The striking difference in recovery rates between the coastal 
(saltwater) birds and the inland birds at once raises the question of 
band loss, which will be taken up in a later section. 
Birds banded after 1930 had not yet lived to ave 1-15 at the 
time this study was undertaken. The small amount of banding work in the 
1920's has created little opportunity for double~crested cormorants to be 
recorded at age intervals past 14-15. This in itself may create a small 
bias in life table analyses that follow. 
Characteristics of the sample studied 
The usable recovery reports of cormorants banded inland 
and on both coasts included 45 found sick or injured, 65 caught or 
captured, 128 known to have been caught by fisherman, 250 found dead, 
and 617 reported shot. These birds had been banded as young up 
through the nesting season of 191. The monthly distribution of these 
shows frequency peaks for both October and April-May in the first 
year of life as wellas in later years in the aggregate (figure 6). 
The same bimodal distribution holds for birds captured in various ways 
(figure 7), but the spring peak is not evident in birds reported shot. 
A rough comparison of the various types of recovery, given 
in table 17, discloses that a significantly higher proportion of 
juvenile birds appears in the shot samples as compared to birds found 
dead, and that a parallel excess of juveniles occurs in coastal-banded 
birds found dead in contrast to inland juveniles found dead. The per- 
centage of juvenile-mortality reports occurring from August 1 to 
November 30 is also significantly higher in the shot sample than in 
those reported found dead. There is thus a clear inference that shot 
samples are biased for age. The hypothesis that fresh-water birds retain 
bands longer than marine birds is not supported by the data on birds 
shot, but it does find confirmation in the data for birds found dead. 
Because 89 per cent of all the inland shot juveniles were reported from 
August 1 to November 30, it is possible that inland hunting pressure on 
young birds serves to mask the band-loss effect we are seeking in age 
ratios of this character. Annual variations in dates banded (inland 
versus coastal) further complicate this picture. 
Composite life tables 
Adult mortality rates.--If we may momentarily disregard the 
possible bias In reports of juvenile birds, we can examine adult 
mortality rates as calculated from reports on birds shot and found 
dead. Table 18 indicates that some inland-banded cormorants are now 
known to survive to at least their fifteenth year of life and that 
their adult mortality rate in this sample averaged about 22-26 per 
cent per annum. Similar data for marine-banded birds (table 19) 
give a maximum life span of only 7 years and a post-juvenile mortality 
rate averaging 39 per cent per annum. The discrepancy between the 
two survival curves is clearly seen in figure 8. 
48 
