Following on the heels of Dennis Pauslon's bold test of
the "fair use doctrine", I have copied the discussion
section from:

KERLINGER, P., M. R. LEIN, and B. J. SEVICK. 1985. Distribution
and population fluctuations of wintering Snowy Owls (Nyctea
scandiaca) in North America. Can. J. Zool. 63: 1829-1834.

Discussion

Our analysis yielded a somewhat different picture of the
distribution and dynamics of wintering populations of the
Snowy Owl than previous reports (e.g., Gross 1947). Numbers
of wintering owls fluctuate dramatically from winter to winter
within both the eastern and western factor areas, but numbers
are usually low. These movements are analogous to the irruptions
(sensu Gauthreaux 1982) of boreal seed-eating birds (Bock and
Lepthien 1976). On the Great Plains of Canada (GPN), and possibly
the northern Great Plains of the United States (GPM), Snowy Owls
are not irruptive. Instead, they are regular migrants in these
regions. The average winter abundance of owls in the GPN is
nearly twice that of any other region and > 10 times that of
most regions. Even in years when owls were not abundant in the
GPN, OWLFREQ was usually greater than or equal to peak years
from most other regions (Fig. 3). The large fluctuations in the
GPN region from 1952 into the early 1960's are undoubtedly a
result of the small number of localities conducting Christmas
counts. After 1963 the fluctuations become greatly damped. The
relative lack of large fluctuations in owl abundance from winter
to winter is also the probable reason for the lack of correlations
among the Great Plains regions.

Evidence for a regular migration of owls to the Great Plains comes
from several sources. Many of the owls that winter in southern
Alberta are territorial for several months (Boxall and Lein 1982).
Furthermore, winter site fidelity, as demonstrated by banding
returns, has been reported for several individual owls (Oeming 1957,
1964; Follen and Leupke 1980). These findings, when combined with
our analyses, suggest that the Great Plains is the prime wintering
area for thousands of Snowy Owls.

The irruptions of owls in eastern and western North America are
the movements interpreted as cyclic invasions by previous
researchers (Shelford 1945; Gross 1947, Chitty 1950). Time series
analysis did not support the contention that owl irruptions occur
precisely at 3 to 4 -year intervals. Although irruptions did occur
frequently at intervals of 3 and 4 years, they were not strictly
periodic. Most importantly, cycles were absent on the northern
Great Plains where owls were more numerous in nearly all winters
than they were in irruption years in either eastern or western
North America. Garsd and Howard (1981) reported a similar absence
of cycles among several species of microtines. Their results are
important because irruptions of Snowy Owls have been causally
linked to microtine fluctuations in the arctic (Shelford 1945;
Chitty 1950). We feel that use of the term 'cyclic" to describe
Snowy Owl populations is unjustified.

Why owl numbers fluctuate synchronously over large geographic areas?
Wintering owls fluctuated in synchrony from the Continental Divide
westward to the Pacific Ocean and from the western Great Lakes
east to the Atlantic Ocean. That invasions were not always continent
wide has been noted by Gross (1947) and Hanson (1971). Hanson 11971)
used a limited series of Christmas Bird Counts to demonstrate
asynchrony between Pacific coast owl migrations and those on the
western edge of the northern Great Plains. Our data show that there
are at least two areas within which migrations were synchronous,
yet independent of the fluctuations of owls in the other geographic
areas. The Great Plains possibly represent a third area of
independent population fluctuations.

Synchronous variation in owl numbers over large geographic areas
requires an explanation. We do not feel that the traditional
explanation, that Snowy Owl irruptions reflect the 3 to 4 -year
cycle in their lemming prey, is tenable for two reasons. First, we
did not find evidence that Snowy Owl irruptions occur at regular
3 to 4 -year periods. Second, the explanation assumes that the
lemming populations fluctuate in synchronously over vast areas of
the Arctic (Shelford 1945; Gross 1947; Chitty 1950). There is no
evidence that supports this assumption. The largest known area
within which lemmings fluctuate synchronously is only about
250 000 ha (reviewed by Maher 1970). Areas of this size would probably
support only a few hundred pairs of breeding Snowy Owls (Pitelka
et al. 1955). We believe that the tundra would be more realistically
viewed as a mosaic of patches varying in size, lemming abundance,
and timing of fluctuations. Even the largest of these patches could
not have acted as a source area for the irruptions recorded in
eastern and western North America. To explain the occurrence of large.
geographically synchronous irruptions, we require a factor operating
on a geographic scale much larger than the patches of the lemming
mosaic. Source areas for Snowy Owl irruptions must necessarily be on
the order of millions of hectares. One possible factor that operates
on such a scale is weather. An investigation of the relationships
between snowfall and temperature conditions on the arctic breeding
grounds in late fall and early winter, and the abundance of Snowy
Owls in southerly locations in the same winters, may provide new
insights.

--
Mike Patterson
Astoria, OR
celata at pacifier.com

Christmas Bird Count Calendar and FAQ for Oregon and Washington
http://home.pacifier.com/~mpatters/cbc/cbc_WAOR_reg.html