Effects
of
Acidic
Deposition
on
Aquatic
Resources
in
the
Southern
Appalachians
with
a
Special
Focus
on
Class
I
Wilderness
Areas
Prepared
for
the
Southern
Appalachian
Mountains
Initiative
(
SAMI)

August
20,
1996
Alan
T.
Herlihy
and
Philip
R.
Kaufmann
Department
of
Fisheries
and
Wildlife
Oregon
State
University,
c/
o
U.
S.
EPA
NHEERL
200
SW
35th
St.
Corvallis,
OR
97333
John
L.
Stoddard,
Dynamac
Corp.
c/
o
U.
S.
EPA
NHEERL
200
SW
35th
St.
Corvallis,
OR
97333
Keith
N.
Eshleman
Appalachian
Environmental
Laboratory
CEES,
University
of
Maryland
System
Frostburg,
MD
21532
Arthur
J.
Bulger,
Department
of
Environmental
Sciences
University
of
Virginia
Charlottesville,
VA
22903
iii
EXECUTIVE
SUMMARY
The
Southern
Appalachian
Mountains
Initiative
(
SAMI)
was
established
to
address
concerns
about
the
adverse
effects
of
air
pollution
on
environmental
resources
in
the
Southern
Appalachians.
Acidic
deposition
and
its
effects
on
surface
waters
is
a
major
air
pollution
issue
in
this
area.
This
report
has
two
main
objectives:

1.
Summarize
the
existing
state
of
knowledge
about
the
effects
of
acidic
deposition
on
surface
water
resources
in
the
Southern
Appalachians.

2.
Evaluate
and
make
recommendations
on
the
use
of
available
methodologies
for
predicting
future
changes
in
the
aquatic
effects
of
acidic
deposition
in
the
Southern
Appalachians.

The
major
findings
of
this
report
are
as
follows.

Factors
Controlling
the
Response
of
Surface
Waters
to
Acidic
Deposition
The
two
major
processes
influencing
the
response
of
surface
waters
in
the
Southern
Appalachians
to
acidic
deposition
are
sulfate/
nitrate
retention
and
base
cation
mobilization.
The
amount
of
watershed
sulfate/
nitrate
retention
controls
how
much
of
the
incoming
anions
from
deposition
reach
the
lake/
stream.
The
degree
of
base
cation
mobilization
controls
the
cation
composition
entering
surface
waters.
If
the
entering
anions
are
all
balanced
by
base
cations,
there
is
little
effect
on
the
acid­
base
status
of
the
water
and
consequently
little
effect
on
aquatic
biota.
However,
if
the
anions
are
balanced
by
significant
concentrations
of
acid
cations
(
H+,
Al),
surface
water
acidity
increases
and
there
can
be
significant
adverse
effects
on
many
aquatic
species.
Base
cation
mobilization
is
controlled
primarily
by
the
composition
of
the
watershed
bedrock
and
soils
and
is
reflected
in
surface
water
acid
neutralizing
capacity
(
ANC).

Current
Status
of
Aquatic
Resources
Lakes
in
the
Southern
Appalachians
are
mostly
reservoirs
and
are
not
very
numerous
(
estimated
total
of
71
lakes
>
4
ha
in
the
southern
Blue
Ridge).
The
Eastern
Lake
Survey
found
no
acidic
lakes;
5%
of
the
lakes
had
ANC

50
µ
eq/
L
and
would
be
considered
sensitive
to
acidic
deposition.

Acidic
and
very
low
ANC
streams
are
located
in
small
(<
20
km2),
upland,
forested
catchments
in
areas
of
base­
poor
bedrock.
The
National
Stream
Survey
estimated
(
in
1986)
that
of
the
62,200
km
of
streams
on
1:
250,000­
scale
maps
in
the
acid­
sensitive
part
of
the
SAMI
region,
815
km
(
1%)
were
acidic
and
4,410
km
(
7%)
had
ANC

50

eq/
L.
In
these
acidic
Southern
Appalachian
streams,
sulfate
and
nitrate
from
atmospheric
deposition
are
the
dominant
source
of
acid
anions,
and
the
low
pH
(
median
=
4.7)
and
high
levels
of
inorganic
monomeric
aluminum
(
median
=
364

g/
L)
are
causing
damage
to
aquatic
biota.
Watersheds
in
the
SAMI
region
are
currently
retaining
significant
proportions
of
the
incoming
sulfur
and
nitrogen
from
acidic
deposition.

The
SAMI
Class
I
wilderness
areas
are
much
more
sensitive
to
acidic
deposition
than
the
region
as
a
whole.
Based
on
geology,
physiography,
and
stream
chemistry,
the
10
Class
I
areas
in
the
SAMI
region
can
be
aggregated
into
four
groups
for
assessing
the
aquatic
effects
of
acidic
deposition:

1.
West
Virginia
Plateau:
Dolly
Sods
and
Otter
Creek
wilderness
areas
2.
Northern
Blue
Ridge:
James
River
Face
wilderness
area
and
Shenandoah
National
Park
iv
3.
Southern
Blue
Ridge:
Great
Smoky
Mountains
National
Park,
Joyce
Kilmer/
Slickrock,
Linville
Gorge,
Shining
Rock,
and
Cohutta
wilderness
areas
4.
Alabama
Plateau:
Sipsey
wilderness
area
In
terms
of
adverse
aquatic
effects
of
acidic
deposition,
the
four
Class
I
groups
can
be
ranked:
West
Virginia
Plateau
>>
Northern
Blue
Ridge
>
Southern
Blue
Ridge
>
Alabama
Plateau.

Streams
in
the
West
Virginia
Plateau
Class
I
areas
had
the
highest
percentage
of
acidic
stream
length
(
53%
in
Otter
Creek,
82%
in
Dolly
Sods),
the
highest
sulfate
and
inorganic
aluminum
concentrations,
and
the
lowest
pH
of
any
of
the
Class
I
areas.
Organic
acids
play
a
significant
role
in
a
portion
(
13%)
of
the
stream
length
in
Otter
Creek,
but
the
rest
of
the
stream
length
in
both
areas
was
dominated
by
acid
anions
from
deposition.
Streams
in
both
Otter
Creek
and
Dolly
Sods
are
heavily
impacted
by
acidic
deposition.
The
Alabama
Plateau
Class
I
area
(
Sipsey)
is
of
least
concern
for
acidic
deposition
impacts
because
sulfate
appears
to
be
at
steady
state
with
deposition
and
this
area
has
higher
streamwater
ANC
concentrations.

The
Plateau
areas
retain
less
sulfate
than
do
the
Blue
Ridge
areas,
so
any
delayed
effect
of
acidic
deposition
will
be
more
pronounced
in
the
Blue
Ridge.
Watershed
soils
in
all
the
Blue
Ridge
wilderness
areas
retain
most
of
the
sulfate
entering
from
deposition.
In
these
areas,
low
ANC
streams
are
common
and
some
acidic
streams
are
found
in
areas
of
resistant
bedrock
and/
or
higher
elevations.
Streams
in
the
northern
Blue
Ridge
areas
have
higher
sulfate
concentrations
than
streams
in
the
southern
Blue
Ridge
areas
and
appear
to
be
more
influenced
by
acidic
deposition.

The
southern
Blue
Ridge
areas
stand
out
from
the
other
Class
I
areas
in
terms
of
having
the
highest
streamwater
nitrate
concentrations
(
lowest
nitrogen
retention).
It
appears
that
nitrate
is
breaking
through
these
watersheds
and
is
entering
streams
in
concentrations
that
approach
and
sometimes
exceed
sulfate
concentrations.

Few
detailed
studies
of
episodic
acidification
have
been
conducted
in
forested
watersheds
in
the
Southern
Appalachian
Mountains.
However,
results
from
several
field
studies
suggest
that
streams
in
the
region
with
antecedent
baseflow
ANC
values
below
about
25

eq/
L
(
2,280
km
or
4%
of
the
1:
250,000­
scale
stream
network)
may
experience
substantial
depressions
in
pH
(>
0.5
units)
and
increases
in
inorganic
aluminum
concentrations
(>
50

g/
L)
during
major
rainstorm
and
snowmelt
periods.

Chemical
properties
of
surface
waters
that
are
most
important
in
influencing
biological
responses
to
acidic
deposition
are
pH,
aluminum,
and
calcium
concentrations.

Many
aquatic
species
cannot
survive,
reproduce,
or
compete
in
acidic
waters.
Thus,
with
increasing
acidity,
the
"
acid­
sensitive"
species
are
lost
and
species
richness
(
the
number
of
species
living
in
a
given
lake
or
stream)
declines.
These
changes
in
aquatic
community
structure
occur
at
chronic
pH
levels
<
6.0B6.5.
Ecosystem
level
processes,
such
as
decomposition,
nutrient
cycling,
and
productivity,
are
fairly
robust
and
are
affected
only
at
relatively
high
levels
of
acidity
(
e.
g.,
chronic
pH
<
5.0B5.5).

Biological
damage
is
occurring
in
this
region
due
to
the
high
levels
of
inorganic
aluminum
and
low
pH
in
acidic
streams.
Direct
quantification
of
the
extent
of
biological
effects,
however,
is
not
possible
from
existing
data.

Current
Trends
in
Acidification
Impacts
There
are
few
long­
term
monitoring
sites
in
the
SAMI
region
from
which
to
draw
conclusions
about
trends.
In
the
control
water­
v
sheds
at
Coweeta
(
North
Carolina
Blue
Ridge)
and
White
Oak
Run
in
Shenandoah
National
Park,
sulfate
concentrations
in
streamwater
have
been
increasing
over
the
last
10­
20
years.
The
sulfate
increases
at
these
sites
are
consistent
with
a
gradual
saturation
of
soils
in
the
region
with
sulfate
from
deposition,
and
have
been
predicted
by
acidification
models.

It
is
likely
that
many
parts
of
the
SAMI
region
have
undergone
increases
in
stream
nitrate
over
the
past
several
decades.
At
Fernow
Experimental
Forest
in
the
West
Virginia
Appalachian
Plateau,
streamwater
nitrate
concentrations
have
increased
from
near
zero
to
50B60

eq/
L
at
baseflow
from
1970
to
the
present.
In
the
Great
Smoky
Mountains
National
Park,
streamwater
nitrate
shows
strong
correlations
with
elevation
and
forest
age,
with
the
highest
concentrations
occurring
at
high
elevations
(
where
deposition
is
highest)
and
in
areas
of
old­
growth
forest,
where
biological
demand
for
nitrogen
is
lowest.

More
recently,
chemical
trends
in
Shenandoah
National
Park
have
been
altered
as
a
result
of
forest
defoliation
by
gypsy
moth
larvae.
Defoliation
has
resulted
in
large
increases
in
streamwater
nitrate,
decreases
in
sulfate,
and
little
change
in
ANC
or
pH
at
baseflow.

Data
do
not
exist
to
determine
trends
in
the
episodic
acidification
of
surface
waters
related
to
atmospheric
deposition
in
the
Southern
Appalachian
Mountains.
However,
episodic
mobilization
of
nitrate
resulting
from
recent
forest
defoliation
by
the
gypsy
moth
caterpillar
has
been
shown
to
exacerbate
episodic
changes
in
ANC,
pH,
and
aluminum
in
streams
in
the
region.

There
are
virtually
no
data
from
which
to
draw
quantitative
conclusions
about
recent
trends
in
biological
condition
due
to
acidification
Methodologies
for
Predicting
Impacts
of
Acidic
Deposition
on
Aquatic
Resources
The
two
dynamic
watershed
models
that
would
be
most
useful
for
a
SAMI
assessment
are
ILWAS
and
MAGIC.
Both
these
models
attempt
to
describe
the
major
processes
controlling
surface
water
response
to
acidic
deposition
and
require
information
on
deposition,
watershed
attributes,
soils,
hydrologic
flow,
and
water
chemistry.
ILWAS
models
more
processes
with
more
compartments
than
MAGIC
and
thus
requires
more
data
input
and
has
a
greater
computational
complexity.
Both
models
were
extensively
tested
in
the
National
Acid
Precipitation
Assessment
Program
(
NAPAP)
and
there
was
no
evidence
that
either
model
was
more
accurate
than
the
other.
Most
of
the
NAPAP
modeling
effort
focused
on
sulfur
dynamics.
A
number
of
models
that
examine
nitrogen
dynamics
have
been
developed
in
the
past
few
years
(
MAGICWAND
MERLIN,
PnET­
CN).

Several
models
are
available
for
predicting
future
episodic
effects
on
surface
waters
in
the
region;
these
models
were
all
developed
by
linking
a
long­
term
acidification
model
(
e.
g.,
MAGIC)
with
a
hydrological
mixing
model
(
e.
g.,
a
two­
component
mixing
model).
Although
major
uncertainties
are
associated
with
applying
any
of
these
models,
the
models
are
expected
to
predict
the
occurrence
of
biologically
significant
episodic
effects
prior
to
the
onset
of
chronic
acidification
effects.

Modeling
of
biotic
effects
of
acidic
deposition
has
focused
mainly
on
fish.
Fish
effects
models
are
either
empirical,
toxicity
based,
or
a
combination
of
the
two.
The
model
used
in
the
NAPAP
assessment
was
based
on
calculating
an
Acid
Stress
Index
(
ASI)
derived
from
surface
water
pH,
ANC,
and
inorganic
aluminum.
The
ASI
values
were
related
to
field
observations
of
fish
distributions
to
set
critical
ASI
values
for
effects
on
sensitive,
intermediate,
and
tolerant
fish
species.
vi
Results
from
NAPAP
Assessments
The
Direct/
Delayed
Response
Project
and
the
NAPAP
Integrated
Assessment
projected
future
acidification
based
on
the
MAGIC
model
at
a
subset
of
the
National
Stream
Survey
sites
in
the
mid­
Appalachians
(
Pennsylvania,
Maryland,
Virginia,
West
Virginia)
and
the
southern
Blue
Ridge.

NAPAP
future
projections
(
50
year)
in
the
mid­
Appalachians
show
that
at
1985
rates
of
deposition,
the
number
of
acidic
streams
is
projected
to
triple,
while
the
number
with
ANC

50

eq/
L
would
increase
by
a
factor
of
1.6.
At
sulfur
deposition
20%
and
30%
greater
than
1985
levels,
the
number
of
acidic
streams
is
projected
to
increase
by
factors
of
5.4
and
6.2,
respectively.
A
20%
to
30%
reduction
in
sulfur
deposition
would
be
necessary
to
prevent
further
acidification.

NAPAP
future
projections
(
50
year)
in
the
southern
Blue
Ridge
show
that
1985
deposition
rates
are
projected
to
increase
the
percentage
of
acidic
streams
from
0
to
10%;
the
number
of
streams
with
ANC

50

eq/
L
is
projected
to
increase
by
a
factor
of
1.6.
At
sulfur
deposition
20%
and
30%
greater
than
1985
levels,
the
number
of
acidic
streams
is
projected
to
increase
to
10%
of
the
modeled
population;
the
number
of
streams
with
ANC

50

eq/
L
would
increase
by
a
factor
of
2.2.
A
30%
to
50%
reduction
in
sulfur
deposition
would
be
necessary
to
prevent
further
acidification

Recommendations
for
SAMI
Assessment
We
recommend
that
MAGIC
be
used
to
model
surface
water
acid­
base
chemistry
and
that
the
PnET­
CN
model
be
used
to
input
time
to
nitrogen
saturation
into
MAGIC.
We
also
recommend
that
the
regression
approach
of
Eshleman
be
applied
to
MAGIC
projections
to
predict
episodic
changes
in
stream
chemistry.
For
modeling
effects
on
fish,
we
recommend
the
acid
stress
index
(
ASI)
approach.
We
have
outlined
four
possible
levels
of
effort
for
the
SAMI
assessment:

1.
Level
1:
Qualitative
summary
of
existing
modeling
information
2.
Level
2:
Re­
run
MAGIC
model
at
existing
regional
network
sites
with
new
SAMI
deposition
scenarios.

3.
Level
3:
Collect
new
field
data
and
run
the
MAGIC
model
for
three
acidic/
low
ANC
sites
in
each
of
the
Class
I
wilderness
area
groups
(
e.
g.,
West
Virginia
Plateau).
Also
collect
the
information
needed
to
run
the
PnET­
CN
nitrogen
model
and
the
episodic
acidification
model.
Do
for
one,
two,
three,
or
all
groups
as
resources
allow.

4.
Level
4:
Do
Level
3
analyses
for
the
regional
network
of
sites
in
the
SAMI
region.

Fish
Effects:
Fish
effects
can
be
added
through
the
ASI
approach
in
levels
2B4.
However,
field
information
on
the
relationship
between
fish
distributions
and
ASI
values
needs
to
be
collected
and
evaluated
to
set
critical
ASI
levels.

Uncertainties
in
the
absolute
magnitudes
and
timing
of
aquatic
effects
projections
are
high,
but
we
have
confidence
in
the
projected
direction
of
change
and
in
the
relative
amounts
of
change.
Thus,
we
are
fairly
comfortable
with
running
the
models
and
making
conclusions
about
the
relative
differences
in
aquatic
effects
among
different
deposition
scenarios.
However,
we
are
uncomfortable
with
using
the
absolute
results
(
e.
g.,
acidic
stream
length
or
stream
length
that
has
lost
fish
due
to
acidic
deposition)
of
these
models
because
of
their
high
uncertainty.
We
are
very
uncomfortable
with
taking
these
absolute
results,
linking
them
together
with
absolute
results
from
other
effects
models
(
e.
g.,
vii
visibility,
ozone),
running
them
all
through
a
socio­
economic
valuation
model,
and
using
some
kind
of
overall
"
cost"
estimate
as
the
decision
making
tool
for
evaluating
different
emissions
scenarios.
viii
ix
TABLE
OF
CONTENTS
Section
Page
Executive
Summary
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
iii
List
of
Illustrations
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
xii
List
of
Tables
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
xiii
Acknowledgements
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
xiv
Glossary
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
xv
1.
INTRODUCTION
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1.1
Terminology
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
1.2
The
SAMI
Region
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
2.
FACTORS
CONTROLLING
AQUATIC
RESPONSE
TO
ACIDIC
DEPOSITION
.
.
.
.
.
.
4
2.1
Overview
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4
2.2
Acidic
Deposition
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4
2.2.1
Current
Pattern
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4
2.2.2
Historical
Evidence
for
Acidification
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7
2.3
Hydrologic
Flow
Path
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8
2.4
Anion
Mobility/
Retention
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8
2.4.1
Sulfur
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8
2.4.2
Nitrogen
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8
2.5
In­
lake
and
In­
stream
Processes
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
10
2.6
Other
Acidity
Sources
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
10
3.
CURRENT
STATUS
OF
AQUATIC
RESOURCES
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
11
3.1
Surface
Water
Acid­
Base
Chemistry
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
11
3.1.1
Regional
Status
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
11
3.1.1.1
Lakes
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
11
3.1.1.2
Streams
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
11
3.1.2
Relationship
of
Stream
ANC
to
Geology
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
14
3.1.3
Detailed
Analysis
of
Class
I
Areas
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
14
3.1.3.1
West
Virginia
Appalachian
Plateau
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
20
3.1.3.2
Northern
Blue
Ridge
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
23
3.1.3.3
Southern
Blue
Ridge
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
25
3.1.3.4
Alabama
Appalachian
Plateau
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
29
3.1.4
Class
I
Area
Summary
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
29
3.2
Episodic
Chemical
Conditions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
33
3.2.1
Synopsis
of
Synthesis
Documents
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
33
3.2.2
Current
Episodic
Impacts
in
the
Southern
Appalachians
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
35
3.2.2.1
Local
Scale
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
35
3.2.2.2
Regional
Scale
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
36
3.3
Biological
Effects
of
Acidic
Deposition
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
36
x
3.3.1
Synopsis
of
Synthesis
Documents
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
36
3.3.1.1
Chemical
Factors
Influencing
Biota
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
38
3.3.1.2
Effects
on
Species
Richness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
38
3.3.1.3
Effects
on
Ecosystem
Level
Processes
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
39
3.3.1.4
Recovery
of
Biological
Communities
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
39
3.3.1.5
Developing
Regional
Models
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
39
3.3.1.6
Documentation
of
Effects
on
Fish
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
40
3.3.2
Current
Biological
Impacts
in
the
Southern
Appalachians
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
40
3.3.2.1
Mid­
Appalachians
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
41
3.3.2.2
Interior
Southeast
Streams
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
42
4.
RECENT
TRENDS
IN
ACIDIFICATION
IN
THE
SOUTHERN
APPALACHIANS
.
.
.
.
44
4.1
Trends
in
Surface
Water
Chemistry
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
44
4.2
Trends
in
Episodic
Effects
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
44
4.3
Trends
in
Biological
Effects
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
45
5.
METHODOLOGIES
FOR
PREDICTING
FUTURE
EFFECTS
OF
ACIDIC
DEPOSITION
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
47
5.1
Surface
Water
ChemistryModels
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
47
5.1.1
Steady­
state
Models
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
47
5.1.2
Dynamic
Models
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
47
5.1.2.1
ILWAS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
48
5.1.2.2
MAGIC
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
48
5.1.3
Nitrogen
Models
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
48
5.1.3.1
Model
of
Acidification
of
Groundwater
in
Catchments
B
With
Aggregated
Nitrogen
Dynamics
(
MAGICBWAND)
.
.
.
.
.
.
.
.
.
.
.
.
.
49
5.1.3.2
Model
of
Ecosystem
Retention
and
Loss
of
Inorganic
Nitrogen
(
MERLIN)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
49
5.1.3.3
Net
Photosynthesis
and
Evapo­
Transpiration
(
PnET)
B
Carbon
and
Nitrogen
(
CN)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
49
5.2
Episodic
Chemical
Models
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
50
5.3
Modeling
the
Effects
of
Acidic
Deposition
on
Fish
Communities
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
50
5.3.1
Empirical
Models
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
50
5.3.2
Toxicity
Models
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
51
5.3.3
Combined
Toxicity­
Field
Models
C
the
Acid­
Stress
Index
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
51
5.3.3.1
Acid
Stress
Index
Model
Structure
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
52
5.3.3.2
Acid
Stress
Index
Model
Evaluation
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
53
6.
PREDICTIVE
STUDIES
OF
ACIDIC
DEPOSITION
EFFECTS
IN
THE
SOUTHERN
APPALACHIANS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
55
6.1
NAPAP
Modeling
Efforts
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
55
6.1.1
Model
Results
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
56
6.1.1.1
Sensitivity
to
Acidification
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
56
xi
6.1.1.2
Projections
for
Current
and
Increased
Deposition
.
.
.
.
.
.
.
.
.
.
.
.
.
.
56
6.1.1.3
Reduced
Deposition
Scenarios
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
58
6.1.1.4
Chemical
Conditions
for
Fish
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
58
6.1.1.5
Uncertainties
in
Model
Projections
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
58
6.2
Nitrogen
Bounding
Study
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
59
7.
RECOMMENDATIONS
FOR
SAMI
ASSESSMENT
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
60
7.1
Model
Recommendations
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
60
7.2
Levels
of
Assessment
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
60
7.2.1
Level
1
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
61
7.2.2
Level
2
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
61
7.2.3
Level
3
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
61
7.2.4
Level
4
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
62
7.2.5
Fish
Assessment
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
62
7.3
Limitations
and
Relationship
to
a
SAMI
Integrated
Assessment
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
62
7.4
Data
Availability
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
62
8.
LITERATURE
CITED
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
64
9.
ANNOTATED
BIBLIOGRAPHY
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
70
APPENDIX
A
B
Stream
Networks
in
the
SAMI
Class
I
Wilderness
Areas
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
83
xii
LIST
OF
ILLUSTRATIONS
Figure
Page
Figure
1
Map
of
the
Southern
Appalachian
study
area
showing
the
location
of
the
Class
I
wilderness
areas
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2
Figure
2
Conceptual
diagram
of
(
a)
major
processes
and
(
b)
hydrologic
flowpaths
that
control
surface
water
acid­
base
chemistry
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5
Figure
3
Annual
1995
volume­
weighted
pH
in
wet
deposition
in
the
United
States
measured
at
the
field
laboratories
in
the
National
Acid
Deposition
Program/
National
Trends
Network
sites
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6
Figure
4
Map
of
the
acid­
sensitive
part
of
the
Southern
Appalachians
sampled
in
the
National
Stream
Survey
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
12
Figure
5
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Dolly
Sods
wilderness
area
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
21
Figure
6
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Otter
Creek
wilderness
area
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
22
Figure
7
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
James
River
Face
wilderness
area
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
24
Figure
8
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Joyce
Kilmer/
Slickrock
wilderness
area
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
27
Figure
9
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Shining
Rock
wilderness
area
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
28
Figure
10
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Cohutta
wilderness
area
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
30
Figure
11
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Sipsey
wilderness
area
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
31
Figure
12
Future
projections
of
stream
ANC
and
pH
after
50
years
for
6
sulfate
deposition
scenarios
for
(
a)
Mid­
Atlantic
Highlands
and
(
b)
Southern
Blue
Ridge
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
57
xiii
LIST
OF
TABLES
Table
Page
Table
1
Estimates
of
Regional
Acid­
base
Status
for
Streams
and
Lakes
in
the
Southern
Appalachians
from
National
Surface
Water
Survey
Data
.
.
.
.
.
.
.
.
.
.
.
.
13
Table
2
Median
Values
for
Major
Ion
Chemistry
in
Streams
in
Class
I
Wilderness
Areas
and
in
the
entire
Southern
Appalachians
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15
Table
3
Surface
Area
and
Stream
Length
at
1:
100,000
and
1:
24,000
Map
Scales
for
Streams
in
the
Southern
Appalachians
and
in
Class
I
Wilderness
Areas
.
.
.
.
.
.
16
Table
4
Observed
and
Estimated
Steady­
state
Sulfur
and
Nitrogen
Concentrations
for
Streams
in
Class
I
Wilderness
Areas
of
the
Southern
Appalachians
.
.
.
.
.
.
.
.
.
18
Table
5
Length
of
Stream
(
km)
in
Various
ANC
Classes
in
Class
I
Wilderness
Areas
in
the
Southern
Appalachians
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
19
Table
6
Summary
of
Acidic
and
Acid­
sensitive
Stream
percentages
in
Class
I
Wilderness
Areas
and
Other
Aggregate
Physiographic
Regions
of
the
Southern
Appalachians
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
32
Table
7
General
Summary
of
Biological
Changes
Anticipated
with
Surface
Water
Acidification,
Expressed
as
a
Change
in
pH
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
37
xiv
ACKNOWLEDGEMENTS
This
study
was
funded
by
the
Southern
Appalachian
Mountains
Initiative
(
SAMI).
We'd
like
to
thank
Rebecca
Kemp,
then
Coordinator
of
SAMI's
Technical
Oversight
Committee,
and
Leslie
Cox,
SAMI
Coordinator,
for
smoothly
handling
all
administrative
aspects
of
the
project
with
SAMI.
We'd
also
like
to
thank
Rick
Webb,
the
SAMI
work
assignment
manager,
for
guidance
and
reviews
of
early
drafts
of
the
report.
The
report
also
benefitted
from
review
comments
from
a
number
of
SAMI
members.
Data
for
the
Class
I
wilderness
areas
was
kindly
provided
by
Rick
Webb
and
Frank
Deviney
from
University
of
Virginia
(
Otter
Creek,
Dolly
Sods,
and
James
River
Face),
Terry
Flum
from
University
of
Tennessee
(
Smokies),
and
Richard
Burns,
Bill
Jackson,
and
David
Wergowske
of
the
U.
S.
Forest
Service
(
Shining
Rock,
Slickrock,
Sipsey
and
Cohutta).
Gary
Lear
from
NADP
provided
the
national
1995
deposition
pH
map.
Finally,
we'd
like
to
acknowledge
Charley
Barrett
for
digitizing
the
Class
I
stream
networks,
GIS
work,
and
map
production,
and
Susan
Christie
for
technical
editing
and
producing
the
final
report.
xv
GLOSSARY
ANC
Acid
Neutralizing
Capacity
ASI
Acid
Stress
Index
DDRP
Direct/
Delayed
Response
Project
DOC
Dissolved
Organic
Carbon
ELS
Eastern
Lake
Survey
EPA
Environmental
Protection
Agency
ETD
Enhanced
Trickle
Down
FISH
Fish
in
Sensitive
Habitats
(
Shenandoah
National
Park
Project)

GSMNP
Great
Smoky
Mountains
National
Park
IA
Integrated
Assessment
(
1990
NAPAP
report)

ILWAS
Integrated
Lake
Watershed
Acidification
Study
LAF
Lake
Acidification
and
Fisheries
model
MAGIC
Model
of
Acidification
of
Groundwater
in
Catchments
MAGIC­
WAND
MAGIC
­
With
Aggregated
Nitrogen
Dynamics
model
MERLIN
Model
of
Ecosystem
Retention
and
Loss
of
Inorganic
Nitrogen
MIDAPP
Mid­
Appalachians
(
as
region
modeled
in
DDRP)

N
Nitrogen
NADP/
NTN
National
Atmospheric
Deposition
Program/
National
Trends
Network
NAPAP
National
Acid
Precipitation
Assessment
Program
NBS
Nitrogen
Bounding
Study
NHEERL
National
Health
and
Environmental
Effects
Research
Laboratory
NSS
National
Stream
Survey
NSWS
National
Surface
Water
Survey
PnET­
CN
Net
Photosynthesis
and
Evapo­
Transpiration
(
PnET)
­
Carbon
and
Nitrogen
(
CN)
model
S
Sulfur
SAMI
Southern
Appalachian
Mountains
Initiative
SBRP
Southern
Blue
Ridge
Province
(
as
region
modeled
in
DDRP
and
IA)

SOS/
T
State
of
Science/
Technology
(
1990
NAPAP
reports)

USGS
United
States
Geological
Survey
VTSSS
Virginia
Trout
Stream
Sensitivity
Study
xvi
1
1.
INTRODUCTION
The
Southern
Appalachian
Mountains
Initiative
(
SAMI)
was
established
to
address
concerns
about
the
adverse
effects
of
air
pollution
on
environmental
resources
in
the
Southern
Appalachians.
Acidic
deposition
and
its
effects
on
surface
waters
is
a
major
issue
associated
with
air
pollution
in
this
area.
For
this
study,
we
define
the
Southern
Appalachians
as
the
Blue
Ridge,
Ridge
and
Valley,
and
Appalachian
Plateau
provinces
(
as
defined
by
Fenneman,
1938)
in
the
states
of
Alabama,
Georgia,
Kentucky,
North
Carolina,
South
Carolina,
Tennessee,
Virginia,
and
West
Virginia
(
Figure
1).
This
report
has
two
main
objectives:

1.
Summarize
the
existing
state
of
knowledge
about
the
effects
of
acidic
deposition
on
surface
water
resources
in
the
Southern
Appalachians.
2.
Evaluate
and
make
recommendations
on
the
use
of
available
methodologies
for
predicting
future
changes
in
the
aquatic
effects
of
acidic
deposition
in
the
Southern
Appalachians
under
alternate
deposition
scenarios.

In
this
report,
we
examine
the
SAMI
region
as
a
whole,
but
the
primary
focus
is
on
the
10
Class
I
wilderness
areas
located
in
the
Southern
Appalachians
(
Figure
1).

A
comprehensive
study
of
the
effects
of
acidic
deposition
in
the
United
States
was
completed
under
the
National
Acid
Precipitation
Assessment
Program
(
NAPAP)
in
1990.
This
study
was
documented
in
the
27
volumes
of
NAPAP's
State
of
Science/
Technology
(
SOS/
T)
report.
The
following
six
reports
pertain
to
our
objectives:

SOS/
T
#
9
Current
Status
of
Surface
Water
Acid­
Base
Chemistry
L.
Baker
et
al.,
1990
SOS/
T
#
10
Watershed
&
Lake
Processes
Affecting
Surface
Water...
Turner
et
al.,
1990
SOS/
T
#
11
Historical
Changes
in
Surface
Water
Acid­
Base
Chem...
Sullivan,
1990
SOS/
T
#
12
Episodic
Acidification
of
Surface
Water
...
Wigington
et
al.,
1990
SOS/
T
#
13
Biological
Effects
of
Changes
in
Surface
Water
...
J.
Baker
et
al.,
1990
SOS/
T
#
14
Methods
for
Projecting
Future
Changes
in
Surface
...
Thornton
et
al.,
1990
These
SOS/
T
reports
contain
detailed
reference
lists
that
document
the
work
done
before
1990.
In
general,
we
cite
the
appropriate
SOS/
T
report
for
our
summaries
and
conclusions
rather
than
all
the
pertinent
pre­
1990
literature.
To
bridge
the
gap
between
1990
and
1995,
we
have
compiled
an
annotated
bibliography
of
pertinent
literature
published
after
1990.
The
annotated
bibliography
focuses
on
surface
waters
in
the
SAMI
region
but
it
also
includes
some
citations
relating
to
acidic
deposition
effects
in
general.

The
first
part
of
this
report
summarizes
our
current
understanding
of
acidic
deposition
effects,
broken
down
into
sections
on
important
processes
(
Section
2),
current
status
of
aquatic
resources
(
Section
3),
and
current
trends
in
acidification
(
Section
4).
Each
section
addresses
surface
water
chemistry,
episodic
effects,
and
biological
effects.
A
2
Figure
1.
Map
of
the
Southern
Appalachian
study
area
showing
the
location
of
the
Class
I
wilderness
areas.
3
review
of
assessment
methodologies
appears
in
Section
5,
followed
by
results
from
other
predictive
assessments
in
Section
6.
Section
7
provides
recommendations
for
the
SAMI
assessment.
The
report
concludes
with
an
annotated
bibliography
of
post­
1990
literature
in
Section
9.

1.1
Terminology
The
terminology
used
to
describe
acidbase
status
can
get
rather
confusing.
In
order
to
avoid
problems,
we
follow
the
conventions
developed
during
NAPAP.
In
brief,
it
is
important
to
understand
the
following
distinctions:

Acidity
is
the
amount
of
acid
in
a
sample,
usually
measured
on
the
pH
scale
(

log[
H+]).

Acid
Neutralizing
Capacity
(
ANC)
is
a
measure
of
the
ability
of
a
water
sample
to
neutralize
acid
inputs
(
determined
by
acid
titration).
ANC
is
used
as
the
main
indicator
of
the
sensitivity
of
a
surface
water
to
acid
inputs.

Acidic
refers
to
a
water
sample
that
has
lost
all
ANC.
By
definition,
we
consider
waters
with
ANC

0
to
be
acidic
(
not
pH
<
7).
Most
waters
with
ANC

0
have
pH
<
5.3.

Acidification
refers
to
declines
in
ANC
or
pH
over
time.
Acidified
waters
had
higher
ANC
or
pH
in
the
past.
A
lake
or
stream
may
have
acidified
but
not
be
acidic
(
e.
g.,
ANC
decline
from
80
to
40

eq/
L).

1.2
The
SAMI
Region
The
SAMI
region
consists
of
three,
somewhat
linear,
southwest­
to­
northeast
trending
physiographic
provinces.
From
east
to
west,
they
are
the
Blue
Ridge
Mountains,
Valley
and
Ridge,
and
Appalachian
Plateau
(
Figure
1).
Streams
are
the
dominant
aquatic
resource
in
this
region.

In
Virginia,
the
Blue
Ridge
Mountains
make
up
a
narrow
complex
ridge
lying
between
the
Piedmont
and
the
Great
Valley
of
the
Valley
and
Ridge
Province
(
Figure
1).
In
the
southern
Blue
Ridge,
the
range
broadens
into
a
range
of
uniformly
dissected
mountains
with
elevations
ranging
from
500
m
to
2000
m
above
sea
level.
The
highest
mountains
in
the
eastern
United
States
are
found
in
the
southern
Blue
Ridge.
Bedrock
typically
is
composed
of
Cambrian
and
Precambrian
metamorphosed
sedimentary
and
complex
gneissic
and
plutonic
rocks.

The
Valley
and
Ridge
Province
(
Figure
1)
is
composed
of
a
series
of
folded
Paleozoic
sedimentary
rock
layers,
resulting
in
a
sequence
of
valleys
separated
by
narrow,
parallel,
linear
ridges.
The
ridges
typically
are
composed
of
more
resistant
sandstones,
whereas
limestone
and
shale
formations
are
common
in
the
valley
bottoms.
The
proportion
of
ridge
area
to
valley
area
is
quite
variable
throughout
the
province.
The
streams
often
show
a
trellis
drainage
pattern
due
to
the
valley
and
ridge
structure.

In
West
Virginia,
the
Appalachian
Plateau
Province
(
Figure
1)
is
separated
from
the
Valley
and
Ridge
Province
to
the
east
by
a
steep
outfacing
escarpment
(
e.
g.,
Allegheny
Front).
The
southern
part
of
the
Appalachian
Plateau
is
often
referred
to
as
the
Cumberland
Plateau.
The
rocks
in
the
Appalachian
Plateau
Province
are
younger
than
those
in
the
other
provinces,
and
clastic
conglomerates,
sandstones
and
shales
predominate.
The
mountains
of
the
Appalachian
Plateau
are
the
result
of
severe
plateau
dissection.
4
2.
FACTORS
CONTROLLING
AQUATIC
RESPONSE
TO
ACIDIC
DEPOSITION
2.1
Overview
Watershed
soils/
geology
and
hydrology
are
the
major
factors
controlling
streamwater
ANC
response
to
acidic
deposition.
In
fact,
on
a
regional
scale,
there
is
no
relationship
between
deposition
and
surface
water
ANC
(
L.
Baker
et
al.,
1990).
Two
watersheds
a
mile
apart
can
have
vastly
different
responses
to
acidic
deposition;
one
can
be
acidic,
while
the
other
is
so
well
buffered
that
deposition
poses
no
problem.
Quantifying
the
effects
on
any
one
stream
will
tell
very
little
about
the
effects
across
the
region.
Thus,
a
study
of
the
effects
of
acidic
deposition
should
be
related
to
the
population
of
streams
in
the
study
region.

The
NAPAP
SOS/
T
Processes
Report
summarizes
the
major
processes
influencing
the
response
of
surface
water
acid­
base
chemistry
to
acidic
deposition.
Figure
2
presents
a
summary
figure
from
that
report
(
Turner
et
al.,
1990).
For
the
SAMI
region,
the
two
most
important
processes
are
base
cation
neutralization
reactions
(
cation
exchange,
geologic
weathering)
and
anion
retention/
mobility
(
Elwood,
1991).
The
synthesis
statement
of
Turner
et
al.
(
1990)
provides
a
summary
of
the
effects
of
acidic
deposition
on
surface
waters:

"
The
magnitude
of
change
in
water
chemistry
parameters
in
response
to
acidic
deposition
and
changes
in
watershed
drainage
chemistry
may
range
from
an
equivalent
increase
in
base
cation
concentrations
to
a
reduction
of
50
or
more

eq/
L
ANC,
to
a
shift
from
a
low
ANC
or
acidic,
organic­
dominated
system,
or
to
a
sulfate­
dominated
system
with
little
change
in
ANC
or
pH.
The
first
response
is
probably
most
common.
In
the
latter
two
cases,
the
net
effect
of
atmospheric
deposition
of
S
on
lake
and
stream
chemistry
is
a
shift
toward
systems
that
are
dominated
by
mineral
acidity
and
that
have
higher
concentrations
of
inorganic
aluminum
which
is
toxic
to
aquatic
organisms."

In
the
SAMI
region,
the
major
control
that
determines
the
ability
of
a
watershed
to
buffer
its
surface
waters
against
the
influence
of
acidic
deposition
is
its
base
cation
mobilization
ability.
Watersheds
that
have
soils
with
high
base
saturation/
cation
exchange
capacity
and/
or
highly
weatherable
geologic
material
have
streams
and
lakes
virtually
unaffected
by
acidic
deposition.
The
abundance
of
available
base
cations
in
these
systems
balances
the
added
sulfate/
nitrate
from
deposition,
preventing
any
depression
of
ANC
or
pH.
Systems
that
can't
balance
the
added
sulfate/
nitrate
from
deposition
with
base
cations
have
decreased
ANC/
pH
and/
or
increased
inorganic
aluminum.
In
the
SAMI
region,
the
most
important
factors
controlling
the
supply
of
base
cations
are
the
watershed
soil
and
the
bedrock
type.

2.2
Acidic
Deposition
2.2.1
Current
Pattern
Relative
to
"
pristine"
deposition,
acidic
deposition
contains
elevated
concentrations
of
sulfate,
nitrate,
acidity,
base
cations,
and
ammonia.
There
are
two
types
of
deposition:
wet
(
e.
g.,
rain,
snow,
sleet,
hail)
and
dry
(
e.
g.,
aerosols,
gases,
particulates).
The
amount
and
content
of
wet
deposition
are
much
better
known
than
the
amount
and
content
of
dry
deposition.
Deposition
throughout
the
SAMI
area
would
be
considered
acidic
(
Figure
3).
In
the
Southern
Appalachians,
the
amount
of
5
Figure
2.
Conceptual
diagram
of
(
a)
major
processes
and
(
b)
hydrologic
flowpaths
that
control
surface
water
acid­
base
chemistry.
Taken
from
Figure
10­
41
in
Turner
et
al.,
1990.
6
7
acidic
deposition
decreases
from
north
to
south.
For
example,
if
we
use
the
volume
weighted
1995
wet
deposition
pH
as
an
indicator
of
acidic
deposition,
the
deposition
in
northern
West
Virginia
(
pH=
4.3)
has
twice
the
acidity
(
pH
is
log
scale)
of
that
in
northern
Alabama
(
pH=
4.6).
Similar
patterns
occur
in
the
deposition
of
sulfate
and
nitrate,
as
well
(
NADP/
NTN,
1996).

On
a
national
scale,
there
is
a
close
relationship
between
sulfate
deposition
and
sulfate
concentration
in
lakes
and
streams
(
L.
Baker
et
al.,
1990).
The
same
general
pattern
is
evident,
although
less
clear,
for
the
relationship
between
nitrate/
ammonia
deposition
and
nitrate
concentration
in
streams
of
the
eastern
United
States
(
Kaufmann
et
al.,
1991).
Lakes
and
streams
in
the
SAMI
region
are
an
exception
to
the
national
deposition/
surface
water
sulfate
relationship.
Sulfate
concentrations
in
these
southeastern
lakes
and
streams
are
much
lower
than
would
be
expected,
given
their
sulfate
deposition
levels,
due
to
substantial
retention
of
sulfate
by
adsorption
in
the
soils
of
their
watershed.

2.2.2
Historical
Evidence
for
Acidification
It
would
be
very
valuable
if
the
current
status
of
acid
sensitive
streams
in
the
region
covered
by
SAMI
could
be
placed
in
a
historical
contextCthat
is,
if
we
could
know
the
acid­
base
status
of
streams
in
the
region
in
pre­
industrial
times.
No
historical
data
are
available
for
streams
in
this
region,
however,
and
the
kinds
of
retrospective
techniques
(
e.
g.,
paleolimnology)
that
yield
historical
records
in
other
regions
(
Sullivan
et
al.,
1992)
are
not
applicable
to
streams.
Historical
data
from
the
Catskill
Mountains,
a
northern
extension
of
the
same
physiographic
province
that
makes
up
much
of
the
SAMI
region,
suggest
that
large
changes
in
streamwater
chemistry
have
occurred
(
Stoddard,
1991).
While
small
streams
in
the
Catskills
show
signs
of
acidification
over
the
past
50
years,
many
large
streams
have
actually
increased
in
ANC
and
base
cation
concentrations,
presumably
due
to
watershed
disturbances.

In
the
absence
of
historical
information
on
stream
chemistry
and
biology
for
the
SAMI
region,
we
are
forced
to
use
a
historical
context
based
on
deposition
and
estimated
deposition.
Husar
et
al.
(
1991)
report
that
the
northern
and
southern
parts
of
the
SAMI
region
probably
have
experienced
differences
in
historical
sulfur
deposition.
In
West
Virginia
and
Virginia,
S
deposition
rates
rose
from
a
background
near
4
kg/
ha/
yr
in
the
late
1800s
to
a
peak
near
30
kg/
ha/
yr
around
1970.
Since
that
time,
S
deposition
rates
have
declined,
although
they
still
remain
high,
relative
to
background
rates
(
ca.
25
kg/
ha/
yr).

In
the
southern
Blue
Ridge,
S
deposition
rates
were
not
significantly
different
from
background
rates
prior
to
the
1950s.
Husar
et
al.
(
1991)
also
report
similar
information
for
Florida,
and
it
seems
likely
that
the
region
from
the
southern
Blue
Ridge
southward
has
undergone
the
same
history.
Beginning
in
the
1960s,
S
deposition
increased
significantly
in
this
region,
to
near
20
kg/
ha/
yr,
and
there
was
no
evidence
of
a
decline
in
deposition
through
1984,
the
last
year
of
data
reported
by
Husar
et
al.
(
1991).

Similar
historical
estimates
for
nitrogen
deposition
are
not
available,
because
of
the
difficulty
of
translating
nitrogen
emissions
rates
into
reliable
estimates
of
deposition
(
Husar,
1986).
Husar
et
al.
(
1991)
do
document
the
changes
in
N
emissions
that
have
occurred
in
the
area
upwind
of
the
southeastern
United
States.
Nitrogen
emissions
began
rising
ca.
1900,
with
a
significant
part
of
the
increase
occurring
after
1940.
Current
rates
of
emission
(
ca.
7
million
tons
of
NO
2
per
year)
in
the
Southeast
are
almost
twice
those
in
the
northeastern
United
States.

2.3
Hydrologic
Flow
Path
8
Watersheds
in
which
the
greatest
proportion
of
the
flow
is
through
shallow,
more
acidic
soils
have
surface
waters
with
lower
ANC
than
watersheds
in
which
a
large
proportion
of
the
flow
is
through
deeper,
more
weatherable
materials.
The
longer
the
flowpath
the
more
intimate
the
contact
with
weathering
soil
and
rock
and
the
longer
the
time
available
to
acquire
solutes
through
biogeochemical
reactions
(
Turner
et
al.,
1990).

2.4
Anion
Mobility/
Retention
Watershed
soils
typically
contain
acid
cations
(
H+,
Aln+)
in
abundance.
The
mobile
anion
hypothesis
explains
how
acid
anions
(
bicarbonate,
sulfate,
nitrate,
organics)
are
needed
to
transport
acid
cations
from
soils
into
lakes
and
streams.
Thus,
cation
leaching
in
soils
is
controlled
by
the
availability
of
mobile
anions.
In
systems
affected
by
acidic
deposition,
elevated
levels
of
SO
4
2

and
NO
3

leach
either
equivalent
amounts
of
base
cations
(
medium
to
high
base
saturation
soils)
or
acid
cations
(
low
base
saturation
soils).
Thus,
the
mobility
of
the
SO
4
2

and
NO
3

through
the
watersheds
is
an
important
factor
controlling
the
response
of
surface
waters
to
acidic
deposition
(
Turner
et
al.,
1990).
If
all
the
SO
4
2

and
NO
3

derived
from
deposition
were
retained
within
the
watershed,
there
would
be
little
surface
water
response
to
acidic
deposition.

2.4.1
Sulfur
Of
the
two
major
anions
in
deposition,
SO
4
2

received
a
major
share
of
the
attention
during
the
NAPAP
process.
In
many
parts
of
the
United
States,
there
is
little
retention
of
SO
4
2

within
the
watershed.
Observed
lake/
stream
SO
4
2

concentrations
are
roughly
the
same
as
calculated
surface
water
SO
4
2

concentrations
assuming
evapoconcentration
of
SO
4
2

in
deposition.
These
surface
waters
are
considered
to
be
in
"
steady­
state"
with
SO
4
2

in
deposition.
A
notable
exception
to
this
pattern
is
represented
by
the
lakes
and
streams
in
the
southeastern
United
States,
where
most
of
the
incoming
SO
4
2

is
retained
in
the
watershed
and
does
not
reach
the
lake
or
stream
(
L.
Baker
et
al.,
1990).
The
H+
and
Aln+
in
soils
thus
does
not
reach
the
lake
or
stream
either.

Soils
adsorb
SO
4
2

according
to
a
concentration
dependent
function.
They
continue
to
adsorb
SO
4
2

until
the
concentration
of
adsorbed
SO
4
2

reaches
equilibrium
with
SO
4
2

in
the
soil
solution.
Different
soils
have
different
equilibrium
points
and
SO
4
2

adsorption
capacities.
The
adsorption
capacity
is
thought
to
be
directly
related
to
the
amount
of
Fe/
Al
oxide
in
soil
and
inversely
related
to
soil
pH
and
organic
matter
content.
The
higher
the
SO
4
2

adsorption
capacity,
the
longer
it
will
take
for
atmospherically
derived
SO
4
2

to
break
through
the
watershed
and
enter
surface
water.
In
the
Southern
Appalachians,
much
of
the
SO
4
2

is
still
being
adsorbed
in
the
watershed
and
the
response
to
atmospheric
deposition
is
considered
to
be
"
delayed."
Watersheds
currently
in
SO
4
2

steady
state
either
never
had
much
SO
4
2

adsorption
capacity
or
the
SO
4
2

adsorption
sites
have
been
filled
by
a
longer
history
of
deposition,
such
that
their
response
to
SO
4
2

deposition
is
considered
"
direct."
A
big
remaining
question
is
the
reversibility
of
SO
4
2

adsorption.
It
is
not
well
known
what
happens
when
SO
4
2

deposition
declines.
It
is
likely
that
SO
4
2

adsorption
will
show
some
degree
of
irreversibility;
SO
4
2

desorption
will
not
follow
the
same
path
as
adsorption
(
Turner
et
al.,
1990).

2.4.2
Nitrogen
In
the
past
several
years,
our
scientific
understanding
of
the
impacts
of
long­
term
N
deposition
to
forested
watersheds
has
undergone
significant
improvement.
Although
S
deposition
has
been
the
focus
of
most
research
on
acid
deposition
effects
in
the
past
decades,
we
now
recognize
that
N
deposition
is
also
a
9
threat
to
the
integrity
of
aquatic
and
terrestrial
systems,
especially
in
areas
that
historically
have
received
elevated
levels
of
N
deposition
(
Aber
et
al.,
1989;
Driscoll
and
Schaefer,
1989;
Murdoch
and
Stoddard,
1992).
The
danger
is
that
relatively
undisturbed
watersheds
aside
from
elevated
deposition,
may
through
time
become
N
saturated.
Nitrogen
saturation
has
been
variously
defined,
but
central
to
all
definitions
is
the
concept
that
the
supply
of
nitrogenous
compounds
from
the
atmosphere
exceeds
the
demand
for
these
compounds
on
the
part
of
watershed
plants
and
microbes
(
Skeffington
and
Wilson,
1988;
Aber
et
al.,
1989).
Under
conditions
of
N
saturation,
forested
watersheds
that
previously
retained
nearly
all
N
inputs,
due
to
a
high
demand
for
N
by
plants
and
microbes,
begin
to
undergo
higher
loss
rates
of
N.
These
losses
may
be
in
the
form
of
leaching
to
surface
waters
or
to
the
atmosphere
through
denitrification.

The
key
processes
of
the
N
cycle
that
affect
acidification
of
soils
and
surface
waters
are
assimilation,
mineralization,
nitrification,
and
denitrification.
Nitrogen
assimilation
is
the
uptake
and
metabolic
use
of
N
by
plants
and
soil
microbes.
Because
N
is
the
most
commonly
limiting
nutrient
in
forest
ecosystems
in
North
America
(
Cole
and
Rapp,
1981;
Vitousek
and
Howarth,
1991),
assimilation
plays
a
key
role
in
the
development
of
N
saturation.
The
form
of
N
used
by
terrestrial
ecosystems
strongly
affects
the
acidifying
potential
of
N
deposition.
Ammonium
uptake
is
an
acidifying
process
(
i.
e.,
uptake
of
NH
4
+

releases
one
mole
of
H+
per
mole
of
N
assimilated),
whereas
biological
uptake
of
NO
3

is
an
alkalinizing
process
(
i.
e.,
uptake
of
NO
3

consumes
one
mole
of
H+
per
mole
of
N
assimilated).

Mineralization
is
the
bacterial
decomposition
of
organic
matter,
releasing
NH
4
+
that
can
subsequently
be
nitrified
to
NO
3

.
Mineralization
is
an
important
process
in
watersheds
as
it
recycles
N
that
would
otherwise
be
tied
up
in
soil
organic
matter
following
the
death
of
plants,
or
as
leaf
litter.
Nitrification
is
the
oxidation
of
NH
4
+
to
NO
3

;
it
is
mediated
by
bacteria
and
fungi
in
both
the
terrestrial
and
aquatic
parts
of
watersheds.
It
is
an
important
process
in
controlling
the
form
of
N
released
to
surface
waters
by
watersheds,
as
well
as
in
controlling
the
acid­
base
status
of
surface
waters.
Nitrification
is
a
strongly
acidifying
process,
producing
two
moles
of
H+
for
each
mole
of
nitrogen
(
NH
4
+)
nitrified.
Because
nitrification
in
forest
soils
transforms
NH
4
+

into
NO
3

,
the
acidifying
potential
of
deposition
(
attributable
to
N)
is
often
defined
as
the
sum
of
NH
4
+
and
NO
3

,
assuming
that
all
N
will
leave
the
watershed
as
NO
3

(
Hauhs
et
al.,
1989).

Denitrification
is
the
biological
reduction
of
NO
3

to
produce
gaseous
forms
of
reduced
nitrogen
(
N
2
,
NO,
or
N
2
O).
Denitrification
is
an
anaerobic
process
(
i.
e.,
it
occurs
only
in
environments
where
oxygen
is
absent),
whose
end
product
is
lost
to
the
atmosphere.
It
is
always
an
alkalinizing
process,
consuming
one
mole
of
H+
for
every
mole
of
N
denitrified.
In
terrestrial
ecosystems,
denitrification
occurs
in
boggy,
poorly
drained
soils,
or
in
anaerobic
microsites.
It
has
traditionally
been
considered
a
relatively
unimportant
process
outside
of
wetlands
(
Post
et
al.,
1985),
although
it
may
be
locally
important
after
such
events
as
spring
snowmelt
and
heavy
rain
storms,
when
soil
oxygen
tension
is
reduced
(
Melillo
et
al.,
1983;
Groffman
et
al.,
1993).
On
a
watershed
scale,
denitrification
is
not
likely
to
be
an
important
sink
for
N
(
Bowden,
1986;
Bowden
and
Bormann,
1986),
because
it
is
limited
by
the
availability
of
anaerobic
sites
(
Klemedtsson
and
Svensson,
1988).

A
number
of
factors
may
predispose
watersheds
to
become
N
saturated,
including
elevated
N
deposition,
stand
age,
and
high
soil
N
pools
(
Stoddard,
1994).
High
rates
of
N
deposition
play
a
clear
role,
as
the
ability
of
10
forest
biomass
to
accumulate
N
must
be
finite.
At
very
high,
long­
term
rates
of
N
deposition,
the
ability
of
forests
and
soils
to
accumulate
N
is
exceeded,
and
the
only
remaining
pathway
for
loss
of
N
(
other
than
runoff)
is
denitrification
High
rates
of
N
deposition
may
favor
increased
rates
of
denitrification,
but
many
watersheds
lack
the
conditions
necessary
for
substantial
denitrification
(
e.
g.,
low
oxygen
tension,
high
soil
moisture,
and
temperature).

Another
important
factor
in
N
loss
from
watersheds
is
the
age
of
the
forest
stands.
A
loss
in
the
ability
to
retain
N
is
a
natural
outcome
of
forest
maturation,
as
both
older
trees
and
those
that
occur
later
in
a
successional
sequence
grow
more
slowly
and
therefore
exhibit
lower
N
demand
(
Vitousek
and
Reiners,
1975).
Uptake
rates
of
N
into
vegetation
are
generally
maximal
around
the
time
of
canopy
closure
for
conifers,
and
somewhat
later
(
and
at
higher
rates)
in
deciduous
forests,
due
to
the
annual
replacement
of
canopy
foliage
in
these
ecosystems
(
Turner
et
al.,
1990).
Finally,
soil
N
status
may
also
affect
N
loss
rates;
where
large
soil
N
pools
exist,
they
imply
that
soil
microbial
processes
that
are
ordinarily
limited
by
the
availability
of
N
are
instead
limited
by
some
other
factor
(
e.
g.,
availability
of
labile
organic
carbon,
or
another
inorganic
nutrient),
and
contribute
to
the
likelihood
that
watersheds
will
leach
NO
3

(
Johnson,
1992;
Joslin
et
al.,
1992).
To
be
N
saturated,
both
the
vegetational
and
soil
microbial
N
demands
of
a
watershed
must
be
fulfilled;
the
existence
of
large
soil
N
pools
suggests
that
the
second
of
these
requirements
may
be
easily
met.

2.5
In­
lake
and
In­
stream
Processes
A
number
of
processes
operating
within
lakes
and
streams
can
modify
their
acid­
base
chemistry,
including
SO
4
2

/
NO
3

retention
mechanisms
and
base
cation
production.
These
processes
are
most
important,
however,
in
waters
with
long
residence
times
(
Turner
et
al.,
1990).
Thus,
they
would
be
expected
to
be
of
minor
importance
in
streams
and
reservoirs
in
the
SAMI
region.

2.6
Other
Acidity
Sources
In
addition
to
acidic
deposition,
the
two
other
major
sources
of
acid
anions
for
lakes
and
streams
are
natural
organic
acids
and
oxidation
of
watershed
sulfur
compounds.
Lakes
and
streams
naturally
acidic
due
to
organic
acids
are
found
in
areas
where
bedrock
and
soils
are
resistant
to
weathering
and
where
there
is
a
large
buildup
of
organic
matter.
These
naturally
acidic
waters
are
typically
colored
from
high
dissolved
organic
carbon
(
DOC)
concentrations
and
are
often
found
in
places
of
low­
relief
terrain
and
poor
drainage
(
Turner
et
al.,
1990).
Lakes
and
streams
may
also
be
acidic
due
to
oxidation
of
naturally
occurring
sulfur
compounds
within
their
watersheds.
This
process
is
similar
to
acidic
deposition
in
that
a
sulfuric
acid
solution
is
carried
into
the
receiving
water,
but
the
sulfur
source
is
internal
to
the
watershed
rather
than
deposited
on
it.
The
source
of
the
sulfur
is
usually
bedrock
that
contains
sulfide
(
e.
g.,
Anakeesta
Formation
in
the
Great
Smoky
Mountains
or
pyrite
(
FeS
2)
associated
with
coal
seams
in
the
Appalachian
Plateau).
Significant
SO
4
2

mobilization
from
internal
sources,
however,
usually
requires
some
kind
of
watershed
disturbance
(
mining
or
road
cut)
to
expose
bedrock
materials
to
water/
air.
11
3.
CURRENT
STATUS
OF
AQUATIC
RESOURCES
3.1
Surface
Water
Acid­
Base
Chemistry
3.1.1.
Regional
Status
As
part
of
NAPAP,
EPA
conducted
a
National
Surface
Water
Survey
(
NSWS)
of
lakes
and
streams
in
acid­
sensitive
areas
of
the
United
States.
In
the
SAMI
area,
lakes
were
sampled
in
the
Southern
Blue
Ridge
subregion
during
the
Eastern
Lake
Survey
(
ELS).
The
National
Stream
Survey
(
NSS)
sampled
streams
throughout
the
SAMI
region,
except
for
the
western
parts
of
the
Appalachian
Plateau
(
Figure
4).
The
western
part
of
the
plateau
was
not
sampled
in
the
NSS
because
existing
data
at
the
time
indicated
that
it
had
high
ANC
(>
400

eq/
L).

The
results
of
the
NSWS
reflect
chronic,
not
worst­
case
episodic,
acid­
base
conditions
(
see
Section
3.2).
Due
to
the
probability
basis
of
the
sample,
and
its
regional
extent,
the
NSWS
is
the
best
picture
we
have
about
the
regional
status
and
extent
of
chronic
acid­
base
chemistry
(
L.
Baker
et
al.,
1990).
For
this
report,
we
have
used
two
ANC
criteria
to
define
acid­
base
status:
acidic
(
ANC

0)
and
low
ANC
(

50

eq/
L).
In
the
SAMI
region,
acidic
streams
are
those
most
affected
by
acidic
deposition;
they
typically
have
pH
in
the
4'
s
and
low
5'
s
and
they
have
elevated
levels
of
inorganic
monomeric
aluminum
that
is
toxic
to
fish.
Surface
waters
with
ANC
between
0
and
50

eq/
L
are
not
chronically
acidic,
but
many
of
them
become
acidic
during
storm
events
or
snowmelt
(
episodes).
During
these
episodes,
they
experience
the
low
pH
and
elevated
aluminum
levels
of
chronically
acidic
systems
(
see
section
3.2).
Acidic
deposition
does
not
usually
cause
substantial
environmental
impacts
in
surface
waters
with
ANC
>
50

eq/
L.

3.1.1.1
Lakes.
In
the
Southern
Blue
Ridge
subregion,
the
ELS
sampled
lakes
in
both
the
Piedmont
and
Blue
Ridge
physiographic
provinces
between
southern
Virginia
and
northern
Georgia
(
L.
Baker
et
al.,
1990).
The
following
analysis
(
Table
1)
reflects
only
those
lakes
in
the
Blue
Ridge
part
of
the
SAMI
region.
The
ELS
lake
population
was
determined
by
listing
all
lakes
(
with
surface
area
>
4
ha)
in
the
study
region
present
on
1:
250,000­
scale
USGS
topographic
maps.
The
ELS
sampled
45
lakes
in
the
Blue
Ridge
during
fall
1984,
after
overturn;
90%
of
the
estimated
total
population
of
71
lakes
were
reservoirs.
It
was
estimated
that
5%
of
the
lakes
(
3
lakes)
had
ANC

50

eq/
L.
No
sample
lakes
were
acidic
and
1%
had
pH
<
6.
The
lakes
had
very
low
SO
4
2

,
NO
3

,
and
DOC
concentrations
(
Table
2).
Given
that
most
of
these
lakes
are
reservoirs,
it
appears
that
they
have
large
enough
drainage
areas
to
buffer
inputs
from
acidic
deposition.
For
SAMI,
lakes
can
thus
be
considered
of
minor
interest
with
respect
to
acidic
deposition
impacts
due
to:

$
The
low
percentage
of
sensitive
systems
$
The
low
lake
density
in
the
study
area
(
for
comparison
there
are
>
2000
lakes
in
the
Adirondack
Mountains
of
New
York)

$
Lack
of
a
significant
lake
resource
in
the
Class
I
areas
Lakes
are
not
discussed
further
in
any
of
the
remaining
sections
of
this
report.

3.1.1.2
Streams.
The
NSS
sampled
the
stream
network
in
the
SAMI
area
present
on
1:
250,000­
scale
USGS
topographic
maps
(
those
with
watershed
areas
<
155
km2).
The
NSS
used
a
randomized,
systematic
approach
to
selecting
stream
segment
sample
sites.
Chemistry
was
sampled
at
both
the
upstream
and
the
downstream
ends
of
each
segment
(
Kaufmann
et
al.,
1991).
In
the
SAMI
study
12
Figure
4.
Map
of
the
acid­
sensitive
part
of
the
Southern
Appalachians
sampled
in
the
National
Stream
Survey.
Areas
outside
the
NSS
study
boundary
were
expected
to
have
ANC
>
400

eq/
L.
13
Table
1.
Estimates
of
Regional
Acid­
base
Status
for
Streams
and
Lakes
in
the
Southern
Appalachians
from
National
Surface
Water
Survey
Data.
For
streams,
the
sampling
unit
was
stream
segments.
Streams
were
sampled
at
both
the
upstream
and
downstream
ends
of
each
segment.
The
sampling
unit
for
lakes
was
individual
lakes.

Region
Percent
ANC

0

eq/
L
Upper
C
Lower
End
Percent
ANC

50

eq/
L
Upper
C
Lower
End
Estimated
Population
Total
Upper
C
Lower
End
Blue
Ridge
Streams
0%
C
0%
12%
C
2%
4,850
C
4,350
Valley
and
Ridge
Streams
Ridges
Valleys
10%
C
0%

0%
C
0%
25%
C
2%

0%
C
0%
3,160
C
2,720
2,990
C
4,350
Appalachian
Plateau
Streams
5%
C
2%
31%
C
24%
8,940
C
8,780
All
Streams
4%
C
1%
20%
C
11%
19,900
C
20,200
Southern
Blue
Ridge
Lakes
0%
5%
71
region,
306
segment
ends
were
sampled,
statistically
representing
19,900
upstream
segment
ends
and
20,200
downstream
ends
(
or
62,000
km
of
stream).
Streams
were
sampled
in
the
spring
during
1985
and
1986.
There
was
a
strong
upstream/
downstream
gradient
in
the
NSS
data;
ANC
increased
in
the
downstream
direction
due
to
increasing
mobilization
of
base
cations.
For
example,
12%
of
the
stream
segments
in
the
Blue
Ridge
had
low
ANC
(

50

eq/
L)
at
the
upstream
ends
versus
only
2%
at
the
downstream
ends
(
Table
1).
Overall,
4%
of
the
stream
segments
were
acidic
at
the
upstream
ends
in
the
SAMI
part
of
the
NSS
study
area
(
1%
at
the
downstream
end).

Of
the
three
physiographic
provinces,
the
Appalachian
Plateau
had
the
highest
proportion
of
acidic
(
5%)
and
low
ANC
(
31%)
streams.
Streams
acidic
because
of
acid
mine
drainage
were
screened
out
of
the
NSS
population
and
are
not
included
in
any
of
the
estimates
presented
here
(
Herlihy
et
al.,
1990).
In
the
Valley
and
Ridge
province,
the
effects
of
acidic
deposition
are
restricted
to
the
ridges.
Low
ANC
streams
are
very
rare
or
absent
(
none
were
observed
in
the
NSS
sample)
in
the
valleys,
due
to
the
easily
weatherable
bedrock
(
e.
g.,
limestone).
Streams
on
the
ridges
are
more
sensitive
to
acidification
due
to
the
resistance
to
weathering
of
the
bedrock
geology.
Thus
ridges
have
less
ability
to
neutralize
acidic
deposition.
At
the
upstream
ends,
10%
of
the
ridge
streams
were
acidic
and
25%
had
low
ANC
(
Table
1).
The
ANC
increases
rapidly
as
streams
flow
down
off
the
ridges.
At
the
downstream
ends,
none
of
the
sample
segments
were
acidic
and
only
2%
had
low
ANC.

There
is
also
a
gradient
in
the
percentage
of
acidic
and
low
ANC
streams
from
the
northern
to
the
southern
parts
of
the
SAMI
region.
Acidic
and
low
ANC
stream
segments
are
more
common
in
Virginia/
West
Virginia
14
(
26%)
than
in
the
six
southern
SAMI
states
(
8%).
This
is
probably
due
to
a
combination
of
higher
S
and
N
deposition
and
lower
SO
4
2

retention
in
the
more
northern
watersheds.

The
NSS
data
from
the
SAMI
region
indicated
that
acidic
streams
and
streams
with
very
low
ANC
were
almost
all
located
in
small
(<
20
km2),
upland,
forested
catchments
in
areas
of
base­
poor,
weathering­
resistant,
bedrock
(
Herlihy
et
al.,
1993).
By
interpolating
between
segment
ends,
NSS
estimates
for
the
whole
SAMI
region
show
that
of
the
62,200
km
of
stream
length
on
1:
250,000­
scale
maps,
815
km
(
1%)
were
acidic
and
4,410
km
(
7%)
had
ANC

50

eq/
L.
Sulfate
and
NO
3

from
atmospheric
deposition
are
the
dominant
sources
of
acid
anions
in
these
acidic
NSS
Southern
Appalachian
streams.
As
a
group,
they
had
low
pH
(
median
=
4.7)
and
high
levels
of
inorganic
monomeric
aluminum
(
median
=
364

g/
L)
that
can
cause
damage
to
aquatic
biota
(
Table
2).

The
NSS
data
also
reflect
distinct
patterns
of
SO
4
2

retention
among
the
three
physiographic
provinces.
Streamwater
SO
4
2

concentrations
in
the
Appalachian
Plateau
were
near
steady
state
with
respect
to
SO
4
2

deposition
whereas
Blue
Ridge
streams
were
retaining
large
amounts
(>
50%)
of
incoming
SO
4
2

deposition.
Valley
and
Ridge
streams
showed
intermediate
levels
of
SO
4
2

retention
(
Herlihy
et
al.,
1993).

3.1.2
Relationship
of
Stream
ANC
to
Geology
Several
investigators
have
described
the
association
between
bedrock
geology
and
the
ANC
of
streams
in
the
northern
Blue
Ridge
(
Lynch
and
Dise,
1985;
Bricker
and
Rice,
1989).
Lynch
and
Dise
(
1985)
described
a
strong
correlation
(
r2
=
0.95)
between
bedrock
distribution
and
streamwater
ANC
in
Shenandoah
National
Park;
percentile
distributions
of
streamwater
ANC
are
distinctly
different
for
the
park's
five
major
bedrock
types.
In
theory,
the
distribution
of
bedrock
types
within
the
region
should
provide
a
basis
for
predicting
the
ANC
of
unsampled
streams
in
the
region,
as
well
as
for
explaining
the
ANC
of
sampled
streams.
In
practice,
however
the
success
of
this
approach
on
regional
scales
is
limited
by
the
generality
of
much
of
the
geologic
information
available
for
the
Southern
Appalachians.
Herlihy
et
al.
(
1993)
had
little
success
in
relating
geological
information
at
the
1:
250,000
map
scale
to
stream
ANC
in
data
from
the
1987
synoptic
component
of
the
Virginia
Trout
Stream
Sensitivity
Study
(
VTSSS).
This
map
scale
was
too
coarse
to
adequately
characterize
small
streams.
Thus,
a
geological
approach
to
predicting
ANC
works
well
in
areas
for
which
geologic
maps
are
available
at
high
levels
of
detail.
Just
recently,
a
detailed
geologic
lithofacies
map
for
the
Southern
Appalachians
has
been
developed
to
provide
regional­
scale
information
on
stream
ANC
(
Peper
et
al.,
1995).
It
may
prove
to
be
a
useful
tool
for
regionalizing
acidic
deposition
impacts,
but
the
predictive
power
of
this
new
map
against
observed
stream
ANC
needs
to
be
tested.

3.1.3
Detailed
Analysis
of
Class
I
Areas
A
major
focus
of
the
SAMI
effort
is
to
evaluate
air
pollution
effects
on
the
Class
I
wilderness
areas
in
the
Southern
Appalachians.
Stream
networks
from
both
1:
100,000­
scale
and
1:
24,000­
scale
USGS
topographic
maps
were
digitized
for
each
of
the
eight
smaller
Class
I
areas
(
see
Appendix
A).
Stream
networks
for
the
two
largest
Class
I
areas
(
two
national
parks)
and
the
region
were
taken
from
the
digitized
version
of
the
1:
100,000­
scale
blue­
line
network
(
from
EPA
River
Reach
File
B
Version
3).

In
the
entire
SAMI
study
region
(
Figure
1),
there
are
just
under
200,000
km
of
streams
at
the
1:
100,000
scale
(
Table
3).
Roughly
60%
of
this
length
is
in
the
acid­
sensitive
part
15
Table
2.
Median
Values
(
with
First
and
Third
Quartiles
in
Parentheses)
for
Major
Ion
Chemistry
in
Streams
in
Class
I
Wilderness
Areas
and
in
the
Entire
Southern
Appalachians.
Year(
s)
of
data
collection
and
number
of
observations
(
n)
are
given
below
the
wilderness
area
name;
data
sources
are
discussed
in
text.

Wilderness
Area
ANC
(

eq/
L)
pH
Sulfate
(

eq/
L)
Nitrate
(

eq/
L)
Chloride
(

eq/
L)
DOC
(
mg/
L)

Dolly
Sods
1994
(
n=
34)

18
(

53
­

3)
4.7
(
4.3­
5.1)
105
(
91­
115)
4
(
2­
6)
11
(
9­
12)
2.2
(
1.7­
3.1)

Otter
Creek
1994
(
n=
63)

28
(

82
­
11)
4.6
(
4.1­
6.0)
129
(
111­
153)
6
(
1­
14)
9
(
8­
10)
2.0
(
0.9­
3.1)

Shenandoah
NP
1981­
1982
(
n=
47)
82
(
21­
120)
6.7
(
6.0­
6.9)
85
(
66­
103)
7
(
3­
23)
28
(
25­
32)
­­­­­

James
River
Face
1991­
1994
(
n=
8)
25
(
22­
44)
6.3
(
6.1­
6.5)
68
(
54­
74)
0
(
0­
0)
19
(
18­
20)
­­­­­

Great
Smoky
Mt.
NP
1994­
1995
(
n=
337)
44
(
24­
64)
6.4
(
6.2­
6.6)
31
(
18­
46)
15
(
6­
29)
14
(
12­
16)
­­­­­

Joyce
Kilmer/
Slickrock
1992­
1994
(
n=
9)
70
(
53­
80)
­­­­­
­­­­­
7
(
6­
11)
­­­­­
­­­­­

Shining
Rock
1992­
1995
(
n=
9)
70
(
65­
80)
6.8
(
6.7­
7.0)
­­­­­
7
(
6­
7)
­­­­­
­­­­­

Cohutta
1992­
1994
(
n=
16)
41
(
26­
56)
6.5
(
6.2­
6.6)
35
(
25­
53)
14
(
9­
21)
24
(
21­
28)
1.8
(
1.4­
2.5)

Sipsey
1991­
1993
(
n=
30)
245
(
120­
699)
7.3
(
6.8­
7.6)
94
(
83­
106)
2
(
1­
3)
33
(
32­
34)
2.2
(
1.6­
2.7)

SAMI
Region
Streams
a
1986
NSS
(
N=
19,940)
172
(
65­
491)
7.1
(
6.5­
7.5)
135
(
62­
229)
16
(
4­
34)
36
(
18­
68)
1.0
(
0.7­
1.7)

Acidic
SAMI
Streams
a
1986
NSS
(
N=
730)

24
(

35
­

24)
4.7
(
4.5­
4.7)
142
(
117­
229)
0.3
(
0.2­
3.5)
16
(
12­
25)
1.4
(
1.0­
1.7)

S.
Blue
Ridge
Lakes
a
1984
ELS
(
N=
71)
152
(
87­
246)
6.8
(
6.7­
7.0)
29
(
23­
36)
1
(
0­
6)
25
(
18­
42)
1.0
(
1.2­
1.5)

a
Regional
estimate
for
SAMI
region
streams
is
calculated
using
National
Stream
Survey
(
NSS)
data
(
Figure
4)
for
the
upstream
segment
end
population
(
extrapolated
from
154
sample
streams).
The
Southern
Blue
Ridge
lake
estimate
is
extrapolated
from
45
lakes
sampled
in
the
Eastern
Lake
Survey
(
L.
Baker
et
al.,
1990).

­­­
Not
measured,
no
data
found.
16
Table
3.
Surface
Area
and
Stream
Length
at
1:
100,000
and
1:
24,000
Map
Scales
for
Streams
in
the
Southern
Appalachians
and
in
Class
I
Wilderness
Areas.

Region
Land
Area
(
km2)
1:
100,000
Scale
Stream
Length
(
km)
1:
24,000
Scale
Stream
Length
(
km)

SAMI
REGION
Blue
Ridge
44,950
40,200
­­­­­

Valley
and
Ridge
78,980
60,600
­­­­­

Appalachian
Plateaus
117,100
97,800
­­­­­

Total
SAMI
region
240,700
198,600
­­­­­

Acid­
sensitive
part
(
NSS
boundary)
142,700
117,800
­­­­­

CLASS
I
AREAS
Dolly
Sods
43
25
29
Otter
Creek
81
57
85
Shenandoah
National
Park
788
259
­­­­­

James
River
Face
35
30
44
Great
Smoky
Mountain
National
Park
2,300
2,035
­­­­­

Joyce
Kilmer/
Slickrock
69
73
159
Linville
Gorge
44
21
38
Shining
Rock
75
44
102
Cohutta
121
140
208
Sipsey
52
70
108
CLASS
I
TOTAL
3,608
2,754
­­­­­

­­­­­
=
not
determined.
17
of
the
Southern
Appalachians
as
defined
by
the
NSS
study
boundary
(
Table
3;
Figure
4).
At
the
same
map
scale,
there
are
2,754
km
of
streams
in
the
10
Class
I
areas
(
1.4%
of
the
total
in
the
SAMI
region);
74%
of
the
total
Class
I
stream
length
is
in
the
Great
Smoky
Mountains
National
Park
and
9%
is
in
Shenandoah
National
Park.
The
eight
smaller
Class
I
areas
have
a
total
length
of
466
km
at
the
1:
100,000
map
scale
and
773
km
at
the
1:
24,000
map
scale
(
Table
3).
Based
on
geology,
physiography,
and
stream
chemistry,
the
10
Class
I
areas
in
the
SAMI
region
can
be
aggregated
into
four
groups
for
assessing
the
aquatic
effects
of
acidic
deposition:

1.
West
Virginia
Plateau:
Dolly
Sods
and
Otter
Creek
wilderness
areas
2.
Northern
Blue
Ridge:
James
River
Face
wilderness
area
and
Shenandoah
National
Park
3.
Southern
Blue
Ridge:
Great
Smoky
Mountains
National
Park
and
Joyce
Kilmer/
Slickrock,
Linville
Gorge,
Shining
Rock,
and
Cohutta
wilderness
areas
4.
Alabama
Plateau:
Sipsey
wilderness
area
For
each
of
these
wilderness
areas,
we
gathered
as
much
information
on
stream
chemistry
as
possible.
Summary
statistics
of
available
chemical
data
can
be
found
in
Table
2.
We
also
estimated
SO
4
2

and
NO
3

steadystate
concentrations
for
each
Class
I
area
by
identifying
the
3
B
5
NSS
sample
streams
closest
to
each
area.
As
part
of
the
Direct/
Delayed
Response
Project
(
DDRP;
Church
et
al.,
1989;
1992),
the
steady­
state
concentrations
of
SO
4
2

and
NO
3

in
streamwater
for
each
NSS
site
were
estimated
by
interpolating
available
precipitation,
runoff,
and
S
and
N
deposition
isopleths.
Sulfate
steady­
state
estimates
incorporated
estimates
of
dry
deposition
The
N
steady­
state
estimates
were
based
on
wet
deposition
only
and
are
almost
certainly
underestimated
(
Table
4).
We
calculated
percent
S
and
N
retention
for
each
Class
I
area
using
the
observed
median
streamwater
concentration
for
each
area
and
the
median
of
the
3
B
5
steady­
state
estimates
from
the
closest
NSS
sites.
Although
these
are
rather
rough
estimates,
they
give
a
reasonably
good
picture
of
the
differences
in
retention
capabilities
of
the
Class
I
areas
and
the
potential
for
delayed
response
to
acidic
deposition.

Available
stream
chemistry
data
were
also
overlain
on
the
digitized
1:
100,000­
scale
stream
network
maps.
The
entire
100,000­
scale
map
stream
length
was
then
ordered
into
the
following
classes,
similar
to
the
ones
developed
by
Herlihy
et
al.
(
1991)
for
NSS
data:

1.
ANC

0
2.
0
<
ANC

50

eq/
L
3.
ANC
>
50

eq/
L
4.
Organic
dominated
(
DOC
>
5
mg/
L)

5.
No
data
Stream
length
between
two
sample
points
with
the
same
class
was
assumed
to
reflect
that
class.
If
the
class
changed
between
sample
points,
the
stream
length
was
split
in
half
to
reflect
each
of
the
sample
points.
Stream
length
between
a
sample
point
and
a
confluence
or
upstream
termination
was
assumed
to
have
the
class
of
the
sample
point.
The
stream
length
in
each
of
the
five
classes
was
then
calculated
from
the
digitized
stream
map
(
Table
5).
A
DOC
level
of
5
mg/
L
was
used
to
distinguish
streams
whose
acid
anion
source
is
primarily
organic,
rather
than
inorganic
anions
from
deposition.
None
of
the
available
data
from
the
Class
I
areas
had
streamwater
SO
4
2

concentrations
indicative
of
a
predominantly
watershed
source
(
e.
g.,
acid
mine
drainage).
Thus,
we
did
not
include
a
18
Table
4.
Observed
and
Estimated
Steady­
state
Sulfur
and
Nitrogen
Concentrations
for
Streams
in
Class
I
Wilderness
Areas
of
the
Southern
Appalachians.
Observed
data
are
the
median
concentrations
from
Table
2.
Steady­
state
values
are
based
on
precipitation,
runoff,
and
deposition
isopleth
maps
that
were
interpolated
for
each
NSS
site
by
the
DDRP.
Value
in
the
table
is
the
median
of
the
3
 
5
NSS
sample
streams
closest
to
each
wilderness
area.
Percent
retention
is
calculated
as
(
steady­
state

observed)/
steadystate

100.

Wilderness
Area
Observed
Sulfate
(

eq/
L)
Steady­
state
Sulfate
(

eq/
L)
Percent
Sulfur
Retention
Observed
NO3+
NH4
(

eq/
L)
Steady­
state
NO3+
NH4
(

eq/
L)
Percent
Nitrogen
Retention
Dolly
Sods
105
177
41%
4
61
93%

Otter
Creek
129
177
27%
6
61
90%

Shenandoah
NP
85
196
57%
7
93
93%

James
River
Face
68
213
68%
0
99
100%

Great
Smoky
Mt.
NP
31
104
70%
15
43
65%

Joyce
Kilmer/
Slickrock
­­­
112
­­­
7
46
85%

Shining
Rock
­­­
86
­­­
7
41
83%

Cohutta
35
103
66%
14
40
65%

Sipsey
94
95
1%
2
60
97%

­­­
Not
measured,
no
data
found.
19
Table
5.
Length
of
Stream
(
km)
in
Various
ANC
Classes
in
Class
I
Wilderness
Areas
in
the
Southern
Appalachians.
This
analysis
is
based
on
the
1:
100,000­
scale
mapped
stream
length.
Year(
s)
of
data
collection
and
number
of
unique
stream
sample
points
(
n)
are
given
below
the
wilderness
area
name;
data
sources
are
discussed
in
the
text.
The
percentages
below
the
lengths
are
based
on
percent
of
the
total
stream
length
with
measured
stream
chemistry
data.

Wilderness
Area
ANC

0
(

eq/
L)
ANC
0­
50
(

eq/
L)
ANC
>
50
(

eq/
L)
Organic
Dominated
No
Data
Dolly
Sods
1994
(
n=
28)
20.6
(
82%)
4.6
(
18%)
0
0
0
Otter
Creek
1994
(
n=
45)
28.5
(
53%
)
8.9
(
17%)
9.4
(
18%)
6.7
(
13%
)
3.9
Shenandoah
NP
1981­
1982
(
n=
47)
­­­

(
6%)
­­­

(
19%)
­­­

(
75%)
­­­
­­­

James
River
Face
1991­
1994
(
n=
8)
0
20.5
(
92%)
1.7
(
8%)
0
7.6
Great
Smoky
Mt.
NP
1994­
1995
(
n=
337)
32
(
4%)
336
(
46%)
368
(
50%)
0
1,300
Joyce
Kilmer/
Slickrock
1992­
1994
(
n=
9)
0
8
(
36%)
14
(
64%)
0
51
Shining
Rock
1992­
1995
(
n=
9)
0
0
10
(
100%)
0
34
Cohutta
1992­
1994
(
n=
15)
0
102
(
96%)
4.8
(
4%)
0
33.2
Sipsey
1991­
1993
(
n=
10)
0
13.6
(
37%)
23.3
(
63%)
0
33.1
­­­
=
Length
estimates
not
made;
percentage
estimates
based
on
sample
percentages.
20
watershed
SO
4
2

source
class
in
this
classification
system.
There
were
a
few
sites
in
the
Great
Smoky
Mountains
National
Park,
however
that
had
SO
4
2

concentrations
somewhat
higher
than
typical,
indicative
of
a
watershed
SO
4
2

contribution
from
the
weathering
of
bedrock
containing
sulfide
(
Anakeesta
formation
see
section
3.1.3.3).

3.1.3.1.
West
Virginia
Appalachian
Plateau.
The
Dolly
Sods
and
Otter
Creek
Class
I
wilderness
areas
are
located
about
25
km
apart
in
the
Appalachian
Plateau
of
northeastern
West
Virginia
(
Figure
1).
Both
areas
are
drained
by
a
third­
order
(
1:
100,000­
scale)
stream.
The
Dolly
Sods
wilderness
(
43
km2)
is
drained
by
Red
Creek
but
the
headwaters
of
the
stream
lie
outside
the
wilderness
area.
The
area
is
mainly
a
plateau
top
dissected
by
Red
Creek
and
its
tributaries.
Elevations
in
the
Dolly
Sods
range
from
2600
to
3900
ft.
The
Otter
Creek
wilderness
(
81
km2)
contains
almost
the
entire
Otter
Creek
watershed.
Elevations
range
from
1800
to
3900
ft.
Both
areas
have
similar
bedrock
geology
(
West
Virginia,
1968),
underlain
by
numerous
formations
including
the
Pottsville,
Allegheny,
Conemaugh,
and
Mauch
Chunk.
These
formations
are
dominated
by
sandstone
and
shale
and
are
all
associated
with
coal
deposits.
The
geology
maps
also
indicate
the
presence
of
Greenbrier
limestone
in
the
downstream
ends
of
each
of
the
major
creeks.
In
addition
to
the
streams
in
the
area,
there
are
7.5
ha
of
small
ponds
in
the
Dolly
Sods
wilderness.
Both
wilderness
areas
also
contain
vernal
pools
and
a
number
of
bogs
and
wetlands
that
serve
as
habitat
for
amphibians
in
the
area.

An
extensive
synoptic
survey
of
chemistry
data
for
the
Dolly
Sods
and
Otter
Creek
wilderness
areas
was
undertaken
by
the
U.
S.
Forest
Service
in
May,
1994.
The
survey
sampled
a
total
of
78
sites
in
Otter
Creek
and
43
sites
in
Dolly
Sods
(
Webb,
1995).
Figures
5
and
6
show
the
ANC
of
the
sample
sites
on
the
100,000­
scale
stream
network.
The
Dolly
Sods
area
is
strongly
affected
by
acidic
deposition
All
the
streams
in
the
Dolly
Sods
had
ANC

50

eq/
L
and
82%
of
the
stream
length
was
acidic
(
Table
5).
The
interquartile
range
in
pH
was
4.3
to
5.1.
None
of
the
streams
had
high
DOC
(
max
=
4.7,
median
=
2.2
mg/
L),
so
organic
acids
are
unlikely
to
be
an
important
source
of
acidity.
Some
of
the
vernal
pools
sampled
in
Dolly
Sods,
however,
did
have
high
DOC
(
max
=
10
mg/
L).
Sulfate
concentrations
(
median
=
105

eq/
L,
Table
2)
were
lower
than
steady­
state
estimates
(
177

eq/
L),
indicating
about
40%
retention
within
the
watershed
(
Table
4).
There
was
no
evidence
of
any
watershed
sources
of
SO
4
2

(
maximum
stream
SO
4
2

=
143

eq/
L).
Concentrations
of
the
other
acid
anions
were
very
low
(
Table
2).

In
the
Otter
Creek
wilderness,
the
majority
of
the
stream
length
was
acidic
(
53%),
17%
had
ANC
between
0
and
50

eq/
L,
and
18%
had
ANC
>
50

eq/
L
(
Table
5;
Figure
6).
Another
13%
of
the
stream
length
was
heavily
influenced
by
organic
anions
(
DOC
between
5
and
8.5
mg/
L).
This
stream
length
is
probably
influenced
more
by
organic
acids
than
acidic
deposition.
For
the
wilderness
area
as
a
whole,
the
interquartile
range
in
pH
was
4.1
B
6.0.
Interpretation
of
this
information
is
complicated
by
the
fact
that
a
limestone
doser
is
operated
on
the
main
stem
of
Otter
Creek
to
ameliorate
the
acidic
conditions
(
see
Figure
6
for
location).
The
resulting
chemical
change
from
the
doser
is
quite
apparent
in
Figure
6.
The
main
stem
of
Otter
Creek
had
high
ANC
for
~
4
km
and
low
ANC
for
~
3
km
downstream
of
the
doser
before
becoming
acidic
again.
This
section
of
the
creek
would
probably
be
acidic
in
the
absence
of
the
limestone
additions.
The
only
naturally
high
ANC
stream
section
is
a
tributary
in
the
northwest
corner
of
the
wilderness
area
(
Figure
6).
Although
SO
4
2

concentrations
were
higher
in
21
Figure
5.
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Dolly
Sods
wilderness
area.
The
dots
indicate
sample
sites.
22
Figure
6.
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Otter
Creek
wilderness
area.
The
dots
indicate
sample
sites.
23
Otter
Creek
than
in
Dolly
Sods
(
median
of
129
vs.
105

eq/
L),
there
still
do
not
appear
to
be
any
significant
watershed
sources
of
SO
4
2

.
The
maximum
SO
4
2

concentration
in
any
stream
in
Otter
Creek
was
224

eq/
L,
only
26%
higher
than
the
expected
steady
state
value
of
177

eq/
L
for
this
wilderness
area
(
Table
5).

In
summary,
both
the
Otter
Creek
and
the
Dolly
Sods
wilderness
areas
in
the
West
Virginia
Plateau
lie
on
base­
poor,
resistant
bedrock
and
have
very
high
percentages
(
70%,
100%,
respectively)
of
acidic
and
low
ANC
streams.
In
the
vast
majority
of
the
systems,
the
acid
anions
in
the
streamwater
are
dominated
by
anions
from
deposition.
Among
acidic
streams
in
these
two
areas,
inorganic
monomeric
aluminum
concentrations
(
interquartile
range
=
52
B
112

g/
L
for
Dolly
Sods,
140
B
348

g/
L
for
Otter
Creek)
were
above
threshold
levels
for
adverse
biological
impacts
(
see
Section
3.3).

It
is
virtually
impossible
to
quantify
the
degree
of
acidification
in
these
systems.
Anecdotal
reports
indicate
that
these
systems
were
"
always
acidic"
or
are
"
naturally
acidic."
It
is
hard
to
assess
what
that
means
quantitatively
(
pH
<
7?,
turns
litmus
paper
red?).
It
is
very
difficult
to
compare
ANC
and
pH
measurements
of
50
years
ago
to
current
measurements.
Also,
these
systems
were
probably
receiving
acidic
deposition
50
years
ago.
A
likely
scenario
is
that
in
pre­
industrial
times,
these
streams
had
very
low
ionic
strength,
almost
distilled
water
supersaturated
with
CO
2
(
ANC
levels
of
0
B
20

eq/
L,
pH
in
the
low
5s,
and
little
aluminum).
If
so,
then
the
degree
of
acidification
over
the
last
150
years
would
be
about
0.5
B
1
pH
unit,
20
B
40

eq/
L
of
ANC,
and
50
B
250

g/
L
of
inorganic
aluminum.
This
would
be
consistent
with
the
historic
pH
and
ANC
changes
reported
from
paleolimnological
analysis
of
sediment
cores
in
many
deposition­
dominated
acidic
lakes
in
the
Adirondack
Mountains
of
New
York
(
Sullivan,
1990,
Sullivan
et
al.,
1992).

3.1.3.2
Northern
Blue
Ridge.
Shenandoah
National
Park
and
the
James
River
Face
wilderness
area
are
both
located
in
the
Blue
Ridge
Mountains
in
the
northern
and
middle
parts
of
Virginia,
respectively
(
Figure
1).
Shenandoah
National
Park
is
a
thin
linear
band
straddling
a
110­
km
segment
of
the
Blue
Ridge
(
788
km2)
over
an
elevation
range
of
600
B
4050
ft
above
sea
level.
Logging,
farming
and
mining
occurred
in
various
parts
of
the
park
before
its
establishment
in
1935.
There
are
five
major
bedrock
formations
in
the
park:
the
Old
Rag
(
a
coarsely
crystalline
granite),
Pedlar
(
feldspathic
granodiorite),
Catoctin
(
a
metamorphosed
basalt),
Hampton
(
phyllite,
shale,
sandstone,
and
quartzite),
and
the
Antietam
(
sandstone
and
quatzite)
(
Lynch
and
Dise,
1985).
As
discussed
in
Section
3.1.2,
there
is
a
strong
relationship
between
geology
and
ANC
in
the
park.
The
Hampton
and
Antietam
formations
in
the
southwestern
part
of
the
park
are
most
resistant
to
weathering
and
have
the
streams
with
the
lowest
ANC.
The
streams
in
the
park
are
virtually
all
first
and
second
order
(
1:
100,000
scale),
draining
east
and
west
off
the
crest
of
the
Blue
Ridge.
The
James
River
Face
wilderness
area
(
35
km2)
is
located
on
the
southern
bank
of
the
James
River,
where
it
cuts
through
the
Blue
Ridge
Mountains.
Elevation
ranges
from
650
to
2950
ft
above
sea
level.
The
streams
are
almost
all
first
order
and
drain
outwards
from
the
high­
elevation
area
in
the
middle
of
the
wilderness
area
(
Figure
7).
The
geologic
formations
in
the
James
River
Face
are
from
the
Chilhowee
Group
(
sandstone
and
quartzite
and
are
similar
to
the
Hampton
and
Antietam
formations
found
in
the
southwestern
part
of
Shenandoah
National
Park.
There
is
also
a
minor
amount
of
the
Pedlar
Formation
in
the
James
River
Face
(
Virginia,
1963).
24
Figure
7.
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
James
River
Face
wilderness
area.
The
dots
indicate
sample
sites.
25
The
most
comprehensive
survey
of
streams
in
Shenandoah
National
Park
was
conducted
by
Lynch
and
Dise
(
1985)
in
1981
and
1982.
Sample
sites
were
not
selected
with
a
statistical
design
and
we
did
not
attempt
to
make
estimates
of
the
lengths
in
various
chemical
classes.
Based
on
flow­
weighted
annual
average
data
from
47
of
these
sites,
covering
70%
of
the
park
area,
Cosby
et
al.
(
1991)
reported
that
6%
of
the
sample
sites
were
acidic
and
25%
had
ANC

50

eq/
L
(
Table
5).
Over
300
streams
were
sampled
in
the
mountains
of
western
Virginia
as
part
of
the
VTSSS
(
Cosby
et
al.,
1991).
Eight
of
these
sites
were
within
or
on
streams
draining
from
the
James
River
Face
wilderness
area
(
Figure
7).
According
to
chemistry
data
from
the
VTSSS
collected
at
these
sites
between
1991
and
1994,
virtually
all
(
92%)
the
stream
length
had
ANC
between
0
and
50

eq/
L
(
Table
5).
Anion
chemistry
in
the
Shenandoah
is
similar
to
that
of
the
James
River
Face;
median
SO
4
2

values
were
85
and
68

eq/
L,
and
median
NO
3

values
were
7
and
0

eq/
L,
respectively
(
Table
2).
ANC
and
pH
are
lower
in
the
James
River
Face
due
to
the
concentration
of
more
resistant
bedrock
in
that
wilderness
area
compared
to
Shenandoah
Park.
Both
areas
retain
most
of
the
incoming
SO
4
2

(
57
B
68%)
and
almost
all
the
incoming
N
(
Table
4).

3.1.3.3.
Southern
Blue
Ridge.
Most
of
the
land
area
and
stream
length
in
the
SAMIregion
Class
I
wilderness
areas
are
in
the
Great
Smoky
Mountains
National
Park
(
GSMNP;
Table
3).
The
GSMNP
(
2,300
km2)
is
located
on
the
Tennessee/
North
Carolina
border
in
an
area
of
uplifted
sedimentary
rock.
Elevations
in
the
park
range
from
900
to
6640
ft.
Bedrock
is
primarily
sandstone
with
some
limestone
(
Elwood
et
al.,
1991).
Many
investigators
have
noted
that
one
of
the
formations
(
Anakeesta)
contains
pyrite
that
can
be
oxidized
to
sulfuric
acid
when
exposed
to
air
and
water
during
watershed
disturbances
(
landslides
road
cuts).
The
state
geologic
map
also
notes
the
presence
of
sulfidic
bedrock
in
other
formations
within
the
park
(
e.
g.,
Copperhill,
Boyd
Gap;
North
Carolina,
1985).

Based
on
a
number
of
biogeochemical
studies,
the
following
conclusions
can
be
made
about
streams
in
GSMNP
(
Silsbee
and
Larson,
1982;
Elwood
et
al.,
1991;
Cook
et
al.,
1994;
Flum
and
Nodvin,
1995).

$
As
in
Shenandoah
National
Park,
bedrock
geology
is
a
good
predictor
of
general
streamwater
ANC
status.

$
Low
ANC
(

50

eq/
L)
streams
are
common
in
the
park.

$
Acidic
streams
exist
in
the
park,
primarily
at
higher
elevations.

$
ANC
tends
to
be
lower
and
NO
3

tends
to
be
higher
at
higher
elevations.

$
Nitrate
and
SO
4
2

concentrations
are
comparable
in
many
streams;
in
higher
elevation
catchments,
NO
3

concentrations
are
often
greater
than
SO
4
2

concentrations.
Streams
in
watersheds
with
a
history
of
logging
have
lower
NO
3

than
streams
in
unlogged
watersheds.
Thus,
N/
forest
dynamics
play
a
major
role
in
controlling
stream
anion
chemistry
in
GSMNP.

$
Most
of
the
incoming
SO
4
2

and
N
is
retained
within
the
watershed
in
the
majority
of
the
studied
streams.

$
Streams
that
were
acidic
and
had
SO
4
2

concentrations
>
~
65

eq/
L
probably
are
influenced
by
sulfide
mineral
weathering
and
are
in
watersheds
containing
the
Anakeesta
Formation.

$
Higher
ANC
systems
(>
100

eq/
L)
probably
are
influenced
by
limestone
weathering
and
are
concentrated
in
the
far
western
end
of
the
park.

On
the
1:
100,000­
scale
maps,
there
are
2,035
km
of
streams.
Elwood
et
al.
(
1991)
26
report
a
length
of
1,173
km
of
streams
in
the
GSMNP
capable
of
supporting
trout
and/
or
small­
mouth
bass.
Flum
and
Nodvin
(
1995)
have
conducted
a
large
survey
of
streamwater
chemistry
in
GSMNP.
Using
their
data,
we
calculated
average
spring
chemistry
for
their
sample
sites
by
averaging
all
March,
April,
and
May
chemistry
data
they
collected
in
1994
and
1995.
These
streams
(
337
sites)
had
very
low
ionic
strength,
and
had
median
SO
4
2

and
NO
3

concentrations
of
31

eq/
L
and
15

eq/
L,
respectively
(
Table
2).
These
medians
signify
a
70%
retention
of
deposition
SO
4
2

and
65%
retention
of
deposition
NO
3

(
Table
4).
These
337
sites
represent
736
km
of
the
2,035
km
of
streams
on
the
100,000­
scale
maps
in
GSMNP.
Of
this
assessed
length,
4%
(
32
km)
was
acidic
and
46%
(
336
km)
had
ANC
between
0
and
50

eq/
L
(
Table
5).

The
expected
steady­
state
SO
4
2

concentration
in
GSMNP
is
104

eq/
L.
Typically,
SO
4
2

concentrations
are
much
lower
than
steady
state
but
streams
with
higher
SO
4
2

do
exist
in
GSMNP,
probably
due
to
bedrock
sulfide
weathering.
In
the
Flum
and
Nodvin
(
1995)
data,
9
of
the
337
sites
had
SO
4
2

>
100

eq/
L
(
maximum=
248

eq/
L),
corresponding
to
about
8
km
of
stream
length
(
1%
of
total
assessed
length).
Of
the
23
stream
sites
that
were
acidic,
11
had
SO
4
2

>
65

eq/
L
and
1
had
SO
4
2

>
100

eq/
L.

The
Joyce
Kilmer/
Slickrock
wilderness
area
(
69
km2)
contains
the
headwaters
of
Slickrock
and
Little
Santeetlah
Creeks
and
is
adjacent
to
the
southwest
boundary
of
the
GSMNP.
As
such,
it
has
geology
and
topography
similar
to
those
of
the
GSMNP
and
the
conclusions
listed
above
for
GSMNP
are
probably
applicable
to
the
Slickrock
wilderness
area.
Elevations
in
the
Slickrock
wilderness
area
range
from
2000
to
5300
ft
and
the
bedrock
geology
is
composed
primarily
of
the
Copperhill
(
metagraywacke,
slate,
schist)
and
Boyd
Gap
Formations
(
slate,
metasiltstone,
and
metagraywacke).
Both
formations
have
notations
on
the
geologic
maps
indicating
that
they
contain
sulfidic
bedrock
(
North
Carolina,
1985).
We
were
able
to
obtain
chemistry
data
for
9
stream
sites
in
the
Slickrock
wilderness
area
from
Bill
Jackson
and
Richard
Burns
with
the
U.
S.
Forest
Service
in
Asheville,
North
Carolina
(
Figure
8;
unpublished
data).
These
9
sites
represent
about
22
km
of
the
73
km
of
streams
on
the
100,000­
scale
map
of
the
area;
about
one­
third
of
the
assessed
stream
length
had
ANC
between
0
and
50

eq/
L
and
the
rest
had
ANC
>
50

eq/
L
(
Table
5).

The
Linville
Gorge
wilderness
area
covers
a
stretch
of
the
Linville
River
roughly
20
km
long
as
it
flows
through
Linville
Gorge.
Other
than
the
Linville
River,
there
are
virtually
no
aquatic
resources
in
this
wilderness
area
(
see
Appendix
A).
It
is
highly
unlikely
that
the
Linville
River
is
acid­
sensitive.
In
any
event,
water
quality
is
not
an
air­
quality
related
value
for
the
Linville
Gorge
wilderness
area
(
Bill
Jackson,
pers.
comm.).
Thus,
we
will
not
assess
its
acid­
base
status
any
further.

The
Shining
Rock
wilderness
area
(
75
km2)
drains
parts
of
the
headwaters
of
the
East
Fork
of
the
Pigeon
River
in
North
Carolina,
just
north
of
the
South
Carolina
border.
Elevations
in
this
area
range
from
3300
to
6000
feet.
The
bedrock
geology
is
primarily
from
the
Tallulah
Falls
Formation,
a
locally
sulfidic
muscovite­
biotite
gneiss
(
North
Carolina
1985).
We
were
able
to
obtain
chemistry
data
for
9
stream
sites
in
the
Shining
Rock
wilderness
area
from
Bill
Jackson
and
Richard
Burns
with
the
U.
S.
Forest
Service
in
Asheville,
North
Carolina
(
Figure
9;
unpublished
data).
All
9
sites
had
ANC
between
50
and
100

eq/
L.
These
9
sites
represented
about
10
km
of
the
44
km
of
streams
on
the
100,000­
scale
map
of
the
area
(
Table
5).
27
Figure
8.
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Joyce
Kilmer/
Slickrock
wilderness
area.
The
dots
indicate
sample
sites.
28
Figure
9.
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Shining
Rock
wilderness
area.
The
dots
indicate
sample
sites.
29
The
Cohutta
wilderness
area
(
121
km2)
drains
the
headwaters
of
the
Conasauga
and
Jacks
rivers
in
northern
Georgia.
Elevation
ranges
from
1000
to
3900
feet
and
the
bedrock
is
composed
primarily
of
mica
and
graphite
schists,
gneiss,
and
amphibolite
(
Georgia,
1976).
David
Wergowske
(
pers.
comm.),
from
the
U.
S.
Forest
Service
Chattahoochee­
Oconee
National
Forest
provided
streamwater
chemical
data
from
15
sites
in
the
Cohutta.
Data
were
collected
during
1992­
1994
and
represent
about
threefourths
of
the
145
km
of
streams
represented
on
the
1:
100,000­
scale
maps
of
the
wilderness
area
(
Figure
10).
Virtually
all
(
96%)
of
the
sampled
stream
length
had
ANC

50

eq/
L
but
was
not
acidic
(
Table
5).
Streamwater
chemistry
in
Cohutta
was
very
similar
to
that
observed
in
the
GSMNP.
Streams
have
very
low
DOC
(
median=
1.8
mg/
L)
and
high
NO
3

concentrations
(
median=
14

eq/
L)
relative
to
the
rest
of
the
SAMI
region.
Roughly
twothirds
of
the
incoming
SO
4
2

and
NO
3

is
retained
within
the
watershed
(
Table
4).
Thus,
watersheds
in
Cohutta
are
"
leaking"
nitrogen.

3.1.3.4
Alabama
Appalachian
Plateau.
The
Sipsey
wilderness
area
(
52
km2)
in
northern
Alabama
is
located
in
the
Cumberland
Plateau
(
southern
part
of
the
Appalachian
Plateau).
The
area
is
drained
by
Sipsey
Fork,
formed
by
the
confluence
of
Thompson
and
Hubbard
creeks.
The
headwaters
of
this
basin
lie
outside
the
wilderness
area
boundary.
Elevation
ranges
from
500
to
1000
ft.
A
study
of
the
stream
chemistry
in
the
Sipsey
wilderness
area
was
undertaken
by
Dr.
G.
Milton
Ward
of
the
University
of
Alabama
during
1991B1993.
He
described
the
results
in
a
series
of
reports
to
the
U.
S.
Forest
Service
(
Ward,
1991,
1992,
1993).
Parts
of
this
wilderness
area
(
around
mainstem
Sipsey
Fork)
are
underlain
by
the
Bangor
limestone
formation
and
the
streams
draining
this
formation
have
such
high
ANC
levels
that
they
are
not
likely
ever
to
be
affected
by
acidic
deposition
(
Table
2,
Figure
11).
Other
parts
of
this
wilderness
area
are
drained
by
bedrock
of
the
Pottsville
(
pebbly
quartzose
sandstone)
and
Parkwood
(
shale,
sandstone)
formations
(
Osborne
et
al.,
1989).
These
streams
have
lower
ANC
and
may
be
sensitive
to
acidic
deposition.
Just
over
half
(
52%)
of
the
100,000­
scale
map
stream
length
in
the
area
was
sampled
for
stream
chemistry
(
Table
2).
Sampling
was
concentrated
on
the
mainstem
Sipsey
Fork
and
major
tributaries.
Within
the
half
of
the
network
that
was
sampled,
63%
(
including
Sipsey
Fork)
had
very
high
ANC
(
200B800

eq/
L)
and
the
other
37%
(
14
km)
had
ANC
levels
of
40B50

eq/
L
(
Figure
11).
In
terms
of
anion
chemistry,
Sipsey
is
the
only
Class
I
area
that
appears
to
be
at
steady
state
with
respect
to
SO
4
2

deposition
(
i.
e.,
little
to
no
watershed
SO
4
2

retention).
Thus,
a
delayed
response
to
S
deposition
is
unlikely
in
this
area.
Nitrate,
on
the
other
hand,
is
largely
retained
in
the
watershed
(
Table
4).

3.1.4
Class
I
Area
Summary
Upon
comparing
results,
it
is
clear
that
the
smaller
Class
I
areas
are
much
more
sensitive
to
acidic
deposition
(
higher
percentages
of
low
ANC
streams)
than
their
specific
regions
as
a
whole
(
Table
6).
In
terms
of
adverse
aquatic
effects
of
acidic
deposition,
the
four
Class
I
groups
can
be
ranked:
West
Virginia
Plateau
>>
Northern
Blue
Ridge
>
Southern
Blue
Ridge
>
Alabama
Plateau.

Streams
in
the
West
Virginia
Plateau
Class
I
areas
had
the
highest
percentage
of
acidic
stream
length
(
53%
in
Otter
Creek,
82%
in
Dolly
Sods),
the
highest
SO
4
2

and
inorganic
aluminum
concentrations,
and
the
lowest
pH
of
any
of
the
Class
I
areas.
They
are
currently
heavily
impacted
by
acidic
deposition.
The
Sipsey
wilderness
area
is
of
least
concern
for
acidic
deposition
impacts
because
30
Figure
10.
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Cohutta
wilderness
area.
The
dots
indicate
sample
sites.
31
Figure
11.
Stream
ANC
classification
for
the
1:
100,000­
scale
USGS
map
stream
network
in
the
Sipsey
wilderness
area.
The
dots
indicate
sample
sites.
32
Table
6.
Summary
of
Acidic
and
Acid­
sensitive
Stream
Percentages
in
Class
I
Wilderness
Areas
and
Other
Aggregate
Physiographic
Regions
of
the
Southern
Appalachians.
Data
for
the
Class
I
wilderness
area
is
from
Table
5
and
is
based
on
the
1:
100,000­
scale
stream
network.
%
total
length
assessed
refers
to
the
percentage
of
the
length
of
streams
in
the
region
with
which
there
were
available
data
to
infer
ANC
status.

REGION
%
Assessed
Length
with
ANC

0
(

eq/
L)
%
Assessed
Length
with
ANC

50
(

eq/
L)
Total
Length
(
km)
%
Total
Length
Assessed
(
had
Data)

WEST
VIRGINIA
PLATEAU
Dolly
Sods
Wilderness
Area
92
100
25
100%

Otter
Creek
Wilderness
Area
53
70
57
93%

Entire
Province
­
NSS
(
n=
37)
5
17
14,100
100%

NORTHERN
BLUE
RIDGE
Shenandoah
National
Park
6**
25**
n=
47**
**

James
River
Face
Wilderness
Area
0
92
30
75%

Entire
Province
­
NSS
(
n=
13)
0
4
8,700
100%

Virginia
Blue
Ridge
­
VTSSS
6**
42**
n=
147**
**

SOUTHERN
BLUE
RIDGE
Great
Smoky
Mt.
National
Park
4
50
2,035
36%

Slickrock
Wilderness
Area
0
36
73
30%

Shining
Rock
Wilderness
Area
0
0
44
23%

Cohutta
Wilderness
Area
0
96
140
77%

Entire
Province
­
NSS
(
n=
55)
0
7
10,800
100%

Entire
Province
­
WINGER
2**
4**
n=
62**
**

ALABAMA
PLATEAU
Sipsey
Wilderness
Area
0
37
70
53%

**
Estimate
reflects
percentage
of
sample
size,
not
length.
Representativeness
of
the
sample
to
the
population
is
unknown
and
may
not
accurately
reflect
the
entire
population.
The
sample
size
used
in
the
estimate
is
shown
in
the
Total
Length
column.

VTSSS
­
Estimates
are
based
on
the
Virginia
Trout
Stream
Sensitivity
Study
data
(
Webb
et
al.,
1989a).

NSS
­
Estimates
based
on
weighted
extrapolation
from
probability
sample
data
from
the
National
Stream
Survey
(
Kaufmann
et
al.,
1988);
stream
network
from
1:
250,000
scale
maps.

WINGER
­
Estimates
are
based
on
study
by
Winger
et
al.,
1987.
33
SO
4
2

appears
to
be
at
steady
state
and
it
has
higher
streamwater
ANC
concentrations.
The
Plateau
areas
retain
less
SO
4
2

than
do
the
Blue
Ridge
areas
so
any
delayed
effect
of
acidic
deposition
will
be
more
pronounced
in
the
Blue
Ridge.
Watersheds
in
both
of
the
Blue
Ridge
areas
retain
the
majority
of
the
entering
SO
4
2

from
deposition.
Low
ANC
streams
are
common
and
some
acidic
streams
are
found
in
areas
of
resistant
bedrock
and/
or
higher
elevations.
The
northern
Blue
Ridge
areas
have
higher
sulfate
concentrations
than
the
southern
Blue
Ridge
areas
and
seem
to
be
more
influenced
by
acidic
deposition.
The
southern
Blue
Ridge
areas
stand
out
from
the
other
Class
I
areas
in
terms
of
having
the
highest
NO
3

concentrations
(
lowest
N
retention
It
appears
that
NO
3

is
breaking
through
these
watersheds
and
is
entering
streams
in
concentrations
that
approach
and
sometimes
exceed
SO
4
2

concentrations.

3.2
Episodic
Chemical
Conditions
Most
of
the
research
on
(
and
modeling
of)
acidification
of
surface
waters
has
dealt
with
the
problem
of
chronic
acidification
over
relatively
long
time
scales
(
years
to
centuries)
and
most
regional
assessments
of
surface
water
acid­
base
status
have
focused
on
average
annual
conditions
or
on
so­
called
"
index"
periods.
However,
transient
changes
in
acid­
base
status
associated
with
hydrological
events
(
i.
e.,
those
occurring
during
snowmelt
and
rainfall
periods)
represent
a
potentially
important
acidification
processC
known
as
episodic
acidificationCin
regions
that
receive
acidic
deposition,
including
the
entire
SAMI
region.
In
this
section
of
the
report,
we
briefly
summarize
the
synthesis
documents
pertaining
to
the
problem
of
episodic
acidification
of
surface
waters,
namely
NAPAP
State
of
Science
and
Technology
Report
12,
entitled
Episodic
Acidification
of
Surface
Waters
Due
to
Acid
Deposition
(
Wigington
et
al.,
1990),
and
the
NAPAP
1990
Integrated
Assessment
Report
(
NAPAP,
1991).
This
summary
focuses
on
the
factors
most
likely
to
control
episodic
acidification
of
surface
waters
and
on
the
techniques
that
can
be
used
to
quantify
current
regional
and
localscale
episodic
acidification
impacts.
In
the
second
part
of
the
section,
we
focus
our
discussion
specifically
on
surface
waters
of
the
Southern
Appalachians,
including
those
in
the
Class
I
wilderness
areas.

3.2.1
Synopsis
of
Synthesis
Documents
Episodic
acidification
is
defined
as
the
process
by
which
lakes
and
streams
experience
short­
term
decreases
in
ANC,
generally
during
hydrological
events
and
over
time
scales
varying
from
several
hours
to
several
weeks.
ANC
depressions
are
usually
accompanied
by
changes
in
concentration
of
one
or
more
of
the
following:
hydrogen
ion,
base
cations,
dissolved
organic
carbon,
sulfate,
nitrate,
and
various
forms
of
dissolved
aluminum.
In
highly
acid­
sensitive
lakes
and
streams
(
chronically
low
levels
of
ANC),
episodic
ANC
depressions
are
usually
accompanied
by
increases
in
hydrogen
ion
concentrations
(
i.
e.,
decreases
in
pH)
and
increases
in
dissolved
aluminum
concentrations.
These
transient
increases
in
hydrogen
ion
and
aluminum
concentrations
can
cause
significant
nonlethal
stress
or
increased
mortality
of
some
species
and
life
stages
of
fish
and
other
aquatic
organisms
(
Wigington
et
al.,
1990).

In
a
thorough
review
and
synthesis
of
results
published
from
studies
of
episodic
acidification
conducted
in
the
United
States,
Canada,
and
Europe,
Wigington
et
al.
(
1990)
were
able
to
draw
four
important
conclusions:

1.
Episodic
acidification
is
a
ubiquitous
process
in
streams
and
drainage
lakes
and
is
not
necessarily
symptomatic
of
anthropogenically
produced
chronic
acidification
(
i.
e.,
episodic
acidification
occurs
both
in
waters
that
are
chronically
acidic
and
in
those
that
are
not).
However,
episodic
and
chronic
acidification
are
clearly
34
related,
as
the
magnitude
of
ANC
depressions
is
usually
larger
in
intermediate
to
high
ANC
systems
than
in
low
ANC
or
acidic
systems,
and
the
lowest
minimum
ANC
and
pH
levels
occur
during
episodes
in
surface
waters
with
low
preepisode
ANC
levels.
These
observations
at
least
qualitatively
support
the
explanation
that
ANC
dilution
is
an
important
component
of
episodic
acidification.

2.
The
characteristics
of
episodes
are
determined
largely
by
changes
in
hydrological
flowpaths
during
rainfall
and
snowmelt
events.
During
baseflow
periods,
streamflow
is
derived
largely
from
drainage
of
relatively
alkaline
groundwater,
whereas
during
stormflow
periods,
both
pre­
event
and
event
water
can
be
routed
through
either
acidic
upper
soil
layers
(
as
subsurface
stormflow)
or
across
the
soil
surface
as
overland
flow
(
as
saturation
overland
flow).
In
drainage
lakes
in
cold
temperate
regions
where
surface
ice
forms,
inverse
thermal
stratification
accentuates
the
episodic
acidification
response
by
restricting
the
hydrological
mixing
of
snowmelt
and
ambient
lake
water
under
ice
cover.

3.
Episodic
acidification
can
be
attributed
to
several
different
anthropogenic
and
natural
processes,
including
direct
inputs
of
mineral
acids
(
sulfuric
and
nitric)
from
the
atmosphere
during
events,
antecedent
deposition
of
mineral
acids
to
watershed
soils
(
chronic
or
antecedent
"
conditioning
hydrological
dilution,
nitrification,
organic
acid
production,
and
deposition
of
neutral
sea
salts.
Any
one
or
several
of
these
mechanisms
may
be
operative
in
a
particular
watershed.
In
many
watersheds
of
the
northeastern
and
mid­
Atlantic
regions
of
the
United
States,
atmospheric
deposition
appears
to
be
operating
in
conjunction
with
several
natural
processes
to
generate
episodes
with
lower
minimum
ANC
and
pH
and
higher
dissolved
aluminum
levels.
Most
likely
atmospheric
deposition
is
affecting
episodic
acidification
in
these
(
and
perhaps
other)
regions
by
providing
direct
inputs
to
surface
waters,
by
conditioning
watersheds
during
antecedent
periods,
and
by
reducing
chronic
ANC
levels.
Thus,
relatively
small
ANC
depressions
are
capable
of
creating
acidic,
high
Al
conditions
in
streams
and
lakes.

4.
Modeling
of
episodic
acidification
has
been
only
moderately
successful,
due
primarily
to
a
lack
of
understanding
of
hydrological
pathways
and
biogeochemical
reactions.
Several
fairly
simple
empirical
models
have
been
successfully
linked
to
surface
water
survey
databases
to
provide
crude
estimates
of
regional
episodic
acidification
impacts
on
streams
in
the
eastern
United
States
and
on
lakes
in
the
Adirondack
Mountains
of
New
York.

While
these
conclusions
were
based
on
a
relatively
large
number
of
field
studies,
the
authors
also
concluded
that
a
clear
scientific
consensus
regarding
the
primary
causes
of
episodic
acidification
had
not
yet
emerged.
In
addition,
the
report
concluded
that
there
was
little
knowledge
of
the
regional
extent
or
environmental
significance
of
episodic
acidification
in
any
of
the
regions
where
field
studies
had
been
conducted
and
that
statistically
rigorous
population
estimates
of
current
or
future
episodic
acidification
impacts
could
not
be
made
with
data
that
were
available
at
the
time.

Finally,
the
possible
biological
significance
of
episodic
acidification
had
not
been
properly
addressed
at
the
time
the
synthesis
document
was
written
(
Wigington
et
al.,
1990).
To
address
these
assessment
issues,
Wigington
et
35
al.
(
1990)
identified
several
critical
knowledge
gaps:

1.
Additional
field
data
collected
with
high
temporal
resolution
in
various
geographic
regions
of
the
United
States
(
in
particular
the
southeastern,
upper
midwestern,
and
western
regions)
that
could
be
used
for
calibrating
simple
models
of
episodic
acidification
for
estimating
current
regional
impacts.

2.
Long­
term
episodic
data
for
describing
the
behavior
of
systems
over
longer
time
periods.

3.
New
techniques
for
quantifying
watershedscale
hydrological
flowpaths.

4.
Better
understanding
of
biogeochemical
processes
involving
nitrogen,
sulfur,
and
natural
organic
acids.

The
1990
NAPAP
assessment
of
current
and
future
acidification
impacts
of
acidic
deposition
recognized
that
assessments
based
solely
on
chemical
conditions
during
index
periods
(
i.
e.,
periods
approximating
average
annual
or
seasonal
conditions)
did
not
account
for
the
worst­
case
chemical
conditions
that
usually
occur
during
extreme
hydrological
events.
However,
due
to
the
rather
large
uncertainties
associated
with
incorporating
episodic
acidification
into
the
regional
assessment
framework,
model­
based
projections
of
watershed
responses
to
emissions
reduction
scenarios
were
not
formulated
to
account
for
changes
in
episodic
acidification
over
long
time
periods.

3.2.2
Current
Episodic
Acidification
Impacts
in
the
Southern
Appalachians
As
summarized
by
Wigington
et
al.
(
1990),
field
studies
of
current
episodic
acidification
impacts
have
been
conducted
on
at
least
two
continents
(
North
America
and
Europe),
including
many
physiographic
regions
of
the
United
States
and
much
of
Canada.
As
previously
observed,
episodic
acidification
has
been
found
to
be
virtually
ubiquitous;
therefore,
episodes
must
be
accounted
for
in
any
overall
scheme
to
assess
the
current
acidbase
status
of
surface
waters,
either
at
local
or
regional
scales.
For
several
regions
of
the
eastern
United
States,
models
have
been
employed
to
predict
current
episodic
acidification
impacts.
Most
of
these
modeling
studies
have
unfortunately
been
based
on
extremely
limited
data
sets.
The
following
paragraphs
summarize
our
current
understanding
of
episodic
acidification
in
the
Southern
Appalachians
at
the
two
scales
of
interestClocal
and
regional.

3.2.2.1
Local
Scale.
Wigington
et
al.
(
1990)
specifically
identified
the
southeastern
United
States
as
a
region
for
which
no
complete
data
sets
on
episodic
acidification
were
available
and
for
which
incomplete
data
were
available
for
fewer
than
10
streams
in
the
entire
region.
Declines
in
pH
and
ANC
were
noted
in
all
studies,
but
only
one
study
reported
minimum
episodic
pH
values
below
5.0
(
Raven
Fork,
North
Carolina).
Since
1990,
however,
data
from
several
intensive
studies
of
episodic
acidification
in
the
Southern
Appalachians
have
been
reported
(
Miller­
Marshall,
1993;
Hyer
et
al.,
1995;
Eshleman
et
al.,
1995;
Nodvin
et
al.,
1995),
from
which
local
episodic
impacts
can
be
quantified.
These
studies
suggest
that
streams
with
antecedent
baseflow
ANC
values
less
than
about
25

eq/
L
may
experience
substantial
depressions
in
pH
(
as
much
as
one
unit)
and
increases
in
dissolved
Al
concentrations
(
perhaps
as
much
as
100

g/
L).
Streams
with
higher
antecedent
baseflow
ANC
values
probably
experience
much
smaller
changes
in
pH
and
Al
concentrations.

Another
interesting
aspect
of
this
problem,
addressed
by
Eshleman
et
al.
(
1995),
was
the
effect
of
insect
defoliations
on
the
episodic
mobilization
of
nitric
acid
and
the
associated
acidification
response.
At
White
Oak
Run
in
36
Shenandoah
National
Park,
a
statistical
analysis
of
13
years
of
data
suggested
that
mean
episodic
depressions
of
ANC
have
increased
by
about
a
factor
of
2
since
the
first
outbreak
of
forest
defoliation
by
the
gypsy
moth
caterpillar
during
the
summer
of
1990;
the
mean
episodic
change
in
NO
3

concentration
also
has
increased
by
about
12

eq/
L,
while
the
mean
episodic
dilution
of
C
B
has
decreased
from

8.5

eq/
L
to

1.7

eq/
L
during
the
same
period.
Episodic
changes
in
SO
4
2

have
remained
the
same,
however.

3.2.2.2
Regional
Scale.
Eshleman
(
1988)
estimated
regional­
scale
episodic
acidification
impacts
in
the
Southern
Appalachians
using
an
empirical
two­
component
mixing
model
with
minimal
calibration.
Eshleman
(
1988)
estimated
that
5B7%
of
the
population
of
Southern
Appalachian
stream
reaches
sampled
during
the
National
Stream
Survey
may
become
acidic
(
ANC
<
0)
during
extreme
hydrological
conditions,
compared
to
0B3%
during
typical
spring
baseflow
conditions.
Recently,
Hyer
et
al.
(
1995)
presented
results
from
an
intensive
study
of
episodic
acidification
of
three
streams
in
Shenandoah
National
Park
(
Virginia)
that
support
the
use
of
the
two­
component
mixing
model
of
ANC
and
provide
data
for
model
parameterization.
No
comparison
of
the
model
parameters
has
been
conducted,
although
the
data
are
available
to
do
so.
Additional
field
data
are
needed,
however
to
determine
the
robustness
of
the
model
(
and
of
the
existing
parameters)
for
quantifying
regional
impacts
in
other
regions
of
the
Southern
Appalachians
(
including
Class
I
areas).
3.3
Biological
Effects
of
Acidic
Deposition
3.3.1
Synopsis
of
Synthesis
Documents
The
NAPAP
SOS/
T
report
on
biological
effects
(
J.
Baker
et
al.,
1990)
states
that
there
is
no
doubt
that
acidification,
at
pH
levels
as
high
as
6.0B6.5,
results
in
changes
in
biological
communities.
All
major
groups
of
aquatic
organisms
have
been
affected,
but
individual
species
differ
greatly
in
acid
tolerance.
Two
readily
observable
consequences
are
(
1)
shifts
in
species
composition
and
(
2)
reduction
in
the
total
number
of
species
in
a
body
of
water.
Ecological
processes
are
more
robust
than
some
acid­
sensitive
species.
A
general
summary
of
the
biological
changes
associated
with
declining
pH
is
shown
in
Table
7.

It
is
more
difficult
to
state
precisely
the
magnitude
of
community
effects
on
a
regional
scale.
The
report
provides
guidelines
and
methods
for
estimating
the
magnitude
of
effects,
given
regional
differences
in
data
availability.
The
use
of
suggested
models
provides
three
significant
advances:
(
1)
improved
definition
of
chemical
ranges
critical
for
specific
biological
responses,
(
2)
quantification
of
the
interactions
among
pH,
aluminum,
and
calcium
(
the
most
important
parameters
for
understanding
biological
responses
in
this
context)
at
realistic,
regional
levels,
and
(
3)
estimates
of
the
current
and
projected
numbers
of
lakes
and
streams
unsuitable
for
fish
due
to
acidification.
However,
given
the
complexity
of
natural
systems,
it
would
be
misleading
to
consider
model
outputs
as
absolute;
although
sources
of
uncertainties
in
models
have
been
identified,
their
influence
on
model
outputs
can
be
only
partly
quantified.

The
discussions
of
biological
effects
cover
all
components
of
aquatic
communities
(
algae,
37
Table
7.
General
Summary
of
Biological
Changes
Anticipated
with
Surface
Water
Acidification,
Expressed
as
a
Change
in
pH.
Taken
from
Table
13­
37
in
the
NAPAP
SOS/
T
report
(
J.
Baker
et
al.,
1990).

pH
Decrease
General
Biological
Effects
6.5
to
6.0
Small
decrease
in
species
richness
of
phytoplankton,
zooplankton,
and
benthic
invertebrate
communities
resulting
from
the
loss
of
a
few
highly
acid­
sensitive
species,
but
no
measurable
change
in
total
community
abundance
or
production
Some
adverse
effects
(
decreased
reproductive
success)
may
occur
for
highly
acid­
sensitive
species
(
e.
g.,
fathead
minnow,
striped
bass)

6.0
to
5.5
Loss
of
sensitive
species
of
minnows
and
dace,
such
as
blacknose
dace
and
fathead
minnow;
in
some
waters
decreased
reproductive
success
of
lake
trout
and
walleye,
which
are
important
sport
fish
species
in
some
areas
Visual
accumulations
of
filamentous
green
algae
in
the
littoral
zone
of
many
lakes,
and
in
some
streams
Distinct
decrease
in
the
species
richness
and
change
in
species
composition
of
the
phytoplankton,
zooplankton,
and
benthic
invertebrate
communities,
although
little
if
any
change
in
total
community
biomass
or
production
Loss
of
a
number
of
common
invertebrate
species
from
the
zooplankton
and
benthic
communities,
including
zooplankton
species
such
as
Diaptomus
silicis,
Mysis
relicta,
Epsichura
lacustris;
many
species
of
snails,
clams,
mayflies,
and
amphipods,
and
some
crayfish
5.5
to
5.0
Loss
of
several
important
sport
fish
species,
including
lake
trout,
walleye,
rainbow
trout,
and
smallmouth
bass;
as
well
as
additional
non­
game
species
such
as
creek
chub
Further
increase
in
the
extent
and
abundance
of
filamentous
green
algae
in
lake
littoral
areas
and
streams
Continued
shift
in
the
species
composition
and
decline
in
species
richness
of
the
phytoplankton,
periphyton,
zooplankton,
and
benthic
invertebrate
communities;
decrease
in
the
total
abundance
and
biomass
of
benthic
invertebrates
and
zooplankton
may
occur
in
some
waters
Loss
of
several
additional
invertebrate
species
common
in
oligotrophic
waters,
including
Daphnia
galeata
mendotae,
Diaphanosoma
leuchtenbergianum,
Asplanchna
priodonta;
all
snails,
most
species
of
clams,
and
many
species
of
mayflies,
stoneflies,
and
other
benthic
invertebrates
Inhibition
of
nitrification
38
5.0
to
4.5
Loss
of
most
fish
species,
including
most
important
sport
fish
species
such
as
brook
trout
and
Atlantic
salmon;
few
fish
species
able
to
survive
and
reproduce
below
pH
4.5
(
e.
g.,
central
mudminnow,
yellow
perch,
and
in
some
waters
largemouth
bass)

Measurable
decline
in
the
whole­
system
rates
of
decomposition
of
some
forms
of
organic
matter,
potentially
resulting
in
decreased
rates
of
nutrient
cycling
Substantial
decrease
in
the
number
of
species
of
zooplankton
and
benthic
invertebrates
and
further
decline
in
the
species
richness
of
the
phytoplankton
and
periphyton
communities;
measurable
decrease
in
the
total
community
biomass
of
zooplankton
and
benthic
invertebrates
in
most
waters.

Loss
of
zooplankton
species
such
as
Tropocyclops
prasinus
mexicanus,
Leptodora
kindtii,
and
Conochilis
unicornis;
and
benthic
invertebrate
species,
including
all
clams
and
many
insects
and
crustaceans
Reproductive
failure
of
some
acid­
sensitive
species
of
amphibians
such
as
spotted
salamanders,
Jefferson
salamanders,
and
the
leopard
frog
zooplankton,
benthic
invertebrates,
fish,
amphibians
and
waterfowl)
and
major
ecological
processes.
However,
quantitative
methods
and
models,
plus
regional
discussions,
cover
only
fish
responses.
The
reasons
for
this
limitation
include
data
availability
(
historical
and
experimental),
as
well
as
the
ability
to
make
direct
linkages
between
water
chemistry
and
toxic
mechanisms
in
fish.

The
major
conclusions
of
this
SOS/
T
report
are
shown
in
italics
(
J.
Baker
et
al.,
1990)
and
discussed
in
the
following
subsections.

3.3.1.1
Chemical
Factors
Influencing
Biota.
The
most
important
chemical
properties
of
surface
waters
influencing
biological
responses
to
acid­
base
chemistry
are
pH,
aluminum,
and
calcium.

Surface
water
pH
is
probably
the
most
important
of
the
three;
decreases
in
pH
(
particularly
at
levels
below
6.0B6.5)
have
been
shown
to
produce
negative
effects
on
many
aquatic
animals.
Species
(
and
life
history
stages
within
species)
differ
greatly
in
tolerance
to
acid
conditions;
in
brook
trout,
for
example,
adults
do
not
typically
disappear
from
streams
until
baseflow
pH
falls
below
5.5,
but
sublethal
effects
on
the
growth
of
young
fish
are
detectable
when
pH
drops
below
6.5.
Increased
H+
concentration
(
decreased
pH)
has
been
shown
to
be
directly
toxic
itself,
but
perhaps
more
importantly
in
nature,
increased
acidity
mobilizes
aluminum,
which
is
toxic
under
acidic
conditions,
even
though
it
is
harmless
under
acid
neutral
conditions.
Animals
typically
tolerate
lower
pH
values
in
the
absence
of
aluminum.
Aluminum
is
the
most
abundant
metal
in
the
Earth's
crust,
and
is
virtually
ubiquitous
in
terrestrial
environments.
The
aqueous
chemistry
of
aluminum
is
complex,
and
it
can
exist
in
multiple
chemical
forms
differing
in
toxicity
in
solution;
inorganic
monomeric
aluminum
is
thought
to
be
the
most
toxic
form.
Organically
bound
aluminum
is
relatively
nontoxic,
so
waters
with
high
organic
content
usually
contain
little
toxic
aluminum.
Concentrations
of
inorganic
monomeric
aluminum
above
30B50

g/
L
cause
adverse
effects
in
the
most
sensitive
organisms;
concentrations
over
100

g/
L
affect
many
organisms.

Many
aquatic
animals
are
more
sensitive
to
acid
conditions
when
calcium
concentrations
are
low.
Calcium
ameliorates
the
negative
effects
of
acid
conditions
by
directly
supporting
the
physiological
processes
and
structures
damaged
by
H+
and
aluminum
in
fish.
Small
changes
in
calcium
concentration
can
produce
substantial
changes
in
response;
most
of
the
benefits
of
increased
calcium
concentration
are
evident
at
concentrations
of
39
150

eq/
L,
and
further
increases
produce
smaller
improvements.

3.3.1.2
Effects
on
Species
Richness.
A
number
of
the
species
that
commonly
occur
in
surface
waters
susceptible
to
acidic
deposition
cannot
survive,
reproduce,
or
compete
in
acidic
waters.
Thus,
with
increasing
acidity,
the
"
acid­
sensitive"
species
are
lost
and
species
richness
(
the
number
of
species
living
in
a
given
lake
or
stream)
declines.
These
changes
in
aquatic
community
structure
occur
at
chronic
pH
levels
<
6.0
B
6.5.

Acid­
sensitive
species
occur
in
all
groups
of
aquatic
organisms.
The
drop
in
species
richness
is
most
dramatic
between
pH
5
and
6.
For
example,
many
species
of
minnows,
zooplankton,
mollusks,
and
mayflies
are
adversely
affected
at
chronic
pH
levels
between
5.5
and
6.
Long­
term
declines
in
pH
from
5.5
to
5.0
result
in
the
loss
of
several
important
sport
fish
species,
including
lake
trout,
rainbow
trout,
walleye,
and
smallmouth
bass.
Brook
trout
generally
cannot
survive
or
reproduce
successfully
at
pH
levels
below
4.8B5.2.
Few
fish
are
found
at
pH
levels
below
4.5.
Both
chronic
and
acute
(
episodic)
acidification
contribute
to
species
loss;
episodic
acidification
is
more
pronounced
in
streams
than
in
lakes.

3.3.1.3
Effects
on
Ecosystem
Level
Processes.
Ecosystem
level
processes,
such
as
decomposition,
nutrient
cycling,
and
productivity
are
fairly
robust
and
are
affected
only
at
relatively
high
levels
of
acidity
(
e.
g.,
chronic
pH
<
5.0
B
5.5).

Among
nonvertebrate
organisms,
acidsensitive
species
lost
in
the
early
stages
of
acidification
are
"
replaced"
to
some
degree
by
acid­
tolerant
organisms;
as
a
result,
the
net
change
in
total
community
productivity
is
relatively
small;
however,
less
information
is
available
for
system­
level
processes
than
for
community
structure,
so
subtle
effects
on
system
processes
may
not
yet
have
been
detected.

3.3.1.4
Recovery
of
Biological
Communities
Relatively
few
studies
have
been
conducted
on
the
recovery
of
biological
communities
following
reductions
of
acid
inputs.
Nevertheless,
it
is
predicted
that,
with
decreasing
acidity,
acid­
sensitive
species
would
reappear
and
species
richness
would
increase.

Much
of
the
information
on
recovery
is
derived
from
liming
experiments,
involving
the
addition,
in
most
cases,
of
calcium
carbonate.
With
few
exceptions,
this
treatment
results
in
improved
water
quality,
to
levels
that
would
support
more
acid­
sensitive
species.
Problems
in
interpreting
these
results
remain.
Although
liming
adds
calcium,
decreases
in
acid
precipitation
may
reduce
surface
water
calcium.
Preliminary
results
from
Norway
suggest
that
fish
community
declines
have
continued
in
the
last
10B15
years
in
regions
where
acid
deposition
has
decreased.
In
these
low
calcium
waters
(
50B100

g/
L),
small
decreases
in
calcium
have
apparently
offset
the
benefits
of
lower
acid
and
aluminum
concentrations.
In
contrast,
preliminary
results
from
Ontario
indicate
improved
biological
status
following
reductions
in
atmospheric
loading;
calcium
levels
there
tend
to
be
higher
than
in
Norway.
Rates
of
recovery
are
poorly
known
in
general,
but
algae
appear
to
recover
more
rapidly
than
other
groups
of
organisms.
Additional
research
on
recovery
(
and
management
steps
to
accelerate
it)
is
much
needed.

3.3.1.5
Developing
Regional
Models.
Laboratory
toxicity
experiments
and
field
surveys
provide
an
adequate
basis
for
quantifying
the
relationship,
on
a
regional
scale,
between
changes
in
pH,
aluminum,
and
calcium
and
acidity­
induced
stress
on
fish
populations.
Thus,
toxicity­
based
models,
field­
based
models,
and
models
that
combine
40
laboratory
and
field
data
can
be
used
to
evaluate
the
biological
significance
of
projected
changes
in
acid­
base
chemistry,
given
alternate
deposition
and
emission
scenarios.

Regional
assessments
of
the
effects
of
surface
water
acidification
on
fish
require
two
components:
the
expected
distribution
of
fish
in
the
area
in
the
absence
of
acidification
effects
(
baseline
status)
and
the
ways
to
predict
changes
in
the
fish
community
as
a
function
of
water
chemistry
changes
(
prediction
The
NAPAP
Integrated
Assessment
developed
two
procedures
for
determining
baseline
status:
(
1)
habitat
evaluations
based
on
fish
habitat
requirements
known
from
the
literature
and
(
2)
field­
based
empirical
models
that
rely
on
statistical
associations
between
fish
status
and
measured
habitat
characteristics
Three
procedures
have
been
developed
for
prediction.
The
first,
using
toxicity
models,
predicts
conditional
mortality
rates,
or
an
acidic
stress
index
(
ASI),
based
on
water
chemistry
and
fish
mortality
in
laboratory
bioassays.
The
second
procedure
uses
fieldbased
empirical
models,
relating
fish
status
to
water
chemistry.
In
the
third
procedure,
models
combining
toxicity
and
field
survey
data
were
developed.
Two
types
of
models
are
available,
Bayesian
models
and
the
Lake
Acidification
and
Fisheries
framework.

Development
of
baseline
status
is
often
more
difficult
and
less
precise
than
prediction.
The
approaches
for
the
NAPAP
Integrated
Assessment
vary
among
regions,
due
to
fish
data
availability.
Methods
for
conducting
regional
assessments
of
potential
effects
on
fish
in
Mid­
Appalachian
streams
include
habitat
evaluation
for
baseline
status,
and
both
toxicity
models
and
field­
based
models
for
prediction
of
acidification
effects.

3.3.1.6
Documentation
of
Effects
on
Fish.
The
loss
of
fish
populations
and/
or
absence
of
fish
species
as
a
result
of
acid­
base
chemistry
has
been
documented
for
some
lakes
and
streams
in
several
regions
of
the
United
States.
Applications
of
fish
response
models
suggest
that
the
percentage
of
NSWS
waters
with
acid­
base
chemistry
unsuitable
for
acidsensitive
fish
species
ranges
from
<
5%
in
the
Upper
Midwest
to
nearly
60%
for
upper
stream
reaches
in
the
Mid­
Atlantic
Coastal
Plain.
An
estimated
23%
of
the
Adirondack
lakes
and
18%
of
the
Mid­
Appalachian
streams
classified
as
potential
brook
trout
habitat
currently
have
acid­
base
chemistry
unsuitable
for
brook
trout
survival.

Good
quality
survey
data
on
the
regional
status
of
fish
communities
in
the
United
States
are
limited;
other
natural
and
human
factors
also
affect
fish
distributions.
As
a
result,
effects
caused
by
acidification
can
be
difficult
to
document.
Nevertheless,
intensive
studies
at
a
small
number
of
sites
in
Mid­
Appalachian
streams
have
documented
toxic
conditions
during
episodes,
fishkills,
and
loss
of
fish
populations
as
a
result
of
increasing
stream
acidity.
An
estimated
18%
of
potential
brook
trout
streams
in
the
Mid­
Appalachians
are
too
acid
for
brook
trout
survival;
in
about
30%
of
the
streams,
ASIs
indicate
that
conditions
are
too
acid
for
more
acid­
sensitive
species
that
might
be
expected
there.

In
the
southern
Blue
Ridge,
few
systems
currently
have
baseflow
pH
levels
detrimental
to
fish.
However,
these
streams
have
low
buffering
capacity,
and
are
vulnerable
during
increases
in
acidity,
especially
episodically.

Acidification
of
mountain
streams
in
the
SAMI
region
apparently
has
consequences
outside
the
SAMI
region.
Declines
in
anadromous
fish
stocks
may
be
explained
partly
by
acidification.
Nearly
60%
of
upper
reaches
of
streams
entering
the
Mid­
Atlantic
Coastal
Plain
are
unsuitable
for
acid­
sensitive
species.
Anadromous
fishes
(
those
that
are
spawned
in
freshwater,
spend
various
amounts
of
time
there,
migrate
to
seawater
where
they
attain
sexual
maturity,
and
then
return
to
freshwater
41
to
spawn)
have
been
shown
to
be
among
the
most
acid­
sensitive
fish
in
their
freshwater
phase.
These
species
include
Atlantic
salmon,
striped
bass,
and
blueback
herring;
all
show
acid­
induced
stress
at
pH
levels
of
6.0B6.5.
The
contribution
of
acidification
to
losses
of
these
species
is
difficult
to
evaluate
because
other
factors
such
as
overfishing
and
habitat
loss
also
affect
them.

3.3.2
Current
Biological
Impacts
in
the
Southern
Appalachians
The
NAPAP
synthesis
documents
(
J.
Baker
et
al.,
1990)
did
not
use
regional
boundaries
perfectly
coincident
with
the
SAMI
region.
However,
two
of
the
regions
overlap
with
the
SAMI
study
area
and
the
conclusions
of
the
NAPAP
study
are
applicable
to
surface
waters
in
the
SAMI
region.
The
NAPAP
mid­
Appalachians
region
covers
the
SAMI
study
area
in
Virginia
and
West
Virginia,
plus
most
of
Pennsylvania,
western
Maryland,
and
the
Catskill
Mountains
of
New
York.
The
NAPAP
Interior
Southeast
region
consists
of
parts
of
the
Piedmont
and
the
SAMI
study
region
(
Figure
1)
south
of
the
Virginia/
North
Carolina,
and
Kentucky/
Tennessee
state
lines.

3.3.2.1
Mid­
Appalachians.
In
the
mountains
of
western
Virginia,
the
high­
elevation,
low
ANC
headwater
streams
in
the
area
have
fish
assemblages
dominated
by
eastern
brook
trout,
the
only
salmonid
native
to
the
region
(
Cosby
et
al.,
1991).
Stocking
with
brown
and
rainbow
trout
is
partly
responsible
for
displacing
brook
trout
from
low­
elevation
reaches,
but
it
has
not
displaced
brook
trout
from
habitats
at
higher
elevations.
The
reaches
farthest
upstream
contain
brook
trout
or
no
fish
at
all.
Farther
downstream,
mottled
sculpins
and
blacknose
dace
co­
occur
with
brook
trout.

The
extent
of
aquatic
biological
responses
to
acidic
deposition
has
not
been
studied
region­
wide
in
the
Mid­
Appalachians.
A
study
in
western
Maryland
measured
aluminum,
calcium,
pH,
and
fish
abundance
in
79
trout
streams
(
Morgan
et
al.,
1990).
The
mean
pH
of
streams
with
fish
was
7.1,
compared
with
5.4
for
fishless
streams;
no
brook
trout
were
found
in
streams
with
baseflow
pH
<
5.5.
In
situ
bioassays
on
brook
trout
and
other
fish
species
in
the
Mid­
Appalachians
have
demonstrated
adverse
affects,
including
altered
spawning
behavior,
reduced
egg
viability,
decreased
hatching
rate,
reduced
survival,
and
reduced
growth.

The
current
status
and
trends
regarding
region­
wide
acidification
effects
in
the
mid­
Appalachians
is
difficult
to
assess
because
of
the
absence
of
large­
scale
fish
surveys.
In
the
SAMI
region,
losses
of
fish
species
associated
with
increasing
stream
acidity
have
been
reported
in
Virginia
(
Little
Stony
Creek
and
St.
Mary's
River;
Webb
et
al.,
1989a,
b)
and
in
West
Virginia
(
Cranberry
River
drainage;
Zurbuch
et
al.,
1986).
In
Pennsylvania,
a
study
of
61
streams
in
the
Laurel
Hills
region
showed
that
16%
of
the
sites
were
fishless.
Fishless
streams
had
significantly
lower
pH
values
than
streams
with
fish
and
they
did
not
have
acid
mine
drainage
impacts.
All
the
fishless
streams
were
in
largely
undisturbed
forests
with
base­
poor
bedrock
types
(
Sharpe
et
al.,
1987).
Also
in
Pennsylvania,
fish
kills
due
to
pH
depressions
during
rainfall
events
have
been
confirmed
in
some
Appalachian
Plateau
streams
Wigington
et
al.,
1993).

A
three­
year
project
on
the
effects
of
acidbase
chemistry
on
fish
communities
in
mountain
streams
in
Virginia
was
started
in
Shenandoah
National
Park
in
1992
(
Bulger
et
al.,
1995).
Both
chronic
and
episodic
acidification
are
occurring
in
these
streams.
Biological
differences
in
low
ANC
versus
high
ANC
streams
include
fish
species
richness,
population
density,
condition
factor,
age,
size,
and
bioassay
survival.
In
particular,
both
episodic
and
chronic
mortality
occurred
in
young
brook
trout
exposed
in
low
ANC
streams,
but
not
in
high
ANC
streams
42
(
MacAvoy
and
Bulger,
1995),
and
blacknose
dace
in
low
ANC
streams
were
in
poor
condition
relative
to
dace
in
high
ANC
streams
(
Dennis
and
Bulger,
1995;
Dennis
et
al.,
1995).
In
the
same
study,
blacknose
dace
and
brook
trout
from
a
low
ANC
stream
were
able
to
detect
and
avoid
acid
pulses
simulating
an
acid
episode
in
the
laboratory;
they
could
also
find
neutral
pH
refugia
(
Newman
and
Dolloff,
1995).
Predictive
models
relating
fish
status
to
future
water
chemistry
are
to
be
produced.

Dennis
(
1995)
lists
26
species
of
fish
found
in
Shenandoah
National
Park
(
based
on
the
park's
Fishery
Management
Plan).
Ranges
of
critical
pH
for
nine
of
these
species
were
reported
by
Baker
and
Christensen
(
1991),
who
estimated
average
pH
thresholds
for
a
variety
of
negative
effects
observed
in
several
studies
for
each
species.
Paine
Run,
which
hosts
three
fish
species,
has
been
studied
intensively
since
1992
as
part
of
the
park's
Fish
in
Sensitive
Habitats
(
FISH)
Project
(
Bulger
et
al.,
1995).
Episodic
pH
values
within
the
critical
range
of
all
nine
species
have
been
recorded
in
this
stream,
and
Paine
Run's
baseflow
pH
(
5.7B6.0)
is
within
the
critical
pH
range
for
two
of
the
nine
species
whose
critical
pH
range
is
reported
by
Baker
and
Christensen
(
1991).
One
of
these
two
species,
blacknose
dace
(
critical
pH
range:
5.6B6.2),
occurs
in
Paine
Run,
but
individuals
from
that
stream
are
significantly
smaller
than
blacknose
dace
from
higher
ANC
streams
(
Dennis
and
Bulger,
1995).
Indeed,
pH
as
low
as
4.11
has
been
recorded
for
a
small
tributary
of
Paine
Run
(
Dennis,
1995).
It
seems
clear
that
elements
of
the
fish
community
of
Shenandoah
Park
are
vulnerable
to
habitat
loss
resulting
from
future
decreases
in
stream
pH.

The
NAPAP
SOS/
T
report
used
two
approaches
to
assess
acidification
effects
in
the
Mid­
Appalachian
region:
(
1)
ASI
values
calculated
from
pH,
calcium,
and
inorganic
monomeric
aluminum
using
NSS
index
chemistry
(
see
Section
5.3.3)
and
(
2)
a
brook
trout
presence/
absence
model
based
on
streams
in
western
Maryland
(
Morgan
et
al.,
1990)
and
southeastern
Pennsylvania
(
Sharpe
et
al.,
1987).

There
are
about
63,000
km
of
total
stream
length
in
the
NSS­
I
target
population
in
the
Mid­
Appalachian
region.
About
24%
(
15,000
km)
of
the
NSS­
I
target
population
stream
length
exhibits
sensitive
fish
ASI
values
>
10,
indicating
chemistry
conditions
unsuitable
for
acid­
sensitive
fish
species.
The
highest
ASI
values
occur
in
the
western
part
of
the
region.

Potential
brook
trout
habitat
in
the
region
was
defined
by
the
following
criteria:
elevation
>
400
m,
stream
gradient
0.4B17%,
and
Strahler
stream
order
(
1:
24,000­
scale
map)
<
4.
The
entire
mid­
Appalachian
region
is
within
the
zoogeographic
range
of
brook
trout,
so
all
streams
meeting
the
criteria
were
assumed
to
be
potential
brook
trout
habitat.
About
37%
(
23,000
km)
of
the
NSS­
I
target
population
can
be
considered
potential
brook
trout
habitat.
Of
these,
18%
(
4,100
km)
have
ASI­
sensitive
values
>
30,
indicating
conditions
unsuitable
for
brook
trout.

3.3.2.2
Interior
Southeast
Streams.
In
Great
Smoky
Mountains
National
Park,
fish
surveys
conducted
periodically
since
the
1930s
indicate
a
steady
decline
in
brook
trout
range;
similar
declines
in
brook
trout
range
have
occurred
elsewhere
in
the
SAMI
region.
The
reasons
include
habitat
loss
through
logging,
overfishing,
and
competition
from
introduced
brown
and
rainbow
trouts.
Acidification
is
not
considered
a
major
factor
yet,
though
brook
trout
do
not
occur
in
streams
whose
pH
values
approach
5.0.
The
only
documented
fish
kills
in
the
area
associated
with
stream
acidity
have
occurred
at
fish
rearing
facilities;
these
incidents
involved
introduced
rainbow
and
brown
trout,
which
are
more
acid
sensitive
than
brook
trout.

No
adverse
effects
of
acidic
deposition
on
biota
have
been
demonstrated
conclusively
so
43
far
in
the
southern
Blue
Ridge
province
(
SBRP),
except
in
fish
hatcheries
supplied
by
Raven
Fork
(
North
Carolina).
Several
fish
kills
of
brown
trout
and
rainbow
trout
in
holding
tanks
supplied
with
water
from
Raven
Fork
have
occurred
during
storms
(
Jones
et
al.,
1983).
Stream
acidification
was
implicated
as
the
cause
of
the
fish
kills
based
on
observed
streamwater
pH/
aluminum
levels
and
the
lack
of
trout
mortality
in
limed
streamwater.
Elsewhere
in
the
region,
the
relationship
established
elsewhere
between
low
pH
and
species
diversity
indicates
that
any
decreases
in
streamwater
pH
may
produce
decreases
in
species
richness.
As
in
the
mountains
of
Virginia,
fish
assemblages
in
the
low­
order,
high­
elevation
streams
at
risk
are
dominated
by
trout
species,
especially
brook
trout
and
rainbow
trout,
joined
(
moving
downstream)
by
mottled
sculpin
and
blacknose
dace,
creek
chub
and
longnose
dace,
then
assemblages
of
introduced
salmonids,
minnows,
suckers,
and
darters.
Aquatic
invertebrate
species
richness
and
pH
are
positively
correlated
in
this
province,
suggesting
that
declines
in
pH
would
result
in
declines
in
benthic
macroinvertebrate
diversity
(
Baker
and
Christensen,
1991).

Rosemond
et
al.
(
1992)
report
strong
relationships
between
measures
of
benthic
invertebrate
community
status
and
water
chemistry
in
Great
Smoky
Mountains
National
Park.
Baseflow
pH
values
were
4.5B6.8,
and
inorganic
monomeric
aluminum
was
3B197

g/
L.
Total
invertebrate
density
(
excluding
the
acid­
tolerant
chironomids)
and
species
richness
were
higher
in
the
high
pH
streams;
these
effects
were
attributed
to
direct
effects
on
invertebrate
survival
rather
than
on
food
availability.

Assessment
of
biological
effects
is
limited
to
ASI
toxicity
models
(
see
Section
5.3.3)
because
no
region­
wide
surveys
of
chemistry
and
fish
status
are
available.
The
highest
acidic
stress
levels
occur
in
the
high­
elevation
areas
of
the
southern
Blue
Ridge.
Extreme
acidic
stress
levels
(
ASI­
sensitive
values
>
50)
occur
much
less
frequently
than
in
other
regions.
About
24%
(
63,000
km)
of
the
NSS­
I
target
population
stream
length
has
ASI­
sensitive
values
over
10,
indicating
conditions
unsuitable
for
acid­
sensitive
species.
Much
of
the
acid­
related
stress
in
this
region
results
from
the
very
low
calcium
levels
which
are
typical
of
regional
geology
classes.

Potential
brook
trout
habitat
in
the
region
was
defined
by
the
following
criteria:
elevation
>
1000
m,
stream
gradient
0.4B17%,
and
Strahler
stream
order
(
1:
24,000­
scale
map)
<
4.
These
criteria
include
only
streams
within
the
Southern
Blue
Ridge
subregion,
but
all
are
within
the
zoogeographic
range
of
brook
trout.
About
46%
(
24,000
km)
of
the
NSS­
I
target
population
can
be
considered
brook
trout
habitat;
of
this
percentage,
11.5%
(
1,400
km)
has
ASI­
sensitive
values
>
30.
Perhaps
10%
of
streams
otherwise
suitable
for
brook
trout
are
unavailable
because
of
acidification
this
estimate,
however,
contains
much
uncertainty.
We
do
not
have
sufficient
information
to
construct
probability
of
presence
models.
44
4.
RECENT
TRENDS
IN
ACIDIFICATION
IN
THE
SOUTHERN
APPALACHIANS
4.1
Trends
in
Surface
Water
Chemistry
Within
the
SAMI
Class
I
areas,
temporal
trend
information
is
available
only
for
streams
in
Shenandoah
National
Park.
In
the
late
1980s,
White
Oak
Run
and
Deep
Run
both
showed
significant
increases
in
SO
4
2

and
decreases
in
pH
(
Ryan
et
al.,
1989).
At
Coweeta,
in
the
North
Carolina
Blue
Ridge,
there
were
no
significant
trends
in
acidity
but
there
was
a
significant
increase
in
SO
4
2

concentration
(
average
of
0.7

eq/
L/
yr)
in
all
control
watersheds
between
1974
and
1982
(
Swank
and
Waide,
1988).
The
SO
4
2

increases
at
these
sites
are
consistent
with
a
gradual
saturation
of
soils
in
the
region
with
SO
4
2

from
deposition,
and
have
been
predicted
by
acidification
models
(
Church
et
al.,
1992).
More
recently,
the
SO
4
2

trends
in
Shenandoah
National
Park
have
been
greatly
altered
as
a
result
of
forest
defoliation
by
gypsy
moth
larvae
(
Webb
et
al.,
1995).
Defoliation
has
resulted
in
large
increases
in
streamwater
NO
3

(
up
to
60

eq/
L),
decreases
in
SO
4
2

,
and
little
change
in
ANC
or
pH
at
baseflow.
During
storms,
however,
the
increased
leaching
of
NO
3

in
these
watersheds
has
led
to
an
increase
in
episodic
acidification,
with
ANC
decreases
during
episodes
nearly
twice
the
magnitude
previously
observed
(
Eshleman
et
al.,
1995).
We
do
not
currently
know
how
the
streams
will
respond
as
the
forests
recover
from
defoliation.

It
is
likely
that
many
parts
of
the
SAMI
region
have
undergone
increases
in
NO
3

over
the
past
several
decades.
There
are
few
data
to
verify
this
pattern,
but
it
is
supported
by
the
few
time­
series
data
that
do
exist,
and
by
the
current
patterns
of
streamwater
NO
3

distribution
in
the
region.
At
Fernow
Experimental
Forest
in
West
Virginia,
NO
3

concentrations
have
increased
from
near
zero
to
50B60

eq/
L
at
baseflow
from
1970
to
the
present
(
Stoddard,
1994).
In
the
Great
Smoky
Mountains
National
Park,
NO
3

shows
strong
correlations
with
elevation
and
forest
age
(
Cook
et
al.,
1994;
Flum
and
Nodvin,
1995),
with
the
highest
concentrations
(
up
to
100

eq/
L)
occurring
at
high
elevations
(
where
deposition
is
highest;
Shubzda
et
al.,
1995)
and
in
areas
of
old­
growth
forest
where
biological
demand
for
nitrogen
is
lowest
(
Stoddard,
1994;
Nodvin
et
al.,
1995).
It
is
very
likely
that
a
combination
of
high
rates
of
N
deposition,
coupled
with
forest
maturation
in
the
park,
have
led
to
accelerated
rates
of
N
loss
in
the
Smokies,
and
that
this
trend
will
continue
as
maturation
of
the
forests
at
lower
elevation
progresses.

We
do
not
know
what
effect
the
presumed
increase
in
streamwater
NO
3

has
had
in
the
Smokies,
but
it
is
likely
to
have
led
to
substantial
cation
depletion
and
ultimately
to
soil
acidification
in
the
region
(
Johnson
and
Lindberg,
1992).
Many
of
the
same
streams
that
now
show
elevated
NO
3

concentrations
are
either
chronically
or
episodically
acidic
(
Flum
and
Nodvin,
1995;
Nodvin
et
al.,
1995).

4.2
Trends
in
Episodic
Effects
It
has
been
proposed
that
episodic
changes
in
surface
water
pH
during
stormflow
conditions
may
represent
an
"
early
warning"
of
sustained,
chronic
acidification
impacts
associated
with
regional­
scale
acidic
deposition.
However,
only
a
few
experimental
or
modeling
studies
of
long­
term
changes
in
episodic
conditions
have
been
conducted,
due
to
a
general
lack
of
data
with
which
to
statistically
evaluate
trends
or
to
parameterize
an
appropriate
predictive
acidification
model
(
Neal
et
al.,
1992;
Hooper
and
Christophersen,
1992;
Eshleman
et
al.,
1995).
One
of
these
studies
45
(
Eshleman
et
al.,
1995)
was
conducted
within
the
SAMI
region
in
Shenandoah
National
Park.

Two
modeling
studies
have
been
conducted
using
the
predictive
acidification
model,
MAGIC
(
see
section
5.1.2.2).
The
study
by
Neal
et
al.
(
1992)
used
MAGIC
in
two­
component
mode
(
i.
e.,
MAGIC
was
calibrated
for
two
distinct
soil
horizons
and
the
resulting
solutions
were
mixed
according
to
known
proportions
based
on
chemical
hydrograph
separations).
The
technique
was
applied
to
the
Afon
Gwy
catchment
in
mid­
Wales,
with
results
demonstrated
in
the
forms
of
(
1)
3­
month
sequences
of
hydrogen
ion
and
inorganic
aluminum
concentrations
and
(
2)
chemical
duration
curves
for
hydrogen
ion
and
aluminum.
The
most
important
finding
in
the
study
was
that
aluminum
concentrations
in
the
stream
did
not
recover
as
rapidly
as
had
previously
been
thought,
in
response
to
reductions
in
sulfur
deposition.
The
second
modeling
study,
by
Hooper
and
Christophersen
(
1992),
of
long­
term
episodic
changes
in
stream
chemistry
at
Panola
Mountain
(
Georgia)
largely
supported
the
hypothesis
that
stormflow
conditions
provide
an
"
early
warning";
while
baseflow
chemistry
gradually
became
acidic
over
a
50­
year
modeling
period,
acidification
of
two
upper
soil
layers
occurred
more
rapidly,
causing
acidic
conditions
during
stormflow
periods.
Although
this
work
was
located
geographically
outside
the
SAMI
region,
the
results
of
Hooper
and
Christophersen
(
1992)
appear
to
provide
a
reasonable
hypothesis
for
long­
term
changes
in
episodic
conditions
in
other
southern
Appalachian
mountain
waters.

The
analytical
study
of
Eshleman
et
al.
(
1995)
involved
the
use
of
a
hydrological
separation
program
and
long­
term
daily
flow
and
weekly
chemical
data
from
White
Oak
Run,
Virginia.
A
statistical
analysis
of
13
years
of
daily
discharge
data
and
weekly
streamwater
composition
data
for
White
Oak
Run
in
Shenandoah
National
Park
was
performed
in
order
to
quantify
episodic
changes
in
composition
and
to
identify
long­
term
trends
in
episodic
acidification
attributable
to
both
natural
and
anthropogenic
processes.
An
objective
hydrological
separation
technique
was
used
to
identify
more
than
100
"
stormflow
baseflow
pairs"
in
the
database,
from
which
episodic
chemical
changes
could
be
quantified.
Univariate
statistical
analysis
suggested
that
mean
episodic
depressions
of
ANC
in
White
Oak
Run
have
increased
by
about
a
factor
of
2
since
the
first
outbreak
of
forest
defoliation
by
the
gypsy
moth
caterpillar
during
the
summer
of
1990;
in
addition,
the
mean
episodic
change
in
NO
3

concentration
has
increased
by
about
12

eq/
L,
while
the
mean
episodic
dilution
of
C
B
has
decreased
from

8.5

eq/
L
to

1.7

eq/
L
during
the
same
period.
Episodic
changes
in
SO
4
2

have
remained
the
same,
however.
The
results
indicate
that
natural
processes
such
as
insect
defoliations
can
contribute
to
episodic
acidification
through
mobilization
of
NO
3

.
Results
from
the
study
did
not
demonstrate
any
longterm
changes
in
episodic
conditions
associated
with
atmospheric
deposition,
presumably
due
to
the
larger,
overwhelming
influence
of
the
gypsy
moth
defoliations.
Other
experimental
and
modeling
studies
of
long­
term
trends
in
episodic
conditions
are
needed
for
the
SAMI
region.

4.3
Trends
in
Biological
Effects
It
is
difficult
to
quantify
trends
in
biological
condition
because
few
datasets
have
good
biological
data
over
time.
It
is
also
difficult
to
differentiate
the
effects
of
acidic
deposition
from
other
anthropogenic
effects.
Thus,
it
is
not
possible
to
make
a
definitive
statement
about
trends
in
biological
effects
related
to
acidic
deposition
in
the
SAMI
region.
There
is
at
least
one
stream
in
the
area,
however,
in
which
fish
declines
are
probably
the
result
of
acidification:
the
St.
Mary's
River
in
George
46
Washington
National
Forest
in
the
Virginia
Blue
Ridge.
Surveys
in
1988,
compared
to
those
in
the
1930s,
showed
declines
in
acid­
sensitive
fish
and
invertebrates,
coincident
with
lower
pH
values
(
Webb
et
al,
1989a,
b).
47
5.
METHODOLOGIES
FOR
PREDICTING
FUTURE
EFFECTS
OF
ACIDIC
DEPOSITION
5.1
Surface
Water
Chemical
Models
A
large
number
of
models
have
been
developed
over
the
years
to
predict
and/
or
explain
surface
water
chemical
response
to
acidic
deposition.
These
models
fall
into
two
general
categories:
steady
state,
and
dynamic.
Most
of
these
models
were
comprehensively
reviewed
and
cited
in
the
NAPAP
SOS/
T
Report
#
14
(
Thornton
et
al.,
1990)
and
the
next
two
subsections
briefly
summarize
those
findings.
Most
of
the
NAPAP
modeling
effort
focused
on
sulfur
chemistry.
It
has
become
more
apparent
over
the
last
5
years,
however,
that
NO
3

is
increasing
in
some
areas
and
needs
to
be
treated
more
explicitly.
Thus,
a
number
of
modeling
efforts
have
recently
been
developed
to
incorporate
N
dynamics
in
future
projections
of
surface
water
acid­
base
chemistry.

5.1.1
Steady­
state
Models
Steady
state
refers
to
the
condition
in
which
outputs
equal
inputs.
In
other
words,
after
a
perturbation
such
as
acidic
deposition,
lake
or
stream
chemistry
(
output)
will
eventually
reach
an
equilibrium
with
the
new
chemical
input
levels.
In
general,
steady­
state
models
use
current
water
chemistry
to
estimate
steady­
state
conditions
for
both
preindustrial
conditions
and
the
eventual
future
condition
for
a
given
level
of
acidic
deposition.
Steadystate
models
have
no
explicit
time
component;
the
lake
or
stream
is
assumed
to
be
in
equilibrium
with
atmospherically
deposited
SO
4
2

in
both
the
past
and
future
scenarios.
Future
changes
in
ANC
are
then
usually
estimated
from
the
change
in
SO
4
2

either
empirically
or
by
using
a
charge­
balance
approach
that
calculates
ANC
decline
as
some
fraction
of
the
change
in
SO
4
2

.
This
fraction
has
been
named
the
Henriksen
F­
factor
(
after
its
inventor)
and
relates
the
proportion
of
added
SO
4
2

that
is
balanced
by
increased
base
cation
concentration.
Thus,
an
F=
1
means
that
all
the
additional
SO
4
2

is
balanced
with
additional
base
cations
(
no
ANC
or
pH
decline).
An
F=
0
indicates
that
no
base
cations
are
released
and
that
all
the
added
SO
4
2

is
balanced
against
a
decline
in
ANC
(
additional
H+
or
Aln+).
As
one
would
expect,
steady­
state
models
using
the
F­
factor
approach
are
very
sensitive
to
the
value
of
F
used
in
the
model,
and
various
approaches
have
been
used
to
estimate
F.
Thornton
et
al.
(
1990)
concluded
that
most
steady­
state
models
were
not
appropriate
for
use
in
the
NAPAP
assessment
because
they
had:
(
1)
unrealistic
assumptions
concerning
the
level
of
watershed
neutralization
(
F­
factor),
(
2)
a
lack
of
biological
relevance
(
no
aluminum
predictions),
and
(
3)
problems
in
applying
the
models
outside
the
regions
for
which
they
were
first
developed.

5.1.2
Dynamic
Models
Whereas
steady­
state
models
have
no
explicit
time
component,
dynamic
models
project
water
chemistry
for
specific
days,
months,
or
years.
Dynamic
models
integrate
our
current
understanding
about
the
hydrological
and
biogeochemical
processes
that
occur
as
acidic
deposition
falls
on
watersheds
and
is
transported
into
lakes
and
streams.
The
modeled
mechanisms
include
hydrologic
flow
routing,
soil­
water
interactions
(
including
soilwater
contact
time),
anion
retention,
base
cation
exchange,
mineral
weathering,
and
other
watershed
processes
(
e.
g.,
vegetative
uptake
and
organic
interactions).
The
mechanisms
are
modeled
with
varying
degrees
of
complexity
among
the
different
dynamic
models
but
they
are
all
are
much
more
complex
and
time
consuming
and
they
have
larger
data
input
requirements
than
steady­
state
48
models.
The
two
dynamic
watershed
models
of
most
use
to
an
assessment
of
the
aquatic
effects
of
acidic
deposition
in
the
SAMI
region
are
the
ILWAS
(
Integrated
Lake­
Watershed
Acidification
Study)
and
MAGIC
(
Model
of
Acidification
of
Groundwater
in
Catchments)
models.

5.1.2.1
ILWAS.
The
ILWAS
model
(
Chen
et
al.,
1983;
Gherini
et
al.,
1985)
was
developed
as
a
research
tool
to
further
understanding
about
the
processes
affecting
acidbase
chemistry
in
Adirondack
lakes.
It
is
a
process­
oriented
model
that
uses
both
equilibrium
and
rate­
limited
expressions
to
describe
the
mass
balances
for
acid­
base
chemistry.
ILWAS
has
three
modules:
canopy,
hydrologic/
soil,
and
within
lake.
ILWAS
is
quite
complex;
it
models
more
processes
with
more
compartments
(
e.
g.,
soil
layers)
than
the
other
dynamic
models.
Thus
it
has
greater
input
data
requirements
and
is
more
time
consuming
to
run
and
calibrate.
ILWAS
requires
daily
meteorological
data,
weekly
or
monthly
deposition
chemistry,
watershed
vegetation
types
and
coverage,
watershed
attributes,
physical/
chemical
soil
data,
lake/
stream
hydrological
data,
and
initial
water
chemistry.
All
in
all,
it
requires
specification
of
more
than
200
parameters,
coefficients
and
initial
conditions
for
model
calibration
Model
results
can
be
output
over
a
range
of
time
steps
from
daily
to
annual.

5.1.2.2
MAGIC.
MAGIC
(
Cosby
et
al.,
1985a,
b;
1986a,
b)
is
a
lumped
parameter
model
of
intermediate
complexity
developed
to
project
the
long­
term
effects
(
decades
to
centuries)
of
acidic
deposition
on
responses
in
average
annual
stream
or
lake
chemistry.
It
was
originally
developed
for
streams
in
Shenandoah
National
Park.
MAGIC
was
formulated
to
be
parsimonious
in
selecting
processes
for
inclusion
in
the
model.
It
assumes
that
only
a
few
key
processes
influence
the
long­
term
response
of
watersheds
to
acidic
deposition.
MAGIC
has
a
separate
hydrologic
flow
routing
component
that
runs
on
a
daily
time
step.
Hydrologic
data
are
averaged
and
the
rest
of
the
MAGIC
model
has
a
monthly
or
annual
time
step.
MAGIC
has
two
soil
layers
and
a
lake/
stream
component,
and
both
equilibrium
and
rate­
limited
expressions
are
used
to
describe
the
mass
balances
for
acid­
base
chemistry.
MAGIC
requires
annual
meteorological/
deposition
data,
watershed
attributes,
physical/
chemical
soil
data,
lake/
stream
hydrological
data,
and
initial
water
chemistry.
Model
results
can
be
output
seasonally
or
annually.
It
does
not
simulate
episodic
responses.

A
regional
MAGIC
model
has
been
developed
to
project
the
acid­
base
status
of
regional
stream/
lake
populations
rather
than
individual
sites
(
Hornberger
et
al.,
1986,
1987).
Regional
MAGIC
is
structurally
and
functionally
similar
to
MAGIC
for
individual
sites.
It
has
the
same
data
input
requirements.
The
major
difference
is
that
in
Regional
MAGIC,
inputs
and
outputs
are
in
the
form
of
regional
distributions,
not
individual
values.
Regional
MAGIC
is
also
calibrated
differently;
a
set
of
simulated
watersheds
is
generated
by
random
sampling
from
the
input
distribution
data
and
then
calibrated
to
fit
an
observed
distribution
of
water
chemistry
from
a
regional
survey.
It
is
important
to
note
that
the
output
projects
how
a
regional
distribution
will
change
over
time
and
reflects
simulated
watersheds.
It
cannot
be
used
to
project
conditions
for
any
specific
watershed
or
geographic
location
within
the
region.

5.1.3
Nitrogen
Models
Three
models
of
nitrogen
dynamics
are
being
used
currently,
or
are
being
developed,
to
assess
the
acid­
base
status
of
surface
waters.
One
of
the
models
(
MAGIC­
WAND)
is
incorporated
directly
into
the
MAGIC
model.
The
others
provide
input
that
can
be
used
to
set
up
the
MAGIC
model
to
run
with
49
the
influences
of
N
deposition
included.
Each
model
is
described
briefly
here.

5.1.3.1
Model
of
Acidification
of
Groundwater
in
Catchments
B
With
Aggregated
Nitrogen
Dynamics
(
MAGIC­
WAND).
MAGIC­
WAND
is
a
direct
modification
of
the
MAGIC
model
that
includes
some
simple
representations
of
the
N
cycle
(
Cosby,
pers.
comm.).
Nitrogen
is
assumed
to
be
present
only
in
solution
in
soil
water.
Nitrogen
inputs
to
the
system
are
in
the
form
of
inorganic
N
added
to
the
soil
solution,
and
are
represented
as
atmospheric
deposition
and
mineralization.
The
rates
for
these
inputs
must
be
entered
a
priori.
Nitrogen
losses
from
the
system
are
represented
as
hydrologic
runoff
and
denitrification
Transformations
included
in
the
model
are
nitrification
and
uptake.
The
primary
limitation
of
MAGIC­
WAND
is
that
it
includes
no
internal
feedbacks
(
e.
g.,
mineralization
rates
must
be
entered
a
priori,
and
do
not
depend
on
changes
in
other
processes
or
on
N
pools).
Its
major
strength
is
that
it
is
directly
incorporated
into
MAGIC,
so
that
direct
outputs
of
the
variables
of
interest
(
e.
g.,
time
series
of
ANC)
can
be
made.

5.1.3.2
Model
of
Ecosystem
Retention
and
Loss
of
Inorganic
Nitrogen
(
MERLIN).
MERLIN
is
a
stand
alone
nitrogen­
cycling
model,
adapted
from
the
Generalized
Ecosystem
Model
(
GEM;
Rastetter
et
al.,
1991).
It
is
probably
the
most
realistic
of
the
three
models
available,
and
therefore
also
requires
the
most
detailed
inputs
(
Cosby,
pers.
comm.).
The
processes
in
MAGIC­
WAND
are
included
in
MERLIN,
not
as
driving
variables,
but
calculated
from
internal
state
variables.
The
carbon
and
N
pools
of
the
terrestrial
compartments
(
photosynthetic
biomass,
wood,
litter,
labile
organic
matter,
refractory
organic
matter,
etc.)
are
modeled
explicitly
and
the
carbon:
nitrogen
ratios
in
these
pools
feed
back
and
affect
the
N
transformation
processes.
Data
on
the
carbon
and
N
pool
sizes
in
each
compartment,
plus
all
the
fluxes
between
compartments,
are
needed
for
calibration.
The
realistic
nature
of
this
model
(
particularly
the
division
of
soil
organic
matter
into
labile
and
refractory
pools)
is
clearly
advantageous.
Its
major
drawback
is
the
detailed
input
data
required
to
run
the
model;
many
of
the
transformation
rates
(
e.
g.,
into
and
out
of
the
soil
organic
matter
pools)
are
not
known
for
many
sites.
This
is
still
a
rapidly
developing
model,
and
the
results
of
its
application
to
any
existing
sites
are
not
yet
available.

5.1.3.3
Net
Photosynthesis
and
Evapo­
Transpiration
(
PnET)
B
Carbon
and
Nitrogen
(
CN)
Model.
PnET­
CN
(
Aber
and
Federer,
1992)
has
evolved
from
a
relatively
simple
model
of
gross
and
net
photosynthesis
to
one
that
now
includes
cycles
of
water,
carbon,
and
nitrogen
in
a
monthly
time
step
(
Aber
et
al.,
in
press).
Because
it
evolved
from
a
"
treegrowing
model,"
it
focuses
on
the
effects
of
forest
development,
and
the
terrestrial
processes
that
accompany
it,
on
N
dynamics.
The
latest
version
of
the
model
includes
litter
decomposition
and
turnover,
as
well
as
transfers
between
terrestrial
compartments
(
roots,
foliage,
wood,
and
soil),
and
incorporates
site
history
(
land
use
disturbance)
into
the
transformation
rates
between
compartments.
The
model
has
been
applied
to
a
large
number
of
sites
with
remarkable
success,
especially
considering
that
it
is
not
truly
"
calibrated"
(
e.
g.,
the
model
parameters
are
not
tuned
so
that
they
reproduce
current
conditions
before
the
model
is
run
to
predict
future
conditions)
to
each
site.
For
example,
it
accurately
reproduces
the
record
of
NO
3

outputs
from
the
Hubbard
Brook
control
watershed
for
the
last
three
decades
(
Aber
et
al.,
in
press).
Although
the
model
requires
a
large
number
of
parameters
to
run,
it
has
been
very
successfully
applied
using
literature
values
for
different
forest
types
(
e.
g.,
hardwoods
vs.
conifers).
The
key
site­
specific
variables
needed
to
run
the
model
successfully
are
climatic
variables
50
(
temperature
and
precipitation),
foliar
N
content,
and
land
use
history.

5.2
Episodic
Chemical
Models
The
1990
NAPAP
assessment
of
current
and
future
acidification
impacts
of
acidic
deposition
recognized
that
assessments
based
solely
on
chemical
conditions
during
"
index
periods"
(
i.
e.,
periods
that
approximate
average
annual
or
seasonal
conditions)
did
not
account
for
the
"
worst­
case"
chemical
conditions
that
usually
occur
during
extreme
hydrological
events.
However,
because
of
the
rather
large
uncertainties
associated
with
incorporating
episodic
acidification
into
the
regional
assessment
framework,
model­
based
projections
of
watershed
responses
to
emissions
reduction
scenarios
were
not
formulated
to
account
for
changes
in
episodic
acidification
over
long
time
periods.

Since
completion
of
the
1990
NAPAP
assessment,
several
modeling
approaches
have
been
utilized
to
predict
future
episodic
chemical
responses
to
a
variety
of
emissions/
deposition
reduction
scenarios
(
Neal
et
al.,
1992;
Hooper
and
Christophersen,
1992;
Eshleman
et
al.,
1995),
although
none
has
yet
been
applied
to
a
watershed
in
the
SAMI
region.
Two
of
the
published
studies
were
conducted
on
individual
watersheds
and
utilized
the
MAGIC
model,
while
one
of
the
studies
was
conducted
at
the
regional
scale
and
utilized
a
regional
modeling
approach.
Depending
upon
the
nature
of
the
integrated
assessment
desired
by
SAMI
(
local
scale
or
regional
scale),
any
one
of
these
three
approaches
could
be
employed
to
address
the
issue
of
episodic
changes
in
surface
water
acid­
base
status
resulting
from
changes
in
deposition
loadings.
Two
key
issues
that
will
need
to
be
addressed
before
conducting
such
an
assessment
are:
(
1)
the
assumptions
inherent
in
each
of
the
modeling
approaches
and
(
2)
the
availability
of
watershed
and
surface
water
data
for
model
calibration
and
testing.

5.3
Modeling
the
Effects
of
Acidic
Deposition
on
Fish
Communities
The
primary
goal
of
the
mathematical/
statistical
models
that
evaluate
the
effects
of
acidic
deposition
on
fish
communities
is
to
predict
future
effects
on
biological
resources
based
on
projected
changes
in
water
chemistry
Few
models
for
regional
assessment
of
acidification
have
been
developed,
and
most
deal
only
with
fish
responses;
most
of
these
deal
with
fish
species
presence
or
absence,
but
more
sophisticated
response
variables
are
increasingly
being
modeled.
This
report
considers
three
kinds
of
models:
field­
based
empirical
models,
toxicity
models,
and
models
that
combine
toxicity
(
laboratory)
data
with
field
observations.

5.3.1
Empirical
Models
Empirical
or
field­
based
models
are
based
on
associations
between
fish
status
and
water
chemistry
as
they
exist
in
nature,
and
are
typically
constructed
from
survey
data.
The
future
values
of
variables
that
turn
out
to
be
most
strongly
correlated
with
fish
status
are
estimated
with
the
model,
and
fish
status
is
predicted
from
its
association
with
those
variables.
These
models
make
two
assumptions
(
1)
the
systems
surveyed
are
in
steady
state,
and
(
2)
the
observed
associations
between
fish
status
and
water
chemistry
accurately
reflect
what
would
occur
over
time.

These
empirical
models
do
not
assume
that
the
variables
most
strongly
associated
with
fish
status
(
or
with
greatest
predictive
capacity)
in
a
given
data
set
directly
control
fish
status
in
a
mechanistic
way,
although
the
mechanistic
linkages
are
often
easy
to
imagine.
For
example,
ANC
often
shows
a
strong
relationship
to
fish
status,
but
it
is
clear
that
fish
do
not
respond
directly
to
ANC;
however,
ANC
controls
pH,
which
controls
aluminum,
51
which
is
toxic.
Likewise,
variables
that
have
well­
understood
mechanistic
effects
on
fish
do
not
always
show
strong
empirical
relationships
with
fish
status
in
survey
data.
An
example
is
calcium,
which
ameliorates
acid
stress;
it
is
not
always
a
good
predictor
of
fish
status,
because
healthy
populations
exist
at
high
as
well
as
low
calcium
concentrations,
in
the
absence
of
acid
stress;
in
the
presence
of
acid
stress,
most
of
the
benefits
of
elevated
calcium
are
achieved
at
150

g/
L,
and
further
increases
produce
smaller
improvements.
Fish
presence
or
absence,
measures
of
fish
population
status,
or
the
number
of
fish
species
(
richness)
in
a
body
of
water
or
region
have
been
estimated
with
empirical
models.
Multiple
regression
analysis
is
usually
the
statistical
tool
used
in
empirical
models,
with
either
a
two­
state
response
variable,
such
as
presence
or
absence,
using
logistic
regression,
or
a
multiple­
state
or
continuous
response
variable,
using
standard
regression.

In
all
cases,
the
data
set
must
be
purged
of
bodies
of
water
which
have
known
explanations
for
fish
loss
other
than
acidification,
if
the
remaining
differences
among
lakes
or
streams
are
to
be
attributed
to
acidification
effects
on
fish.
In
this
connection,
two
other
physical
attributes
of
bodies
of
water
must
be
taken
into
account
in
modeling
efforts:
size
and
elevation.
For
example,
larger
lakes
and
lakes
at
lower
elevation
tend
to
support
more
species
of
fish,
all
else
being
equal.
For
the
SAMI
region,
focusing
on
mountain
streams,
size
and
elevation
are
probably
correlated
to
some
extent,
so
smaller
streams
and
streams
at
higher
elevation
are
expected
to
host
fewer
species.
5.3.2
Toxicity
Models
Toxicity
models
are
covered
in
some
detail
in
this
synopsis
because
they
provide
the
opportunity
for
near­
term
assessments
of
acidification
effects
on
fish
populations
in
the
SAMI
region.
J.
Baker
et
al.
(
1990)
define
toxicity
models
in
the
context
of
acidification
as
mathematical
functions
fitted
to
fish
responses
in
laboratory
studies
with
constant
levels
of
pH,
inorganic
monomeric
aluminum,
and
calcium.
Regression
models
are
then
used
to
estimate
the
acidification
effects
on
fish
associated
with
aluminum,
pH,
and
calcium
levels
measured
in
the
field.
Because
they
are
based
on
variables
with
known
toxic
(
aluminum
and
pH)
or
mitigating
(
calcium)
effects,
the
mechanistic
linkages
between
fish
response
and
acid­
base
chemistry
are
established
at
the
outset.
This
is
an
advantage,
but
two
important
disadvantages
to
toxicity
models
remain:
(
1)
the
field
survey
data
used
as
input
frequently
do
not
capture
the
extreme
events
to
which
sensitive
life
stages
of
fish
may
be
exposed
in
nature
and
(
2)
the
model
output
is
best
at
estimating
percent
mortality
in
laboratory
studies
and
cannot
easily
be
interpreted
directly
as
a
population
level
response.
The
primary
advantage
of
toxicity
models
is
that
they
can
deal
with
the
joint
effects
of
pH,
aluminum,
and
calcium.

5.3.3
Combined
Toxicity­
Field
Models
C
the
Acid­
Stress
Index
The
approach
used
to
model
the
effects
of
acidity
on
fish
in
the
1990
NAPAP
assessment
converted
the
joint
effects
of
pH,
inorganic
monomeric
aluminum,
and
calcium
into
a
single,
biological
response
variable,
called
the
acid
stress
index
(
ASI).
Three
criteria
were
used
for
the
data
sets
used
to
create
the
ASI
toxicity
models.
The
first
was
that
aluminum,
calcium,
and
pH
were
measured
as
part
of
the
experimental
design.
Secondly,
the
data
sets
were
based
on
observations
of
fish
mortality
as
opposed
to
sublethal
stresses;
the
reasons
52
being
that
mortality
is
most
often
used
as
an
observational
endpoint,
and,
while
sublethal
stresses
are
important,
they
are
difficult
to
interpret
as
population
level
effects.
Lastly,
the
data
sets
were
based
on
observations
of
early
life
stages;
the
reasons
being
that
there
are
large
number
of
such
studies,
early
stages
are
usually
the
most
sensitive
stage,
and
early
life
stage
mortality
has
a
clear
effect
on
population
recruitment.

Toxicological
experimental
elements
of
the
Lake
Acidification
and
Fisheries
(
LAF)
framework
(
Mount
et
al.,
1988a,
b)
were
specifically
designed
to
produce
results
that
could
be
incorporated
into
regression
models;
thus
the
LAF
data
are
an
important
resource
for
the
ASI
models.
The
four
freshwater
species
used
in
the
LAF
framework
were
brook
and
rainbow
trouts,
smallmouth
bass,
and
white
sucker.
Four
levels
of
fish
sensitivity
to
acidity
were
modeled:
tolerant,
intermediate,
sensitive,
and
anadromous.
Since
not
all
fish
species
potentially
affected
by
acidification
can
be
practically
modeled,
three
species
from
the
LAF
data
were
used
in
toxicity
models
as
reasonable
representations
of
the
range
of
responses
possible
in
fish
communities.

1.
Tolerant
Toxicity
model:
21­
day
survival
of
brook
trout
fry.

2.
Intermediate
Toxicity
model:
8­
day
survival
of
smallmouth
bass
alevins.

3.
Sensitive
Toxicity
model:
21­
day
survival
of
rainbow
trout
fry.

It
became
necessary
to
incorporate
an
even
more
sensitive
life
form
after
it
was
discovered
that
anadromous
species
(
those
that
are
spawned
in
freshwater,
spend
various
amounts
of
time
there,
then
migrate
to
seawater,
attain
sexual
maturity,
then
return
to
freshwater
to
spawn)
are
among
the
most
acid­
sensitive
fish
in
their
freshwater
phase,
even
more
sensitive
than
rainbow
trout.
Example
of
such
species
include
Atlantic
salmon,
striped
bass,
and
blueback
herring;
all
show
acid­
induced
stress
at
pH
levels
of
6.0B6.5.
Thus
a
fourth
toxicity
model
was
developed
specifically
for
use
in
the
Mid­
Atlantic
Coastal
Plain,
using
4­
day
survival
of
blueback
herring
alevins
as
the
basis
for
the
anadromous
model.
This
fourth
model
would
be
relevant
only
in
estimating
effects
of
acidification
outside
the
SAMI
region.

5.3.3.1
Acid
Stress
Index
Model
Structure.
The
proportion
of
fish
surviving
a
toxic
exposure
was
regressed
as
a
function
of
the
values
of
pH,
Al,
and
Ca,
using
maximum
likelihood
regression;
considerable
literature
exists
to
support
this
choice
of
model;
the
predicted
variable
was
converted
to
percent
mortality
(
0B100%),
and
adjusted
for
"
background"
(
not
due
to
acid
stress)
mortality.
The
mortality
percentage
due
to
acid
stress
is
referred
to
as
"
conditional
mortality
rate"
and
called
the
acid
stress
index
(
ASI).
Thus
higher
values
of
ASI
indicate
greater
stress
caused
specifically
by
acidification.
It
is
a
characteristic
of
the
water
whose
chemistry
is
modeled.

For
each
toxicity
data
set,
alternative
models
structures
were
evaluated
using
various
combinations
of
pH,
inorganic
Al,
and
Ca,
plus
squared
and
interactive
terms.
The
final
coefficients
included
for
each
of
the
four
models
were
based
on
three
criteria.

1.
Visual
(
graphical)
examination
of
the
model
fit
to
the
data
set
and
realism
of
the
model
output
outside
the
range
of
chemistry
variables
tested.

2.
Statistical
significance
of
model
parameters
(
only
variables
significant
at
p
<
0.05
were
included).

3.
Model
r2,
calculated
by
the
goodness­
of­
fit
chi­
square
for
the
intercept­
only
model
(
X)
to
the
goodness­
of­
fit
chi­
square
for
the
full
model
(
Y).
53
The
variables
used
in
the
toxicity
models
are
pH
in
standard
units,
Al
in

g/
L,
and
Ca
in

eq/
L.
The
model
r2
values
were
0.70,
0.80,
0.79,
and
0.78
for
tolerant,
intermediate,
sensitive
and
anadromous
models,
respectively.
The
intercept
and
coefficients
for
each
of
the
four
models
may
be
found
in
Table
13­
42
of
SOS/
T
13
(
J.
Baker
et
al.,
1990).
The
generalized
model
structure
is:

ASI
(%
mortality)
=
100/[
1
+
exp(
a
+
b
i
x
i)]

where
x
i
are
values
of
the
three
chemical
variables
and
b
i
are
their
coefficients.
Some
restrictions
were
placed
on
model
output.
The
ASI
was
assumed
to
be
zero
at
pH
>
8.0,
and
at
Ca
concentrations
>
2000

eq/
L;
inorganic
Al
concentrations
were
assumed
to
be
zero
at
pH
>
6.5.

5.3.3.2
Acid
Stress
Index
Model
Evaluation
In
regional
applications
of
the
toxicity
models,
outputs
are
intended
to
represent
index
responses
of
tolerant,
intermediate,
and
sensitive
fish
species,
rather
than
just
the
species
for
which
the
models
were
developed.
Index
or
baseflow
levels
of
pH,
Al,
and
Ca
were
used
to
calculate
ASI
values
for
individual
water
bodies.
To
determine
the
suitability
of
this
approach,
two
types
of
evaluations
were
executed:
interspecies
comparisons
and
comparison
to
field
population
responses.
For
interspecies
comparisons,
the
focus
was
on
the
applicability
of
the
sensitive
model
to
estimates
of
effects
on
other
sensitive
species.
Agreement
between
the
experimental
mortalities
of
common
shiner
and
blacknose
dace,
typical
of
a
group
of
sensitive
cyprinid
fish,
and
the
output
of
the
sensitive
model
predictions
of
mortality
was
considered
adequate.

ASI
values
were
also
compared
to
field
observations
of
fish
population
responses.
Predictions
(
ASI
values)
derived
from
lab
studies
do
not
include
natural
variation
in
chemistry
or
other
variables
which
affect
survival
in
nature.
Nor
do
these
predictions
consider
biological
compensation
(
positive
effects
on
the
survivors
due
to
reduced
density
resulting
from
deaths
of
some
of
the
population
Nevertheless,
it
is
clear
that
the
ASI
has
value
as
an
index,
based
on
comparisons
with
three
data
sets
from
(
1)
Adirondack
lakes,
(
2)
Ontario
lakes,
and
(
3)
western
Maryland
streams.
For
lakes,
the
likelihood
of
presence
of
brown
bullhead
was
about
50%
when
the
ASI­
tolerant
values
were
about
30;
thus,
waters
with
an
ASI
value
>
30
would
appear,
on
average,
to
have
acid­
base
chemistry
unsuitable
for
the
survival
of
brown
bullheads
Brook
trout
were
lost
from
waters
(
likelihood
of
occurrence
=
50%)
at
ASItolerant
values
above
10.
The
likelihood
of
occurrence
was
50%
when
ASI­
intermediate
values
were
near
80
for
lake
trout
and
ASI­
sensitive
values
were
near
80B90
for
common
shiner.

The
relationship
between
ASI
and
fish
presence/
absence
clearly
differs
between
streams
and
lakes.
Stream
fish
populations
appear
to
be
affected
at
much
lower
values
of
ASI.
Since
ASI
values
are
calculated
from
the
index
(
baseflow)
chemistry
values,
this
difference
in
response
probably
reflects
the
more
significant
role
of
episodic
acidification
in
streams
versus
lakes.
As
a
result,
both
brook
trout
and
blacknose
dace
presence/
absence
appeared
to
be
best
represented
by
the
sensitive
model
in
streams,
with
reference
ASI
values
of
>
30
and
>
10
for
brook
trout
and
blacknose
dace,
respectively.

Although
there
is
not
a
one­
to­
one
relationship
between
ASI
and
population­
level
effects,
the
NAPAP
study
concluded
that
toxicity­
based
model
outputs
can
be
useful
as
an
index
of
probable
acidification
stress
at
the
population
level.
As
a
result,
reference
levels
of
ASI
were
associated
with
the
following
fish
responses
for
streams
(
Table
13­
43
in
SOS/
T
13,
J.
Baker
et
al.,
1990):
54
$
ASI­
intermediate
values
>
30:
Loss
of
all
species
$
ASI­
sensitive
values
>
30:
Loss
of
brook
trout
$
ASI­
sensitive
values
>
10:
Loss
of
acidsensitive
species,
e.
g.
minnows
It
is
very
important
to
note
that
ASI
models
constructed
from
lab
experiments
taken
by
themselves
appear
to
underestimate
the
acidification
stress
experienced
by
populations
in
nature.
For
example,
streams
that
have
lost
brook
trout
populations
have
index
values
of
pH,
Al,
and
Ca
that
would
result
in
as
little
as
30%
mortality
of
the
more
sensitive
rainbow
trout
fry
in
lab
bioassays
(
see
above,
ASIsensitive
>
30,
associated
with
loss
of
rainbow
trout
populations).
This
is
no
doubt
caused
in
part
by
temporal
and
spatial
variation
in
chemistry,
e.
g.,
episodes
of
poor
water
quality
which
occur
during
sensitive
life
stages
of
fish.
It
would
be
preferable
to
include
episodes
into
predictions
of
fish
status,
but
regional­
scale
data
sets
are
lacking,
and
exposures
during
episodes
are
themselves
difficult
to
characterize.
Thus,
given
the
uncertainties
and
limitations
regarding
chemical
variation
in
nature,
the
NAPAP
Assessment
used
ASI
values
based
on
index
chemistry,
but
interpreted
them
in
light
of
reference
ASI
values
based
on
field
observations
as
indices
of
probable
population
effects.
55
6.
PREDICTIVE
STUDIES
OF
ACIDIC
DEPOSITION
EFFECTS
IN
THE
SOUTHERN
APPALACHIANS
6.1
NAPAP
Modeling
Efforts
The
NAPAP
efforts
to
model
future
changes
consisted
of
the
Direct/
Delayed
Response
Project
(
DDRP)
and
the
1990
Integrated
Assessment
(
IA).
In
the
SAMI
region,
both
efforts
focused
on
the
potential
effects
of
acidic
deposition
on
surface
water
ANC
during
the
spring
baseflow
index
period
defined
and
sampled
by
the
NSS.
The
time
horizon
for
these
analyses
was
50
years,
starting
in
1985.
The
primary
processes
modeled
were
(
1)
the
retention
of
deposited
sulfur
within
watersheds
and
(
2)
the
supply
of
base
cations
from
watersheds
to
surface
waters.
These
two
processes
had
been
identified
by
a
National
Academy
of
Sciences
panel
as
the
most
important
watershed
processes
affecting
or
mediating
long­
term
surface
water
acidification
(
NAS,
1984).

The
DDRP
(
Church
et
al.,
1989,
1992)
and
IA
(
NAPAP,
1991)
modeled
streamwater
chemistry
responses
to
various
deposition
scenarios
in
two
regions
of
the
eastern
United
States:
the
Mid­
Appalachians
(
MIDAPP)
and
the
Southern
Blue
Ridge
Province
(
SBRP).
The
SBRP
modeling
region
is
entirely
within
the
SAMI
region,
but
it
does
not
include
parts
of
the
Blue
Ridge
located
in
Virginia
and
northern
North
Carolina.
The
MIDAPP
modeling
region
contains
all
the
acid­
sensitive
SAMI
study
region
in
Virginia
and
West
Virginia,
plus
large
parts
of
Pennsylvania
and
the
Catskill
region
of
New
York.

Three
watershed
models
were
applied
to
streams
in
the
SBRP:

1.
The
Model
of
Acidification
of
Groundwater
in
Catchments,
MAGIC
(
Cosby
et
al.,
1985a,
b;
1986a,
b).
2.
The
Enhanced
Trickle
Down
(
EDT)
model
(
Lee,
1987;
Nikolaidis
et
al.,
1988;
Schnoor
et
al.,
1986).

3.
The
Integrated
Lake­
Watershed
Acidification
Study
(
ILWAS)
model
(
Chen
et
al.,
1983;
Gherini
et
al.,
1985).

These
models
integrated
the
current
understanding
of
how
various
watershed
processes
interact
and
respond
to
acidic
deposition.
The
three
models
used
common
datasets
for
forcing
functions
(
e.
g.,
rainfall,
runoff,
atmospheric
deposition)
and
used
data
aggregated
from
the
DDRP
soils
database
for
state
variables
such
as
soil
physical
and
chemical
characteristics
(
Church
et
al.,
1992).

Although
absolute
values
of
modeled
concentrations
of
constituents
varied
in
comparisons
for
individual
watersheds,
Church
et
al.
(
1989)
reported
that
projections
of
changes
in
ANC
among
the
models
were
remarkably
consistent
on
a
regional
basis.
Rather
than
repeating
costly
and
time
consuming
duplicative
modeling
analyses,
Church
et
al.
(
1992)
used
only
the
MAGIC
model
for
projections
in
the
MIDAPP.
They
reported
that
MAGIC
was
by
far
the
easiest,
quickest,
and
least
expensive
to
apply.
For
the
same
reasons,
NAPAP
(
1991)
used
MAGIC
as
the
primary
method
for
projecting
and
comparing
the
chemical
responses
of
SBRP
and
MIDAPP
streams
to
various
deposition
scenarios
in
the
1990
Integrated
Assessment
Report.

The
DDRP
and
IA
modeling
efforts
were
intended
to
estimate
the
general
direction
and
the
relative
magnitude
of
possible
future
changes
in
surface
water
acid­
base
chemistry
under
alternative
scenarios
of
atmospheric
sulfur
deposition
(
Church
et
al.,
1992).
The
projections
of
these
models
were
not
intended
to
be
forecasts
of
future
conditions,
but
tools
56
to
evaluate
the
relative
effectiveness
of
alternative
emission
controls.
The
modeled
stream
population
for
all
three
assessments
is
the
downstream
ends
of
NSS
stream
segments
with
drainage
areas
<
30
km2.
In
addition,
only
streams
with
ANC
<
200

eq/
L
and
no
substantial
watershed
sources
of
sulfate
or
chloride
were
included
in
the
MIDAPP
population.
These
criteria
eliminated
about
30%
of
the
NSS
target
population
in
the
SBRP
and
42%
in
the
MIDAPP,
and
restricted
the
model
inferences
to
the
subset
of
streams
likely
to
respond
to
acidic
deposition
inputs
over
the
next
several
decades.
However,
the
projections
were
calibrated
on
the
downstream
ends
of
the
NSS
stream
reaches
and
estimate
the
chemistry
for
these
ends.
ANC
at
the
lower
ends
of
these
streams
was
generally
higher
than
at
the
upstream
ends
(
Table
1),
so
these
projections
overestimate
ANC
and
underestimate
changes
in
ANC
for
the
upper
ends.
For
these
reasons,
the
model
baseline
conditions
will
not
agree
with
the
NSS
current
status
estimates
that
include
streams
over
a
broader
size
and
disturbance
spectrum.

6.1.1
Model
Results
The
basic
results
and
conclusions
of
the
IA
agree
with
the
DDRP
results
of
Church
et
al.
(
1989,
1992),
and
primarily
the
IA
results
appear
in
the
following
outline
of
future
modeling
projections.
Furthermore,
we
have
borrowed
freely
from
Church
et
al.
(
1989,
1992)
and
NAPAP
(
1991)
in
the
wording
of
most
of
these
conclusions.
Both
the
DDRP
and
the
IA
modeled
a
projection
of
current
deposition
and
defined
it
as
the
1985
level.
All
deposition
scenarios
extend
to
50
years
beyond
1985.
The
basic
results
and
conclusions
of
the
future
modeling
efforts
are
as
follows.

6.1.1.1
Sensitivity
to
Acidification.
Three
percent
of
the
modeled
MIDAPP
stream
population
currently
is
acidic
and
15%
has
ANC

50

eq/
L.
The
streams
in
the
MIDAPP
highlands
with
lowest
ANC
are
generally
restricted
to
small,
forested
watersheds
at
relatively
high
elevations
(>
300
m).
These
streams
are
expected
to
be
the
most
sensitive
to
acidic
deposition.
None
of
the
streams
sampled
by
the
NSWS
in
the
SBRP
was
chronically
acidic,
but
12%
of
the
modeled
subpopulation
had
ANC

50

eq/
L.
Streams
in
the
SBRP
may
be
responsive
to
changes
in
acidic
deposition,
but
the
response
is
delayed
more
than
that
of
MIDAPP
streams
because
watershed
soils
are
retaining
more
sulfate
from
acidic
deposition.
Church
et
al.
(
1989)
estimated
that
their
watersheds
are
currently
retaining
three­
quarters
of
the
atmospherically
deposited
sulfur
on
the
average,
but
projected
that
their
soils
will
become
more
saturated
with
sulfur
over
time.
As
a
result,
they
project
that
watershed
sulfur
retention
will
decline,
increasing
sulfate
concentrations
in
the
surface
waters
of
this
region.
The
response
is
projected
to
be
marked
over
the
next
50
years
at
either
current
or
increased
levels
of
sulfur
deposition,
as
are
decreases
in
streamwater
ANC.
Superimposed
upon
this
effect
is
a
relatively
minor
acidification
effect
of
base
cation
depletion
(
Church
et
al.,
1989).

6.1.1.2
Projections
for
Current
and
Increased
Deposition.
Streams
in
the
MIDAPP
and
SBRP
are
projected
to
acidify
(
decrease
in
ANC
over
time)
under
current
levels
of
deposition
(
Figure
12).
Streamwater
concentrations
of
sulfate
are
projected
to
increase
substantially
over
the
next
50
years,
accelerating
both
cation
leaching
from
soils
and
the
projected
acidification
of
surface
waters.
Median
ANC
of
streams
in
the
MIDAPP
and
SBRP
would
decrease
about
20

eq/
L
over
50
years
at
current
levels
of
atmospheric
sulfur
deposition,
because
less
sulfate
is
being
retained
by
soils
over
time.
ANC
would
change
by
about
3
to
4

eq/
L
in
addition
to
the
20

eq/
L
decrease
for
each
57
Figure
12.
Future
projections
of
stream
ANC
and
pH
after
50
years
for
6
sulfate
deposition
scenarios
for
(
a)
Mid­
Atlantic
Highlands
(
includes
Pennsylvania,
Maryland,
and
New
York)
and
(
b)
Southern
Blue
Ridge.
Taken
from
Figures
4.4­
7
and
4.4­
9
in
1990
NAPAP
Integrated
Assessment.
CSC
are
the
current
(
1985)
simulated
conditions
at
year
0.
The
six
future
deposition
scenarios
are:


50%:
Current
levels
for
5
years,
ramp
down
50%
over
10
years,
then
constant.


30%:
Current
levels
for
5
years,
ramp
down
30%
over
10
years,
then
constant.


20%:
Current
levels
for
5
years,
ramp
down
20%
over
10
years,
then
constant.
0:
Current
deposition
level
for
50
years.
+
20%:
Ramp
up
20%
above
current
levels
over
25
years,
then
constant.
58
+
30%:
Ramp
up
30%
above
current
levels
over
25
years,
then
constant.
59
kg/
ha/
yr
change
from
current
sulfur
deposition
over
50
years.
Under
current
levels
of
deposition
over
50
years,
46%
of
streams
would
experience
ANC
declines
of

10

eq/
L;
15%
of
streams
would
experience
pH
declines
of

0.3
units.

In
the
MIDAPP,
a
decline
in
the
regional
median
ANC
of
about
25

eq/
L
is
projected
over
50
years.
After
50
years
at
current
levels
of
deposition,
10%
of
MIDAPP
streams
are
projected
to
be
acidic
and
24%
would
have
ANC

50

eq/
L
(
Figure
12).
At
sulfur
deposition
20%
and
30%
greater
than
current
levels,
the
number
of
acidic
MIDAPP
streams
is
projected
to
increase
by
factors
of
5.4
and
6.2,
respectively,
over
the
next
50
years.
In
the
SBRP,
projections
suggest
that
after
50
years
at
current
levels
of
deposition,
about
10%
of
the
streams
will
be
acidic
and
15%
will
have
ANC

50

eq/
L
(
Figure
12).
A
20%
increase
in
deposition
would
result
in
an
increase
to
10
B
12%
acidic
streams
and
23
B
26%
with
ANC

50

eq/
L.

6.1.1.3
Reduced
Deposition
Scenarios.
With
reduced
deposition,
projected
changes
in
streamwater
sulfate
will
ultimately
respond
to
the
level
of
deposition.
Model
projections
suggest
that
a
20
B
30%
reduction
in
sulfur
deposition
would
be
necessary
to
prevent
further
acidification
in
MIDAPP
streams
over
the
next
50
years;
reductions
of
30
B
50%
would
be
required
in
the
SBRP.
Because
of
the
substantial
amount
of
sulfate
adsorption
in
SBRP
soils,
a
delay
in
the
reduction
of
sulfur
deposition
would
be
expected
to
result
in
acidification
greater
than
that
which
would
occur
with
no
delay.
These
effects
would
occur
beyond
the
50­
year
time
frame
modeled.

6.1.1.4
Chemical
Conditions
for
Fish.
The
projected
acidic
stress
indices
for
sensitive
fish
species
suggest
that
biological
conditions
in
low
ANC
MIDAPP
streams
will
deteriorate
under
current
deposition
levels.
The
simulated
current
percentage
of
MIDAPP
stream
segments
with
downstream
end
water
chemistry
unsuitable
for
sensitive
fish
species
(
e.
g.,
rainbow
trout,
blacknose
dace)
is
21%.
With
a
30%
increase
in
sulfur
deposition,
this
percentage
is
projected
to
increase
to
34%.
With
a
50%
decrease
in
sulfur
deposition,
it
would
remain
at
21%.
A
25­
year
delay
in
the
reduction
of
sulfur
deposition
is
projected
to
increase
the
number
of
MIDAPP
streams
unsuitable
for
sensitive
fish
species
by
6%
during
the
50­
year
modeling
period.
If
unsuitable
chemical
conditions
eliminate
reproducing
populations
of
fish
and
other
aquatic
biota
from
some
streams,
they
will
have
to
recolonize
these
streams
from
other
locations
and
may
not
return
until
well
after
chemical
conditions
permit.
Estimates
of
fish
response
were
not
made
for
streams
in
the
SBRP
because
the
IA
researchers
did
not
find
field
information
relating
fish
status
to
stream
acidbase
chemistry
and
felt
that
the
uncertainties
of
extrapolating
models
developed
elsewhere
were
too
great
(
NAPAP,
1991).

6.1.1.5
Uncertainties
in
Model
Projections
Uncertainties
in
the
projections
for
timing
of
aquatic
effects
are
high
because
of
uncertainties
in
sulfate
adsorption­
desorption
dynamics
and
in
the
relative
importance
of
weathering
versus
cation
exchange
processes
(
NAPAP
1991).
Additional
uncertainty
results
from
the
fact
that
acidification
from
nitrogen
deposition
was
not
considered
in
the
modeling.
However,
limited
field
observations
of
ANC
trends
in
MIDAPP
streams
are
in
the
low
end
of
the
predicted
range
of
modeled
ANC
changes.
Conclusions
regarding
future
effects
from
acidic
deposition
are
considerably
more
uncertain
in
the
SBRP
than
in
the
MIDAPP,
primarily
because
of
the
difficulty
of
predicting
future
changes
in
sulfur
retention
by
soils.
Effects
are
likely
to
occur
beyond
the
50­
year
time
frame
used
in
the
modeling.
60
6.2
Nitrogen
Bounding
Study
Most
of
the
efforts
at
dynamic
modeling
of
watershed
acidification
that
were
carried
out
as
part
of
NAPAP
(
see
section
6.1)
did
not
consider
the
potential
influence
of
changes
in
watershed
nitrogen
retention
on
model
forecasts
The
Nitrogen
Bounding
Study
(
NBS;
Van
Sickle
and
Church,
1995),
whose
results
were
also
reported
as
part
of
the
Acid
Deposition
Standard
Feasibility
Study
(
U.
S.
Environmental
Protection
Agency,
1995)
was
an
attempt
to
bracket
the
potential
effects
that
nitrogen
saturation
might
have
on
future
acidification.
The
NBS
used
the
modeling
techniques
of
the
DDRP,
several
different
scenarios
for
future
nitrogen
and
sulfur
deposition
and
hypothetical
time
sequences
for
nitrogen
saturation
to
predict
the
percentages
of
streams
that
would
be
either
chronically
acidic
(
ANC
<
0)
or
susceptible
to
episodic
acidification
(
ANC
<
50

eq/
L)
in
the
future.

In
the
case
of
the
Mid­
Appalachian
region,
the
study
predicted
that
8%
of
streams
would
become
acidic
at
current
rates
of
nitrogen
deposition
and
forecasted
rates
of
sulfur
deposition
(
roughly
half
of
1985
rates),
if
nitrogen
saturation
occurs
in
50
years.
Approximately
40%
of
streams
would
be
episodically
acidic
under
the
same
scenario.
If
the
time
to
nitrogen
saturation
is
expanded
to
250
years,
current
rates
of
N
deposition
would
lead
to
chronic
acidification
in
4%
of
streams
and
episodic
acidification
in
28%
of
streams.

The
NBS
predicted
much
lower
rates
of
acidification
in
the
southern
Blue
Ridge,
where
4%
(
chronic)
and
16%
(
episodic)
of
streams
would
be
acidified
with
a
time
to
saturation
of
50
years.
With
a
250­
year
time
to
saturation,
the
models
predicted
no
streams
would
be
chronically
acidic,
and
14%
would
be
susceptible
to
episodic
acidification.
These
modeling
results
are
somewhat
at
odds
with
information
about
the
current
status
of
streams
in
the
SBRP.
Some
streams
in
Great
Smoky
Mountains
National
Park
are
acidic
and
have
high
nitrate
concentrations.
These
streams
tend
to
occur
at
high
elevations,
whereas
the
DDRP
sampling
design
was
based
on
the
lower
elevation
downstream
segment
end
population.
Thus,
the
NBS
results
based
on
the
DDRP
are
biased
toward
lower
elevation
streams
(
Cook
et
al.,
1994),
leading
to
more
conservative
forecasts
for
future
acidity.
61
7.
RECOMMENDATIONS
FOR
SAMI
ASSESSMENT
7.1
Model
Recommendations
We
recommend
that
the
SAMI
assessment
use
the
MAGIC
model
for
any
quantitative
future
projections
of
surface
water
acid­
base
status.
The
steady­
state
models
are
not
appropriate
for
the
Southern
Appalachians.
In
addition
to
the
arguments
against
steady­
state
models
listed
in
section
5.1.1,
streams
in
the
SAMI
region
are
not
at
steady
state.
The
time
required
to
reach
steady
state
is
one
of
the
major
unknowns
that
needs
to
be
modeled
in
the
SAMI
assessment.
Of
the
two
dynamic
watershed
models
(
ILWAS
and
MAGIC)
suitable
for
SAMI,
MAGIC
is
by
far
the
easiest,
quickest,
and
least
expensive
to
apply.
The
two
models
do
have
some
real
differences
in
terms
of
some
projections
(
especially
among
alternate
deposition
scenarios
in
the
southern
Blue
Ridge).
Both
models
were
extensively
tested
during
the
NAPAP
assessment
and
there
was
no
evidence
that
either
model
was
more
accurate
than
the
other.
Although
ILWAS
is
more
complex
and
models
more
processes
and
compartments,
complexity
does
not
necessarily
make
the
model
results
more
accurate.
Therefore,
it
seems
more
cost­
effective
to
select
the
more
economical
MAGIC
model.

We
recommend
the
use
of
the
PnET­
CN
model
as
the
nitrogen
model
for
use
in
further
assessments,
because
of
its
track
record
of
applicability
to
a
variety
of
sites
with
relatively
little
site­
specific
information.
It
would
require
some
regional
compilation
of
climate
data
and
land
use
histories,
as
well
as
the
collection
of
foliar
nutrient
samples
from
all
sites
of
interest
(
a
fairly
straightforward
task).
One
of
the
primary
outputs
of
PnET­
CN
is
the
time
to
nitrogen
saturation.
This
information
could
easily
be
incorporated
into
runs
of
the
MAGIC
model,
to
produce
desired
forecasts
of
acid­
base
status,
given
various
scenarios
of
sulfur
and
nitrogen
deposition
(
as
well
as
climate
change,
if
desired).

To
make
projections
about
the
effects
of
episodic
acidification
on
streams
in
the
SAMI
region,
we
recommend
linking
the
regression
model
approach
of
Eshleman
(
1988)
to
MAGIC
model
projections.
This
step
would
provide
a
first
approximation
of
the
likely
magnitude
of
episodic
impacts.
The
regression
model
is
fairly
robust,
can
be
applied
on
a
regional
basis,
and
has
already
been
shown
to
be
linkable
to
MAGIC.
The
regression
equations,
however,
should
be
recalibrated
using
new
data
from
studies
in
the
Shenandoah
and
Great
Smoky
Mountains
National
Parks.

For
doing
future
projections
of
the
effects
of
acidic
deposition
on
fish,
we
recommend
using
the
acid
stress
index
(
ASI)
approach
used
in
the
NAPAP
assessment.
Calcium,
pH,
and
aluminum
values
can
be
obtained
from
MAGIC
output
to
calculate
ASIs
under
alternate
deposition
scenarios.
Critical
ASI
values
can
then
be
obtained
by
comparing
current
fish
distributions
to
current
ASI
levels.

7.2
Levels
of
Assessment
For
SAMI,
we've
devised
four
levels
of
effort
for
doing
an
aquatic
effects
assessment.
Each
additional
level
will
require
more
resources
but
will
give
a
more
complete
assessment.
The
final
choice
of
an
assessment
approach
will
require
decisions
about
the
nature
and
objective
of
the
assessment.
For
example,
will
the
assessment
focus
on
streams
in
the
entire
SAMI
region,
the
acid­
sensitive
part
of
the
region,
or
just
the
Class
I
areas?
Should
the
assessment
just
look
at
sulfur
deposition
or
sulfur
and
nitrogen?
How
quantitative
does
the
answer
need
to
be?
How
different
are
the
various
emission
management
options
or
deposition
scenarios?
As
the
SAMI
assessment
is
still
in
its
formative
stages,
our
recommended
levels
span
a
wide
range
of
62
options.
The
final
decision
will
depend
heavily
on
the
final
choice
of
the
SAMI
assessment
objective.

7.2.1
Level
1
A
very
simple
qualitative
assessment
could
be
made
by
examining
the
existing
results
of
the
DDRP
and
NAPAP
Integrated
Assessment.
For
example,
Figure
12
in
this
report
gives
a
reasonably
good
idea
of
the
relative
effects
of
various
sulfur
deposition
scenarios
on
the
regional
population
of
streams
in
the
mid­
Appalachians
and
southern
Blue
Ridge.
This
analysis
focuses
on
the
region
as
a
whole,
not
the
Class
I
areas.
As
the
Class
I
areas
comprise
some
of
the
most
sensitive
systems
in
the
SAMI
region,
inferences
to
them
could
be
made
by
examining
what
happens
to
the
most
sensitive
streams
in
the
regional
distribution.

7.2.2
Level
2
An
analysis
that
would
require
no
new
collection
of
data
would
involve
redoing
the
DDRP/
NAPAP
Assessment
at
the
existing
DDRP
sites
using
new
deposition
scenarios
provided
by
SAMI.
The
data
currently
exist
to
run
MAGIC
at
all
these
sites
and
to
calculate
fish
ASI
values
using
MAGIC
output.
This
activity
would
provide
a
quantitative
regional
estimate
of
stream
acid­
base
status
for
each
of
the
SAMI
deposition
scenarios.
This
level
has
two
major
shortcomings.
One
drawback
is
that
it
really
does
not
handle
changes
in
nitrogen
deposition/
dynamics;
the
assessment
would
remain
driven
by
sulfur
deposition.
Secondly,
as
in
level
1,
level
2
focuses
on
the
region
as
a
whole,
not
the
Class
I
areas
that
may
be
of
primary
interest
to
SAMI.
As
in
level
1,
some
inference
to
the
Class
I
areas
could
be
made
by
examining
what
happens
to
the
most
sensitive
streams
in
the
region,
but
it
would
not
be
very
quantitative
or
applicable
to
any
specific
Class
I
area.

7.2.3
Level
3
Level
3
efforts
would
be
geared
toward
using
the
most
current
methodologies
for
making
the
projections
of
the
aquatic
effects
of
acidic
deposition.
The
analysis
should
be
done
by
Class
I
wilderness
area
groups
(
e.
g.,
West
Virginia
Plateau,
southern
Blue
Ridge).
It
could
be
done
for
just
one
group
or
as
many
of
the
groups
as
resources
allow.
We
would,
however,
make
the
Sipsey
wilderness
area
group
a
low
priority
and
focus
efforts
in
the
other
three
areas.
We
would
recommend
field
efforts
to
collect
new
data
on
soils,
watershed
attributes,
and
hydrology
for
three
low
ANC
watersheds
(
along
the
available
ANC
gradient)
in
each
group.
Where
water
chemistry
information
is
sparse
or
lacking
(
see
7.4),
a
probability
sample
of
30
B
50
streams
in
the
wilderness
area
group
should
be
conducted
to
aid
in
regionalizing
to
the
entire
area
group.
This
task
would
be
relatively
simple
in
the
smaller
areas
but
might
require
some
extra
effort
in
the
GSMNP.
Data
necessary
to
run
the
PnET­
CN
nitrogen
model
and
episodic
acidification
model
(
foliar
chemistry,
land
use,
and
hydrologic
information)
should
be
collected
at
the
same
time.
MAGIC
and
PnET­
CN
models
would
be
used
to
make
future
projections
about
average
annual
streamwater
chemistry.
The
episodic
model
would
be
linked
to
the
MAGIC
output
to
predict
episodic
acidity.
ASI
values
could
then
be
calculated
for
both
episodic
and
chronic
conditions
to
estimate
the
effects
on
fish.
All
the
models
would
need
to
be
calibrated
and
run
under
the
SAMI
deposition
scenarios.
This
approach
would
allow
for
the
best
available
projections
for
the
modeled
Class
I
wilderness
area
groups.
63
7.2.4
Level
4
A
level
4
analysis
would
consist
of
conducting
the
level
3
analysis
for
each
of
the
DDRP
stream
sites
(
level
2)
to
allow
regional
projections
that
take
both
nitrogen
and
sulfur
dynamics
into
account.
This
work
would
involve
collecting
additional
information
at
these
sites
to
run
PnET­
CN.

7.2.5
Fish
Assessment
A
fish
assessment
could
be
added
to
Level
2,
3,
or
4
quite
easily
using
the
ASI
approach.
However,
field
information
on
the
relationship
between
fish
distributions
and
ASI
values
needs
to
be
collected
and
evaluated
in
order
to
set
critical
ASI
levels
for
sensitive,
intermediate
and
tolerant
fish
species.

7.3
Limitations
and
Relationship
to
a
SAMI
Integrated
Assessment
One
of
the
conclusions
of
the
NAPAP
assessment
was
that
"
Uncertainties
in
the
absolute
magnitudes
and
timing
of
aquatic
effects
projections
are
high,
but
we
have
confidence
in
the
projected
direction
of
change
and
in
the
relative
amounts
of
change."
We
believe
that
this
conclusion
has
two
important
implications
for
the
SAMI
assessment:

1.
Emission
management
options
or
deposition
scenarios
that
result
in
very
small
changes
in
sulfur
and
nitrogen
deposition
will
not
cause
distinguishable
differences
in
aquatic
effects.
The
resolution
of
the
models
is
not
that
high.

2.
Transferring
absolute
aquatic
model
projections
(
e.
g.,
miles
of
acidic
streams)
into
a
broader
integrated
effects
model
has
a
large
potential
for
significant
errors.
The
models
are
reasonably
good
for
evaluating
the
relative
differences
of
different
deposition
scenarios.
Thus,
we
are
fairly
comfortable
with
running
the
models
and
making
conclusions
about
the
relative
differences
in
aquatic
effects
among
various
deposition
scenarios.
However,
we
are
uncomfortable
with
using
the
absolute
results
(
e.
g.,
length
of
acidic
streams)
of
these
models.
We
are
very
uncomfortable
with
taking
these
absolute
results,
linking
them
together
with
absolute
results
from
other
effects
models
(
e.
g.,
visibility,
ozone),
running
them
all
through
a
socioeconomic
valuation
model,
and
using
some
kind
of
overall
"
cost"
estimate
as
the
decision
making
tool
for
evaluating
different
emissions
scenarios.
From
the
aquatic
perspective,
future
projections
about
miles
of
acidic
streams
or
miles
of
streams
that
will
have
impaired
or
absent
fish
populations
have
very
large
uncertainties
due
to
model
structure,
sensitivity,
calibration,
and
definition
of
the
target
population.

7.4
Data
Availability
The
surface
water
chemistry
and
watershed
data
required
to
run
the
MAGIC
model
exist
at
35
streams
in
the
southern
Blue
Ridge
and
36
streams
in
the
mid­
Appalachians
that
were
sampled
in
the
NSS
and
DDRP.
Note
that
only
17
of
the
36
streams
in
the
mid­
Appalachians
were
in
the
SAMI
states
of
West
Virginia
or
Virginia.
Data
also
exist
to
run
the
MAGIC
model
in
the
streams
of
Shenandoah
National
Park
where
the
MAGIC
model
was
developed.
We
are
not
aware
of
any
other
locations
in
the
SAMI
region
where
sufficient
watershed,
soils,
water
chemistry,
and
hydrologic
data
are
available
or
are
in
a
form
ready
to
run
MAGIC.

Available
water
chemistry
data
sets
exist
in
many
of
the
Class
I
areas
in
varying
degrees
of
resolution.
The
magnitude
of
the
"
No
Data"
column
in
Table
5
relates
the
degree
of
completeness.
High­
resolution
data
sets
exist
for
Otter
Creek,
Dolly
Sods,
and
the
Shenandoah
National
Park.
Data
sets
with
good
resolution
exist
for
James
River
Face
and
Cohutta.
More
complete
data
need
to
be
collected
in
Sipsey,
Slickrock,
and
Shining
64
Rock
wilderness
areas.
Data
exist
for
359
stream
sites
in
the
Great
Smoky
Mountains
National
Park.
These
sites,
however,
repre
sent
only
about
one­
third
of
the
stream
length
in
the
park.
A
population­
level
assessment
is
likely
to
require
a
probability
survey
of
the
park's
2,035
km
of
streams
(
Table
3).
65
8.
LITERATURE
CITED
Aber,
J.
D.,
K.
J.
Nadelhoffer,
P.
Steudler,
and
J.
M.
Melillo.
1989.
Nitrogen
saturation
in
northern
forest
ecosystems.
BioScience
39:
378­
386.

Aber,
J.
D.
and
C.
A.
Federer.
1992.
A
generalized,
lumped­
parameter
model
of
photosynthesis,
evapotranspiration
and
net
primary
production
in
temperate
and
boreal
forest
ecosystems.
Oecologia
92:
463­
474.

Aber,
J.
D.,
S.
V.
Ollinger
and
C.
T.
Driscoll.
In
press.
The
ratio
of
measured
to
predicted
steady­
state
N
cycling
rates
as
an
indicator
of
nitrogen
saturation
in
forest
ecosystems.
Ecological
Modelling.

Baker,
J.
P.,
D.
P.
Bernard,
S.
W.
Christensen,
M.
J.
Sale,
and
Others.
1990.
Biological
Effects
of
Changes
in
Surface
Water
Acid­
Base
Chemistry.
Report
13.
In:
National
Acid
Precipitation
Assessment
Program,
Acidic
Deposition:
State
of
Science
and
Technology,
Volume
II.
National
Acid
Precipitation
Assessment
Program,
Washington,
D.
C.

Baker,
J.
P.,
and
S.
W.
Christensen.
1991.
Effects
of
acidification
on
biological
communities
in
aquatic
ecosystems.
Pages
83­
106
in:
D.
F.
Charles,
ed.
Acidic
Deposition
and
Aquatic
Ecosystems:
Regional
Case
Studies.
Springer­
Verlag,
New
York.

Baker,
L.
A.,
P.
R.
Kaufmann,
A.
T.
Herlihy,
and
J.
M.
Eilers.
1990.
Current
Status
of
Surface
Water
Acid­
Base
Chemistry.
Report
9.
In:
National
Acid
Precipitation
Assessment
Program,
Acidic
Deposition:
State
of
Science
and
Technology,
Volume
II.
National
Acid
Precipitation
Assessment
Program,
Washington,
D.
C.

Bowden,
W.
B.
1986.
Gaseous
nitrogen
emissions
from
undisturbed
terrestrial
ecosystems:
An
assessment
of
their
impacts
on
local
and
global
nitrogen
budgets.
Biogeochemistry
2:
249­
279.

Bowden,
W.
B.,
and
F.
H.
Bormann.
1986.
Transport
and
loss
of
nitrous
oxide
in
soil
water
after
forest
clear­
cutting.
Science
233:
867­
869.

Bricker,
O.
P.,
and
K.
C.
Rice.
1989.
Acidic
deposition
to
streams:
a
geology
based
method
predicts
their
sensitivity.
Environ.
Sci.
Technol.
23:
379­
385.

Bulger,
A.
J.,
C.
A.
Dolloff,
B.
J.
Cosby,
K.
N.
Eshleman,
J.
R.
Webb,
and
J.
N.
Galloway.
1995.
The
Shenandoah
National
Park:
Fish
In
Sensitive
Habitats
(
SNP:
FISH)
Project:
An
integrated
assessment
of
fish
community
responses
to
stream
acidification.
Water
Air
Soil
Pollut.
85:
309­
314.

Chen,
C.
W.,
S.
A.
Gherini,
J.
D.
Dean,
R.
J.
M.
Hudson,
and
R.
A.
Goldstein.
1983.
Modeling
of
Precipitation
Series.
Volume
9.
Ann
Arbor
Sciences,
Butterworth
Publishers,
Boston,
Massachusetts.
175
pp.

Church,
M.
R.,
K.
W.
Thornton,
P.
W.
Shaffer,
D.
L.
Stevens,
B.
P.
Rochelle,
G.
R.
Holdren,
M.
G.
Johnson,
J.
J.
Lee,
R.
S.
Turner,
D.
L.
Cassell,
D.
A.
Lammers,
W.
G.
Campbell,
C.
I.
Liff,
C.
C.
Brandt,
L.
H.
Liegel,
G.
D.
Bishop,
D.
C.
Mortenson,
S.
M.
Pierson,
and
D.
D.
Schmoyer.
1989.
Direct/
Delayed
Response
Project:
Future
Effects
of
Long­
term
Sulfur
Deposition
on
Surface
Water
Chemistry
in
the
Northeast
and
Southern
Blue
Ridge
Province.
EPA/
600/
3­
89/
061.
U.
S.
Environmental
Protection
Agency,
Washington
D.
C.
887
pp.

Church,
M.
R.,
P.
W.
Shaffer,
K.
W.
Thornton,
D.
L.
Cassel,
C.
I.
Liff,
M.
G.
Johnson,
D.
A.
Lammers,
J.
J.
Lee,
G.
R.
Holdren,
J.
S.
Kern,
L.
H.
Liegel,
S.
M.
Pierson,
D.
S.
Stevens,
B.
P.
Rochelle,
and
R.
S.
Turner.
1992.
Direct/
Delayed
Response
Project:
Future
Effects
of
Long­
term
Sulfur
Deposition
on
Stream
Chemistry
in
the
Mid­
Appalachian
Region
of
the
Eastern
United
States
EPA/
600/
R­
92/
186.
U.
S.
Environmental
Protection
Agency,
Washington
D.
C.

Cole,
D.
W.
and
M.
Rapp.
1981.
Elemental
cycling
in
forest
ecosystems.
Pages
341­
409
in
D.
E.
Reichle,
ed.
Dynamic
Properties
of
Forest
Ecosystems.
Cambridge
Press,
New
York.

Cook,
R.
B.,
J.
W.
Elwood,
R.
R.
Turner,
M.
A.
Bogle,
P.
J.
Mulholland,
and
A.
V.
Palumbo.
1994.
Acidbase
chemistry
of
high­
elevation
streams
in
the
Great
Smoky
Mountains.
Water
Air
Soil
Pollut.
72:
331­
356.

Cosby,
B.
J.,
G.
M.
Hornberger,
J.
N.
Galloway,
and
R.
F.
Wright.
1985a.
Modeling
the
effects
of
acid
deposition:
Assessment
of
a
lumped
parameter
model
of
soil
water
and
streamwater
chemistry.
Water
Resour.
Res.
21:
51­
63.

Cosby,
B.
J.,
G.
M.
Hornberger,
J.
N.
Galloway,
and
R.
F.
Wright.
1985b.
Time
scales
of
catchment
acidification:
A
quantitative
model
for
estimating
freshwater
acidification.
Environ.
Sci.
Technol.
19:
1144­
1149.
66
Cosby,
B.
J.,
G.
M.
Hornberger,
E.
B.
Rastetter,
J.
N.
Galloway,
and
R.
F.
Wright.
1986a.
Estimating
catchment
water
quality
response
to
acid
deposition
using
mathematical
models
of
soil
ion
exchange
processes.
Geoderma
38:
77­
95.

Cosby,
B.
J.,
G.
M.
Hornberger,
R.
F.
Wright,
and
J.
N.
Galloway.
1986b.
Modeling
the
effects
of
acid
deposition:
Control
of
long­
term
sulfate
dynamics
by
soil
sulfate
adsorption.
Water
Resour.
Res.
22:
1283­
1292.

Cosby,
B.
J.,
P.
F.
Ryan,
J.
R.
Webb,
G.
M.
Hornberger,
and
J.
N.
Galloway.
1991.
Mountains
of
Western
Virginia.
Pages
297­
318
in
D.
F.
Charles,
ed.
Acidic
Deposition
and
Aquatic
Ecosystems:
Regional
Case
Studies.
Springer­
Verlag,
New
York.

Dennis,
T.
E.
1995.
Acidification
Effect
on
Fish
in
Shenandoah
National
Park.
M.
S.
Thesis,
Department
of
Environmental
Sciences,
University
of
Virginia,
Charlottesville.

Dennis,
T.
E.,
and
A.
J.
Bulger.
1995.
Condition
factor
and
whole­
body
sodium
concentration
in
a
freshwater
fish:
evidence
of
acidification
stress
and
possible
ionoregulatory
over­
compensation.
Water
Air
Soil
Pollut.
85:
377­
382.

Dennis,
T.
E.
,
S.
E.
MacAvoy,
M.
B.
Steg
and
A.
J.
Bulger.
1995.
The
association
of
water
chemistry
variables
and
fish
condition
in
streams
of
Shenandoah
National
Park
(
USA).
Water
Air
Soil
Pollut.
85:
365­
370.

Driscoll,
C.
T.,
and
D.
A.
Schaefer.
1989.
Background
on
nitrogen
processes.
Pages
4.1­
4.12
in:
J.
L.
Malanchuk
and
J.
Nilsson,
eds.
The
Role
of
Nitrogen
in
the
Acidification
of
Soils
and
Surface
Waters.
No.
1989:
10.
Nordic
Council
of
Ministers,
Copenhagen,
Denmark.

Elwood,
J.
W.
1991.
Southeast
overview.
Pages
291­
295
in
D.
F.
Charles,
ed.
Acidic
Deposition
and
Aquatic
Ecosystems:
Regional
Case
Studies.
Springer­
Verlag,
New
York.

Elwood,
J.
W.,
M.
J.
Sale,
P.
R.
Kaufmann,
and
G.
F.
Cada.
1991.
The
Southern
Blue
Ridge
Province.
Pages
319­
366
in
D.
F.
Charles,
ed.
Acidic
Deposition
and
Aquatic
Ecosystems:
Regional
Case
Studies.
Springer­
Verlag,
New
York.

Eshleman,
K.
N.
1988.
Predicting
regional
episodic
acidification
of
surface
waters
using
empirical
techniques.
Water
Resour.
Res.
24:
1118­
1126.
Eshleman,
K.
N.,
T.
D.
Davies,
M.
Tranter,
and
P.
J.
Wigington,
Jr.
1995.
A
two­
component
mixing
model
for
predicting
regional
episodic
acidification
of
surface
waters
during
spring
snowmelt
periods.
Water
Resour.
Res.
31:
1011­
1021.

Eshleman,
K.
N.,
L.
M.
Miller­
Marshall,
and
J.
R.
Webb.
1995.
Long­
term
changes
in
episodic
acidification
of
streams
in
Shenandoah
National
Park,
Virginia
(
U.
S.
A.).
Water
Air
Soil
Pollut.
85:
517­
522.

Fenneman,
N.
M.
1938.
Physiography
of
the
Eastern
United
States.
McGraw­
Hill
Book
Co.,
Inc.,
New
York.

Flum,
T.,
and
S.
C.
Nodvin.
1995.
Factors
affecting
streamwater
chemistry
in
the
Great
Smoky
Mountains,
USA.
Water
Air
Soil
Pollut.
85:
1707­
1712.

Georgia.
1976.
Geologic
Map
of
Georgia.
1:
500,000
scale.
Georgia
Department
of
Natural
Resources,
Atlanta.

Gherini,
S.
A.,
L.
Mok,
R.
J.
Hudson,
G.
F.
Davis,
C.
W.
Chen,
and
R.
A.
Goldstein.
1985.
The
ILWAS
model:
Formulation
and
application.
Water
Air
Soil
Pollut.
26:
425­
459.

Groffman,
P.
M.,
D.
R.
Zak,
S.
Christensen,
A.
Mosier,
and
J.
M.
Tiedje.
1993.
Early
spring
nitrogen
dynamics
in
a
temperate
forest
landscape.
Ecology
74:
1579­
1585.

Hauhs,
M.,
K.
Rost­
Siebert,
G.
Raben,
T.
Paces,
and
B.
Vigerust.
1989.
Summary
of
European
data.
Pages
5­
1
to
5­
37
in:
J.
L.
Malanchuk
and
J.
Nilsson,
eds.
The
Role
of
Nitrogen
in
the
Acidification
of
Soils
and
Surface
Waters.
No.
1989:
10.
Nordic
Council
of
Ministers,
Copenhagen,
Denmark.

Herlihy,
A.
T.,
P.
R.
Kaufmann,
M.
E.
Mitch,
and
D.
D.
Brown.
1990.
Regional
estimates
of
acid
mine
drainage
impact
on
streams
in
the
Mid­
Atlantic
and
Southeastern
United
States.
Water
Air
Soil
Pollut.
50:
91­
107.

Herlihy,
A.
T.,
P.
R.
Kaufmann,
and
M.
E.
Mitch.
1991.
Chemical
characteristics
of
streams
in
the
Eastern
United
States:
II.
Sources
of
acidity
in
acidic
and
low
ANC
streams.
Water
Resour.
Res.
27:
629­
642.

Herlihy,
A.
T.,
P.
R.
Kaufmann,
M.
R.
Church,
P.
J.
Wigington,
Jr.,
J.
R.
Webb,
and
M.
J.
Sale.
1993.
The
effects
of
acidic
deposition
on
streams
in
the
67
Appalachian
Mountain
and
Piedmont
region
of
the
mid­
Atlantic
United
States.
Water
Resources
Res.
29:
2687­
2703.

Hooper,
R.
P.,
and
N.
Christophersen.
1992.
Predicting
episodic
stream
acidification
in
the
southeastern
United
States:
Combining
a
longterm
acidification
model
and
the
end­
member
mixing
concept.
Water
Resour.
Res.
28:
1983­
1990.

Hornberger,
G.
M.,
B.
J.
Cosby,
and
J.
N.
Galloway.
1986.
Modelling
the
effects
of
acid
deposition:
uncertainty
and
spatial
variability
in
estimation
of
long­
term
sulfate
dynamics
in
a
region.
Water
Resour.
Res.
22:
1293­
1302.

Hornberger,
G.
M.,
B.
J.
Cosby,
and
R.
F.
Wright.
1987.
Analysis
of
historical
surface
water
acidification
in
southern
Norway
using
a
regionalized
conceptual
model
(
MAGIC).
In:
M.
B.
Beck,
ed.
Systems
Analysis
in
Water
Quality
Management.
Pergamon
Press,
New
York.

Husar,
R.
B.
1986.
Emissions
of
sulfur
dioxide
and
nitrogen
oxides
and
trends
for
eastern
North
America.
Pages
48­
92
in
Acid
Deposition:
Longterm
Trends.
National
Academy
Press,
Washington,
D.
C.

Husar,
R.
B.,
T.
J.
Sullivan,
and
D.
F.
Charles.
1991.
Historical
trends
in
atmospheric
sulfur
deposition
and
methods
for
assessing
long­
term
trends
in
surface
water
chemistry.
In:
D.
F.
Charles,
ed.
Acidic
Deposition
and
Aquatic
Ecosystems:
Regional
Case
Studies.
Springer­
Verlag,
New
York.

Hyer,
K.
E.,
J.
R.
Webb,
and
K.
N.
Eshleman.
1995.
Episodic
acidification
of
three
streams
in
Shenandoah
National
Park,
Virginia
(
U.
S.
A.).
Water
Air
Soil
Pollut.
85:
523­
528.

Johnson,
D.
W.
1992.
Nitrogen
retention
in
forest
soils.
J.
Environ.
Qual.
21:
1­
12.

Johnson,
D.
W.,
and
S.
E.
Lindberg,
eds.
1992.
Atmospheric
Deposition
and
Forest
Nutrient
Cycling.
Ecological
Studies.
Springer­
Verlag,
New
York.

Jones,
H.
C.,
J.
C.
Noggle,
R.
C.
Young,
J.
M.
Kelley,
H.
Olem,
R.
J.
Ruane,
R.
W.
Pasch,
G.
J.
Hyfantis,
and
W.
J.
Parkhurst.
1983.
Investigations
of
the
cause
of
fishkills
in
fish­
rearing
facilities
in
Raven
Fork
Watershed.
TVA/
ONR/
WR­
83­
9.
Tennessee
Valley
Authority,
Office
of
Natural
Resources,
Division
of
Air
and
Water
Resources,
Knoxville,
TN.
Joslin,
J.
D.,
J.
M.
Kelly,
and
H.
Van
Miegroet.
1992.
Soil
chemistry
and
nutrition
of
North
American
spruce­
fir
stands:
Evidence
for
recent
change.
J.
Environ.
Qual.
21:
12­
30.

Kaufmann,
P.
R.,
A.
T.
Herlihy,
J.
W.
Elwood,
M.
E.
Mitch,
W.
S.
Overton,
M.
J.
Sale,
J.
J.
Messer,
K.
A.
Cougan,
D.
V.
Peck,
K.
H.
Reckhow,
A.
J.
Kinney,
S.
J.
Christie,
D.
D.
Brown,
C.
A.
Hagley,
and
H.
I.
Jager.
1988.
Chemical
Characteristics
of
Streams
in
the
Mid­
Atlantic
and
Southeastern
United
States,
Volume
I:
Population
Descriptions
and
Physico­
Chemical
Relationships.
EPA/
600/
3­
88/
021a.
U.
S.
Environmental
Protection
Agency,
Washington,
D.
C.
397
pp.

Kaufmann,
P.
R.,
A.
T.
Herlihy,
M.
E.
Mitch,
and
W.
S.
Overton.
1991.
Chemical
characteristics
of
streams
in
the
Eastern
United
States:
I.
Synoptic
survey
design,
acid­
base
status
and
regional
chemical
patterns.
Water
Resour.
Res.
27:
611­
627.

Klemedtsson,
L.,
and
B.
H.
Svensson.
1988.
Effects
of
acid
deposition
on
denitrification
and
N2Oemission
from
forest
soils.
Pages
343­
362
in:
J.
Nilsson
and
P.
Grennfelt,
eds.
Critical
Loads
for
Sulphur
and
Nitrogen.
No.
1988:
15.
Nordic
Council
of
Ministers,
Copenhagen,
Denmark.

Landers,
D.
H.,
W.
S.
Overton,
R.
A.
Linthurst,
and
D.
F.
Brakke.
1988.
Eastern
Lake
Survey:
Regional
estimates
of
lake
chemistry.
Environ.
Sci.
Technol.
22:
128­
135.

Lee.
S.
1987.
Uncertainty
Analysis
for
Long­
term
Acidification
of
Lakes
in
Northeastern
U.
S.
A.
PhD
Thesis.
University
of
Iowa,
Iowa
City.

Lynch,
D.
D.,
and
N.
B.
Dise
1985.
Sensitivity
of
Stream
Basins
in
Shenandoah
National
Park
to
Acid
Deposition.
USGS
Water
Resources
Investigations
Report
85­
4115.
U.
S.
Geological
Survey,
Washington
D.
C.

MacAvoy,
S.
E.
and
A.
J.
Bulger.
1995.
Survival
of
brook
trout
(
Salvelinus
fontinalis)
embryos
and
fry
in
streams
of
different
acid
sensitivity
in
Shenandoah
National
Park,
USA.
Water
Air
Soil
Pollut.
85:
439­
444.

Melillo,
J.
M.,
J.
D.
Aber,
P.
A.
Steudler,
and
J.
P.
Schimel.
1983.
Denitrification
potentials
in
a
successional
sequency
of
northern
hardwood
forest
stands.
Environ.
Biogeochem.
Ecol.
Bull.
(
Stockholm)
35:
217­
228.

Miller­
Marshall,
L.
M.
1993.
Mechanisms
Controlling
Variation
in
Stream
Chemical
68
Composition
During
Hydrologic
Episodes
in
the
Shenandoah
National
Park,
Virginia.
M.
S.
Thesis,
Department
of
Environmental
Sciences,
University
of
Virginia,
Charlottesville.
165
pp.

Morgan,
R.
P.,
II,
A.
J.
Janicki,
C.
K.
Murray,
M.
A.
Pawlowski,
and
M.
J.
Pindar.
1990.
Western
Maryland
Stream
Survey:
Relationship
between
fish
distributions,
acidification,
and
watershed
characteristics.
Final
Contract
Report
to
the
Maryland
Department
of
Natural
Resources,
Annapolis,
Md.

Mount,
D.
R.,
J.
R.
Hockett
and
W.
A.
Gern.
1988a.
Effect
of
long­
term
exposure
to
acid,
aluminum,
and
low
calcium
of
adult
brook
trout
(
Salvelinus
fontinalis)
II.
Vitellogenesis
and
osmoregulation.
Can.
J.
Fish.
Aquat.
Sci.
45:
1633­
1642.

Mount,
D.
R.,
C.
G.
Ingersoll,
D.
D.
Gulley,
J.
D.
Fernadez,
T.
W.
LaPoint,
and
H.
L.
Bergman
1988b.
Effect
of
long­
term
exposure
to
acid,
aluminum,
and
low
calcium
of
adult
brook
trout
(
Salvelinus
fontinalis).
I.
Survival,
growth,
fecundity,
and
progeny
survival
Can.
J.
Fish.
Aquat.
Sci.
45:
1623­
1632.

Murdoch,
P.
S.,
and
J.
L.
Stoddard.
1992.
The
role
of
nitrate
in
the
acidification
of
streams
in
the
Catskill
Mountain
of
New
York.
Water
Resour.
Res.
28:
2707­
2720.

National
Academy
of
Sciences
(
NAS).
1984.
Acid
Deposition:
Processes
of
Lake
Acidification.
Summary
of
Discussion.
National
Research
Council
Commission
on
Physical
Sciences,
Mathematics,
and
Resources.
Environmental
Studies
Board,
Panel
on
Processes
of
Lake
Acidification.
National
Academy
Press,
Washington,
D.
C.
11
pp.

NADP/
NTN,
National
Atmospheric
Deposition
Program
(
NRSP­
3)/
National
Trends
Network.
1996.
NADP/
NTN
Coordination
Office,
Natural
Resource
Ecology
Laboratory,
Colorado
State
University,
Fort
Collins,
Colorado.

National
Acid
Precipitation
Assessment
Program
(
NAPAP).
1991.
The
U.
S.
National
Acid
Precipitation
Assessment
Program
1990
Integrated
Assessment
Report.
NAPAP
Office
of
the
Director,
722
Jackson
Place,
N.
W.,
Washington,
D.
C.,
20503.
520
pp.

Neal,
C.,
A.
Robson,
B.
Reynolds,
and
A.
Jenkins.
1992.
Prediction
of
future
short­
term
stream
chemistry
C
a
modeling
approach.
J.
Hydrol.
130:
87­
103.
Newman,
K.
and
A.
Dolloff.
1995.
Response
of
blacknose
dace
(
Rhinichthys
atratulus)
and
brook
char
(
Salvelinus
fontinalis)
to
acidified
water
in
a
laboratory
stream.
Water,
Air,
and
Soil
Pollution
85:
371­
376.

Nikolaidis,
N.
P.,
H.
Rajaram,
J.
L
Schnoor,
and
K.
P.
Georgakakos.
1988.
A
generalized
soft
water
acidification
model.
Water
Resour.
Res.
24:
1983­
1996.

Nodvin,
S.
C.,
H.
Van
Miegroet,
S.
E.
Lindberg,
N.
S.
Nicholas,
and
D.
W.
Johnson.
1995.
Acidic
deposition,
ecosystem
processes,
and
nitrogen
saturation
in
a
high
elevation
Southern
Appalachian
watershed.
Water
Air
Soil
Pollut.
85:
1647:
1652.

North
Carolina.
1985.
Geologic
Map
of
North
Carolina.
Department
of
Natural
Resources
and
Community
Development,
Raleigh.

Osborne,
W.
E.,
M.
W.
Szabo,
C.
W.
Copeland,
and
T.
L.
Neathery.
1989.
Geologic
Map
of
Alabama,
1:
500,000
scale,
Geological
Survey
of
Alabama
Special
Map
221.

Peper,
J.
D.,
A.
E.
Grosz,
T.
H.
Kress,
T.
K.
Collins,
G.
B.
Kappesser,
C.
M.
Hyber,
and
J.
R.
Webb.
1995.
Acid
Deposition
Sensitivity
Map
of
the
Southern
Appalachian
Assessment
Area:
Virginia,
North
Carolina,
South
Carolina,
Tennessee,
Georgia,
and
Alabama.
U.
S.
Geological
Survey
On­
line
Digital
Data
Series
Open
File
Report.

Post,
W.
M.,
J.
Pastor,
P.
J.
Zinke,
and
A.
G.
Stangenberger.
1985.
Global
patterns
of
soil
nitrogen
storage.
Nature
317:
613­
616.

Rastetter,
E.
B.,
M.
G.
Ryan,
G.
R.
Shaver,
J.
M.
Melillo,
J.
Nadelhoffer,
J.
E.
Hobbie
and
J.
D.
Aber.
1991.
A
general
biogeochemical
model
describing
the
reponses
of
C
and
N
cycles
in
terrestrial
ecosystems
to
changes
in
CO2,
climate,
and
N
deposition.
Tree
Physiology
9:
101­
126.

Rosemond,
A.
D.,
Reice,
S.
R.,
Elwood,
J.
W.,
and
Mulholland,
P.
J.
1992.
The
effects
of
stream
acidity
on
benthic
invertebrate
communities
in
the
southeastern
United
States.
Fresh.
Biol.
27(
2):
193­
209.

Ryan,
P.
F.,
J.
N.
Galloway,
B.
J.
Cosby,
G.
M.
Hornberger,
and
J.
R.
Webb.
1989.
Changes
in
the
chemical
composition
of
streamwater
in
two
catchments
in
the
Shenandoah
National
Park,
Virginia,
in
response
to
atmospheric
deposition
of
sulfur.
Water
Resour.
Res.
25:
2091­
2099.
69
Schnoor,
J.
L.,
N.
P.
Nikolaidis,
and
G.
E.
Glass.
1986.
Lake
resources
at
risk
to
acidic
deposition
in
the
Upper
Midwest.
J.
Water
Pollut.
Control
Fed.
58:
139­
148.

Sharpe,
W.
E.,
V.
G.
Leibfried,
W.
G.
Kimmel
and
D.
R.
DeWalle.
1987.
The
relationship
of
water
quality
and
fish
occurrence
to
soils
and
geology
in
an
area
of
high
hydrogen
and
sulfate
ion
deposition.
Water
Resour.
Bull.
23:
37­
46.

Shubzda,
J.,
S.
E.
Lindberg,
C.
T.
Garten,
and
S.
C.
Nodvin.
1995.
Elevational
trends
in
the
fluxes
of
sulphur
and
nitrogen
in
throughfall
in
the
Southern
Appalachian
mountains:
Some
surprising
results.
Water
Air
Soil
Pollut.
85:
2265­
2270.

Silsbee,
D.
G.,
and
G.
L.
Larson.
1982.
Water
quality
of
streams
in
the
Great
Smoky
Mountains
National
Park,
Hydrobiologia
89:
97­
115.

Skeffington,
R.
A.,
and
E.
J.
Wilson.
1988.
Excess
nitrogen
deposition:
Issues
for
consideration.
Environ.
Pollut.
54:
159­
184.

Stoddard,
J.
L.
1991.
Trends
in
Catskill
stream
water
quality:
Evidence
from
historical
data.
Water
Resour.
Res.
27:
2855­
2864.

Stoddard,
J.
L.
1994.
Long­
term
changes
in
watershed
retention
of
nitrogen:
Its
causes
and
aquatic
consequences.
Pages
223­
284
in
L.
A.
Baker,
ed.
Environmental
Chemistry
of
Lakes
and
Reservoirs.
Advances
in
Chemistry
Series,
No.
237.
American
Chemical
Society,
Washington,
DC.

Sullivan,
T.
J.
1990.
Historical
Changes
in
Surface
Water
Acid­
Base
Chemistry
in
Response
to
Acidic
Deposition.
NAPAP
Report
11.
In:
National
Acid
Precipitation
Assessment
Program,
Acidic
Deposition:
State
of
Science
and
Technology.
Volume
II.
National
Acid
Precipitation
Assessment
Program,
Washington,
D.
C.

Sullivan,
T.
J.,
R.
S.
Turner,
D.
F.
Charles,
B.
F.
Cumming,
J.
P.
Smol,
C.
L.
Schofield,
C.
T.
Driscoll,
B.
J.
Cosby,
H.
B.
J.
Birks,
A.
J.
Uutala,
J.
C.
Kingston,
S.
S.
Dixit,
J.
A.
Bernert,
P.
F.
Ryan,
and
D.
R.
Marmorek.
1992.
Use
of
historical
assessment
for
evaluation
of
process­
based
model
projections
of
future
environmental
change:
Lake
acidification
in
the
Adirondack
mountains,
New
York,
USA.
Environ.
Pollut.
77:
253­
262.

Swank,
W.
T.,
and
J.
B.
Waide.
1988.
Characterization
of
baseline
precipitation
and
stream
chemistry
and
nutrient
budgets
for
control
watersheds.
Pages
57­
79
in
D.
A.
Crossley,
Jr.,
ed.
Forest
Hydrology
and
Ecology
at
Coweeta.
Springer­
Verlag,
New
York.

Thornton,
K.
W.,
D.
Marmorek,
and
P.
F.
Ryan.
1990.
Methods
for
Projecting
Future
Changes
in
Surface
Water
Acid­
Base
Chemistry.
NAPAP
Report
14.
In:
National
Acid
Precipitation
Assessment
Program,
Acidic
Deposition:
State
of
Science
and
Technology.
Volume
II.
National
Acid
Precipitation
Assessment
Program,
Washington,
D.
C.

Turner,
R.
S.,
and
others.
1990.
Watershed
and
Lake
Processes
Affecting
Surface
Water
Acid­
Base
Chemistry.
Report
10.
In:
National
Acid
Precipitation
Assessment
Program,
Acidic
Deposition:
State
of
Science
and
Technology.
Volume
II.
National
Acid
Precipitation
Assessment
Program,
Washington,
D.
C.

U.
S.
Environmental
Protection
Agency.
1995
Acid
Deposition
Standard
Feasibility
Study:
Final
Report
to
Congress.
U.
S.
Environmental
Protection
Agency,
EPA/
430/
R­
95/
001a,
Washington,
DC,
120
pp.

Van
Sickle,
J.
and
M.
R.
Church.
1995.
Methods
for
Estimating
the
Relative
Effects
of
Sulfur
and
Nitrogen
Deposition
on
Surface
Water
Chemistry.
U.
S.
Environmental
Protection
Agency,
EPA/
600/
R­
95/
172,
Washington,
DC,
121pp.

Virginia.
1963.
Geologic
Map
of
Virginia.
Department
of
Conservation
and
Economic
Development,
Commonwealth
of
Virginia,
Richmond.

Vitousek,
P.
M.,
and
R.
W.
Howarth.
1991.
Nitrogen
limitation
on
land
and
in
the
sea:
How
can
it
occur?
Biogeochemistry
13:
87­
115.

Vitousek,
P.
M.
and
W.
A.
Reiners.
1975.
Ecosystem
succession
and
nutrient
retention:
A
hypothesis.
Bioscience
25:
376­
381.

Ward,
G.
M.
1991.
A
Longitudinal
Survey
of
the
Water
Chemistry
of
Sipsey
Fork,
Bankhead
National
Forest,
Alabama.
Final
Report
for
U.
S.
Forest
Service,
Contract
4­
4146­
0369.

Ward,
G.
M.
1992.
A
Longitudinal
Survey
of
the
Water
Chemistry
of
Sipsey
Fork,
Bankhead
National
Forest,
Alabama.
Final
Report
for
U.
S.
Forest
Service,
Order
Number
40­
4146­
2­
0362.

Ward,
G.
M.
1993.
A
Survey
of
the
Water
Chemistry
of
Sipsey
Fork,
Bankhead
National
Forest,
70
Alabama.
Final
Report
for
U.
S.
Forest
Service
Order
Number
40­
4146­
3­
0364.

Webb,
J.
R.
1995.
Synoptic
Surveys
of
Surface
Water
Chemical
Conditions
in
Dolly
Sods
and
Otter
Creek
Wilderness
Areas:
Data
Report.
Report
submitted
to
U.
S.
Forest
Service,
Monongahela
National
Forest,
Elkins,
West
Virginia.

Webb,
J.
R.,
B.
J.
Cosby,
J.
N.
Galloway,
and
G.
M.
Hornberger.
1989a.
Acidification
of
native
brook
trout
streams
in
Virginia,
Water
Resour.
Res.
25:
1367­
1377.

Webb,
J.
R.,
P.
E.
Bugas,
B.
J.
Cosby,
J.
N.
Galloway,
G.
M.
Hornberger,
J.
W.
Kaufman,
L.
O.
Mohn,
P.
F.
Ryan
and
P.
P.
Smith.
1989b.
Acidic
deposition
and
the
status
of
Virginia's
wild
trout
resource.
pp
228­
233.
In:
Symposium
Proceedings,
Wild
Trout
IV,
U.
S.
Fish
and
Wildlife
Service.

Webb,
J.
R.,
B.
J.
Cosby,
K.
Eshleman,
and
J.
Galloway.
1995.
Change
in
the
acid­
base
status
of
Appalachian
Mountain
catchments
following
forest
defoliation
by
the
Gypsy
Moth.
Water
Air
Soil
Pollut.
85:
535­
540.

West
Virginia.
1968.
Geologic
Map
of
West
Virginia.
1:
250,000
scale.
West
Virginia
Geological
and
Economic
Survey,
Charleston.

Wigington,
P.
J.,
Jr.,
T.
D.
Davies,
M.
Tranter,
and
K.
N.
Eshleman.
1990.
Episodic
Acidification
of
Surface
Waters
Due
to
Acidic
Deposition.
NAPAP
Report
12.
In:
National
Acid
Precipitation
Assessment
Program,
Acidic
Deposition:
State
of
Science
and
Technology.
Volume
II.
National
Acid
Precipitation
Assessment
Program,
Washington,
D.
C.
Wigington,
P.
J.,
Jr.,
et
al.
1993.
Episodic
Acidification
of
Streams
in
the
Northeastern
United
States:
Chemical
and
Biological
Results
of
the
Episodic
Response
Project.
EPA/
600/
R­
93/
190.
U.
S.
Environmental
Protection
Agency,
Washington,
D.
C.

Winger,
P.
V.,
and
others.
1987.
Sensitivity
of
highelevation
streams
in
the
Southern
Blue
Ridge
Province
to
acidic
deposition.

Zurbuch,
P.
E.,
R.
Menendez,
and
J.
E.
Woodrum.
1986.
West
Virginia
Dogway
Fork
Project
Cooperative
Acid
Precipitation
Mitigation
Program,
Annual
Report
1986.
Cooperative
Agreement
No.
14­
16­
009­
85­
945.
Report
from
the
West
Virginia
Department
of
Natural
Resources
to
the
Eastern
Energy
and
Land
Use
Team,
US
Fish
and
Wildlife
Service,
Kearneysville,
WV.
71
9.
ANNOTATED
BIBLIOGRAPHY
This
annotated
bibliography
provides
up­
to­
date
information
on
papers
published
since
the
1990
NAPAP
SOS/
T
reports,
focusing
particularly
on
articles
of
regional
significance
to
SAMI
and
on
those
that
present
significant
new
general
findings.

Baker,
J.
P.,
J.
Van
Sickle,
C.
J.
Gagen,
D.
R.
DeWalle,
W.
E.
Sharpe,
R.
F.
Carline,
B.
P.
Baldigo,
P.
S.
Murdoch,
D.
W.
Bath,
W.
A.
Kretser,
H.
A.
Simonin,
and
P.
J.
Wigington,
Jr.
1996.
Episodic
acidification
of
small
streams
in
the
northeast
United
States:
Effects
on
fish
populations.
Ecol.
Appl.
6:
422­
437.

This
paper
largely
summarizes
the
results
from
the
experimental
field
studies
of
fish
populations
in
the
13
ERP
streams
first
reported
on
by
Wigington
et
al.
(
1993).

Baker,
L.
A.,
A.
T.
Herlihy,
P.
R.
Kaufmann,
and
J.
M.
Eilers.
1991.
Acidic
lakes
and
streams
in
the
United
States:
The
role
of
acidic
deposition.
Science
252:
1151­
1154.

A
short
article
taken
from
the
NAPAP
SOS/
T
report
9
examining
the
likely
causes
of
acidity
in
lakes
and
streams.
It
also
summarizes
the
number
and
proportions
of
lakes
and
streams
in
the
U.
S.
likely
to
be
acidic
as
a
result
of
acidic
deposition.

Bazemore,
D.
E.,
K.
N.
Eshleman,
and
K.
J.
Hollenbeck.
1994.
The
role
of
soil
water
in
stormflow
generation
determined
from
natural
tracer
and
hydrometric
techniques.
J.
Hydrol.
162:
47­
75.

A
field
study
that
combined
natural
tracer
methods
and
hydrometric
observations
was
conducted
to
estimate
the
contributions
of
pre­
event
soil
water
to
stormflow
in
a
small
forested
catchment
in
Shenandoah
National
Park,
Virginia.
A
three­
component
model
using
two
naturally
occurring
tracers
(
oxygen­
18
and
chloride)
was
used
to
show
that
pre­
event
soil
water
dominated
total
storm
runoff
and
peak
runoff
during
two
major
rainfall
events.
The
authors
also
illustrated
that
quantitative
error
analyses
are
advisable
in
chemical
and
isotopic
hydrograph
separation
studies.
Field
observations
suggested
that
the
pre­
event
water
response
could
be
attributed
to
a
threshold­
type
expansion
of
hillslope
source
areas.
Results
of
the
study
are
significant
for
understanding
episodic
acidification,
because
the
chemical
composition
of
the
three
end­
members
(
soil
water,
groundwater,
and
throughfall)
may
vary
dramatically
in
small
headwater
systems.

Bishop,
K.
H.
1991.
Episodic
Increases
in
Stream
Acidity,
Catchment
Flow
Pathways
and
Hydrograph
Separation.
PhD
dissertation.
Department
of
Geography,
University
of
Cambridge,
U.
K.

Detailed
field
investigations
(
including
hydroisotopic
separations
using
oxygen­
18)
were
conducted
at
Loch
Fleet
(
Scotland)
and
Svartberget
(
Sweden)
to
test
the
hypothesis
that
acid
episodes
are
due
primarily
to
the
rapid
displacement
of
pre­
event
water
from
the
catchment
during
stormflow
events.
This
hypothesis
was
confirmed
by
the
field
investigation,
although
the
study
was
unable
to
identify
the
specific
mechanisms
of
runoff
generation
(
e.
g.,
saturation
overland
flow,
subsurface
stormflow)
in
the
two
watersheds.
The
author
also
concluded
that
non­
Darcian
flow
(
i.
e.,
macropore
flow)
could
explain
the
field
observations
of
rapid
displacement
of
pre­
event
water
and
increases
in
stream
acidity
during
storm
events.

Bradford,
D.,
C.
Swanson,
and
M.
Gordon.
1992.
Effects
of
low
pH
and
aluminum
on
two
declining
species
of
amphibians
in
the
Sierra
Nevada,
California.
Jour.
of
Herpetology
26:
369­
377.

Bradford,
D.
F.,
and
M.
S.
Gordon.
1992.
Aquatic
amphibians
in
the
Sierra
Nevada:
Current
status
and
potential
effects
of
acidic
deposition
on
populations
Final
report,
Contract
No.
A932­
139,
California
Air
Resources
Board,
Sacramento,
CA.

Bradford,
D.
F.,
C.
Swanson,
and
M.
Gordon.
1994.
Effects
of
low
pH
and
aluminum
on
amphibians
at
high
elevation
in
the
Sierra
Nevada,
California.
Can.
J.
Zool.
72:
1272­
1279.

Bradford,
D.
F.,
M.
Gordon,
D.
Johnson,
R.
Andrews,
and
W.
B.
Jennings.
1994.
Acidic
deposition
as
an
unlikely
cause
of
amphibian
population
declines
in
the
Sierra
Nevada,
California.
Biological
Conservation
69:
155­
161.

Toxicity
testing
indicated
that
amphibians
are
at
little
risk
from
low
pH
in
water
acidified
to
a
pH
level
of
5.0,
with
Al
concentrations
of
39
B
80

g/
L.
However,
the
authors
observed
sublethal
effects
(
reduced
growth
rate
and
earlier
hatching)
for
pH
as
high
as
5.25
at
the
Al
concentrations
tested.
Field
survey
findings
implied
that
acidic
deposition
is
unlikely
to
have
been
a
cause
of
72
recent
amphibian
population
declines
in
the
Sierra
Nevada.

Bulger,
A.
J.,
L.
Lien,
J.
Cosby,
and
A.
Henriksen.
1993.
Brown
trout
(
Salmo
trutta)
status
and
chemistry
from
the
Norwegian
Thousand
Lake
Survey:
Statistical
Analysis.
Can.
J.
Fish.
Aquat.
Sci.
50:
575­
585.

The
relationship
between
atmospheric
sulfate
and
terrestrial
calcium
is
the
main
control
of
acidification
processes,
but
neither
one
is
a
good
predictor
of
fish
response;
this
paper
provides
a
useful
explanation
of
this
as
a
general
phenomenon,
not
limited
to
Norway.
It
also
provides
an
explanation
linking
toxic
mechanisms
and
empirical
responses
of
fish
populations
in
nature.

Bulger,
A.
J.,
C.
A.
Dolloff,
B.
J.
Cosby,
K.
N.
Eshleman,
J.
R.
Webb,
and
J.
N.
Galloway.
1995.
The
"
Shenandoah
National
Park:
Fish
in
Sensitive
Habitats
(
SNP:
FISH)"
Project:
An
integrated
assessment
of
fish
community
responses
to
stream
acidification.
Water
Air
Soil
Pollut.
85:
309­
314
This
paper
presents
a
summary
to
date
of
the
overall
findings
of
a
three­
year
project
on
fish
community
effects
and
acid­
base
chemistry
in
mountains
streams
in
Virginia.
Both
chronic
and
episodic
acidification
are
occurring
in
these
streams.
Biological
differences
in
low
ANC
versus
high
ANC
streams
include
fish
species
richness,
population
density,
condition
factor,
age,
size,
and
bioassay
survival.
Predictive
models
relating
fish
status
to
future
water
chemistry
will
be
produced
by
this
project.

Carline,
R.
F.,
D.
R.
Dewalle,
W.
E.
Sharpe,
B.
A.
Dempsey,
C.
J
Gagen,
and
B.
Swistock.
1992.
Water
chemistry
and
fish
community
responses
to
episodic
stream
acidification
in
Pennsylvania,
USA.
Environ.
Pollut.
78:
45­
48.

Five
undisturbed,
wooded
streams
were
studied
on
the
Northern
Appalachian
Plateau
of
Pennsylvania
for
9
months
to
determine
chemical
changes
and
fish
responses
that
occur
during
episodic
storm
runoff.
Wild
brook
trout
were
found
in
all
streams,
although
only
a
remnant
population
existed
in
the
most
acidic
stream.
Sculpins
(
Cottus
bairdi
or
C.
cognatus)
were
collected
only
in
the
two
streams
with
the
least
severe
episodes.
Mortality
of
brook
trout
and
sculpins
in
in­
situ
bioassays
ranged
from
0
to
about
80%
among
streams
during
acidic
episodes
and
was
positively
related
to
concentrations
of
total
dissolved
Al.
Some
displaced
trout
were
found
near
groundwater
seeps,
where
pH
was
higher
and
dissolved
Al
lower
than
in
the
main
channel.

Church,
M.
R.,
and
J.
Van
Sickle.
1995.
Potential
Relative
Effects
of
Nitrogen
and
Sulfur
Deposition
on
Regional
Scale
Acidification
of
Surface
Waters
in
the
Eastern
United
States.
U.
S.
Environmental
Protection
Agency,
Washington,
DC.

The
Nitrogen
Bounding
Study
used
the
MAGIC
model
to
predict
the
future
acid­
base
status
of
regional
populations
of
surface
water
sites
under
several
different
scenarios
for
future
nitrogen
and
sulfur
deposition,
and
several
different
time
series
for
nitrogen
saturation.

Church,
M.
R.,
P.
W.
Shaffer,
K.
N.
Eshleman,
and
B.
P.
Rochelle.
1990.
Potential
future
effects
of
current
levels
of
sulfur
deposition
on
stream
chemistry
in
the
Southern
Blue
Ridge
Mountains,
United
States.
Water
Air
Soil
Pollut.
1990:
39­
48.

This
article
concludes
that,
although
little
change
in
surface
water
chemistry
is
likely
to
have
occurred
due
to
acidic
deposition
in
the
region
to
date,
soils
are
currently
retaining
a
majority
of
atmospherically
deposited
S,
and
it
is
likely
that
marked
increases
in
surface
water
SO4
2

will
occur.
Such
increases
could
be
accompanied
by
significant
surface
water
acidification
(
loss
of
ANC).

Church,
M.
R.,
P.
W.
Shaffer,
K.
W.
Thornton,
D.
L.
Cassell,
C.
I.
Liff,
M.
G.
Johnson,
D.
A.
Lammers,
J.
J.
Lee,
G.
R.
Holdren,
J.
S.
Kern,
L.
H.
Liegel,
S.
M.
Pierson,
D.
L.
Stevens,
B.
P.
Rochelle,
and
R.
S.
Turner.
1992.
Direct/
Delayed
Response
Project:
Future
Effects
of
Long­
Term
Sulfur
Deposition
on
Stream
Chemistry
in
the
Mid­
Atlantic
Region
of
the
Eastern
United
States.
EPA/
600/
R­
92/
186.
U.
S.
Environmental
Protection
Agency,
Washington,
D.
C.

This
is
the
last
in
the
series
of
DDRP
reports
detailing
the
results
of
empirical
and
dynamic
modeling
of
acidification
of
the
regional
population
of
surface
waters
in
several
regions.
For
this
report,
the
results
of
the
Appalachian
Plateau,
Valley
and
Ridge,
and
Piedmont
provinces
are
presented
together.

Cook,
R.
B.,
J.
W.
Elwood,
R.
R.
Turner,
M.
A.
Bogle,
P.
J.
Mulholland,
and
A.
V.
Palumbo.
1994.
Acidbase
chemistry
of
high­
elevation
streams
in
the
Great
Smoky
Mountains.
Water
Air
Soil
Pollut.
72:
331­
356.
73
Cook
et
al.
report
on
seasonal
data
from
eight
streams
in
Great
Smoky
Mountains
National
Park
on
differing
types
of
bedrock.
They
point
out
both
the
effects
of
the
Anakeesta
formation
on
stream
chemistry
(
high
SO4
2

and
acidic
waters)
and
the
effects
of
high­
elevation,
old
growth
forests
on
NO3

concentrations.

Davies,
T.
D.,
P.
J.
Wigington,
Jr.,
M.
Tranter,
and
K.
N.
Eshleman.
1992.
`
Acidic
episodes'
in
surface
waters
in
Europe.
J.
Hydrol.
132:
25­
69.

This
paper
largely
presents
a
summary
of
Section
3.2
of
Wigington
et
al.
(
1990),
NAPAP
SOS/
T
Report
12.

Dennis,
T.
E.,
and
A.
J.
Bulger.
1995.
Condition
factor
and
whole­
body
sodium
concentration
in
a
freshwater
fish:
evidence
of
acidification
stress
and
possible
ionoregulatory
over­
compensation.
Water
Air
Soil
Pollut.
85:
377­
382.

Acidification
negatively
affects
ion
regulation
in
fish.
This
paper
shows
that
fish
in
a
low
ANC
stream
appear
to
be
able
to
ion­
regulate
adequately
at
current
acid­
base
conditions,
but
it
also
suggests
that
the
added
metabolic
cost
of
ion
regulation
in
low
ANC
streams
versus
high
ANC
streams
my
be
responsible
for
slower
growth
and
smaller
sized
individuals
in
the
low
ANC
stream.

Dennis,
T.
E.
,
S.
E.
MacAvoy,
M.
B.
Steg
and
A.
J.
Bulger.
1995.
The
association
of
water
chemistry
variables
and
fish
condition
in
streams
of
Shenandoah
National
Park
(
USA).
Water
Air
Soil
Pollut.
85:
365­
370.

This
paper
demonstrates
a
strong
relationship
between
water
chemistry
and
fish
condition
factor,
across
nine
montane
streams
in
Virginia.
The
variables
most
strongly
related
to
fish
condition
were
those
associated
with
ANC
and
SO4
2

,
with
condition
factor
poorest
in
low
ANC
streams.

DiStefano,
R.
J.,
R.
J.
Neves,
L.
A.
Helfrich,
and
M.
C.
Lewis.
1991.
Response
of
the
crayfish
Cambarus
bartonii
bartonii
to
acid
exposure
in
southern
Appalachian
streams.
Can.
J.
Zool.
69(
6):
1585­
1591.

Tolerance
of
crayfish
to
acidity
was
rather
high
(
96
hr
LC50,
range
2.43
B
2.85
for
early
juveniles
to
adults).
Lowering
the
water
temperature
increased
acid
tolerance.
Nevertheless,
the
authors
state
that
gradually
increasing
acidity
and
loss
of
watershed
buffering
capacity
could
produce
sublethal
effects,
such
as
altered
reproductive
activity
or
changes
in
early
life
history
stages
and
more
sensitive
molt
cycle
stages,
that
could
damage
crayfish
populations.

Dow,
C.
L.,
D.
R.
DeWalle,
J.
A.
Lynch,
and
W.
E.
Sharpe.
1994.
Blizzard's
effects
on
Appalachian
stream
chemistry
assessed.
EOS
Trans.
Amer.
Geophys.
Union
75:
389­
390.

This
short
paper
provides
a
description
of
the
hydrochemical
changes
observed
in
three
small
streams
in
Pennsylvania
during
an
extreme
rainon
snow
event
that
followed
the
"
blizzard
of
1993."
This
storm
deposited
approximately
90
cm
of
snow
on
the
study
catchments
and
was
followed
by
about
7
cm
of
rainfall,
which
caused
the
highest
sustained
(
monthly)
discharge
in
the
Susquehanna
River
basin.
ANC
and
pH
depressions
(
and
elevated
Al
concentrations)
during
the
event
were
quite
large,
although
their
absolute
magnitude
was
not
found
to
vary
significantly
from
previously
monitored
conditions
of
lesser
(
hydrological)
magnitude.
Nitrate
and
organic
anion
concentrations
were
significantly
higher
than
those
recorded
previously,
however,
which
supported
the
explanation
that
flows
were
attributable
to
movement
of
water
through
shallow
soil
horizons.

Downey,
D.
M.,
C.
R.
French,
and
M.
Odom.
1994.
Low
cost
limestone
treatment
of
acid
sensitive
trout
streams
in
the
Appalachian
Mountains
of
Virginia.
Water
Air
Soil
Pollut.
77:
49­
77.

Based
on
their
results
in
three
streams,
the
authors
suggest
that
approximately
88%
of
native
trout
streams
in
Virginia,
which
average
29

eq/
L
ANC
reduction
from
acid
deposition,
could
be
temporarily
restored
using
single
application
liming.

Driscoll,
C.
T.,
and
R.
Van
Dreason.
1993.
Seasonal
and
long­
term
temporal
patterns
in
the
chemistry
of
Adirondack
lakes.
Water
Air
Soil
Pollut.
67:
319­
344.

A
long­
term
(
1982­
present)
monitoring
program
of
17
Adirondack
lakes
was
conducted
to
examine
temporal
patterns
in
acid­
base
chemistry.
The
monitoring
results
indicated
relatively
uniform
SO4
2

concentrations
among
the
lakes,
while
ANC
variations
were
largely
determined
by
differences
in
base
cations.
Lakes
in
the
southern
and
western
Adirondacks
showed
the
highest
levels
of
NO3

,
during
both
peak
and
baseflow
conditions.
From
an
episodic
acidification
perspective,
this
paper
is
significant
since
it
concluded
that
Adirondack
lakes
with
ANC
values
<
100

eq/
L
74
during
baseflow
conditions
may
experience
decreases
in
ANC
to
values
near
or
below
0.

Eckhardt,
B.
W.
and
T.
R.
Moore.
1990.
Controls
on
dissolved
organic
carbon
concentrations
in
streams,
southern
Quebec.
Can.
J.
Fish.
Aquat.
Sci.
47(
8):
1537­
1544.

DOC
concentrations
in
42
streams
draining
small
catchments
in
the
Appalachian
Uplands
and
St.
Lawrence
Lowlands
were
consistently
related
to
the
percentage
of
wetland
in
the
catchment.
The
results
indicate
that
stream
DOC
concentrations
may
be
predicted
from
easily
obtained
catchment
variables,
such
as
percent
wetland.

Englin,
J.
E.,
T.
A.
Cameron,
R.
E.
Mendelsohn,
G.
A.
Parsons,
and
S.
A.
Shankle.
1991.
Valuation
of
damages
to
recreational
trout
fishing
in
the
upper
Northeast
due
to
acidic
deposition.
Technical
Report.
NTIS
Order
No.
DE91012029/
GAR.
Battelle
Pacific
Northwest
Laboratories.

This
report
documents
methods
used
by
NAPAP
to
model
the
economic
value
of
changes
in
recreational
fishing
due
to
acidic
deposition.

Eshleman,
K.
N.
1992.
Comment
on
"
The
episodic
acidification
of
Adirondack
lakes
during
snowmelt
by
Douglas
A.
Schaefer
et
al.
Water
Resour.
Res.
28:
2869­
2873.

This
paper
is
a
technical
comment
on
an
earlier
paper
published
in
Water
Resources
Research
that
dealt
with
the
significance
of
the
linear
relationship
observed
by
Eshleman
(
1988)
between
ANCindex
and
ANCminimum
for
a
group
of
lakes
in
the
Adirondacks
of
New
York.
The
author
argued
that
the
observed
linear
relationship
is
statistically
and
physically
significant
and
can
be
applied
to
an
entire
population
of
lakes
in
the
Adirondacks.

Eshleman,
K.
N.
1995.
Predicting
Regional
Episodic
Acidification
of
Streams
in
Western
Maryland.
Report
CBRM­
AD­
95­
7.
Maryland
Department
of
Natural
Resources,
Annapolis.

This
report
used
the
empirical
linear
regression
model
first
proposed
by
Eshleman
(
1988)
and
data
from
two
studies
of
episodic
acidification
in
the
Appalachian
Plateau
to
provide
an
estimate
of
the
number
and
length
of
streams
in
western
Maryland
that
are
episodically
acidic
(
ANC
<
0)
during
stormflow
periods.
The
report
concluded
that
about
50%
more
streams
(
11%
of
the
population)
are
acidic
during
episodes
than
are
acidic
during
spring
baseflow
conditions
(
7%
of
the
population)
in
the
region;
the
population
used
corresponds
to
those
streams
sampled
during
the
Maryland
Synoptic
Stream
Chemistry
Survey
(
MSSCS)
in
1987.
An
important
uncertainty
is
whether
the
linear
regression
model
can
be
applied
across
all
bedrock
types
found
in
western
Maryland.

Eshleman,
K.
N.,
P.
J.
Wigington,
Jr.,
M.
Tranter,
and
T.
D.
Davies.
1992.
Modelling
episodic
acidification
of
surface
waters:
The
state
of
the
science.
Environ.
Pollut.
77:
287­
295.

This
paper
largely
summarizes
Chapter
5
of
Wigington
et
al.
(
1990),
NAPAP
SOS/
T
Report
12.

Eshleman,
K.
N.,
P.
J.
Wigington,
Jr.,
T.
D.
Davies,
and
M.
Tranter.
1995.
A
two­
component
mixing
model
for
predicting
regional
episodic
acidification
of
surface
waters
during
spring
snowmelt.
Water
Resour.
Res.
31:
1011­
1021.

A
two­
component
mixing
model
of
ANC
was
used
to
explain
two
observed
features
related
to
the
episodic
acidification
of
Adirondack
and
Catskill
mountain
surface
waters
during
spring
snowmelt:
(
1)
maximum
episodic
declines
in
ANC
are
largest
in
high
ANC
systems
and
increase
linearly
with
antecedent
ANC
and
(
2)
relative
depressions
in
ANC
attributable
to
dilution
of
base
cations
are
larger
in
high
ANC
systems.
The
model
was
calibrated
using
snowmelt
data
for
10
Adirondack
lakes
and
was
then
applied
to
the
regional
population
of
lakes
described
by
the
National
Lake
Survey.
The
model
was
also
linked
to
an
empirical
acidification
model
for
predicting
the
future
extent
of
episodically
acidic
(
ANC
<
0)
lakes
in
the
Adirondacks,
given
various
emissions
control
strategies.
Model
predictions
indicated
that
40%
reductions
in
sulfuric
acid
concentrations
will
not
restore
to
positive
values
the
ANC
of
all
lakes
that
are
currently
acidic
during
spring
snowmelt.

Eshleman,
K.
N.,
L.
M.
Miller­
Marshall,
and
J.
R.
Webb.
1995.
Long­
term
changes
in
episodic
acidification
of
streams
in
Shenandoah
National
Park,
Virginia
(
U.
S.
A.).
Water
Air
Soil
Pollut.
85:
517­
522.

A
statistical
analysis
of
13
years
of
daily
discharge
data
and
weekly
streamwater
composition
data
for
White
Oak
Run
(
WOR)
in
Shenandoah
National
Park
was
performed
in
order
to
quantify
episodic
changes
in
composition
and
to
identify
long­
term
trends
in
episodic
acidification
attributable
to
both
natural
and
anthropogenic
processes.
An
objective
hydrological
separation
technique
was
75
used
to
ident
i
fy
mor
e
than
100
"
stormflow/
baseflow
pairs"
in
the
database,
from
which
episodic
chemical
changes
could
be
quantified.
Univariate
statistical
analysis
suggested
that
mean
episodic
depressions
of
ANC
in
WOR
have
increased
by
about
a
factor
of
2
since
the
first
outbreak
of
forest
defoliation
by
the
gypsy
moth
caterpillar
during
the
summer
of
1990;
in
addition,
the
mean
episodic
change
in
nitrate
concentration
has
increased
by
about
12

eq/
L,
while
the
mean
episodic
dilution
of
CB
has
decreased
from

8.5

eq/
L
to

1.7

eq/
L
during
the
same
period.
Episodic
changes
in
SO4
2

have
remained
the
same,
however.
The
results
indicate
that
natural
processes
such
as
insect
defoliations
can
contribute
to
episodic
acidification
through
mobilization
of
NO3

.

Flum,
T.,
and
S.
C.
Nodvin.
1995.
Factors
affecting
streamwater
chemistry
in
the
Great
Smoky
Mountains,
USA.
Water
Air
Soil
Pollut.
85:
1707­
1712.

Flum
et
al.
summarize
the
results
of
synoptic
stream
chemistry
surveys
carried
out
in
Great
Smoky
Mountains
National
Park
during
the
1990s.
They
focus
on
differences
among
different
types
of
bedrock
and
former
land
use
types
(
e.
g.,
old
growth
vs.
second
growth
forests)
and
across
elevations.
ANC
is
lowest
and
NO3

is
highest
in
the
least
disturbed
(
i.
e.,
old
growth)
and
highest
elevation
watersheds.

Gagen,
C.
J.,
W.
E.
Sharpe,
and
R.
F.
Carline.
1994.
Downstream
movement
and
mortality
of
brook
trout
(
Salvelinus
fontinalis)
exposed
to
acidic
episodes
in
streams.
Can.
J.
Fish.
Aquat.
Sci.
51(
7):
1620­
1628.

This
article
describes
a
study
of
two
streams
with
severe
acidic
episodes
and
two
with
less
severe
acidic
episodes.
Episodes
of
low
pH
and
high
Al
concentration
were
associated
with
net
downstream
movement
and
increased
mortality
of
radio­
tagged
brook
trout.
Study
populations
moved
downstream
hundreds
of
meters
in
the
streams
with
more
severe
acidic
episodes
(
pH
<
5.0
and
Al
>
200

g/
L.
Median
downstream
movement
in
spring
was
250
and
900
m
after
20
days
for
fish
in
the
more
acidic
streams;
one­
third
of
the
fish
were
found
dead
during
this
time.
They
found
no
net
movement
and
no
dead
fish
in
the
reference
streams.
Lower
stream
discharge
in
fall
studies
was
associated
with
less
severe
acidic
episodes,
less
net
movement,
and
no
mortality.
Water
samples
collected
at
individual
fish
locations
indicated
that
few
fish
avoided
adverse
effects
of
acidic
episodes
by
remaining
in
microhabitats
with
higher
H
and
lower
Al
concentration.

Griffith,
M.
B.,
and
S.
A.
Perry.
1993.
Colonization
and
processing
of
leaf
litter
by
macroinvertebrate
shredders
in
streams
of
contrasting
pH.
Freshwat.
Biol.
30(
1):
93­
103.

This
article
describes
a
leaf
pack
study
of
four
central
Appalachian
streams.
Leaf
litter
processing
rates
were
fastest
in
the
neutral
streams,
slowest
in
the
acidic
stream,
and
intermediate
in
the
most
alkaline
stream.
Slower
processing
rates
in
the
acidic
stream
were
associated
with
lower
total
shredder
biomass,
made
up
predominantly
by
small
leuctrid
and
nemourid
stoneflies.
The
differences
in
processing
rates
between
the
more
alkaline
stream
and
the
neutral
streams
appeared
to
be
related
to
taxonomic
differences
in
the
shredder
assemblages.
Insects
were
dominant
in
the
neutral
streams,
and
amphipods
were
dominant
in
the
more
alkaline
stream.

Heath,
R.
H.,
J.
S.
Kahl,
S.
A.
Norton,
and
I.
J.
Fernandez.
1992.
Episodic
acidification
caused
by
the
seasalt
effect
in
coastal
Maine
streams,
USA.
Water
Resour.
Res.
28:
1081­
1088.

This
field
study
examined
the
causes
and
characteristics
of
episodic
acidification
of
low­
order
streams
in
Acadia
National
Park
(
Maine),
based
on
samples
collected
during
stormflow
conditions
using
automated
water
samplers
and
analyzed
for
ANC,
pH,
DOC,
Si,
and
major
ions.
The
primary
conclusion
of
the
study
was
that
the
sea­
salt
effect
is
the
dominant
cause
of
episodic
acidification
of
streams
in
coastal
Maine,
owing
to
a
depression
in
the
Na:
Cl
ratio
in
surface
waters
by
precipitation.
No
other
contributors
to
episodic
acidification
(
SO4
2

,
NO3

,
organic
acids)
were
able
to
explain
the
magnitude
of
ANC
changes
observed.

Herlihy,
A.
T.,
P.
R.
Kaufmann,
and
M.
E.
Mitch.
1991.
Chemical
characteristics
of
streams
in
the
Eastern
United
States:
II.
Sources
of
acidity
in
acidic
and
low
ANC
streams.
Water
Resour.
Res.
27:
629­
642.

This
is
part
2
of
a
two­
part
paper
that
summarizes
the
results
of
EPA's
National
Stream
Survey,
a
probability
survey
of
stream
chemistry
in
the
mid­
Atlantic
and
southeastern
United
States.
This
part
of
the
paper
describes
the
chemical
classification
scheme
used
to
attribute
sources
of
acidity
to
organics,
mining,
and
deposition.
It
then
uses
the
76
probability
design
of
the
Survey
to
make
population
estimates
of
stream
numbers
in
the
different
acid
source
classes.

Herlihy,
A.
T.,
P.
R.
Kaufmann,
M.
R.
Church,
P.
J.
Wigington,
Jr.,
J.
R.
Webb,
and
M.
J.
Sale.
1993.
The
effects
of
acidic
deposition
on
streams
in
the
Appalachian
Mountain
and
Piedmont
region
of
the
mid­
Atlantic
United
States.
Water
Resour.
Res.
29:
2687­
2703.

This
article
is
an
assessment
of
the
effects
of
acidic
deposition
on
streams
of
the
mid­
Appalachian
and
Piedmont
region
of
the
U.
S.
It
summarizes
all
available
survey
information
on
the
effects
of
acidic
deposition
on
chronic
and
episodic
stream
chemistry,
and
includes
biological
effects.

Hooper,
R.
P.,
and
N.
Christophersen.
1992.
Predicting
episodic
stream
acidification
in
the
southeastern
United
States:
Combining
a
longterm
acidification
model
and
the
end­
member
mixing
concept.
Water
Resour.
Res.
28:
1983­
1990.

This
paper
addresses
the
problem
of
predicting
episodic
stream
acidification
through
the
use
of
a
long­
term
acidification
model
(
MAGIC)
linked
to
a
end­
member
mixing
model.
The
results
of
the
modeling
effort
C
applied
to
the
Panola
Mountain
Research
Watershed
(
Georgia)
C
predicted
that
the
two
upper
soil
levels
will
dramatically
acidify
within
the
next
50
years,
causing
the
streamwater
to
become
unsuitable
for
sensitive
aquatic
biota
for
much
of
the
year.

Hyer,
K.
E.,
J.
R.
Webb,
and
K.
N.
Eshleman.
1995.
Episodic
acidification
of
three
streams
in
Shenandoah
National
Park,
Virginia
(
U.
S.
A.).
Water
Air
Soil
Pollut.
85:
523­
528.

The
short­
term
acidification
of
three
streams
in
Shenandoah
National
Park
was
studied
to
quantify
the
magnitude
of
chemical
changes
accompanying
stormflow
conditions
and
to
evaluate
the
contributions
of
individual
ions
to
changes
in
streamwater
ANC.
An
important
element
of
the
study
was
the
unique
bedrock
geology
of
each
of
the
catchments.
Data
from
25
storm
events
were
analyzed
using
the
method
of
Molot
et
al.
(
1989).
Depressions
in
pH
and
ANC
were
observed
in
all
three
streams,
although
ANC
only
became
negative
in
the
most
acid­
sensitive
stream,
Paine
Run,
which
drains
a
catchment
underlain
by
silici­
clastic
bedrock.
Similar
to
other
studies,
SO4
2

and
NO3

concentrations
usually
increased
during
storm
events
in
all
streams,
although
base
cation
concentrations
typically
increased.
Minimum
values
of
ANC
during
storm
events
were
best
predicted
by
a
linear
model
using
antecedent
baseflow
ANC
as
the
independent
variable.

Johnson,
D.
W.,
and
S.
E.
Lindberg,
eds.
1992.
Atmospheric
Deposition
and
Forest
Nutrient
Cycling.
Ecological
Studies.
Springer­
Verlag,
New
York.

This
book
summarizes
the
results
of
the
Integrated
Forest
Study
(
IFS),
and
includes
detailed
information
on
the
IFS
sites
in
Great
Smoky
Mountains
National
Park
(
3
sites)
and
at
Coweeta.
The
focus
was
on
detailed
measurements
of
deposition
(
wet,
dry,
and
cloud),
soil
solution
chemistry,
and
other
aspects
of
nutrient
cycles
in
forests.
This
work
clearly
identifies
the
high­
elevation
parts
of
Great
Smoky
Mountains
National
Park
as
the
highest
deposition
areas
in
the
United
States.

Johnson,
D.
W.,
W.
T.
Swank,
and
J.
M
Vose.
1993.
Simulated
effects
of
atmospheric
deposition
on
nutrient
cycling
in
a
mixed
deciduous
forest.
Biogeochem.
23:
169­
196.

Using
the
NuCM
(
Nutrient
Cycling
Model)
model,
the
authors
investigated
the
effects
of
three
S
deposition
scenarios
on
biogeochemical
cycling
of
N,
P,
S,
K,
Ca,
and
Mg
in
a
mixed
deciduous
forest
at
Coweeta,
North
Carolina.
Ecosystem
S
and
SO4
2

leaching
were
almost
entirely
controlled
by
SO4
2

adsorption
via
the
nature
of
the
Langmuir
adsorption
isotherm
used
in
NuCM.
Both
the
simulations
and
field
data
show
that
the
ecosystem
is
becoming
more
S
saturated.
Varying
S
deposition
had
very
little
effect
upon
simulated
vegetation
growth,
nutrient
uptake,
or
N
cycling
but
had
a
strong
effect
on
base
cation
and
P
leaching.
S
deposition
effects
on
soil
exchangeable
pools
of
these
elements,
however,
were
minimal
due
to
the
size
of
these
pools
relative
to
the
fluxes.

Kahl,
J.
S.,
S.
A.
Norton,
T.
A.
Haines,
E.
A.
Rochette,
R.
H.
Heath,
and
S.
C.
Nodvin.
1992.
Mechanisms
of
episodic
acidification
of
low­
order
streams
in
Maine,
U.
S.
A.
Environ.
Pollut.
78:
37­
44.

Observed
chemical
changes
during
hydrological
events
in
low­
order
streams
in
Maine
were
examined
in
order
to
understand
the
mechanisms
of
episodic
acidification
in
this
region.
It
was
generally
observed
that
five
processes
contribute
to
episodic
depressions
in
pH
and
ANC:
(
1)
increases
in
nitric
acid
concentrations,
(
2)
increases
in
77
organic
acidity,
(
3)
increases
in
the
anion
fraction
of
SO4
2

,
(
4)
increases
in
acidity
due
to
the
salt
effect
in
soils,
and
(
5)
hydrological
dilution
of
ANC
by
increased
stream
discharge.
The
chemical
composition
of
individual
precipitation
events
was
found
to
be
irrelevant
in
the
generation
of
acidic
episodes,
with
the
exception
of
those
primarily
attributable
to
high
loadings
of
neutral
salts
within
the
coastal
region
of
Maine.

Kaufmann,
P.
R.,
A.
T.
Herlihy,
M.
E.
Mitch,
and
W.
S.
Overton.
1991.
Chemical
characteristics
of
streams
in
the
Eastern
United
States:
I.
Synoptic
survey
design,
acid­
base
status
and
regional
chemical
patterns.
Water
Resour.
Res.
27:
611­
627.

This
is
part
1
of
a
two­
part
paper
that
summarizes
the
results
of
EPA's
National
Stream
Survey,
a
probability
survey
of
stream
chemistry
in
the
mid­
Atlantic
and
southeastern
United
States.
This
part
of
the
paper
describes
the
probability
design
of
the
Survey
and
discusses
the
regional
findings
in
terms
of
population
estimates
of
condition.

Kaufmann,
P.
R.,
A.
T.
Herlihy,
and
L.
A.
Baker.
1992.
Sources
of
acidity
in
lakes
and
streams
of
the
United
States.
Environ.
Pollut.
77:
115­
122.

This
paper
presents
and
defends
a
classification
scheme
designed
to
attribute
sources
of
acidity
to
the
lakes
and
streams
of
EPA's
National
Surface
Water
Survey.

Kobuszewski,
D.
M.,
and
S.
A.
Perry.
1993.
Aquatic
insect
community
structure
in
an
acidic
and
a
circumneutral
stream
in
the
Appalachian
Mountains
of
West
Virginia.
J.
Freshwat.
Ecol.
8(
1):
37­
45.

The
authors
report,
in
comparing
benthic
macroinvertebrates
in
two
West
Virginia
streams,
that
density
did
not
differ
between
the
acidic
and
circumneutral
stream,
but
species
richness
and
evenness
were
higher
in
the
neutral
stream.
Plecoptera
and
shredders
dominated
the
acidic
stream.

Kobuszewski,
D.
M.,
and
S.
A.
Perry.
1994.
Secondary
production
of
Rhyacophila
minora,
Ameletus
sp.,
and
Isonychia
bicolor
from
streams
of
low
and
circumneutral
pH
in
the
Appalachian
Mountains
of
West
Virginia.
Hydrobiologia
273(
3):
163­
169.

This
article
describes
a
study
of
three
acidic
streams
and
one
circumneutral
stream
in
Randolph
County,
West
Virginia.
Differences
in
secondary
production
of
these
species
were
associated
with
differences
in
macroinvertebrate
community
structure.

Kutka,
F.
J.
1994.
Low
pH
effects
on
swimming
activity
of
Ambystoma
salamander
larvae.
Environ.
Toxicol.
Chem.
13(
11):
1821­
1824.

Results
suggest
that
salamander
embryos
and
larvae
may
suffer
from
pH
levels
below
5.0,
even
though
these
levels
are
not
directly
lethal.
Because
of
their
sensitivity,
the
authors
recommend
activity
tests
with
amphibian
larvae
for
use
in
risk
assessments.

Larsen,
G.
L.,
S.
E.
Moore,
and
B.
Carter.
1995.
Ebb
and
flow
of
encroachment
by
nonnative
rainbow
trout
in
a
small
stream
in
the
Southern
Appalachian
Mountains.
Trans.
Am.
Fish.
Soc.
124:
613­
622.

The
results
of
this
study
suggest
that
encroachment
by
rainbow
trout
(
nonnative)
can
exhibit
considerable
ebb
and
flow
in
steep
gradient
streams
of
the
Great
Smoky
Mountains
National
Park.
Results
also
suggest
substantial
evolutionary
adaptation
by
brook
trout
to
the
hydrological
conditions
in
these
small
streams.
We
found
this
paper
to
be
interesting,
because
similar
resistance
to
rainbow
trout
invasion
in
headwater
streams
has
been
attributed
to
the
greater
tolerance
of
brook
trout
to
the
typically
more
acidic
conditions
in
steep
headwaters.

Lawrence,
G.
B.,
M.
B.
David,
and
W.
C.
Shortle.
1995.
A
new
mechanism
for
calcium
loss
in
forest­
floor
soils.
Nature
378:
162­
164.

Concentrations
of
root­
available
calcium
have
declined
in
the
northeastern
United
States
over
the
last
60
years.
Based
on
data
collected
in
red
spruce
forests,
the
authors
propose
a
mechanism
where
aluminum
(
mobilized
in
the
mineral
soil
by
acid
deposition)
is
transported
into
the
forest
floor
in
a
reactive
form
that
reduces
storage
of
calcium
and
thus
its
availability
for
root
uptake.
This
would
result
in
potential
stress
to
trees,
increased
forest
demand
for
calcium,
and
decreased
calcium
runoff
to
surface
waters
(
decreasing
surface
water
ANC).

Loer,
S.
C.,
and
J.
L.
West.
1992.
Microhabitat
selection
by
brook
and
rainbow
trout
in
a
southern
Appalachian
stream.
Trans.
Am.
Fish.
Soc.
121(
6):
729­
736.

We
were
interested
in
this
paper
because
it
deals
with
the
decreased
range
of
native
brook
trout
in
the
Great
Smoky
Mountains
National
Park,
which
78
has
been
attributed
to
competition
with
introduced
rainbow
trout
C
a
fish
not
as
tolerant
of
acidity
as
brook
trout.
This
study
reports
that
young­
of­
theyear
and
to
some
extent
age
1,
brook
trout
shifted
to
habitat
positions
farther
from
overhead
cover
when
rainbow
trout
numbers
were
reduced.

MacAvoy,
S.
E.
and
A.
J.
Bulger.
1995.
Survival
of
brook
trout
(
Salvelinus
fontinalis)
embryos
and
fry
in
streams
of
different
acid
sensitivity
in
Shenandoah
National
Park,
USA.
Water
Air
Soil
Pollut.
85:
439­
444.

This
paper
demonstrates
significantly
higher
mortality
of
hatchery
stock
brook
trout
embryos
and
fry
in
field
bioassays
in
low
ANC
streams
compared
to
high
ANC
streams
in
Virginia,
both
chronically
and
during
acidic
episodes.

McQuattie,
C.
J.,
S.
L.
Stephenson,
and
P.
J.
Edwards.
1993.
Effect
of
stream
acidity
on
decomposition
of
sugar
maple
(
Acer
saccharum)
and
red
oak
(
Quercus
rubra)
leaves.
Ohio
J.
Sci.
93(
2):
48.

Leaf
decomposition
of
both
species
was
more
rapid
in
a
pH
5.6
stream
than
in
a
pH
3.2
stream.
Authors
state
that
differences
were
probably
due
to
increased
numbers
or
types
of
aquatic
microorganisms
found
in
the
pH
5.6
stream.
Acidity
level
appeared
to
have
a
direct
effect
on
cuticular
wax
structure
of
the
leaves.

Miller­
Marshall,
L.
M.
1993.
Mechanisms
Controlling
Variation
in
Stream
Chemical
Composition
During
Hydrologic
Episodes
in
the
Shenandoah
National
Park,
Virginia.
MS
thesis.
Department
of
Environmental
Sciences,
University
of
Virginia,
Charlottesville.
165
pp.

This
thesis
describes
both
a
field
investigation
of
episodic
acidification
at
two
sites
in
Shenandoah
National
Park
and
a
statistical
analysis
of
longterm
weekly
stream
chemical
composition
data
for
four
sites
in
the
Park.
A
major
result
of
the
study
was
that
sulfuric
acid
increases
and
base
cation
dilution
are
the
primary
mechanisms
contributing
to
the
losses
of
ANC
in
streams
in
the
Park
during
hydrological
events,
although
there
is
evidence
that
nitric
acid
has
recently
become
a
more
important
contributor
to
ANC
losses
in
these
streams.
These
conclusions
were
based
on
application
of
an
approach
to
partitioning
ANC
losses
developed
by
Molot
et
al.
(
1989)
and
on
a
newer
approach
known
as
the
Response
Sector
Model
(
RSM).

Morgan,
R.
P.,
II,
C.
K.
Murray,
and
K.
N.
Eshleman.
1994.
Episodic
water
chemistry
changes
in
a
western
Maryland
watershed.
Report
CBRM­
AD­
94­
8.
Maryland
Department
of
Natural
Resources,
Annapolis.
125
pp.

This
report
describes
a
study
of
episodic
acidification
at
six
sites
in
the
Big
Run
watershed
in
western
Maryland.
A
total
of
23
hydrological
events
(
including
storm
fronts,
thunderstorms,
and
snowmelts)
were
sampled
over
a
3­
year
period
(
1989­
92).
Antecedent
ANC
was
found
to
be
an
excellent
predictor
(
r2
=
0.91)
of
ANCminimum
at
the
six
sites
and
ANC
depressions
were
largely
attributable
to
increases
in
sulfuric
acid
concentrations
and
dilution
of
base
cations.
Nitric
and
organic
acids
were
secondary
contributors
to
the
loss
of
ANC.

Murdoch,
P.
S.,
and
J.
L.
Stoddard.
1992.
The
role
of
nitrate
in
the
acidification
of
streams
in
the
Catskill
Mountains
of
New
York.
Water
Resour.
Res.
28:
2707­
2720.

The
authors
of
this
paper
presented
a
statistical
analysis
of
long­
term
(
23
B
70­
year
records)
chemical
data
from
19
streams
and
rivers
in
the
Catskill
mountains
of
New
York
revealing
that
SO4
2

concentrations
are
generally
declining,
while
NO3

concentrations
are
increasing.
The
NO3

trend
appears
to
be
attributable
largely
to
higher
concentrations
of
NO3

during
peak
flow
conditions
(
particularly
spring
snowmelt).
The
magnitude
of
episodic
nitrate
increases
has
increased
since
about
1970.

Neal,
C.,
A.
Robson,
B.
Reynolds,
and
A.
Jenkins.
1992.
Prediction
of
future
short­
term
stream
chemistry
C
a
modeling
approach.
J.
Hydrol.
130:
87­
103.

This
paper
addresses
the
problem
of
predicting
future
short­
term
stream
chemistry
in
acidic
and
acid­
sensitive
streams
under
various
deposition
scenarios.
The
approach
used
a
hydrograph
separation
based
on
ANC,
coupled
to
the
twocomponent
version
of
MAGIC
and
ALCHEMI.
The
technique
was
applied
to
the
Afon
Gwyn
catchment
in
mid­
Wales,
with
results
demonstrated
in
the
forms
of
(
1)
3­
month
sequences
of
hydrogen
ion
and
inorganic
aluminum
concentrations
and
(
2)
chemical
duration
curves
for
hydrogen
ion
and
aluminum.

Newman,
K.,
and
A.
Dolloff.
1995.
Responses
of
blacknose
dace
(
Rhinichthys
atratulus)
and
brook
char
(
Salvelinus
fontinalis)
to
acidified
water
in
a
laboratory
stream.
Water
Air
Soil
Pollut.
85:
371­
376.
79
This
study
was
based
on
Shenandoah
National
Park
fish.
Both
species
actively
avoided
an
acid
pulse
(
ambient
pH
7.1
shifted
to
5.1)
by
moving
to
a
pH­
neutral
refuge,
demonstrating
at
least
the
potential
of
two
regionally
important
species
to
recognize
waters
with
either
lethal
or
sublethal
acid­
base
chemistry
and
seek
better
water
quality.

Nikolaidis,
N.
P.,
P.
K.
Muller,
J.
L.
Schnoor,
and
H.
L.
Hu.
1991.
Modeling
the
hydrogeochemical
response
of
a
stream
to
acid
deposition
using
the
enhanced
trickle­
down
model.
Res.
J.
WPCF.
63(
3):
220­
227.

The
enhanced
trickle
down
model
was
applied
to
White
Oak
Run,
a
second­
order
stream
in
Shenandoah
National
Park
in
Virginia.
Model
results
demonstrated
the
"
delayed
response"
of
this
system,
projecting
that
SO4
desorbed
from
the
soils
after
40
B
50
years,
assuming
no
reductions
in
the
current
deposition.
This
finding
has
significant
policy
implications,
because
reduction
in
deposition
levels
will
not
result
in
a
quick
recovery
of
this
system.

Nodvin,
S.
C.,
H.
Van
Miegroet,
S.
E.
Lindberg,
N.
S.
Nicholas,
and
D.
W.
Johnson.
1995.
Acidic
deposition
ecosystem
processes,
and
N
saturation
in
a
high­
elevation
southern
Appalachian
watershed.
Water
Air
Soil
Pollut.
85:
1647­
1652.

This
paper
describes
a
field
study
of
biogeochemical
processes
in
the
Noland
Divide
watershed
in
the
Great
Smoky
Mountains
National
Park
(
GSMNP).
The
watershed
is
dominated
by
a
high­
elevation
spruce­
fir
forest
that
receives
the
highest
loadings
of
atmospheric
S
deposition
in
North
America.
Watershed
NO3

export
was
extremely
high,
exacerbating
the
loss
of
base
cations
from
the
system.
Stream
ANC
was
extremely
low
(

10
B
20

eq/
L)
and
depressions
in
pH
of
a
full
unit
for
as
long
as
several
weeks
were
associated
with
stormflow
conditions.
Nitrate
concentrations
actually
exceeded
SO4
2

concentrations
during
episodes,
which
was
attributed
to
substantial
SO4
2

retention
by
sorption
and
N
saturation.
The
authors
conclude
that
episodic
and
chronic
acidification
are
moderated
by
sulfate
sorption,
but
exacerbated
by
N
(
nitric
acid)
release
from
the
watershed
soils.

Norton,
S.
A.,
R.
F.
Wright,
J.
S.
Kahl,
and
J.
P.
Scofield.
1992.
The
MAGIC
simulation
of
surface
water
acidification
at,
and
first
year
results
from,
the
Bear
Brook
watershed
manipulation,
Maine,
USA.
Environ.
Pollut.
77:
279­
286.
A
manipulation
experiment
involving
the
addition
of
ammonium
sulfate
to
one
of
a
pair
of
monitored
catchments
(
East
and
West
Bear
Brooks)
in
eastern
Maine
was
performed;
this
paper
describes
results
from
the
first
year
of
the
manipulation
experiment
and
the
chemical
patterns
evident
in
the
3.5
years
of
antecedent
monitoring.
The
most
important
aspect
of
the
study
from
the
perspective
of
assessing
episodic
acidification
was
that
stormflow
conditions
in
the
treated
catchment
were
associated
with
lower
ANC
and
pH
and
higher
Al
concentrations
than
those
observed
prior
to
the
manipulation
experiment.
These
results
provide
additional
evidence
that
the
episodic
hydrochemical
responses
of
catchments
are
at
least
in
part
a
function
of
the
atmospheric
deposition
loading.

O'Brien,
A.
K.,
and
K.
N.
Eshleman.
1996.
Episodic
acidification
of
a
coastal
plain
stream
in
Virginia.
Water
Air
Soil
Pollut.
89:
291­
316.

This
study
examined
the
episodic
acidification
of
an
acid­
sensitive
coastal
plain
stream
in
Virginia.
All
storms
sampled
showed
increases
in
SO4
2

,
with
highest
concentrations
near
peak
discharge;
small
increases
in
base
cations
and
DOC
also
were
found
for
most
storms.
Increases
in
sulfuric
acid
concentrations
were
the
primary
causes
of
ANC
depressions,
although
organic
acids
also
contributed
to
ANC
loss
during
winter/
spring
rainstorms.
The
chemical
results
also
lend
support
to
the
hypothesis
that
saturation
overland
flow
is
the
dominant
mechanism
for
transport
of
solutes
to
the
stream
during
stormflow
conditions.

O'Brien,
A.
K.,
K.
C.
Rice,
M.
M.
Kennedy,
and
O.
P.
Bricker.
1993.
Comparison
of
episodic
acidification
of
mid­
Atlantic
upland
and
coastal
plain
streams.
Water
Resour.
Res.
29:
3029­
3039.

Field
studies
of
episodic
acidification
were
conducted
in
five
mid­
Atlantic
watersheds
in
three
physiographic
provinces:
Coastal
Plain,
Valley
and
Ridge,
and
Blue
Ridge.
ANC
depressions
were
largest
in
watersheds
underlain
by
reactive
bedrock,
compared
to
those
underlain
by
quartzites
or
unconsolidated
quartz
sands
and
cobbles.
Results
from
the
study
clearly
demonstrated
that
sulfuric
acid
can
contribute
to
episodic
ANC
depressions
in
the
region,
compared
to
the
northeastern
United
States
where
nitric
acid
is
the
more
dominant
contributor.

Pinder,
M.
J.,
and
R.
P.
Morgan.
1994.
Interactions
of
pH
and
habitat
on
cyprinid
distributions
in
80
Appalachian
streams
of
Maryland.
Trans.
Am.
Fish.
Soc.
124(
1):
94­
102.

The
authors
related
water
chemistry,
physical
habitat,
and
watershed
characteristics
to
cyprinid
distributions
in
56
Appalachian
streams
in
Maryland.
Eleven
of
these
streams
had
pH

5.3,
ANC
of

50

eq/
L,
and
lacked
cyprinids.
Gradient
was
the
primary
factor
affecting
cyprinid
presence
in
streams
that
had
pH
6.49
or
higher.
This
indicates
that
some
streams
with
pH
5.3
or
less
would
have
lacked
cyprinids
in
the
absence
of
acidification,
but
that
cyprinid
distributions
are
affected
by
factors
related
to
stream
acidification.

Qualls,
R.
G.,
and
B.
L.
Haines.
1992.
Biodegradability
of
dissolved
organic
matter
in
forest
throughfall,
soil
solution,
and
stream
water.
Soil
Sci.
Soc.
Am.
J.
56:
578­
586.

More
than
95%
of
DOC
and
DON
leached
by
throughfall
was
removed
by
soils
under
oak
hickory
forest
in
the
southern
Appalachians.
Soil
adsorption
rather
than
decomposition
seems
responsible
for
most
DON
and
DOC
removal.

Robson,
A.,
K.
Beven,
and
C.
Neal.
1992.
Towards
identifying
sources
of
subsurface
flow:
A
comparison
of
components
identified
by
a
physically
based
runoff
model
and
those
determined
by
chemical
mixing
techniques.
Hydrol.
Proc.
6:
199­
214.

This
paper
addresses
the
problem
of
dynamic
modeling
of
ANC
in
streams
using
a
modified
version
of
TOPMODEL,
in
which
fixed
concentrations
are
assumed
for
the
two
endmembers
deep
groundwater
flow
and
saturation
overland
flow
components.
Results
using
the
modified
version
of
TOPMODEL
were
then
compared
with
a
2­
month
synthetic
record
of
ANC
(
from
pH)
for
a
small,
spruce­
forested
catchment
in
Wales.
The
comparison
is
encouraging,
except
during
periods
when
the
saturation
overland
flow
component
is
a
combination
of
rainwater
and
subsurface
water
(
and
thus
the
exact
chemical
composition
of
this
mixture
is
not
well
defined).
In
addition,
the
results
indicated
that
the
assumption
of
constant
composition
end­
members
is
not
valid
over
long
time
periods.

Rosemond,
A.
D.,
Reice,
S.
R.,
Elwood,
J.
W.,
and
Mulholland,
P.
J.
1992.
The
effects
of
stream
acidity
on
benthic
invertebrate
communities
in
the
southeastern
United
States.
Fresh.
Biol.
27(
2):
193­
209.
This
paper
reports
on
strong
relationships
between
measures
of
the
benthic
invertebrate
community
and
water
chemistry
in
Great
Smoky
Mountains
National
Park.
Baseflow
pH
values
were
4.5
B
6.8,
and
inorganic
monomeric
aluminum
was
3
B
197

g/
L.
Total
invertebrate
density
(
excluding
the
acid­
tolerant
chironomids)
and
species
richness
were
higher
in
the
high
pH
streams;
these
effects
were
attributed
to
direct
effects
on
the
invertebrates
rather
than
on
food
availability.

Rosseland,
B.
O.,
and
M.
Staurnes.
1994.
Physiological
mechanisms
of
toxic
effects
to
acidic
water
and
ecophysiological
and
ecotoxicological
approach.
In:
C.
E.
W.
Steinberg
and
R.
F.
Wright,
eds.
Acidification
of
Freshwater
Ecosystems:
Implications
for
the
Future.
John
Wiley
and
Sons,
Ltd.

This
paper
is
a
recent
review
of
toxic
mechanisms
and
fish
sensitivities
in
acidic
water.
More
up­
to­
date
information
is
provided
than
was
available
in
1990­
91,
but
the
essential
roles
of
aluminum
in
ion­
regulation
failure
at
the
cellular/
organism
level,
and
the
loss
of
early
life
stages
at
the
population
level,
have
not
changed.
New
relevant
areas
of
emphasis
are
identified:
(
1)
effects
on
sensory
organs,
(
2)
avoidance
responses,
(
3)
sublethal
effects
on
growth,
and
(
4)
the
greater
vulnerability
of
fish
in
streams
versus
lakes.

Rosseland,
B.
O.,
I.
A.
Blakar,
A.
J.
Bulger,
F.
Kroglund,
A.
Kvellestad,
E.
Lydersen,
D.
H.
Oughton,
B.
Salbu,
M.
Staurnes,
and
R.
Vogt.
1992.
The
mixing
zone
between
limed
and
acidic
river
waters:
Complex
aluminum
chemistry
and
extreme
toxicity
for
salmonids.
Environ.
Pollut.
78(
1):
3­
8.

This
paper
contains
an
explanation
of
one
of
the
important
consequences
of
liming,
which
is
being
increasingly
considered
and
used
in
the
SAMI
region.
The
mixing
zone
downstream
of
the
confluence
between
limed
and
unlimed
acidic
waters
has
unstable
aluminum
chemistry,
and
the
water
can
be
even
more
toxic
than
the
unlimed
acidic
tributary,
even
though
the
pH
and
calcium
are
higher
and
the
inorganic
monomeric
aluminum
is
lower.
This
phenomenon
might
occur
at
any
confluence
where
the
mixing
waters
have
very
different
pH
values;
it
can
create
a
very
toxic
zone
unpassable
by
fish.

Schaefer,
D.
A.,
and
C.
T.
Driscoll.
1993.
Identifying
sources
of
snowmelt
acidification
with
a
watershed
mixing
model.
Water
Air
Soil
Pollut.
67:
345­
365.
81
Two­
component
separations
of
discharge
into
old
(
soil
and
ground
water)
and
new
(
snowmelt
water)
were
performed
using
ionic
tracers
for
11
Adirondack
watersheds;
old
water
contributions
ranged
from
66%
to
90%
and
no
relationship
between
old
water
%
and
soil
depth
(
to
till)
was
found.
It
was
therefore
concluded
that
the
magnitude
of
episodic
acidification
in
Adirondack
watersheds
is
primarily
governed
by
watershed
chemical
interactions,
not
by
the
varying
proportions
of
"
old"
water
during
snowmelt.

Schaefer,
D.
A.,
C.
T.
Driscoll,
Jr.,
R.
Van
Dreason,
and
C.
P.
Yatsko.
1992.
Reply.
Water
Resour.
Res.
28:
2874­
2876.

This
paper
was
a
reply
to
the
technical
comment
of
Eshleman
(
1992),
which
again
argued
that
the
linear
relationship
observed
by
Eshleman
(
1988)
between
ANCindex
and
ANCminimum
for
a
group
of
lakes
in
the
Adirondacks
of
New
York
is
a
statistical
artifact.

Shanley,
J.
B.,
and
N.
E.
Peters.
1993.
Variations
in
aqueous
sulfate
concentrations
at
Panola
Mountain,
Georgia.
J.
Hydrol.
Amst.
146:
361­
382.

Aqueous
SO4

concentrations
were
measured
in
incident
precipitation,
canopy
throughfall,
stemflow
soil
water,
groundwater,
and
streamwater
at
three
locations
in
a
41­
ha
forested
watershed
at
Panola
Mountain
State
Park
in
the
Georgia
Piedmont.
Canopy
throughfall,
stemflow,
and
runoff
from
a
bedrock
outcrop
in
the
headwaters
were
enriched
in
SO4

relative
to
incident
precipitation
due
to
washoff
of
dry
deposition
that
accumulated
between
storms.
Streamwater
SO4

concentrations
during
base
flow
were
controlled
by
low
SO4

groundwater
discharge.
As
flow
increased,
an
increasing
proportion
of
shallow,
high­
sulfate
groundwater
and
soil
water
contributed
to
streamflow.
The
dominant
control
on
stream
SO4

concentration
shifted
from
SO4

retention
by
adsorption
in
the
mineral
soil
at
baseflow
to
mobilization
of
SO4
2

from
the
upper,
organic­
rich
horizons
of
the
soil
at
high
flow.

Shubzda,
J.,
S.
E.
Lindberg,
C.
T.
Garten,
and
S.
C.
Nodvin.
1995.
Elevational
trends
in
the
fluxes
of
sulphur
and
nitrogen
in
throughfall
in
the
Southern
Appalachian
mountains:
Some
surprising
results.
Water
Air
Soil
Pollut.
85:
2265­
2270.

This
study
evaluated
deposition
and
throughfall
chemistry
along
an
elevational
gradient
in
Great
Smoky
Mountains
National
Park.
Deposition
rates
increased
with
elevation,
driven
largely
by
a
large
input
from
cloudwater
above
1700
m.
Nitrogen
concentrations
in
throughfall
at
the
highest
elevations
were
30%
lower
than
at
lower
elevation
sites.
The
authors
hypothesize
that:
(
1)
lower
rates
of
dry
deposition
among
declining
forests
at
high
elevations
and/
or
(
2)
higher
rates
of
canopy
uptake
of
N
at
high
elevations
are
responsible
for
the
lower
N
in
throughfall
at
high
elevations.

Stoddard,
J.
L.
1994.
Long­
term
changes
in
watershed
retention
of
nitrogen:
Its
causes
and
aquatic
consequences.
Pages
223­
284
in
L.
A.
Baker,
ed.
Environmental
Chemistry
of
Lakes
and
Reservoirs.
Advances
in
Chemistry
Series,
No.
237.
American
Chemical
Society,
Washington,
DC.

This
paper
does
not
deal
specifically
with
the
SAMI
region,
but
it
does
present
some
results
for
Great
Smoky
Mountains
National
Park,
Fernow,
Coweeta
and
other
sites.
The
paper
proposes
a
taxonomic
scheme
for
classifying
watersheds
according
to
their
stage
(
Stage
0
through
Stage
3)
of
nitrogen
saturation.
The
control
watersheds
at
Fernow
are
among
the
few
United
States
sites
at
Stage
2,
while
the
Noland
Divide
watershed
in
Great
Smoky
Mountains
National
Park
is
the
only
Stage
3
site
identified.

Tranter,
M.,
T.
D.
Davies,
P.
J.
Wigington,
Jr.,
and
K.
N.
Eshleman.
1994.
Episodic
acidification
of
freshwater
systems
in
Canada
C
physical
and
geochemical
processes.
Water
Air
Soil
Pollut.
72:
19­
39.

This
paper
largely
summarizes
Section
3.3
of
Wigington
et
al.
(
1990),
NAPAP
SOS/
T
Report
12.

Van
Sickle,
J.,
J.
P.
Baker,
H.
A.
Simonin,
B.
P.
Baldigo,
W.
A.
Kretser,
and
W.
E.
Sharpe.
1996.
Episodic
acidification
of
small
streams
in
the
northeast
United
States:
Effects
on
fish
mortality
during
field
bioassays.
Ecol.
Appl.
6:
408­
421.

This
paper
largely
summarizes
the
results
from
the
field
bioassays
in
the
13
ERP
streams
first
reported
on
by
Wigington
et
al.
(
1993).

Wang,
D.,
Y.
Xia,
D.
Zhuang,
B,
Liu,
X.
Li,
Q.
Kuang,
and
S.
Wang.
1992.
A
study
of
effects
of
water
acidification
on
aquatic
organisms
of
different
trophic
levels.
Acta
Sci.
Circumstant.
12(
1):
91­
96.
82
This
study
assessed
the
effects
of
low
pH
on
aquatic
organisms
at
various
trophic
levels,
including
zooplankton,
mollusks,
aquatic
insects,
algae,
and
heterotrophic
bacteria.
The
threshold
of
adverse
effects
appeared
to
be
pH
5.0
B
5.5
for
most
of
these
organisms.

Webb,
J.
R.,
F.
A.
Deviney,
J.
N.
Galloway,
C.
A.
Rinehart,
P.
A.
Thompson,
and
S.
Wilson.
1994.
The
Acid­
base
Status
of
Native
Brook
Trout
Streams
in
the
Mountains
of
Virginia:
A
Regional
Assessment
Based
on
the
Virginia
Trout
Stream
Sensitivity
Study.
Report
submitted
to
Virginia
Department
of
Game
and
Inland
Fisheries,
Charlottesville.

This
report
provides
an
assessment
of
the
acidbase
status
of
streams
in
Virginia
that
support
reproducing
populations
of
native
brook
trout
(
Salvelinus
fontinalis)
based
on
streamwater
composition
data
collected
through
June
1993.
The
report
describes
(
1)
a
stream
classification
system
based
on
present
acid­
base
status
and
bedrock
geology,
(
2)
trend
analysis
of
changes
in
streamwater
composition
during
the
period
of
study
(
1987­
1993),
and
(
3)
development
and
application
of
several
models
for
assessing
current
impacts
and
predicting
future
impacts
given
a
range
of
sulfur
emission
reduction
levels.
An
important
aspect
of
the
study
was
the
calibration
of
the
linear
regression
model
relating
ANCminimum
to
ANCindex
for
the
population
of
streams;
this
technique
provided
a
worst­
case
analysis
of
acidification
impacts
under
baseflow
conditions.

Webb,
J.
R.,
B.
J.
Cosby,
K.
N.
Eshleman,
and
J.
N.
Galloway.
1995.
Change
in
the
acid­
base
status
of
Appalachian
Mountain
catchments
following
forest
defoliation
by
the
gypsy
moth.
Water
Air
Soil
Pollut.
85:
535­
540.

Long­
term
monitoring
of
the
chemical
composition
of
White
Oak
Run
in
Shenandoah
National
Park
(
Virginia)
indicates
that
changes
in
acid­
base
status
are
associated
with
forest
defoliation
by
the
gypsy
moth
caterpillar.
Increasing
concentrations
of
nitrate,
chloride,
base
cations,
and
hydrogen
ion,
as
well
as
decreasing
concentrations
of
sulfate
and
ANC
were
observed.
In
addition,
record
high
levels
of
hydrogen
ion
and
record
low
levels
of
ANC
indicate
that
the
magnitude
of
episodic
acidification
has
increased
in
this
acid­
sensitive
stream.

Wigington,
P.
J.,
Jr.,
T.
D.
Davies,
M.
Tranter,
and
K.
N.
Eshleman.
1992.
Comparison
of
episodic
acidification
in
Canada,
Europe
and
the
United
States.
Environ.
Pollut.
78:
29­
35.

This
paper
largely
summarizes
Section
3.4
of
Wigington
et
al.
(
1990),
NAPAP
SOS/
T
Report
12.

Wigington,
P.
J.,
Jr.,
et
al.
1993.
Episodic
Acidification
of
Streams
in
the
Northeastern
United
States:
Chemical
and
Biological
Results
of
the
Episodic
Response
Project.
EPA/
600/
R­
93/
190.
U.
S.
Environmental
Protection
Agency,
Washington,
D.
C.

This
report
of
a
major
field
project
conducted
by
cooperators
for
the
U.
S.
EPA
addressed
uncertainties
about
the
occurrence,
causes,
and
biological
effects
associated
with
episodic
acidification
of
streams
in
the
northeastern
United
States.
The
project
consisted
of
intensive
studies
of
chemical
and
biological
effects
in
13
streams
draining
forested
watersheds
in
three
study
regions:
the
northern
Appalachian
plateau
in
Pennsylvania,
the
Catskill
Mountains
in
New
York,
and
the
Adirondack
Mountains
in
New
York.
Automated
sampling
and
in­
situ
monitoring
equipment
was
used
to
determine
the
chemical
composition
of
the
13
streams
from
fall
1988
through
spring
1990.
Biological
studies
focused
on
brook
trout
and
native
forage
species
and
utilized
in­
situ
bioassays,
radio
transmitter
studies
of
fish
movement
and
fish
population
monitoring.

Results
from
the
chemical
studies
clearly
indicated
the
occurrence
of
acidic
episodes
with
pH
<
5
and
inorganic
monomeric
Al
concentrations
>
150

g/
L
in
at
least
two
streams
in
each
region.
ANC
depressions
were
shown
to
result
from
a
complex
interaction
of
multiple
ions,
but
base
cation
dilution
was
important
in
all
regions.
Organic
acid
pulses
were
significant
in
Adirondack
streams,
while
nitric
acid
pulses
dominated
the
response
of
the
Catskill
and
Adirondack
streams.
Only
the
Pennsylvania
streams
were
regularly
affected
by
episodic
pulses
of
sulfuric
acid.

Results
from
the
in­
situ
field
bioassays
indicated
that
mortality
was
significantly
higher
during
acidic
episodes
than
during
nonacidic
conditions.
In
addition,
a
multiple
logistic
regression
model
was
used
to
relate
bioassay
mortality
to
summary
statistics
of
time­
varying
stream
chemical
composition
In
general,
the
modeling
results
indicated
that
mortality
could
be
adequately
predicted
using
a
single
index
of
inorganic
monomeric
Al
concentration
during
stormflow
periods;
however,
a
higher
percentage
of
variation
83
in
mortality
was
explained
when
pH
and
Ca
variables
were
added
to
the
model.

Results
from
the
fish
population
studies
clearly
indicated
net
downstream
movement
of
fish
during
events
and
movement
of
trout
into
alkaline
refugia
C
processes
that
could
potentially
mitigate
the
toxic
effects
associated
with
individual
acid
episodes.
However,
the
study
results
showed
that
ERP
streams
with
suitable
chemical
conditions
during
low
flow,
but
moderate
to
severe
conditions
during
high
flow
had
higher
mortality
and
lower
brook
trout
density
and
biomass
compared
to
the
nonacidic
"
control"
streams.
In
general,
trout
abundance
was
reduced
and
acid­
sensitive
species
were
absent
from
ERP
streams
with
median
pH
<
5.0
B
5.2
and
inorganic
Al
concentrations
>
100
B
200

g/
L
during
high
flows.
Therefore,
the
authors
concluded
that
episodic
acidification
can
have
long­
term
effects
on
fish
populations
and
communities
in
small,
acid­
sensitive
streams.

Wigington,
P.
J.,
Jr.,
J.
P.
Baker,
D.
R.
DeWalle,
W.
A.
Kretser,
P.
S.
Murdoch,
H.
A.
Simonin,
J.
Van
Sickle,
M.
K.
McDowell,
D.
V.
Peck,
and
W.
R.
Barchet.
1996.
Episodic
acidification
of
small
streams
in
the
northeast
United
States:
Episodic
Response
Project.
Ecol.
Appl.
6:
374­
388.

This
paper
largely
summarizes
the
study
areas
and
objectives
of
the
Episodic
Response
Project,
which
were
first
reported
on
by
Wigington
et
al.
(
1993).
Wigington,
P.
J.,
Jr.,
D.
R.
DeWalle,
P.
S.
Murdoch,
W.
A.
Kretser,
and
H.
A.
Simonin.
1996.
Episodic
acidification
of
small
streams
in
the
northeast
United
States:
Ionic
controls
of
episodes.
Ecol.
Appl.
6:
389­
407.

This
paper
largely
summarizes
the
results
from
the
study
of
chemical
changes
in
the
13
ERP
streams
first
reported
on
by
Wigington
et
al.
(
1993).
84
Appendix
A
Stream
Networks
in
the
SAMI
Class
I
Wilderness
Areas
As
part
of
the
analyses
done
for
this
report
on
the
SAMI
Class
I
wilderness
areas,
we
digitized
the
stream
networks
for
the
eight
smaller
Class
I
areas.
We've
included
these
maps
in
this
appendix
in
case
they
may
be
useful
for
the
SAMI
assessment
or
other
future
SAMI
efforts.
The
stream
networks
were
digitized
from
the
finest
map
scale
available
either
a
1:
24,000­
scale
topographic
map
or
a
Forest
Service
Wilderness
area
map.
The
presence/
absence
of
a
given
stream
on
the
1:
100,000­
scale
topographic
map
is
also
given
on
the
maps
in
this
appendix.
Thus,
a
choice
in
map­
scale
networks
is
available.
85
Figure
A­
1.
Stream
network
in
Dolly
Sods
wilderness
area.
86
Figure
A­
2.
Stream
network
in
Otter
Creek
wilderness
area.
87
Figure
A­
3.
Stream
network
in
James
River
Face
wilderness
area.
88
Figure
A­
4.
Stream
network
in
Joyce
Kilmer/
Slickrock
wilderness
area.
89
Figure
A­
5.
Stream
network
in
Linville
Gorge
wilderness
area.
90
Figure
A­
6.
Stream
network
in
Shining
Rock
wilderness
area.
91
Figure
A­
7.
Stream
network
in
Cohutta
wilderness
area.
92
Figure
A­
8.
Stream
network
in
Sipsey
wilderness
area.
