
literature
either
except
anecdotal
reports
that
it
is
detrimental
to
the
development
of
either
xenopus
laevis
or
rana
pipiens.

One
thing
I
did
note
was
several
people
felt
like,
and
these
are
big
commercial
suppliers
of
these
stock
sources
for
these
frogs,
said
that
they
feel
that
laboratory
reared
and
conditioned
frogs
actually
adapt
quite
well
to
being
in
a
flow­
through
system.

You
wouldn't
want
to
take
wild
caught
frogs
or
tads
and
put
them
in
a
system.
They
probably
wouldn't
deal
with
it
as
well.

The
question
I
had
for
you
was
what
kind
of
water
is
it.
Is
it
reconstituted
R
O
treated
water?

DR.
TIETGE:
No.
The
water
we
use
is
for
the
most
part
unmodified
Lake
Superior
water,
which
is
a
relatively
low
conductivity
water.

It
goes
through
several
treatment
­­
well,
filtration
steps
and
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
312
UV
sterilization.
But
beyond
that,
it
is
Lake
Superior
water.

DR.
GREEN:
It
is
potable,
chlorinated,
filtered
water
that
goes
through
UV
sterilization?

DR.
TIETGE:
It
is
not
chlorinated.

DR.
GREEN:
Chloraminated?

DR.
TIETGE:
It
is
not
chloraminated.

This
is
taken
from
the
bottom
of
Lake
Superior.
We're
located
right
on
the
lake.
And
we
have
an
intake
that's
about
a
quarter
of
a
mile,
I
think,
from
the
laboratory
out
into
the
lake
where
there
is
a
specialized
filter
system
that
is
built
into
the
lake.

And
then
the
water
comes
in,
and
then
there
is
further
filtration,
and
then
the
final
step
is
UV
sterilization.
There
is
no
chemical
additives
in
the
water
whatsoever.

DR.
GREEN:
I
know
this
is
being
detailed,
but
if
we
were
to
try
and
reproduce
this
system
or
recommend
that
something
like
this
be
used,
when
you
talk
about
filtered,
do
you
know
­­
you
are
filtering
particulates.
Do
you
know
what
size
­­
the
reason
being
is
all
the
runoff
that
goes
into
the
lake
you
want
to
make
sure
you
get
pathogens
like
­­

DR.
TIETGE:
We
have
several
types
of
filters
involved
in
our
system.
First
of
all,
in
the
lake
itself,
there
is
a
graded,
gravel
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
313
sandbag
that
has
fairly
crude,
I
think
maybe,
25
micron,
I
think,
is
the
cutoff
range
for
that.
But
when
it
comes
into
the
building,
it
goes
through
some
five
micron
filters,
at
which
point
it
is
UV
sterilized.

And
then
prior
to
introduction
to
the
system,
it
goes
through
a
more
typical
laboratory
filter
that
I
think
is
probably
two
and
a
half
or
five
microns
or
in
that
range.

One
thing
I
meant
to
mention
earlier
in
terms
of
efficiency
is
from
experience,
I
can
tell
you
that
these
systems
require
much
less
labor.
And
we
actually
elected
to
conduct
a
static
renewal
study
about
two
years
ago
for
very
specific
purposes.

And
after
we
got
into
it,
we
really
regretted
it
because
it
was
so
much
work.
These
systems
require
very
low
maintenance.

DR.
GREEN:
That's
what
makes
me
a
bigger
fan
too,
low
maintenance.

DR.
TIETGE:
There
are
numerous
universities
and
contract
laboratories
that
use
these.
It
isn't
necessary
to
use
our
lake
water
supply.
There
is
numerous
examples
of
waters.

Of
course,
I
think
that,
if
you
wanted
to,
you
could
modify
a
water
with
additives,
salts,
whatever
using
the
appropriate
technology.

DR.
ROBERTS:
Dr.
Kloas.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
314
DR.
KLOAS:
Did
you
perform
­­
not
just
for
the
performance
of
the
animals,
this
is
quite
convincing.
I
agree
that
it
would
grow
and
so
on.

But
do
you
have
any
comparative
study
using
positive
controls
for
instance
in
endocrine
disruption,
estradiol
treatment,
something
like
this,
where
you
can
compare
static
renewal
system
with
flow
through
and
the
sensitivity
of
both
exposure
regimes?

DR.
TIETGE:
No,
because
we
don't
do
static
renewal.
Our
work
primarily
is
focused
on
thyroid
axis
disruption.
So
we've
used
numerous
chemicals
to
inhibit
or
to
stimulate
metamorphosis.

But
we
have
also
run
a
couple
of
chemicals,
nonaromatizable
androgen,
and
we
have
also
­­
we
did
see
androgenic
effects.
We
have
also
run
estradiol
where
we
do
see
­­
off
the
top
of
my
head,
I
don't
remember
the
concentrations
of
the
study,
but
I
know
that
we
do
have
feminization
in
that
case.

DR.
KLOAS:
Because
there
is
always
some
concern
about
if
it's
the
same
or
if
you
lose
sometimes
­­
maybe
you
can
also
lose
by
permanent
loading
of
humero
(
ph)
compliments,
you
may
lose
a
little
bit
sensitivities.

DR.
TIETGE:
Loss
of
sensitivity
based
on
what?
I'm
sorry.

DR.
KLOAS:
In
comparison
to
static
renewal
system.
I
can't
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
315
compare.
We
don't
have
flow
through.
But
if
you
compare,
some
people
found
­­
I'm
aware
of
one
or
two
studies
in
a
lab
with
a
flow
through.
They
use
same
concentrations
of
estradiol
but
are
getting
much
lower
sensitivity
concerning
feminization
of
xenopus
and
also
ranid
species.

DR.
TIETGE:
I
think
I'm
familiar
with
at
least
some
of
the
work
you
are
talking
about.
I'm
not
sure
if
they
used
­­
that
was
a
positive
control.
I'm
not
sure
that
they
had
analytical
verification
of
that.

We
can
talk
about
that.
But
I'm
not
sure.

One
of
the
things
that
also
I
should
mention
has
to
do
with
the
rate
of
development
in
some
of
the
studies
that
we
reviewed
for
the
white
paper.
The
developmental
rates
for
xenopus
were
very
long,
in
our
opinion.

And
in
our
laboratory,
using
the
flow
through
systems,
we
typically
have
metamorphosis
well
underway
at
seven
weeks.
About
50
days
post
fertilization.
And
it
is
usually
completed
within
three
or
four
days
after
that.

Certainly,
within
56
days,
we
are
generally
done
with
metamorphosis
at
that
point,
unless
there
is
a
chemical
effect,
and
then,
of
course,
that
is
another
story.
But
I'm
referring
to
controls
in
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
316
that
case.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
Joe,
I
agree
with
your
recommendations
for
a
flow­
through
system
as
well
as
recommending
the
proportional
diluters.

But
it
seems
that
proportional
diluters
scare
a
lot
of
people.
It
must
be
all
those
tubes
and
boxes
or
something.
I
was
just
wondering
if
there
are
alternatives
that
you
are
aware
of
in
terms
of
peristaltic
pump
systems
­­

DR.
TIETGE:
Yeah.
During
the
period
that
those
were
designed,
pump
systems
weren't
quite
as
reliable
as
they
are
today.

There
are
many,
many
reliable
pump
systems
that
can
be
used.

And
some
of
our
newer
designs
do
not
use
the
proportional
flow
through
devices.
They
have
simple
dilution
or,
I
should
say,
solution
cells
that
serve
as
a
stock
for
a
peristaltic
pump
that
might
have
six
or
eight
lines
coming
out
of
it
or
other
pumps.

We
have
very
many
systems
in
our
laboratory
that
represent
different
design
ideas.

I
think
when
you
get
into
a
more
complex
exposure
paradigm,

then
the
system
has
to
be
more
complicated
to
accomplish
it.
But
pumps
work
fine,
actually.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
317
DR.
ROBERTS:
Any
other
comments?
Dr.
Skelly.

DR.
SKELLY:
Just
real
quickly,
Dr.
Tietge.

We're
talking
about
this
I
assume
because
this
is
going
to
be
what
EPA
recommends
as
people
go
forward
or
requires.

I
think
Dr.
Bradbury
or
somebody
said
before
that
there
doesn't
necessarily
need
to
be
a
requirement
that
people
that
want
to
do
experiments
that
are
going
to
be
evaluated
and
used
by
EPA
to
make
judgments
will
use
one
method
or
another.

I
thought
it
would
be
helpful
if
you
could
comment
on
that.

DR.
TIETGE:
I
think
in
the
absence
of
a
validated
protocol,

one
needs
to
adhere
to
a
standard.
And
the
standard
is
right
now
this
most
appropriate
ASTM.
There
is
the
flow
through
recommendation.

And
there
is
the
static
renewal
recommendation.

I
think
if
you
take
a
look
at
the
two,
it
is
more
efficient
to
do
the
flow
through.
Because
for
ends
that
you
need
to
utilize
for
your
statistical
design,
I
think
you
would
be
in
a
lot
of
trouble
with
a
static
test.

DR.
ROBERTS:
Dr.
Skelly
I
think
wanted
to
respond.

DR.
SKELLY:
Just
to
follow
up.
Maybe
the
economics
of
working
in
a
government
agency
versus
a
university
can
be
a
little
bit
different
where
labor
can
be
quite
cheap
in
a
university
setting,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
318
people
getting
credits
or
whatever,
whereas
equipment
costs
whatever
it
costs.

I
guess
what
I'm
wondering
is
will
the
EPA
have
an
all
flowers
can
bloom
sort
of
attitude
towards
this?
Or
are
they
going
to
say
flow
through
or
nothing.

DR.
STEEGER:
I
think
one
of
the
things
we're
looking
to
see
happen
is
that
the
studies
that
are
conducted
for
regulatory
purposes
are
consistent
or
as
consistent
as
they
can
be
with
our
guidelines.

And
although
we
don't
have
any
guidelines
right
now
for
amphibian
studies,
we
do
have
the
A
50
guidelines
for
water
quality
standards.
And
we
would
hope
that
whatever
studies
are
conducted
adhere
to
those
guidelines.

DR.
BRADBURY:
I
think
some
of
the
sort
of
minimal
issues
in
terms
of
data
quality
that
we're
talking
about
are
the
ASTM
guidelines.
They
are
measured
concentrations.

As
Dr.
LeBlanc
mentioned
earlier
today,
if
it
is
a
static
renewal,
you
probably
need
to
do
more
analytical
chemistry
in
terms
of
numbers
of
samples
over
time
than
one
would
have
to
do
in
a
flow
through
test
to
ensure
that
you
really
know
what
your
concentrations
are.
Because
that's
one
of
the
other
challenges
of
a
static
renewal,

can
be,
depending
on
the
chemical
and
toxicokinetics,
that
aspect.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
319
So
I
think
it
is
more
of
the,
if
you
will,
performance
issues,

loading,
DO,
ammonia,
measuring
the
concentrations
that
are
the
real
important
issues
to
the
extent
that
some
studies
would
be
done
because
people
read
these
documents
and
think
they
would
be
like
­­

they
would
like
to
pursue
some
research,
they
will
have
some
ideas
on
the
kinds
of
quality
indicators
that
would
be
important
for
the
agency
to
use
­­
looking
at
­­
when
they
look
at
the
open
literature
to
the
extent
the
registrant
may
be
performing
some
of
these
studies.

That's
sort
of
a
different
venue
in
which
some
of
these
studies
potentially
could
be
done,
if
they
are
done,
which
gets
at
what
Tom
was
saying
in
terms
of
the
kind
of
guidelines
the
agency
requires
for
data
that's
submitted
as
part
of
registration
or
reregistration.

DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
Concerning
atrazine
effect,
obtained
up
to
now
they
have
been
just
obtained
in
static
renewal
systems.
Maybe
metabolism
of
atrazine
and
so
on
could
also
play
a
role
for
getting
the
effect.

If
there
is
really
concern
about
atrazine
and
­­
I
agree
with
several
advantages.
And
from
a
logical
point
of
view,
I
agree
completely.
I
would
also
like
to
have
flow
through
in
my
lab
to
try
doing
that.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
320
But
there
is
concern
about
what
was
done
and
results
obtained
up
to
now.
I
think
to
repeat
this
should
contain
both
at
least
static
renewal
and
maybe
an
advance
study
using
the
flow
through.
Because
otherwise,
we
cannot
rule
out
if
there
is
no
effect
in
a
flow
through
anymore.

DR.
BRADBURY:
One
point.
I
guess
we
would
be
curious
to
hear
the
panel's
response
in
terms
of
ASTM
loading
guidelines,

ammonia,
D
O,
those
kind
of
attributes
that
have
to
be
met
in
terms
of
evaluating
the
quality
of
the
study
regardless
if
it
was
done
flow
through
or
static
renewal.

I
think
the
agency
in
putting
out
the
document
to
get
a
response
was
first
understanding
your
impressions
about
basic
data
quality
parameters.
How
you
get
there.
It
could
be
static
renewal.
It
could
be
flow
through.

DR.
ROBERTS:
Dr.
Delorme,
then
Dr.
Green.

DR.
DELORME:
Dr.
Bradbury
just
answered
my
question.
He
said
that
you
had
to
meet
the
water
quality
standards
that
have
been
set
out.

But
just
as
another
note
for
the
panel,
as
somebody
who
does
risk
assessments
and
gets
studies
in,
typically,
when
you
get
a
study
in
on
fish
or
invertebrates,
they
have
to
provide
information
on
the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
321
water,
which
includes
the
presence
of
other
contaminants,
be
they
heavy
metals,
pesticides
or
whatnot.

So
the
water
is
characterized
typically
at
the
labs
that
do
the
studies,
whether
it
be
an
industry
lab
or
a
contracting
company.

DR.
ROBERTS:
Dr.
Green.

DR.
GREEN:
I
think
Dr.
Kloas's
point
is
well
taken
too.
There
may
be
something
inherent
with
the
static
renewal
system
that
mimics
wildlife
pond
situations
closer
than
the
flow
through.
Particularly,

with
regards
to
ammonia
and
other
things
in
the
water
at
higher
levels.

And
some
of
the
studies
we
looked
at
yesterday
clearly
had
higher
levels
of
ammonia.
And
that
might
play
a
role
in
interaction
with
atrazine
that
if
we
put
them
in
flow
through
systems
and
we
adhere
to
the
guidelines,
we
might
not
see
that.

If
it's
feasible,
it
seems
a
reasonable
thing
to,
if
you
don't
get
results
in
the
flow
through
system
that
are
repeatable,
that
perhaps
a
static
renewable
system
would
be
on
a
smaller
scale
something
to
look
at
as
well.

DR.
ROBERTS:
Dr.
Bradbury.

DR.
BRADBURY:
I
want
to
get
clarification.

If
our
question
at
hand
is
to
evaluate
the
hypothesis
that
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
322
atrazine
can
cause
developmental
effects,
sort
of
putting
on
my
toxicology
hat,
I
would
like
to
try
to
get
a
system
that
gives
me
the
cleanest
possible
way
to
interpret
whether
or
not
I'm
getting
a
toxicological
signal.
And
then
I'll
get
more
complex
as
I
need
to.

I
guess
I
would
like
to
get
a
response
to
that.

DR.
ROBERTS:
I
think
Dr.
Green's
point
is
if
the
answer
is
if
you
get
a
negative
response,
it
is
possible
that
atrazine
under
conditions,
under
something
less
than
clean
conditions
that
may
be
more
relevant,
in
fact,
to
the
environment
might
give
a
positive
response,
you
might
want
to
check
that
before
you
stop,
I
guess.

I'm
not
speaking
for
Dr.
Green.
But
that's
what
I
think
the
point
is.

DR.
GREEN:
I
think
that's
right.
Otherwise,
we'll
be
open
scientifically
to
the
criticism
that
it
wasn't
a
static
and
you
get
negative
results
on
a
flow
through,
and
we're
back
to
phase
one
again.

DR.
BRADBURY:
Hopefully
if
some
of
our
discussions
about
the
hill
criteria
hold
up,
we'll
­­

DR.
ROBERTS:
Dr.
Richards.

DR.
RICHARDS:
This
is
partly
a
question,
partly
response.

I
think
maybe
to
step
back
a
bit,
some
of
the
people
I
have
heard
over
the
last
few
days
express
that
the
static
was
more
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
323
appropriate
because
maybe
it
more
mimicked
a
pond­
type
situation.

I
was
trying
to
think
back.
I
don't
know
a
great
deal
about
the
development
of
the
ASTM
standards,
but
when
I
think
of
all
the
species
of
fish
that
have
been
used
in
the
development
of
those
standards,
how
many
of
them
might
you
place
into
the
category
of
being
species
that
might
inhabit
these
mesotrophic
or
eutrophic
sorts
of
ponds
be
exposed
to
relatively
high
ammonia
levels,
and
that
might
be
considered
a
normal
environment,
as
would
be
xenopus
and
possibly
rana
pipiens
too.

DR.
STEEGER:
Our
test
species
are
chosen
again
because
of
their
ability
to
be
raised
under
laboratory
conditions.

There
are
large
mouth
bass
and
blue
gill
sunfish.
There
are
other
choices
we
get.
Fathead
minnows.

They
can
live
under
eutrophic
conditions.
But
those
are
challenges
that
can
impact
the
signal
that
Steve
is
talking
about
in
terms
of
just
trying
to
create
a
situation
where
we
have
removed
as
many
of
those
variables
as
possible
to
have
as
clean
a
signal
as
possible,
is
what
the
agency
uses
for
regulatory
purposes.

Otherwise,
it
gets
to
be
very
difficult
to
sort
out
what
the
real
cause
of
effect
was.

DR.
RICHARDS:
In
partial
response,
am
I
hearing,
then,
that,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
324
yes,
other
organisms
like
fatheads
and
so
forth
that
do
live
often
times
in
mesotrophic
or
eutrophic
situations
seem
to
respond
very
well
and
repeatably
in
flow
through
conditions.

DR.
ROBERTS:
I
guess
in
response
I
would
just
say
that
if
you
find
yourself
in
the
situation
where
you
use
flow
through
and
you
don't
get
a
response,
the
question
will
be,
and
I
think
I
would
be
sort
of
compelled
to
go
back
and
see
whether
or
not
the
reason
you
don't
get
a
response
has
something
to
do
with
flow
through
versus
static
conditions,
and
then
the
question
will
come
up,
which
is
more
relevant.

I
don't
know
­­
of
course,
until
you
understand
the
basis
for
the
difference
between
static
and
flow
through,
I
don't
suppose
you
could
answer
that
question,
but
we
can
envision
the
situation
where
you
would
be
sort
of
drawn
down
that
path.

And
if
you
can't
duplicate
it
in
static
conditions,
you
can
just
say,
well,
we
can't
duplicate
it
under
static
or
flow
through.

If
you
get
it
under
static,
you
don't
get
it
under
flow
through,

then
you
are
going
to
have
to
figure
out
why
and
figure
out
which
one
is
more
relevant.

I
guess
we're
just
saying
that.

DR.
BRADBURY:
Can
I
ask
one
more
question?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
325
Let's
say
we
go
down
a
path,
you
do
a
flow
through
and
you
don't
get
the
effect.
Maybe
that
has
something
to
do
with
the
physics
of
a
static
renewal
is
what
is
important
to
triggering
whatever
event
occurs.

This
is
just
because
I
couldn't
remember
the
dialogue.
Would
it
still
be
the
recommendation
of
the
panel
that
one
would
try
to
meet
ASTM
standards
in
running
that
static
renewal?

DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
I
think
it
should
be
the
same
conditions
whereas
people
obtained
the
effect.

I'm
completely
surprised
if
I
can
­­
25
animals
and
75
fetus
or
something.
It
is
hard
to
handle.
But
the
effect
already
observed
have
been
used
in
volumes
of
four
liters
or
something
like
that.

I
think
it
wouldn't
make
sense
to
use
another
static
renewal
system.
Really
in
agreement
with
ASTM
standards.
I'm
a
dirty
endocrinologist.
I'm
sorry.
I
would
like
to
have
a
study
repeated
under
the
same
conditions.
Whatever
it
might
make
any
difference.

I
would
like
to
have
flow
through
and
to
see
that
there
are
results.
But
if
you
change
in
another
way
and
doing
static
renewal
under
different
conditions
and
you
can't
repeat
again,
it
is
the
same
question.
If
you
can't
repeat
it
in
flow
through,
then
you
are
again
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
326
next
step
and
one
step
backwards.

DR.
BRADBURY:
I
just
have
this
­­
I
get
a
feeling
that
I'm
going
down
a
path
that
gets
us
back
to
some
of
the
white
paper's
analysis
that
it
was
very
difficult
to
interpret
the
responses
from
the
previous
studies
because
of
concern
over
the
ammonia
levels,
the
feeding
issues,
the
DO
issues.

Then
we're
being
asked
to
try
to
replicate
scenarios
that
­­
I'm
trying
to
track
the
­­
I
feel
like
we've
started
getting
into
a
do­
loop.

DR.
ROBERTS:
I
guess
Dr.
Green
can
respond.

We
got
I
think
some
responses
by
at
least
a
couple
different
labs.
I
don't
know
they
all
had
real
high
ammonia
levels.
I
don't
know
that
that
­­
I
don't
know
that
you're
going
to
have
to
really
make
awful
conditions
to
get
­­

DR.
BRADBURY:
I
think
some
guidance
on,
if
not
ASTM
guidelines,
where
do
you
feel
a
reasonable
static
renewal
in
terms
of
those
issues,
at
least
bracket
that.

DR.
ROBERTS:
Sure.
Dr.
Green.

DR.
GREEN:
I
don't
think
our
intent
was
to
recreate
the
dirty
water
with
a
static
renewal
system.
That
wasn't
it
at
all.

We
have
static
renewal
systems
that
house
400
and
500
adult
frogs.
No
matter
how
hard
we
work
on
it,
we
cannot
produce
ASTM
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
327
clean
water
to
that
degree.

I
think
what
is
important
is
the
dynamics
of
the
animal
sitting
and
reabsorbing
the
drug
that
they
are
absorbing
and
excreting
for
the
few
hours
that
they
do
it
in
between
water
changes
and
for
the
gradual
increase
in
ammonia.

And
certainly
they
don't
have
to
go
up
to
the
toxic
lethal
levels
that
one
of
those
papers,
I
believe,
had
in
it,
but
that
system
alone
would
be
something
I
think
­­
would
be
a
justifiable
attempt
to
reproduce
the
results
they
got
in
the
static
system.

Try
to
keep
the
water
as
clean
as
possible.
But
you
are
never
going
to
get
it
as
clean
as
you
would
with
the
flow
through
system.
I
would
say
change
the
water
as
they
did
in
the
static
renewal
system,

as
frequently
as
you
need
to
to
keep
the
ammonia
down
and
all
that
stuff.

And
it
is
still
going
to
go
up
enough
that
it
is
only
speculative
that
the
interactions
in
the
water
chemistry
dynamics
will
be
there
in
a
way
that
they
are
not
there
in
the
flow
through
system.

DR.
ROBERTS:
Dr.
LeBlanc.

DR.
LEBLANC:
It
just
seems
like
we're
getting
into
the
realm
of
what
if,
what
if
we
don't
­­
what
if
we
get
a
negative
result,
what
do
we
do.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
328
I
just
don't
know
that
we
should
be
crossing
that
bridge.
I
recognize
you
need
recommendations
and
you
are
looking
for
advice
and
help.
But
I
think
we
should
proceed
with
the
assumption
that
we
will
replicate
the
results.

And
if
we
don't,
at
that
point
we
have
to
look
at
the
new
data
as
compared
to
the
old
data
and
make
some
judgments
as
to
why
perhaps
we
can't
replicate
the
results.
And
then
design
some
basic
experiments
to
address
those
variables.
Leave
it
at
that.

DR.
ROBERTS:
I
want
to
ask
if
there
are
there
other
comments
by
panel
members
regarding
any
of
the
topics
­­
this
topic
and
the
discussions
we
have
had
over
the
last
few
days,
any
points
that
are
important
you
think
that
have
not
been
made
as
yet
but
that
would
be
important
to
make
and
introduce
into
the
minutes.

Dr.
LeBlanc.

DR.
LEBLANC:
Quick
technical
point.
The
experimental
design
that
was
up
there,
you
had
a
positive
control
of
estradiol.
We
have
seen
some
experiments
with
DHT.
I
don't
see
any
reason
why
we
would
be
looking
at
DHT
as
a
positive
control.

But
we
are
interested,
I
think,
in
some
anti­
androgenic
effects,

not
necessarily
due
to
the
ability
of
atrazine
to
compete
at
the
receptor
level,
but
perhaps
to
interfere
with
conversion
of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
329
testosterone
to
DHT
or
just
to
overall
cripple
the
synthesis
of
hormones.

So
you
might
give
consideration
to
a
control
for
anti­
androgenic
properties.
I
guess
you
would
have
to
look
at
an
anti­
androgen.
And
one
that
I
have
seen
in
the
literature
with
frogs
and
I
think
Darcy
or
Sherril
mentioned,
it
was
cyproterone
acetate.

That
would
be
something
to
consider.

DR.
ROBERTS:
Dr.
Kloas.

DR.
KLOAS:
I
can
add
three
suggestions.
You
can
use
cyproterone
acetate,
which
works
as
an
anti­
androgen,
or
you
can
use
para
para
DDE.
Also,
vinclozolin
would
work
also.

DR.
ROBERTS:
Dr.
Gibbs.

DR.
GIBBS:
I
still
have
lingering
concerns
that
far
too
individuals
use
the
found,
the
experimental
larval
populations
in
some
of
these
laboratory
studies.

It
is
not
my
area
of
expertise,
but
as
a
population
geneticist,
it
really
concerns
me
when
three
pairs
are
used.
One
onymous
individual
with
three
pairs
will
potentially
skew
an
entire
experiment.

I
think
boosting
numbers
up
into
the
10s
of
pairs,
I
think
it
seems
like
a
good
idea
to
me.

DR.
SKELLY:
Just
a
point
that
has
been
touched
on
a
little
bit,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
330
but
I
wanted
to
make
sure
it
made
it
clearly
into
the
record.

That
is
that
it
may
be
standard
practice
in
toxicology,
but
I
was
struck
by
the
kind
of
fluctuating
use
of
different
levels
of
statistical
units.

In
some
cases,
people
were
using
individuals.
In
some
places,

people
are
using
tanks
as
their
units
of
analysis,
even
though
all
the
individuals
are
raised
in
group
tanks.

And
in
ecology,
that's
just
totally
tabu.
You
don't
do
that.
If
things
are
raised
in
a
tank,
they
can
influence
each
other.
And
you
don't
use
individuals
as
replicates
in
analysis
if
they
came
out
of
the
same
tanks.

I
don't
pretend
an
SAP
can
change
practice
in
an
entire
field.

But
if
that
is
standard
operating
practice
and
if
we
are
going
to
get
into
ecological
relevance
at
some
point
and
ecologists
are
going
to
be
evaluating
this
stuff,
they
are
going
to
care
about
that
a
whole
lot.

That
alone
will
eliminate
a
paper
from
peer
reviewed
literature
in
ecology.

DR.
ROBERTS:
Any
other
suggestions
from
panel
members?

Dr.
Bradbury,
I
think
the
panel
has
given
you
our
best
advice
on
this
subject.
Are
there
any
final
clarifications
or
follow­
up
questions?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
331
DR.
BRADBURY:
No.
I
think
it
has
been
a
very
helpful
discussion
this
afternoon
and
mid
morning.
I
appreciate
all
the
hard
work
and
input.

DR.
ROBERTS:
I
think
as
came
through
in
our
comments,
I
think
the
panel
was
very
impressed
with
the
job
that
the
agency
had
done
on
the
white
paper
and
your
analysis
and
your
thoughts
on
how
to
move
forward.

I
think
we
thought
in
general
you
did
a
very
good
job
with
that.

We
had
some
suggestions
here
and
there
for
things
maybe
to
add
and
consider.
But
overall,
we
thought
the
agency
did
an
excellent
job
on
a
very
difficult
subject
area.

Before
we
close
the
meeting,
I
would
also
like
to
thank
our
public
commenters.
We
had
many
of
them
that
travelled
sometimes
a
great
distance
to
come
and
speak
to
us
and
share
with
us
their
data
and
their
viewpoints
on
this
subject.
We
always
welcome
different
perspectives
and
viewpoints
on
the
issues
that
we
face.
And
we
thank
the
public
commenters.

I
would
also
like
to
thank
the
hearty
soles
in
the
audience
who
have
stayed
with
us
now
for
a
number
of
days
through
our
presentations
and
our
discussions.
I
like
to
thank
you.

And
of
course,
always,
I
would
like
to
thank
the
SAP
staff
for
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
332
putting
this
meeting
together.
They
assembled
an
outstanding
panel
of
experts,
got
us
all
here,
got
us
all
the
documents
we
needed
and
have
been
very
supportive.

Finally,
I
would
like
to
thank
the
panel
members,
a
very
impressive
panel.
Your
expertise
was
obvious.
The
fact
you
came
prepared
was
obvious.
We
have
had
excellent
discussion.
Excellent
questions
right
from
the
beginning.
I
think
you
guys
did
a
terrific
job.

Do
we
have
any
other
announcements
before
we
close
the
session?
Paul,
anything
you
need
to
say?

MR.
LEWIS:
Just
a
few
remarks.
I
want
to
first
thank
Dr.

Roberts
for
serving
as
our
chair
for
the
past
three
days.
Did
an
excellent
job
in
managing
the
process
for
this
meeting.

Members
of
the
panel,
it
was
a
pleasure
working
with
all
of
you.
And
looking
forward
as
we
work
in
writing
our
report
that
for
the
public's
interest
will
be
available
in
about
four
weeks,
available
on
our
SAP
web
site
and
also
in
the
docket.

And
I
thank
the
members
of
the
public
for
listening
to
our
discussion
and
for
the
contributions
they
made
as
part
of
the
deliberations
that
we
had
in
the
past
three
days.

Finally,
my
colleagues
at
EPA,
both
in
EFED
(
ph)
and
my
1
2
3
4
5
6
7
8
9
10
11
12
333
colleagues
with
the
SAP
staff
in
working
with
me,
pleasure
working
with
you
in
getting
this
meeting
off
the
ground.
Thank
you.

Dr.
Roberts.

DR.
ROBERTS:
Immediately
following
the
close
of
this
session,
I
would
like
the
panel
to
meet
in
the
meeting
room
for
a
short
closed
session
just
to
discuss
the
logistics
of
writing
up
our
minutes.

As
soon
as
we're
finished
here,
if
we
could
convene
there
for
a
short
meeting,
that
would
be
great.

If
there
no
other
announcements
or
no
other
topics
to
discuss,
I
would
like
to
now
close
this
session
of
the
FIFRA
Scientific
Advisory
Panel.

(
Thereupon,
the
session
concluded
at
6
p.
m.)
1
2
3
4
5
6
7
8
334
CERTIFICATE
OF
STENOTYPE
REPORTER
I,
Frances
M.
Freeman,
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.

_______________________________

FRANCES
M.
FREEMAN
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
­­
335
I­
N­
V­
O­
I­
C­
E****
****
I­
N­
V­
O­
I­
C­
E****

FRANCES
M.
FREEMAN
43773
Sunset
Terrace
Ashburn,
VA
20147
703/
726­
6944
TODAY'S
DATE:
6/
29/
03
DATE
TAKEN:
6/
19/
03
CASE
NAME:
Fifra
conf
DEPONENTS:

TOTAL:
PAGES:
481,
plus
59
pgs
OT
ATTORNEY
TAKING
DEPO:

COPY
SALES
To:
Mr.

DELIVERY:
7
COMPRESSED:

DISK:

E­
MAIL:
no
EXHIBITS:
none
TRIAL
DATE:

**
SIGNATURE:
