Unlike humans today, the earliest human ancestors were lanky-armed and short-legged. However, around two million years ago, our ancestors' legs grew longer. Why this happened is a mystery that intrigues Herman Pontzer from Washington University, St Louis. He wondered if leg length might affect the energetic costs of locomotion; that is, the amount of energy animals use to walk and run. To investigate, Pontzer built a mathematical walking model from scratch, called LiMb, to try and find out what affects the energetic costs of locomotion. Having already tested his model in humans, he wanted to see if it could predict the costs of moving around in four-legged animals too(p. 484). Not only could a model like LiMb tell scientists about what influences how much energy living animals use to get around, but it might also give them some clues as to why ancient humans developed longer legs.

The LiMb model has two steps: first the model estimates limb muscle forces,which accelerate the body, by taking into account a leg's length, its movements during walking, and stride frequency. The second step uses this information to predict how much energy is used. To measure leg movements during walking and running, Pontzer put his human volunteers, and obliging dogs and goats, to work on the treadmill. Using reflective markers attached to the limbs, he could track how each limb moved. He put this information into the model to estimate the muscle forces, and then the energy cost.

However, Pontzer needed to see if his model was hitting the mark, so he compared his model's predictions with measurements he made on his volunteers. First he measured forces produced by the legs during running using a force plate incorporated into the treadmill and compared these measurements with the model's predictions, finding that his measurements were a close match.

Then, to find out if the LiMb model was also accurately predicting the cost of locomotion in two- and four-legged creatures, he measured how much energy the humans, dogs and goats were using as they ambled along. Each individual wore a mask which collected the air they breathed out: because oxygen consumption is directly related to energy use, the amount of oxygen left in the exhaled air told Pontzer how much energy each animal had used. He found that the model did a very good job of predicting the cost of walking and running in humans, dogs and goats, showing that it works equally well for animals with four legs as well as two. However, the predictions weren't an exact match. This is probably because the model doesn't take into account everything that can affect locomotion, such as subtle differences in walking styles or variation in muscle fibre length.

Finally, because the model uses leg length to calculate cost, it probably has an important effect on the energy used to walk and run. `All things being equal, leg length is one of the major determinants of cost', says Pontzer,adding that if two animals are identical except for leg length, longer legs are more efficient. As for our ancestors, it's possible that longer legs helped them walk more efficiently.

Pontzer, H. (
). Predicting the energy cost of terrestrial locomotion: a test of the LiMb model in humans and quadrupeds.
J. Exp. Biol.