Incomplete estimates of genetic diversity within species: Implications for DNA barcoding

Abstract

DNA barcoding has greatly accelerated the pace of specimen identification to the species level, as well as species delineation. Whereas the application of DNA barcoding to the matching of unknown specimens to known species is straightforward, its use for species delimitation is more controversial, as species discovery hinges critically on present levels of haplotype diversity, as well as patterning of standing genetic variation that exists within and between species. Typical sample sizes for molecular biodiversity assessment using DNA barcodes range from 5 to 10 individuals per species. However, required levels that are necessary to fully gauge haplotype variation at the species level are presumed to be strongly taxon‐specific. Importantly, little attention has been paid to determining appropriate specimen sample sizes that are necessary to reveal the majority of intraspecific haplotype variation within any one species. In this paper, we present a brief outline of the current literature and methods on intraspecific sample size estimation for the assessment of COI DNA barcode haplotype sampling completeness. The importance of adequate sample sizes for studies of molecular biodiversity is stressed, with application to a variety of metazoan taxa, through reviewing foundational statistical and population genetic models, with specific application to ray‐finned fishes (Chordata: Actinopterygii). Finally, promising avenues for further research in this area are highlighted.

Can metabarcoding resolve intraspecific genetic diversity changes to environmental stressors? A test case using river macrozoobenthos

“…Higher numbers of unique haplotypes could be induced into the dataset through sequencing errors, which would emphasize the application of an even higher filtering threshold on metabarcoding datasets. However, the increased number of haplotypes found through metabarcoding can also be real haplotype variants in one specimen, which cannot be determined through single specimen barcoding due to the underlying sequencing method (Phillips et al, 2019).…”

Peer Review #2 of "HACSim: an R package to estimate intraspecific sample sizes for genetic diversity assessment using haplotype accumulation curves (v0.1)"

“…Haplotype accumulation curves paint a picture of observed standing genetic variation that exists at the species level as a function of expended sampling effort (Phillips et al, 2015(Phillips et al, , 2019. Haplotype sampling completeness can then be gauged through measuring the slope of the curve, which gives an indication of the number of new haplotypes likely to be uncovered with additional specimens collected.…”

“…Introducing greater statistical rigour into DNA barcoding appears to be the clear way forward in this respect (Nielsen and Matz, 2006;Čandek and Kuntner, 2015;Luo et al, 2015;Phillips et al, 2019). The introduction of computational approaches for automated species delimitation such as Generalized Mixed Yule Coalescent (GMYC) (Pons et al, 2006;Monaghan et al, 2009;Fujisawa and Barraclough, 2013), Automatic Barcode Gap Discovery (ABGD) (Puillandre et al, 2011) and Poisson Tree Processes (PTP; (Zhang et al, 2013)) has greatly contributed to this endeavour in the form of web servers (GMYC, ABGD, PTP) and R packages (GMYC: Species' LImits by Threshold Statistics, splits (Ezard et al, 2017)).…”

“…In this paper, we take a middle-ground approach that is an alternative to probing sampling completeness on the basis of haplotype accumulation curve slope measurement. To this end, iteration is applied to address the issue of relative sample size determination for DNA barcode haplotype sampling completeness, a technique suggested by Phillips et al (2019). Given that specimen collection and processing is quite a laborious and costly endeavour (Cameron et al, 2006;Stein et al, 2014), the next most direct solution to an otherwise blind search strategy is to employ computational simulation that approximates specimen collection in the field.…”

Gene flow and species delimitation in fishes of Western North America: Flannelmouth (Catostomus latipinnis) and Bluehead sucker (C. Pantosteus discobolus)

“…given the semipermeable boundaries now recognized in species. Despite being cofounded by various problems, single-gene methods still predominate in the literature, often with sample sizes 48 that do not capture intraspecific haplotype variability, an issue scarcely parameterized (Phillips et al, 2019). Additionally, single-locus delimitation methods fall under several broad categories, 50 yet each suffers from limitations not easily overcome (Dellicour & Flot, 2018).…”