Using Natural Populations of Threespine Stickleback to Identify the Genomic Basis of Skeletal Variation
MetadataShow full item record
Across vertebrates, skeletal shapes are diverse, and much of this variation appears to be adaptive. In contrast, the early developmental programs of these structures are highly conserved across vertebrates. The question then becomes where in the conserved genetic programs of skeletal development does variation lie to direct diversity? In threespine stickleback, rapid changes in head and body shape have been documented during repeated and independent invasions of oceanic fish into freshwater habitats in regions deglaciated approximately 13,000 years ago. However, recent research indicates that similar phenotypic and genetic divergence can occur in decades. A remaining challenge is to link stickleback population genomic variation to causal genes that underlie such rapid phenotypic evolution. Here I use genome wide association studies (GWAS) in natural populations of stickleback to uncover genomic regions that contribute to variation of two dermal bone derived traits, lateral plate number and opercle shape. The decrease of lateral plate body armor and change in opercle bone shape, important for feeding mechanics, are classically associated with freshwater divergence. GWAS has recently begun to be used in natural populations but is still under scrutiny for performance among different populations. Using a population of phenotypically variable stickleback in Oregon, GWAS proved an effective method to uncover genomic regions and genetic variants known to contribute to lateral plate number and opercle shape, as well as new genomic regions and candidate genes not previously implicated in phenotypic variation. Although successful, using similar methods on decades old stickleback populations in Alaska revealed the challenges that accompany controlling population structure created by strong natural selection. Together, I found that although lateral plate number and opercle shape rapidly evolve in a coordinated fashion during adaptation from marine to freshwater environments, phenotypic variation is largely driven by independent genetic architectures. However, in very rapidly evolving populations, despite this independence of genetic architecture, the genetic variants contributing to the traits co-localize to similar genomic regions. This finding could be either biological or methodological which highlights the promise and limitations of using GWAS to identify genetic variation that gives rise to phenotypic diversity. This dissertation includes unpublished co-authored material.