2026/07/01 by Samuel Lewis, Sally Maxwell, Michael Shackleton +8
paper · doi:10.1002/aqc.70433
ABSTRACT Floodplain wetlands are declining globally in extent and ecological condition. Changing climates, increasing agricultural activity and invasive species are among the key threats contributing to biodiversity loss, with native fish species increasingly in decline and listed as threatened or endangered. Floodplain wetlands present unique sampling challenges requiring bespoke methods to detect species across complex habitats. Environmental DNA (eDNA) metabarcoding has rarely been assessed against traditional fish sampling techniques in floodplain wetlands. We compared seine netting, fyke netting and eDNA metabarcoding across 34 floodplain wetlands in south‐eastern Australia. We assessed the effects of known constraints to eDNA detection: turbidity (10–450 NTU) and sample volume (2.3–7.6 L) on metabarcoding taxa richness. eDNA provided the most comprehensive detection of fish taxa overall. Common, abundant taxa were reliably detected by all methods, whereas rare species were predominantly detected using eDNA. Turbidity constrained filtration volume through filter clogging but did not compromise species detection; filtered water volume was not correlated with species richness. Sequence read depth showed a significant positive correlation with netting catch‐per‐unit‐effort at the species level, suggesting potential utility for ranking sites by relative fish abundance or biomass. eDNA provides complementary information to traditional netting methods, detecting species missed by physical capture while detecting all species captured by nets. This makes eDNA particularly valuable as a reconnaissance tool for initial biodiversity assessment and spatial prioritisation of sampling effort, complemented by physical capture to assess population demographics including age structure, size distributions, condition and sex ratios essential for understanding recovery and decline.