were 61 and 15 per cent respectively. Since nonhunting mortality 
does in nature coincide at least in part with hunting mortality, 
the per cent dying from nonhunting in the above table will tend to 
be too high. 
It seems to me remarkable that a 20 per cent adult hunting 
mortality should result from these calculations. While I suspect 
that this result is too low, there seems to be no way of verifying 
this at the present time. Higher hunting rates would make for declin- 
ing populations--a possibility that seems quite real but ia likewise 
unprovable at this writing. | 
Redhead populations other than Utah's.-—The weak point in 
these analyses, as 1 see it, lies In the local origin of the data; 
a productivity estimate from Iowa, banding statistics frcm Fish and 
Wildlife Service operations in Utah, bag information from Hochbaum's 
work at Delta, Manitoba. Perhaps certain segments of the redhead 
population have been subjected to mortality rates of the order 
postulated here. It surely does not follow at this time that the 
data hold for all segments of this species' population. As a 
critical test for this, I assembled 223 mortality reports of red- 
heads banded outside of Utah up through the year 1941 (table 3.). 
Even when birds banded in the hunting season are included in the 
sample, a very high proportion of: the reports cover birds shot in 
their first year of life. Although this proportion may be lowered 
slightly by subsequent reports of birds shot as adults, we mst 
ecnclude that the extraordinary population turnover of redheads 
pictured for Utah birds is not a local phenomenon. 
The importance of a correct adult mortality rate in 
these calculations should not be overlooked. All samples cited 
here in this connection were small. The mean mortality rate of 
Utah-banded redheads (5), per cent, table 33, part A) was based 
on 61 birds; that for the non-Utah birds (7 per cent, table 3h) 
was based on 7. Seventy-three others banded as adults gave a 
mean annual mortality rate of 62 per cent. It is entirely possi- 
ble that these rates reflect not only sampling errors but also 
changes in hunting regulations. A 55 per cent overall adult mor- 
tality rate seems like a reasonable yardstick under the circum 
stances. 
All this evidence leads me to conclude that redheads in 
North America have in the past sustained a juvenile hunting kill 
of the order of 50 per cent, that the adult kill has been 20 or 
30 per cent and that annual mortality rates have been about 70 
per cent for juveniles and about 55 per cent per year for adults. 
Under these conditions, the population would remain stable only if 
each pair produced about 3.6 young as of September 1. 
80 
