The distribution of traditional breeding dens on the Varanger Peninsula (70-71°N) in northernmost Fennoscandia indicates that this area once harboured a large Arctic fox population. Early 20th century naturalists regarded the coastal tundra of the Fennoscandian Low Arctic to be a stronghold for the species. At the start of our research in 2004, however, the local Arctic fox population was critically small and most neighbouring populations had been extirpated. Here, we synthesize the results of 11 years of research to highlight ecosystem drivers behind the critical state of the Arctic fox in Low-Arctic Fennoscandia. We identify two fundamental drivers: (1) an increasingly climate-driven irregularity of the lemming cycle and (2) a management-and climate-driven increase in the abundance of red fox that is subsidized by more ungulate carrion. Arctic fox reproductive success is low when lemmings are scarce (despite high vole abundance), while red foxes exclude Arctic foxes from high-quality breeding territories in summer and from marine and terrestrial carrion in winter. Red fox culling on Varanger Peninsula may have prevented the extirpation of the Arctic fox population. However, one decade after the onset of this management action the Arctic fox population has failed to increase either because the action has been insufficient or because demographic and environmental stochasticity has precluded a positive response. We discuss options for future research and management of the Arctic fox in the Fennoscandian Low Arctic.

Visually foraging planktivorous fish are prey of visual predators, and their foraging behaviour may be affected by light levels both in terms of gain and risk. The large seasonal change in day length throughout a subarctic summer at 698N was used to show the influence of light on diel vertical migration (DVM) and shoaling patterns in a planktivorous fish assemblage consisting of two species (Coregonus lavaretus and Coregonus albula). Under the midnight sun in June, night and daytime behaviour was similar, with extensive shoaling and limited DVM. With increasingly darker nights towards autumn, the fish dispersed during the dark hours and showed more extensive DVM. Throughout the changing light regime of both the day and the season, the planktivores consistently chose depths with light levels compatible with visual foraging and reduced predation risk as revealed from reactive distance modelling of coregonids and their salmonid predators. The findings support the hypothesis that behavioural decisions are based on a trade-off between foraging rate and predation risk, and increased predator avoidance behaviour towards autumn suggests that this trade-off is state-dependent. Résumé :Les poissons planctonophages qui chassent à vue sont les proies de prédateurs visuels et leur comportement de recherche de nourriture peut être affecté par l'intensité lumineuse, tant en ce qui concerne les gains que les risques. L'important changement saisonnier de la longueur du jour au cours de l'été subarctique à 698N a permis de montrer l'influence de la lumière sur la migration verticale journalière (DVM) et les patrons de formation de bancs dans un peuplement de poissons planctonophages consistant de deux espèces (Coregonus lavaretus et Coregonus albula). Sous le soleil de minuit en juin, les comportements de jour et de nuit sont semblables avec une importante formation de bancs et une DVM ré-duite. Avec les nuits de plus en plus sombres à l'approche de l'automne, les poissons se dispersent durant les heures d'obscurité et font une DVM plus marquée. Pendant tout le changement de régime lumineux, tant journalier que saisonnier, les planctonophages choisissent toujours des intensités lumineuses compatibles avec la recherche visuelle de nourriture et la réduction du risque de prédation; c'est ce que révèle la modélisation de la distance de réaction des corégones et de leurs prédateurs salmonidés. Ces observations appuient l'hypothèse qui veut que les décisions comportementales soient basées sur un compromis entre le taux de recherche de nourriture et le risque de prédation; l'augmentation du comportement d'évitement des prédateurs vers l'automne laisse croire que ce compromis varie en fonction des conditions.[Traduit par la Rédaction]

Snow covers the ground over large parts of the world for a substantial portion of the year. Yet very few methods are available to quantify biotic variables below the snow, with most studies of subnivean ecological processes relying on comparisons of data before and after the snow cover season. We developed a camera trap prototype to quantify subnivean small mammal activity. The trap consists of a camera that is attached facing downward from the ceiling of a box, which is designed to function as a snow-free tunnel. We tested it by placing nine traps with passive infrared sensors in a subarctic habitat where snow cover lasted for about 6 months. The traps were functional for the whole winter, permitting continuous data collection of site-specific presence and temporal activity patterns of all three small mammal species present (the insectivorous common shrew, Sorex araneus, the herbivorous tundra vole, Microtus oeconomus, and the carnivorous stoat, Mustela erminea) as well as abiotic conditions (presence/absence of snow cover and subnivean temperature). Based on their successful functioning (only 6% of the photographs appeared empty or were of poor quality, whereas ca 80% were of small mammals and the remaining of birds and invertebrates), we discuss how the new camera trap can enable subnivean studies of small mammal communities. This greatly increases the temporal resolution and extent of data collection and thereby provides unpreceded opportunities to understand population and food web dynamics in ecosystems with snow cover.