綜合論述

台灣昆蟲 31: 117-131 (2011) Formosan Entomol. 31: 117-131 (2011)

Evolution of Asexual Queen Succession System in

Kenji Matsuura*

Laboratory of Insect Ecology, Graduate School of Environmental Science, Okayama University, Okayama 700-8530, Japan

ABSTRACT

The evolution and maintenance of sexual reproduction is believed to involve important tradeoffs. The queens of social insects are faced with a dilemma over the costs and benefits of sexual and . Asexual reproduction by a queen doubles her contribution to the gene pool. However, overuse of asexual reproduction reduces the genetic diversity of the offspring and thus the ability of the colony to adapt to environmental stress. Recent research suggests that queens of some Reticulitermes termites can solve this tradeoff by the conditional use of sexual and asexual reproduction, whereby queens produce neotenic (secondary) queens by but use sexual reproduction to produce workers and alates (sexual imagos). I also discuss proximate physiological mechanism and genetic background of the asexual queen succession system in the termites.

Key words: thelytoky, caste differentiation, genetic diversity, queen succession, queen pheromone, sex ratio

Introduction altruistic members. Hughes et al. (2008) compared female mating frequencies in and inclusive fitness eusocial and demonstrated theory predicts that high relatedness, that monandry was the ancestral state possibly generated by genetic factors such when arose in Hymenoptera as , inbreeding, or clonal and that mating by queens with multiple reproduction, is critical to the evolution of males is always derived. High levels of eusociality. However, relatedness among polyandry occur only in whose workers within colonies of some species is workers have lost reproductive totipotency. relatively low due to sexual reproduction, Similarly, polygyny is also derived. This including outbreeding, multiple mating indicates that high relatedness has played (polyandry), and multiple reproductive a decisive role in the females (polygyny), which lowers the and strongly supports inclusive fitness potential inclusive fitness gains for theory.

*Corresponding email: [email protected] Evolution of Asexual Queen Succession System in Termites 117 Living in social groups has not only has been regarded as an unusual case advantages but also drawbacks. For example, with little adaptive significance in nature. infectious diseases can potentially spread Even after the adaptive significance of more easily among group members than parthenogenesis was recognized, solitary-living individuals. Transmission is researchers believed that the function of more likely in groups because individuals parthenogenesis in termites was no more live at high densities and have frequent than “the best of a bad job”, that is, social contact. In addition, group members females used parthenogenesis only when are close relatives and thus susceptible to they failed to mate with males (Matsuura the same parasite infections (Schmid- and Nishida, 2001; Matsuura et al., 2002). Hempel, 1998). This can be understood Most recently, however, it was revealed using an analogy to human agriculture that parthenogenesis in some termite (e.g., Muira et al., 2008). Selective breeding species plays a much greater role than has of crops for desirable traits and against been previously understood. In the Japanese the undesirable ones leads to monocultures, subterranean termite R. speratus (Matsuura or entire farms of nearly genetically identical et al., 2009) and the North American plants. Little to no genetic diversity makes subterranean termite R. virginicus (Matsuura crops extremely susceptible to widespread and Vargo, in prep.), queens exclusively disease. Low genetic variance and closely use parthenogenesis to produce secondary spaced individuals may be the reasons (neotenic) queens. On the other hand, why the avian flu causes devastating queens produce workers and alates by effects on poultry farming and why plant outcrossing with the primary king. This hoppers wipe out acres of rice paddies at model of asexual queen succession (AQS) once. Both human agriculture and social or conditional use of sexual and asexual insect colonies face risks associated with reproduction can be studied to understand high density and low genetic diversity. the advantages and disadvantages of Perhaps the best solution to the thelytoky in termites. dilemma over the costs and benefits of sexual and asexual reproduction is to use Thelytoky in eusocial insects both modes of reproduction conditionally Thelytoky is a type of parthenogenesis and therefore to experience the advantages in which females are produced from of both. Indeed, recent studies have unfertilized eggs. Among the 12,500 uncovered unusual modes of reproduction and 2,400 termite species in the world, both in the cursor (Pearcy thelytoky has been reported in only nine et al., 2004), Wasmannia auropunctata ant species and seven termite species. (Fournier et al., 2005), and Vollenhovia Thus, the percentages of species with emeryi (Ohkawara et al., 2006), and in the thelytokous capability are 0.07% and termite Reticulitermes speratus (Matsuura 0.29% in ants and termites, respectively. et al., 2009), in which advantage is taken These values are much lower than 2%, of the social caste system to use sex for which is the percentage of all insect somatic growth, but parthenogenesis for species with thelytokous capabilities. By germ line production. taking advantage of asexual reproduction, The capability of parthenogenesis in queens of eusocial insects are able to Isoptera was first reported by Light in realize a twofold advantage over sexual 1944. However, the reproductive biology of reproduction by allowing the transmission termite parthenogenesis has not been of twice the number of genes to offspring. examined in detail beyond its notional This leads to the very interesting question documentation until recently. This is of why thelytoky is so rare among eusocial largely because parthenogenetic reproduction insects.

118 台灣昆蟲第三十一卷第二期 One possible explanation is that only Reticulitermes termites have a fixed a limited number of studies have carefully number of chromosomes (2n = 42). examined the possibility of thelytokous Karyotypic chromosome observations showed parthenogenesis in eusocial insects. that parthenogenetic offspring are diploid, Thelytoky may be more difficult to detect with 2n = 42 chromosomes in R. speratus. in eusocial insects than in other insects Although termite sex determination mecha- because its presence is sometimes concealed nisms are not completely understood, or suppressed by social structure; the males commonly appear heterogametic in mechanisms that cause this are described termites (Roisin, 2001). Interchange multiples in the next section. Indeed, findings by L. (chains or rings of chromosome) in the Keller of the conditional use of asexual male is common in termites reproduction in ants arose by pure (Syren and Luykx, 1977). Chromosome serendipity (Keller, 2007). Future studies multivalents have been observed in male may reveal thelytoky in many more meiosis in R. speratus, generating a eusocial species. Possibly the rarity of multiple-X, multiple-Y system (Matsuura, thelytoky in eusocial insects is attributable 2002). Because of the XY sex determi- to the importance of genetic diversity nation system, termite parthenogenesis among colony members. While asexual produces only female progeny. reproduction by a queen would increase The genotypes of parthenogenetic within-colony relatedness, the resulting offspring depend on the mode of partheno- reduction in genetic diversity within genesis (Templeton, 1982). Thelytokous colonies would lower homeostasis in colonies. parthenogenesis can be categorized into For example, low genetic diversity colonies two major cytological divisions, “apomixis are more afflicted by disease than (ploidy stasis)” and “automixis (ploidy genetically diverse colonies in bumblebees restoration).” In apomictic partheno- (Baer and Schmid-Hempel, 2001), honeybees genesis, known as clonal reproduction in (Seeley and Tarpy, 2007) and leaf-cutting , the features of meiosis are either ants (Hughes and Boomsma, 2004). In entirely or partially lacking. Only one honeybees, high-diversity colonies maintain maturation division takes place in the egg more uniform temperatures in their brood and this division is equational. The nests than did the uniform ones due to a offspring retain the genetic constitution of system of genetically based task speciali- the mother (excluding mutations), and zation (Jones et al., 2004). Thus, reduction heterozygosity is maintained in subse- in genetic diversity would render the quent generations. colonies less resilient to environmental Importantly, parthenogenesis of termites perturbation. A comprehensive survey of is not clonal. Thelytoky in termites, such species capable of thelytokous partheno- as R. speratus and R. virginicus, is genesis is a necessary next step in the accomplished by automixis with terminal effort to understand the costs and benefits fusion, in which two haploid pronuclei that of sexual and asexual reproduction for divide at meiosis ΙΙ fuse (Fig.1). Thus, eusocial insects. offspring are homozygous for a single maternal allele at all loci that did not Mechanism of termite parthenogenesis crossover, whereas offspring have the Research on chromosome numbers of same genotype as their mother at loci Isoptera showed that higher termites where crossover occurred. This causes a (Termitidae) are karyotypically uniform rapid reduction of heterozygosity (Matsuura (2n = 42), while lower termites are more et al., 2004). variable, with diploid numbers ranging from 28 to 56 (Bergamaschi et al., 2007).

Evolution of Asexual Queen Succession System in Termites 119 Fig. 1. Scheme of ploidy restoration by automixis with terminal fusion. This type of parthenogenesis results in rapid reduction of heterozygosity. r: recombination rate. From Matsuura et al. (2004).

Asexual queen succession (AQS) in colonies and lower lifetime fecundity in Reticulitermes speratus inbred colonies. Like most subterranean In most termite species, one king and termites, Reticulitermes species have one queen usually found colonies. In cryptic nesting habits with transient, termites, especially in lower termites, it hidden royal chambers underground or has long been believed that the inbreeding deep inside wood, making it difficult to cycles of generations of neotenic repro- reliably collect reproductives. Therefore, ductives propagate the colony after the the breeding system of subterranean death of the primary king and queen termites has been primarily estimated by (Bartz, 1979). Evidence for inbreeding genotyping workers or culturing labora- depression in termites is mounting, such tory colonies rather than censuses of field as higher mortality in inbred incipient colonies.

120 台灣昆蟲第三十一卷第二期 Fig. 2. A royal chamber of Reticulitermes speratus, where a single primary king (PK: arrowed) and dozens of secondary queens reproduce.

Reticulitermes speratus is the most wing buds differentiated from nymphs common termite in Japan. To obtain (updated after Matsuura et al., 2009; Fig. reproductives from a sufficient number of 3). natural colonies, we collected more than Sexual reproduction can lead to 1,000 nests in the field. We successfully important conflicts between sexes and found the royal chambers, where repro- within genomes. In monogamous termites, ductives and young broods were protected, conflicts between the primary king and of 55 colonies. In nearly all cases, primary queen can arise over parental investment kings were continuously present but and genetic contribution to offspring. Our primary queens had been replaced by an finding that primary queens are replaced average of 62.73 ± 8.54 (mean ± s.e.m, n = much earlier than primary kings in R. 55) secondary queens (updated after speratus leads to a paradox if the Matsuura et al., 2009; Fig. 2). These secondary queens are the daughters of the results indicate that primary kings live primary king. King-daughter inbreeding much longer than primary queens; should result in uneven genetic contri- replacement of the primary king is rare, bution to the secondary offspring (offspring whereas replacement of the primary queen of secondary queens) by the primary king is the rule at a certain point in colony and queen. The life-for-life relatedness of a development. All of the 3,450 secondary primary king to the secondary offspring queens collected from field colonies were produced by king-daughter inbreeding is nymphoid, i.e., neotenic reproductives with 5/8, while relatedness of the primary

Evolution of Asexual Queen Succession System in Termites 121 Fig. 3. Differentiation pathways of primary and secondary reproductives in the genus Reticulitermes. In field colonies of R. speratus, secondary queens always differentiate from nymphs but never from workers.

queen to the secondary offspring is 1/4 toward females (Investment ratio of male: when the primary king and the primary 0.415 ± 0.02SE). This inconsistency between queen are unrelated. Because male king-daughter inbreeding and sex investment primary reproductives are 2.5 times more ratio in alate production suggests that related to offspring than is the primary there is a different breeding system in queen under this system, colonies are which the king and queen have more expected to bias alate (new primary equal genetic contributions to offspring. reproductives that disperse) production in Together with Edward L. Vargo, we favor of males. Contrary to this prediction examined the genotypes within nests of R. under this breeding system, the alate sex speratus and uncovered an extraordinary ratio is slightly but significantly female- mode of reproduction (Matsuura et al., biased in this species (numerical ratio of 2009). Secondary queens are almost male = 0.43 ± 0.02SE). Because of the exclusively produced parthenogenetically larger size of females relative to males, the by the founding primary queens, whereas biomass sex ratio was even more biased workers and alates were produced by

122 台灣昆蟲第三十一卷第二期 Fig. 4. Proportion of parthenogenetically produced offspring (black) and sexually produced offspring (white) in secondary queens, workers and alates. Secondary queens have only maternal (primary queens’) alleles, while workers and alates have both maternal and paternal alleles indicating conditional use of sexual and asexual reproduction. MS: molecular standard. From Matsuura et al. (2009).

sexual reproduction (Fig. 4, 5). By using primary queens to workers (r = 0.49, parthenogenesis to produce secondary SEjackknife = 0.04) and to alate nymphs (r = queens, primary queens are able to retain 0.58, SEjackknife = 0.079) is not significantly the transmission rate of their genes to different from 0.5, the value expected descendants while maintaining genetic between a female and her sexual offspring. diversity in the workers and new primary This is twice the expected genetic reproductives even after the primary contribution queens would make to colony queen is replaced. The relatedness of the members under king-daughter inbreeding

Evolution of Asexual Queen Succession System in Termites 123 Fig. 5. Scheme for the breeding system with asexual queen succession in termites. This breeding system enables the primary queen to maintain her full genetic contribution to the next generation, while avoiding any loss in genetic diversity from inbreeding.

(rprimary queen to king-daughter offspring = 0.25). The offspring produced by outcrossing Parthenogenetic production of secondary between the king and parthenogenetic queens allows R. speratus to undergo queens may have greater fitness than queen succession without inbreeding (Fig. those produced by king-daughter inbreeding. 5). Heterozygosity of workers in colonies The production of secondary queens headed by secondary queens was as high through conditional parthenogenesis (Ho = 0.733) as that expected for offspring effectively extends the reproductive life of produced by outcrossing of the primary the primary queen, greatly expanding her king and the primary queen (He = 0.736: reproductive capacity (Fig. 7). This process Fig. 6). Likewise, there was no significant of queen succession allows the colony to reduction of heterozygosity in alate nymphs boost its size and possibly its growth rate produced in colonies with secondary without suffering any loss in genetic queens (Fig. 6). Further evidence of the diversity or diminishing the transmission lack of inbreeding in R. speratus colonies rate of the queen’s genes to her grand is provided by the low inbreeding coefficient offspring, feats that would not be possible of workers, which did not differ sig- if secondary queens were produced by nificantly from zero (FIT = 0.014, SEjackknife normal sexual reproduction. = 0.048, over all loci). The lack of consanguineous mating in this breeding Purging: another benefit of AQS system may also benefit primary kings. A faster rate of accumulation of

124 台灣昆蟲第三十一卷第二期 Fig. 6. Proportion of heterozygous loci (black) and homozygous loci (white) in secondary queens, workers and alates. The amount of heterozygosity expected for offspring produced by outcrossing of the primary king and the primary queen are indicated by arrowheads. From Matsuura et al. (2009).

deleterious mutations is a major cost of noptera undergo, eliminating the trans- asexual reproduction. In haplodiploid mission of deleterious recessive alleles to organisms, deleterious alleles are directly the sexual offspring. exposed to selection each generation in the haploid males, and there is no masking Genetic caste determination and AQS effect of dominance. Therefore, purging The question of how reproductives selection will cause a more rapid decrease and sterile workers differentiate within in the frequency of deleterious alleles at eusocial groups has long been a core issue haplodiploid loci than at comparable loci in the study of social insects. Recent in diploid organisms (Goldstein, 1994). studies have shown that not only environ- In termites, paradoxically, asexual mental factors but also genetic factors queen succession can function to purge affect caste differentiation. Genetic influences deleterious mutations. Parthenogenetic on queen-worker differentiation are essential offspring are homozygous for a single to the conditional use of sexual and maternal allele at all loci due to terminal asexual reproduction. In the termite R. fusion. Therefore, deleterious recessive genes speratus, parthenogens are strongly biased are exposed to selection in homozygous to develop into secondary queens, suggesting parthenogens. Parthenogens carrying homo- that differentiation into this caste is zygous recessive deleterious alleles should genetically influenced, possibly by whether not be able to survive or develop into individuals are heterozygous or homo- functional secondary queens. The obligate zygous at certain loci. The AQS system can occurrence of parthenogenesis in the work only if parthenogens have priority to normal life cycle of this species can become secondary queens. Why is it that eliminate recessive deleterious genes in nymphs produced by parthenogenesis can every generation, much like the genetic exclusively differentiate into secondary purging that haploid males of Hyme- queens when there are numerous sexually

Evolution of Asexual Queen Succession System in Termites 125 Fig. 7. Schematic diagram of asexual queen succession. PK: primary king, PQ; primary queen. SQ: secondary queen. SK: secondary king. P: parthenogenesis. S: sexual reproduction. Modified from Matsuura (2010).

produced nymphs at the same time? specialize on the other tasks required for Without any genetic influence on caste colony growth and survival. Pheromones differentiation, it seems impossible for this produced by reining queens have long to happen. A genetic system of homo- believed to be the prime factor inhibiting zygous advantage to be secondary queens the differentiation of new reproductive makes it possible. In addition, a stable AQS individuals. However, there has been very system requires the genetic advantage to little progress in the chemical identifi- be determined by an independent (unlinked) cation of such inhibitory pheromones. multilocus genotype. A single locus system Recently, I and my colleagues identified a cannot discriminate parthenogens and volatile inhibitory pheromone produced by sexually produced offspring, and thus AQS female neotenics (secondary queens) that is impossible, if the founding pair had the acts directly on target individuals to same allele at the locus. A multilocus suppress the differentiation of new female system provides a rigorous mechanism by neotenics (Matsuura et al., 2010). We which only parthenogens can develop into identified n-butyl-n-butyrate and 2-methyl- secondary queens, because terminal fusion 1-butanol as the active components of the yields progeny of near-total homozygosity. inhibitory pheromone. An artificial phero- An analysis of the relationship between mone blend consisting of these two the reproductive dominance of female compounds had a strong inhibitory effect neotenics obtained from experimentally similar to live neotenics. Surprisingly, the orphaned colonies and their genotypes at same two volatiles are also emitted by five microsatellite loci showed that homo- eggs, playing a role both as an attractant zygosity at two loci influenced the priority to workers and an inhibitor of repro- to differentiate into neotenics (Yamamoto ductive differentiation. This dual pro- and Matsuura, submitted data). These duction of an inhibitory pheromone by results suggest the existence of a multi- female reproductives and eggs probably locus queen determination system and reflects the recruitment of an attractant explain why parthenogens have genetic pheromone as an inhibitory pheromone priority to become neotenics in this termite and may provide a mechanism ensuring species. honest signaling of reproductive status with a tight coupling between fertility and Termite queen pheromone and AQS inhibitory power. Most recently, we found The hallmark of social insects is their that the queen pheromone also functions caste system: reproduction is primarily as a signal regulating colony-level egg monopolized by queens while workers production (Yamamoto and Matsuura, in

126 台灣昆蟲第三十一卷第二期 press). duction and thelytoky are m and (1-c) m, From the view point of queen phero- respectively, where c is the cost of mone, the most parsimonious evolutionary thelytoky. The number of male Mm and origin of AQS might be a loss of the female Mf alates produced by the colony receptor of queen pheromone in homo- founded by the mutant female are given zygous progenies at a certain locus. In by other words, the origin of the queen Mm = xms succession gene might be a recessive Mf = xm (1-s) + (1-x) (1-c) m deleterious gene. The individuals with where s is the sex ratio (male ratio) of homozygous sets of the gene lose the alates produced by the mutant colony. The queen pheromone receptor and thus have number of male Rm and female Rf alates the developmental priority to differentiate produced by the colonies founded by the into neotenics. resident females are given by Rm = (F-1) xms* Thelytoky and alate sex ratio in Rf = (F-1) xm (1-s*) + (F-1) (1-x) (1-c) m termites where s* is the sex ratio of alates In R. speratus, females that fail to pair produced by the resident colonies. with males found colonies cooperatively The alates of the next generation form with female partners or even alone, and P monogamous pairs (a male and a female) reproduce by parthenogenesis (Matsuura and reproduce sexually. The proportion of and Nishida 2001, Matsuura et al., 2002). mutant males in the P pairs is equal to the The diploid female progeny reproduced by proportion of mutant males in the total thelytokous parthenogenesis develop in number of males in the population. There- the same way as sexually produced offspring. fore, The number of mutant males and Parthenogenetic reproduction is advantageous, mutant females which formed mo- even if ultimately it may be inferior to nogamous pairs are P Mm / (Mm + Rm) and sexual reproduction in terms of long term P Mf / (Mf + Rf ), respectively. fitness. It enables females to reproduce in The males which failed to mate with the event that a mating partner cannot be females die, whereas the females which found. Therefore facultative parthenogenesis, failed to mate with males still found where both sexual and asexual colony colonies by alone or by female-female pairs foundation is sometimes possible, may and reproduce by thelytoky (Matsuura and explain why alate sex ratio is female biased Nishida, 2001). The number of females in some termite species. which reproduce by thelytoky in maleless Contrary to the intuitive expectation, colonies is P (1-x)/x, and thus the total an ESS model predicts that maleless number of mutant females reproducing by colony foundation does not cause female- thelytoky is given by biased alate sex ratios (Matsuura and P [(1-x)/x] [Mf / (Mf + Rf )] Dobata, in prep.). We consider a situation Then the monogamous pairs produce where F females found colonies in a m alates by sexual reproduction. On the population. We assume that one female is other hand, females of the maleless colonies a mutant and F-1 females are resident. produce (1 - c) m alates. In sexual repro- The sex ratio of alates produced by a duction, a parent male and female share colony is determined by the genotype of the genetic contribution to the alates of the founder female. Founder females the next generation Therefore, the expected produce alates by sexual reproduction and genetic contribution (i.e., fitness) of a thelytoky at the proportion of x and 1-x, mutant to the next generation is given by respectively. The total numbers of alates W(s, s*) = 1/2 [P Mm / (Mm + Rm)]m + produced by the foundresses’ sexual repro- 1/2 [P Mf / (Mf + Rf )]m + P [(1-x)/x] [Mf /

Evolution of Asexual Queen Succession System in Termites 127 (Mf + Rf )] (1-c) m queens do not require sperm from mates W(s, s*) achieves its maximal value to produce diploid (female) offspring. with respect to s when s = s*, yielding the Nevertheless, they retain sexual repro- ESS condition duction for production of workers and have thus the genetic diversity in the ∗ ∂W (s, s ) worker force is maintained. By using = 0 alternative modes of reproduction for the ∂s ∗ s=s queen and worker castes, genetic diversity As a consequence, the ESS sex ratio is in the worker population can be maintained. s*=1/2. As discussed above, reduced genetic This ESS model shows that the diversity in the worker force may be occurrence of maleless colony foundation detrimental for the colony because it leads does not cause sex-asymmetric fitness to reduced defense against parasites, less values of alates regardless of pairing efficient division of labor, and a decreased efficiency, although thelytoky enables range of environmental conditions that a females to found colonies even when they colony can tolerate. These costs are akin to failed to mate with males. those thought to lead to the instability of This result provides an important parthenogenetic reproduction in nonsocial implication that the actual female-biased organisms. Thus, sexual reproduction might alate sex ratio of R. speratus cannot be have important benefits for colony function explained by maleless colony foundation through increased defense against but should be explained by any other parasites, more efficient division of labor, mechanism. and an increased range of environmental An alternative hypothesis for the conditions that a colony can tolerate. female-biased alate sex ration in AQS Queens of these species take advantage of species is that the relatedness asymmetry the social caste system to use sex for caused by sex-specific parent-offspring producing workers, which amounts to mating biases investment toward the sex somatic growth but parthenogenesis, which of inbred parent (Kobayashi et al. in prep). does not involve the evolutionary cost of AQS system balances the genetic contri- sex is used for germ line production. bution to the next generation between the In R. speratus, only the non dispersing primary king and the primary queen for as neotenic (secondary) queens that develop long as the primary king is alive. However, within established colonies are produced inbreeding is inevitable after the primary asexually, whereas alates (adult primary king’s death, and should result in uneven queens) are produced sexually. Partheno- genetic contribution to the offspring by the genesis with terminal fusion results in primary king and queen. Thus even a near complete homozygosity, which should small bias in genetic contribution to next reduce the viability and fitness of the generation can cause an unequal alate sex secondary queens. However, the cones- ratio (Kobayashi et al. in prep). quences for secondary queens in this termite species may be minimal, because Comparison of the conditional use of secondary queens stay in the natal nest thelytoky between ants and termites protected and cared for by the existing worker force, unlike independent colony Asexual royal succession in R. founding by the primary king and queen. speratus is in some ways analogous to the Contrary to neotenic queens, alates of conditional use of sex found in C. termites disperse and found colonies cursor, W. auropunctata, and Vollenhovia independently, and are thus subjected to a emeryi. In these ant and termite species, number of environmental contingencies in

128 台灣昆蟲第三十一卷第二期 which genetic diversity is likely to be Fournier D, Estoup A, Orivel J, Foucaud J, advantageous. Jourdan H, Le Breton J, Keller L. Although conditional use of sexual 2005. Clonal reproduction by males and asexual reproduction is currently and females in the little fire ant. known only in three ants and a single Nature 435:1230-1234. termite species, breeding systems involving Goldstein DB. 1994. Deleterious mutations conditional parthenogenesis may occur in and the evolution of male haploidy. other eusocial insects. To date, the ability Am Nat 144: 176-183. to reproduce parthenogenetically has been Hughes WOH, Oldroyd BP, Beekman M, reported in unmated females in seven Ratnieks FLW. 2008. Ancestral mo- termite species from four different families, nogamy shows kin selection is key to in which the production of neotenic re- the evolution of eusociality. Science placement reproductives is common. This 320: 1213-1216. raises the possibility that the conditional Jones JC, Myerscough MR, Graham S, use of sex and parthenogenesis could be Oldroyd BP. 2004. Honey nest widespread within this ecologically and thermoregulation: diversity promotes economically important group of social stability. Science 305: 402-404. insects. In addition, the genetic system of Keller L. 2007. Uncovering the biodiversity homozygous advantage to secondary queens of genetic and reproductive systems: may provide an ideal opportunity to Time for a more open approach- identify the queen determination gene in American Society of Naturalists E. O. termites. Wilson award winner address. Am Nat 169: 1-8. Acknowledgments Light SF. 1944. Parthenogenesis in termites of the genus Zootermopsis. Univ Calif I thank C. Himuro, T. Yokoi, Y. Publ Zool 43: 405-412. Yamamoto and T. Yashiro for research Matsuura K. 2002. A test of the assistance; K. Tsuji, E. L. Vargo, S. Dobata, haplodiploid analogy hypothesis in the E. Hasegawa, J. Yoshimura and K. termite Reticulitermes speratus Kobayashi gave helpful advices. This work (Isoptera: Rhinotermitidae). Ann Entomol was supported by the Japan Society for Soc Am 95: 646-649. the Promotion of Science and the Program Matsuura K. 2010. Sexual and asexual for Promotion of Basic Research Activities reproduction in termites. pp 255-277. for Innovative Biosciences. In: Bignell DE, Roisin Y, Lo N (eds). Biology of Termites: a Modern Synthesis. References Springer. Matsuura K, Fujimoto M, Goka K. 2004. Baer B, Schmid-Hempel P. 2001. Unexpected Sexual and asexual colony foundation consequences of polyandry for parasitism and the mechanism of facultative and fitness in the bumblebee, Bombus parthenogenesis in the termite terrestris. Evolution 55: 1639-1643. Reticulitermes speratus (Isoptera, Bartz SH. 1979. Evolution of eusociality in Rhinotermitidae). Insectes Soc 51: termites. Proc Natl Acad Sci USA 76: 325-332. 5764-5768. Matsuura K, Fujimoto M, Goka K, Nishida Bergamaschi S, Dawes-Gromadzki TZ, T. 2002. Cooperative colony foundation Scali V, Marini M, Mantovani B. 2007. by termite female pairs: altruism for Karyology, mitochondrial DNA and survivorship in incipient colonies. the phylogeny of Australian termites Anim Behav 64: 167-173. Chromosome Res 15: 735-753. Matsuura K, Himuro C, Yokoi T,

Evolution of Asexual Queen Succession System in Termites 129 Yamamoto Y, Vargo EL, Keller L. Pearcy M, Aron S, Doums C, Keller L. 2010. Identification of a pheromone 2004. Conditional use of sex and regulating caste differentiation in parthenogenesis for worker and queen termites. Proc Natl Acad Sci 107: production in ants. Science 306: 1780- 12963-12968. 1783. Matsuura K, Nishida T. 2001. Comparison Roisin Y. 2001. Caste sex ratios, sex of colony foundation success between linkage, and reproductive strategies sexual pairs and female asexual units in termites. Insectes Soc 48: 224-230. in the termite Reticulitermes speratus Schmid-Hempel P. 1998. Parasites in (Isoptera: Rhinotermitidae). Popul Ecol social insects. Princeton: Princeton 43: 119-124. University Press. 392 pp. Matsuura K, Vargo EL, Kawatsu K, Seeley TD, Tarpy DR. 2007. Queen promis- Labadie PE, Nakano H, Yashiro T, cuity lowers disease within honeybee Tsuji K. 2009. Queen succession through colonies. Proc R Soc B-Biol Sci 274: asexual reproduction in termites. 67-72. Science 323: 1687-1687. Syren RM, Luykx P. 1977. Permanent Muira WM, Wong GKS, Zhang Y, Wang J, segmental interchange complex in the Groenen MAM, Crooijmans R, Megens termite Incisitermes schwarzi. Nature HJ, Zhang H, Okimoto R, Vereijken 266:167-168. A and others. 2008. Genome-wide Templeton AR. 1982. The prophecies of assessment of worldwide chicken SNP parthenogenesis. pp 75-102. In: Dingle genetic diversity indicates significant H, Hegmann JP (eds). Evolution and absence of rare alleles in commercial Genetics of Life Histories. Springer breeds. Proc Natl Acad Sci USA 105: Varlag, Berlin. 17312-17317. Yamamoto Y, Matsuura K. Queen phero- Ohkawara K, Nakayama M, Satoh A, mone regulates egg production in a Trindl A, Heinze J. 2006. Clonal repro- termite. Biol Lett, in press. duction and genetic caste differences in a queen-polymorphic ant, Vollenhovia Received: April 17, 2011 emeryi. Biol Lett 2: 359- 363. Accepted: April 20, 2011

130 台灣昆蟲第三十一卷第二期 無性生殖的白蟻蟻后繼承系統之演化

Kenji Matsuura*

日本岡山大學 環境科學研究所 昆蟲生態研究室

摘 要

有性生殖的演化和維持被認為有重要的代價權衡參與其中，社會性昆蟲的王后必 須在有性生殖與無性生殖的代價和利益中進行抉擇。王后可經由無性生殖加倍其貢獻 至基因庫，然而過度的無性生殖將降低後代的遺傳多樣性，並因此降低群體對於環境 壓力的適應能力。近年的研究顯示散白蟻屬 Reticulitermes 的一些蟻后可以有條件的 運用有性與無性生殖來解決此代價權衡，蟻后經由孤雌生殖產生下一世代的蟻后，工 蟻則經由有性生殖產生。我也將討論無性生殖的白蟻蟻后繼承系統可能的生理機制和 潛在的遺傳背景。

關鍵詞： 產雌孤雌生殖、階級分化、遺傳多樣性、蟻后繼承、蟻后費洛蒙、性別比例。

*論文聯繫人 Corresponding email: [email protected] Evolution of Asexual Queen Succession System in Termites 131