Long-term trends of bald eagles in winter on the Skagit River, Washington

摘要

We present a long-term assessment of a bald eagle (Haliaeetus leucocephalus) population from a wintering site in the U.S. Pacific Northwest. Eagles were counted weekly between 1978 and 2000 on a 19-km reach of the Skagit River, and oil a contiguous 24-km upriver reach between 1982 and 2000. Eagles typically occurred in greatest numbers from mid-December until the end of January: with the peak during the first 3 weeks of January. Peak 1-day counts varied from 77 in 1983-84 to 506 in 1991-92. Annual eagle detections increased since the first counts were made in 1978. but this trend has not been linear. Detections increased most rapidly from 1987-92. but have averaged about 36% below the 1991-92 peak during the last 6 years. The increase in eagle detections has been most pronounced on the upriver reach. Increases in chum salmon (Oncorhynchus keta) spawning along this reach may have resulted in proportionately greater eagle use of this area. Before 1989. the size of the Skagit chum salmon run was positively correlated with eagle detections (r = 0.81. P = 0.002). whereas frequency of peak flow events on the Skagit was negatively correlated (r = -0.81, P = 0.003). These relationships have weakened since 1988 (r = 0.58, P = 0.05, and r = 0.40. P> 0.10). Low salmon escapements on several nearby rivers between 1989-93 coincided with a peak in eagle detections on the Skagit. A subsequent rebound in these other salmon runs coincides with a return to lower eagle detections on the Skagit. The proportion of subadult eagles in the Skagit population has declined over the last 2 decades, and may reflect changes in the age structure of the regional population in the Pacific Northwest.