| Many flowering plants are capable of reproducing by self-fertilization and outcrossing. The purpose of this study was to investigate the ecological and genetic significance of such a mixed mating system in both selfing and outcrossing populations of the winter annual plant, Leavenworthia crassa Rollins (Cruciferae).; Monitoring of natural populations and experimentation were used to study the following: genetic control of floral traits associated with breeding system; rates of outcrossing; patterns of natural fertilization, seed abortion and seed production; relative success of imposed natural, selfing, and outcrossing breeding systems; and female, male and interaction effects on reproduction. An experimental protocol for the study of non-random transmission in plants is also described.; In a phytotron experiment there was significant variation both within and among populations for flower size, and fruit and seed set in the absence of pollinators, indicating that there is a genetic component to such traits. Populations characterized by these traits had different outcrossing rates: a population with large scented flowers had an outcrossing rate of 0.25 whereas a population with small unscented flowers had an outcrossing rate of 0.09. Plants in outcrossing populations tended to abort more seeds and produced larger seeds than plants in selfing populations. There were non-random effects on ovule position within an ovary for fertilization, abortion and seed weight and these effects were stronger in outcrossing populations.; Natural pollination resulted in reproductive success equal to or greater than that of selfing or outcrossing, achieved primarily through lower seed abortion. The relative success of the three breeding system within a plant was not independent. Selfing and outcrossing populations showed similar responses to the three imposed breeding systems.; Maternal effects predominate throughout reproduction and appear to have both a genetic and environmental component. Paternal effects appear for fertilization and fruit set, and also appear to have a large environmental component. Female-male interaction effects on fruit set can not be explained by self-incompatibility or outcrossing depression. There is no evidence for mate choice in this species.; An experimental and analytical model for studying non-random transmission in plants is presented. Its two key components are (1) separation of reproduction into pollination, fertilization and seed maturation phases and (2) the application of crossing designs with several treatments for both males and females. Interpretation of possible results is discussed in light of sexual selection theory. |
