Pollination, seed predation and floral character evolution in Silene latifolia.

摘要

Current theories of floral evolution are based on the assumption that floral characters influence offspring production, either through maternal (ovules) or paternal (pollen) success through the attraction of pollinators. However, flowers may be subject to seed predation by larvae of pollinators as in the case of Silene latifolia in its native European habitat, where moths of the genus Hadena pollinate the flowers as well as oviposit onto the ovary of the female flowers. To assess the relationship between floral characters, pollination and seed predation, I used three approaches. First, the impact of female floral morphology on female reproductive success was examined in two natural populations of Silene latifolia growing in Spain (Chapter 2). This was accomplished through comparing seed number with floral characters. This revealed selection on flower number only. A comparison of flowers that were pollinated with those which were not revealed that pollinated flowers had a significantly lower average height than non-pollinated flowers, and that pollinated flowers had a significantly larger average corolla diameter. There were no statistically significant differences between predated and non-predated flowers. However, there were several trends in the data, which suggest that floral characters may be important for seed predation. Secondly, male reproductive success was evaluated in terms of its relationship with floral morphology and location within one natural population over two years (Chapter 3). This was done through the use of marker-based paternity analysis using maximum likelihood models to estimate effects of floral characters. There was stabilizing selection on flower length in one year, and corolla diameter in the other. Finally, I grew an experimental, greenhouse population of S. latifolia and subjected the plants to differing levels of pollination and seed predation (Chapter 4). Pollination and seed predation both impacted the pattern of flower production, the number and size of flowers, and the life span of non-pollinated flowers. I conclude that while pollination and seed predation have impacts on the fitness of plants, the relationships between fitness and floral morphology is inconsistent between years and sites, and therefore may not lead to evolutionary changes in floral morphology.