Integrating citizen-science data with movement models to estimate the size of a migratory golden eagle population

Estimating population size is fundamental to conservation and management. Population size is typically estimated using survey data, computer models, or both. Some of the most extensive and often least expensive survey data are those collected by citizen-scientists. A challenge to citizen-scientists is that the vagility of many organisms can complicate data collection. As a result, animal-movement effects on data collection can adversely affect modeling of those data. Thus, it would be helpful to develop methods that integrate citizen-science datasets with models that account for animal movement. We used hawk-count data collected by citizen-scientists to estimate the number of golden eagles (Aquila chrysaetos canadensis) migrating through Pennsylvania, USA. To do this, we designed a computer model to simulate migratory flights of eagles to estimate what proportion of the population is available (i.e., within visible range or close enough) to be counted at hawk-count sites in Pennsylvania. We then conducted a multi-state mark-recapture analysis to estimate detection probability (i.e., the rate at which birds within visible range are observed) of migrating eagles. Finally, we used availability rates and detection probabilities to adjust raw hawk-count data to produce estimates of population size. Our models suggest that 24% (±14; mean ± SE) of migrating golden eagles are available to be counted at hawk-count sites, and that 55% (±1.6) of the available eagles are detected by hawk-count observers. We estimate that 5122 (±1338) golden eagles migrate annually through Pennsylvania. Our analysis provides the first quantitative estimate of the size of the eastern golden eagle population, and we demonstrate the utility of one approach to use citizen-science data to address a pressing conservation goal-population size estimation.