Flock sizes and sex ratios of canvasbacks in Chesapeake Bay and North Carolina

Knowledge of the distribution, size, and sex ratios of flocks of wintering canvasbacks (Aythya valisineria) is fundamental to understanding the species' winter ecology and providing guidelines for management. Consequently, in winter 1986-87, we conducted 4 monthly aerial photographic surveys to investigate temporal changes in distribution, size, and sex ratios of canvasback flocks in traditional wintering areas of Chesapeake Bay and coastal North Carolina. Surveys yielded 35mm imagery of 194,664 canvasbacks in 842 flocks. Models revealed monthly patterns of flock size in North Carolina and Virginia, but no pattern of change in Maryland. A stepwise analysis of flock size and sex ratio fit a common positive slope (increasing proportion male) for all state-month datasets, except for North Carolina in February where the slope was larger (P lt 0.001). State and month effects on intercepts were significant (P lt 0.001) and confirmed a previously identified latitudinal gradient in sex ratio in the survey region. There was no relationship between flock purity (% canvasbacks vs. other species) and flock size except in North Carolina in January, February, and March when flock purity was related to flock size. Contrasting characteristics in North Carolina with regard to flock size (larger flocks) and flock purity suggested that proximate factors were reinforcing flocking behavior and possibly species fidelity there. Of possible factors, the need to locate foraging sites within this large, open-water environment was hypothesized to be of primary importance. Comparison of January 1981 and 1987 sex ratios indicated no change in Maryland, but lower (P lt 0.05) canvasback sex ratios (proportion male) in Virginia and North Carolina.