Androdioecy in white mangrove (Laguncularia racemosa): Maintenance of a rare breeding system through plant-pollinator interactions

Androdioecy is a rare breeding system with two breeding types: male-flowered plants and hermaphroditic-flowered plants (with both male and female function). The question as to how males can be maintained in populations with hermaphrodites has resulted in mathematical models that address the conditions necessary to maintain males and that predict the frequency of males. I constructed a new model for androdioecy, based on evolutionarily stable strategy (ESS) assumptions, that incorporates pollinator foraging behavior. I applied the model to several Floridian populations of insect-pollinated White Mangrove (Combretaceae: Laguncularia racemosa (L.) Gaertn. f.). I estimated model parameters with data from pollination studies and pollinator observations, specifically addressing relative pollen fecundity, inbreeding depression, outcrossing rate, and relative pollinator visitation rate. A fitness advantage is conferred on male plants through increased pollen fecundity, but there are no significant differences in fitness between males and hermaphrodites in terms of fruit set, seedling emergence, or seedling survival. Inbreeding depression was stronger, and outcrossing rates were higher, in androdioecious populations relative to hermaphroditic populations. Several insect species, including Apis mellifera (honeybee), significantly depressed outcrossing rates in androdioecious populations, but only had a marginal effect on the outcrossing rates in hermaphroditic populations. In general, pollinators visited male plants more frequently than hermaphroditic plants, which increased male fitness relative to hermaphrodites in androdioecious populations. Male frequency predictions generated by my model are in agreement with observed male frequencies in three of the four populations tested. This suggests that males may have reached equilibrium frequencies and that androdioecy can be considered an evolutionarily stable strategy in these populations. According to model predictions, conditions in the fourth population cannot support the observed male frequency (0.42); males are predicted to decline until an equilibrium frequency is achieved (< 0.10).