17 

Fig. 10. A small wetland highly preferred by nesting female canvasbacks in the Minnedosa study area. 
Mixed Broods 
It is not always possible to identify redhead ducklings 
in canvasback broods, especially when the birds are young 
or the light is poor. Broods in which the number of red- 
head ducklings was not accurately determined were omitted 
from final tabulations. However, numbers of redhead duck- 
lings per canvasback brood (Table 26) are probably con- 
servative. Variation between numbers of redheads in Class 
I and II broods is probably a result of some redheads in 
the younger age classes being overlooked. The most signifi- 
cant feature of the data for most years is the close relation 
between the number of redhead ducklings per canvasback 
brood and the density of redhead breeding pairs for cor- 
responding years. 
Brood counts indicated that more canvasback nests were 
parasitized by redheads during the middle and late portions 
of the nesting season than during the early portion (Table 
27). Wide variation in numbers of redhead ducklings from 
nests hatched after 16 July might have been due to the small 
sample size for that period. 
Juvenile Mortality 
Brood sizes of Class II ducklings were used to determine 
juvenile mortality (Table 28) because canvasback broods 
of Class III usually separated into two or more groups and 
were difficult to recognize as a unit. Inasmuch as the dif- 
ference in mean size between Class II and Class III broods 
(minus redheads) was less than 0.5 duckling (Table 25), 
mortality appeared to be low. Actual juvenile mortality 
from nestlings to fledglings is probably depicted accurately 
by the data of Table 28. Juvenile mortality averaged 21 % 
during the 12-year study, although there were marked year- 
to-year variations. These data do not include losses of en- 
tire broods. 
