The choice of universal primers and the characteristics of the species mixture determine when DNA metabarcoding can be quantitative

Abstract

DNA metabarcoding is a technique used to survey biodiversity in many ecological settings, but there are doubts about whether it can provide quantitative results, that is, the proportions of each species in the mixture as opposed to a species list. While there are several experimental studies that report quantitative metabarcoding results, there are a similar number that fail to do so. Here, we provide the rationale to understand under what circumstances the technique can be quantitative. In essence, we simulate a mixture of DNA of S species with a defined initial abundance distribution. In the simulated PCR, each species increases its concentration following a certain amplification efficiency. The final DNA concentration will reflect the initial one when the efficiency is similar for all species; otherwise, the initial and final DNA concentrations would be poorly related. Although there are many known factors that modulate amplification efficiency, we focused on the number of primer-template mismatches, arguably the most important one. We used 15 common primers pairs targeting the mitochondrial COI region and the mitogenomes of ca. 1,200 insect species. The results showed that some primers pairs produced quantitative results under most circumstances, whereas some other primers failed to do so. In conclusion, depending on the primer pair used in the PCR amplification and on the characteristics of the mixture analysed (i.e., high species richness, low evenness), DNA metabarcoding can provide a quantitative estimate of the relative abundances of different species.

Evaluating next‐generation sequencing (NGS) methods for routine monitoring of wild bees: Metabarcoding, mitogenomics or NGS barcoding

“…PCR amplification efficiency is primarily (73%) influenced by the number of template‐primer mismatches (Piñol et al, ), and therefore, the selection of primers will greatly influence the quantitative output (Piñol et al, ). While testing 15 common universal COI primer pairs, Piñol et al () found a significant relationship between DNA concentration pre‐ and post‐PCR for the vast majority of primers (14/15) although R 2 values were variable. The primer pair used in our study performed relatively well, even though other primers performed better (e.g., ArF5 & ArR5, Gibson et al, ).…”

Promises and pitfalls of using high‐throughput sequencing for diet analysis

“…For example, the 16S‐targeting Coleop primers (Epp et al, ) recovered a wider dietary diversity in insectivorous bats than the COI targeting ZBJ‐Art primers (Zeale, Butlin, Barker, Lees, & Jones, ), and yet, the percentage of species‐level assignments was considerably lower in the former (Alberdi et al, ). Finally, in silico testing the potential taxonomic biases of different primers might be a promising approach to select the most appropriate primer pairs (Piñol, Senar, & Symondson, ).…”

A guide to the application of Hill numbers to DNA‐based diversity analyses

“…Recently conducted research has clearly documented how many biological and technical distortion factors introduce numerous biases that break the relation between the actual biomass distribution in the system and the relative amount of DNA sequences obtained in the final results (Barnes & Turner, ; Lamb et al, ). This is partly due to primer amplification biases, and thus sequencing probability, due to the different binding affinity between primers and target sequences (Piñol, Senar, & Symondson, ). Finally, both PCR sequencing and DNA sequencing can generate artifactual DNA sequences that do not exist in the actual biological system, which results in increased false positive rate and inflated diversity (Alberdi et al, ).…”

“…Occupancy‐modelling approaches have traditionally been applied to overcome such detectability biases (MacKenzie, Nichols, Hines, Knutson, & Franklin, ). In molecularly characterized systems, the issue of detectability is even more complex, because in addition to the biological distortions of environmental DNA (Barnes & Turner, ), there is another important source of bias produced by uneven primer amplification rates (Piñol et al, ). While eDNA representativeness assessment might be too complex to model (Alberdi et al, ; Barnes & Turner, ), amplification biases can be measured in silico (Piñol et al, ) and using mock communities (Lamb et al, ).…”