FLUCTUATIONS OF DEER MICE IN ONTARIO IN RELATION TO SEED CROPS

Abstract

We trapped deer mice in Algonquin Park, Ontario, Canada, over a 36-year period. Using information-theoretic methods, we examined relationships among weather variables, sugar maple seed crops, and mouse populations. Deer mouse populations were naturally regulated with stochastic variation in growth rate from year to year influenced strongly by autumn seed crops mediated by temperature during the previous summer and snowfall during the previous winter. Population peaks followed heavy seed crops but did not occur at regular intervals. Demographic characteristics differed between peak and nonpeak populations. Overwinter losses were considerably less entering peak years than in other years. Although overwintered adults were only slightly more numerous in the spring of peak years, they were heavier than in nonpeak years. Breeding began earlier in peak years and in years with high March temperatures. The proportion of young was greater in the spring of peak years but less than in nonpeak years in summer. Apparent survival was greatest in spring and early summer of peak years. Immigration in summer was also greatest in peak years. A combination of the above factors resulted in rapid population growth in spring and early summer of peak years. However, young grew slowly in peak summers and most did not reach breeding mass in contrast to nonpeak years. Breeding was drastically reduced for all sex and age groups in August and September of peak years but continued all summer in nonpeak years. Intensity of late summer breeding was positively related to the current seed crop. Apparent survival declined sharply in late summer of peak years but not in nonpeak years. Thus, declines in breeding and apparent survival were responsible for ending population peaks. We interpret these recurrent events principally as responses to variation in food supply.