DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
282
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
7.0
DISCUSSION
The
primary
objectives
of
this
study
were
two
fold:
1)
evaluate
in
an
independent
lab
a
short­
term
reproduction
test
developed
by
EPA
(
2001)
for
Tier
1
screening
of
endocrine
disrupting
chemicals
(
EDCs)
with
the
fathead
minnow,
and
2)
compare
this
test
with
two
other
assays,
one
a
non­
spawning
14­
day
adult
fathead
minnow
screen
(
OECD
2001)
and
the
other
a
shortened
version
of
the
assay
described
by
EPA
(
2001),
with
respect
to
an
ability
to
identify
EDCs
operating
by
different
modes
of
action.
The
effectiveness
of
the
ability
of
the
three
different
short­
term
screening
assays
to
detect
substances
that
interact
adversely
with
endocrine
systems
in
fish
(
estrogen,
androgen,
and
thyroid
systems)
were
assessed
after
exposure
to
four
chemical
agents
representing
different
endocrine
disruptive
modes
of
action,
with
the
ultimate
goal
of
selecting
a
preferred
assay
to
be
used
in
Tier
1
screening.
In
general,
the
Tier
1
screen
should
be
relatively
quick
and
simple
to
run,
have
the
ability
to
detect
multiple
modes
of
action
of
EDCs,
and
be
readily
interpretable.
The
information
gleaned
from
a
Tier
1
assay
will
be
used
to
determine
whether
the
chemical
under
investigation
could
adversely
affect
the
endocrine
system
of
fish
and,
if
necessary,
direct
the
focus
of
further
Tier
2
testing.

The
four
test
chemicals
used
in
this
study
have
been
previously
tested
by
USEPA­
MED
using
the
21­
day
protocol,
providing
a
reference
data
set
to
allow
comparison
of
results
with
the
present
study
and
indicate
the
transferability
of
the
short­
term
fish
screening
assay.
Before
discussing
transferability
and
performance
of
the
various
protocols,
a
comparison
is
made
of
control
values
for
selected
endpoints
reported
in
this
study
with
those
of
previous
studies.
This
comparison
is
intended
to
provide
some
context
for
quantitative
comparisons
between
current
and
past
studies,
in
addition
to
comparison
of
the
relative
differences
within
a
study
among
exposed
and
non­
exposed
fish.

7.1
Fecundity
The
fecundity
endpoint
is
a
measure
of
the
number
of
eggs
produced
by
an
individual
female.
Arguably,
fecundity
is
the
most
useful
indicator
of
the
general
reproductive
condition
of
mature
fish
as
it
reflects
successful
integration
of
a
variety
of
physiological
processes.
As
a
reference
for
daily
fecundity
averages
in
fathead
minnows,
a
summary
is
presented
in
Table
7.1.
The
values
in
this
table
were
taken
from
previous
EPA­
MED
studies
(
Ankley
et
al.
2001,
2002,
2003)
using
the
same
21­
day
protocol
as
that
used
in
this
study.
Also
included
are
representative
fecundity
data
from
other
studies
using
pair
breeding
minnows
or
alternative
group
breeding
configurations.
Based
on
this
comparison,
the
average
daily
fecundity
of
minnows
maintained
during
the
course
of
this
study
appear
to
be
two
to
three
times
higher
than
those
typically
observed
at
the
EPA­
MED
laboratory,
but
within
the
upper
range
reported
in
other
studies.
The
increased
fecundity
in
the
present
study
appears
directly
related
to
the
larger
size
of
females
(
Table
7.1).
There
is
a
trend
toward
increased
body
size
among
females
during
the
study,
which
is
reflected
in
the
progressively
higher
daily
fecundity.
This
increased
size
is
to
be
expected,
as
fecundity
in
batch
spawning
is
a
direct
function
of
body
size
(
Wootton
1990).
Importantly,
the
daily
variation
in
fecundity
observed
in
this
study
(%
CV)
appears
to
be
consistent
with
that
reported
in
past
studies.
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
283
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
Table
7.1.
Average
Daily
Fecundity
in
Control
Fathead
Minnows
(
mean,
mean
±
SD)

Fish
Average
Fecundity
Reference
or
Weight
eggs
/
female
/
day
(
±
SD)
2­
18
average
(
g)
Pair
Breeding
Group
Breeding
Benoit
&
Carlson
1977
1.5
34
­
52
a
Kramer
et
al.
1998
nr
9.6
 
15.7
b
Harries
et
al.
2000
nr
34
 
132c
Jensen
et
al.
2001
nr
19
±
1
Ankley
et
al.
2001
nr
20.5
±
1d
Ankley
et
al.
2001
nr
16.4
±
5.4e
Ankley
et
al.
2002
nr
20.5
±
5.7
Ankley
et
al.
2003
1.75
38
f
Ankley
et
al.
unpublished
nr
57
g
This
study:
Methoxychlor
controls
1.59
32.5
±
2.5
Trenbolone
controls
1.72
43.7
±
9.5
Flutamide
controls
1.82
55.7
±
9.2
Fadrozole
controls
2.13
66.0
±
5.0
Note:
All
data
are
based
on
21­
day
cumulative
egg
production
except
where
noted
below.
nr
=
not
reported.
aRange
of
values
from
daily
spawnings
(
average
of
11­
16
spawns)
using
a
variety
of
minnow
barns;
includes
non­
adhering
eggs
laid
(
3
males
/
6
females
per
tank).
bDaily
average
calculated
from
range
reported
per
female
after
13
spawning
days
(
3
males
/
3
females
per
tank).
cRange
reported
in
the
Harries
et
al.
2000
study
from
two
separate
experiments
(
including
tank
water
and
solvent
control
values).
dControl
values
reported
with
methoxychlor
results.
eControl
values
reported
with
methyl­
testosterone
results
(
12
day
data
collection).
fEstimated
from
Figure
3
in
the
Ankley
et
al.
2003
paper.
gEstimated
from
unpublished
control
data
collected
during
flutamide
studies.

7.2
Histology
and
Gonadosomatic
index
(
GSI)

The
gonadosomatic
index
(
GSI)
is
generally
considered
a
good
measure
of
gonad
maturation
and
spawning
readiness
and
is
based
on
the
broad
assumption
that
proportionally
larger
gonads
indicate
greater
development
(
West
1990).
The
GSI
in
female
fathead
minnows
can
also
be
quite
variable
depending
on
the
time
since
last
spawning
(
Jensen
et
al.
2001).
Control
values
for
GSI
in
females
throughout
this
study
ranged
between
10%
and
14%,
and
for
males,
between
1.1%
and
1.4%.
Although
the
male
GSI
values
are
consistent
with
previous
studies,
the
female
GSI
values
are
slightly
higher
than
those
observed
in
past
EPA­
MED
studies,
which
appears
to
vary
between
8%
and
12%
(
Jensen
et
al.
2001;
Ankley
et
al.
2001).
This
difference
may
be
related
to
the
larger
size
of
female
fish
used
in
the
present
study.
The
use
of
the
GSI
assumes
a
linear
relationship
between
gonad
weight
and
body
weight,
which
may
not
be
consistent
across
a
broad
range
of
body
weights,
as
negative
or
positive
allometric
effects
may
skew
the
values
(
Weatherly
1990).

Histology
is
a
qualitative
to
semi­
quantitative
tool
used
to
describe
alterations
and
localize
specific
changes
in
organ
morphology.
Histological
methods
have
been
recognized
as
the
most
accurate
approach
for
staging
reproductive
development
in
fish
(
West
1990).
However,
a
primary
difficulty
in
applying
histological
analysis
is
that
interpretation
of
changes
in
tissues
may
vary
from
one
investigator
to
another.
In
the
present
study,
quantitative
aspects
of
the
histological
examination
have
been
emphasized
as
much
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
284
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
as
possible
to
reduce
investigator
bias.
As
a
comparison
to
the
analysis
used
in
this
study,
the
histological
examination
of
the
normal
fathead
minnow
reproductive
system
described
in
Jensen
et
al.
(
2001)
is
perhaps
the
most
relevant.
With
respect
to
ovarian
development,
the
Jensen
et
al.
(
2001)
study
described
approximately
35%
to
55%
of
the
oocytes
as
being
in
Stage
I
development
(
subdivided
into
1a
and
1b
in
the
current
study),
and
the
remaining
stages
varying
between
5%
and
25%
of
the
total
number
of
oocytes.
Notably,
the
oocytes
in
a
late
vitellogenic
stage
(
equivalent
to
Stage
4
in
the
current
study)
varied
systematically
with
time
since
last
spawning,
with
approximately
35%
of
the
oocytes
at
Stage
4
two
days
after
spawning
and
12%
the
same
day
after
spawning.
This
pattern
of
oocyte
growth
is
consistent
with
the
findings
from
control
fish
sampled
during
the
present
study.
Typically,
oocytes
in
Stage
4
development
comprised
20%
of
the
sample,
and
were
often
highly
variable
in
number,
probably
reflecting
ovaries
removed
from
females
at
different
post­
spawning
intervals.

7.3
Vitellogenin
and
Plasma
Steroid
Levels
The
biological
role
of
VTG
and
regulation
of
its
biosynthesis
is
well
established
in
fish,
including
fathead
minnows.
Synthesis
of
VTG
is
under
estrogen
control
and
both
male
and
female
fish
can
be
induced
to
synthesize
VTG
after
estrogen
exposure
or
after
exposure
to
estrogen
mimics.
In
sexually
mature
control
minnows,
it
is
expected
that
VTG
levels
are
low
to
negligible
in
males
and
in
the
mg/
mL
concentration
range
in
spawning
females.
Mean
values
of
VTG
in
spawning
females
from
previous
EPA­
MED
studies
appear
to
vary
from
12
 
27
mg/
mL
(
Table
7.2).
Other
studies
have
reported
much
lower
values
(
for
example
Harries
et
al.
2001),
which
are
likely
due
to
differences
in
analytical
methods,
specifically
the
use
of
carp
anti­
VTG
antibodies
as
opposed
to
fathead
minnow
anti­
VTG
used
in
past
EPA­
MED
studies.
In
the
present
study,
a
commercially
available
kit
for
VTG
detection
was
used,
which
used
carp
anti­
VTG.
The
average
values
for
VTG
in
control
males
using
this
kit
were
very
low,
as
expected,
and
are
similar
or
slightly
higher
than
those
observed
in
previous
studies
(
Table
7.3).
The
VTG
levels
in
control
females
ranged
from
1
mg/
mL
to
6
mg/
mL,
or
roughly
2
to
4
times
lower
than
those
from
past
EPA­
MED
studies.
This
difference
is
likely
attributable
to
the
use
of
carp
versus
minnow
anti­
VTG.
It
also
appears
inter­
female
variation
in
VTG
levels
are
greater
in
the
present
study
(
e.
g.,
%
CV
up
to
95%,
Table
7.2)
than
that
found
in
previous
EPA
studies.
As
mentioned
previously
in
the
results
section,
a
large
part
of
this
variation
is
due
to
some
control
females
having
exceptionally
low
VTG
levels
(
<
2%
to
5%
of
the
sample
mean;
see
figures
E21,
E14M
VTG­
1;
E21,
E14FL
VTG­
1).
For
these
fish,
it
is
questionable
whether
they
would
be
able
to
successfully
spawn
despite
an
external
appearance
of
sexual
maturity.
This
variation
may
represent
an
accepted
inevitability
with
protocols
using
group
spawners,
in
which
it
is
difficult
to
identify
individual
females
that
are
not
laying
eggs
during
pre­
exposure
observations.

Besides
VTG,
the
measurement
of
plasma
levels
of
E2,
11­
KT,
and
T
can
be
used
as
endpoints
to
assess
endocrine
status
in
male
and
female
fish,
provided
there
is
some
knowledge
about
normal
population
levels
when
the
samples
are
taken.
Plasma
levels
of
E2
have
limited
utility
in
males,
because
levels
of
this
steroid
are
usually
low
or
non­
detectable.
The
measurement
of
plasma
E2
is
most
useful
in
sexually
maturing
females
because
of
the
gradual
rise
in
this
hormone
during
the
period
of
vitellogenesis,
and
the
cyclic
nature
of
the
levels
during
the
reproductive
cycle
(
Jensen
et
al.
2001).
In
male
fishes,
11­
KT
is
the
sex
steroid
characteristic
of
sexual
maturity,
and
reduced
levels
are
considered
synonymous
with
reproductive
dysfunction.
Very
low
levels
of
11­
KT
(
e.
g.,
pg/
mL
range)
are
sometimes
reported
in
adult
female
fish
(
Jensen
et
al.,
2001;
Simpson
and
Wright
1977),
although
the
physiological
significance,
if
any,
is
not
understood.
Testosterone
levels
are
normally
much
lower
than
11­
KT
in
males
at
time
of
spawning
(
Jensen
et
al.,
2001),
although
in
females,
levels
of
T
can
approach
E2
levels,
making
this
a
useful
endpoint
to
measure
in
assessing
the
endocrine
status
of
sexually
mature
fish.
An
important
concern
with
single
measurements
of
plasma
sex
steroids
in
mature
fish
is
the
relevance
of
one
measurement
to
reproductive
function
or
dysfunction.
As
an
aid
in
determining
mode
of
action,
steroid
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
285
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
Table
7.2.
Average
Plasma
Markers
in
Control
Female
Fathead
Minnows
(
mean
or
mean
±
SD)

Vitellogenin
Estrogen
Testosterone
11­
Keto
E2
/
T
mg
/
ml
ng
/
ml
ng
/
ml
ng
/
ml
Ratio
2­
18
average
Jensen
et
al.
2001
12­
17
5.97
±
1.1
3.08
±
0.34
0.36
±
0.11
1.94
Harries
et
al.
2000
0.1
 
0.35a
nd
nd
nd
Makynen
et
al.
2000
nd
3.52
±
1.22
3.79
±
3.63
nr
0.93
Ankley
et
al.
2001b
13
9
3.25
nr
2.77
Ankley
et
al.
2001bc
25
4.5
3.5
nr
1.29
Ankley
et
al.
2002b
19.5
5.5
5.5
nr
1.0
Ankley
et
al.
2003b
27
4
4.25
nr
0.94
Ankley
unpublishedd
16
9
6
nr
1.50
This
study:
Methoxychlor
4.93
±
4.07
1.73
±
1.38
4.34
±
7.48
0.839
±
0.441
0.4
Trenbolone
1.29
±
1.16
1.36
±
1.38
1.33
±
1.33
nd
1.02
Flutamide
4.92
±
2.14
1.23
±
1.26
1.33
±
1.06
nd
0.92
Fadrozole
5.95
±
2.25
2.86
±
1.49
1.10
±
0.94
0.071
±
0.19
2.6
WA
2­
18
study
avg.

nd
=
not
determined;
nr
=
not
reported
aDetermined
using
Carp
anti­
VTG
ab.
bValues
estimated
from
figures.
cControl
values
reported
with
Methyl­
testosterone
results
(
12
d
data
collection).
d
Estimated
from
unpublished
control
data
collected
during
Flutamide
studies.

Table
7.3.
Average
Plasma
Markers
in
Control
Male
Fathead
Minnows
(
mean
or
mean
±
SD)

Vitellogenin
Estrogen
Testosterone
11­
Keto
mg
/
ml
ng
/
ml
ng
/
ml
ng
/
ml
2­
18
average
Jensen
et
al.
2001
0.004
±
0.001
0.4
±
0.13
9.11
±
0.92
33.1
±
3.14
Harries
et
al.
2000
0.00003a
nd
nd
nd
Makynen
et
al.
2000
nd
0.8
±
0.16
5.32
±
4.82
7.09
±
9.11
Ankley
et
al.
2001b
<
0.0001
0.1
±
0.13
7.5
±
0.92
29
±
3.14
Ankley
et
al.
2001bc
<
0.0001
0.3
8
28
Ankley
et
al.
2002b
<
0.0001
12.5
40.
Ankley
et
al.
2003b
nd
0.2
12
32
Ankley
unpublishedd
<
0.02
0.15
25
45
This
study:
Methoxychlor
0.00011
±
0.00015
0.31
±
0.09
5.21
±
3.0
35.21
±
16.81
Trenbolone
0.0029
±
0.0036
0.32
±
0.27
1.41
±
0.60
12.93
±
18.17
Flutamide
0.0024
±
0.0046
0.19
±
0.09
2.22
±
1.17
16.88
±
8.48
Fadrozole
0.0023
±
0.0026
0.069
±
0.09
2.21
±
1.11
19.77
±
15.11
WA
2­
18
study
avg.

nd
=
not
determined
aDetermined
using
Carp
anti­
VTG
anti­
bodies.
Male
VTG
from
solvent
controls
were
<
0.00007
mg/
mL
bValues
estimated
from
figures.
cControl
values
reported
with
Methyl­
testosterone
results
(
12
d
data
collection).
d
Estimated
from
unpublished
control
data
collected
during
Flutamide
studies.
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
286
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
ratios,
specifically
E2
/
T
in
females,
may
be
more
useful.
The
basis
for
this
analysis
is
the
hypothesis
that
sex
steroid
ratios,
as
opposed
to
their
absolute
values,
are
more
important
in
determining
sexual
differentiation
and,
perhaps,
sexual
development
(
Jalabert
et
al.
2000).
In
fathead
minnows,
there
does
appear
to
be
a
correlation
between
plasma
E2
and
T
levels
in
female
fish,
with
E2
/
T
concentration
ratios
between
1
and
2
being
typical
(
Jensen
et
al.
2001).

The
mean
sex­
steroid
levels
observed
in
control
female
and
male
fish
from
the
current
study
are
summarized
in
Table
7.2
and
7.3.
The
levels
for
E2
and
T
in
females
are
lower
than
those
previously
observed
in
EPA­
MED
studies
by
a
factor
of
roughly
2
to
3.
However,
the
E2
/
T
ratios
are
similar
to
past
studies
and
generally
ranged
between
1
and
2,
as
to
be
expected.
The
exception
to
this
observation
was
control
female
fish
sampled
during
the
methoxychlor
studies,
which
had
unusually
high
T
levels,
lowering
the
ratio
to
less
than
0.5
(
Table
7.2).
The
T
levels
in
male
control
fish
sampled
during
this
same
time
were
also
unusually
high
(
Table
7.3),
suggesting
the
cause
may
be
analytical
as
opposed
to
biological
(
e.
g.,
assay
error
rather
than
induction).
This
may
be
explained
by
the
use
of
an
ELISA­
based
method
for
analyzing
steroids
as
opposed
to
the
RIA­
based
method
used
in
the
EPA­
MED
studies.
The
sex­
steroid
levels
in
male
fish
were,
in
general,
closer
to
values
reported
in
previous
EPA
studies
although
T
levels
in
this
study
are
approximately
4
to
5
times
lower
than
those
previously
observed.

7.3.1
Test
Results:
Transferability
and
Inter­
Protocol
Comparison
During
the
course
of
this
study,
an
enormous
quantity
of
raw
data
was
collected
and
summarized
in
the
statistical
analysis
presented
in
the
previous
results
section.
A
number
of
measured
parameters,
such
as
the
chemistry
values,
serve
to
establish
that
the
environmental
conditions
and
measured
exposure
levels
were
within
acceptable
limits.
Other
parameters
such
as
body
weight
and
condition
factor
provide
some
indication
of
the
general
health
status
of
the
fish.
In
the
latter
case,
it
was
anticipated
that
condition
factor
would
be
unaffected
by
the
chemical
treatments,
as
sublethal
doses
were
employed
that
have
been
previously
established
to
be
well­
tolerated
in
EPA­
MED
studies.
In
general,
the
results
of
this
study
essentially
confirm
this
expectation.
Although
in
some
treatments
a
significant
difference
in
body
was
noted,
this
appeared
to
be
a
random
occurrence.
With
regard
to
other
biological
endpoints,
a
number
of
the
measurements
concerning
fertilization
success,
hatchability,
and
larval
survivability
/
development
indicated
these
were
comparatively
insensitive
parameters,
as
consistently
high
values
were
recorded
(>
90%)
regardless
of
chemical
treatment.
For
the
protocols
incorporating
reproduction,
overall
fecundity
either
expressed
as
total
eggs
or
the
daily
average
was
the
most
useful
indicator
of
reproductive
performance.

In
considering
the
merits
of
the
different
test
protocols
in
this
study,
it
is
important
to
identify
the
endpoints
that
appear
to
be
the
most
sensitive
and
pertinent
to
the
evaluation
of
mode
of
action.
Therefore,
the
following
discussion
focuses
on
fecundity,
plasma
VTG
and
sex
steroids,
GSI,
and
histology
as
the
key
endpoints
for
comparison.
For
each
chemical
treatment,
a
table
is
presented
that
summarizes
the
results
as
either
not
significantly
different
(­­­)
from
control
values,
or
if
a
significant
difference
was
observed,
the
direction
of
the
change
higher
( )
or
lower
( ).
For
simplification,
doseresponse
aspects
of
the
treatments
are
ignored
(
e.
g.,
differences
between
high
and
low
treatments).
Also,
in
some
of
the
sex
steroid
analysis,
it
appeared
that
exceptionally
high
variability
prevented
a
statistically
significant
conclusion,
despite
obviously
large
differences
in
the
sample
means.
In
these
instances,
a
"~"
symbol
is
shown
on
the
tables
to
distinguish
these
conclusions
from
those
that
achieved
statistical
significance.

7.3.2
Methoxychlor
(
mode
of
action:
weak
estrogen)

Methoxychlor
is
a
relatively
non­
persistent
organochlorine
insecticide
that
has
been
demonstrated
to
be
a
weak
estrogen
mimic
in
fish
and
both
an
estrogen
and
anti­
androgen
proprieties
in
mammals.
It
is
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
287
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
registered
for
insect
control
in
post­
harvest
applications
in
a
variety
of
crops,
foliar
treatment,
soil,
and
seed
treatment.
It
is
most
commonly
used
to
control
black
flies,
houseflies,
and
mosquitoes
in
association
with
human
habitation
(
EPA
1992).

The
results
of
the
methoxychlor
tests
summarized
below
indicate
a
general
lack
of
response
in
the
measured
endpoints.
The
notable
exception
was
reduced
fecundity
at
the
high
exposure
during
the
21­
day
spawning
protocol.
In
contrast,
fecundity
was
not
statistically
different
during
the
14­
day
protocol.
However,
problems
with
male
mortality
during
the
first
week
of
the
exposure
reduced
the
sample
size
to
two
replicates
at
termination,
and
combined
with
high
variation
in
fecundity,
diminished
the
statistical
power
of
this
test.
With
regard
to
the
biochemical
endpoints,
some
differences
were
observed
at
the
low­
and
mid­
exposure
levels
but
not
at
the
high,
making
the
biological
significance
of
the
findings
questionable.
Similarly,
histological
changes
were
observed
at
the
lower
treatments
but
not
at
the
highexposure
rate
in
the
non­
spawning
protocol.
With
the
exception
of
fecundity,
no
significant
effects
on
measured
endpoints
were
observed
at
the
high­
exposure
rate.

The
decreased
fecundity
at
the
high­
exposure
level
is
in
agreement
with
a
previous
study
by
Ankley
et
al.
(
2001),
which
also
observed
a
significant
reduction
in
fecundity
at
an
exposure
rate
of
5
µ
g/
L.
Changes
in
sex­
steroid
levels
and
VTG
induction
were
also
observed.
In
females,
estradiol
levels
were
decreased,
whereas
in
males,
decreased
testosterone
and
11­
KT
and
increased
VTG
were
documented
(
Ankley
et
al.
2001).
These
additional
changes
were
not
observed
in
the
present
study;
however,
high
variability
in
these
measurements
made
it
difficult
to
assess
statistical
significance.
Additionally,
the
confounding
variable
of
the
mycobacterium
infestation
discovered
after
the
test
may
also
be
a
contributing
factor
to
the
unusually
high
variability
observed
in
the
biochemical
endpoints.

Table
7.4.
Methoxychlor
Summary
Exposure
level
(
measured
range
in
µ
g/
L)
Endpoint
Low
(
0.8)
Mid
(
2.0)
High
(
3.22
­
4.1)
E­
21
E­
14
NS­
14
NS­
14
E­
21
E­
14
NS­
14
Fecundity
­­­
­­­
nd
nd
 
­­­
nd
VTG
­­­
F 
­­­
­­­
­­­
­­­
­­­
Estradiol
­­­
M 
­­­
­­­
­­­
­­­
­­­
Testosterone
­­­
­­­
F 
F 
­­­
­­­
­­­

11­
keto­
testost.
F 
­­­
­­­
­­­
­­­
­­­
­­­
GSI
Females
­­­
­­­
­­­
­­­
­­­
­­­
­­­
GSI
Males
­­­
­­­
­­­
­­­
­­­
­­­
­­­
Histology­
Females
­­­
­­­
­­­
­­­
­­­
­­­
­­­
Histology­
Males
­­­
­­­
Y
Y
­­­
­­­
­­­

­­­
not
significantly
different
from
controls
Y
a
significant
effect
observed
M
male
F
female
 
or
 
significantly
greater
or
less
than
controls.
nd
not
determined.
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
288
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
7.3.3
Trenbolone
(
anabolic
steroid)

Trenbolone
is
a
potent
steroid
that
is
often
used
in
the
cattle
industry
to
promote
growth
and
was
selected
to
be
used
in
this
study
as
a
model
androgenic
compound.
In
this
study,
trenbolone
exposure
caused
pronounced
changes
in
many
of
the
test
endpoints
at
the
high,
mid,
and
in
some
instances,
low
concentrations
(
Table
7.5).
These
results
are
as
one
would
expect
from
a
potent
anabolic
steroid
and
are
summarized
below.

In
comparing
the
test
results
among
the
three
different
protocols,
fecundity
results
between
the
14­
and
21­
day
spawning
assay
at
the
high
and
low
concentrations
were
similar.
Plasma
biochemical
measurements
were
also
similar
between
the
two
spawning
assays,
with
the
only
differences
being
female
testosterone
plasma
levels.
In
these
tests,
the
results
of
the
high­
exposure
group
showed
a
significant
decrease
in
female
testosterone
levels
in
the
14­
day
assay
but
not
in
the
21­
day
spawning
assay.
The
opposite
was
seen
in
the
low
treatments,
in
which
a
significant
decrease
in
female
testosterone
plasma
levels
was
observed
in
the
21­
day
assay
but
was
not
observed
in
the
14­
day
assay.
Regarding
the
nonspawning
assay,
biochemical
results
were
similar
to
both
the
14­
and
21­
day
spawning
protocols
at
the
high
concentration.
However,
at
the
low
concentrations,
the
14­
day
non­
spawning
assay
observed
a
significant
decrease
in
female
vitellogenin
and
testosterone,
and
male
estradiol,
testosterone,
and
11­
KT
steroid
plasma
levels.
This
significant
difference
was
not
observed
in
the
low­
treatment
groups
of
the
14­
day
and
21­
day
spawning
protocols
(
the
21­
day
low
treatment
had
reduction
in
female
testosterone
levels
only).
Regarding
GSI,
no
significant
differences
were
observed
in
the
non­
spawning
assay,
whereas
GSI
in
females
was
significantly
greater
than
that
in
controls
in
both
treatment
groups
for
both
the
14­
and
21­
day
spawning
protocols.
Female
histology
results
were
similar
at
high
and
low
concentrations
in
both
14­
and
21­
day
spawning
protocols.
At
high
concentrations,
significant
differences
were
observed
in
females
for
both
spawning
protocols.
This
significant
difference
in
female
histology
was
not
observed
at
the
high
concentration
in
the
14­
day
non­
spawning
protocol.
Male
histology
results
were
generally
similar
among
the
three
protocols
in
that
no
significant
differences
were
observed
between
treatments
and
controls.
The
only
exception
was
the
observed
significant
difference
at
the
high
concentration
in
the
14­
day
spawning
protocol.

The
results
from
this
study
are
similar
to
results
from
previous
studies
using
the
21­
day
spawning
protocol
(
Ankley
et
al,
2003)
with
regard
to
a
significant
reduction
in
female
fecundity
and
plasma
steroid
levels
(
testosterone,
estradiol,
vitellogenin
concentrations).
However,
this
study
did
not
observe
significant
increases
in
estradiol
and
vitellogenin
plasma
concentrations
in
male
fish,
as
was
observed
from
EPA­
MED
studies.
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
289
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
Table
7.5.
Trenbolone
Summary
Exposure
level
(
measured
range
in
µ
g/
L)
Endpoint
Low
(
0.041­
0.71)
Mid
(
0.45)
High
(
0.60­
0.86)
E­
21
E­
14
NS­
14
NS­
14
E­
21
E­
14
NS­
14
Fecundity
­­­
­­­
nd
nd
 
 
nd
VTG
­­­
­­­
F 
F 
F 
F 
F 
Estradiol
­­­
­­­
M 
F 
F 
F 
~
F 
Testosterone
F 
­­­
~
F/
M 
~
F/
M 
­­­
F 
F 
11­
keto­
testost.
­­­
­­­
M 
­­­
­­­
­­­
­­­
GSI
Females
 
 
­­­
­­­
 
 
­­­
GSI
Males
­­­
­­­
­­­
­­­
­­­
 
­­­
Histology­
Females
­­­
­­­
­­­
­­­
Y
Y
­­­
Histology­
Males
­­­
­­­
­­­
­­­
­­­
Y
­­­

­­­
not
significantly
different
from
controls
Y
a
significant
effect
observed
M
male
F
female
 
or
 
significantly
greater
or
less
than
controls.
nd
not
determined.

Female
body
weights
increased
in
E­
21
high
treatment.
Histology:
E­
21
test
females
had
 
in
atretic
follicles
and
 
in
corpus
lutea
at
high
treatment.
E­
14
test
females
exhibited
similar
effects
plus
inhibition
of
stage
3­
4
progression
of
oocytes
(
progression
to
late
vitellogenesis).
In
E­
14
males,
there
was
an
 
in
microscopic
staging.

7.3.4
Flutamide
(
anti­
androgen)

Flutamide
is
an
anti­
androgen
and
therapeutic
agent
that
is
used
primarily
to
treat
prostate
cancer
(
Sogani
et
al.
1984).
It
is
considered
to
be
a
pro­
drug,
and
bioactivation
to
the
2­
hydroxy
metabolite
is
believed
necessary
for
full
biological
activity
(
Simard
et
al.
1986).
In
this
study,
the
effects
of
flutamide
exposure
on
test
endpoints
were
similar
to
methoxychlor
in
that
relatively
little
or
no
effect
was
observed.
A
notable
exception
was
in
the
fecundity
data,
which
showed
a
decrease
at
the
high­
exposure
level
in
both
the
14­
and
21­
day
spawning
protocols.
This
result
was
consistent
with
the
histological
analysis,
which
indicated
an
increase
in
atretic
follicles
in
these
fish.
With
regard
to
sex­
steroid
levels,
increased
levels
of
testosterone
in
females
and
11­
KT
in
males
are
noteworthy.
The
former
result
indicates
a
decrease
in
the
E2
/
T
ratio
in
females,
which
may
be
associated
with
decreased
fecundity,
although
sex­
steroid
levels
were
unaffected
in
the
14­
day
spawning
protocol.
It
should
also
be
noted
that
no
statistically
significant
differences
in
the
endpoints
were
observed
in
the
non­
spawning
protocol
fish.

The
relatively
weak
activity
of
flutamide
in
the
fathead
minnow
protocols
may
be
the
result
of
limited
bioactivation
compared
with
mammals
or
inadequate
exposure
duration.
The
latter
point
is
supported
by
the
more
pronounced
effects
observed
on
sex­
steroid
levels
in
the
21­
day
females
compared
to
14­
day
spawning
females.
In
general,
the
results
from
the
21­
day
spawning
protocol
in
this
study
are
consistent
with
a
previous
unpublished
study
(
Ankley
2002)
using
a
similar
exposure
protocol.
In
the
earlier
study,
fecundity
was
decreased
at
an
exposure
level
of
640
µ
g/
L.
Similarly,
testosterone
levels
in
females
were
increased
with
an
apparent
decrease
in
E2
levels
(
statistical
significance
unknown).
Thus,
flutamide
exposure
appeared
to
decrease
the
E2
/
T
ratio
in
females.
Other
sex­
steroid
levels
did
not
appear
to
be
significantly
altered,
nor
was
GSI
affected.
There
did,
however,
appear
to
be
a
slight
induction
of
VTG
in
males
at
the
640­
µ
g/
L
exposure
level,
which
was
not
observed
in
the
current
study.
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
290
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
Table
7.6.
Flutamide
Summary
Exposure
level
(
measured
range
in
µ
g/
L)
Endpoint
Low
(
43­
47)
Mid
(
260)
High
(
504­
519)
E­
21
E­
14
NS­
14
NS­
14
E­
21
E­
14
NS­
14
Fecundity
­­­
­­­
nd
nd
F 
F 
nd
VTG
­­­
­­­
­­­
­­­
­­­
­­­
­­­
Estradiol
­­­
­­­
­­­
­­­
M 
M 
­­­
Testosterone
M 
­­­
­­­
­­­
F 
­­­
­­­

11­
keto­
testost.
­­­
­­­
­­­
­­­
M 
­­­
­­­
GSI
Females
­­­
­­­
­­­
­­­
­­­
­­­
­­­
GSI
Males
­­­
­­­
­­­
­­­
­­­
­­­
­­­
Histology­
Females
­­­
­­­
­­­
­­­
Y
Y
­­­
Histology­
Males
­­­
­­­
­­­
­­­
­­­
­­­
­­­

­­­
not
significantly
different
from
controls
Y
a
significant
effect
observed
M
male
F
female
 
or
 
significantly
greater
or
less
than
controls.
nd
not
determined.

Histology:
E­
21,
14
test
females
had
 
in
atretic
follicles.

7.3.5
Fadrozole
(
aromatase
inhibitor,
indirect
anti­
estrogen)

Fadrozole
is
a
competitive
inhibitor
of
CYP19
(
aromatase),
a
key
enzyme
in
the
biosynthetic
pathway
for
estradiol
synthesis.
Estrogen
synthesis
inhibitors
may
also
be
considered
indirect
anti­
estrogens.
In
this
study,
fadrozole
exposure
caused
pronounced
changes
in
several
test
endpoints
after
the
high
and
in
some
instances,
mid
and
low
exposure
levels.
As
expected
with
an
aromatase
inhibitor,
estradiol
levels
in
females
were
decreased
in
the
high­
exposure
groups
and
in
the
low­
exposure
group
from
the
21­
day
spawning
protocol.
This
decrease
in
estradiol
was
also
associated
with
reduced
fecundity
and
VTG
in
females.
Testosterone
levels
were
elevated
in
females,
causing
an
overall
large
decrease
in
the
E2
/
T
ratio.
Histological
examination
of
the
ovaries
was
also
consistent
with
the
biochemical
and
fecundity
results,
with
significantly
reduced
staging
(
preponderance
of
immature
oocytes)
observed
in
nearly
all
treatments
and
protocols.
Fadrozole
effects
on
male
fish
were
less
obvious,
except
for
consistent
increases
in
testosterone
and
11­
KT
levels.
The
GSI
in
males
from
the
21­
day
protocol
was
elevated,
which
was
associated
with
an
increase
in
tubule
diameter
determined
during
histological
examination.

In
comparing
the
test
results
from
the
different
protocols,
the
biochemical,
fecundity,
and
histological
results
were
similar
between
the
14­
and
21­
day
spawning
protocols,
although
estradiol
levels
were
decreased
at
the
low­
exposure
level
only
in
the
21­
day
protocol.
With
regard
to
the
non­
spawning
protocol,
it
is
notable
that
no
effects
were
seen
at
the
low
treatment,
which
is
in
contrast
to
the
results
from
the
protocols
that
include
spawning.

The
results
from
this
study
are
in
close
agreement
with
a
previous
study
using
the
21­
day
spawning
protocol
(
Ankley
et
al.
2002).
In
that
study,
fadrozole
exposure
altered
the
biochemical
endpoints
in
females
in
a
manner
nearly
identical
to
that
observed
in
this
study,
including
decreases
in
estradiol
at
10,
50,
and
2
µ
g/
L,
and
decreases
in
VTG
at
10
and
50
µ
g/
L
fadrozole
exposures.
Fecundity
was
reduced
at
2
µ
g/
L
and
spawning
ceased
within
2
days
at
the
higher
treatment
levels
(
Ankley
et
al.
2002).
Histological
analysis
of
the
ovaries
was
similar
to
the
current
study,
with
essentially
the
absence
of
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
291
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
oocytes
beyond
Stage
3
development
(
Ankley
et
al.
2002).
In
males,
the
GSI,
testosterone,
and
11­
KT
levels
were
increased
at
the
high­
treatment
rates,
which
is
also
consistent
with
results
from
the
present
study.

Table
7.7.
Fadrozole
Summary
Exposure
level
(
measured
range
in
µ
g/
L)
Endpoint
Low
(
5.1­
5.5)
Mid
(
31.7)
High
(
55.7­
60.0)
E­
21
E­
14
NS­
14
NS­
14
E­
21
E­
14
NS­
14
Fecundity
­­­
­­­
nd
nd
 
 
nd
VTG
F 
F 
­­­
F 
F 
F 
F 
Estradiol
F 
­­­
­­­
­­­
nd
F 
F 
Testosterone
F,
M 
F,
M 
­­­
­­­
F,
M 
F,
M 
~
M 
11­
keto­
testost.
M 
M 
­­­
­­­
M 
M 
~
M 
GSI
Females
­­­
 
­­­
­­­
 
 
 
GSI
Males
­­­
­­­
­­­
­­­
 
­­­
­­­
Histology­
Females
Y
Y
­­­
Y
Y
Y
Y
Histology­
Males
­­­
­­­
­­­
­­­
Y
­­­
­­­

­­­
not
significantly
different
from
controls
Y
a
significant
effect
observed
M
male
F
female
 
or
 
significantly
greater
or
less
than
controls.
nd
not
determined.

Female
body
weights
greater
in
high
treatment
E­
21,14.
Histology:
E­
21,14;
NS­
14
test
females
exhibited
less
advanced
staging
( 
pre­
vitellogenic
oocytes)
at
the
mid
and
high
doses
and
both
increased
numbers
of
atretic
follicles.
E­
21
males
had
 
tubule
diameter.

7.4
Conclusion
The
results
suggest
the
four
endocrine­
disruptive
chemicals
tested
in
this
study
can
be
divided
into
two
groups:
weak­
acting
with
little
measured
effect
(
methoxychlor
and
flutamide)
or
strong­
acting,
causing
pronounced
effects
(
trenbolone
and
fadrozole).
It
is
interesting
to
note
that
the
weak­
acting
chemicals
require
bioactivation
for
biological
effect.
Undoubtedly,
both
inter­
species
(
e.
g.,
mammals
versus
fish)
and
inter­
individual
differences
contribute
to
the
extent
of
metabolite
formation,
and
for
pharmaceutical
agents
such
as
flutamide,
it
is
unclear
whether
the
active
metabolite
binds
to
fish
androgen
receptors
(
Wells
and
Van
Der
Kraak
2000).
Nonetheless,
this
distinction
can
be
useful
in
determining
which
protocol
is
superior
as
a
screening
protocol,
since
it
likely
both
strong
and
weakly
active
compounds
will
be
encountered
during
large­
scale
screening
efforts.

7.5
Protocol
Selection
The
short­
term
reproduction
assay
(
21­
day
protocol)
was
demonstrated
to
be
the
most
comprehensive
and
sensitive
protocol
for
use
as
a
fish
endocrine­
disruptor
screening
assay.
However,
the
14­
day
spawning
protocol
shows
good
promise
as
a
cost­
effective
alternative.
The
reasons
for
this
can
be
summarized
as
follows:

 
Consistency
in
overall
response
to
the
21­
day
protocol
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
292
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
 
Insufficient
need
for
quantitative
pre­
exposure
fecundity
measurements
 
Reduced
costs
compared
with
the
21­
day
protocol
 
Lack
of
response
at
the
low­
exposure
level
and
frequent
negative
responses
at
the
highexposure
level
in
the
non­
spawning
protocol.
This
observation
suggests
a
higher
risk
of
false­
negative
results,
particularly
with
weak­
acting
endocrine
disruptors.

Throughout
the
study,
the
14­
day
protocol
gave
results
that
were
generally
consistent
with
the
21­
day
protocol,
including
fecundity.
In
nearly
all
chemical
treatments
and
exposure
levels,
the
conclusions
drawn
from
the
fecundity
measurements
were
identical
for
both
the
14­
day
and
21­
day
treatments,
with
the
notable
exception
of
methoxychlor.
For
the
most
part,
results
from
sex­
steroid
analysis
were
similar,
with
a
slight
tendency
for
more
obvious
results
in
the
21­
day
protocol
with
flutamide
and
fadrozole.
With
regard
to
the
pre­
exposure
evaluation
period,
quantitative
fecundity
estimates
did
not
improve
the
power
of
the
statistical
analysis,
which
is
the
primary
limitation
in
the
use
of
fecundity
measurements.
Thus,
the
effort
placed
in
quantitative
pre­
exposure
fecundity
measurements
would
be
better
served
if
used
to
increase
the
sample
size
of
exposure
replicates.
Since
this
increase
may
be
impractical
for
many
laboratories
relying
on
proportional
diluters
for
toxicant
delivery,
the
qualitative
approach
used
in
the
14­
day
protocol
is
a
reasonable,
yet
cost
effective
alternative.
It
is
recommended
that
additional
development
of
the
abbreviated
14­
day
spawning
protocol
be
performed
to
evaluate
how
well
the
assay
performs
in
comparison
to
the
full
21­
day
protocol
with
weaker
acting
and/
or
high
log­
P
compounds.

When
comparing
the
results
between
the
non­
spawning
protocol
and
those
incorporating
reproduction,
it
is
clear
that
all
protocols
would
provide
similar
conclusions
with
strong­
acting
endocrine
disruptors
tested
at
high­
exposure
levels.
However,
comparison
of
the
results
from
the
weak­
acting
compounds
and
results
from
all
low­
exposure
treatments
suggests
the
non­
spawning
protocol
is
less
sensitive
and
may
falsely
conclude
that
a
chemical
is
not
an
endocrine
disruptor.
For
example,
all
measured
endpoints
were
negative
in
the
non­
spawning
assay
during
the
flutamide
exposure,
yet
the
high­
exposure
level
clearly
reduced
fecundity
and
elevated
androgen
levels
in
both
sexes
in
the
spawning
protocols.
In
the
lowfadrozole
exposure,
the
non­
spawning
assay
gave
completely
negative
results,
whereas
the
spawning
protocols
exhibited
a
pattern
of
effects
typical
of
anti­
estrogens.
Thus,
we
cannot
recommend
the
nonspawning
protocol,
as
the
potential
exists
for
a
high
false­
negative
rate
with
weak­
acting
compounds
and
negative
results
due
to
testing
an
insufficiently
low
concentration.

7.6
Transferability
Comparison
of
the
results
from
this
study
with
previous
EPA­
MED
efforts
using
the
21­
day
spawning
protocol
suggest
the
protocols
are
readily
transferable,
in
that
similar
conclusions
regarding
mode
of
action
can
be
made
from
the
experimental
results.
The
consistency
in
the
results
between
the
current
and
past
studies
is
most
obvious
at
the
high­
treatment
rates
and
with
measurements
of
fecundity,
sex
steroids,
and
histology.
In
a
few
instances
in
the
present
study,
notably
with
the
methoxychlor
results,
a
lack
of
response
was
observed
in
some
endpoints,
whereas
a
significant
change
was
noted
in
a
previous
study.
However,
unusually
high
variation
in
some
parameters,
such
as
fecundity
during
the
methoxychlor
exposures,
reduced
the
sensitivity
of
this
and
other
measurements.
In
general,
higher
variability
(
and
hence,
reduced
statistical
power)
was
a
greater
problem
at
lower
exposure
levels
throughout
the
study
and
in
particular
with
weak­
acting
chemicals.
Thus,
for
acceptable
transferability
of
the
protocol,
proper
dose
selection
is
critical.
Greatest
reproducibility
of
the
results
would
be
expected
to
occur
at
the
highest
possible
dose
level
that
does
not
cause
unacceptable
lethality
during
the
course
of
the
study.
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
293
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
8.0
REFERENCES
Ankley,
G.
T.,
K.
M.
Jensen,
M.
D.
Kahl,
J.
J.
Korte,
and
E.
A.
Makynen.
2001.
"
Description
and
evaluation
of
a
short­
term
reproduction
test
with
the
fathead
minnow
(
Pimephales
promelas)."
Environmental
Toxicology
and
Chemistry
20,
1276­
1290.

Ankley,
G.
T.,
K.
M.
Jensen,
E.
A.
Makynen,
M.
D.
Kahl,
J.
J.
Korte,
M.
W.
Hornung,
T.
R.
Henry,
J.
S.
Denny,
V.
S.
Leino,
M.
C.
Wilson,
P.
C.
Cardon,
and
L.
E.
Gray.
2003.
"
Effect
of
the
Androgenic
Growth
Promoter
17­
 ­
Trenbolone
on
Fecundity
and
Reproductive
Endocrinology
of
the
Fathead
Minnow."
Environmental
Toxicology
and
Chemistry,
Vol.
22
No.
6,
pp.
1350­
1360.

Ankley,
G.
T.,
M.
D.
Kahl,
K.
M.
Jensen,
M.
W.
Hornung,
J.
J.
Korte,
E.
A.
Makynen,
and
R.
L.
Leino.
2002.
"
Evaluation
of
the
aromatase
inhibitor
fadrozole
in
a
short­
term
reproduction
assay
with
the
fathead
minnow
(
Pimephales
promelas)."
Toxicological
Sciences
67,
121­
130.

Benoit,
D.
A.,
and
R.
W.
Carlson,.
1977.
"
Spawning
Success
of
Fathead
Minnows
on
Selected
Artificial
Substrates,"
The
Progressive
Fish
Culturist
vol.
39,
no
2,
pp.
67­
69.

Chapman,
P.
F.,
M.
Crane,
J.
Wiles,
F.
Noppert,
and
E.
McIndoe.
1996.
Improving
the
quality
of
statistics
in
regulatory
ecotoxicity
tests.
Ecotoxicology
5:
169­
186.

Daniel,
W.
W.
1978.
Applied
Nonparametric
Statistics.
Houghton
Mifflin
Company,
Boston.
503
pp.

EPA
(
U.
S.
Environmental
Protection
Agency).
1987.
Guidelines
for
the
Culture
of
Fathead
Minnows
Pimephales
promelas
for
Use
in
Toxicity
Tests.
U.
S.
Environmental
Protection
Agency,
Environmental
Research
Laboratory,
Duluth,
MN.
EPA/
600/
3­
87/
001.
January
1987.

EPA
(
U.
S.
Environmental
Protection
Agency).
1992.
National
Study
of
Chemical
Residues
in
Fish.
EPA,
Office
of
Science
and
Technology,
Standards
and
Applied
Science
Division,
V.
2
EPA­
823­
R­
92­
008B.

EPA
(
U.
S.
Environmental
Protection
Agency).
2001.
A
Short­
term
Method
for
Assessing
the
Reproductive
Toxicity
of
Endocrine
Disrupting
Chemicals
Using
the
Fathead
Minnow
(
Pimephales
promelas).
U.
S.
Environmental
Protection
Agency,
Mid
Continent
Ecology
Division,
EPA/
600/
R­
01/
067.

Harries,
J.
E.,
T.
Runnalls,
E.
Hill,
C.
A.
Harris,
S.
Maddix,
J.
P.
Sumpter,
and
C.
R.
Tyler.
2000.
"
Development
of
a
reproductive
performance
test
for
endocrine
disrupting
chemicals
using
pair­
breeding
fathead
minnows
(
Pimephales
promelas),"
Environmental
Science
&
Technology,
vol.
34,
no.
14,
pp.
3003­
3011.

Jalabert,
B.,
J.
F.
Baroiller,
B.
Breton,
A.
Fostier,
F.
Le
Gac,
Y.
Guiguen,
and
G.
Monod.
2000.
"
Main
Neuro­
endocrine,
Endocrine
and
Paracrine
Regulations
of
Fish
Reproduction,
and
Vulnerability
to
Xenobiotics,"
Ecotoxicology,
Vol.
9,
No.
1­
2,
pp.
25­
40.

Jensen,
K.
M,
J.
J.
Korte,
M.
D.
Kahl,
M.
S.
Pasha,
G.
T.
Ankley
2001.
"
Aspects
of
basic
reproductive
biology
and
endocrinology
in
the
fathead
minnow
(
Pimephales
promelas)"
Comparative
Biochemistry
and
Physiology
Part
C
Toxicology
&
Pharmacology
,
vol.
128C,
pp.
127­
141.

Kahl
M.
D.,
C.
L.
Russom,
D.
L.
DeFoe,
and
D.
E.
Hammermeister.
1999.
Saturation
Units
for
Use
in
Aquatic
Bioassays.
Chemosphere,
v.
39,
p.
539­
551.
DRAFT
EPA
WA
3­
8
(
Report
of
WA
2­
18
Study)
294
Preliminary
Data:
May
be
subject
to
change
following
QA
and
Management
review
July
30,
2003
Kramer,
V.
J.,
S.
Miles­
Richardson,
S.
L.
Pierens,
J.
P.
Giesy,
J.
P.
1998.
"
Reproductive
impairment
and
induction
of
alkaline­
labile
phosphate,
a
biomarker
of
estrogen
exposure,
in
fathead
minnows
(
Pimephales
promelas)
exposed
to
waterborne
17beta­
estradiol."
Aquatic
Toxicology,
vol.
40,
pp.
335­
360.

Makynen,
E.
A.,
M.
D.
Kahl,
K.
M.
Jensen,
J.
E.
Tietge,
K.
L
Wells,
G.
Van
Der
Kraak,
G.
T.
Ankley.
2000.
"
Effects
of
the
mammalian
antiandrogen
vinclozolin
on
development
and
reproduction
of
the
fathead
minnow
(
Pimephales
promelas)."
Aquatic
Toxicology
vol.
48,
p.
461­
475.

OECD.
(
2001
Draft).
Guidelines
for
the
Testing
of
Chemicals
Draft
Proposal
for
a
New
Guideline:
Short­
term
Reproductive
Test
with
the
Fathead
Minnow
for
Identification
of
Endocrine
Disruption
Chemicals.
OECD
Environmental
Health
and
Safety
Publications.
Series
on
Testing
and
Assessment.

Simard,
J.,
I.
Luthy,
J.
Guay,
A.
Belanger,
and
F.
Labrie.
1986.
"
Characteristics
of
Interaction
of
the
Antiandrogen
Flutamide
with
the
Androgen
Receptor
in
Various
Target
Tissues."
Molecular
and
Cellular
Endocrinology
44,
261­
270.

Simpson,
T.
H.,
and
R.
S.
Wright.
1977.
"
A
Radioimmunoassay
for
11­
oxotestosterone:
its
Application
in
the
Measurement
of
Levels
in
Blood
Serum
of
Rainbow
Trout
(
S.
Gairdneri).,"
Steroids
No.
29,
pp.
383­
398.

Snedecor,
G.
W.,
and
W.
G.
Cochran.
1980.
Statistical
Methods.
The
Iowa
State
University
Press,
Ames,
Iowa.
507
pp.

Sogani,
P.
C.,
M.
R.
Vagaiwala,
W.
F.
Whitmore.
1984.
"
Experience
with
Flutamide
in
Patients
with
Advanced
Prostatic­
Cancer
without
Prior
Endocrine
Therapy."
Cancer
vol.
54
no.
4:
744­
750.

Weatherley,
A.
H.
1990.
"
Approaches
to
Understanding
Fish
Growth."
Transactions
of
the
American
Fisheries
Society,
Vol.
119,
No.
4,
pp.
662­
672.

Wells,
K.,
and
G.
Van
Der
Kraak.
2000.
"
Differential
binding
of
endogenous
steroids
and
chemicals
to
androgen
receptors
in
rainbow
trout
and
goldfish."
Environmental
Toxicology
and
Chemistry
19,
2059­
2065.

West,
G.
1990,
"
Methods
of
Assessing
Ovarian
Development
in
Fishes
­
a
Review,"
Australian
Journal
of
Marine
and
Freshwater
Research,
Vol.
41,
No.
2,
pp.
199­
222.

Wootton,
R.
J.
1990.
Ecology
of
Teleost
Fishes.
Chapman
and
Hall.
Fish
and
Fisheries
Series
1.
ISBN:
0­
412­
31720­
6
