Transoceanic Migration, Spatial Dynamics, and Population Linkages of White Sharks

Abstract

The large-scale spatial dynamics and population structure of marine top predators are poorly known. We present electronic tag and photographic identification data showing a complex suite of behavioral patterns in white sharks. These include coastal return migrations and the fastest known transoceanic return migration among swimming fauna, which provide direct evidence of a link between widely separated populations in South Africa and Australia. Transoceanic return migration involved a return to the original capture location, dives to depths of 980 meters, and the tolerance of water temperatures as low as 3.4 degrees C. These findings contradict previous ideas that female white sharks do not make transoceanic migrations, and they suggest natal homing behavior.

Full text

subtracting each judge’s rating of the typical
American from his or her rating of the typical
compatriot for each NCS item. Assuming that
cultures agree on the typical American, this proce-
dure in effect subtracts the bias plus a constant and
leaves a potentially better estimate of national
character. We standardized the differences as T
scores, using difference score normative values from
the worldwide sample, excluding the United States.
The difference scores were highly correlated with
NCS scores (rs 0 0.65 to 0.91, P G 0.001) and
provided essentially the same results. ICCs between
difference scores and NEO-PI-R observer ratings
ranged from –0.44 for England to 0.48 for Lebanon
(median, 0.03). ICCs between differences scores and
NEO-PI-R self-reports ranged from –0.47 for Russia
to 0.53 for Poland (median, 0.01). For the five
factors, correlati ons with observer ratings across
cultures ranged from 0.08 to 0.23, and those with
self-reports ranged from –0.37 to 0.23. These results
suggest that the lack of correspondence between
NEO-PI-R and NCS profiles is not simply due to
different sta ndards of evalua tion in different
cultures. A different issue concerns the reference-
group effect (28), according to which self-reports
and observer ratings of individuals are implicitly
made by reference to the distribution of scores in the
rater’s culture. Such an effect would tend to make
aggregate personality scores uniform for all cultures,
and the failure to find correlations with NCS factors
would be due to a lack of variation in aggregate
NEO-PI-R means. However, NEO-PI-R means in fact
vary systematically across cultures and show strong
correlations across methods and with other culture-
level variables (12, 14). Thus, the reference-group
effect cannot explain the failure to find correlations
with NCS scales.
28. S. J. Heine, D. R. Lehman, K. P. Peng, J. Greenholtz,
J. Pers. Soc. Psychol. 82, 903 (2002).
29. F. van de Vijver, K. Leung, J. Pers. 69, 1007 (2001).
30. D. L. Hamilton, T. L. Rose, J. Pers. Soc. Psychol. 39,
832 (1980).
31. T. W. Adorno, E. Frenkel-Brunswik, D. J. Levinson, R. N.
Sanford, The Authoritarian Personality [Norton, New
York, 1969 (original work published 1950)].
32. F. H. Allport, The Nature of Prejudice (Houghton
Mifflin, New York, 1954).
33. R.R.M. receive s royalties from the Revised NEO
Personality Inventory. This research was supported
in part by the Intramural Research Program of NIH,
National Institute on Aging. Czech participation was
supported by grant 406/01/1507 from the Grant
Agency of the Czech Republic and is related to
research plan AV AV0Z0250504 of the Institute of
Psychology, Academy of Sciences of the Czech
Republic. S.G.’s participation was supported by the
Turkish Academy of Sciences. Burkinabe
`
and French
Swiss participation was supported by a grant from
the Swiss National Science Foundation to J.R. The
data collection in Hong Kong was supported by
Research Grants Council Direct Allocation Grants
(DAG02/03.HSS14 and DAG03/04.HSS14) awarded
to M.Y. Data collection in Malaysia was supported by
Univesiti Kebangsaan Malaysia Fundamental Re-
search Grant 11JD/015/2003 awarded to K.A.M.
Portions of these data were presented at the 113th
Convention of the American Psychological Association,
August 2005, Washington, DC. For helpful comments
on the manuscript, we thank Y. H. Poortinga; for their
assistance on this project we t hank F. Abal, L. de
Almeida, S. Baumann, H. Biggs, D. Bion, A. Butkovic
´
,
C. Y. Carrasquillo, H. W. Carvalho, S. Catty, C.-S.
Chan,A.Curbelo,P.Duffill,L.Etcheverry,L.Firpo,J.
Gonzalez, A. Gramberg, H. Harrow, H . Imuta, R.
Ismai l, R. Kamis, S. Ka nnan, N. Messoulam, F. Molina,
M. Montarroyos Calegaro, S. Mosquera, J. C. Munene,
V. Najzrova, C. Nathanson, D. Padilla, C. N. Scollon, S. B.
Sigurdardottir, A. da Silva Bez, M. Takayama, T. W.
Teasdale, L. N. Van Heugten, F. Vera, and J. Villamil.
Supporting Online Material
www.sciencemag.org/cgi/content/full/310/5745/96/DC1
Materials and Methods
References
Tables S1 and S2
Appendix S1
11 July 2005; accepted 31 August 2005
10.1126/science.1117199
Transoceanic Migration, Spatial
Dynamics, and Population
Linkages of White Sharks
Ramo
´
n Bonfil,
1
*
Michael Mey
¨
er,
2
Michael C. Scholl,
3
Ryan Johnson,
4
Shannon O’Brien,
1
Herman Oosthuizen,
2
Stephan Swanson,
2
Deon Kotze,
2
Michael Paterson
2
.
The large-scale spatial dynamics and population structure of marine top
predators are poorly known. We present electronic tag and photographic
identification data showing a complex suite of behavioral patterns in white
sharks. These inclu de coastal return migrations and the fastest known
transoceanic return migration among swimming fauna, which provide direct
evidence of a link between widely separated populations in South Africa and
Australia. Transoceanic return migration involved a return to the original capture
location, dives to depths of 980 meters, and the tolerance of water temperatures
as low as 3.4-C. These findings contradict previous ideas that female white sharks
do not make transoceanic migrations, and they suggest natal homing behavior.
Great white sharks (Carcharodon carcharias)
occupy the apex of most marine food webs in
which they occur. Their major centers of abun-
dance are in the coastal waters of California–
Baja California, Australia–New Zealand, South
Africa, and, formerly, the Mediterranean Sea
(1–3). Management and conservation of this
threatened species (4, 5) have been limited,
partly because its space utilization and mi-
grations and the linkages between popula-
tions were poorly understood and difficult to
research until the development of sophisticated
telemetry instruments and high-resolution ge-
netic markers for the species (6–9). Long be-
lieved to primarily be shelf inhabitants, white
sharks are now known to be more pelagic and
to travel from California to Hawaii (6). Males
are assumed to move between distant popula-
tions, whereas females have been assumed to
be nonroving and philopatric (9).
We tagged white sharks off the Western
Cape of South Africa between June 2002 and
November 2003 with pop-up archival satellite-
transmitting (PAT) tags (n 0 25), near-real-time
satellite tags (from here onward, Bsatellite
tags[)(n 0 7), and acoustic tags (n 0 25) in
order to study their spatial dynamics (table S1).
Using high-resolution photographic identifica-
tion techniques, we have recorded the daily
presence or absence of individual white sharks
off Gansbaai (34-39¶S, 019-24¶E; Western
Cape) since October 1997 (10).
Electronic tagging and photographic identi-
fication records reveal complex spatial dynam-
ics in white sharks, which we categorized into
four behavioral patterns: rapid transoceanic re-
turn migrations, frequent long-distance coastal
return migrations, smaller-scale patrolling, and
site fidelity. A white shark performed a previ-
ously unknown fast transoceanic return migration
spanning the entire Indian Ocean, swimming
coast-to-coast from South Africa to Australia
and back. This È380-cm total length (TL;
measured as a straight line from the tip of the
snout to the end of the upper caudal lobe)
female shark (number P12), PAT-tagged on 7
November 2003 off Gansbaai, traveled in 99
days to a location 2 km from shore and 37 km
south of the Exmouth Gulf in Western Aus-
tralia (22-01¶05µS, 113-53¶13µE; Fig. 1A).
This shark_scourseofÈ11,100 km (11)en-
tailed a counterclockwise displacement of more
than 750 km off the southern tip of Africa,
followed by a remarkably direct path toward
northwestern Australia, indicating that white
sharks do not need oceanic islands as gate-
ways for transocea nic migrations, as previ-
ously hypothesized (12). Shark P12 traveled
at a minimum speed of 4.7 km hour
j1
during
its migration to Australia (13), which is the
fastest sustained long-distance speed known
among sharks (14 –17 ) and comparable to
1
Wildlife Conservation Society, 2300 Southern Boulevard,
Bronx, NY, 10460, USA.
2
Marine and Coastal Manage-
ment Branch, Department of Environmental Affairs and
Tourism, Private Bag X2, Roggebaai 8012, Cape Town,
Western Cape, South Africa.
3
White Shark Trust, Post Of-
fice Box 1258, Strand Street 6, Gansbaai 7220, Western
Cape, South Africa; and Department of Zoology, Univer-
sity of Cape Town, Rondebosch 7700, Western Cape,
South Africa.
4
Department of Zoology and Entomology,
University of Pretoria, Pretoria 0002, South Africa.
*To whom corresponde nce should be ad dressed.
E-mail: rbonfil@wcs.org
.Present address: Sea Technology Services, Ground
Floor, Foretrust House, Martin Hammerschlag Way,
Cape Town, Western Cape, South Africa.
R EPORTS
7 OCTOBER 2005 VOL 310 SCIENCE www.sciencemag.org
100

th at of some of the fastest-swimming tunas
(18, 19). Records obtained through photo-
graphic identification revealed the return of
P12 from Australia back to its original tagging
site on 20 August 2004 (Fig. 2 and fig. S1),
evidencing site fidelity and an outstanding
navigational ability. Shark P12 performed the
fastest transoceanic return migration recorded
among marine fauna (14, 20), taking just
under 9 months to complete a circuit of more
than 20,000 km. Logged records from the
photographic identification study show that
P12 is a seasonal visitor (from June to De-
cember) to the Gansbaai area (table S2). It
has been recorded during 38 different days
spanning 1999–2004, suggesting that it is a
South African shark and that its transoceanic
return migration could be common. A second
PAT-tagged shark (unsexed, È200- to 230-cm
TL; number P3) traveled to an offshore lo-
cation 242 km SE of Port Elizabeth, where its
tag detached on 26 December 2003, in what
might have been the first leg of a migration
toward Australia (Fig. 1A).
Fig. 1. Transoceanic
migration of a white
shark from South Af-
rica to northwestern
Australia and possible
first leg of a second
transoceanic-migrating
shark. (A) Positions of
(dots) and track fol-
lowed by (black line)
shark P12 during coast-
al and transoceanic
movement; geolocation-
estimated positions
were corrected using
SST data to derive
positions shown (11).
The first leg of another
possible transoceanic
migration to Australia
(or an offshore move-
ment toward the north-
east coast of South
Africa) is shown by the
pop-up location of the
PAT tag from shark P3
(blue line and square).
Tagging and pop-up
dateswereasfollows:
for P12, 7 November
2003 and 28 February
2004; for P3, 14 April
2003 and 25 December
2003. SST is an average
composite at 4 km
resol utio n for daily
Moderate Resolution
Imaging Spectro-
radiometer data
from 23 November
2003to28February
2004. Southwest In-
dian Ridge shown as
white depth contours
(100 to 2000 m). The
scale bar represents
5000 km; the white
arrow marks the tag
deployment location.
(B) Differential time-
at-depth patterns dur-
ing the coastal and
oceanic legs of shark
P12’s trip, showing a
bimodal pattern with a
strong preference for
the depths of 0.0 to
0.5 m and 500 to 750 m
during transoceanic
travel. (C) Minimum
(black line and squares)
and maximum (bright
blue line) depths and
minimum temperature
(orange dots) visited during the coastal and oceanic phases of movement; all data are in 6-hour periods.
A
B
C
R EPORTS
www.sciencemag.org SCIENCE VOL 310 7 OCTOBER 2005
101

Transoceanic return migration is previously
unknown in white sharks and only suspected in
other chondrichthyans. Our results provide
direct evidence of a physical link between
two of the most important and widely sepa-
rated white shark populations, and they con-
firm philopatry in white sharks. They also
prove that female white sharks are capable of
transoceanic migrations and indicate that the
sex-biased dispersal of this species (9) is not
necessarily based on differences in the pro-
clivity of either sex to undertake transoceanic
migrations, but is probably attributable to
differences in how these migrants become
reproductively integrated into the Brecipient[
population. In light of our data, the transmis-
sion of nuclear, and not mitochondrial, genetic
material between South Africa and Australia
(9) could be explained if (i) both sexes make
transoceanic migrations, but only males repro-
duce in the recipient population, and/or (ii)
females make transoceanic migrations and
mate with males from the recipient popula-
tion, only to return to their original location to
give birth. Indeed, the migration of P12 from
South Africa to Australia corresponds to what
is thought to be the mating season in this re-
gion (21). An eventual return of this shark to
give birth in South Africa would prove natal
homing in white sharks, as has been sug-
gested for other shark species (22, 23), and
would support recent theories about the simi-
larity of reproductive strategies among a wide
range of marine taxa (24).
The mechanisms used by P12 to navigate to
Australia and back remain unknown; aside
from a few shallow seamounts on the South-
west Indian and Ninety East Ridges, there are
no other topographic features that could be
used for orientation on the route it followed
(Fig. 1A). We analyzed the satellite-transmitted
summary data to reveal the diving pattern of
P12 and found that during eastward trans-
oceanic migration, it made frequent deep dives,
reaching record maximum depths (980 m) (25),
experienced record ambient temperatures of
3.4-C, and spent 18% of the time at depths
of 500 to 750 m (Fig. 1, B and C). This shark
spent considerably more time (61%) just be-
low the surface (0.0 to 0.5 m) while in oceanic
waters than when in coastal waters (23%),
swimming most of the time (66%) above 5 m
during this trip. A strong preference for sur-
face swimming during oceanic travel is a be-
havioral pattern previously unreported in white
sharks (1, 2, 6, 26 ). We speculate that, like
many other vertebrates (14), white sharks could
be using visual stimuli such as celestial cues
as an important navigational mechanism in
addition to, or instead of, following gradients
in Earth_smagneticfieldasiscommonlyac-
cepted behavior for sharks (27).
Great white sharks undertake long-distance
return migrations along the South African coast
with relative frequency, as revealed by the track-
ing of satellite tags and by PAT tag pop-up
locations (Fig. 3 and fig. S2). They travel from
high-abundance sites in the Western Cape
(28, 29) to waters as far as 92000 km away off
kwaZulu-Natal and beyond, using underwater
routes along the continental shelf, then return
to their original tagging sites off the Western
Cape after 4 to 6 months. A 284-cm TL female
(S1) was fitted with a satellite tag in Mossel
Bay (34-08¶S, 22-07¶E) on 24 May 2003 and
completed the first tracked long-distance return
migration for a chondrichthyan, moving in
65 days to waters northeast of Delagoa Bay
(Mozambique) and outside the South African
Economic Exclusive Zone, where white sharks
are legally protected (Fig. 3). S1 returned to
Mossel Bay 162 days after being tagged, and
was photographed with its transmitter still at-
tached. Shark S2, a 310-cm TL female double-
tagged with satellite and acoustic tags in Mossel
Bay on 31 May 2003, was tracked for 46
days to the Tugela Bank, then recorded by our
acoustic-tag bottom monitors back in Mossel
Bay 123 days after being tagged (Fig. 3). In
total, 25% of tagged sharks that yielded in-
formation moved from the Western Cape to
kwaZulu-Natal and beyond, and 12.5% showed
return migrations (Fig. 3 and fig. S2). The high
proportion of immature white sharks (table S1,
Fig. 3, and fig. S2) moving to the rich en-
vironment of the Tugela Bank (30, 31) suggests
that these long-distance coastal return migra-
tions might be feeding-related events.
Records obtained from satellite and PAT tags
reflect additional spatial dynamics patterns in
white sharks, including smaller-scale patrolling
behavior and site fidelity (Fig. 3 and figs. S3 and
S4). These patterns and the return migrations
described above suggest a wider and more com-
plex range of behavioral patterns in white sharks
than was previously thought to exist. The dis-
covery of a trans–Indian Ocean return-migrating
white shark after a relatively low tagging effort,
in addition to its periodic absence from Gansbaai
as evidenced through photographic records, im-
plies that the Australian and South African pop-
Fig. 2. Photogr aphic identi fi-
cation records of shark P12 at
tagging (7 November 2003) and
upon return to the tagging loca-
tion at Gansbaai (20 A ugust
2004) after its transoceanic mi-
gration to Western Australia. (A)
Trailing edge of the first dorsal
fin, showing a unique notch
pattern allowing identification;
the white lines connect corre-
sponding notches in both photo-
graphs. (B) Right side of the first
dorsal fin, with magnified details
(left insets) showing a unique
black pigmentation pattern aid-
ing identification.
R EPORTS
7 OCTOBER 2005 VOL 310 SCIENCE www.sciencemag.org
102

ulations maintain a physical link within a single
generation and that this return migration might
be more common than is presently known.
Our studies show that we do not have a full
understanding of the ways in which identified
populations are connected. The movement of a
female to a region of Australia known for the
presence of Australian white sharks and its
return to South Africa, in conjunction with pre-
vious genetic studies, implies that earlier hypothe-
ses about sex-biased dispersal might need to be
modified. Males are currently considered to be
the ones who move between populations (9), but
our data suggest that the connectivity between
populations could be facilitat ed also or exclu-
sively by females. The return of females mating
in Australia to give birth in South Africa would
be consistent with genetic analyses; the finding
of a rare male of South African Borigin[ in
Australia (9) might reflect equally rare birthing
in Australia by South African females.
The discoveries presented here and our lack
of evidence of sex- or size-related patterns of
space utilization in white sharks underscore the
need for additional research. Multidisciplinary
studies integrating population genetic analyses
and electronic tagging, as well as the devel-
opment of improved monitoring instruments,
should be encouraged.
Long-distance and transoceanic migrations
expose great whites to increased risk of mor-
tality as they leave domestically protected
waters in South Africa/Australia and travel into
neighboring or remote countries, sometimes
located across entire ocean basins. An in-
creasing global demand for shark products
(32), coupled with our findings, suggests that
global protective measures, such as the recent
listing of the white shark in CITES Appendix
2 (CITES, Convention on International Trade
in Endangered Species of Wild Fauna and
Flora), are warranted to ensure the effective-
ness of local protective legislation currently in
place in a handful of countries.
References and Notes
1. A. P. Klimley, D. G. Ainley, Eds., Great White Sharks:
The Biology of Carcharodon carcharias (Academic
Press, San Diego, CA, 1996).
2. L. J. V. Compagno, Sharks of the World. An Annotated
and Illustrated Catalogue of Shark Species Known to
Date. Vol. 2. Bullhead, Mackerel and Carpet Sharks
(Heterodontiformes, Lamniformes and Orectolobiformes).
FAO Species Catalogue for Fishery Purposes No. 1 (Food
and Agriculture Organization of the United Nations,
Rome, 2001).
3. A. Soldo, I. Jardas, Periodicum Biologorum 104, 195
(2002).
4. L. J. V. Compagno, M. A. Marks, I. K. Fergusson,
Environ. Biol. Fish 50, 61 (1997).
5. C. Hilton-Taylor, Compiler, 2002 IUCN Red List of
Threatened Species (IUCN, Gland, Switzerland, 2000).
6. A. M. Boustany et al., Nature 415, 35 (2002).
7. B. A. Block, H. Dewar, C. Farwell, E. D. Prince, Proc.
Natl. Acad. Sci. U.S.A. 95, 9384 (1998).
8. R. L. Johnson et al., paper presented at the meeting
on Conservation Research of Great White Sharks,
New York, 20 to 22 January 2004 (Wildlife Conser-
vation Society, New York, 2004).
9. A. T. Pardini et al., Nature 412, 139 (2001).
10. Materials and methods are available as supporting
material on Science Online.
11. The positions estimated from archived light-level data
using geolocation algorithms provided by the manu-
facturer of the tags were corrected using satellite sea
surface temperature (SST) data with the method
described in the supporting online material.
12. G. Cliff, L. J. V. Compagno, M. J. Smale, R. P. Van Der
Elst, S. P. Wintner, S. Afr. J. Sci. 96, 365 (2000).
13. The widely accepted definition of migration is ‘‘the
act of moving from one spatial unit to another’’ (14).
This definition is general enough to be applicable to
all animal taxa, independently of spatiotemporal
scales, and includes at its core individual migration.
14. R. R. Baker, Migration: Paths Through Time and Space
(Hodder and Stoughton, London, 1982).
15. S. A. Eckert, B. S. Stewart, Environ. Biol. Fish 60, 299
(2001).
16. N. E. Kohler, P. A. Turner, Environ. Biol. Fish. 60, 191
(2001).
17. I. G. Priede, Fish. Res. 2, 201 (1984).
18. B. A. Block et al., Science 293, 1310 (2001).
19. T. Itoh, S. Tsuji, A. Nitta, Fish. Bull. (Wash. D.C.) 101,
514 (2003).
20. D. Inagake et al., Bull. Nat. Res. Inst. Far Seas Fish 38,
53 (2001).
21. M. P. Francis, in (1), pp. 157–172.
22. R. E. Hueter, M. R. Huepel, E. J. Heist, D. B. Keeney, J.
Northw. Atl. Fish. Sci. 35, 239 (2005).
23. K. A. Feldheim, S. H. Gruber, M. V. Ashley, Proc. R.
Soc. London Ser. B 269, 1655 (2002).
24. P. Cury, Can. J. Fish. Aquat. Sci. 51, 1664 (1994).
25. Previous reports of record diving depths of 1280 m
for white sharks (2) are based on the capture of one
specimen in a longline set at that depth; however, to
our knowledge, there is no evidence that the shark
was caught at 1280 m as opposed to anywhere else
along the water column.
26. F. G. Carey et al., Copeia 2, 254 (1982).
27. T. P. Quinn, Trends Ecol. Evol. 9, 277 (1994).
28. L. J. V. Compagno, D. A. Ebert, M. J. Smale, Guide to
the Sharks and Rays of Southern Africa (Struik, Cape
Town, South Africa, 1989).
29. R. L. Johnson, thesis, University of Pretoria, South
Africa (2003).
30. S. T. Fennessy, S. Afr. J. Mar. Sci. 14, 263 (1994).
31. S. T. Fennessy, S. Afr. J. Mar. Sci. 14, 287 (1994).
32. S. C. Clarke, thesis, University of London, UK (2003).
33. We thank the Natal Sharks Board and particularly S.
Dudley, G. Cliff, K. Cox, and W. Harrison for valuable
fieldwork assistance and helpful discussions in the
satellite tag study and S. Dudley for assistance in the
design and supervision of the acoustic tag study;
B. Mangold, C. Masterton, S. Parsons, P. Koen, D.
Woodborne, and P. Fre
´
on for the health maintenance
of sharks; R. and J. Portway, L. Staverees, D. Reynolds, T.
Keswick, and M. Rutzen for support and assistance with
fieldwork; L. Drapeau for Geographic Information Sys-
tems assistance; Smit Marine for maintaining bottom
receivers; M. N. Bester for supervision and D. Sadie for
conception of the acoustic tag study; and the Roe Foun-
dation, Wildlife Conservation Society, the South African
Government, International Fund for Animal Welfare,
World Wide Fund for Nature, and Professional Associ-
ation of Diving Instructors–Aware for financial support.
Supporting Online Material
www.sciencemag.org/cgi/content/full/310/5745/100/
DC1
Materials and Methods
SOM Text
Figs. S1 to Fig. S4
Tables S1 and S2
References
16 May 2005; accepted 21 July 2005
10.1126/science.1114898
Fig. 3. Northeastward long-distance return migrations of South African white sharks. The figure
shows the tracks of two satellite-tagged sharks showing long-distance return migrations and
crossing to Mozambique. Shark S1 (black trace) left Mossel Bay after tagging (24 May 2004); moved
rapidly to Bird Island, residing within a limited area (385 km
2
) for 27 days; and continued northeast
along the shelf edge, then in oceanic waters beyond the Agulhas Current, reaching Mozambique 65
days after tagging. Transmissions ceased 11 days later, to resume on Bird Island 62 days later, then
at the original tagging location on 2 November 2003. Shark S2 (white trace), tagged on 31 May 2003
with satellite and acoustic tags, traveled steadily along the coast to the Tugela Bank in 37 days,
where it ceased transmitting 9 days later and was recorded by acoustic bottom receivers back in
Mossel Bay on 1 October 2004. The red star indicates the tagging location; the dashed line indicates
projected movement during long periods without transmissions.
R EPORTS
www.sciencemag.org SCIENCE VOL 310 7 OCTOBER 2005
103