Trophic interactions in population cycles of voles and lemmings: A model-based synthesis

Abstract

Trophic interactions between rodents and their food resources, as well as between rodents and their specialist and generalist predators, are strong candidates for explaining the well-known population cycles of northern voles and lemmings. Here, we link the demography of northern rodents together with their population dynamics by proposing a novel set of stage-structured population models for plants, rodents and predators as well as the interactions between these. Our models are formulated in terms of demographic rates and represented as discrete time-step matrices. They are parameterized with values assumed to be provided by best-available field data. Our models predict, in accordance with earlier studies, multiannual rodent cycles which are driven by feedback interaction with specialist predators (assumed to resemble small mustelids). The cyclic population dynamics change to seasonal density fluctuations when specialist predators become inefficient hunters, or when a rodent population is preyed upon by an abundant generalist predator assemblage (assumed to resemble birds of prey). Furthermore, food resources are found to be important in winter only, and do not by themselves cause population cycles, neither for voles nor for lemmings (the latter of which live on the fairly unproductive tundra habitat). Special attention is given to geographical gradients observed in dynamic properties of rodent fluctuations in Fennoscandia and in Hokkaido, Japan. Model-generated dynamics are analyzed by adopting an autoregressive time series approach, through which possible scenarios are provided for explaining these geographic gradients in the rodent dynamics. Our results are discussed on the basis of recent theoretical and empirical developments within the field of rodent population ecology.