MANEN W VAN (2006) Use of winter-roosts, breeding numbers and reproduction. LIMOSA 79 (2): 53-62.

Long-eared Owls Asio otus were studied in the northern Netherlands in 1986-2001, to evaluate how diet composition affects the use of winter roosts, numbers of breeding pairs, and reproduction. The study area (2500 ha) is dominated by farmland (both arable and grassland), complemented with dry heathland, mixed woodlots, some small villages and the edge of a town. The area offers almost unlimited roosting and nesting possibilities and foraging habitat is diverse.
      Diet composition was studied by collecting pellets at winter roosts or in breeding territories. Owls were counted on winter-roosts every ten days. From 1986-1996 presumably all roosts were known. Territories were mapped by visiting potential breeding sites in the early evening, six times in January-April to detect displaying birds, and three times in May-July for begging young. Nests were searched for in suitable breeding habitat.
      Throughout the study period Short-tailed Voles Microtus arvalis were the most important prey species (69%; Fig. 1). Other prey included Wood Mouse Apodemus sylvaticus (14%), Common Vole Microtus agrestis (5.5%), Bank Vole Clethrionomys glareolus (3.5%), birds (3.0%) and Harvest Mouse Micromys minutus (2.8%). Other species were represented in the diet by less than 1%. The proportion of voles in the diet was highest in late summer, declined in autumn and winter, and was lowest in spring (Fig. 2).
      The number of winter roosts was nearly constant, but the number of owls per roost varied between years. The number of wintering owls depended on the proportion of Short-tailed Voles in their winter diet (October-December; Fig. 3). The use of winter-roosts was described by the period during which more than 50% of the maximum number of owls was present at the roost. There was a tendency that roosts were established later and deserted earlier in years with higher proportions of Short-tailed Voles in their winter diet. When the owls spent fewer days at the communal roosts, and left roosts earlier, more breeding territories were occupied (Fig. 4) and more young produced in the following breeding season.
      The number of breeding territories was related to the percentage of Short-tailed Voles in the diet in January-March (Fig. 5). There was also a correlation between the average number of young fledged per pair in the preceding breeding season and the number of breeding pairs in the subsequent spring, indicating that local offspring tend to remain in the study area. No correlations were found between breeding parameters and the proportion of voles in the summer diet. Percentage of successful pairs and number of fledglings per brood were positively correlated with the sum of the proportion of voles and Wood Mice in the summer diet (Fig. 6). Nevertheless, the exact role of prey species other than Short-tailed Vole during the breeding season remains unclear.
      In three years the numbers of owls increased considerably outside the breeding season, indicating immigration (Fig. 7). This only occurred in autumns and winters preceding highly productive breeding seasons (Fig. 8). In most years, numbers decreased during autumn and winter (by up to 85%), due to mortality but possibly also due to emigration. The Netherlands are situated near the western edge of the area in which individual Longeared Owls track fluctuations in Microtus abundance. I argue that communal roosting of owls is linked to migratory populations, and that the ultimate function is related to the possibility to find a breeding partner.

[pdf only for members] [dutch summary]