A Multivariate Study of the Relationship between Dispersal and Demography in Populations of Microtus ochrogaster in Eastern Kansas

Abstract

Stepwise multiple regression analysis was used to determine which demographic variables explained the most variation in numbers of dispersers and rates of dispersal in prairie vole (Microtus ochrogaster) populations. Dispersal was estimated by the removal of all voles appearing on two treatment grids that were live-trapped biweekly. Population density was important in explaining variation in numbers of dispersers of all age classes. Season was important for adult males and females. It was concluded that dispersal was density-independent, in that a constant proportion of the population emigrated at all densities. Different demographic variables were important in explaining the rates of dispersal of each age class. A principal components analysis reduced the original set of demographic variables to four components. The first principal component had high positive loadings for population density variables; PC-II had high negative loadings for survival; PC-III was a reflection of seasonality, with a high positive loading for autumn; PC-IV had a high negative loading for male reproductive activity. There were statistically significant positive correlations between the factor scores of PC-II and rates of dispersal of adult males and females, and a significant negative correlation between PC-II and the rate of dispersal of juvenile males. There were significant positive correlations between PC-IV and rates of dispersal of adult males, adult females and subadult females. Except for juvenile females, number of dispersers of all age classes was correlated positively with the factor scores of PC-I. PCII was correlated positively with the number of dispersers of both subadult males and females, and number of subadult male dispersers was correlated positively with PC-III.