VARIATION IN NUMBERS IN A HOUSE MOUSE-POPULATION HOUSED IN A LARGE OUTDOOR ENCLOSURE - SEASONAL FLUCTUATIONS

(1) The seasonal fluctuations in a population of mice housed in an outdoor pen (5.7 .times. 14.6 m) were studied for 12 years at 4-weekly intervals (''months''). Throughout the period, the number of shelters and the supplies of food remained constant. (2) Total numbers usually peaked between February and May (late summer-early autumn in the Southern Hemisphere). The monthly variations in total population size were largely due to monthly variations in the numbers of juveniles recruited. (3) The number of juveniles recruited during the 12-year period represented 8.0% of the possible total. This shortfall was due to the failure of many females to become pregnant (52%), and to losses, between birth and weaning, of young born to those females which did become pregnant (85% of the estimated number born). Monthly variations in the numbers recruited were probably due to variations in the number of adult females breeding; this peaked in December-January (mid-summer). The peak in the number of adult females breeding was preceeded by a peak in the percentage of the adult female population breeding in October-November (early summer). (4) The monthly mortality rate for the 12-year period was about 20%. In most years the rate was similar at all seasons. (5) Mean ages of both males and females were higher in winter than in summer because more young were recruited in warm weather than in the cold. (6) About 32% of the males and 19% of the females recruited into the pen population moved from their home shelters to neighboring shelters. Both the numbers moving and the numbers living in foreign shelters were similar in all months. (7) Although the mean number of mice in shelter was lowest in winter, the mean number of mice in the most crowded nest-box peaked in mid-winter. (8) Of the females that reached adulthood 43% were breeding on one or more occasions. The percentage breeding was highest among cohorts recruited between August and November (late winter-spring). (9) The mean life-spans of males and females recruited into the pen (4.0 and 5.4 months, respectively) were 33% and 43% as long as those of males and females reared in laboratory cages (25 .times. 15 .times. 12.5 cm). The life-spans of cohorts entering the population between September and November (spring) and between March and May (autumn) were slightly, but significantly, longer than those of cohorts entering the population between December and February (summmer) and between June and August (winter). (10) Body weights at 4 months of age were greater for cohorts of males and females recruited into the population in August and September (spring) than for cohorts recruited at other seasons. (11) All the fluctuations observerd in the population must have been due to factors other than fluctuations in the availability of food and shelter. We suggest that these factors were seasonal fluctuations in ambient temperature and seasonal changes in the number of adult females in the population because both these factors are known to affect the percentage of adult females breeding.