Evaluation of Wintering Waterfowl Population Aerial Survey Methodology

Abstract

Suitably designed wildlife population surveys account for methodological biases to produce defensible estimates. Although North American waterfowl populations have been aerially surveyed for nearly a century, accounting for biases remains challenging. We evaluated wintering waterfowl aerial population surveys conducted by the Alabama Wildlife and Freshwater Fisheries Division in the Tenneesee River Valley and Mobile coast, bay, and delta during November-February 2021–2024. We quantified detection probability and count variability within aerial surveys and used those estimates with an availability index to assess the relative influence of sources of error and survey design in aerial surveys. Detection probability was highest (0.71-0.95) for geese (Branta spp.) and lowest (0.24-0.70) for dabbling ducks (tribe Anatini) but was also influenced by seat position, observer experience, and group size. Detectability had lower influence on simulated estimates than availability or count variability. While assessing availability is challenging, managers can address counting variation by conducting replicate surveys.