"
Beaulieu,
Stephen
M."
<
steveb@
rti.
org>
11/
23/
2005
01:
15
PM
To
Ashley
Allen/
DC/
USEPA/
US@
EPA
Cc
"
Mansfield,
Carol
A."
<
carolm@
rti.
org>,
Brett
Snyder/
DC/
USEPA/
US@
EPA
Subject
RE:
Status
of
Peer
Review
of
316(
b)
documents
Ashley,

Below
are
the
final
comments
from
Dr.
Nisbet.

Have
a
Happy
Thanksgiving.

Regards,

Steve
From
Dr.
Nisbet:

My
remaining
comments
relate
to
question
15:

The
PPM
modeling
approach
begins
with
a
calibration
procedure
whereby
fecundity
expressed
as
eggs/
pound
is
adjusted
until
the
dominant
eigenvalue
of
the
transition
matrix
equals
the
inherent
growth
rate
of
the
population
based
on
alpha
values
in
Myers
et
al.
(
1999).
Given
that
independent
estimates
of
fecundity
are
available
for
most
species,
does
the
panel
believe
that
this
calibration
step
is
an
essential
part
of
the
PPM
approach?
Would
the
PPM
model
be
equally
valid
if
the
transition
matrix
were
defined
with
independent
estimates
of
fecundity
rather
than
values
derived
from
calibration?
Also,
the
fecundity
estimate
derived
by
calibration
is
expressed
as
a
single
constant
value
of
eggs/
pound
that
is
applied
to
all
ages
above
a
predetermined
age
of
fecundity.
Is
it
biologically
realistic
to
use
a
single
constant
rather
than
a
set
of
age
specific
fecundity
values?

There
are
three
questions
here:

a)
Given
that
independent
estimates
of
fecundity
are
available
for
most
species,
does
the
panel
believe
that
this
calibration
step
is
an
essential
part
of
the
PPM
approach?

RE:
I
do
believe
this
calibration
to
be
a
very
important
part
of
the
approach,
and
that
it
should
be
used
wherever
data
permit
it.
During
the
conference
call,
there
was
some
confusion
about
the
interpretation
of
the
parameter
alpha
based
on
descriptions
in
the
paper
by
Myers
et
al.
An
exchange
of
correspondence
with
Steve
Newbold
establishes
that
the
conceptual
problems
are
resolved,
though
there
may
be
some
technical
details
requiring
further
work.
I
would
be
happy
to
interact
directly
with
EPA
if
further
advice
is
required.
b)
Would
the
PPM
model
be
equally
valid
if
the
transition
matrix
were
defined
with
independent
estimates
of
fecundity
rather
than
values
derived
from
calibration?

RE:
In
view
of
my
answer
to
(
a)
­
no.

c)
Is
it
biologically
realistic
to
use
a
single
constant
rather
than
a
set
of
age
specific
fecundity
values?

RE:
This
simplification
is
unlikely
to
cause
problems
­
especially
as
the
modeling
objective
is
focused
on
equilibrium
dynamics.

S.
M.
Beaulieu
steveb@
rti.
org
