Systematics of prokaryotes: the state of the art

The term taxonomy is often used synonymously with systematics but it should be regarded more as a specific part of the latter and comprises the orderly arrangements of (defined) units in addition to the nomenclature, i.e. labelling of these units defined by classification, and also identification of these units defined by classification and labeled by nomenclature. Similar to all biological disciplines, taxonomic approaches in microbiology aim at the establishment of a system that mirrors the order in nature as closely as possible with the ultimate goal to describe the whole evolutionary order back to the origin of life. With the recognition of molecular markers present in all organisms (here in particular the small subunit rRNAs, ssRNSs), the achievement of this goal has become more and more feasible and the generation of gene and increasing numbers of genome sequences allow nowadays the generation of large amounts of data and often a very detailed insight into the genetic potential of prokaryotes. The possibility to generate whole genome sequences in a very short period of time leads to a strong tendency to base the taxonomic system more and more on sequence data. However, a comprehensive understanding of all the information behind sequence data is lagging far behind their accumulation. Genes and genomes may (or may not) function only in a given environment, with the cell as basic entity for the display of this potential. Prokaryotic taxonomy still has its focus on the whole organism. In this context, natural selection drives evolution selecting the existing phenotypes and it is the phenotype that exhibits this process both in a given cellular and also environmental context. The term polyphasic taxonomy, which was coined almost 40 years ago and aimed at the integration of many levels of information (from molecular to ecological data) thereby allowing a more holistic view, should be revisited in the light of the enormous potential of the novel information associated with large data sets.