Long Read | Baobabs tell of an ancient African past

These giant trees can grow to be as old as 2 500 years and dominate areas in southern Africa where people lived in ancient times. They may also help us understand our climate-challenged present.

There is a trail, 1 000 years old, linking Vilanculos on the coast of Mozambique with the home of the little golden rhino, Mapungubwe, which calls to be travelled end to end to get a sense of times past, 600 years before the first settlers arrived in the Cape.

These places and the stories of the people who lived here, transacting along the trail in gold, iron, ivory, ceramics, cloth and glass beads – from as far away as Arabia, India and China – lay unearthed for centuries, only in recent decades to be rediscovered and celebrated.

As you make your way along the trail, much of which is unchanged from a millennium ago, you cannot but notice an omnipresent feature of the landscape then and now, which shaped where and how people lived, and was given life by people who helped propagate and populate it.

This is the baobab, which can grow as old as 2 500 years and so dominates some areas where people lived in ancient times that there can be hundreds in a small area. By journey’s end, we would not only have eaten the baobab but drunk it, too.

Besides Mapungubwe, we’d also visit the ancient sites of Thulamela in the northern Kruger National Park, and Manyikeni and Chibuene in Mozambique, the latter being the main port of entry for southern African through the Indian Ocean, from about 1 250 to 1 000 years ago. These sites have been uninhabited for as long as 800 years. The people moved on, but the baobabs, perhaps no more than saplings then, are still there.


On Mapungubwe hill, where previously only royalty and courtiers resided in southern Africa’s first stratified society, guide Cedric Setlako showed where the little rhino, and much golden treasure besides, were found in the 1930s by white farmers probing local mythology that said treasure was hidden in the area.

A gold rush frenzy by archaeologists and treasure hunters followed, the University of Pretoria taking a cache of artefacts, but it was only after the advent of democracy that the remarkable richness of Mapungubwe culture became generally known, a case study of how ideology shaped and guided research and questions of ownership versus custodianship of the past.

Mapungubwe is a fine but relatively little visited park compared with better known ones such as the Kruger. It has a bounty of animals, especially elephants, many of which moved in from Botswana after the park was established in 1995. It has a unique, gnarly topography and splendid views overlooking the confluence of the Shashe and Limpopo rivers, which demark the countries of South Africa, Botswana and Zimbabwe.

Related article:

A long time ago, a couple of million years, the Zambezi flowed into what is now the Makgadikgadi pans and then down the Shashe into the Limpopo. This is a special place of confluence. On earlier visits the Shashe, with its wide bed of white sand, has always been dry, but now it held a lot of water that covered much of its expanse.

Mapungubwe was the centre of a trading network from about 1 000 to 800 years ago when, for reasons unclear, Great Zimbabwe, 350km to the north, took over as the trading epicentre that dominated the region until about 500 years ago.

The park probably has as many baobabs as elephants. What would these sentinels, which were here perhaps as long ago as a millenium before the little rhino and other golden treasures were buried with royalty on Mapungubwe hill, tell us if they could?

December 2019: Guide Cedric Setlako shares the many secrets and treasures of Mapungubwe, the centre of a trading network a millennium ago, with visitors. Archaeological digs have unearthed woven textiles incorporating baobab bark. (Photograph by Kevin Davie)


We headed east to Thulamela, an ancient site in the very northern part of the Kruger National Park, camping for a night under a giant baobab at the Pafuri River Camp on the border of the park.

Manager Martin Erlank, who grew up on a farm alongside Mapungubwe and knows the area well, had many stories to tell. With a squat physique, vaguely suggestive of the giant tree, Erlank sells baobabs from as little as R50, but this would be a useless purchase on our part as it would be unlikely to survive winter in our hometown Johannesburg.

What does a young baobab look like? Erlank pointed to one that is 13 years old, just a spindly tall thing with no presence at all. He explained that baobabs grow relatively quickly to their full height and then spread out from there.

Erlank described what happens when a baobab dies, something he witnessed 40 years ago. The dead tree lay, felled, for a few days and then exploded.

December 2019: The rock walls at Thulamela in the northern Kruger National Park, which dates back 800 years, have been reconstructed in recent times. (Photograph by Lynn Morris)
December 2019: The rock walls at Thulamela in the northern Kruger National Park, which dates back 800 years, have been reconstructed in recent times. (Photograph by Lynn Morris)

Flowers from the tree above us fell during the night, including on our tent, with a gentle splat. It is worth knowing that Adansonia digitata, the formal name of the tree, is neither male nor female. Its flower, which is pollinated by bats and hawk moths, not bees, has male and female parts. Yet for reasons still unclear, some trees typically do not produce fruit and are seen as male while those that do are thought of as female.

Thulamela is not signposted, but with a bit of trial and error we came to a parking place with a set of interpretative boards that tell its story. Discovered by a park ranger in 1983, archaeology has dated its beginnings to about 800 years ago. Somewhat later, about 450 years ago, residents began building rock walls to fortify what is already a natural fortress overlooking the Levuvhu River.

The rock walls lay out the social order of the place. The boards explain that this conforms with the ideology of the VhaVenda people today: the sacred leader called the Khosi lived in a hilltop complex in ritual seclusion out of view of the commoners below. “Monoliths and horns of sacred cattle were often placed on the walls surrounding his residence; these functioned as a symbol of justice and defence.”

The graves of two VhaVenda royals are marked and have prominence on the site. 

Masons from the area rebuilt the rock walls over a 14-month period in a project sponsored by mining firm Gold Fields, so visitors can get a real sense of Thulamela’s life and times. Tours can be arranged, but we had the place to ourselves. It was just us and the site’s sentinels: giant baobabs, some in rows, sprawling, watching, guarding, not revealing.


We crossed into Mozambique at the Pafuri border post and were subjected to a little scam by officials on the Mozambique side, who charged us 100% tax on the additional fuel we were carrying, heeding advice that there’d be no petrol between Pafuri and the coast. The actual duty should have been 26%, but the confusion of forms and language meant that these petty bureaucrats were able to shake us down for the rest. What tithes and tribute applied, one wonders, in times past as items moved between the coast and the hinterland?

A thousand years ago, there was no bridge to cross the Limpopo here. This is still the case for hundreds of kilometres of the river. At low levels you simply engage 4×4 and drive across. Lourens Jooste, a missionary who runs the Dumela campsite on the Limpopo, where crocodiles lurk, reckoned the river would still be low enough to drive across. But when we arrived, a man who was helping to operate a makeshift pontoon indicated that the water was waist high.

You drive your vehicle up ramps on to the pontoon, which six or seven men then walk across. There were long poles on the pontoon, presumably to be used when the water is high. 

“Are there crocodiles?”

“Only when the water is high,” one of the men, waist-deep in the river, answered. 

The trail leaves the Limpopo River here, at Mapai. What followed was sections of wide dirt road and narrow jeep track winding through the bush. We had an up-to-date Mozambique map and Google Maps, which differed on key details. Not that the difference mattered as both were hopelessly wrong for long sections of this part of the trip.


Research indicated that there was another ancient site we could visit, Manyikeni, about 50km inland from Vilanculos. Manyikeni, which has multiple spellings and was marked on Google Maps in three different places, was not easy to find. But with the help of locals, we eventually came across signposts that narrowed our options until we found it.

Manyikeni, by the way, is only 3km off the present, modern dirt road.

There was a small museum, which was locked. A man, Phillips, appeared, saying the key was in Vilanculos. His English was as good as our Portuguese. Plaques on the site told its story, but also in Portuguese. Phillips showed us around the site that at one time connected to Mapungubwe and, later, to Great Zimbabwe about 400km away. At one spot he pointed to a large, hollowed out area that he indicated was where water was sourced.

Archaeology dates the occupation of Manyikeni between 800 and 400 years ago. 

The rock walls were not as imposing as at Thulamela but, as Martin Hall and Rebecca Stefoff point out in Great Zimbabwe (2006), the area is not as endowed with rocks for use as building material. The baobab sentinels, too, were not nearly as impressive in number or size as elsewhere.

The most striking tree at Manyikeni, nowhere near the size of the monsters at Thulamela, had markings about a foot apart from bottom to top. It seemed that pegs had at one time been driven into it, perhaps to create a ladder. A baobab expert later confirmed this had been the likely usage. Trees with pegs in them can still be seen to this day.

The sites we visited in cases predated the Great Zimbabwe empire, which came to dominate the region. Isaac Samuel, in a recent article on the Uncensored Opinion website titled “Great Zimbabwe and the stone ruins of southern Africa”, captures much of its grandeur and influence.

“The massive stone edifices of Great Zimbabwe are the ruins of what was once one of southern Africa’s largest cities; just one of the city’s sections – the great enclosure –  was constructed with 14ft-thick walls using over a million granite blocks raised to a height of over 33ft. The city is estimated to have produced up to nine million ounces of gold at its height, with population estimates peaking at 18 000 for a region covering seven square kilometres.

 “At its height, the city’s dwellers ate on Chinese porcelain, dressed in Swahili and Indian silk and wore gold and copper jewellery, both foreign and locally manufactured, and exported large amounts of gold and ivory,” Samuel writes.

Related article:

“Unlike its peers however, Great Zimbabwe has been the subject of much controversy, perhaps the most of any medieval city from which has arisen a debate at the intersection of race, colonialism and nationalism where the most disparate groups of people were ever attributed to a settlement – from Phoenicians to Arabs to Assyrians to Indians to Europeans – all of whom were living tens of thousands of miles from the Bantu-speaking Shona people who built not just the city itself but the over 1 000 similar cities in the regions now located in the modern countries of Botswana, Zambia, Mozambique South Africa and Zimbabwe itself.”


Accommodation inquiries at Vilanculos took us to Baobab Beach, a lodge with a massive baobab to welcome visitors. The lodge is on the beach, where fishermen use dhows and dugouts. The tidal range is impressive and there are views across to the island of Magaruque and the archipelago.

Chibuene, about 5km down the beach, had been for hundreds of years the southernmost port on an Indian Ocean trading network stretching up the African coast to Arabia, India and China beyond. It had taken over from Raphta, off the Tanzanian coast, which had been the terminus of this trade 2 000 years ago.

The site was occupied during two distinct phases. The earlier phase of occupation dates to about 1 000 years ago, the second from around 1450 at the time of Great Zimbabwe.

“Chibuene is the most southerly first millennium port known to have been a participant in the Indian Ocean trade on Africa’s eastern seaboard. It is also the earliest on the southern coast, with its role as a port lasting from at least the seventh to the mid-10th century AD,” says Wits University’s Marilee Wood in a paper titled “The glass of Chibuene, Mozambique: new insights into early Indian ocean trade, 2012”.

“The large numbers of glass beads and vessel shards recovered bear witness to its role as an important trading hub. Its connection to an extensive interior trading network can be seen in the large numbers of glass beads, identical to those from Chibuene, that are found in a hinterland stretching over 1 500km from east to west,” Wood and her co-authors say.

“Based on the number of exotic goods found there, it has been recognised as the port of entry for trade goods, especially glass beads, for Mozambique and the interior – including the Shashe-Limpopo region, the Zimbabwe plateau and Botswana as far west as the Tsodilo Hills – up to about AD 1000.” 

We negotiated through Vilanculos’ busy market and sandy streets, heading south by a somewhat circuitous route to an area close to the shore of homes and a few lodges. A sign read “Chibuene Historical Heritage. Gateway to Great Zimbabwe.” It provided a cellphone number and email contact details. We drove 100m to a closed gate, behind which was a small museum similar to the one we’d seen at Manyikeni. It was also locked and there was no one to rouse. Baobabs were in attendance. Of course. 

But where was Chibuene, this great trade terminus? We took various routes to the seafront but found no physical evidence of anything resembling a port. We phoned the cellphone number, coming up with the same language barrier as at Manyikeni. We drove to a nearby lodge to ask directions. It turned out we had been at the right place so we drove back to the museum, parked and walked down to the beach.

Related article:

The scene was no different to the rest of the coastline: a dhow or two, a couple of dugouts and a modern speedboat. But walking towards what appeared to be a cove that would give a bit more shelter to moored boats, we saw that the shore, which was raised at the point, had been cut away, presumably by strong currents, exposing middens of shellfish several metres thick. People had – for hundreds of years – broken open their shellfish here, harvesting the flesh and checking for pearls before discarding the shells.

“Although Chibuene has generally been considered a remote extension of the trade on the eastern African coast, the substantial numbers of glass beads and shards unearthed at the site, as well as large numbers of beads in the interior, indicate that Chibuene was not like the fledgling coastal communities developing in the north,” says Wood.

“The copious exotic finds at Chibuene underline its role, for at least two and a half centuries, as an important and active trading port that was the hub linking Indian Ocean trade with an extensive interior trading network. When the port location moved elsewhere in the mid to late 10th century, Chibuene lost its raison d’être and was either abandoned or became a quiet backwater.”

This is Chibuene now. A single fisherman arrived in his dugout and a father and his two children walked along the beach. It was otherwise deserted.


What had the couriers eaten while carrying tradeables to market and back? At Baobab Beach, a manager, Marina Ferraris, suggested that baobab seeds had at least in part nourished these ancient travellers. She offered to introduce us to a friend knowledgeable in the history and economy of the region.

Bernhard Wiemer, an economist, political analyst and long-time Mozambique resident, said over lunch that while he had not personally seen it, pilots reported a trail of baobabs lining the ancient trade route inland, people having helped to distribute the trees along the trail. Later, I read of researchers who look for clusters of baobabs to locate these ancient sites.

In these parts, Mozambicans not only eat the baobab seed but they drink it, too. Malambe is almost a national drink. So popular is it, in fact, that we could order cold baobab juice from the bar just a few metres away.

A glass soon arrived, like a light brown chocolate milkshake. But malambe is not like a milkshake or sugary drink; it is closer to kombucha. Citrus-flavoured, it is a little bitter with a slight fermented fizz.

Health websites proclaim the baobab seed to be a superfood like broccoli, oranges, eggs, apricots and bananas. “A 15 gram serving of baobab powder is the equivalent to eating as much fibre as three cups of broccoli, the same vitamin C as one orange, the equivalent calcium of three eggs, as much potassium as four dried apricots and the same magnesium as nearly one banana,” says Limpopo-based EcoProducts, which has partnered with Zimbabwean firm B’Ayoba to supply baobab powder from southern Africa. 

Related article:

The FoodNavigator website reported in January that B’Ayoba sources baobab powder from 5 000 harvesters in Zimbabwe, which have five million baobabs, as well as from South Africa.

Pretty much every Mozambican market in the region sells the chalk-like substance. We even saw buckets of malambe being sold on the main road. A woman at one roadside stall took a branch and smashed open the pod, revealing the white strips used to make malambe.

In the market at Vilanculos we bought a bag, making the rookie error of buying seed that was a little too pink for the untrained eye. You want it more white than pink. We also overpaid by 400% or so, the plastic bag-full should have cost no more than 10 meticais, about 2.5c.

Clara Waziwei, a manager at Baobab Beach, took us through the process of making malambe. The tartaric acid-like white chalk is crushed into a powder. Water is added and strained. Then a spoon of sugar and/or honey is stirred in before the drink is put in a fridge overnight to be drunk the next day.

Malambe is her favourite drink and she has several baobabs growing in her garden from the discarded seeds, some about five years old.


The managing director of EcoProducts B’Ayoba is Sarah Venter, who has a doctorate in baobab ecology and runs the Baobab Foundation, which aims to promote and protect the tree. Venter lives in the municipality of Makhado (Louis Trichardt) and talks about certain individual trees with such familiarity that you could be forgiven for thinking they were her personal friends.

Studies in Mali, West Africa and Australia have shown a symbiotic relationship between baobab and man. Up to 300 separate uses for the tree have been identified; people use the tree as a resource and in the process help spread it.

Baobab fruit is unique in that it is dry when ripe, says Venter. Other fruits are wet when ripe, giving them a short life. Baobab fruit is easy to carry. Waziwei at Baobab Beach also told of it being carried to be eaten as a snack.

Venter says people have helped distribute the baobab, not through active domestication but by eating the white portion of the pod and discarding the seed. DNA studies suggest a West African origin for the tree, it moving in part through human assistance across the continent and to western Madagascar, coastal Arabia, India, Australia and, in more recent slavery times, Brazil.

Venter says archeological digs at Mapungubwe have recovered baobab fibres woven into strings and cords. The younger trees are the most sought-after for bark harvesting. The scars grow with the tree, so marks from perhaps 1 000 years ago are prominent on the larger trees today.

November 2010: The largest baobab in the Mapungubwe National Park. (Photograph by Gallo Images/ Go!/ Villiers Steyn)
November 2010: The largest baobab in the Mapungubwe National Park. (Photograph by Gallo Images/ Go!/ Villiers Steyn)

She says stone-wall settlements in the area, including sites with unknown names that have not been researched, can have “tons of baobab trees” from people eating the fruit and discarding the seeds, leaving it “tangled with baobabs”.

Xander Antonites, an archeologist at the University of Pretoria, has studied several sites in the Limpopo valley and Soutpansberg, including those that are not the main site or capital, to get a better sense of the complex and changing human occupation patterns in the ancient past.

Antonites said the people who lived there were cattle-rearing. The sites generally produce more remains of goats and sheep, but this could mean that these animals were more frequently consumed compared with cattle. He added that not enough was known about their agriculture as little work has been done on what crops they grew.

He agrees that humans have influenced the dispersion of baobabs, but stresses that this is not a one-to-one correlation as other factors are likely to have been in play. Human-assisted dispersal is also seen in the case of the marula, he says. People sit under the tree, eat the fruit, discard the seed and make another marula.


Unlike other trees, baobabs do not have rings and cannot be dated using this technique. So they are mute in this respect, too. But in recent years, scientists have been using radiocarbon dating to accurately date the trees. A cored sample is taken under authorised and controlled conditions by drilling a narrow hole into the tree, the core allowing for its entire history to be dated.

In this way, the tree tells of the climatic conditions during its past. A study titled “A 1 000-year carbon isotope rainfall proxy record from South African baobab trees”, with Stephan Woodborne of iThemba Labs, a laboratory for accelerator-based science in Gauteng, as the lead author, says this proxy record “is in good agreement with a 200-year tree ring record from Zimbabwe, and it indicates the existence of a rainfall dipole between the summer and winter rainfall areas of South Africa. The wettest period was c. AD 1075 in the Medieval Warm Period, and the driest periods were c. AD 1635, c. AD 1695 and c. AD 1805 during the Little Ice Age.”

Radiocarbon dating of baobabs is allowing archeologists to match climatic conditions to what they unearth. For instance, Thomas Huffman and Woodborne used baobab data to confirm an early 14th-century drought associated with the abandonment of Mapungubwe. But they add, in “Archaeology, baobabs and drought: Cultural proxies and environmental data from the Mapungubwe landscape, southern Africa, 2015”, that “this abandonment, however, owed as much to cultural factors as to environmental pressures”.

While the trees in this way tell of the past, they are also speaking of the present. An 800-year-old tree at Klein Baloyi near Musina has been hooked up by Woodborne to its own weather station to record present-day climatic conditions.

Antonites said the tree has the potential to tell us about long-term climate change, such as changes in rainfall and temperature. 

“Baobabs are, in a sense, living records of the past – museums that store information about the environmental backdrop of the ancient societies we study. During the lifespan of these trees, numerous cultures and societies have come and gone, and the trees contain information about them. But since they are living, they also store information about the present,” he said.

“Archaeologists often conduct thought experiments, where you try and imagine what future archaeologists would say about us in 2020. Seen in the context of current accelerated climate change, these trees may actually chart the demise of the world as we know it.”

Related article:

We had overnighted at this site, with its spectacular baobabs and outsized rock domes, coincidentally on our journey along the trail. Archaeological work by the University of Pretoria over the past few years at Klein Baloyi, also a satellite, dates occupation to about 50 years earlier than Mapungubwe.

A radiocarbon dating project of 13 monumental baobabs in southern Africa, published in the scientific journal Nature Plants in 2018, said that nine of the 13 oldest and five of the six largest individuals studied had died, or at least their oldest parts or stems had collapsed and died, over the past 12 years. The authors said the cause of the mortalities is still unclear.

Media reports subsequently speculated that climate change could be the reason for the deaths. Venter, though, sees it differently. Baobabs are 75% water. As they get larger and larger, they can struggle to contain their own weight.

She says the giant trees develop a hollowed out trunk as an architectural device to spread and share the weight. But they are also known to collapse and appear dead, but then keep on growing.

Only a few have died, says Venter, of the few million out there. Her bet, in fact, is that as climate change does wipe out numerous species, the resilient baobab will be the last tree standing.

If you want to republish this article please read our guidelines.