274 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol 123. No. 2. June 2011 
Canada Geese, as strong territoriality is more 
common (Mowbray et al. 2002, Reiter 2009). The 
lack of association between nest density and fox 
den occupancy combined with evidence of a 
strong correlation of nest density and hatch date in 
our study suggested that spring phenology and 
nesting geese in good physiological condition (as 
represented by nest density and hatch date) best 
described among-ycar variation in success of 
dispersed Canada Goose nests. 
The bird-lemming hypothesis and, more gener¬ 
ally. the alternative-prey hypothesis as applied to 
the system we studied, assumes arctic fox 
populations specialize on lemmings when lem¬ 
mings are abundant, functionally respond and 
switch to nesting birds when lemmings decline, 
and this interaction occurs at some relevant spatial 
scale. Arctic fox, under this assumption, are 
present and depredate goose nests even when 
lemming abundance has declined. However, if 
alternative prey, particularly nesting birds, is not 
abundant when lemming abundance decreases, 
foxes may not breed and become nomadic or 
substantially increase the size of their home range 
(Wrigley and Hatch 1976. Eberhardt et al. 1983. 
Eide et al, 2004). The extent of increased arctic 
fox mortality or emigration and the magnitude of 
a decline in fox abundance, following a lemming 
peak, likely depends on timing of the lemming 
population crash (Tannerfeldt et al. 1994, An- 
gerbjom et al. 1999). abundance of alternative 
food resources (natural [Roth 2003J or man-made 
[Eberhardt et al. 1982. 1983]), prevalence of 
disease (Kaplan 1985. Ballard el al. 2001), and 
how foxes respond to these factors. 
The relationship between arctic fox abundance 
and changes in lemming abundance may be 
confounded by the effects of alternative prey 
resources off our study area. Bahr (1989) reported 
high year-to-year variability in the presence of 
lemmings in the arctic fox diet at Nestor One, but 
little variability in the presence of adult geese or 
goose eggs. This suggests arctic fox are lemming 
specialists and do not compensate for reduced 
lemming abundance through increased consump¬ 
tion of goose eggs on the Nestor One study area. 
The La Perouse Bay Lesser Snow Goose and 
Ross's Goose colony, which is <20 km from 
Nestor One and has 2> 20,000 nesting pairs of 
Snow and Ross’s geese, provides a much higher 
density of potential food for arctic fox than 
Canada Geese at Nestor One, when lemming 
abundance is low. Human habitation sites may 
also attract foxes during low primary prey years 
(Eberhardt et al. 1982. 1983). The town of 
Churchill. Manitoba is <60 km from our Cape 
Churchill study area and may provide a reliable, 
concentrated food resource (e.g., an open dump; 
which operated during our study) for arctic foxes. 
The magnitude of the effect of variation in 
predator abundance and their primary prey 
(lemmings) on alternative prey (goose eggs) may 
change depending on whether the funcuonal 
response of predators is to remain in the local 
area, switching to bird nests (i.e., a closed 
predator population), or to disperse following a 
decline of lemmings. Ultimately, predator, prey, 
and alternative-prey dynamics in this system may 
be operating at a larger spatial scale than our 
study area, highlighting the importance of con¬ 
sidering spatial scale when evaluating the bird¬ 
lemming hypothesis. 
Nest success of Canada Geese near Cape 
Churchill exhibited strong annual variation and 
high arctic fox den occupancy was associated with 
low nest success. Previous studies of predator 
pressure on northern nesting geese in simple trophic 
systems supported the predictions of the bird¬ 
lemming hypothesis (Wilson and Bromley 2001; 
BSty et al. 2001, 2002) with geese experiencing 
reduced nest survival in years following a lemming 
peak. Our relatively long-term study (12 years) was 
conducted in a landscape inhabited by geese with 
dispersed-nesting ecology and our results were not 
consistent with predications of the bird-lemming 
hypothesis. We suggest that, at least for dispersed- 
nesting geese, the spatial distribution of predators 
may be most important to nest survival, regardless 
of the abundance of small mammals in the local 
ecosystem. Further research is needed to understand 
factors influencing the spatial distribution and 
abundance of arctic fox, particularly in light of 
rapidly increasing numbers of Lesser Snow Geese 
and Ross’s Geese throughout this region of arctic 
and subarctic Canada (Jefferies et al. 2006), and the 
appropriate spatial scale at which to evaluate the 
dynamics of these trophic interactions. 
ACKNOWLEDGMENTS 
We thank the Mississippi Flyway Council Technics 
Section, Manitoba Conservation, Minnesota Departnwnl of 
Natural Resources, Iowa Department of Natural Resource' 
Missouri Department of Conservation. Arkansas Game and 
Fish Commission. U.S. Fish and Wildlife Service, Canadian 
Wildlife Service, and Parks Canada (particularly the staff of 
Wapusk National Park), for support throughout this work 
in particular, M. M. Gillespie. Garth Ball. A. H. Raedeke. 
