The evolution and functional significance of stigma-height dimorphism in flowering plants

Genetic polymorphisms, have been widely used as model systems for evolutionary studies of adaptation and natural selection. This thesis investigates a distinct plant sexual polymorphism—stigma-height dimorphism (SHD)—and is the first comprehensive analysis of the nature of SHD, its taxonomic distribution, and its evolution and functional significance in animal-pollinated plant populations.; SHD is characterized by discrete variation in stigma height with anthers positioned similarly in flowers of long (L)- and short (S)-styled morphs. The lack of reciprocity between stigmas and anthers in species with SHD distinguishes it from the well-known sexual polymorphism heterostyly. I compared floral morphology, incompatibility type and population morph-ratios of taxa with SHD vs. distyly. The systems typically differ in incompatibility status and morph ratios, yet SHD, like heterostyly, probably functions to increase proficient pollen transfer between floral morphs.; I examined features of SHD in populations of the Mediterranean geophyte Narcissus. SHD is widespread in the genus making it valuable for studying questions related to the evolution and function of the dimorphism. Self-incompatible populations of N. assoanus had morph ratios ranging from isoplethy (1:1) to L-biased, while populations of self-compatible N. dubius were L-biased or monomorphic-L. There were no morph-specific differences in the species in outcrossing rates or female fertility suggesting that differences in pollen transfer and levels of assortative and disassortive mating contribute to the variable morph ratios. A computer mating model provided evidence supporting this hypothesis.; To investigate the role of SHD in the evolution of distyly, pollen transfer and male fertility were examined in experimental populations of the annual herb Eichhornia paniculata. Pollen transfer between morphs was typically greater than among individuals of the same morph. Mating patterns partially supported the validity of the Lloyd & Webb (1992) model of the evolution of distyly with SHD as an intermediate stage. The conditions of the model were satisfied when outcross transfers were considered but were met less often when selfing was included. While SHD seems likely to have been an inter-mediate step in the evolution of heterostyly in some groups, its occurrence in non-heterostylous families indicates that this is not always true.