J. Gurnell, THE EFFECTS OF FOOD AVAILABILITY AND WINTER WEATHER ON THE DYNAMICS OF A GREY SQUIRREL POPULATION IN SOUTHERN ENGLAND, Journal of Applied Ecology, 33(2), 1996, pp. 325-338
1. The population ecology of grey squirrels Sciurus carolinensis livin
g in a 9-ha oak Quercus robur wood in southern England was studied bet
ween 1976 and 1987 using live-trapping techniques. Trapping was carrie
d out in winter, spring and summer. The availability of tree seeds dur
ing the autumn of each year, and the severity of cold weather over eac
h winter were also measured to examine their effects on squirrel popul
ation dynamics. 2. Capture probabilities of squirrels in winter, and t
o a lesser extent in spring, were inversely related to food availabili
ty and data from these two seasons were not considered dependable. The
analyses concentrated on the summer populations. 3. The long-term ave
rage summer density of squirrels was high at 8.8 ha(-1) (SE 3.41 ha(-1
)) demonstrating that the oak wood was high quality habitat for grey s
quirrels. Over 10 of the 12 years, summer densities were remarkably si
milar, ranging between 7 ha(-1) and 10 ha(-1) (mean 8.5 ha(-1), SE 0.9
5 ha(-1)). However, numbers were driven upwards in 1977 to a density a
pproaching 18 ha(-1) and downwards in 1982 to a density of about 3 ha(
-1): a 6-fold difference. 4. In good seed years, breeding starts in De
cember, in poor seed years the start of breeding is deferred until the
spring. There was no or very little spring breeding in 5 years when f
ood supplies were poor. Female reproductive success was positively ass
ociated with food availability. Partial correlation analysis showed th
at the level of association was not improved when the effects of winte
r weather were taken into account. The number of new adult females in
the summer population was positively associated with food availability
but there was no association between new males and food. This suggest
s that food availability is more important to breeding females than br
eeding males. 5. Persistence from summer to winter was positively asso
ciated with food availability, but persistence from summer to spring a
nd to the following summer were not. Partial correlation analyses show
ed that the severity of winter weather tended to mask the effects of f
ood availability on persistence and the partial correlation coefficien
ts were higher when the effects of weather were held constant. Adult f
emales had the highest persistence between summers (52%, n = 11, SE 6%
), followed by juvenile females (38%, n = 7, SE 8%), adult males (36%,
n = 11, SE 4%), and juvenile males (21%, n = 7, SE 7%). The persisten
ce of adult males but not females was inversely related to the initial
number of males present. 6. In males, there was no significant differ
ence in winter body mass between years and winter body mass was not as
sociated with food availability, although there were very few data for
winters when the food supply was good. In 7 years when the food suppl
y was poor to moderate, there was an increase in mean body mass betwee
n summer and winter in 4 years, and a decrease in 3 years. 7. The data
were explored using ordination techniques; first a standardized princ
ipal component analysis and then the canonical form of principal compo
nent analysis or redundancy analysis. The analysis was carried out on
years derived from MNA of males and females and various combinations o
f the environmental variables: food availability (FOOD) and the severi
ty of winter weather (TEMP). The ordination biplots clearly showed the
high correlation between MNA males and MNA females and that FOOD was
the most important environmental variable. TEMP on its own had no effe
ct but FOOD x TEMP was important. 8. Following on the exploratory data
analysis and for predictive purposes, a general linear model between
the numbers of squirrels in the summer populations and sex, FOOD and F
OOD x TEMP as explanatory variables accounted for 77% the variance in
squirrel numbers among years. 9. This study shows that tree seed avail
ability is the most important factor limiting grey squirrel densities,
but this factor both positively and negatively interacts with the sev
erity of winter weather to affect grey squirrel population dynamics.