Monitoring priority British mammals with environmental DNA methods
File(s)
Author(s)
Priestley, Victoria
Type
Thesis
Abstract
Environmental DNA methods have the potential to be game-changing for surveys of rare and elusive species, with just a few copies of eDNA shed into the environment enough to reveal presence. Proof of concept studies show that mammal eDNA is detectable in water, but the probability of detection is prohibitively low (averaging 23%). This thesis aims to bridge the research-application gap by developing and testing methods for three conservation priority species: the hazel dormouse Muscardinus avellanarius; Eurasian otter Lutra lutra; and water vole Arvicola amphibius.
A multiple linear regression model of eDNA detection probability was developed, validated with data for 27 British mammals, and used to predict and understand detection probabilities for the most threatened British mammals. Species-specific assays were designed to amplify short fragments of mtDNA by quantitative PCR and in combination with novel eDNA sampling methods directly compared with traditional survey techniques: a clean substrate to line survey equipment and ‘catch’ urine as a sample; passive samplers in water, soil associated with field signs and a device to sample the entire water column.
The model explained 44% of the variation in detection probability with just three independent variables: population density, weight, and time spent in the water and streamlines our understanding of, and ability to predict, the availability of eDNA in water for mammals. Environmental DNA methods outperform traditional approaches for the detection of all three species. For surveys of hazel dormouse, there was a 12-fold efficiency saving, even with sex-specific differences in eDNA availability in urine. For water vole, eDNA surveys worked outside of the accepted survey season and were less likely to result in false- negative results. Eurasian otter eDNA is no longer elusive, with detection enhanced by a new hydrolysis probe. Results indicate that soil samples can validate Eurasian otter presence where field signs are inconclusive. For water, passive samplers made of sand are as good as surface samples at detecting otter eDNA but filtering at intervals <600m on at least two consecutive days and from the entire water column is the most sensitive.
For the detection of British mammals by eDNA, this thesis contributes a probability of detection model, new methods for three priority species, and a decision tool to support future research in this field. The research concludes by encouraging others to present and market eDNA methods, not only to the academic community but in a relatable way for policy decision-makers. Only by doing so can research be translated into practice at a meaningful pace. A jointly owned; by researchers and policy decision- makers, probability of eDNA detection tool would unify best practices in eDNA research and policy. The strategic aim for eDNA research should be that it fades to a footnote as one of many tools applied to address the grand challenges in conservation science.
A multiple linear regression model of eDNA detection probability was developed, validated with data for 27 British mammals, and used to predict and understand detection probabilities for the most threatened British mammals. Species-specific assays were designed to amplify short fragments of mtDNA by quantitative PCR and in combination with novel eDNA sampling methods directly compared with traditional survey techniques: a clean substrate to line survey equipment and ‘catch’ urine as a sample; passive samplers in water, soil associated with field signs and a device to sample the entire water column.
The model explained 44% of the variation in detection probability with just three independent variables: population density, weight, and time spent in the water and streamlines our understanding of, and ability to predict, the availability of eDNA in water for mammals. Environmental DNA methods outperform traditional approaches for the detection of all three species. For surveys of hazel dormouse, there was a 12-fold efficiency saving, even with sex-specific differences in eDNA availability in urine. For water vole, eDNA surveys worked outside of the accepted survey season and were less likely to result in false- negative results. Eurasian otter eDNA is no longer elusive, with detection enhanced by a new hydrolysis probe. Results indicate that soil samples can validate Eurasian otter presence where field signs are inconclusive. For water, passive samplers made of sand are as good as surface samples at detecting otter eDNA but filtering at intervals <600m on at least two consecutive days and from the entire water column is the most sensitive.
For the detection of British mammals by eDNA, this thesis contributes a probability of detection model, new methods for three priority species, and a decision tool to support future research in this field. The research concludes by encouraging others to present and market eDNA methods, not only to the academic community but in a relatable way for policy decision-makers. Only by doing so can research be translated into practice at a meaningful pace. A jointly owned; by researchers and policy decision- makers, probability of eDNA detection tool would unify best practices in eDNA research and policy. The strategic aim for eDNA research should be that it fades to a footnote as one of many tools applied to address the grand challenges in conservation science.
Version
Open Access
Date Issued
2022-01-06
Date Awarded
01/09/2023
Copyright Statement
Creative Commons Attribution NonCommercial NoDerivatives Licence
Advisor
Savolainen, Vincent
Arnold, Richard
Sponsor
Thomson Environmental Consultants (Firm)
Natural Environment Research Council (Great Britain)
Publisher Department
Life Sciences
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)