Applying occupancy estimation with detection probability models to environmental DNA survey data for the flattened musk turtle and Black Warrior waterdog

Role of Geodata Crawler:
Geodata Crawler was used to measure watershed characteristics at about 1,000 stream locations where eDNA samples were collected. Watershed charateristics that were measured included slope, base-flow index, stream order, watershed area, canopy cover, impervious survaces, and geologic characteristics. Data were collected at multiple spatial scales: watershed, 300 m site radius, and local-riparian scale with a 300 m site radius and a 100 m riparian buffer. These covariates were used to predict species occupancy and detection probabilities.

de Souza L, Godwin J, Larson ER. 2016. Applying occupancy estimation with detection probability models to environmental DNA (eDNA) survey data for the flattened musk turtle (Sternotheus depressus) and Black Warrior waterdog (Necturus alabamensis). In prep for Biological Conservation.

Funding Source:
Alabama Department of Conservation and Natural Resources, Section 6 grants

Proposal Abstract:
The Black Warrior waterdog (Necturus alabamensis) and flattened musk turtle (Sternotherus depressus) are aquatic species endemic to the upper Black Warrior River basin in Alabama. Their ranges overlap in identical habitat, which includes permanent clear water streams with abundant submerged slab rock, ledges, and crevices. Habitat fragmentation, water quality degradation, and stream impoundments have led to population extirpation and declines for both species. In this study we conducted field surveys and employed eDNA techniques to detect the presence of these aquatic species. We selected sites to survey based on current or historic presence of one or both species, and sampled during both cool and warm seasons. Incorporating eDNA techniques in species surveys enables the ability to rapidly sample long stream reaches or multiple sites where conditions would prohibit conventional trapping, and also facilitates monitoring for rare, cryptic, or otherwise difficult to collect species. eDNA samples were analyzed using qPCR with six replicates per sample. Positive and negative controls were included to ensure no contamination. Samples were considered positive once they passed four criteria. In addition to the field and eDNA surveys, we applied occupancy estimation with detection probability modeling to evaluate factors associated with presence and detectability of both species. In particular, the Black Warrior waterdog is cool-season active, while the flattened musk turtle is warm-season active, and season of sampling affected eDNA detection probabilities for both species. Utilizing various techniques for status surveys will ensure we provide more robust information on species distributions, occupancy, and population trends.

Professional Presentations:
2016. Southeast Partners in Amphibian and Reptile Conservation (SEPARC) Annual Meeting. Nauvoo, Alabama.