| Summary: | Collared lemmings (Dicrostonyx groenlandicus) fluctuate periodically in abundance on
the Kent Peninsula, Northwest Territories, Canada. I tested whether predation was necessary to
(1) limit the lemming population during a peak and decline summer (1996 and 1997), and (2)
cause the population decline. The nesting success of geese also fluctuates at this location. I
tested whether the relative number of goose nests lost could be predicted from functional and
numerical responses of nest predators to lemming density.
I reduced predation on lemmings with a fence and an overhead mesh of monofilament
line over 11 ha, completed during the lemming increase (summer 1995). I used mark-recapture
and radio-telemetry to investigate demography in this "Exclosure" and three Control areas, and
estimated winter predation from droppings and abundance of predators.
Density increased in Exclosure relative to Controls in both summers. Survival was
significantly higher within Exclosure during the lemming decline only. Neither proportions of
reproductive animals nor net movements differed significantly between treatments. I conclude
that predation was a necessary limiting factor in the peak and decline summers, but that the
magnitude of limitation was greater in the decline.
Density declined over winter on all sites. The next summer (1997), the decline
accelerated on controls but was reversed within Exclosure. Summer survival was lower in the
decline than in the peak. Lemmings stopped reproducing early in the peak summer. If there was
no winter breeding, winter survival was high. The contribution of predation to the winter decline
depends on the extent to which mortality factors were compensatory. I conclude that predation
may not have been necessary to cause the decline; instead the decline was initiated by cessation
of reproduction. However, predation accelerated and extended the decline the next summer.
From observed responses of nest predators to lemming density, I predicted that the
number of goose nests depredated should be lowest during lemming peaks. This was not true in
past lemming cycles in this region (1989 - 1994). I conclude that nest loss cannot be predicted
from the phase of the lemming cycle alone, and make testable predictions of how other factors
should affect nest loss.
|
|---|
