Migratory behaviour and the persistence of ecotypes

How do ecotypes develop and maintain a coherent identity on the way to becoming species? We are surrounded by many, identifiably different, species, each having some coherent set of properties that distinguish them from related species. Yet it is difficult to understand how these properties come to be associated with each other in the early stages of speciation. In a landmark paper, Felsenstein (1981) highlighted two defining properties of species: adaptations to their environments and a tendency to mate with others in the same species (assortative mating). Felsenstein went on to show that it is not difficult to explain how either of these properties alone are shaped by natural selection, but a fundamental problem is to discover how the genes underlying environmental adaptation and assortative mating become associated with one another in the genomes of diverging taxa. This must take place for coherent new species to emerge but is opposed by the genetic reshuffling (recombination) that occurs between generations in sexual species and breaks apart genetic associations ("linkage disequilibrium"). How can these associations be maintained? The answer must involve selection that favours the associations, but progress in understanding how has been inhibited by limited study of the major sources of selection and when they act during the life cycle. For many years Felsenstein's paper was largely overlooked because for most species we lacked the necessary genetic resolution to test his theory by identifying the genes that define species. Recent developments in genomic technology have reduced this difficulty. Cutting edge studies of ecotypes, local variants within species that maintain their identities through time and space, show that some genes underlying environmental adaptation and assortative mating are held together in "structural variants", genomic regions where recombination is prevented (eg "inversions", where the direction of the DNA is reversed). However, these studies cannot yet explain how such regions originate, or are held together across the whole genome with other such regions, to form the genetic associations that define ecotypes. In animals, ecotypes often differ in migratory behaviour (e.g. one ecotype migrates but another does not). The physiological demands of migration could be an important source of selection that drives a wedge between populations and allows the accumulation of suites of associated traits to suit different migration strategies. The three-spined stickleback is a common, widely distributed and genome-sequenced fish that provides an excellent model to quantify the differences between ecotypes and carry out ambitious experiments to understand how these differences might be maintained by the selection caused by variation in migration. In the North Atlantic, migratory and resident stickleback ecotypes breed alongside each other in many locations, yet the differences between them are maintained by some combination of assortative mating and natural selection, likely due to differences in migratory behaviour. We will quantify the genomic differences between the ecotypes, using state-of-the-art approaches, and investigate how these genes contribute to the adaptive differences between ecotypes in assortative mating and migratory behaviour. We will go on to use a novel and ambitious programme of experiments to understand how selection shapes and maintains the differences between the ecotypes. We will weave together the different strands of our work to shed new light on the fundamental problem of how nascent differences between taxa are maintained at the earliest stages of divergence, advancing our fundamental understanding of biodiversity.