The australopithecines were several species of human relatives that lived in Africa between 5 million and 1.5 million years ago. One of the best-represented species of australopithecines is Australopithecus afarensis, which is known from Ethiopia and Kenya between 3.9 million and 3 million years ago. Although it is not the earliest hominin species, A. afarensis provides some of the earliest evidence about the evolution of bipedality in our lineage.
Human feet are very different from ape feet. The differences are in the toes, the "arches" of the midfoot and the ankle.
1. An ape foot bears an opposable big toe and they can grasp objects or branches with their feet. Humans have a stout big toe in line with the other toes, and our feet function as a stiff lever that lets us "toe off" powerfully into our next step. The rest of the toes are long and curving in apes, relatively short in humans.
2. Ape feet are relatively flat and flexible in the middle section of the foot, letting them conform (and sometimes grip around) the surface they are on. Human feet have strong ligaments and bones packed into arch structures, both along the length of the foot (the longitudinal arch) and across the foot from side-to-side (the transverse arch).
3. Apes have ankle joints that can flex upward (dorsiflex) more than human ankles, allowing them to climb more effectively on vertical tree trunks. Humans are more limited in this movement and have ankles that are directed more toward front-to-back movement with less side-to-side mobility.
Paleoanthropologists have found evidence about the feet of A. afarensis from fossil footprint tracks, at a place called Laetoli, Tanzania. These footprints are nearly 3.7 million years old, and were uncovered by Mary Leakey during the late 1970's.
View Larger Map
The marker shows the location of the Laetoli footprint trail, in Tanzania. Laetoli is just south of Olduvai Gorge and west of the Ngorongoro crater.
The footprint trackways are preserved in a series of overlapping surfaces that are the result of falls of volcanic ash, very quickly cemented by rainfall. These layers preserve the footprints of many kinds of animals, including ancient horses, birds and hominids. The best preserved hominin footprints are in trail G, representing two hominids walking in a straight line for a distance of about five meters.
The Laetoli footprints are strikingly human-like, suggesting that these early hominins had feet that were functionally equivalent to ours. The imprints of the toes show that the big toe of A. afarensis was in line with the others. The big toe did not jut out from the foot as these creatures walked, it would have helped "toe off" each successive step. Many of the prints show that the feet were arched, which suggests an anatomy similar to the human anatomy in which the midfoot functions as a lever (White 1980; White and Suwa 1987).
The paths have short spaces between successive prints, indicating a slow walking speed. One of the tracks is substantially smaller than the other, and since the two individuals matched each other stride-for-stride, it is likely that the smaller individual set the pace, with the larger one following (Leakey and Hay 1979). The larger track appears to preserve the footprints of at least two individuals, one literally following in the footsteps of the other.
Fossil foot bones from A. afarensis add detail to what we've learned from the Laetoli footprint trails. For example, Carol Ward, Bill Kimbel and Don Johanson  described one of the bones of the midfoot from Hadar, Ethiopia, that represents A. afarensis. This bone, the fourth metatarsal, is the one that connects the fourth toe to the bones of the ankle. In humans, these ankle bones are higher, and transfer the body's weight downward into the arching midfoot. So the fourth metatarsal has to be slightly twisted as it arches down toward the lateral (outside) side of the foot. A chimpanzee's foot is much flatter, so the bone doesn't twist. Ward and colleagues found that the A. afarensis bone was twisted in a humanlike way.
Hadar is the place where one of the most famous skeletons in the world was found: the "Lucy" skeleton, discovered by Don Johanson in 1974. Lucy presents evidence across much of her skeleton for a humanlike manner of walking. Jeremy DeSilva and Zach Throckmorton  looked at the tibia (shin bone) of this skeleton, along with other fossil tibiae of A. afarensis, to try to determine whether their ankle bones were structured to create a rearfoot arch like humans. What they found was interesting: Most A. afarensis tibiae met the ankle in a humanlike orientation, but they varied. Lucy's ankle in particular looked like her feet were relatively flatter. Just as human feet vary in their shape, this early species of hominins varied as well.
Ape feet are made for climbing. Their flexibility allows the sole of the foot to conform to a tree trunk or branch, giving them more friction and a stronger grip. Ape ankle joints allow them to walk up a trunk while holding it with their arms, which means their feet need to carry weight while they are flexed upward and tilted to the side. Humans who are good climbers tend to shimmy up a tree by gripping the trunk tightly between their legs. Our feet just don't work very well when climbing a truly vertical surface. Rock climbers look for footholds where they can place their feet; loggers can use metal spikes clamped to their boots.
What about A. afarensis? Jeremy DeSilva  examined the ankle joints preserved for this species and found that they did not have the orientation that chimpanzee ankles do. Where chimpanzees can bear weight effectively with their feet tilted to the side, for A. afarensis this orientation just wouldn't have worked. If these early hominins needed to climb trees — which might explain the powerful arm bones of some individuals — they must have done it in a different way than chimpanzees and other apes.
Yet, the feet of A. afarensis were not entirely like ours. One major difference is that their toe bones were curved more than ours (Stern and Susman 1983). This curvature is less than the great curvature of the foot bones in chimpanzees, and may be correlated to the relatively long length of australopithecine toes compared to body size. Or, curved bones may just be a retention from their ape ancestry that hadn't been eliminated because it didn't impede walking. The curvature of the bones and a few other apelike features of the A. afarensis foot show that human feet did not instantly evolve in one moment of our evolution. Instead, human feet represent a long evolutionary history with different steps at different times.