Abstract
Accurate species detection underpins conservation by informing distribution, population trends, and community composition, which guide monitoring and management strategies. As biodiversity declines globally, there is an urgent need for detection tools that can be used to efficiently survey large areas while minimising logistical and financial costs. Reptiles are particularly vulnerable to decline, with 21% of species worldwide facing extinction and 86.2% of Aotearoa New Zealand species classified as Threatened or At Risk. Detecting reptiles is challenging due to their cryptic behaviour and the coarse, resource-intensive nature of traditional detection methods. Environmental DNA (eDNA) offers a promising alternative, enabling detection of elusive or low-density species without direct observation and providing a faster, less labour-intensive, and potentially more cost-effective approach. However, from a global perspective, its application to terrestrial reptiles remains limited, and few studies have quantitatively compared eDNA with traditional methods.
I investigated the feasibility of using eDNA to detect lizards in terrestrial environments across various species and population densities and quantitatively compared the detection sensitivity and cost-efficiency with traditional techniques. Seven eDNA tools were trialled within a predator-proof grassland enclosure containing only the Threatened Kapitia skink (Oligosoma salmo), including six modified from existing lizard detection devices and one commercially developed by Wilderlab Ltd. Three tools: customised tracking tunnels, pipes, and artificial cover object (ACO) swabs, were the most efficient at detecting lizard eDNA. These were further tested across high-, medium-, and low-density skink populations within the enclosure and at two field sites with multispecies assemblages: a remnant indigenous shrubland in Canterbury and a subalpine open shrubland on Rakiura. I also compared the detection sensitivity of the eDNA ACO swab method with the traditional visual ACO method on Rakiura and evaluated actual and projected costs for eDNA and visual ACO surveys across one front-country and two back-country sites.
My results demonstrate that terrestrial eDNA tools reliably detect lizard presence across various species, habitats, and population densities, including at low-density. Compared with traditional visual ACO surveys, eDNA achieved higher detection sensitivity and greater cost-efficiency, identifying more species more frequently and reducing survey costs by up to 3.2-fold. These efficiencies were largely driven by reduced staffing requirements, including fewer personnel, shorter field time, and reliance on field workers rather than specialist herpetologists; advantages that were particularly pronounced in remote environments where repeat visits are challenging.
My research provides the first comprehensive evaluation, and comparison of multiple terrestrial eDNA tools for lizard detection in Aotearoa’s grassland habitats, and the first quantitative assessment of their detection performance and cost-efficiency relative to a traditional method. The findings demonstrate that eDNA offers a scalable, non-invasive, and cost-effective lizard detection method. The three eDNA tools have broad applications for species detection, range delineation, habitat-use assessments, long-term monitoring, and citizen science. Integrating eDNA with conventional methods may provide the most robust strategy, where eDNA guides placement of traditional methods needed to obtain demographic and abundance data. These tools offer wide utility in conservation, biosecurity, ecological research, and ecological impact assessments, with broad applicability across other environments and reptile taxa globally.