People: John Hutchinson
Despite their unseemly bulk, elephants can hit high speeds - but use an unusual style.
New papers expanding on or related to the original study
- Ren, L., Miller, C., Lair, R., Hutchinson, J.R. 2010. Integration of biomechanical compliance, leverage, and power in elephant limbs. Proc Nat Acad Sci USA, Early Edition (online). See Biomechanical research reveals that elephants move like 4x4 vehicles
- Zioupos, P., Cook, R.B., Hutchinson, J.R. 2009. Response: More thoughts on the relationship between apparent and material densities in bone. Journal of Biomechanics 42:794-795. [pdf]
- Hutchinson, J.R. 2009. Response: Of ideas, dichotomies, methods, and data- how much do elephant kinematics differ from those of other large animals? Journal of Experimental Biology 212:153-154. [pdf]
- Zioupos, P., Cook, R.B., Hutchinson, J.R. 2008. Some basic relationships between cortical and cancellous bone. Journal of Biomechanics 41:1961-1968. [pdf]
- Hutchinson, J.R., C.E. Miller, G. Fritsch, T. Hildebrandt. 2008. The anatomical foundation for multidisciplinary studies of animal limb function: examples from dinosaur and elephant limb imaging studies. pp. 23-38 in H. Endo and R. Frey (eds.), Anatomical Imaging Techniques: Towards a New Morphology. Berlin: Springer-Verlag. [pdf]
- Ren, L., M. Butler, C. Miller, D. Schwerda, M. Fischer, J.R. Hutchinson. 2008. The movements of limb segments and joints during locomotion in African and Asian elephants. Journal of Experimental Biology 211:2735-2751. pdf and associated files (.zip archive) [see Response to G.S. Paul commentary below]
- Miller, C.M., C. Basu, G. Fritsch, T. Hildebrandt, J.R. Hutchinson. 2008. Ontogenetic scaling of foot musculoskeletal anatomy in elephants. J. Roy. Soc. Interface 5:465-476. pdf and associated files (.zip archive)
- Ren, L. and J.R. Hutchinson. 2008. The three-dimensional locomotor dynamics of African (Loxodonta africana) and Asian (Elephas maximus) elephants reveal a smooth gait transition at moderate speed. J. Roy. Soc. Interface 5:195–211. pdf and associated files (.zip archive)
- Weissengruber, G.E., G.F. Egger, J.R. Hutchinson, H.B. Groenewald, L. Elsässer, D. Famini, G. Forstenpointner. 2006. The structure of the cushions in the feet of African Elephants (Loxodonta africana). Journal of Anatomy 209: 781-792. [pdf]
- Hutchinson, J.R., D. Schwerda, D. Famini, R.H.I. Dale, M. Fischer, R. Kram. 2006. The locomotor kinematics of African and Asian elephants: changes with speed and size. Journal of Experimental Biology 209: 3812-3827. [pdf]
- Have you ever seen an elephant...run? RVC Press Release, 2006
Stanford Press Release
March 31 2003
CONTACT: DawnLevy, News Service: (650) 725-1944, email@example.com
Relevant Web URLs:
- Full web version of this story at the Stanford Report
- Stanford University Neuromuscular Biomechanics Lab: www.stanford.edu/group/nmbl/
- Thai Elephant Conservation Center: www.thailandelephant.org
Speedy elephants use a biomechanical trick to 'run' like Groucho
A study published in the April 3 issue of Nature solves a longstanding mystery about elephant speeds by clocking the animals at 15 miles per hour. That's faster than reliable observations of 10 mph top speeds but slower than speculations of 25 mph. But do fast-moving elephants really "run"?
Even at fast speeds, it might seem to the casual observer that elephants don't run. Their footfall pattern remains the same as that in walking, and never do all four feet leave the ground at the same time - a hallmark of running. But biomechanists are finding that an elephant's center of mass appears to bounce at high speeds. If that turns out to be true, an elephant's gait meets the biomechanical definition of running.
Biomechanists have dubbed this gait "Groucho running" after the silly, crouched walk of Groucho Marx. They say the elephants seem to bend their limbs slightly in order to move their bodies more smoothly. This research may provide insight into the biomechanical tricks that help large animals, from extinct dinosaurs to obese people, overcome the physical forces that restrict their motion.
"We do find evidence that elephants run in a sense," said first author John Hutchinson, a Stanford postdoctoral research fellow in the Department of Mechanical Engineering. "It's an intermediate sort of gait, but it looks like what we biomechanically would call running. They don't leave the ground, which is the classical definition, but they do seem to bounce, which is the biomechanical definition."
Last year Hutchinson co-authored another Nature paper that used a computer model of physical forces to show that Tyrannosaurus rex probably was too big to run quickly. For his recent paper, he teamed up with Dan Famini, a veterinary student at the University of California-Davis; Richard Lair, an adviser and international relations director at the Thai Elephant Conservation Center; and Rodger Kram, an associate professor of kinesiology and applied physiology at the University of Colorado-Boulder. They focused on an extant biggie rather than an extinct one: the Asian elephant (Elephas maximus), which can tip the scales at more than 4 tons.
From Africa USA to Thailand
In 1997, Hutchinson, Kram and Famini were all at UC-Berkeley. Kram, the first with colleagues at Harvard to measure the rate of oxygen consumption in walking elephants, was advising Hutchinson and Famini about "normal" elephant biomechanics during the duo's kinematic experiments with African elephants at Six Flags Marine World in Vallejo, Calif. Earlier, Kram had noted that elephants preferred to walk at a slow but efficient speed that gave them what he called the "best gas mileage."
Hutchinson began to correspond with Lair, the author of Gone Astray: The Care and Management of the Asian Elephant in Domesticity, published by the United Nations Food and Agriculture Organization in 1997. Once an elephant trainer at Marine World/Africa USA in Redwood City, Calif., before the park relocated to Vallejo, Lair moved to Thailand in 1980 to help save Asian elephants from extinction. He has trained Asian elephants for films, notably Disney's Operation Dumbo Drop, and now works at the Thai Elephant Conservation Center, which provided crucial support for the Nature study.
"[Hutchinson asked] if I thought Thai elephants could run faster than the speeds he and Dan had got from U.S. zoo and circus elephants [about 10 mph]," recalled Lair in an e-mail interview. "I said that I knew they could because I had timed them much faster at the Surin Elephant Round-up in northeast Thailand in 1984."
Thanks to a traveling fellowship from the Journal of Experimental Biology, Hutchinson and Famini went to Thailand in 2000 and 2001 to put some elephants to the test.
For their experiments, Hutchinson palpated the animals' limbs to find their joints, and then the duo marked the joints with large dots of water-soluble, nontoxic paint. They videotaped 188 trials of 42 Asian elephants walking and running through a 100-foot course and measured their speed with photosensors and video analysis.
The average walking speed was 4.5 mph. But 32 of the elephants moved faster than previously documented - up to 15 mph. Three were especially fleet of foot, exceeding 15 mph - 50 percent faster than anyone had ever reliably recorded, Hutchinson said.
Past references gave anecdotes, not data. The result was a lot of confusion about elephant speeds.
"The vast majority of statements regarding the maximum speed of African elephants descend from one of two apocryphal hunches dating back over 60 years," Famini wrote in an e-mail.
Said Hutchinson: "Here we actually have the videotape and data to back it up, whereas with an anecdote, like some big game hunter clocking an elephant with a speedometer on a car, it's just not reliable."
Seeing was believing - these elephants were fast. "When I saw the speed trap times and videos I was convinced," Kram wrote in an e-mail. "I ran the mile in 4:30 when I was in high school and I am still a competitive Master's runner. I can only just barely sprint as fast as the fastest elephants we measured."
To run or not to run - that is still the question
So what turns a walk into a run? It isn't just speed, although that plays a part.
Kinematically, one thing that distinguishes walking from running is the footfall pattern. Typical quadrupeds use a walk at slow speeds, a trot at medium speeds and a gallop at fast speeds.
In the footfall pattern of a trot, diagonal limbs contact the ground at the same time. "So a quadruped goes left hind, right front together and then right hind, left front together," Hutchinson explained. "It's acting like a biped."
In contrast, in the footfall pattern of a gallop, the two hindlimbs touch the ground one after the other, followed by a pause, after which the two forelimbs touch the ground one at a time. If an animal's feet are on the ground less than half of the time, Hutchinson said, it meets the kinematic definition of running.
But elephants are weird because no matter how fast they go, their footfall pattern doesn't change. They use a walking footfall pattern even at 15 mph, the researchers found. That pattern has the left hind foot moving first, followed by a brief pause, after which the left front foot moves. Then there's a long pause, after which the same thing happens on the right side.
An all-aerial phase - where no feet are touching the ground - also kinematically differentiates running from walking. But elephants never have all their feet off the ground.
"Elephants probably don't run with an aerial phase because it would be too mechanically stressful on their bodies," said Hutchinson. It turns out that a lot of other animals - including running birds like chickens, emus and rheas - have limbs that release elastic strain energy like the rebound of a stretched rubber band without ever getting propelled so forcefully that all feet are off the ground at the same time.
That led biomechanists to redefine running more than 30 years ago to better describe the physical forces at work, Hutchinson said. "We're just beginning to understand which animals can break the rules and bounce without leaving the ground - and how they do it."
A deeper biomechanical mechanism may explain running better than the aerial phase frequently observed. All legged land animals, Hutchinson said, "whether they have two legs, four legs, six legs or even in the case of a centipede, 42 legs," use the same mechanism to switch from a walk to a run. That switch often occurs at the same relative point, or Froude number, which is a dimensionless measurement that gives an animal's speed relative to its hip height. So even though a cockroach has shorter legs than an elephant, in terms of how many body lengths it can move in a certain amount of time, it may still scurry with greater relative speed than a charging elephant.
"At the same Froude number, any animal, regardless of size, should be moving with the same mechanism," Hutchinson said. "It should be exerting itself in relatively the same way." A Froude number of 0.01 is slow for any animal; a Froude number of 20, fast. Most animals should switch from a walk to a run at about the same Froude number, at 0.5 or so, he said. Animals shift from a walk to a run because at faster speeds walking becomes less energetically efficient, or more mechanically stressful, than running, he said.
"The stunning thing about our elephants is they were going at a Froude number as fast as 3.4, which is over three times the dimensionless speed that an elephant should be switching from walking to running," Hutchinson said. "A horse would be well into a gallop by this point. But the elephants were still using a walking footfall pattern."
If you think of the body abstractly as just a stick swinging back and forth as it moves, he explained, its center of mass moves differently during walking compared to running. "Walking is a stiff, pendulum-like gait; the limb stays pretty straight and swings back and forth. Running is a bouncing gait in which the limb actually compresses and bounces back with a spring."
The researchers' kinematic measurements suggest that fast-moving elephants may switch from a pendulum-like gait to a bouncing gait. If they do, they fit the biomechanical definition for running.
But there's only one way to find out for sure. The animals would have to move across a force platform - a special device that registers the forces that elephants exert on the ground - to see if their center of mass swings like an inverted pendulum (as in walking) or bounces like a spring (running).
"That's a problem because the force platforms that are generally available would break if an elephant ran across them," Hutchinson said. "That's been the obstacle for years. That's one reason why no one has ever done it."
He and Kram are building a prototype force platform in Colorado to answer once and for all if elephants can run. So there's still time to place your bets.
By Dawn Levy
It is generally thought that elephants do not run, but there is confusion about how fast they can move across open terrain and what gait they use at top speed. Here we use video analysis to show that some Asian elephants (Elephas maximus Linn.) can move at surprisingly high speeds of up to 6.8 m/s (25 kph) and that, although their gait might appear to be a walk even at this speed, some features of their locomotion conform to definitions of running.
Simplified Summary of Our Work
It was poorly known how fast elephants can move, and whether they run in any sense. To solve these two mysteries, we worked with 42 Asian elephants in Thailand that were known for being healthy and quick. We used video analysis to measure the average speed of each elephant across the middle 10 meters of a 30 meter long course. We found that the elephants routinely went much faster than previously documented (~4 m/s, 10 mph, or 16 kph), up to speeds as high as 6.8 m/s (~15 mph, 25 kph). At these faster speeds, the elephants still did not leave the ground, seeming like they might have still been walking even at such fast speeds. Yet we looked more closely at their motions at those speeds and found several subtle lines of evidence that -- in a strict biomechanical sense -- they were running, not walking. Running is a bouncing gait rather than a rigid-legged gait, so it seems likely that elephants “bounce” at faster speeds. We cannot conclusively state that elephants can run, because we need measurements from force platforms to do that, and unfortunately conventional force platforms would break if used with elephants. However, it seems likely that elephants can run, and the force platform studies are being prepared.
- Download The Nature Paper: Are fast moving elephants really running? (PDF)
- Supplementary Information:
- News Article in Nature (see Links tab for more)
What did you find out?
- We dispel the myth that, at least in the case of Asian elephants, they can move at 25 mph. 15 mph (6.8 m/s or 25 kph) seems to be near the limit of their speed.
- We were surprised to discover hints that elephants "bounce", fitting the modern biomechanical definition of running. We have not proven this, but it seems likely.
Other FAQs (see below for definitions)
Why did we do it? The basic principles governing locomotion in animals larger than a horse remain poorly understood. Elephants are the best choice to begin studying as they are the largest land animals. Besides, elephants are strange and fascinating animals that are a joy to work with.
Why should the public care? Asian elephants are endangered; we should learn from them while healthy, active populations are still available to study. The general principles that we can learn from large animals like elephants apply even to humans, as these principles are about factors such as size, speed, strength, and stability. Also, the public should know about the way biomechanists define running today: as a bouncing spring-mass system rather than as necessarily having an aerial phase.
How did we do it? The elephants were motivated to move fast with food rewards, comfort, competition, cheering, playful chasing, other positive reinforcement, etc. One elephant, Pop, was so excited by the experiments that she had to be held back from running down the course, and broke free at full speed twice (once making one of our speed records). The scope of the project was broad: we selected 42 elephants that were known to be quick and healthy from over 300 potential subjects in Thai elephant camps. Thus we feel that we have found a reasonably close estimate of the maximum speed of Asian elephants, although it is likely some few can move slightly faster.
What next? Force platforms are being designed to definitively solve the running question. We also hope to work with African elephants as some have said they are faster than Asian elephants. Computer models can be used to test what features limit elephant speed and gait. Later, research on other large animals such as hippos, rhinos, and giraffes is needed.
Details for elaboration
SPEEDS: We provided the first well-documented study of the locomotion of elephants, the largest land animals. This subject has been surrounded by confusion and misinformation. The confusion about elephant locomotion extends back to the late 1800s, involving famous researchers such as Eadweard Muybridge (who was among the first to use cinematography, and so won a bet for Leland Stanford that all four legs of a horse leave the ground during a trot) and continuing to the present, when studies of animal locomotion cite widely different maximum speeds (and gaits) for elephants.
The myth that all elephants can move at 25 mph (and still walk!) mainly dates back to one anecdote from Roy Chapman Andrews (famed American Museum of Natural History paleontologist/archaeologist and a model for Indiana Jones). We still hear assurances from people familiar with fast African elephants that they can move 25 mph, but speeds are difficult to measure accurately. (we have a longer version of this speed story in essay form posted on the website)
This 15 mph speed is not that fast, but is 50% faster than previous reliable estimates at ~10 mph. For comparison, a typical high-school sprinter at 20 mph, or a champion Olympian 100m dash at 25 mph average speed, could outrun an Asian elephant without difficulty. An average human would be fairly evenly matched with a fast, healthy elephant in a footrace on level ground.
RUNNING: Elephants might be running at their faster speeds. It depends on the definition of "running" applied. The old, classical definition familiar to most people is that running is when all feet are off the ground at once (called an aerial phase) during a stride. Biomechanical engineers and biologists favor a more mechanistic definition today, which has come into favor over the last 40 years. This kinetic definition models the center of mass of the body as a rigid pendulum-like mechanism (in walking) vs. a bouncy spring-like mechanism (in running), explainable as follows:
In walking, kinetic energy and potential energy are out-of-phase: when the foot is on the ground, the center of mass swings up to its highest point (trading kinetic energy for potential energy), then back down again (regaining kinetic energy) when the foot leaves the ground.
In running, kinetic energy and potential energy are in-phase, so they cannot be exchanged. Instead, as the center of mass descends once the foot hits the ground, the leg(s) becomes compressed, storing elastic strain energy in its tissues (tendons, muscles, ligaments, bones, etc.). Rather than the center of mass being highest at the middle of ground contact, it is lowest, and the "leg-spring" is maximally compressed. Then, later in the step, the leg-spring rebounds, providing energy to raise the center of mass back up. Visualizing a pogo stick's gait might help you understand how running works. If the kinetic energy provided during bouncing on the ground is high enough, the body may leave the ground for an aerial phase. But that aerial phase is not necessary, and in many animals it does not occur.
Crabs and other many-legged animals bounce without leaving the ground, and lately scientists have found that other animals do this, including birds running at medium speeds, and humans who "Groucho run" by running with very flexed knees. Elephants might be doing this. We found three lines of evidence that suggest this:
In elephants, the dimensionless speed or Froude number, a ratio of the kinetic energy (or inertial forces) and potential energy (or gravitational forces) during locomotion, reaches values as high as 3.4. Theoretically, at Froude numbers greater than 1.0 animals should have to switch from walking to running, as the forces drawing them off their circular (pendular) path exceed the gravitational forces pulling them down. Indeed, most animals switch from walking to running at a Froude number of ~0.5, because it is either mechanically less stressful or energetically more efficient to switch gaits at a lower speed. But elephants surpass Froude numbers of 0.5, 1.0, and higher without their motions showing an obvious change of gait.
For the Froude number, think about it this way: if you are late for the bus and start walking faster as you realize that you might miss it, at some speed you cannot walk any faster. If you go faster, you must leave the ground because you fly right off the ground (inertial / centripetal forces overcome gravitational forces holding you down). "Groucho running" can be done without an aerial phase, unlike normal running, because the flexed legs are springier and can stay on the ground longer than more rigid legs. Humans can "Groucho run" to a Froude number of about the same as a really fast, small elephant: ~3. This is the same Froude number that a typical Boston marathon runner might use, in that case jogging rather than "Groucho running". In the upcoming Kentucky Derby, at 40 mph those horses will be using Froude numbers of about 20; much more relative exertion than any human runner or elephant can do. A cheetah at 70 mph might have a Froude number of about 30 or more.
Another explanation is that if a small child walks alongside a tall adult, because their legs are shorter, they must switch from a walk to a run at a lower speed than the adult. The Froude number is equal to the speed-squared, divided by hip height and gravity, so this makes sense. Shorter legs (lower hips) must switch gaits at slower speeds, for the same reasons as explained above. Elephants have long legs for their size, so perhaps this helps them move fast, maybe even run, without leaving the ground.
Clues that elephants may run at faster speeds include our observation that each foot spends less than 50% of the time on the ground (the duty factor). In most animals, this duty factor would require elephants to use an aerial phase -- e.g., a human whose feet are each only on the ground for 40% of the stride, a duty factor of 0.40, are both off the ground for 20% of the time, requiring an aerial phase. But the evenly spaced footfall pattern of elephants (see below) keeps at least one foot evenly on the ground, even at higher speeds with duty factors as low as 0.37. At one point in their stride, a fast elephant supports itself on only one leg. If it is a forelimb, that leg is held vertical and rigid, while the hindlimbs are airborne. If it is a hindlimb, it is more bent and compliant (looking like a spring), while the front of the elephant is airborne.
This ties into our Fig. 1b. While its foot is on the ground, the shoulder joint goes up then down regardless of speed, and the hip joint does the same thing until higher speeds, but then goes down and often back up. This is a strange mechanism that has suggested to some that the forelimbs are walking while the hindlimbs are running. We don't know if that's actually possible. And this is one reason why we can't say that elephants are definitely running, as the two sets of legs work oppositely and could change the mechanics in a major, unanticipated way. But I think it's rather likely, as the Froude number is so high, the duty factor is so low, and the hindlegs are so bent. We might be wrong, though; locomotion works in mysterious, complex ways.
For elephants, if they do indeed run, then this means they bounce. "Bounce" is not a word one often thinks about when considering elephants. "Rigid" or cumbersome more often come to mind. But if we are correct and the hints we see from the motions are hints of running, then indeed elephants do bounce.
On the running vs. walking definition, we see this as an opportunity to show the public that the old definition of running (aerial phase) is not satisfactory, and has fallen out of favor in the last 40 years. The modern definition shows that the center of mass is lowest at the midpoint of a step in running, vs. highest in walking. This requires that the legs act like a spring in running, vs. like a rigid pole in walking. Imagine a bouncing ball compared to a walking toy soldier and you can see the difference. When one realizes that animals of 2, 4, 6, or even 42 legs use this bouncing mechanism at higher speeds, and even switch from walking to that gait at the same relative speed (Froude number), one can see why the biomechanical definitions work better to unite how these animals work. Many of these animals (especially those with many legs) do not leave the ground at faster speeds, but still do use bouncing mechanisms.
One piece of evidence that might seem contradictory to our "running" speculation is the way the feet hit the ground. The same sequence (Fig. 1a in our paper) applies at all speeds for all elephants: 1 (brief pause), then 2, longer pause, 3 (brief pause), then 4 (imagine a 4-legged dance step, 1-2, 3-4). In elephants this 1-2, 3-4 pattern involves the legs on the same side of the body (1-2 = left side, 3-4 = right side for example). This footfall pattern is called a "lateral sequence singlefoot with lateral couplets" (Muybridge helped define this sequence; at faster speeds he called it the "amble", which is now an outdated term). It was previously recorded almost exclusively in animals that lacked an aerial phase and were thought to be walking. Yet as we now know from more studies of animals, that walking definition does not work well, because a walking footfall pattern does not necessarily coincide with a rigid, toy soldier-like gait.
Gait = a discrete pattern of locomotor movements. Running, walking, and hopping are the three main terrestrial gaits.
Stride = a complete cycle of footfalls, e.g., from the time the left hind foot hits the ground to the next time it hits the ground. In a walking biped, a stride consists of 2 steps, whereas in a quadruped it consists of 4 steps.
Kinetic energy = related to the mass and velocity of a body (e.g., inertia or momentum), it is the total energy propelling the animal in the direction of its velocity heading.
Potential energy = the energy an animal or other object has because of its height from the ground, its mass, and gravity.
Center of mass = a point at which the entire mass of an animal or other object can be abstracted as lying in 3D space.
Duty factor = the fraction of a stride that a given foot is on the ground.
Classical running definition = a gait in which there is an aerial phase, or the duty factor is less than 0.5. Walking therefore involves no aerial phase, and a duty factor 0.5 or more.
Kinetic running definition ("running" as defined here and in science in general today) = a gait in which the center of mass is lowest at the middle of ground contact during a stride. This mechanistic definition models the animal's limb and body as a spring-mass (bouncing) system, storing and releasing elastic strain energy to save energy while kinetic and potential energies oscillate in-phase and the body moves forward.
Kinetic walking definition = a gait in which the center of mass is highest at the middle of ground contact during a stride. This mechanistic definition models the animal's limb and body as a rigid pendular system (technically, an inverted pendulum). Kinetic and potential energies oscillate out-of-phase to save energy while the body moves forward.
"Groucho running" = running with more bent limbs. This unusual form of running gait often lacks an aerial phase. It is known among many animals with more than four legs, as well as some four-legged animals (lizards, sheep, opossums, and others) and bipeds (birds at medium speeds, comedic humans such as the namesake Groucho Marx).
Shuffling Through the Past:
The muddled history of the study of elephant locomotion
By Dan Famini (UC-Davis), with help from John R. Hutchinson (was at Stanford; now at RVC)
(copyright D. Famini and J.R. Hutchinson, 2003)
[This essay is intended for educational use, not as a scientific publication, and we ask that contents of this paper are only used in other venues with our permission]
As the largest extant animal on land, there is a mystique about elephants that makes them an obvious and interesting topic of study. Unfortunately the locomotion of elephants is poorly understood, even basic parameters such as maximum velocity (Table 1). This is largely because finding active elephants and accurately measuring their speeds is a difficult, and potentially dangerous, undertaking. Many accessible elephants of captivity are sedentary compared to their wild counterparts, and seem incapable of providing information on the species' peak locomotor performance. On the other hand, the conditions and the elephants found in the field are inhospitable towards accurate and reliable measurement techniques. Consequently, the majority of data pertaining to elephant locomotion, particularly maximum velocity, comes not from scientific information, but from the educated guesses of individuals with elephant experience. Anecdotes are subjective, and anecdotes regarding elephants are particularly prone to overestimation due to the impressive stature of the animals. Inaccuracy in both estimating elephant speeds and in referencing the estimates of earlier authors has created an inflated and confused idea of the speed of elephants. What gait(s) elephants use at their fastest speeds (and in some cases any speed) also is a subject of much confusion in the popular and scientific literature. We review the history of these speed estimates and gait descriptions briefly here in order to clarify this muddled history of study. Hutchinson et al. (2003,2006) provide a modern assessment of both speed and gait in Asian and African elephants.
The earliest known study of elephant locomotion was conducted by Eadweard Muybridge (1899) who captured several trials of an Asian Elephant using high speed photography. Muybridge tried to obtain "as fast a speed as vigorous persuasion could induce -- equivalent to a mile in somewhat less than seven minutes (3.4 m/s)." Later analysis of these same images estimated the fastest photographed trial at 3.8 m/s [] (Alexander 1979). Muybridge made no claim to have captured an elephant's top performance, and for the maximum possible velocity Muybridge quotes an estimate by Sir Samuel Baker (1890). Sir Baker had written a two volume memoir of big game hunts around the world. Within these narratives, Baker comments that the "African elephant is capable of a speed of 15 miles an hour (6.7 m/s), which it could keep up for 200-300 yards, after which it would travel at about 10 miles per hour (4.5 m/s)."
A significantly faster estimate of 24.5 mph (11.2 m/s) comes from a 1937 article in Natural History by Roy Chapman Andrews. The article's focus is actually about the speed of deer bot flies, Cephanemyia pratti, which Andrews claims can reach a velocity of 818 mph. That preposterous statement has since been proven to be erroneous (May 1999). Within Andrews' article there is a large chart containing the maximum speeds of many animals, people, and vehicles. A large variety of methods are used, predominantly automobile speedometers, but no method is ascribed to the estimate for elephants. The only details Andrews gives are that the elephant was charging over a length of 120 yards. Howell (1944) claims that Andrews' estimate is "based on the experience of one of the staff." Andrews was the Director of the of Natural History at the time, so it is entirely plausible that his information came from one of his employees, or that Howell was simply referring to Andrews himself. Andrews also mentions that "Dr. William K. Gregory, of the of Natural History, has revealed an interesting relationship between anatomy and speed." Gregory did publish a related article (1912) but it is written from an evolutionary context, with reports on sizes and ratios of the muscles and bones of many animals. It contains no speed estimates for elephants. The 11.2 m/s "estimate" from Andrews' chart, without any substantiation, became the standard elephant speed estimate for at least forty years.
One of the most widely cited sources for the maximum velocity of elephants is Howell's book, Speed in Animal's (1944). Howell quotes both Andrews' estimate of 11.2 m/s over 120 yards and Muybridge (1877) for 6.7 m/s over 200-300 yards (which is actually Baker's estimate). These numbers were portrayed as the best available estimates, yet Howell recognizes that "There are no authentic figures on the maximum speed of an elephant" (p. 52). Many authors of books about elephants or animals in general (and not locomotion specifically) refer to these values. Breeland (1948) states that "one African elephant has been timed with a stop watch at 24 mph for 120 yards." Breeland mentions both Howell's book and the journal Natural History in his selected references, but it is unclear where he got the information regarding this timed trial. Andrews had noted all of the times measured by watches, but the elephant's time was not one of them. Bourliere (1964) quotes the 11.2 m/s as from Howell, without any disclaimer. A rounded form of the 11.2 m/s estimate appears in Beebe (1968), Van Gelder (1969), Borst (1969), and (1977). All four of these works are about animals or elephants in general, and none offer any further explanation or source for the maximum velocity estimate.
The 7 and 11 m/s estimates are not called into question until a journal article about high-speed buffalo and elephant locomotion by Alexander (1979). He points out that "it seems likely that the casual observers who have seen elephants charging will be liable to exaggerate their speeds." There is no mention of Andrews or Baker, and the numbers are presented as estimates that "Howell reports". However, Alexander (1979) does measure an elephant's speed from video analysis at 4-4.5 m/s; the first credible estimate since Muybridge (1899). Garland (1983) also expressed doubts about the existing estimate of 40 km/h (~11.2 m/s). He alludes that "W.P.Coombs (pers. comm.) doubts the credibility of the higher figure so I have used 35 km/h (9.7 m/s)." However, provides no rationale or support for why he chose to particular value. Christiansen (2002) went so far as to exclude elephants from his scaling study of mammal maximum speed and limb proportions, "owing to their apomorphic appendicular anatomy and mode of locomotion." (p.687). This was a view also held by Paul (1998), who claimed that "Because elephants are unable to run, and must always walk with at least one foot contacting the ground, they should not be able to exceed ~20 km/h. A racing large Asian elephant bull (in a video) won with a peak speed of 19 km/h" (p.259). Paul also asserted that elephant-sized extinct animals, including dinosaurs, refuted the popular notion that large animals cannot be fast, suggesting that elephants were not reliable indicators of the size limits on speed and gait. In contradiction, in another speed-scaling study Iriarte-Díaz (2002) argued that larger animals were very limited in their range of locomotor performance. The latter study entered speed data of 9.45 and 7.02 m/s for African and Asian elephants respectively, citing two mammalogy texts and (1983) as sources. Have anecdotes surreptitiously become data? Are they trustworthy? An analogy could be made that allowing sloppy speed estimates into a biomechanics study is akin to a morphometrics study measuring a bone's length from across a museum hall with a ruler. Standards for these data should be carefully considered by authors and reviewers.
Doubts regarding older anecdotal quotes have only caused greater confusion and the replacement of old anecdotal information with new anecdotes. In the last decade sources not specifically about locomotion have reported a variety of maximum velocities for African elephants. Jackson (1990) and Groning (2000) reported 25 km/hr (6.9 m/s). Estes (1991) and Spinage (1994) reported 30 km/hr (8.3 m/s). Le Rue III (1994) quotes 40 km/hr (11.2m/s), but remarkably the number did not descend from Andrews (1937) but rather from a Dr. Griznemek who reportedly clocked the value with a Land Rover speedometer. One of the newest books on elephants contains the most extreme claim yet as S. Alexander (2000) states that "a mature African bull can achieve speeds 30 mph (13.3 m/s)." While S. Alexander refers to Ivan T. Sanderson for quotes on the velocity of Asian elephants (max of 15 mph (6.67 m/s)), there is no reference offered for the velocity of African elephants. As might be expected, the greater overestimates are being found in less scientific sources. However, these exaggerations also reflect the lack of consensus within the expert community, and most importantly the dearth of scientific studies on the subject.
Evaluating the way in which elephants move is far more difficult than measuring speed. In addition to the limited access to elephants, the complexity of quadrupedal movement is a challenge to measure, understand, or describe. There are few scientific reports on elephant locomotion, and none have fully quantified and described the kinematic and kinetic variables that define gaits. The lack of accurate and detailed information has been compounded by ambiguous and conflicting gait terminology. Consequently, a wide variety of gaits have been attributed to elephants. These include the walk, running walk, run, shuffling run, shuffling trot, shuffle, amble, ambling walk, rack, and pace. In addition there is a lack of consistency as to what the different gait names mean. Even with this many terms for elephant gaits, the same word often has different meanings depending on the author.
Muybridge (1899) is the oldest known source on the classification of quadrupedal gaits in general, including elephant gaits. He states that elephants only "walk" and "amble" (p. 67 of 1957 edition). Muybridge defines gaits in terms of footfall pattern and number of limbs supporting the animal at different times throughout a stride. In both the walk and amble each foot moves sequentially with the hindfoot moving before the forefoot on each side. The number of supporting limbs alternates between 2 and 3 in the walk, and between 1 and 2 in the amble. Muybridge states that "Practically, [the amble] is an accelerated walk." Muybridge claims that one of his photo sequences (plate #27) of an Asian elephant walking is an amble, but it appears to actually have at least 2 limbs of support at all times.
Howell (1944) substitutes the term "running walk" for Muybridge's "amble". In addition, Howell differentiates between "slow" movements and their quicker counterparts. Throughout a stride, a "slow walk" has 4, 3 or 2 limbs supporting the animal at different instances. A "fast walk" has either 3 or 2 supporting limbs. A "slow running walk" has 3, 2 or 1 supporting limbs, and a "fast running walk" has 2 or 1 supporting limbs. In apparent contradiction, Howell states that elephants rely "exclusively upon the walk or its more speedy equivalent, the running walk, which permits it to keep at least two feet always upon the ground" (p.52). Howell views the sequence of footfalls with the left foot first, calling the Left Front, Right Hind, Right Front, Left Hind sequence used by elephants a "diagonal walk" (p.228). Further, he explains that, "The running walk comes natural to very few animals. In diagonal sequence, it is the exclusive speed gait of the elephant" (p. 232).
Bourliere (1964) returns to Muybridge's term of "amble" claiming it as "the chief manner of rapid locomotion" for Proboscidea (p.4). However, Bourliere states that "in the slow amble, the body is alternately supported by two or three diagonal or lateral legs, but the swifter the animal moves, the more the diagonal pattern is replaced by a lateral one" (p.4). This deviates in both limb support and footfall sequence from the Muybridge's definition of "amble". Alexander (2000) also describes what appears to be a Muybridge "amble". On page 52 he simply states that elephants can "only walk or amble, keeping at least one foot always on the ground."
The term "amble" has also been used to describe a completely different gait, in which both limbs on the same side move synchronously. This gait is used commonly by giraffes and camels, and Muybridge called this gait the "rack" or "pace". In fact, Howell chooses to use the term "running walk" for an accelerated "walk" at least in part to avoid confusion as the term amble is "sometimes used to designate the pace" (p.231). Van Gelder (1969) states that "Elephants, giraffes, camels, hyenas, and some young dogs move both legs on the same side at the same time in a gait called pacing, or ambling" (p.35).
Beebe (1968) claims that elephants cannot run because "elephant legs do not have the spring necessary for running." Eltringham (1982) also states that elephants cannot run, but because "in the accepted sense since it must keep one foot on the ground at all times." Even so, he uses the words "shuffling run" to describe fast elephant movement. According to Eltringham, the shuffling run is distinct from "a fast extended walk during which it takes maximum strides." Similarly, Spinage (1994) reports that "the elephant can neither jump, trot, canter, or gallop" and "movement is restricted to a walk" (p.46). However, he a calls the elephants fastest movement a "shuffling trot" (p. 43). To avoid self-contradiction Spinage must consider the "trot" and "shuffling trot" different types of movement, yet he offers no definitions or means to discern the two gaits. Gale (1974) also concurs that elephants "cannot leap, trot, gallop, or canter", calling the elephant gait a "shuffle."
Gambaryan (1974) uses the term "fast walk" that would seem to be synonymous with Howell's running walk. However, Gambaryan's observations about the footfall pattern are actually more along the lines of Bourliere's. He describes that "a change in the rhythm of locomotion toward a rack is characteristic for elephants during slow movement, while a reverse switch of rhythm is typical during accelerated motion" (p. 167). Gambaryan goes on to knowingly (but inappropriately) use "running". According to his own definition, "running is a form of high-speed motion with a stage of flight in the air. This does not apply to elephants, but still we think it is worthwhile calling the fastest form of this animal's locomotion running" (p. 168).
In a paper on fast locomotion, Alexander (1979) also uses the term running to describe elephant locomotion. Alexander (1982) describes eight different quadrupedal running gaits. He clarifies that in bipeds, running differs from walking in that when running the leg acts as a spring with the hip reaching its lowest point during mid-stance. In contrast, in walking the hip reaches its highest point at mid-stance. It is implied that these distinctions hold for quadrupeds. Alexander explains that the "amble is the unusual running gait of elephants." In Alexander's "amble", the limbs have relative phase relationships "the same as for a typical quadrupedal walk" of 0.0, 0.25, 0.5, 0.75 with forelimbs following hind on the same side, as per Muybridge. He does not state what other gaits an elephant may be capable of, just that "Elephants generally use the amble instead of the trot and have no faster gait" (p.100). Estes (1991) uses the term "ambling walk" to define elephants' only gait, but since he offers no definition it is unclear if this is synonymous with other authors' amble or walk. In general, "amble" has become a sloppily applied and vague term that might best be abandoned in gait analysis.
Hildebrand (1985) uses limb phase relationships to distinguish different quadrupedal running gaits. Like Alexander's term "duty factor", Hildebrand also distinguishes running from walking based on whether the foot is in contact with the ground for more than or less than half of the stride. However, Hildebrand has his own term called the "singlefoot" for gaits in which "consecutive foot falls of the four feet are about equally spaced in time" (p. 40). Hildebrand distinguishes between diagonal and lateral sequences, but his definitions for the terms are the opposite of Howell's. Hildebrand's gait terminology consists of all three descriptors: walking/running, gait and order of sequence (lateral vs. diagonal). Thus, according to Hildebrand, "the running singlefoot in lateral sequence is a smooth gait used by elephants" (p.40).
Another term, charging, while not a scientifically recognized phrase, has a significant impact on how elephant information has been presented. It is no coincidence that all of the fastest velocity overestimates all mention that the maximum velocity is of a "charging" elephant (Baker, Andrews, Le Rue III, etc.). The phrase "charging" is not used or defined in any of the texts exploring animal locomotion. Where the phrase has been employed it appears to be used as if it were an explanation for an otherwise illogically fast speed.
A superb example of the power of anecdotal misinformation comes from Sir J. E. Tennet's 1867 book The Wild Elephant. On page 41 Tennet makes the assertion that "I am disposed to think that the elephant is too weighty and unwieldy to leap" and that the elephant cannot "gallop", but rather "shuffles." The reason for Tennet's lack of conviction regarding his idea that elephants cannot jump is an account in the 1866 Colombo Observer claiming that an angered bull elephant "fairly leaped the barrier, of some fifteen feet in high, only carrying away the top cross beam with a great crash." Tennet was skeptical of the claim, and questioned the accuracy of the report. In the book's preface is a letter from the office of the Observer who did some further investigation, reporting that "the result is the usual one whenever exact measurements are substituted for guess-work." The barrier the elephant scaled had not been a full 15 feet, but 12 feet in full, and only 9 below the top bar. In addition there was a 2.5 foot mound behind the fence, meaning the elephant only had to scale a meager 6.5 feet. Elephants, particularly a full-grown "Tusker", are quite capable of having limbs over 6 feet in length, rendering a 6.5 foot barrier little barrier at all. Tennet's correct conclusion that elephants could not leap had been undermined by unreliable observational misinformation. Tennet's situation was not unique. For as long as elephant locomotion has been studied there have been over-estimations of elephant's abilities. While modern technology has replaced visual guesses with guesses from Land Rover speedometers today, we are still hindered by "guess-work" and in a severe shortage of "exact measurements."
The understanding of elephant locomotion has been doubly stymied. First, there is the inaccessibility of the animals themselves, which has led to more anecdotal guesses and the spread of misinformation. Second, the lack of a consistent language to describe the movements of elephants has hindered the spread of what knowledge does exist. The purpose of our studies (Hutchinson et al., 2003,2006) were to accurately quantify and describe the basic kinematics of elephants and inspect whether they showed signs of a gait transition. Future work is needed to test whether they truly run, why their speed and gait are limited, and how their unusual locomotion evolved, but some clarity regarding the speed and gait of fast elephants is emerging at last.
- Andrews, R.C. (1937) Wings Win. Natural History, (October 1937) v. 40 p. 559-65.
- Alexander, R. McN. (1979) Mechanical stresses in fast locomotion of buffalo (Syncerus caffer) and elephant (Loxodonta africana). Journal of Zoological Society of London, (1979) v. 189 p. 135-44.
- Alexander R. McN. (1982) Locomotion of Animals. Blackie & Son Ltd., Glascow UK.
- Alexander, S (2000) The Astonishing Elephant. Random House, New York.
- Baker, S.W. Sir. (1890) Wild Beasts and Their Ways. 2 volumes. Macmillan, New York.
- Beebe, B.F. (1968) African Elephants. David McKay Co, Inc., New York.
- Borst, J. (1969) A Field Guide to The Larger Mammals of Africa. Houghton Milon Co., Boston.
- Bourliere, F. (1964) The Natural History of Animals, 3rd ed. Alfred E Knopf., New York.
- Breeland, O.P. (1948) Animal Facts and Fallacies. Harper & Bros., New York.
- Christiansen, P. (2002) Locomotion in terrestrial mammals: the influence of body mass, limb length and bone proportions on speed. Zoological Journal of the Linnaean Society (2002) v. 136 p. 685-714.
- Eltringham, S.K. (1982) Elephants. Blandford Press, Poole Dorset, UK.
- Estes, R.D. (1991) The Behavior Guide to African Mammals..
- Famini, D., , J.R., and Kram, R. (1999) Locomotion kinematics of African elephants. American Zoologist (1999) v. 39 p.84A.
- Gale U.T. (1974) Burmese Timber Elephant. Trade Corp. 9, Printer: Toppan Print, Co.
- Gambaryan, P.P. (1974) How Mammals Run: Anatomical Adaptations. Wiley, New York.
- Garland, T. (1983) The relation between maximal running speed and body mass in terrestrial mammals. Journal of the Zoological Society of (1983) v. 199 p. 157-70.
- Gregory, W. (1912) Notes on principles of quadrupedal locomotion and the mechanisms of limbs in hoofed mammals. Annals of the New York of Sciences (1912) v. 22 p. 267-294.
- Groning, K. (1998) Elephants. Kohemann Verlagsgesellschaft.
- Hildebrand, M. (1985) Functional Vertebrate Morphology: Chapt. 3 - Walking and Running. The Belknap Press.
- Howell, A.B. (1944) Speed in Animals. of Press, Chicago.
- Hutchinson, J.R., D. Schwerda, D. Famini, R.H.I. Dale, M. Fischer, R. Kram. 2006. The locomotor kinematics of African and Asian elephants: changes with speed and size. Journal of Experimental Biology (2006) v.209 p.3812-3827.
- Hutchinson, J.R., Famini, D., Lair, R., and Kram, R. (2003) Are fast-moving elephants really running? Nature (2003) v. 422 p. 493-494.
- Hutchinson, J.R., Famini, D., Kram, R., and Lair, R. (2002) Do elephants run? American Zoologist v. 41 p. 1479.
- Iriarte-Díaz, J. (2002) Differential scaling of locomotor performance in small and large terrestrial mammals. Journal of Experimental Biology v. 205 p.2897-2908.
- Jackson, P. (1990) Endangered Species: Elephants. Chartwell Books.
- Le Rue III, L. (1994) A Portrait of the Animal World. Todtri Prod. Limited, New York.
- May, M. (1999) Speed demons. The Sciences (January/February 1999) p. 16-18.
- McKay, G. (1973) Behavior and Ecology of the Asiatic Elephant in S.E. Asia. Smithsonian Institute Press< Washington.
- Muybridge, E. (1899) Animals In Motion. Dover Publications Inc., New York.
- Paul, G.S. (1998) Limb design, function and running performance in ostrich-mimics and tyrannosaurs. Gaia v. 15 p. 257-70.
- Sikes, S. (1971) The Natural History of the African Elephant. American Isevier Pub. Co, New York.
- Sikes, S. (1994) Elephants. T & AD Poyser Ltd.
- Tennet, Sir J. E. (1867) The Wild Elephant. Longmans, Green, and co.
- Van Gelder, R.G. (1969) Biology of Mammals. Charles Scribner's Sons. New York.
- Williams, J.H. (1950). Elephant Bill. Doubleday, Garden City, NY.
Table 1 Summary of Elephant Speed Accounts
|Le Rue III||1994||1||1.6||0.44|
|Alexander, S||2000||2.5||4||1.11||Asian normal|
|Alexander, S||2000||3||4.8||1.33||avg. Asian speed (quotes Ivan T Sanderson|
|Estes||1991||6 to 8||1.6-2.2||avg.|
|Le Rue III||1994||5||8||2.22||from walking along side|
|Alexander, R.McN.||1979||8.55||13.68||3.80||Muybridge photos|
|Paul||1998||12||20||5.56||max; claims 19 kph from Asian in filmed race|
|Baker||1890||15||24||6.67||200-300 yards; Asian and African|
|Gale||1974||15||24||6.67||Asian max (100 yards)|
|Spinage||1994||15||24||6.67||African shuffling walk|
|Alexander, S||2000||15||24||6.67||Asian max (quotes Sanderson)|
|Hutchinson et al.||2003, 2006||15||24||6.80||Asian and African near-maximal speed|
|Spinage||1994||20||32||8.89||max Asian (also says Asian is slower?!?)|
|1983||21.88||35||9.72||knockdown of Howell|
|Van Gelder||1969||25||40||11.11||120 yards|
|Borst||1969||25||40||11.11||avg. speed is 4 mph|
|Le Rue III||1994||25||40||11.11||charging (land rover)|
|Alexander, S||2000||30||48||13.33||African max|
|Howell||1944||see Muybridge and Andrews|
Thanks to all of the people and institutions that have helped with this long term project!
We especially want to thank:
- All of the elephants
- The wonderful elephant owners, mahouts, and other staff who assisted us in the field in Thailand
- Thai Elephant Conservation Center, National Elephant Institute, and Forest Industry Organization
- Stanford University: Department of Mechanical Engineering, Biomechanical Engineering Division
- University of California: Department of Integrative Biology
- University of Colorado: Department of Applied Physiology and Kinesiology
- J.R.H. appreciates funding from a Journal of Experimental Biology Travelling Fellowship (2000) and the National Science Foundation under a grant awarded in 2001.
- Heidi Riddle for having connected the US side with the Thai side
- Stephanie Sanchez and Betty Tzeng for crucial assistance during data collection
- Robert Full (University of California) for use of equipment and facilities
- Steve Gatesy (Brown Univ.) for a helpful review of a paper draft
Images and Movies From Our Research On Fast-Moving Elephants
Please ask for permission before using any of these media, and in all cases we ask that due credit be given as listed.
This 17 year old male “tusker” elephant was appropriately named, as he was among the largest elephants we worked with in Thailand, at about 2800 kg (6100 lbs.). He and Nong Pop (below) had a friendly competition going for two days during our experiments, repeatedly beating each others’ times. We had thought Pop had won, until recently we re-analyzed our video of one of what we thought was a slower trial for Big. It turned out to be the winner actually, at 6.8 m/s (15 mph; 24 kph), beating out Pop’s fastest time by a nose. (Credit for images: John R. Hutchinson
Big Fast 6.6m/s (15mph; 24kph):
2. Nong Pop
The female elephant Nong Pop, at Surin, Thailand, was at the time of our studies about 7 years old. The movie included here is one of our fastest recorded trials with any elephant. Nong Pop particularly seemed to enjoy the “races”; she had to be firmly held back or she would start moving before we were ready. She needed little if any encouragement and went straight for the 30 meter course when released; it was difficult to get her to agree to walk slowly! She broke away at least once, chasing behind our fastest elephant “Big” and getting lots of laughs, plus one of her fastest times! (Credit for images: John R. Hutchinson
Steady walk (MOV) movie; 3183 kb)
Fast: 6.6 m/s (MOV) (15 mph; 24 kph)
3. 3D animations
We’ve also had the pleasure of meeting a computer animator, Karen Johnson from the Savannah College of Art and Design, who has used our videos as inspiration to design amusing 3D movies of elephants moving at slow, fast, and ridiculously fast (airborne; fantasy only) speeds. You can visit her website at: www.paintedthorn.com.
Airborne elephant: (watch closely; it has 4 feet off the ground -- not from a real video, just imagined)
Photos of fast elephants from our research, courtesy of Richard Lair:
Miscellaneous other images, courtesy of John R. Hutchinson: