Can theory improve the scope of quantitative metazoan metabarcoding?

Abstract

Using high-throughput sequencing approaches to quantify biodiversity has a number of hurdles, in particular that the number of reads for a given taxon may not be proportional to the number of individuals of that taxon in a sample. Here, we consider whether summary statistics generated in the course of population genetic analyses (such as estimates of haplotype diversity and mutation rate) may be useful in reverse inference of the number of individuals input to an assay. Although our results show that these statistics-combined with the observed number of segregating sites and number of haplotypes in the assay-may be informative, there remain significant concerns about the ability to "metabarcode" a sample and infer relative species abundance.