Limited dispersal and in situ diversification drive the evolutionary history of Rasborinae fishes in Sundaland

AbstractAimSea‐level changes have long been put forward to explain the colonization of Southeast Asian islands by freshwater aquatic organisms. We examined the relative impact of Sundaland geology since the Oligocene and of Pleistocene Eustatic Fluctuations on the mitochondrial lineage diversification of a species‐rich subfamily of freshwater fishes widely distributed in Southeast Asia. We specifically tested if the expansion of exposed lands and increased island connectivity during Pleistocene low sea levels (the paleoriver hypothesis) induced bursts of diversification.LocationSundaland.TaxonRasborinae (Actinopterygii, Cypriniformes, Danionidae).MethodsWe aggregated 1,017 cytochrome oxidase I sequences and 79 mitogenomes to delineate Molecular Operational Taxonomic Units (MOTUs) and further reconstruct a time‐calibrated phylogeny of Rasborinae. Ancestral area estimations were conducted using both island and paleoriver partitioning to examine the impact of island connectivity during Pleistocene sea‐level changes on dispersal. Temporal trends of diversification are explored through statistical selection of best‐fit models.ResultsThe origin of Sundaland mitochondrial lineages is dated at c. 33 Ma and four major clades are identified, which diversified between c. 31 and 22 Ma. The Island of Borneo and North Sunda paleoriver are identified as the source of Sundaland Rasborinae. Geographical patterns of lineage divergence indicate that most divergence events occurred within islands and diversification under constant birth rate models are the most likely for all clades.ConclusionsThe geographical and historical context of diversification of mitochondrial lineages in Rasborinae provides little support for the paleoriver hypothesis. The onset of isolation of Borneo from mainland Asia triggered the initial diversification of the group (c. 31–22 Ma). The late colonization of Java and Sumatra occurred through several independent dispersal events, poorly explained by Pleistocene sea‐level changes and frequently followed by in situ diversification.