Upload
vuongmien
View
216
Download
0
Embed Size (px)
Citation preview
ISSN 1403-5227
LICENTIATAVHANDLING
ZOOLOGISKA
INSTITUTIONEN Stockholms Universitet S-106 91 Stockholm
Social dominance and personality in male fowl (Gallus gallus domesticus)
Författare: Anna Favati Ämne: Etologi Löpnummer: 2013:4
Arbetets titel:
Social dominance and personality in
male fowl (Gallus gallus domesticus)
Författarens namn: Anna Favati
Handledare: Olof Leimar
Forskningsämne: Etologi
Granskningskommitté: Charlie Cornwallis, Sven Jakobsson, Bodil Elmhagen
Datum för seminarium: 2013-‐05-‐13
INDEX
ABSTRACT ............................................................................................................................... 1
INTRODUCTION ..................................................................................................................... 3
METHODS ................................................................................................................................ 6
PAPER I ................................................................................................................................................ 6 PAPER II ............................................................................................................................................... 9
RESULTS ................................................................................................................................ 10
PAPER I .............................................................................................................................................. 10 PAPER II ............................................................................................................................................. 11
DISCUSSION .......................................................................................................................... 13
CONCLUSION ........................................................................................................................ 17
ACKNOWLEDGEMENT ....................................................................................................... 17
REFERENCES ........................................................................................................................ 19
LIST OF PAPERS
This thesis is based on the following manuscripts, referred to by their roman numerals.
Paper I
Favati, A., Leimar, O., Radesäter, T. and Løvlie, H. Social ‘states’ and personality: Stability
in social status and individual characteristics explain consistency in behavioural responses.
Submitted manuscript
Paper II
Favati, A., Leimar, O. and Løvlie, H. Aggressiveness and exploration predicts social status in
male domestic fowl (Gallus gallus domesticus). Manuscript
1
1
Social dominance and personality in male fowl (Gallus gallus domesticus)
Anna Favati
Department of zoology, Stockholm university, SE-106 91 Stockholm, Sweden
ABSTRACT
Individuals in social species commonly form dominance relationships among each other, and
are often observed to differ in behaviour depending on their social status. However, whether
such behavioural differences are a consequence of dominance position, or also a cause to it,
remains unclear. In this thesis I therefore investigated two perspectives of the relationship
between social dominance and personality in the domestic fowl (Gallus gallus domesticus), a
social species that forms relatively stable dominance hierarchies. In paper I I investigated the
influence of social status on the expression and consistency of behaviours by experimentally
changing status between repeated personality assays. The level of vigilance, activity and
exploration changed with social status, while boldness and territorial crows appeared as
stable individual properties, independent of status. These results showed that social status
contribute to both variation and consistency in behavioural responses. Social status should
therefore be taken into account when investigating and interpreting variation in personality.
In paper II I showed that behaviour in a novel arena test and during encounter with an
opponent can predict social status, more specifically that fast exploration and aggressiveness
predicted a dominant social position. Together, these results highlight the dynamics of the
two-way relationship between social position and individual behaviour and indicate that
individual behaviour can both be a cause and a consequence of social status.
Kruijt 1964
2
3
INTRODUCTION
The take-off point of this licentiate thesis is the notion that not all individuals of a group
behave the same. Individuals of group living species often form dominance relationship
among each other, characterised by repeated outcome in favour of one participant of dyadic
agonistic interactions (Chase 1980). Socially dominant individuals commonly enjoy
increased access to resources, such as mating partners, which typically results in higher
reproductive success (Andersson 1994). Beside differences in social behaviour, like who is
more aggressive or perform more mating behaviours (e.g. Korzan et al. 2006, McGhee and
Travis 2010), there are also a few examples of behavioural differences between dominant and
subordinate individuals in non-social contexts (e.g. Colleter and Brown 2011, Dahlbom et al.
2011, David et al. 2011). For example, explorative great tits (Parus major) outcompete less
explorative ones (Verbeek et al. 1996), and bold three-spined sticklebacks (Gasterosteus
aculeatus) are more aggressive than shyer individuals (Huntingford 1976). However, there
are also a couple of examples of negative correlations, and no correlations between
explorative behaviour or boldness and dominance (Gomez-Laplaza 2002). In the mountain
chickadee (Poecile gambeli) less explorative individuals have higher chances of becoming
dominant, and boldness does not differ between dominant and subordinate individuals (Fox et
al. 2009). Interestingly, variation in such behavioural differences as activity, exploration,
boldness and aggression has been demonstrated to be consistent across time and context
(Wilson et al. 1994, Dall et al. 2004, Sih et al. 2004b, Groothuis and Carere 2005). Consistent
individual differences in behaviour have been described in a large number of species in
multiple taxa (Gosling 2001), and have been referred to as ‘animal personality’ (Dall et al.
2004) or ‘temperaments’ (Reale et al. 2007). When multiple behavioural traits are correlated
across time or context, the term ‘behavioural syndromes’ (Sih et al. 2004a) is used, and when
focus lays on the relationship between behavioural consistency and stress responses, the
related concept of ‘coping styles’ is often applied (Koolhaas et al. 1999). By scoring
individuals in personality tests, they can be categorised along a continuous axis of specific
personality traits, like boldness or exploration, and an individual is then said to have a
'behavioural type' or 'personality type' (e.g. more or less explorative, Sih et al. 2004b). The
research field of animal personality is still young, although fast growing. Nevertheless, there
are still major gaps in the understanding of why there is personality variation, including
questions about the mechanisms behind stable behavioural responses and the evolution and
4
maintenance of behavioural polymorphism (Dingemanse and Wolf 2010, Reale et al. 2010).
Current ideas of how variation is maintained include negative frequency dependence (Dall et
al. 2004, Wolf and Weissing 2010), selection regimes that fluctuate over time (Dingemanse
and de Goede 2004), and various trade-offs, for example between survival and reproductive
success (Smith and Blumstein 2008), or between growth and mortality (Stamps 2007).
Behavioural consistency within the individual has been suggested to be maintained by
physiological limitations in plasticity (Sih et al. 2004b), or by stability in physiological or
environmental states that changes slower than behaviour, and that affects the costs and
benefits of the individual's actions (e.g. body size, territory size or stable social hierarchies;
Houston and McNamara 1999, Dall et al. 2004, Dingemanse and Wolf 2010, Luttbeg and Sih
2010, Wolf and Weissing 2010). To learn more about the significance and evolution of
animal personality, studies of both through survival and reproductive success of different
personality types are requested (Dingemanse and Reale 2005, Smith and Blumstein 2008).
Social dominance typically affects reproductive success positively (Andersson 1994), and as
mentioned above, often seem to correlate to various personality traits. It would therefore be
interesting to further study this relationship. Not least, we should also aim to investigate the
causality behind this relationship, to reveal possible fitness consequences of various
personality traits.
There are four partly nonexclusive scenarios that may describe the causality of the
relationship between individual behaviour and social status. First, behavioural response may
be a consequence of the current social position (Cornwallis and Birkhead 2008). If so, we
expect a change in dominance status to lead to a change in behavioural response. This would
also mean that social status may represent a state that gives rise to consistent differences in
behaviour (Dall et al. 2004). Nevertheless, there are very few studies that have
experimentally altered social status to study if a change in behavioural response follows (but
see Cornwallis and Birkhead 2008). Second, social status and behavioural responses may
share a common underlying cause such as stress reactivity (Øverli et al. 2007, Carere et al.
2010). In this case we expect behaviours to be correlated to social status, but showing less
plasticity when status is changed. Third, and not exclusive from the second, the personality
type may affect an individuals' possibility to obtain a dominant social position (e.g. Dahlbom
et al. 2011). In this case, we expect to predict future social status by means of behavioural
responses in non-social contexts. Phenotypic traits like body size or size of armaments
etcetera have traditionally been investigated in this scenario and have been shown to predict
5
social status (Andersson 1994, Berglund et al. 1996). The potential link between behaviour
and future status is on the other hand much less investigated and the role for behavioural
differences to influence establishment of status is still poorly explored (but see e.g. Øverli et
al. 2004, Korzan et al. 2006, Dahlbom et al. 2011). And last, there is the remaining possibility
that variation in behavioural responses or personality traits is uncorrelated to variation in
social status. In this case, we expect no difference in the behavioural responses between
dominant and subordinate individuals, and behaviours to be unaffected by changes in social
position. Revealing the causal relationship between behavioural differences and social status
is essential for further understanding of how variation in personality types affect fitness, and
ultimately for understanding of the evolution of personality types. Each scenario described
above has been examined to some extent, but there is a lack of studies exploring more than
one of the scenarios outlines above in the same animal model. In this thesis, I aim to study the
causality of the relationship between behaviour and social status by an experimental
approach, using the male domestic fowl (Gallus gallus domesticus) as my study species.
Study species
The domestic fowl (Gallus gallus domesticus) is, like its wild ancestor the red junglefowl
(Gallus gallus; Fumihito et al. 1994), a group living species. Natural groups are typically
constituted by a dominant male, a few subordinate males and several females, and group sizes
ranges from pairs to around 10-15 individuals (McBride et al. 1969, Collias and Collias
1996). The sex ratio varies, but is generally equal or slightly female biased (e.g. 2:3-2:5,
males to females), probably due to higher predation on subordinate males in the periphery of
the group (McBride et al. 1969, Collias and Collias 1996). Both males and females form
(most often linear) social hierarchies defined by threats, attacks and avoidance, and males
always dominate over females (Banks 1956, Collias et al. 1994). The fowl is suitable for
studies of the potential link between behaviour and social status for three main reasons. First,
the status of an individual is rather stable and can persist for several years, but last on average
approximately one breeding season in free-ranging flocks of red junglefowl (Collias and
Collias 1996). This is a considerable part of the lifespan of males, which lives up to 5.5 years
in semi-natural flocks exposed to predation (Collias and Collias 1996). However, changes in
groups over shorter time spans can occur and cause changes in the status of individuals, for
example if the dominant male is predated (McBride et al. 1969). Second, dominant males
have higher reproductive success compared to subordinate males (Pizzari and Birkhead 2000,
Wilson et al. 2008), thus the prerequisite for evolutionary consequences of a relationship
6
between social dominance and personality traits. Third, dominant and subordinate male fowl
differ quantitatively in behaviour. Dominant males not only perform more courtship
behaviour (Cheng and Burns 1988), but also crow more (a signal of territoriality and
dominance; Leonard and Horn 1995), and spend more time being vigilant compared to
subordinate males (Cornwallis and Birkhead 2008). However, the relationship between
personality traits and social status is unclear.
METHODS
The study population is kept at Tovetorp research station (Björnlunda, Sweden) in six mixed-
sex (sex ratio ~ 1:1), mixed-age (1-10 yrs) groups (15-20 individuals per group) in outdoor
aviaries (approximately 6x10x2 m) with access to chicken houses. All aviaries are furnished
with perches and dust baths, and the animals have ad libitum access to regular chicken food
and water. The experiments were performed during the breeding season (may-september) in
2007 (n males = 36), 2011 (n = 48, of which 9 were reused from 2007) and 2012 (n = 50).
Personality assays
There were no personality assays specifically developed for chickens when I started the work
with this thesis. Therefore, I developed two novel arena tests to measure activity and
explorative behaviour. The design of these were inspired by the open field test commonly
used in applied ethology to measure variation in fear in domestic fowl (often a small barren
arena, e.g. 1 m. in diameter; Jones et al. 1995) and by the novel arena tests commonly used to
measure variation in exploration propensity in passerine birds (commonly a room furnished
with five artificial trees; Verbeek et al. 1994). The arena used in paper I consisted of an
outdoor roofed arena (3x6 m), with peat floor and five artificial trees (50 cm high). The trees
were added to prevent immediate overview of the arena, encouraging the fowl to explore it
(Fig. 1a). The arena was divided in 8 equal-sized subareas by drawing subtle lines in the
substrate, and the number of subareas visited was used as a measure of exploration
propensity. A personality test should optimally be designed to reflect variation in individual
behaviour within a population (Reale et al. 2007). However, around half of the males retained
the maximum exploration score of visiting 8 subareas in the novel arena. In paper II I
therefore used a larger arena (~ 7x10 m oval, divided into 36 equal-sized subareas; Fig. 1b)
placed in the forest, without a roof, in order to approve of more variation in explorative
behaviour. The forest floor was naturally covered with leafs, herbs and bushes, as in the
7
natural habitat of Red junglefowl (Collias et al. 1964). A full description of the behavioural
responses measured in the novel arena tests are given below.
Figure 1. The novel arena-test used as a personality assay in (a) paper I, and (b) paper II.
Startle tests are commonly used to estimate variation in risk-taking behaviour within the
research field of personality (van Oers et al. 2004, Bell 2005, Quinn and Cresswell 2005,
Briffa et al. 2008, Rudin and Briffa 2012). In general a stressor (a sound, a predator model
etc.) is presented, and thereafter the latency until the individual retain regular activities like
feeding, is measured (e.g. van Oers et al. 2004). The fowl utter alarm calls (loud cackling
vocalisations) in response to terrestrial predators, to which flock mates react on by vigilance
(Evans et al. 1993). In paper II I therefore used a recorded male fowl alarm call as a startle.
The latency until the male retained normal activities after the call was played, thus stopped
being vigilant and resumed feeding, was used as a measure of how alert the male was, or in
other words how risky the individual appreciated the environment to be.
Morphological measures
Male body size and comb size (a fleshy red ornament on the head) often correlate positively
with social status in male fowl (Ligon et al. 1990; Zuk and Johnsen 2000; Parker et al. 2002),
and are also shown to interfere with the establishment of social status (Parker et al. 2002).
Males I used, were therefore matched for these morphological features in both studies (to the
nearest 10%). Come size, approximated by comb length and body size, approximated by
tarsus length and body weight, were measured prior to the experiments, at the day of the pair
setup (paper I), or at the day of isolation (paper II).
(a) (b)
8
Paper I
To study the influence of social status on the expression and consistency of behaviours, I
experimentally changed status of male fowl between repeated personality assays. Pairs of
matched males were set up, and the dominance order of the two males was determined. After
acclimatising to the new social situation, a novel arena test was performed in order to
investigate their boldness (latency to enter the arena), activity (no. of subarea transitions) and
explorative behaviour (no. of subareas visited). The males' vigilance (proportion of time a
male spent being vigilant) was also registered, as was the number of crows (a vocal territorial
display, Collias et al. 1964, Leonard and Horn 1995). Then pair members were
experimentally changed so that the two previously dominant males formed a new pair, and
the two subordinate males another pair, forcing one male in each pair to change social status,
as there can only be one dominant male and one subordinate male in each pair (Fig. 2; see
Cornwallis and Birkhead 2007, 2008 for previous use of this design). After two days of
acclimation to the new social context, the novel arena test was performed once again in order
to estimate the influence of change of social status on behaviour.
Figure 2. Schematic picture of the design used when manipulating social status of male fowl. Males were either
A) dominant across both pair set-ups, B) increased status from subordinate to dominant, C) reduced status from
dominant to subordinate, or D) were subordinate over both pair set-ups.
Setup 1 Setup 2
A
B
C
D
B
D
A
C
A: Dominant -‐ Dominant
C: Dominant – Subordinate
B: Subordinate – Dominant
D: Subordinate -‐ Subordinate
9
Paper II
To investigate the role for personality on establishment of status, I scored individual
behaviour in a novel arena test and a startle test (play back alarm sound of a conspecific, see
above). In paper II I added the latency to visit at least five subareas as a complementary
measure of exploration propensity (Verbeek et al. 1996). Besides this additional measure, the
behavioural measures were the same as in paper I. All males were isolated from same-sex
competitors for five days prior to the behavioural tests in order to standardise their previous
social experience. After personality was assessed, males were set up in duels resulting in one
male becoming dominant and the other subordinate. The agonistic interaction between males
started spontaneously they were placed in front of each other, and the initial reaction to the
duel situation was used to score the males' aggressiveness from 1-4 (Fig. 3). The males
aggressiveness was scored between 1 (least aggressive) and 4 (most aggressive) as follows:
(1) fleeing away from the other male, (2) straight body posture, no avoidance or attraction to
the other male, (3) crouched body posture and dropped wing or raised hackles (Collias 1943,
Kruijt 1964) or (4) directly – within 2 seconds – approaching the other male (Fig. 3). All
duels ended with one of the opponents retreating and no longer retaliating further attacks
from the opponent. A minimum of five occasions of avoidance within an hour by the same
male was used to define a male as subordinate, and the other male as dominant.
Although I only measured behavioural responses once, and thus did not explicitly evaluate
their repeatability, similar responses have been found to be consistent within the individual in
an earlier study of male fowl (paper I). Moreover, a number of studies in other species have
shown that explorative behaviour and boldness typically show substantial repeatability over
time (e.g. Verbeek et al. 1994, Dingemanse et al. 2002, Quinn and Cresswell 2005). I
therefore consider the single measurement in this study as useful for determination of
personality types.
10
Figure 3. Illustration of body postures used for scoring aggressiveness. Picture of a male having a (a) non-
aggressive body posture (upright body posture, score 2) and (b) aggressive body posture (i.e. raised hackles, a
crouched body position and dropped wing, score 3).
SUMMARY OF RESULTS
Paper I
There were several behavioural differences observed between dominant and subordinate
males. Dominant males were more vigilant and active and crowed more compared to
subordinate males, but did not differ significantly in exploration and boldness in the first trial
of the novel arena test (see Fig. 1 in paper I). The main result of this study was however that
some of these responses (vigilance, activity and exploration, Fig. 4 b-d; Table 2 in paper I)
were plastic and changed in frequency when social status was changed, while others were not
(boldness and crowing Fig. 4a, 4e). At the same time, there was a significant relationship
between behaviour in the first and the second trial for all behavioural responses. In other
words, the individual behaviours we recorded were partly stable over time, and change of
status.
Paper II
Among the behaviours observed in the novel arena test in paper II, exploration was the only
response that could forecast social dominance. The quantitative difference in exploration
between future dominant and subordinate males was fairly large; dominant males explored
the novel arena were more than twice as fast as the subordinate males (dominant
243.92±57.16 s., subordinate 500±96 s., Fig. 5). In contrast, there were no significant
(a) (b)
11
differences in activity, vigilance, alertness after being startled, or crowing between future
dominant and subordinate males (Fig. 5). Aggressiveness in the first seconds of duels could
also predict a dominant position (Fig. 6), and males that showed a higher aggressiveness
more often won the duel. Exploration and aggressiveness were not significantly correlated,
thus not forming a behavioural syndrome along a proactive-reactive axis (Table 1), but
instead independently predicted a dominant social position.
Figure 4. Behavioural response of dominant and subordinate males during the first and second test trial of a
repeated novel arena test. (a) Boldness (whether the males entered the novel arena within 30 s. or not) was stable
within the individual across changes in social status. (b-d) Males that had a stable status across trials showed
more stable behavioural responses of exploration, activity and vigilance compared to males that changed social
status, irrespective of their social status. (e) The number of crows uttered in the second trial was significantly
●
●●
●
●
●●
●
304050607080
trial 1 trial 2
Bold
ness
(a)
●
●●
●
●
●
●
●
0.4
0.5
0.6
0.7
0.8
0.9
trial 1 trial 2
Vigi
lanc
e
(b)
●
●
●
●
●
●
●●
10
15
20
25
30
35
40
trial 1 trial 2
Activ
ity
● DominantSubordinate
Stable statusChanged status
(c)
●
●
●
●
●
●
●
●
5
6
7
8
trial 1 trial 2
Expl
orat
ion
(d)
●
●
●
●
●
●
● ●
0
5
10
15
20
25
trial 1 trial 2
Cro
win
g
(e) ●
●
DominantSubordinateStable statusChanged status
12
affected by previous, but not current, social status, and no clear effects of the change in status were detected.
Values presented are mean ± SE for (b-e), but for (a) the percentages of males that entered the arena within 30
seconds ± 95% CI, are given.
Figure 5. Difference in behavioural response in the novel arena test (a-e) and the startle test (f) between males
that later became dominant (filled dots), or subordinate (empty dots), respectively. Future dominant males were
more explorative compared to subordinate males. All other differences were non-significant. Mean values ± SE
are given.
●
●
0
2
4
6
8
10
Late
ncy
to e
xplo
re5
suba
reas
(min
)
Dom Sub
(a)
●
●
0
10
20
30
40
Activ
ity (%
of t
ime)
Dom Sub
(b)
●
●
0
5
10
15
20
25
30
No.
of s
ubar
eas
visi
ted
Dom Sub
(c)
●
●
10
15
20
25
30
Num
ber o
f cro
ws
Dom Sub
(d)
●
●
50
60
70
80
90
100
Vigi
lanc
e (%
of t
ime)
Dom Sub
(e)
●
●
0
40
80
120
160
Vigi
lanc
e af
ter s
tartl
e (s
)
Dom Sub
(f)
13
Figure 6. Difference in aggressiveness score between males that subsequently became dominant and
subdominant. Males that later won the duel and obtained a dominant social position more often had a higher
aggressiveness score compared to the later subordinate male. However, both males of a pair had equal
aggressiveness score in nearly half of the cases.
To sum up the main results of both papers, dominant and subordinate males differed
quantitatively in several behavioural aspects in non-social contexts. These differences were
partly a result from the current social status (paper I), but also existed prior to the
establishment of social status (paper I and paper II). Together the results reveal a two-way
causality between social position and individual behaviour.
DISCUSSION
In this thesis, I investigated the causality of the relationship between individual behaviour and
social status in male fowl. Dominant and subordinate males differed in several behavioural
responses in a non-social context, and by an experimental approach I showed that the current
social status primarily causes these differences (paper I). However, in paper II I showed that
whether a male end up being dominant or subordinate is not random with respect to
personality types, but influenced by males' personality.
Dominant males were more vigilant, explorative and active, and the occurrence of these
behaviours changed after the experimental change in social status (Fig. 1b-d, Table 2 in
paper I). This meant that individual males adjusted their level of expression of these
behaviours according to their current social status, supporting the first scenario of the
causality of the relationship between social dominance and behaviour. In a study of male
fowl by Cornwallis and Birkhead (2008), vigilance was similarly found to be a plastic
response which frequency changed when social status was changed. It has earlier been
# of
due
ls
0
2
4
6
8
10
12
Dom>Sub Sub>Dom Dom=Sub
14
demonstrated in male fowl that dominants suppress the behaviour of subordinates (Leonard
and Horn 1995, Johnsen et al. 2001). The quick change in behaviour following a change in
social status (paper I) indicates that subordinate males have an intrinsic capability of
becoming dominant and behaving thereafter, but that such behaviour seem to be suppressed
by the presence of a more dominant male (Wilson et al. 2009). In addition, all males behaved
in a dominant manner already after a very short period of isolation from other males
(personal observation, paper II; see also Pizzari et al 2007 for rapid changes in male
reproductive physiology associated with alteration of status). This indicates that variation in
responses to personality assays do not necessarily reflect fixed behavioural types, but may
reflect the current social position of the individuals. This is in congruence with studies of
male rats (Rattus norvegicus) brought up under various social regimes, showing that social
position affects behaviour also in non-social test situations (Arakawa 2006). The strong effect
of social status indicates that status may act like a ‘state’; a slow changing condition that
infers stability also in behavioural response (Wilson et al. 1994, Dall et al. 2004). Despite the
intuitiveness of the idea of stabilising (external) states causing consistent behaviour, there is a
lack of empirical studies explicitly testing this idea. However, there is one example; a study
investigating Siamese fighting fish (Betta splendens), which demonstrated that holding a
territory stabilise individual behaviour (Dzieweczynski and Forrette 2013), a phenomenon
closely related to social position. Part of the behaviour remained stable across the change in
status in our study, which means that other states and/or heritability also play a role in
shaping individual behaviour. In the study by Cornwallis and Birkhead (2008), no such
behavioural consistency was detected across a change in social status. Behaviour was
monitored in a social context in their study, and it therefore appears that social status has an
even stronger effect in the presence of conspecifics, overriding individual differences. In
another study, male vocalisations (Nelson et al. 2008) were shown to correlate across
functional contexts only when the males were observed in their (stable) social groups, further
strengthening this conclusion.
Dominant and subordinate males differed in crow rate during the first novel arena trial, and
little plasticity was observed in this behaviour when social status was changed. Individual
males seemed to vary in their underlying propensity to crow and this supports the second
scenario outlined previously where social status and behavioural responses share a common
underlying cause. A potential underlying factor for crowing and social dominance could be
testosterone, which regulates development of crowing frequency in male fowl (Davis and
15
Domm 1943). Further, males fowl with a high testosterone level have shorter attack latencies
and more often become dominant (Johnsen and Zuk 1995). Indeed, a high crow rate in the
novel arena test was predictive of dominance, although the study in paper I was not
specifically designed to test predictors of status. The latter result was not repeated in paper
II, thus the predictive value of crow rate for social status is questioned. A potential
explanation for this discrepancy could be that the social settings prior to the novel arena test
affects crowing in the arena. In paper I, males were housed in pairs, which forces the
subordinate male to reduce its crowing frequency, since dominant males punish subordinate
males that crow (Leonard and Horn 1995). This might have reinforced the difference in crow
rate between males that were high or intermediate in relative dominance ability (that were
dominant in one or two pair setups) and males that were repeatedly found in subordinate
positions (subordinate in both trials, Fig. 4). In contrast, males were individually caged for
five days prior to the arena test in paper II, thus no males were suppressed in their crowing.
In paper II I performed additional tests to investigate the possibility that behavioural type
could predict future social status, thus testing the third causality scenario; that personality
affects social status. Males that were either more explorative or aggressive had increased
chances of obtaining a dominant position. The results indicate that pre-existing differences in
behavioural type affected the outcome of duels, in other words that not all individuals had
equal chances of obtaining a dominant position. These traits were uncorrelated and equally
strong predictors of social status, suggesting that multiple factors are important for winning a
fight. Scoring high in these traits is considered proactive on the proactive-reactive-continuum
(Koolhaas et al. 1999, Sih et al. 2004b), and the results are therefore in congruence with
studies of other species, which show a general pattern of more proactive individuals having
higher chances of becoming dominant (e.g. Huntingford 1976, Verbeek et al. 1996, Colleter
and Brown 2011, Dahlbom et al. 2011, David et al. 2011). However, there are a couple of
examples of less bold or less explorative individuals having higher chances of obtaining
dominant positions (Gomez-Laplaza 2002, Fox et al. 2009). In a study of great tits, the
success of individuals differing on a proactive-reactive axis differed depending on the
intensity of intraspecific competition over space and food (Dingemanse et al. 2004),
indicating that fitness of different personality types might be context dependent. Fitness of
different personality types under various environmental conditions, as well as the exact
pathway through which these personality traits affect the outcome of duels warrants further
investigations.
16
The remaining suggested scenario, that there is no causal relationship between social status
and behavioural response, was observed for only one behavioural response: boldness. There
was no difference in latency to enter the novel arena between males of different social status,
and the change in social status did not alter the latency to enter the arena in the second trial.
However, a recent study of sea anemones (Actinia equine; Rudin and Briffa 2012) showed
that boldness (measured as response to a startle) decreased after the experience of losing a
fight. This indicates that the social environment can affect also boldness in some cases.
The results of this thesis clearly show that social status must be taken into account during
studies of individual behaviour in group-living animals. It is preferable that the social status
of the individual, and the eventual changes of the same, is known during the course of the
study. Other properties of the individual like quality of territory or partner, properties may
potentially affect individual behaviour in a test context in a corresponding way in solitary
species or species with a less clear hierarchical structure of the dominance relationships.
Future prospects
Under natural conditions, several aggressive interactions may follow after one another. In a
socially hierarchical species like the fowl, individuals need to settle the dominance
relationships to all other group members of the same sex, for example when an individual
transfers from one flock to another (Collias and Collias 1996). The ability to recover from a
loss, or the ‘boost’ obtained by a win, have proved to affect the outcome of subsequent duels,
the so-called ‘winner- and loser effects’ (Chase et al. 1994), but it is unknown whether the
magnitude of these effects depend on personality type. In future studies I therefore aim to
investigate how several interactions after each other affect male fowl of different personality
types to improve our understanding of how personality type affect fitness through social
dominance.
17
CONCLUSION
In this thesis I aimed to disentangle the commonly observed relationship between variation in
behaviour and social status. I conclude that behavioural differences between dominant and
subordinate male fowl were partly plastic and changed with changes in social status, but also
that the social position of an individual could be a consequence of pre-existing differences in
behavioural style. Plasticity of behavioural responses according to the current social position
have been previously shown in the domestic fowl (Cornwallis and Birkhead 2008), but it is
the first time the effect of personality type on the establishment of social status is shown in
the species. Together, the results of paper I and II reveal a two-way causality between social
position and individual behaviour. This leads to the conclusions that social position should be
taken into account when investigating and interpreting variation in personality, as it could
otherwise act as a confounding factor, and that individual personality type should vice versa
be considered in studies of the underlying reasons of variation in social status.
18
ACKNOWLEDGEMENTS
Först och främst vill jag rikta ett stort och varmt TACK till min nuvarande och före detta
handledare Olof Leimar och Tommy Radesäter samt min medförfattare och mentor Hanne
Løvlie, självklart för all hjälp med arbetet med de två studierna i denna avhandling, men
också för att ni uppmuntrar och utmanar mig att utvecklas till en självständig forskare.
Hanne, du ställer alltid upp när jag behöver dryfta stort som smått inom forskningsprojektet,
en riktig klippa, och du har dessutom initierat mig till hönsens fascinerande värld, vilket jag
innerligt glad för. Olle, dig vill jag särskilt tacka för att du generöst delar med dig av
statistisk kunskap och för att du med knivskarp skärpa hjälper mig att förstå de data jag
samlar in.
Jag vill också rikta ett hjärtligt tack till all personal på Tovetorps forskningsstation; Nisse,
Thomas, Sven och "damerna i köket", för att ni alltid är hjälpsamma med allsköns spörsmål,
och för våra glada fikastunder. Utan Minna, Siw, Berit och Anette hade det inte blivit
mycket till ordning på allt administrativt runtomkring, så det vill jag tacka er för. Martin och
Karolina, ni är de bästa rumskompisar man kan önska, i det finaste rummet på hela
institutionen, och jag känner mig verkligen glad när jag varje morgon jäktad ramlar in på
rummet till ert sällskap. Tack också till Sandra, Bea, Kristin, Johanna, Alexandra,
Alexander och alla andra härliga kolleger på Zootis, för alla skratt och alla intressanta
vetenskapliga och ovetenskapliga diskussioner kring lunchbordet (men måste de verkligen
varje gång inkludera någon form av bajsprat? ;).
Tack mamma, för att du alltid stöttar mig och får mig att tro på mig själv. Tack Rikard och
hela tjocka släkten för alla roliga släktträffar som har en omedelbar avslappnande effekt, trots
kakafonin. Tack Magnus för din outsinliga kärlek som bär mig genom tuffa perioder, och för
att du osjälviskt ger mig utrymme att jobbajobbajobba när det behövs. Du är bäst! Och till
sist, tack mina älskade pojkar Vide, Grim och Björn för att ni finns, och gör mitt liv så
mycket bättre!
19
REFERENCES
Andersson, M. 1994. Sexual selection. Princeton University Press, New Jersey.
Arakawa, H. 2006. Changes in the pattern of exploratory behavior are associated with the emergence of social dominance relationships in male rats. Developmental Psychobiology 48:39-47.
Banks, E. M. 1956. Social-organization in red jungle fowl hens (Gallus-gallus-subsp) Ecology 37:239-248.
Bell, A. M. 2005. Behavioural differences between individuals and two populations of stickleback (Gasterosteus aculeatus). Journal of Evolutionary Biology 18:464-473.
Berglund, A., A. Bisazza, and A. Pilastro. 1996. Armaments and ornaments: An evolutionary explanation of traits of dual utility. Biological Journal of the Linnean Society 58:385-399.
Briffa, M., S. D. Rundle, and A. Fryer. 2008. Comparing the strength of behavioural plasticity and consistency across situations: animal personalities in the hermit crab Pagurus bernhardus. Proceedings of the Royal Society B-Biological Sciences 275:1305-1311.
Carere, C., D. Caramaschi, and T. W. Fawcett. 2010. Covariation between personalities and individual differences in coping with stress: Converging evidence and hypotheses. Current Zoology 56:728-740.
Chase, I. D. 1980. Social-Process and Hierarchy Formation in Small-Groups - a Comparative Perspective. American Sociological Review 45:905-924.
Chase, I. D., D. Bartolomeo, and L. A. Dugatkin. 1994. Aggressive interactions and inter-contest interval: how long do winners keep winning? Animal behaviour 48:393-400.
Cheng, K. M. and J. T. Burns. 1988. Dominance relationship and mating behavior of domestic cocks – a model to study mate-guarding and sperm competition in birds. Condor 90:697-704.
Colleter, M. and C. Brown. 2011. Personality traits predict hierarchy rank in male rainbowfish social groups. Animal Behaviour 81:1231-1237.
20
Collias, N., E. Collias, and R. I. Jennrich. 1994. Dominant red jungle fowl (Gallus-gallus) hens in an unconfined flock rear the most young over thier lifetime. Auk 111:863-872.
Collias, N. E. 1943. Statistical Analysis of Factors Which Make for Success in Initial Encounters between Hens. The American Naturalist 77:519-538.
Collias, N. E. and E. C. Collias. 1996. Social organization of a red junglefowl, Gallus gallus, population related to evolution theory. Animal Behaviour 51:1337-1354.
Collias, N. E., E. C. Collias, and P. Saichuae. 1964. A field study of the Red jungle fowl (Gallus gallus) in Southeast Asia and India. American Zoologist 4:272-273.
Cornwallis, C. K. and T. R. Birkhead. 2007. Changes in sperm quality and numbers in response to experimental manipulation of male social status and female attractiveness. American Naturalist 170:758-770.
Cornwallis, C. K. and T. R. Birkhead. 2008. Plasticity in reproductive phenotypes reveals status-specific correlations between behavioral, morphological, and physiological sexual traits. Evolution 62:1149-1161.
Dahlbom, S. J., D. Lagman, K. Lundstedt-Enkel, L. F. Sundstrom, and S. Winberg. 2011. Boldness Predicts Social Status in Zebrafish (Danio rerio). Plos One 6:7.
Dall, S. R. X., A. I. Houston, and J. M. McNamara. 2004. The behavioural ecology of personality: consistent individual differences from an adaptive perspective. Ecology Letters 7:734-739.
David, M., Y. Auclair, and F. Cezilly. 2011. Personality predicts social dominance in female zebra finches, Taeniopygia guttata, in a feeding context. Animal Behaviour 81:219-224.
Davis, D. E. and L. V. Domm. 1943. The influence of hormones on the sexual behavior of domestic fowl. Pages 169–181) Essays in biology in honor of Herbert M. Evans. University of California Press, Berkeley.
Dingemanse, N. J., C. Both, P. J. Drent, and J. M. Tinbergen. 2004. Fitness consequences of avian personalities in a fluctuating environment. Proceedings of the Royal Society of London Series B-Biological Sciences 271:847-852.
Dingemanse, N. J. and P. de Goede. 2004. The relation between dominance and exploratory behavior is context-dependent in wild great tits. Behavioral Ecology 15:1023-1030.
21
Dingemanse, N. J. and D. Reale. 2005. Natural selection and animal personality. Behaviour 142:1159-1184.
Dingemanse, N. J. and M. Wolf. 2010. Recent models for adaptive personality differences: a review. Philosophical Transactions of the Royal Society B-Biological Sciences 365:3947-3958.
Dzieweczynski, T. L. and L. M. Forrette. 2013. Reproductive state but not recent aggressive experience influences behavioral consistency in male Siamese fighting fish. Acta Ethologica 16:31-40.
Evans, C. S., L. Evans, and P. Marler. 1993. On the meaning of alarm calls - functional reference in an avian vocal system. Animal Behaviour 46:23-38.
Fox, R. A., L. D. Ladage, T. C. Roth, and V. V. Pravosudov. 2009. Behavioural profile predicts dominance status in mountain chickadees, Poecile gambeli. Animal Behaviour 77:1441-1448.
Fumihito, A., T. Miyake, S. I. Sumi, M. Takada, S. Ohno, and N. Kondo. 1994. One subspecies of the Red junglefowl (Gallus gallus gallus) suffies as the matriarchic ancestor of all domestic breeds. Proceedings of the National Academy of Sciences of the United States of America 91:12505-12509.
Gomez-Laplaza, L. M. 2002. Social status and investigatory behaviour in the angelfish (Pterophyllum scalare). Behaviour 139:1469-1490.
Gosling, S. D. 2001. From mice to men: What can we learn about personality from animal research? Psychological Bulletin 127:45-86.
Groothuis, T. G. G. and C. Carere. 2005. Avian personalities: characterization and epigenesis. Pages 137-150.
Houston, A. I. and J. M. McNamara. 1999. Models of adaptive behaviour: an approach based on state. Cambridge University Press, New York.
Huntingford, F. A. 1976. Relationship between anti-predator behavior and aggression amonf conspecifics in 3-apined stickleback, Gasterosteus aculeatus. Animal Behaviour 24:245-260.
Johnsen, T. S. and M. Zuk. 1995. Testosterone and aggression in male red jungle fowl. Hormones and Behavior 29:593-598.
22
Johnsen, T. S., M. Zuk, and E. A. Fessler. 2001. Social dominance, male behaviour and mating in mixed-sex flocks of red jungle fowl. Behaviour 138:1-18.
Jones, R. B., H. J. Blokhuis, and G. Beuving. 1995. Open-field and tonic immobility responses in domestic chicks of 2 genetic lines differing in their propensity to feather peck. British Poultry Science 36:525-530.
Koolhaas, J. M., S. M. Korte, S. F. De Boer, B. J. Van Der Vegt, C. G. Van Reenen, H. Hopster, I. C. De Jong, M. A. W. Ruis, and H. J. Blokhuis. 1999. Coping styles in animals: current status in behavior and stress-physiology. Neuroscience and Biobehavioral Reviews 23:925-935.
Korzan, W. J., Ø. Øverli, and C. H. Summers. 2006. Future social rank: forecasting status in the green anole (Anolis carolinensis). Acta Ethologica 9:48-57.
Kruijt, J., P. 1964. Ontogeny of social behaviour in Burmese red junglefowl (Gallus gallus Spadiceus) Bonnaterre. Behaviour Supplement:I-IX, 1-201.
Leonard, M. L. and A. G. Horn. 1995. Growing in relation to status in roosters. Animal Behaviour 49:1283-1290.
Ligon, J. D., R. Thornhill, M. Zuk, and K. Johnson. 1990. Male male competition, ornamentation and the role of testosterone in sexual selection in red jungle fowl. Animal Behaviour 40:367-373.
Luttbeg, B. and A. Sih. 2010. Risk, resources and state- dependent adaptive behavioural syndromes. Philosophical Transactions of the Royal Society B-Biological Sciences 365:3977-3990.
McBride, G., I. P. Parer, and F. Foenander. 1969. The social organisation and behaviour of the feral domestic fowl. Baillière, Tindall & Cassell, London.
McGhee, K. E. and J. Travis. 2010. Repeatable behavioural type and stable dominance rank in the bluefin killifish. Animal Behaviour 79:497-507.
Nelson, X. J., D. R. Wilson, and C. S. Evans. 2008. Behavioral Syndromes in Stable Social Groups: An Artifact of External Constraints? Ethology 114:1154-1165.
Parker, T. H., R. Knapp, and J. A. Rosenfield. 2002. Social mediation of sexually selected ornamentation and steroid hormone levels in male junglefowl. Animal Behaviour 64:291-298.
23
Pizzari, T. and T. R. Birkhead. 2000. Female feral fowl eject sperm of subdominant males. Nature 405:787-789.
Quinn, J. L. and W. Cresswell. 2005. Personality, anti-predation behaviour and behavioural plasticity in the chaffinch Fringilla coelebs. Behaviour 142:1377-1402.
Reale, D., N. J. Dingemanse, A. J. N. Kazem, and J. Wright. 2010. Evolutionary and ecological approaches to the study of personality. Philosophical Transactions of the Royal Society B-Biological Sciences 365:3937-3946.
Reale, D., S. M. Reader, D. Sol, P. T. McDougall, and N. J. Dingemanse. 2007. Integrating animal temperament within ecology and evolution. Biological Reviews 82:291-318.
Rudin, F. S. and M. Briffa. 2012. Is boldness a resource-holding potential trait? Fighting prowess and changes in startle response in the sea anemone, Actinia equina. Proceedings of the Royal Society B-Biological Sciences 279:1904-1910.
Sih, A., A. Bell, and J. C. Johnson. 2004a. Behavioral syndromes: an ecological and evolutionary overview. Trends in Ecology & Evolution 19:372-378.
Sih, A., A. M. Bell, J. C. Johnson, and R. E. Ziemba. 2004b. Behavioral syndromes: An integrative overview. Quarterly Review of Biology 79:241-277.
Smith, B. R. and D. T. Blumstein. 2008. Fitness consequences of personality: a meta-analysis. Behavioral Ecology 19:448-455.
Stamps, J. A. 2007. Growth-mortality tradeoffs and 'personality traits' in animals. Ecology Letters 10:355-363.
van Oers, K., P. J. Drent, P. de Goede, and A. J. van Noordwijk. 2004. Realized heritability and repeatability of risk-taking behaviour in relation to avian personalities. Proceedings of the Royal Society of London Series B-Biological Sciences 271:65-73.
Verbeek, M. E. M., A. Boon, and P. J. Drent. 1996. Exploration, aggressive behavior and dominance in pair-wise confrontations of juvenile male great tits. Behaviour 133:945-963.
Verbeek, M. E. M., P. J. Drent, and P. R. Wiepkema. 1994. Consistent individual differences in early exploratory behavior of male great tits. Animal Behaviour 48:1113-1121.
24
Wilson, D. R., K. L. Bayly, X. J. Nelson, M. Gillings, and C. S. Evans. 2008. Alarm calling best predicts mating and reproductive success in ornamented male fowl, Gallus gallus. Animal Behaviour 76:543-554.
Wilson, D. R., X. J. Nelson, and C. S. Evans. 2009. Seizing the Opportunity: Subordinate Male Fowl Respond Rapidly to Variation in Social Context. Ethology 115:996-1004.
Wilson, D. S., A. B. Clark, K. Coleman, and T. Dearstyne. 1994. Shyness and boldness in humans and other animals. Trends in Ecology & Evolution 9:442-446.
Wolf, M. and F. J. Weissing. 2010. An explanatory framework for adaptive personality differences. Philosophical Transactions of the Royal Society B-Biological Sciences 365:3959-3968.
Zuk, M. and T. S. Johnsen. 2000. Social environment and immunity in male red jungle fowl. Behavioral Ecology 11:146-153.
Øverli, Ø., W. J. Korzan, E. Hoglund, S. Winberg, H. Bollig, M. Watt, G. L. Forster, B. A. Barton, E. Overli, K. J. Renner, and C. H. Summers. 2004. Stress coping style predicts aggression and social dominance in rainbow trout. Hormones and Behavior 45:235-241.
Øverli, Ø., C. Sorensen, K. G. T. Pulman, T. G. Pottinger, W. Korzan, C. H. Summers, and G. E. Nilsson. 2007. Evolutionary background for stress-coping styles: Relationships between physiological, behavioral, and cognitive traits in non-mammalian vertebrates. Neuroscience and Biobehavioral Reviews 31:396-412.
1
Paper I
1
Social ‘states’ and personality: Stability in social status and
individual characteristics explain consistency in behavioural
responses
Anna Favati1, Olof Leimar1, Tommy Radesäter1 and Hanne Løvlie1,2*
1 Department of Zoology, Stockholm University, 106 91 Stockholm, Sweden
2 Department of Physics, Chemistry and Biology, Linköping University, 581 83 Linköping,
Sweden
*Corresponding author: Hanne Løvlie, [email protected]
2
ABSTRACT
Stability of ‘state’ has been suggested as an underlying factor explaining behavioural
stability and animal personality (i.e., variation among, and consistency within individuals in
behavioural responses), but the possibility that stable social relationships represent such
states remains unexplored. Here we investigated the influence of social status on the
expression and consistency of behaviours by experimentally changing social status between
repeated personality assays. We used male domestic fowl (Gallus gallus domesticus), a
social species that forms relatively stable dominance hierarchies, and showed that
behavioural responses were strongly affected by social status, but also by individual
characteristics. The level of vigilance, activity and exploration changed with social status,
while boldness appeared as a stable individual property, independent of status. Further,
variation in vocalisation predicted future social status, indicating that individual behaviours
can both be a predictor and a consequence of social status, depending on the aspect in
focus. Our results illustrate that social states contributed to both variation and stability in
behavioural responses, and should therefore be taken into account when investigating and
interpreting variation in personality.
Keywords: behavioural syndromes; intra-‐sexual selection; phenotypic plasticity; social
dominance; chicken.
3
1. INTRODUCTION
Understanding variation in phenotypes is a key issue in evolutionary biology [1]. The topic
includes the study of animal personality, in the form of consistent individual differences in
behaviour, across time and/or situations [2], also referred to as temperaments [3] or coping
styles [4]. The phenomenon has been described in a large number of species in multiple taxa
[5], but there are still major gaps in our understanding of why there is personality variation,
including unanswered questions about the mechanisms behind stable behavioural
responses and the evolution and maintenance of behavioural polymorphism [6, 7].
Properties or circumstances of an individual that may affect the costs and benefits of its
behaviours, such as size or energy reserves are sometimes referred to as ‘states’ [8].
Stability of such states is theoretically predicted to produce stability in behavioural
responses which, in combination with between-‐individual variation in state, gives rise to
variation in behavioural types, and thus personality [2, 6, 9]. More broadly, any state
changing more slowly than behaviour per se is predicted to cause short-‐term stability in
state-‐dependent behavioural responses [2]. Together with a positive feedback system,
stability in state may also generate long-‐term stability of behavioural types [9]. Social
relationships such as pair bonds or status hierarchies could be one, however not yet
investigated, example of such states. If social positions or relationships constitute stable
states, intra-‐individual stability of behavioural responses would follow as a consequence,
whereas the behavioural responses should change when the social state of an individual is
changed. Despite the intuitiveness of these predictions, the importance of social states for
personality variation has not yet been empirically tested.
In social species, social relationships often take the form of dominance hierarchies, which in
turn are based on repeated outcomes in favour of one participant of dyadic agonistic
interactions [10]. Socially dominant individuals commonly enjoy increased access to
resources, such as mating partners, which typically results in a positive relationship between
social status and reproductive success [11]. Aggression and the ability to dominate
conspecifics can correlate positively with boldness, exploration and active stress handling,
thereby defining a ‘proactive’ behavioural style of the reactive-‐proactive coping style
continuum [4, 12]. For example, explorative great tits (Parus major) win more fights
compared to less explorative ones [13], and bold three-‐spined sticklebacks (Gasterosteus
4
aculeatus) are more aggressive than shyer individuals [14]. On the other hand, boldness and
exploration have been found to have a negative, or no correlation, with social dominance
[15, 16]. This indicates that the relationship between personality and social status can be
species-‐specific, but also that there are currently limitations to our understanding of the
relationship between them.
In principle, there are three possible scenarios for observed correlations between social
dominance and personality traits. First, different social positions can be associated with
different behavioural tendencies [17]. These differences can manifest themselves also
outside the social group, and thus influence responses in personality assays [18]. In this
scenario, it is expected that behavioural responses are flexible, adjusting to the current
social position, at least to some extent. Second, differences in behaviour can directly
influence the chance of obtaining a certain social position (e.g., aggression [19]). In this
scenario, certain behavioural types are more likely to be found in specific social positions.
However, and crucially, behavioural responses are independent of experimental change of
social positions. Third (and partially overlapping with the previous), behaviour and social
position can have a common underlying cause [12, 20]. In such a case, both personality
traits and the underlying cause (e.g., hormonal state [20]) might predict social position, but
behaviour is not necessarily altered when the social position is changed. Although there is
some support for each of the three scenarios separately, they have not previously been
tested simultaneously, leaving the causality between behaviour and social status unclear.
Our aim here is therefore to investigate the issue by means of an experimental approach
where social status of individual male domesticated fowl (Gallus gallus domesticus) is
changed and behaviour scored in personality assays. If a change in dominance status leads
to a change in behavioural response, we can conclude that social status may represent a
state that gives rise to consistent differences in behaviour.
Male fowl are suitable for investigating the potential link between variation in personality
and social status for two main reasons. First, dominance hierarchies are relatively stable; the
status of an individual persists for approximately one breeding season in free-‐ranging flocks
of fowl [21], but changes in groups over shorter time spans can occur and cause changes in
the status of individuals (e.g., if the dominant male is predated [22]). Therefore, status can
be considered as a slowly changing yet not permanent state. Second, dominant and
5
subordinate male fowl differ quantitatively in behaviour. For example, dominant males crow
more, as a signal of territoriality, and spend more time being vigilant compared to
subordinate males [17, 23]. Nevertheless, the causal relationship between these
behavioural differences and status is untested.
2. METHODS
Animals and housing
The study took place during the breeding seasons (May -‐ September) of 2007 and 2011 at
Tovetorp Research station, Stockholm University, Sweden. We used 84 males (2007: n = 36;
2011: n = 48) of an old Swedish game breed of fowl (Gallus gallus domesticus, ‘Gammal
svensk dvärghöna’) used to human handling and kept under semi-‐natural conditions in
mixed-‐sex, mixed-‐age (1-‐9 years) groups. Data were collected during the hours of day when
the birds are most active (0530-‐1100 and 1500-‐1900 hours local time, see [24] for further
information on variation in daily behavioural patterns). Groups of four males (ngroups = 21)
were matched for morphological measures (< 10% difference in comb and body size) to
reduce the effect these variables can have on the establishment of status [19, 25, 26]. Comb
size and body size were estimated as comb length and tarsus length respectively, measured
with a digital calipeter to the nearest tenth of a millimetre. Body size was also estimated as
body weight measured with a digital scale to the nearest gram. To generate males of
different social positions, two pairs of males were randomly chosen from each of the
matched groups of four males (npairs = 42). Pairs of males were housed in outdoor aviaries
(approximately 3 x 3 meters) that were visually, but not vocally, isolated from other birds.
This set up resulted in one of the males in each pair becoming dominant, and the other
subordinate. A minimum of five observed successive submissive behaviours within 2 hours
(avoiding the other male when he approached) by a male defined that male as subordinate
[27]. Males within a pair had not been housed together for at least 2 weeks prior to the
trials, reducing any effects earlier encounters may have had on the establishment of social
status [28, 29]. All birds had ad libitum access to food and water, and all aviaries had dust
baths and perches. Experiments were conducted according to ethical requirements in
Sweden (Linköping Ethical committee, ethical permission number 60-‐10).
Behavioural assays and manipulations of social status
6
Two days after the pair was formed, which was enough time for males to establish a
dominance relationship without reversals, each bird was exposed to its first ‘Novel arena’
test (referred to as trial 1) in a 3 x 6 meter large outdoor area connected to the aviary by a
door. The order of testing of the males in a pair was random with respect to social
dominance. The floor in the arena was covered with dark brown peat to create novelty (in
comparison to the sandy floor of the birds’ home pens) and divided into 8 subareas by
drawing subtle lines in the floor substrate to facilitate estimates of exploration (see below).
The arena had five artificial bushes (green plastic spruce trees, approx. 70 x 50 cm) in order
to obstruct direct view of the arena and thus encourage exploration (see S.I., figure S1). The
cage mate of the focal male was removed from the aviary by herding him into an empty
adjacent pen two minutes before the test started. The door between the aviary and the test
arena was opened, and focal males that entered the arena within 30 seconds were
considered bold, while those who did not were considered shy (i.e., a ‘boldness’ spectrum
split into two categories, becasue of a bimodal distribution of the latencies). In cases when a
male did not voluntarily enter the arena within 2 min, he was gently herded into it by the
observer. The observer sat outside the arena, visible to the male during the test. Behaviours
and vocalisations were recorded for 15 min, starting when the male entered the arena.
Vigilance (head above shoulder height, eyes open) and behaviours categorised as non-‐
vigilance (eating, preening, resting and dust bathing, behaviours that all occurred in lower
frequencies) were recorded every 30s, and the percentage of recordings a male was vigilant
during the test was calculated (hereafter named ‘vigilance’). The number of unique subareas
a male visited (1-‐8) divided by the total number of possible subareas (8) was used as a
measure of ‘exploration’ propensity. The number of times a male entered a subarea he had
previously visited, was noted as a measure of ‘activity’. Number of crows were recorded ad
libitum (‘crowing’). Crowing is a species-‐specific vocalisation related to territoriality in male
fowl [30]. In addition to crowing, some males uttered alarm calls, which are loud and
distinct warning vocalisations typically uttered when an individual is startled, for example
when a predator appears [31], but these few vocalisations were not analysed further.
When the behaviour of all four males of a matched group had been scored in the first novel
arena test, pair members were experimentally changed so that the two previously dominant
males formed a new pair, and the two subordinate males another pair. The new pair setup
7
forced one male in each pair to change social status, as there can only be one dominant
male and one subordinate male in each pair [see 18 for previous use of this design]. After
two days of acclimation to the new social context, the novel arena test was performed a
second time in order to estimate the influence of change of social status on behaviour [26].
We refer to this test as trial 2. For the second pair setup, the number of males was reduced
to 82 due to the death of one male in 2007 and injury of one male in 2011. A. F. conducted
all observations.
Statistical analyses
Our general approach to the statistical modeling of the behavioral responses was to first
determine suitable Box-‐Cox transformations of the continuously distributed responses, in
order to achieve homogeneous variances. The outcome of this procedure was that the
measure vigilance did not need transformation, the transformation (𝑥 + 1)!.!" was used for
activity and crowing, and the transformation 𝑥! for exploration. We applied these
transformations in all analyses, including model selection and Bayesian MCMC modeling,
but we show the untransformed variables in the figures. The binary measure of boldness
was treated as a discrete variable. Based on our experimental design, with males being
divided into matched groups, we used the group as a random effect in mixed models of the
behavioral responses.
Difference in behavioural responses of dominant and subordinate males in trial 1
To investigate differences in behavioural response between dominant and subordinate
males (n = 84), we analysed the effect of social status on each response (boldness, vigilance,
activity, exploration and crowing) in trial 1, by fitting mixed models with the trial 1 social
status as a fixed effect and the group as a random effect (Bayesian Markov Chain Monte
Carlo generalized linear mixed models, MCMCglmm, [32]).
Individual characteristics versus the effect of social status on behavioural responses
Behavioural response in the second trial can be dependent on behavioural response in the
first trial (i.e., personality), current social status (thus behaviour is a plastic response) or the
previous social status (suggesting a common underlying variable). In order to disentangle
the effect of social status and personality on a behavioural response, we analysed the effect
8
of both ‘current status’ (obtained in the second pair setup, ‘dominant’ vs. ‘subordinate’) and
‘previous status’ (obtained in the first pair setup, ‘dominant’ vs. ‘subordinate’), and
behavioural response in the first trial on each of the responses in the second trial by
Bayesian MCMCglmm (n = 82). Random effect and error distributions were the same as in
the previous analyses (above). To determine final models, we used bidirectional stepwise
selection choosing the model with the lowest AIC value using Generalized linear mixed
model [33]. The minimal model for each response in the second trial was set to include the
corresponding response from the first trial, current status and previous status. The range of
models used in stepwise search of the best fitted model included all response variables in
the first trial, current status and previous status. We also estimated repeatability with and
without taking social status into account (see S.I.).
Factors predicting social status
Comb size, body size (tarsus length, and body weight) and age are traits that may affect the
establishment of status in male fowl [19]. We investigated whether the initial matching of
males for these properties was successful in both pair setups by testing for the difference in
comb size, body size, body weight and age between dominant and subordinate males in
each pair by means of a two-‐sided, one-‐sample student’s t-‐test. The level of this analysis
was thus the pair (ntrial1 = 42, ntrial2 = 40). To investigate whether behaviour may predict
social status, we tested if behavioural response in the first novel arena test could predict the
outcome of the duel in the second pair setup. Difference between dominant and
subordinate males in the binary response boldness was tested with Fisher’s exact test, while
difference in vigilance, activity, exploration and crowing between dominant and subordinate
male of each pair, were tested with a two-‐sided one-‐sample student’s t-‐test (npairs = 41).
All statistics analyses were conducted using R version 2.15.1 [34].
3. RESULTS
Differences in behavioural responses of dominant and subordinate males
Dominant males were more vigilant and active, and crowed more compared to subordinate
males, but did not differ significantly in boldness and exploration (table 1, figure 1).
Individual characteristics versus the effect of social status on behavioural responses
9
Analyses of factors affecting the observed behavioural responses of males in the second trial
showed that all response variables investigated had a statistically significant positive
relationship with the corresponding response in the first trial (table 2). At the same time, all
responses but boldness and crowing, were affected by the current social position (table 2,
illustrated by figure 2 showing variation in vigilance, for other variables see S.I. figure S2).
Vigilance, activity and exploration were flexible behaviours and changed to a large extent
with changed social status (table 2, figure 3b-‐d). This was especially pronounced for
vigilance (figure 3b), while for example boldness did not follow changes in status (figure 3a).
Males that had changed status clearly altered the time they spent being vigilant: males with
reduced status from dominant to subordinate showed a decrease in vigilance, while an
increase in vigilance was seen in males that improved their status from subordinate to
dominant (figure 3a). Males with stable status across the two experimental pair setups
(males that remained either dominant, or subordinate) showed a stable between-‐test
response (figure 3a). The number of crows uttered by males was dependent on the previous
social status: males that previously had been dominant crowed more in the second trial
compared to other males, irrespective of their current social status (table 2, figure 3e).
There were additional statistically significant factors influencing the behavioural responses
of males in the novel arena test of trial 2 (table 2). Part of variation in vigilance was
explained by the number of crows a male uttered in the first trial, thus males that crowed
more in the first trial were more vigilant in the second trial (table 2). In a similar way,
activity was partly explained by vigilance in trial 1, and crowing was affected by activity in
trial 1 (table 2).
Factors predicting social status
None of the measured morphological traits, or age, had a detectable effect on the
establishment of social status in this experiment, indicating that the experimental matching
for these features was successful (first pair setup: tarsus length, t41 = -‐0.080, p = 0.94; body
weight, t41 = 0.36, p = 0.72; comb length, t41 = -‐1.03, p = 0.31; age, t41 = 0.49, p = 0.63,
second pair setup: tarsus length, t40 = 0.76, p = 0.45; body weight, t40 = 0.0072, p = 0.99;
comb length, t40 = 1.71, p = 0.09; age, t40 = -‐0.14, p = 0.89). On the other hand, the number
of crows a male uttered in the first trial predicted future social status in the second pair
10
setup; males that crowed more had a higher probability of becoming dominant compared to
males that crowed less (t40 = 2.59, p = 0.01). No other behavioural responses predicted
future status (boldness, p = 0.58; vigilance, t40 = 0.41, p = 0.68; activity, t40 = -‐0.73, p = 0.47;
exploration, t40 = 0.78, p = 0.44).
4. DISCUSSION
Our experiment showed that behavioural responses of male domestic fowl in behavioural
assays are affected both by the social status of an individual and by its individual
characteristics. Our results support the general interpretation of behaviours being plastic
but also showing constraints in flexibility, in line with the few empirical studies where both
possibilities have been investigated [35, 36]. How the two factors social position and
individual characteristics explain variation and stability in a behavioural response depends
on the behavioural response in focus, and will be discussed below.
We introduced three scenarios for how a relationship between social position and
behavioural responses can come about. By exploring different aspects of behaviour of male
fowl across changes in social position, we found experimental support for two of the three
scenarios, namely that the social position in itself influences behaviour, and that social
position and behaviour may share a common, underlying cause. According to the first of
these, frequencies of behaviours are expected to change when social status changes,
thereby putting a limit to the consistency of the individual responses. We found this to be
the case for three of the recorded responses: activity, exploration and vigilance. Activity and
exploration are commonly used behavioural dimensions in personality studies [3], but the
effect of social status on their expression has previously not been investigated. In the fowl,
dominant males guard territories [30]. The increased activity and exploration observed
when a male becomes dominant may thus be explained by an increased propensity to patrol
a territory and its surroundings. Our results support the general idea that individuals
obtaining dominant social positions are more often observed to behave in a proactive way,
rather than in a reactive manner [4, 13, 20]. They further suggest that an individual’s degree
of activity and exploration is partly a consequence of the current social situation, and stays
constant if social status is unchanged, but changes when social status changes. This is in line
11
with the reasoning of Dall et al. [2], who suggested that stability in various states can
underlie consistency in behavioural responses.
Vigilance decreases predation risk by facilitating early detection of predators and can be
considered a response to an individual’s assessment of risks in its environment [37]. Both in
our study and as previously shown [17], dominant males were more vigilant, and the degree
of vigilance changed dramatically when status changed, which can be interpreted as a risk-‐
averse response to protect increased assets (e.g., access to females as sexual partners) [6].
Such a response follows the principle of asset protection, in that individuals should increase
their cautiousness when assets increase [38, 39]. However, it has been argued that
differences in state should converge over time, as cautiousness is assumed to lead to a
decrease of assets [9]. We do, however, not expect increased access to females -‐ as a
consequence of high social status -‐ to be negatively affected by an increased vigilance, thus
the relationship between asset and behaviour in this case need not function as a negative
feedback mechanism that causes a convergence of individual states. Our results indicate
that social status could act as a stabilising condition, or ‘state’, generating at least short-‐
term stability in behavioural responses [39, 40]. Whether long-‐term stability is achievable in
a similar way remains to be studied. This could happen if, for example, early experiences of
dominance positions lead to long-‐term individual differences in physiology or morphology
that later in life affect the chances of obtaining a certain social position.
We also found support for the scenario that behaviour and social dominance share common
underlying causes. The number of crows uttered by a male was correlated with both the
previous social status and the behavioural type of an individual, indicating that individual
males vary in their underlying propensity to crow and that this propensity can influence the
chances of becoming dominant. The observation that dominant males crow more than
subordinates has been reported previously in the species [17, 23], but to our knowledge our
study is the first observation of an effect of social position on number of crows uttered
outside of a social context. Also, we found that when two dominant, or two subordinate
males, were paired up in preparation for the second trial, the male that had crowed more in
the preceding novel arena test more often became dominant. Our observations thus
suggest that crowing frequency partly reflects an underlying characteristic that influences
the establishment of social status [20, 41, 42]. Note, however, that crowing is not a
12
commonly observed behaviour during the interactions that determine social status; it rather
appears as a signal of the outcome [23, 43]. A potential underlying factor for crowing and
social dominance could be testosterone, which regulates development of crowing frequency
in male fowl [44]. A higher testosterone level during male-‐male interaction has been found
to be associated to shorter attack latencies and an increase in the probability of obtaining
high social status in male domestic fowl [45]. Further studies of the relationship between
crowing, behaviour during duels and hormonal status could help elucidate why crowing in a
non-‐social situation not only signals high social status, but also predicts it.
Our experimental design did not allow an explicit and thorough test of the remaining
scenario; that behaviour per se leads to a certain social position, because males in our study
had already obtained a social position prior to the first personality test. Behavioural
responses from personality assays have sometimes proved to be predictive of future social
status [4, 13], but that possibility is not mutually exclusive with the two other scenarios
presented here, and thus remains to be investigated further.
Of the responses studied here, only the level of boldness was unaffected by changes in
social status. Dominant males tended to be bolder than subordinate males in the first trial,
but there was no change in boldness when status was changed. Compared to the other
behaviours we investigated, boldness therefore appears to be less flexible, in being
independent of changes in social status. However, a recent study of sea anemones (Actinia
equina [46]) showed that the experience of a fight, and in particular for the losing part,
decrease boldness (latency to retain regular activity after being startled). This indicates that
the social environment can affect also boldness.
In addition to the sometimes striking effect of social status on behavioural responses in our
study, there was always a remaining correlation between the responses of individuals across
the two test occasions. This consistency in behaviour across social situations is an
expression of individual characteristics that fall under the general heading of animal
personality, but the precise nature and general significance of these characteristics is an
open question. Studies on other species have shown that heritable components, together
with early life experiences (e.g., maternal effects), can partly explain variation and stability
of individual behaviour [47, 48]. Less attention has been directed to how intrinsic or
13
environmental factors later in life may shape personality. An recent experimental study on
great tits (Parus major) showed that breeding effort is a potential example of a non-‐genetic
cause of personality variation [49], in a broadly similar way as we found here for social
positions. Our results emphasises the need for controlling for stability in the social context
that is of importance for the species in question (e.g., social status and group composition),
and possibly stability also in other contexts or dynamic states, to further improve our
understanding of causes and consequences of variation in personality.
ACKNOWLEDGEMENTS
We are grateful to N. Andbjer for game keeping and L. Nauman for field assistance. OL was
supported by a grant (621-‐2010-‐5437) from the Swedish Research Council and by the
Strategic Research Program Ekoklim at Stockholm University. HL was supported by a
fellowship from the Swedish Research Council and by a grant from the Längmanska
Foundation.
14
Tables
Table 1. Behavioural responses of dominant and subordinate male fowl in the first trial of a
novel arena test (see main text for further details).
response dominant subordinate PMCMC
boldness 64% 45% 0.060 vigilance 0.79 ± 0.02 0.55 ± 0.03 <0.001 activity 23.00 ± 2.36 12.81 ± 1.97 <0.001 exploration 7.26 ± 0.17 6.69 ± 0.24 0.066 crowing 14.79 ± 1.56 8.55 ± 1.52 <0.001 Values presented are mean ± SE for all responses except boldness, where the percentages of bold males are
shown. PMCMC-‐values < 0.05 are highlighted in bold.
Table 2. Factors explaining variation in behavioural responses of male fowl in the second
trial of a novel arena test (see main text for further details).
response trial 2 response trial 1 status trial 1 status trial 2 additional variables included in model (from trial 1)#
boldness 2.11*** -‐0.45 -‐0.37 vigilance 0.27** -‐0.05 0.18*** crowing, 0.08*
activity 0.42*** -‐0.06 0.23** vigilance, -‐0.84**; boldness, -‐0.14 exploration 0.37** -‐11.77 85.38* boldness, 70.46 crowing 0.37*** 0.33*** 0.09 activity, 0.22*; boldness, -‐0.15 Values presented are posterior means of model parameters in the final MCMCglmm analyses, with effects of
status given as dominant-‐subordinate differences. See main text for the variable transformations that were
used in the model fitting. The star symbols indicate statistical significance: ‘*’ PMCMC < 0.05, ‘**’ PMCMC<0.01,
‘***’ PMCMC<0.001. #Inclusion of certain behavioural responses from trial 1 contributed to a better model fit.
15
Figures
Figure 1. Behavioural response of dominant and subordinate males in the first trial of a novel arena test.
Dominant males (‘dom’, filled dots) (a) tended to be bolder (i.e., more often entered the arena within 30
seconds), (b) were more vigilant (i.e., had higher frequency of time spent being vigilant), (c) were more active
(i.e., conducted more sub-‐area transitions), (d) tended to be more explorative (i.e., visited more sub-‐areas),
and (e) crowed more often (i.e., number of crows) compared to subordinate males (‘sub’, open dots) (table 1).
Values are given as mean ± SE, except for (a), where percentages of males that entered the arena within 30
seconds ± 95% CI, are given.
●
●
20
30
40
50
60
70
80
Boldness
dom sub
(a)
●
●
0
0.2
0.4
0.6
0.8
1.0
Vigilance
dom sub
***(b)
●
●
0
10
20
30
40
Activity
dom sub
***(c)
●
●
4
5
6
7
8
Exploration
dom sub
(d)
●
●
0
4
8
12
16
20
Crowing
dom sub
**(e)
16
Figure 2. The effect of current social status, and previous behavioural response (in trial 1) on vigilance in a
second novel arena test (trial 2). Males that were dominant (filled dots, solid line) spent more time being
vigilant compared to males that were subdominant (open dots, dashed line). At the same time, vigilance was
repeatable within individuals (demonstrated by a positive slope, table 2). Social status of males presented in
the figure was based on social status obtained in trial 2 (see main text for further details). See S.I. for similar
figures for other behaviours (figure S2).
●
●
●
●
●
●
●
●
●
● ●●
●
●●
●●
●
●
●
●●
●
●
●
●●
●
●●
●
● ●
●
●
●●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
0
0.2
0.4
0.6
0.8
1.0
0 0.2 0.4 0.6 0.8 1.0
Vigi
lanc
e tri
al 2
Vigilance trial 1
●
●
DominantSubordinate
17
Figure 3. Behavioural response of male fowl of different social status during the first (trial 1) and second test
trial (trial 2) in a repeated novel arena test. (a) Males of different status did not differ in boldness (i.e., equal
percentage of dominant, filled dots, and subordinate, open dots, males entered the arena within 30 seconds),
and stable (solid lines) vs. changed (dashed lines) social status had no significant effect on variation in
boldness. Dominant males were more (b) vigilant (frequency of time spent being vigilant), (c) active (i.e.,
conducted more sub-‐area transitions) and (d) exploratory (i.e., visited more sub-‐areas), compared to
subordinate males. Males that had a stable status across trials showed more stable behavioural responses
compared to males that changed social status, irrespective of their social status. (e) The number of crows
uttered in the second trial was significantly affected by previous, but not current, social status. Values
presented are mean ± SE, except for (a), where percentages of males that entered the arena within 30 seconds
± 95% CI, are given.
●
●●
●
●
●●
●
304050607080
trial 1 trial 2
Bold
ness
(a)
●
●●
●
●
●
●
●
0.4
0.5
0.6
0.7
0.8
0.9
trial 1 trial 2
Vigi
lanc
e
(b)
●
●
●
●
●
●
●●
10
15
20
25
30
35
40
trial 1 trial 2
Activ
ity
● DominantSubordinate
Stable statusChanged status
(c)
●
●
●
●
●
●
●
●
5
6
7
8
trial 1 trial 2
Expl
orat
ion
(d)
●
●
●
●
●
●
● ●
0
5
10
15
20
25
trial 1 trial 2
Cro
win
g
(e) ●
●
DominantSubordinateStable statusChanged status
18
REFERENCES
1. Mitchell-‐Olds T., Willis J.H., Goldstein D.B. 2007 Which evolutionary processes influence
natural genetic variation for phenotypic traits? Nat Rev Genet 8(11), 845-‐856.
(doi:10.1038/nrg2207).
2. Dall S.R.X., Houston A.I., McNamara J.M. 2004 The behavioural ecology of personality:
consistent individual differences from an adaptive perspective. Ecol Lett 7(8), 734-‐
739. (doi:10.1111/j.1461-‐0248.2004.00618.x).
3. Reale D., Reader S.M., Sol D., McDougall P.T., Dingemanse N.J. 2007 Integrating animal
temperament within ecology and evolution. Biol Rev 82(2), 291-‐318.
(doi:10.1111/j.1469-‐185X.2007.00010.x).
4. Koolhaas J.M., Korte S.M., De Boer S.F., Van Der Vegt B.J., Van Reenen C.G., Hopster H.,
De Jong I.C., Ruis M.A.W., Blokhuis H.J. 1999 Coping styles in animals: current status
in behavior and stress-‐physiology. Neurosci Biobehav Rev 23(7), 925-‐935.
(doi:10.1016/S0149-‐7634(99)00026-‐3).
5. Gosling S.D. 2001 From mice to men: What can we learn about personality from animal
research? Psychol Bull 127(1), 45-‐86. (doi:10.1037//0033-‐2909.127.1.45).
6. Dingemanse N.J., Wolf M. 2010 Recent models for adaptive personality differences: a
review. Phil Trans R Soc B 365(1560), 3947-‐3958. (doi:10.1098/rstb.2010.0221).
7. Reale D., Dingemanse N.J., Kazem A.J.N., Wright J. 2010 Evolutionary and ecological
approaches to the study of personality. Phil Trans R Soc B 365(1560), 3937-‐3946.
(doi:10.1098/rstb.2010.0222).
8. Houston A.I., McNamara J.M. 1999 Models of adaptive behaviour: an approach based on
state. New York, Cambridge University Press.
9. Luttbeg B., Sih A. 2010 Risk, resources and state-‐ dependent adaptive behavioural
syndromes. Phil Trans R Soc B 365(1560), 3977-‐3990. (doi:10.1098/rstb.2010.0207).
10 Chase I.D. 1980 Social-‐Process and Hierarchy Formation in Small-‐Groups -‐ a Comparative
Perspective. Am Sociol Rev 45(6), 905-‐924. (doi:Doi 10.2307/2094909).
11. Andersson M. 1994 Sexual selection. New Jersey, Princeton University Press.
12. Coppens C.M., de Boer S.F., Koolhaas J.M. 2010 Coping styles and behavioural flexibility:
towards underlying mechanisms. Phil Trans R Soc B 365(1560), 4021-‐4028.
(doi:10.1098/rstb.2010.0217).
19
13. Verbeek M.E.M., Boon A., Drent P.J. 1996 Exploration, aggressive behavior and
dominance in pair-‐wise confrontations of juvenile male great tits. Behaviour 133(11-‐
12), 945-‐963. (doi:10.1163/156853996X00314).
14. Huntingford F.A. 1976 Relationship between anti-‐predator behavior and aggression
among conspecific in 3-‐spined stickleback, Gasterosteus Aculeatus. Anim Behav
24(MAY), 245-‐260.
15. Gomez-‐Laplaza L.M. 2002 Social status and investigatory behaviour in the angelfish
(Pterophyllum scalare). Behaviour 139(11-‐12), 1469-‐1490.
(doi:10.1163/15685390260514726).
16. Fox R.A., Ladage L.D., Roth T.C., Pravosudov V.V. 2009 Behavioural profile predicts
dominance status in mountain chickadees, Poecile gambeli. Anim Behav 77(6), 1441-‐
1448. (doi:10.1016/j.anbehav.2009.02.022).
17. Cornwallis C.K., Birkhead T.R. 2008 Plasticity in reproductive phenotypes reveals status-‐
specific correlations between behavioral, morphological, and physiological sexual
traits. Evolution 62(5), 1149-‐1161. (doi:10.1111/j.1558-‐5646.2008.00346.x).
18. Arakawa H. 2006 Changes in the pattern of exploratory behavior are associated with the
emergence of social dominance relationships in male rats. Dev Psychobiol 48(1), 39-‐
47. (doi:10.1002/dev.20114).
19. Ligon J.D., Thornhill R., Zuk M., Johnson K. 1990 Male male competition, ornamentation
and the role of testosterone in sexual selection in red jungle fowl. Anim Behav 40(2),
367-‐373. (doi:10.1016/s0003-‐3472(05)80932-‐7).
20. Øverli Ø., Korzan W.J., Hoglund E., Winberg S., Bollig H., Watt M., Forster G.L., Barton
B.A., Overli E., Renner K.J., et al. 2004 Stress coping style predicts aggression and
social dominance in rainbow trout. Horm Behav 45(4), 235-‐241.
(doi:10.1016/j.yhbeh.2003.12.002).
21. Collias N.E., Collias E.C. 1996 Social organization of a red junglefowl, Gallus gallus,
population related to evolution theory. Anim Behav 51, 1337-‐1354.
(doi:10.1006/anbe.1996.0137).
22. McBride G., Parer I.P., Foenander F. 1969 The social organisation and behaviour of the
feral domestic fowl. London, Baillière, Tindall & Cassell.
23. Leonard M.L., Horn A.G. 1995 Growing in relation to status in roosters. Anim Behav
49(5), 1283-‐1290. (doi:10.1006/anbe.1995.0160).
20
24. Løvlie H., Pizzari T. 2007 Sex in the morning or in the evening? Females adjust daily
mating patterns to the intensity of sexual harassment. Am Nat 170(1), E1-‐E13.
(doi:10.1086/518180).
25. Parker T.H., Ligon J.D. 2002 Dominant male red junglefowl (Gallus gallus) test the
dominance status of other males. Behav Ecol Sociobiol 53(1), 20-‐24.
(doi:10.1007/s00265-‐002-‐0544-‐5).
26. Cornwallis C.K., Birkhead T.R. 2007 Changes in sperm quality and numbers in response to
experimental manipulation of male social status and female attractiveness. Am Nat
170(5), 758-‐770. (doi:10.1086/521955).
27. Cloutier S., Beaugrand J.P., Lague P.C. 1996 The role of individual differences and
patterns of resolution in the formation of dominance orders in domestic hen triads.
Behav Processes 38(3), 227-‐239. (doi:10.1016/s0376-‐6357(96)00034-‐4).
28. McBride G. 1958 The measurement of aggressiveness in the domestic hen. Anim Behav
6(1-‐2), 87-‐91. (doi:10.1016/0003-‐3472(58)90015-‐0).
29. Cloutier S., Beaugrand J.P., Lague P.C. 1995 The effect of prior victory or defeat in the
same site as that of subsequent encounter on the determination of dyadic
dominance in the domestic hen. Behav Processes 34(3), 293-‐298. (doi:10.1016/0376-‐
6357(95)00006-‐g).
30. Collias N.E., Collias E.C. 1967 A field study of Red jungle fowl in Northe-‐central India.
Condor 69(4), 360-‐&. (doi:10.2307/1366199).
31. Evans C.S., Evans L., Marler P. 1993 On the meaning of alarm calls -‐ functional reference
in an avian vocal system. Anim Behav 46(1), 23-‐38. (doi:10.1006/anbe.1993.1158).
32. Hadfield J.D. 2010 MCMC Methods for Multi-‐Response Generalized Linear Mixed
Models: The MCMCglmm R Package. J Stat Softw 33(2), 1-‐22.
33. Venables W.N., Ripley B.D. 2002 Modern Applied Statistics with S. Fourth ed. New York,
Springer.
34. R Development Core Team 2012 R: A language and environment for statistical
computing. Vienna, R Foundation for Statistical Computing.
35. Johnson J.C., Sih A. 2007 Fear, food, sex and parental care: a syndrome of boldness in
the fishing spider, Dolomedes triton. Anim Behav 74(5), 1131-‐1138.
(doi:10.1016/j.anbehav.2007.02.006).
21
36. Briffa M., Rundle S.D., Fryer A. 2008 Comparing the strength of behavioural plasticity
and consistency across situations: animal personalities in the hermit crab Pagurus
bernhardus. Proc R Soc B 275(1640), 1305-‐1311. (doi:10.1098/rspb.2008.0025).
37. Lima S.L. 1998 Stress and decision making under the risk of predation: Recent
developments from behavioral, reproductive, and ecological perspectives. Stress and
Behavior 27, 215-‐290.
38. Wolf M., van Doorn G.S., Leimar O., Weissing F.J. 2007 Life-‐history trade-‐offs favour the
evolution of animal personalities. Nature 447(7144), 581-‐584.
39. Clark C.W. 1994 Antipredator behavior and the asset-‐protection principle. Behav Ecol
5(2), 159-‐170. (doi:10.1093/beheco/5.2.159).
40. Wilson D.S., Clark A.B., Coleman K., Dearstyne T. 1994 Shyness and boldness in humans
and other animals. Trends Ecol Evol 9(11), 442-‐446. (doi:10.1016/0169-‐
5347(94)90134-‐1).
41. Korzan W.J., Øverli Ø., Summers C.H. 2006 Future social rank: forecasting status in the
green anole (Anolis carolinensis). Acta Ethologica 9(1), 48-‐57. (doi:10.1007/s10211-‐
006-‐0015-‐5).
42. Dahlbom S.J., Lagman D., Lundstedt-‐Enkel K., Sundstrom L.F., Winberg S. 2011
Boldness Predicts Social Status in Zebrafish (Danio rerio). Plos One 6(8), 7.
(doi:e23565 10.1371/journal.pone.0023565).
43. Nelson X.J., Wilson D.R., Evans C.S. 2008 Behavioral Syndromes in Stable Social Groups:
An Artifact of External Constraints? Ethology 114(12), 1154-‐1165.
(doi:10.1111/j.1439-‐0310.2008.01568.x).
44. Davis D.E., Domm L.V. 1943 The influence of hormones on the sexual behavior of
domestic fowl. In Essays in biology in honor of Herbert M Evans (pp. 169–181).
Berkeley, University of California Press.
45. Johnsen T.S., Zuk M. 1995 Testosterone and aggression in male red jungle fowl. Horm
Behav 29(4), 593-‐598. (doi:10.1006/hbeh.1995.1288).
46. Rudin F.S., Briffa M. 2012 Is boldness a resource-‐holding potential trait? Fighting
prowess and changes in startle response in the sea anemone, Actinia equina. Proc R
Soc B 279(1735), 1904-‐1910. (doi:10.1098/rspb.2011.2418).
22
47. van Oers K., de Jong G., van Noordwijk A.J., Kempenaers B., Drent P.J. 2005 Contribution
of genetics to the study of animal personalities: a review of case studies. Behaviour
142, 1185-‐1206.
48. Fidler A.E., van Oers K., Drent P.J., Kuhn S., Mueller J.C., Kempenaers B. 2007 Drd4 gene
polymorphisms are associated with personality variation in a passerine bird. Proc R
Soc B 274(1619), 1685-‐1691. (doi:10.1098/rspb.2007.0337).
49. Nicolaus M., Tinbergen J.M., Bouwman K.M., Michler S.P.M., Ubels R., Both C.,
Kempenaers B., Dingemanse N.J. 2012 Experimental evidence for adaptive
personalities in a wild passerine bird. Proc R Soc B 279(1749), 4885-‐4892.
(doi:10.1098/rspb.2012.1936).
23
Paper II
24
Aggressiveness and exploration predicts social dominance in male
domestic fowl (Gallus gallus domesticus).
Anna Favati1*, Olof Leimar1 and Hanne Løvlie1,2
1 Department of Zoology, Stockholm University, 10691 Stockholm, Sweden
2 Department of Physics, Chemistry and Biology, Linköping University, 581 83 Linköping, Sweden
*Corresponding author, [email protected]
25
ABSTRACT
Individuals in social species commonly form dominance relationships, where dominant
individuals enjoy greater access to important resources compared to subordinates. Recent
work on animal personality (i.e. consistent variation in individual behavioural responses)
show that personality can covary with social status, thus suggest that besides morphological
traits, also behavioural variation can play an important role in establishment of status. We
investigated whether behavioural responses in non-‐social contexts could predict the
outcome of duels of pairs of morphologically matched pairs of male fowl (Gallus gallus
domesticus). Male fowl form social hierarchies only partly explained by variation in
morphological traits, and with unknown influences of variation in behaviours. Several
behaviours covary with social status in male fowl, however it is unclear whether such a
relationship is a consequence of the dominance position, or a cause of it. We show that male
fowl that more quickly explored a novel arena were more likely to obtain a dominant
position. In addition, and independent of exploration propensity, males that were more
aggressive at the start of a duel also had greater chances of becoming dominant. Our results
therefore show that multiple, uncorrelated behavioural traits, can predict social status.
Keywords: behavioural syndromes; intra-‐sexual selection; social status; social hierarchy;
chicken.
26
INTRODUCTION
In social species, relationships often take form of dominance hierarchies, commonly
described as the repeated win of one individual over another (Chase 1980). Ornaments and
other morphological features, like body size, typically contribute to the assessment of an
antagonist, but do not fully predict social rank (Andersson 1994). Behavioural predictors of
dominance have received much less attention, even though dominant and subordinate
individuals commonly differ quantitatively in behaviour. For example, aggression often (but
not always) shows a positive correlation with social status (Verbeek et al. 1996, Kim and Zuk
2000, McGhee and Travis 2010). Aggression often includes a heritable component, can have
a positive effect on reproductive success and shows intra-‐individual stability (Smith and
Blumstein 2008). Behaviours that are repeatable within the individual, but varies among
individuals, are also known as personality traits describing an individual's personality type
(Sih et al. 2004b). Dominance and aggression are sometimes included in the concept of
coping styles, a composite personality classification typically describing individuals by their
speed of exploration, boldness, and stress hormone profiles (Koolhaas et al. 1999).
Individuals scoring high on these behaviours are said to have a 'proactive' coping style, while
those scoring low are defined as having a 'reactive' coping style (Benus et al. 1987, Koolhaas
et al. 1999, Schjolden and Winberg 2007). These traits are sometimes positively correlated
within individuals (Verbeek et al. 1996, Øverli et al. 2004), but not always (Gomez-‐Laplaza
2002, Weiss et al. 2007, Fox et al. 2009, Weiss et al. 2009). Individual differences in coping
with challenges experienced during behavioural assays may reflect variation in coping also
with the challenge of meeting a same-‐sexed conspecific in a duel over social dominance. To
date, there are only a few studies exploring the predictive power of variation in personality
types for social status. Some of these studies have used lines selected for variation in
proactivity or aggressiveness, and indicate that proactive and aggressive individuals have
higher chances of obtaining a high social status (Benus et al. 1990, Pottinger and Carrick
2001). However, studies of the relationship between natural variation in personality and
social status are still rare, and the results are inconclusive; either more proactive individuals
or more reactive ones may become dominant (Verbeek et al. 1996, Øverli et al. 2004, Korzan
et al. 2006, Fox et al. 2009, Dahlbom et al. 2011, David et al. 2011).
27
We have tested the predictive power of behavioural responses in a non-‐social context on the
establishment of social status in male domestic fowl (Gallus gallus domesticus). The fowl is a
group-‐living species that readily forms social hierarchies where individuals obtain clear social
roles (Kruijt 1964). Despite the long history of studies on social hierarchies in the fowl, traits
determining status are not fully understood. Ornaments (e.g. the comb, a fleshy red
ornament on the head), and other morphological features (like body size) often contribute
to assessment of the antagonist, but do not fully predict the social status of an individual
(Ligon et al. 1990, Zuk and Johnsen 2000, Parker et al. 2002, Cornwallis and Birkhead 2007).
Our hypothesis is that there is scope for also behavioural variation to explain variation in
establishment of social status in the fowl. We therefore investigated if variation in behaviour
in a novel arena test, and aggressive behaviour at the start of a duel, could predict the
outcome of duels of morphologically matched males. Our expectation is that more proactive
or aggressive males have a higher chance of becoming dominant.
METHODS
Study population and housing
The study took place during the breeding seasons (June -‐ August) of 2011 and 2012 at
Tovetorp Research station, Stockholm University, Sweden. The population of an old Swedish
game breed of fowl (‘Gammalsvensk dvärghöna’, Gallus gallus domesticus) is kept in several
mixed-‐sex, mixed-‐age (1-‐10 years) groups under semi-‐natural conditions, and are used to
human handling. All males were marked with numbered leg rings to facilitate identification.
In order to standardise the social position of experimental males, single males were housed
together with one female in outdoor aviaries (3 x 3 meters) for 4-‐5 days before the first
behavioural test. Aviaries were visually, but not vocally, isolated from other birds. This set-‐
up resulted in that all males behaved like dominant birds prior to observations (Parker and
Ligon 2002). The experiments followed ethical requirements for animal experimentation in
Sweden (Linköping Ethical committee, ethical permission number 60-‐10).
Personality assays
To investigate variation in activity and explorative behaviour among males, 50 males were
tested singly in a Novel Arena test (NA). NA is a commonly used personality test to measure
28
exploration and activity, and the fowl show repeatability in both traits in similar tests (Favati
et al., unpublished (paper I)). A large oval shaped arena of a 10 x 7 m fenced area in a
deciduous forest was divided into 36 even-‐sized subareas by a grid of thin wooden sticks.
The male was placed in a cage of chicken wire connected to the arena by a wire door. The
door to the start cage was immediately opened, after which the behaviour of the male was
recorded for 20 minutes. The occurrence of ‘activity’ (i.e. movement of the legs of the bird,
either by walking or scratching) was recorded every 15s. The number of subareas visited at
least once by a male (between 1-‐36) was used as a measure of ‘area use’, and the latency to
visit at least 5 subareas was used as measures of ’exploration‘ propensity. The limit of five
subareas was chosen in order to obtain a continuously distributed variable, and a majority of
the males visited five subareas within the time limit of the experiment. The observer sat
outside the arena, visible to the male. The number of territorial crows uttered was also
recorded ('crowing'), as was the proportion of time a male spent being vigilant (‘vigilance’,
head above shoulder height, recorded every 15s.). At the end of the NA, a recorded alarm
call of a conspecific was played-‐back using an iPhone connected to a portable loudspeaker
placed next to the arena (Philips Mighty Mini). Alarm calls are loud cackling vocalisations
typically uttered in response to terrestrial predators, and which flock mates react on by
vigilance (Evans et al. 1993). The latency until the male retained normal activities after the
call was played, thus stopped being vigilant and resumed feeding after the startle, was used
as a measure of how alert and wary the male was ('alertness'). Males that did not resume
feeding within 10 minutes received the maximum score of 10 minutes.
Aggressiveness and social status
To investigate whether variation in aggressiveness affects establishment of social dominance
in the domestic fowl, aggressiveness was estimated during dyadic interactions not more
than one day after the personality assays. Pairs of males (npairs = 25) were simultaneously
placed one meter from each other in an outdoor aviary (2 x 3 m) where none of the males
had previously been housed. The males of each pair were matched for comb size and body
size in order to minimise the influence of these traits on the establishment of social status
(<10% difference). Body size and comb size were estimated by tarsus length and body
weight, and comb length, respectively. Males within a pair had not been housed together for
at least 2 weeks prior to trials, reducing any effects earlier encounters may have had on the
29
establishment of social status (McBride 1958, Cloutier et al. 1995). The duels were filmed
using a digital camcorder (Sony DCR-‐VX1000E). The agonistic interaction between males
started spontaneously when the two males of a pair were placed in front of each other. The
intensity of fowl fights typically ranges from aggressive displays (raised hackles, dropped
wing and ‘waltzing’ movements (Collias 1943, Kruijt 1964)) and surrender of one male
without any physical attacks, to repeated attacks with the feet or the beak until one male
retreats. The initial behaviour (aggressiveness) of each male was classified along a four grade
scale as either 1: ‘surrending’ (fleeing away from the other male), 2: ‘non-‐aggressive’
(straight body posture, no avoidance or attraction to the other male), 3: ‘aggressive’
(crouched body posture and dropped wing or raised hackles (Collias 1943)) or 4: ‘very
aggressive’ (directly – within 2 seconds – approaching the other male). All duels ended with
one of the opponents retreating and no longer retaliating further attacks from the
opponent. We used a minimum of five occasions of avoidance within an hour by the same
male to define a male as subordinate, and the other male as dominant. Overt aggression
lasted not longer than a few minutes, and there were no duels involving injuries other than
minor bleeding from the comb.
Statistical analyses
To investigate whether behaviour in the novel arena test could predict the outcome of the
duels, we compared a generalized linear model with binary response containing all possible
explanatory variables (activity, area use, exploration, crowing, vigilance, alertness) with a
null model containing no explanatory variables (only the intercept), using log-‐likelihood Χ2-‐
statistics (deviance). To obtain only one measure per pair, we used the difference in
response between the two males of each pair, and randomly altered between taking the
dominant score minus the subordinate score and vice versa. The outcome of the duels
(dominant or subordinate) was used as response variable and at the level of the pair (n =
25). To determine a final model, we used bidirectional stepwise model search, choosing the
model with the lowest AIC using generalized linear modelling (Venables and Ripley 2002).
The fit of the final model was tested using the log-‐likelihood Χ2-‐statistic for the final model
compared to the null model. To evaluate the predictive power of the model we used the
Nagelkerke r2 (Nagelkerke 1991). The strength of the models was further examined by
computing classification tables using the data the models were built on. The cut-‐off value to
30
classify each males' social status was set to 0.5 based on the equal amount of dominant and
subordinate males (Elliott and Woodward 2007). The same approach as described above was
used in order to analyse whether behaviour in NA could predict aggressiveness during the
first seconds of the duel, this time with aggressiveness as a binary response variable (0 =
non-‐aggressive, score 1-‐2; 1 = aggressive, score 3-‐4). Whether aggressiveness could predict
social dominance was analysed by a logistic regression with aggressiveness as explanatory
factor and social dominance as response variable. Finally, pairwise correlations using
Pearson’s product moment correlation were performed in order to investigate the
relationship between all behavioural responses measured.
In order to achieve normally distributed variables, the transformation √x was used for area
use and crowing, log(x) for exploration, and x0.25 for alertness, based on visual inspection of
histograms. These transformations were used for all analyses. R version 2.15.1 was used for
all statistic analyses (R Development Core Team 2012).
RESULTS
A full model containing all behavioural responses of the novel arena test showed that the
responses could forecast the social dominance in upcoming duels (Χ2 = 12.573, df = 6, p =
0.050). We therefore proceeded with model selection which resulted in the following final
model: social status ~ exploration + vigilance + crowing. This model significantly predicted
social status (Χ2 = 11.662, df = 3, p = 0.0086, r2 = 0.50), and the only significant term was
exploration; males that more quickly explored the arena more often won the following duel
(exploration, Χ2 = 5.50, df = 1, p = 0.019; vigilance, Χ2 = 2.43, df = 1, p = 0.12; crowing, Χ2 =
3.16, df = 1, p = 0.075, Fig. 1). This model accurately classified the males in 84% of the cases
(Table 2). Behaviour in NA could not predict aggressiveness at the start of the duel (Χ2 = 7.61,
df = 6, p = 0.27), and this was confirmed by pairwise correlations: no behaviour in NA had a
significant correlation with aggressiveness (Table 1). However, aggressiveness predicted
social dominance: more aggressive males had increased chances of becoming dominant (Χ2 =
8.21, df = 1, p = 0.0042, r2 = 0.37, Fig. 2). The logistic model of aggression correctly classified
the status of males in 80% of the cases (Table 2).
31
DISCUSSION
Socially dominant individuals commonly enjoy increased access to resources, such as mating
partners, which typically results in a positive relationship between social status and
reproductive success and thereby individual fitness (Andersson 1994). We here showed that
the behavioural type affected the chances of obtaining a high socials status for individuals of
the same apparent physical quality.
No strong correlations were observed between behavioural responses observed in the novel
arena test, when exposed to a startle and during duels, other than the one between activity
and exploration. In other words, we found no support for an overall behavioural syndrome
categorising individuals along a proactive-‐reactive continuum, which is sometimes found in
other species (Sih et al. 2004a). Therefore, we discuss each behavioural trait separately.
Dominant male fowl differ quantitatively from subordinate males in their behaviour, for
example they crow more (Favati et al. , Leonard and Horn 1995, Johnsen et al. 2001).
Crowing is an auditory display, which signals dominance and territoriality, and dominant
males often interfere with subordinate males that attempts to crow (Collias et al. 1964,
Leonard and Horn 1995). Dominant males are also more vigilant (Pizzari 2003, Cornwallis
and Birkhead 2008), perhaps due to mate guarding (Artiss and Martin 1995). However, we
found no support of pre-‐existing differences between future dominant and subordinate
individuals in neither how much a male crowed, nor how much time he spent being vigilant
in NA. The differences observed in other studies (Favati et al. , Leonard and Horn 1995,
Johnsen et al. 2001, Pizzari 2003, Cornwallis and Birkhead 2008) therefore primarily seem to
be an effect of the current social position rather than pre-‐existing differences in behavioural
type (Favati et al. , Cornwallis and Birkhead 2008). Among all behavioural traits observed in
the novel arena test, including exposure to a startle, only exploration seemed to have any
substantial relationship to future social status, as more explorative males more often
became dominant. Exploration was strongly correlated to activity and the number of
subareas visited, and the results suggest that this vigorousness brought advantages to the
duel situation, although not through a direct relationship with aggressiveness. This adds to
the general pattern of more proactive individuals being found in dominant positions,
although only a few have investigated the relationship between exploration and social
32
dominance, and in a some cases with opposite results (Verbeek et al. 1996, Dingemanse and
de Goede 2004, Fox et al. 2009, David et al. 2011).
That more aggressive individuals win more fights has been demonstrated in earlier studies of
various species (e.g. (Verbeek et al. 1996)) including the domestic fowl, where individuals
that initiate the first attack more often also win the fight (females, Ligon et al. 1990, males,
Cloutier and Newberry 2000). Aggression is however sometimes considered a property of
the interaction between individuals, in which case the individual aggressive response
depends on the actions of the opponent (Francis 1988). For instance, whether an attack is
initiated or not is not a purely individual decision, but also depends on the actions of the
opponent. Alternatively, as adopted in the current study, one can attempt to estimate
individual aggressiveness and distinguish it from social dominance. Thus, individual
aggressiveness can be estimated by the very initial reaction of an individual when meeting
an opponent. By basing our measure of aggressiveness on how the males react to the
fighting situation during the very first seconds of the interaction, it seems unlikely that an
evaluation of the opponent already has taken place. Using this approach, we found that an
initial aggressive position forecasts the outcome of the duel in male fowl, which has
previously been shown only in female fowl (Foreman and Allee 1959, Wilson 1974).
When males differed in aggressiveness within a pair, the male with higher aggressiveness
scores almost always won the fight (Figure 2). However, in around half the cases both males
initially adopted an aggressive body posture (i.e. received the same aggressiveness score),
which decreases the strength of the overall predictability. Despite this, aggressiveness still
appeared as a useful predictor of status, as it was used to correctly predict the outcome in
80 % of the duels. Explorative behaviour comes forth as an equally strong predictor of status
(Table 2). Overall, our results show that when combatants were matched for morphological
characters, behavioural traits did play an important role in the establishment of social status.
We targeted aggressiveness and exploration as the prior attributes, or personality traits,
most useful for prediction of a dominant social position. Over the last decade, reports on
fitness effects of behavioural types have accumulated; not least studies showing that
personality affects reproductive success (Dingemanse and Reale 2005, Smith and Blumstein
2008). Since dominance is important for sexual success in male fowl (McBride et al. 1969,
33
Collias and Collias 1996, Pizzari and Birkhead 2001), our results suggest that variation in
personalities in male fowl have fitness consequences via the establishment of social status.
ACKNOWLEDGEMENTS
We wish to thank A. Hennyey, D. Molin and L. Neuman for field assistance, and N. Andbjer
for game keeping.
34
Tables
Table 1. The relationship between behavioural responses in behavioural assays and initial aggressiveness
in duels, of male fowl.
Exploration Area use Activity Crowing Vigilance Alertness Aggressiveness Exploration -‐
Area use -‐0.70* -‐ Activity -‐0.51* 0.78* -‐
Crowing -‐0.21 0.14 -‐0.073 -‐ Vigilance -‐0.34 0.25 0.10 0.24 -‐
Alertness -‐0.153 -‐0.14 -‐0.089 0.018 0.381 -‐ Aggressiveness -‐0.038 -‐0.048 -‐0.021 -‐0.43 0.016 -‐0.055 -‐
Pearson’s correlation coefficients (r) are shown. * denotes correlations that remained significant at the p <
0.05 level after sequential Bon Ferroni correction.
35
Table 2. Classification table comparing observed dominance status with dominance status predicted by
final logistic models.
Observed Predicted Percentage
Sub Dom correct
Novel arena Sub 10 2 0.83 Dom 2 11 0.85
total 0.84
Aggressiveness Sub 7 5 0.58
Dom 0 13 1.00
total 0.80
Dominance status of males (subordinate, 'Sub', and dominant, 'Dom') was successfully classified by the final
logistic regression model of behavioural responses in the novel arena test in 84% of the cases, and in 80% of the
cases by the logistic model of aggressiveness during the start of the duel (in bold).
36
Figure 1. Differences in behavioural response in behavioural assays between future dominant and subordinate
male fowl. Males that in a later duel became dominant (filled dots) (a) were more explorative but (b) were not
more active, (c) did not have a larger area use, (d) did not crow more, and (e) were not more vigilant or (f)
more alert compared to subordinate males (open dots). Mean values ± SE are given. * = p < 0.05.
●
●
0
2
4
6
8
10La
tenc
y to
exp
lore
5 su
bare
as (m
in)
Dom Sub
(a)*
●
●
0
10
20
30
40
Activ
ity (%
of t
ime)
Dom Sub
(b)
●
●
0
5
10
15
20
25
30
No.
of s
ubar
eas
visi
ted
Dom Sub
(c)
●
●
10
15
20
25
30
Num
ber o
f cro
ws
Dom Sub
(d)
●
●
50
60
70
80
90
100
Vigi
lanc
e (%
of t
ime)
Dom Sub
(e)
●
●
0
40
80
120
160
Vigi
lanc
e af
ter s
tartl
e (s
)
Dom Sub
(f)
37
Figure 2. The distribution of male fowl duels with respect to the difference in aggressiveness between the
contestants. The winning, dominant male either had a higher aggressiveness score compared to the losing,
subordinate part (Dom>Sub, black column), a lower score (Sub<Dom, white column) or the same
aggressiveness score (Dom=Sub, grey column). When threat score differed between the two males of a pair,
the male with the higher aggressiveness score more often won (12 cases vs 1 case).
# of
due
ls
0
2
4
6
8
10
12
Dom>Sub Sub>Dom Dom=SubDifference in aggressiveness
38
REFERENCES
Andersson, M. 1994. Sexual selection. Princeton University Press, New Jersey.
Artiss, T. and K. Martin. 1995. Male vigilance in white-‐tailed ptarmigan, Lagopus
leucurus – mate guarding or predator detection. Animal Behaviour 49:1249-‐1258.
Benus, R. F., S. Dendaas, J. M. Koolhaas, and G. A. Vanoortmerssen. 1990. Routine
formation and flexibility in social and nonsocial behavior of aggressive and
nonaggressive male-‐mice. Pages 176-‐193.
Benus, R. F., J. M. Koolhaas, and G. A. Vanoortmerssen. 1987. Individual-‐differences in
behavioral reaction to a changing environment in mice and rats. Pages 105-‐122.
Chase, I. D. 1980. Social-‐Process and Hierarchy Formation in Small-‐Groups -‐ a
Comparative Perspective. American Sociological Review 45:905-‐924.
Cloutier, S., J. P. Beaugrand, and P. C. Lague. 1995. The effect of prior victory or defeat in
the same site as that of subsequent encounter on the determination of dyadic
dominance in the domestic hen. Behavioural Processes 34:293-‐298.
Cloutier, S. and R. C. Newberry. 2000. Recent social experience, body weight and initial
patterns of attack predict the social status attained by unfamiliar hens in a new
group. Behaviour 137:705-‐726.
Collias, N. E. 1943. Statistical Analysis of Factors Which Make for Success in Initial
Encounters between Hens. The American Naturalist 77:519-‐538.
Collias, N. E. and E. C. Collias. 1996. Social organization of a red junglefowl, Gallus gallus,
population related to evolution theory. Animal Behaviour 51:1337-‐1354.
Collias, N. E., E. C. Collias, and P. Saichuae. 1964. A field study of the Red jungle fowl
(Gallus gallus) in Southeast Asia and India. American Zoologist 4:272-‐273.
Cornwallis, C. K. and T. R. Birkhead. 2007. Changes in sperm quality and numbers in
response to experimental manipulation of male social status and female
attractiveness. American Naturalist 170:758-‐770.
Cornwallis, C. K. and T. R. Birkhead. 2008. Plasticity in reproductive phenotypes reveals
status-‐specific correlations between behavioral, morphological, and physiological
sexual traits. Evolution 62:1149-‐1161.
Dahlbom, S. J., D. Lagman, K. Lundstedt-‐Enkel, L. F. Sundstrom, and S. Winberg. 2011.
Boldness Predicts Social Status in Zebrafish (Danio rerio). Plos One 6:7.
39
David, M., Y. Auclair, and F. Cezilly. 2011. Personality predicts social dominance in
female zebra finches, Taeniopygia guttata, in a feeding context. Animal Behaviour
81:219-‐224.
Dingemanse, N. J. and P. de Goede. 2004. The relation between dominance and
exploratory behavior is context-‐dependent in wild great tits. Behavioral Ecology
15:1023-‐1030.
Dingemanse, N. J. and D. Reale. 2005. Natural selection and animal personality.
Behaviour 142:1159-‐1184.
Elliott, A. and W. Woodward. 2007. Statistical Analysis Quick Reference Guidebook.
Sage, Thousand Oaks.
Evans, C. S., L. Evans, and P. Marler. 1993. On the meaning of alarm calls -‐ functional
reference in an avian vocal system. Animal Behaviour 46:23-‐38.
Favati, A., O. Leimar, T. Radesater, and H. Løvlie. Stability in social ‘state’ and
personality: social status and individual characteristics explain consistency in
behavioural responses. Unpublished.
Foreman, D. and W. C. Allee. 1959. A correlation between posture stance and outcome in
paired contests of domestic hens. Animal Behaviour 7:180-‐184.
Fox, R. A., L. D. Ladage, T. C. Roth, and V. V. Pravosudov. 2009. Behavioural profile
predicts dominance status in mountain chickadees, Poecile gambeli. Animal
Behaviour 77:1441-‐1448.
Francis, R. C. 1988. On the relationship between aggression and social-‐dominance.
Ethology 78:223-‐237.
Gomez-‐Laplaza, L. M. 2002. Social status and investigatory behaviour in the angelfish
(Pterophyllum scalare). Behaviour 139:1469-‐1490.
Johnsen, T. S., M. Zuk, and E. A. Fessler. 2001. Social dominance, male behaviour and
mating in mixed-‐sex flocks of red jungle fowl. Behaviour 138:1-‐18.
Kim, T. and M. Zuk. 2000. The effects of age and previous experience on social rank in
female red junglefowl, Gallus gallus spadiceus. Animal Behaviour 60:239-‐244.
Koolhaas, J. M., S. M. Korte, S. F. De Boer, B. J. Van Der Vegt, C. G. Van Reenen, H. Hopster,
I. C. De Jong, M. A. W. Ruis, and H. J. Blokhuis. 1999. Coping styles in animals: current
status in behavior and stress-‐physiology. Neuroscience and Biobehavioral Reviews
23:925-‐935.
40
Korzan, W. J., Ø. Øverli, and C. H. Summers. 2006. Future social rank: forecasting status
in the green anole (Anolis carolinensis). Acta Ethologica 9:48-‐57.
Kruijt, J., P. 1964. Ontogeny of social behaviour in Burmese red junglefowl (Gallus gallus
Spadiceus) Bonnaterre. Behaviour Supplement:I-‐IX, 1-‐201.
Leonard, M. L. and A. G. Horn. 1995. Growing in relation to status in roosters. Animal
Behaviour 49:1283-‐1290.
Ligon, J. D., R. Thornhill, M. Zuk, and K. Johnson. 1990. Male male competition,
ornamentation and the role of testosterone in sexual selection in red jungle fowl.
Animal Behaviour 40:367-‐373.
McBride, G. 1958. The measurement of aggressiveness in the domestic hen. Animal
Behaviour 6:87-‐91.
McBride, G., I. P. Parer, and F. Foenander. 1969. The social organisation and behaviour
of the feral domestic fowl. Baillière, Tindall & Cassell, London.
McGhee, K. E. and J. Travis. 2010. Repeatable behavioural type and stable dominance
rank in the bluefin killifish. Animal Behaviour 79:497-‐507.
Nagelkerke, N. J. D. 1991. A note on a general definition of the coefficient of
determination. Biometrika 78:691-‐692.
Parker, T. H., R. Knapp, and J. A. Rosenfield. 2002. Social mediation of sexually selected
ornamentation and steroid hormone levels in male junglefowl. Animal Behaviour
64:291-‐298.
Parker, T. H. and J. D. Ligon. 2002. Dominant male red junglefowl (Gallus gallus) test the
dominance status of other males. Behavioral Ecology and Sociobiology 53:20-‐24.
Pizzari, T. 2003. Food, vigilance, and sperm: the role of male direct benefits in the
evolution of female preference in a polygamous bird. Behavioral Ecology 14:593-‐
601.
Pizzari, T. and T. R. Birkhead. 2001. For whom does the hen cackle? The function of
postoviposition cackling. Animal Behaviour 61:601-‐607.
Pottinger, T. G. and T. R. Carrick. 2001. Stress responsiveness affects dominant-‐
subordinate relationships in rainbow trout. Hormones and Behavior 40:419-‐427.
R Development Core Team 2012. R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna.
41
Schjolden, J. and S. Winberg. 2007. Genetically determined variation in stress
responsiveness in rainbow trout: Behavior and neurobiology. Brain Behavior and
Evolution 70:227-‐238.
Sih, A., A. Bell, and J. C. Johnson. 2004a. Behavioral syndromes: an ecological and
evolutionary overview. Trends in Ecology & Evolution 19:372-‐378.
Sih, A., A. M. Bell, J. C. Johnson, and R. E. Ziemba. 2004b. Behavioral syndromes: An
integrative overview. Quarterly Review of Biology 79:241-‐277.
Smith, B. R. and D. T. Blumstein. 2008. Fitness consequences of personality: a meta-‐
analysis. Behavioral Ecology 19:448-‐455.
Venables, W. N. and B. D. Ripley. 2002. Modern Applied Statistics with S. Fourth edition.
Springer, New York.
Verbeek, M. E. M., A. Boon, and P. J. Drent. 1996. Exploration, aggressive behavior and
dominance in pair-‐wise confrontations of juvenile male great tits. Behaviour
133:945-‐963.
Verbeek, M. E. M., P. J. Drent, and P. R. Wiepkema. 1994. Consistent individual
differences in early exploratory behavior of male great tits. Animal Behaviour
48:1113-‐1121.
Weiss, A., M. Inoue-‐Murayama, K. W. Hong, E. Inoue, T. Udono, T. Ochiai, T. Matsuzawa, S.
Hirata, and J. E. King. 2009. Assessing Chimpanzee Personality and Subjective Well-‐
Being in Japan. American Journal of Primatology 71:283-‐292.
Weiss, A., J. E. King, and W. D. Hopkins. 2007. A cross-‐setting study of chimpanzee (Pan
troglodytes) personality structure and development: Zoological parks and Yerkes
National Primate Research Center. American Journal of Primatology 69:1264-‐1277.
Wilson, R. H. 1974. Agonostic postures and latency to first interaction during initial pair
encounters in red jungle fowl, Gallus-‐gallus. Animal Behaviour 22:75-‐&.
Zuk, M. and T. S. Johnsen. 2000. Social environment and immunity in male red jungle
fowl. Behavioral Ecology 11:146-‐153.
Øverli, Ø., W. J. Korzan, E. Hoglund, S. Winberg, H. Bollig, M. Watt, G. L. Forster, B. A.
Barton, E. Overli, K. J. Renner, and C. H. Summers. 2004. Stress coping style predicts
aggression and social dominance in rainbow trout. Hormones and Behavior 45:235-‐
241.