From Survival to Selection: The Genomic Landscape of Serpentine Adaptation in Monkeyflowers

Abstract

Adaptation to patchily distributed, physiologically extreme environments provides a powerful framework for studying the genetic architecture of local adaptation. Serpentine soils, characterized by skewed calcium-to-magnesium ratios, low nutrient content, poor water retention, and elevated heavy metals, impose severe selective pressures on plants, often leading to rapid and repeated evolution of tolerance traits. This dissertation investigates the genetic basis of serpentine soil tolerance in Mimulus guttatus and its close relative M. nudatus, integrating quantitative trait locus (QTL) mapping, experimental evolution, and population genomics.In Chapter 1, I map the genetic basis of serpentine tolerance in M. guttatus and M. nudatus using survival-based assays of F₂ mapping populations. Both species exhibit strong differential survival on serpentine soils, and remarkably, both share a major-effect QTL on the q-arm of chromosome 13. Despite this shared locus, additional minor-effect QTLs differ between species, suggesting a combination of genetic conservation and divergence. Sequence divergence analyses reject the hypothesis that recent introgression explains the shared major locus, instead supporting retention of ancestral variation or convergent evolution. These findings suggest that adaptation to serpentine soils may involve both shared and lineage-specific genetic solutions. In Chapter 2, I develop a modified Evolve-and-Resequence (E&R) framework to quantify fitness effects of serpentine tolerance alleles in M. guttatus. Starting with an F₂ population derived from a serpentine-tolerant and non-tolerant inbred line, I track allele frequency changes over two generations under selection in serpentine and control soils. Whole-genome sequencing of pooled progeny enables estimation of marginal fitness for tens of thousands of SNP windows. Results confirm strong selection on the chromosome 13 locus and reveal additional loci on chromosomes 5, 7, 11, and 14 that contribute to tolerance, many of which were not detected in survival-based mapping. This multi-generation experiment highlights the polygenic nature of adaptation and underscores the value of using comprehensive fitness-based approaches over survival alone. In Chapter 3, I assess how recombination rate and linked selection shape patterns of nucleotide diversity across Mimulus genomes. Despite theoretical expectations that diversity should be lower in regions of low recombination due to background selection, the relationship is weak in M. guttatus, likely due to the extreme baseline diversity levels and heterogeneous gene density across the genome. Gene-rich, high-recombination regions show very slightly elevated diversity, suggesting that gene density, not recombination alone, modulates the impact of linked selection. Collectively, this work shows that serpentine tolerance in Mimulus is shaped by both large-effect loci and a broader polygenic background, with evolutionary paths shaped by both shared ancestry and independent adaptation. By combining QTL mapping, experimental evolution, and comparative genomics, this dissertation provides new insights into how selection alters the architecture of the genome.