Assessment of stream macroinvertebrate communities with eDNA is not congruent with tissue‐based metabarcoding

Abstract

Freshwater habitats are particularly sensitive to changes caused by anthropogenic stress, including habitat degradation, pollution and alteration of water inputs and flow (Dudgeon et al., 2006).Bioindicators are species or taxonomic groups that respond predictably to environmental changes (Cairns & Pratt, 1993) and are often used to evaluate the degree of disturbance or habitat impairment (Norris & Hawkins, 2000). Aquatic macroinvertebrates are commonly used bioindicators in both lentic and lotic systems (Bonada, Prat, Resh, & Statzner, 2006;Cairns & Pratt, 1993); however, they present challenges to traditional morphological identification as some immature stages cannot be reliably identified. The difficulty of identifying juvenile or damaged specimens can result in a significant

Improved freshwater macroinvertebrate detection from environmental DNA through minimized nontarget amplification

“…Analysis with universal primer pair BF2/BR2 recovered mostly nontarget taxa, especially diatoms, similar to reported observations from other studies using universal COI primers (Beentjes et al., 2019; Deiner et al., 2016; Gleason et al., 2020; Hajibabaei, Porter, Robinson, et al, 2019; Macher et al., 2018), leaving only a small number of reads and OTUs of target taxa recovered from the extracted eDNA. The amplification of nontarget DNA from extracted eDNA samples was consistently observed across all 20 samples.…”

“…A similar comparison was performed by Gleason et al. (2020) in Canada, also reporting only 12% of the eDNA reads being assigned to benthic invertebrate taxa as opposed to 99% of the reads in bulk samples. Similarly, Beentjes et al.…”

Ethanol eDNA Reveals Unique Community Composition of Aquatic Macroinvertebrates Compared to Bulk Tissue Metabarcoding in a Biomonitoring Sampling Scheme

“…Once samples were filtered (approximately 4-20 min per sample), we placed the filter in a sterile petri dish and stored in the samples at −80 • C until DNA extraction. We used a CTAB extraction protocol for ethanol eDNA (see [25,29]), quantified and normalized DNA concentration of all samples to 5 ng/µL.…”

“…This can account for the absence of some species and a subsequent reduction in biodiversity observed in previous studies comparing sampling methods [24]. While some research suggests that eDNA metabarcoding water samples is an appropriate substitute for kick-net sampling (e.g., [15,21,22]), other studies have preferred bulk sample metabarcoding [24,25]. Gleason et al [25] compared invertebrate community composition between tissue samples and eDNA samples and found that eDNA mostly amplified algae, with minimal overlap in invertebrate community composition between paired bulk tissue and eDNA samples collected from the same location.…”

“…Zizka et al [20] found that not enough DNA from mollusc families was being released, resulting in them being undetected. However, these comparisons have not used the same analysis method as Gleason et al [25] that compares the different techniques at the local sample level, as opposed to the regional pooled study scale. The local sample level is important to more accurately distinguish differences in community composition between methods.…”