Habitat selection in forest owls: The roles of vegetation structure, prey density and competitors

Abstract

The study was aimed to the seasonal distribution of calling males of the Pygmy Owl (Glaucidium passerinum) and the Tawny Owl (Strix aluco) depending on the structure of forest habitats, the abundance of small mammals and the presence of other owls. Moreover, the ratio of the significance of these three predictors depending on the season was determined. The materials were collected annually from 2001 to 2011 in the southwest of Moscow Region (55.4594 N, 37.1795 E). Owls counts and captures of small mammals were carried out twice a year — in spring and autumn. To characterize forest habitats, we used forest inventory materials, own data of geobotanical descriptions and other sources. The model plot was divided into 105 squares with a side of 200 meters. Thus, we determined the presence of owls in different seasons, small mammals’ abundance and habitats characteristics for each square. We used a machine learning technique (Boosted Tree Classifier) for estimation of different predictors’ influence on the seasonal distribution of calling owls. We created four models with different dependent variables: the annual occupation of the selected squares by Pygmy Owl in spring (1) and autumn (2); the annual occupation of the selected squares by Tawny Owl in spring (3) and autumn (4). The independent variables for each model were: 20 parameters describing the forest structure; the abundance of small mammals at each site in a certain season (15 parameters); presence in the square or next to it of one’s own or another species of owls as well as both species in the previous season (6 parameters). The distribution and maximum number of calling males of owls in spring in local populations was determined primarily by the presence of the most favorable sites in suitable forest habitats. The total contribution to the final model of the distribution of variables associated with various parameters of forest habitats was more than 55% for both species. At the same time, among them, the upper tree layer characteristics were the most significant for both species. The widespread suggestion that the main factor influencing on territory occupation by birds of prey is high main prey abundance was not confirmed by our study. Apparently, owls are guided primarily by certain parameters of habitats and then by the favorable food supply, and all this is corrected by the presence of other myophagous predators.