Salmonid
Distributions
and
Temperature
Contents
Abstract
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
Introduction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
What
is
a
"
distribution"?
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2
Ontogenetic
variation
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2
Life
history
variation
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
Historical
vs.
contemporary
vs.
potential
distribution
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
Examples
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4
What
are
the
direct
effects
of
temperature?
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
12
Examples
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
12
What
are
indirect
effects
of
temperature
on
fish
distributions?
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
13
Biotic
interactions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
13
Habitat
size
and
isolation
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
13
What
is
meant
by
"
scale"
and
"
level?"
At
what
"
level"
should
we
be
concerned
with
temperature
criteria
to
protect
fish
distributions?
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
13
Information
on
thermal
relationships
of
salmonids
comes
from
a
variety
of
laboratory
and
field
studies
 
how
do
we
integrate
work
conducted
at
different
scales
or
levels
of
organization
(
e.
g.,
population
vs.
individuals)?
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
14
EPA
Fish
and
Temperature
Database
Matching
System
(
FTDMS)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15
Point
Observation
of
Cold­
Water
Fish
in
relation
to
Temperature
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15
Is
unoccupied
habitat
relevant
to
temperature
requirements
of
salmonids?
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15
Conclusion
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
17
Literature
Cited
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
17
1
Salmonid
Distributions
and
Temperature
Issue
Paper
2
Salmonid
Distributions
and
Temperature
Prepared
as
Part
of
Region
10
Temperature
Water
Quality
Criteria
Guidance
Development
Project
Jason
Dunham,
Jeff
Lockwood,
and
Chris
Mebane
Abstract
Distributions
of
native
salmonid
fish
in
the
Pacific
Northwest
are
strongly
tied
to
temperature
conditions
in
their
habitat.
Salmonid
populations
have
declined
in
conjunction
with
thermal
changes
and
the
loss
and
fragmentation
of
large
and
interconnected
cold­
water
habitats.
Temperature
affects
the
health
of
not
only
individual
fish
but
also
entire
populations
and
groups
of
species.
Temperature
changes
have
obvious
direct
effects,
and
also
interact
with
other
factors
to
indirectly
affect
salmonids.

The
best
way
to
protect
existing
populations
and
restore
depleted
populations
is
to
create
temperature
criteria
that
explicitly
consider
salmonids'
temperature
requirements
at
different
times
and
places.
Natural
temperature
conditions
must
be
preserved
whenever
possible.
Because
current
fish
distributions
and
populations
are
significantly
reduced
from
their
historical
numbers,
protection
and
restoration
of
their
thermal
environment
must
often
extend
beyond
the
boundaries
of
their
existing
or
suitable
habitat.

Attempts
to
set
temperature
criteria
must
balance
what
is
known
and
not
known
about
the
habitat
and
biological
requirements
of
salmonids.
Full
consideration
of
current
and
potential
fish
distribution
and
habitat,
including
thorough
documentation
of
assumptions
and
knowledge
gaps,
is
needed
in
establishing
and
implementing
temperature
criteria
to
support
healthy
(
viable,
productive,
and
fishable)
salmonid
populations.

Introduction
Under
natural
conditions,
freshwater
salmonid
habitat
is
defined
by
physical
and
chemical
characteristics
of
the
environment,
including
water
quality,
flow,
geological
and
topographic
features
of
the
stream
and
its
valley,
and
cover
(
National
Research
Council
1996).
Common
factors
influencing
fish
distribution
include
size
and
accessibility
of
suitable
habitat,
connectivity
between
areas
of
suitable
habitat,
biological
interactions,
and
"
historical"
factors
(
e.
g.,
postglacial
dispersal
and
geographic
barriers)
(
Matthews
1998).
Many
of
these
factors
act
directly
or
indirectly
with
temperature
to
determine
the
distribution
of
a
species.
This
is
especially
true
for
cold­
water
fishes
such
as
salmonids.

This
paper
is
not
intended
to
be
an
exhaustive
review
of
the
status
or
declines
in
salmonid
2
Salmonid
Distributions
and
Temperature
populations
or
distributions.
These
are
widely
documented
elsewhere.
We
briefly
review
some
examples
of
declines
in
salmonid
populations
and
habitats
to
provide
some
context
for
these
issues,
but
our
focus
is
not
on
declines
per
se.
Furthermore,
this
issue
paper
is
not
intended
to
be
an
exhaustive
review
of
the
effects
of
temperature
on
salmonid
distributions
in
the
Pacific
Northwest
(
see
McCullough
1999).
Rather,
it
is
intended
to
describe
a
basic
framework
for
thinking
about
salmonid
distributions
and
appropriate
biological
criteria
to
protect
salmonid
populations
from
adverse
effects
of
altered
factors
affecting
thermal
regimes.

This
paper
describes
in
a
question­
and­
answer
format
five
main
issues
related
to
salmonid
distributions
and
temperature
criteria:

1.
Definition
of
a
"
distribution"
2.
Direct
effects
of
temperature
3.
Indirect
effects
of
temperature
4.
Relevance
of
scale
5.
Importance
of
unoccupied
habitat
What
is
a
"
distribution"?

Often,
the
word
"
distribution"
is
used
without
reference
to
what
is
specifically
meant.
Like
any
other
organism,
salmonid
fishes
(
and
temperatures)
are
not
distributed
equally
across
landscapes.
Within
stream
basins,
limits
to
fish
distributions
may
be
obvious,
but
even
within
continuous
areas
of
suitable
habitat,
discontinuities
in
distributions
may
arise
(
Angermeier
et
al.
in
press,
Dunham
et
al.
in
press).

A
common
example
of
"
distribution"
for
animals
can
be
found
in
popular
bird
identification
and
field
guides.
Distribution
maps
for
birds
often
cover
broad
areas.
In
some
cases,
ranges
of
different
"
races"
or
recognized
subspecies
are
distinguished.
Within
these
areas,
it
is
obvious
that
birds
do
not
occur
everywhere.
For
example,
a
wading
bird
may
only
be
found
in
wetland
areas,
though
it
is
broadly
distributed
across
the
continent
(
because
wetlands
are
broadly
distributed).
Furthermore,
this
bird
may
only
be
found
in
particular
kinds
of
wetlands
(
those
with
sufficient
cover
and
food
to
support
reproduction).
This
bird
may
be
found
in
different
areas,
depending
on
the
season.
Birds
may
appear
in
"
unusual"
habitats
while
migrating,
or
may
shift
habitat
use
from
year
to
year,
depending
on
climate
(
wet
vs.
dry
years).
Similar
analogies
apply
to
salmonid
fishes.
There
are
several
things
to
consider
when
using
the
term
"
distribution"
for
salmonids:
ontogenetic
variation;
life
history
variation;
and
historical,
contemporary,
and
potential
distribution.

Ontogenetic
variation.
"
Ontogenetic
variation"
refers
to
changes
in
habitat
use
during
the
life
cycle
of
an
individual.
Here,
the
term
"
life
cycle"
refers
to
the
sequence
of
events
(
egg
6
alevin
6
parr
6
smolt
6
juvenile
6
adult)
that
must
occur
within
an
individual's
life
for
successful
reproduction.
Ideally,
temperature
criteria
established
for
salmonids
should
address
spatial
and
temporal
distribution
of
thermal
habitats
that
protect
all
life
stages.
3
Salmonid
Distributions
and
Temperature
Habitat
requirements
vary
considerably
as
salmonids
begin
their
lives
as
eggs
in
(
or
on)
the
substrate
and
progress
through
developmental
stages
to
reproduction
as
an
adult.
Different
life
stages
may
have
different
thermal
requirements
(
Magnuson
et
al.
1979;
Physiology
issue
paper).
However,
thermal
requirements
may
also
overlap
considerably
among
life
stages.
Furthermore,
some
life
stages
are
relatively
insensitive
to
temperature
whereas
others
(
such
as
egg
incubation)
are
extremely
sensitive
(
see
Physiology
issue
paper).

Life
stage
requirements
may
be
tied
to
specific
spatial
or
temporal
frames.
Many
salmonids'
life
stages
may
use
certain
habitats
only
on
a
seasonal
or
intermittent
basis.
For
example,
the
timing
of
migration
and
spawning
for
most
species
is
strongly
tied
to
temperature
(
Bjornn
and
Reiser
1991).

Often,
assessments
for
salmonids
focus
on
the
distribution
of
areas
used
for
spawning
and
early
rearing
(
Dunham
et
al.
2001).
Even
though
the
importance
of
spawning
and
rearing
habitat
is
obvious,
other
components
of
the
life
cycle
may
be
key
to
viability
or
productivity,
particularly
for
species
with
obligate
life
histories.
Such
habitats
can
include
migratory
corridors,
feeding
areas,
and
seasonal
refuges
(
Northcote
1997).
In
many
species,
loss
or
severe
degradation
of
these
habitats
can
cause
extinction
even
if
spawning
and
rearing
habitats
are
in
good
condition.
An
obvious
example
is
extinction
of
migratory
salmonid
populations
that
used
spawning
habitats
now
blocked
by
dams.
As
of
1991,
at
least
106
major
populations
of
salmon
and
steelhead
on
the
West
Coast
of
the
United
States
had
become
extinct,
with
inadequate
fish
passage
at
dams
a
primary
cause
(
Nehlsen
et
al.
1991).

Life
history
variation.
Life
history
refers
to
how
an
individual
completes
the
life
cycle.
Salmonids
may
adopt
a
"
resident"
or
"
migratory"
life
history.
Resident
fish
remain
very
close
to
their
natal
habitats
throughout
their
life
cycle,
whereas
migratory
fish
use
a
much
broader
range
of
habitat.
Each
of
these
broad
categories
has
its
own
variations.
For
example,
spawning
migrations
vary
by
time
and
location
(
e.
g.,
summer
vs.
winter
steelhead;
fall
vs.
winter
chinook).
The
length
of
juvenile
residence
in
natal
areas
may
also
be
important
(
e.
g.,
"
stream"
vs.
"
ocean"
type
chinook).

Some
species
have
relatively
fixed
life
cycles
and
life
history
patterns
(
e.
g.,
pink
salmon,
Groot
and
Margolis
1991);
others
exhibit
considerable
variation
or
polymorphism
(
e.
g.,
cutthroat
trout).
Most
Pacific
salmon
die
after
spawning,
whereas
most
species
of
trout
and
char
do
not
(
iteroparous).
Some
species,
subspecies,
races,
or
populations
have
flexible
life
histories
(
referred
to
as
"
facultative");
others
have
fixed
life
history
patterns
(
referred
to
as
"
obligatory")
(
Rieman
and
Dunham
1999).
Species
in
the
latter
category
may
be
less
resistant
to
environmental
change.

Historical
vs.
contemporary
vs.
potential
distribution.
Both
fish
distributions
and
stream
temperatures
can
be
considered
in
terms
of
"
historical,"
"
contemporary,"
or
"
potential"
distribution.
Historical
refers
to
the
distribution
of
native
salmonids
before
European
settlement.
Contemporary
refers
to
the
present
distribution
of
native
salmonids.
Potential
refers
to
the
distribution
of
native
salmonids
we
would
expect
if
natural
habitat
conditions
were
restored
to
the
fullest
extent
possible,
given
the
current
natural
capacity
(
Ebersole
et
al.
1997)
of
the
system.
In
other
words,
potential
distribution
allows
for
the
possibility
that
physical
systems
have
been
altered
such
that
historical
distributions
are
no
longer
attainable.
Widespread
declines
of
salmonids
observed
in
most
areas
4
Salmonid
Distributions
and
Temperature
(
Nehlsen
et
al.
1991,
Lee
et
al.
1997,
Thurow
et
al.
1997)
suggest
that
many
streams
are
not
currently
at
their
full
natural
potential
or
capacity.
5
Salmonid
Distributions
and
Temperature
A
primary
concern
of
managers
is
protecting
or
restoring
fish
distributions
that
maximize
population
viability
(
most
recently
reviewed
by
McElhany
et
al.
2000).
Many
efforts
are
under
way
to
define
thermal
habitat
potential
using
predictive
physical
models
(
reviewed
by
Bartholow
2000).
Prediction
of
physical
responses
is
complex,
but
is
much
simpler
than
predicting
biological
responses.

Restoration
of
the
physical
system
(
temperature,
thermal
regime)
should
be
considered
together
with
biological
requirements
(
viability,
productivity)
of
a
species.
The
physical
potential
of
a
system
constrains
what
can
be
achieved
biologically.
There
are
four
possible
scenarios
in
which
physical
system
potential
and
biological
requirements
or
potential
are
considered:

1.
System
potential
attained,
biological
goal
attained.
This
is
the
best
of
all
worlds,
where
protection
to
maintain
existing
conditions
would
be
a
prudent
management
option.

2.
System
potential
attained,
biological
goal
not
attained.
This
is
a
situation
where
nothing
can
be
done
to
enhance
the
potential
of
the
natural
system
to
attain
a
biological
goal.

3.
System
potential
not
attained,
biological
goal
not
attained.
This
is
a
situation
where
enhancement
of
system
potential
could
result
in
a
biological
benefit.

4.
System
potential
not
attained,
biological
goal
attained.
This
is
a
situation
where
enhancement
of
system
potential
could
result
in
a
biological
benefit,
but
the
current
state
of
the
biological
system
is
satisfactory
from
a
regulatory
viewpoint.

It
may
be
difficult
to
balance
the
attainment
of
biological
goals
versus
physical
system
potential,
but
the
answer
is
essential
to
long­
term
viability
and
productivity
of
salmonid
populations.
In
reality,
these
four
scenarios
represent
extremes
along
a
continuum
of
biological
requirements
and
physical
system
potential.
In
practice,
it
is
much
easier
to
define
physical
system
potential
than
to
define
"
how
much
is
enough?"
from
a
biological
perspective.
Thus,
it
may
be
difficult
to
discern
different
scenarios
based
on
biological
requirements.
In
practice,
most
management
to
date
has
focused
on
system
potential.

Defining
of
system
potential
can
be
challenging.
First,
it
is
critical
to
realize
that
perspectives
on
attainment
of
system
potential
may
depend
on
scale.
For
example,
a
local
reach
of
stream
may
be
at
system
potential,
but
part
of
a
larger
degraded
system
in
need
of
restoration.
Second,
it
is
difficult,
if
not
impossible,
to
restore
all
aquatic
habitats
to
their
historic
condition.
There
usually
are
insufficient
data
to
definitively
document
"
historic"
conditions,
but
even
limited
information
on
historic
habitat
conditions
and
fish
populations
can
provide
a
useful
perspective.
Such
determination
involves
finding
what
is
"
irreversible"
(
e.
g.,
removal
of
major
dams
and
urban
centers)
and
what
can
likely
be
accomplished
through
basin
management.

Examples.
The
historical
and
contemporary
distributions
of
resident
and
anadromous
fish
have
been
documented
in
the
Columbia
River
Basin
(
CRB)
by
the
Interior
Columbia
Basin
Ecosystem
6
Salmonid
Distributions
and
Temperature
Management
Project
(
Figures
1
to
7).
About
12,452
km
of
the
16,935
km
of
streams
that
originally
were
accessible
are
now
blocked
(
Quigley
and
Arbelbide
1997),
including
some
large
subbasins
and
many
smaller
watersheds.
Other
factors
contributing
to
the
decline
of
7
Salmonid
Distributions
and
Temperature
Figure
1.
Columbia
Basin
fall
chinook
distribution.
8
Salmonid
Distributions
and
Temperature
Figure
2.
Columbia
Basin
bulltrout
distribution.
9
Salmonid
Distributions
and
Temperature
Figure
3.
Columbia
Basin
spring
chinook
distribution.
10
Salmonid
Distributions
and
Temperature
Figure
4.
Columbia
Basin
redband
trout
distribution.
11
Salmonid
Distributions
and
Temperature
Figure
5.
Columbia
Basin
westslope
cutthroat
trout
distribution.
12
Salmonid
Distributions
and
Temperature
Figure
6.
Columbia
Basin
steelhead
distribution.
13
Salmonid
Distributions
and
Temperature
Figure
7.
Columbia
Basin
Yellowstone
cutthroat
trout
distribution.
14
Salmonid
Distributions
and
Temperature
salmonids
in
the
CRB
are
habitat
loss
(
including
thermal
degradation),
harvest,
and
direct
and
indirect
effects
of
hatcheries
(
Lichatowich
1999).
The
current
known
and
predicted
distribution
of
steelhead
trout
in
the
CRB
encompasses
46%
of
the
historical
range.
For
chinook
salmon,
the
current
known
and
predicted
distribution
encompasses
28%
of
the
historical
range
for
stream­
type
chinook
and
29%
for
ocean­
type
chinook
(
Quigley
and
Arbelbide
1997).
For
many
of
these
species
and
populations,
it
is
likely
that
both
system
potential
and
biological
goals
are
not
attained
(
scenario
3).
In
other
words,
widespread
enhancement
of
system
physical
potential
to
minimize
adverse
effects
of
altered
temperature
conditions
is
needed
in
the
region.

What
are
the
direct
effects
of
temperature?

Temperature
may
constrain
the
distribution
of
fish
through
direct
effects
on
physiological
function.
If
temperatures
are
too
warm,
metabolic
rates
may
rise
to
the
point
at
which
energy
intake
(
e.
g.,
food
consumption)
is
insufficient
to
maintain
basic
physiological
functions.
Growth
ceases,
and
compounding
effects
of
temperature
may
result
in
death.
Cold
temperatures
may
also
be
important,
particularly
where
growing
seasons
are
short
and
fish
must
endure
a
long
season
(
e.
g.,
winter)
of
scarce
resources
(
Shuter
and
Post
1990).
For
salmonids
in
the
Pacific
Northwest,
the
concern
is
unsuitably
warm
summer
temperatures.
Currently,
there
are
no
criteria
that
directly
address
excessively
cool
temperatures.

Examples.
Studies
of
thermal
effects
on
regional
salmonid
distributions
are
numerous
(
see
McCullough
1999).
These
studies
use
a
wide
variety
of
indicators.
At
larger
scales,
it
is
common
to
use
climate
indicators
of
thermal
regimes,
such
as
air
temperature,
elevation,
or
geographic
location
(
e.
g.,
Meisner
1990,
Flebbe
1994,
Keleher
and
Rahel
1996,
Dunham
et
al.
1999).
Geographic
variation
in
distribution
of
fish
populations
is
typically
studied
with
these
large­
scale
indicators.
At
finer
scales,
air
temperature
can
be
a
poor
indicator
of
fish
distributions.
Within
streams,
variation
in
local
climate
is
minimal,
but
variation
in
water
temperatures
is
often
obvious.
For
example,
geographic
variation
in
the
distribution
of
cutthroat
trout
is
strongly
tied
to
climate
gradients
(
Dunham
et
al.
1999).
At
a
smaller
scale
within
streams,
water
temperature
is
the
best
indicator
(
in
terms
of
water
temperature)
of
suitable
conditions
for
fish
(
Dunham
1999).
At
even
smaller
scales
(
e.
g.,
stream
reach
or
unit)
it
may
be
possible
to
distinguish
habitat
use
patterns,
if
local
variation
in
water
temperature
is
large
enough
to
elicit
a
biologically
significant
response
(
e.
g.,
Torgerson
et
al.
1999;
Ebersole
et
al.,
in
press).

Most
attempts
to
relate
salmonid
distributions
to
temperature
are
based
on
air
temperatures,
which
are
widely
available.
Air
temperature
and
groundwater
temperature
are
known
to
be
related
(
either
directly
or
indirectly),
which
is
believed
to
explain
the
association
between
air
temperatures
and
salmonid
distributions
on
a
regional
scale
(
e.
g.,
>
104
m,
or
6th
field
hydrologic
unit
code;
see
Rieman
et
al.
1997).
However,
air
temperature
generally
has
only
a
weak
direct
influence
on
surface
water
temperature
(
Poole
and
Berman
in
press).

More
recent
studies
have
focused
on
direct
associations
between
fish
distributions
and
surface
water
temperatures.
As
digital
data
loggers
and
remote
sensing
(
e.
g.,
forward­
looking
infrared
ideography,
Torgerson
et
al.
1999)
become
more
accessible,
reliance
on
indirect
measures
of
aquatic
thermal
regimes
(
e.
g.,
regional
climatic
or
air
temperatures)
will
be
less
necessary.
15
Salmonid
Distributions
and
Temperature
What
are
indirect
effects
of
temperature
on
fish
distributions?

The
direct
effects
of
temperature
are
obvious,
but
indirect
effects
can
be
important
as
well.
Multiple
stressors
(
see
Multiple
Stressors
issue
paper)
can
modify
the
effect
of
temperature
on
probability
of
survival
under
different
thermal
regimes.
In
colder
seasons,
fish
may
be
vulnerable
to
warm­
blooded
predators,
including
birds
and
mammals
(
Conduce
et
al.
1998).
In
warm
seasons,
thermal
stress
may
similarly
render
fish
susceptible
to
predators,
competitors,
or
disease.
Patterns
of
habitat
use
may
change.
Abundance
of
prey
may
also
change
(
e.
g.,
Li
et
al.
1994).
Interactions
of
these
factors
with
temperature
may
affect
fish
distributions
and
responses
to
temperature,
as
in
the
following
examples.

Biotic
interactions.
The
response
of
a
species
to
a
given
thermal
environment
can
be
modified
dramatically
by
biotic
interactions
(
e.
g.,
competition,
disease,
predation)
within
or
among
species.
In
some
cases,
this
influence
may
affect
the
distribution
of
a
species.

Within
a
species,
temperature
may
affect
important
life
history
attributes,
such
as
size
and
age
at
emigration
and
return
times
for
spawning
adults.
Within
cohorts,
intraspecific
competition
for
limited
resources
(
e.
g.,
food,
shelter,
mates)
may
be
affected,
possibly
leading
to
variation
in
competitive
ability
and
fitness
of
individuals
and
changing
patterns
of
growth
and
survival.
The
subtle
influences
of
these
factors
on
the
distribution
of
fish
within
aquatic
habitats
has
not
been
documented
in
the
published
literature.

There
is
better
evidence
for
the
influence
of
temperature
on
distribution
of
fishes
among
different
species.
In
studies
by
Reeves
et
al.
(
1987),
juvenile
steelhead
production
was
the
same
at
water
temperatures
of
53.6­
59
°
F
(
12­
15
°
C)
whether
red
shiners
were
present
or
not.
At
warmer
temperatures
(
66.2­
71.6
°
F
[
19­
22
°
C]),
steelhead
production
was
lower
when
shiners
were
present
than
when
shiners
were
absent.
Additional
examples
can
be
found
in
the
Behavior
and
Multiple
Stressors
issue
papers.

Habitat
size
and
isolation.
Thermal
gradients
often
result
in
erratic
distribution
of
fish
populations
(
Dunham
et
al.
2001).
Changes
in
the
size
and
distribution
of
habitats
result
in
habitat
fragmentation,
which
has
been
documented
for
several
species
(
e.
g.,
Rieman
and
Dunham
2000).
Generally,
as
habitat
size
decreases
and
isolation
increases,
the
occurrence
of
fish
decreases.
For
bull
trout
(
Salvelinus
confluentus)
and
cutthroat
trout
(
Oncorhynchus
clarki)
limited
evidence
suggests
these
species
are
unlikely
to
be
found
in
watersheds
with
surface
areas
of
less
than
roughly
105
ha
(
Dunham
et
al.,
in
press).
Available
data
are
not
sufficient
to
propose
minimum
area
requirements
for
any
species
(
Rieman
and
Dunham
2000).
Populations
in
smaller
habitats
are
assumed
to
be
more
vulnerable
to
chance
extinction,
or
extinction
caused
by
deterministic
factors
such
as
replacement
by
competitors
or
land
use
impacts
(
see
McElhany
et
al.
2000).

What
is
meant
by
"
scale"
and
"
level?"
At
what
"
level"
should
we
be
concerned
with
temperature
criteria
to
protect
fish
distributions?

Temperature
can
affect
fishes
at
several
scales
and
levels.
Scale
refers
to
the
space
or
time
dimensions
of
a
problem,
whereas
level
refers
to
the
ways
in
which
physical
or
biological
processes
are
organized
(
for
details,
see
Allen
1998).
A
local
population
may
be
considered
as
a
16
Salmonid
Distributions
and
Temperature
level
of
biological
organization,
for
example.
Local
populations
for
salmonids
may
correspond
to
the
distribution
of
spawning
and
rearing
areas
(
Dunham
et
al.
in
press).
In
terms
of
scale,
local
populations
can
occupy
very
small
or
very
large
watersheds.
Thus,
"
scale"
and
"
level"
are
not
exactly
synonymous.

Research
on
salmonid
habitat
has
addressed
a
wide
variety
of
spatiotemporal
scales
and
levels.
In
the
1970s
and
1980s,
research
often
focused
on
fishery
production.
Numerous
studies
addressed
the
relationship
between
standing
crop
of
salmonids
and
site­
specific
habitat
characteristics
(
e.
g.,
pools,
cover,
substrate).
These
models
were
often
limited
by
their
lack
of
transferability
in
space
or
time
and
poor
predictive
ability
(
Fausch
et
al.
1988).
Existing
EPA
temperature
criteria
(
U.
S.
EPA
1998)
are
site­
specific
and
do
not
address
larger
scale
issues
of
landscape
processes
and
fish
distributions.

Recent
models
have
addressed
aquatic
habitat
at
larger
spatial
scales
(
Johnson
and
Gage
1997;
see
Spatial­
Temporal
Issue
Paper)
and
focused
more
on
patterns
of
species
diversity,
distribution,
and
occurrence
than
on
standing
crop
(
Dunham
et
al.
in
press,
Angermeier
et
al.
2001).
Larger
scale
approaches
to
salmonid
habitat
focus
on
both
habitat
characteristics
and
the
spatial
context
of
a
habitat
in
the
landscape
(
Rieman
and
Dunham
1999).
Part
of
the
motivation
for
a
larger
scale
approach
is
the
need
for
models
that
address
habitat
requirements
at
the
population
level.
Population­
level
concerns
(
e.
g.,
occurrence,
persistence,
diversity)
are
increasingly
critical
in
this
region
as
the
list
of
threatened,
endangered,
and
sensitive
salmonids
grows.

Information
on
thermal
relationships
of
salmonids
comes
from
a
variety
of
laboratory
and
field
studies
 
how
do
we
integrate
work
conducted
at
different
scales
or
levels
of
organization
(
e.
g.,
population
vs.
individuals)?

One
important
issue
related
to
scaling
is
the
connection
between
laboratory
studies
of
thermal
tolerance
and
thermal
habitat
use
in
the
field.
EPA
criteria
for
temperature
(
Federal
Register
1998,
Brungs
and
Jones
1977)
are
based
on
laboratory
tests
of
individual
fish
responses
to
temperature.
These
experiments
provide
a
mechanistic
basis
for
understanding
the
effects
of
temperature
on
individual
fish.
In
the
laboratory,
a
rigorous
experimental
design
can
isolate
the
effects
of
specific
factors
and
test
for
interactions
among
factors.

It
is
difficult
to
extrapolate
results
obtained
under
laboratory
conditions
to
the
field,
where
many
uncontrolled
factors
interact
simultaneously.
Nonetheless,
it
is
in
the
field
that
temperature
has
a
potentially
important
role.
Although
it
is
sometimes
possible
to
conduct
large­
scale
field
experiments,
field
studies
more
often
involve
analyses
of
correlation
or
association
between
factors,
such
as
fish
distribution
and
temperature.
Obviously,
such
correlations
do
not
necessarily
constitute
cause
and
effect.
Development
of
temperature
criteria
must
therefore
involve
a
combination
of
approaches,
including
laboratory
experiments
and
field
studies.
Integrating
pattern
and
process
across
multiple
scales
or
levels
of
biological
organization
is
essential
for
correct
ecological
inference
(
Werner
1998).
Following
are
several
examples.
17
Salmonid
Distributions
and
Temperature
EPA
Fish
and
Temperature
Database
Matching
System
(
FTDMS).
A
simple
example
of
integrating
field
and
laboratory
studies
comes
from
results
of
the
EPA's
FTDMS,
Eaton
et
al.
1995).
Eaton
et
al.
(
1995)
found
a
close
correspondence
between
laboratory­
derived
thermal
tolerance
limits
and
maximum
water
temperatures
in
the
field.
For
salmonids,
maximum
water
temperatures
in
the
field
were
33.8­
39.2
°
F
(
1­
4
°
C)
cooler
than
limits
indicated
by
laboratory
studies.
The
fact
that
fish
distributions
in
the
field
corresponded
to
cooler
temperatures
suggests
that
sublethal
effects
may
be
important.
This
may
occur
when
temperature
directly
or
indirectly
acts
as
one
of
multiple
stressors
(
see
Temperature
Interaction
issue
paper).

Point
Observation
of
Cold­
Water
Fish
in
Relation
to
Temperature
Sometimes
simple
observations
of
fish
in
unusually
warm
water
are
used
to
support
(
or
reject)
proposed
temperature
criteria
or
thresholds.
Such
observations
do
not
indicate
anything
about
individual
fitness
or
population
health.
Furthermore,
they
ignore
the
essential
chain
of
inference
that
should
be
made
using
both
laboratory
and
field
observations
(
Werner
1998).
Observations
of
fish
in
"
unusual"
(
or
any)
conditions
should
be
interpreted
in
a
probabilistic
context.
For
example,
given
the
observed
thermal
regime,
what
is
the
probability
that
a
given
species
will
occur?
Determining
this
requires
information
(
and
evidence)
on
a
fuller
range
of
thermal
conditions
and
is
much
more
informative.
Salmonid
fish
may
occasionally
occur
in
"
hot"
water,
but
in
general
they
are
much
more
likely
to
occur
when
temperatures
are
cooler.
This
is
illustrated
by
the
wide
range
of
temperatures
where
salmonids
and
other
species
are
observed
to
occur
(
Figure
8).

A
useful
perspective
can
be
found
in
humans'
use
of
high­
temperature
environments.
In
many
cultures,
it
is
common
to
engage
in
recreational
or
ritual
use
of
steam
baths,
saunas,
sweat
lodges,
hot
springs,
or
other
extremely
warm
microclimates.
Although
limited
use
of
these
environments
is
common,
they
are
by
no
means
suitable
in
the
long
term;
the
negative
health
effects
are
obvious
to
humans.
Fish,
like
humans,
will
occasionally
be
found
in
thermal
habitats
that
are
unsuitable
for
long­
term
(
and
sometimes
even
short­
term)
health.
In
some
cases,
short
forays
into
physiologically
stressful
habitats
may
provide
a
net
benefit.
Many
prey
organisms
make
use
of
predator­
free
space,
which
often
exists
at
the
extremes
of
physiological
tolerance
for
predators
(
e.
g.,
Rahel
et
al.
1994).

Is
unoccupied
habitat
relevant
to
temperature
requirements
of
salmonids?

Thermal
habitat
can
be
utilized
at
a
variety
of
spatial
and
temporal
scales
(
see
also
Spatial/
Temporal
issue
paper).
Spatial
and
temporal
variation
in
the
availability
of
thermal
habitat
may
be
an
important
constraint.
Temperature
criteria
should
address
all
temperatures
likely
to
be
used
by
fish,
not
just
upper,
lower,
or
"
optimal"
temperatures.
It
is
the
full
range
of
thermal
variability
that
provides
a
context
for
continued
evolution
of
species
(
Lichatowich
1999).
Distribution
of
"
habitat"
can
extend
well
beyond
that
which
is
currently
occupied
by
a
species
or
population.

Because
of
natural
variation
in
space
and
time,
most
fish
occupy
landscapes
with
a
considerable
amount
of
suitable
but
unoccupied
habitat.
Unoccupied
habitat
is
a
natural
consequence
of
extinction
and
recolonization,
natural
habitat
succession,
and
human
influences
on
fish
populations
and
habitats
(
Reeves
et
al.
1995,
Rieman
and
Dunham
2000).
Therefore,
the
distribution
of
habitat
needed
by
fish
may
extend
well
beyond
that
which
is
currently
occupied.
18
Salmonid
Distributions
and
Temperature
19
Salmonid
Distributions
and
Temperature
This
is
especially
true
for
most
threatened
and
endangered
species
because
current
distributions
are
reduced
or
declining.

Unoccupied
habitat
is
a
potentially
controversial
issue,
particularly
because
it
can
be
difficult
to
identify
areas
needing
protection.
When
fish
are
present,
the
choice
of
habitat
to
protect
or
restore
can
be
relatively
obvious.
When
fish
are
not
present,
the
choice
must
be
guided
by
information
on
historical
and
potential
distributions
of
fish
and
suitable
habitat,
and
potential
sources
of
natural
recolonization.

Conclusion
Salmonids
in
the
Pacific
Northwest
evolved
in
habitats
with
large
amounts
of
cold,
clean
water.
Their
life
histories
and
ecology
are
strongly
tied
to
natural
thermal
regimes.
Region­
wide
declines
in
salmonids
have
paralleled
the
loss
and
fragmentation
of
formerly
large
and
interconnected
cold
water
habitats,
and
changes
in
thermal
regimes
(
see
Spatial/
Temporal
issue
paper).
Temperature
is
widely
appreciated
as
an
important
factor
affecting
not
only
the
health
of
individual
fish,
but
also
entire
populations
and
species
assemblages.
Direct
effects
of
temperature
may
be
obvious,
or
temperature
may
interact
with
other
important
variables
to
indirectly
affect
salmonids.

Temperature
criteria
that
explicitly
consider
the
thermal
requirements
of
salmonids
at
multiple
spatial
and
temporal
scales
(
see
also
Spatial/
Temporal
issue
paper),
and
the
connection
between
salmonids
and
natural
thermal
regimes,
will
be
most
protective
of
existing
populations
and
offer
a
means
for
restoring
depressed
populations.
In
many
cases,
protection
and
restoration
of
thermal
habitat
must
extend
beyond
the
current
boundaries
of
existing
occupied
and/
or
suitable
habitat,
because
current
fish
distributions
are
significantly
reduced
from
their
historical
extent.

Attempts
to
set
temperature
criteria
must
balance
what
is
known
and
not
known
about
the
physical
system
potential
and
biological
requirements
of
salmonids.
Consideration
of
unknown
factors
is
essential
in
determining
precautionary
measures
to
avoid
adverse
effects
related
to
temperature
criteria.
General
guidelines
for
assessing
salmonid
populations
(
e.
g.,
McElhany
et
al.
2000)
provide
a
means
for
determining
the
"
knowns"
and
"
unknowns."
Full
consideration
of
the
weight
of
evidence
about
current
and
potential
fish
distribution
and
physical
system
potential,
including
thorough
documentation
of
assumptions
and
knowledge
gaps,
is
needed
in
establishing
and
implementing
temperature
criteria
to
support
healthy
(
viable,
productive,
and
fishable)
salmonid
populations.
20
Salmonid
Distributions
and
Temperature
Literature
Cited
Allen
TFH.
1998.
The
landscape
"
level"
is
dead:
Persuading
the
family
to
take
it
off
the
respirator.
In:
Peterson
DL,
Parker
VT,
eds.
Ecological
Scale:
Theory
and
Applications.
New
York:
Columbia
University
Press,
pp.
35­
54.

Angermeier
PL,
Krueger
KL,
Dolloff
CA.
In
Press.
Discontinuity
in
stream­
fish
distributions:
Implications
for
assessing
and
predicting
species
occurrences.
In:
Scott
JM,
Heglund
PJ,
Samson
F,
Haufler
J,
Morrison
M,
Raphael
M,
Wall
B,
eds.
Predicting
Species
Occurrences:
Issues
of
Accuracy
and
Scale.
Covelo,
CA:
Island
Press.

Bartholow
JM.
2000.
The
stream
segment
and
stream
network
temperature
models:
A
self­
study
course.
Version
2.0.
U.
S.
Geological
Survey
Open
File
Report
99­
112.

Bjornn
TC,
Reiser
DW.
1991.
Habitat
requirements
for
salmonids.
Am
Fish
Soc
Special
Pub
19:
83­
138.

Brungs
WA,
Jones
BR.
1977.
Temperature
criteria
for
freshwater
fish:
Protocol
and
procedures.
Environmental
Research
Laboratory­
Duluth,
Office
of
Research
and
Development,
U.
S.
Environmental
Protection
Agency,
EPA­
600/
3­
77­
061,
Duluth,
MN.

Cunjak
RA,
Prowse
TD,
Parrish
DL.
1998.
Atlantic
salmon
(
Salmo
salar)
in
winter:
"
The
season
of
parr
discontent?"
Can
J
Fish
Aquat
Sci
55(
Suppl.
1):
161­
180.

Dunham
JB.
1999.
Temperature
criteria
for
Oregon's
Lahontan
cutthroat
trout.
Report
to
Oregon
Department
of
Environmental
Quality,
Portland,
OR.

Dunham
JB,
Rieman
BE.
1999.
Metapopulation
structure
of
bull
trout:
Influences
of
physical,
biotic,
and
geometrical
landscape
characteristics.
Ecol
Appl
9:
642­
655.

Dunham
JB,
Rieman
BE,
Peterson
JT.
In
Press.
Patch­
based
models
to
predict
species
occurrence:
Lessons
from
salmonid
fishes
in
streams.
In:
Scott
JM,
Heglund
PJ,
Morrison
M,
Raphael
M,
Haufler
J,
Wall
B,
eds.
Predicting
Species
Occurrences:
Issues
of
Accuracy
and
Scale.
Covelo,
CA:
Island
Press.

Dunham
JB,
Peacock
MM,
Rieman
BE,
Schroeter
RE,
Vinyard
GL.
1999.
Local
and
geographic
variability
in
the
distribution
of
stream­
living
Lahontan
cutthroat
trout.
Trans
Am
Fish
Soc
128:
875­
889.

Eaton
JG,
et
al.
1995.
A
field
information­
based
system
for
estimating
fish
temperature
tolerances.
Fisheries
20(
4):
10­
18.

Ebersole
JL,
Liss
WJ,
Frissell
CA.
In
Press.
Relationship
between
stream
temperature,
thermal
refugia,
and
rainbow
trout
Oncorhynchus
mykiss
abundance
in
arid­
land
streams
of
the
northwestern
United
States.
Ecol
Freshwater
Fish.

Ebersole
JL,
Liss
WJ,
Frissell
CA.
1997.
Restoration
of
stream
habitats
in
the
western
United
States:
Restoration
as
reexpression
of
habitat
capacity.
Environ
Manage
21:
1­
14.

Fausch
KD,
Hawkes
CL,
Parsons
MG.
1988.
Models
that
predict
standing
crop
of
stream
fish
from
habitat
variables:
1950­
85.
USDA
Forest
Service
General
Technical
Report
PNW­
GTR­
213.
USDA
Forest
Service
Pacific
Northwest
Research
Station,
Portland,
OR.

Flebbe
PA.
1994.
A
regional
view
of
the
margin:
Salmonid
abundance
and
distribution
in
the
southern
Appalachian
mountains
of
North
Carolina
and
Virginia.
Trans
Am
Fish
Soc
123:
657­
667.
21
Salmonid
Distributions
and
Temperature
Groot
C,
Margolis
L,
eds.
1991.
Pacific
salmon
life
histories.
Vancouver:
University
of
British
Columbia
Press.

Johnson
LB,
Gage
SH.
1997.
Landscape
approaches
to
the
analysis
of
aquatic
ecosystems.
Freshwater
Biol
37:
113­
132.

Keleher
CJ,
Rahel
FJ.
1996.
Thermal
limits
to
salmonid
distributions
in
the
Rocky
Mountain
region
and
potential
habitat
loss
due
to
global
warming:
a
geographic
information
system
(
GIS)
approach.
Trans
Am
Fish
Soc
125:
1­
13.

Lee
D,
Sedell
CJ,
Rieman
B,
Thurow
R,
Williams
J.
1997.
U.
S.
Forest
Service,
Pacific
Northwest
Research
Station,
Portland,
OR,
Report
Number
PNW­
GTR­
405.

Li
HW,
Lamberti
GA,
Pearsons
TN,
Tait
CK,
Li
JL,
Buckhouse
JC.
1994.
Cumulative
effects
of
riparian
disturbances
along
high
desert
trout
streams
of
the
John
Day
Basin,
Oregon.
Trans
Am
Fish
Soc
123:
627­
640.

Lichatowich
J.
1999.
Salmon
without
rivers:
A
history
of
the
Pacific
salmon
crisis.
Covelo,
CA:
Island
Press.

Magnuson
JJ,
Crowder
LB,
Medvick
PA.
1979.
Temperature
as
an
ecological
resource.
Am
Zool
19:
331­
343.

Matthews
WJ.
1998.
Patterns
in
freshwater
fish
ecology.
New
York:
Chapman
and
Hall.

McCullough
DA.
1999.
A
review
and
synthesis
of
effects
of
alterations
to
the
water
temperature
regime
on
freshwater
life
stages
of
salmonids,
with
special
reference
to
chinook
salmon.
Prepared
for
the
U.
S.
Environmental
Protection
Agency,
Region
10,
Seattle,
WA.

McElhany
P,
Ruckleshaus
M,
Ford
MJ,
Wainwright
T,
Bjorkstedt
E.
2000.
Viable
salmon
populations
and
the
recovery
of
evolutionarily
significant
units.
U.
S.
Dept.
of
Commerce,
NOAA
Technical
Memorandum
NMFS­
NWFSC­
42.

Meisner
JD.
1990.
Effect
of
climate
warming
on
the
southern
margins
of
the
native
range
of
brook
trout,
Salvelinus
fontinalis.
Can
J
Fish
Aquat
Sci
47:
1065­
1070.

National
Research
Council.
1996.
Upstream
 
Salmon
and
society
in
the
Pacific
Northwest.
Washington,
DC:
National
Academy
Press.

Nehlsen
W,
Williams
JE,
Lichatowich
JA.
1991.
Pacific
salmon
at
the
crossroads:
Stocks
at
risk
from
California,
Oregon,
Idaho
and
Washington.
Fisheries
16(
2):
4­
21.

Northcote
TG.
1997.
Potamodromy
in
salmonidae
 
living
and
moving
in
the
fast
lane.
N
Am
J
Fish
Manage
17:
1029­
1045.

Poole
GC,
Berman
CH.
In
Press.
An
ecological
perspective
on
in­
stream
temperature:
Natural
heat
dynamics
and
mechanisms
of
human­
caused
thermal
degradation.
Environ
Manage.

Quigley
TM,
Arbelbide
SJ,
eds.
1997.
An
assessment
of
ecosystem
components
in
the
interior
Columbia
basin
and
portions
of
the
Klamath
and
Great
Basins:
vol.
3.
Gen.
Tech.
Rep.
PNW­
GTR­
405.
Portland,
OR:
U.
S.
Department
of
Agriculture,
Forest
Service,
Pacific
Northwest
Research
Station.

Rahel
FJ,
Nutzman
JW.
1994.
Foraging
in
a
lethal
environment:
Fish
predation
in
hypoxic
waters
of
a
stratified
lake.
Ecology
75:
1246­
1253.
22
Salmonid
Distributions
and
Temperature
Reeves
GH,
Benda
LE,
Burnett
KM,
Bisson
PA,
Sedell
JR.
1995.
A
disturbance­
based
ecosystem
approach
to
maintaining
and
restoring
freshwater
habitats
of
evolutionarily
significant
units
of
anadromous
salmonids
in
the
Pacific
Northwest.
In:
Nielsen
JL,
ed.
Evolution
and
the
aquatic
ecosystem:
defining
unique
units
in
population
conservation.
American
Fisheries
Society
Symposium
17,
Bethesda,
MD.
pp.
334­
349.

Reeves
GH,
Everest
FH,
Hall
JD.
1987.
Interactions
between
the
redside
shiner
(
Richardsonius
balteatus)
and
the
steelhead
trout
(
Salmo
gairdneri)
in
western
Oregon:
The
influence
of
water
temperature.
Can
J
Fish
Aquat
Sci
44:
1603­
1613.

Rieman
BE,
Dunham
JB.
2000.
Metapopulations
and
salmonids:
A
synthesis
of
life
history
patterns
and
empirical
observations.
Ecol
Freshwater
Fish
9:
51­
64.

Rieman
BE,
Lee
DC,
Thurow
RF.
1997.
Distribution,
status,
and
likely
future
trends
of
bull
trout
within
the
Columbia
River
and
Klamath
Basins.
N
Am
J
Fish
Manage
17:
1111­
1125.

Shuter
BJ,
Post
JR.
1990.
Climate,
population
viability,
and
the
zoogeography
of
temperate
fishes.
Trans
Am
Fish
Soc
119:
314­
336.

Thurow
RF,
Lee
DC,
Rieman
BE.
1997.
Distribution
and
status
of
seven
native
salmonids
in
the
Interior
Columbia
River
Basin
and
portions
of
the
Klamath
River
and
Great
Basins.
N
Am
J
Fish
Manage
17:
1094­
1110.

Torgersen
CE,
Price
DM,
Li
HW,
McIntosh
BA.
1999.
Multiscale
thermal
refugia
and
stream
habitat
associations
of
chinook
salmon
in
northeastern
Oregon.
Ecol
Appl
9:
301­
319.

U.
S.
Environmental
Protection
Agency.
1998.
National
recommended
water
quality
criteria;
republication.
Federal
Register
63:
68354­
68364.

Werner
EE.
1998.
Ecological
experiments
and
a
research
program
in
community
ecology.
In:
Resetarits
WJ,
Bernardo
J,
eds.
Experimental
ecology:
Issues
and
perspectives.
New
York:
Oxford
University
Press.
