The Peopling of South America
The Peopling of South America

The Peopling of South America

While questions still outnumber answers, new findings from archaeology, genetics, and other disciplines are revealing surprising insights into the early cultures of the most recently populated continent.

Shawna Williams
Shawna Williams
Sep 1, 2020


In 2007, while searching for signs of ancient human inhabitants of the Andes mountain range, more than 4,300 meters above sea level, Kurt Rademaker came across a field littered with chunks of obsidian—some of them fashioned into tools. “There were hundreds and hundreds of them,” recalls Rademaker, then a graduate student at the University of Maine. “Adjacent to this open-air workshop, just up above it on the hillside, was a beautiful rock shelter. . . . I just had a gut feeling that it was the kind [of site] I had been looking for.” 

A decade earlier, Rademaker’s advisor, archaeologist Daniel Sandweiss, had made the unexpected discovery of flakes and tools made of obsidian, a volcanic rock, in excavations at one of the oldest archaeological sites in South America. Quebrada Jaguay, which dates to 13,000–11,000 years ago, sits along the Peruvian coast, where there are no natural deposits of obsidian. Sandweiss and his colleagues had analyzed the elements in the artifacts and found that they were likely from a source in the nearby Andes. The researchers suggested that ancient people at Quebrada Jaguay had probably split their time between the coast and the highlands. Curious about the early inhabitants of the Andes, Rademaker used the obsidian finding as a clue and began searching for early highland sites in an area with deposits of the rock. 

Rademaker struck pay dirt with his 2007 discovery: the site had clear evidence of human activity, including beads made of bone and hundreds of stone tools. Carbon dating of food residue and collagen from animal bones on the floor of the rock shelter, dubbed Cuncaicha, pinned that activity to as early as 12,500 years before present (BP), more than 1,000 years before the oldest previously dated Andean archaeological sites.

In late 2018, two studies compared human genomes from across the Americas to infer that there were multiple waves of migration from North to South America.

The finding of such early occupation of the Andes upended some assumptions about the ancient people of the Americas, says Rademaker, now a professor at Michigan State University. For example, scientists generally thought that before 11,000 or so years ago, the Andes would have been covered with glaciers or otherwise too cold for humans to inhabit. But in addition to their archaeological work on the site, analyses of the geology at Cuncaicha revealed that “ice was already well in retreat way before people arrived,” Rademaker says, and the environment would have been about as warm as it is today, temperate enough for human habitation.

Researchers had also previously thought that the low-oxygen environment of the high Andes would have forced ancient people to stick to lower altitudes after their initial migration into the continent, says Rademaker. Yet Quebrada Jaguay and other nearby sites along the coast have been dated to about the same time period as Cuncaicha. “That suggested a more rapid colonization process of a very challenging environment,” he says. 

Research on Cuncaicha is just one of the many recent inquiries into the peopling of South America that are challenging previous assumptions about how migrations, cultures, and ancestries shaped that continent’s human landscape. In addition to archaeological clues, genetic investigations of ancient and modern humans are yielding intriguing new details about when and how the continent was populated. But many fundamental questions remain unresolved, such as when people first arrived in South America and how they spread across the continent, and a coherent picture of its prehistory has yet to emerge.

A shifting archaeological picture

When anthropologist Tom Dillehay, now at Vanderbilt University, began working at a site called Monte Verde in southern Chile in 1977, most archaeologists thought the first humans moved into South America from North America about 11,000 years ago, he says. But in 1988, he and a colleague published an analysis of traces of human occupation at Monte Verde—including stone tools, mastodon bones with cut marks, and hearths with traces of burned plants—that were dated to 12,500 BP. (They also unearthed charcoal and what might have been stone tools from a layer dated to 33,000 BP, but Dillehay now says he doesn’t think the evidence is strong enough to support occupation of the site that early.) More recently, Dillehay’s group reported new excavations that unearthed human-linked materials dated to between 18,000 and 15,000 BP.

That timeline positioned Monte Verde as the oldest archaeological site in South America, says Lars Fehren-Schmitz, a biological anthropologist at the University of California, Santa Cruz, who was not involved in Dillehay’s work. The site “shows that people must have reached southern South America quite quickly after they even entered North America.” While the timing is unclear of humans’ first migrations out of a region known as Beringia—which spanned parts of present-day Siberia and Alaska—and into North America, most archaeologists put the date at no more than about 20,000 years ago. The dating of Monte Verde suggests, then, that people made it from one end of the New World to the other—about 10,000 miles—remarkably quickly. Several other sites in South America similarly point to a quick dispersal, Fehren-Schmitz notes. “We actually have a bunch of archaeological sites that are securely dated that fall into the time range of 14,000 to 12,000 BP, that tell us that people were more or less all over South America quite early after the initial peopling of the Americas in general.”

There’s wide variation in the style of artifacts found at different early sites, such as projectile points and stone tools used for cutting and scraping, indicating that groups on the continent quickly developed diverse cultures and technologies, Dillehay says. “People adapted to a very wide range of different environmental zones and habitats,” exploiting the various food sources they offered. 

Some studies point to an even earlier occupation of South America than that indicated by Dillehay’s most recent work at Monte Verde. For example, research on a site in Brazil yielded stone artifacts associated with charcoal remnants, suggesting the possibility of cooking fires, that researchers carbon dated to between 35,000 and 28,000 BP. And a site in Uruguay called Arroyo del Vizcaíno, dated to between 30,000 and 27,000 BP, yielded more than 1,000 bones of ancient animals, some with apparent cut marks. It’s not widely accepted that such remnants truly represent evidence of ancient human activity. In some cases, what are presented as artifacts likely aren’t really human-made, says Dillehay; other times, genuine artifacts might have been carried by floodwaters into layers that don’t reflect their true ages.

Eric Boëda, an archaeologist at Paris University Nanterre, says the lack of consensus that humans occupied the Americas before the period known as the Last Glacial Maximum, which extended from about 26,500 to 20,000 years ago, reflects “denialism” in the field. Boëda, who has worked at archaeological digs all over the world, including some of the Brazilian sites with purported human activity dating to more than 20,000 years ago, says artifacts that would be quickly accepted as human-made tools if they turned up in Europe or Africa are discounted when they are discovered in Brazil because of a longstanding belief that the Americas weren’t home to any people until more recently.

DIGGING DEEP: Ancient sites such as the Cuncaicha rock shelter (left) in Peru and Lapa do Santo in Brazil, where archaeologists unearthed this skull (right, top) and projectile point (right, bottom), are adding new details to the picture of South America’s prehistory.

Boëda says there’s not enough evidence yet to understand how and when people first reached South America or what subsequent migrations might have taken place to or within the continent. One possibility, he says, is that people came from East Asia through Alaska via an ice-free corridor more than 24,000 years ago, before complete glaciation made overland travel through North America impossible during the Last Glacial Maximum. (See “Where Early South American Migrants Came From” below.)

In the view of André Strauss, an archaeologist at the University of São Paulo in Brazil, such an early occupation is possible—but he says he thinks that such an extraordinary claim requires more evidence. “I don’t find any of the evidence available at this point [convincing] without a doubt,” he writes in an email to The Scientist. One gaping hole in the evidence is the lack of human remains in the Americas from before the Last Glacial Maximum, he adds—“not a single tooth!”

Indeed, the oldest human remains found in South America thus far date to just 10,000 years ago or so. But while these fossils don’t shed much light on how long the continent has been occupied, they can reveal a lot about its ancient inhabitants. For example, skeletons at the Brazilian cave Lapa do Santo are among the earliest remains found in South America, and they paint an intriguing picture of the people who called it home, says Strauss, who’s worked extensively on the site. As he and his colleagues reported in 2016, some skeletons at Lapa do Santo revealed signs of elaborate mortuary rituals involving cutting apart bodies, burning parts of them, and burying several individuals together. And while archaeologists have long envisioned early South Americans as highly mobile hunter-gatherers, he says, isotopic analyses revealed that the people at Lapa do Santo consumed food and water from a circumscribed area that he estimates to be around 1,000 square kilometers, and were thus relatively settled.

As archaeologists continue to search for more human remains to help flesh out the continent’s prehistory, researchers are also mining modern genomes, together with ancient DNA from sites such as Lapa do Santo, for data that reveal more about ancient South Americans.

Evidence of Early Human Presence in South America

Excavations of South American sites containing traces of ancient human activity have suggested that humans reached the southern region of the continent at least 14,500 years before present (BP)—remarkably quickly after first entering the Americas—and that they soon developed diverse technologies across different sites. But the picture yielded by these archaeological investigations is a patchwork, leaving open key questions, such as whether the first humans migrated south along the Pacific coast or by some other route. The history is further complicated by disputed claims (examples marked by red headers with asterisks) that certain sites reflect a much earlier occupation of the continent beginning more than 20,000 BP.



Site: Rock shelter at 4,480 meters in elevation dated to ~12,400 BP

Contains: Remains of plants and animals consumed as food and other human-made debris; human remains; stone tools

Significance: Oldest known site in high Andes; shows humans had adapted to high altitudes

Quebrada Jaguay

Site: Remains of a seasonally occupied fishing village dated to ~13,000–11,000 BP

Contains: Seafood remnants; hearths; tools made of obsidian and other types of stone

Significance: Demonstrates that people were using marine food sources and, together with Cuncaicha, that coastal people had contact with the Andean highlands

Monte Verde

Site: Settlement dated to ~18,500–15,000 BP 

Contains: Remains of plants and animals consumed as food; charcoal; wooden artifacts; stone tools

Significance: Pushed back the date of earliest known human occupation of the Americas by as much as 5,000 years

Lapa do Santo

Site: Cave with signs of human activity dated to as early as ~12,700–11,700 BP

Contains: Remains of 50 people, dating as far back as 10,600–9,700 BP, who were buried at the site; stone tools; rock art; animal remnants

Significance: Yielded ancient DNA for analysis and new insight into early cultures

Arroyo del Vizcaíno*

Site: Assembly of more than 1,000 animal bones dated to ~30,000–27,000 BP

Contains: Bones of giant sloths and other large animals, some with apparent cut marks that may indicate they were butchered by humans; purported stone tools

Criticisms include: The bones could have been carried to the site by flowing water rather than human activity; the scenario the authors propose (including human transport of large, killed animals) is not consistent with the way known hunter-gatherer groups operate.

Toca do Sítio do Meio*

Site: Rock shelter with signs of human occupation dated to ~35,000–28,000 BP

Contains: Charcoal remnants; purported stone tools

Criticisms include: Rocks resembling stone tools could have come about through natural geological processes or been made by monkeys. 

See full infographic: WEB | PDF

Clues in the genome

In late 2018, two studies compared human genomes from across the Americas to infer that there were multiple waves of migration from North to South America. One of the studies, led by geneticist David Reich of Harvard University, analyzed both ancient and modern genomes and surmised that before European settlers arrived, there were four waves of migration into South America. One wave was made up of people genetically related to the Clovis culture, distinguished by a style of tools found across much of North America, the researchers concluded. Beginning around 9,000 years ago, that lineage was partly replaced by a separate wave of migrants to South America.

The second study, led by Eske Willerslev of the University of Copenhagen, included 15 ancient genomes unearthed in North and South America. It found evidence for two waves of migration into the southern continent, one about 14,000 BP from North America, followed by a dispersal of people who had diverged from North American groups and spent time in Central America. While the results are somewhat similar to those of the paper from Reich and colleagues, Willerslev says his study didn’t find the widespread replacement of the first migrants that Reich’s did. Rather, Willerslev’s data suggested the two groups intermixed with one another.

Willerslev’s team also turned up a puzzling signal in 10,000-year-old skeletons from Brazil: between 2 percent and 6 percent of their genomes were more closely related to modern-day inhabitants of the Andaman Islands in Southeast Asia than to any other population. Willerslev points out that the ancestors of the Andaman Islanders once occupied a much broader swath of East Asia, making it plausible that some could have crossed the land bridge into North America—although that leaves open the question of why their genes haven’t appeared in any ancient or modern North American samples. As for the possibility that they could have crossed the Pacific by boat, “in many people’s minds, at least, it’s a very unrealistic scenario . . . to have done a journey like that” so early in prehistory, Willerslev says. How genetic traces of this population have shown up in Brazil and nowhere else in the Americas “is still a mystery,” he adds.

Thanks to newer technologies and approaches, archaeologists can now mine more information from sites once they’re identified.

Other recent genetic studies have homed in on what occurred in particular regions during South America’s prehistory. In a study published shortly before the Reich and Willerslev papers, for example, a team led by Anna Di Rienzo of the University of Chicago compared the whole-genome sequences of ancient human remains found in the Andes with the sequences of modern people living in the Andes and nearby lowlands. While that study didn’t indicate when people might first have begun settling the Andes, it did reveal that the highland and lowland populations began to diverge around 9,000 years ago, says coauthor John Lindo, an anthropologist now at Emory University. “That’s pretty early to start [having established populations] at 9,000 years, I think,” he says. “That’s pretty amazing.” 

Fehren-Schmitz agrees, noting that in contrast to Eurasia, which experienced multiple mass migrations of humans, South American prehistory overall shows a pattern of populations staying relatively stationary after the initial peopling. In the case of the Andes and nearby coastal regions, he says, “the [west] coast is actually only a very narrow stretch of land and [then] the Andes directly start, so we’re talking about distances between populations of a day or two days of travel—like fifty to one hundred miles. And still we see patterns of genetic distinctness between these groups, which is quite stunning.” 

Based on current evidence, says Tábita Hünemeier, a population and evolutionary geneticist at the University of São Paulo in Brazil who conducts genetic studies of indigenous Brazilian groups, present-day Andeans seem to have descended from the wave of migrants that replaced the Clovis-related population, together with a more recent wave that occurred around 4,000 BP. After the highland and lowland populations divided around 9,000 years ago, the highland population further split into northern and southern populations around 5,800 BP, a study led by Reich and Fehren-Schmitz recently suggested.

Less is known about the dynamics of ancient Amazonian groups, says Hünemeier, but genetic evidence points to an initial arrival of people in the first large migration event. That population appears to have been replaced by a second wave of migration into the continent, and linguistic evidence indicates these people quickly differentiated into different groups beginning around 9,000 years ago, leading to a panoply of languages, more than 100 of which still exist today. The current goal of her work, Hünemeier says, is “to figure out how [people in the second group] settled Amazonia and how they split,” as well as “which was the original group that came to Amazonia.” 

In a study published earlier this year, she and her colleagues found modern genetic traces of a population known as the Tupí that dominated the Brazilian coast during the 15th century but had supposedly been driven extinct by European conquerors. Modern people known as the Tupiniquim, who live in cities and do not speak an indigenous language, have long identified themselves as descendants of the Tupí, and the researchers found that they do indeed have signals of indigenous ancestry, along with European and African heritage, in their genomes. Furthermore, their indigenous ancestry was distinct from that of any other modern groups who have been studied. 

The team’s analysis points to a migration of Tupí people from central Amazonia northeast toward Brazil’s coast around 2,000–3,000 years ago, followed by a later migration of a separate Tupí group who traveled southeast but also ultimately expanded along the coast. They were not the first people to reach the Brazilian coast; archaeological sites with distinctive shell mounds known as sambaquis indicate that others had settled the coast 10,000–8,000 years ago, but they appear to have been completely replaced by the Tupí. “We don’t know where [the people who made the shell mounds] came from . . . and which language they used to speak,” says Hünemeier, who is now working with ancient DNA to try to learn more about the prehistoric coastal group. 

Genetic Insights About the First South Americans

Two studies published in late 2018, one led by David Reich of Harvard University (results depicted in cool colors) and the other by Eske Willerslev of the University of Copenhagen (results in warm colors), compared ancient and modern genomes from across the Americas to infer that there were multiple waves of migration from the northern continent to the southern one.

See full infographic: WEB | PDF

Looking forward, looking back

Researchers’ understanding of South America’s earliest inhabitants would benefit from the discovery of more human remains, researchers say, but such finds have been hard to come by—which itself raises questions, Dillehay says, of what happened to the bodies of ancient people. “One major thing is to understand the mortuary practices [of ancient people in the Americas] a little better,” he says.

With or without remains, more archaeological sites in parts of Central America and the northwest coast of South America are needed to reveal how people migrated onto and within the continent, and how many waves of migration there were, Boëda notes. The oldest widely accepted archaeological sites in Central America date to around 11,500 BP, and the oldest human remains in the area to about 6,000 BP. The region’s acidic soil inhibits preservation of human remains, and some archaeologists suspect that rising sea levels at the end of the Last Glacial Maximum may have swallowed some of the earliest signs of human activity. 

Thanks to newer technologies and approaches, archaeologists can now extract more information from sites once they’re identified. These methods include genome analysis and stable isotope analysis, which allows researchers to determine whether an individual’s early diet was made up of plants and animals from the local area or from further afield. This in turn can reveal whether a person was from the area where their remains were found or had migrated there. With these new methods, says Rademaker, “each [new] site is going to tell us something incredibly useful.” 

Where Early South American Migrants Came From

The widely, though not universally, held thinking among researchers in the field is that indigenous Americans came from East Asia at a time when sea levels would have been low enough to form an inhabitable region known as Beringia that spanned what’s now eastern Siberia and western Alaska. According to the Beringian standstill hypothesis, glaciers long prevented movement further into the Americas, so the founding population remained in the region for thousands of years. When the ice melted enough to permit passage, the Beringians moved fast. 

“I think we have pretty compelling evidence from the Y DNA studies and the mtDNA [mitochondrial DNA] studies that roughly between 15,400 or so years ago and maybe 14,300 . . . we have an expansion of lineages,” says Ben Potter, an archaeologist at Liaocheng University in China. “We have a star-like radiation” of groups that moved into different areas of North America. A caveat, he says, is that no human remains from this period have been found, and there is thus no fossil evidence to show where exactly the earlier initial expansion originated or where the genetic isolation took place.

Some studies have challenged the majority view that that expansion represented humans’ first foray into the Americas. In a 2017 Nature paper, for example, paleontologist Kathleen Holen, then at the San Diego Natural History Museum, and colleagues reported finding 130,000-year-old mastodon bones at a site in California that appeared to have been smashed with human-made stone tools. The study was met with skepticism by researchers who said the bones could instead have been broken by other means, such as modern construction equipment. 

In July of this year, an international team led by Eske Willerslev of the University of Copenhagen garnered headlines with another report in Nature, this one of purportedly human-made tools dating to 33,000 BP in a cave in Mexico. But critics questioned whether the stones in question had truly been shaped by humans. 

Another disputed idea, known as the Solutrean hypothesis, is based on similarities between tools made by North America’s early Clovis culture and those made by Europe’s Solutrean people between 22,000 and 17,000 BP. Its main proponents, Bruce Bradley of the University of Exeter and the late Dennis Stanford of the Smithsonian National Museum of Natural History, suggested that some Solutreans could have migrated to North America via boats that hugged an ice bridge between the two continents (World Archaeol, 36:459–78, 2004). Other researchers have pointed out what they see as multiple problems with the hypothesis, including the 7,000-year gap between Solutreans’ use of the tools in Europe and the first Clovis site in North America, and a lack of genetic evidence that any early Americans had European ancestry.

The Beringian standstill hypothesis continues to undergo revisions as new studies emerge. In 2018, for example, Potter and his colleagues reported a genetic analysis of 11,500-year-old remains of a baby girl found buried in Alaska at a site called Upward Sun. The infant belonged to a previously unknown lineage that split from the ancestors of today’s Native Americans between 22,000 and 18,100 years ago, the researchers suggested. Potter notes that the potential for interbreeding between known populations and extinct lineages such as this can complicate genetic models that estimate how long ago populations diverged, and he says he thinks future discoveries like that made at Upward Sun are likely. “There probably are more of those out there that we just haven’t detected yet,” he says. 

Clarification (September 4): The language in the infographic has been updated to reflect the fact that Cuncaicha is the oldest known site in the high Andes, not the Andean region.

Correction (September 8, 2020): The language of the article has been changed to reflect the fact that Monte Verde is in southern Chile, but not near the country's southern tip. The first map has also been updated to correctly show Monte Verde's location. The Scientist regrets the errors.