970 Swimming

References Reeves, R.R., and Chaudhry, A.A. (1998). Status of the Indus River dol-
phin Platanista minor. Oryx 32(1), 35–44.
Anderson, J. (1879). Anatomical and Zoological Researches: Comprising an
Rice, D.W. (1998). Marine Mammals of the World, Systematics and
Account of Zoological Results of the Two Expeditions to Western Yunnan
Distribution. Special Publications of the Society of Marine
in 1868 and 1875; and a Monograph of the Two Cetacean Genera,
Mammalology, Lawrence, Kansas.
Platanista and Orcella [sic]. Bernard Quaritch, London.
Sinha, R.K., Smith, B.D., Sharma, G., Prasad, K., Choudhury, B.C.,
Braulik, G.T. (2006). Status assessment of the Indus River dolphin,
Sapkota, K., Sharma, R.K. and Behera, S.K. (2000). Status and dis-
Platanista gangetica minor, March–April 2001. Biol. Conserv. 129(4),
tribution of the Ganges Susu, Platanista gangetica, in the Ganges
579–590.
river system of India and Nepal. In “Biology and Conservation of
Braulik, G.T., Bhatti, Z.I., Ehsan, T., Hussain, B., Khan, A.R., Khan,
Freshwater Cetaceans in Asia, IUCN Species Survival Commission
A., Khan, U., Kundi, K., Rajput, R., Reichert, A.P., Northridge, S.P.,
Occasional Paper Series 23”, (R.R. Reeves, B.D. Smith, and
Bhaagat, H.B., and Garstang, R. (2012). Robust abundance estimate
T. Kasuya, Eds), pp. 54–61, Gland, Switzerland.
for endangered river dolphin subspecies in South Asia. Endanger.
Smith, B.D., Ahmed, B., Edrise, M., and Braulik, G. (2001). Status of the
Species Res. 17, 201–215.
Ganges river dolphin or shushuk Platanista gangetica in Kaptai Lake
Braulik, G.T., Arshad, M., Noureen, U., and Northridge, S.P. (2014).
and the southern rivers of Bangladesh. Oryx 35, 61–72.
Habitat fragmentation and species extirpation in freshwater ecosys-
Smith, B.D., Haque, A.K.M., Hossain, M.S., and Khan, A. (1998). River
tems; causes of range decline of the Indus River dolphin (Platanista
dolphins in Bangladesh: Conservation and the effects of water devel-
gangetica minor). PLoS One 9(7), 101657.
opment. Environ. Manage. 22(3), 323–335.
Braulik, G.T., Barnett, R., Odon, V., Islas-Villanueva, V., Hoelzel, A.R.,
Smith, B.D., Braulik, G., Strindberg, S., Ahmed, B., and Mansur, R.
and Graves, J.A. (2015). One species or two? Vicariance, lineage
(2006). Abundance of Irrawaddy dolphins (Orcaella brevirostris) and
divergence and low mtDNA diversity in geographically isolated
Ganges river dolphins (Platanista gangetica gangetica) estimated
populations of South Asian River dolphin. J. Mamm. Evol. 22(1),
using concurrent counts from independent teams in waterways of
111–120.
the Sundarbans mangrove forest in Bangladesh. Mar. Mamm. Sci.
Cassens, I., Vicario, S., Waddell, G.V., Balchowsky, H., Van Belle, D.,
22(2), 1–21.
Wang, D., Fan, C., Mohan, R.S.L., Simões-Lopes, P.C., Bastida,
Smith, B.D., Diyan, M.A.A., Mansur, R.M., Mansur, E.F., and Ahmed,
R., Meyer, A., Stanhope, M.J., and Milinkovitch, M.C. (2000).
B. (2010). Identification and channel characteristics of cetacean
Independent adaptation to riverine habitats allowed survival of
hotspots in waterways of the eastern Sundarbans mangrove forest,
ancient cetacean lineages. Proc. Natl. Acad. Sci. U.S.A. 97(21),
Bangladesh. Oryx 44(02), 241–247.
11343–11347.
Yamasaki, F., and Kito, K. (1984). A morphological note on the intestine
Choudhary, S., Smith, B.D., Dey, Subhasish, Shushant, Dey, and Prakash,
of the boutu, with emphasis on its length and ileo-colic transition
S. (2006). Conservation and biomonitoring in the Vikramshila
compared with other platanistids. Sci. Rep. Whales Res. Inst. Tokyo 35,
Gangetic Dolphin Sanctuary. Oryx 40(2), 189–197.
165–172.
Choudhary, S., Dey, S., Dey, S., Sagar, V., Nair, T., and Kelkar, N. (2012). River
dolphin distribution in regulated river systems: Implications for dry-sea-
son flow regimes in the Gangetic basin. Aquat. Conserv. 22(1), 11–25.
Herald, E.S., Brownell Jr., R.L., Frye, F.L., Morris, E.J., Evans, W.E.,
and Scott, A.B. (1969). Blind river dolphins: First side-swimming SWIMMING
cetaceans. Science 166, 1408–1410.
Jensen, F.H., Rocco, A., Mansur, R.M., Smith, B.D., Janik, V.M., and
Madsen, P.T. (2013). Clicking in shallow rivers: Short-range echolo-
Terrie M. Williams
cation of Irrawaddy and Ganges river dolphins in a shallow, acousti-
cally complex habitat. PLoS One 8(4), e59284. The primary mode of locomotion for marine mammals with the
Kasuya, T. (1972). Some informations on the growth of the Ganges dol- exception of polar bears is swimming. For dolphins, porpoises and
phin with a comment on the Indus dolphin. Sci. Rep. Whales Res. Inst. whales, as well as manatees and dugongs, it is the only form of loco-
24, 87–108.
motion. The duration of swimming among these mammals may be
Kelkar, N., Krishnaswamy, J., Choudhary, S., and Sutaria, D. (2010).
Coexistence of fisheries with river dolphin conservation. Conserv.
as short as several seconds when moving between foraging patches
Biol. 24(4), 1130–1140. or as long as several months during seasonal migrations across
Mansur, R.M., Alom, Z., Smith, B.D., and Akhtar, F. (2014). Monitoring entire ocean basins. Although swimming by marine mammals often
the Mortality of Freshwater Cetaceans in the Sundarbans, appears effortless, it is in reality a delicate balance between precise
Bangladesh. In “Rivers for Life - Proceedings of the International body streamlining, exceptional thrust production by specialized
Symposium on River Biodiversity: Ganges-Brahmaputra-Meghna propulsive surfaces, and locomotor efficiency (Fig. 1).
River System”, (R.K. Sinha, and B. Ahmed, Eds), pp. 124–128. IUCN
S Paudel, S., Pal, P., Cove, M.V., Jnawali, S.R., Abel, G., Koprowski, J.L.,
and Ranabhat, R. (2015). The Endangered Ganges River dolphin
I. Hydrodynamics and Body Streamlining
One the most characteristic features of marine mammals other
Platanista gangetica gangetica in Nepal: Abundance, habitat and con- than polar bears is a streamlined body shape. This is not surpris-
servation threats. Endanger. Species Res. 29(1), 59. ing when one considers the forces that the animal has to overcome
Pilleri, G., and Zbinden, K. (1973–74). Size and ecology of the dolphin
in order to move through water. During swimming a force, termed
population (Platanista indi) between the Sukkur and Guddu bar-
rages, Indus River. Invest. Cetacea 5, 59–69. drag, acts backwards on the swimmer’s body resisting its forward
Pilleri, G., Marcuzzi, G., and Pilleri, O. (1982). Speciation in the motion. The equation describing total body drag is given by
Platanistoidea: Systematic, zoogeographical and ecological observa- 1 2
tions on recent species. Invest. Cetacea 14, 15–46. Total Drag = ρV ACd (1)
Reeves, R.R., and Brownell, R.L. (1989). Susu Platanista gangetica 2
(Roxburgh, 1801) and Platanista minor Owen, 1853. In “Handbook where ρ is the density of the fluid, V is the velocity of the fluid
of Marine Mammals, River Dolphins and Larger Toothed Whales”, relative to the body, A is a characteristic area of the body, and Cd
(S.H. Ridgeway, and R. Harrison, Eds) Sir, Vol. 4, pp. 69–100. is the drag coefficient (a factor that takes into account the shape of
Academic Press, London.
 Swimming 971

these morphological measurements, termed the Fineness Ratio, can

be written
maximum body length
Fineness Ratio =
maximum body diameter (2)
The optimum fineness ratio that results in minimum drag with
maximum accommodation for volume is 4.5. Calculations of the
fineness ratio for a wide variety of marine mammals show that
many species have body shapes that conform to the ideal hydrody-
namic range (Fig. 2). A review paper by Fish (1993) showed that
many cetaceans, pinnipeds and sirenians have body shapes with
Fineness Ratios that range from 3.0 to 8.0. The species examined
included seals, sea lions and odontocete whales that are consid-
ered by many to typify a streamlined body profile. Even mysticete
whales with enlarged heads and jaws specialized for filter feeding
maintain a streamlined body profile when swimming (Fig. 2). This
hydrodynamic profile is lost when the animals open their mouths to
feed, which immediately causes a marked increase in drag, a reduc-
tion in forward speed, and a concomitant elevation in energetic
costs (Acevedo-Gutierrez et al., 2002; Goldbogen et al., 2006, 2012).
Despite nearly ideal body streamlining, all marine mammals
must contend with drag forces when moving through the water.
These forces can be a considerable challenge for the swimmer and
will influence how quickly the animal will be able to move. It is
apparent from Eq. (1) that the velocity of the swimmer will have
Figure 1 A bottlenose dolphin (Tursiops truncatus) swimming at a large impact on total body drag. As the swimmer moves faster,
high speed on the water surface. The generation of waves by the dolphin’s body drag increases exponentially. An example of the relationship
movements leads to increases in both body drag and energetic costs during between total body drag and velocity is presented in Fig. 3 for the
surface swimming (Photo by T.M. Williams). sea otter (Williams, 1989). Whether the sea otter swims on the
water surface or submerged, drag increases with velocity. Body
position clearly affects the level of force encountered such that at all
comparable swimming speeds body drag is higher for the sea otter
the swimmer). There are four primary types of drag that contrib- moving on the water surface than when it is swimming submerged.
ute to total body drag, (1) skin friction drag which is a tangential The same results have been found for other swimmers including
force resulting from shear stresses in the water sliding by the body, humans and harbor seals. In general, body drag for a swimmer
(2) pressure drag which is a perpendicular force on the body associ- moving on or near the water surface is 4–5 times higher than the
ated with the pressure of the surrounding fluid, (3) wave drag that level of drag encountered by the submerged swimmer moving at
occurs when a swimmer moves on or near the water surface, and (4) the same speed (Hertel, 1966). Much of this increase in drag at the
induced drag that is associated with water deflection off of hydrofoil water surface is due to energy wasted in the formation of waves.
surfaces such as fins, flukes or flippers. Of these, pressure drag is the This can be avoided if the swimmer is able to submerge to a depth
component most influenced by body streamlining in marine mam- equivalent to three body diameters. For a seal or small whale with a
mals. The more streamlined a body, the lower the pressure drag and maximum body diameter of 1 m, this would mean changing swim-
consequently the lower the total drag of the swimmer. ming position to at least three meters in depth to avoid wave drag
Mammals whose lifestyles or foraging habits involve prolonged and the consequent elevation in total body drag. This is one of the
periods of swimming typically have streamlined body shapes. In reasons that swimming is comparatively difficult for humans—our
contrast to the lanky appearance and appendages of terrestrial entire performance takes place on the water surface where wave
mammals, marine mammals tend to have a reduced appendicular drag, and hence total body drag, is the highest.
The ability to swim submerged for prolonged periods is one of
skeleton and characteristic tear-drop body profile. External features
that may disrupt water flow across the body are also reduced or the most important adaptations for increasing swimming efficiency
S
absent in many species of marine mammal. These features include and performance in marine mammals. The sea otter provides an
the pinnae (external ears), limbs, and long fur. In highly specialized excellent example of the advantage provided by this adaptation. Sea
swimmers, such as dolphins and humpback whales, microanatomical otters restrict prolonged periods of surface swimming to speeds less
characteristics of the skin and the shape of the leading edges of the than 0.8 m s−1 and to a maximum body drag of 4.2N (Fig. 3). For
fins, respectively, may help to direct the flow of water in a controlled high speed swimming, sea otters change to a submerged mode of
manner across body surfaces. All of these adaptations prevent the locomotion. In doing so, drag is reduced by 3.5 times and the sea
onset of turbulence in the water surrounding the swimmer, thereby otter is able to reach speeds of 1.4 m s−1 before body drag once again
reducing total drag. exceeds 4.0N. Thus, behavioral changes by the sea otter take into
Hydrodynamic theory describes the streamlined body shape as account the differences in drag associated with body position in the
one in which a rounded leading edge slowly tapers to the tail, and water column and allows the animal to extend its range of swim-
total length is 3–7 times maximum body diameter. The ratio of ming speeds. Several other behavioral strategies, such as porpoising
 972 Swimming

(A)

f.r. 4.3

(B)

f.r. 7.6
(C)

(D)
f.r. 4.5

f.r. 6.8
Figure 2 Body shapes and fineness ratios for cetaceans. Shapes can range from the robust bowhead whale, Balaena mysticetus (A) to the long thin
tapered body of the rorqual whales (B) and beaked whales (D). The killer whale (Orcinus orca, C) has the optimum shape in terms of fineness ratio and
streamlining. From Berta and Sumich (1999), “Marine Mammals: Evolutionary Biology”, Academic Press.

6 stroke cycle. This capability is found in highly adapted marine spe-

cies such as pinnipeds and cetaceans. It contributes to an increase in
locomotor efficiency in marine mammals especially when compared
5
to the inefficient drag-based swimming styles of humans and terres-
trial mammals (Fig. 4).
Body drag (Newtons)

4 Marine mammals use a wide variety of swimming styles to

move through the water (Table 1). The most terrestrial species of
3 this group, the polar bear and sea otter, swim by alternate strokes
of the forelimbs or hind limbs, respectively. Polar bears use a dog-
style of forelimb paddling with the hind limbs dragged passively
e

2
c

behind or used as an aid to steering. Sea otters are unique among

rfa

d
Su

ge marine mammals in their ability to lie on their backs during surface

er
1 bm swimming. Propulsion is provided by either the simultaneous or
Su
alternate strokes of the hind limbs. When on the surface sea otters
0 can also swim ventral surface (belly) down using the hind paws for
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 propulsion. The front paws are held against the submerged chest
1
Velocity (ms ) and do not play a role in propulsion during this mode of swimming.
Figure 3 A comparison of body drag for surface and submerged sea otters Stroke frequency has been measured for swimming sea otters and
S (Enhydra lutris) in relation to swimming speed. Note that at all comparable ranges from approximately 30 to 80 strokes per minute while swim-
speeds, body drag of the sea otter on the water surface is higher than when ming on the water surface (Williams, 1989).
the otter is submerged. The dashed line denotes the preferred swimming Polar bears and sea otters are the only marine mammals that
speeds of surface and submerged sea otters. From Williams, 1989. rely primarily on drag-based modes of swimming. These modes
have two distinct phases during the stroke cycle, a power phase
when thrust is produced, and a recovery phase when the foot or paw
and wave-riding, are also used by marine mammals to avoid ele- is repositioned for the next stroke. During the power phase the paw
vated body drag while swimming and will be discussed below in or foot is moved backward relative to the body. Drag created by this
Swimming Speed and Behavior. motion is subsequently translated into thrust and the animal moves
forward through the water. The enlarged hind flippers of sea otters
II. Kinematics and fore paws of polar bears enables the animals to increase propul-
A hallmark of marine mammal swimming is the use of lift-based sive efficiency by moving a large mass of water during this power
propulsion that allows thrust to be generated through the entire phase. The recovery phase of the stroke is only used to bring the
 Swimming 973

(A) limb back to its starting position and occurs without the generation
T D of thrust. Because thrust is produced only during part of the stroke
cycle, drag-based modes of swimming are comparatively inefficient.
When sea otters want to move quickly through the water they
switch to an undulatory mode of swimming involving dorsoven-
d tral body flexion and simultaneous movements of paired hind flip-
pers. The tail and hind flippers are held straight back and trail the
undulatory movements of the trunk. Stroke frequency of sea otters
L
remains relatively constant at 55 strokes per minute during sub-
(B) d
merged undulatory swimming, which suggests that underwater
T D speed is elevated by increasing stroke amplitude.
Dolphins and whales use undulatory modes of propulsion, in a
manner similar to that of submerged sea otters. The primary pro-
pulsive movements of all cetaceans occur in the vertical plane
with the posterior third of the body undulating in a dorsoventral
direction. Termed, thunniform swimming or carangiform swim-
L
d
ming with a semi lunate tail, this mode of locomotion is character-
(C) ized by an undulatory wave that travels with increasing amplitude
T D down the body, caudal peduncle and finally the flukes (Fig. 5).
“Semi-lunate” refers to the crescent shape of the flukes. This mode
of propulsion is shared by other fast swimming vertebrates includ-
ing tuna, hence the name “thunniform.” Undulatory propulsion
in cetaceans is considered highly efficient and can generate high
levels of thrust on both the upstroke and downstroke. There is
L
no recovery phase; thus, propulsion can be produced through-
d
out the stroke cycle. Stroke frequency using this mode of swim-
(D) ming varies with the speed and size of the cetacean. The range of
stroke frequencies for bottlenose dolphins swimming and diving is
T D
30–180 strokes min−1.
Stroke frequency decreases with increasing body size among
Figure 4 Swimming modes for semi-aquatic and marine mammals. The the cetaceans. The largest species of swimming mammal, the
muskrat (A) is a semi-aquatic mammal that uses drag-based propulsion by 100 t blue whale, uses stroke frequencies that are only one-tenth
paddling its hind feet. Otariids (B), phocid seals (C) and cetaceans (D) use of the range observed for bottlenose dolphins. A measurement
lift-based propulsion that may involve fore flippers (sea lion), lateral body of the stroke frequency of blue whales ascending during a dive
undulation (seal) or dorsoventral undulation (dolphin). Major forces on the was 6–10 strokes min−1 (Williams et al., 2000). Interestingly, the
animals and propulsive surfaces are shown. T denotes thrust, and D shows much smaller manatee, a 500 kg herbivore, also uses a dorsoven-
the direction of body drag on the animals. L and d illustrate lift and drag tral style of propulsion and displays a range of stroke frequencies
forces on the appendages, respectively. From Fish, 1993, with permission. (5–10 strokes min−1) similar to that of blue whales when it is slowly
cruising through topical waters. Maximum stroke frequency for this
comparatively sedentary sirenian is 27 strokes min−1.

TABLE 1 A Comparison of Swimming Characteristics for Four Major Classes of Marine Mammals

Sea otter Otariid Phocid Small

cetacean
Routine <0.8 (surface)
Speed (m s−1) <1.4 (submerged) 2.0–3.0 1.2–2.0 2.0–4.0
Sprints to 4.0 Sprints to 10.0 S
Hydrodynamics Surface/Submerged Submerged Submerged Submerged
Kinematics
Mode Paddle, Row (surface) Pectoral Lateral Dorsoventral
Undulate (submerged) Carangiform Thunniform
Energetics
(COT measured) 12.0 (surface)
COT predicted 6.0 (submerged) 2.3–4.0 2.3–4.0 2.1–2.9

The energetic Cost of Transport (COT) was measured for animals swimming in a flume or freely swimming in open
water. The ratio of these values and the predicted values for fish of similar body mass are presented. COT values and
ratios are for routine cruising speeds only.
 974 Swimming

other advantages are provided by fore flipper propulsion. These

include stability at slow speeds and maneuverability at high speeds.
Fluke tip
Consequently, otariids are champion underwater acrobats and are
capable of rapid changes in direction and acceleration.
Phocid seals and walruses differ from the otariids in terms of
Dorsal fin swimming style, and rely on alternate sweeps of the hind flippers
Fluke hinge for propulsion. In addition to the flippers, the posterior half of the
Peduncle body flexes during each stroke with the result that body flexion pro-
vides nearly 90% of the change in amplitude during the stroke cycle.
In phocid seals both hind flippers are swept in the same direction as
the posterior portion of the body during each half of the stroke cycle.
The leading flipper remains closed and the trailing flipper maximally
expands during the sweep to one side. Once the flippers have moved
(A) to the maximum lateral position, the flippers switch their open and
closed positions in preparation for the reverse lateral sweep. By revers-
Fluke tip ing the role of each flipper during lateral sweeps, one flipper is able to
1.2
provide thrust while the other flipper recovers. The result, once again,
Dorsal fin is the ability to produce propulsive thrust during the entire stroke
Peduncle
cycle. Stroke frequency in phocids increases linearly with swimming
1.0 speed. For harbor seals trained to swim at 1.0 to 1.4 m s−1 in a water
flume stroke frequency ranged from 60 to 78 strokes min−1. This com-
Position (m)

pares with the relatively slow stoke frequency (10–55 strokes min−1) of
0.8 free-ranging Weddell seals diving beneath the Antarctic ice.
Fluke III. Energetics
hinge
0.6 The energetic cost of swimming has been measured using a wide
variety of techniques with numerous species of semi-aquatic and
marine mammals. Smaller swimmers such as minks, muskrats, river
0.4 otters, sea otters, seals and sea lions have been studied while they
3.5 3.7 3.9 4.1 4.3 4.5 4.7 swam against a current in water flumes. Similar to placing a human
(B) Time (sec) on a treadmill, flume studies have enabled scientists to measure how
Figure 5 Video image (A) and range of movement (B) of four anatomi- much energy a swimmer expends while moving at different speeds.
cal sites during a single stroke for a swimming bottlenose dolphin. Colored Often oxygen consumption is measured during these tests by using
squares in the picture correspond to the line colors illustrating the move- a face mask or metabolic hood connected to an oxygen gas analyzer.
ments for each site. Note the traveling wave as it passes from the dorsal fin By training animals to breathe into the metabolic hood, expired
(dark blue) to the peduncle (red), fluke hinge (green) and finally the fluke tip respiratory gases can be collected and analyzed for oxygen con-
(pink). From Skrovan et al., 1999, with permission. tent. For larger, more powerful swimmers like dolphins and whales
most flumes are not adequate in terms of size or challenging water
speeds. Instead, investigators have relied on novel techniques for
Swimming by pinnipeds differs markedly between the eared determining the energetic cost of swimming in cetaceans. Methods
seals (otariids) and the true seals (phocids). Otariids use pectoral have included using trained dolphins that match their swimming
appendages to generate propulsive forces during swimming, with speed to that of a moving boat in open water (Williams et al., 1992)
the hind flippers trailing passively or occasionally used for steer- or having whales swim to metabolic stations where expired gases
ing. In this way, sea lions and fur seals resemble penguins and sea can be collected for analysis (Worthy et al., 1987; Otani et al., 2001).
turtles during swimming. Detailed kinematic analyses have been To compare swimmers of different size, it is useful to convert
conducted for California sea lions swimming in a flume (Feldkamp, the metabolic measurements into a cost of transport. Defined as the
1987). These studies revealed three distinct phases to the stroke, amount of fuel it takes to transport one unit of body weight over
(1) the power phase, (2) a paddle phase, and (3) a recovery phase. a unit distance, the cost of transport is analogous to the fuel rat-
S The majority of thrust is produced during the paddle phase when ing of an automobile. In this case, the cost of transport indicates the
the fore flippers are quickly and forcibly moved from the outer “gas per mile” used by the swimmer rather than the “miles per gas”
water flow to the sides of the animal’s body. Stroke frequency for achieved by automobiles. The total cost of transport is calculated
these sea lions increased with swimming speed and ranged from from the following equation:
15 to 50 strokes min−1 as the animals increased speed from 0.5 Oxygen consumption
to 3.0 m s−1. In addition to stroke frequency, sea lions increase the Totalcost of transport = (3)
Swimming speed
amplitude of the fore flipper stroke during high speed swimming.
When viewed in cross-section, the fore flipper of the sea lion where oxygen consumption is in mLO2 kg−1 s−1 and speed
resembles a hydrofoil. This specialized shape allows the flipper to is in m s−1 which results in a cost of transport in mLO2 kg−1 m−1.
produce thrust during both the power and recovery phases of the These values are usually converted to an energetic term and
stroke cycle. As found for cetaceans, the specialized flipper move- expressed as Joules expended per kg of body mass per meter trav-
ments of otariids result in thrust production throughout the eled (J kg−1 m−1). The conversion calculation assumes a caloric
stroke cycle and contribute to overall locomotor efficiency. Several equivalent of 4.8 kcal L−1 of oxygen consumed and a conversion
 Swimming 975

100 As illustrated in Fig. 6, the energetic cost of swimming for

marine mammals is greater than predicted for salmonid fish of sim-
ilar body size. Despite specialization of the body and propulsive
Total Cost of Transport MIN (J. kg–1.m–1)

surfaces for aquatic locomotion, the cost of transport for swimming

Semi Aquatic Mammals
by seals and sea lions is 2.3 to 4.0 times higher than predicted for
swimming fish. Values for cetaceans are somewhat lower, and range
10 from 2.1 to 2.9 times predicted values. These differences between
marine mammals and fish are due in part to the amount of energy
Phocid Seals
Steller Sea Lion
expended for maintenance functions, particularly thermoregulation
and the support of a high core body temperature. As endotherms,
Sea Lions
mammals expend more energy to support the production of endog-
Bottlenose
enous heat than ectothermic fish. In addition, many marine mam-
1 Killer Whales
Dolphin mals show exceptionally high metabolic rates while resting in water
in comparison to terrestrial mammals resting in air. A consequence
Gray Whale of these high maintenance costs is an overall increase in the total
energy expended during swimming, especially when compared to
Fish fish. It remains to be seen how manatees with their low core body
0.1 temperatures and low resting metabolic rates (Scholander and
0.1 1 10 100 1000 10000
Irving, 1941) will compare to other mammalian groups.
Body Mass (kg) IV. Swimming Speeds and Behavior
Figure 6 Total energetic cost of transport in relation to body mass Although body size varies by nearly 3 orders of magnitude among
for different classes of swimmers. Marine mammals include grey seals marine mammals from the 20 kg sea otter to the 122,000 kg blue
(Haliochoerus grypus) and harbor seals (Phoca vitulina) (circles), whale, routine swimming is limited to a surprisingly narrow range
California sea lions (Zalophus californianus) and Steller sea lions of speeds. Many species of marine mammal routinely swim between
(Eumetopias jubatus) (diamonds), bottlenose dolphins (triangle), killer whales approximately 1.0 to 3.6 m s−1 regardless of body size (Fig. 7). Within
(squares), and a gray whale (Eschrichtius robustus) (downward-point- this range, pinnipeds generally select slower routine traveling speeds
ing triangle). The least squares regression through the data points for marine than cetaceans, and mysticete whales swim slower than odontoc-
mammals is presented in the text as Eq. (4). This regression is compared to the etes. For example, average swimming speed for a wide variety of
regressions for swimming semi-aquatic mammals (upper solid line), and the otariids and phocids ranges from 1.3 to 2.0 m s−1. The massive mys-
predicted regression for salmonid fish (lower solid line). The values used here ticete whales are only slightly faster; routine speeds for this group of
are equivalent to Total COTMIN. From Williams, 1999, with permission. marine mammals ranges from 2.1 to 2.6 m s−1. Although they are not
the largest marine mammals, odontocetes tend to move the fastest
during routine travel. The slowest of the odontocetes represented in
factor of 4.187 × 103 J kcal−1. Often the minimum cost of transport Fig. 7 is the beluga whale with a routine speed of 1.8 m s−1. In com-
(COTMIN) occurring at routine swimming speeds is used for com- parison, the killer whale demonstrates the fastest routine speed of
paring the energetic cost of swimming among marine mammals. the marine mammals measured to date and averages 3.6 m s−1 during
COTMIN for a wide variety of mammalian swimmers indicate casual swimming. These speeds are even more remarkable when com-
that this form of locomotion is energetically expensive for mammals pared to the efforts of humans. The routine speed of humans during
compared to fish. The total COTMIN for swimming mammals can freestyle swimming is approximately 1.0 m s−1, about the same speed
also be separated into two distinct groups, the semi-aquatic mam- as a sea otter swimming under water.
mals and the marine mammals (Williams, 1999; Fig. 6). Swimming As would be expected, the sprinting speeds of marine mammals
costs for semi-aquatic mammals such as minks, muskrats, and are considerably faster than routine cruising speeds, and show much
humans are two to five times higher than observed for marine variation among the species measured. Most of the information
mammals. These high energetic swimming costs are attributed to a regarding sprint swimming performance in marine mammals is for
wide variety of factors including elevated body drag associated with cetaceans. However, the speed of adult Weddell seals chasing fish
a surface swimming position (Fig. 3) and low propulsive efficiency beneath sea ice has been measured at more than 4.0 m s−1 during the
associated with drag-based propulsion (Fig. 4). hunt. Among cetaceans, sprint speeds are even higher. The range of
Mammals specialized for swimming demonstrate comparatively sprinting speeds measured for mysticete whales is 4.1 to 13.3 m s−1 S
lower energetic costs. Total COTMIN in relation to body mass for (Fig. 7); sprint swimming by odontocetes is within the upper end
swimming marine mammals ranging in size from a 21 kg California of this range and averages 6.1 to 12.5 m s−1. The short-finned pilot
sea lion to a 15,000 kg gray whale is described by whale, called the cheetah of the deep sea, cruises at 2.0 m s−1 and
can sprint up to 9.0 m s−1 when foraging (Aguilar de Soto et al.,
Total COTMIN = 7.79 mass−0.29 (4)
2008). Killer whales remain the fastest of the odontocetes and can
where the cost of transport is in J kg m and body mass is in −1 −1 sprint at 12.5 m s−1; sustained speeds of orca are closer to 3.7 m s−1
kilograms. The style of swimming used by marine mammals does as recorded during prolonged (30 min) chases in pursuit of bluefin
not affect this cost of transport relationship. Species and swimming tuna (Guinet et al., 2007). Note that the extreme sprint speeds of
styles represented in this equation include sea lions using pectoral these odontocetes is nearly six times faster than the maximum per-
fins for propulsion, phocid seals using lateral undulation of paired formance of human swimmers in Olympic sprint competitions.
hind flippers, and odontocete and mysticete whales using dorsoven- Because marine mammals must periodically surface to breathe,
tral undulation of flukes. they are subject to high levels of drag associated with the effects of
 976 Swimming

14

12

10
Swimming speed (m/s)

0
ag .s.
er l.

’s

ale

VS
ck
Bo Fin
d

d.

Bo otte .
.

ler .w.
l
al

.s.
Ga rhte .s.

s.l.

Bry ke

ay

Wh mm p.

Be d.
a

n
ea

Sp ed d
ttle d d
mp Sei
.

ea

Fa lug

ma
Ho os s

de
lap rn f
Se

Blu
ba
f
No ctic f

Gr

ite on
Co rbor

se

Kil se k
wh

AU
tS

An gos

Mi

wh

Hu
no
d
ell

ok
an

-si
Ha

l
tar
a
dd

Hu
ph

lap
We
Ele

Ga

Phocids Otariids Mysticetes Odontocetes Man-Made

Figure 7 Swimming speeds for marine mammals. Routine speeds of phocid seals, otariids, mysticetes, odontocetes, a sirenian, humans and autonomous
underwater vehicles (AUVs) are shown by the bars. Filled circles above the bars denote the sprinting speeds recorded for each species. Note the similar range
of routine speeds for these marine mammals regardless of body size.

wave formation and splashing especially during high speed swim- sequential dives that allows the animals to remain submerged except
ming. To help minimize body drag and energetic costs during these for brief surface intervals to breathe (Davis et al., 2001).
surface intervals, marine mammals have developed a number of
unique behavioral strategies to accommodate breathing while swim- V. Field Measurements and the Special Case of
ming fast. Porpoising is one such highly visible behavioral strategy Swimming at Depth
used by small cetaceans and some pinnipeds moving at high speed Most of our information about swimming in marine mammals
near the water surface (Au and Weihs, 1980). Rather than stroke is from animals moving near the water surface. Yet, the major-
continuously, the animals leap into the air and simply avoid the ele- ity of swimming by these animals occurs at depth in conjunction
vated wave drag that occurs when swimming near the water sur- with diving. When descending or ascending during a dive, marine
face to breathe. Theoretically, this behavior results in an energetic mammals must contend with changes in buoyant forces, hydro-
savings to the animal, although the cost of surface swimming ver- static pressure, as well as body drag. As discussed above, drag forces
sus leaping has yet to be measured. Wave-riding is another strategy resist both forward progression and limb movements of the swim-
that enables the swimmer to avoid the work of continuous stroking mer. In contrast, buoyant forces act in a vertical direction in the
while moving near the water surface. In a study involving bottle- water column and result from the weight, volume and compressi-
nose dolphins trained to swim freely or wave-ride next to a moving bility of the tissues and air spaces of the animal’s body. Hydrostatic
boat, investigators found that heart rate, respiration rate and ener- pressure is a function of the weight of the water column above the
getic cost were reduced for animals riding the bow wave of the boat marine mammal.
(Williams et al., 1992). This behavior enabled the dolphins to nearly The magnitude of buoyant forces and hydrostatic pressure on
S double their forward traveling speed with only a 13% increase in the swimming marine mammal will depend on where in the water
energetic cost. Consequently, it is not surprising that marine mam- column activity takes place. Hydrostatic pressure progressively
mals routinely ride waves generated by the wind, surf, the wake of increases by 1 ATM for every 10.1 m an animal descends. This will
boats and even large whales. What appears to be an amusing activ- have a profound effect on compressible spaces or tissues, and hence
ity also provides an energetic benefit to the swimmer. buoyancy of the animal, especially for marine mammals that may
Although energetically advantageous when swimming near the descend and ascend hundreds of meters during the course of a dive.
water surface, both wave-riding and porpoising have been described In addition, seasonal changes in blubber content, pregnancy and
for only a limited number of marine mammal species moving at high lactation will have an effect on the overall buoyancy of the marine
speeds. These locomotor strategies are not possible during slow mammal (Webb et al., 1998).
transit, in large marine mammals such as elephant seals and whales, A consequence of the interrelationships between depth, buoy-
or in polar regions where ice covers the water surface. Instead, ancy, hydrostatic pressure and body drag is that the physical forces
transit swimming is often accomplished by a sawtooth series of influencing an animal that is swimming horizontally near the water
 Swimming 977

Figure 8 Percentage of time spent gliding during descent in relation to dive depth for five species of marine mammal. Except for the dolphin, the depth
range was determined by the free-ranging behavior of the animals during foraging. Although the sea otter is the smallest species and demonstrated the least
amount of gliding, body size of pinnipeds and cetaceans per se does not dictate gliding duration. From Williams et al., 2015, with permission.

surface are very different from those encountered by the diving ani- descending to 540 m beneath the Antarctic sea ice. Nearly 80% of
mal moving vertically through the water column. Detailed studies the descent of diving seals is spent passively gliding rather than
on diving bottlenose dolphins and elephant seals have shown that actively swimming on dives exceeding 200 m in depth (Fig. 8).
the animals are positively buoyant near the water surface, and that The ascent portion of a dive requires more effort by these
buoyancy decreases as the animal descends during the dive. For marine mammals when compared to the descent. The beginning
example, the buoyancy of a bottlenose dolphin changes from posi- of the ascent represents the period of greatest swimming effort for
tive (+ 24N) when near the water surface to negative (− 26N) once many mammalian divers. During this period, pinnipeds and ceta-
the animal exceeds 70 m in depth (Skrovan et al., 1999). ceans may use sequential, large amplitude strokes to begin mov-
These changes in buoyancy are associated with changes in lung ing upward (Fig. 10). As the ascent continues, the physical forces
compression due to the increase in hydrostatic pressure as marine impacting the diver are once again altered with hydrostatic pres-
mammals descend on a dive. As dolphins, whales and seals dive, sure decreasing on ascent. Consequently, the lungs reinflate and
hydrostatic pressure increases, the lungs progressively collapse the buoyancy of the marine mammal increases. Swimming behav-
instigating a marked change in overall buoyancy. These changes in ior reflects these changes such that the initial continuous stroking
physical forces with depth affect both the locomotor behavior and phase is followed by a stroke and glide mode of swimming, and
energetics of the marine mammal as it sequentially descends and finally a brief glide to the water surface.
ascends (Fig. 8). An interesting contrast to the seals, dolphins, and blue whales is
Until recently, it was not possible to observe the swimming the right whale. This species, considered the “right” whale to hunt
modes of marine mammals during deep dives. With the develop- because their carcasses tend to float, are comparatively buoyant.
ment of miniaturized video cameras and instrumentation worn by The right whale displays some of its most powerful fluke strokes
free-ranging marine mammals, new information about swimming at the beginning of descent as it counteracts large positive buoyant
at depth has been obtained (Figs 8–10). Videos and accelerome- forces at the start of a dive (Nowacek et al., 2001). The advantage of S
ter recordings have revealed that bottlenose dolphins, elephant this positive buoyancy occurs during the ascent, when the animals
seals, Weddell seals and blue whales (Williams et al., 2000) as well are able to glide to the surface and reduce the number of energeti-
as deep diving beaked whales (López et al., 2015) switch between cally costly strokes.
different modes of swimming during the dive much like terres- By altering the mode of swimming to account for changes in
trial mammals switch between gaits. Dive descents usually begin the physical forces that occur during a dive, marine mammals are
with a period of continuous stroking. Once marine mammals reach able to conserve limited oxygen reserves during submergence.
70–80 m in depth they often change to a passive glide for much of Studies investigating the metabolic rates of Weddell seals diving
the remainder of the descent (Fig. 8). For deep divers such as pho- from an ice hole found that the incorporation of prolonged glides
cid seals and beaked whales, these gliding periods can be quite long. enabled seals to reduce the energetic cost of individual dives by
For example, prolonged gliding periods exceeded 6 min for north- 9%–60% (Williams et al., 2000, 2004). Such an energetic savings
ern elephant seals traveling to nearly 400 m and Weddell seals could make the difference between completing the dive aerobically
 978 Swimming

Figure 9 Instrumentation used to monitor swimming behavior in marine mammals. On the left, a bottlenose dolphin carries a video camera to record its
swimming movements during deep dives (Photo courtesy of Kevin McDonnell). On the right, a polar bear wears a collar equipped with a 3-axis accelerome-
ter to measure swimming strokes (Photo by T.M. Williams).

Figure 10 A typical foraging dive for a Weddell seal feeding on Antarctic silverfish. The 3-dimensional dive plot shows depth in relation to distance from
a breathing hole and is color coded for stroke frequency. Silverfish captures (black stars) occurred on the ascent phases of a rollercoaster foraging segment
at depth. Three primary swimming behaviors- gliding (blue), stroke-and-glide (intermittent blue and green), and constant, high amplitude stroking prior
to and during fish capture and initial ascent (green to orange-red) as recorded by an accelerometer are presented in the inset boxes. Each dot represents 1 s.
From Williams et al., 2015, with permission.
 Systematics 979

or anaerobically, and can increase the time available for hunting or Nowacek, D.P., Johnson, M.P., Tyack, P.L., Shorter, K.A., McLellan,
avoiding predators. W.A., and Pabst, D.A. (2001). Buoyant balaenids: The ups and downs
In summary, these studies demonstrate that swimming can be of buoyancy in right whales. Proc. R. Soc. Lond. B 268, 1811–1816.
energetically expensive for mammals. Marine adapted species includ- Otani, S., Naito, Y., and Kawamura, A. (2001). Oxygen consumption and
swim speed of the harbor porpoise Phocoena phocoena. Fish. Sci. 67,
ing sea otters, pinnipeds, and cetaceans have undergone marked
894–989.
morphological, physiological and behavioral changes to increase
Scholander, P.F., and Irving, L. (1941). Experimental investigations
their swimming efficiency. An especially important adaptation that on the respiration and diving of the Florida manatee. J. Cell. Comp.
distinguishes marine mammals from semi-aquatic mammals is the Physiol. 17(2), 169–191.
ability to remain submerged for prolonged periods when swimming. Skrovan, R.C., Williams, T.M., Berry, P.S., Moore, P.W., and Davis, R.W.
However, prolonged submergence also requires specialized physio- (1999). The diving physiology of bottlenose dolphins (Tursiops trun-
logical responses associated with oxygen loading and utilization as catus) II. Biomechanics and changes in buoyancy at depth. J. Exp.
described in Diving Physiology. A major benefit of these adapta- Biol. 202, 2749–2761.
tions is a capacity for aquatic performance by marine mammals that Webb, P.M., Crocker, D.E., Blackwell, S.B., Costa, D.P., and LeBoeuf, B.J.
far exceeds those of semi-aquatic mammals and the best Olympic (1998). Effects of buoyancy on the diving behavior of northern ele-
phant seals. J. Exp. Biol. 201, 2349–2358.
efforts of humans. It is important to note that in recent years rapid
Williams, T.M. (1989). Swimming by sea otters: Adaptations for low
changes in the marine environment have challenged these locomo-
energetic cost locomotion. J. Comp. Physiol. A 164, 815–824.
tor adaptations as marine mammals swim longer distances in search Williams, T.M. (1999). The evolution of cost efficient swimming in
of prey and more favorable thermal conditions. It is no wonder that marine mammals: Limits to energetic optimization. Phil. Trans. R.
the polar bear with its poorly streamlined body shape, inefficient Soc. Lond. B 354, 193–201.
dog-paddle kinematics, and surface swimming position, is especially Williams, T.M., Fuiman, L.A., and Davis, R.W. (2015). Locomotion and
threatened as it faces progressively longer oceanic forays. the cost of hunting in large, stealthy marine carnivores. Integr. Comp.
Biol. 55(4), 673–682. doi:10.1093/icb/icv025.
Williams, T.M., Davis, R.W., Fuiman, L.A., Francis, J., Le Boeuf, B.J.,
See Also the Following Articles
Horning, M., Calambokidis, J., and Croll, D.A. (2000). Sink or swim:
Bow-riding ■ Diving Behavior ■ Diving Physiology ■ Energetics ■ Strategies for cost-efficient diving by marine mammals. Science 288,
Streamlining 133–136.
Williams, T.M., Friedl, W.A., Fong, M.L., Yamada, R.M., Sedivy, P., and
Haun, J.E. (1992). Travel at low energetic cost by swimming and
References wave-riding bottlenose dolphins. Nature 355, 821–823.
Acevedo-Gutierrez, A., Croll, D.A., and Tershy, B.R. (2002). High feed- Williams, T.M., Fuiman, L.A., Horning, M., and Davis, R.W. (2004).
ing costs limit dive time in the largest whales. J. Exp. Biol. 205, The cost of foraging by a marine predator, the Weddell seal
1747–1753. Leptonychotes weddellii: Pricing by the stroke. J. Exp. Biol. 207,
Aguilar de Soto, N., Johnson, M.P., Madsen, P.T., Díaz, F., Domínguez, I., 973–982.
Brito, A., and Tyack, P. (2008). Cheetahs of the deep sea: Deep forag- Worthy, G., Innes, S., Braune, B., and Stewart, R. (1987). Rapid acclima-
ing sprints in short-finned pilot whales off Tenerife (Canary Islands). tion of cetaceans to an open-system respirometer. In “Approaches in
J. Anim. Ecol. 77(5), 936–947. Marine Mammal Energetics”, (A. Huntly, D. Costa, G. Worthy, and
Au, D., and Weihs, D. (1980). At high speeds dolphins save energy by M. Castellini, Eds), pp. 115–126. Soc. Marine Mammalogy Special
leaping. Nature 284, 548–550. Publication 1. Society Marine Mammalogy, Lawrence, KS.
Davis, R.W., Fuiman, L.A., Williams, T.M., and LeBoeuf, B.J. (2001).
Three-dimensional movements and swimming activity of a northern
elephant seal. Comp. Biochem. Physiol. A 129, 759–770.
Feldkamp, S.D. (1987). Foreflipper propulsion in the California sea lion,
Zalophus californianus. J. Zool. (Lond.) 212(1), 117–136.
SYSTEMATICS
Fish, F.E. (1993). Influence of hydrodynamic design and propulsive
mode on mammalian swimming energetics. Aust. J. Zool. 42, 79–101. Annalisa Berta
Goldbogen, J.A., Calambokidis, J., Shadwick, R.E., Oleson, E.M.,
McDonald, M.A., and Hildebrand, J.A. (2006). Kinematics of for- Systematics is the study of biological diversity that has as its pri-
aging dives and lunge-feeding in fin whales. J. Exp. Biol. 209,
mary goal the reconstruction of phylogeny, the evolutionary or
1231–1244.
Goldbogen, J.A., Calambokidis, J., Croll, D.A., McKenna, M.F., Oleson, genealogical history of particular group of organisms (e.g., spe-
E., Potvin, J., Pyenson, N.D., Schorr, G., Shadwick, R.E., and Tershy, cies). Because of its emphasis on phylogeny, this discipline is often
B.R. (2012). Scaling of lunge-feeding performance in rorqual whales: referred to as phylogenetic systematics or cladistics. Other related S
Mass-specific energy expenditure increases with body size and goals of systematics include determination of the times at which
progressively limits diving capacity. Funct. Ecol. 26(1), 216–226. species originated and became extinct and the origin and rate of
doi:10.1111/j.1365-2435.2011.01905.x. change in their characters. An important component of systematics
Guinet, C., Domenici, P., de Stephanis, R., Barrett-Leonard, L., Ford, is taxonomy that involves the identification, description, nomencla-
J.K.B., and Verborgh, P. (2007). Killer whale predation on bluefin ture, and classification of organisms. Systematics provides a frame-
tuna: Exploring the hypothesis of the endurance-exhaustion tech- work for interpreting patterns and processes in evolution using
nique. Mar. Ecol. Prog. Ser. 347, 111–119.
explicit, testable hypotheses.
Hertel, H. (1966). Structure, Form and Movement. Reinhold Publishing
Corporation, New York, NY. The rapid pace of research on marine mammals has resulted
López, L.M.M., Miller, P.J.O., Aguilar de Soto, N., and Johnson, M. in renewed interest in their systematics. Phylogenetic systematic
(2015). Gait switches in deep-diving beaked whales: Biomechanical methodology as introduced here has gained near universal accept-
strategies for long-duration dives. J. Exp. Biol. 218, 1325–1338. ance. [For a general introduction see Baum and Smith (2013) and
doi:10.1242/jeb.106013. for more detailed discussion of methods see Felsenstein (2004).]