Longer is Not Always Better: Optimizing Barcode Length for Large-Scale Species Discovery and Identification

Area of Science:

• Genomics
• Biodiversity Science
• Molecular Ecology

Background:

• Accelerating species discovery, understanding global biodiversity, and developing effective biomonitoring require efficient, scalable, and cost-effective methods for sorting millions of specimens.
• Current methods often rely on museum specimens yielding low-quality DNA, necessitating techniques largely insensitive to DNA quality.
• Mini-barcodes show promise for these criteria, but their efficacy for species-level sorting compared to full-length barcodes remains under investigation.

Purpose of the Study:

• To compare the performance of mini-barcodes against full-length barcodes for species-level sorting.
• To assess the impact of mini-barcode length and position on species delimitation accuracy.
• To evaluate the utility of mini-barcodes for large-scale species discovery and biomonitoring.

Main Methods:

• In silico generation of mini-barcodes (100 bp, 200 bp, 300 bp windows and established primer-minibarcodes) from full-length barcodes of ~30,000 specimens (5500 species) and ~98,000 specimens (>20,000 species).
• Testing species-level congruence between morphospecies and molecular operational taxonomic units (mOTUs) using three delimitation techniques (Poisson Tree Process, Automatic Barcode Gap Discovery, Objective Clustering).
• Analysis of barcode length and position effects on congruence and specimen identification accuracy.

Main Results:

• No significant difference in species- or specimen-level identification performance between full-length and mini-barcodes (>$>$200 bp).
• Very short mini-barcodes (<200 bp) showed poor performance, particularly those near the 5' end of the Folmer region.
• Mean congruence between morphospecies and mOTUs was ~75% for barcodes >200 bp, with congruent mOTUs containing ~75% of specimens.
• Approximately 10% of specimens caused most conflicts, highlighting them for re-examination and efficient conflict resolution.

Conclusions:

• Moderate length mini-barcodes (>$>$200 bp) are as effective as full-length barcodes for species-level sorting and identification.
• Mini-barcodes are suitable for large-scale species discovery, identification, and metabarcoding applications without compromising information content.
• The findings support the use of mini-barcodes for efficient and cost-effective biodiversity assessment and biomonitoring.