Sexual Variations Among Plants of a Perfect-Flowered Species

Abstract

Numerous environmental factors can influence the female and male reproductive success of hermaphroditic plants. These factors include, among others, the sexual status of neighboring plants, temporal changes in the number of seeds set in the population, and pollinator foraging. This study explores the interaction of such factors with the phenotypic gender of individual plants in a natural population of cardinal flower, Lobelia cardinalis. By measuring fecundity directly and estimating fecundation by assuming random mating, we show that the reproductive success of an individual is likely to be influenced by its phenotypic gender relative to others in the population, pollinator foraging, the timing of anthesis, temporal seed-set patterns in the population, and individual variation in seed and pollen production. Because the protandrous flowers are borne on an acropetally maturing raceme, the phenotypic gender of any individual plant begins as entirely male and changes to entirely female by the end of flowering. During the transition period from entirely male to entirely female, each plant exhibits considerable fluctuations in phenotypic gender. Moreover, the same trend holds for the entire population in which the percentage of open flowers in the staminate phase changes from 100% early in the season to 0% late in the flowering season but with unpredictable variations during mid-season. Mean seed set per day per pistillate-phase flower is also not constant during the flowering season. Much of the variation in daily seed set is attributable to the foraging behavior of the pollinator, the ruby-throated hummingbird. Supplementary hand-pollinations did not increase seed set relative to naturally pollinated plants, suggesting that seed production in this population is not limited by pollen reception. Considering this result and the environmental effects on male reproductive success, there appears to be greater certainty of success through the female than through the male reproductive functions. Although we have demonstrated that numerous environmental factors could partially obscure the relationship between a plant's phenotypic gender and its reproductive success through the female and male reproductive functions, our results indicate that the environment probably could not completely uncouple this relationship. This correlation between phenotypic and functional gender is likely because critical determinants of both gender estimates are differences among plants in seed and pollen production. These results are important because most theoretical studies of plant sexuality assume a strong correspondence between morphological and functional measures of sexuality. Our results, however, cannot be considered conclusive even for L. cardinalis plants; consequently, additional studies are needed.