Abstract
1,324 vanishing bearings of individual pigeons, completely inexperienced in homing, were recorded during 100 releases at 36 sites symmetrically distributed around their home loft near Würzburg at distances ranging from 7 to 180 km. Two directional components could be derived from the data: (a) a weak but significant ‘homeward component’ pointing to the loft site and (b) a more pronounced ‘compass component’ pointing to the northwest (Fig. 1). With the latter kind of tendency the pigeons clearly demonstrated what earlier had been described as ‘preferred compass direction’ (PCD). Homeward directedness at the 36 differently situated release sites is negatively correlated with angular divergence between PCD and homeward direction (Fig. 2). Compass preferences derived from different quartets of symmetrically arranged sites were similar. Over a period of 11 years the PCD varied significantly in time but did not change fundamentally (Figs. 3, 4). Interrelations between various parameters of initial-orientation data were investigated by correlation analyses (Table 3). PCDs have not only been found in the Würzburg area but around many other loft sites as well. They are, however, differently pronounced, and in experienced pigeons they are generally weaker than in first-flight birds. Various reasons (partly evident, partly thought possible) for variability of PCDs, and for their absence in some series of experiments, are discussed. As a PCD persists, and even tends to become clearer, after elimination of homeward orientation by olfactory deprivation, it is concluded that it reflects directional tendencies which are independent of the process of site localization. A great deal of what is called ‘release-site biases’, i.e., site-specific deviations from the beeline course towards home, can be understood on a PCD basis. The functional background of the PCD is unknown. Hypothetical advantages for the pigeons’ homing strategy are considered.