Wildlife population survey methods: challenges and opportunities in the 21st century

We consider how advances in technology are affecting the methods used to assess wildlife abundance, and speculate on future methodological needs. We anticipate that aerial transect surveys will become more widespread, with the use of drones, high-resolution images, and automated software to identify animals. Analysis methods will need to address mis-identification of detections. We also expect acoustic surveys to become widespread. These might involve terrestrial or marine acoustic arrays that detect and localise calling animals. We expect increased use of tagging devices to generate availability time series data for more species and more individuals, together with concurrent development of availability process models and abundance estimation methods that incorporate them. Increasingly, we expect model-based methods to replace design-based methods for estimating abundance, so that animal density can be related to covariates, and so that changes in density can be modelled and interpreted. We also expect to see development of spatial models for analysing capture-recapture data (spatially-explicit capture-recapture). Combining data from multiple autonomous sources (e.g. sonar combined with digital aerial) to produce abundance estimates will become more common. We discuss analytical developments we believe will be needed to keep pace with advances in data acquisition.