Abstract
<jats:p>Sandy beaches and coastal dunes form a narrow ecotone characterised by steep environmental gradients and often intense human use. Efficient biodiversity assessment in these systems requires understanding how community composition varies along the coastline and how much spatial replication is needed for comparable surveys. We used environmental DNA metabarcoding to assess invertebrate diversity from a 55-transect field survey along the North Sea coast. Each transect covered four habitats: lower intertidal, upper intertidal, incipient dunes, and established dunes. We quantified multi-site operational taxonomic unit (OTU) sharing using zeta diversity, examined relationships between OTU detection frequency and transect incidence, and used coverage-based rarefaction and extrapolation to derive habitat-specific benchmarks for sampling effort. Zeta retention declined with increasing order in all habitats, but more slowly in the intertidal than in the dunes, which indicates greater compositional overlap among intertidal transects and greater heterogeneity among dune transects. Detection frequency was positively associated with transect incidence in all habitats, with the steepest slopes in the dunes. Dune assemblages were dominated by low-incidence OTUs, although OTUs detected more consistently across replicate sediment samples also tended to occur in more transects. Intertidal habitats reached approximately 90% sample coverage with 8&ndash;9 transects, whereas established and incipient dunes required 16 and 19 transects, respectively. These findings show that intertidal sediment eDNA assemblages in the study area can be characterised with lower sampling effort, whereas more heterogeneous dune assemblages require denser spatial replication. By linking multi-site sharing, detection&ndash;incidence scaling, and sample coverage, this study provides practical sampling targets for sediment eDNA surveys across beach&ndash;dune ecosystems.</jats:p>