Dr. John K Kelly

A common but incorrect belief: Biological evolution occurs too slowly to be observed. Consequently, most applications of evolutionary theory are historical in nature. A theory is tested by comparing its predictions to extant patterns of variation in nature, either within or across species. These tests can be convincing, but the most direct way to test any scientific theory is to develop an explicit prediction for what should happen and then see if this does happen. In situations where the genetic composition of a population changes at a rapid pace, it is possible to directly test the dynamical predictions of evolutionary models. One of my primary research interests is to construct and test models that predict observable changes in the genetic composition of populations. 

Our research group investigates a number of questions in evolutionary genetics, but the most basic is why everyone is different from everyone else. Extensive individual variation is the rule in nature. Oftentimes, a substantial fraction of this variation has a genetic basis. If natural selection tends to favor an optimal genotype, then why do so many different genotypes persist? We are adapting modern genomic technologies to address this classical question. Most of our experimental work (field and laboratory) focuses on the plant species Mimulus guttatus, but we collaborate on projects involving insects, mice, and a great diversity of other plant species.