Incorporating Aerial Imagery into Waterfowl Population Estimates

With advances in digital photography, aerial imagery has become a common method of estimating wildlife population size. This method has typically been used when wildlife congregate in large groups and can be photographed efficiently. When wildlife are distributed widely across the landscape, a series of aerial images have been taken along transects to estimate population size. From September through January of 2014–2016, we conducted aerial waterfowl surveys using a quadrat method to determine population size of waterfowl using floodplain lakes, wetlands, and other areas of the Illinois River valley. We established a grid of 359 quadrats (260 ha) in the 100 year floodplain and randomly selected 50 quadrats each week from a high and low density strata. We placed a single transect in each quadrat running diagonally from the NE to SW corner and obtained aerial images using an automated camera affixed to the fuselage. We manually searched each aerial image and enumerated individuals by species or lowest taxonomic unit. We compared the corresponding aerial estimate with the extrapolated estimate from aerial images and estimated an error rate (difference between aerial image count and aerial observer estimate) for all major guilds. Error rates were high (waterfowl, 106% ± 40%; ducks 111% ± 43%; geese, 224% ± 72%, swans, 80% ± 72%) and estimates from aerial images averaged 206% higher than aerial estimates. A power analysis will be conducted to determine the minimum number of aerial images needed to produce a coefficient of variation of less than 15%. Once the minimum number of images is determined, the feasibility of the method can be analyzed by taking into account the additional time and cost required to take and analyze that amount of aerial images.