Frozen Antarctic path for dispersal initiated parallel host-parasite evolution on different continents.

After the break-up of Gondwana dispersal of organisms between America, Australia and Africa became more complicated. One of the possible remaining paths led through Antarctica, that was not yet glaciated and it remained habitable for many organisms. This favourable climate made Antarctica an important migration corridor for organisms with good dispersal ability, such as Aculeata (Hymenoptera), till the Oligocene cooling. Here we tested how cooling of Antarctica impacted global dispersal of Aculeata parasites (Strepsiptera: Xenidae). Our data set comprising six nuclear genes from a broad sample of Xenidae. Bayesian dating was used to estimate divergence times in phylogenetic reconstruction. Biogeography was investigated using event-based analytical methods: likelihood-based dispersal-extinction-cladogenesis and Bayesian models. The Bayesian model was used for reconstruction of ancestral host groups. Biogeographical methods indicate that multiple lineages were exchanged between the New World and the Old World + Australia until the Antarctica became completely frozen over. During the late Paleogene and Neogene periods, several lineages spread from the Afrotropics to other Old World regions and Australia. The original hosts of Xenidae were most likely social wasps. Within one lineage of solitary wasp parasites, parallel switch to digger wasps (Sphecidae) occurred independently in the New World and Old World regions. The biogeography and macroevolutionary history of Xenidae can be explained by the combination of dispersal, lineage extinction and climatic changes during the Cenozoic era. A habitable Antarctica and the presence of now-submerged islands and plateaus that acted as a connection between the New World and Old World + Australia provided the possibility for biotic exchanges of parasites along with their hymenopteran hosts. Although Xenidae are generally host specialists, there were significant host switches to unrelated but ecologically similar hosts during their evolution. There is little or no evidence for cophylogeny between strepsipteran parasites and hymenopteran lineages.