Genomic data reveal new species and the limits of mtDNA barcode diagnostics to contain a global pest species complex (Diptera: Tephritidae: Dacinae)

The Oriental fruit fly Bactrocera dorsalis (Hendel), a global pest that can decimate regional fruit industries and elicit international quarantines, has been the subject of considerable taxonomic confusion. Previous phylogenetic work revealed that B. dorsalis is part of a monophyletic clade containing 12 species. We present restriction site-associated DNA sequencing (RAD-seq) genomic data for 2,292 specimens, which unequivocally supports the delimitation of two new species, here described as Bactrocera borneoensissp. n. Doorenweerd & San Jose and B. incognitasp. n. Doorenweerd & San Jose. We additionally obtained 1,985 Cytochrome C oxidase I (COI) sequences for a subset of the specimens to see which species can be diagnosed with this mtDNA marker and conclude that B. dorsalis, B. incognita, B. carambolae Drew & Hancock, B. raiensis Drew & Hancock, B. occipitalis (Bezzi) and B. kandiensis Drew & Hancock cannot be identified reliably using COI due to introgression-but the newly described species B. borneoensis can be identified using COI. The supposed innocuous species B. raiensis distribution is underestimated in Asia and Africa. Bactrocera kandiensis COI genotypes occur in African flies, but RAD-seq data confirm that these are B. dorsalis with introgressed B. kandiensis COI. The phylogenomic dataset brings new light to the extent of the B. dorsalis s.l. clade and the morphological and molecular confusion based on COI. This will have ramifications for ecological data-including host and distribution ranges-associated with B. dorsalis s.l. clade species, pest identification protocols and our understanding of the economic importance of the various species in the clade.

Automated summary

This study used RAD-seq and COI sequences to investigate the taxonomic confusion of the Oriental fruit fly, identifying two new species and resolving the identification issues of other closely related species. The findings have implications for ecological data and pest identification protocols.