Kalisz 9726980 A plant's mating system (the degree to which it outcrosses rather than self pollinates) is a vital determinant of genetic variability at the population level. While theoretical studies have demonstrated that either complete outcrossing or complete selfing are stable evolutionary states, mixed mating (whereby progeny are sired by a mixture of self and outcross pollen) is not as easily explained. Yet in nearly half of the animal pollinated plants studied, intermediate outcrossing rates (those between 20-80%) are observed. The current study utilizes the winter annual, Collinsia verna, the Blue-eyed Mary, as a model organism to examine the consequences of variability of pollinator activity over the flowering season and its influence on the evolution of self pollination as a back-up strategy (i.e., reproductive assurance selfing). Using clonal replicates of individual plants in a series of integrated field and controlled environment experiments, the correlations between specific floral attributes conferring delayed selfing, fruiting success via self pollination, and the degree of inbreeding depression expressed by those individuals will be quantified. These measurements will be interpreted in the light of a novel field bioassay which allows us to quantify the frequency of pollinator failure from the perspective of paired flowers which are experiencing common resource conditions. Field estimates of pollinator failure and selfing ability will be compared with estimates of the natural outcrossing rates derived for three populations across the flowering season and across two years. The proposed research addresses the links between selective pressures on delayed selfing via pollinator mediated pollen limitation, the expression of mixed mating, and inbreeding depression expressed in natural populations of Collinsia verna. The actual, physical process of reproductive assurance selfing has been characterized for very few species. Q uantifying the timing of self-pollination within a flower, and the associated morphological traits will address an important issue in evolutionary ecology: how developmental processes (like the timing of anther and pistil elongation) contribute to selfing rates and how they evolve. Moreover, this research will be the first to test the mixed mating model of Schoen and Brown (1991) which indicates that when environmental variation in pollinator availability exists, reproductive assurance is more important in driving the evolution of a mixed mating system than is inbreeding depression. As a result, this work will be of interest to plant geneticists, crop breeders and conservationists who work with species which may become pollen-limited due to unusual weather patterns or regional epidemics in key pollinators (e.g. as has happened recently to honeybee populations).
