The signalling function of eyespan in stalk-eyed flies (Diptera: Diopsidae)

Genetic models of the handicap theory of sexual selection propose that ornaments signal heritable male quality, so females mating with the most ornamented males acquire fitness benefits for their offspring. Male ornaments are predicted to have coevolved with female preference to be larger, and so more costly. The key prediction made by the handicap hypothesis is that male sexual traits have evolved heightened condition dependence, a result of the higher differential cost of ornaments relative to other traits. I investigated evidence for condition-dependent sexual ornaments and found little support from well-designed experiments. Most studies had neglected to 1) compare condition dependence in sexual traits with suitable non-sexual controls, 2) adequately account for body size variation, and 3) assess individuals under a range of stresses representative of those experienced in nature. There was also a dearth of experimental studies exploring the genetic basis of condition dependence. I used experiments with stalk-eyed flies to examine predictions made by condition-dependent handicap models of sexual selection. Cyrtodiopsis dalmanni is highly sexually dimorphic for eyespan, and females exhibit strong mating preferences for males with large eyespans. Condition was varied experimentally by manipulating larval food availability. I found that male eyespan was more sensitive to changes in condition than female eyespan and other non-sexual traits. Male eyespan also showed a great increase in standardized phenotypic variance under stress, unlike non-sexual traits. These patterns persisted before and after controlling for body size. In contrast, there was no heightened condition dependence of male eyespan in Sphyracephala beccarri, a species without female mate choice for exaggerated male eyespan and only minor sex differences in eyespan. The genetic basis of ornament condition dependence was investigated in C. dalmanni by comparing the performance of distinct genotypes (inbred lines) along a gradient of environmental stress. Lines that produced a large ornament in one environment tended to do so in others. Stress also amplified these differences between genotypes leading to an increase in the genetic variance of the male ornament. Such patterns were less marked in non-sexual traits, and persisted after controlling for size. I looked for positive correlations between ornaments and viability by assessing the genetic correlations between male eyespan expression and four components of fitness (male fertility, female fecundity, and male and female longevity). I found no evidence that females obtain genetic benefits, other than male attractiveness, for their offspring by mating with well-ornamented males. However, body size-corrected male eyespan was negatively correlated with female longevity. This was unexpected and does not provide support for "good genes" benefits of sexual selection. Possible reasons for such findings (or lack thereof) are discussed.