The reproductive success of female red squirrels (Tamiasciurus hudsonicus preblei Howell) was studied from Spring 1988 to Spring 1991 in a jack pine (Pinus banksiana) forest in central Alberta, Canada (54.20$\sp\circ$N, 114.45$\sp\circ$W). There were two major objectives: (1) document the natural variation and potential correlates of reproductive success within each year, and (2) conduct experimental manipulations to test the relative importance of several potential factors in affecting reproductive success. Using a combination of live-trapping, visual observations, and radio telemetry, the reproductive success of mother squirrels was measured by documenting the fates of individual offspring, starting from birth.

Factors such as parturition date, previous breeding experience and territory area were relatively unimportant in determining a female's reproductive success, whereas litter size (range 1-4) had a significant influence. Also, the mean diameter of trees on the territory (inversely related to tree density) was related to offspring survival to weaning and beyond. Females allowed their offspring to reside on the natal territory until offspring were able to obtain their own territory. However, approximately half of dispersing offspring acquired territories on or adjacent to the mother's territory, with no sex-biased patterns being evident. This suggests that females suffer little from the close settlement of progeny and/or that long-distance emigration away from the natal area is too costly.

Manipulations of the overwinter food hoard of females were conducted using natural food items (pine cones) and artificial, high-energy items (sunflower seeds). Females with hoard additions produced significantly higher litter sizes than control animals. A census of the cones cropped by squirrels during the autumn hoarding period suggested that individuals were not limited in their hoard size by the absolute number of cones.

The importance of territory quality to reproductive success was explored through a series of removal experiments. These experiments indicated that (1) neither non-breeding or breeding females responded to habitat vacancy by relocating their entire territory, (2) squirrels did not preferentially visit or settle on vacant territories that originally belonged to males, breeding females or non-breeding females, and (3) females with previous breeding experience shifted their territories in order to encompass neighbouring, vacant areas. These new areas were then allocated to offspring in the following year.