14 
Groupings of broods have been recorded in other Canada 
goose populations (Williams and Marshall 1937; Miller and 
Collins 1953; Naylor 1953; Geis 1956; Sherwood 1967; and 
Hanson and Eberhardt 1971); yet MacInnes (1962) and 
Mickelson (1975) did not consider brood flocks common in 
the populations of small northern Canada geese they 
studied. Hanson and Eberhardt (1971) observed mixing 
mainly after goslings were 4-5 weeks old. Geis (1956) sug- 
gested crowded conditions in the rearing areas might in- 
crease brood flocking. 
A comparison of the average number of goslings per 
brood (3.99) with the average number of goslings hatched 
per successful pair (4.2) indicates a maximum gosling sur- 
vival rate of 95%. The actual rate might have been lower 
because the fate of over 18% of the eggs in 1976 was 
unknown. The low hatching to fledging mortality (5% ) at 
Buldir is probably caused by the absence of mammalian 
predators and the abundance of buffer prey species for avian 
predators. Canada goose gosling survival rates of 90% or 
greater were found at four of the six locations cited by 
MaclInnes (1962). 
There has been considerable discussion about the use of 
family group counts to estimate waterfowl production. 
Sherwood (1967) found that pairs with grouped broods re- 
mained together in winter and exhibited family behavior. 
He pointed out that counts of such assemblages would bias 
estimates of productivity. In a different study area, Raveling 
(1969) demonstrated that families of Canada geese re- 
mained intact in winter, and Raveling and Lumsden (1977) 
presented the rationale for using such counts to predict pro- 
duction. 
We believe that most broods at Buldir were probably with 
their own parents when they fledged. We make this assump- 
tion on the basis of observations of color-marked birds, the 
size of broods (one to seven goslings per family), and the 
behavior of groups (alighting together) thought to be fami- 
lies (Raveling 1968). 
Mortality Factors— Eggs 
Loss of entire clutches. — Predation of nests in Canada 
geese is highly variable, with losses ranging from 5.0 to 
64.6% in most populations (Table 7). No Aleutian Canada 
goose nests were found completely destroyed by predators; 
desertion accounted for all the unsuccessful nests (9%). This 
rate is intermediate compared with other populations 
(Table 7). In 1974 one nest was lost in an earth slide, a 
potentially serious hazard on the unstable volcanic island. 
Loss of partial clutches. — Glaucous-winged gulls were 
observed eating goose eggs once, and 10 eggs with large 
holes like those expected after gull predation (Mickelson 
1975) were also found. Gulls are also known to swallow 
whole eggs (D. Raveling, personal communication). Most 
eggs with holes were found away from nests and would have 
been recorded as “unknown fate.” A comparison of hatch- 
ing success of goose nests at various distances from gull con- 
centrations (e.g., gull nesting colonies and loafing areas) 
Table 10. Nesting and hatching success of Aleutian Canada 
geese at various distances from glaucous-winged gull 
colonies, Buldir Island, Alaska, 1975-76. 



Distance to Nesting Hatching 
gull colony Number of — success Number of — success@ 
(m) nests (%) __ nests (%) 
0-50 70 (56)> 87 63 (56) 81] 
51-100 34 (27) 91 29 (26) 80 
Over 100 21 (17) 95 20 (18) 77 
Total 125 112 


aSuccessful nests only. 
bPercentage of total in parentheses. 
indicates no significant differences, but nesting success in- 
creased slightly as the distance from gulls increased (Table 
10). Abundant prey items were available to gulls; they took 
at least eight species of birds, several species of fish, insects, 
and berries (Trapp 1979). 
Herring gulls (Larus argentatus), glaucous gulls (L. 
hyperboreus), glaucous-winged gulls, and parasitic jaegers 
were major predators of goose eggs at other areas (Barry 
1956; Angstadt 1961; MacInnes 1962; Mickelson 1975; 
Bromley 1976; Raveling and Lumsden 1977), Proximity of 
geese to areas frequented by predators (Cooper 1978), 
abundance of buffer prey species (Angstadt 1961), and dis- 
turbance of incubating geese by humans (MacInnes 1962; 
Mickelson 1975) are factors that affect predation rates. 
Mortality Factors—Goslings and Adults 
Predation. — We were unable to directly assess the extent 
of gull and jaeger predation on young goslings because we 
avoided the nesting areas from onset of hatching until most 
broods were 2-3 weeks old. Nevertheless, we saw a gull pick 
up and drop, apparently unharmed, a Class Ib gosling, and 
a parasitic jaeger was seen diving at a brood of Class Ib 
goslings. 
On the Yukon-Kuskokwim Delta, Alaska, most losses of 
B. c. minima goslings occurred during their first 2 weeks 
of life (Mickelson 1975). In the same study, parasitic jaegers 
were able to take only Class Ia goslings and were not im- 
portant predators. Gulls and jaegers affected goslings 
similarly in the Northwest Territories (MacInnes 1962). 
Bald eagles occasionally took flightless goslings at Buldir. 
Food remains from the single active eagle eyrie were 
examined periodically throughout the summer and fall of 
1974 and 1975, and the remains of only one goose was 
found, a banded gosling in 1975. Three wings of adult geese 
were found in an eagle eyrie at Buldir in 1963 (Jones 1963). 
As many as five eagles hunted geese frequently in early to 
mid September after most other prey species had left Buldir. 
The fresh remains of five molting geese killed by eagles were 
found during the study, and old bones and wing feathers 
of 10 geese were also found. Peregrine falcons and snowy 
owls may have taken geese, but no direct evidence was 
found, 
