GENDER DIFFERENCES IN THE EFFECTS OF FLORAL SPUR LENGTH MANIPULATION ON FITNESS IN A HERMAPHRODITE ORCHID

Selection on floral traits in hermaphrodite plants can operate through either male or female sex function. However, selection is typically measured for the female component only, and when selection through male function is measured, it is usually done only in terms of pollen removal from flowers. We tested whether selection on floral spur length might differ between genders in the orchid Satyrium longicauda by measuring the effect of spur length manipulations on pollen receipt and export of color-labeled pollen to conspecific stigmas. This was done in two populations with different pollinator faunas to determine whether pollinator context influenced the relationship between spur length and gender-specific fitness components. In one population, spur length manipulation had opposite effects on sex functions, while in the other, both pollen export and receipt were reduced by spur shortening. Furthermore, measures of male function through pollen removal and pollen export did not always correspond. These results suggest that selection on floral spur length can differ between genders and that this gender bias may be affected by the pollinator context. Furthermore, interpretation of selection through male function based only on pollen removal can lead to erroneous conclusions. The context dependence of gender differences in selection could be an important factor generating geographically structured divergence in floral traits and, ultimately, speciation.