Quantifying Contributions of Climate, Geography, and Gene Flow to Divergence: a Case Study for Three North American Pines

Bolte, Constance

ABSTRACT: Long-lived species of trees, especially conifers, often display weak patterns of reproductive isolation, but clear patterns of local adaptation and phenotypic divergence. Discovering the evolutionary history of these patterns is paramount to a generalized understanding of speciation and the processes that confer population persistence versus those that compromise adaptive potential under rapidly changing environments. Forest trees have long generation times and low migratory potential making them especially vulnerable to population fragmentation and reductions of genetic diversity due to insufficient tracking of niche optima and adaptational lags. Within clades of the genus Pinus, evolutionary histories appear to be riddled with hybridization (i.e., interspecific gene flow), periods of isolation, and effective population size changes that co-occur with major shifts in climate. Quantifying the relative contributions of each of these factors to divergence and changes to genetic diversity requires a multidisciplinary approach involving historical species distributional modeling, demographic inference, and associations of genetic structure to climate and geography. This dissertation focuses on identifying drivers of divergence and explaining differing levels of reproductive isolation across three ecologically and economically valuable North American pine species (Pinus pungens, P. rigida, and P. taeda). First, we inferred demographic histories and found the recurrence of interspecific gene flow between P. pungens and P. rigida, as well as population size reductions during the last glacial period, to be important contributors to the mode and tempo of previously documented reproductive isolation between these species. Seasonality and elevation associated with both genetic and distributional differences indicating ecological divergence was also important to the divergences among the three focal species, but the relationship of P. taeda to the other two species remains enigmatic. Next, we illustrate how genomic patterns of differentiation across genic and intergenic regions can explain differing levels of reproductive isolation through pairwise assessments and mapping RADseq contigs to the annotated genome of P. taeda. Finally, in estimating the extent of hybridization and genetic diversity in shared forest stands of P. pungens and P. rigida, we discovered a general lack of hybridization at present and low genetic diversity in southern, trailing edge populations. Striking congruences across results, various methods employed, and work previously performed for the genus Pinus all provide support for emerging hypotheses related to forest tree speciation and biodiversity. This dissertation also presents useful information for forest conservation and management planning. At present, the adaptive potential of P. pungens, a montane pine with highly fragmented populations, is low based on genetic diversity estimates, its current distribution, and restricted levels of interspecific gene flow.