SPRING DECLINES IN MICROTUS-PENNSYLVANICUS AND THE ROLE OF STEROID-HORMONES

1. Spring declines are a common feature of small mammal demography. We tested the hypothesis that meadow voles from populations experiencing severe spring declines should exhibit severe stress responses as indicated by high free corticosterone levels. 2. Three populations in southern Ontario were live-trapped intensively in the spring of 1985, all animals were bled each time, and measurements of total corticosterone, corticosterone-binding globulin (CBG), and total androgen (testosterone and dihydrotestosterone) were obtained by radioimmunoassay. 3. Population density was correlated to stress responses: in populations with the highest densities, both males and females had the highest total corticosterone, lowest CBG, and highest free corticosterone levels; males had the highest wounding levels. However, these populations declined the least. In contrast, the population with the lowest density had males and females with the lowest total corticosterone, highest CBG, and lowest free corticosterone levels and males with the lowest wounding levels. This population declined the most. 4. Total corticosterone and CBG levels in males were about half those in females, and in females varied significantly with reproductive condition, with lactating and pregnant females having the highest levels. Only in pregnant females was body mass positively related to total corticosterone and CBG levels and this is probably related to impending parturition. In males, androgen levels were similar among populations and not correlated to total corticosterone or CBG levels nor to body mass. In females, androgen concentrations were only 4% that of males. In males, CBG, free corticosterone, androgen levels, and wounding rates showed significant repeatability. 5. We conclude that stress responses were not related to population demography in our study and reject the Christian stress hypothesis and the adaptive stress hypothesis of Lee and Cockburn as explanations for spring declines in this species. We propose an adaptive model to account for the differences in the hormonal response to breeding between species in which the males are semelparous and those in which males are iteroparous.