Two males, one female: triplet-style mating behaviour in the Darwin wasp Xorides ater (Gravenhorst, 1829) (Hymenoptera, Ichneumonidae, Xoridinae) in the Swiss Alps

The Darwin wasp Xorides ater is a solitary idiobiont ectoparasitoid of wood-boring beetle larvae. It occurs across the Palaearctic, and in Switzerland seems to be restricted to areas above 980 m altitude. Very little is known about its biology and behaviour. We here report three separate instances of a female X. ater mating simultaneously with two males, observed on the same occasion on a Picea abies woodpile near Lenk, Switzerland. Only two copulae that involved one male and one female were observed at the same time, which might indicate that this mating mode is rather common in X. ater. We illustrate our observations with photo and video material and compare it to mating behaviour known from other species of Darwin wasps.


Introduction
Hymenoptera have a haplo-diploid reproductive mode (Crozier 1975) and, although thelytoky is not uncommon in the order, most species reproduce sexually through union between a haploid male and a diploid female. The latter then lays eggs that are either haploid (male-producing) or diploid (female-producing) through her ability to control access of sperm to the egg as it passes down the oviduct and past the spermatheca (Gerber and Klostermeyer 1970;Cole 1981;Strand 1989). Female hymenopterans are thus able to produce viable offspring without having to mate, although the offspring then would only be of the male sex. This mode of sex determination, known also as arrhenotokous parthenogenesis, is believed to have had a great bearing on the success of the order (Gauld and Bolton 1988;Heimpel and De Boer 2008). It is especially through the ability of mated females to control the sex of the eggs they lay that the ensuing complex behavioural and evolutionary ecology of parasitoid Hymenoptera has been so well-studied, albeit in a very limited number of species (Godfray 1994;Wajnberg et al. 2008).
Detailed courtship studies have been made only in a minute fraction of Darwin wasp species, but it is usual for female parasitoid Hymenoptera to be receptive to mating for only a short time following their emergence (Quicke 1997). Courtship is typically rather brief but can involve elaborate cues. It is usually initiated by volatile pheromones emanating from females (Quicke 1997;McClure et al. 2007) and may then involve wing-fanning "songs" (Bredlau and Kester 2019), male-male struggles over access to females (Eggleton 1991), or the transfer of contact pheromones from the antennae of the male to those of the female (Steiner et al. 2010).
It is by no means unusual for several males of a range of parasitoid wasps to simultaneously court a female, and aggregations of competing males are commonly observed, especially in parasitoids of wood-boring insect larvae such as Pimplinae (Theronia maculosa Kreiger, 1906, M. Giannota observations) and Rhyssinae (Eggleton 1991). However, to the best of our knowledge, only a single male has ever been observed to achieve copulation at any one time, even though females are sometimes receptive to repeated matings with the same or a different male for a short time afterwards (Quicke 2015;M. R. Shaw observations). With the exception of a very brief mention of Rhyssa persuasoria Linnaeus, 1758 being observed with two males apparently attached, though this was under conditions of disturbance whereby a large courting aggregation was being deliberately dismantled (Baker 1992), we have not encountered reports of more than one male achieving simultaneous copulation (whether ejaculation or not) with a single female in any species of Ichneumonoidea. We have never observed it ourselves during many decades of extensive rearing and field observations and, on that basis, we presume that it is highly unusual for it to occur.
It was therefore a surprise to observe three instances of a female of the xoridine ichneumonid Xorides ater Gravenhorst, 1829 in copula with two males, and apparently comfortably so for a prolonged time. During the same observation period, we observed only two standard pairings in the same location. In this brief report, we describe and illustrate these observations.

Study species
Xorides ater (Fig. 1) is a member of the putative sister group to all other Darwin wasps, the subfamily Xoridinae (Klopfstein et al. 2019;Bennett et al. 2019). By analogy with congeners X. brachylabis Kriechbaumer, 1889and X. irrigator Fabricius, 1793(Chrystal and Skinner 1931, and X. corcyrensis Kriechbaumer, 1894 (Sharifi and Javadi 1971), X. ater can safely be presumed to be an idiobiont parasitoid (i.e., one that kills or permanently paralyses its host at the time of oviposition) whose solitary larva develops on the body of its beetle larva, prepupa or pupa host. The parasitoid would pass the winter in its cocoon; however, neither the host range nor the developmental biology of X. ater is properly understood, nor is there any report of its courtship and mating behaviour.

Circumstances of observations
During a period of general collecting of Ichneumonoidea near Lenk in the Swiss Canton of Bern, we came across an exposed woodpile of cut and partly split spruce logs (Picea abies), originally up to about 20 cm diameter before being split. The woodpile was located just below Läheweid, about 600 m NW of the cable car station Bettelberg, Stoos (46°26'40"N, 7°25'31"E) at about 1510 m altitude. The observations took place on 3.vii.2019 in the early afternoon from 13.00 hrs to 15.00 hrs of a humid and intermittently sunny day. The dimensions of the woodpile, which had a corrugated iron cover, was approximately 10 × 2 × 1 m and ran length-ways from West to East with one side in shade (at 23 °C) and with the majority of insect activity observed on the shaded side. From the insect species seen on the woodpile and their activity, we deduced that it had been present for at least two years.
Photographs and short video recordings were taken with an Olympus EM-5 Mark II with a 60 mm macro-Olympus lens and later viewed and processed in Adobe Photoshop 2020 and Adobe Illustrator 2020. The composite video was edited in iMovie 10.2.3.

Activity and mating observations
Arriving at the woodpile, we immediately observed a very high level of Ichneumonoidea activity, at one point counting more than 40 males in flight in front of the shaded side of the woodpile. Not only males but also females were present, and closer examination revealed that most of the individuals belonged to the ichneumonid X. ater ( Fig. 1) with fewer of the similarly sized (but koinobiont endoparasitoid) braconid Helcon nunciator Fabricius, 1793. At least some of the females of X. ater present were clearly freshly emerged, as matings were occurring, though there were also some females resting without being courted that may or may not have already mated. A few males of X. ater were observed at rest but more often they were flying close to the woodpile and apparently in search of females, although without evident aggregation. Although some adult parasitoids were seen on the sunny (southern) side of the woodpile, insects there were too highly active to be easily observed or collected, and most of our observations were made on the shaded (northern) side, where insects were more abundant and easier to observe.
In the braconid H. nunciator, we observed about three matings, each between one male and one female. In X. ater, on the other hand, over a period of about two hours we saw three instances of a female X. ater with two males attached in apparent mating mode ( Fig. 2A, B; Suppl. material 1: Movie 1). Only two copulations involving a single individual of each sex by this species were observed during the same period (Fig. 2C, D; Suppl. material 1: Movie 1). Unfortunately, no courtships or commencements of copulation were seen, so whether the attachment of the two males in the triplets happens simultaneously or sequentially is unknown.
In addition to the fact that the triplet-style matings appeared to happen rather easily and possibly even  regularly, several things struck us. In the first place, the union seemed to be satisfying all three participants as, overall, the assemblages were as quiescent as the two normal copulae seen at the same time and appeared to persist for an equivalent period (at least 5-10 minutes, with separation occurring only as we collected them). Second, there was no evident jostling or attempt by one male to dislodge the other; indeed, they seemed extremely relaxed, with legs twitching and often held outstretched and clear of everything (usually at least one hind leg touched the substrate, but sometimes all legs were completely free). The females, too, seemed unperturbed and made no attempt to dislodge either male. Third, the positioning of the genitalia appeared to be normal, although the precise mechanism of attachment of males to the female's sixth sternite (hypopygium) in the triplets could not be investigated in any detail. However, the single paramere (= harpe, sensu Schulmeister 2001) that was visible appeared to be normally deployed (Fig. 2B) and was presumably attached to the female's sixth sternite by means of the usual pincer action of the cuspis and digitus of the volsella. The other side of the male external genitalia was not visible, and how it was deployed or indeed whether insemination would be possible, could not be deduced.

Ecological observations
In addition to copulation behaviour, some information on the ecology of Xorides ater can be given here. First, we were interested in the altitudinal distribution of the species in Switzerland and thus obtained data from museum collections (Table 1). All 71 specimens found were collected above 980 m altitude, and the narrow flight time from late June to early August indicates that X. ater is univoltine. Furthermore, three out of the 26 collection events included at least a dozen males, while females were always only collected one or two at a time, adding support to our observation of high male abundance (Table 1).
The few and mostly unconfirmed host records that can be found for X. ater in the literature all include cerambycid beetles in conifers (Hilszczański 2003;Sheng and Wen 2008). The only suitably sized adult beetles seen at our site were the three species of Callidium (Coleoptera: Cerambycidae) that occur in Switzerland: Callidium aeneum De Geer, 1775, C. coriaceum Paykull, 1800, and C. violaceum Linnaeus, 1758. However, there was no direct evidence that any of these was serving as host.
While the females of X. ater seen were rather uniform in size, males were substantially more variable in that respect, suggesting that hosts of different sizes had been accepted, but whether of the same or different host species is unknown.

Discussion
Copulatory position in X. ater and other Hymenoptera In the majority of Hymenoptera, the male genitalia are orthandrous, resulting in a mating stance whereby the male is mounted on the top of the female, the same way up and with head-to-head orientation and the apex of his metasoma curled below that of the female (Schulmeister 2001). The major exception is the Tenthredinoidea and Xyelinae, in which the male genitalia are strophandrous (twisted through 180°), resulting in a coplanar end-to-end stance with the two sexes looking in opposite directions (or positioned co-ventrally), as is typically seen in Lepidoptera. The database Taxapad (Yu et al. 2016) gives references to some 333 papers in which courtship and/ or mating in Ichneumonoidea is at least mentioned, but it is beyond the scope of this work to attempt a detailed review of these, and we restrict ourselves to an outline of the most usual copulatory stances seen, and a more detailed comparison between species of Xorides. In most Ichneumonoidea, especially when the sexes are of comparable size so that the heads can be more or less adjacent, the male is usually mounted on top of the female and holds her with his legs (Fig. 3A). In this position, some degree of antennal contact during mating is usual. Indeed, the antennae of males are furnished with tyloids in species of approximately ten subfamilies of Darwin wasps (Quicke 2015), which appear to play a role in antennal communication during courtship and mating (e.g., Bin et al. 1999;Bordera and Hernandez-Rodriguez 2003;Steiner et al. 2010); an extreme example of this is seen in the diplazontine Syrphoctonus tarsatorius Panzer, 1809 ( Fig. 3B; Steiner et al. 2010). In other cases, the male may slip back, foregoing antennal contact ( Fig. 3C; also see Benelli et al. 2019 for the campoplgeine Campolex capitator Aubert, 1960). When the sexes are dissimilar in size, the male is usually the smaller and, because he is necessarily further back, antennal contact is often literally out of reach during copulation itself. In such cases, the male might start by climbing onto the female's back to perform antennal courtship, then sliding back as the female signals receptiveness, as we have seen in the rhyssine Rhyssa persuasoria ( Fig. 3D; S. Klopfstein observations).
Once genital attachment is achieved, it seems not unusual for males, particularly in species with a large sexual size difference, to fall further back and flip through 180 ° so that the orientation changes to one with the heads looking in opposite directions, and the male on his back with respect to the female (Fig. 2C). In such cases, there is usually little advantage in the male using many of his legs to hold onto the substrate, and they are sometimes all left free, with only his genitalia attached to anything (Suppl. material 1: Movie 1). This orientation of the male's body was seen in both of the normal copulae (Fig. 2C) and all three triplets ( Fig. 2A). It has been reported also in the tryphonine subgenus Netelia (Netelia) Gray, 1860 (Vance 1927) although, in another species of the same subgenus, Shaw (2001) recorded a more normal maleon-top stance, suggesting some degree of plasticity or variation in the group. Recently, we have seen a similar stance to that encountered in X. ater in an unidentified species of the ophionine genus Enicospilus in Australia (Fig. 3E), though in this case the male did twist round to grasp the substrate rather than hang free. Mating behaviour has been described for two other species of Xorides, in both cases from captive material in the laboratory. Chrystal and Skinner (1931) observed several pairings of X. brachylabis, noting that in courtship the male stroked the female's antennae with his own and, having approached slowly, mounted the female and grasped her with all six legs while stroking her thorax with his antennae throughout the copulation, which lasted from 15 to 40 minutes. This head-to-head alignment with the male mounted on top corresponds to our Fig. 3A. The mating behaviour of a single pair of Xorides corcyrensis described by Sharifi and Javadi (1971) differs in several respects but the male, although perhaps a little further back, was also essentially mounted head-to-head on the female. During copulation the mounted male placed his fore and mid legs on the female's mesoscutum and scutellum, with the fore legs sometimes suspended in the air.
Although unfortunately we did not witness the initiation of copulation, the position finally arrived at by X. ater in the copulations we observed, whether involving one or two males, differed markedly from these literature descriptions of union in other Xorides species. As in the case of Netelia mentioned above, this indicates that uniformity within a genus should not necessarily be expected. It is possible that the copulatory position of X. ater may be one factor facilitating matings involving two males, as more space around the female's external genitalia is exposed than in a male-on-top stance. However, additional mating observations under natural circumstances are needed to assess whether triplet-style matings are common in the genus or whether they are restricted to X. ater.
Are triplet-style matings of adaptive significance or just a side-effect of scramble competition?
Although males certainly predominated, we were able to find several females of X. ater on the woodpile that were not mating and that did not seem to attract much attention at all from the numerous males. Although they were not necessarily receptive, some females did appear freshly emerged, judging from the distended membranous separation of metasomal tergites from sternites and from particles of wood dust on them not yet cleaned away. There is thus no reason to believe that the triplet-style mating was a unique result of particularly high male density and a shortage of females, and we did not observe anything resembling scramble competition among the many males present.
Whether or not insemination was taking place in the triplets and, if so, whether it was from one or both of the males, are key questions that unfortunately we are not able to resolve at present. There are, however, several interesting considerations pertaining to this that can be briefly summarised here. If sperm transfer from both males takes place, is it a fitness strategy of the female to permit or even encourage simultaneous mating by two males, perhaps thereby ensuring a higher genetic mix for her progeny, or offsetting her risk of mating with an infertile male? By decreasing mating time while simultaneously maximising overall sperm transfer, a triplet-style mating strategy may lower the cost (for example the increased risk of predation, or simply the time investment) that would be associated with mating successively with more than one male. From the male side, is it just an unintended consequence of the scramble competition between males frequently observed in parasitoids of wood-inhabiting insects (e.g. Eggleton 1991), or is there a trade-off in avoiding conflict? Is there more than just passive sperm competition? Or, if insemination does not occur (in spite of the female's presumed belief and subsequent behaviour), and if triplet-style sterile matings are common, does it constitute a fitness impediment carried by X. ater by producing a higher proportion of males than would be optimal?
It would be interesting to know of additional observations of similar behaviour, whether in Xoridinae or in other ichneumonoid or hymenopteran groups. Further work is needed to elucidate whether or not sperm transfer from one or both males was successfully achieved in the triplets that evidently occur in X. ater at least on occasion, and possibly regularly, and any morphological modifications that might facilitate it. Supplementary material 2

Figure S1
Authors: Mark R. Shaw, Madalene Giannotta, Andrés F. Herrera-Flórez, Seraina Klopfstein Data type: Raw Images Explanation note: Two raw images used to generate the stacked composite image, Fig. 2A, in the manuscript. Copyright notice: This dataset is made available under the Open Database License (http://opendatacommons. org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited. Link: https://doi.org/10.3897/alpento.5.64803.suppl2