Potential constraints on evolution: sexual dimorphism and the problem of protandry in the butterfly Bicyclus anynana.

The earlier mean adult emergence between males and females, protandry, has been well studied mathematically and in comparative studies. However, quantitative and evolutionary genetic research on protandry is scarce. The butterfly, Bicyclus anynana exhibits protandry and here we selected for each of the different combinations of male and female development time in this species, thus including direct selection on protandry (i.e., FAST, fast males and fast females; SLOW, slow males and slow females; FMSF, fast males and slow females; and SMFF, slow males and fast females). After eight generations of selection there was no significant response for increased or decreased protandry, whereas selection for increased or decreased development time in both sexes (FAST or SLOW) was successful. Continued selection (> 30 generations) for decreased or increased protandry showed a significant difference between the FMSFC and SMFFC lines (subscript c for continued selection), which was of the same magnitude as the nonsignificant difference observed between the FMSF and SMFF lines at generation eight. This indicated that the initial selection was successful, but that the difference between the lines did not increase with continued selection. Our results also indicate that the genetic covariance across sexes for development time is near unity. Interestingly, lines selected for decreased protandry (SMFF) had lower egg-to-adult survival, and broods from these lines had lower rates of egg hatching. This suggests that interactions with fertility might constrain certain directions of change in patterns of protandry. Moreover, selection yielded a change in the ratio of male to female development time for slow lines, suggesting that some amount of sex-specific genetic variance for development time is still present in this population. The FMSFC line showed the largest effect of selection on protandry, mainly through an effect on female developmental time. Lastly, our results show that temperature has an effect on the amount of protandry in the selected lines. These results are discussed in relation to the ecology of this species and the evolution of protandry.