How many replicates to accurately estimate fish biodiversity using environmental DNA on coral reefs?

Abstract

Quantifying the diversity of species in rich tropical marine environments remains challenging. Environmental DNA (eDNA) metabarcoding is a promising tool to face this challenge through the filtering, amplification, and sequencing of DNA traces from water samples. However, the reliability of biodiversity detection from eDNA samples can be low in marine environments because eDNA density is low and certainly patchy in this vast, heterogenous and dynamic environment. So, the number of sampling replicates and filtered volume necessary to obtain accurate estimates of biodiversity in rich tropical marine environments using eDNA metabarcoding is still unknown. Here, we used a paired sampling design of 30L per replicate on 68 reef transects from 8 sites in three tropical regions and identified fish Molecular Taxonomic Units (MOTUs) using a 12S marker. We quantified local biodiversity variation as MOTU richness, compositional turnover and compositional nestedness between replicated pairs of seawater samples. We report strong turnover of MOTUs between replicated pairs of samples undertaken in the same location, time, and conditions. Paired samples contained non-overlapping assemblages rather than subsets of one-another. As a result, localised diversity accumulation curves showed that even 6 replicates (180L) in the same location underestimated local diversity (for an area <1km). However, sampling of regional diversity using ~25 replicates in variable locations (often covering 10s of km) achieved saturation of biodiversity accumulation curves. Our results demonstrate high variability of diversity estimates perhaps arising from heterogeneous and local distribution of eDNA distribution in seawater or highly skewed frequencies of eDNA traces. This high compositional variability has consequences for using eDNA to monitor temporal and spatial biodiversity changes of local assemblages. Future biomonitoring efforts could be strongly undermined by a high level of false-negative detections under low replication protocols. We reveal the need to increase replicates or increase sampled water volume to better inform management of marine biodiversity using eDNA.

Cross-ocean patterns and processes in fish biodiversity on coral reefs through the lens of eDNA metabarcoding

“…However, our site-based and station-based accumulation curves do not reach plateaus suggesting that our sampling effort was not sufficient to exhaustively estimate fish biodiversity for each site (electronic supplementary material, analyses, figure S5) and station (electronic supplementary material, analyses, figure S6). Twenty-five replicates (so, 12 stations in case of field duplicates) could accurately estimate biodiversity regionally due to high local turnover [53]. A higher number of eDNA samples would be necessary here to reach MOTU accumulation per site and station.…”

Disentangling the components of coastal fish biodiversity in southern Brittany by applying an environmental DNA approach

“…However, replication quality and sampling effort may vary according to the studied environment because biotic and abiotic conditions can influence degradation, deposition and detection of eDNA (Stewart, 2019). For example in highly diverse tropical marine ecosystems where eDNA shedding and decay rates tend to be higher, partially due to higher water temperature (Jo et al, 2019), a study by Stauffer et al (2021) highlighted strong dissimilarity in species composition between filtration replicates (e.g., dissimilarity β jac mean = 0.729 ± 0.102 in the West Indian Ocean; β jac mean = 0.528 ± 0.146 in the Caribbean Sea, with a major contribution of the turnover component for both sampling areas) and required additional sampling efforts to ensure robust biodiversity estimates.…”

Airborne environmental DNA captures terrestrial vertebrate diversity in nature

“…It further emphasises the need for incorporation of field replicates, as is recommended for other eDNA substrate types 36 . Studies of airborne eDNA monitoring of insects have not incorporated and contrasted field replicates 37,38 , however in studies on marine eDNA, a similar patchy distribution of DNA has been observed with paired field replicates not detecting the exact same community 39,40 . In the present study, we found no differences in the vertebrate composition between microhabitats, nor sampling sites.…”