Curating BOLD records via Bayesian phylogenetic assignments enables harmonization of regional subgeneric classifications and cryptic species detection within the genus Tachina (Diptera: Tachinidae)

Abstract

Abstract DNA barcoding is a popular tool for species identification of insect samples, with important applications for plant protection. However, Barcode of Life Database queries related to tachinid samples can produce ambiguous results due to contradictory species assignments. Therefore, we applied Bayesian phylogenetic inference to 811 Tachina Meigen, 1803 COI-5P sequences containing all contradictory species assignments of Tachina Barcode Index Numbers. Based on phylogenetic assignment, the assemble species by automatic partitioning (ASAP) analysis, the MonoPhy test and analyses of genetic divergence, we identified 805 sequences representing species of the Tachina genus, whereas six sequences were incorrectly assigned to Tachina. The dataset contained 21 described species and at least seven unidentified species. Additionally, we found that the described species Tachina marklini Zetterstedt, 1838, includes two cryptic species based on ASAP partitioning of sequences, phylogenetic inference and geographical metadata. Furthermore, we harmonized conflicting subgeneric classifications of the Palearctic and Nearctic by employing different subgenera assignments from both partitions. We aimed at monophyletic subgenera partitions. We largely confirmed the validity of the Palearctic Tachina subgenera and Nowickia as a subgenus of Tachina. However, Palearctic Nowickia species formerly allocated to the outdated genus Fabriciella appeared to constitute one monophyletic group with the subgenus Rhachogaster. We conclude that this set of bioinformatics tools is suitable for curating COI-5P-based DNA barcode libraries. Although, curation is limited due to possible species introgression, horizontal gene transfer, incomplete lineage sorting and human errors the resulting dataset enabled the revision of subgenera concepts and the unambiguous cryptic species detection.