1 
reliably taken as breeding in the province in which they were banded. 
A. S. Hawkins and L. K. Sowls point out to me that the magnitude of 
mallard flights to molting lakes still requires much study. In this 
species at least, the breeding season cannot reliably be assumed to 
extend through July and August. | 
With these limitations in mind one may, however, conclude 
that on some occasions this technique will enable us to compare the 
relative size of migratory bird populations from mortality data 
alone. Given banding operations of sufficient magnitude, the right 
time, and the right geographic location, this formia should be useful: 
BoCa 
A>-—_——? 
ACh 
where A and B are two regional populations to be compared for rela- 
tive size 
A 
e¢ = the percentage of birds banded in region A and recovered 
in a reference area C 
oy 
Q 
il 
the percentage of birds banded in region B and recovered 
in reference area C 
the number of birds banded in C and recovered in area A 
Q 
i) 
i} 
= the number of birds banded in C and recovered in area B 
(>) 
o 
’ 
ounay 
The percentage of a banded sample shot during the first 
hunting season to which the birds were exposed is termed the first- 
season recovery rate (or the seasonal recovery rate). While this 
rate appears to be a good index of hunting pressure, it does vary 
from one banding station to another, the fluctuations being rather 
‘well synchronized from one year to the next. The juvenile first- 
season recovery rate may in some years be less than the adult one. 
Under such circumstances the difference can be used to furnish a 
minimal estimate of the amount of nonhunting mortality affecting 
the juvenile population before the start of the hunting season. 
Seasonal recovery rates for adult mallards appear to be correlated 
with the overall annual mortality rates. Their use as an easily 
computed mortality index offers considerable promise in wildlife 
management. : 3 
Under special conditions rarely encountered at the 
present time, the mortality data on banded birds can be used to 
calculate the relative abundance of two geographic populations of 
a given species. The conditions necessary for this analysis are 
described and a formila given to obtain the results. 
135 

