Abstract
Geographic isolation plays a pivotal role in speciation by restricting gene flow between populations through distance or physical barriers. However, the speciation process is complex, influenced by the interplay between dispersal ability and geographic isolation, as seen in "great speciators" - bird species that simultaneously have broad island distributions but high levels of subspecific diversity. Comparing genomic population differentiation in species that occupy both continental and island settings can reveal the effects of different forms of geographic isolation and validate if the primary mechanism proposed to catalyse a great speciator pattern, that is, dispersal reduction following island colonisation, has occurred. The highly diverse white-eye family Zosteropidae includes several great speciators, including the silvereye (Zosterops lateralis), with 16 subspecies (11 occurring on islands), distributed on the Australian continent and numerous southwest Pacific islands. We compared continental and island patterns of divergence using whole genome and morphological data. Australian mainland populations showed a low genetic population structure, lack of isolation by distance patterns and low morphological diagnosability, suggesting that the species' dispersal propensity in a continental setting is sufficient to overcome multiple forms of geographic barriers and large geographic distances. In contrast, except for island populations less than 200 years old, most island populations were highly genomically structured with clearer morphological diagnosability even if separated by relatively short geographic distances. The inferred reduction of dispersal propensity in island situations is consistent with the proposed model of great speciator formation on islands. Our phylogenomic analyses also allowed resolution of the silvereyes' evolutionary position, showing their relatively early emergence (~1.5 Mya) within the rapidly radiating Zosteropidae, while population-level analyses demonstrated where morphological subspecies and genomic data align and disagree. However, the silvereye example also shows how uncertainties about relationships remain when reconstructing evolutionary history in rapidly radiating groups, even when whole genome data is available. Altogether, our results show how within-species genomic and morphological patterns measured over broad spatial scales and with varying geographic contexts can help reveal when particular stages of speciation such as great speciators are likely to emerge.