Proc. Natl. Acad. Sci. USA Vol. 90, pp. 8412-8414, September 1993 Evolution Fighting for a harem of queens: Physiology of reproduction in Cardiocondyla male ants JURGEN HEINZE AND BERT HOLLDOBLER Theodor-Boveri-Institut, Lehrstuhl ftr Verhaltensphysiologie und Soziobiologie, Am Hubland, D 97074 Wiirzburg, Federal Republic of Germany Communicated by Edward 0. Wilson, June 17, 1993

ABSTRACT Several species of the ant genus Cardio- incurring a different risk: that the queens mate with additional condyla produce dimorphic males, which exhibit sharply dif- males and use his sperm only to fertilize worker-destined ferent mating strategies. Winged males typically disperse to eggs, which again implies zero direct fitness. Male ants are mate outside the nest, whereas wingless, ergatoid males stay in thus under strong selective pressure to husband their sperm the nest and aggressively employ their mandibles against carefully, and as expected, they behave differently from competing ergatoid males to monopolize the virgin queens males of other animals: they act rather timidly and engage in eclosing in the nest. Such aggressive mating strategy would only some scramble competition with no evidence of male fight- be adaptive if the males had unlimited sperm supply. Histo- ing. Ant males are rather invariably characterized by a logical studies showed that, contrary to the rule in the Hy- voluminous thorax, large wings and genitalia, and a small menoptera order, the ergatoid Cardiocondyla males are indeed head with large eyes and more or less imperfectly developed able to produce sperm during their entire adult life. Winged mandibles (17)-i.e., their morphology is adapted to locating males, on the other hand, have only a limited sperm supply young queens and to mating rapidly but almost never to since spermatogenesis ceases in the late pupal stage. serious fighting. As in other groups ofHymenoptera (18-20), however, there Sperm cells are minute and cheaply produced, but eggs are are exceptions in which ant males are morphologically aber- relatively large, rich in nutrients, and thus expensive. Be- rant. Males of several species of Hypoponera (18) and Car- cause ofthese differences between male and female gametes, diocondyla (21) are large-headed, wingless, and endowed with the two sexes generally follow different reproductive strat- strong, heavily sclerotized mandibles. In some taxa, these egies. Females tend to be choosy in selecting their mates, as "'ergatoid" males co-occur with normal winged males, and to pair offwith the wrong male would cause the loss ofa large whereas the latter behave as ordinary ant males do, deadly amount of energy invested in eggs and/or broodcare. Males fighting has been observed among the ergatoid males of in contrast, if they have sperm either stored in abundance or Cardiocondyla wroughtonii, Cardiocondyla nuda, and other continuously renewable, are expected to mate rather indis- species. In C. wroughtonii, ergatoid males both kill male criminately with as many females as possible and often to callows with their saber-shaped mandibles and smear secre- compete aggressively with one another for access to females tions that induce worker aggression onto the cuticle of adult (1, 2). There are, of course, many examples of this kind, but rivals (22-24). Adult males of C. nuda tolerate each other but how often are males so supplied and how often is their attempt to kill all freshly eclosing rivals by biting (25). Thus, behavior really like this? Leaving aside possible abundant colonies typically contain only one or a few adult ergatoid prestorage ofsperm, renewable sperm is not the rule in insect males, which monopolize the newly eclosing queens. Fighting males (3-5). If sperm is limited, however, natural selection among winged Cardiocondyla males has never been observed, theory predicts a less intensely aggressive competitive strat- although, as in other ant species, there may be scramble egy. The sharply dimorphic males of the ant genus Cardio- competition (26). Winged males inseminate nestmate queens if condyla provide a test for this prediction. Their differences they have the chance to do so, but most of their sexual may also illuminate male strategies in many other groups. activities occur during mating flights. Consequently, winged In the Hymenoptera order (e.g., bees, wasps, and ants), the males are rarely attacked by ergatoid males (22, 23). testes degenerate by the time males reach sexual maturity, To struggle for a harem evidently would not make much and adult males therefore have only a limited amount of sense with respect to limited sperm supply. Multiple matings sperm available (6-9). Though in some parasitoid wasps (10) of ergatoid males have repeatedly been observed (22, 23). and ants (unpublished observations) males may regain mating However, it was not reported whether sperm was actually capability after resting, there is no evidence that "sperm transferred during all of the copulations. Therefore, we replenishment" is correlated with continual spermatogene- investigated spermatogenesis and male insemination ability sis. In contrast, even in parasitoid wasps the testes decrease in several species of Cardiocondyla. in size soon after eclosion (11), and the males' ability to inseminate is limited (12, 13). The evolutionary consequences of this extraordinary con- MATERIALS AND METHODS dition on male mating strategies were examined in detail for Live colonies of C. nuda were collected in Barranco de Las ants (14, 15). The sperm cells produced by an ant male suffice Angustias, La Palma, Canary Islands, and housed in the to inseminate only one or a few queens: the published record laboratory in small plastic boxes with a plaster floor. Ants for an ant male is 10 successful copulations (16). Ant males were fed three times a week with diluted honey and pieces of thus face the problem of either investing all of their sperm in cockroaches. Dissections were carried out according to the one single queen, risking the loss of all their direct fitness technique described by Buschinger and Alloway (27). His- when this queen dies without having produced sexual off- tological studies were performed with freshly fixed material spring; or they can spread their sperm over several queens, from this population, and material from B.H.'s ant collection. For histological investigations, whole ants were fixed in The publication costs ofthis article were defrayed in part by page charge alcoholic Bouin solution, embedded in Spurr's low-viscosity payment. This article must therefore be hereby marked "advertisement" medium (28), and serially sectioned at 6 ,Am with glass knives in accordance with 18 U.S.C. §1734 solely to indicate this fact. and a Reichert-Jung microtome. The sagittal sections were 8412 Downloaded by guest on September 24, 2021 Evolution: Heinze and Hdlldobler Proc. Natl. Acad. Sci. USA 90 (1993) 8413 attached to egg-albumin-coated slides and stained on a hot Furthermore, the ecological factors that favor the evolu- plate with Mallory's solution (1% methylene blue/1% azur II tion of ergatoid males in some but not all Cardiocondyla are in 1% sodium borate). obscure, and it is unknown why ergatoid males completely replaced winged males in some species, while in others a male RESULTS AND DISCUSSION polymorphism is maintained. Male ergatoidy is obviously correlated with stationary intranidal mating, a strategy that, Behavioral observations showed that ergatoid males of C. although rather common among ants, is as yet comparatively nuda are capable ofinseminating a surprisingly large number little understood. As colonies of Cardiocondyla are typically of queens. Dissection of two samples, containing 16 and 38 highly polygynous (22-26), a male that obtains a harem not queens, respectively, that eclosed in the presence of single only inseminates sisters but probably also less closely related males showed that all were inseminated. Our monitoring also young queens. Thus, inbreeding is avoided, which in many revealed that one male was still capable oftransferring sperm ants is deleterious because ofthe production ofsterile diploid at the relatively old age of 60 days and suggested that in males (33). As Cardiocondyla queens may disperse on the contrast to other Hymenopteran males, sperm supply may be wing after mating (26) and large colonies may split into unlimited in ergatoid males of Cardiocondyla. several colony buds (25, 34), by inseminating all queens in a By histological studies, we found that spermatogenesis colony, a male very effectively increases its reproductive continues in ergatoid males of C. nuda, C. cf. mauritanicus, success. However, to exclusively produce ergatoid males and C. wroughtonii throughout their whole life span. Mature becomes a risky strategy when average colonies are small and sperm cells were already available to the males immediately on average rear only few queens. In these cases, the pro- after eclosion, and the testicular follicles were fully intact and duction of winged males with the capability to disperse and contained all stages of spermatogonia even in males more to mate with the young queens produced in other colonies than 4 months old (Fig. 1). To our knowledge this is the first might be advantageous. As yet, ecological requirements and case of continuing sperm production within the order Hy- sociometric parameters, which might help to understand menoptera. On the other hand, in winged males of C. male reproductive strategies, are wroughtonii and C. emeryi, as in winged males of 25 addi- unknown for the various tional species of nine subfamilies (for a complete list, see ref. Cardiocondyla species. 29), spermatogenesis was completed shortly after eclosion, In any case, the correlation between reproductive physi- their follicles were collapsed, and all sperm cells were stored ology and behavior in Cardiocondyla males convincingly in the seminal vesicles. corraborates a central paradigm of behavioral ecology and It is especially remarkable that the two male reproductive suggests a correlate of fighting and high polygamy versus strategies co-occur within some species and that both winged scramble competition and monogamy that may be tested in males with completed spermatogenesis and ergatoid males other groups of animals. with continuing spermatogenesis may be reared within the same colony. We thank M. Obermayer for her excellent assistance in histology. The basis for this striking polymorphism is K. Yamauchi, Gifu University, kindly provided a copy of a forth- completely unknown. According to karyotype analyses (un- coming manuscript, passages of which were translated by Y. Imam- published results; see also ref. 26) and rearing experiments ichi. W. D. Hamilton (Oxford University) and E. 0. Wilson (Har- (25), ergatoid Cardiocondyla males are haploid as are winged vard University) made helpful comments on an earlier draft of this ant males, and both male morphs can develop from uninsem- manuscript. These studies were supported by the Deutsche Fors- inated eggs. Queen polymorphism, a similar phenomenon chungsgemeinschaft (project 18 ofthe Sonderforschungsbereich 251, that is better studied, is genetically mediated in at least two University of Wurzburg, Germany). ant species but probably environmentally or pheromonally controlled in others (30-32). 1. Bateman, A. J. (1948) Heredity 2, 349-368. 2. Trivers, R. L. (1972) in Sexual Selection and the Descent of Man, ed. Campbell, B. (Aldine, Chicago), pp. 136-179. 3. Engelmann, F. (1970) The Physiology of Insect Reproduction (Pergamon, Oxford). 4. de Wilde, J. & de Loof, A. (1973) in The Physiology ofInsecta, ed. Rockstein, M. (Academic, New York), pp. 12-95. 5. Dumser, J. B. (1980) Annu. Rev. Entomol. 25, 341-369. 6. Bishop, G. H. (1920) J. Exp. Zool. 31, 267-286. 7. Holldobler, B. (1966) Z.-Vgl. Physiol. 52, 430-455. 8. Wheeler, D. E. & Krutzsch, P. H. (1992) J. Morphol. 211, 307-312. 9. Keller, L. & Passera, L. (1992) Behav. Ecol. Sociobiol. 31, 359-366. 10. Nadel, H. & Luck, R. F. (1985) Ann. Entomol. Soc. Am. 78, 410-414. 11. Wilkes, A. (1965) Can. Entomol. 97, 647-657. 12. Griffiths, N. T. & Godfray, H. C. J. (1988) Behav. Ecol. So- ciobiol. 22, 211-217. 13. Gordh, G. & DeBach, P. (1976) Can. Entomol. 108, 583-589. 14. Holldobler, B. & Bartz, S. (1985) in Ecology and Sociobiology, eds. Holldobler, B. & Lindauer, M. (Sinauer, Sunderland, MA), pp. 237-257. 15. Holldobler, B. & Wilson, E. 0. (1990) The Ants (Harvard Univ. Press, Cambridge, MA). 16. Winter, U. & Buschinger, A. (1986) Entomol. Gener. 11, 125-137. FIG. 1. Longitudinal section through a 10-day-old ergatoid male 17. Wheeler, W. M. (1910) Ants: Their Structure, Development of C. nuda. (Upper) Whole habitus showing fully developed follicles and Behavior (Columbia Univ. Press, New York). of the testes (T). (Lower) Closeup of the testis, showing different 18. Hamilton, W. D. (1979) in Sexual Selection and Reproductive stages of the spermatogenesis: SG, spermatogonia; S, mature sperm Competition in Insects, eds. Blum, M. S. & Blum, N. A. cells. (Academic, New York), pp. 167-220. Downloaded by guest on September 24, 2021 8414 Evolution: Heinze and Holldobler Proc. Natl. Acad. Sci. USA 90 (1993)

19. Kukuk, P. F. & Schwarz, M. (1988) Pan Pac. Entomol. 64, 27. Buschinger, A. & Alloway, T. M. (1978) Insectes Soc. 25, 131-137. 339-350. 20. Danforth, B. N. (1991) Behav. Ecol. Sociobiol. 29, 235-247. 28. Spurr, A. R. (1969) J. Ultrastruct. Res. 26, 31-43. 21. Kugler, J. (1983) Isr. J. Entomol. 17, 1-21. 29. Hoildobler, B. & Engel-Siegel, H. (1982) Psyche 89, 113-132. 22. Kinomura, K. & Yamauchi, K. (1987) J. Ethol. 5, 75-81. 30. Buschinger, A. (1975) Naturwissenschaften 62, 239. 23. Stuart, R. J., Francoeur, A. & Loiselle, R. (1987) Naturwis- 31. Heinze, J. & Buschinger, A. (1989) Insectes Soc. 36, 139-155. senschaften 74, 548-549. in Biology and Evolution of 24. Yamauchi, K. & Kawase, N. (1992) Naturwissenschaften 79, 32. Buschinger, A. & Heinze, J. (1992) 274-276. Social Insects, ed. Billen, J. (Leuven Univ. Press, Leuven, 25. Heinze, J., Kuhnholz, S., Schilder, K. & Holidobler, B. (1993) Belgium), pp. 11-23. Insectes Soc. 40, in press. 33. Ross, K. G. & Fletcher, D. J. C. (1985) Evolution 39, 888-903. 26. Yamauchi, K. & Kinomura, K. (1993) in Evolution of Insect 34. Stuart, R. J. (1990) in Social Insects and the Environment, eds. Societies, eds. Inoue, T. & Yamane, S. (Hakuhinsha, Tokyo), Veeresh, G. K., Mallik, B. & Viraktamath, C. A. (Oxford; in press. IBH, New Delhi), p. 242. Downloaded by guest on September 24, 2021