Are some endosymbionts more prone to horizontal transfer and recombination?
One explanation for the observed patterns could be the relative ease with which supergroup A Wolbachia and Cardinium can undergo horizontal transfer and recombination as compared to supergroup B Wolbachia and Arsenophonus . This essentially means that the former two endosymbionts would encounter previously existing bacterial infections within their hosts which would increase the opportunity for recombination among the pre-existing and the new bacterial strains. Recombination would then create newer allele variants. This is indeed borne out by the results in table 1 which depicts the number of unique alleles found in this study among theWolbachia infections. In all about 84% (71 out of a possible 85 alleles) of the A supergroup infection are unique. Whereas, about 75% (56 out of a possible 75 alleles) are unique in B supergroupWolbachia infections. Furthermore, as indicated in table 2, the number of within supergroup recombination detected in the A supergroup strains (8 instances) far outnumber the B supergroup Wolbachia , where none were detected. This is in spite of horizontal transfer of the entire B supergroup ST’s (ST-541 and ST-559) to taxonomically unrelated hosts (Table 1). An expected outcome of such pervasive horizontal transfer and resulting recombination would have been an increase in sequence diversity in the A supergroup strains, especially, if the source of recombination had been infections outside the community. This does not seem to be the case as the A supergroup infection show less than expected pairwise distance (2.67%) when compared with the B supergroup infections (4.17%). This indicates that the sources of recombination must be from infections within this community. In other words, the standing sequence variation of the A supergroup infections is being partitioned across the community-wide arthropod taxa into newer recombinants with resulting increase in allele diversity but not overall sequence divergence. Moreover, what follows from this relatively low pairwise divergence of the A supergroup infections is that this horizontal transfer and recombination must have been recent or rapid enough for any post-recombination sequence variation to accumulate. This indicates that the A supergroup infections are either better at horizontal transfer across the community or are presently undergoing such rapid transfers as has been suggested by Werren et al (1995). On the other hand, the B supergroup Wolbachia infections show relatively diverged strains with low rates of recombination indicating much stable infections. Since, little is known about the biological characteristic of the different Wolbachia supergroups, other than sequence divergence, it is difficult to speculate whether there are supergroup specific effects on their hosts. For example, it is not known whether any supergroup infections exclusively infect any specific arthropod taxa or whether any supergroup make hosts more prone to horizontal transfer? Therefore, we concentrate on specific trophic interactions of the hosts themselves and try to explain why supergroup A infections show such extensive horizontal spread.