Abstract
Environmental DNA (eDNA) metabarcoding is a promising tool for monitoring marine biodiversity, but remains underutilised in Africa. In this study, we evaluated the ability of aquatic eDNA metabarcoding as a tool for detecting biodiversity associated with a South African kelp forest, an ecosystem that harbours high diversity of species, many of which are endemic, but are also sensitive to changing environmental conditions and anthropogenic pressures. Using fine-scale spatial (1 m and 8 m) and temporal (every four hours for 24 h) sampling of aquatic environmental DNA and targeting two gene regions (mtDNA COI and 12S rRNA), metabarcoding detected 880 OTUs representing 75 families in the broader metazoan community with 44 OTUs representing 24 fish families. We show extensive variability in the eDNA signal across space and time and did not recover significant spatio-temporal structure in OTU richness and community assemblages. Metabarcoding detected a broad range of taxonomic groups, including arthropods, ascidians, cnidarians, echinoderms, ctenophores, molluscs, polychaetes, ichthyofauna and sponges, as well as Placozoa, previously not reported from South Africa. Fewer than 3% of OTUs could be identified to species level using available databases (COI = 19 OTUs, 12S = 11 OTUs). Our study emphasizes that kelp-forest associated biodiversity in South Africa is understudied, but that with careful consideration for sampling design in combination with increased barcoding efforts and the construction of regional databases, eDNA metabarcoding will become a powerful biomonitoring tool of kelp-forest associated biodiversity.