A breakdown of the recoveries using bandings 
disclosed no difference in survival between T2 Gavortiag Gece aa 
Alaska and Cariada and 128 banded in the northern part of the Unit ed 
States. A group of 86 males appeared to have a higher mean mortality 
cg than 75 females but the difference was not statistically signifi- 
cant. 
Hand-reared birds.—-Samples of hand-reared birds were examined 
in passing. Among these, the percentage shot during the first year of 
life did not vary with the month of banding (June, July, or August). 
It did, however, markedly differ from that for wild-reared birds, 
Preliminary data also indicate that the adult survival rate of wild 
birds may not be attained by these hand-reared individuals until the 
third year of life (table 28). By this time, however, only 5 per cent 
of those reported are still alive. It is difficult to appraise the 
bias in the mortality curves thus constructed for game-farm birds. 
The birds that I examined were released from California to Delaware 
and as far north as Alberta and Manitoba. Errington and Albert (1936) 
have analyzed in detail one release in Iowa; similar analyses for other 
releases would undoubtedly be instructive. The sample studied in 
table 28 included not only birds raised on game farms but a small 
number reared from wild eggs by the Delta Waterfowl Research Station. 
The latter appeared to possess the migratory behavior of wild-reared 
individuals; they also represented nearly half of the 2) hand-reared 
individuals known to be alive at the start of the third year of life, 
Survival of adult birds 
Hunters! reports.--A reasonable approximation of an adult 
mortality rate can be obtained by analyses of reports for birds that 
were unaged at the time of banding. If the mortality reports for the 
first year are discarded in such a series, the remaining reports refer 
to adults, the ages being x + 1, x + 2, and so on. A sample of this 
type is illustrated in table 29. To form it, I used 181 recoveries 
accumlated by F. C. Lincoln for birds he banded in Illinois in 1922, 
228 more of Lincoln's for Illinois birds banded in 1926, and 821 banded 
in Missouri by L. V. Walton and John Brocker from 1922 to 1928. This— 
Mississippi Valley subsample, which included. some nonhunting reports, 
had a mean mortality rate of 6.5 per cent per annum. Although many 
of these birds were banded in years when bands may not have been very 
gurable, I failed to detect any important evidence of band loss. 
(This would show up as abnormally high mortality rates during the 
early ages in the life table.) A second subsample ecnsisted of 1762 
males banded mostly by A. J. Butler in British Columbia and by G. M. 
Benson and his associates at Malheur National Wildlife Refuge in 
Oregon; these birds were found to have a mean adult mortality rate 
of 48.5 per cent. A few more mortality reports in the Pacific coast 
sample can probably be expected. These should total less than 0.2 
per cent of the total and should ‘not appreciably affect the mortality 
rates calculated. Evidence against the existence of differential 
mortality rates for male and female mallards is given in Chapter XIV. 
69 
