Content uploaded by Kyle J. Hartman
Author content
All content in this area was uploaded by Kyle J. Hartman on Jun 11, 2014
Content may be subject to copyright.
Hydrobiologia 515: 203–213, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands. 203
Drought effect on stream morphology and brook trout (Salvelinus
fontinalis) populations in forested headwater streams
James P. Hakala2& Kyle J. Hartman1,∗
1Wildlife & Fisheries Resources Program, West Virginia University, 322 Percival Hall, Morgantown,
WV 26506-6125, U.S.A.
E-mail: hartman@wvv.edu (∗Author for correspondence)
2Georgia Department of Natural Resources, Fisheries Management Section, 231 Fish Hatchery Rd.,
Summerville, GA 30747, U.S.A.
Received 7 February 2003; in revised form 4 September 2003; accepted 10 September 2003
Abstract
We assessed the effect of a severe drought in 1999 upon stream morphology and brook trout (Salvelinus fontinalis)
populations in seven headwater streams in the Greenbrierand Potomac River watersheds, West Virginia. During the
drought, stream discharge was 96% lower than in years of normal precipitation. As a result, habitat availability and
quality over all study streams was significantly lower. Riffle area was greatly reduced (−54%) relative to available
pool area (−2%). Fine sediment levels (<0.063 mm) significantly increased within spawning substrate (p=0.01).
Water temperature and dissolved oxygen were adequate (mean 15.8 ◦C, >6.0 mg l−1, respectively) for brook trout
survival in all streams during the drought. Brook trout populations were significantly reduced (adult 60%, Young-
of-the-year 67%), and individual fish had significantly lower body condition during the drought relative to the
post-drought period. Reductions in brook trout density and population condition during, and in the-post drought
period, were related to spatially-limited food resources and/or increased fine sediment levels, but not to degraded
water quality. Fisheries managers should consider the effect of periodic drought on brook trout populations and
consider short-term harvest restrictions to abet recovery after such stochastic events.
Introduction
In the northern Appalachians, first- or second-order
streams are among the most affected hydrologically
during drought conditions. Droughts in these systems
occur frequently and are likely responsible for struc-
turing fish populations therein (Cowx et al., 1984).
Often brook trout (Salvelinus fontinalis) are the only
fish present in these Appalachian headwater systems.
Brook trout are a fall spawning salmonid, with spawn-
ing activity in eastern West Virginia typically begin-
ning in early October and peaking in late October (per-
sonal observation). Redds (nests) are constructed in
gravel substrate possibly in areas of upwelling ground-
water discharge (Curry et al., 1995). Eggs develop
within the gravel over the winter months, hatching in
late winter to early spring.
In 1998, we began a two-year study to examine the
habitat features associated with brook trout in fores-
ted headwater streams in eastern West Virginia. The
objectives of this study were to document habitat and
corresponding brook trout population parameters in
the hopes of identifying key habitat features in this
region. During 1999 a drought occurred presenting
us the opportunity to assess the effects of this natural
perturbation upon habitat quality and availability, and
the resulting effect upon brook trout populations. The
purpose of this paper was to identify changes in stream
morphology and brook trout populations associated
with this drought.
Study area
The study area was located within the east central
portion of the Monongahela National Forest, West Vir-
ginia. We examined habitat and trout populations in
seven, first-and second-order headwater streams loc-
204
ated in the East Fork of the Greenbrier River watershed
and the adjacent, North Fork of the South Branch of
the Potomac River drainage. Stream elevations range
from 900–1150 m above sea level. Study streams were
not stocked and all supported wild brook trout popu-
lations. In one of the seven study streams (Little Low
Place), 19 habitat improvement structures (k-dams, v-
dams) were in place that generally formed deep pools
in their lee.
Drought conditions were experienced over much
of the Eastern United States in 1999. The National
Oceanic and Atmospheric Administration (NOAA) re-
ported the U.S. nationwide had experienced the second
warmest and 22nd driest year on record (1900–1999)
(NOAA, 1999). Drought intensity varied over the na-
tion, with West Virginia experiencing its 12th driest
year in the 100-year data set (NOAA, 1999). Records
for the study region indicated 1999 to be the 5th driest
on record (1899–1999). Annual mean discharge meas-
ured 6 km downstream of the study area (Greenbrier
River at Durbin, WV) was 38% lower than the 58-
year average (Ward et al., 2000). The Palmer Drought
Index (Palmer, 1968) and Standardized Precipitation
Index (McKee et al., 1993) classified the 1999 drought
as ‘moderate’ to ‘severe’ over the region (NOAA,
1999). By all accounts, the study area in 1999 exper-
ienced harsh drought conditions that were reflected in
extremely low flows in all streams regionally.
Materials and methods
We established study reaches within each of the 7
streams to quantify habitat conditions and brook trout
populations. Three 50-m sample reaches were per-
manently established at the lower, middle and upper
portions of each study stream. Sample reach size was
expanded to 100 m (Armour et al., 1983) after the
first of four sample periods to better account for nat-
urally high variation in stream fish densities (Platts
& Nelson, 1988). After expansion to 100 m reaches,
total sample reach length averaged 18% (range 9–
29%) of total study stream length with distances of
300–1000 m between reaches depending upon stream
length.
We conducted a quantitative assessment of phys-
ical instream habitat in each sample reach within
all seven study streams. Habitat assessments were
conducted near baseflow conditions over a one-week
period in late June 1999 and again in June 2000. Hab-
itat data were collected using Dolloff et al. (1997)
habitat-based sample design. Three habitat unit types
(pool, riffle or glide) were identified (Bisson et al.,
1981) and habitat unit dimensions were measured
using standard methodologies (Dolloff et al., 1997;
Overton et al., 1997). Maximum water depth within
pools was measured and pool habitats were rated sub-
jectively for their potential to provide cover for fish
using rating criteria modified from Platts et al. (1983).
All suitable spawning substrate was measured and
quantified as an area by measuring length and width of
the unit (Bozek & Rahel, 1991; MaGee et al., 1996).
Substrate ranging from 4 to 30 mm in diameter, in at
least 7-cm depth of low velocity water (generally tails
of pools) was considered suitable brook trout spawn-
ing substrate based upon our observationsof spawning
pairs in these systems.
Substrate samples were collected from brook trout
spawning areas so inferences about compositional
changes could be related back to impacts on brook
trout reproductive potential. Substrate samples were
collected in fall 1998 and fall 1999 to assess compos-
itional changes during the drought. Samples were col-
lected from known brook trout spawning areas (pres-
ence of redds) using a plastic grain scoop (Hakala,
2000) in a modification of the shovel technique (Grost
et al., 1991). Substrate samples were taken to a depth
of 8.0 cm which is the average maximum depth egg
pockets have been found for brook trout of similar size
(<250 mm) as those found in this study (Young et al.,
1989). An average of six substrate samples (range 5–
8) were collected annually from each study stream.
No spawning substrate samples were collected from
Mullenax Run in either year because of an inabil-
ity to locate actively used spawning areas. This was
likely a function of naturally low brook trout dens-
ities in that stream. The stream was excluded from
further substrate analyses in both years. Sawmill Run
was not used as a study stream until after fall 1998,
thus substrate samples were not collected in 1998.
Substrate samples were collected and processed as
described in Hakala (2000). Sediment sizes analyzed
were >32.0, 16.0, 8.0, 4.0, 2.0, 1.0, 0.5, 0.25, 0.125
and <0.063 mm. Percent composition of each size
class was calculated based upon sample weight and a
randomzed block ANOVA (Friedman’s test) was used
to test the null hypothesis of ‘no difference’ in stream
morphologic characteristics between years. Individual
streams were used as the blocking unit. A significance
level of 0.05 was used for this, and all subsequent
statistical analyses.
205
Stream discharge was measured with cross-
sections using standard USGS techniques (USGS,
1977). Discharge data were collected once a year dur-
ing near summer baseflow conditions. Discharge was
calculated three times at the same location over a
15-min period. All three measures of discharge were
averaged to come up with a measure of total stream
discharge. Average discharge over all streams was
compared between drought and non-drought periods
using a Wilcoxon signed ranks test.
Water temperature was recorded every hour to the
nearest 0.01 ◦C using Onset Corporation HOBO tem-
perature loggers. Loggers were affixed to the stream
bottom in a PVC housing attached by chain to a rebar
anchor. Analysis of variance was used to test for sig-
nificant differences in water temperature between the
drought and non-drought period.
Dissolved oxygen (DO) and percent DO saturation
were measured at least monthly from (July through
September) during the 1999 drought and again in the
following non-drought year. Measures were recorded
at a single pool within each sample reach. Data were
collected during the day when water temperatures
were at their expected peak.
Daily precipitation data were collected from
NOAA weather observatories located in Bartow, WV
and Elkins, WV. The Bartow station was close to all
study streams (<15 km), but lacked a long-term data
series. The Elkins station was approximately 35 km
north of the study streams, but was reflective of pre-
cipitation patterns and provided a long-term data set
for comparison. Total precipitation between January 1
and August 31 was recorded for each year of the study
(1998–2000). Precipitation data were used to establish
that drought conditions existed in 1999, and that nor-
mal precipitation amounts fell in both 1998 and 2000
(Bell et al., 2000). To assess the relative magnitude of
the drought, precipitation data at the Bartow station
was compared to 30-year average precipitation data
collected at the Elkins station.
We conducted fish population surveys in October
1998, July 1999 October 1999, and July 2000 in all
sample reaches. Fish were collected using a back-
pack electrofisher and a standard three-pass removal
technique. Captured brook trout were measured to
the nearest 1.0 mm (total length, TL) and weighed
to the nearest 1.0 g. Brook trout condition was cal-
culated using the formula C=W/L3×1 00 000
(where C=condition; L=total fish length; W=fish
weight) (Anderson & Gutreuter, 1983). We discrimin-
ated between young-of-the-year (YOY) and age 1 and
Table 1. Total precipitation recorded in Bartow
and Elkins, WV between January 1 and August 31
for 1998, 1999 and 2000.
Year Location of observatory
Bartow Elkins Year
(study area) (cm)
(cm)
1998 78.0 82.0
1999 58.6 62.5
2000 74.8 78.9
30-year average n/a 79.2
n/a =No long-term data available.
older (adult) fish through analysis of length-frequency
histograms and verified these determinations using
scale-aging techniques (DeVries & Frie, 1996).
The program CAPTURE (White et al., 1982) was
used to calculate separate maximum-likelihood popu-
lation estimates and 95% CI for YOY and adult brook
trout. If the model assumption of ‘equal catch probab-
ility’ was not satisfied (rarely the case) for a particular
population estimate, then we bootstrapped the raw
total catch using the average catch efficiency for those
estimates satisfying the model’s assumption. Catch ef-
ficiency was calculated as the quotient of the actual
number of fish captured, and the derived population
estimate for a sample reach (Lohr & West, 1992). If
less than 30 fish (YOY or adult) were captured in a
given sample reach then the actual number captured
was used in lieu of an estimate (Riley & Fausch, 1992).
Brook trout length, weight, and condition were
analyzed separately for adults and YOY. Statistical
comparisons were made between drought (summer
1999 and fall 1999) and non-drought sample periods
(fall 1998 and summer 2000). A Non-parametric ran-
domized block (streams) ANOVA (Friedman’s test)
was used to test for differences in brook trout length,
weight and condition between the periods. Linear re-
gression was used to test if pool area influenced mean
brook trout density during the drought and in the
non-drought period.
Results
Precipitation data indicated that 1999 was a drought
year and both 1998 and 2000 were typical years for
precipitation amounts (Table 1). Drought induced low
flows resulted in reduced quantity and quality of hab-
itat for brook trout (Table 2). Average stream wetted
206
width was significantly (p<0.0001) smaller (−32%)
during the drought period (summer 1999) than meas-
ured in the non-drought period (summer 2000). Re-
ductions in wetted stream width were primarily attrib-
uted to a 54% loss of riffle area during the drought
(p<0.0001) as compared to a non-significant reduc-
tion of pool habitat area (p=0.3). Accompanying
decreases in wetted stream area were significant re-
ductions in average riffle and pool depth, and average
maximum pool depth. Riffle habitats were 60% shal-
lower during the 1999 drought period (p<0.0001),
while average pool depths were reduced only 27%
(p<0.0001). Average maximum pool depths were
also significantly lower (p<0.0001) during the
drought (−24%). Moreover, significant declines in
overall pool cover quality between the drought and
non-drought period were measured (p<0.0001).
No significant (p=0.9) difference existed in the
amount of available spawning substrate between the
drought and non-drought periods. However, signific-
ant increases in the percent (from 0.9% to 1.3% by
weight) of fine sediment (<0.063 mm) over all streams
were detected between spawning substrate samples
collected in the year before the drought and during the
drought year (p=0.01).
Water conditions. Discharge during the drought
(1999) was extremely low in all study streams, with
some streams experiencing flow cessation in depos-
itional areas. Discharge over all streams during this
period was 96% lower than measured in the non-
drought year of 2000 (p=0.008). Average discharge
for all streams during the drought was 0.0027 m3s−1
(range =0.0002–0.0041 m3s−1), while considerably
higher stream discharges were recorded in the non-
drought period under normal precipitation regimes
(average =0.0659 m3s−1, range =0.0118–0.2835).
Water temperature averaged over all streams
(15.7 ◦C) was significantly higher during the drought
(June 1–August 31, 1999) than in the non-drought
period (2000) (14.8 ◦C) (p<0.0001), but water tem-
peratures remained suitable for brook trout. The range
in water temperature over all streams was larger dur-
ing the drought (1999) (9.4–22.9 ◦C) than occurred in
the non-drought period of 2000 (9.8–21.2◦C). Highest
maximum water temperatures were recorded in Mul-
lenax Run. Average daytime summer water temperat-
ure at Mullenax Run during the drought was 17.5 ◦C
(range =9.4–22.9 ◦C). Fifteen percent (162 observa-
tions) of the daytime temperature readings exceeded
Table 2. Mean (±95% CI) physical habitat parameters over all
streams for both years. Differences between years are relative
to the non-drought year of 2000 and significant differences are
denoted with an asterisk (∗).
Habitat parameter Sample period
1999 2000 % difference
between years
Average stream wetted 1.7 ±0.1 2.5 ±0.1 −32%∗
width (m) n=317 n=293
Average riffle depth 4.0 ±0.3 10.0 ±0.7 −60%∗
(cm) n=141 n=129
Average pool depth 16.0 ±1.1 22.0 ±1.3 −27%∗
(cm) n=134 n=138
Average maximum pool 25.0 ±1.8 33.0 ±1.9 −24%∗
depth (cm) n=140 n=138
Average pool rating 2.3 ±0.2 3.0 ±0.3 −23%∗
(1–5) n=140 n=138
Pool area 9.0 ±1.4 9.2 ±1.4 −2%
(m2)n=134 n=138
Riffle area 12.8 ±1.7 28.0 ±5.4 −54%∗
(m2)n=141 n=129
Spawning area 9.9 ±1.3 10.4 ±3.5 −5%
(m2)n=317 n=293
20.0 ◦C. No other study stream exceeded 20 ◦Cforany
significant amount of time during the drought.
Dissolved oxygen was suitable for brook trout
survival at all times measured during drought and
non-drought conditions. Dissolved oxygen averaged
over all streams from July to September in 1999 was
7.2 mg l−1(n=40). Minimum DO concentrations
measured in all streams during the drought were above
6.0 mg l−1.
Fish populations. With significant reductionsin hab-
itat quantity, reduced flows and increasing sediment
levels during the drought, resident brook trout popu-
lations experienced negative population effects. In our
summer and fall population surveys, the influence of
the drought on brook trout density was not assessed
until the post-drought period in summer 2000. Density
of adults over all streams was 60% lower in summer
2000 than densities recorded in summer 1999 at the
beginning of the drought (p=0.0003) (Tables 3 and
4). Young-of-the-year density followed a similar trend
as adults, with 67% fewer YOY over all streams in the
post drought period (p<0.0001) (Tables 3 and 4).
During the drought (summer through fall 1999)
brook trout condition declined slightly (−3%; p<
207
0.0001) (Table 5). Average adult length was not sig-
nificantly different (p=0.2), however, body weight
was significantly lower over all streams in the same
comparison (−4%; p=0.006). Comparison of aver-
age adult body condition between the drought (sum-
mer 1999) and non-drought (summer 2000) periods
showed a 10% lower Fulton’s Condition during the
drought (Table 4). This indicates brook trout were in
poorer condition during the drought with little or no
net growth of individuals in the population during the
drought period (Table 5).
Drought impacts on YOY brook trout size and
condition were more pronounced than for adults. Al-
though YOY brook trout entered the summer of each
year in similar condition (Table 5) body condition of
YOY declined 16% (p<0.0001) over the course of
the 1999 drought (summer 1999 – fall 1999) (Table 5).
This decline in condition during the drought may not
be unusual, but we did not sample in summer of 1998
or fall 2000 preventing comparison of body condition
changes of similar times in non-drought years. How-
ever, in summer comparison YOY length and weight
were both significantly (p<0.0001 for both) lower
during the drought (summer 1999) than was recorded
in the non-drought period (summer 2000) (Table 5).
Young-of-the-year length was 11% (7 mm) shorter and
body weight was 31% lower during the drought as
compared to non-drought conditions
Adult brook trout density between streams was as-
sociated with amount of residual pool area during the
drought (R2=0.4872), yet the relationship lacked a
robust p-value (p=0.08). However, the relationship
was again analyzed using fall adult brook trout abund-
ance and pool area measured during the previous sum-
mer. The relationship was stronger (R2=0.6514),
and statistically significant (p=0.02). Young-of-the-
year abundance during the drought showed no such
relationship in summer and fall comparison. Analysis
of brook trout abundance in the year following the
drought showed no relationship between either adult
or YOY abundance and pool area. No other habitat,
or water quality variable measured was significantly
related to brook trout density.
Discussion
This study demonstrated that during the 1999 drought,
habitat quality and quantity were generally reduced for
brook trout inhabiting first and second-order streams.
During the drought the body condition of brook trout
declined and thereafter brook trout densities declined
significantly. Although there are many possible mech-
anisms for this declining abundance we feel that di-
minished food availability during the drought (Kaller,
2001) and the resulting reductions in condition and
potential overwinter mortality may have been respons-
ible for the apparent declines in the population fol-
lowing the drought. Empirical relationships between
adult brook trout density and pool area during the
drought indicate the importanceof this habitat type for
maintaining brook trout populations during drought
periods.
Drought affect on stream morphology. Headwater
stream morphology was altered over all streams as a
result of drought conditions experienced in 1999. Des-
pite extreme flow reductions (−96%), stream wetted
width, average and maximum water depths, and pool
area were reduced no more than 1/3 that measured
under non-drought conditions. In contrast, riffle area
and average riffle depth were 54% and 60% lower, re-
spectively during drought conditions; with riffle area
reduced more than 25 times that of pool area. This
disproportionate decrease in riffle morphology has
been documented elsewhere (Kraft, 1972) and can be
explained by the inherent morphological differences
between pools and riffles.
With proportionally larger reductions in riffle area
during the drought period, the importance of pool hab-
itat as refuge during low flows is exemplified (Kraft,
1972; Binns, 1994; Huntingford et al., 1999; El-
liot, 2000). If such habitat is limited, potential stream
carrying capacity is likewise affected (Elliot, 1993;
Lestelle et al., 1993). We found that during the drought
year, adult brook trout density was significantly re-
lated to pool area in streams while during non-drought
periods this relationship was not significant. This sug-
gests that during drought, pools are critical to the
survival of brook trout in these first- and second-order
headwater streams.
The relationship between brook trout density and
pool area during the drought was age dependent. Un-
like adults, YOY abundance during the drought was
not a function of available pool area, a difference
likely attributed to the inherent depth requirements
of larger fish. Elliot (1987) studying population reg-
ulation of brown trout (Salmo trutta) in two English
streams, noted that drought affected older fish (and
presumably larger fish) more than it did younger in-
dividuals. Our data support Elliott (1987) and Lestelle
208
Table 3. Total brook trout density (fish ha−1) for all seven streams for each of the
four sample periods (range in parentheses).
Age Sample period
Fall 1998 Summer 1999 Fall 1999 Summer 2000
Age (non-drought) (drought) (drought) (non-drought)
YOY 5 214 12 017 8 19713 922
(65–2083) (689–3436) (433–2774) (0–2113)
Adult 21 113 19 692 12 64717 957
(465–4789) (160–6877) (47–3850) (77–3024)
et al. (1993) and indicate that adult carrying capacity is
limited by available pool area during low flow periods.
Flow reductions diminish the quality of existing
fish cover within streams (Curtis, 1959; Kraft, 1972;
Wesche, 1974; Hunt, 1979; Randolph & White, 1984;
Titus & Mosegaard, 1992), as was documented in
this study. Lower pool cover ratings observed dur-
ing drought conditions reflect changes in water ve-
locity, undercut bank availability, and lower water
depths. These data suggest that resident brook trout in
headwater streams suffer reduced habitat quantity and
quality under moderate to severe drought conditions.
Drought affects on brook trout. The influence of
drought upon brook trout abundance was not appar-
ent until following the drought, as population surveys
(early July 1999) were likely conducted before the
full impact of the drought on population density could
be realized. Fish surviving the summer drought likely
had diminished body fat reserves, as is evident by
a measured decline in mean adult body weight over
the drought period. Body fat reserves are crucial to
overwinter survival (Thompson et al., 1991; Hutch-
ings, 1994; Sogard & Olla, 2000). As a result, adult
brook trout may have suffered higher mortality during
the winter months, following the rigors of spawning,
which substantially depletes lipid reserves (Hutchings,
1994). Further, smaller fish utilize lipid reserves at
a higher rate (Miranda & Hubbard, 1994), resulting
in lower overwinter survival (Smith & Griffith, 1994;
Quinn & Peterson, 1996). Meyer & Griffith (1997)
reported that in situ cage experiments, brook trout
overwinter survival was higher for larger (mean TL =
124 mm) fish relative to smaller individuals (mean
TL =101 mm). Hence, individuals suffering slower
drought-induced growth potentially had higher over-
winter mortality as a consequence of lower winter
survival attributed to reduced body condition entering
the winter. (Table 4).
As expected, adult body condition was signific-
antly lower (10%) during the drought than in the
following non-drought year. Our measures of body
condition may not have adequately reflected the dis-
proportionate reduction in body energy comparing
mass and length during the drought. Reductions in
adult body condition were attributed primarily to loss
of body weight. Further, true body condition was prob-
ably lower than we measured because as fish lose
lipids they uptake water (Sogard & Olla, 2000) and
thus, condition as measured by length-weight based
indices, would tend to not relate well to total body
energy levels.
There were significantly fewer YOY brook trout
in summer 2000 (67% fewer) than were captured the
previous summer at the beginning of the drought. This
may be attributed to a number of factors. First, sig-
nificantly fewer adults were present over all streams
in fall 1999, certainly leading to reduced egg depos-
ition. Elliot et al. (1997) reported brown trout egg
production was reduced 73–83% as a result of adult
female loss during drought episodes. In addition to
fewer adults, those present were of lower body con-
dition, perhaps resulting in some adults failing to
reproduce (Wydoski & Cooper, 1965) or producing
fewer gametes. Adams & Huntingford (1997) demon-
strated arctic char (Salvelinus alpinus) egg production
was positively related to lipid reserves. Further, it
is possible eggs produced by poorer condition brook
trout were of lower quality and had lower survival to
emergence success as shown in other species (Laine
& Rajasilta, 1999). Low water levels during the fall
spawning period may have lead to the desiccation of
eggs, and/or aided the growth of fungus (i.e., Sap-
rolegnia spp.) in the redd leading to increased egg
mortality (Argent & Flebbe, 1999). We surmise one or
209
Table 4. Difference in adult and YOY density, body condition, body weight, and body length
between drought and non-drought sample periods. An "ns" indicates no significant difference
(p>0.05). (D =density; K =‘Fulton’ type body condition; W =body weight; L =body
length).
Sample period Adult YOY
DKWLDKWL
(%) (%) (%) (%) (%) (%) (%) (%)
Summer 1999 and 2000 −60 −10 −29 −8−67 ns −31 −11
(drought) and (non-drought)
Fall 1998 and Fall 1999 −58 −2ns ns ns −5−19 −5
(non-drought) & (drought)
more of these factors contributed to the observed re-
duction in post-drought YOY brook trout abundance.
Mechanisms responsible for observed population
changes. With both water temperature (Fry et al.,
1946; MacCrimmon & Campbell, 1969; Meisner,
1990) and dissolved oxygen (Spoor, 1990) remain-
ing suitable for brook trout survival, we propose
drought-induced changes in the physical habitat were
the causal mechanism behind the observed declines
in brook trout condition and abundance. Reduc-
tions in available habitat may have created conditions
whereby spatially-limited food resources caused de-
creased growth and/or survival as demonstrated for
other salmonids (Mason & Chapman, 1965; Chap-
man, 1966; Egglishaw, 1967; Jonsson et al., 1998).
Salmonids are territorial, establishing and defending
territories early in life (Mason & Chapman 1965;
Chapman, 1966; Allen, 1969; LeCren, 1973; Slaney
& Northcote, 1974; Elliot, 1990). With less wet-
ted habitat available there were presumably fewer
available ‘optimal foraging’ locations (Fausch, 1984).
Salmonids not holding a feeding territory tend to ex-
perience reduced growth, survival (Chapman, 1966;
Elliot, 1990) and/or are forced from the area by ter-
ritorial holding individuals (Mason, 1976). Symons
(1971) showed in the absence of a feeding territ-
ory, juvenile Atlantic salmon (Salmo salar)grewat
a rate two-thirds that of fish holding a feeding ter-
ritory. Additionally, droughts have been shown to
reduce benthic macroinvertebrate densities in streams
(Extence, 1981; Iversen et al., 1978). A concurrent
study in our streams also found lower abundance of
benthic macroinvertebrates during the drought than
under typical flow conditions (Kaller, 2001). Lower
wetted habitat during drought would also reduce the
surface area available for interception of falling ter-
restrial insects, potentially reducing availability of the
other primary prey of brook trout in headwater streams
(Sweka & Hartman, 2001). Therefore, brook trout in
the streams studied likely experienced reduced growth
and/or survival during the 1999 drought as a result of
documented reductions in habitat availability.
Compounding decreased territorial space, reduc-
tions in invertebrate drift rates and prey availability
during the drought, presumably aided in the devel-
opment of spatially-restricted food resources. Prey
availability is a function of density (Allen, 1969) and
drift rate. Organisms in the drift are the principal food
source for stream salmonids (Chapman, 1966; Elliot,
1973), thus reduced flow rates (96% decline in dis-
charge) and diminished stream area (−32%) would
tend to diminish prey delivery rates through reduced
drift and reduced capture of falling terrestrial insects.
As riffle areas are the primary source of in-stream food
production for streams (Waters, 1982; Hawkins et al.,
1983), decreases in riffle area measured in this study
(−54%) would have certainly resulted in salmonid
food supply reductions. Kaller (2001), conducting
macroinvertebrate research in the same streams used
in this study, reported significant declines in benthic
invertebrate density over the drought period further
bolstering our conclusions.
Coincident with the drought, was increased sedi-
mentation in brook trout spawning gravel. Others have
reported similar increases in stream siltation during
low flow periods (Kraft, 1972; Wright, 1992). Further,
substrate permeability has been related to salmonid
survival to emergence for species larger than brook
trout (Cooper, 1965; Peterson, 1978, Reiser & White,
1988). Hakala (2000) demonstrated within the same
streams as this study, a significant inverse relationship
between percent sediment <0.063 mm in brook trout
210
Table 5. Mean adult and YOY brook trout length (TL, mm ±95% CI) and weight (g,
±95% CI) and Fulton’s condition for all streams for each sample period (95% CI in
parentheses).
Age Fall 1998 Sample period Summer 2000
(non-drought) Summer 1999 Fall 1999 (non-drought)
(drought) (drought)
Adult
Length 121 ±2 125 ±2 124 ±1 136 ±2
Weight 17.8 ±1.1 19.7 ±0.6 18.9 ±0.6 27.9 ±1.6
Condition 90.29 ±1.54 90.95 ±0.73 88.57 ±5.46 101.03 ±4.66
N 521 1040 648 608
YOY
Length 70 ±258±167±165±1
Weight 3.6 ±0.2 2.2 ±0.1 2.9 ±0.1 3.2 ±0.1
Condition 98.15 ±4.89 110.10 ±2.18 92.85 ±2.11 109.58 ±2.19
N 148 581 397 301
redds and YOY brook trout abundance the following
year. Laboratory studies by Argent & Flebbe (1999)
using coarser sediment classes than in our study also
demonstrated reduced emergence of brook trout under
increasing fine sediment levels. Consequently, the pos-
sibility exists that reduced substrate quality reduced
brook trout egg and alevin survival in the spawning
bed further contributing to the observed decrease in
YOY abundance in the post-drought period.
We conclude that significant reduction in food pro-
ducing area, reductions in macroinvertebrate abund-
ance (Kaller, 2001), low flows resulting in decreased
spawning habitat availability, and possible increases
in fine sediment (<0.063 mm) levels related to the
drought, caused the observed decreases in brook trout
density and overall body condition. As stated pre-
viously, the impact of these factors on brook trout
abundance and body condition did not fully develop
until some time after the fall spawning and winter
months.
Declines in brook trout abundance following the
drought could have been related to factors other than
food limitation and related mortality. For example,
emigration, predation, or angling could account for
declines in brook trout abundance associated with the
drought. However, emigration in four of the seven
streams during the droughtwas prevented by a number
of dewatered riffle habitats in lower reaches. Preda-
tion could not be quantified, but few mammalian signs
were noted near the streams and in ongoing studies (S.
Owen, West Virginia University, pers. commun.) no
sign of fish remains were found in scats of raccoons
(Procyon lotor). Fishing pressure in these streams was
light to non-existent (personal observation) and since
angling is a size-selective activity, it could account for
declines in larger adults, but would not account for the
proportional declines in YOY abundance.
Based on the lack of research indicating down-
stream migration of salmonids in response to reduced
water levels (Kraft, 1972; Elliott & Hurley, 1998;
Huntingford et al., 1999), the observed obstacles to
migration found in our streams, and the lack of evid-
ence for predation or angling to account for changes
in the population, we speculate the observed reduc-
tions in brook trout density were primarily a function
of mortality within the study streams. Bolstering our
contention that reduced food availability during the
drought was responsible for declines in abundance and
condition is the fact that brook trout in this region
are food limited. Sweka & Hartman (2001) found that
even under typical flows, brook trout in the study re-
gion consumed only about 7% of predicted ad libitum
rations. Although some studies have shown rapid re-
colonization by benthic macroinvertebrates following
drought (Miller & Golladay, 1996; Wood & Petts,
1999; Ledger & Hildrew, 2001), these were believed
to occur via drift from upstream reaches. In the first-
and second-order streams studied here, rapid recol-
onization via drift is probably not possible because
typically the areas upstream of our study sites were
intermittent in nature. Several studies have shown that
benthic macroinvertebrate densities declined in up-
211
per stream reaches following droughts (Iversen et al.,
1978; Extence, 1981; Cowx et al., 1984; Morrison,
1990; Flecker & Fefarek, 1994) and macroinvertebrate
densities in our study streams declined significantly
during the drought (Kaller, 2001). In similar headwa-
ter streams in this region brook trout receive >50% of
annual energy intake from terrestrial insects (Sweka
& Hartman, 2001). If terrestrial insects are of similar
importance to brook trout in our streams we would still
expect the drought to have reduced their availability
to brook trout by reducing stream area and effectively
reducing the area available for falling insects to enter
the stream drift. Thus, the low feeding rates coupled
with the likelihood of reduced prey availability sug-
gest that following the drought brook trout were in
poor condition and mortality rates may have been elev-
ated. This population response appeared to affect both
young and old fish and hence was not explainable by
size-dependent ulterior hypotheses such as angling.
Conclusions
Proportionally larger decreases in riffles relative to
pool habitat, and reduced flow velocity potentially
combined to limit macroinvertebrate production and
subsequent availability to trout during the drought.
Consequently, brook trout confined to pool habitats
were likely forced to compete for spatially-restricted
food resources. We believe observed decreases in
brook trout body condition and density during and fol-
lowing the drought were a function of spatially-limited
food resources and not a function of degraded water
quality or emigration from the systems.
This study highlights the importance of pool hab-
itat to brook trout during drought conditions. Fishery
managers should recognize this relationship and per-
haps consider increasing available pool habitats (i.e.,
habitat improvement) in streams strongly influenced
by drought conditions. The ability to institute such
improvements over a large-scale is likely cost prohibi-
tive. However, the use of habitat improvementssuch as
large woody debris additions for increasing pool hab-
itat is one low cost option for selected systems. Redu-
cing harvest of adult fish the year following a drought
and increasing available pool habitat would potentially
expedite recovery of the population through greater
adult/YOY survival and increased egg production.
Such measures would potentially increase the stability
of brook trout populations after such stochastic events.
Acknowledgements
The authors thank P. M. Mazik, R. P. Morgan II, J. T.
Petty and C. A. Dolloff for their comments and tech-
nical review of this research. We thank M. Kaller, J.
Sweka, M. Sipe and R. Cook for their invaluable field
efforts. W. Thayne and J. Sweka provided statistical
support. R. Morgan II and the staff at the Appalachian
Environmental Laboratory donated water sampling
supplies and quantitative analysis of water samples.
This project was funded by the U.S. Forest Service
Monongahela National Forest, the WESTVACO Cor-
poration, and from the McIntire-Stennis Act to West
Virginia University (WVU). All research was conduc-
ted in accordance with WVU Animal Care and Use
Committee approved protocol #9810-02.
References
Adams, C. E., & F. A. Huntingford, 1997. Growth, maturation and
reproductive investment in Artic charr. Journal of Fish Biology
51: 750–759.
Allen, R. K., 1969. Limitations on production in salmonid pop-
ulations in streams. In Northcote, T. G. (ed.), Symposium on
Salmon and Trout in Streams. H.R. MacMillan Lectures in
Fisheries, University of British Columbia, Vancouver, Canada:
3–18.
Anderson, R. O. & S. J. Gutreuter, 1983. Length, weight, and
associated structural indices. In Nielsen, L. A. & D. L. John-
son (eds), Fisheries Techniques. American Fisheries Society,
Bethesda, Maryland: 283–300.
Argent, D. G., & P. A. Flebbe, 1999. Fine sediment effects on brook
trout eggs in laboratory streams. Fisheries Research 39: 253–262.
Armour, C. L., K. P. Burnham & W. S. Platts, 1983. Field methods
and statistical analyses for monitoring small salmonid streams.
U.S. Fish and Wildlife Service FWS/OBS-83/33, 200 pp.
Bell, V. A., J. M. Elliot & R. J. Moore, 2000. Modeling the effects of
drought on the population of brown trout in Black Brows Beck.
Ecological Modeling 127: 141–159.
Binns, N. A., 1994. Long-term responses of trout and macrohabitats
to habitat management in a Wyoming headwater stream. North
American Journal of Fisheries Management 14: 87–98.
Bisson, P.A., J.L.Nielsen, R.A.Palmason&L.E.Grove,
1981. A system of naming habitat types in small streams, with
examples of habitat utilization by salmonids during low stream-
flow. In Armantrout, N. B (ed.), Acquisition and Utilization
of Aquatic Habitat Inventory Information. American Fisheries
Society, Western Division, Bethesda, Maryland: 62–73.
Bozek, M. A. & F. J. Rahel, 1991. Assessing habitat requirements of
young Colorado River cutthroat trout by use of macrohabitat and
microhabitat analyses. Transactions of the American Fisheries
Society 120: 571–581.
Chapman, D. W., 1966. Food and space as regulators of salmonid
populations in streams. American Naturalist 100 (913): 345–355.
Cooper, A. C., 1965. The effect of transported stream sediments on
survival of sockeye and pink salmon eggs and alevin. Interna-
tional Pacific Salmon Fisheries Commission Bulletin 18.
212
Cowx, I. A., W. O Young & J. M. Hellawell, 1984. The influence
of drought on the fish and invertebrate populations in an upland
stream in Wales. Freshwater Biology 14: 165–177.
Curry, R. A., D. L. G. Noakes & G. E. Morgan, 1995. Groundwa-
ter and the incubation and emergence of brook trout. Canadian
Journal of Fisheries and Aquatic Sciences 52: 1741–1749.
Curtis, B., 1959. Changes in a river’s physical characteristics un-
der substantial reductions in flow due to hydroelectric diversion.
California Fish & Game 45: 181–188.
DeVries, D. R. & R. V. Frie, 1996. Determination of age and growth.
In Murphy, B.R. & D. W. Willis (eds), Fisheries Techniques, 2nd
edition. American Fisheries Society, Bethesda, Maryland: 483–
512.
Dolloff, C. A., H. E. Jennings & M. D. Owen, 1997. A com-
parison of basinwide and representative reach habitat survey
techniques in three southern Appalachian watersheds. North
American Journal of Fisheries Management 17: 339–347.
Egglishaw, H. J., 1967. Food, growth and population structure of
salmon. Freshwater Salmon Fisheries Research 38: 1–32.
Elliot, J. M., 1973. The food of brown trout and rainbow trout
(Salmo trutta and Salmo gairdneri) in relation to the abundance
of drifting invertebrates in a mountain stream. Oecologia. 12:
329–347.
Elliot, J. M., 1987. Population regulation in contrasting popula-
tions of trout Salmo trutta in two lake district streams. Journal
of Animal Ecology 56: 83–98.
Elliot, J. M., 1990. Mechanisms responsible for population regula-
tion in young migratory trout: the role of territorial behavior III.
Journal of Animal Ecology 59: 803–818.
Elliot, J. M., 1993. The self-thinning rule applied to juvenile sea-
trout, Salmo trutta. Journal of Animal Ecology 62: 371–379.
Elliot, J. M., 2000. Pools as refugia for brown trout during two
summer droughts: trout responses to thermal and oxygen stress.
Journal of Fish Biology 56: 938–948.
Elliot, J. M. & M. A. Hurley, 1998. Population regulation in adult,
but not juvenile, resident trout in a Lake District stream. Journal
of Animal Ecology 67: 280–286.
Elliot, J. M., M. A. Hurley & J. A. Elliot, 1997. Variable effects
of droughts on the density of a sea-trout Salmo trutta population
over 30 years. Journal of Applied Ecology 34: 1229–1238.
Extence, C. D. A., 1981. Effect of drought on benthic macroin-
vertebrate communities in a lowland river. Hydrobiologia 83:
217–224.
Fausch, K. D., 1984. Profitability stream position for salmonids:
relating specific growth rate to net energy gain. Canadian Journal
of Zoology 62: 441–451.
Flecker, A. S. & B. Feifarek, 1994. Disturbance and the temporal
variability of invertebrate assemblages in two Andean streams.
Freshwater Biology 31: 131–142.
Fry, F. E. J., J. S. Hart & K. F. Walker, 1946. Lethal temperature
relations for a sample of speckled trout, Salvelinus fontinalis.
University of Toronto Studies, Biological Series 54. Publication
of the Ontario Fisheries Research Laboratory 66: 9–35.
Grost, R. T., W. A. Hubert & T. A. Wesche, 1991. Field comparison
of three devices used to sample substrate in small streams. North
American Journal of Fisheries Management 11: 347–351.
Hakala, J. H., 2000. Factors influencing brook trout (Salvelinus
fontinalis) abundance in forested headwater streams with em-
phasis on fine sediment. M.Sc. thesis. West Virginia University,
Morgantown, West Virginia.
Hawkins, C. P., M. L. Murphy, N. H. Anderson, & M. A. Wilzbach,
1983. Density of fish and salamanders in relation to riparian can-
opy and physical habitat in streams of the northwestern United
States. Canadian Journal of Fisheries and Aquatic Sciences 40:
1173–1185.
Hunt, R. L., 1979. Removal of woody streambank vegetation to
improve trout habitat. Technical Bulletin 115. Wisconsin Dept.
Nat. Res., Madison, Wisconsin.
Huntingford, F. A., D. Aird, P. Joiner, K. E. Thorpe, V. A. Braith-
waite & J. D. Armstrong, 1999. How juvenile Atlantic salmon,
Salmo salar L., respond to falling water levels: experiments in an
artificial stream. Fisheries Management and Ecology 6: 357–364.
Hutchings, J. A., 1994. Age- and size-specific costs of reproduction
within populations of brook trout, Salvelinus fontinalis.Oikos
70: 12–20.
Iversen, T. M., P. Wiberg-Larsen, S. Birkholm & F. S. Hansen, 1978.
The effect of partial and total drought on the macroinvertebrate
communities of three small Danish streams. Hydrobiologia 60:
235–242.
Jonsson, N., B. Jonsson & L. P. Hansen, 1998. The relative role
of density-dependent and density-independent survival in the life
cycle of Atlantic salmon, Salmo salar Journal of Animal Ecology
67: 751–762.
Kaller, M. D., 2001. Effects of sediment upon benthic macroinver-
tebrates in forested northern Appalachian streams. M.Sc. thesis,
West Virginia University, Morgantown, West Virginia.
Kraft, M. E., 1972. Effects of controlled flow reduction on a trout
stream. Journal of the Fisheries Research Board of Canada 29:
1405–1411.
Laine, P. & M. Rajasilta, 1999. The hatching success of Baltic
herring eggs and its relation to female condition. Journal of
Experimental Marine Biology and Ecology 237(1): 61–73.
LeCren, E. D., 1973. The population dynamics of young trout in
relation to density and territorial behavior. Rapp. P.V. Reun. Int.
Explor. Mer 164: 241–246.
Ledger, M. E. & A. G. Hildrew, 2001. Recolonization by the benthos
of an acid stream following a drought. Archiv für Hydrobiologie
152: 1–17.
Lestelle, L. C., M. L. Rowse & C. Weller, 1993. Evaluation of nat-
ural stock improvement measures for Hood Canal coho salmon.
Tech. Rep. No. 93–1. Point No Point Treaty Council, Kingston,
Washington.
Lohr, S. C. & J. L. West, 1992. Microhabitat selection by brook and
rainbow trout in a southern Appalachian stream. Transactions of
the American Fisheries Society 121: 729–736.
MacCrimmon, H. R. & J. S. Campbell, 1969. World distribution
of brook trout, Salvelinus fontinalis. Journal of the Fisheries
Research Board of Canada 26: 1699–1725.
MaGee, J. P., T. E. McMahon & R. F. Thurow, 1996. Spatial Vari-
ation in Spawning Habitat of cutthroat trout in a sediment-rich
stream basin. Transactions of the American Fisheries Society
125: 768–779.
Mason, J. C., 1976. Response of underyearling coho salmon to
supplemental feeding in a natural stream. Journal of Wildlife
Management 40 (4): 775–788.
Mason, J. C. & D. W. Chapman, 1965. Significance of early emer-
gence, environmental rearing capacity, and behavioral ecology of
juvenile coho salmon in stream channels. Journal of the Fisheries
Research Board of Canada 22: 173–190.
McKee, T. B., N. J. Doesken & J. Kleist, 1993. The relationship
of drought frequency and duration to time scales. Preprints, 8th
Conference on Applied Climatology, 17–22 January, Anaheim,
California: 179–184.
Meisner, J. D., 1990. Potential loss of thermal habitat for brook
trout, due to climactic warming, in two southern Ontario streams.
Transactions of the American Fisheries Society 119: 282–291.
213
Meyer, K. A. & J. S. Griffith, 1997. First-winter survival of rainbow
trout and brook trout in the Henrys Fork of the Snake River,
Idaho. Canadian Journal of Zoology 75: 59–63.
Miller, A. M. & S. W. Golladay, 1996. Effects of spates and dry-
ing on macroinvertebrate assemblages of an intermittent and a
perennial stream. Journal of the North American Benthological
Society 15: 670–689.
Miranda, L. E. & W. D. Hubbard, 1994. Length-dependent winter
survival and lipid composition of age-0 largemouth bass in Bay
Springs Reservoir, Mississippi. Transactions of the American
Fisheries Society 123: 80–87.
Morrison, R. S., 1990. Recolonization of four small streams in
central Scotland following drought conditions in 1984. Hydro-
biologia 208:261–267.
NOAA, 1999. Climatological data. National Oceanic and Atmo-
spheric Administration National Climate Data Center, Cornell
University, Ithaca, New York.
Overton, C. K., S. P. Wollrab, B. C. Roberts & M. A. Radko,
1997. R1/R4 (Northern/Intermountain Regions) fish and fish
habitat standard inventory procedures handbook. USFS General
Technical Report INT-GTR-346.
Palmer, W. C., 1968. Keeping track of crop moisture conditions,
nationwide: the new Crop Moisture Index. Weatherwise 21: 156–
161.
Peterson, R. H., 1978. Physical characteristics of Atlantic salmon
spawning gravel in some New Brunswick streams. Fisheries and
Marine Service Technical Report 785.
Platts, W. S. & R. L. Nelson, 1988. Fluctuations in trout populations
and their implications for land-use evaluation. North American
Journal of Fisheries Management 8: 333–345.
Platts, W. S., W. F. Megahan & G. W. Minshall, 1983. Methods
for evaluating stream, riparian, and biotic conditions. U.S. Forest
Service General Technical Report INT-138.
Quinn, T. P. & N. P. Peterson, 1996. The influence of habitat
complexity and fish size on overwinter survival and growth of in-
dividually marked juvenile coho salmon (Oncorhynchus kisutch)
in Big Beef Creek, Washington. Canadian Journal of Fisheries
and Aquatic Sciences 53: 1555–1564.
Randolph, C. L. & R. G. White, 1984. Validity of the wetted peri-
meter method for recommending instream flows for salmonids
in small streams. Research project technical completion report,
Montana Water Resource Center, Montana State University,
Bozeman, Montana.
Reiser, D. W & R. G. White, 1988. Effects of two sediment size-
classes on survival of steelhead and chinook salmon eggs. North
American Journal of Fisheries Management 8: 432–437.
Riley, S. C. & K. D. Fausch, 1992. Underestimation of trout pop-
ulation size by maximum likelihood removal estimates in small
streams. North American Journal of Fisheries Management 12:
768–776.
Slaney, P. A. & T. G. Northcote, 1974. Effects of prey abundance on
density and territorial behavior of young rainbow trout (Salmo
gairdneri) in laboratory stream channels. Journal of the Fisheries
Research Board of Canada 31: 1201–1209.
Smith, R. W. & J. S. Griffith, 1994. Survival of rainbow trout during
their first winter in the Henrys Fork of the Snake River, Idaho.
Transactions of the American Fisheries Society 123: 747–756.
Sogard, S. M. & B. L. Olla, 2000. Endurance of simulated winter
conditions by age-0 walleye pollock: effects of body size, water
temperature and energy stores. Journal of Fish Biology 56: 1–21
Spoor, W. A., 1990. Distribution of fingerling brook trout,
Salvelinus fontinalis (Mitchell), in dissolved oxygen concentra-
tion gradients. Journal of Fish Biology 36: 363–373.
Sweka, J. A. and K. J. Hartman, 2001. Fall and winter brook trout
prey selection and daily ration. Proceedings of the Annual Con-
ference of the Southeastern Association of Fish and Wildlife
Agencies 55: 8–22.
Symons, P. E. K., 1971. Behavioral adjustment of population dens-
ity to available food by juvenile Atlantic salmon. Journal of
Animal Ecology 40 (3): 569–587.
Thompson, J. M., E. P. Bergersen, C. A. Carlson & L. R. Kaeding,
1991. Role of size, condition, and lipid content in the over-
winter survival of age-0 Colorado squawfish. Transactions of the
American Fisheries Society 120: 346–353.
Titus, R. G. & H. Mosegaard, 1992. Fluctuating recruitment and
variable life history of migratory brown trout, Salmo trutta L., in
a small unstable stream. Journal of Fish Biology 41: 239–255.
U.S. Geological Survey (USGS), 1977. National Handbook of Re-
commended Methods for Water-Data Acquisition. U.S. Geolo-
gical survey, Washington, D.C.
Ward, S. M., B. C. Taylor & G. R. Crosby, 2000. Water resources
data, West Virginia, water year 1999. U.S. Geologic Survey
Water-Data Report WV-99-1.
Waters, T. F., 1982. Annual production by a stream brook char pop-
ulation and its principal food. Environmental Biology of Fishes
7 (2): 165–170.
Wesche, T. A., 1974. Relationship of discharge reductions to avail-
able trout habitat for recommending suitable streamflows. Water
Resources Research Institute, University of Wyoming. Com-
pletion Report to Office of Water Research and Technology &
Wyoming Game and Fish Commission, Project B-023-WYO.
White, G. C., D. R. Anderson, K. P. Burnham & D. L. Otis, 1982.
Capture-recapture and removal methods for sampling closed
populations. Los Alamos National Laboratory LA-8787-NERP,
Los Alamos, New Mexico.
Wood, P. J. & G. E. Petts, 1999. The influence of drought on chalk
stream macroinvertebrates. Hydrological Processes 13: 387–399.
Wright, J. F., 1992. Spatial and temporal occurrences of inverteb-
rates in a chalk stream, Berkshire, England. Hydrobiologia 248:
11–30.
Wydoski, R. S. & E. L. Cooper, 1965. Maturation and fecundity
of brook trout from infertile streams. Journal of the Fisheries
Research Board of Canada 235 (5): 623–619.
Young, M. K., W. A. Hubert & T. A. Wesche, 1989. Substrate altera-
tion by spawning brook trout in a southeastern Wyoming stream.
Transactions of the American Fisheries Society 118: 379–385.
