Abstract
Wind energy development is increasing in the United States and Canada and may affect bald (Haliaeetus leucocephalus) and golden eagle (Aquila chrysaetos) populations through direct mortality. Wind farms on ridge‐tops may present a greater mortality threat because of the importance of these features for migrating raptors. Regulators in both countries recommend methods for preconstruction surveys, so that eagle use and collision risk can be assessed at potential construction sites. We obtained hourly count data collected during the autumn seasons of 1990–2014 from 22 ridge‐top raptor‐migration monitoring sites in the Pacific, Eastern, and Central flyways of North America. We simulated 18 different survey protocols and effort levels based on survey guidance by repeatedly drawing quasi‐random subsamples from continuously collected data in a manner that imitated the recommended protocols, comparing the number of eagles seen in simulated counts to the eagles seen in all data collected to assess the effectiveness of those protocols in correctly estimating the eagle passage rate (eagles detected per hour of counts) at each site. Multihour point counts conducted on a weekly basis were found to be ineffective at correctly estimating eagle passage both within and across years. Counts conducted in the afternoon were found to be more effective than those conducted in the morning. We demonstrate that full‐day counts, conducted weekly during the peak period of eagle passage, were much more effective at estimating eagle passage rate. Although current guidance recommends the performance of such full day counts, this result underscores their importance in preconstruction surveys at potential wind sites as an essential tactic to estimate eagle passage and risk of collision with energy infrastructure.