by Meave Leakey
This loss was felt most keenly for the elephant data, which were studied by a palaeontologist on Patterson’s team named Vince Maglio. Elephants are huge and robust, so they tend to erode out slowly and stay on the surface for years. Vince found and collected almost all of the best elephant specimens known from Lothagam—but for many of these, we simply do not have provenance data. We therefore had to make do without many of the most complete specimens for our study. The elephant story at Lothagam is one that would turn out to be one of the most interesting, and these beautiful but worthless specimens will always be the source of great frustration and regret for me.
One memorable day quite late in the season, Kamoya and I were putting the final few specimens on our photographs in the heat of high noon. We were in the Kaiyamung, a small area of fossil-rich sediments in the western reaches of the site, and we were using several of the small-scale photos since the enlarged ones did not extend this far. We had the hang of this at last. All the photos in the series were painstakingly marked with tiny pinpricks, each one representing a fossil whose accession number had been inked on the back of the photo.
A particularly strong and sudden gust of wind unexpectedly snatched the photographs out of my hands, and they took off in an up-current, over the next hill. We tore after them, gathering them up as we went, but the last photograph continued to elude us, snatched out of reach by the capricious wind each time that one of us bent to grasp it. We scrambled over rocky mound after rocky mound until we could grab the renegade photo. I looked at a flushed and dishevelled Kamoya—usually slow and deliberate as he meticulously scours every inch of a surface for bones. We simultaneously exclaimed with newfound respect and admiration: “I didn’t know you could be so tough!” We collapsed in the shade, exhausted and laughing with relief that our precious provenances were safe. Back in camp, I carefully transferred all our specimen numbers onto tracing paper so I had some sort of second record should another dust devil or unanticipated misadventure catch us unawares.
This was not long before pricking a hole in a photograph as the sole record of where a fossil was collected would become quaintly archaic. We still use photographs to this day—but all of our provenance data are also now recorded with GPS coordinates and backed up on computer. It was while we were at Lothagam that I first came across global positioning. Richard flew in for a weekend with his chief pilot, Phil Matthews, who proudly demonstrated his new Trimble device—one of the earliest commercially available GPS models, which was huge in comparison to the pocket pieces that would follow. Coming hot on the heels of our epic chase up and down Lothagam’s rocky peaks, a large portion of that day’s entry in the field diary was devoted to extolling the virtues of this wonderful new machine and what a difference it would make if only we had one. But this was all still in the future.
One of the main intellectual challenges for me when we first arrived at Lothagam was to make sense of the many huge geological faults. These are, of course, one of Lothagam’s greatest assets since the faulting is what has exposed such a long time sequence in a relatively small area. However, the faults do make it harder to interpret what’s going on, especially for the uninitiated. “How can you tell which is up and which is down?” I wrote despairingly in my field diary after Kenyan geologist Patrick N’gang’a, Kay Behrensmeyer, and I had walked the entire length of the site for three hot, exhausting days. A case of too many cooks, perhaps, for they vehemently disagreed about whether we were picking up the main marker bed or something else as we followed it through the southern exposures. As for me, I no longer knew what to believe.
Lothagam’s geology was first described by Bryan Patterson and William Sill during the 1967 American expedition. While still a student at Harvard, Kay did some further studies in 1968 when she was considering doing her doctoral thesis on Lothagam’s geology. Kay had gone on to do her PhD at Koobi Fora, studying the processes of fossilisation (part of the field of study known as “taphonomy”), but she agreed to show us around Lothagam at the beginning of our first season of work there. Her excellent maps formed the basis of the work that Frank Brown’s former student Craig Feibel compiled when he joined us for subsequent field seasons.
The problem for both Patterson’s team and Kay was that they had enormous trouble finding suitable material for dating. Twenty-five years later, we encountered the exact same problem. The main marker bed, called the Marker Tuff, is a thick prominent bed of distinctive pinkish-red rock full of sparkling crystals with a secure date that showed the lower sediments were older than 6.54 million years. But other than this, dating material eluded everyone. When Craig later joined us, he searched high and low to find pumices suitable for dating but only found tiny bits of material that appeared to have been washed into small ponds. In desperation, he sent some of these off to Ian McDougall, an Australian geophysicist. Craig was banking on the slim chance that Ian might possibly be able to analyse these minuscule samples with newly available dating techniques and the new machines that had recently been installed in his lab. To our amazement and delight, Ian succeeded. He came back to us with a series of dates that indicated that the fossiliferous sediments at Lothagam range in age from more than 7.5 to less than 3.5 million years—exactly the time interval when the earliest hominins are thought to have split from our common ancestor with apes.
Armed with Craig’s new dates and my newly trained eye, I would see a much clearer picture of Lothagam’s geology were I to retrace my earlier steps with Kay and Patrick at the end of four field seasons. The site is an uplifted fault block about ten kilometres long and six kilometres wide. Two roughly parallel ridges oriented in a north-south direction protect lower-lying exposures in a central valley between them. These exposures slope gently down to the north and south from a sandy saddle of much younger deposits, which bridges the parallel hills to form a distinctive H shape. The site is named after the highest easternmost ridge, Lothagam Hill. In the Turkana language, losthagam describes something rough, varied, and heterogeneous, an apt description that eloquently captures the area’s colourful assortment of rocks. There are extensive conglomerates, some with huge boulders and others with smaller stones, pebbles, and cobbles cemented together with sandstones. Basalt horizons cap the top and sandwich the conglomerate layers. One of these is a spectacular outcrop of columnar basalt, impressive in the perfect symmetry of the hexagonal pillars arranged in eerily neat rows.
Lothagam Hill is a horst, the geological term for a block of sediments left standing undisturbed while the sediments on either side are downfaulted. Consequently, although the horst is the highest topographical feature of the site, it is also the oldest, deposited roughly between 14.2 and 9.1 million years ago. Had there not been the disruptive tectonic activity, the oldest sediments would be underneath all the consecutively younger layers. But not at Lothagam, where oldest is at the top, and the subsequent layers of sediments are found not only at a lower level at the sides of the horst, but they are also tilted, split into separate units, and interrupted by erosional gullies. No wonder it was so confusing!
Craig renamed the ancient sediments older than nine million years the Nabwal Arangan beds after the nabwal arangan or “red waterhole,” a muddy pool of startling rouge that sits in a gorge cutting through the horst. The only fossils recovered in the Nabwal Arangan beds were fossilised wood, so we did not spend much time prospecting there. The view from the top of the horst is spectacular, however, and sometimes we got up early on Sundays to climb it. The hill towers over the rest of Lothagam, and to the east, the lake glistens in the sun in ever-varying shades of jade, and on a clear day, one can also see the hills flanking the opposite shore. High, steep-sided sand dunes continually form along the base of the horst as the strong easterly winds dump sand from the vast surrounding desert. To the west is a panoramic view of the entire site between the horst and the parallel lava-capped ridge that forms the western boundary.
Cutting across the exposures to the west of the horst are successively younger exposure
s. The sediments are all tilted at a steep angle with long sloping pavements of hard rock terminating in almost vertical cliffs. Craig named the next oldest group of beds in the sequence the Nawata Formation. Nawata is a Turkana name for a distinctive type of grass that grows in the seasonal Nawata River, which drains the northern half of Lothagam’s central valley and cuts through these sedimentary deposits. The grass looks as though it is made from a cluster of miniature bottlebrushes bound together at their base and mounted on top of a long narrow stalk, the whole composition burnished a rich gold by the desert heat. Apart from a few doum palms at the northern end of the riverbed, the ornate nawata grass is virtually the only vegetation in the hot, barren, sedimentary badlands.
Deposition of the Nawata sediments began soon after 9.1 million years ago and continued for nearly four million years. Some eighty metres above the base of the sequence, there is a distinctive layer of bright brick-red sediment called the Red Marker. Craig divided the Nawata Formation into two members, the Lower and Upper Nawata, and picked the Red Marker to indicate the top of the Lower Nawata sequence. The Upper Nawata begins with a layer of volcanic ash, which is the Marker Tuff that we were following that memorable day with Kay and Patrick, and lies directly above the Red Marker. The volcanic eruption that produced the ash of the Marker Tuff happened 6.54 million years ago, giving us lower (9.1 million) and upper (6.6 million) chronological boundaries for the Lower Nawata beds. Another very distinctive marker bed, the Purple Marker, defines the top of the Upper Nawata sequence. Ian McDougall’s analyses gave us a good series of dates for the Lower Nawata beds that allowed us to identify the age of the fossils very precisely. But we had enormous trouble identifying dating intervals for the Upper Nawata beds. Apart from the Marker Tuff date of 6.54 million years at the base of the sequence, there are no other secure dates for the Upper Nawata beds, including for the Purple Marker tuff. This dearth of dates is terribly frustrating—all the more so because the Upper Nawata was a time of great upheaval.
The sedimentary layers in the Nawata Formation are mostly conglomerates, mudstones, and sandstones. Tellingly, the sandstones are “upward fining”—with coarser particles at the bottom and finer ones at the top. This is the typical deposition of a slow-flowing river system, where the largest, heavier particles settle out first. With the buildup of sediment on the edges of this ancient meandering river that ran across Lothagam all those millions of years ago, the outer curves of the riverbed got shallower, and the river shifted course slightly as it cut a new and deeper channel. The water on the edge slowed down, and progressively finer materials settled on top of the coarse layers laid down by what was once the fast-flowing centre of the channel. As the river wove its way back and forth on the floodplain, it cut a new channel through sediments laid down earlier; the pattern that the sediments form is called braiding for its resemblance to a plait we might braid in our hair.
Differences in thickness and composition of the sedimentary layers show that there were fluctuations in the volume of water and the speed of the river throughout this time period. Massive fossil reefs of the Nile oyster, Etheria elliptica, laid down in sandstone channels are common throughout the beds of the Lower Nawata, and their presence demonstrates that the river complex was flowing all year round as these molluscs thrive only in well-oxygenated fresh water. We also found thin limestone layers with minute bivalve molluscs called ostracods in them. Interpreting the rocks, Craig told us that this used to be a lush riverine environment with broad shallow channels, back swamps, and oxbows formed as the meandering river cut new channels. The landscape was a mosaic of floodplain savannas dissected by gallery woodland tracing the river’s path, which provided a rich and bountiful habitat that could support a large diversity of life. During most of this time, a large perennial river was flowing, but the fossil soils showed that during two intervals, around 6.7 million years ago and 5.2 million years ago, deposition of sediments slowed practically to nothing, which suggested intervals that were considerably dryer.
Above the Nawata Formation beds, a new set of sediments, the Apak Member, form the earliest sequence of the Nachukui Formation. The geology of the Apak is a bit more complex than that of the Nawata. Thick beds of multistory deposits, each characterized by coarse-grained sandstones at the bottom and built up with progressively finer materials and mudstones towards the top, tell their own story. This was also a river system, but one that was faster flowing and with less back swamps. The absence of the Nile oyster suggests that the river might have been seasonal, and Craig believed it represented a different drainage system from the Nawata, perhaps ancestral to the modern Kerio River that currently lies to the east of Lothagam.
At the end of the Apak interval, some 4.2 million years ago, there was a volcanic eruption that led to the creation of the Lothagam Basalt, which tops the easternmost of Lothagam’s parallel ridges. Just before the volcano that spurted this basalt erupted, the river system that had persisted through much of the Apak was replaced by a lake. There are impressive beds of distorted, squashed snails where the red-hot ribbon of thick magma encroached inexorably on the lakeshore and preserved them for posterity.
As you clamber up the exposed lake beds at the top of the Apak onto the Lothagam Basalt, you are rewarded with a spectacular view across the younger beds to the west. The ancient lake, which begins in the uppermost beds of the Apak Member with the squashed snails, continues through the Mururongori Member. These lake beds are important because they correlate with the distinctive olive-green fine clays found all around the basin. They are all that is left of the huge Lonyumun Lake, which marks the beginning of the sediments in Omo-Turkana Basin as we know it today. I’ll come back to the story of their geology later on for we come across the waxing and waning of a lake in the basin over and over again in the years to come. At any rate, with the geology of Lothagam unlocked at last, we were fast learning that most of the evolutionary action was going on before the vast Lonyumun Lake swamped the landscape some 4.1 million years ago.
The sediments in the Nawata Formation and Apak Member contain the majority of fossils, and the story that the fossils tell echoes Craig’s interpretation of the geology. From the vantage point of modern-day Lothagam’s bare and barren rocks, it is hard to imagine such a diverse and watery world. Hippos were everywhere—they turned out to be the most frequently preserved mammals when we tallied up the fossils at the end of four field seasons, making up 27 percent of the mammalian fauna. And no other East or Central African fossil locality rivals Lothagam for the diversity, abundance, or quality of preservation of turtles and crocodiles.
Turtles from the Nawata Formation include at least six different species. By far the most common was a brand-new genus and species of side-necked turtle, Turkanemys pattersoni, so named because these turtles fold their necks sideways within the protective shell rather than tucking them in as other turtles do. The type specimen for this turtle was found by Patterson’s team. It is exquisitely preserved and almost complete, with the skull, mandible, and most of the skeleton tucked inside the shell. We found countless other specimens of this turtle and were initially puzzled by the fact that almost all of them were adults of roughly the same age. This pattern was very similar to that seen today at modern nesting sites, where large numbers of adult turtles congregate to breed, always returning unerringly to the same place year after year.
Five species of crocodilians also rubbed scaly shoulders, several of which grew to formidable lengths. The most common was a very large broad-nosed crocodile. The biggest skull and mandible we found turned out to be roughly a metre long when the pieces were reconstructed. Only the year before, Samira had been assigned a statistics project, and it just so happened that she had picked crocodiles as her subject. Studying different types of modern crocodiles, she worked out that the species all have similar ratios between head, body, and tail lengths probably because the weight must be evenly distributed for the animal to be able to move efficiently, especially in water. This entailed some rat
her perilous practical data collection perched on the walls separating the holding pens in the reptile park at the museum in Nairobi while juggling camera and measuring stick. My support for the project suffered rather grave doubts when she came to my office telling me how close she came to falling in when a crocodile lunged at her! We plugged our big specimen into her ratio to see how long it might have been. The result was a whopping twenty-one feet—about the same size as big male saltwater crocodiles in Australia, the largest reptiles alive today.
It is mind-boggling to think that this aquatic environment could support so many of these large formidable predators. They must have had a high degree of specialization in different ecological niches because there must have been a remarkably abundant and diverse range of prey for them to eat. Among this bountiful prey were large numbers of fish. In the Lower Nawata beds, the fossil fish were all small-sized species that eat other fish and vegetation and thrive in well-vegetated bays and swampy shallows. In the Upper Nawata, these fish become less common, and by the Apak, the fossil fish represent species that prefer more oxygenated water consistent with a faster-flowing river system. We also found a number of birds, which, apart from ostriches, were all species at home in an aquatic environment. The terrestrial fauna included many species typically inhabiting well-watered, well-vegetated habitats, and we were finding a considerable diversity of monkeys that would have thrived in the swamps and the gallery woodland along the river.