the United States that are adjacent to Canada should be completed 
before the opening of the Canadian hunting season. However, in 
order to have larger samples available for analysis, pre-season 
bandings in the United States were included even though hunting 
had begun in adjacent areas of Canada. Relatively few Canadian 
recoveries were obtained from pre-season bandings in the United 
States. 
Only direct recoveries (those occurring during the first hunting 
season following banding) are listed in tables 1 and 2. Since the 
objectives of pre-season banding relate to the hunting kill, only 
ducks believed to be shot were included. In addition to recoveries 
reported as shot, those reported as found dead during the hunting 
season were used, because these also usually reflect shooting mor- 
tality. Banded Samples that yielded fewer than five recoveries in 
each age group each year were grouped by State or Province. Data 
from States or Provinces having no major banding location and a 
small combined total from miscellaneous locations were omitted. A 
single location includes all bandings in a 10' block of latitude 
and longitude which is an area of approximately 8 x 12 miles in the 
northern United States and southern Canada. 
Adequacy of the Sample 
The size of banded samples to be attained in the pre-season 
program is defined in the Continental Banding Program (U. S. Fish 
and Wildlife Service, 1959) as follows: "Sufficient birds should 
be banded at each station to annually yield a minimum of 50 direct 
recoveries from birds of the year and 50 from adults for each of 
the important species. Since 100 direct recoveries from each of the 
two age groups would allow for more accurate determination of differ- 
ences in mortality from one station to the next, this figure will be 
held as an objective." Only a few stations attained the minimum 
recovery quotas. In 1959, Tule Lake National Wildlife Refuge in 
California and Rice Lake National Wildlife Refuge in Minnesota were 
the only stations to obtain 50 recoveries for each age group of 
mallards. In 1960, no banding station attained these quotas for 
mallards. No station or combination of adjacent stations reached 
black duck quotas in either year. Station to station variability 
in recovery rates and relative recovery rates cannot be accurately 
evaluated because sample sizes are too small. Since a measure of 
this variability is necessary in order to determine how well the 
data reflect the characteristics of the entire pre-season population, 
the objectives of the program were not fully attained. Much of the 
apparent variability in tables 1 and 2 may be due to inadequate 
samples rather than to true differences in the characteristics of 
the populations being sampled. 
