JAE_550.fm Page 980 Tuesday, November 6, 2001 9:45 AM
Journal of Animal
Ecology 2001
70, 980 – 986
Intraspecific resource partitioning in brown trout:
the temporal distribution of foraging is determined
by social rank
Blackwell Science Ltd
ANDERS ALANÄRÄ*, MARTIN D. BURNS† and NEIL B. METCALFE†
*Department of Aquaculture, Swedish University of Agricultural Sciences, SE-901 83 Umeå, Sweden; and
†Fish Biology Group, Division of Environmental & Evolutionary Biology, Graham Kerr Building, Institute of
Biomedical and Life Sciences, Glasgow University, Glasgow G12 8QQ, UK
Summary
1. Animals can reduce the competition for a limiting resource by temporal segregation,
whereby individuals exploit the resource at different times. However, the pay-offs may
vary predictably over time, and it can be predicted that (a) more dominant competitors
should gain access to resources at the preferred times and (b) the degree of temporal
segregation will vary with the intensity of competition.
2. Here we show experimentally that individual brown trout Salmo trutta (L.) made
sequential use of foraging areas, with dominant individuals feeding mainly at the most
beneficial times of dusk and the early part of the night while more subordinate fish fed
at other times.
3. However, the degree of overlap in foraging times between high-ranking fish was
dependent on energetic demands. At low temperatures (when requirements were low)
the temporal activity patterns of top-ranking fish were synchronized, with foraging concentrated at the preferred times. In contrast, when temperature was raised to increase
energetic requirements, activity patterns showed strong temporal segregation: the
most dominant fish remained predominantly nocturnal, whereas second-ranking fish
became increasingly diurnal.
4. This is the first experimental demonstration of shifts in the daily pattern of activity
caused by varying intensity of intraspecific competition.
Key-words: antipredator behaviour, competition, diel activity rhythms, Salmonid,
Salmo trutta.
Journal of Animal Ecology (2001) 70, 980–986
Introduction
Competition for limiting resources in the natural world
has led to the evolution of diverse adaptations to mitigate its effects. Animals can reduce the impact of competition by partitioning the resources, through habitat,
dietary and temporal segregation (Schoener 1974).
An intriguing example is where different individuals
exploit the same resource but at different times, so that
competitors rarely meet. Such resource partitioning
has received most attention at the level of interspecific
competition (e.g. Cotton 1998; Kronfeld-Schor &
Dayan 1999), but the same processes can also occur
© 2001 British
Ecological Society
Correspondence: Dr Anders Alanärä, Department of Aquaculture, Swedish University of Agricultural Sciences, SE-901
83 Umeå, Sweden. E-mail: [email protected]. Tel:
+ 46 90 786 76 77. Fax: + 46 90 12 37 29.
within species. Such interactions become more complex when requirements for resources change over time,
so leading to shifts in exploitation rates and hence
the strength of competition: at low levels of resource
exploitation the niche overlap between competitors is
predicted to be much broader than when resources are
limiting (Keddy 1989). Here we examine (using trout as
a model) whether individuals of different competitive
status show temporal partitioning in their use of
foraging areas, and whether the degree of temporal
partitioning varies as predicted with temperaturedependent changes in nutritional requirements.
Juvenile brown trout live in streams, where the most
profitable feeding positions are those adjacent to areas
of fast flow and hence high rates of invertebrate drift
(Fausch 1984). However, feeding exposes the fish to
predators (Martell & Dill 1995), and trout often hide
in refuges between feeding bouts. Daytime field
JAE_550.fm Page 981 Tuesday, November 6, 2001 9:45 AM
981
Intraspecific
resource
partitioning in
brown trout
observations indicate that the same feeding position
may be used sequentially by different trout, with more
dominant fish displacing others (Bachmann 1984).
Moreover, individual wild trout appear to differ in their
daily activity patterns (Giroux et al. 2000) and the feeding activity of tank-reared fish shows that some fish
may be predominantly nocturnal while others in the
same group are diurnal (Alanärä & Brännäs 1997;
Brännäs & Alanärä 1997).
The profitability of feeding will vary with time of day.
The availability of drifting food is usually lowest during
the day (Elliott 1967) but the fish’s ability to detect that
food increases with light intensity (Fraser & Metcalfe
1997). Predation risk is greatest, however, during the
day (Metcalfe, Fraser & Burns 1999). Feeding at twilight or night while spending the day hiding in refuges
therefore minimizes the mortality risk per unit of food
obtained (Metcalfe et al. 1999), but may not provide a
sufficiently high daily intake when food demands
increase. Therefore trout and related species have been
observed to be almost exclusively nocturnal at low temperatures but become more day-active as their nutritional needs increase, such as when temperature rises
(Fraser, Metcalfe & Thorpe 1993; Heggenes et al. 1993)
or the food availability or their own nutritional state
decreases (Metcalfe, Fraser & Burns 1998; Metcalfe et
al. 1999). Such changes in the amount of time that must
be spent foraging will influence the temporal partitioning of foraging sites; we can predict that the strength
of this partitioning will increase with the intensity of
foraging.
Here we demonstrate status-dependent intraspecific
segregation of feeding times in trout, with the most
dominant fish feeding at the preferred times of low light
intensity. Moreover, we provide one of the first
intraspecific examples of an increased food demand
causing a temporal shift from overlapping to partitioned resource exploitation among competitors: the
greater the energetic requirements of dominant individuals, the more the subdominants shifted foraging
activity to the less preferred time of day.
Methods
© 2001 British
Ecological Society,
Journal of Animal
Ecology, 70,
980– 986
The study fish (the progeny of wild fish) were reared in
a holding tank under routine husbandry conditions.
The experiment was conducted in a temperaturecontrolled room, using standard rearing tanks (1 m2,
water depth 30 cm) modified to create 12 × 130 cm
stream channels that ended in a 0·5-m2 refuge (Fig. 1).
Each tank was screened off from surrounding light by
black PVC plastic sheeting. Fluorescent lights above
the tanks provided illumination during the day (200 lx
in open stream areas), dawn/dusk (1 lx) and night
(0·1 lx, representative of full moon conditions (Fraser
& Metcalfe 1997)). The light : dark cycle was 11 : 11 h,
plus 1 hour of dawn and dusk.
A pump created a continuous flow of water through
the channel (Fig. 1). Up to five pellets of food were
A
B
Hiding area
Feeding
area
C
D
D
Fig. 1. Experimental tank design, showing (A) water inlet
and screen to prevent fish from entering the foraging area
from upstream, (B) food delivery point, (C) food trap and (D)
PIT-tag antennae recording individual fish passing in or out
of the foraging area. Fish were unable to enter the triangular
area in the centre of the tank.
dropped into a feeding tube by an automatic feeder
at 10-min intervals 24 h a day. The tube released the
pellets 2 cm below the water surface in front of the
water inflow so that the food drifted with the current
as it slowly sank to the tank floor. Uneaten food was
diverted into a drain by a 2-cm-high strip of PVC plastic on the bottom of the tank at the bend of the channel
(Fig. 1). The area between the food delivery tube and
this food trap is referred to as the foraging area.
In order to simulate the dichotomy faced by wild
stream-living salmonids between a safe refuge and a
risky foraging area, we manipulated the degree of concealment offered by the two habitats. Fish were relatively concealed in the refuge, which had dark green
walls and floor and an opaque overhead cover, but were
conspicuous in the foraging area due to its white walls
and mid-grey floor; when given a choice in the absence
of food, trout generally avoid paler substrates and illuminated areas.
Each experimental trial consisted of two phases: the
determination of social status within the group, and the
recording of foraging activity patterns. Prior to the first
phase, fish were taken from the stock tank and selected
visually to create groups of five fish of the same age but
covering a range of sizes. The purpose with this size
range was to promote the development of natural
social hierarchies, since body size in general is a good
indicator of the previous social status within the
population (Metcalfe et al. 1990; Metcalfe, Wright &
Thorpe 1992). They were then anaesthetized, weighed
and given both an identifying dyemark on their fins
and a Passive Integrated Transponder (PIT) tag with a
unique identification code, injected into the body cavity.
The relative social rank of fish within each group was
then assessed by visual observations, using a procedure
based on that described by Metcalfe et al. (1989).
Briefly, the fish were placed in the foraging area and
JAE_550.fm Page 982 Tuesday, November 6, 2001 9:45 AM
982
A. Alanärä,
M.D. Burns &
N.B. Metcalfe
© 2001 British
Ecological Society,
Journal of Animal
Ecology, 70,
980–986
prevented from entering the refuge area by a net barrier. The downstream part of the foraging area was provided with a temporary overhead cover in order to give
shelter to fish not actively foraging. Observations of
social status commenced after the fish had settled in the
tank for about 24 h. The rate of aggression in newly
formed groups of salmonids declines rapidly after
group formation, due to the formation of stable social
relationships (O’Connor, Metcalfe & Taylor 1999), and
in the present study no aggressive acts were observed
between fish. Thus, to quantify status, we used a combination of spatial position and the amount of food
obtained when this was a limiting resource; it has been
shown previously that there is a high degree of concordance in the social rankings produced by these two
methods, and they correlate strongly with measures
based on observed aggression (Metcalfe et al. 1989;
Johnsson & Bjornsson 1974). During observation days
the automatic feeder was switched off. Observations
were carried out four times a day (twice within an hour,
both in the morning and in the afternoon). At the
beginning of each observation spatial positions were
scored as follows. The fish occupying the area nearest
the food delivery point was given a score of 3 points,
whereas fish in the open channel farther downstream
were given 1 point and fish hiding in the shelter 0
points. After the position score of each fish was
recorded, a single pellet of food was released down the
feeding pipe and the fish that obtained it scored 1 point.
The outcome was scored similarly for a further four
pellets, released at 1-minute intervals. This protocol
generated four observations on fish positions and foraging scores for 20 pellets per day. Position and foraging scores were then summed to produce one status
score for that day (maximum value = 32). The fish with
the highest combined score was considered to be of
highest status if it scored at least 20 points more than
the second-ranking fish. If a top-ranking fish was not
apparent after 1 day, the procedure was repeated the
next day. Once a top-ranked fish had been identified, it
was removed from the group and placed in the central
refuge area of the tank (Fig. 1). Observations then continued the next day to determine the next-highest ranking fish, which was then removed and the process
repeated. Differences in status between the two lowest
ranked fish were difficult to determine as they usually
scored very few points even in the absence of the more
dominant individuals. They were therefore given the
same rank, i.e. rank 4.
The cover over the downstream part of the foraging
area was removed, as was the net barrier separating the
foraging area from the central refuge, so that all five fish
were together and able to move between the refuge and
foraging area. The automatic feeder was switched on,
and the fish were left undisturbed for 14 days. Movements between the foraging area and the refuge were
registered using a PIT tag registration system (Brännäs
et al. 1994) consisting of two antennae positioned in
the channel between the foraging area and the refuge
(Fig. 1). The tank floor was made white between the
two antennae to discourage fish from remaining there.
All passages were recorded and stored on computer for
later analysis; the direction of movement was evident
from the sequence in which the two antennae were
triggered.
After 14 days the fish were removed from the tank,
anaesthetized and re-weighed in order to determine
growth rates, which were calculated as daily growth
coefficients: DGC = 100 (final weight1/3 – initial
weight1/3)/days elapsed (Cowey 1992).
In order to investigate the effect of food demand on
diel feeding patterns, we manipulated the temperature
(and hence energetic requirements) while keeping the
food availability constant. Therefore six replicate
groups of five fish each were tested at 7·1 °C (SD = 0·4)
(hereafter referred to as 7 °C), and six groups with different fish were tested at 14·2 °C (SD = 0·3) (hereafter
14 °C). Fish were acclimatized to the experimental
temperatures for at least 1 week prior to the start of the
experiments. Activity patterns over the last 10 days of
each trial were analysed for the number, duration and
timing of trips into the foraging area. Time of day
was categorized into dawn (06.00–06.59 h), morning
(07.00–13.59 h), afternoon (14.00–18.59 h), dusk
(19.00–19.59 h), early night (20.00–24.59 h) and late
night (01.00–06.59 h).
Results
In general the fish spent the majority of their time in the
refuge but made short, relatively frequent visits to the
foraging area. Foraging effort showed marked diel
variation, and the pattern was influenced strongly by
social rank as well as by temperature (Fig. 2). These
activity patterns were analysed by calculating the (arcsin transformed) percentage of time spent by each fish
in the foraging area in each of the six diel periods
(dawn, morning, afternoon, dusk, early and late night);
percentage time foraging was then used in a repeated
measures  with time period as the within-subject
effect and rank and temperature as between-subject
effects. Initial weight varied significantly between dominance ranks with more dominant fish tending to be
larger (Table 1; , 7 °C: F3,25 = 9·94, P < 0·001;
14 °C: F3,25 = 4·11, P = 0·017). However the relationship within a group between rank and size was only
moderate: the Spearman’s rank correlation between
social rank and size rank averaged 0·733 (n = 12
groups). Therefore both rank and initial weight (as a
covariate) were used in the initial analyses, in order to
evaluate their independent effects on foraging activity
both within and between treatments (i.e. temperatures). These results showed that, while there were
strong effects of social rank (see below), neither initial
weight nor its interaction terms had significant effects
on patterns of activity, and so they were dropped from
the analysis in order to reduce the number of parameters and so clarify the patterns among the data.
JAE_550.fm Page 983 Tuesday, November 6, 2001 9:45 AM
983
Intraspecific
resource
partitioning in
brown trout
Fig. 2. Diel distribution of mean number of trips into (dots) and time spent in (bars + SD) the foraging area for brown trout of
different social rank (1– 4) at (a) 7 °C and (b) 14 °C. Shaded bars indicate hours during night, hatched bars dawn and dusk, and
unfilled bars daytime; n = 6 fish for all ranks except rank 4 where n = 12.
Table 1. Mean values (± SD) for initial weight and daily growth coefficient (DGC) for brown trout of different social rank (1 =
most dominant) held at 7 and 14 °C, respectively
Social rank
1
2
3
4
Temperature: 7 °C
Number of fish
Initial weight (g)
DGC
5†
43·8 ± 5·0
1·21 ± 0·14
6
37·1 ± 6·2
0·88 ± 0·50
6
34·1 ± 8·2
0·39 ± 0·42
12
26·9 ± 5·4
0·50 ± 0·51
Temperature: 14 °C
Number of fish
Initial weight (g)
DGC
6
29·6 ± 2·9
3·13 ± 0·42
6
23·0 ± 5·5
1·78 ± 1·04
6
22·4 ± 6·1
0·87 ± 0·89
12
18·2 ± 8·1
0·49 ± 1·00
†Note that one fish died during the course of the experiment.
© 2001 British
Ecological Society,
Journal of Animal
Ecology, 70,
980– 986
There was a highly significant effect of dominance
rank, with more dominant fish spending a greater percentage of time in the foraging area (Fig. 2, F3,50 = 8·44,
P < 0·001). There was no overall effect of temperature
(F1,50 = 0·019, NS), nor any interaction between rank
and temperature (F3,50 = 0·81, NS). The usage of the
foraging area varied significantly across the 24 h (effect
of time of day: F5,46 = 7·47, P < 0·001). More interestingly, this diel pattern was strongly dependent on the
dominance rank of the fish (rank × time interaction:
F5,48 = 26·23, P < 0·001). Activity in top-ranked fish
peaked in the period from dusk (when fish made many
JAE_550.fm Page 984 Tuesday, November 6, 2001 9:45 AM
984
A. Alanärä,
M.D. Burns &
N.B. Metcalfe
Fig. 2. continued
© 2001 British
Ecological Society,
Journal of Animal
Ecology, 70,
980–986
Mean minutes per hour,
subdominants
24
(a)
20
16
12
8
0
10
20
30
40
50
24
Mean minutes per hour,
subdominants
short foraging trips) until midnight, then declined to
reach minimal levels from dawn until midday (Fig. 2).
Subdominant (second-ranked) fish showed a contrasting pattern, with much less diel variation and significant amounts of daytime activity. Movement rates
peaked at both dawn and dusk, and time in the foraging
area was generally greatest from late night to midday.
Foraging activity by the lower-ranked fish was much
lower and showed little diel variation apart from a
tendency towards greater activity in the late part of
the night.
There was a significant effect of temperature on
the diel pattern of activity, with greater use of the
foraging area during the day at the higher temperature
(temperature × time interaction: F5,46 = 2·65, P = 0·035);
however, this was least pronounced in the dominant
fish (temperature × time × rank interaction: F5,48 = 3·42,
P = 0·01). This was examined further by comparing
the relative use of the foraging area by dominant and
subdominant fish at each hour of the day. At 7 °C there
was a positive correlation between the mean proportion of time that dominants and subdominants used
the foraging area (Fig. 3a; Spearman rank correlation
coefficient rsp = 0·656, n = 24–h periods, P = 0·001),
(b)
20
16
12
8
0
10
20
30
40
50
Mean minutes per hour, dominants
Fig. 3. Relationship between the mean proportion of time
that dominant (rank 1) and subdominant (rank 2) fish spent
in the foraging area each hour of the day. Black symbols
indicate hours at night, grey symbols dawn and dusk, and
unfilled symbols daytime. (a) At 7 °C the fish foraged at
similar times of day, whereas (b) at 14 °C they showed
temporal segregation. See text for statistical analysis.
JAE_550.fm Page 985 Tuesday, November 6, 2001 9:45 AM
985
Intraspecific
resource
partitioning in
brown trout
although dominants showed greater avoidance of the
feeding area during the day than did subdominants.
However, this pattern was changed at the higher temperature, when there was a significant negative correlation
between dominants and subdominants in their relative usage of the foraging area (Fig. 3b; rsp = – 0·819,
n = 24, P < 0·001). This indicates that they partitioned the resource on a temporal basis: dominant fish
used the foraging area more at dusk and night, while
subdominants used it at dawn and during the day.
These trends are also shown by the relative amount
of time that fish of different rank were present in the
foraging area by day vs. by night. The percentage diurnal activity was calculated as 100Td /(Td + Tn ), where Td
is the mean percentage of time that an individual fish
spent in the foraging area during the day and Tn is the
corresponding figure for the night (dawn and dusk
periods were ignored in this analysis). A value of 50%
indicates that the fish were equally active by day as by
night, while 0% and 100% indicate exclusively nocturnal and diurnal activity, respectively. Dominant fish
were predominantly nocturnal irrespective of temperature, the percentage diurnal activity being 14·3 ± 5·9%
at 7 °C and 18·7 ± 4·4% at 14 °C. In contrast, subdominants became increasingly diurnal at higher temperatures (37·4 ± 9·4% at 7 °C and 70·7 ± 12·5% at 14 °C).
There was therefore a significant difference between
dominants and subdominants in the extent to which
they were diurnal (two-way  on arcsin-transformed
percentage diurnal; effect of rank: F1,19 = 14·13,
P = 0·001; effect of temperature: F1,19 = 4·89, P = 0·039).
Growth rates were analysed using a two-way 
with rank and temperature as factors and initial weight
as a covariate. The strongest effect was that of social
rank (Table 1; F3,50 = 11·85, P < 0·001), with higherranking fish growing much faster, presumably as a
result of their increased time spent foraging. While
there was a significant overall effect of temperature
(F1,50 = 5·18, P = 0·027), the extent to which the higher
temperature produced faster growth was dependent
on rank, as dominant fish grew much faster at 14 °C
whereas the growth of the lowest-ranking fish was
unchanged (Table 1; status × temperature interaction,
F3,50 = 4·10, P = 0·011). Initial weight had no effect on
growth rates once social status was taken into account
(F1,50 = 1·01, NS).
Discussion
© 2001 British
Ecological Society,
Journal of Animal
Ecology, 70,
980– 986
Usage of the foraging area was clearly non-random,
with consistent differences between social ranks of fish
in their foraging intensity over the course of the day.
Dominant fish foraged most from dusk until dawn,
with the intensity of foraging decreasing through the
night. They thus foraged at the time that is thought to
minimize the predation risk incurred per unit of food
obtained (Metcalfe et al. 1999), and would have been
able to feed reasonably effectively given the ‘full moon’
condition (Fraser & Metcalfe 1997). Their greater
foraging intensity in the early rather than in the late
stages of the night is presumably related to the fact that
they started the night with a relatively empty stomach.
It was noticeable that at the higher temperature, when
the time required to empty the gut was shorter than
the daylength (Elliott 1975), their foraging activity
increased as dusk approached, a trend also found by
Brännäs & Alanärä (1997) and Metcalfe et al. (1999).
The higher temperature used in this experiment is
almost exactly that at which brown trout can achieve
their maximum growth rate, provided that food is
unlimited; however, their maintenance costs and hence
food demands are also greater than at the lower temperature (Elliott 1976). Both dominant and subdominant fish were evidently able to increase their food
intake at 14 °C sufficiently to allow an increased
growth rate, whereas more subordinate fish showed no
increase in growth. However, an increase in exploitation rate by superior competitors has effects on those of
poorer competitive ability. At the lower temperature,
subdominants were able to make extensive use of the
foraging area at night, albeit by exploiting the second
half of the night when dominant fish were relatively
inactive. The better competitors were thus able to avoid
feeding during the day, so on a broad temporal scale
their activities were synchronized. However, the more
prolonged nocturnal foraging by dominant fish at the
higher temperature was associated with a shift by subdominants towards diurnal foraging, presumably
because they were unable to gain access to the foraging
area at the preferred times. As predicted, this resulted
in a very strong temporal segregation of foraging times.
As a result, subdominant fish were able to increase their
growth rate at the higher temperature despite the
increased food requirements of dominants, but at the
expense of an increase in their perceived predation
risk (through greater daytime foraging; Metcalfe
et al. 1999).
It might be argued that the differences in diel activity
patterns were due not to competition but to contrasting endogenous rhythms, as has been documented in
other species of fish (Sánchez-Vázquez, Madrid &
Zamora 1995; Brännäs & Alanärä 1997; SánchezVázquez et al.). However, this appears to be unlikely
since the diel patterns of foraging activity were correlated with dominance status, and subdominants were
apparently forced to change their activity schedule
with an increase in food demand. Moreover, the patterns were not a consequence of body size-dependent
risk taking (Johnsson 1993; Damsgård & Dill 1998;
Metcalfe et al. 1998), since body size did not explain
any of the variation in activity patterns if status was
taken into account. Instead it appears that changing
intensities of competition for resources causes animals
of differing rank to alter their timing of daily activity.
This has been documented before at an interspecific
level (Craig & Douglas 1984) but not, to our knowledge, within a species. Pronounced individual variation
in daily foraging patterns, observed in trout (Bachmann
JAE_550.fm Page 986 Tuesday, November 6, 2001 9:45 AM
986
A. Alanärä,
M.D. Burns &
N.B. Metcalfe
1984; Giroux et al. 2000) and possibly widespread in
animals, may be the result of a complex trade-off
between ease of access to those resources and diel variation in the risks of foraging.
Acknowledgements
A Postdoctoral Fellowship from the Swedish Council
for Forestry and Agricultural Research (SJFR) enabled
A.A. to conduct the study in Glasgow. We thank
John Laurie for fish husbandry and BOCM Pauls for
provision of fish food.
References
Alanärä, A. & Brännäs, E. (1997) Diurnal and nocturnal
feeding activity in arctic char and rainbow trout. Canadian
Journal of Fisheries and Aquatic Sciences, 54, 2894 –
2900.
Bachmann, R.A. (1984) Foraging behavior of free-ranging
wild and hatchery brown trout in a stream. Transactions of
the American Fisheries Society, 113, 1– 32.
Brännäs, E. & Alanärä, A. (1997) Is diel dualism in feeding
activity influenced by competition between individuals?
Canadian Journal of Zoology, 75, 661– 669.
Brännäs, E., Lundqvist, H., Prentice, E., Schmitz, M.,
Brännäs, K. & Wiklund, B.S. (1994) Use of the passive
integrated transponder (PIT) in a fish identification and
monitoring system for fish behavioral studies. Transactions
of the American Fisheries Society, 123, 395 – 401.
Cotton, P.A. (1998) Temporal partitioning of a floral resource
by territorial hummingbirds. Ibis, 140, 647– 653.
Cowey, C.B. (1992) Nutrition: estimating requirements of
rainbow trout. Aquaculture, 100, 177–189.
Craig, J.L. & Douglas, M.E. (1984) Temporal partitioning of
nectar resources in relation to competitive asymmetries.
Animal Behaviour, 32, 624 – 625.
Damsgård, B. & Dill, L.M. (1998) Risk-taking behavior in
weight-compensating coho salmon, Oncorhynchus kisutch.
Behavioral Ecology, 9, 26 – 32.
Elliott, J.M. (1967) Invertebrate drift in a Dartmoor stream.
Archiv für Hydrobiologie, 63, 202 – 237.
Elliott, J.M. (1975) Number of meals in a day, maximum
weight of food consumed in a day and maximum rate of
feeding for brown trout, Salmo trutta L. Freshwater Biology,
5, 287– 303.
Elliott, J.M. (1976) The energetics of feeding, metabolism and
growth of brown trout (Salmo trutta L.) in relation to body
weight, water temperature and ration size. Journal of Animal
Ecology, 45, 923 – 948.
Fausch, K.D. (1984) Profitable stream positions for salmonids:
relating specific growth rate to net energy gain. Canadian
Journal of Zoology, 62, 441– 451.
Fraser, N.H.C. & Metcalfe, N.B. (1997) The costs of becoming
nocturnal: feeding efficiency in relation to light intensity
in juvenile Atlantic salmon. Functional Ecology, 11, 385 –
391.
Fraser, N.H.C., Metcalfe, N.B. & Thorpe, J.E. (1993) Temperature-dependent switch between diurnal and nocturnal
© 2001 British
Ecological Society,
Journal of Animal
Ecology, 70,
980–986
foraging in salmon. Proceedings of the Royal Society of
London B, 252, 135 –139.
Giroux, F., Ovidio, M., Philippart, J.-C. & Baras, E. (2000)
Relationship between the drift of macroinvertebrates and
the activity of brown trout in a small stream. Journal of Fish
Biology, 56, 1248 –1257.
Heggenes, J., Krog, O.M.W., Lindås, O.R., Dokk, J.G. &
Bremnes, T. (1993) Homeostatic behavioural responses in a
changing environment: brown trout (Salmo trutta) become
nocturnal during winter. Journal of Animal Ecology, 62,
295 – 308.
Johnsson, J.I. (1993) Big and brave: size selection affects
foraging under risk of predation in juvenile rainbow trout,
Oncorhynchus mykiss. Animal Behaviour, 45, 1219–1225.
Johnsson, J.I. & Björnsson, B.Th (1994) Growth hormone
increases growth rate, appetite and dominance in juvenile
rainbow trout, Oncorhynchus mykiss. Animal Behaviour,
48, 177–186.
Keddy, P.A. (1989) Competition. Chapman & Hall, London.
Kronfeld-Schor, N. & Dayan, T. (1999) The dietary basis for
temporal partitioning: food habits of coexisting Acomys
species. Oecologia, 121, 123 –128.
Martel, G. & Dill, L.M. (1995) Influence of movement by
coho salmon (Oncorhynchus kisutch) parr on their detection
by common mergansers (Mergus merganser). Ethology, 99,
139 –149.
Metcalfe, N.B., Fraser, N.H.C. & Burns, M.D. (1998) Statedependent shifts between nocturnal and diurnal foraging
in salmon. Proceedings of the Royal Society of London B,
265, 1503 –1507.
Metcalfe, N.B., Fraser, N.H.C. & Burns, M.D. (1999) Food
availability and the nocturnal vs. diurnal foraging trade-off
in juvenile salmon. Journal of Animal Ecology, 68, 371–381.
Metcalfe, N.B., Huntingford, F.A., Graham, W.D. &
Thorpe, J.E. (1989) Early social status and the development
of life-history strategies in Atlantic salmon. Proceedings of
the Royal Society of London B, 236, 7–19.
Metcalfe, N.B., Huntingford, F.A., Thorpe, J.E. & Adams,
C.E. (1990) The effects of social status on life-history
variation in juvenile salmon. Canadian Journal of Zoology,
68, 2630 – 2636.
Metcalfe, N.B., Wright, P.J. & Thorpe, J.E. (1992) Relationships between social status, otolith size at first feeding and
subsequent growth in Atlantic salmon (Salmo salar).
Journal of Animal Ecology, 61, 585–589.
O’Connor, K.I., Metcalfe, N.B. & Taylor, A.C. (1999) Does
darkening signal submission in territorial contests between
juvenile Atlantic salmon, Salmo salar? Animal Behaviour,
58, 1269 –1276.
Sánchez-Vázquez, F.J., Azzaydi, M., Martínez, M., Zamora, S.
& Madrid, J.A. (1998) Annual rhythms of demand-feeding
activity in sea bass: evidence of a seasonal phase inversion
of the diel feeding pattern. Chronobiology International, 15,
607 – 622.
Sánchez-Vázquez, F.J., Madrid, J.A. & Zamora, S. (1995)
Circadian rhythms of feeding activity in sea bass, Dicentrarchus labrax L. dual phasing capacity of diel demand-feeding
pattern. Journal of Biological Rhythms, 10, 256–266.
Schoener, T.W. (1974) Resource partitioning in ecological
communities. Science, 185, 27–39.
Received 2 March 2001; accepted 18 June 2001