USGS Home
Contact USGS
Search USGS

Birds are not alone in their ability to navigate long distances. Fish, mammals, and even insects make migratory journeys. But the clarion honking of geese moving in huge skeins across the vault of the heavens, the twittering of migrants filtering down out of the night sky, the flocks of newly arrived birds filling woodlands, fields, and mudflats makes us most aware of the seasonal movements of birds and fills us with awe and wonder as to how such a magnificent event can be accomplished season after season, year after year, with such unerring precision.

Of the three kinds of information necessary for navigation, we know something about the environmental cues that birds use to orient their migratory flight in the proper direction. On the other hand, there also is well-supported experimental evidence that birds use neither the positions of the sun or the stars to know where they are or where they are to go. It has been shown, however, that birds must learn both the location of the wintering area as well as the location of the breeding area in order to navigate properly, but we have no idea what information they are learning. Nor do we know what cues birds use to know the location of their migratory destination when they are in their wintering locale, often thousands of miles away. The recapture of banded birds at the same places along the route of the migratory journey in subsequent years suggests that some species also learn the location of traditional stop-over sites, but how they do that remains a mystery.

Vector Navigation

Displacement studies in the Western Hemisphere using several species of buntings also demonstrated that birds recognized they had been moved and could fly appropriate, yet unique, routes to return to their normal range. Yet adult Hooded Crows transported latitudinally by over 600 km from wintering grounds in the eastern Baltic to northwestern Germany failed to recognize this displacement. In the spring they oriented properly but migrated to Sweden, west of their normal breeding range. This species used vector navigation, but did not know the location of its traditional destination. Since it is generally accepted that migratory behavior evolved independently again and again in different bird populations, a single explanation to fit all cases perhaps should not be expected.

Orientation Cues

Landmarks are useful as a primary navigation reference only if the bird has been there before. For cranes, swans, and geese that migrate in family groups, young of the year could learn the geographic map for their migratory journey from their parents. But most birds do not migrate in family flocks, and on their initial flight south to the wintering range or back north in the spring must use other cues. Yet birds are aware of the landscape over which they are crossing and appear to use landmarks for orientation purposes. Radar images of migrating birds subject to a strong crosswind were seen to drift off course, except for flocks migrating parallel to a major river. These birds used the river as a reference to shift their orientation and correct for drift in order to maintain the proper ground track. That major geographic features like Point Pelee jutting into Lake Erie or Cape May at the tip of New Jersey are meccas for bird-watchers only reflects the fact that migrating birds recognize these peninsulas during their migration. Migrating hawks seeking updrafts along the north shore of Lake Superior or the ridges of the Appalachians must pay attention to the terrain below them in order to take advantage of the energetic savings afforded by these topographic structures.

Since humans learned to use celestial cues, it was only natural that studies were undertaken to demonstrate that birds could use them as well. Soon after the end of the Second World War, Gustav Kramer showed that migratory European Starlings oriented to the azimuth of the sun when he used mirrors to shift the sun's image by ninety degrees in the laboratory and obtained a corresponding shift in the birds' orientation. Furthermore, since the birds would maintain a constant direction even though the sun traversed from east to west during the day, the compensation for this movement demonstrated that the birds were keeping time. They knew what orientation to the sun was appropriate at 9 a.m. They knew what different angle was appropriate at noon, and again at 4 p.m. It has been recently shown that melatonin secretions from the light-sensitive pineal gland on the top of the bird's brain are involved in this response. Not only starlings but homing pigeons, penguins, waterfowl, and many species of perching birds have been shown to use solar orientation. Even nocturnal migrants take directional information from the sun. European Robins and Savannah Sparrows that were prevented from seeing the setting sun did not orient under the stars as well as birds that were allowed to see the sun set. Birds can detect polarized light from sunlight's penetration through the atmosphere, and it has been hypothesized that the pattern of polarized light in the evening sky is the primary cue that provides a reference for their orientation.

Using the artificial night sky provided by planetariums demonstrated that nocturnal migrants respond to star patterns. (quite analogous to Kramer's work on solar orientation, Franz Sauer demonstrated that if the planetarium sky is shifted, the birds make a corresponding shift in their orientation azimuth. Steve Emlen was able to show that the orientation was not dependent upon a single star, like Polaris, but to the general sky pattern. As he would turn off more and more stars so that they were no longer being projected in the planetarium, the bird's orientation became poorer and poorer. While the proper direction for orientation at a given time is probably innate, Emlen was able to show that knowing the location of "north" must be learned. When young birds were raised under a planetarium sky in which Betelgeuse, a star in Orion of the southern sky, was projected to the celestial north pole, the birds oriented as if Betelgeuse was "north" when they were later placed under the normally orientated night sky, even though in reality it was south!

Radar studies have shown that birds do migrate above cloud decks where landmarks are not visible, under overcast skies where celestial cues are not visible, and even within cloud layers where neither set of cues is available. The nomadic horsemen of the steppes of Asia used the response of lodestones to the Earth's magnetic field to find their way, and the hypothesis that migrating birds might do the same was suggested as early as the middle of the nineteenth century. Yet it was not until the mid-twentieth century that Merkel and Wiltschko demonstrated in a laboratory environment devoid of any other cues that European Robins would change their orientation in response to shifts in an artificial magnetic field that was as weak as the Earth's natural field. Although iron-containing magnetite crystals are associated with the nervous system in homing pigeons, Northern Bobwhite, and several species of perching birds, it is unknown whether they are associated with the sensory receptor for the geomagnetic cue. An alternate hypothesis for the sensory receptor suggests that response of visual pigments in the eye to electromagnetic energy is the basis for geomagnetic orientation. It has been shown, however, that previous exposure to celestial orientation cues enhances the ability of a bird to respond more appropriately when only geomagnetic cues are available.

Radar observations indicate that birds will decrease their air speed when their ground speed is augmented by a strong tail wind. We also know that birds can sense wind direction as gusts ruffling the feathers stimulate sensory receptors located in the skin around the base of the feather. Since there are characteristic patterns of wind circulation around high and low pressure centers at the altitude most birds migrate, it has been hypothesized that birds could use these prevailing wind directions as an orientation cue. However, there presently is no experimental support for this hypothesis.

The sense of smell in birds was considered for a long time to be poorly developed, but more recent evidence suggests that some species can discriminate odors quite well. If the olfactory nerves of homing pigeons are cut, the birds do not return to their home loft as well as birds whose olfactory nerves were left intact. A similar experiment has demonstrated that European Starlings with severed olfactory nerves returned less often than unaffected control birds even at distances as great as 240 km from their home roosts. And even more interesting, when these starlings returned to the nesting area the following spring, the starlings with nonfunctioning olfactory nerves returned at a significantly lower frequency than the other starlings.

Considering the array of demonstrated and suggested cues that birds might use in their orientation, it is clear that they rely upon a suite of cues rather than a single cue. For a migrating bird this redundancy is critical, since not all sources of orientation information are equally available at a given time, nor are all sources of information equally useful in a given situation.