Student Projects from FSU, Winter 2000
The Timing of Open (Out-crossed) and Closed (Selfed) Flowers in Violets
Elizabeth Boyd

Viola septemloba is a perennial plant with a specialized breeding system in which it produces two distinct flower types. The chasmogamous flowers are typical flowers that open and have attractive structures and rewards for pollinators. The cleistogamous flowers are much smaller flowers that self fertilize without ever opening. Previous work has shown that the cleistogamous flowers are much less expensive to produce then the chasmogamous flowers and that the progeny of the cleistogamous flowers do not suffer from inbreeding depression. Knowing this, the question is why does the plant maintain a mixed breeding system that includes chasmogamous flowers? There must be some fitness advantage associated with the seeds produced by chasmogamous flowers. This fitness advantage may be achieved through one of two routes. The short term selection hypothesis is that seeds from chasmogamous flowers have an immediate fitness payoff in terms of better performance than their cleistogamous counterparts. The long term selection hypothesis is that seeds from chasmogamous flowers have more variation in genotype than seeds from cleistogamous flowers (which, resulting from selfing, are identical), and this variation will allow them to perform better than their cleistogamous counterparts in situations of environmental variability. The model predicts the values of the fitness advantage of chasmogamous flowers that are necessary to maintain mixed breeding for given probabilities of these two types of selection.

CybeRose note: Seeds from cleistogamous flowers are not likely to end up far from the seed parent. Those from chasmogamous flowers aren't much more likely to stray, unless birds (?) carry off the pods. It would be helpful to know whether the pollen grains and seeds differ between the two types of flowers. Differences in nutritient supply contributes at least a little to what we think of as inbreeding depression and hybrid vigor. Experiments with co-pollination have shown that inbreeding depression can be prevented by the use of pollen that cannot fertilize ova in the target flower. E.g., sunflower pollen on maize. In addition, primitive plants (e.g., liverworts, mosses) spend most of their life-cycles in the gametophyte "generation" which cannot exploit heterosis. They don't suffer from inbreeding depression any more than the violets mentioned. Another question: do the progeny of the chasmogamous flowers of the same plants suffer inbreeding depression?

See: Knight & Waller (1987) for somewhat contradictory results with Impatiens capensis.