RAPID OR DELAYED TRACKING OF MULTIANNUAL VOLE CYCLES BY AVIAN PREDATORS

1. The relationship between the yearly densities of avian predators and their microtine prey in western Finland was studied. Predator densities were determined by checking nest-boxes in forests [Tengmalm's owl (Aegolius funereus), hawk owl (Surnia ulula) and Ural owl (Strix uralensis)] and by searching for nests in farmland [short-eared owl (Asio flammeus), long-eared owl (Asio otus) and kestrel (Falco tinnunculus)]. Tengmalm's, Ural and hawk owls overwinter in Fennoscandia, long-eared owls are partially migratory, and short-eared owls and kestrels are migratory. 2. Prey densities were estimated by snap-trapping in spring (early May) and autumn (mid-September), and by snow-tracking in late February to early March (early spring) and late November to early December (late autumn). 3. The breeding densities of Tengmalm's owls in two study areas were significantly correlated with trap indices of voles in the prevailing spring and in the preceding autumn (6-month lag), but not in the preceding spring (1-year lag). Tengmalm's owl breeding densities in one study area covaried with track indices of voles in the early spring (i.e. the settling period of owls), but not with those in the late autumn. 4. The yearly breeding densities of Ural owls and hawk owls, and the wintering densities of hawk owls were positively related to spring trap indices of voles, but not to indices in the previous autumn and spring. 5. The breeding densities of long-eared owls and short-eared owls were dependent on vole abundances both in the current spring and the preceding autumn, but not in the preceding spring. Kestrel breeding densities fluctuated in accordance with spring vole abundances. 6. The densities of most avian predators tracked rapidly, without obvious time lags, vole abundances at the time the birds of prey settled on their territories. This rapid tracking is mostly based on vole-supply dependent immigration and emigration. Densities of most avian predators did not lag 9 months behind prey densities which, in theory, may drive 3-4-year vole cycles (May 1981).