Behavioural Processes 108 (2014) 71–79
Contents lists available at ScienceDirect
Behavioural Processes journal homepage: www.elsevier.com/locate/behavproc
Male-male sexual behavior in Japanese quail: Being “on top” reduces mating and fertilization with females Elizabeth Adkins-Regan ∗ Cornell University, Department of Psychology, 218 Uris Hall, Ithaca, NY 14853-7601, United States
a r t i c l e
i n f o
Article history: Received 2 May 2014 Received in revised form 16 September 2014 Accepted 17 September 2014 Available online 27 September 2014 Keywords: Japanese quail Coturnix japonica Same-sex sexual behavior Dominance Fertilization Sperm competition Reproductive tactic
a b s t r a c t Male Japanese quail (Coturnix japonica) engage in vigorous same-sex sexual interactions that have been interpreted as aggressive behavior reflecting dominance relationships. The consequences of this behavior for reproductive success, and whether it is a form of competition over mating and fertilization, are unclear. Three experiments were conducted to determine the effect of seeing or interacting with another male on a male’s subsequent mating and fertilization success with females. A vigorous interaction with another male in which the subject performed more cloacal contact movements (movements to try to make contact with the other bird’s cloacal opening) reduced subsequent mating and fertilization success with a female to a similar extent as a prior mating with a different female. Receiving one or more cloacal contacts from another male was less detrimental for subsequent success. The mere presence of another (stimulus) male delayed mating initiation in those male subjects that approached the stimulus first instead of the female. These results do not support the idea that the male “on top” in male–male sexual interactions is the dominant bird who goes on to achieve greater reproductive success. Instead, the results are consistent with male–male sexual behavior as an occasionally costly by-product of strong mating motivation. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Same-sex sexual behavior has been observed in many species and in both wild and domestic animals (Bailey and Zuk, 2009). Numerous hypotheses have been proposed to explain such behavior (Vasey and Sommer, 2006; Poiani, 2010). Some of these concern possible adaptive functions to help explain how the behavior is maintained over evolutionary time even though it has similar costs as heterosexual behavior (for example, sexually transmitted diseases or increased predation risk) but without the obvious direct fitness benefits of mating with the other sex. Sexual behavior between males has often been hypothesized to be a form of competitive dominance establishment that, like some other forms of male–male competition over mating, would be expected to have benefits in the form of greater reproductive success for the dominant individual. On the other hand, it is also possible that male–male sexual behavior is non-adaptive or an epiphenomenon of a high level of mating activity. Male–male sexual behavior occurs in a number of birds (MacFarlane et al., 2010) but because many of the reports are
∗ Tel.: +1 607 255 3834; fax: +1 607 255 8433. E-mail address: [email protected] http://dx.doi.org/10.1016/j.beproc.2014.09.027 0376-6357/© 2014 Elsevier B.V. All rights reserved.
from wild populations there are few experiments testing causal hypotheses about the consequences of the behavior for reproductive success. In the experiments reported here, the subjects are Japanese quail. Domestic male quail housed individually typically initiate mating immediately when placed with a female, and even those that are slower to mate usually initiate mating within the first 5 min if they mate at all (Schein et al., 1972). Mating consists of grabbing the feathers of the female’s head or neck, mounting (placing both feet on her back) and reaching back with the wings spread to try to achieve cloacal contact (termed a “cloacal contact movement”) (Fig. 1A and supplemental video 1). Because males of this species mate reliably and rapidly, Japanese quail are the most important avian model for analysis of the neural and hormonal mechanisms of copulatory behavior (Adkins-Regan, 1996; Ball and Balthazart, 2004). Males are very sexually active with both sexes, however. When tested with another male, a vigorous interaction occurs in which each bird tries to mount the other and achieve cloacal contact, occasionally succeeding (Adkins, 1974) (Fig. 1B and supplemental video 2). Beginning with Kuo (1960), these male–male interactions have nearly always been interpreted as fighting or aggressive conflict (e.g., Tsutsui and Ishii, 1981; Ramenofsky, 1984; Hirschenhauser et al., 2008). The male that is more often “on top,” i.e., the one that performs more mounts or cloacal contact movements, has
72
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
Fig. 1. Japanese quail sexual behavior. (A) A cloacal contact movement by a male (on the left) in a mating interaction with a female. See also supplemental video 1. (B) A cloacal contact movement by the male “on top” in an interaction with another male. Image captured from video. See also supplemental video 2, from which the image was captured.
been repeatedly described as dominant (e.g., Riters and Balthazart, 1998). Mating attempts with unreceptive females have a similarly vigorous and aggressive quality, and forced copulations are common (Adkins-Regan, 1995; Ophir and Galef, 2003). Although the sexual and aggressive behavior of wild Japanese quail has not been reported, the closely related European quail (Coturnix coturnix) has had a reputation for centuries as a highly “salacious” bird famous for its “lust” (Ray, 1678, p. 170). This suggests that the vigorous sexual tendencies are species-typical
characteristics of both members of the genus and not an artifact of the recent domestication of the Japanese quail. Reports from the field of C. coturnix as well as observations of genetically near-wild C. japonica in outdoor enclosures suggest flexible mating systems combining short-term pair relationships, mate switching and extra-pair matings (Nichols 1991; Rodriguez-Teijeiro et al., 2003; Sardà-Palomera et al., 2011). Males do not incubate the eggs or care for the chicks, which is consistent with the results of a comparative analysis of birds showing an association between more frequent male–male sexual behavior and absence of paternal care (MacFarlane et al., 2010). The consequences for reproductive success of the male–male sexual interactions of Japanese quail are unclear. Females have been shown to prefer the males described as the subordinates in the interactions (Ophir and Galef, 2003) but which male succeeds in fertilizing more eggs is not known and forced copulation can override female preference (Adkins-Regan, 1995). Here three questions are addressed. First, what is the influence of the presence of another male, but without allowing a direct interaction, on a male’s mating and fertilization success with a female (Experiment 1)? If the other male, a stimulus male, is perceived as a reproductive competitor but one that is unable to access the female (a subordinate status), the mating male (the subject) might be expected to show the enhanced mating and fertilization of a dominant individual. If, on the other hand, the stimulus male is a sexual opportunity or sexual stimulus, the effect on the subject could be negative (the subject has a conflict over which bird to mate with first), especially if the subject approaches the stimulus first instead of the female, or positive (the other male adds additional sexual stimulation and enhances the subject’s sexual motivation), especially if the subject approaches the female first. See Table 1 for a summary of the hypotheses and predictions for this and the following two experiments. Second, how is a male’s mating and fertilization success affected by a prior sexual interaction with another male? If those interactions are dominance encounters, and the dominant individual is the male “on top,” the dominant would be predicted to be successful at mating with and fertilizing females (Experiment 2). Alternatively, the male “on top” might be exhausted or sexually depleted (satiated) and less successful in a subsequent encounter with a female. Third, what is the effect of receiving cloacal contact (CC) from another male on the CC recipient’s mating and fertilization success (Experiment 3)? If those males are subordinates, a detrimental effect would be predicted for their success in a subsequent encounter with a female, both for behavioral reasons (the stress of defeat) and because of the role of male foam in sperm competition. Male Japanese quail produce a large amount of meringue-like foam from a special gland. The foam is passed to the cloaca along with the sperm during insemination. Foam enhances a male’s fertilization success in a competitive mating situation with another male (Finseth et al., 2013). Because male–male interactions also occasionally result in insemination, it is possible that the inseminator’s foam might reduce the fertilization success of the recipient when the latter then mates with a female. If so, that would suggest that copulating with another male could be a competitive reproductive tactic, a hypothesis supported in a study of Razorbills (Alca torda) that found that mounting other males was positively correlated with success at extra-pair copulations with females (Wagner, 1996). Alternatively, quail same-sex sexual behavior could be a tactic by the animal being mounted to divert the mounter from mating with females or to make the mounter waste his ejaculate (including foam, in the case of quail), a different sperm competition tactic (Jamieson and Craig, 1987; Birkhead and Møller, 1992). In this hypothesis the male “on the bottom” is not affected negatively with respect to success with females, and instead the male “on top”
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
73
Table 1 Hypotheses tested in the three experiments and their predictions for the mating and fertilization success of the male subjects (S) with female targets. Stim = stimulus male; CC = cloacal contact; ↑ = increase; ↓ = decrease; 0 = no change. Experiment
Hypothesis
Predicted effect on Mating
Fertilization
1
Stim is a competitor; S is dominant Stim is a sexual opportunity; S approaches Stim first Stim is a sexual stimulus; S approaches female first
↑ ↓ ↑
↑ 0 ↑
2
S is on top; S is dominant S is on top; S is then exhausted/depleted
↑ ↓
↑ ↓
3
CC recipient (S) is subordinate/defeated Stim’s foam interferes with CC recipient’s sperm
↓ 0
↓ ↓
would experience the negative effect that is tested in Experiment 2. 2. Methods 2.1. Animals Quail were hatched from fertilized eggs obtained from CBT Farms, Chestertown, MD. Beginning at age 4 weeks, all birds were housed individually on a 16-h light:8-h dark cycle to initiate and maintain reproductive activity. Subjects in the experiments were at least 8 weeks old, sexually mature (developed foam glands in males, regular egg laying by females), and sexually experienced (had participated in non-experimental mating trials to screen for ability to mate). Males and females were housed so that they could not see each other, and males used in the same test were not from adjacent cages. All animal use and procedures were approved by the Cornell University IACUC. 2.2. Assessment of mating behavior, insemination and fertilization success The mating sequence consists of grabbing the feathers of the other bird’s head or neck, mounting (placing both feet on its back) and reaching back with the wings spread (cloacal contact movement, or CCM, defined as just described). Some but not all CCMs result in contact between the two birds’ cloacal openings, the definition of cloacal contact, or CC. Initiating mating was defined as head grabbing, and head grab latency is the time from the beginning of the mating opportunity to the first head grab. Completing mating was defined as cloacal contact followed immediately by cessation of the mating attempt (Adkins-Regan, 1995). Because males transmit sperm and foam at the same time, the presence of the highly visible white foam in the female’s cloaca or in the drop pan under her home cage following mating confirms that insemination occurred (Adkins, 1974; Adkins-Regan, 1995). Not all copulatory attempts result in a completed mating, not all completed matings result in insemination, and not all inseminations fertilize any eggs (many do not) (Adkins-Regan, 1995). Thus there are three components to copulation-related male reproductive success, each of which is contingent on the prior component and subject to failure: (1) mating success (initiated and completed mating), (2) insemination (depositing sperm plus foam in the female) and (3) fertilization success (one or more of the eggs laid by the female after mating is fertilized). Female Japanese quail can store sperm for up to 11 days (Sittmann and Abplanalp, 1965; Birkhead and Fletcher, 1994). Eggs are laid in the late afternoon or early evening, and nearly 24 h elapse between the times a particular ovum is fertilized and the time that egg is laid. The first egg that could possibly be fertilized by a mating is the egg laid the day after mating, which, if laid late in the day,
would be found the next morning, on day 2 after mating on day 0. Therefore, eggs were collected daily in the mornings of days 2 through 11. Eggs were stored at 7.2 ◦ C prior to incubation at 37.5 ◦ C and approximately 30% relative humidity. After 1 week of incubation, they were broken open to check for the presence of an embryo. Only one early embryonic death was detected, which was counted as a fertilized egg. 2.3. Statistical analyses Proportions of trials with initiated mating, completed mating, or insemination were analyzed with binomial tests (for within-male comparisons) or Fisher’s exact tests (for betweenmale comparisons). Head grab latencies were analyzed with Wilcoxon signed-ranks tests (for within-male comparisons) or Mann–Whitney U-tests (for between-male comparisons). Egg fertilization outcomes have distributions that are markedly zero heavy and sometimes bimodal. Typically up to half of single inseminations fail to fertilize any eggs, and those that do fertilize between 1 and a maximum of 10 eggs, with a mean of about 4–5 fertilized eggs (Adkins-Regan, 1995). Because such distributions cannot be modeled using parametric statistics, non-parametric tests were also used to analyze those results. 3. Experiment 1: Effect of the presence of another male on mating and fertilization success 3.1. Methods A within-male design was used to compare success with female mating targets as a function of whether there was another male present and visible during the trial (a stimulus male, M trials), another female present (a stimulus female, F trials), or no third bird present (control, C trials). Male subjects (N = 31) had been pre-screened to eliminate any that failed to attempt mating with females. Female stimuli and mating targets were drawn from a pool of 96 and male stimuli from a pool of 35. Each subject male underwent three consecutive daily trials (i.e., one trial per day for 3 days), one of each type, with the three trial types in counterbalanced order and a different female mating target each day. Trials took place in an otherwise empty room in a wire mesh cage 1.8 m × 0.8 m × 0.9 m (height) with three smaller wire mesh holding cages inside, one for the subject male in the rear corner, one for the female mating target in the other rear corner, and one for the stimulus bird in the center front (which was empty for C trials). The locations of the male subject and the female mating target alternated between left and right rear corners across trials. To begin the trial each bird was placed in its small holding cage and remained there for 2 min (Phase I). The male subject and female target were 1.6 m apart, and the stimulus bird was 1.1 m from each of them. All birds could see each other
74
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
Fig. 2. Experiment 1. (A) Percentage of M (male stimulus) trials, F (female stimulus) trials and C (no stimulus bird, control) trials in which males initiated mating (head grabbed, clear bars), completed mating (gray bars) and produced a foam positive mating (black bars) with the target female. (B) Box plots of head grab latencies in M, F and C trials for those males that did initiate mating. Maximum possible is 180 s (the trial duration). M trials: another male was present during the mating opportunity with the target female; F trials: another female was present; C (control) trials: no third bird was present. *p < 0.05 compared with their C trials.
during Phase I but could not directly interact. After 2 min the subject male and female mating target were released to begin Phase II by pulling on ropes attached to their holding cages from behind a blind. The stimulus bird was never released but all three birds continued to have visual access to each other. The male subject’s behavior was observed from behind the blind and the trial was terminated 3 min after release or once the male completed mating, whichever occurred first, in order to limit the number of inseminations to one by preventing repeat mating. The subject’s head grab latency (time from release of the subject and the female mating target to his first head grab) was recorded along with whether the male completed mating and whether the male made contact with the stimulus bird’s cage or the released target female first. Successfully mated females were immediately checked for foam in the cloaca to confirm insemination, and if none was seen their drop pans were checked for foam for the next hour. Eggs laid by females with foam were collected and incubated to assess fertilization. The experiment was first carried out with nine male subjects and then repeated twice with a new set of 11 males each time.
3.2. Results The percentages of M, F and C trials in which males initiated mating (head grabbed), completed mating (ceased mating immediately following a cloacal contact), and inseminated the female (achieved a foam positive mating) in Phase II are shown in Fig. 2A. Mating completion was high and did not differ by trial type (M trials vs. C trials: p = 0.63; F trials vs. C trials: p = 0.11). Confirmed inseminations occurred in fewer trials than completed matings defined by behavior, as expected, but again did not differ by trial type (M trials vs. C trials: p = 0.77; F trials vs. C trials: p = 0.39). Across all subjects, head grab latencies in their M and C trials did not differ (Fig. 2B) (T = 153, N = 30, p > 0.1), but head grab latencies of the 13 males that made contact with the stimulus male’s cage first instead of the target female were longer than those of the 18 males that made contact with the target female first (medians 11 s vs. 4 s, respectively, U = 65, p = 0.04). Males were significantly slower to start head grabbing the target female in F trials than in C trials (Fig. 2B) (T = 88, N = 27, p = 0.013). Fourteen of the 31 males approached the caged stimulus female first instead of the target female that had been released. The head grab latencies of those 14 males were significantly longer than those of the 17 males that approached the target female first (medians 21 s vs. 6 s, respectively, U = 36, p = 0.001). Head grab latencies of males that
approached the target female first were not significantly longer than their latencies in their C trials (medians 6 s vs. 2 s, respectively, T = 45, N = 13, p > 0.9). Neither mating completion nor insemination differed for males that approached the stimulus first vs. those that did not in either the M or F trials (all p > 0.1). Egg fertilization outcomes of foam positive trials were similar for the three trial types. The percentages of trials producing at least one fertilized egg were 50% for M trials, 36% for F trials, and 45% for C trials. Within-male comparisons of numbers of fertilized eggs where both trials were foam positive confirmed the absence of any statistically significant differences as a function of trial type (Fig. 3) (M trials vs. C trials: medians 0 and 0, T = 31.5, N = 11, p = 0.89; F trials vs. C trials: medians 0 and 2, T = 13.5, N = 9, p = 0.30; M trials vs. F trials: medians 1 and 0, T = −17, N = 9, p = 0.57). The same comparisons of numbers of fertilized eggs but including all trials, regardless of whether they were foam positive, and entering zeros for foam negative or failed mating trials, also showed no significant differences (all medians = 0; M trials vs. C trials: T = −68, N = 17, p = 0.71; F trials vs. C trials: T = 24.5, N = 12, p = 0.27; M trials vs. F trials: T = −26, N = 13, p = 0.19). Between-male comparisons were made for M and for F trials between subjects that approached the male or female stimulus first and those that approached the target female first. The numbers of eggs fertilized did not differ between males that approached the stimulus first vs. those that did not in either the M or F trials (all p > 0.3).
4. Experiment 2: Effect of performing cloacal contact movements on another male on mating and fertilization success 4.1. Methods To test whether performing cloacal contact movements on another male impacted mating and fertilization success, male subjects were allowed to engage in a direct interaction with another male in which they were biased to be “on top” before mating with a target female. Prior to beginning the experiment, a pilot study was conducted with 10 males to see if larger body size (weight and tarsus length) would predict the outcome of interactions, which would allow a within-male design in which subjects were “on top” in some trials and “on the bottom” in others. The 10 subjects were tested in 2 min trials on 2 consecutive days in the same cage as in Experiment 1 (with small holding cages removed) with stimulus males that were either smaller or larger. Subject males had red
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
75
Fig. 3. Experiment 1. Within-male comparisons of number of eggs fertilized in foam positive (confirmed insemination) matings in M vs. C trials (A) and F vs. C trials (B). Each dot is the fertilization outcome with one target female; lines connect outcomes of the same male. M trials: another male was present during the mating opportunity with the target female; F trials: another female was present; C (control) trials: no third bird was present. There were no statistically significant differences between trial types (see text).
bands on both legs so they could be visually distinguished from the stimulus males while interacting. In all cases the subjects interacted vigorously with the other male, but size did not predict either which male performed more cloacal contact movements (CCMs), which achieved cloacal contact (CC), or which was “on top” more of the time. In five of the 10 trials where the subject was larger, the subject did perform more CCMs, but in three the stimulus male performed more, and in the remaining two trials both males performed the same number of CCMs. In seven of the 10 trials where the subject was smaller, the subject performed more CCMs, in two the stimulus male performed more, and in the remaining trial neither bird succeeded in performing any CCMs. Furthermore, in very few trials was there a consistent male “on top” Instead, males often changed roles during an interaction and alternated between performing CCMs or CCs and being a recipient of CCMs or CCs. In light of those pilot results, Experiment 2 used an alternative way of ensuring that the subjects would perform more CCMs than the stimuli. The alternative capitalized on the fact that all males are prescreened for mating performance in tests with females. During these prescreening trials, some males fail to attempt mating with females. Because of the similarity between male–female and male–male behavior, it was expected that the males that failed to attempt mating with females would also be unlikely to initiate sexual behavior with males, and so would be “on the bottom” more often in interactions with the subject males. Consequently, males that failed to attempt mating with females were used as stimulus males in Experiment 2. The subjects used in Experiment 2 were 24 males. None of the birds (subjects, stimuli, female mating targets) had been in any prior experiment or pilot study. Each subject underwent three consecutive daily trials (i.e., one trial per day for 3 days), one trial of each of three types in counterbalanced order: (1) subject interacts with stimulus male before mating with a mating target female (M trials), (2) subject interacts with a stimulus female before mating with a mating target female (F trials), and (3) no interaction prior to mating with a mating target female (control, C trials). The testing cage was the same as for Experiment 1 but without any small cages inside. Each trial proceeded in two phases as follows. For Phase I, the male subject and stimulus male (M trials) or the male subject and stimulus female (F trials) were placed in the testing cage. The mating target female was not in the room during Phase 1. The two birds,
subject and stimulus, were allowed to interact for 2 min, observing them from behind the blind. During the male–male interactions the number of times each male performed a CCM on the other male was recorded. Achieving cloacal contact (CC) was recorded separately. In F trials the male was allowed to mate with the stimulus female. For C trials the subject spent the 2 min alone in the testing cage. After 2 min the male or female stimulus was removed from the room and the target female was added to the testing cage. The trial continued for another 5 min (Phase II). The subject’s head grab latency and whether he completed the mating was recorded. Males were allowed to mate a second time in Phase II if sufficiently motivated and capable. Eggs were collected and incubated from all target females where the subjects had completed mating. 4.2. Results Two subjects never initiated mating with the target female in Phase II, even in their control trials, and were dropped from the analysis, leaving 22 males as subjects. Due to an error, one male did not get an F trial. All other subjects initiated mating with the stimulus female in Phase I of the F trials. The procedure of selecting as stimuli for the M trials those males that had failed to copulate with females when prescreened did succeed in biasing a majority of the male–male interactions in Phase I. Although all but one stimulus male did respond and participate vigorously in those interactions, subjects performed more CCMs than stimuli did in 15 of the 22 trials (in 12 of those, stimuli performed no CCMs), stimuli performed more in five trials, and the two males performed equal numbers in two trials (p < 0.05, sign test). Cloacal contacts (CCs) occurred in nine of the male–male interactions (43%) and were performed by the subject in three, the stimulus in four, and both males in two. The percentages of M, F and C trials in which males initiated and completed mating in Phase II are shown in Fig. 4A. Subjects were less likely to initiate mating and less likely to complete mating in M trials than in their C trials (initiate mating: p = 0.008; complete mating: p = 0.027). The six subjects that received CCs were not less likely to mate than subjects that did not receive CCs (1/6 vs. 9/16 failed to mate, respectively). When subjects did initiate mating, they had longer head grab latencies in M trials than in their C trials (T = 1, N = 8, p = 0.016) (Fig. 4B). This overall pattern of results was seen regardless of whether trials in which the subject did not
76
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
Fig. 4. Experiment 2. (A) Percentage of M trials, F trials and C trials in which males initiated mating (head grabbed, clear bars) and completed mating (gray bars) with the target female. (B) Box plots of head grab latencies in M, F and C trials for those males that did initiate mating. Maximum possible is 300 s (the trial duration). M trials: males interacted with another male prior to mating with the target female; F trials: males mated with a female prior to mating with a second target female; C (control) trials: no prior interaction. *p < 0.05, **p < 0.01, compared to their C trials.
perform more CCMs than the stimulus in Phase I were excluded; the same measures remained significantly different except for head grab latency, where the N became very small because of the failures to initiate mating in M trials (initiate mating: p = 0.023; complete mating: p = 0.04; head grab latency: T = 0, N = 5, p = 0.063). Both subjects that received CCs and those that did not had longer head grab latencies in their M trials than in their C trials (subjects receiving CCs: T = 0, N = 6, p = 0.03; subjects not receiving CCs: T = −1, N = 10, p = 0.004). Subjects were also less likely to initiate and complete mating in F trials than in their C trials (initiate mating: p = 0.021; complete mating: p = 0.012) (Fig. 4A). If they initiated mating, they were slower to head grab in F trials than in their C trials (T = 2, N = 8, p = 0.023) (Fig. 4B). There were no significant differences between measures in M vs. F trials (p > 0.5 for all comparisons). For fertilization success, because of the substantial number of failures to initiate mating with target females in M and F trials, there were too few within-male comparisons between completed mating trials for statistical analysis. Therefore analyses were done by including all trials and scoring mating failure trials as 0 eggs fertilized. Overall, only 23% of the males fertilized any egg in their M trials, and only 9% in their F trials, whereas they fertilized at least one egg in 41% in their C trials. Overall, the number of eggs fertilized did not differ between their M and C trials (Fig. 5A) (both medians = 0, T = 19, N = 11, p = 0.24). If, however, only the 15 males that performed more CCMs are included (i.e., only those relevant to testing the hypothesis), then those males fertilized fewer eggs in their M trials than in their C trials (medians 0 and 1.5, T = 2.5, N = 7, p = 0.047). Across all subjects, fewer eggs were fertilized in their F trials than in their C trials (Fig. 5B) (both medians = 0, T = 4, N = 10, p = 0.014). None of the five subjects that performed a CC fertilized an egg in their M trials, but two of the six subjects that received a CC did. Only two other males in the experiment managed to fertilize any eggs in their M trials. More trials with a CC would be needed to tell whether performing a CC reduces fertilization success. In each trial type, some males mated twice (60% in M trials, 30% in F trials, and 53% in C trials). Only in C trials were there enough males that completed mating at all to compare fertilization outcomes between single and double matings. The outcomes tended to be greater for double matings but not significantly so (median numbers fertilized = 3 for double matings vs. 0 for single matings, U = 17, p = 0.071). Even with double matings, three of those males did not fertilize any eggs.
5. Experiment 3: Effect of receiving cloacal contacts from another male on mating and fertilization success 5.1. Methods Whereas Experiment 2 focused on the effect of performing more CCMs than the stimulus male, Experiment 3 focused on the effect of being the recipient of cloacal contact (CC) so that the effect of possible transfer of foam from the male performing the CC could be better determined. Experienced successful maters (N = 18) were chosen for the male–male interactions. The outcome of the trial (which male received the CCs) determined which male was the subject that then had the mating opportunity with the target female. All males had been used as subjects in Experiment 2, but over 5 months had elapsed since then and they were not assigned to the same male or female as before. For each trial, two males interacted for up to 2 min in the testing cage and CCs by either were recorded (Phase I). If there were no CCs, the trial was terminated. If either male received a CC, the other male was immediately removed and the target female (from a pool of 50 regularly laying females) was introduced (Phase II). The trial then continued for a maximum of 10 min or until the male completed mating, whichever occurred first. Phase II was 10 min instead of the 5 min in Experiment 2 in order to increase the likelihood that a completed mating would occur so that fertilization success could be determined. In order to generate enough trials in which one male was the recipient of a CC, five to nine of these male–male (M) trials were conducted each day for 5 days, pairing males in different combinations for a total of 37 trials. Each of the 18 males also had a control trial (C trial) on a day between male–male trial days. Thus males were used on more than 1 day but not more than once in a day. For control trials, the male was simply placed in the testing cage with a new target female. Eggs were collected and incubated from all target females where the subjects had completed mating. 5.2. Results In 19 of the 37 trials, neither male achieved a CC. In the other 18 trials (49% of the trials), ten different males were CC recipients. All recipients then successfully completed mating with the target female. Only six of those 10 males completed a mating in their C trials. While those outcomes are not significantly different (p = 0.13), it is clear that, as in Experiment 2, receiving a CC did not reduce mating success with a female.
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
77
Fig. 5. Experiment 2. Within-male comparisons of number of eggs fertilized in M vs. C trials (A) and F vs. C trials (B). For clarity lines connecting cases of zero eggs fertilized in both trial types are omitted (10 double zero cases for M vs. C and 11 double zero cases for F vs. C). M trials: males interacted with another male prior to mating with the target female; F trials: males mated with a first female prior to mating with a second target female; C (control) trials: no prior interaction. *p < 0.05 compared to their C trials. #p < 0.05 compared to their C trials, excluding cases where the subject did not perform more cloacal contact movements than the stimulus.
Fig. 6. Experiment 3. Within-male comparisons of number of eggs fertilized in M vs. C trials. M trials: subjects received a cloacal contact from another male prior to mating with the target female; C (control) trials: no prior interaction. Males fertilized similar numbers of eggs in their two trial types.
To compare fertilization outcomes in the two kinds of trials, incomplete mating trials were scored as 0 eggs fertilized, and for the three males that received a CC in more than one M trial, the mean number of eggs fertilized in those repeated trials was used in the analysis. Males produced at least one fertilized egg in similar percentages of their M and C trials (50% of each) and they did not fertilize a significantly different number of eggs in the two kinds of trials (Fig. 6) (medians 1 and 0.75, T = 11.5, N = 7, p = 0.69). 6. Discussion A prior interaction with another male did not increase mating or fertilization success. In addition, receiving cloacal contacts from another male (being “on the bottom”) did not reduce mating or fertilization success. Instead, a prior interaction with another male in which the subject performed more cloacal contact movements caused those subjects to be slower to initiate mating and
decreased the probability that they would complete a mating, leading to reduced fertilization success. These detrimental effects on reproductive success were similar to those that occurred when the prior interaction was with a stimulus female. Being “on top” in a sexual encounter interfered with the success of a subsequent mating attempt regardless of whether that sexual activity was with another male or with a female. This is consistent with the idea that the males “on top” with either sex are then usually too fatigued or sexually depleted to mate successfully. These results fail to support the hypothesis that males “on top” are dominants that go on to have increased reproductive success, with males “on the bottom” subordinates having decreased reproductive success. They even argue against the assumption that the male–male interactions are dominance encounters producing clearly dominant and subordinate individuals. Instead, to the extent that males could be classed as “on top” or “on the bottom” and either male did better reproductively, it seemed to be the males “on the bottom,” because they experienced little or no detrimental effect on subsequent mating and fertilization success. If these male–male interactions reflect any competitive mating tactic, the tactic suggested by the results is that males engage other males in performing CCMs and CCs in order to reduce the performers’ success with females (Jamieson and Craig, 1987; Birkhead and Møller, 1992). Further research would be required before knowing if this hypothesis has merit. CCs occurred in less than half of trials and instead both males spent most of their time trying to perform CCMs. Furthermore, a different hypothesis, that male–male sexual behavior is an incidental by-product of strong mating motivation, is fully consistent with the results of the experiments. That is, males may have been selected to mate rapidly, vigorously and fairly indiscriminately, which although costly if the other bird happens to be another male, would be sufficiently beneficial whenever the other bird is a female to have a net adaptive advantage. Reports of other avian species also question the assumption that being “on top” in male–male sexual encounters indicates dominant status or that there are consistent dominants and subordinates in such encounters (e.g., Bertran and Margalida, 2003). House finches (Carpodacus mexicanus) that were subordinate based on other behavior were the ones mounting the dominants (McGraw and Hill, 1999). Among lek-breeding ruffs (Philomachus pugnax), there are three different male behavioral and plumage phenotypes, one of which, faeder, is a female mimic. Faeders are mounted by the other male types but also sometimes mount them (Jukema
78
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
and Piersma, 2006). A notable exception comes from a study of Razorbills (Alca torda), colonially breeding socially monogamous seabirds. Extra-pair copulation attempts by males are a common reproductive tactic, many males mount both sexes, and mountings of other males (but not being mounted by males) predicted successful extra-pair copulations with females (Wagner, 1996). A comparative analysis of male–male sexual behavior in birds concluded that higher frequencies of the behavior are associated with mating and parental care systems that produce selection for frequent mating opportunities (MacFarlane et al., 2010). The extrapair matings reported in the Japanese quail’s sibling species, C. coturnix, the frequent forced copulations in Japanese quail, and the remarkably short latencies to initiate mating also suggest that male–male sexual behavior might be a result of selection for rapidly and vigorously seizing any copulatory opportunity. Although the mounting males suffer a cost if mating immediately after a male–male encounter, there might be an overall benefit at other times. The success of the pre-selection of stimulus males in Experiment 2 means that males that are more vigorous at mating with females are also more vigorous at performing CCMs on other males. Studies with insects also suggest that high mating rate and frequent male–male sex may be genetically correlated traits (Burgevin et al., 2013). In Experiment 2 it was also interesting that the male stimuli pre-selected for failing to initiate mating with females nonetheless almost always responded to the male subjects’ copulatory attempts with attempts of their own. One male’s foam reduces the fertilizing success of another male when both have mated with a female (Finseth et al., 2013). Experiment 3 did not provide evidence, however, that receiving the foam of the male performing the cloacal contacts reduced the fertilizing ability of the recipient to any substantial extent. Although it is unlikely that the other male’s foam did not get mixed with the subject’s foam and sperm to exert any effect, that possibility cannot be ruled out, especially if females have any control over where inseminations are placed in their reproductive tracts, which is unknown in this species. Also, a greater number of trials would be needed to detect a small effect, especially given the likelihood that not all of the cloacal contacts resulted in insemination, just as in matings with females. In a similar vein, in Experiment 2 there were too few M trials with completed matings to tell whether males that performed a CC with the other male then fertilized fewer eggs (whether they wasted or reduced an ejaculate). What is clear is that those that vigorously engage in male–male sexual behavior are then much more likely to fail to mate at all, so that ejaculate amount is a secondary factor in fertilization success. In Experiment 1, there was a significant delay in initiating mating when the subject approached the stimulus male or female first instead of the target female. While this did not affect mating completion or insemination in the testing environment, in a less restricted environment such a delay would give the target female adequate time to leave and avoid the male’s mating attempt. Thus the results of Experiment 1, like those of Experiment 2, indicate a potential cost of male–male behavior. In addition, the observations in Experiment 1 of males that preferred to approach the stimulus male first and, in Experiment 2, of males that had shown no interest in mating with females in prescreening but engaged in sexual behavior with other males, together raise the possibility of a small percentage of males that consistently prefer other males to females for sexual behavior, as has been reported in sheep (Perkins and Fitzgerald, 1997). Additional research with choice tests of unscreened males would be needed to investigate this possibility. One less interesting explanation for the observed results of these experiments is that males simply mistake other males for females, depleting both their sperm and energy in the fruitless mating attempts. A combination of poor sex discrimination and high
mating rate has been hypothesized to account for male–male sexual behavior in both birds (especially those with sexually monomorphic plumage) and insects (Wagner, 1996; Burgevin et al., 2013). With greater experience, male Drosophila engage in less male–male sexual behavior (Bailey et al., 2013). Japanese quail plumage is only slightly sexually dimorphic. Indeed, evidence suggests that sexually inexperienced male Japanese quail learn to discriminate between females and males for mating through greater success with females (Nash and Domjan, 1991). The quail in the experiments reported here were sexually experienced, however, and in the absence of a female, even these experienced males readily engaged in sexual behavior with other males. The results of the Razorbill study also failed to support mistaken identity as a source of male–male mounting (Wagner, 1996). Thus, it seems unlikely that mistaken identity can account for all of these results. Male Japanese quail provide unique advantages for research on same-sex sexual behavior. Because of the high mating rate with both sexes and the ease of determining insemination and fertilization success, they allow experimental testing of hypotheses about same-sex behavior. The experiments reported here provide a new interpretation of their male–male behavior and both question the assumption that the behavior has anything to do with dominance and show that the male “on top” does not have a short-term advantage in reproductive success. Their reputation as “highly salacious” creatures is centuries old but many questions remain about the evolution and function of their mating habits. Acknowledgements Support came from NSF IBN-9514088 and IBN-0130986 during this work. The author thanks Michael Fairchild, Emiko MacKillop, and Andrew Mong for assistance with Experiment 1, Tim Van Deusen for animal care management, Findley Finseth for discussion of Experiment 3, Nicole Baran for comments on the manuscript and two reviewers for their helpful comments that improved the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.beproc.2014.09.027. References Adkins, E.K., 1974. Electrical recording of copulation in quail. Physiol. Behav. 13, 475–477. Adkins-Regan, E., 1995. Predictors of fertilization in the Japanese quail, Coturnix japonica. Anim. Behav. 50, 1405–1415. Adkins-Regan, E., 1996. Neuroanatomy of sexual behavior in the male Japanese quail from top to bottom. Poult. Avian Biol. Rev. 7, 193–204. Bailey, N.W., Zuk, M., 2009. Same-sex sexual behavior and evolution. Trends Ecol. Evol. 24, 439–446. Bailey, N.W., Hoskins, J.L., Green, J., Ritchie, M.G., 2013. Measuring same-sex sexual behavior: the influence of the male social environment. Anim. Behav. 86, 91–100. Ball, G.F., Balthazart, J., 2004. Hormonal regulation of brain circuits mediating male sexual behavior in birds. Physiol. Behav. 83, 329–346. Bertran, J., Margalida, A., 2003. Male–male mountings in polyandrous bearded vultures (Gypaetus barbatus): an unusual behaviour in raptors. J. Avian Biol. 34, 334–338. Birkhead, T.R., Fletcher, F., 1994. Sperm storage and the release of sperm from the sperm storage tubules in Japanese quail Coturnix japonica. Ibis 136, 101–104. Burgevin, L., Friberg, U., Maklakov, A.A., 2013. Intersexual correlation for same-sex sexual behavior in an insect. Anim. Behav. 85, 759–762. Birkhead, T.R., Møller, A.P., 1992. Sperm Competition in Birds: Evolutionary Causes and Consequences. Academic Press, London. Jukema, J., Piersma, T., 2006. Permanent female mimics in a lekking shorebird. Biol. Lett. 2, 161–164. Finseth, F.R., Iacovelli, S.R., Harrison, R.G., Adkins-Regan, E.K., 2013. A non-semen copulatory fluid influences the outcome of sperm competition in Japanese quail. J. Evol. Biol. 26, 1875–1889.
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79 Hirschenhauser, K., Wittek, M., Johnston, P., Möstl, E., 2008. Social context rather than behavioral output or winning modulates post-conflict testosterone responses in Japanese quail (Coturnix japonica). Physiol. Behav. 95, 457–463. Jamieson, I.G., Craig, J.L., 1987. Male–male and female–female courtship and copulation behavior in a communally breeding bird. Anim. Behav. 35, 1251–1253. Kuo, Z.Y., 1960. Studies on the basic factors in animal fighting: III. Hormonal factors affecting fighting in quails. J. Genet. Psychol. 96, 217–223. MacFarlane, G.R., Blomberg, S.P., Vasey, P.L., 2010. Homosexual behaviour in birds: frequency of expression is related to parental care disparity between the sexes. Anim. Behav. 80, 375–390. McGraw, K.J., Hill, G.E., 1999. Induced homosexual behaviour in male house finches (Carpodacus mexicanus): the “Prisoner Effect.”. Ethol. Ecol. Evol. 11, 197–201. Nash, S., Domjan, M., 1991. Learning to discriminate the sex of conspecifics in male Japanese quail (Coturnix coturnix japonica) – tests of biological constraints. J. Exp. Psychol.-Anim. Behav. Process. 17, 342–353. Nichols, C.R., 1991. A comparison of the reproductive and behavioral differences in feral and domestic Japanese quail. Master’s Thesis. University of British Columbia. Ophir, A.G., Galef, B.G., 2003. Female Japanese quail that ‘eavesdrop’ on fighting males prefer losers to winners. Anim. Behav. 66, 399–407. Perkins, A., Fitzgerald, J.A., 1997. Sexual orientation in domestic rams: some biological and social correlates. In: Ellis, L., Ebertz, L. (Eds.), Sexual Orientation: Toward Biological Understanding. Westport, Conn, Praeger, pp. 107–128. Poiani, A., 2010. Animal Homosexuality: A Biosocial Perspective. Cambridge University Press, Cambridge UK.
79
Ramenofsky, M., 1984. Agonistic behavior and endogenous plasma hormones in male Japanese quail. Anim. Behav. 32, 698–708. Riters, L.V., Balthazart, J., 1998. Behavioral evidence for individual recognition in Japanese quail. Behaviour 135, 551–578. Rodriguez-Teijeiro, J.D., Puigcerver, M., Gallego, S., Cordero, P.J., Parkin, D.T., 2003. Pair bonding and multiple paternity in the polygamous common quail Coturnix coturnix. Ethology 109, 291–302. Sardà-Palomera, F., Puigcerver, M., Vinyoles, D., Rodriguez-Teijeiro, J.D., 2011. Exploring male and female preferences, male body condition, and pair bonds in the evolution of male sexual aggregation: the case of the Common Quail (Coturnix coturnix). Can. J. Zool. 89, 325–333. Schein, M.W., Diamond, M., Carter, C.S., 1972. Sexual performance levels of male Japanese quail (Coturnix coturnix japonica). Anim. Behav. 20, 61–67. Sittmann, K., Abplanalp, H., 1965. Duration and recovery of fertility in Japanese quail (Coturnix coturnix japonica). Brit. Poult. Sci. 6, 245–250. Tsutsui, K., Ishii, S., 1981. Effects of sex steroids on aggressive behavior of adult male Japanese quail. Gen. Comp. Endocrinol. 44, 480–486. Vasey, P.L., Sommer, V., 2006. Homosexual behavior in animals: topics, hypotheses and research trajectories. In: Sommer, V., Vasey, P.L. (Eds.), Homosexual Behaviour in Animals: An Evolutionary Perspective. Cambridge University Press, Cambridge, UK, pp. 3–44. Wagner, R.H., 1996. Male–male mountings by a sexually monomorphic bird: mistaken identity or fighting tactic? J. Avian Biol. 27, 209–214. Ray, J., 1678. The Ornithology of Francis Willughby. Royal Society, London.
Contents lists available at ScienceDirect
Behavioural Processes journal homepage: www.elsevier.com/locate/behavproc
Male-male sexual behavior in Japanese quail: Being “on top” reduces mating and fertilization with females Elizabeth Adkins-Regan ∗ Cornell University, Department of Psychology, 218 Uris Hall, Ithaca, NY 14853-7601, United States
a r t i c l e
i n f o
Article history: Received 2 May 2014 Received in revised form 16 September 2014 Accepted 17 September 2014 Available online 27 September 2014 Keywords: Japanese quail Coturnix japonica Same-sex sexual behavior Dominance Fertilization Sperm competition Reproductive tactic
a b s t r a c t Male Japanese quail (Coturnix japonica) engage in vigorous same-sex sexual interactions that have been interpreted as aggressive behavior reflecting dominance relationships. The consequences of this behavior for reproductive success, and whether it is a form of competition over mating and fertilization, are unclear. Three experiments were conducted to determine the effect of seeing or interacting with another male on a male’s subsequent mating and fertilization success with females. A vigorous interaction with another male in which the subject performed more cloacal contact movements (movements to try to make contact with the other bird’s cloacal opening) reduced subsequent mating and fertilization success with a female to a similar extent as a prior mating with a different female. Receiving one or more cloacal contacts from another male was less detrimental for subsequent success. The mere presence of another (stimulus) male delayed mating initiation in those male subjects that approached the stimulus first instead of the female. These results do not support the idea that the male “on top” in male–male sexual interactions is the dominant bird who goes on to achieve greater reproductive success. Instead, the results are consistent with male–male sexual behavior as an occasionally costly by-product of strong mating motivation. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Same-sex sexual behavior has been observed in many species and in both wild and domestic animals (Bailey and Zuk, 2009). Numerous hypotheses have been proposed to explain such behavior (Vasey and Sommer, 2006; Poiani, 2010). Some of these concern possible adaptive functions to help explain how the behavior is maintained over evolutionary time even though it has similar costs as heterosexual behavior (for example, sexually transmitted diseases or increased predation risk) but without the obvious direct fitness benefits of mating with the other sex. Sexual behavior between males has often been hypothesized to be a form of competitive dominance establishment that, like some other forms of male–male competition over mating, would be expected to have benefits in the form of greater reproductive success for the dominant individual. On the other hand, it is also possible that male–male sexual behavior is non-adaptive or an epiphenomenon of a high level of mating activity. Male–male sexual behavior occurs in a number of birds (MacFarlane et al., 2010) but because many of the reports are
∗ Tel.: +1 607 255 3834; fax: +1 607 255 8433. E-mail address: [email protected] http://dx.doi.org/10.1016/j.beproc.2014.09.027 0376-6357/© 2014 Elsevier B.V. All rights reserved.
from wild populations there are few experiments testing causal hypotheses about the consequences of the behavior for reproductive success. In the experiments reported here, the subjects are Japanese quail. Domestic male quail housed individually typically initiate mating immediately when placed with a female, and even those that are slower to mate usually initiate mating within the first 5 min if they mate at all (Schein et al., 1972). Mating consists of grabbing the feathers of the female’s head or neck, mounting (placing both feet on her back) and reaching back with the wings spread to try to achieve cloacal contact (termed a “cloacal contact movement”) (Fig. 1A and supplemental video 1). Because males of this species mate reliably and rapidly, Japanese quail are the most important avian model for analysis of the neural and hormonal mechanisms of copulatory behavior (Adkins-Regan, 1996; Ball and Balthazart, 2004). Males are very sexually active with both sexes, however. When tested with another male, a vigorous interaction occurs in which each bird tries to mount the other and achieve cloacal contact, occasionally succeeding (Adkins, 1974) (Fig. 1B and supplemental video 2). Beginning with Kuo (1960), these male–male interactions have nearly always been interpreted as fighting or aggressive conflict (e.g., Tsutsui and Ishii, 1981; Ramenofsky, 1984; Hirschenhauser et al., 2008). The male that is more often “on top,” i.e., the one that performs more mounts or cloacal contact movements, has
72
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
Fig. 1. Japanese quail sexual behavior. (A) A cloacal contact movement by a male (on the left) in a mating interaction with a female. See also supplemental video 1. (B) A cloacal contact movement by the male “on top” in an interaction with another male. Image captured from video. See also supplemental video 2, from which the image was captured.
been repeatedly described as dominant (e.g., Riters and Balthazart, 1998). Mating attempts with unreceptive females have a similarly vigorous and aggressive quality, and forced copulations are common (Adkins-Regan, 1995; Ophir and Galef, 2003). Although the sexual and aggressive behavior of wild Japanese quail has not been reported, the closely related European quail (Coturnix coturnix) has had a reputation for centuries as a highly “salacious” bird famous for its “lust” (Ray, 1678, p. 170). This suggests that the vigorous sexual tendencies are species-typical
characteristics of both members of the genus and not an artifact of the recent domestication of the Japanese quail. Reports from the field of C. coturnix as well as observations of genetically near-wild C. japonica in outdoor enclosures suggest flexible mating systems combining short-term pair relationships, mate switching and extra-pair matings (Nichols 1991; Rodriguez-Teijeiro et al., 2003; Sardà-Palomera et al., 2011). Males do not incubate the eggs or care for the chicks, which is consistent with the results of a comparative analysis of birds showing an association between more frequent male–male sexual behavior and absence of paternal care (MacFarlane et al., 2010). The consequences for reproductive success of the male–male sexual interactions of Japanese quail are unclear. Females have been shown to prefer the males described as the subordinates in the interactions (Ophir and Galef, 2003) but which male succeeds in fertilizing more eggs is not known and forced copulation can override female preference (Adkins-Regan, 1995). Here three questions are addressed. First, what is the influence of the presence of another male, but without allowing a direct interaction, on a male’s mating and fertilization success with a female (Experiment 1)? If the other male, a stimulus male, is perceived as a reproductive competitor but one that is unable to access the female (a subordinate status), the mating male (the subject) might be expected to show the enhanced mating and fertilization of a dominant individual. If, on the other hand, the stimulus male is a sexual opportunity or sexual stimulus, the effect on the subject could be negative (the subject has a conflict over which bird to mate with first), especially if the subject approaches the stimulus first instead of the female, or positive (the other male adds additional sexual stimulation and enhances the subject’s sexual motivation), especially if the subject approaches the female first. See Table 1 for a summary of the hypotheses and predictions for this and the following two experiments. Second, how is a male’s mating and fertilization success affected by a prior sexual interaction with another male? If those interactions are dominance encounters, and the dominant individual is the male “on top,” the dominant would be predicted to be successful at mating with and fertilizing females (Experiment 2). Alternatively, the male “on top” might be exhausted or sexually depleted (satiated) and less successful in a subsequent encounter with a female. Third, what is the effect of receiving cloacal contact (CC) from another male on the CC recipient’s mating and fertilization success (Experiment 3)? If those males are subordinates, a detrimental effect would be predicted for their success in a subsequent encounter with a female, both for behavioral reasons (the stress of defeat) and because of the role of male foam in sperm competition. Male Japanese quail produce a large amount of meringue-like foam from a special gland. The foam is passed to the cloaca along with the sperm during insemination. Foam enhances a male’s fertilization success in a competitive mating situation with another male (Finseth et al., 2013). Because male–male interactions also occasionally result in insemination, it is possible that the inseminator’s foam might reduce the fertilization success of the recipient when the latter then mates with a female. If so, that would suggest that copulating with another male could be a competitive reproductive tactic, a hypothesis supported in a study of Razorbills (Alca torda) that found that mounting other males was positively correlated with success at extra-pair copulations with females (Wagner, 1996). Alternatively, quail same-sex sexual behavior could be a tactic by the animal being mounted to divert the mounter from mating with females or to make the mounter waste his ejaculate (including foam, in the case of quail), a different sperm competition tactic (Jamieson and Craig, 1987; Birkhead and Møller, 1992). In this hypothesis the male “on the bottom” is not affected negatively with respect to success with females, and instead the male “on top”
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
73
Table 1 Hypotheses tested in the three experiments and their predictions for the mating and fertilization success of the male subjects (S) with female targets. Stim = stimulus male; CC = cloacal contact; ↑ = increase; ↓ = decrease; 0 = no change. Experiment
Hypothesis
Predicted effect on Mating
Fertilization
1
Stim is a competitor; S is dominant Stim is a sexual opportunity; S approaches Stim first Stim is a sexual stimulus; S approaches female first
↑ ↓ ↑
↑ 0 ↑
2
S is on top; S is dominant S is on top; S is then exhausted/depleted
↑ ↓
↑ ↓
3
CC recipient (S) is subordinate/defeated Stim’s foam interferes with CC recipient’s sperm
↓ 0
↓ ↓
would experience the negative effect that is tested in Experiment 2. 2. Methods 2.1. Animals Quail were hatched from fertilized eggs obtained from CBT Farms, Chestertown, MD. Beginning at age 4 weeks, all birds were housed individually on a 16-h light:8-h dark cycle to initiate and maintain reproductive activity. Subjects in the experiments were at least 8 weeks old, sexually mature (developed foam glands in males, regular egg laying by females), and sexually experienced (had participated in non-experimental mating trials to screen for ability to mate). Males and females were housed so that they could not see each other, and males used in the same test were not from adjacent cages. All animal use and procedures were approved by the Cornell University IACUC. 2.2. Assessment of mating behavior, insemination and fertilization success The mating sequence consists of grabbing the feathers of the other bird’s head or neck, mounting (placing both feet on its back) and reaching back with the wings spread (cloacal contact movement, or CCM, defined as just described). Some but not all CCMs result in contact between the two birds’ cloacal openings, the definition of cloacal contact, or CC. Initiating mating was defined as head grabbing, and head grab latency is the time from the beginning of the mating opportunity to the first head grab. Completing mating was defined as cloacal contact followed immediately by cessation of the mating attempt (Adkins-Regan, 1995). Because males transmit sperm and foam at the same time, the presence of the highly visible white foam in the female’s cloaca or in the drop pan under her home cage following mating confirms that insemination occurred (Adkins, 1974; Adkins-Regan, 1995). Not all copulatory attempts result in a completed mating, not all completed matings result in insemination, and not all inseminations fertilize any eggs (many do not) (Adkins-Regan, 1995). Thus there are three components to copulation-related male reproductive success, each of which is contingent on the prior component and subject to failure: (1) mating success (initiated and completed mating), (2) insemination (depositing sperm plus foam in the female) and (3) fertilization success (one or more of the eggs laid by the female after mating is fertilized). Female Japanese quail can store sperm for up to 11 days (Sittmann and Abplanalp, 1965; Birkhead and Fletcher, 1994). Eggs are laid in the late afternoon or early evening, and nearly 24 h elapse between the times a particular ovum is fertilized and the time that egg is laid. The first egg that could possibly be fertilized by a mating is the egg laid the day after mating, which, if laid late in the day,
would be found the next morning, on day 2 after mating on day 0. Therefore, eggs were collected daily in the mornings of days 2 through 11. Eggs were stored at 7.2 ◦ C prior to incubation at 37.5 ◦ C and approximately 30% relative humidity. After 1 week of incubation, they were broken open to check for the presence of an embryo. Only one early embryonic death was detected, which was counted as a fertilized egg. 2.3. Statistical analyses Proportions of trials with initiated mating, completed mating, or insemination were analyzed with binomial tests (for within-male comparisons) or Fisher’s exact tests (for betweenmale comparisons). Head grab latencies were analyzed with Wilcoxon signed-ranks tests (for within-male comparisons) or Mann–Whitney U-tests (for between-male comparisons). Egg fertilization outcomes have distributions that are markedly zero heavy and sometimes bimodal. Typically up to half of single inseminations fail to fertilize any eggs, and those that do fertilize between 1 and a maximum of 10 eggs, with a mean of about 4–5 fertilized eggs (Adkins-Regan, 1995). Because such distributions cannot be modeled using parametric statistics, non-parametric tests were also used to analyze those results. 3. Experiment 1: Effect of the presence of another male on mating and fertilization success 3.1. Methods A within-male design was used to compare success with female mating targets as a function of whether there was another male present and visible during the trial (a stimulus male, M trials), another female present (a stimulus female, F trials), or no third bird present (control, C trials). Male subjects (N = 31) had been pre-screened to eliminate any that failed to attempt mating with females. Female stimuli and mating targets were drawn from a pool of 96 and male stimuli from a pool of 35. Each subject male underwent three consecutive daily trials (i.e., one trial per day for 3 days), one of each type, with the three trial types in counterbalanced order and a different female mating target each day. Trials took place in an otherwise empty room in a wire mesh cage 1.8 m × 0.8 m × 0.9 m (height) with three smaller wire mesh holding cages inside, one for the subject male in the rear corner, one for the female mating target in the other rear corner, and one for the stimulus bird in the center front (which was empty for C trials). The locations of the male subject and the female mating target alternated between left and right rear corners across trials. To begin the trial each bird was placed in its small holding cage and remained there for 2 min (Phase I). The male subject and female target were 1.6 m apart, and the stimulus bird was 1.1 m from each of them. All birds could see each other
74
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
Fig. 2. Experiment 1. (A) Percentage of M (male stimulus) trials, F (female stimulus) trials and C (no stimulus bird, control) trials in which males initiated mating (head grabbed, clear bars), completed mating (gray bars) and produced a foam positive mating (black bars) with the target female. (B) Box plots of head grab latencies in M, F and C trials for those males that did initiate mating. Maximum possible is 180 s (the trial duration). M trials: another male was present during the mating opportunity with the target female; F trials: another female was present; C (control) trials: no third bird was present. *p < 0.05 compared with their C trials.
during Phase I but could not directly interact. After 2 min the subject male and female mating target were released to begin Phase II by pulling on ropes attached to their holding cages from behind a blind. The stimulus bird was never released but all three birds continued to have visual access to each other. The male subject’s behavior was observed from behind the blind and the trial was terminated 3 min after release or once the male completed mating, whichever occurred first, in order to limit the number of inseminations to one by preventing repeat mating. The subject’s head grab latency (time from release of the subject and the female mating target to his first head grab) was recorded along with whether the male completed mating and whether the male made contact with the stimulus bird’s cage or the released target female first. Successfully mated females were immediately checked for foam in the cloaca to confirm insemination, and if none was seen their drop pans were checked for foam for the next hour. Eggs laid by females with foam were collected and incubated to assess fertilization. The experiment was first carried out with nine male subjects and then repeated twice with a new set of 11 males each time.
3.2. Results The percentages of M, F and C trials in which males initiated mating (head grabbed), completed mating (ceased mating immediately following a cloacal contact), and inseminated the female (achieved a foam positive mating) in Phase II are shown in Fig. 2A. Mating completion was high and did not differ by trial type (M trials vs. C trials: p = 0.63; F trials vs. C trials: p = 0.11). Confirmed inseminations occurred in fewer trials than completed matings defined by behavior, as expected, but again did not differ by trial type (M trials vs. C trials: p = 0.77; F trials vs. C trials: p = 0.39). Across all subjects, head grab latencies in their M and C trials did not differ (Fig. 2B) (T = 153, N = 30, p > 0.1), but head grab latencies of the 13 males that made contact with the stimulus male’s cage first instead of the target female were longer than those of the 18 males that made contact with the target female first (medians 11 s vs. 4 s, respectively, U = 65, p = 0.04). Males were significantly slower to start head grabbing the target female in F trials than in C trials (Fig. 2B) (T = 88, N = 27, p = 0.013). Fourteen of the 31 males approached the caged stimulus female first instead of the target female that had been released. The head grab latencies of those 14 males were significantly longer than those of the 17 males that approached the target female first (medians 21 s vs. 6 s, respectively, U = 36, p = 0.001). Head grab latencies of males that
approached the target female first were not significantly longer than their latencies in their C trials (medians 6 s vs. 2 s, respectively, T = 45, N = 13, p > 0.9). Neither mating completion nor insemination differed for males that approached the stimulus first vs. those that did not in either the M or F trials (all p > 0.1). Egg fertilization outcomes of foam positive trials were similar for the three trial types. The percentages of trials producing at least one fertilized egg were 50% for M trials, 36% for F trials, and 45% for C trials. Within-male comparisons of numbers of fertilized eggs where both trials were foam positive confirmed the absence of any statistically significant differences as a function of trial type (Fig. 3) (M trials vs. C trials: medians 0 and 0, T = 31.5, N = 11, p = 0.89; F trials vs. C trials: medians 0 and 2, T = 13.5, N = 9, p = 0.30; M trials vs. F trials: medians 1 and 0, T = −17, N = 9, p = 0.57). The same comparisons of numbers of fertilized eggs but including all trials, regardless of whether they were foam positive, and entering zeros for foam negative or failed mating trials, also showed no significant differences (all medians = 0; M trials vs. C trials: T = −68, N = 17, p = 0.71; F trials vs. C trials: T = 24.5, N = 12, p = 0.27; M trials vs. F trials: T = −26, N = 13, p = 0.19). Between-male comparisons were made for M and for F trials between subjects that approached the male or female stimulus first and those that approached the target female first. The numbers of eggs fertilized did not differ between males that approached the stimulus first vs. those that did not in either the M or F trials (all p > 0.3).
4. Experiment 2: Effect of performing cloacal contact movements on another male on mating and fertilization success 4.1. Methods To test whether performing cloacal contact movements on another male impacted mating and fertilization success, male subjects were allowed to engage in a direct interaction with another male in which they were biased to be “on top” before mating with a target female. Prior to beginning the experiment, a pilot study was conducted with 10 males to see if larger body size (weight and tarsus length) would predict the outcome of interactions, which would allow a within-male design in which subjects were “on top” in some trials and “on the bottom” in others. The 10 subjects were tested in 2 min trials on 2 consecutive days in the same cage as in Experiment 1 (with small holding cages removed) with stimulus males that were either smaller or larger. Subject males had red
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
75
Fig. 3. Experiment 1. Within-male comparisons of number of eggs fertilized in foam positive (confirmed insemination) matings in M vs. C trials (A) and F vs. C trials (B). Each dot is the fertilization outcome with one target female; lines connect outcomes of the same male. M trials: another male was present during the mating opportunity with the target female; F trials: another female was present; C (control) trials: no third bird was present. There were no statistically significant differences between trial types (see text).
bands on both legs so they could be visually distinguished from the stimulus males while interacting. In all cases the subjects interacted vigorously with the other male, but size did not predict either which male performed more cloacal contact movements (CCMs), which achieved cloacal contact (CC), or which was “on top” more of the time. In five of the 10 trials where the subject was larger, the subject did perform more CCMs, but in three the stimulus male performed more, and in the remaining two trials both males performed the same number of CCMs. In seven of the 10 trials where the subject was smaller, the subject performed more CCMs, in two the stimulus male performed more, and in the remaining trial neither bird succeeded in performing any CCMs. Furthermore, in very few trials was there a consistent male “on top” Instead, males often changed roles during an interaction and alternated between performing CCMs or CCs and being a recipient of CCMs or CCs. In light of those pilot results, Experiment 2 used an alternative way of ensuring that the subjects would perform more CCMs than the stimuli. The alternative capitalized on the fact that all males are prescreened for mating performance in tests with females. During these prescreening trials, some males fail to attempt mating with females. Because of the similarity between male–female and male–male behavior, it was expected that the males that failed to attempt mating with females would also be unlikely to initiate sexual behavior with males, and so would be “on the bottom” more often in interactions with the subject males. Consequently, males that failed to attempt mating with females were used as stimulus males in Experiment 2. The subjects used in Experiment 2 were 24 males. None of the birds (subjects, stimuli, female mating targets) had been in any prior experiment or pilot study. Each subject underwent three consecutive daily trials (i.e., one trial per day for 3 days), one trial of each of three types in counterbalanced order: (1) subject interacts with stimulus male before mating with a mating target female (M trials), (2) subject interacts with a stimulus female before mating with a mating target female (F trials), and (3) no interaction prior to mating with a mating target female (control, C trials). The testing cage was the same as for Experiment 1 but without any small cages inside. Each trial proceeded in two phases as follows. For Phase I, the male subject and stimulus male (M trials) or the male subject and stimulus female (F trials) were placed in the testing cage. The mating target female was not in the room during Phase 1. The two birds,
subject and stimulus, were allowed to interact for 2 min, observing them from behind the blind. During the male–male interactions the number of times each male performed a CCM on the other male was recorded. Achieving cloacal contact (CC) was recorded separately. In F trials the male was allowed to mate with the stimulus female. For C trials the subject spent the 2 min alone in the testing cage. After 2 min the male or female stimulus was removed from the room and the target female was added to the testing cage. The trial continued for another 5 min (Phase II). The subject’s head grab latency and whether he completed the mating was recorded. Males were allowed to mate a second time in Phase II if sufficiently motivated and capable. Eggs were collected and incubated from all target females where the subjects had completed mating. 4.2. Results Two subjects never initiated mating with the target female in Phase II, even in their control trials, and were dropped from the analysis, leaving 22 males as subjects. Due to an error, one male did not get an F trial. All other subjects initiated mating with the stimulus female in Phase I of the F trials. The procedure of selecting as stimuli for the M trials those males that had failed to copulate with females when prescreened did succeed in biasing a majority of the male–male interactions in Phase I. Although all but one stimulus male did respond and participate vigorously in those interactions, subjects performed more CCMs than stimuli did in 15 of the 22 trials (in 12 of those, stimuli performed no CCMs), stimuli performed more in five trials, and the two males performed equal numbers in two trials (p < 0.05, sign test). Cloacal contacts (CCs) occurred in nine of the male–male interactions (43%) and were performed by the subject in three, the stimulus in four, and both males in two. The percentages of M, F and C trials in which males initiated and completed mating in Phase II are shown in Fig. 4A. Subjects were less likely to initiate mating and less likely to complete mating in M trials than in their C trials (initiate mating: p = 0.008; complete mating: p = 0.027). The six subjects that received CCs were not less likely to mate than subjects that did not receive CCs (1/6 vs. 9/16 failed to mate, respectively). When subjects did initiate mating, they had longer head grab latencies in M trials than in their C trials (T = 1, N = 8, p = 0.016) (Fig. 4B). This overall pattern of results was seen regardless of whether trials in which the subject did not
76
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
Fig. 4. Experiment 2. (A) Percentage of M trials, F trials and C trials in which males initiated mating (head grabbed, clear bars) and completed mating (gray bars) with the target female. (B) Box plots of head grab latencies in M, F and C trials for those males that did initiate mating. Maximum possible is 300 s (the trial duration). M trials: males interacted with another male prior to mating with the target female; F trials: males mated with a female prior to mating with a second target female; C (control) trials: no prior interaction. *p < 0.05, **p < 0.01, compared to their C trials.
perform more CCMs than the stimulus in Phase I were excluded; the same measures remained significantly different except for head grab latency, where the N became very small because of the failures to initiate mating in M trials (initiate mating: p = 0.023; complete mating: p = 0.04; head grab latency: T = 0, N = 5, p = 0.063). Both subjects that received CCs and those that did not had longer head grab latencies in their M trials than in their C trials (subjects receiving CCs: T = 0, N = 6, p = 0.03; subjects not receiving CCs: T = −1, N = 10, p = 0.004). Subjects were also less likely to initiate and complete mating in F trials than in their C trials (initiate mating: p = 0.021; complete mating: p = 0.012) (Fig. 4A). If they initiated mating, they were slower to head grab in F trials than in their C trials (T = 2, N = 8, p = 0.023) (Fig. 4B). There were no significant differences between measures in M vs. F trials (p > 0.5 for all comparisons). For fertilization success, because of the substantial number of failures to initiate mating with target females in M and F trials, there were too few within-male comparisons between completed mating trials for statistical analysis. Therefore analyses were done by including all trials and scoring mating failure trials as 0 eggs fertilized. Overall, only 23% of the males fertilized any egg in their M trials, and only 9% in their F trials, whereas they fertilized at least one egg in 41% in their C trials. Overall, the number of eggs fertilized did not differ between their M and C trials (Fig. 5A) (both medians = 0, T = 19, N = 11, p = 0.24). If, however, only the 15 males that performed more CCMs are included (i.e., only those relevant to testing the hypothesis), then those males fertilized fewer eggs in their M trials than in their C trials (medians 0 and 1.5, T = 2.5, N = 7, p = 0.047). Across all subjects, fewer eggs were fertilized in their F trials than in their C trials (Fig. 5B) (both medians = 0, T = 4, N = 10, p = 0.014). None of the five subjects that performed a CC fertilized an egg in their M trials, but two of the six subjects that received a CC did. Only two other males in the experiment managed to fertilize any eggs in their M trials. More trials with a CC would be needed to tell whether performing a CC reduces fertilization success. In each trial type, some males mated twice (60% in M trials, 30% in F trials, and 53% in C trials). Only in C trials were there enough males that completed mating at all to compare fertilization outcomes between single and double matings. The outcomes tended to be greater for double matings but not significantly so (median numbers fertilized = 3 for double matings vs. 0 for single matings, U = 17, p = 0.071). Even with double matings, three of those males did not fertilize any eggs.
5. Experiment 3: Effect of receiving cloacal contacts from another male on mating and fertilization success 5.1. Methods Whereas Experiment 2 focused on the effect of performing more CCMs than the stimulus male, Experiment 3 focused on the effect of being the recipient of cloacal contact (CC) so that the effect of possible transfer of foam from the male performing the CC could be better determined. Experienced successful maters (N = 18) were chosen for the male–male interactions. The outcome of the trial (which male received the CCs) determined which male was the subject that then had the mating opportunity with the target female. All males had been used as subjects in Experiment 2, but over 5 months had elapsed since then and they were not assigned to the same male or female as before. For each trial, two males interacted for up to 2 min in the testing cage and CCs by either were recorded (Phase I). If there were no CCs, the trial was terminated. If either male received a CC, the other male was immediately removed and the target female (from a pool of 50 regularly laying females) was introduced (Phase II). The trial then continued for a maximum of 10 min or until the male completed mating, whichever occurred first. Phase II was 10 min instead of the 5 min in Experiment 2 in order to increase the likelihood that a completed mating would occur so that fertilization success could be determined. In order to generate enough trials in which one male was the recipient of a CC, five to nine of these male–male (M) trials were conducted each day for 5 days, pairing males in different combinations for a total of 37 trials. Each of the 18 males also had a control trial (C trial) on a day between male–male trial days. Thus males were used on more than 1 day but not more than once in a day. For control trials, the male was simply placed in the testing cage with a new target female. Eggs were collected and incubated from all target females where the subjects had completed mating. 5.2. Results In 19 of the 37 trials, neither male achieved a CC. In the other 18 trials (49% of the trials), ten different males were CC recipients. All recipients then successfully completed mating with the target female. Only six of those 10 males completed a mating in their C trials. While those outcomes are not significantly different (p = 0.13), it is clear that, as in Experiment 2, receiving a CC did not reduce mating success with a female.
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
77
Fig. 5. Experiment 2. Within-male comparisons of number of eggs fertilized in M vs. C trials (A) and F vs. C trials (B). For clarity lines connecting cases of zero eggs fertilized in both trial types are omitted (10 double zero cases for M vs. C and 11 double zero cases for F vs. C). M trials: males interacted with another male prior to mating with the target female; F trials: males mated with a first female prior to mating with a second target female; C (control) trials: no prior interaction. *p < 0.05 compared to their C trials. #p < 0.05 compared to their C trials, excluding cases where the subject did not perform more cloacal contact movements than the stimulus.
Fig. 6. Experiment 3. Within-male comparisons of number of eggs fertilized in M vs. C trials. M trials: subjects received a cloacal contact from another male prior to mating with the target female; C (control) trials: no prior interaction. Males fertilized similar numbers of eggs in their two trial types.
To compare fertilization outcomes in the two kinds of trials, incomplete mating trials were scored as 0 eggs fertilized, and for the three males that received a CC in more than one M trial, the mean number of eggs fertilized in those repeated trials was used in the analysis. Males produced at least one fertilized egg in similar percentages of their M and C trials (50% of each) and they did not fertilize a significantly different number of eggs in the two kinds of trials (Fig. 6) (medians 1 and 0.75, T = 11.5, N = 7, p = 0.69). 6. Discussion A prior interaction with another male did not increase mating or fertilization success. In addition, receiving cloacal contacts from another male (being “on the bottom”) did not reduce mating or fertilization success. Instead, a prior interaction with another male in which the subject performed more cloacal contact movements caused those subjects to be slower to initiate mating and
decreased the probability that they would complete a mating, leading to reduced fertilization success. These detrimental effects on reproductive success were similar to those that occurred when the prior interaction was with a stimulus female. Being “on top” in a sexual encounter interfered with the success of a subsequent mating attempt regardless of whether that sexual activity was with another male or with a female. This is consistent with the idea that the males “on top” with either sex are then usually too fatigued or sexually depleted to mate successfully. These results fail to support the hypothesis that males “on top” are dominants that go on to have increased reproductive success, with males “on the bottom” subordinates having decreased reproductive success. They even argue against the assumption that the male–male interactions are dominance encounters producing clearly dominant and subordinate individuals. Instead, to the extent that males could be classed as “on top” or “on the bottom” and either male did better reproductively, it seemed to be the males “on the bottom,” because they experienced little or no detrimental effect on subsequent mating and fertilization success. If these male–male interactions reflect any competitive mating tactic, the tactic suggested by the results is that males engage other males in performing CCMs and CCs in order to reduce the performers’ success with females (Jamieson and Craig, 1987; Birkhead and Møller, 1992). Further research would be required before knowing if this hypothesis has merit. CCs occurred in less than half of trials and instead both males spent most of their time trying to perform CCMs. Furthermore, a different hypothesis, that male–male sexual behavior is an incidental by-product of strong mating motivation, is fully consistent with the results of the experiments. That is, males may have been selected to mate rapidly, vigorously and fairly indiscriminately, which although costly if the other bird happens to be another male, would be sufficiently beneficial whenever the other bird is a female to have a net adaptive advantage. Reports of other avian species also question the assumption that being “on top” in male–male sexual encounters indicates dominant status or that there are consistent dominants and subordinates in such encounters (e.g., Bertran and Margalida, 2003). House finches (Carpodacus mexicanus) that were subordinate based on other behavior were the ones mounting the dominants (McGraw and Hill, 1999). Among lek-breeding ruffs (Philomachus pugnax), there are three different male behavioral and plumage phenotypes, one of which, faeder, is a female mimic. Faeders are mounted by the other male types but also sometimes mount them (Jukema
78
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79
and Piersma, 2006). A notable exception comes from a study of Razorbills (Alca torda), colonially breeding socially monogamous seabirds. Extra-pair copulation attempts by males are a common reproductive tactic, many males mount both sexes, and mountings of other males (but not being mounted by males) predicted successful extra-pair copulations with females (Wagner, 1996). A comparative analysis of male–male sexual behavior in birds concluded that higher frequencies of the behavior are associated with mating and parental care systems that produce selection for frequent mating opportunities (MacFarlane et al., 2010). The extrapair matings reported in the Japanese quail’s sibling species, C. coturnix, the frequent forced copulations in Japanese quail, and the remarkably short latencies to initiate mating also suggest that male–male sexual behavior might be a result of selection for rapidly and vigorously seizing any copulatory opportunity. Although the mounting males suffer a cost if mating immediately after a male–male encounter, there might be an overall benefit at other times. The success of the pre-selection of stimulus males in Experiment 2 means that males that are more vigorous at mating with females are also more vigorous at performing CCMs on other males. Studies with insects also suggest that high mating rate and frequent male–male sex may be genetically correlated traits (Burgevin et al., 2013). In Experiment 2 it was also interesting that the male stimuli pre-selected for failing to initiate mating with females nonetheless almost always responded to the male subjects’ copulatory attempts with attempts of their own. One male’s foam reduces the fertilizing success of another male when both have mated with a female (Finseth et al., 2013). Experiment 3 did not provide evidence, however, that receiving the foam of the male performing the cloacal contacts reduced the fertilizing ability of the recipient to any substantial extent. Although it is unlikely that the other male’s foam did not get mixed with the subject’s foam and sperm to exert any effect, that possibility cannot be ruled out, especially if females have any control over where inseminations are placed in their reproductive tracts, which is unknown in this species. Also, a greater number of trials would be needed to detect a small effect, especially given the likelihood that not all of the cloacal contacts resulted in insemination, just as in matings with females. In a similar vein, in Experiment 2 there were too few M trials with completed matings to tell whether males that performed a CC with the other male then fertilized fewer eggs (whether they wasted or reduced an ejaculate). What is clear is that those that vigorously engage in male–male sexual behavior are then much more likely to fail to mate at all, so that ejaculate amount is a secondary factor in fertilization success. In Experiment 1, there was a significant delay in initiating mating when the subject approached the stimulus male or female first instead of the target female. While this did not affect mating completion or insemination in the testing environment, in a less restricted environment such a delay would give the target female adequate time to leave and avoid the male’s mating attempt. Thus the results of Experiment 1, like those of Experiment 2, indicate a potential cost of male–male behavior. In addition, the observations in Experiment 1 of males that preferred to approach the stimulus male first and, in Experiment 2, of males that had shown no interest in mating with females in prescreening but engaged in sexual behavior with other males, together raise the possibility of a small percentage of males that consistently prefer other males to females for sexual behavior, as has been reported in sheep (Perkins and Fitzgerald, 1997). Additional research with choice tests of unscreened males would be needed to investigate this possibility. One less interesting explanation for the observed results of these experiments is that males simply mistake other males for females, depleting both their sperm and energy in the fruitless mating attempts. A combination of poor sex discrimination and high
mating rate has been hypothesized to account for male–male sexual behavior in both birds (especially those with sexually monomorphic plumage) and insects (Wagner, 1996; Burgevin et al., 2013). With greater experience, male Drosophila engage in less male–male sexual behavior (Bailey et al., 2013). Japanese quail plumage is only slightly sexually dimorphic. Indeed, evidence suggests that sexually inexperienced male Japanese quail learn to discriminate between females and males for mating through greater success with females (Nash and Domjan, 1991). The quail in the experiments reported here were sexually experienced, however, and in the absence of a female, even these experienced males readily engaged in sexual behavior with other males. The results of the Razorbill study also failed to support mistaken identity as a source of male–male mounting (Wagner, 1996). Thus, it seems unlikely that mistaken identity can account for all of these results. Male Japanese quail provide unique advantages for research on same-sex sexual behavior. Because of the high mating rate with both sexes and the ease of determining insemination and fertilization success, they allow experimental testing of hypotheses about same-sex behavior. The experiments reported here provide a new interpretation of their male–male behavior and both question the assumption that the behavior has anything to do with dominance and show that the male “on top” does not have a short-term advantage in reproductive success. Their reputation as “highly salacious” creatures is centuries old but many questions remain about the evolution and function of their mating habits. Acknowledgements Support came from NSF IBN-9514088 and IBN-0130986 during this work. The author thanks Michael Fairchild, Emiko MacKillop, and Andrew Mong for assistance with Experiment 1, Tim Van Deusen for animal care management, Findley Finseth for discussion of Experiment 3, Nicole Baran for comments on the manuscript and two reviewers for their helpful comments that improved the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.beproc.2014.09.027. References Adkins, E.K., 1974. Electrical recording of copulation in quail. Physiol. Behav. 13, 475–477. Adkins-Regan, E., 1995. Predictors of fertilization in the Japanese quail, Coturnix japonica. Anim. Behav. 50, 1405–1415. Adkins-Regan, E., 1996. Neuroanatomy of sexual behavior in the male Japanese quail from top to bottom. Poult. Avian Biol. Rev. 7, 193–204. Bailey, N.W., Zuk, M., 2009. Same-sex sexual behavior and evolution. Trends Ecol. Evol. 24, 439–446. Bailey, N.W., Hoskins, J.L., Green, J., Ritchie, M.G., 2013. Measuring same-sex sexual behavior: the influence of the male social environment. Anim. Behav. 86, 91–100. Ball, G.F., Balthazart, J., 2004. Hormonal regulation of brain circuits mediating male sexual behavior in birds. Physiol. Behav. 83, 329–346. Bertran, J., Margalida, A., 2003. Male–male mountings in polyandrous bearded vultures (Gypaetus barbatus): an unusual behaviour in raptors. J. Avian Biol. 34, 334–338. Birkhead, T.R., Fletcher, F., 1994. Sperm storage and the release of sperm from the sperm storage tubules in Japanese quail Coturnix japonica. Ibis 136, 101–104. Burgevin, L., Friberg, U., Maklakov, A.A., 2013. Intersexual correlation for same-sex sexual behavior in an insect. Anim. Behav. 85, 759–762. Birkhead, T.R., Møller, A.P., 1992. Sperm Competition in Birds: Evolutionary Causes and Consequences. Academic Press, London. Jukema, J., Piersma, T., 2006. Permanent female mimics in a lekking shorebird. Biol. Lett. 2, 161–164. Finseth, F.R., Iacovelli, S.R., Harrison, R.G., Adkins-Regan, E.K., 2013. A non-semen copulatory fluid influences the outcome of sperm competition in Japanese quail. J. Evol. Biol. 26, 1875–1889.
E. Adkins-Regan / Behavioural Processes 108 (2014) 71–79 Hirschenhauser, K., Wittek, M., Johnston, P., Möstl, E., 2008. Social context rather than behavioral output or winning modulates post-conflict testosterone responses in Japanese quail (Coturnix japonica). Physiol. Behav. 95, 457–463. Jamieson, I.G., Craig, J.L., 1987. Male–male and female–female courtship and copulation behavior in a communally breeding bird. Anim. Behav. 35, 1251–1253. Kuo, Z.Y., 1960. Studies on the basic factors in animal fighting: III. Hormonal factors affecting fighting in quails. J. Genet. Psychol. 96, 217–223. MacFarlane, G.R., Blomberg, S.P., Vasey, P.L., 2010. Homosexual behaviour in birds: frequency of expression is related to parental care disparity between the sexes. Anim. Behav. 80, 375–390. McGraw, K.J., Hill, G.E., 1999. Induced homosexual behaviour in male house finches (Carpodacus mexicanus): the “Prisoner Effect.”. Ethol. Ecol. Evol. 11, 197–201. Nash, S., Domjan, M., 1991. Learning to discriminate the sex of conspecifics in male Japanese quail (Coturnix coturnix japonica) – tests of biological constraints. J. Exp. Psychol.-Anim. Behav. Process. 17, 342–353. Nichols, C.R., 1991. A comparison of the reproductive and behavioral differences in feral and domestic Japanese quail. Master’s Thesis. University of British Columbia. Ophir, A.G., Galef, B.G., 2003. Female Japanese quail that ‘eavesdrop’ on fighting males prefer losers to winners. Anim. Behav. 66, 399–407. Perkins, A., Fitzgerald, J.A., 1997. Sexual orientation in domestic rams: some biological and social correlates. In: Ellis, L., Ebertz, L. (Eds.), Sexual Orientation: Toward Biological Understanding. Westport, Conn, Praeger, pp. 107–128. Poiani, A., 2010. Animal Homosexuality: A Biosocial Perspective. Cambridge University Press, Cambridge UK.
79
Ramenofsky, M., 1984. Agonistic behavior and endogenous plasma hormones in male Japanese quail. Anim. Behav. 32, 698–708. Riters, L.V., Balthazart, J., 1998. Behavioral evidence for individual recognition in Japanese quail. Behaviour 135, 551–578. Rodriguez-Teijeiro, J.D., Puigcerver, M., Gallego, S., Cordero, P.J., Parkin, D.T., 2003. Pair bonding and multiple paternity in the polygamous common quail Coturnix coturnix. Ethology 109, 291–302. Sardà-Palomera, F., Puigcerver, M., Vinyoles, D., Rodriguez-Teijeiro, J.D., 2011. Exploring male and female preferences, male body condition, and pair bonds in the evolution of male sexual aggregation: the case of the Common Quail (Coturnix coturnix). Can. J. Zool. 89, 325–333. Schein, M.W., Diamond, M., Carter, C.S., 1972. Sexual performance levels of male Japanese quail (Coturnix coturnix japonica). Anim. Behav. 20, 61–67. Sittmann, K., Abplanalp, H., 1965. Duration and recovery of fertility in Japanese quail (Coturnix coturnix japonica). Brit. Poult. Sci. 6, 245–250. Tsutsui, K., Ishii, S., 1981. Effects of sex steroids on aggressive behavior of adult male Japanese quail. Gen. Comp. Endocrinol. 44, 480–486. Vasey, P.L., Sommer, V., 2006. Homosexual behavior in animals: topics, hypotheses and research trajectories. In: Sommer, V., Vasey, P.L. (Eds.), Homosexual Behaviour in Animals: An Evolutionary Perspective. Cambridge University Press, Cambridge, UK, pp. 3–44. Wagner, R.H., 1996. Male–male mountings by a sexually monomorphic bird: mistaken identity or fighting tactic? J. Avian Biol. 27, 209–214. Ray, J., 1678. The Ornithology of Francis Willughby. Royal Society, London.
