Browser upgrade warning
 
Business Gateway - Research Intelligence
Business Gateway home page Research Intelligence home page
 
 
 
Tell us what you think about our website
ISSUE 22 - Nov 2004  | Past Issues  |  Future Issues

KEYWORDS: WALKING, EVOLUTION OF BIPEDALISM, COMPUTER SIMULATION, EARLY HOMINIDS,
MODERN PRIMATES, METABOLIC COSTS

Six years ago a Liverpool University scientist caused a stir with his computer simulation of a 3.6M year old skeleton walking. Today, he can tell us a lot more about 'Lucy' and other hominids, thanks to a wealth of data on contemporary humans and apes, and sophisticated new computer models.

Walking tall after all

Walking on two legs is not uniquely human: apes, lizards and birds do it from time to time, and even cockroaches can run on two legs. What sets us apart from other species is our unique ability to walk or run upright for very long distances.

We don't really understand how or why this evolved. Did foraging for raw materials - for food, fire and tools - encourage bipedalism, or did bipedalism make it possible to forage further and further afield? We can't hope to answer such questions without first knowing how bipedalism evolved.

If we can solve this fundamental puzzle, we should learn a lot more about the evolution of humans and other primates - and this is what drives Professor Robin Crompton's pioneering research at the University of Liverpool.

Mixed messages

 
The most complete fossil skeleton of an early ancestor, Australopithecus afarensis. This particular skeleton, known as ?Lucy?, is about 3.16 million years old.
The most complete fossil skeleton of an early ancestor, Australopithecus afarensis. This particular skeleton, known as ‘Lucy’, is about 3.16 million years old.

There are clues in every hominid skeleton found, but we are still learning how to interpret them. 'Lucy' and her kind, who lived in Africa between 3.6M and 2.9M years ago, are a classic example of the difficulties entailed. "Lucy had a small brain, and a face and teeth which were more like a chimp's than a human's", Robin Crompton explains. "She also had a long trunk, long arms, short legs and grasping toes, whereas humans have short trunks and long legs … However, Lucy's knees closely resembled those of a modern humans and the arrangement of the ankle bone was very similar, too. So, how did she walk? Bent-kneed and bent-hipped like a chimp, or erect, like a human?"

 
These two graphs show the predicted power output from the ?Lucy? simulations. Left: erect human bipedal gait. Right: Bent hip, bent knee (BHBK) human bipedalism of the type proposed for ?Lucy? by some scientists. In BHBK walking energy is stored at the knee and ankle, predicting increased heat load. Lucy?s mechanical effectiveness in BHBK walking would have been very low, but she could have been an effective upright walker.
These two graphs show the predicted power output from the ‘Lucy’ simulations. Left: erect human bipedal gait. Right: Bent hip, bent knee (BHBK) human bipedalism of the type proposed for ‘Lucy’ by some scientists. In BHBK walking energy is stored at the knee and ankle, predicting increased heat load. Lucy’s mechanical effectiveness in BHBK walking would have been very low, but she could have been an effective upright walker.

In 1998 Robin Crompton's Primate Evolution & Morphology Research Group hit the world's headlines when it produced a computer simulation of Lucy walking upright. Russell Savage, Li Yu and Wang Weijie had used a mechanical engineering technique to predict how a model with Lucy's body proportions and build would respond to the different joint motions associated with walking like a human, like a chimp, and like a human mimicking a chimp. The simulations indicated that Lucy could not have walked bipedally the way a chimp does. She could have walked like a human simulating a chimp; however, the mechanical joint power required was nearly double that required for erect walking and the computer model predicted a risk of overheating. So Robin Crompton concluded that Lucy had almost certainly walked upright.

New approach

Some scientists were not convinced by computer simulations, so the research group decided to approach the problem from a different angle. Dr Tanya Carey began by collecting detailed data on the physiological costs of erect and bent-hip, bent-knee walking in contemporary humans. Her study confirmed the computer model's predictions: in bent-hip, bent-knee posture the energy costs of locomotion double; the core body temperature rises so much that it is necessary to rest one and a half times as long as the time spent walking. This would have limited the distances Lucy could safely cover, and that, in turn, would have restricted her ability to forage for food.

Following this study, Robin Crompton set up a state of the art laboratory equipped with sophisticated gait analysis equipment, scanning and computing facilities designed for comprehensive analyses of the physiology and mechanics of human gait.

His research group extended its analyses to other great apes, which were filmed in captivity and in the wild and given an opportunity to walk over force plates to provide contact data; in the process, the group discovered that orang-utans' style of bipedal walking is the closest to modern humans'. The group also analysed film of a chimp from Kenya who was discovered living in captivity in an upright cage so small that there was no room to walk on all fours. "As a result, his body has been 'super-trained' to do things it would not do naturally - rather like a gymnast", says Robin Crompton. "While still not as efficient as the orang, 'Poko' came a close second. This has given us valuable insights into constraints on the plasticity of a chimp's body."

Hopping, skipping, tip-toeing

Two years ago, Robin Crompton and Dr Bill Sellers started collaborating with scientists who had created the first computer model capable of predicting the metabolic costs of bipedal gait. Due to the more limited computing capabilities available at the time, they had been forced to utilise unrealistic foot and leg lengths. With financial support from Britain's Bioengineering & Biosciences Research Council (BBSRC), the collaborators set out to create new computer models, incorporating recent advances in computing to achieve much higher levels of biological realism.

 
A hybrid multi-rigid-body/Finite elements forwards dynamics model with a 4-component foot, driven by 12 lower limb muscle groups to predict metabolic costs. It calculates and applies under-foot sagittal stress to the ground to propel the whole body, as in real human walking.
A hybrid multi-rigid-body/Finite elements forwards dynamics model with a 4-component foot, driven by 12 lower limb muscle groups to predict metabolic costs. It calculates and applies under-foot sagittal stress to the ground to propel the whole body, as in real human walking.

One model is a redevelopment of Dr Alberto Minetti's original metabolic costs model, which now works with realistic foot and leg lengths, and 'evolves' muscle activation patterns best suited to its proportions. "Essentially, you tell the model which anatomical data to access - modern human, orang-utan, Australopithecus or whatever, and the initial conditions you want to set - for instance, which muscles should contract in the first half step", says Robin Crompton. "Then you hit the 'run' button and watch it simulate bipedal locomotion on-screen. It can generate any kind of bipedalism it sees as feasible. It worked out for itself that modern humans can progress by hopping, skipping, walking, running - and even going on tip-toe.

"The models can handle single queries or multiple queries in sequence or in parallel. You can ask them to simulate the style of bipedal locomotion which is fastest, or least likely to stress the bones in the short, medium or long term, or cheapest in terms of oxygen consumption, and so on. The oxygen consumption which the model predicts for different types of modern human locomotion is within 5-15% of the real values. This is much more accurate than any other model to date."

The other model works at a more detailed level, and concentrates on predicting the forces under each part of the foot which drive walking, and the joint loads which result from its action. "Until now the foot has usually been represented like an inflexible stick, but it's got at least twelve basic components and it's very flexible", Robin Crompton explains.

Lucy's options

The new models tell us a lot more about the way walking evolved. They confirm that bent-hip, bent-knee walking would have been very expensive for Lucy. "Her body form is designed to give her good static stability", says Robin Crompton. "However, she could also have walked upright, and the models tell us that this would have been fairly energy-efficient.

"Due to their long trunks and short legs, hominids like Lucy would not have been as capable as modern humans - or Homo ergaster, who lived about 1.8 million years ago - when it came to carrying loads on their shoulders or in their hands. Nonetheless, walking erect, Lucy could have hand-carried loads weighing up to 60% of her body weight for less than the cost she would have incurred walking bent-hip, bent-knee but unloaded.

"On balance, I think it's most likely that Lucy walked erect, was able to travel some distance to secure food - and may therefore have enjoyed a fairly varied diet. But Lucy and her kind hadn't achieved the full potential of bipedal gait for inexpensive long-distance travel and load-carrying."

View publications relating to this story.

FURTHER INFO:Visit www.liverpool.ac.uk/premog/premog-research.htm or contact Robin Crompton at rhcromp@liverpool.ac.uk.

Download PDF of this page Download PDF of this issue Download Adobe Reader 6.0

Back to top