91av

IN THE high desert south-east of Puebla, Mexico, lies a sloping apron of pockmarked volcanic rock. It’s a silent and sun-drenched landscape, close enough to Mexico City for the volcano Popocatépetl to be visible in the distance. The site – an abandoned quarry – doesn’t look like a battlefield, but that’s what it has become. If a team of geoarchaeologists from Liverpool John Moores University in the UK are right, about 160 of the pockmarks on the quarry floor are human footprints – footprints some 30,000 years older than they ought to be.

The footprints are the new front line of a long-running battle over the peopling of the Americas, the last great land mass to be occupied by humans. Until recently most archaeologists believed the first arrivals – and the ancestors of today’s Native Americans – were big-game hunters from north-east Asia who crossed the Bering land bridge linking Siberia and Alaska about 13,400 years ago, as the last ice age waned. Within a few hundred years these “Clovis” people – named after the site in New Mexico where their distinctive stone tools were discovered – had spread right across the continent.

In the past decade or so numerous researchers have made discoveries that challenge the Clovis First theory, and archaeologists now agree that there is clear evidence of human occupation at least 1000 years before the Clovis hunters arrived. But to claim that humans were roaming the Americas 30,000 years earlier still is a different matter altogether.

The Liverpool team, led by Silvia Gonzalez, discovered the prints in the summer of 2003 while on a field trip to clarify the dating of the surrounding area, known as the Valsequillo Basin. Since the 1960s, the basin has been a rich source of archaeological finds, including stone tools alongside the bones of mammoths, mastodons and camels. Some archaeologists believe that those artefacts are good evidence of a human presence in North America long before the Clovis occupation. But difficulties with the basin’s precise chronology have long bedevilled researchers.

In the 1960s, for example, archaeologists discovered a stone scraper associated with mollusc shells carbon dated to 22,000 years ago. That link, however, remains controversial. Similar deposits recently produced radiocarbon ages from 10,000 to 40,000 years, but other studies have yielded ages ranging from 200,000 to 600,000 years – only adding to the confusion, as this was before modern humans even evolved.

The footprints could prove decisive. Imprinted into a layer of volcanic rock called the xalnene ash (pronounced salnene), they must have been made shortly after a nearby volcano called Cerro Toluquilla erupted, before the ash hardened into rock. Gonzalez, who has studied ancient human footprints elsewhere, was the first to spot them. “It felt like a thunderbolt in my mind,” she says.

The thunderbolt flashed because she knew that the xalnene ash lay under sediments that were 22,000 years old. That would mean that the people who made them predated the Clovis culture by 10,000 years, maybe more. And that, as Gonzalez’s colleague Matthew Bennett realised once he too concluded the footprints were real, “would require a complete rethink of the timing, route and origin of the colonisation of the Americas”.

Gonzalez’s team announced their findings in July 2005 at the Royal Society Summer Science Exhibition in London. By then they had identified 269 prints, 164 of which were human, the rest from animals. From the size of the footprints they calculated that the largest was an adult a strapping 190 centimetres tall; the smallest a child of just 117 centimetres. Some of the prints formed curving trails tracing the shoreline of an ancient lake, the shifting sediments of which had preserved them. The most stunning thing about the prints, though, was their age. They were an astonishing 40,000 years old.

The researchers should have been elated, but their mood was tense. Gonzalez knew full well that they were entering controversial territory. A handful of previous studies have yielded comparable dates, including excavations at Topper, South Carolina, and Pedra Furada, Brazil, but all have encountered intense criticism. “We’re in for a fight,” Gonzalez predicted. She was right.

Footprints, or not?

First of all, they knew the authenticity of the footprints would be challenged. Human prints are rare and hard to validate and are generally only accepted if they conform to a rigorous set of criteria developed after Mary Leakey’s 1978 discovery of 3.6-million-year-old hominin footprints in volcanic ash near Laetoli, Tanzania (for details, see Diagram).

To perform the tests, Bennett took high-resolution laser scans of the prints and transformed them into contour maps. “You can see what it really looks like, not what your eye wants it to look like,” he explains. When they ran the Leakey test on the scans, many matched the established criteria. Convinced that the footprints were real, the team’s next challenge was to date them.

Gonzalez sent samples of the footprint layer and overlying sediments to various specialist dating labs, including one at the University of Oxford, to be analysed using a suite of high-tech methods. Two key findings pinned down the dates. At Oxford, Tom Higham used a technique called accelerator mass spectrometry, a highly sensitive form of radiocarbon dating, to find out the age of mollusc shells from sediments above the ash. His results: 38,000 years, give or take 500.

Jean-Luc Schwenninger, also at Oxford, dated the xalnene layer itself. He used optically stimulated luminescence (OSL) dating to analyse the orange-coloured grains of baked lake sediments that lace the ash. OSL measures the radiation a crystal has absorbed since it was last heated or exposed to light. His date of 38,000 years plus or minus 8600 bracketed Higham’s results. Call the footprints 40,000 years old and you shouldn’t be far off (Quaternary Science Reviews, vol 25, p 201).

The implications are huge. If people were living in the Americas 40,000 years ago, they couldn’t have come from Siberia – it was the height of the last ice age and the Bering land bridge was icebound and impassable. It’s anyone’s guess who they were, where they hailed from, or how they got here.

“If people were here 40,000 years ago they weren’t from Siberia”

“Surely there were several migrations,” says Gonzalez. She thinks groups of boat-builders spread up the Pacific coast from southern Asia, exploiting coastal resources and leapfrogging from shelter to shelter all the way to North America. If that sounds unlikely, then consider the fact that people from south-east Asia crossed hundreds of kilometres of ocean to colonise Australia around 50,000 years ago.

If Gonzalez is right, the first Americans were not Siberians but people who may have looked like south Asians or Australian Aborigines. Gonzalez even thinks she has found a descendant of those pioneers, in the form of a delicately featured skull found near Mexico City in 1959. Peñon Woman lived 13,000 years ago, according to Oxford’s radiocarbon lab, making hers the oldest human skull in the Americas.

A new era

Peñon Woman has a long, narrow skull – unlike both ancient and modern Native Americans, whose skulls are typically short and broad (unfortunately there are no known Clovis skulls to make a comparison with). Nor is she alone. Kennewick man, a 9300-year-old skeleton from Washington state, also has a long narrow skull. What is more, “in the National Museum of Anthropology in Mexico City, there’s material from 10,000 bodies from the Pleistocene to the 18th century,” says Gonzalez. She is now performing craniometric and genetic studies on the 27 oldest. Three are more than 10,000 years old, and share Peñon Woman’s skull shape. Gonzalez hopes to trace the genetic roots of the earliest inhabitants of Mexico, which she views as a key crossroads in the peopling of the Americas.

Gonzalez is not ready to publish her team’s genetic results, but she sees evidence mounting on the side of humans coming to the Americas from somewhere other than north-western Asia, and long before Clovis. “It’s very difficult to ignore,” she says. “A new era is coming.”

A few researchers, including dating specialist Guaciara dos Santos of the University of California, Irvine, support a very early human presence in the Americas, and welcomed the team’s findings. Many others were politely sceptical, pointing out that such a paradigm-breaking discovery requires independent validation. Some were downright dismissive. One of those was Paul Renne, director of the University of California’s Berkeley Geochronology Center, who had checked out the footprints while working at a nearby, potentially pre-Clovis site called Hueyatlaco. “I’ve seen them up close and personal,” he told Nature in July 2005, “and I don’t think they are footprints.” In December he followed that warning shot with a seemingly devastating broadside, also published by Nature (vol 438, p E7).

Renne used argon-argon dating on the volcanic lapilli – millimetre-sized black beads – that comprise much of the ash. By bombarding them with neutrons, he could calculate dates from the ratio of two argon isotopes. A member of Gonzalez’s team, Simon Kelly of the Open University in Milton Keynes, UK, had tried the same technique in 2004. He got such a wide range of ages for the beads, from 1.3 to 4.6 million years, that he concluded the results were unreliable. Renne, however, had no such problems. The nine beads he studied all yielded the same age, 1.3 million years.

What’s more, Renne found another way to date the ash. As volcanic rocks solidify after an eruption, they may record the direction of Earth’s magnetic field, stored in grains of iron oxide. When Renne’s graduate student Joshua Feinberg measured the polarity of the beads, he found that it pointed in the opposite direction to today’s magnetic field. Earth’s most recent magnetic flip-flop occurred 790,000 years ago, with another a million years earlier, which fits with the argon-argon date of 1.3 million years.

If Renne’s dates are right, Gonzalez’s claim is in trouble. The idea of people roaming the New World 40,000 years ago might be believable – but 1.3 million years isn’t. Modern humans only appeared around 100,000 years ago. True, Homo ergaster thrived in what is now the republic of Georgia and Homo erectus occupied much of Asia as early as 1.8 million years ago. While it’s not absolutely impossible that those ancient, small-brained humans made it to the Americas, there’s absolutely no evidence that they did.

Renne believes this leaves two possibilities. Either the footprints are the sign of “shockingly” old hominins or they’re not footprints. He and co-researcher Mike Waters, director of the Center for the Study of the First Americans at Texas A&M University in College Station, believe the prints are simply eroded quarrying marks. “We were at the site one day and two locals came down and said, “Hey! We made those marks,” Waters says. “What [Gonzalez] needs to do is to peel back a layer of quarry and see if the footprints continue.”

Game over? Not according to Gonzalez and her colleagues. They have studied and scanned the entire site in minute detail, and scoff at the idea that they can’t distinguish between quarrying marks, which are fresh and sharp-edged, eroded patches, and the footprints. They are equally sure about their dates.

At the footprint site, Gonzalez hefts a chunk of ash. It’s clear that the rock is far from uniform. While it is mostly made up of the black beads Renne dated, the rock is also laced with white, orange and yellow fragments. It looks as if buckets of multicoloured beach sand got stirred into the ash. That’s not so far from the truth, says Gonzalez. Her team is convinced that the eruption that spewed the xalnene ash started deep beneath the lake. When magma met the saturated rock and sediment under the lake, the detonation pulverised layer after layer of rock, including earlier volcanic deposits. “The explosion that produced the xalnene ash was like an atomic bomb,” Gonzalez says. “When it blew, it brought all of the particles with it, particles that are very different geochemically. Renne dated the black bits and got 1.3 million. We dated the orange bits and got 40,000 years. The youngest particles are going to be the age of the volcano.”

Gonzalez can also explain Feinberg’s geomagnetic results. Even if they prove to be correct, she points out, Earth’s magnetic field reversed itself quite recently, although only for about 1500 years. That reversal, known as the Laschamp Excursion, occurred 40,000 years ago – exactly when they believe Cerro Toluquilla erupted.

Renne is not impressed. Certainly the ash contains some “extraneous rock fragments”, or xenoliths, he says, but these are “a red herring” when it comes to dating. He’s sure that the black beads are basalt from the eruption itself. “If what I am dating are xenoliths,” he says, “I would like to know where the primary magmatic material is.” As for the Laschamp Excursion, Renne admits it could explain his geomagnetic results, but says it amounts to special pleading.

Gonzalez does not dispute that the beads Renne dated are primary magmatic material. However, she says, the eruption mixed older materials into the melt, producing a hybrid magma. That may explain Kelly’s wide range of dates, and Renne’s as well.

Renne is the most vocal critic of the footprint research, but he is not the only one. Many specialists, including Tim Jull at the University of Arizona in Tucson and Andrew Calvert of the US Geological Survey in Menlo Park, California, are wary of the dates produced by OSL. “OSL has had mixed success so far,” says Jull. Renne’s work, he says, “calls the entire OSL dating into question”. That’s the kind of comment that frustrates Bennett. After studying the geology of the basin for three years, he’s convinced that it is Renne’s date that is questionable; it just doesn’t fit with what they find in the field.

A few kilometres from Cerro Toluquilla, Bennett stands in the broiling sun pointing at a dark smudge layered into a bank of chalky rock. He says it’s a crucial new finding. The smudge, he explains, is where the xalnene deposit ends. “We can trace this ash layer all the way back to Cerro Toluquilla,” he says. The chalky rocks above and below it are sediments from the ancient lake. “Here we have a continuous record of sedimentation below, through, and beyond the xalnene.”

They have dated the sediment layers above the ash from 9000 down to 38,000 years. If Renne is right, more than a million years of sediments need to be accounted for. Despite three years of geological work in the basin, Gonzalez and her colleagues have found no sign of those missing rocks. And here, at the edge of the ash layer, there’s a seamless transition right through the volcanic episode.

That’s supported by new OSL dating of layers beneath the ash. According to Gonzalez, those sediments are more than 38,000 years old, but far less than 1.3 million. “Not to put it too bluntly, Bennett says, “it blows the million years out of the water.”

The journey has just begun

It’s late afternoon at the footprint site. Gonzalez, Bennett and their colleague David Huddart are packing their gear. They’ll return to their labs soon, taking with them new laser scans detailing the entire site, plus dozens of cores of xalnene ash and the sediments surrounding it. They will use them for dating, for geomagnetic studies, and to reconstruct the basin’s ancient environment.

As the sun sinks, the site is transformed. In the slanting light, the scattered footprints and curving trails stand out. Suddenly it’s easy to picture figures emerging from the shadows – a tall young man, a narrow-faced woman, a playful child skirting the lakeshore, their feet sinking into the moist, gritty ash from the still-smouldering volcano.

“As the sun goes down, it’s easy to imagine human figures emerging from the shadows”

Asked how the controversies over the peopling of the Americas can ever be resolved, dos Santos quotes archaeologist Niède Guidon, director of Brazil’s Serra da Capivara National Park: “Let’s talk less and dig more.”

That’s exactly what Gonzalez and her colleagues plan to do. “We need to date the whole basin, using different types of dating methods according to what we find there,” she says. “We need to make sure that everything makes sense.” And they have identified several sites where they expect to uncover undisturbed footprints.

Between working out a coherent chronology for the Valsequillo Basin, decoding DNA from 10,000-year-old bones and returning to dig for pristine footprints, Gonzalez has embarked on a long and daunting journey. “I know that this will be for the rest of my life,” she says. But if it’s a journey that solves one of the biggest mysteries in human evolution, it will have been one worth making.