THE Amazon rainforest is so vast that it boggles the imagination. A person could enter at its eastern edge, walk 3000 kilometres directly west and still not come out from under the vast canopy.
This haven for about 10 per cent of the world鈥檚 species has long been regarded as wild and pristine, barely touched by humanity, offering a glimpse of the world as it was before humans spread to every continent and made a mess of things. It is painted in sharp contrast to the logged forests of Europe and the US.
Advertisement
But it now seems this idea is completely wrong. Far from being untouched, we are coming to realise that the landscape and ecosystem of the Amazon has been shaped by humanity for thousands of years. Long before the arrival of Europeans in the Americas, the Amazon was inhabited, and not just by a handful of isolated tribes. A society of millions of people lived there, building vast earthworks and cultivating multitudes of plants and fish.
We don鈥檛 fully understand why this flourishing society disappeared centuries ago, but their way of life could give us crucial clues to how humans and the rainforest could coexist and thrive together 鈥 even as Brazil鈥檚 new government threatens to destroy it.
Some of the first Europeans to explore the Amazon in the 1500s reported cities, roads and cultivated fields. The Dominican friar Gaspar de Carvajal chronicled an expedition in the early 1540s, in which he claimed to have seen sprawling towns and large monuments. But later visitors found no such thing.
The mysteries and rumours didn鈥檛 die down. The most famous is the legend of El Dorado, a fabled city replete with gold. Many explorers died failing to find the non-existent city. In the early 20th century, British soldier and geographer Percy Fawcett became convinced that there was a lost city he dubbed 鈥淶鈥 in the Mato Grosso region of Brazil. Fawcett disappeared in 1925 searching for it.
The idea of the pristine Amazon became gradually more established. And then when US archaeologist Betty Meggers explored the region from the 1940s, she argued that the Amazon鈥檚 lushness was deceptive and could never have supported many people. Most of the soils are acidic and poor in nutrients.
The tide wouldn鈥檛 start turning until the 1990s. One important challenge to the notion of the untouched wilderness came when archaeologist , now at the University of Florida in Gainesville, arrived in the Amazon for the first time. He spent 1993 living in a village of the Kuikuro people in the Upper Xingu region of Brazil. 鈥淲ithin about two weeks, the chief brought me up to a big ditch, which we determined was a big moat associated with a palisade, with a village that was at least 10 times the size of a contemporary village,鈥 he says. 鈥淭hen he took me to another and another.鈥
Clearly there had been many more people living here in the past, who were building on a much grander scale than their descendants. How was this possible if the Amazon wasn鈥檛 suited to a large population?
One crucial clue had been but wasn鈥檛 fully appreciated: terra preta, or Amazonian dark earth. Patches of this black, unusually fertile soil are dotted throughout the rainforest. Since the 1960s, soil scientists have suspected the terra preta became so fertile as a result of prolonged cultivation by people: in other words, that it was artificial. It was made from accumulated waste such as fish bones and seeds, which were burned to produce charcoal, according to of the National Institute of Amazonian Research in Manaus, Brazil. Levis suspects that at first this was a waste management practice. 鈥淏ut then later they probably realised that these soils were fertile.鈥
鈥淎ncient Greece had nothing like this, even at its peak鈥
Terra preta helps explain how the Amazon soils could have supported large societies, but it is only part of the story. For one thing the dark earths only date from the past 2000 years, and the Amazon has been inhabited for much longer than that.
The first colonists came to the Americas from east Asia between 17,000 and 13,000 years ago 鈥 although an including artefacts and burnt areas has been dated to 18,500 years ago, which implies an earlier entry. Either way, some of them reached the Amazon quickly, according to genetic evidence published in August, which compared DNA samples from 672 Amazonian people and reconstructed their population history. between 17,000 and 13,500 years ago (Molecular Biology and Evolution, vol 35, p 2719).
By 9000 years ago, the Amazonians had domesticated a number of crops. One of the most important was manioc, also known as cassava, which was . 鈥淲e now know that many of the most important crops in the world have been domesticated in Amazonia,鈥 says Jos茅 Iriarte of the University of Exeter, UK. Within a few thousand years, people began changing the forest in a big way. Around 4500 years ago, edible plants become more common in the fossil record, indicating that forest inhabitants selectively planted trees to produce fruit. around this time too.
Today, to some extent. They include sweet potato, tobacco, pineapple, hot peppers and even a palm tree: the peach palm. Fish were also a major source of food. Archaeological investigations at a site called Hatahara, where the Amazon and Negro rivers meet, by people living there between about 750 and 1230 AD. A staggering 37 species were on the menu, as well as freshwater turtles. A vast fish farm, complete with zigzag-shaped weirs to trap fish, has been discovered in Llanos de Moxos in northern Bolivia, at the south-west corner of the Amazon basin.
Home to millions
The increasing variety of foods enabled a rise in population that began about 5000 years ago, peaking around 1000 years ago. Quite what the population reached remains debated. 鈥淥ur estimations just for the southern rim of the Amazon is from 1 million to 5 million people,鈥 says Iriarte. Based on the quantity of terra preta, it has been estimated that there were at least 8 million people in the Amazon in 1492. US geographer William Denevan, in his 1992 paper , argued that 鈥渢here is substantial evidence鈥 that the Native American landscape of the early sixteenth century was a humanized landscape almost everywhere鈥 (Annals of the Association of American Geographers, vol 82, p 369). Denevan has suggested that the .
Since Heckenberger鈥檚 finds in the early 1990s, it has become clear that these societies also built some spectacular structures. Their remains just weren鈥檛 obvious. Many of them were swallowed by the forest after Europeans arrived, bringing with them diseases which decimated the native peoples. Explorers like Fawcett had gone looking for 鈥減yramids and cobbled streets鈥, says Heckenberger, but the Amazonians didn鈥檛 work with particularly durable materials. 鈥淭hey basically didn鈥檛 use stone at all, nor did they have metal.鈥 Instead their primary building material was soil.
Heckenberger has spent 25 years mapping out the lost settlements of the Kuikuro鈥檚 ancestors, the oldest of which are at least 1500 years old. There are about 20 towns and villages a few kilometres apart, each built around a central plaza that could be 150 metres across. Some are surrounded by defensive ditches up to 5 metres deep that stretch 2.5 kilometres. The villages were once surrounded by fields in which crops like manioc were grown. They seem to have been .
By around 1200-1300 AD, the villages were linked up with a network of broad, straight roads constructed of compressed earth. 鈥淭hey had this grid-like latticework of roads across the entire forested Upper Xingu basin. Ancient Greece had nothing like this, even at its peak,鈥 he says. Major ones were 10 to 15 metres wide, demarcated by raised curbs.
There may be a dreadful irony here. In his book , journalist David Grann suggested that these sites, known as Kuhikugu, might be the city that Fawcett was trying to find when he disappeared. Fawcett may even have walked right through it without recognising the earthworks for what they represented.
Such large-scale construction wasn鈥檛 unusual. Since 2000, Alceu Ranzi of the Federal University of Acre in Brazil has discovered a host of geoglyphs 鈥 large shapes marked out by earthworks 鈥 in Acre state, Brazil. Each geoglyph is a geometric form, often a square or circle, surrounded by a ditch. No cultural objects like pottery are found in them, suggesting they were kept clean, possibly for ritual purposes. They were often 100-300 metres across. Iriarte compares them to the Avebury stone circle in the UK, a site that often surprises visitors with its scale. By 2009, it was clear that had spread over hundreds of kilometres of forest, suggesting the existence of a thriving and complex society. Today, 450 geoglyphs have been found in Acre, years old. The culture persisted until not long before the arrival of Europeans (see 鈥Who lived there鈥).


The finds keep coming. In March 2018, Iriarte鈥檚 team reported an analysis of satellite imagery of the upper Tapajs river basin in Brazil in search of traces of settlements. They found 81, ranging from tiny hamlets to a town that occupied 20 hectares, many surrounded by ditches. Combined with the geoglyphs, this suggests that the entire southern Amazon, a belt 1800 kilometres long, was
There is disagreement on how to interpret the settlements. To Iriarte, 鈥渋t doesn鈥檛 look to be like a big urban centre鈥, but rather 鈥渧illages of pretty much similar size, spaced out in the landscape鈥. However, the villages in Acre are densely connected by roads. 鈥淚t鈥檚 amazing to see all these straight roads going from one village to another,鈥 says Iriarte.
The same is true for the villages that Heckenberger uncovered, and it has led him to argue that we should think of them as a non-traditional kind of city, a bit like the garden cities of the UK. 鈥淭hese were not anything at all like European cities or anything that Western historical experience had encountered in ancient Greece or Mesopotamia,鈥 he says. 鈥淭hey basically wove their cities out of the forest.鈥 The carefully planned network of roads suggests a central authority, and each village a 鈥渘eighbourhood鈥 of the diffuse cityscape rather than an independent settlement. Each cluster of villages had a central site with a big plaza 鈥 maybe a political or ritual centre. Close to this were the largest residential sites, with smaller residential centres further out.
In fact, the Amazonian settlements can be thought of as a mirror-image of 鈥渢raditional鈥 cities. 鈥淛ust about everything they do and use is the alter ego of what we are accustomed to seeing,鈥 says Heckenberger. 鈥淭hey farmed fish and trees, rather than wheat, barley and cattle.鈥 Whereas Eurasian farmers relied on a few domesticated species, Amazonian agriforesters used about 100. And their cities were more distributed, less centralised than ours.
Heckenberger says we could learn a lot from the way the lost Amazonian societies were organised. 鈥淭hat is probably the best way for large sedentary populations to exist in the tropical forests of the Amazon,鈥 he says. 鈥淭hey really nailed it.鈥
So far, modern society hasn鈥檛 managed the Amazon rainforest anything like as well. Since 1970, about 20 per cent of the rainforest has been cleared, mainly to make way for cattle farms. Although the rate of deforestation has slowed, that could change with the election of Jair Bolsonaro as president of Brazil in October. He has said he plans to strip powers from Brazil鈥檚 environment agencies.
Bolsonaro鈥檚 election has alarmed those studying the Amazon. Levis worries that the discovery of the forest civilisations could be deliberately misinterpreted as meaning that it is OK to chop it down and exploit it. 鈥淧eople were producing [food] in small batches and in agroforestry systems,鈥 says Levis. 鈥淲hat is happening now is that the forest is being destroyed with exotic crops like soy beans and also cattle ranching, on a very large scale.鈥 The old system seems to have been sustainable for thousands of years 鈥 the modern one is destructive and contributes to climate change.
Even just in archaeological terms, the Amazon still has a lot to offer. Iriarte estimates that three-quarters of it hasn鈥檛 been investigated. More lost cities may lay undiscovered. 鈥淭he whole north-west Amazon, the Colombian, Ecuadorean and Peruvian, is pretty much unexplored,鈥 he says. 鈥淚t鈥檚 going to give us more surprises.鈥
Who lived there?
Little is known about the earliest inhabitants of the Amazon. Genetic evidence suggests that a few males monopolised a lot of females, indicating that they lived in societies dominated by polygynous chiefs. And it shows some populations were shaped by the practice of linguistic exogamy, where males marry a female who speaks a different language.
The rest is mostly a mystery. So is the fate of these civilisations. Their population peaked about a thousand years ago, and numbers were already falling when Europeans arrived, at which point they nose-dived. We don鈥檛 know what caused that initial decline or why many of the societies changed drastically in the 500 years before Europeans arrived. Around this time, the geoglyph-building people seem to have been replaced by people building circular villages 鈥 unless they were the same people who changed their practices.
The population could have fallen because the societies reached an ecological limit. Or they could have been fighting 鈥 there seems to have been an increase in the number of defensive structures such as palisades around that time.

