APPLICATIONS OF ENVIRONMENTAL DNA METABARCODING ACROSS DIFFERENT MARINE ENVIRONMENTS

The use of environmental DNA (eDNA) metabarcoding to reveal marine biodiversity in water samples has been widely tested and applied. However, its effectiveness across different marine and transitional environments and management contexts is still under investigation. This thesis addresses the challenge of interpreting eDNA signals as management-relevant information, particularly in situations where conventional surveys are costly or logistically challenging. Three case studies are presented, examining a coastal lagoon, a network of marine protected areas and an offshore Fish Aggregating Device (FAD). The objective was to determine whether eDNA can: (i) capture benthic community structure from surface waters in a tidally influenced lagoon, (ii) investigate biodiversity patterns consistent with conservation zoning across multiple Italian MPAs once strong spatial effects are accounted for, and (iii) provide complementary detections when comparing active filtration with passive samplers around a tropical FAD, including at different depths. In the lagoon case study, surface eDNA provided a picture of benthic assemblage, even though only superficial water was collected. Clear spatial differentiation was detected, which was consistent with known environmental gradients. The time series also revealed significant temporal variability. Several non-indigenous taxa were detected, illustrating the value of eDNA for early warning in transitional systems where benthic sampling is difficult. In the multi-MPA study, location emerged as the primary driver of community differences; after accounting for this strong spatial signal, protection level showed a smaller but detectable association with community composition. Conventional richness alone was an ineffective indicator of zones, whereas phylogenetic diversity was a more reliable predictor of higher levels of protection. In the offshore FAD study, active filtration and passive samplers produced overlapping yet complementary assemblages. Both approaches recovered predictable depth structuring and substantial sampling day variability, indicating that replication in time and along depth is essential. Each method contributed unique detections, supporting their combined use at remote offshore sites with limited data. This thesis shows that eDNA is most informative when study design fits both the place and the question. Across our case studies, we examined how spatial and temporal replication, the sampling position in the water column, and the mode of sample collection alter the biodiversity signal. These effects varied in size and direction across systems, and were sometimes considerable. We therefore recommend adjusting replication in space and time, choosing an appropriate depth or layer, and selecting the collection approach to match the site and objective. The cases presented in this thesis are intended to help build clear, objective studies that establish baselines, enable comparisons across sites or periods, and support management decisions without over-interpreting the data.