Answers are grounded in this paper and cite their sources.
Tests abundance estimation in white-tailed deer with drone and ground-based methods, suggesting improved accuracy.
Methods to estimate abundance from unmarked populations of animals are widely used because of their efficiency and presumed accuracy. These methods account for imperfect detection by comparing observed data against the assumed distribution of the entire population. Traditional tests of whether these methods produce reliable estimates for wild populations fall short because they have not known the true population size or been able to directly observe animal movement during surveys. Drones with high‐resolution imaging may be well poised to observe wildlife surveys as they are conducted to enumerate the number of animals available for sampling and to observe responsive animal movement. We flew a thermal‐imaging drone above ground‐based nocturnal spotlight surveys of white‐tailed deer Odocoileus virginianus to test the following: 1) whether traditional ground‐based distance sampling accurately estimates abundance, 2) whether this system meets the assumption of perfect detection on the transect line, 3) whether deer exhibit responsive movement and whether such behavior leads to biased population estimates, and 4) whether measurement error for distance to detected animals or group size leads to biased abundance estimates. We found that abundance estimates from traditional ground‐based surveys were 10% lower than estimates from drone surveys. This bias largely stemmed from imperfect detection on the transect line, which was caused by responsive movement, distraction by other animals, and brushy landcover. Measurement error existed but was not strongly biased, resulting in negligible effects on abundance estimation. In summary, drone surveys enabled a direct evaluation of foundational distance sampling assumptions and their effects on abundance estimation. Pairing drone flights with traditional wildlife surveys offers a scalable framework for validating unmarked abundance estimation methods and improving the reliability of population estimates.
No takes yet. Share an insight, caveat, or question.
Delaney et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: