Abstract
We are in the midst of an ongoing sixth mass extinction, where thousands of critically endangered species have been lost in the last century. This rapid collapse of biodiversity is predominantly driven by a dramatic increase in the human population, consumption of resources and land use change. There is a dire need for immediate global action against this pattern of decline in biodiversity. Amphibians are the most threatened vertebrate group worldwide, with an estimated 40% at risk of extinction. This group are dependent on very specific habitat conditions and as such are vulnerable to even minor changes in their abiotic environment. New Zealand’s three species of native Leiopelma are listed as both nationally and globally threatened. Gathering accurate population data is paramount for effective conservation actions against such threatened species. Effective management of these species is hampered by the limited population abundance data.
In this thesis, the potential use of a trained detection dog for Leiopelma was investigated. Detection dogs are recognised as effective tools for monitoring endangered species. A detection dog trained on Leiopelma may greatly improve our ability to accurately gather population data and improve conservation actions for these species.
After undergoing rigorous training, a detection dog was able to accurately and consistently locate Leiopelma under testing conditions. The dog scored the best possible scores for efficiency and sensitivity, with reliability and correct indication rates of 100%, and a miss rate of only 0.05% during performance testing. This study provides a proof of concept for Leiopelma detection in the field for conservation purposes. The training of a Leiopelma detection dog marks the beginning of an effective new mode of monitoring and surveying of New Zealand’s native frogs.
The mode of communication used by Leiopelma is currently unknown. Unlike other anurans, these species do not use acoustics or visual cues. It has been theorised that these species may be communicating via pheromones, used for intra-species communication. The fact that the detection dog easily located Leiopelma suggests that Leiopelma emits volatile organic compounds (VOCs). Therefore, the VOC profiles of Leiopelma were characterised using gas chromatography-mass spectrometry (GC-MS) with headspace sampling with solid phase microextraction (SPME). Five measurable compounds (Pentanal, D-Limonene, (+)-Cuparene, trans-Geranyl acetone, and α-Pinene) of interest were identified, however, no consistent patterns or difference in species-specific VOCs were identified.
To transition the Leiopelma detection dog to working in field conditions, GetXent tubes, a new commercial detection training aid, were investigated to determine their suitability as detection training aids for use on Leiopelma. The GetXent tubes were unable to adsorb and/or emit consistent and measurable VOCs from the headspace of Leiopelma as measured by GC-MS. Lindt Dark chocolate, which has a known and stronger VOC profile, was subsequently used to evaluate whether the tubes could absorb and release VOC for use as a training aid. As these tubes were unable to absorb representative VOC profiles from either Leiopelma or chocolate, it can be concluded that these tubes are ineffective as detection training aids.
The results in this thesis suggest that detection dogs are more sensitive for detecting Leiopelma VOCs than GC-MS analysis. Therefore, the failure to identify species-specific VOC profiles in our GC-MS analysis does not rule out species-specific VOC profiles or the use of pheromone communication by these species. The results outlined in this study highlight the continued value of dogs for detection work, not only in conservation, but in any field of detection where the target odours are trace or complex. Given the inaccurate sampling of both Leiopelma and chocolate VOCs exhibited by GetXentTM tubes in this study, the most promising approach for training a Leiopelma detection dog is to train with live specimens. It is hoped that the detection dog trained during the course of this study will improve Leiopelma monitoring and population data, in turn improving the quality of conservation actions directed at these species.