Insectes Sociaux

, Volume 59, Issue 1, pp 61–65

Males, but not females, mate with multiple partners: a laboratory study of a primitively eusocial wasp Ropalidia marginata


  • M. C. Shilpa
    • Centre for Ecological SciencesIndian Institute of Science
    • Centre for Ecological SciencesIndian Institute of Science
    • Department of EntomologyPurdue University
  • S. Samudre
    • Centre for Ecological SciencesIndian Institute of Science
  • R. Gadagkar
    • Centre for Ecological SciencesIndian Institute of Science
    • Evolutionary and Organismal Biology UnitJawaharlal Nehru Centre for Advanced Scientific Research
Research Article

DOI: 10.1007/s00040-011-0188-4

Cite this article as:
Shilpa, M.C., Sen, R., Samudre, S. et al. Insect. Soc. (2012) 59: 61. doi:10.1007/s00040-011-0188-4


The intense interest in social Hymenoptera, on account of their elaborate sociality and the paradox of altruism, has often suffered from considerable gender imbalance. This is partly due to the fact that worker behaviour and altruism are restricted to the females and partly because males often live off the nest. Yet, understanding the males, especially in the context of mating biology is essential even for understanding the evolution of sociality. Mating patterns have a direct bearing on the levels of intra-colony genetic relatedness, which in turn, along with the associated costs and benefits of worker behaviour, are central to our understanding of the evolution of sociality. Although mating takes place away from the nest in natural colonies of the primitively eusocial wasp Ropalidia marginata, mating can be observed in the laboratory if a male and a female are placed in a transparent, aerated plastic container, and both wasps are in the range of 5–20 days of age. Here, we use this setup and show that males, but not females, mate serially with multiple partners. The multiple mating behaviour of the males is not surprising because in nature males have to mate with a number of females, only a few of whom will go on to lay eggs. The reluctance of R. marginata females to mate with multiple partners is consistent with the expectation of monogamy in primitively eusocial species with totipotent females, although the apparent discrepancy with a previous work with allozyme markers in natural colonies suggesting that females may sometimes mate with two or three different males remains to be resolved.


Social waspsRopalidia marginataMating behaviourMultiple matingSingle mating


Social hymenopteran females usually mate with one or more males early in their life and store the received sperm in their spermatheca. This sperm is used in all future fertilizations required for laying female-destined, diploid eggs. Females in this insect order also lay unfertilized haploid eggs, which develop parthenogenetically and give rise to males. In female social Hymenoptera, both the number of mating partners and the effective number of matings vary and are associated with a variety of costs and benefits. Single mating increases the relatedness among the colony members, which may sometimes promote social harmony (though not always, see Ratnieks et al., 2006), while multiple mating is thought to bring the advantages of increased supply of sperm, genetic diversity among the brood (resulting in genetic polyethism and increased immunity) and avoidance of the sex locus load, the costs of evolving specialized strategies for choosing or resisting mates, etc. (Strassmann, 2001; Crozier and Fjerdingstad, 2001). Female single mating is a common female reproductive strategy (effective paternity is often little over 1) for stingless bees (Peters et al., 1999), bumble bees (Schmid-Hempel and Schmid-Hempel, 2000), most ants and most vespine and polistine wasps (reviewed in Strassmann, 2001). However, multiple mating is abundantly found in honeybees (Estoup et al., 1995; Moritz et al., 1995; Oldroyd et al., 1997; Palmer and Oldroyd 2000), some fungus-growing ants (Bekkevold et al., 1999; Bekkevold et al., 1999; Murakami et al., 2000; Fjerdingstad and Boomsma, 1998; 2000), harvester ants (Cole and Wiernasz, 1999) and is occasionally found in vespine wasps (Muralidharan et al., 1986; Ross 1985, 1986). In all, high polyandry has evolved on nine occasions in the social Hymenoptera and about two-thirds of species are monandrous (Hughes et al., 2008a, b).

Males in social hymenopterans usually emerge from the pupa with mature sperm, and additional sperm production does not take place in the adult stage (Dumser, 1980; the exception: ergatoid males of the ant genus Cardiocondyla, Heinze and Hölldobler, 1993). Hymenopteran males are haploid (producing clonal sperm), short-lived and often referred to as sperm packets, as they eclose to perform the sole function of mating. Single mating associated with leaving a mating plug or spermatophore in the female genital tract is a usual phenomenon for many social hymenopteran males (reviewed in Boomsma et al., 2005), especially in stingless bees, honeybees and some ants. Multiple mating is found in males of bumble bees, social wasps and ants (reviewed in Boomsma et al., 2005).

In the primitively eusocial, polistine wasp Ropalidia marginata (Gadagkar, 2001), females either become workers in their natal nests or leave to found a new nest of their own. In the event of a loss of natural queen or experimental queen removal in an established nest, one of the workers becomes aggressive and eventually becomes the next queen of the colony (Gadagkar, 2001). Although many workers in a colony may mate, mating itself is not essential for females to obtain direct reproductive fitness; unmated females can develop their ovaries, suppress the ovarian development of other, even mated females and can even function as the sole egg layer of a colony, at least for a while (Chandrashekara and Gadagkar, 1991). R. marginata males, on the other hand, live in their natal nest for about a week and then leave to lead a nomadic life. Males appear to wait in the foraging areas and mate with the foraging females (Gadagkar, 2001).

Given the difficulty of studying mating behaviour in natural conditions, Sen et al. (2010) have recently described mating behaviour of R. marginata in a laboratory setup and showed that male and female wasps of 5–20 days of age readily mate in transparent, aerated, closed plastic containers. Using this setup, Sen et al. (2010) and Shilpa et al. (2010) have shown that for either sex, the probability and mate choice are not influenced by body size or nestmateship of the partner. Here, we used the same setup to ascertain whether, when given a chance, males and females would mate multiply (sequentially with multiple partners). We show that under this setup, males often mate multiply but females do not mate with any more than one partner.

Materials and methods

R. marginata nests with large number of pupae were collected in and around Bangalore (13°00′N 77°32′E), India. All adults, larvae and eggs were removed from the nests, and the nests with pupae were maintained in ventilated plastic boxes and monitored daily for the presence of newly eclosed wasps. All newly eclosed males and females were removed from the nest and kept in isolation in transparent, ventilated plastic bottles (22 cm × 11 cm × 11 cm) with ad libitum food and water for 5–20 days until they were used for the experiments.

To study multiple mating in males, we introduced one male and one female simultaneously into a fresh plastic bottle (as described above). The pair was observed for 1 h or till they performed long conjugation behaviour (Sen et al., 2010), whichever was earlier. Every occurrence and duration of all mating behaviours: unsuccessful attempt to mount—UAM, mount—MO (male climbs on to the back of the female), short conjugation—SC (up to 5 s long connection of abdominal tips of the male and the female in the mounted condition without any sperm transfer) and long conjugation—LC (locking of abdominal tips for more than 20 s that usually includes sperm transfer) were recorded (for a detailed description of each of these behaviours please see Sen et al., 2010). If the pair performed long conjugation behaviour, then the female was removed from the bottle immediately and a new virgin female was introduced into the bottle. Similar behavioural observation was carried out for the second female for 1 h or until the pair performed long conjugation behaviour. We continued to remove the previous female and introduce a new female until the male stopped performing the long conjugation behaviour. Behavioural observations were conducted for every pair. All females, who took part in long conjugation behaviour, were then dissected to check for the presence of sperm in the spermatheca. Multiple mating behaviour was studied for 19 males.

To study female multiple mating behaviour, we attempted to follow a similar procedure but females who performed the LC behaviour with one male did not take part in the same behaviour with another male in pilot studies. Therefore, we followed a different protocol to study if females remate (Fig. 1). We first introduced a virgin male and female into the plastic container. The pair was observed for 1 h or until the performance of the long conjugation behaviour, whichever was earlier. Once a female performed the LC behaviour with a male, we immediately removed the male and introduced the second male (this time was considered as 0 h). The second male was removed after 1 h and a third male was introduced after 2 h from the time of the LC with the first male. After an hour with the third male, the female was kept in isolation for 1 h and then the fourth male was introduced for 1 h. The fifth male was introduced after 24 h of the first LC behaviour for 1 h. All males were virgins. Behavioural observations were carried out with every male, and every occurrence and duration of the mating behaviours (UAM, MO, SC and LC) were recorded. This protocol was tested for 20 females who showed LC behaviour with the first male.
Fig. 1

Protocol to test multiple mating in females. The female and the first male were introduced simultaneously. Immediately after long conjugation (LC) behaviour, the first male was removed and the second male was introduced. The time of LC with the first male was considered as 0 h, and introducing and removing time of subsequent males were adjusted accordingly. Timing of introduction of different males is presented at the top, while the lower panel shows the removal time of the males

To check if the females needed a longer time interval between two matings, we used an additional set of 20 females who showed LC behaviour with the first male. For this set, we paired the females with a new virgin male for 1 h after every 24 h for 4 more days. Thus, every female was paired with five males with a 24-h time interval. Behavioural observations were carried out for each pair for every occurrence and duration of the mating behaviours (UAM, MO, SC and LC).

To test if multiple mating occurs between the same pair, we left another set of 20 pairs of wasps in the mating container for 1 h after the first performance of long conjugation.


Males mate multiply

Most males in our study were seen to mate multiply, i.e., sequentially with multiple partners. Out of 19 males, 4 mated (successfully transferred sperm) with 4 different females each, 5 males mated with 3 females each, 7 males mated twice and 3 males mated singly; 10 males did not mate at all (Fig. 2a). Considering only the males that mated with at least one female, about 84% mated with at least two females, about 60% with at least three females and about 50% mated with at least four females (Fig. 2b). The durations of mount and long conjugation behaviour were not significantly different across different partners in the mating sequences (Mann–Whitney U test, all P values were more than 0.1; Fig. 3a, b). However, latency to mount (time difference between introducing a pair and start of first mount behaviour) and latency to long conjugation behaviour were significantly longer for the first partner compared to all subsequent partners (Mann–Whitney U test, all P values <0.008, α set as 0.008 after Bonferroni correction due to six comparisons; Fig. 3c, d). The latency for either behaviour was not significantly different between the second, third or fourth pair (Mann–Whitney U test, all P values >0.1; Fig. 3c, d).
Fig. 2

Male multiple mating. a Number of males performing long conjugation (LC) behaviour with consecutive females. b Probability of successful mating (sperm transfer) at subsequent mating sequences. Bars carrying same letters are not significantly different from each other (Fisher’s exact test, P > 0.05)
Fig. 3

Male multiple mating: durations of a mount, b long conjugation, c latency to mount and d latency to long conjugation across different female partners in the mating sequence. Bars carrying different letters are significantly different from each other (Mann–Whitney U test at P < 0.008)

Females did not mate with multiple partners

We did not find any occurrence of serial mating (with different partners) by the females (n = 20 for hourly interval; n = 20 for daily interval). Some females took part in unsuccessful attempt to mount, mount and short conjugation with the subsequent males (Fig. 4), but long conjugation behaviour did not take place. All females took part in MO and SC behaviours with the first male (note, only those females who took part in LC with the first male were chosen for this analysis), but only a few participated in these behaviours with the subsequent males. The proportion of females who took part in UAM, MO and SC behaviours did not differ between the second and fifth males (Fisher’s exact test, all P values were more than 0.005, α set as 0.005 after Bonferroni correction due to 10 comparisons).
Fig. 4

Proportion of R. marginata females taking part in an unsuccessful attempt to mount (UAM), mount (MO) and short conjugation (SC) behaviour with successive males after a hourly and b daily intervals. All females performed long conjugation (LC) behaviour with only the first males

Repeated mating between the same pair is rare

If the male was not removed immediately after the first LC, then 5-day-old females occasionally performed LC with the same male again (unpublished pilot observations). But in this study, none of the pairs performed LC more than once. None of the 20 tested pairs performed any mating related behaviour after the first LC within the 1 h after the first LC.


Mating is neither necessary nor sufficient for female wasps to obtain direct reproductive fitness in R. marginata (Chandrashekara and Gadagkar, 1991). Unlike many other social hymenopteran species, where only reproductives (or gynes) mate, in R. marginata many females mate but most mated females seldom get an opportunity to lay eggs. It is therefore not surprising that we found males to mate with multiple females in this study, because mating with a single female does not assure fitness. It should be emphasized that in this study, we not only recorded mating behaviour (long conjugation) but we also confirmed successful sperm transfer by the presence of sperm in the spermathecae of the females. Thus, males appear to ejaculate only a part of their sperm load into each female. By transferring sperm to many females, R. marginata can maximize their chances of siring offspring. Indeed, male R. marginata face a serious conundrum—most females they mate with will never become queens and lay female-destined eggs, but the males have no way of telling which females will do so. Distributing their sperm across many females appears to be their best bet.

The observed reluctance of females to mate with multiple males is consistent with monoandry seen in most other polistine wasps including Ropalidia revolutionalis (Henshaw and Crozier, 2004; Hughes et al., 2008). However, a previous study using allozyme markers to genotype naturally mated mothers and their daughters had shown that R. marginata females may sometimes mate with two or three males in nature (Muralidharan et al., 1986). It is possible that the previous study was inaccurate because of the relative low power of allozyme markers. On the other hand, the present study may be inaccurate because of differences between what the females might do under laboratory conditions where they are presented with a series of males in quick succession in the same environment and what they might do in nature where they have access to multiple males during different foraging trips, which may be weeks apart and hundreds of metres apart. Hence, we would like to reserve judgement on the exact situation with regard to single/multiple mating in R. marginata, pending a more powerful study with microsatellite markers, which we now have initiated.


This study was supported by the Department of Science and Technology, Department of Biotechnology and Ministry of Environment and Forests, Government of India. RS was supported by a senior research fellowship of the Council of Scientific and Industrial Research. RS and RG designed the study; MCS and SS made the quantitative behavioural observations; RS dissected the female wasps; RS and MCS analysed the data; RG supervised the study; RS, MCS and RG co-wrote the paper. We thank three anonymous referees for useful comments that helped improve this paper.

Copyright information

© International Union for the Study of Social Insects (IUSSI) 2011