Original article Jay D. Evans Parentage and sex allocation in the facultatively polygynous ant Myrmica tahoensis Received: 3 March 1998 / Accepted after revision: 20 June 1998 Abstract Most social groups have the potential for reproductive con¯ict among group members. Within insect societies, reproduction can be divided among multiple fertile individuals, leading to potential con¯icts between these individuals over the parentage of sexual ospring. Colonies of the facultatively polygynous ant Myrmica tahoensis contain from one to several mated queens. In this species, female sexuals were produced almost exclusively by one queen. The parentage of male sexuals was more complex. In accordance with predic- tions based on worker sex-allocation preferences, male- producing colonies tended to have low levels of genetic relatedness (i.e., high queen numbers). Correspondingly, males were often reared from the eggs of two or more queens in the nest. Further, over half of the males pro- duced appeared to be the progeny of fertile workers, not of queens. Overall investment ratios were substantially more male biased than those predicted by genetic re- latedness, suggesting hidden costs associated with the production of female sexuals. These costs are likely to include local resource competition among females, most notably when these individuals are adopted by their maternal nest. Key words Sex allocation á DNA ®ngerprinting á Myrmica tahoensis á Ants á Microsatellite DNA Introduction Inclusive ®tness theory (Hamilton 1964) provides a framework for explaining the diversity of social behav- iors exhibited in the animal kingdom. Speci®cally, members of social groups are predicted to cooperate and compete in a manner that enhances their genetic con- tributions to future generations (inclusive ®tness; e.g., Hamilton 1972; Crozier and Pamilo 1996). Extensions of kin-selection theory (Trivers and Hare 1976; Keller and Reeve 1994) have applied Hamilton's insight to con¯icts over reproduction in colonies of social Hymenoptera (ants, bees, and wasps; reviewed by Seger 1991). These con¯icts, broadly speaking, fall into one of two classes. First, fertile colony members might be in direct con- ¯ict over reproduction. Unequal rates of reproduction among nestmate queens (reproductive skew; Keller and Reeve 1994) are common in ants and other social insects (e.g., Reeve and Ratnieks 1993; Bernasconi and Keller 1996). Reproductive skew can aect both overall colony relatedness levels (by decreasing the eective number of queens) and the costs and bene®ts received by queens when they share nests. A laboratory study involving polygynous colonies of the imported ®re ant, Solenopsis invicta, suggested that particular queens are able to dominate the production of sexual ospring (Ross 1988). Such a skew in reproduction has implications for both the coexistence of queens in polygynous nests (Keller and Reeve 1994) and sex-allocation patterns across colonies. Another form of reproductive competition involves the parentage of male sexuals. In many social insects, workers retain active ovaries and therefore have the potential to compete directly with their queens for the production of male-destined haploid eggs (Bourke 1988). Fertile workers might be in con¯ict with both their queens and nestmate workers over male parentage. The outcome of these con¯icts can be assessed through pre- cise estimates of male parentage (Evans 1993). Second, colony members might be in con¯ict less di- rectly over the ways in which colony resources are allo- Behav Ecol Sociobiol (1998) 44: 35±42 Ó Springer-Verlag 1998 J. D. Evans Department of Biology, University of Utah, Salt Lake City, UT 84112, USA Department of Entomology, University of Georgia, Athens, GA 30602, USA Present address: J.D. Evans (&) Department of Entomology, 410 Forbes Building, University of Arizona, Tucson, AZ 85721, USA e-mail: jevans@ag.arizona.edu, Tel.: +1-520-6215855, Fax: +1-520-6211150