Improving eDNA detection probabilities for monitoring aquatic species

Abstract

Reliable monitoring data is important for managing biodiversity. Yet this reliability is largely dependent on the efficiency of the methods used to detect individuals or species. When species are rare, cryptic or elusive, sensitive tools are needed to ensure adequate detection rates. Much interest has been given to the application of environmental DNA (eDNA) for targeted species detection and biodiversity assessments. This method’s sensitivity, non-invasiveness and rapid collection of samples makes it an appealing tool for monitoring, but despite the rising interest, there are many factors that can impact upon eDNA detection and these require investigation. This research examines several factors affecting eDNA detection in the field and the laboratory in order to improve eDNA detection probabilities. Specifically, field sampling and choice of methods during analysis were examined to show how these affected DNA recoveries. Spatial and temporal patterns of eDNA production, degradation and transport were also investigated to understand eDNA dynamics in natural systems and how these relate to estimates of species abundance. Through field investigations, the results of traditional detection were compared with eDNA detection to objectively assess each method’s performance. This research focused on lotic systems and used three invasive fish species of management concern as target species. The research found that the choice of DNA capture, preservation and extraction methods can significantly affect DNA yield and that the most commonly used eDNA methods are not necessarily the most cost-efficient. By investigating three fish species with contrasting spatial distributions (benthic, pelagic, benthopelagic) across two seasons, a significant effect of season on eDNA concentration was found, with higher DNA detection rates in spring than autumn. Sampling location (surface vs subsurface water sampling) did not significantly affect eDNA concentration in all three species studied. These findings underscore the importance of considering sampling and analytical methods as well as the target species’ ecology and behaviour to maximize eDNA detection success. Improving detection rates require a good understanding of eDNA production, persistence and transport. From the field experiments, the research herein demonstrated that time-dependent changes in eDNA production seen in closed systems was demonstrable in flowing systems, but only when water samples were taken close to the source (within 150 m). There is increased stochasticity in eDNA detection when target species are present at low densities, making it challenging to draw inferences on abundance and proximity. However, because eDNA concentration drops off quickly only a few meters from source, strong and consistent eDNA signals strongly suggests local presence when species are rare. This thesis found that the eDNA method can significantly improve the detection of rare, elusive species, but like any other survey tool, its detection is imperfect. Thus, it is best used in conjunction with other survey methods to enhance detection rates and increase confidence in the monitoring results. This thesis underscores the importance of considering sampling and analytical methods, species’ ecology and behavioural patterns as well as the potential effects of physical site characteristics to accurately draw inferences from eDNA data and improve eDNA detection success.

Date of Award	2018
Original language	English
Supervisor	Dianne GLEESON (Supervisor), Elise FURLAN (Supervisor), Mark Lintermans (Supervisor) & Heleena Bamford (Supervisor)