architecture. As such, conspecifics share numerous characteristics by which they can be grouped, even when evidence of interbreeding is limited or unknown.
By lacking a mechanism that regularly homogenizes the features of different organisms, strictly asexual organisms continuously diverge from one another as independent lineages. And although the classification of these lineages is undoubtedly useful, it could be argued that any criteria for delineating asexual species, e.g., possessing of a particular suite of phenotypic traits or attaining a prescribe degree of DNA similarity, are arbitrary, inconsistent across taxa, and biologically meaningless.
Acknowledging the problems associated with classifying groups of asexual organisms into discrete species, the situation with bacteria is even worse. Bacteria reproduce asexually, yet they are also capable of obtaining genes from other organisms, even those of different kingdoms. Moreover, the amounts, types, and sources of imported genes can vary among lineages, allowing gene transfer to blur the boundaries of bacterial groups at every taxonomic level and in ways that are impossible to predict. And if patterns of vertical descent are obscured in varied and unknown ways, then the systematic classification of bacteria might not be possible [see Rossello-Mora and Amann (2001), Lan and Reeves (2001), Young (2001), and Cohan (2002) for current reviews on the concept of bacterial species].
There is a clear advantage to examining the process of diversification in bacteria, which is the availability of complete sequences from hundreds of genomes whose relationships range in type from members of the same nominal species to representatives of groups that diverged billions of years ago. These new data allow us to follow the origin and ancestry of every gene in a genome to resolve the degree to which gene transfer has shaped the contents of bacterial genomes and has obscured the history of bacterial groups at different phylogenetic depths.
There are several means by which bacteria can acquire genes: by conjugal transfer, by phage-mediated insertions and by the update of native DNA from the outside sources (Ochman et al., 2000; Redfield, 2001). But given the diversity of mechanisms that are capable of planting virtually any gene in virtually any organism, bacterial genomes remain small (on the order of 500–10,000 kb) and are not simply arbitrary assortments of genes of mixed heritage. Although bacteria might be bombarded constantly with foreign genes, only evolutionarily relevant events of transfer, i.e., those resulting in genes that persist, are evident from the contents of extent genomes.
With the completion of each bacterial genome sequence, there is a