The impact of intragenomic rRNA variation on metabarcoding‐derived diversity estimates: A case study from marine nematodes

Abstract

AbstractIntragenomic rRNA variation is a critical concern for eukaryotic metabarcoding studies, due to its potentially confounding effects on species delimitation and biodiversity estimates derived from ‐Omics data. In the present study, we assessed patterns associated with 18S rRNA metabarcoding loci in marine nematodes, including characterization of intragenomic rRNA gene variants (number of variants and abundance profiles) and aspects of datasets that can obscure biological signals (e.g., amplification of nontarget DNA, ambiguous taxonomy assignments). We estimated amplicon sequence variants (ASVs) using DADA2 from an 18S rRNA metabarcoding dataset (Illumina MiSeq) generated from individual marine nematodes. Illumina data were analyzed in conjunction with nematode morphological identifications and nearly full‐length 18S reference sequences (~1,600 bp Sanger barcodes) generated for a subset of the same specimens. Our results indicated that levels of intragenomic rRNA variation appeared to vary widely across nematode taxa (irrespective of phylogenetic clades or ecological feeding groups) and that coamplification of nontarget DNA was common (relic DNA, gut contents, etc.). The DADA2 pipeline appeared to produce a biologically accurate profile of intragenomic rRNA variants in nematodes that was consistent with “Head‐Tail” patterns (of dominant vs. minor rRNA gene variants) identified in previous studies. Although intragenomic rRNA variation appears to be ubiquitous in marine nematodes, nematode identifications were highly congruent across our three methods for species delimitation (traditional morphological taxonomy, Sanger DNA barcoding, and high‐throughput metabarcoding). In spite of pervasive intragenomic variation and high copy number of rRNA genes, the most abundant ASVs in metabarcoding datasets are likely to represent true species barcodes and thus confer an accurate view of extant biodiversity. However, our findings also emphasize the importance of applying bioinformatic filtering techniques and developing well‐curated reference databases in order to better link rRNA molecules with specimen‐level data and alleviate the confounding effects of intragenomic gene variants in studies of microbial eukaryotes.