KINEMATIC ANALYSIS OF THE LOCOMOTION OF THE
POLAR BEAR (URSUS MARITIMUS, PHIPPS, 1774) IN
NATURAL AND EXPERIMENTAL CONDITIONS
S. RENOUS, J.-P. GASC and A. ABOURACHID
Laboratoire d'Anatomie Comparée,
d'Histoire Naturelle, 55 rue Buffon,
75005 Paris, France)
The striking ability of the polar bear to travel on ice or frozen snow is tentatively
features involved in the locomotor behaviour of the animal. A comparison
with the brown bear shows the specific
in gaits, leg movement and in ground contact
structures. It is suggested
that these specific
in polar environment.
During slow gaits, polar bear hind limbs are maximally extended. The legs are able to
resist the transfer of mass during the contralateral limb swing phase. This results in a walk
with swaying hips. The polar bear uses transverse gallop to improve stability, whereas the
brown bear uses rotary gallop. The polar bear is comfortable on slippery
the brown bear is reluctant to move on it.
Proximodistal alternation of pads and large zones with hair constitute the main charac-
teristics of the plantar and palmar soles of the polar bear. These features may constitute a
for the drainage of water from the feet, the reinforcing
and an increase in the area of contact (snowshoe). The drainage is produced by two kinds
of structures: the superficial
network of the epidermis
of the pads and the hair between the
pads. These hirsute zones absorb the liquid which is drained off the pads by the animal's
weight during the stance phase. The hairs are also present in the regions of the soles where
thrusts are transmitted to the ground.
polar bear, Ursus maritimus,
brown bear, Ursus arctos.
The polar bear shows a spectacular ability to move with different gaits on
snow, a substrate with weak cohesiveness, and ice, a substrate with a low
friction coefficient. The ability to run on ice suggests a control of dynamic
balance due to limb coordination and efficient morphofunctional systems
to match the specific characteristics of the substrate. Few papers deal with
polar bear locomotion. According to HARINGTON
(1970), and VAN WORMER
(1966), Ursus maritimus is able to walk, gallop and perhaps pace, but it
never trots. DAGG (1973), in her review of mammalian gaits, emphasizes
the confusing data concerning bears. In the polar bear, a good grip on the
ice may be explained by the organisation of locomotor cycle of fore and
hind limbs in slow and fast gaits. The structure of the palmar and plantar
surfaces may also contribute to the ability to move on snow and ice.
For this reason, the present paper analyzes the action of the locomotor
system of the polar bear on several substrates. The analysis is based upon
films taken in natural conditions and a set of experiments performed on
zoo animals. A comparison with the brown bear, Ursus arctos, allows the
interpretation of the specific locomotor characteristics of the polar bear.
MATERIALS AND METHODS
Two polar bears (Ursus maritimus), a male of 900 kg and a female of
approximately 500 kg, were filmed at the zoo of Vincennes (National Mu-
seum of Natural History) using a Sony HI8 video camera (25 frames/sec).
These kinematic data were complemented by other films of Ursus mar-
itimus (producer: Rémi Marion) showing locomotion under natural condi-
tions at Churchill (Manitoba, Canada) and various parts of the polar region
(Canal + Video: l'ours blanc). In order to visualize movements of the
different skeletal elements of the limbs during locomotion, we used artic-
ulated skeletons and single bones (Museum national d'Histoire naturelle
(M.N.H.N.), Comparative Anatomy collections). Plaster casts of the palmar
and plantar surfaces of a polar bear were obtained from the Laboratory of
Zoology (Mammals and Birds) of M.N.H.N. The footprints of two brown
bears (Ursus arctos; 700 kg) were taken in Vincennes Zoo for comparison.
The analysis of the gaits is based upon the record of periodic events: the
duration of one complete motion cycle (i.e., stride) starting with the place-
ment of a reference limb (often a hind limb) on the ground, the duty factor,
time lags between diagonal or homolateral limbs, fore and hind leads, the
leading foot being the second of a pair touching down in each couplet
Film sequences of locomotion on soft or frozen snow, ice (in the field) or
concrete surfaces at the zoo were selected to analyze the gaits. In the se-
quences selected, the bears moved at steady speed (measured by the regular
displacement of the hip during approximately ten cycles) and perpendicular
to the direction of the camera to ensure a lateral view. The duration of
the total limb cycle, the stance phase, and the time lag between footfalls
of diagonal limbs (for example: right fore limb and left hind limbs RF-LH
and left front and right hind limb LF-RH) and between homolateral limbs
(for example: RF-RH and LF-LH) were measured to specify the type of co-
ordination. Each contact lasts the same fraction of the cycle and the cycles
showing large differences in limb contact duration were eliminated. The
characteristics of the gaits were finally identified according to HILDEBRAND
(1966, 1976, 1977).
As most of the proximal parts of the limbs are covered by the fur, the
estimation of the anatomical position of the joints was based upon the
observation of the bulbing of the large muscles in moving animals. The
successive articular joints of the fore and hind limbs were identified in
sequences showing lateral, front or back views of the polar bears. The
coordinates of the joints and the angular variations were analyzed with a
digitizing video system (Optimas software).
Footprints of the polar and brown bears were obtained at the zoo on
various substrates. To reach the outside area from their cage, the animals
had to pass through two rooms separated by narrow doors. The floor of the
first room was covered by a mat that was impregnated with paint. Footprints
were recorded on various substrates (sheets of smooth and very strong paper,
sheets of absorbent and thick paper, perspex) that were firmly attached to the
floor of the second room. Sony HI8 video cameras using 25 frames/sec were
fixed into the rooms outside of the reach of the animals. The experiments
were repeated at regular intervals to accustom the bears to the experimental
set up and to obtain a continuous gait without trampling on the substrate.
In all mammals (JENKINS, 1971; FISCHER, 1992; ROCHA-BARBOSA et al.,
1996a, 1996b), each limb has a periodic motion. Cineradiographical analy-
sis of the locomotion of the guinea pig clearly shows the basic component
of the mammalian locomotor cycle (ROCHA-BARBOSA et al., 1996; fig. 1).
Each cycle consists of a stance and a swing phase (GOSLOW et al., 1981).
The touchdown is followed by a yield subphase of the stance (E2) charac-
terized by the activity of extensor muscles (ORSAL, 1987), which act against
gravity and inertial effects. An elastic storage of energy may occur at this
moment in muscular and tendinous elements. The propulsive subphase (E3)
starts when the proximal element of the articular chain, transmitting part
of the body mass, passes over the point of support. An extension of the
limb, resulting from the action of the extensor muscles, causes the thrust
against the ground to create the propulsive force. When the limb takes off
the ground (F) at the beginning of the swing, it is accelerated forward and
flexed by the action of the flexor muscles. It is extended and decelerated
(El) by the action of the extensor muscles to contact the ground again for
a new cycle.
Polar bears observed at the zoo or in the wild used slow, moderate or fast
walks and slow run and pace (fig. 2) as defined by HILDEBRAND
leg sequence was always lateral, the fore footfall following the ipsilateral
hind footfall (fig. 3). The gait used in the zoo was always moderate pacing
walk. In the films analysed, polar bears moved on hard snow or on fine
snow using all the gaits, from slow walk to running pace. On ice, they
used moderate walk, which was also observed on soft snow when the snow
sank beneath their feet. An unusual locomotor behaviour was observed on
ice: the hindfeet were used as in bipedal locomotion propeling the body
forward, while the forefeet slided on the ice. A very long sequence of a
young, showed a cub playing with an ice ball with the forefeet, using only
the hindfeet for locomotion.
To analyse the movement of the fore and hind limbs, we chose a cycle of
fast walk in lateral view (fig. 4A and B; duty factor: 0.63; interval between
the fore footfall and thc ipsilateral hind footfall in percent of the cycle:
13%) lasting approximatively 1.48 second and beginning with the step of
the left forelimb. The skeleton segments were superimposed on the outlines
of the limbs.
The stance began with a special subphase (FE) lasting 20% of the support
phase, which was characterized by a gradual extension of the limb. This
was followed by a short yield subphase (E2) 0.24 s, 25% of the stance)
occuring after the shoulder or the hip passed vertically over the hand or foot
contacts. The yield subphase was marked by little flexion of the shoulder,
the elbow and the wrist joints in the fore limb and by a larger flexion of
the knee and the ankle joints in the hind limb. The hip was more or less
stable. The beginning of the yield of a limb corresponded to the moment of
the take-off of the contralateral limb. Consequently, at the beginning of the
swing phase of one limb a part of the body mass was transferred to the other
limb, which was in extended position, its skeletal elements constituting a
Fig. 1. Locomotor
cycle. A: typical phases of the hind limb locomotor
cycle of a guinea pig.
E2 and E3 (stance), F and El (swing) during a trot, with corresponding general movement
of the leg (1, 2, 3 and 4). The minimal value of the sum (in rad) of the main angles of
the chain, ilio femoral (a), femoro tibial (b) and tibio tarsal (c), define the limits of both
phases of stance and of swing. B: a comparison
with the same cycle of a walking polar bear.
of full extension.
Fig. 2. Gait diagrams of the polar bear. A: slow walk, B: medium slow walk, C: fast walk.
Stance phase is indicated by a thick line and swing phase by the absence of a line. Time is
given in percent of the reference cycle (LH). LF, left fore limb; LH, left hind limb; RF, right
fore limb; RH, right hind limb. -
Fig. 3. Graph of distribution
of symmetrical gaits of polar bears on different substrates sho-
wing the use of a lateral sequence (from HILDEBRAND,
vertical column. The propulsion subphase (E3) lasted approximately 55% of
the stance and was characterized by an extension of the wrist joint, whereas
the shoulder joint was flexed. The elbow was locked at maximal extension.
The hind limb showed an extension of the hip and ankle joints whereas the
knee was flexed (fig. 4).
The swing of the fore limb began by flexion of all joints and raising
the limbs above the ground due to the simultaneous action of the flexor
muscles (F). This flexion was accompanied by a forward acceleration of
the limb. Maximal flexion of the leg coincides with maximal ventroflexion
of the hand. For the hind limb, the knee was maximally flexed at take
off and was extended afterward (fig. 4, frame 10, 11 and 1 compared to
frames 2 and 3). The foot was elevated a little above the ground by a
moderate flexion of the hip joint and a large flexion of the ankle. The end
of the flexion subphase of the swing was difficult to specify. Considering
the locked extension of the knee at 180° as a criterion for the onset of the
extension subphase (El), the flexion of the hind limb lasted about 70% of
the swing phase. Extension of the fore limb was performed by extension
of all three main joints. Extension of the hind limb involved a stable ankle,
the maximal extension of the knee joint and some extension of the hip joint.
A rear view of the polar bear during fast walk (fig. 5) revealed that during
the swing phase the femur was endorotated moving the flexed knee inward
and the foot outward. Then the knee was extended and exorotation of the
femur moved the foot medially. The foot was then prepared to land with
its axis parallel to the body axis (fig. 5).
A front view of the polar bear (fig. 6) suggests that the humerus was
rotated inward at the end of the stance phase, the elbow being brought
laterally, the wrist outward and the fingers inward. This rotation continued
during the flexion subphase of the swing. A rolling up movement of the
hand associated with ventroflexion was observed. The extension of the
swing was characterized by protraction and outer rotation of the humerus,
resulting in limb adduction. Just before the touch down, the palmar surface
was oriented medially. Then, the hand contacted the ground with its outer
border and total contact of the palmar surface was achieved by outward
rotation of the wrist.
Captivity was not a favourable condition to observe running polar bears.
In the wild, polar bears may gallop on any kind of substrate. When they
accelerated to jump over an obstacle, they used only one stride of gallop.
This gait followed a transverse sequence, with one diagonal pair of feet
moving almost synchronously, while the other diagonal pair moved inde-
pendently (fig. 7). A flight period, which occured after the stance of a fore
limb (the right for all the cycles studied), corresponded to a general flexion
(fig. 7, frames 17 and 18) of the four limbs, named flexed suspension by
DAGG & DE Vos (1968), crossed flight by GAMBARYAN
(1974), and gath-
ered suspension by HILDEBRAND, (1977). It occured at the end of the gallop
cycle, before the touch down of a hind limb (left in all cases) which began
the next cycle. This flight was very short (about 8% of the cycle studied).
In the observed transverse gallop of the polar bear, limbs succession was
caracterized by: left hind, right hind, left fore and right fore limbs, the
fore and hind leads being on the same side of the body. In the example
Fig. 4. Slow walk of the polar bear. A: lateral view of a sequence. The direction of the
arrows under the right fore and hind limbs indicates the stance (downward)
and the swing
(upward) phases. B: corresponding gait diagram. The stages of the progression (1 to 11)
are indicated on the diagram. LF, left fore limb; LH, left hind limb; RF, right fore limb;
RH, right hind limb; El, extension subphase of the swing; E2, yield period; E3, propulsion
period of the swing; FE, full extension
during a preliminary subphase
illustrated in fig. 7, the duty factor was 35, the time lag in the asynchronous
diagonal (LH-RF) represented 55% of the cycle and only 3.75% in the other
diagonal (RH-LF). The hind and the fore leads were 24.3 and 27% of the
cycle, respectively. The total hind and fore limb contact durations were
respectively 62.1 and 64.8% of the cycle.
No good lateral views of the animals during a gallop were available. On
the 3/4 views, only one side of the bear could be studied, RH for the hinds
limbs and RF for the fore limbs (the leading). The hind limb behaved as
usual: footfall with the knee flexed followed by flexion of the different
limb joints. E3, which approximately began when the hip passed straight
over the foot contact (fig. 7, frames 6 and 7), corresponded to extension
of the joints. This indicated participation of the hip, knee and ankle in the
propulsive action of the limb. The knee and ankle joints flexed during F
(fig. 7, frames 17 and 18) before extension during El (fig. 6, frames 19 and
20) contributing to an increase in stride length.
The fore limb was almost completely extended at touch down. E2 seemed
to be very short (fig. 7, frames 10 and 11) and E3 long (fig. 7, frames 12
to 16). The E3 subphase suggested a movement of the scapula creating
an extension of the shoulder joint. This movement was concomitant to a
maximal extension of the elbow and an extension of the wrist. The right
fore limb was the only limb which made contact during the last part of
the gallop cycle, before suspension (fig. 7A). The subphase E3 of the right
fore limb followed the synchronous propulsive action of the diagonal limbs
RH-LF. The action of this limb may increase the efficiency and regularity of
propulsion and also contribute to the stability of such large animals (fig. 7,
frames 14, 15 and 16).
The study of all the sequences of galloping showed a relation between the
duration of fore and hind limb contacts with speed. Whatever the velocity,
fore and hind contacts represented more or less the same percentage of
the cycle (fig. 8A). For low and medium velocities, the fore contact lasted
approximately 60% of the hind contact, indicating that the leading limb
was only briefly in contact with the ground. At higher velocities, the fore
contact lasted 140% of the hind contact, indicating a relative lengthening of
the time the leading limb is in contact with the ground. In these gallops, the
unsupported interval in gathered position was always short and represented
between 0 and 20% of the cycle duration (fig. 8B). Sometimes there was
no unsupported period and the gathered limb position coincided with the
contact of one or two limbs, as is generally the case for heavy animals with
Fig. 5. Rear view of a sequence of slow walk of the polar bear (12 frames). The direction
of the arrows under the feet indicates the stance (downward)
and the swing (upward) phases.
White arrows: right foot, black arrows: left foot.
a limited flexibility of the vertebral axis (HILDEBRAND, 1977). The hind
and fore leads tended to increase when the duty factor decreased for each
limb, in relation with the increasing speed (fig. 8C).
Structure of the contact areas and function
The polar bear has a plantigrade stance. Studies of the hand and foot of a
cadaver and of plaster casts of palmar and plantar surfaces provided informa-
tion on their structure (snowshoe-shaped, alternation of different shapes).
The hand (fig. 9) showed two regions with resistant but elastic surfaces
formed by local thickening of the subepidermal tissues: dermal reticulum
and adipose panniculus. Proximally, a large semilunar pad was transversally
spread. Its inner part (close to the first finger) was well developed, whereas
its outer part (towards the fifth finger) was narrower. Within this large pad
five separate but confluent pads could be distinguished separated by weak
depressions. Distally, a second area corresponded to the line of five digital
pads. Digital and palmar pads were separated by a deep depression occu-
pied by a short thick brush-like fur. Proximal to the palmar pad, hairs were
radially set from a central pole, the shortest oriented forward covering the
posterior margin of the palmar pad. Very long hairs constituted a crown
around the hand, increasing the contact area with the ground. They passed
between the digital pads, covering their inner and outer margins. The skin
linked the fingers along half of their length forming a web under these hairs.
The hand did not possess carpal pads. The sole of the foot was longer and
narrower (W/L = 0.05) than the sole of the hand. The foot was di-
vided into six regions: 3 lines of pads and 3 fur areas. The first proximally
located pad was ovoid and corresponded to the heel. The second was a large
plantar semilunar pad, transversally spread, formed by five fused units. Its
outer part (fifth toe) was better developed than its inner part (first toe). The
third region was formed by a line of five digital pads. A triangular area of
short thick brush-like hairs separated the heel and the plantar pads. A deep
depression occupied by long hairs separated the plantar and the digital pads.
Long hairs constitute a crown around the foot sole surface. The surface of
the foot and hand pads has a granular appearance due to the outgrowth of
the outermost keratinized epidermal cells.
The footprints of palmar and plantar soles showed the same features on
any substrate. The prints of the palmar pad of the hand, and the short hairs
on its outline, suggest that this pad exerted the main contact during the
Fig. 6. Front view of a sequence
of slow walk of the polar bear ( 10
frames). The direction of
the arrows under the hands indicates the stance (downward)
and the swing (upward) phases.
White arrows: left hand, black arrows: right hand.
first footfall. The digital pads were also well delimited, except for the first
finger, which was not always visible. The fur located behind the palmar
pad, first contacting the ground, showed a circular print. This print, visible
during a slow walk, changed in shape during a fast walk and included the
region directly behind the palmar pad. The dense fur between the digital
and palmar pads was saturated with paint. Streams of paint oriented in the
direction of the movement were visible between the digital pads. Other
prints revealed a draining of the liquid by the long peripheral hairs towards
the body axis. The prints of the hind paw showed differences on four points.
The posterior part of the sole always indicated a heel contact area that was
larger than the area of the pad and which included the medial area of short
hairs. The metatarsal pads constituting the plantar pad showed five distinct
centers of pressure. The area between the plantar pad and the heel acted as
a sponge saturated with paint. The liquid was drained mainly in a backward
direction by the long peripheral hairs.
Comparison with brown bears
A comparison with the gallop of the brown bear (GAMBARYAN, 1974) was
made by aligning similar kinetic events during a cycle of both species
(fig. 10). The brown bear has a rotary gallop, with the fore and hind leads
on opposite sides (LH, RH, RF, LF). As in the polar bear, the gathered
suspension was produced by the thrust on the fore limbs. The flight starts
after the left fore limb stance for the brown bear, but after the right fore
limb stance for the polar bear. The two diagonals of the brown bear were
separated, while they were more synchronous in the polar bear. The fore
limb stride was longer than the hind limb stride in the brown bear.
There are marked differences between the foot pads of both species. Com-
pared with the plantar and palmar soles of the polar bear (fig. 11 ) the pads
of the brown bear were larger and the hairy regions were reduced: carpal
pads were present and on the foot, the plantar pad was much larger. Unlike
polar bears, the claws of the brown bear are in contact with the ground.
The hand prints of the brown bear suggested that the main contact was
provided by the palmar and the digital pads, because their surface was
Fig. 7. Gallop cycle of polar bear. A: 3/4 view
of a gallop
cycle of a polar bear. The direction
of the arrows under the feet and hands indicates the beginning
of the stance (downward)
the swing (upward) phases. B: succession of the limbs touch down indicated by the numbers
1, 2, 3 and 4. C: corresponding gait diagram. The stages of the progression (1 to 20)
are indicated on the diagram and correspond to the frames of the sequence (1 frame =
0.04 s). El, extension subphase of the swing; E2, yield period; E3, propulsion subphase;
period of the swing; FE, full extension
during a preliminary subphase
of the stance;
LF, left fore limb; LH, left hind limb; RF, right fore limb; RH, right hind limb.
immediately dry after contact. The outer carpal pad print (toward finger V)
was present in a slow walk but missing in a faster walk. The liquid was
drained by the surface structure of the pads as soon as they were pressed
against the ground for the first time. The hairs between the digital and
the palmar pads also assisted draining. Tracks made by the claws become
visible after several locomotor cycles. The large and bare plantar sole of
the foot, which is crossed by a transverse groove, gave a more limited print
which corresponded to the plantar pad, the posterior part of the heel and a
strip on the outer side of the foot. The digital pads and the claw contacts
were visible. Liquid was drained by the polygonal network of epidermal
cells on the pads surface. The pressure exerted on these pads drove the
liquid toward the less protuding regions where hairs were concentrated.
In the slow or the slow medium walk of the polar bear, which is charac-
terized by a lateral sequence (HILDEBRAND, 1966), the pendular action of
the limbs seems to be predominant. The propulsive action by the muscles
is probably weak. In this context, the flexion of the shoulder joint in E3
reveals the anticipated activity of the flexor muscles which normally act
only during the swing. The knee shows the same behaviour. The elbow
being locked in maximal extension during E3, the propulsive action of the
fore limb is generated by the wrist and the hand. In contrast, the hip and
.the ankle both play this role for the hind limb. One of the main features
of the symmetrical gaits is the presence of a subphase FE, corresponding
to the setting of the limb in a pre-strain condition, in particular the hind
limb which must support a large percentage of the body mass. Maximally
extended, the leg forms a column which is able to resist the mass transfered
on it, at the beginning of the contralateral limb swing phase. The direction
of the ground reaction force is probably subvertical. The stiff and extended
hind leg in subphase FE results in a walk with swaying hips.
Fig. 8. Variation of several parameters of the asymmetrical gaits of the polar bear'(from
1977). A: duration of fore and hind contact intervals as percent of the stride
interval. B: Gait graph relating hind support to midtime lag. The point surrounded by a
circle corresponds to the gallop presented in fig. 7. C: hind and fore leads relation to stance
phase (duty factor). Regression
by means of least squares.
. , - - -
Fig. 9. Ventral
aspect of the right hand and foot of the polar bear and their footprints during
slow walk. I, II, III, IV and V, digits of the hand and the foot. A: right hand, d.p., digital
pads; h., hairy regions between the palmar and digital pads and behind the palmar pad; p.p.,
palmar pad. B: right foot, d.p., digital pads; h., pad of the heel; Ih., long hairs between
plantar and digital pads; Ish., longest hairs of the periphery; pl.p., plantar pad; sh., short hairs
between the heel and the plantar pad. C: right front leg footprint,
D: right hind leg footprint.
Fig. 10. Gallop cycle of a brown bear (Gambaryan, 1974). A: lateral view. The direction
of the arrows under the feet and hands, associated with LH, RH, LH and RF, indicates the
beginning of the stance (downward) and the swing (upward) phases. B: succession of the
limbs touch down. The succession of the limbs touch down is indicated by the numbers 1,
2, 3 and 4. C: gait diagram (20 stages). The stages of the progression
(1 to 20) are indicated
on the diagram and correspond to the frames of the sequence (1 frame = 0.04 s). On the
diagram of gait, the black lines and the dotted lines respectively represent the stance and the
swing phases. LF, left fore limb; LH, left hind limb; RH, right fore limb; RH, right hind
During the asymmetrical gaits, the propulsive activity of the hind limb
seems to increase, as demonstrated by the extension of all joints in E3.
The fore limb action is also amplified, and plays an important role in the
forward traction of the body. The shoulder musculature seems to mainly
operate in this way, assisted by the elbow. Both hind limbs probably have
the same propulsive effect. However, the leading fore limb seems to have a
greater propulsive action than the trailing limb. It is the only leg in contact
with the ground during the E3 subphase supporting the entire body mass,
just before the gathered suspension.
The transverse gallop used by the polar bear improves stability by increas-
ing the triangle of support, while the rotary gallop of the brown bear seems
more suited for manoeuvrability and climbing by the action of contralateral
(1974) related the larger stride length of the fore limb
of brown bears to climbing. The difference in gallop may also be related
to geometrical features: the body of the polar bear is more elevated and the
limbs tend to have equal length.
Our experiments and observations at the Zoo revealed that the behaviour
of the brown bear was quite different from that of the polar bear when the
animals were faced with a slippery substrate. Brown bears were reluctant
to walk on perspex sheets and they slided on them. The polar bears has no
difficulty with walking on perpex and left footprints which were identical
to those on rough paper. This behavioural difference may be explained by
the increase of hairy areas in the polar bear, which improve grip on the
The ability of the polar bear to travel at different gaits on ice and frozen
snow are probably the result of adaptive features at different structural levels.
The first level corresponds to the general aspect of the organism. The huge
body has a long trunk set above the ground by stiff limbs. The second level
concerns the use of the legs. The peculiar yield system contributes to a firm
grip on the slippery surface. At the end of the swing phase, the limb is
moved downward vertically forming a column by full extension of the joints.
This is a preparation for the transfer of a large part of the body mass before
the take-off of the contralateral limb occurs (subphase FE of the stance).
Fig. 11. Ventral aspect of the right hand and foot of a brown bear (Pocock, 1914) and their
footprints. A: right hand: c.p., carpal pads; d.p., digital pads; h., hairy; p.p. palmar pad.
B: right foot: d.p., digital pads; pl.p., plantar sole; h., hairy zone. C: right front leg footprint.
D: right hind leg footprint.
This loading results in some flexion of the leg (subphase E2 of cycle) which
is stopped by the activity of the extensor muscles (GOSLOW et al., 1973,
ORSAL, 1987). The FE strategy not only resists the mass transfered on a
leg, but probably favours a large contact of feet and hands with the ground
at the begining of E2. The third level is related to the structures for contact.
Plantigrady, including the contribution of the inner margin of the soles,
increases the surface supporting the body. The phalanges, particularly the
first, are laid flat on the ground as are the tarsal or the carpal, the metatarsal
or the metacarpal elements. Alternation of pads and hairy zones is a basic
characteristic of the plantar and palmar soles. The pads are orientated
perpendicular to the axis of the sole and to the direction of motion. There
are three hairy zones: short hairs are located posterior to the large pads
(palmar or plantar), long hairs occupy the area between the large pads and
the digital pads, and the longest hairs on the outer borders constitute a fold,
which increases the snowshoe-like surface of the feet. The structure of the
soles in polar bears creates an increase of the friction between the animal
and the substrate. This feature allows a good exchange of forces and avoids
aquaplanning and thus preserves the propulsive component.
The adaptation of the polar bear to travel on ice is probably the fitting
of the different structural systems acting at these three distinct levels of the
We are grateful to Mr J.J. Petter and Mme Leclerc-Cassan, who authorized
the experiments with thc polar and the brown bears in the zoo of Vincennes
(National Museum of Natural History) and also Mr R. Marion for the loan
of many video records and also for helpful information collected during ex-
peditions to the polar regions. Financial support was provided by a contract
between the Museum of Natural History and the Michelin Company.
DAGG, A.I., 1973. Gaits in mammals. Mammal Rev. 3,4: 135-154.
A.I. & A. DE VOS,
1968. The walking gaits of some species of Pecora. J. Zool. Lond.
FISCHER, M.S., 1992. Kinematics of the locomotion of the hyrax and morphological adapta-
tions. A new approach to the locomotion
of smaller mammals. Am. J. Phys. Anthrop.
GAMBARYAN, P., 1974. How mammals run. Wiley, New York
(Orig. publ. in Russian, 1972).
& D.G. STUART,
1973. The cat step cycle: hind limb
angles and muscle lengths during a restraining locomotion. J. Morph. 141: 1-42.
HARINGTON, C.R., 1970. Polar bear. Canadian Wildlife Service: Hinterland Who's Who.
HILDEBRAND, M., 1962. Walking, running and jumping. Am. Zool. 2: 151-155.
HILDEBRAND, M., 1966. Analysis of the symmetrical gaits of tetrapods. Folia Biotheoretica
HILDEBRAND, M., 1976. Analysis of tetrapod gaits: general considerations and symmetrical
gaits. In: R.M. Herman et al. (Eds): Advances in Behavioral Biology:
Press, New York.
M., 1977. Analysis of asymmetrical
gaits. J. Mammal. 58: 131-156.
F.A. JR., 1971. Limb posture and locomotion in the Virginia opossum (Didelphis
marsupialis) and in other non-cursorial mammals. J. Zool. Lond. 165: 303-315.
ORSAL, D., 1987. Analyse des coordinations intra et inter-appendiculaires
au cours de
chez le chat thalamique.
Thèse de doctorat d'Etat, univ Paris.
POCOCK, R.I., 1914. On the feet and other external features of the Canidae and Ursidae.
Zool. Soc. Lond. 2: 913-941.
ROCHA-BARBOSA, O., S. RENOUS & J.P. GASC,
1996. Comparison of the fore and hind
in the symmetrical
and asymmetrical gaits of a caviomorph
domestic guinea pig, Cavia
porcellus (Linné, 1758) (Rodentia, Caviidae).
Ann. Sci. Nat.
(Zool.) 17 (4): 149-165.
ROCHA-BARBOSA, O., S. RENOUS & J.P. GASC,
à la course chez le cobaye
Cavia porcellus (Mammifère, Rongeur, Caviomorphe).
Bull. Soc. Zool. Fr. 121 (1):