Rapid molecular species identification of mammalian scat samples using nanopore adaptive sampling

Abstract

Accurate taxonomic species identification is essential to the study of mammals. Despite this necessity, rapid and accurate identification of cryptic, understudied, and elusive mammals remains challenging. Traditional barcoding of mitochondrial genes is standard for molecular identification but requires time-consuming wet-lab methodologies. Recent bioinformatic advancements for nanopore sequencing data offer exciting opportunities for noninvasive and field-based identification of mammals. Nanopore adaptive sampling (NAS), a polymerase chain reaction (PCR)-free method, selectively sequences regions of DNA according to user-specified reference databases. Here, we utilized NAS to enrich mammalian mitochondrial genome sequencing to identify species. Fecal DNA extractions were sequenced from 9 mammals, several collected in collaboration with Minnesota Tribal Nations, to demonstrate utility for NAS barcoding of noninvasive samples. By mapping to the entire National Center for Biotechnology Information mammalian mitochondrial reference genome database and bioinformatically analyzing highly similar matches, we successfully produced species identifications for all fecal samples. Eight of 9 species identifications matched previous PCR or animal/fecal appearance-based identifications. For the ninth species, our genetic data indicate a misidentification stemming from the original study. Our approach has a range of applications-particularly in field-based wildlife research, conservation, disease surveillance, and monitoring of wildlife trade. Of importance to Minnesota tribes is invasive species monitoring, detections, and confirmation as climate impacts cause changes in biodiversity and shifts in species distributions. The rapid assessment techniques described here will be useful as new introductions and range expansions of native and invasive species may first be detected by the presence of signs such as scat rather than direct observations and will be helpful for chronically understaffed tribal natural resources agencies.

Keywords: MinION; fecal DNA; mitochondrial DNA; molecular barcoding; nanopore sequencing; phylogenetics; species identification.

Fig. 1.

Species identification workflow using NAS. Mammal feces is collected from the environment and whole genomic DNA is extracted. DNA extracts are prepared for Oxford Nanopore Sequencing following protocols for genomic DNA (see Materials and methods). NAS enriches for sequencing of mammalian mitochondrial genomes. Sequenced reads are bioinformatically analyzed to determine species. Figure created with BioRender.com.

Fig. 2.

Various bioinformatic methods can be utilized to produce species identification. Samples or the species identification are indicated with the black star. Panel A shows a neighbor-joining phylogenetic tree generated by the BOLD database using COI (accession numbers provided in Supplementary Data SD2). Panel B shows a phylogenetic tree generated by RAxML using Cytb sequences from Sample 7 and closely related species from NCBI. Panel C shows the full mitochondrial genome of Sample 2 in a RAxML phylogenetic tree with closely related species from NCBI. Panel D shows the Pavian visualization of Kraken 2 database matches. Reads from Sample 5 matched with 3 species, Catopuma temminckii, Lynx rufus, and Lepus americanus. The majority of the reads matched to Lynx rufus.

Fig. 3.

Complete and near-complete mitochondrial genomes from Peromyscus leucopus (Sample 7) and Canis lupus (Sample 2), respectively. Both mitochondrial genomes were sequenced using NAS. Canis lupus mitochondrial genome is near-complete, with approximately 911 bps missing from Cytb and small sections of D-loop control region and 12S rRNA. Select annotations include genes and rRNA. Figure created with BioRender.com and Geneious Prime software.