Optimizing Marine eDNA Protocols for Rare Megafauna: A Case Study on Dugongs in the Western Indian Ocean

Assessing habitat use, spatial distribution, and site fidelity is critical for the conservation of threatened marine species, yet monitoring rare and elusive species in these types of complex environments remains difficult. This limitation is particularly critical for dugong (Dugong dugon) populations in the Western Indian Ocean, which are extremely small, fragmented, and difficult to monitor, posing a major conservation challenge. This pilot study aimed to develop a cost-effective and field-optimized eDNA workflow based on species-specific qPCR for large-scale monitoring of this threatened sirenian in shallow tropical coastal ecosystems. Field sampling was conducted in Mozambique, which hosts the only remaining large dugong population in the region. Using a species-specific qPCR assay independently developed and validated for this study, we evaluated how sampling design and sample processing influence dugong eDNA detection under natural conditions. Sampling was conducted along survey tracks using different sampling and DNA extraction strategies to evaluate three key variables of the detection protocol-(i) vertical sampling position (surface, near-bottom, and sediment), (ii) filtration volume (1.5, 30, and 60 L), and (iii) field versus laboratory DNA extraction-resulting in a total of 11 sampling-extraction strategies. At every visually confirmed location, at least one of the 11 sampling-extraction strategies resulted in a positive detection, including up to 2 h after the last visual sighting. All strategies detected dugong DNA except those based on sediment sampling. qPCR detection was influenced by an interaction between filtration volume and vertical sampling position: at low filtration volume (1.5 L), detection was higher near the bottom than at the surface, whereas this difference disappeared at 60 L. Conversely, increasing filtration volume improved detection at the surface but had little effect near the bottom. Laboratory extractions markedly outperformed field extractions. The sampling strategy yielding the highest detection rate-large-volume near-bottom filtration followed by laboratory-based sample processing-detected dugong DNA in all cases where a visual sighting had occurred. Overall, these results show that filtration volume can partly compensate for differences in detection associated with vertical sampling position and provide a robust and field-applicable framework for eDNA-based dugong monitoring, supporting scalable and long-term surveillance efforts in tropical coastal ecosystems.