The Association Between Personality Traits, Morphological Traits and Alternative Mating Behaviour in Male Endler's Guppies, Poecilia wingei

Abstract

Alternative mating behaviour, personality traits and morphological characters are predicted to be correlated. Bolder, larger and more colourful males are expected to preferentially court females, while shy, small and drab-coloured individuals are predicted to sneak copulations. We used males of Endler's guppy, Poecilia wingei, to test this association over a long temporal period (hence including ontogenetic changes) and under two social environments (male-biased and female-biased). We found that personality traits (exploration, boldness, activity) of P. wingei males were highly repeatable across long time spans, but they were not correlated (formed no behavioural syndrome). Male age and social environment had no effect on any personality trait, despite their effects on alternative mating behaviour. Young males with higher activity levels were more likely to attempt sneaking. In older fish, there was an association between orange coloration, courtship and boldness, but this was not observed in young males. Our results suggest that alternative mating behaviour is more flexible than personality traits and is independent of them. Non-colour-based morphological traits (gonopodium length, body length, caudal straps length, dorsal fin length) were not correlated with any particular mating behaviour.

Full text

RESEARCH PAPER
The Association Between Personality Traits, Morphological
Traits and Alternative Mating Behaviour in Male Endler’s
Guppies, Poecilia wingei
Radomil

Re
zucha & Martin Reichard
Institute of Vertebrate Biology, Academy of Sciences of the Czech Republic, Brno, Czech Republic
Correspondence
Martin Reichard, Institute of Vertebrate
Biology, Academy of Sciences of the Czech
Republic, Kv
etn
a 8, 603 65 Brno, Czech
Republic.
E-mail: reichard@ivb.cz
Received: December 18, 2015
Initial acceptance: January 20, 2016
Final acceptance: February 21, 2016
(J. Schneider)
doi: 10.1111/eth.12491
Keywords: alternative mating tactic, animal
personality, boldness, experience, sexual
selection, social environment
Abstract
Alternative mating behaviour, personality traits and morphological char-
acters are predicted to be correlated. Bolder, larger and more colourful
males are expected to preferentially court females, while shy, small and
drab-coloured individuals are predicted to sneak copulations. We used
males of Endler’s guppy, Poecilia wingei, to test this association over a long
temporal period (hence including ontogenetic changes) and under two
social environments (male-biased and female-biased). We found that per-
sonality traits (exploration, boldness, activity) of P. wingei males were
highly repeatable across long time spans, but they were not correlated
(formed no behavioural syndrome). Male age and social environment had
no effect on any personality trait, despite their effects on alternative mat-
ing behaviour. Young males with higher activity levels were more likely
to attempt sneaking. In older fish, there was an association between
orange coloration, courtship and boldness, but this was not observed in
young males. Our results suggest that alternative mating behaviour is
more flexible than personality traits and is independent of them. Non-col-
our-based morphological traits (gonopodium length, body length, caudal
straps length, dorsal fin length) were not correlated with any particular
mating behaviour.
Introduction
Alternative mating behaviour (AMB) describes intra-
sexual differences in mating behaviour among indi-
viduals within a population to maximize their
reproductive success. AMB is primarily performed by
individuals with different competitive abilities and
forms an important aspect of intrasexual competition
for fertilization success (Oliveira et al. 2008). AMB
can be fixed for life (Shuster & Wade 1991) but often
is highly flexible (Candolin 2004), and its expression
and success are affected by a range of environmental
and demographic factors (reviewed in Kokko & Ran-
kin 2006; Taborsky 1994).
The role of morphological traits (e.g. size and col-
our) on the adoption of AMB is well established
(Andersson 1994). In contrast, more abstract sources
of interindividual variation related to AMB, such as
cognitive and personality traits, remain relatively
neglected (Sih 2013; but see Smith et al. 2015),
despite their potential to affect expression and success
of the AMB. While it was demonstrated that different
behavioural types had different reproductive (e.g.
Dingemanse et al. 2005) and mating success (e.g.
Reaney & Backwell 2007), our understanding of the
links between personality traits and sexual selection
remains surprisingly limited (Schuett et al. 2010; but
see Godin & Dugatkin 1996; Schuett et al. 2011).
Personality traits possess a degree of plasticity (Gal-
hardo et al. 2012; Thomson et al. 2012), and individ-
uals can adjust their behaviour on the basis of
external cues while still consistently differing from
each other (Mathot et al. 2012). Social learning (Frost
et al. 2007) and social experience (Modlmeier et al.
2014) should enable males to modulate their repro-
ductive behaviour according to female responses.
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH 1
Ethology

Hence, males frequently rejected by females may ben-
efit from adjusting their behaviour to support success-
ful sneaking, while males receiving female interest
may benefit from a different experience-based modu-
lation of their behaviour. Personalities constrain the
overall magnitude of these changes, but plasticity of
their expression enables to modulate particular beha-
viours over time in relation to current individual con-
dition and social environment (Piyapong et al. 2010;
Thomson et al. 2012). Hence, there is scope for adap-
tive exploitation of variable social environment by
males through modulation of personality-related
behaviours (Bergm€
uller & Taborsky 2010).
We investigated a potential link between individual
personality traits and AMB in the sneaker/courter
complex and its response to contrasting social envi-
ronments using Endler’s guppy, Poecilia wingei, a small
poeciliid fish. It is a sister species of the common
guppy (Poecilia reticulata), with a largely similar suite
of reproductive behaviours. Male P. wingei either
court females using sigmoid displays and solicit coop-
erative mating or try to circumvent female choice by
sneak (coercive) copulations. In P. reticulata, AMB is
affected by diverse demographic and environmental
factors (Houde 1997). Individual males vary greatly in
their use of courting and sneaking and differ in the
frequency of their switching (Houde 1997), implying
that the use of AMB is flexible (Magurran & Magellan
2007). There is clear and repeatable interindividual
variation in particular personality traits in P. reticulata
(e.g. Budaev 1997; Brown & Irving 2014).As with
many other taxa, guppy personality traits may be
grouped into correlated suites of behaviours called
behavioural syndromes (BSs) (Smith & Blumstein
2012). Personality traits also have considerable conse-
quences for guppy individual fitness (Smith & Blum-
stein 2010) and, in another poeciliid fish, Gambusia
holbrooki, personality can be modulated by a modifica-
tion of environmental factors (Sinclair et al. 2014).
How social conditions affect expression of personal-
ity traits is less clear (but see Piyapong et al. 2010).
Male P. wingei are capable of modifying their repro-
ductive behaviour according to their phenotypic
traits, social environment and intensity of sexual
selection (

Re
zucha & Reichard 2014, 2015). It is not
known whether there is any link between personality
and AMB and, if so, whether personality traits and
the use of AMB can be modulated. In this study, we
tested the following: (1) the existence of personalities
by testing repeatability of the main behavioural traits
(exploration, boldness towards predator and general
activity level) in P. wingei males and their possible for-
mation into a BS; (2) the link between personality
traits (particularly boldness, as indicated by predator
inspection) and AMB, accounting for individual
morphological traits; and (3) consistent changes in
individual personality traits after 5 wk of exposure
to contrasting social environments (female-biased 9
male-biased environment).
We predicted differences in personality traits and
formation of distinct BSs between sneakers and cour-
ters. We expected that more conspicuous individuals
(i.e. with more carotenoid coloration) would have a
greater propensity towards predator inspection and
would court females rather than sneak (Godin &
Dugatkin 1996; Jirotkul 2000; Kiritome et al. 2012).
Sneakers, being generally less colourful and thus less
able to attract females (Houde 1997; Oliveira et al.
2008), were predicted to be more active in order to
increase female encounter rate. Males from a female-
biased environment (with no male rivals) were pre-
dicted to be bolder, but less active and explorative due
to their unrestricted access to females and the lack of
competition with rivals.
Material and Methods
Experimental Animals
Experimental fish came from our breeding stock com-
posed of outbred descendants of P. wingei imported
from Laguna de los Patos (northern Venezuela) in
2007. Over a period of 1 mo, all emerging juveniles
were collected from the stock aquarium (120 l) and
isolated in a separate 72-l aquarium. Juveniles were
frequently sexed, and females were removed and kept
separately.
Sixty-five males (approximately 18.5 wk old, sexu-
ally mature) were collected from the 72-l male stock
aquarium and individually housed in 2-l plastic aqua-
ria enriched with artificial plants. The aquaria were
visually separated from each other. Fish experienced a
natural light regime, and additional light was pro-
vided for 12 h a day (08:00–20:00) by a 40-W Sun
Glo daylight spectrum fluorescent tube. Water tem-
perature fluctuated between 22 and 25°C. Males were
fed daily, and water was partly exchanged every 2 wk
(for full details, see

Re
zucha & Reichard 2015).
Behaviour Trials Treatments
Behavioural trials were conducted in the same order
for all subjects before and after change in the social
environment, as recommended by Bell (2013). An
alternative is to randomize the order of trials for each
subject to minimize any potential ‘carry-over effects’
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH2
Personalities in Endler’s Guppy Males R.

Re
zucha & M. Reichard

(Logue et al. 2009; Dochtermann 2010; Bell 2013).
Carry-over effect may result in false short-term beha-
vioural correlations; none was observed in our study,
minimizing (tough not excluding) the possibility of
carry-over effects (see ‘Results’).
The first test of male behaviour (Trial 1) was per-
formed after approximately 1 wk of isolation. After all
Trial 1 tests were completed, two females were added
into 33 randomly assigned aquaria to create a female-
biased social environment (FBE: 1 male, 2 females).
In the remaining 32 aquaria, a single male and single
female from the stock population were added to the
focal male to create a male-biased social environment
(MBE: 2 males, 1 female). The second test of focal
male behaviour (Trial 2) was completed after approxi-
mately 5 wk of experience of the particular social
environment (Fig. 1). Males had unrestricted access
to females (and rivals in the MBE) and gained mating
experience in their particular social environment. The
treatment was maintained by replacing any dead
female or non-experimental male. Focal males were
not replaced; natural mortality of focal males caused a
minor decrease in sample size during the second trials
(FBE, n =22; MBE, n =17). Upon completion of Trial
2, males were returned to their treatment aquarium
until their natural death.
Exploration Trials
Individual explorative behaviour was tested twice: at
the age of approximately 19 wk (Exploration 1) and
31 wk (Exploration 2). We adopted a test of emer-
gence from a refuge to study individual’s exploration
level, as commonly used in fish personality studies
(e.g. Brown et al. 2005; Fabre et al. 2014).Trials took
place between 09:00 and 17:00 Central European
Time in a 43.5-l aquarium (70 925 925 cm), with
three sides covered with opaque barriers to avoid dis-
turbing the fish. The bottom of the aquarium was cov-
ered with a 1-cm layer of river gravel. The water
depth was 5 cm and ambient temperature matched
that of holding aquaria (23–25°C). The water was aer-
ated, but aeration was paused during behavioural
observations. Illumination was provided by a 25-W
(a) (b)
(c) (d)
Fig. 1: Association between personality traits and male phenotypes. Relationship between area of orange coloration and boldness (number of preda-
tor inspection events) in young (r
2
=0.005) (a) and older males (r
2
=0.21) (b). Relationship between activity (number of zones crossed) and sneaking
(number of sneaking attempts) in young inexperienced males (r
2
=0.07) (c) and older males (r
2
=0.004) (d).
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH 3
R.

Re
zucha & M. Reichard Personalities in Endler’s Guppy Males

Sun Glo daylight spectrum bulb positioned above the
aquarium, which also received indirect natural day-
light.
A randomly chosen focal male was placed in a
refuge represented by a plastic box [16 (h) 915
(w) 98 (d) cm]. After 2 min, small doors
(4.5 94 cm) were opened and the time taken to
swim out from the refuge to the open arena by a full
body length was recorded. Males that remained in the
refuge for 10 min were considered as minimally
explorative, and a score of 600 s was assigned to
them.
Open Field Test: Boldness Towards Predator and
General Activity
Individual levels of boldness and activity were tested
at the age of approximately 20 wk (Boldness 1) and at
the age of 32 wk (Boldness 2), that is in each case
approximately 1 wk after the exploration trials. An
open field test under the risk of potential predation
was used to test individual behaviour towards preda-
tor and general activity. A slightly modified set-up
from Frommen et al. (2009) was used. The same
aquarium was used as in the exploration test (43.5 l,
70 925 925 cm with three sides covered with opa-
que barriers). The aquarium was divided by lines
drawn on the front side into five compartments (0–4).
Compartment 0 was the starting compartment, com-
partments 1, 2 and 3 were neutral, and Compartment
4 was the stimulus section with a fish predator. Com-
partments 0 and 4 were separated by glass dividers.
The divider between compartments 3 and 4 was cov-
ered with a black opaque lid to prevent the predator
from being seen; however, diffusion of predator olfac-
tory cues was not prevented. Commercially obtained
blue acara (Andinoacara pulcher (Gill 1858), Cichlidae),
a generalized predator sympatric to P. wingei, was
used as a potential predator. A male acara (total
length 11 cm) was placed into Compartment 4
30 min prior to the onset of trials. Predator fish
remained in a stable position and did not show any
signs of stress or aggressive behaviour. A single indi-
vidual of the blue acara was used across all replicates.
A randomly chosen focal male was gently released
into Compartment 0 of the test aquarium. After 2-
min acclimatization, the glass divider was lifted to
allow free access to the test arena (compartments 0–
3). At the same time, the opaque barrier covering
divider of the predator compartment was lifted to
allow visual contact with the predator. Focal male
behaviour was scored using JWatcher 1.0 (Blumstein
& Daniel 2007) for 10 min. The following behaviours
were scored: percentage of time spent in each com-
partment, number of movements between compart-
ments (a score of male activity) and number of
individual inspection events towards the predator
(boldness). Inspection events were defined as slow,
hesitant moves by the focal male oriented towards the
predator and were scored irrespective to the focal
male position.
Courtship Behaviour Trials
Male mating behaviour was tested at the age of
approximately 25 wk (Courtship 1) and 32 wk
(Courtship 2). At Courtship 1, males were virgin and
had no prior experience with an adult receptive
female. Mating behaviour was scored in a 6-l aquar-
ium with three sides covered with an opaque barrier
to minimize disturbance. Light was provided by a 25-
W Sun Glo daylight spectrum bulb positioned above
the aquarium. Females used in the experiment were
collected from a virgin female aquarium and left with
a group of 3 adult non-experimental males from the
stock aquarium for 1 d prior to the experimental trials
to standardize their receptivity (

Re
zucha & Reichard
2014). During Courtship 2 trials (after exposure to
social environment treatments), focal males were iso-
lated from all conspecifics for 1.5 d prior to testing to
standardize their mating effort. All replicates were
completed between 09:00 and 16:00, that is later than
2 h after sunrise and 2 h before sunset.
A female was allowed to settle for 5 min in the test
aquarium. Then, a randomly chosen focal male was
released to the aquarium and left for 10 min to settle.
After settling, male mating behaviour was scored for
15 min by a single observer using JWatcher 1.0. The
number and duration of sigmoid displays (courtship),
number of gonopodial thrusts (sneaking), number of
gonopodial swings [loading gonopodium with sperm
(Pilastro & Bisazza 1999)], number of gonopore nips
[possibly analysing female cues (Herdman et al.
2004)] and general interest in females were scored.
General interest was defined as time (in seconds)
when male’s head was oriented towards the female
and there was a maximum distance of 10 cm between
the male and female. After 15 min, the focal male
was gently captured and returned to his home
aquarium.
Photographing and Analysis of Colour Pattern
Focal males were photographed at the age of approxi-
mately 19 and 32 wk. A detailed description is given
in

Re
zucha & Reichard (2015). In brief, a male was
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH4
Personalities in Endler’s Guppy Males R.

Re
zucha & M. Reichard

gently released into a photo-aquarium, fixed in a
stable position against front glass and quickly pho-
tographed from both sides. A dark case and flash illu-
mination were used to standardize conditions and a
small ruler provided a scale. All pictures were taken in
the shortest possible time, typically <4 min. Images
were processed in Fiji (Schindelin et al. 2012) by
measuring the area of carotenoid (light orange to dark
red), melanin (dark black and fuzzy black) and irides-
cent (mostly green, blue, purple and silver, formed by
guanine crystals in iridiophores) patches, and by
counting the number of separate carotenoid and mel-
anin spots. Each colour was expressed as a propor-
tional area of lateral body projection (excluding eye,
gonopodium and fins, with the exception of the cau-
dal fin). The length of the lower and upper coloured
part of the caudal fin (strap; black and carotenoid col-
our combined), gonopodium length, body length and
length of the dorsal fin were also measured from the
photographs. The measurement of morphological
traits was completed blind with respect to behavioural
data.
Data Analyses
All statistical analyses were performed in the R envi-
ronment (R Development Core Team 2009). Repeata-
bility was tested using intraclass correlation in the
psych package, using ICC1 (standard repeatability sensu
Lessells & Boag (1987)) and ICC3 (standardized for
overall change in trait mean between time points,
time treated as a fixed factor). For measures of person-
ality traits that can only be expressed after emergence
from the refuge (activity, boldness), the frequency of
behaviour was expressed as a rate per unit time.
An information theoretical approach (Burnham &
Anderson 2002) was used to test the effects of predic-
tors on behavioural traits. A set of biologically plausi-
ble models was constructed, with various level of
complexity. Before applying statistical models, data
exploration was undertaken as recommended by Zuur
et al. (2009). Collinearity among explanatory vari-
ables and among three response variables (personality
traits) was tested using variance inflation factors
(VIF). The VIF quantifies how much the variance
increased due to collinearity among explanatory vari-
ables. Values of global variance inflation factor (GVIF)
lower than 2–10 are considered to indicate an accept-
able level of multicollinearity (Zuur et al. 2009).
There was no collinearity (all GVIF <3) among vari-
ables selected for model construction. Null models
(i.e. intercept-only models) were always included,
and the fit of predictor models (in the default stat
package) was always compared with the null model.
A total of 31 models of varying complexity were com-
pared (Appendix 1). The models were selected on the
basis of our hypotheses and their alternatives with a
biological relevance. Where appropriate, Poisson
(counts) and gamma (time) distribution (generalized
linear models, GLM) and log-transformation to nor-
malize data (linear models, LM) were considered to
improve the fit of the models. Fits of alternative mod-
els (listed in Appendix 1) were compared using the
Akaike information criterion corrected for small sam-
ple size (AIC
c
)(MuMIn package). Model weights were
calculated from relative likelihoods. Models within
DAIC <2 were considered to have comparable support
(Burnham & Anderson 2002). When the null model
was included in the set of the best candidate models
(i.e. within DAIC <2), models with predictor vari-
ables were considered to have no explanatory power.
Only analyses where the null models were not within
the set of the best models were considered further. If
more than one parameterized model gained a support
within DAIC <2, they were subjected to a model
averaging procedure (using MuMIn package) and esti-
mates from averaged models are presented. Original
parameters of all models from the set of best models
are presented in Table S1. In models within DAIC <2
(used for model averaging), the residuals were exam-
ined to ensure model assumptions were met. Models
for Trial 1 and Trial 2 for each personality had identi-
cal structure but were treated separately.
The effect of male age on personality traits (pairwise
difference in trait values between trials 1 and 2) was
analysed using a paired t-test. The effect of social envi-
ronment on personality traits (male-biased vs. female-
biased, i.e. only data from Trial 2 for each trait) was
compared using ANOVA. All p-values are for two-
tailed test.
Results
All personality traits (exploration, boldness, activity)
were significantly repeatable (Table 1a). Coloration
and morphological traits were also highly repeatable
across measurements, with a stronger repeatability
when the ICC3 was used (i.e. accounting for popula-
tion-level change in trait expression between trials 1
and 2) (Table 1b). Courtship behaviour, in contrast,
was less repeatable. Only general male interest and
gonopodial swings were repeatable. Sigmoid displays,
sneaking attempts and gonopore nipping were not
repeatable at all (Table 1c) because males responded
to the change in their social environment (see

Re
zucha & Reichard 2014).
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH 5
R.

Re
zucha & M. Reichard Personalities in Endler’s Guppy Males

There was no BS. Personality traits were not associ-
ated with each other (Pearson correlation: exploration
and boldness: r=0.094, p =0.482; exploration and
activity: r=0.111, p =0.406; boldness and activity:
r=0.147, p =0.272).
Male exploration (time to emergence from hide)
was not associated with any morphological or court-
ship traits; the null model had highest explanatory
power (Appendix 1). Male age (paired t-test,
t
36
=0.55, p =0.59) and social environment (log-
transformed, F
1,36
=0.82, p =0.37) had no effect on
male exploration.
Male boldness towards the predator was not associ-
ated with any morphological or courtship trait in
young males (Appendix 1). In older males (Trial 2),
more carotenoid coloration (Fig. 1a) and a higher ten-
dency to court females were positively associated with
boldness towards predators (Table 2a). Male age
(paired t-test, t
36
=0.87, p =0.39) and social environ-
ment (F
1,36
=0.33, p =0.57) had no effect on male
propensity to inspect predator (decrease by DAIC 1.5–
3.0 for paired comparisons between otherwise identi-
cal models), implying that social treatment did not
affect male boldness levels.
In young males, activity was positively associated
with the rate of sneaking (Fig. 1b); the effect of male
size was retained in the final averaged model
(Appendix 1), but was not significant (Table 2b). This
association ceased in sexually experienced males
(Fig. 1d, Appendix 1). Male age (paired t-test,
t
36
=0.33, p =0.75) and social environment (log-
transformed data, F
1,36
=0.01, p =0.93; decrease by
DAIC 1.5–3.0 for paired comparisons between other-
wise identical models) had no effect on male activity
level.
There was no population-level change in any
personality trait values between young and old
males (exploration: F
1,32
=0.16, p =0.69, boldness:
F
1,32
=0.03, p =0.83, activity: F
1,32
=0.21, p =0.65).
A temporal change in courtship behaviour (and its
response to social environment) has been reported
elsewhere (

Re
zucha & Reichard 2014).
Discussion
Repeatable Behaviour But a Lack of BS
All personality traits (exploration, boldness towards
predator, general activity) of male P. wingei were
highly repeatable despite the change in social envi-
ronment during the 3 mo of the study. This finding is
consistent with data on personalities in the closely
related P. reticulata (Budaev 1997; Smith & Blumstein
2010; Brown & Irving 2014) and other taxa (Bell et al.
2009). At the same time, no BS was detected, suggest-
ing the independence of different personality traits in
focal males. This is surprising, because BSs encom-
passing boldness towards predator, refuge use and
activity are common (e.g. Mazu
e et al. 2015; Wilson
& Godin 2009 but see McEvoy et al. 2015), and hid-
ing in a refuge is a form of protection and a certain
level of boldness is needed to leave the refuge and
Table 1: Repeatability of male traits between the first and second trials
ICC1 p ICC3 p
(a) Personality traits
Exploration 0.29 0.036 0.28 0.042
Inspection 0.39 0.006 0.39 0.007
Activity 0.38 0.007 0.38 0.009
(b) Morphological traits
Orange coloration (%) 0.46 0.001 0.46 0.001
Black coloration (%) 0.30 0.033 0.32 0.023
Iridescent coloration (%) 0.09 0.703 0.37 0.009
Dot asymmetry 0.13 0.210 0.14 0.210
Orange dots (N) 0.67 0.001 0.73 0.001
Black dots (N) 0.43 0.004 0.41 0.005
Dorsal fin length 0.35 0.001 0.62 <0.001
Lower sword length 0.69 <0.001 0.76 <0.001
Upper sword length 0.60 <0.001 0.86 <0.001
Gonopodium length 0.91 <0.001 0.91 <0.001
Total length 0.77 <0.001 0.78 <0.001
Standard length 0.85 <0.001 0.85 <0.001
(c) Mating behaviour traits
General interest 0.32 0.018 0.37 0.008
Gonopodial swings 0.28 0.026 0.29 0.032
Gonopore nipping 0.15 0.170 0.17 0.140
Sigmoid display duration 0.19 0.900 0.39 0.600
Gonopodial thrusts 0.01 0.520 0.01 0.570
ICC1 is standard repeatability sensu Lessells & Boag (1987) and ICC3
denotes repeatability standardized for overall change in population trait
mean between time points.
Table 2: Model-averaged estimates (and their SE) from the top models
(by >2 AIC) and their statistical significance for (a) boldness towards
predators in older males and (b) activity of young males
Estimate SE z-score p Value
(a) Boldness of old males
Intercept 0.402 0.297 1.31 0.191
Sigmoid displays 0.028 0.013 2.07 0.039
Percentage of orange 0.103 0.042 2.38 0.018
(b) Activity of young males
Intercept 21.088 28.641 0.73 0.469
Gonopodial thrustsa 4.949 2.270 2.13 0.033
Total length 1.574 1.759 0.87 0.383
a
log(x+1)-transformed prior analysis.
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH6
Personalities in Endler’s Guppy Males R.

Re
zucha & M. Reichard

start exploration. Behavioural syndromes are often
linked with aggression to conspecifics and appear to
be promoted by the same underlying mechanisms
(Caramaschi et al. 2013), responsible for the coupling
of behavioural traits, including reproductive beha-
viour (Mutzel et al. 2013). This enables females to
judge male’s mating or parental quality (Barnett et al.
2012; Stein & Bell 2012). However, particular BSs can
also increase male reproductive success irrespective of
female choice (Sih et al. 2014). While we suggest that
the lack of correlation between exploration and bold-
ness towards predator in our data set was a real phe-
nomenon, it is possible that larger sample size
(Garamszegi et al. 2012) or inclusion of positional
data during predator inspection events and their use
in the analysis would strengthen the association.
Our findings underscore strong context-specific dif-
ferences in boldness (Coleman & Wilson 1998). The
boldness needed to emerge from a safe refuge was
unrelated to the boldness needed for gaining informa-
tion about a potential threat in P. wingei. Similarly,
Carter et al. (2012) did not observe any relationship
between response to a threatening stimulus (snake)
and reaction to a novel object (new food item) in wild
chacma baboons (Papio urcinus). We also did not
record any association between exploration and gen-
eral activity level, despite frequent correlation between
them reported elsewhere (e.g. Fraser et al. 2001; Sih
et al. 2004; but see Conrad et al. 2011; Sinclair et al.
2014). It is possible that the link between exploration
and activity is adaptive only in certain environments
and under particular circumstances. Both exploration
and activity are also prone to be strongly state depen-
dent, implying the existence of such an association
only under specific conditions, such as low energy
reserves (Wolf & Weissing 2010). Individual behaviour
also seems to be largely influenced by the presence of
predators, which may promote behavioural correla-
tions on the one hand (Bell & Sih 2007) and uncover
behavioural flexibility on the other (David et al.
2014). One caveat of our study was that our measure
of activity level was not independent of boldness
towards predator, potentially resulting in a false signif-
icant association. However, these two behaviours were
not correlated in our study.
Another reason for the lack of correlations among
personality traits may stem from the paucity of selec-
tive pressure in captivity (Archard & Braithwaite
2010), with predation risk being considered a major
source of personality trait correlations (Bell 2005; Bell
& Sih 2007). Domestication has a clear effect on fish
behaviour (Metcalfe et al. 2003; Huntingford 2004)
and may have therefore also affected the outcome of
this study. Experimental males were kept in a pre-
dictable environment without predation or other
external sources of mortality for several generations,
similar to that of domestic strains. While Bleakley
et al. (2006) demonstrated that individuals of inbred
P. reticulata strains retained the capacity to respond
appropriately to predator cues, we acknowledge that a
lack of BS may arise from relaxed selection in a
benign captive environment.
No Effect of Social Environment on Male Personality
Social environment did not affect individual male per-
sonality over the period of 5 wk, leading to the con-
clusion that it has limited impact on personality traits
modulation, contrary to, for example, abiotic factors
(Sinclair et al. 2014). This result corroborates the
finding of overall relative stability of individual per-
sonality traits despite changes in surrounding social
conditions, such as reported in the water strider
Aquarius remigis (Sih & Watters 2005). In contrast, sex
ratio affected individual boldness in P. reticulata males
who were bolder in male-biased than in female-biased
groups (Piyapong et al. 2010). Nevertheless, Piyapong
et al.’s (2010) study tested only the actual level of
boldness in various contexts, without taking individ-
ual experience or ontogenetic changes into account.
Personality, AMB and Morphological Traits
There was some association between personality and
AMB. In young virgin males, more active individuals
were more likely to attempt sneak copulations than
less active individuals. This is consistent with the link
between male activity levels and mate searching.
Sneaking is less time- (Parker 1974) and energy (Sar-
gent 1985)-consuming than courtship, and sneaker
males can spend more time and energy on searching
and chasing females to sneak copulations. Among
older males, individuals with a larger proportion of
orange, carotenoid-based coloration had a higher ten-
dency to court females and were bolder towards
predators, as has been found in male P. reticulata
(Godin & Dugatkin 1996).Guppy males with brighter
coloration are, as in many other taxa, considered to
be more viable due to higher intrinsic quality (Godin
& McDonough 2003; Locatello et al. 2006), preferred
by females for solicited matings (Alexander & Breden
2004), but also preferentially targeted by predators
(Godin & McDonough 2003). Male courtship is costly
and makes males conspicuous to predators. Brighter
males can therefore invest in predator inspection and
acquire information about potential threats and hide
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH 7
R.

Re
zucha & M. Reichard Personalities in Endler’s Guppy Males

or, alternatively, deter predator attacks (Dugatkin &
Godin 1992). The link between courtship and bold-
ness towards predator was only manifested in older
males. It may be speculated that social experience, in
terms of interaction with females, is needed to prop-
erly develop boldness in courting colourful males.
The morphological traits tested (gonopodium
length, body length, coloured caudal straps length
and length of dorsal fin) were not important in deter-
mining individual choice of AMB in P. wingei. This
contrasts with a positive correlation between gonopo-
dium length and number of gonopodial thrusts
reported for P. reticulata (Reynolds et al. 1993),
although the association was only apparent under
low light conditions. Similarly, Karino & Kobayashi
(2005) reported a positive link between the rate of
gonopodial thrusts and caudal fin length in P. reticu-
lata. The difference may simply reflect variation
among populations with variable selection regimes.
Alternatively, the lack of correlation between AMB
and morphological traits in P. wingei despite their
presence in some P. reticulata populations may arise
from slight differences in their mating behaviour
(Poeser et al. 2005) and lower level of intra- and
intersexual aggression in P. wingei males.
Conclusion
Personality traits of Poecilia wingei males were highly
repeatable, despite no BS being identified. There was
no general difference in any personality trait between
Trial 1 and Trial 2 or between the two social environ-
ments, that is living with a rival, or females did not
make males bolder, more explorative or more active.
Young active males were more likely to attempt sneak
copulations. Among older fish, more orange, courting
males were more likely to inspect predators irrespec-
tive of their social environment –but in virgin males
this was not observed. The resulting temporal change
of individual AMB in the course of 5 wk and relative
stability of personality traits suggest higher flexibility
of AMB. It seems that personality traits are relatively
fixed in a short timescale compared with more flexible
AMB. It is notable because a dramatic change in the
social environment was predicted to at least slightly
influence the expression of personality traits assuming
that different traits are favoured in different environ-
ments. As a result, social environments in this experi-
ment only influenced AMB while personality traits
remained unchanged. Taken together, our results sug-
gest that AMB is independent of personality traits as
there was no personality characteristic predisposing
males to adopt either courting or sneaking as a gen-
eral strategy.
Acknowledgements
We are grateful to Radim Bla
zek, Matej Pola
cik and
Milan Vrt
ılek for their help with experiments, discus-
sions and comments, to Ji
r
ı Farka
c for his help with
fish maintenance and to Rowena Spence for com-
ments and linguistic correction. This paper was sub-
mitted via Axios Review, with Alison Bell and three
anonymous reviewers providing insightful feedback
on the article. Financial support came from Czech
Science Foundation GACR P505/12/G112. All appli-
cable international, national and institutional guideli-
nes for the care and use of animals were followed.
Experimental procedures were approved by the ethi-
cal committees of the IVB and Ministry of Agriculture
(CZ 62760203) and are in accordance with Czech
legal requirements. R

R and MR hold a licence for con-
ducting experimental work on vertebrates.
Conflict of interest
The authors declare no conflict of interest.
Literature Cited
Alexander, H. J. & Breden, F. 2004: Sexual isolation and
extreme morphological divergence in the Cuman
a
guppy: a possible case of incipient speciation. J. Evol.
Biol. 17, 1238—1254.
Andersson, M. 1994: Sexual Selection. Princeton Univ.
Press, London.
Archard, G. A. & Braithwaite, V. A. 2010: The importance
of wild populations in studies of animal temperament.
J. Zool. 281, 149—160.
Barnett, C. A., Thompson, C. F. & Sakaluk, S. K. 2012:
Aggressiveness, boldness and parental food provisioning
in male house wrens (Troglodytes aedon). Ethology 118,
984—993.
Bell, A. M. 2005: Behavioural differences between individ-
uals and two populations of stickleback (Gasterosteus
aculeatus). J. Evol. Biol. 18, 464—473.
Bell, A. 2013: Randomized or fixed order for studies
of behavioral syndromes? Behav. Ecol. 24,
16—20.
Bell, A. M. & Sih, A. 2007: Exposure to predation gener-
ates personality in threespined sticklebacks (Gasterosteus
aculeatus). Ecol. Lett. 10, 828—834.
Bell, A. M., Hankison, S. & Laskowski, K. L. 2009: The
repeatability of behaviour: a meta-analysis. Anim.
Behav. 77, 771—783.
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH8
Personalities in Endler’s Guppy Males R.

Re
zucha & M. Reichard

Bergm€
uller, R. & Taborsky, M. 2010: Animal personality
due to social niche specialisation. Trends Ecol. Evol. 25,
504—511.
Bleakley, B. H., Martell, C. M. & Brodie, E. D. 2006: Varia-
tion in anti-predator behavior among five strains of
inbred guppies, Poecilia reticulata. Behav. Genet. 36,
783—791.
Blumstein, D. T. & Daniel, J. C. 2007: Quantifying Behav-
ior: The JWatcher Way. Sinauer Associates Inc,
Sunderland.
Brown, C. & Irving, E. 2014: Individual personality traits
influence group exploration in a feral guppy population.
Behav. Ecol. 25,95—101.
Brown, C., Jones, F. & Braithwaite, V. 2005: In situ exami-
nation of boldness–shyness traits in the tropical poeciliid,
Brachyraphis episcopi. Anim. Behav. 70,1003—1009.
Budaev, S. V. 1997: “Personality” in the guppy (Poecilia
reticulata): a correlational study of exploratory behavior
and social tendency. J. Comp. Psychol. 111, 399—411.
Burnham, K. P. & Anderson, D. R. 2002: Model Selection
and Multimodel Inference: A Practical Information-The-
oretic Approach. Springer, New York.
Candolin, U. 2004: Effects of algae cover on egg acquisition
in male three-spined stickleback. Behaviour 141, 1389
—1399.
Caramaschi, D., Carere, C., Sgoifo, A. & Koolhaas, J. M.
2013: Neuroendocrine and automatic correlates of ani-
mal personalities. In: Animal Personalities: Behavior,
Physiology, and Evolution, 1st edn (Carere, C. &
Maestripieri, D., eds). The Univ. of Chicago Press, Chi-
cago and London, pp. 353—377.
Carter, A. J., Marshall, H. H., Heinsohn, R. & Cowlishaw,
G. 2012: How not to measure boldness: novel object
and antipredator responses are not the same in wild
baboons. Anim. Behav. 84, 603—609.
Coleman, K. & Wilson, D. S. 1998: Shyness and boldness
in pumpkinseed sunfish: individual differences are con-
text-specific. Anim. Behav. 56, 927—936.
Conrad, J. L., Weinersmith, K. L., Brodin, T., Saltz, J. B. &
Sih, A. 2011: Behavioural syndromes in fishes: a review
with implications for ecology and fisheries manage-
ment. J. Fish Biol. 78, 395—435.
David, M., Salignon, M. & Perrot-Minnot, M. J. 2014:
Shaping the antipredator strategy: flexibility, consis-
tency, and behavioral correlations under varying preda-
tion threat. Behav. Ecol. 25, 1148—1156.
Dingemanse, N. J., Both, C., Drent, P. J. & Tinbergen, J.
M. 2005: Pairs of extreme avian personalities have high-
est reproductive success. J. Anim. Ecol. 74, 667—674.
Dochtermann, N. A. 2010: Behavioral syndromes: carry-
over effects, false discovery rates, and a priori hypothe-
ses. Behav. Ecol. 21, 437—439.
Dugatkin, L. A. & Godin, J. G. J. 1992: Prey approaching
predators: a cost-benefit perspective. Ann. Zool. Fennici.
29, 233—252.
Fabre, N., Garc
ıa-Galea, E. & Vinyoles, D. 2014: Boldness
is related to the development of the cephalic crest in the
male of the river blenny Salaria fluviatilis (Asso, 1801).
Curr. Zool. 60, 373—380.
Fraser, D. F., Gilliam, J. F., Daley, M. J., Le, A. N. &
Skalski, G. T. 2001: Explaining leptokurtic movement
distributions: intrapopulation variation in boldness and
exploration. Am. Nat. 158, 124—135.
Frommen, J. G., Mehlis, M. & Bakker, T. C. M. 2009:
Predator-inspection behaviour in female three-spined
sticklebacks Gasterosteus aculeatus is associated with sta-
tus of gravidity. J. Fish Biol. 75, 2143—2153.
Frost, A. J., Winrow-Giffen, A., Ashley, P. J. & Sneddon, L.
U. 2007: Plasticity in animal personality traits: does
prior experience alter the degree of boldness? Proc. Biol.
Sci. 274, 333.
Galhardo, L., Vitorino, A. & Oliveira, R. F. 2012: Social
familiarity modulates personality trait in a cichlid fish.
Biol. Lett. 8, 936—938.
Garamszegi, L., Mark
o, G. & Herczeg, G. 2012: A meta-
analysis of correlated behaviours with implications for
behavioural syndromes: mean effect size, publication
bias, phylogenetic effects and the role of mediator vari-
ables. Evol. Ecol. 26, 1213—1235.
Godin, J. G. J. & Dugatkin, L. A. 1996: Female mating pref-
erence for bold males in the guppy, Poecilia reticulata.
Proc. Natl Acad. Sci. USA 93, 10262—10267.
Godin, J. G. J. & McDonough, H. E. 2003: Predator prefer-
ence for brightly colored males in the guppy: a viability
cost for a sexually selected trait. Behav. Ecol. 14, 194—
200.
Herdman, E. J. E., Kelly, C. D. & Godin, J. J. 2004: Male
mate choice in the guppy (Poecilia reticulata): do males
prefer larger females as mates? Ethology 110,
97—111.
Houde, A. E. 1997: Sex, Color, and Mate Choice in Gup-
pies. Princeton, Univ. Press, Princeton, New Jersey.
Huntingford, F. A. 2004: Implications of domestication
and rearing conditions for the behaviour of cultivated
fishes. J. Fish Biol. 65, 122—142.
Jirotkul, M. 2000: Male trait distribution determined alter-
native mating tactics in guppies. J. Fish Biol. 56, 1427—
1434.
Karino, K. & Kobayashi, M. 2005: Male alternative mating
behaviour depending on tail length of the guppy, Poe-
cilia reticulata. Behaviour 142, 191—202.
Kiritome, A., Sato, A. & Karino, K. 2012: Influences of
orange spot patterns and operational sex ratio on male
mating behaviors in the guppy Poecilia reticulata.
Ichthyol. Res. 59, 304—313.
Kokko, H. & Rankin, D. J. 2006: Lonely hearts or sex in
the city? Density-dependent effects in mating systems.
Philos. Trans. R. Soc. Lond. B Biol. Sci. 361, 319—334.
Lessells, C. M. & Boag, P. T. 1987: Unrepeatable repeatabil-
ities: a common mistake. Auk 104, 116—121.
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH 9
R.

Re
zucha & M. Reichard Personalities in Endler’s Guppy Males

Locatello, L., Rasotto, M. B., Evans, J. P. & Pilastro, A.
2006: Colourful male guppies produce faster and more
viable sperm. J. Evol. Biol. 19, 1595—1602.
Logue, D. M., Mishra, S., McCaffrey, D., Ball, D. & Cade,
W. H. 2009: A behavioral syndrome linking courtship
behavior toward males and females predicts reproduc-
tive success from a single mating in the hissing cock-
roach, Gromphadorhina portentosa. Behav. Ecol. 20, 781
—788.
Magurran, A. E. & Magellan, K. 2007: Behavioural pro-
files: individual consistency in male mating behaviour
under varying sex ratios. Anim. Behav. 74, 1545—
1550.
Mathot, K. J., Wright, J., Kempenaers, B. & Dingemanse,
N. J. 2012: Adaptive strategies for managing uncertainty
may explain personality-related differences in beha-
vioural plasticity. Oikos 121, 1009—1020.
Mazu
e, G. P. F., Dechaume-Moncharmont, F. X. & Godin,
J. G. J. 2015: Boldness-exploration behavioral syn-
drome: interfamily variability and repeatability of per-
sonality traits in the young of the convict cichlid
(Amatitlania siquia). Behav. Ecol. 26, 900—908.
McEvoy, J., While, G. M., Sinn, D. L., Carver, S. & Wap-
stra, E. 2015: Behavioural syndromes and structural
and temporal consistency of behavioural traits in a
social lizard. J. Zool. 296,58—66.
Metcalfe, N. B., Valdimarsson, S. K. & Morgan, I. J. 2003:
The relative roles of domestication, rearing environ-
ment, prior residence and body size in deciding territo-
rial contests between hatchery and wild juvenile
salmon. J. Appl. Ecol. 40, 535—544.
Modlmeier, A. P., Laskowski, K. L., DeMarco, A. E., Cole-
man, A., Zhao, K., Brittingham, H. A., Mcdermott, D. R.
& Pruitt, J. N. 2014: Persistent social interactions beget
more pronounced personalities in a desert-dwelling
social spider. Biol. Lett. 10, 20140419.
Mutzel, A., Dingemanse, N. J., Araya-Ajoy, Y. G. & Kem-
penaers, B. 2013: Parental provisioning behaviour plays
a key role in linking personality with reproductive suc-
cess. Proc. Biol. Sci. 280, 20131019.
Oliveira, R. F., Taborsky, M. & Brockmann, H. J. 2008:
Alternative Reproductive Tactics: An Integrative
Approach. Cambridge Univ. Press, New York.
Parker, G. A. 1974: Courtship persistence and female-
guarding as male time investment strategies. Behaviour
48, 157—184.
Pilastro, A. & Bisazza, A. 1999: Insemination efficiency of
two alternative male mating tactics in the guppy Poecilia
reticulata. Proc. Biol. Sci. 266, 1887—1891.
Piyapong, C., Krause, J., Chapman, B. B., Ramnarine, I.
W., Louca, V. & Croft, D. P. 2010: Sex matters: a social
context to boldness in guppies (Poecilia reticulata).
Behav. Ecol. 21,3—8.
Poeser, F. N., Kempkes, M. & Isbr€
ucker, I. J. H. 2005:
Description of Poecilia (Acanthophacelus)wingei n. sp. from
the Par

ıa Peninsula, Venezuela, including notes on Acan-
thophacelus Eigenmann, 1907 and other subgenera of
Poecilia Bloch and Schneider, 1801 (Teleostei, Cyprin-
odontiformes, Poeciliidae). Contrib. Zool. 74,97—115.
R Development Core Team 2009: R: A language and envir-
onment for statistical computing. R Foundation for Sta-
tistical Computing, Vienna, Austria.
Reaney, L. T. & Backwell, P. R. Y. 2007: Risk-taking
behavior predicts aggression and mating success in a fid-
dler crab. Behav. Ecol. 18, 521—525.
Reynolds, J. D., Gross, M. R. & Coombs, M. J. 1993: Envi-
ronmental conditions and male morphology determine
alternative mating behaviour in Trinidadian guppies.
Anim. Behav. 45, 145—152.

Re
zucha, R. & Reichard, M. 2014: The effect of social
environment on alternative mating tactics in male
Endler’s guppy, Poecilia wingei. Anim. Behav. 88,
195—202.

Re
zucha, R. & Reichard, M. 2015: Strategic exploitation of
fluctuating asymmetry in male Endler’s guppy courtship
displays is modulated by social environment. J. Evol.
Biol. 28, 356—367.
Sargent, R. C. 1985: Territoriality and reproductive trade-
offs in the threespine stickleback, Gasterosteus aculeatus.
Behaviour 93, 217—226.
Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V.,
Longair, M., Pietzsch, T., Preibisch, S., Rueden, C., Saal-
feld, S., Schmid, B., Tinevez, J. Y., White, D. J., Harten-
stein, V., Eliceiri, K., Tomancak, P. & Cardona, A. 2012:
Fiji: an open-source platform for biological-image analy-
sis. Nat. Methods 9, 676—682.
Schuett, W., Tregenza, T. & Dall, S. R. X. 2010: Sexual
selection and animal personality. Biol. Rev. 85, 217—
246.
Schuett, W., Godin, J. G. J. & Dall, S. R. X. 2011: Do female
zebra finches, Taeniopygia guttata, choose their mates
based on their “Personality”? Ethology 117,908—917.
Shuster, S. M. & Wade, M. J. 1991: Equal mating success
among male reproductive strategies in a marine isopod.
Nature 350, 608—610.
Sih, A. 2013: Frontiers on the interface between behav-
ioral syndromes and social behavioral ecology. In: Ani-
mal Personalities: Behavior, Physiology, and evolution,
1st edn (Carere, C. & Maestripieri, D., eds). Univ. of Chi-
cago Press, Chicago and London, pp. 221—251.
Sih, A. & Watters, J. V. 2005: The mix matters: beha-
vioural types and group dynamics in water striders.
Behaviour 142, 1417—1431.
Sih, A., Johnson, J. C. & Bell, A. 2004: Behavioral syn-
dromes: an ecological and evolutionary overview.
Trends Ecol. Evol. 19, 372—378.
Sih, A., Chang, A. T. & Wey, T. W. 2014: Effects of beha-
vioural type, social skill and the social environment on
male mating success in water striders. Anim. Behav. 94,
9—17.
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH10
Personalities in Endler’s Guppy Males R.

Re
zucha & M. Reichard

Sinclair, E. L. E., de Souza, C. R. N., Ward, A. J. W. & See-
bacher, F. 2014: Exercise changes behaviour. Funct.
Ecol. 28, 652—659.
Smith, B. R. & Blumstein, D. T. 2010: Behavioral types as
predictors of survival in Trinidadian guppies (Poecilia
reticulata). Behav. Ecol. 21, 919—926.
Smith, B. R. & Blumstein, D. T. 2012: Structural consis-
tency of behavioural syndromes: does predator training
lead to multi-contextual behavioural change? Beha-
viour 149, 187—213.
Smith, C., Phillips, A. & Reichard, M. 2015: Cognitive abil-
ity is heritable and predicts the success of an alternative
mating tactic. Proc. Biol. Sci. 282, 20151046.
Stein, L. R. & Bell, A. M. 2012: Consistent individual dif-
ferences in fathering in threespined stickleback Gasteros-
teus aculeatus. Curr. Zool. 58,45—52.
Taborsky, M. 1994: Sneakers, satellites, and helpers: para-
sitic and cooperative behavior in fish reproduction. Adv.
Study Behav. 23,1—100.
Thomson, J. S., Watts, P. C., Pottinger, T. G. & Sneddon, L.
U. 2012: Plasticity of boldness in rainbow trout, Oncor-
hynchus mykiss: do hunger and predation influence risk-
taking behaviour? Horm. Behav. 61, 750—757.
Wilson, A. D. M. & Godin, J. G. J. 2009: Boldness and
behavioral syndromes in the bluegill sunfish, Lepomis
macrochirus. Behav. Ecol. 20, 231—237.
Wolf, M. & Weissing, F. J. 2010: An explanatory frame-
work for adaptive personality differences. Philos. Trans.
R. Soc. Lond. B Biol. Sci. 365, 3959—3968.
Zuur, A. F., Ieno, E. N., Walker, N. J., Saveliev, A. A. &
Smith, G. M. 2009: Mixed Effects Models and Exten-
sions in Ecology with R. Springer, New York.
Appendix 1: Set of all candidate models, their degrees of freedom (df), their relative fits as compared to the best fitting model (DAIC) and their
relative eight (w) for each personality trait. The best fitting models are indicated in bold. In cases where a model other than the null model was the
best fitting, other models with DAIC <2.0 (used in the model averaging) are also indicated in bold.
Model parameters df
Emergence from refuge Boldness towards predator Activity levels
Young males Older males Young males Older males Young males Older males
DAIC wDAIC wDAIC wDAIC wDAIC wDAIC w
Null model (intercept-only) 2 0.00 0.13 0.00 0.12 0.58 0.08 12.54 0.00 2.56 0.06 0.00 0.11
Orange % 3 2.11 0.04 1.41 0.06 2.23 0.03 1.87 0.09 4.55 0.02 1.39 0.06
Orange % +Total length 4 4.26 0.02 3.03 0.03 2.76 0.03 4.20 0.03 5.90 0.01 3.38 0.02
Total length 3 2.05 0.05 1.58 0.05 1.05 0.06 14.78 0.00 3.80 0.03 1.81 0.05
Orange % +Black % +Total length 5 4.01 0.02 5.21 0.01 5.07 0.01 5.88 0.01 8.29 0.00 6.02 0.01
General interest 3 2.03 0.05 1.86 0.05 2.13 0.03 12.00 0.00 4.77 0.02 0.65 0.08
Sigmoid displays 3 1.10 0.07 1.06 0.07 1.32 0.05 3.00 0.05 4.33 0.03 2.36 0.04
Gonopodial thrusts 3 1.21 0.07 2.21 0.04 2.70 0.03 14.67 0.00 0.00 0.22 2.22 0.04
Gonopodial swings 3 0.94 0.08 1.28 0.06 2.00 0.04 12.40 0.00 3.71 0.04 2.15 0.04
Gonopore nibs 3 1.72 0.05 0.83 0.08 0.75 0.07 14.64 0.00 2.61 0.06 2.35 0.04
General interest +Total length 4 4.26 0.02 3.75 0.02 3.04 0.02 14.37 0.00 5.87 0.01 2.65 0.03
General interest +Orange % 4 4.25 0.02 3.30 0.02 3.80 0.02 4.05 0.03 6.84 0.01 2.83 0.03
General interest +Orange % +Total length 5 6.58 0.00 5.25 0.01 4.80 0.01 6.54 0.01 8.04 0.00 4.99 0.01
Sigmoid displays 4 3.07 0.03 2.42 0.03 1.29 0.05 5.35 0.02 5.44 0.01 4.32 0.01
Sigmoid displays +Orange % +Total length 5 5.37 0.01 3.94 0.02 2.98 0.02 2.40 0.07 7.64 0.00 5.88 0.01
Sigmoid displays +Orange % 4 3.30 0.02 2.56 0.03 3.00 0.02 0.00 0.22 6.41 0.01 3.74 0.02
Gonopodial thrusts +Total length 4 3.30 0.02 3.90 0.02 3.26 0.02 17.03 0.00 1.50 0.11 4.03 0.02
Gonopodial thrusts +Orange % 4 3.29 0.02 3.66 0.02 4.37 0.01 3.82 0.03 2.28 0.07 3.79 0.02
Gonopodial thrusts +Total length +Orange % 5 5.49 0.01 5.39 0.01 5.01 0.01 6.33 0.01 3.88 0.03 5.81 0.01
Gonopodium length 3 1.85 0.05 1.93 0.04 0.00 0.10 14.76 0.00 4.54 0.02 2.24 0.04
Gonopodial thrusts +Gonopodium length 4 3.23 0.03 4.22 0.01 2.18 0.03 17.02 0.00 2.21 0.07 4.64 0.01
Gonopodial thrusts +Gonopodium length +
Total length
5 5.55 0.01 6.21 0.01 2.96 0.02 19.53 0.00 3.49 0.04 6.38 0.00
Gonopodium length +Total length 4 4.11 0.02 3.81 0.02 0.66 0.07 17.12 0.00 5.40 0.02 3.95 0.02
Black % 3 1.26 0.07 1.64 0.05 2.62 0.03 10.50 0.00 4.40 0.02 2.11 0.04
Black % +Total length 4 3.18 0.03 2.99 0.03 2.98 0.02 12.86 0.00 5.92 0.01 4.10 0.01
Iridescence 3 2.24 0.04 1.04 0.07 1.21 0.06 11.95 0.00 4.58 0.02 2.33 0.04
Iridescence +Total length 4 4.37 0.01 3.00 0.03 2.28 0.03 14.20 0.00 5.76 0.01 4.32 0.01
Orange % +Black % +Iridescence +
Total length
6 6.47 0.00 7.49 0.00 6.92 0.00 8.47 0.00 10.62 0.00 8.62 0.00
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH 11
R.

Re
zucha & M. Reichard Personalities in Endler’s Guppy Males

Appendix 1: (Continued)
Model parameters df
Emergence from refuge Boldness towards predator Activity levels
Young males Older males Young males Older males Young males Older males
DAIC wDAIC wDAIC wDAIC wDAIC wDAIC w
Orange % +Black % +Total length +
Sigmoid displays
6 5.56 0.01 6.33 0.00 5.39 0.01 5.03 0.02 10.07 0.00 8.62 0.00
Orange % +Black % +Total length +
Gonopodial thrusts
6 5.09 0.01 7.68 0.00 7.41 0.00 8.11 0.00 6.26 0.01 8.64 0.00
Orange % +Black % +Total length +
Sigmoid displays +Gonopodium length
7 8.09 0.00 8.84 0.00 4.81 0.01 7.87 0.00 12.41 0.00 11.53 0.00
Orange % +Black % +Total length +
Gonopodial thrusts +Gonopodium length
7 7.72 0.00 10.86 0.00 17.57 0.00 10.45 0.00 8.62 0.00 11.47 0.00
Supporting Information
Additional supporting information may be found in the online version of this article:
Table S1. Model parameters, their standard errors (SE), z-values and their statistical significance (p) for the
set of best fitting models (deltaAIC <2).
Ethology 122 (2016) 1–12 ©2016 Blackwell Verlag GmbH12
Personalities in Endler’s Guppy Males R.

Re
zucha & M. Reichard