
Because of the heterogeneity encountered on all of the larger breeding 
grounds, contours following lines of gross changes in waterfowl breeding densities 
are being established in re-stratifying the major habitat areas. For instance, the 
25,000 square mile Yukon-Kuskokwim Delta, formerly all included in one stratum, 
was divided into two strata in 1955, and probably will be further modified next year. 
In some instances much smaller units can be practicably divided where definite 
lines of demarcation between varying waterfowl densities are apparent. 
In addition to classifying the various breeding areas more adequately, 
there is a major reconciliation yet to be made between aerial counts and ground 
counts of the same transects. In the Ft. Yukon area, for example, the aerial 
count, made June 7 and 8 averaged 5.9 pair of ducks per square mile for the six 
transects on and near the study area. On June 8 and 9a complete ground census 
of eight square miles averaged 35.5 pair per square mile. Both counts tallied 
the same nine species exclusive of scoters and old squaws. That the number of 
ducks counted from the air was considerably fewer than the total resident population 
was verified by the brood count later in the season. The number of broods actually 
counted averaged 5.8 per square mile, essentially the same as the total pairs of 
ducks counted earlier from the air. These figures leave no margin for unsuccess- 
ful nesters and broods missed, of which there were undoubtedly many. On the 
other hand, the 35.5 pair per square mile counted from the ground may be high due 
to duplication when birds flushed to anotherpond and were re-counted. A 
comparison of the relative abundance of the various species as tabulated in aerial 
and ground counts indicates that about half of the ducks are unidentified in an 
aerial census (Table II.), Further indications are that most of these are puddle 
ducks, and probably many puddle ducks are missed entirely in aerial counts, 
because of their drab plimage and considerable time spent on land where they are 
relatively inconspicuous. These observations in Alaska correlate with those from 
some of the intensive air-ground study areas in the Canadian prairie provinces. 
Aerial brood surveys were attempted in three of the areas where 
ground studies were conducted, but too few broods were observed to be of value 
in assessing production, or in comparing results with the ground surveys. 
Ground 
Production studies indicated relatively good brood sizes on the Ft. 
Yukon Flats compared with the 1954 production. With the exception of pintails, 
brood sizes at Selawik were comparable to the Ft. Yukon data in those species 
tallied frequently enough for the figures to be valid. Table IV. summarizes 
brood data. On the Copper River, where major emphasis was directed toward 
goose banding, only pintail broods were observed frequently enough to compare 
with past years. Because it has been a common observation among some tech- 
nicians on various study areas in Alaska that duck broods of age Class II and 
above tend to intermingle much as geese do, all age classes were averaged 
together in considering average brood size. Particularly with the late nesting 
species, too few Class III broods were observed to be of significance anyway. 
