Population Ecology of Alberta Red Squirrels

Abstract

Temporal changes and spatial differences in population structure and density of red squirrels (Tamiasciurus hudsonicus) were investigated in 3 habitats near Rochester, Alberta. Between 1966 and 1969, red squirrels were more abundant in spruce (Picea mariana and Picea alba) stands (161—684/100 ha), less abundant in jack pine (Pinus banksiana) stands (86—264/100 ha), and least abundant in aspen (Populus tremuloides) woods (0—99/100 ha). Summer populations fluctuated between 67 (1966) and 151 (1968) per 100 ha of mixed habitat, but spring populations were more stationary (32—49/100 ha of mixed habitat). Ovulation rates of adult and yearling squirrels and the percentage of yearling ♀ ♀ which bred varied annually. Variation in annual reproductive rate (2.4—4.4 young/♀) was correlated with the date of onset of breeding, and the number of days of snow cover subsequent to 1 January. Juvenile sex ratios were 1:1 but there was a trend toward preponderance of ♂ ♂ in older age classes. Although sex ratios on the study areas were close to 1:1 during early summer, late—summer mortality of adult ♀ ♀ resulted in a preponderance of ♂ ♂ (60%) during the remainder of the year. Equality of sex ratio was restored during the breeding season when ♂ ♂ experienced high mortality rates. After fall dispersal, the percentage of adults was higher in the spruce (78—85%) than in pine (30—50%) but virtually 100% of the squirrels occupying aspen habitat were juveniles. Changes in age ratios suggested that juveniles suffered higher overwinter losses than did adults. Spring age ratios were not related to fall age ratios because of differential annual survival of juvenile squirrels. Dispersal movements occurred in spring and fall; squirrels were relatively sedentary the remainder of the year. The spring shuffle involved breeding ♂ ♂ visiting territories of ♀ ♀ and redistribution of squirrels from deciduous and pine forest into spruce forest. Almost all fall dispersants were young of the year. High adult densities in spruce stands prompted emigration of most juveniles from these areas. Fewer juveniles emigrated from jack pine where adult densities were lower and territories were available. Juveniles apparently dispersed into the large tracts of aspen—dominated deciduous forest. Survival of squirrels varied seasonally and annually, among habitats and between sex and age cohorts. Heavy mortality of adult ♀ ♀ and nestlings occurred between parturition and weaning (July—September). The overwinter period (October—February) was one of high juvenile losses. During the 3—mo breeding season (March—May), ♂ ♂ experienced higher mortality rates than ♀ ♀. Annual survival was highest in the spruce and lowest in the aspen. Survival rates were highest during 1966—1967, °30% lower in 1967—1968, and °44% lower in 1968—1969. Survival of juveniles was lower in all years than survival of adults. In the wild an average of 639 meristematic buds, and the seeds from 35 pine cones were consumed by a single squirrel each day. Captive feeding trials indicated that an average of 176 black spruce cones or 123 white spruce cones per day were necessary to sustain 1 squirrel when offered no supplementary foods. Territorial boundaries in spruce and pine did not change detectably between 1966 and 1969. Within monotypic habitats, territories were similar in size. Territories in white spruce were smallest (x = 0.24 ha), those in mixed spruce larger (x = 0.35 ha), and those in jack pine much larger (x = 0.66 ha). Territory size may be related to cone supply in poor mast years. In 1967, a year of poor cone mast, sufficient numbers of cones were produced per territory to provide one—half of the annual energy requirement of 1 squirrel. The stability of spring populations, despite the change in mast crop from very poor (1966 and 1967) to medium—heavy (1968), implies regulation of squirrel densities at levels which preclude critical food shortage. We hypothesize that territorial behavior is the mechanism which regulates densities about a mean determined by food supply in years of poor cone mast, and that it operates through density—dependent mortality by excluding nonterritorial squirrels from winter cone supplies. Similarities in red squirrel densities (or territory sizes) in similar forest types recorded by numerous authors in different areas and different years imply consistency of territory size within particular habitats. We believe that variation in numbers of squirrels harvested primarily reflects variation in annual production.