Flyway, but it seems more probable that the difference reflects a 
lesser vulnerability of young ducks which migrate to the Pacific 
Flyway, leaving a larger number available to be taken there as 
adults. This hypothesis corresponds to the vulnerability pattern 
suggested below for mallards banded in Alaska and northern Canada. 
In considering the latitudinal variation between the direct and 
indirect recoveries discussed above, we find that significantly more 
direct than indirect recoveries were reported from north of 50° 
latitude. Relatively fewer direct recoveries were obtained between 
34° and 49°, but the proportion was about equal to or slightly 
higher than that of indirect recoveries south of 34°. If we ignore 
recoveries north of 49° and calculate proportions from the remainder 
(mostly migrating birds), we find that the proportion was slightly 
lower for direct than indirect recoveries between 44° and 49° of 
latitude, significantly lower between 39° and 44°, about the same 
between 34° and 39°, and significantly higher south of 34°. The 
pattern of direct recoveries in migration and wintering areas thus 
indicated a more southerly distribution than that of indirect 
recoveries. 
In general, the distribution of direct recoveries of immature 
mallards was more like that of adult females than of adult males. 
However, immature birds tended to be shot further south than either 
adult males or females, and more often in the Mississippi than in 
the Central Flyway. Immature® and adults of both sexes followed 
similar annual fluctuations in distribution of recoveries (tables 41 
through 45). 
Although the dissimilarities between sex and age groups have 
been stressed, it should be noted that in most cases the differences, 
although they may be significant, were relatively small (less than 3 
to 5 percent), and that sex and age groups from the same breeding 
area have similar distributions of recoveries. By far the most pro¬ 
nounced difference in kill distribution of sex or age groups that 
bandings of locals have demonstrated is the high proportion of 
immature mallards, relative to adults, killed on the breeding grounds. 
Recovery Rates 
Direct recovery rates have varied among reference areas and 
years. Because of the variation, which is related to hunting pressure 
the recoveries are more precise in measuring the distribution of the 
kill than the proportions of populations that migrate to various areas 
21 
