| Variation in population sex ratio and size may be the result of pollinator behavior, parental fitness, and germination differences as a result of the relative fitness of the sexes. Populations of the gynodioecious Lobelia siphilitica vary in sex ratio from 0% female to nearly 100% female and in size from a few individuals to several hundred. Understanding what causes population sex ratio and size variation can help explain how gynodioecy is maintained through differential fitness of the sexes, which is measured by comparing the maternal fitness of female plants to the maternal fitness of hermaphrodite plants. Both pollinator behavior and germination success may be altered by variation in population size and sex ratio. In experimental populations of 30 and 90 plants, 10% and 80% female, I found that large arrays received higher total pollinator visitation, though individuals in small arrays showed higher levels of multiple paternity. Additionally, individuals in the hermaphrodite-biased arrays received more total pollinator visits and produced a higher number of seeds per fruit. Finally, plant sex influenced total seed production, where female plants overall produced more seeds than hermaphrodites, suggesting a mechanism for female maintenance. To determine the effect of population sex ratio, size, and levels of inbreeding, which may influence the sexes differently, I measured germination in seeds collected in natural and experimental populations, and in crosses of varying degrees of inbreeding. I found that seeds collected from small natural populations had higher germination than those in large, and no variation in germination was explained by population sex ratio. In seeds collected in experimental populations, I found no effect of population sex ratio, size, or plant sex on proportion germinated seeds. In the manipulated crosses, I found no effect of inbreeding on germination, though females consistently produced heavier seeds. These results suggest support for the maintenance of females in gynodioecious populations as well as an explanation for why large, highly female populations of L. siphilitica are not found in nature. |