My research focuses on the evolution and ecology of plant mating systems. In particular I am interested in the molecular mechanisms and genetic consequences of shifts from outbreeding mating systems to those that involve self-fertilization.
I am also interested in the conservation genetics of threatened and endangered plant species, such as the species of Leavenworthia which are endemic to the cedar glade habitats of middle Tennessee.
Other work in my lab focuses on the evolutionary maintenance of genetic polymorphisms such as the white/yellow/purple flower color polymorphism found in Leavenworthia stylosa.
Self-fertilization has evolved many times, in almost every plant family, yet the ecological and genetic factors underlying this transition are still unclear. My PhD work focused on the evolution of the mating system of Aquilegia canadensis, a plant which appears to be adapted to outcrossing via hummingbird pollination, but is actually highly self-fertilizing.
Self-fertilization may be advantageous if it ensures seed set when pollinators and/or mates are scarce (reproductive assurance), but may be detrimental if selfing decreases a plant's ability to outcross (seed discounting). Although reproductive assurance is often invoked as reason why self-fertilization should evolve, I have shown that the benefits of reproductive assurance are greatly outweighed by seed discounting and inbreeding depression in Aquilegia canadensis.
Are the high levels of selfing, and large fitness costs of inbreeding depression found in Ontario populations of Aquilegia canadensis typical of populations throughout the species range How do factors such as pollinator fauna, population size & density and other ecological factors vary across the species range? I compared populations in the center of the species range (Virginia and North Carolina) with Ontario populations which are near the northern periphery of the species range. Contrary to the general expectation, northern peripheral populations were not smaller or less dense than central populations, and both regions appeared to experience similarly low rates of pollinator visitation. Plants in central populations produce larger flowers, and dramatically larger herkogamy (the distance between anthers and stigmas), however, this marked geographic variation in key floral traits does not reflect evolutionary differentiation in the mating system. Populations in both regions are highly selfing, despite strong inbreeding depression.
Many traits, such as the separation of anthers and stigmas are known to have large effects on the mating system, and seem to be highly variable in populations of Aquilegia canadensis. Using quantitative genetic approaches, I examined the heritability of herkogamy to examine whether selection could operate on herkogamy to cause evolutionary change in the mating system.
My research as a postdoc in Lynda Delph's Lab investigated the conflict between natural selection and sexual selection on growth and reproduction in Silene latifolia. In this sexually dimorphic species, males produce many small flowers, and females produce fewer larger flowers. However, natural selection on life-history traits may constrain sexual selection on flower size and number. For example, floral traits and leaf traits tend to be genetically correlated such that plants producing many small flowers also produce thin leaves which are prone to water loss. Hence, natural selection imposed by water availability may influence floral trait evolution, and flower size and/or number may be sub-optimal in one or both sexes. We investigated this through several experiments, using both natural variation in these traits, and by extending this variation through artificial selection in the greenhouse.
We performed several experiments on artificial selection lines that differ in flower size and number (many small flowers vs. few large flowers). We planted these selection lines in the field to quantify fitness in terms of reproductive success (sexual selection for floral traits) and survival (natural selection on physiological traits). We also compared the performance of the selection lines in a greenhouse experiment in which we manipulated water and nutrient levels.