Genetic diversity is the foundation of all biological variation. An approach for sustainable use and protection of genetic diversity is continuous sampling over space and time, i.e. monitoring. It is important to consider genetic changes over contemporary time frames, since most human perturbations have occurred within the last century. Modern molecular tools now enable genome-wide diversity monitoring, also in non-model species.

The work included in this thesis utilizes theoretical and molecular tools to monitor genomic diversity over microevolutionary time frames using salmonid fishes as models. First, the capacity for substructured populations to retain genetic variation following population extinctions was theoretically assessed. Models of effective population size (N_e) relevant to salmonids were used. Further, spatio-temporal genetic patterns of the highly substructured brown trout (Salmo trutta) were empirically estimated. Wild populations were studied using whole-genome sequencing, primarily of pools of individuals (Pool-seq). The brown trout is characterized by a large and complex genome, and genomic resources have, until recently, been lacking. One central aim of this thesis was therefore to evaluate the benefit of using Pool-seq data for monitoring genetic diversity in this species. To this end, disparate natural populations were studied that are, in part, previously described using classic genetic markers. First, I hypothesized that a Pool-seq-only approach developed for non-model species that lack reference genomes could be used to detect population differentiation between two scenarios of coexisting populations. In a second step, two different cases of populations in the wild – one experimental release and one case of protected populations – were monitored over nearly four decades (5-6 brown trout generations) using Pool-seq data. I asked what the levels of diversity and divergence among populations are, whether changes could be detected over contemporary time and if they could be attributed to adaptation.

Paper I demonstrates that the effect of extinction on the rate of diversity change in population systems is more complex than previously recognized. Diversity loss is most prominent when migration within the population system is limited, which suggests that highly substructured population systems, e.g., many salmonids, are particularly vulnerable to population extinction. The utility of Pool-seq for monitoring brown trout populations over contemporary time is demonstrated for the three different cases of brown trout populations (Papers II-IV). Paper II confirms the ability of a Pool-seq-only approach to detect subtle population differentiation. Paper III identifies genome-wide levels of hybridization between populations introduced to a new environment and signs of adaptation in genes putatively involved in metabolism. Paper IV detects significant allele frequency shifts over a limited number of generations. Potentially adaptive change is also identified, with regions containing genes possibly associated to immunity, skin pigmentation, and reproduction (Paper IV).

This thesis demonstrates the benefit of modern theoretical and molecular tools for monitoring diversity in highly substructured population systems. These tools are relevant for advancing population genetic knowledge, as well as for sustainable management and conservation of a wide range of species.