Comparative population dynamics of small and large mammals in the Northern Hemisphere

We compared the patterns of population dynamics between small and large mammals using time series models and principal component analysis (PCA).  The Kalman filter was fit to 23 time series estimates of large mammal abundance and 39 time series estimates of small mammal abundance from the northern temperate region to estimate the parameters describing the strengths of density dependence and environmental stochasticity.  We also calculated long-term mean and variance of population growth rate (pgr) for each population.  Eight different parameters were used in PCA to characterize and classify the patterns of population dynamics.  Small and large mammal populations were separated into two groups in the space defined by the first two principal components, and the variances of pgr and environmental stochasticity were the main factors differentiating the two groups.  Long-term means of pgr did not differ between small and large mammals; however, the variances of pgr and environmental stochasticity and strengths of direct and indirect density dependence of small mammals were greater than those of large mammals.  The strength of direct density dependence was positively related to the strength of environmental stochasticity in large mammals.  Therefore, small mammal populations appeared to be more variable and experience stronger intrinsic regulation.  Demographic, mechanistic approaches are needed to elucidate different population dynamic patterns between small and large mammals.