Sex-specific life-history strategies amongst immature jumping spiders: differences in body parameters and behaviour

Abstract

Selection forces often generate sex-specific differences in various traits closely related to fitness. While in adult spiders (Araneae), sexes often differ in colouration, body size, antipredator or foraging behaviour, such sex-related differences are less pronounced amongst immatures. However, sex-specific life-history strategies may also be adaptive for immatures. Thus, we hypothesized that, among spiders, immature individuals show different life-history strategies that are expressed as sex-specific differences in body parameters and behavioural features, and also in their relationships. We used immature individuals of a protandrous jumping spider, Carrhotus xanthogramma, and examined sex-related differences. Results showed that males have higher mass and larger prosoma than females. Males were more active and more risk-tolerant than females. Male activity increased with time, and larger males tended to capture the prey faster than small ones, while females showed no such patterns. However, females reacted to the threatening abiotic stimuli more with the increasing number of test sessions. In both males and females, individuals with better body condition tended to be more risk-averse. Spiders showed no sex-specific differences in inter-individual behavioural consistency and in intra-individual behavioural variation in the measured behavioural traits. Finally, we also found evidence for behavioural syndromes (i.e. correlation between different behaviours), where in males only the activity correlated with the risk-taking behaviour, but in females all the measured behavioural traits were involved. The present study demonstrates that C. xanthogramma sexes follow different life-history strategies even before attaining maturity.

Full text

Current Zoology, 2022, XX, 1–17
https://doi.org/10.1093/cz/zoac069
Advance access publication 6 September 2022
Original Article
© The Author(s) 2022. Published by Oxford University Press on behalf of Editorial Ofce, Current Zoology.
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons.org/
licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For
commercial re-use, please contact journals.permissions@oup.com
Received 14 April 2022; accepted 2 September 2022
Sex-specific life-history strategies among immature
jumping spiders: differences in body parameters and
behavior
LászlóMezőfia,*,, ViktorMarkóa, Dóra ÁgnesTaranyib and GáborMarkóa,
aInstitute of Plant Protection, Hungarian University of Agriculture and Life Sciences, Budapest 1118, Hungary;
bInstitute of Viticulture and Enology, Hungarian University of Agriculture and Life Sciences, Budapest 1118, Hungary
*Address correspondence to László Mezőfi. E-mail: mezofilaszlo@gmail.com
Handling editor: Zhi-YunJia
Abstract
Selection forces often generate sex-specific differences in various traits closely related to fitness. While in adult spiders (Araneae), sexes often
differ in coloration, body size, antipredator, or foraging behavior, such sex-related differences are less pronounced among immatures. However,
sex-specific life-history strategies may also be adaptive for immatures. Thus, we hypothesized that among spiders, immature individuals show
different life-history strategies that are expressed as sex-specific differences in body parameters and behavioral features, and also in their rela-
tionships. We used immature individuals of a protandrous jumping spider, Carrhotus xanthogramma, and examined sex-related differences. The
results showed that males have higher mass and larger prosoma than females. Males were more active and more risk tolerant than females.
Male activity increased with time, and larger males tended to capture the prey faster than small ones, while females showed no such patterns.
However, females reacted to the threatening abiotic stimuli more with the increasing number of test sessions. In both males and females,
individuals with better body conditions tended to be more risk averse. Spiders showed no sex-specific differences in interindividual behavioral
consistency and in intraindividual behavioral variation in the measured behavioral traits. Finally, we also found evidence for behavioral syndromes
(i.e., correlation between different behaviors), where in males, only the activity correlated with the risk-taking behavior, but in females, all the
measured behavioral traits were involved. The present study demonstrates that C. xanthogramma sexes follow different life-history strategies
even before attaining maturity.
Key words: activity, behavioral syndrome, boldness, intraindividual variability, repeatability, sexual dimorphism
Animals often face life-history trade-offs during their life-
time due to different internal or external constraints, gen-
erating inter- and intraspecic variations in traits that are
tightly linked with their tness (Alonzo and Kindsvater 2008;
Chapin 2017). Differential reproductive investment among
sexes can also generate variability (regarding many tness-re-
lated traits) among individuals. Therefore, due to their differ-
ent investment, sexes may face different selection pressures,
resulting in physiological, morphological, and behavioral dif-
ferences (Slatkin 1984). Sexually dimorphic traits are shaped
either by sexual selection or by natural selection. However,
many traits are often exposed simultaneously to both forms
of selective forces (Hosken and House 2011). Generally, the
effect of net selection (sum of selection forces) is stronger in
males (Winkler et al. 2021). This seems to allow populations
to adapt faster to new environmental challenges (Winkler et
al. 2021). Therefore, selection may shape the traits of females
and males in different ways resulting in sex-specic life histo-
ries across several interlinked traits (Hämäläinen et al. 2018;
Tarka et al. 2018).
Animal behavior can be an adaptive and exible response
to various ecological and environmental challenges and can
directly affect specic tness components (Moiron et al.
2020). Recently, an increasing number of studies have focused
on behavioral ecology of different arthropod taxa, including
spiders, to understand inter- and intraindividual behavioral
variations and behavioral correlations (i.e., behavioral syn-
dromes) (Kralj-Fišer and Schuett 2014; Modlmeier et al.
2015). Similarly to vertebrates, certain arthropod taxa have
consistent interindividual behavioral differences (also referred
to as animal personality: the temporal variation of the same
behavioral trait) (Bell et al. 2009; Réale and Dingemanse
2012). Behavioral consistency (in one behavior and/or among
correlated behaviors) might favor individuals and, through
them, populations or species in an adaptive manner, depend-
ing on the current ecological situation (Dingemanse and
Réale 2005; Réale and Dingemanse 2012; Jandt et al. 2014).
However, compared with vertebrates, this phenomenon in
arthropods has received much less attention. Nevertheless,
some good examples emphasize the adaptive signicance
of consistent behavioral differences in certain spider spe-
cies as compared to other species. Larinioides sclopetarius
(Araneidae) easily colonize urban habitats, possibly due to
personality, that is, consistent boldness and increased activ-
ity in a novel environment (Kralj-Fišer and Schneider 2012;
Kralj-Fišer et al. 2017). Besides this, certain arthropod taxa
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

2Current Zoology
can form behavioral syndromes when behavioral traits meas-
ured in 2 or more functionally different ecological situations/
contexts correlate with each other (Sih et al. 2004; Royauté
et al. 2014; Michalko et al. 2017). For example, in a shing
spider (Dolomedes triton, Pisauridae), voracity toward heter-
ospecic prey often positively correlates with precopulatory
sexual cannibalism. Although consuming potential mates is
not necessarily adaptive, it seems that high levels of vorac-
ity (in different situations) can improve the tness of spiders
by increasing their adult size and fecundity (Johnson and Sih
2005). Apart from this, some level of behavioral plasticity
that allows individuals of a species to react rapidly to an
emerging environmental challenge by altering behavior can
also be advantageous (Dingemanse and Réale 2005; Snell-
Rood 2013). Sometimes individuals show high similarity
in the behavioral mean, but there may be individual differ-
ences in behavioral predictability (Stamps et al. 2012). For
example, in an araneophagous jumping spider, Portia labiata
(Salticidae), the intraindividual variability (IIV) in boldness
may increase in the presence of a conspecic (Chang et al.
2019). Higher IIV in a behavior indicated lower predictability
(i.e., individuals behave less consistently), which may be more
advantageous for P. labiata in a dangerous situation (Chang
et al. 2019). These examples indicate that arthropods, par-
ticularly spiders, are suitable model organisms to provide a
unique insight into behavioral variations.
Among spiders, sexual dimorphism is very common in
several features such as morphology, behavior, and life his-
tory. Sex-specic morphological and behavioral traits can be
genetically determined (Kralj-Fišer et al. 2019, 2021; Chang
et al. 2020; Cordellier et al. 2020) and generate specic conse-
quences for life history. For example, spiders usually practice
protandry, that is, a male matures earlier and has a shorter
lifespan than a female (Klein 1988; Uhl et al. 2004; Foelix
2011; Kralj-Fišer et al. 2014). Sexual maturation can often
change the males’ appearance and behavior (Sullivan and
Morse 2004; Framenau 2005; Cordellier et al. 2020). Besides
these differences, there are sex-specic variations in metabolic
rate (Kotiaho 1998), immune response (Rádai et al. 2018),
body size (Head 1995; Prenter et al. 1998), or other morpho-
logical traits (Albín et al. 2018). Furthermore, spiders exhibit
sexual dimorphism in features related to behavioral ecology,
for example, in aggressiveness (Kralj-Fišer et al. 2017, 2019),
boldness (Sweeney et al. 2013), the behavioral mean, and
even in the temporal pattern of locomotor activity (Schmitt
et al. 1990; Krumpálová and Tuf 2013; Mező et al. 2019).
Most behavioral studies of invertebrates have focused on
adult individuals where sex-specic differences (e.g., mor-
phological traits) are clearly expressed. However, a few stud-
ies have shown that the manifested behavioral traits may not
always be the same during the individual lifespan as behavioral
trait expression can change with development and experience
(Carducci and Jakob 2000; Dingemanse et al. 2010; DiRienzo
and Montiglio 2016). For example, the effect of the environment
played an essential role in shaping behavioral traits in Marpissa
muscosa (Salticidae). Individuals of this species grown in a phys-
ically enriched environment tended to be more exploratory
(Liedtke et al. 2015), while social enrichment improved their
cognitive abilities (Liedtke and Schneider 2017). These studies
highlight the relationship between environmental conditions and
behavioral exibility during the ontogeny of immature spiders,
but there is still a knowledge gap on the sex-related differences
in immature stages.
The protandrous jumping spider, Carrhotus xanthogramma
(Latreille, 1819), is a euryphagous species that is widely dis-
tributed from Europe to the Far East and can be the domi-
nant hunting spider in the canopy level of pome fruit orchards
(Mező et al. 2020; WSC 2021). Although the immature
stages have a fairly similar pattern and coloration, adults
show marked sexual dimorphism in these features (males are
much darker than females, but the opisthosoma of males is
usually a richer red brown), and according to Kim and Lee
(2014), males are generally slightly shorter in body length.
Additionally, a recent study found sex-related behavioral
differences in adults: Females were more active than males
(Mező et al. 2019).
Therefore, we aimed to detect sex-specic differences
regarding the most crucial tness-related traits in the seem-
ingly uniform immature individuals (i.e., penultimate and
antepenultimate instars) of our chosen model organism, C.
xanthogramma. We tested for sex-specic differences in some
body parameters and behavioral variability across time (both
at inter- and intraindividual levels) and functionally differ-
ent ecological situations (i.e., behavioral syndrome). We sup-
posed that sex-specic differences might already be expressed
before maturity due to genetically determined physiological
background (i.e., sex-specic life-history trade-offs). In that
case, we expected sex-specic differences in some specic
tness-related traits and their correlation structure. Finally,
we also tested the relationship between body parameters and
behavioral traits. If any sex-specic differences were detected
regarding the body parameters or behavior, we also predicted
a close link among these traits suggesting sex-specic differ-
ences in the life-history trade-off structure.
Materials and Methods
Test animals and animal housing
We collected C. xanthogramma individuals from 3 insecti-
cide-free apple orchards, that is, from 3 spatially isolated C.
xanthogramma populations. We labeled the sampling sites by
their closest village as follows: Csány (“Site_Cs”, 47°38ʹ25″N,
19°46ʹ24″E) and Madocsa (“Site_M1”, 46°40ʹ42″N,
18°58ʹ21″E; “Site_M2”, 46°40ʹ48″N, 18°58ʹ31″E). Both
Site_Cs and Site_M1 had been abandoned for several years,
while the orchard Site_M2 was sprayed yearly with a con-
tact fungicide combination (Vegesol-eReS, BVN Növényvédő
Ltd., active ingredients: copper hydroxide + sulfur + sun-
ower oil, dose: 5L/ha, applied: 12 April 2017) and here, the
vegetation between the rows was mown 2 times a year. All
the sampled sites were surrounded by other apple orchards
and crop elds.
We collected 31, 31, and 30 immature (mainly penultimate)
individuals by beating the branches of randomly chosen apple
trees between the 11 and 15 of September 2017 in Site_Cs,
Site_M1, and Site_M2, respectively. For the beating, we used
a beating funnel 70cm in diameter and a ~120-cm-long beat-
ing stick, and we collected the spiders that fell from the tree.
In the laboratory, we housed the collected spiders individu-
ally in plastic Petri dishes (height: 16 mm, outer diameter:
61mm) and placed them in random order on plastic trays
after labeling them by a specic ID. We covered the sides of
the Petri dishes with white tape to avoid possible disturbance
from spiders in the neighboring Petri dishes. On 20 September
2017, we placed and thereafter kept the spiders in our behav-
ioral laboratory in a standard environment (temperature
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

Mezőfi et al. · Dimorphic strategies among immatures of a jumping spider 3
[mean, ±standard deviation {SD}]: 19.26 °C, ± 1.39 °C; rel-
ative humidity [mean, ± SD]: 42.26%, ±5.41%; photoperiod
[L:D]: 16:8h). To synchronize their hunger level, on the rst
acclimatization day (21 September 2017), we fed each spider
with 3 ightless fruit ies Drosophila hydei. Apart from this,
the spiders received food only during the behavioral assays
(see later), resulting in controlled hunger levels in all sessions.
Additional water supply was restricted during the experiments
to keep the relative humidity low because C. xanthogramma
mainly prefers sunny and dry habitats (Szinetár 2006). We
started the behavioral assays on 26 September 2017, so the
spiders had enough time to acclimatize to the new conditions
in the behavioral laboratory.
A few days before the rst behavioral assay session (on the
18th of September) and after the last (on the 24th of October)
behavioral assay session, we measured the mass of each spider
using an analytic scale (Kern-PCB 250-3). None of the indi-
viduals molted over the course of the experiment. At theend
of the study, we also measured the prosoma width of the
preserved specimens with 0.04mm accuracy using an ocu-
lar micrometer calibrated with a stage micrometer. Carrhotus
xanthogramma overwinters mostly as a penultimate or ante-
penultimate (i.e., last instars right before the adult stage).
Thus, in autumn, it can be collected at a relatively uniform
ontogenetic stage (Markó and Keresztes 2014). Within the
genus Carrhotus, only C. xanthogramma occurs in Europe
(Nentwig et al. 2021), thus, immatures could not be confused
with other species. However, to check the taxonomic identity
of the individuals, we kept alive 10 randomly selected spiders
(5 females and 5 males) and fed them with D. hydei until
their nal molts. Finally, we preserved the raised and imma-
ture individuals in 70% ethanol, and using a binocular ster-
eomicroscope (Leica MZ6), we conrmed their taxonomic
identity after Nentwig et al. (2021). We determined the sex of
each immature (penultimate or antepenultimate) spider after
the following criteria—specimens with a slightly enlarged pal-
pal tarsus were considered as males, while specimens with a
pre-epigyne (undeveloped, thus, just visible external genitalia
of the females) but no enlarged palpal tarsus were considered
as being females. We excluded the data of the 2 specimens (1
from Site_Cs and 1 from Site_M1) that did not have either
enlarged palpal tarsus or pre-epigyne (i.e., earlier instars)
from our analyses. In accordance with the ndings of Markó
and Keresztes (2014), we found that most of the population
(60%) was female.
Behavioral assays
We assessed activity, risk-taking, and prey capture behavior
for each spider individual on 3 consecutive days and repeated
these assays weekly for 4 weeks in the same order. We per-
formed our assays between 0900 and 1600h (as that was the
most active period for these animals) based on the model spe-
cies’ circadian activity (Mező et al. 2019). We recorded the
assays using recording platforms (Kaiser RS 10 copy stands)
equipped with a camera (Panasonic HC-X920 HD), which
permitted the recording of the behavioral traits of several
individuals simultaneously. During the assays, we also contin-
uously recorded the temperature. To minimize human distur-
bance, only one person remained in the behavioral laboratory
to manipulate the spiders, handle the cameras, and check the
assays. During the experimental series, we measured each
individual’s different behaviors in the same Petri dish to avoid
generating unwanted random noise in a novel environment.
Measuring activity
First, we positioned the trays holding the Petri dishes with
spiders on the copy stands. After that, we waited 5min before
we started recording the activity of the spiders. For spiders,
the Petri dishes represented the environment in which we
measured their activity. The activity was recorded on every
experimental day for 30min. Later, we analyzed the record-
ings using the software ToxTrac version 2.84 (Rodriguez et al.
2018) to calculate the ‘Activity rate’, that is, the total time of
the spiders was engaged in locomotion divided by the assay’s
length.
Measuring risk-taking behavior
We positioned the plastic trays containing the spiders on the
copy stands and waited 5min before starting. Thereafter, to
expose the spiders to a physical stimulus that they may per-
ceive as a potential threat, we gently lifted the plastic tray
5cm high and dropped it back on the copy stand. To test the
short-term plasticity of risk-taking behavior, we repeated this
process 10min later, that is, we tested the risk taking of the
spiders twice (2 intra-assay trials) in one assay. Ten minutes
after the second “startling” process, we stopped recording the
behavior of spiders.
We calculated the “Freezing duration” in seconds as
the time between receiving the startling stimulus and the
moment when the spiders started moving again after freez-
ing. The tested individuals demonstrated 3 types of behavio-
ral responses to the stimulus. First, spiders froze immediately
after receiving the stimulus (majority, 91.6% of all cases);
second, before freezing, they ran about for a short time (≤
1s) (minority, 6.4%), while the third continued their activ-
ity without demonstrating any clear response to the stimulus
(rare, 2%). We omitted the time of running for the individu-
als who ran before freezing as it was negligible compared to
these individuals’ total time of freezing. Individuals that did
not respond to the stimulus by freezing received a zero value
(seconds), and individuals without any movement in the rst
or second intra-assay trial, received a value of 601 (seconds)
for the corresponding trial. An individual with a low Freezing
duration was considered risk tolerant (i.e., a bold one), while
the one with a high Freezing duration score was considered
risk averse (i.e., a shy one).
Measuring prey capture latency
For testing the spiders’ willingness to attack, we gently tipped
over the top of the Petri dishes, and using featherweight
entomology forceps, we inserted 3 ightless D. hydei in the
dishes quickly. After that, we immediately placed the dish on
the copy stand under the camera and recorded the spiders’
hunting. We continuously fed the spiders on a particular plas-
tic tray by the previous method and placed them under the
camera in the same order as they were on the tray. The whole
process was recorded, and after placing the last spider under
the camera, we recorded their activity for further 30min. We
checked the Petri dishes 24h later, counted the number of
live fruit ies, and removed both the living and the dead ones.
The same observer assessed all the video recordings and
noted the exact time when the spiders were placed under the
camera and when each spider successfully caught the rst, the
second, and the third fruit y. Then, we calculated the capture
latency (i.e., the time in seconds between the placement of
the spider under the camera and the capture of the prey) for
each of the ies that were captured within the rst 30min.
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

4Current Zoology
Individuals catching their prey just when the prey was offered
received a value of 1, 2, and 3 (seconds), and individuals who
caught the prey later than 30min of observation but within
24h scored as 1,801, 1,802, and 1,803s for catching the rst,
second, and third fruit y, respectively. However, most spiders
caught the rst prey very early, and many spiders caught the
third y later than 30min. Thus, we computed the arithme-
tic mean of the capture of the rst, second, and third prey
(hereinafter “Capture latency”) to characterize the spiders’
willingness to attack. In 3 cases, the spiders did not catch the
third y at all, thus, these observations were excluded from
the analyses.
Statistical analyses
All statistical analyses were performed within the R (v.3.5.3.)
statistical environment (https://www.R-project.org/). All the
analyzed data with their R script are provided in the les
entitled “Supplementary data” and “Supplementary codes,”
respectively. To nd out which predictors explain a signicant
amount of variance in the behavior of the spiders used in our
study, we built various linear (nonmixed and mixed) models
(described in more detail below) using the “lme4” R package
(https://CRAN.R-project.org/package=lme4). To statistically
control the unwanted effect of the temperature, rst, we built
Linear Models (LMs) where the behavioral traits were the
response variables, and the temperature was the predictor.
In the case of Freezing durations, we used log transforma-
tion to approach a normal distribution. After transformation,
although the Freezing duration was right-censored (601s; see
above), diagnostic statistics did not indicate a noticeable devi-
ation from the assumptions of linear models. In our further
models, we used the residuals of the abovementioned LMs as
response variables.
Analyzing the body parameters
First, we calculated a set of new variables, which made it pos-
sible to analyze the variation in body parameters of spiders
properly. “Mean mass” was the arithmetic mean of the mass
(in mg) of the spiders measured at the beginning (“Initial
mass”) and the end (“Final mass”) of the experiment. We
computed the “Relative mass change” as follows: (Final mass
– Initial mass)/ Initial mass × 100. To estimate the “Body con-
dition” of the individuals, we used the residuals of the regres-
sion of (log) Mean mass on (log) Prosoma width (Jakob et al.
1996).
Using LMs, we ran separate analyses with Initial mass,
Relative mass change, Prosoma width, and Body condition as
response variables, while the Site, Sex, and their interaction
were entered as predictor variables. In Relative mass change,
we used log transformation [log(Relative mass change + 100)]
to approach the normal distribution.
Analyzing the behavioral mean and the
relationship with the body parameters
To test what affects the behavioral means, we ran various
linear mixed-effects models (LMMs), separately, in which
the given behavioral variable was entered as a response var-
iable into the model. The initial LMMs comprised the Site,
Session number, and Sex as predictor variables and the spe-
cic ID numbers as a random factor. In the Freezing duration,
the model also comprised the Trial number (intra-assay tri-
als, a 2-level factor) as a predictor variable. The R syntaxes
of the initial (both sex included) models were as follows:
lmer(Activity rate res. ~ Site + Session nr. + Sex + (1|Spider.
ID)), lmer(Freezing duration res. ~ Site + Session nr. + Sex +
Trial nr. + (1|Spider.ID)), and lmer(Capture latency res. ~ Site
+ Session nr. + Sex + (1|Spider.ID)).
Preliminary results showed that Sex interfered with specic
body parameters and behavioral mean (see the Results). So,
we analyzed the behavioral data of females and males sep-
arately. Furthermore, we entered 2 new predictor variables,
Body condition and Prosoma width, into the sex-specic
mixed models. For model summaries, see the Supplementary
Material.
LMMs are powerful tools for analyzing complex datasets,
for example, in behavioral ecology or evolution, since model
estimates are usually robust to violations of distributional
assumptions (Schielzeth et al. 2020). Nevertheless, before
interpreting the model outcomes, we performed numerous
model diagnostic statistics to avoid misleading results based
on statistical artifacts. Following the recommendations of
Garamszegi et al. (2014), we checked the assumptions about
the distribution of residuals (normality and homogeneity),
and we calculated variance ination factors (VIF) to examine
the issues about multicollinearity. In the case of LMs, based
on our models, we computed the η2 values (with 90% CIs due
to the one-tailed tests) for our predictors using the R packages
“sjstats” (https://CRAN.R-project.org/package=sjstats) and
“MBESS” (https://CRAN.R-project.org/package=MBESS).
Regarding the LMMs, following the procedure of Garamszegi
et al. (2014), we estimated the statistical signicance of the
focal predictors using the likelihood ratio tests (full models
versus restricted models without the given predictor), where
the signicance was described by the probability function
ofthe chi-square distribution (at df = 1). Finally, we calcu-
lated the effect sizes (Cramer’s V) with 95% CIs for each focal
relationship (Garamszegi et al. 2014).
Calculating behavioral repeatability and estimating
behavioral syndrome structure
As every spider was tested 4 times, we calculated their behav-
ior’s repeatability (as a proxy of interindividual consist-
ency) and evaluated the correlational relationships between
the measured behavioral trait variables. For calculating
these estimates, we ran multivariate Markov Chain Monte
Carlo Generalized Linear Mixed Models (MCMCglmms)
for females and males separately, using the Bayesian
“MCMCglmm” R package (Hadeld 2010). In these mul-
tivariate models, our mean-centered behavioral variables
(Activity rate, Freezing duration in the rst and second trials,
and Capture latency) were (together) the response variables,
while specic IDs were coded as random effect. In order to
statistically control the putative effect of temporal repetition,
we entered the Session number as a predictor variable (as
a covariate) into the models. Applying such a model struc-
ture was necessary because (consistency) repeatability esti-
mates were calculated (after Nakagawa and Schielzeth 2010)
using the variance components obtained from these models,
and ignoring time-related change might lead to biased esti-
mates of repeatability (Biro and Stamps 2015). Following
the approach of Dingemanse and Dochtermann (2013), we
used these multivariate models to decompose phenotypic (co)
variances into inter- and intraindividual components and,
as Dingemanse and Dochtermann (2013) recommended, we
used only the interindividual (co)variance components to
evaluate the relationship between each measured behavioral
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

Mezőfi et al. · Dimorphic strategies among immatures of a jumping spider 5
trait variable (i.e., between-individual correlations were cal-
culated). In these models, we set the intraindividual covari-
ance to 0 since the different behavioral traits were not tested
at the same time (Dingemanse and Dochtermann 2013). We
used a weakly informative inverse gamma prior and speci-
ed our MCMCglmms with 1 300 000 iterations, 300 000
iterations “burn-in” and a thinning interval of 1,000. Both
for the repeatability estimates and the correlation coefcients
(effect sizes), 95% credible intervals were calculated based on
the posterior mode of their estimates. In order to evaluate
the (dis)similarity of the behavioral syndrome structure of the
sexes, we performed Mantel’s test on the sex-specic matrices
of the posterior correlation coefcients using the “mantel”
function of the R package “ecodist” (https://CRAN.R-project.
org/package=ecodist). Following the methods of Royauté et
al. (2015), we calculated ∆r, the average difference in pairwise
correlations between sexes, and, as Royauté et al. (2015), ∆r
values are interpreted based on the following scale: 0 < |∆r|
< 0.2, no to low effect; 0.2 < |∆r| < 0.5, medium effect; 0.5 <
|∆r|, strong effect. Regarding the behavioral correlations with
the highest |∆r|, we illustrate these relationships using the
posterior modes of our random effects (i.e., best linear unbi-
ased predictors—BLUPs) from our multivariate models after
Houslay and Wilson (2017). Finally, we determined statistical
support for covariances (correlations) by differences in devi-
ance information criteria (DIC) values. Thus, as Dingemanse
and Dochtermann (2013) proposed, we compared the DIC of
constrained (inter- and intraindividual covariances were set to
0) and unconstrained models (only intraindividual covariance
was set to 0) for a better t. Signicant behavioral correla-
tions (based on nonoverlap of the CI with 0) were accepted
when 5< DIC constrained − DIC unconstrained (Kralj-Fišer et
al. 2017). For model summaries and detailed results, see the
Supplementary Material.
Analyzing the intraindividual behavioral variability
As the behavioral trait variables were measured multiple
times, we computed the residual individual standard devia-
tion (riSD) as a proxy of IIV. IIV values refer to behavioral
predictability in the following way: The higher the value of
IIV, the lower the behavioral predictability. Therefore, we
calculated the riSD values after the procedure proposed by
Stamps et al. (2012). First, we tted LMMs in which we
incorporated a temporal reaction norm. In these models, the
response variable was the given behavioral trait variable, the
xed effect was the Session number (i.e., time, altogether 4
sessions), and the random effects were represented by Session
number as a random slope and by specic ID as a random
intercept (R syntax: lmer(Behavioural trait res. ~ Session nr.
+ (Session nr.|Spider.ID))). Then we extracted the residuals
of the models and computed the riSD index values. Finally,
we entered the riSD values (as response variables) into LMs
where the predictor variables were represented by the Site,
Sex, and their interaction.
Results
Variability in body parameters
Based on the η2 values, the largest amount of variance was
accounted for by Sex concerning the Initial mass (21.8 %)
and Prosoma width (16.4 %); that is, immature females had
lower mass and a narrower prosoma than immature males
in C. xanthogramma (Table 1 and Figure 1). We also found
sex-specic variance in the Relative mass change of the imma-
ture spiders. The body mass changed differently, as females
gained (mean ± SD: 6.67±16.08 %; N = 53) while males
maintained or slightly lost (mean ± SD: −0.75 ± 12.64 %;
N = 36) body mass during the study (Table 1 and Figure 1).
Generally, larger individuals lost while smaller ones gained
mass; thus, Initial mass negatively correlated with Relative
mass change (Pearson’s r = −0.64; P < 0.001). Furthermore,
we detected interpopulation variation in the Initial mass and
Body condition but not in the other body parameters of the
immature individuals (Table 1 and Supplementary Figure S1).
Variability in behavior
Robust sex-specic differences were detected in C. xantho-
gramma regarding the Activity rate and Freezing duration
(Table 2). Males were more active and took a higher risk
(shorter Freezing duration) toward a potentially threatening
abiotic stimulus than the females (Figure 2).
Furthermore, in contrast with the females, males’ activity
increased with time, and the corresponding effect size and the
associated CIs indicated medium to strong effects (Table 2
and Supplementary Figure S2). Regarding the Freezing dura-
tion, though both females and males reacted more sensitively
(small to strong effect) to the second intra-assay startling stim-
uli (Figure 2), in the long term, only the females reacted to the
threatening abiotic stimuli more and more sensitively (strong
effect, the Freezing duration increased with time), while males
did not show such pattern (Table 2 and Supplementary Figure
S2). Finally, male individuals tended to catch the offered prey
faster with the increasing number of test sessions (marginal
relationship with a small to strong effect, the Capture latency
decreased with time) (Table 2 and Supplementary Figure S2).
We detected a marginally signicant positive relationship
(with a small effect) between the Prosoma width and the
Activity rate in males but not in females (Table 2 and Figure
3). We found in both sexes that Freezing duration was related
positively (small to strong effect) to the Body condition and
related negatively (only marginally with a small to medium
effect) to the Prosoma width, in both cases with a stronger
effect in males than females (Table 2 and Figure 3). A sig-
nicant negative relationship (with a small to strong effect)
was found between the Capture latency and Prosoma width
in males, that is, larger individuals tended to catch the prey
faster (Table 2 and Figure 3). No such relationship was
found in females (Table 2 and Figure 3). Finally, our anal-
yses revealed a small to strong effect of the collecting site
on the Freezing duration (regarding females and males) and
the Capture latency (regarding only females) (Table 2 and
Supplementary Figure S3).
Behavioral repeatability, correlation structure, and
IIV
Both females and males showed behavioral repeatability
regarding all of the measured behavioral traits (Activity
rate, Freezing duration in the rst and second trials, and
Capture latency) (Table 3). In females, we found evidence
for a behavioral syndrome involving all measured behav-
ioral traits: Activity rate was negatively associated with
Freezing duration (a measure of risk taking) and with
Capture latency, while a positive relationship was found
between the latter 2 behavioral traits (Figure 4). In males,
we found a signicant (negative) correlation only between
the Activity rate and Freezing duration (measured in the
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

6Current Zoology
Table 1. Site- and sex-specific differences in body parameters in Carrhotus xanthogramma immatures (N = 90).
Response variable Predictors df Sum of Sq. Mean of Sq. F-value P-value Effect size (η2)* CIlower CIupper
Initial mass (mg) Site 2 286.54 143.27 3.491 0.035 0.059 0.003 0.166
Sex 1 1,050.31 1,050.31 25.591 <0.001 0.218 0.111 0.350
Site:Sex 2 42.02 21.01 0.512 0.601 0.009 NA 0.058
Residuals 84 3,447.50 41.04 NA NA NA NA NA
Relative mass change Site 2 0.08 0.04 2.299 0.107 0.050 NA 0.133
Sex 1 0.09 0.09 5.046 0.027 0.054 0.003 0.152
Site:Sex 2 0.00 0.00 0.010 0.990 0.000 NA NA
Residuals 83 1.47 0.02 NA NA NA NA NA
Prosoma width (mm) Site 2 0.32 0.16 2.979 0.056 0.055 NA 0.152
Sex 1 0.95 0.95 17.766 <0.001 0.164 0.066 0.290
Site:Sex 2 0.03 0.02 0.272 0.762 0.005 NA 0.040
Residuals 84 4.50 0.05 NA NA NA NA NA
Body condition Site 2 0.41 0.21 6.179 0.003 0.123 0.028 0.230
Sex 1 0.05 0.05 1.358 0.247 0.014 NA 0.084
Site:Sex 2 0.10 0.05 1.433 0.244 0.028 NA 0.101
Residuals 84 2.81 0.03 NA NA NA NA NA
Note: Effect size values are bolded if the related CI does not overlap with zero.
*Small effect: 0.02 ≤ η2 < 0.13; Medium effect: 0.13 ≤ η2 < 0.26; Large effect: 0.26 ≤ η2.
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

Mezőfi et al. · Dimorphic strategies among immatures of a jumping spider 7
second trial) (Figure 4). Based on the differences in DIC
values (female ∆DIC = 30.844; male ∆DIC = 9.945), the
unconstrained models were substantially more well sup-
ported by the data than the models in which the interindi-
vidual covariances were set to 0 (Table 3), thus signicant
correlations were accepted. However, the Mantel test did
not provide clear evidence for sex-specic behavioral syn-
dromes (Figure 4) as a marginally signicant overall cor-
relation (Mantel test 1 000 000 permutations: r = 0.924;
P = 0.084) was found between the behavioral correlation
matrices. Nevertheless, the highest sex-related differences
in correlation estimates (∆r) were found in the following
behavioral trait combinations: Activity rate versus Capture
latency (∆r = 0.300) and Freezing duration (measured in
the second trial) versus Capture latency (∆r = −0.247)
(Table 3 and Figure 5). None of the predictors (Site, Sex,
and their interaction) had an effect on IIV (i.e., behavioral
predictability) (Table 4).
Discussion
We tested for sex-specic differences in the essential t-
ness-related traits and their correlation structure in C. xan-
thogramma immatures. In general, we found sex-specic
differences in specic body parameters and behavioral traits
between females and males before reaching their reproductive
stage, suggesting different life-history strategies.
Testing the sexual dimorphism in specic body parame-
ters, we found that the immature C. xanthogramma males
had a wider prosoma and greater initial body mass than the
female conspecics (Figure 1). Our results contrasted with
the general pattern among spiders regarding their body size
because adult spider males are often smaller than females
(Head 1995). Though a few exceptions with a reversed pat-
tern exist, for example, in salticids (Prenter et al. 1999; Lim
and Li 2004), the adult males of the species studied here are
also smaller than the adult female conspecics (Kim and Lee
2014). One possible explanation for our ndings could be
Figure 1. Sexual differences in body parameters (A—Initial mass; B—Relative mass change; C—Prosoma width; D—Body condition) in Carrhotus
xanthogramma immatures. The distance between the box bottom (first quartile) and top (third quartile) corresponds to the interquartile, while the
whisker shows the nonoutlier range. The red diamond and the bold horizontal line indicate the mean and median values, respectively. Data points were
jittered horizontally. Effect size (η2) and related 90% CI are displayed on the corresponding panel. Effect size values are bolded if CI does not overlap
with zero.
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

8Current Zoology
Table 2. Results from linear mixed models, in which the behavioral traits of Carrhotus xanthogramma immatures (N = 90) were the response variables.
Response variable Model Predictors Likelihood Chi2 PLikelihood ratio Effect size (Cramer’s V)* CIlower CIupper
Activity rate Both sex included Site 1.087 0.581 0.110 −0.100 0.310
Sex 8.259 0.004 0.303 0.102 0.480
Session nr. 7.723 0.006 0.293 0.091 0.471
Only females Site 0.198 0.906 0.061 −0.211 0.323
Session nr. 0.500 0.480 0.096 −0.176 0.355
Body condition 0.666 0.415 0.111 −0.162 0.368
Prosoma width 0.216 0.642 0.063 −0.208 0.326
Only males Site 2.634 0.268 0.271 −0.064 0.550
Session nr. 12.110 <0.001 0.580 0.311 0.763
Body condition 0.297 0.586 0.091 −0.245 0.407
Prosoma width 3.148 0.076 0.296 −0.036 0.569
Freezing duration Both sex included Site 9.604 0.008 0.327 0.128 0.500
Sex 7.067 0.008 0.280 0.078 0.461
Session nr. 26.869 <0.001 0.546 0.383 0.677
Trial nr. 16.234 <0.001 0.425 0.239 0.581
Only females Site 6.303 0.043 0.342 0.081 0.558
Session nr. 44.182 <0.001 0.905 0.840 0.944
Trial nr. 7.592 0.006 0.375 0.119 0.584
Body condition 6.599 0.010 0.350 0.090 0.565
Prosoma width 2.720 0.099 0.225 −0.046 0.464
Only males Site 14.689 <0.001 0.639 0.393 0.800
Session nr. 0.000 0.995 0.001 −0.328 0.330
Trial nr. 9.617 0.002 0.517 0.227 0.723
Body condition 10.311 0.001 0.535 0.251 0.735
Prosoma width 3.504 0.061 0.312 −0.018 0.581
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

Mezőfi et al. · Dimorphic strategies among immatures of a jumping spider 9
that the female sample collected in autumn comprised pro-
portionally more antepenultimate individuals than the male
population. A previous study (Markó and Keresztes 2014)
detected a sex-biased temporal asynchrony in the population
structure of C. xanthogramma, implying differential timing of
maturity. It was observed that adult males were more numer-
ous in April while females in May (Markó and Keresztes
2014), suggesting that adult males reached adulthood and
became ready for reproduction earlier than females (i.e., C.
xanthogramma shows protandry), which mating strategy
could result in female-biased sexual size dimorphism (smaller
males relative to females) in spiders (Maklakov et al. 2004).
Also, in our reared individuals, the mean developmental time
(calculating from the time of collection, mean ± SD, in days)
was shorter in males than females (56.2± 8.93; N = 5 vs.
102.4±27.57; N = 5).
It seems that body parameters concerned are subject to
opposing evolutionary forces. On one hand, selection acts
for protandry, that is, rapidly developing males (with a
decreased male size) have higher tness, especially in web-
builder spiders (Head 1995). But on the other hand, a larger
size may also increase male tness as a heavier male is usu-
ally more successful than a smaller one in a direct compet-
itive context (Kasumovic and Andrade 2009; Kasumovic
et al. 2011). Our results (Figure 1) might be explained by
the fact that the primary objective of adult males is to nd
a mate for copulation as soon as possible, thus they often
do not feed at all or only occasionally (Givens 1978; Foelix
2011). Therefore, accumulating additional nutrient reserves
(i.e., greater body mass) before maturity might provide
them with an adaptive advantage. Furthermore, mating
success is often associated with male size (Sivalinghem et al.
2010; Golobinek et al. 2021), and for example, in another
jumping spider Phidippus clarus, heavier males were more
successful in intraspecic male–male competition (Elias
et al. 2008). As sex-specic selection forces favor males
with larger body size (Fernández-Montraveta and Moya-
Laraño 2007), to maximize their tness outputs, accumu-
lating nutrient reserves before maturity could be crucial for
males. In contrast, females with a relatively longer lifespan
than males may, over the longer adulthood, compensate for
their slower rate of weight gain during immaturity. This
temporal asynchrony in growth patterns between females
and males could be sourced by the differential reproductive
investment, affecting behavioral and feeding patterns.
Consistent individual differences in behavior are linked
with consistent individual differences in energy metabolism
(Biro and Stamps 2010; Holtmann et al. 2017a). In our study,
males were more active and bolder than females (Table 2 and
Figure 2), which may be explained by the different physio-
logical backgrounds of the sexes. Spider males usually show
higher metabolic rates than females (Schmitz 2004; Walker
and Irwin 2006; but see Kotiaho 1998). Higher level of met-
abolic activity often associates with a shorter lifespan (Réale
et al. 2010; Kralj-Fišer and Schuett 2014). Similarly to our
results, Chapman et al. (2013) found that rock pool prawn
(Palaemon elegans, Palaemonidae) males were more active
and bolder than the females, which usually live for twice as
long as males. Field observations of Markó and Keresztes
(2014) implied that C. xanthogramma males mature earlier
and have a shorter lifespan than females. These results suggest
that immature males’ higher activity and boldness stem partly
from their assumed higher rate of metabolism.
Response variable Model Predictors Likelihood Chi2 PLikelihood ratio Effect size (Cramer’s V)* CIlower CIupper
Capture latency Both sex included Site 8.650 0.013 0.310 0.110 0.486
Sex 2.335 0.126 0.161 −0.048 0.356
Session nr. 1.091 0.296 0.110 −0.099 0.310
Only females Site 4.370 0.112 0.284 0.018 0.513
Session nr. 0.029 0.864 0.023 −0.246 0.289
Body condition 0.937 0.333 0.132 −0.141 0.386
Prosoma width 0.388 0.533 0.085 −0.187 0.345
Only males Site 3.355 0.187 0.305 −0.026 0.576
Session nr. 3.335 0.068 0.304 −0.027 0.575
Body condition 0.066 0.798 0.043 −0.290 0.366
Prosoma width 7.976 0.005 0.471 0.168 0.692
Note: Signicance levels (P) and effect sizes (Cramer’s V) originated from the corresponding likelihood ratio test that compared the model t of the full model and the reduced model after excluding the given
predictor. The 95% CIs around effect sizes originated from the parametric bootstrap performed on data simulated according to the model’s predictions. Effect size values are bolded if the related CI does not
overlap with zero.
*Small effect: 0.1 ≤ V < 0.3; Medium effect: 0.3 ≤ V < 0.5; Large effect: 0.5 ≤ V.
Table 2. Continued
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

10 Current Zoology
In this study, immature C. xanthogramma males showed
higher activity than immature females, which results were
in contrast with a previous study, reporting that the adult
males were less active than adult females (Mező et al.
2019). One possible explanation for this pattern might
be that the C. xanthogramma individuals change their
strategy with maturation: Immature males forage much
more intensively and accumulate nutrient reserves, while
immature females, to protect their condition, remain more
cautious and less active. In contrast, mature males feed
only occasionally and use up their nutrient reserves, while
mature females feed actively to gain more and more energy
to increase their egg production. Ontogenic behavioral
shift was observed, for example, in the Sydney funnel-web
spider (Atrax robustus, Atracidae), where adult females
reacted more intensively to an aversive stimulus than the
juveniles (Duran et al. 2022). Furthermore, behavioral
changes were also documented among ontogenetic stages
in other arthropod taxa, suggesting that individuals could
respond exibly according to their current physiological
requirements (e.g., Gyuris et al. 2012; Niemelä et al. 2012;
Kralj-Fišer and Schuett 2014).
In males, both the activity and the willingness to attack a
prey increased (but the latter was only marginally) with the
repeated test sessions (Table 2 and Supplementary Figure S2).
We observed that males did not gain body weight in the course
of this study. This was probably because the feeding regime (3
Drosophila/week) followed did not meet the males’ (possibly
higher) nutritional needs (see Figure 1). Hunger can increase
activity (Walker et al. 1999), and activity is closely linked with
the metabolic rate (Schmitz 2004; Walker and Irwin 2006).
Thus, the physiological differences and environmental con-
ditions could be responsible for the males’ increased activity
and the willingness to attack modulated by the individuals’
current hunger level. A self-excitation process could be gener-
ated in males due to the strong link between physiology and
behavior. Their assumed higher (base) metabolic rate (see pre-
viously) would result in a higher level of hunger and induce
a higher level of activity to nd prey, which would further
increase the metabolic rate and generate other physiological
and behavioral consequences.
A prior negative experience can reduce the degree of bold-
ness (Frost et al. 2007). Therefore, accordingly, both females
and males increased their latency to movement initiation in
the risk-taking assay (in the short term, see Table 2 and Figure
2). However, in contrast to males, the Freezing duration of
females was also signicantly increased with the number of
test sessions (Table 2 and Supplementary Figure S2). A similar
(but insignicant) trend was also observed in Philodromus
albidus (Philodromidae) females (Michalko et al. 2017),
which could be considered as an effect of sensitization to a
threatening stimulus (Blumstein 2016), causing the manifes-
tation of a risk-averse behavior among females.
Several studies (e.g., Royauté et al. 2014; Ingle et al. 2018;
Michalko and Řežucha 2018) have shown a relationship
between body parameters and behavior. We also found a sig-
nicant relationship between Freezing duration and condition
both in females and males—individuals in better condition
tended to be more risk averse (i.e., showing longer Freezing
duration) (Table 2 and Figure 3). These results support the
“asset-protection principle,” which proposes that an individ-
ual with a better body condition should be more risk averse
than an individual with a poorer body condition (Clark 1994;
Kralj-Fišer and Schuett 2014; Moran et al. 2021). Royauté et
al. (2014) also observed decreasing boldness with increasing
body condition in the jumping spider, E. militaris, although
the direction of the relationship between boldness and con-
dition may depend on the actual experimental or ecological
context (Johnson and Sih 2007). Besides this, the response to a
threatening stimulus could be based mainly on some individ-
ual-specic traits, such as body size and sex (i.e., being male
or having a larger body size could initiate a bolder response)
(Table 2; Figure 3). Regarding Capture latency, the negative
relationship between behavior and body size could play an
essential role in the hunting decisions but only in males (Table
2; Figure 3).
In this study, an interpopulation variation was detected
for tness-related traits such as Initial mass, Body condi-
tion, Freezing duration, and Capture latency (Tables 1 and
2; Supplementary Figures S1 and S3). Supporting our results,
Michalko and Dvoryankina (2019) recently communicated
that certain traits of another spider species could vary along
Figure 2. Sexual differences in behavioural traits (A—Activity rate; B—Freezing duration; C—Capture latency) of Carrhotus xanthogramma immatures.
The distance between the box bottom (first quartile) and top (third quartile) corresponds to the interquartile, while the whisker shows the nonoutlier
range. The red diamond and the bold horizontal line indicate mean and median values, respectively. Data points were jittered horizontally. Effect size
(Cramer’s V) and related 95% CI are displayed on the corresponding panel. Effect size values are bolded if CI does not overlap with zero. AR, Activity
rate; FD, Freezing duration; CL, Capture latency.
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

Mezőfi et al. · Dimorphic strategies among immatures of a jumping spider 11
a spatial gradient even within a single orchard. Individuals
from the orchard centers can be larger than those from the
edges. Developmental diet quality can affect the condition
(Taylor et al. 2011) and, according to the meta-analysis by
Moran et al. (2021), a low-quality diet increases boldness in
several ecological contexts. Thus, the differences in the quali-
tative or quantitative composition of the arboreal arthropod
(potential prey) assemblages of the different sampling sites
could explain the detected interpopulation variations. Besides
this, personality-matching habitat choice (Holtmann et al.
2017b) cannot be excluded.
Sex-related differences in repeatability, behavioral correla-
tions, IIV, and their mechanistic background have not been
adequately studied in the invertebrate literature. Therefore,
more studies are needed to ll these gaps in knowledge. Here,
the tested behavioral traits showed moderate repeatability
Figure 3. The linear relationship between a tested behavioral trait and body parameters such as the Body condition (A, C, E) and Prosoma width (B,
D, F) in female and male Carrhotus xanthogramma immature individuals. Solid lines indicate significant (PLikelihood ratio < 0.05; effect size CI excludes 0)
relationships, while dashed lines indicate marginally significant (PLikelihood ratio between 0.05 and 0.1) relationships. AR, Activity rate; FD, Freezing duration;
CL, Capture latency. Colors represent the sexes (orange—female, blue—male).
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

12 Current Zoology
Table 3. Results of multivariate MCMCglmms testing for interindividual correlations (estimate with 95% CIs) of the measured behavioral traits (AR,
Activity rate; FD_I and FD_II, Freezing duration in the first and second trial; CL, Capture latency) in Carrhotus xanthogramma immatures by sex.
Females
Unconstrained model DIC 2242.34
Constrained model DIC 2273.184
AR FD_I FD_II CL
AR 0.370 (0.244; 0.512)
FD_I −0.621 (−0.815; −0.219) 0.226 (0.143; 0.381)
FD_II −0.622 (−0.808; −0.269) 0.664 (0.423; 0.858) 0.360 (0.224; 0.475)
CL −0.501 (−0.703; −0.060) 0.407 (0.036; 0.705) 0.555 (0.096; 0.748) 0.310 (0.202; 0.481)
Males
Unconstrained model DIC 1508.846
Constrained model DIC 1518.791
AR FD_I FD_II CL
AR 0.337 (0.207; 0.519)
FD_I −0.376 (−0.749; 0.027) 0.232 (0.128; 0.419)
FD_II −0.412 (−0.773; −0.044) 0.528 (0.071; 0.801) 0.294 (0.157; 0.458)
CL 0.048 (−0.525; 0.359) 0.451 (−0.011; 0.765) 0.299 (−0.255; 0.654) 0.385 (0.201; 0.528)
∆r (male r − female r)
AR FD_I FD_II CL
AR
FD_I −0.012
FD_II 0.141 −0.072
CL 0.300 0.112 −0.247
Note: Repeatability estimates (with 95% CIs) of each behavioral trait were shown in the diagonals. ∆r represents the average effect size of the difference
in correlation coefcients between sexes. Unconstrained (covariance within individuals was set to 0) and constrained models (covariances between and
within individuals were set to 0) were compared. Correlations and repeatability estimates were calculated from the unconstrained models and signicant
correlations (based on nonoverlap of the CI with 0) were accepted when 5< DIC constrained − DIC unconstrained. Signicant correlations and repeatability
estimates are bolded. |∆r| values >0.2 are indicated in bold. For model summaries and detailed results, see the Supplementary Material.
Figure 4. The correlation structure (A—female; B—male) of the measured behavioral traits in Carrhotus xanthogramma immatures. Results of
multivariate MCMCglmms testing for interindividual correlations. AR, Activity rate; FD_I and FD_II, Freezing duration in the first and second trial; CL,
Capture latency.
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

Mezőfi et al. · Dimorphic strategies among immatures of a jumping spider 13
in both sexes (Table 3), reecting consistent interindividual
behavioral differences, which ndings t the concept of ani-
mal personality (Réale and Dingemanse 2012). A meta-ana-
lytic study (Bell et al. 2009) found that among invertebrates,
the behavior of females is more repeatable than that of con-
specic males. However, the direction of the sex differences
in repeatability may depend on the specic behavior being
considered (Bell et al. 2009). In the present study, sex-specic
differences were not found concerning the repeatability esti-
mates as their credible intervals were overlapped (Table 3).
Sex-specic behavioral syndromes were reported in cer-
tain invertebrate taxa, for example, in eld crickets (Gryllus
integer, Gryllidae), and it seems that these syndromes were
partially driven by genetics (Royauté et al. 2021). According
to Royauté et al. (2021), these kinds of sex-specic behavio-
ral syndromes allow the independent evolution of behavio-
ral dimorphism. Though we did not nd clear evidence for
sex-specic syndromes, the links between functionally dif-
ferent behavioral traits suggest that the sexes differ slightly
in the structure of the behavioral syndrome (Figures 4 and
5). In both sexes, individuals with higher activity tended
to be more risk tolerant (i.e., have shorter Freezing dura-
tion), while the relationship between the rst and second
trials of the Freezing duration reected high (short term)
intraindividual consistency. More active and bolder females
also captured prey faster than the shyer ones, but such a
relationship was not found in males (Figures 4 and 5). The
same relationship between Activity rate and Capture latency
for females, but not for males, was found in E. militaris
(Royauté et al. 2015). Consistent interindividual differences
might coexist with developmental plasticity (Kralj-Fišer and
Schneider 2012). Here, both females and males showed con-
sistent interindividual behavioral differences, but we found
weaker or undetectable relationships between certain behav-
ioral traits in males compared to the females. This might
be explained that sexual maturation can often change the
males’ appearance and behavior (Sullivan and Morse 2004;
Framenau 2005; Cordellier et al. 2020). Hence, some behav-
ioral traits of C. xanthogramma males might be more plas-
tic during development than the behavior of females. In our
study, the relationships between the examined traits were
usually stronger in females (Table 3 and Figure 4), which
contrasted with the general patterns observed in a meta-an-
alytic study (i.e., tend to be stronger in males) focusing on
similar behavioral traits in vertebrates (Garamszegi et al.
2012).
The sexes showed similar IIV regarding all of the measured
behavioral traits (Table 4). However, using right-censored
data (e.g., Freezing duration) might lead to biased estimates
of IIV (Stamps et al. 2012), which could eventually result
in that sex-specic differences in behavioral predictability
remain hidden.
As sexes differ in their life-history optima and reproduc-
tive role, unsurprisingly, sexual dimorphism can be observed
in certain tness-related traits and their complex physiologi-
cal, behavioral, and genetic backgrounds (Hämäläinen et al.
2018; Tarka et al. 2018). The present study implied that even
immature females and males might have different life-history
strategies with different, sex-specic consequences. Thus,
before maturation, females tend to be less active, take less
risk, and more sensitive to alarming stimuli than males. In
contrast, males could follow a “live fast, die young” life-his-
tory strategy. Immature males forage much more intensively
to increase their body size faster because selection forces favor
the heavier and larger males after maturation.
Acknowledgments
The authors would like to thank Kristóf Bársony, Rebeka
Saliga, and Anna Sándor for their assistance, and Petr Dolejš
for providing literature. We also thank Mathieu Videlier and
the anonymous reviewers for their constructive comments.
Figure 5. Sexual differences in interindividual behavioral correlations (A—Activity rate vs. Capture latency; B—Freezing duration in the second trial vs.
Capture latency). Plots represent linear relationships between behavioral traits with highest difference in correlation estimates between sexes (Δr, see
Table 3.). Values (best linear unbiased predictors—BLUPs) were extracted from multivariate MCMCglmms using the posterior modes of random effects
(specific IDs). Solid lines indicate significant (effect size CI excludes 0) relationships. Colors represent the sexes (orange—female, blue—male). Note the
absence of significant relationships between behavioral traits in males.
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

14 Current Zoology
Table 4. Results from linear models, in which the riSD values (a proxy of IIV) of behavioral trait variables of Carrhotus xanthogramma immatures (N = 90) were the response variables.
Response variable Predictors df Sum of Sq. Mean of Sq. F-value P-value Effect size (η2)* CIlower CIupper
Activity rate riSD Site 2 24.95 12.47 0.327 0.722 0.008 NA 0.044
Sex 1 4.43 4.43 0.116 0.734 0.001 NA 0.039
Site:Sex 2 33.33 16.66 0.437 0.647 0.010 NA 0.053
Residuals 84 3,200.52 38.10 NA NA NA NA NA
Freezing duration,
rst trial riSD
Site 2 0.49 0.25 0.793 0.456 0.018 NA 0.073
Sex 1 0.06 0.06 0.195 0.660 0.002 NA 0.045
Site:Sex 2 0.18 0.09 0.293 0.747 0.007 NA 0.041
Residuals 84 26.16 0.31 NA NA NA NA NA
Freezing duration,
second trial riSD
Site 2 0.15 0.08 0.369 0.693 0.009 NA 0.048
Sex 1 0.00 0.00 0.002 0.963 0.000 NA NA
Site:Sex 2 0.33 0.17 0.795 0.455 0.018 NA 0.073
Residuals 84 17.41 0.21 NA NA NA NA NA
Capture latency riSD Site 2 41 785.46 20 892.73 1.432 0.245 0.033 NA 0.104
Sex 1 4,745.51 4,745.51 0.325 0.570 0.004 NA 0.055
Site:Sex 2 24 643.61 12 321.81 0.844 0.434 0.020 NA 0.078
Residuals 81 1 182 056.53 14 593.29 NA NA NA NA NA
*Small effect: 0.02 ≤ η2 < 0.13; Medium effect: 0.13 ≤ η2 < 0.26; Large effect: 0.26 ≤ η2.
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

Mezőfi et al. · Dimorphic strategies among immatures of a jumping spider 15
Funding
This study was supported by the National Research,
Development, and Innovation Ofce of Hungary (K112743).
Ethics Statement
Our experiments comply with the ASAB/ABS guidelines for
the use of animals. We performed experiments with arthro-
pod species that are not protected, and no permission from an
ethical committee was needed. We minimized the effect on the
population size of the used spiders by reducing sample sizes
while maintaining sufcient statistical power.
References
Albín A, Aisenberg A, Simó M, Dolejš P, 2018. Sexual dimorphism in the
spinning apparatus of Allocosa senex (Araneae: Lycosidae), a wolf
spider with a reversal in typical sex roles. J Arachnol 46:207–213.
Alonzo SH, Kindsvater HK, 2008. Life-history patterns. In: Fath B,
editor. Encyclopedia of Ecology. 2nd edn. Vol 3. Oxford: Elsevier,
130–136.
Bell AM, Hankison SJ, Laskowski KL, 2009. The repeatability of
behaviour: a meta-analysis. Anim Behav 77:771–783.
Biro PA, Stamps JA, 2010. Do consistent individual differences in met-
abolic rate promote consistent individual differences in behavior?
Trends Ecol Evol 25:653–659.
Biro PA, Stamps JA, 2015. Using repeatability to study physiological
and behavioural traits: ignore time-related change at your peril.
Anim Behav 105:223–230.
Blumstein DT, 2016. Habituation and sensitization: new thoughts
about old ideas. Anim Behav 120:255–262.
Carducci JP, Jakob EM, 2000. Rearing environment affects behaviour
of jumping spiders. Anim Behav 59:39–46.
Chang CC, Connahs H, Tan ECY, Norma-Rashid Y, Mrinalini D, et al.,
2020. Female spider aggression is associated with genetic under-
pinnings of the nervous system and immune response to pathogens.
Mol Ecol 29:2626–2638.
Chang CC, Klomp DA, Norma-Rashid Y, Li D, 2019. Consistency in
boldness expression varies with ecological context in a jumping
spider. Ethology 125:724–732.
Chapin KJ, 2017. Arthropod life history. In: Vonk J, Shackelford T,
editors. Encyclopedia of Animal Cognition and Behavior. Cham:
Springer.
Chapman BB, Hegg A, Ljungberg P, 2013. Sex and the syndrome: indi-
vidual and population consistency in behaviour in rock pool prawn
Palaemon elegans. PLoS ONE 8:e59437.
Clark CW, 1994. Antipredator behavior and the asset-protection prin-
ciple. Behav Ecol 5:159–170.
Cordellier M, Schneider JM, Uhl G, Posnien N, 2020. Sex differ-
ences in spiders: from phenotype to genomics. Dev Genes Evol
230:155–172.
Dingemanse NJ, Dochtermann NA, 2013. Quantifying individual var-
iation in behaviour: mixed-effect modelling approaches. J Anim
Ecol 82:39–54.
Dingemanse NJ, Kazem AJN, Réale D, Wright J, 2010. Behavioural
reaction norms: animal personality meets individual plasticity.
Trends Ecol Evol 25:81–89.
Dingemanse NJ, Réale D, 2005. Natural selection and animal personal-
ity. Behaviour 142:1159–1184.
DiRienzo N, Montiglio PO, 2016. The contribution of developmental
experience vs. condition to life history, trait variation and individ-
ual differences. J Anim Ecol 85:915–926.
Duran LH, Wilson DT, Rymer TL, 2022. Behaviour of the Sydney fun-
nel-web spider Atrax robustus over different contexts, time, and
stimuli. Toxicon: X 13:100093.
Elias DO, Kasumovic MM, Punzalan D, Andrade MCB, Mason AC,
2008. Assessment during aggressive contests between male jumping
spiders. Anim Behav 76:901–910.
Fernández-Montraveta C, Moya-Laraño J, 2007. Sex-specic plasticity
of growth and maturation size in a spider: implications for sexual
size dimorphism. J Evol Biol 20:1689–1699.
Foelix R, 2011. Biology of Spiders. 3rd edn. New York, USA: Oxford
University Press.
Framenau VW, 2005. Gender specic differences in activity and home
range reect morphological dimorphism in wolf spiders (Araneae,
Lycosidae). J Arachnol 33:334–346.
Frost AJ, Winrow-Giffen A, Ashley PJ, Sneddon LU, 2007. Plasticity in
animal personality traits: does prior experience alter the degree of
boldness? Proc R Soc B Biol Sci 274:333–339.
Garamszegi LZ, Markó G, Herczeg G, 2012. A meta-analysis of cor-
related behaviours with implications for behavioural syndromes:
mean effect size, publication bias, phylogenetic effects and the role
of mediator variables. Evol Ecol 26:1213–1235.
Garamszegi LZ, Mueller JC, Markó G, Szász E, Zsebők S et al., 2014.
The relationship between DRD4 polymorphisms and phenotypic
correlations of behaviors in the collared ycatcher. Ecol Evol
4:1466–1479.
Givens RP, 1978. Dimorphic foraging strategies of a salticid spider
Phidippus audax. Ecology 59:309–321.
Golobinek R, Gregorič M, Kralj-Fišer S, 2021. Body size, not personal-
ity, explains both male mating success and sexual cannibalism in a
widow spider. Biology 10:189.
Gyuris E, Feró O, Barta Z, 2012. Personality traits across ontogeny in
rebugs Pyrrhocoris apterus. Anim Behav 84:103–109.
Hadeld JD, 2010. MCMC methods for multi-response generalized
linear mixed models: the MCMCglmm R package. J Stat Softw
33:1–22.
Hämäläinen A, Immonen E, Tarka M, Schuett W, 2018. Evolution
of sex-specic pace-of-life syndromes: causes and consequences.
Behav Ecol Sociobiol 72:50.
Head G, 1995. Selection on fecundity and variation in the degree of
sexual size dimorphism among spider species (Class Araneae).
Evolution 49:776–781.
Holtmann B, Lagisz M, Nakagawa S, 2017a. Metabolic rates, and not
hormone levels, are a likely mediator of between-individual differ-
ences in behaviour: a meta-analysis. Funct Ecol 31:685–696.
Holtmann B, Santos ESA, Lara CE, Nakagawa S, 2017b. Personality-
matching habitat choice, rather than behavioural plasticity, is a
likely driver of a phenotype–environment covariance. Proc R Soc B
Biol Sci 284:20170943.
Hosken DJ, House CM, 2011. Sexual selection. Curr Biol
21:R62–R65. https://www.sciencedirect.com/science/article/pii/
S0960982210015198?via%3Dihub.
Houslay TM, Wilson AJ, 2017. Avoiding the misuse of BLUP in behav-
ioural ecology. Behav Ecol 28:948–952. https://academic.oup.com/
beheco/article/28/4/948/3059669.
Ingle K, Horváth A, Gallé-Szpisjak N, Gellért L, Csata E et al., 2018.
The effects of overwintering and habitat type on body condition
and locomotion of the wolf spider Pardosa alacris. Acta Oecol
89:38–42.
Jakob EM, Marshall SD, Uetz GW, 1996. Estimating tness: a compar-
ison of body condition indices. Oikos 77:61–67.
Jandt JM, Bengston S, Pinter-Wollman N, Pruitt JN, Raine NE et al.,
2014. Behavioural syndromes and social insects: personality at
multiple levels. Biol Rev 89:48–67.
Johnson JC, Sih A, 2005. Precopulatory sexual cannibalism in shing
spiders Dolomedes triton: a role for behavioral syndromes. Behav
Ecol Sociobiol 58:390–396.
Johnson JC, Sih A, 2007. Fear, food, sex and parental care: a syndrome
of boldness in the shing spider Dolomedes triton. Anim Behav
74:1131–1138.
Kasumovic MM, Andrade MCB, 2009. A change in competitive
context reverses sexual selection on male size. J Evol Biol
22:324–333.
Kasumovic MM, Mason AC, Andrade MCB, Elias DO, 2011. The rela-
tive importance of RHP and resource quality in contests with own-
ership asymmetries. Behav Ecol 22:39–45.
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

16 Current Zoology
Kim ST, Lee SY, 2014. Arthropoda: Arachnida: Araneae: Clubionidae,
Corinnidae, Salticidae, Segestriidae. Invertebr Fauna Korea
21(31):1–186.
Klein W, 1988. Erfassung und bedeutung der in den apfelanlagen aufget-
retenen spinnen (Araneae) als nützlinge im grossraum Bonn (Doctoral
dissertation). Bonn: Rheinischen Friedrich-Wilhelms-Universität.
Kotiaho JS, 1998. Sexual differences in metabolic rates of spiders. J
Arachnol 26:401–404.
Kralj-Fišer S, Čelik T, Lokovšek T, Šuen K, Šiling R et al., 2014.
Development and growth in synanthropic species: plasticity and
constraints. Naturwissenschaften 101:565–575.
Kralj-Fišer S, Hebets EA, Kuntner M, 2017. Different patterns of
behavioral variation across and within species of spiders with dif-
fering degrees of urbanization. Behav Ecol Sociobiol 71:125.
Kralj-Fišer S, Laskowski KL, Garcia-Gonzalez F, 2019. Sex differences
in the genetic architecture of aggressiveness in a sexually dimorphic
spider. Ecol Evol 9:10758–10766.
Kralj-Fišer S, Schneider JM, 2012. Individual behavioural consistency
and plasticity in an urban spider. Anim Behav 84:197–204.
Kralj-Fišer S, Schneider JM, Kuntner M, Laskowski K, Garcia-Gonzalez
F, 2021. The genetic architecture of behavioral traits in a spider.
Ecol Evol 11:5381–5392.
Kralj-Fišer S, Schuett W, 2014. Studying personality variation in inver-
tebrates: why bother? Anim Behav 91:41–52.
Krumpálová Z, Tuf IH, 2013. Circadian rhythms of ground living spi-
ders: mechanisms of coexistence strategy based on the body size.
Pol J Ecol 61:575–586.
Liedtke J, Redekop D, Schneider JM, Schuett W, 2015. Early environ-
mental conditions shape personality types in a jumping spider.
Front Ecol Evol 3:134.
Liedtke J, Schneider JM, 2017. Social makes smart: rearing conditions
affect learning and social behaviour in jumping spiders. Anim Cogn
20:1093–1106.
Lim MLM, Li D, 2004. Courtship and male-male agonistic behav-
iour of Cosmophasis umbratica Simon, an ornate jumping spider
(Araneae: Salticidae) from Singapore. Rafes B Zool 52:435–448.
Maklakov AA, Bilde T, Lubin Y, 2004. Sexual selection for increased male
body size and protandry in a spider. Anim Behav 68:1041–1048.
Markó V, Keresztes B, 2014. Flowers for better pest control? Ground
cover plants enhance apple orchard spiders (Araneae), but not nec-
essarily their impact on pests. Biocontrol Sci Techn 24:574–596.
Mező L, Markó G, Kovács P, Markó V, 2019. Circadian rhythms in
the locomotor activity of the spiders Carrhotus xanthogramma
(Salticidae) and Philodromus cespitum (Philodromidae): temporal
patterns and sexual differences. Eur J Entomol 116:158–172.
Mező L, Markó G, Nagy C, Korányi D, Markó V, 2020. Beyond pol-
yphagy and opportunism: natural prey of hunting spiders in the
canopy of apple trees. Peer J 8:e9334.
Michalko R, Dvoryankina V, 2019. Intraspecic phenotypic variation
in functional traits of a generalist predator in an agricultural land-
scape. Agr Ecosyst Environ 278:35–42.
Michalko R, Košulič O, Řežucha R, 2017. Link between aggressiveness and
shyness in the spider Philodromus albidus (Araneae, Philodromidae):
state dependency over stability. J Insect Behav 30:48–59.
Michalko R, Řežucha R, 2018. Top predator’s aggressiveness and mes-
opredator’s risk-aversion additively determine probability of pre-
dation. Behav Ecol Sociobiol 72:105.
Modlmeier AP, Keiser CN, Wright CM, Lichtenstein JL, Pruitt JN,
2015. Integrating animal personality into insect population and
community ecology. Curr Opin Insect Sci 9:77–85.
Moiron M, Laskowski KL, Niemelä PT, 2020. Individual differences in
behaviour explain variation in survival: a meta-analysis. Ecol Lett
23:399–408.
Moran NP, Sánchez-Tójar A, Schielzeth H, Reinhold K, 2021. Poor
nutritional condition promotes high-risk behaviours: a systematic
review and meta-analysis. Biol Rev 96:269–288.
Nakagawa S, Schielzeth H, 2010. Repeatability for Gaussian and
non-Gaussian data: a practical guide for biologists. Biol Rev
85:935–956.
Nentwig W, Blick T, Bosmans R, Gloor D, Hänggi A et al., 2021. Spiders
of Europe, Version 01.2021. doi:10.24436/1.
Niemelä PT, Vainikka A, Hedrick AV, Kortet R, 2012. Integrating
behaviour with life history: boldness of the eld cricket Gryllus
integer during ontogeny. Funct Ecol 26:450–456.
Prenter J, Elwood RW, Montgomery WI, 1998. No association between
sexual size dimorphism and life histories in spiders. Proc R Soc B
Biol Sci 265:57–62.
Prenter J, Elwood RW, Montgomery WI, 1999. Sexual size dimorphism
and reproductive investment by female spiders: a comparative
analysis. Evolution 53:1987–1994.
Rádai Z, Németh Z, Barta Z, 2018. Sex-dependent immune response in
a semelparous spider. Naturwissenschaften 105:39.
Réale D, Dingemanse NJ, 2012. Animal Personality. eLS. doi:
10.1002/9780470015902.a0023570.
Réale D, Garant D, Humphries MM, Bergeron P, Careau V et al., 2010.
Personality and the emergence of the pace-of-life syndrome concept
at the population level. Phil Trans R Soc B Biol Sci 365:4051–4063.
Rodriguez A, Zhang H, Klaminder J, Brodin T, Andersson PL et al.,
2018. ToxTrac: a fast and robust software for tracking organisms.
Methods Ecol Evol 9:460–464.
Royauté R, Buddle CM, Vincent C, 2014. Interpopulation variations in
behavioral syndromes of a jumping spider from insecticide-treated
and insecticide-free orchards. Ethology 120:127–139.
Royauté R, Buddle CM, Vincent C, 2015. Under the inuence: suble-
thal exposure to an insecticide affects personality expression in a
jumping spider. Funct Ecol 29:962–970.
Royauté R, Hedrick A, Dochtermann NA, 2021. Sex-specic behav-
ioral syndromes allow the independent evolution of behavioral
dimorphism. EcoEvoRxiv.
Schielzeth H, Dingemanse NJ, Nakagawa S, Westneat DF, Allegue H et
al., 2020. Robustness of linear mixed-effects models to violations
of distributional assumptions. Methods Ecol Evol 11:1141–1152.
Schmitz A, 2004. Metabolic rates during rest and activity in differ-
ently tracheated spiders (Arachnida, Araneae): Pardosa lugubris
(Lycosidae) and Marpissa muscosa (Salticidae). J Comp Physiol B
174:519–526.
Schmitt A, Schuster M, Barth FG, 1990. Daily locomotor activity pat-
terns in three species of Cupiennius (Araneae, Ctenidae): the males
are the wandering spiders. J Arachnol 18:249–255. https://www.
jstor.org/stable/3705430.
Sih A, Bell AM, Johnson JC, Ziemba RE, 2004. Behavioral syndromes:
an integrative overview. Q Rev Biol 79:241–277.
Sivalinghem S, Kasumovic MM, Mason AC, Andrade MCB, Elias DO,
2010. Vibratory communication in the jumping spider Phidippus
clarus: polyandry, male courtship signals, and mating success.
Behav Ecol 21:1308–1314.
Slatkin M, 1984. Ecological causes of sexual dimorphism. Evolution
38:622–630.
Snell-Rood EC, 2013. An overview of the evolutionary causes
and consequences of behavioural plasticity. Anim Behav
85:1004–1011.
Stamps JA, Briffa M, Biro PA, 2012. Unpredictable animals: individ-
ual differences in intraindividual variability (IIV). Anim Behav
83:1325–1334.
Sullivan HL, Morse DH, 2004. The movement and activity patterns
of similar-sized adult and juvenile crab spiders Misumena vatia
(Araneae, Thomisidae). J Arachnol 32:276–283.
Sweeney K, Gadd RDH, Hess ZL, McDermott DR, MacDonald L et al.,
2013. Assessing the effects of rearing environment, natural selec-
tion, and developmental stage on the emergence of a behavioral
syndrome. Ethology 119:436–447.
Szinetár Cs, 2006. Pókok – Keresztespókok, farkaspókok, ugrópókok
és rokonaik a Kárpát-medencében. Budapest, Hungary: Kossuth
Kiadó.
Tarka M, Guenther A, Niemelä PT, Nakagawa S, Noble DWA, 2018.
Sex differences in life history, behavior, and physiology along
a slow-fast continuum: a meta-analysis. Behav Ecol Sociobiol
72:132.
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022

Mezőfi et al. · Dimorphic strategies among immatures of a jumping spider 17
Taylor LA, Clark DL, McGraw KJ, 2011. Condition dependence of
male display coloration in a jumping spider Habronattus pyrri-
thrix. Behav Ecol Sociobiol 65:1133–1146.
Uhl G, Schmitt S, Schäfer MA, Blanckenhorn W, 2004. Food and
sex-specic growth strategies in a spider. Evol Ecol Res 6:523–540.
Walker SE, Irwin JT, 2006. Sexual dimorphism in the metabolic rate of two
species of wolf spider (Araneae, Lycosidae). J Arachnol 34:368–373.
Walker SE, Marshall SD, Rypstra AL, Taylor DH, 1999. The effects
of hunger on locomotory behaviour in two species of wolf spider
(Araneae, Lycosidae). Anim Behav 58:515–520.
Winkler L, Moiron M, Morrow EH, Janicke T, 2021. Stronger net
selection on males across animals. bioRxiv 2021(04):16–440171.
doi:10.1101/2021.04.16.440171.
WSC, 2021. World Spider Catalog, Version 22.0. doi:10.24436/2.
Downloaded from https://academic.oup.com/cz/advance-article/doi/10.1093/cz/zoac069/6692897 by guest on 05 October 2022