1 HYBRIDISATIONINLADYBIRDS 1 INTERSPECIFIC HYBRIDISATION IN LADYBIRDS (COL.: COCCINELLIDAE) Michael E.N. Majerus DepartmentofGenetics, UniversityofCambridge,DowningStreet,CambridgeCB23EH. Introduction THE MOST COMMONLY used definition of a biological species is based upon effective reproduction. Members of a sexually reproducing species do, or potentially could, inter-breed amongst themselves, but not with members of other species (Mayr, 1963). This definition implies not only that matings occur, but that viable and fertile offspring are produced. Matings between individuals ofputatively different species have long interested biologists for their potential in helping to unravel the evolutionary mechanisms of speciation, the nature of barriers to mating and so gene-flow between populations, and their possible role in the origin ofnew species. Matings between members ofdifferent species are a waste ofreproductive resources for both partners. The waste for females is generally greater than for males, because females produce large, heavily resourced gametes, while males produce small gametes containing little more than a nucleus and a "tail". However, the cost to males is not negligible. Not only does a male that mates with a female ofanother species waste gametes, he wastes energy in mating, he wastes time that might be used in more productive ways such as feeding, resting or courting more appropriate females, and, as he is less mobile while copulating, he may be more prone to predation. He also runs the risk ofcontracting disease from his mating partner. Ofcourse females are liable to these latter costs as well. Ladybirds that mate with non-conspecific partners suffer all these costs. Mating in Adalia bipunctata, a highly promiscuous species, generally lasts for several hours. Both sexes expend considerable energy while mating. The female almost invariably shows rejection behaviour when mounted by a male, stretching up with her back legs, kicking back at the male's genitalia and running while swaying her abdomen from side to side. This rejection behaviour may last for many minutes before the female accepts the male. The male has to cling on while the female is trying to dislodge him. Thereafter, he is energetic throughout the copulation, twisting in a corkscrew motion during the first phase ofmating, and rocking rapidly from side to side at regular intervals later (Ransford & Majerus in prep.). Mating ladybirds are less mobile than those that are not mating, and, in Coccinella 7-punctata at least, are consequently more prone to parasitisation by the bracionid wasp Dinocampus coccinellae (Majerus, unpublished data). Several species of coccinellid {A. bipunctata, A. 10-punctata, C. 11- punctata, Hippodamia convergens, Tytthaspis 16-punctata, and two African species, Exochomusfulvimanus and E. concaius) are also known to bear a 12 ENTOMOLOGIST'S RECORD, VOL. 109 25.1.1997 mite, Coccipolipus hippodamiae This mite is transmitted from one ladybird . to another during mating and has a detrimental effect on host fecundity and fertility (McDaniel & Morrill, 1969; Husband, 1984; Majerus, 1994; Hurst et al, 1995; Webberley etal, in prep.). Given these potential costs to mating with a member of another species, selection should promote mechanisms of species recognition in animals. That this is the case is supported by the existence of a huge array of such mechanisms and by the fact that interspecific matings are very rare in animals. However, they do occur occasionally. This paper contains details of a number of interspecific matings involving ladybirds, and discusses the reasons for the occurrence ofthis apparently ineptbehaviour. Observations ofhybrid matings The records of interspecific matings contained herein comprise three categories defined by the situation of the mating ladybirds. Some observations are of ladybirds mating in the field. Others are of matings between individuals that had recently been collected and had been placed together in the same container. The third category involves ladybirds that were maintained in the laboratory, under specific conditions, for some period before being placed together specifically to determine whether they would copulate. Fieldobservations Details of 18 field observations ofinterspecific matings are given in Table 1. Most of the pairs concerned were collected. These were kept in captivity to determine the outcome of the copulation. The ladybirds were kept individually in Petri-dishes, and fed daily on pea aphids Acyrthosiphon pisum, black bean aphids Aphisfahae or oyster scale Pseudochermesfraxini depending on the food preference of the female involved. The dishes containing females were examined for eggs daily. If eggs had been laid, the female was removed to a clean dish. Eggs were counted and monitored for signs ofembryonic development. Small pea aphids were added to any dishes containing eggs that showed signs of development. If the eggs hatched, larvae were reared on pea aphids, using techniques previously described (Majerus et al, 1989). The external morphology of fourth instar larvae and pupae was examined to compare with that of the parental species. Any progeny that reached adulthood were examined in detail. The sex, pronotal and elytral colour patterns and other external morphological features were noted while these ladybirds were alive. Attempts were then made to back- cross them to an opposite sex individual of the parental species that they most resembled. In the case of progeny that resembled C. 7-punctata, adults were overwintered before back-crossing, as most British individuals of this species require a dormant period before becoming reproductively mature. Eggs from any successful matings were collected and their hatch rates . HYBRIDISATIONINLADYBIRDS 13 recorded as an indication offertility level. Once the ladybirds had died, their genitalia were examined to check sex and look for abnormalities that might be indicative ofan interspecific hybrid origin. Brief details of the outcomes of these treatments for each of the pairs collected are given in Table 1 Table 1. Records ofhybrid matings observed in the field. + Indicatesthatcopulationwasalreadyinprogresswhenthepairwas firstseen. * Suspectedthattheprogenyweretheresultofthefemalehavingpreviouslymatedwithamaleof herownspecies. 14 ENTOMOLOGIST'S RECORD, VOL. 109 25.1.1997 Obsen'ations ofrecently-collected ladybirds Observations of interspecific matings between recently-collected individuals involve ladybirds that were either put into the same container in the field, or ladybirds that were sent in to the Cambridge Ladybird Survey and were placed in the same container, when they were unpacked, before sorting. In all cases, the matings involved took place less than 72 hours after the ladybirds were collected from the wild. The ladybirds involved were treated in the same way as those collected having been observed mating in the wild. Details ofthese pairings appear in Table 2. Table 2. Records ofhybrid matings between recentlycollectedcoccinellids 1. A female C. 5-punctata and a male C. 11-punctata sent in the same box from the Spey Valley, Scodand, in June 1987 by Patricia Duncan, mated soon after arrival. They subsequently remated on three occasions. The female C. 5-punctata produced eggs, however, all the progeny were normal C. 5-punctata, the ensuing stockbeing maintainedfortwo more generations withoutproducing any abnormal descendants. 2. Male E. 4-pustulatus repeatedly attempted to mate with both a C. renipustulatus (sex not determined) and a male A. bipunctata. Mating attempts were apparently unsuccessful (Davies,pe/-^. comm.). 3. Male C. bipustulatus and female E. 4-pustulatus collected Lakenheath, May 1987, mated within a few minutes ofcollection and stayed in copula for over two hours. No eggsresulted. 4. Male C. 7-punctata placed in container with a number of C. 11-punctata collected earlier the same day, attempted to mate with several ofthem, and then apparently succeeded in copulating with one, remaining in copula for approximately 45 minutes. Some 73 eggs were produced, but all progeny were normal C. 11- punctata. 5. A female A. bipunctata and a male A. 10-punctata collected from overwintering sites in Cambridge, 11 December 1991, mated the following day while still in collecting box. The pair copulated for 75+ minutes. No eggs resulted for 11 days. Thereafter, afew infertile eggs were laid. 6. MaleA. obliterata mated with afemale C. 7-punctata inthe laboratory, for over an hour, shortly afterbeingcollectedfromthe wild. Nofertileeggsresulted. 7. A sample of T. 16-punctata, swept from grass with a few C. 7-punctata, at King's Forest, Suffolk, in June 1994, were placed in the same box. A male T. 16-punctata mated with a female C. 7-punctata shortly afterwards. The pair were isolated to allow close scrutiny. The pair remained in copula for 71 minutes, the female did not appearto reject the male. 8. A male C. magnifica and a female C. 7-punctata collected together in Germany by Hinrich Schulenberg, mated. The pair was sent to John Sloggett in Cambridge. No eggs resulted (Sloggett,per^. comm.). 5 HYBRIDISATIONINLADYBIRDS 1 Laboratoij obsei'vations It is known that isolating A. 10-punctata or A. bipunctata adults from opposite sex conspecific individuals, for a period of two or more weeks, while keeping them active (temperature 12-28°C) and well-fed on an appropriate aphid diet, will lead these individuals to mate with partners of the other species more readily than is normal (Lusis, unpublished data communicated by LA. Zakharov; Ireland et al., 1986; Majerus & Keams, 1989; O'Donald & Majerus, 1993; Majerus, 1994). It was considered possible that this procedure might induce other coccinellid species to copulate. Therefore, at various times over the last ten years, when culturing two or more species of coccinellid, the chance to test this possibility has arisen. In these cases, known virgin females of one species that were in reproductive condition and had been fed on an excess of a principal food {sensu Hodek, 1973) for at least two weeks, were offered males ofa different species in Petri-dishes, a single pair being placed in each dish. The ladybirds were kept under observation and their behaviour recorded. Copulations that occurred were timed. Pairs were separated after parting, and the females were treated as described above. The details of pairs of species tested and the results obtained are given in Table 3. Results and discussion Matings observed in the wild Eighteen interspecific hybrid pairs were recorded in the wild. Of these, 15 were between congeneric species. Ofthe others, two were between members of the sister genera Exochomus and Chilocorus. Only one pair involves species that would not be considered closely related, that being the mating betweenE. 4-pustulatus and C. 7-punctata. The initial mounting of the female by the male was only observed in five of the matings. It was notable that in none of these cases did females show strong initial rejection ofthe male, and insertion ofthe sipho occurred within two minutes. In the mating between E. 4-pustulatus and C. 7-punctata, the female began to show vigorous rejection behaviour (kicking back with her hind legs) some 16 minutes after being mounted. The male disengaged after about a minute. The lack of initial rejection of males by females suggests that the females may have been reproductively mature, and ready to oviposit, but that they had been unmated for some period. Females (whether virgin or not) that have been kept isolated from males for a significant period show less rejection behaviour than those that have mated in the last two or three days (Ireland et al, 1986). In the case ofthe female C. 7-punctata that began to reject the E. 4-pustulatus male after initially appearing to accept him, her response may have been triggeredby some genitalial incompatibility. 16 ENTOMOLOGIST'S RECORD, VOL. 109 25.L1997 Table3. Results ofattempts to induce interspecific hybrid matingsbetween coccinellids byisolatingfemalesfortwoweeksor more before offeringamaleofa differentspecies Male 7 HYBRIDISATIONINLADYBIRDS 1 development, indicating that zygotes perished very early in embryogenesis. Conversely, some of the eggs from the two matings between E. 4-pustulatus and C. bipustulatus did begin to develop, turning grey after three days. The deduction in the case of these two pairs is that the eggs produced were the product of the observed hybrid matings, but that the genetic constitution of the species is sufficiently divergent that the zygotes produced suffer developmental breakdown. The possibility that the zygotes failed to hatch because they became cramped in the eggs before they were ready to hatch, as is known to occur in some hybrids between different species ofhawkmoth (Newman, 1965) was considered. However, this does not seem probable given that the two pairs were reciprocal, unless the hybrid zygotes were larger than those ofeitherparent. The remaining eight pairs produced eggs that did hatch. The egg hatch rates of seven of these were high (>70%). Examination of the progeny from these seven crosses revealed all to be apparently normal individuals of the maternal species. The back-crosses all produced high egg hatch rates, and progeny that were normal. The deduction that the females involved in the hybrid matings had previously been mated by one or more conspecific males is obvious. The final cross, between a male A. 10-punctata and a female A. bipunctata produced 183 eggs of which only five hatched. These were all raised to adulthood. From examination of the larvae and adults it was obvious that these were true hybrids. The fourth instar larvae were paler than those ofA. bipunctata, but had the abdominal spotting pattern ofthis species rather than that ofA. 10-punctata. The patterns of the adults were variable. Two had pronotal patterns similar to those of the typical form of A. bipunctata, the others having pronota more like those ofA. 10-punctata, but with the spots fused together. The ground colour ofthe elytra (at three weeks old), was orange in all five cases, and was thus akin to that of A. 10- punctata. The elytra ofone was marked with a single bold black central spot as in typical A. bipunctata. The other four had small dark-brown dots on the elytra. Three had six of these dots, the other having ten, positioned as in a typical A. 10-punctata. The legs of all five hybrids were brown as in A. 10- punctata. Externally, two of the progeny appeared male, the other three appearing female. Attempts to back-cross these progeny failed to produce issue. Three ofthe individuals (one male and two females) did appear to mate normally, however, none of the eggs laid as a result showed any sign of embryonic development. Dissection of the two males showed each to have severely under- developed testes. The sipho of both males appeared indistinguishable from that of a normal A. bipunctata. The ovaries of the females were variable. In one the ovary was extremely under-developed. In the second the left ovary appeared normal, while the right ovary was only partly developed, and the ovarioles were somewhat distorted. The ovaries of the third appeared to be Ig ENTOMOLOGIST'S RECORD, VOL. 109 25.1.1997 normal The genitalia of the three females were all similar. As with female hybrids of these two species described by Ireland et al. (1986), the genitalia were more similar to those ofA. bipunctata than those ofA. 10-puctata, but in all three the infundibulum was unique, and readily distinguishable from that of either parental species. It is questionable whether, in back-cross matings, males of either parental species would be able to insert their sipho and manufacture a spermatophore efficiently in one ofthese female hybrids. Recently collected ladybirds The observations of interspecific matings between ladybirds recently collected extend the range of species which will attempt or actually mate together. Again most of the pairings observed were of congeneric species, however, two were between individuals from different genera. Both of these matings involved female C. 7-punctata, one pairing with a male Aphidecta obliterata, and the other mating, most incongruously with a male T. 16- punctata. Because of the difference in size between the sexes in these two pairings, and in particular the minute size of the male T. 16-punctata, the genital contact was examined both during the copulation and at the moment of withdrawal. In both cases the impression was that the male's sipho was fully embedded, an impression that was reinforced as the siphos were withdrawn. There are two possible explanations for the occurrence of interspecific hybrid matings between recently collected individuals. First, males may be stimulated by the close proximity of, or contact with, other coccinellids, particularly if they have not encountered females in the wild for some time. Reproductively mature male ladybirds are known to mount other coccinellids, and some other beetles, rather indiscriminately of species or sex, or even of whether the individual being mounted is alive or dead (Majerus, 1994). Second, recently collected females placed with males may be prepared to accept non-conspecific partners if the female has not copulated for some time. This may be the case when population density is low, and appropriate food is common so that members of a species do not become concentrated together on a few patches of resource-rich habitat. These two possibilities are not mutually exclusive. In three of the cases (numbers 1, 3 and 6) it seems probable that what might be termed "frustrated female syndrome" was the primary cause of the interspecific pairings. With respect to pair 1, both Coccinella 5-punctata and C. 11-punctata were recorded as being rare in the Spey Valley in 1987 (Duncan, pers. comm.). Indeed, the female C. 5-punctata in question is one & of only two recorded from that region since the early 1950s (Majerus Fowles, 1989; Majerus, 1994). That said, the fact that this female produced normal C. 5-punctata progeny suggests that she was not a virgin. In the case ofpair 3, the female species, Exochomus 4-pustulatus, was unusually scarce in the Lakenheath area in 1987. The fact that the female that accepted a HYBRIDISATIONINLADYBIRDS 19 C. bipustulatus male failed to produce any eggs suggests that she was devoid ofconspecific sperm. The same argument applies to the case ofthe female C. 7-punctata that mated withA. obliterata (pair 6). The instances ofa male E. 4-pustulatus which attempted to mate with both a Chilocorus renipustulatus and a male A. bipunctata (pair 2) and of the , male C. 7-punctata that attempted to mate with several C. 11-punctata (pair 4) are both probably the result ofmale "randiness". Certainly the male E. 4- pustulatus attempted to mount any ladybird he came into contact with. The other two pairings fit less well with either the female frustration syndrome or male randiness explanations. The pairing of an A. bipunctata with an A. 10-punctata only 24 hours after they were removed from overwintering sites is certainly difficult to explain. Usually female A. bipunctata have to be fed for several days after removal from overwintering sites before they mate. This is probably because ovaries that become dormant in the winter have to be reactivated and resourced. The fact that no eggs were laid by the female for 11 days after the mating suggests that the female was not bearing ovaries with mature eggs. The T. 16-punctata that mated with a female C. 7-punctata was one of 39 collected with five C. 7- punctata by sweeping long grass. All 44 ladybirds were placed in the same small perspex container. Both species were very common at the site making the female frustration extremely unlikely. It is possible that the male may have been stimulated by the proximity ofconspecific females in the box, and simply mounted the wrong partner. In such a case, the female would usually reject a ladybird of another species. That this female C. 7-punctata did not reject the male may be a consequence of his very much smaller size. The female may simply not have realised what was going on. None of these pairings produced true hybrid offspring. In all cases, either no eggs were laid, or the eggs were infertile, or the eggs were fertile but the progeny produced were normal individuals ofthe maternal species. Laboratoiypairings Most of the pairings set up were ofcongeneric species. The exceptions were crosses ofA. bipunctata with both P. 14-punctata and A. obliterata, and of E. 4-pustulatus and the two British Chilocorus species. The protocol of keeping females isolated from males for two weeks or longer proved successful in inducing hybrid mating in all cases except one. The general conclusion is that the rejection behaviour that female ladybirds of most species show towards non-conspecific males can be broken down easily. Interestingly, there was some indication that sympatric species mated less readily than those that do not have overlapping distributions. Thus, the pairs of species that hybridised most easily were Anatis ocellata and A. labiculata from England and the USA respectively, and Harmonia 4-punctata and H. axyridis from England and Japan respectively. The reason for this may be that part of the female's mate recognition system involves a rejection 20 ENTOMOLOGIST'S RECORD, VOL. 109 25.1.1997 response to males of other species that they encounter reasonably frequently in evolutionary time. This part of the recognition system is only likely to extend to sympatric species that share similar habitat requirements. The one pairing that did not produce a copulation involved female C. magnifica that were placed with male C. 7-punctata. This failure is of particular interest because the reciprocal pairing did produce matings, although no eggs resulted. Coccinella magnifica is a myrmecophile, usually living in close proximity to nests of the wood ant Formica rufa (Donisthorpe, 1920; Pontin, 1960; Majerus, 1989). It is morphologically very similar to C. 7-punctata, and often occurs with it. Speculation on the evolutionary divergence of these two species has concentrated on the ecological questions ofwhy C. magnifica lives close to ants, and why it does not live elsewhere (see Majerus, 1994 for review). Less consideration has been given to the way in which gene flow between the two species was arrested. The fact that female C. magnifica would not accept male C. 7- punctata, even when the females had been prevented from mating for a considerable period, suggests that the mate recognition system of females of this species is stronger than that ofmost. Only two of the successful pairings (C. magnifica male x C. 7-punctata female and C. 7-punctata male x C. 11-punctata) failed to produce eggs. However, the eggs from six of the pairings did not show any sign of development, indicating either that the eggs had not been fertilised, or that the zygotes perished early in embryogenesis. A small proportion of eggs developed from seven of the hybridisations. However, in only two of these, the reciprocal crosses ofA. ocellata and A. labiculata did any ofthe eggs hatch, and in both instances all the larvae died in the first instar. Interestingly the two reciprocal sets of crosses between non-sympatric congeneric species (A. ocellata x A. labiculata and H. 4- punctata x H. axyridis all showed some embryonic development, perhaps suggesting that the divergence between these species may not be as great as that between C. 7-punctata and either C. magnifica or C. 11-punctata. When are interspecific hybridmatings likely to occur in the wild? The ease with which females can be induced to mate with males from other species suggests that interspecific hybrid matings may occur in the wild whenever females of a species do not encounter males very frequently. This may occur in three situations. First, when a species is at very low density, males and females may simply not come across each other very often. This may be the result of a species either being very rare, or being less rare but distributed widely and thinly across the environment. Second, there is a period towards the end of the main reproductive season in Britain, but before the new generation of adults have emerged from their pupae and become reproductively mature, when the sex ratio of adults in mating condition may become strongly female biased. This is because the