Performance of eDNA capture methods for monitoring fish biodiversity in a hyper-tidal estuary

Environmental DNA (eDNA) has become an established and efficient method for monitoring biodiversity in aquatic systems. However, there is a need to compare and standardise sampling methods across ecosystem types, particularly complex ecosystems such as estuaries where unique challenges for monitoring fish populations are present due to fluctuating environmental factors. Here, we compare fish biodiversity metrics obtained from eDNA metabarcoding data using four different eDNA filtering methods: three manual filtering methods with different pore sizes (0.45, 1.2 and 5 µm) and a newly established passive method, the metaprobe. The study was applied across a salinity gradient in a hyper-tidal estuarine ecosystem. Overall, 44 fish species were detected across the four methods used. The 0.45 µm filter recovered the highest richness (39 species), then the metaprobe method (35), followed by the 1.2 µm (34) and 5 µm (33) filters. Filter performance between salinity gradients revealed that the 0.45 µm and the 1.2 µm methods recovered the highest species richness across all sampled zones. The 0.45 µm also had the most consistent detection probabilities using representative species from each zone. While the 0.45 µm method appeared to be the optimal method, each of the methods can be considered as a viable and comparable option for biomonitoring in dynamic ecosystems such as estuaries and rivers. In particular, the passive metaprobe (used in a freshwater system for the first time here) performed well in comparison to the manual filtering methods despite a short deployment time. This study provides critical insights for optimising fish biodiversity assessments using eDNA metabarcoding in estuarine ecosystems, providing a valuable framework for future monitoring efforts in similar systems worldwide.