Abstract
Many species show subdivision into phenotypically and genetically differentiated forms that are associated with fine-scale habitat variation. These ecotypes may represent an intermediate stage to the formation of new species, and thus offer key models for understanding the process of speciation. Open questions remain with respect to how local adaptations, historical contingencies and components of genome architecture interact in ecotype formation. The current thesis aimed to study ecotypic subdivision in a comparative framework controlling for a similar biogeographic context. We studied five species of marine fishes from the North East Atlantic and Mediterranean Sea: the European anchovy (Engraulis encrasicolus), the long-snouted seahorse (Hippocampus guttulatus), the big-scale sand smelt (Atherina boyeri), the grey wrasse (Symphodus cinereus), and the broadnosed pipefish (Syngnathus typhle). These species occur in a variety of different habitats along the marine-lagoon ecological gradient, and comparing their evolutionary histories has the potential to reveal important aspects related to ecotype formation. We wished to characterise the relative roles of ecology, historical contingencies and genomic architecture in determining the evolutionary trajectories of ecotype pairs in each species. Using whole-genome sequencing data, we aimed to test (i) whether genetic differences were associated with different habitat types, and (ii) how these are maintained in the presence of gene flow. (iii) We evaluated the extent to which the genomic architecture participates in maintaining ecotypic differentiation, and (iv) whether these differences originated from new mutations, standing genetic variation, or introgressed variation. Finally, we aimed (v) to characterise the historical context of ecotypic divergence. In Chapter I, we study ecotypic structure in E. encrasicolus - a highly mobile pelagic species showing marine and coastal ecotypes at a wide geographic scale. We identified multiple structural variants (SVs) that underlie ecotypic differentiation and which were likely introgressed from a third lineage in the Southern Atlantic Ocean. In Chapter II, we study two SVs segregating in H. guttulatus, which differentiate geographical and ecotype lineages. Our results show that these correspond to large chromosomal inversions representing ancient intraspecific polymorphisms, which are subject to different evolutionary dynamics and contribute differently to ecotype formation. Finally, in Chapter III, we compare the eco-geographic patterns and associated genome architectures of ecotypes in all five species. We found that ecotype structure was generally more pronounced in the Mediterranean as compared to the Atlantic, likely indicating the influence of a shared biogeographic history. Moreover, the comparison of divergence landscapes across species revealed that large SVs, such as chromosomal inversions, are consistently involved in ecotypic differentiation. Due to their suppressive effects on recombination, SVs maintain allelic combinations and could act as barriers to gene flow between diverging lineages experiencing gene flow. Although a single SV might not be sufficient for ensuring reproductive isolation, the build-up of linkage disequilibrium among multiple SVs could help strengthen reproductive isolation, although it remains unclear whether this is a sufficient condition for speciation to complete.