Proceeding global and local environmental changes require innovative conservation concepts and animal model systems which integrate empirical and genomic data. Odonata (dragonflies and damselflies) are a highly suitable model system for the evaluation of ecosystems and on top hold an evolutionary key position at the base of winged insects. The rapid (r)evolution of genetic methods in the last two decades has allowed the transition of traditional ecological field model systems (which cannot or only with difficulty be kept in the laboratory in contrast to traditional genetic model organisms) to model systems of integrative research. This raises the possibility of closing the gap between ecological factors and their genetic consequences and / or causes. Here, the order Odonata plays a prominent role based on the solid foundation of extensive ecological data and now numerous genetic research approaches for dragonflies. Consequently, the presented dissertation deals with two odonate species with very different ecological niches, almost opposing adaptation strategies and distributional dynamics. These are on the one hand Lestes macrostigma, an endangerd, stenotopic damselfly adapted to temporary, brackish ecosystems with spatially very limited distribution. The other target species is the cosmopolitan, abundant and migratory Pantala flavescens, inhabiting all kinds of temporary waters created by local rains as an ecological generalist. Due to their temporary nature, both habitats are particularly prone to climate change, but they differ significantly in their spatio-temporal occurrence. The resulting different adaptation strategies and distributional dynamics of both species were therefore genetically characterized in this work on (i) temporal, (ii) local and (iii) global levels to compare potential “winners” and “losers” in terms of the current insect extinction and rapidly changing environmental conditions:A long-term local monitoring of a population of L. macrostigma in southern France revealed highly fluctuating population sizes resulting in genetic bottlenecks. These results were associated with fluctuating precipitation and drought periods and were compared to populations covering the entire distribution range of this species and are discussed in a geographic context. The global comparison approach reveals that the fragmented population structure is reflected in the genetics of this species and that populations are even genetically isolated. Furthermore, two conservation units were discovered in Spain and France requiring the re-evaluation of the conservation status of these populations by the relevant authorities. Finally, a new dispersal model of this species was proposed, suggesting that this species originated in Asia with a subsequent westward expansion.A first global population study on the cosmopolitan dragonfly species P. flavescens surprisingly revealed contradicting geographic structuring based on mitochondrial and nuclear data. Despite their extraordinary migratory potential, island populations of P. flavescens show genetic signals of local adaptation processes, going along with phenotypic differences and behavioral adaptations. In particular, the investigated population on Easter Island is genetically isolated and depleted whereas all other populations show remarkably high intraspecific diversity. However, nucleotide and amino acid diversity indicate saturation of the CO1 barcode gene fragment, highlighting the limited suitability of marker for population genetic studies as local adaptations could not be detected.To complement single marker gene analyses and to pave the way for genomic studies on insect migration, the complete mitochondrial genome of an Easter Island indivdual of P. flavescens was characterized and compared to mitogenomes of the key odonate species Anax imperator, Ischnura elegans, and Megaloprepus caerulatus. These new mitogenome data are a highly valuable resource for future comparative mitogenome studies on the population level and for insect migration. Migration of flying insects has so far only been studied in detail on two phylogenetically very derived model systems (Danaus plexippus and Locustra migratoria). Since dragonflies are phylogenetically placed on the base of winged insects, studies onP. flavescens open up the possibility of tackling the question of the evolution of migration and its genetic mechanisms.