them to early death or was the result of a circumst 
ance that had al 
determined the fate of these individuals, Thus, injuries sek es 
from pole traps or. gun shot would fulfill these conditions and bias 
the sample for a longevity study. Excludi 
were recovered as follows; a ng these 19, other adults 
When Recovered Number Recovered 
a re lies 2 ee 
Same year ll 
l year later 
2 years later 
3 years later 
l years later 
5 years later 
6 years later 
HKOrHNHWw 
I interpret this mortality series as probably baised for 
birds recovered within one year of capture. The fact that 30 out of 
38 died within one year after banding should be noticed in connection 
with the assumption stated earlier in this chapter that captured 
horned owls can usually be regarded as dead ones. The 8 birds shot 
one or more years later appear to represent a reasonable confirma- 
tion of the low adult mortality rate for adults obtained in the life 
table above. This seems to be as far as we can go at the present time. 
| Age Ratios and Productivity 
If this species breeds at the age of 2 years, a hypothetical 
population such as that given in colum lt, of table 51 could remain 
stable if each pair raised 1.2 young successfully to September 1, 
This hypotnesis is hard to verify at the present time. 
If all horned owls do breed at the end of one year, popula- 
tion balance in this species would be achieved either by a fairly 
high percentage of nesting failures or very low early survival of the 
young. The first of these alternatives seem possible. Fitch (190) 
has shown that 3 nests out of 5 on a California study area failed 
because of weather, predation, and unsanitary nest conditions; a much 
more intensive nesting study or an extensive inquiry is still needed 
to settle this interesting question. 
Population Dynamics 
The possible existence of a continental gradient in the mean 
size of cluthces of great horned owls seems to be suggested in the 
ornithological literature. When sufficiently large samples of banding 
recoveries are available, regional subadult and adult survival rates 
may have to be correlated with regional variations in the mean number 
of eggs and young produced. The small samples assembled in the present 
study originated mostly from banding work in the northern part of the 
United “tates and the southern part of Canada. 
117 
