-------
Q.
TS
"c
/.o
7 -
65 -
6-
c jr
5 -
4.5 -
A -
9
-1_
o
i
1
e Q °
f JLJL
i
Lj-J o
o
o
o
o
i i T
< * \
I ri ?
-1- t.V.V.'...
0 ° '
J
0
0
T
X
o
o
T
1
o
0
o
-(a)
_
-
—
-
Q.
7.5
7
6.5 -
5.5
5
4.5
4
Figure 4. Box plots of (a) initial pH, (b) minimum pH, and (c) pH change, for episodes in
ERP streams. Line in box indicates median; upper and lower borders of box
show 25th and 75th quartiles; whiskers indicate 10th and 90th percentiles; circles
represent observations beyond 10th and 90th percentiles.
XXXVII
-------
aoo -
700 -
.£ 4°° '
< 300 -
••§ 200 -
100 -
0 -'
fYY"l —
UUU
^r TOO -
1 600-
£ QY) -
— ^
400 -
1 *n
3 300 "
c
1 200-
2 100 -
700 -
CD 600 -
0)
cB 500 -
O 400 -
2" 300 -
J! 200 -
.1 100 -
U
•4Cf\ ~
-1UU
a
r
a
JL
^
r
X
1
m
cr
S
(0
m
0
I „ ' JL
O ....•:•.'.•.•.•.•
n 0 SspS
T R iJiii
£x&!& t- -.--.-.-.- 1 1
1 o — i — 1
r > j ; 1 Q t y : . L
— 1 T 1 — "—i — ' — F5 — i — • r • — P — i— •— i — | p p
o
B
T - ° I
Q •.-..»..... 1
1- 8 mm
:;:&•?;: o " sli!g;;;
1 a i 8 IT*;
1 ,, T,... o & o ,,".,,
O t 1 ^ T T- ["••''"] °
O Q fa f~ •SB^Bn IAvj||J.IJJ _J_ fl ' O
O
o o °
?
T °
f UJ x
CO
-
_
•(fa)
-
-
-
:(c)
-
-
-
Figure 5. Box plots of (a) initial estimated Allm, (b) maximum estimated Alim, and
(c) estimated Al,m change, for episodes in ERP streams. Line in box indicates
median; upper and lower borders of box show 25th and 75th quartiles; whiskers
indicate 10th and 90th percentiles; circles represent observations beyond 10th
and 90th percentiles.
xxxviii
-------
3. Buck Creek and Linn Run were not chronically acidic but had severe acidic episodes with
low pH and ANC and high Alim for long durations.
4. Bald Mountain Brook and Seventh Lake Inlet had episodes of moderate severity (low pH
and ANC, but moderate Alim levels) and duration.
5. Biscuit Brook and Benner Run experienced acidic episodes but they were of short
duration with moderate pH levels and relatively low Aljm concentrations.
6. Fly Pond Outlet, Black Brook, High Falls Brook, and Baldwin Creek were classified as
nonacidic (with ANC always > 0, except for one brief excursion below 0 in Baldwin
Creek) and had relatively high pH and low Aljm throughout the study period.
The ERP was designed to examine the chemical and biological responses of streams to episodic
acidification. The study ideally would have been conducted during a period when snowmelt and
rainstorms generated hydrologic events with normal or above normal flows. In practice, the
hydrologic conditions during the study varied from region to region. In the Adirondacks, winter
conditions allowed significant snowpacks to develop during both winters of the ERP. In each
case, snowpacks > 50 cm developed, with meltwaters contributing to episodes during late winter
and early spring periods. Spring and fall rainstorms also generated hydrologic events. In the
Catskills, below normal snowpacks developed during both winters. Most of the episodes were
generated by rainstorms. A similar situation developed in Pennsylvania. No significant snow-
packs accumulated in either of the winters during the ERP. Because of the small snowpacks in
Pennsylvania and the Catskills, episodic conditions in these regions may not have been as severe
during the ERP as during years with normal snowpack accumulations. However, large rainstorms
during the spring and late winter did produce high streamflows and major episodes.
Ionic Controls on ANC Depressions
Episodic ANC depressions result from complex interactions of multiple ions that are controlled by
natural processes and acidic deposition. The importance of major ions in determining the mini-
mum ANC during episodes varies among regions and among streams (Table 2). In the Adiron-
dacks, base cation decreases and organic acid (A~) increases most consistently contributed to
ANC depressions during episodes. Nitrate increases were also very important contributors to
episodes in Adirondack streams. In the Catskills, base cation decreases, NO3~ increases, and A~
increases were all consistently important contributors to episodic ANC depressions. For most of
the Catskill streams, base cation decreases were the highest ranked contributors, NO3~ increases
were the second most important contributor, and A~ increases were the third most important ion
changes. In the Pennsylvania streams, SO42~~ increases and base cation decreases frequently
xxxix
-------
Table 2. Number of Episodes for Which Ion Changes Contributed to ANC Depressions and
Mean Rank of Ion Change Contributions (1 = most important) to ANC Depressions
Region/Stream
Adirondacks
Bald Mountain Brook
Buck Creek
Fly Pond Outlet
Seventh Lake Inlet
Catskllls
Biscuit Brook
Black Brook
East Br. Neversink River
High Falls Brook
Pennsylvania
Baldwin Creek
3enner Run
Linn Run
Roberts Run
Stone Run
No.
Episodes*
14
18
21
18
23
11
20
25
13
21
25
24
23
No. Episodes Ion Changes
Contributed to ANC Depressions
so/-
10
10
18
9
1
4
7
5
8
19
15
10
20
N03-
5
15
11
13
20
9
20
13
7
4
12
10
11
cr
9
12
8
8
5
4
9
7
6
3
7
7
11
A"0
12
17
14
16
21
9
19
22
5
13
17
20
19
2CBd
14
17
21
17
18
11
15
23
7
10
24
21
11
Al
0
1
4
0
0
0
0
2
3
2
1
3
2
Mean Rank of Importance13 of Ion
Changes to ANC Depressions
so42-
2.6
3.1
2.3
2.3
3.0
3.2
3.4.
2.0
1.1
1.1
2.3
2.1
1.5
N03-
2.4
2.4
2.8
2.3
1.5
2.2
2.2
1.9
2.4
3.0
3.2
2.8
3.3
cr
3.9
4.3
3.6
3.9
3.4
3.5
3.4
3.4
3.0
2.0
3.0
3.1
3.4
A"0
2.2
1.9
2.8
1.8
2.3
2.6
1.9
2.4
2.6
2.3
2.8
2.2
2.6
2CBd
1.4
1.8
1.0
2.0
1.7
1.1
1.8
1.2
1.3
1.9
1.1
1.1
1.5
Ale
-
3.0
4.0
-
-
-
3.5
2.3
3.0
2.0
3.0
3.5
0 With enough ion measurements to perform analysis.
b Mean rank for episodes in which ion changes made contributions to ANC depressions.
0 A~ = organic acids,
d 2CB = sum of base cations.
e - = no episodes measured in which Al ion changes contributed to ANC depressions.
xl
-------
made strong contributions to episodic ANC depressions. Organic acids, on average, were the
third most important contributor to ANC depressions in all Pennsylvania streams. Nitrate and Cl~
increases made minor contributions to ANC depressions in the Pennsylvania streams.
From a biological perspective, the greatest concern is about major episodes—those that generate
low pH and ANC and high Alim for relatively long periods of time (days to weeks). Ion behavior
that controlled ANC depressions during major episodes recorded in each stream was not neces-
sarily the same as the ion changes most important to smaller, more frequent episodes. In the
Adirondacks, the only region to have major snowpacks, the most severe episodes generally
occurred during spring snowmelt. Nitrate was a more important contributor to ANC depressions
during the major snowmelt episodes than during the other episodes recorded. During major
Adirondack episodes, base cation decreases were usually the most important contributors to ANC
depressions, and A~ and NO3~ changes pontributed to episodes and had similar ranks (2 or 3) of
importance to ANC depressions (Table 3). One Adirondack stream experienced a major snow-
melt episode in which an increase in NO3~ was the most important ion change. Three of the four
Catskill streams experienced one episode influenced by snowmelt in which NO3~ pulses were the
first ranked contributor and base cation decreases were the second ranked ion change (Table 4).
Within regions, the major episodes generated by rainstorms had more variable ionic controls than
did snowmelt episodes. The large rain-induced Adirondack episodes occurred in the spring and
fall (Table 3). In these episodes, increases in A~ and decreases in base cations were the most
important ion changes to ANC depressions. In major rain-driven spring episodes in the Catskills,
decreases in base cations or increases in NO3~ were typically the first or second ranked ion
changes contributing to ANC depressions (Table 4). For the major episodes in Pennsylvania,
base cation decreases and SO42~ increases were usually the most important or second most
important ion changes contributing to ANC depressions (Table 5). Two Pennsylvania streams
experienced major rain-induced episodes in which A~ increases were the most important ion
changes. Nitrate pulses were the second ranked ion change in two major episodes in another
Pennsylvania stream.
Acidic deposition, as evidenced by stream water SO42~ and NO3~ during episodes, contributed
significantly to the occurrence of acidic episodes with low pH and-high Al levels in all three study
areas. Although base cation decreases were frequently the most important ion change that
occurred during episodes, base cation decreases alone cannot create acidic stream water condi-
xli
-------
Table 3. Rank (1 = most important) of Ion Changes Contributing to ANC Depressions
During Episodes with Lowest Minimum ANC Values in Adirondack Streams
Stream
Bald Mountain Brook
Buck Creek
Fly Pond Outlet
Seventh Lake Inlet
Start Date*
3/12/90
3/27/89
5/16/90
9/20/89
11/16/89
11/16/89
3/12/90
4/1/90
5/16/90
4/3/89
3/27/89
3/12/90
4/10/90
4/4/89
4/3/90
3/12/90
3/27/89
11/16/89
5/16/90
5/6/89
Hydrologic
Stimulusb
Snowmelt
Snowmelt
Rain
Rain
Rain
Rain
Snowmelt
Snowmelt
Rain
Snowmelt
Snowmelt
Snowmelt
Snowmelt
Snowmelt
Snowmelt
Snowmelt
Snowmelt
Rain
Rain
Rain
ANC
Minimum
(/zeq/L)
-21
-20
-13
-13
-6
-30
-29
-28
-26
-25
17
19
25
31
39
-20
-18
-9
-9
-8
Rank of Ion Changes Contributing to
Episodic ANC Depressions
1C
2CB
2CB
2CB
2CB
2CB
A~
N03-
2CB
A-
2CB
2CB
2CB
2CB
2CB
2CB
2CB
2CB
A-
A-
A~
2C
N03-
A~
A"
so42-
A"
N03-
2CB
A"
2CB
A"
N03-
A"
A"
so42-
so/-
N03-
A"
N03-
2CB
2CB
3C
A"
NO3"
A"
cr
2CB
A"
N03-
N03-
Al
A"
N03-
so42-
NO37A~
NO37A"
A'
N03-
2CB
N03- .
N03-
4d
—
cr
—
—
cr
—
cr
—
cr
—
—
—
—
—
cr
* Episodes with incomplete ion chemistry not included in analysis.
b Snowmelt includes rain-on-snow events.
c SCB = sum of base cations; A" = organic acids.
d — a no ion change that contributed to ANC depression.
xlii
-------
Table 4. Rank (1 = most important) of Ion Changes Contributing to ANC Depressions
During Episodes with Lowest Minimum ANC Values in Catskill Streams
Stream
Biscuit Brook
1
Black Brook
East Branch
Neversink River
High Falls Brook
-
Start Date8
4/10/90
5/5/89
1/25/90
9/19/89
3/12/90
10/17/89
5/5/89
4/10/90
9/18/89
11/16/89
10/18/89
9/19/89
4/10/90
1/25/90
10/31/89
5/5/89
10/19/89
1/25/90
2/22/90
4/10/90
Hydrologic
Stimulus'5
Rain
Rain
Snowmelt
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Snowmelt
Rain
Rain
Rain
Snowmelt
Rain
Rain
ANC
Minimum
Ueq/L)
-6
-2
-2
-2
-2
17
18
25
36
36
-45
-39
-37
-34
-32
12
21
25
27
28
Rank of Ion Changes Contributing to
Episodic ANC Depressions
1C
N03-
2CB
NO3~
A"
N03-
2CB
2CB
2CB
2CB
2CB
A~
A~
N03-
N03-
A"
2CB
SCB
N03-
2CB
SCB
2C
scB
N03-
A-
N03-
A~
N03-
N03-
N03-
A-
N03-
SCB
SCB
2CB
A~
2CB
N03-
N03-
2CB
N03-
N03-
3°
A~
A~
2CB
2CB
—
A~
A~
A~
N03-
A~
N03-
so42-
A~
SCB
NOg-
A~
A~
A"
A"
A~
4d
_
—
—
—
—
—
_
__
—
N03-
—
—
so42-
—
—
—
—
a Episodes with incomplete ion chemistry not included in analysis.
b Snowmelt includes rain-on-snow events,
0 2CB = sum of base cations; A" = organic acids.
d — = no ion change that contributed to ANC depression.
xliii
-------
Table 5. Rank (1 = most important) of Ion Changes Contributing to ANC Depressions
During Episodes with Lowest Minimum ANC Values in Pennsylvania Streams
Start Date0
Baldwin Creek
Benner Run
Linn Run
Roberts Run
Stone Run
3/30/89
5/6/89
6/20/89
3/24/89
2/15/89
11/16/89
5/14/89
5/9/89
2/2/90
2/15/90
3/4/89
4/17/89
11/16/89
2/14/89
4/1/90
6/20/89
6/3/89
6/25/89
2/2/90
3/29/89
6/3/89
1/16/90
2/2/9Q
2/15/90
4/10/90
Hydrologic
Stimulus
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
ANC
Minimum
(Meq/L)
-1
4
4
5
6
-15
-10
-9
-6
-5
-17
-15
-13
-11
-10
-52
-45
-39
-33
-28
-41
-38
-33
-31
-25
Rank of Ion Changes Contributing to
Episodic ANC Depressions
1b
2CB
2CB
A~
so42-
2CB
so42-
so42-
so42-
2CB
so42-
2CB
2CB
2CB
2CB
so42-
2CB
2CB
2CB
2CB
2CB
so42-
so42-
A~
so42-
2CB
2b
cr
N03-
N03-
Al
so42-
A"
A~
2CB
so42-
A-
N03-
—
N03-
so42-
2CB
A"
so42-
A~
A-
A"
2CB
2CB
so42-
2CB
so42-
3b
_
cr
cr
cr
—
^^
—
A"
_
2CB
so42-
—
A"
N03-
A~
A"
N03-
Al
—
A'
A~
—
N03-
—
4°
—
__
—
__
—
,
—
A"
—
cr
—
N03-
—
cr
—
—
—
—
—
__
—
* Episodes with incomplete ion chemistry not included in analysis.
b 2CB = sum of base cations; A~ = organic acids.
c _ = no ion change that contributed to ANC depression.
xliv
-------
tions during episodes. Organic acid pulses, a natural source of acidity, also were important
contributors to ANC depressions in Adirondack streams and, to a lesser extent, in the Catskill and
Pennsylvania streams. However, SO42~ or NO3~ pulses during episodes in study streams from all
three regions augmented the base cation decreases and organic pulses to create episodes with
lower pH and ANC and higher Al concentrations than would have occurred from natural proc-
esses alone. In addition, large baseline concentrations of SO42~ and NO3~ reduced episodic
minimum ANC, even when these ions did not change during episodes (Figure 6).
We examined episodic acidification in 13 streams that were specifically selected for the ERP. It is
not possible to make direct regional estimates of episodic acidification from these data. However,
the results should be reasonably representative of episodic acidification responses and processes
in similar streams in the Catskills, the Adirondacks, and the Northern Appalachian Plateau of
Pennsylvania.
Effects on Episodic Acidification on Fish Populations in Streams
Results from the ERP clearly demonstrate that episodic acidification can have long-term adverse
effects on fish populations. Streams that experienced episodic acidification had significantly
(p < 0.05) lower levels of brook trout density and biomass than nonacidic streams. Differences in
trout abundance between episodically and chronically acidic streams were not statistically
significant (p > 0.05). Brook trout density and biomass were also significantly (p < 0.001)
correlated with the qualitative rankings of stream chemical severity (r = 0.83 for brook trout
density and 0.92 for biomass; Figure 7). With one exception, acid-sensitive fish populations
occurred only in ERP streams with median weekly pH > 6.0 (Black Brook, High Falls Brook,
Biscuit Brook, Benner Run, Baldwin Run) and were absent from streams that were chronically
acidic or that experienced moderate to severe episodic acidification (Bald Mountain Brook,
Seventh Lake Inlet, East Branch Neversink River, Linn Run, Stone Run, Roberts Run). Likewise,
young-of-the-year brook trout were abundant (indicative of successful reproduction during the
study period) in Fly Pond Outlet, Black Brook, High Falls Brook, Biscuit Brook, Benner Run, and
Baldwin Creek, but were absent or rare in Bald Mountain Brook, Buck Creek, Seventh Lake Inlet,
East Branch Neversink River, and Linn Run.
In situ bioassays demonstrated that fish exposed to episodic acidification can experience signifi-
cant mortality. Fish mortality during bioassays was significantly (p < 0.05) lower in reference
streams (Fly Pond Outlet, Black Brook, High Falls Brook, Baldwin Run, Benner Run) than in
xlv
-------
. (a)
x
Cl A
ofiA
250 -
200 -
150 -
100 -
50 -
J
OAri
oUU -
250 -
9IY1 -
150 -
100 -
50 -
. (b
Catskills
° o
• i ^ i •• ...-. i ^ H
, (e
Catskills
-f
'"f" -+- Jin
S042' NO ' Cl" A"
! ^00 -T
- 250 -
- 200 -
- 150 -
- 100 -
- 50 -
0_
' ^nn -T
- 250 -
200 -
- 150 -
- 100 -
50 -
0 -
. (c)
Pennsylvania
JL
7
t.;fi^:.-y "*" —3P—
; , ; (f)
o Pennsylvania
*
™BK :--:?:;;';
T
o
I .-,,- ^l3.:.-:-:-l • 1 ' ' Ifll '
SO4" NO 3 Cl" A"
Figure 6. Anion concentrations for (a) Adirondack, (b) Catskill, and (c) Pennsylvania streams at the beginning of episodes,
and anion concentrations for (d) Adirondack, (e) Catskill, and (f) Pennsylvania streams at the time of minimum ANC.
Organic acid (A~) concentrations are estimates. Line in box indicates median; upper and lower borders of box show
25th and 75th quartiles; whiskers indicate 10th and 90th percentiles; circles represent observations beyond 10th
and 90th percentiles.
-------
4001
300-
05
7:200
CD
.Q
100 H
p
p
1 1 1 1
2345
Severity of Stream Chemistry (relative ranking)
(a)
3-
"c?
XT
In
CO
O-l
p
p
A
A
C
P
(b)
2345
Severity of Stream Chemistry (relative ranking)
Figure 7. (a) Mean density (number per 0.1 ha) and (b) biomass (kg/0.1 ha) of brook trout
(average of fall 1988, 1989, and spring 1989, 1990 surveys) in ERP study streams
ranked according to overall severity of chemical conditions in the stream (from
most severe, 1, to least severe, 6). Rankings are explained in the text. Streams
are labeled by region: A = Adirondacks, C = Catskills, and P = Pennsylvania.
xlvii
-------
nonreference streams. Mortality rates were significantly higher in bioassays with acidic episodes
than in bioassays with ANC > 0, but similar to bioassays that had chronically acidic conditions
(ANC :£ 0). Results were consistent across all fish species (Figure 8).
Chemical conditions toxic to fish occurred at some time during the study in all ERP streams
except the five reference streams. Maximum observed mortality rates were generally highest in
those streams with the most severe chemical conditions: > 80% in East Branch Neversink River,
Stone Run, Buck Creek, and Linn Run; between 40% and, 50% in Roberts Run and Bald Mountain
Brook; 20% to 30% in Seventh Lake Inlet and Biscuit Brook; and < 20% in all five reference
streams. Because bioassays were not conducted during periods with the most severe chemistry
in each stream, these results provide only a qualitative indicator of the toxicity of ERP streams. All
streams also had at least one bioassay with low mortality (< 10%), indicating that toxic conditions
do not occur throughout the year.
A net downstream movement of radio-tagged brook trout was observed in all streams and study
periods that experienced stressful chemical conditions (e.g., Figure 9). Little to no net down-
stream movement occurred in streams with relatively high pH (> 5.1-5.2) and low Aljm levels
(< 150-160 /ig/L) throughout the study period. Downstream fish movement either was associ-
ated with chronically acidic conditions at the start of the experiment or coincided with the
occurrence of one or more episodes with Alim > 160 fiQ/L for 1.5 or more days. Radio-tagged
brook trout died when exposed to high concentrations of aluminum during an episode in Linn
Run. During the same episode, radio-tagged trout survived if they avoided exposure to peak Al
levels.
All ERP streams except the East Branch of the Neversink River had chemical conditions during
low flow considered suitable for fish survival and reproduction (pH ^ 6.0; Aljm < 60 yjeq/L). Thus,
stream assessments based solely on chemical measurements during low flow do not accurately
predict the status of fish communities in small streams.
Episode Characteristics that Determine the Severity of Effects on Fish
Inorganic aluminum was the single best predictor of fish mortality during in situ bioassays.
Calcium, pH, and DOC were also important predictors of brook trout mortality; at a given Alim
concentration, lower mortality occurred in bioassays with lower minimum pH, higher calcium, and
higher DOC.
xlviii
-------
x
Brook Trout
Common Pool
(a)
90
80
70
60
50
40
30
20
10
0
I
Reference Acidic Chronic
Episode Acidity
Sculpin
(b)
g 90 -
>\
£ 70 -
2 60 -
| 50 -
40 -
P 30 -
1 20 -
S 10 -
0 -
©
-e—
T
Reference Acidic Chronic
Episode Acidity
All Brook Trout
*
O
•o
CM
100
90
80
70
60
50
40
30
20
10
0
(c)
0
O
T
i i i
Reference Acidic Chronic
Episode Acidity
Dace
(d)
"3^
^
-*J
"o
i
o
o
-Q
I
O
CM
100
90
80
70
60
50
40
30
20
10
0
T
I
1
-
-
-
-
_ -p
-
-
Reference Acidic
Episode
Figure 8. Box plots of percent mortality after 20 days in in situ bioassays classified
according to ANC: reference (nonacidic with ANC always > 0); chronic acidity
(ANC always ^ 0); and acidic episode (initial ANC > 0 with at least two
consecutive values < 0 during 20-day period). Line in box indicates median %
mortality; upper and lower borders of box show 25th and 75th quartiles; whiskers
indicate 10th and 90th percentiles; circles represent observations beyond 10th and
90th percentiles.
xlix
-------
^^
P n -
d U
1 -100-
E
2 -200-
o
5 -300 -
1 -400 •
.CO
"6
Q) -500 "
/
n 1 ' • d
WV^t
7 7 i
*
/
Hj-CH
[ 1
6
•
4 4 4 W ^ . I
' i ^ ™ n « .„. r: n
*-A.A-T N
6
HD-t>Q-CWl-tHI- i>
-A,%
NA-.
r-A*A-A-i
3 3
§ -100-
E
.A-A-A 0 _2QQ _
0
, 5 -300-
3 Q
^ -400 -
,co
. T3 crrvA
1 1 1 1 Q) -OUU -|
14 14 14 13 12 " I
_._„„, x ^ P^ — JL — 1 J__ J.
^ ^^ryo*.
i— -U
I 12 VM
1<5 f-A-*-6-}-^..
I 1 ^
10 9 V,
/
/
.-A-A-4
\
I
4 ^^^
*•
4
i i » i i b
10
15
20 25
30
10
15
20
25
(b)
30
600 '
400 -
200-
5 10 15 20 25
Days from October 16,1988
(c)
30
600 -\
400 •
200
10
15
A
i
20
25
Days from March 13,1989
(d)
30
Figure 9. Total dissolved aluminum (Altd) concentrations at continuous monitoring stations and median net movement of
brook trout in Linn Run (triangles) and Baldwin Creek (squares). Negative values for movement indicate down-
stream relative to initial location. Sample size and ± 1 SE (vertical bars) are shown for fish movement on repre-
sentative days (Source: Gagen, 1991).
-------
Brook trout density and biomass in ERP streams were significantly (p < 0.05) correlated with both
stream pH and Aljm concentrations, although rank correlations were slightly higher for pH than for
Aljm. Because pH and Alim are highly correlated, it is difficult, based on field data alone and given
the small number of study streams, to distinguish the relative importance of pH and Alim.
The relationships between Alim and pH, and between Alim and ANC, varied among ERP streams.
High levels of Alim (> 200^g/L) commonly occurred in Linn Run at pH < 5.4, but only at.pH
< 5.2 in Stone Run, Bald Mountain Brook, Buck Creek, and Seventh Lake Inlet, and at pH < 5.0
in Roberts Run and East Branch Neversink River. [No Aljm levels above 200/fg/L occurred in the
remaining six ERP streams.] Thus, predictions of potential effects on fish based solely on pH or
ANC may be misleading.
In general, fish exposed for longer periods of time to high Aljm levels had higher mortality rates.
However, the single best predictor of brook trout and sculpin mortality was the time-weighted
median Alim concentration during the 20-day bioassay period, as opposed to more complex
expressions of chemical exposure incorporating peak levels and duration.
Blacknose dace are more sensitive to high Aljm than brook trout or sculpin. Dace mortality was
best predicted by an integrated function of duration and Alim concentrations. High mortality is
expected to occur with Alim > 250 j«g/L for two or more days.
During radiotelemetry studies, some brook trout were able to move downstream or into alkaline
microhabitats and avoided exposure to low pH and high Al during episodes. However, in most
cases, the majority of radio-tagged fish were exposed to relatively high, potentially lethal Al levels
(Figure 10).° Fish behavioral avoidance and the occurrence of refugia can partially, but not
entirely, mitigate the adverse effects of episodic acidification on fish populations. Recolonization
from groundwater seeps and more alkaline tributary streams can maintain low densities of fish in
streams that experience toxic episodes, but is not sufficient to sustain fish densities oFFiomass at
levels near those expected in the absence of adverse acid-base chemistry.
Brook trout are fairly mobile, frequently moving more than 1 km. Thus, of the fish species
common in small headwater streams, brook trout are best able to take advantage of refugia.
Sculpin, by contrast, generally move only short distances (e.g., < 10 m in a summer). Sculpin
were as, or more, tolerant of high Alim in bioassays than brook trout. Yet, sculpin populations are
absent from streams that maintain low densities of brook trout. We hypothesize that toxic
li
-------
600
400 -
sT 200 -I
0
0
10 15 20
Days from March 3,1989
25
30
Figure 10. Median stream total dissolved AI (Altd) concentration at locations of radio-
tagged fish (triangles; vertical bars = ± 1 SE) and Altd concentrations of
concurrent samples collected at the Linn Run continuous monitoring station
(broken line) (Source: Gagen, 1991).
episodes have a more severe and long-lasting effect on sculpin populations, compared to brook
trout populations, because of differences in fish mobility.
Major Conclusions
• Episodes were a common occurrence in the study streams of all three regions, and acidic
episodes were common when ANC values were < 50 |ieq/L immediately before the episode.
When acidic episodes occurred, they were accompanied by depressed pH levels and elevated
Aljm concentrations.
• Acidic deposition, as evidenced by stream water SO42~ and NO3~ during episodes, contributed
significantly in two ways to the occurrence of acidic episodes with low pH and high AI levels in
all three regions. Pulses of SO42~ (in Pennsylvania streams) and NO3~ (in Catskill and
Hi
-------
Adirondack streams) during episodes augmented natural processes to create episodes with
lower ANC and pH and higher Aljm concentrations than would have occurred from natural
processes alone. In addition, large baseline concentrations of SO42~ (all regions) and NO3~
(Catskills and Adirondacks) reduced episodic minimum ANC levels, even when these ions did
not change or decreased slightly during episodes.
• Episodic acidification can have long-term adverse effects on fish populations. As a result,
stream assessments based solely on chemical measurements during low flow do not accur-
ately predict the status of fish communities in small streams.
*
• Fish exposed to low pH and high Aljm for longer periods of time experienced higher mortality.
Time-weighted median Aljm concentration was the single best predictor of brook trout
mortality.
• Fish behavioral avoidance only partially mitigated the adverse effects of episodic acidification
in small streams and was not sufficient to sustain fish density or biomass at the levels
expected in the absence of acidic episodes.
References
Gagen, C.J. 1991. Direct Effects of Acidic Runoff Episodes on the Distribution and Abundance of
Fishes in Streams of the Northern Appalachian Plateau. Ph.D. Thesis. The Pennsylvania State
University, college Station, PA. 127 pp.
Ranalli, A.J., B.P. Baldigo, D.A. Horan-Ross, R.V. Allen, and P.S. Murdoch. 1993. Site Descrip-
tions, Data and Data Collection Methods, and Quality Assurance/Quality Control Procedures
Used in the Study of Episodic Stream Acidification in Four Catskill Mountain Streams, 1988-
1990. Open File Report OF 93-137. U.S. Geological Survey, Albany, NY. 80 pp.
liii
-------
-------
SECTION 1
INTRODUCTION
Episodic acidification refers to short-term decreases of acid neutralizing capacity (ANC) that often
accompany rainstorm runoff and snowmelt events (Wigington et al., 1990). Typically, if a stream
or lake experiences an episodic ANC decrease—an episode—there are accompanying changes
in the concentrations of other chemical solutes, such as H+ and Al. Short-term exposures to low
pH and elevated Al during episodes can cause significant stress and increased mortality of
aquatic biota (J. Baker et al., 1990a).
The Episodic Response Project (ERP), originally funded as a component of the U.S. Environ-
mental Protection Agency's (EPA) Aquatic Effects Research Program (AERP), was designed to
address major knowledge gaps concerning episodic acidification. With the exception of the ERP,
AERP research focused mostly on chronic acidification, that is, long-term changes in surface
water acid-base chemistry and associated changes in biological communities. The National
Surface Water Survey (NSWS) measured the current, chronic, acid-base chemical status of sur-
face waters in regions of the United States potentially sensitive to acidic deposition (Linthurst et
al., 1986; Landers et al., 1987; Kaufmann et al., 1988). The Direct/Delayed Response Project
(DDRP) projected future changes in annual average surface water chemistry for a range of alter-
native future deposition scenarios (Church et al., 1989, 1992). Biological responses to chronic
changes in acid-base chemistry were evaluated as part of the Little Rock Lake whole-lake acidifi-
cation experiment (Brezonik et al., 1986; Watras and Frost, 1989) and other AERP projects.
Because of these and other studies, a relatively sound knowledge base exists regarding chronic
acidification and its effects.
However, many uncertainties remain regarding the occurrence, causes, and biological conse-
quences of episodic acidification. For example, the NSWS used index samples collected at
specific times of the year and specific locations to assess the current chemical status of surface
waters. The DDRP projected future chemical changes based on index chemistry and a variety of
soil and watershed characteristics. These projects yielded valuable research results that were
used to address critically important policy questions related to acidic deposition. However,
assessments based on index conditions have not adequately reflected the influence of episodes.
It has not been possible to accurately estimate the worst-case chemical conditions associated
with episodes for the populations of surface waters potentially sensitive to acidic deposition. The
processes and forcing factors controlling episodes have remained poorly quantified. Further-
1
-------
more, a very important factor is that the degree and severity of biological effects of episodic
acidification, especially population-level effects on fish, has remained uncertain.
1.1 PROJECT OBJECTIVES
The ERP was conducted to address key uncertainties regarding episodic acidification. .The
project objectives were as follows (Thornton et a!., 1988):
1. Determine the magnitude, duration, frequency, and characteristics of episodic chemical
changes that accompany hydrological events (snowmelt and rainstorms) in streams.
2. Evaluate the effects of episodic acidification on fish populations in streams.
3. Define key characteristics of episodes that determine the severity of effects on fish
populations.
4. Develop and calibrate regional models of episodic chemistry that link atmospheric
deposition to biologically relevant chemistry during episodes.
1.2 PROJECT SCOPE
Episodic acidification is of concern in regions of the United States susceptible to adverse impacts
to aquatic ecosystems because of acidic deposition. The ERP research efforts focused on three
priority areas: the Northern Appalachian region of Pennsylvania and the Catskill Mountains and
Adirondack Mountains of New York. These regions were selected because of three factors: the
importance of their stream resources, the likely occurrence of acidic episodes, and the high
probability that episodes are an important factor affecting the current status of fish populations in
these areas (Thornton et al., 1988; Wigington et al., 1990). Wigington et al. (1990) concluded that
(1) in the United States, acidic episodes occur most often in the Northeast and Mid-Atlantic
regions, including the Adirondacks, Catskills, and Northern Appalachian region of Pennsylvania,
and (2) atmospheric deposition seems to have the greatest influence on minimum ANC and pH of
episodes in these areas. All three regions have a relatively high percentage of chronically acidic
surface waters [14% of the lakes in the Adirondacks and 6% of the upper stream reaches in the
Poconos/Catskills and Northern Appalachian Plateau regions (Linthurst et al., 1986; Kaufmann et
al., 1988)]. In addition, preliminary modeling analyses by Eshleman (1988) estimated that the
numbers of acidic systems in these regions could increase 3-6 times, depending on the region.
Episodic acidification occurs in both lakes and streams. ERP research efforts were conducted
only in streams, however, for three reasons. First, the available evidence suggested that episodes
-------
probably have more significant impacts on fish populations in streams than in lakes (J. Baker et
al., 1990a). Episodes in lakes are often accompanied by a high degree of spatial variability that
may mitigate the effects on fish and other biota. Second, sampling episodic chemical conditions
and biological responses to episodes is much more difficult in lakes than in streams, because of
the spatial variability in lakes as well as the safety problems associated with collecting samples
during spring snowmelt on lakes with thin ice covers. Third, episodic responses in streams are
more directly linked to the watersheds. Therefore, understanding episodic acidification in streams
is a necessary precursor to understanding the more complex episodic responses of lakes.
ERP field research did not directly address the regional extent of episodic acidification.
Eventually, to aid in decisions regarding acidic deposition controls, we would like to know the
number, area, and geographic distribution of waters that experience episodes severe enough to
cause significant adverse biological effects. As part of the ERP planning process, we decided,
however, that (1) insufficient information was available to determine what characteristics of
episodes were most important in controlling biological effects and, therefore, most important to
measure, and (2) regional surveys of episode chemistry, similar to the NSWS assessment of
chronic acidity, are not logistically feasible because of the transient nature of episodes (Thornton
et al., 1988). For these reasons, the ERP was designed as an intensive study, rather than an
extensive regional survey. Detailed field monitoring and research were conducted at a relatively
small number of selected sites. Five types of monitoring data were collected at the ERP study
sites: (1) atmospheric deposition quantity and quality, (2) stream discharge, (3) stream chemistry,
(4) fish population status and fish responses to episodes, and (5) qualitative information on
stream benthic invertebrates.
Results of the ERP research are presented in this report. In addition, ERP data are being used to
develop and refine models with which, together with the NSWS data, it may be possible to esti-
mate the regional extent of biologically significant episodes. The results from these model
development efforts are not included in this report but will be described in subsequent journal
articles.
1.3 PROJECT ORGANIZATION
The ERP was a cooperative research effort involving scientists from a number of different institu-
tions and agencies. Field research was conducted by regional cooperators in three areas of the
United States:
-------
1. The Adirondack Lakes Survey Corporation (Walter A. Kretser and Howard A. Simonin,
principal investigators) in the Adirondack Mountains of New York.
2. The U.S. Geological Survey (Peter S. Murdoch, principal investigator) in the Catskill
Mountains of New York.
3. The Environmental Resources Research Institute at Pennsylvania State University
(David R. DeWalle, principal investigator) in the Northern Appalachian Plateau of
Pennsylvania.
The U.S. EPA Environmental Research Laboratory in Corvallis, Oregon (ERL-C), served as the
project lead. The U.S. EPA Environmental Monitoring Systems Laboratory in Las Vegas, Nevada
(EMSL-LV), developed and guided the ERP quality assurance/quality control (QA/QC) program.
EMSL-LV also developed the field-based data management system. ERL-C performed central
database management functions. The Pacific Northwest Laboratory designed and coordinated
deposition monitoring for the project, working in cooperation with the U.S. EPA Atmospheric
Research and Exposure Assessment Laboratory, Research Triangle Park, North Carolina.
1.4 REPORT SCOPE AND ORGANIZATION
This report describes the major findings of the field component of the ERP and thereby addresses
the first three of the four objectives listed in Section 1.1. The stream chemistry section describes
the occurrence and characteristics of episodes and explains how ion changes during episodes
may indicate the causes of episodic acidification. Biology results include measures of fish
toxicity, behavioral responses, and population status in the study streams in relation to stream
chemistry and selected episode characteristics. An important objective of the repcrt is to compare
these chemical and biological responses among the three ERP regions. More intensive analyses of
the data collected for any one region are being conducted by the individual regional cooperators, for
publication in the peer-reviewed literature.
Funding constraints have hindered completion of a full analysis of the ERP data set. The ERP budget
was reduced substantially in the third and final year of the project (1990), during the final stages of the
original National Acid Precipitation Assessment Program (NAPAP). As a result, most field research
was completed, but few resources were left for data analyses by cooperators or ERL-Corvallis staff.
The analyses presented in this document have been accomplished through the persistent efforts of a
relatively small group of people. Consequently, the report contents reflect very difficult choices
regarding the most critical analyses to be performed and presented. Two criteria guided the prioritiza-
tion of work for the report. First, analyses should address the ERP objectives as directly as possible.
Secondly, the work should be a logical step on which subsequent analyses could be based. Some
-------
analyses originally planned could not be conducted. For example, the report provides only a brief
description of deposition characteristics during the ERP study period. However, analyses that
explicitly examine the linkage between acidic deposition and episodic acidification could not be
included. Some of the biology sections rely on summaries of analyses conducted by the field
cooperators, separately for each region, rather than analysis of the integrated data set.
Additional work involving ERP data is continuing. Modeling analyses, to address the fourth ERP
objective, are being conducted by Keith N. Eshleman (University of Virginia), Trevor D. Davies
(University of East Anglia, UK), Martyn Tranter (University of Bristol, UK), and P.J. Wigington, Jr.
(ERL-C). Furthermore, both the individual field cooperators and ERL-Corvallis staff are planning future
analyses of the ERP data. Results from these efforts will be published in peer-reviewed journals.
The remaining sections of this report are organized as follows:
• Chapter 2, Importance of Episodic Acidification, provides a brief review of the current
understanding and uncertainties regarding the extent, severity, causes, and biological
effects of episodic acidification.
• Chapter 3, Study Areas, provides a general description of the regions studied as well as the
specific study sites (including information on atmospheric deposition).
• Chapter 4, Methods, provides an overview of the field, analytical, and data analysis
methods used to collect and interpret the ERP data, including procedures for QA/QC.
• Chapters 5 and 6 present and discuss project results, organized by topic area:
Chapter 5, Stream Hydrology and Chemistry
- Chapter 6, Effects on Fish
• Chapter 7, Summary and Conclusions, reiterates the major findings and conclusions of the
ERP.
-------
-------
SECTION 2
IMPORTANCE OF EPISODIC ACIDIFICATION
This chapter provides a brief overview of our current understanding of the extent, severity, causes,
and biological importance of episodic acidification in streams. It provides the background infor-
mation necessary for interpreting the results from the ERP and identifies the key knowledge gaps
and uncertainties that the ERP was designed to address. This summary is based primarily on the
recent NAPAP state of science and technology reports on episodic acidification (Wigington et al.,
1990) and the biological effects of acidification (J. Baker et al., 1990a).
We define acidification as a decrease in surface water ANC. Acidic waters have no ability to
neutralize strong acids, that is, ANC < 0. Episodic acidification, therefore, is the process by
which a lake or stream experiences a transient, short-term decrease in ANC, usually during a
hydrologic event, over a time scale of hours to weeks. A hydrologic event is an increase in
stream discharge or flow resulting from rainfall or snowmelt. Thus, episodes make up the subset
of hydrologic events during which ANC decreases; during acidic episodes, ANC < 0.
Although acidification and episodes are defined on the basis of a decrease in ANC, changes also
occur in a large number of other water chemistry variables. Of particular concern for aquatic
biota are the decreases in pH and the increases in inorganic aluminum that accompany surface
water acidification. Laboratory experiments have demonstrated that both low pH and elevated
levels of inorganic aluminum are toxic to fish and other biota (Schofield and Trojnar, 1980; Baker
and Schofield, 1982; Clarke and LaZerte, 1985; Ingersoll, 1986; France and Stokes, 1987; Holtze
and Hutchinson, 1989).
2.1 CHARACTERISTICS OF EPISODES
Figure 2-1 presents a hydrograph (representation of discharge over time) and a chemograph
(representation of chemical concentration over time) for an idealized hydrologic event and an
associated episode. It illustrates several important episode characteristics. Episode magnitude
has two components: (1) the maximum change in the concentration of ANC or other chemical
variable of interest, such as pH or inorganic aluminum, represented by the change from point A to
point B along the Y-axis in the figure, and (2) the absolute minimum value (or maximum, for
variables such as aluminum that increase during episodes) that occurs during the event (point B).
-------
t
CD
O)
CO
o
(0
— Discharge
- ANCorpH
O
O
Time
Figure 2-1. Variation of streamflow and ANC (or pH) for an idealized episode.
Episode duration is the amount of time that the ANC or pH depression (or elevated aluminum)
lasts, that is, the time from point A to point C along the X-axis. For some purposes, duration is
also specified as the amount of time below (or above) a given threshold value, such as ANC
s 0 or biologically significant thresholds of pH and inorganic aluminum. Available information,
discussed further in Section 2.4, indicates that episode magnitude and duration are the two most
important features determining the severity of the toxic effects on aquatic biota.
Aquatic systems rarely behave, however, in the idealized fashion illustrated in Figure 2-1. Multiple
rainstorms or combinations of rain and snowmelt frequently produce complex hydrographs and
chemographs that make it difficult to describe episode magnitude and duration. As a result, one
of the most challenging tasks facing scientists is to select an appropriate index of episode
chemical severity that can be consistently applied to assess the biological significance and
impacts of episodic acidification in streams.
The nature of the hydrologic events that occur in a watershed provides important information
about the possible characteristics of episodes in that system. For example, the duration of an
event also defines the maximum possible episode duration. In northerly regions, where large
snowpacks accumulate, hydrological events in the spring may last several weeks or more as the
snow gradually melts. Rainstorms, on the other hand, produce hydrological events ranging from
8
-------
r
several hours to several days in low-order streams. Thus, episodes in small streams associated
with individual rainstorms are likely to be no more than a few days long. Rainstorm generated
episodes may last several weeks in large rivers. During the spring snowmelt period, shorter,
more severe episodes may be superimposed on longer snowmelt episodes.
Numerous researchers have reported strong correlations between stream discharge and concen-
trations of selected chemistry variables, such as pH, ANC, and sulfate (Johnson et al., 1969;
Lewis and Grant, 1979; Sharpe et al., 1984; Eshleman, 1985). In such situations, the stream
hydrograph may also be used to estimate, with quantifiable uncertainty, stream chemistry and
episode magnitude. Several limitations to this approach have been noted, however. In particular,
concentration-discharge relationships often vary seasonally with differences in antecedent
conditions (e.g., the amount of time since the last storm event) and with the rising and falling limb
of the discharge curve. For example, for a given level of discharge and pH, inorganic aluminum
concentrations are generally higher during the early phases of snowmelt, as water levels rise,
than during later phases of the event (Sullivan et al., 1987; Hooper and Shoemaker,' 1985).
Possible explanations include changes in hydrological flowpaths over the course of the event (see
Section 2.3) and the redissolution of aluminum from the streambed during the initial pH decline
(Norton et al., 1992). Patterns of chemical changes during events and concentration-discharge
relationships vary among regions as well as among stream systems within a given region
(Wigington et al., 1990).
Frequency refers to the number of occurrences of episodes, of a given magnitude and duration,
during a specified period (e.g., year). Because of their linkage to hydrological events, episodes
are stochastic or probabilistic in nature. Some years are dry and tend to have few significant
hydrological events and therefore few potential episodes. In other years, wetter conditions pre-
vail, allowing a greater opportunity for hydrological events and episodes to occur more frequently.
Only by monitoring episodic acidification in an aquatic system for several years can the full range
of the system's response be reasonably assessed.
Additional complexity arises from the spatial variability of chemical conditions during episodes.
Seeps, springs, and tributaries that deliver water to streams can create refugia with chemical
characteristics quite different from those in the main body of water. In addition, in most stream
systems, ANC and pH generally increase with increasing distance downstream (Johnson et al.,
1981; Driscoll et al., 1987b; Kaufmann et al., 1988). As stream order and watershed size
increase, the contribution of groundwaters, with higher ANC and pH, to stream baseflow tends to
-------
increase (Winter, 1984; Freeze and Cherry, 1979). One would also expect episodes to be less
severe in higher order streams, although relatively few studies have been conducted (e.g.,
Hooper and Shoemaker, 1985) on longitudinal trends in episodic chemistry to confirm this
assumption.
2.2 EXTENT AND SEVERITY OF THE PROBLEM
Episodic acidification is widespread. Throughout the United States, Canada, and Europe,
streams and drainage lakes usually experience at least some loss of ANC during hydrological
events (Wigington et al., 1990). Although some exceptions exist, that is, systems that experience
no change or even an increase in pH or ANC during events, these systems or episodes represent
a small proportion of those studied.
Although the occurrence of episodes is almost universal in the United States, episode magnitude
varies widely both among regions and among systems within the same region (Figure 2-2).
According to available studies, streams and lakes in the Northeast most consistently experience
episodes with low minimum values of ANC and pH. Every northeastern state for which data are
available (New York, Vermont, New Hampshire, Maine, and Massachusetts) has streams or lakes
that experience episodes with minimum pH < 5.0 and/or minimum ANC < 0 (with pre-episode
conditions of pH 5.5 or greater). In the southwestern Adirondacks, some streams with baseflow
ANC > 200^eq/L may become acidic during snowmelt events (Galloway et al., 1987), and many
Adirondack streams and lakes with pre-episode ANC up to 75 fieq/L become acidic during epi-
sodes (Colquhoun et al., 1984; Schaefer et al,, 1990). Episodes with minimum ANC < 0 or mini-
mum pH < 5.0 have also been recorded in the Mid-Atlantic region, specifically in the Valley and
Ridge area of Pennsylvania and the Atlantic Coastal Plain of Maryland (Lynch et al., 1986; Sharpe
et al., 1984; Correll et al., 1987). Outside the Northeast and Mid-Atlantic regions, most of the
episodes recorded have not resulted in acidic conditions, although only limited data are available.
Despite a large amount of published information, significant knowledge gaps remain concerning
the extent and severity of episodic acidification, primarily because measuring episodic acidi-
fication is a difficult and expensive task. Adequately capturing the temporal dynamics and spatial
variations in chemistry during episodes requires large investments in field personnel and instru-
mentation. In many of the existing studies, samples were collected too infrequently to provide
highly accurate data on episode magnitude and duration. Studies that use sampling intervals that
10
-------
cr
(U
o
<
o
V)
'a.
UJ
Minimum
300
250
200
150
100
50
0
-50
-100
A New England
O Adirondacks, NY
• Catskills Mtns, NY
A N, App. Plateau/Ridge & Valley, PA
100-50 0
I
50
T I I I
100 150 200 250 300
Pre — episode ANC
in
Q.
CD
•o
O
.—
LU
Figure 2-2.
8
7 -
6 -
A New England
O Adirondacks, NY
• Catskills Mtns, NY
A N. App. Plateau/Ridge & Valley, PA
O
__ O
OA O
r>O
HOO AO
OA « •oo" •
O
8
Pre —episode pH
Minimum ANC and pH for the episode with the lowest recorded minimum ANC
or pH value for streams and lakes in the Northeast (including Pennsylvania).
Each symbol represents an individual stream or lake (data from Wigington et
al., 1990).
11
-------
are long in comparison to the rate of chemical change are likely to underestimate the total
magnitude of the change, the minimum ANC, and, to some degree, the duration of episodes.
Furthermore, variables of interest, such as inorganic aluminum, often were not routinely
measured.
At present, there are not enough data for any region of the United States to estimate the regional
extent, magnitude, duration, and frequency of episodic acidification. In particular, it has proven to
be logistically infeasible to conduct statistically rigorous surveys that would allow for population
estimates of episode chemistry comparable to the regional population estimates of chronic acidity
provided by the National Surface Water Survey (MSWS) (Linthurst et al., 1986; Landers et al.,
1987; Kaufmann et al., 1988).
The best approach for achieving population-level estimates of episodic acidification will probably
be to link improved regional databases with simple modeling approaches. Eshleman (1988) con-
ducted a preliminary analysis, using a two-component mixing model together with the NSWS
chemistry data, to generate worst-case estimates of the minimum ANC during episodes in
streams in six regions of the eastern United States (Table 2-1). His results suggest that the
estimated proportion of acidic streams (ANC < 0) would increase by 40-640%, compared to the
NSWS results in the six regions, if the effects of episodic acidification were taken into account.
Neither the model nor the model coefficients used by Eshleman (1988) have been fully validated,
however. Thus, these results must be considered preliminary. Data from the ERP are being used
to refine and improve these and other models, with the objective of providing improved regional
estimates of the extent and magnitude of episodic acidification.
2.3 CAUSES OF EPISODIC ACIDIFICATION
Several processes and factors influence the nature and severity of episodic acidification. Physical
factors, such as watershed hydrology, as well as chemical interactions are important.
2.3.1 Watershed Hydrology
The changes that occur in water flowpaths through the watershed during hydrological events are
an important determinant of the characteristics of episodes in receiving streams and lakes.
During periods of baseflow, relatively alkaline water is derived from the lower part of the mineral
soil and deeper groundwater storage zones (Velbel, 1985) (see Figure 2-3). During hydrologic
events, on the other hand, water is routed primarily through upper soil layers (Potter et al., 1988;
Chen et al., 1984), which are more acidic because of natural processes or acidic deposition.
12
-------
Table 2-1. Population Estimates of the Number and Proportion of Acidic Reaches (ANC
< 0) Based on Index Conditions and Worst-Case Episodic Conditions
Using the Two-Component Mixing Model (from Eshleman 1988)
Subregion
Poconos/Catskills (1D)
Southern Blue Ridge (2As)
Valley and Ridge (2Bn)
Northern Appalachian
Plateau (2Cn)
Ozarks/Ouachitas (2D)
Southern Appalachians (2X)
Piedmont (3A)
Chesapeake Area (3B)
Florida (3C)
Node3
(lower)
(upper)
(lower)
(lower)
(upper)
(lower)
(upper)
(lower)
(upper)
(lower)
(upper)
(lower)
(upper)
(lower)
(upper)
(lower)
(upper)
Index Conditions
(ANC < 0
Number of
Reaches
0
209
Ob
Ob
636
326
499
Ob
Ob
Ob
121
Ob
Ob
772
1,334
225
678
Proportion
(%)
0.0
6.4
Ob
Ob
4.9
3.8
5.8
Ob
Ob
Ob
2.5
Ob
Ob
6.8
11.8
14.5
39.2
Episodic Conditions
(ANC < 0)
Number of
Reaches
157
746
39
47
1,126
2,379
3,224
Ob
75
243
364
Ob
Ob
1,727
3,132
520
963
Proportion
(%)
4.8
23.0
2.2
0.3
8.6
28.0
37.2
Ob
1.8
4.8
7.4
Ob
Ob
15.3
27.8
33.5
55.7
Preliminary estimates are given for both upper and lower sampling nodes for nine NSWS subregions.
No acidic reaches were sampled; although the best estimate is zero, the upper 95% confidence bound on the estimate
does not preclude a certain number of acidic systems in the target population.
13
-------
ATMOSPHERE
!
!
.c
I
.c
I
o
Snow
Rain
FC
FOREST CANOPY
Throughfall
SP SNOWPACK
Rain or
Snowmelt
Snowmelt
WS WATERSHED SURFACE
Infiltration
Qel
OL ORGANIC LAYER
Qe2
Matrix Flow
Qpl
UMS UPPER
MINERAL SOIL
Qe3
o
TO
1
O
f
I
, Matrix Flow
Qe4
LMS LOWER
MINERAL SOIL
Qp3
Deep Percolation
Qe5
< w
{11
Event
STREAM or LAKE
Water
DCS DEEP
GROUNDWATER STORAGE
Qp4
Qp5
Figure 2-3. Episodic acidification conceptual model. ln = direct Input; Qen = event water;
Q = pre-event water (from Wigington et al., 1990).
14
-------
Differences in the characteristics of these two water sources will determine the degree of epjsodic
acidification.
A considerable body of literature has demonstrated that streamwater during events is composed
of (1) event water derived directly from the rainstorm or snowmelt driving the hydrologic event
(flowpaths Qe1-Qe4 in Figure 2-3) and (2) pre-event water residing in the watershed before the
rainstorm or snowmelt occurred (Qp1-Qp4 in Figure 2-3) (Wigington et al., 1990). Results from
natural isotope studies suggest that pre-event water typically accounts for 30-90% of the water
transported during hydrological events (Hooper and Shoemaker, 1986; Eshleman, 1988).
As water moves through the watershed system, it may be chemically altered by numerpus bio-
geochemical reactions. Basic differences exist between the opportunities for event water and for
pre-event water to be modified by these biogeochemical processes. By definition, event water
moves rapidly through the system and generally has less opportunity to interact with soils and
rock. For watersheds receiving acidic deposition, this means that event water is less likely to be
neutralized before being delivered to a waterbody (Winter, 1984). Pre-event water resides in the
soil/geologic complex for a longer period of time, increasing the opportunity for neutralization
reactions to take place (Anderson and Bowser, 1986). The combination of the various sources of
event and pre-event water determines the net chemical characteristics of water delivered to a
stream or lake at any moment in time.
2.3.2 Natural Processes that Contribute to Episodic Acidification
Four important natural processes can produce episodes: (1) dilution, (2) nitrification, (3) organic
acid production, and (4) the sea-salt effect. If present, weathering and oxidation of mineral S
within the watershed can also induce episodic acidification. However, we focus our discussion on
the four more commonly important processes.
Dilution occurs when precipitation water of low ionic strength (event) is mixed with higher ionic
strength pre-event water in soil water or in a lake or stream. For dilution to occur, modification of
the precipitation water by chemical reactions in watershed soils must be limited by either little
physical contact or a short interaction time. Dilution can lower ANC, but dilution alone cannot
cause an acidic episode. Episodes caused by dilution are characterized by decreases in base
cation concentrations (sodium, potassium, calcium, and magnesium), which are derived largely
from watershed soils.
15
-------
Nitrification is the biological conversion of organic nitrogen and reduced forms of inorganic nitro-
gen, such as ammonium, into nitrate. The end products of these reactions include hydrogen ions
(H+) as well as nitrate anions (NO3~). Nitrification typically occurs in the organic layers of soils.
Thus, waters moving through these upper soil layers during events are likely to flush out any
accumulations of nitrate and acidity from nitrification. Episodes caused by nitrification are char-
acterized by relatively high nitrate concentrations (Galloway et al., 1987; Peters and Driscoll,
1987).
Organic acids are also produced in soil organic layers, as byproducts of the partial decomposi-
tion of organic matter. A high proportion of the organic acids in wetland systems is exported
directly to surface waters. In forest soils, on the other hand, organic acids generated in the
organic layer tend to be adsorbed onto the mineral surfaces in the lower horizon mineral soils.
When flowpaths bypass these mineral soils, however, as during some events, forest soils also can
export organic acids (Turner et al., 1990). Hence, the production of organic acids in the organic
layer and the minimal adsorption of these acids during transit to the watercourse are prerequisites
for episodes stimulated by organic acids. Episodes caused by organic acid production typically
are characterized by an increase in the streamwater organic anion content and, depending on the
ion exchange characteristics of the organic layer, a possible decrease in base cation concentra-
tions.
The sea-salt effect occurs when base cations from atmospheric deposition displace H+ ions from
the soil complex; these H+ ions may then be exported to streams and lakes. Two conditions are
required for this process to be a significant episode-generating mechanism. First, there must be
an ample supply of neutral salts in wet or dry deposition. Therefore, the sea-salt effect generally
occurs only in areas along oceanic coastlines. Second, H+ ions must be readily available on the
soil cation exchange complex. Thus, the sea-salt effect can cause episodic acidification only in
watersheds with soils that have been chronically acidified by either natural or anthropogenic
factors. The chronic acidification of soils by acidic deposition, therefore, could theoretically lead
to episodes stimulated by sea salts, or episodes of greater magnitude, in regions where otherwise
the atmospheric deposition of neutral salts would have relatively little impact on surface water
acidity. In the United States, the sea-salt effect has been documented in coastal streams of
Maine (Health et al., 1992; Kahl et al., 1992); in Europe, there have been numerous examples of
streamwater acidic episodes caused by the sea-salt content of particular rainfall events (Harriman
and Wells, 1985; Langan, 1985, 1987; Howells and Brown, 1987).
16
-------
2.3.3 Role of Acidic Deposition
Acidic deposition can contribute to episodic acidification by (1) providing direct inputs of acidic
waters to surface waters, (2) conditioning watersheds, via the accumulation of sulfate and nitrate
anions and H+ from atmospheric deposition in the upper layers of watershed soils during rela-
tively dry periods, and (3) lowering the chronic ANC of some systems and thereby lowering the
minimum ANC values attained during episodes.
A
Several mechanisms exist through which acidic waters from atmospheric deposition can be
delivered to streams and lakes without significant interaction or neutralization within the
watershed. For example, the snow and ice cover on a lake melts directly into the lake, with no
watershed contact. Snowmelt and rain on snow may travel over frozen surfaces to a receiving
stream or lake. Precipitation (or throughfall) may fail directly into a water body, or travel through
macropores in the soil with minimal soil contact. Under certain circumstances, all of these
flowpaths can be important, resulting in a more or less direct transfer of acidic deposition into
episodic acidification.
The second way in which acidic deposition can influence episodes is by conditioning watersheds,
much like the natural processes of nitrification and organic acid production described in Section
2.3.2. Sulfate and nitrate anions and the associated H+ ions delivered to a watershed by wet and
dry deposition may be stored in upper soil layers during'dry periods and flushed from the system
during especially large hydrological events (Lynch and Corbett, 1989; Turner et al., 1990). By
increasing NH4+ pools in watershed soils, atmospheric deposition of NH4+ may also contribute to
nitrification (Aber et al., 1989).
Finally, in some watersheds with low pre-episode ANC levels (< 25 ^eq/L), chronically high
sulfate concentrations appear to create conditions that allow other processes to create acidic,
high-aluminum conditions with relatively small ANC depressions during episodes (Galloway et al.,
1987). Nitrate pulses, organic acid pulses, and the sea-salt effect are examples of processes that
work in conjunction with high chronic sulfate concentrations to create acidic episodes.
Both natural and anthropogenic processes, therefore, may cause episodic acidification. In some
areas of the United States, especially in the Northeast and Mid-Atlantic regions, sufficient evi-
dence exists to conclude that acidic deposition has significantly increased the severity (minimum
ANC and pH; maximum aluminum) of episodes in some waters (Wigington et al., 1990). The rela-
17
-------
tive importance of acidic deposition and other factors varies among regions, among surface
waters, and even among events within a given lake or stream.
Episodes induced by organic acids have been documented or implicated in a number of studies
in the United States, for example, Raven Fork, North Carolina (Jones et al., 1983), Camp Creek,
Washington (Lefohn and Klock, 1985), several small streams near the coast of Maine (Haines,
1987; Haines et al., 1990; Kahl et al., 1992), and wetland and upland streams in the Bickford
Watershed, Massachusetts (McAvoy, 1989), as well as in some catchments in Canada and
Europe (Borg, 1986; Kerekes et al., 1986a,b; Seip et al., 1979). In other instances, organic acids
have been discounted as major contributors to episodes, for example, in Pancake Creek in the
Adirondacks, New York (Driscoll et al., 1987a,c), and in the inlets and outlets to Harp Lake,
Ontario (Servos and Mackie, 1986; LaZerte and Dillon, 1984).
In the United States, nitrate seems to have the greatest influence on episodes in the Northeast,
where it is the anion most consistently associated with episodes, especially during the winter and
spring. For example, Schaefer et al. (1990) concluded that nitrate was an important contributor to
episodic ANC depressions in Adirondack lakes with relatively low pre-episode ANC (< 25 ^eq/L).
However, the relative contributions of acidic deposition and natural nitrification to these nitrate
pulses are somewhat uncertain. Galloway et al. (1987) calculated that the mass of nitrate accu-
mulated in snowpack was sufficient to account for all of the nitrate transported into two of their
three study lakes in the Adirondacks. On the other hand, studies by Peters and Driscoll (1987)
and Rascher et al. (1987) suggest that in the western part of the Adirondacks, nitrate is generated
primarily by nitrification within the forest floor.
Aber et al. (1989) proposed a unified theory that accounts for the relative influence of nitrogen
pools within forest systems and nitrogen deposited by atmospheric deposition. They defined
nitrogen saturation as a situation in which the inputs of atmospheric nitrogen exceed the demand
for this element by watershed plants. In addition, Aber et al. (1989) hypothesized that nitrate can
accumulate in forest soils in areas receiving elevated levels of nitrogen inputs from the atmos-
phere that exceed the nitrogen requirements of forests. Atmospheric inputs of nitrogen add to
forest nitrogen pools, including the forest floor. In areas receiving large inputs of nitrogen via
atmospheric deposition, nitrate liberated by nitrification of organic nitrogen stored in the forest
floor, then, is a product both of background nitrogen in the forest and nitrogen introduced by
atmospheric deposition.
18
-------
In watersheds with'elevated nitrogen deposition regimes, the transition of forests from a condition
in which virtually all nitrogen received from atmospheric deposition is. retained by the forest to a
condition in which stream waters draining the forest have high nitrate concentration year round
(nitrogen saturation) can be considered as a series of stages (Stoddard, in press; Kahl et al.,
1993). According to the hypothesis of nitrogen saturation stages, increased nitrate export during
episodes is an indicator of an intermediate stage of nitrogen saturation.
Sulfate pulses have been identified as an important contributor to ANC depressions during epi-
sodes in many of the streams studied in Pennsylvania, the Southeast, and the Upper Midwest
(Jones et al., 1983; Schnoor et al., 1984; Lynch and Corbett, 1989; Barker and Witt, 1990; Elwood
et al., 1991). In some instances, in the Southern Blue Ridge Province and the Upper Midwest,
these sulfate pulses can be linked, at least partially, to internal mineral sources of sulfate in the
watershed (Elwood et al., 1991; Schnoor et al., 1984). Studies in Pennsylvania, oh the other
hand, suggest that atmospheric deposition is the primary source of sulfate, stored in the water-
shed during relatively dry periods and then exported during wet periods ( Barker and Witt, 1990;
Lynch and Corbett, 1989).
The importance of various causes of episodic acidification also seems to vary depending on the
pre-episode ANC level in the stream or lake. In the Adirondacks, Schaefer et al. (1990) found that
decreases in base cations (i.e., dilution) were very important contributors to episodic ANC depres-
sions in lakes with large initial ANC levels. In lakes with low pre-episode ANC (< 25 /teq/L),
however, nitrate increases were more important than base cation decreases. Other studies in the
United States and Canada have confirmed this pattern (Kennedy et al., 1989; Molot et al., 1989).
More research is needed in ail regions of the United States and elsewhere to better quantify the
effects of acidic deposition and natural processes on episodic acidification. In addition, our cur-
rent understanding of how flowpaths change during episodes is largely qualitative. These uncer-
tainties imply major knowledge gaps that limit our ability to quantitatively project episodic acidi-
fication and the importance of acidic deposition. This report is a step in addressing these gaps.
2.4 BIOLOGICAL EFFECTS
In general, we assume that acidic episodes are biologically significant. But definitive, lasting
effects on biological communities resulting specifically from episodic acidification, as opposed to
chronic acidification, have been demonstrated in relatively few systems (J. Baker et al., 1990a).
19
-------
Many uncertainties remain concerning the biological significance of episodic acidification and the
characteristics of episodes that are most important in determining biological effects.
2.4.1 Evidence for Effects
There is no doubt that fish and other organisms can experience significant mortality when
exposed to adverse chemical conditions for even short time periods. Numerous laboratory
bioassays have demonstrated that, in waters with sufficiently low pH and/or elevated aluminum,
mortalities can occur within 1 to 5 days (Schofield and Trojnar, 1980; Baker, 1981; France and
Stokes, 1987; Holtze and Hutchinson, 1989; Schweinforth et al., 1989). In addition, increased
mortality of fish and benthic invertebrates exposed in situ to episodic acidification has been
observed in several field studies (Hultberg, 1977; Sharpe et al., 1983; Andersson and Nyberg,
1984; Harvey and Whelpdale, 1986; Schofield et al., 1986; Lacroix and Townsend, 1987; Gunn,
1989). In general, the more severe the conditions, the more rapid the toxic response. The more
acid-sensitive the organism, the more likely it is to be significantly affected during short-term
exposures.
Although it is clear from both laboratory and field bioassays that short-term exposures to low pH
and elevated aluminum can increase mortality, examples of studies that definitively link episodic
acidification to population-level effects in the field are relatively rare. This may reflect in part
(1) the difficulty in distinguishing between effects from short-term and long-term acidification,
(2) the complexity of population responses to acidification, and (3) the need to account for the
coincident physical stresses on fish during periods of high discharge, which can have direct
adverse effects and also increase fish sensitivity to adverse chemical conditions. Evidence for the
importance of episodes includes the following:
• Records of fish kills during acidic episodes, although limited primarily to fish kills of
Atlantic salmon and seatrout (sea running brown trout) in Norwegian rivers (e.g.,
Jensen and Snekvik, 1972; Leivestad et al., 1976; Hesthagen, 1989) and fish kills in
hatcheries (Schofield and Trojnar, 1980; Jones et al., 1983).
• A few observations of changes in biological communities through time that apparently
resulted from episodic acidification, specifically long-term declines in Atlantic salmon
and seatrout populations associated with the fish kills in Norwegian rivers noted above
(Leivestad et al., 1976) and the observed loss of acid-sensitive benthic invertebrates in
stream systems in Sweden (Engblom and Lingdell, 1984) and Ontario (Hall and Ide,
1987) that now experience episodic acidification.
20
-------
Whole-system experiments demonstrating the response of benthic invertebrate commu-
nities (increased drift and loss of acid-sensitive species) and periphyton (increased
periphyton biomass) to short-term pulses of acid and aluminum (Planas and Moreau,
1986; Ormerod et al., 1987; Weatherley et al., 1988; Hall, 1990).
Comprehensive studies combining bioassay data on species sensitivity with information
on life history dynamics and variations in surface water chemistry to estimate the
potential importance of episodes.
- Gunn (1987, 1989) and coworkers evaluated the effects of episodes on lake trout
populations in Ontario lakes and concluded that episodic acidification during
snowmelt did not appear to be a major factor responsible for the observed long-
term decline of some populations.
- France and Stokes (1987), on the other hand, concluded that observed pH depres-
sions in Plastic Lake, Ontario, were likely to cause significant mortality of the
amphipod Hyalella azteca, severe enough to cause a long-term reduction in popula-
tion abundance.
- Lacroix and coworkers (Lacroix, 1985, 1989a,b; Lacroix et al., 1985; Lacroix and
Townsend, 1987) conducted detailed studies of Atlantic salmon survival and growth
in Nova Scotia rivers. Loss rates of Atlantic salmon parr (age 0+ and age 1 +)
increased during periods of minimum pH associated with fall rains and increased
water flow. Mortality rates during in situ bioassays also increased following pH
declines associated with fall rainstorm events. Based on these results, Lacroix
(1989b) concluded that pH decreases below 4.6-4.7 for periods longer than 20
days, or several days at pH 4.4-4.6, severely reduce parr densities and can
completely eliminate year-classes of both age 0+ and age 1 + parr in Nova Scotia
rivers.
2.4.2 Episode Characteristics That Influence Biological Response
We would expect the following factors to influence the effects of episodic acidification on aquatic
biota: duration, magnitude, frequency, rate of change, chemical composition, timing, and spatial
extent and distribution. Unfortunately, the data available to evaluate and quantify these
relationships are quite limited, highly diverse, and difficult to integrate.
Several investigators have concluded that episode duration and magnitude are probably the most
important factors controlling the toxic effects of episodes; for example:
• Curtis et al. (1989) exposed brook trout embryos continuously for 90 days (through the
feeding fry stage) to low pH and to four variations of intermittent cycles. The primary
determinant of fish survival was the peak H+ concentration (minimum pH). For inter-
mittent exposures using the same minimum pH, mortality correlated principally with
exposure duration, rather than with the frequency of the pulse or the length of the
recovery period.
21
-------
• Gagen and Sharpe (1987a) observed that significant mortality of yearling brook trout in
Linn Run, Pennsylvania, occurred only when total dissolved aluminum exceeded 200
figIL for at least one day (Figure 2-4). Lower concentrations or shorter durations had
minimal effect on fish survival.
• Schweinforth et al. (1989) found that survival of smallmouth bass and rainbow trout fry
exposed to 4-hour, 12-hour, and 36-hour pulses of acid and aluminum was closely
correlated with measures of both exposure magnitude and duration. In general,
measures of minimum pH and maximum aluminum were better predictors of survival
than were integrated expressions of dose (concentration over time). In all cases, fish
survival was distinctly better after brief 4-hour pulses than after 12- or 36-hour
exposures.
Although the results vary among species, life stages, and test conditions, increases in exposure
duration, particularly in the range of 0.5 to 5 days, and in exposure magnitude clearly affect the
severity of biological response. Acidic episodes with ANC < 0 fieq/L and pH < 5.0 (and associ-
ated levels of inorganic aluminum) lasting 2-5 days have been shown to adversely affect a diver-
sity of aquatic biota. Episodes of moderate severity (pH < 5.5 for 1-2 days) may cause signifi-
cant effects on organisms that are relatively acid sensitive, including many species of minnows
and mayflies (J. Baker et al., 1990a).
7
6
f 5
CO
rt) .
0. 4
**—
o
5 3
Q
1
0
No
Mortality
Percent
Mortality
O
100
iooO
0
100 , 200 300 400
Median Al Concentration of Peak (ug/L)
500
600
Figure 2-4. Brook trout mortality in Linn Run (Pennsylvania) as a function of duration and
median total dissolved Al concentration of coincident peaks in Al
concentration. The hand-fitted curve separates data into regions of Al
concentration and duration where mortality is and is not expected (source:
Gagen and Sharpe, 1987a).
22
-------
The three most important chemical parameters affecting fish survival in acidified waters are pH
(the H+ ion), inorganic aluminum, and calcium (Schofield and Trojnar, 1980; Brown, 1982a,b,
1983; Baker and Schofield, 1982; Chester, 1984; Ingersoll, 1986; J. Baker et al., 1990a). During
episodes, concentrations of these three ions often do not change in synchrony. For example,
although pH and inorganic aluminum levels generally are highly correlated (Driscoll et al., 1980),
aluminum levels for a given pH have been observed to be higher during the early stages of spring
melt than later in the melt and higher during midwinter thaws than during the major spring melt
(Hooper and Shoemaker, 1985; Hendershot et al., 1986; see Section 2.1). Base cations, calcium
included, often experience an initial brief peak in concentration during events, followed by
substantial dilution and a decline in concentration (Johannessen et al., 1980). These shifts in the
relative concentrations of H+, aluminum, and calcium through time and among different events
may further complicate attempts to predict episode toxicity.
J. Baker et al. (1990a) present an integrated index, termed the acidic stress index, that expresses
the combined effects of pH, inorganic aluminum, and calcium on fish mortality. Although they
were developed using chronic, constant-exposure bioassays, these functions also may be useful
for interpreting fish responses to episodes. J. Baker et al. (1990a) developed separate indices, for
several different fish species of varying acid sensitivity. All are based on laboratory bioassays
with fish larvae soon after swim-up and the initiation of feeding. The model for brook trout fry, 'for
example, is as follows:
ASI =
100
1 + exp[-23.49 + 5.35*pH + (2.97 x 10'3)*Ca - (1.93 x 10'3)*AI]
where ASI is the acidic stress index, equivalent to the expected percent mortality of brook trout fry
when exposed to constant levels of pH, inorganic aluminum, and calcium for 21 days. Calcium
(Ca) is expressed in fieq/L and inorganic aluminum (Al) m/ig/L.
The sensitivity of fish and other organisms to acidic conditions varies with age and among life
stages. Generally, early life stages (eggs and larvae/fry) tend to be more sensitive than older
individuals (Kwain, 1975; Baker and Schofield, 1985). In many cases, these highly sensitive life
stages occur only at certain times of the year and in specific locales. Brook trout, for example,
show a strong preference for spawning in areas of groundwater upwelling with fairly high pH
(Johnson and Webster, 1977). Brook trout eggs and yolk-sac fry may be reared, therefore, in an
environment with pH substantially higher than in the overlying water column. Only after the fry
23
-------
emerge from the spawning bed, which occurs during spring or slightly before snowmelt, may
brook trout early life stages be exposed to acidic waters and episodes. Thus, the timing and
spatial distribution of episodes relative to the occurrence of sensitive life stages may be critical to
population success.
Some organisms (e.g., adult fish) are able to detect and avoid acidic waters. Avoidance reactions
to potentially lethal waters with low pH and/or elevated aluminum have been demonstrated experi-
mentally in the laboratory (van Coillie et al., 1983; France, 1985; Gunn and Noakes, 1986; Redder
and Maly, 1986; Peterson et af., 1989) and observed in some field situations. For example, Muniz
and Leivestad (1980) observed behavior of brown trout in a single tributary stream with higher pH
during snowmelt in the River Gjov, Norway. Fish confined in cages in the main river died within
one week. During the reacidiflcation of Cranberry Lake in the Adirondacks, New York, following
an earlier treatment of lime, large numbers of the brook trout that had been stocked in the lake
were collected in the lake outlet emigrating from the lake as the pH dropped from 6.5 to 5.0
during a period of increased discharge (Gloss et al., 1989). More than 50% of the fish in the lake
were lost due to emigration. Finally, several field studies have demonstrated that acid-sensitive
stream benthic invertebrates (e.g., the mayfly Baetis rhodani) often respond to acid or aluminum
stress with dramatic increases in downstream drift (Hall et al., 1980, 1985; Bernard, 1985;
Raddum and FJellheim, 1984; Ormerod et al., 1987; Weatherley et al., 1988; Hopkins et al., 1989).
At least in the initial stages of response, this increase in drift probably reflects behavioral
avoidance by organisms attempting to escape adverse chemical conditions.
Thus, the degree to which episodes affect fish and other biota may be largely a function of (1) the
mobility of the organism, (2) the availability and accessibility of alternate habitats or refuge areas
that are less acidic and otherwise suitable for survival, and (3) the ability of biota to recolonize the
area after the episode without long-term adverse effects on the fish population or biological com-
munity. In streams, refugia may include tributaries with higher pH, areas further downstream
where pH levels are generally higher, or zones of groundwater inflow.
Although theoretically sound, most statements regarding the biological importance of the timing
and spatiaf extent of episodic acidification have not been specifically tested and proven in the
field. In addition, comprehensive studies of the effects of episodic acidification on fish
communities in streams are lacking. This report begins to address these uncertainties.
24
-------
SECTION 3
STUDY AREAS
ERP field research took place in three regions, the Adirondack Mountains of New York, the
Catskill Mountains of New York, and the Northern Appalachian Plateau of Pennsylvania. In each
region, we conducted intensive studies of the chemistry and biological effects of episodes at 4-5
sites (Figure 3-1). Study streams were selected based on the following criteria:
• Range of baseflow chemistry, but focusing on streams likely to be adversely affected
by episodic acidification (i.e., those with baseflow ANC between 0 and
• Range of expected episode severity, based on existing data or preliminary water
chemistry samples collected during site selection.
• Physical suitability for fish survival and reproduction, and similar size and quality fish
habitat.
• Indigenous fish populations, of the species of interest, present in at least some part of
the stream system at some time of the year.
• At least one reference stream per region, with similar physical and baseflow chemical
characteristics as the other streams but relatively stable pH during events and no
anticipated adverse effects on fish populations in the stream.
Because of the intensity of field activities, logistical considerations also influenced site selection.
These factors included ease and availability of access, ease of installing chemical and hydrologi-
cal monitoring equipment, ease of fish sampling and monitoring, proximity to central research
facility or field station, and proximity of sites to each other, to minimize travel times between sites.
The sites selected for the ERP are considered generally representative of other streams in regions
with baseflow ANC between 0 and 100 ^eq/L. Brook trout occurred in all 13 ERP streams. Brief
descriptions of these regions and the physical characteristics of the ERP study streams and
associated watersheds are provided in the following subsections.
3.1 CHARACTERISTICS OF STUDY SITES
3.1.1 Adirondack Mountains
The Adirondack Mountains form a large (24,000 km2) forested upland and mountainous region in
northern New York (Driscoll et al., 1991). Geologically, the Adirondack Mountains are among the
oldest landforms in eastern North America and are similar in origin to the Laurentian Shield of
Canada. The bedrock is predominantly granitic gneisses and metasedimentary rocks. Surficial
25
-------
79
44° -
42° -
40° -
EXPLANATION
O EPISODIC RESPONSE
STUDY SITE
/ADIRONDACK |
/ MOUNTAIN /
/ REGION /
0
CATSKILL
MOUNTAIN
REGION
\ / MASS
;
PENNSYLVANIA
h
50 100 MILES
0 50 100 150 KILOMETERS
Figure 3-1. Location of the three Episodic Response Project regions (adapted from Ranalli
et al., 1993).
26
-------
deposits in the region are the result of glacial activity. The soils, developed from this glacial flllf
are typically acidic Spodisols. The thickness of the glacial till and soil layers generally decreases
with increasing elevation.
The four ERP study streams are located in the southwestern highlands area of the Adirondacks
(Figure 3-2), where elevations range from 400 to 800 m. All the streams are in the Oswegathchie-
Black watershed, within 10 km of Eagle Bay, New York. The southwestern Adirondacks and
Oswegathchie-Black watershed, in particular, have the, highest proportion of low pH (< 5.0) and
fishless lakes in the Adirondacks (Kretser et al., 1989). In addition, numerous prior studies have
documented the adverse effects of chronic acidification on fish communities in the area (Driscoll
et al., 1980; Schofield and Driscoll, 1987; J. Baker et al., 1990b).
The study area is surrounded almost entirely by state forest lands within Adirondack State Park
that have not been logged or burned since the late 1800s. The principal forest cover type is a
northern hardwood/conifer rnixture, including sugar maple (Acer saccharum), red maple (Acer
rubrum), yellow birch (Betula alleghaniensis), and American beech (Fagus grandifolia) interspersed
with red spruce (P/'cea rubens), balsam fir (Abies balsamea), eastern hemlock (Tsuga canadensis),
and eastern white pine (Pinus strobus).
The bedrock, surficial geology, and soil types in the ERP watersheds are typical of the Adirondack
region; differences among the four watersheds appear relatively minor (Kretser et al., in press).
Soils are predominantly well drained, loamy soils overlaying bedrock and sandy glacial till ranging
in depth from 20 to 66 cm. All of these soils are quite acidic in the upper organic-rich horizons
and low in available nutrients and base saturation. The Fly Pond Outlet and Bald Mountain Brook
watersheds contain the highest percentage of deep soils (Becket and Skerry series). Both water-
sheds contain deep soils to the southeast and thin, rocky, steep terrain to the north and west.
The Buck Creek catchment is characterized by steep terrain with numerous boulders, rock
ledges, and thin soils (Becket-Lyman series). The Seventh Lake Inlet watershed is a combination
of moderately sloping terrain with deep bouldery soils (24% Becket-Skerry) and bedrock outcrops.
Three streams—Buck Creek, Bald Mountain Brook, and Seventh Lake Inlet—are typical first or
second order Adirondack streams (Table 3-1). Minimum watershed elevations range from 560 m
to 570 m for the four streams; maximum elevations range from 710 m to 775 m. All of the
Adirondack study streams are relatively small and short (< 2.7 km), compared with the study
streams in the Catskills (3.7-8.4 km) and Pennsylvania (5.0-8.8 km).
27
-------
Adirondacks
ERP Study Streams
1.25
1.25 2.5
mi
2.5
2.5
5 km
Seventh Lake inlet
Eagle Bay
Fourth Lake
Bald Mmi
'' Brook
Figure 3-2. Location of Adirondack study streams.
28
-------
Table 3-1. Physical Characteristics of the Four ERP Study Streams in the Adirondack
Mountain Region, New York
Stream Gage:
Latitude (N)
Longitude (W)
Watershed
Area (km2)
Max. elevation (m)
Min. elevation (m)
Lake or pond present
Wetland present
Soil series
Stream
Order '
Length (km)
Gradient (m/km)
Fly Pond
Outlet
43°45'05"
74°54'34»
0.9
710
563
Yes
Yes
Lyman
Becket
1
0.8
9
Bald Mountain
Brook
43°45'03"
74°54'39"
1.8
715
570
No
Yes
Lyman
Becket
1
2.2
25
Buck Creek
43°44'39"
74°43'20"
3.1
775
560
No
Minor
Becket-Lyman
Becket-Skerry
2
2.1
50
Seventh Lake
Inlet
43°45'49"
74°42'11"
6.4
725
570
Yes
Yes
Lyman
Becket-Skerry
Becket-Lyman
Adams-Croghan
2
2.7
31
As its name implies, Fly Pond Outlet is the outlet of a small pond. It is the smallest of the ERP
streams, with a watershed area of 0.9 km2, mean stream width of 1.6 m, and mean stream depth
of 6 cm in the intensively studied reach. The stream gradient, 9 m/km, is also the lowest of the
four Adirondack ERP streams. Substrate in the study reach is primarily sand, gravel, and silt. Fly
Pond (2.4 ha), upstream of the study area, is the primary water source for Fly Pond Outlet.
Several wetland areas also occur in the Fly Pond Outlet watershed above the ERP study reach.
With circumneutral water chemistry, Fly Pond Outlet was the reference stream for ERP biological
studies in the Adirondacks.
Bald Mountain Brook, immediately adjacent to Fly Pond Outlet, is nearly twice as large (watershed
area 1.8 km2, mean width 3.0 m) and more than twice as steep (gradient 25 m/km) as Fly Pond
Outlet. Stream substrate consists primarily of gravel with large amounts of cobble and boulder.
The Bald Mountain Brook study reach, at 143 m long, is the shortest of the ERP study reaches. A
natural falls at the upper end of the study reach restricts fish movements further upstream.
Immediately below the study reach, Bald Mountain Brook and Fly Pond Outlet join, flowing from
29
-------
this point 1.2 km. downstream before draining into West Lake. Extensive wetland areas occur in,
the upper portions of the Bald, Mountain Brook watershed.
Buck Creek and Seventh Lake Inlet drain directly into Seventh Lake; the ERP chemistry study
sites were located too m and 700 m upstream from the lake In Buck Creek and Seventh Lake
Inlet, respectively. Seventh Lake Inlet is the largest of the Adirondack ERP streams, with a
drainage area of 6.4 km2 and mean width of 5 m in the study reach. Three small ponds and
numerous beaver ponds occur in the upper watershed. Gravel and boulder are the major sub-
strate types; the numerous boulders in the study reach provide excellent cover for fish. Buck
Creek drains an area of 3.1 km2. The mean width at the study reach was 3.1 m, with a relatively
steep gradient of 50 m/km. The pool/riffle ratio (53%) was the highest among the four Adirondack
study streams. Gravel is the major substrate type, with large amounts of cobble and boulder.
3.1.2 Catskill Mountains
The Catskill Mountains, which form the northeastern end of the Appalachian Plateau, are the
uplifted remnants of a massive Devonian delta that fed into a shallow inland sea to the west
(Stoddard and Murdoch, 1991). The high ridges are predominantly erosional remains, and the
physiography suggests that the relief is due largely to stream action. The bedrock consists of
sandstone and interspersed conglomerates (60%) and mudstone or siltstone (40%). Soils are
generally Entisols and Inceptisols.
Surface waters within the Catskills are almost entirely streams; ponds and lakes are rare,
especially in the upper watersheds. Precipitation, stream discharge, and stream chemistry have
been monitored in a number of Catskill streams since the early 1930s (Murdoch, 1988; Stoddard
and Murdoch, 1991; Murdoch and Stoddard, 1992). However, relatively few studies have been
conducted on the potential effects of acidification on stream biota. Although many streams in the
region exhibit ANC < 0 during high-flow events, few are chronically acidic (Murdoch, 1988).
Four first, second, or third order streams were selected for the ERP: the upper reaches of the
East Branch of the Neversink; Biscuit Brook and High Falls Brook, both of which are tributaries to
the West Branch of the Neversink; and Black Brook, a tributary to the Beaverkill (Ranalli et al.,
1993) (Figure 3-3). Black Brook and High Falls Brook are relatively well buffered. Both streams
were considered reference streams for the ERP, although High Falls Brook can experience large
ANC declines during major storms.
30
-------
Catskills
ERP Study Streams
mi
eposition Site
Valley
East Branch
Neversink River
Figure 3-3. Location of Catskill study streams.
31
-------
Low-order basins in the Catskills are generally characterized by steep gradients; thin, glaciated
soils with extensive bedrock outcrop; and bedrock or large cobbles in stream channels. Stream
gradients tend to be steeper than in comparable Adirondack Mountain streams, and Catskill
streams typically flow through fewer wetlands and ponds than Adirondack streams. In general,
the Catskill ERP study streams drained larger watersheds (3.7-23.1 km2), occurred at higher
elevations (591-1,280 m), and were substantially longer (3.7-8.4 km) than the Adirondack ERP
study streams (Tables 3-1 and 3-2). The East Branch of the Neversink was the largest stream
studied in the ERP.
Surficial deposits in the headwater valleys of the Catskills are a combination of continental-glacier-
derived deposits oriented in the direction of regional ice movement and reworked or secondarily
scoured alpine-glacier deposits oriented perpendicular to the valley. Rich (1934) reported thick till
and thinner ground moraine from repeated glaciations in most of the upper Beaverkill and upper
East Branch Neversink valleys and pockets of thick drift in the lower High Falls and Biscuit Brook
basins. Morainal loops, consisting of unstratified till deposited at the end of local valley glaciers,
occur midway up the Biscuit Brook basin and in the headwaters of the East Branch Neversink
and Beaverkill basins. Many glacial deposits in the headwaters are thin, and many streams are
incised to bedrock.
Surficial material in the Catskills consists primarily (at least 90%) of local rock and sediment. The
percentage of exotics, particularly carbonate-bearing sediments, varies locally and could be a sig-
nificant factor explaining chemical differences among streams. Way (1972) found calcite and
pyrite in the shale along Route 17 to the south of the study area and along Route 28 to the north-
east. However, no evidence of pyrite has been found in the ERP watersheds.
Soils in the Catskills are generally categorized in the Arnot-Oquaga-Lackawanna association,
which are excessively drained to well-drained soils mainly on steep slopes (Ranalli et al., 1993).
Soils in the ERP study area are varied, but are predominantly shallow boulder soils on steep
slopes and are conducive, therefore, to rapid precipitation runoff; they are also moderately to
extremely acidic. The Lackawanna soils in the East Branch of the Neversink watershed contain a
fragipan of very low permeability at depths of 45 cm to 90 cm. Runoff from these soils is rapid.
The Biscuit Brook and High Falls Brook watersheds contain mainly Arnot and Oquaga soils, well
drained and moderately permeable with no fragipan; these soils range in thickness from 35 cm to
60 cm. The Black Brook watershed contains Wellsboro soils on gentle slopes near the stream
channel; these are deep, bouldety loam soils that contain a fragipan below a depth of 50 cm.
32
-------
Table 3-2. Physical Characteristics of the Four ERP Study Streams in the Catskill
Mountain Region, New York.
Stream Gage:
Latitude (N)
Longitude (W)
Watershed
Area (km2)
Max. elevation (m)
Min. elevation (m)
Lake or pond present
Wetland present
Soil series
Stream
Order
Length (km)
Gradient (m/km)
Black
Brook
42°00'42"
74°36'13"
3.7
1,140
681
No
No
Wellsboro
1
3.7
45
High Falls
Brook
41°58'40"
74°31'21"
7.1
1,170
591
No
No
Arnot
Oquaga
2
4.9
60
Biscuit
Brook
•
41°59'43"
74°30'05"
9.8
1,120
628
No
No
Arnot
Oquaga
2
4.5
45
East Branch
Neversink
41°58'01"
74°26'54"
23.1
1,280
651
No
No
Lackawanna
3
8.4
26
Alluvial deposits are present in the flood plains of Black Brook and the East Branch Neversink
near the ERP chemistry monitoring sites.
All the ERP Catskill watersheds are covered largely by northern hardwood forest, including
American beech (Fagus grand/folia), sugar maple (Acersaccharum), and yellow birch (Betula
alleghaniensis) (Stoddard and Murdoch, 1991). Hemlock (Tsuga canadensis) also occur through-
out the region, primarily along stream banks. The upper watersheds became a state preserve in
1870, and many of the Catskill watersheds contain first-growth forests at elevations above 730 m.
3.1.3 Northern Appalachian Plateau in Pennsylvania
Most of the western portion of Pennsylvania falls within the Northern Appalachian Plateau, an area
of high hills and low mountains ranging in elevation from 180 m to more than 1,000 m. Forest
cover of oak and mixed northern hetrdwoods dominates the landscape in areas of high elevation
and high relief. More than 90% of the stream basins in the region are underlain by sandstones,
shales, and siltstones (L. Baker et al., 1990). Unlike the Adirondack and Catskill regions, the
Northern Appalachian Plateau areas in Pennsylvania were not glaciated during the late
Wisconsinan era.
33
-------
Five streams were studied for the ERP: Linn Run and Baldwin Creek in the Laurel Hill area of
southwestern Pennsylvania and Roberts Run, Stone Run, and Benner Run in the northcentral part
of the state (Figures 3-4 and 3-5). Several prior studies have evaluated episodic acidification and
the toxic effects of episodes on fish in streams in the Laurel Hill area (Sharpe et al., 1983, 1984;
Gagen and Sharpe, 1987a,b). Periodic fish kills of trout stocked in the spring have been reported
in Linn Run, on the western slope of Laurel Hill, since the early 1960s (Sharpe et al., 1984). Linn
Run State Park pumped well water into Linn Run during the periods April 7 to June 27, 1989, and
April 3 to June 22, 1990. The pumping, which was below the sampling station, was done to
maintain trout that were stocked 0.5 km below the ERP study reach on Linn Run during these
pdriods. Baldwin Creek and Benner Creek in southwestern and northcentral Pennsylvania,
respectively, served as reference streams for the ERP in the two areas.
All the ERP study streams in Pennsylvania are second order (DeWalle et al., 1993). Baldwin
Creek drains an area of 5.3 km2; watershed areas for the other four streams are similar, ranging
between 10.0 km2 and 11.6 km2 (Table 3-3). Watershed elevations range from maximum eleva-
tions of 701-893 m in the five watersheds to 439-582 m at the ERP stream gages. Stream gradi-
ents are moderate in the three northcentral streams (35-39 m/km), but steeper in the two Laurel
Hill streams (58 m/km in Linn Run and 112 m/km in Baldwin Creek). Except for Linn Run, the
stream channels appear stable, with stones ranging from 20 cm to 60 cm commonly covering the
bottom. Areas with sandy deposits and exposed bedrock also occur. The stream bed in Linn
Run consists of boulders, smaller rocks, and gravel, similar in size to the substrate in the other
study streams, but not well anchored and less stable.
Both the northcentral and southwestern study areas are underlain primarily by sandstone with
some shale bedrock. Dinicola (1982) provided additional detail on the geology of Laurel Hill, an
anticlinal mountain (oriented northeast-southwest) within the Allegheny Mountain system. Major
rock units exposed in the Laurel Hill study area include Allegheny group sandstones and shales,
Pottsville group sandstones, Mauch Chunk shale, Loyalhanna limestone, and Pocono sandstone.
Soils in the region are acidic, and are composed of residual material weathered from noncal-
careous bedrock of sandstone and shale (DeWalle et al., 1993). Soils on the upper part of the
study basins, which approach 30% slope, are sandy loam in texture and extremely rocky. Major
soil series in the Laurel Hill watersheds are Calvin, Gilpin, Dekalb, and Cavode. Hazleton and
Clymer soils occur in the three northcentral study watersheds; Cookport soils also occur in the
Roberts Run and Stone Run watersheds in northcentral Pennsylvania.
34
-------
North Central Pennsylvania
ERP Study Streams
Frenchville
• <-i
-Lecontes Mills
Figure 3-4. Location of three study streams in north central Pennsylvania.
35
-------
Southwest Pennsylvania
ERP Study Streams
2.5
2.5
mi
10 km
^Deposition Site
nn Ru
v
Figure 3-5. Location of two study streams in southwest Pennsylvania.
36
-------
Table 3-3. Physical Characteristics of the Five ERP Study Streams in the Northern
Appalachian Plateau, Pennsylvania
Stream Gage:
Latitude (N)
Longitude (W)
Watershed
Area (km2)
Max. elevation (m)
Min. elevation (m)
Lake or pond present
Wetland present
Soil series
Stream
Order
Length (km)
Gradient (m/km)
Benner
Run
40°56'04"
78°01 '22"
11.3
749
582
No
No
Hazleton
Clymer
2
5.4
36
Roberts
Run
41°10'12"
78°24'22"
10.7
733
509
No
Yes
Cookport
Hazleton
Clymer
2
8.1
39
Stone
Run
41°05'52"
78°26'48"
11.6
701
494
No
No
Cookport
Hazleton
Clymer
2
6.4
35
Baldwin
Creek
40°21 '05"
79°03'04"
5.3
832
439
No
No
Calvin
Gilpin
2
5.0
112
Linn
Run
40°08'40"
79°12'37"
10.0
893
573
No
No
Calvin
Dekalb
Cavode
2
8.8
58
The five basins are completely forested with large pole-sized to mature deciduous tree species
predominating (DeWalle et al., 1993). The stands on the study watersheds appear to have
regenerated from logging conducted early this century. Some timber cutting has been conducted
on several of the basins. Salvage logging of scattered trees killed by gypsy moth defoliation was
conducted in some of the upland areas of the Roberts Run catchment during the period of data
collection. Tree defoliation and mortality, caused by gypsy moths, was also observed on the
Stone Run watershed. However, no salvage logging was performed during the study periods. In
1973 and 1974, clearcuts were made in the uplands of Benner Run, which cover approximately
12% of the total watershed area. Mo detailed forest inventory exists for the study watersheds.
However, red oak (Quercus rubra), scarlet oak (Quercus coccinea), chestnut oak (Quercus
prinus), red maple (Acer rubrum), and black cherry (Prunus serotina) are common species of
deciduous trees. Eastern hemlock (Tsuga canadensis) are common in the riparian areas of the
streams. Eastern white pine (Pinus strobus) is scattered in the uplands. Extensive wetland areas
exist in the Roberts Run watershed.
37
-------
3.2 DEPOSITION
As discussed in Section 1.4, we were unable to perform analyses to evaluate the role of atmos-
pheric deposition characteristics on episodic acidification in the study streams. This section pro-
vides a brief summary of wet deposition measured at ERP deposition stations.
Precipitation sampling began in August 1988 at Biscuit Brook in the Catskills, in December 1988
at Moss Lake in the Adirondacks, and in February 1989 at Linn Run in Southwest Pennsylvania.
Periods of continuous sampling ended in January 1990 at Linn Run, in May 1990 at Biscuit Brook,
and in June 1990 at Moss Lake. Data from the Pennsylvania State University MAP3S site were
collected during the period January 1989 through June 1990. Figures 3-6 through 3-9 present
monthly ion wet deposition and precipitation water equivalent.
The central and southwest Pennsylvania deposition collection sites received the largest monthly
loadings of H+ and SO42" and had the greatest monthly variations of these ions (Figures 3-8 and
3-9) (Barchet, 1991). Distinct seasonal fluctuations in ion concentrations were clearly evident only
at Moss Lake. In general, SO42" concentrations were highest during the warmer months and low-
est during the winter months. At the other sites, large month-to-month changes in ion concentra-
tions masked seasonal variation. Sulfate and NO3" together accounted for more than 80% of the
anions at all sites. Sulfate contributed a higher percentage of the total anion concentrations
during summer than in winter.
Wet deposition, which occurs intermittently, is one of two atmospheric pathways for chemical
inputs to watersheds (Barchet, 1991). The other pathway, dry deposition, occurs continuously.
However, the ERP did not attempt to measure dry deposition. Research at NOAA dry-deposition
sites near some of the ERP sites suggests that, over an annual period, dry deposition of S (SO2
and SO42" aerosol) and N (HNO3 and NO3") species is a factor of 0.30 to 1.00 for S and 0.48 to
0.81 for N of the wet deposition rate (Barchet, 1991). As the distance from emission sources
increases, the relative contribution- of dry deposition to total deposition decreases. For the ERP,
the Pennsylvania sites would be expected to have higher proportions of dry deposition than the
Adirondack sites, with the Catskill sites intermediate.
38
-------
o
30
20-
10-
Precipitation
15
en
E
5H
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
D)
£
250-
S04= Deposition
M
\
u
f
'<
'<
r\
'<
'
/
R
J
>
n
-
'
n|
ff
v\
/
'
$
CS|
D)
E
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
150-
100-
a>
E
50-
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
10
7
O)
E
5-
Na-h Deposition
ft
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
50
E
25-
. Ca++ Deposition
m
E
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
H+ Deposition
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
500
250-
o>
E
N03— Deposition
i
fl
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
100
50-
NH4+ Deposition
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
100-
50-
K+ Deposition
RrilJlnn
-
n
/
'
J
nrlplHrf
R
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
5
2.5-
Mg+-f Deposition
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
Figure 3-6. Monthly precipitation water equivalent depth and precipitation-weighted
average ion deposition for the Moss Lake wet deposition collection site in the
Adirondacks (adapted from Barchet 1991). Two precipitation values shown
are for rain gage (left) and the greater of rain gage or wet deposition bucket
(right).
39
-------
15
I
£
10-
5-
0-
H+ Deposition
n
ffl
JH
n '
' /
.
r
x
/ r
'HP
HP
i
1
n
X
x
X
X
X
X
/
X
X
X
7
x
/
X ,
/\
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
(N
P
250-
0-
S04= Depos
nn
Ml/I
HH
R
(jffl;
:ition
''\
X y
X y
>ln
/ y
-
'
ffl
I
•-i
^
01
E
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
500
250-
N03- Deposition
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
150
HijUfpHmmMmrjmHHmHM i
100-
en
E
50-
Cl— Deposition
n
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
100
50-
NH4+ Deposition
R
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
^
€25-
o>
E
Nd-1- Deposition
r
rl
;
nlxlm
*<* 100-
4
O)
£ 50-
K+ Deposition
,
0,
t
t
t
P
'
^
^
n
n
H
iPl
x
^
X
/
'
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
30
20-
o>
E
10-
Ca++ Deposition
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
10-i
Mg++ Deposition
o>
£
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
Figure 3-7. Monthly precipitation water equivalent depth and precipitation-weighted
average ion deposition for the Biscuit Brook wet deposition collection site in
the Catskills (adapted from Barchet 1991). Two precipitation values shown
are for rain gage (left) and the greater of rain gage or wet deposition bucket
(right).
40
-------
20
20-
c
g 10-
Precipitation
_
r
||
!
!
^^
;
_
; t
•. •
"'$
q n
j
'ppl
rO
1
r
!
1
-V. 10-
cn
E
H+ Deposition
p.
Rpfln
_
'
/
'
'
P
Pit
flPI rf
fjflnHBflrt
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
cs
{
< 500-
D>
E
S04= Deposition
n
nnl/L
T
/
/
/
'
n
'
PI
/
' Rul
^Pl^PHnMH
^
x
x
/
*>»
<
4
O)
£
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
500-
250-
N03- Depo:
FIR
R
/
n
sition
X
^
7
'
n
jffld!
_
'
P
/
n
/
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
? 50-
E
\
O)
E 25-
0-
Cl- Deposition
r j
MR
R
,-,
/
/
/
/
\
\
1
I
ffl
'
x
^
x
H
/^
1
r
/
n
,
x
'
/
7
^
/
/
100-
?
E
•< 50-
o>
E
o-
NH4-I- Depos
n
'
nn^P
'
'
'
/
,
(
'
it
^
'
X
f
/
-
ion
1
-
n
/j
Hlxi
WnR
r
,
/
/
/
'{
np]
nriH
~
s
/
/
'
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
5-
Na-H Deposition
_
'
*
/
PI
7
/
/
/
'
R
/
^
(
*
'
'
<
n
Y
\
R
/
/
/
'
ra PI
N Y
m
PI
/
/
/
/
/
{
T
1
/
x
'
/
__.
'
'
/
/
/
'
/
s
'
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
75
50-
01
E
25-
K+ Deposition
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
75
50-
0>
E 25H
Ca++ Deposition
^,Hc
nri
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
10-
Mg++ Deposition
a
nJw
^ /
' lyjf J rll/llyl
p]R
:nffl^
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
Figure 3-8. Monthly precipitation water equivalent depth and precipitation-weighted
average ion deposition for the Pennsylvania State University wet deposition
collection site in central Pennsylvania (adapted from Barchet 1991). Two
precipitation values shown are for rain gage (left) and the greater of rain gage
or wet deposition bucket (right).
41
-------
30
30-
20-
E
0
10-
0-
Precfpitation
„!
fl
B
;|
n
1
0
k
7 20-
^
D>
E 10-
H+ Deposition
r-t
0H:
MM/
nHI/l .
J
0"
:
n PI
Ma
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
•X. 500-
D>
E
S04= Deposi
PI
. n
'R,
H
'H
Mon
11
lu
^ 500-
^
£ 250-
N03- Deposition
J
1
1
llmi
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
^ 100-
fc
o»
E 50-
Cl— Deposition
R
n
/
'R
1
R
-,
x Rrj
'Bfflpi
ASONdJFMAMJJASONDJFMAMJJ
1988 1989 1990
7* 100-
O)
E 50-
NH4+ Deposition
p
J
I!
'/
fci
fVj
X. 5-
D)
E
Na4- Dep
PI
osi-
n
3
ic
/
^
<
'
>n
"
'
n
H
3
'
^
1
en
E
ASONDJFMAMJJASONDJFMAMJJ
1986 1989 1990
100
50-
K+ Deposition
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
I 25-
cn
0-
Cd++ Depo.
1
sit
p
ic
7
>n
B
/
1
7 5-
D)
E 2.5-
0-
Mg++ De
n
pos
D;
iti
~\n
/
Dn
k
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
ASONDJFMAMJJASONDJFMAMJJ
1988 1989 1990
Figure 3-9. Monthly precipitation water equivalent depth and precipitation-weighted
average ion deposition for the Linn Run wet deposition collection site in
southwest Pennsylvania (adapted from Barchet 1991). Two precipitation
values shown are for rain gage (left) and the greater of rain gage or wet
deposition bucket (right).
-------
SECTION 4
METHODS
This section describes the field, analytical, and data analysis methods used to collect the ERP
data. The introductory paragraphs provide an overview of the (QA/QC) program. The sections
that follow briefly summarize the methods for each major component of the ERP: atmospheric
deposition (Section 4.1), stream discharge (Section 4.2), stream chemistry (Section 4.3),
biological effects (Section 4.4), and database management (Section 4.5). Further information on
the ERP methods and QA/QC procedures are provided in the Methods Reports prepared by each
regional cooperator (Kretser et al., in press; Ranalli et al., 1993; DeWalle et al., 1993), and in the
ERP quality assurance project plan (Peck et al., 1988).
The primary purpose of the QA/QC program, an integral part of the ERP, was to maximize the
probability that data collected during the project met or exceeded data quality objectives, and
thus would allow scientifically sound interpretations of the data. Peck et al. (1988) established
objectives for seven data quality indicators (representativeness, completeness, comparability,
precision, bias, detection limits, and tolerable background levels). Procedures were established
to (1) ensure that collection and measurement techniques were standardized among all partici-
pants, (2) monitor performance of the various measurement systems being used in the ERP to
maintain statistical control and to provide rapid feedback so that corrective measures could be
taken before data quality was compromised, (3) allow for periodic assessment of the performance
of measurement systems and their components, and (4) verify and validate that reported data are
sufficiently representative, unbiased, and precise so as to be suitable for their intended use.
Appendix B summarizes the performance of ERP water quality laboratories. Additional QA/QC
information is included subsections that follow.
4.1 ATMOSPHERIC DEPOSITION
4.1.1 Wet Deposition
This section briefly summarizes procedures used to collect wet deposition samples. Section 3.2
provides an overview of wet deposition characteristics during the study period. Additional details
on deposition monitoring are described in Barchet (1991). "
43
-------
For each of the three study regions, a precipitation sampling site was selected according to stan-
dard criteria for deposition sites (Barchet, 1991). The sites, selected in August 1988, were located
in clearings within deciduous or mixed deciduous-coniferous forests. Site installations, coordin-
ated and supervised by each watershed research group, were completed in early 1989. Most
sites operated through June 1990, except for the Linn Run site, which ceased operation at the
end of January 1990. Field operation procedures were the same as those followed at sites in the
Multistate Atmospheric Power Production Pollution Study (MAP3S) Precipitation Chemistry
Network (Dana, 1982).
Figure 4-1 illustrates the locations of the three wet deposition monitoring sites established
specifically for the ERP in relation to other ERP rainfall measurement and stream chemistry moni-
toring activities being carried out in the vicinity of those sites. A wet deposition monitoring site
that was part of the U.S. Department of Energy's (DOE) MAP3S Precipitation Chemistry Network
was used to provide wet deposition data for the central Pennsylvania streams in the Northern
Appalachian Plateau study region [designated Pennsylvania State University (PSU) site]. The
Moss Lake site in the Adirondacks was located about 6 km southeast of the Big Moose monitor-
ing site in the Electric Power Research Institute's Operational Evaluation Network (OEN). The
Biscuit Brook site in the Catskill Mountains study area was co-located along with an NADP site,
which operates on a weekly sampling protocol. Biscuit Brook is 81 km north-northwest of West
Point in the Hudson River valley, which is the only other NADP monitoring site within 100 km.
Linn Run is 5 km east-northeast of the Linn Run ERP watershed, 20 km south-southwest of the
Baldwin Creek ERP watershed, and 45 km north-northeast of a wet deposition monitoring site in
Laurel Hill State Park (elevation 616 m). The PSU site is 15 km north of the Leading Ridge NADP
site.
Each precipitation sampling site was equipped with an Aerochem Metrics automatic wet-only pre-
cipitation collector and a Belfort recording, weighing rain gauge. The collection bucket in the
precipitation sampler was exposed only when rain or snow were detected. Each site was visited
at least once per week (even during dry periods) to check on the operation of the collector and
rain gauge. A daily precipitation sampling protocol was adopted for this study. Samples were
sent monthly to DOE's Pacific Northwest Laboratory (PNL) for chemical analysis and archiving.
Precipitation samples received by PNL from the ERP sites were merged with other samples being
analyzed as part of the MAP3S Precipitation Chemistry Network. The analytical operations of
PNL's precipitation chemistry laboratory, including data entry and archiving, are described in
44
-------
N
Moss Loke
2 km
Bold Ml Brook o Fly Pond
7th Lake Inlet
0 •
7th Lake Inlet
Bold Mt Brook
Buck Creek
• Precipitation chemistry
* Recording rain gauge
o Stream chemistry
01
Roberts Run
o
o
Stone Run
Benner Run
o
Black Brook
20 km
Penn State
Biscuit Brook
if Biscuit Brook
High Falls
o
Baldwin Creek
o
E Br Neversink
o
2 km
Linn Run
0*
Linn Run
• Precipitation chemistry
• Recording rain gauge
o Stream chemistry
• Precipitation chemistry
• Recording rain gauge
o Stream chemistry
Figure 4-1. Locations of ERP wet deposition monitoring sites and other ERP rainfall measurement and stream chemistry sites
(from Barchet, 1991).
-------
detail in a series of standard operating procedures (SOPs) in Barchet (1991). Table 4-11 lists the
analysis methods employed. For each analytical procedure, measurements of blanks were used
to assess contaminants; QC standards were used to evaluate accuracy. Duplicate samples were
used to determine the precision of the analyses.
4.1.2 Snow Measurements
Measurements of snowpack water equivalence and chemistry were made as part of the ERP but
are not included in this report. During both winter/spring periods of the ERP, investigators in
each region conducted snow surveys to determine the water equivalent and chemical characteris-
tics of seasonal snowpacks. In the Adirondacks, snow core transects were established in the
Bald Mountain Brook and Seventh Lake Inlet watersheds. Snow cores were collected weekly
during 1989 and biweekly during 1990. In the Catskills, snowpack was sampled biweekly during
the period of accumulation and weekly during spring snowmelt in the Black Brook, Biscuit Brook,
and East Branch Neversink River watersheds. A similar procedure was followed to sample snow-
packs in each of the Pennsylvania watersheds. Additional details regarding snow measurements
are available in DeWalle et al. (1993), Kretser et al. (in press), and Ranalli et al. (1993).
4.2 STREAM DISCHARGE
To allow the continuous measurement of stream discharge, a stream gauging' station was located
on each of the study streams. In the Adirondacks and Catskills, stream stage was measured at
15-minute intervals with a pressure transducer connected to a Campbell™ CR-10 data logger.
The transducer was located inside a 2-in PVC pipe that was bolted to bedrock in the stream
channel as close as possible to a nonrecording staff gauge. A stage-discharge relationship was
established for each stream by measuring discharge over a wide range of stages and plotting the
discharge against the stage at which the discharge occurred. This relationship was used to
convert stage measurements into discharge. Investigators in Pennsylvania used strip chart, water
level recorders instead of pressure transducers to measure stream stage continuously. The
accuracy of automated stream stage measurements was checked routinely by comparing the
stage measurements of the pressure transducers or water level recorders to staff gauges located
nearby in the streams.
Since this section contains so many tables, we have placed them all at the end of the section, beginning on page 61.
46
-------
4.3 STREAM CHEMISTRY
This section describes the methods used to collect water samples (Section 4.3.1) and to perform
chemical analyses (Section 4.3.2). Section 4.3.3 presents analyses used to estimate inorganic
monomeric aluminum for water samples for which aluminum speciation was not performed.
Section 4.3.4 briefly summarizes in situ measurements made during the ERP but not included in
the report.
4.3.1 Water Sampling
A vital part of the ERP was collecting water samples often enough to characterize episodes that
occurred during the study period. Because of the uncertain occurrence of episodes, ERP investi-
gators used automated pumping samplers to collect samples during periods of high flow. Con-
trolled by a data logger, the Instrumentation Specialties Company (ISCO™) pump samplers were
programmed to collect samples at fixed time intervals (2-6 hours) or at specified changes of
stage level (e.g., 0.05 ft). Water samples generally were retrieved from the pump samplers within
24 hours of collection. During periods of time when problems occurred with the data loggers pr
samplers, grab samples were collected by field crews during episodes to the extent possible. In
addition, streamwater samples were collected manually on a weekly basis when field crews made
visits to service instrumentation. During radio telemetry studies of fish movement during epi-
sodes, additional grab samples were collected. Each day during the telemetry study, a sample
was collected at any location where a tagged fish was observed.
All water samples were collected in polyethylene bottles with unlined plastic caps. The samples
were packed in ice and transported to laboratories in coolers. In general, one field blank sample
was collected and submitted with each batch of samples to monitor container contamination.
4.3.2 Laboratory Methods
All the regional cooperators had their own laboratories for analysis of samples collected in their
regions. Sample handling procedures are summarized in Table 4-2. Methods and equipment
used in the analysis of water samples are summarized in Table 4-3. Most samples were analyzed
for the full suite of analytes presented in Table 4-3. However, selected samples associated with
radiotelemetry work (subsection 4.4.2.1) were analyzed only for pH, ANC, and total dissolved
aluminum. Selected episode samples were analyzed for total monomeric and organic monomeric
47
-------
aluminum in addition to the other measured variables. Aluminum speciation was based upon the
methods of Barnes (1975) and Driscoll (1984) and is described completely in Pagano (1990). In
the laboratory, blanks, check samples, and performance evaluation samples were routinely used.
Laboratory data verification was also routinely performed using cation/anion balances and
conductivity computation checks. If the measured cation/anion balance did not agree with the
theoretical balance within 10%, or if conductivity calculated from individual ion measurements did
not agree with measured conductivity within 15%, the sample was completely reanalyzed.
On a monthly basis, the ERP laboratories received performance evaluation samples from the EPA
Environmental Monitoring Systems Laboratory in Las Vegas (EMSL-LV). These samples con-
sisted of two natural and three synthetic samples and covered the ranges of ionic values
anticipated to be measured in the study. In addition, the ERP laboratories participated three
times a year in a round-robin audit program administered by the Long Range Transboundary Air
Pollution (LRTAP) Program. Water chemistry performance objectives and performance evaluation
sample results are presented in Appendix B.
4.3.3 Estimation of Inorganic Monomeric Aluminum
As noted, concentrations of inorganic monomeric Al (Aljm) were measured in only a subset of the
water samples collected. However, from a biological perspective, it is very important to include
this chemical parameter in the characterization of episodes and biological study periods. There-
fore, regression models were developed to estimate Al(m based on other measured chemical and
physical variables. Separate models were developed for each study stream. Separate models
were also evaluated for each season, for streams and seasons with sufficient data (Tables 4-4 and
4-5). Measurements of Alim occurred primarily during periods of intensive biological studies in the
spring and fall. Thus, season-specific models were considered only for fall 1988, spring 1989, fall
1989, and spring 1990.
Predictor variables considered included discharge, total dissolved Al, ANC, pH, DOC, sulfate,
nitrate, chloride, sum of base cations, and conductivity. The log-transformed and nontransformed
forms of each variable were included in maximum r2 analyses with both log- and nontransformed
Aljm, to identify sets of variables strongly associated with in-stream variations in Aljm. Based on
these results, four to ten alternative models were selected for testing, including models with one
or more interaction terms. For streams and seasons with sufficient data (n ^ 30), the data set
48
-------
was randomly divided into two-thirds for model calibration and one-third for model testing. Each
candidate model was calibrated using ordinary least squares regression, and observed and
predicted levels of Alim were then compared for both the calibration data set and the test data set.
The following criteria and statistics were used to select the final model:
• Lowest model mean square error, calculated using nontransformed Aljm data for
consistency among models.
• Lowest mean square error for the test data set (sum of the squared differences
between observed and predicted nontransformed Aljm divided by one less than the
number of observations).
• No statistically significant decrease in fit for the test data set compared to the
calibration data (based on the ratio of the mean square errors for the two data sets).
• All predictor variables in the model significant at p < 0.1, and preferably p < 0.05.
• Absence of significant autocorrelation, based on the Durbin-Watson D statistic (Neter
and Wasserman, 1974).
• Absence of any highly influential data points, based on visual examination of residual
plots and scatter plots comparing predicted and observed Alim.
• Model well behaved when applied to the full data set (i.e., no unreasonable outliers or
unusually high or low predicted values of Aljm), based on scatter plots of predicted Al|
versus pH and discharge.
The final model, used to estimate Alim, was calibrated to the full data set.
For seasons with enough data (n > 10), the above steps were repeated to select the best model
for estimating Aljm levels within that season. The need for a season-specific model, as opposed
to the full-data (overall) model, was assessed by:
• Comparing the mean square errors for the season-specific and full-data models,
calculated specifically for Aljm measurements during that season.
• T-test and signed rank tests to determine whether the full-data model estimates of Alim
for the season of interest were significantly biased.
If the season-specific model provided a significantly (p < 0.05) better fit than the full-data model,
or if the estimates from the full-data model were significantly biased, then the season-specific
model was used to estimate Alim levels for that season. For each stream, if any season-specific
model resulted in a significant improvement, then season-specific models were used for all
seasons with enough data to calibrate a season-specific model, as long as the season-specific
model did not perform significantly worse than the full-data model. For all other seasons and
49
-------
water samples, Alim estimates were based on the full-data model. For those water samples with
missing observations for one or more of the predictor variables, back-up models were used, that
is, the best possible model involving only those predictor variables with no missing observations.
Some Alim values (calculated as the difference between the measured total monomeric Al and
measured organic monomeric Al) were zero or negative. Thus, for models involving log-trans-
formed Al, a constant, 25 /tg/L* was added to alt inorganic Al observations before model calibra-
tion, and then subtracted from the model predictions.
For analyses of biological effects, the regression models just described were used to estimate
Alim only for those water samples where total and organic monomeric Al (required to calculate
AIIm) were not measured. However, episode characterization analyses (See Section 6) involved
only predicted Aljm values.
Tables 4-6 through 4-8 present the regression models that were used to estimate Aljm during time
periods for which direct Alim measurements were not made. When available, a season specific
model was employed. Otherwise, an overall model was used. If samples were missing one or
more predictor variables, a simpler, back-up model was used to estimate Alinv Text, tables, and
figures throughout the report identify whether directly measured Alim or estimated Aljm is being
reported.
4.3.4 In Situ Measurements
In situ measurements of stream pH, conductivity, water temperature, and air temperature were
made using U.S. Geological Survey (USGS) mini-monitors rented from the USGS Instrumentation
Facility in Mississippi. The mini-monitor data have not beqn used in primary data analyses of this
report. In some regions, such as the Adirondacks, regional investigators had good success with
the mini-monitor performance and were able to use the data to assist in interpreting biological
results. However, in other regions, the performance of the mini-monitors was more sporadic.
Each regional cooperator will use these data after a case-by-case evaluation of data quality.
4.4 BIOLOGICAL EFFECTS
Assessments of the biological effects of epispcjes in the study streams involved four major
activities:
50
-------
1. Conducting in situ bioassays to quantify the toxic effects of episodic acidification on
fish.
2. Tracking fish movements during episodes using radiotelemetry and fish traps.
3. Monitoring fish population density and biomass.
4. Supplementing the existing populations of fish in the study reaches with transplants
from other nearby stream reaches.
The methods employed for each of these research components are described in the following
subsections.
Brook trout (Salvelinus fontinalis) was selected as the ERP target fish species. Brook trout is the
sportfish most widely distributed in the small, low-ionic strength, headwater streams in the
Adirondacks, Catskills, and Pennsylvania that are most susceptible to episodic as well as chronic
acidification (Pfeiffer and Festa, 1980; Sharpe et al., 1987; Driscoll et al., 1991; Stoddard and
Murdoch, 1991). Although brook trout are relatively acid tolerant (Baker and Christensen, 1991),
this species is one of those mostly widely affected by acidification and acidic deposition, because
of its distribution in a particularly sensitive habitat (J. Baker et al., 1990a). Thus, the status of
brook trout populations and brook trout toxicity were selected as the primary indicators of effects
of episodes on fish.
In s/fu bioassays were also conducted with at least one forage fish species, in addition to brook
trout, in each stream. The forage species varied among regions and streams because of
differences in native fish fauna. Blacknose dace (Rhinichthys atratulus), a highly acid-sensitive fish
species, was used for bioassays in the Adirondacks; slimy sculpin (Cottus cognatus) were used in
the Catskills; and either slimy sculpin or mottled sculpin (Cottus bairdi) were used in
Pennsylvania, depending on the stream. All fish .species caught during stream surveys (see Sec-
tion 4.4.3) were recorded and included in the assessment of fish community status in the stream.
4.4.1 In Situ Fish Toxicity
In situ bioassays were conducted in the spring and fall, for a total of four test periods: fall 1988,
spring 1989, fall 1989, and spring 1990. Test periods were selected to coincide with the expected
timing of spring snowmelt and fall rainstorms. Experiments in the spring were often delayed,
however, by the difficulties caused by high flows and/or ice cover on the streams. The test
periods for each bioassay are presented in Tables 4-9 to 4-14.
51
-------
All fish used in bioassays were wild strains, collected by electrofishing (see Section 4.4.3) from
local streams with similar habitat. The objective was to test yearling brook trout in the spring,
young-of-the-year brook trout in the fall, and adult or yearling forage fish in both periods. How-
ever, fish availability to some degree limited the ages and sizes of fish tested. Thus, the size (and
presumably age) offish varied (see Tables 4-9 to 4-14). We assumed all fish tested (of a given
species) were equally sensitive; differences in fish size or age were not accounted for in statistical
analyses.
Most tests involved organisms selected from .a common pool; that is, fish of the appropriate
species and size were collected from one or more circumneutral streams, combined into a
common pool of fish for that region, and then randomly distributed from this pool into the
bioassay test chambers in each stream in the region. Common pool tests were used to compare
toxic responses among streams, since the groups of organisms tested in each stream presum-
ably were equally sensitive to acid-aluminum stress. In some streams and test periods, bioassays
were conducted with resident fish, collected from the population of fish residing in the stream
(and/or from other study streams that experience episodic acidification). Comparisons between
common pool and resident fish bioassays can be used to determine whether resident fish may
have adapted to conditions in the stream, and thus be more tolerant of acid-aluminum stress,
than the common pool fish collected from circumneutral streams.
Test fish were held in each stream in 4-L polyethylene jugs with at least two 7 x 12-cm openings,
covered with 1- or 2-mm mesh fiberglass screening, to ensure adequate water exchange (after
Johnson et al., 1987). The number offish per jug varied (generally between 4 and 7), depending
on the size of the fish. The total fish weight per container never exceeded 20 g wet mass in the
Adirondacks and Catskills and 35 g in the Pennsylvania streams, to avoid stress from crowding.
In addition, the largest fish in any bioassay container did not exceed the length of the shortest
fish by more than 50%, to minimize problems with cannibalism. Multiple jugs (between 3 and 6)
were used per test, so that at least 20 fish of each species were tested during most bioassays
(see Tables 4-9 to 4-14). The jugs were placed within 0.6-m x 0.5-m x 0.3-m cages constructed of
wood (5 cm x 5 cm) covered with 6.4- or 12-mm plastic screening. These cages were sub-
merged and anchored near the continuous water quality monitoring station in a pool area or
downstream of a large obstruction, such as a large rock, so that the test fish were not subjected
to high current velocities. The jugs further reduced the velocity of the currents to which fish were
exposed.
52
-------
The numbers of fish in each bioassay container that were alive, dead, or visibly stressed (i.e., that
had lost their equilibrium) were counted every 1-5 days. Generally, fish mortalities were checked
daily during periods of storm or snowmelt events when water quality conditions were changing
rapidly. Most tests lasted for a period of 20-40 days. Occasionally, tests were terminated after
< 20 days if units in acidified waters experienced high levels of mortality (> 50-80%) or if weather
conditions became unsuitable (e.g., in the late fall/early winter). Fish were not fed during the
experiment; therefore, with longer exposure times, fish may have been increasingly stressed by
starvation. For this reason, most analyses rely on the percent mortality observed after 20 days.
Percent mortalities are compared to the measured and estimated (see Section 4.3) water chemis-
try for the stream during the bioassay period, in particular, the magnitude and duration of epi-
sodes of low pH and elevated Aljrn levels. For these response surface analyses, fish mortalities
are expressed as the combined result from all jugs tested within a given stream, for a given
species, fish source, age group, and test period. More detailed analyses, for example, assessing
changes in mortality over time, are being conducted by the cooperating scientists in each region
and will be published in subsequent journal papers.
Continuous records of stream chemistry were required to estimate episode magnitude and dura-
tion and median chemistry during each bioassay period. Two alternative methods were consid-
ered: (1) interpolation between chemistry samples and (2) concentration-discharge relationships.
Low regression coefficients (r2 < 0.25) indicated poor correlations between Aljm concentrations
and discharge in some streams and seasons. For this reason, direct interpolation was selected
as the primary basis for estimating the continuous chemical record. Chemical concentrations at
any point in time were estimated from linear interpolation between measured values (both grab
and event samples). Periods with inadequate chemistry data, for interpolating with reasonable
confidence, were identified by comparing plots of interpolated chemistry over time to the con-
tinuous discharge record. Bioassay periods with inadequate coverage were deleted from subse-
quent analyses (n = 8, all of which were fall 1988 bioassays in the Catskills). The time-weighted
median value for a bioassay period was defined as the chemical concentration which half the time
was exceeded and half the time concentrations were below, based on the interpolated chemical
record.
53
-------
4.4.2 Fish Movement
Individual fish movements were tracked using radiotelemetry in each region during the fall and
spring (fall 1989 and spring 1989 and 1990 in all regions; plus fall 1988 in Pennsylvania streams).
In addition, fish traps were used in selected Adirondack streams to monitor upstream and down-
stream fish movements. Both types of information provide insight into changes in fish movement
patterns during episodes and the ability of fish to behaviorally avoid stressful chemical conditions.
4.4.2.1 Radiotelemetry
Miniature radio transmitters (1-3 g) were placed in 6-15 adult brook trout per stream per
experiment. The location of each fish was then determined daily or every other day using a
portable radio receiver. Fish movements were tracked for a period of about 30 days or longer, or
until the experiment was ended because of transmitter malfunction or regurgitation. Studies were
conducted in at least one reference stream and one nonreference stream in each region per
season (spring or fall). The timing and duration of each study, and streams and number and size
of brook trout involved, are summarized in Tables 4-15 to 4-17. Only brook trout movements
were evaluated, because forage fish (blacknose dace and sculpin) in the study streams were too
small to carry the radio transmitters without adverse effects on the fish.
Brook trout for use in the radiotelemetry studies were collected using electrofishing (see Section
4.4.3) from nearby streams of similar size and habitat. In the Adirondacks and Pennsylvania, all
fish collected were combined into a single common pool and randomly assigned to the study
streams for tracking. Studies in the Catskills in spring and fall 1989 involved both common pool
and resident fish (collected from the study stream). Brook trout weighed 30 g or more (averaging
45-70 g in most studies; see Tables 4-15 to 4-17), so that the transmitter weight never exceeded
5% of the fish's body weight.
Initially, radio transmitters were placed in the stomachs of the fish. Fish were anesthetized and
the transmitter gradually pushed down the esophagus. Because of problems with regurgitation,
In later seasons (fall 1989 and spring 1990), transmitters were also surgically implanted in some
fish. Fish were anesthetized and transmitters inserted into the abdominal cavity through a
10-15 mm mid-ventral incision, which was then closed with three sutures. Fish were held for at
least one day after ingestion and three days after surgical implantation before being released into
the study streams.
54
-------
Patterns of fish movement were assessed relative to chemical conditions at the ERP stream
monitor (see Section 4.3) and individual fish locations. When fish moved > 25 m over a 24-hour
period, water samples were collected by hand from both the initial and final fish location, and
analyzed for ANC, pH, and total dissolved Al.
4.4.2.2 Fish Traps
In some Adirondack streams, fish traps were used to monitor the timing and numbers offish
moving upstream and downstream in relation to stream chemistry. Three bi-directional fish traps
were installed in fall 1988, at the lower end of the study reaches in Fly Pond Outlet and Bald
Mountain Brook and 80 m below the confluence of Fly Pond Outlet and Bald Mountain Brook (see
Section 3). A fourth trap was installed in spring 1989 at the upstream end of the Fly Pond Outlet
study reach. A natural falls at the upper end of the Bald Mountain Brook study reach restricted
fish movement further upstream. All traps extended across the entire stream reach and presum-
ably caught all fish moving either upstream or downstream.
Fish traps were operated mid-October to 10 December 1988, 1 May to 1 December 1989, and-1
May to 30 June 1990. When in operation, fish traps were checked daily; all fish caught were
checked for fin clips, measured, weighed, and then released in the direction in which they were
moving, either upstream or downstream of the trap.
Fly Pond Outlet and Bald Mountain Brook were the smallest of the ERP study streams. Fish traps
were not constructed on other streams because of costs and potential operational problems
during periods of peak discharge.
4.4.3 Fish Community Surveys
Fish communities in each study stream were surveyed to quantify the density of brook trout and
the occurrence of other fish species. Sampling dates are presented in Table 4-18. Fish were
collected using portable backpack electroshockers (DC battery powered units in the Adirondacks
and battery and gasoline powered AC/DC units using DC current in the Catskills and Pennsyl-
vania). Study reaches, 100-300 m in length, near the ERP stream chemistry monitor were
isolated with blocking seines and sampled with at least three electrofishing passes. All fish
collected in each pass were identified to species and counted. Brook trout were individually
weighed and measured for total length. Forage fish" (blacknose dace and sculpin) were weighed
55
-------
individually or as a group. Brook trout and blacknose dace were also checked for identifying
marks; unmarked fish were fin clipped. After all electrofishing passes and fish processing were
completed, the fish caught were redistributed throughout the study reach.
Brook trout population density (number per unit of stream area) was estimated using the Zippin
removal method (Zippin, 1958; Everhart et al., 1975). Additional electrofishing passes (up to 6)
were conducted until the 95% upper confidence limit was < 20% of the brook trout population
estimate. Population estimates were also calculated for other fish species when sufficient
numbers of fish were caught. Electrofishing efficiency and the accuracy of fish population
estimates were evaluated by stocking and then surveying a known number of marked brook trout
in a separate 50-100-m stream section. Capture efficiencies ranged between 66% and 97%
(Table 4-19).
Additional, qualitative surveys of fish communities (< 3 electrofishing passes) were conducted at
other stream locations and times. In particular, longitudinal variations in fish community compo-
sition In Catskill and Pennsylvania study streams were evaluated by sampling at multiple (2-6)
locations in the mainstream as well as nearby tributary streams (in June 1989 in the Catskills and
May 1989 in Pennsylvania). Water chemistry samples were collected by grab sampling at each
site and analyzed for ANC, pH, conductivity, and total dissolved Al.
4.4.4 Fish Transplants
Some streams had distinctly lower fish biomass than others. To ensure that differences among
streams were not related to problems with fish access (e.g., natural fish barriers), additional brook
trout and forage fish (blacknose dace and sculpin) were transplanted into each study reach to
achieve an initial, comparable level of fish density per stream. Fish density target levels were
selected to be 50% (30%-80% in the Catskills) of densities observed in circumneutral streams in
the area of similar size and habitat. Brook trout targets were 2.0 g/m2, 0.9-2.4 g/m2, and 75
fish/0.1 ha in the Adirondacks, Catskills, and Pennsylvania, respectively. Table 4-20 shows
transplant dates and the number of fish stocked in each stream.
in most cases, resident fish populations in the study streams were supplemented with additional
fish, as needed, to raise fish density to the desired target. Fish for stocking were collected, using
electrofishing, from other similar nearby streams, combined into a single common pool, and then
distributed randomly among streams. In the Adirondacks, during one fish transplant (April 1989
56
-------
for brook trout arid September 1989 for blacknose dace) resident populations in the reference
stream (Fly Pond Outlet) were depleted, and the stream restocked with fish from other streams,
so that reference stream fish experienced a similar level of stocking and handling stress as fish in
nonreference streams. All stocked fish were fin clipped to distinguish common pool transplants
from resident fish. For this report, however, only changes in total fish biomass (by species,
resident and transplanted fish combined) are presented.
4.4.5 Other Biological Studies
Other biological research conducted as part of the ERP, but not reported on here, includes
special studies in some regions relating to fish reproductive success and qualitative surveys of
benthic macroinvertebrates. The ERP study design does not explicitly address the effects of
episodic acidification on fish reproductive success. However, supplemental studies were con-
ducted by some cooperators. In the Adirondacks, hatchery-reared brook trout feeding fry were
tested with in situ bioassays in spring 1989 and 1990. In Pennsylvania, brook trout eggs and sac
fry, from wild brook trout artificially spawned in the laboratory, were exposed in situ from October
1989 through March 1990. Spawning surveys were also conducted in Pennsylvania streams in
fall 1989 to identify the occurrence and location of brook trout redds (areas in the substrate where
adult brook trout have spawned and embryos have incubated). Redds were sampled for pH,
ANC, total dissolved Al, dissolved oxygen, and temperature during winter, and then excavated in
mid-February (before fry emergence) to evaluate embryo survival. Because each of these special
studies was conducted in only one region, neither the methods nor the results are described in
this report. Simonin et al. (1993) summarize the Adirondack brook trout fry bioassays; Fiss (1991)
presents results from the Pennsylvania egg and sac fry bioassays and spawning surveys. No
special studies relative to fish reproduction were conducted in Catskill streams.
Qualitative data were collected on the composition of the benthic macroinvertebrate community in
each ERP stream to provide general information on the suitability of fish food supplies. Each
stream was sampled at least three times; on each date, five samples were collected from riffle
areas with a 0.1-m2 Surber sampler. Organisms were counted and identified to family. Results
from these qualitative surveys are not summarized in this report, but are available within the ERP
database (see Section 4.5.4) and presented in various reports and papers by the regional cooper-
ators (DeWalle et al., 1991; Kretser et al., 1991; Murdoch et al., 1991).
57
-------
4.5 DATABASE MANAGEMENT
The ERP was designed to record a wide variety of data pertaining to episodic acidification in streams,
ranging from simple stage height measurements to complex chemical analyses and biological beha-
vior observations. These data were collected by three different research groups in three separate
regions. The database resulting from these activities had to accurately and efficiently store the project
data, and it had to be designed so that the data quality would be readily apparent during future
analyses. Data gathering methods had to be carefully coordinated to ensure that the resulting data
could be pulled together into a single database for analysis. This section describes the methods used
to manage ERP data during the project and those used to develop the final ERP database.
Databases designed to ease the chore of data entry are not necessarily the best databases for data
analysis. Therefore, the overall ERP database was built in two phases: a data entry phase and a
database construction phase. Data quality was assessed during and after each phase. Software was
developed by the Lockheed Engineering and Science Company (LEMSCO) of Las Vegas, Nevada, to
provide the tools for hand data entry of stream chemistry, biology, and research site characterization
data, to allow the regions to assess the quality of these data, and to apply data quality flags. The
USGS in Doraville, Georgia, also developed software to control USGS mini-monitors and data loggers
used in the collection of electronic time series stream chemistry and hydrology data. Batelle's Pacific
Northwest Laboratory (PNL) in Richland, Washington, handled wet deposition and precipitation data
(Barchet, 1991). ManTech Environmental Technology, Inc. (METI), of Corvallis, Oregon, combined
these data into the final database by integrating the files from the regional data entry efforts into a
consistent structure.
4.5.1 Data Entry
Data entry was handled separately in each ERP region, and by PNL. Each regional cooperator was
responsible for entering chemical, biological, and site characterization data into the regional ERP
database, and for performing data quality checks. Each region also transferred data-from the data
loggers to computer files and formatted these data into structures suitable for addition to the final
database. Regional files were submitted periodically, and at the end of the project, to METI data
management personnel for inclusion in the overall ERP database.
58
-------
4.5.2 Database Construction
Once data entry was complete, the data files from each region and from PNL were combined into a
consistent structure. Files were merged, and some fields specific to data entry were removed to
simplify the data structure. The ERP database is described in the Episodic Response Project
Database User's Guide (see Appendix A).
4.5.3 Data Quality Assurance
Database quality was maintained throughout the project through the use of internal database
consistency checks (within sample checks) and by comparing current versions of the database with
past versions of the database. During the data entry phase, each region was responsible for checking
data quality by using the tools provided in the data entry software package. Anion/cation balances
were checked, as well as theoretical versus measured conductivity. These types of checks served the
dual purpose of guarding against laboratory analysis problems as well as data entry problems, since
problems detected by the checks could be attributable to either source of error. Once the regional
databases were delivered for inclusion into the overall database, data management personnel became
responsible for database integrity. Backups of each version of the database were maintained at each
phase of data processing to provide a method to recover lost data or correct data processing errors.
The data validation step served as a final check for data entry and data processing errors.
Data validation consisted of identifying outliers within the database through examining relationships
among water chemistry variables. However, just because a datum is an outlier does not mean that it
is incorrect. Therefore, a datum was not considered suspect unless it appeared to be an outlier in
many of the methods. In most instances in which a datum was considered suspect, it had previously
been flagged as part of the data verification step.
The methods used to identify outliers were patterned after the validation procedures used for the
National Stream Survey (NSS) (Kaufmann et al., 1988). However, the procedures were updated to
reflect the collection of large amounts of data at individual streams, as in the ERP, instead of small
amounts of data collected at a large number of streams across a region, as in the NSS. Data were
first categorized by their type: snow lysimeter, snow core, and stream grab samples. Within these
classifications, each chemical constituent was investigated by region and by stream within region.
Graphical methods were the primaiy tool used for identifying outliers. Summary statistics and box-
plots (Chambers et al., 1983) were used to initially identify outliers. Scatterplot matrices and
59
-------
Spearman correlation matrices were also used to identify pairs of chemical parameters with empirical
linear or curvilinear relationships. These pairs were plotted in more detail and were studied carefully to
identify outliers. In general, the plots studied include pH versus ANC, pH and ANC versus major
cations, pH and ANC versus aluminum (total dissolved, organic monomeric, and inorganic mono-
meric), pH and ANC versus major anions, and conductivity versus sum of anions and sum of cations.
Plots between major cations and anions were also studied when a good relationship existed. Plots of
chemical variables and discharge were also examined.
The outliers were then tabulated and investigated more closely. First, ERL-C personnel discussed
identified outliers with the field cooperators. If the conditions suggested that the sample values were
in error, they were corrected if possible. Data that could not be corrected in this fashion, but had
been identified as in error, were removed from the database. Once the rejected data had been
removed, the revised data set was again subjected to these procedures.
4.5.4 Data Availability
The ERP data and the database user's guide (see Appendix A for a list of ERP publications) will be
available to the public in late 1993 through the National Technology Information Service (NTIS), U.S.
Department of Commerce, 5285 Port Royal Rd., Springfield, VA 22161.
60
-------
Table 4-1. Parameters of Interest and Associated Measurement Methodology for the Wet
Deposition Chemistry Element, Episodic Response Project .
Parameter
Precipitation Volume
pH at 25°C
Conductivity at 25°C
Calcium
Magnesium
Sodium
Potassium
Ammonium
Chloride
Nitrate
Sulfate
Method
Rain gauge (Belfort type or tipping
bucket)
Electrometric; pH meter and combin-
ation electrode
Electrolytic; conductivity cell and
meter
Inductively coupled argon plasma
emission spectroscopy
Inductively coupled argon plasma
emission spectroscopy
Inductively coupled argon plasma
emission spectroscopy
Inductively coupled argon plasma
emission spectroscopy
Colorimetric, automated (phenate)
Ion chrotnatography
Ion chromatography
Ion chromatography
References
Topol and Ozdimir, 1986
EPA 150.6; Peden et al., 1986;
Dana, 1982
EPA 120.6; Peden et al., 1986;
U.S. EPA, 1987
EPA 200.7; U.S. EPA, 1987
EPA 200.7; U.S. EPA, 1987
EPA 200.7; U.S. EPA, 1987
EPA 200.7; U.S. EPA, 1987
EPA 350.7; U.S. EPA, 1987;
Peden et al., 1986
EPA 300.6; Peden et al., 1986;
U.S. EPA, 1987
EPA 300.6; Peden et al., 1986;
U.S. EPA, 1987
EPA 300.6; Peden et al., 1986;
U.S. EPA, 1987
61
-------
Table 4-2. Sample Handling and Holding Times, Stream Chemistry Component, Episodic
Response Project (adapted from Peck et al., 1988)
Analytes
Calcium
Magnesium
Potassium
Sodium
Calcium
Magnesium
Potassium
Sodium
Aluminum, total dissolved
Aluminum, monomeric
Aluminum, nonlabile
Chloride
Nitrate
Sulfate
Silica
Dissolved organic carbon
(DOC)
Ammonium
Acid neutralizing capacity
(ANC)
PH
Conductivity
Processing
Filtration
(0.45 /*m)b
Filtration
(0.45 ftm)c
Filtration
(0.10/fm or
0.22 pm)
Filtration
(0.10 pm or
0.22 jUm)
Filtration
(0.10/im or
0.22 urn)
Filtration
(0.45 fim)
Filtration
(0.45 urn)
None
Preservation3
pH < 2 with HNO3; store at 4°C
None; store at 4°C with no head-
space
pH < 2 with HNO3; store at 4°C
Extraction into methyl-isobutyl-
ketone (MIBK); store at 4°C
Cation exchange; extraction into
methyl-isobutyl-ketone (MIBK);
store at 4°C
None; store at 4°C with no head-
space
pH < 2 with H2SO4; store at 4°C
None; store at 4°C with no head-
space
Holding Time
Days
(Preserved)
28
28
28
28
14
14
14
14
14
14
14
14
14
14
14
14
14
14
7
14
a Bulk samples maintained at approximately 4°C until processed and preserved.
Applicable for samples being analyzed by atomic absorption spectroscopy.
£
Applicable for samples being analyzed by ion chromatography.
-------
Table 4-3. Analytes of Interest and Associated Measurement Methodology, Stream Chemistry
Component, Episodic Response Project (adapted from Peck et al., 1988)
Parameter
Summary of Method
References
Field Measurements
pH, in situ
Conductivity at 25°C, in situ
Temperature, in situ
pH, field
Conductivity at 25°C, field
Electrometric (USGS mini-monitor), with
combination electrode
Electrolytic (USGS mini-monitor)
Thermometric; Thermistor (USGS mini-monitor)
Electrometric, with glass combination electrode
Electrolytic, conductivity cell and meter
EPA 150.2 (modified); Metcalf et al.,
1988
EPA 120.6 (modified); Metcalf et al.,
1988
EPA 170.1; Metcalf et al., 1988
EPA 150,6 (modified); Hagley et al.,
1988
EPA 120.6 (modified); Hagley et al., '
1988
Laboratory Measurements
Calcium, dissolved3
Magnesium, dissolved3
Potassium, dissolved3
Sodium, dissolved3
Ammonium, dissolved3
Aluminum, total dissolved1"
Aluminum, monomericb
Aluminum, nonlabileb
(organic monomeric)
Chloride, dissolved3
Nitrate, dissolved3
Sulfate, dissolved3
Silica, dissolved3
Carbon, dissolved3 organic (DOC)
PH
Conductivity at 25°C
Acid Neutralizing Capacity (ANC)
Atomic absorption spectroscopy (flame); Ion
chromatography
Atomic Absorption Spectroscopy (flame); Ion
chromatography
Atomic absorption spectroscopy (flame); Ion
chromatography
Atomic absorption spectroscopy (flame); Ion
chromatography
Colorirnetric, automated (phenate)
Digestion at pH < 1; analysis by atomic
absorption spectroscopy (furnace)
Complexation with 8-hydroxy-quinoline,
extraction into methyl-isobutyl-ketone, analysis
by atomic absorption spectroscopy (furnace)
Cation exchange, complexation with 8-hydroxy-
quinoline, extraction into methyl-isobutyl-ketone,
analysis by atomic absorption spectroscopy
(furnace)
Ion chromatography
Ion chromatography
Ion chromatography
Automated colorirnetric (molybdate blue)
UV-promoted persulfate oxidation, IR detection
Electrometric: pH meter with glass combination
electrode
Electrolytic, with conductance cell and meter
Acidimetric titration with Gran plot analysis
EPA 200.6; U.S. EPA 1987; EPA 300.7;
Peden et al., 1986
EPA 200.6; U.S. EPA, 1987; EPA
300.7; Peden et al., 1986
EPA 200.6; U.S. EPA, 1987; EPA
300.7; Peden et al., 1986
EPA 200.6; U.S. EPA, 1987; EPA
300.7; Peden et al., 1986
EPA 350.7; U.S." EPA, 1987; Peden et
al., 1986
EPA 202.2; U.S. EPA, 1987
EPA 202.2 (modified); Driscoll, 1984;
U.S. EPA, 1987
EPA 202.2 (modified); Driscoll, 1984
EPA 300.6; U.S. EPA, 1987
EPA 300.6; U.S. EPA, 1987
EPA 300.6; U.S. EPA, 1987-
EPA 370.1 (modified); U.S. EPA, 1987
EPA 41 5.2; U.S. EPA, 1987
EPA 150.6; U.S. EPA, 1987
EPA 120.6; U.S. EPA, 1987
EPA 310.1 (modified); U.S. EPA, 1987
Dissolved is defined as passing through a 0.45 fim pore filter.
1 Dissolved is defined as passing through a 0.10 ^m or 0.22 fim pore filter.
63
-------
Table 4-4. Dates Used to Define Seasons for Inorganic Monomeric Aluminum (Al|m)
Analyses
Season
Fall 1988
Spring 1989
Fall 1989
Spring 1990
Adirondacks
10/1 -12/15
4/1 - 6/30
9/15-12/15
4/1 - 6/30
Catskills
10/1 -12/31
3/1 - 5/30
9/15-12/31
3/1 - 5/30
Pennsylvania
10/1 -12/31
3/1 - 5/30
9/15-12/31
3/1 - 4/30
Table 4-5. Number of Measurements of Inorganic Monomeric Aluminum (Aljm) per Season
and Stream
Stream
Adirondacks:
Fly Pd. Outlet
Bald Mt. Brook
Buck Creek
Seventh L. Inlet
Catskills:
Black Brook
High Falls Brook
Biscuit Brook
E. Br. Neversink
Pennsylvania:
Benner Run
Roberts Run
Stone Run
Baldwin Creek
Linn Run
Fall
1988
0
0
0
0
0
0
0
0
0
5
1
19
23
Spring
1989
4
3
3
4
2
5
24
32
30
28
24
29
35
Fall
1989
16
24
28
27
8
7
19
22
15
17
16
11
12
Spring
1990
11
16
24
17
0
13
23
28
23
28
44
17
44
All
(total)
64
73
92
84
10
27
69
86
68
79
86
76
114
64
-------
Table 4-6. Models Selected to Estimate Inorganic Monomeric Al (Aljm) from Other
Measured Physical and Chemical Variables: Adirondacks
Application3
Bald Mt. Brook
Overall
Fall 1989
Spring 1990
Back-up
Buck Creek
Overall
Fall 1989
Spring 1990
Back-up
Fly Pond Outlet
Overall
Seventh Lake Inlet
Overall
Fall 1989
Spring 1990
Back-up
Model, Allm =b
-98.85 + 0.6760(Altd) + 2.741 (ANC) -
0.014181(ANC*Altd)
10**(10.65 - 1.804(log Altd) - 1.134(pH) +
0.09356(DOC) + 0.0091 (2CB))
-69.43 + 0.9728(Altd) - 39.62(DOC)
10** (5.083 - 0.5371 (pH)) - 25
-76.83 + 0.9199(Altd) - 25.29(DOC)
236.5 - 6.455(ANC) - 1.124(Q)
-101.2 + 0.4544(Ay + 4.478(NO3) -
14.79(DOC)
10**(5.661 - 0.6520(pH)) - 25
No significant models; median measured
values used for the prediction
393.9 + 0.5049(Altd) - 15.07(DOC) -
66.00(pH)
10**(5.2317 - 0.5696(pH) - 0.0028(Q)) - 25
-81 .23 + 0.7989(Altd) - 20.59(DOC)
10**(5.522 - 0.6405(pH)) - 25
R2
0.83
0.82
0.93
0.58
0.91
0.73
0.97
0.80
-
0.84
0.69
0.91
0.57
n
73
24
16
73
90
28
24
90
68
84
27
17
84
RMSEC
62.1
52.9
30.8
80.0
52.8
46.7
30.0
102.7
37.6
•
40.1
41.5
22.3
59.9
Season specific models were applied for the seasons listed. Otherwise, the overall (full-data) model was used. For
those samples with missing values for one or more predictor variables, the back-up model was used to estimate Alim.
Units for variables used in models: ANC, 2CB, NO3 = ^eq/L; Aljm, Altd = /jg/L; DOC = mg/L; pH = pH units; Q = cfs.
2CR = sum of base cations.
^
RMSE = [{sum of squared prediction errors)/n] .
65
-------
Table 4-7. Models Selected to Estimate Inorganic Monomeric Al (Al,m) from Other
Measured Physical and Chemical Variables: Catskills
Application8
Biscuit Brook
Overall
Spring 1989
Fall 1989
Spring 1990
Back-up
Black Brook
Overall
Back-up
East Branch
Neversink River
Overall
Spring 1989
Fall 1989
Spring 1990
Back-up
High Falls Brook
Overall
Back-up
Model, Alim =b
127.5 + 0.4526(Altd) - 11.43(DOC) +
1.316(804) -0.8396(SCB) + 8.55(COND)
1820 + 1.699(ANC) + 8.773(COND) - 675.4
(logSO4) - 248.5(log 2CB)
-0.4018 + 0.1983(Altd)
61 .10 +0.321 1 (Altd) - 4.158(ANC)
-81.89 + 5.950(COND) + 2.193(ANC)
9
-965.3 + 1.959(Q) + 455.5(logSO4)
396.2 - 59.55(pH)
435.2 - 226.2(pH) - 10.88(DOC) +
221.7(logAltd) + 231.1 (log NO3)
69.04 +0.3929(Altd) - 4.693(ANC)
-530.2 + 3.771 (SO4) + 9.953(COND)
-2177 + 360.4(pH) + 13.13(NO3)
974.9 - 183.7(pH) + 2.891 (NO.,)
17.52 + 0.3797 (Altd) - 16.15(DOC) +
0.71 94(Q)
204.0 + -41.65(pH) + 7.110(COND) -
0.6156(SCB)
R2
0.87
0.93
0.55
0.91
0.58
0.76
0.60
0.82
0.87
0.71
0.82
0.56
0.78
0.59
n
57
20
17
17
64
10
10
65
29
20
23
75
23
27
RMSEC
14.5
7.9
11.1
12.7
68.6
11.7
14.9
43.0
37.7
26.2
42.0
69.9
11.7
15.3
Season specific models were applied for the seasons listed. Otherwise, the overall (full-data) model was used. For
thosa samples with missing values for one or more predictor variables, the back-up model was used to estimate Alim.
Units for variables used in models: ANC, 2CB, SO4, NO3 = /ieq/L; Alim, Altd = ^ug/L; DOC = mg/L; pH = pH units; Q
- cfs; COND = /zmhos/cm. 2BC = sum of base cations; COND = conductivity.
RMSE ** [(sum of squared prediction errors)/n]V4.
66
-------
Table 4-8. Models Selected to Estimate Inorganic Monomeric Al (Al|m) from Other
Measured Physical and Chemical Variables: Pennsylvania
Application3
Baldwin Creek
Overall
Back-up
Benner Run
Overall
Spring 1 989
Fall 1989
Spring 1990
Back-up
Linn Run
Overall
Fall 1989
Spring 1989
Fall 1989
Spring 1 990
Back-up
Roberts Run
Overall
Spring 1989
Fall 1989
Spring 1 990
Back-up
Stone Run
Overall
Spring 1 989
Fall 1989
Spring 1990
Back-up
Model, Alim =b
1 1.60 + 8.804(Q) - 1.406(pH*Q)
5.486 + 1.104(Q)
543.3 - 89.52(pH)
10** (5.347 - 0.6588(pH) + 0.0586(DOC) -
0.0056(Q)) - 25
10**(5.131 - 0.5985(pH)) - 25
405.9 - 89.40(pH) + 0.7474(20^
-44.70 + 78.53(log Q)
10**(5.2479 - 0.0027(Altd) - 0.5742(pH) +
0.0007(Alte,*pH)) - 25
1098 + 0.3423(Altd) + 3.270(Q) - 182.4(pH)
10** (4.8071 + 0.3953(log Altd) + 0.2106(log Q) -
1.536(log2CB))-25
10**(4.53 + 0.001 6(Altd) -0.4743(pH)) -25
1 1 41 + 0.5969(Altd) - 1 51 .2(pH) - 0.7081 (2CB)
-140.4 + 311.6(logQ)
10**(5.163 - 0.571 8(pH) - 0.002969(Q)) - 25
10**(6.892 - 0.8474(pH) - 0.3631 (logQ)) - 25
1 0** (5.209 - 0.5797(pH)) - 25
10**(2.249 + 9.429(log DOC) - 1.859(log
DOC*pH)) - 25
10**(3.742 - 0.3638(pH) + 0.0036(304) -
0.001 7(2CR))- 25
10**(7.353 - 0.9702(pH) - 0.0439(Q) +
0.0090(pH*Q)) - 25
10** (0.2465 + 0.8306(log Altd) + 1.432(log Q) -
0.5088(log AIM*log Q)) - 25
-55.85 + 2.879 (Altd) - 13.25(Q) -0.1498(Altd*Q)
1 758 - 303.0(pH) + 3.391 (Q)
1 849 - 31 4.5(pH) + 1 1 .75(NOJ - 3.974(CI) +
0.5912(20^
R2
0.79
0.72
0.83
0.92
0.92
0.80
0.45
0.90
0.97
0.96
0.98
0.80
0.51
0.92
0.62
0.96
0.82
0.92
0.96
0.95
0.96
0.88
0.91
n
74
74
68
30
15
22
67
107
19
35
11
42
111
76
28
17
28
79
83
23
15
43
86
RMSE°
\9.8
11.3
14.9
9.3
12.9
9.8
27.4
57.6
37.1
26.2
20.8,
66.5
118.1,
28.2
20.7
9.9
23.0
27.1
35.8
21.1
9.3
34.4
39.9
Season specific models were applied for the seasons listed. Otherwise, the overall (full-data) model was used. For those samples
with missing values for one or more predictor variables, the back-up model was used to estimate Alim.
Units for variables used in models: 2CB, SO4, Cl = peq/L; Alim> Al,d = ftg/L; DOC = mg/L; pH = pH units. 2CB = sum of base
cations.
• T/i
RMSE = [(sum of squared prediction errors)/n] .
67
-------
Table 4-9. fn Situ Bioassays with Brook Trout in Adirondack Streams
o>
00
Stream*
FALL 1988
FPO
8MB
BCK
SLI
SPRING 1989
FPO
BMB
BCK
SLI
FALL 1989
FPO
BMB
BCK
SLI
FPO
BMB
BCK
SLI
Source
of flshb
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
Starting
Date
1 1/01/88
11/01/88
1 1/01/88
1 1/01/88
5/26/89
5/26/89
5/26/89
5/26/89
9/27/89
9/27/89
9/27/89
9/27/89
10/06/89
10/06/89
10/06/89
10/06/89
Ending
Date
12/01/88
12/01/88
12/01/88
12/01/88
6/20/89
6/20/89
6/20/89
6/20/89
1 1/06/89
1-1/06/89
11/06/89
1 1/06/89
11/06/89
11/06/89
1 1/06/89
11/06/89
Total
No, Days
30
30
30
30
25
25
25
25
40
40
40
40
31
31
31
31
Total
No. Fish
20
20
15
20
20
20
20
20
20
20
20
. 20
20
20
20
20
Length (mm)
Mean
80.2
74.8
104.1
70.7
75.6
76.4
78.6
77.4
73.2
74.4
67.0
69.4
68.4
67.4
70.3
67.2
STD
9.0
8.9
13.5
10.9
6.5
7.7
7.8
8.1
10.9
9.9
12.1
12.4
7.2
6.9
7.9
9.8
Weight (g)
Mean
4.00
3.60
10.54
3.25
3.60
4.31
4.64
3.50
3.05
3.45
2.95
4.80
2.20
2.40
3.00
5.50
STD
1.46
1.27
3.43
1.07
0.55
1.03
1.12
0.71
1.32
1.47
1.87
2.98
0.89
0.82
1.03
2.35
1 FPO = Fly Pond Outlet; BMB = Bald Mountain Brook; BCK = Buck Creek; SLJ = Seventh Lake Inlet.
' CP = -common pool (see text for explanation).
-------
Table 4-9. In Situ Bioassays with Brook Trout in Adirondack Streams (Continued)
Stream*
SPRING 1990
FPO
BMB
BCK
SLI
Source
of fish"
CP
CP
CP
CP
Starting
Date
4/30/90
4/30/90
4/30/90
4/30/90
Ending
Date
5/30/90
5/30/90
5/30/90
5/30/90
Total
No. Days
30
30
30
30
Total
No. Fish
20
20
20
20
Length (mm)
Mean
73.2
77.2
76.0
75.6
STD
6.5
7.5
6.0
10.3
Weight (g)
Mean
2.65
3.72
3.35
2.90
STD
0.75
1.35
0.81
1.48
FPO a Fly Pond Outlet; BMB = 5a!d Mountain Brook; BCK = Buck Greek; SU = Seventh Lake Inlet.
1 CP = common pool (see text for explanation).
0)
CO
-------
Table 4-10. In Situ Bioassays with Brook Trout in Catskill Streams
Stream"
FALL 1988
Black
Biscuit
EBrNS
Black
High Falls
Biscuit
EBrNS
SPRING 1989
Black
High Falls
Biscuit
EBrNS
High Falls
EBrNS
Source
offishb
Black
Black
Biscuit
Biscuit
EBrNS
EBrNS
Black
Black
Black
Black
PIN
PIN
PIN
PIN
H.Falls
EBrNS
Starting
Date
11/10/88
11/10/88
11/03/88
11/03/88
1 1/04/88
1 1/04/88
1 1/30/88
11/30/88
11/30/88
1 1/30/88
4/08/89
4/06/89
4/06/89
4/06/89
4/06/89
4/06/89
Ending
Date
12/15/88
12/15/88
12/14/88
12/14/88
12/15/88
12/15/88
12/15/88
12/15/88
12/14/88
12/15/88
4/27/89
4/27/89
4/27/89
4/27/89
4/27/89
4/27/89
Total
No. Days
35
35
41
41
41
41
15
15
14
15
19
21
21
21
21
21
Total
No. Fish
19
9
21
13
14
6
20
20
20
20
21
21
21
16
21
21
Length (mm)
Mean
75.9
129.3
76.0
145.3
95.8
150.7
69.5
77.2
80.1
80.4
74.8
71.5
73.8
76.9
69.0
68.8
STD
10.2
12.5
15.9
31.0
15.6
10.1
6.8
9.8
9.1
. 12.4
6.4
6.3
9.3
5.1
6.4
6.6
Weight (g)
Mean0
-
57.1
--
42.0
-
--
3.05
3.48
4.32
4.41
4.27
2.94
3.50
3.69
2.38
2.62
STD°
—
9.3
—
14.7
--
-
0.55
0.92
1.20
1.80
0.39
0.87
1.20
0.79
0.74
0.80
EBrNS = East Branch Neversink River.
PIN = Pigeon Creek; WBrNS = West Branch Neversink River.
- = no measurements were made offish weight.
-------
Table 4-10. In Situ Bioassays with Brook Trout in Catskill Streams (Continued)
Stream3
SPRING 1989
High Falls
Biscuit
EBrNS
High Falls
EBrNS
High Falls
EBrNS
FALL 1989
Black
High Falls
Biscuit
EBrNS
High Falls
EBrNS
High Falls
EBrNS
High Falls
EBrNS
Source
of fishb
(cont.)
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
H.Falls
EBrNS
WBrNS
WBrNS
WBrNS
WBrNS
H.Falls
EBrNS
WBrNS
WBrNS
H.Falls
EBrNS
Starting
Date
5/04/89
5/04/89
5/05/89
5/19/89
5/19/89
5/19/89
5/19/89
10/04/89
10/04/89
10/04/89
10/04/89
10/05/89
10/07/89
1 1/08/89
11/08/89
11/08/89
1 1/07/89
Ending
Date
5/17/89
5/17/89
5/16/89
6/24/89
6/24/89
6/24/89
6/24/89
11/09/89
1 1/08/89
1 1/08/89
1 1/07/89
1 1/08/89
11/07/89
1 1/30/89
11/30/89
1 1/30/89
11/30/89
Total
No. Days
13
13
11
36
36
36
36
36
35
35
34
34
31
22
22
22
23
Total
No. Fish
21
21
21.
21
21
21
21
21
21
21
21
21
21
21
21
21
21
Length (mm)
Mean0
—
-
71.8
78.4
80.5
73.4
77.0
66.6
65.6
64.7
65.2
72.0
68.0
69.5
67.7
72.2
69.4
STD°
-
-
5.3
7.4
6.1
6.9
6.5
9.6
7.9
7.2
7.5
5.6
4.4
7.4"
7.1
6.0
5.5
Weight (g)
Mean0
--
-
3.89
2.64
3.55
1.95
2.63
2.52
2.19
2.00
2.14
2.81
2.19
2.86
2.62
2.98
2.95
STD°
—
--
0.76
0.88
1.06
0.67
0.86
1.17
0.87
0.77
0.65
0.68
0.51
0.85
0.74
0.72
0.67
EBrNS = East Branch Neversink River.
' PIN a Pigeon Creek; WBrNS = West Branch Neversink River.
! - = no measurements made of fish length of weight
-------
Table 4-10. In Situ Bioassays with Brook Trout In Catskill Streams (Continued)
Stream11
SPRING 1990
High Falls
Biscuit
High Falls
Biscuit
EBrNS
High Falls
EBrNS
Source
offishb
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
H.Falls
EBrNS
Starting
Date
3/09/90
3/09/90
4/06/90
• 4/06/90
4/06/90
4/06/90
4/06/90
Ending
Date
4/06/90
4/06/90
5/06/90
5/06/90
5/06/90
5/06/90
5/06/90
Total
No. Days
28
28
30
30
30
30
30
Total
No. Fish
21
21
21
21
21
21
21
Length (mm)
Mean
73.1
69.9
74.3
75.0
74.4
74.5
66.2
STD
3.8
5.6
6.1
5.3
5.9
4.6
4.3
Weight (g)
Mean
3.19
2.86
2.90
3.12
3.37
2.76
2.71
STD
0.60
0.57
0.77
0.55
0.83
0.44
0.73
M a EBrNS = East Branch Neversink River.
b PIN = Pigeon Creek; WBrNS = West Branch Neversink River.
-------
Table 4-11. In Situ Bioassays with Brook Trout in Pennsylvania Streams
Stream
FALL 1988
Banner
Roberts
Stone
Baldwin
Linn
SPRING 1989
Benner
Roberts
Stone
Baldwin
Linn
Source
of fish*
Benner
Benner
Roberts
Stone
Benner
Roberts
Stone
Baldwin
Baldwin
Benner
Benner/
Roberts
Roberts
Stone
Baldwin
Linn
Starting
Date
10/11/88
10/05/88
10/05/88
10/05/88
10/06/88
10/06/88
10/06/88
10/13/88
10/13/88
10/13/88
2/21/89
2/23/89
2/22/89
2/27/89
2/28/89
Ending
Date
11/16/88
11/10/88
11/10/88
11/10/88
11/10/88
11/10/88
11/10/88
11/18/88
11/18/88
11/18/88
4/04/89
4/03/89
4/03/89
4/07/89
4/06/89
Total
No. Days
36
36
36
36
35
35
35
36
36
36
42
39
40
39
37
Total
No. Fish
20
21
39
22
20
35
33
36
37
14
21
21
21'
21
42
Length (mm)
Mean
68.8
67.7
66.6
66.3
62.0
64.4
65.5
74.5
73.4
61.3
63.5
112.6
109.5
159.8
86.9
STD
22.8
20.0
8.5
7.5
12.8
7.2
8.6
17.9
20.3
6.7
7.6
203.2
203.9
279.0
22.0
Weight (g)
Mean6
-
-
—
-
--
-
--
-
--
• -
--
2.25
2.30
3.00
3.20
3.00
STDb
—
~
-
—
--
-
—
—
-
--'
--
0.62
0.48
0.00
0.63
--
1 CP = common pool (see text for explanation).
1 - = no measurements made offish weight.
-------
Table 4-11. In Situ Bioassays with Brook Trout in Pennsylvania Streams (Continued)
Stream
SPRING 1989
Roberts
Stone
Linn
FALL 1989
Benner
Roberts
Stone
Baldwin
Linn
SPRING 1990
Benner
Stone
Benner
Stone
Source
offish8
(cont.)
Roberts
Stone
Linn
Benner
Stone
Stone
CP
CP
CP
CP
CP
CP
Starting
Date
3/14/89
3/14/89
3/20/89
10/30/89
10/31/89
10/31/89
1 1/02/89
1 1/02/89
3/02/90
3/02/90
3/22/90
3/22/90
Ending
Date
4/03/89
4/03/89
4/06/89
1 1/20/89
1 1/20/89
11/20/89
11/22/89
11/22/89
3/22/90
3/22/90
4/11/90
4/11/90
Total
No. Days
20
20
17
21
20
20
20
20
20
20
20
20
Total
No. Fish
21
21
21
35
35
35
35
35
35
35
35
35
Length (mm)
Mean
69.0
68.0
79.8
60.9
60.8
62.1
72.8
69.1
74.3
76.2
84.9
82.1
STD
6.4
7.6
6.9
7.5
4.6
8.3
8.3
9.4
7.6
6.6
9.4
8.3
Weight (g)
Mean"
2.50
3.10
-
2.37
2.31
2.46
3.40
2.91
3.49
4.40
4.88
5.57
STDb
0.52
0.74
-
0.94
0.72
1.09"
0.98
1.04
1.31
1.14
1.55
1.56
1 CP = common pool (see text for explanation).
1 - = no measurements made offish weight
-------
Table 4-12. In Situ Bioassays with Blacknose Dace in Adirondack Streams
Streama
FALL 1988
FPO
BMB
BCK
SLI
SPRING 1989
FPO
BMB
BCK
SLI
FPO
BMB
BCK
SLI
FALL 1989
FPO
BMB
BCK
SLI
Source
of fishb
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
Starting
Date
11/01/88
11/01/88
11/01/88
11/01/88
5/08/89
5/08/89
5/08/89
5/08/89
5/18/89
5/18/89
5/18/89
5/18/89
9/27/89
9/27/89
9/27/89
9/27/89
Ending
Date
12/01/88
12/01/88
12/01/88
12/01/88
5/18/89
5/18/89
5/18/89
5/18/89
6/28/89
6/28/89
6/28/89
6/28/89
1 1/06/89
1 1/06/89
1 1/06/89
1 1/06/89
Total
No. Days
30
30
30
30
10
10
10
10
41
41
41
41
40
40
40
40
Total
No. Fish
20
20
20
20
21.
21
22
21
21
21
21
21
20
20
20
20
Length (mm)
Mean
70.6
72.8
71.2
71.6
37.4
37.4
39.5
38.9
38.3
38.0
38.2
37.1
61.4
62.2
.62.0
61,2
STD
3.4
5.2
6.8
6.9
3.6
4.1
3.4
3.8
4.8
3.7
4.1
4.5
5.7
6.2
6.3
6.9
Weight (g>
Mean0
2.80
4.10
3.93
3.82
-
0.50
0.50
0.50
0.67
--
0.53
~
1.90
2.20
2.60
6.00
STDC
0.62
0.97
1.03
1.10
.. .
, 0.00
0.00
0.00
0.29
-
0.14
-
0.85
.0.77
0.88
2.29
FPO = Fly Pond Outlet; BMB = Bald Mountain Brook; BCK = Buck Creek; SU = Seventh Lake Inlet.
CP = common pool (see text for explanation.
1 - = no measurements made of fish weight
-------
Table 4-12. In Situ Bioassays with Blacknose Dace in Adirondack Streams (Continued)
Stream*
SPRING 1990
FPO
BMB
BCK
SLI
Source
offish"
CP
CP
CP
CP
Starting
Date
4/30/90
4/30/90
4/30/90
' 4/30/90
Ending
Date
5/30/90
5/30/90
5/30/90
5/30/90
Total
No. Days
30
30
30
30
Total
No. Fish
20
20
20
20
Length (mm)
Mean
59.4
59.5
61.9
61.9
STD
3.6
5.7
4.7
5.7
Weight (g)
Mean
1.45
2.20
2.40
2.35
STD
0.51
0.70
0.75
0.75
1 FPO = Fly Pond Outlet; BMB = Bald Mountain Brook; BCK = Buck Creek; SU = Seventh Lake Inlet.
1 CP = common pool (see text for explanation).
-------
Table 4-13. In Situ Bioassays with Slimy Sculpin in Catskill Streams
Stream*
FALL 1988
Black
Biscuit
SPRING 1989
Black
High Falls
Biscuit
EBrNS
High Falls
Biscuit
EBrNS
FALL 1989
Black
High Falls
Biscuit
EBrNS
High Falls
EBrNS
Source
of fishb
Black
Biscuit
PIN
PIN
PIN
PIN
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
H.Falls
EBrNS
Starting
Date
11/10/88
11/03/88
4/08/89
4/06/89
4/06/89
4/06/89
5/04/89
5/04/89
5/07/89
10/04/89
10/04/89
10/04/89
10/04/89
10/05/89
10/07/89
Ending
Date
12/15/88
12/14/88
4/27/89
4/27/89
4/27/89
4/27/89
5/17/89
5/17/89
5/16/89
1 1/09/89
1 1/08/89
1 1/08/89
11/07/89
1 1/08/89
11/07/89
Total
No. Days
35
41
19
21
21
21
13
13
9
36
35
35
34
34
31
Total
No. Fish
21
20
21
21
21
19
21
21
21
21
21
21.
21
21
21
Length (mm)
Mean0
62.5
91.5
63.2
59.2
61.8
59.4
-
-
80.0
63.8
64.5
62.1
65.3
63.8
71.0
STDC
18.2
21.9
7.8
9.5
8.7
6.6
-
~
-
8.2
7.3
8.2
6.3
5.7
6.3
Weight (g)
Mean0
-
--
--
2.10
2.60
2.47
-
"
8.00
2.64
2.38
2.29
2.57
2.10
3.60
STD°
-
—
-
-
1.09
1.14
0.82
-
-
--
1.22
0.86
0,92
0.81
0.62
1.36
a EBrNS = East Branch Neversink River.
PIN = Pigeon Creek; WBrNS = West Branch Neversink River.
c - = no measurements made offish length or weight.
-------
Table 4-13. In Situ Bioassays with Slimy Sculpin in Catskill Streams (Continued)
Stream*
FALL 1989
High Falls
EBrNS
High Falls
EBrNS
Source
offishb
(cont.)
WBrNS
WBrNS
H.Falls
EBrNS
Starting
Date
11/08/89
11/08/89
11/08/89
11/07/89
Ending
Date
11/30/89
11/30/89
11/30/89
11/30/89
Total
No. Days
22
22
22
23
Total
No. Fish
21
21
21
21
Length (mm)
Mean
65.6
68.0
67.9
68,4
STD
6.6
8.9
6.4
5.6
Weight (g)
Mean
2.76
3.24
3.05
3.33
STD
1.04
1.14
0.80
1.02
a EBrNS = East Branch Neversink River.
b PIN = Pigeon Creek; WBrNS = West Branch Neversink River.
00
-------
Table 4-14. In Situ Bioassays with Sculpin in Pennsylvania Streams
Stream
FALL 1988
Benner
Roberts
Stone
Linn
Baldwin
• Linn
SPRING 1989
Benner
Roberts
Stone
Baldwin
Linn
SPRING 1990
Benner
Stone
Benner
Stone
Species
Slimy
Slimy
Slimy
Slimy
Mottled
Mottled
Slimy
Slimy
Slimy
Mottled
Mottled
Slimy
Slimy
Mottled
Mottled
Source
of Fisha
Benner
Benner
Benner
Benner
Baldwin
Baldwin
SixMile
SixMile
SixMile
Linn
Baldwin
CP
CP
CP
CP -
Starting
Date
10/11/88
10/05/88
10/06/88
10/13/88
10/13/88
10/13/88
2/21/89
2/22/89
2/22/89
2/27/89
2/28/89
3/02/90
3/02/90
3/02/90
3/02/90
Ending
Date
11/16/88
11/10/88
11/10/88
11/18/88
11/18/88
11/18/88
4/04/89
4/03/89
4/03/89
4/07/89
4/06/89
3/22/90
3/22/90
3/22/90
3/22/90
Total
No. Days
36
36
35
36
36
36
42
40
39
39
37
20
20
20
20
Total
No. Fish
55
47
27
21
25
23
21
21
21
21
21
35
35
35
35
Length (mm)
Mean
40.2
52.7
66.8
67.6
68.6
70.8
55.1
56.6
100.3
96.5
100.1
69.4
69.2
70.0
70.7
STD
17.0
21.9
8.9
8.0
7.2
13.0
4.5
5.4
206.0
206.8
206.1
8.0
7.8
8.0
10.1
Weight (g)
Meanb
-
-
-
..
..
•• -
1.81
2.00
..
1.07
-
3.34
4.03
3.37
4.29
STDb
-
--
-
..
-
-
0.68
0.00
--
0.27
-,
1.51
1.29
1.37
1.84
CP = common pool (see text for explanation).
' - = no measurements made of fish weight.
-------
Table 4-14. In Situ Bioassays with Sculpin in Pennsylvania Streams (Continued)
Stream
SPRING 1990
Banner
Stone
Benner
Stone
Species
(cont)
Slimy
Slimy
Mottled
Mottled
Source
of Fish"
CP
CP
CP
CP
Starting
Date
3/22/90
3/22/90
3/22/90
3/22/90
Ending
Date
4/11/90
4/11/90
4/1 1/90
4/1 1/90
Total
No. Days
20
20
20
20
Total
No. Rsh
35
35
35
35
Length (mm)
Mean
69.2
65.5
76.2
69.8
STD
7.1
7.7
9,3
7.3
Weight (g)
Mean
3.17
3.31
4.80
3.71
STD
1.07
1.13
1.86
1.27
CP - common pool (see text for explanation).
-------
Table 4-15. Radiotelemetry Studies of Brook Trout Movement in Adirondack Streams
Stream8
SPRING 1989
FPO
SLI
FALL 1989
FPO
8MB
BCK
SPRING 1990
FPO
BCK
Starting
Date
5/22-26/89
5/22-23/89
1 0/04/89
10/04/89
10/04/89
5/04/90
5/04/90
Ending
Date
6/20/89
6/20/89
11/03/89
11/03/89
1 1/03/89
6/05/90
6/05/90
Total
No. Days
25-29
28-29
30
30
30
32
32
Total
No. Fish
8
7
9
10
9
6
6
Length (mm)
Mean
182,0
174.1
182.1
181.2
181.1
170.8
175.3
STD
28.6
29.2
27.8
27.2
21.2
11.7
8.3
Weight (g)
Mean
55.8
54.7
62.1
65.3
,56.7
44.0
46.2
STD
27.6
38.2
36.8
38.8
24.5
7.6
7.4
00
' FPO = Fty Pond Outlet; SU = Seventh Lake Inlet; 8MB = Bald Mountain Brook; BCK = Buck Creek.
-------
Table 4-16. Radiotelemetry Studies of Brook Trout Movement in Catskill Streams
Stream8
SPRING 1989
High Falls
EBrNS
FALL 1989
High Falls
EBrNS
SPRING 1990
High Falls
Biscuit
Source
offish"
PIN
H. Falls
PIN
EBrNS
WBrNS
H. Falls
WBrNS
EBrNS
WBrNS
WBrNS
Starting
Date
4/12/89
4/12/89
4/12/89
4/12/89
10/17//89
10/17/89
10/17/89
10/17/89
3/24/90
3/24/90
Ending
Date
4/27/89
4/27/89
4/28/89
4/28/89
1 1/25/89
1 1/25/89
11/28/89
1 1/28/89
5/23/90
5/22/90
Total
No. Days
15
15
16
16
39
39
42
42
60
59
Total
No. Fish
4
5
4
5
8
5
4
9
10
11
Length (mm)
Mean
182.2
225.4
179.8
185.4
220.8
222.6
202.8
191.9
221.5
203.8
STD
9.4
11.5
6.3
13.5
32.4
18.3
15.9
8.7
17.6
18.2
Weight (g)
Mean
57.2
99.6
50.8
63.2
111.6
112.4
83.2
71.6
99.7
80.8
STD
6.7
16.6
2.8
15.0
47.3
20.5
21.2
11.7
25.4
18.1"
03
to
1 EBrNS = East Branch Neversink River.
' PIN = Pigeon Creek; WBrNS = West Branch Neversink River.
-------
Table 4-17. Radiotelemetry Studies of Brook Trout Movement in Pennsylvania Streams
Stream
FALL 1988
Baldwin
Linn
SPRING 1989
Baldwin
Linn
FALL 1989
Benner
Stone
SPRING 1990
Benner
Stone
Starting
Date
10/15-16/88
10/15-16/88
3/02-03/89
3/02-03/89
11/06/89
11/06/89
3/07/90
3/07/90
Ending
Date
11/14/88
11/14/88
4/07/89
4/06/89
11/28/89
1.1/30/89
4/06/90
4/05/90
Total
No. Days
29-30
29-30
35-36
34-35
22
24
30
29
Total
No. Fish
10
10
15
15
14
15
10
10
Length (mm)
Mean
172.9
188.4
176.7
179,1
174,8
179.2
178.8
193.5
STD
13.9
25.6
20.5
19.0
16.7
14.8
25.9
25.9
Weight (g)
Mean
55.5
72.2
45.8
48.2
48.5
53.0
60.8
65.8
STD
16.6
31.6
17.8
14.0
15.4
16.4
30.2
22.9
-------
Table 4-18. Sampling Dates for Quantitative Surveys of Fish Communities in ERP Study Reaches
Region
Adirondacks
Catskills
Pennsylvania
1988
9/22-10/12
10/10-12/12
7/12-14
11/15-22
1989
4/24-26
6/21-27
9/01-12
11/07-14
5/1-6/20
1/13-29
8/22-28
10/13-11/27
3/29^/6
6/28-30
8/22-25
11/28-12/1
1990
4/19-24
6/12-14
6/2-6/6
4/3-10
5/31-6/5
Table 4-19. Electrofishing Efficiency Checks
Region
Adirondacks
Catskills
Pennsylvania
Date
6/89
6/89
10/89
N/A
No.
Marked
Fish
25
40
40
50
No.
Passes
3
3
3
4
No. Fish
Caught
17
39
26
39
Population
Estimate
N/Aa
39
28
40 (± 4)
Capture
Efficiency
68%
97%
66%
78%
Accuracy
Population
Estimate
N/A
97%
70%
80%
Data not available (not calculated or reported by regional cooperator).
84
-------
Table 4-20. Fish Transplants
Region
Adirondacks
Catskills
Pennsylvania
Date
September 1988
April 1989
September 1989
November 1989
May-June 1989
October 1988 and
February 1989
Stream
Fly Pond Outlet
Bald Mt. Brook
Buck Creek
Seventh L. Inlet
Fly Pond Outlet
Bald Mt. Brook
Buck Creek
Seventh L. Inlet
Fly Pond Outlet
Bald Mt. Brook
Buck Creek
Seventh L. Inlet
Fly Pond Outlet
Bald Mt. Brook
Buck Creek
Seventh L. Inlet
High Falls
Biscuit Brook
E. Br. Neversink
Benner Run
Roberts Run
Stone Run
Baldwin Creek
Linn Run
Brook Trout Stocked
Number
0
30
103
65
51
26
112
100
t -
43
42
50
26
44
77
28
67
Weight6 (g)
0
126
820
804
358
183
974
866
718
680
770
--
--
—
Forage Fish Stocked8
Number
57
30
86
122
71
44
113
185
46
29
81
126
32
21
82
106
0
29
50
50
100
50
100
100
Weightb (g)
154
103
203
323
103
104
210
323
139
80
205
298
83
63
193
278
0
180
280
-
--
—
00
01
a Blacknose dace in Adirondack streams; slimy sculpin in Catskill streams and in Benner, Roberts, and Stone; and mottled sculpin in Baldwin and Linn.
- = no measurements made offish weight.
-------
-------
SECTION 5
STREAM HYDROLOGY AND CHEMISTRY
In Section 5, we describe the hydrologic and chemical responses of the ERP streams during the
course of the study. Section 5.1 provides the climatic setting during the ERP and summarizes the
occurrence of major hydrologic events. Section 5.2 includes an overview of the chemical charac-
teristics of the ERP streams and describes their chemical responses during episodes. Subsection
5.2.1 describes episodic changes of ANC, pH, and Alinv In Subsection 5.2.2, we explore the role
that major ions played in controlling episodes in ERP streams.
5.1 HYDROLOGIC RESPONSE
5.1.1 Climatic Conditions
We can obtain an overview of the climatic conditions during the ERP by examining long-term
records for climatic stations located near the ERP watersheds (USGS 1991; and data from the
NOAA National Climatic Data Center, Reston, Virginia). In the Adirondacks, nearby climatic
stations are located at Old Forge, Big Moose, and Stillwater Reservoir. For the period of summer
(July, August and September) 1988 through spring (April, May and June) 1989, precipitation was
75% to 125% of the 30-year normal. Summer 1989 rainfall was 125% to 150% of normal. From
fall (October, November, December) 1989 to spring 1990, precipitation was near or slightly above
normal. Temperatures during summer 1988 were above normal, whereas the winter (January,
February, March) 1988 through summer 1989 period had near normal temperatures. During fall
1989, temperatures were below normal and during winter and spring 1990, temperatures were
above normal.
Slide Mountain and Liberty are climatic stations near the ERP sites in the Gatskills. For summer
1988, fall 1988, winter 1989, and summer 1989, precipitation was 75% to 125% of normal. Pre-
cipitation during spring 1989 was 150% to 200% of normal. Snowfall accumulations during the
winter of 1989 were at a record low. Fall 1989 through spring 1990 had precipitation depths that
were near or slightly above normal. Temperatures for summer 1988, winter 1989, and winter 1990
were above-normal, whereas temperatures during spring 1989 and spring 1990 were near normal.
Both fall 1988 and fall 1989 had below normal temperatures.
87
-------
Phillipsburg is the climatic station closest to the ERP streams in central Pennsylvania. Two
stations, Laurel Mountain and Donegal, are near the southwest Pennsylvania study sites. Both
parts of Pennsylvania experienced similar meteorological conditions during the ERP. Summer
1988 through winter 1989 had precipitation levels that were 75% to 125% of normal. Spring 1989
precipitation levels were 125% to 150% of normal and summer 1989 levels returned to 75% to
125% of normal. Precipitation for fall 1989 through spring 1990 was approximately normal.
Temperatures during winter 1989 and winter 1990 were above normal. Temperatures during all
other seasons were normal or below normal.
5.1.2 Major Hydrologic Events
The ERP was designed to examine the chemical and biological responses of streams to episodic
acidification. The study ideally would have been conducted during a period when snowmelt and
rainstorms generated hydrologic events with normal or above normal flows. In practice, the
hydrologic conditions during the study varied from region to region. In the Adirondacks, winter
conditions allowed significant snowpacks to develop during both winters of the ERP. In each
case, snowpacks > 50 cm developed, with meltwaters contributing to episodes during late winter
and early spring periods. Spring and fall rainstorms also generated hydrologic events. In the
Catskills, very small snowpacks developed during both winters. Most of the episodes were
generated by rainstorms. A similar situation developed in Pennsylvania. No significant
snowpacks accumulated in either of the winters during the ERP. Hydrologic events were
generated almost exclusively by rainstorms. However, a number of the rainstorms were quite
large and generated hydrologic events with large peak flows and relatively long durations.
All Adirondack streams had similar numbers of hydrologic events in response to snowmelt and
rainstorms (Figures 5-11 through 5-4). Fly Pond-outlet, the reference stream, had the smallest
streamflows, and the least difference between the minimum and maximum flows (Tables 5-11 to
5-4). These levels reflect the influence of Fly Pond as the source of the stream. The other
Adirondack streams also reached very low flows (< 0.01 m3/s) during the summer months and
median flow rates were no greater than 0.06 ms/s (Tables 5-1, 5-2, and 5-4). However, these
streams experienced maximum flows of 0.6 to 4.1 m3/s during the snowmelt periods of late winter
and early spring. During the first year of study, the largest hydrologic event occurred during
snowmelt, beginning approximately March 27 and lasting 7 days. During 1990, major hydrologic
Since this section contains so many tables and figures, we have placed them all at the end of the section. Figures
appoar first, beginning on page 109, then tables, beginning on page 161.
88
-------
events occurred during early winter, but the main snowmelt period began the second week of
March, with warming temperatures and rain.
In the Catskill Mountains, the 1989 spring snowmelt was much smaller than normal, with a period
of moderately high discharge beginning February 21 and continuing through a small rainstorm in
early April. The most sustained high flows occurred in late March and early April, a period of high
air temperatures but little precipitation. A high flow period in late March was caused mainly by
melting ice, rather than snowmelt. Snowpack accumulations during winter 1990 were also much
smaller than normal. However, in late January 1990, a melt event did occur. In addition, a small
hydrologic event in March 1990 was influenced by snowmelt. Rainstorms generated a number of
major hydrologic events throughout the year (Figures 5-5 to 5-8). High flows ranged from 1.7 to
41.7 m3/s (Tables 5-5 to 5-8). The East Branch of the Neversink River, because of its large
drainage area, had the highest levels of discharge during hydrologic events (Figure 5-7). How-
ever, much smaller differences in flow existed between the streams during periods of low and
moderate discharge (Tables 5-5 to 5-8).
During both winter periods in the Pennsylvania study areas, snowpack levels were very low and
had minimal influence on hydrologic events recorded during the study. However, major rain-
storms occurred during spring 1989 and to a lesser degree in spring 1990 (Figures 5-9 to 5-13).
Maximum discharge ranged from 2.0 to 8.6 m3/s, and minimum flow rates were from < 0.01 to
0.04 m3/s (Tables 5-9 to 5-13). Linn Run had the greatest streamflows and the flashiest response
to rainstorms (Table 5-11).
5.2 STREAM CHEMISTRY
Tables 5-1 through 5-13 summari2:e the overall maximum and minimum values of major chemical
variables recorded in the ERP streams. In addition, we created a subset of the ERP database that
represents weekly sampling of the ERP streams. The purpose of this activity was to be able to
report the chemical characteristics of the ERP streams in a format similar to that commonly found
in the literature from less intensive (non-episodic) stream chemistry studies. We chose a weekly
sample for each stream that was closest in time to Tuesday at 10 AM (reference time). In all
regions, a sample was available within 1 or 2 days of the designated reference time. Then, we
made statistical summaries for the one-year period within the study (1989: 2/1/89 through
1/31/90) with the most complete weekly data record (Tables 5-1 to 5-13). Throughout the report,
we refer to the values reported from these procedures as weekly statistics (e.g., weekly median).
89
-------
The weekly ANC, pH, and Al data are virtually complete for all streams in all regions. For
Pennsylvania streams, major ion data are missing for several weeks. The missing data are from
the period November 1989 through January 1990, a time with relatively high discharge and epi-
sodes. Therefore, the Pennsylvania weekly statistics are somewhat biased away from the higher
flow regimes. However, the weekly summaries still provide a qualitative picture of the major ion
characteristics of the Pennsylvania streams.
As indicated in Section 3, one or two streams in each region served as references for the bio-
logical studies. The biological reference streams are Fly Pond Outlet in the Adirondacks, Black
Brook and High Falls Brook in the Catskills, and Baldwin Creek and Benner Run in Pennsylvania.
The discussion that follows will show that the water chemistry in the reference streams was
generally favorable for fish and other aquatic biota during the ERP (see Section 6). However, as
discussed in Subsection 5.2.1 , Benner Run experienced occasional acidic episodes of short dura-
tion, but had the best water chemistry of the three northcentral Pennsylvania streams. The
reference streams do not serve a strict reference or control function for ERP hydrochemical
studies. Instead, the reference streams allow ion changes in fairly well-buffered systems to be
examined along with and compared to ion changes in the streams that have acidic, high Al
episodes.
AH streams in the Adirondacks had 1989 weekly median ANC values of at least 10 ^eq/L and
median pH values > 5.3 (Tables 5-1 to 5-4). Weekly median Aljm values ranged from 20 to
165/ig/L. Fly Pond Outlet, the reference stream for biological studies, had the highest pH and
ANC values, and the lowest Alim values. Fly Pond was also the most alkaline of the 13 ERP study
streams. Buck Creek had the lowest pH and ANC values, and the highest Aljm levels. Bald
Mountain Brook and Seventh Lake Inlet weekly median ANC and pH values were only slightly
greater than values for Buck Creek.
In the Catskills, Black Brook and High Falls Brook had similar weekly median ANC (> 95
and pH values (^ 6.6). During the study period, the East Branch Neversink River was a chronic-
ally acidic system with a median ANC of -6 ^eq/L and a median pH of 4.91 (Table 5-7). Not sur-
prisingly, East Branch Aljm values are the highest of the Catskill streams. Linn Run in Penn-
sylvania and Buck Creek in the Adirondacks had weekly median Aljm values that were just slightly
less than the East Branch median. Biscuit Brook had intermediate median pH and ANC levels
compared to the other Catskills study streams (Table 5-5).
90
-------
Two of the Pennsylvania streams v/ere chronically acidic during 1989. The weekly median ANC
values of both Roberts Run and Stone Run were < 0, and their weekly median pH values were
< 5.4 (Tables 5-12 and 5-13). Although Linn Run's weekly median ANC was 30 ^ueq/L and its
weekly median pH was 6.24, it had the highest median Aljm value of any stream in Pennsylvania.
Baldwin Creek and Benner Run were the two reference streams for biological research. Both of
these streams had weekly median ANC levels above 0, but Baldwin Creek had higher levels of pH
and ANC. Neither stream had median Aljm levels above 11 //g/L.
Because episodes occur during periods of high streamflow, a useful way to examine the data is
to compare periods of high streamflow (during episodes) with periods of low flow (not during
episodes or seasons with high flow, e.g. snowmelt). We developed flow duration curves of each
stream using hourly streamflow dsita for the entire study period, and we calculated the 20th
percentile streamflow to represent low-flow conditions and the 95th percentile streamflow to
represent high-flow conditions associated with moderate to large episodes. Then, we sorted the
stream chemistry database by the streamflow at the times water samples were collected, and we
calculated the median values of major chemical variables at streamflows < 20th percentile (low
flow) and > 95th percentile (high flow). For each stream, we compared the discharge distributions
< 20th percentile and > 95th percentile in the discharge database and in the stream
chemistry database and found them to be very similar.
Tables 5-14 and 5-15 summarize the median water chemistry variables for low flow and high flow.
The numbers presented do not attempt to deal with the seasonal variation described in Sub-
sections 5.2.1 and 5.2.2, but they represent an approach that provides a simplified comparison of
chemistry during major episodes (high flow) among streams. With the exception of the East
Branch Neversink River, all the ERP streams had ANC values > 0 and pH and Aljm values suitable
to support fish populations during low flow (see Section 2). In general, the low flow median pH
and ANC were slightly greater than 1989 weekly median pH and ANC (Figures 5-14 and 5-15).
However, low flow Alim concentrations were considerably smaller than the 1989 weekly median
values (Figure 5-16). During high flow, five of the streams had median ANC values < -10 fieq/L,
median pH < 4.8, and Aljm concentrations > 195y«g/L (Table 5-15 and Figures 5-14 to 5-16).
Distinctive regional characteristics are evident in the ERP stream major ion data. The five
Pennsylvania streams had the highest overall SO42~ concentrations, and had four of the five
highest median SO42~ concentrations (Tables 5-9 to 5-15). Benner Run, with the lowest median
SO42~ value of Pennsylvania streams, had the greatest total variation of any ERP stream. The
Adirondack streams had the largest DOC concentrations, the Pennsylvania streams had the
91
-------
lowest concentrations, and the Catskills had intermediate concentrations (Tables 5-1 to 5-15).
The Adirondack and Catskill streams had similar median and maximum concentrations of NO3~.
The NO3~ concentrations of the Pennsylvania streams were much smaller than the NO3~ concen-
trations of the Catskill and Adirondack streams. Although the Pennsylvania streams are located
farthest from the coast, they generally had the highest Cl~ concentrations. Fly Pond Outlet also
had large Cl~ concentrations. The primary study reach of Fly Pond Outlet is located immediately
below a stretch of stream that parallels a paved road, and the stream may have been influenced
by road salt. Calcium was the dominant cation in all ERP streams. As expected, the sum of base
cations for all streams was directly related to ANC.
5.2.1 Episodic ANC, pH, and Aljm
The sampling strategy employed in the ERP allowed most episodes (and virtually all major epi-
sodes) that occurred during the study period to be reasonably well characterized, at least with
regard to worst-case chemical conditions (Figures 5-1 to 5-13). Sometimes the total duration,
especially the recession stages, of hydrologic events were not sampled as intensively as the
rising limb and peak of the storm hydrograph. Also, each region experienced hydrologic events
that were unsampled or had few samples. Nevertheless, the ERP database represents one of the
most intensive temporal stream sampling efforts (related to acidic deposition) to be conducted in
the United States.
Episodes that occurred during the ERP were individually identified. The senior author evaluated
the time series of hydrologic and chemical data, displayed in Figures 5-1 through 5-13 (but on a
much larger time scale), and separated out 19 to 33 episodes per stream that had been sampled
with sufficient intensity to allow the initial conditions and the lowest ANC and pH and the highest
Alfm values to be identified (Figure 5-17). If episodes occurred sequentially, the two episodes
were separately Identified even if the stream chemistry and streamflow had not returned to pre-
episode conditions from the first episode in the series. For most analyses, ANC was used as the
primary chemical variable for establishing the initial and minimum points of the episode;
The ERP streams exhibited a wide range of ANC depressions during episodes (Figure 5-18a).
Aquatic systems with high initial ANC values tended to have larger ANC depressions. However,
the lowest episodic ANC values occurred in streams with initial ANC values of < 50 /*eq/L For
episodes in which pH levels dropped to 5 or below, initial pH values were < 5.8 (Figure 5-18b).
The magnitude of pH changes during episodes was greatest for episodes with mid-range initial
pH values (5.5 to 6.8). Episodic changes of Aljm were highly variable (Figure 5-18c). Estimated
92
-------
ERRATA
Page 92, third paragraph, line 3.
Change "19 to 33 episodes" to "13 to 26 episodes".
-------
-------
increases of 100 to 200 /ng/L were fairly common. Episodes with higher initial Alim concentrations
tended to have the highest maximum Aljm concentrations.
A comparison of the episodic chemical characteristic of streams in the three ERP regions reveals
that Roberts Run, Stone Run, Buck Creek, and the East Branch Neversink River had the lowest
initial episode ANC levels (Figure 5-19). The starting ANC values were often < 0. These streams
also had relatively small ANC changes during episodes, but they reached the lowest absolute
ANC levels of any of the ERP streams (Figure 5-19). Fly Pond Outlet, High Falls Brook, and Black
Brook had the highest initial ANC values, the greatest ANC depressions, and the highest mini-
mum ANC values among the ERP streams (Figure 5-19). Biscuit Brook and Baldwin Creek occa-
sionally experienced acidic episodes. Most of the other ERP streams often experienced acidic
conditions during episodes. The behavior of pH during episodes was similar to that exhibited by
ANC (Figure 5-20). Buck Creek, East Branch Neversink River, Roberts Run, and Stone Run~had
the lowest minimum pH levels, commonly with values < 4.8.
Aluminum chemistry patterns during episodes in ERP streams were somewhat different from
those observed for ANC and pH. Bald Mountain Brook, Buck Creek, Seventh Lake Inlet, East
Branch Neversink River, Linn Run, Roberts Run, and Stone Run all had estimated initial Alim
concentrations > lOO^g/L (Figure 5-21) during some episodes. Of these streams, Linn Run
exhibited the greatest Alim increases during episodes. Buck Creek, East Branch Neversink River,
Linn Run, and Stone Run consistently had the highest estimated maximum Aljm concentrations
(commonly > 350 fig/L) during episodes.
ERP results demonstrate that streams with the lowest ANC levels do not necessarily have the
highest Aljm concentrations. All Adirondack streams have a similar relationship between ANC and
estimated Aljm (Figure 5-22). Some variation in the Aljm-ANC relationship is evident among the
Catskill streams. However, the Pennsylvania streams have very distinctive Aljm-ANC relationships
(Figure 5-22). Linn Run and Stone Run tend to mobilize Al at higher ANC levels than the other
Pennsylvania streams, an indication of differences in substances and processes controlling Al
concentrations among the streams.
Plots of ANC, pH and estimated Aljm at the beginning of episodes and at the time of minimum
episodic ANC illustrate the variance of episodic chemistry responses in the three regions during
the course of the study (Figures 5-23 through 5-35). In all regions, the initial conditions of
episodes have strong influences on the episodic behavior of ANC, pH and Alim.
93
-------
In the Adirondacks, initial ANC and pH tended to be high during the summer months, to decrease
during the fall months, and to be at a minimum during snowmelt of late winter and early spring
(Figures 5-23 to 5-26). Initial Allm concentrations were low in the summer and high during late
winter and early spring. Episodic pH and ANC decreases tended to be greatest in the summer
and early fall, but the lowest pH and ANC occurred during episodes in the spring, winter, and late
fall. Summer had the lowest initial and episodic Aljm concentrations. Episodes with the lowest
minimum ANC and pH generally occurred during March 1989 and 1990 in response to snowmelt
or rain and snowmelt. Buck Creek was an exception, however. Its most severe episode occurred
in response to a rainstorm in November 1989. Snowmelt runoff produced episodes with the
longest durations and with the highest Alim concentrations.
The Catskill streams exhibited less seasonally of initial and minimum episodic ANC and pH,
probably because of small snowpacks and a relatively uniform distribution of hydrologic events
(Figures 5-27 to 5-30). For Biscuit Brook and the East Branch Neversink River, estimated Alim
values were the highest in the winter and spring periods. However, East Branch Alim concentra-
tions were much greater than Biscuit Brook Aljm concentrations. Estimated episodic Aljm concen-
trations in Black Brook and High Falls Brook were generally < 100 fig/L. In Biscuit Brook, the
episodes with the lowest minimum values of ANC and pH and the largest concentrations of esti-
mated Al!fn occurred during episodes generated by spring rainstorms in May 1989 and April 1990
(Figure 5-27). A snowmelt-influenced acidic episode with elevated Aljm concentrations started on
January 25, 1990. A fall episode during September 1989 produced acidic conditions but failed to
generate high Alim concentrations. The East Branch Neversink River, with the exception of a few
fall or late summer episodes had acidic conditions at the beginning of episodes (Figure 5-29).
The lowest minimum ANC and pH levels occurred during episodes in October 1988, September
and October 1989, January 1990, and April 1990. The winter and spring episodes had large
maximum Alim concentrations (> 400 ,«g/L). However, the fall episodes had much lower maxi-
mum Al(m concentrations (< 300 fig/L).
*
Spring and late winter rainstorms produced most of the major episodes in the Pennsylvania
streams during the ERP (Figures 5-31 to 5-35). Also, a few acidic episodes were recorded during
fall periods. However, unlike the Catskill streams, when fall episodes occurred in the Pennsyl-
vania streams, they generated Aljm concentrations similar to those of spring episodes. In Roberts
Run and Stone Run (central Pennsylvania), ANC levels at the beginning of episodes were typically
less than 0. Episodes with very low levels of minimum ANC or pH, or high levels of maximum
Al(m occurred in March, May, June, and November 1989, and January and February 1990
94
-------
(Figures 5-34 and 5-35). The episodes with the lowest minimum ANC values in Roberts Run and
Stone Run occurred in June 1989. Pre-episode ANC values in Linn Run (southwest Pennsylvania)
were usually positive. Linn Run episodes with low pH and ANC levels or high Aljm concentrations
occurred throughout the spring and fall periods (Figure 5-33). Episodes during March and April
1989 had the lowest minimum ANC values.
The study streams can be divided into six ranked classes of chemical severity, based on ANC,
pH, and Alim behavior during episodes and low-flow periods.
1. The East Branch of the Neversink River was chronically acidic (median ANC < 0) and
had strong episodic ANC and pH depressions. It had the longest sustained durations
of severe chemical conditions (low pH and ANC and high Aljm) of the ERP streams.
2. Stone Run and Roberts Run were also chronically acidic, but pH and Aljm levels were
not as extreme as in the East Branch of the Neversink River. These streams experi-
enced severe episodes with low pH and high Aljm. During summer low-flow periods,
ANC levels were positive.
3. Buck Creek and Linn Run were not chronically acidic but had severe acidic episodes
with low pH and ANC and high Aljm for long durations.
4. Bald Mountain Brook and Seventh Lake Inlet had episodes of moderate severity (low
pH and ANC, but moderate Aljm levels) and duration.
5. Biscuit Brook and Benner Run experienced acidic episodes but they were of short
duration with moderate pH levels and relatively low Alim concentrations.
6. Fly Pond Outlet, Black Brook, High Falls Brook, and Baldwin Creek were classified as
nonacidic (with ANC always > 0, except for one brief excursion below 0 in Baldwin
Creek) and had relatively high pH and low Aljm throughout the study period.
The biological significance of these chemical severity classes is discussed in Section 6.
The shaded areas in Figures 5-1 to 5-13 identify the time periods during which brook trout
bioassays were conducted (see Section 6.1). In the Adirondacks, the bioassay experiments
occurred during major episodes. However, no bioassays were conducted during snowmelt
periods, which generally had episodes with the lowest ANC and pH levels and the highest Aljm
values. The Catskill bioassays included some of the most severe episodes measured in the
Catskill streams. Bioassay experiments in Pennsylvania occurred during major episodes, but
episodes with lower pH and ANC levels and higher Aljm levels occurred at other times during the
study.
95
-------
5.2.2 Contributions of Major Ions to Episodic Acidification
With the exception of organic acids (A"), major ions in streamwater were measured by the
methods described in Section 4.3.2. These measurements provided SO42~, NO3~, Cl~, and base
cation data for the ERP streams. However, DOC measurements do not yield information about
the ionic charge of organic acids, which is required to evaluate the role of organic acids in
controlling episodes. Therefore, the ionic charge of A~ was determined indirectly.
In the Adirondacks, the definition of ANC was used as the basis for determining A" (Munson and
Gherini, 1991).
ANC = [HCO31 + 2[CO32~] + [OH! + [other proton acceptors] - [proton donors]
ANC = ZCB - ZCA
(5-1)
(5-2)
where: ZCB = the sum of base cation molarities times their equivalence charge. Base cations
are 2[Ca] + 2[Mg] + [Na] + [K| + [NH4] + X[AI],
where: X = charge of Al.
ZCA = the sum of the strong acid anions times the equivalence charge. Acid anions
are 2[SO4 + [NO3] + [Cl] + Y[A],
where: Y = charge of A.
For each water sample with complete ion chemistry, the following expression was calculated:
DANC =
- ZCA) - Gran ANC-
(5-3)
D
ANC
an estimate of A" for a water sample, because the Gran ANC measurement includes the
Influence of organic acids, whereas the ICB - ZCA term does not include a value for A~. This
approach assumes that DANC is not the result of analytical error and that the charge associated
with Al can be determined. The charge of Al, as represented by X in the ICB definition, was
estimated by the following linear expression (Eq. 5-4), which is based on data presented by
Sullivan et al. (1989):
X = 6.5 - 0.875(pH)
(5-4)
96
-------
For each water sample, the calculated Al charge (X) was applied to the total dissolved Al value to
calculate the equivalent concentration of Al.
For any given water sample, DANC was too variable to provide a reasonable estimate of A".
Therefore, a regression approach was used to develop empirical models, by ERP region, that
estimate A~ (DANC) as a function of DOC and pH in a manner similar to that of Wilkinson et al.
(1992). Because DOC and pH were measured directly, a regression equation based on these
analytes will provide a more stable estimate of A~. The best model for the Adirondack ERP
streams is:
A~ = -33.64 + 6.407(pH) + 15.08(DOC) - 1.653(pH*DOC)
R2 = 0.65
(5-5)
We used Eq. 5-5 to estimate A for the Adirondack streams. We could not develop statistically
significant models for the Catskills (R2 = 0.20) or Pennsylvania (R2 = 0.06). There are at least
two reasons that could explain why a sound empirical model could be developed in the Adiron-
dacks but not in the Catskills or Pennsylvania. First, the concentrations and ranges of DOC are
both much greater in the Adirondack streams. Second, the source of the organics is likely to shift
during the course of episodes because of changes in hydrologic flowpaths. Most empirical rela-
tionships used to estimate organics from anion deficit have been developed for databases of
regional surveys of lakes or streams. Typically, these waterbodies have been sampled during
similar hydrologic conditions and would therefore have organics contributed from fairly uniform
parts of their watersheds. In the Adirondacks, the organics concentrations are higher and the
characteristics of the organics may be more uniform in all hydrologic source areas than in the
Catskills and Pennsylvania.
Consequently, we decided to use a literature-based estimate of organic carbon charge-density to
estimate A~ for Catskill and Pennsylvania streams. We reviewed charge-density values tabulated
by Kahl et al. (1989) and Wilkinson et al. (1992) from studies of a wide variety of streams and
lakes in North America and Europe. We selected a mid-range charge-density value of 5 ^eq/mg
of DOC to estimate A".
A traditional anion deficit approach was not employed because dissolved inorganic carbon (DIG)
was not measured in all regions, and when DIG was measured, the pH - DIG relationship did not
97
-------
uniformly conform to the theoretical, open-atmosphere relationship for these two variables
(Stumm and Morgan, 1981). Therefore, an accurate calculation of [HCO3] could not be made.
Figures 5-36 through 5-48 show the concentrations of major ions and stream discharge at the
beginning of each episode and at the time of minimum ANC (ANCmjn) during the episode. The
time of minimum ANC was used as a reference point because these ion changes are used in
Section 5.2.2.1 to examine the causes of episodic ANC depressions. The reader is advised to
examine the scales of the y-axes carefully because they are not uniform from stream to stream or
from ion to ion. The graphs have been designed to show the maximum amount of ion variation.
In the Adirondacks, SO42~ concentrations do not appear to have consistent seasonal patterns
(Figures 5-36 to 5-39). However, SO42~ is most likely to increase during episodes that occur in
the summer and early fall months. Nitrate does exhibit a strong seasonal pattern in the Adiron-
dack ERP streams. Initial and ANCmin NO3~ concentrations are much greater in the winter and
early spring episodes. Also, NO3~ increased more frequently during episodes in the winter and
early spring than during the remainder of the year. Chloride concentrations and episodic
changes were fairly constant during the course of the study. Concentrations of A~ were much
greater in the Adirondack streams than in the Catskill or Pennsylvania streams. Typically, A~
increased during episodes throughout the year, but the greatest increases tended to occur during
episodes in the late summer and fall. Base cations usually decreased during episodes through-
out the year, with the largest initial concentrations and changes occurring in the summer and
early fall. Fly Pond Outlet experienced the greatest decreases in base cations (Figure 5-38).
Similar to the Adirondacks, SO42~ concentrations and changes in Catskill episodes did not exhibit
much seasonality (Figures 5-40 to 5-43). Unlike the Adirondacks, the Catskill streams frequently
exhibited large SO42~ decreases during episodes. Nitrate concentrations routinely increased
during episodes in all streams except High Falls Brook (Figure 5-43). Episodic NO3~ increases in
High Falls Brook were concentrated in the winter and spring. Initial and ANCmin NO3~ concentra-
tions were somewhat greater during the winter and spring than in other seasons. Although A~
increased during episodes year round, the most dramatic increases tended to occur in late
summer and fall. Chloride tended to decrease slightly during episodes, with Black Brook having
the largest CF decreases (Figure 5-41). Base cations also decreased during most, but not all,
episodes in the Catskill streams. As expected, Black Brook and High Falls Brook had the largest
base cation decreases during episodes (Figures 5-41 and 5-43).
98
-------
Pennsylvania streams had the highest SO42 concentrations (Figures 5-44 to 5-48). All the
Pennsylvania streams had SO42~ increases, some quite large, during episodes. Baldwin Creek
(fall and early spring), Linn Run (fall and early spring), and Benner Run (throughout the year) had
the largest SO42~ pulses (Figures 5-44 to 5-46). However, strong seasonal SO42~ patterns were
not evident in any of the Pennsylvania streams. Nitrate concentrations in the Pennsylvania
streams were much smaller than those in the Catskill or Adirondack streams, and the central
Pennsylvania streams (Benner Run, Roberts Run, and Stone Run) had lower concentrations of
NO3~than did the southwest Pennsylvania streams (Baldwin Creek and Linn Run). In central
Pennsylvania, Roberts Run and Stone Run had a mixture of small NO3~ increases and decreases
during episodes, whereas Benner Run consistently had NO3~ decreases. In southwest Pennsyl-
vania, NO3~ decreases were common during episodes in both Linn Run and Baldwin Creek
(Figures 5-46 and 5-44). Concentrations of A~ typically increased during Pennsylvania episodes.
However, A~ concentrations and changes did not exhibit any seasonal patterns during the study.
Although both Cl~ increases and decreases occurred during episodes in the Pennsylvania
streams, decreases were more common. Benner Run consistently showed some fairly large
decreases in Cl~ during both fall and spring (Figure 5-45). As a rule, changes in base cations
during episodes followed the typical pattern of decreasing during episodes in all streams.
However, some notable exceptions occurred. Base cations increased fairly often during episodes
in Baldwin Creek, Benner Run, and Stone Run. Linn Run, which most consistently had base
cation decreases during episodes, had the greatest decreases during late spring (Figure 5-46).
5.2.2.1 Ion Changes
To evaluate the role of ion changes on episodic ANC depressions, we once again depend on the
CBCA definition of ANC (see Eq. 5-2). From Eq. 5-2, we can develop the following expression:
AANC = ASCB - AZCA.
(5-6)
For the ion change analyses, we slightly modified CB by excluding Al, which allows a separate
evaluation of the effects of traditional base cations (Ca, Mg, Na, K, NH4) and Al. Consequently,
we can expand Eq. 5-6 to the following:
AANC = A2CB + AAI - ASO4 - ANO3 - ACI - AA
(5-7)
99
-------
We used Eq. 5-7 as the basis for ranking the importance of ion changes for each episode. All
computations were performed for the greatest AANC of an episode. Ion changes were calculated
in equivalents. Inspection of Figures 5-36 to 5-48 reveals that ion changes can contribute to ANC
depressions (-AANC) or reduce ANC depressions. Decreases of base cations (-A2CB) or Al
(-AA1) and increases of acid anions (+ASO42~ +ANO3~ +ACF, +AA~) contribute to ANC, depres-
sions. For each episode, all of the ion changes were evaluated to determine if they contributed to
the ANC depression. All ion changes that contributed to episodic ANC depression were tabulated
and ranked. The largest magnitude ion change was assigned a rank of 1 (most important con-
tributor to the ANC depression). The second largest ion change was assigned a rank of 2, and
so forth.
Regional and individual stream characteristics emerge upon examination of the number of
episodes for which ion changes contributed to ANC depressions and the mean rank of ion
change contributions to ANC depressions (Table 5-16). In the Adirondacks, base cation
decreases and A~ increases most consistently contributed to ANC depressions during episodes.
For Fly Pond Outlet, base cation decreases were consistently the most important contributor to
ANC depressions (for episodes with sufficient data to allow ion change analysis). Base cation.
decreases were also the highest mean contributor to episodes in the other Adirondack streams
except Seventh Lake Inlet. On average, A" increases were the most important ion changes in
Seventh Lake Inlet and were the second most important ion changes to ANC depressions in the
other Adirondack streams. Nitrate increases were also very important contributors to episodes in
Adirondack streams, with mean ranks just slightly less than A~ increases.
In the Catskills, base cation decreases, NO3~ increases and A~ increases were all consistently
important contributors to episodic ANC depressions (Table 5-16). On average, With the exception
of the East Branch Neversink River, base cation decreases were the highest ranked contributors,
NO3~ increases were the second most important contributor, and A~ increases were the third
most important ion changes. Decreases in base cations were especially important in Black Brook
and High Falls Brook. For the East Branch Neversink River, base cation decreases were the most
inportant ton changes and A~ increases were the second most important ion changes to ANC
depressions. Sulfate increases rarely contributed to episodic ANC depressions in the Catskill
streams. When SO42~ increases did contribute to ANC depressions, they were typically the
second or third mean ranked ion change.
100
-------
In the Pennsylvania streams, SO42 increases and base cation decreases frequently made strong
mean contributions to episodic ANC depressions (Table 5-16). In Linn Run and Roberts Run,
base cation decreases were consistently the most important ion change contributing to episodic
ANC depressions. However, in Baldwin Creek, Benner Run, and Stone Run, SO42~ increases
contributed to ANC depressions in more episodes than did base cation-decreases. Organic acids
(A~), on average, were the third most important contributor to ANC depressions in all Pennsyl-
vania streams. Nitrate and Cl~ increases occurred less frequently than did SO42~ increases, base
cation decreases, or A~ increases and generally made less important contributions to ANC
depressions in the Pennsylvania streams.
In all three regions, Al typically increased during episodes and reduced ANC depressions.
However, during a few episodes in some of the streams, especially in Pennsylvania, Al concen-
trations did decrease and make modest contributions to ANC depressions (Table 5-16).
An overview of ion changes is useful in developing a general understanding of the variance of
episodic chemistry among the ERP streams. However, we are not equally interested in all
episodes. We are most interested in those episodes that result in conditions harmful to aquatic
biota. Because we are focusing on ANC in this section, we should be most interested in the
severe episodes (lowest ANCmin) that result in acidic conditions for relatively long periods of time.
In the Adirondacks, these type of episodes usually occurred during snowmelt periods. In Penn-
sylvania and the Catskills, large rainstorms induced episodes with the lowest ANC values.
For most streams in the Adirondacks, the episodes with the lowest ANC and pH values and high-
est Al concentrations occurred during the two snowmelt periods (1989 and 1990) and during
spring rainstorms (Table 5-17). A rainstorm in November 1989 produced the most severe episode
in Buck Creek. Fall rainstorms also produced major episodes in Bald Mountain Brook and
Seventh Lake Inlet. Chemical behavior generally was similar among the Adirondack streams for
the snowmelt episodes (Table 5-17). Sulfate and Cl~ concentrations decreased, reducing ANC
depressions; base cation decreases were the first ranked contributors to ANC depressions; A~
and NO3~ increases positively contributed to episodes and were second or third ranked (Table
5-17; Figures 5-36 to 5-39). During one snowmelt-driven episode in Buck Creek, NO3~ increases
were the most important ion change.
The ion changes controlling ANC depressions were more variable in Adirondack study streams
during major rain-driven episodes than during major snowmelt episodes (Table 5-17). For Bald
101
-------
Mountain Brook, base cation decreases were the most important ion changes during the two
major rain-driven episodes, which occurred during fall 1989. Sulfate and A" were the second or
third ranked ion changes, but NO3~ increases did not contribute to the ANC depressions of these
episodes. Snowmelt was the driving force behind all the major Fly Pond Outlet episodes. In
Buck Creek and Seventh Lake Inlet, A~ increases made the most important contributions to ANC
depressions during major rain-driven episodes, regardless of whether they occurred in the spring
or fall. Base cations decreases or NO3~ increases were the second or third most important ion
change during these episodes.
Unlike the Adirondacks, most of the major episodes in the Catskills streams occurred in response
to spring or fall rainstorms (Table 5-18). Sulfate and Cl~ decreases reduced ANC depressions
during the most episodes in Catskill streams (Figures 5-41 and 5-43). During the major episodes
in High Falls Brook and Black Brook, base cation decreases were the most important contributors
and NO3~ increases (once an A~ increase) were the second ranked contributors to episodic ANC
depressions (Table 5-18). Increases in A~ were consistently the third ranked ion change in High
Falls Brook and Black Brook.
For Biscuit Brook, NO3~ increases were the most important ion change in two of three of the
major spring episodes (Table 5-18). Base cation decreases or an A~ increase were the second
most important ion change. In the third spring episode, base cation decreases were the most
important ion change; a NO3~ increase was the second ranked ion change. Nitrate was also the
most important contributor, followed by base cation decreases, to the major spring episode that
occurred in the East Branch Neversink River.
During the major fall episodes in the East Branch Neversink River, A~ pulses were the most
important contributors to ANC depressions (Table 5-18). Base cation decreases were second
ranked and NO3~ or SO42~ increases were the third ranked ion changes. Similarly, an A~ increase
was the most important ion change during the major fall episode measured in Biscuit Brook
(Table 5-18). However, a NO3~ pulse was the second most important ion change.
The ionic responses of the Catskill streams were fairly uniform during the January 1990 snowmelt
influenced episodes (Table 5-18). For all streams except Black Brook, NO3~ pulses were the first
ranked and A~ increases or base cation decreases were the second ranked contributors to ANC
depressions.
102
-------
All the major episodes in Pennsylvania streams were generated by rainstorms in the winter and
spring (Table 5-19). Base cation decreases were usually the most important ion change for major
episodes in Roberts Run and Linn Run. In Roberts Run, A~ pulses tended to be the second
ranked ion change. In episodes in Linn Run, the second ranked ion changes were increases in
NO3~, pulses of SO42~, or decreases in base cations. During one 1990 episode in Linn Run, a
SO42~ increase was the first ranked ion change controlling the episodic ANC depression. Base
cation decreases commonly were the most important ion change in Baldwin Creek episodes.
However, SO42~ and A~ pulses were each the first ranked ion change in a major episode. Sulfate
pulses were commonly the most important ion change during major episodes in Benner Run and
Stone Run; base cation decreases or A~ pulses were also the most important ion change in one
or more episodes.
Major episodes in the Pennsylvania streams often had fewer ion changes contributing to ANC
depressions than did the major episodes in the Catskills or Adirondacks (Tables 5-17 to 5-19).
Whereas the Catskill and Adirondack streams rarely had fewer than three ion changes that
contributed to ANC depressions, 40% of the major episodes in Pennsylvania had only two (or in
one instance, one) ion changes that controlled ANC depressions. Benner Run had the greatest
number of episodes with only two ion changes contributing to ANC depressions. In each case
SO42~ pulses were the most important ion change and A~ increases were the second ranked ion
change. During one major episode in Linn Run, base cation decreases, alone, controlled the
ANC depression.
Another way to examine the importance of ion changes is to look for patterns of ion change in
relationship to the minimum ANC (ANCmin) of episodes. In the Adirondacks, SO42~ tends to
decrease during acidic episodes, whereas NO3~ and A~ increases are likely to be the greatest
during acidic episodes (Figure 5-49). In the Catskills, SO42~ tends to decrease in episodes with a
wide range of episodic ANCmin levels (Figure 5-50). Nitrate increases are greatest in acidic or low
ANC episodes. Organic acid changes may also be somewhat greater in low ANC or acidic epi-
sodes in the Catskills. No obvious relationships exist between SO42~ changes and ANCmin, NO3~
changes and ANCmjn, or A~ changes and ANCmin in Pennsylvania (Figure 5-51). In the Adiron-
dacks and the Catskills, base cation decreases are smaller for acidic episodes than nonacidic
episodes. However, the Pennsylvania streams experienced much larger base cation decreases
during acidic episodes than did the Adirondack and Catskill streams (Figures 5-49 to 5-51).
103
-------
5.2.2.2 Ion Concentrations
In Subjection 5.2.2.1, we explore the role that ion changes played in controlling episodes in the
ERP streams. As can be seen from the data, these changes are very important in determining
episodic acid-base chemistry. However, the ion changes must be linked with the absolute, con-
centrations of ions to obtain a complete picture of the ionic control of minimum episodic ANC. In
all three regions, SO42~ typically had the greatest concentrations of any of the acid anions at all
times during episodes (Figure 5-52). Based on a CBCA (Eq. 5-2) definition of ANC, SO42~ is a
major contributor to the ANCmjn of a stream or lake, even if it remains unchanged or decreases
during episodes. Sulfate is especially dominant in the Pennsylvania streams.
The role of NO3~ in episodic chemistry is another example of the importance of examining both
absolute concentrations of ions and ion changes. In Pennsylvania, NO3~ pulses can be the
second most important ion change during episodes (e.g., Linn Run). However, these changes
are an order of magnitude smaller than the base cation decreases that are typically the most
important ion change (Figure 5-46). Whether at the beginning or at the ANCmin of episodes, the
NO3~ concentrations in Linn Run or the other Pennsylvania streams are much smaller than in the
Adirondack and Catskill streams (Figure 5-52). In the Catskill streams, both NO3~ increases and
absolute concentrations are important (Figure 5-52; Tables 5-16 and 5-18). The NO3~ concentra-
tions are sufficiently great at the beginning and at ANCmin to make a significant impact on ANC.
In the Adirondacks, NO3~ pulses are not typically the primary control of episodic ANC, but the
NO3~ concentrations, because of their relatively high absolute levels, depressed ANC values at
the beginning and during the course of episodes (Figure 5-52; Tables 5-16 and 5-17). Even
though A~ typically had somewhat greater increases during episodes in Adirondack streams,
NO3~ at times had a greater effect on minimum ANC because of its greater concentrations. This
was especially true during late winter and early Spring.
In the Adirondacks and Catskills, Cl~ concentrations were generally the smallest of the major ions
(Figure 5-52). In the Pennsylvania streams, however, Cl~ concentrations in general were greater
than NO3~ or A~ concentrations at the beginning of episodes and at ANCmjn. Therefore, Cl~ may
have exerted an influence on the acid-basis status of the water chemistry of the Pennsylvania
study streams.
As previously stated, A~ concentrations in the Catskill and Pennsylvania streams were generally
low (Figure 5-52). In these two areas, the influence of A~ on ANCmjn was primarily because of A~
104
-------
changes during episodes. In the Adirondack streams, however, A concentrations were great
enough for both absolute concentrations and for ion changes to have an important influence on
ANCmin.
5.2.2.3 Role of Acidic Deposition
Evaluation of ion changes and concentrations, as presented in Section 5.2.2, is a useful way to
understand processes controlling episodes. Data presented in the previous subsection show that
episodic ANC depressions are a result of complex interactions of multiple ions. However, there is
the question of what these ionic behaviors mean with regard to the effects of atmospheric deposi-
tion. Organic acids are one common natural source of acidity during episodes. Although base
cation decreases are generally viewed as largely a natural process, acidic deposition can
influence the magnitude of the effect. During hydrological events, and associated episodes, the
source of streamflow changes from groundwater, which is dominant during baseflow periods, to
surface soils and other shallow sources (Wigington et al., 1990; Swistock et al., 1989). Base
cation decreases are a manifestation of the differences in cation concentrations of the major
sources of streamflow before and during episodes. Because acidic deposition can contribute to
the depletion of base cations in surface soils more rapidly than deeper soils, it can, accentuate
base cation decreases that occur during episodes (Wigington et al., 1990).
In the ERP watersheds, there is no evidence of internal sources of SO42~. Therefore, SO42~ in the
streams is predominantly, if not solely, a consequence of atmospheric deposition. Nitrate in the
ERP watersheds probably is derived from a combination of natural N cycling and atmospheric
deposition (Aber et al., 1989, 1991). Significant NO3~ pulses from forested watersheds are very
unlikely without large deposition of atmospheric N (Stoddard, in press). Watersheds in the
Adirondacks and Catskills, in which stream water concentrations of NO3~ are quite high, have in
all likelihood been significantly affected by atmospheric deposition of N. In the remainder of this
section and the report, we ascribe all SO42~ effects and a portion of the NO3~ effects to acidic
deposition. Given these assumptions, we see evidence of both natural processes and acidic
deposition making important contributions to episodic acidification in the Adirondack, Catskiil, and
Pennsylvania study streams.
For all episodes recorded (large and small) in the three ERP regions, base cation decreases were
commonly the most important ion change contributing to ANC depressions. In the Catskills and
Adirondacks, the largest base cation decreases occurred during nonacidic episodes. Base cation
105
-------
decreases were especially dominant for episodes in circumneutral streams in the study (Black
Brook, Fly Pond Outlet, and High Falls Brook). However, some base cation decreases in the
Pennsylvania streams were very large even during acidic episodes.
Organic acid concentrations were greatest in the Adirondack streams, and A~ increases during
episodes were commonly the second and sometimes the first ranked ion changes during major
episodes. Organic acid contributions tended to be greatest in episodes with acidic or low ANCmin
values. Organic acids were generally less important in Pennsylvania and Catskill episodes.
However, A~ increases were the most important ion changes during several fall episodes in two
Catskill streams. Organic acid pulses were commonly the second most important, and occasion-
ally the most important, ion changes during episodes in the Pennsylvania streams, even though
the magnitudes of A~ changes were small.
Sulfate was the dominant acid anion (C^ virtually at all times in the study streams in all three
regions. For both acidic and nonacidic episodes in the Catskills, SO42~ concentrations tended to
decrease, thereby reducing ANC depressions. In the Adirondacks, SO42~ concentrations were
observed to increase or decrease during episodes. However, SO42~ concentrations tended to
decrease during acidic episodes. Sulfate concentrations in the Pennsylvania streams were
greater than in Catskill or Adirondack streams. Based on the CBCA ANC definition, these large
SO42~ concentrations (in all regions) provided an acidity base upon which other ion changes
further modified the acid-base status of the stream waters. Another major difference between the
Pennsylvania streams and the Adirondack and Catskill streams is the importance of SO42~ pulses
in Pennsylvania episodes. In some major episodes, SO42~ pulses were the most important ion
change contributing to ANC depressions in three of the five Pennsylvania streams and were con-
sistently the most important fon change in two of the streams.
Nitrate behavior and contributions to episodes were different in each of the three regions. Nitrate
concentrations were much greater in the Catskill and Adirondack streams than in Pennsylvania
streams. In the Catskills, NO3~ concentrations typically increased during episodes, particularly
during acidic episodes. In two of the Catskilt streams, Biscuit Brook and East Branch Neversink
River, NO3~ was the most important ion change contributing to ANC depressions during several
episodes. In the Adirondacks, NO3~ increases were typically the third ranked fon change during
episodes. The largest NO3~ increases occurred during acidic episodes. Even though episodic
changes of NO3~ were relatively small, NO3~ concentrations were large throughout Adirondack
106
-------
episodes. Nitrate pulses in Pennsylvania streams were generally small and made little
contribution to ANC depressions.
Dow (1992) calculated S and N input-output budgets for the Pennsylvania streams. Four of the
five watersheds exported 66-78% of S received as wet and dry deposition. The fifth basin,
Benner Run, exported only 30% of deposited S. Dow attributed this phenomenon to the Benner
Run watershed soils having higher SO42~ adsorption capacities than the other Pennsylvania
watersheds. Benner Run is also the watershed in which SO42~ pulses during episodes were most
frequently important contributors to ANC depressions. The Pennsylvania watersheds exported
only 2-15% of the N deposited, with the southwest streams having the greatest NO3~ export, a
finding supporting the general lack of large NO3~ pulses during episodes in Pennsylvania.
Stoddard (in press) has proposed stages to describe the progression of nitrogen saturation in
watersheds based on temporal patterns of NO3~ leaching in surface waters. Stage 0 is one of
very low, or immeasurable, concentrations during most of the year, and measurable concentra-
tions only during snowmelt or during spring rain storms. The Pennsylvania watersheds are
probably in Stage 0 of this scheme. At Stage 1, the seasonal pattern typical of Stage 0 water-
sheds is amplified. In Stage 2, NO3~ concentrations increase during periods of baseflow and the
seasonal cycle of NO3~ is dampened. The Catskill ERP streams may be in this stage. The ERP
streams in the Adirondacks may be approaching the mid to later phases of Stage 2. In Stage 3,
the watershed becomes a net source of N rather than a sink. Watersheds in Stage 3 have very
high NO3~ concentrations and lack a coherent seasonal pattern in NO3~ concentrations.
Other ion changes were less important to episodes measured in the ERP. Chloride changes and
concentrations usually had relatively little impact on episodes. However, for a few episodes in
some Pennsylvania streams, Cl~ increases were the second ranked ion change. In virtually all
episodes, especially those that were acidic, Al increases tended to buffer ANC depressions.
Ion behavior that controlled ANC depressions during the five major episodes recorded in each
stream during the study period was not necessarily the same as the ion changes most important
to smaller, more frequent episodes. In the Adirondacks, the only region to have major snow-
packs, the most severe episodes generally occurred during spring snowmelt. During these epi-
sodes, base cation decreases were usually the most important contributions to ANC depressions,
and A~ and NO3~ changes positively contributed to episodes and were of similar rank (2 or 3).
Buck Creek experienced one major snowmelt episode in which an increase in NO3~ was the most
107
-------
important ion change. The Catskills had one episode influenced by snowmelt in which NO3~
pulses were the first ranked contributor and base cation decreases were the second ranked ion
change for three of the four streams.
Within regions, the major episodes generated by rainstorms had more variable ionic controls than
did snowmelt episodes. The large rain-induced Adirondack episodes occurred in the spring and
fall. In these episodes, increases in A~ and decreases in base cations were the most important
ion changes to ANC depressions. In major spring episodes in the Catskills, decreases in base
cations or increases in NO3" were typically the first or second ranked ion-changes contributing to
ANC depressions. For a series of Catskill streams (including the ERP streams) that had been
monitored for a number of years, Murdoch and Stoddard (1993) concluded that decreases in
ANC are controlled primarily by dilution of base cations during the fall and early summer and by
increases in NO3~ during the initial spring snowmelt. For the major episodes in Pennsylvania,
base cation decreases and SO42~ increases were usually the most important or second most
important ion changes contributing to ANC depressions. Stone Run and Baldwin Creek both
experienced major rain-induced episodes in which A" increases were the most important ion
changes. In Linn Run, NO3" pulses were the second ranked ion change in two major episodes.
Atmospheric deposition of SO42~ and NO3~, as evidenced by stream water SO42~ and NO3~
during episodes, has contributed significantly to the occurrence of acidic episodes with low pH
and high Al levels in all three study areas. Base cation decreases are often the most important
ion change that occurs during minor and major episodes. However, base cation decreases alone
cannot create acidic stream water conditions during episodes. Organic acid pulses are also
important contributors to ANC depressions in the Adirondack streams and, to a lesser extent, in
the Catskill and Pennsylvania streams. In all three study areas, SO42~ or NO3~ pulses during
episodes augment the natural processes to create episodes with lower pH and ANC and higher Al
concentrations than would have occurred from natural processes alone. Furthermore, the large
baseline concentrations of SO42~ and NO3~ reduce episodic minimum ANC, even when these
ions do not change during episodes.
108
-------
1.6 -
-« 1-2'
"1 -8H
0 -4H
AM, •,
VvJAv^_ .1 a JllL^JLuWL.^.
luLA
-(a)
100
200 300
400
500
600
700
600
100
X
a.
7
6.5 -
6 -
5.5 -
5 -
4.5
100
200 300
400
500.
600
-(C)
700
-40
100
200
300
400
500
600
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-1. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for Bald
Mountain Brook, Adirondacks. Shaded areas designate periods when brook
trout bioassays were conducted.
109
-------
2 -
•B 1 "I
S" .SH
100
200
300
400
500
600
T—'
700
(a)
100 200
300
400
500
600
700
6.75 -
6.25 -
X 5.75 =
5.25 -
4.75 -
4.25
100
200 300
400
500
600
700
O
100
200 300
400
500
600
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-2. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for Buck
Creek, Adirondacks. Shaded areas designate periods when brook trout
bioassays were conducted.
110
-------
-0)
"I
a
.2
.15 -
.1 -
.05 -
100
-(a)
200 300
400
500
600
I ' '
700
600
500 :
400 -
300 -
200 -
100 -
-(b)
O^M
100
200 300
400
500
600
700
Q.
7.4 -
7 -
6.6 -
6.2 -
5.8 -
5.4
0
100
200
300
400
500
600
700
500
2" 400 -
8" 300 -
o
200 -
100 -
0
100
-(d)
200 300
400
500
600
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-3. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for Fly Pond
Outlet, Adirondack*;. Shaded areas designate periods when brook trout
bioassays were conducted.
111
-------
W
a
4
3 -
2 -
1 -
0
•"I"—"I 1 1
100
200
400
600
-(a)
700
100
200 300
400
500
600
700
X
Q.
7
6.5 -
6 -
5.5 -
5 -
4.5
100
I | 1 t I I | I I I I J I I T- I | I I I |- II 1 I | I
200 300 400 500 600 700
-(C)
o
100
200 300
400 .500
600
-(d)
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-4. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for Seventh
Lake Inlet, Adirondacks. Shaded areas designate periods when brook trout
bioassays were conducted.
112
-------
40
H
a
4 -
3 -
2 -
1 -
100
200
300
400
500
600
-(a)
700
200
150 -
100 -
50 -
(b)
0
100
200
300
400
500
600
700
6.8
6.4 -
6 -
5.6 -
5.2 -
4.8
-(C)
100
200
300
400
500
600
700
o
60
50 -
40 -
30 -
20 -
10 -
0 -
-10
100
-(d)
200
300
400
500
600
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-5. (a) Mean daily discharge, (b) estimated Aijm, (c) pH, and (d) ANC for Biscuit
Brook, Catskills. Shaded areas designate periods when brook trout bioassays
were conducted.
113
-------
o
100
200 300
400
500
600
700
100
200
300
400
500
600
700
100
200 300
400
500
600
700
O
160 -
120 -
80 -
40 -
100
200 300
400
500
-(d)
600
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-6. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for Black
Brook, Catskills. Shaded areas designate periods when brook trout bioassays
were conducted.
114
-------
JOT
a
35
30 -
25 -
20 -
15 -
10 -
5 -
100
200
-(a)
:•
uLujwA^
300
400
500
600
700
500
100
200
300
400
500
600
700
5.6
5.4 -
5.2 -
' 5 -
4.8 -
4.6 -
4.4 -
4.2
100
200
300
400
500
600
700
O
20
10 H
0 --•
-10 -
-20 -
-30 -
-40 -
-50
0
100
200
300
400
500
-(d)
600
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-7. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for East
Branch Neversink River, Cats kills. Shaded areas designate periods when
brook trout bioassays were conducted.
115
-------
1.6 -
0 .4-
100
200
300
400
500
600
700
-(a)
300 -
.-, 250 -
"B) 200-
•3= 150 -
•i 100-
< 50-
n -
^ f
"
X
Li .^^/l^/jfefjM »J „.*
*$rl "^.
';» 1 ' •S'l ^^y
0 '*'•'*. ! "* '*•>.<•'•
, '
'
m « * ? »
\VL,&4_~, j^SHs^L.a,
100
200 300
700
(b)
7.2
6.8 -
6 -
5.6 -
5.2
-(C)
100
200
300
400
500
600
700
O
100
200 300
400
500
600
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-8. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for High Falls
Brook, Catskills. Shaded areas designate periods when brook trout bioassays
were conducted.
116 .
-------
"E
1.6 -
1.2 -
.8 -
.4 -
100
200 300
400
500
600
-(a)
700
100
Q- 80-
§ 60 -
g 40 -
< 20 H
100
200 300
400
500
600
(b)
700
X
Q.
7.2
6.8 -
6.4 -
6 -
5.6 -
5.2
100
200
300
400
500
600
-(C)
700
100
^ 80 -
"o- 60 -
|t 40 -
O 20 -
-20
100
200 300
400
500
600
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-9. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for Baldwin
Creek, Pennsylvania. Shaded areas designate periods when brook trout
bioassays were conducted.
117
-------
1.4
1.2-
c| .8-
o
.6 -
.4 -
.2 -
100
200
300
400
500
600
700
-(a)
200
6.4 -
6 -
5,6 -
5.2 -
4.8
-(C)
100
200 300
400
500
600
700
O
-20
100
200 300
400
500
600
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-10. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for Benner
Run, Pennsylvania. Shaded areas designate periods when brook trout
bioassays were conducted.
118
-------
O
; 100
200
300
400
500
600
700
100
200 300
Q.
7
6.5 -
6 -
5.5 -
5 -
4.5
100
200 300
400
500
600
700
CT
1
O
-40
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-11. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for Linn
Run, Pennsylvania. Shaded areas designate periods when brook trout
bioassays were conducted.
119
-------
.
n
O
2
1.6 -
1.2 -
.8 -
.4 -
100
200
300
400
500
-'(a)
600
700
100
200
300
400
500
600
700
Q.
6.4 -
6 -
5.6 -
5.2 -
4.8 -
4.4
100
200
300
400
500
-(C)
600
700
1
o
<
60
40 -
20 -
0 -••
-20 -
-40 -
-60
100
200
300
400
500
600
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-12. (a) Mean daily discharge, (b) estimated Alim, (c) pH, and (d) ANC for Roberts
Run, Pennsylvania. Shaded areas designate periods when brook trout
bioassays were conducted.
120
-------
c-5-
O
2.5
2 -
1 -
.5 -
-r-r-*
100
200
300
400
500
600
(a)
700
700
D)
100
200
300
400
500
600
700
6.4
6 -
5.6 -
5.2 -
4.8 -
4.4
-(C)
0
100
200
300
400
500
600
700
CT
o>
O
30
10 -
-10 -
-30 -
-50
100
200 300
400
500
600
-(d)
700
10/1/88 1/1/89 4/1/89 7/1/89 10/1/89 1/1/90 4/1/90
Date and Days Since Project Initiation
Figure 5-13. (a) Mean daily discharge, (b) estimated Aljm, (c) pH, and (d) ANC for Stone
Run, Pennsylvania. Shaded areas designate periods when brook trout
bioassays were conducted.
121
-------
300
O
CM
VI
O
O
eg
"•a
CD
200 -
150 -
100 -
50 -
0 -
-50
O Adirondacks
D Catskills
A Pennsylvania
(a)
-i | i | i | i | * | i | i—
-50 0 50 100 150 200 250 300
1989 Weekly Median ANC
250
lO
CD
AI
O
200 -
150 -
100 -
CD
O 50 -
c
eg
T3
CD
o -
-50
O Adirondacks
D Catskills
A Pennsylvania
(b)
-50 0 50 100 150 200 250
1989 Weekly Median ANC
Figure 5-14. (a) Comparison of 1989 weekly median ANC (^eq/L) and median ANC for
discharge < 20th percentile, and (b) comparison of weekly median ANC with
median ANC for discharge > 95th percentile for all ERP streams.
122
-------
V\
7.5
6.5 -\
6H
.0
Q. 5.5 ^
4.5 H
4
O Adirondacks
D Catskills
A Pennsylvania
4.5 5 5.5 6 6.5 7
1989 Weekly Median pH
(a)
7.5
7.5
7 -
"5 R * -
95th percentile for all ERP streams.
123
-------
S
CM
VI
O
i
T3
Q
200
175 -
150 -
125 -
100 -
75 -
50 -
25 -
O Adirondacks
D Catskills
-A Pennsylvania
O
D
A
(a)
0 25 50 75 100 125 150 175 200
1989 Weekly Median Aljm ftig/L)
.G 400 -
ur>
CJ
Al
9* 300 -
200 -
c 100 -\
OJ
T3
CD
O Adirondacks
D Catskills
A Pennsylvania
(b)
0 100 200 300 400
1989 Weekly Median Aljm ftig/L)
Figure 5-16. (a) Comparison of 1989 weekly median estimated Alim and median estimated
AIim for discharge < 20th percentile, and (b) comparison of weekly median
estimated Aljm with median estimated Alim for discharge > 95th percentile
for all ERP streams.
124
-------
t
o
<
Initial ANC
Single Episode
Last ANC
measurement
of episode
Minimum ANC
(a)
Time
t
o
<
Episode 1
Back-to-Back Episodes
Episode 2
(b)
Time
Figure 5-17. Idealized diagram of (a) a single episode and (b) back-to-back episodes to
illustrate the approach used to identify episodes and their characteristics.
125
-------
400
7.5
to
05
-100
-100 0 1dO 200 3(JO 400
Initial ANG ftieq/L)
7 -
oo
I I ' I ~I ' I ' i ' I '
4 4.5 5 5;5 6 6.5 7 7.5
Initial pH
3
E
S
0 200 400 600 800
initial Ai,m(Mfl/L)
Figure 5-18.
(a) Initial and minimum
in all ERP streams.
(b) initial and minimum pH, and (c) initial and maximum estimated Alim for episodes
-------
400
(a)
Figure 5-19. Box plots of (a) initial ANC, (b) minimum ANC, and (c) ANC change, for
episodes in ERP streams. Line in box indicates median; upper and lower
borders of box show 25th and 75th quartiles; whiskers indicate 10th and
90th percentiles; circles represent observations beyond 10th and 90th
percentiles.
127
-------
CL
/.o -
7-
65 -
b -
O.b
fa-
4.5 -
A -
f
J_
O
A
4 0 ° fitf lilipl j1
f JL JL * i iii
C- o °
i °
" i
o J
o
o
JLr1! 8 °
........ u o
5 .nut.....
1 -^ •'
O
(a)
CL
E
1
7.5
7
6.5
6
5.5
5
4.5
4
(b)
CD
P>
o
u.o -
-0.5 -
-1.5 -
-2-
-2.5 -
•s'J
?
o
o
0
m
<£
73
To
I
1 I
^ *-
ET,T,7T^ i::.......f
> J_ £ — !-• — 1
'•*•"!"•" o
+-• vv ^ \^ >^ V f~ C^ ^
n\ jc ^f- ^c ^c _x. c ^— ^ c
— OOCO m co
-J g tfl ;
-C .» m
5 m 0 S c w "J
J a> 2
-d en
0
DC
(c)
m
0)
Figure 5-20. Box plots of (a) initial pH, (b) minimum pH, and (c) pH change, for episodes
in ERP streams. Line in box indicates median; upper and lower borders of
box show 25th and 75th quartiles; whiskers indicate 10th and 90th
percentiles; circles represent observations beyond 10th and 90th
percentiles.
128
-------
ouu
700 -
^T 600 -
H s°°-
E 400 -
"^ 300 -
•S ?oo -
"c
100 -
0 -"
JWl -n
IT 700 -
Jl 600 -
E 500 ~
~J3
400 -
13 oUU
x 200-
2 100 -
CD 600 -
O)
§ 500 -
6 400 -
IIP 300 -
§? 200 -
.i 100 -
0
a
T
s
I
,.,.
V
r
X
o
g
m
0
I
••
o
[
I
1
0
o
[
J-, -
a
Buck Creek •
o
o " o
T
^ SSB8:
slis
.... i f
*1 - _J
T
o „'„
T -v_
~~ X 8
1 .JT, o » • o r":
e— p ^"^ — J0" ig^l T E^
o • " 8
r~*~i | \
- - % tT:d * o t
• I...... ,,1 f--~--:-l ^.x,,, 4 Eft-fc:::! T E|
s « -g is -g -g -s
S-E2 S'wS £c
3 0) CO CD a) CD O
o ^ 8 5 ^ i£ -I £
|> I S « f of
*• > HI T
Q) "***
CO,
$
F-$ \ 1
^ E j. .--d I
1=1
o
o |
Q 1
S H i o
b o
F o
- o
0 °
t
-------
*
^3.
_£
<
_
§
E
8
700-
600 -
500-
400 -
300 -
200-
100-
r» -
o Bald Mtn. Brook
nn a Buck Creek
A Fly Pond Outlet
o 0 Seventh Lake Inlet
n
nn$?n
n _o ,.
o^ yip n
^lasl.
&
O
G^ O
^^^ W8^ -AA A
(a)
-60-40-20 0 20 40 60 80 100
Minimum ANC (u.eq/L)
O)
—
-J?
<
E
"s
^5
TOO -
600 -
500 -
400 -
300-
200 -
100 -
o-
o Biscuit Brook
a Black Brook
A E. Br. Neversink
A o High Falls Brook
A A
^o
A A.
.
"^
AA O
A °00
c^^Ms -j
•
0
OJk&^OD
1 • 1 • t • 1 • 1 • 1 _ • 1 •
(b)
-60-40-20 0 20 40 60 80 100
Minimum ANC (u.eq/L)
700 -
~ eco-
1 500-
f «*>-
E 300 -
1 OT-
^ 100 -
o Baldwin Creek
a Benner Run
A Linn Run
O Roberts Run
+ Stone Run
-P—
-------
s
o
I
180
150 -
120 -
90 -
60 -
30 -
0 -
-30 -
-60
CO CO
co co
o in
co co cn
co co co
cn co co
^ c\i i-
o o co
05
CO
05
9°
CD
CJ> O) CJ>
go go co
o co r-
- CM CM
CD CD r^-
o
f
CO
0)
O CD
§ 5
O
CM
CO
—t 1—
O O
CO CO
to 55
(a)
Q.
7.5
7 -
6.5 -
6
5.5 -
5 -
4.5 -
4
(b)
CO CO
co co
a :e
55 o
CO CO
I 1
(-; ^
oo
CD O>
co go
CO
O)G)O>O)
cocococo
C7)
co
O)O)O
cococn
o
cn
CO
o
cn
CD
CO
800
700 -
600 -
400 -
300 -
20° '
100 -
o -
-100
CO
CO
(C)
CO
CO
25
co
§
CO
co
cn
1
cn
co
cn
i
55
CO
c3
cn
co
cn
co
CVl
CO
cn
co
o
§1
cn
co
o
o>
co
o
cn
o
S>
c\i
T- m
Start Date
Figure 5-23. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Aljm during episodes in Bald Mountain Brook, Adiron-
dacks. Distances between episode start dates on the x-axis are not linear.
Open circles designate initial values, and closed circles designate minimum
or maximum values.
131
-------
180
150 -
2* 120 -
"B- 90 -
Ji 60 -
O 30 -
< ° '
-30 -
-60
(a)
co co co co en o> en o> en o> en en en o) en o o o o o o o
cococ^cococococococjOcocococococnc^cncj>cnc|)cn
o co !*••• co r*^ co T- ° o o
io
10 to
n:
CL
7.5
7 -
6.5 -
6 -
5.5 -
5 -
4.5 -
{b)
CO CO CO
GO CO CO
O CO h-
EM — T—
c> £ o
co en
co go
cocpcocococo
OOOOOOoOOO
COCOCOOC7)C7>O)C7>G)G>
O O CD CO V" CM T— O CO CO
9S-5:s:e!Sa:SSS
» as 55 s ^
co
is ift
800
700 -
600 -
500 -
400 -
300 -
200 -
100 -
0 -
-100
(c)
cococococncncncnencncn
cocococococococococooo
CM^i-CMCVlSJlI^i-CMC!
o?5SrS'^'loiop;S;co
en
co
en
co
o>
co
en
co
o
o>
3
co
§ 5
o o
g> g>
c\l ^=
o o
o en
o
5?
CD
ii3 55 in
Start Date
Figure 5-24. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Aljm during episodes in Buck Creek, Adirondacks. Dis-
tances between episode start dates on the x-axis are not linear. Open
circles designate initial values, and closed circles designate minimum or
maximum "values.
132
-------
I
co oo co co en en en
m t~ '
00 00
S
en en o> o> o> o>
CD o co o r^» ^
10
o G> 0> en
CO CO 00 CO CO
•^ en CD o CD
55 S> o' o r^
O O
CD O3
CD v
^ c\i
O O
O5 CJ)
C\l CO
O
CD
Is*.
§'§
7.5
7 -
6.5 -
6 -
O- 5.5 -
5 -
4.5 -
4
(b)
CO CO CO CO CDCDCDCDCDCDCDCDCD CD CD CD CD CD O O O O OOO O
CO CO CO CO CO CO GO GO CO GO GO GO GO GO CO , GO GO CO CD CD O O) C3> ' CD CD CD
C5 •*"" ^J" LO Is* ^4" "<~~ tQ ^^ CO C3 Is"- ^" ^" CD CO C? CO CD *r~ C\J CO ^^ C3 CO r*"*
O> O O ' CO ^" ^ " CD CD r** [**- O5 C35 O O i— •*— CM CO ^" ^" IO IO IO
^^
— 1
t
F
<
ouu -
700 -
600 -
500 -
400 -
300 -
200 -
100 -
o -
uu
CO OO CO CO O> O> O> O) ~ O> CDCDO>O> OO5CDO CDOO
CO CO CO CO CO CO CO CO CO CO CO CO 00 CO CO CO CO CO O> CD
c^ if" i^f 10 r*** "*J" ^~* cis c^ co CD r^* ^d~ ^t* o? CD cs CD CD ^~
OTOO^ oo^^^cocDr^r^00 CDCDOO *- T- CM
o o o o o o
CD CD CD CD O) CD
c\3 c^ c5 c5 co N-,
co ^ ^ w5 to u5
(C)
Start Date
Figure 5-25. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Alim during episodes in Fly Pond Outlet, Adirondacks.
Distances between episode start dates on the x-axis are not linear. Open
circles designate initial values, and closed circles designate minimum or
maximum values.
133
-------
r
180
150 -J
CD
:±
O
I
90 -
60 -
30 -
0 -
-30 -
-60
(a)
CO OO CO CO CO O> O) 0) O) O) O) O> CJ) C7> O) O) O> O O
90 coco coco co co oo co QO co £000 co co 00 co cn <3>
^3 CO O) I""*** ^J C3 IO CO t**"'- *—* CO C3 Tj" ^"^ ^3 ^D CD CD ^~
cn22oocMcococo'""5i::;o°cnS5o?:^CM
0000
cn cn cn cn
CM ^ O CO
0
cn
CD
55
in in
CL
7.5
7 -
6.5 -
6 -
5.5 -
5 -
4.5 -
(b)
CO CJ
S 5
cococncncncncncnencncncncnen
GO oo oo co oo co op GO co GO GO co co co
P^cvSoincof^-!— cDO^^oocBcb
CM CO CO CO
u5 n
o o o o o o o
cn cn cn cn cn cn cn
c\i ^ o co CD
in in in
D)
800
700 -
600 -
500 -
400 -
300 -
200 -
100 -
0 -
-100
S
(C)
co co co co cn cn cncncncncncncncncncnoooooo
cococooooooooooooocooooooooooococncncncncncn
cn t-- CM
S § §
u3 £5 .
j^ j^ CM
CO CO CO
• o T- ^- oj co
c\i ^ o
^ ^~
co
o
o
in in in
Start Date
Figure 5-26. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Alim during episodes in Seventh Lake Inlet, Adiron-
dacks. Distances between episode start dates on the x-axis are not linear.
Open circles designate initial values, and closed circles designate minimum
or maximum values.
134
-------
CD
O
<
180
150 -
120 -
90 -
60 -
30 -
0 -
-30 -
-60
(a)
GO O O
2 s §
^ CM CO
COO>COO>O>O>O>COCOCOCOCOCOO OOOOOOO OO
gcococopioacopspocooscooscocncggcncococgcoco
•P in O
in m
inincocoOTcoooi-CMi-
in o CD
CM CM
r-- o o
CO 00
o o
O CO
in in
Q.
7.5
7 -
6.5 -
6 -
5.5 -
5 -
4.5 -
4
(b)
CO 0>
00 00
^ CM 00
COOOOOOOCOOO
03 go oo go oo 03 en
oj cvi r*^ CD co ''™~ .f"*"«
O O O O Q O
CM ^ T^ CM CM CM CO
OOOO
O> CO G) O) CO CO
f^- o o -^- o co
in
in m
800
700 -
600 -
400 -
300 -
200 -
100 -
0 -
-100
(c)
OOOO OOOOOOO) OO
tt5 00 '"*"' '"rt JVV *™*^ 'v^ i~f^ r** 'v^ 'v^ /v%
.
Cj ^
CM 00
co OT co co co co
in ^— 2> o
-------
180
150 -
^T 12° '
B- 90 -
0>
.=L 60 -
30 -
0 -
-30 -
-60
O
i
O)
O>
co
in
O)
§
CD
U)
CO
t~-
en
CO
O)
a>
co
CD
o>
90
co
•-. • r- s\i £y —
*- ^ ^ ?: CM
o
.1
o
i
a.
7.5
7
6.5
6
5.5
5
4.5
4
(b)
a>
§9
a>
co
10
o>
go
o
go
co
co
co
en
to
i
92
§
o
co
O)O>O>O>OG)O>
CO CO CO CO CO CO CO CO CO
^— I^lOOOOCOlOh^O)
O) O) O O O .O
CO 00 O) O) CD O)
(c)
3
5
;s ^ ^ c\j
Start Date
Figure 5-28. (a) Initial and minimum ANG, (b) initial and minimum pH, and (c) initial and
maximum estimated Aljm during episodes in Black Brook, Catskills. Dis-
tances between episode start dates on the x-axis are not linear. Open
circles designate initial values, and closed circles designate minimum or
maximum values.
136
-------
180
150 -
O
<
90 -
60 -
30 -
0 -
-30 -
-60
(a)
GOCOG>O0)O)G)OO
O> h"* O CD CO T" IO O CD
O
h**
OG>
O> CO
G)
• O)
U)
o> o>
§1
O<3
25
7.5
7 -
6.5 -
6 -
5.5 -
5 -
4.5 -
4
(b)
OO
O)
coooco
oocococogoco
O1OOOOOOO.O
co cj> ' cj> gj> go c3> CJD cj> o>
CO U5 O CO C\l ^3 r** C3 CO
-100
CO GO 00 GO
5> i^ O CD
^ s: £^ e?
O i— CM 00
codcioocoooooooco
S5 o co K O) oo i-
m LO co
o
o>
§ I
•r- CM
o
s?
tS
CM
o
s
s
"5
Start Date
Figure 5-29. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Aljm during episodes in East Branch Neversink River,
Catskills. Distances between episode start dates on the x-axis are not
linear. Open circles designate initial values, and closed circles designate
minimum or maximum values.
137
-------
CO
:±
O
I
180
150 -
120 -
90 -
60 -
30 -
-30 -
-60
(a)
coG)CT)G) en en en en en en en en en en en en en CD o o o
iniOt— in cninoxfo) CM CM in en •»— en co co -i— r^. in CM
"*"- eo *" *" ^tnr^ojpjcn^oo o *" i— •«— csi T^ ^ c\i
o o
T- P:
co co
o o
C5 C3
o
o
in
a.
7.5
7 -
6.5 -
6 -
5.5 -
5 -
4.5 -
(b)
GO en o) o> o> en
03 co co co co cp
c»
en
0) 0) o>
o
c>
TO
e» in o
— s; Qi IT i; *X —•
10 in h- en en en °
o
o
O>
o
in
800
700 -
^ 600 -
==i 500 -
§ 400-
i 30° "
^" 200 -
100 -
o -
-100
(C)
CO O)
CO CO
O) O
CO CO
O> 0) CD 0> 0>
CO CO CO CD CO
O) O> O) O) 0> 0> 0> O O O O O Q O
CO CO CO CO CO CO CO 0) 0) 0) 0) 0) 0> 0>
o
O>
to in in
in
*t in
Start Qate
Figure 5-30. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Aljm during episodes in High Falls Brook, Catskills. Dis-
tances between episode start dates on the x-axis are not linear. Open
circles designate initial values, and closed circles designate minimum or
maximum values.
138
-------
CD
180
150
120
90
60
O 30
? o
-30
-60
(a)
CO
S5
CO
§
00
S2
10
CO
I
o>
£5
in
o>
00
O)
I
CO
O)
1
52
CO
CO
u5
CD
CO
O
CJ
CO
Q.
7.5
7
6.5
6
5.5
5
4.5
4
(b)
CO
oo
S2
10
CO
1
en
SS
in
Si
o>
S5
o
1
CO
go
CO
U5
o>
00
o
o
9?
CM
CNi
^^
— 1
1
F
•^
ouu -
700 -
600 -
500 -
400 -
300 -
200 -
100 -
o -
-100 -
......
0/23/88 -
• A m
^-'-""-'^
CO CO O
CO CO CO
in o in
i: ^ Si
«,
3/20/89 -
_ <^5\^=__4fc===__
-^rr- ^r^.
en en en en en
CO CO CO CO CO
TJ- o ^ CD f-
co oo "> w CD
•__
"•***"
6/20/89
9/22/89 •
2/2/90 •
(c)
Start Date
Figure 5-31. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Aljm during episodes in Baldwin Creek, Pennsylvania.
Distances between episode start dates on the x-axis are not linear. Open
circles designate initial values, and closed circles designate minimum or
maximum values.
139
-------
0)
O
180
150 -
120 -
90 -
60-
30 -
o-
-30 -
-60
(a)
CO CO O) CD O)
CO CO CO CO CO
55 ^ ^ |i §s
co co
5 S i
en en en en
co co co co
o -r- in en
en en en o> en
^ to in en Si
en o
co en
in en
0 2
en en
c\! in
o o
en en
i- CM
od
cncn
Q.
7.5
7 -
G.S -
6 -
5.5 -
5 -
4.5 -
4
(b)
5 §
O> G) O5 O> 05 O) O) CD G) CD CD CD O)
comcocooococococogcogco
o ci o ci
en en en en
tvi in ^ o3
o o
en en
^ i?5
a
1
800
700 -
_^ 600 -
^ 500-
3- 400 -
e 300 -
^ 200 -
100 -
0 -
-100
CO CO G) CD CD CD G) CD G) CD CD CD CD CD CD
cpcococpcpcpcocococpcocpcp— —
Start Date
Figure 5-32. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Alim during episodes in Benner Run, Pennsylvania. Dis-
tances between episode start dates on the x-axis are not linear. Open
circles designate initial values, and closed circles designate minimum or
maximum values.
140
-------
180
150 -
120 -
90 -
O 30 -
I o-
-30 -
-60
(a)
OO OO GO GO GO CO CO O) O> O) CO CO CO CO CO CO CO CO CO GO CO CO O O O O
co go oo co co go go op co SS 9?. 92 92 oo oo co co 90 €9. go po go co co o CD
CM 00 tO CO CO CD ^ T"- *cj- Tj* "^J* CO r**" T— IO CO tO tO O CM CO CD CO i— to O
CL
7.5
7 -
6.5 -
6 -
5.5 -
5 -
4.5 -
4
(b)
00 CO OO CO CO O) O) O>
CVJOOtOCO OCD^J-f—
CO CO GO 00 GO 00 00
GO 00
io 10
i
go oo
CO CO
o> -en
o> T-
o o
cn cn
in o
10 m 10
^ ^ CO
cn
800
700 -
600 -
500 -
400 -
E 300 -
^ 200 -
100 -
o -
-100
(c)
co oo oo co coocncncn cn cn cn cn ocncncn
COCOcOCO COCOCOCOCO COCOOOCO COCOCOOO
cvlco inco o> CD Tf T— ^ ^^cnr^- i— in o> in
co co
to o
S° °5 =;
oo
Start Date
Figure 5-33. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Aljm during episodes in Linn Run, Pennsylvania. Dis-
tances between episode start dates on the x-axis are not linear. Open
circles designate initial values, and closed circles designate minimum or
maximum values.
141
-------
!;• W §
CM CM CO CO
-* o
3^ ^
incMcoina>cMO>inT-
CM CM QJ jr; CM
CO O> O> T— T—
in
C\3
a.
7.5
7 -
6.5 -
6 -
5.5 -
5 -
4.5 -
4
(b)
CO
S°
coco
O> C7)
— —
cocog>
co
ocnoooooooo
£OCOCJOC^>G>C^)£^>C^>€7>C^>
cMcoinocMcnin-i— ino
_J
800
700 -
600 -
500 -
400 -
g 300 -
^ 200 -
100 -
0 -
-100
(c)
COCDO)O>O>O>0>OO>O>a)O>O>O>O>OCJ>O>OOOOOOO
cooococooooooocooooococococococooococncncncnocncn
:
^ J: c\3 CM
Start Date
Figure 5-34. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Aljm during episodes in Roberts Run, Pennsylvania.
Distances between episode start dates on the x-axis are not linear. Open
circles designate initial values, and closed circles designate minimum or
maximum values.
142
-------
180
150 -
^ 120-
"B- 90 -
CD
rt 60 -
O
<
30 -
0 -
-30 -
-60
(a)
GO
OO
CO CD CO
00 00 CO
IO O CO
CO
CO
oo
CD
op
CD
op
CO
oo
If) CO
CD
go
in
CD
go
c5
CO
CO
go
CO
go
co
f- CM CM
CD
oo
O>
go
in
o
CD
o o
CD CD
CM CO
o
in
si
in
^
X
Q.
7.5
7 -
6.5 -
6 -
5.5 -
5 -
4.5 -
(b)
COO)C?>CDO>CDCniCDCDCDCDCDO>C>CDOOO OOOOO
00 CO CO CO CO CO CO CO CO CO CO CO CO CO CO CD CD CD CD CO CD CO CD
-r-LOOCOCOT-LOCOincOCVlTj'CDCMlOCOT— OJCDLOi— IOO
CM CM
LO
CD CD CD O> T— i—
CM
CVi
CM
800
700 -
600 -
500 -
400 -
300 -
200 -
100 -
0 -
-100
(C)
CO CO CD CF> CD CO O> CD CO
COOOCOCOCOCOCOCOCO
^W§!£3|!^is35in
CD
9°
CD CD
00 00
CM ^=
CD
00
CO
CO O
00 CO
iS cS
o
CD
o
CD
CM
CM
CO
IO IO IO
in
CD CD CD CD
o
OT
5>
CM
o
O>
o
CO
o
CD
o
Start Date
Figure 5-35. (a) Initial and minimum ANC, (b) initial and minimum pH, and (c) initial and
maximum estimated Alim during episodes in Stone Run, Pennsylvania. Dis-
tances between episode start dates on the x-axis are not linear. Open
circles designate initial values, and closed circles designate minimum or
maximum values.
143
-------
A
CM
ff
160
140 -
120 -
100 -
80
'(a)
-r 100
80 -
.§, 60 -
'eo
40 -
20 -
0
Sa
(b)
25
t~ 20 -
15 -
'_ 1°~
O 5 "
0
2)
55 55
'(c)
80
60 -
40 -
20 =
S
5
25
CO
a
SCO
250 -
200 -
150 :
100
a
— 1 H
^ ..H
.01 -
.001
?5 55 Bj
-(e)
%
Si Si
05
Start Date
Figure 5-36.
2-
(a) QOf, (b) NO3~ (c) C\~, (d) estimated A~ (e) 2CB, and (f) discharge at
the beginning of episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in Bald Mountain Brook, Adirondacks.
Distances between episode start dates on the x-axis are not linear.
144
-------
160
J, 14°-^
c\J 120 H
100
"(a)
150
\j
I
•m
O
100 -
50 -
O
15
12 -
6_
3-
O 35
as
35 35
S3
trt
s - s
ffj
(b)
"(0)
120
90 -
eq -
30 -
c\)
ift 25 u5
(d)
CQ
a
300
250 -
200 -
150 -
100
a a»
55 Bs 25
(e)
1 -
"E -1 -
O .01 -
.001
05
"5
35
Start Date
Figure 5-37. (a) SO42~ (b) NO3~ (c) Cl~ (d) estimated A", (e) 2CB, and (f) discharge at
the beginning of episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in Buck Creek, Adirondacks. Distances
between episode start dates on the x-axis are not linear.
145
-------
160
140 -
120 -
100 -
80
a
t
o
-T. 100
80 -
60 -
n «-
2 20 -
100
75 -
80 -
25 -
i
a as 5
a.
s
S
'(a)
'§§§§
(b)
f*
5
"(c)
m
a
75
so -
25 -
750
600 -
450 -
300 -
150
'.§ 5 5 S
§ § §
(d)
3s
S S 5
(e)
O) O> O)
^ ro QD
c| -H
O -01 H
,001
"(t)
Start Date
Figure 5-38.
2~
(a) SO4~, (b) NO3- (c) C\~, (d) estimated A~ (e) 2CB, and (t) discharge at
the beginning o'f episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in Fly Pond Outlet, Adirondacks. Distances
between episode start dates on the x-axis are not linear.
146
-------
CM
160 -
140 -
133 -
100
__I 1_
10 -
o
"(a)
T~ T— <\J •»— T11 T- T-
^ ^= ol
-------
sr
CM
120 -
100 -
80
(a)
"> 120
90 -
30-
*
(b)
F5
a
20 -
15~
« 10 -
O 5 -
40
3D -
33 -
10 -
t^
i
11
1 -
.1 -
J01
'(f)
oS
c5
»
w
e\l
Start Date
Figure 5-40.
2~
(a) SO4~ (b) NO3~ (c) CI" (d) estimated A~ (e) 2CB, and (f) discharge at
the beginning of episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in Biscuit Brook, Catskills. Distances
between episode start dates on the x-axis are not linear.
148
-------
CM
sf
160
140 -
120 -
100
(a)
100
75 -
50-
25-
cvl
(b)
o
25
20 -
15 -
10 -
5 -
u5
35
"(C)
I
50
40 -
30 -
20 -
10 -
55
(d)
I
CO
o
400
350 -
300 -
250 -
200 -
150
,§,
O
.01
fis
S
(e)
5
(f)
25
Start Date
Figure 5-41. (a) SO42~ (b) NO3~ (c) Cl~ (d) estimated A', (e) 2CB, and (f) discharge at
the beginning of episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in Black Brook, Catskills. Distances
between episode start dates on the x-axis are not linear.
149
-------
-T. 140 -
120 -\
ff
80
T 130
90 -
60 -
30 -
0
5 5
'(a)
O F5 3
S § 5
(b)
o
s
20 -
15 -
10 -
5 -
0
fa
=v==8=-^-=t--«-^g
'(c)
23
60
^5 -
30 -
15 -
0
^ 10 -
to
^ 1 H
O ,1 H
.01
i S
iS 05 O5 5
C\l O^ CT ^ ^
(d)
c\l fa c\) ft S5 ^
Start Date
Figure 5-42. (a) SO42- (b) NOg" (c) CF, (d) estimated A~ (e) 2CB, and (f) discharge at
the beginning of episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in East Branch Neversink River, Catskills.
Distances between episode start dates on the x-axis are not linear.
150
-------
-j- 150
fr 125 -
c\l 100 -
75
120
90 -
60 -
30 -
"(a)
25 Ss
Ss
=^§
(b)
O
25
20 -
10 -
"(c)
60
<=d 45 -
I 30-
'*< 15 -
55 25
§ S
(d)
—, 40°
t350 -
300 -
O 200 -
W 150
O
.1 -
.01
Figure 5-43.
5
(f)
25 55 iS
Start Date
(a) SO42~ (b) NO3", (c) CF, (d) estimated A~ (e) SCB, and (f) discharge at
the beginning of episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in High Falls Brook, Catskills. Distances
between episode start dates on the x-axis are not linear.
151
-------
-------
CM
8"
200 -
150 -
100 -
50
'(a)
100
75 -
50 -
* a».a» 5fc§§§5S§s>SS§?**5
^ ^ 5
"(c)
so
40 -
30 -
2) -
10 -
250
a
1 -
.1 -
.01
25 a s.a» 5.5
(d)
ift ift 25
(f)
Start Date
Figure 5-45. (a) SO42' (b) NO3~ (c) Cr, (d) estimated A~, (e) 2CB, and (f) discharge at
the beginning of episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in Benner Run, Pennsylvania. Distances
between episode start dates on the x-axis are not linear.
153
-------
225 -
125 -J-p-
-1 1 I—
40 -
30 -
i -H
(b)
O
60
40 -
23 -
ft it5 ift 5 S
"(c)
20 -
10 -
(d)
m
a
400
350 -
300 -
230 -
200 -
150
ift 2s
(e)
> ft!
(f)
,01
Start Date
Figure 5-46. (a) SO42" (b) NO3~ (c) C\~, (d) estimated A~, (e) 2CB, and (f) discharge at
the beginning of episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in Linn Run, Pennsylvania. Distances
between episode start dates on the x-axis are not linear.
154
-------
-^ 250
g- 225 -
3, 200 -
<<. 175 -
cT 19° ~
W 125
«—•—«
"(a)
10 -
'«
o
S)
§
(b)
60
93 -
40-
-=• 30 -
O 20 -
10
"(c)
43
30 -
20 -
10 -
(d)
?= Si
275
25
5 fe
$! si si
00 175 -
W
125
?= Si
o
.01
g Si Si
(f)
Start Date
Figure 5-47. (a) SO42~ (b) NO3~ (c) C\~, (d) estimated A~ (e) 2CB, and (f) discharge at
the beginning of episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in Roberts Run, Pennsylvania. Distances
between episode start dates on the x-axis are not linear.
155
-------
CM
233 -
180 -
140 -
100
-(a)
50
40 -
3L 3°"
•= 33 -
O »
3D
20 -
10 -
300
240 -
a 21°-
O 180 -
W 150
>*-•
'(0)
(d)
(e)
10
€ 1
rT .1 H
.01
"(f)
Figure 5-48.
Si
Start Date
2~
(a) SO4~ (b) NO3- (c) C\~, (d) estimated A", (e) 2CB, and (f) discharge at
the beginning of episodes (open circles) and at the time of minimum ANC
(closed circles) during episodes in Stone Run, Pennsylvania. Distances
between episode start dates on the x-axis are not linear.
156
-------
o
6
cy\
OU
40 -
30 -
20 -
10 -
•
o -
"
-20.-
on
~ou
•*
Cf\
DU
50 -
~
40 -
20 -
10 :
-10 -
on
. i .1 . i i i i i • i .
:
I O
^ cu
! O
c^&\
' x*k f*\ C5
<9&Jo^OO O
to
Q
8R
o i
• i ' i ' i ' i ' i • i '
50 0 50 100 150 200 250 3(
Minimum ANC (fieq/L)
i i i i i i i i i i i i i
00!
Oi
OjO
to
. » o °° o o
^ntLjn& ^ O
Sx5o ^
oo0o 0
- cy\ -1
.(a) ^
50 -
- S * "
" ^ X -
r «" 20 -
c5 10 -
.C n
O u
O "10 ~
" z -20-
: -30 -
._
f" "*rU ~~l
yo -£
Cf\ _!
.(c) 5°.
— f\
o -
^^,
: | -so -
f -100 -
O)
| -150 -
O
tf -200 -
W
r250 -
onn -
i I . i . t . i , i , i .
O!
4
od
o ! o
^^r
^^^^80 o
Uu^^C) U U
(P^ 0 ° °
o P o
. o
' 1 ' 1 ' 1 ' 1 ' 1 * 1 '
W 0 50 100 150 200 250 3C
Minimum ANC (ueq/L)
. i . i . i , i , i , i ,
QgL-
^waHk^^^
imQ^>^
^9 Q °
C§Q o0 ° o
• 8 °
o
jo
-50 0 50 100 150 2200 250 300
Minimum ANC (ueq/L)
-50 0 50 100 150 200 250 300
Minimum ANC (u.eq/L)
Figure 5-49. Relationship of (a) SO42~ (b) NO3~ (c) estimated A", and (d) 2CB changes
during episodes to minimum ANC, Adirondacks.
157
-------
10 -
o-
-10 -
^-20H
o
W -GOH
II I I I I I
cP
°
o o
o o
c9
T I 1 I T~ 1 I I i I
-60 -20 20 60 100 140
Minimum ANC (ueq/L)
(a)
CO
-co
o
40 -
10 -
-10 -
I I
o
0
o
ogoo
u
0
o
II T F I 1 I I I i
-60 -20 20 60 100 140
Minimum ANC (u.eq/L)
(b)
30 -
I 2°~
? OH
to
-20 -
•30
-60 -20 20 60 100 140
Minimum ANC (u.eq/L)
O)
60 -
0-
-30 -
-60 -
-90 -
-120 -
-150 -
-180
n—i—i—i—i—i—i—i—i—i—
-60 -20 20 60 100 140
Minimum ANC (ueq/L)
(d)
Figure 5-50. Relationship of (a) SO42~, (b) NO3~ (c) estimated A , and (d) SCB changes
during episodes to minimum ANC, Catskills.
158
-------
-60-40-20 0 20 40 60 80 100
Minimum ANC (i^eq/L)
-60-40-20 0 20 40 60 80 100
Minimum ANC (ueq/L)
CD
O>
cc
_c
O
-60^0-20020406080100
Minimum ANC
-60 ^10 -20 0 20 40 60 80 100
Minimum ANC {u.eq/L)
Figure 5-51. Relationship of (a) SO42~, (b) NO3~ (c) estimated A~, and (d) SCB changes
during episodes to minimum ANC, Pennsylvania.
159
-------
O)
o
*5inn
P- 250 -
g 200 -
«
o 150 -
"c
<
o5 i°o -
its
c
50 -
.—» Qnn
•g.
nj
3, 250 -
o
Z 200 -
E
= 150 -
E
'c
ii 100 -
OS
vU
« 50 -
.9
c ^
u
. 0)
Adirondacks
Q
^-
L- ._:_.; ; ;l
I L: :J~3
1 8
-S- ®
±Q
- ~ ... ._
T
, (d)
Adirondacks
©
-I. P
|_ I T:....j
:^$xx^™ ^Sl.— , iinj^j-'-J
~ir tru ^y 1
2-
SO4 N03 Gl A"
.
250 -
200 -
150 -
100 -
50 -
250 -
200 -
150 -
100 -
50 -
. (b
Catskills
> 4. ^
^T 0
t •_, I _ j_
^ V ^^^^^p^^w
, (e
Catskills
j.
^p ^
_§_
J _£_ pA^
S042" NO g Cl" A'
ow -
250 -
200 -
150 -
100 -
50 -
n -
Pennsylvania
JL
T
1
| t . t o
E^'S'!^!^) fffln'BJP^Bn
J : . : : : v>
250 -
200 -
150 -
100 -
50 -
0
Pennsylvania
Figure 5-52. Anion concentrations for (a) Adirondack, (b) Catskill, and (c) Pennsylvania streams at the beginning of episodes,
and anion concentrations for (d) Adirondack, (e) Catskill, and (f) Pennsylvania streams at the time of minimum ANC.
A~ concentratibns are estimates. Line in box indicates median; upper and lower borders of box show 25th and 75th
quartiles; whiskers indicate 10th and 90th percentiles; circles represent observations beyond 10th and 90th
percentiles.
-------
Table 5-1. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiles of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Aljm for Bald Mountain Brook, Adirondacks (unless specified, units are ^eq/L)
Overall
minimum
Overall
maximum
1 989 weekly
minimum6
1989 weekly
25th percentile
1989 weekly
50th percentile
1989 weekly
75th percentile
1989 weekly
maximum
n for weekly
statistics
Q
(m3/s)
0.01
2.86
0.01
0.02
0.02
0.05
2.86
51
PH
4.60
6.87
4.80
5.42
6.15
6.53
6.87
52
ANC
-21
117
-10
7
26
56
99
52
so42-
88
146
93
112
124
131'
146
52
N03-
3
86
11
25
34
54
77
52
cr
3
22
7
8
9
9
13
52
DOC
L(m9/L)
1.7
11.1
1.7
2.6
3.2
3.7
6.9
51
Ca2+
71
157
92
101
111
120
142
52
Mg2+
20
65
28
36
42
50
65
52
Na+
16
69
22
34
42
53
70
52
K+
4
25
4
8
8
9
12
. 52
Altd
(«g/i-)
37
730
72
129
176
300
609
52
Al a
Alim
C"9/L)
0
571
0
25
55
95
338
52
0)
Measured or estimated values.
2/1/89 through 1/31/90.
-------
Table 5-2. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiles of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Al!m for Buck Creek, Adirondacks (unless specified, units are //eq/L)
Overall
minimum
Overall
maximum
1989 weekly
minimum13
1989 weekly
25th percentile
1989 weekly
50th percentile
1989 weekly
75th percentile
1 989 weekly
maximum
n for weekly
statistics
Q
(m3/s)
<0.01
4.13
< 0.01
0.02
0.03
0.08
0.58
51
PH
4.43
6.72
4.71
4.98
5.38
5.84
6.54
52
ANC
-30
101
-22
0
10
21
71
52
so42-
108
167
129
142
147
152
167
52
N03-
11
123
12
20
37
58
100
52
cr
5
19
9
11
11
12
14
52
DOC
(mg/L)
2.6
14.9
2.6
3.9
4.7
5.4
9
52
Ca2+
92
164
103
117
124
134
151
52
Mg2+
25
52
27
34
37
44
52
52
Na+
16
68
25
31
37
47
68
52
K+
4
28
4
9
10
11
13
52
Al,d
C«g/L)
160
954
160
291
396
522
823
52
Al a
Alim
N/L)
0
690
0
94
165
279
536
52
ro
Measured or estimated values.
2/1/89 through 1/31/90.
-------
Table 5-3. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiles of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Al,m for Fly Pond Outlet, Adirondacks (unless specified, units are
Overall
minimum
Overall
maximum
1 989 weekly
minimum13
1989 weekly
25th percentile
1989 weekly
50th percentile
1989 weekly
75th percentile
1989 weekly
maximum
n for weekly
statistics
Q
(m3/s)
0.01
0.26
0.01
0.01
0.01
0.02
0.08
47
PH
5.51
7.38
6.12
6.85
7.08
7.19
7.38
52
ANC
17
442
46
151
225
282
375
52
so42-
80
154
97
115
121
128
154
52
N03-
7
93
11
25
32
48
86
52
or
12
88
18
25
32
36
69
52
DOC
(mg/L)
2.2
18.2
2.2
3.0
3.9
4.8
7.8
52
Ca2+
143
422
175
227
258
285
314
52
Mg2+
35
111
48
62
76
88
109
52
Na+
32
116
49
77
92
100
116
52
K+
7
29
7
12
13
14
17
52
Altd
(«g/L)
60
485
60
98
127
169
369
52
Al a
Alim
C«g/L)
0
142
0
20
20
20
107
52
0)
CO
Measured or estimated values.
2/1/89 through 1/31/90.
-------
Table 5-4. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiles of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Al)m for Seventh Lake Inlet, Adirondacks (unless specified, units are /ueq/L)
Overall
minimum
Overall
maximum
1989 weekly
minimumb
1989 weekly
25th percentile
1989 weekly
50th percentile
1989 weekly
75th percentile
1 989 weekly
maximum
n for weekly
statistics
Q
(m3/s)
<0.01
4.12
0.01
0.03
0.06
0.16
2.60
51
PH
4.63
6.74
4.89
5.28
5.67
5.99
6.56
52
ANC
-20
146
-5
7
' 21
30
70
52
so42-
107
184
140
140
148
154
184
52
N03-
7
106
7
14
26
45
78
52
CT
4
16
9
10
11
12
15
52
DOC
(mg/L)
2.6
17.8 •
3.7
4.8
6.2
6.9
11.5
52
Ca2+
107
2Q2
116
130
139
150
174
52
Mg2*
32
64
34
39
43
50
63
52
Na+
18
52
22
28
33
41
52
51
K+
4
27
4
8
9
10
14
52
Altd
(«g/L)
118
846
190
290
333
407
572
52
AC
fcg/L)
0
485
0
64
93
158
264
52
o
Measured or estimated values.
2/1/89 through 1/31/90.
-------
Table 5-5. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiles of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Aljm for Biscuit Brook, Catskills (unless specified, units are ^
Overall
minimum
Overall
maximum
1989 weekly
minirnurnb
1 989 weekly
25th percentile
1 989 weekly
50th percentile
1989 weekly
75th percentile
1 989 weekly
maximum
n for weekly
statistics
Q
(m3/s)
< 0.01
9.57
0.03
0.08
0.20
0.46
2.84
52
PH
4.97
6.63
5.34
5.89
6.11
6.29
6.50
51
ANC
-6
55
7
22
28
33
50
51
S042~
98
149
111
128
132
134
144
51
N03-
1
129
11
23
33
45
68
49
cr
10
27
13
18
19
19
21
51
DOC
(mg/L)
0.3
7.5
1.0
1.0
1.8
2.2
4.8
51
Ca2+
98
198
108
127
137
144
166
51
Mg2+
31
63
36
48
50
53
58
51
Na+
6
38
9
14
16
17
22
51
K+
4
21
5
5
6
6
11
51
Altd
C«g/L)
13
332
13
24
41
55
166
49
Al a
Alim
(«g/L)
. 0
159
0
1
10
15
92
51
O)
01
Measured or estimated values.
2/1/89 through 1/31/90.
-------
Table 5-6. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiies of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Alim for Black Brook, Catskills (unless specified, units are /teq/L)
Overall
minimum
Overall
maximum
1 989 weekly
minimum13
1989 weekly
25th percentile
1 989 weekly
50th percentile
1989 weekly
75th percentile
1 989 weekly
maximum
n tor weekly
statistics
Q
(m3/s)
0.01
1.6
0.01
0.05
0.08
0.17
0.52
42
PH
5.44
7.04
5.77
6.44
6.67
6.82
7.04
52
ANC
10
176
50
94
118
137
176
50
so42-
100
146
106
135
139
141
146
51
N03-
10
108
28
37
41
45
72
51
cr
5
23
15
16
17
18
23
51
DOC
(mg/L)
0.7
8.1
0.7
1.0
1.3
2.0
7.9
49
Ca2+
136
294
182
223
235
256
288
51
Mg2+
31
66
40
48
52
55
63
51
Na+
6
39
10
10
11
12
26
51
K+
4
27
5
5
6
7
16
51
Al»d
W-)
2
539
2
11
18
35
306
49
Al a
Mlim
feg/L)
9
72
0
10
17
20
49
52
O)
a Measured or estimated values.
b 2/1/89 through 1/31/90.
-------
Table 5-7. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiles of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Al)m for East Branch Neversink River, Catskills (unless specified, units are /teq/L)
Overall
minimum
Overall
maximum
1989 weekly
rninimumb
1989 weekly
25th percentile
1989 weekly
50th percentile
1 989 weekly
75th percentile
1 989 weekly
maximum
n for weekly
statistics
Q
(m3/s)
0.01
41.7
0.03
0.06
0.55
1.23
37.4
49
PH
4.27
5.43
4.45
4.78
4.91
5.0
5.35
52
ANC
-45
21
-31
-9
-6
-3
10
52
S042~
68
143
84
115
118
119
124
52
N03-
4
106
13
15
32
39
67
52
cr
10
28
11
15
16
17
19
52
DOC
(mg/L)
0.3
11.7
1.0
1.6
2.0
2.5
6.3
49
Ca2+
37
124
51
62
66
75
80
52
Mg2+
28
67
30
45
49
55
61
52
Na+
7
87
8
13
15
17
37
52
K+
4
20
4
5
6
8
13
52
A"td
(«g/L)
94
737
101
148
191
240
562
51
Alima
(«g/L)
36
505
0
99
156
188
380
52
Measured or estimated values.
1 2/1/89 through 1/31/90.
-------
Table 5-8. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Qualifies of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Al)m for High Fails Brook, Catskilis (unless specified, units are/ieq/L)
Overall
minimum
Overall
maximum
1989 weekly
minimum13
1989 weekly
25th percentile
1989 weekly
50th percentile
1989 weekly
75th percentile
1989 weekly
maximum
n for weekly
statistics
Q
(m3/s)
0.02
2.38
0.04
0.09
0.20
0.35
2.21
52
PH
5.40
7.07
5.86
6.33
6.61
6.83
7.07
49
ANC
10
172
38
72
95
122
172
52
so42-
91
153
91
129
133
136
146
52
N03-
14
106
17
25
31
37
61
52
or
9
23
12
16
17
18
20
52
DOC
(mg/L)
0.9
8.0
0.9
1.4
1.8
2.7
4.3
49
Ca2*
119
277
119
196
209
235
277
51
Mg2+
26
60
26
42
46
50
57
52
Na+
8
62
10
13
14
16
33
51
K+
3
23
3
6
6
7
12
52
Altd
feg/L)
10
258
6
12
24
40
258
50
Al!ma
(«g/i-)
0
84
0
0
6
11
84
52
a
03
Measured or estimated values.
2/1/89 through 1/31/90.
-------
Table 5-9. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiles of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Al,m for Baldwin Creek, Pennsylvania (unless specified, units are /*eq/L)
Overall
minimum
Overall
maximum
1 989 weekly
minimum13
1989 weekly
25th percentile
1 989 weekly
50th percentile
1989 weekly
75th percentile
1 989 weekly
maximum
n for weekly
statistics0
Q
(m3/s)
0.02
1.95
0.03
0.07
0.08
0.13
0.32
49
pH
5.37
7.03
5.62
6.11
6.29
6.40
6.70
52
ANC
_H
97
3
19
31
40
68
52
so42-
102
288
102
185
194
203
253
40
N03-
3
39
10
16
17
22
34
40
cr
12
63
12
34
40
43
50
40
DOC
(mg/L)
0.3
6.2
0.3
1.0
1.4
2.3
4.0
40
Ca2+
84
244
133
153
160
174
199
40
Mg2+
64
151
74
81
83
87
102
40
Na+
11
31
13
19
21
22
24
40
K+
9
37
13
15
17
20
25
40
Altd
(«g/L)
1
102
1
17
26
35
87
52
Alima
(ug/L)
0
84
10
11
11
12
22
49
O)
CO
Measured or estimated values.
2/1/89 through 1/31/90.
Missing values for period 11/89 through 1/31/90.
-------
Table 5-10. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Qualities of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Al(m for Benner Run, Pennsylvania (unless specified, units are /*eq/L)
Overall
minimum
Overall
maximum
1989 weekly
minimum13
1 989 weekly
25th percentile
1 989 weekly
50th percentile
1 989 weekly
75th percentile
1989 weekly
maximum
n for weekly
statistics0
Q
(m3/s)
0.04
2.01
0.06
0.10
0.14
0.22
0.93
52
PH
4.91
6.53
5.06
5.86
6.02
6.20
6.52
52
ANC
-15
54
-6
7
11
14
26
50
so/-
59
204
59
72
87
96
183
40
N03-
1
25
1
8
10
11
21
39
cr
25
102
32
53
64
80
102
40
DOC
(mg/L)
0.6
9.6
0.7
1.2
1.6
2.2
9.4
40
Ca2+
54
168
61
73
77
81
129
40
Mg2*
38
81
44
46
48
50
81
40
Na+
15
54
25
35
39
44
54
40
K+
13
30
13
18
19
20
30
40
A"td
(«g/L)
4
232
4
15
24
40
192
52
Al a
Alim
C"9/L)
0
133
0
0
5
19
102
52
Measured or estimated values.
b 2/1/89 through 1/31/90.
0 Missing values for period 11/89 through 1/31/90.
-------
Table 5-11. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiles of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Al)m for Linn Run, Pennsylvania (unless specified, units are /teq/L)
Overall
minimum
Overall
maximum
1 989 weekly
minimum6
•1989 weekly
25th percent! le
1 989 weekly
50th percent! le
1 989 weekly
75th percentile
1989 weekly
maximum
n for weekly
statistics0
Q
(m3/s)
< 0.01
8.62
<0.01
0.11
0.17
0.40
1.16
49
PH
4.69
6.73
4.84
5.37
5.97
6.24
6.70
52
ANC
-21
85
-21
0
11
30
85
52
so42-
165
272
190
207
218
233
272
40
N03-
6
55
7
15
19
26
55
39
cr
17
66
19
25
30
33
61
40
DOC
(mg/L)
0.5
7.7
0.6
1.5
1.9
2.7
5.2
40
Ca2+
103
248
121
160
175
196
248
40
Mg2+
37
152
48
58
61
67
76
40
Na+
12
68
15
20
25
33
59
38
K+
4
32
8
10
11
12
19
40
AltH
C"9/L)
7
654
10
22
40
147
564
52
Alim3
v*g/L)
0
805
0
28
63
163
805
52
Measured or estimated values.
2/1/89 through 1/31/90.
Missing values for period 11/89 through 1/90.
-------
Table 5-12. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiles of Discharge, pH, ANC, Major
Ions, DOC, Alfd, and Al|m for Roberts Run, Pennsylvania (unless specified, units are/teq/L)
Overall
minimum
Overall
maximum
1989 weekly
minimumb
1989 weekly
25th percentile
1989 weekly
50th percentile
1 989 weekly
75th percentile
1989 weekly
maximum
n for weekly
statistics0
Q
(m3/s)
<0.01
2.81
0.03
0.08
0.14
0.31
1.34
50
PH
4.59
6.51
4.66
5.07
5.33
5.70
6.40
52
ANC
-52
57
-30
-11
. -2
3
46
52
so42-
135
310
135
170
178
190
208.
40
N03-
0
16
1
4
6
9
14
38
cr
16
50
15
22
25
36
50
40
DOC
(mg/L)
0.8
7.4
1.1
1.8
2.6
3.4
5.5
40
Ca2+
69
161
72
92
99
105
125
40
Mg2+
58
110
61
68
75
79
87
40
Na*
8
34
10
17
19
22
28
40
K*
10
29
10
15
17
18
28
40
Al,d
("9/L)
23
420
26
54
93
123
420
52
Alima
C«g/L)
8
301
7
54
98
145
225
52
Measured or estimated values.
1 2/1/89 through 1/31/90.
Missing values for period 11/89 through 1/31/90.
-------
Table 5-13. Overall Maximum and Minimum, and 1989 Weekly Maximum, Minimum, and Quartiles of Discharge, pH, ANC, Major
Ions, DOC, Altd, and Alim for Stone Run, Pennsylvania (unless specified, units are /teq/L)
Overall
minimum
Overall
maximum
1989 weekly
minimum11
1989 weekly
25th percentile
1989 weekly
50th percentile
1989 weekly
75th percentile
1 989 weekly
maximum
n for weekly
statistics0
Q
(m3/s)
< 0.01
3.39
0.01
0.09
0.13
0.25
0.97
41
PH
4.53
6.26
4.78
5.08
5.31
5.77
6.09
50
ANC
-41
26
-26
-8
-5
7
22
50
so42-
55
243
131
162
184
199
225
41
N03-
0
21
0
1
2
3
7
41
cr
11
74
19
25
28
32
49
41
DOC
(mg/L)
0.6
7.2
0.6
1.1
2.1
2.6
5.0
41
Ca2+
74
137
77
92
96
103
117
41
Mg2+
47
155
53
62
64
67
77
41
Na+
9
47
12
18
21
23
26
41
K+
7
29
8
11
12
14
29
52
A'td
C"9/L)
12
720
12
39
104
212
424
52
Alima
C«g/i-)
0
496
0
33
136
270
463
50
Measured or estimated values.
2/1/89 through 1/31/90.
Missing values for period 11/89 through 1/90.
-------
Table 5-14. Median Values of Discharge, pH, ANC, Major Ions, DOC, Altd, and AIIm for Streamflows £ 20th Percentile during the
ERP*
Stream
Bald Mountain Brook
Buck Creek
Fly Pond Outlet
Seventh Lake Inlet
Biscuit Brook
Black Brook
E. Branch Neversink
High Falls Brook
Baldwin Creek
Benner Run
Linn Run
Roberts Run
Stone Run
Q
(m3/s)
0.01
0.01
0.01
0.02
0.07
0.02
0.04
0.07
0.05
0.06
0.06
0.05
0.02
PH
6.59
6.20
7.22
6.19
6.31
6.82
4.91
6.87
6.44
6.21
6.40
6.05
6.02
ANC
65
42
290
46
33
147
-9
147
47
27
47
15
12
so/-
103
140
109
132
134
140
115
135
198
64
217
148
132
N03-
30
26
32
24
26
37
14
26
15
9
21
6
3
cr
9
11
28
11
20
17
14
17
45
91
30
34
25
DOC
(mg/L)
3.2
4.6
3.0
6.2
1.4
1.1
2.4
1.8
2.3
2.2
2.5
2.7
1.3
Ca2+
111
131
284
138
135
266
62
246
179
84
211
114
94
Mg2+
50
43
90
49
51
59
45
52
85
46
64
75
54
Na+
54
52
102
42
18
12
16
16
22
48
25
24
24
K+
9
10
14
9
5
5
6
6
18
22
11
19
13
Al,d
(«g/L)
125
254
99
284
23
12
157
15
20
25
22
41
28
Al b
Alim
(«9/L)
43
47
21
59
3
13
58
0
11
0
0
20
10
Units = /ieq/L unless otherwise specified.
Measured or estimated values.
-------
Table 5-15. Median Values of Discharge, pH, ANC, Major Ions, DOC, Altd, and Al,m for Streamflows > 95th Percentile during the
ERPa
Stream
Bald Mountain Brook
Buck Creek
Fly Pond Outlet
Seventh Lake Inlet
Biscuit Brook
Black Brook
E. Branch Neversink
High Falls Brook
Baldwin Creek
Benner Run
Linn Run
Roberts Run
Stone Run
Q
(m3/s)
0.44
0.70
0.12
1.79
1.90
0.66
10.53
1.25
0.42
0.85
1.15
0.99
0.97
pH
4.77
4.57
6.19
4,78
5.38
6.10
4.50
6.04
5.89
5.23
5.12
4.71
4.80
ANC
-10
-22
46
-4
8
50
-25
34
8
-3
-4
-25
-16
S042~
118
126
125
132
115
122
107
116
208
121
210
189
209
N03-
69
70
56
66
62
50
48
38
17
7
19
6
2
cr
9
10
29
11
15
17
15
17
32
40
23
22
26
DOC
(mg/L)
4.6
7.5
6.0
7.2
3.6
3.7
5.2
4.1
2.4
3.0
2.4
2.9
2.8
Ca2+
93
111
173
135
132
181
57
164
160
75
133
84
94
Mg2+
27
29
43
37
48
40
40
39
83
50
53
67
64
Na+
21
22
49
22
13
10
10
12
19
28
17
16
18
K+
12
13
14
12
8
11
11
9
21
18
11
17
14
A'td
C«g/i-)
586
778
367
589
149
115
420
119
52
97
409
172
328
Alimb
C"g/L)
331
375
21
225
70
34
347
21
19
68
387
196
426
ol
Units ^eq/L unless otherwise specified.
Measured or estimated values.
-------
Table 5-16. Number of Episodes for Which Ion Changes Contributed to ANC Depressions
and Mean Rank of Ion Change Contributions (1 = most important) to ANC
Depressions
Region/Stream
Adfrondacks
Bald Mountain Brook
Buck Creek
Fly Pond Outlet
Seventh Lake Inlet
Catskllls
Biscuit Brook
Black Brook
East Br, Neversink River
High Falls Brook
Pennsylvania
Baldwin Creek
Benner Run
Linn Run
Roberts Run
Stone Run
No.
Episodes3
14
18
21
18
23
11
•20
25
13
21
25
24
23
No. Episodes Ion Changes
Contributed to ANC Depressions
so42-
10
10
18
9
1
4 .
7
5
8
19
15
10
20
NO-f
5
15
11
13
20
9
20
13
7
4
12
10 ,
11
cr
9
12
8
8
5
4
9
7
6
3
7
7
11
A~
12
17
14
16
21
9
19
22
5
13
17
20
19
2CB
14
17
21
17
18
11
15
23
7
10
24
21
11
Al
0
1
4
0
0
0
0
2
3
2
1
3
2
Mean Rank of Importance15 of Ion
Changes to ANC Depressions
so42-
2.6
3.1
2.3
2.3
3.0
3.2
3.4
2.0
1.1
1.1
2.3
2.1
1.5
N03-
2.4
2.4
2.8
2.3
1.5
2.2
2.2
1.9
2.4
3.0
3.2
2.8
3.3
cr
3.9
4.3
3.6
3.9
3.4
3.5
3.4
3.4
3.0
2.0
3.0
3.1
3.4
A~
2.2
1.9
2.8
1.8
2.3
2.6
1.9
2.4
2.6
2.3
2.8
2^2
2.6
ZGB
1.4
1.8
1.0
2.0
1.7
1.1
1.8
1.2
1.3
1.9
1.1
1.1
1.5
Al°
-
3.0
4.0
-
-
—
3.5
2.3
3.0
2.0
3.0
3.5
* With enough ion measurements to perform analysis.
b Mean rank for episodes in which ion changes made contributions to ANC depressions.
c - «B no episodes measured in which Al ion changes contributed to ANC depressions.
176
-------
Table 5-17. Rank (1 = most important) of Ion Changes Contributing to ANC Depressions
During Episodes with Lowest Minimum ANC Values in Adirondack Streams
Stream
Bald Mountain Brook
Buck Creek
Fly Pond Outlet
Seventh Lake Inlet
Start Datea
3/12/90
3/27/89
5/16/90
9/20/89
11/16/89
11/16/89
3/12/90
4/1/90
5/16/90
4/3/89
3/27/89
3/12/90
4/10/90
4/4/89
4/3/90
3/12/90
3/27/89
11/16/89
5/16/90
5/6/89
Hydrologic
Stimulus6
Snowmelt
Snowmelt
Rain
Rain
Rain
Rain
Snowmelt
Snowmelt
Rain
Snowmelt
Snowmelt
Snowmelt
Snowmelt
Snowmelt
Snowmelt
Snowmelt
Snowmelt
Rain
Rain
Rain
ANC
Minimum
(A*q/L)
-21
-20
-13
-13
-6
-30
-29
-28
-26
-25
17
19
25
31
39
-20
-18
-9
-9
-8
Rank of Ion Changes Contributing to
Episodic ANC Depressions
2CB
2CB
.2CB
2CB
2CB
A~
N03-
2CB
A~
2CB
2CB
SCB
2CB
2CB
2CR
2CB
2CB
A-
A~
A~
N03-
A-
A~
S042-
A-
N03-
2CB
A~
2CB
A"
N03-
A~
A~
so42-.
so42-
N03-
A~
N03-
SCB
2CB
A~
NO3~
—
A~
cr
2CB
A-
N03-
N03-
Al
A~
N03-
so42-
N037A-
N037A~
A~
N03-
2CB
N03-
N03-
.c
4
—
cr
—
, —
—
or
—
cr
—
cr
- —
—
—
—
—
—
—
cr
* Episodes with incomplete ion chemistry not included in analysis.
b Snowmelt includes rain-on-show events.
c — = no ion change that contributed to ANC depression.
177
-------
Table 5-18. Rank (1 = most important) of Ion Changes Contributing to ANC Depressions
During Episodes with Lowest Minimum ANC Values in Catskill Streams
Stream
Biscuit Brook
Black Brook
East Branch
Neversink River
High Falls Brook
Start Date"
4/10/90
5/5/89
1/25/90
9/19/89
3/12/90
10/17/89
5/5/89
4/10/90
9/18/89
11/16/89
10/18/89
9/19/89
4/10/90
1/25/90
10/31/89
5/5/89
10/19/89
1/25/90
2/22/90
4/10/90
Hydrologic
Stimulus'5
Rain
Rain
Snowmelt
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Snowmelt
Rain
Rain
Rain
Snowmelt
Rain
Rain
ANC
Minimum
Cueq/L)
-6
-2
-2
-2
-2
17
18
25
36
36
-45
-39
-37
-34
-32
12
21
25
27
28
Rank of Ion Changes Contributing to
Episodic ANC Depressions
1
N03-
2CB
N03~
A"
N03-
2CB
scB
2CB
2CB
2CB
A"
A"
N03-
N03-
A-
2CB
2CB
N03-
2CB
2CB
2
2CB
N03-
A-
N03-
A-
N03-
NQ3-
NCvf
A"
N03-
2CB
2CB
2CB
A~
SCB
N03-
N03-
SCB
N03-
NO3~
3
A"
A"
2CB
2CB
—
A-
A-
A-
N03-
A-
N03-
so42-
A"
2CB
N03-
A~
A~
A~
A-
A~
4C
—
—
—
—
—
—
—
—
—
—
N03-
—
so42-
—
—
—
—
—
* Episodes with incomplete ion chemistry not included in analysis.
b Snowmelt includes rain-on-snow events.
0 _ s no ion change that contributed to ANC depression.
178
-------
Table 5-19. Rank (1 = most important) of Ion Changes Contributing to ANC Depressions
During Episodes with Lowest Minimum ANC Values in Pennsylvania Streams
Stream
Baldwin Creek
Benner Run
-
Linn Run
Roberts Run
Stone Run
Start Date*
3/30/89
5/6/89
6/20/89
3/24/89
2/15/89
11/16/89
5/14/89
5/9/89
2/2/90
2/15/90
3/4/89
4/17/89
11/16/89
2/14/89
4/1/90
6/20/89
6/3/89
6/25/89
2/2/90
3/29/89
6/3/89
1/16/90
2/2/90
2/15/90
4/1O/90
Hydrologic
Stimulus
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
Rain
ANC
Minimum
Ueq/L)
-1
4
4
5
6
-15
-10
-9
-6
-5
-17
-15
-13
-11
-10
-52
-45
-39
-33
-28
-41
-38
-33
-31
-25
Rank of Ion Changes Contributing to
Episodic ANC Depressions
1
SCB
SCB
A~
so42-
SCB
so42-
so42-
so42-
SCB
so42-
SCB
SCB
SCB
SCB
so42-
SCB
SCB
SCB
SCB
SCB
S042-
so42-
A~
so42-
SCB
2
cr
N03-
N03-
Al
so42-
A"
A~
2CB
so42-
A"
N03-
—
N03-
so42-
SCB
A"
so42-
A-
A~
A~
SCB
SCB
so42-
SCB
so/-
3
—
cr
cr
cr
—
—
—
A~
—
2CB
so42-
—
A~
N03-
A~
—
A~
N03~
Al
—
A-
A"
—
N03-
—
4b
—
—
—
—
—
—
—
—
—
A~
—
ci-
—
N03~
—
cr
—
—
—
—
—
—
—
—
* Episodes with incomplete ion chemistry not included in analysis.
b — = no ion change that contributed to ANC depression.
179
-------
-------
SECTION 6
EFFECTS ON FISH
In Section 6, we describe results from studies of the effects of stream chemistry and episodic
acidification on fish. Section 6.1 presents results from in situ bioassays, which measure direct
toxic effects on individual fish. Section 6.2 discusses fish movements in response to stream
chemistry, based on radiotelemetry tracking of individual fish locations and fish traps that
monitored upstream and downstream fish movements. Section 6.3 analyzes results from surveys
of fish community composition and brook trout abundance in relation to stream chemistry.
Summary conclusions regarding the effects of episodic acidification on fish populations in ERP
streams are presented in Section 7.
6.1 DIRECT TOXICITY
In situ bioassays were conducted each fall and spring (fall 1988, spring 1989, fall 1989, spring
1990) during periods when major events and episodic acidification were expected to occur.
Tables 6-11 through 6-6 summarize fish mortality and chemical conditions during each experi-
ment. Figures 6-11 through 6-3 illustrate cumulative changes in fish mortality and changes in
stream chemistry through time during three example bioassay periods. As discussed in Section
4.4.1, analyses of bioassay results focus on the cumulative percent mortality at the end of 20
days.
In situ bioassays measure the toxic effects of stream chemistry during the bioassay period. They
also serve as indicators of the relative toxicity of ERP study streams. As expected, median fish
mortality was significantly (p < 0.05) lower in reference streams (Fly Pond Outlet, Black Brook,
High Falls Brook, Benner Run, and Baldwin Creek) than in nonreference streams (Table 6-7;
Figure 6-4). Mortality after 20 days never exceeded 20% in reference streams, but ranged
between 0 and 100% in nonreference streams. In general, higher levels of mortality occurred in
streams with more severe chemical characteristics, as described in Section 5.2.1. For example,
maximum levels of brook trout mortality measured during ERP bioassays were > 80% in Buck
Creek, East Branch Neversink River, Stone Run, and Linn Run; between 40% and 50% in Bald
Mountain Brook and Roberts Run; and 20% to 30% in Seventh Lake Inlet and Biscuit Brook
(Tables 6-1 to 6-3). In most streams, bioassay periods did not coincide with the most severe
Since this section contains so many tables and figures, we have placed them all at the end of the section. Figures
appear first, beginning on page 200, then tables, beginning on page 261.
181
-------
chemical conditions (lowest pH, highest Alim; see Figures 5-1 to 5-13); thus, the maximum mortal-
ity levels recorded are a conservative estimate of maximum stream toxicity. In every stream, low
levels of mortality (< 10%) occurred during at least some bioassays, indicating that chemical
conditions were not toxic throughout the year.
We analyzed the bioassay results to determine the specific chemical conditions associated with
high levels offish mortality. Three issues were addressed: (1) the association between fish
mortality and a simple classification of stream chemistry based on ANC, (2) the chemical variables
most closely correlated with levels of fish mortality, and (3) the influence of exposure magnitude
and duration on fish mortality.
Bioassay periods were classified according to ANC as nonacidic (ANC always >
chronically acidic (ANC always :£ 0 ^ueq/L), and acidic episode (initial ANC > 0 fteq/L with at least
two consecutive values < 0 fieq/L during the 20-day period). Fish exposed to acidic episodes
experienced significantly (p < 0.10) higher mortality than did fish in nonacidic bioassays; mortality
levels were similar (not significantly different) in bioassays with acidic episodes versus those in
the chronically acidic class (Table 6-8; Figure 6-5). Results were consistent across all fish
species, and for common pool fish as well as common pool and resident fish combined.
What chemical variables appear most important in controlling the level of fish mortality? We
examined three: pH, Aljm, and various forms of the acidic stress index (ASI), an integrated index
that combines pH, Alim, and Ca. The ASI was developed by J. Baker et al. (1990a) from labora-
tory bioassay data; it is equivalent to the expected fish mortality (percent) caused by exposure to
constant levels of a given pH, Alim, and Ca combination in a controlled laboratory environment
(see Section 2.4.2). Three different forms of the ASI were evaluated:
1. ASIt, based on a laboratory bioassay with brook trout swim-up fry.
2. ASIS, based on a laboratory bioassay with rainbow trout swim-up fry.
3. ASIt,ad, based on a laboratory bioassay with adult brook trout.
Equations for calculating these indices are presented in Table 6-9. Equation forms and coeffici-
ents were estimated using regression analysis by J. Baker et al. (1990a). Observe that ASIt.ad is
calculated from Aljm alone; neither pH nor Ca were statistically significant (p < 0,05).
Simple summary statistics were calculated for each chemical variable for each bioassay period:
time-weighted median values for pH, Alim, and each ASI; minimum pH; and maximum values for
182
-------
Alim and each ASI. Both median Aljm and Iog10[median Aljm] were considered. In addition,
separate analyses were conducted for the observed percent mortality and an adjusted mortality,
corrected for background mortality using Abbott's formula (Ashton, 1972; Hewlett and Plackett,
1979):
Q = (Q' - C)/(1 - C)
(6-1)
where Q = the adjusted percent fish mortality (i.e., the percentage of fish expected to die during
the bioassay if no other sources of mortality were operating besides the acidification-related
chemical stress).
Q' = the observed percent mortality.
C = the "control" mortality in the absence of toxicant stress.
Because the ERP bioassays included no true control, C was defined as the minimum observed
mortality in a given bioassay set (i.e., bioassays initiated at approximately the same time, with the
same species and fish source, and in a given region).
The association between fish mortality and each chemical variable was evaluated using bivariate
plots (e.g., Figures 6-6 to 6-14) and regression analysis (ordinary least squares regression and
maximum likelihood logistic regression). All plots indicated a relatively high level of scatter. How-
ever, overall, median Alim (untransformed) was the best predictor of fish mortality. The logistic
model of fish mortality as a function of median Alim had the highest regression coefficient (r2) for
brook trout and blacknose dace and second highest r2 for sculpin (Table 6-10). For sculpin,
ASIt.ad, calculated from Alim (see Table 6-9), provided a slightly better fit (r2 = 0.63) than median
Aljrn (r2 = 0.61). Adjustments for background mortality resulted in only minor model improvement
(see Table 6-10). Based on these results, subsequent analyses focused on more complex
expressions of fish exposure to Aljm and used only observed (unadjusted) fish mortality.
Previous studies, reviewed in Section 2.4.2, suggest that both exposure magnitude (concentra-
tion) and duration (days of exposure) are important characteristics that influence the severity of
effects of episodes on fish. Higher median concentrations of Aljm were associated with higher fish
mortality in ERP bioassays, as demonstrated by the plots and regression analyses just discussed
(Figures 6-6, 6-11, and 6-13; Table 6-10). In addition, fish exposed for longer periods of time to a
183
-------
given Alim concentration generally had higher mortality rates (Figures 6-15 and 6-16). However,
any given bioassay can be characterized by a continuum of concentration-duration combinations,
as illustrated in Figures 6-17 to 6-22, and it is difficult to distinguish what specific concentration or
duration results in increased fish mortality. Furthermore, patterns of chemical conditions during
bioassays with high mortality were highly variable (Figure 6-23).
We examined five alternative expressions of aluminum exposure:
1 . Time-weighted median Aljm concentration.
2. Maximum Al!m concentration.
3. The total duration of time (days) during the bioassay when Aljm concentrations
exceeded three different threshold concentrations: 100, 200, and 300 ag/L (TDUR100,
TDUR200, TDUR300).
4. The maximum sustained period of time when Aljm levels exceeded the three threshold
concentrations (> 100, 200, and 300 fig/L) (MDUR100, MDUR200, MDUR300).
5. An integrated function of concentration and duration (INT), essentially the area or
integral under the curves in Figure 6-23 above each threshold Aljrn concentration (100,
200, and 300 /zg/L). Because of the wide range of values for this integrated concen-
tration-duration function, values were Iog10 transformed (LINT100, LINT200, LINT300).
Analyses for blacknose dace also considered the entire area under the curve (Ali(
0 figjL, LINTO).
im
All calculations of exposure were based on linear interpolations between measured or estimated
Al,m levels in water chemistry samples, as described in Section 4.4.1.
Multivariate logistic regression analyses were used to identify which of these expressions of
chemical exposure were important predictors of fish mortality. Median and minimum pH and
median Ca and DOC were included as additional independent variables. Laboratory bioassays
have shown that high levels of Ca and DOC can reduce the toxicity of Alim (Brown, 1983;
Peterson et al., 1989). The toxic effects of low pH may contribute to fish mortality; also Aljm
toxicity varies with pH (Fivelstad and Leivestad, 1984; Clark and LaZerte, 1985; Sadler and
Lynam, 1987). Comparisons among alternative models were based on the deviance statistic for
logistic regression [-2(log likelihood function); Agresti, 1990], the variance of the residuals
(predicted minus observed mortality), a logistic-regression analog for r2 (Agresti, 1990) equivalent
to the proportion of the variation in mortality explained by the model, and results from stepwise
regression analyses (SAS Institute, 1989). Analyses were conducted for common pool brook
trout bioassays (all three regions combined), blacknose dace (Adirondacks only), and sculpin
(slimy and mottled sculpin combined in both Catskill and Pennsylvania streams). Region was
184
-------
included as an independent variable for brook trout and sculpin analyses, to determine whether
mortality-chemistry relationships varied among the ERP regions. Sculpin analyses included a
species variable to distinguish between slimy and mottled sculpin.
For brook trout, stepwise regression selected large numbers of variables (> 10) as statistically
significant (p < 0.05) predictors of mortality. However, models with more than 3-5 variables were
highly unstable, that is, different sets of predictor variables yielded models with very similar
statistics (deviance, variance of residuals, and r2). The best 1-variable model predicted brook
trout mortality as a function of median Aiim, with a model r2 value of 0.59 (Table 6-11; Figure
6-24). The best 3-variable model predicted mortality as a function of median Al,m, minimum pH,
and median Ca, with r2 = 0.72 (Table 6-12). The best 5-variable model predicted mortality as a
function of median Alim, minimum pH, median Ca, median DOC, and a Catskill regional variable
(indicating lower mortality in Catskill bioassays than in Adirondack or Pennsylvania bioassays at a
given combination of Aljm, pH, Ca, and DOC levels), with r2 = 0.79 (Table 6-13). As an alternative
to logistic regression, we also conducted a stepwise, weighted ordinary least squares regression,
using the logit transformation of percent mortality and weighted by the inverse of the variance of
the logit-transformed mortality (McCullagh and Nelder, 1989). A 3-variable model was selected,
predicting mortality as a function of median AIim, minimum pH, and median Ca, with parameter
estimates similar to those in the 3-variable logistic regression model (Table 6-14).
Observe that none of the Aljryi duration variables (TDUR and MDUR) or integrated concentration-
duration variables (LINT) were selected as significant predictors of brook trout mortality except in
higher order models (> 5 predictor variables). Also, as expected, Aljm is positively associated
with brook trout mortality (higher mortality at higher Aljmj and Ca is negatively associated (lower
mortality with higher Ca). However, somewhat unexpectedly, pH is positively associated with
mortality (higher mortality with higher pH, for given levels of Aljm and Ca), suggesting that the
primary influence of pH on brook trout is as a modifier of Aljm toxicity, rather than a direct
contributor to toxic effects (at least within the range of conditions that occurred within ERP
bioassays).
For blacknose dace, stepwise logistic regression selected only a single-variable model, predicting
percent mortality as a function of the integrated function of Al concentration and duration for Aljm
concentrations > 200 ^g/L (r2 = 0.97; Table 6-15; Figure 6-25). No other variables were signifi-
cant at p < 0.05 after LINT200. Based on the criterion of minimum deviance, the LINT200 vari-
able provided a distinctly better model fit than any other single-variable model or any multiple-
185
-------
variable model relying solely on simple summary statistics such as median or minimum/maximum
*
inorganic Al or pH.
For sculpin, stepwise regression selected a 4-variable model predicting sculpin mortality as a
function of median Alim> a regional variable (distinguishing between Catskill and Pennsylvania
bioassays), a species variable (distinguishing between slimy and mottled sculpin), and an Alirn
duration variable, TDUR300, the total duration of time with Alim > 300 /tg/L. However, the model
improvement gained by adding variables 3 and 4 (species and TDUR300) was relatively small;
model r2 values are 0.80, 0.81, and 0.83 for the best 2-, 3-, and 4-variable models, respectively.
Thus, as for brook trout, variables that explicitly account for exposure duration (TDUR, MDUR, and
LINT) were relatively poor predictors of sculpin mortality in ERP bioassays. Median Alim was the
single best predictor (r2 = 0.61), although mortality levels were substantially higher in
Pennsylvania bioassays than in Catskill bioassays at a given median Alim concentration (Figures
6-26 and 6-27). Table 6-16 provides parameter estimates and model statistics for the 2-variable
model, predicting sculpin mortality as a function of median Alim and region.
Simonin et al. (1993) conducted similar multivariate regression analyses using ERP in situ bio-
assay data for the Adirondacks, although with some differences in specific predictor variables
considered. Brook trout mortality was best correlated with a 2-variable model that included DOC
and a concentration-duration variable for Alim (median Alim during the episode times the duration
of the episode). The best overall model for blacknose dace included only one variable, median
Alim. They also suggested that the time to initiation of an episode was important. Bioassay fish
that had been in streams 15-24 days before the occurrence of an episode survived better than
fish exposed to an episode soon after the start of a bioassay. Longer times prior to episodic
acidification may allow fish to recover from handling stress and acclimate to stream chemistry.
What chemical variables and exposure characteristics are most important in controlling the effects
of episodes on fish mortality? The ERP database is unique—fish bioassays were conducted in a
relatively large number of streams (n = 13) during multiple seasons in combination with intensive
chemical monitoring. However, despite the quality of the database, our ability to address this .
question is still limited. Episode chemistry is highly variable and difficult to characterize.
Furthermore, many of the key features (pH, Al, episode magnitude, and duration) are highly corre-
lated in the field, making it difficult to distinguish effects of individual factors. Controlled experi-
ments are needed to better delineate the relative importance of episode magnitude, duration, total
186
-------
r
exposure (concentration times dose), timing of exposure, and each chemical variable and their
interactive effects. However, analysis of the ERP database has demonstrated the following:
• Chemical conditions toxic to fish occurred at some time during the study in all ERP
streams except the five reference streams (Fly Pond Outlet, Black Brook, High Falls
Brook, Benner Run, and Baldwin Creek). Maximum observed mortality rates were gen-
erally highest in those streams with the most severe chemical conditions: > 80% in
Buck Creek, East Branch Neversink River, Stone Run, and Linn Run; between 40% and
50% in Bald Mountain Brook and Roberts Run; and 20% to 30% in Seventh Lake Inlet
and Biscuit Brook (Tables 6-1 to 6-3).
• All streams also had some bioassays with low mortality (< 10%), suggesting that toxic
conditions do not occur throughout the year.
• Episodic acidification can be toxic to fish. Mortality rates were significantly higher
during bioassays with acidic episodes than in bioassays with ANC > 0, but similar (not
significantly different) in chronically acidic bioassays (ANC < 0 throughout the
bioassay) and bioassays with acidic episodes (Figure 6-5).
• Inorganic Al was the most important chemical variable for predicting fish mortality. Ca,
pH, and DOC were also important predictors of brook trout mortality (Tables 6-13 and
6-14).
• Fish exposed for longer periods to high Aljm had higher mortality rates (Figures 6-15
and 6-16). However, somewhat surprisingly, a simple summary statistic, time-weighted
median Alim, usually provided as good or better predictions of fish mortality than did
more complex expressions of chemical exposure incorporating peak levels as well as
duration. For brook trout and sculpin, median Aljm was the best single-variable pre-
dictor of mortality (Tables 6-11 and 6-16).
• Mortality of blacknose dace was predicted best by an integrated function of duration
and Aljm concentrations > 200 figlL. High mortality (> 50%) is predicted to occur at
LINT > about 2.0 (Figure 6-25), equivalent to an Aljm concentration of 300 jug/L for one
day or 250 ^wg/L for two days.
• Based on their toxic responses to Aljm, the relative sensitivity of the fish species tested
in ERP bioassays, from most to least sensitive, is blacknose dace > brook trout > mot-
tled and slimy sculpins. Median Aljm levels predicted to cause 50% mortality in in situ
bioassays are about 120 ,ag/L for blacknose dace, 200 fig/L for brook trout, 200 /tg/L
for sculpin in Pennsylvania streams, and 300 fig/L for sculpin in Catskill streams.
• For both brook trout and sculpin, relationships between mortality and stream chemistry
were significantly different among ERP regions. (Blacknose dace bioassays were con-
ducted in only one region, the Adirondacks.) Reasons for inter-regional differences are
not known, but could include differences in the sensitivity of fish strains, the acclimation
of native fish to low ionic strength and acidic conditions, stream chemistry not
accounted for in our analyses, or experimental protocols.
187
-------
6.2 FISH MOVEMENT
Chemical conditions within low-order, softwater stream systems, such as the ERP study sites, can
be quite variable spatially. ANC and pH levels often increase gradually as waters flow down a
drainage system (e.g., Driscoll et al., 1987b; Kaufmann et al., 1988). Even within a given stream
reach, groundwater inflows and tributaries may provide microhabitats with chemistry distinctly
different, often more alkaline, than the main stream. If fish move into such areas during episodes,
the chemistry measured at the ERP continuous monitoring site may be a poor index of the chemi-
cal conditions to which free-ranging fish are actually exposed.
The locations of individual brook trout were tracked using radiotelemetry (Section 4.4.2.1) to
determine whether fish respond to episodic acidification by moving downstream or into more
alkaline microhabitats. In selected streams in the Adirondacks, fish traps were also used to
monitor up- and downstream fish movements in relation to stream chemistry (Section 4.4.2.2).
This section summarizes results from these studies as analyzed by and presented in Gagen
(1991) for Pennsylvania [also, Carline et al. (1992) and DeWalle et al. (1991)], by Murdoch et al.
(1991) for the Catskills, and by D. Bath [Adirondack Lake Survey Corporation (ALSC), Ray Brook,
New York, 1992, pers. comm.] for the Adirondacks [also Kretser et al. (1991)]. The objective was
to evaluate how fish behavior influenced the effects of episodic acidification on fish populations in
the ERP study streams.
A net downstream movement of radio-tagged brook trout occurred in all streams and study peri-
ods that experienced stressful chemical conditions (Table 6-17), but little to no net downstream
movement occurred in streams with relatively high pH (> 5.1-5.2) and low Alim (< 150-160 ^g/L)
throughout the study period (Table 6-18). Downstream fish movement either was associated with
chronically acidic conditions at the start of the experiment (pH < 5.2 and/or Alim > 150-200 fig/L)
or coincided with the occurrence of one or more major episodes (Aljm > 160 ^g/L for 1.5 or more
days; see Table 6-17). Minor episodes (such as in Stone Run, fall 1989, with a maximum Alim of
156/ig/L) caused no discernible net downstream movement (Table 6-18).
Telemetry studies were generally conducted concurrently with one or more in situ bioassays (see
Section 6.1). All study periods in which bioassays indicated high levels of brook trout mortality
(5: 50%) had significant net downstream movement of radio-tagged brook trout (Buck Creek,
spring 1990; East Branch Neversink, spring 1989; Linn Run, spring 1989; and Stone Run, spring
1990; Table 6-17). However, net downstream movement also occurred when bioassays indicated
188
-------
nontoxic conditions (3-10% mortality; Bald Mountain Brook and Buck Creek, fall 1989; Linn Run
1988; Table 6-17). Thus, brook trout appear to exhibit behavioral responses even when exposed
to sublethal pH and Al levels. Study periods with no net downstream movement also had rela-
tively low mortality rates as indicated by bioassays (0-23%; Table 6-18).
Results from the fish traps in Bald Mountain Brook and Fly Pond Outlet (Adirondacks) were hot as
clear, however. No brook trout were caught in the traps at pH < 5.0, moving up- or downstream
(Tables 6-19 and 6-20). Furthermore, the largest monthly downstream movement of trout was
recorded in May 1989 in Fly Pond Outlet at pH > 6.0, presumably as a result of the brook trout
stocking that occurred in April 1989 as part of the fish transplant experiments (see Table 4-17).
Fifty percent of the trout caught in the trap in May had been transplanted into Fly Pond Outlet in
April. At pH > 5.0, brook trout were caught moving both up- and downstream, although more
than twice as many fish moved downstream as upstream. Likewise, most (95%) blacknose dace
caught in the traps were moving downstream (Table 6-19). The largest monthly downstream
movement of dace occurred.in Bald Mountain Brook in May 1989, following stocking of dace in
April and coincident with an early May episode that lasted 15 days.
Gagen (1991) conducted a detailed analysis of individual fish responses and chemical exposures
during radiotelemetry studies in Linn Run (spring 1989) and Stone Run (spring 1990), Pennsyl-
vania. Both study periods had significant net downstream fish movement (Figures 6-28 and 6-29)
and high levels offish mortality in bioassays (91-93%); also, one-third of the radio-tagged fish in
each stream were recovered dead during the study period. In contrast, no radio-tagged fish were
recovered dead in other radiotelemetry studies in Pennsylvania and mortality rates in coincident
bioassays were relatively low (0-23%).
In Linn Run, spring 1989, total dissolved aluminum (Altd) levels at the start of the study were
relatively low (Altd < 100 ,«g/L) but major episodes, with Altd > 350 fig/L at the continuous
monitoring station (CMS), occurred on days 3-5, 12-15, 18-20, 22-24, and 26-29 (Figure 6-28).
By day 4, one day after the onset of the first episode, radio-tagged fish occurred in locations with
Altd levels 100-200 ^ag/L less than at the CMS, although most fish remained within 200 m of the
CMS and were still exposed to relatively high Al levels (Altd > 200 ^g/L; Figures 6-30 and 6-31).
Two fish were recovered dead on day 6. By day 1 1 , before the second major episode, 3 of 12
radio-tagged trout had moved 500-700 m downstream into an area where ANC was higher rela-
tive to the mainstream. During the episode on day 12, three fish appeared in areas with relatively
low Altd (^ 100 ftg/L), whereas the remaining fish were exposed to Altd > 300 fig/L (Figure 6-31).
189
-------
One of the fish that avoided exposure to high Al had moved into an alkaline microhabitat within
200 m of the continuous monitoring site; another had moved almost 1,600 m downstream. Two
fish were recovered dead on day 12 and another on day 15, for a total of 5 dead fish out of the
15 released. In general, fish recovered dead were those exposed to high Altd levels during
episodes (Figure 6-32); fish that survived avoided exposure to peak Al levels (e.g., Figure 6-33).
No major episodes occurred during the spring 1990 telemetry study in Stone Run, but Al levels
remained relatively high throughout most of the study period (Altd 100-300 //g/L, Figure 6-29;
median Al,m 218/ig/L, Table 6-17). Fish were released at sites with ANC and Altd levels similar to
those at the CMS (Altd about 150 ^g/L). Two days after the release, many of the fish had moved
to areas with slightly lower Altd (60-100 ^g/L), at sites near the CMS (< 50 m downstream), to
almost 400 m downstream (Figure 6-34). Stream chemistry remained relatively stable through
day 8. By this time, 9 of 10 fish had moved 400 m or more downstream and 5 of these fish were
in areas with Altd < 50 ftglL. Three of the 10 fish released were recovered dead during the study
period: 2 on day 10 and 1 on day 16.
Murdoch et al. (1991) analyzed results from the spring 1989 telemetry study in the East Branch of
the Neversink River and High Falls Brook in the Catskills. Significant net downstream movement
occurred in East Branch Neversink, associated with chronically low pH (< 5.0) and elevated Al
levels (Altd 250-300/tg/L, Aljm 157-284 yug/L), while brook trout in well-buffered High Falls Brook
(pH > 6.0, Altd < 100/^g/L) exhibited a small net upstream movement (Figure 6-35). Water
samples collected at fish locations in High Falls Brook indicated relatively uniform chemical
quality, similar to that at the CMS. No significant springs or tributaries enter High Falls Brook
between the confluence of the stream with the West Branch of the Neversink and the waterfall 1
km above the CMS. The study area within the East Branch of the Neversink River, on the other
hand, contains several springs and tributaries with higher pH and ANC than the main channel
(Figure 6-36). Radio-tagged fish were released near the confluence of one such spring (pH 6.2)
and two fish remained at that junction throughout the study, in a pool with relatively high pH.
Likewise, in the fall 1989 telemetry study (also in the East Branch of the Neversink), several radio-
tagged fish moved into tributaries with higher pH than the main channel (Murdoch et al., 1991).
In the Adirondacks, significant net downstream movement (> 100 m) occurred in Buck Creek in
fall 1989 and spring 1990 associated with low pH (median values 5.2 and 4.8) and elevated Aljm
levels (median values 187 and 308 fig/L; Table 6-17). Fish began moving downstream within
three days of the start of the study in fall 1989; by day 16, all but two brook trout had migrated
190
-------
downstream into Seventh Lake (pH > 6.0; D. Bath, ALSC, pers. comm.). The two fish that
remained in Buck Creek appeared at the mouth of a small tributary. When pH levels in Buck
Creek increased to > 5.0, several of the trout that had migrated to Seventh Lake moved back into
the lower reaches of the creek, but remained downstream of the CMS. In spring 1990, four of six
radio-tagged trout moved downstream into Seventh Lake by day 6. The remaining fish migrated
into a small tributary with pH 6.2 and Altd < 200 fig/L, where they remained throughout the study
period.
Therefore, in each region and study, some radio-tagged fish were able to avoid exposure to the
high Al and low pH measured at the CMS by moving downstream or into more alkaline micro-
habitats. Aluminum levels measured at these fish locations were often substantially less than
those measured at the CMS (by 100-200//g Altd/L or more). However, other fish, and in most
cases the majority of radio-tagged fish, were exposed to relatively high, potentially lethal Al levels.
Two factors may limit the degree to which fish can avoid exposure to adverse chemical condi-
tions: (1) the availability of suitable microhabitats and other "refuge areas" with higher pH and
lower Al levels and (2) the ability of fish to detect and actively avoid stressful chemical conditions.
The availability of suitable refuge areas is highly variable among and within stream systems,
because of variations in soil and bedrock chemistry and hydrologic flowpaths (DeWalle et al.,
1991). Many small headwater streams have little or no groundwater input, and as a result, few if
any alkaline microhabitats. Other systems, such as the East Branch of the Neversink River, tend
to be more spatially variable (see Figure 6-36). The occurrence of numerous groundwater seeps
and more alkaline tributaries may be responsible, at least in part, for the maintenance of a
residual trout population in the East Branch of the Neversink despite low pH and high Al levels in
the stream during much of the year (see Section 6.3).
Fish movements into areas of higher pH and lower Al may result from either active or passive fish
behavior. Fish may be able to detect stressful chemical conditions and actively seek out suitable
refuge areas. As an alternative, fish stressed by exposure to low pH/high Al may move randomly,
but then remain in areas of higher pH/lower Al that they happen upon—a passive avoidance
behavior. Likewise, fish that are physiologically stressed may be less able to maintain their
position in the stream and more susceptible to being carried downstream by the current, a
passive behavior that would result in net downstream movement.
191
-------
The ERP data are not sufficient to distinguish between active and passive fish behavior. However,
Gagen (1991) hypothesized that the fish movements downstream and into alkaline microhabitats
were largely passive rather than active, directed avoidance behavior. He cites two observations in
support of this hypothesis. Radio-tagged fish in Linn Run and Stone Run did not take full advan-
tage of the available microhabitats to avoid exposures to toxic chemical conditions. Also, net
downstream movements resulted primarily from reduced upstream movement rather than
increased downstream movement, relative to fish in reference streams. Individual fish in reference
streams also moved substantial distances downstream, but these downstream movements were
offset by other individuals moving upstream, with little to no change in median fish location. If fish
behavioral responses are largely passive, rather than active, a lower percentage of fish would be
expected to successfully locate refuge areas and avoid exposure to episodic acidification.
We should be aware that the ERP telemetry studies involved relatively large fish (170-220 mm in
length; see Tables 4-15 to 4-17). Smaller fish, young-of-the-year, and especially swim-up fry are
probably less mobile and may be less able to avoid adverse chemical conditions.
Behavioral avoidance may both mitigate and accentuate the effects of episodic acidification on
fish populations in streams. If fish move into refuge areas during episodes and then return, fish
behavioral responses would be expected to reduce the overall effect on the fish population.
However, if fish move out of the stream system during episodes but do not return, or return in
smaller numbers, and if fish move in response to sublethal pH/AI levels, as suggested by the
bioassay-telemetry comparisons, then the population-level effects of episodic acidification would
be greater than effects predicted based on mortality tests alone. Downstream fish migration may
be an important mechanism of fish population loss, especially if upstream migrations are inhibited
by steep gradients, waterfalls, or temporary log and debris dams.
6.3 EFFECTS ON FISH POPULATIONS
The number of fish in a stream reach at any given time reflects the dynamic balance between fish
recruitment (reproduction and immigration) and fish loss (mortality and emigration). Episodic
acidification can increase fish mortality and emigration, as demonstrated in Sections 6.1 and 6.2,
and may also decrease fish reproduction (J. Baker et al., 1990). To what degree, therefore, does
episodic acidification result in long-term declines in fish population density and biomass in
streams?
192
-------
All ERP study streams except the East Branch of the Neversink River had chemical conditions
during low flow considered suitable for fish survival and reproduction. The median pH measured
during low flow (20th percentile low flow; see Section 5.2) was 4.9 in the East Branch of the
Neversink, but ranged from 6.0 to 7.2 in the other 12 ERP streams (Table 5-14). Median Aljm
concentrations during low flow were < 60 fig/L in all streams (Table 5-14). Thus, if stream
assessments were based solely on chemical measurements during low flow, we would predict
that all but one of the study sites should support a diverse and healthy fish community.
However, most streams had substantially lower pH and higher Alim during periods of increased
discharge (Figures 5-15 and 5-16). For example, median pH values associated with the 95th
percentile high flow were 4.5-4.8 In Buck Creek, Bald Mountain Brook, Seventh Lake Inlet, East
Branch of the Neversink, Roberts Run, and Stone Run (Table 5-15). Median Aljm levels during
high flow were > 300 ^ag/L in Buck Creek, Bald Mountain Brook, East Branch of the Neversink,
Linn Run, and Stone Run. We would expect these streams to support fewer fish and fewer fish
species compared to ERP streams that maintain relatively high pH and low Aljm during events.
For example, Fly Pond Outlet, Biscuit Brook, Black Brook, High Falls Brook, Benner Run, and
Baldwin Creek all had high flow median Aljm < 100/^g/L.
In Section 5.2.1, ERP study streams were ranked according to the overall severity of stream
chemistry during the study period, from most to least severe, as follows:
1 . The East Branch of the Neversink River was chronically acidic (median ANC < 0) and
had the most severe chemical conditions (lowest pH, highest Alim for the longest
duration).
2. Stone Run and Roberts Run were also chronically acidic, but pH and Aljm levels were
not as extreme as in the East Branch of the Neversink.
3. Buck Creek and Linn Run were not chronically acidic but had severe acidic episodes,
with low pH and high Aljm for long durations.
4. Bald Mountain Brook and Seventh Lake Inlet had episodes of moderate severity.
5. Biscuit Brook and Benner Run experienced acidic episodes but they were of short
duration, and as a result, would be expected to have only minor effects on fish
populations.
6. Fly Pond Outlet, Black Brook, High Falls Brook, and Baldwin Creek were classified as
nonacidic (with ANC always > 0, except for one brief excursion below 0 In Baldwin
Creek) and had relatively high pH and low Alim throughout the study period.
193
-------
We can compare fish community composition and brook trout abundance in each stream to these
qualitative rankings, as well as to summary statistics of stream chemistry during high flow, to
evaluate the effects of episodic acidification on fish populations.
Ten fish species were caught in ERP streams (Table 6-21). Species were classified as acid-sensi-
tive, intermediately sensitive, or acid-tolerant, based on data on the effects of low pH and high Al
on fish summarized by Baker and Christensen (1991). Acid-sensitive fish species (blacknose
dace, slimy sculpin, or mottled sculpin) were caught in six streams: Buck Creek, Black Brook,
High Falls Brook, Biscuit Brook, Benner Run, and Baldwin Creek. With the exception of Buck
Creek, these streams all maintain relatively high pH (median pH > 5.2) and low Alim (^ 70 fig/L)
during high flow (Table 5-15) and were ranked as 5 or 6 in terms of the severity of stream chem-
istry. Dace caught in Buck Creek occurred only in the lower section of the stream immediately
above Seventh Lake.2 In contrast, the East Branch of the Neversink River, Stone Run, and Linn
Run, three streams with severe chemical conditions during high flow (median Alim > 300 ^g/L;
rankings 1, 2, and 3, respectively), had only a single, acid-tolerant species, brook trout. Roberts
Run (ranking 2) also had only brook trout, except for a single brown trout specimen caught
during the study. All other streams had two or more fish species. These results support the
hypothesis that chemical conditions during high flow can eliminate acid-sensitive fish populations
and alter fish community composition.
Brook trout occurred in all 13 ERP streams. However, brook trout density and biomass were
lower in streams with lower pH and higher AIjm levels during high flow (Figures 6-37 to 6-40).
Rank correlation analyses (Steel and Torrie, 1980) indicate a significant association (p < 0.05)
between brook trout abundance and both median weekly and high flow chemistry values for both
pH and Alim (Table 6-22). Rank correlations with median weekly pH provided the highest corre-
lation coefficients (r = 0.85 for brook trout density and 0.89 for biomass), but in general, no single
variable stands out distinctly and consistently as the best predictor of brook trout density or
biomass. Correlations between weekly and high flow chemistry (see Figures 5-15 and 5-16) and
between pH and Aljm make it difficult to distinguish the relative importance of these chemical
features based on observational data alone.
study reach in Buck Creek was only 100 m above Seventh Lake, a oircumneutral lake with a diverse fish community
(see Section 3.1.1). It is likely that many of the acid-sensitive and intermediately sensitive fish caught in Buck Creek were
temporary migrants from Seventh Lake. The study reach in Seventh Lake Inlet was 700 m upstream of the lake.
194
-------
Brook trout density and biomass were also significantly (p < 0.001) correlated with the qualitative
rankings of stream chemical severity (rank correlation coefficients, r = 0.83 for brook trout density
and 0.92 for biomass; Figure 6-41). Statistical comparisons between ranks were not possible
because of the small number of streams per rank. Thus, streams were grouped as nonacidic
(ranking 6; n == 4)< episodically acidic (rankings 3-5; n = 6), and chronically acidic (rankings 1-2;
n = 3). Analysis of variance and the Tukey-Kramer multiple comparison test (Miller, 1986) pro-
vided the same results. Both brook trout density and biomass were significantly (p < 0.05) higher
in nonacidic streams than in streams that had acidic episodes or were chronically acidic. How-
ever, brook trout density and biomass were not significantly different in streams with episodic
acidification compared to chronically acidic streams. Average levels of brook trout density
(fish/0.1 ha) in the three groups were 215 for nonacidic streams, 82 in streams with acidic
episodes, and 46 in chronically acidic streams. Likewise, average values for brook trout biomass
(g/0.1 ha) were 2355 in nonacidic streams, 1256 in streams with acidic episodes, and 704 in
chronically acidic streams.
No regional differences were evident in the relationships between brook trout abundance and
stream chemistry (see Figures 6-37 to 6-41), although several site-specific variations deserve
notice:
• Linn Run stands out as a distinct outlier in plots of brook trout abundance as a function
of stream pH (median weekly and high flow pH; Figures 6-37 and 6-38). Brook trout
density and biomass in Linn Run (4 fish/0.1 ha and 253 g/0.1 ha) are lower than
expected given the measured median weekly pH of 5.97 and median high flow pH of
5.12. As discussed in Section 5, Linn Run has the flashiest response to storm events
(Figure 5-11), experiences the greatest increase in Aljn1 during episodes, and tends to
mobilize higher levels of Aljm at a given ANC and pH than do other ERP streams
(Figure 5-22). Aljm concentrations > 200 ,«g/L commonly occur in Linn Run at pH
levels < 5.4, but only at pH < 5.2 in Stone Run, Bald Mountain Brook, Buck Creek,
: and Seventh Lake Inlet, and-at pH < 5.0 in Roberts Run and East Branch of the
Neversink River. [No Aljm values > 200 ,«g/L were recorded for the remaining six ERP
streams.] These differences in the pH-AI relationship probably account for the
occurrence of Linn Run as an outlier in trout abundance-stream pH relationships. We
believe that low levels of brook trout density and biomass in Linn Run result primarily
from the very high concentrations of Aljm that occur in the stream during storm events
(maximum value, 805 figlL; Table 5-11).
Bald Mountain Brook is an outlier on plots of brook trout abundance as a function of
both high flow Aljm and high flow pH (Figures 6-38 and 6-40), although median weekly
Aljm and pH levels provide reasonable predictions of trout density and biomass in the
stream. Trout density and biomass in Bald Mountain Brook (159 fish/0.1 ha and 1609
g/0.1 ha) are higher than expected given median high flow values of pH 4.77 and Aljm
331 fig/L. Reasons for the relatively high fish abundance are uncertain, although two
factors may play some role. Visual comparisons of Figures 5-1 to 5-13 suggest that
chemical conditions in Bald Mountain Brook recover (i.e., return to near baseflow levels
195
-------
of pH and Alim) more rapidly following events than in other ERP streams, such as Buck
Creek, Seventh Lake Inlet, and Roberts Run. Thus, fish in Bald Mountain Brook may
be exposed to adverse chemical conditions for relatively short periods of time. In
addition, immediately below the ERP study site, Bald Mountain Brook merges with Fly
Pond Outlet, a stream with circumneutral pH and high trout density (292 fish/0.1 ha).
Fish may migrate from Fly Pond Outlet to Bald Mountain Brook, particularly during low
flow periods. The potential importance of spatial variability within the drainage system
is discussed further later in this section.
• Brook trout density and biomass were distinctly higher in Black Brook (309 fish/0.1 ha,
3469 g/0.01 ha) and Fly Pond Outlet (292 fish/0.01 ha, 2364 g/0.01 ha) than in other
streams, such as Baldwin Creek (133 fish/0.1 ha, 1759 g/0.01 ha) and High Falls Brook
(127 fish/0.01 ha, 1827 g/0.1 ha), with similar chemistry (i.e., relatively high pH and low
Alim even during events). Such differences may reflect the dominant influence of other
habitat features on trout productivity in streams with suitable acid-base chemistry. For
example, Fly Pond Outlet and Black Brook had the highest Ca levels (median weekly
concentrations of 235 and 258 fieq/L, respectively) among ERP streams (compared to
66-209 fteq/L in other ERP streams; see Tables 5-1 to 5-13).
The above analyses are based on average values for trout density and biomass in each stream,
calculated from fish surveys conducted in fall 1988, spring 1989, fail 1989, and spring 1990. Fish
abundance varied among sampling dates (Table 6-23), although in general, the pattern of inter-
stream differences in trout density and biomass were relatively consistent over time (Figures 6-42
to 6-44).
Brook trout were transplanted into ERP study reaches to establish consistent, initial levels of fish
biomass in all streams, by region (see Section 4.4.4). These transplants had short-term effects
(over several months) on brook trout density or biomass in some streams, but no discernible
long-term effects (Figures 6-42 to 6-44). Over time, brook trout populations returned to levels
observed before the transplants and were generally consistent with density and biomass levels
i
expected given the chemical conditions in the streams.
Transplant experiments were also conducted with forage fish: blacknose dace in the Adirondacks
and slimy or mottled sculpin in Catskill and Pennsylvania streams. Few to none of these fish
were recovered in subsequent surveys in streams with median pH < 5.2 and Alim > 100-150 ^wg/L
during high flow (Buck Creek, Bald Mountain Brook, Seventh Lake Inlet, East Branch of the
Neversink River, Linn Run, Stone Run, and Roberts Run), whereas larger numbers of stocked fish
were caught in Fly Pond Outlet, Biscuit Brook, Baldwin Creek, and Benner Run. Sculpin are
difficult to capture with electrofishing; thus, population abundance and stocking recovery rates
could not be quantified. Average recovery rates for blacknose dace (percent of stocked fish
caught in the next survey) were 32% in Fly Pond Outlet, 7% in Bald Mountain Brook, 2% in
196
-------
Seventh Lake Inlet, and 0 in Buck Creek (see Figure 6-45). Large numbers of blacknose dace
remained only in Fly Pond Outlet (median high flow pH 6.2, Aljm 20
Chemical conditions in
Bald Mountain Brook, Seventh Lake Inlet, and Buck Creek resulted in high and rapid rates of
dace loss, through fish mortality arid/or emigration. Of the dace stocked into Bald Mountain
Brook, 25-35% were captured in downstream fish traps (Section 6.2). These results support the
conclusion that low pH and elevated Alim during high flow can eliminate acid-sensitive fish
populations.
Many investigators have concluded that fish early life stages and reproduction are critical periods
for fish populations in acidified lakes and streams (J. Baker et al., 1990a). Detailed studies of
effects on reproduction were not conducted as part of the ERP. However, the occurrence of
young-of-the-year fish in field surveys is indicative of reproductive success (during the period of
study). No young-of-the-year brook trout were caught in any survey in Buck Creek, Seventh Lake
Inlet, East Branch of the Neversink River, or Linn Run — all streams with low pH and/or high Aljm
during high flow. Small numbers of young-of-the-year trout were caught in fall 1989 in Bald
Mountain Brook, but none in fall 1988 or spring 1989, 1990, and 1991. Young-of-the-year trout
were abundant in Fly Pond Outlet, Black Brook, High Falls Brook, Biscuit Brook, Benner Run, and
Baldwin Run. Numbers of young-of-the-year caught in Stone Run and Roberts Run varied among
sampling years (Figure 6-46). Gagen (1991) concluded that year-to-year variations in reproduc-
tive success in Stone Run and Roberts Run resulted from annual variations in precipitation and
associated variations in episode severity during spring, when brook trout fry emerge from the
gravel. A strong year class occurred in 1988, coincident with below normal precipitation during
winter and spring 1988, whereas higher than normal precipitation in winter/spring 1989 resulted in
more severe episodes, decreased reproductive success, and a small year class.
Associations between stream chemistry and the occurrence of young-of-the-year fish, as well as
fish community composition, were confirmed in longitudinal profiles of fish distribution and stream
chemistry in Catskill and Pennsylvania ERP streams (Figures 6-47 to 6-53) .3 Ypung-of:the-year
trout and acid-sensitive fish species were caught only in stream sections and tributaries with pH
> 5.5-6.0 at the time of the survey. For example, the East Branch of the Neversink River supports
a sparse population of brook trout but no young-of-the-year trout or sculpin. Deer Shanty Brook
(pH 4.9), a tributary to the East Branch of the Neversink River, has an unnamed tributary with a
Adirondack ERP streams were relatively short, most with lakes or barriers to fish movement not very far upstream or
downstream from the ERP study reach (see Section 3.1.1). For this reason, we have not presented longitudinal profiles for
Adirondack streams.
197
-------
high pH (6.3) that supports a moderate population of adult and juvenile brook trout as well as
slimy sculpin (Figure 6-47). This isolated population of sculpin is of particular interest because
the nearest known source of sculpin is Flat Brook, a tributary to the East Branch of the Neversink
nearly 3 km downstream from the gauging station (Murdoch et al., 1991). Gagen (1991) con-
ducted a detailed study of several tributaries to Linn Run. Grove Run, E-Run, and Cabin Run had
total dissolved Altd concentrations < 20 fig/L and pH > 6.2 during a 4-day episode in April 1987,
Whereas Alld levels peaked at 425 and 525 jug/L at two sites in Linn Run (Figure 6-48). All three
tributaries support sculpin populations and had reasonable numbers of young-of-the-year brook
trout.
The occurrence of tributaries or microhabitats with suitable acid-base chemistry may mitigate, to
some degree, the effects of episodic acidification on fish populations. Fish from such areas may
immigrate into and recolonize the mainstream during low flow. Telemetry studies confirmed that
some fish are able to avoid exposure to adverse chemical conditions by moving into alkaline
refuge areas (Section 6.2). However, streams with low pH and high Aljm during episodes consis-
tently supported fewer fish and fewer fish species than did nonacidic streams (Figures 6-37 to
6-41; Table 6-21). In addition, there is no evident association between fish population status and
the degree of spatial variability in stream chemistry within the drainage system. For example,
despite the occurrence of tributary streams with high pH, Linn Run still supports very few brook
trout. Therefore, we conclude that recolonization from alkaline tributaries or microhabitats can
maintain low densities of fish in streams that experience toxic episodes, but is not sufficient to
sustain fish densities and biomass at levels expected' in the absence of adverse acid-base
chemistry.
Sculpin are absent from streams that support low numbers of brook trout (Table 6-21). Yet, ERP
In situ bioassays indicate that sculpin are as tolerant, or more so, of low pH and high Aljm than
brook trout (Section 6.1). Gagen (1991) proposed that differences in the population-level effects
of episodic acidification on sculpin and brook trout result from differences in fish mobility.
Phinney (1975) observed that brook trout can rapidly move more than 1 km to recolonize stream
sections. Brown and Downhower (1982) reported that summer movements of mottled sculpin
were generally < 10 m. Thus, sculpin would be expected to recolonize stream reaches at a
much slower rate than brook trout and, as a result, the effects of toxic episodes on sculpin
populations would be more severe and long lasting. Sculpin populations are acid sensitive (see
Table 6-21; Baker and Christensen, 1991), even though individual sculpin were relatively acid
tolerant in the in situ bioassays.
198
-------
The ERP results demonstrate that streams with suitable conditions during low flow, but adverse
chemical conditions during high flow, support fewer fish species and substantially lower brook
trout numbers and biomass than do nonacidic streams. Streams that experienced acidic epi-
sodes had significantly lower numbers and biomass of brook trout than nonacidic streams; differ-
ences between chronically and episodically acidic streams were not statistically significant. In
general, reduced trout abundance occurred in ERP streams with median high flow pH < 5.0 and
Alim > 100-200 fj.g/L. Acid-sensitive fish species were absent from streams with median high flow
pH < 5.2 and Alim > 100 fig/L.
199
-------
8
o
2!
«Q
C
3
o>
o. o
o> sr
o o
•2. 3
> CD
I a
o> 3
s §•
o 5"
§ CD
I 5
s
I!
3 IT
to
o 5'
U (O
o
-------
.~. 1332"
W
0-~- 1184
.E, 1036'
0) 888'
P
}« 740'
296"
148-
0"
6
X
Q.
4-
432'
384'
336
^ 288'
A 24°
E 192'
^ 144"
96"
48"
0"
100"
80"
•^ 60-
I 401
20
(a)
(b)
(c)
(d)
30MAR90 04APR90 09APR90 14APR90 ' 19APR90 24APR90 29APR90 04MAY90 09MAY90 14MAY90
Date
Figure 6-2. In situ bioassay in East Branch of the Neversink River, Catskills, in spring
1990: (a) discharge, (b) pH, (c) Al,m, and (d) cumulative percent mortality of
brook trout over time.
201
-------
279
248
.£. 217
CO 188
•g 124-
62'
31'
0
T
O.
(a)
(b)
(c)
(d)
24FEB89 01MAR89 06MAR89 11MAR89 16MAR89 21MAR89 26MAR89 31MAR89 05APR89 10APR89
Date
Figure 6-3. In situ bioassay in Linn Run, Pennsylvania, in spring 1989: (a) discharge,
(b) pH, (c) Alim, and (d) cumulative percent mortality of brook trout over time.
202
-------
Brook Trout
(a)
s?
•*->
jg
o
o
I
o
IN
1 UW
80 -
60 -
40 -
20 -
0 -
0
T
1
0
6
Reference Other-
Streams Streams
Sculpin
(b)
15
s
o
"U
1
O
CM
i wtj
80 -
60 -
40 -
20 -
0 -
O
T
o
r^\
©,
w
Reference Other
Streams Streams
Blacknose Dace
(c)
1 UW -
^^v
«
^ 80 -
o 60 -
o
s 40 -
o
y 20 -
N 'n-
"K
/~\
\~t
Reference Other
Streams Streams
100
o
All Fish
(a)
80 -
60 -
40 -
o
T 20
o
(N
0
O
T
Reference Other
Streams Streams
Figure 6-4. Box plots of percent mortality after 20 days in in situ bioassays in reference
and nonreference streams. Line in box indicates median % mortality; upper
and lower borders of box show 25th and 75th quartiles; whiskers indicate 10th
and 90th percentiles; circles represent observations beyond 10th and 90th
percentiles.
203
-------
I*
o
-*-•
l_
o
2
o
I
o
CM
Brook Trout
Common Pool
(a)
100
90
80
70
60
50
40
30
20
10
0
_L
Reference Acidic Chronic
Episode Acidity
Sculpin
(b)
^100 -
g 90 -
^ 80 -
~ 70 -
-2 60 -
o 50 -
^ 40 -
o* 30 -
TJ
I 20 -
8 10 -
0 -
T
s~\
w
— &-
Reference Acidic Chronic
Episode Acidity
All Brook Trout
«
N«^
"o
-f-»
o
o
TJ
I
O
CM
(c)
100
90
80
70
60
50
40
30
20
10
0
O
O
T
Reference Acidic Chronic
Episode Acidity
100
90
80
70
60
50
o 30
I 20
8 10
Dace
(d)
Reference Acidic
Episode
Figure 6-5. Box plots of percent mortality after 20 days in m situ bioassays classified
according to ANC: reference (nonacidic with ANC always > 0); chronic
acidity (ANC always £ 0); and acidic episode (initial ANC > 0 with at least two
consecutive values < 0 during 20-day period). Line in box indicates median %
mortality; upper and lower borders of box show 25th and 75th quartiles;
whiskers indicate 10th and 90th percentiles; circles represent observations
beyond 10th and 90th percentiles.
204
-------
&
o
c
s
0)
Q.
100
.
80-
-
60-
40-
-
20-
-
0
p
p
c
c A
C
A
A A A A
A
A
C
PC A
CAP A
P A
PAPA A
A r
fT» " ^ p p
fPOC<~ ATA A f f"
0
100 200
Median Alim(^g/L)
(a)
300
^
1
"£
QJ
0)
D_
100-
80-
•
60-
40-
20-
n-
P p
c
c A
C
A
A A A A
A
A
C
PC A
C A A P
P A
PA • A p A
p p C A p p
r. s^. rr. r. p&£" r &
0 100 200 300 400
Maximum Alim(Mg/L)
500
(b)
600
Figure 6-6. Bivariate plot of percent survival of brook trout (common pool fish only) after
20 days in in situ bioassays as a function of (a) time-weighted median Aljm
(ug/L) and (b) maximum measured or estimated AIjm (ug/L) during the 20-day
period. Bioassays conducted in Adirondack streams denoted by A, Catskills
by C, and Pennsylvania by P.
205
-------
£•
1
0
E
t*"1
CD
0.
100-
80-
-
60-
.
40-
20-
•
0-
p
p
c
AC
C
A
AA A A
A
A
C
A c P
A pc A
A P
A p A P A
C pp ftp
r. r. A er. p r. rrtrnr. A
i • i • i ' i ' • i
45678
(a)
Median pH
100-
80-
&
1 60-
Q
| 40-
Q_
20-
o-
PP
c
CA
c
A
A A A A
A
A
C
A C P
A P C A
A P
A P A Ap
p&FP p
r. r. A rp n ITTA rnr A A
1 I ' 1 ' i ' i
4 5 67 8
Minimum pH
(b)
Figure 6-7. Bivariate plot of percent survival of brook trout (common pool fish only) after
20 days in in situ bioassays as a function of (a) time-weighted median pH and
(b) minimum measured pH during the 20-day period. Bioassays conducted in
Adirondack streams denoted by A, Catskills by C, and Pennsylvania by P.
206
-------
e
Q
^
^
8
CD
Q_
iuo-
80-
60-
40-
20-
0-
PP
c
c A
C
A
A A A
A
A
C
1 A
I A
<
P
(a)
10
20
Median ASIt
30
40
i
o
§
£
100-
80-
60
40-
20-
o-
P p
c
A c
C
A
A A A A
A
A,
C
E A
6 p A
S.
8 P A
effi^ A r r
^..rr., i « | i | l •( r | i | i j — -. --J1- -i f'
0 10 20 30 40 50 60 70 8C
(b)
Maximum ASlt
Figure 6-8. Bivariate plot of percent survival of brook trout (common pool fish only) after
20 days in in situ bioassays as a function of (a) time-weighted median acidic
stress index based on brook trout fry (ASIt, see Table 6-9) and (b) maximum
ASI, during the 20-day period. Bioassays conducted in Adirondack streams
denoted by A, Catskills by C, and Pennsylvania by P.
207
-------
&
1
£
"c
S
CO
CL
TOO
80
60
40-
20-
•
0-
P
c
g
c
A A A
A
A
C
PC *
C A P A
P A
P AA p A
P P P P
Acrrr p r a PA AT
(a)
10 20 30 40 50 60 70 80 90 100
Median ASL
100-
80-
9
•e 60-
5
t
^
§ 40-
S
DL
20-
o-
1
P
c
£
c
A
AA
A
A
P c A
C A B
P A
P A £
& P n p P c~. r^fe
1 ' T ' I ' I ' I ' I ' I ' I ' i " ' "T
0 20 30 40 50 60 70 80 90 10
(b)
Maximum AS
Figure 6-9. Bivariate plot of percent survival of brook trout (common pool fish only) after
20 days in in s/fu bioassays as a function of (a) time-weighted median acidic
stress index based on rainbow trout fry (ASIS, see Table 6-9) and (b) maximum
ASls during the 20-day period. Bioassays conducted in Adirondack streams
denoted by A, Catskills by C, and Pennsylvania by P.
208
-------
£
•c
o
E
c
CD
2
CL
100-
80-
60-
40-
p
p
c
c A
C
A
/», A A
A
A
C
6 A
20 -W3 A
PA
% A
! ks-.
o-
tptT A r
1^ ' 1 • 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1
01 23456789
(a)
Median ASIt_ad
i"
•2
O
C
,£
100-
80-
60-
40-
20-
.
o-
p
c
c
A
A A A A
A
A
C
P A
£ A P
SAP A
ffiC A p
at CA
1 ' 1 ' 1 ' 1 ' 1
0 10 20 30'40
p
c
A
P
' I
5(
(b)
Maximum ASIt_ad
Figure 6-10. Bivariate plot of percent survival of brook trout (common pool fish only) after
20 days in in situ bioassays as a function of (a) time-weighted median acidic
stress index based on brook trout adults (ASIt_ad, see Table 6-9) and (b) maxi-
mum ASI,.ad during the 20-day period. Bioassays conducted in Adirondack
streams denoted by A, Catskills by C, and Pennsylvania by P.
209
-------
100-
80-
1 60H
§ 40-
20-
o-i
0
"KK
A A
A A
(a)
100
200
300
Median
I
£
1
o>
a.
100-
80-
60-
-
40-
20-
o-
A A
A A
A A
A
A
D 100 200 300 400 50
(b)
Maximum AlimC"g/L)
Figure 6-11. Bivariate plot of percent survival of blacknose dace after 20 days in in situ
bioassays as a function of (a) time-weighted median AISm (ftg/L) and (b) maxi-
mum measured or estimated Aljm (pg/L) during the 20-day period. Bioassays
conducted in Adirondack streams denoted by A.
210
-------
100
A A
(a)
o
£
80-
60
40
20-
T~
4
1001
80-
o
60
40-
20
o-
~T~
5
-AA A_
A AA
Median pH
~K~K
AA
A
I
5
Minimum pH
M_
~r
8
8
Figure 6-12. Bivariate plot of percent survival of blacknose dace after 20 days in in situ
bioassays as a function of (a) time-weighted median pH and (b) minimum
measured pH during the 20-day period. Bioassays conducted in Adirondack
streams denoted by A.
211
-------
100
80
i
60-
40-
20-
0-
(a)
100 200
Median Al
\
300
100-
80-
JeoJ
§ 40H
20 -I
IT P i
C
r r
(b)
100
200 300
Maximum Alim(Mg/L)
400
500
Figure 6-13. Bivariate plot of percent survival of slimy and mottled sculpin after 20 days in
in situ bioassays as a function of (a) time-weighted median Alim Oag/L) and
(b) maximum measured or estimated Aljm (fig/L) during the 20-day period.
Bioassays conducted in Catskill streams denoted by C and Pennsylvania by P.
212
-------
100
80
*
I e°~
1
§ 40
t
20
.„:
p
c p
p
p
p
c P c p c
r r P fflTP P K r. r. r. r: p
i • i ' i ' I
4 5 6 7
(a)
Median pH
100
o
80-
60
40-
20-
• c
_EP_
C
crrpr:
p
n fc n
(b)
5 6
Minimum pH
Figure 6-14. Bivariate plot of percent survival of slimy and mottled sculpin after 20 days in
in situ bioassays as a function of (a) time-weighted median pH and (b) mini-
mum measured pH during the 20-day period. Bioassays conducted in Catskill
streams denoted by C and Pennsylvania by P.
213
-------
r
100 H
80-
f*'
o
c
-------
100
80-
t60
o
40-
03
Q_
20-
o
o
o
o
o
o
25
75
125
175
225
275
325
375
Aljm threshold
Figure 6-16. Boxplots of 20-day percent mortality of brook trout in in situ bioassays with
Aljm concentrations continuously above selected thresholds for periods less
than or equal to 8 days (dashed line) or longer than 8 days (solid line). Boxes
indicate 25th through 75 percentile distribution of observed percent mortality
for that group of bioassays; vertical lines (whiskers) extend 1.5 times the
interquartile range; and open circles are individual outlier data points;
horizontal lines join median values for each group.
215
-------
20-
18-
| 16
•f 14-
% 12-
_|
•5 10-1
*co*
Q
Q
8-
6-
4-
2-
0-
0
100
200
300
400
500
Al
im
Figure 6-17. Concentration duration curves (total period of time, days, with Alim levels
above each threshold concentration, /*g/L) for brook trout in situ bioassays
with low mortality (< 10%, dashed lines) and high mortality (> 40%, solid
lines) in Adirondack streams.
216
-------
I
40%, solid
lines) in Cats kill streams.
217
-------
CD
O
CD
O
"at
I
Q
20
18-
16
14
12-
10
8-
6
4-
2
0
0
100
200
300
400
500
Al,
im
Figure 6-19. Concentration duration curves (total period of time, days, with Afjm levels
above each threshold concentration, ftg/L) for brook trout in situ bioassays
with low mortality (< 10%, dashed lines) and high mortality (> 40%, solid
lines) in Pennsylvania streams.
218
-------
20 H
18-
CD 16
o
CO
•a 14
0) IH
i
CD 12
£
o
"nT
o
10
8-
6-
4-
I
-| 1 1
-i r
100
200
300
400
500
Al:
im
Figure 6-20. Concentration duration curves (total period of time, days, with Aljm levels
above each threshold concentration, ftg/L) for blacknose dace in situ bio-
assays with low mortality (< 10%, dashed lines) and high mortality (> 40%,
solid lines) in Adirondack streams.
219
-------
20 -
18-
16-
1
I
I
I
II
II
n
ti
n
it
n
a
it
n
ti
n
11
n
n
11
IB
in
ti
in
n \
n\
i
t
i
i
i
i
\\ \
\\ \
\\ v
\\ v
\\
\\\
1 "" V SX\
I _ _v. X.
0
100
200
Al
300
400
500
im
Figure 6-21. Concentration duration curves (total period of time, days, with Aljm levels
above each threshold concentration, /ig/L) for sculpin in situ bioassays with
low mortality (< 10%, dashed lines) and high mortality (> 40%, solid lines) in
Catskill streams.
220
-------
CD
O
CO
•a
CD
I
CD
of
o
20
18
16
14
12
10
8
6
4
2
0
0
100
200
300
400
500
AI
im
Figure 6-22. Concentration duration curves (total period of time, days, with Alim levels
above each threshold concentration, fig/L) for sculpin in situ bioassays with
low mortality (< 10%, dashed lines) and high mortality (> 40%, solid lines) in
Pennsylvania streams.
221
-------
FPO 1989SPRNG F18
500
400
300
200
MO
26MSH89
0=MortaBy 0=AdLmafelV
31MW89
05JUN89
1UUN89
16JUN89
FPO 1890SPRING Fl
0=Mortalty 0=A$ mortaly
500
400
300
200
100
30APR90
05MMSO
16MATCO
2IMSI90
0SQCIB9
HOCIB9
160CIB9
220CIB9
270CIB9
SU 1989 SPRING SIB
500
400
300
200
0=Mortafy . 0=Adj. mortally
31MAYB9
05JUN89
1UUN89
16JUN89
BIS 1969 FALL 1
0=Adf.mcxti/
500
400
300
200
100
0
A!'m("&4
1 1 ~ i I~ 1
WOCIB9 090CIB9 140CIB9 190CTB9 250CTB9
Date
BIS 1990SPRNG 9
^
400
300
200
K»
0=A4mortety
09MAR90 14MAR90 19MAR90 25MAffiO 3CMAf«0
Date
Figure 6-23. Inorganic Al concentrations (pg/L) during each common pool brook trout in
situ bioassay. Open circles, connected by linear interpolation, indicate both
measured and estimated Alim values. Bioassays are ordered from low to high
mortality (20-day percent mortality). Adjusted percent mortality (see text for
explanation), stream, season, year, and bioassay code are also indicated
(page 1 of 9).
222
-------
BIX 1989FAIL 1
0=Mortafiy 0=Adj. mortal/
HFL 1989FAU. 1
0=Moriaiy
0=
500
400
300
200
100
0
040
HFL
500
400
300
200
100
0
*.**)
^-~^_
o-e »- ««*» — <&*—&&' ^
i I ; — I i
CTB9 090CT89 140C1B9 190C1B9 250C
Date
400
300
200
100
0
TO 040
1969RLL 5 0=Mortaty 0=Acj. morfeiy HFL
fltafe*)
„
400
300
200
100
0
*.(*)
°^^^^ Jfe-_
1 1 I 1 1
CTM 090CIS9 140CT69 190CIB9 250CTB9
Date
1969SPR1NG 31 0=Horta% 0=Adj. mortally
/HtaMD
o — — o © --o ' o
07NOV89
13NCN89
24NOV89
29NOV89
29Mfly89
04JUN89
09JUN89
HFL 1990SPRING 11
500
400
300
200
100
0
06VR90
0=M
-------
T1S 1989FALL 1
500
400
300
200
100
0=Motfaty
0=AJj. mortally
(HOOTS
090C1B9
MOC1B9
190CIB9
250CT89
US
500
400
300
200
100
0
1989SPRJNG 31
0=Mocfely 0=A4 mortal
24MAY89
29MAY89
04JUN89
OSJUN89
Bffl 1990STOG 4
SCO
400
300
200
KB
0=MortaRy 0=Adj. mortaffly
02MAR90
07MAR90
12MAR90
18MAR90
23MAR90
500
400
300
200
100
130CT38
1A
2.7=Mort*y
190CIB8
240CT88
290CTB8
03W88
500
400
SCO
200
100
22MAB90
4A
2fl=
27HAR90
01APR90
06APR90
12APR90
BWN 1989 RLL 3
500
400
300
200
100
29=Mort#y
02NOV89
07NOB9 13NOV89
24NOV89
Figure 6-23. Inorganic Al concentrations (ug/L) during each common pool brook trout in
situ bioassay. Open circles, connected by linear interpolation, indicate both
measured and estimated Aljm values. Bioassays are ordered from low to high
mortality (20-day percent mortality). Adjusted percent mortality (see text for
explanation), stream, season, year, and bioassay code are also indicated
(page 3 of 9).
224
-------
IW 1999FALL 3
23=MortaIty 0=Adj. mortal/
500
400
300
200
100
0
02NCVB9
07NCM89
13NOV89
24NOV89
1990SPBNQ 10
4.8=Mortaty
500
400
300
200
WO
06APR90
11APR90
21APR90
27APR90
Date
BMB 1989 BUI MIA
5=Mortality 5=Adj. mortality
500
400
300
200
100
060C1B9
110CT89
160CIB9
220CIB9
270CIB9
RBS 1988RALL 1A
500
400
300
200
100
0
050CIB8
100CT68
160CIB8
210CI88
260CTB8
BCK 1989FA1I CIA
500
400
300
200
ttO
0
10=Morta%
060CTO9
110CTO9
160CIB9
220CIB9
270CT89
FPO 1988EML .R
500
400
300
200
WO
01NOV88
05NOV88
11NOVB8
Date
17NOW8
22NCV88
Figure 6-23. Inorganic Al concentrations G«g/L) during each common pool brook trout in
situ bioassay. Open circles, connected by linear interpolation, indicate both
measured and estimated Aljm values. Bioassays are ordered from low to high
mortality (20-day percent mortality). Adjusted percent mortality (see text for
explanation), stream, season, year, and bioassay code are also indicated
(page 4 of 9).
225
-------
FPO
500
400
300
200
100
0
OK)
SU
500
400
300
200
100
0
27$
HFL
500
400
300
200
100
0
Oft!
mfflL F1A W=MortaSiy W=Atf. mortal
"IhW
--^— — ^\^^—
1 1 1 1
CTB9 110C1B9 160CIB9 22QCTO9 270C
Dale
1969 fail SI t5=Mbrtaiy 0=Adj. mortal
«i»W
^^-- — -
3>89 020CTO9 070CIB9 120CI89 180C
Date
1989SPf8JG 23 19=Mortaty 0=/4 mortals
%M)
0 S> _»-« ^— 0
>R89 11APR89 17APR89 22APR89 27API
Date
fy BW
500
400
300
200
100
0
HB9 130
y BNP
500
400
300
200
KB
0
TB9 HOC
If FPO
500
400
300
200
WO
0
R89 27SE
1 1988 FAIL 1 11.1=Mortafy 8£=A4 mortaffly
IliWi
i i i i
CTB8 190CTB8 240C1B8 290G1B8 03NCVB8
Date
1938 BUI 1B 15=MQ(1a% RT=Ao}. nxxfaE
AtaWD
o»« — S o
y
JI88 160CTB8 210CIB8 ' 270CIB8 OtNO/88
Dale
1969 WL F1 20=ltefaiy 55=/>4 mortal^
AlinW
^— ^ ^ ^
?89 020CIB9 070CIB9 120CIB9 180CTO
Date
Figure 6-23. Inorganic Al concentrations («g/L) during each common pool brook trout in
situ bioassay. Open circles, connected by linear interpolation, indicate both
measured and estimated Al!m values. Bioassays are ordered from low to high
mortality (20-day percent mortality). Adjusted percent mortality (see text for
explanation), stream, season, year, and bioassay code are also indicated
(page 5 of 9).
226
-------
SU 1988BUI SI
500
400
300
200
WO-
01NCW88
20=Motta«yr 111=^. mortaty
1988FALL 1B
2U=MortaHy
530
400
300
200
WO
0
22NOV88
130CIB8
190CIB8
240CTB8
290CIB8
03NCM38
BIS 1969 SPRING 25
218=Morfeily
300
200
100
06APR89
11APR89
17APR89
22APR89
27APR89
SU «90SPRING SI
25=Mortal/ 25=Ac|. moriafiy
500
400
300
200
100
30APR90
2IMAWO
SIN 1988FAII 1A
500
400
300
200
100
25=Mortaiy 17.1=Acj. morta^r.
060CI88
120CTB8
170CTB8
220CIB8
270CTB8
US '!989fi«i 5
500
400
300
200
100
07NOV89
13NOV89
18NOV89
24NOVB9
29NOV89
Figure 6-23. Inorganic Al concentrations Cug/L) during each common pool brook trout in
situ bioassay. Open circles, connected by linear interpolation, indicate both
measured and estimated Aljm values. Bioassays are ordered from low to high
mortality (20-day percent mortality). Adjusted percent mortality (see text for
explanation), stream, season, year, and bioassay code are also indicated
(page 6 of 9).
227
-------
EAB
MI
35=Mortaity 235=
BCK 1989SPRWG C1B
45=Mortality
=*4 matlty
500
«0
300
200
100
0
27SEP89
020CIB9
070CTB9
120CIB9
500
400
300
200
KM
26MW89
31HAY89
05JUN89
1UJN89
16JUN33
BCK
500
300
a»
100
1990SFHNG
50=A4nxx«y
30APR90
KTOSO
16MAYSO
2IMAWO
500
,400
300
200
100
01NOV88
11NCM88
17NOWB8
22NCWB8
500
«0
300
200
100
0
1969SFHNG
50=MoriaJiy 50=Adj. mortaly
31MAV89
05JUN89
Date
1UUN89
16JH89
BMB 1990SPHNG Ml
•SOO-
400
300
200
100
50=Morta/Jy
30APR90
05MAV90
10MAY90
Date
16M/WSO
2IMAY80
Figure 6-23. Inorganic Al concentrations (ftg/L) during each common pool brook trout in
situ bioassay. Open circles, connected by linear interpolation, indicate both
measured and estimated Aljm values. Bioassays are ordered from low to high
mortality (20-day percent mortality). Adjusted percent mortality (see text for
explanation), stream, season, year, and bioassay code are also indicated
(page 7 of 9).
228
-------
BCK 1989RIL Cl
500
400
300
200
K»
27SEP89
020CIB9
55=Mortaft/ 47.1=Ac| mortal/
070C1B9
120CTB9
180CIB9
T1S 1990SPRN6 10
500
300
200
100
06APR90
11APR90
6a7=Mortaty 6a7=Acj.mortaty
1 1 :
16APR90 2IAPR90 27APR90
BCK 1868FWL CI
503
400
300
200
80=Mortaty
01NOVB3
06NOV88
11NCWB8
17NOV88
22NOV88
06APR89
11APR89
17APR89
=tq. morta%
85.7=MoitfiV 85.7=Adj. mortality
05MAY89
08MAY89
11MWB9
14MAY89
17MAV89
SIN
500
400
300
200
K»
1990SPHNG 4
02MAR90
07MAR90
91.4=Matafty 91.4=A4 mcrtaly
12MAR90
Date
18MAR90
23MAR90
Figure 6-23. Inorganic Al concentrations (ug/L) during each common pool brook trout in
situ bioassay. Open circles, connected by linear interpolation, indicate both
measured and estimated Aljm values. Bioassays are ordered from low to high
mortality (20-day percent mortality). Adjusted percent mortality (see text for
explanation), stream, season, year, and bioassay code are also indicated
(page 8 of 9).
229
-------
SIN 1990SPHNG 4A
943=Mortaty 9i.1=Adj. mxfeity
22MAR90
27MARM
OlAPfSO
12APR90
Figure 6-23. Inorganic Al concentrations (ug/L) during each common pool brook trout in
situ bioassay. Open circles, connected by linear interpolation, indicate both
measured and estimated Aljm values. Bioassays are ordered from low to high
mortality (20-day percent mortality). Adjusted percent mortality (see text for
explanation), stream, season, year, and bioassay code are also indicated
(page 9 of 9).
230
-------
100-
-
80-
t60'
t
0
E -
_i__r
(D
§40-
Q_
20-
-
0
0
o
° x°
X
X
X
X
X
X
X
X
X
X
X
0 /
00 Px 0
X
o
/
/
/
o
X
X
o
3 0 °
o - o ° ^X o
3 0 _ x -' "
0 OQ,Q^" o
'' 0 0
00 00
QKBDO 0 O 0 0 00
0
50 100 150
Median A
200
250
300
irn
Figure 6-24. 20-day percent mortality of brook trout as a function of the time-weighted
median Al,m concentration (ug/L): observed mortality (open circles) and
predicted mortality (dashed line) based on best single-variable logistic model
(see Table 6-11).
231
-------
100-1
80-
i* 60
e
o
CD
§ 40
Q_
20-
" _' ..... IT'-O
0-^
0
1 2
LINT200
Figure 6-25. 20-day percent mortality of blacknose dace as a function of the log of the
integral of concentration duration for Aljm > 200 //g/L: observed mortality
(open circles) and predicted mortality (dashed line) based on best overall
logistic model (see Table 6-15). Eight of 16 bioassays had observed mortality
= 0.
232
-------
100-1
80-
^60
CO
•c.
o
CD
£
£
40 -I
20-
0-
oo®
0
50 100 150 200 250 300
Median AIim(^g/L)
Figure 6-26. 20-day percent mortality of sculpin (mottled and slimy) in Pennsylvania
streams as a function of the time-weighted median Aljm concentration
observed mortality (open circles) and predicted mortality (dashed line) (see
Table 6-16).
233
-------
r
100-f
80-
60-
o
•E
o>
-------
"F" n
L_ U
a -100-
E
•§ -200-
o
? -300 -
Q)
z
c -400 -
.2
33 -snn -
' ' ' A 4 4 \a; ^
n I ' • A I I ' - •— n
^H^4-4«^
7 7 4 I ' I 4.
j- A-A-A-| \
6 6
l-O-t^CHO^-CHP^ ^" £>
A I
"\ I ^
3 '
3
§ -100-
E
.A-A-A ffi _200 .
0
J -300 -
3 CD
~?
c -400 -
.03
14 14 14 13 12 ° I
14 v\ T
VA
1
AA!
.'" A"A J"> .
12 WAvi
15 ~ p*.*~"
10 £
"N
\
A—.
/
\
\
\
4 '^_A_i,
f
4
10 15 20 25 30
10
15
20
25
(b)
30
ro
w
01
600 '
400 •
200 -
5 10 15 20 25
Days from October 16,1988
(C)
30
600 '
400 '
200 -
.
V;
*;
5 10 15 20 25
Days from March 3,1989
30
Figure 6-28. Total dissolved aluminum concentrations at continuous monitoring stations and median net movement of brook trout
in Linn Run (triangles) and Baldwin Creek (squares). Negative values for movement indicate downstream relative to
initial location. Sample size and ± 1 SE (vertical bars) are shown for fish movement on representative days (Source:
Gagen, 1991).
-------
*c*
Jf~i
».g
e
CD
O
2
0
c
cd
s
14| jt i12 ._
"
-200 -
-400 -
-600 -
-800 •
"5 .-innn -
i^f i i1^^, 1
iK-*1- t •"-::f-i^L_^p--'
14 13 13 12
i
c'"^^
10
9 (a) _
HE
I
CD
§
^>
o
Z
.23
73
CD
to 10 0 ? ? ? ?
0-
-200-
-400-
-600-
-800-
-1000-
a [J o if a-^L^a J a J u J u 0
\
i
\
\
i
A
\
\
1
' 10 9 I6 5 6 4 /4
\| I ,\ I I I I/2
vj / ^-^J & J A— a
1 ' *^ 1* 1 1 ^— i 1
10
15
20
25
30
10 15
(b)
20 25 30
ro
CO
o
600 n
400 -
200 -
5 10 15 20
Days from November 6,1989
(c)
600 i
400 •
ra
-£ 200 •
(d)
5 10 15 20 25 30
Days from March?, 1990
Fiaure 6-29 Total dissolved aluminum concentrations at continuous monitoring stations and median net movement of brook trout
in Stone Run (triangles) and Benner Run (squares). Negative values for movement indicate downstream relative to
initial location. Sample size and ± 1 SE (vertical bars) are shown for fish movement on representative days (Source:
Gagen, 1991).
-------
ovu
400 -
^>
^
•o
5* 200-
n -
1
*i |»
;» /K \ i\
I \ ff\ \ /"\
*• i i • •
» \ i » / *
* . i \ i ' i
/ 41 ; 4^^ V' f\
1 i ' 4
/ <-Aj ^444 ,
*-± v 4 a
1 4T
^
/ x
' 'i
/ ',
/ i
/ \
.;
/L
4 T
0
10
15
20
25
Days from March 3,1989
30
Figure 6-30. Median stream total dissolved Al concentration at locations of transmitter-
equipped fish (triangles; vertical bars = ± 1 SE) and Alw concentra-
tions of concurrent samples collected at the Linn Run continuous monitoring
station (broken line) (Source: Gagen, 1991).
237
-------
500
400
5" 300
**-**
Jr 200
<
100
o
°0
o
o
o
A A
A
A
O
J*LlP£P
v v£> o
Day No
•12
A4
0
il II 1 i
(a)
^r
O
0
O
r>
00
Day No.
(b)
-2000 -1600 -1200 -800 -400 0
Stream Distance (m)
Figure 6-31. Linn Run Altd concentrations and ANC at individual fish locations versus
position along stream length on days 0, 4, 11, and 12. Day 0 was March 3,
1989. Solid symbols, to the right of the vertical bar, indicate concurrent data
from the Linn Run continuous monitoring station. Stream distance is relative
to the continuous monitoring station and negative numbers correspond to
downstream (Source: Gagen, 1991).
238
-------
400-
o>
_r 200-
0
0
5 10
Time (d)
15
Fish #1
Continuous monitor
fish #15
Figure 6-32. Linn Run Altd concentrations during the spring 1989 telemetry study for
samples collected at the continuous monitoring station versus samples
collected at the locations of two radio-tagged fish that died. The last plotted
Al value indicates when the fish was found dead (Source: DeWalle et al.,
1991).
239
-------
400-
Jr 200 H
0
Time (d)
Fish #7
Continuous monitor
fish #23
Figure 6-33. Linn Run Altd concentrations during the spring 1989 telemetry study for
samples collected at the continuous monitoring station versus samples
collected at the locations of two radio-tagged fish that moved to microhabitats
or refuges with lower Al and survived (Source: DeWalle et al., 1991).
240
-------
'•j*
B>
•o
•»••
<
^—^
_i
o-
0>
O
^
Day No.
250 -i
200 -
150
100 -
Rfl
u\j
0 -
u
O A
o A/!^
D
0 ° V
q,.
1 1 r—i
v ° nn B
v a
^-7
V
v
V V V
*16
A 0
2
1 1 .1 1 11 1
Day No.
40 -i
20 -
0 -
-20 -
-4O -
V
V
w v
V V
0 0 0 _. ^^
00
^
0
» A
•16
8
T
(a)
(b)
-2400 -2000 -1600 -1200 -800 -400 0
Stream Distance (m)
Figure 6-34. Stone Run Altd concentrations and ANC at individual fish locations versus
position along stream length on days 0, 2, 8, and 16 of spring 1990 telemetry
study. Day 0 was March 7, 1990. Solid symbols, to the right of the vertical
bar, indicate concurrent data from the Stone Run continuous monitoring
station. Stream distance is relative to the continuous monitoring station;
negative numbers correspond to downstream (Source: Gagen, 1991).
241
-------
E
*••
o
o
2
JC
(D X H'9h Fa"S T°tal A'
+ Neversink Total Al
D Monomeric Al
** ^=— F1 ' * O Organic Al
11111111111
High Falls Brook
PH East Branch
Neversink River
' " "^
1 i i i i i i i I i i
Discharge High Falls Brook
\ /v East Branch
\ / \ Neversink River
- ^^—J V^—_ ___
1 1 1 1 1 ~T 1 1 1 1 1 —
(a)
(b)
(c)
(d)
6 8 10 12 14 16 18 20 22 24 26 28
April 1989
Figure 6-35. Net downstream movement of radio-tagged fish, AItd concentrations, pH, and
stream discharge during the spring 1989 radiotelemetry study in High Falls
Brook and East Branch Neversink River (Source: Murdoch et al., 1991).
242
-------
Intermittent /
Braid
Ud\ /pH5.0
-/ —•
GAGE
pH4.9
Point of
fish release
Spring
pH6.2
pH 4.7 -K
Direction
of flow
pH 4.7
Springs
pH4.8
Tray Mill Brook
o
h
500
V
1000 Feet
i
o
Fish dam
pH4.9
100 200 Meters
Figure 6-36. Longitudinal changes in stream pH along the East Branch Neversink River
during the spring 1989 radiotelemetry study (Source: Murdoch et al., 1991).
243
-------
r
400 -r
(a)
300-
nf
200-
100-
o-l
pA
P
~1 ' 1 '
5 6
Median weekly pH
i
7
8
41
5
§,
of
m
3-
0-i
PA
(b)
!| '] 'I !t B, it*. I. '.
i
7
8
Median weekly pH
Figure 6-37. (a) Mean density (number/0.1 ha) and (b) biomass (kg/0.1 ha) of brook trout
(average of fall 1988, 1989, and spring 1989, 1990 surveys) in ERP study
streams as a function of median weekly pH (Tables 5-1 to 5-13). Streams
labeled by region: A = Adirondacks, C = Catskills, and P = Pennsylvania.
244
-------
400
300-
TT200
Q>
100-
p
(a)
567
Median pH during high flows
8
4-f
CO
w
a
g
in
O-l
A
P
A P
(b)
567
Median pH during high flows
\
8
Figure 6-38. (a) Mean density (number/0.1 ha) and (b) biomass (kg/0.1 ha) of brook trout
(average of fall 1988, 1989, and spring 1989, 1990 surveys) in ERP study
streams as a function of median pH during the 95th percentile high flow
(Table 5-15). Streams labeled by region: A = Adirondacks, C = Catskills, and
P = Pennsylvania.
245
-------
4UO-
^300-
j=
B,
-200-
S3
E
3
^y»
100-
o-
c
A
£ P
C
P
A
AP - A
C
P
1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 '' 1 ' il ' ,;l
0 20 40 60 80 100 120 140 160 181
(a)
Median weekly Alim{Aig/L)
4-
ffi 1-
(H
0 20
40
Kb)
I
60
80 100 120
Median weekly Alim
140
i ' I
160 180
Figure 6-39. (a) Mean density (number/0.1 ha) and (b) biomass (kg/0.1 ha) of brook trout
(average of fall 1988, 1989, and spring 1989, 1990 surveys) in ERP study
streams as a function of median weekly inorganic Al concentration Cwg/L)
(Tables 5-1 to 5-13). Streams labeled by region: A = Adirondacks, C =
Catskills, and P = Pennsylvania.
246
-------
4UO-
300-
"ST
x:
I200"
E
3
Z
100-
0-
c
A
A
D
C p
C
p
A.
P A
C
P
V/ | ' • I ' I ' I ' I ' I
0 100 200 300 400 50
(a)
Median Alim (ug/g, during high flows
IF3
T-
c>
c
p
A
C P
(b)
100
200
Median A!
jm
300 400
, during high flows
500
Figure 6-40. (a) Mean density (number/0.1 ha) and (b) biomass (kg/0.1 ha) of brook trout
(average of fall 1988, 1989, and spring 1989, 1990 surveys) in ERP study
streams as a function of median Alim (wg/L) during the 95th percentile high
flow (Table 5-15). Streams labeled by region: A = Adirondacks, C =
Catskills, and P = Pennsylvania.
247
-------
400-1
300-
co
"J200-
|
z
100-
A
P
1 ; 1 1 1 T
23456
Severity of Stream Chemistry (relative ranking)
(a)
4i
i
it-
o-i
p
p
A
A
C
P
2345
Severity of Stream Chemistry (relative ranking)
(b)
Figure 6-41. (a) Mean density (number per 0.1 ha) and (b) biomass (kg/0.1 ha) of brook
trout (average of fall 1988, 1989, and spring 1989, 1990 surveys) in ERP study
streams ranked according to overall severity, of chemical conditions in the
stream (from most severe, 1, to least severe, 6). Rankings are explained in
the text; also see Section 5.2.1. Streams labeled by region: A = Adirondacks,
C = Catskills, and P = Pennsylvania.
248
-------
B-B-B BCK *• A-A BMB *^* * FPO
SLI
CO
6001
500
400-
Q,
-£300
JD
Z200
100-
0
\
\
\
\
\
\
—*
A
-Q..
(a)
JUN88 SEP88 DEC88 MAR89 JUN89 SEP89 DEC89 MAR90 JUN90 SEP90
ODD BCK *• A-A BMB *-* * FPO
SLI
3-
co
eo"
s
g
in
H
(b)
JUN88 SEP88 DEC88 MAR89 JUN89 SEP89 DEC89 MAR90 JUN90 SEP90
Figure 6-42. Variations in brook trout (a) density (numbers per 0.1 ha) and (b) biomass
(kg/0.1 ha) over time in Adirondack ERP streams. Cubic spline smoothing
curves shown to help identify trends in brook trout population abundance.
Arrows indicate approximate dates of fish transplants into ERP streams.
249
-------
600-
500-
•5-400-
o
"300H
2200-
100-
o-i
B-B-H BIS
BI_K *-"* * HFL 9-O-e TIS
(a)
JUN88 SEP8S DECS8 MAR89 JUN89 SEP89 DEC89 MAR90 JUN90 SEP90
4i
3-
ca
2-
03
s
I
CO
1-
D a D BIS
BLK *-* * HFL
TIS
(b)
o-i, . . . . .
JUN88 SEP88 DEC88 MAR89 JUN89 SEP89 DEC89 MAR90 JUN90 SEP90
Figure 6-43. Variations in brook trout (a) density (numbers per 0.1 ha) and (b) biomass
(kg/0.1 ha) over time in Catskill ERP streams. Cubic spline smoothing curves
shown to help identify trends in brook trout population abundance. Arrows
indicate approximate dates of fish transplants into ERP streams.
250
-------
BNR *- ^A BWN *-* * LNN
RBS
STN
6001
500-
400-
.Q
200-
100
(a)
JUN88 SEP88 DEC88 MAR89 JUN89 SEP89 DEC89 MAR90 JUN90 SEP90
BNR A- /^A BWN *-* * LNN
RBS
STN
41
3-
s
g
(H
1-
(b)
JUN88 SEP88 DEC88 MAR89 JUN89 SEP89 DEC89 MAR90 JUN90 SEP90
Figure 6-44. Variations in brook trout (a) density (numbers per 0.1 ha) and (b) biomass
(kg/0.1 ha) over time in Pennsylvania ERP streams. Cubic spline smoothing
curves shown to help identify trends in brook trout population abundance.
Arrows indicate approximate dates of fish transplants into ERP streams.
251
-------
CO
o
o>
CO
CO
CO
E
o
500
400
300
200
100
0
500
400
300
200
100
0
Fly Pond Outlet
MB Population
i i Stocked
Bald Mountain Brook
Buck Creek
(c)
Seventh Lake Inlet
(d)
Sep
88
Apr
89
Jun
89
Sep
89
Date
Nov
89
Apr
90
Jun
90
Jun
91
Figure 6-45. Biomass (g/m ) of blacknose dace caught and stocked into Adirondack ERP
streams. All streams sampled 8 times, September 1988 to June 1991.
(Source: D. Bath, AL.SC, pers. comm.).
252
-------
120
CO
o>
.fi
=J
Z
Roberts
Run
Benner
Run
Baldwin
Creek
Figure 6-46. Density of age-0 (young-of-the-year) brook trout in ERP streams in
Pennsylvania during May/June 1988 to 1990. No age-0 brook trout were
caught in Linn Run. (Source: Gagen, 1991).
253
-------
Black Brook
pH 6.6
pH6.7
GAGE
(a)
Biscuit Brook
No sculpin / /. ,. .
present' //£*?
1 Mile
- 0 t Kilometer
pH6.3
(C)
N
High Falls Brook
pH6.7
PH6.8
East Branch Neversink
(b)
1 Mile
1 Kilometer
(d)
• Young of year
o No young of year
n Brown trout
v Gage
x Few trout
v Sculpin
# 200-meter survey reach
Figure 6-47. Longitudinal profiles of fish distribution and pH along the length of the four
ERP Catskill study streams during August 1989 (Source: Murdoch et al.
1991).
254
-------
BRK
-I-MS
age-0 age-1+
# fish/0.1 ha
0.6
2.4
2.8
Linn Run
Downstream
Grove Run
0.3 1-4 0.2
Linn Run
Upstream
(a)
600
400
200
01234 Days
Linn Run
Downstream
Linn Run
Upstream
Grove Run
(b)
Figure 6-48. Median density of brook trout (BRK) age-0 and age-1+ and mottled sculpin
(MS) captured on five sampling dates in 1987 at five sampling sites in the Linn
Run drainage (a). Total dissolved aluminum concentrations during an April
1987 acidic episode (b). Arrow indicates direction of stream flow. (Source:
Gagen, 1991).
255
-------
r
to Loyalhanna Creek
pH6.7
Altd9
Linn Run
• Brook trout
T Mottled sculpin
o No fish
Continuous Monitor
pHS.O
Altd 304
N
pH4.7
Altd 535
1 km
Rgure 6-49. Linn Run drainage network showing fish community composition and stream
pH and total dissolved aluminum concentrations. All water samples were
collected on May 23, 1989, when stream discharge at the continuous monitor
was 0.033 m3/s/km2. Arrow indicates direction of stream flow. Dotted box
indicates area enlarged in Figure 6-48. (Source: DeWalle et al., 1991).
256
-------
pH4.3
Altd 484
N
Continuous Monitor
Stone Run
• Brook trout
o No fish
Lick Run
Figure 6-50. Stone Run drainage network showing fish community composition and stream
pH and total dissolved aluminum concentrations. All water samples were
collected on May 18, 1989 when stream discharge at the continuous monitor
was 0.063 m3/s/knn2. Arrow indicates direction of stream flow. (Source:
DeWalle et al., 1991).
257
-------
pH4.5
Altd119
Continuous Monitor
Trout Run
Roberts Run
• Brook trout
o No iish
N
Figure 6-51. Roberts Run drainage network showing fish community composition and
stream pH and total dissolved aluminum concentrations. All water samples
were collected on May 18, 1989, when stream discharge at the continuous
monitor was 0.072 m /s/km2. Arrow indicates direction of stream flow.
(Source: DeWalle et al., 1991).
258
-------
Conemaugh R,
1 km
N
Continuous Monitor
Baldwin Creek
Brook trout
White sucker
Blacknose dace
Creek chub
Mottled sculpin
Figure 6-52. Baldwin Creek drainage network showing fish community composition and
stream pH and total dissolved aluminum concentrations. All water samples
were collected on May 22, 1989, when stream discharge at the continuous
monitor was 0.032 m /s/km2. Arrow indicates direction of stream flow.
(Source: DeWalle et al., 1991).
259
-------
Black Moshannon Creek
pH5.7
Benner Run
Brook trout
Slimy sculpin
Blacknose dace
Brown trout
N
Continuous
Monitor
1 km
Figure 6-53. Benner Run drainage network showing fish community composition and
stream pH and total dissolved aluminum concentrations. All water samples
were collected on May 19,1989, when stream discharge at the continuous
monitor was 0.107 m /s/km2. Arrow indicates direciton of stream flow.
(Source: DeWalle et al., 1991).
260
-------
Table 6-1. Fish Mortality and Chemistry (Median, Range) During In Situ Bioassays with Brook Trout in Adirondack Streams3
FALL 1988
FPO
BMB
BCK
SLI
SPRING 1989
FPO
BMB
BCK
SLI
FALL 1989
FPO
BMB
BCK
SLI
FPO
BMB
BCK
SLI
Source
of Fish0
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
Starting
Date
11/01/88
11/01/88
11/01/88
11/01/88
5/26/89
5/26/89
5/26/89
5/26/89 '
9/27/89
9/27/89
9/27/89
9/27/89
10/06/89
10/06/89
10/06/89
10/06/89
Total
Days
30
30
30
30
25
25
25
25
40
40
40
40
31
31
31
31
20-Day Period
pH
6.52 (6.20-6.84)
4.97 (4.64-5.52)
4.77 (4.64-5.10)
4.95 (4.65-5.42)
7.06 (7.06-7.13)
5.92 (5.83-6.33)
5.28 (5.27-5.48)
5.59 (5.55-5.85)
7.07 (7.04-7.22)
6.33 (6.15-6.51)
5.38 (5.36-5.49)
5.61 (5.67-5.90)
7.04 (6.58-7.22)
6.26 (5.05-6.51)
5.31 (4.58-5.49)
5.81 (4.84-5.90)
Alim
20 (20-20)
198 (106-415)
299 (192-497)
210 (97-341)
10 (0-28)
31 (0-72)
132 (0-165)
83 (9-115)
42 (20-71)
22 (17-33)
173 (143-249)
54 (31-88)
35 (6-142)
19 (17-421)
176 (153-359)
44 (31-250)
% Mort
10.0
50.0
80.0
20.0
0.0
50.0
45.0
0.0
20.0
35.0
55.0
15.0
10.0
5.0
10.0
0.0
Full Experiment
PH
6.51 (6.20-6.84)
4.97 (4.64-5.52)
4.79 (4.64-5.10)
4.94 (4.65-5.42)
7.08 (7.06-7.15)
6.19 (5.83-6.36)
5.38 (5.27-5.48)
5.68 (5.55-5.85)
7.00 (6.58-7.22)
6.01 (5.05-6.51)
5.20 (4.58-5.49)
5.62 (4.84-5.90)
7.00 (6.58-7.22)
6.01 (5.05-6.51)
5.20 (4.58-5.49)
5.62 (4.84-5.90)
Alim
20 (20-20)
198 (106-415)
329 (192-497)
210 (97-341)
19 (0-28)
39 (0-110)
100 (0-165)
51 (9-115)
39 (6-142)
41 (17-421)
200 (143-359)
83 (31-250)
37 (6-142)
46 (17-421)
199 (153-359)
83 (31-250)
% Mort
15.0
50.0
86.7
25.0
15.0
60.0
45.0
0.0
30.0
45.0
60.0
25.0
25.0
15.0
15.0
5.0
ro
o
Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for AlirJ
in any one chemistry sample.
FPO = Fly Pond Outlet; SLI = Seventh Lake Inlet; BMB = Bald Mountain Brook; BCK = Buck Creek.
CP = common pool (see Section 4.4.1 for explanation).
-------
Table 6-1. Fish Mortality and Chemistry (Median, Range) During In Situ Bioassays with Brook Trout in Adirondack Streams
(Continued)3
Stream6
SPRING 1990
FPO
BMB
BCK
SLI
Source
of Fish0
CP
CP
CP
CP
Starting
Date
4/30/90
4/30/90
4/30/90
4/30/90
Total
Days
30
30
30
30
20-Day Period
PH
6.58 (6.19-6.91)
5.65 (4.74-5.87)
4.93 (4.54-5.07)
5.17 (4.74-5.37)
Alfcn
25 (10-107)
52 (15-277)
269 (196-434)
157 (87-236)
%Mort
0.0
50.0
50.0
25.0
Full Experiment
PH
6.62 (6.19-6.91)
5.20 (4.74-5.87)
4.82 (4.54-5.07)
5.17 (4.74-5.41)
A'im
20 (10-107)
136 (15-316)
296 (196-446)
164 (87-280)
%Mort
5.0
65.0
70.0
25.0
a Medians are time-weighted values, based on interpolated chemistry throughout the bloassay period. Maximum and minimum are the highest and lowest values measured (or estimated for Ai) In
any one chemistry sample.
FPO = Fly Pond Outlet; SU = Seventh Lake Inlet; BMB = Bald Mountain Brook; BCK = Buck Creek.
0 CP = common pool (see Section 4.4.1 for explanation).
10
-------
Table 6-2. Fish Mortality and Chemistry (Median, Range) During In Situ Bioassays with Brook Trout in Catskill Streams8
Stream6
FALL 1988
Black
Biscuit
EBrNS
Black
High Falls
Biscuit
EBrNS
SPRING 1989
Black
High Falls
Biscuit
EBrNS
High Falls
EBrNS
Source
of Fish0
Black
Black
Biscuit
Biscuit
EBrNS
EBrNS
Black
Black
Black
Black
PIN
PIN
PIN
PIN
H.Falls
EBrNS
Starting
Date
11/10/88
11/10/88
1 1/03/88
11/03/88
1 1/04/88
11/04/88
11/30/88
11/30/88
1 1/30/88
1 1/30/88
4/08/89
4/06/89
4/06/89
4/06/89
4/06/89
4/06/89
Total
Days
35
35
41
41
41
41
15
15
14
15
19
21
-21
21
21
21
20-Day Period
PHd
6.22 (6.08-6.75)
6.22 (6.08-6.75)
5.91 (5.59-6.21)
5.91 (5.59-6.21)
4.84 (4.59-5.05)
4.84 (4.59-5.05)
—
—
—
—
—
5.96 (5.86-6.77)
5,87 (5.51-S.06)
4.81 (4.56-4.96)
5.96 (5.86-6.77)
4.81 (4.56-4.96)
Al d
Alim
19 (0-43)
19 (0-43)
13 (13-71)
13 (13-71)
118 (81-207)
118 (81-207)
—
—
—
—
—
5 (0-17)
13 (6-25)
231 (173-312)
5 (0-17)
231 (173-312)
% Mortd
5.3
0.0
4.8
0.0
7.1
66.7
—
—
—
—
—
19.0
23.8
81.3
0.0
4.8
Full Experiment
PH
6.42 (6.08-6.75)
6.42 (6.08-6.75)
5.99 (5.59-6.21)
5.99 (5.59-6.21)
4.84 (4.59-5.08)
4.84 (4.59-5.08)
6.38 (6.33-6.53)
6.26 (6.24-6.38)
5.99 (5.98-6.01)
5.03 (5.07-5.08)
6.42 (6.52-6.83)
6.35 (5.86-6.77)
5.87 (5.51-6.28)
4.83 (4.56-4.96)
6.35 (5.86-6.77)
4.83 (4.56-4.96)
A'im
19(0-43)
19 (0-43)
13 (10-71)
13 (10-71)
137 (81-207)
137 (81-207)
23 (17-43)
0 (0-2)
11 (10-16)
134 "(136-1 46)
12(5-20)
6 (0-17)
13 (0-25)
229 (173-312)
6 (0-17)
229 (173-312)
% Mort
5.3
11.1
19.0
23.1
7.1
83.3
5.0
0.0
5.0
15.0
23.8
19.0
23.8
81.3
0.0
4.8
ro
0)
CO
Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for Allm)
in any one chemistry sample.
EBrNS = East Branch Neversink River.
PIN = Pigeon Creek; WBrNS = West Branch Neversink River.
— = bioassay did not last 20 days; therefore, no statistics are provided for the 20-day period.
-------
Table 6-2. Rsh Mortality and Chemistry (Median, Range) During In Situ Bloassays with Brook Trout In Catskill Streams
(Continued)*
Stream15
SPRING 1989
High Falls
Biscuit
EBrNS
High Falls
EBrNS
High Falls
EBrNS
FALL 1989
Black
High Falls
Biscuit
EBrNS
High Falls
EBrNS
High Falls
EBrNS
High Falls
EBrNS
Source
of Fish"
(cont.)
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
H.Falls
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
H.Falls
EBrNS
WBrNS
WBrNS
H.Falls
EBrNS
Starting
Date
5/04/89
5/04/89
5/05/89
5/19/89
5/19/89
5/19/89
5/19/89
10/04/89
10/04/89
10/04/89
10/04/89
10/05/89
10/07/89
1 1/08/89
1 1/08/89
1 1/08/89
1 1/07/89
Total
Days
13
• 13
20e
36
36
36
36
36
35
35
34
34
31
22
22
22
23
20-Day Period
PHd
—
—
4.60 (4.43-4.78)
6.38 (6.02-6.78)
4.74 (4.61-4.92)
6.38 (6.02-6.78)
4.74 (4.61-4.92)
6.55 (5.44-6.78)
6.64 (5.47-6.95)
6.20 (5.04-6.32)
4.88 (4.27-5.12)
6.65 (5.47-6.95)
4.86 (4.27-5.12)
6.48 (6.00-6.73)
4.74 (4.52-4.90)
6.48 (6.00-6.73)
4.74 (4.52-4.88)
Al d
Alim
—
—
295 (223-505)
1 (0-9)
148 (101-237)
1 (0-9)
148 (101-237)
17 (10-73)
12(0-34)
12 (7-55)
162 (138-237)
4(0-34)
162 (138-237)
11 (0-28)
211 (175-264)
11 (0-28)
210 (175-264)
%Mortd
—
—
85.7
0.0
0.0
0.0
0.0
0.0
0,0
0.0
0.0
9.5
4.8
0.0
28.6
0.0
0.0
Full Experiment
pH
6.19 (5.40-6.50)
5.47 (4.97-5.92)
4.60 (4.43-4.78)
6.55 (6.02-6.78)
4.67 (4.50-4.92)
6.55 (6.02-6.78)
4.67 (4.50-4.92)
6.55 (5.44-6.86)
6.58 (5.47-6.95)
6.04 (5.04-6.34)
4.71 (4.27-5.12)
6.55 (5.47-6.95)
4.71 (4.27-5.12)
6.48 (6.00-6.73)
4.74 (4.52-4.90)
6.48 (6.00-6.73)
4.74 (4.52-4.90)
Alim
7 (0-83)
51 (27-145)
295 (223-505)
0 (0-32)
177 (86-286)
0 (0-32)
177 (86-286)
20 (10-73)
3 (0-34)
12 (5-55)
184 (138-243)
3 (0-34)
184 (138-243)
11 (0-28)
211 (175-264)
11 (0-28)
211 (175-264)
% Mort
0.0
0.0
85.7
4.8
47.6
4.8
33.3
0.0
0.0
0.0
9.5
9.5
4.8
0.0
38.1
0.0
4.8
ro
o>
Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for Al^
in any one chemistry sample.
EBrNS = East Branch Neversink River.
PIN = Pigeon Creek; WBrNS = West Branch Neversink River.
— = bioassay did not last 20 days; therefore, no statistics are provided for the 20-day period.
Bioassay terminated after 11 days due to high mortality.
-------
Table 6-2. Fish Mortality and Chemistry (Median, Range) During In Situ Bioassays with Brook Trout in Catskill Streams
(Continued)3
Stream6
SPRING 1990
High Falls
Biscuit
High Falls
Biscuit
EBrNS
High Falls
EBrNS
Source
of Fishc
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
H.Falls
EBrNS
Starting
Date
3/09/90
3/09/90
4/06/90
4/06/90
4/06/90
4/06/90
4/06/90
Total
Days
28
28
30
30
30
30
30
20-Day Period
pH
6.51 (5.95-6.66)
5.85 (5.30-6.44)
6.43 (5.87-6.64)
5.68 (5.03-5.86)
4.65 (4.36-4.85)
6.43 (5.87-6.64)
4.65 (4.36-4.85)
Aljm
7(0-60)
32(21-117)
4 (0-32)
32 (13-159)
252 (182-423)
4 (0-32)
252 (182-423)
% Mort
0.0
0.0
0.0
4.8
66.7
0.0
52.4
Full Experiment
PH
6.55 (5.95-6.66)
5.91 (5.30-6.44)
6.54 (5.87-6.74)
5.76 (5.03-6.10)
4.68 (4.36-4.85)
6.54 (5.87-6.74)
4.68 (4.36-4.85)
Alim
8 (0-60)
32(6-117)
2 (0-32)
25 (0-159)
213 (50-423)
2 (0-32)
213 (50-423)
% Mort
4.8
. 4.8
0.0
4.8
66.7
0.0
57.1
Ol
Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for Al^
in any one chemistry sample.
EBrNS = East Branch Neversink River.
PIN = Pigeon Creek; WBrNS = West Branch Neversink River.
-------
Table 6-3. Fish Mortality and Chemistry (Median, Range) During In Situ Bioassays with Brook Trout in Pennsylvania Streams"
Stream
FALL 1988
Banner
Roberts
Stone
Baldwin
Linn
SPRING 1989
Benner
Roberts
Stone
Baldwin
Linn
Source
of Rsh
Benner
Benner
Roberts
Stone
Benner
Roberts
Stone
Baldwin
Baldwin
Benner
Benner/
Roberts
Roberts
Stone
Baldwin
Linn
Starting
Date
10/11/88
10/05/88
10/05/88
10/05/88"
10/06/88
10/06/88
10/06/88
10/13/88
10/13/88
10/13/88
2/21/89
2/23/89
2/22/89
2/27/89
2/28/89
Total
Days
36
36
36
36
35
35
35
36
36
36
42
39
40
39
37
20-Day Period
PH
6.14 (5,96-6.53)
5.69 (4.86-6.51)
5.69 (4.86-6.51)
5.69 (4,86-6.51)
6.12 (6.02-6,26)
6.12 (6.02-6.26)
6.12 (6.02-6.26)
6.78 (6,63-6.94)
5.92 .(5. 12-6.39)
5.92 (5.12-6.39)
5.82 (5.55-6.09)
5.19 (4.95-5.31)
5.08 (4.97-5.15)
5.99 (5.95-6.09)
5.63 (5.03-5.97)
A'im
0 (0-10)
28 (2-202)
28 (2-202)
28 (2-202)
0 (0-9)
0 (0-9)
0 (0-9)
9 (0-11)
55 (0-458)
55 (0-458)
15 (0-46)
115 (104-171)
212 (208-326)
14 (13-14)
200 (62-453)
%Mort
15.0
9.5
10.3
4.5
25.0
5.7
30.3
11.1
2.7
21.4
0.0
42.9
76.2
9.5
92.9
Full Experiment
PH
5.94 (5.54-6.53)
5.69 (4.85-6.51)
5.69 (4.86-6.51)
5.69 (4.86-6.51)
5.44 (5.19-6.26)
5.44 (5.19-6.26)
5.44 (5.19-6.26)
6.69 (6.25-6.94)
5,78 (4.98-6.39)
5.78 (4.98-6.39)
5.82 (4.91-6.09)
5.21 (4.59-5.31)
5.18 (4.63-5.47)
5.99 (5.37-6.16)
5.38 (4.95-5.97)
A'im
6(0-47)
28 (2-202)
28 (2-202)
28 (2-202)
5 (0-189)
5 (0-189)
5 (0-189)
9 (0-26)
79 (0-500)
' 79 (0-500)
17 (0-152)
117 (104-233)
173 (0-443)
14 (5-95)
208 (62-494)
%Mort
35.0
14.3
10.3
4.5
45.0
5.7
30.3
19.4
18.9
21.4
42.9
52.4
90.5
52.4
97.6
0)
Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for /M|J
in any one chemistry sample.
-------
Table 6-3. Fish Mortality and Chemistry (Median, Range) During In Situ Bioassays with Brook Trout in Pennsylvania Streams
(Continued)3
Stream
SPRING 1989
Roberts
Stone
Linn
FALL 1989
Benner
Roberts
Stone
Baldwin
Linn
SPRING 1990
Benner
Stone
Benner
Stone
Source
of Fishb
(cont.) -
Roberts
Stone
Linn
Benner
Stone
Stone
CP
CP
CP
CP
CP
CP
Starting
Date
3/14/89
3/14/89
3/20/89
10/30/89
10/31/89
10/31/89
11/02/89
11/02/89
3/02/90
3/02/90
3/22/90
3/22/90
Total
Days
20
20
17
21
20
20
20
20
20
20
20
20
20-Day Period
pH
5.10 (4.59-5.27)
5.22 (4.63-5.47)
5.30 (4.95-5.60)
5.96 (4.91-6.19)
5.55 (4.81-5.77)
5.90 (5.34-6.03)
6.30 (6.23-6.46)
5.84 (4.99-6.15)
6.06 (5.12-6.22)
5.18 (4.80-6.05)
5.92 (5.16-6.17)
5.13 (4.74-5.85)
AU
127 (111-233)
149 (0-443)
208 (151-494)
11 (2-131)
71 (48-233)
8 (0-156)
12 (10-17)
44 (19-496)
2 (0-101)
204 (0-324)
9 (0-90)
239 (12-496)
% Mort
28.6
19.0
85.7
14.3
'34.3
22.9
2.9
2.9
0.0
91.4
2.9
94.3
Full Experiment
pH
5.10 (4.59-5.27)
5.22 (4.63-5.47)
5.30 (4..95-5.60)
5.96 (4.91-6.19)
5.55 (4.81-5.77)
5.90 (5.34-6.03)
6.30 (6.23-6.46)
5.84 (4.99-6.15)
6.06 (5.12-6.22)
5.18 (4.80-6.05)
5.92 (5.16-6.17)
5.13 (4.74-5.85)
AL
127 (111-233)
149 (0-443)
208 (151-494)
11 (2-131)
71 (48-233)
8 (0-156)
12 (10-17)
44 .(19-496)
2 (0-101)
204 (0-324)
9 (0-90)
239 (12-496)
% Mort
28.6
19.0
85.7
-
14.3
34.3
22.9
2.9
2.9
0.0
91.4
2.9
94.3
Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for Alim)
in any one chemistry sample.
CP = common pool (see Section 4.4.1 for explanation).
-------
Table 6-4. Rsh Mortality and Chemistry (Median, Range) During In Situ Bioassays with Blacknose Dace in Adirondack Streams8
Stream5
FALL 1988
FPO
BMB
BCK
SLI
SPRING 1989
FPO
BMB
BCK
SLI
FPO
BMB
BCK
SLI
FALL 1989
FPO
BMB
BCK
SLI
Source
of Fishc
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
CP
Starting
Date
1 1/01/88
1 1/01/88
1 1/01/88
1 1/01/88
5/08/89
5/08/89
5/08/89
5/08/89
5/18/89
5/18/89
5/18/89
5/18/89
9/27/89
9/27/89
9/27/89
9/27/89
Total
Days
30
30
30
30
10
10
10
10
41
41
41
41
40
40
40
40
20-Day Period
PHd
6.52 (6.20-6.84)
4.97 (4.64-5.52)
4.77 (4.64-5,10)
4,95 (4.65-5.42)
—
—
—
—
7.03 (7.02-7.08)
5.84 (5.67-6.11)
5.20 (5.13-5.42)
5.54 (5.44-5.67)
7.07 (7.04-7.22)
6.33 (6.15-6.51)
5.38 (5.36-5.49)
5.61 (5.67-5.90)
Al d
Alim
20 (20-20)
198 (106-415)
299 (192-497)
210 (97-341)
—
—
—
—
14 (0-107)
35 (18-163)
149 (113-207)
97 (72-135)
42 (20-71)
22 (17-30)
173 (143-249)
54 (31-88)
%Moif
0.0
90.0
100.0
95.0
—
—
—
—
0.0
0.0
52.4
0.0
0.0
0.0
90.0
0.0
Full Experiment
PH
6.51 (6.20-6.84)
4.97 (4.64-5.52)
4.79 (4.64-5.10)
4.94 (4.65-5.42)
6.64 (6.61-6,78)
5.04 (5.02-5.15)
4,76 (4.75-4.83)
4.96 (4.93-5.08)
7.06 (6.47-7.15)
6.12 (5.67-6.36)
5.38 (5.13-6.01)
5.68 (5,44-6.11)
7.00 (6.58-7.22)
6.01 (5.05-6.51)
5.20 (4.58-5.49)
5.62 (4.84-5.90)
Alim
20 (20-20)
198 (106-^15)
329 (192-497)
210 (97-341)
20 (20-20)
172 (151-230)
334 (327-397)
171 (156-224)
19 (0-107)
39 (0-163)
92 (0-207)
51 (9-135)
39 (6-142)
41 (17-421)
200 (143-359)
83 (31-250)
%Mort
0.0
90.0
100.0
95.0
0.0
100.0
100.0
95.2
14.3
0.0
66.7
0,0
'
0.0
0.0
100.0
10.0
CO
a Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period, Maximum and minimum are the highest and lowest values measured (or estimated for Aljm)
in any one chemistry sample.
FPO = Fly Pond Outlet; BMB = Bald Mountain Brook; BCK = Buck Creek; SU = Seventh Lake Inlet.
° CP = common pool (see Section 4.4.1 for explanation).
— = bioassay did not last 20 days; therefore, no statistics are provided for the 20-day period.
-------
Table 6-4. Fish Mortality and Chemistry (Median, Range) During In Situ Bioassays with Blacknose Dace in Adirondack Streams
(Continued)3
Stream6
SPRING 1990
FPO
BMB
BCK
SLI
Source
of Fishc
CP
CP
CP
CP
Starting
Date
4/30/90
4/30/90
4/30/90
4/30/90
Total
Days
30
30
30
30
20-Day Period
pH
6.58 (6.19-6.91)
5.65 (4.74-5.87)
4.93 (4.54-5.07)
5.17 (4.74-5.37)
Alim
25 (10-107)
52 (15-277)
269 (196-434)
157 (87-236)
% Mort
0.0
80.0
100.0
95.0
Full Experiment
PH
6.62 (6.19-6.91)
5.20 (4.74-5.87)
4.82 (4.54-5.07)
5.17 (4.74-5.41)
Alim
20 (10-107)
136 (15-316)
296 (196-446)
164 (87-280)
% Mort
0.0
100.0
100.0
100.0
8
CO
Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for Alim)
in any one chemistry sample.
b FPO = Fly Pond Outlet; BMB = Bald Mountain Brook; BCK = Buck Creek; SU = Seventh Lake Inlet.
CP = common pool (see Section 4.4.1 for explanation).
-------
Table 6-5. Fish Mortality and Chemistry (Median, Range) During In Situ Bioassays with Slimy Sculpin in Catskill Streams8
Stream13
FALL 1988
Black
Biscuit
SPRING 1989
Black
High Falls
Biscuit
EBrNS
High Falls
Biscuit
EBrNS
FALL 1989
Black
High Falls
Biscuit
EBrNS
High Falls
EBrNS
Source
of Fish0
Black
Biscuit
PIN
PIN
PIN
PIN
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
WBrNS
KFalls
EBrNS
Starting
Date
11/10/88
11/03/88
4/08/89 '
4/06/89
4/06/89
4/06/89
5/04/89
5/04/89
5/04/89
10/04/89
10/04/89
10/04/89
10/04/89
10/05/89
10/07/89
Total
Days
35
41
19
21
21
21
13
13
9
36
35
35
34
34
31
20-Day Period
PHd
6.22 (6.08-6.75)
5.91 (5.59-6.21)
—
5.96 (5.86-6.77)
5.87 (5.51-6.06)
4.81 (4.56-4.96)
—
—
—
6.55 (5.44-6.78)
6.64 (5.47-6.95)
6.20 (5.04-6.32)
4.88 (4.27-5.12)
6.65 (5.47-6.95)
4.86 (4.27-5.12)
Al d
Alim
19 (0-43)
13 (13-71)
—
5 (0-17)
13 (6-25)
231 (173-312)
—
—
—
17 (10-73)
12(0-34)
12 (7-55)
162 (138-237)
4(0-34)
162 (138-237)
%Mortd
0.0
0.0
—
0.0
4.8
57.9
—
—
—
0.0
0.0
4.8
4.8
0.0
0.0
Full Experiment
PH
6.42 (6.08-6.75)
5.99 (5.59-6.21)
6.42 (6.52-6.83)
6.35 (5.86-6.77)
5.87 (5.51-6.28)
4.83 (4.56-4.96)
6.19 (5.40-6.50)
5.47 (4.97-5.92)
4.52 (4.45-4.69)
6.55 (5.44-6.86)
6.58 (5.47-6.95)
6.04 (5.04-6.34)
4.71 (4.27-5.12)
6.55 (5.47-6.95)
4.71 (4.27-5.12)
A'im
19(0-43)
13 (10-71)
12 (5-20)
6 (0-17)
13 (0-25)
229 (173-312)
7 (0-63)
51 (27-145)
301 (232-505)
20 (10-73)
3(0-34)
12 (5-55)
184 (138-243)
3(0-34)
184 (138-243)
%Mort
0.0
0.0
0.0
0.0
4.8
68.4
0.0
0.0
4.8
0.0
0.0
4.8
9.5
0.0
0.0
ro
a Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for Al(m)
in any one chemistry sample.
d
EBrNS = East Branch Neversink River.
PIN = Pigeon Creek; WBrNS = West Branch Neversink River.
— = bioassay did not last 20 days; therefore, no statistics are provided for the 20-day period.
-------
Table 6-5. Fish Mortality and Chemistry (Median, Range) During In Situ Bioassays with Slimy Sculpin in Catskill Streams
(Continued)3
Streamb
FALL 1989
High Falls
EBrNS
High Falls
EBrNS
Source
of Fishc
WBrNS
WBrNS
H.Falls
EBrNS
Starting
Date
11/08/89
1 1/08/89
1 1/08/89
1 1/07/89
Total
Days
22
22
22
23
20-Day Period
pH
6.48 (6.00-6.73)
4.74 (4.52-4.90)
6.48 (6.00-6.73)
4.74 (4.52-4.88)
AL
11 (0-28)
211 (175-264)
11 (0-28)
210 (175-264)
%Mort
0.0
0.0
0.0
0.0
Full Experiment
PH
6.48 (6.00-6.73)
4.74 (4.52-4.90)
6.48 (6.0.0-6.73)
4.74 (4.52-4.90)
Alim
11 (0-28)
211 (175-264)
11 (0-28)
211 (175-264)
% Mort
0.0
4.8
0.0
4.8
Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for Allm)
in any one chemistry sample.
EBrNS = East Branch Neversink River.
PIN = Pigeon Creek; WBrNS = West Branch Neversink River.
ro
-------
Table 6-6. Rsh Mortality and Chemistry (Median, Range) During In Situ Bioassays with Scuipin in Pennsylvania Streams3
Stream
FALL 1988
Benner
Roberts
Stone
Linn
Baldwin
Linn
SPRING 1989
Benner
Roberts
Stone
Baldwin
Linn
SPRING 1990
Benner
Stone
Benner
Stone
Species
Slimy
Slimy
Slimy
Slimy
Mottled
Mottled
Slimy
Slimy
Slimy
Mottled
Mottled
Slimy
Slimy
Mottled
Mottled
Source
of Fishb
Benner
Benner
Benner
Benner
Baldwin
Baldwin
SixMile
SixMile
SixMile
Linn
Baldwin
CP
CP
CP
CP
Starting
Date
10/11/88
10/05/88
10/06/88
10/13/88
10/13/88
10/13/88
2/21/89
2/22/89
2/22/89
2/27/89
2/28/89
3/02/90
3/02/90
3/02/90
3/02/90
Total
Days
36
36
35
36
36
36
42
40
39
39
37
20
20
20
20
20-Day Period
PH
6.14 (5.96-6.53)
5.69 (4.86-6.51)
6.12 (6.02-6.26)
5.92 (5.12-6.39)
6.78 (6.63-6.94)
5.92 (5.12-6.39)
5.82 (5.55-6.09)
5.03 (4.83-5.31)
5.08 (4.97-5.15)
5.99 (5.95-6.09)
5.63 (5.03-5.97)
6.06 (5.12-6.22)
5.18 (4.80-6.05)
6.06 (5.12-6.22)
5.18 (4.80-6.05)
Alim
-
0 (0-10)
28 (2-202)
9 (0-9)
55 (0-458)
9(0-11)
55 (0-458)
15 (0-46)
126 (104-198)
212 (208-326)
14 (13-14)
200 (62-453)
2 (0-101)
204 (0-324)
2 (0-101)
204 (0-324)
%Mort
1.8
4.3
7.4
0.0
0.0
0.0
0.0
33.3
47.6
0.0
57.1
0.0
48.6
0.0
45.7
Full Experiment
PH
5.94 (5.54-6.53)
5.69 (4.86-6.51)
5.44 (5.19-6.26)
5.78 (4.98-6.39)
6.69 (6.25-6.94)
5.78 (4.98-6.39)
5.82 (4.91-6.09)
5.16 (4.59-5.31)
5.18 (4.63-5.47)
5.99 (5.37-6.16)
5.38 (4.95-5.97)
6.06 (5.12-6.22)
5.18 (4.80-6.05)
6.06 (5.12-6.22)
5.18 (4.80-6.05)
A'im
6(0-47)
28 (2-202)
5 (0-189)
79 (0-500)
9 (0-26)
79 (0-500)
17 (0-152)
128 (104-233)
173 (0-443)
14 (5-95)
208 (62-494)
2 (0-101)
204 (0-324)
2 (0-101)
204 (0-324)
% Mort
30.9
38.3
7.4
0.0
0.0
30.4
0.0
81.0
100.0
33.3
100.0
0.0
48.6
0.0
45.7
ro
Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for Alim) in any
one chemistry sample.
CP = common pool (see Section 4.4.1 for explanation).
-------
Table 6-6. Fish Mortality and Chemistry (Median, Range) During In Situ Bioassays with Sculpin in Pennsylvania Streams
(Continued)3
Stream
SPRING 1990
Benner
Stone
Benner
Stone
Species
Slimy
Slimy
Mottled
Mottled
Source
of Fishb
CP
CP
CP
CP
Starting
Date
3/22/90
3/22/90
3/22/90
3/22/90
Total
Days
20
20
20
20
20-Day Period
PH
5.92 (5.16-6.17)
5.13 (4.74-5.85)
5.92 (5.16-6.17)
5.13 (4.74-5.85)
Alim
9 (0-90)
239 (12-496)
9 (0-90)
239 (12-496)
% Mort
11.4
91.4
0.0
42.9
Full Experiment
PH
5.92 (5.16-6.17)
5.12 (4.74-5.85)
5.92 (5.16-6.17)
5.13 (4.74-5.85)
Alim
9 (0-90)
239 (12-496)
9 (0-90)
239 (12-496)
% Mort
30.9
38.3
7.4
0.0
Medians are time-weighted values, based on interpolated chemistry throughout the bioassay period. Maximum and minimum are the highest and lowest values measured (or estimated for Alim) in any
one chemistry sample.
CP = common pool (see Section 4.4.1 for explanation).
ro
-------
Table 6-7. Fish Mortality (% Mortality after 20 Days) in In Situ Bioassays in Reference and Nonreference Streams
Species (Group)
All fish Mean
(Std. Dev.)
(n=)
Brook trout Mean
(all) (Std. Dev.)
(n=)
Brook trout Mean
(common pool) (Std. Dev.)
(n=)
Blacknose dace Mean
(common pool) (Std. Dev.)
(n=)
Sculpin3 Mean
(common pool) (Std. Dev.)
(n=)
% Mortality
Reference Stream
3.1
(5.6)
(n = 46)
4.8
(6.6)
(n = 27)
5.3
(7.2)
(n = 17)
0.0
(0.0)
(n = 4)
1.2
(3.4)
(n = 11)
Nonreference Streams
33.0
(33.6)
(n = 84)
30.6
(30.2)
(n = 53)
32.1
(30.7)
(n = 32)
58.5
(45.0)
(n = 12)
26.2
(28.3)
(n = 15)
Statistical Comparisons: Wilcoxon Rank
Sums (p-value)
0.0001
0.0001
0.0009
0.0441
0.0017
Mottled and slimy sculpin combined.
-------
Table 6-8. Fish Mortality (% Mortality after 20 Days) in In Situ Bioassays Classified by ANC as Nonacidic (Reference Bioassays),
Acidic Episodes, and Chronically Acidic3
Species (Group)
Brook trout Mean
(all) (Std. Dev.)
(n=)
Brook trout Mean
(common pool) (Std. Dev.)
(n=)
Blacknose dace Mean
(common pool) (Std. Dev.)
(n=)
Sculpinb Mean
(common pool) (Std. Dev.)
(n=)'
% Mortality
Nonacidic
8.0
(13.3)
(n = 33)
10.2
(15.3)
(n = 23)
5.8
(17.5)
(n = 9)
0.9
(1.9)
(n = 11)
Acidic
Episode
31.6
(30.9)
(n = 14)
29.7
(27.9)
(n = 8)
92.0
(7.6)
(n = 5)
20.6
(31.7)
(n = 3)
Chronically
Acjdic
31.4
(31.4)
(n = 12)
44.1
(36.9)
(n = 4)
(n = 0)
23.1
(26.5)
(n = 6)
Statistical Comparisons: Wilcoxon Rank Sums (p-value)
Nonacidic vs.
Acidic Episode
0.0017
0.0257
0.001 1
0.0951
Acidic Episode vs.
Chronically Acidic
0.8769
0.4954
—
0.9999
ro
•Nl
01
Nonacidic: ANC always > 0 ^eq/L
Chronically acidic: ANC always
Acidic episode: Initial ANC > 0, with at least two consecutive ANC values £ 0/
-------
Table 6-9. Equations for Calculating Acidic Stress Index (from J. Baker et al. 1990a)£
Model based on laboratory bioassay with brook trout swim-up fry:
ASIt =
100
\-3\
1 + exp[-23.49 + 5.35pH + (2.97 X lO^Ca - (1.93 x 10~Jy\//m]
Model based on laboratory bioassay with rainbow trout fry:
ASIS =
100
1 + exp[-8.90 + 1.56pH + (4.08 x 10"3)Ca - (7.04 x
,
Model based on laboratory bioassay with adult brook trout:
ASIt_ad = 100(1 - P/0.9627), where
P =
1
1 + exp[-3.25 + (1.30 x
Ca
Aljm units ftg/L of inorganic monomeric Al; pH as pH units.
276
-------
Table 6-10. Regression Coefficients (r2) from Single-Variable Linear and Logistic Regression of 20-Day Percent Mortality as a
Function of Time-Weighted Median Values of pH, Alim («g/L), and Three Acidic Stress Indices for Brook Trout
(Common Pool and All Fish), Blacknose Dace, and Slimy and Mottled Sculpin Using Both Observed and Adjusted
Mortality3
Response Variables
Trout, common, unadjusted
mortality (n = 49)
Trout, common, adjusted
mortality (n = 49)
All trout, unadjusted
mortality (n = 74)
All trout, adjusted mortality
(n = 74)
Dace, unadjusted mortality
(n = 16)
All sculpin, unadjusted
mortality (n = 32)
Slimy sculpin, unadjusted
mortality (n = 24)
Dependent Variables and Model Types
Linear
Al|m
0.56
0.60
0.47
0.48
0.75
0.55
0.45
Linear
L°910 Alim
0.41
0.46
0.36
0.36
0.73
0.38
0.28
Linear
pH
0.37
0.38
0.25
0.25
0.65
0.27
0.23
Linear
ASIS
0.31
0.33
0.29
0.27
0.50
0.41
0.31
Linear
ASIt
0.29
0.31
0.14
0.16
0.45
0.01
0.01
Linear
ASIt.ad
0.58
0.61
' 0.46
0.48
0.54
0.63
0.57
Logistic
Alim
0.59
0.64
0.49
0.51
0.79
0.61
0.52
Logistic
PH
0.41
0.43
0.25
0.25
0.77
0.14
0.13
ro
-j
ASI, see Table 6-9. Regression results presented as r2 values for linear regression and logistic-regression analog for r2 (Agresti, 1990). See text for .definition of
adjusted mortality. Measured Aljm is used where available; otherwise Alim is estimated as described in Section 4.3.3.
-------
Table 6-11. Logistic Regression Model Predicting 20-Day Percent Mortality of Brook Trout
(Common Pool) as a Function of the Time-Weighted Median Al,m Concen-
tration
Variable
Intercept
Median Alim (ag/L)
Parameter
Estimate
-2.707
0.0136
Standard
Error
0.1434
0.0089
Pr > Chi-Square
0.0001
0.0001
Model deviance = 887.0
Variance of residuals = 323.7
r2 = 0.592
Table 6-12. Logistic Regression Model Predicting 20-Day Percent Mortality of Brook Trout
(Common Pool): Best 3-Variable Model.
Variable
Intercept
Median Alim («g/L)
Minimum pH
Median Ca (aeq/L)
Parameter
Estimate
-10.86
0.0191
1.853
-0.0174
Standard
Error
1.34
0.0016
0.263
0.0028
Pr > Chi-Square
0.0001
0.0001
0.0001
0.0001
Model deviance = 826.1
Variance of residuals = 221 .9
r2 = 0.720
278
-------
Table 6-13. Logistic Regression Model Predicting 20-Day Percent Mortality of Brook Trout
(Common Pool): Best 5-Variable Model.
Variable
Intercept
Median Al,m (wg/L)
Minimum pH
Median Ca («eq/L)
Median DOC (mg/L)
Catskill Region
Parameter
Estimate
-8.334
0.0206
1.501
-0.0141
-0.2589
-1 .31 1 1
Standard
Error
1.32
0.0016
0.256
0.0029
0.0635
0.263
Pr > Chi-Square
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
Model deviance = 793.5
Variance of residuals = 1 66.3
r2 = 0.790
Table 6-14. Ordinary Least Squares Regression Model Predicting the Logit Transformation
of the 20-Day Percent Mortality of Brook Trout (Common Pool): Best Overall
Model.
Variable
Intercept
Median Alim (wg/L)
Minimum pH
Median Ca (^eq/L)
Parameter
Estimate
-6.177
0.0124
1.085
-0.0136
Standard
Error
1.91
0.0022
0.367
0.0039
Pr > Chi-Square
0.0023
0.0001
0.0050
0.0012
Variance of residuals = 256.6a
r2 = 0.6513
Model statistics calculated based on predicted probability of fish mortality rather than logit transformed mortality, for
comparison to logistic models.
279
-------
Table 6-15. Logistic Regression Model Predicting 20-Day Percent Mortality of Blacknose
Dace
Variable
Intercept
Log (integral concentration-duration for
AI > 200 fig/L)
Parameter
Estimate
-7.838
4.111
Standard
Error
2.399
1.060
Pr > Chi-Square
0.001 1
0.0001
Model deviance = 102.0
Variance of residuals = 52.9
r2 = .975
Table 6-16. Logistic Regression Model for Predicting 20-Day Percent Mortality of Sculpin
Variable
Intercept
Median AI fcg/L)
Region
Parameter
Estimate
-5.677
0.0189
1.837
Standard
Error
0.424
0.0016
0.319
Pr > Chi-Square
0.0001
0.0001
0.0001
Model deviance = 448.0
Variance of residuals = 118.1
r2 = .799
280
-------
Table 6-17. Chemical Conditions during Telemetry Studies with Net Downstream Movement of Brook Trout
stream
Bald Mt.
Brook
Buck
Creek
Buck
Creek
East Br.
Neversink
Linn Run
Linn Run
Stone
Run
Dates
10/04-
11/03/89
10/04-
11/03/89
5/07-
6/05/90
4/13-
4/28/89
10/16-
11/14/88
3/03-
4/02/89
3/07-
4/05/90
Net
Movement
(m)a
-50
-150
-200
-300
-200
-400
-950
Study Period
PH
Mediand
6.10
5.21
4.82
4.80
6.28
5.29
5.16
Range
5.05-6.51
4.58-5.49
4.54-6.16
4.72-4.97
4.98-r6.41
4.95-5.81
4.80-6.05
Alim c/jg/u
Mediand
28
187
308
186
69
252
218
Range
17-421
153-359
22-446
157-284
0-500
62-494
0-324
Ca (jueq/L)
Mediand
112
121
105
76
179
151
95
Range
102-124
110-129
96-162
74-SO
139-192
113-169
74-120
Discharge (cfs)
Mediand
1.0
0.9
9.1
38.7
5.0
19.9
6.4
Range
0.7-12.9
0.6-19.8
>*,'
2.2-49.0
18-102.8
2.8-31.2
6.9-64.4
4.8-13.2
Episode or Period Associated with Initiation of Movement13
Episode/
Period
Episode
#13
Initial
conditions
Initial
conditions
Initial
conditions
Episode
#2
Episode
#9
Initial
conditions
Min.
PH
5.05
5.41-
5.49
4.86
4.72
5.12
5.03
5.17
Max.Alim
(pg/L)
421
143-153'
317
284
458
357
211
Duration (d)
Alim>160
1.55
-
4.0
4.0
2.00
3.40
4.0
Alim>200
0.93
-
4.0
4.0
0.71
3.14
1.09
Al|m>300
0.00
-
1.52
0.00
0.45
1.48
0.00
Bioassay
%
Mortalit
V
5%
10%
50%
81%
3%
93%
91%
ro
CO
" From figures presented in D. Bath. Adirondack Lake survey Corporation, Ray Brook, New York, pers. comm., 1992; Murdoch et al. (1991), and Oagen (1991).
Episodes defined as described in Section 5.2.1. Initial conditions indicates that fish movement started within one to a few days after fish were released; chemistry summaries are for first
four days.
' No chemistry samples analyzed during first four days of study; range indicates values for samples collected immediately before and after four-day period.
Time-weighted median.
-------
Table 6-18. Chemical Conditions during Telemetry Studies with Little, If Any, Net Downstream Movement of Brook Trout
Stream
Fly Pond Outlet
Fly Pond Outlet
Fly Pond Outlet
Seventh Lake
Inlet
High Falls Brook
Baldwin Creek
Baldwin Creek
Benner Run
Benner Run
Stone Run0
Dates
5/22-6/20/89
10/05-11/03/89
5/05-6/05/90
5/23-6/20/89
4/13-4/27/89
10/16-11/14/88
3/03-4/02/89
11/06-11/28/89
3/07-4/06/90
11/06-11/30/89
PH
Median®
7.08
6.97
6.68
• 5.68
6.26
6.44
5.99
5.99
6.03
5.55
Range
7.02-7.15
6.58-7.22
6.19-7.00
5.55-5.85
5.86-6.77
6.28-6.94
5.37-6.16
4.91-6.19
5.12-6.22
5.24-6.03
Bioassay
% Mortality
0
10%
0
0
19%
11%
9%
14%
0
23%
Alim ("9A-)b
Median0
19
39
16
51
9
2
14
9
3
59.
Range
0-28
6-142
10-42
9-115
2-12
0-26
5-95
0-131
0-101
0-166
Ca (Keq/L)
Median*
235
253
197
124
201
125
160
66
83
116
Range
225-290
194-299
152-241
121-128
198-204
84-244
142-184
63-109
61-167
103-132
Discharge (cfs)
Median*
0.5
0.5
1.2
2.5
6.4
1.6
5.8
5.9
4.6
3.2
Range
0.4-1.0
0.2-1.7
0.8-4.7
2.3-5.1
5.7-11.5
1.2-6.8
5.8-68.7
4.3-31.6
4.0-10.0
1.4-6.4
to
00
to
Time-weighted median.
Estimated Aljm, as described in Section 4.3.3.
Episode #15 occurred during study but resulted in no net movement: Alim > 100//g/L for 2.6 days; maximum Alim = 156/
-------
Table 6-19. Number of Brook Trout and Blacknose Dace Caught in Fish Traps Moving
Upstream and Downstream in Bald Mountain Brook and Fly Pond Outlet in
Relation to Average Daily Stream Minimonitor pHa
Stream
Bald Mountain
Brook
Bald Mountain
Brook
Fly Pond Outlet
Minimonitor
PH
< 5.0
> 5.0
> 6.0
Brook Trout
Upstream
0
10
23
Downstream
0
31
54
Blacknose Dace
Upstream
0
0
2
Downstream
16
8
19
Source: D. Bath, ALSC, pers. comm.
Table 6-20. Number of Days Brook Trout Were Captured in Fish Traps in Bald Mountain
Brook and Fly Pond Outlet in Relation to Stream Minimonitor pHa
Stream
Bald Mountain
Brook
Bald Mountain
Brook
Fly Pond Outlet
Daily
Minimonitor
PH
< 5.0
> 5.0
> 6.0
No
Movement
43
217
233
Downstream
Movement
0
23
49
Upstream
Movement
0
7
23
Both
Directions
0
1
9
Fish traps were operational in both streams for 296 days; source: D. Bath, ALSC, pers. comm.
283
-------
Table 6-21. Species Caught In ERP Streams8
Stream
East Branch Neversink River
Roberts Run
Stone Run
Linn Run
Buck Creek
Bald Mountain Brook
Seventh Lake Inlet
Biscuit Brook
Benner Run
Fly Pond Outlet
Black Brook
High Falls Brook
Baldwin Creek
Stream Chemistry
Severity
Ranking15
1
2
2
3.
3
4
4
5
5
6
6
6
6
Median pH
Low Flow
4.91
6.05
6.02
6.40
6.20
6.59
6.19
6.31
6.21
7.22
6.82
6.87
6.44
High Flow
4.50
4.71
4.80
5.12
4.57
4.77
4.78
5.38
5.23
6.19
6.10
6.04
5.89
Median Al^ («g/L)c
Low Flow
58
20
10
0
47
43
59
3
0
20
13
0
11
High Flow
347
196
426
387
375
331
225
70
68
20
34
21
19
Fish Species Acid Sensitivity*1
Acid Tolerant
Brook trout
Brook trout
Brook trout
Brook trout
Brook trout
Mudminnow
Brook trout
Brook trout
Mudminnow
Brook trout
Brook trout
Brook trout
White sucker
Yellow perch
Brook trout
Brook trout
Brook trout
Intermediate
(Brown trout)*
Creek chub
(Brown trout)6
Creek chub
Creek chub
Darter sp.
Brown trout
Brown trout
Creek chub
Brown trout
Acid Sensitive
Blacknose dace
Slimy sculpin
Slimy sculpin
Slimy sculpin
Slimy sculpin
Mottled sculpin
ro
oo
Streams ranked by severity of stream chemistry; fish species classified according to their acid sensitivity. Species caught using etectrofishing—all sampling dates, except presence/absence of species
involved in transplant experiments (brook trout, blacknose dace, and sculpin) based only on pre-transplant survey in fall 1988.
Severity of stream chemistry ranked from 1 (most severe) to 6 (least severe); see text for explanation.
Using measured and estimated Alim combined, as described in Section 4.3.3.
Acid sensitivity classified based on Figure 4.2 in Baker and Christensen (1991). Species classified as acid tolerant if at least some populations are expected to survive in lakes with pH < 4.7 (i.e.,
lower range for critical pH < 4.7); intermediately sensitive, with lower pH range between 4.8 and 5.0 inclusive; and acid sensitive if species is expected to occur only in
waters with pH > 5.0. Mottled sculpin were not included in Figure 4.2 and were assumed to be in the same acid sensitivity class as slimy sculpin, based on their similar
mortality levels in ERP in situ bioassays (Section 6.1). Unknown darter species caught in Seventh Lake Inlet arbitrarily classified as intermediate sensitivity.
Only one brown trout was caught, in total, during all sampling dates. Thus, this species occurs in very low abundance.
-------
Table 6-22. Rank Correlation Coefficients (and Associated P-Values for Test of Zero
Correlation) Relating Brook Trout Density and Biomass (Mean Values from
Spring and Fall Samples) to Annual Median Stream Chemistry (Weekly
Samples) and Median Chemistry during Low and High Stream Discharge
Median Chemistry
Weekly pH
High flow pH
Low flow pH
Weekly Aljm
High flow Aljm
Low flow Alim
Trout Density
+0.852
(0.0002)
+0.725
(0.0050)
+0.728
(0.0027)
-0.610
(0.0269)
-0.703
(0.0073)
-0.055
(0.8585)
Trout Biomass
+0.890
(0.0001)
+0.830
(0.0005)
+0.780
(0.0017)'
-0.687
(0.0095)
-0.808
(0.0008)
-0.126
(0.6808)
285
-------
r
Table 6-23. Density and Biomass of Brook Trout Measured in ERP Streams8
Stream
Buck Creek
Bald Mountain
Brook
Fly Pond Outlet
Seventh Lake
Inlet
Biscuit Brook
Date
09/22/88
04/25/89
06/22/89
09/11/89
11/10/89
04/24/90
06/14/90
06/15/91
10/11/88
04/24/89
06/22/89
09/01/89
11/07/89
04/19/90
06/13/90
06/15/91
10/12/88
04/24/89
06/21/89
09/12/89
11/08/89
04/19/90
06/13/90
06/15/91
09/22/88
04/25/89
06/27/89
09/07/89
11/14/89
04/20/90
06/12/90
06/15/91
10/10/88
12/12/88
05/30/89
08/22/89
10/29/89
06/02/90
Density (number/0.1 ha)
Caught
4
31
69
57
83
39
26
6
162
117
194
171
135
135
113
90
516
308
216
236
167
115
239
89
49
69
67
71
69
27
43
24
59
60
87
126
141
86
Estimate
4
31
71
57
86
43
26
6
180
122
198
185
135
140
122
90
522
329
222
236
167
115
256
89
51
71
67
71
69
27
44
26
60
67
102
133
152
94
95% Upper
Confidence
Limit
14
35
77
59
96
55
28
8
216
140
212
212
140
158
144
104
530
352
233
242
173
121
282
153
58
78
70
74
71
28
49
26
62
77
116
141
161
103
Biomass
(g/0.1 ha)
Caughtb
265
344
1208
1255
1092
417
367
120
1441
1140
2068
1946
1423
1068
1505
1230
3452
2014
2121
1882
1986
1219
1896
1179
1125
1063
1629
1788
1418
418
838
646
—
—
1475
—
1858
1879
Data from Gagen (1991) for Pennsylvania streams; D. Bath (ALSC, pers. comm.) for Adirondack streams; B. Baldigo
(USGS, pers. comm.) for Catskill streams.
— = fish weights not measured.
286
-------
Table 6-23. Density and Biomass of Brook Trout Measured in ERP Streams (Continued)2
Stream
Black Brook
High Falls Brook
East Branch
Neversink River
Benner Run
Baldwin Creek
Linn Run
Date
11/10/88
06/20/89
08/28/89
10/25/89
11/27/89
06/04/90
11/15/88
05/01/89
08/22/89
10/13/89
06/02/90
06/07/89
08/23/89
10/16/89
06/06/90
07/14/88
11/15/88
03/30/89
06/29/89
08/22/89
11/28/89
04/06/90
06/05/90
07/12/88
11/22/88
04/06/89
06/28/89
08/25/89
12/01/89
04/03/90
05/31/90
07/12/88
11/22/88
04/06/89
06/28/89
08/25/89
12/01/89
04/03/90
05/31/90
Density (number/0.1 ha)
Caughtb
233
231
528
354
205
331
147
82
197
187
68
28
35
13
13
218
142
60
87
111
53
62
180
84
111
100
166
187
129
92
108
2
2
4
9
4
2
7
Estimate
268
241
554
395
233
331
155
85
235
193
75
45
36
14
14
227
146
62
91
116
55
62
189
85
111
114
172
193
132
98
119
11
2
2
4
9
4
2
7
95% Upper
Confidence
Limitb
297
256
"579
433
272
336
165
92
277
203
88
71
38
15
17
240
153
72
102
127
64
66
202
87
116
140
182
203
140
112
139
—
—
—
—
15
—
9
Biomass
(9/0.1 ha)
Caughf
3582
2749
—
3697
—
3849
2217
1397
—
2073
1620
394
—
195
283
1976
—
393
957
1582
744
587
2062
1976
—
393
957
1582
744
587
2062
—
—
50
252
278
196
—
311
Data from Gagen (1991) for Pennsylvania streams; D. Bath (ALSC, pers. comm.) for Adirondack streams; B. Baldigo
(USGS, pers. comm.) for Catskill streams.
— = data not obtained from regional cooperators.
Q
— = fish weights not measured.
287
-------
Table 6-23. Density and Biomass of Brook Trout Measured in ERP Streams (Continued)£
Stream
Roberts Run
Stone Run
Date
07/14/88
11/15/88
03/29/89
06/30/89
08/22/89
11/30/89
04/10/90
06/05/90
07/14/88
11/17/88
03/29/89
06/30/89
08/22/89
11/29/89
04/05/90
06/05/90
Density (number/0.1 ha)
Caught
84
51
64
38
53
36
29
42
159
71
59
85
95
66
39
49
Estimate
85
51
68
38
53
38
29
42
166
78
61
88
100
69
39
48
95% Upper
Confidence
Limit
87
55
80
42
57
47
33
45
178
92
71
96
109
76
44
53
Biomass
(g/0.1 ha)
Caught
1942
742
978
758
1036
578
569
751
1965
944
641
1350
1539
1220
1010
945
Data from Gagen (1991) for Pennsylvania streams; D. Bath (ALSO, pers. oomm.) for Adirondack streams; B. Baldigo
(USGS, pers. oomm.) for Catskill streams.
288
-------
SECTION 7
SUMMARY AND CONCLUSIONS
During the Episodic Response Project (ERP), we studied episodic acidification and the effects of
episodic acidification on fish in 13 streams in the Adirondack Mountains, the Catskill Mountains,
and the Northern Appalachian Plateau in Pennsylvania for approximately 20 months, starting in
fall 1988 and concluding in spring 1990. Intensive field efforts were successful in collecting
hydrologic, chemical, and biological data with which to accomplish project objectives. This
section summarizes findings from the ERP by objective; major conclusions of the project follow.
Objective 1. Determine the magnitude, duration, frequency, and characteristics of episodic
chemical changes that accompany hydrological events (snowmelt and rainstorms) in
streams.
Episodes were a common phenomenon in all three regions. We recorded from 19 to 33 episodes
in each of the study streams. Episode durations ranged from less than 1 day to more than 10
days. There was a direct relationship between ANC level at the beginning of episodes and the
magnitude of ANC depressions that occurred during episodes. When conditions were nearly
acidic at the beginning of episodes, ANC depressions tended to be smaller than during episodes
in which initial ANC values were higher. However, the small ANC depressions in low ANC sys-
tems often produced acidic episodes with the lowest minimum ANC values. Acidic episodes
occurred in at least some of the streams in all regions and were common where ANC values of
streams immediately before the episodes were £ 50 fieq/L. When acidic episodes occurred, they
were accompanied by depressed pH levels and elevated Aljm concentrations.
The occurrence of episodes was most common during the spring, winter, and fall months.
Summer streamflow levels were low and were generally not very responsive to rainstorms. In
most of the streams, three to five episodes developed much more severe chemical conditions
(lower minimum ANC and pH and higher Al) than did the remainder of the episodes during the
study.
In the Adirondacks, near normal snowpacks developed each winter of the study, and snowmelt,
or rain on snow, produced episodes with the lowest ANC and pH values and the highest concen-
trations of Aljm recorded. For Buck Creek, Bald Mountain Brook, and Seventh Lake Inlet,
minimum ANC concentrations during snowmelt were typically < -20 ,aeq/L; minimum pH values
were < 4.6; maximum Aljm concentrations in Buck Creek exceeded 600 fig/L.
289
-------
Snowpack accumulations were much smaller than normal in the Catskills. The majority of major
episodes in the Catskills occurred in response to rainstorms. Spring rainstorms generated acidic
episodes with minimum ANC values of -6 fieq/L in Biscuit Brook and -45 jueq/L in the East
Branch of the Neversink River; maximum Al|m concentrations were 457 /ug/L in the East Branch
and 166/^g/L in Biscuit Brook. Episodes with similar ANC levels in the fall mobilized much less
Al. The minimum episodic pH was 5.0 in Biscuit Brook and 4.3 in East Branch Neversink.
Virtually no snowpacks developed in Pennsylvania during the study period. Spring and late
winter rainstorms produced most of the major episodes in the Pennsylvania streams. Benner Run
and Baldwin Run were the two reference (highest ANC) streams in Pennsylvania. During the
most severe episodes, Linn Run, Roberts Run, and Stone Run had minimum ANC values ranging
from -17 to -52 fieq/L, minimum pH values ranging from 4.7 to 4.5, and maximum Alim concentra-
tions ranging from 270 to 707 fig/L. Similar amounts of Al mobilized in major spring and fall
episodes.
Because of the small snowpacks in Pennsylvania and the Catskills, episodic conditions in these
regions may not have been as severe during the ERP as during years with normal snowpack
accumulations. However, large rainstorms during the spring and late winter produced high
streamflows and major episodes.
The relationship between ANC and Aljm during episodes was fairly consistent among Adirondack
streams and among Catskill streams. However, the Pennsylvania streams had distinctive AI/ANC
relationships. For example, Linn Run had higher Aljm concentrations than did Roberts Run at a
given ANC level.
Evaluation of ion change and concentration data shows that episodic ANC depressions are a
result of complex interactions of multiple ions. These data also provide evidence that both natural
processes and acidic deposition make important contributions to episodic acidification in the
Adirondack, Catskill, and Pennsylvania study streams.
Sulfate in the streams is predominantly, if not solely, a consequence of atmospheric deposition,
because the ERP watersheds do not have internal sources of S. Nitrate in the ERP watersheds
probably is derived from a combination of natural N cycling and atmospheric deposition.
Watersheds in the Adirondacks and Catskills, in which stream water concentrations of NO3~ are
quite high, have in all likelihood been significantly affected by atmospheric deposition of N.
290
-------
For all episodes recorded (large and small) in the three ERP regions, base cation decreases were
commonly the most important ion change contributing to ANC depressions. In the Catskills and
Adirondacks, the largest base cation decreases occurred during nonacidic episodes. Base cation
decreases were especially dominant for episodes in circumneutral streams in the study (Black
Brook, Fly Pond Outlet, and High Falls Brook). However, some base cation decreases in the
Pennsylvania streams were very large even during acidic episodes.
Organic acid (A~) concentrations were greatest in the Adirondack streams, and A~ increases
during episodes were commonly the second and sometimes the first ranked ion changes during
major episodes. Organic acid contributions tended to be greatest in episodes with acidic or low
ANCmjn values. Organic acids were generally less important in Pennsylvania and Catskill
episodes. However, A~ increases were the most important ion changes during several fall
episodes in two Catskill streams. Organic acid pulses were commonly the second most impor-
tant, and occasionally the most important, ion changes during episodes in the Pennsylvania
streams, even though the magnitudes of A~ changes were small.
Sulfate was the dominant acid anion (C^ virtually at all times in the study streams in all three
regions. For both acidic and nonacidic episodes in the Catskills, SO42~ concentrations tended to
decrease, thereby reducing ANC depressions. In the Adirondacks, SO42~ concentrations were
observed to increase or decrease during episodes. However, SO42~ concentrations tended to
decrease during acidic episodes. Sulfate concentrations in the Pennsylvania streams were
greater than in Catskill or Adirondack streams. Based on the CBCA ANC definition, these large
SO42~ concentrations (in all regions) provided an acidity base upon which other ion changes
further modified the acid-base status of the stream waters. Another major difference between the
Pennsylvania streams and the Adirondack and Catskill streams is the importance of SO42~ pulses
in Pennsylvania episodes. In some major episodes, SO42~ pulses were the most important ion
change contributing to ANC depressions in three of the five Pennsylvania streams (Baldwin
Creek, Benner Run, and Stone Run) and were consistently the most important ion change in two
of the streams.
Nitrate behavior and contributions to episodes were different in each of the three regions. Nitrate
concentrations were much greater in the Catskill and Adirondack streams than in Pennsylvania
streams. In the Catskills, NO3~ concentrations typically increased during episodes, particularly
during acidic episodes. In two of the Catskill streams, Biscuit Brook and East Branch Neversink
River, NO3~ was the most important ion change contributing to ANC depressions during several
291
-------
episodes. In the Adirondacks, NO3 increases were typically the third ranked ion change during
episodes. The largest NO3~ increases occurred during acidic episodes. Even though episodic
changes of NO3~ were relatively small, NO3~ concentrations were large throughout Adirondack
episodes. Nitrate pulses in Pennsylvania streams were generally small and made little
contribution to ANC depressions.
Other ion changes were less important to episodes measured in the ERP. Chloride changes and
concentrations usually had relatively little impact on episodes. However, for a few episodes in
some Pennsylvania streams, Cl~ increases were the second most important ion change
contributing to ANC depressions. In virtually all episodes, especially those that were acidic, Al
increases tended to reduce ANC depressions.
From a biological perspective, we are most concerned about major episodes—those that gener-
ate low pH and ANC and high Alim for relatively long periods of time (days to weeks). Ion
behavior that controlled ANC depressions during major episodes recorded in each stream was
not necessarily the same as the ion changes most important to smaller, more frequent episodes.
In the Adirondacks, the only region to have major snowpacks, the most severe episodes generally
occurred during spring snowmelt. During these episodes, base cation decreases were usually
the most important contributions to ANC depressions, and A~ and NO3~ changes positively con-
tributed to episodes and were of similar rank (2 or 3). One Adirondack stream experienced a
major snowmelt episode in which an increase in NO3~ was the most important ion change. Three
of the four Catskill streams experienced one episode influenced by snowmelt in which NO3~
pulses were the first ranked contributor and base cation decreases were the second ranked ion
change.
Within regions, the major episodes generated by rainstorms had more variable ionic controls than
did snowmelt episodes. The large rain-induced Adirondack episodes occurred in the spring and
fall. In these episodes, increases in A~ and decreases in base cations were the most important
Jon changes to ANC depressions. In major spring episodes in the Catskills, decreases in base
cations or increases in NO3" were typically the first or second ranked ion changes contributing to
ANC depressions. For the major episodes in Pennsylvania, base cation decreases and SO42~
increases were usually the most important or second most important ion changes contributing to
ANC depressions. Two Pennsylvania streams experienced major rain-induced episodes in which
A" increases were the most important ion changes. In another Pennsylvania stream, NO3" pulses
were the second ranked ion change in two major episodes.
292
-------
Acidic deposition, as evidenced by stream water SO42 and NO3 during episodes, contributed
significantly to the occurrence of acidic episodes with low pH and high Al levels in all three study
areas. Although base cation decreases are often the most important ion change that occurs
during episodes, base cation decreases alone cannot create acidic stream water conditions
during episodes. Organic acid pulses, a natural source of acidity, are also important contributors
to ANC depressions in the Adirondack streams and, to a lesser extent, in the Catskill and
Pennsylvania streams. However, SO42~ or NO3~ pulses during episodes in study streams from all
three regions augment the natural processes to create episodes with lower pH and ANC and
higher Al concentrations than would have occurred from natural processes alone. In addition,
large baseline concentrations of SO42~ and NO3~ reduced episodic minimum ANC, even when
these ions did not change during episodes.
We examined episodic acidification in 13 streams that were specifically selected for the ERP. It is
not possible to make direct regional estimates of episodic acidification from these data. However,
the results should be reasonably representative of episodic acidification responses and processes
in similar streams in the Catskills, the Adirondacks, and the Northern Appalachian Plateau of
Pennsylvania.
Objective 2. Evaluate the effects of episodic acidification on fish populations in streams.
Results from the ERP clearly demonstrate that episodic acidification can have long-term adverse
effects on fish populations. Streams with suitable chemistry during low flow (pH > 6.0), but low
pH (< 5.0-5.2) and high Aljm levels (> 100-200 fig/L) during high flow, supported substantially
lower numbers and biomass of brook trout than nonacidic ERP streams and lacked acid-sensitive
fish populations, such as blacknose dace and sculpin.
Three streams were nonacidic (ANC > 0) throughout the period of study (Fly Pond Outlet, Black
Brook, High Falls Brook). Baldwin Creek had only one brief acidic episode with a minimum ANC
of -1 /teq/L. These four streams had higher brook trout biomass (average values for each stream,
1.8-3.5 kg/0.1 ha) than all other ERP streams (0.3-1.6 kg/0.1 ha), and higher brook trout density
(127-309 fish/0.1 ha) than in all but one other ERP stream (4-159 fish/0.1 ha).
Three streams were chronically acidic during the period of study (i.e., median weekly ANC < 0)
(East Branch Neversink River, Stone Run, Roberts Run). All three supported low numbers (24-71
fish/0.1 ha) and biomass of brook trout (0.3-1.1 kg/0.1 ha).
293
-------
Two streams, Buck Creek and Linn Run, were not chronically acidic but commonly had acidic
episodes lasting 10 or more days during hydrologic events. The density and biomass of trout in
these streams (4-47 fish/0.1 ha and 0.25-0.73 kg/0.1 ha) were substantially less than in nonacidic
ERP streams and similar to levels in the three chronically acidic ERP streams.
The four remaining ERP streams (Benner Run, Biscuit Brook, Seventh Lake Inlet, Bald Mountain
Brook) had chemical conditions of intermediate severity (less frequent and shorter duration acidic
episodes) and intermediate levels of brook trout density and biomass (58-158 fish/0.1 ha; 1.2-1.6
kg/0.1 ha).
As a group, streams that experienced episodic acidification had significantly (p < 0.05) lower
levels of brook trout density and biomass than nonacidic streams. Differences in trout abundance
between episodically and chronically acidic streams were not statistically significant (p > 0.05).
With one exception, acid-sensitive fish populations occurred only in ERP streams with median
weekly pH > 6.0 (Black Brook, High Falls Brook, Biscuit Brook, Benner Run, Baldwin Run) and
were absent from streams that were chronically acidic or that experienced moderate to severe
episodic acidification (Bald Mountain Brook, Seventh Lake Inlet, East Branch Neversink River, Linn
Run, Stone Run, Roberts Run).
Likewise, young-of-the-year brook trout were abundant (indicative of successful reproduction
during the study period) in Fly Pond Outlet, Black Brook, High Falls Brook, Biscuit Brook, Benner
Run, and Baldwin Creek, but were absent or rare in Bald Mountain Brook, Buck Creek, Seventh
Lake Inlet, East Branch Neversink River, and Linn Run. Young-of-the-year brook trout were
abundant in Stone Run and Roberts Run in 1988, associated with below normal precipitation
levels in winter/spring 1988, but .absent or rare in 1989, following higher than normal precipitation
and presumably more severe episodes in winter/spring 1989.
In situ bioassays demonstrated that fish exposed to episodic acidification can experience signifi-
cant mortality. Fish mortality during bioassays was significantly (p < 0.05) lower in reference
streams (Fly Pond Outlet, Black Brook, High Falls Brook, Baldwin Run, Benner Run) than in non-
reference streams. Mortality rates were significantly higher in bioassays with acidic episodes than
in bioassays with ANC > 0 ^ueq/L, but similar to bioassays that had chronically acidic conditions
(ANC :£ 0). Results were consistent across all fish species.
294
-------
Chemical conditions toxic to fish occurred at some time during the study in all ERP streams
except the five reference streams. Maximum observed mortality rates were generally highest in
those streams with the most severe chemical conditions: > 80% in Buck Creek, East Branch
Neversink River, Stone Run, and Linn Run; between 40% and 50% in Bald Mountain Brook and
Roberts Run; 20% to 30% in Seventh Lake Inlet and Biscuit Brook; and < 20% in all five reference
streams. Because bioassays were not conducted during periods with the most severe chemistry
in each stream, these results provide only a qualitative indicator of the toxicity of ERP streams. All
streams also had at least one bioassay with low mortality (< 10%), indicating that toxic conditions
do not occur throughout the year.
A net downstream movement of radio-tagged brook trout was observed in all streams and study
periods that experienced stressful chemical conditions, but little to no net downstream movement
occurred in streams with relatively high pH (> 5.1-5.2) and low Alim (< 150-160 fig/L) throughout
the study period. Downstream fish movement either was associated with chronically acidic condi-
tions at the start of the experiment or coincided with the occurrence of one or more episodes with
Alim > 160 fig/L for 1.5 or more days. Radio-tagged brook trout died when exposed to high con-
centrations of Al during an episode in Linn Run. During the same episode, radio-tagged trout
survived if they avoided exposure to peak Al levels.
Objective 3. Define key characteristics of episodes that determine the severity of effects on
fish populations.
Inorganic Al was the single best p*redictor of fish mortality during in situ bioassays. Calcium, pH,
and DOC were also important predictors of brook trout mortality; at a given Aljm concentration,
lower mortality occurred in bioassays with lower minimum pH, higher Ca, and higher DOC.
Brook trout density and biomass in ERP streams were significantly (p < 0.05) correlated with both
stream pH and Aljm concentrations, although rank correlations were slightly higher for pH than
Alim. Because pH and Alim are highly correlated, it is difficult, based on field data alone and
given the small number of study streams (n = 13), to distinguish the relative importance of pH
and Alim.
The relationships between Aljm and pH, and between Aljm and ANC, varied among ERP streams.
High levels of Aljm (> 200 /*g/L) commonly occurred in Linn Run at pH < 5.4, but only at pH <
5.2 in Stone Run, Bald Mountain Brook, Buck Creek, and Seventh Lake Inlet, and at pH < 5.0 in
Roberts Run and East Branch Neversink River. [No Aljm levels above 200 //g/L occurred in the
295
-------
remaining six ERP streams.] Thus, predictions of potential effects on fish based solely on pH or
ANC may be misleading.
Fish exposed for longer periods of time to high Alim levels, in general, had higher mortality rates.
However, the single best predictor of brook trout and sculpin mortality was the time-weighted
median Alim concentration during the 20-day bioassay period, as opposed to more complex
expressions of chemical exposure incorporating peak levels and duration.
Blacknose dace are more sensitive to high Aljm than brook trout or sculpin. Dace mortality was
best predicted by an integrated function of duration and Aljm concentrations. High mortality is
expected to occur with Aljm > 250 pg/L for two or more days.
During radiotelemetry studies, some brook trout were able to move downstream or into alkaline
microhabitats and avoided exposure to low pH and high Al during episodes. However, in most
cases, the majority of radio-tagged fish were exposed to relatively high, potentially lethal Al levels.
Fish behavioral avoidance and the occurrence of refugia can partially, but not entirely, mitigate
the adverse effects of episodic acidification on fish populations. Recolonization from groundwater
seeps and more alkaline tributary streams can maintain low densities of fish in streams that
experience toxic episodes, but is not sufficient to sustain fish densities or biomass at levels near
those expected in the absence of adverse acid-base chemistry.
Brook trout are fairly mobile, frequently moving more than 1 km. Thus, of the fish species
common in small headwater streams, brook trout are best able to take advantage of refugia.
Sculpin, by contrast, generally move only short distances (e.g., < 10 m in a summer). Sculpin
were as or more tolerant of high Aljm in bioassays than brook trout. Yet, sculpin populations are
absent from streams that maintain low densities of brook trout. We hypothesize that toxic epi-
sodes have a more severe and long-lasting effect on sculpin populations, compared to brook
trout populations, because of differences in fish mobility.
Better delineation of the specific characteristics of episodes and stream chemistry that are most
Important in controlling effects on fish and fish populations would require controlled experiments,
involving specific combinations of chemical variables and exposure magnitude, duration, and
timing.
296
-------
Objective 4. Develop and calibrate regional models of episodic chemistry that link
atmospheric deposition to biologically relevant chemistry during episodes.
We have not reported the results of modeling investigations in this report. They will be presented
in future journal articles.
Major Conclusions
Episodes were a common occurrence in the study streams of all three regions, and acidic
episodes were common when ANC values were < 50 neq/L immediately before the episode.
When acidic episodes occurred, they were accompanied by depressed pH levels and elevated
Aljm concentrations.
• Acidic deposition, as evidenced by stream water SO42 and NO3 during episodes, contributed
significantly in two ways to the occurrence of acidic episodes with low pH and high Al levels in
all three regions. Pulses of SO42~ (in Pennsylvania streams) and NO3~~ (in Catskill and
Adirondack streams) during episodes augmented natural processes to create episodes with
lower ANC and pH and higher Aljm concentrations than would have occurred from natural
processes alone. In addition, large baseline concentrations of SO42~ (all regions) and NO3~~
(Catskills and Adirondacks) reduced episodic minimum ANC levels, even when these ions did
not change or decreased slightly during episodes.
• Episodic acidification can have long-term adverse effects on fish populations. As a result,
stream assessments based solely on chemical measurements during low flow do not accur-
ately predict the status of fish communities in small streams.
• Fish exposed to low pH and high Aljm for longer periods of time experienced higher mortality.
Time-weighted median Aljm concentration was the single best predictor of brook trout mortality.
• Fish behavioral avoidance only partially mitigated the adverse effects of episodic acidification
in small streams and was not sufficient to sustain fish density or biomass at the levels
expected in the absence of acidic episodes.
297
-------
-------
REFERENCES
Aber, J.D., K.J. Nadelhoffer, P. Steadier, and J.M. Melillo. 1989. Nitrogen saturation in northern
forest ecosystems—hypotheses and implications. Bioscience 39:378-386.
Aber, J.D., J.M. Melillo, N.J. Nadelhoffer, J. Pastor, and R.D. Boone. 1991. Factors controlling
nitrogen cycling and nitrogen saturation in northern temperate forest ecosystems. Ecol.
Appl. 1:303-315.
Agresti, A. 1990. Categorical Data Analysis. John Wiley & Sons, New York.
Anderson, M.P., and C.J. Bowser. 1986. The role of groundwater in delaying lake acidification.
• Water Resour. Res. 22:1101 -1108.
Andersson, P., and P. Nyberg. 1984. Experiments with Brown Trout during Spring in Mountain
Streams at Low pH and Elevated Levels of Iron, Manganese and Aluminum. Report No. 61.
Institute of Fresh Water Research, Drottningholm, Stockholm. Pages 34-47.
Ashton, W.D. 1972. The Logit Transformation with Special Reference to its Uses in Bioassay. No.
32 of Griffin's Statistical Monographs and Courses, edited by Alan Stuart. Hafner Publishing
Co., New York.
Baker, J.P. 1981. Aluminum Toxicity to Fish as Related to Acid Precipitation and Adirondack
Surface Water Quality. Ph.D. Thesis. Cornell University, Ithaca, NY. 441 pp.
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 Eco-
systems: Regional Case Studies. Springer-Verlag, New York.
Baker, J.P., and C.L. Schofield. 1982. Aluminum toxicity to fish in acidic waters. Water Air Soil
Pollut. 18:289-309.
Baker, J.P., and C.L. Schofield. 1985. Acidification impacts on fish populations: a review.
Pages 183-221 in D.D. Adams and W.P. Page, eds. Acid Deposition Environmental, Eco-
nomic, and Policy Issues. Plenum Press, New York.
Baker, J.P., DP. Bernard, S.W. Christensen, M.J. Sale, J. Freda, K.J. Heltcher, D.R. Marmorek, L.
Rowe, P.F. Scanlon, G.W. Suter II, W.J. Warren-Hicks, and P.M. Welbourn. 1990a. Biologi-
cal Effects of Changes in Surface Water Acid-Base Chemistry. NAPAP Report 13. Acidic
Deposition: State of Science and Technology. National Acid Precipitation Assessment
Program, Washington, D.C.
Baker, J.P., S.A. Gherini, S.W. Christensen, C.T. Driscoll, J. Gallagher, R.K. Munson, R.M. Newton,
K.H. Reckhow, and C.L. Schofield. 1990b. Adirondack Lakes Survey: An Interpretive Anal-
ysis of Fish Communities and Water Chemistry, 1984-87. Adirondack Lakes Survey Corpora-
tion, Ray Brook, NY.
Baker, LA, P.R. Kaufmann, A.T. Herlihy, and J.M. Eilers. 1990. Current Status of Surface Water
Acid-Base Chemistry. NAPAP Report 9. In: Acidic Deposition: State of Science and Tech-
nology. National Acid Precipitation Assessment Program, Volume II, Washington, D.C.
299
-------
Baldigo, B.P. U.S. Geological Survey, Water Resources Division, Box 1669, Albany, NY 12201,
USA (personal communication).
Barchet, W.R. 1991. Episodic Response Project: Wet Deposition at Watersheds in Three
Regions of the Eastern United States. PNL-7876, NTIS DE92-004709. Bateile Pacific
Northwest Laboratory, Richland, WA.
Barker, J.L., and E.G. Witt. 1990. Effects of Acidic Precipitation on the Water Quality of Streams in
the Laurel Hill Area, Somerset County, Pennsylvania, 1983-86. Water Resources
Investigations Report 89-4113. U.S. Geological Survey. Harrisburg, PA. 72pp.
Barnes, R.B. 1975. The determination of specific forms of aluminum in natural water. Chem.
GeoL 15:177-191.
Bath, D.W. 1992. Adirondack Lake Survey Corporation, Ray Brook, New York. Personal commu-
nication.
Bernard, D.P. 1985. Impact of Stream Acidification on Invertebrates: Drift Response to in situ
Experiments Augmenting Aluminum Ion Concentrations. M.S. Thesis. University of British
Columbia, Vancouver. 121 pp.
Borg, H. 1986. Metal speciation in acidified mountain streams in central Sweden. Water Air Soil
Pollut 30:1007-1014.
Brezonik, P.L, LA Baker, J.R. Eaton, T.M. Frost, P. Garrison, T.K. Kratz, J.J. Magnuson, W.J.
Rose, B.K. Shephard, W.A. Swenson, C.J. Watras, and K.E. Webster. 1986. Experimental
acidification of Little Rock Lake, Wisconsin. Water Air Soil Pollut. 31:115-121.
Brown, D.J.A. 1982a. The effect of pH and calcium on fish and fisheries. Water Air Soil Pollut
18:343-351.
Brown, D.J.A. 1982b. Influence of calcium on the survival of eggs and fry of brown trout (Salmo
trutta) at pH 4.5. Bull. Environ. Contam. Toxicol. 28:664-668.
Brown, D.J.A. 1983. Effect of calcium and aluminum concentrations on the survival of brown
trout (Salmo trutta) at low pH. Bull. Environ. Contam. Toxicol. 30:582-587.
Brown, L., and J.F. Downhower. 1982. Summer movements of mottled sculpins, Cottus bairdi
(Pices: Cottidae). Cope/a 1982(2) :45CM153.
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.
Chambers, J.M., W.S. Cleveland, B. Kleiner, and P.A. Tukey. 1983. Graphical Methods for Data
Analysis. Duxbury Press, Boston.
Chen, C.W., S.A. Gherini, N.E. Peters, P.S. Murdoch, R.M. Newton, and R.A. Goldstein. 1984.
Hydrologic analyses of acidic and alkaline lakes. Water Resour. Res. 20:1875-1882.
.Chester, P.F. 1984. Chemical balance in fishless lakes. Proc. R. Soc. Lond. B Biol.
305:564-565.
300
-------
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.
Future Effects of Long-Term Sulfur Deposition on Surface Water Chemistry in the Northeast
and Southern Blue Ridge Province (Results of the Direct/Delayed Response Project).
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. 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-Appalachian Region of the Eastern United States.
EPA/600/R-92/186. U.S. Environmental Protection Agency, Washington, D.C. 384 pp.
Clarke, K.L., and B.D. LaZerte. 1985. A laboratory study of the effects of aluminum and pH on
amphibian eggs and tadpoles. Can. J. Fish. Aquat. Sci. 42:1544-1551.
Colquhoun, J., W. Kretser, and M. Pfeiffer. 1984. Acidity Status Update of Lakes and Streams in
New York State. New York State Department of Environmental Conservation, Albany, NY,
139 pp.
Correll, D.L., J.J. Miklas, A.H. Hines, and J.J. Schafer. 1987. Chemical and biological trends
associated with acidic atmospheric deposition in the Rhode River watershed and estuary.
Water Air Soil Pollut. 35:63-86.
Curtis, L.R., W.K. Seim, L.K. Siddens, D.A. Meager, R.A. Carchman, W.H. Carter, and G.A.
Chapman. 1989. Role of exposure duration in hydrogen ion toxicity to brook (Satvelinus
fontinalis) and rainbow (Salmo gairdneri) trout. Can. J. Fish. Aquat. Sci. 46:33-40.
Dana, M.T. 1982. The MAP3S Precipitation Chemistry Network Site Operator's Manual. Revision
4. PNL-4703. Battelle Pacific Northwest Laboratory, Richland, WA.
DeWalle, D.R., C.J. Gagen, M.C. Jones, R.F. Carline, W.E. Sharpe, W.G. Kimmel, B.R. Dempsey,
T.A. Pagano, J.L Baliiet, D.G. Galeone, F.F. Fiss, R.B. Warren, Jr., J.A. Lynch, and E.S.
Corbett. 1991. Episodic Acidification and Associated Fish and Benthic Invertebrate
Responses in Five Northern Appalachian Streams: An Interim Report of the Episodic
Response Project. EPA/600/3-91/035. U.S. Environmental Protection Agency, Corvallis, OR.
DeWalle, D.R., B.R. Swistock, C.L. Dow, W.E. Sharpe, and R.F. Carline. 1993. Episodic
Response Project—Northern Appalachian Plateau: Site Description and Methodology.
EPA/600/R-93/023. NTIS PB93-149755. Cooperative Agreement No. CR814566012, U.S.
EPA Environmental Research Laboratory, Corvallis, OR.
Dinicola, R.S. 1982. Geologic Controls on the Sensitivity of Headwater Streams to Acid Precipita-
tion in the Laurel Highlands. M.S. Thesis. Pennsylvania State University, University Park.
Dow, C.L.A. 1992. Sulfur and Nitrogen Budgets on Five Forested Appalachian Plateau Basins.
M.S. Thesis. Pennsylvania State University, University Park.
Driscoll, C.T. 1984. A procedure for fractionation of aqueous aluminum in dilute acidic waters.
Intern. J. Env. Anal. Chem. 16:267-283.
301
-------
Driscoll, C.T., J.P. Baker, J.J. Bisogni, Jr., and C.L. Schofield. 1980. Effect of aluminum
speciation on fish in dilute acidified waters. Nature 284:161-164.
Driscoll, C.T., C.C. Cosentini, and R.M. Newton. 1987a. Processes regulating episodic
acidification of an Adirondack, NY, stream. Pages 244-246 in Extended Abstracts,
GEOMON, International Workshop on Geochemistry and Monitoring in Representative Basins.
Malostranske n. 19, 118 21 Praha 1, Czechoslovakia.
Drfscoll, C.T., C.P. Yatsko, and F.J. Unangst. 1987b. Longitudinal and temporal trends in the
water chemistry of the North Branch of the Moose River. Biogeochemistry 3:37-61.
Driscoll, C.T., B.J. Wyskowski, C.C. Cosentini, and M.E. Smith. 1987c. Processes regulating
temporal and longitudinal variations in the chemistry of a low-order woodland stream in the
Adirondack region of New York. Biogeochemistry 3:225-241.
Driscoll, C.T., R.M. Newton, C.P. Gubala, J.P. Baker, and S.W. Christensen. 1991. Adirondack
Mountains. Pages 133-202 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.
Engblom, E., and P. Lingdell. 1984. The mapping of short-term acidification with the help of
biological pH indicators. Instit. Freshw. Res. Drottningholm 61:60-68.
Eshleman, K.N. 1985. Hydrochemical Response of a New England Watershed to Acid Deposition.
Ph.D. Dissertation. Massachusetts Institute of Technology, Cambridge. 180 pp.
Eshleman, K.N. 1988. Predicting regional episodic acidification of surface waters using empirical
models. Water Resour. Res. 24:1118-1126.
Everhart, W.H., A.W. Eipper, and W.D. Youngs. 1975. Principles of Fishery Science. Cornell
University Press, Ithaca, NY.
Fiss, F.C. 1991. Survival and Development of Brook Trout Embryos in Streams that Undergo
Episodic Acidification. M.S. Thesis. Pennsylvania State University, University Park.
Fivelstad, S., and H. Leivestad. 1984. Aluminium toxicity to Atlantic salmon (Salmo salar L.) and
brown trout (Salmo trutta L.): mortality and physiological response. Report No. 61:69-77.
Institute of Freshwater Research Drottningholm, National Swedish Board of Fisheries, Lund.
France, R.L. 1985. Low pH avoidance by crayfish (Orconectes virilis): evidence for sensory
conditioning. Can. J. Zoo/. 63:258-262.
France, R.L., and P.M. Stokes. 1987. Life stage and population variation in resistance and
tolerance of Hyalella azteca (Amphipoda) to low pH. Can. J. Fish. Aquat. Sci. 44:1102-1111.
Freeze, R.A., and J.A. Cherry. 1979. Groundwater. Prentice-Hall, Inc., Englewood Cliffs, NJ. 604
pp.
302
-------
Gagen, C.J. 1991. Direct Effects of Acidic Runoff Episodes on the Distribution and Abundance of
Fishes in Streams of the Northern Appalachian Plateau. Ph.D. Thesis. The Pennsylvania
State University, College Station, PA. 127 pp.
Gagen, C.J., and W.E. Sharpe. 1987a. Influence of acid runoff episodes on survival and net
sodium balance of brook trout (Salvelinus fontinalis) confined in a mountain stream. Pages
220-230 in H. Witters and O. Vanderborght, eds. Ecophysiology of Acid Stress in Aquatic
Organisms. Annals. Soc. R. Zoo/. Belg. 117(Supplement 1):219-230.
Gagen, C.J., and W.E. Sharpe. 1987b. Net sodium loss and mortality of three salmonid species
exposed to a stream acidified by atmospheric deposition. Bull. Environ. Contam. Toxicol.
39:7-14.
Galloway, J.N., G.R. Hendrey, C.L. Schofield, N.E. Peters, and A.H. Johannes. 1987. Processes
and causes of lake acidification during spring snowmelt.in the westcentral Adirondack
Mountains, New York. Can. J. Fish. Aquat. Sci. 44:1595-1602.
Gloss, S.P., C.L. Schofield, R.L. Spateholts, and B.A. Plonski. 1989. Survival growth,
reproduction, and diet of brook trout (Salvelinus fontinalis) stocked into lakes after liming to
mitigate acidity. Can. J. Fish. Aquat. Sci. 46:277-286.
Gunn, J.M. 1987. The Role of Episodic Acidification in the Extinction of Lake Charr (Salvelinus
namaycush) Populations. Ph.D. Thesis. University of Guelph, Guelph, Ontario.
Gunn, J.M. 1989. Survival of lake charr (Salvelinus namaycush) embryos under pulse exposure
to acidic runoff water. Pages 23-45 in J.A. Nriagu, ed. Aquatic Toxicology and Water
Quality Management. John Wiley & Sons, New York.
Gunn, J.M., and D.L.G. Noakes. 1986. Avoidance of low pH and elevated Al concentrations by
brook charr (Salvelinus fontinalis) alevins in laboratory tests. Water Air Soil Pollut.
30:497-503.
Hagley, C.A., C.L. Mayer, and R. Hoenicke. 1988. National Surface Water Survey: National
Stream Survey (Phase I, Southeastern Screening, and Episodes Pilot Surveys). Field Opera-
tions Report. EPA/600/4-88/023. U.S. Environmental Protection Agency, Las Vegas,
Nevada.
Haines, T.A. 1987. Atlantic salmon resources in the northeastern United States and the potential
effects of acidification from atmospheric deposition. Water Air Soil Pollut. 35:37-48.
Haines, T.A., S.A. Norton, J.S. Kahl, C.W. Fay, S.J. Pauwels, and C.H. Jagoe. 1990. Intensive
Studies of Stream Fish Populations in Maine. EPA/600/3-90/043. U.S. EPA Environmental
Research Laboratory, Corvallis, Oregon. 339 pp.
Hall, R.J. 1990. Relative importance of seasonal, short-term pH disturbances during discharge
variation on a stream ecosystem. Can. J. Fish. Aquat. Sci. 47:2261-2274.
Hall, R.J., G.E. Likens, S.B. Fiance, and G.R. Hendrey. 1980. Experimental acidification of a
stream in Hubbard Brook Experimental Forest, New Hampshire. Ecology 61:976-989.
Hall, R.J., C.T. Driscoll, G.E. Likens, and J.M. Pratt. 1985. Physical, chemical, and biological
consequences of episodic aluminum additions to a stream. Limnol. Oceanogr. 30:212-220.
303
-------
Hall, R.J., and P.I. Ide. 1987. Evidence of acidification effects on stream insect communities in
central Ontario between 1937 and 1985. Can. J. Fish. Aquat. Sci. 44:1652-1657.
Harriman, R., and D.E. Wells. 1985. Causes and effects of surface water acidification in
Scotland. J. Inst Water Pollut. Control 84:215-224.
Harvey, H.H., and D.M. Whelpdale. 1986. On the prediction of acid precipitation events and their
effects on fishes. Water Air Soil Pollut. 30:579-586.
Heath, R.H., J.S. Kahl, and S.A. Norton. 1992. Episodic stream acidification caused by atmos-
pheric deposition of sea salts at Acadia National Park, Maine, United States. Water Resour.
Res, 28:1081-1088.
Hendershot, W.H., A. Dufresne, H. Lalande, and F. Courchesne. 1986. Temporal variation in
aluminium speciation and concentration during snowmelt. Water Air Soil Pollut. 31:231-237.
Hesthagen, T. 1989. Episodic fish kills in an acidified salmon river in southwestern Norway.
Fisheries 14:10-17.
Hewlett, P.S., and R.L. Plackett. 1979. An Introduction to Quanta! Responses in Biology. Edward
Arnold, London.
Holtze, K.E., and N.J. Hutchinson. 1989. Lethality of low pH and Al to early life stages of six fish
species inhibiting Precambrian Shield waters in Ontario. Can. J. Fish. Aquat. Sci.
46:1188-1202.
Hooper, R.P., and C.A. Shoemaker. 1985. Aluminum mobilization in an organic headwater
stream: temporal variation and mineral dissolution disequilibria. Science 229:463-465.
Hooper, R.P., and C.A. Shoemaker. 1986. A comparison of chemical and isotopic hydrograph
separation. Water Resour. Res. 2244:1444-1454.
Hopkins, P.S., K.W. Kratz, and S.D. Cooper. 1989. Effects of an experimental acid pulse on
invertebrates in a high altitude Sierra Nevada stream. Hydrobiologia 171:45-58.
Howells, G., and D.J.A. Brown. 1987. The Loch Fleet Project, SW Scotland. Trans. R. Soc.
Edinburgh. Earth Sci. 78:241-248.
Hultberg, H. 1977. Thermally stratified acid water in late winter—a key factor inducing self-
accelerating processes which increase acidification. Water Air Soil Pollut. 7:279-294.
Ingersoll, C.G. 1986. The Effects of pH, Aluminum, and Calcium on Survival and Growth of Brook
Trout (Salvelinus fontinalis) Early Life Stages. Ph.D. Dssertation. University of Wyoming,
Laramie, WY.
Jensen, K.W., and E. Snekvik. 1972. Low pH levels wipe out salmon and trout populations in
southernmost Norway. Ambio 1 -.223-225.
Johannessen, M., A. Skarveit, and R.F. Wright. 1980. Streamwater chemistry before, during and
after snowmelt. Pages 224-225 in D. Drabltfs and A. Tollan, eds. Ecological Impact of Acid
Precipitation. SNSF Project. Proceedings of an International Conference, March 11-14,
Sandefjord, Norway.
304
-------
Johnson, D.W., and D.A. Webster. 1977. Avoidance of low pH in selection of spawning sites by
brook trout (Salvelinus fontinalis). J. Fish. Res. Board Can. 34:2215-2218.
Johnson, D.W., H.A. Simonin, J.R. Colquhoun, and P.M. Flack. 1987. In situ toxicity tests of
fishes in acid waters. Biogeochemistry 3:181-208.
Johnson, N.M., G.E. Likens, F.H. Bormann, D.W. Fisher, and R.S. Pierce. 1969. A working model
for the variation in stream water chemistry at the Hubbard Brook Experimental Forest, New
Hampshire. Water Resour. Res. 5:1353-1363.
Johnson, N.M., C.T Driscoll, J.S. Eaton, G.E. Likens, and W.H. McDowell. 1981. Acid rain,
dissolved aluminum, and chemical weathering at the Hubbard Brook Experimenal Forest,
New Hampshire. Geochim. Cosmochim. Acta 45:1421-1437.
Jones, H.C., III, J.C. Noggle, R.C. Young, J.M. Kelley, H. Olem, R.J. Ruane, R.W. Pasch, G.J.
Hufantis, 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.
Kahl, J.S., S.A. Norton, R.K. McRae, T.A. Haines, and R.B. Davis. 1989. The influence of organic
acidity on the acid-base chemistry of surface waters in Maine, USA. Water Air Soil Pollut.
46:221-233.
Kahl, J.S., S.A. Norton, T.A. Haines, E.A. Rochette, R.H. Heath, and S.C. Nodvin. 1992. Mechan-
isms of episodic acidification in low-order streams in Maine, USA. Environ. Pollut. 78:37-44.
Kahl, J.S., S.A. Norton, I.J. Fernandez, K.J. Nadelhoffer, C.T. Driscoll, and J.D. Aber. 1993.
Experimental inducement of nitrogen saturation at the watershed scale. Environ. Sci.
Technol. 27:565-567.
Kaufmann, P.R., A.T. Herlihy, J.W. Elwood, M.E. Mitch, W.S. Overton, M.J. Sale, J.J. Messer, K.A.
Cpugan, 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/600/3-88/021 a. U.S. Environmental Protection Agency, Washington, D.C. 397 pp.
Kennedy, M.M., K.C. Rice, and O.P, Bricker. 1989. The effects of storm events on episodic
acidification of headwater streams on different bedrock types in Maryland and Virginia.
Abstract. EOS 70:1116.
Kerekes, J., S. Beauchamp, R. Tordon, and T. Pollock. 1986a. Sources of sulphate and acidity in
wetlands and lakes in Nova Scotia. Water Air Soil Pollut. 31:207-214.
Kerekes, J., S. Beauchamp, R. Tordon, C. Tremblay, and T. Pollock. 1986b. Organic versus
anthropogenic acidity in tributaries of the Kejimkujik watersheds in western Nova Scotia. .
Water Air Soil Pollut. 31:165-173.
Kretser, W.A., J. Gallagher, and J. Nicolette. 1989. Adirondack Lakes Study 1984-1987: An
Evaluation of Fish Communities and Water Chemistry. Adirondack Lakes Survey Corporation,
Ray Brook, NY.
305
-------
Kretser, W.A., H.A. Simonin, D.W. Bath, B.P. Baldigo, D. DiTommaso, J. Gallagher, and M. Olson.
1991. Episodic Acidification and Associated Fish and Benthic Invertebrate Responses of
Four Adirondack Headwater Streams: An Interim Report of the Episodic Response Project.
EPA/600/3-91/036. U.S. Environmental Protection Agency, Corvallis, OR.
Kretser, W.A., H.A. Simonin, D.W. Bath, J. Gallagher, and M. Olson. In press. Episodic Acidifica-
tion and Associated Fish and Benthic Invertebrate Responses of Four Adirondack Headwater
Streams: Methods Report. EPA/600/ / Cooperative Agreement No 815189, U.S.
EPA Environmental Research Laboratory, Corvallis, OR.
Kwain, W. 1975. Effects of temperature on development and survival of rainbow trout, Salmo
gairdneri, in acid waters. J. Fish. Res. Board Can. 32:493-497.
Lacroix, G.L 1985. Survival of eggs and alevins of Atlantic salmon (Salmo salar) in relation to
stream chemistry of interstitial water in redds in some acidic streams of Atlantic Canada.
Can. J. Fish. Aquat ScL 42:292-299.
Lacroix, G.L. 1989a. Ecological and physiological responses of Atlantic salmon in acidic organic
rivers of Nova Scotia, Canada. Water Air Soil Pollut. 46:375-386.
Lacroix, G.L. 1989b. Production of juvenile Atlantic salmon (Salmo salar) in two acidic rivers of
Nova Scotia. Can. J. Fish. Aquat Sci. 46:2003-2018.
Lacroix, G.L., and D.R. Townsend. 1987. Responses of juvenile Atlantic salmon (Salmo salar) to
episodic increases in acidity of Nova Scotia rivers. Can. J. Fish. Aquat. Sci. 44:1475-1484.
Lacroix, G.L., D.J. Gordon, and D.J. Johnston. 1985. Effects of low environmental pH on the
survival, growth, and ionic composition of post-emergent Atlantic salmon (Salmo salar). '
Can. J. Fish. Aquat Sci. 42:768-775.
Landers, D.H., J.M. Eilers, D.F. Brakke, W.S. Overton, P.E. Kellar, M.E. Silverstein, R.D.
Schonbrod, R.E. Crowe, R.A. Linthurst, J.M. Omernik, S.A. Teague, and E.P. Meier. 1987.
Characteristics of Lakes in the Western United States. Volume I. Population Descriptions
and Physico-Chemical Relationships. EPA/600/3-86/054a. U.S. Environmental Protection
Agency, Washington, D.C. 176pp.
Langan, S.J. 1985. Atmospheric Deposition, Afforestation and Water Quality at Loch Dee, SW
Scotland. Ph.D. Dissertation. University of St. Andrews, Scotland.
Langan, S.J. 1987. Episodic acidification of streams at Loch Dee, SW Scotland. Trans. R. Soc.
Edinburgh. Earth Sci. 78:393-397.
LaZerte, B.D., and P.J. Dillon. 1984. Relative importance of anthropogenic versus natural
sources of acidity in lakes and streams of central Ontario. Can. J. Fish. Aquat Sci.
41:1664-1677.
Lefohn, A.S., and G.O. Klock. 1985. The possible importance of forest soil processes in defining
surface water pH depresisons. J. Air Pollut. Control Assoc. 35:632-637.
Leivestad, H., G. Hendrey, IP. Muniz, and E. Snekvik. 1976. Effects of acid precipitation on
freshwater organisms. Pages 87-111 in F.H. Braekke, ed. Impact of Acid Precipitation on
Forest and Freshwater Ecosystems in Norway. SNSF Project FR 6/76. As, Norway.
306
-------
Lewis, W.M., and M.C. Grant. 1979. Changes in the output of ions from a watershed as a result
of the acidification of precipitation. Ecology 60:1093-1097.
Linthurst, R.A., D.H. Landers, J.M. Eilers, D.F. Brakke, W.S. Overton, E.P. Meier, and R.E. Crowe.
1986. Characteristics of Lakes in the Eastern United States. Volume I. Population
Descriptions and Physico-Chemical Relationships. EPA/600/4-86/007a. U.S. Environmental
Protection Agency, Washington, D.C. 136 pp.
Lynch, J.A., and E.S. Corbett. 1989. Hydrologic control of sulfate mobility in a forested
watershed. Water Resour. Res. 25:1695-1703.
Lynch, J.A., C.M. Hanna, and E.S. Corbett. 1986. Predicting pH, alkalinity, and total acidity in
stream water during episodic events. Water Resour. Res. 22:905-912.
McAvoy, D.C. 1989. Episodic response of aluminum chemistry in an acid-sensitive
Massachusetts catchment. Water Resour. Res. 25:233-240.
McCullagh, P., and J.A. Nelder. 1989. Generalized Linear Models. 2nd edition. Chapman and
Hall, London and New York.
Metcaif, R.C., M.A. Stapanian, R. Hoenicke, L.J. Arent, D.V. Peck, J.R. Baker, and W.L Kinney.
1988. Continuous electrochemical monitoring of snow inelt in Utah, USA. Pages 179-190
in T. Thomsen, H. S0gaard, and R. Braithwhite, eds. Seventh Northern Research Basins
Symposium/Workshop: Applied Hydrology in the Development of Northern Basins. Danish
Society for Arctic Technology, Greenland Technical Organization, Copenhagen, Denmark.
Miller, R.G., Jr. 1986. Beyond Analysis of Variance, Basics of Applied Statistics. John Wiley and
Sons, New York.
Molot, L.A., P.J. Dillon, and B.D. LaZerte. 1989. Changes in ionic composition of streamwater
during snowmelt in central Ontario. Can. J. Fish. Aquat. Sci. 46:1658-1666.
Muniz, I.P., and H. Leivestad. 1980. Acidification effects on freshwater fish. Pages 84-92 in D.
Drabl^s and A. Tollan, eds. Ecological Impact of Acid Precipitation. SNSF Project.
Proceedings of an International Conference, March 11-14, Sandefjord Norway.
Munson, R.K., and S.A. Gherini. 1991. Processes influencing the acid-base chemistry of surface
waters. Pages 9-34 in D.F. Charles, ed. Acidic Deposition and Aquatic Ecosystems:
Regional Case Studies. Springer-Verlag, New York.
Murdoch, P.S. 1988. Chemical Budgets and Stream-Chemistry Dynamics at Biscuit Brook, Catskill
Mountains, New York, 1984-85. Water Resources Investigations Report 88-4035. U.S.
Geological Survey, Reston, VA.
Murdoch, P.S., and J.L Stoddard. 1992. The role of nitrate in acidification of streams in the
Catskill Mountains of New York. Water Resour. Res. 28:2707-2720.
Murdoch, P.S., and J.L. Stoddard. 1993. Chemical characteristics and temporal trends in eight
streams of the Catskill Mountains, New York. Water Air Soil Pollut. 67:367-395.
307
-------
Murdoch, P.S., C.E. Bonitz, K.W. Eakin, A.J. Ranalli, and E.G. Witt. 1991. Episodic Acidification
and Associated Fish and Aquatic Invertebrate Responses in Four Catskill Mountain Streams:
Interim Report of the Episodic Response Project. Open-File Report 90-566. U.S. Geological
Survey, Albany, NY.
Neter, J., and W. Wasserman. 1974. Applied Linear Statistical Models. Richard D. Irwin, Inc.,
Homewood, IL
Norton, S.A., J.C. Brownlee, and J.S. Kahl. 1992. Artificial acidification of a nonacidic and an
acidic headwater stream in Maine, USA. Environ. Pollut. 77:123-128.
Ormerod, S.J., N.S. Weatherley, P. French, S. Blake, and W.M. Jones. 1987. The physiological
response of brown trout Salmo trutta to induced episodes of low pH and elevated aluminium
in a Welsh hill-stream. Ann. Soc. R. Zoo/. Belg. 117:435-447.
Pagano, T.A. 1990. Evaluation of a Method to Fractionate Aluminum in Surface Waters and Inter-
pretation of the Speciation of Aluminum in Streams during Episodes of Acidic Runoff. M.S.
Thesis. Pennsylvania State University, University Park.
Peck, D.V., J.L Engels, K.M. Howe, and J.E. Pollard. 1988. Aquatic Effects Research Program:
Episodic Response Project Integrated Quality Assurance Plan. EPA 600/X-88/274. U.S.
Environmental Protection Agency, Las Vegas, NV.
Pedder, S.C.J., and E.J. Maly. 1986. The avoidance response of groups of juvenile brook trout,
Salvelinus fontinalis, to varying levels of acidity. Aquat. Toxicol. 8:111-119.
Peden, M.E., S.R. Bachmann, C.J. Brennan, B. Demir, K.O. James, B.W. Kaiser, J.M. Lockarcl,
J.E. Rothert, J. Sauer, L.M. Skowron, and M.J. Slater. 1986. Methods for Collection and
Analysis of Precipitation. Contract Report 381. Illinois State Water Survey, Champaign.
Peters, N.E., and C.T. Driscoll. 1987. Sources of acidity during snowmelt at a forested site in the
west-central Adirondack Mountains, New York. Pages 99-108 in Forest Hydrology and
Watershed Management. fAHS-AISH Publication No. 167. International Association of
Hydrological Science, Vancouver, B.C.
Peterson, R.H., K. Coombs, and J. Power. 1989. Responses of several fish species to pH
gradients. Can. J. Zoo/. 67:1566-1572.
Pfeiffer, M., and P. Festa. 1980. Acidity Status of Lakes in the Adirondack Region of New York in
Relation to Fish Resources. FW-P142 (12/9). New York State Department of Environmental
Conservation, Albany.
Phinney, D.E. 1975. Repopulation of an eradicated stream section by brook trout. Trans. Am.
Fish. Soc. 104(4):685-687.
Planas, D., and G. Moreau. 1986. Reaction of lotic periphyton to experimental acidification.
Water Air Soil Pollut. 30:681-686.
Potter, F.J., J.A. Lynch, and E.S. Corbett. 1988. Source areas contributing to the episodic
acidification of a forested headwater stream. J. Contam. Hydrol. 3:293-305.
308
-------
Raddum, G.G., and A. Fjellheim. 1984. Acidification and early warning organisms in freshwater in
western Norway. Verh. Int. Verein. Limnol. 22:1973-1980.
Ranalli, A.J., B.P. Baldigo, D.A. Horan-Ross. R.V. Allen, and P.S. Murdoch. 1993. Site Descrip-
tions, Data and Data Collection Methods, and Quality Assurance/Quality Control Procedures
Used in the Study of Episodic Stream Acidification in Four Catskill Mountain Streams, 1988-
1990. Open File Report OF 93-137. U.S. Geological Survey, Albany, NY. 80pp.
Rascher, C.M., C.T. Driscoll, and N.E. Peters. 1987. Concentration and flux of solutes from snow
and forest floor during snowrnelt in the most central Adirondack region of New York.
Biogeochemistry 3:209-224.
Rich, J.L. 1934. Glacial Geology of the Catskills. New York State Museum Bulletin 299.
Sadler, K., and S. Lynam. 1987. Some effects on the growth of brown trout from exposure to
aluminum at different pH levels. J. Fish. Biol. 3:209-219.
SAS Institute. 1989. SAS/STAT User's Guide. Version 6, 4th Ed., Vol. 2.
Schaefer, D.A., C.T. Driscoll, Jr., R. van Dreason, and C.P. Yatsko. 1990. The episodic
acidification of Adirondack lakes during snowrnelt. Water Resour. Res. 26:1639-1648.
Schnoor, J.L., W.D. Palmer, Jr., and G.E. Glass. 1984. Modeling impacts of acid precipitation for
northeastern Minnesota. Pages 155-173 in J.L. Schnoor, ed. Modeling of Total Acid
Precipitation Impacts. Butterworth Publishers, Toronto, Canada.
Schofield, C.L., and C.T. Driscoll. 1987. Fish species distribution in relation to water quality
gradients in the North Branch of the Moose River Basin. Biogeochemistry 3:63-85.
Schofield, C.L., and J.R. Trojnar. 1980. Aluminum toxicity to fish in acidified waters. Pages
347-366 in T.Y. Toribara, M.VV. Miller, and P.E. Morrow, eds. Polluted Rain. Plenum Press,
New York.
Schofield, C.L., S.P. Gloss, and D. Josephson. 1986. Extensive Evaluation of Lake Liming,
Restocking Strategies, and Fish Population Response in Acidic Lakes Following
Neutralization by Liming. Interim Progress Report NEC-86/18. Air Pollution and Acid Rain.
U.S. Fish and Wildlife Service, Washington, D.C.
Schweinforth, R.L., R.C. Young, J.R. Wright, Jr., R.W. Garber, J.E. Persall, and T.A. McDonough.
1989. Effects of,pH and Aluminum on Early Life Stages ofSmallmouth Bass and Rainbow
Trout. TV-68154A. Tennessee Valley Authority, Nashville, TN.
Seip, H.M., G. Abrahemsen, E.T. Gjessing, and A. Stuanes. 1979. Studies of Soil, Precipitation,
and Runoff Chemistry in Six Small Natural Plots (Mini-Catchments). SNSF Project IR 46/79,
As, Oslo, Norway.
Servos, M.R., and G.L Mackie. 1986. The effect of short-term acidification during the spring
snowrnelt on selected Mollusca in south-central Ontario. Can. J. Zoo/. 64:1690-1695.
Sharpe, W.E., W.G. Kimmel, E.S. Young, Jr., and D.R. DeWalle. 1983. In situ bioassays offish
mortality in two Pennsylvania streams acidified by atmospheric deposition. Northeastern
Environ. Sci. 2:171 -178.
309
-------
Sharpe, W.E., D.R. DeWalle, R.T. Leibfried, R.S. Dinicola, W.G. Kimmel, and L.S. Sherwin. 1984.
Causes of acidification of four streams on Laurel Hill in southwestern Pennsylvania. J.
Environ. Qua/. 13:619-631.
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.
Sirnonin, H.A., W.A. Kretser, D.W. Bath, M. Olson, and J. Gallagher. 1993. In situ bioassays of
brook trout (Salvelinus fontinalis) and blacknose dace (Rhinichthys atratulus) in Adirondack
streams affected by episodic acidification. Can. J. Fish Aquat Sci. 50:902-912.
Steel, R.G.D., and J.H. Torrie.
Hill, New York. 633 pp.
1980. Principles and Procedures of Statistics. 2nd ed. McGraw-
Stoddard, J.L. In press. Long-term changes in watershed retention of nitrogen: its causes and
aquatic consequences. In L.A. Baker, ed. Environmental Chemistry of Lakes and Reser-
voirs. Advances in Chemistry Series No. 237, American Chemical Society, Washington, D.C.
Stoddard, J.L., and P.S. Murdoch. 1991. Catskill Mountains. Pages 237-271 in D.F. Charles, ed.
Acidic Deposition and Aquatic Ecosystems: Regional Case Studies. Springer-Verlag, New
York.
Stumrn, W., and J.J. Morgan. 1981. Aquatic Chemistry, An Introduction Emphasizing Chemical
Equilibria in Natural Waters. John Wiley and Sons, New York.
Sullivan, T.J., H.M. Seip, and R.D. Vogt. 1987. Episodic changes in the aluminum concentrations
and other important parameters in stream arid soil water. Report No. 85-07-01-1. Senter for
Industriforskning, Blindern, Oslo, Norway, pp. 25.
Sullivan, T.J., C.T. Driscoll, S.A. Gherini, R.K. Munson, R.B. Cook, D.F. Charles, and C.P. Yatsko.
1989. Influence of aqueous aluminum and organic acids on measurement of acid neutral-
izing capacity in surface waters. Nature 338:408-410.
Swistock, B.R., D.R. DeWalle, and W.E. Sharpe. 1989. Sources of acidic stormflow in an
Appalachian headwater stream. Water Resour. Res. 25:2139-2147.
Thornton, K.W., J.P. Baker, D. Marmorek, D. Bernard, M.L. Jones, P.J. McNamee, C. Wedeles,
and K.N. Eshleman. 1988. Episodic Response Project Research Plan. U.S. Environmental
Protection Agency, Washington, D.C. 225 pp.
Topol, L.E., and S. Odzimir. 1986. Operations and Maintenance Manual for Precipitation
Measurement Systems. Revision 2. EPA 600/4-82-042b. U.S. Environmental Protection
Agency, Research Triangle Park, NC.
Turner, R.S., R.B. Cook, H.V. Meigerot, D.W. Johnson, J.W. Elwood, O.P. Bricker, S.E. Lindbergh,
and G.M. Hornberger. 1990. Watershed and Lake Processes Affecting Surface Water Acid-
Base Chemistry. NAPAP Report 10. Acidic Deposition: State of Science and Technology.
National Acid Precipitation Assessment Program, Washington, D.C.
310
-------
U.S. Environmental Protection Agency. 1987. Handbook of Methods for Acid Deposition Studies:
Laboratory Analyses for Surface Water Chemistry. EPA/600/4-87/026. U.S. Environmental
Protection Agency, Office of Research and Development, Washington, D.C.
U.S. Geological Survey (USGS). 1991. National Water Summary 1988-89: Hydrologic Events and
Floods and Droughts. USGS Water Supply Paper No. 2375, Reston, Virginia.
Van Coillie, R., C. Thellen, P.G.C. Campbell, and Y. Vigneault. 1983. Effets Toxiques de
I'Aluminium Chez les Salmonides en Relation avec des Conditions Physico-chimiques
Acides. Rapp. Tech. Can. Sci. Halieut. et Aquat. No. 1237. Ministere des Peches et des
Oceans, Quebec, Canada.
Velbel, M.A. 1985. Geochemical mass balance and weathering rates in forested watersheds of
the southern Blue Ridge. Am. J. Sci. 285:904-930.
Watras, C.J., and T.M. Frost. 1989. Little Rock Lake (Wisconsin): perspectives on an experi-
mental ecosystem approach to seepage lake acidification. Arch. Environ. Contam. Toxicol.
18:157-165.
Way, J.H. 1972. A More Detailed Discussion of the Depositional Environmental Analyses - Middle
and Upper Devonian Sedimentary Rocks, Catskill Mountain Area, New York. Ph.D. Thesis.
Rensselaer Polytechnic Institute, Troy, NY.
Weatherley, N.S., S.J. Ormerod, S.P. Thomas, and R.W. Edwards. 1988. The response of macro-
invertebrates to experimental episodes of low pH with different forms of aluminum, during a
natural spate. Hydrobiologia 169:225-232.
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. Acidic Deposition: State of
Science and Technology. National Acid Precipitation Assessment Program, Washington,
D.C. 200 pp.
Wilkinson, K.J., H.G. Jones, P.G.C. Campbell, and M. Lachance. 1992. Estimating organic contri-
butions to surface water acidity in Quebec (Canada). Water Air Soil Pollut. 61:57-74.
Winter, T.C. 1984. Modeling the interrelationship of groundwater and surface water. Pages
89-119 in J.L. Schnoor, ed. Modeling of Total Acid Precipitation Impacts. Butterworth
Publishers, Boston.
Zippin, C. 1958. The removal method of population estimation. J. Wildlife Manage. 22(1):82-90.
311
-------
-------
APPENDIX A
BIBLIOGRAPHY OF ERP PUBLICATIONS
Barchet, W.R. 1991. Episodic Response Project: Wet Deposition at Watersheds in Three
Regions of the Eastern United States. PNL-7876, NTIS DE92-004709. Batelle Pacific
Northwest Laboratory, Richland, WA.
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.
DeWalle, D.R., B.R. Swistock, C.L Dow, W.E. Sharpe, and R.F. Carline. 1993. Episodic
Response Project—Northern Appalachian Plateau: Site Description and Methodology.
EPA/600/R-93/023. NTIS PB93-149755. Cooperative Agreement No. CR814566012, U.S.
EPA Environmental Research Laboratory, Corvallis, OR.
Dow, C.L A. 1992. Sulfur and Nitrogen Budgets on Five Forested Appalachian Plateau Basins.
M.S. Thesis. Pennsylvania State University, University Park.
Fiss, F.C. 1991. Survival and Development of Brook Trout Embryos in Streams that Undergo
Episodic Acidification. M.S. Thesis. Pennsylvania State University, University Park.
Fiss, F.C., and R.F. Carline. 1993. Survival of brook trout embryos in three episodically acidified
streams. Trans. Am. Fish. Soc. 122:268-278.
Gagen, C.J. 1991. Direct Effects of Acidic Runoff Episodes on the Distribution and Abundance of
Fishes in Streams of the Northern Appalachian Plateau. Ph.D. Thesis. The Pennsylvania
State University, College Station, PA.
Gagen, C.J., W.E. Sharpe, and R.F. Carline. 1993. Mortality of brook trout, mottled scuipins, and
slimy scuipins during acidic episodes. Trans. Am. Fish. Soc. 122:616-628.
Kretser, W.A., H.A. Simonin, D.W. Bath, B.P. Baldigo, D. DiTommaso, J. Gallagher, and M. Olson.
1991. Episodic Acidification and Associated Fish and Benthic Invertebrate Responses of
Four Adirondack Headwater Streams: An Interim Report of the Episodic Response Project.
EPA/600/3-91/036. U.S. Environmental Protection Agency, Corvallis, OR.
Kretser, W.A., H.A. Simonin, D.W. Bath, J. Gallagher, and M. Olson. In press. Episodic Acidifica-
tion and Associated Fish and Benthic Invertebrate Responses of Four Adirondack Headwater
Streams: Methods Report. EPA/600/ / Cooperative Agreement No 815189, U.S.
EPA Environmental Research Laboratory, Corvallis, OR.
McDowell, M.K., and Ming Hu. In review. Episodic Response Project Database User's Guide.
U.S. Environmental Protection Agency, Corvallis, OR.
Murdoch, P.S., and J.L Stoddard. 1992. The role of nitrate in acidification of streams in the
Catskill Mountains of New York. Water Resour. Res. 28:2707-2720.
Murdoch, P.S., C.E. Bonitz, K.W. Eakin, A.J. Ranalli, and E.G. Witt. 1991. Episodic Acidification
and Associated Fish and Aquatic Invertebrate Responses in Four Catskill Mountain Streams:
313
-------
Interim Report of the Episodic Response Project.
Survey, Albany, NY.
Open-File Report 90-566. U.S. Geological
Pagano, T.A. 1990. Evaluation of a Method to Fractionate Aluminum in Surface Waters and Inter-
pretation of the Speciation of Aluminum in Streams during Episodes of Acidic Runoff. M.S.
Thesis. Pennsylvania State University, University Park.
Peck, D.V., J.L Engels, K.M. Howe, and J.E. Pollard. 1988. Aquatic Effects Research Program:
Episodic Response Project Integrated Quality Assurance Plan. EPA 600/X-88/274. U.S.
Environmental Protection Agency, Las Vegas, NV.
Simonin, H.A., W.A. Kretser, D.W. Bath, M. Olson, and J. Gallagher. 1993. In situ bioassays of
brook trout (Sah/elinus fontinalis) and blacknose dace (Rhinichthys atratulus) in Adirondacks
streams affected by episodic acidification. Can. J. Fish Aquat. Sci. 50:902-912.
Thornton, K.W., J.P. Baker, D. Marmorek, D. Bernard, M.L. Jones, P.J. McNamee, C. Wedeles,
and K.N. Eshleman. 1988. Episodic Response Project Research Plan. U.S. Environmental
Protection Agency, Washington, D.C.
314
-------
APPENDIX B
WATER CHEMISTRY PERFORMANCE EVALUATION SAMPLES AND BLANKS
B.1 PERFORMANCE EVALUATION SAMPLES
Synthetic and well-characterized natural performance evaluation (PE) samples were utilized as
single-blind samples to monitor water chemistry analyses for the ERP. Natural PE samples'were
characterized through repeated use in a large interlaboratory round-robin program sponsored by
the Long Range Transboundary Atmospheric Pollutant (LRTAP) program. For most analytes,
target values for natural samples were estimated as the mean of median values reported for each
round-robin study. Each sample was used in at least three rounds of the LRTAP round-robin.
For synthetic samples, target values represented the theoretical concentration of analyte in the
sample.
Performance objectives for precision and bias and total error were developed for each analysis,
following the approach of Hunt and Wilson (1986). These objectives were used as control criteria
for laboratory quality control samples, and for review of individual performance evaluation results.
Two natural PE samples (FN-09 from Seventh Lake, New York, and FN-10 from Big Moose Lake,
New York) were taken to the field and processed through the sample collection devices to pro-.
vide information on the effects of sample collection and handling on analytical performance.
Target values for these two samples were supplied to the laboratories to allow them to monitor
their own performance using a consistent set of samples among all three regions throughout the
entire duration of the project. Additional information about the performance evaluation sample
program for the ERP is presented in the quality assurance project plan (Peck et al., 1988).
At the conclusion of the ERP, results of PE samples were reviewed for outlying observations, and
those outliers that could be attributed to a definite cause (e.g., calibration problems, sample
contamination, or dilution) were excluded from statistical summarization. A total of 15 data values
from the Northern Appalachians region, 36 from the Adirondacks region, and 62 from the Catskills
region were considered invalid and excluded. Other errors, such as reporting errors, were
corrected and included in the summarization. All data (including values considered invalid) from
the PE samples are included as part of the overall ERP database.
In instances where internal quality control or data verification activities indicated the need to
reanalyze samples or batches of samples, associated PE samples were not reanalyzed. Thus the
315
-------
performance indicated from the PE samples may be conservative in some cases, because some
PE sample results are no longer associated with routine sample analyses. As the results show,
there were very few instances where an individual laboratory consistently achieved all
performance objectives for an individual analyte. The fact that all three laboratories, each
employing the same QA/QC program and utilizing identical QC samples, could not achieve the
established performance objectives suggests these objectives may be too stringent for routine
implementation over an extended period of time. The objectives established for the ERP were
more stringent than those established for other monitoring programs, such as the National
Surface Water Survey. For the ERP, project management and principal investigators considered
monitoring and assessment of long-term analytical performance to be more important than
performance within individual analytical batches. These objectives were based on a desire to
control long-term variability in individual laboratory performance and facilitate comparability of
chemical data among the three ERP regions. Thus, consistency in performance among the three
regions is considered more important than the failure or success of any individual laboratory in
meeting performance objectives.
Performance evaluation sample results for organic monomeric aluminum (Alom), total monomeric
aluminum (Al^), and total dissolved aluminum (Altd) are presented in Tables B-1 and B-21. For
the monomeric aluminum analytes, sample sizes for most PE samples were relatively small,
except for the two field-handled samples (FN-09 and FN-10). In addition, natural samples were
not well characterized for monomeric aluminum species, and target values were developed based
on the results of ERP analyses alone, by calculating a weighted mean value assuming unequal
sample sizes and variances among laboratories, following Taylor (1987). For Alom and Altm,
results from the Northern Appalachians laboratory are also applicable to samples collected in the
Catskills region. Organic Al and Altm data from all three regions may be subject to considerable
imprecision at higher concentrations (> 0.400 mg/L; Table B-1). For Altm, data from the
Adirondack region may be subject to positive bias at moderate concentrations (0.400 mg/L),
whereas data from the other two regions may be subject to negative bias at very high concentra-
tions (0.750 mg/L). For Altd (Table B-2), data from all three regions may be subject to
considerable imprecision at higher concentrations.
For pH (Table B-3), performance evaluation sample data indicate little bias and good precision for
acidic samples (pH < 5.00), and for more circumneutral samples, based on results of natural per-
1
Alt tables for this appendix follow the text.
316
-------
formance evaluation samples. All three regions experienced both bias and precision problems
with two synthetic samples having circumneutral target values. These samples were prepared
from carbonate compounds and were subject to considerable effects from CO2 exchange during
both transport and analysis in the laboratories. These results are presented here to illustrate the
effects that CO2 can have on pH measurements of circumneutral samples of low ionic strength (if
not protected from atmospheric exchange). Results from natural performance evaluation samples
should be used to evaluate laboratory performance in the circumneutral range. Based on these
results, data from the Northern Appalachians region are subject to more imprecision in the
circumneutral range than data from the other two regions. Peck and Metcalf (1991) provide
additional discussion of the problems associated with these synthetic samples.
Results for ANC analyses of PE samples (Table B-4) indicate data from the Northern
Appalachians region may be subject to some negative bias and larger imprecision than data from
the other two regions. Specific conductance data (Table B-5) are generally acceptable for all
three regions, although some unexplained problems (possible dilution or cross-contamination)
were noted with one of the field-handled samples (FN-09) from the Catskills region. As similar
problems were not observed for other PE samples, this is not believed to have an impact on
overall data quality. Analysis of dissolved inorganic carbon (DIG) was optional for the ERP, and
thus was not subject to the same level of quality control as other analytes. Results of PE samples
(Table B-6) indicate that DIG data from the Adirondacks may be subject to greater imprecision
than data from the Catskills region, possible due to larger variations in atmospheric carbon
dioxide concentrations in the Adirondacks laboratory affecting unprotected samples.
Results of dissolved organic carbon (DOC) analysis (Table B-7) did not indicate any major
problems with performance. There were indications of possible field handling effects in PE
samples from both the Northern Appalachians and the Catskills regions, and some evidence of
imprecision at lower concentrations from the Northern Appalachians region.
In general, no major problems were indicated from the results of cation analyses (Tables B-8
through B-12). No problems were indicated with Ca2+ (Table B-8) or Mg2+ (Table B-9) analyses.
Sodium (Table B-10) and K+ (Table B-11) results from the Northern Appalachians and Catskills
regions region show greater imprecision than data from the Adirondacks region. Ammonium was
an optional analyte and was not measured in the Catskills region. Results from PE samples
(Table B-12) indicate NH4+ data from both the Northern Appalachians and Adirondacks regions
may be more imprecise than expected at higher concentrations.
317
-------
Performance evaluation sample results for major anions indicate that Cl data from the Northern
Appalachians region (Table B-13) may be more imprecise than data from the other two regions.
Some imprecision was indicated at very low (0.3 mg/L) and very high (15 mg/L) concentrations in
the Adirondacks region, and at very high concentrations (15 mg/L) in the Catskills region,
possibly a result of dilution errors. Nitrate data from the Northern Appalachians region appear to
be more imprecise than data from the other two regions (Table B-14), and there may be some
influence of field handling in NO3~ data from the Adirondacks region. No major problems were
indicated for SO42~ data (Table B-15).
No major problems were identified for SiO analyses in any region (Table B-16). Results of some
PE samples indicate that data from the Adirondacks may be more imprecise than data from the
other two regions.
In summary, the performance evaluation sample results indicate that data from all three regions
are generally comparable (with the exception of aluminum species data) and may be combined
for certain types of analyses and interpretative activities. Systematic errors in the data set are not
an important concern for almost all analytes, and there are only a few analytes where imprecision
may affect data interpretation or its use in modeling activities. Another point of interest is that the
field-handled samples did not indicate large effects of sample collection or handling on either
random or systematic errors. This result indicates that the major source of measurement error is
the laboratory, and that laboratory-only QC data sets can provide a reasonable assessment of
data quality with respect to collection and measurement error.
B.2 BLANKS
Tables B-17 to B-19 summarize analyses of ERP field blanks. These results indicate that appro-
priate container cleaning and sample handling procedures were used during the project and that
sample contamination was not a significant problem.
B.3 REFERENCES
Peck, D.V., and R.C. Metcalf. 1991. Dilute neutral pH standard of known conductivity and acid
neutralizing capacity. Analyst 116:221 -231.
Hunt, D.T.E., and A.L Wilson. 1986. The Chemical Analysis of Water: Principles and Techniques.
Royal Society of Chemistry, London, England. 683 pp.
318
-------
Peck, D.V., J.L Engels, K.M. Howe, and J.E. Pollard. 1988. Aquatic Effects Research Program:
Episodic Response Project Integrated Quality Assurance Plan. EPA/600/X-88/274. U.S.
Environmental Protection Agency, Las Vegas, NV.
Taylor, J.K. Quality Assurance for Chemical Measurements. Lewis Publishers, Inc., Chelsea,
Michigan. 328 pp.
319
-------
Table B-1. Performance Evaluation Sample Summary for the Episodic Response Project:
Organic Monomeric and Total Monomeric Aluminum (mg/L)
Performance objectives: MDLa = 0.006; precision and bias = ±0.006 or ±3%; total error = 0.018
or 10%.
Lot Target Value n Mean Std. Dev. Biasb
Precision0 Total Errord
Adirondacks Region
Organic Monomeric Aluminum
LN-14
LN-13
LN-15
FN-09
FN-10
0.001
0.034
0.058
0.007
0.042
7
7
4
8
9
0.007
0.033
0.068
0.019
0.048
0.019
0.019
0.027
0.017
0.026
+0.006
-0.001
+0.010
+0.012
+0.006
Catskills and Northern Appalachians Regions6
LN-13
FN-09
FN-10
0.034
0.007
0.042
6
18
15
0.035
0.010
0.041
0.018
0.005
0.006
+0.001
+0.003
-0.001
±0.040
±0.038
±0.054
±0.034
±0.052
±0.036
±0.010
±0.012
0.030
0.037
0.064
0.046
0.058
0.037
0.013
0.013
Total Monomeric Aluminum
Adirondacks Region
LN-14
SYN-1
LN-13
LN-15
SYN-2
SYN-3
FN-09
FN-10
Catskills
LN-14
SYN-1
LN-13
SYN-2
SYN-3
FN-09
FN-10
* MDL =
0.010
0.050
0.058
0.146
0.400
0.750
0.041
0.158
and Northern
0.010
0.050
0.058
0.400
0.750
0.013
0.158
method detection
7
4
7
4
4
4
7
9
0.015
0.049
0.048
0.157
0.468
0.755
0.013
0.177
0.024
0.024
0.035
0.028
0.127
0.062
0.015
0.031
+0.005
-0.001
-0.010
+0.011
+ 17%
+ 1%
±0.028
+0.019
±0.048
±0.048
±0.070
±0.056
±54%
±16%
±0.030
±0.062
0.053
0.049
0.080
0.067
71%
17%
0.002
0.081
Appalachians Regions6
6
10
6
10
8
18
16
limit.
0.003
0.044
0.059
0.363
0.649
0.013
0.171
Bias = Mean — Target Value. Relative bias =
j Precision = 2 x standard
deviation.
Relative
0.001
0.006
0.009
0.021
0.079
0.006
0.016
-0.007
-0.006
+0.001
-9%
-14%
0.000
+0.013
[(Mean - Target Value) * Target Value]
precision = 2 x
[(Standard deviation *
\j-s!<-strtttN f.~i11rtt«nMi-< 1— ll lf\+
±0.002
±0.012
±0.018
±12%
±24%
±0:012
±0.032
x 100.
Mean) x 100].
anrJ U/llonn l-tOf*R\
0.009
0.018
0.019
21%
38%
0.012
0.045
Organic monomeric and total monomeric aluminum samples collected in the Catskills region were sent to the Northern
Appalachian region laboratory for analysis.
320
-------
Table B-2. Performance Evaluation Sample Summary for the Episodic Response Project:
Total Dissolved Aluminum (mg/L)
Performance objectives: MDLa = 0.006; precision and bias = ±0.006 or ±3%; total error = 0.018
or 10%.
Lot Target Value n Mean Std. Dev. Biasb Precision0 Total Errord
Adirondacks Region
LN-14
SYN-1
LN-13
LN-15
SYN-2
SYN-3
FN-09
FN-10
0.010
0.050
0.070
0.183
0.400
0.750
0.041
0.187
7
7
7
4
7
7
15
17
0.008
0.040
0.104
0.207
0.382
0.718
0.051
0.208
0.005
0.007
0.015
0.032
0.025
0.030
0.016
0.046
-0.002
-0.010
+0.034
+0.024
-4%
-4%
+0.010
+0.021
±0.010
±0.014
±0.030
±0.064
±13%
±8%
±0.032
±0.092
0.008
0.004
0.064
0.088
17%
12%
0.042
0.113
Catskills Region
LN-14
SYN-1
LN-13
LN-15
SYN-2
SYN-3
FN-09
FN-10
0.010
0.050
0.070
0.183
0.400
0.750
0.041
0.187
8
12
10
4
9
8
11
12
0.012
0.036
0.083
0.231
0.359
0.718
0.037
0.163
0.011
0.007
0.017
0.029
0.058
0.099
0.013
0.030
+0.002
-0.014
+0.013
+0.048
-10%
-4%
-0.004
-0.024
' ±0.022
±0.014
±0.034
±0.058
±32%
±28%
±0.026
±0.060
0.024
0.000
0.047
0.106
42%
32%
0.030
0.084
Northern Appalachians Region
LN-14
SYN-1
LN-13
SYN-2
SYN-3
FN-09
FN-10
0.010
0.050
0.070
0.400
0.750
0.041
0.187
10
16
10
16
13
19
19
0.014
0.048
0.052
0.392
0.759
0.041
0.171
0.009
0.006
0.011
0.040
0.122
0.012
0.024
+0.004
-0.002
-0.018
-2%
+1%
0.000
-0.016
±0.018
±0.012
±0.022
±20%
±32%
±0.024
±0.048
0.022
0.014
0.040
22%
33%
0.024
0.064
MDL = method detection limit.
Bias = Mean - Target Value. Relative bias = [(Mean - Target Value) -H Target Value] x 100.
Precision = 2 x standard deviation. Relative precision = 2 x [(Standard deviation •*• Mean) x 100].
Total error = Ibias (or relative bias}' + precision (or relative precision), following Hunt and Wilson (1986).
321
-------
Table B-3. Performance Evaluation Sample Summary for the Episodic Response Project:
pH (pH units)
Performance objectives: MDL — not appliocable; precision and bias = ±0.08; total error = 0.25.
Lot Target Value n Mean Std. Dev. Bias
Precision0 Total Errord
Adirondacks Region
SYN-1
LN-15
LN-13
SYN-2
LN-14
SYN-3
FN-09
FN-10
Catskills
SYN-1
LN-15
LN-13
SYN-2
LN-14
SYN-3
FN-09
FN-10
Northern
SYN-1
LN-15
LN-13
SYN-2
LN-14
SYN-3
FN-09
FN-10
* MQL =
4.40
5.52
5.95
6.53
7.07
7.49
6.83
5.15
Region
4.40
5.52
5.95
6.53
7.07
7.49
6.83
5.15
7
4
7
7
7
7
16
17
11
4 -
8
10
10
11
12
13
4.43
5.70
5.90
6.39
6.97
7.33
6.93
5.20
4.42
5.49
5.89
6.08
7.06
6.99
6.72
5.26
0.02
0.17
0.12
0.09
0.24
0.22
0.11
0.03
0.04
0.03
0.05
0.21
0.06
0.16
0.19
0.24
+0.03
+0.18
-0.05
-0.14
-0.10
-0.16
+0.10
+0.05
+0.02
-O.03
-0.06
-0.45
-0.01
-0.50
-0.11
+0.11
±0.04
±0.34
±0.24
±0.18
±0.48
±0.44
±0.22
±0.06
±0.08
±0.06
±0.10
±0.42
±0.12
±0.32
±0.38
±0.48
0.07
0.52
0.29
0.32
0.58
0.60
0.32
0.11
0.10
0.09
0.16
-0.87
0.13
-0.82
0.49
0.59
Appaichians Region
4.40
5.52
5.95
6.53
7.07
7.49
6.83
5.15
method detection
14
4
10
13
10
14
21
19
limit
4.48
5.71
5.89
6.29
6.90
7.10
6.78
5.22
Bias — Mean — Target Value. Relative bias =
0.05
0.12
0.08
0.20
0.16
0.24
0.11
0.07
+0.08
+0.19
-0.06
-0.24
-0.17
-0.39
-0.05
+0.07
±0.10
±0.24
±0.16
±0.40
±0.32
±0.48
±0.22
±0.14
0.18
0.43
0.22
0.64
0.49
0.87
0.27
0.21
•
[(Mean - Target Value) -s- Target Value] x 100.
° Precision « 2 x standard deviation. Relative precision = 2 x [(Standard deviation -4- Mean) x 100].
Total error = 'bias (or relative bias)! + precision (or relative precision), following Hunt and Wilson (1986).
322
-------
Table B-4. Performance Evaluation Sample Summary for the Episodic Response Project:
Acid Neutralizing Capacity faeq/L) - •. .
Performance objectives: MDLa = not applicable; precision and bias = ±2.5 or ±3%; total error =
7.5 or 10%.
Lot Target Value n Mean Std. Dev. Bias6 Precision0 Total Errord
Adirondacks Region
SYN-1
LN-15
SYN-2
LN-13
SYN-3
LN-14
FN-09
FN-10
-39.9
7.7
15.0
20.5
140.0
223.3
152.3
-O.8
7
4
7
7
7
7
16
17
-37.6
10.3
16.5
24.9
144.3
228.6
157.0
+0.9
4.6
2.9
1.4
5.9
8.7
16.4
3.8
3.9
+2.3
+2.6
+ 1.5
+4.4
+3%
+2%
+3%
+1.7
±9.2
±5.8
±2.8
±11.8
±12%
±14%
±4%
±7.8
11.5
8.4
4.3
16.2
15%
16%
12%
9.5
Catskills Region
SYN-1
LN-15
SYN-2
LN-13
SYN-3
LN-14
FN-09
FN-10
-39.9
7.7
15.0
20.5
140.0
223.3
152.3
-0.8
12
4
10
10
13
10
12
13
-39.2
11.7
17.0
23.3
133.0
224.5
147.0
1.8
Northern Appalachians Region
4.1
2.0
3.1
3.5
7.1
2.4
10.4
3.9
+0.7
+4.0
+2.0
+2.8
-5%
+0.5%
-4%
+2.6
±8.2
±4.0
±6.2
±7.0
±11%
±2%
±14%
±7.8
8.9
8.0
8.2
9.8
16%
2%
18%
10.4
SYN-1
LN-15
SYN-2
LN-13
SYN-3
LN-14
FN-09
FN-10
-39.9
7.7
15.0
20.5
140.0
223.3
152.3
-0.8
12
4
13
8
14
8
21
19
-47.2
4.8
9.2
15.5
137.3
242.1
153.2
-5.2
4.7
2.2
6.1
2.6
10.7
14.8
7.0
3.6
-7.3
-2.9
-5.8
-5.0
-2%
+8%
+1%
-4.4
±9.4
±4.4
±12.2
±5.2
±16%
±12%
±9%
±7.2
16.7
7.3
18.0
10.2
18%
20%
10%
11.6
= method detection limit
Bias = Mean - Target Value. Relative bias = [(Mean - Target Value) -s- Target Value] x 100.
Precision = 2 x standard deviation. Relative precision = 2 x [(Standard deviation * Mean) x 100].
Total error = (bias (or relative bias)! + precision (or relative precision), following Hunt and Wilson (1986).
323
-------
Table B-5. Performance Evaluation Sample Summary for the Episodic Response Project:
Specific Conductance (pS/cm @ 25°C)
Performance objectives: MDLa = not applicable; precision and bias = ±1 or ±3%; total error =
3 or 10%.
Lot Target Value n Mean Std. Dev. Biasb Precision0 Total Errord
Adirondacks Region
SYN-1
SYN-2
LN-13
LN-15
SYN-3
LN-14
FN-09
FN-10
Catskills
SYN-1
SYN-2
LN-13
LN-15
SYN-3
LN-14
FN-09
FN-10
Northern
SYN-1
SYN-2
LN-13
LN-15
SYN-3
LN-14
FN-09
FN-10
? MDL-
1.4 7
17.0 7
22.2 7
25.9 4
45.5 7
54.1 7
52.1 16
23.5 17
Region
1.4 9
17.0 10
22.2 7
25.9 4
45.5 10
54.1 10
52.1 12
23.5 13
Appalachians Region
1.4 13
17.0 14
22.2 10
25.9 4
45.5 14
54.1 10
52.1 21
23.5 19
method detection limit.
2.0
16.7
22.5
25.8
46.7
54.2
50.1
22.8
2.0
16.3
22.2
25.6
45.2
53.2
47.8
22.6
1.9
16.1
22.0
25.4
44.8
53.1
48.1
22.3
Bias " Mean - Target Value. Relative bias =
0.1 -
0.6
0.2
0.2
1.8
0.7
0.7
0.5
6.2
1.2
0.1
0.1
1.4
0.6
2.2
0.4
0.2
1.0
0.5
0.3
2.7
0.8
1.2
0.5
+0.6
-0.3
+0.3
-0.1
+3%
+0.2%
-4%
-0.7
+0.6
-0.7
0.0
-0.3
-1%
-2%
-8%
-0.9
+0.5
-0.9
-0.2
-0.5
-2%
-2%
-8%
-1.2
[(Mean - Target Value) * Target
±0.2
±1.2
±0.4
±0.4
±8%
±3%
±3%
±1.0
±0.4
±2.4
±0.2
±0.2
±6%
±2%
±9%
±0.8
+0.4
±2.0
±1.0
±0.6
±12%
±3%
±5%
±1.0
Value] x 100.
i4inn _•_ KAn
-------
Table B-6. Performance Evaluation Sample Summary for the Episodic Response Project:
Dissolved Inorganic Carbon (mg C/L)
Performance objectives: MDLa = 0.05; precision and bias = ±0.05 or ±3%; total error = 0.15 or
10%.
Lot Target Value n Mean Std. Dev. Biasb
Precision0 Total Errord
Adirondacks Region
SYN-1
SYN-2
LN-15
LN-13
SYN-3
LN-14
FN-09
FN-10
0.12
0.30
0.37
0.46
1.80
2.82
2.06
0.49
Catskills Region
SYN-1
SYN-2
LN-15
LN-13
SYN-3
LN-14
FN-09
FN-10
0.12
0.30
0.37
0.46
1.80
2.82
2.06
0.49
7
7
4
7
5
7
12
13
7
7
0
6
6
6
6
5
0.19
0.41
0.34
0.56
2.05
3.18
2.21
0.38
0.15
0.37
0.61
1.80
3.10
2.20
0.58
0.23
0.12
0.16
0.18
0.23
0.20
0.13
0.10
0.02
0.04
0.01
0.05
0.04
0.19
0.12
+0.07
+0.11
-0.03
+0.10
+14%
+ 13%
+7%
-0.11
+0.03
+0.07
+0.15
0%
+ 10%
+7%
+0.09
Northern Appalachians Region
Dissolved inorganic carbon was not measured at this laboratory.
±0.46
±0.24
±0.32
±0.36
±22%
±13%
±12%
±0.20
±0.04
±0.08
±0.02
±6%
±3%
±17%
±0.24
0.53
0.35
0.35
0.46
36%
26%
19%
0.31
0.07
0.15
0.17
6%
13%
24%
0.33
a
b
c
d
MDL = method detection limit.
Bias = Mean - Target Value. Relative bias = [(Mean - Target Value) •*• Target Value] x 100.
Precision = 2 x standard deviation. Relative precision = 2 x [(Standard deviation + Mean) x 100].
Total error = (bias (or relative bias)l + precision (or relative precision), following Hunt and Wilson (1986).
325
-------
Table B-7. Performance Evaluation Sample Summary for the Episodic Response Project:
Dissolved Organic Carbon (mg C/L)
Performance objectives: MDLa = 0.1; precision and bias = ±0.1 or ±5%; total error = 0.3 or
15%.
Lot Target Value n Mean Std. Dev. Biasb
Precision0
Total Errord
Adirondacks Region
SYN-1
LN-14
LN-15
SYN-2
LN-13
SYN-3
FN-09
FN-10
2.5
3.8
3.9
5.0
6.1
10.0
4.4
3.3
Catskills Region
SYN-1
LN-14
LN-15
SYN-2
LN-13
SYN-3
FN-09
FN-10
2.5
3.8
3.9
5.0
6.1
10.0
4.4
3.3
7
7
4
7
7
7
16
17
14
7
0
14
8
13
11
12
2.6
4.2
4.2
5.2
6.4
10.1
4.1
3.5
2.7
4.3
4.9
6.2
10.0
4.2
3.9
0.1
0.1
0.1
0.2
0.1
0.3
0.2
0.2
0.1
0.2
0.2
0.1
0.3
0.3
0.3
+4%
+ 11%
+8%
+4%
+5%
+1%
-7%
+6%
+8%
+13%
-2%
+2%
0%
-5%
+ 18%
Northern Appalachians Region
±8%
±5%
±5%
±8%
±3%
±6%
±10%
±11%
±7%
±9%
±8%
±3%
±6%
±14%
±15%
12%
16%
13%
12%
8%
7%
17%
17%
15%
22%
10%
5%
6%
19%
33%
SYN-1
LN-14
LN-15
SYN-2
LN-13
SYN-3
FN-09
FN-10
* MDL
b Bias
2.5 13 2.9
3.8 10 4.3
3.9 4 4.0
5.0 13 5.4
6.1 10 6.1
10.0 13 10.3
4.4 21 4.2
3.3 19 3.6
= method detection limit
« Mean - Target Value. Relative bias =
j Precision « 2 x standard deviation,. Relative
0 Total
0.3
0.4
0.2
0.3
0.3
0.4
0.5
0.5
+ 16%
+ 13%
+3%
+8%
0%
+3%
-5%
+9%
±21%
±19%
±10%
±11%
±10%
±8%
±24%
±28%
37%
32%
13%
19%
10%
11%
29%
37%
[(Mean - Target Value) •*• Target Value] x 100.
precision = 2
error « Ibias (or relative bias)! + precision (or relative
X [(Standard deviation -5- Mean) x 100].
precision), following
Hunt and Wilson (1986).
326
-------
Table B-8. Performance Evaluation Sample Summary for the Episodic Response Project:
Calcium (mg/L)
Performance objectives: MDLa = 0.01; precision and bias = ±0.01 or ±3%; total error = 0.03 or
10%.
Lot Target Value n Mean Std. Dev. Biasb Precision0 Total Errord
Adirondacks Region
LN-13
LN-15
SYN-1
SYN-2
LN-14
SYN-3
FN-09
FN-10
2.08
2.31
2.50
4.50
5.59
7.00
5.02
1.82
Catskills Region
LN-13
LN-15
SYN-1
SYN-2
LN-14
SYN-3
FN-09
FN-10
2.08
2.31
2.50
4.50
5.59
7.00
5.02
1.82
7
4
7
6
7
7
16
17
8
4
10
10
8
10
12
13
2.15
2.26
2.58
4.60
5.23
7.10
5.09
1.81
2.05
2.22
2.53
4.57
5.51
7.04
4.51
1.78
0.06
0.09
0.07
0.08
1.30
0.08
0.23
0.06
0.04
0.14
0.09
0.19
0.19
0.18
0.44
0.05
4-3%
-2%
4-3%
4-2%
-6%
4-1%
4-1%
-1%
-1%
-4%
4-1%
4-2%
-1%
4-1%
-10%
-2%
Northern Appalachians Region
±6%
±8%
±5%
±3%
±50%
±2%
±9%
±7%
±4%
±13%
±7%
±8%
±7%
±5%
±20%
±6%
9%
10%
5%
56%
3%
10%
8%
5%
17%
8%
10%
8%
6%
30%
8%
LN-13
LN-15
SYN-1
SYN-2
LN-14
SYN-3
FN-09
FN-10
b'pDL
Kan
2.08
2.31
2.50
4.50
5.59
7.00
5.02
1.82
= method detection
= Moan _ Tarnot Ual
10
4
13
14
10
14
21
19
limit.
Ilia Ral
2.08
2.38
2.48
4.54
5.48
7.01
4.87
1.74
ativa hiats = fttj
0.11
0.17
0.16
0.23
0.14
0.25
0.24
0.13
loan _ TJarnot
0%
4-3%
-1%
+ t%
-2%
0%
-3%
-4%
\fa\t ic*\ -*- Tsirrto'1
±11%
±14%
±13%
±10%
±5%
±7%
±10%
±15%
t Valnol y mn
11%
17%
14%
11%
7%
7%
13%
19%
Precision = 2 x standard deviation. Relative precision = 2 x [(Standard deviation -s- Mean) x 100].
Total error = Ibias (or relative bias)l + precision (or relative precision), following Hunt and Wilson (1986).
327
-------
Table B-9. Performance Evaluation Sample Summary for the Episodic Response Project:
Magnesium (mg/L)
Performance objectives: MDLa = 0.01; precision and bias = ±0.01 or ±3%; total error = 0.03 or
10%.
Lot Target Value n Mean Std. Dev. Bias13
Precision0 Total Errord
Adfrondacks Region
LN-15
LN-13
SYN-1
SYN-2
LN-14
SYN-3
FN-09
FN-10
Catskills
LN-15
LN-13
SYN-1
SYN-2
LN-14
SYN-3
FN-09
FN-10
Northern
LN-15
LN-13
SYN-1
SYN-2
LN-14
SYN-3
FN-09
FN-10
* MDL =
0.36
0.48
0.50
0.80
1.47
1.50
0.78
0.29
Region
0.36
0.48
0.50
0.80
1.47
1.50
0.78
0.29
Appalachians
0.36
0.48
0.50
0.80
1.47
1.50
0.78
0.29
method detection
4
7
7
7
7
7
16
17
4
8
10
11
8
11
12
12
Region
4
10
12
14
10
13
20
19
limit
0.37
0.48
0.52
0.83
1.46
1.51
0.82
0.30
0.35
0.48
0.52
0.82
1.46
1.52
0.73
0.29
0.34
0.47
0.51
0.82
1.42
1.46
0.78
•0.28
Bias = Mean - Target Value. Relative bias =
° Precision = 2 x standard
deviation.
Total error = 'bias (or relative bias)!
Relative
0.01
0.01
0.04
0.04
0.04
0.04
0.06
0.02
0.02
0.02
0.01
0.02
0.04
0.05
0.06
0.02
0.02
0.02
0.01
0.03
0.04
0.16
0.02
0.02
[(Mean - Target
precision = 2 x
+3%
0%
+4%
+4%
-1%
+ 1%
+5%
+0.01
-3%
0%
+4%
+2%
-1%
+1%
-6%
0.00
-6%
-2%
+2%
+2%
-3%
-3%
0%
-0.01
±5%
±4%
±15%
±10%
- ±5%
±5%
±15%
±0.04
±11%
±8%
±4%
±5%
±5%
±7%
±16%
±0.04
±12%
±8%
±4%
±7%
±6%
±22%
±5%
±0.04
8%
4%
19%
14%
6%
6%
20%
0.05
14%
8%
8%
7%
6%
8%
22%
0.04
18%
10%
6%
9%
9%
25%
5%
0.05
Value) * Target Value] x 100.
[(Standard deviation •*• Mean) x 100].
+ precision (or relative precision), following
Hunt and Wilson (1986).
328
-------
Table B-10. Performance Evaluation Sample Summary for the Episodic Response Project:
Sodium (mg/L)
Performance objectives: MDLa = 0.01; precision and bias = ±0.01 or ±3%; total error = 0.03 or
10%.
Lot Target Value n Mean Std. Dev. Biasb Precision0 Total Errord
Adirondacks Region
LN-13
LN-15
SYN-1
LN-14
SYN-2
SYN-3
FN-09
FN-10
0.54
0.70
0.80
1.56
3.50
6.00
2.59
0.57
Catskills Region
LN-13
LN-15
SYN-1
LN-14
SYN-2
SYN-3
FN-09
FN-10
0.54
0.70
0.80
1.56
3.50
6.00
2.59
0.57
7
4
7
7
7
7
16
17
8
4
11
8
11
10
12
13
0.54
0.69
0,82
1.57
3.55
6.04
2.59
0.56
0.53
0.67
0,80
1.56
3.47
5.85
2.41
0.57
0.02
0.01
0.10
0.02
0.23
0.21
0.05
0.06
0.02
0.01
0.03
0.04
0.06
0.14
0.20
0.03
0%
-1%
+3%
+1%
+ 1%
+ 1%
0%
-2%
-2%
-4%
0%
0%
-1%
-3%
-7%
0%
Northern Appalachians Region
±7%
±3%
±24%
±3%
±13%
±7%
±4%
±21%
±8%
±3%
±8%
±5%
±3%
±5%
±17%
±11%
7%
4%
27%
4%
14%
8%
4%
23%
10%
7%
8%
5%
4%
8%
24%
11%
LN-13
LN-15
SYN-1
LN-14
SYN-2
SYN-3
FN-09
FN-10
u MDL
o Bias
0.54
0.70
0.80
1.56
3.50
6.00
2.59
0.57
8
4
13
10
14
13
21
19
0.55
0.54
0,81
1.41
3.43
6.08
2.44
0,56
0.01
0.16
0.08
0.23
0.37
0.15
0.31
0.10
2%
-23%
1%
-10%
-2%
1%
-6%
-2%
±4%
±59%
±20%
±33%
±22%
±5%
±25%
±36%
6%
82%
21%
43%
24%
6%
31%
38%
= method detection limit.
= Mean - Target
Value. Relative bias =
Precision = 2. x standard deviation,.
Total
error = Ibias (or
relative bias)!
Relative
[(Mean — Target
precision = 2 x
Value) -=- Target
Value] x 100.
[(Standard deviation -H Mean) x 100].
+ precision (or relative precision), following
Hunt and Wilson (1986).
329
-------
Table B-11. Performance Evaluation Sample Summary for the Episodic Response Project:
Potassium (mg/L)
Performance objectives: MDLa = 0.01; precision and bias = ±0.01 or ±3%; total error = 0.03 or
10%.
Lot Target Value n Mean Std. Dev. Bias
Precision0 Total Errord
Adirondacks Region
SYN-1
LN-13
SYN-2
LN-15
SYN-2
LN-14
FN-09
FN-10
Catskills
SYN-1
LN-13
SYN-2
LN-15
SYN-2
LN-14
FN-09
FN-10
Northern
SYN-1
LN-13
SYN-2
LN-15
SYN-2
LN-14
FN-09
FN-10
* MDL-
0.25 7
0.32 7
0.35 7
0.42 4
0.60 7
0.73 7
0.47 16
0.32 17
Region
0.25 10
0.32 8
0.35 11
0.42 4
0.60 11
0.73 8
0.47 12
0.32 13
Appalachians Region
0.25 13
0.32 8
0.35 13
0.42 4
0.60 14
0.73 10
0.47 21
0.32 19
method detection limit.
0.25
0.31
0.34
0.42
0.59
0.72
0.44
0.31
0.25
0.32
0.36
0.41
0.60
0.74
0.42
0.30
0.24
0.33
0.36
0.38
0.59
0.71
0.44
0.30
BJas •* Moan - Target Value. Relative bias =
Dro^lcTon K 9 V etanrlarri Hov/iatinn RAlativo
0.02
0.01
0.07
0.01
0.04
0.01
0.01
0.03
0.02
0.01
0.04
0.02
0.02
0.04
0.05
0.02
0.02
0.01
0.03
0.04
0.04
0.05
0.04
0.06
0.00
-0.01
-3%
0.00
-2%
-1%
-6%
-0.01
0.00
0.00
+3%
-2%
0%
+1%
-11%
-0.02
-0.01
0.01
+3%
-10%
-2%
-3%
-6%
-0.02
±0.04
±0.02
±41%
±5%
±14%
±3%
4%
±0.06
±0.04
±0.02
±22%
±10%
±7%
±11%
24%
±0.04
±0.04
±0.02
±17%
±21%
±14%
±14%
18%
±0.12
0.04
0.03
44%
5%
16%
4%
10%
0.07
0.04
0.02
25%
12%
7%
12%
35%
0.06
0.05
0.03
20%
31%
16%
17%
24%
0.14
[(Mean - Target Value) * Target Value] x 100.
nrof!i«5i'nn = 9 v lYRtanrlarH riouiation -s- MoarA * mm
. n^VIJWWll fc-^ OVOIIV4WIVI U^VIdUWII.
Total error = I bias (or relative bias) I
+ precision (or relative precision), following Hunt and Wilson (1986).
330
-------
Table B-12. Performance Evaluation Sample Summary for the Episodic Response Project:
Ammonium (mg/L)
Performance objectives: MDLa = 0.01; precision and bias = ±0.01 or ±5%; total error = 0.03 or
15%.
Lot Target Value n Mean Std. Dev. Biasb Precision0 Total Errord
Adirondacks Region
LN-14
LN-15
LN-13
SYN-1
SYN-2
SYN-3
FN-09
FN-10
0.01
0.03
0.04
0.10
0.25
0.50
0.01
0.04
7
4
7
7
7
6
15
16
0.00
0.04
0.06
0.10
0.24
0.48
0.02
0.04
0.02
0.01
0.01
0.01
0.03
0.08
0.03
0.02
-0.01
+0.01
+0.02
0.00
-4%
-4%
+0.01
0.00
Catskills Region
Ammonium was not determined for this region.
Northern Appalachians Region
±0.04
±0.02
±0.02
±0.02
±25%
±33%
±0.06
±0.04
0.05
0.03
0.04
0.02
29%
37%
0.07
0.04
LN-14
LN-15
LN-13
SYN-1
SYN-2
SYN-3
FN-09
FN-10
0.01
0.03
0.04
0.10
0.25
0.50
0.01
0.04
10
4
10
13
13
14
21
18
0.01
0.03
0.05
0.10
0.23
0.46
0.01
0.03
0.005
0.003
0.01 '
0.02
0.02
0.04
0.01
0.01
0.00
0.00
+O.01
0.00
-8%
-8%
0.00
-0.01
±0.01
±0.01
±0.02
±0.04
±17%
±17%
±0.02
±0.02
0.01
0.01
0.03
0.04
25%
25%
0.02
0.03
MDL = method detection limit
Bias = Mean - Target Value. Relative bias = [(Mean - Target Value) -e- Target Value] x 100.
Precision = 2 x standard deviation. Relative precision = 2 x [(Standard deviation H- Mean) x 100].
Total error = Ibias (or relative bias)! + precision (or relative precision), following Hunt and Wilson (1986).
331
-------
Table B-13. Performance Evaluation Sample Summary for the Episodic Response Project:
Chloride (mg/L)
Performance objectives: MDLa = 0.01; precision and bias = ±0.01 or ±3%; total error = 0.03 or
10%.
Lot Target Value n Mean Std. Dev. Biasb Precision0 Total Errord
Adirondacks Region
LN-13
LN-15
SYN-1
LN-14
SYN-2
SYN-3
FN-09
FN-10
0.36
0.50
0.80
2.09
5.00
15.0
3.96
0.30
Catskills Region
LN-13
LN-15
SYN-1
LN-14
SYN-2
SYN-3
FN-09
FN-10
0.36
0.50
0.80
2.09
5.00
15.00
3.96
0.30
7
4
7
7
7
7
16
17
7
4
10
7
13
12
11
12
0.32
0.49
0.75
1.99
4.77
14.7
3.93
0.32
0.37
0.50
0.79
2.04
5.09
15.55
3.92
0.35
0.02
0.03
0.02
0.04
0.13
1.1
0.13
0.05
0.01
0.003
0.02
0.05
0.19
1.48
0.23
0.04
-11%
-2%
-6%
-5%
-5%
-2%
-1%
+7%
+3%
0%
-1%
-2%
+2%
+4%
-1%
+17%
Northern Appalachians Region
±13%
±12%
±5%
±4%
±5%
±15%
±7%
±31%
±5%
±1%
±5%
±5%
±7%
±19%
±12%
±23%
24%
14%
11%
9%
10%
17%
8%
38%
8%
1%
6%
7%
9%
23%
13%
40%
LN-13
LN-15
SYN-1
LN-14
SYN-2
SYN-3
FN-09
FN-10
* MOL
b Rl«,o
0.36
0.50
0.80
2.09
5.00
15.00
3.96
0.30
— method detection
«• KArian Tor*"*«+ Wai
10
4
15
10
14
12
21
19
limit
li IA QA
0.36
0.49
0.74
2.04
4.78
15.26
3.82
.0.33
tatiifA l^i^e* — f/
0.03
0.06
0.09
0.11
0.59
1.02
0.55
6.03
'K^Aon Tar*"iAl
0%
-2%
-8%
-2%
-4%
+2%
—4%
+10%
•\/alii^ _i. Tarn At
±17%
±24%
±24%
±11%
±25%
±13%
±29%
±18%
Vain^i v -inn
17%
26%
32%
13%
29%
15%
33%
28%
j Precision = 2 x standard deviation. Relative precision = 2 x [(Standard deviation •*• Mean) x 100].
Total error -Ibias (or relative bias)l + precision (or relative precision), following Hunt and Wilson (1986).
332
-------
Table B-14. Performance Evaluation Sample Summary for the Episodic Response Project:
Nitrate (mg/L)
Performance objectives: MDLa = 0.006; precision and bias = ±0.006 or ±3%; total error = 0.018
or 10%.
Lot Target Value n Mean Std. Dev. Bias
Precision0 Total Errord
Adirondacks Region
LN-14
LN-13
SYN-1
LN-15
SYN-2
SYN-3
FN-09
FN-10
0.055
0.126
0.300
1.415
3.000
6.000
1.063
0.886
Catskills Region
LN-14
LN-13
SYN-1
LN-15
SYN-2
SYN-3
FN-09
FN-10
0.055
0.126
0.300
1.415
3.000
6.000
1.063.
0.886
7
7
7
4
7
7
16
17
7
7
10
4
12
12
11
12
0.019
0.142
0.309
1.442
2.980
5.960
0.979
0.945
0.050
0.120
0.312
1.438
2.918
5.722
0.925
0.899
0.047
0.053
0.017
0.080
0.118
0.360
0.062
0.069
0.000
0.014
0.034
0.016
0.104
0.289
0.057
0.023
-0.036
+0.016
+3%
+2%
-1%
-1%
-8%
+7%
-0.005
-0.006
+4%
+2%
-3%
-5%
-13%
+2%
±0.094
±0.106
±11%
±11%
±8%
±12%
±13%
±15%
±0.000
±0.028
±22%
±2%
±7%
±10%
±12%
±5%
0.140
0.122
14%
13%
9%
13%
21%
22%
-0.005
0.034
26%
4%
10%
15%
25%
7%
Northern Appalachians Region
LN-14
LN-13
SYN-1
LN-15
SYN-2
SYN-3
FN-09
FN-10
* MDL-=
Rin-c =
0.055
0.126
0.300
1.415
3.000
6.000
1.063
0.886
method detection
Moan — Tar-not \/fll
10
10
15
4
13
12
19
19
limit.
no Rol
0.046
0.120
0.313
1.147
2.985
6.019
0.898
0.855
atiua hias = lYKj
0.034
0.053
0.034
0.021
0.166
0.867
0.091
0.102
loan — Tarnot
-0.009
-0.006
+4%
-19%
-0.5%
+0.3%
-16%
-4%
Vail 10^ -=- Tarnat
±0.068
±0.106
±22%
±4%
±11%
±29%
±20%
±24%
Vah ml -x inn
0.077
0.106
26%
23%
12%
29%
36%
28%
Precision = 2 x standard deviation. Relative precision = 2 x [(Standard deviation -s- Mean) x 100].
Total error = I bias (or relative bias)! -
i + precision (or relative precision), following Hunt and Wilson (1986).
333
-------
Table B-15. Performance Evaluation Sample Summary for the Episodic Response Project:
Sulfate (mg/L)
Performance objectives: MDLa = 0.06; precision and bias = ±0.06 or ±3%; total error = 0.18 or
10%.
Lot Target Value n Mean Std. Dev. Biasb Precision0 Total Errord
Adirondacks Region
SYN-1
LN-13
SYN-2
LN-15
LN-14
SYN-3
FN-09
FN-10
Catskills
SYN-1
LN-13
SYN-2
LN-15
LN-14
SYN-3
FN-09
FN-10
Northern
SYN-1
LN-13
SYN-2
LN-15
LN-14
SYN-3
FN-09
FN-10
* MDL =
2.00 7
5.60 7
6.00 7
6.45 4
8.80 7
10.00 7
6.36 16
5.65 17
Region
2.00 12
5.60 7
6.00 13
6.45 4
8.80 7
10.00 10
6.36 11
5.65 13
Appalachians Region
2.00 11
5.60 10
6.00 14
6.45 4
8.80 10
10.00 15
6.36 21
5.65 19
method detection limit.
1.96
5.54
5.87
6.32
8.80
9.87
6.45
5.87
2.02
5.63
5.92
6.30
8.94
9.93
6.35
5.90
1.78
5.38
5.60
6.11
8.71
9.67
6.26
5.67
Bias - Mean - Target Value. Relative bias =
Pro^Id/Mi = *> V <5*anHarH riftvlatinn Rnlativo
0.04
0.08
0.06
0.08
0.18
0.14
0.07
0.17
0.03
0.08
0.08
0.39
0.15
0.09
0.32
0.10
0.09
0.20
0.34
0.12
0.37
0.60
0.48
0.34
[(Mean - Target
nrof^i-cinn = 9 V
-2%
-1%
-2%
-2%
0%
-1%
+ 1%
+4%
+1%
0.5%
-1%
-2%
+2%
-1%
-0.2%
+4%
-11%
-4%
-7%
-5%
-1%
-3%
-2%
+0.4%
Value) -s- Target
UStanriarH riovia
±4%
±3%
±2%
±2%
±4%
±3%
±2%
±6%
±3%
+3%
±3%
±12%
±4%
±2%
±10%
±3%
±10%
±7%
±12%
±4%
±8%
±12%
±15%
±12%
Value] x 100.
ttinn .4- MoarA v 1Om
6%
4%
4%
4%
4%
4%
3%
10%
4%
4%
4%
14%
6%
3%
10%
7%
21%
11%
19%
9%
9%
15%
17%
12%
Total error - Ibias (or relative bias)! + precision (or relative precision), following Hunt and Wilson (1986).
334
-------
Table B-16. Performance Evaluation Sample Summary for the Episodic Response Project:
Silica (mg/L)
Performance objectives: MDLa = 0.05; precision and bias = ±0.05 or ±3%; total error = 0.15 or
10%.
Lot Target Value n Mean Std. Dev. Biasb Precision0 Total Errord
Adirondacks Region
SYN-1
LN-13
LN-14
SYN-2
LN-15
SYN-3
FN-09
FN-10
0.70
1.10
2.15
3.00
4.16
7.00
4.07
3.47
Catskills Region
SYN-1
LN-13
LN-14
SYN-2
LN-15
SYN-3
FN-09
FN-10
0.70
1.10
2.15
3.00
4.16
7.00
4.07
3.47
7
7
7
7
4
7
15
17
8
8
8
8
0
7
11
13
0.66
1.10
2.24
3.05
4.32
7.06
4.33
3.74
0.76
1.19
2.35
3.15
6.70
4.18
3.74
0.04
0.08
0.19
0.13
0.30
0.47
0.34
0.23
0.07
0.08
0.07
0.13
0.15
0.29
0.22
-0.04
0.00
+4%
+2%
+4%
+1%
+6%
+8%
0.06
0.09
+9%
+5%
-4%
+3%
+8%
Northern Appalachians Region
±0,08
±0.16
±17%
±8%
±14%
±13%
±16%
±12%
±0.14
±0.16
±6%
±8%
±5%
±14%
±12%
0.12
0.16
21%
10%
18%
14%
22%
20%
0.20
0.25
15%
13%
9%
17%
20%
SYN-1
LN-13
LN-14
SYN-2
LN-15
SYN-3
FN-09
FN-10
0.70
1.10
2.15
3.00
4.16
7.00
4.07
3.47
11
10
10
16
4
13
21
19
0.72
1.06
2.00
2.88
3.65
6.7'2
4.02
3.34
0.07
0.07
0.10
0.16
0.18
0.31
0.27
0.28
+0.02
-0.04
-7%
-4%
-12%
-4%
-1%
-4%
±0.14
±0.14
±10%
±11%
10%
±9%
±13%
±17%
0.16
0.18
17%
15%
22%
13%
14%
21%
r MDL = method detection limit.
Bias = Mean - Target Value. Relative bias = [(Mean - Target Value) •*• Target Value] x 100.
° -Precision = 2 x standard deviation. Relative precision = 2 x [(Standard deviation + Mean) x 100].
Total error = bias (or relative bias)! + precision (or relative precision), following Hunt and Wilson (1986).
335
-------
Table B-17. Field Blank Sample Results for Measured Variables for the Adirondacks
Constituent
characteristics
Aluminum, organic monomeric (mg/L)
Aluminum, total dissolved (mg/L)
Aluminum, total monomeric (mg/L)
ANC fceq/L)
Calcium (mg/L)
Chloride (mg/L)
Specific conductance (uS)
Dissolved inorganic carbon (mg/L)
Dissolved organic carbon (mg/L)
Potassium (mg/L)
Magnesium (mg/L)
Sodium (mg/L)
Ammonium (mg/L)
Nitrate (mg/L)
pH (pH units)
Silica (mg/L)
Suifate (mg/L)
Number of
Observations
69
128
71
128
105
107
128
74
106
105
105
105
107
107
128
107
107
Mean
-0.0003
-0.0019
0.0007
2.7480
0.0845
0.0364
1.6581
0.1421
0.2539
0.0113
0.0084
0.0076
0.0111
0.0016
5.6928
0.0238
0.0718
Minimum
-0.0340
-0.0228
-0.0360
-5.8300
-0.0470
-0.0990
0.7600
-0.1320
-0.2200
-0.0510
-0.0160
-0.0530
-0.0150
-0.0538
5.1800
-0.1780
0.0300
Maximum
0.0480
0.0299
0.0490
68.0900
5.6040
0.5160
18.4200
0.9950
1.4100
0.8870
0.3000
0.3160
0.2770
0.1448
6.8200
2.3660
1.0830
Standard
Deviation
0.0171
0.0078
0.0165
7.0736
0.5637
0.0766
2.1185
0.1435
0.3928
0.0909
0.0332
0.0526
0.0326
0.0259
0.1930
0.2626
0.1207
Lower 95%
Cl
-0.0344
-0.0173
-0.0321
-11.2483
-1.0332
-0.1153
-2.5338
-0.1439
-0.5248
-0.1690
-0.0574
-0.0967
-0.0536
-0.0498
5.3109
-0.4967
-0.1675
Upper 95%
Cl
0.0338
0.0135
0.0335
16.7443
1.2022
0.1.881
5.8499
0.4280
1.0326
0.1916
0.0742
0.1119
0.0757 ,
0.0530
6.0748
0.5444
0.3111
Table B-18. Field Blank Sample Results for Measured Variables for the Catskills
Constituent Number of
characteristics Observations Mean
Aluminum, organic monomeric (mg/L)
Aluminum, told dissolved (mg/L)
Aluminum, told monomeric (mg/L)
ANC («eq/L)
Calcium (mg/L)
Chloride (mg/L)
Specific conductance (uS)
Dissolved inorganic carbon (mg/L)
Dissolved organic carbon (mg/L)
Potassium (mg/L)
Magnesium (mg/L)
Sodium (mg/L)
Nitrate (mg/L)
pH (pH units)
Silica (mg/L)
6
52
*8
60
. 64
3
65
45
61
62
63
64
2
66
41
0.0217
0.0189
0.0581
1.4810
0.0484
0.1197
1.1620
0.2320
0.7367
0.0134
0.0116
0.0377
0.0440
5.6941
0.3098
Minimum
0.0030
-0.0010
0.0020
-9.0500
0.0050
0.0890
0.6900
0.0700
0.1800
0.0100
0.0100
0.0100
0.0430
5.1900
0.0100
Maximum
0.091
0.411
0.420
10.650
0.700
0.174
2.430
0.940
10.400
0.100
0.040
0.840
0.045
6.370
6.920
Standard
Deviation
0.03426
0.06786
0.14631
4.16079
0.09195
0.04718
0.34633
0.15938
1.29975
0.0128Q
0.00515
0.10804
0.00141
0.21218
1.07605
Lower 95%
Cl
-0.0622
-0.1173
-0.2793
-6.8418
-0.1353
-0.0305
0.4703
-0.0890
-1.8623
-0.0122
0.0013
-0.1782
0.0379
5.2705
-1.8634
Upper 95%
Cl
0.1055
0.1550
0.3955
9.8038
0.2321
0.2698
1.8537
0.5530
3.3357
0.0390
0.0219
0.2535
0.0501
6.1177
2.4829
336
-------
Table B-19. Field Blank Sample Results for Measured Variables for Pennsylvania
Constituent
characteristics
Number of Standard
Observations Mean Minimum Maximum Deviation
Lower 95% Upper 95%
Cl . Ci
Aluminum, organic monomeric (mg/L)
Aluminum, total dissolved (mg/L)
Aluminum, total monomeric (mg/L)
ANC (ueq/L)
Calcium (mg/L)
Chloride (mg/L)
Specific conductance («S)
Dissolved organic carbon (mg/L)
Potassium (mg/L)
Magnesium (mg/L)
Sodium (mg/L)
Ammonium (mg/L)
Nitrate (mg/L)
pH(pH units)
Silica (mg/L)
Sulfate (mg/L)
4
24
4
24
9
9
9
8
9
9
9
9
9
25
9
8
0.0060
0.0022
0.0008
-5.4025
0.0308
0.0243
1.4211
0.6500
-0.4418
0.0034
0.0132
0.0060
0.0182
5.5996
0.1111
0.0256
0.0000
-0.0010
0.0000
-12.3300
0.0080
0.0000
1.0200
0.0900
-2.0000
0.0000
0.0010
0.0000
0.0000
4.7800
0.0500
0.0000
0.0220
0.0070
0.0030
4.6100
0.0640
0.0510
3.2100
1.5100
0.0120
0.0070
0.0650
0.0120
0.0440
5.8400
0.200
0.039
0.0107
0.0021
0.0015
3.5674
0.0196
0.0190
0.6844
0.4663
0.8834
0.0023
0.0203
0.0030
0.0173
0.2237
0.0547
0.0125
-0.0237
-0.0021
-0.0034
-12.7653
-0.0136
-0.0186
-0.1270
-0.4253
-2.4403
-0.0018
-0.0328
-0.0008
-0.0208
5.1389
-0.0125
-0.0033
0.0357
0.0065
0.0049
1.9603
0.0752
0.0673
2.9693
1.7253
1.5567
0.0086
0.0592
0.0128
0.0573
-
0.2347
0.0545
337 -tr'u.S. GOVERNMENT PRINTING OFFICE: 1994 - 550-001/80328
-------
-------