Although high-throughput sequencers (HTS) have largely displaced their Sanger counterparts, the short read lengths and high error rates of most platforms constrain their utility for amplicon sequencing. The present study tests the capacity of single molecule, real-time (SMRT) sequencing implemented on the SEQUEL platform to overcome these limitations, employing 658 bp amplicons of the mitochondrial cytochrome c oxidase I gene as a model system. By examining templates from more than 5000 species and 20,000 specimens, the performance of SMRT sequencing was tested with amplicons showing wide variation in GC composition and varied sequence attributes. SMRT and Sanger sequences were very similar, but SMRT sequencing provided more complete coverage, especially for amplicons with homopolymer tracts. Because it can characterize amplicon pools from 10,000 DNA extracts in a single run, the SEQUEL can reduce greatly reduce sequencing costs in comparison to first (Sanger) and second generation platforms (Illumina, Ion). SMRT analysis generates high-fidelity sequences from amplicons with varying GC content and is resilient to homopolymer tracts. Analytical costs are low, substantially less than those for first or second generation sequencers. When implemented on the SEQUEL platform, SMRT analysis enables massive amplicon characterization because each instrument can recover sequences from more than 5 million DNA extracts a year.

Type specimens have high scientific importance because they provide the only certain connection between the application of a Linnean name and a physical specimen. Many other individuals may have been identified as a particular species, but their linkage to the taxon concept is inferential. Because type specimens are often more than a century old and have experienced conditions unfavourable for DNA preservation, success in sequence recovery has been uncertain. This study addresses this challenge by employing next‐generation sequencing (NGS) to recover sequences for the barcode region of the cytochrome c oxidase 1 gene from small amounts of template DNA. DNA quality was first screened in more than 1800 century‐old type specimens of Lepidoptera by attempting to recover 164‐bp and 94‐bp reads via Sanger sequencing. This analysis permitted the assignment of each specimen to one of three DNA quality categories – high (164‐bp sequence), medium (94‐bp sequence) or low (no sequence). Ten specimens from each category were subsequently analysed via a PCR‐based NGS protocol requiring very little template DNA. It recovered sequence information from all specimens with average read lengths ranging from 458 bp to 610 bp for the three DNA categories. By sequencing ten specimens in each NGS run, costs were similar to Sanger analysis. Future increases in the number of specimens processed in each run promise substantial reductions in cost, making it possible to anticipate a future where barcode sequences are available from most type specimens.

The nuclear genomes of most animal species include NUMTs, segments of the mitogenome incorporated into their chromosomes. Although NUMT counts are known to vary greatly among species, there has been no comprehensive study of their frequency/attributes in the most diverse group of terrestrial organisms, insects. This study examines NUMTs derived from a 658 bp 5’ segment of the cytochrome c oxidase I (COI) gene, the barcode region for the animal kingdom. This assessment is important because unrecognized NUMTs can elevate estimates of species richness obtained through DNA barcoding and derived approaches (eDNA, metabarcoding). This investigation detected nearly 10,000 COI NUMTs ≥ 100 bp in the genomes of 1,002 insect species (range = 0–443). Variation in nuclear genome size explained 56% of the mitogenome-wide variation in NUMT counts. Although insect orders with the largest genome sizes possessed the highest NUMT counts, there was considerable variation among their component lineages. Two thirds of COI NUMTs possessed an IPSC (indel and/or premature stop codon) allowing their recognition and exclusion from downstream analyses. The remainder can elevate species richness as they showed 10.1% mean divergence from their mitochondrial homologue. The extent of exposure to “ghost species” is strongly impacted by the target amplicon’s length. NUMTs can raise apparent species richness by up to 22% when a 658 bp COI amplicon is examined versus a doubling of apparent richness when 150 bp amplicons are targeted. Given these impacts, metabarcoding and eDNA studies should target the longest possible amplicons while also avoiding use of 12S/16S rDNA as they triple NUMT exposure because IPSC screens cannot be employed.