Content uploaded by Leonardo Castilho
Author content
All content in this area was uploaded by Leonardo Castilho on May 12, 2021
Content may be subject to copyright.
Individual preference functions exist without overall preference in a
tropical jumping spider
Leonardo B. Castilho
a
,
*
, Regina H. Macedo
a
, Maydianne C. B. Andrade
b
a
Departamento de Zoologia, Universidade de Brasília, DF, Brazil
b
Departments of Biological Sciences and Ecology &Evolutionary Biology, University of Toronto at Scarborough, Toronto, ON, Canada
article info
Article history:
Received 15 May 2019
Initial acceptance 1 July 2019
Final acceptance 3 October 2019
MS. number: A19-00341R
Keywords:
mate choice
personality
sexual selection
Female mate choice is a widespread and well-recognized phenomenon. Nevertheless, individual varia-
tion in female preference has not yet received the same attention, although such preferences can have
important effects on evolutionary dynamics. Here we assess and compare population- and individual-
level female preferences for male ornaments and size in the tropical jumping spider Hasarius adansoni
in two sets of laboratory experiments. First, we paired females with a single male and quantified
receptive behaviours (e.g. receptive posture, number of copulations) and unreceptive behaviours (e.g.
attacking the male, running away from a male). We assessed whether these male traits were related to
offspring quality and quantity to determine whether there was selection on female preferences. Then, we
paired different females with three different-sized males, one per day, and scored similar behaviours to
measure preference functions (relationship between male traits and female receptivity). Generally, the
population of females did not show a consistent average preference for male traits, despite our finding
that females mated to larger males produced more offspring. However, at the individual level, females
showed different preference functions for male size, such that some females preferred larger males,
while others preferred smaller males. We discuss these data in terms of the causes and consequences of
individual preference functions, highlighting the importance of including individual preference functions
in future studies that focus on sexual selection and how individual preference can maintain phenotypic
variation in wild populations.
©2019 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
Sexual selection models predict that animals should not mate
randomly (Andersson, 1994). As the sex that invests more heavily in
each reproductive event, females are usually ‘choosier’and, ac-
cording to a number of theoretical models, are expected to mate
nonrandomly with males in ways that will enhance their repro-
ductive output or the fitness of their offspring (Andersson, 1994;
Bateman, 1948). The evidence supporting nonrandom mating by
females is remarkable and encompasses virtually all main taxo-
nomic groups (Andersson, 1982; Candolin, 2003; Cotton, Small, &
Pomiankowski, 2006; Kirkpatrick, 1982; Kokko, Jennions, &
Brooks, 2006; Møller &Alatalo, 1999; Ronald, Fern
andez-Juricic,
&Lucas, 2012; von Schantz, Bensch, Grahn, Hasselquist, &
Wittzell, 1999). Much of this work examines average preferences
that are expected to impose consistent selection on male traits,
although these population-level patterns could be coincident with
significant variation at the level of individuals (Wagner, 1998).
Individual sexual preferences are a less frequently explored
facet of sexual selection. The idea that individuals consistently
differ in the way they perform specific behaviours is now well
recognized, and recent research has shown that such differences
may be adaptive (Bell, 2007; Dall, Houston, &McNamara, 2004;
Dingemanse &R
eale, 2005; R
eale, Reader, Sol, McDougall, &
Dingemanse, 2007; Sih, Bell, &Johnson, 2004; Sih, Bell, Johnson,
&Ziemba, 2004; Stamps &Groothuis, 2010; Wolf &Wessing,
2010). However, distinct behavioural types (i.e. consistent individ-
ual differences in behaviour) and different behavioural syndromes
(i.e. suites of correlated behaviours across situations; Bell, 2007; Sih
&Johnson et al., 2004), most recently called animal ‘personalities’
(Sih &Johnson et al., 2004), have been assessed in a relatively
limited number of behaviours. Aggression, environment explora-
tion and fear response to novel or threatening situations (usually
called boldness) are among the most common behaviours assessed
in the personalities paradigm (Carter, Goldizen, &Tromp, 2010;
Castilho &Macedo, 2016; Dingemanse, Both, van Noordwijk,
*Correspondence: L. Braga Castilho, Departamento de Zoologia, IB, Universidade
de Brasilia, 70910-900, Brasília, DF, Brazil.
E-mail address: leonardobcastilho@gmail.com (L. B. Castilho).
Contents lists available at ScienceDirect
Animal Behaviour
journal homepage: www.elsevier.com/locate/anbehav
https://doi.org/10.1016/j.anbehav.2019.11.016
0003-3472/©2019 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
Animal Behaviour 160 (2020) 43e51
Rutten, &Drent, 2003; Møller, 2010; Rabaneda-Bueno, Aguado,
Fern
andez-Montraveta, &Moya-Lara~
no, 2014; Smith &Blumstein,
2010; Zhao, Hu, Liu, Chen, &Sun, 2016). The influence of consistent
differences in behaviour in a sexual selection context, however, has
been relatively neglected (Schuett, Tregenza, &Dall, 2010). To date,
there is some evidence that females may vary in their propensity to
engage in copulations (Godin &Dugatkin, 1995; Keiser,
Lichtenstein, Wright, Chism, &Pruitt, 2018; Kralj-Fi
ser, Schneider,
&Kuntner, 2013), or extrapair copulations (Forstmeier, 2007) and
in the direction of mate choice (Ah-King &Gowaty, 2016; Edwards,
Melville, Joseph, &Keogh, 2015; Forstmeier &Birkhead, 2004; also
see ; Jennions &Petrie, 1997). For instance, in the zebra finch,
Taeniopygia guttata, some females prefer high song rates while
others prefer low song rates, although song rate is related to
average male attractiveness (Forstmeier &Birkhead, 2004). In other
cases, such as brown-headed cowbirds, Molothrus ater, variation in
female visual and auditory systems affects perception and links to
individual preferences for multimodal male display traits (Ronald
et al., 2017, 2018). The pattern of female receptivity as a function
of male phenotype defines a preference function that indicates the
rank order of preference for males with particular trait values
(Jennions &Petrie, 1997). The causes of distinct mating preference
functions may vary, but female past experience (e.g. previous sex-
ual encounters) and assortative mating are among the most
commonly reported (e.g. Baldauf, Engqvist, Ottenheym, Bakker, &
Thünken, 2013; Bel-Venner, Dray, Allain
e, Menu, &Venner, 2008;
Fowler-Finn &Rodríguez, 2011; Hebets, 2003;Hoefler, Persons, &
Rypstra, 2008;Johnson &Basolo, 2003; Keiser et al., 2018; Kralj-
Fi
ser et al., 2013).
Given the possibility that animals can vary intrinsically in their
sexual preferences, evaluating the parameters that reflect
population-level preference functions may not accurately reflect
the detailed processes of sexual selection that act within a species.
A comprehensive understanding of sexual selection can only be
achieved with the assessment of the full spectrum of sexual pref-
erences, encompassing both population and individual levels, and
considering possible individual variations that might influence
mate choice and preference functions (Ah-King &Gowaty, 2016;
Wagner, 1998).
Here, we explored the scope of sexual selection at population
and individual levels in the tropical jumping spider Hasarius
adansoni. Jumping spiders are excellent models to assess many
aspects of behaviour, including sexual behaviour, since most spe-
cies perform complex displays (Elias, Maddison, Peckmezian,
Girard, &Mason, 2012), which usually imply mate choice (Busso
&Rabosky, 2016; Girard, Elias, &Kasumovic, 2015), and are easy
to capture and maintain in controlled conditions. Hasarius adansoni
males have courtship displays with visual and vibrational signals,
and females exhibit a stereotyped receptive posture when accept-
ing a male for copulation (Castilho, Andrade, &Macedo, 2018). In
the present laboratory experiment, we first paired each female with
a randomly assigned, single male to test whether there was a
population-level preference function for females based on male
size or ornament dimensions. Then, we compared offspring quan-
tity and quality as a function of paternal traits to determine
whether there were any benefits for choosiness (i.e. if different
males would convey different fitness benefits for the female).
Finally, we paired females with multiple males in sequence to
measure individual female preference functions (Wagner, 1998).
We also assessed possible causes for such individual variation, by
testing whether female size or mating experience influenced fe-
males’sexual preferences.
Since our study species performs sexual displays, we predicted
that females would exhibit a population-level preference, being
more receptive to males siring more and/or better offspring than
others (e.g. larger males). Given that individual variation is com-
mon in animals, and individual variation in preference functions
can coexist with a population-level preference (e.g. high versus low
song rate in zebra finches; Forstmeier &Birkhead, 2004), we also
predicted that individual females would differ in their preference
functions.
METHODS
Morphology and Mating Behaviour
Hasarius adansoni (Salticidae: Araneae) is a common tropical
spider with sexually dimorphic coloration. Females are black and
brown, while males are primarily black with bright white patches
on their palps (paired, anterior-most appendages, which are also
the copulatory organs in male spiders; see Figure 1 in Castilho et al.,
2018). The white patch is revealed to females during courtship and
may be used as a visual signal (Castilho et al., 2018), as is typical for
sexual ornaments used in courtship displays in jumping spiders
(Foelix, 2011). Males also produce vibrational signals during
courtship, classified as tremulation, a type of substrate-borne signal
in which the male vibrates his abdomen without touching the
substrate. The vibration energy is transmitted through his legs to
the substrate and the receiver (Castilho et al., 2018). Male
H. adansoni usually court in bouts, that is, a single couple may go
through the whole process (i.e. courtshipecopulation/rejection)
many times in a few hours (Castilho et al., 2018).
Receptive females respond to the male's courtship by continu-
ally orienting towards the male as he courts, and eventually
adopting a receptive posture (i.e. staying motionless with legs
curled and vibrating the abdomen), which is required for mating to
proceed (Castilho et al., 2018). Nonreceptive females may attack the
male, rapidly retreat from the male, or simply ignore his copulation
attempt by not adopting a receptive posture (Castilho et al., 2018).
The former behaviours were classified as receptive behaviours,
while the latter were classified as nonreceptive behaviours.
General Capture and Handling
A total of 161 animals (71 females and 90 males) were captured
in the urban environment around the city of Brasília, in central
Brazil (15
46
0
47
00
S, 47
55
0
47
00
W) and brought to the Laborat
orio de
Comportamento Animal at the Universidade de Brasília main
campus. Spiders were maintained in glass vials (9 4.5 cm) with a
piece of wet cotton and were fed with 15 adult Drosophila spp. and
one Gryllus sp. cricket nymph every 4e7 days. Males and females
Figure 1. Paired copulatory organs (secondary sexual organs, or palps) of male
Hasarius adansoni after dissection from the body to allow measurement. We measured
the area of the whole palp (green outline) and the area of the white patch (potential
ornament, purple outline).
L. B. Castilho et al. / Animal Behaviour 160 (2020) 43e5144
captured as juveniles were reared until adulthood and experiments
were conducted only with adult, sexually mature animals. We were
able to distinguish between adult and juvenile females by the
presence of a visible epigynum (adult female's external genital
structure), assessed under a stereomicroscope. Adult males were
identified by their fully developed palps. Animals captured as
adults were used only in a subset of experiments (see below).
A pilot study showed that after a few copulations, females
become globally less willing to copulate, even with males they had
previously chosen. In the same pilot study, we found that males
were still willing to copulate after many copulations. Given these
pilot results, we reared virgin females for use in general preference
mating trials in which females are paired with only one male (see
below). However, since individual preference measures required
assessing the difference in responses to multiple males from a
single female, without actually mating (see below), and to avoid a
very small sample size for these experiments, we allowed
nonvirgin females to enter individual preference experiments. Our
focus here was to examine individual females' responses to distinct
males, while maintaining each female's willingness to copulate
constant. This was achieved with our procedure, while allowing us
to keep the sample size large enough to produce robust models. It is
also possible that after the first copulation, females change their
preference. We do not think this is the case, since we did not find
any consistent difference in preference functions for females
captured as adults versus those captured as young (although our
design did not allow for formal tests of this hypothesis). Even if this
were the case, it would not diminish the importance of a possible
individual preference function variation and could be interpreted
as the cause of such variation.
Permits for collection and maintenance of animals were issued
by Instituto Chico Mendes (ICMBio), Brazil. Animals were handled
as carefully as possible, and at the end of the project, all remaining
animals were released in the same city where they were initially
collected. All protocols described adhered to the legal requirements
of Brazil and Canada.
Measuring Spider Phenotype
Before mating trials, we chilled all spiders inside a vial that was
in contact with ice for a few minutes, until the spiders stopped
moving. We took their weight with a 0.001 g precision balance and
also photographed them with a ruler in the image to use as a scale
to later determine absolute body size (i.e. carapace width) using the
software ImageJ (Schneider, Rasband, &Eliceiri, 2012). Photographs
were taken with a digital USB 2.0 stereomicroscope (manufactured
by DigiMicro). Individual condition was also estimated using the
residuals of a regression of animal mass on animal size. In spiders,
well-fed animals are usually in better condition (i.e. larger and
heavier), and this affects their ability to court and compete for
mates, and might affect female choice (e.g. Cotton et al., 2006; Elgar
&Fahey, 1996; Hoefler et al., 2008; Kasumovic &Andrade, 2006;
Kasumovic, Brooks, &Andrade, 2009). Moreover, female condi-
tion (i.e. size, infection status, energy available, etc.) might affect
choice (Cotton et al., 2006). Although the condition index we used
has been criticized (Jordi &Andy, 2009), it has also been endorsed
(Schulte-Hostedde et al., 2005), and has been considered the best
condition index for spiders (Jakob, Marshall, &Uetz, 1996).
After the males’natural deaths, we photographed their palps to
measure the percentage of the palp area covered by the white patch
as an estimate of ornament elaboration (Fig. 1). Palps were ampu-
tated at the base to facilitate measurement, and the mean of both
palps was used as our index except in a few cases where one palp
was damaged by the procedure.
EXPERIMENT 1: GENERAL FEMALE PREFERENCES AND
OFFSPRING QUALITY
Mating trials took place in a mating arena, which consisted of an
acrylic square box, measuring 13 13 4 cm. The arena had two
opaque dividers in opposite corners, so two animals could be held
at the same time without visual contact until their release. Since
our arena was made from a rigid acrylic material, any substrate-
borne vibrations made by the spiders were likely to have been
significantly attenuated while in the arena (e.g. see Elias, Mason, &
Hoy, 2004). We are aware that using a softer substrate would have
allowed females to better assess males’vibration signals, which
could modify the shape of preference functions. Most of our spi-
ders, however, were captured on the walls of buildings, so we
believe our experimental set-up closely matches the conditions
they encounter naturally.
In each trial, animals were placed inside the arena and separated
by the dividers for 1 h (acclimation), after which the dividers were
removed and the animals were allowed to see and interact with
each other for 3 h. We videotaped all trials and, from the videos,
collected the following variables to measure female receptivity: (1)
number of copulations; (2) total copulation time across all suc-
cessful copulations; (3) number of female unreceptive behaviours
(i.e. attacking the male, running away from the male or not
adopting receptive posture when in front of a displaying male); and
(4) percentage of copulation attempts by the male that were un-
successful. One copulation attempt was defined as a courting male
getting close to a female and touching her (i.e. trying to engage in
palp insertion). These variables were chosen based on our obser-
vations indicating that, in H. adansoni, no copulations occur if the
male does not court and females control the total number of cop-
ulations (Castilho et al., 2018).
Before each trial, the arena was cleaned with soap, water and
alcohol to remove any pheromonal cues that could have been left
behind by previous animals. Since sexual ornaments of some
jumping spiders reflect light in the ultraviolet spectrum (e.g.
Bulbert, Hanlon, Zappettini, Zhang, &Li, 2015), we conducted ex-
periments under a natural light-simulating lamp (Arcadia Bird
lamp, Model FB 36). Trials in which it was clear that animals did not
see each other (N¼12) were excluded from analyses. It was easy to
detect the moment when animals saw each other, since they usu-
ally turned their median eyes to one another quickly to engage in
eye contact, and visually followed each other using cephalothoracic
movements following the initial contact.
After mating trials, females were kept in the same housing
conditions as described above until they produced eggs. The
number of eggsacs per female and the number of hatched offspring
per eggsac were counted. After spiderlings emerged from the
eggsac, they were kept individually in small glass vials
(~5 1.5 cm). Half of each brood was left unfed, providing us with
data to conduct survival analyses to assay juvenile provisioning and
fitness under stress (Cox proportional hazard model, described
below), and the other half was used in a feeding performance trial.
Since newly dispersed jumping spiders usually rely on their own
predation ability to survive (Richman &Jackson, 1992), we
measured feeding performance as an offspring quality indicator.
We created a protocol to measure feeding performance in young
jumping spiders, using springtails (Collembola) as the model prey.
Feeding performance trials consisted of placing individual spider-
lings in a petri dish with a single live springtail. We replaced the
springtail if it died before being captured by the spiderling. Since
static prey do not attract spiders, we kept the springtails moving by
touching them with a paintbrush every time they were immobile.
We kept the spiders within the petri dish by gently pushing them
back inside with a paintbrush when they tried to escape. Such
L. B. Castilho et al. / Animal Behaviour 160 (2020) 43e51 45
protocol usually ended with the springtail being captured and
eaten by the spider. One caveat to this methodology is that the
paintbrush could alter the behaviour of the spider or the springtail.
However, this seems unlikely for two reasons: (1) we did not
observe any change in spider behaviour when the paintbrush was
touching the springtail, in contrast to when the spider was stalking
the springtail; and (2) although this method was used throughout
the experiment, we found a decrease in predation performance in
later broods (see Results), showing biological correlations were
measurable using this technique.
Each feeding trial was videotaped, and the following variables
were extracted from the videos as measurements of feeding per-
formance: (1) latency to start moving towards prey, once oriented
towards it; and (2) speed while approaching prey (in mm/s). Our
springtail populations were obtained from an independent seller in
Toronto, Canada. Springtails were kept in consistent housing con-
ditions throughout our experiments and bred successfully under
these conditions. They were held at room temperature in two small
boxes with soil, and provisioned with water and food (live yeast)
every other day.
Experiment 2: Individual Preference
We used a subsample of 23 adult males to establish the mean
and standard errors of body size in the population. The same person
measured all males, with repeatability of these measurements
assessed on a subsample of males that were measured twice
(R¼0.92, F
1,18
¼24.3, P<0.0001). Once all 23 animals were
measured, we were able to classify any male in one of three body
size categories: average males (between -1SE and þ1SE of the
average); small males (<1SE); large males (>þ1SE). Before each
trial, we also measured the females.
We used standard errors as cutoff points to separate male sizes
because this resulted in a more or less equivalent number of small,
average and large animals for the experimental trial, which was our
main goal with the size classification. On average, males in the large
male group were larger than those in the average group, which in
turn were larger than males in the small group (large:
2.21 ±0.12 mm; average: 2.11 ±0.09 mm; small: 1.89 ±0.14 mm),
and these differences were significant (ANOVA: F
2,46
¼29.2,
P<0.0001). This shows that our classification correctly divided
males into three distinct size groups.
Mating trials were similar to those described in experiment 1,
however, each female was presented with three different-sized
males (small, average and large) in random order and on three
separate days (one male presented per day), to prevent excessive
stress and fatigue from handling. Since one copulation may alter
the chance of a female copulating again (Castilho et al., 2018), ex-
periments were interrupted right before mating took place or after
1 h, if no copulations occurred.
From the videos, we recorded the following female behaviours
as measurements of female receptivity: (1) number of unreceptive
behaviours (i.e. attacking or running awayfrom the male, even if he
did not attempt copulation); (2) number of rejections (i.e. number
of displays performed by the male that did not result in a receptive
posture by the female); (3) percentage of copulation attempts by
the male that were rejected; and (4) occurrence of copulations
(coded as 1 or 0). If the animals tried to copulate, but were sepa-
rated by us, we assigned the number ‘1’. If not, we assigned the
number ‘0’. The occurrence of copulation was easy to observe, since
a courting male would approach a female in receptive posture and
mount her. As soon as she turned her abdomen up to facilitate palp
insertion and the male was about to insert, they were separated.
Trials in which males did not court females at least once were
excluded from further analysis (N¼2/51 trials).
Statistical Analysis
Reducing variables’dimensionality
Principal component analyses (PCA) were used to reduce data
dimensionality. Females’preference variables (e.g. number of at-
tacks towards males, number of copulations, etc.) and offspring
feeding performance variables along with eggsac number (to con-
trol for any early versus late brood effects) were reduced by PCA
analysis (Tables 1e3). The number of principal components we
used in our analyses was based on the percentage of total variance
explained and consideration of the biological significance of the
PCs. As a general rule, we were conservative and used PCs that
together explained at least 75% of the total variance in our data sets.
General female preferences
General linear models were used to create female preference
functions, with variables of female preference (extracted from the
principal components) as response variables and the following
male quality variables as predictors: male mass, size, condition and
percentage of white patch cover on palp.
Young survival
A Cox proportional hazard rate model was used to perform a
survival analysis of the offspring. The model uses brood number as
afixed effect and brood identity nested in female identity as
random factors. Female Cox proportional hazard coefficients were
then regressed on male quality variables to assess the influence of
male quality on survival of each female's offspring.
Young feeding performance
Male quality, measured as male size and condition, was also
inserted in linear models as predictors, with offspring feeding
performance (extracted from the axes of the PCA) as response
variables, to test for any effect of male quality on offspring preda-
tory performance. We also regressed number of offspring produced
against male quality, to assess the influence of male quality on
offspring number.
Note, however, that for experiment 1, not every predictor was
used in every analysis, since this would lead to collinearity prob-
lems in some models (e.g. male size, mass and/or condition were
sometimes correlated). We also used percentage of white patch
cover as a predictor variable, since total patch area was usually
correlated with other variables, especially male size. To summarize,
we used male weight, condition and percentage of white patch
cover as predictors in a model that had female acceptance as
response. The same predictors were used in another model with
female rejection as response (this separation was due to the PCA
results, see below). Another model had offspring feeding perfor-
mance as response, and male size and condition as predictors. We
also regressed number of young produced with male condition, size
and percentage of white patch cover. Offspring survival was
Table 1
Component loadings from a PCA with different variables used to measure female
Hasarius adnasoni preference for males
Variable PC1 PC2 PC3 PC4
Number of copulations 0.92 0.32 0.16 0.18
Total copulation time 0.91 0.36 0.06 0.18
Unreceptive behaviours 0.32 0.84 0.44 0.02
% Rejections 0.64 0.55 0.54 0.01
Cumulative % variance explained 55% 86% 98% 100%
Unreceptive behaviours include attacking or running away from the male or not
adopting a receptive posture in the presence of a displaying male.
L. B. Castilho et al. / Animal Behaviour 160 (2020) 43e5146
regressed against male condition, size and percentage of patch
cover.
Individual female preferences
To measure within-individual consistency and between-
individual variation in mate preference, we used the reaction
norms approach to build individual female preference functions for
male body size (see details of this approach in Dingemanse, Kazem,
R
eale, &Wright, 2010). Female acceptance was included as a
response variable, male size centralized by its mean was included
as a fixed continuous variable and female identity was included as a
random factor. With this approach, each female's intercept can be
interpreted as the mean sexual responsiveness of that female, and
each female's slope can be interpreted as the female's behavioural
plasticity due to differences in male size (Dingemanse et al., 2010).
Male size was chosen as a male quality of interest since it is easy to
measure and compare between males and is strongly correlated
with male weight (which is correlated with condition) and with
palp white patch area.
Preference variation due to quality
To test for differences in preference functions due to female
quality, we built regression models with the values of females’in-
tercepts or slopes of the reaction norm model as responses, and
female size as predictors. Since such relationships appeared to be
nonlinear, we performed general additive models (GAM) with the
local weighted linear regression (LOWESS) smoothing technique as
described by Zuur, Ieno, Walker, Saveliev, and Smith (2009).
Previous sexual experience
To evaluate the effect of previous sexual experience on female
preference functions, we also tested the effect of male presentation
order on the intercepts and slopes of female preference functions
through analysis of variance (ANOVA).
All analyses were performed in R (R Core Team, 2014) using the
package ‘nlme’and ‘nlme4’to build GLMs and the package ‘gam’to
build GAMs.
RESULTS
General Female Preference
The PCA of female general mate preference resulted in a first
principal component heavily and positively loaded on number of
copulations and total copulation time and negatively loaded on
percentage of failed copulation attempts. The second component
loaded heavily only on unreceptive behaviours. Together, both
components explained 86% of the total variance (Table 1). Since the
first component did not explain a large percentage of variance, and
the second only loaded on unreceptive behaviours, we used the
first principal component as a measure of female acceptance and
the raw values of unreceptive behaviours as a measure of female
rejection.
Female acceptance was not related to male mass, condition or
percentage of white patch cover (GLM:
b
weight
¼154.98, P¼0.15;
b
condition
¼44.13, P¼0.81;
b
patch
¼0.86, P¼0.85, N¼11).
Similarly, female rejection was unrelated to male weight or con-
dition, and only weakly affected by percentage of white patch cover
(GLM:
b
weight
¼437.83, P¼0.48;
b
condition
¼1968.50, P¼0.13;
b
patch
¼0.73, P¼0.053, N¼11). A closer evaluation of the per-
centage of white patch cover revealed no further relationship with
female rejection (Spearman rank correlation: r
S
¼0.27, N¼11,
P¼0.42).
Offspring Quality and Quantity
We had access to a total of 53 eggsacs, with an average of 21
spiderlings per sac. The PCA with offspring feeding performance
variables yielded a first principal component heavily and positively
loaded on eggsac production order and heavily and negatively
loaded on spiderlings' speed towards prey. The second component
loaded heavily and positively on spiderlings’latency to start mov-
ing towards prey (Table 2). The two principal components
explained 78% of the total variance.
Male quality and the first and second principal components for
offspring feeding performance yielded no significant relationships
(PC1:
b
male size
¼1.32, P¼0.14;
b
male condition
¼30.64, P¼0.67,
N¼18; PC2:
b
male size
¼0.05, P¼0.91;
b
male condition
¼24.12,
P¼0.55, N¼18). The only male quality variable related to the total
number of offspring was male size, with larger males producing
more offspring (GLM:
b
male condition
¼151.1, P¼0.94;
b
patch
¼64.2, P¼0.38;
b
size
¼159.5, P¼0.0007, N¼17). We
found that quality of adult males did not influence offspring sur-
vival, since the female random coefficients, extracted from the Cox
proportional hazard mixed model, were not related to male con-
dition, size or percentage of white patch cover (
b
condition
¼105.894,
P¼0.14;
b
size
¼0.55, P¼0.5;
b
patch
¼0.19, P¼0.93, N¼17).
Individual Female Preferences
The first principal component for individual female preference
explained 63% of the total variance and was strongly and positively
correlated with number of unreceptive behaviours, number of re-
jections and percentage of rejections. The first component was also
moderately and negatively correlated with the presence of copu-
lations. Then, we considered this first component as a measure of
female rejection. The second component explained another 21% of
the variance and was moderately and positively correlated with
presence of copulations and only weakly related to other variables
(Table 3). Since the first component did not explain a large part of
the variance, and the second was highly influenced only by pres-
ence of copulations, we used the first principal component in one
behavioural reaction model and the raw values of presence of
copulations in a separate behavioural reaction model.
The first model, using the first component, had no significant
general effect of male size on female rejection (
b
¼0.09, N¼46,
P¼0.89). This shows that, overall, females did not choose males
based on their size. However, significant differences in slopes
(likelihood ratio test:
c
21
¼13.20, P¼0.0014) showed that females
differed in how much they preferred different-sized males (Fig. 2).
However, these preference functions slopes cancelled each other
out at the level of the population (or general preference) because
some females preferred larger males while others preferred smaller
males.
For the behavioural binomial reaction norm using presence of
copulation as the response variable, we included female size as a
fixed effect in the model to ensure normality of random effects. The
effect of slopes was nonsignificant (likelihood ratio test:
c
21
¼4.2,
P¼0.12), while the effect of the intercept was significant (
c
21
¼4.9,
P¼0.02). Thus, we further analysed the model with random
Table 2
Component loadings from a PCA with different variables used to measure feeding
performance of H. adansoni young
Variables PC1 PC2 PC3
Latency to move towards prey 0.28 0.96 0.06
Speed moving towards prey 0.79 0.22 0.57
Eggsac number 0.8 0.12 0.58
Cumulative % variance explained 45% 78% 100%
L. B. Castilho et al. / Animal Behaviour 160 (2020) 43e51 47
intercepts only. Similar to our first reaction norm model, we had no
significant effect of male or female size on the probability of
copulation (male size:
b
¼2.6, P¼0.3; female size:
b
¼13.9,
N¼38, P¼0.11). This shows that, overall, male size did not affect
the probability of copulation, but females differed intrinsically in
their probability of copulation. Although the difference in slopes
was nonsignificant, the correlation between slopes and intercepts
was positive and high (R¼0.9), showing a pattern similar to that
found in our first model: females differed in their propensity to
copulate with different-sized males, but some females preferred
large males while others preferred small males, which could mask
female preferences if one measures only population-level
correlations.
Preference Variation due to Quality
Our GAMs did not reveal any relationship between female size
and female average sexual responsiveness (F
1,7. 5
¼1.5, P¼0.24) or
between female size and female sexual behavioural plasticity.
(F
1,7. 5
¼1.5, P¼0.25). This shows that female preference functions
cannot be predicted by female quality as assessed by body size.
Also, these nonsignificant effects are probably not very conserva-
tive. Female sexual responsiveness and plasticity were extracted
from random terms of mixed modes. Since there is an intrinsic
uncertainty in those terms (Houslay &Wilson, 2017), which we did
not account for, our results probably overestimate the relationship
between female size and sexual behaviour.
Previous Sexual Experience
Previous female interaction did not affect sexual responsiveness
or plasticity of females, since females first presented to large, small
or medium males did not differ in their intercepts or slopes in the
behavioural reaction model (ANOVA: slope: F
2,14
¼0.51, P¼0.608;
ANOVA: intercept: F
2,14
¼0.517, P¼0.607). This shows that
different sexual experiences (within the experimental paradigm)
did not dictate preferences of individual females. Similarly to our
GAMs, the values of intercept and slope were extracted from a
mixed model and have an intrinsic uncertainty related to them,
which was not accounted for. Thus, our nonsignificant Fvalues are
probably overestimated.
DISCUSSION
We found that female H. adansoni did not prefer any particular
male trait at the population level, despite the fact that larger males
produced more offspring. However, we found that females strongly
preferred males of different sizes at the individual level. These in-
dividual preferences, however, were arranged in a way that
cancelled each other out, eliminating any general preference
function. Variation in female preferences may arise from assorta-
tive mating (reviewed by Bel-Venner et al., 2008; Cotton et al.,
2006), variable social and life history (Bierbach, Sommer-Trembo,
Hanisch, Wolf, &Plath, 2015; Dukas, 2005; Fowler-Finn &Rodrí-
guez, 2011; Gaskett, Herberstein, Downes, &Elgar, 2004; Johnson &
Basolo, 2003; Place, Todd, Penke, &Asendorpf, 2010; Rundus, Bie-
muller, DeLong, Fitzgerald, &Nyandwi, 2015), variation in female
perception of male traits (Ronald et al., 2012) or selection for
genetically compatible mates (Landry, Garant, Duchesne, &
Bernatchez, 2001; Tregenza &Wedell, 2000; see Kelly, 2018 for a
review of all these topics), although we have not tested these ideas
directly. The effect of individual preferences could be significant,
including the maintenance of variation in male size due to sexual
selection (Kelly, 2018), despite higher fitness of larger males.
Different mating preferences may result from variation in fe-
male quality, and females of high quality are expected to choose
males of similar high quality in many species, either because they
can more easily bear the cost of choosiness or because they are
more capable of choosing high-quality males (Cotton et al., 2006;
Kelly, 2018). However, our results suggest that neither female
average sexual responsiveness nor the direction of female prefer-
ence for male traits was predicted by female size (which is corre-
lated with mass, and thus could be a measure of ‘quality’;L.
Castilho, personal observation). Thus, we believe differences in
female quality do not explain the highly variable preference func-
tions we observed in this study. It is possible, however, that less
ephemeral female characteristics guide female preferences. For
instance, females might prefer males that best match their per-
sonalities (e.g. Kralj-Fi
ser et al., 2013; Pog
any et al., 2018), if male
size is somehow correlated with these characteristcs. Also, other
factors that might influence individual quality, such as age, para-
sites and mutation loads, have been shown to change female
preference functions (Cotton et al., 2006).
Across taxa, one plausible reason for variation in female pref-
erence is a difference in benefits provided by males, when females
face different environmental or social contexts. For instance, in a
study with the wax moth (Achroia grisella), females that copulated
with more attractive males produced faster-growing and heavier
Table 3
Component loadings from a PCA with different variables used to measure female
H. adansoni preference for distinct sized males
Variables PC1 PC2 PC3 PC4
Unreceptive behaviours 0.83 0.38 0.35 0.23
0.87 0.39 0.1 0.3
Rejections
0.83 0.25 0.48 0.16
% Rejections
0.63 0.7 0.3 0.1
Presence of copulations
a
Cumulative % variance explained 63% 84% 96% 100%
Unreceptive behaviours included attacking or running away from the male or not
adopting a receptive posture in the presence of a displaying male.
a
Coded as 1 or 0.
–0.4 –0.2 0 0.2
Male size centralized b
y
the mean
–3
–2
–1
0
1
Female preference
Figure 2. Preference functions for individual Hasarius adansoni females as a function
of male size. Each dot represents a distinct mating trial. Each female underwent three
trials with different males, and each male was classified in one of three size categories
(small, average or large). Thus, each continuous line represents the preference function
for a specific female. Male size is centralized by its mean, so the intercept (vertical
dotted line) represents the preference of the females for an average-sized male. The
horizontal thick dotted line represents the population preference function.
L. B. Castilho et al. / Animal Behaviour 160 (2020) 43e5148
offspring when food was abundant and temperature was close to
optimum (Jia &Greenfield, 1997). However, when food availability
was reduced and temperatures were not optimal, females that
mated with less attractive males produced higher-quality offspring.
Other taxa are known for mate choice that varies based on their
own genotypes, and traits that indicate the genotype of a potential
mate. Disassortative mating leads to mating with partners with
dissimilar, but compatible genes, which in turn enhances offspring
heterozygosity (Landry et al., 2001; Tregenza &Wedell, 2000).
These examples may help explain why individual preference
functions evolve and are maintained in animals, including
H. adansoni, since different choices may yield alternative benefits,
given the female's genotype and her environment.
Such changes in the benefits provided by males could also help
explain the initially counterintuitive result showing that larger
males were not generally preferred, although they sired more
offspring. If female fitness is not entirely dependent on offspring
quantity, then females might choose smaller males if the benefits
generated by such a choice outweigh the risk of producing fewer
offspring.
The results of our study with H. adansoni may reflect the pre-
dictions of some models, which suggest that females should
constantly alternate between strategies of producing numerous
versus better offspring, especially when environmental conditions
vary (Fischer, Taborsky, &Kokko, 2011; Sinervo, Svensson, &
Comendant, 2000). Although our experimental conditions were
stable, the conditions each animal experienced while developing
are unknown. In addition, we measured offspring survival and
feeding performance and could not detect such a trade-off. How-
ever, trade-offs could occur relative to other factors influencing
offspring quality. For instance, predator avoidance could be an
important aspect of the quality of offspring (Johnson &Basolo,
2003) that could trade off with quantity.
We did not find an effect of previous sexual experience on fe-
male preference function. It is plausible, however, that experiences
earlier in life could have an effect in this regard (Kelly, 2018). For
instance, Dukas (2005) found that Drosophila courted by large
males early in life became less choosy; while Fowler-Finn and
Rodríguez (2011) showed that female treehoppers became
choosier after having been exposed to male vibrations in their
preferred frequency. Since some of our females were captured after
the last instar, we cannot rule out the possibility that males
encountered by these females earlier in life may have affected adult
mate preference and choosiness (Bierbach et al., 2015; Dukas, 2005;
Fowler-Finn &Rodríguez, 2011; Place et al., 2010; Santangelo, 2015;
Schuett et al., 2010; Swaddle, Cathey, Correll, &Hodkinson, 2005).
The consequences of individual variation are as important as the
reasons for which they evolve. Forstmeier and Birkhead (2004)
provided evidence that, in the zebra finch, some females prefer
orange-beaked males and others prefer red-beaked males. Simi-
larly, some females prefer more aggressive males, while others
prefer less aggressive ones. The variation in the preferred traits was
so high that the authors could not detect a general population
preference for beak colour or aggressiveness (but they did find one
for song rates). Such contradictory findings could have led to the
wrong conclusion that such characters are not being targeted by
sexual selection, had the authors not also assessed individual
preferences. In the present study, we found a similar pattern in
female preference for male size in the jumping spider H. adansoni.
Female preferences are related to male size, however, some females
prefer small males, while others prefer large males. The individual
preference functions were arranged in a way that cancelled each
other out, so that the net selection for size was close to zero (e.g. no
significant slope in the model). Interestingly, such a system may
also help maintain different male sizes in the population, since no
directional selection would arise for male size. In fact, the coeffi-
cient of variation for male size in our population (CV ¼11, 08%) was
above the average reported for most animal species in a compre-
hensive review reported by McKellar and Hendry (2009).
Future studies should also help us understand the evolutionary
importance of individual sexual preferences. Some studies
conclude that sexual selection alone has the potential to promote
speciation (Higashi, Takimoto, &Yamamura, 1999; Turner &
Burrows, 1995), and the possibility that distinct female prefer-
ences are heritable (Jennions &Petrie, 1997) reinforces this para-
digm. The idea that sexual selection plays a large role in speciation
is still under debate (Maan &Seehausen, 2011), but we believe that
distinct female preferences could potentially facilitate speciation in
the long term, since this would maintain the needed genetic
variance.
A comparison of population and individual levels of mate pref-
erence is an important step towards developing and testing
comprehensive hypotheses in sexual selection, since this is a
growing field and yet under-represented in the behavioural liter-
ature (Ah-King &Gowaty, 2016; Wagner, 1998). Our study also has
the advantage of targeting the behaviour of a tropical species, most
of which are extremely under-represented in the sexual selection
literature (e.g. Macedo, 2010). Because the tropical and temperate
regions differ in their selection pressures (e.g. temperature, biodi-
versity, photoperiod, seasonality, etc.; Willig, Kaufman, &Stevens,
2003), information about tropical species can help us understand
the variability and types of adaptations across a larger pool of
species, strengthening the conceptual pillars of sexual selection
theory.
Our results contribute to the accumulating evidence that mate
preference is more labile than previously thought. Actually, mate
preference is a flexible behaviour that may vary with environment
and the choosy sex genotype, since partner-derived benefits are
also expected to be similarly flexible (Candolin, 2003). Interest-
ingly, depending on how individual preference functions are ar-
ranged, it may be impossible for a population preference function
to be assessed, an important alert for empiricists seeking to
describe sexual preferences in a species. In studies in which a
population appears to display no sexual preferences, individual
preferences might be playing an important role, and such a
conclusion can only be assessed when individual preference func-
tions are accounted for. Note, however, that even when a
population-level preference is present, individual-level preferences
can also play an important role in sexual selection and evolution
(e.g. Forstmeier &Birkhead, 2004), and can, at least in principle,
help in the process of speciation if such variation is heritable (e.g.
Gray &Cade, 2000).
Declaration of Interest
None.
Acknowledgments
We thank the Coordenaç~
ao de Aperfeiçoamento de Pessoal de
Nível Superior (CAPES), Brazilian federal government agency under
the Ministry of Education, Brazil, the Conselho Nacional de
Desenvolvimento Científico e Tecnol
ogico (CNPq), Brazilian Na-
tional Council for Scientific and Technological Development is an
organization of the Brazilian federal government under the Min-
istry of Science and Technology, Brazil and the Emerging Leaders in
the American Program (ELAP), Canada for the scholarships pro-
vided to Leonardo B. Castilho, CNPq for a fellowship for Regina H.
Macedo, and the Canada Research Chairs program, Canada and
Canadian Foundation for Innovation, Canada for infrastructure and
L. B. Castilho et al. / Animal Behaviour 160 (2020) 43e51 49
support for Maydianne C. B. Andrade. We thank Dr Rosana Tidon
and her staff for providing food for the spiders. We are also grateful
to all the staff in the Andrade lab for the academic support given
during the time data for this project was being collected, and to
everyone who helped collect spiders for this project, especially
Vitor Renan. We also thank two anonymous referees for helpful
comments on previous versions of this paper.
References
Ah-King, M., & Gowaty, P. A. (2016). A conceptual review of mate choice: Stochastic
demography, within-sex phenotypic plasticity, and individual flexibility. Ecol-
ogy and Evolution, 6(14), 4607e4642. https://doi.org/10.1002/ece3.2197.
Andersson, M. (1982). Female choice selects for extreme tail length in a widowbird.
Nature, 299(28), 818e820.
Andersson, M. (1994). Sexual selection. Princeton, NJ: Princeton University Press.
Baldauf, S. A., Engqvist, L., Ottenheym, T., Bakker, T. C. M., & Thünken, T. (2013). Sex-
specific conditional mating preferences in a cichlid fish: Implications for sexual
conflict. Behavioral Ecology and Sociobiology, 67(7), 1179e1186. https://doi.org/
10.1007/s00265-013-1543-4.
Bateman, A. J. (1948). Intra-sexual selection in Drosophila.Heredity, 2, 349e368.
Bel-Venner, M. C., Dray, S., Allain
e, D., Menu, F., & Venner, S. (2008). Unexpected
male choosiness for mates in a spider. Proceedings of the Royal Society B: Bio-
logical Sciences, 275(1630), 77e82. https://doi.org/10.1098/rspb.2007.1278.
Bell, A. (2007). Future directions in behavioral syndromes research. Proceedings of
the Royal Society B: Biological Sciences, 274, 755e761. https://doi.org/10.1098/
rspb.2006.0199.
Bierbach, D., Sommer-Trembo, C., Hanisch, J., Wolf, M., & Plath, M. (2015). Person-
ality affects mate choice: Bolder males show stronger audience effects under
high competition. Behavioral Ecology, 26(5), 1314e1325. https://doi.org/10.1093/
beheco/arv079.
Bulbert, M. W., Hanlon, J. C. O., Zappettini, S., Zhang, S., & Li, D. (2015). Sexually
selected UV signals in the tropical ornate jumping spider, Cosmophasis
umbratica may incur costs from predation. Ecology and Evolution, 5(4), 914e920.
https://doi.org/10.1002/ece3.1419.
Busso, J., & Rabosky, D. (2016). Disruptive selection on male reproductive poly-
morphism in a jumping spider, Maevia inclemens.Animal Behaviour, 120,1e10.
https://doi.org/10.1016/j.anbehav.2016.07.028.
Candolin, U. (2003). The use of multiple cues in mate choice. Biological Reviews, 78,
575e595.
Carter, A. J., Goldizen, A. W., & Tromp, S. A. (2010). Agamas exhibit behavioral
syndromes: Bolder males bask and feed more but may suffer higher predation.
Behavioral Ecology, 21(3), 655e661. https://doi.org/10.1093/beheco/arq036.
Castilho, L. B., Andrade, M. C. B., & Macedo, R. H. (2018). Mating and egg-laying
behavior of Hasarius adansoni (Araneae: Salticidae) and the influence of sex-
ual selection. Journal of Arachnology, 46(3), 398e403. https://doi.org/10.1636/
joas16-091.1.
Castilho, L. B., & Macedo, R. F. (2016). Behavioral types and syndromes in the
neotropical passarine blue-black grassquit (Volatinia jacarina). Oecologia Aus-
tralis, 20(3), 28e36. https://doi.org/10.4257/oeco.2016.2003.04.
Cotton, S., Small, J., & Pomiankowski, A. (2006). Sexual selection and condition-
dependent mate choice. Current Biology, 16(17), R755eR756. https://doi.org/
10.1016/j.cub.2006.08.022.
Dall, S. R., Houston, A. I., & McNamara, J. M. (2004). The behavioural ecology of
personality: Consistent individual differences from an adaptive perspective.
Ecology Letters, 7(8), 734e739. https://doi.org/10.1111/j.1461-
0248.2004.00618.x.
Dingemanse, N. J., Both, C., van Noordwijk, A. J., Rutten, A. L., & Drent, P. J. (2003).
Natal dispersal and personalities in great tits (Parus major). Proceedings of the
Royal Society of London, Series B: Biological Sciences, 270(1516), 741e747. https://
doi.org/10.1098/rspb.2002.2300.
Dingemanse, N. J., Kazem, A. J. N., R
eale, D., & Wright, J. (2010). Behavioural reaction
norms: Animal personality meets individual plasticity. Trends in Ecology &
Evolution, 25(2), 81e89. https://doi.org/10.1016/j.tree.2009.07.013.
Dingemanse, N., & R
eale, D. (2005). Natural selection and animal personality.
Behaviour, 142,1159e1184 .
Dukas, R. (2005). Learning affects mate choice in female fruit flies. Behavioral
Ecology, 16(4), 800e804. https://doi.org/10.1093/beheco/ari057.
Edwards, D. L., Melville, J., Joseph, L., & Keogh, S. (2015). Ecological divergence,
adaptive diversification, and the evolution of social signaling traits: An empir-
ical study in arid Australian lizards. American Naturalist, 186(6), E144eE161.
https://doi.org/10.1086/683658.
Elgar, M. A., & Fahey, B. F. (1996). Sexual cannibalism, competition, and size
dimorphism in the orb-weaving spider Nephila plumipes Latreille (Araneae :
Araneoidea). Behavioral Ecology, 7(2), 195e198.
Elias, D. O., Maddison, W. P., Peckmezian, C., Girard, M. B., & Mason, A. C. (2012).
Orchestrating the score: Complex multimodal courtship in the Habronattus
coecatus group of Habronattus jumping spiders (Araneae: Salticidae). Biological
Journal of the Linnean Society, 105(3), 522e547. https://doi.org/10.1111/
j.10958312.2011.01817.x.
Elias, D. O., Mason, A. C., & Hoy, R. R. (2004). The effect of substrate on the efficacy of
seismic courtship signal transmission in the jumping spider Habronattus
dossenus (Araneae: Salticidae). Journal of Experimental Biology, 207(23),
4105e4110. https://doi.org/10.1242/jeb.01261.
Fischer, B., Taborsky, B., & Kokko, H. (2011). How to balance the offspring quality
quantity tradeoff when environmental cues are unreliable. Oikos, 120(2),
258e270. https://doi.org/10.1111/j.1600-0706.2010.18642.x.
Foelix, R. F. (2011). Reproduction. In R. F. Foelix (Ed.), Biology of spiders (pp.
218e260). Oxford, U.K.: Oxford University Press.
Forstmeier, W. (2007). Do individual females differ intrinsically in their propensity
to engage in extra-pair copulations? PLoS One, 2(9). https://doi.org/10.1371/
journal.pone.0000952. e952.
Forstmeier, W., & Birkhead, T. (2004). Repeatability of mate choice in the zebra
finch: Consistency within and between females. Animal Behaviour, 68(5),
1017e1028. https://doi.org/10.1016/j.anbehav.2004.02.007.
Fowler-Finn, K. D., & Rodríguez, R. L. (2011). Experience-mediated plasticity in mate
preferences: Mating assurance in a variable environment. Evolution, 66(2),
459e468.
Gaskett, A., Herberstein, M., Downes, B., & Elgar, M. (2004). Changes in male mate
choice in a sexually cannibalistic orb-web spider (Araneae: Araneidae).
Behaviour, 141,1197e1210.
Girard, M., Elias, D., & Kasumovic, M. (2015). Female preference for multimodal
courthip: Multiple signals are important for male mating success in a peacock
spider. Proceedings of the Royal Society B: Biological Sciences, 282(1820), 12e14.
https://doi.org/10.1098/rspb.2015.2222.
Godin, J.-G., & Dugatkin, L. A. (1995). Variability and repeatability of female mating
preferences in the guppy. Animal Behaviour, 49(6), 1427e1433.
Gray, A. D., & Cade, H. W. (2000). Sexual selection and speciation in field crickets.
Proceedings of the National Academy of Sciences of the United States of America,
97(26), 14449e14454. https://doi.org/10.1073/pnas.97.26.14449.
Hebets, E. A. (2003). Subadult experience influences adult mate choice in an
arthropod: Exposed female wolf spiders prefer males of a familiar phenotype.
Proceedings of the National Academy of Sciences of the United States of America,
100(23), 13390e13395.
Higashi, M., Takimoto, G., & Yamamura, N. (1999). Sympatric speciation by sexual
selection. Nature, 402, 523e526.
Hoefler, C. D., Persons, M. H., & Rypstra, A. L. (20 08). Evolutionarily costly courtship
displays in a wolf spider: A test of viability indicator theory. Behavioral Ecology,
19(5), 974e979. https://doi.org/10.1093/beheco/arn055.
Houslay, T. M., & Wilson, A. J. (2017). Avoiding the misuse of BLUP in behavioural
ecology. Behavioral Ecology, 28(4), 948e952. https://doi.org/10.1093/beheco/
arx023.
Jakob, E., Marshall, S., & Uetz, G. (1996). Estimating fitness: A comparison of body
condition indicies. Oikos, 77(1), 61e67.
Jennions, M., & Petrie, M. (1997). Variation in mate choice and mating preferences:
A review of causes and consequences. Biological Reviews of the Cambridge
Philosophical Society, 72, 283e327.
Jia, F., & Greenfield, M. D. (1997). When are good genes good? Variable outcomes of
female choice in wax moths. Proceedings of the Royal Society of London - Series B:
Biological Sciences, 264(1384), 1057e1063. https://doi.org/10.1098/rspb.1997.
0146.
Johnson, J. B., & Basolo, A. L. (2003). Predator exposure alters female mate choide in
the green swordtail. Behavioral Ecology, 14(5), 619e625. https://doi.org/10.1093/
beheco/arg046.
Kasumovic, M. M., & Andrade, M. C. B. (2006). Male development tracks rapidly
shifting sexual versus natural selection pressures. Current Biology, 16(7),
R242eR243. https://doi.org/10.1016/j.cub.2006.03.017.
Kasumovic, M. M., Brooks, R. C., & Andrade, M. C. B. (2009). Body condition but not
dietary restriction prolongs lifespan in a semelparous capital breeder. Biology
Letters, 5(5), 636e638. https://doi.org/10.1098/rsbl.2009.0335.
Keiser, C. N., Lichtenstein, J. L., Wright, C. M., Chism, G. T., & Pruitt, J. N. (2018).
Personality and behavioral syndromes in insects and spiders. In A. C
ordoba-
Aguilar, D. Gonz
alez-Tokman, & I. Gonz
alez-Santoyo (Eds.), Insect behavior: From
mechanisms to ecological and evolutionary consequences (pp. 236e256). Oxford,
U.K.: Oxford University Press.
Kelly, C. D. (2018). The causes and evolutionary consequences of variation in female
mate choice in insects: The effects of individual state, genotypes and environ-
ments. Current Opinion in Insect Science, 27,1e8. https://doi.org/10.1016/
j.cois.2018.01.010.
Kirkpatrick, M. (1982). Sexual selection and the evolution of female choice. Evolu-
tion, 36(1), 1e12.
Kokko, H., Jennions, M., & Brooks, R. (2006). Unifying and testing models of sexual
selection. Annual Review of Ecology, Evolution and Systematics, 37(1), 43e66.
https://doi.org/10.1146/annurev.ecolsys.37.091305.110259.
Kralj-Fi
ser, S., Schneider, J., & Kuntner, M. (2013). Challenging the aggressive spill-
over hypothesis: Is pre-copulatory sexual cannibalism a part of a behavioural
syndrome? Ethology, 119(8), 615e623. https://doi.org/10.1111/eth.12111.
Landry, C., Garant, D., Duchesne, P., & Bernatchez, L. (2001). ‘Good genes as het-
erozygosity’: The major histocompatibility complex and mate choice in Atlantic
salmon (Salmo salar). Proceedings of the Royal Society of London, Series B: Bio-
logical Sciences, 268(1473), 1279e1285. https://doi.org/10.1098/rspb.2001.1659.
Maan, M. E., & Seehausen, O. (2011). Ecology, sexual selection and speciation.
Ecology Letters, 14,591e602.
Macedo, R. (2010). Behavioral ecology of tropical animals (1st ed., Vol. 42). London,
U.K.: Academic Press.
McKellar, A. E., & Hendry, A. P. (2009). How humans differ from other animals in
their levels of morphological variation. PLoS One, 4(9), 1e5.
L. B. Castilho et al. / Animal Behaviour 160 (2020) 43e5150
Møller, A. P. (2010). Interspecific variation in fear responses predicts urbanization in
birds. Behavioral Ecology, 21(2), 365e371. https://doi.org/10.1093/beheco/
arp199.
Møller, A. P., & Alatalo, R. V. (1999). Good-genes effects in sexual selection. Pro-
ceedings of the Royal Society of London, Series B: Biological Sciences, 266,85e91.
Place, S. S., Todd, P. M., Penke, L., & Asendorpf, J. B. (2010). Humans show mate
copying after observing real mate choices. Evolution and Human Behavior, 31(5),
1e6. https://doi.org/10.1016/j.evolhumbehav.2010.02.001.
Pog
any,
A., Vincze, E., Szurovecz, Z., Kosztol
anyi, A., Barta, Z., Sz
ekely, T., et al. (2018).
Personality assortative female mating preferences in a songbird. Behaviour, 155,
481e503.
R Core Team. (2017). R: A language and environment for statistical computing. Vienna,
Austria: R Foundation for Statistical Computing. https://www.R-project.org/.
Rabaneda-Bueno, R., Aguado, S., Fern
andez-Montraveta, C., & Moya-Lara~
no, J.
(2014). Does female personality determine mate choice through sexual canni-
balism? Ethology, 120(3), 238e248. https://doi.org/10.1111/eth.12197.
R
eale, D., Reader, S., Sol, D., McDougall, P., & Dingemanse, N. (2007). Integrating
animal temperament within ecology and evolution. Biological Reviews, 82,
291e318. https://doi.org/10.1111/j.1469-185X.2007.00010.x.
Richman, D. B., & Jackson, R. R. (1992). A review of the ethology of jumping spiders
(Araneae, Salticidae). Bulletin of the British Arachnological Society, 9(2), 33e37.
Ronald, K., Ensminger, A., Shawkey, M., Lucas, J., & Fern
andez-Juricic, E. (2017).
Testing a key assumption in aninal communication: Between-individual vari-
ation in female visual systems alters perception of male signals. Biology Open, 6,
177 1e1783. https://doi.org/10.1242/bio.028282.
Ronald, K. L., Fern
andez-Juricic, E., & Lucas, J. R. (2012). Taking the sensory
approach: How individual differences in sensory perception can influence mate
choice. Animal Behaviour, 84(6), 1283e1294.
Ronald, K., Fern
andez-Juricic, E., & Lucas, J. (2018). Mate choice in the eye and ear of
the beholder? Female multimodal sensory configuration influences her pref-
erences. Proceedings of the Royal Society B: Biological Sciences, 285(1878), 1e10.
Rundus, A., Biemuller, R., DeLong, K., Fitzgerald, T., & Nyandwi, S. (2015). Age-
related plasticity in male mate choice decisions by Schizocosa retrorsa wolf
spiders. Animal Behaviour, 107, 233e238.
Santangelo, N. (2015). Female breeding experience affects parental care strategies of
both parents in a monogamous cichlid fish. Animal Behaviour, 104,31e37.
https://doi.org/10.1016/j.anbehav.2015.03.004.
von Schantz, T., Bensch, S., Grahn, M., Hasselquist, D., & Wittzell, H. (1999). Good
genes, oxidative stress and condition-dependent sexual signals. Proceedings of
the Royal Society of London, Series B: Biological Sciences, 266,1e12.
Schilte-Hostedde, A., Zinner, B., Millar, J. S., & Hickling, G. (2005). Restitution of
massesize residuals: Validating body condition indices. Ecology, 86(1), 155e163.
Schneider, C. A., Rasband, W. S., & Eliceiri, K. W. (2012). NIH image to ImageJ: 25
years of image analysis. Nature Methods, 9(7), 671e675.
Schuett, W., Tregenza, T., & Dall, S. R. X. (2010). Sexual selection and animal per-
sonality. Biological Reviews of the Cambridge Philosophical Society, 85(2),
217e246. https://doi.org/10.1111/j.1469-185X.2009.00101.x.
Sih, A., Bell, A. M., & Johnson, C. (2004). Behavioral syndromes: An ecological and
evolutionary overview. Trends in Ecology &Evolution, 19(7), 372e378.
Sih, A., Bell, A. M., Johnson, C., & Ziemba, R. (20 04). Behavioral syndromes: An
integrative overview. Quarterly Review of Biology, 79(3), 241e277.
Sinervo, B., Svensson, E., & Comendant, T. (2000). Density cycles and an offspring
quantity and quality game driven by natural selection. Nature, 406(6799),
985e988. https://doi.org/10.1038/35023149.
Smith, B. R., & Blumstein, D. T. (2010). Behavioral types as predictors of survival in
Trinidadian guppies (Poecilia reticulata). Behavioral Ecology, 21(5), 919e926.
https://doi.org/10.1093/beheco/arq084.
Stamps, J., & Groothuis, T. (2010). Developmental perspectives on personality: Im-
plications for ecological and evolutionary studies of individual differences.
Philosophical Transactions of the Royal Society of London, Series B: Biological
Sciences, 365(1560), 4029e4041. https://doi.org/10.1098/rstb.2010.0218.
Swaddle, J. P., Cathey, M. G., Correll, M., & Hodkinson, B. P. (2005). Socially trans-
mitted mate preferences in a monogamous bird: A non-genetic mechanism of
sexual selection. Proceedings of the Royal Society of London, Series B: Biological
Sciences, 272(1567), 1053e1058. https://doi.org/10.1098/rspb.2005.3054.
Tregenza, T., & Wedell, N. (2000). Genetic compatibility, mate choice and patterns of
parentage: Invited review. Molecular Ecology, 9(8), 1013e1027. https://doi.org/
10.1046/j.1365-294x.2000.00964.x.
Turner, G. F., & Burrows, M. T. (1995). A model of sympatric speciation by sexual
selection. Proceedings of the Royal Society of London, Series B: Biological Sciences,
260, 287e292. https://doi.org/10.1098/rspb.1995.0093.
Wagner, W. E. (1998). Measuring female mating preferences. Animal Behaviour, 55,
1029e1042.
Willig, M. R., Kaufman, D. M., & Stevens, R. D. (2003). Latitudinal gradients of
biodiversity: Pattern, process, scale, and synthesis. Annual Review of Ecology,
Evolution and Systematics, 34,273e309.
Wolf, M., & Wessing, F. (2010). An explanatory framework for adaptive personality
differences. Philosophial Transactions of the Royal Society B, 365(1560),
3959e3968. https://doi.org/10.1098/rstb.2010.0215.
Zhao, Q. S., Hu, Y. B., Liu, P. F., Chen, L. J., & Sun, Y. H. (2016). Nest site choice: A
potential pathway linking personality and reproductive success. Animal
Behaviour, 118,97e103. https://doi.org/10.1016/j.anbehav.2016.05.017.
Zuur, A. F., Ieno, E. N., Walker, N. J., Saveliev, A. A., & Smith, G. M. (2009). Things are
not always linear; Additive modeling. In A. F. Zuur, E. N. Ieno, N. J. Walker,
A. A. Saveliev, & G. M. Smith (Eds.), Mixed effects models and extensions in
ecology with R (pp. 35e69). New York, NY: Springer.
L. B. Castilho et al. / Animal Behaviour 160 (2020) 43e51 51
