Adult female hamsters require long and sustained exposures to heterospecific males to avoid interspecific mating

Abstract

Interspecific mating normally decreases female fitness. In many species, females avoid heterospecific males innately or by imprinting on their parents. Alternatively, adult females could learn to discriminate against heterospecific males after exposure to such males. For example, Syrian hamster (Mesocricetus auratus) females learn to discriminate between conspecific males and Turkish hamster (M. brandti) males during adulthood by exposure to males of both species. Adult females not previously exposed to Turkish hamster males will mate similarly with conspecific and heterospecific males. However, in a previous study we showed that exposure to a heterospecific male and a conspecific male for 8 days led to mating avoidance and aggression towards the heterospecific male. Here we conducted two experiments to investigate how much exposure to the heterospecific male was required for females to avoid mating with the heterospecific male (Experiment 1) and how long that avoidance lasted in the absence of continuous exposure to heterospecific stimuli (Experiment 2). Fast and durable learning would indicate the evolution of an efficient avoidance response. In Experiment 1, females were exposed to a heterospecific male for 1, 4 h, 4 or 8 days and then paired with that male. We found more avoidance of interspecific mating after 4 or 8 days of exposure than after 1 or 4 h of exposure. In Experiment 2, females were exposed to a heterospecific male for 8 days and then paired with that male either 10 min later or 8 days later. We found that after an 8-day delay females were highly sexually receptive to the heterospecific male. Additionally, a comparison between the current experiments and a previous study indicates that female Syrian hamsters do not require concurrent exposure to a conspecific male and a heterospecific male to learn to avoid interspecific mating; exposure to a heterospecific male is sufficient.

Full text

RESEARCH ARTICLE
Adult female hamsters require long and sustained
exposures to heterospecific males to avoid interspecific
mating
Javier delBarco-Trillo •Robert E. Johnston
Received: 21 April 2010 / Accepted: 9 September 2010 / Published online: 19 September 2010
!Springer Science+Business Media B.V. 2010
Abstract Interspecific mating normally decreases female fitness. In many species,
females avoid heterospecific males innately or by imprinting on their parents. Alterna-
tively, adult females could learn to discriminate against heterospecific males after exposure
to such males. For example, Syrian hamster (Mesocricetus auratus) females learn to dis-
criminate between conspecific males and Turkish hamster (M. brandti) males during
adulthood by exposure to males of both species. Adult females not previously exposed to
Turkish hamster males will mate similarly with conspecific and heterospecific males.
However, in a previous study we showed that exposure to a heterospecific male and a
conspecific male for 8 days led to mating avoidance and aggression towards the hetero-
specific male. Here we conducted two experiments to investigate how much exposure to
the heterospecific male was required for females to avoid mating with the heterospecific
male (Experiment 1) and how long that avoidance lasted in the absence of continuous
exposure to heterospecific stimuli (Experiment 2). Fast and durable learning would indicate
the evolution of an efficient avoidance response. In Experiment 1, females were exposed to
a heterospecific male for 1, 4 h, 4 or 8 days and then paired with that male. We found more
avoidance of interspecific mating after 4 or 8 days of exposure than after 1 or 4 h of
exposure. In Experiment 2, females were exposed to a heterospecific male for 8 days and
then paired with that male either 10 min later or 8 days later. We found that after an 8-day
delay females were highly sexually receptive to the heterospecific male. Additionally, a
comparison between the current experiments and a previous study indicates that female
Syrian hamsters do not require concurrent exposure to a conspecific male and a hetero-
specific male to learn to avoid interspecific mating; exposure to a heterospecific male is
sufficient.
J. delBarco-Trillo !R. E. Johnston
Department of Psychology, Cornell University, Uris Hall, Ithaca, NY 14853, USA
J. delBarco-Trillo (&)
Department of Evolutionary Anthropology, Duke University, 128 Biological Sciences Building,
Box 90383, Durham, NC 27708, USA
e-mail: delbarcotrillo@gmail.com
123
Evol Ecol (2011) 25:391–401
DOI 10.1007/s10682-010-9429-7

Keywords Interspecific mating !Species discrimination !Learning !Memory !Lordosis !
Aggression
Introduction
When females mate with a heterospecific male (i.e., a male from a different species), they
normally produce no offspring or unfit offspring (Gro
¨ning and Hochkirch 2008); thus there
is strong selective pressure on females to avoid interspecific mating. To avoid interspecific
mating, females may reduce their sexual receptivity and/or increase their aggression
towards males of closely-related species (Andersson 1994). This discriminative behavior
against heterospecifics can be innately determined (Hebets 2007) or it may require some
type of learning during early development, e.g. imprinting (Irwin and Price 1999; Servedio
et al. 2009). Learning to discriminate between conspecific and heterospecific individuals
also occurs during adulthood in male fruit flies and male guppies (Dukas 2004; Magurran
and Ramnarine 2004; Dukas 2008), but not in female guppies (Haskins and Haskins 1949).
In fact, learning to avoid interspecific mating by females during adulthood has only been
shown in Syrian hamsters, Mesocricetus auratus (delBarco-Trillo et al. 2010).
Estrous female Syrian hamsters show a preference for the scent of conspecific males
over the scent of Turkish hamster males, M. brandti (delBarco-Trillo et al. 2009a,b). In
addition, activity in the posterior medial amygdala, a brain region that has been implicated
in species discrimination (Meredith and Westberry 2004), is higher in estrous female
Syrian hamsters exposed to the scent of conspecific males than in females exposed to the
scent of heterospecific males (delBarco-Trillo et al. 2009a). These findings indicate that
females can discriminate between conspecific and heterospecific males. However, Syrian
hamster females do not show an innate mating avoidance of heterospecific males (delB-
arco-Trillo et al. 2009b; delBarco-Trillo and Johnston 2010). Such interspecific mating
would be costly to Syrian hamster females because the failure to produce hybrids seems to
occur after fertilization due to differences in the number of chromosomes between both
species. Consequently, females mating with a heterospecific male may lose not only a
reproductive event but the next estrous cycle can also be significantly delayed (Todd et al.
1972; Murphy 1977).
Syrian hamster females that have never encountered heterospecific males are receptive
to and do not show any sign of aggression towards such males (delBarco-Trillo et al.
2009b; delBarco-Trillo and Johnston 2010). However, when Syrian hamster females were
exposed for 8 days to a conspecific male and a heterospecific male across wire-mesh
partitions and then paired sequentially with those same two males, the females were
aggressive and not receptive when paired with the heterospecific male whereas females
were always receptive when paired with the conspecific male (delBarco-Trillo et al. 2010).
We do not know whether this learned response necessarily requires a relatively long time
(8 days) or whether a much shorter time (e.g., a few hours of exposure) can be equally
effective.
The amount of time required for females to learn to discriminate against heterospecific
males may help us understand the evolution of such a learned response. We argue that
learning to avoid mating with heterospecific males through direct interactions or exposure
to stimuli from heterospecific males, instead of innately avoiding heterospecific males, can
be an adaptive strategy in groups of emergent or recently formed species which distri-
butions repeatedly fluctuate over time so that there are recurring episodes of overlapping
392 Evol Ecol (2011) 25:391–401
123

and no-overlapping distributions among pairs of those species. Under this evolutionary
scenario, each different species will be evolving separately and interactions between pairs
of species will be possible only during periods of overlapping distributions. Any innate
response toward heterospecific males established during a period of overlapping distri-
butions may not be adaptive during the following episodes of overlapping distributions if
the stimuli used by females to discriminate against heterospecific males have significantly
evolved during the time that both species were separated (Magurran and Ramnarine 2005).
In addition, an innate heterospecific avoidance may cause the rejection of some conspecific
males. Thus, during long periods of allopatry, innate preferences might be selected against
because they would reduce mating opportunities or increase mating costs. An alternative
and possibly more stable strategy is an adaptive response in which females learn to dis-
criminate against males that are significantly different than conspecific males (Bergman
and Feldman 1995; Dukas 2009). Such a strategy would remain adaptive independently of
how much males of the other species change during non-overlapping periods. How fast
females learn to discriminate against heterospecific males may depend on the rate of
interactions between any two species. If interactions are relatively frequent, females are
predicted to require shorter exposures to heterospecific males in order to discriminate
against them and to maintain such a learned response once in place. Alternatively, if
interactions between the two species are relatively infrequent, females are predicted to
require longer exposures to heterospecific males to discriminate against them and to be
more likely to forget the learned response in the absence of continuous heterospecific
stimuli. We base our predictions on the assumption that infrequent episodes of overlapping
distributions will provide (a) fewer opportunities for the adaptive establishment of a rapid
and persistent response, and (b) more opportunities for the evolved response to become less
efficient during the long episodes of non-overlapping distributions, i.e., when there is not
an adaptive value for that response.
The Mesocricetus genus is a good model to test the above predictions because it consists
of species which current distributions and phylogenetic relationships indicate that such
distributions have fluctuated over time (Neumann et al. 2006). For example, even though
M. auratus is more closely related to M. raddei than to M. brandti, the distribution of
M. brandti is between those of M. auratus (to the South) and M. raddei (to the Northeast)
(Neumann et al. 2006). Also, there are two separate populations of M. brandti separated by
the distribution of M. raddei. The two species considered in this study, M. auratus and
M. brandti, at present seem to have non-overlapping geographical distributions (Neumann
et al. 2006) even though M. auratus distribution has not yet been fully clarified (Gatter-
mann et al. 2001); however, it is unclear what even quite recent distributions were, given
that M. auratus distribution has severely declined very recently (in evolutionary terms) due
to loss of habitat (caused by agriculture) and deliberate destruction by humans (Gattermann
et al. 2001). In addition, M. brandti has been reported to be distributed not only to the
North but also to the Southwest of the M. auratus known distribution (Hamar and
Schutowa 1966), offering more opportunities for episodes of overlapping distributions
since their inception as separate species, approximately 2.7 million years ago (Neumann
et al. 2006).
We conducted two experiments to determine how much exposure to a heterospecific
male is necessary for Syrian hamster females to avoid interspecific mating (Experiment 1)
and to determine whether females that were initially exposed to a heterospecific male still
discriminate against that male after a long period of no exposure to heterospecific stimuli
(Experiment 2). In Experiment 1, females were exposed to a heterospecific male across a
wire-mesh partition for 1, 4 h, 4 or 8 days and then paired with the same heterospecific
Evol Ecol (2011) 25:391–401 393
123

male. We measured sexual receptivity and aggression towards the heterospecific male. We
predicted that longer exposures would lead to higher degrees of avoidance against the
heterospecific male. In Experiment 2, female subjects were first exposed to a heterospecific
male across a wire-mesh partition for 8 days. Then, these females were paired with the
familiar heterospecific male after either 10 min of separation or 8 days of separation. We
predicted that females would avoid mating with the heterospecific male even after a period
without exposure to the heterospecific male.
We also compared the results of the present study (in which females were exposed for
8 days only to a heterospecific male) to those of our previous study, in which females were
concurrently exposed for 8 days to a conspecific male and a heterospecific male (delBarco-
Trillo et al. 2010). Liley (1966) proposed that in order to learn to avoid interspecific
mating, an individual must be simultaneously exposed to both conspecifics and hetero-
specifics (i.e., as in a sympatric context). Alternatively, exposure to a heterospecific male
could be sufficient for females to learn to avoid interspecific mating.
Methods
Animals
All animals were born and raised in captivity at Cornell University, Ithaca, NY. Hamsters
were weaned at 30 days of age and housed individually in solid bottom polycarbonate
cages (45 924 914.5 cm) with sani-chip bedding material and constant access to water
and food (Prolab 1000, Agway, Syracuse, NY). Turkish and Syrian hamsters were main-
tained in separate rooms with independent air intake and exhaust. Turkish hamsters were
maintained on a 16L:8D light–dark schedule with lights off between 10:00 and 18:00 h
(Eastern Standard Time). Syrian hamsters were maintained on a 14L:10D light–dark
schedule with lights off between 09:00 and 19:00 h. Different light cycles were simply due
to husbandry reasons (Turkish hamsters seem to require longer light cycles to reproduce
successfully in captivity). Experiments were run between 10:00 and 13:00 h. We used dim
and indirect light to allow videotaping and observations. The video camera was located
approximately one meter from the cage. To avoid family effects, no siblings (of either sex)
were used in the same group.
To determine that a female was in estrus on a specific day, a conspecific male hamster
was placed inside the female’s home cage. If lordosis occurred within 15 s, the female was
considered in estrus (no mounting occurred during estrous testing). Lordosis is a stereo-
typical position in Mesocricetus spp. in which the female arches her back (down in the
center), spreads her legs, raises her tail, and remains immobile from seconds to minutes. If
no lordosis was observed, the female was retested on the following days (up to 3 days)
until lordosis occurred. Given the regular, 4-day estrous cycle in Syrian hamsters (Lisk
1985), once the day of estrus was determined, we could always determine the subsequent
estrous days. In this study, females were always tested 4 days after determining the timing
of their estrous cycles (i.e., during their next estrous day).
Experiment 1. Different exposure times
We used arenas (50 938.1 921.5 cm; Allentown, model# PCT4SHT) divided into two,
similarly-sized sections (24.85 938.1 921.5 cm) by a wire-mesh partition with 1-cm
2
openings. This type of barrier is sufficient for hamsters to learn the multi-odor individual
394 Evol Ecol (2011) 25:391–401
123

identity of other conspecifics (Johnston and Peng 2008). We started each test by placing
one Turkish hamster male in one of the two sections and then placed a Syrian hamster
female in the other section. The female could perceive and interact in a limited manner
with the male through the wire-mesh partition. We used soiled bedding from each animal’s
home cage as substrate for its respective section of the arena. Each animal had constant
access to its own water and food. Animals lived undisturbed in this arena for either 1 h
(N=9), 4 h (N=11), 4 days (N=9), or 8 days (N=10). The mismatched sample sizes
across treatments were unintentional and due to a counting error. Arenas were placed in a
room different than the colony and testing rooms; the light cycle in this room was the same
as in the rooms in which the females were housed (i.e., 14L:10D light–dark). After the
exposure treatment (1, 4 h, 4, or 8 days), the female and the male were transferred to
individual cages lined with clean bedding and both animals were moved to the testing
room. The pairing trial started 10 min after the animals had been transferred to their
individual cages. The male was transferred to the female’s cage in a clean, 500-ml plastic
beaker. The trial lasted 5 min. We did not interfere with the behavior of animals during
these trials. We videotaped all trials. We scored the recorded videos to determine the
latency to lordosis, the duration of lordosis, the latency to aggression, and the duration of
aggression. We considered any attempted biting or tumbling fights as signs of aggression.
If a female did not show lordosis or aggression during a test, we scored the latency for that
behavior as 300 s. If the female was aggressive and not receptive to the heterospecific
male, the female was immediately paired with an unfamiliar conspecific male for 5 min to
determine that the female was indeed in estrus. Such estrus-checking trials were not used
for any statistical analyses.
Experiment 2. Delaying the time between exposure and testing
In this experiment, after a female and a heterospecific male had lived in the adjacent
sections of the arena for 8 days, the female and the male were separated and transferred to
individual cages lined with clean bedding. Females were tested with the familiar hetero-
specific male after 10 min of separation (‘10-min delay group’; N=10) or 8 days of
separation (‘8-day delay group’; N=10). The 10-min delay group was the 8-day group in
Experiment 1. We interspersed testing of the 10-min delay and 8-day delay groups to avoid
any order effects. In the ‘8-day delay group’ (N=10), after the male and the female were
transferred to individual cages, they were then returned to their original colony rooms.
Eight days later, the two animals were moved to the testing room. After 10 min in the
testing room, the male was transferred to the female’s cage in a clean, 500-ml plastic
beaker. That is, there was a delay of 8 days from the time that the female and the male
were separated to the time that they were tested together. Trials were videotaped and
scored as indicated in Experiment 1.
In Experiment 1 and Experiment 2 we did not use a control in which a female Syrian
hamster was housed with a conspecific male and then paired with that same conspecific
male. We considered that the additional number of animals required for such controls was
not warranted because in a previous study (delBarco-Trillo et al. 2010) we housed female
Syrian hamsters next to two conspecific males for 8 days and then sequentially paired the
female with both conspecific males; in all pairings the female was as sexually receptive
and not aggressive as when a receptive female is paired with an unfamiliar conspecific
male. We did not have any reason to expect a different response by females when housed
with only one conspecific male.
Evol Ecol (2011) 25:391–401 395
123

Statistics
In Experiment 1 we used an ANOVA to determine if any of the variables (lordosis latency,
lordosis duration, aggression latency, and aggression duration) differed among the four
groups (1-hour, 4-hours, 4-days, and 8-days of exposure). If so, we used Tukey pairwise
comparisons for the post-hoc tests.
In Experiment 2 we used independent t-tests to determine if any of the variables
(lordosis latency, lordosis duration, aggression latency, and aggression duration) differed
between the two groups (10-min delay and 8-day delay). We used the Levene’s test for
equality of variances; when equal variances were not assumed, we report the estimated
degrees of freedom.
We tested the assumption of normality using the Kolmogorov–Smirnov test. When
necessary, we log transformed data to meet normality. In both experiments values are
shown as means ±SEM. We used SPSS 14 for Windows for all statistical analyses.
Results
Experiment 1. Different exposure times
Lordosis latency was greater with longer durations of exposure of females to heterospecific
males (1 h: 67.89 ±30.71 s; 4 h: 68.64 ±24.9 s; 4 days: 186.33 ±39.89 s; 8 days:
225.3 ±38.09 s; F
3,38
=6.98, P=0.001). Significant differences were found in 1-hour
vs. 4-day groups (P=0.028), 1-hour vs. 8-day groups (P=0.007), 4-hour vs. 4-day
groups (P=0.032), and 4-hour vs. 8-day groups (P=0.007) using Tukey pairwise
comparisons.
The duration of lordosis towards a heterospecific male was significantly longer after 1 or
4 h of exposure to heterospecific males than it was in females exposed to heterospecific
males for 4 or 8 days (F
3,38
=5.94, P=0.002; Fig. 1a). The duration of lordosis in naı
¨ve
females (i.e., females that were never exposed to heterospecific males before testing)
measured in previous studies (delBarco-Trillo et al. 2009b; delBarco-Trillo and Johnston
2010) was similar to the duration of lordosis in females in the 1-hour group (t
9.8
=1.47,
P=0.17) and the 4-hour group (t
11.8
=1.66, P=0.12; Fig. 1a); but higher compared to
females in the 4-day and 8-day groups (P\0.05; Fig. 1a).
Copulation within the 5 min trials occurred in 55.6% of the trials (5 out of 9) in the
1-hour group, in 54.6% of the trials (6 out of 11) in the 4-hour group, in 11.1% of the trials
(1 out of 9) in the 4-day group, and in 10% of the trials (1 out of 10) in the 8-day group.
These proportions were not significantly different between groups (two-tailed Fisher’s
exact test: P[0.05). No lordosis during the 5 min trials was observed in 11.1% of the
trials (1 out of 9) in the 1-hour group, in 9.1% of the trials (1 out of 11) in the 4-hour group,
in 33.3% of the trials (3 out of 9) in the 4-day group, and in 70% of the trials (7 out of 10)
in the 8-day group. Such absence of lordosis differed statistically between the 8-day group
and both the 1-hour group (two-tailed Fisher’s exact test: P=0.02) and the 4-hour group
(P=0.008). In all of the trials in which the female did not exhibit lordosis towards the
heterospecific male, the female was highly receptive when she was paired immediately
afterwards with an unfamiliar conspecific male (274.6 ±3.15 s of lordosis out of 300 s).
The latency to engage in aggressive behavior by female Syrian hamsters towards male
Turkish hamsters did not significantly differ in the 1-hour (175 ±43.67 s), 4-hour
(160.73 ±39.06 s), 4-day (72 ±33.1 s), or 8-day groups (115.9 ±38.74 s; F
3,38
=1.37,
396 Evol Ecol (2011) 25:391–401
123

P=0.27). The duration of aggression increased progressively in the 1-hour
(29.16 ±13.67 s), 4-hour (35.81 ±13.95 s), 4-day (63.93 ±18.46 s), and 8-day groups
(79.84 ±31 s) but there were no significant differences among the four groups
(F
3,38
=1.33, P=0.28; Fig. 1b). All these durations were significantly higher than those
of naı
¨ve females in previous studies (delBarco-Trillo et al. 2009b; delBarco-Trillo and
Johnston 2010; Mann–Whitney tests: P\0.05; Fig. 1b).
Experiment 2. Delaying the time between exposure and testing
Lordosis latency was significantly shorter in the group with a 8-day delay between
exposure to the heterospecific male and pairing with that male (67.5 ±26.8 s; t
18
=3.39,
P=0.003) than in the group with a 10-min delay (225.3 ±38.09 s). The duration of
lordosis was significantly longer in the 8-day delay group (169.4 ±30.24 s; t
18
=-3.35,
Fig. 1 Duration of lordosis (a) and aggression (b) by female Syrian hamsters (Mesocricetus auratus) after
different durations of exposure to a heterospecific male (Turkish hamster, M. brandti): 1, 4 h, 4, or 8 days.
As a reference, the dashed lines indicate the average duration of lordosis and aggression shown in previous
studies by naı
¨ve females, i.e., females that were never exposed to heterospecific males before testing
(delBarco-Trillo et al. 2009b; delBarco-Trillo and Johnston 2010). Different letters denote a statistically
significant difference. Values represent mean ±SEM
Evol Ecol (2011) 25:391–401 397
123

P=0.004; Fig. 2a) than in the 10-min delay group (40.3 ±23.98 s). The duration of
lordosis in naı
¨ve females measured in previous studies (delBarco-Trillo et al. 2009b;
delBarco-Trillo and Johnston 2010) was similar than for females in the 8-day delay group
(t
10.9
=1.15, P=0.28; Fig. 2a).
Copulation within the 5 min trials occurred in 10% of the trials (1 out of 10 trials) in the
group with a 10-min delay between exposure to the heterospecific male and pairing with
that male whereas it occurred in 70% of the trials (7 out of 10) in the 8-day delay group
(two-tailed Fisher’s exact test: P=0.02). No lordosis occurred in 10% of the trials in the
8-day delay group compared to 70% of the trials in the 10-min delay group (P=0.02). In
the only trial in the 8-day delay group in which the female did not show lordosis towards
the heterospecific male, the female was highly receptive when she was immediately paired
with an unfamiliar conspecific male (256 s of lordosis out of 300 s).
The latency to fight did not differ significantly between the 10-min delay group
(115.9 ±38.73 s) and the 8-day delay group (189.2 ±45.32 s; t
18
=-1.23, P=0.24).
Fig. 2 Duration of lordosis (a) and aggression (b) by female Syrian hamsters (Mesocricetus auratus)
which, after being exposed to a heterospecific male (Turkish hamster, M. brandti) for 8 days, were then
tested with that male 10 min later or 8 days later. As a reference, the dashed lines indicate the average
duration of lordosis and aggression shown in previous studies by naı
¨ve females, i.e., females that were never
exposed to heterospecific males before testing (delBarco-Trillo et al. 2009b; delBarco-Trillo and Johnston
2010). Different letters denote a statistically significant difference. Values represent mean ±SEM
398 Evol Ecol (2011) 25:391–401
123

The duration of aggression was shorter in the 8-day delay group (17.26 ±9.42 s) than in
the 10-min delay group (79.84 ±31 s), but this difference was not significantly different
(t
10.7
=1.93, P=0.08; Fig. 2b). The duration of aggression in both groups was signifi-
cantly higher than for naı
¨ve females in previous studies (delBarco-Trillo et al. 2009b;
delBarco-Trillo and Johnston 2010; Mann–Whitney tests: P\0.05; Fig. 2b).
Discussion
Our results indicate that, in female Syrian hamsters, a longer exposure to a heterospecific
male leads to a greater avoidance of interspecific mating (Experiment 1). Furthermore,
avoidance of interspecific mating can be lost if females are not periodically exposed to
stimuli from heterospecific males (Experiment 2). These experiments reinforce the results
of a previous study in which we showed that exposure to heterospecific males during
adulthood can influence subsequent female behavior towards such males (delBarco-Trillo
et al. 2010).
Our results agree with an evolutionary scenario in which the two species have had (a)
few interactions during their evolutionary history, and consequently few opportunities for a
faster and more resilient learned response to evolve, and/or (b) a relatively long period of
isolation since their last episode of overlapping distributions, resulting in a reduced effi-
ciency of that response (i.e., long time to start and reduction of the response when the
heterospecific stimuli disappears).
An alternative explanation is that evolution has not favored the learned response that we
suggest, and that instead females respond negatively to the abnormal behavior of the
heterospecific male. However, if this was the case, we should not have seen any significant
differences among groups (since heterospecific males appeared to behave similarly in all
groups). We can also rule out the possibility that females will become unreceptive toward
any male after a long exposure to him because in a previous study females exposed to
conspecific males (either kin, non-kin, losers or winners) for 8 days were sexually
receptive to them (delBarco-Trillo et al. 2010).
Experiments using pairs of closely-related Drosophila species have shown that male
flies exposed for 1 h to a heterospecific female later did not court other heterospecific
females (Dukas 2004,2008). Interestingly, in another study in which the exposure phase
was only 10 min, male fruit flies did not learn to avoid heterospecific females (Kandul
et al. 2006), even though there were several methodological differences between these
studies. The only other vertebrate system in which learning to avoid interspecific mating
during adulthood has been investigated was carried out with guppies (Poecilia spp.) and
these studies indicate that exposure to heterospecifics during several days is required for
accurate avoidance of interspecific mating (Haskins and Haskins 1949; Liley 1966; Ma-
gurran and Ramnarine 2004). For example, Poecilia reticulata males will show courtship
towards both conspecific and heterospecific (P.picta) females at random in allopatric
populations (Magurran and Ramnarine 2004). Up to 3 days of exposure to heterospecific
females does not have any significant effect on males, but after 4–6 days of exposure to
heterospecific females, males learn to avoid courting such females (Magurran and
Ramnarine 2004). In contrast to our study, female guppies seem unable to learn to avoid
interspecific mating, and behave similarly towards conspecific and heterospecific males
even after several days of exposure to heterospecific males (Haskins and Haskins 1949).
We predicted that female hamsters that had been exposed to a heterospecific male for
8 days would discriminate against this heterospecific male even after a period without
Evol Ecol (2011) 25:391–401 399
123

exposure to the heterospecific male. In contrast, we found that the learned avoidance of
interspecific mating requires maintained exposure to heterospecific males. A significant
period of time without such exposure (8 days in our study) resulted in the duration of
lordosis being similar to that shown by females that had never been exposed to hetero-
specific males. These results can be interpreted in two ways. First, females exposed to
heterospecific males may be forming a short-term discriminative memory for heterospe-
cific males. As long as females are exposed to heterospecific males, that type of memory
can be sustained. However, if heterospecific stimuli are not available to females for some
duration of time (8 days in this study), memory for such stimuli can be lost. Syrian
hamsters can retain memories of conspecific individuals at least for 10 days (Johnston
1993); however, we do not know if memories about heterospecific individuals are retained
that long or for a shorter amount of time. Alternatively, exposure to a heterospecific male
may change the physiological state of the female, for example by altering concentrations of
hormones in blood or neurotransmitters and receptors in the brain that modulate behavior
(Mucignat-Caretta et al. 2006). After 8 days without any heterospecific stimuli, physio-
logical levels of the relevant substances may return to baseline levels so that females
behave again as naı
¨ve females do, that is, responding similarly towards conspecific and
heterospecific males. Further research on the physiological mechanisms underlying the
results obtained in this study will be important for advancing our understanding of
avoidance of interspecific mating.
Our results indicate that lordosis and aggressive behaviors directed towards hetero-
specific males can be independent of one another. For example, lordosis duration in
females after 1 or 4 h of exposure did not differ compared to females without any previous
exposure to heterospecific males (delBarco-Trillo et al. 2009b; delBarco-Trillo and
Johnston 2010; Fig. 1a). In contrast, aggression was elicited in females after 1 or 4 h of
exposure to a heterospecific male compared to naı
¨ve females (which do not show any signs
of aggression towards heterospecific males; Fig. 1b). That is, short exposures to hetero-
specific males are sufficient to elicit aggression in females but such exposures may not
influence lordosis behavior.
Liley (1966) proposed that in order to learn to avoid interspecific mating, an individual
must be simultaneously exposed to both conspecifics and heterospecifics. In a previous
study we showed that female Syrian hamsters that were exposed to both a conspecific male
and a heterospecific male for 8 days learned to avoid interspecific mating (delBarco-Trillo
et al. 2010). However, we did not show whether exposure to a heterospecific male could be
sufficient for females to learn to avoid interspecific mating. In the present study, females
were exposed only to a heterospecific male and we also found that 8 days of exposure
significantly reduced lordosis behavior and increased aggressive behavior towards such
males. Thus these results show that females do not discriminate against heterospecific
males by making a comparison between conspecific males and heterospecific males.
Instead, it is the exposure to the stimuli from heterospecific males that results in females
learning to avoid interspecific mating.
Acknowledgments This work was supported by NIMH grant NIMH 5 R01 MHO58001-08 to
R. E. Johnston. The experiments here described comply with the current laws of the USA. All research was
conducted with approval from Cornell University’s Institutional Animal Care and Use Committee (protocol
#1993-0120).
Conflict of interest The authors declare that they have no conflict of interest.
400 Evol Ecol (2011) 25:391–401
123

References
Andersson M (1994) Sexual selection. Princeton University Press, Princeton
Bergman A, Feldman MW (1995) On the evolution of learning: representation of a stochastic environment.
Theor Popul Biol 48:251–276
delBarco-Trillo J, Johnston RE (2010) Fluoxetine does not prevent interspecific mating between two
hamster species. Physiol Behav 99:684–686
delBarco-Trillo J, Gulewicz K, Johnston RE (2009a) Medial amygdala involvement in discrimination of
same-species and closely-related-species male stimuli in estrous female Mesocricetus hamsters. Behav
Neurosci 123:758–763
delBarco-Trillo J, Gulewicz K, Segal A, McPhee ME, Johnston RE (2009b) Can captivity lead to inter-
species mating in two Mesocricetus hamster species? J Zool 278:308–312
delBarco-Trillo J, McPhee ME, Johnston RE (2010) Adult female hamsters avoid interspecific mating after
exposure to heterospecific males. Behav Ecol Sociobiol 64:1247–1253
Dukas R (2004) Male fruit flies learn to avoid interspecific courtship. Behav Ecol 15:695–698
Dukas R (2008) Learning decreases heterospecific courtship and mating in fruit flies. Biol Lett 4:645–647
Dukas R (2009) Dynamics of learning in the context of courtship in Drosophila persimilis and D. pseud-
oobscura. Anim Behav 77:253–259
Gattermann R, Fritzsche P, Neumann K, Al-Hussein I, Kayser A, Abiad M, Yakti R (2001) Notes on the
current distribution and the ecology of wild golden hamsters (Mesocricetus auratus). J Zool 254:
359–365
Gro
¨ning J, Hochkirch A (2008) Reproductive interference between animal species. Q Rev Biol 83:257–282
Hamar M, Schutowa M (1966) Neue daten u
¨ber die geographische vera
¨nderlichkeit und die entwicklung der
gattung Mesocricetus Nehring, 1898 (Glires, Mammalia). Z Saeugetierkunde 31:237–251
Haskins CP, Haskins EF (1949) The role of sexual selection as an isolating mechanism in three species of
poeciliid fishes. Evolution 3:160–169
Hebets EA (2007) Subadult female experience does not influence species recognition in the wolf spider
Schizocosa uetzi stratton 1997. J Arachnol 35:1–10
Irwin DE, Price T (1999) Sexual imprinting, learning and speciation. Heredity 82:347–354
Johnston RE (1993) Memory for individual scent in hamsters (Mesocricetus auratus) as assessed by
habituation methods. J Comp Psychol 107:201–207
Johnston RE, Peng A (2008) Memory for individuals: hamsters (Mesocricetus auratus) require contact to
develop multicomponent representations (concepts) of others. J Comp Psychol 122:121–131
Kandul NP, Wright KM, Kandul EV, Noor MAF (2006) No evidence for learned mating discrimination in
male Drosophila pseudoobscura. BMC Evol Biol 6:54
Liley NR (1966) Ethological isolating mechanisms in four sympatric species of poeciliid fishes. Behav
Suppl 13:1–197
Lisk RD (1985) The estrous cycle. In: Siegel HI (ed) The hamster: reproduction and behavior. Plenum Press,
New York, pp 23–51
Magurran AE, Ramnarine IW (2004) Learned mate recognition and reproductive isolation in guppies. Anim
Behav 67:1077–1082
Magurran AE, Ramnarine IW (2005) Evolution of mate discrimination in a fish. Curr Biol 15:R867–R868
Meredith M, Westberry JM (2004) Distinctive responses in the medial amygdala to same-species and
different-species pheromones. J Neurosci 24:5719–5725
Mucignat-Caretta C, Colivicchi MA, Fattori M, Ballini C, Bianchi L, Gabai G, Cavaggioni A, Della Corte L
(2006) Species-specific chemosignals evoke delayed excitacion of the vomeronasal amygdale in freely-
moving female rats. J Neurochem 99:881–891
Murphy MR (1977) Intraspecific sexual preferences of female hamsters. J Comp Physiol Psychol 91:
1337–1346
Neumann K, Michaux J, Lebedev V, Yigit N, Colak E, Ivanova N, Poltoraus A, Surov A, Markov G, Maak
S, Neumann S, Gattermann R (2006) Molecular phylogeny of the cricetinae subfamily based on the
mitochondrial cytochrome band 12S rRNA genes and the nuclear vWF gene. Mol Phylogenet Evol
39:135–148
Servedio M, Sæther S, Sætre G-P (2009) Reinforcement and learning. Evol Ecol 23:109–123
Todd NB, Nixon CW, Mulvaney DA, Connelly ME (1972) Karyotypes of Mesocricetus brandti and
hybridization within the genus. J Hered 63:73–77
Evol Ecol (2011) 25:391–401 401
123