Flower sex ratio, pollinator abundance, and the seasonal pollination dynamics of a protandrous plant
Ecology Washington D C uary; 82(1): 127-144
The separation between staminate and pistillate phases in protandrous, long-lived hermaphroditic flowers results in a seasonal shift in sex ratio from a pure male to a pure female flower population. In theory, temporal sex ratio associated with dichogamy may affect pollination, promoting the evolution of sexual specialization in plants when pollinator visitation does not limit seed output. Alstroemeria aurea, a perennial herb native to the temperate forests of southern South America, bears strongly protandrous flowers. Over a three-year period, I assessed the influence of seasonal changes in flower sex ratio and pollinator abundance on the pollination dynamics of this plant species. Workers of Bombus dahlbomii accounted for [>]90% of all visits to A. aurea flowers and were the most efficient pollinators on a per-visit basis. As the season progressed, pollen loads carried by bumble bees decreased by almost one order of magnitude, and the amount of pollen deposited per visit decreased as well from [>]40 to [<]4 pollen grains from the beginning to the end of the flowering season. As flowering patches became more female, the number of male-phase flowers visited uninterruptedly during each bumble bee foraging bout decreased, while the number of sequential visits to female-phase flowers increased. A laboratory experiment of manipulating dried bumble bees showed that the number of male- and female-phase flowers visited in uninterrupted sequence strongly influences the amount of pollen deposited on a flower's stigma. Pollen deposition in this experiment closely mimicked the relationship between sex ratio and the observed seasonal decline in per-visit pollen deposition in the field. Because overall rate of bumble bee visitation increased monotonically over the flowering season, total pollen receipt per flower also increased during the first half of the season. Nevertheless, pollen deposition decreased during the second half of the season when visitation rates saturated pollen removal, and low pollen availability associated with an increasingly female-biased sex ratio apparently became the limiting factor in pollen transfer. Despite pronounced seasonal changes in pollinator and pollination levels, comparisons of seed output between open- and hand-pollinated flowers showed no evidence that seed production was ever limited by total pollen receipt. Because flower sex ratio influenced pollinator visitation and pollination patterns, and pollinator availability did not limit seed output, protandry should favor the evolution of female and male specialists flowering at the tails of the flowering season. Neither unisexuality nor any other form of sexual polymorphism is found in A. aurea, however. Likewise, I found no evidence that plants might be compensating for a seasonal shift in sex ratio through differential allocation into pollen and ovule production within flowers or to relative duration of male and female sexual phases. These results suggest that a seasonal shift in operational sex ratio within a dichogamous species can profoundly affect pollination without necessarily driving the evolution of floral polymorphism.