Abstract
Circulating levels of testosterone in adults have mainly evolved as a consequence of selection on males for increased levels, while levels of circulating testosterone in females may be an indirect consequence of selection on males. A review of the literature revealed that intense directional selection for high levels of circulating testosterone in birds is likely to be mainly due to direct selection on males. A comparative study of testosterone levels in birds revealed a strong positive relationship between peak testosterone levels in adult females and peak levels in males. The slope of this relationship was significantly less than unity, implying that the testosterone levels in females have been reduced in species with high levels in males. An analysis of the order of evolutionary events suggested that peak concentration of testosterone in females changed after peak concentrations of testosterone in males. Females in colonial species of birds had significantly higher circulating peak testosterone levels compared to females of solitary species, and relative levels after controlling for the effects of peak levels in males were also larger, suggesting that any costs of high testosterone levels in females are particularly likely in colonial birds. Direct selection on male circulating testosterone levels may increase the costs that females incur from high testosterone titers. For example, high female levels may negatively affect ovulation and laying and may also affect the levels of testosterone that females deposit in their eggs and hence the exposure of pre- and post-hatching offspring to testosterone. This in turn may affect not only offspring behavior, but also offspring development and the trade-offs between growth, development of immune function, and behavior in offspring.