Preexisting Male Traits Are Important in the Evolution of Elaborated Male Sexual Display

The evolutionary cause of elaborate male sexual display traits remains controversial despite extensive recent research. R. A. Fisher is credited with developing the most widely discussed models: the good genes hypothesis (Fisher, 1915; see also Han~ilton and Zuk, 1982; Maynard Smith, 1976; Zahavi, 1975, 1977) and runaway selection (Fisher, 1930, 1954; see also Arnold, 1983; Heisler, 1985; Lande, 1981,1987). The good genes hypothesis has gained support from models showing how male traits and good genes preferences could coevolve (Houle and Kondrashov, 2002; Iwasa and Pomiankowski, 1999; Pomiankowski, 1987, 1988) and, most importantly, by strong empirical support (Goransson et al., 1990; Hasselquist et al., 1996; Hill, 1991; Hoikkala et al., 1998; Kempenaers r t a/., 1992; Moore, 1994; Norris, 1993; Partridge, 1980; Petrie, 1994; Reynolds and Gross, 1992; von Schantz et al.,' 1989; Welsh et al., 1998; Wilkinson et al., 1998). Similar strong empirical support is lacking for runaway selection (Ryan, 1997). Developing on a largely separate track has been preexisting preference (Burley, 1985) and related models (Basolo and Endler, 1995; Ryan and Rand, 1990). In these models, the females commonly have preferences for male traits that are not currently expressed in males. Males that appear with novel traits suited to that preference are selected by these females. These preexisting preference models differ from all other sexual selection models because the female preference evolves as a pleiotropic side effect rather than from the benefits of mate choice (Burley, 1985) and may involve maladaptive female preferences (Ryan and Rand, 1990). Preexisting preferences are not coevolutionary and do not require genetic correlations between traits and preferences. This is seen by some (Kirkpatrick and Ryan,