Of Biogeography, Fishes and Kelp: Environmental DNA Metabarcoding the Great African Seaforest

Abstract

Aim

The biodiversity associated with kelp forest ecosystems is understudied globally, including in South Africa, where kelp forests have expanded their range under climate change. This study used environmental DNA (eDNA) metabarcoding to assess fish biodiversity across an expansive range (~1000 km) of South Africa's kelp forests, contributing to a contemporary fish biodiversity baseline. We examined whether this baseline reflects existing biogeographical patterns, including the leading edge of an expanding kelp forest.

Location

South Africa.

Methods

We conducted eDNA metabarcoding, targeting the mitochondrial 12S rRNA gene for fishes, on 192 aquatic eDNA samples collected at eight sites spanning two ecoregions (Southern Benguela and Agulhas) and three kelp forest types (Namaqua, Cape and Agulhas). Intra-site spatiotemporal samples were collected to capture variability in eDNA signals at small spatial (< 600 m) and temporal (~24 h) scales.

Results

In total 140 operational taxonomic units (OTUs) from 39 marine fish families were detected, with 40 OTUs resolved to genus and 25 to species, encompassing a wide range of functional groups. OTU richness increased from west to east and differed significantly (p < 0.05) between ecoregions. Communities were significantly different (p < 0.05) across all biogeographical and intra-site predictors. While OTU accumulation curves did not fully saturate, spatiotemporal sampling at sites with lower species richness captured a larger proportion of extrapolated diversity. Notably, De Hoop, located at the kelp forest's leading edge, exhibited particularly distinct community patterns.

Main Conclusions

This research demonstrates that eDNA metabarcoding effectively detects diverse fish communities in a marine biodiversity and climate change hotspot, revealing biogeographic and local structuring patterns. However, gaps in the 12S reference database limit taxonomic resolution, highlighting the need to expand reference sequences. Continued development of eDNA metabarcoding across a range of spatiotemporal scales will enhance our understanding of biodiversity dynamics in South Africa's coastal ecosystems.