Although genetic variation in characters closely related to fitness is expected to either become depleted by selection or masked by environmental variation, "good gene" models of sexual selection require moderate to high heritabilities of secondary sexual characters to explain the occurrence of costly female mate preferences. In this study, I investigated whether the estimated heritability of a condition-dependent secondary sexual character (i.e., the white forehead badge) in the collared flycatcher varied depending on environmental conditions experienced during offspring growth. The data were collected over a period of 14 years making it possible to exploit natural variation in natal conditions. In addition, natal conditions were experimentally altered through brood size manipulations. During unfavorable conditions caused by generally poor weather or experimentally enlarged brood size, no significant heritability based on father-sons regressions could be demonstrated (0.19 less than or equal to h(2) less than or equal to 0.27). In contrast, sons reared during years with favorable weather or in experimentally reduced broods significantly resembled their fathers (0.44 less than or equal to h(2) less than or equal to 0.65). In addition, the heritability estimates declined with increasing maternal age. The strong effect of natal environmental condition on the estimated heritability of forehead badge size suggests that the potential genetic benefit from mate choice vary according to environmental conditions (e.g., the benefit is reduced during unfavorable rearing conditions). Because sons reared during poor conditions have probably experienced a natal environment different from that experienced by their fathers, the low heritability estimates obtained under poor conditions seem to be caused by low additive genetic variation expressed in such environments and/or a low genetic correlation between the expression of the trait in the two different environments (i.e., good vs. bad). Both of these explanations imply the presence of genotype-by-environment interactions. If such interactions frequently affect the expression of secondary sexual characters, this may offer an explanation of the high heritabilites sometimes reported for such traits, despite their exposure to long-term directional selection.
