MetaBridge project: sediment processing comparisons for animal biodiversity through metabarcoding

Benthic biodiversity surveys are a cornerstone of marine environmental monitoring. As blue economy sectors are expanding into the deep-sea , DNA metabarcoding offers prospects to survey biodiversity in a more scalable and holistic manner. However, sampling and processing methods for sediments still lack standardisation and calbiration compared to traditional morphology-based methods for surveying macroinvertebrate animals . Aiming to evaluate the performance of sample processing strategies typically employed by researchers and environmental consultants, we compared and assessed the recovery and complementarity of the mitochondrial cytochrome c oxidase subunit I (“cox1”, “CO1” or “COI”) and small subunit ribosomal RNA (16S and 18S) markers , across sandy and muddy sediments, of the macro-, meio- and microbenthic communities. Using DNA metabarcoding targeting metazoan (morphology, cox1, 18SV1V2), protist (18SV1V2, 18SV4 ), and prokaryotic (16S) communities sampled at 20 stations around two North Sea oil & gas extraction platforms, we show that standard size-class sorting through sieving (1 mm mesh) proved effective in recovering macrofauna in sandy sediments, albeit with small effect sizes . However, it performed poorly in muddy sediments, demonstrating the need for alternative approaches using smaller mesh sizes, even for recovering macrofauna. While elutriation (enriching for organism sizes >32 µm) improved macro- and meiofauna detection across sediment types, the small effect sizes relative eDNA and the increased processing time and costs emphasize the trade-offs between efficiency and feasibility in large-scale biomonitoring. Assessing community composition of metazoan and microbial organisms across sampling methods showed that sieving amplified non-target planktonic taxa and an associated loss of recovery of benthic taxa in protist and prokaryotic datasets. Elutriation led to similar community compositions as directly targeting eDNAfor prokaryotes, suggesting that comprehensive microbial communities may be recovered if artificial seawater is used . Compared to size-sorted datasets, direct eDNA methods showed poorer recovery and congruence with morphology-based inventories to macrofauna biotic indices. However, direct eDNA extraction from 10g (vortex cell lysis) and triplicates from 0.5g using homogenisation (Precellys cell lysis) of sediment resulted in comparable richness and community compositions across taxonomic compartments. This underlines that direct eDNA methods require increased spatial replication rather than larger individual sample volume to achieve high biodiversity coverage. This work provides a basis for standardizing sampling strategies in eDNA-based benthic surveillance.