Homing in pelagic birds: a pilot experiment with white-chinned petrels released in the open sea.
ABSTRACT During the breeding period white-chinned petrels (Procellaria aequinoctialis) repeatedly perform long foraging trips in the open ocean from their breeding island, and are able to home with an astonishing precision. The orientation mechanisms involved are not yet known. By analogy with those used by desert ants moving in a similarly "featureless" environment, one can hypothesise that petrels may home using path-integration. We displaced 11 white-chinned petrels 725-785km from their burrows to the open sea, preventing them from using visual and magnetic route-based information. Three birds carried satellite transmitters. Our results showed that they can home rather efficiently in such conditions.
- SourceAvailable from: Bruno Poucet[show abstract] [hide abstract]
ABSTRACT: This paper reviews spatial memory processes in three highly evolved taxa: hymenoptera, birds and mammals. In these three taxa, the goal location can be memorized egocentrically as a vector specifying the head-referred direction and the distance to the goal, and/or exocentrically as a view specifying the spatial layout of the surrounding landmarks perceived by the animal when standing at the goal. The egocentric coding process requires a path-integration mechanism to update the memorized goal location as a function of the animal's current position. Changes of direction are estimated allothetically (by reference to an external compass) in hymenoptera, idiothetically (on the basis of internal movement-related information) in mammals, and probably in both ways in birds. Computer simulations have shown that path-integration is very sensitive to random errors occurring in idiothetic but not in allothetic estimations. When using the exocentric coding process, hymenoptera store the bearings and angular sizes of landmarks in a compass-oriented colour snapshot taken at the goal. They may then return to the goal by moving so as to reduce the discrepancy between the current view of landmarks and the memorized snapshot. In mammals, this process can be accounted for by a neurobiologically plausible model which highlights the fundamental role of exploration of the environment. The way this process is implemented in birds is less clear.Behavioural Processes. 01/1995;
- [show abstract] [hide abstract]
ABSTRACT: Antarctic procellariiform seabirds forage over vast stretches of open ocean in search of patchily distributed prey resources. These seabirds are unique in that most species have anatomically well-developed olfactory systems and are thought to have an excellent sense of smell. Results from controlled experiments performed at sea near South Georgia Island in the South Atlantic indicate that different species of procellariiforms are sensitive to a variety of scented compounds associated with their primary prey. These include krill-related odors (pyrazines and trimethylamine) as well as odors more closely associated with phytoplankton (dimethyl sulfide, DMS). Data collected in the context of global climatic regulation suggest that at least one of these odors (DMS) tends to be associated with predictable bathymetry, including upwelling zones and seamounts. Such odor features are not ephemeral but can be present for days or weeks. I suggest that procellariiforms foraging over vast distances may be able to recognize these features reflected in the olfactory landscape over the ocean. On the large scale, such features may aid seabirds in navigation or in locating profitable foraging grounds. Once in a profitable foraging area, procellariiforms may use olfactory cues on a small scale to assist them in locating prey patches.Biological Bulletin 05/2000; 198(2):245-53. · 1.23 Impact Factor
Article: Olfaction in BirdsAdvances in The Study of Behavior - ADVAN STUDY BEHAV. 01/1999; 28:247-332.
Behavioural Processes 61 (2003) 95–100
Homing in pelagic birds: a pilot experiment with white-chinned
petrels released in the open sea
Simon Benhamou∗, Joël Bried, Francesco Bonadonna, Pierre Jouventin
CNRS-CEFE, Behavioural Ecology Group, F-34293 Montpellier Cedex 5, France
Received 6 August 2002; received in revised form 19 November 2002; accepted 20 November 2002
During the breeding period white-chinned petrels (Procellaria aequinoctialis) repeatedly perform long foraging trips in the
open ocean from their breeding island, and are able to home with an astonishing precision. The orientation mechanisms involved
sea, preventing them from using visual and magnetic route-based information. Three birds carried satellite transmitters. Our
results showed that they can home rather efficiently in such conditions.
© 2002 Elsevier Science B.V. All rights reserved.
Keywords: Homing; Navigation; Path-integration; Petrel
Long distance homing in birds has been the sub-
ject of intensive research for several decades. The
orientation mechanisms involved have been tackled
with many potential informational sources, making it
difficult to disentangle the orientation mechanisms re-
ally involved. To address this question in a simplified
context, a suitable means consists in studying hom-
ing performances of pelagic birds. The open ocean
constitutes a kind of “featureless” environment due
to the lack of topographic cues, even if there exist
some diffuse information sources such as sea surface
temperatures and odours (Bonadonna et al., 2002).
White-chinned petrels (Procellaria aequinoctialis) are
E-mail address: firstname.lastname@example.org (S. Benhamou).
pelagic birds that seem to be particularly suitable for
long distance homing studies. During the incubation
period, they repeatedly leave their breeding sites on
small isolated oceanic islands and fly for several days
and nights over several thousands of kilometres above
the ocean searching for food at very distant foraging
sites (Catard and Weimerskirch, 1999; Weimerskirch
et al., 1999). They then usually return to their nests
along nearly straight line homing paths.
Like breeding petrels, desert ants (Cataglyphis sp.)
return to the small target corresponding to their nest
with an astonishing precision after a long and sin-
uous foraging trip in an apparently featureless envi-
ronment (see Wehner, 1998). Desert ants are able to
update in memory the direction and distance of their
nest by path-integration based on translational infor-
mation provided by the optical flow from the ground
and rotational information information provided by
a sun compass (Wehner et al., 1996). Such an abil-
ity to home by path-integration has been shown in
0376-6357/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved.
S. Benhamou et al./Behavioural Processes 61 (2003) 95–100
several species of mammals and arthropods (review
in Benhamou and Poucet, 1995), and occasionally in
birds (geese walking short distances: Von Saint Paul,
1982). Path-integration is a site-independent orienta-
tion mechanism that can be efficient in long distance
homing only in species able to rely on compass-based
directional information (Benhamou et al., 1990). Birds
can rely on a sun compass (Schmidt-Koenig, 1990) as
well as on geomagnetic and other types of compass
(Wiltschko and Wiltschko, 1990, 1996). Thus, petrels
might home by path-integration using translational in-
formation provided by the optic flow from the sea sur-
face and sun or magnetic compass-based directional
ments they actively perform, but not passive linear
displacements (at least in the absence of visual cues),
homing experiments, in which animals are passively
displaced in the dark to a distant release site, can
show whether these animals have the capacity of re-
lying on site-dependent information to home. Only a
few homing experiments have been performed on pe-
trels. The homing times obtained in a series of such
experiments by Griffin (1940) and Billings (1968) on
Leach’s storm-petrels, Oceanodroma leucorhoa, and
by Kenyon and Rice (1958) on Laysan albatrosses,
Diomedea immutabilis, supported the hypothesis that
these birds are able to navigate toward their breeding
site using site-dependent information, even if the hom-
ing times are long enough to indicate that the birds
probably did not home along a straight course. In con-
trast, Matthews (1953, 1964) showed that Manx shear-
waters, Puffinus puffinus, can home efficiently only
in sunny conditions. Except for some experiments by
Griffin (1940) and Matthews (1953, 1964), however,
birds were released at coastal locations, so that poten-
tially familiar topographic landmarks were available
at the starting point. In these previous homing experi-
ments, the lack of data about the homing paths made it
rather difficult to determine the actual ability of birds
to home efficiently by taking the shortest route. To our
knowledge, Dall’Antonia et al. (1995) conducted the
only experiment on petrels focusing on homing paths,
using a direction recorder. Their results showed that
Cory’s shearwaters, Calonectris diomedea, can use
site-dependent information to home efficiently. In this
experiment, the only bird that was prevented from us-
ing possibly familiar site-dependent landmark-based
information (being released far away from the coast)
was nevertheless able to home along a rather straight
path from a release site located about 100km from its
Eventually, whereas the similarity of the situations
encountered by petrels and desert ants suggests that
petrels might rely on path-integration when perform-
ing long distance homing over the open sea, the exper-
iments conducted on petrels for the last decades glob-
ally suggest that they can use a navigation mechanism
based on site-dependent information. To get a clearer
image, we conducted a pilot homing experiment with
white-chinned petrels. Eleven birds were displaced far
away from their breeding island without the possi-
bility to use visual or magnetic route-based informa-
tion to feed their path-integrator, and released in the
open sea. Three of them were equipped with satel-
lite transmitters (Argos PTT) to record their homing
2. Material and methods
We conducted our study on Possession Island,
Crozet archipelago, southern Indian Ocean (colony
location: 46.4◦S, 51.9◦E), in mid-December 2000.
This is a particularly suitable place for studying the
homing abilities of petrels at sea because there are no
other land masses in a radius of 1000km and the clos-
est mainland is more than 2000km away (see Fig. 1).
Hence, the white-chinned petrels that breed on Crozet
shuttle back and forth between their burrows and the
open sea, flying over very long distances without
seeing any landmark. At this time of the year, both
parents alternate incubation and foraging at sea, each
stint lasting about 8 days (Jouventin et al., 1985). We
monitored the shifts between the incubating partners
in 60 burrows and selected 11 birds that returned to
their burrow from foraging in the open sea for the
36h prior to the homing experiment. Such birds were,
therefore, supposed to be highly motivated to home
and to incubate rather than to forage once released at
The 11 selected petrels were kept individually in
was glued on each petrel’s head just after capture and
removed just before release. It prevented the bird from
using a magnetic compass during the passive displace-
S. Benhamou et al./Behavioural Processes 61 (2003) 95–100
Fig. 1. Path of the satellite-tracked white-chinned petrel that did not return to its breeding site. H: home (46.4◦S, 51.9◦E), on Possession
Island, Crozet archipelago; RS: release site, labeled with a star, 730km away from home. C: Crozet archipelago, K: Kerguelen archipelago,
M: Marion and Prince Edward Islands, SA: South Africa.
ment by inducing a field intensive enough to overrun
the geomagnetic field around the bird’s head. Thus,
being passively displaced without the possibility to
rely both on visual and geomagnetic information, pe-
path. Three petrels were equipped with miniaturised
Argos satellite transmitters (Microwave PTT100-33g)
in order to record their homing paths. About 30h af-
ter capture, the birds were released individually (every
15min.) at night about mid-way between Crozet and
Kerguelen archipelagos (725 to 785km eastward from
home: 48.3◦S, 61.4◦E–48.4◦S, 61.8◦E). Night releases
constituted a safe procedure, because just released
birds may display unusual behaviour that prompt giant
petrels (Macronectes sp., that often follow ships) to at-
tack them, but these petrels do not fly at night. To min-
imise the impact of the experiment on the reproductive
success of the displaced birds, their single eggs were
placed in an incubator and transitorily replaced in the
burrows by plaster eggs of similar size and weight. We
did not know the fate of the eggs following our ex-
periment. Even unmanipulated white-chinned petrels
breeding on Possession Island have a very low proba-
bility (about 16%) of being successful because of the
presence of introduced black rats (Rattus rattus; see
Bried and Jouventin, 1999). Fortunately, there exist
several islands without rats (e.g. East Island, 18km
east of Possession Island) where white-chinned petrels
can reproduce with higher success.
Data were analysed globally in terms of homing
times, and for the petrels equipped with satellite trans-
mitters, in terms of homing paths. The least reliable
fixes, based on fewer than four emissions (Argos data
classes A and B), were discarded prior to analysis. The
orientation performance was measured by the straight-
ness index D/L (ranging from 0 to 1) which is the ratio
of the beeline distance between the release site and the
home over the path length, and constitutes a reliable
estimate of the orientation efficiency (see Benhamou
and Bovet, 1992).
Three out of the eleven petrels released at sea did
not return to their burrow, but presumably preferred to
remain in the open sea. This is clearly shown by one of
these birds that carried a satellite transmitter (Fig. 1).
Released at 730km from its burrow, this petrel first
flew over a 1910km loop lasting 61h before coming
back within 20km from its release site. Then, it started
moving for about two days above subantarctic waters
S. Benhamou et al./Behavioural Processes 61 (2003) 95–100
Fig. 2. Paths of the two satellite-tracked white-chinned petrels that
returned to their breeding site. H: home (46.4◦S, 51.9◦E); RS:
release site, labeled with a star, 750km away (a) or 770km away
(b) from home.
(45–47◦S, 63–65◦E), presumably foraging according
to its low speed (about 12km/h on average) along an
about 500km path, and then moved progressively to-
ward sub-tropical waters near the South-African coast
weeks after being released, and was still there 10 days
later when the transmitter stopped emitting.
Two of the eight petrels that returned to their
burrows were equipped with satellite transmitters
(Fig. 2). Excepted when they were about to reach
Crozet archipelago, these birds were never in visual
contact with a land at any time of their way to home.
One petrel, released 750km away, homed in 67h
along a 1565km path (Fig. 2a), whereas the other
one, released 770km away, homed in 95h along a
2310km path starting with a 805km loop lasting
about 36h which led it to come at about 15km from
its release site (Fig. 2b). The homing path efficiency
(D/L) was therefore equal to 0.48 for the former, and
equal to 0.33 (including the loop) or 0.51 (excluding
the loop) for the latter.
Complementary information was provided by the
recapture times of the other six petrels. One of them
was recaptured only 68h after release, but the other
five were recaptured after 8–13 days. In fact, such
long recapture delays should be considered with cau-
tion because they could be far longer than actual
homing times. For example, one of the petrel tracked
by satellite (Fig. 2b) eventually homed in 59h (ex-
cluding the initial 36h loop) but stayed at home
only a very short time (so that its burrow was empty
when we checked it) and was recaptured at its burrow
only 18 days later, after it had foraged at sea, for a
few days near the Antarctica pack zone (59◦–60◦S,
64–65◦E) and later for 10 days on a single spot in
subantarctic waters (43.5◦S, 57◦E) about 500km from
Of 11 birds experimentally displaced more than
700km away, eight returned to their nests located on
a small isolated oceanic island. Three of them homed
in a few days. The probability of reaching the goal by
choosing a homing direction at random is very low,
less than 0.05 (assuming that the breeding island can-
not be detected at a distance greater than 100km, cor-
responding to an angular sector from the release site
less than 15◦). As the petrels were prevented from
integrating their outward movement, this result indi-
cates that they used some site-dependent information
to home. In addition, the tracked bird which did not re-
turn to its breeding site appeared to desert rather than
to be lost: it did not wander at random, but moved to-
ward waters close to the South African coast, that is
the main wintering area of white-chinned petrels, and
one of the usual foraging areas of the Crozet popula-
tion during the incubation period (Weimerskirch et al.,
The efficiency of the other two tracked birds was
rather low (D/L ≤ 0.5) with respect to that of the
six individuals tracked by Catard and Weimerskirch
(1999) that returned to their burrow on Crozet
archipelago after having performed spontaneous for-
aging flights during the incubation period (D/L = 0.9
on the average). One explanation would be that visual
and magnetic route-based information, although not
necessary to home, would nevertheless enable petrels
S. Benhamou et al./Behavioural Processes 61 (2003) 95–100
to improve their homing efficiency when they can
rely on it. In addition, when they actively perform
the outward paths during their spontaneous forag-
ing activities, petrels can probably gain other useful
locational and/or directional cues from other infor-
mational sources such as odours. Indeed, petrels are
endowed with a well-developed sense of smell (re-
view in Roper, 1999) and can use it in foraging (e.g.
Verheyden and Jouventin, 1994; Nevitt et al., 1995;
Nevitt, 2000) and in burrow recognition (Bonadonna
et al., 2001; Bonadonna and Bretagnolle, 2002), and
they might use it as well in long-distance orientation
(Bonadonna et al., 2002). In our experiment, as the
petrels were locked up in closed opaque boxes (with
just a few small holes for breathing) during the out-
ward trip to the release site, they had probably some
difficulties for perceiving odours correctly. We cannot
also exclude that the white-chinned petrels we tracked
homed less efficiently because they experienced un-
favourable wind conditions. Indeed, the direction of
the prevailing winds in the southern Indian Ocean
in December (mainly eastward) did not favour the
adoption of straight homing paths by petrels released
east of Crozet because they had to home with head
wind. Wandering albatrosses prefer to fly with tail
or side winds, and take sinuous paths when flying
with head wind (Jouventin and Weimerskirch, 1990;
Weimerskirch et al., 2000). The six white-chinned
petrels tracked by Catard and Weimerskirch (1999)
did not home from east of Crozet, and so experienced
more favourable wind patterns.
To conclude, this pilot experiment clearly showed
that white-chinned petrels are able to home us-
ing only site-dependent information. The role of
path-integration in natural conditions, i.e. in homing
after a spontaneous foraging path, remains uncertain.
Further experiments are needed to address this point,
and to determine which type of site-dependent infor-
mation is used by homing petrels. The data provided
by Argos satellite tracking were insufficiently accu-
rate and frequent (about a fix every 2h) to enable us to
determine the type of the information used by homing
petrels based on the detailed path analysis proposed by
Benhamou and Bovet (1992). The micro-GPS loggers
recently developed (Steiner et al., 2000; Weimerskirch
et al., 2002) are able to provide satellite location data
with unprecedented precision and frequency, and thus
open very promising perspectives in this field.
This study was supported by the Polar French In-
stitute (IPEV). We are sincerely grateful to Florence
Hesters, Amanda Searby, and Greg Cunningham for
their help in the field. The experiment complied with
the French current rules regarding the use of animals
in experiments (licence #34-100) and was approved
by the IPEV’s Ethical Committee.
Benhamou, S., Bovet, P., 1992. Distinguishing between elementary
orientation mechanisms by means of path analysis. Anim.
Behav. 43, 371–377.
Benhamou, S., Poucet, B., 1995. A comparative analysis of spatial
memory processes. Behav. Proc. 35, 113–126.
Benhamou, S., Sauvé, J.-P., Bovet, P., 1990. Spatial memory
in large scale movements: efficiency and limitation of the
egocentric coding process. J. Theor. Biol. 145, 1–12.
Billings, S.M., 1968. Homing in Leach’s petrel. Auk 85, 36–43.
Bried, J., Jouventin, P., 1999. Influence of breeding success on
mate fidelity in long-lived birds: an experimental study. J. Av.
Biol. 30, 392–398.
Bonadonna, F., Benhamou, S., Jouventin, P., 2002. Orientation
in “featureless” environments: the extreme case of pelagic
birds. In: Berthold, P., Gwinner, E. (Eds.), Avian Migration.
Bonadonna, F., Bretagnolle, V., 2002. Smelling home: a good
solution for burrow finding in nocturnal petrels. J. Exp. Biol.
Bonadonna, F., Spaggiari, J., Weimerskirch, H., 2001. Could
osmotaxis explain blue petrels ability in returning to their
burrows at night? J. Exp. Biol. 204, 1485–1489.
Catard,A., Weimerskirch, H.,
Proceedings of the 22nd International Ornithological Congress,
Birdlife South Africa, Johannesburg, pp. 3008–3023.
Dall’Antonia, L., Dall’Antonia, P., Benvenuti, S., Ioalè, P., Massa,
B., Bonadonna, F., 1995. The homing behaviour of Cory’s
shearwaters (Calonectris diomedea) studied by means of a
direction recorder. J. Exp. Biol. 198, 359–362.
Griffin, D.R., 1940. Homing experiments with Leach’s petrels.
Auk 57, 61–74.
Jouventin, P., Mougin, J.L., Stahl, J.C., Weimerskirch, H., 1985.
Comparative biology of the burrowing petrels of the Crozet
islands. Notornis 32, 157–220.
Jouventin, P., Weimerskirch, H., 1990. Satellite tracking of
wandering albatrosses. Nature 343, 746–748.
Kenyon, K.W., Rice, D.W., 1958. Homing of Laysan albatrosses.
Condor 60, 3–6.
Matthews, G.V., 1953. Navigation in the Manx shearwater. J.
Exp. Biol. 30, 370–396.
S. Benhamou et al./Behavioural Processes 61 (2003) 95–100
Matthews, G.V., 1964. Individual experience as a factor in the
navigation of Manx shearwaters. Auk 81, 132–146.
Nevitt, G.A., 2000. Olfactory foraging by Antarctic Procellariiform
seabirds: life at high Reynolds numbers. Biol. Bull. 198, 245–
Nevitt, G.A., Veit, R.R., Kareiva, P., 1995. Dimethyl sulphide as
a foraging cue for Antarctic Procellariiform seabirds. Nature
Roper, T.J., 1999. Olfaction in birds. Adv. Stud. Behav. 28, 247–
Schmidt-Koenig, K., 1990. The sun compass. Experientia 16, 336–
Steiner, I., Bürgi, C., Werffeli, S., Dell’Omo, G., Valenti, P.,
Tröster, G., Wolfer, D.P., Lipp, H.P., 2000. A GPS logger and
software for analysis of homing in pigeons and small mammals.
Physiol. Behav. 71, 589–596.
Verheyden, C., Jouventin, P., 1994. Olfactory behavior of foraging
Procellariiformes. Auk 111, 285–291.
Von Saint Paul, U., 1982. Do geese use path-integration for walking
home? In: Papi, F., Wallraff, H.G. (Eds.), Avian Navigation,
Springer-Verlag, Berlin, pp. 298–307.
Wallraff, H.G., 2001. Navigation by homing pigeon: updated
perspective. Ethol. Ecol. E 13, 1–48.
Wehner, R., 1998. Navigation in context: grand theories and basic
mechanisms. J. Avian Biol. 29, 370–386.
Wehner, R., Michel, B., Antonsen, P., 1996. Visual navigation in
insects: coupling of egocentric and geocentric information. J.
Exp. Biol. 199, 129–140.
Weimerskirch, H., Bonadonna, F., Bailleul, F., Mabille, G.,
Dell’Omo, G., Lipp, H.P., 2002. GPS tracking of foraging
albatrosses. Science 295, 1259.
Weimerskirch, H., Catard, A., Prince, P.A., Cherel, Y., Croxall,
J.P., 1999. Foraging of white-chinned petrels Procellaria
aequinoctialis at risk: from the tropics to Antarctica. Biol.
Conserv. 87, 273–275.
Weimerskirch, H., Guionnet, T., Martin, J., Shaffer, S.A., Costa,
D.P., 2000. Fast and fuel efficient? Optimal use of wind by
flying albatrosses. Proc. R. Soc. Lond. B 267, 1869–1874.
Weimerskirch, H., Jouventin, P., Mougin, J.-L., Stahl, J.-C., Van
Beveren, M., 1985. Banding recoveries and the dispersal of
seabirds breeding in the French Austral and Antarctic territories.
Emu 85, 22–32.
Wiltschko, W., Wiltschko, R., 1990. Magnetic orientation and
celestial cues in migratory orientation. Experientia 46, 342–352.
Wiltschko, W., Wiltschko, R., 1996. Magnetic orientation in birds.
J. Exp. Biol. 199, 29–34.