by Meave Leakey
All primates have a manipulative hand—but ours is by far the most mobile. We owe this dexterity in large part to the shared evolution of the primate order. For example, the mobility in our shoulders and elbows originally evolved for swinging through the trees. And our five-digit configuration of strong, flexible fingers evolved for grasping branches. But because these traits were also very useful for manipulative tasks, we improved upon them. To these shared primate advantages in our upper body configuration, we added another, by freeing the hand from its locomotive constraints.
Such a high degree of dexterity boils down to three principal features: the proportions of our fingers, an opposable thumb, and the mobility afforded by the morphology of the joints at the wrist, elbow, and shoulder. The African apes tend to have long curved fingers and reduced thumbs because this configuration gives the best grip on an assortment of different-sized branches. A colobine, the most skilled suspensory arboreal monkey, barely has a thumb at all. Conversely, monkeys that spend most of their time on the ground tend to have reduced fingers like baboons and patas monkeys do. The relative proportion of the thumb to the index finger is part of what allows for a precision grip between the tip of the index finger and the tip of the thumb. Because an ape’s thumb is short relative to its other digits, the thumb only reaches part of the way up the index finger whereas ground-dwelling monkeys have digit proportions better suited to achieving some sort of modified precision grip. But the full precision grip also requires flexibility in the joints to allow the rotation of the thumb—something that we excel at above all other primates.
Manual dexterity vastly improved the efficiency with which an individual could gather food and also allowed for further developments such as the manufacture of effective stone-tool kits and the ability to carry and store foraged food (not drawn to scale).
The ability to rotate the thumb and some of the other fingers is the secret to our vastly superior array of grips. We can perform, among others, the scissor grip (which many humans use to smoke a cigarette with), the precision grip (where we use the thumb against one or several other fingers for such tasks as threading a needle, holding a key to unlock a door, picking up a pencil), and power grips (which we use to grasp a ball, open a jar, and wield a hammer). The only grips available to apes such as gorillas involve folding the fingers over an object—a hook grip to brandish a branch (or, in our case, carry a basket) and a less accurate version of the precision grip between the thumb pad and the side of the index finger to grasp small food items. Another key difference is that our grips are much stronger.
What would have driven these changes in our hand morphology? I believe that here again the clue is food; like bipedality, the evolution of manual dexterity was driven by selective forces related to feeding. With a dexterous, flexible hand, it is possible for an animal to effectively collect small objects such as grass seeds, small fruits, and tiny insects. If collected efficiently, these can provide an important component of the diet, and in times of extreme food shortages and climate stress, minute adaptations that permit more efficient food acquisition can literally make all the difference between life and death, as we saw with the Galápagos finches.
There is a fascinating analogue among modern monkeys, many of which are able to collect small seeds and fruits with their manipulative, dexterous fingers. Few can do this as fast and efficiently as modern gelada baboons. These baboons live only in the Ethiopia Highlands and feed exclusively on grass. After the rains, they eat the new green shoots; later, they eat the grass flowers; and after the grasses have flowered, they eat the seeds. When the grass becomes dry and withered, they dig up the grass tubers and feed on these. So they have a continuous source of food regardless of seasonal changes. Among modern primates, the geladas have hand proportions closest to our own and can almost achieve a precision grip. Ancestors of the modern geladas were common in the Turkana Basin and elsewhere in the Late Pliocene and Pleistocene. Their dental morphology and analyses of the stable carbon isotopes in their fossilised tooth enamel show that these ancestral geladas, like their modern counterparts, were committed grazers. They had already evolved hand proportions that are similar to those of gelada baboons today, and manual dexterity would have been an essential component in their successful feeding.
The flexibility that provides the ability to rotate the fingers depends on the shape of the collection of bones that make up the wrist joint, so it is these bones, together with the finger bones, that we are most interested in finding to learn about the evolution of manual dexterity in hominins. Over time, we would expect fingers that originally resembled an ape’s long curved phalanges to become shorter and straighter, which would reduce the difference in proportions between index-finger and thumb length. And the tightly interlocking wristbones that characterize an ape should give way to wristbones that allow some rotation and flexibility. The problem is that hand bones are often destroyed before they are fossilised, and their pebblelike shapes are notoriously hard to discern, making manual dexterity an extremely elusive trait to chart through time.
Incredibly, almost all the hand bones of A. afarensis are known from Hadar. These bones show that the proportions of the finger bones (phalanges) were quite close to those of our own hands but that the fingers, although much shorter, were strong and quite curved like a chimpanzee’s. Australopithecus afarensis had strong flexible fingers with which to firmly grasp branches. On the other hand, the bones of the wrist show that some movement between the joints was possible, although it is less than for humans, so the power grip would have been only weakly developed in afarensis. In other words, the fingers became shorter before major changes in the morphology of the wrist.
Although the hand of A. afarensis was considerably less dexterous than that of later hominins, its shorter fingers and partial wrist rotation revealed that the changes in hominin hands leading to greater dexterity had already begun a million years before the earliest stone tools known at the time, 2.6 million years ago. I was deeply curious to know whether the hand of A. anamensis was as manipulative as that of A. afarensis or if it would be more primitive. If we could possibly find a capitate (the most diagnostic of the wristbones), what would it look like? Against all odds, we had found one of these very bones at South Turkwell in my first year in charge of the fieldwork. It was Wambua who spotted a number of hand bones that included a beautifully preserved capitate. This looked exactly like the A. afarensis capitates, and at 3.5 million years, it is the same age as A. afarensis at Laetoli.
I had a second unanswered question regarding a tooth that could lay to rest any lingering doubts I might have about whether the Kanapoi fossils justifiably deserved their own separate species. When Tim White initially found Ar. ramidus, one particular specimen had convinced him that Ar. ramidus was much more primitive than A. afarensis. This was the small and rather insignificant-looking fragment of a baby’s jaw that contained a deciduous (milk) lower first molar and caught my attention on my visit to Addis for its remarkable similarity to a chimpanzee’s. The Ethiopian collections also included deciduous lower first molars belonging to A. afarensis, and these looked quite distinct. I was curious to know whether the lower first milk molar of A. anamensis would be more like that of A. afarensis or Ar. ramidus.
To answer both these questions—what the anamensis deciduous first molar would look like and how dexterous its hands would be—I needed fossils that are both rare and difficult to find because they are very small and fragile. I thought it most unlikely that I would be this lucky. The chances of coming across one are remote, and our best hopes lay in discovering one (as Tim had) in the small and delicate jaw of a baby. Hand bones are also small, and the even smaller wristbones are very rarely recovered as they are hard to recognise. A single phalanx that we found in 1995 showed us that the anamensis fingers were similar to those of afarensis. This phalanx was slightly curved, not very long, and had well-developed ridges for the attachment of strong flexing muscles; anamensis, like afarensis, had a hand that was s
till capable of firmly grasping branches when climbing. But we still had no idea of the morphology of the wrist, and finding any of the hand bones, let alone the exact bone we wanted, was like looking for a needle in a very large haystack.
When we returned to Kanapoi for the 1996 field season, it was these two elusive bones that we sought above all others. Taking a leaf out of Tim White’s book, we decided to try his “hill crawl” method on a particularly prolific area where we had found so many isolated hominin teeth that we dubbed it the “graveyard.” This area was too large to excavate and sieve given the time and funds we had available. Clutching numbered collecting bags, we spread out in a line on our hands and knees on the stony ground about half a metre apart. Then we inched forward through the morning, eyes glued to the ground as we scoured every centimetre and collected every single bone fragment we could find. Back at camp, I sorted through the fossils. A hill crawl enabled us to localise our sieves and excavations around points where we found hominins, which saved us from having to sieve the whole area. One patch on the top of a small hill looked particularly promising. Several unworn and apparently unerupted anamensis teeth were scattered on the slope of the hill, seemingly coming from a horizon not far from the top. We hoped that if we made an excavation here, we might be able to determine whether the bones were coming from a discrete soil layer. If this were so, we could concentrate our searches on a much smaller area. As we dug deeper into the hill, the compacted soil became extremely hard, and we doused it in precious water in the evening to soften it for removal the following day. It turned out that the fossils were not so localised, but to our amazement and disbelief, our excavation was rewarded with the discovery of one of the very bones we were looking for.
After hours of tedious digging through very hard rock, we began to uncover more isolated adult teeth as well as the milk teeth of a young juvenile. As more of these tiny, delicate, beautifully preserved teeth were recovered, I hoped against hope for the elusive lower milk molar. We found the tiny, erupted, and worn deciduous teeth including incisors, canines, and molars from the upper and lower jaws as well as unworn, exquisitely preserved adult teeth that had not yet erupted. We couldn’t believe our luck when we also found a lower deciduous first molar!
Even more astonishing and improbable was Wambua’s finding the jaw of another baby at a different site some distance away—and the jaw had a deciduous first molar firmly rooted in it! We could now compare our baby first molars with Tim’s Ar. ramidus specimen. As we peered through the microscope back in Nairobi, we noticed that ours were intermediate in shape between Ardipithecus ramidus and Australopithecus afarensis, although they resembled afarensis more than they did the more primitive ramidus. This was reassuring, because we had postulated that anamensis belonged to the australopithecine genus, not Ardipithecus.
When we subsequently and unexpectedly found a capitate while sieving an area close to the main graveyard site, I was ecstatic. The capitate also confirmed what our initial analysis had told us. While the capitate of afarensis was in many ways more similar to that of modern humans than that of chimpanzees, the new anamensis capitate from Kanapoi still retained some chimpanzee features not observed in the younger afarensis. In particular, its morphology suggested that very little movement would have been possible between the wristbones—and much less movement than afarensis had. In many ways, the capitate had a morphology that was halfway between a chimpanzee’s and that of afarensis. Now we could say with some certainty that the changes in hand morphology that ultimately led to a uniquely dexterous hand came gradually after bipedality was firmly established.
Our last field season at Kanapoi concluded in 1997 and coincided with the news that NASA had successfully landed on Mars, and once again, Nzube was transfixed and disbelieving. The Pathfinder Mission was even more incredible to him than Armstrong’s moon landing three decades earlier. Parachutes to slow the descent as the projectile hurtled into the Martian atmosphere? Airbags to help Sojourner bounce to a safe landing? He merely chuckled at our descriptions. And then images were transmitted directly into our primitive camp on a laptop screen. Such technology seemed all the more impossible! It was an endlessly fascinating subject, and we soon found that we were all caught up in following Sojourner’s progress through the field season while we gazed up at the stars above our camp. Everyone was enthralled by the ancient Martian floodplain, Ares Vallis, where the rover landed, which was not too different from our own remote surroundings where we were engaged in a similar task of imagining a past landscape by deciphering the fingerprints of rocks. We greedily waited for more images to be downloaded in Nairobi and brought to camp by a visiting scientist. Inspired by the spirit of exploration, we bent more determinedly to our own mission of discovery of the beginning of manual dexterity—the very adaptation that enabled modern humankind to develop the technical capability for space exploration.
10
Open-Country Survivors
After the wonderful discoveries of Australopithecus anamensis at Kanapoi and Allia Bay, I was faced with a decision—should I look for the very rare Late Miocene sites of Turkana that preserve sediments older than five million years, such as Lothagam, to try to find even earlier hominins from whom A. anamensis may have evolved? Or should I look for new sites between four and three million years and resolve the question of what A. anamensis evolved into? Either plan would bring me closer to the issue of whether or not there was as much early hominin diversity as I believed there must have been. But this was not an easy decision, not least because a lot was being discovered elsewhere in the meantime. Fascinating new fossils were emerging from different parts of Africa that were shaking all our views.
At the heart of most of the ensuing controversies lies the question of early hominin diversity, made all the more intractable by the rarity of any hominin fossils older than four million years and the even more striking rarity of any ape fossils for the last six million years. There is far more known about hominin ancestry than about the equally long parallel lineage that gave rise to the chimpanzee—we only have one such example of fossil ape from the whole of Africa. This singular specimen is of a chimpanzee that lived somewhere between 545,000 and 284,000 years ago. I would guess that the reason for this incredible paucity is because of the different habitats of the two groups: the fossil sites we sample in the East African Rift are either riverine forest or more open woodland and savannah. None of them come close to the thickly forested habitats preferred by the modern apes of Central Africa. Nevertheless, I would dearly love to find an ape of equivalent age to any of the early hominins. But even more, I would love to find the ancestor of A. anamensis.
While we have good reason to think that A. afarensis probably evolved from A. anamensis, it is far less clear where A. anamensis came from. Tim White’s Ardipithecus ramidus, the surprising 4.4-million-year-old fossils he showed me on my visit to Addis Ababa with my new Kanapoi discoveries, is one of the currently known possible contenders. Ar. ramidus is, as we would expect, more apelike and primitive than A. anamensis. It has smaller molars with much thinner enamel and longer canines than A. anamensis has.
Since his early work with Don Johanson that joined the Laetoli and Hadar fossils into the single species Australopithecus afarensis, Tim has strongly argued that there was one single main lineage of early hominins. In 2006, Tim and his colleagues described new fossils of A. anamensis from two sites in Ethiopia, Assa Issie and Aramis (the same site as Ar. ramidus but in younger sediments). At a similar age to Kanapoi, these fossils added nine new individuals of this species and extended its range by one thousand kilometres. Because they were found at Aramis in the same area as Ar. ramidus and A. afarensis, these three species represent a time-successive series from one place. This poses the question of whether this is one evolutionary lineage—or whether Ar. ramidus represents the diversity I keep expecting to see at this early age. It is certainly Tim’s view that Ar. ramidus is the direct ancestor of A. anamensis. However, for this to be true, a lot of
evolutionary change would have had to happen in a very compressed amount of time—only 200,000 years. Ar. ramidus is found at 4.4 million years, and the oldest specimens of A. anamensis lived at approximately 4.2 million. Yet the changes in the canine-premolar complex between the two are quite dramatic, and I remain unconvinced about their phylogenetic relationship.
But what about fossils much older than Ar. ramidus? Since the late 1990s, three new candidates for an earlier potential australopithecine ancestor have been discovered. These show a hodgepodge of ape and human traits. At this early age, so close to the split from our common ancestor with chimpanzees, it is incredibly difficult to distinguish from the fragmentary bones available whether they represent ape or hominin. And in any case, what are the defining characters that set hominin apart from ape? Bipedalism, an obvious divider, would not have arrived in an instant. Similarly, the canine would have reduced in size only gradually.
In fact, we do see this gradual process among the early australopithecines. The male canine has already evolved to the point that it occludes (bites down) directly on the lower premolar rather than overlapping outside this lower tooth (honing) as it does in apes and monkeys. Nevertheless, reflecting its ancestral condition in apes, an early male australopithecine like A. anamensis still has a rather large set of canines. Moreover, although the canine is no longer honing, the lower premolar still has some enamel extending down the outer surface of the root. When the canines hone, as they do in apes, they are continually sharpened as they wear against the first lower premolar, and in order to preserve this tooth as a sharpening device, the enamel surface extends partway down the outside of the root. In the more recent australopithecines, the extension of the enamel on the lower premolar recedes as the canine decreases in size, and its crown gets bigger as it increasingly takes on a chewing function.