Evolution of species in a dynamic landscape.

Summary The geographical structuring of populations is a central issue in evolutionary theory. Various analyses have helped to understand its influence on the evolution of species. Nevertheless, the landscape, defined as the set of biotic and abiotic elements that characterize the spatial structuring of populations, is considered as static by formal models of evolution. However, many geological, climatic, ecological and anthropic processes change the landscape at various scales of space and time. Populations are thus repeatedly subjected to fragmentation and mergers, varying their ecological interactions as well as the nature, intensity and conditions of action of the evolutionary mechanisms at work. The objective of this thesis is to determine, with the help of theoretical models, the influence of these landscape dynamics on the evolution of species. At the micro-evolutionary level, we show that probability and time of allele fixation are strongly altered by even simple landscape dynamics. At the meso-evolutionary level, we show that allopatry and sympatry can act in concert in the speciation process, and that complex landscape dynamics can easily generate radiation. Our work reveals that in order to understand species evolution, it is necessary to take into account the way populations are and have been fragmented, but also the time scales associated with the dynamics of this spatial structuring: these determine the combinations and interactions of evolutionary mechanisms producing new species and altering the structure of metacommunities.