Regulation of Numbers in an Island Population of the California Vole, A Problem in Community Dynamics

Abstract

A population of California voles, Microtus californicus, living on Brooks Island in San Francisco Bay was studied for 13 years. During the first 2 years following its establishment, the performance of the population was significantly different from that in the subsequent 11 years. The 11—year pattern that developed is characterized by: (1) annual peaks in abundance; (2) a delay of up to 5 months between the return of wet season conditions and population growth; and (3) an alternating pattern of "high" and "low" winter densities. Reproduction shows a strongly seasonal pattern, beginning 1—2 months after the start of the rainy season and ending with the desiccation of the vegetation in June. In contrast to this, reproduction started promptly with the autumn rains during the initial 2—year colonization phase, and limited summer reproduction occurred as well. Within the breeding season, litter size and number of corpora lutea produced per ovulation episode show marked seasonal changes which appear to be independent of annual variations in demographic detail and condition of the mice. Males are more vagile than females, although both show a reduction of movements during the main breeding period. The incidence of wounding was positively correlated with reproductive condition in males, but not in females. Mortality schedules were studied by means of population age structure, recruitment rates, and sex ratios. Apparent aging ceases for almost 5 months from about 1 month after the dry season begins to 60 days after it ends. This period is associated with heavy mortality, particularly among old males and young females. Consequently, late fall and early winter sex ratios favor males. In general, however, females survive longer than males, and during the breeding season the sex ratio progressively favors females. Recruits first appear 1—2 months into the wet season, but initially their numbers are very few and are insufficient to reverse population declines. Rapid population growth in late spring is largely dependent on reproduction by these early recruits. Physiological condition was measured by study of body—weight changes, particularly ratios of body weight to body length, ectoparasite loads, including scabies mite infections (mange), and molting patterns. The body—weight data reveal that wet season condition, especially in spring, is much better than dry season. Surprisingly, males begin to recover during the winter period whereas females continue their decline. Moreover, the whole pattern of weight losses is more severe in poor survival summers than in the alternate ones. A flea—mite index reveals a strong seasonal pattern which seems to be induced by vole demography; during the colonization period, ectoparasites were barely noticeable. It took 4 years for scabies to reach significant levels of infestation. Highest incidence generally occurs in winter, and an outbreak occurred in the seventh winter; subsequently, the mange mites returned to endemic levels. Molt activity showed two strongly marked seasonal peaks, one in the late spring—early summer and one in the autumn. Males molt at a slower rate than females; female molt is strongly inhibited by reproduction. The autumn molt begins before the wet season arrives suggesting a photoperiodic cue. During colonization, molts were completed more quickly than afterwards. The quasi 2—year cycle of regularly recurring peaks every year and alternating high and low winter densities which came to characterize the Brooks Island vole population is considered to be the result of a regulation process in which a multiplicity of factors interact to achieve regulation. Both density—unresponsive and responsive factors are involved: unresponsive–rainfall pattern, photoperiod and/or gonadotrophic potency of diet, heavy autumn dews, and possibly winter sex ratios favoring males; responsive–dry season resource levels, long—term physiological damages, and ectoparasite load, including scabies infections. The first two factors in each category are probably the essential ingredients. During the colonization phase, regulation involved the additional factors of emigration and possibly intraspecific strife of some sort. Finally, a model is proposed to explain the longer term cycles in abundance presumably characterizing mainland populations. It is suggested that a 2—year cycle (peak every other year) would be generated without the involvement of mammalian predation. Such a cycle would result from the action of all the essential factors proposed for Brooks Island, plus a supposedly more severe dry season and much more extensive emigration. Longer term cycles probably require the further addition of mammalian predators to this complex. Thus, changes in numbers in this species can be explained most satisfactorily in terms of the interacting effects of at least six variables.