Abstract
Detecting rare species is often a necessity for conservation and management, yet challenging for many field survey methods. Environmental DNA (eDNA) is a highly promising solution that has been shown to outperform many established survey methods. Macquarie perch (Macquaria australasica) is an endangered native species that has declined significantly in range and abundance. Detection of M. australasica was compared with an abundant alien fish species (Oncorhynchus mykiss) using eDNA and three conventional survey methods: gill nets, electrofishing and fyke nets. eDNA occupancy estimates for both fish species were compared using four different models to investigate what effect these differences have on false positives and false negatives for the rare and common fish species. These models used unadjusted eDNA detections in water samples, eDNA detections that have been screened using a limit of detection method to remove potential false positives, eDNA data supplemented with a second survey method, or eDNA data augmented with sequencing of positive polymerase chain reaction replicates. eDNA surveying as a single detection method was found to be more efficient and sensitive compared with each capture method separately and combined. Occupancy estimates for the common and rare species did not vary significantly between the four site occupancy-detection models, suggesting that supplementary data may not have as much effect on occupancy estimates compared with other approaches such as temporal or spatial sampling. We conclude that eDNA outperforms the three established survey methods for both a rare and common freshwater fish species. Although there was no significant effect of augmenting eDNA survey methods with other survey data, additional data may improve confidence in detection, and provide confirmatory evidence for unexpected or new detections of a species.
Original language | English |
---|---|
Pages (from-to) | 173-184 |
Number of pages | 12 |
Journal | Aquatic Conservation: Marine and Freshwater Ecosystems |
Volume | 31 |
Issue number | 1 |
Early online date | Nov 2020 |
DOIs | |
Publication status | Published - Jan 2021 |