We monitored distribution and abundance of food resources and determined food habits of nonbreeding bald eagles (Haliaeetus leucocephalus) on the northern Chesapeake Bay, as a preliminary step toward examining food-base effects on bald eagle distribution and abundance. To correctly interpret our food habits results, we first examined biases of 2 commonly-used food habits techniques, pellet analysis and food remains collection, through feeding trials with 2 captive bald eagles. Eagles were fed a variety of food items found on the northern Bay. Egested pellet contents and frequency of remains were compared with actual diet. We also examined efficacy of direct observation by observing eagles in high-use foraging areas. We found pellet analysis accurately indicated the species of birds and mammals eaten, but overrepresented medium-sized mammals and underrepresented large carrion in percent occurrence results. Fish were poorly represented in pellets. Eagles rarely produced pellets after eating fish, suggesting that pellet egestion rate, defined as the number of pellets produced per eagle per night, can serve as an index to relative use of birds and mammals. Food remains collection was highly biased toward birds, medium~sized mammals, and large, bony fish. Direct observation was labor intensive and required close proximity of the observer for unbiased identification of food items. Observation may be the only means of documenting eagles' use of small, soft-bodied fish.

We used direct observation, pellet analysis, and pellet formation rates to determine bald eagle food habits from December 1986 through April 1988. We monitored fish abundance by gill netting and waterfowl abundance by aerial surveys over this same period. Fish and waterfowl abundance varied reciprocally; waterfowl numbers peaked in winter and fish numbers peaked in spring and late summer. Bald eagles responded to differences in food abundance with diet shifts. Canada geese (Branta canadensis), mallard (Anas platyrhynchos), and white-tailed deer (Odocoileus virginianus) carrion were primary foods from November through February. Cold-stressed gizzard shad (Dorosoma cepedianum) were captured frequently by eagles below a hydroelectric dam on the Susquehanna River in November and December, and also were taken frequently throughout the study area during a winter when ice cover was extensive. Shad were not commonly available during a milder winter. From April through September, bald eagles fed on a variety of fish species, primarily gizzard shad, channel catfish (Ictalurus punctatus), Atlantic menhaden (Brevoortia tyrannus), white perch (Morone americana), American eel (Anguilla rosfrata), and yellow perch (Perca flavescens). The 4 most commonly consumed fish species also were the most commonly gill netted species. At least 25% of all fish taken were scavenged. Live fish were most abundant at the water's surface in shallow water. Bald eagles' use of live fish reflected this availability; water depth at live fish capture sites was less than at sites where fish of dead or unknown status were taken. Eagles foraged most intensively within 1 hour of sunrise. A second smaller peak in foraging activity was observed in early afternoon.