In the present report we address questions about winter distribution patterns and survival rates of North American mallards Anas platyrhynchos. Inferences are based on analyses of banding and recovery data from both winter and preseason banding period. The primary wintering range of the mallard was dividded into 45 minor reference areas and 15 major reference areas which were used to summarize winter banding data. Descriptive tables and figures on the recovery distributions of winter-banded mallards are presented.

Using winter recoveries of preseason-banded mallards, we found apparent differences between recovery distribution of young versus adult birds from the same breeding ground reference areas. However, we found no sex-specific differences in winter recovery distribution patterns. Winter recovery distributions of preseason-banded birds also provided evidence that mallards exhibited some degree of year-to-year variation in wintering ground location. The age- and sex-specificity of such variation was tested using winter recoveries of winter-banded birds, and results indicated that subadult (first year) birds were less likely to return to the same wintering grounds the following year than adults. Winter recovery distributions of preseason-banded mallards during 1950-58 differed from distributions in 1966-76. These differences could have resulted from either true distributional shifts or geographic changes in hunting pressure.

Survival and recovery rates were estimated from winter banding data. We found no evidence of differences in survival or recovery rates between subadult and adult mallards. Thus, the substantial difference between survival rates of preseason-banded young and adult mallards must result almost entirely from higher mortality of young birds during the approximate period, August-January. Male mallards showed higher survival than females, corroborating inferences based on preseason data. Tests with winter banding and band recovery data indicated some degree of year-to-year variation in both survival and recovery rates, a result again consistent with inference from preseason data. Some evidence indication geographic variation in survival rates; however, there were no consistent directional differences between survival rates of mallards from adjacent northern versus southern areas, or eastern versus western areas. In some comparisons, Central Flyway mallards exhibited slightly higher survival rates than mallards from other flyways.

Weighted mean estimates of continental survival rates were computed for the period 1960-77 from both winter banding data and preseason banding of adults. Resulting estimates differed significantly for males, but not for females, and the magnitude of the difference between point estimates was relatively small, even for males. The direction of the difference between these estimates was predicted correctly from previous work on the effects of heterogeneous survival an d recovery rates on band recovery model estimates. The similarity of survival estimates from these two independent data sets supports the believe that biases in these estimates are relatively small.