Predators drive the lemming cycle in Greenland

A recent study conducted in eastern Greenland and published in the October 31 issue of the Science magazine provides new understanding of the dynamics of arctic lemming populations. Olivier Gilg and Ilkka Hanski from the University of Helsinki, Finland, and Benoît Sittler from the University of Freiburg, Germany, combined long-term field observations and mathematical modelling of what is probably the simplest vertebrate predator-prey community in the world. In the study area in the Karup Valley, at 72o North, there is just one prey species, the collared lemming, which is preyed upon by four species of predators, the stoat, the arctic fox, the snowy owl and the long-tailed skua.

Ecologists have known of the regular cyclic dynamics of arctic lemmings and boreal voles since the publication of a classic study by Charles Elton in 1924. Dozens of hypotheses have been proposed to explain the cyclic rodent dynamics, which are spectacular enough to have drawn the attention of laymen as well as researchers.

For a long time the mechanism behind the rodent cycles was thought to be in the biological attributes of the rodents themselves. It was thought that social interactions among individuals in low-density versus high-density populations would be sufficiently different to select for fast-reproducing docile individuals versus competitive but poorly reproducing individuals. It was proposed that shifting selection pressures would lead to changes in the growth rate of populations and ultimately to cyclic dynamics.

Other researchers have been impressed by the impact that high-density rodent populations may have on their food resources. Lack of food, or lack of high-quality food, was thought to feed back to the breeding performance of the rodents and cause the collapse of high-density populations – after which another cycle could have started. Reasonable as these ideas may appear, empirical studies have provided only limited support to these mechanisms, though ecologists now also recognize that different mechanisms may be in play in different geographical regions to generate what appears to be similar population dynamics.

The study by Gilg, Hanski and Sittler points to the critical role that predators play in shaping up the lemming dynamics. All predators in Greenland but the stoat adjust quickly their numbers in relation to the current numbers of lemmings, and hence the rate of predation by these species is much stronger the higher the lemming density. In fact, the mortality imposed by these predators becomes so high that it alone suffices to stop the lemming population ever reaching densities greater than 10 lemmings per ha. Predation is so intense during the long summer days in Greenland that lemming populations typically decline. Its only during the winter and under thick snow cover that lemmings have a chance to increase in numbers.

Though the impact of the fox, the owl and the skua is spectacular in summer, it is the fourth predator, the stoat, that really sets the pace of lemming dynamics. The stoat is a resident predator, hunting lemmings also under snowbeds in winter. The rate of reproduction of the stoat is however much lower than that of the lemming, and hence the numbers of the stoat lag behind the numbers of the prey. But when the other three predators have stopped the lemming population increase, the stoat will catch up and, according to the predictions of the mathematical model, is the main culprit causing the crash of the lemming population to a very low density. Then follows the collapse of the stoat population itself, and a new cycle is started.

The significance of the new study by Gilg, Hanski and Sittler is in the convincing demonstration of how predator-prey interactions drive the regular 4-year lemming cycle in Greenland. Thus it is not the response of lemmings to their own density, nor interaction with food resources, which is of importance in Greenland, but it is the interaction with the stoat that drives the cycle, while the other predators are responsible of setting an upper limit to lemming population density.

Though these results from an arctic region cannot be straightforwardly applied to all other regions in the world where small rodents exhibit cyclic dynamics, it is clear that predation is now the top contender for explaining the decade-old puzzle of lemming and vole cycles.

See also: 31 October issue of Science, Paper No. 22, Cyclic Dynamics in a Simple Vertebrate Predator-Prey Community.

For more information, please, contact Professor Ilkka Hanski, University of Helsinki, tel: +358 9 191 57745, e-mail: ilkka.hanski@helsinki.fi