Summary
Olfactory navigation by birds
In contrast to earlier navigation hypotheses, based as they are on theoretical constructs deduced from our knowledge of the physical world, the notion of olfactory navigation is an unexpected outcome of empirical research. Referring to sceptical articles on the issue in this journal and elsewhere (e. g. Schmidt‐Koenig 1985, 1987, 2001, Wiltschko 1996), and in order to fill a gap in a recent review on avian navigation (Wiltschko & Wiltschko 1999), I describe the most instructive experiments providing evidence that birds are able to home by utilizing atmospheric trace gases perceived by the sense of smell. (1) When released in an unfamiliar distant area, homing pigeons with bisected olfactory nerves fly considerable distances, but fail to approach the home site (Fig. 1, 2, 3). Largely analogous treatments in control birds and experimentals make it extremely unlikely that the failures are due to non‐olfactory side‐effects. (2) Elimination of trace gases from the inhaled air by means of charcoal filters prior to release, combined with nasal anaesthesia upon release, prevents initial homeward orientation, whereas nasal anaesthesia alone (after smelling of natural release‐site air) does not (Fig. 5). (3) Pigeons exposed to natural air at one site and released, without access to natural air, at a quite different site, fly in a direction corresponding to homeward from the site of exposure, but not from the current actual position (Fig. 6). (4) Long‐term screening from winds in an aviary at home prevents subsequent homeward orientation from distant sites. Deflecting or reversing winds in a home aviary results in accordingly deflected or reversed orientation (Fig. 7). (5) From areas made familiar by previous flights homing is possible also on a non‐olfactory basis. This can be explained in terms of the utilisation of visual landscape features.
In as far as related experiments were conducted using reliable methods, the results are unequivocal. On the whole, they can be understood only provided that the birds are able to deduce their position relative to the home site from atmospheric trace gases, and that this ability requires previous opportunity to correlate current wind conditions with simultaneous olfactory conditions at the home site over a lengthy period of time. As an attempt to explain the underlying system, a working hypothesis is presented which postulates that (a) long‐range gradients exist in the ratios among several airborne trace substances and that (b) their directions can be derived, at the home site, from changes of ratios in dependence on wind direction. Atmospheric hydrocarbons investigated by means of gas chromatography in an area covering 400 km in diameter did in fact include such postulated ratio gradients (Fig. 8). Their directions were fairly stable even under varying conditions of weather and winds. Correlations among gradient directions and changes of ratios according to wind directions were also found, but the long‐term angular relationships have not yet ...