Bartonelloses are neglected emerging infectious diseases caused by facultatively intracellular bacteria transmitted between vertebrate hosts by various arthropod vectors. The highest diversity of Bartonella species has been identified in rodents. Within this study we focused on the edible dormouse (Glis glis), a rodent with unique life-history traits that often enters households and whose possible role in the epidemiology of Bartonella infections had been previously unknown. We identified and cultivated two distinct Bartonella sub(species) significantly diverging from previously described species, which were characterized using growth characteristics, biochemical tests, and various molecular techniques including also proteomics. Two novel (sub)species were described: Bartonella grahamii subsp. shimonis subsp. nov. and Bartonella gliris sp. nov. We sequenced two individual strains per each described (sub)species. During exploratory genomic analyses comparing two genotypes ultimately belonging to the same species, both factually and most importantly even spatiotemporally, we noticed unexpectedly significant structural variation between them. We found that most of the detected structural variants could be explained either by prophage excision or integration. Based on a detailed study of one such event, we argue that prophage deletion represents the most probable explanation of the observed phenomena. Moreover, in one strain of Bartonella grahamii subsp. shimonis subsp. nov. we identified a deletion related to Bartonella Adhesin A, a major pathogenicity factor that modulates bacteria-host interactions. Altogether, our results suggest that even a limited number of passages induced sufficient selective pressure to promote significant changes at the level of the genome.