The Sensory Basis of Bird Navigation

F EW natural phenomena have excited more wonder, guesswork, and debate than the long distance flights of migratory birds. Many species travel thousands of miles from breeding grounds to winter range, often crossing hundreds of miles of open ocean. Some young birds apparently make their first migration without adults of the same species to guide them. The central problem is one of navigation: how does the bird know where to fly? Some receptor or receptors must be excited by an aspect of the bird's environment which is related to the direction of its goal. Visually perceived landmarks constitute one such aspect of the environment, but there are many migrations in which some other type of orientation must be involved. For instance, some birds cross large bodies of water in cloudy weather when no landmarks are visible, while in other species the young migrate before they have had an opportunity to learn anything about the route. It seems, at first glance, that those aspects of the environment which are functions of the geographical relationship between the bird and its goal are outside the sensitivity range of its receptors. Are unsuspected geographical cues available to birds, or is their sensory spectrum more extensive than we know? No completely satisfactory answer to this question can yet be given, for direct studies of migration have seldom passed beyond the descriptive stage. Valikangas (1933) has shown that ducks raised in Finland from English eggs migrated south in their first autumn with Finnish birds and that several returned the next spring to their "foster homes" in Finland. This experiment indicates a lack of any inheritance committing the bird to a specific route. Thienemann (1930), Skovgaard (1929), and Schenk (1909) reared young storks in captivity and released them in the autumn after most wild storks had departed. Several of the young birds were recovered on or not far from the species' normal migration route. Rowan (1931) reported similar results with crows, although Drost (1939) found that when young sparrowhawks were treated in this way, they generally remained near the point of release. These experiments confirm the ornithologists' impression that young birds of many species can make their first migratory flight without the guidance of experienced adults. Rowan (1932, 1938a, 1938b), Schildmacher (1933, 1934), Wolfson (1941), and others have studied the effects of endocrine secretions on the time of migration They have learned much about the internal stimulus necessary to initiate migration, but such studies have not thrown much light on the sensory processes by which birds guide themselves on long migratory flights. Most of our knowledge of the sensory basis of bird navigation has been obtained from homing experiments; for in these experiments it is possible to control, to some extent, the conditions under which the bird is to make its long distance flight. Consequently this review will be concerned almost entirely with such work. The term "homing" has been used to cover (1) return of birds shipped in captivity to a distant release point, and (2) a natural migratory flight, often hundreds of miles long, in which birds leave one region and then return to it the following season. Austin (1940) and Stoner (1941), for example, have used the term in the latter sense, while Thompson (1926) and others have made a sharp distinction between what might be called "natural homing" and the "artificial" homing experiments, such as those described below, in which birds are captured, shipped to a distance and released. Among the "artificial" homing experiments should be included the remarkable feats of the domestic and non-migratory racing pigeon. It seems reasonable to assume with WVarner (1931) that the same or similar means of navigation are used in both types of homing, and furthermore that if we knew how a bird finds its way home when shipped 1000 miles from its nest, we should be in a better position to explain annual migratory flights of similar extent. Many other animals have migrations comparable