Chapter XIV.--Dynamics of Mallard Populations 
Effect of Hunting on Adult Mallard Mortality Rates 
What percentage of the North American waterfowl population 
is annually harvested by hunting? Attempts to get this percentage 
directly have not yet been satisfactory. The total kill is also dif-~ 
ficult to sample, and the amount of crippling losses is still wknown. 
Bellrose (1945) has published an interesting estimate that the total 
annual kill of the continental duck population is about 11 per cent, 
This statistic was based on (1) first-season recovery rates of water- 
fowl banded in Canada by Ducks Unlimited; (2) a correction factor 
allowing for 75 per cent of recovered bands being reported; and (3) 
an estimate that hunters retrieve only 75 per cent of the ducks they 
kill, 25 per cent being lost by "crippling." It must be obvious 
from the very considerable fluctuations in adult mallard mortality 
rates studied in relation to hunting in this chapter that Bellrose's 
estimate is probably too low. It is dangerous, of course, to generalize 
from the specific (in this case, adult mallards) to a continental 
population embracing all the waterfowl in North America. 
Bellrose's commendable attempt can, however, be shown to 
have a real or potential weakness at each step of his calculations: 
(1) the recovery rates used seem to include birds banded during the 
hunting season; notice in table 55 (Chapter XII) the difference in 
mallard rates I use for Ducks Unlimited birds and those quoted there 
from Bellrose's (19h) paper; (2) the ¢orrection for unreported bands 
involves an unwarranted assumption pointed out in Chapter XII; (3) 
the 33 per cent correction factor for crippling loss (equivalent to 
25 per cent of all the birds dying from hunting, although frequently 
mentioned by waterfowl biologists, nevertheless remains a guess, 
More recently Bellrose and Chase (1950) have presented some 
interesting manipulations of band-recovery data that "indicate" hunt- 
ing mortality annually affected 1 per cent of the 30,912 mallards 
banded at Lake Chautauqua, Illinois, from 1939 to 194. This con- 
clusion importantly rests upon an experiment in which a control 
series of 2);2 standard bands yielded a recovery rate of 8.7 per 
cent in contrast to 200 special reward bands that yielded a recovery 
rate of 25 per cent. The implication, that only 35 per cent of 
Canadian and U. S. hunters report the bands they obtain, would ap- 
pear to be potentially affected by fatrly large sampling errors, 
and I regard this Illinois result as a wholly preliminary one. When 
the final results of this reward-band experiment by the Fish and 
Wildlife Service and its cooperators are reported from other parts 
of the continent, the samples should be large enough to permit more 
confident estimates of the fraction of mallard populations bagged by 
hunters. Eleven per cent is clearly too low, and 1 per cent is 
quite possibly too high. | 
1h 
