Content uploaded by Martin Plath
Author content
All content in this area was uploaded by Martin Plath
Content may be subject to copyright.
ORIGINAL PAPER
Mustached males in a tropical poeciliid fish: emerging female
preference selects for a novel male trait
Ellen McCoy &Norbert Syska &Martin Plath &
Ingo Schlupp &Rüdiger Riesch
Received: 13 December 2010 / Revised: 26 January 2011 / Accepted: 28 January 2011
#Springer-Verlag 2011
Abstract One possible mechanism for the (co-)evolution
of seemingly novel male traits and female preferences for
them is that males exploit pre-existing female biases, and
livebearing fishes (Poeciliidae) have been at the forefront of
this research for almost two decades. Here, using 13
poeciliid species from four different genera, we tested
whether mustache-like rostral filaments found in males of
some Mexican molly (Poecilia sphenops) populations
could have evolved due to exploitation of a pre-existing
female bias. While Mexican mollies were the only species
with a significant female association preference for mus-
tached males, we also did not find any species exhibiting
significant aversion for mustached males; rather, variance in
female preference scores was large throughout. For exam-
ple, more than 25% of females spent twice as much time
with the mustached male compared to the non-mustached
male in most species, but even 31% of Mexican molly
females spent more time near the non-mustached male.
Hence, a comparison of the strength of preference was
inconclusive. We discuss the possibility that the female
preference of P. sphenops for mustached males could be
due to a female pre-existing bias (sensu lato), even if
population means were not significant for species other than
P. sphenops. This highlights the importance of distinguishing
between population means and individual preferences when
interpreting mate choice, and thus, adds depth to the concept
of mating preferences as a motor for evolutionary change.
Keywords Female choice .Mate preferences .Pre-existing
bias .Sexual selection .Character mapping
Introduction
Sexual selection, e.g., through female mate choice, can
drive male trait exaggeration, leading to sexual dimorphism
(Darwin 1871; Andersson 1994; Ryan and Rand 1993;
Rios-Cardenas and Morris 2011). There is ample evidence
for pre-existing sensory biases in females (sensu Endler
1992), and males might evolve traits to exploit those biases
(sensory exploitation; Ryan 1990; Ryan and Keddy-Hector
1992;Shaw1995; Endler and Basolo 1998). As a
prominent example, complex mating calls (‘whine’plus
‘chuck’) in túngara frogs (Physalaemus pustulosus) are
thought to have evolved as a consequence of pre-existing
female preferences (for auditory tuning, see Ryan et al.
1990; Wilczynski et al. 2001; Ryan 1985 for a review), and
females of other species (like Physalaemus coloradorum)
prefer conspecific calls with an artificially added hetero-
Communicated by T. Czeschlik
E. McCoy :N. Syska :I. Schlupp :R. Riesch
Department of Zoology, University of Oklahoma,
730 Van Vleet Oval,
Norman, OK 73019, USA
N. Syska
Unit of Evolutionary Biology & Systematic Zoology,
University of Potsdam,
Karl-Liebknecht Str. 24-25,
14476 Potsdam, Germany
M. Plath
Department of Ecology & Evolution, University of Frankfurt,
Siesmayerstr. 70a,
60054 Frankfurt, Germany
R. Riesch (*)
Department of Biology & W. M. Keck Center for Behavioral Biology,
North Carolina State University,
127 David Clark Labs,
Raleigh, NC 27695-7617, USA
e-mail: ruedigerriesch@web.de
E. McCoy
College of Veterinary Medicine, Texas A&M University,
College Station, TX 77843, USA
Behav Ecol Sociobiol
DOI 10.1007/s00265-011-1154-x
specific ‘chuck’, even though conspecific males are not
known to produce ‘chucks’as part of their natural
repertoire (Ryan and Rand 1993,1995,1999).
Livebearing fishes (family Poeciliidae) are prime exam-
ples for the evolution of male traits through pre-existing
biases (Basolo 1990a,b,1995a,b; Rosenthal and Evans
1998; Rodd et al. 2002; Rios-Cardenas and Morris 2011).
For example, swords (colorful extensions of the lower
caudal fin in the male sex) within the genus Xiphophorus
are thought to have evolved in response to a pre-existing
female preference for this trait in poeciliids (Basolo 1990a,
b,1995a,b;Basolo2002; but see Rosenthal and Evans
1998), even though swordtail-predators (like Astyanax
mexicanus, Characidae) also prefer males with long
swords as prey (Rosenthal et al. 2002). However, not all
poeciliid females express this preference (Basolo 2002;
Rosenthal et al. 2002; Witte and Klink 2005)andfemales
mayevenshowdisdainforswords(Basolo2002;Wong
and Rosenthal 2006).
In the present study, we use a comparative approach to
investigate the potential role of pre-existing bias for the
evolution of an unusual trait in male poeciliids: mustache-like
rostral filaments, as found in some populations of Mexican
mollies, P. sphenops (Schultz and Miller 1971; Schlupp et al.
2010). We could show beforehand that the mustache-like
structures on the rostrum of P. sphenops consist of epidermal
outgrowths on the edges of scales that appear to have no
sensory function (for functional analyses using EM micros-
copy see Schlupp et al. 2010). When given a choice between
mustached and mustache-less males, wild-caught P. sphe-
nops females prefer the mustached phenotype over males
that had their mustaches removed, suggesting that inter-
sexual selection by female choice indeed drives the evolution
of this unusual trait (Schlupp et al. 2010).
In the present study, we tested for a female preference
for males exhibiting (artificially added) mustache-like
rostral filaments in 13 species of poeciliid fishes from
various phylogenetically distant lineages using association
Fig. 1 Mate choice experiments using video-animated males with or
without mustache-like rostral filaments: aexample of video images
used for X. hellerii;bcladogram of the 13 poeciliid species included
in this study colors correspond to character history reconstruction
based on maximum parsimony analysis; cresults from the female
choice experiments depicting the strength of preference (SOP, see
main text), the percentage of females that spent at least twice as much
time near mustached males (SOP ≥0.25), and paired t-tests on
association time
Behav Ecol Sociobiol
preference tests and video-animated stimuli (Fig. 1). Fe-
male association preferences are good indicators of female
mate choice, e.g., in green swordtails (Xiphophorus hellerii,
Walling et al. 2010). We show that females tend to exhibit a
preference for “mustached”males in most species of
Poecilia, as indicated by positive strength of preference
(SOP) values. However, a significant preference for
“mustached”males was only discovered in P. sphenops—
the species in which males naturally express the trait.
Hence, it would seem that the female preference for the trait
emerged from a latent tendency to associate with mustached
males that at least within the genus Poecilia, was already
present prior to the expression of the actual trait.
Material and methods
Study organisms and their maintenance
We examined female preferences in 13 species of poeciliid
fishes (Fig. 1b). Mexican mollies (P. sphenops) stemmed
from El Limon, Veracruz, Mexico [35.8 ±1.9 mm standard
length (SL), N=16]. Two further representatives of “short-
fin mollies”were included: Atlantic mollies (Poecilia
mexicana) originated from Río Amatan in Tabasco in
southern México (35.9±0.9 mm, N=20) and from site VI/
17 (Río Purificacion) in Tamaulipas in central México
(35.9±1.6 mm, N= 15), and mangrove mollies (Poecilia
orri) came from Roatan Island off Honduras (33.4± 2.0 mm,
N=14). We included two representatives of “long-fin
mollies”: Sailfin mollies (Poecilia latipinna)werecollected
from Lincoln Park in Texas, USA (41.5±1.9 mm, N=12),
while Tamesi mollies (Poecilia latipunctata)werecollected
near Ciudad Mante in Tamaulipas (internal lab code III/9),
México (34.1±1.8 mm, N= 14). Gynogenetic Amazon
mollies (Poecilia formosa;populationfromLincolnPark,
37.6± 1.5 mm, N=14; and from VI/17, 38.5±2.5 mm, N=
10) are an all-female species (see below) and were either
tested with P. mexicana stimulus males (see above), or with
P. latipinna videos (Lincoln Park, 43.5±1.3 mm, N=15;
population from VI/17, 40.0 ±2.6 mm, N=9). Guppies
(Poecilia reticulata), stemmed from a feral population in
San Antonio in Texas, USA (17.6± 1.7 mm, N=15). We
included two species of Limia: laboratory-raised sulfur
limia (Limia sulphurophila) from the sulfidic Zufrada
Springs (35.4 ± 3.2 mm, N=9) and an aquarium stock of
Hispaniola limia (Limia perugiae) from the Dominican
Republic (28.6± 1.2 mm, N=14). Two species of Xipho-
phorus were obtained from the Xiphophorus genetic stock
center in San Marcos (Texas, USA); green swordtails
(Xiphophorus hellerii) stemmed from Río Sarabia near
Oaxaca in México (33.0± 2.2 mm, N=13); variable platyfish
(Xiphophorus variatus) came from Encino in Tamaulipas,
México (34.8±0.7 mm, N=13). Finally, we included western
mosquitofish (Gambusia affinis), which originated from a
feral population that thrives in a pond at the Aquatic
Research Facility of the University of Oklahoma, USA
(36.8±0.9 mm, N= 15), and widemouth mosquitofish
(Gambusia eurystoma), which are endemic to the highly
sulfidic Baños del Azufre near Teapa in Tabasco, Mexico
(26.8±1.2 mm, N=12).
The Amazon molly (P. f o r m o s a ) is an all-female,
ameiotic species that originated through natural hybridiza-
tion of a P. mexicana-like female and a P. latipinna-like
male around 100,000 generations ago (Avise et al. 1991;
Schartl et al. 1995). Amazon mollies are gynogenetic and
thus rely on sperm of “host species”to trigger embryogen-
esis, but normally no paternal DNA is transmitted; hence,
inheritance is clonal (Schlupp 2005). P. formosa naturally
forms mixed-species shoals with one or both of two host
species (Schlupp and Ryan 1996). In this unusual mating
system, Amazon mollies need to interact with heterospe-
cific males. Interestingly, female mating preferences have
been recorded even though indirect (genetic) benefits of
mate choice can be ruled out (Marler and Ryan 1997;
Tobler et al. 2006; Poschadel et al. 2009). Given its unique
ancestry as a natural hybrid of a long-fin and a short-fin molly,
we included the Amazon molly to test for the influence of
‘conflicting’genomes on the specificity of a female prefer-
ence. We predicted that if both ancestral species (P. latipinna
and P. m e x i ca n a ) differed in their preference for bearded
males, Amazon mollies should show an intermediate
preference score relative to its parental species.
Preference data for Mexican mollies (P. sphenops) were
reanalyzed from a previous study (Schlupp et al. 2010). All
species/populations included in this study were maintained
as randomly outbred stocks in tanks (250–1,000 l) in a
greenhouse of the University of Oklahoma in Norman. The
tanks contained naturally growing algae and other sub-
merged plants as well as a variety of naturally occurring
invertebrates like chironomid larvae, copepods, and amphi-
pods, on which the fish could feed. In addition, they were
fed ad libitum amounts of flake food every 2 days. All fish
used in this study were sexually mature and had interacted
with the opposite sex; thus, all females were most likely
pregnant.
Mate-choice experiments
We used video playback techniques to assess female
preference (or avoidance) of mustached and non-mustached
males. To generate the video stimuli, we employed estab-
lished protocols (see Schlupp et al. 2010). In short, digital
images of individual males were taken while a male was
swimming in a small tank using a Nikon D70 digital
camera. The image was then extracted from the background
Behav Ecol Sociobiol
in Adobe® Photoshop Elements 6.0, duplicated, and
mustache-like rostral filaments were pasted onto the snout
region of one copy of the male picture (Fig. 1a). Conse-
quently, the pairs of animation showed two identical males
differing only in the presence or absence of a mustache
(Fig. 1a; Schlupp et al. 2010).
The resulting images were then animated and converted
to an AVI-file (resolution 800× 600) using “Pencil v.0.4.4b”
for Mac. A straight movement of the pictures from left to
right and right to left was generated in front of a uniform
white background. The animations were 14 s long: twice
6 s for the distance of 28.5 cm on the screen forth and back,
each followed by an invisible turn of 1 s. Simultaneous
playback was performed using two identical computer
monitors (Belinea 10 30 40) with a Matrox Millenium
G400 Dual Head graphic card. The monitor refresh rate was
85 Hz, and the AVI-films were run in infinite loops during
the experiment using the “Monitor Plug-In”that is part of
the Viewer software package (BIOBSERVE GmbH).
All females were only exposed to video stimuli of their
own males, the lone exception being the all-female Amazon
mollies. As a result of their hybrid origin, we tested
Amazon mollies with stimuli of both parental species.
However, individual Amazon mollies were only presented
with one set of video stimuli: either a set of videos
depicting P. latipinna males or a set of videos depicting P.
mexicana males.
Measuring female preferences
The monitors were placed on either side of a test tank (60 ×
40×30 cm). Water level was maintained at 25 cm, which
was also the height of the monitors, and the water
temperature was 25°C. Using the “Zone Definition”
function of the Viewer software, the tank was virtually
divided into three equal-sized sections: A central neutral
zone and two preference zones near the monitors. The
behavior and movement of each focal female was moni-
tored and tracked using a Sony camera (Exwave HAD)
placed above the setup, connected to the PC running the
Viewer program. To increase the contrast between fish and
background (i.e., making it easier for the viewer software to
track the focal fish), white opaque Plexiglas sheets covered
the bottom and long sides of the tank.
A test female was placed into a transparent Plexiglas
cylinder (8.5 cm diameter) in the middle of the neutral
zone. The female was allowed to acclimate for 5 min, then
the cylinder was carefully removed, and the experiment was
remotely started. The Viewer software now tracked the
focal fish for an observation period of 5 min, and recorded
the time spent in each preference zone. To detect side
biases, the female was immediately placed into the cylinder
again, the video playbacks were switched, and the
procedure was repeated. After the test, the SL of the female
was measured to the closest millimeter. Females were then
transferred into another tank, so that each female was only
tested once.
For some species (e.g., both Gambusia sp.), focal
females were often smaller than 25 mm (SL), which is the
lowest cut-off for reliable tracking with the viewer system.
For these females, times spent in either preference zone
were recorded using stopwatches, while visually tracking
fish movement on-screen of the PC monitor while running
the Viewer program.
We decided a priori to exclude side biases if females
spent more than 80% of their time in the same compartment
of the tank (i.e., the females did not follow the initially
preferred male during the course of the trial), and trials with
low response (<50% of the time inside the preference
zones). We assumed such females not to be motivated to
choose (Landmann et al. 1999). Side biases occurred in 14
cases (out of 408 trials total). No trials were excluded due
to low response.
Statistical analyses
The statistical analyses were carried out on a species-wise
level. We tested that populations did not differ in strength
of preference in those cases in which more than one
population per species were included (independent samples
t-tests, P. mexicana t
32
=0.23, P=0.82; P. formosa tested
with P. mexicana t
22
=0.52, P=0.61; P. formosa tested with
P. latipinna t
22
=-0.64, P=0.53). In a first step, we tested for
female preferences within each species separately; to do so,
association times near both types of stimulus males were
compared using paired t-tests. In a second approach, we
compared the proportion of females that preferred the
mustached-male across species (0 =female spent more time
near non-mustached male; 1=female spent more time near
mustached male) using a logistic regression with species
and log-transformed female SL as covariates. The interac-
tion term of ‘SL by species’was not significant (B= 0.04,
SE=0.44, Wald=0.01, df =1, P=0.93) and was therefore
removed from the final model (−2 log likelihood =311.19).
We further compared the proportion of females that spent
approximately twice as much time or more with the
mustached male (i.e., SOP≥0.25) across species using a
similar model as before. Again, the interaction term of ‘SL
by species’was not significant (B=−0.32, SE=0.44, Wald=
0.55, df=1, P=0.46) and was therefore removed from the
final model (−2 log likelihood=260.67). Finally, we
compared the strength of preference [SOP, (time spent near
mustached male–time spent near non-mustached male)/
(time near mustached+time near non-mustached)] across
species using a general linear model, while including
female SL as a covariate. However, neither SL (F
1, 204
=
Behav Ecol Sociobiol
0.46, P=0.50) nor the interaction term of ‘SL by species’
(F
12, 204
=0.47, P=0.93) had a significant effect; thus, we
excluded SL from the final model.
To investigate how Amazon mollies would behave
relative to their parental species, we first screened the two
parental species (P. mexicana and P. latipinna) for differ-
ences in SOPs using independent t-test (t
45
=0.36, P=0.72),
and then conducted a one-sample t-test to compare Amazon
molly SOPs with the mean SOP of both parental species
(mean SOP= 0.0685).
We used a phylogenetic approach to map character
states, i.e., strength of preference (Basolo 1995b; Shaw
1995). A cladogram of the included taxa was created
(Fig. 1b); relative topologies were inferred from recent
molecular phylogenetic analyses of the entire family
Poeciliidae and of the genus Poecilia (Ptacek and Breden
1998; Breden et al. 1999; Hrbek et al. 2007). However, it is
important to note that the original phylogenies our
cladogram is based upon clearly warrant additional phylo-
genetic analysis.
For heuristic purposes, we also traced the potential
character history (i.e., the tendency to spend more time with
a mustached male) on our cladogram by conducting
maximum parsimony analysis as implemented in Mesquite
2.74 (1997–2010 W. & D. Maddison). For this analysis,
population means of SOP were assigned to one of four
different discrete categories [(0) ‘aversion’SOP <−0.01; (1)
‘neutral’−0.01<SOP < 0.01; (2) ‘weak preference’0.1 >
SOP>0.01; (3) ‘strong preference’SOP> 0.1].
All reported variables are given as means ± standard error
(SE). Unless noted otherwise, all statistical analyses were
conducted using SPSS 16.0.2 for Mac (2008 SPSS Inc.).
Results
When testing for a female preference for mustached males
within each species separately, we found females to spend
significantly more time near the mustached male in P. sphenops
(t
15
=2.38, P=0.031), but not in any other taxon examined (t≤
1.34, P≥0.19; Fig. 1c). Congruently, there was a significant
difference among species in the proportion of females that
spent more time with the mustached male (logistic regression:
B±SE=−0.09± 0.04, Wald=5.60, df=1, P=0.018). SL had no
significant effect in the logistic regression (B±SE=−2.08±
1.16, Wald=3.20, df=1, P=0.074). Species still differed
significantly from one another when P. sphenops was removed
from the analysis (species B±SE=−0.08± 0.04, Wald= 5.93,
df=1, P=0.015; SL B±SE=−2.01± 1.16, Wald=3.00, df=1,
P=0.083). On the other hand, when comparing the proportion
of females that spent at least twice as much time with the
mustached male (SOP≥0.25), the significant difference
between species disappeared (B±SE=−0.07± 0.04, Wald=
2.67, df=1, P=0.10; SL B±SE=−1.44± 1.23, Wald=1.37,
df=1, P= 0.24). Yet, when plotting this percentage on a
cladogram based on the phylogenetic relationships among the
observed taxa, it appears that proportionally more fish of the
genus Poecilia had a tendency to strongly prefer mustached
males (i.e., at least 25% of the females examined per species).
However, the consistency of this pattern is broken up by the
low values for P. f o r m o s a and high proportions of X. variatus
and G. eurystoma females (Fig. 1c). Interestingly, even though
the SOP for Amazon mollies was negative relative to positive
SOPs in P. mexicana and P. latipinna, Amazon mollies did not
differ significantly from the mean SOP value of the two
parental species (one-sided t-test, t
47
=−1.97; P=0.054).
No clear directional pattern of trait evolution was discern-
able when simply plotting SOP values on the cladogram; e.g.,
the second largest SOP value following P. sphenops was not
found in the closely related other “short-fin mollies”P.
mexicana and P. orri, but rather in Poecilia latipunctata (a
member of the “long-fin molly”clade; Fig. 1). There was
considerable variance within most examined species (see
standard errors in Fig. 1c), suggesting that females of those
species did not behave uniformly. For example, females in
G. eurystoma showed a net preference close to zero along
with pronounced variance (Fig. 1c); in other words, some
females preferred the mustached male while others preferred
the non-mustached one. Due to this pronounced variance
within most species, no significant difference among the
examined taxa in the expression of the female preference
(strength of preference) was detected (one-way ANOVA
F
12,217
=0.56, P=0.87).
Finally, when applying maximum parsimony an interesting
pattern emerged: for all genera other than Poecilia,itwas
equally likely that the ancestral state was ‘aversion’or a
‘weak’tendency to associate with the mustached males.
Furthermore, there was an almost equal number of decreases
(resulting in an aversion) and increases in the strength of the
preference. At the same time, maximum parsimony analysis
predicted that within the genus Poecilia, the ancestral state is
a‘weak’positive SOP (Fig. 1b), with two independent
increases resulting in ‘strong’positive SOPs (P. sphenops
and P. latipunctata) and one decrease resulting in a ‘neutral’
SOP in P. orri. Overall, this analysis clearly reveals the lack
of a linear drive towards increasing preference strength (that
is, there is no linear movement from character state 0 to 1 to
2 to 3), instead, the pattern jumps from 0 to 1, 0 to 2, 2 to 3
and/or 2 to 1 (Fig. 1b).
Discussion
In our current study, we tested for a potential role of pre-
existing female bias for mustached males within livebearing
fishes (family Poeciliidae). If pre-existing biases (sensu
Behav Ecol Sociobiol
stricto) played a role here, we would have expected a
female preference for mustached males also in some close
relatives of P. sphenops (e.g., within the “short-fin molly”-
clade; Fig. 1b, c), but no such pattern was discernible.
However, it appears that—with a few exceptions—mainly
females of the genus Poecilia tend to associate with a
mustached rather than a non-mustached male. This indi-
cates that the female preference for mustached mollies in P.
sphenops could be an emerging preference that built on an
already present tendency for mustached males even in
species other than P. sphenops. As we will argue below, if
one applied a slightly broader definition of pre-existing bias
(sensu lato) that takes into account variability of female
preferences within populations rather than interpreting
population means as species-specific stereotypic traits that
apply to all individuals, we actually find some support for it
in our data set.
This leads to the question as to how this novel trait and
the emerging female preference for it could have arisen.
The theory of sensory exploitation posits that female
preferences predate male traits, and preferences evolve
under selection other than sexual selection (Ryan 1990;
Shaw 1995; Endler and Basolo 1998; Rios-Cardenas and
Morris 2011). For example, the female preference for male
orange coloration in Trinidadian guppies (P. reticulata)is
thought to have evolved as a feeding preference [Rodd et
al. 2002; see also Madden and Tanner (2003) for male
bower decoration matching color preferences of females'
preferred food items in the regent bowerbird, Sericulus
chrysocephalus]. Likewise, it is tempting to speculate that
P. sphenops females may perceive the mustache as if the
male is ingesting food, thus indicating the presence of
resources, and making it more likely for females to
approach. In various invertebrates, nuptial feeding can
evolve when the male exploits the female's foraging
motivation in a sexual context [e.g., Sakaluk (2000) for
crickets; Bilde et al. (2007) and Andersen et al. (2008) for
spiders, Pisaura mirabilis].
Then again, in a logical extension to this hypothesis, one
would have expected other female poeciliids to also exhibit
the preference. Why then did only P. sphenops females
express a significant preference for male mustaches? We
can only speculate at this point—possible explanations
range from thus far unknown differences in feeding ecology
to differences in the sensory environment in which female
preferences and the male trait have co-evolved. For
example, if mustaches were really perceived as worm- or
insect larvae-like prey, then a preference for male mustaches
would be more likely to evolve in insectivorous species. It
could be that P. sphenops relies more strongly on invertebrate
prey as opposed to more algivorous and detritovorous
Poecilia species (e.g., Darnell 1962; Winemiller 1993;
Kramer and Bryant 1995; Tobler 2008), but future gut-
content analyses will need to be conducted to answer this
question. The fact that no preference (nor even a positive
mean SOP) was detected in the insectivorous Gambusia
species does not preclude this hypothesis, but also does not
lend any support to it.
In this context, it is important to note that all species
investigated in the present study did show a large degree of
variance in their SOP scores (including P. sphenops,in
which 5 out of 16 females actually had a negative SOP).
While for a variety of animals, female mate choice has been
shown to vary from trial to trial (e.g., Jennions and Petrie
1997; Bell et al. 2009); several studies have demonstrated
that it is usually highly repeatable in poeciliids (e.g., Godin
and Dugatkin 1995; Brooks 1996,2002; Jennions and
Petrie 1997; Brooks and Endler 2001; Aspbury and Basolo
2002) in particular when video stimuli are presented
(Kodric-Brown and Nicoletto 1997). Still, future experi-
ments will need to confirm that the variance detected in our
study is indeed based on repeatable, and thus, potentially
heritable (but see Jennions and Petrie 1997; Brooks and
Endler 2001; Brooks 2002) individual performance or
whether it stems from ad hoc variability, in which females
sometimes prefer a mustache and sometimes not. If the
former were true, then those females with a heritable
preference for mustached males would be easy targets for
Fisherian runaway selection in the Mexican molly. Yet
again, even if the latter were true and female preferences
would vary unpredictably over time, an emerging male trait
(e.g., a mustache) may persist in a population as long as
there is (a) no uniform and strong female discrimination
against mustached males, (b) no strong natural selection
against them via other selective agents (e.g., predators), or
(c) no genetic drift events immediately removing the novel
trait from a population in which it arises. In fact, in many
poeciliids, females prefer rare male phenotypes (e.g.,
Schlupp et al. 1999; Eakley and Houde 2004), and females
are known to copy the mate choice of other females (e.g.,
Witte and Ueding 2003; Godin and Hair 2009). Conse-
quently, one could even make the argument that a novel
trait like male mustaches is actually expected to (at least
initially) spread until it reaches a certain frequency within a
population. Future experiments will need to examine this
further, for example, by measuring individual repeatability
of preference scores or by conducting selection experiments
to test whether the preference for mustached males can be
artificially selected for in species in which the trait does not
naturally occur.
The confusing case of the Amazon molly
We originally predicted that unisexual Amazon mollies,
being of hybrid origin, would show an SOP intermediate to
that of the two parental species P. latipinna and P.
Behav Ecol Sociobiol
mexicana. However, our results do not support this,
because preferences of Amazon mollies were opposite in
direction (albeit not significantly different) compared to
both parental species. How can this be explained? One
could argue that since Amazon mollies are the product of a
single hybridization event (Stöck et al. 2010) and have
clonal inheritance (Schlupp and Riesch 2011), the two
individuals that actually hybridized could have been
representatives of the fraction of individuals that exhibit(ed)
a tendency to avoid mustached males. For this scenario
to work, one would have to assume that Amazon mollies
lacked the capability to evolve [as proposed by the often
postulated hypothesis that assumes that they (and similar
species) are F
1
s frozen in time; e.g., Vrijenhoek 1979;
Rogers and Vamosi 2010]. However, recent studies have
called this result into question. For example, Amazon
mollies exhibit extensive MHC loci polymorphism, albeit
reduced relative to the two parental species (Schaschl et
al. 2008; Lampert et al. 2009),andtheyalsodiffer
morphologically and behaviorally depending on which
sexual species they co-occur with (M. Tobler, I. Schlupp
unpublished data).
Another hypothetical argument could revolve around
the fact that Amazon mollies have existed for more than
100,000 generations (Schartl et al. 1995;seeLampertand
Schartl 2008 for an even higher estimate). So, couldn't it
be that both parental species had historically been
characterized by an aversion to the mustache-like trait? If
that were the case, and again assuming that Amazon
mollies lack the ability for evolutionary change, the
tendency to prefer the mustache would then have evolved
in the parental species after the hybridization but remained
the ancestral preference in asexual Amazon mollies.
Again, we think that this is extremely unlikely because
as we argued above, there is quite convincing evidence
that Amazon mollies are capable of evolutionary change.
Furthermore, the other lineages of Poecilia that we tested
in the present study are clearly older than this particular
hybridization event (e.g., Hrbek et al. 2007), so if this
scenario were true, the tendency of some females to prefer
the mustache must have independently evolved in all
Poecilia only after the hybridization event. Once more,
this scenario seems highly unlikely.
What could be a factor, however, is that Amazon
mollies have previously been demonstrated to avoid
certain parasitized males that the two parental species
do not discriminate against (Tobler et al. 2006), while at
the same time not exhibiting the same significant
preferences for novel male traits that P. latipinna and P.
mexicana show (Schlupp et al. 1999). Potentially,
Amazon mollies are simply unsure as to whether the
unknown male trait (i.e., mustache-like rostral filaments)
is an indication of parasitization and therefore rather
‘play it safe’by showing signs of avoidance behavior.
This scenario is clearly plausible, but our data are
unfortunately not suitable to specifically test for this.
We thus simply have to admit that at present, we do not
have a compelling explanation for the pattern our study
uncovered.
Conclusions
The present study highlights the importance of under-
standing a female preference not as an absolute trait in
the sense that it is either “present”or “absent”.Rather,a
preference can arise as a tendency, even a very weak
one, that is acted upon by selection in nonlinear ways to
become stronger in some species, remain the same in
others, or even become weaker (e.g., P. o r r i )insome
cases. Indeed the majority of reviews on preference–trait
coevolution stress the importance of considering “prefer-
ence functions”(e.g., Jennions and Petrie 1997;Brooks
and Endler 2001; Brooks 2002)or“strength of preference”
(e.g., Van de Weerd et al. 1998; Higgins and Waugaman
2004); yet, the fact that preferences are a continuum rather
than discrete categories, and that small fractions of
females within a population may actually exhibit a
preference even if the population mean is neutral or
negative, is often overlooked in studies of mate choice.
Our results, therefore, add further depth to the concept of
an individual's preference as a motor for evolutionary
change.
Furthermore, we do not think that the pre-existing bias
hypothesis requires that close relatives share the bias to the
same degree or intensity as the species in which the male
trait has originated (Endler and Basolo 1998). Rather, it is
important that the origin of the female bias in whatever
strength must have predated the origin of the male trait.
Hence, even though the female preference for a mustache in
Mexican mollies is clearly not based on a pre-existing bias
in a way that close relatives share a significant preference, it
does seem to emerge from some latent tendency for
mustached males already present in a certain proportion of
females in some close relatives (and other poeciliids).
Whether this tendency to associate with mustached males is
linked to food acquisition remains to be tested. For
example, experiments investigating whether P. sphenops
females are indeed more likely to approach a male if it
carries food items in his mouth (especially if the female is
starved) are underway.
Finally, it is important to stress that the details of
preference evolution may change as more species are added
to the analysis and future studies provide us with a more
robust phylogenetic tree, but the general conclusions (as
stated above) are likely to be retained.
Behav Ecol Sociobiol
Acknowledgments We thank R. B. Langerhans for his introduction to
Mesquite and helpful discussions, as well as D. McLennan, and an
anonymous reviewer for comments that greatly helped to improve this
manuscript. The Mexican Government and the Municipal of Tacotalpa
kindly issued permits to collect fish. Financial support came from NSF
(DEB-0743406; to I. S.), and DFG (Deutsche Forschungsgemeinschaft;
PL 470/1–2; to M.P.).
References
Andersson M (1994) Sexual selection. Princeton University Press
Andersen T, Bollerup K, Toft S, Bilde T (2008) Why do males of the
spider Pisaura mirabilis wrap their nuptial gifts in silk: female
preference or male control? Ethology 114:775–781
Aspbury AS, Basolo AL (2002) Repeatable female preferences,
mating order and mating success in the poeciliid fish, Heteran-
dria formosa. Behav Ecol Sociobiol 51:238–244
Avise JC, Trexler JC, Travis J, Nelson WS (1991) Poecilia mexicana
is the recent female parent of the unisexual fish P. formosa.
Evolution 45:1530–1533
Basolo AL (1990a) Female preference for a male sword length in the
green swordtail, Xiphophorus helleri (Pisces: Poeciliidae). Anim
Behav 40:332–338
Basolo AL (1990b) Female preference predates the evolution of the
sword in swordtail fish. Science 250:808–810
Basolo AL (1995a) A further examination of a pre-existing bias
favouring a sword in the genus Xiphophorus. Anim Behav
50:365–375
Basolo AL (1995b) Phylogenetic evidence for the role of a preexisting
bias in sexual selection. Proc R Soc B 259:307–311
Basolo AL (2002) Congruence between the sexes in preexisting
receiver responses. Behav Ecol 13:832–837
Bell AM, Hankison SJ, Laskowski KL (2009) The repeatability of
behaviour: a meta-analysis. Anim Behav 77:771–783
Bilde T, Tuni C, Elsayed R, Pekar S, Toft S (2007) Nuptial gifts of
male spiders: sensory exploitation of the female's maternal care
instinct or foraging motivation? Anim Behav 73:267–273
Breden F, Ptacek MB, Rashed M, Taphorn D, Figueiredo CA (1999)
Molecular phylogeny of the live-bearing fish genus Poecilia
(Cyprinodontiformes: Poeciliidae). Mol Phylogenet Evol 12:95–104
Brooks R (1996) Copying and the repeatability of mate choice. Behav
Ecol Sociobiol 39:323–329
Brooks R (2002) Variation in female mate choice within guppy
populations: population divergence, multiple ornaments and the
maintenance of polymorphism. Genetica 116:343–358
Brooks R, Endler JA (2001) Female guppies agree to differ:
phenotypic and genetic variation in mate-choice behavior and
the consequences for sexual selection. Evolution 55:1644–1655
Darnell RM (1962) Fishes of the Rio Tamesi and related coastal
lagoons in east-central Mexico. Publ Inst Marine Sci, Univ Texas
8:299–365
Darwin C (1871) The descent of man and selection in relation to sex.
John Murray, London
Eakley AL, Houde AE (2004) Possible role of female discrimination
against ‘redundant’males in the evolution of colour pattern
polymorphism in guppies. Proc R Soc B 271:S299–S301
Endler JA (1992) Signals, signal conditions, and the direction of
evolution. Am Nat 139:125–153
Endler JA, Basolo AL (1998) Sensory ecology, receiver biases and
sexual selection. Trends Ecol Evol 13:415–420
Godin J-GJ, Dugatkin LA (1995) Variability and repeatability of
female mating preference in the guppy. Anim Behav 49:1427–
1433
Godin J-GJ, Hair KPE (2009) Mate-choice copying in free-ranging
Trinidadian guppies (Poecilia reticulata). Behaviour 146:1443–
1461
Higgins LA, Waugaman RD (2004) Sexual selection and variation: a
multivariate approach to species-specific calls and preferences.
Anim Behav 68:1139–1153
Hrbek T, Seckinger J, Meyer A (2007) A phylogenetic and
biogeographic perspective on the evolution of poeciliid fishes.
Mol Phylogenet Evol 43:986–998
Jennions MD, Petrie M (1997) Variation in mate choice and mating
preferences: a review of causes and consequences. Biol Rev
72:283–327
Kodric-Brown A, Nicoletto PF (1997) Repeatability of female choice
in the guppy: response to live and videotaped males. Anim
Behav 54:369–376
Kramer DL, Bryant MJ (1995) Intestine length in the fishes of a
tropical stream: 2. Relationships to diet—the long and short of a
convoluted issue. Environ Biol Fish 42:129–141
Lampert KP, Schartl M (2008) The origin and evolution of a unisexual
hybrid: Poecilia formosa. Philos Trans R Soc B 363:2901–
2909
Lampert KP, Fischer P, Schartl M (2009) MHC variability in the
clonal Amazon molly, Poecilia formosa: is copy number less
important than genotype? Mol Ecol 18:1124–1136
Landmann K, Parzefall J, Schlupp I (1999) A sexual preference in the
Amazon molly, Poecilia formosa. Environ Biol Fish 56:325–
331
Madden JR, Tanner K (2003) Preferences for coloured bower
decorations can be explained in a nonsexual context. Anim
Behav 65:1077–1083
Marler CA, Ryan MJ (1997) Origin and maintenance of a female
mating preference. Evolution 51:1244–1248
Poschadel J, Plath M, Schlupp I (2009) Divergent female preference in
a clonal fish. Acta Ethol 12:55–60
Ptacek MB, Breden F (1998) Phylogenetic relationships among the
mollies (Poeciliidae: Poecilia:Mollienesia group) based on
mitochondrial DNA sequences. J Fish Biol 53:64–81
Rios-Cardenas O, Morris MR (2011) Precopulatory sexual selection in
poeciliid fishes. In: Evans J, Pilastro A, Schlupp I (eds) Ecology
and Evolution of Poeciliid Fishes. University of Chicago Press,
Chicago, IL
Rodd FH, Hughes KA, Grether GF, Baril CT (2002) A possible non-
sexual origin of mate preference: are male guppies mimicking
fruit? Proc R Soc B 269:475–481
Rogers SM, Vamosi SM (2010) Frozen F1's amidst a masterpiece of
nature: new insights into the rare hybrid origin of gynogenesis
in the Amazon molly (Poecilia formosa). Mol Ecol 19:5086–
5089
Rosenthal GG, Evans CS (1998) Female preferences for swords in
Xiphophorus helleri reflects a bias for large apparent size. Proc
Natl Acad Sci USA 95:4431–4436
Rosenthal GG, Wagner WE Jr, Ryan MJ (2002) Secondary reduction
of preference for the sword ornament in the pygmy swordtail
Xiphophorus nigrensis (Pisces: Poeciliidae). Anim Behav 63:37–
45
Ryan MJ (1985) The túngara frog: a study in sexual selection and
communication. University of Chicago Press, Chicago, IL
Ryan MJ (1990) Sexual selection, sensory systems and sensory
exploitation. Oxf Surv Evol Biol 7:157–195
Ryan MJ, Keddy-Hector A (1992) Directional patterns of female mate
choice and the role of sensory biases. Am Nat 139:S4–S35
Ryan MJ, Rand AS (1993) Sexual selection and signal evolution—the
ghost of biases past. Phil Trans R Soc B 340:187–195
Ryan MJ, Rand AS (1995) Female responses to ancestral advertise-
ment calls in the túngara frog. Science 269:390–392
Behav Ecol Sociobiol
Ryan MJ, Rand AS (1999) Phylogenetic influence on mating call
preference in female túngara frogs, Physalaemus pustulosus.
Anim Behav 57:945–956
Ryan MJ, Fox JH, Wilczynski W, Rand AS (1990) Sexual selection
for sensory exploitation in the frog Physalaemus pustulosus.
Nature 343:66–67
Sakaluk SK (2000) Sensory exploitation as an evolutionary origin to
nuptial food gifts in insects. Proc R Soc B 267:339–343
Schartl M, Wilde B, Schlupp I, Parzefall J (1995) Evolutionary origin
of a parthenoform, the Amazon molly Poecilia formosa, on the
basis of a molecular genealogy. Evolution 49:827–835
Schaschl H, Tobler M, Plath M, Penn DJ, Schlupp I (2008)
Polymorphic MHC loci in an asexual fish, the amazon molly
(Poecilidae formosa; Poeciliidae). Mol Ecol 17:5220–5230
Schlupp I (2005) The evolutionary ecology of gynogenesis. Annu Rev
Ecol Evol Syst 36:399–417
Schlupp I, Riesch R (2011) Evolution of unisexual reproduction. In:
Evans J, Pilastro A, Schlupp I (eds) Ecology and evolution of
Poeciliid Fishes. University of Chicago Press, Chicago, IL
Schlupp I, Ryan MJ (1996) Mixed-species shoals and the maintenance
of a sexual-asexual mating system in mollies. Anim Behav
52:885–890
Schlupp I, Riesch R, Tobler M, Plath M, Parzefall J, Schartl M (2010)
A novel, sexually selected trait in poeciliid fishes: female
preference for mustache-like, rostral filaments in male Poecilia
sphenops. Behav Ecol Sociobiol 64:1849–1855
Schlupp I, Waschulewski M, Ryan MJ (1999) Female preferences for
naturally occurring novel male traits. Behaviour 136:519–527
Schultz RJ, Miller RR (1971) Species of the Poecilia sphenops
complex (Pisces: Poeciliidae) in Mexico. Copeia 1971:282–
290
Shaw K (1995) Phylogenetic tests of the sensory exploitation model of
sexual selection. Trends Ecol Evol 10:117–120
Stöck M, Lampert KP, Möller D, Schlupp I, Schartl M (2010)
Monophyletic origin of multiple clonal lineages in an asexual
fish (Poecilia formosa). Mol Ecol 19:5204–5215
Tobler M (2008) Divergence in trophic ecology characterizes
colonization of extreme habitats. Biol J Linn Soc 95:517–528
Tobler M, Plath M, Burmeister H, Schlupp I (2006) Black spots and
female mating preferences in a sexual/asexual mating complex
(Poecilia, Poeciliidae, Teleostei). Behav Ecol Sociobiol 60:159–165
Van de Weerd HA, Van Loo PLP, Van Zutphen LFM, Koolhaas JM,
Baumans V (1998) Strength of preference for nesting material as
environmental enrichment for laboratory mice. Appl Anim Behav
Sci 55:369–382
Vrijenhoek RC (1979) Factors affecting clonal diversity and coexis-
tence. Am Zool 19:787–797
Walling CA, Royle NJ, Lindström J, Metcalfe NB (2010) Do female
association preferences predict the likelihood of reproduction?
Behav Ecol Sociobiol 64:541–548
Wilczynski W, Rand AS, Ryan MJ (2001) Evolution of calls and
auditory tuning in the Physalaemus pustulosus species group.
Brain Behav Evol 58:137–151
Winemiller KO (1993) Seasonality of reproduction by livebearing
fishes in tropical rain-forest streams. Oecologia 95:266–276
Witte K, Klink KB (2005) No preexisting bias in sailfin molly
females, Poecilia latipinna, for a sword in males. Behaviour
142:283–303
Witte K, Ueding K (2003) Sailfin molly females (Poecilia latipinna)
copy the rejection of a male. Behav Ecol 14:389–395
Wong BBM, Rosenthal GG (2006) Female disdain for swords in a
swordtail fish. Am Nat 167:136–140
Behav Ecol Sociobiol