Abstract
The gigantea species group of the genus Colocasiomyia de Meijere (Diptera: Drosophilidae) is among the four aroid-breeding species groups in this genus; however, it differs from the remaining three groups in the host use: all the flies in this group use plants from the subfamily Monsteroideae instead of from the subfamily Aroideae. So far, we have not resolved the phylogenetic relationship within this group, making it difficult to trace its geographical origin, pattern of species diversification and history of host plant use. In this study, we reconstructed the phylogenetic relationships within the C. gigantea group using DNA sequences of eight (two mitochondrial and six nuclear) gene markers, and we inferred the ancestral areas and host plants of the group based on the resulting phylogeny. According to the results, the C. gigantea group may have diverged from its sister group (i.e., the C. cristata group) through vicariance between the northeastern Oriental region and Sundaland + Wallacea, and the subsequent diversification of the C. gigantea group occurred mostly in the northeastern Oriental region, although an Oriental-to-Sundaland dispersal was followed by vicariance between these two areas, which finally gave rise to the C. gigantea-C. scindapsae lineage in the latter area. We inferred the most likely ancestral host plant of the C. gigantea group to be of the genus Rhaphidophora Hassk, with possible subsequent shifts to Scindapsus Schott and/or Epipremnum Schott plants. We discuss the potential for the egg filaments in the C. gigantea group to be used as a model system for comparative studies in pollination mutualism and developmental genetics concerning tubulogenesis.