Counts of visible migrants at traditional watchsites throughout North America provide an opportunity to augment population-monitoring efforts for birds of prey. We analyzed hourly counts of migrating raptors at one inland (Hawk Mountain Sanctuary, Pennsylvania) and one coastal (Cape May Point, New Jersey) watchsite in northeastern North America. Hourly counts of migrants have been collected for 38 years at Hawk Mountain Sanctuary and for 28 years at Cape May Point. We compared effort-adjusted, arithmetic-mean passage rates to five geometric-mean indexes for 12 species. We used reparameterized polynomial regression to estimate trends in the indexes and to test the significance of trends from 1976–1978 (average index over three-year period) to 2001–2003. Effort-adjusted, arithmetic-mean indexes corresponded to more sophisticated indexes on the complete data sets but did not perform well on simulated data with missing observation days. We recommend the use of a regression-based, date-adjusted index for the analysis of hawk-count data. This index produced trends similar to other geometric-mean indexes, performed well on data sets simulating reduced sampling frequency, and outperformed other indexes on data sets with large blocks of missing observation days. Correspondence between trends at the watchsites and trends from Breeding Bird Surveys (BBSs) suggests that migration counts provide robust estimates of population trends for raptors. Furthermore, migration counts allow the monitoring of species not detected by BBS and produce trends with greater precision for species sampled by both methods. Analysis of migration counts with appropriate methods holds considerable promise for contributing to the development of integrated strategies to monitor raptor populations.
Detección de Tendencias Poblacionales en Aves de Presa Migratorias