TorpeDNA: a fit-for-purpose eDNA sampling device for marine biodiversity monitoring across applications and scales.

Background: Environmental DNA (eDNA) has revolutionized biodiversity monitoring by enabling species detection from microbes to megafauna without direct observation or capture. Yet most marine eDNA collection systems remain expensive and labor-intensive, limiting their deployment at scale or by non-specialists. To address these challenges, we developed TorpeDNA, a compact, low-cost, and user-friendly eDNA sampler designed to operate under diverse marine conditions, including high-speed towing and citizen-science use. This proof-of-concept study evaluates its performance, scalability, and limitations across three contrasting marine environments encompassing different trophic levels, taxa, and user contexts. Methods: Three case studies were conducted. (1) In Northern Brittany (France), TorpeDNA (20 µm filters) was compared with large-capacity filtration capsule (0.45 µm) for characterising eukaryotic and vertebrates' communities using metabarcoding. (2) Along the French Atlantic coast, TorpeDNA was benchmarked against serial laboratory filtration (10-0.45 µm fraction) to assess bacterial assemblages near aquaculture and wastewater gradients. (3) In the South-West Pacific, TorpeDNA was deployed aboard a Results: Across all studies, TorpeDNA generated biodiversity profiles broadly comparable to those from fine-pore or serial filtration methods. However, some notable differences were observed between methods, largely attributable to the different filter pore sizes and water volumes processed, which influenced the preferential capture or exclusion of host-associated or sediment-bound microbes, larger plankton forms, and/or rare taxa. TorpeDNA effectively recovered dominant microbial, phytoplankton, and zooplankton assemblages, reproducing known ecological and biogeographic gradients, from temperate to tropical plankton turnover and coastal-to-offshore transitions. Vertebrate DNA detection was limited in both methods, likely reflecting low natural eDNA concentrations, insufficient water volume processed, or marker inefficiency and highlighting the need for further optimization of sampling duration, flow rate, and assay sensitivity. Conclusion: TorpeDNA achieves a practical balance between usability, versatility across trophic levels, and integration across spatial and methodological scales. Its robust, low-cost design and compatibility with different filter types and standard molecular workflows make it scalable across platforms, from small to large research vessels to citizen-science initiatives and offering a fit-for-purpose tool for global, inclusive, and scalable ocean biodiversity sampling.