Sensitive detection methods are essential for weed biosecurity monitoring, because weed populations are small at the onset of a new incursion, and again during final stages of eradication. Detection methods based on environmental DNA (eDNA) are reportedly highly sensitive to low density target organisms in sample areas. We trialled the use of eDNA for detection of the Prohibited Matter freshwater weed Amazon frogbit, Limnobium laevigatum (Humb. & Bonpl. ex Willd.) in waterways, during eradication activities on four infestations in NSW: two in ponds and lagoons, and two in intermittently flowing streams.
We developed a species-specific quantitative PCR (qPCR) assay for detection of Amazon frogbit and coupled this with our universal qPCR assay to ensure quality control of the testing process. We found that Amazon frogbit could be consistently detected at all sites where it was present. EDNA could also be detected at intermittent stream sites downstream from infestations. At sites upstream from infestations we detected no Amazon frogbit eDNA, while positive results for the universal assay ensured that non-detections were not the result of technical failures. At sites in ponds and lagoons, we found that remnant DNA of the weed remained as a legacy of its presence for at least 8 months after the last plants were observed. At sites in intermittently flowing streams, legacy eDNA was detectable for similar durations but at lower concentrations than in lakes and lagoons.
We conclude that our eDNA method is sensitive and reliable for detection of Amazon frogbit infestations in various waterways, but it can give false inferences of weed presence following eradication due to longevity of remnant DNA in waterways. Our ongoing research is aimed at developing environmental genetic methods that detect live plants only, avoiding the confounding signals of legacy eDNA. This would enable use of eDNA as a tool to demonstrate proof-of-freedom.
Keywords: surveillance, molecular detection, eDNA degradation, eDNA persistence