Content uploaded by Rodney J. Nakamoto
Author content
All content in this area was uploaded by Rodney J. Nakamoto on Jul 13, 2016
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=utaf20
Download by: [170.144.8.79] Date: 13 July 2016, At: 11:28
Transactions of the American Fisheries Society
ISSN: 0002-8487 (Print) 1548-8659 (Online) Journal homepage: http://www.tandfonline.com/loi/utaf20
Seasonal and Among-Stream Variation in Predator
Encounter Rates for Fish Prey
Bret C. Harvey & Rodney J. Nakamoto
To cite this article: Bret C. Harvey & Rodney J. Nakamoto (2013) Seasonal and Among-Stream
Variation in Predator Encounter Rates for Fish Prey, Transactions of the American Fisheries
Society, 142:3, 621-627, DOI: 10.1080/00028487.2012.760485
To link to this article: http://dx.doi.org/10.1080/00028487.2012.760485
Published online: 28 Mar 2013.
Submit your article to this journal
Article views: 274
View related articles
Citing articles: 6 View citing articles
Transactions of the American Fisheries Society 142:621–627, 2013
American Fisheries Society 2013
ISSN: 0002-8487 print / 1548-8659 online
DOI: 10.1080/00028487.2012.760485
NOTE
Seasonal and Among-Stream Variation in Predator
Encounter Rates for Fish Prey
Bret C. Harvey* and Rodney J. Nakamoto
U.S. Forest Service, Pacific Southwest Research Station, 1700 Bayview Drive,
Arcata, California 95521, USA
Abstract
Recognition that predators have indirect effects on prey popu-
lations that may exceed their direct consumptive effects highlights
the need for a better understanding of spatiotemporal variation
in predator–prey interactions. We used photographic monitoring
of tethered Rainbow Trout Oncorhynchus mykiss and Cutthroat
Tro ut O. clarkii to quantify predator encounter rates for fish in
four streams of northwestern California during winter–spring and
summer. To estimate maximum encounter rates, provide the clear-
est contrast among streams and seasons, and provide an empirical
estimate of a key parameter in an individual-based model of stream
salmonids, we consistently placed fish in shallow microhabitats that
lacked cover. Over 14-d periods, predators captured fish at 66 of
the 88 locations where fish were placed. Eight species of birds (in-
cluding two species of owls) and mammals were documented as
capturing fish. Thirty-six percent of the predator encounters oc-
curred at night. Predator encounter rates varied among streams
and between seasons; the best-fitting model of survival included a
stream ×season interaction. Encounter rates tended to be higher
in larger streams than in smaller streams and higher in winter–
spring than in summer. Conversion of predator encounter rates
from this study to estimates of predation risk by using published
information on capture success yielded values similar to an inde-
pendent estimate of predation risk obtained from calibration of an
individual-based model of the trout population in one of the study
streams. The multiple mechanisms linking predation risk to popu-
lation dynamics argue for additional effort to identify patterns of
spatiotemporal variation in predation risk.
Predators affect prey populations directly by consumption
and indirectly through a variety of nonconsumptive effects, such
as alteration of habitat selection and diel activity patterns. Non-
consumptive effects of predators can have greater effects on prey
demographics than consumptive effects (Preisser et al. 2005),
suggesting that overall predator effects may be more impor-
tant to prey population dynamics than traditional ecological
theory suggests. Fully recognizing the potential significance of
*Corresponding author: bharvey@fs.fed.us
Received July 3, 2012; accepted December 13, 2012
Published online March 28, 2013
predation to prey population dynamics highlights the need for
understanding the magnitude of predation risk and its spatiotem-
poral variation. For stream fishes, high rates of fish consump-
tion by various endothermic predators have been observed (e.g.,
Alexander 1979; Heggenes and Borgstrøm 1988; Dolloff 1993),
along with significant annual variation in the presence–absence
of important predators. A variety of studies have addressed the
influence of local habitat features (e.g., cover, depth, and wa-
ter velocity) on predation risk, while advances in long-term
monitoring of tagged fish have allowed large-scale studies of
survival in general (e.g., Berger and Gresswell 2009; Xu et al.
2010). However, both in general and for purposes of fish popu-
lation modeling (e.g., Railsback et al. 2009), it would be useful
to know more about reach-scale and shorter-term temporal vari-
ation in predation risk.
In this study, we sought to examine spatiotemporal variation
in predator encounter rates for fish occupying four streams in
northwestern California. Our specific objectives included de-
tection of seasonal and diel patterns in predator encounters and
the identification of predators. We also sought to empirically
estimate a parameter in the individual-based stream trout model
of Railsback et al. (2009). This model utilizes a stream reach-
scale parameter that represents the minimal rate of survival of
predation risk from nonaquatic predators. Because this param-
eter cannot be routinely measured and is highly uncertain, it is
commonly adjusted in the model calibration process to match
model results to empirical observations.
STUDY SITES
We made observations in Jacoby and Little Jones creeks,
which both drain forested catchments in northwestern Cali-
fornia. In the study reach (at an elevation of about 250 m),
Jacoby Creek is a second-order stream draining 10–15 km2of
621
Downloaded by [] at 11:28 13 July 2016
622 HARVEY AND NAKAMOTO
second-growth forest and grassland. Red alder Alnus rubra and
bigleaf maple Acer macrophyllum dominate much of the ripar-
ian zone; coast redwood Sequoia sempervirens and Douglas-fir
Pseudotsuga menziesii provide most of the forest cover in the
catchment and also commonly occur in the riparian zone. The
active stream channel in the reach averages about 4 m wide,
with a gradient of 1.5%. Water temperature ranges from 9◦Cto
16◦C in summer and from 4◦Cto12
◦C in winter. Streamflow
in the study reach averages less than 0.05 m3/s in the summer
and approximately 0.75 m3/s in the winter. The Rainbow Trout
Oncorhynchus mykiss is the only fish species in the Jacoby
Creek study reach. The stream also supports semiaquatic ver-
tebrates, including the coastal giant salamander Dicamptodon
tenebrosus, northern red-legged frog Rana aurora, Pacific Coast
aquatic garter snake Thamnophis atratus, and coastal tailed
frog Ascaphus trueii. Both the coastal giant salamander and the
Pacific Coast aquatic garter snake are known to prey on fish.
Little Jones Creek is a third-order tributary of the Middle Fork
Smith River in northwestern California, draining about 27 km2
of steep, forested terrain. The Little Jones Creek reach used in
this study drains 15–20 km2of mostly second-growth forest.
Red alder dominates the riparian vegetation. The active stream
channel is about 8 m wide, and stream gradient in the study
reach averages 1.8%. Water temperature ranges from 11.5◦Cto
16◦C in the summer and from 3◦Cto11
◦C in winter. Streamflow
averages 0.15 m3/s in the summer and 2.5 m3/s in the winter.
We also included two first-order tributaries of Little Jones
Creek in this study. The first (informally named “Big Head
Creek”) enters Little Jones Creek 1.6 km upstream of the con-
fluence of Little Jones Creek and the Middle Fork Smith River
and drains 2.7 km2(<2.0 km2in the study reach). The second
(informally named “Weejak Creek”) enters Little Jones Creek
3.3 km upstream of the Middle Fork Smith River–Little Jones
Creek confluence and drains 1.7 km2. Both tributaries have av-
erage streamflows of less than 0.01 m3/s in the summer and
0.1 m3/s or less in the winter. The Cutthroat Trout O. clarkii
is the only fish species in the Little Jones Creek catchment.
Semi-aquatic vertebrates include the coastal giant salamander,
foothill yellow-legged frog Rana boylii, Pacific Coast aquatic
garter snake, and coastal tailed frog.
METHODS
We initially attempted to quantify predator encounter rates by
using artificial lures. The behavior of potential predators in the
vicinity of the lures was recorded via the photographic methods
described below. In each of three different approaches, we used
artificial lures designed to resemble 150-mm FL Rainbow Trout.
Each lure had articulations just anterior and posterior to the
dorsal fin and had a soft plastic caudal fin, which gave the lure
an apparently natural swimming motion when tethered in water
velocities of 5–10 cm/s. The treble hooks on each lure were
removed and replaced with split shot to position the lure just
below the water’s surface with a horizontal orientation. In the
first approach, single lures were positioned in shallow (<10 cm),
slow-moving water by attachment to a 0.5-m-long monofilament
line secured at the upstream end to a small metal stake driven
into the streambed. In the second approach, we tethered three
lures at each monitoring location; the lures were separately
secured by lines anchored 10 cm apart. Finally, we attempted
to attract predators by constructing an apparatus in which three
lures “responded” to predators by exhibiting short movements.
This device incorporated an infrared motion detector, a battery-
powered servo motor, and a suspended counterweight. Lures
were connected to other parts of the apparatus by at least 3 m of
monofilament line to minimize the influence of the apparatus on
predator behavior. When the mechanism was triggered by the
infrared sensor, the servo motor pulled and released attachment
lines multiple times, moving the lures approximately 15 cm
with each cycle. Over 17–42 d of testing, the three approaches
described above failed to attract predators, although cameras
recorded raccoons Procyon lotor and great blue herons Ardea
herodias in the vicinity of the lures.
The failure of artificial lures to attract predators that we
had observed consuming fish at the study sites (e.g., great blue
herons and belted kingfishers Ceryle alcyon)promptedanin-
vestigation of live-fish tethering methods. Extensive daily be-
havioral observations of tethered live fish revealed that (1) the
tethered fish remained quiescent except when disturbed at close
range; (2) a simple tether arrangement in unobstructed habi-
tat eliminated the risk of entanglement; and (3) tethered fish
remained in good condition after 5–7 d in place.
After establishing the effectiveness of the method, we
monitored tethered fish with remote cameras to assess predator
encounter rates across streams and seasons. Trout used in the
experiment (Rainbow Trout in Jacoby Creek and Cutthroat
Trout in Little Jones Creek and its tributaries) were collected
from the study reaches by electrofishing; we assumed that the
difference in trout species between Jacoby and Little Jones
creeks did not influence the results. Fish averaged 114 mm FL
(SD =17); this size reflected our desire to minimally affect
fish populations while using fish that were large enough to be
relatively vulnerable to avian and mammalian predators. After
receiving anesthesia, fish were tethered via a 30-cm monofil-
ament line to a 340-g lead weight that was partially buried in
the substratum. The line was attached to the fish through the
musculature immediately anterior to the insertion of the dorsal
fin. All fish were observed until they had completely recovered
from the anesthesia. We placed tethered fish in shallow locations
(depth =8–15 cm) with low water velocity (0–5 cm/s) and
gravel or sand substratum to maximize their vulnerability and
to allow approximation of the minimum survival parameter in
the individual-based model of Railsback et al. (2009). We also
anticipated that consistent placement of fish in vulnerable loca-
tions would preclude any interaction between tethering artifacts
and the independent variables of interest; such interactions have
been a concern in some previous studies that have used prey
tethering (Barshaw and Able 1990; Aronson et al. 2001). We
Downloaded by [] at 11:28 13 July 2016
NOTE 623
positioned cameras about 1.5 m from the locations of tethered
fish. Each camera was mounted on a metal stake so that the cam-
era was about 60 cm above the water’s surface. Cameras were
triggered with a passive infrared motion sensor; we set cameras
to record five images approximately 0.75 s apart when triggered.
We set a maximum of six tethered fish in each stream at any
time, with tether locations separated by at least 20 m of stream
length. We visited tether locations at 5–7-d intervals. Fish that
survived over an interval were released. Survival of two fish
over successive intervals at one location was classified as an
observation of “no predation.” In some cases, logistical con-
straints dictated that observations of no predation constituted
less than 10–14 d. If predators preferentially visited locations
of prior success in capturing fish, this could affect our measure-
ments; therefore, each location provided only one observation,
regardless of outcome. We anticipated that the density and distri-
bution of tethered fish would preclude any interactions between
predator density and tethering, which have been an issue in some
smaller-scale, short-term studies (Kneib and Scheele 2000). Ob-
servations for each combination of stream and season spanned
25–34 d. We made winter–spring observations from 20 January
to 16 May 2011 and summer observations from 14 July to 17
August 2011. The cameras recorded the date and time of preda-
tion events (so that survival time could be quantified), and the
photos allowed us to identify predators. We summarized the data
by building survival curves for each combination of stream and
season. We also distinguished daytime versus nighttime preda-
tion events, with daytime defined as extending from 1 h before
sunriseto1haftersunset.
We used the Kaplan–Meier estimator (Therneau and
Grambsch 2010) to construct survival curves for the eight
combinations of stream and season. To explore the influence
of stream and season on survival, we used Cox regression
(Therneau and Grambsch 2010) with stream and season as
dummy variables. The raw data for these analyses were ob-
served survival times, including observations of no predation
over known time spans. We used Akaike’s information criterion
(AIC) to compare five models of survival: (1) a null model (no
covariates); (2) a model with stream as the independent vari-
able; (3) a model with season as the independent variable; (4)
a model that included both stream and season; and (5) the full
model, which included stream, season, and a stream ×season
interaction. Using the Cox regression results for the full model,
we also computed hazard ratios to contrast results by season
and stream size (Jacoby and Little Jones creeks versus the two
tributaries of Little Jones Creek). Laplante-Albert et al. (2010)
provide a more detailed description of the general approach to
survival analysis used here.
RESULTS
The field methods appeared to be generally effective. Parallel
to our preliminary observations, we never observed fish straining
at the end of their tethers except immediately after the tethering
FIGURE 1. Two examples of photo-documented predator encounters for teth-
ered fish: (upper panel) a belted kingfisher capturing a fish during the daytime
and (lower panel) a western screech-owl capturing a fish at night.
procedure or during the release procedure. Photographs pro-
vided evidence that encounter rates with some predators were
not increased by the tethering procedure. Potential predators,
including the great blue heron, American marten Martes amer-
icana, American black bear Ursus americanus, American mink
Neovison vison, raccoon, and North American river otter Lon-
tra canadensis, were recorded by the cameras as passing within
0–2 m of tethered fish, apparently without detecting them. All
surviving fish were released in good condition. All of the pho-
tographed predators were identifiable to species (Figure 1).
In total, we documented 66 predator encounters (i.e., with
prey being removed) and 22 instances of no predation. On six
occasions (three in winter–spring and three in summer), fish
were not recovered from the tether location, but no photographs
of predation events were recorded. These six fish were probably
removed by predators that did not trigger the infrared motion
sensor of the camera; ectothermic predators, such as coastal
Downloaded by [] at 11:28 13 July 2016
624 HARVEY AND NAKAMOTO
giant salamanders or Pacific Coast aquatic garter snakes, may
have been responsible.
Although the study was limited to four streams in two small
catchments, we documented encounters by eight species of
avian and mammalian predators. Birds were responsible for
41 (62%) of the 66 documented predator encounters: belted
kingfisher (18 prey captures), western screech-owl Megascops
kennicottii (11 captures), great blue heron (6 captures), com-
mon merganser Mergus merganser (3 captures), barred owl
Strix varia (2 captures), and red-tailed hawk Buteo jamaicen-
sis (1 capture). Raccoons were responsible for 18 captures, and
North American river otters were responsible for seven captures.
Twenty-four (36%) of the 66 total encounters occurred at night.
Four predator species captured fish at night: barred owls (100%
of captures were at night), western screech-owls (36% at night),
North American river otters (57% at night), and raccoons (78%
at night).
Survival curves by stream and season revealed noteworthy
spatiotemporal variation (Figures 2, 3). Between-season differ-
ences varied among streams; for example, the largest stream
included in the study (Little Jones Creek) exhibited mod-
est differences between seasons (Figure 2), in contrast to the
differences observed for the smallest stream (Weejak Creek;
Figure 3). As this result suggests, the Cox regression model that
included season, stream, and the season ×stream interaction
had the strongest support, as indicated by AIC (Table 1). Al-
though the significant season ×stream interaction demands
caution in interpreting main effects, computation of hazard ra-
tios suggested that the risk of encountering predators was about
2.8 times greater in winter–spring than in summer. Comparison
of the two larger streams with the two smaller streams suggested
that the risk of predator encounter was about 2.9 times greater
in the larger streams.
DISCUSSION
Tethering experiments require careful interpretation (e.g.,
Barbeau and Scheibling 1994; Post et al. 1998). The method
used here probably measures prey detection reasonably well
for several predators (e.g., kingfishers, great blue herons, and
TABLE 1. Comparison of five models of fish survival (based on tethered
Rainbow Trout and Cutthroat Trout) in four small streams of northwestern
California. The difference in Akaike’s information criterion (AIC) indicates
the difference in model fit between the given candidate model and the best-fitting
model (i.e., the model with the lowest AIC value). Akaike weights (w) reflect
the relative likelihoods of the models (Burnham and Anderson 2002).
Model AIC w
Null (no covariates) 26.1 <0.0001
Season 17.4 0.0002
Stream 15.7 0.0004
Season +stream 4.5 0.0947
Season +stream +
(season ×stream)
0 0.9047
FIGURE 2. Kaplan–Meier survival curves by season for fish in Jacoby Creek
(tethered Rainbow Trout) and Little Jones Creek (tethered Cutthroat Trout),
northwestern California, 2011. Symbols indicate photo-documented prey cap-
tures or observations of surviving fish. Shaded symbols indicate nighttime preda-
tion events (screech-owl =western screech-owl; kingfisher =belted kingfisher;
merganser =common merganser).
owls) that almost certainly detected and attacked quiescent
fish. In some cases, tethering may have affected prey detection
and capture; although raccoons were photographed on several
occasions in which they did not detect fish, they were probably
over-represented in our data set because tethering appeared
to enhance their probability of capturing fish. To estimate
predation risk for free-swimming fish based on the observations
presented here, information on predators’ capture success in the
pursuit of free-swimming fish is needed. A variety of previous
observations of capture success indicate that the survival curves
presented here, which reflect predator encounter rates, would
require significant modification to reflect predation risk. For
Downloaded by [] at 11:28 13 July 2016
NOTE 625
FIGURE 3. Kaplan–Meier survival curves by season for fish (tethered Cut-
throat Trout) in two first-order tributaries of Little Jones Creek (Weejak and
Big Head creeks), northwestern California, 2011. Symbols indicate photo-
documented prey captures or observations of surviving fish. Shaded sym-
bols indicate nighttime predation events (screech-owl =western screech-owl;
kingfisher =belted kingfisher).
example, pied kingfishers Ceryle rudis had 19% capture success
when feeding on fish along the shoreline of Lake Malawi
(Johnston 1989). Abbruzze and Ritchison (1997) reported that
six radio-tagged eastern screech-owls Megascops asio were
23% successful in 35 attacks on prey that included birds, insects,
crayfish, small mammals, leeches, and fish. For several species
of wading birds feeding on fish in shallow-water areas that
lacked habitat complexity, the capture success averaged 31%
(Lantz et al. 2010). In addition, common mergansers feeding
on the smolts and fry of Coho Salmon O. kisutch had a capture
success rate of 36%, but they succeeded in subduing and eating
only 18% of the prey they pursued (Wood and Hand 1985).
Prior application of an individual-based model to the Cut-
throat Trout population in Little Jones Creek (Harvey and
Railsback 2009, 2012) provides context for the empirical obser-
vations of predator encounter rates presented here, as the model
includes a parameter that represents the daily survival rate for
fish in the habitat offering the lowest survival (Railsback et al.
2009). In application of the model to the Cutthroat Trout pop-
ulation in Little Jones Creek, calibration using multiple years
of empirical data on age-specific abundance and size yielded an
estimate of 98.7% for the minimum daily survival parameter.
From the current study, the loss of 66 out of 88 fish yields a
predator encounter rate of 75% over 14 d. Applying a capture
success rate of 35%—a conservative estimate according to the
literature reviewed above—to this encounter rate would yield a
mortality rate of 22.5% and therefore a survival rate of 77.5%
over 14 d (this assumes that all of the live fish we recovered
would have survived for a complete observation period). This
rate converts to a daily survival of 97.8%. If we exclude fish that
were captured by raccoons from the number of fish encountered
by predators (i.e., because the probability that raccoons will de-
tect and pursue fish is almost certainly overestimated in this data
set), the same exercise produces a daily survival rate of 98.5%.
Although this exercise necessarily relies on a highly specula-
tive estimate of overall capture success rate from the literature
to convert encounter rates to capture rates, we find encourag-
ing the correspondence between the two distinct approaches to
estimation of predation risk.
Our findings suggest that for fish in the streams we studied,
there is a significant chronic risk from a variety of predators.
Because fish are unlikely to be able to perceive and avoid several
of the predators we observed prior to a prey capture attempt, the
results indicate that predation risk could have persistent effects
on habitat selection by fishes. Such effects are exemplified
by Power’s (1984) observation that predation risk from birds
prevented herbivorous fishes from occupying and feeding in
shallow-water areas within a Panamanian stream, thus leading
to “bathtub rings” of algae in pools. The cost of risk-sensitive
habitat selection may be less severe for drift-feeding fishes,
such as salmonids, in that some stream habitats may offer
both relative safety and superior foraging opportunities. For
example, large elements in stream channels (e.g., boulders and
woody debris) can provide cover and cause local streambed
scour that increases water depth. Both cover and depth can
reduce predation risk (e.g., Harvey and Stewart 1991), while
pool habitat can provide the most favorable feeding conditions
for relatively large fish in small streams (Rosenfeld and Boss
2001). Fish seeking to minimize energy expenditure rather than
to maximize foraging efficiency, as may be the case at cold
temperatures (Cunjak 1996), may commonly encounter habitats
that offer both low predation risk and favorable energetic con-
ditions in the form of microhabitats with low water velocity and
cover that provides concealment. However, these observations
do not preclude an important role for the indirect effects
of predation risk on salmonid population dynamics because
Downloaded by [] at 11:28 13 July 2016
626 HARVEY AND NAKAMOTO
low-risk habitat and foraging opportunities do not consistently
overlap.
Our observations suggested a greater risk of predator en-
counters in winter–spring than in summer for fish in the small
streams we studied. Where this pattern applies, its effect on
predation risk would be compounded by any additional nega-
tive effects of water temperature on fish swimming performance
(Webb 1978) and the associated consequences for the suscepti-
bility of fish to endothermic predators, as suggested by several
authors (e.g., Fraser et al. 1993; Cunjak 1996; Reeves et al.
2010). Predators may shift toward smaller stream channels in
winter–spring, when high streamflows make prey detection and
capture in larger channels more challenging. Another possibility
is that endothermic predators may increase their focus on fish
when the availability of alternative prey declines. For example,
piscivory by owls may increase in winter, when some of their
terrestrial prey are hibernating.
The preponderance of daytime predator–prey encounters we
observed (64% of encounters) corresponds with previous con-
clusions that fish in other lotic systems face lower predation
risk at night. Using information on predator diet, density, and
energetics, Metcalfe et al. (1999) estimated that primarily noc-
turnal predators were responsible for 10.5% of the predation on
juvenile salmon in Scottish rivers. For the present study, the ex-
clusion of raccoons, which seem unlikely to have a high capture
success with free-swimming fish in continuous stream systems,
would lower the percentage of encounters occurring at night
from 36% to 21%. The potential seems great for spatiotempo-
ral variation in predator assemblages to result in variation in
diel risk patterns for stream fish. For example, our observations
suggest that owls can be important nocturnal predators of fish
in some streams, but the density and distribution of piscivorous
owls probably vary dramatically. The ongoing range expansion
of the barred owl in western North America may be causing
changes in the risk environment of stream fishes, amphibians,
and crustaceans. The flexibility in diel behavior exhibited by
salmonid fishes (e.g., Metcalfe et al. 1999; Reeves et al. 2010)
and the potential consequences of predation risk for popula-
tion dynamics (e.g., Railsback and Harvey 2011) suggest that
diel variation in predation risk deserves attention in population
modeling.
These initial observations of predator encounter rates in
streams indicate a lower risk for fish in smaller streams, but
this result may have been strongly influenced by specific fea-
tures of the streams included in our study. For example, the two
tributaries of Little Jones Creek had sharply different patterns of
fish survival in winter. This difference may relate to the extent
of riparian vegetation closely overhanging the stream, which
could be a useful covariate in future studies. In some settings,
an overall pattern of decreasing risk upstream could to some
extent offset detrimental features of upstream habitat, such as
the risk of habitat loss from stream drying and lower food avail-
ability, as indicated by lower growth rates upstream (Harvey
1998).
This study revealed, within a geographically limited area, a
broad array of predators and substantial spatiotemporal variation
in predator encounter rates for stream fish. Broader observations
may identify key predator–prey combinations and geographic
and seasonal variation in predator encounter rates that could
help to explain differences in fish behavior and population dy-
namics. While our goal of informing fish population models led
us to focus on these predator–prey interactions from the per-
spective of vulnerable prey, more information on the behavior
and capabilities of specific piscivores would clearly be useful in
improving our understanding of terrestrial–aquatic linkages.
ACKNOWLEDGMENTS
Megan Arnold, Michael Helmair, and Jason White assisted
with fieldwork. Linda Long, C.J. Ralph, and Bill Zielinski as-
sisted with analysis of photographs to identify predators. The
City of Arcata provided access to the Arcata Community Forest.
Sylvia Mori assisted with statistical analyses. The manuscript
benefited from reviews by Jason Dunham and anonymous
individuals.
REFERENCES
Abbruzzese, C. M., and G. Ritchison. 1997. The hunting behavior of eastern
screech-owls (Otus asio). Pages 21–32 in J. R. Duncan, D. H. Johnson, and
T. H. Nicholls, editors. Biology and conservation of owls of the northern
hemisphere: second international symposium. U.S. Forest Service General
Technical Report NC-190.
Alexander, G. R. 1979. Predators of fish in coldwater streams. Pages 153–170
in H. R. Stroud and H. Clepper, editors. Predator–prey systems in fisheries
management: proceedings of the international symposium on predator-prey
systems in fish communities and their role in fisheries management. Sport
Fishing Institute, Washington, D.C.
Aronson, R. B., K. L. Heck Jr., and J. F. Valentine. 2001. Measuring predation
with tethering experiments. Marine Ecology Progress Series 214:311–312.
Barbeau, M. A., and R. E. Scheibling. 1994. Procedural effects of prey tethering
experiments: predation of juvenile scallops by crabs and sea stars. Marine
Ecology Progress Series 111:305–310.
Barshaw, D. E., and K. W. Able. 1990. Tethering as a technique for assess-
ing predation rates in different habitats: an evaluation using juvenile lobsters
Homarus americanus. U.S. National Marine Fisheries Service Fishery Bul-
letin 88:415–417.
Berger, A. M., and R. E. Gresswell. 2009. Factors influencing Coastal Cutthroat
Trout (Oncorhynchus clarkii clarkii) seasonal survival rates: a spatially con-
tinuous approach within stream networks. Canadian Journal of Fisheries and
Aquatic Sciences 66:613–632.
Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel
inference: a practical information-theoretic approach, 2nd edition. Springer-
Verlag, New York.
Cunjak, R. A. 1996. Winter habitat of selected stream fishes and potential
impacts from land-use activity. Canadian Journal of Fisheries and Aquatic
Sciences 53(Supplement 1):267–282.
Dolloff, C. A. 1993. Predation by river otters (Lutra canadensis) on juvenile
Coho Salmon (Oncorhynchus kisutch) and Dolly Varden (Salvelinus malma)
in southeast Alaska. Canadian Journal of Fisheries and Aquatic Sciences
50:312–315.
Fraser, N. H. C., N. B. Metcalfe, and J. E. Thorpe. 1993. Temperature-dependent
switch between diurnal and nocturnal foraging in salmon. Proceedings of the
Royal Society of London B 252:135–139.
Downloaded by [] at 11:28 13 July 2016
NOTE 627
Harvey, B. C. 1998. Influence of large woody debris on retention, immigration,
and growth of Coastal Cutthroat Trout (Oncorhynchus clarki clarki) in stream
pools. Canadian Journal of Fisheries and Aquatic Sciences 55:1902–1908.
Harvey, B. C., and S. F. Railsback. 2009. Exploring the persistence of stream-
dwelling trout populations under alternative real-world turbidity regimes with
an individual-based model. Transactions of the American Fisheries Society
138:348–360.
Harvey, B. C., and S. F. Railsback. 2012. Effects of passage barriers on demo-
graphics and stability properties of a virtual trout population. River Research
and Applications 28:479–489.
Harvey, B. C., and A. J. Stewart. 1991. Fish size and habitat depth relationships
in headwater streams. Oecologia 87:336–342.
Heggenes, J., and R. Borgstrøm. 1988. Effect of mink, Mustela vison Schreber,
predation on cohorts of juvenile Atlantic Salmon, Salmo salar L., and Brown
Trout, S. trutta L., in three small streams. Journal of Fish Biology 33:885–894.
Johnston, D. W. 1989. Feeding ecology of pied kingfishers on Lake Malawi,
Africa. Biotropica 21:275–277.
Kneib, R. T., and C. E. H. Scheele. 2000. Does tethering of mobile prey measure
relative predation potential? an empirical test using Mummichogs and grass
shrimp. Marine Ecology Progress Series 198:181–190.
Lantz, S. M., D. E. Gawlik, and M. I. Cook. 2010. The effects of water depth
and submerged aquatic vegetation on the selection of foraging habitat and
foraging success of wading birds. Condor 112:460–469.
Laplante-Albert, K. A., M. A. Rodr´
ıguez, and P. Magnan. 2010. Quantifying
habitat-dependent mortality risk in lacustrine fishes by means of tethering
trials and survival analyses. Environmental Biology of Fishes 87:263–273.
Metcalfe, N. B., N. H. C. Fraser, and M. D. Burns. 1999. Food availability and
the nocturnal vs. diurnal foraging trade-off in juvenile salmon. Journal of
Animal Ecology 68:371–381.
Post, J. R., E. A. Parkinson, and N. T. Johnston. 1998. Spatial and temporal
variation in risk to piscivory of age-0 Rainbow Trout: patterns and population
level consequences. Transactions of the American Fisheries Society 127:932–
942.
Power, M. E. 1984. Depth distributions of Armored Catfish: predator-induced
resource avoidance? Ecology 65:523–528.
Preisser, E. L., D. I. Bolnick, and M. F. Benard. 2005. Scared to death? the ef-
fects of intimidation and consumption in predator-prey interactions. Ecology
86:501–509.
Railsback, S. F., and B. C. Harvey. 2011. Importance of fish behaviour in
modelling conservation problems: food limitation as an example. Journal of
Fish Biology 79:1648–1662.
Railsback, S. F., B. C. Harvey, S. K. Jackson, and R. H. Lamberson. 2009.
InSTREAM: the individual-based stream trout research and environmental
assessment model. U.S. Forest Service General Technical Report PSW-GTR-
218.
Reeves, G. H., J. B. Grunbaum, and D. W. Lang. 2010. Seasonal variation in diel
behaviour and habitat use by age 1 +steelhead (Oncorhynchus mykiss)in
Coast and Cascade Range streams in Oregon, U.S.A. Environmental Biology
of Fishes 87:101–111.
Rosenfeld, J. S., and 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.
Therneau, T. M., and P. M. Grambsch. 2010. Modeling survival data: extending
the Cox model. Springer-Verlag, New York.
Webb, P. W. 1978. Temperature effectson acceleration of Rainbow Trout, Salmo
gairdneri. Journal of the Fisheries Research Board of Canada 35:1417–1422.
Wood, C. C., and C. M. Hand. 1985. Food-searching behaviour of the common
merganser (Mergus merganser) I: functional responses to prey and predator
density. Canadian Journal of Zoology 63:1260–1270.
Xu, C. L., B. H. Letcher, and K. H. Nislow. 2010. Size-dependent survival of
Brook Trout Salvelinus fontinalis in summer: effects of water temperature
and stream flow. Journal of Fish Biology 76:2342–2369.
Downloaded by [] at 11:28 13 July 2016
