Telemetry‐based aerial surveys for estimating abundance of sparse populations

Abstract

Aerial surveys are commonly used to collect data for estimating abundance of a variety of large animals. In the context of aerial surveys, common abundance estimators, including distance sampling, closed population mark‐resight models, and sightability models, all require visually scanning a large area, usually along transects. For sparse populations, such extensive surveys may be inefficient and costly. We evaluated abundance estimators proposed by Rivest et al. (1998) and Clement et al. (2015) in which animals are located using telemetry instead of extensive visual searches, thereby reducing flight time. We conducted aerial surveys of Sonoran pronghorn in southwest Arizona in 2017 and evaluated the bias and precision of abundance estimates returned by both estimators. We also compared the costs of telemetry surveys to sightability surveys, a more common method. Both telemetry‐based methods generated unbiased abundance estimates for a population of known size. However, in a population of unknown size, there was evidence of a non‐random distribution of telemetry collars and downward bias in abundance estimates. The telemetry‐based methods dramatically increased the sighting rate of Sonoran pronghorn during flights and reduced total flight time by >90%, a substantial time and cost savings. However, telemetry‐based approaches incur an additional cost of maintaining telemetry collars on animals in the population. For Sonoran pronghorn and other species, telemetry‐based abundance estimators may offer significant cost savings over estimators requiring extensive visual searches, if there are not meaningful deviations from model assumptions. Telemetry‐based methods will be most useful if the cost of collaring animals is low, if animals are sparse or hard to detect, if the study has a short duration, and if the study species has a loose social structure and tends to form large groups.