Distinct coral reef habitat communities characterized by environmental DNA metabarcoding

Abstract

AbstractCoral reefs are biodiversity hotspots, places of high endemicity and provide essential services to billions of people globally. With increasing threats to these reefs worldwide, there is a need to implement faster, more efficient ways to monitor spatial and temporal patterns of biodiversity. Environmental DNA (eDNA) metabarcoding offers a promising tool to address this issue, as it has revolutionized our ability to monitor biodiversity from complex environmental samples such as seawater. However, the capacity for eDNA to resolve fine scale shifts in community composition across habitats in seascapes is yet to be fully explored. Here, we applied eDNA metabarcoding using the rRNA 18S Universal eukaryote assay to explore differences in community profiles between samples collected from the lagoon and reef slope habitats across more than 170 km of the Ningaloo Coast World Heritage Area in Western Australia. We recovered 2061 amplicon sequence variants that comprised of 401 taxa spanning 14 different metazoan phyla such as cnidarians, poriferans, molluscs, algae, worms, and echinoderms. Our results revealed strong clustering of samples by habitat type across the length of the reef. Community dissimilarity (beta diversity) between samples collected from the reef slope and lagoon habitats was high and was driven largely by a strong rate of spatial turnover, indicating a distinct set of taxa representing each reef zone community. We also detected a strong pattern of isolation by distance within our slope samples, suggesting that communities are spatially stratified across the length of the reef. Despite high connectivity due to regular flushing of the lagoon environment, our results demonstrate that metabarcoding of seawater eDNA from different habitats can resolve fine scale community structure. By generating multi-trophic biodiversity data, our study also provided baseline data for Ningaloo from which future changes can be assessed.