THE POLLINATION BIOLOGY OF PANICULATE AGAVES: DOCUMENTING THE IMPORTANCE OF MALE FITNESS IN PLANTS.

Abstract

Recently, it was hypothesized that pollinator selectivity for large inflorescences has led to the evolution of monocarpic reproduction in agaves. To test this hypothesis, fruit set and stalk length data were collected for six species of paniculate agaves and two species of spicate agaves. Regression results, for all of the paniculate species and for one species of the spicate agaves, showed no significant correlation between fruit set and stalk length. These results do not support the pollinator selectivity hypothesis. The validity of the assumptions was then examined, utilizing the results from hand pollination, pruning, and tie down experiments. These results imply that factors other than pollinator availability might be important in determining fruit set in agaves. In an effort to determine the relative importance of resource and pollinator limitation to fruit set in Agave chrysantha, three treatments were applied to entire plants: (1) natural pollination, (2) natural and hand pollination, and (3) hand pollination. When fruits were collected, there were no significant differences between percent fruit set for the three treatments, implying that fruit set is not limited by the availability of pollinators. When additional plants were pruned to reduce the total number of flowers by approximately one-half, thereby doubling the amount of resources available to each flower, the percent fruit set was twice that for control plants, implying that percent fruit set is energy limited. It is common for plants that exhibit resource limited fruit set to have relatively low fruit-to-flower ratios. This is surprising, since it appears that the resources expended for production of these "excess flowers" could be allocated to fruit maturation and thereby increase fruit production. Four hypotheses explaining the fruit set in Agave mckelveyana are tested. Results from pruning, bagging, and hand pollination experiments indicate that the apparently "excess flowers" do not contribute to fruit production (female fitness). Additional data on nectar production imply that these flowers act primarily as pollen donors and contribute only to male fitness. The importance of pollen donation (male fitness) in determining fruit set is examined for hermaphroditic, monecious, and dioecious plants.