AbstractThe Australasian region is a centre of biodiversity and endemism, mainly based on the tropical climate in combination with the large amount of islands. During the Pleistocene, islands of the Sahul Shelf (Australia, New Guinea, Aru Islands) had been part of the same land mass, while islands within the Wallacea (Lesser Sunda Islands, Moluccas, Sulawesi etc.) remained isolated. We investigated biogeographical avian diversification patterns of two species complexes across the Wallacea and the Sahul Shelf: the Eclectus Parrot Eclectus roratus Wagler, 1832, and the Rainbow Lorikeet Trichoglossus haematodus Linnaeus, 1771. Both species are represented by a large number of described geographical subspecies. We used mitochondrial cytochrome b (cyt b) sequences for phylogenetic and network analysis to detect biogeographic roles of islands and avian diversification patterns. The number of threatened taxa in this region is increasing rapidly and there is an urgent need for (sub-)species conservation in this region. Our study provides first genetic evidence for treating several island taxa as distinct species. In both species complexes similar genetic patterns were detected. Genetic diversification was higher across the islands of the Wallacea than across the islands of the Sahul Shelf. Divergence in E. roratus can be dated back about 1.38 million years ago, whereas in the younger T. haematodus it was 0.80 million years ago. Long distance dispersal was the most likely event for distribution patterns across the Wallacea and Sahul Shelf. The geographic origin of the species-complex Eclectus roratus spp. is supposed to be Wallacean, but for the species-complex Trichoglossus haematodus spp. it is supposed to be non-Wallacean. Trichoglossus euteles, so far considered a distinct species, clearly belongs to the Trichoglossus-haematodus-complex. The only case of sympatry in the complex is the distribution of T. (h.) euteles and T. h. capistratus on Timor, which means a rapid evolution from one ancestor into two distinct species within only 800,000 years. For all other taxa a Checkerboard distribution pattern is present. In this complex, 8 taxa are already treated as separate species (del Hoyo et al. 2014). Based on genetic evidence, the following populations are supported to represent phylogenetic units: (1) N New Guinea (haematodus) incl. Biak (rosenbergii), Bismarck Archipelago (massena), and New Caledonia (deplanchii); (2) Flores (weberi); (3) E Australia (moluccanus) incl. Aru Islands (nigrogularis) and S New Guinea (caeruleiceps); (4) N Australia (rubritorquis); (5) Timor 1st lineage (capistratus) incl. Sumba (fortis); (6) Bali and Lombok (mitchellii); (7) Sumbawa (forsteni); (8) Timor 2nd lineage (euteles). Those 8 phylogenetic units are not identical to the 8 species listed by del Hoyo et al. (2014). Several populations on smaller islands are under decline, a separate species status may lead to a higher conservation status in both species complexes, which are currently listed as “Least Concern”. Eclectus roratus is currently treated as monospecific. Based on genetic evidence, the following populations are suggested being treated as valid species: (1) Sumba (Eclectus cornelia), (2) Tanimbar Islands (E. riedeli), (3) Moluccas (E. roratus), and (4) New Guinea (E. polychloros incl. Aru Islands (E. aruensis), and Solomon Island (E. solomonensis).