The refinement of this analysis represents a study which 
could not “be incorporated into my own project. It should be obvious 
that so-called sampling limits or confidence limits can eventually - 
be computed for..line AB. Once these are established,“a given first— 
season recovery rate could be determined: each.spring and a. reasonable 
set of statistics: derived for the past hunting season. Thus if. 
figure 20 were based on precise data and not crude ones, a recovery 
rate of 12 per cent would indicate an annual mortality rate of about 
50 per cent and a loss of about 30 per cent of the population due 
to hunting. Under such conditions and with precise data, the rela- 
tionship of reported to unreported-kill should be readily computable. 
It is my feeling that the dynamics of adult mallard popula- 
tions: do, in fact, involve statistics with these relationships and 
that the ultimate determination of point A on the ordinates of 
Bremne such as figure 20 can be readily determined. 
‘Population Dynamics in the Mallard 
: Contemporary neating statistics.--From Je B. isa (asks) 
Bice we have seen that in the redhead, clutches averaged 9.8 eggs, 
that "YS. ‘per cent of these produced young, and that juvenile losses .- 
to August 20 or September 1] averaged about 30 per cent. In the 
mallard, recorded mean clutch size is lower-~about 7.8 (table 2)) 
with 55-70 per cent of the nests hatching. The observed brood size 
for downy young is about 7.6 (table 26), and 20 per. cent is the 
maximum mortality thus far reported from the time the young are 
downy to the point when they are able to fly (Stoudt 1946). A 
separate estimate, calculated in Chapter XII from recovery-rates 
in-table’55, suggests a juvenile mortality rate of about 27 per 
cent from the time of banding to the start of hunting. Thus for 
these two species of waterfowl, the summer mortality of young 
birds Bppaers to be roughly similar--about 25-35 per cent. 
The dneaiats implication of this comparison is that 
redheads have larger clutches and lower nesting success in contrast 
to smaller clutches and higher nesting success in the mallard. 
The contrast is hardly a fair one. The redheadidata come from an 
intensively studied area in Iowa, the mallard data from-extensive 
studies scattered over a continent. Lyle K. Sowls (in litt.) sends 
me the following unpublished facts about mallard egg “gets intensively 
studied at Delta, Manitoba: "Of 51 mallard nests known to have com- 
plete sets, the average was 9. Of 23 complete sets prior to June 15, 
the average was°10.0."'. Whitherby, Jourdain, Ticehurst and Tucker 
(1939) report that clutch size is "usually about 10 or 12, but may 
be any number from about 7 or 8 to 16." Bent's (1923) figures are 
substantially the same. There are thus good reasons to believe 
that a mean clutch size of 7.8 eggs.for this species must refer 
importantly to renests- tapetch clutch size will average lower 
than first sneer | 
157 
