Parentally acquired differences in resource acquisition ability between brown trout from alternative life history parentage

Abstract

Dominance hierarchies, where they exist, affect individual food acquisition ability and fitness, both of which have the potential to influence life history pathways. Juvenile salmonids exhibit clear dominance hierarchies in early life. As one of the drivers for the adoption of alternative life histories in salmonids is the relative rate of resource acquisition, there is potential for juvenile behaviour to influence the subsequent life history strategy of the individual. Lacustrine brown trout, Salmo trutta, exhibit a multitude of life histories which includes among others the piscivorous (ferox) life history where individuals grow to large size and have delayed maturity and benthivorous and pelagic life histories where individuals grow to much smaller sizes, however mature earlier. Using a number of observable characteristics of dominance, this study compared differences in behaviour between size-matched pairs of progeny, reared under common garden conditions which are derived from alternative, co-existing life history strategy parents. We found that first-generation progeny of ferox trout were more aggressive, acquired more food, had lighter skin pigmentation and held more desirable positions than the progeny of benthivorous brown trout in an experimental stream system. Ferox trout progeny were dominant over benthivorous brown trout progeny in 90% of trials in dyadic contests. Given such clear differences in dominance, this study indicates that parentally acquired dominance-related differences, passed through either, or both, of genetic and nongenetic (e.g. maternal effects) means, are likely a contributing factor to the continued maintenance of distinct life history strategies of brown trout.

Full text

Ecol Freshw Fish 2016; 1–8 wileyonlinelibrary.com/journal/e
|
1
© 2016 John Wiley & Sons A/S.
Published by John Wiley & Sons Ltd
Accepted: 5 November 2016
DOI: 10.1111/e.12323
ORIGINAL ARTICLE
Parentally acquired dierences in resource acquision ability
between brown trout from alternave life history parentage
Marn R. Hughes1 | Travis E. Van Leeuwen1 | Peter D. Cunningham2 | Colin E. Adams1
1Scosh Centre for Ecology and the Natural
Environment, University of Glasgow, Glasgow,
UK
2Wester Ross Fisheries Trust, Harbour Centre,
Gairloch, UK
Correspondence
Marn R. Hughes, Scosh Centre for Ecology
and the Natural Environment, University of
Glasgow, Glasgow, UK.
Email: m.hughes.4@research.gla.ac.uk
Funding informaon
European Union’s INTERREG IVA Programme.
Abstract
Dominance hierarchies, where they exist, aect individual food acquision ability and
tness, both of which have the potenal to inuence life history pathways. Juvenile
salmonids exhibit clear dominance hierarchies in early life. As one of the drivers for the
adopon of alternave life histories in salmonids is the relave rate of resource acqui-
sion, there is potenal for juvenile behaviour to inuence the subsequent life history
strategy of the individual. Lacustrine brown trout, Salmo trua, exhibit a multude of
life histories which includes among others the piscivorous (ferox) life history where
individuals grow to large size and have delayed maturity and benthivorous and pelagic
life histories where individuals grow to much smaller sizes, however mature earlier.
Using a number of observable characteriscs of dominance, this study compared dif-
ferences in behaviour between size- matched pairs of progeny, reared under common
garden condions which are derived from alternave, co- exisng life history strategy
parents. We found that rst- generaon progeny of ferox trout were more aggressive,
acquired more food, had lighter skin pigmentaon and held more desirable posions
than the progeny of benthivorous brown trout in an experimental stream system.
Ferox trout progeny were dominant over benthivorous brown trout progeny in 90% of
trials in dyadic contests. Given such clear dierences in dominance, this study indi-
cates that parentally acquired dominance- related dierences, passed through either,
or both, of genec and nongenec (e.g. maternal eects) means, are likely a contribut-
ing factor to the connued maintenance of disnct life history strategies of brown
trout.
KEYWORDS
behaviour, dominance hierarchies, ferox trout, life history strategy, Salmo trua
1 | INTRODUCTION
Individual behaviour can determine mang and reproducve success
(Cowlishaw & Dunbar, 1991; Dewsbury, 1982), social status (Gilmour,
DiBasta, & Thomas, 2005), and the quality of resources (Nakano,
1995) and territories acquired (Fox, Rose, & Myers, 1981; Wells,
1977). Thus, an animal’s individual behaviour plays a signicant role in
successful development, longevity and overall tness. Aggression and
territoriality are commonly displayed in animals where individuals con-
test a limited resource in an ecosystem, with successful individuals fre-
quently being the most dominant (Ellis, 1995). Such contests can also
inuence the social rank of animals and frequently result in hierarchies
within a populaon. Such dominance hierarchies are found across taxa
including birds (Dingemanse & de Goede, 2004; Ens & Goss- Custard,
1984), sh (Abbo, Dunbrack, & Orr, 1985; Metcalfe, Hunngford,
Graham, & Thorpe, 1989), mammals (Creel, 2001; Seyfarth, 1976) and

2
|
HUGHES ET AL.
invertebrates (Cole, 1981; Röseler, Röseler, Strambi, & Augier, 1984)
with more dominant individuals generally beer at acquiring resources
than subordinate individuals. In birds, dominant American redstarts,
Setophaga rucilla (L.), regularly exclude subordinates from high-
quality habitats and are therefore generally a larger size as a result of
the increased food acquision (Marra & Holmes, 2001). In mammals,
dominance rank inuences both food access and breeding success.
For example, in Red deer, Cervus elaphus (L.), higher ranking individuals
have greater reproducve success and acquire higher quality resources
than subordinates (Appleby, 1980; Cluon- Brock, Albon, & Guinness,
1986). Among sh, compeon for feeding territories in stream dwell-
ing juvenile salmonids has proven a useful model in the study of social
dominance (Adams, Hunngford, Turnbull, & Beae, 1998; Burton,
Hoogenboom, Armstrong, Groothuis, & Metcalfe, 2011; Cus, Adams,
& Campbell, 2001; Höjesjö, Armstrong, & Griths, 2005; Hunngford,
Metcalfe, Thorpe, Graham, & Adams, 1990; Metcalfe et al., 1989;
Hunngford, 2004; Preston, Taylor, Adams, & Migaud, 2014; Van
Leeuwen, Hughes, Dodd, Adams, & Metcalfe, 2015).
In juvenile Atlanc salmon, Salmo salar (L.), aggression and dom-
inance correlate posively with territory quality and food acquisi-
on, with dominant individuals more likely to achieve higher growth
and faster development. This has, in turn, been shown to inuence
the life history of the individual (Metcalfe, 1998). For example,
dominance status has been shown to inuence the age at which
juvenile Atlanc salmon undergo smolng (the physiological and
morphological preparaon for salmonids to enter sea water) and
thus migrate into marine waters, with dominant individuals migrat-
ing at a younger age than subordinates (Metcalfe, 1998; Metcalfe
et al., 1989).
Juvenile brown trout, Salmo trua (L.), display similar dominance
hierarchies to those displayed in the closely related, Atlanc salmon
(Harwood, Armstrong, Griths, & Metcalfe, 2002). They also exhibit
a multude of life history strategies with some individuals remaining
in natal streams their enre life (river residency), while others migrate
to lakes (aduvial potamodromy) or into marine waters (anadromy).
Life history strategy is partly dependent on the life history of the par-
ent (through genec and nongenec maternal eects; Van Leeuwen
et al., 2015) and the relave rates of resource acquision (see review
by Dodson, Aubin- Horth, Thériault, & Páez, 2013), which is likely
dependent on the behaviour of the individual. One understudied
life history of S. trua is the ferox life history paern. Ferox trout
manifest as lacustrine dwelling, piscivorous trout (Grey, 2001; Grey,
Thackeray, Jones, & Shine, 2002) which grow to large size (Mangel &
Abrahams, 2001), exhibit delayed maturaon (Campbell, 1971) and
increased longevity (Mangel & Abrahams, 2001). Considered one of
the many life history types adopted by the highly variable S. trua spe-
cies complex, ferox trout are associated with large, deep oligotrophic
lakes and the presence of Arcc charr Salvelinus alpinus L. (Campbell,
1979; Greer, 1995; Hughes, Dodd, Maitland, & Adams, 2016). Ferox
trout have also been described as reproducvely isolated and genet-
ically disnct from brown trout expressing other life history traits in
sympatry in some lakes in Scotland and Ireland (Duguid, Ferguson, &
Prodöhl, 2006; Ferguson & Mason, 1981; Ferguson & Taggart, 1991;
McVeigh, Hynes, & Ferguson, 1995; Prodöhl, Taggart, & Ferguson,
1992). As a result, ferox trout have been variously classied as ei-
ther one of the many adopted life history types or as a disnct spe-
cies Salmo ferox Jardine, 1835; a nomenclature recognised by the
Internaonal Union for the Conservaon of Nature (IUCN; Freyhof
& Koelat, 2008).
Given that the adopon of piscivory in sh is oen limited by an in-
dividual’s gape size (Mielbach & Persson, 1998; Persson, Andersson,
Wahlström, & Eklöv, 1996), increasing growth through food acquisi-
on to reach a crical size threshold at which they can access sh
prey would benet piscivorous species by enabling them to exploit
larger prey items (and thus sh prey) at an earlier developmental stage.
Therefore, given the advantages of dominance rank on food acquisi-
on in juvenile salmonids, we hypothesise that if the adopon of a
life history pathway is parentally derived, either genecally or through
maternal eects, then juvenile progeny of ferox trout will be more
dominant than progeny of brown trout, adopng the more common
foraging strategy of lacustrine trout exploing macrobenthic inverte-
brates (hereaer benthivores).
To test for dierences in dominance- related traits between life his-
tories, we reared progeny of sympatric ferox and benthivorous brown
trout in the laboratory under common garden condions and com-
pared commonly used indicators of dominance behaviour (aggression,
skin colour, spaal posion and food acquision) in size- matched, dy-
adic contests in an experimental stream.
2 | METHODS
2.1 | Broodstock collecon
Broodstock (three ferox trout females and three ferox trout males,
three benthivorous brown trout females and three benthivorous
brown trout males) were captured using fyke nets and electroshing
between 1 October and 12 November 2013 from two tributaries
in the Loch Maree catchment, Scotland. Reciprocal hybrid crosses
between a male ferox and female benthivorous brown trout and
a female ferox trout with a male benthivorous brown trout were
made. However, due to high mortality of hybrids during early
development, they could not be used for this experiment. For
future studies, the high mortality of hybrid crosses should be noted
and may infer a level of hybrid inviability. Due to the rarity of ferox
trout (Duguid et al., 2006), the likelihood of collecng ripe females
during spawning me is low. At capture, sh were classied as
ferox trout or benthivorous brown trout based on size: ferox trout
(40–80 cm); benthivorous brown trout (20–35 cm; Campbell, 1979)
and prior knowledge from local angling groups about spawning lo-
caons. Classicaon of the two life history types was conrmed
by subsequent stable isotope analysis of egg samples. Mature sh
were transported to holding tanks on the Coulin Estate, Kinlochewe,
Scotland, where they were held in two large, 2,000- L tanks supplied
with river water and assessed daily for ripeness. On 14 November
2013, all sh were anaesthesed, bloed dry, and their eggs or
sperm extruded by abdominal massage. Eggs were ferlised by

|
3
HUGHES ET AL.
randomly selected males of the same life history type to create two
full sibling families of each life history.
2.2 | Egg rearing, hatching and sh husbandry
Ferlised eggs were transported to the Scosh Centre for Ecology
and the Natural Environment (SCENE), Loch Lomond, Scotland. Each
family was reared separately in mesh baskets held in clear plasc tanks
(50 × 30 × 15 cm) in a temperature controlled environment cham-
ber (temp 7.4 ± 0.1°C) using water on a paral recirculaon system.
Water was pumped directly from Loch Lomond before being chan-
nelled through a free- standing lter unit and a single in- line UV steri-
liser. Eggs were held in complete darkness unl hatching. Eggs were
examined daily, with any dead eggs carefully removed. Egg size (two-
dimensional surface area [mm2]) was measured using image soware
(ImageJ) from photographs of eggs taken 48 hours aer ferlisaon.
The ming of emergence of successive developmental stages, eye pig-
mentaon, hatch and yolk absorpon, was recorded as the number of
degree days (DD; calculated as the sum of the daily mean water tem-
perature each day) since ferlisaon. Ferox trout eggs hatched from 5
January to 8 January 2014, with eggs from benthivorous brown trout
hatching from 7 January to 11 January 2014. Following hatching,
alevins were raised under an ambient photoperiod. Aer complete
yolk absorpon, fry were fed a standard commercial salmon pellet
(Biomar, Aarhus, Denmark) at approximately 3% body wt. day−1. On 5
April 2014, sh were moved to larger 175- L radial ow circular tanks.
At this stage, families were mixed to create two groups of equal den-
sity (160 sh in each of the two tanks) for each life history.
2.3 | Stable isotope analysis
Stable isotope analysis was conducted to conrm the dierent forag-
ing strategies of the broodstock used (Jensen et al., 2012). Previous
stable isotope analysis on ferox trout in Loch Ness indicated ferox
trout would have a signicantly elevated δ15N signature compared to
benthivorous brown trout feeding on zooplankton or macroinverte-
brates (Grey et al., 2002).
Eggs (n = 4) from each family were randomly selected from each
batch during stripping (total n = 16) and dried for 96 hr at 48°C in a
drying oven. The dried ssue was ground to a ne powder using a
pestle and mortar. Approximately 50% of each sample was then lipid
extracted as follows: 15 mg of ground ssue was soaked in a 2:1
(by volume) chloroform:methanol solvent mixture. Aer 20 min, the
sample was centrifuged (3,000 rpm for 5 min), the supernatant dis-
carded and the process was repeated unl the solvent ran clear. The
lipid extracted samples were then dried for a further 96 hr at 48°C
in a drying oven. Nonlipid extracted and lipid extracted samples
(n = 32) were measured (7–9 mg) into n capsules (standard weight
5 × 3.5 mm). Carbon (δ13C) and nitrogen (δ15N) stable isotope raos
were determined by connuous ow isotope rao mass spectrome-
try (CF- IRMS), using a Costech ECS 4010 elemental analyser coupled
to a ThermoFisher Scienc Delta XP- Plus IRMS at the NERC Life
Sciences Mass Spectrometry Facility.
2.4 | Behavioural trials
Experimental sh were introduced into one of een compartments
(60 × 60 × 60 cm) of an arcial ume channel at the Scosh Centre
for Ecology and the Natural Environment, Rowardennan, Scotland.
Each compartment was paroned by plasc mesh mounted to
a wooden frame. Homogeneous substrate (gravel/pebble) was
distributed throughout each compartment with a single (10 × 5 cm)
rock located in the centre of the arena. The rock was used as an
indicator of opmal habitat for the experimental sh as juvenile salmo-
nids have been shown to hold central posions in streams behind such
structures (Metcalfe, Valdimarsson, & Morgan, 2003). A 30- cm tube
terminated immediately upstream of this central posion into which
food pellets were introduced (so increasing the likelihood of compe-
on between the two sh). The observaon area, located in the centre
of the ume, allowed observers to view the sh behind glass without
interfering with the shes natural behaviour (Figure 1).
Dyad behavioural trials were conducted using one individual of ferox
origin and one of benthivorous brown trout origin. As dominance in
salmonids has been shown to be signicantly aected by body size
(Hunngford et al., 1990), sh pairs were size matched by length to
the nearest mm to reduce any size eects and potenally any residual
maternal nutrional eects. One sh from each pair was randomly
marked with alcian blue on the dorsal n to allow disncon be-
tween individuals in each compartment to be made. Thirty pairwise
trials were conducted over a three- week period. Once introduced to
the arena, the sh pair were le for 48 hr to acclimate. Food pellets
were introduced ve mes daily throughout this acclimaon period
via feeding tubes to further accentuate the opmal spaal posion
in each compartment. Following acclimaon, sh were observed four
mes daily for 6 min at 09:00, 11:00, 13:00, 15:00 h over 2 days.
Four behavioural characteriscs of dominance in salmonid sh were
measured: overt aggression rate, sh colouraon, spaal posion and
food acquision (Adams, Hunngford, Turnbull, Arno, & Bell, 2000;
Adams et al., 1998; Kilsen et al., 2009; Metcalfe et al., 1989; Nicieza
& Metcalfe, 1999).
Aggressive interacons, sh colour and spaal posion were
scored during an inial 3- minute period during each observaon.
Aer this 3- minute period, a single food pellet was introduced to the
FIGURE1 Schematic diagram of the artificial stream used for
fish observations. Trial arenas (TA), feeding tubes (FT) and rocks (R)
used for indicator of prime habitat, propeller (P), water flow (WF) and
central observation area (OA)

4
|
HUGHES ET AL.
chamber and the acquision of this food item by either sh recorded.
A second 3- minute observaon followed the introducon of the food
pellet during which aggression, sh colour and spaal posion were
recorded. Fish displaying aggressive behaviour during observaons
were scored (+1) for each aggressive display. Five common character-
isc overt aggressive behaviours were recorded during observaons:
n nips, chases, mouth gapes, n displays and sh displacements
(Adams et al., 1998, 2000; Kilsen et al., 2009; Metcalfe et al., 1989).
As body colouraon is a well- known indicator of stress in salmo-
nids (Kilsen et al., 2009), the ank colour of each sh was recorded
at the beginning and end of each observaon. As subordinates tend to
be darker in colouraon, sh with dark bodies were scored negavely
(−1), and those with lighter colouraon scored posively (+1), interme-
diate colouraon received a score of (0).
The behavioural arena was marked into three equal- sized units in
the x, y and z dimensions, using marbles on the substrate and markings
on the observer glass, to indicate paron boundaries. Thus, a total
of 27 cuboids within each chamber were dened and each allocated
a score according to their proximity to the central opmal locaon.
Cuboids immediately below the feeding tube above the substrate were
classed as “opmal” locaons and were given a posive score (+1).
Cuboids on the periphery of “opmal” locaons were neutrally scored
(0) with cuboids located in the corners of the arena being scored neg-
avely (−1). In addion to this, sh observed lying on the substrate
or mesh paron received an addional negave score of (−1). If sh
were highly acve during observaons, every visited cuboid score was
recorded and an average calculated to provide a single score for each
sh.
Food acquision following pellet introducon was scored as fol-
lows. If a sh made no aempt at acquiring the food pellet, it scored
0; if a sh aempted to acquire the food pellet but was unsuccessful,
it scored +1, a sh that was successful in acquiring the food pellet
scored +2.
2.5 | Stascal analysis
Dierences in egg size and the ming of developmental milestones
between ospring of each life history parentage were tested using
ANOVA.
To test for dierences in mean δ13C and δ15N raos between ferox
trout and benthivorous trout eggs, a Welch’s t- test was conducted on
both lipid extracted and nonlipid extracted values.
To test whether the outcome of the number of winning contests
between ferox trout and benthivorous brown trout was equally distrib-
uted, a chi- square goodness- of- t test was conducted. Normality and
homogeneity of data were tested using the Shapiro–Wilk test and the
Anderson–Darling test respecvely. Due to a failure of normality, non-
parametric tests were used in subsequent analysis. Correlaons be-
tween the four variables (aggression, skin colour, spaal posion and
food acquision) were described using a nonparametric Spearman’s
rank correlaon test, before being summarised with a principal com-
ponent analysis (PCA). An overall dominance score was obtained by
extracng PC scores for each sh from PC1 which weighted posively
for aggression, colour, spaal posion and food acquision (Table 3).
To test univariate aggression rates, food acquision rates, colour index
and spaal scores between ferox trout and brown trout, a nonpara-
metric Mann–Whitney U- test (two- sample Wilcoxon rank sum) was
used. Dominance scores were compared using a Welch two- sample
t- test. All stascal analysis was performed using R version 3. 3. 1
stascal soware (R Core Team, 2016).
3 | RESULTS
Brown trout eggs (n = 8) were signicantly more depleted in δ15N
(t = −35.4, df = 13.1, p < 0.01) than ferox trout eggs (n = 8). There was
no signicant dierence in δ13C (t = 1.4, df = 12.3 p = 0.2) between
brown trout and ferox trout eggs (Figure 2). Similarly, lipid extracted
samples from brown trout eggs (n = 8) were signicantly more depleted
in δ15N (t = −35.2, df = 13.8, p < 0.01) than ferox trout eggs (n = 8), and
there was no signicant dierence in δ13C (t = 0.2, df = 12.8, p = 0.84).
There was a signicant dierence in egg surface area (mm2) be-
tween life history types (F1,197 = 82.06, p < 0.001) with ferox trout
progeny having a greater egg surface area than brown trout progeny
(p < 0.001; Table 1).
There was no signicant dierence in development me taken
to reach successive stages (me to eyed stage [F2,3 = 1.1, p = 0.51];
TABLE1 Egg number and egg surface area (mm2 ± SE) of each
family of ospring
Family Life history Egg number
Egg area
(mm2) ± SE
1FX 122 30.67 ± 0.17
2FX 267 32.45 ± 0.18
5BT 328 30.44 ± 0.19
6BT 462 26.83 ± 0.21
FIGURE2 Mean (±SE) stable isotope ratio of carbon and nitrogen
in benthivorous trout (○) and ferox trout eggs (●) from the Loch
Maree catchment

|
5
HUGHES ET AL.
me to hatch [F2,3 = 1.1, p = 0.51]; and me to swim- up [F2,3 = 6.5,
p = 0.13]) between ospring types (Table 2).
There was a strong, posive correlaon between all four observed
indicators of dominance (Table 3). PC1 weighted all four variables
highly and posively and summarised 58% of the variaon. Thus, high
PC1 scores characterised individuals with high levels of aggression,
lighter skin colouraon, which held beer spaal posions in the arena
and had greater rates of food acquision. Ferox trout progeny had the
higher dominance score of the pair (based on PC1) in 27 of 30 trials.
This is greater than would be expected by chance (χ2 = 41.85, df = 1,
n = 60, p < 0.01). The mean aggression rate of ferox trout progeny was
14.8 ± 2.51 (mean ± SE) compared with benthivorous brown trout
progeny 5.0 ± 1.32 (mean ± SE; Mann–Whitney U- test: W = 222.0,
p < 0.01). The mean food acquision rate of ferox trout progeny was
5.6 ± 0.82 (mean ± SE), higher than benthivorous brown trout progeny
3.2 ± 0.76 (mean ± SE; Mann–Whitney U- test: W = 294.0, p < 0.05).
The mean spaal posion score of ferox trout progeny was higher
5.3 ± 1.17 (mean ± SE), compared with −15.0 ± 2.76 (mean ± SE) for
benthivorous brown trout progeny (Mann–Whitney U- test: W = 94.5,
p < 0.01). The mean body colouraon score of ferox trout progeny
was higher 11.4 ± 1.03 (mean ± SE), compared with −3.43 ± 1.77
(mean ± SE) for benthivorous brown trout progeny (Mann–Whitney
U- test: W = 83.5, p < 0.01). The mean dominance score of ferox trout
progeny was 1.04 ± 0.10 (mean ± SE), compared with −1.04 ± 0.17
(mean ± SE) for benthivorous brown trout progeny (Welch two- sample
t- test df = 45, p < 0.01; Figure 3).
4 | DISCUSSION
This study demonstrates a very clear dierence in dominance behav-
iour in juvenile S. trua derived from parents expressing two alter-
nave life history strategies; a piscivorous, ferox life history strategy
and a benthivorous life history strategy. Although all experimental sh
were reared at the same densies, fed the same type and quanty of
food and were size matched prior to behavioural observaons, the
progeny of ferox trout parents were more aggressive, acquired more
food and held more advantageous spaal posions in the stream than
benthivorous brown trout ospring. Thus, overall, progeny of ferox
trout exhibited considerably higher levels of dominance than the
progeny of benthivorous trout in the experiment.
Given that all individuals in this experiment were reared from eggs
under common condions, the high levels of dominance exhibited by
the progeny of ferox trout, compared with trout of benthivorous pa-
rental origin, indicate that the expression of behavioural dominance
was parentally acquired.
As the adopon of a piscivorous diet in many sh species, in-
cluding S. trua, is dependent upon gape size, it is generally thought
that individuals must achieve a minimum size threshold before they
may access sh prey (Campbell, 1979; Mielbach & Persson, 1998;
Persson et al., 1996).
The wealth of literature on salmonids shows that the dominance
rank expressed has consequences for the life history pathway ul-
mately adopted in later life (Metcalfe, 1998; Metcalfe et al., 1989).
High dominance status individuals acquire higher quality territories
and take a greater share of resources, and in parcular exhibit higher
levels of food acquision, than those of lower dominance status. Thus,
in condions of limited food availability, salmonid sh exhibing high
behavioural dominance grow faster than those of lower dominance
TABLE2 The developmental me (measured in degree days
(°C day−1) of three key developmental stages: eye pigmentaon
present in eggs, day of hatch and “swim- up” stage
Family Life history
Eyed- egg
stage Hatch Swim- up
1FX 239.1 492.9 852.2
2FX 239.1 468.9 843.2
5BT 227.5 484.8 869.4
6BT 239.1 461 861
Aggression
Behavioural traits
Colour Posion Food PC1
Aggression p < 0.01 p < 0.01 p < 0.01 0.41
Colour 0.51 p < 0.01 p < 0.01 0.57
Posion 0.37 0.71 p < 0.01 0.56
Food 0.46 0.38 0.48 0.44
TABLE3 Pairwise Spearman’s
correlaon coecients and PC1
coecients from PCA, for all four
behavioural traits observed. All four traits
were signicantly correlated with one
another (p < 0.01), with the rst principal
component summarising 58% of the
variaon
FIGURE3 The mean (±SE) extracted PC1 scores (overall
dominance score) PCA of four variables (aggression, colour, spatial
position and food acquisition) of benthivorous brown trout and ferox
trout offspring

6
|
HUGHES ET AL.
(Adams et al., 1998; Cus et al., 2001; Höjesjö et al., 2005; Metcalfe
et al., 1989; Nakano, 1995; Van Leeuwen et al., 2015; Wells, 1977).
One consequence for S. trua individuals for which the piscivorous
(ferox) life history is a possible strategy is that if foraging resources are
limited, then high dominance status is likely to result in higher growth
and the aainment of the size threshold for piscivory at an earlier
age than would be possible for individuals expressing low dominance
status. Thus, one logical conclusion is that S. trua adopng a ferox
life history strategy are more likely to be drawn from individuals with
a high dominance status, allowing them to acquire a higher share of
the foraging resources and thus reach the size threshold for piscivory
more quickly.
Whether adult S. trua adopng a ferox life history strategy have
a higher dominance status than those adopng a benthivorous life
history strategy was not tested in this study. The rst- generaon o-
spring of parents adopng a ferox life history strategy were, however,
very clearly and consistently expressing a higher dominance status
than the rst- generaon ospring of parents adopng a benthivorous
life history strategy.
The mechanism through which behavioural dominance status was
acquired from their parents was not clearly idened in this study.
There are, however, at least two alternaves that are not mutually
exclusive. First, dominance status may be genecally inherited, with
the behavioural traits that confer dominance simply passed to o-
spring from their parents. This explanaon is certainly plausible as
inheritance of foraging behaviour is well known from other species
(Ferguson & Noakes, 1982, 1983; Kamler, 2005). Supporng this ex-
planaon is evidence, for at least some populaons, that S. trua ex-
hibing a ferox life history strategy are genecally disnct from those
adopng a benthivorous life history strategy, when the two are in
sympatry (Duguid et al., 2006; Ferguson & Mason, 1981; Ferguson &
Taggart, 1991; McVeigh et al., 1995; Prodöhl et al., 1992). It may also
be noted that as well as being genecally disnct and thus reproduc-
vely isolated in sympatry, all ferox trout examined in sucient detail,
thus far, in Ireland and Scotland are derived from the same lineage
(McKeown et al., 2010, and references therein). Thus, selecon may
have occurred in an ancestral populaon and not convergently in cur-
rent ones. If this is the mechanism for cross- generaonal transmission
of behavioural traits resulng in dominance, this suggests that these
traits have been posively selected for in populaons that express
high proporons of the ferox life history strategy, and less so (or not
at all) in populaons expressing a high frequency of the benthivory life
history strategy. There are currently no populaon genec data for
S. trua expressing the two life history strategies from the Loch Maree
catchment to provide support (or not) for this possible mechanism.
Alternavely, dominance traits may be transmied across gener-
aons through maternal eects. In the experiment reported here, we
tried to control for maternal eects on dominance operang through
body size by matching individuals, as size is known to be strongly
correlated with dominance in salmonids (Hunngford et al., 1990).
Despite this, it is highly plausible that some maternal eects that we
did not control for, may be responsible for the transmission across
generaons reported here. Leblanc, Benhaïm, Hansen, Kristjánsson,
and Skúlason (2011), for example, showed that maternally derived egg
size in Arcc charr, Salvelinus alpinus, inuences juvenile behaviour,
morphology and foraging acvity. There was a dierence in egg size
between life history origins, in the experiment reported here, and it
is thus conceivable that although juveniles were size matched in the
dyad experiment, that some egg size or nutrional legacy was con-
ferred on one life history group that did accrue to the other.
This study demonstrates at least one trait driving the mainte-
nance of the ferox life history type is inherited from one generaon
to the next. Given that ferox trout are rare (Duguid et al., 2006), live
in low densies in the wild (Thorne, MacDonald, & Thorley, 2016) and
are highly sought aer by recreaonal anglers, these ndings have
implicaons for the management of this rare life history type. As most
internaonal conservaon policy is focussed at the species level, less
protecon is aorded to populaons or specic life history types, as a
component part of a larger species complex. Therefore, conservaon
strategies risk overlooking important rare phenotypes, such as ferox
trout. The inability to agree on a taxonomic classicaon for ferox
trout among anglers, sheries managers and academics as either an
adopted life history within the broader S. trua species complex or
as a disnct species S. ferox has important conservaon implicaons
(Ferguson, 2004; Freyhof & Koelat, 2008). We therefore propose
that conservaon strategies must include consideraon of the com-
plex natural processes which give rise to these rare life history types
to eecvely protect the full range of biodiversity for the future.
ACKNOWLEDGEMENTS
We thank two anonymous reviewers’ for extremely useful comments.
A special thanks to Dr Jennifer Dodd and Dr Oliver Hooker and to
Roddy Legge, Kevin McNeil, Ben Rushbrooke, Cory Jones and Kyle
McFarlane for their assistance during eld collecon. Thanks to Dr
Jason Newton and Dr Rona McGill at the NERC Life Sciences Mass
Spectrometry Facility for their help and advice on stable isotope anal-
ysis. We thank Stuart Wilson and other sta at the Scosh Centre for
Ecology and the Natural Environment, for assistance with animal hus-
bandry tasks. This work was supported by funding from the European
Union’s INTERREG IVA Programme (Project 2859 “IBIS”) managed by
the Special EU Programmes Body.
REFERENCES
Abbo, J. C., Dunbrack, R. L., & Orr, C. D. (1985). The interacon of size
and experience in dominance relaonships of juvenile steelhead trout
(Salmo gairdneri). Behaviour, 92, 241–253.
Adams, C., Hunngford, F., Turnbull, J., Arno, S., & Bell, A. L. Y. (2000). Size
heterogeneity can reduce aggression and promote growth in Atlanc
salmon parr. Aquaculture Internaonal, 8, 543–549.
Adams, C. E., Hunngford, F. A., Turnbull, J. F., & Beae, C. (1998).
Alternave compeve strategies and the cost of food acquision in
juvenile Atlanc salmon (Salmo salar). Aquaculture, 167, 17–26.
Appleby, M. C. (1980). Social rank and food access in red deer stags.
Behaviour, 74, 294–309.
Burton, T., Hoogenboom, M. O., Armstrong, J. D., Groothuis, T. G., &
Metcalfe, N. B. (2011). Egg hormones in a highly fecund vertebrate:

|
7
HUGHES ET AL.
do they inuence ospring social structure in compeve condions?
Funconal Ecology, 25, 1379–1388.
Campbell, R. N. (1971). The growth of brown trout Salmo trua L. in north-
ern Scosh lochs with special reference to the improvement of sher-
ies. Journal of Fish Biology, 3, 1–28.
Campbell, R. N. (1979). Ferox trout, Salmo trua L., and charr, Salvelinus
alpinus (L.), in Scosh lochs. Journal of Fish Biology, 14, 1–29.
Cluon-Brock, T. H., Albon, S. D., & Guinness, F. E. (1986). Great expecta-
ons: dominance, breeding success and ospring sex raos in red deer.
Animal Behaviour, 34, 460–471.
Cole, B. J. (1981). Dominance hierarchies in Leptothorax ants. Science, 212,
83–84.
Cowlishaw, G., & Dunbar, R. I. (1991). Dominance rank and mang success
in male primates. Animal Behaviour, 41, 1045–1056.
Creel, S. (2001). Social dominance and stress hormones. Trends in Ecology &
Evoluon, 16, 491–497.
Cus, C. J., Adams, C. E., & Campbell, A. (2001). Stability of physiolog-
ical and behavioural determinants of performance in Arcc char
(Salvelinus alpinus). Canadian Journal of Fisheries and Aquac Sciences,
58, 961–968.
Dewsbury, D. A. (1982). Dominance rank, copulatory behavior, and dier-
enal reproducon. Quarterly Review of Biology, 57, 135–159.
Dingemanse, N. J., & de Goede, P. (2004). The relaon between domi-
nance and exploratory behavior is context- dependent in wild great ts.
Behavioral Ecology, 15, 1023–1030.
Dodson, J. J., Aubin-Horth, N., Thériault, V., & Páez, D. J. (2013). The evo-
luonary ecology of alternave migratory taccs in salmonid shes.
Biological Reviews, 88, 602–625.
Duguid, R. A., Ferguson, A., & Prodöhl, P. (2006). Reproducve isolaon and
genec dierenaon of ferox trout from sympatric brown trout in Loch
Awe and Loch Laggan, Scotland. Journal of Fish Biology, 69, 89–114.
Ellis, L. (1995). Dominance and reproducve success among nonhuman
animals: a cross- species comparison. Ethology and Sociobiology, 16,
257–333.
Ens, B. J., & Goss-Custard, J. D. (1984). Interference among oystercatchers,
Haematopus ostralegus, feeding on mussels, Mylus edulis, on the Exe
estuary. The Journal of Animal Ecology, 53, 217–231.
Ferguson, A. (2004). The importance of idenfying conservaon units:
brown trout and pollan biodiversity in Ireland. Proceedings of the Royal
Irish Academy, 104B, 33–41.
Ferguson, A., & Mason, F. M. (1981). Allozyme evidence for reproducvely
isolated sympatric populaons of brown trout Salmo trua L. in Lough
Melvin, Ireland. Journal of Fish Biology, 18, 629–642.
Ferguson, M. M., & Noakes, D. L. (1982). Genecs of social behaviour in
charrs (Salvelinus species). Animal Behaviour, 30, 128–134.
Ferguson, M. M., & Noakes, D. L. (1983). Movers and stayers: Genec
analysis of mobility and posioning in hybrids of lake charr, Salvelinus
namaycush, and brook charr, S. fonnalis (Pisces, salmonidae). Behavior
Genecs, 13, 213–222.
Ferguson, A., & Taggart, J. B. (1991). Genec dierenaon among the
sympatric brown trout (Salmo trua) populaons of Lough Melvin,
Ireland. Biological Journal of the Linnean Society, 43, 221–237.
Fox, S. F., Rose, E., & Myers, R. (1981). Dominance and the acquision
of superior home ranges in the lizard Uta stansburiana. Ecology, 62,
888–893.
Freyhof, J., & Koelat, M. (2008). Salmo ferox. The IUCN Red List of
Threatened Species, 2008, e.T135577A4150683.
Gilmour, K. M., DiBasta, J. D., & Thomas, J. B. (2005). Physiological causes
and consequences of social status in salmonid sh. Integrave and
Comparave Biology, 45, 263–273.
Greer, R. (1995). Ferox trout and arcc charr: a predator, its pursuit and its
prey. London: Swan-Hill Press.
Grey, J. (2001). Ontogeny and dietary specializaon in brown trout (Salmo
trua L.) from Loch Ness, Scotland, examined using stable isotopes of
carbon and nitrogen. Ecology of Freshwater Fish, 10, 168–176.
Grey, J., Thackeray, S. J., Jones, R. I., & Shine, A. (2002). Ferox trout (Salmo
trua) as Russian dolls’: complementary gut content and stable iso-
tope analyses of the Loch Ness foodweb. Freshwater Biology, 47,
1235–1243.
Harwood, A. J., Armstrong, J. D., Griths, S. W., & Metcalfe, N. B. (2002).
Sympatric associaon inuences within- species dominance relaons among
juvenile Atlanc salmon and brown trout. Animal Behaviour, 64, 85–95.
Höjesjö, J., Armstrong, J. D., & Griths, S. W. (2005). Sneaky feeding by
salmon in sympatry with dominant brown trout. Animal Behaviour, 69,
1037–1041.
Hughes, M. R., Dodd, J. A., Maitland, P. S., & Adams, C. E. (2016). Lake
bathymetry and species occurrence predict the distribuon of a lacus-
trine apex predator. Journal of Fish Biology, 88, 1648–1654.
Hunngford, F. A. (2004). Implicaons of domescaon and rearing condi-
ons for the behaviour of culvated shes. Journal of Fish Biology, 65,
122–142.
Hunngford, F. A., Metcalfe, N. B., Thorpe, J. E., Graham, W. D., & Adams,
C. E. (1990). Social dominance and body size in Atlanc salmon parr,
Salmo salar L. Journal of Fish Biology, 36, 877–881.
Jensen, H., Kiljunen, M., & Amundsen, P. A. (2012). Dietary ontogeny and
niche shi to piscivory in lacustrine brown trout Salmo trua revealed
by stomach content and stable isotope analyses. Journal of Fish Biology,
80, 2448–2462.
Kamler, E. (2005). Parent–egg–progeny relaonships in teleost shes:
an energecs perspecve. Reviews in Fish Biology and Fisheries, 15,
399–421.
Kilsen, S., Schjolden, J., Beitnes-Johansen, I., Shaw, J. C., Ponger,
T. G., Sørensen, C., Braastad, B. O., Bakken, M., & Øverli, Ø. (2009).
Melanin- based skin spots reect stress responsiveness in salmonid
sh. Hormones and Behavior, 56, 292–298.
Leblanc, C. A. L., Benhaïm, D., Hansen, B. R., Kristjánsson, B. K., & Skúlason,
S. (2011). The importance of egg size and social eects for behaviour
of Arcc charr juveniles. Ethology, 117, 664–674.
Mangel, M., & Abrahams, M. V. (2001). Age and longevity in sh, with
consideraon of the ferox trout. Experimental Gerontology, 36,
765–790.
Marra, P. P., & Holmes, R. T. (2001). Consequences of dominance- mediated
habitat segregaon in American Redstarts during the nonbreeding sea-
son. The Auk, 118, 92–104.
McKeown, N. J., Hynes, R. A., Duguid, R. A., Ferguson, A., & Prodöhl, P.
A. (2010). Phylogeographic structure of brown trout Salmo trua in
Britain and Ireland: glacial refugia, postglacial colonizaon and origins
of sympatric populaons. Journal of Fish Biology, 76, 319–347.
McVeigh, H., Hynes, R., & Ferguson, A. (1995). Mitochondrial DNA dif-
ferenaon of sympatric populaons of brown trout, Salmo trua L.,
from Lough Melvin, Ireland. Canadian Journal of Fisheries and Aquac
Sciences, 52, 1617–1622.
Metcalfe, N. B. (1998). The interacon between behavior and physiology
in determining life history paerns in Atlanc salmon (Salmo salar).
Canadian Journal of Fisheries and Aquac Sciences, 55, 93–103.
Metcalfe, N. B., Hunngford, F. A., Graham, W. D., & Thorpe, J. E. (1989).
Early social status and the development of life- history strategies in
Atlanc salmon. Proceedings of the Royal Society of London Biological
Sciences, 236, 7–19.
Metcalfe, N. B., Valdimarsson, S. K., & Morgan, I. J. (2003). The relave
roles of domescaon, rearing environment, prior residence and body
size in deciding territorial contests between hatchery and wild juvenile
salmon. Journal of Applied Ecology, 40, 535–544.
Mielbach, G. G., & Persson, L. (1998). The ontogeny of piscivory and
its ecological consequences. Canadian Journal of Fisheries and Aquac
Sciences, 55, 1454–1465.
Nakano, S. (1995). Individual dierences in resource use, growth and em-
igraon under the inuence of a dominance hierarchy in uvial red-
spoed masu salmon in a natural habitat. Journal of Animal Ecology, 64,
75–84.

8
|
HUGHES ET AL.
Nicieza, A. G., & Metcalfe, N. B. (1999). Costs of rapid growth: the risk
of aggression is higher for fast- growing salmon. Funconal Ecology, 13,
793–800.
Persson, L., Andersson, J., Wahlström, E., & Eklöv, P. (1996). Size- specic
interacons in lake systems: predator gape limitaon and prey growth
rate and mortality. Ecology, 77, 900–911.
Preston, A. C., Taylor, J. F., Adams, C. E., & Migaud, H. (2014). Surface feeding
and aggressive behaviour of diploid and triploid brown trout Salmo trua
during allopatric pair- wise matchings. Journal of Fish Biology, 85, 882–900.
Prodöhl, P. A., Taggart, J. B., & Ferguson, A. (1992). Genec variability within
and among sympatric brown trout Salmo trua populaons: mul- locus
DNA ngerprint analysis. Hereditas, 117, 45–50.
R Core Team (2016). R: A language and environment for stascal compung.
Vienna, Austria: R Foundaon for Stascal Compung. URL: hps://
www.R-project.org/.
Röseler, P. F., Röseler, I., Strambi, A., & Augier, R. (1984). Inuence of in-
sect hormones on the establishment of dominance hierarchies among
foundresses of the paper wasp, Polistes gallicus. Behavioral Ecology and
Sociobiology, 15, 133–142.
Seyfarth, R. M. (1976). Social relaonships among adult female baboons.
Animal Behaviour, 24, 917–938.
Thorne, A., MacDonald, A. I., & Thorley, J. L. (2016). The abundance of
large, piscivorous Ferox Trout (Salmo trua) in Loch Rannoch, Scotland.
Pe erJ, 4, e2646.
Van Leeuwen, T. E., Hughes, M. R., Dodd, J. A., Adams, C. E., & Metcalfe,
N. B. (2015). Resource availability and life- history origin aect
compeve behavior in territorial disputes. Behavioral Ecology, 27,
385–392.
Wells, K. D. (1977). Territoriality and male mang success in the green frog
(Ranaclamitans). Ecology, 58, 750–762.