Shorter, better, faster, stronger? Comparing the identification performance of full-length and mini-DNA barcodes for apid bees (Hymenoptera: Apidae)

Abstract

Apid bees are essential ecosystem pollinators, occurring worldwide and comprising over 5900 species. Although they are identified mainly using morphology, DNA barcoding has been explored since its proposal as a supplementary tool in bee taxonomy. Smaller regions of barcode markers—mini-barcodes—were also successfully employed in corbiculate bee identification, but the performance of mini-barcodes was only tested in a narrow taxonomic scope. Here, we scrutinized all 18167 apid bee cox1 sequences from the Barcode of Life Data System to provide an overview of the available data, search for barcoding gaps at genus level, test if full-length and mini-barcode regions perform similarly in specimen identification, and flag bee taxa that may benefit from studies implementing DNA barcodes. Our dataset encompassed five subfamilies, 25 tribes, 71 genera, and 1012 species, although it was biased towards corbiculate tribes. Most of the surveyed genera showed good performance in the barcoding gap analyses. Moreover, full-length and mini-barcodes displayed a similar probability of correct identification, demonstrating that both marker types are equivalent in bee identification. Finally, we discuss some examples to show how full-length and mini-barcodes can help solve taxonomic inconsistencies and foment future studies of apid bees.