Synchrony and Scaling in Dynamics of Voles and Mice in Northern Japan

One hundred and seventy-six time series of the Japanese wood mouse (Apodemus speciosus) and 185 time series of the grey-sided vole (Clethrionomys rufocanus) spanning 31 years (1962–1992) were studied with respect to synchrony and spatial correlation in population dynamics. The time series were collected at fixed sites as part of the rodent census program of the Japanese Forestry Agency. The survey locations cover a region of 115 by 270 km in northern Hokkaido, Japan. The average correlation between the time series was 0.34 (ci95% = 0.31, 0.36) for the grey-sided vole and 0.16 (ci95% = 0.13, 0.18) for the Japanese wood mouse. This average correlation was decomposed spatially using spatial autocorrelation techniques and the nonparametric covariance function. The lower region-wide correlation in the latter species was found to be due to the spatial covariance dropping more rapidly with distance. The spatial scale of the dynamics was measured by the L0 correlation length (the distance at which the covariance is equal to that between two randomly chosen sites within the region). This distance was estimated to be ∼50 km for the grey-sided vole and 20–30 km for the Japanese wood mouse. The L0 correlation length was linked to the scaling of the underlying ecological processes through a simple epiphenomenological model motivated by diffusion theory (the exponential covariance model). A review of the ecology of the rodent species indicates that the spatial scale of the pattern of fluctuation of the grey-sided vole is more extensive than can be accounted for by a process of dispersal of voles. Predator movement and regulation of the vole by predators are possible causes of the pattern of spatial covariance. The Japanese wood mouse has a scale of dynamics slightly larger than what may be accounted for by the movement of individuals. This species constitutes a negligible part of predators’ diet. Food resource dynamics may be an important regulatory factor and a source of spatial covariance. We interpret the results as indicating that the natural scales of regulation are species specific and can be large relative to the scale of local populations.