Environmental DNA reflects spatial distribution of a rare turtle in a lentic wetland assisted colonisation site

<p>Conservation translocations require robust post-release monitoring to evaluate their success, which can be challenging to implement and maintain. Monitoring techniques that can account for the dispersal and cryptic nature of translocated animals are necessary to provide critical information on persistence and distribution. In this study, we developed a highly sensitive environmental DNA (eDNA) assay specific to the Critically Endangered western swamp turtle (<em>Pseudemydura umbrina</em>), a species currently undergoing trials of assisted colonisation. Actively filtering sufficient volumes of water in lentic systems is difficult due to high concentrations of clogging particulates, therefore we assessed the viability of passive sampling in a controlled environment by submerging filter membranes and directly extracting DNA. Active sampling detected <em>P. umbrina</em> with a 97.6% detection rate, whereas passive sampling resulted in an 8.3% detection rate. We then used a fine-scale eDNA sampling design and radio tracked translocated <em>P. umbrina</em> at the assisted colonisation wetland to investigate eDNA dispersal and spatial monitoring resolution. We detected <em>P. umbrina</em> at 42% (7 / 17) of eDNA sample sites, and the probability of a positive eDNA detection was negatively associated with the distance of <em>P. umbrina</em> from the sampling site, indicating limited eDNA dispersal from the source. Systems with low natural mixing and limited eDNA dispersal provide an opportunity for high resolution spatial and temporal monitoring via targeted eDNA approaches. This is beneficial for monitoring rare species in these systems, as such high-resolution results can provide insights on species presence, distribution, and microhabitat use.</p>