Dynamics and Regulation of Red Squirrel (Tamiasciurus Hudsonicus) Populations

This paper describes red squirrel population dynamics on two intensive study areas (Camp and Main) in mixed—forest types near Rochester, Alberta. It also examines local and regional population fluctuations, and their relationship to cone crops and weather factors. Adult red squirrel numbers on our study areas varied little from 1967 to 1968. Yearly differences in reproductive output were caused principally by changing ovulation and pregnancy rates. Mean litter size increased significantly from 3.4 in 1967 to 4.3 in 1968; while the per cent adult females breeding increased significantly from 67 to 88. A life—table analysis of age—ratio data estimated mean annual mortality among juveniles (postweaning) at 67%; and an adult mortality rate of 34% for yearlings and 61% for older cohorts. A time—specific estimate of adult mortality on the Camp study area during the year starting summer 1967 was 21%. Red squirrel territories appeared to be of two distinct types: (1) defended winter food caches which were subsequently abandoned during the summer, and (2) "prime" territories in which a specific area was defended year round. During the summers of 1967 and 1968, 31% and 26% of study—area adult populations occupied prime territories. The distribution of prime territories chiefly reflected the presence of mature seed—producing conifers, and hence a potential year—round food supply. Deciduous areas were particularly important in overwintering the juvenile cohort. Fur returns were used as population indices in determining synchrony and periodicity of red squirrel fluctuations. Fluctuations tend to occur synchronously over much of the Prairie Provinces. Furthermore, mean intervals between peak years in Alberta (2.6) and Saskatchewan (2.8) were significantly shorter than in random series of comparable length, while the 2.9—year mean interval in Manitoba approached significance. A statistically significant correlation was found between white spruce cone crops and red squirrel populations in Alberta as indexed by annual fur harvests. The negative correlation between cone crops in late summer and rainfall during summer of the preceding year was almost significant. This supported the widely held view that weather factors influence bud differentiation and hence determine the size of the following year's cone crop. Our contention that cone production may in this way provide a vehicle through which weather affects squirrel populations was supported by a number of statistically significant negative correlations between Alberta and Saskatchewan fur returns and the preceding year's summer rainfall. Such a mechanism would explain the observed widespread synchrony of population fluctuations, since extensive regional weather patterns could be involved.