during the first part of the incubation period. During the latter 
part of the period, males may desert their mates and associate with 
other males. Stotts and Davis, from seven examples of male attend¬ 
ance after incubation began, found that desertions by the male 
averaged about midway in the incubation period. At this point, 
aerial crews may again see paired or small groups of birds which 
consist entirely of males. Later the postbreeding males may congre¬ 
gate in larger groups in conjunction with nonbreeding males. This 
phase in the nesting season should be marked by flocked birds once 
again, as in the early phase. However, the number of observations 
of paired birds should be lower than in the early phase. 
In addition to the numbers of pairs and flocked birds, the 
number of groups of three may have significance in nesting phenology. 
Groups of three are likely to be males in an early stage of post¬ 
breeding congregation, and would not be expected in the early phase 
when groupings are predominantly pairs and large flocks. 
The relations of group frequencies to the stages of the nesting 
season are summarized below: 
Stage of the 
nesting season Aerial observations 
Early More pairs. 
Fewer singles. 
Fewer groups of 3. 
More Large groups (5 or more). 
Middle Pairs intermediate. 
Singles intermediate. 
Groups of 3 intermediate. 
Fewer large groups. 
Late Fewer pairs. 
More singles. 
More groups of 3. 
More large groups (5 or more). 
Counts of black ducks, owing to their low density over large 
areas, are probably more subject to bias from phenological differences 
than are those of other major species of waterfowl. With our present 
sampling procedure, one or more flocks of birds may increase the count 
inordinately. Removing flocked ducks from the count and including 
only pairs and lone ducks (doubled) is probably not an adequate solu¬ 
tion to the problem. Surveys of black ducks are complicated by the 
6 
