Gaps and Data Ambiguities in DNA Reference Libraries: A Limiting Factor for Molecular-Based Biodiversity Assessments Using Annelids as a Case Study.

Regional DNA reference libraries are essential to improve the accuracy of molecular-based biodiversity assessments, species identification and conservation. However, these libraries are often incomplete, limiting the full potential of molecular tools. In this study, we evaluated the completeness of DNA barcode reference data for annelids from the Red Sea, Arabian Gulf and Gulf of Oman and examined its implications for biodiversity assessments. A database of 2291 worldwide annelid species and 3131-4047 Molecular Operational Taxonomic Units (MOTUs) was compiled from the Barcode of Life Data System (BOLD). Two regional checklists -OBIS (498 species) and Wehe & Fiege (892 species)-were cross-referenced against this database to identify coverage gaps and taxonomic inconsistencies. Only 24% and 23% of species in each checklist, respectively, had corresponding barcodes, and just three species were sampled from the region. Additionally, 43% of BOLD's Barcode Index Numbers (BINs) revealed taxonomic ambiguities. To further assess local annelid biodiversity, we analysed a metabarcoding dataset from 135 Autonomous Reef Monitoring Structures (ARMS) deployed on shallow reefs in the region. This yielded 5375 Amplicon Sequence Variants (ASVs), with 55% classified only to class or phylum level. Of the 1350 MOTUs identified, either none to just 14 species-level identifications were found depending on the taxonomic classification method and database, 10 of which appear to be cryptic species complexes. Based on proposed MOTU/species ratios, we estimate approximately 992 annelid species in the ARMS dataset, highlighting underexplored regional diversity. Manual inspection of clustered ASVs also revealed potential pseudogene artifacts, taxonomic misidentifications up to the class level and underestimation of species matches due to discordant MOTUs. These findings underscore the urgent need to expand regional reference libraries, apply integrative taxonomy and implement refined, user-defined MOTU clustering algorithms to improve molecular biodiversity assessments in the region.