Abstract
Molecular population genetics is one of the fastest growing areas of biological research, being of central importance to understanding human history, diversity, and the potential for personalized medicine. The field uses the fact that historical processes such as change in population size, rates and pathways of migration, and natural selection, all leave distinct footprints in an organism’s DNA. Modern population genetics involves a rich arsenal of statistical tools that use DNA sequence data to make inferences about the evolutionary history of organisms and the genetic basis of their traits. My dissertation focuses on identifying the importance of selective and non-selective forces at different, hierarchically nested, biological levels in spatially structured (meta-)populations. In particular, my dissertation research explores how the evolutionary process is affected by the reality that populations of organisms are clustered in space. My dissertation research focused on a long-term metapopulation (25+ years now) of the angiosperm Silene latifolia and its sterilizing fungal pathogen Microbotryum lychnidis-dioicae. I have attempted to develop a multifaceted research program that includes fine-scale iterative sampling, where every individual within a extant population up to an expected genetic self-assignment asymptote of 50 individuals, from approximately 1/3 of 800 extant populations, distributed across three of nine metapopulation sub-sections: approximately 2000 individuals collected in 2008, 2010, and 2012. I have developed a high-throughput multi-plexed microsatellite genotyping protocol (20 microsatellites, from a panel of roughly 80 candidate loci, identified to encompass sufficient information for 1) assignment of individuals identified as colonists through an annual census, and 2) identification of recent migration amongst extant populations). Finally, I have worked to develop statistical genetic software that allow the use of molecular marker data derived from different genomic regions, even if they have very different evolutionary histories and rates of mutation.
This dissertation has derived a number of interesting inferences concerning the consequences of spatial population structure. Molecular marker specific evolutionary processes, and its concomitant variation, will determine both the type and accuracy of population genetic inferences available. The expectation of an inverse relationship between fitness and F will not always hold, and this discrepancy might be attributed to processes associated with population structure. Historical contingency will play a dominant role in determining the role of selective processes determining the quantity and distribution of neutral molecular variation, as well as asynchrony in the co-evolutionary process in spatially structured host-pathogen systems. Mean levels of cyto-nuclear disequilibrium are generally quite stable over up to seven generations of the focal plant species, though specific associations are quite labile and are highly affected by drift-like processes. And finally, intra-demic selection can be active in “everyday”, natural metapopulations, through hard selection acting on differential colonization, where different levels of selection are not necessarily in opposition to one another. Taken together, my dissertation aims to provide a unique window into the current and historical factors distributing population genetic diversity, but also how this structure will effect the evolution of the system itself. Continued long-term monitoring, iterative sampling, application of newer sequencing technologies, and the development of newer analytical methodologies will continue to show interesting biological dynamics taking place as the result of spatial population structure.
