Investigating Phylogenetic Relationships, Hybridization, and Evolutionary Origins in Boechera and Related Genera (Brassicaceae)

Abstract

The angiosperm genus Boechera of the mustard family Brassicaceae comprises over 100 recognized species that are most diverse in western North America, but also extend to eastern North America and the Russian Far East. The development of genetic and genomic tools in Boechera has been greatly facilitated due to its small genome (~200Mb) and close evolutionary relationship with Arabidopsis, making it the subject of intense research over the last two decades. Boechera has also become a model system for studying ecological and evolutionary genomics as well as for the study of apomixis. All previous phylogenetic studies have had limited success in resolving species relationships within the genus. Although Boechera is estimated to be only 2.5 million years old, it has diversified to form ~80 sexual diploid species, making it one of the most rapid diversifications recorded among plants. Boechera and the closely related genus Phoenicaulis are the only lineages of flowering plants known to produce apomictic diploid taxa, which have formed repeatedly whenever sexual species have encountered one another. These apomictic diploids account for approximately 20% of the diversity within Boechera. In turn, when these apomictic diploids encounter sexual species of Boechera they readily expend their unreduced pollen, resulting in the formation of apomictic triploids (accounting for an astounding 60% of Boechera diversity). Together, these crossing events have resulted in the formation of more than 400 documented apomictic hybrids and a profoundly complex taxonomy. This dissertation, by utilizing rigorous phylogenetic methods, sets out to advance our understanding of the evolutionary relationships of Boechera and the tribe Boechereae, including the origins of intergeneric and intrageneric hybrids. Chapter 2 utilizes the recent effective application of sequence data from target enrichment approaches to resolve the evolutionary relationships of Boechera and closely related genera. Utilizing the Hybpiper pipeline to analyze two combined bait sets: Angiosperms353 (with broad applicability across flowering plants), and a Brassicaceae-specific bait set (designed for use in the mustard family), relationships for 101 samples representing 81 currently recognized species were inferred from a total of 1114 low-copy nuclear genes using both supermatrix and species coalescence methods. The analyses resulted in a well-resolved and highly supported phylogeny of the tribe Boechereae, which is divided into two major clades, one comprising all western North American species of Boechera, the other encompassing the eight other genera of the tribe. This understanding of relationships within Boechera resulted in the recognition of three core clades that are further subdivided into robust regional species complexes. This study is the first broadly sampled, well-resolved phylogeny for most known sexual diploid Boechera. This effort provides the foundation for a new phylogenetically-informed taxonomy of Boechera that is crucial for its continued use as a model system. Chapter 3 explores the origin of Boechera tiehmii, a putative intergeneric hybrid between Boechera lemmonii (widely distributed in alpine habitats from Alaska to California) and Nevada holmgrenii (endemic to the western Great Basin of Nevada). Boechera tiehmii is quite rare, currently known only from the central Sierra Nevada in California and the Carson Range in Washoe County, Nevada. To explore this hypothesis, this study leverages a unique methodological approach that integrates results from 15 microsatellite loci from a dataset of 5000+ samples from the Boechera Microsatellite Website (BMW), a phylogeny of sexual diploid members of Boechera and closely related genera inferred using 1114 phased Hyb-Seq loci, and detailed morphological observations. This study supports the hypothesis that B. tiemhii is indeed an intergeneric hybrid between B. lemmonii and N. holmgrenii. The placement of phased hybrid tips within the sexual diploid phylogeny is unequivocal, the microsatellite loci have near perfect additivity, and B. tiehmii is shown to be morphologically intermediate between its proposed parents. By affirming the hybrid nature of B. tiehmii, this work contributes to a deeper understanding of the evolutionary processes that shape biodiversity within the Boechereae tribe. It highlights the importance of applying a broad range of analytical tools when addressing taxonomic ambiguities in groups where hybridization is rampant. Chapter 4 builds on chapter 2 by including seven phased apomictic Boechera taxa in a phylogenetic analysis together with the 91 sexual Boechereae dataset already available. A modified version of HybPhaser was developed here and used to phase Hyb-Seq loci from two baits sets, Angiosperms353 (with broad applicability across flowering plants) and a Brassicaceae-specific bait set (designed for use in the mustard family) from seven hybrid samples across all 1114 loci. Phylogenies were constructed with ng-RAxML and wASTRAL and resulted in the successful phasing of 1114 loci for all seven putative hybrid or apomictic taxa. This study resulted in a phylogeny that incorporated 14 phased hybrid tips into an existing sexual diploids-only phylogeny. Combining phased hybrid loci with an existing phylogeny demonstrates not only the utility of phasing multiple bait sets but also displays the complicated evolutionary history of Boechera. By exploring the mechanisms of hybridization and the implications of reproductive strategies, we aim to uncover the underlying genetic and ecological factors that contribute to the remarkable diversity observed in Boechera and its related groups. The ongoing evolution of Boechera exemplifies the significant roles of hybridization and apomixis in shaping biodiversity and adaptive landscapes within flowering plants. Chapter 5 is a study of Boechera pusilla––an extremely rare mustard taxon known from a single population located on a metamorphic rock outcrop in Fremont County, Wyoming. To investigate the evolutionary history of the Fremont County rockcress, we conducted a cytogenetic analysis, as well as morphological and microsatellite genotype studies of nearly all available herbarium collections. Our cytogenetic analyses revealed B. pusilla to be an apomictic triploid (n = 2n = 21), and microsatellite analyses showed that its triploid genome comprises B. oxylobula (restricted to west-central Colorado at elevations of 2000–3200 m), B. lemmonii (widespread alpine species generally occurring above 3000 m in the southern Rocky Mountains), and B. “wyomingensis”, an undescribed sexual diploid largely confined to southern Wyoming at elevations ranging from 1600–2800 m. Given the current distribution ranges of these taxa, the most parsimonious explanation for the origin of B. pusilla involves an initial hybridization event between B. oxylobula and B. lemmonii in western Colorado––an apomictic diploid bridge taxon that migrated north into Wyoming––followed by a subsequent hybridization with B. “wyomingensis” to produce the Fremont County rockcress. Our data suggests that the rarity of B. pusilla is more likely due to its recent origin, rather than to any inherent genetic limitations. In a world with limited conservation resources and given that there are scores of similarly formed apomictic hybrids in Boechera, we propose that the future may be better served by using these resources to protect the rarer, sexual diploid taxa.