Abstract : The diversity of living organisms is essential for their capacity to evolve and adapt to environmental changes. Therefore, determining the factors responsible for the origin of diversity and for the maintenance of the genetic variance observed remains central and fundamental research objective. The aim of this thesis was to understand the evolutionary factors maintaining neutral polymorphism. Since the influence of evolutionary processes such as natural selection and genetic drift are complex, we developed complementary experimental and theoretical approaches in order to disentangle their contributions. Using a biological model consisting of the bacterium Escherichia coli and the social amoeba Dictyostelium discoideum enable us to study the natural variability of interactions between the two species. In the second part of this work, we studied the bacterial traits involved in this natural variability. We showed that bacteria carrying virulence genes were resistant to grazing by amoeba, a result which was in agreement with the coincidental evolution hypothesis of virulence factors. We then focus on population genetics aspects of our biological system. In coevolution experiments, we followed temporal allele frequency variations over 300 bacterial generations under four sets of environmental conditions: with or without biotic factor and with or without spatial structure. Our results did not differ from genetic drift predictions. The aim of theoretical model we developed was to address the demographic stochasticity effects on neutral allele fixation probability and time to fixation. We found that fixation probability and the time to fixation were affected by the demographic stochasticity compared with a model using a population of constant size (Moran model).