W E Ruth. Bruce Coleman 
Weather could drive the cycle indi- 
rectly. Lightning-caused fire is the sin- 
gle most important initiator of boreal 
forest succession, and early successional 
forests provide ideal hare habitat. 
Nearly fifty years ago, Wallace Grange, 
an early student of snowshoe hare ecol- 
ogy, suggested that the cycle is driven 
by cyclic wildfires, and there is evidence 
to support this theory. A strong correla- 
tion, for example, exists between the 
periodicity of wildfire in the Canadian 
boreal forest and the periodicity ol Can- 
ada lynx pelt returns recorded by the 
Hudson’s Bay Company. 
The connection, however, is probably 
not as direct as Grange supposed. In 
western Alaska, the hare cycle is pro- 
nounced in riparian willow habitats, 
which are produced by spring flooding 
and ice scouring, but fire in the region is 
a relatively rare event and is probably 
not in phase with the hare cycle. More- 
over, if a continent-wide fire cycle did 
exist, it could hardly provide the expla- 
nation for regional hare cycles that are 
often out of phase with each other. Nev- 
ertheless, there is no question that by 
creating large expanses of ideal hare 
habitat, fire strongly affects the poten- 
tial for extreme expressions of the cycle. 
Flooding and insect outbreaks are other 
forms of disturbance that can set boreal 
forest succession in motion and thus 
provide the environment in which the 
hare cycle operates. 
Population ecologists now generally 
agree that the hare cycle is density de- 
pendent, but they have widely divergent 
views as to what initiates the decline in 
the population once the number of hares 
per unit area in a region approaches 
some critical upper density. One school 
of thought holds that the hares have 
evolved intrinsic population regulation 
mechanisms driven by aggression. In- 
creased aggression between hares at 
times of high density may lead either to 
psycho-physiological stress in the sub- 
missive animal, which causes failure of 
the adrenal gland and, ultimately, death 
or to dispersal of submissive individuals 
into habitats where the probability of 
survival is low. 
Another school maintains that rather 
than controlling their own numbers, the 
hares are regulated by food shortages, 
disease, parasitism, predation, or some 
combination of such constraints. A 
model suggested by some ecologists pre- 
dicts that an increase in hares is fol- 
lowed by an increase in hare pred- 
ators — for example, Canada lynx, red 
fox, great horned owl, and goshawk — 
until the hare population can no longer 
withstand the predation rate and 
crashes. Following the disappeaiance of 
the hares, predators starve or emigrate 
to more rewarding hunting grounds. Re- 
leased from intense predation, the hare 
population rebounds, and another oscil- 
lation of the cycle begins. Although ap- 
pealing because of its biological and 
mathematical elegance, this model does 
not appear to be a sufficient account of 
the cycle’s mechanism. Given the hares’ 
tremendous capacity for population in- 
crease, even the combined toll of all 
their potential predators would be un- 
likely to contain an expanding hare pop- 
ulation unless some more fundamental 
constraint had first initiated the decline. 
Increased predation, disease, and para- 
sitism may aggravate the hares’ decline, 
but they cannot cause it. 
A more likely explanation seems to be 
that the decline phase is initiated when 
the hares destroy their preferred winter 
food, the twigs and bark of disturbance- 
adapted early successional trees and 
shrubs. The most fascinating aspect of 
the cycle, however, is, not that the hare 
population overshoots its food supply 
and subsequently crashes because of 
malnutrition, but the length of time — at 
least one hundred years and probably 
much longer — that the subarctic popu- 
lations have exhibited an approximately 
ten-year periodicity. 
Ecologists G. Evelyn Hutchinson of 
Yale University and Robert May of 
Princeton University have proposed that 
such cycles are a result of the delay with 
which species respond to changes in 
their populations or in the populations of 
interacting species. If an herbivore pop- 
ulation, for example, is slow to adjust to 
a decline in the plant populations on 
At the peak of the ten-year snowshoe 
hare cycle, the woody plants preferred 
by the animals are overbrowsed. 
In response, the plants produce 
shoots unpalatable to the hares. 
The deterioration in the food supply 
appears to last long enough to cause 
a crash in the hare population. 
which it feeds, it tends to overshoot the 
food supply and undergo a malnutrition- 
caused, overcompensatory crash. The 
extent of the crash and the speed of the 
recovery depend, in part, on just how 
long the delay is, relative to the periodic- 
ity of the cycle. Chaotic, violent oscilla- 
tions, which can lead to the extinction of 
the herbivore, begin to occur if the delay 
is long enough. Shorter delays produce 
less extreme oscillations. When, for ex- 
ample, some property of the vegetation 
recovery process retards the recovery of 
the herbivore population for a length of 
time equivalent to one-fourth the period 
of the cycle, the oscillations will be 
marked enough to produce a severe 
crash of the herbivore population, but 
not so severe as to prevent the herbivore 
from rebuilding to its previous peak. 
Hutchinson and May propose that the 
rise and fall of the snowshoe hare is an 
example of this kind of cycle. According 
to this theory, since the period of the 
hare cycle ranges from eight to twelve 
years, the hares should take two to three 
years to respond to the recovery of the 
plants. My evidence suggests that this is 
the case. 
What sort of mechanism could con- 
W E Ruth 
48 
