The incidence and diversity of Wolbachia in gallwasps (Hymenoptera; Cynipidae) on oak.
Rokas A., Atkinson RJ., Nieves-Aldrey J-L., West SA., Stone GN.
Wolbachia bacteria infect approximately 20% of all insect species, and cause a range of alterations to host reproduction, including imposition of thelytoky. The incidence and phenotypic impact of Wolbachia remains to be established in many insect taxa, and considerable research effort is currently focused on its association with particular reproductive modes and the relative importance of the various pathways via which infection occurs. Gallwasps represent an attractive system for addressing these issues for two reasons. First, they show a diversity of reproductive modes (including arrhenotoky, thelytoky and cyclical parthenogenesis) in which the impact of Wolbachia infection can be examined. Second, they occupy two intimately linked trophic niches (gall-inducers and inquilines) between which there is potential for the horizontal exchange of Wolbachia infection. In the arrhenotokous gallwasp lineages screened to date (the herb-galling 'Aylacini' and the rose-galling Diplolepidini), Wolbachia infection always induces thelytoky. The impact of Wolbachia in other arrhenotokous clades, and in the cyclically parthenogenetic clades remains unknown. Here we use polymerase chain reaction (PCR) screening and sequence data for two Wolbachia genes (wsp and ftsZ) to examine the prevalence and incidence of Wolbachia infection in 64 species (a total of 609 individuals) in two further tribes: the arrhenotokous inquilines (tribe Synergini), and the cyclically parthenogenetic oak gallwasps (tribe Cynipini). We ask: (i) whether Wolbachia infection has any apparent impact on host reproduction in the two tribes and (ii) whether there is any correlation between Wolbachia infection and the apparent lack of an arrhenotokous generation in many oak gallwasp life cycles. We show: (i) that Wolbachia infection is rare in the Cynipini. Infected species show no deviation from cyclical parthenogenesis, and infection is no more common in species known only from a thelytokous generation; (ii) that there is a higher incidence of infection within the arrhenotokous inquilines, and generally in gallwasp tribes without cyclical parthenogensis; (iii) all Wolbachia-positive inquiline species are known to possess males, implying either that Wolbachia infection does not result in loss of sex in this tribe or, more probably, that (as for some rose gallwasps) Wolbachia infection leads to loss of sex in specific populations; and (iv) although we find some inquilines and gall inducers to be infected with Wolbachia having the same wsp sequence, these hosts are not members of the same gall communities, arguing against frequent horizontal transmission between these two trophic groups. We suggest that exchange may be mediated by the generalist parasitoids common in oak galls.