Homing Behavior

Abstract

Homing behavior is a form of animal behavior—an animal's capacity to return to a specific place, whether a nest meters off or a natal beach an ocean away. It is a problem in spatial cognition: to steer home an animal must know the direction to travel and, for true navigation, where it is relative to the goal. This article develops four mechanisms research has separated—path integration, which tracks a home vector by summing an animal's movements; a solar, stellar, or magnetic compass, which supplies a stable heading; a map sense read from the geomagnetic field, which supplies position; and geomagnetic imprinting, by which long-lived migrants relocate their birthplace. How these cues are weighted, and whether a true bicoordinate magnetic map exists, remains contested. A worked example and three interactive demonstrations accompany the text.

Keywords: homing behavior, path integration, animal navigation, magnetoreception, geomagnetic imprinting

A desert ant that has wandered a hundred meters from its nest on a looping foraging trip does not retrace its steps. When it finds food it turns and runs almost straight back, along a path it never walked, to within a body length of the hole it came from. A Manx shearwater carried in a box to a continent it has never seen, then released, sets a course for its burrow and arrives. A sockeye salmon, years after leaving the gravel of a particular stream, finds its way back from the open Pacific to spawn in that same stream. These are all homing: the return of an animal to a defined place from which it has been displaced.

This article treats homing as a problem in spatial cognition rather than a single instinct. The questions are mechanical. How does an animal keep track of where home is while it is moving away from it? How does it hold a straight course once it has decided which way to go? How does it know where it is when it has been displaced to unfamiliar ground? Each question corresponds to a distinct capacity—a running home vector, a compass, and a map—and much of the science of navigation is the work of pulling these apart and testing them one at a time.

What Homing Behavior Is

Homing behavior is the ability of an animal to return to a particular location—a nest, a den, a breeding site, or a home range—after it has moved or been moved away from it. In MeSH it is classified under animal behavior, and the category spans an enormous range of scales: the few meters a wasp travels back to its burrow, the tens of kilometers a homing pigeon covers from an unfamiliar release site, and the thousands of kilometers a sea turtle crosses to the beach where it hatched.

The field's central distinction is between compass orientation and true navigation. An animal using compass orientation can hold a fixed heading—it can head north—but cannot correct for being displaced sideways, because it has no sense of where it is. An animal capable of true navigation can be taken to a site it has never visited, in a direction it was not expecting, and still work out the correct way home (Lohmann et al., 2007). The difference is the difference between a compass and a compass plus a map, and establishing which animals have which has driven the displacement experiments that define the discipline (Walcott, 1996).

Path Integration: the Running Home Vector

The simplest way to always know the way home is to never lose track of it. In path integration, an animal continuously combines the direction and distance of each segment of its outbound journey into a single running estimate—the home vector—that points straight back to the start and shrinks as it returns. No landmarks are required; the computation runs on the animal's own motion. The desert ant Cataglyphis, foraging across featureless salt pans where landmarks are scarce, is the classic model. Rudiger Wehner and Martin Muller showed that after a long, winding search an ant computes a nearly exact return vector and runs it home, then switches to a local search once the vector is used up (Muller & Wehner, 1988).

Path integration requires two measurements taken continuously and multiplied together: a direction for each step and a distance travelled along it. The ant takes its directional reference from the pattern of polarized skylight—a sky compass—and gauges distance with an internal step-counter, or pedometer, that literally tallies strides (Wehner, 2003). Because the estimate is built by accumulating small increments, it also accumulates small errors, which is why an ant that reaches the end of its home vector without finding the nest begins a systematic spiral search rather than continuing blindly. The same strategy, scaled up and run on different sensors, reappears across the animal kingdom, and its logic—summing one's own movements to keep a vector home—is the content of the worked example below.

Compasses: Sun, Stars, and the Magnetic Field

A home vector is only as good as the directional reference it is built on, and animals draw that reference from several independent compasses. The sun compass uses the sun's azimuth, corrected for its movement across the sky by an internal clock; shift an animal's clock by a few hours and its bearing rotates by a predictable angle, the signature experiment that proves a sun compass is in use. Night-migrating birds read a star compass from the rotation of the night sky about the celestial pole. And many animals carry a magnetic compass that reads direction from the geomagnetic field.

The magnetic compass was first demonstrated unambiguously in a bird by Wolfgang and Roswitha Wiltschko, who showed that European robins orient to an artificial magnetic field and—crucially—that their compass reads the inclination of the field lines rather than their polarity, responding to the axis of the field and its angle to gravity rather than to north versus south (Wiltschko & Wiltschko, 1972). In the homing pigeon, William Keeton provided the complementary evidence: bar magnets glued to a pigeon's back disrupted its orientation under overcast skies, when the sun compass was unavailable, showing that the two compasses back each other up (Keeton, 1971). Half a century of work since has localized a light-dependent magnetic compass to cryptochrome photoreceptors and a dedicated visual-processing pathway in the night-migratory bird brain (Mouritsen, 2018). Insects, by contrast, hold their course largely on celestial cues, whose memory-based use in navigation Thomas Collett and Matthew Collett traced through the visual systems of bees and ants (Collett & Collett, 2002).

The Map Sense: Magnetic Maps and Geomagnetic Imprinting

A compass cannot explain true navigation, because knowing which way is north does not tell a displaced animal where it is. For that an animal needs a map—a way to read its position from environmental gradients—and the leading candidate is the geomagnetic field itself. The field varies predictably over the Earth's surface in two largely independent ways: its intensity and its inclination (the angle the field lines make with the horizontal) both change with latitude. An animal able to sense both has, in principle, a bicoordinate magnetic map: two gradients that intersect to specify a location (Lohmann et al., 2022).

Kenneth Lohmann and Catherine Lohmann demonstrated such a map in hatchling loggerhead sea turtles, showing that turtles exposed to the magnetic signatures of different oceanic locations swam in the directions that would keep them within their migratory gyre—responding to magnetic coordinates as positional information, not merely as a compass (Lohmann et al., 2007). The same capacity was shown in an invertebrate when Larry Boles and Lohmann found that spiny lobsters displaced to unfamiliar sites oriented homeward using magnetic cues alone (Boles & Lohmann, 2003). The map's resolution is coarse—good for ocean-basin navigation, not for pinpointing a stream mouth—so long-distance natal homing is thought to use the magnetic map for the open-sea leg and other cues, such as olfaction, for the final approach (Dittman & Quinn, 1996).

How an animal relocates the exact site of its birth after years away is answered by geomagnetic imprinting. Lohmann, Nathan Putman, and Catherine Lohmann proposed that young animals learn the magnetic signature of their natal site and, at reproductive age, navigate back to the coordinates that match the learned value (Lohmann et al., 2008). The hypothesis unifies salmon and sea turtles under one mechanism (Lohmann & Lohmann, 2019), and it makes a testable prediction: because the field drifts over time, the geographic target should shift with it. Putman and colleagues confirmed exactly this for Pacific salmon, whose return routes tracked year-to-year changes in the geomagnetic field (Putman et al., 2013).

The Neural Basis of the Home Vector

The home vector of path integration is not a metaphor; it is computed in neural tissue, and in insects the circuitry has been mapped in detail. The central complex, a set of midline neuropils in the insect brain, contains a ring of neurons that encodes the animal's current heading as a bump of activity—a biological compass needle—and downstream circuitry that can combine this heading signal with a measure of distance to accumulate a vector. Stanley Heinze has characterized the central complex as the brain's navigation hub, the site where compass information and self-motion are integrated into steering commands (Heinze, 2017).

In vertebrates the corresponding machinery is the hippocampal-entorhinal spatial system. Bruce McNaughton and colleagues framed path integration as the computational core of the mammalian cognitive map, arguing that grid cells and head-direction cells provide precisely the metric of direction and distance that integrating a home vector requires (McNaughton et al., 2006). The convergence is striking: an insect with a brain of a million neurons and a mammal with billions solve the same geometric problem—keeping a running sum of where the animal has been—with recognizably similar ingredients, a heading signal and a distance signal combined over time.

Worked Example

Path integration reduces to vector addition, and a short outbound trip makes the arithmetic concrete. Measure position in meters east and north of the nest, and record each leg of a foraging ant's winding outbound path as a displacement with an east component and a north component. The home vector is found by summing the legs to get the net displacement, then reversing it: the nest lies in the direction opposite to the animal's accumulated travel, at the straight-line distance of that accumulated displacement. Consider a three-leg outbound trip:

The ant has walked a total path of 10 + 6 + 3 = 19 m, but its net displacement from the nest is only (12 m east, 5 m north). The straight-line distance home is the magnitude of that displacement, √(12² + 5²) = √169 = 13 m—a clean 5–12–13 right triangle. The direction to the accumulated position is a bearing of arctan(12 ÷ 5) = 67.4° (measured clockwise from north), so the home vector points the opposite way, at 67.4° + 180° = 247.4°, for 13 m. The ant turns to 247.4°, runs 13 m, and arrives—having covered the shortcut it never walked outbound. The signature of the strategy is the gap between path length and vector length: the more a route meanders, the shorter the home vector is relative to the distance travelled, and the greater the saving. The PathIntegrationDemo lets the reader build an outbound path leg by leg and watch the home vector update. The numbers are illustrative; the structure—sum the legs, reverse the total—is the content.

Discussion

The mechanisms of homing are best understood not as rivals but as a layered toolkit, each suited to a different scale and circumstance. Path integration is self-contained and always available but drifts with accumulated error, which makes it ideal for the short, landmark-poor foraging trips of an ant and unreliable over thousands of kilometers (Muller & Wehner, 1988). A compass corrects nothing about position but holds a course indefinitely, and animals redundantly equip themselves with several—solar, stellar, magnetic—so that cloud or nightfall never leaves them without one (Keeton, 1971; Wiltschko & Wiltschko, 1972). The magnetic map supplies the one thing a compass cannot, position, and it is what lifts an animal from orientation to genuine navigation (Lohmann et al., 2007).

Seen this way, the spectacular feats of natal homing are not a separate magic but a combination of these ordinary parts. A salmon or a turtle uses a magnetic map, imprinted in youth, to navigate to the right stretch of coast, and then switches to finer, local cues for the last stage of the journey (Lohmann et al., 2008; Dittman & Quinn, 1996). The layering is itself the design: coarse and robust at long range, fine and fallible at short range, with a handoff between them.

Cognitive Implications

For cognitive science, homing is a tractable model of spatial representation. The compass-plus-map decomposition is, in effect, the animal's version of the distinction between a heading and a position—between knowing which way one faces and knowing where one stands—and the displacement experiment is a clean behavioral assay for which of the two an animal possesses (Walcott, 1996). Path integration, in particular, is a continuous state-estimation problem of exactly the kind studied in the mammalian cognitive map, and the finding that grid and head-direction cells supply its ingredients ties the insect and vertebrate accounts to a common computational description (McNaughton et al., 2006; Heinze, 2017).

The magnetic map also reframes a long-standing question about the senses. That animals extract a two-coordinate position from a field humans cannot perceive at all shows that the perceptual world an organism navigates can include dimensions entirely outside our own, and that spatial cognition is built on whatever reliable gradients the environment offers (Lohmann et al., 2022). Geomagnetic imprinting adds a developmental dimension: the map is not merely sensed but learned, early and durably, and recalled across years and ocean basins—a memory feat that invites direct comparison with other forms of early, lasting learning (Lohmann et al., 2008).

Current Directions

The most active front concerns the map sense in migratory birds. Having established the magnetic compass, the field has turned to whether birds also read a magnetic map for position, and to what its coordinates are. Nikita Chernetsov and colleagues showed that Eurasian reed warblers can use magnetic declination—the angle between magnetic and geographic north—to solve the longitude problem, a coordinate long thought inaccessible from the field (Chernetsov et al., 2017). Dmitry Kishkinev and colleagues then demonstrated, by virtually displacing birds with an altered field, that reed warblers extrapolate geomagnetic cues beyond their familiar range to correct course—direct evidence of a true navigational map in a songbird (Kishkinev et al., 2021).

A second direction probes the information homing animals actually use by withholding it. Oliver Padget and colleagues showed that shearwaters denied the normal sensory record of their outward journey took longer, detouring homing routes, implying that the outbound trip itself supplies navigational information the birds would otherwise use (Padget et al., 2022). The open questions span levels: the biophysical identity of the magnetic receptor remains unsettled (Mouritsen, 2018), the resolution and update rate of the magnetic map are not fully known, and how the several compasses and the map are weighted and combined in a single decision is an active modeling problem.

Common Misconceptions

Homing animals simply retrace their outbound route.

Path integration lets an animal compute a direct home vector and take a novel shortcut it never walked, rather than reversing its winding path. The desert ant runs almost straight home across ground it did not cross on the way out (Muller & Wehner, 1988).

A magnetic compass is enough to navigate home.

A compass gives direction but not position, so it cannot correct for displacement. True navigation additionally requires a map sense to establish where the animal is relative to its goal (Lohmann et al., 2007).

The bird magnetic compass detects north and south like a hand compass.

The avian magnetic compass is an inclination compass: it reads the angle the field lines make with the horizontal, not their polarity, and distinguishes poleward from equatorward rather than north from south (Wiltschko & Wiltschko, 1972).

Natal homing means the animal remembers landmarks from birth.

Long-distance natal homing is explained by geomagnetic imprinting—learning the magnetic signature of the birthplace—not by remembering visual landmarks across an ocean. Local cues guide only the final approach (Lohmann et al., 2008).

Glossary

Bicoordinate map.

A positional map built from two independent environmental gradients whose intersection specifies a location, as geomagnetic intensity and inclination do.

Central complex.

A midline region of the insect brain that encodes heading as a ring of neural activity and integrates it with distance to compute a home vector.

Compass orientation.

The ability to hold a fixed heading using a directional cue, without knowledge of one's position relative to a goal.

Geomagnetic imprinting.

The learning of the magnetic signature of a natal site early in life and its later use to navigate back to the matching coordinates.

Home vector.

The continuously updated straight-line direction and distance from an animal's current position back to its starting point, maintained by path integration.

Homing behavior.

The return of an animal to a specific location, such as a nest or natal site, after moving or being displaced away from it.

Inclination compass.

A magnetic compass that reads direction from the angle the field lines make with the horizontal rather than from their polarity.

Magnetic declination.

The local angle between magnetic north and geographic north, usable as a cue to longitude.

Magnetic inclination.

The angle the geomagnetic field lines make with the horizontal, which varies predictably with latitude.

Magnetoreception.

The sensing of the Earth's magnetic field, used by animals for both compass direction and positional information.

Natal homing.

The return of an adult animal to the specific site where it was born or hatched in order to reproduce.

Path integration.

The continuous summation of an animal's own movements to maintain a running estimate of its position relative to a starting point.

Sun compass.

A directional reference derived from the sun's azimuth, corrected for the sun's apparent movement by an internal clock.

True navigation.

The ability to orient toward a goal from an unfamiliar location, requiring both a map sense for position and a compass for direction.

Key Researchers

Stanley Heinze

(Lund University). Neuroscientist who maps the insect central complex as the brain's navigation hub, identifying the circuitry that encodes heading direction and supports vector-based path integration. ORCID

William T. Keeton

(1933–1980; Cornell University). Zoologist who pioneered controlled homing-pigeon release experiments, showing that bar magnets and overcast skies disrupt initial orientation and thereby establishing that pigeons use both a sun compass and a magnetic compass. Wikipedia

Dmitry Kishkinev

(Keele University). Behavioral ecologist whose virtual-displacement experiments provided evidence that migratory reed warblers possess a true navigational map, extrapolating geomagnetic cues to correct for displacement far outside their normal range. ORCID

Kenneth J. Lohmann

(University of North Carolina at Chapel Hill). Biologist who demonstrated magnetic maps and geomagnetic imprinting in sea turtles and salmon, showing that animals read latitude and longitude from the geomagnetic field to home on their natal site. ORCID

Henrik Mouritsen

(Carl von Ossietzky University of Oldenburg). Neurobiologist who linked long-distance avian navigation to a light-dependent magnetic compass, implicating cryptochrome photoreceptors and a dedicated forebrain visual pathway in the night-migratory bird brain. ORCID

Rudiger Wehner

(University of Zurich, emeritus). Neuroethologist who established path integration as the core of desert-ant navigation, quantifying the vector memory and sky-compass cues an animal uses to compute a straight homeward course. Wikidata

Frequently Asked Questions

What is homing behavior?

Homing behavior is an animal's ability to return to a specific place—a nest, den, or natal site—after moving or being displaced away from it. In MeSH it is classified as a form of animal behavior, and it spans scales from an insect's return to its burrow to a sea turtle's ocean crossing to its hatching beach (Lohmann et al., 2007).

How does an animal know the way home?

There are several mechanisms. Path integration keeps a running home vector by summing the animal's own movements; a compass (solar, stellar, or magnetic) supplies a stable heading; and a map sense, often read from the geomagnetic field, supplies position. Most homing animals combine two or more of these (Muller & Wehner, 1988; Lohmann et al., 2022).

What is path integration?

Path integration is the continuous combination of the direction and distance of every segment of an outbound trip into a single running estimate of the straight-line way back to the start. It needs no landmarks, running instead on the animal's own motion, and it lets a desert ant take a direct shortcut home across ground it never crossed (Muller & Wehner, 1988).

What is the difference between a compass and a map in navigation?

A compass supplies direction—it lets an animal hold a heading—but it says nothing about position, so an animal with only a compass cannot correct for being displaced sideways. A map sense supplies position, telling the animal where it is relative to its goal. True navigation requires both (Lohmann et al., 2007; Walcott, 1996).

How do animals use the Earth's magnetic field to navigate?

Two ways. As a compass, many animals read direction from the field; the bird compass notably reads inclination rather than polarity. As a map, animals read position from the field's intensity and inclination, which vary predictably with location, giving a bicoordinate estimate of where they are (Wiltschko & Wiltschko, 1972; Lohmann et al., 2022).

How does a salmon find its home stream after years at sea?

By geomagnetic imprinting. A young salmon learns the magnetic signature of its birthplace and, years later, navigates back to the coordinates that match the learned value, using the magnetic map for the oceanic leg and finer cues such as smell for the final approach into the stream (Lohmann et al., 2008; Dittman & Quinn, 1996).

Where in the brain is homing computed?

In insects, the central complex encodes the animal's heading as a ring of activity and integrates it with distance to form a home vector. In mammals, the hippocampal-entorhinal system—grid cells and head-direction cells—supplies the same metric of direction and distance for the cognitive map (Heinze, 2017; McNaughton et al., 2006).

Can birds really navigate like using a GPS?

Recent evidence suggests some can perform true navigation. Virtually displacing migratory reed warblers with an altered magnetic field shows they extrapolate geomagnetic cues beyond their familiar range to correct course, and they can use magnetic declination to judge longitude—the components of a genuine magnetic map, though coarser and slower than a satellite fix (Kishkinev et al., 2021; Chernetsov et al., 2017).

References