LETTER Effectiveness of sexual selection in removing mutations induced with ionizing radiation Jacek Radwan Institute of Environmental Sciences, Jagiellonian University, ul. Gronostajowa 7, 30-387 Cracow, Poland Correspondence: E-mail: radwan@eko.uj.edu.pl Abstract Because of the production of males, sexual populations are expected to incur a 50% cost in potential growth rate. However, theory predicts that sexual competition between males can compensate for this cost by decreasing the mutation load of sexual populations. To test this hypothesis, I induced mutations in male bulb mites with ionizing radiation and subjected their progeny (F 1 ) to two selective regimes differing in opportunity for sexual selection. Mutations which were not removed by selection acting on the F 1 decreased embryonic viability in the F 2 . Viability was significantly higher in the treatment in which there was an opportunity for sexual selection than in the treatment in which sexual selection was experimentally eliminated. The results indicate that sexual selection can increase population fitness and, at least partly, compensate for the cost of sex. Keywords Condition dependence, deleterious mutations, good genes, ionizing radiation, mainten- ance of sex. Ecology Letters (2004) 7: 1149–1154 INTRODUCTION The evolutionary consequences of anisogamy have been extensively explored by evolutionary biologists for two main reasons. First, the reproductive success of the sex, which produces many small gametes, i.e. males, is expected to increase with the number of mating partners (Bateman 1948). This leads to the sexual competition underlying sexual selection (Darwin 1871; Andersson 1994). Secondly, population growth is limited by the reproductive rate of the sex producing fewer and larger gametes, i.e. females. Therefore, asexual lineages can in theory reproduce twice as fast as sexual ones (Maynard Smith 1978), raising a question about mechanisms compensating for the Ôcost of producing malesÕ (Maynard Smith 1978; West et al. 1999; Rice 2002). The areas of sexual selection and the evolution of sex have developed mostly in isolation, but recent theoretical advances have shown that sexual selection can have important consequences for the evolution of sex (Agrawal 2001; Siller 2001). Sexual selection leads to evolution of traits that increase the male reproductive success, such as weapons used in contests over access to females, or ornaments increasing male attractiveness to females (Darwin 1871; Andersson 1994). The latter mechanism has been subject to much debate, as it is not clear why females should show preferences for particular male phenotypes (reviewed in Andersson 1994; Kokko et al. 2003; Tomkins et al. 2004b). Zahavi (1975) proposed that such traits reveal male genetic quality, and therefore mate choice can provide choosy females with genetic benefits in terms of improved progeny fitness. The revealing function of sexual traits depends on their costs (Zahavi 1975; 1991; Grafen 1990). Costs of production and maintenance of sexual traits make their expression dependent on the phenotypic condition of individuals, and in consequence on genes that affect condition (Andersson 1982, 1986; Rowe & Houle 1996). As phenotypic condition is likely to be affected by many loci, condition-dependence of sexual traits may explain why these traits show considerable additive genetic variance (V a ) (Pomiankowski & Møller 1995), despite the expectation that sexual selection should deplete V a . Many loci imply that condition is a large target for mutations (Houle 1998), and therefore genetic variance in condition may be maintained by a continuous influx of deleterious mutations. Indeed, Houle & Kondrashov (2002) have shown that evolution of costly male displays and female preferences may proceed when genetic variance in condition is maintained solely by deleterious mutations. However, how male displays are influenced by mutation has been little investigated. Møller & Ecology Letters, (2004) 7: 1149–1154 doi: 10.1111/j.1461-0248.2004.00681.x Ó2004 Blackwell Publishing Ltd/CNRS