Contribution of olfactory navigation and non-olfactory pilotage to pigeon homing

Summary By means of training flights (mostly flock releases), two groups of young homing pigeons were made familar with a larger area asymmetrically extending from the loft toward SW and NE, respectively. Thereafter, birds of both groups were released within each of the training areas with which one sort of birds was familiar (F⁺), the other unfamiliar (F^-). Even the F⁺ pigeons had never been released at the test site itself (nearest previous release 10 km apart). Half of each group was allowed to smell environmental odors (O⁺); the other half breathed charcoal-filtered air during transportation and at the release site until a few minutes before release when they were deprived of olfaction by intranasal application of Xylocain (O^-). The two test sites were 53 km distant from home. There was little difference in initial orientation as well as in homing performance between pigeons that were allowed to smell natural air and were familiar with the area (F⁺O⁺) and those that were privileged in only one respect (F⁺O^- and F^-O⁺). Yet if none of the preconditions was met (F^-O^-), performances were drastically reduced. The findings show that pigeons make use of two independent homing methods, olfactory navigation (presumably based on a navigational map) and non-olfactory pilotage (presumably based on a topographical map). The latter method is restricted to a more or less familiar area determined by individual experience. Its boundaries are poorly defined and can be estimated by the experimenter only in rough approximation. Within this area, the homing system takes advantage of more or less redundant inputs. Outside of it, olfactory information seems indispensable.