Comparing infrared imagery to traditional methods for estimating ringed seal density

When conducting aerial surveys of wildlife populations, a number of animals go undetected by observers. The use of infrared imagery may offer a solution to improve detection rates and reduce visibility bias. Our objective was to compare the use of an infrared camera system with traditional methods using direct visual observations for estimating density of ringed seals (Pusa hispida) on ice. We conducted aerial surveys of ringed seals in the areas of Pond Inlet and western Hudson Bay in the Canadian Arctic during spring 2016 and 2017. Infrared videos were recorded within a 250-m-wide strip beneath the aircraft and potential seals were verified using corresponding visual images from a digital single-lens reflex camera. Results from strip-transect analysis of infrared observations were compared with results from strip-transect analysis of direct visual observations and results from line-transect analysis of a combined data set of infrared and direct visual observations. The different methods produced different density estimates with the infrared strip-transect and line-transect methods producing similar results that were, on average, approximately 2-3 times greater than strip-transect analyses of observer data. Use of infrared imagery provides several advantages over traditional methods because it allows for more efficient and reliable detection of seals, eliminates the need for large teams of trained observers, and simplifies the process of data analysis and density estimation. (c) 2019 The Wildlife Society.