Estimating Sightability for Helicopter Surveys Using Surrogates of White‐Tailed Deer

Abstract

ABSTRACTOn large management units where terrain allows observation of white‐tailed deer (Odocoileus virginianus) from the air, helicopter surveys may provide managers with cost‐effective and accurate estimates of population abundance; however, imperfect detection of deer introduces negative bias. This can result in potentially inappropriate management recommendations (e.g., harvest prescriptions). Sightability estimators are designed to model detection heterogeneity based on factors that affect observer detection of target animals. Sightability models are widely used to estimate ungulate abundance; however, previous researchers have developed sightability models using radio‐collared animals, potentially introducing bias during model development attributed to unaccounted for movement between initial and follow‐up flights. Therefore, our objectives were to model detection probability using surrogates of white‐tailed deer as a function of covariates including distance from transect, vegetative obstruction, and light conditions to improve population estimates derived from helicopter surveys and apply our final sightability model in operational helicopter surveys. We conducted our study in 2019 on a 399.4‐km2 study area within a 1,210‐km2 cattle ranch in central Florida, USA. We placed 3‐dimensional archery targets as surrogates for white‐tailed deer at assigned locations unknown to observers across a combination of stratified bins of vegetative obstruction and distances from the transect. Our top model indicated that distance from the transect and vegetative obstruction negatively affected detection of deer. Detection probability on the flight transect ranged from 0.95 (95% CI = 0.89–0.98) for vegetative obstruction 0–25% and distance from transect 0–25 m to 0.05 (95% CI = 0.01–0.18) for vegetative obstruction 76–100% and distance from transect 100–125 m. We applied our sightability model to operational surveys on 3 units, which produced population estimates averaging 26% higher than those derived from uncorrected counts. Observers simultaneously recorded live deer during operational flights, of which 60% of deer groups were observed while stationary. Our methods using stationary surrogates to develop region‐specific sightability models could be used in other study areas to account for imperfect detection of white‐tailed deer. © 2021 The Wildlife Society.