Upstream fishway performance by Atlantic salmon (Salmo salar) and brown trout (Salmo trutta) spawners at complex hydropower dams — is prior experience a success criterion?

Abstract

Passage of hydropower plants by upstream-migrating salmonid spawners is associated with reduced migration success, and the need for knowledge of fish behavior downstream of dams is widely recognized. In this study, we examined fishway passage of landlocked Atlantic salmon (Salmo salar) in River Klarälven, Sweden, and brown trout (Salmo trutta) in River Gudbrandslågen, Norway, and the influence of prior experience on passage success in 2012 and 2013. Fishway trap efficiency varied from 18% to 88% and was influenced by river discharge. Most salmon (81%) entered the fishway trap on days without spill, and salmon moved from the turbine area to the spill zone when there was spill, with small individuals showing a stronger reaction than large fish. Analysis of fish with and without prior trap experience showed that a higher percentage of the “naïve” fish (70% of salmon and 43% of the trout) entered the fishway traps than the “experienced” ones (25% of the salmon and 15% of the trout). Delays for fish that entered the trap ranged from 3 to 70 days for salmon and 2 to 47 days for trout.

Full text

ARTICLE
Upstream fishway performance by Atlantic salmon (Salmo salar)
and brown trout (Salmo trutta) spawners at complex hydropower
dams —is prior experience a success criterion?
1
Anna Hagelin, Jon Museth, Larry Greenberg, Morten Kraabøl, Olle Calles, and Eva Bergman
Abstract: Passage of hydropower plants by upstream-migrating salmonid spawners is associated with reduced migration
success, and the need for knowledge of fish behavior downstream of dams is widely recognized. In this study, we examined
fishway passage of landlocked Atlantic salmon (Salmo salar) in River Klarälven, Sweden, and brown trout (Salmo trutta)in
River Gudbrandslågen, Norway, and the influence of prior experience on passage success in 2012 and 2013. Fishway trap effi-
ciency varied from 18% to 88% and was influenced by river discharge. Most salmon (81%) entered the fishway trap on days
without spill, and salmon moved from the turbine area to the spill zone when there was spill, with small individuals show-
ing a stronger reaction than large fish. Analysis of fish with and without prior trap experience showed that a higher per-
centage of the “naïve”fish (70% of salmon and 43% of the trout) entered the fishway traps than the “experienced”ones (25%
of the salmon and 15% of the trout). Delays for fish that entered the trap ranged from 3 to 70 days for salmon and 2 to
47 days for trout.
Résumé : Le passage à travers des centrales hydroélectriques par des salmonidés en montaison de frai est associé à une
réduction du succès de migration, et il est largement reconnu que des connaissances sur le comportement des poissons en
aval de barrages sont nécessaires. Nous examinons le passage dans des passes migratoires de ouananiches (Salmo salar)dans
le fleuve Klarälven (Suède) et de truites brunes (Salmo trutta)danslefleuve Gudbrandsdalslågen (Norvège), et l’influence de
l’expérience antérieure sur le succès de passage en 2012 et 2013. L’efficacité de piégeage des passes migratoires variait de
18 % à 88 % et était influencée par le débit du fleuve. La plupart des saumons (81 %) entraient dans le piège de la passe durant
des jours sans déversement et les saumons se déplaçaient de la zone de turbinage à la zone de déversement quand il y avait
un déversement, les petits spécimens présentant une réaction plus forte que les gros poissons. L’analyse des poissons avec
et sans expérience de pièges antérieure montre qu’un plus grand pourcentage de poissons « naïfs » (70 % des saumons et
43 % des truites) que de poissons « expérimentés » (25 % des saumons et 15 % des truites) entrait dans les pièges des passes.
Les retards des poissons qui entraient dans un piège allaient de 3 à 70 jours pour les saumons et de 2 à 47 jours pour les
truites. [Traduit par la Rédaction]
Introduction
Habitat fragmentation is a major threat to biodiversity world-
wide and an important topic in conservation biology (Nilsson
et al. 2005;Noss and Daly 2006). Hydroelectric development, in
which dams block, partly or completely, upstream migration of
fishes such as salmonids, is one common cause of habitat frag-
mentation (Nilsson et al. 2005;Clay and Eng 2017). As a result,
fishways have often been implemented to restore connectivity
in these fragmented rivers, with varying degrees of success
(Mallen-Cooper and Brand 2007). The different kinds of fishways
vary in their physical characteristics, and these differences may
not only facilitate passage of certain species of fish over others,
they may also favour passage of a subset of individuals within a
species due to individual differences in behaviour and (or)
physiological status of the fish (Hinch and Bratty 2000;Pon et al.
2009). Moreover, it is not only the physical characteristics of the
actual fishway that affect successful dam passage, but also the
environment characterizing the headwater and tailrace above
and below the hydropower station and dam, which often varies
over time, so that even well-designed fishways may function
poorly in certain situations, resulting in delayed or disrupted
dam passage (Larinier 2001). This may be especially true during
high flow conditions due to turbulent “white water”in the tail-
race section, inadequate attraction flows at fishway entrances
relative to turbine outflows and spillwater release, as well as the
placement of fishways at hydraulically unsuitable locations
(Larinier 2001;Larinier et al. 2002). Successful passage of dams
might therefore be lower than expected, and fishways may not
Received 14 August 2019. Accepted 3 August 2020.
A. Hagelin,* L. Greenberg, O. Calles, and E. Bergman. River Ecology and Management Research Group RivEM, Department of Environmental and Life
Sciences, Karlstad University, S-651 88 Karlstad, Sweden.
J. Museth and M. Kraabøl.
†
Norwegian Institute for Nature Research (NINA), NO-2624 Lillehammer, Norway.
Corresponding author: Anna Hagelin (email: anna.hagelin@lansstyrelsen.se).
*Present address: County administrative Board of Västra Götaland, Länsstyrelsen Västra Götalands län, S-403 40 Gothenburg, Sweden.
†
Present address: Multiconsult Norway AS, P.O. 0265 Skøyen, N-0213 Oslo, Norway.
1
This article is being published as part of the special issue “Conservation, Ecology, and Evolution of Nonanadromous Atlantic Salmon”, arising from an
international symposium on landlocked Atlantic salmon held 17–20 June 2018, at the Ecology and Evolutionary Ethology of Fishes Conference in
Montréal, Quebec.
Copyright remains with the author(s) or their institution(s). Permission for reuse (free in most cases) can be obtained from copyright.com.
Can. J. Fish. Aquat. Sci. 00: 1–11 (0000) dx.doi.org/10.1139/cjfas-2019-0271 Published at www.nrcresearchpress.com/cjfas on 4 September 2020.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
1
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.

mitigate the fragmentation of the river caused by hydroelectric
dams and power plants (Roscoe and Hinch 2010).
Anadromous salmon and brown trout (Salmo trutta) use consid-
erable amounts of energy during migration, and it is important
that they use energy efficiently, minimizing swimming costs
when possible (Bernatchez and Dodson 1987). Fishways with low
functionality may have severe consequences for both individual
fitness and population resilience. Studies in the Fraser River, Brit-
ish Columbia, Canada, showed that sockeye salmon (Oncorhyncus
nerka) that failed to pass a fishway had high levels of physiologi-
cal stress, made repeated attempts to locate and enter a fishway,
and exhibited long periods of intense swimming activity, often
swimming in unfavourable areas with high water velocities and
turbulence (Hinch and Bratty 2000;Cooke et al. 2006;Young
et al. 2006). The successful fish, on the other hand, often entered
the fishway on their first attempt, spent little time searching for
the fishway entrance, and selected routes through areas where
water velocity was low and thus less energy-consuming. These
results for sockeye salmon suggest that failure to pass fishways
may be due to a combination of poor route choices, elevated
stress, and physical exhaustion. Thus, demanding migratory
routes, combined with challenging fish passage solutions, may
not only cause exhaustion and stress, but could ultimately also
act as bottlenecks for survival and passage success of migrating
fish (Stevens and Black 1966;Peake and Farrell 2004). Prolonged
searches for a passage route past an obstacle also expose fish to
predators and anglers, which may cause surviving fish to con-
sume even more energy in their struggle, energy that otherwise
could be used for reproduction (Kinnison et al. 2001,2003). As
anadromous species feed less during their riverine migratory
phase, energy-consuming migrations may increase mortality and
act as evolutionary selection forces (Bernatchez and Dodson
1987;Jonsson and Jonsson 2011).
Upstream-migrating salmon and trout often seek areas with
high velocities and flows. This behaviour is thought to be an evo-
lutionary mechanism that ensures spawning success by enabling
the fish to negotiate along the main river stem and reach suitable
spawning grounds (Ferguson et al. 2002). Hence, discharge of
water is probably the single most important factor for attracting
upstream migratory salmonids to fishway entrances, and large
variation in the operation scheme of flow and (or) spill releases
from dams may hinder or delay upstream migration (Larinier
2001 and references within; Rivinoja et al. 2001;Williams et al.
2012 and references within). In many cases, the main reason for
upstream passage failure often occurs when turbine and (or) spill-
way discharge flow masks attraction flow from fishway entrances
(Vegar and Kraabøl 1996;Karppinen et al. 2002;Thorstad et al.
2003). For example, ascending fish may be steered towards blind
alleys and interrupted or delayed in their upstream migration if
attraction flow from turbine or spillway discharge is high rela-
tive to fishways or if spill is released far from the fishway
(Bjornn and Peery 1992;Kraabøl 2012). Therefore, application of
optimal spillway manipulations seems to be a potentially cost-
effective mitigative action to improve fishway performance. To
survive, reproduce, and maintain natural adaptations to the
preregulated river environment, fish rely on the ability to make
decisions that encourage behavioural patterns that lead to suc-
cessful passage (Brown and Laland 2003;Laland et al. 2003).
Thus, any fish passage solution should consider the behaviour
of the fish, including their cognitive abilities and motivational
state, as seen in Goerig and Castro-Santos’(2017) study of brook
trout (Salvelinus fontinalis), where they found individual variabili-
ty in attempts to pass even after accounting for the effects of
hydraulics, diel period, and physiology. Although individuals
respond directly to the hydrodynamic and hydraulic environ-
ment in the tailrace section, behavioural responses may also
depend on previous experience and the ability to learn from
prior experiences during passage attempts. Kieffer and Colgan
(1992), for example, describe how fish learn about their sur-
roundings through trial and error, but also by observing the
behaviour of conspecifics. Thus, the cognitive abilities of fish
may enable them to modify their search behaviour from previ-
ous experiences. If so, this may entail searching for, or avoiding,
certain objects or areas (Goodyear 1973).
Studies of fish passage typically involve some sort of tagging,
which in turn involves capturing and handling of fish. Handling
the fish during capture, anaesthesia, tagging procedures, and
transportation is associated with an increase in stress and may
result in abnormal behaviour (Thorstad et al. 2008 and references
within). Whether or not the stress generated from handling is
related to learning, with subsequent effects on passage, is in
need of further investigation. Nevertheless, stress associated
with handling may not only have a negative effect on passage suc-
cess, but may also lead to faulty evaluations of the performance
of a given fish passage solution, potentially leading to poor man-
agement decisions (Nyqvist et al. 2017).
In this study, we investigated how mature wild spawners of
large-bodied, landlocked Atlantic salmon (Salmo salar)inthe
River Klarälven, Sweden, and brown trout in the River Gud-
brandsdalslågen, Norway, performed during their attempts to
locate and ascend fishways. The purpose of the study was to
(i) assess fishway trap efficiency at both study sites; (ii) compare
the behaviour and performance of salmon and trout translocated
from successful fishway entries with behavior and performance
of ascending salmon and trout without previous experience from
successful entries of the fishway; and (iii) describe and analyze
behaviour and performance of adult salmon approaching a large
and complex hydroelectric dam and power station in relation to
spillway and turbine outlets in the tailrace area. The shortage of
experimental studies on these topics represents a major knowl-
edge gap, and there is a need for descriptive as well as experimen-
tal studies dealing with the behavioural details during the
migratory phase of salmonids in regulated rivers.
Materials and methods
The study was conducted at the Forshaga dam in the River Klar-
älven and at the Hunderfossen dam in the River Gudbrandsdal-
slågen (Fig. 1). Both dams are associated with a run-of-the-river
power plant and a dam section equipped with several spillways.
The environmental conditions are characterized by several
waterways across the dam and great variation in water discharge
during the migratory season. The River Klarälven was investi-
gated during 2012 and 2013 and the River Gudbrandsdalslågen
during 2013.
Study area in Sweden (salmon)
The River Klarälven (catchment area 11 800 km
2
)stretches460km
through Norway and Sweden before it empties into Lake Vänern
(5650 km
2
), Sweden’s largest lake. This river is the major spawning
and nursery river for landlocked Atlantic salmon in Lake Vänern.
The mean annual discharge at the river outlet is 162.5 m
3
·s
–1
,witha
mean annual high of 690 m
3
·s
–1
(www.smhi.se). During the two
study years, the mean discharge for June to September was 265
m
3
·s
–1
in 2012 and 165 m
3
·s
–1
in 2013 (Fig. 2). The River Klarälven has
been dammed for hydropower purposes since the beginning of the
1900s (Piccolo et al. 2012), and today there are 11 hydropower plants
in the river, of which nine are situated in Sweden and two in
Norway. The main available spawning areas occur above the eighth
dam, within a free-flowing reach of 140 km. There are no known
spawning grounds below the first hydropower plant.
All hydropower dams located in the Swedish portion of the
River Klarälven lack fishways and are not passable for upstream-
migrating fish. Instead, the upstream-migrating salmonid spawners
are collected in a fishway equipped with a trap at the lowermost
power plant in Forshaga, 25 river km upstream of Lake Vänern
(Fig. 3). The maximum intake capacity at Forshaga power plant is
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
2 Can. J. Fish. Aquat. Sci. Vol. 00, 0000
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.

Fig. 1. The catchment areas for Lake Mjösa, River Gudbrandsdalslågen, Norway, and Lake Vänern, River Klarälven, Sweden. Map made
using ArcMap 10.5.
Fig. 2. Discharge (m
3
·s
–1
) in the River Klarälven during the migration period in 2012 (solid black line) and 2013 (broken black line) and in
the River Gudbrandsdalslågen in 2013 (solid grey line).
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Hagelin et al. 3
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.

163 m
3
·s
–1
, and water is spilled through four spillways and a log
chute distributed across the dam, depending on river discharge
(Fig. 3). Once captured, the fish are transported by truck and
released upstream of the eighth power plant where the spawning
grounds are situated (see also Hagelin et al. 2016 for details of the
system).
There are two entrances to the fishway, one facing the spill
area and one facing the turbine outflow area. Both entrances lead
the fish to a single large pool into which auxiliary water is
released (Figs. 3 and 4). From each entrance, fish can ascend a
Denil fishway to a false weir that empties into a downward-
sloping tube that leads them into an indoor collecting basin
where the fish are held up to a week until transported (Fig. 4).
The fishway discharge is about 1 m
3
·s
–1
in the ladder but the
attraction flow can be up to 3 m
3
·s
–1
. All spawners caught in the
trap are netted, sorted by species and origin (hatchery-reared or
wild, sorted by the absence of an adipose fin in reared fish), and
measured for body length and sex before transporting them fur-
ther upstream.
The fish trap in Forshaga was open 106 days between 11 June
and 27 September 2012 and 110 days between 21 May and 3 Octo-
ber 2013. From 2004 to 2013, an annual mean number of 628
(range: 292–1031) wild and 592 (range: 124–992) hatchery-reared
Atlantic salmon were caught in the trap (data from Fortum Gen-
eration AB).
Study area in Norway (trout)
The River Gudbrandsdalslågen (catchment area 11 500 km
2
)
is the major spawning and nursery river for the landlocked and
large-bodied brown trout in Lake Mjøsa (365 km
2
). The mean
annual discharge is 248 m
3
·s
–1
, with a mean annual high of
630 m
3
·s
–1
(Fig. 2). A 78 km river section is available for ascend-
ing trout, of which 62 km is situated upstream of the dam and
reservoir at the Hunderfossen power plant. In this study, we
focused on trout movements downstream the dam. Out of 17
major and minor spawning areas recorded in the river, 10 are
located upstream of the dam (Kraabøl and Arnekleiv 1998);
the rest are located downstream of the power plant. In this
study, however, we tagged fish in the fishway (termed experi-
enced) and immediately downstream of the dam (termed
naïve), so these fish are likely looking to reach areas further
upstream.
Hunderfossen dam was constructed in 1960–1964 (Fig. 5).
The maximum intake capacity of Hunderfossen power plant
is 320 m
3
·s
–1
. After passing the turbines, the water is led back
to the river, 4.4 km downstream of the dam. Spill water is
released through seven spill gates, one timber gate, and one
ice and trash spillway (Fig. 5).
The fishway (1.8 m
3
·s
–1
) at Hunderfossen dam consists of a pool-
and-weir section (from the entrance to the fish trap) and a Denil
section (from the fish trap to the outlet) (Fig. 5). From 2004 to
2013, the trap caught an annual mean number of 508 (range: 305–
685) brown trout. The fishway empties directly into a deep pool
below the dam, where there are three different entrances (Fig. 5).
Radiotagging and tracking Atlantic salmon in Sweden
In 2012 we tagged 16 wild Atlantic salmon (Table 1) caught in
fyke nets in Lake Vänern, 5 km f rom the river mo uth. The mean
length of the tagged salmon was 71 cm (range: 61–79 cm; SD =
5.0 cm). Th e fish were tagged on a boat and then either released at
the capture site in the lake (n= 10) or transported in a tank and
released 5 km upstream of the river mouth (a total transport dis-
tance of 10 km) (n= 6). Fish were tagged on four occasions: 19
and 26 June and 3 and 6 July. All fish in 2012 were tagged before
they reached the fishway in Forshaga and are hereinafter
referred to as “naïve”.
In 2013, 20 wild Atlantic salmon (Table 1) were caught in fyke
nets at the same location as in 2012. The fish were removed from
the fyke net, tagged, and then immediately released back into
the lake. We tagged fish on four occasions: 19, 24, and 27 June
and 8 July. The mean length of the tagged salmon was 73 cm
(range: 57–86 cm, SD = 5.5 cm). In addition, 20 “experienced”fish
Fig. 3. Aerial photo (Google Maps) of the power plant at Forshaga, River Klarälven, showing the location of the three turbines, four
upward-opening spill gates, and one downward-opening log chute. There are two entrances to the fishway. The area depicted with “1”
represents the spill area, and “2”represents the turbine outflow area, with the dotted line showing the border between the two areas.
The three stars show the location of the telemetry antennas. [Colour online.]
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
4 Can. J. Fish. Aquat. Sci. Vol. 00, 0000
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.

(Table 1)fromthefish trap in Forshaga were also tagged and then
transported by truck in an aerated tank and released 4 km down-
stream of the power plant. The mean length of the experienced
tagged salmon was 74 cm (range: 62–85 cm; SD = 6.1 cm). Again,
the fish caught in the lake were regarded as “naïve”fish, lacking
experience from entering the fish trap in 2013, whereas the fish
caught in the fishway trap were treated as experienced. The
experienced fish were all tagged on 28 June.
Fig. 4. Schematic diagram of the fish trap in Forshaga.
Fig. 5. Aerial photo (Google Maps) of the power plant at Hunderfossen, Gudbrandsdalslågen, showing the intake to the power station and
the placement of the seven spillways and the fishway solutions. The fish trap is located between the pool and weir section (from the
fishway entrances to the fish trap; in green) and the Denil section (from the fish trap to the outlet; in blue). [Colour online.]
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Hagelin et al. 5
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.

All fish were inspected for injuries and measured to the nearest
centimetre (total fish length, L
T
), after which they were tagged
with external radio transmitters (model F2120, Advanced Teleme-
try Systems (ATS), Isanti, Minnesota, USA), with a mortality signal
that becomes activated after 8 h of no movements. The tags
weighed 16 g, which is well below the recommended maximum
of 2% of total body mass (Winter 1983;Thorstad et al. 2000), and
measured 21 mm 52 mm 11 mm, lying flat against the fish’s
body. During the tagging procedure, in which two coated wires
are pierced through the area below the dorsal fin, fish were kept
in a dark plastic tube filled with river or lake water. As soon as
the tagging was done, the fish were put in a large tank to monitor
their recovery. To minimize further stress, the fish were treated
without anesthetics as described by Finstad et al. (2005).Allfish
were tagged in temperatures below 19 °C.
To track the tagged fish, we used stationary loggers and man-
ual tracking. We placed three stationary data loggers, model
R4500S (ATS), connected to six-element Yagi antennas, around
the power plant to detect movements in the area downstream of
the dam. One was placed near the turbine outlet, one at the spill
gate closest to the fish trap, and one at the spill gate furthest
away from the fish trap (Fig. 3). Before tagging we used the sta-
tionary loggers to provide a signal map of the area so the we
could tell, more specifically, where the fish resided. Manual
tracking, using receiver model R4000 (ATS) and a three-element
Yagi antenna, was conducted approximately every third day
from the time of release until the fish had either been caught in
the trap or had moved back downstream to the lake. Manual
tracking was primarily done on foot but also by boat and covered
the area from the power plant to the river mouths. We also had
additional antennas at the three mouths of the River Klarälven to
detect river entry and downstream movements into Lake Vänern.
Radio-tagging and tracking trout in Norway
During the spawning migration season in 2013, artificial fre-
shets were used to attract and capture brown trout at spillway 1
on the eastern side of Hunderfossen dam (Kraabøl 2012;Fig. 5).
When flow through spillway 1 exceeds 10 m
3
·s
–1
, ascending trout
gather in the deep pool below the dam (Fig. 5). Trout were
trapped in pots or stranded after spillway closure and netted and
secured as quickly as possible in a 1 m deep pool. A total of 44
brown trout were captured, tagged, and released into the deep re-
cipient pool immediately below the fishway (Table 1;Fig. 5). All
individuals were tagged with Floy anchor tags, and a subsample
was radio-tagged (n= 9) on 9 an d 29 August 20 13. Th e mean length
of the tagged trout was 68 cm (range: 48–88 cm; SD = 9.9 cm).
These fish were termed “naïve”as they had not entered the fish
trap during 2013. In addition, 73 experienced brown trout were
captured in the fish trap (Fig. 5;Table 1). All individuals were
tagged with Floy anchor tags, and a subsample was radio-tagged
(n= 10). These individuals were released downstream of the dam
in the same pool as the naïve fish. The mean length of the experi-
enced trout was 72.5 cm (range: 48–87 cm; SD = 9.6 cm).
For all fish, origin (hatchery-reared or wild, sorted by the ab-
sence of an adipose fininrearedfish) of the fish was noted, and
total fish length (L
T
) was measured to the nearest centimetre.
The hatchery-reared fish were stocked as 2-year-olds (lengths
20–25 cm) and had been at least three growth seasons in Lake
Mjøsa before they returned as spawners. Previous studies have
shown no effect of origin on downstream or upstream migration
and behavior of trout at the Hunderfossen dam (Kraabøl 2012),
but we nevertheless included origin in the analysis of the trout in
this study. The Floy anchor tags (Floy Tag, Seattle, Washington,
USA) were inserted in front of the dorsal finbyaFloypistol.The
unanesthetized fish were kept in a dark plastic tub filled with
river water during tagging. Radio-tagging followed the same pro-
cedure as in Sweden. After tagging and measuring, all tagged na-
ïve trout were carried in opaque, dark plastic bags filled with
water and released in the deep pool below the dam. Experienced
trout were put in a large fish tank and transported with a crane
to the deep pool downstream of the fishway. The handling time
was approximately the same for the naïve and experienced fish.
All fish were ta gged in temperatures below 19 °C.
Tagged trout (Floy and radio tags) caught in the fish trap were
identified on a daily basis by the staff at Hunderfossen trout
hatchery. To follow the radio-tagged fish, we used manual track-
ing. The fish were positioned by using an ATS R4500s receiver
and a handheld antenna from land (there are roads along both
sides of the river, running from the dam to Lake Mjøsa). The posi-
tion of trout was identified on average every 2.7 days (i.e., 23
times in the period 30 August to 1 November 2013).
Statistics
We calculated fishway efficiency (Eberstaller et al. 1998), or to
be more exact fishway trap efficiency, for the fishways in For-
shaga (Atlantic salmon) and Hunderfossen (brown trout). Fish-
way trap efficiency was defined as the proportion of tagged fish,
residing in the area, that entered the fishway and were captured
in the fish trap. In Forshaga, only fish actively approaching the
dam were considered residing in the area, whereas in Hunderfos-
sen all tagged fish were assumed to be residing in the area, since
they were captured and released immediately below the fishway.
Since experienced fish had already passed the fishway once, we
only used recaptures of naïve salmon and trout (one year at Hun-
derfossen and both years at Forshaga) to estimate fishway trap ef-
ficiency, as this measurement best describes efficiency under
“normal”conditions.
To make comparisons between naïve and experienced fish,
we also calculated fishway trap efficiency for tagged fish (i.e.,
capture rate for naïve fish and recapture rate for experienced
fish) that were detected near the dam. These calculations were
based on data from 2013 for both naïve and experienced Atlan-
tic salmon and brown trout. Analysis of these data was done
using a binary logistic regression model BSTEP, exploring the
relationship between the probability (P) of observing captures
in the fishway at the dam in Forshaga (Atlantic salmon) and
Hunderfossen (brown trout) and three different fish state
Table 1. Overview of the tagged Atlantic salmon and brown trout during 2012–2013 in
the River Klarälven (Sweden) and the River Gudbrandsdalslågen (Norway).
Year
Sex Length (cm)
Female Male Mean Range SD
Naïve salmon 2012 —— 64–79 4.5
Experienced salmon 2013 13 7 74.5 62–85 6.4
Naïve salmon 2013 10 10 72.9 57–86 5.5
Experienced brown trout 2013 52 21 72.5 48–87 9.6
Naïve brown trout 2013 32 12 68.3 48–88 9.9
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
6 Can. J. Fish. Aquat. Sci. Vol. 00, 0000
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.

variables, namely the level of experience (experienced or naïve
fish), sex, and length. In addition, origin (wild or hatchery-
reared) was included as a variable in the brown trout analysis.
The movement between areas 1 and 2 in the River Klarälven
(Fig. 3) was analysed for naïve Atlantic salmon from 2012 and
2013 using a generalized linear model. “Occupancy of an area”
was treated as a binary response variable and year as a factor. The
spill difference during the time period between manual tracking
occasions was quantified as either absolute or relative spill (i.e.,
absolute spill was set as maximum spill minus the minimum
spill, and relative spill was set as the ratio of mean spill to mean
flow (Q)), and both were used as covariates together with fish
length.
We also tested for behavioural differences between naïve and
experienced fish. The behaviours tested were delay and motivation.
Delay (i.e., the number of days from entering the area below the
power plant (Fig. 3) (Sweden) or from being tagged (Norway) to enter-
ing the fish trap) was analyzed using a Mann–Whitney test. For At-
lantic salmon we used data from 2012 and 2013, and for brown trout
we used data from 2013. Motivation, expressed as number of
attempts to enter the trap, was calculated both totally and per day
andtestedusingaMann–Whitney test. An attempt is defined as an
occasion when the fish resided in the vicinity of the fishway en-
trance, positioned by the automatic loggers. This metric was only
computed for Atlantic salmon in Klarälven in 2012 and 2013.
A separate success rate was also analysed for the Atlantic
salmon studied in 2012. Here, we compared the performance of
naïve Atlantic salmon that were treated in two different ways:
(i) tagged and released directly into the lake or (ii) tagged, trans-
ported, and then released into the river. This analysis was done
using a binary logistic regression BSTEP. Whether or not the fish
made attempts and (or) passed the fishway was used as the
response variable, and handling of fish (released directly versus
transported and released) and length were treated as factors. All
statistical analyses were carried out in IBM SPSS Statistics 24.
All handling of fish in Sweden was performed in agreement
with the animal welfare permit No. 2013/85 from the Swedish
Board of Agriculture and in Norway in agreement with animal
welfare permit No. 2013/116588.
Results
Fishway trap efficiency
For Atlantic salmon at Forshaga, 11 of the 16 radio-tagged naïve
salmon made attempts to enter the fishway in 2012 (i.e., they
resided in the area). Two of the 11 salmon eventually entered the
fishway and were captured in the collecting basin in the fish trap.
The transmitters switched to mortality mode for five of the
salmon, indicating the fish died, became inactive, or lost their
transmitters before entering the fishway. Thus, fishway trap effi-
ciency in 2012 was 18% if the fish with mortality signals are
included and 33% if they are excluded. In 2013, 18 of the 20 naïve
radio-tagged salmon made attempts to enter the fishway (i.e.,
they resided in the area). Of the 18 salmon, 14 were captured in
the trap, and two either died or lost their transmitter (for the
same reasons as outlined above). Thus, fishway trap efficiency in
2013 was 78% if the fish with mortality signals are included and
88% if they are excluded. The average discharge for the migration
period (i.e., when the fish were in the area) was higher in 2012
(276 m
3
·s
–1
)thanin2013(121m
3
·s
–1
), and consequently there was
also more spill in 2012 (Fig. 6). Most fish (81%) entered the trap on
days without spill (Figs. 6aand 6b).
For brown trout at Hunderfossen in 2013, 19 of the 44 naïve
tagged (Floy and radio) trout were captured in the trap in the fish-
way. Thus, fishway trap efficiency was 43%.
Naïve versus experienced
There was a significant effect of level of experience (naïve ver-
sus experienced) on the number of salmon that entered the fish
trap in Forshaga in 2013 (Z=7.5,P= 0.006). Of the 20 tagged naïve
salmon, 70% were captured in the trap as compared with 25% of
the 20 experienced ones (Table 2;Fig. 7). Further, 10% of the naïve
and 45% of the experienced salmon ceased upstream migration.
Based on manual tracking and information from the dataloggers,
they moved downstream and left the area after tagging and
release, and they did not move upstream to the power plant area.
A similar difference in success rate between naïve and experi-
enced fish was found for brown trout at Hunderfossen dam (Z=
10.35, P= 0.001). Out of 44 tagged naïve brown trout, 43% were
captured in the fish trap, whereas only 15% of the experienced
(n= 73) ones were captured in the fish trap (Fig. 7). Manual posi-
tioning of the radio-tagged fish revealed that 11% of the naïve and
50% of the experienced brown trout ceased migration after tag-
ging and release. We found no significant effects of sex and fish
length on the success rate for salmon or trout (Table 2). In addi-
tion, we found no significant effects of origin (wild or hatchery-
reared) for trout.
Searching behaviour
Based on telemetry data, we found that salmon searched for a
passage route in waterways releasing the highest water dis-
charges. When the majority of water discharge was released
through the turbines, the salmon were mainly observed in tur-
bine zone 2, and when the majority of water was released
through the spill gates, the fish were more often located in spill
zone 1 (Fig. 3). Periods of high spill water release (Z=21.04,
P≤0.001) and increasing spill (Z=4.71,P= 0.03) stimulated the
fish to move to zone 1 in both 2012 and 2013 (Figs. 6aand 6b;Table 3).
There was also a significant difference between years (Z= 8.68,
P= 0.003), where more fish moved to the spill zone in 2012, and an
effect of fish length (Z= 5.14, P= 0.023), where smaller fish
responded by searching more actively towards the spill than
larger individuals (Table 3).
Delays
Naïve salmon were delayed by a median of 15 days (3–70 days),
whereas the experienced salmon were delayed by 7 days (7–70 days).
For brown trout, the median delay for experienced and naïve trout
was 26 (2–47) and 25 (4–43) days, respectively. The difference in
delays between experienced and naïve fish was not significant for
either species (Mann–Whitney Utest: U=36.5,N
1
=16,N
2
=5,
P=0.771forsalmon;U=95.0,N
1
=19,N
2
=11,P= 0.698 for trout).
We found no differences in total number of attempts to enter
the fishways between the naïve (median 1 attempt) and experi-
enced (median 2 attempts) salmon (Mann–Whitney Utest: U=
87.0, N
1
= 18, N
2
=11,P= 0.444). The same was true if expressed as
the number of attempts per day, with a median of 0.027
attempts·day
1
for the experienced fish and 0.037 attempts·day
1
for the naïve ones (Mann–Whitney Ustatistics: U=94.0,N
1
= 18,
N
2
=11,P= 0.822).
Handling
There was a significant difference in the number of attempts
between the two groups of naïve salmon that had been handled
differently in 2012 (logistic regression, Z=4.49,P= 0.034). Out of
the 10 salmon that were released immediately after tagging,
seven made attempts to enter the fishway, and two succeeded in
entering the trap. Out of the six salmon that were transported to
the river, two made attempts to enter the fishway, but none man-
aged to enter it.
Discussion
All migratory salmonid populations are dependent on spawners
reaching spawning grounds, and therefore in regulated rivers, it is
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Hagelin et al. 7
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.

Fig. 6. The percentage of salmon (of the total number of salmon in the area at that time; depicted by the histobars) observed in the
vicinity of the turbine outlet (area 2 in Fig. 3) at Forshaga in 2012 (a)and2013(b). Arrows represent times of salmon entering the fish trap.
[Colour online.]
Table 2. Results for the binary logistic regression model for the probability of captures
in the fish trap, showing the Wald statistic and the odds ratios (exp(
b
)) with 95% CI for
captures in the fish traps.
Model Predictor Coefficients
Wald statistic
exp(
b
) 95% CIZ P
Salmon Intercept 0.85 3.02 0.082 0.43
Experienced 0 1
Naïve 1.95 7.50 0.006 7 0.43–28.17
Trout Intercept 0.27 0.81 0.37 1.32
Experienced 0 1
Naïve 1.44 10.35 0.001 4.22 1.75–10.12
Note: A backward stepwise likelihood procedure suggested that the variables “sex”and “individual
fish length”should be excluded from the model. The variables “experienced”or “naïve”refer to the
salmon and trout that either passed or failed to pass a fishway.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
8 Can. J. Fish. Aquat. Sci. Vol. 00, 0000
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.

paramount that spawners can successfully pass dams to reach
them. Our study of Atlantic salmon and brown trout showed that
the situations for single dam passages in two regulated rivers could
potentially be improved as passage success varied in Klarälven and
was unsatisfactorily low in Gudbrandsdalslågen. Our analyses also
showed that previous experience, defined as prior successful entry
into the fishway traps, had a negative effect on passage success for
the same fishway later in the season.
Reproductive success is of fundamental importance for all pop-
ulations, and for migrating fish it is tightly linked with both
migration success and timing (Dingle 1996). Within this context,
well-functioning fishways are required, where functionality not
only involves successful passage (i.e., the proportion of fish that
enter and pass a fishway), but also passage with little delay. If
delay is substantial, reproductive success may be negatively
impacted due to aborted spawning migrations, reduced windows
of opportunity to mate due to delayed arrival at spawning sites,
and forced spawning at suboptimal and over-crowded areas
below dams (Gorsky et al. 2009;Holbrook et al. 2009). While the
observed delays for salmon (7–15 days) and trout (25–26 days)
passing the dams at Forshaga and Hunderfossen may not have se-
rious consequences for reproductive success, although one may
question this for trout, there is seldom only one dam for fish to
pass to reach their spawning grounds. In the River Klarälven, the
fish would need to pass eight dams to reach the spawning
grounds in Sweden and an additional three to reach the Norwe-
gian spawning grounds. The cumulative delay associated with so
many dams would undoubtedly have a negative effect on repro-
ductive success, and thus the current truck and transport system
seems to be the only viable alternative.
A well-functioning fishway must work well under a variety of
flow conditions. We found that mean fishway (trap) efficiency in
River Klarälven varied greatly, ranging from 18% in a high flow
year to 88% in a year of normal flow conditions. Our measure of
efficiency in Gudbrandsdalslågen may be an underestimate, as
we assumed all of the fish remained in the dam area after release
(i.e., we did not radio-track all fish). Even if this is the case, the ef-
ficiency in Gudbrandsdalslågen (47%) was higher than a previous
report of 21% to 39% for other years (Kraabøl 2012). Fish used in
these previous reports (Kraabøl 2012) were all considered naïve.
Hence, there was large annual variation in fishway performance
at both dams, which have run-of-the-river hydropower plants.
Fishways at run-of-the-river hydropower plants have previously
been shown to have large interannual variation in their passage
performance, which has been ascribed to differences in flow condi-
tions, resulting from different patterns of spill in relation to flow
(Rivinoja et al. 2001;Lundqvist et al. 2008). We found, for example,
that Atlantic salmon were more likely to enter the fishway when
spill was low or nil and moved from the tailrace to the spill area
when spill increased, a pattern also reported by Rivinoja et al.
(2001). While we did not study in detail the behaviour of the trout
at Gudbrandsdalslågen, previous studies here have estimated fish-
way performance to be optimal at flows of 2–20 m
3
·s
–1
, suboptimal
up to 180 m
3
·s
–1
, and completely dysfunctional at higher flows
(Jensen and Aass 1995). In Klarälven, we also saw a size-dependent
effect of spill, where small fish responded by searching more
actively towards the spill area than large fish. We can only specu-
late as to why this occurred, but it may be related to large individu-
als being more successful at holding position (Fleming 1996;
Fleming et al. 1997).
Fig. 7. Percentage of naïve and experienced individuals of (a) Atlantic salmon and (b) brown trout that succeeded in entering the fish trap
at Forshaga and Hunderfossen dams, respectively. Total sample sizes are indicated above each histobar.
Table 3. Results for the generalized estimating equations model for occupancy in an area, showing the Wald statistic and
the odds ratios (exp(
b
)) with 95% CI.
Model Predictor Coefficients
Wald statistic
exp(
b
) 95% CIZ P
Salmon Intercept 6.36 4.30 0.067 578.27
Year 1.65 8.68 0.003 0.19 2.75 to 0.55
Length 0.93 5.14 0.023 0.91 0.17 to 0.13
Spill (m
3
·s
1
) 2.58 21.04 0.000 13.20 1.48 to 3.68
Spill increase (m
3
·s
1
) 0.01 4.71 0.030 1.01 0.00 to 0.26
Note: Variables in the model are years 2012 and 2013 and individual fish length. Total spill and spill increase were measured during the
time between two tracking occasions.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Hagelin et al. 9
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.

The low efficiency at high spill observed in the River Klarälven
will require one or more countermeasures to reduce the likeli-
hood that fish become attracted away from fishway entrances.
One way to increase fishway trap efficiency would be to establish
operational schemes for releasing spill water near the fish trap
entrance during spawning migrations. Such a spill plan was
recently implemented in Klarälven (in 2018), but to date it has
not been evaluated. Another countermeasure that one might
consider is installation of physical screens along the tailrace to
hinder fish from moving away from fishway entrances towards
spill or turbine discharge areas. At the Pitlochry Dam, Scotland,
Webb (1990) found that 55% of the ascending Atlantic salmon
passed fishways successfully without screens, but 100% passed af-
ter screens had been installed (Gowans et al. 1999).
Catching, handling and transporting fish may affect the up-
stream migration behaviour and success due to stress (Jokikokko
2002;Portz et al. 2006) and exhaustion (Cooke and Hinch 2013).
Many studies looking at passage success have still used fish col-
lected from passage facilities (i.e., “experienced”)fish that have
been exposed to some degree of handling (e.g., Roscoe et al. 2011;
Bunt et al. 2012). Unfortunately, the difficulties associated with cap-
turing naïve spawners in large lakes and rivers often necessitates
the need to use fish captured in fishways (i.e., it is the only viable
option). Our study indicates that there may be biases associated
with using “experienced”fish. We do not know the reason for the
lower passage success for experienced fish than for naïve fish.
Nevertheless, it is important to consider that prior experience as
defined here includes more than experience entering a fishway; it
also includes effects from handling or transport procedures as well
as other experiences such as delays in the fishway traps or exhaus-
tion from migrating in a portion of the river for a second time. We
believe that a likely explanation for the difference in fishway trap
efficiency for naïve and experienced fish, which is consistent with
the results for the 2 years of study in the River Klarälven and for the
River Gudbrandsdalslågen, may be related to stress and that stress
reactions may have over-ruled the effects of prior success. Both na-
ïve and experienced fish were subjected to stress associated with
handling and tagging, but the experienced fish were also subjected
to stress associated with being held in the collection basin and, for
salmon, transportation from the fishway trap to the release site in
the river. Further support for an effect of stress can be seen in the
response of naïve and experienced salmon directly after release,
where 10% of the naïve salmon and 45% of the experienced ones
ceased migration after tagging and release and did not move up to
the power plant area.
Even if a tenable explanation for this difference in behavior
and performance of naïve and experienced salmon and trout is
stress-related, we cannot rule out the possibility that this differ-
ence may be related to learning as well. Previous experimental
studies have provided indications that fish can remember nega-
tive experiences for a substantial amount of time (Odling-Smee
and Braithwaite 2003;Yue et al. 2004) and that the primary func-
tion of fear and stress is to help animals avoid danger (Paul et al.
2005) and thereby avoid certain places (Portavella et al. 2004;Yue
et al. 2004). In the study conducted in 2012, we found that a larger
percentage of the salmon released directly into the lake
attempted to enter the fishway (70%) than fish transported 10 km
into the stream before released (33%). Thus, it once again seems
likely that the difference in behaviour and performance of naïve
and experienced fish is stress-related and (or) related to second-
ary effects such as energy consumption, where stress associated
with transportation alone may be sufficient to produce behaviou-
ral differences.
Overall, our results suggest that evaluations of fishway per-
formance (efficiency) should be tempered with caution, depend-
ing on the source of individuals used in such studies. More
research is needed to understand the reason for the difference we
observed, but the potential bias associated with prior experience
appears to be general, as we obtained similar results with two dif-
ferent species in two different river systems. Our results also
underscore the need to consider fishway efficiency during multi-
ple years, presumably related to interannual differences in flow
conditions. There are many ways to deal with interannual flow
variation, such as establishing a spill plan and using screens to
influence route choice. Other possibilities are to increase attrac-
tion flow at the traps, design fishways with multiple entrances, or
in some cases, in particular in large rivers, to construct more than
one fish passage solution and in that way cover the broad range of
flow conditions that the fish face (Larinier 2001).
Fishway efficiency and fish behaviour during passage is
affected by a number of factors, factors that may cause cumula-
tive responses, and it is difficult to single out the importance of
each factor. We have, in this study, attempted to illustrate the im-
portance of prior experience to fishway efficiency, but we also
recommend further research to continue to investigate and pin-
point the different factors affecting fishway efficiency to reduce
losses during migration in the future.
Acknowledgements
We thank Anders Andersson and Daniel Nilsson for help with
the fieldwork in the River Klarälven, Måns Bagge for support in
catching fish in Lake Vänern, and Klas Jarmuszewski and the crew
at Fortum Generation AB for practical support at the fish trap. The
crew at the Hunderfossen power plant and trout hatchery are also
thanked for their contributions to trapping, maintaining, and
transporting the trout. The study in Sweden was carried out as part
of the EU, European Regional Development Fund, Interreg project
“Free passage for landlocked Lake Vänern salmon”. The study in
Norway was funded by the Norwegian Research Council (NRC)
through the MILJOE2015 program (thematic area: Water), which
supports the RIVERCONN project (grant No. 221454) and the
SAFEPASS-project (grant No. 244022).
References
Bernatchez, L., and Dodson, J.J. 1987. Relationship between bioenergetics
and behavior in anadromous fish migrations. Can. J. Fish. Aquat. Sci. 44(2):
399–407. doi:10.1139/f87-049.
Bjornn, T.C., and Peery, C.A. 1992. A review of literature related to move-
ments of adult salmon and steelhead past dams and through reservoirs
in the lower snake river. Technical Report FWS-TR-92-1. University of
Idaho –Moscow Dept. of Fish and Wildlife Resources.
Brown, C., and Laland, K.N. 2003. Social learning in fishes: a review. Fish
Fish. 4(3): 280–288. doi:10.1046/j.1467-2979.2003.00122.x.
Bunt, C.M., Castro-Santos, T., and Haro, A. 2012. Performance of fish passage
structures at upstream barriers to migration. River Res. Appl. 28(4): 457–
478. doi:10.1002/rra.1565.
Clay, C.H. and Eng, P. 2017. Design of fishways and other fish facilities. CRC
Press. doi:10.1201/9781315141046.
Cooke, S.J., and Hinch, S.G. 2013. Improving the reliability of fishway attrac-
tion and passage efficiency estimates to inform fishway engineering, sci-
ence, and practice. Ecol. Engineer. 58:123–132. doi:10.1016/j.ecoleng.2013.
06.005.
Cooke,S.J.,Hinch,S.G.,Crossin,G.T.,Patterson,D.A.,English,K.K.,
Shrimpton, J.M., et al. 2006. Physiology of individual late-run Fraser River
sockeye salmon (Oncorhynchus nerka) sampled in the ocean correlates with
fate during spawning migration. Can. J. Fish. Aquat. Sci. 63(7): 1469–1480.
doi:10.1139/F06-042.
Dingle, H. 1996. Migration: the biology of life on the move. Oxford Univer-
sity Press, Oxford.
Eberstaller, J., Hinterhofer, M., and Parasiewicz, P. 1998. The effectiveness of
two nature-like bypass channels in an upland Austrian river. In Migration
and fish bypasses. Edited by M. Jungwirth, S. Schmutz and S. Weiss. Fish-
ing News Books, Oxford. pp. 363–383.
Ferguson, J.W., Williams, J.G., and Meyer, E. 2002. Recommendations for
improving fish passage at the Stornorrfors Power Station on the Umeälven,
Umeå, Sweden. Department of Commerce, National Marine Fisheries Serv-
ice, Northwest Fisheries Science Center, Seattle, Washington.
Finstad, B., Økland, F., Thorstad, E.B., Bjørn, P.A., and McKinley, R.S. 2005.
Migration of hatchery-reared Atlantic salmon and wild anadromous
brown trout post-smolts in a Norwegian fjord system. J. Fish Biol. 66(1):
86–96. doi:10.1111/j.1095-8649.2004.00581.x.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
10 Can. J. Fish. Aquat. Sci. Vol. 00, 0000
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.

Fleming, I.A. 1996. Reproductive strategies ecology and evolution of Atlantic
salmon. Rev. Fish Biol. Fish. 6(4): 379–416. doi:10.1007/BF00164323.
Fleming,I.A.,Lamberg,A.,andJonsson,B.1997.Effectsofearlyexperience
on the reproductive performance of Atlantic salmon. Behav. Ecol. 8(5):
470–480. doi:10.1093/beheco/8.5.470.
Goerig, E., and Castro-Santos, T. 2017. Is motivation important to brook
trout passage through culverts? Can. J. Fish. Aquat. Sci. 74(6): 885–893.
doi:10.1139/cjfas-2016-0237.
Goodyear, C.P. 1973. Learned orientation in the predator avoidance behavior
of mosquitofish, Gambusia affinis. Behaviour, 45(3–4): 12–223. doi:10.1163/
156853974X00642.PMID:4707170.
Gorsky, D., Trial, J., Zydlewski, J., and Mccleave, J. 2009. The effects of smolt
stocking strategies on migratory path selection of adult Atlantic salmon
in the Penobscot river. Maine. North Am. J. Fish. Manage. 29(4): 949–957.
doi:10.1577/M08-068.1.
Gowans, A.R.D., Armstrong, J.D., and Priede, I.G. 1999. Movements of adult
Atlantic salmon in relation to a hydroelectric dam and fish ladder. J. Fish
Biol. 54(4): 713–726. doi:10.1111/j.1095-8649.1999.tb02028.x.
Hagelin, A., Calles, O., Greenberg, L., Piccolo, J., and Bergman, E. 2016. Spawn-
ing migration of wild and supplementary stocked landlocked Atlantic
salmon (Salmo salar). River Res. Appl. 32(3): 383–389. doi:10.1002/rra.2870.
Hinch, S.G., and Bratty, J. 2000. Effects of swim speed and activity pattern
on success of adult sockeye salmon migration through an area of diffi-
cult passage. Trans. Am. Fish. Soc. 129(2): 598–606. doi:10.1577/15488659
(2000)129<0598:EOSSAA>2.0.CO;2.
Holbrook, C.M., Zydlewski, J., Gorsky, D., Shepard, S.L., and Kinnison, M.T.
2009. Movements of prespawn adult Atlantic salmon near hydroelectric
damsinthelowerPenobscotriver.Maine.NorthAm.J.Fish.Manage.
29(2): 495–505. doi:10.1577/M08-042.1.
Jokikokko, E. 2002. Migration of wild and reared Atlantic salmon (Salmo
salar L.) in the river Simojoki, Northern Finland. Fish. Res. 58(1): 15–23.
doi:10.1016/S0165-7836(01)00364-2.
Jonsson, B., and Jonsson, N. 2011. Habitats as template for life histories. In Ecol-
ogy of Atlantic salmon and brown trout. Springer, Dordrecht. pp. 256–271.
Jensen, A.J., and Aass, P. 1995. Migration of a fast-growing population of
brown trout (Salmo trutta L.) through a fish ladder in relation to water
flow and water temperature. Regul. Rivers Res. Manage. 10(2–4): 217–228.
doi:10.1002/rrr.3450100216.
Karppinen, P., Mäkinen, T.S., Erkinaro, J., Kostin, V.V., Sadkovskij, R.V.,
Lupandin, A.I., and Kaukoranta, M. 2002. Migratory and route-seeking
behaviour of ascending Atlantic salmon in the regulated River Tuloma.
Hydrobiologia, 483(1/3): 23–30. doi:10.1023/A:1021386319633.
Kieffer,J.D.,andColgan,P.W.1992.Theroleoflearninginfish behaviour.
Rev. Fish Biol. Fish. 2(2): 125–143. doi:10.1007/BF00042881.
Kraabøl, M., and Arnekleiv, J.V. 1998. Registrerte gytelokaliteter for storørret
i Gudbrandsdalslågen og Gausa med sideelver. NTNU Vitenskapsmuseet.
Report Zoological Series 1998-2. [In Norwegian with English abstract.]
Kinnison, M.T., Unwin, M.J., Hendry, A.P., and Quinn, T.P. 2001. Migratory
costs and the evolution of egg size and number in introduced and indige-
nous salmon populations. Evolution, 55(8): 1656–1667. doi:10.1111/j.0014-
3820.2001.tb00685.x.PMID:11580025.
Kinnison, M.T., Unwin, M.J., and Quinn, T.P. 2003. Migratory costs and contem-
porary evolution of reproductive allocation in male chinook salmon. J. Evol.
Biol. 16(6): 1257–1269. doi:10.1046/j.1420-9101.2003.00631.x.PMID:14640417.
Kraabøl, M. 2012. Reproductive and migratory challenges inflicted on mi-
grantbrowntrout(Salmo trutta L.) in a heavily modified river. Doctoral
dissertation. Department of biology. Norwegian University of Science and
Technology.
Laland, K.N., Brown, C., and Krause, J. 2003. Learning in fishes: from three-
second memory to culture. Fish Fish. 4(3): 199–202. doi:10.1046/j.1467-
2979.2003.00124.x.
Larinier, M. 2001. Dams, fish and fisheries: opportunities, challenges and
conflict resolution. FAO Fisheries Technical Paper No. 419, FAO, Rome.
pp. 45–90.
Larinier, M., Travade, F., and Porcher, J.P. 2002. Fishways: biological basis,
design criteria and monitoring. Bulletin Francais de la Peche et de la Pis-
culture, 364:1–208.
Lundqvist, H., Rivinoja, P., Leonardsson, K., and McKinnell, S. 2008. Upstream
passage problems for wild Atlantic salmon (Salmo salar L.) in a regulated
river and its effect on the population. Hydrobiologia, 602(1): 111–127.
doi:10.1007/s10750-008-9282-7.
Mallen-Cooper, M., and Brand, D.A. 2007. Non-salmonids in a salmonid fish-
way: what do 50 years of data tell us about past and future fish passage?
Fish. Manag. Ecol. 14(5): 319–332. doi:10.1111/j.1365-2400.2007.00557.x.
Nilsson, C., Reidy, C.A., Dynesius, M., and Revenga, C. 2005. Fragmentation
and flow regulation of the world’s large river systems. Science, 308(5720):
405–408. doi:10.1126/science.1107887.PMID:15831757.
Noss, R.F., and Daly, K.M. 2006. Incorporating connectivity into broad-scale
conservation planning. In Connectivity Conservation. Cambridge Univer-
sity Press. pp. 587–617. doi:10.1017/CBO9780511754821.026.
Nyqvist, D., Nilsson, P.A., Alenäs, I., Elghagen, J., Hebrand, M., Karlsson, S.,
et al. 2017. Upstream and downstream passage of migrating adult Atlan-
tic salmon: remedial measures improve passage performance at a hydro-
power dam. Ecol. Eng. 102: 331–343. doi:10.1016/j.ecoleng.2017.02.055.
Odling-Smee, L., and Braithwaite, V.A. 2003. The role of learning in fish ori-
entation. Fish Fish. 4(3): 235–246. doi:10.1046/j.1467-2979.2003.00127.x.
Paul, E.S., Harding, E.J., and Mendl, M. 2005. Measuring emotional processes
in animals: the utility of a cognitive approach. Neurosci. Biobehav. Rev.
29(3): 469–491. doi:10.1016/j.neubiorev.2005.01.002.
Peake, S.J., and Farrell, A.P. 2004. Locomotory behaviour and post-exercise
physiology in relation to swimming speed, gait transition and metabo-
lism in free-swimming smallmouth bass (Micropterus dolomieu). J. Exp.
Biol. 207(9): 1563–1575. doi:10.1242/jeb.00927.PMID:15037650.
Piccolo,J.J.,Norrgård,J.R.,Greenberg,L.A.,Schmitz,M.,andBergman,E.
2012. Conservation of endemic landlocked salmonids in regulated rivers:
a case-study from Lake Vänern, Sweden. Fish Fish. 13(4): 418–433.
doi:10.1111/j.1467-2979.2011.00437.x.
Pon, L.B., Hinch, S.G., Cooke, S.J., Patterson, D.A., and Farrell, A.P. 2009.
Physiological, energetic and behavioural correlates of successful fishway pas-
sage of adult sockeye salmon Oncorhynchus nerka in the Seton River, British
Columbia. J. Fish Biol. 74(6): 1323–1336. doi:10.1111/j.1095-8649.2009.02213.x.
Portavella, M., Torres, B., and Salas, C. 2004. Behavioral/systems/cognitive
avoidanceresponseingoldfish: emotional and temporal involvement of
medial and lateral Telencephalic Pallium. J. Neurosci. 24(9): 2335–2342.
doi:10.1523/JNEUROSCI.4930-03.2004.PMID:14999085.
Portz,D.E.,Woodley,C.M.,andCech,J.J.2006.Stress-associatedimpactsof
short-term holding on fishes. Rev. Fish Biol. Fish. 16(2): 125–170.
doi:10.1007/s11160-006-9012-z.
Rivinoja, P., McKinnell, S., and Lundqvist, H. 2001. Hindrances to upstream
migration of Atlantic salmon (Salmo salar) in a northern Swedish river
caused by a hydroelectric power-station. Regul. Rivers Res. Manag. 17(2):
101–115. doi:10.1002/rrr.607.
Roscoe, D.W., and Hinch, S.G. 2010. Effectiveness monitoring of fish passage
facilities: historical trends, geographic patterns and future directions.
Fish Fish. 11(1): 12–33. doi:10.1111/j.1467-2979.2009.00333.x.
Roscoe, D.W., Hinch, S.G., Cooke, S.J., and Patterson, D.A. 2011. Fishway pas-
sage and post-passage mortality of up-river migrating sockeye salmon in
the Seton River, British Columbia. River Res. Appl. 27(6): 693–705.
doi:10.1 002/rra.13 84 .
Stevens, E.D., and Black, E.C. 1966. The effect of intermittent exercise on
carbohydrate metabolism in rainbow trout, Salmo gairdneri.J.Fish.Res.
Board Canada. 23(4): 471–485. doi:10.1139/f66-039.
Thorstad, E.B., Økland, F., and Finstad, B. 2000. Effects of telemetry trans-
mitters on swimming performance of adult Atlantic salmon. J. Fish Biol.
57(2): 531–535. doi:10.1111/j.1095-8649.2000.tb02192.x.
Thorstad, E.B., Økland, F., Kroglund, F., and Jepsen, N. 2003. Upstream
migration of Atlantic salmon at a power station on the River Nidelva.
Southern Norway. Fish. Manage. Ecol. 10(3): 139–146. doi:10.1046/j.1365-
2400.2003.00335.x.
Thorstad, E.B., Økland, F., Aarestrup, K., and Heggberget, T.G. 2008. Factors
affecting the within-river spawning migration of Atlantic salmon, with
emphasis on human impacts. Rev. Fish Biol. Fish. 18(4): 345–371.
doi:10.1007/s11160-007-9076- 4.
Vegar, J.A., and Kraabøl, M. 1996. Migratory behaviour of adult fast-growing
brown trout (Salmo trutta, L.) in relation to water flow in a regulated Nor-
wegian river. Reg. River Res. Manage. 12(1): 39–49. doi:10.1002/(SICI)1099-
1646(199601)12:1<39::AID-RRR375>3.0.CO;2-% 23.
Webb, J. 1990. Behaviour of adult Atlantic salmon ascending the rivers Tay
and Tummel to Pitlochry dam. Scottish Fish. Res. Rep. 48.
Williams, J.G., Armstrong, G., Katopodis, C., Larinier, M., and Travade, F.
2012. Thinking like a fish: a key ingredient for development of effective
fish passage facilities at river obstructions. River Res. Appl. 28(4): 407–
417. doi:10.1002/rra.1551.
Winter, J.D. 1983. Underwater telemetry. In Fisheries techniques. Edited by
L.A. Nielsen and D.L. Johnson. American Fisheries Society, Bethesda. pp. 371–
395.
Young, J.L., Hinch, S.G., Cooke, S.J., Crossin, G.T., Patterson, D.A., Farrell, A.P.,
et al. 2006. Physiological and energetic correlates of en route mortality for
abnormally early migrating adult sockeye salmon (Oncorhynchus nerka)inthe
Thompson River, British Columbia. Can. J. Fish. Aquat. Sci. 63(5): 1067–1077.
doi:10.1139/f06-014.
Yue, S., Moccia, R., and Duncan, I.J. 2004. Investigating fear in domestic rain-
bow trout, Oncorhynchus mykiss, using an avoidance learning task. Appl.
Anim. Behav. Sci. 87(3–4): 343–354. doi:10.1016/J.APPLANIM.2004.01.004.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Hagelin et al. 11
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 192.165.21.4 on 12/18/20
For personal use only.