Data on 12 factors presumed to influence the distribution of aggressive nest defence in I 1 1 species of waders (incubation-sharing by the parents, number of parents present near the nest, incubation time, nest habitat, breeding latitude, body mass, wing loading, wing structure, detectability on the nest, predator regime, coloniality and alternative prey) were collected from literature and field researchers. Body mass and number of parents present on the nest temtory (within response range when avian predators appear) explain 50% of the variation in aggressive defence behaviour. The results support the notion that ecological conditions l i e predation pressure are important in shaping wader parental care systems, with implications for mating systems. Altogether, the investigated factors explain around 70% of the variation in the samples. Future research on the level of individuals is suggested in order to explain the remaining variation.

White-throated Sparrow, Zonotrichia albicol-WINGRELD, J. C. 1985. Short-term changes in plasma lis (Gmelin). Evolution 29:6 11-621. levels of hormones during establishment and de-THORPE, W. H. 196 1. Bird song. Cambridge Univ. fense of a breeding territory in male Song Spar-Press, London.

Communal roosts are an important aspect of Turkey Vulture (Cathartes aura) biology, but remain inadequately studied. We observed the use patterns of Turkey Vultures at a communal roost in northeastern Iowa from their arrival on 17 March 2005 to their departure on 19 October 2005. The roost was on a forested hillside and vultures roosted below the forest canopy in 10 live deciduous trees, spending an estimated 10–16 h per day in the roost. The birds also used nearby pre‐ and postroost perch sites, and formed pre‐ and postroost kettles near the roost site. The number of vultures using the roost ranged from about 20 early in the season to a peak of 281 on 4 October. The apparent pattern was one of moderate and fluctuating numbers early in the 7‐month season, increasing numbers in the middle of the season, and high and fluctuating numbers late in the season. Hatching‐year vultures began to appear at the roost during the last week of August. The mean monthly time of departure from the roost ranged from 3 min before sunrise to 131 min after sunrise. Vultures departed the roost significantly later in the morning during the summer (June, July, and August) than during other months (April, May, September, and October), probably due to differences in soaring conditions (longer thermal generation times during the summer) and available foraging time (longer days in the summer).