Sexual Selection for Aesthetic Traits in Species with Biparental Care

Abstract

I hypothesize two processes by which sexual selection may generate differential reproductive success (RS) in species with biparental care. The differential-access hypothesis posits that an individual's own desirability as a mate affects its access to potential mates. By selecting mates strategically, desirable individuals enhance quantity and/or quality of offspring produced. The differential-allocation hypothesis states that an individual's own mating attractiveness affects the amount of parental investment it is able to secure from a mate. Organisms will provide above-average (for their sex) parental investment (PI) to obtain and/or maintain relatively attractive mates. The fitness of attractive individuals can be augmented through the production of offspring that receive above-average PI; alternatively, attractive individuals can restrict their own per-offspring investment and save reproductive effort for future use. In the latter case, attractive individuals will have longer life spans and higher RS than unattractive individuals of the same sex. These two hypotheses are not mutually exclusive, though some of the predictions they generate are competing. I investigated these hypotheses using the zebra finch (Poephila guttata), a typically monogamous Australian estrildid. Attractiveness of the birds was manipulated by placing plastic bands of various colors on the legs of birds. Previous experiments had established the relative attractiveness of these colors to opposite-sex zebra finches. In the banded-male experiment (BME), males were randomly assigned bands of red (attractive to females), or green (unattractive to females), or orange (neutral, the color of zebra finch legs); females were not color-banded. In the banded-female experiment (BFE), females were randomly assigned black (attractive to males), or blue (unattractive to males), or orange bands; males were not color-banded. The duration of the BME was 22 mo, that of the BFE, 15 mo. Results of both experiments indicated a stong effect of band color (attractiveness) on RS. In the BME, attractive males achieved approximately twice the RS of other color types. In the BFE, attractive females had twice as many offspring as unattractive females; females of intermediate attractiveness had intermediate RS. The differential-allocation hypothesis predicts that attractive individuals live longer than unattractive individuals of the same sex. This occurred in both experiments. Additional evidence supporting this hypothesis is found in the tendency of attractive males to become simultaneous bigamists. It appears that attractive males are sometimes able to transcend a behavioral polygyny threshold by restricting investment in the offspring of one female sufficiently to invest in offspring of a second female. Females monogamously mated to attractive males had significantly higher RS than polygynously mated females. Attractive females showed no parallel tendency toward bigamy, a finding compatible with the fact that female zebra finches typically incur higher parental expenditure than males. Not all the results of these experiments can be explained by the differential-allocation hypothesis alone. However, the results inconsistent with this hypothesis can be explained by the differential-access hypothesis or by the simultaneous occurrence of the two. In the BME, differential RS of males did not result solely from differential survival of color types. Among males that survived past the end of the experiment, attractive males had approximately twice the number of offspring as other males. In both experiments there were trends (not always significant) for the mates of attractive individuals to have high RS. Moreover, in both experiments mortality patterns of the sex that was not color-banded were intermediate between those predicted by the two hypotheses. These experiments demonstrate the existence of considerable potential for the operation of sexual selection in typically monogamous species. Moreover, they indicate that aesthetic or arbitrary (nonfunctional) mate preferences held by individuals of both sexes (not just females) are sufficient to affect the mate-getting abilities, mortality patterns, and RS of the bearers of traits of variable attractiveness.