The African continent still hosts a unique diversified megafaunal community. However, phylogeographical patterns of African species have not yet been largely studied, as compared to the North American and European species. Moreover, nowadays, the African ecosystems suffer of considerable anthropogenic pressures and of severe climatic modifications. Wildlife population fragmentation resulting from habitat loss, drought, poaching and diseases is currently threatening many African species survivals. The present thesis aimed at investigating the phylogeography and the population genetic structure of the African buffalo (Syncerus caffer) at different spatio-temporal scales (phylogeographical scale vs demographic scale), based on different molecular markers (mtDNA D-Loop region, 17 microsatellites and a large set of Single Nucleotide Polymorphisms (SNPs)). The African buffalo is an emblematic key species and provides a powerful model to enhance our understanding of the African biogeography and the species conservation requirements. The aim of the first part of the present work was to reconstruct the evolutionary history of the African buffalo based on the study of the mtDNA D-Loop hypervariable region. More particularly, we aimed at studying the impact of the Quaternary climatic fluctuations on the species distribution (i.e. phylogeographical scale). Moreover, one of our purposes was also to investigate the taxonomic controversies linked to the extreme within species morphological variability, using molecular tools. From a sample set including all four morphologically recognized subspecies roaming the sub-Saharan African continent (hereafter called ecophenotypes), two taxonomic units were supported by genetics. The genetic discontinuity was located between the West-Central (S. c. nanus, S. c. brachyceros and S. c. aequinoctialis) and the South-Eastern populations (S. c. caffer). The low amount of genetic differentiation within each of these two units, or lineages, was attributed to a recent (in evolutionary term) Pleistocene expansion in both lineages, with rapid adaptation to a variety of habitats. Using both microsatellites and a large set of SNPs, the second part of the present thesis aimed at investigating the impact of human activities and recent climatic changes on the population structure of the species (i.e. demographic scale). While mtDNA did not allowed to distinguish finer sub-structuration within the two identified lineages, suggesting high female gene flow at an evolutionary timescale, these last molecular markers allowed to identify eight different populations at the continental level: two in West-Central Africa, three in Eastern and three in Southern Africa. Both ancient (Neolithic revolution) and recent anthropogenic activities were proposed to have shaped the demographic population structure observed in Southern and Eastern Africa. Recently, two of these populations were shown to be under significant genetic drift, following severe demographic bottlenecks. In these two cases, the sharp reduction in the size of the populations was proposed to be linked to disease eradication campaigns and to overharvesting during civil wars. Nevertheless, they did not display a significant loss in heterozygosity, indicating that they are still genetically healthy. Interestingly, at the contact region between the S. c. caffer and the S. c. aequinoctialis savanna ecophenotypes, a population displaying a shared genetic pool was also identified, corroborating the existence of hybrids between the different forms (intermediate morphological characteristics). In conclusion, using various genetic approaches and molecular markers, the present work gained insights into the taxonomy and the evolutionary history of the African buffalo, as well as in its conservation management requirements.