Genetic variation is the raw material of evolution. The sources of this variation within a population, and its maintenance within a species, have been mysterious since the birth of the field of evolutionary genetics. In this work, I study divergently adapted freshwater and marine populations of the threespine stickleback (Gasterosteus aculeatus) as an evolutionary model to track the origin of adaptive genetic variation and to describe the evolutionary processes maintaining variation across the genome. The stickleback is a small fish with a large geographic range encompassing the northern half of the Northern Hemisphere and composed of coastal marine habitats, freshwater lakes, and river systems. Populations of stickleback adapt rapidly to changes in habitat, and fossil evidence suggests that similar adaptive transitions have been ongoing in this lineage for at least ten million years. In this work, I develop a significant extension of restriction site-associated DNA sequencing (RAD-seq) to generate phased haplotype information to estimate gene tree topologies and divergence times at thousands of loci simultaneously. I find anciently derived clades of variation associated with marine and freshwater habitats in genomic regions involved in recent adaptive divergence; some divergence times extend to over ten million years ago. This history of adaptive divergence has had profound effects on genetic variation elsewhere in the genome: chromosomes harboring freshwater-adaptive variants retain anciently derived variation in linked genomic regions, while marine chromosomes have much more recent ancestry. I present a conceptual model of asymmetric selective and demographic processes to explain this result, which will form a nucleus for future research in this species. Lastly, by incorporating genome-wide recombination rates estimated from multiple genetic maps, I describe a recombination landscape that is favorable to the maintenance of marine-freshwater genomic divergence. Low recombination rates in key chromosomal regions condense widespread divergence of the physical genome, encompassing many megabases, into a small number of Mendelian loci. Combined, my results demonstrate the interconnectedness of evolutionary processes taking place on ecological and geological timescales. The genetic variation available for adaptive evolution today is a product of the long-term evolutionary history of a species.

Reproduction is among the most dynamic and rapidly evolving genetic systems across sexual taxa. However, the evolutionary mechanisms that drive reproductive traits during the early stages of species divergence are relatively unknown. Using a systems-level, comparative functional genomics approach, I investigate the role of selection, drift, and genomic architecture in promoting the rapid divergence of reproductive systems in Drosophila. I develop a new comparative genomics database, flyDIVaS (Divergence and Selection in Drosophila), an updateable database for identifying patterns and processes involved in species conservation and divergence. I show that tissue-specific genes play a disproportionate role in driving species level divergence and, in particular, that genes specific to male reproductive tissue are among the most rapidly evolving. Using two deeply sequenced populations of Drosophila melanogaster, I reveal that adaptation is widespread among male-specific genes and identify local signatures of selection that have evolved in less than 500 years on sperm motility. In Drosophila mercatorum, a fruit fly capable of facultative parthenogenesis, I find recent signatures of protein changes involved with centrosomal and meiotic functions, and identify early genomic signatures of male degeneration. Among laboratory strains of Drosophila melanogaster, I identify adaptive signatures on neurogenic genes that have recently been domesticated during the last century . Finally, I explore the role of genomic architecture in shaping such reproductive functional systems by developing a novel hypothesis that rapid changes in behavior, such as those found in diverse mating preferences, is a function of the size of the neurogenome. The results presented in this dissertation point toward the importance of selection, drift, and genomic architecture in driving rapid functional change which, together, promote the generation of species diversity via the formation of reproductive barriers in early species divergence. / Biology

Gracilariaceae is a worldwide distributed family that includes numerous economically important species. Currently, five genera are recognized in the Gracilariaceae: Gracilariophila (parasitic), Curdiea, Melanthalia, Gracilaria, and Gracilariopsis. Some species of Gracilaria were taxonomically transferred to Hydropuntia. However, this genus is quite controversial. High-throughput sequencing (HTS) techniques has led to an increase in studies using complete organellar genomes, which have been used to infer phylogenetic relationships in Rhodophyta and the investigation of other aspects of red algal genomes, including gene synteny and horizontal gene transfers (HGT). HTS also facilitated the search for extrachromosomal plasmids and its influence in the organellar genomes by HGT. We applied HTS to assemble and annotate organellar genomes (mitochondria and chloroplast) from seven species of Gracilariaceae using Illumina HiSeq 2500 platform. We also received raw reads of 31 samples of Gracilariaceae from Dr. Goia Lyra that were analysed and included in our work. We used these data, combined with published genomes, to infer phylogenies and compare the genome architecture of these species representing the main lineages in Gracilariaceae. The mitochondrial and chloroplast genomes were highly conserved in gene synteny among the species, and variation mainly occurred in regions where insertions of plasmid-derived sequences (PDSs) were found, which were similar to known red algae extrachromosomal plasmids. In mitochondrial genomes, the PDS insertions were in two regions where the transcription direction changes: between cob and trnL genes, and trnA and trnN genes. PDS insertions in chloroplast genome were in different positions, but generally found between psdD and rrs genes. The bacterial leuC/leuD operon was found in Gracilaria tenuistipitata, G. chilensis, M. intermedia chloroplasts genomes, and also in G. vermiculophylla extrachromosomal plasmid. Phylogenetic trees show two different origins of leuC/leuD: genes found in chloroplasts and plasmids were close to proteobacteria, and genes encoded in the nucleus are close to Viridiplantae and cyanobacteria. Gracilariaceae may be a good model to study the impact of PDS in genome evolution due to the frequent presence of these sequences inserted in organellar genomes. Our phylogenetic analyses demonstrated similar evolutionary histories between the chloroplast and mitochondrial genomes. However, chloroplast phylogeny was better resolved with full support. Our taxonomical sampling supports the presence of three main lineages: Melanthalia/Curdiea, Gracilariopsis and Gracilaria. Melanthalia intermedia was sister to a monophyletic clade including Gracilaria and Gracilariopsis, which were resolved as monophyletic genera. Furthermore, the characteristics of organellar genome architecture, Gracilariopsis and Gracilaria genera are also supported by the loss of the plastid gene petP in Gracilaria and the rearrangement position of the gene trnH in the mitochondrial genome. Beside this, we found no support for the genus Hydropuntia as originally proposed / A família Gracilariaceae está globalmente distribuída e inclui várias espécies economicamente importantes. Atualmente, cinco gêneros são reconhecidos em Gracilariaceae: Gracilariophila (parasita), Curdiea, Melanthalia, Gracilaria e Gracilariopsis. Algumas espécies de Gracilaria foram taxonomicamente transferidas para Hydropuntia. Entretanto, esse gênero é bastante controverso. Técnicas de sequenciamento de alta performance (HTS) levaram a um aumento de estudos usando genomas organelares completos, que têm sido usados para inferir relações filogenéticas em Rhodophyta e na investigação de outros aspectos dos genomas de algas vermelhas, incluindo sintenia gênica e transferências horizontal de genes (HGT). O HTS também facilitou a busca por plasmídeos extracromossômicos e sua influência nos genomas organelares por HGT. Nós utilizamos HTS para montar e anotar genomas organelares (mitocôndrias e cloroplastos) de sete espécies de Gracilariaceae usando a plataforma Illumina HiSeq 2500 e recebemos sequências de 31 amostras Gracilariaceae da Dr. Goia Lyra que foram montadas, anotadas e incluídas em nossas análises. Utilizamos esses dados, combinados com genomas publicados, para inferir filogenias e comparar a arquitetura do genoma dessas espécies representando as principais linhagens em Gracilariaceae. Os genomas mitocondrial e plastidial são altamente conservados na sintenia gênica e a variação ocorreu principalmente em regiões onde foram encontradas inserções de sequências derivadas de plasmídeos (PDS), similares aos plasmídeos extracromossômicos conhecidos de algas vermelhas. Nos genomas mitocondriais, as inserções de PDS estavam em duas regiões onde a direção da transcrição muda: entre os genes cob e trnL e os genes trnA e trnN. As inserções de PDS no genoma do cloroplasto estavam em posições diferentes, mas geralmente encontradas entre os genes psdD e rrs. O operon bacteriano leu/leuD foi encontrado nos genomas dos cloroplastos de Gracilaria tenuistipitata, G. chilensis, M. intermedia e também no plasmídeo de G. vermiculophylla. As árvores filogenéticas mostram duas origens diferentes de leuC/leuD: os genes encontrados no cloroplasto e no plasmídeo estavam próximos de proteobactérias, e os genes codificados no núcleo estavam próximos de Viridiplantae e cianobactérias. Gracilariaceae pode ser um bom modelo para estudar o impacto de PDS na evolução de genomas devido à presença frequente de inserções PDS em genomas organelares. Nossas análises filogenéticas demonstraram histórias evolutivas similares entre cloroplasto e mitocondria. No entanto, a filogenia de cloroplasto foi melhor resolvida com valores máximos de Bootstrap em todos os ramos. Nossa amostragem taxonômica corrobora a presença de três linhagens principais: Melanthalia/Curdiea, Gracilariopsis e Gracilaria. Melanthalia intermedia aparece como grupo-irmão do clado monofilético incluindo Gracilaria e Gracilariopsis, que foram resolvidos como gêneros monofiléticos. Além disso, baseado nas características da arquitetura do genoma organelar, os gêneros Gracilariopsis e Gracilaria se distinguem pela perda do gene plastidial petP em Gracilaria e pela posição de rearranjo do gene trnH no genoma mitocondrial. Nós não encontramos evidências para a permanencia o gênero Hydropuntia como originalmente proposto