“…That is, studies have generally been designed to ask whether most eggs within the population are found on plant species on which offspring generally do best; the studies do not ask how oviposition preference in an individual female is related to performance of her offspring on the host she chose as compared with other hosts (e.g., Knerer & Atwood, 1973;Courtney, 1981;Rausher, 1982). These population-level studies have been important in testing whether factors other than larval growth rates and pupal masses can contribute to selection for oviposition preference (e.g., Smiley, 1978;Price et al, 1980;Atsatt, 1981a, b;Courtney, 1981;Rausher, 1981;Singer, 1983Singer, , 1984Williams, 1983). As Via (1986) recently emphasized, however, the evolutionary problem is now to understand how different genes for preference and performance are distributed among individuals, not as population averages.…”