by Meave Leakey
On the other hand, the Kanapoi material was more similar in other characters to afarensis than to modern chimpanzees and gorillas. The similarities in the two collections included thicker tooth enamel than is typical of modern African apes. This was not unexpected, since it most likely reflected dietary trends associated with the long-term changes from closed to more open habitats that we had revealed at Lothagam.
Like A. afarensis (and modern African apes), the Kanapoi hominins show a large difference in size between males and females. It was the marked size difference that had led Johanson to initially think that he had more than one species at Hadar. But when he looked at the size difference between male and female modern gorillas (the most sexually dimorphic modern African primates), the range of variation in the Hadar collections was not exceptional. When we compared the enormous sockets for the canines in the lower jaw of Ngui’s mandible with the small canines of Nzube’s beautiful mandible, we also initially wondered if these specimens could be from more than one species. The same comparison with modern gorillas, however, convinced us that we had only one highly sexually dimorphic species.
A trend of evolving characters can be traced from the Kanapoi material to Lucy’s kin. In afarensis, the canine sexual dimorphism is reduced, but the high body-size dimorphism is retained. This suggests a social system characterized by high levels of male-male competition that did not depend upon the use of fearsome-looking canine teeth. This could be explained by the move to bipedality, which allowed the development of alternative means of display, such as chest beating or branch flourishing. But it made our fossils that still showed a high degree of canine dimorphism all the more interesting.
Although there were many characters that clearly linked them, the two collections were readily distinguished due to the more primitive nature of the Kanapoi fossils. Our next step was to check these features on both the specimens from Laetoli, which at 3.6 million years are slightly older than those from Hadar, and the new Allia Bay fossils, which at 3.9 million years are slightly younger than those from Kanapoi. Grouped this way, the evolutionary changes sprang even more clearly into focus. We were looking at fossils from four populations of decreasing geological age that displayed a distinct trend towards less apelike and more hominin-like characters. Should all these specimens be included in one species or should we name a new species for the earlier specimens since these were so clearly distinct from A. afarensis?
Our findings confronted us with a pressing dilemma that is a recurring source of much of the controversy in palaeontology. When do you decide to name a new species, and when do you lump things together? In reality, there is no clear line drawn in the sand to divide one group from another. Evolution works more gradually than that. Modern gulls in Northern Europe give a particularly elegant example of this. Two distinct species—the herring gull and lesser black-backed gull—both trace their ancestry to a Siberian species. This original bird spread slowly to the east and the west, all the while diverging into a chain of subspecies. Each one of these subspecies can breed with those adjacent to them. The two extremes of the populations, the herring and lesser black-backed gulls, overlap in Northern Europe but they cannot interbreed. These birds form what evolutionists call a classic “ring species,” for the birds from all the subspecies represent a continuum of traits. They demonstrate just how arbitrary a dividing line between species can be. We faced a different sort of arbitrary division with our hominin fossils, a continuum of traits across time rather than across a geographical space.
We agonised over how we should interpret these fossils. Lucy and the other younger fossils from Hadar clearly bore the most differences to our fossils at the older sites of Kanapoi. These differences were testimony to the million years that separated the majority of these specimens. The Allia Bay and Laetoli fossils, intermediate in age, showed intermediate forms. If our fossils had been around when Johanson and White did their analysis, their decision to pick an intermediate “average individual” for their holotype for A. afarensis might not have made much sense to them. But significant finds are often made years or even decades apart, and previous names hold until they don’t make sense anymore. We would have to build on what had gone on before and decide how the Kanapoi hominins differed from the Laetoli holotype of Australopithecus afarensis, LH-4. After much study and careful thought, we took the plunge and decided that all our fossils, including those from Allia Bay, were sufficiently different from all the A. afarensis material to deserve being called a new species that was ancestral to afarensis in the genus Australopithecus. We published a preliminary paper in Nature in 1996 that described and named the fossils. We picked the name Australopithecus anamensis (anam means “lake” in the Turkana language). At that time, anamensis represented the oldest known biped, and we received much publicity and acclaim around the world.
* * *
IN 2019, the publication of a new and fascinating discovery from the Godaya Valley in Ethiopia challenged our interpretation that anamensis-afarensis is indeed a single evolving lineage. The largely complete 3.8-million-year-old cranium was found by Yohannes Haile-Selassie in 2016 and is informally dubbed MRD as a shortened version of its full accession number in the National Museum of Ethiopia. It shows for the first time what an adult male A. anamensis could have looked like. Haile-Selassie and his colleagues suggest that, despite being younger than the material from Kanapoi and Allia Bay, some of MRD’s features are more primitive than those in these earlier A. anamensis fossils, which would be inconsistent with their belonging to a single evolving lineage that would become afarensis. Most striking is the forward position of the cheekbones. As further evidence to debunk the single-lineage hypothesis, the authors also present a second 3.9-million-year-old frontal bone (part of the forehead) that they assign to A. afarensis rather than A. anamensis. If they are correct, there was a period of more than 100,000 years when the two species coexisted, which would render it impossible for them to be a single lineage. However, several aspects of these finds still need to be clarified. MRD’s cheekbones are very distinctive, but the authors used a high degree of digital reconstruction to correct distortions and estimate missing parts, so the forward-projecting cheekbones are notable for how smoothed they are, with little sign of the original bone surface. While such digital reconstructions are excellent ways of visualizing what is missing or distorted in a fossil, interpretations based on them always deserve stronger scrutiny. Moreover, these findings are reliant on a single specimen so any variation within the species cannot be assessed, and as we have seen in the Kanapoi, Allia Bay, and Hadar fossil collections, not all individuals from these ancient species looked alike. Lastly, the frontal bone is not a good diagnostic bone, with one looking much like another, so establishing the contemporaneity of anamensis and afarensis on the basis of this fossil fragment will need corroboration. To my mind, this discovery is enormously significant and of huge interest in and of its own. It is consistent with the theory I have long held that there is likely to be more rather than less diversity in a radiation of early bipeds. Yet for the time being, I remain open-minded as to whether the weight of evidence that they present is sufficient to definitively debunk the theory that afarensis evolved from anamensis.
* * *
MEANWHILE, BACK IN 1995, I was extremely curious about the environment our new anamensis lived in. In great contrast to Lothagam, the sediments we were working in at Kanapoi covered a mere snippet of time in geological terms—just 100,000 years. Craig Feibel, who had made that flying visit during our first field season, joined us to study the geology, and he named these sediments the Kanapoi Formation.
A river system believed to be ancestral to the modern Kerio River snaked its way northwards through the ancient volcanic landscape for fifty thousand years some 4.1 million years ago. Conveniently, two volcanic eruptions sandwich this time interval and allow us to have remarkably accurate dates for it. The enormous Lonyumun Lake then flooded the area as it expanded southward, and we can see this in the
claystones and siltstones that make up the middle of the sequence. These oldest riverine layers correlate with the uppermost sediments in the Apak Member at Lothagam—remember where the squashed snails at the Lonyumun lakeshore were sizzled and distorted into a permanent record by the erupting Lothagam basalt?
The top third of the Kanapoi sequence records the establishment of a second river system as Lonyumun Lake receded, and the Kerio River delta advanced northward apace with the retreating lakeshore. A third ashfall, the Kanapoi Tuff, dated at 4.07 million years ago, was deposited in this deltaic complex and filled the bays and flood basins between the many distributary channels of the northward-advancing delta. This ash layer is clearly visible today where it blankets a large area of deposits to the north and east of our campsite.
Not far from Kanapoi, the modern Kerio River is a verdant ribbon slicing through the barren desert plains. While working at Kanapoi, we occasionally visited the Kerio on Sundays. Along with plenty of birdlife and smaller mammals, we always found prodigious quantities of fresh elephant dung underfoot even though the pachyderms have not been seen outside the green riverine belt in years and the river is strongly seasonal. The extreme contrast between the dry barren landscape at Kanapoi and the lush vegetation along the riverbanks vividly demonstrated to us just how localised the conditions we reveal at particular fossil sites could be—only a few miles away, conditions support a completely different biota. This could have been equally true in ancient times, for the fossils that we discover are always tied to a water system that is responsible for the sedimentation process preserving them.
Fossil soils, or palaeosols, add another dimension to our understanding about the past climate and habitat. In the aftermath of a storm, angry rivers, red with silt and tossing about logs and other vegetable detritus, are a force to watch. When the rivers swell beyond the narrow confines of their usual course, they burst their banks, and the furiously flowing water slows as it spreads out over the floodplain. Their burden of sand and silt dumps out of the water and forms a new layer of soil. This is quickly colonised by a new season of grasses, shrubs, and trees, and when these die and decompose, rich organic matter is added to the topsoil before a new rainy season begins the cycle all over again. In this way, soils develop, layer upon layer, at an incredibly slow pace, and the creation of a well-formed palaeosol takes hundreds of thousands of years. This means that the soils serve as an excellent proxy through time of the conditions under which they were created, although, as with all other geological records on land, there are large gaps from all those periods when the sediments were blown away by winds or carried away by water before hardening or being capped by rocks and sandstones.
Jonathan Wynn, whom I never once saw in the field without his trademark red bandana bound about his long and shaggy brown hair, is passionate about palaeosols. Jonathan first joined us in 1995 to study the palaeosols at Kanapoi when he was still a graduate student at the University of Oregon. Jonathan found that in general the Kanapoi palaeosols ranged from poorly drained vertisols (clay soils that crack open into a patchwork surface of yawning gaps in the dry season) that supported grasslands to well-drained alluvial soils that supported gallery woodland along the courses of open streams. In other words, the habitat was a mosaic setting of open and closed habitats, not just a dry savanna plain. Kanapoi at that time probably had a semiarid climate that averaged somewhere between 350 to 600 millimetres of annual rainfall, and the landscape was probably similar to that found today in the vicinity of the Omo Delta at the north end of Lake Turkana. The palaeoenvironmental information from Kanapoi provided the earliest evidence of hominins venturing into relatively open habitats.
The climate was also not unlike that at Kanapoi today. We know this from a comparison of modern and fossil rodents. One of our best hominin specimens was recovered in a horizon that had numerous fragments of the tiny skulls, teeth, and skeletal elements of bats and mice. This was Nzube’s mandible that Richard helped excavate and what we eventually chose as the type specimen for Australopithecus anamensis. These small mammals are particularly good palaeoenvironmental indicators, so to recover a sample in the same horizon as a hominin specimen is exceptionally fortuitous. Kiptalam Cheboi was a crew member famous for his skill at finding the smallest of specimens, and he devoted countless hours to sieving for these beautiful minute bones. Many years later, our colleague Fredrick Kyalo Manthi sampled the modern rodents and shrews from Kanapoi for his PhD dissertation in 2006. When Kyalo (pronounced Cha-lo) compared his modern collection with the fossil sample, he found many resemblances that suggested they lived in similar habitats.
But before then, John Harris once again joined me in the study of the Kanapoi fauna. He shared my burning curiosity to know how the Kanapoi faunal assemblage would differ from Lothagam’s. We found the Kanapoi mammals to be remarkably diverse—during four field seasons, we collected more than fifty species. Although few of these were new to science, the assemblage was instructive. While Lothagam witnessed the demise of many Miocene genera and the first appearance of several new creatures, a suite of these more modern species carried the day at Kanapoi. Indeed, none of the primitive large mammalian genera that characterize the Late Miocene at Lothagam were recorded at Kanapoi. Thus Kanapoi provides a link between the first appearances of more modern species in the upper levels at Lothagam and the more progressive species found in later sediments elsewhere in the Turkana Basin.
All told, the picture we built at Kanapoi revealed a mosaic habitat of woodland flanking the river and in the fan of the fertile river delta that gave way to semiarid grassland. This mosaic setting was frequented by species that are precursors of many of the African game animals today. The herbivores were much as we would expect following the huge shift in climate witnessed at Lothagam: when tallied up, the number of grazing species outnumbered browsing species by a ratio of two to one. The grazers were also more common: we found more specimens of grazing species in a ratio of three grazers to each browser, which suggests a far greater abundance of the grass-eating mammals. Assuming that the dietary adaptations of these herbivores are similar to those of their modern counterparts, their fossils indicate a relatively dry climate and a mixture of woodland and open grassland, again confirming the general trends of opening and drying environments observed at Lothagam.
The Kanapoi pigs, like those in the upper horizons at Lothagam, were grazers too. And our discoveries upturned conventional wisdom on the classification of one of these pigs, Notochoerus jaegeri, which had hereto been thought a completely different genus, the earlier Nyanzachoerus common at Lothagam. Pigs are excellent markers of evolutionary changes through time, and having accurate interpretations of their diversity backed up with good diagnostic fossils is extremely important.
The lower jaws of Notochoerus resemble the shovel of a mechanical digger. They have small widely separated peglike incisors protruding from the front edge of their rather broad and flat mandibles. In contrast, the lower jaw of Nyanzachoerus is more like a garden trowel: large closely aligned incisors form a sharp scooped edge ideal for rootling. A series of well-preserved mandibles and a single skull from Kanapoi clearly demonstrated that the previous generic affiliation was incorrect and that Nyanzachoerus jaegeri should be referred to the genus Notochoerus.
This skull very nearly never saw the light of day. It was already an incredibly fragile specimen when Nzube found it close to the hominin tibia during sieving for the missing piece of leg shaft. His discovery coincided with a visit from Richard, who was very frank about the extremely poor shape it was in. “You are wasting your time!” was his blunt assessment, and while he gave me some helpful advice on how to proceed, he wanted no part of the excavation. Undeterred, I devoted several days of meticulous excavation to the skull. Then I gave it a liberal application of Bedacryl to ready it for plastering and carefully covered it with rocks and branches from wait-a-bit thornbushes to protect it from the marauding efforts of hungry hyaenas as it dried overnight. The next morning, I was
dismayed to find that a hyaena had indeed tried to get at it. The persistent beast had managed to remove much of the protection before giving up. It caused considerable damage to this crucial specimen, which I needed if I was to persuade the pig fraternity that the earliest notochoere was actually Notochoerus jaegeri. Still, the shape of its cheek teeth showed conclusively that this reclassification was needed.
Higher up the food chain, and therefore much rarer, are the carnivores, and we collected an impressive assemblage at Kanapoi that was larger and more diverse than any found at other East African Pliocene sites—eight species in eight genera that represented five families. They beautifully document the first post-Miocene radiation of endemic African species. In contrast to other carnivore assemblages of the same age, the Kanapoi carnivores include only species whose immediate forebearers are found in Africa. These formidable predators thrived off rich pickings, which included anamensis.
* * *
BUT WHAT EXACTLY was this early hominin, and how different was it to Lucy’s kin, Australopithecus afarensis? Most intriguing was the question of how the anamensis hand would differ from a chimpanzee’s. This is critical to establish when our ancestors acquired the dexterous hands that are so essential to all the fine and fiddly tasks that were prerequisites to our ancestors’ ability to make stone tools, which in turn fuelled the greater brainpower that led to the progression of technological advancements that brought us to where we are today.
In the early days of human palaeontology, scientists assumed that our large brains were the key feature that set us apart from apes and, fooled by a forgery—Piltdown Man—thought that they had the fossil evidence to prove it. Discoveries of bipedal hominins with small brains in South Africa showed how wrong they were while Australopithecus afarensis eventually proved that the big brain came after bipedalism and increased dexterity. But the idea that manual dexterity is a result of freeing the hands from the constraints of walking on all fours remained an assumption with no fossils to show for it.