Public COI barcode libraries for New World phlebotomine sand flies: taxonomic coverage, vector-identification...

New World phlebotomine sand flies include medically important vectors and taxonomically complex species groups. Public COI barcode libraries are increasingly used for species identification, biodiversity surveys and vector surveillance, yet their systematic reliability has not been evaluated at continental scale. Here we integrate GenBank and BOLD COI records for New World Phlebotominae to assess taxonomic and geographic coverage, metadata quality, barcode-gap performance, BIN consistency and phylogenetic signal in priority species complexes. We curated 6,200 COI or COI-like records, including 5,562 records assigned to binomial species. The analysis-ready dataset comprised 5,133 filtered records, 3,671 unique haplotypes and 224 species distributed among 20 operational genus-level database labels, all mapped to the Galati et al. (2025) world-checklist hierarchy for supraspecific interpretation. Metadata completeness was uneven: 57.03% of records had country information, 54.40% had coordinates, 64.02% had a BOLD BIN, 82.65% had voucher information and 69.55% had a reference title. Among 203 species evaluable for barcode-gap analysis, 123 showed a barcode gap and 80 showed absent or overlapping gaps. Integrating barcode gap, BINs, geography and focal maximum-likelihood trees classified 103 species as supported by COI for species assignment, 38 as barcode identification risks, 34 as candidates for cryptic diversity or structured populations, 19 as probable species complexes or COI-limited cases, 9 as high-priority species-boundary problems and 21 as insufficiently sampled. Focal IQ-TREE analyses of six critical groups showed strict non-monophyly in 29 of 35 evaluated species. Nearest-neighbour analysis recovered 137 priority species pairs for systematic review; 59 were placed in the same Galati subgenus and 32 in the same species series or group. COI is therefore useful for many New World sand fly species, but its strongest current value is as an audit tool for reference libraries and as a source of evolutionary hypotheses about species groups, sibling species and geographically structured lineages. We provide a reproducible framework for converting public barcode data into a vector-identification and systematic-triage agenda for Phlebotominae.