Key points are not available for this paper at this time.
ABSTRACT There is an expanding body of evidence that environmental DNA (eDNA) can serve as a reliable alternative to traditional assessments of biodiversity and ecological quality. Riverine benthic ecosystems represent one such habitat, holding significant promise for ecological health evaluations using eDNA. Diatoms have typically been assessed in these environmental biofilms through both molecular and conventional methods. However, a wide diversity of life has not been targeted previously, which may serve as important indicators of water quality. To be fully integrated into existing monitoring programs, it is essential to demonstrate the reliability of eDNA‐based assessments. This entails developing unbiased methodologies that capture total DNA across the entire community. DNA extraction from environmental samples is critical in analyzing microbial communities; nevertheless, current workflows often focus on individual kingdoms or communities. In this study, we investigated how extraction methodologies can bias the analysis of microbial community composition using amplicon sequencing at a cross‐kingdom level in river phytobenthos samples. We tested four commercially available DNA extraction methodologies on 23 freshwater benthic biofilm samples collected across a pH and conductivity gradient. Quantitative PCR and metabarcoding of four amplicons (16S, 18S, ITS, and rbcL), targeting bacterial, eukaryotic, fungal, and phototrophic communities, were employed to assess the impact of the DNA extraction kits on community evaluation. This study revealed a high level of similarity between methods incorporating mechanical lysis, which exhibited higher PCR and sequencing success rates as well as increased cross‐kingdom richness and differential abundance compared to chemical and enzymatic lysis alone. However, the origin of the samples, rather than the extraction methodology, emerged as the most significant factor linking them. We recommend utilizing mechanical lysis to optimize cross‐kingdom recovery from environmental samples. Nonetheless, the strong correlation between sample origin and extraction method implies that existing data gathered through alternative methodologies remain valid for informing future monitoring practices.
Newbold et al. (Thu,) studied this question.