How Birds Find Their Way: The Strange Sensory World of Avian Navigation

In October 2022, a bar-tailed godwit fitted with a satellite tag flew 13,560 kilometers from Alaska to Tasmania in eleven days, without stopping, without eating, and without sleep in any normal sense. The bird arrived within a few kilometers of where its species winters every year. This was the longest non-stop flight ever recorded for any animal. It is also one of the more ordinary feats of avian navigation, which routinely accomplishes things that would be impressive even with GPS.

How birds navigate has been one of the great unsolved problems in biology for two centuries, and what makes it interesting is that the answer is not one mechanism but at least four, used together, with different birds emphasizing different combinations and switching between them depending on conditions. The picture that has emerged is that a migrating bird is using something like multi-sensor fusion, but with senses we do not have and only partly understand.

The sun compass

The simplest navigation aid, used by virtually all migrating birds, is the sun. Birds can determine compass direction from the sun's position, corrected for the time of day. Gustav Kramer demonstrated this in 1949 with caged starlings: when the time of day was artificially shifted using indoor lighting, the birds' attempted migration direction shifted by the corresponding compass angle. They were using the sun, and they had a remarkably accurate internal clock.

The clock matters. The sun's azimuth changes by roughly fifteen degrees per hour, which means a one-hour clock error produces a fifteen-degree heading error. Over thousands of kilometers of flight, this would be ruinous. Birds maintain their internal time-keeping accurately enough to use the sun across long flights, and the clock recalibrates against the day-night cycle continuously.

The star map

For night-flying birds, including most songbirds, the relevant reference is the night sky. Stephen Emlen demonstrated in the 1960s using indigo buntings in planetarium experiments that birds navigate from a learned map of the rotational pattern of the stars around the celestial pole. The buntings did not memorize specific constellations; they learned the center of rotation. When Emlen rotated the planetarium sky around a different center, the birds reoriented to that fictitious pole.

The learning happens in the nest. Buntings raised in a planetarium with the stars rotating around Betelgeuse instead of Polaris, when later released and tested, navigated relative to Betelgeuse. The map is innate in form (find the rotational center) but populated by experience. The implication is that a bird raised on a different planet, with different stars, would learn that planet's celestial pole.

The magnetic compass and the strange physics of cryptochromes

The most unusual part of avian navigation is the magnetic sense. Birds detect Earth's magnetic field, which is a useful feat in itself, but the mechanism is genuinely weird: it appears to involve quantum chemistry happening in the bird's eye, in molecules called cryptochromes.

The proposed mechanism, which has accumulated significant evidence over the last two decades, is the radical pair model. When a photon hits a cryptochrome molecule in the retina, it knocks an electron loose, creating a pair of radicals (molecules with unpaired electrons). The spins of those radicals are quantum-mechanically entangled, and the entanglement is sensitive to the local magnetic field. As the bird turns its head, the magnetic field's angle relative to the molecule changes, and so does the chemistry of the radical pair. The bird, in other words, sees the magnetic field as a faint visual pattern overlaid on the world.

This mechanism predicts specific things, and the predictions hold up. The bird's magnetic sense should require light, which it does. It should be sensitive to weak radio-frequency interference at specific frequencies that disrupt the radical pair's spin coherence, which it is. It should work as an inclination compass (sensing the angle of field lines) rather than a polarity compass (sensing magnetic north vs south), which it does. And it should be located in the eye, specifically in cells that contain cryptochrome, which it is.

The European robin, the model organism for this work, loses its magnetic compass when its right eye is covered but not when its left eye is covered. The directional information is processed asymmetrically in the brain. We do not fully understand why.

The infrasound map

Beyond the compass senses, birds appear to use a "map" of acoustic landmarks at frequencies humans cannot hear. Ocean waves striking continental shelves generate infrasound at around 0.1 Hz with characteristic spectra that vary by coastline. Mountains, weather fronts, and volcanic activity produce signature infrasound patterns at distances of hundreds or thousands of kilometers. Pigeons disoriented during a 1969 race across the Susquehanna basin, for example, were eventually correlated with a sonic boom from a Concorde test flight that had disturbed the local infrasound landscape.

The sense allows birds to identify their location by a kind of acoustic fingerprint: even far inland, the rumble of a particular ocean coastline arriving from a particular direction is distinctive enough to localize. Whether this is a primary sense or a secondary check on other senses is unclear. Different bird species seem to weight it differently.

The olfactory map

Pigeons, in particular, seem to use smell. Floriano Papi's experiments in the 1970s showed that homing pigeons released from unfamiliar locations relied on local odors to determine which direction to fly. Pigeons with their olfactory nerves cut were systematically worse at homing from new locations, but unaffected at familiar ones. The proposal is that pigeons learn an "olfactory map" of their home region (this odor comes from the south, this one from the east) and use it for initial orientation.

This is one of the more contested findings in avian navigation. Different research groups have replicated it inconsistently. The current view is probably that olfactory navigation is real but secondary, used by some species as a check on the magnetic compass when the visual cues are absent.

Why so many senses

The natural question is why birds have evolved four or five overlapping systems for the same problem. The answer is that none of them works in all conditions. The sun compass fails on overcast days; the star compass fails when clouds cover the sky; the magnetic compass fails near magnetic anomalies and during solar storms; the infrasound map fails in noise; the olfactory map fails in unfamiliar regions. By weighting them differently depending on conditions, a bird can navigate in fog, in dust storms, over featureless ocean, and across the magnetic anomalies of the polar regions. Each mechanism is fallible; the combination is not.

Multi-sensor fusion is, perhaps unsurprisingly, also how aircraft navigate today. The bird does it with a brain that masses about a gram, on senses partly grounded in quantum mechanics, learned partly in the nest and partly innately. We did not work most of this out until the 1990s and we are still working out the details. There is something humbling in the thought that the godwit landing at the right Tasmanian estuary after eleven days on the wing knew exactly where it was, by means we are still discovering.