Factors affecting the territory distribution of SkylarksAlauda arvensisbreeding on lowland farmland

Abstract
The distribution of breeding Skylarks on lowland farmland was examined on a monthly basis on 13 farms in southern England in 1996. General linear models identified crop type and field area, shape and boundary characteristics as significant independent predictors of Skylark territory numbers in most or all months. Models predicted territory numbers best during the height of the breeding season and less well during territory establishment in March and abandonment in July. From March to May, crop height had no significant effect on Skylark territory distribution, but in June and July it had a highly significant quadratic effect, models suggesting an optimal vegetation height of around 0.55 m in these months. Set-aside held high territory densities, permanent pasture low densities. Spring cereals held higher territory densities than winter cereals. This could best be explained by differences between the two cereal types in crop structure. There was evidence of a positive correlation between crop height diversity and territory density at the farm scale. The results are discussed with reference to the species' recent severe population decline.
Keywords