Access to this full-text is provided by The Royal Society.
Content available from Royal Society Open Science
This content is subject to copyright.
royalsocietypublishing.org/journal/rsos
Research
Cite this article: Miln C, Ward AJW, Seebacher
F. 2021 Social rank and not physiological capacity
determines competitive success in zebrafish
(Danio rerio). R. Soc. Open Sci. 8: 210146.
https://doi.org/10.1098/rsos.210146
Received: 29 January 2021
Accepted: 15 March 2021
Subject Category:
Organismal and evolutionary biology
Subject Areas:
behaviour/physiology
Keywords:
sprint speed, locomotor performance, social
hierarchy, feeding, zebrafish
Author for correspondence:
Frank Seebacher
e-mail: frank.seebacher@sydney.edu.au
Electronic supplementary material is available
online at https://doi.org/10.6084/m9.figshare.c.
5359401.
Social rank and not
physiological capacity
determines competitive
success in zebrafish
(Danio rerio)
Clare Miln, Ashley J. W. Ward and Frank Seebacher
School of Life and Environmental Sciences A08, University of Sydney, NSW 2006, Australia
AJWW, 0000-0003-0842-533X; FS, 0000-0002-2281-9311
Competition for resources shapes ecological and evolutionary
relationships. Physiological capacities such as in locomotor
performance can influence the fitness of individuals by
increasing competitive success. Social hierarchy too can affect
outcomes of competition by altering locomotor behaviour or
because higher ranking individuals monopolize resources.
Here, we tested the hypotheses that competitive success is
determined by sprint performance or by social status. We
show that sprint performance of individuals measured
during escape responses (fast start) or in an accelerated
sprint test did not correlate with realized sprint speed while
competing for food within a social group of five fish; fast
start and accelerated sprint speed were higher than realized
speed. Social status within the group was the best predictor
of competitive success, followed by realized speed. Social
hierarchies in zebrafish are established within 7 days of their
first encounter, and interestingly, there was a positive
correlation between social status and realized speed 1 and 4
days after fish were placed in a group, but not after 7 days.
These data indicate that physiological performance decreases
in importance as social relationships are established. Also,
maximal physiological capacities were not important for
competitive success, but swimming speed changed with
social context.
1. Introduction
Obtaining food is most fundamental for survival and fitness in
heterotrophic organisms, and competition is a principal ecological
and evolutionary driver [1]. The mechanisms that determine
competitive success therefore also reveal the primary drivers
© 2021 The Authors. Published by the Royal Society under the terms of the Creative
Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits
unrestricted use, provided the original author and source are credited.
underlying ecological relationships [2]. Differences in locomotor capacity between individuals may explain
success in scramble competition, although speed may trade off with food detection accuracy [3,4].
However, within a social group, individuals do not necessarily display their maximal physiological
capacities, but alter behaviour depending on the social context [5,6]. Hence, locomotor capacity and
social status may either independently or interactively determine competitive success in social animals.
Our aim was to test this relationship between locomotor performance and social status in zebrafish
(Danio rerio)competingforfood.
Sprint performance is a repeatable trait that is thought to be under selection because of the
advantages it confers for predator escape, prey capture, territorial defence and survival [7]. Sprints are
characterized by short (less than 30 s) bursts of high-powered locomotion that are mediated by fast
glycolytic muscle fibres [8]. In fish, maximal sprint performance is often measured as fast starts in
response to an external stimulus eliciting an escape response [9]. Alternatively, sprint performance can
be measured as the fastest speed attained during constant acceleration in a swimming flume [10].
However, animals under natural circumstances often do not move maximally, and performance is
modulated by context [11]. For example, collared lizards used 25–90% of maximal sprint speed to
forage, escape predators or defend territories [12].
Social relationships may influence locomotor behaviour [13], and they may alter the importance of
locomotor performance for competitive success. It is possible that social status per se takes precedence
over locomotor capacity in determining competitive success. In zebrafish, a social hierarchy among
unfamiliar animals is established within 7 days [14]. This establishment period provides a suitable
context to test for the relative importance of locomotor capacity and social status in determining
competitive success, because social status would be relatively unimportant before a hierarchy is
established. We therefore tested the hypothesis that (i) immediately after fish are placed in a group,
social status is less important than sprint performance in determining competitive success within the
group, but this relationship is reversed after 7 days. Because sprint performance may vary with
context, we tested the hypothesis that (ii) sprint performance measured in individual fish in an arena
or flume reflects sprint performance realized by individuals within the group. Lastly, sprint
performance may determine social status or, vice versa, the realized speed in the group may be
modified by social status as the hierarchy is established, so we tested the hypothesis that (iii) realized
speed in the group is positively associated with social status.
2. Material and methods
2.1. Study animals
All animal handling and experimental procedures were conducted with the approval of the University of
Sydney Animal Ethics Committee (approval number: 2016/982). Zebrafish, Danio rerio, form social
hierarchies that are stable [15] and therefore are a good model to test hypotheses regarding the fitness
consequences of social hierarchies. Adult short-fin zebrafish of mixed sex were purchased from a
commercial supplier (Marine Fish Direct, Narellan, NSW, Australia) and housed in plastic tanks
(0.64 × 0.45 × 0.23 m; 2–3 fish per litre) until the start of experiments (1–2 weeks). Each tank contained
a sponge filter connected to an air pump (AC-9908; Resun, China), and we conducted a 30% water
change once a week. Water temperature was 23°C (± 0.5°C) during all experiments, the light cycle
was 12 L : 12 D h, and fish were fed flake food (Wardley Total Tropical Flake Blend; Hartz Mountain
Company, Secaucus NJ, USA) until satiation once daily. Fish were marked individually with a
coloured elastomer tag injected subcutaneously (NorthWest Marine Technology Inc., Shaw Island,
WA, USA). Each fish within an experimental group (see below) was marked with a different colour,
which was readily distinguishable by direct observation and from video recordings. Fish were given
at least 3 days to recover from the tagging procedure before starting experiments; after 3 days all fish
fed and behaved normally (i.e. engaged in social behaviour and spent most of their time cruising near
the water surface). Experimental fish were of mixed sex (with two or three fish of each sex per group)
to provide a more natural context and to avoid bias from using a single sex only [16]. In zebrafish,
both males and females engage in dominance behaviour [15]. All fish were weighed with an
electronic balance (Navigator NV, Ohaus, USA), and photographed (with a Exilim EX-ZR200A, Casio,
USA camera; electronic supplementary material, figure S1) to determine the standard length (in
Tracker Video Analysis and Modelling Tool 4.01; Open Source Physics, www.opensourcephysics.org).
royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 8: 210146
2
2.2. Sprint swimming performance
We measured swimming performance in all experimental fish (i.e.10 groups of five fish each, see below).
There are different measurement techniques of sprint speed in the literature [17], and to avoid bias, we
included two different measures. Hence, sprint speed was determined as an escape response and a
constant acceleration test as described previously [10]. Escape responses were induced by a
mechanical stimulus to the tail of the fish in a shallow tank (405 × 600 mm, 30 mm water depth). We
filmed the escape response from above at 120 frames s
−1
(with a GoPro Hero 6, San Mateo CA, USA),
and a submerged 300 mm ruler served as a scale. We analysed videos (in Tracker Video Analysis and
Modelling software, 4.01, Open Source Physics, www.opensourcephysics.org) using the fish centre of
mass as the tracking point. Centre of mass was defined as the location at 0.35 body lengths from the
tip to the snout [18]. The purpose of measuring fast starts is to determine maximal capacities. Hence,
four escape responses were analysed for each fish and the fastest speed was used in the analysis,
which we refer to as ‘fast start’[17,19]. Escape responses were only used when the movement of fish
was not impeded by the edge of the tank.
The second method consisted of determining the maximum attained speed (U
10
) in a 10 s constant
acceleration trial in a cylindrical, Blazka-style swimming flume (21 mm diameter × 150 mm length)
[10]. Each flume was fitted tightly over the intake end of a submersible pump (12 V DC, iL500; Rule,
Hertfordshire, UK) that drew water through the flume. The flow was adjusted by changing the
voltage input to the pumps with a DC-regulated power supply (NP9615; Manson Engineering
Industrial, Hong Kong). A flow meter (DigiFlow, 6710M, Savant Electronics, Taichung, Taiwan) fitted
to the outlet of the pump measured flow velocity in real time. Before the test, fish swam at an initial
flow rate of 0.05 m s
−1
for 30 s, after which water flow was continuously increased by an average of
0.05 m s
−1
every 1 s until fish could no longer maintain their position in the water column and fell
back against a mesh net. The maximum flow speed was recorded when the fish was still holding its
position in the water column and used as maximum sprint performance.
2.3. Competitive ability
To measure competitive success and social interactions, we assembled experimental groups (n=10
groups) of five fish each. Fish within groups were size matched so that they were within 2 mm
standard length of each other, and individuals had not been in contact with each other before the
assembly of the groups. Each group was housed in a separate tank (0.30 × 0.21 × 0.18 m). For trials, all
five fish per tank were placed underneath a perforated 1 l cylindrical container, which allowed visual
inspection of the tank but prevented egress from the container. A piece of fish flake food (Wardley
Total Tropical Flake Blend; Hartz Mountain Company, Secaucus, NJ) was then placed approximately
20 cm away from the perforated container. After 5 s, the container was removed allowing the fish to
move freely around the tank. Fish were filmed (GoPro Hero 6, San Mateo CA, USA, at 120 frames s
−1
)
from just before the container was lifted, and the identity of the fish that obtained the food was
recorded. This procedure was repeated 10 times consecutively per day for 7 days. In preliminary
trials, we determined the best size of the food flake so that it was visible to the fish, but small enough
for only one fish to obtain it. During filmed trials, water depth was lowered to 7 cm, and between
trials fish groups remained in their tanks and water levels were increased to 15 cm when fish were
not filmed. After the food trials each day fish were fed to satiation. From the videos, we also
determined the speed at which fish travelled towards the food (realized speed). We defined realized
speed as the maximum speed calculated across two video frames while a fish was moving after the
container was lifted and until the food item was eaten by any fish. Note that the maximum speed
attained during competitive trials was significantly correlated with the average speed (permutational
analysis [see below] p< 0.0001), and this relationship did not differ between days ( p= 0.69; electronic
supplementary material, figure S2). We used maximum speed as a representation of realized speed
because it provides a better comparison with sprint speed, which represents maximum physiological
capacity.
2.4. Social interactions
Before each competitive trial, we filmed fish (with a GoPro Hero 6, San Mateo CA, USA, at 120 frames s
−1
)
moving freely and undisturbed in their home tank for 1 min; the first footage was taken immediately after
fish were assembled into experimental groups on the first day. From videos, we recorded aggressive
royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 8: 210146
3
interactions between individuals. An aggressive interaction was defined as a charge by an individual
towards another fish, or a nip towards another individual that elicited a response in the receiving
individual. We used the method described in [20] to determine social scores, where a fish received a
point (+1) for each incident where it was the aggressor, and a negative score (−1) for each aggressive
interaction where it was the receiver. The total score for each fish per day was calculated by adding
together the aggressor and receiver scores, and fish with higher scores were considered to be more
dominant within the social hierarchy.
2.5. Statistical analyses
We analysed all data with permutational analyses in the R [21] package lmPerm [22]. Permutational
analyses are a randomization procedure that use the data per se for statistical inference rather than
assuming underlying distributions, which makes them more parsimonious and independent from
underlying assumptions [23]. The p-value in permutational analyses is calculated as the number of
randomized datasets that show the same or greater effect as the measured dataset, divided by the
total number of permutations; the p-value therefore denotes the probability that the observed effects
could be obtained randomly. Hence, permutational analyses do not have any associated test statistic
and are often preferable to frequentist statistics [24]. The sample size was 50 fish in 10 groups of five
fish each, and we give p-values and degrees of freedom (p
df
).
In all models, we used ’tank’(=group) as a random blocking factor. In the event, ’tank’was not
significant in any of the comparisons (all p> 0.1). To test the relationship between sprint speed and
realized speed, we used realized speed as the dependent variable and fast start or U
10
speed and
‘day’as the dependent variables. We used a fast start or U
10
speed, realized speed and social score as
independent variables to test their influence on competitive success. Lastly, we tested whether realized
speed influenced social score by using the social score as the dependent variable and realized speed
and ‘day’as independent variables. When ‘day’was significant in the whole dataset, we repeated the
analysis using subsets of data for each day to test on which days there were significant effects. We
analysed data from days 1, 4 and 7 only. Effect sizes of different predictors on competitive success
were calculated as Cohen’s d, and 95% confidence intervals were determined by bootstrapping in the
R package ‘boot’[25].
3. Results
3.1. Fast starts or U
10
did not predict realized speed
Realized speed of individuals while moving towards the food within the group of five fish (n=10
groups) was lower than fast start (figure 1a) and U
10
speed (electronic supplementary material, figure
0 0.5 1.0 1.5
0
0.5
1.0
fast start s
p
eed (m s–1)
realized speed (m s–1)
day 1
day 4
day 7
147
0.3
0.4
0.5
da
y
(a)(b)
Figure 1. Sprint speed during competition and in an arena. Fast start speed measured individually in an arena was greater than the
fastest speed realized by individuals when moving towards a food source within a group of five fish (realized speed) (a). There was
no statistically significant relationship between fast start and realized speeds on day 1 (red circles), day 4 (blue circles) or day 7
(green circles). However, realized speed increased from day 1 to day 7 (b). Sample size was 10 groups of five fish on each day, the
solid line shows x= y, and means ± 95% CI are shown in (b).
royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 8: 210146
4
S1a), and there was no statistically significant association between realized speed and fast start speed
(p
1,137
= 0.3) or U
10
speed ( p
1,137
= 0.9). However, between days 1 and 7 realized speed increased
significantly by 0.19 m s
−1
on average ( p< 0.0001
2,137
; figure 1b). Additionally, fish with greater fast
start speed reduced their realized speed to a disproportionally greater extent than fish with slower
fast start speed ( p
1,137
< 0.0001), and this pattern was more pronounced on day 1 than on days 4 and
7(p
2,137
< 0.0001; electronic supplementary material, figure S3).
3.2. Realized speed and social score predicted competitive success
Realized speed ( p
1,135
= 0.004) and social score ( p
1,135
< 0.0001) had a positive effect on competitive
success, but fast start speed ( p
1,135
= 0.9) and U
10
speed ( p
1,135
= 0.1) did not influence competitive
success (figure 2; electronic supplementary material, figure S4). The effect size of social score on
competitive success was greater than that of realized speed (95% CI; figure 2). There was no
significant effect of day on these relationships ( p
2,135
= 0.7).
3.3. Social score was associated with realized speed only initially
There was a significant positive relationship between social score and realized speed ( p
1,137
< 0.0001), and
there was a significant effect of day ( p
2,137
< 0.0001) (figure 3). The association between social score and
realized speed was significant on day 1 (p
1,48
=0.03) and on day 4 (p
1,48
= 0.005), but not on day 7
(p
1,48
=0.3).
0.2 0.4 0.6 0.8
0
2
4
6
8
10
realized speed (m s
–1
)
–10 –5 0 5 10 15
0
2
4
6
8
10
social score
competitive success
(no. food items)
day 1
day 4
day 7
0.4 0.6 0.8 1.0 1.2 1.4
0
2
4
6
8
10
fast start speed (m s
–1
)
social score
realized speed
U10
fast start
–0.2
0
0.2
0.4
0.6
0.8
effect size
(on competitive success)
(a)
(d)
(b)(c)
Figure 2. Predictors of competitive success. Social score (a) and speed realized by individuals when moving towards a food source
within a group of five fish (realized speed) (b) were positively related to competitive success, measured as the number of food items
obtained by an individual when competing in a group of five fish for 10 consecutive trials per day. Fast start speed was not
associated with competitive success (c), and there were no differences between days (day 1 = red circles; day 4 = blue circles;
day 7 = green circles). Social score had the greatest effect on competitive success (effect size) (d), followed by realized speed,
but speed achieved during a 10 s accelerated spring test (U
10
) or fast start speed did not have an effect on competitive
success. Sample size was 10 groups of five fish on each day, and significant regression lines (± 95% CI) are shown ((a):
Y=−0.73 + 7.35x, R
2
= 0.13; (b): Y= 2.01 + 0.32x, R
2
= 0.39).
royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 8: 210146
5
4. Discussion
We have shown that position within a social hierarchy is the most important determinant of competitive
success, although realized speed while competing does contribute to success. However, maximal
physiological capacities did not lead to a competitive advantage, and realized speed while competing
in a group was less than maximal performance.
Group living is advantageous because it protects individuals from predation, and information
transfer between individuals in the group increases access to resources [26,27]. Group fission and
fusion impacts population dynamics, and interactions between individuals and their relative positions
within the group can influence fitness [28–30]. Social hierarchies within groups are common, and
more dominant individuals within a group have greater competitive success and access to resources
[31]. However, dominance can also come at the cost of increased stress and energy expenditure
resulting from increased aggression to maintain social position [32]. Physiological capacities are often
associated with group dynamics [4,33], and increased capacities may promote success in aggressive
encounters and dominance [34].
The consequences of social hierarchies are particularly important when competing for limited
resources such as food, which can lead to differences in fitness. However, even though subordinate
individuals experience restricted access to resources, they remain within the group because of the
benefits of group living [35]. It is therefore advantageous for individuals and the group as a whole to
reduce levels of aggression within a group. Individuals may therefore adjust their behaviour to
minimize the costs and maximize the benefits of group living [36]. Subordinates within a group may
suppress their potential physical performance to avoid challenging dominant individuals and thereby
avoid aggressive encounters [37]. We show that even while being established, social hierarchy plays
an important role in food acquisition. Despite having the potential swimming capacity to win limited
items of food, we showed that social status within the group had a greater effect on success to obtain
food. Subordinates within a group may benefit from forgoing opportunities to obtain food to reduce
conflict with the more dominant fish [38].
It is interesting that animals rarely perform maximally under natural or naturalistic situations [11,12],
which is likely to be the result of trade-offs with other fitness-related traits [39]. Fish with greater fast start
speeds showed a disproportionally greater reduction to realized speed than fish with lower fast start
speeds, indicating that fish tended to conform to the group regardless of their physiological capacities.
The observation that realized speed increased during the week indicates that the stimulus to compete
did not decrease during that period. Movement while competing may be interpreted within a
conceptual framework that is similar to economic escape theory. Economic escape theory posits that
predator escape responses are determined by the trade-off between the costs of fleeing (e.g. missed
opportunities and energetic investment) and the costs of not fleeing (loss of lifetime fitness) [40].
While competing for food, the cost of obtaining food would include increased swimming speed and
energetic investment, and the increased likelihood of subsequent conflict with others in the group.
The cost of not obtaining food would be a decrement in nutrition. These two costs would need to be
–10 –5 0 5 10
0
0.2
0.4
0.6
0.8
social score
realized speed (m s–1)
–10 –5 0 5 10
0
0.2
0.4
0.6
0.8
social score
–10 –5 0 5 10
0
0.2
0.4
0.6
0.8
social score
(a)(b)(c)
Figure 3. Social score predicted realized speed. There were weak but statistically significant positive relationships between realized
speed and social score on day 1 (a) and day 4 (b), but not on day 7 (c). Sample size was 10 groups of five fish on each day, and
significant regression lines (± 95% CI) are shown ((a): Y=−3.86 + 12.71x, R
2
= 0.096; (b): Y=−7.21 + 18.98x, R
2
= 0.15).
royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 8: 210146
6
traded off while competing in a group. Dominant individuals may minimize swim speed to save energy
while still obtaining food, and subordinate individuals could adjust swim speed according to that of
more dominant fish. Also, the stimulus to escape a predator (fast starts) may be far stronger than to
obtain food. These dynamics could explain why realized speeds are lower than maximal speeds, and
why the social score is related to the realized speed at least early when hierarchies are established.
Ethics. All animal handling and experimental procedures were conducted with the approval of the University of Sydney
Animal Ethics Committee (approval no.: 2016/982).
Data accessibility. Data has been deposited in Dryad (https://doi.org/10.5061/dryad.v15dv41tr) [41].
Authors’contributions. F.S., C.M. and A.J.W.W. designed the study; C.M. conducted experiments; F.S. wrote the manuscript
and analysed data; F.S., C.M. and A.J.W.W. edited the manuscript
Competing interests. We declare we have no competing interests
Funding. This work was supported by an Australian Research Council Discovery grant (DP180103036) to F.S.
References
1. Bolnick DI. 2004 Can intraspecific competition
drive disruptive selection? An experimental test
in natural populations of sticklebacks. Evolution
58, 608–618.
2. Ward AJW, Webster MM, Hart PJB. 2006
Intraspecific food competition in fishes. Fish
Fish. 7, 231–261. (doi:10.1111/j.1467-2979.
2006.00224.x)
3. David M, Gillingham MAF, Salignon M,
Laskowski KL, Giraldeau L-A. 2014 Speed–
accuracy trade-off and its consequences in a
scramble competition context. Anim. Behav. 90,
255–262. (doi:10.1016/j.anbehav.2014.02.009)
4. Killen SS, Marras S, Nadler L, Domenici P. 2017
The role of physiological traits in assortment
among and within fish shoals. Phil.
Trans. R. Soc. B 372, 20160233-13. (doi:10.
1098/rstb.2016.0233)
5. Ward AJW, Herbert-Read JE, Schaerf TM,
Seebacher F. 2018 The physiology of leadership
in fish shoals: leaders have lower maximal
metabolic rates and lower aerobic scope. J. Zool.
305,73–81. (doi:10.1111/jzo.12534)
6. Hansen MJ, Ligocki IY, Zillig KE, Steel AE,
Todgham AE, Fangue NA. 2020 Risk-taking and
locomotion in foraging threespine sticklebacks
(Gasterosteus aculeatus): the effect of nutritional
stress is dependent on social context. Behav.
Ecol. Sociobiol. 74,1–12. (doi:10.1007/s00265-
019-2795-4)
7. Husak JF. 2016 Measuring selection on
physiology in the wild and manipulating
phenotypes (in terrestrial nonhuman
vertebrates). Compr. Physiol. 6,63–85. (doi:10.
1002/cphy.c140061)
8. Gundersen K. 2011 Excitation-transcription
coupling in skeletal muscle: the molecular
pathways of exercise. Biol. Rev. 86, 564–600.
(doi:10.1111/j.1469-185X.2010.00161.x)
9. Domenici P, Hale ME. 2019 Escape responses of
fish: a review of the diversity in motor control,
kinematics and behaviour. J. Exp. Biol. 222,
jeb166009. (doi:10.1242/jeb.166009)
10. Simmonds AIM, Miln C, Seebacher F. 2019
Zebrafish (Danio rerio) as a model for sprint
exercise training. Zebrafish 16,1–7. (doi:10.
1089/zeb.2018.1646)
11. Wilson RS, Husak JF, Halsey LG, Clemente CJ.
2015 Predicting the movement speeds of
animals in natural environments. Integr.
Comp. Biol. 55, 1125–1141. (doi:10.1093/icb/
icv106)
12. Husak JF, Fox SF. 2006 Field use of maximal
sprint speed by collared lizards (Crotaphytus
collaris): compensation and sexual selection.
Evolution 60, 1888–1895.
13. Akbaripasand A, Krkosek M, Lokman PM, Closs
GP. 2014 Does social status within a dominance
hierarchy mediate individual growth, residency
and relocation? Oecologia 176, 771–779.
(doi:10.1007/s00442-014-3038-z)
14. Ricci L, Summers CH, Larson ET, O’Malley D,
Melloni Jr RH. 2013 Development of
aggressive phenotypes in zebrafish: interactions
of age, experience and social status. Anim.
Behav. 86, 245–252. (doi:10.1016/j.anbehav.
2013.04.011)
15. Paull GC, Filby AL, Giddins HG, Coe TS, Hamilton
PB, Tyler CR. 2010 Dominance hierarchies in
zebrafish (Danio rerio) and their relationship
with reproductive success. Zebrafish 7, 109–117.
(doi:10.1089/zeb.2009.0618)
16. Woitowich NC, Beery AK, Woodruff TK. 2020 A
10-year follow-up study of sex inclusion in the
biological sciences. eLife 9,1–8. (doi:10.7554/
eLife.56344)
17. Oufiero CE, Garland Jr T. 2009 Repeatability and
correlation of swimming performances and size
over varying time-scales in the guppy (Poecilia
reticulata). Funct. Ecol. 23, 969–978. (doi:10.
1111/j.1365-2435.2009.01571.x)
18. Turesson H, Satta A, Domenici P. 2009
Preparing for escape: anti-predator posture
and fast-start performance in gobies.
J. Exp. Biol. 212, 2925–2933. (doi:10.1242/jeb.
032953)
19. Domenici P, Blake R. 1997 The kinematics and
performance of fish fast-start swimming. J. Exp.
Biol. 200, 1165–1178.
20. Sloman KA, Taylor AC, Metcalfe NB, Gilmour KM.
2001 Effects of an environmental perturbation
on the social behaviour and physiological
function of brown trout. Anim. Behav. 61,
325–333.
21. R Core Development Team. 2020 R: A language
and environment for statistical computing.
Vienna, Austria: R Foundation for Statistical
Computing, Version 3.6.3.
22. Wheeler RE, Torchiano M. 2016 Permutation
tests for linear models in R. R Package Version
2.1.0.
23. Drummond GB, Vowler SL. 2012 Different tests
for a difference: how do we do research?
J. Physiol. (Lond) 590, 235–238. (doi:10.1113/
jphysiol.2011.225235)
24. Ludbrook J, Dudley H. 1998 Why permutation
tests are superior to t and F tests in biomedical
research. Am. Stat. 52, 127–132.
25. Canty A, Ripley B. 2020 R package ‘boot’version
1.3. CRAN Repository.
26. Krause J, James R, Franks DW, Croft DP. 2015
Animal social networks. Oxford, UK: Oxford
University Press.
27. Ward AJW, Webster MM. 2016 Sociality: The
behaviour of group-living animals. Geneva,
Switzerland: Springer International Publishing.
28. Conradt L, Roper TJ. 2000 Synchrony and social
cohesion: a fission-fusion model. Proc. R. Soc.
Lond. B 267, 2213–2218.
29. Radford AN, Majolo B, Aureli F. 2016 Within-
group behavioural consequences of between-
group conflict: a prospective review, Table 1.
Proc. R. Soc. B 283, 20161567. (doi:10.1098/
rspb.2016.1567)
30. del Mar Delgado M, Miranda M, Alvarez SJ,
Gurarie E, Fagan WF, Penteriani V, di Virgilio A,
Morales JM. 2018 The importance of individual
variation in the dynamics of animal collective
movements. Phil. Trans. R. Soc. B 373,
20170008-10. (doi:10.1098/rstb.2017.0008)
31. Filby AL, Paull GC, Bartlett EJ, Van Look K, Tyler
CR. 2010 Physiological and health consequences
of social status in zebrafish (Danio rerio).
Physiol. Behav. 101, 576–587.
32. Border SE, DeOliveira GM, Janeski HM, Piefke TJ,
Brown TJ, Dijkstra PD. 2019 Social rank, color
morph, and social network metrics predict
oxidative stress in a cichlid fish. Behav. Ecol. 30,
490–499. (doi:10.1093/beheco/ary189)
33. Seebacher F, Krause J. 2017 Physiological
mechanisms underlying animal social
behaviour. Phil. Trans. R. Soc. B 372, 20160231-
8. (doi:10.1098/rstb.2016.0231)
34. Killen SS, Mitchell MD, Rummer JL, Chivers DP,
Ferrari MCO, Meekan MG, McCormick MI.
2014 Aerobic scope predicts dominance
during early life in a tropical damselfish. Funct.
royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 8: 210146
7
Ecol. 28, 1367–1376. (doi:10.1111/1365-2435.
12296)
35. Hoare DJ, Couzin ID, Godin JGJ, Krause J. 2004
Context-dependent group size choice in fish.
Anim. Behav. 67, 155–164.
36. Teles MC, Dahlbom SJ, Winberg S, Oliveira RF.
2013 Social modulation of brain monoamine
levels in zebrafish. Behav. Brain Res. 253,
17–24. (doi:10.1016/j.bbr.2013.07.012)
37. Gómez-Laplaza LM, Morgan E. 2003 The
influence of social rank in the angelfish,
Pterophyllum scalare, on locomotor and feeding
activities in a novel environment. Lab. Anim. 37,
108–120.
38. Wong MYL, Munday PL, Buston PM, Jones GP.
2008 Fasting or feasting in a fish social
hierarchy. Curr. Biol. 18, R372–R373. (doi:10.
1093/beheco/arm141)
39. Lailvaux SP, Husak JF. 2017 Predicting
life-history trade-offs with whole-organism
performance. Am. Zool. 57, 325–332. (doi:10.
1093/icb/icx073)
40. Cooper Jr WE, Samia DSM, Blumstein DT. 2015
Fear, spontaneity, and artifact in economic
escape theory: a review and prospectus. Adv.
Stud. Behav. 47, 147–179. (doi:10.1016/bs.asb.
2015.02.002)
41. Miln C, Ward AJW, Seebacher F. 2021
Data from: Social rank and not
physiological capacity determines
competitive success in zebrafish (Danio rerio).
Dryad Digital Repository (doi:10.5061/dryad.
v15dv41tr)
royalsocietypublishing.org/journal/rsos R. Soc. Open Sci. 8: 210146
8
Available via license: CC BY 4.0
Content may be subject to copyright.
Content uploaded by Frank Seebacher
Author content
All content in this area was uploaded by Frank Seebacher on May 12, 2021
Content may be subject to copyright.
