Hydrobiologia (2009) 623:113–125
DOI 10.1007/s10750-008-9652-1
PRIMARY RESEARCH PAPER
Habitat effects on invertebrate drift in a small trout stream:
implications for prey availability to drift-feeding fish
Elaine S. Leung Æ Jordan S. Rosenfeld Æ
Joanna R. Bernhardt
Received: 5 June 2008 / Revised: 3 October 2008 / Accepted: 8 November 2008 / Published online: 2 December 2008
Ó Springer Science+Business Media B.V. 2008
Abstract In this study, we focused on the drivers of
micro- and mesohabitat variation of drift in a small
trout stream with the goal of understanding the factors
that influence the abundance of prey for drift-feeding
fish. We hypothesized that there would be a positive
relationship between velocity and drift abundance
(biomass concentration, mg/m3) across multiple spatial scales, and compared seasonal variation in
abundance of drifting terrestrial and aquatic invertebrates in habitats that represent the fundamental
constituents of stream channels (pools, glides, runs,
and riffles). We also examined how drift abundance
varied spatially within the water column. We found no
relationship between drift concentration and velocity
at the microhabitat scale within individual pools or
riffles, suggesting that turbulence and short distances
between high- and low-velocity microhabitats minimize changes in drift concentration through settlement
in slower velocity microhabitats. There were also
minimal differences in summer low-flow drift
abundance at the mesohabitat scale, although drift
concentration was highest in riffle habitats. Similarly,
there was no differentiation of drifting invertebrate
community structure among summer samples collected from pools, glides, runs, and riffles. Drift
concentration was significantly higher in winter than
in summer, and variation in drift within individual
mesohabitat types (e.g., pools or riffles) was lower
during winter high flows. As expected, summer
surface samples also had a significantly higher
proportion of terrestrial invertebrates and higher
overall biomass than samples collected from within
the water column. Our results suggest that turbulence
and the short length of different habitat types in small
streams tend to homogenize drift concentration, and
that spatial variation in drift concentrations may be
affected as much by fish predation as by entrainment
rates from the benthos.
Keywords Invertebrate drift Drift-feeding Spatial variation Trout Salmonids Streams
Handling editor: Robert Bailey
E. S. Leung (&) J. R. Bernhardt
British Columbia Conservation Foundation, #206,
17564 56A Avenue, Surrey, BC, Canada V3S 1G3
e-mail: [email protected]
J. S. Rosenfeld
British Columbia Ministry of the Environment,
2202 Main Mall, Vancouver, BC, Canada V6T 1Z4
e-mail: [email protected]
Introduction
Drift, the downstream transport of invertebrates
suspended in the water column (Shearer et al.,
2003), is a defining attribute of stream ecosystems.
Drift affects the recolonization of denuded habitats
following disturbance (Townsend & Hildrew, 1976;
123
114
Sagar, 1983) and strongly influences invertebrate
community structure and composition (Hansen &
Closs, 2007). One of the most important trophic
consequences of variation in drift is to directly affect
the availability of food for drift-feeding fishes
(Elliott, 1967; Sagar & Glova, 1988; Dedual &
Collier, 1995; Hayes et al., 2000), thereby influencing both habitat selection and individual growth
(Hayes et al., 2007). Drift dynamics should also have
a profound effect on fish production and the efficiency of transfer of primary consumer biomass to the
fish trophic level. It is therefore important to understand (1) how drift varies seasonally, since this will
affect seasonal variation in prey availability for fish,
and (2) how drift varies at micro- and mesohabitat
scales within a stream channel, since this will
determine habitat choice and the energetic benefits
that fishes experience in different habitats, as well as
local fish abundance and production.
Average fish biomass and production as well as the
spatial distribution of fish have been found to be
positively correlated with drift abundance (e.g., Elliott, 1973; Wilzbach et al., 1986; Ensign et al., 1990;
Shannon et al., 1996). Predictability of drift distribution within a stream may therefore allow fish to
identify the most profitable feeding locations (Hansen
& Closs, 2007). Since increased prey availability
enhances fish growth (e.g., Rosenfeld et al., 2005) and
survival increases with body size (Holtby et al., 1990),
food availability is an important index of habitat
quality for fish (Nislow et al., 1998). Terrestrial
invertebrates also represent a significant component
of drift, particularly at the stream surface (Sagar &
Glova, 1992) and are an important energy source for
fish (Elliott, 1973; Allan et al. 2003; Romaniszyn
et al., 2007). Managing the production of drift-feeding
fishes (e.g., stream salmonids) should therefore be
informed by an understanding of the habitat factors
that influence the abundance of both terrestrial and
aquatic prey (Romaniszyn et al., 2007).
Research on invertebrate drift has largely focused
on temporal variation in drift, the effects of photoperiod, water temperature, predators, life-cycle stage
(reviewed in Brittain & Eikeland, 1988) and the roles
of disturbance and resource competition in drift
initiation (Statzner et al., 1987; Ramirez & Pringle,
1998; Siler et al., 2001). However, fewer studies have
investigated microhabitat variation in drift within a
stream and the factors that drive this variation. Drift
123
Hydrobiologia (2009) 623:113–125
has been observed to increase with velocity (Brittain
& Eikeland, 1988) since invertebrates are dislodged
as the scouring effect of flow increases with stream
discharge (Ciborowski, 1983; Mackay, 1992), and
once entrained in the water column resettlement rates
of drifting invertebrates are lower in swift water
(Bond et al., 2000). Since velocity and habitat type
may influence drift abundance (Reisen & Prins, 1972)
and thus prey availability for fish, understanding how
these factors drive differences in drift concentration
between habitats is vital to understanding the ecology
and production of species that feed on drift.
In order to better understand the relationship
between habitat structure and variation in drift abundance in small streams, we compared the abundance
of aquatic and terrestrial invertebrate drift across
different habitat types in the summer and winter in
Husdon Creek, a small cutthroat trout stream in
coastal British Columbia, Canada. Since the fundamental constituent habitat types in small streams—
pools, glides, runs, and riffles—differ in depth and
velocity (Peterson & Rabeni, 2001), we anticipated
that drift abundance might also differ between habitat
types. Based on previous research and the assumption
that higher velocities lead to higher entrainment and
lower settling rates of drifting invertebrates, we
predicted (1) that there would be a positive correlation
between drift concentration and water velocity at
multiple spatial scales, (2) that the highest drift
abundances would therefore occur in riffle habitats,
and (3) that drift abundance in the winter would be
higher than in the summer due to increased winter
discharge and water velocity. Given the hierarchical
nature of stream habitats, we also expected that
variation in drift within a single habitat type (e.g., a
pool) would be less than variation between habitats
(e.g., pools vs. riffles). In addition to these predicted
relationships between absolute drift abundance and
habitat factors, we further hypothesized that the
community structure of drifting invertebrates would
respond to habitat (e.g., Cellot, 1996), and (1) that
terrestrial drift, and therefore total drift abundance,
would be higher at the water surface than in the water
column, and (2) that potential differences in benthic
invertebrate community structure between habitats
(e.g., pools vs. riffles) would result in differences in
the taxonomic composition of drifting invertebrates
associated with different habitat types. Finally, to
understand the importance of fish predation relative to
Hydrobiologia (2009) 623:113–125
abiotic drivers of drift abundance, we applied simple
bioenergetic modeling to assess the potential for driftfeeding fishes to reduce drifting invertebrate abundance in Husdon Creek.
Methods
Study site
Drift samples and habitat measurements were collected from a 1 km reach of Husdon Creek, a small
coastal stream on the Sunshine Coast of British
Columbia, 50 km north of the city of Vancouver,
Canada. Husdon Creek has an average bankfull width
of 3.4 m, a summer baseflow of approximately 20–
30 l/s, and drains a 3.4 km2 watershed of second
growth forest. The reach of stream used for experiments had approximately 75% canopy cover, a 1%
gradient, substrate dominated by gravel and sand, and
abundant large woody debris (LWD; 0.37 pieces of
LWD per linear meter) in a forced pool-riffle channel
(Montgomery et al., 1995). Husdon Creek typifies the
small stream habitat where juvenile anadromous
coastal cutthroat trout are most abundant (Rosenfeld
et al., 2000), and the fish community consists of
cutthroat trout in three age classes (young-of-the-year
YOY, 1?, and 2? fish) and YOY coho salmon.
Juvenile cutthroat trout and coho densities are
relatively high, averaging 0.92 and 0.2 fish/m2,
respectively (J. Rosenfeld, unpublished data).
Drift sampling
Variation between habitat types
As part of a broader study to understand the physical
habitat attributes of small coastal streams (Rosenfeld
et al., 2000), and the fitness consequences of different
habitats to juvenile trout (Rosenfeld & Boss, 2001),
we surveyed different habitat types in Husdon Creek
and selected five representative pool, glide, run, and
riffle habitats for drift sampling (total of 20 channel
units). Habitat types were defined as pools
(0% gradient, low current velocity, and deep), glides
(0–1% gradient, slow current velocity, and minimal
water surface turbulence), runs (1–2% gradient, high
current velocity, and turbulent flow), or riffles (1–3%
gradient, high current velocity, water surface broken
115
by protruding substrata, and shallow) as described in
Johnston & Slaney (1996) and Moore et al. (1997).
Drift nets consisted of a 60-cm long bag of 250-lm
Nitex screen attached to a 17.5 cm by 17.5-cm square
collar molded from a PVC pipe. A slightly smaller but
shallower drift net (17 cm by 5 cm) made of the same
materials was fitted into the larger drift net, and could
be adjusted vertically within the larger net to sample
the surface 1–2 cm of the water column, while the
larger net simultaneously sampled subsurface flow.
Water depth and velocity were measured using a
Marsh–McBirney model 2000 flow meter in both the
larger and smaller drift nets when the nets were set and
lifted, in order to calculate the total volume of water
filtered by a drift net over the duration of each set.
In order to ensure that drift samples collected on
the same day were set and lifted within a relatively
narrow time frame, we sampled summer drift over
2 days during summer low-flow discharge (approximately 30 l/s). Ten drift nets were set in separate
channels units between 10:20 and 11:00 on July 21,
2001 and lifted between 15:15 and 16:15, and 10 drift
nets were set on July 22 from 11:30 to 12:20 and
lifted from 16:30 to 17:40.
The durations of sets ranged from 4.9 to 5.3 h on
July 21 and from 4.9 to 5.5 h on July 22. Set and lift
times were chosen to represent the daylight hours
when juvenile cutthroat trout and coho salmon were
observed to forage in Husdon Creek, and to avoid predawn (5:07) and post-dusk (21:19) peaks in drifting
invertebrates that may not be available to drift-feeding
fishes at low light levels (juvenile trout and salmon
were not observed to forage at night; J. Rosenfeld pers.
obs.). Since the number of invertebrates collected in
drift nets is extremely sensitive to disturbance that can
suspend benthic organisms, drift nets were set from
upstream to downstream, and extreme care was taken
to minimize disturbance and avoid stepping in the
stream when setting drift nets. In addition, we placed
250-lm Nitex screen covers over the mouths of drift
nets immediately after they were set, so that we could
initiate sampling by removing covers more or less
simultaneously and avoiding temporal artefacts. Final
contents of drift nets were rinsed into a bucket and
then transferred into a sampling jar using a 250-lm
sieve and preserved in 5% formalin.
We sampled winter drift on December 20, 2001,
when discharge was near bankfull (approximately
440 l/s). We sampled three pools, two glides, two
123
116
runs, and two riffles using a single drift net that
included surface and subsurface drift (i.e., we did not
collect surface and subsurface drift separately). Drift
was sampled between 9:00 and 15:30 and the sets
ranged from 8 to 30 min. Winter drift sets were of
shorter duration than summer samples because of the
higher concentration of suspended organic matter at
high discharge that clogged the drift nets over longer
sets. Although the winter drift sets were short, high
water velocities resulted in nets filtering a volume of
water similar to the longer duration summer sets, so
that short set times should not contribute excessively
to variation in drift abundance (e.g., Culp et al., 1994).
Variation within habitat types
In order to determine the variation in invertebrate
drift abundance within a single channel unit, we
collected 14 drift samples from each of the single
pool and riffle channel units. Four equally spaced
transects were established across each channel unit
and drift nets were set evenly spaced along each
transect. Transects were sampled on four different
days (August 29, 30, 31, and September 1, 2001) to
prevent upstream drift nets from depleting drift
collected in downstream transects; day of sampling
was later used as a covariate in analysis to test for a
day effect on drift abundance (no significant day
effect was detected). Drift nets were set between 9:30
and 11:30, and lifted between 15:30 and 16:30 from a
portable wooden bridge set across the stream channel
to prevent disturbance associated with walking in the
stream. Invertebrate drift samples were preserved as
described above.
Estimation of invertebrate biomass
Invertebrates in drift samples were sorted from
detritus in the lab at 169 magnification and preserved
in 70% ethanol. Most aquatic invertebrates were
identified to genus using Merritt & Cummins (1984),
with the exception of chironomids, which were
identified to subfamily, and terrestrial invertebrates,
which were identified to family. Invertebrate length
was measured to the nearest 0.05 mm using a
binocular microscope equipped with a drawing tube
that projected images of invertebrates onto a digitizing pad (Roff & Hopcroft, 1986), and invertebrate
biomass (mg dry weight) was then estimated using
123
Hydrobiologia (2009) 623:113–125
published length–weight regressions for aquatic
(Meyer, 1989; Benke et al., 1999; Sabo et al., 2002)
and terrestrial (Edwards, 1966; Gowing & Recher,
1984; Sample et al., 1993) invertebrates.
Data analysis
Drift abundance was calculated as both drift biomass
(mg/m3; total biomass of invertebrates collected in a
net divided by water volume filtered) and drift density
(number of invertebrates/m3). Rather than expressing
drift in terms of concentration, many researchers (e.g.,
Bacon et al., 2005; Romero et al., 2005) have
expressed prey availability in the drift as a total flux
(or drift rate) of invertebrates (the product of concentration and discharge, either through the mouth of a
drift net at the microhabitat scale, or through an entire
stream cross section). This unfortunately confounds
the independent effects of drift concentration and
discharge on total prey availability. Total flux and
concentration are both important metrics of prey
availability; total invertebrate flux through a drift net
or stream cross-section will influence capacity (the
total number of fish that can be supported), while
concentration will influence the energy intake and
growth rate that individual fish experience when
exposed to a fixed volume of water. Although we
briefly consider the influence of total flux on capacity,
we present most data in terms of biomass concentration (mg/m3), since biomass can readily be converted
to flux if discharge is known, and the influence of flux
on individual consumption is well described elsewhere (e.g., Hughes & Dill, 1990; Hayes et al., 2007).
(Note that drift biomass, a measure of standing crop,
and drift rate, biomass of drift passing through a
vertical cross-section per unit time, are both distinct
from drift production rate, which is the biomass of
drift entering the water column per m2 of stream bed
per unit time. Drift production is considerably more
difficult to measure in the field than drift biomass
(e.g., see Romaniszyn et al. 2007).)
Although our sampling design was inherently
hierarchical—we sampled variation in drift within
habitats (pool vs. riffle), between habitats (pool, glide,
run, and riffle), and between seasons (summer low flow
vs. winter high flow)—we do not formally analyze it as
a nested analysis, partly because the design is unbalanced. Logistic constraints limited us in measuring the
drift variation within a single pool and riffle, and
Hydrobiologia (2009) 623:113–125
117
sampling was conducted on a single day during winter
high flows. Instead, we use a combination of analyses
at different scales to evaluate our predictions.
We used linear regressions to test for a relationship
between drift abundance and velocity in the summer
at different spatial scales, i.e., between different
habitat types (n = 20), and within habitat types, by
analyzing replicate samples collected along four
transects nested within an individual pool (n = 14)
and riffle (n = 14). We used one-way analyses of
variance (ANOVA) to test for differences in drift
abundance between mesohabitats in the summer
(n = 5 each for pools, glides, runs, and riffles) and
winter (n = 2 per habitat type except n = 3 for
pools). Using a t-test, we also compared drift
abundance in riffles to all other habitat types combined in the summer to increase power by reducing
variation in drift abundance. Similarly, we compared
drift in pools to all other habitat types combined in
the winter. Differences in drift abundance between
surface and subsurface samples as well as the
proportion of terrestrial versus aquatic biomass were
tested using a Wilcoxon two-sample test and t-test,
respectively. We used a t-test to assess seasonal
differences in drift abundance and a Wilcoxon twosample test to examine seasonal differences in the
coefficient of variation for drift biomass. We tested
for greater variance in drift abundance between
different habitat types than within habitat types using
Levene’s test for homogeneity of variance.
We performed correspondence analyses on summer drift samples to (1) assess differences in drifting
invertebrate community structure between habitats
(n = 20 for subsurface samples) and (2) to assess
differences in community structure between surface
and subsurface drift samples (n = 40). We ordinated
untransformed abundance data (numbers/m3) so as to
reflect differences in both taxonomic composition and
absolute abundance of different prey (Jackson, 1993).
All data were analyzed using SAS 9.1 (SAS
Institute Inc., 2002). Data were log10 transformed
where necessary to meet assumptions of normality
and homogeneity of variance.
drift abundance, we used simple bioenergetic modeling and average size and density of fish in pool habitats
in Husdon Creek (from Rosenfeld et al., 2000) to
estimate bioenergetic demand within a typical pool
relative to the flux of invertebrate drift into it. We
assumed a 12-h (daylight) foraging period, a wetted
pool area of 8.3 m2 (3.6 m long by 2.3 m wide, the
average for Husdon Creek), and a combined density of
YOY cutthroat trout and coho salmon in pool habitat
of 1.01 fish/m2, and densities of yearling and older
(year 1? to 2?) cutthroat trout in 80–100 and 110–
190 mm fork length size classes of 0.15 and 0.13 fish/
m2, respectively (Rosenfeld et al., 2000). This translated into eight YOY and 1.3 and 1.1 each of the older
size classes in a representative 8.3 m2 pool. We used
average measured weights in each size class of 1.45,
8.2, and 32.4 g, respectively, to calculate bioenergetic
demand in each size class at 10°C assuming either
25% or 50% of maximum daily consumption (MDC;
calculated using equations from Elliot, 1976) to
bracket a realistic range of consumption rates by
juvenile salmonids. We then estimated YOY trout and
salmon bioenergetic demand as a percent of total
energy flux (drift) into the pool, as well as for all age
classes of fish combined. We assumed energy content
of drifting invertebrates of 21,790 J/g dry weight
(Cummins & Wuycheck, 1971).
We estimated the flux of drifting invertebrates into
the pool using an average measured summer discharge
of 0.022 m3/s over a drift foraging period of 12 h, and
assumed a drift concentration of 0.103 mg/m3 dry
weight (the average observed in pool habitats in
Husdon Creek during this study). Since the drift
concentrations we measured in pools in Husdon Creek
may be partially depleted through fish predation, we
also used the higher average observed drift concentration in riffles (0.37 mg/m3) to provide an upper
bound on maximum potential energy influx into pools.
Estimation of drift consumption by juvenile
salmon relative to drift supply
Average velocities increased from pools, to glides, to
runs, and riffles along a gradient of decreasing habitat
depth; average velocities of drift net sets showed a
similar pattern (Table 1). Velocities were also higher
across all mesohabitat types during the winter high
In order to determine whether consumption by driftfeeding fish has the potential to significantly impact
Results
Relationship between velocity and drift
123
118
Hydrobiologia (2009) 623:113–125
discharge sampling (Table 1). Although we predicted
a positive relationship between drift concentration
and velocity for drift samples collected within
individual pool and riffle habitats, there was no
correlation between drift abundance and velocity at
the microhabitat scale (Fig. 1) within either the pool
(r2 = 0.004, P = 0.84, n = 14) or riffle habitat
(r2 = 0.007, P = 0.78, n = 14). There was a significant positive relationship between drift and velocity
at the mesohabitat scale (i.e., between multiple
channel units; Fig. 2; r2 = 0.31, P = 0.01, n = 19),
but the correlation was relatively weak and became
non-significant (P = 0.09) with the removal of the
highest velocity data point (upper right point in
Fig. 2).
Fig. 1 Relationship between drift biomass (mg/m3) and
velocity (m/s) at the within-habitat scale. Data points represent
drift samples collected across multiple transects in a single
pool or riffle habitat at summer low flow
Habitat and seasonal effects on drift abundance
Despite our expectation of higher drift abundance in
high velocity habitats, there were no significant
differences in drift biomass among pools, glides, runs,
and riffles, in the summer (Table 2; F3,16 = 1.61,
P = 0.23) or winter (F3,5 = 5.09, P = 0.056). However, variability was high, and summer drift biomass in
riffles was significantly higher than all other habitat
types combined (Fig. 3; t = -2.10, P = 0.0498).
Similarly, pools had significantly higher biomass than
all other habitat types in the winter (t = -4.46,
P = 0.0029).
As expected, drift abundance varied spatially
within the water column (Fig. 4). Surface samples
had significantly higher biomass (z = 2.58, P = 0.01)
and a higher proportion of terrestrial invertebrates
(t = -2.54, P = 0.015) than subsurface samples.
Total drift abundance also varied seasonally, with
both higher drift biomass (t = -3.79, P = 0.001)
and a greater proportion of terrestrial invertebrates
Fig. 2 Relationship between drift biomass (mg/m3) and
velocity (m/s) at the mesohabitat scale. Data points represent
drift sampled from a single pool, glide, run or riffle during
summer low flow
(t = -3.60, P = 0.001) during the winter sampling
(Fig. 5). Terrestrial and aquatic invertebrate drift
biomasses were (mean ± SD) 0.05 ± 0.06 and
0.13 ± 0.15, respectively, in the summer and
0.36 ± 0.25 and 0.24 ± 0.17, respectively, during
Table 1 Average velocity (m/s) ± standard deviation at drift net set locations and in different habitat types in the summer and
winter
Habitat type
Pool
Glide
Run
Riffle
Average
Average drift net velocities
Summer
0.11 ± 0.10
0.21 ± 0.09
0.20 ± 0.05
0.35 ± 0.12
0.22 ± 0.10
Winter
0.38 ± 0.19
0.30 ± 0.20
0.46 ± 0.21
0.53 ± 0.03
0.42 ± 0.10
Average habitat velocities
Summer
0.05 ± 0.01
0.12 ± 0.03
0.16 ± 0.03
0.22 ± 0.03
0.14 ± 0.07
Winter
0.25 ± 0.12
0.27 ± 0.11
0.31 ± 0.18
0.45 ± 0.11
0.32 ± 0.09
123
Hydrobiologia (2009) 623:113–125
119
Table 2 Mean drift biomass (mg/m3) ± standard deviation and mean density (number/m3) ± standard deviation in different habitat
types in the summer and winter
Habitat type
Summer
Winter
3
Biomass (mg/m )
3
Density (per m )
Biomass (mg/m3)
Density (per m3)
24.58 ± 7.16
Pool
0.10 ± 0.08
1.51 ± 0.57
1.06 ± 0.46
Glide
0.15 ± 0.15
2.82 ± 1.35
0.32 ± 0.08
8.14 ± 2.87
Run
0.10 ± 0.09
1.73 ± 0.39
0.40 ± 0.18
16.59 ± 7.47
Riffle
0.37 ± 0.35
3.20 ± 1.58
0.39 ± 0.07
12.66 ± 0.30
Fig. 3 Mean drift biomass (mg/m3) in different habitat types
in the summer. Error bars represent one standard deviation
Fig. 5 Average contribution of terrestrial and aquatic drift to
total biomass (mg/m3) in the summer and winter. Error bars
represent one standard deviation for total drift biomass
all habitats (t = 5.68, P = 0.01, for n = 4 habitat
types). The CV between habitat types was also lower
in the winter (CV = 0.70, n = 9) than in the summer
(CV = 1.18, n = 20) indicating greater homogeneity
of drift concentration when stream discharge is high.
Variance in drift abundance within and between
habitat types
Fig. 4 Average contribution of terrestrial and aquatic drift to
total biomass (mg/m3) in surface versus subsurface drift
samples at summer low flow. Error bars represent one standard
deviation for total drift biomass
winter sampling. The coefficient of variation for drift
biomass within habitat types was significantly lower
in the winter (mean CV ± SD = 0.33 ± 0.13) than
in the summer (mean CV ± SD = 0.91 ± 0.08) in
We predicted that variance in summer drift within a
single habitat type would be lower than the variance
between different habitat types. However, the variance
in biomass between different mesohabitats (CV =
1.18, n = 20, representing variation between single
drift samples collected from each of the five pools, five
glides, five runs, and five riffles) was not significantly
different from variance within a single pool (CV =
0.97, n = 14) or riffle (CV = 0.61, n = 14; F2,45 =
1.97, P = 0.15 for a comparison of all variances).
123
120
Hydrobiologia (2009) 623:113–125
These results collectively indicate that variation in
drift concentration within individual channel units
(e.g., individual pools or riffles) is similar to the
variation between different habitat types.
Habitat effects on drift community structure
Given well-documented differences in benthic (e.g.,
Huryn & Wallace, 1987) and drift (e.g., Cellot, 1996)
community between different habitats (e.g., pools vs.
riffles), we anticipated some differentiation of drifting invertebrate community structure between habitat
types. However, correspondence analysis demonstrated broad overlap of drift community structure
for subsurface samples collected in different habitat
types (Fig. 6), and habitats could not be differentiated
by the presence or abundance of particular taxa.
These taxonomic results, along with the observation
of similar variation in drift abundance at multiple
spatial scales, supports the inference that turbulence
and downstream transport processes tend to spatially
homogenize drift in small streams.
In contrast, correspondence analysis demonstrated
distinct differences in community structure between
surface and subsurface drift samples (Fig. 7), with
Fig. 7 Ordination of surface (open diamonds) and subsurface
summer drift samples (filled squares) in taxonomic space.
Dimension 1 is negatively correlated with invertebrates of
terrestrial origin (Collembolans (Collem), spiders (Arach), and
adult flies (Dipt-A)) and positively correlated with common
aquatic taxa (chironomids (Chiro), caddisfly larvae (Trichop),
and Tanypodinae chironomids (Tanypod))
surface drift differentiated by the presence of spiders,
collembolans, and dipteran adults, while subsurface
drift was characterized by a greater abundance of
orthoclad and tanytarsini chironomids.
Bioenergetic demands of juvenile fish relative
to drift flux
Fig. 6 Ordination of summer drift subsurface samples from
pool, glides, runs, and riffles in taxonomic space (taxa labels
represent the location of taxa with the largest scores on
Dimension 1 and Dimension 2; mayfly larvae (Ephem), spiders
(Arach), caddisfly larvae (Trichop), Leptophlebiid mayflies
(Lepto), and Nemourid stoneflies (Nemour)). There is no
apparent taxonomic differentiation of samples from different
habitat types
123
The estimated combined bioenergetic demand of
YOY cutthroat trout and YOY coho salmon in a
typical pool in Husdon Creek (assuming eight 1.45 g
YOY fish per pool) ranged from 45% to 90% of total
energy flux into the pool (Table 3), assuming prey
consumption at 25%–50% of a fish’s physiological
maximum at 10°C, and an invertebrate drift concentration of 0.1 mg/m3 (the average measured in pool
habitat). If drift concentrations are assumed to be the
average measured in riffle habitat (i.e., assuming that
measured concentrations in pools already reflect
depletion by fish consumption), the percent consumption by YOY drops to 13–25% of total drift flux
over a 12-h period. However, if bioenergetic demand
of older fish (year 1? and 2? cutthroat trout at
realistic densities of 0.28 fish/m2) is included, then
bioenergetic demand ranges from 36% to 71% of
Hydrobiologia (2009) 623:113–125
121
Table 3 Estimated flux of invertebrate drift into a typical pool
(g dry weight per 12 h day) in Husdon Creek, and the estimated
bioenergetic demand (g dry weight of invertebrates per 12 h)
by YOY salmonids and all juvenile salmonids combined,
assuming consumption rates at 25% or 50% of Maximum Daily
Consumption (MDC), and average drift concentrations from
pools (0.10 mg/m3) or riffles (0.37 mg/m3)
Biomass flux into
pool (g dry wt.)
YOY bioenergetic
demand (g dry wt.)
All sizes bioenergetic
demand (g dry wt.)
YOY consumption
as % of flux
All sizes consumption
as % of flux
25% MDC
0.098
0.044
0.125
45%
128%
50% MDC
Riffle drift
0.098
0.088
0.25
90%
255%
25% MDC
0.351
0.044
0.125
13%
36%
50% MDC
0.351
0.088
0.25
25%
71%
Pool drift
total drift flux assuming high (riffle) drift concentrations (Table 3).
Discussion
Despite half a century of research on the ecology of
invertebrate drift, our understanding of how habitat
affects drifting invertebrates lags far behind our
knowledge of habitat effects on the fish that consume
them (e.g., Elliot, 1994; Rosenfeld, 2003). A variety
of complex bioenergetic foraging models have been
developed to predict habitat selection and growth of
drift-feeding fishes (e.g., Hughes & Dill, 1990; Van
Winkle et al., 1998; Hayes et al., 2007), but reliably
parameterizing the abundance and dynamics of drifting invertebrate prey remains one of the primary
sources of model uncertainty. Relatively few studies
have investigated how drift varies between discrete
habitat types within a stream, at scales that are
relevant to habitat selection and use by individual fish
(see Nielsen, 1992; Keeley & Grant, 1997; Hansen &
Closs, 2007 for notable exceptions). Given the
importance of invertebrate drift as an energy flux in
stream ecosystems and a food source for drift-feeding
fishes (e.g., salmonids; Elliott, 1967; Sagar & Glova,
1988; Dedual & Collier, 1995; Hayes et al., 2000),
understanding how habitat, flow, and riparian conditions affect drift are vital for the effective
management of stream habitats and the fish that
depend on them (Elliott, 1973; Wilzbach et al., 1986).
Discharge and water velocity are key hydrological
factors associated with increased invertebrate drift
(O’Hop & Wallace, 1983; Brittain & Eikeland,
1988), both in terms of biomass or numbers of
invertebrates/m3, or the total mass flux of suspended
invertebrates carried in the water column. Support for
our expectation of a positive relationship between
velocity and drift abundance (concentration) was
limited, despite previous observations of higher
invertebrate drift in higher velocity habitats (Brittain
& Eikeland, 1988). There was only a weak positive
relationship between drift abundance and velocity at
the mesohabitat scale (Fig. 2), although riffle habitats
had higher drift abundance than lower velocity
habitat types. However, there was no relationship at
all between drift abundance and velocity at the
within-habitat scale (Fig. 1).
Lack of a positive relationship between velocity
and drift at the microhabitat scale suggests that
turbulence and short distances between high- and
low-velocity microhabitats within a channel unit (e.g.,
a single pool) minimize drift settlement in slower
microhabitats. With the exception of higher drift
concentrations in summer riffle habitats, there was
surprisingly little difference in drift abundance
between habitat types (Fig. 2). This indicates, in
conjunction with the lack of taxonomic differentiation
of drift from different habitat types, that turbulence
and downstream transport processes tend to homogenize drift between sequential channel units as well as
within them. This effect may be most pronounced in
small streams like Husdon Creek, where the average
channel unit length (4.0 ± 2.0 m for 44 channel units
surveyed over a 176 m reach) is short relative to
typical invertebrate drift distances (typically 3–10 m
in small streams; e.g., Elliot, 1971). However, spatial
variation in drift may be well differentiated in larger
streams and rivers where the greater length of
different habitats facilitates differentiation. For
123
122
instance, Stark et al. (2002) found consistent habitat
and depth differentiation of drift abundance in a larger
river, and Cellot (1996) also found differences in drift
community structure and abundance between mainstem and side channel habitat.
Our conclusions with respect to the homogenization of stream drift need to be tempered by several
considerations. First, we did not sample microhabitats
within a channel unit repeatedly over time (c.f.
Hansen & Closs, 2007). Therefore, we cannot determine whether the microspatial variation in drift
concentration we observed within channel units is
consistent over time, or the outcome of unpredictable
stochastic variation (e.g., turbulence). It remains
possible that invertebrate drift is locally concentrated
by hydraulics or surface and subsurface eddy lines in
streams, rather than being homogenized by turbulence
as we suggest above. Coarse metrics such as water
column velocity may not adequately represent
hydraulics that are relevant to drift transport but are
easily detected by fish (e.g., Crowder & Diplas, 2002).
The increased drift we observed during higher
velocity winter flow was consistent with our prediction and past studies (e.g., O’Hop & Wallace, 1983).
The expectation of higher invertebrate drift concentration with increasing velocity may be valid when
associated with rising discharge, which initiates the
movement of bed load and suspension of organic
matter and invertebrates (O’Hop & Wallace, 1983;
Gibbins et al., 2007). In contrast, spatial differences
in microhabitat velocity during stable summer low
flows may have limited effects on drift concentration
because benthic disturbance is minimal. However,
differences in microhabitat velocity will have large
effects on the total flux of prey drifting through
different microhabitats and should therefore strongly
influence habitat selection by fish, as discussed later.
The contribution of terrestrial invertebrates to drift
was significant in both summer and winter, as
observed by many others (e.g., Romero et al., 2005;
Romaniszyn et al., 2007), and total drifting invertebrate abundance was consequently higher at the water
surface (see Sagar & Glova, 1992 for similar results).
The proportion of terrestrial invertebrate biomass was
considerably higher in the winter than summer
(Fig. 5), which was surprising since arboreal invertebrate abundance and activity is generally higher
during the summer (Romero et al., 2005). However,
consideration of seasonal variation in drift needs to
123
Hydrobiologia (2009) 623:113–125
be tempered by the source and delivery mechanisms
of invertebrate prey. Our observation of elevated
terrestrial inputs during high winter discharge are
likely the result of rising water levels flooding gravel
bars and stream banks that are colonized by terrestrial
invertebrates at low flow (Angermeier & Karr, 1983),
rather than terrestrial drop of invertebrates from the
riparian canopy (O’Hop & Wallace, 1983). Terrestrial and aquatic drift abundance will also depend on
the timing of drift sampling relative to peak discharge
and the duration of low flows before and high flows
after a storm event. In our study, we sampled winter
drift on only one occasion at high flow during early
winter. In general, drift concentrations are usually
lowest during periods of prolonged high discharge
(Shaw & Minshall, 1980; Sagar & Glova, 1992;
Hansen & Closs, 2007), and winter drift concentrations may very well decrease in coastal streams as
prolonged high flows flush out invertebrates and
benthic organic matter accumulated during low flows
(e.g., O’Hop & Wallace, 1983).
We based our expectation of a positive relationship between drift and velocity on the assumption
that entrainment and suspension of benthic invertebrates are higher in faster water and once entrained
drift should remain in suspension for longer at higher
velocities (Bond et al., 2000). To some extent this
assumption was supported by our observation of
higher drift concentrations in riffle habitats, and the
independent observation by Hansen & Closs (2007)
that drift abundance is positively correlated with riffle
length. However, drift concentration in the water
column will respond to factors that influence both the
exit as well as the entry of invertebrates into the water
column. While velocity may influence the entrainment of invertebrates into the drift, predation by fish
can have a much larger effect on the removal of drift
than settling of drift in low-velocity microhabitats.
The energetic benefit that a drift-feeding fish
experiences in any given microhabitat will be jointly
determined by prey concentration and the volume of
water passing through a fish’s reactive field (Hughes &
Dill, 1990; Hill & Grossman, 1993). Our results
suggest that microhabitat effects on prey concentration vary in a much less predictable way than either
physical constraints on the volume of water that a fish
can scan (imposed by water depth, velocity, and
channel configuration; Hughes & Dill, 1990; Keeley &
Grant, 1997; Grossman et al., 2002), or depletion of
Hydrobiologia (2009) 623:113–125
123
drift by upstream conspecifics. Our simple bioenergetic estimates of consumption indicate that energetic
demand by juvenile salmon and trout in Husdon Creek
pools is sufficient to consume a substantial proportion
of the invertebrate flux into pools, even at higher drift
concentrations (Table 3). This suggests that predation
by fish in pools can act as an effective filter to remove
drifting invertebrates in smaller streams, as suggested
by Waters (1962) and studies demonstrating consumption of a significant proportion of total
invertebrate production by trout (e.g., Huryn, 1996;
Kawaguchi & Nakano, 2001). Accounting for the
effects of consumption by fish (e.g., Hughes, 1992;
Hayes et al., 2007) should therefore greatly improve
our understanding of spatial variation in drift abundance, particularly in small streams where velocities
are relatively low and a significant proportion of the
water column can be searched by drift-feeding fishes.
However, downstream increases in velocity (Rosenfeld et al., 2007) and decreases in densities of juvenile
salmonids along the river continuum (e.g., Rosenfeld
et al., 2000) should result in decreasing efficiency of
transfer of energy from drifting invertebrates to the
fish trophic level. Our drift sampling may also have
underestimated daily drift flux into pools because we
did not include the dawn and dusk spikes in drift
abundance and consumption observed by some
researchers (e.g., Sagar & Glova, 1988; but not others,
e.g., Allan, 1981; Dedual & Collier, 1995).
were also likely influenced through consumption by
juvenile salmonids present at moderate densities
([1 fish/m2), which may tend to reduce measured
drift abundance. Conservative bioenergetic modeling
indicates that juvenile trout at these densities locally
consume a minimum of 13–36% of invertebrate drift
at summer low flow.
Drift biomass or numbers/m3 remain measures of
standing crop that are affected by both the rate of drift
production (recruitment into the water column) and
the rate of drift removal (by settling or fish predation). At present there are few accurate field estimates
of drift production rates or how they differ between
habitats, although some field experiments to directly
measure areal recruitment rates from the benthos are
in progress (K. Shearer, pers. comm.) or completed
(e.g., Romaniszyn et al. 2007). In the short term,
stream ecologists can use drift concentrations measured in different habitats (e.g., pools vs. riffles) as a
surrogate of production rate and the availability of
prey for drift-feeding fish. In the long term, reliable
estimates of drift production and models that incorporate transport and consumption (c.f. Hayes et al.,
2007) are needed to understand the true dynamics of
invertebrate drift abundance.
Conclusion
References
Surprisingly, there was no strong positive relationship
between drift concentration (mg/m3) and velocity in
Husdon Creek, both at the mesohabitat (pool–glide–
run–riffle) scale and the within-habitat scale. This
suggests that in small streams with short channel
units, drift tends to be homogenized by turbulence
and downstream transport processes. This is supported by the lack of taxonomic differentiation of
drift in different habitat types. Our study suggests that
the microspatial cues that fish use to identify the most
profitable areas for feeding on drift are associated
more with maximizing the total flux of drift available
to them, as suggested by many drift-foraging models
(e.g., Hughes & Dill, 1990; Hill & Grossman, 1993),
rather than maximizing drift concentration. However,
the realized drift concentrations that we measured
Allan, J. D., 1981. Determinants of diet of brook trout (Salvelinus fontinalis) in a mountain stream. Canadian Journal
of Fisheries and Aquatic Sciences 38: 184–192.
Allan, J. D., M. S. Wipfli, J. P. Caouette, A. Prussian & J.
Rodgers, 2003. Influence of streamside vegetation on
inputs of terrestrial invertebrates to salmonid food webs.
Canadian Journal of Fisheries and Aquatic Sciences 60:
309–320.
Angermeier, P. L. & J. R. Karr, 1983. Fish communities along
environmental gradients in a system of tropical streams.
Environmental Biology of Fishes 9: 117–135.
Bacon, P. J., W. S. C. Gurney, W. Jones, I. S. McLaren & A. F.
Youngson, 2005. Seasonal growth patterns of wild juvenile fish: partitioning variation among explanatory
variables, based on individual growth trajectories of
Atlantic salmon (Salmo salar) parr. Journal of Animal
Ecology 74: 1–11.
Benke, A., A. D. Huryn, L. A. Smock & J. B. Wallace, 1999.
Length–mass relationships for freshwater macroinvertebrates in North America with particular reference to the
Acknowledgments We would like to thank Leonardo Huato
for assistance in the field and Diane Srivastava for providing
laboratory space. This research was partly funded by the
British Columbia Conservation Corps and Forest Renewal, BC.
123
124
southeastern United States. Journal of the North American
Benthological Society 18: 308–343.
Bond, N. R., G. L. W. Perry & B. J. Downes, 2000. Dispersal
of organisms in a patchy stream environment under different settlement scenarios. Journal of Animal Ecology
69: 608–619.
Brittain, J. E. & T. J. Eikeland, 1988. Invertebrate drift—A
review. Hydrobiologia 166: 77–93.
Cellot, B., 1996. Influence of side-arms on aquatic macroinvertebrate drift in the main channel of a large river.
Freshwater Biology 35: 149–164.
Ciborowski, J. J. H., 1983. Influence of current velocity, density,
and detritus on drift of two mayfly species (Ephemeroptera). Canadian Journal of Zoology 61: 119–125.
Crowder, D. W. & P. Diplas, 2002. Vorticity and circulation:
spatial metrics for evaluating flow complexity in stream
habitats. Canadian Journal of Fisheries and Aquatic Sciences 59: 633–645.
Culp, J. M., G. J. Scrimgeour & C. E. Beers, 1994. The effect
of sample duration on the quantification of stream drift.
Freshwater Biology 31: 165–173.
Cummins, K. W. & J. C. Wuycheck, 1971. Caloric equivalents
for investigations in ecological energetics. Internationale
Vereinigung fur Theoretische und Angewandte Limnologie Verhandlungen 18: 1–158.
Dedual, M. & K. J. Collier, 1995. Aspects of juvenile rainbowtrout (Oncorhynchus mykiss) diet in relation to food-supply during summer in the lower Tongariro River, New
Zealand. New Zealand Journal of Marine and Freshwater
Research 29: 381–391.
Edwards, C. A., 1966. Relationships between weights, volumes, and numbers of animals. In Graff, O. & J. E.
Satchell (eds), Progress in Soil Biology: Proceedings of
the Colloquium on Dynamics of Soil Communities.
North-Holland Publishing Company, Amsterdam: 585–
591.
Elliott, J. M., 1967. Food of trout (Salmo trutta) in a Dartmoor
stream. Journal of Applied Ecology 4: 59–71.
Elliot, J. M., 1971. The distance traveled by drifting invertebrates in a Lake District stream. Oecologia 6: 350–379.
Elliott, J. M., 1973. Food of brown and rainbow-trout
(Salmo trutta and S. gairdneri) in relation to abundance
of drifting invertebrates in a mountain stream. Oecologia
12: 329–347.
Elliot, J. M., 1976. The energetics of feeding, metabolism and
growth of brown trout (Salmo trutta L.) in relation to body
weight, water temperature and ration size. Journal of
Animal Ecology 45: 923–948.
Elliot, J. M., 1994. Quantitative Ecology and The Brown Trout.
Oxford University Press, Oxford.
Ensign, W. E., R. J. Strange & S. E. Moore, 1990. Summer
food limitation reduces brook and rainbow-trout biomass
in a southern Appalachian stream. Transactions of the
American Fisheries Society 119: 894–901.
Gibbins, C., V. Damià, R. J. Batalla & C. M. Gomez, 2007.
Shaking and moving: low rates of sediment transport
trigger mass drift of stream invertebrates. Canadian
Journal of Fisheries and Aquatic Sciences 64: 1–5.
Gowing, G. & H. F. Recher, 1984. Length–weight relationships
for invertebrates from forests in south-eastern New South
Wales. Australian Journal of Ecology 9: 5–8.
123
Hydrobiologia (2009) 623:113–125
Grossman, G., P. A. Rincon, M. D. Farr & R. E. Ratajczak,
2002. A new optimal foraging model predicts habitat use
by drift-feeding stream minnows. Ecology of Freshwater
Fish 11: 2–10.
Hansen, E. A. & G. P. Closs, 2007. Temporal consistency in
the long-term spatial distribution of macroinvertebrate
drift along a stream reach. Hydrobiologia 575: 361–371.
Hayes, J. W., J. D. Stark & K. A. Shearer, 2000. Development
and test of a whole-lifetime foraging and bioenergetics
growth model for drift-feeding brown trout. Transactions
of the American Fisheries Society 129: 315–332.
Hayes, J. W., N. F. Hughes & L. H. Kelly, 2007. Process-based
modelling of invertebrate drift transport, net energy intake
and reach carrying capacity for drift-feeding salmonids.
Ecological Modelling 207: 171–188.
Hill, J. & G. Grossman, 1993. An energetic model of microhabitat use for rainbow trout and rosyside dace. Ecology
74: 685–698.
Holtby, L. B., B. C. Andersen & R. K. Kadowaki, 1990.
Importance of smolt size and early ocean growth to
interannual variability in marine survival of coho salmon
(Oncorhynchus kisutch). Canadian Journal of Fisheries
and Aquatic Sciences 47: 2181–2194.
Hughes, N. F., 1992. Selection of positions by drift-feeding
salmonids in dominance hierarchies: model and test for
Arctic grayling (Thymallus arcticus) in subarctic mountain streams, interior Alaska. Canadian Journal of
Fisheries and Aquatic Sciences 49: 1999–2008.
Hughes, N. F. & L. M. Dill, 1990. Position choice by driftfeeding salmonids: model and test for Arctic grayling
(Thymallus arcticus) in subarctic mountain streams, interior Alaska. Canadian Journal of Fisheries and Aquatic
Sciences 47: 2039–2048.
Huryn, A. D., 1996. An appraisal of the Allen paradox in a
New Zealand trout stream. Limnology and Oceanography
41: 243–252.
Huryn, A. D. & J. B. Wallace, 1987. Local geomorphology as a
determinant of macrofaunal production in a mountain
stream. Ecology 68: 1932–1942.
Jackson, D. A., 1993. Multivariate analysis of benthic invertebrate communities: the implication of choosing
particular data standardizations, measures of association,
and ordination methods. Hydrobiologia 268: 9–26.
Johnston, N. T. & P. A. Slaney, 1996. Fish habitat assessment
procedure. Watershed Restoration Technical Circular 8.
Kawaguchi, Y. & S. Nakano, 2001. Contribution of terrestrial
invertebrates to the annual resource budget for salmonids
in forest and grassland reaches of a headwater stream.
Freshwater Biology 46: 303–316.
Keeley, E. R. & J. W. A. Grant, 1997. Allometry of diet
selectivity in juvenile Atlantic salmon (Salmo salar).
Canadian Journal of Fisheries and Aquatic Sciences 54:
1894–1902.
Mackay, R. J., 1992. Colonization by lotic macroinvertebrates—A review of processes and patterns. Canadian
Journal of Fisheries and Aquatic Sciences 49: 617–628.
Merritt, R. W. & K. W. Cummins, 1984. An Introduction to the
Aquatic Insects of North America. Kendall/Hunt, Dubuque.
Meyer, E., 1989. The relationship between body length
parameters and dry mass in running water invertebrates.
Hydrobiologia 117: 191–203.
Hydrobiologia (2009) 623:113–125
Montgomery, D. R., J. M. Buffington, R. Smith, K. M. Schmidt
& G. Pess, 1995. Pool spacing in forest channels. Water
Resources Research 31: 1097–1105.
Moore, K., K. Jones & J. Dambacher, 1997. Methods for
Stream Habitat Surveys. Oregon Department of Fish and
Wildlife, Corvalis.
Nielsen, J. L., 1992. Microhabitat-specific foraging behaviour,
diet, and growth of juvenile Coho salmon. Transactions of
the American Fisheries Society 121: 617–634.
Nislow, K. H., C. Folt & M. Seande, 1998. Food and foraging
behavior in relation to microhabitat use and survival of
age-0 Atlantic salmon. Canadian Journal of Fisheries and
Aquatic Sciences 55: 116–127.
O’Hop, J. & J. B. Wallace, 1983. Invertebrate drift, discharge,
and sediment relations in a southern Appalachian headwater stream. Hydrobiologia 98: 71–84.
Peterson, J. T. & C. F. Rabeni, 2001. Evaluating the physical
characteristics of channel units in an Ozark stream.
Transactions of the American Fisheries Society 130:
898–910.
Ramirez, A. & C. M. Pringle, 1998. Invertebrate drift and
benthic community dynamics in a lowland neotropical
stream, Costa Rica. Hydrobiologia 386: 19–26.
Reisen, W. K. & R. Prins, 1972. Some ecological relationships
of invertebrate drift in Praters Creek, Pickens County,
South Carolina. Ecology 53: 876–884.
Roff, J. C. & R. R. Hopcroft, 1986. High precision microcomputer based measuring system for ecological research.
Canadian Journal of Fisheries and Aquatic Sciences 43:
2044–2048.
Romaniszyn, E. D., J. J. Hutchens & J. B. Wallace, 2007.
Aquatic and terrestrial invertebrate drift in southern
Appalachian mountain streams: implications for trout
food resources. Freshwater Biology 52: 1–11.
Romero, N., R. E. Gresswell & J. L. Li, 2005. Changing patterns in coastal cutthroat trout (Oncorhynchus clarki
clarki) diet and prey in a gradient of deciduous canopies.
Canadian Journal of Fisheries and Aquatic Sciences 62:
1797–1807.
Rosenfeld, J. S., 2003. Assessing the habitat requirements of
stream fishes: an overview and evaluation of different
approaches. Transactions of the American Fisheries
Society 132: 953–968.
Rosenfeld, J. S. & S. Boss, 2001. Fitness consequences of
habitat use for juvenile cutthroat trout: energetic costs and
benefits in pools and riffles. Canadian Journal of Fisheries
and Aquatic Sciences 58: 585–593.
Rosenfeld, J. S., M. Porter & E. A. Parkinson, 2000. Habitat
factors affecting the abundance and distribution of juvenile cutthroat trout and coho salmon. Canadian Journal of
Fisheries and Aquatic Sciences 57: 766–774.
Rosenfeld, J. S., T. Leiter, G. Lindner & L. Rothman, 2005.
Food abundance alters habitat selection, growth, and
habitat suitability curves for juvenile coho salmon.
Canadian Journal of Fisheries and Aquatic Sciences 62:
1691–1701.
Rosenfeld, J. S., J. R. Post, G. Robins & T. Hatfield, 2007.
Hydraulic geometry as a physical template for the River
Continuum: application to optimal flows and longitudinal
125
trends in fish habitat. Canadian Journal of Fisheries and
Aquatic Sciences 64: 755–767.
Sabo, J. L., J. L. Bastow & M. E. Power, 2002. Length-mass
relationships for adult aquatic and terrestrial invertebrates
in a California watershed. Journal of the North American
Benthological Society 21: 336–343.
Sagar, P. M., 1983. Invertebrate recolonization of previously
dry channels in the Rakaia River. New Zealand Journal of
Marine and Freshwater Research 17: 377–386.
Sagar, P. M. & G. J. Glova, 1988. Diel feeding periodicity,
daily ration and prey selection of a riverine population of
juvenile chinook salmon, Oncorhynchus tshawytscha
(walbaum). Journal of Fish Biology 33: 643–653.
Sagar, P. M. & G. J. Glova, 1992. Invertebrate drift in a large,
braided New Zealand river. Freshwater Biology 27:
405–416.
Sample, B. E., R. J. Cooper, R. D. Greer & R. C. Whitmore,
1993. Estimation of insect biomass by length and width.
American Midland Naturalist 129: 241–247.
SAS Institute, 2002. Statistical Analysis System, Version 9.1.
SAS Institute, Carey, NC.
Shannon, J. P., D. W. Blinn, P. L. Benenati & K. P. Wilson,
1996. Organic drift in a regulated desert river. Canadian
Journal of Fisheries and Aquatic Sciences 53: 1360–1369.
Shaw, D. W. & G. W. Minshall, 1980. Colonization of an
introduced substrate by stream macroinvertebrates. Oikos
34: 259–271.
Shearer, K. A., J. D. Stark, J. W. Hayes & R. G. Young, 2003.
Relationships between drifting and benthic invertebrates
in three New Zealand rivers: implications for drift-feeding
fish. New Zealand Journal of Marine and Freshwater
Research 37: 809–820.
Siler, E. R., J. B. Wallace & S. L. Eggert, 2001. Long-term
effects of resource limitation on stream invertebrate drift.
Canadian Journal of Fisheries and Aquatic Sciences 58:
1624–1637.
Stark, J. D., K. A. Shearer & J. W. Hayes, 2002. Are aquatic
invertebrate drift densities uniform? Implications for salmonid foraging models. Verhandlungen Internationale
Vereinigung für theoretische und angewandte Limnologie
28: 988–991.
Statzner, B., J. M. Elouard & C. Dejoux, 1987. Field experiments on the relationship between drift and benthic
densities of aquatic insects in tropical streams (Ivory
Coast) III. Trichoptera. Freshwater Biology 17: 391–404.
Townsend, C. R. & A. G. Hildrew, 1976. Field experiments on
drifting, colonization and continuous redistribution of
stream benthos. Journal of Animal Ecology 45: 759–772.
Van Winkle, W., H. I. Jager, S. F. Railsback, B. D. Holcomb,
T. K. Studley & J. E. Baldrige, 1998. Individual-based
model of sympatric populations of brown and rainbow
trout for instream flow assessment: model description and
calibration. Ecological Modelling 110: 175–207.
Waters, T. F., 1962. A method to estimate the production rate
of stream bottom invertebrates. Transactions of the
American Fisheries Society 91: 243–250.
Wilzbach, M. A., K. W. Cummins & J. D. Hall, 1986. Influence
of habitat manipulations on interactions between cutthroat
trout and invertebrate drift. Ecology 67: 898–911.
123