A Cytochrome C Oxidase I Barcode distance threshold for genus-level identification of pest fruit flies (Tephritidae)

Accessible and accurate methods of pest identification are a cornerstone of pest management. When identifications are not feasible using morphology, molecular methods such as DNA Barcoding provide an option to complete a diagnosis. However, previous studies have shown that some pests cannot be identified to species using DNA Barcodes. In such cases, a genus-level identification could guide subsequent response to pests, but few studies have examined the performance of DNA Barcoding methods to diagnose pest genera as opposed to pest species. We datamined 6,582 previously published Cytochrome C Oxidase I (COI) Barcode region sequences for 12 genera of tephritid fruit flies that include the six main pest genera. After filtering short sequences and duplicate haplotypes, a dataset with 3,077 sequences remained for a pairwise-distance and tree-based genus-level identification test. Our results show that a 5% pairwise distance threshold provides reliable genus-level identifications using sequences of 400 to 658 base pairs. This threshold can easily be implemented with widely available tools such as the Basic Local Alignment Search Tool (BLAST) algorithm. The 5% distance cutoff value is set conservatively and could be refined further by including more species and genera in the reference set. We additionally find that the 3՛ end of the COI Barcode region has greater pairwise distances between genera than the 5՛ end. Our findings provide an additional use of the COI Barcode region for identification that can readily be implemented in pest fruit fly management protocols.