ing accessory elements, such as mobilizable nonconjugative plasmids. However, the ability to hitchhike (mobilizability) may well be an evolved character of the hitchhiking element.
Can coincidental evolution also explain transformation and transformability (competence)? The mechanisms that naturally transforming bacteria have for picking up free DNA from the environment, protecting it from destruction by restriction enzymes, and incorporating it into their genomes are complex and highly evolved. A number of hypotheses have been presented for the selective pressures responsible for the evolution of transformation. One of these is consistent with a coincidental evolution hypothesis. In accord with this hypothesis, transformation evolved as a mechanism for acquiring food (nucleotides) from the external environment, and recombination is a coincidental side effect of DNA entering the cell (Redfield, 1993). Alternatively, it has been proposed that transformation evolved specifically as a mechanism to acquire genes from without as templates to repair double-stranded breaks (Michod et al., 1988; Wojciechowski et al., 1989, but see also Mongold, 1992; Redfield, 1993). To these hypotheses, we would like to add the following, not mutually exclusive alternative. We postulate that the ability to acquire genes from other organisms as a source of variation for adaptive evolution is the selective pressure responsible for the evolution and maintenance of transformation. In this interpretation, transformation evolved and is maintained through processes similar to the one proposed for the evolution of mutators (Taddei et al., 1997; Tenaillon et al., 1999), but in this case, the source of variation is external rather than internal. An analogous mechanism may also favor the evolution and maintenance of transposons, integrons, and other elements or processes that, once acquired, augment the rate at which variability is generated. (J. Smith, personal communication).
We have argued that the mechanisms of adaptive evolution are quantitatively and qualitatively different in bacteria than they are in sexual eukaryotes. This difference is primarily a consequence of the frequency of homologous gene recombination being low in bacteria and high in sexual eukaryotes and of the phylogenetic range of gene exchange being broad in bacteria and narrow in contemporary eukaryotes. Also contributing to this difference is the prominent role of viruses, plasmids, and other infectiously transmitted accessory genetic elements as bearers and vectors of genes responsible for adaptive evolution and their seemingly negligible role in this capacity in contemporary eukaryotes.
Clearly, both of these regimes of recombination and horizontal gene transfer are associated with groups of organisms that have been success-