Animals display an amazingly reliable, accurate and sometimes mysterious capacity to navigate. When this is done over large distances such as during migrations, without the goal in sight or the cognitive ability to encode all geographical information, we must ask how they achieve such a feat, especially considering the need to correct for wind or oceanic currents and in the absence of discernible landmarks.

The earliest roots of bird navigation research lie in 1882 when Viguier suggested that birds use the spatial distribution of geomagnetic variables such as total intensity and inclination to determine their position. By contrast, Darwin proposed that pigeons tracked their progress during outbound journeys and retraced or integrated this path to return to their original location (referenced in Wiltschko and Wiltschko 2001). These two views manifest in current research still as debate continues regarding the respective importance of outbound route integration versus the capacity for instantaneous location determination by use of external references; the latter in fact constitutes construction of a sort of map. There might be some ontogenetic differentiation in the use of these strategies. Wiltschko and Wiltschko (1978) found that immature homing pigeons could not home if they were subjected to a disrupted magnetic field on the outbound journey. This implies that, at an early stage, there may be recording of information about the outbound journey and integration of this information to calculate a return course, using magnetoreception.

After these questions were posed, it was found that various wild birds were able to return to sites from which they were displaced, including without recourse to familiar visual landmarks. Pigeons too were found to be able to home albeit with decreasing reliability from increasing distances of displacement. It was inferred from this that test pigeons were employing familiar landmarks to a greater extent than wild birds in homing (referenced in Wiltschko and Wiltschko 2001).

At this stage, the field was still hampered by inept tools for research but in the 1950’s several important advances were made. First, Griffin (1952) articulated a framework in which to think about homing, which included three types: (1) use of familiar landmarks or random search until familiar landmarks are found, (2) flying in a fixed direction and (3) inference of the homeward direction and flight in this direction. Orientation types two and three are confounded as both involve the animal showing knowledge of direction in an absolute sense, whether a different direction is followed to reach home (type 3) is another matter. Today these definitions are consolidated and we consider true navigation as a slightly modified version of the third of Griffin’s orientation types. Here, an animal ascertains the direction in which it must travel in order to home (which requires location information) and then travel in it. Integrating an aspect of landmark use into true navigation, Graue (1963) conceived the mosaic map, where animals remember salient landmarks’ spatial relationships but are guided overall by other cues. The extent of landmark use (and so-called piloting; also influence of beaconing when the goal is in sight) remains an active area of research today.

In 1953, Kramer conceived possibly the most important model in navigation: the map-and-compass model. This states that in order to navigate, an animal must be aware of its current and goal locations (map) so calculate a compass course. Then, it must employ a compass of sorts to orient itself, choosing a direction of flight to enact this compass course. Put another way, some internal or external reference must allow alignment of flight direction and compass course as dictated by the two locations on the map – it’s no use knowing you need to go south if you don’t know which way that is! The importance of Kramer’s work is attested in navigation research being phrased in terms of effects on the map or the compass step.

When Kramer put forward his model, he had already posited the sun-compass as a compass mechanism (Kramer 1950). Clock-shift experiments (seminally, Schmidt-Koenig 1958) have extensively probed this mechanism and show that deflections in homing behaviour can be induced in at least a qualitatively predictable way by fast- or slow-shifting the circadian rhythm of birds (e.g. Keeton 1979). The mechanism underlying the sun-compass is thought to be a time-compensated awareness of the position of the sun. Since the sun-arc varies with latitude, the compass is developed on the basis of experience instead of some innate programme and is calibrated with magnetic sensory information (Wiltschko and Wiltschko 1980).

Several studies have looked at the developmental mechanism and plasticity of celestial cues linked with magnetic ones. For example, a study by Cochran, Mouritsen and Wikelski (2004) relatively recently refuted the accepted notion that twilight and star compasses (analogous to sun compasses for nocturnal homing animals or migrants) are calibrated to geomagnetic compasses. Instead, the study showed that exposure of Catharus thrushes to a deflected magnetic field during twilight caused them to follow a deflected flight course on the subsequent night. This showed that the geomagnetic compass was in fact recalibrated and stellar cues were ignored. However, the system is plastic and the geomagnetic compass was recalibrated that night, meaning the deflected flight course was only observed for one night (the night after the experimental treatment). It is worth noting briefly here that Kramer’s mechanism of the sun-compass has largely successfully excluded other hypotheses on navigation, for example inertial navigation, where the route followed by an animal is used to retrace (referened in Wiltschko and Wiltschko 2001).

How sun-compass navigation is achieved cognitively and to what extent the sun is used over other references, such as magnetism, other celestial cues and attendant cues such as polization of skylight is another area of debate (see Guilford and Taylor 2014 for concepts). For example, Walcott and Green (1974) as well as other studies (Visalberghi and Alleva 1979; Walcott 1977) found that magnetic cues are sufficient to disrupt homing using the sun-compass as application of a magnetic field using a Helmholtz coil around a pigeon’s head could make them reverse flight direction despite flying in sunlight. Other studies have found that birds can still navigate on overcast days, which might suggest a role for magnetic or attendant sun cues (Wiltschko and Wiltschko 2001).

Little is known about magnetoreception - the sensory ability of animals to detect the orientation and/or strength of a magnetic field – especially at a physiological level. This is probably as magnetic fields pass through biological tissue so sensory receptors could be located anywhere and also because we have no magnetoreception as humans. These two factors collude and mean we have no scientific or intuitive insight into where to start looking for magnetoreceptors (Johnsen and Lohmann 2005). Efforts by some to find these seemed successful until a later rebuttal (Treiber et al. 2012 and references therein), which found the putative magnetite-based receptors in the rostral subepidermis of the homing pigeon (Columbia livia) beak to in fact be macrophages as confirmed by immunohistology. The mechanisms that have been proposed as a way to detect the geomagnetic field are electromagnetic induction using the ampullae of Lorenzini, chemical magnetoreception through (bio-) chemical reactions that are modulated by magnetic fields and biogenic magnetite. The latter enjoys the most parsimony and empirical evidence. Single-domain Fe3O4 crystals of 50nm diameter could exert torque or pressure on mechanoreceptors presenting a feasible and established mechanism for sensory reception. Moreover, magnetotactic bacteria are well studied and are able to align according to geomagnetic field lines (reviewed by Bazylinski and Frankel 2004). Though the reported discovery of magnetite-based receptors in birds such as the bobolink (Semm and Beason 1990) and fish like the rainbow trout (Walker et al. 1997) has been called into question, the neurophysiological work preceding Treiber and colleagues’ (2012) refutation cannot be ignored. For example, Walker and colleagues’ recordings from the ros V nerve of the rainbow trout (Oncorynchus mykiss) showed responses of the nerve’s spiking activity to magnetic field intensity.

Despite failures in sensory physiology, we do know from behavioural assays and observational data that molluscs, crustaceans, insects and the five vertebrate classes possess magnetoreception (Wiltschko and Wiltschko 1995). Detection of the Earth’s magnetic field lines, an alternative or ancillary mechanism to use of the sun or its attendant cues, such as polarized light or casting of shadows (Guilford and Taylor 2014), could be used to maintain a heading in the context of homing and migration. Utility for such a compass comes from its ability to provide an animal with constantly updated information as to its direction regardless of its absolute instantaneous position on the Earth (paraphrased from Guilford and Taylor p. 136). In this way, magnetoreception very much constitutes the compass component of the map-and-compass model posited by Kramer (1950).

However, consider as well that differences in the strength or orientation of the geomagnetic field with varying declination may also allow some animals to use magnetism for location (Johnsen and Lohmann 2005). This is corroborated by evidence, for example in turtles (Chelonia mydas), which are known to be able to distinguish magnetic fields characteristic of different geographic locations. Lohmann and colleagues (2004) placed turtles from Melbourne Beach, Florida in a cloth harness in a pool and induced an magnetic field characteristic of one either 337km north or south of the test site and observed the turtles swim in opposite directions in order to return to their home site; these turtles were several years old. Evidence exists for these sensory capabilities existing in other animals including salamanders and lobsters (Johnsen and Lohmann 2005).

Birds show at least circumstantial indications of using magnetic clues to distinguish location or declination too. Fransson and colleagues (2001) conducted experiments with Thrush nightingales (Luscinia luscinia) to test similar capabilities. Naturally, migration implies a huge energetic cost and, as such, birds are required to accumulate energy (fat) deposits. This is usually only 20-30% of lean body mass to minimise predation and flight cost but before crossing ecological barriers such as the Sahara fuel loading is greater. Fransson et al. showed that in test nightingales exposed to a magnetic field to mimic that of Egypt, before the Sahara crossing, increase in body weight was significantly higher than the controls (n=8 for both; p=0.004). Such controversies over the relative role of magnetic, solar, other celestial and attendant cues in the construction of the compass persist.

The map component, perhaps unsurprisingly, is also debated, with some clinging onto the notion of a geomagnetic system enabling location of animals via field line orientation and field intensity. Evidence points to birds being guided by atmospheric volatiles. Work by Wallraff (cited in Wiltschko and Wiltschko 2001) posits a system where two volatiles co-occur and intersect at an angle that is not too acute. In this way, birds sense volatiles’ respective scalar concentrations and infer their position. This emerged from an early experiment conducted by Wallraff (1970) testing the effect of the horizon in navigation wherein he prevented full view of the horizon using a palisade and allowed view of the horizon using a glass screen. Surprisingly, those birds that were sheltered from the winds were unable to orient correctly towards home but those who had had the horizon blocked from view could do so correctly. This betrayed the importance of olfactory cues, then labelled ‘atmospheric factor[s]’ by Wallraff. Such a system could account for release-site biases, wherein vanishing bearings deviate from expectation due to an uneven distribution of the volatiles used for mapping as observed by many experimenters (Gagliardo 2013 and references therein).

Many lines of evidence corroborate the importance of olfaction in building the map component in the navigational apparatus of birds. For example, Benvenuti and Wallraff (1985) transported pigeons in airtight containers to a false release site and allowed them to breathe air there for a period before taking them to another release site (transported in pure air), applying nasal anaesthesia and releasing the birds. The true release site was diametrically opposite (approximately) of the false release site with respect to home. The anosmic birds flew along the opposite bearing to the correctly homing one. This indicates that olfactory cues are implicated in navigation and further assays testing the role of compass deflections identified the role of the atmospheric volatiles in the map component. Further unilateral assays of olfaction confirmed its importance. For example, the sectioning of the olfactory nerve or treatment of the nasal mucosa with zinc-sulphate so as to degenerate olfactory receptor neurons both impaired navigational ability as compared with controls (such as the sectioning of the trigeminal nerve in the former assay; Benvenuti and Gagliardo 1996; Papi et al. 1972).

An obvious scepticism of the atmospheric volatile theory is the doubt that volatiles could persist in reliable gradients in the atmosphere over such large scales as to guide navigation. This has also been measured and confirmed directly via gas chromatography by Wallraff (2000) who found that the ratios of atmospheric volatiles were reliable predictors of position over a 200km distance around a pigeon loft site in Germany; behavioural displacement experiments had been conducted here. Indirect evidence also looks promising. In has been proposed for example, that the volatiles present in the atmosphere may be biogenic and therefore we would expect seasonal and ecological variation in their abundance and therefore reliability in guiding navigation. As such, birds have been found to navigate more reliably in areas of high ecological diversity and during seasons when plants are in bloom (see Gagliardo 2013 and references therein).

The discussion of navigation could extend to further topics such as migration, how this is done in immature birds as opposed to experienced ones, the roles of star and twilight cues and an in-depth exploration of attendant sun-cues such as polarized sky-light or shadows on landscapes. These are not covered in the scope of such a brief essay.

To summarise the main points above, bird navigation, since its modern inception in the 1950’s has many persistent controversies and disputes (Alerstam 2006). The dispute as to the relative importance of sun, magnetic and other sensory stimuli, how these are received, cognitively processed or integrated and how this occurs over the developmental timescale of the animal are unclear. Advances and modern techniques like gas chromatography confirming suspected olfactory modalities of map-sensing empower us to answer these questions. What is more, a thorough review of cue-conflict experiments is needed if we are to answer the simplest but most important questions herein encountered.

References

  • Alerstam 2006 Science
  • Bazylinski and Frankel 2004 Nature Reviews Microbiology
  • Benvenuti and Gagliardo 1996 J. Exp. Biol.
  • Benvenuti and Wallraff 1985 J. Comp. Physiol.
  • Cochran, Mouritsen and Wikelski 2004 Science
  • Fransson et al. 2001 Nature
  • Gagliardo 2013 J. Exp. Biol.
  • Graue 1963 Ohio Journal of Science
  • Griffin 1952 Biological Reviews of the Cambridge Philosophical Society
  • Guilford and Taylor 2014 Animal Behaviour
  • Johnsen and Lohmann 2005 Nature Reviews Neuroscience
  • Keeton 1979 Pigeon Navigation in Neural Mechanisms of Behaviour in the Pigeons
  • Kramer (1953) Journal für Ornithologie
  • Kramer 1950 Naturwissenschaften
  • Lohmann et al. 2004 Nature
  • Papi et al. 1972 Monit. Zool. Ital.
  • Schmidt-Koenig (1958) Zeitschrift für Tierpschologie
  • Semm and Beason 1990 Brain Res. Bull.
  • Treiber 2012 Nature
  • Visalberghi and Alleva 1979 Biol. Bull.
  • Walcott 1977 J. Exp. Biol.
  • Walcott and Green 1974 Science
  • Walker et al. (1997) Nature
  • Wallraff 1970 Z. Vgl. Physiol.
  • Wallraff 2000 J. Theor. Biol.
  • Wiltschko and Wiltschko 1978 Naturwissenschaften
  • Wiltschko and Wiltschko 2001 Animal Behaviour
  • Wiltschko, R. & Wiltschko, W. 1995 Magnetic Orientation in Animals. Springer, Berlin, Germany