Content uploaded by Rosli Hashim
Author content
All content in this area was uploaded by Rosli Hashim
Content may be subject to copyright.
ORIGINAL PAPER
Eyespan reflects reproductive quality in wild stalk-eyed
flies
Samuel Cotton ÆJennifer Small ÆRosli Hashim ÆAndrew Pomiankowski
Received: 3 April 2008 / Accepted: 15 January 2009 / Published online: 29 January 2009
ÓSpringer Science+Business Media B.V. 2009
Abstract Handicap models of sexual selection propose that females use male sexual
ornaments as a cue in mate choice because they reflect commodities that increase female
fitness, either directly or indirectly. In contrast to studies on vertebrates, most investiga-
tions of ornaments in insects and other invertebrate taxa have been conducted under
laboratory conditions. There is a pressing need to address questions relating to sexual
signalling of quality in natural populations, as the arbitrary and uniform environments
found in the laboratory fail to reflect the world under which animals have evolved. We
investigated associations between male ornaments (exaggerated eyespan), attractiveness,
and reproductive quality in a wild population of the sexually ornamented stalk-eyed fly,
Teleopsis dalmanni. We also explored the relationship between eyespan and reproductive
quality in females to evaluate the potential for sexually antagonistic selection on eyespan.
We show that eyespan is a generic correlate of reproductive quality, acting as a reliable
mirror of variation in reproductive fitness in both sexes. Our findings suggest that male
ornaments signal commodities that are of interest to females in the natural environment in
which they, and mate preferences for them, have evolved. In addition, the covariance
between female eyespan and reproductive output suggests that the former may be a reliable
cue of quality in its own right. Our data provide important insights into the evolutionary
forces that shape the evolution of exaggerated eyespan in wild populations of this species.
Keywords Sexual selection Ornaments Wild Stalk-eyed fly Testis
Accessory glands Fecundity Fertility
S. Cotton (&)J. Small A. Pomiankowski
Research Department of Genetics, Evolution and Environment, University College London,
Wolfson House, 4 Stephenson Way, London NW1 2HE, UK
e-mail: s.cotton@ucl.ac.uk
R. Hashim
Institute of Biological Science, University of Malaya, 50603 Kuala Lumpur, Malaysia
A. Pomiankowski
CoMPLEX, University College London, Gower Street, London WC1E 6BT, UK
123
Evol Ecol (2010) 24:83–95
DOI 10.1007/s10682-009-9292-6
Introduction
The handicap theory of sexual selection proposes that females use male sexual ornaments
as a cue in mate choice because such traits are costly and reflect the phenotypic quality of
their bearer (Zahavi 1975; Andersson 1986; Pomiankowski 1987,1988; Grafen 1990;
Iwasa et al. 1991; Iwasa and Pomiankowski 1994,1999). Poor quality males are unable to
bear the cost, or reap the benefits, of possessing a large sexual trait, and hence only high
quality males can afford to signal at high levels (Grafen 1990; Iwasa et al. 1991; Iwasa and
Pomiankowski 1994,1999; Getty 2006). Variation in male quality can be environmentally
and/or genetically derived. Either source is sufficient to drive female choice for orna-
mented males, through either direct benefits to the female, and/or indirect benefits to her
offspring (Iwasa and Pomiankowski 1999; Kokko et al. 2006).
Laboratory studies are frequently used to evaluate the currency being signalled by
ornaments (reviewed in Andersson 1994; Johnstone 1995; Maynard Smith and Harper
2003; Cotton et al. 2004a). They are an essential tool in our understanding of ornament
evolution, as they allow systematic investigations to be performed with a high degree of
replication and control. However, such studies can be criticised because the arbitrary and
uniform environments found in the laboratory may be misleading and have obscure effects
on components of fitness (Charmantier and Sheldon 2006; Ellegren and Sheldon 2008). In
addition, since laboratory experiments typically consider results drawn from only one or a
few discrete environments, the average effect of natural environmental variation on wild-
type individuals is unclear. Correlations between traits are known to be particularly sen-
sitive to the environment in which they are measured (reviewed in Sgro
`and Hoffmann
2004). Studies addressing the relationships between traits therefore need to consider not
only correlations derived under relatively benign laboratory environments, but also those
that exist in the resource-limited and stressful conditions experienced by field populations.
The advantages of the narrow and invariant nature of laboratory environments are therefore
also disadvantages, as they fail to reflect the world under which animals have evolved. It is
therefore crucial to address questions relating to sexual signalling of quality in nature.
While there are a number of demonstrations in wild vertebrate populations that orna-
ments reflect components of fitness that are likely to be of interest to females (Andersson
1994), there are remarkably few in insects and other invertebrate taxa. Some studies have
reported potential benefits of mating with larger or older males (e.g. Zuk 1988), or with
well-ornamented males advertising good territories (Greenfield 1997), but examples in
nature of ornamental signals of other male phenotypic qualities are, surprisingly, rather
rare. An exception is that of Simmons et al. (2005), who report that calling song char-
acteristics reflect aspects of immune function in a wild cricket population. The relative
paucity of sexual selection studies in wild insect populations contrasts starkly with our
knowledge of such systems under laboratory conditions.
In the Malaysian stalk-eyed fly, Teleopsis dalmanni (Meier and Baker 2002), the eyes
are displaced laterally from the head on elongate ‘eye-stalks’ in both sexes (Baker et al.
2001). Eyespan (the distance between the eyes) is sexually dimorphic, being much
enlarged in males as a result of sexual selection. T. dalmanni form nocturnal aggregations
on root hairs that hang underneath the eroded banks of forest streams (Burkhardt and de la
Motte 1985; Wilkinson and Dodson 1997). Males fight for control of these roosting sites,
and males with the largest eyespan usually win such male-male contests (Panhuis and
Wilkinson 1999). Females also prefer to alight and mate on root hairs controlled by males
with large eyespans (Wilkinson and Reillo 1994; Wilkinson and Dodson 1997; Hingle
et al. 2001a,b; this paper).
84 Evol Ecol (2010) 24:83–95
123
Many diverse laboratory studies have investigated the signalling function of exagger-
ated eyespan in T. dalmanni. Male eyespan is highly condition-dependent, becoming
greatly reduced with larval stress relative to other traits (dietary stress—David et al. 1998;
Cotton et al. 2004b,c; temperature shocks and desiccation—Bjorksten et al. 2001). It also
has a strong genetic basis, and the relative performance of genotypes is maintained across
different environments, preserving and amplifying the genetic signal broadcast by orna-
ments (David et al. 2000).
In the laboratory, well-ornamented males also develop larger internal reproductive
organs and exhibit higher fertility than small eyespan males when housed with multiple
females (Rogers et al. 2008). The size of the accessory glands, paired structures involved in
the production of spermatophores, is strongly related to male reproductive success in T.
dalmanni, as it determines when sexual maturity is attained (Baker et al. 2003), and
covaries both phenotypically (Baker et al. 2003; Rogers et al. 2005a) and genetically
(Rogers et al. 2005b) with male mating frequency. These glands become depleted with
repeated mating (Rogers et al. 2005a), and their size provides a physiological cap on male
mating rate. High mating rates are beneficial to males, as they attempt to copulate with
females in their roosting aggregations during the brief dusk and dawn periods. Testis size is
likewise an important determinant of male reproductive success in T. dalmanni as the
number of sperm stored in a female’s spermathecae is reported to correlate positively with
the testis size of her mate (Fry 2006).
Despite the wealth of information on the causes and correlates of eyespan in the lab-
oratory, very little is known about its signalling function under natural conditions in this
important sexual selection model. Here, we describe data collected from wild caught stalk-
eyed flies, and examine the relationships between eyespan and measures of attractiveness
and reproductive quality. Male and female eyespan show positive genetic covariance
(Wilkinson 1993), and such inter-sexual genetic correlations may retard ornament evolu-
tion if natural selection on the trait in females counters any sexually selected advantage in
males (Fisher 1958; Lande 1980; Lande and Arnold 1985). So we also evaluated the
relationship between female eyespan and reproductive output (fecundity and fertility)
under field conditions, in order to assess the potential for sexually antagonistic selection on
eyespan. Our data provide important and surprising insights into the evolutionary forces
that shape the evolution of eyespan in wild populations of stalk-eyed flies.
Materials and methods
Fieldwork
Two phases of fieldwork were carried out in Ulu Gombak, Peninsular Malaysia (3°190N,
101°450E). In March and April 2005 we collected flies after dusk (2000–2300 hours) from
aggregation sites that formed on roothairs hanging from the eroded banks of seven tributaries
of the Gombak river (tributaries were not re-sampled). Individual aggregations (N=58)
were captured in their entirety by enclosing them in a plastic bag. Flies (N=69 and 96 for
males and females, respectively) were anaesthetized on ice shortly after capture. All indi-
viduals were measured for eyespan and thorax length to an accuracy of 0.01 mm using a
monocular microscope and the image analysis program NIH Image (Version 1.55; National
Institute of Health, Bethesda, MD, USA). Eyespan was defined as the distance between the
outer tips of the eyes, and thorax length was measured ventrally from the anterior face of the
head along the midline to the joint between the metathoracic legs and the thorax.
Evol Ecol (2010) 24:83–95 85
123
In a second phase of fieldwork at the same location (July and August 2006), we captured
females as above (N=36) and maintained them individually on a diet of pureed banana
(replenished every other day) in 400 ml containers with a moist tissue paper base onto
which females could lay eggs. The tissue paper base was replaced every other day and
stored on a damp cotton pad in a Petri dish for fertility scoring (see below). Following egg
collection, females were measured for eyespan and thorax length (as above). Time in
captivity, and thus the egg collection period, was variable across females (range of time in
captivity =2–26 days).
Male attractiveness and reproductive quality
Nocturnal roosting sites comprised both mixed sex and single sex aggregations. Usually
only one or occasionally two males were found in mixed-sex aggregations, whereas female
numbers were more variable (see ‘‘Results’’). In mixed sex aggregations there exists a high
potential for male mating success as the vast majority of matings occur in lekking
aggregations during the dusk and dawn periods (Lorch et al. 1993; S. Cotton and J. Small,
personal observation). While lone males sometimes sneak matings with females in other
males’ harems (S. Cotton and J. Small, personal observation), the potential for mating
success of males from single-sex aggregations is much lower from those with harems. This
difference between mixed- and single-sex aggregations represents the most basic form of
sexual selection. We therefore quantified sexual selection on male eyespan through the
association with both the number of females in his harem and through membership of
mixed-sex aggregations.
Male reproductive quality was estimated using accessory gland and testis size. The male
reproductive organs were dissected into phosphate buffered saline (PBS), uncoiled, placed
on a graticule under a microscope and photographed digitally (Baker et al. 2003). Previous
work has shown that length and area measurements are highly correlated in both accessory
glands and testes (Rogers et al. 2005a). Each accessory gland and testis was therefore
measured from the digitized images using NIH Image software by tracing a midline that
longitudinally bisected each organ.
Female reproductive quality
Female reproductive quality was estimated using four aspects of fecundity. First, the ovaries
of females from the initial phase of fieldwork (2005) were dissected out in PBS, gently
broken open and the number of mature eggs counted (defined as stages 12–14 using King’s
standard stages of oogenesis, King 1970). Second, females collected in 2006 were scored for
the number of eggs laid over their time in captivity. We also scored the number of eggs
fertilized in a female’s clutch. Fertilised T. dalmanni eggs typically hatch 2–3 days after
being laid, leaving only the outer shell of the chorion; unhatched eggs appear full with the egg
still inside the chorion (Baker et al. 2001). Hatching success of eggs was determined visually
under a dissecting microscope after 5 days of incubation. Some unhatched eggs showed
segmental striations and other signs of development and were classified as fertile.
Statistical analyses
Sexual selection on male eyespan was estimated in two ways: (1) through associations
between male eyespan and the number of females in the harem, and (2) through logistic
86 Evol Ecol (2010) 24:83–95
123
regression of the potential for mating success (0 for males from male-only aggregations, 1
for males from mixed sex aggregations) on a male’s eyespan, using a binomial error
structure. The significance of a male’s eyespan as a predictor of his potential mating
success was evaluated using a likelihood ratio test of explained deviance. We also asked
whether any effect of male eyespan persisted after controlling for allometry by entering
thorax length as a covariate estimator of body size, and asking whether the subsequent
addition of eyespan explained significant amounts of deviance using a likelihood ratio test.
Given our a priori knowledge from laboratory studies that female fitness is increased
directly by associating with large eyespan males with large reproductive organs (Rogers
et al. 2008), we also tested for associations between male reproductive quality (accessory
glands and testis size) and sexual selection (as defined above).
In both males and females the relationships between eyespan and aspects of repro-
ductive quality were evaluated in GLMs. Eggs were collected from each female every
2 days, for up to 26 in the investigation of female reproductive output, so we also included
female identity as a random factor (shrunk by REML estimation) in all models. To
investigate whether fecundity or fertility showed any systematic changes over time in
captivity, and to control for such effects during tests of other variables, we also included
assay period as a continuous factor in the GLMs (qualitatively identical outcomes were
obtained if assay period was classified as an ordinal variable—data not shown). Since
many eggs were unfertilized in the investigation of female reproductive output, we also
evaluated the relationship between female eyespan and the relative number of eggs fer-
tilized during each collection period by (1) including fecundity as a covariate in GLMs,
and (2) computing the proportion of eggs fertilized from the eggs laid. Note that the
estimate of the number of fertile eggs was in some instances reduced because eggs were
occasionally enveloped by mould growth during the 5 days prior to the assay of fertility
(and hence their fertility was inestimable). So when including fecundity as a covariate or
computing the proportion of eggs fertilized, we used the number of eggs from which
fertility was calculated as the covariate/denominator. To permit reasonable estimates of the
relative number of eggs fertilized, analyses were restricted to those females which laid [10
eggs over their duration of captivity.
To control for allometric scaling we also constructed GLMs using thorax length as a
covariate measure of body size, and asked whether the subsequent inclusion of eyespan
explained significant amounts of the error variance.
All statistical analyses were performed using JMP software (version 5.0.1a, SAS
Institute Inc.).
Results
Male attractiveness
Most leks (22/34) had only a single male (mean ±SD =1.39 ±0.56; range =1–3),
whereas female numbers were more variable (mean ±SD =2.09 ±1.57; range =1–7).
We asked whether large eyespan males attracted more females, taking both the largest and
average value when there was more than one male in an aggregation. The number of
females attracted to a lek did not increase significantly with male eyespan in single male
leks (F=2.66, df =1,35, P=0.11) but did with both the largest (largest F=5.41,
df =1,13, P=0.037) and the average (F=12.23, df =1,13, P=0.004) male eyespan
in multi-male leks. When aggregations with zero females were excluded from this analysis,
Evol Ecol (2010) 24:83–95 87
123
none of these relationships were significant (single male eyespan F=0.20, df =1,19,
P=0.66; largest male eyespan F=1.75, df =1,10, P=0.22; average male eyespan
F=3.88, df =1,11, P=0.08).
To extend our analysis, we used the presence or absence of females as a metric of male
attractiveness and found significant sexual selection on male eyespan in single male leks
(logistic regression v
2
=11.71, df =1, P\0.001, Fig. 1). Similar sexual selection
operated on the male with the largest male eyespan (v
2
=4.53, df =1, P=0.033) and the
average male eyespan in multi-male leks (v
2
=9.51, df =1, P=0.002). So males with
large eyespan were significantly more likely to attract females than males with smaller
eyespans, who tended to roost in male-only aggregations. A part of this sexual selection
arose due to the non-allometric component of male ornamentation, as the eyespan effect
persisted after including thorax length as a covariate estimate of body size for single male
aggregations (v
2
=3.75, df =1, P=0.052), and the largest male (v
2
=3.83, df =1,
P=0.051) in multi-male aggregations.
Male reproductive quality
Male eyespan explained significant amounts of variation in male reproductive quality
(Table 1); males with larger eyespans had larger accessory glands length and testes
(Fig. 2a). Such associations may be, in part, due to allometric scaling. We therefore
controlled for allometry by using thorax length as a covariate measure of body size. After
removing the covariance associated with body size, male eyespan still explained significant
amounts of variation in accessory glands and testis size (Table 1); males with dispropor-
tionately large eyespans had larger reproductive organs (Fig. 2b; partial correlation
coefficients, accessory glands r=0.39; testes r=0.37). In parallel with the non-random
distribution of females with respect to male eyespan phenotype (described above), we
found that males in mixed sex aggregations had significantly larger reproductive organs
than those from male-only roosts (accessory gland v
2
=4.58, df =1, P=0.032; testes
v
2
=8.78, df =1, P=0.003).
Fig. 1 Sexual selection on male eyespan, measured through membership of either a single sex aggregation
(potential for mating success =0) or a mixed sex aggregation (potential for mating success =1). Open
circles represent data from individual males, whereas the dotted line represents the overall probability of
achieving mating success (i.e. the sexual selection function)
88 Evol Ecol (2010) 24:83–95
123
Table 1 The effect of eyespan on variation in reproductive quality in males
Eyespan measure Reproductive quality F-ratio df P
Absolute (1) Accessory glands 50.64 1,59 \0.001
(2) Testes 15.67 1,61 \0.001
Body-size controlled (1) Accessory glands 10.23 1,57 0.002
(2) Testes 9.59 1,59 0.003
Male reproductive quality was estimated using the size of (1) the accessory glands, and (2) the size of the
testes. Body size controlled measures are the effects and significance of the reduction of error variance after
the addition of eyespan into a model already containing thorax length
Fig. 2 The relationship between amale eyespan and bresidual male eyespan and mean male reproductive
organ size. Accessory glands, closed circles; testes, open circles. Linear regression lines are shown for
heuristic purposes
Evol Ecol (2010) 24:83–95 89
123
Female reproductive quality
Female eyespan explained significant amounts of variation in all measured aspects of
female reproductive quality (Table 2). Large eyespan females had more eggs within their
ovaries (Fig. 3a), laid more eggs (Fig. 3c) and laid more fertile eggs (Fig. 3d) than females
with smaller eyespan. Our results did not arise through differential mortality or time in
captivity with respect to eyespan, as there was no covariance between eyespan and either
of these factors (eyespan-proportion of time survived in captivity r=0.03, P=0.86;
eyespan-total time in captivity r=-0.06, P=0.77).
Strikingly, we found that, on average, only around one-third of eggs laid by females
were fertile (mean ±SE proportion of fertilized eggs per clutch =0.36 ±0.06). We
found that the number of eggs laid did not change over time in captivity (F=0.51,
df =1,232, P=0.70). However, both the number of fertile eggs laid (F=8.41,
df =1,76, P=0.005) and the proportion of fertile eggs laid (mean =1.3% per day,
F=5.62, df =1,71, P=0.020) declined over time, suggesting that females were
becoming increasingly sperm-limited because of their isolation from males. This effect
does not account for the low overall fertility seen in our sample, as the average proportion
of fertile eggs laid was only 55.2% during the first 4 days in captivity.
When we estimated fertility as both the relative number of fertile eggs with fecundity as
a covariate, and the proportion of eggs fertilised, we found that female eyespan was a
significant explanatory variable (Table 2) as larger females had higher relative fertility and
a greater proportion of their clutch fertilized (Fig. 3e, f; Table 2). After controlling for
allometry, we found that female eyespan still explained significant variation in the number
of eggs in a female’s ovaries (Table 2), wherein females with large eyespans for their body
size had higher ovarian fecundity (Fig. 3b; partial correlation coefficient, r=0.32). We
found no independent effect of female eyespan on the number of eggs laid or on the
number of fertile eggs (Table 2). There was also no effect of body-size controlled female
eyespan on the relative number or proportion of eggs fertilized (Table 2).
Table 2 The effect of eyespan on variation in reproductive quality in females
Eyespan measure Reproductive quality F-ratio df P
Absolute (1) Eggs in ovaries 14.56 1,92 \0.001
(2) Fecundity 12.37 1,203 \0.001
(3) Fertility 13.52 1,70 \0.001
(4) Relative fertility 8.16 1,67 0.006
(5) Proportion of fertile eggs 4.29 1,65 0.042
Body-size controlled (1) Eggs in ovaries 9.71 1,88 0.003
(2) Fecundity 0.02 1,202 0.90
(3) Fertility 2.36 1,69 0.13
(4) Relative fertility 1.44 1,66 0.23
(5) Proportion of fertile eggs 0.75 1,64 0.39
Female reproductive quality was estimated through the following measures: (1) the number of mature eggs
dissected from the ovaries, (2) fecundity, the number of eggs laid, (3) fertility, the number of fertilised eggs
laid, (4) relative fertility, the number of fertilised eggs laid after controlling for variation in fecundity, and
(5) the proportion of fertilised eggs. Body size controlled measures are the effects and significance of the
reduction of error variance after the addition of eyespan into a model already containing thorax length
90 Evol Ecol (2010) 24:83–95
123
Discussion
We investigated the associations between male ornaments, attractiveness, and reproductive
quality in a wild insect population. We found that male ornaments signal accessory gland
and testis size, traits that are likely to be of interest to females in the natural environment in
which they and mate preferences evolved. We also explored the relationship between
female eyespan, the homologue of the male ornament, and female reproductive quality to
evaluate the potential for sexually antagonistic selection on the ornamental trait. Our
discovery that the homologous trait in females was indicative of female fecundity and
fertility suggests that female eyespan is positively associated with female reproductive
quality, and thus male ornament evolution is unlikely to be constrained by sexually
antagonistic selection.
Males with larger eyespans were more likely to attract a harem. Since eyespan is
strongly allometric (Baker and Wilkinson 2001), this effect may have arisen as a correlate
Fig. 3 The relationships between afemale eyespan and bresidual female eyespan and the number of
mature eggs in the ovaries, cfemale eyespan and the number of eggs laid per day, dfemale eyespan and the
number of fertile eggs laid per day, efemale eyespan and the relative number of fertilized eggs, and ffemale
eyespan and the proportion of fertilized eggs, visualized using least squares mean (LSM) estimates of
reproductive output from models containing female identity and assay period. Linear regression lines are
shown for heuristic purposes
Evol Ecol (2010) 24:83–95 91
123
of sexual selection for large body size. However, we show for the first time that there is
sexual selection on the non-allometric component of male ornaments, suggesting that
eyespan is the specific target of female mate choice in this species. We also measured the
relationship between male eyespan and two important aspects of male reproductive quality,
accessory gland and testis size. Males with large eyespans had larger accessory glands and
testes, and these associations persisted after controlling for body size, suggesting that they
are not derived from simple rules of allometric scaling.
The relationship between eyespan and accessory gland size suggests that males with
proportionately large ornaments are able to mate more frequently, as accessory gland size is
phenotypically and genetically correlated to mating rate under laboratory conditions (Baker
et al. 2003; Rogers et al. 2005a,b). Moreover, mating with large eyespan males may result in
higher female fertility, since the number of sperm stored in a female’s spermathecae is
reported to correlate positively with the testis size of her mate (Fry 2006). Given that few
sperm are stored following a single mating (*65 Wilkinson et al. 2005;*140 Rogers et al.
2006), females must copulate repeatedly to achieve maximal fertility (Baker et al. 2001).
Indeed, only around one-third of eggs laid by wild-caught females were fertile and females
became less fertile the longer they were kept in isolation from males, suggesting that sperm-
limitation is severe under natural conditions. So there must be strong selection on females to
acquire more gametes, for example by associating with males who can mate at high fre-
quencies and/or deliver more sperm per ejaculate. In line with this we found that females
preferred to associate with males bearing the largest reproductive organs.
Females with large eyespans had higher reproductive output than small eyespan
females, as female eyespan covaried significantly with the number of mature oocytes, the
number of eggs laid, and the number of fertile eggs laid. Females with large eyespans for
their body size also contained significantly more mature oocytes, but we did not find any
relationships between the number of laid or fertilized eggs and the non-allometric com-
ponent of female eyespan. However, we suspect that this was due to low sample size and a
relative lack of power to detect such effects. The covariance between measures of
fecundity and female eyespan is likely to arise indirectly through the common dependence
of these traits on female condition.
In the laboratory, high quality females can suffer from more acute sperm-limitation, as
females with large eyespan have higher fecundity but a lower percentage of their eggs are
fertilized following a single mating compared to less productive small eyespan females
(Rogers et al. 2006). We found the opposite pattern in wild females, with larger eyespan
females having proportionately more of their eggs fertilized. We offer two non-mutually
exclusive explanations for this discrepancy. First, larger females may mate at a higher
frequency than smaller females, thereby alleviating their sperm deficit, and second, males
may strategically allocate more of their ejaculate to larger, more fecund, females (Rogers
et al. 2006). Given that female eyespan is not used during female-female encounters
(Al-khairulla et al. 2003), both of these mechanisms imply that there is male mate choice
for females with large eyespans. These hypotheses merit future investigation.
Positive relationships between primary and secondary male sexual traits are expected
when resources are non-limiting, but trade-offs should emerge under stressful conditions
when resource distribution is more critical, such as in nature (Sgro
`and Hoffmann 2004).
The observed positive covariance between ornaments and reproductive quality in
T. dalmanni is at odds with a number of recent empirical studies that have reported either
the opposite (e.g. Simmons and Emlen 2006), or that relationships between such traits are
complicated and often contingent on other factors such as social status (e.g. Cornwallis and
Birkhead 2008). Most previous studies have been conducted under laboratory conditions or
92 Evol Ecol (2010) 24:83–95
123
with domesticated species, so it is currently difficult to evaluate their generality, as rela-
tionships between primary and secondary sexual traits are unknown in the environments in
which they evolved.
Stalk-eyed fly external morphology is fixed at eclosion and is determined in large part
by the amount of resources obtained during larval growth stages. In contrast, the repro-
ductive organs develop only after emergence from the pupa, with flies not becoming
mature for at least 3 weeks post eclosion, so their growth is contingent on resources
acquired during the adult stage (Baker et al. 2003; Rogers et al. 2008). Given the differ-
ences in developmental timing between eyespan and reproductive organs, these traits seem
unlikely to be subject to the resource-based trade-off hypotheses that have come to
dominate the literature. However, eyespan in both sexes is highly condition-dependent
(Cotton et al. 2004b,c), as are male reproductive organs (accessory glands and testes—
Baker et al. 2003; Rogers et al. 2008) and fecundity, which is severely reduced by food
stress during adult life (Hingle et al. 2001b). A more likely explanation is therefore that the
positive relationships arise through the dependence of both eyespan and primary sexual
traits on overall quality, with individuals in high condition being better able to buffer
themselves against environmental stress, and producing both large eyespans and high
reproductive output.
In this study we have shown that eyespan is a generic correlate of reproductive quality
and acts as a reliable mirror of variation in reproductive fitness in both sexes, before and
after controlling for covariation with body size. Given that eyespan is genetically corre-
lated across the sexes in T. dalmanni (Wilkinson 1993), our findings suggest that there is
unlikely to be sexually antagonistic selection on male and female eyespan. Together, these
findings add weight to the view that ornaments evolve from traits that were originally
sensitive to underlying quality (Fisher 1958; Cotton et al. 2004c), and may account for the
repeated evolution of eyespan among the stalk-eyed flies (Baker and Wilkinson 2001).
Acknowledgments This work was supported by awards from the Association for the Study of Animal
Behaviour (A.P. and R.H.), the Royal Society (A.P. and R.H.), the BBSRC (grants to K. Fowler, A.P., T.
Chapman and H. Smith; studentship to J.S.), the Department of Biology, University College London (J.S.
and A.P.), and a NERC Fellowship (S.C.). We thank two anonymous reviewers for comments on a previous
version of the manuscript, and staff at the Gombak Valley Field Research Centre, University of Malaya,
Kuala Lumpur for assistance.
References
Al-khairulla H, Warburton D, Knell RJ (2003) Do the eyestalks of female Diopsid flies have a function in
intrasexual aggressive encounters? J Insect Behav 16:679–686. doi:10.1023/B:JOIR.0000007703.
84691.ad
Andersson M (1986) Evolution of condition-dependent sex ornaments and mating preferences; sexual
selection based on viability differences. Evol Int J Org Evol 40:804–816. doi:10.2307/2408465
Andersson M (1994) Sexual selection. Princeton University Press, Princeton
Baker RH, Wilkinson GS (2001) Phylogenetic analysis of sexual dimorphism and eyespan allometry in
stalk-eyed flies (Diopsidae). Evol Int J Org Evol 55:1373–1385
Baker RH, Ashwell RIS, Richards TA et al (2001) Effects of multiple mating and male eye span on female
reproductive output in the stalk-eyed fly Cyrtodiopsis dalmanni. Behav Ecol 12:732–739. doi:10.1093/
beheco/12.6.732
Baker RH, Denniff M, Futerman P et al (2003) Accessory gland size influences time to sexual maturity and
mating frequency in the stalk-eyed fly, Cyrtodiopsis dalmanni. Behav Ecol 14:607–611. doi:10.1093/
beheco/arg053
Bjorksten TA, Pomiankowski A, Fowler K (2001) Temperature shock during development fails to increase
the fluctuating asymmetry of a sexual trait in stalk-eyed flies. Proc R Soc Lond B Biol Sci 268:1503–
1510. doi:10.1098/rspb.2001.1575
Evol Ecol (2010) 24:83–95 93
123
Burkhardt D, de la Motte I (1985) Selective pressures, variability, and sexual dimorphism in stalk-eyed flies
(Diopsidae). Naturwissenschaften 72:204–206. doi:10.1007/BF01195763
Charmantier A, Sheldon BC (2006) Testing genetic models of mate choice evolution in the wild. Trends
Ecol Evol 21:417–419. doi:10.1016/j.tree.2006.06.001
Cornwallis CK, Birkhead TR (2008) Plasticity in reproductive phenotypes reveals status-specific correla-
tions between behavioral, morphological, and physiological sexual traits. Evol Int J Org Evol 62:1149–
1161. doi:10.1111/j.1558-5646.2008.00346.x
Cotton S, Fowler K, Pomiankowski A (2004a) Do sexual ornaments demonstrate heightened condition-
dependent expression as predicted by the handicap hypothesis? Proc R Soc Lond B Biol Sci 271:771–
783. doi:10.1098/rspb.2004.2688
Cotton S, Fowler K, Pomiankowski A (2004b) Condition dependence of sexual ornament size and variation
in the stalk-eyed fly Cyrtodiopsis dalmanni (Diptera: Diopsidae). Evol Int J Org Evol 58:1038–1046
Cotton S, Fowler K, Pomiankowski A (2004c) Heightened condition dependence is not a general feature of
male eyespan in stalk-eyed flies (Diptera: Diopsidae). J Evol Biol 17:1310–1316. doi:10.1111/j.
1420-9101.2004.00754.x
David P, Hingle A, Greig D et al (1998) Male sexual ornament size but not asymmetry reflects condition in
stalk-eyed flies. Proc R Soc Lond B Biol Sci 265:2211–2216. doi:10.1098/rspb.1998.0561
David P, Bjorksten T, Fowler K et al (2000) Condition-dependent signalling of genetic variation in stalk-
eyed flies. Nature 406:186–188. doi:10.1038/35018079
Ellegren H, Sheldon BS (2008) Genetic basis of fitness differences in natural populations. Nature 452:
169–175. doi:10.1038/nature06737
Fisher RA (1958) The genetical theory of natural selection. Clarendon Press, Oxford (reprinted 1999,
Oxford Univ Press, Oxford)
Fry C (2006) Juvenile hormone mediates a trade-off between primary and secondary sexual traits in stalk-
eyed flies. Evol Dev 8:191–201. doi:10.1111/j.1525-142X.2006.00089.x
Getty T (2006) Sexually selected signals are not similar to sports handicaps. Trends Ecol Evol 21:83–88.
doi:10.1016/j.tree.2005.10.016
Grafen A (1990) Sexual selection unhandicapped by the fisher process. J Theor Biol 144:473–516. doi:
10.1016/S0022-5193(05)80087-6
Greenfield MD (1997) Sexual selection in resource defense polygamy: lessons from territorial grasshoppers.
In: Chloe J, Crespi B (eds) The evolution of mating systems in insects and arachnids. Cambridge
University Press, Cambridge, pp 75–88
Hingle A, Fowler K, Pomiankowski A (2001a) Size-dependent mate preference in the stalk-eyed fly
Cyrtodiopsis dalmanni. Anim Behav 61:589–595. doi:10.1006/anbe.2000.1613
Hingle A, Fowler K, Pomiankowski A (2001b) The effect of transient food stress on female mate preference
in the stalk-eyed fly Cyrtodiopsis dalmanni. Proc R Soc Lond B Biol Sci 268:1239–1244. doi:
10.1098/rspb.2001.1647
Iwasa Y, Pomiankowski A (1994) The evolution of mate preferences for multiple handicaps. Evol Int J Org
Evol 48:853–867. doi:10.2307/2410492
Iwasa Y, Pomiankowski A (1999) Good parent and good genes models of handicap evolution. J Theor Biol
200:97–109. doi:10.1006/jtbi.1999.0979
Iwasa Y, Pomiankowski A, Nee S (1991) The evolution of costly mate preferences. II. The ‘‘handicap’’
principle. Evol Int J Org Evol 45:1431–1442. doi:10.2307/2409890
Johnstone RA (1995) Sexual selection, honest advertisement and the handicap principle: reviewing the
evidence. Biol Rev Camb Philos Soc 70:1–65. doi:10.1111/j.1469-185X.1995.tb01439.x
King RC (1970) Ovarian development in Drosophila melanogaster. Academic Press, New York
Kokko H, Jennions MD, Brooks R (2006) Unifying and testing models of sexual selection. Annu Rev Ecol
Syst 37:43–66. doi:10.1146/annurev.ecolsys.37.091305.110259
Lande R (1980) Sexual dimorphism, sexual selection, and adaptation in polygenic characters. Evol Int J Org
Evol 34:292–305. doi:10.2307/2407393
Lande R, Arnold SJ (1985) Evolution of mating preference and sexual dimorphism. J Theor Biol 117:651–
664. doi:10.1016/S0022-5193(85)80245-9
Lorch P, Wilkinson GS, Reillo RP (1993) Copulation duration and sperm precedence in the stalk-eyed fly
Cyrtodiopsis whitei. Behav Ecol Sociobiol 32:303–311. doi:10.1007/BF00183785
Maynard Smith J, Harper D (2003) Animal signals. Oxford University Press, Oxford
Meier R, Baker RH (2002) A cladistic analysis of Diopsidae (Diptera) based on morphological and DNA
sequence data. Insect Syst Evol 33:325–336
Panhuis TM, Wilkinson GS (1999) Exaggerated eyespan influences male contest outcome in stalk-eyed flies.
Behav Ecol Sociobiol 46:221–227. doi:10.1007/s002650050613
94 Evol Ecol (2010) 24:83–95
123
Pomiankowski A (1987) Sexual selection: the handicap principle does work—sometimes. Proc R Soc Lond
B Biol Sci 231:123–145
Pomiankowski A (1988) The evolution of female mate preferences for male genetic quality. Ox Surv Evol
Biol 5:136–184
Rogers DW, Baker RH, Chapman T et al (2005a) Direct and correlated responses to artificial selection on
male mating frequency in the stalk-eyed fly Cyrtodiopsis dalmanni. J Evol Biol 18:642–650. doi:
10.1111/j.1420-9101.2004.00860.x
Rogers DW, Chapman T, Fowler K et al (2005b) Mating-induced reduction in reproductive organ size in the
stalk-eyed fly Cyrtodiopsis dalmanni. BMC Evol Biol 5:37. doi:10.1186/1471-2148-5-37
Rogers DW, Grant CA, Chapman T et al (2006) The influence of male and female eyespan on fertility in the
stalk-eyed fly Cyrtodiopsis dalmanni. Anim Behav 72:1363–1369. doi:10.1016/j.anbehav.2006.03.027
Rogers DW, Denniff M, Chapman T et al (2008) Male ornament size signals reproductive quality in the
stalk-eyed fly Teleopsis dalmanni. BMC Evol Biol 8:236
Sgro
`CM, Hoffmann AA (2004) Genetic correlations, trade-offs and environmental variation. Heredity
93:241–248. doi:10.1038/sj.hdy.6800532
Simmons LW, Emlen DJ (2006) Evolutionary trade-off between weapons and testes. Proc Natl Acad Sci
USA 103:16346–16351. doi:10.1073/pnas.0603474103
Simmons LW, Zuk M, Rotenberry JT (2005) Immune function reflected in calling song characteristics in a
natural population of the cricket Teleogryllus commodus. Anim Behav 69:1235–1241. doi:
10.1016/j.anbehav.2004.09.011
Wilkinson GS (1993) Artificial selection alters allometry in the stalk-eyed fly Cyrtodiopsis dalmanni
(Diptera: Diopsidae). Genet Res Camb 62:213–222
Wilkinson GS, Dodson GN (1997) Function and evolution of antlers and eye stalks in flies. In: Chloe J,
Crespi B (eds) The evolution of mating systems in insects and arachnids. Cambridge University Press,
Cambridge, pp 310–328
Wilkinson GS, Reillo PR (1994) Female choice response to artificial selection in an exaggerated male trait
in a stalk-eyed fly. Proc R Soc Lond B Biol Sci 225:1–6. doi:10.1098/rspb.1994.0001
Wilkinson GS, Amitin EG, Johns PM (2005) Sex-linked correlated responses in female reproductive traits to
selection on male eye span in stalk-eyed flies. Integr Comp Biol 45:500–510. doi:10.1093/icb/45.3.500
Zahavi A (1975) Mate selection—a selection for a handicap. J Theor Biol 53:205–214. doi:10.1016/
0022-5193(75)90111-3
Zuk M (1988) Parasite load, body size, and age of wild-caught male field crickets (Orthoptera: Gryllidae):
effects on sexual selection. Evol Int J Org Evol 42:969–976. doi:10.2307/2408912
Evol Ecol (2010) 24:83–95 95
123
