Disentangling bias for non-destructive insect metabarcoding

Abstract

A fast and reliable method for obtaining a species-level identification is a fundamental requirement for a wide range of activities, from plant protection and invasive species management to biodiversity assessments and ecological studies. For insects, novel molecular techniques such as DNA metabarcoding have emerged as a rapid alternative to traditional morphological identification, reducing the dependence on limited taxonomic experts. Until recently, molecular techniques have required a destructive DNA extraction, precluding the possibility of preserving voucher specimens for future studies, or species descriptions. Here we paired insect metabarcoding with two recent non-destructive DNA extraction protocols, to obtain a rapid and high-throughput taxonomic identification of diverse insect taxa while retaining a physical voucher specimen. The aim of this work was to explore how non-destructive extraction protocols impact the semi-quantitative nature of metabarcoding, which alongside species presence/absence also provides a quantitative, but biased, representation of their relative abundances. By using a series of mock communities representing each stage of a typical metabarcoding workflow we were able to determine how different morphological (i.e., insect biomass and exoskeleton hardness) and molecular traits (i.e., primer mismatch and amplicon GC%), interact with different protocol steps to introduce quantitative bias into non-destructive metabarcoding results. We discuss the relevance of taxonomic bias to metabarcoding identification of insects and potential approaches to account for it.

Optimizing insect metabarcoding using replicated mock communities

“…Such findings were made by, e.g., Marquina et al (2019) and Nielsen et al (2019) when comparing different methods of DNA purification using lysate from the same bulk sample. Reproducibility was also evaluated by Martoni and collaborators (2022), who used mock communities to evaluate taxonomic bias introduced by the choice of PCR primers and by the library preparation method. However, to our knowledge, the current study is the first to examine variation between biological replicates of multi-species communities.…”

“…Differences in species’ biomass and physical structure cause unequal contribution of DNA to the sequenced pool. Added to this, PCR errors and primer biases towards certain taxonomic groups can obscure estimates of species diversity and abundances, as demonstrated through metabarcoding of samples with known composition (Braukmann et al, 2019; Elbrecht and Leese, 2015; Martoni et al, 2022). Multiple solutions have been proposed to address these challenges.…”

Depicting environmental gradients from Malaise trap samples: Is ethanol‐basedDNAmetabarcoding enough?

“…A striking subject of today's (and the future's) research concerns the advancing methodology of non‐destructive DNA extractions. Numerous studies dedicated to the development of non‐destructive methodologies for sequencing are emerging, showing that it is possible to extract DNA (although in smaller quantities) from specimens while keeping their structural integrity intact (Batovska et al, 2021; Carew et al, 2018; Kirse et al, 2022; Marquina et al, 2022; Martins et al, 2019; Martoni et al, 2022; Nielsen et al, 2019). Such protocols roughly consist of leaching DNA from whole individuals by temporarily submerging them in a digestive buffer (Castalanelli et al, 2010; Krosch & Cranston, 2012; Nielsen et al, 2019; Porco et al, 2010; Wong et al, 2014).…”

The spread of Carpophilus truncatus is on the razor's edge between an outbreak and a pest invasion

“…However, we have highlighted that the parameres exhibit a degree of variability that is not considered in any key, and the perception of their shape is strongly affected by the methodology with which this character is evaluated. This variability, together with the molecular correspondence between C. truncatus and Australian samples 29 , 77 , creates a need for the taxonomic position of C. jarijari to be re-evaluated. The unreliability of some morphological discriminating characters in this genus, if used alone, has been known for a long time 32 .…”