Ancient Genomic Analysis of Wild and Domestic Sheep

Abstract

This thesis examines the natural history of domesticated sheep using ancient DNA data. The domestic sheep (Ovis aries) is a descendant of the wild sheep species mouflon (Ovis gmelini) found in Europe and Asia. Domestication is thought to have happened in Southwest Asia approximately ten thousand years before present. Data has been generated by extracting and sequencing DNA from subfossil remains. Chapter 1 briefly introduces domestic sheep and two wild sheep species. The domestication of sheep is also discussed, along with archeology relevant to samples analysed as part of this thesis. Previous ancient DNA studies of sheep are also considered to review the current state of the field. Chapter 2 details the generation of data. Methodologies used to prepare, extract, amplify and sequence ancient DNA are detailed. The results of subfossil screening are given, in which species are determined along with the percentage of DNA that is endogenous to the sample. The results of deeper sequencing efforts are also recorded, summarising the mitochondrial and genomic datasets generated. The efficacy of mitochondrial capture is also detailed and this technique was used to generate part of the mitochondrial data set. Chapter 3 focuses on mitochondrial analyses of ancient sheep. Previous studies on both modern and ancient sheep mitochondria are discussed to give perspective. Phylogenies containing both ancient and modern sequences are included and ancient samples are assigned to haplogroups where possible. The temporal and geographic distribution of mitochondrial haplogroups is described and analysed. Bayesian analysis was used to date the divergence between and within haplogroups. Chapter 4 describes analyses of genomic data. The topics of sheep population structure and selection in sheep are both introduced. Population structure of ancient sheep is investigated through exploratory analyses and admixture between distinct populations is tested based on the initial results. Several ancient wild sheep are also included. These samples allow for investigation of the domestication process and which wild populations are most similar to early domesticates. A model of sheep population history is established. Genomic regions of selection in Neolithic populations are also explored using a sliding window approach based on Fst divergence from mouflon genomes. These analyses found genes of interest that may have been selected for in Neolithic sheep populations in the millennia after domestication. Some of these genes are related to phenotypes that have previously been shown to be under selection in both sheep and other livestock, such as pigmentation. Chapter 5 summarises the conclusions from each previous chapter. The limitations of the analyses are also discussed. Finally, potential future work after this thesis is discussed. This includes some potential methodologies that may be used to address the limitations of this study, as well as future work that may allow for questions to be answered that this thesis has raised.