Taylor’s law (TL) asserts that the variance in a population density is a power-law function of its mean: log(variance) = a + b log(mean). The slope b lies between 1 and 2 in most observed TLs, and an underlying mechanism determining slope b is one of major questions about TL. I showed in a previous study that densities of the Hokkaido vole satisfied temporal and spatial forms of TL, and demonstrated that time series generated by the Gompertz model reproduced the form of temporal and spatial TLs, but with slopes that were significantly steeper than the slopes estimated from data (Cohen and Saitoh 2016, Ecology 97: 3402-3413). In this paper, I analyzed effects of interpopulation dispersal and spatial synchrony of population dynamics on the slope b using the Gompertz model. When each population was independent, the spatial and the temporal slopes were higher than 2. When interpopulation dispersals were introduced, the temporal slopes were lowered to the interval between 1 and 2. However, the spatial synchrony was required to lower the spatial slopes.