Karupelv Valley Project - A long term study on lemming cycles in the North East Greenland National Park

Initiated in 1988, this project primarily aims at a monitoring of a lemming population in a simplified high arctic ecosystem. Such long term data are regarded as a prerequisite to any realistic approach to understanding microtine cycles.

The study site encompasses the coastal lowlands of Karupelv Valley on Traill Island in North East Greenland. It is located at approximately 72°30' north and 24° west, and is part of the North East Greenland National Park that provides freedom from human disturbance. Vegetation belongs to the northern tundra and is discontinuous, with great parts of the area that are barren.

The selection of this area was mainly governed by the fact that it supports one of the most simple communities in existence in which lemmings occur, collared lemming (Dicrostonyx groenlandicus) being the only microtine present in this part of Greenland. It is the staple prey of a very limited number of predators. These include stoats and arctic foxes among the mammals present year-round and snowy owls as well as long tailed skuas that are present only on a seasonal basis.

As in any long term study, routine surveys had to be designed simply, with samplings easy to operate. The basic approach adopted is a systematic recording of winter nests that are easy to detect and monitor following snow melt. By applying this systematic search on a greater area (1000 ha), a relative merit of this method is that it provides also more data during so called lemming lows, when other methods like trapping fail to produce enough observations. In addition, one may trace the use of the nests for subnivean reproduction (pellets of juveniles) as well as whether it was taken over by a stoat (lining with fur, scats of stoats).
Field work done on a yearly basis is concentrated on the snow free period. Besides these winter nest surveys, additional data collected consist in the record of patterns displayed by predators. These includes mapping of breeding among avian predators (snowy owls, long tailed skuas) as well as the checking of occupancy of fox dens. Breeding of migratory birds (barnacle geese, waders, etc.) is also assessed on a yearly basis.

Some follow up studies that specifically document patterns in summer time are also conducted in this study site (for example study on Arctic foxes). These investigations often involve close co-operation with other institutions, such as Olivier Gilg's modelling approach performed as PhD project under the auspices of the Universities of Helsinki and Montpellier).

This long term monitoring of lemming abundance in Karupelv Valley has evidenced the reality of the cycling fluctuations in North East Greenland, with up to 30-fold differences in absolute number of winter nests between outbreaks and deep depressions (nearly 4000 nests in peak phases against only around 100 in depressions). The observations of now 3 cycles suggest that the length of the cycle may be 4 or 5 years, but several patterns may depart from cycle to cycle. Evidence is also provided that population outbreaks result from intense subnivean reproduction. Observations on predation rates by stoats on winter nests clearly suggest a delayed reaction by more than one year on the outbreaks of the lemmings. Snowy owls were breeding in only 5 out of 15 years. Reproduction success among long tailed skuas was also closely related to the abundance of lemmings.

As similar approaches were also adopted in a parallel monitoring carried out since 1995 by Thomas Berg (Danish Polar Center) in Zackenberg station some 250 km north of Karupelv Valley, these studies will also contribute assessing degrees of synchrony on a regional scale.

Results of this study have been published in the Science Magazine in 2003.