Population cycles in voles and lemmings: density dependence and phase dependence in a stochastic world

I review the regular multiannual population fluctuations in voles and lemmings of northern latitudes. Periodically fluctuating small rodent populations all exhibit a clear two-dimensional density-dependent structure. This implies both a direct and a delayed annual density dependence, and suggests that either a predator-prey type of interaction or a specialised plant-herbivore type of interaction (but not both) may be the underlying cause of these multiannual density cycles. Clear-cut experimental testing relating to these propositions is, however, lacking. A two-dimensional annual density-dependent structure is typically non-linear in a way which may be modelled as a threshold type of non-linearity and interpreted as a phase dependence in the density-dependent structure (implying that the density-dependent structure is different in the increase and the decrease phase). Two clear geographic gradients in the annual density-dependent structure are reviewed: the Fennoscandian gradient and the Hokkaidian gradient. The seasonal nature of the annual density-dependent structure is furthermore reviewed: for voles in both Fennoscandia and Hokkaido the direct annual density dependence during the winter (measured per time unit) is concluded to be the strongest. I close with a survey of the main challenges within the field of 'small rodent cycles' - the greatest of which is suggested to be the integration of demography and population dynamics. I recommend to look at other populations for potentially applicable model systems. I conclude that multiannual population cycles seen in voles and lemmings will continue to be a strong source of conceptual and methodological developments within the field of population ecology.