Active Versus Passive eDNA Sampling Across Depths at a Tropical Fish‐Aggregating Device ( FAD )

Environmental DNA (eDNA) is a valuable tool for monitoring fish biodiversity, particularly in pelagic environments, where con-ventional methods are difficult to implement. Fish aggregating devices (FADs), floating structures widely used by tropical tunafishers, are known to attract numerous pelagic fish species, making them ideal sampling locations. In this study, we assessed pe-lagic fish diversity at a FAD off the coast of Bali, Indonesia, using eDNA metabarcoding. We compared different eDNA samplingstrategies with the primary aim of evaluating the effectiveness of eDNA metabarcoding for assessing pelagic fish biodiversity,while testing and refining accessible, low-cost protocols suitable for remote and logistically constrained environments. Samplingwas conducted over three consecutive days at four depths (1, 10, 40, and 60 m) using two distinct eDNA collection methods:active filtration and a custom-designed passive system consisting of 3D-printed cylinders filled with sterile gauze, mounted onunit holders that allow the simultaneous deployment of triplicate samples at each depth. A total of 66 samples were collected, andmetabarcoding was performed using an available primer pair targeting fish 12S mitochondrial DNA, with sequencing performedon an Illumina NovaSeq platform. Across both sampling methods, a total of 39 fish Operational Taxonomic Units (OTUs) weredetected, of which 25 were shared between active and passive approaches, and 31 were assigned to the species level. The twosampling methods yield overlapping assemblages dominated by epipelagic taxa commonly reported at FADs, indicating that bothapproaches are suitable for characterizing fish communities in this environment. Importantly, the integration of passive and ac-tive sampling provided a practical balance between deployment flexibility and taxonomic detection. Beyond technical validation,standardized eDNA protocols can be effectively transferred and applied in contexts where logistical constraints are extremelyrelevant, thereby supporting the development of biodiversity monitoring programs centered around ecologically relevant featuressuch as FADs.