1. The densities of common cranes Grus grus in 10 zones of a wintering area were approximately proportional to the amounts of food resources, but some overuse of the zones with highest food densities was observed, i.e. greater numbers of birds than expected used these zones. The distribution resembled ideal free distribution only after numbers of cranes had exceeded carrying capacity. The seasonal pattern of settlement deviated from ideal free distribution. During the early and late part of the season, when there were fewer birds at the study site, cranes preferred to forage as close to the roost as possible provided that there was enough food, instead of selecting the zones further away with highest food densities. 2. However, 12 individually marked cranes differed in their competitive ability and foraging area selection. Larger adult birds were dominant in aggressive encounters, displacing subdominant cranes from good feeding positions. Food intake rate of dominant cranes tended to be higher than the hock average, the difference increasing with rank. Dominant cranes preferred to forage in the zones with highest food densities and had higher absolute daily food intakes. The relative pay-offs of different phenotypes changed across zones with different food densities: subordinate birds could not increase their intake rate at the richest zones as much as dominants. 3. The average daily food intake of a crane was thus positively correlated with both the quality of the foraging zone and the dominance rank of the bird. These results fulfil most assumptions and predictions of the interference phenotype-limited distribution model, although truncation of phenotypes between zones was imperfect.
