498 0169-5347/98/$ – see front matter © 1998 Elsevier Science. All rights reserved. PII: S0169-5347(98)01513-4 TREE vol. 13, no. 12 December 1998 T he evolution of sexually selected traits is driven by contests over mating op- portunities and/or by mate choice. In most species, males can increase their reproductive suc- cess by increasing their number of mating partners, whereas fe- males increase their reproductive success mainly by improving the quality of their mates. Conse- quently, females are usually more selective about their mates than are males. In this article we will, for simplicity, discuss only female choice and male–male compe- tition, although males occasion- ally constitute the more choosy sex and females the more competi- tive one. The issues we discuss will, however, apply to either case. It is generally thought that win- ners of male–male competition are of superior quality and that it would be in the females’ interest to mate with these males 1–3 . Thus, dominance per se or traits re- flecting it, such as large body size, heavy weaponry and intense sig- nals of fighting ability (such as sta- tus badges; Box 1), are expected to be important cues in female choice. Indeed, female preference for dominant males has been found in many species and, occasionally females even incite competition between males and then mate with the winner 3 . However, although it is often assumed that fe- males gain by mating with dominant males, few studies have actually investigated the fitness consequences. Furthermore, it is important to bear in mind that there are several possible ways in which the two processes of sexual selection can interact (Fig. 1). An increasing number of studies are showing that domi- nant males are not always pre- ferred by females (Table 1). Here, we summarize some of the pres- ent evidence of the main benefits and costs of choosing dominant males. Direct benefits and costs Do dominant males provide high quality resources? In many species where males compete for resources necessary for attracting females (e.g. suit- able breeding sites), dominant males monopolize high quality re- sources and consequently enjoy a mating advantage 2 . However, in polygynous species, dominance per se can be a poor predictor of the breeding situation for a spe- cific female because she might have to share resources with other females. In such cases, other cues, such as territory qual- ity and number of females already settled, can be important for fe- male mating decisions 4 . Thus, when resources provided by males are crucial for female reproduc- tive success, females should base their mate choice on the quality of the resource provided directly, rather than on male dominance position (i.e. the ability to monopolize re- sources). Several experimental studies have shown that females seem to base their mate choice mainly on the qual- ity of the territory or nest site 5–7 rather than on male traits. Although female attraction to essential resources could be the reason why males compete for access to such re- sources, male dominance per se need not necessarily be used as an important cue in female choice. Do dominant males provide better parental care? In organisms with paternal or biparental care, the qual- ity of male care can be important for the number and qual- ity of offspring produced. Females might therefore in- crease their fitness by selecting good parents 8 . It is often assumed that dominant males are of higher overall quality and that they can consequently provide superior paren- tal care, as was demonstrated in a freshwater goby (Pado- gobius martensi ) 9 . However, the results of a recent study show that in the closely related sand goby (Pomatoschistus minutus), winners of male–male competition did not pro- vide better parental care than losers 10 . Furthermore, in the sand goby, females did not prefer winners of male con- tests, but instead preferred males that provided good pa- ternal care. As a result, these females managed to bring a larger proportion of their eggs to hatching 10 . Therefore, traits important in male–male competition might not always be the same as those that are important to females selecting a mate. REVIEWS Should females prefer dominant males? Anna Qvarnström and Elisabet Forsgren It is generally believed that success in male–male competition genuinely reflects high quality and that female preference for dominant males should therefore be widespread. However, recent studies suggest that male dominance is not always attractive and that it does not necessarily predict superior parental quality, better genes or other forms of benefit to females. In fact, the costs of choosing a dominant male can sometimes outweigh the benefits. When traits selected by male–male competition do not reflect overall mate quality, females are expected to use other choice cues and might occasionally prefer subordinate males. Thus, male–male competition and female choice can sometimes work in different, or even opposing, directions. Anna Qvarnström is at the Dept of Zoology, Uppsala University, Villavägen 9, S-75236 Uppsala, Sweden (anna.qvarnstrom@zoologi.uu.se); Elisabet Forsgren is at the Dept of Marine Ecology, Göteborg University, Kristineberg Marine Research Station, S-45034 Fiskebäckskil, Sweden. Box 1. Causes and consequences of dominance ‘Dominance’ can be defined as success in contests. By killing, driving away or using other means of intimidating individuals, dominant individuals exclude at least some of their rivals from access to mates or resources crucial for attracting mates. Dominance hierarchies are often settled by relative body size, aggres- siveness, size of weaponry or signals of fighting ability (badges of status). Such morphological and behavioural traits are costly because they can cause increased predation risk, increased energy stress and/or increased disease susceptibility. If it is only the individuals of relatively high quality that are able to bear the cost of dominance 37,38 , the position in the hierarchy per se or traits indicating dominance will reliably reflect certain aspects of mate quality. However, because motivation (i.e. how much an individual values the con- tested resource) is also important, the best fighter will not always win a contest. Thus, dominance hierarchies could be unstable: an individual that is dominant in a social contest (e.g. successful in fights over food or resting sites) may not necessarily be successful in contests over mates. Furthermore, because indi- viduals value their resources increasingly with their time of ownership, previ- ous site knowledge and order of occupancy will affect the outcome of territorial disputes.