Northern Harriers Have a Geographically Broad Four-Year Migration Cycle

Abstract

We used observation data from major hawk-watch sites in northeastern North America to assess multiyear cycles in breeding productivity of the Northern Harrier (Circus hudsonius). We found that autumn totals of migrating harriers for 1983–2017 autocorrelated positively at the 4-yr lag for interior, but not Atlantic coastal, counts. The autocorrelation was especially high for harriers counted in October. The autumn counts of south-bound harriers were not correlated with counts of north-bound harriers in the previous spring. Interannual differences in atmospheric circulation during migration in 1983–2017, as expressed by the autumn North Atlantic Oscillation Index, correlated with the autumn harrier counts but did not have a 4-yr cycle. The temporal pattern revealed in this study should be considered when using counts of autumn migrants to infer regional population trends. Future research could focus on the cause of the cyclic pattern in harrier numbers. One obvious question concerns the relationship between spring populations of meadow voles (Microtus pennsylvanicus), which are important prey for harriers during breeding, and harrier numbers reported during the subsequent autumn migration.