Male mate preference for female eyespan and fecundity in the stalk-eyed fly, Teleopsis dalmanni

Abstract

Traditional views of sexual selection view males as the indiscriminate sex, competing for access to choosy females. It is increasingly recognized that mating can also be costly for males and they are therefore likely to exhibit choice in order to maximize their reproductive success. Stalk-eyed flies are model species in sexual selection studies. Males are sperm limited and constrained in the number of matings they are able to partake in. In addition, variation in female fecundity has been shown to correlate positively with female eyespan, so eyespan size could provide males with a reliable signal of female reproductive value. We examined male mate preference in the wild in the stalk-eyed fly, Teleopsis dalmanni. In addition, we set up experiments in the laboratory allowing males a choice between females that varied in 1) eyespan (a proxy for fecundity) and/or 2) fecundity (manipulated through diet). We found that males exhibited preference for large eyespan females, both in the wild and laboratory studies. As well as using female eyespan as a mating cue, males were also able to assess female fecundity directly. Changes in fecundity among large eyespan females caused corresponding changes in male mate preference, whereas changes in the fecundity of small eyespan females had limited effect on their attractiveness. These results show that male mate preferences are a prevalent feature of a canonical example of female mate choice sexual selection and that males use multiple cues when they assess females as potential mates.

Full text

© The Author 2014. Published by Oxford University Press on behalf of
the International Society for Behavioral Ecology. All rights reserved. For
permissions, please e-mail: journals.permissions@oup.com
The ocial journal of the
ISBE
International Society for Behavioral Ecology
Behavioral
Ecology
Original Article
Male mate preference for female eyespan and
fecundity in the stalk-eyed fly, Teleopsis dalmanni
Alison J.Cotton,
a,b,c
SamuelCotton,
a
JenniferSmall,
a
and AndrewPomiankowski
a,b
a
Department of Genetics, Evolution & Environment, University College London, Gower Street,
London WC1E 6BT, UK,
b
CoMPLEX, University College London, Gower Street, London WC1E 6BT, UK,
and
c
Bristol Zoological Society, c/o Bristol Zoo Gardens, Clifton Down, Clifton, Bristol BS8 3HA, UK
Received 9 May 2014; revised 29 September 2014; accepted 2 October 2014.
Traditional views of sexual selection view males as the indiscriminate sex, competing for access to choosy females. It is increasingly
recognized that mating can also be costly for males and they are therefore likely to exhibit choice in order to maximize their reproduc-
tive success. Stalk-eyed flies are model species in sexual selection studies. Males are sperm limited and constrained in the number of
matings they are able to partake in. In addition, variation in female fecundity has been shown to correlate positively with female eye-
span, so eyespan size could provide males with a reliable signal of female reproductive value. We examined male mate preference in
the wild in the stalk-eyed fly, Teleopsis dalmanni. In addition, we set up experiments in the laboratory allowing males a choice between
females that varied in 1)eyespan (a proxy for fecundity) and/or 2)fecundity (manipulated through diet). We found that males exhib-
ited preference for large eyespan females, both in the wild and laboratory studies. As well as using female eyespan as a mating cue,
males were also able to assess female fecundity directly. Changes in fecundity among large eyespan females caused corresponding
changes in male mate preference, whereas changes in the fecundity of small eyespan females had limited effect on their attractive-
ness. These results show that male mate preferences are a prevalent feature of a canonical example of female mate choice sexual
selection and that males use multiple cues when they assess females as potential mates.
Key words: fecundity, female ornament, male mate preference, mate choice, sexual selection, stalk-eyed fly.
INTRODUCTION
Sexual selection has been viewed classically as competition among
undiscriminating males for access to choosy females (Darwin 1871;
Bateman 1948; Trivers 1972). However, it is increasingly recog-
nized that this perspective is too simplistic and that mating can
be constrained or costly for males (Dewsbury 1982; Bonduriansky
2001; Webberley and Hurst 2002; Wedell et al. 2002; Andrade
2003). Males should, therefore, allocate their matings prudently so
as to maximize their reproductive success. This leads to the pre-
diction that it will often pay males, as for females, to discriminate
between individuals when choosing a mate (Bonduriansky 2001).
Parker (1983) noted that choosiness is favored when high vari-
ance in quality exists in the opposite sex; if there is little variation
in mate quality, then there will be few advantages accruing from
mate preference (Parker 1983; Gwynne 1991). In promiscuous spe-
cies, males are expected to select females on the basis of fecundity
(Bonduriansky 2001). Nonetheless, directional male mate prefer-
ence is expected to evolve for traits that reflect female reproductive
value even when signaling compromises female viability and male
mate preference is costly (Servedio and Lande 2006; Nakahashi
2008). It has been suggested that males may use female body size
as a proxy for fecundity, as size and fecundity tend to be corre-
lated (Honěk 1993). In several polygynous species, females display
ornament-like traits that may have initially evolved as a correlated
response to selection on homologous male ornaments (Lande and
Arnold 1985). These may subsequently have been co-opted as tar-
gets of male mate preference, given that such traits in females are
conspicuous, easy to evaluate, and often reflect aspects of female
quality linked to fecundity (Amundsen 2000). Male mating prefer-
ence for female ornaments has been shown in several insect species
where the ornament is thought to be an indicator of fecundity, like
dance flies (Funk and Tallamy 2000; LeBas et al. 2003) and the
mosquito Sabethes cyaneus (South and Arnqvist 2011).
The evolution of male mate preference is also likely to be
influenced by the constraints and costs arising from mating
(Bonduriansky 2001). If males are able to mate cost free with every
female they encounter, there is little reason for discrimination to
evolve. For preference to be favored, male mating investment must
be subject to limits. It is increasingly recognized that sperm and
ejaculates are not unlimited resources, but rather they become
depleted by repeated mating. This sets limits to the mating rate
Address correspondence to A.Pomiankowski. E-mail: ucbhpom@ucl.ac.uk.
Behavioral Ecology (2014), 00(00), 1–10. doi:10.1093/beheco/aru192
Behavioral Ecology Advance Access published November 14, 2014
by guest on November 17, 2014http://beheco.oxfordjournals.org/Downloaded from

Behavioral Ecology
(Dewsbury 1982; Wedell et al. 2002) and to investment in subse-
quent matings (Preston etal. 2001; Wedell etal. 2002). Mating pref-
erence for particular female phenotypes implies lower male mating
success (Servedio and Lande 2006; Nakahashi 2008). As increasing
numbers of males court or mate with attractive females, each has a
smaller chance of success (i.e., of mating and of paternity). On the
other hand, the costs associated with finding and assessing females
must not be prohibitive (Nakahashi 2008). Selection will favor the
evolution of male mate preference when the costs of mate search-
ing and sampling are not high, for example, when the distribution
of females is clumped and males are able to assess them easily
(Forsberg 1987).
The Malaysian stalk-eyed fly Teleopsis dalmanni (Diopsidae;
Diptera) is an important model species for studies of sexual selec-
tion (Wilkinson and Dodson 1997; Maynard Smith and Harper
2003). Stalk-eyed flies are characterized by having their eyes dis-
placed laterally from the head on elongate “eyestalks,” in both
sexes. Eyespan (the distance between the eyes) is sexually dimorphic
in T.dalmanni, being much enlarged in males as a result of sexual
selection. In natural populations, T.dalmanni form nocturnal lekking
aggregations on root hairs that hang underneath the eroded banks
of rainforest streams (Burkhardt and de la Motte 1985; Wilkinson
and Dodson 1997; Cotton et al. 2010). Males fight for control of
these roosting sites (Wilkinson and Dodson 1997; Small etal. 2009),
and females prefer to alight on root hairs controlled by males with
large eyespan (Wilkinson and Reillo 1994; Hingle et al. 2001a,
2001b; Cotton etal. 2010). The vast majority of matings occur in
these aggregations during the dawn and dusk period, when males
attempt to mate with females in their harem (Burkhardt and de
la Motte 1988; Lorch etal. 1993; Small et al. 2009; Cotton etal.
2010). So T.dalmanni appears to be a textbook example of harem-
based polygyny, with choosy females and competitivemales.
However, evidence suggests that there is a high potential for male
mate preference in T.dalmanni. Laboratory experiments have shown
that female fecundity is sensitive to environmental (dietary) stress
(Hingle et al. 2001a), suggesting that variance in female quality is
high. This is borne out in the wild, where female fecundity is highly
variable (Cotton etal. 2010). Moreover, female eyespan is an accu-
rate and reliable indicator of female fecundity in wild females, even
after controlling for its covariation with body size (Cotton et al.
2010). Female eyespan in T. dalmanni, like the ornamental homo-
logue in males, is prominent and easily assessable suggesting that it
may serve as a useful cue for males to use in mating decisions. In
the related African stalk-eyed fly species Diasemopsis meigenii, males
mate for longer and transfer more sperm to females with larger eye-
span (Harley etal. 2013). As female eyespan is positively correlated
with fecundity in D.meigenii, males could gain a selective advantage
by investing more in large eyespan females.
In addition, there is evidence that male stalk-eyed flies suer con-
straints on multiple mating. In T.dalmanni, male mating frequency
is correlated both phenotypically and genetically with the size of
the accessory glands, the paired internal organs involved in sper-
matophore production (Baker et al. 2003; Rogers, Baker, et al.
2005). Accessory glands become depleted with successive matings
(Rogers, Chapman, et al. 2005). When males with larger acces-
sory glands are allowed to mate with multiple females over a short
period of time, they confer higher fertility on females, probably
because they mate at a higher rate (Rogers etal. 2008). In D.mei-
genii, males with larger eyespan (and hence larger accessory glands
and testes) show smaller reductions in spermatophore size and
number of sperm transferred over successive matings, relative to
the performance of small eyespan males (unpublished data). Taken
together, these experimental results suggest that there are limita-
tions on the mating rate of males that have attracted many females
to their lek sites. In the dawn period, when most mating occurs,
males have about 20–30 min in which to mate with females in their
harem. Yet typically, some females are observed to disperse from
lek sites before mating with the dominant male. Under such con-
straints, we expect that there is the opportunity for males to direct
their mating attempts toward those females that have the highest
reproductivevalue.
Given these features of the mating system in T. dalmanni, and
the fact that the harem-based mating system in this species allows
males to choose among groups of females with relatively low costs
of search and assessment, we hypothesized that males should dis-
criminate between females on the basis of their fecundity. We
investigated the potential for male mate preference in the flies’
native habitat in the Malaysian rainforest and also under labora-
tory conditions. This combined approach allowed us to examine
male mate choice both under biologically realistic conditions and
under controlled, experimentally manipulated conditions. In the
latter case, we manipulated the eyespan of experimental females in
order to assess whether this trait is used in male choice and alter-
ing the adult diet of females in order to vary female fecundity in a
controlled manner. We then used a series of mate choice tests to
examine whether males were primarily using female eyespan as an
indicator of fecundity or were able to assess fecundity directly.
METHODS
Male mate preference in field conditions
Field data were collected during 2 phases of fieldwork carried out
on a T. dalmanni population at Ulu Gombak, Peninsular Malaysia
(3°19′N, 101°45′E) during July/August in 2006 and 2007. Lek sites
(exposed root hairs) on the banks of a tributary of the Gombak river
were identified after dusk. Observations of male mating behavior
at focal harems were conducted during the following dawn period,
starting at approximately 06:55 when flies were still quiescent and
ending at approximately 07:45 when flies had usually dispersed into
the forest. In a few instances (N=3), more than 1 male was present
on the root hair. In order to obtain information from a single focal
male per harem, we carefully removed the additional males without
disturbing the other individuals (the most easily removed male was
chosen). We noted the harem size (number of females present) and
the frequency of successful matings with each female, defined as a
copulation ≥ 30 s, as shorter copulations do not usually result in
transfer of a spermatophore (Lorch etal. 1993; Rogers etal. 2006).
Females within each harem were categorized by inspection by
observers into large and small eyespan classes. Observers (S.C.and
J.S.) were experienced in judging fly size from prior experience with
field populations of T.dalmanni. Arelative measure was used to cat-
egorize females within a lek. Medium-sized eyespan females were
classified as small when the lek contained larger females and large
when there were no larger eyespan females. Data from harems in
which there were no observable dierences in female size were
excluded from the analysis. Eyespan of the focal male was mea-
sured in situ noninvasively using standardized digital photographs.
Images were taken with a digital SLR camera (Canon EOS 350D)
through a 180-mm macro lens set to its minimum focal distance,
which creates a fixed distance between the camera and the subject.
The focal male was kept perpendicular to the camera by keeping
Page 2 of 10
by guest on November 17, 2014http://beheco.oxfordjournals.org/Downloaded from

Cotton etal. • Male mate preference, female ornament, and fecundity
both eye bulbs in focus. Eyespan was then estimated from the
size of the resultant image relative to a known standard, photo-
graphed under identical conditions. This method is highly accurate
compared with controlled measurements of the same individuals
(repeatability > 0.93; Lessells and Boag 1987; Small etal. 2009).
Laboratory studies—origin of flies and
generation of flies for experiments
The flies used were from a population collected in Ulu Gombak,
Peninsular Malaysia (3°19′N, 101°45′E) in 2005 (S.C. and A.P.).
They have since been maintained in the laboratory in cage culture
(>200 individuals to minimize inbreeding) at 25 °C on a 12:12 h
light:dark cycle and fed pureed sweet corn twice weekly. Fifteen-
minute artificial dawn and dusk periods were created by illumina-
tion from a single 60-W bulb, at the start and end of the lightphase.
To obtain experimental flies, we collected eggs from the cage
cultures over a 3-week period and reared larvae on a variable
amount of pureed sweet corn to ensure high variance in eyespan
(David et al. 1998; Cotton et al. 2004a). On eclosion, flies were
anesthetized on ice and measured for their eyespan, defined as
the distance between the outermost lateral edges of the eye bulbs
(Cotton etal. 2004a). Following Rogers etal. (2006), females were
separated into large and small size classes, defined as having eye-
spans >5.8 or <5.4 mm, respectively. Intermediate size females
were discarded. To control for the well-documented eects of male
eyespan on female mating behavior (Wilkinson and Reillo 1994;
Hingle et al. 2001a, 2001b), we used only large eyespan males
(>8.5 mm) with a low sample variance (N=36, mean ± standard
deviation [SD]=8.94 ± 0.31 mm) in the subsequent assays of pref-
erence. In addition, previous work has shown that small eyespan
females are less able to discriminate among variation in male eye-
span (Hingle etal. 2001a), and this eect may extend to male abil-
ity to discriminate. Finally, because small eyespan males have fewer
opportunities to mate under field conditions (Cotton etal. 2010),
their mate preferences may dier compared with large eyespan
males. Though this is a topic of interest, it is beyond the scope of
the current study.
Male mate preference for female eyespan
Male mate preference under laboratory conditions was examined
using a specially designed cage. This comprised two 500-mL trans-
parent plastic pots, one inverted on top of the other and separated
by a pair of opaque removable partitions (Figure 1). A roosting
string hung from the ceiling of the upper pot extending down to
near the base of the lower pot. The base of the test cage contained
moist cotton wool to maintain high humidity. The focal male and
the pair of tester females, 1 large and 1 small, were introduced
during late afternoon on the day prior to the assay. The male was
placed in the upper half of the cage and the females in the lower
half of the cage, and the partitions were inserted to keep the sexes
segregated (Figure1).
At the beginning of the dawn period on the following morn-
ing, the partitions were removed allowing the flies to interact. Male
mating behavior was observed for 30 min. A mating was defined
as a copulation lasting ≥ 30 s (as in the field study). Asample of
N = 36 males were assayed for mate preference. Large and small
females were drawn at random from a population of each type of
tester female (N > 25 for each type). Tester females were therefore
used more than once in the trials. However, they were never used
more than once in any 48-h period.
All individuals used in the experiment were nonvirgins, hav-
ing been kept in mixed-sex groups prior to the mate preference
experiment. They were collected over a short period of time (3
weeks) and so were of similar age when used in the experiment
(~8+ weeks). Female T.dalmanni are highly promiscuous and mate
at high frequencies (Wilkinson et al. 1998; Rogers, Baker, et al.
2005). Female reproductive life span is in the order of months
(Wilkinson and Reillo 1994; Reguera etal. 2004), so the incidence
of female virgins under natural conditions is rare, which we aimed
to mimic in this design. Females lay eggs continually after reach-
ing maturity, irrespective of whether they have mated and/or are
fertilized (Baker et al. 2001; Reguera et al. 2004). Moreover, the
frequency of matings among mated females has no detectable
eect on egg output (Baker et al. 2001). So, we consider details
of female mating history prior to the experiment will have had a
minimal influence on the egg-laying rate. Likewise, male virgins
are also rare in nature, as they also mate promiscuously and live
for months (Wilkinson and Reillo 1994; Reguera etal. 2004). So,
we also housed our experimental males in mixed-sex groups. This
procedure further allowed comparison between field and labora-
tory experiments.
To examine male fitness benefits of mating with large and small
females, we measured the fecundity of a sample of females using a
previously developed protocol (Cotton et al. 2006), after the mate
preference assays were complete. Tester females (N=23 large and
N = 22 small females) were housed individually in 500-mL pots
with a roosting string hanging from the top and moist tissue paper
and a food tray at the base. The tissue and food were removed from
the containers every 2–3days and all the eggs on both substrates
were counted. Females were allowed to acclimatize in their new
pots for 3days, and their fecundity was measured over the subse-
quent 11days (N=5 collections per female).
Roosting
string
Focal
male
Tester
females
Removable
partitions
Figure1
Apparatus used for male mate preference assays in laboratory experiments.
A focal male was placed in the upper section and 2 tester females in the
lower section. The sexes were separated by removable partitions (cardboard)
until testing commenced. Asingle string resembling a rootlet runs the whole
length of the cage, providing a suitable roosting site.
Page 3 of 10
by guest on November 17, 2014http://beheco.oxfordjournals.org/Downloaded from

Behavioral Ecology
Male mate preference for female eyespan and
fecundity
We examined the relative importance of female eyespan and fecun-
dity in determining male mate preference under laboratory condi-
tions. Large and small female eyespan classes were set up as defined
above. Fecundity was experimentally manipulated by placing females
on a reduced quality diet for 2 weeks prior to the start of the mate
preference assays and then throughout the remainder of the experi-
ment (N=50 for each eyespan class). The reduced diet consisted of
20% corn to 80% sucrose. The eect of this diet on fecundity was
compared with that for females on a high-quality diet of 80% corn
to 20% sucrose. To ensure that all the food had the same viscos-
ity, an indigestible bulking agent, carboxymethycellulose (3% w/v),
was added to the sucrose (25% w/v) solution (Rogers et al. 2008).
Half of the females from each eyespan class were placed on each
diet. To characterize the eect of the diet manipulation on egg pro-
duction, we assessed the fecundity of females after the completion
of the mate preference assays. Eggs were collected (as above) from
pots (N = 15) each containing 10 females (in each pot, all females
were either on the high or reduced quality diet). Eggs were collected
every 2–4days over a 15-day period (N=6 collections per female).
We examined male mate preference across 5 dierent treatment
groups (Table 1) using the same mate preference assays described
above. Treatments 1 and 2 were control treatments for high and
reduced fecundity, respectively, while varying eyespan. Treatments
3 and 4 controlled for large and small eyespan, respectively, while
varying fecundity. Treatment 5 manipulated both fecundity and
eyespan, by giving each focal male the choice between a large
female with reduced fecundity and a small female with high fecun-
dity. Asample size of N=28 males was set up for each treatment.
Statistical analysis
Male mate preference was assessed using an index based on the
dierence between the observed and expected numbers of copula-
tions with large females. In the field study, P
Field
was calculated for
individual males and allowed for multiple copulations and variable
numbers of large and small females on the lek,
Pc
n
Nc c
i
i
i
t
FieldL
L
LS
=−










+
=
∑
1
1
,
where c
L
is the number of copulations with large eyespan females
and c
S
is the number of copulations with small eyespan females, n
Li
is the number of large eyespan females in the harem at the time
of mating i, and N
i
is the total number of females in the harem at
the time of mating i. This index takes into account the changing
composition of leks through time, as females occasionally flew away
between matings by the focal male. P
Field
is 0 under random mating,
P
Field
> 0 for males showing mating preference for large eyespan
females, and P
Field
< 0 for males showing preferences for small eye-
span males. The minimum/maximum values of P
Field
always dier
by 1 but are not necessarily symmetric about 0 due to the distribu-
tion of large and small females on the lek throughtime.
P
Field
values were not normally distributed, so we tested whether
the mean of the distribution of individual male P
Field
values was
dierent from 0 using a Wilcoxon signed-rank test. In harems in
which more than 1 mating was observed, we tested whether the
observed probability of a mating with a large eyespan female in
the ith mating attempt was more likely than expected by chance,
calculating the expected mating probability n
Li
/N
i
for that mating.
We used a repeated measures approach to evaluate variation in
preference across matings by testing whether the mean P
Field(mating
i)
− P
Field(mating i+1)
value was significantly dierent from 0 using a
Wilcoxon signed-ranktest.
In the studies of male mate preference under laboratory condi-
tions, we used a similar index of male mate preference. This was
simpler as females could not depart from the test cage, and there
was always 1 large and 1 small female perpot,
P
cc
cc
Lab
LS
LS
=
−
+2( )
.
As for the field index, P
Lab
equals 0 under random mating, P
Lab
>
0 for preference for large females, and P
Lab
< 0 for preference for
small females. But in this case, the minimum/maximum values of
P
Lab
are ±0.5 and are symmetric about 0.We used the same index
when the 2 females diered in fecundity, substituting the copulation
rate of females with high (c
H
) or reduced (c
R
) fecundity for those of
large (c
L
) and small (c
S
) females. As before, we tested whether the
distribution of individual P
Lab
scores had a mean that was signifi-
cantly dierent from 0 using Wilcoxon signed-rank tests and used
similar procedures to investigate whether there was change in pref-
erence across subsequent matings.
In the second laboratory study in which females diered in
fecundity and eyespan, we examined the ability of males to distin-
guish high fecundity females, using female eyespan as a covariate.
Pairs of treatments in which fecundity diered but the eyespan of
both females was large (treatment 3) or small (treatment 4) were
combined, after separate analysis of each treatment.
In the first laboratory experiment in which females diered
in eyespan alone, we examined potential fecundity dierences
between large and small females by performing a general linear
model (GLM) on the number of eggs laid per female (5 repeated
measures), nesting female identity within the eyespan variable (large
or small). In the second laboratory experiment, we evaluated the
eect of the diet manipulation by examining fecundity of females
on the 2 diet treatments. To do this, we performed a GLM on each
of the number of eggs laid per pot (6 repeated measures), nesting
pot identity within diet manipulation.
All statistical analyses were performed using JMP V.10.0.0 (SAS
Institute, Cary, NC).
RESULTS
Male mate preference in field conditions
Just over half of observed males mated multiply under field conditions
at dawn (13/25) with a mating frequency, mean ± SD=1.52 ± 0.51
Table1
Examination of multiple signals used in mate choice
Treatment Female eyespan Female fecundity
1 Large or small Both high
2 Large or small Both reduced
3 Both large High or reduced
4 Both small High or reduced
5 Large or small Reduced (large eyespan)
or high (small eyespan)
Attributes of paired females presented to focal males in each treatment
group. Females potentially diered in eyespan (large or small) and/or
fecundity (high or reduced).
Page 4 of 10
by guest on November 17, 2014http://beheco.oxfordjournals.org/Downloaded from

Cotton etal. • Male mate preference, female ornament, and fecundity
(range 1–2). Males preferred to mate with large eyespan females
(P
Field
: mean ± standard error [SE]= 0.24 ± 0.08, Wilcoxon signed-
rank=94.00, degrees of freedom [df]=24, P=0.0074; Figure2).
In harems in which the focal male mated twice, we compared
the eyespan of females across the 2 matings. In the first (Wilcoxon
signed-rank = 42.50, df = 12, P = 0.0012) and second (Wilcoxon
signed-rank=30.50, df=12, P=0.0303) matings, males copulated
more often with large eyespan females than expected given their
frequency in the harem. There was no significant dierence in the
strength of mate preference across the 2 mating attempts (Wilcoxon
signed-rank=−7.50, df=12, P=0.2500).
We found no relationship between male eyespan and P
Field
(F
1,15
 = 0.0023, P = 0.9622), suggesting that our results are not
confounded by any eect of the focal male’s eyespan on male or
female behavior. The number or types of female in a focal male’s
harem could have influenced his ability to express mate preference.
However, we found no relationship between the proportion of large
eyespan females in the harem and P
Field
(F
1,15
=1.5589, P=0.2310)
or between harem size and P
Field
(F
1,23
=0.6375, P=0.4328).
Male mate preference for female eyespan
Within the half-hour period allowed, most males (33/36) mated
multiply (mating frequency: mean ± SD = 3.97 ± 1.63, range
1–7), well in excess of what was typical under natural conditions.
Males preferred to mate with large eyespan females (P
Lab
: mean
± SE = 0.18 ± 0.05, Wilcoxon signed-rank = 155.50, df = 35,
P = 0.0010; Figure 3). This result was not contingent on singly
mated males, as there was still preference for large eyespan females
when singly mated males were excluded (P
Lab
 = 0.19 ± 0.05,
Wilcoxon signed-rank=137.50, df =32, P= 0.0006). There was
no relationship between male eyespan and P
Lab
(F
1,30
 = 0.4871,
P=0.4906), suggesting that by only using large eyespan males, we
had removed any potential confounding eect of variation in eye-
span of the focalmale.
When preference was examined in sequential matings,
we found preference for large eyespan females in the first
(P
Lab
 = 0.17 ± 0.08, N = 36, Wilcoxon signed-rank = 111.00,
P=0.0438) and second (P
Lab
=0.20 ± 0.08, N=33, Wilcoxon
signed-rank = 110.50, P = 0.0212) matings. But the patterns
of the third (P
Lab
 = 0.16 ± 0.09, N = 29, Wilcoxon signed-
rank = 67.50, P = 0.0951) and subsequent matings (Wilcoxon
signed-rank < 34.50, P > 0.2080) were not significantly dif-
ferent than expected under random mating. This may partly
have been due to the reduced sample size of males that mated
more often. But it could have reflected a decline in preference
among males that mated more often. When this was explicitly
tested however we found no association between preference
0
–1.0 - –0.8 –0.8 - –0.6 –0.6 - –0.4 –0.4 - –0.2 –0.2 - 0.0
Male mate preference (P
Field
)
0.0 - 0.2 0.2 - 0.4 0.4 - 0.6 0.6 - 0.8 0.8 - 1.0
2
4
6
Frequency
8
10
12
Figure2
Frequency distribution of P
Field
, the preference function of wild males. P
Field
accounts for the harem size, the number of large and small females available, and
the dynamic changes in these variables between matings. P
Field
=0 indicates no preference, P
Field
< 0 indicates a preference for small eyespan females, and
P
Field
> 0 indicates a preference for large eyespan females.
Page 5 of 10
by guest on November 17, 2014http://beheco.oxfordjournals.org/Downloaded from

Behavioral Ecology
and the number of matings a male engaged in (F
1,34
= 3.0292,
P = 0.0908).
Females from the large eyespan group laid 36% more eggs per
day than those from the small eyespan group (F
1,180
 = 9.4249,
P = 0.0025; mean ± SE daily egg output, large = 4.76 ± 0.49,
N=23; small=3.49 ± 0.50, N=22).
Male mate preference for female fecundity and
eyespan
We manipulated fecundity through diet, using either a reduced
(20% corn) or high-quality diet (80% corn). Females on the reduced
diet manipulation laid fewer eggs than those on the high-quality
diet (F
1,75
 = 111.7045, P < 0.0001; mean ± SE daily egg output,
reduced=0.2640 ± 0.02, N=7, high quality=1.99 ± 0.10, N=8).
In line with the first laboratory study of male mate prefer-
ence (see above), most males mated multiply (128/138), but at an
even higher rate (mating frequency: mean ± SD = 6.33 ± 3.12,
range 1–16). In the control treatments where diet was standard-
ized but eyespan varied (treatments 1 and 2, Table 1), there was
a significant preference for large eyespan females (P
Lab
: mean
± SE = 0.25 ± 0.08, Wilcoxon signed-rank = 307.50, df = 55,
P= 0.0029). We found a significant dierence between the 2 diet
treatments, with stronger male mate preference for large eyespan
females when females were on the high-quality diet (P
Lab
(high-
quality diet) = 0.41 ± 0.10, P
Lab
(reduced diet) = 0.09 ± 0.11,
χ
2
(df = 1) = 4.4544, N=56, P=0.0348; Figure4).
We further investigated how males responded to dierences
in female fecundity by varying diet and standardizing eyespan
(treatments 3 and 4, Table 1). Overall, there was a significant
preference for fecund females in the absence of eyespan varia-
tion (P
Lab
=0.21 ± 0.09, Wilcoxon signed-rank= 217.50, df=54,
P = 0.0275). There was no dierence in preference when males
were presented with either 2 large or 2 small eyespan females
(
χ
1
2
1 4418=
.,
N=55, P=0.2298).
Finally, we studied how changes in fecundity aected prefer-
ence for large eyespan females. Comparing male mate prefer-
ence when only large eyespan females diered in diet (treatments
1 and 5, Table 1), we found overall preference was for large eye-
span females (P
Lab
 = 0.24 ± 0.08, Wilcoxon signed-rank = 252.50,
df = 54, P = 0.0031), and this was stronger when the large eye-
span female had high fecundity (P
Lab
(treatment one)=0.41 ± 0.10,
P
Lab
(treatment 5) = 0.06 ± 0.11,
χ
2
(df = 1) = 4.9108, N = 55,
P = 0.0267). In contrast, comparing male mate preference when
only small eyespan females diered in diet (treatments 2 and 5,
Table1), we found no overall preference for large eyespan females
(P
Lab
 = 0.07 ± 0.08, Wilcoxon signed-rank = 87.50, df = 54,
P = 0.3743) or any dierence in preference between the treat-
ments when the small eyespan female had high or low fecundity
(
χ
1
2
0= 0.041 ,
N=55, P=0.8396).
DISCUSSION
Mate preference by males is predicted when there is exploitable
variation in female quality, limited male mating capacity, and low
costs of finding and assessing mates (Bonduriansky 2001). These
conditions are met in the stalk-eyed fly T. dalmanni: females vary
considerably in fecundity (Cotton et al. 2010), males have limited
ability to mate multiply over short periods of time (Rogers, Baker,
0
–0.5 - –0.4 –0.4 - –0.3 –0.3 - –0.2 –0.2 - –0.1 –0.1 - 0.0
Male mate preference (P
Lab
)
0.0 - 0.1 0.1 - 0.2 0.2 - 0.3 0.3 - 0.4 0.4 - 0.5
2
4
6
Frequency
8
10
12
14
Figure3
Frequency distribution of P
Lab
, the preference function of laboratory males, when given the choice of mating with either a large or small female. In the
laboratory assays, there is no dynamic change in the number of females available because females cannot leave the test arena. For P
Lab
, P=0 indicates no
preference, P
Lab
< 0 indicates a preference for small eyespan females, and P
Lab
> 0 indicates a preference for large eyespan females.
Page 6 of 10
by guest on November 17, 2014http://beheco.oxfordjournals.org/Downloaded from

Cotton etal. • Male mate preference, female ornament, and fecundity
et al. 2005; Rogers, Chapman, et al. 2005), and the lek mating
system means that several females congregate closely with males
who can choose the order in which they mate with them. These
flies are well known for female mate preference for males with
large eyespans (Wilkinson and Reillo 1994; Cotton etal. 2010), so
we hypothesized that males might also use variation in eyespan to
assess females. This is logical because female eyespan has several
of the properties usually associated with sexual ornaments. Firstly,
female eyespan in sexually dimorphic stalk-eyed fly species is an
exaggerated trait in comparison to sexually monomorphic species
(Baker and Wilkinson 2001). Secondly, female eyespan is sensitive
to stress compared with nonsexual traits (e.g., wing size), even after
controlling for body size (Cotton etal. 2004b). In addition, female
eyespan is a reliable indicator of fecundity even after controlling for
the influence of body size (Cotton etal. 2010).
Male mate preference in T.dalmanni was investigated both in the
wild and in controlled laboratory experiments. The vast majority
of previous studies examining male mate preferences, especially
in Drosophila, have been performed in the laboratory (e.g., Byrne
and Rice 2006; Edward and Chapman 2013), and thus we have
remained largely ignorant on the importance of male mate prefer-
ence under natural conditions. Indeed, some authors have noted
that laboratory studies involving simultaneous choice tests may
result in inflated preference estimates as the abundance of “mates”
they encounter in the lab far exceeds that found in the wild (Barry
and Kokko 2010). Thus, our study provides important data on male
mating preferences under natural conditions. In the wild, males
arrive at lek sites at early dusk. Males fight to be the sole lek holder,
with the largest male typically being successful (Small etal. 2009).
Females then arrive and choose which lek to join and roost on, with
large eyespan males attracting more females (Cotton et al. 2010).
Males defend their harem during dusk from intrusions and mating
attempts by other nonlek holding males. The majority of the mat-
ings occur the following morning before flies disperse (Burkhardt
and de la Motte 1988; Lorch etal. 1993). In our wild leks, we found
that males with multifemale harems mated more frequently with
the largest eyespan females in the harem. This eect was indepen-
dent of harem size. We defined large female eyespan as the larg-
est eyespan available for the focal male to mate with, on the basis
that his assessment would be among those females in his harem.
In parallel, under experimentally controlled laboratory conditions,
we confirmed male mate preference for large female eyespan. In
these experiments, we gave males limited choices between pairs
of females, one with large and the other with small eyespan, and
restricted the dispersal of females. A large proportion of copula-
tions in the wild occur within 20–30 min of dawn (Lorch et al.
1993), and our laboratory experiments mirrored this, considering
only a 30-min window. We also constrained male eyespan in the
laboratory experiment, only using males that had large eyespan. So,
male preference may be dierent in small eyespan males though
our data from the field (where male eyespan was unconstrained) did
not reveal any variation in preference with male eyespan.
Given that female eyespan covaries positively with body size
(David et al. 1998), we cannot discount that our observation of
male mate preference for female eyespan arose indirectly from male
mate preference for large-bodied females. However, 2 lines of evi-
dence suggest that eyespan, rather than body size, is likely to be the
main cue that males use in their choice of mate. First, flies assess
each other face-on, meaning that the laterally elongated eyestalks
are more readily assessed than body size, which would require flies
High Fecundity
Male mate preference (P
Lab
)
–1.5
–1
–0.5
0
0.5
1
1.5
Low Fecundity
Treatment
Figure4
The eect of fecundity on male mate preference (independent of eyespan). Male mate preference for large eyespan females (P
Lab
), when females were fed a
high-quality diet and had high fecundity (treatment one) or a reduced diet and had low fecundity (treatment 2). There was stronger male mate preference
when females were on the high-quality diet. The line represents the mean preference of the 2 diet treatments. *P<0.05.
Page 7 of 10
by guest on November 17, 2014http://beheco.oxfordjournals.org/Downloaded from

Behavioral Ecology
to be oriented perpendicular to each other. Second, female eyespan
is a condition-dependent trait (Cotton etal. 2004b) and is a more
accurate signal of fecundity than body size alone (Cotton et al.
2010). It is also possible that cues other than eyespan, such as subtle
behavioral cues or chemical signals (Thomas 2011), influence mat-
ing behavior and might allow females to indicate their reproduc-
tive value to males. The disentanglement of such highly correlated
traits is a general problem faced by researchers of mate preference
in sexually dimorphic species (Hedrick and Temeles 1989).
Previous work has shown that males sire more ospring (fertile
eggs) following a single mating with a large eyespan female (Rogers
etal. 2006). Similarly, in wild-caught T.dalmanni (Cotton etal. 2010)
and in this laboratory experiment, female eyespan was a good indi-
cator of fecundity, and this is also the case in a related stalk-eyed fly
species, D. meigenii (Harley etal. 2013). These findings echo other
studies that have demonstrated male mate preference for females
with large body size or for female ornamental trait values that are
good predictors of female fecundity (Amundsen 2000; Amundsen
and Forsgren 2001; Bonduriansky 2001; Doutrelant et al. 2008;
Baldauf et al. 2011; Potti et al. 2013). Our own and these other
studies suggest that males with mating preference for large orna-
ments will, all things being equal, sire more ospring compared
with males who mate at random. However, fecund females are
likely to attract more matings by males, thereby increasing the
potential for sperm competition and diluting the gain in paternity
stemming from any particular male or particular mating. This has
been the subject of some theoretical consideration (Servedio and
Lande 2006; Nakahashi 2008), and our experimentation does not
directly address the range of fitness benefits and costs that follow
from male mate preference. For instance, males choosing larger
females may have a reduced likelihood of fertilizing any particular
egg, but this may be compensated by the fact that larger females
lay more eggs. The exact relationship between female ornaments,
female quality, and fitness needs to be elucidated through experi-
mentation in which the paternity gain to a male mating with an
attractive, large eyespan female is compared with a mating with a
less attractive small eyespan female. Afull analysis will also need to
consider the genetic and environmental inputs to female fecundity
and any interaction between these 2 factors. This will require fur-
ther analysis both in the field as well as using laboratory manipula-
tive experimentation.
In a further experiment, we manipulated diet as a way of alter-
ing fecundity independently of female eyespan (Hingle etal. 2001a,
2001b). Flies on reduced quality food had relatively low fecundity.
By constraining female eyespan while manipulating diet, we were
able to show male mate preference for females with higher fecun-
dity per se. These results lend support to the idea that males are
using multiple cues when assessing females. Perhaps males detect
the distension of the female abdomen that occurs when it harbors
many mature eggs. Another possibility is that females signal their
fecundity through scent as has been shown in other insects (Peeters
et al. 1999; Mitra and Gadagkar 2012) or other sensory modali-
ties. The use of multiple cues in mate preference decisions, such
as visual, chemical, and behavioral signals, has been the focus of
much interest with a key question being what information they sig-
nal (Candolin 2003; Bro-Jørgensen 2010).
These dietary manipulations also showed that the strength of
preference for fecundity dierences induced by diet did not dif-
fer when both tester females had large eyespan or both had small
eyespan. However, there were interactions between female eyespan
and fecundity. Male mate preference was weakened when the large
eyespan female was put on a reduced quality diet, but there was no
eect on preference of moving the small eyespan female between
diets. These results imply that fecundity dierences have a greater
eect on the attractiveness of large eyespan females than on that of
small eyespan females. However, this needs to be verified by further
investigation, involving direct measures of individual fecundity. In
this context, it is vital to further investigate how fecundity dier-
ences alter preferences among males in thewild.
It could be argued that the distribution of observed copula-
tions results from female behavior rather than male mate prefer-
ence, for example, if large females are more eager to mate. Indeed,
large females do need to mate more frequently than small females
although this has been interpreted as a reflection of their higher
fecundity and hence their need for more copulations to oset the
chronic sperm limitation typical of this species (Baker etal. 2001;
Rogers, Baker, et al. 2005; Cotton et al. 2010). However, several
lines of evidence suggest that eects of female behavior cannot
account (entirely) for the mating biases reported here. If females
compete among themselves for access to a male, then one might
expect that females with the largest eyespans would prevail, and
biased mating distributions would result from intrasexual competi-
tion rather than male mate preference. However, there is no evi-
dence that female eyespan influences contest outcome in female
T. dalmanni (Al-khairulla et al. 2003). In addition, observations of
lek sites reveal no obvious evidence that females compete for access
to males on the lek and it is indeed males who exhibit patrolling
behavior (personal observation). Likewise, we found no evidence
that harem size or the proportion of large females in the harem
correlated with preference in our wild experiment. In addition, in
our laboratory experiments, male eyespan was controlled to avoid
strong female mate preference influencing the outcome. Although
it is not possible to eliminate female eects, it seems likely that the
patterns in our data result primarily from male-controlled biases in
mating.
We have shown, using a combination of field studies and con-
trolled laboratory experiments, that males from a well-known
model species of harem-based polygyny exhibit strong preference
for female traits that indicate fecundity. We also provide evidence
that males can directly assess fecundity when variation in morpho-
logical traits associated with male mate preference is controlled
for. Males use multiple cues in their mate assessment. Future work
should capitalize on these initial findings and seek to explain the
variation that exists in male mate preference and estimate how this
aects the strength of sexual selection on male sexual ornaments.
The eect of male eyespan and condition on male mate preferences
(i.e., whether small eyespan males exhibit a dierence in preference)
should also be examined as condition-dependent male mate prefer-
ences could occur. Future work should also endeavor to understand
the cues used by females to attract male mating to provide a more
complete picture of how sexual selection operates in this species.
FUNDING
We acknowledge support from a variety of sources: A.J.C. by
a UCL IMPACT Studentship administered by the CoMPLEX
Systems Biology Programme, J.S.by a Biotechnology and Biological
Sciences Research Council Studentship, S.C. by a Natural
Environment Research Council (NERC) Research Fellowship (NE/
E012620/1), and A.P. by grants from Engineering and Physical
Sciences Research Council (EP/F500351/1, EP/I017909/1) and
NERC (NE/G00563X/1).
Page 8 of 10
by guest on November 17, 2014http://beheco.oxfordjournals.org/Downloaded from

Cotton etal. • Male mate preference, female ornament, and fecundity
The authors thank H. Eager, C. Liedtke, L. Bellamy, and M. Földvári
for help with data collection, K. Fowler for comments on the article,
and R. Hashim and the sta at the Ulu Gombak Field Research Centre,
University of Malaya, Kuala Lumpur, for their assistance.
Handling editor: Paco Garcia-Gonzalez
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.
Amundsen T. 2000. Why are female birds ornamented? Trends Ecol Evol.
15:149–155.
Amundsen T, Forsgren E. 2001. Male mate choice selects for female color-
ation in a fish. Proc Natl Acad Sci USA. 98:13155–13160.
Andrade MCB. 2003. Risky mate search and male self-sacrifice in redback
spiders. Behav Ecol. 14:531–538.
Baker RH, Ashwell RIS, Richards TA, Fowler K, Chapman T,
Pomiankowski A. 2001. Eects of multiple mating and male eye span
on female reproductive output in the stalk-eyed fly, Cyrtodiopsis dalmanni.
Behav Ecol. 12:732–739.
Baker RH, Denni M, Futerman P, Fowler K, Pomiankowski A, Chapman
T. 2003. Accessory gland size influences time to sexual maturity and
mating frequency in the stalk-eyed fly, Cyrtodiopsis dalmanni. Behav Ecol.
14:607–611.
Baker RH, Wilkinson GS. 2001. Phylogenetic analysis of sexual dimor-
phism and eye-span allometry in stalk-eyed flies (Diopsidae). Evolution.
55:1373–1385.
Baldauf SA, Bakker TCM, Kullmann H, Thünken T. 2011. Female nuptial
coloration and its adaptive significance in a mutual mate choice system.
Behav Ecol. 22:478–485.
Barry KL, Kokko H. 2010. Male mate choice: why sequential choice can
make its evolution dicult. Anim Behav. 80:163–169.
Bateman AJ. 1948. Intra-sexual selection in Drosophila. Heredity. 2:349–368.
Bonduriansky R. 2001. The evolution of male mate choice in insects: a syn-
thesis of ideas and evidence. Biol Rev. 76:305–339.
Bro-Jørgensen J. 2010. Dynamics of multiple signalling systems: animal
communication in a world in flux. Trends Ecol Evol. 25:292–300.
Burkhardt D, de la Motte I. 1985. Selective pressures, variability, and
sexual dimorphism in stalk-eyed flies (Diopsidae). Naturwissenschaften.
72:204–206.
Burkhardt D, de la Motte I. 1988. Big ‘antlers’ are favoured: female choice
in stalk-eyed flies (Diptera, Insecta), field collected harems and laboratory
experiments. J Comp Physiol A. 162:649–652.
Byrne PG, Rice WR. 2006. Evidence for adaptive male mate choice in the
fruit fly Drosophila melanogaster. Proc R Soc B. 273:917–922.
Candolin U. 2003. The use of multiple cues in mate choice. Biol Rev.
78:575–595.
Cotton S, Fowler K, Pomiankowski A. 2004a. Condition dependence of
sexual ornament size and variation in the stalk-eyed fly Cyrtodiopsis dal-
manni (Diptera: Diopsidae). Evolution. 58:1038–1046.
Cotton S, Fowler K, Pomiankowski A. 2004b. Heightened condition depen-
dence is not a general feature of male eyespan in stalk-eyed flies (Diptera:
Diopsidae). J Evol Biol. 17:1310–1316.
Cotton S, Rogers DW, Small J, Pomiankowski A, Fowler K. 2006.
Variation in preference for a male ornament is positively associated with
female eyespan in the stalk-eyed fly Diasemopsis meigenii. Proc R Soc B.
273:1287–1292.
Cotton S, Small J, Hashim R, Pomiankowski A. 2010. Eyespan reflects
reproductive quality in wild stalk-eyed flies. Evol Ecol. 24:83–95.
Darwin C. 1871. The descent of man and selection in relation to sex.
London: John Murray.
David P, Hingle A, Greig D, Rutherford A, Pomiankowski A, Fowler K.
1998. Male sexual ornament size but not asymmetry reflects condition in
stalk-eyed flies. Proc R Soc B. 265:2211–2216.
Dewsbury DA. 1982. Ejaculate cost and male choice. Am Nat. 119:601–610.
Doutrelant C, Grégoire A, Grnac N, Gomez D, Lambrechts MM, Perret P.
2008. Female coloration indicates female reproductive capacity in blue
tits. J Evol Biol. 21:226–233.
Edward DA, Chapman T. 2013. Variation in male mate choice in Drosophila
melanogaster. PLoS One. 8:e56299.
Forsberg J. 1987. A model for male mate discrimination in butterflies.
Oikos. 49:46–54.
Funk DH, Tallamy DW. 2000. Courtship role reversal and deceptive sig-
nals in the long-tailed dance fly, Rhamphomyia longicauda. Anim Behav.
59:411–421.
Gwynne DT. 1991. Sexual competition among females: what causes court-
ship-role reversal? Trends Ecol Evol. 6:118–121.
Harley E, Birge LM, Small J, Tazzyman SJ, Pomiankowski A, Fowler K.
2013. Ejaculate investment and attractiveness in the stalk-eyed fly,
Diasemopsis meigenii. Ecol Evol. 3:1529–1538.
Hedrick AV, Temeles EJ. 1989. The evolution of sexual dimorphism in ani-
mals: hypotheses and tests. Trends Ecol Evol. 4:136–138.
Hingle A, Fowler K, Pomiankowski A. 2001a. The eect of transient food
stress on female mate preference in the stalk-eyed fly Cyrtodiopsis dalmanni.
Proc R Soc B. 268:1239–1244.
Hingle A, Fowler K, Pomiankowski A. 2001b. Size-dependent mate prefer-
ence in the stalk-eyed fly Cyrtodiopsis dalmanni. Anim Behav. 61:589–595.
Honěk A. 1993. Intraspecific variation in body size and fecundity in insects:
a general relationship. Oikos. 66:483–492.
Lande R, Arnold SJ. 1985. Evolution of mating preference and sexual
dimorphism. J Theor Biol. 117:651–664.
LeBas NR, Hockham LR, Ritchie MG. 2003. Nonlinear and correla-
tional sexual selection on ‘honest’ female ornamentation. Proc R Soc B.
270:2159–2165.
Lessells CM, Boag PT. 1987. Unrepeatable repeatabilities: a common mis-
take. Auk. 104:116–121.
Lorch PD, Wilkinson GS, Reillo PR. 1993. Copulation duration and sperm
precedence in the stalk-eyed fly Cyrtodiopsis whitei (Diptera: Diopsidae).
Behav Ecol Sociobiol. 32:303–311.
Maynard Smith J, Harper D. 2003. Animal signals. Oxford: Oxford
University Press.
Mitra A, Gadagkar R. 2012. Queen signal should be honest to be involved
in maintenance of eusociality: chemical correlates of fertility in Ropalidia
marginata. Insect Soc. 59:251–255.
Nakahashi W. 2008. Quantitative genetic models of sexual selection by
male choice. Theor Popul Biol. 74:167–181.
Parker GA. 1983. Mate quality and mating decisions. In: Bateson P, editor.
Mate choice. New York: Cambridge University Press. p. 141–166.
Peeters C, Monnin T, Malosse C. 1999. Cuticular hydrocarbons cor-
related with reproductive status in a queenless ant. Proc R Soc B.
266:1323–1327.
Potti J, Canal D, Serrano D. 2013. Lifetime fitness and age-related female
ornament signalling: evidence for survival and fecundity selection in the
pied flycatcher. J Evol Biol. 26:1445–1457.
Preston BT, Stevenson IR, Pemberton JM, Wilson K. 2001. Dominant rams
lose out by sperm depletion. Nature. 409:681–682.
Reguera P, Pomiankowski A, Fowler K, Chapman T. 2004. Low cost of
reproduction in female stalk-eyed flies, Cyrtodiopsis dalmanni. J Insect
Physiol. 50:103–108.
Rogers DW, Baker RH, Chapman T, Denni M, Pomiankowski A, Fowler
K. 2005. Direct and correlated responses to artificial selection on male
mating frequency in the stalk-eyed fly Cyrtodiopsis dalmanni. J Evol Biol.
18:642–650.
Rogers DW, Chapman T, Fowler K, Pomiankowski A. 2005. Mating-
induced reduction in accessory reproductive organ size in the stalk-eyed
fly Cyrtodiopsis dalmanni. BMC Evol Biol. 5:37.
Rogers DW, Denni M, Chapman T, Fowler K, Pomiankowski A. 2008.
Male sexual ornament size is positively associated with reproductive
morphology and enhanced fertility in the stalk-eyed fly Teleopsis dalmanni.
BMC Evol Biol. 8:236.
Rogers DW, Grant CA, Chapman T, Pomiankowski A, Fowler K. 2006.
The influence of male and female eyespan on fertility in the stalk-eyed fly,
Cyrtodiopsis dalmanni. Anim Behav. 72:1363–1369.
Servedio MR, Lande R. 2006. Population genetic models of male and
mutual mate choice. Evolution. 60:674–685.
Small J, Cotton S, Fowler K, Pomiankowski A. 2009. Male eyespan and
resource ownership aect contest outcome in the stalk-eyed fly, Teleopsis
dalmanni. Anim Behav. 78:1213–1220.
South SH, Arnqvist G. 2011. Male, but not female, preference for an orna-
ment expressed in both sexes of the polygynous mosquito Sabethes cyaneus.
Anim Behav. 81:645–651.
Thomas ML. 2011. Detection of female mating status using chemical sig-
nals and cues. Biol Rev. 86:1–13.
Trivers RL. 1972. Parental investment and sexual selection. In: Campbell
B, editor. Sexual selection and the descent of man, 1871–1971. Chicago
(IL): Aldine Publishing Co. p. 136–179.
Page 9 of 10
by guest on November 17, 2014http://beheco.oxfordjournals.org/Downloaded from

Behavioral Ecology
Webberley KM, Hurst GD. 2002. The eect of aggregative overwintering
on an insect sexually transmitted parasite system. J Parasitol. 88:707–712.
Wedell N, Gage MJG, Parker GA. 2002. Sperm competition, male pru-
dence and sperm-limited females. Trends Ecol Evol. 17:313–320.
Wilkinson GS, Dodson GN. 1997. Function and evolution of antlers and
eye stalks in flies. In: Choe J, Crespi B, editors. The evolution of mating
systems in insects and arachnids. Cambridge (UK): Cambridge University
Press. p. 310–328.
Wilkinson GS, Kahler H, Baker RH. 1998. Evolution of female mating
preferences in stalk-eyed flies. Behav Ecol. 9:525–533.
Wilkinson GS, Reillo PR. 1994. Female choice response to artificial selection
on an exaggerated male trait in a stalk-eyed fly. Proc R Soc B. 255:1–6.
Page 10 of 10
by guest on November 17, 2014http://beheco.oxfordjournals.org/Downloaded from