Strict monandry in the ponerine army ant genusSimopelta suggests that colony size and complexity drivemating system evolution in social insectsDANIEL J. C. KRONAUER,* SEAN O’DONNELL,† JACOBUS J. BOOMSMA‡ andNAOMI E. PIERCE**Museum of Comparative Zoology, Harvard University, 26 Oxford Street, Cambridge, MA 02138, USA, †Animal BehaviorProgram, Department of Psychology, Box 351525, University of Washington, Seattle, WA 98195, USA, ‡Centre for SocialEvolution, Department of Biology, University of Copenhagen, Universitetsparken 15, 2100 Copenhagen, DenmarkAbstractAltruism in social insects has evolved between closely related full-siblings. It is thereforeof considerable interest why some groups have secondarily evolved low within-colonyrelatedness, which in turn affects the relatedness incentives of within-colony cooperationand conflict. The highest queen mating frequencies, and therefore among the lowestdegrees of colony relatedness, occur in Apis honeybees and army ants of the subfamiliesAenictinae, Ecitoninae, and Dorylinae, suggesting that common life history features suchas reproduction by colony fission and male biased numerical sex-ratios have conver-gently shaped these mating systems. Here we show that ponerine army ants of the genusSimopelta, which are distantly related but similar in general biology to other army ants,have strictly monandrous queens. Preliminary data suggest that workers reproduce inqueenright colonies, which is in sharp contrast to other army ants. We hypothesize thatdifferences in mature colony size and social complexity may explain these strikingdiscrepancies.Keywords: colony fission, eusociality, Formicidae, mating frequency, microsatellites, maleparentageReceived 12 July 2010; revision received 29 October 2010; accepted 4 November 2010IntroductionSingle mating (monandry) by females was the ancestralcondition in all major lineages of eusocial insects(Boomsma 2007; Hughes et al. 2008). This is importantfor two reasons: first, it means that relatedness among afemale’s offspring was high, which favoured the evolu-tion of eusociality (Crozier & Pamilo 1996; Boomsma2007, 2009; Crozier 2008), and second, that females min-imized the cost of mating. Multiple mating (polyandry)is thought to be costly for females because it increasesthe risk of predation and contracting diseases, and itaugments time and energy expenditure. Furthermore,males can physically harm females during copulation(e.g. Kamimura 2008) and storing ejaculates from multi-ple males can be associated with immunity costs (Baeret al. 2006) and reduced sperm viability because ofsperm competition (den Boer et al. 2010). In line withthese risks, monandry or low levels of facultative poly-andry prevail in the social Hymenoptera (Strassmann2001; Hughes et al. 2008; Boomsma et al. 2009). Somenotable exceptions exist, however, and substantial efforthas been directed towards understanding the adaptivevalue of the high levels of obligate polyandry in thesesystems (Crozier & Page 1985; Boomsma & Ratnieks1996; Palmer & Oldroyd 2000; Strassmann 2001; Crozier& Fjerdingstad 2001; Boomsma et al. 2009). Most promi-nent among these are the army ants of the dorylomorphsubfamilies Aenictinae, Ecitoninae, and Dorylinae(abbreviated as the AenEcDo army ants throughout thismanuscript; Kronauer et al. 2007a; Kronauer 2009), theAtta and Acromyrmex leaf-cutting ants (Villesen et al.2002; Sumner et al. 2004), the Pogonomyrmex harvesterCorrespondence: Daniel J. C. Kronauer,E-mail: [email protected] 2010 Blackwell Publishing LtdMolecular Ecology (2011) 20, 420–428 doi: 10.1111/j.1365-294X.2010.04945.xants (Cole & Wiernasz 2000; Pol et al. 2008), the Vespulawasps (Foster & Ratnieks 2001), and the Apis honeybees(Oldroyd et al. 1998; Tarpy et al. 2004).While the precise parameters in each case are still notfully understood, four factors seem particularly impor-tant in explaining the evolution of polyandry in socialHymenoptera, three of which are directly related toincreased genetic diversity within colonies: first, geneti-cally diverse colonies might be more resistant to para-site and pathogen infections (Hamilton 1987; Shermanet al. 1988); second, genetic diversity might increasehomeostasis via more effective task allocation (Oldroyd& Fewell 2007); third, multiple mating might, undercertain circumstances, alleviate the cost of producingdiploid, sterile males when queens mate with malesthat carry an identical sex-locus allele (Page 1980).Finally, queens of species with large reproductive out-put might mate with numerous males because a singlemale may not provide sufficient sperm (Cole 1983). Theoverall trend suggests that obligate polyandry hasmainly evolved in species with large and complex soci-eties, which is in accordance with hypotheses one, two,and four (Boomsma et al. 2009). The application ofhypothesis three, on the other hand, has largely beenlimited to explaining the idiosyncratic mating system ofhoneybees (Page 1980; Tarpy & Page 2001, 2002; but seeKronauer et al. 2007a for army ants).Among the groups with obligate polyandry, armyants and honeybees stand out in that their queens haveby far the highest mating frequencies, with typical matenumbers somewhere between 10 and 40 (Tarpy et al.2004; Kronauer et al. 2007a). Two strikingly convergentaspects of the army ant and honeybee mating systemmight at least partially account for the extreme queen-mating frequencies: reproduction by colony fission, andhighly male biased numerical sex-ratios (Kronauer et al.2004, 2007a; Kronauer 2009). In monogynous specieswith colony fission, a mother colony invests substan-tially in a single or very few daughter queens, whichretain a large part of the colony’s worker force. Thismode of reproduction is rare among most groups ofeusocial Hymenoptera, but is the rule in army ants, hon-eybees, stingless bees, and many ponerine ants (Peeters& Ito 2001; Boomsma et al. 2009). Dependent colonyfounding, where workers accompany young queens,also occurs in many polygynous ant species (Keller1991) and epiponine wasps (Strassmann et al. 1997), butthe social dynamics in such systems are very differentfrom monogynous species with colony fission (Peeters& Ito 2001). Models suggest that under colony fission inmonogynous species, it might be advantageous forqueens to avoid extreme colony genotypes by matingmultiply and thereby decreasing the genetic variancebetween colonies (Page 1980; Sherman et al. 1988;Kronauer et al. 2007a;