95th ESA Annual Meeting (August 1 -- 6, 2010)

COS 72-7 - Modeling effects of pollinator behavior on pollen limitation and hermaphrodite selfing rate in gynodioecious populations

Wednesday, August 4, 2010: 3:40 PM
412, David L Lawrence Convention Center
James David Stone, Biology and Wildlife, University of Alaska Fairbanks, Fairbanks, AK and Yoh Iwasa, Department of Biology, Kyushu Univerity, Fukuoka, Japan
Background/Question/Methods

Sex ratios of gynodioecious plant populations, where females and hermaphrodites coexists, depend on several genetic and ecological factors.  Key ecological factors include hermaphrodite selfing rate and the degree of pollen limitation in females.  Both of these may be pollinator-mediated, yet theoretical models of the maintenance of gynodioecy explicitly incorporating pollinator behavior are lacking.  We use analytical models to explore how changes in pollinator behavior influence hermaphrodite selfing rates and pollen limitation in gynodioecious populations.  We also investigate how changes in sex ratio affect the relationship between pollinator behavior, selfing rate, and pollen limitation.  The model is used to predict the influence of pollinator behavior on the relative fitness of hermaphrodites and females and its long-term influence on population sex ratio.

Results/Conclusions
Our results formalize the relationship between pollinator behavioral parameters, selfing rate, and pollen limitation in simplified gynodioecious populations.  Hermaphrodite selfing rate increased monotonically with some parameters of pollinator behavior such as pollen load decay, increased length of pollinator visitation, female-biased sex ratios, and when pollinators preferred females. On the other hand, the number of pollinator visits did not affect selfing rate, as long as hermaphrodites did not engage in autonomous selfing.  Outcross pollen limitation in either sex was highest when pollinators preferentially visited the other sex and when bout lengths were a large fraction of the total inflorescence.  If pollinators preferred one sex, their disparity in siring success was highest when pollinator bout lengths and the number of pollinator visits were intermediate.  This region of the parameter space corresponds with strong selection for the preferred sex.  Frequent and long visitation bouts saturated both sexes with pollen, regardless of pollinators’ preference for a particular sex.  This corresponded to weaker advantage for the preferred sex.  Increased pollinator preference for a particular sex expanded the sensitive region of parameter space corresponding to disparate fitness over a broader range of parameter values.  We show that the explicit modeling of pollinator behavioral parameters can clarify pollinators’ role in the maintenance and evolution of plant mating and breeding systems.