History, chance, and adaptation: The evolution of Silene vulgaris in its native and introduced ranges

Species expanding into new ranges will experience stochastic effects of colonization and must either be preadapted or evolve adaptations to survive and reproduce in a novel environment. Biological invasions provide natural experiments for investigating these issues. This dissertation used the 200 year old invasion of North America by Silene vulgaris, a weedy plant native to Europe, to address how invasions are impacted by (1) evolutionary history in the native range, (2) the genetic composition of the invasion inoculum, (3) the match in ecological niche between ranges, and (4) post-invasion evolution. A phylogenetic investigation of chloroplast DNA supported a history of range expansion in Europe from Mediterranean glacial refugia. However, there was no detectable signature of isolation among native range geographic regions in this locus. In a multi-locus analysis of the nuclear genome, amplified fragment length polymorphisms (AFLP) revealed five divergent demes that were regionally structured in Europe. Both cpDNA and AFLP supported a genetically diverse inoculum, probably the result of multiple introductions. Founder effects were evident in shifted deme frequencies and in the mismatching of genotypes from the ecological niche predicted by demes in the native range. As the invasion progressed, S. vulgaris expanded its range to fill and eventually exceed the predicted niche, consistent with a scenario of adaptive evolution. This hypothesis was tested by planting 1600 individuals from both ranges into common gardens in Ontario and Virginia. North American genotypes were genetically divergent for several traits, including having a faster reproductive maturity, and greater overall fecundity and survival compared to European genotypes. Phenotypic clines with latitude were also evident among families collected from both ranges. Incorporating null expectations using AFLP confirmed that divergence between and within ranges was due in part to post-invasion evolution. However, demes themselves also differed in performance regardless of continent of origin. Coupled with the shift in deme frequencies due to founder effect, this indicated that total phenotypic divergence resulted from a combination of adaptive and stochastic processes. Collectively, these results demonstrate that evolution during invasion is a multi-dimensional process, affected by prior evolutionary history, chance sampling events, and adaptation to the introduced environment.