See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/227998438

Antarctic Peninsula killer whales (Orcinus orca) hunt seals and a

penguin on floating ice

Article in Marine Mammal Science · November 2007

DOI: 10.1111/j.1748-7692.2007.00163.x

CITATIONS READS

85 7,096

6 authors, including:

Ingrid Natasha Visser Thomas G Smith

67 PUBLICATIONS 1,216 CITATIONS

EMC Eco marine Corporation
98 PUBLICATIONS 6,117 CITATIONS
SEE PROFILE
SEE PROFILE

Santiago Imberti
Asociación Ambiente Sur
49 PUBLICATIONS 533 CITATIONS

SEE PROFILE

All content following this page was uploaded by Santiago Imberti on 16 December 2019.

The user has requested enhancement of the downloaded file.

MARINE MAMMAL SCIENCE, **(*): ***–*** (*** 2007)

C 2007 by the Society for Marine Mammalogy

DOI: 10.1111/j.1748-7692.2007.00163.x

Antarctic peninsula killer whales (Orcinus orca) hunt seals

and a penguin on floating ice
INGRID N. VISSER
Orca Research Trust,
P. O. Box 402043
Tutukaka, 0153, New Zealand
E-mail: [email protected]

THOMAS G. SMITH
E.M.C. Eco Marine Corporation,
5694 Camp Comfort Road,
Beaulac-Garthby, Quebec G0Y 1B0, Canada

IAN D. BULLOCK
Tegfan, Caerbwdi,
St. David’s, Pembrokeshire SA62 6QP, United Kingdom

GEOFFREY D. GREEN
12 Chemin Fosbery,
Chelsea, Quebec J9B 2G6, Canada

OLLE G. L. CARLSSON
Svanegatan,
Lund, 222 24 Sweden

SANTIAGO IMBERTI
Rivadavia 780,
(9400) Rio Gallegos, Argentina

In 1979, Smith et al. (1981) observed a group of seven killer whales (Orcinus orca)
conduct a coordinated attack on a crabeater seal (Lobodon carcinophaga) by creating
a wave to wash the hauled-out seal off floating ice. Since then a number of similar
events have been witnessed along the Antarctic Peninsula. The increased tourist ship
traffic in this area has provided an opportunity for many observers equipped with
cameras and video recorders to document a number of these.
Given that it is often difficult to gain an accurate description of events involving
a number of killer whales moving quickly and typically under water, in an ice-filled
environment, the availability of good quality video recordings is often the best way
of documenting such a fast-moving incident. The vantage point is of importance and
those observations from the bow of a large ship, which can maneuver to within a short
distance of the animals, are of significantly better quality than those at water level.
Still photography, in addition to video, is of particular importance in any attempt to
identify individual whales or groups.

1
2 MARINE MAMMAL SCIENCE, VOL. **, NO. **, 2007

Here we present six “events” (some involving multiple attacks) by killer whales
on prey that were hauled out on ice floes, one of which we describe in detail. Prey
kills were confirmed by the presence of body parts, blood or fat, either in the mouth
of a killer whale or in the water. Across all incidents we estimated the total size of
the ice floes by comparing them to the average known size of the seal; 2.5 m for
crabeater and Weddell (Leptonychotes weddellii) and 3 m for leopard (Hydrurga leptonyx)
seals (Nowak 1999). When an accurate estimate of the total floe size was difficult or
impossible, an approximate estimate of diameter is provided.
The sex/age class of the individual killer whales was determined based on relative
size (e.g., calves were defined as less than a third the size of adult females, juveniles
were approximately half the size of adult females). Adult males had dorsal fins that
were at least twice the size of adult female dorsal fins; however, given that it can be
difficult to determine non-sprouted sub-adult males (i.e., the dorsal fin had not yet
begun rapid pubescent growth in height) from females, all animals of this size were
classified together as adult female/sub-adult male (AF/SAM).
Of the six events described, five involved seals and one involved a penguin (Table 1).
All seals were adults and during each attack they moved to the approximate center of
the ice floes. One event involved two different species of seals (leopard and Weddell)
on two separate ice floes but within 10 m of each other. Of the three events involving
crabeater seals, there was one probable and two definite kills. Leopard seals were
involved twice; one was washed into the water and killed, the other remained on the
ice floe. One Weddell seal was attacked five times and dislodged from the ice floe
by waves induced by the killer whales each time, but no definite kill was confirmed
(Table 1). One lone adult Adélie penguin was attacked twice; however, it remained
on the ice floe (Table 1) and was not killed.
The killer whale group size ranged from five to seven individuals, with the excep-
tion of the event involving the Adélie penguin, where a single killer whale mounted
the attacks. The age/sex structure for the groups is given in Table 1.
Quantity of sea ice covering the surface is typically described as a fraction, where
10/10ths would be complete coverage. All of the events occurred in areas where there
was at least 2/10th ice cover. In all cases the prey was on a piece of first-year sea ice.
These ice floes varied in initial size from 4.5 to 18 m wide. In two cases (12 November
1979 and 14 January 2000) the ice was thicker than 1 m. In all the other cases the
ice had melted to less than <0.5 m thick, and the ice floes were very low to the water
surface, making them easily fractured by waves generated by the killer whales.
In four of the five events involving attacks on seals, the prey, after being dislodged
from the ice floe, either regained the original ice floe or, in three cases, moved to emerge
on another nearby ice floe. During two attacks the dislodged seals were carried in
the mouth of a killer whale but were either released, escaped, or were deposited, still
alive, on another ice floe.
On 15 January 2006, a coordinated event involving a crabeater seal in the Grand
Didier Channel at 65◦ 47 S, 065◦ 02 W, on the west side of the Antarctic Peninsula,
was observed. We summarize the most significant incidents in Table 2 and provide
further details here. We had a high and close vantage point on the bow of a tourist
ship, the Lindblad Expeditions MS National Geographic Endeavour. The seas were
 Table 1. Comparison of six attacks by killer whales made on three seal and one penguin species, on ice floes during the period 1979 to 2006.

Ice Ice type Initial Number Observer,

Approximate Killer cover and floe Number of times platform,
duration whale groupa Prey within thickness size of waves prey recording
Date (min) Location composition species 1 km2 (m) (m) observed dislodged Kill method
◦ 
12 November 17 64 53 S, 1 AM, 4 Crabeater 4/10 1st year 8 (diam) 1 1 Probable Smith et al.
1979 062◦ 53 W AF/SAM, seal (>1.5 m) (1981)
Paradise Bay 1 j, 1 c ship, photo
18 January 23 65◦ 15 S, 3 AF/SAM, Leopard 2–3/10 1st year 25 (diam) 6 5 Yes J. Drennan
1998 064◦ 16 W 2j seal (<0.5 m) zodiac, video
Argentine
Islands
14 January 36 65◦ 14 S, 2 AM, 2 Leopard 3–4/10 1st year 12 (diam) 0 0 No GDG, OGC, SI
2000 064◦ 10 W AF/SAM seal (>1.5 m) zodiac,

NOTES
Yalour Islands 1 j, 2 c photo, video
Weddell 16 × 5 6 5 No
seal
25 November 6 65◦ 10 S, 1 sprouting Adélie 5/10 1st year 5 × 2 3 0 No IDB land, visual
2004 064◦ 30 W M penguin (>1.0 m)
Penola Strait
25 December 45 64◦ 57 S, 1 AM, 3–4 Crabeater 2/10 1st year 5 × 6 6 3 Yes GDG, SI
2004 063◦ 22 W AF/SAM, seal (<1.0 m) ship, photo
S end Gerlache 2c
Strait
15 January 32 65◦ 47 S, 5 AF/SAM, Crabeater 2/10 1st year 15 × 10 11 3 Yes TGS, INV
2006 065◦ 02 W 1 j, 1 c seal (<0.5 m) ship, photo,
Grand Didier video
Chanel
a
AM = adult male, AF/SAM = adult female/subadult male, j = juvenile, and c = calf.

3
4 MARINE MAMMAL SCIENCE, VOL. **, NO. **, 2007

Table 2. Summary of hunting tactics used by a group of seven killer whales, Orcinus orca,
during a 28-min observation of an attack on an adult crabeater seal, Lobodon carcinophaga,
lying on a piece of first year sea ice (see Table 1, 15 January 2006, for location and group
composition details).

Elapsed Floe size

time (min) (meters) Behavior Result
+3.12 15 × 10 Killer whales submerge to create Broke into five pieces
Wave 1
3.56 Wave no. 2
4.10 5 wide Seal leaves broken floe and swims Seal on new floe
right, to new floe
4.12 25 × 15 Wave no. 3 Floe reduced by 30%
5.37 18 wide Wave no. 4 Floe reduced by 20%
6.36 16 wide Wave no. 5 Floe reduced by 20%
8.34 13 wide Wave no. 6 No floe reduction
9.24 12 × 13 Wave no. 7, washes only west floe No floe reduction
10.30 12 × 13 Wave no. 8 Floe reduced by 10%
10.30–11.30 Group actively steering floe into
open water and clearing
obstructing brash
11.40 11 × 12 Wave no. 9, washes across floe Floe reduced by 30%
12.00 7×8 Group steering floe into open
water
13.00 Killer whale clears adjacent floes Floe in open water
by surfacing and diving
between floes
14.56 Seal rolls onto back (submissive)
15.16 7×8 Wave no. 10, washes across floe Seal washed off far
side of floe
15.36 Killer whale seen with seal in
mouth
18.43 6 wide Seal reappears on new floe
19.36 Wave no. 11 Floe reduced by 50%
19.42 Seal back in the water
20.48 4 wide Seal hauling out onto floe, but Killer whale takes seal
dragged back in by hind
flippers
25.32 One killer whale with seal body Kill confirmed
part in mouth
28.04 Group moving out of area

calm with 2/10th ice cover consisting mostly of first-year ice floes of varying size, all
<0.5 m thick and in an advanced state of melt. At approximately 2330 we sighted
killer whales spyhopping beside an ice floe, on which an adult male crabeater seal was
lying. The floe was approximately 15 × 10 m and in an advanced state of melt. We
watched and filmed for approximately 30 min as the group of killer whales, consisting
of five AF/SAM, one juvenile, and one calf, mounted a number of coordinated attacks,
dislodged the seal from four different ice floes, and ultimately killed it. The ship was
approximately 1 km distant when the killer whales were first sighted but gradually
closed to less than 100 m without any noticeable reaction to the ship by either the
 NOTES 5

whales or the seal. The event was recorded onto high definition video (HDV) and we
refer to timings as elapsed time, e.g., +12.30 min.
At +3.12 min, after a period of spyhopping, bubble blowing, and moving around
the ice floe on which the seal was hauled out, two AF/SAM killer whales submerged
suddenly. They were approximately 15 m away from the ice floe and simultaneously
swam directly at and under it, creating a significant wave (hereinafter referred to as
wave-washing), which tipped the ice floe and washed over it. This did not dislodge
the seal but broke the ice floe into five pieces (as illustrated in Fig. 1A), leaving the
seal on a smaller fragment less than 5 × 5 m. Shortly afterwards (+4.10 min) the
seal went into the water of its own accord and swam quickly to emerge on a larger
floe (25 × 15 m), which was approximately 100 m away from the seal’s original
platform.
From +5.37 to +14.56 min the killer whales created a further six wave-washes,
none of which dislodged the seal but sequentially reduced the ice floe to approximately
7 × 8 m. During this period the killer whales were also seen to actively move the ice
floe into an area of open water free of adjacent ice floes or ice debris. It is apparent
from the video footage that the whales were sometimes actively pushing the ice
floe with their bodies, particularly their rostrums, although it appears that ice floe
movement was also created by individual whales submerging and creating vortices
at the very edge of the ice floe (as illustrated in Fig. 1B, C). The whales also blew
bubbles underwater near the edge of the ice creating turbulence, which was seen
to clear ice debris away from the edge. On one occasion a killer whale was clearly
seen to separate the target ice floe from an adjacent pan. It swam fast, submerged
between the ice floes, thereby making a wave that moved them apart. At the end of
this maneuver, it surfaced and then submerged on the far side of the obstructing ice
floe, which created a vortex and then pulled it away further (Fig. 1C).
Near the end of this period it was evident that the crabeater seal was severely
agitated. Its chest was heaving, its jaws were opening and closing, and it eventually
turned completely on its back.
At +14.58 min, one AF/SAM killer whale remained in position with its rostrum
against the ice floe while four AF/SAM killer whales moved away from the ice floe
with the seal on it. These four killer whales reappeared simultaneously, approximately
20 s later in line-abreast with all submerged just under the surface. All four were
coordinated-swimming, with their left sides orientated towards the surface. A trail of
bubbles emanated from each of the animals blowholes as they accelerated and passed
directly under the ice floe, two on each side of the stationary killer whale (Fig. 2).
This generated a large wave, which tipped the ice floe initially towards the wave,
then as the wave poured over and crested under the ice, it pivoted and tilted the ice
in the other direction where the attacking whales were now waiting (as illustrated
in Fig. 1D). The breaking wave washed the seal into the water at +15.19 min. A
short segment of HDV can be viewed at http://www.orcaresearch.org/dvd. The seal
was then seen to be jerked down vertically, presumably grabbed by one of the killer
whales, submerged in position on the far side of the ice floe. At +15.37 min the
seal was lifted out of the water in the mouth of an AF/SAM killer whale. Given the
short time frame (i.e., 18 s) between when the seal entered the water and when it
6 MARINE MAMMAL SCIENCE, VOL. **, NO. **, 2007

Figure 1. The use of waves by killer whales for (A) reducing ice floe size, (B) clearing the
ice around the floe, (C) moving the target floe into open water, and (D) wave washing the prey
into the ocean. (Illustrations by IDB).
 NOTES 7

Figure 1. Continued.

was observed in the mouth of the killer whale, we assume this was the same seal. At
+15.39 min the seal, which was still in the mouth of the killer whale, was taken
underwater.
At +18.4 min, the seal egressed onto an ice floe approximately 6 m wide, and less
than a body length from the whales yet several hundred meters from the original
ice floe. At +19.42 min the whales created a wave that again washed the seal into
the ocean (this now being the 11th and last wave observed during the event). At
8 MARINE MAMMAL SCIENCE, VOL. **, NO. **, 2007

Figure 2. Four submerged killer whales swim two on either side of a stationary killer
whale (blowing a bubble burst). They created a wave, which washed the adult crabeater seal
off the ice floe, in Grand Didier Channel, W. Antarctic Peninsula on 15 January 2006. (Photo
from video, by INV).

+20.48 min, the seal attempted to climb onto a third ice floe but was pulled back
into the water by an AF/SAM killer whale, which had grabbed it by the hind flippers.
At +25.32 min, parts of the body of the seal were seen in the mouth of an AF/SAM
killer whale, and we could confirm that the seal had been killed. Observations ended
at +28.04 min as the group of killer whales moved slowly away from the ship.
We were afforded a unique opportunity to observe the details of coordinated killer
whale hunting. After the initial period of spyhopping and looking at the seal, the killer
whales engaged in two types of wave-creating behavior as well as bubble-blowing,
which appeared to have three separate ends, ice movement, ice fragmentation and ice
tipping. The frequent submergence of individual whales at the very edge of the ice floe
began to actively move it in a concerted direction. These waves did not break over the
ice floe but instead created sufficient turbulence to drive it in a consistent direction,
sometimes straight, sometimes rotating it. This behavior as well as bubble-blowing
was also used to move ice debris or adjacent ice floes clear of the ice floe carrying
the seal. Larger breaking waves created by killer whales swimming towards and then
under the ice floe resulted in breaking up the fragile melting ice and reducing it
significantly in size.
Once the floe was a smaller size and had been steered into open water, a coordinated
attack by five AF/SAM killer whales (four swimming while one remained in position)
created a breaking wave that tipped the ice floe (resulting in the seal being displaced).
As the four killer whales approached the ice, streams of bubbles were seen emanating
from their blowholes. It is possible, but we have no direct evidence, that they were
emitting whistles while making these bubble-streams. Similar bubble-streams have
been associated with whistles for bottlenose dolphins, Tursiops truncatus (Caldwell
and Caldwell 1972). It cannot be ruled out that whistles may help coordinate the
animals during such an attack, yet we could not determine the role of the stationary
killer whale.
 NOTES 9

As the four killer whales approached the ice floe they all swam on their sides
(Fig. 2), possibly to protect their dorsal fins and enable them to pass very close to the
underside of the ice floe. By swimming on their sides they augmented the plane of
their bodies, which deflected the water, and possibly increased the size of the wave
they were creating. It appears that the group had not only learned to place the ice
floe in open water where the chances of the seal escaping were minimized, but they
could enhance their success by reducing ice floe size, using the waves to dislodge
their prey, and also increase the size of the waves to maximize their effect. In the
December 2004 event at least two (probably three) killer whales were also involved
in coordinated wave making, and they too swam on their sides.
In several of our documented attacks the killer whales did not kill the prey imme-
diately. For at least three of the five events involving seals, the seal was washed into
the water, captured, and carried in the mouth of an adult whale, after which it was
either released, escaped, or was deposited alive onto another ice floe. Such manipula-
tion of the prey (intentional or not) might serve in part as training, social learning,
or coaching for the young animals—see full definitions of each in Caro and Hauser
(1992). The immediate cost to the teacher (i.e., potential loss of the prey cf. to imme-
diately killing and consuming prey) may have long-term benefits if the naı̈ve pupil(s)
can learn and therefore later participate in the coordinated hunt (Caro and Hauser
1992). Complex hunting strategies, such as those described here, necessarily involve a
lengthy apprenticeship for the young group members. Such training could manifest
itself through imitation or assisted practice as has been observed for killer whales
(e.g., see Guinet 1991, and references therein). However, Baird and Dill (1995) found
that, for “transient” killer whales off the west coast of North America, prolonged prey
handling time (i.e., from catch until consume) did not necessarily represent training
opportunities for younger whales, but it was characterized by behaviors typical of
social-play behavior.
The attack on a penguin by a single juvenile killer whale is the only penguin
attack using wave-washing methods reported in the many years during which we
have recorded attacks on seals in this area.
Where possible, photo-identification quality images were obtained of the killer
whales. These were compared to the Antarctic Killer Whale Identification Catalogue
(www.akwic.org). This catalogue contains over 100 photographically identified
Antarctic killer whales. Perhaps due to the limited number of images obtained
from each event listed here, only one adult male has been resighted (from the
14 January 2000 encounter to a subsequent record on 23 December 2001, where
no hunting was observed), approximately 64 km to the south of the observation
recorded here (Visser, unpublished data). However, despite only one match to the
catalogue, we can still determine that all of the killer whales that have been in-
volved in the seal attacks that we have documented here are Type B killer whales,
as described by Pitman and Ensor (2003), in that they are gray and white and
have large eye patches, oriented parallel to the body axis. The Type B killer whales
are thought to specialize in hunting for pinnipeds and the data presented here re-
flect that. We could not ascertain the type of killer whale involved in the penguin
attack.
 10 MARINE MAMMAL SCIENCE, VOL. **, NO. **, 2007

Observations of killer whale predation and prey handling are rarely documented in
the Antarctic habitat. With such a small sample size it is difficult to draw strong con-
clusions. Nonetheless, we speculate that such complex coordinated hunting behaviors
are culturally transmitted (i.e., passed down [or carried forward] over generations).
Although this wave-washing hunting method may be novel to this area (all obser-
vations were within 61 km, if distances are measured in a straight-line), we cannot
exclude the possibility of its existence elsewhere because of the paucity of observa-
tions in the greater Antarctic area. We anticipate that this noticeable behavior will
be seen more frequently as Antarctic ship traffic increases in the future, and this will
hopefully lead to a more accurate assessment of the extent of this behavior among
killer whales.

ACKNOWLEDGMENTS
A number of people generously contributed documentation, which aided greatly in the
analysis of these events. We thank Cameron Kepler for his detailed notes, Joe Drennan for
his 1998 video record, Juliet Shrimpton for her photographs, and Lindblad Expeditions for
making the additional January 2006 video record available, which was filmed by Erika Jannson.
Terry Hardie assisted with frame-grabs from the video footage by INV. The officers and
expedition staff aboard the Lindblad Expeditions’ MS National Geographic Endeavour, as always,
did everything possible to afford us the best view of the whales. We are particularly grateful
to Captain Leif Skog and his deck officers for their enthusiasm, skillful ship maneuvering, and
respectful approach to the wildlife that we were observing. Trey Byus and Larry Prussin of
Lindblad Expeditions generously facilitated the scientific reporting. James Estes, Joel Berger,
Jo Berghan, Kate Norton, and two anonymous reviewers suggested constructive comments
towards improving the manuscript. INV’s research is supported through the Orca Research
Trust, Cetacean Society International, and the Whale and Dolphin Conservation Society.

LITERATURE CITED
BAIRD, R. W., AND L. M. DILL. 1995. Occurrence and behaviour of transient killer whales:
Seasonal and pod-specific variability, foraging behaviour, and prey handling. Canadian
Journal of Zoology 73:1300–1311.
CALDWELL, D. K., AND M. C. CALDWELL. 1972. The world of the bottlenose dolphin, Philadel-
phia, J.B. Lipincott C., 157 pp.
CARO, T. M., AND M. HAUSER. 1992. Is there teaching in nonhuman animals? The Quarterly
Review of Biology 67:151–174.
GUINET, C. 1991. Intentional stranding apprenticeship and social play in killer whales (Orcinus
orca). Canadian Journal of Zoology 69:2712–2716.
NOWAK, R. M. 1999. Walker’s mammals of the world. The John Hopkins University Press,
Baltimore, MD.
PITMAN, R. L., AND P. ENSOR. 2003. Three forms of killer whales (Orcinus orca) in Antarctic
waters. Journal of Cetacean Research and Management 5:131–139.
SMITH, T. G., D. B. SINIFF, R. REICHLE AND S. STONE. 1981. Coordinated behavior of killer
whales, Orcinus orca, hunting a crabeater seal, Lobodon carcinophagus. Canadian Journal of
Zoology 59:1185–1189.
Received: 29 January 2007
Accepted: 30 May 2007

View publication stats