Is a colonial-era drop in CO₂ tied to regrowing forests?

Did the massive scale of death in the Americas following colonial contact in the 1500s affect atmospheric CO₂ levels? That’s a question scientists have debated over the last 30 years, ever since they noticed a sharp drop in CO₂ around the year 1610 in air preserved in Antarctic ice.

That drop in atmospheric CO₂ levels is the only significant decline in recent millennia, and scientists suggested that it was caused by reforestation in the Americas, which resulted from their depopulation via pandemics unleashed by early European contact. It is so distinct that it was proposed as a candidate for the marker of the beginning of a new geological epoch—the “Anthropocene.”

But the record from that ice core, taken at Law Dome in East Antarctica, shows that CO₂ starts declining a bit late to match European contact, and it plummets over just 90 years, which is too drastic for feasible rates of vegetation regrowth. A different ice core, drilled in the West Antarctic, showed a more gradual decline starting earlier, but lacked the fine detail of the Law Dome ice.

Which one was right? Beyond the historical interest, it matters because it is a real-world, continent-scale test of reforestation’s effectiveness at removing CO₂ from the atmosphere.

In a recent study, Amy King of the British Antarctic Survey and colleagues set out to test if the Law Dome data is a true reflection of atmospheric CO₂ decline, using a new ice core drilled on the “Skytrain Ice Rise” in West Antarctica.

Precious tiny bubbles

In 2018, scientists and engineers from the British Antarctic Survey and the University of Cambridge drilled the ice core, a cylinder of ice 651 meters long by 10 centimeters in diameter (2,136 feet by 4 inches), from the surface down to the bedrock. The ice contains bubbles of air that got trapped as snow fell, forming tiny capsules of past atmospheres.

The project’s main aim was to investigate ice from the time about 125,000 years ago when the climate was about as warm as it is today. But King and colleagues realized that the younger portion of ice could shed light on the 1610 CO₂ decline.

“Given the resolution of what we could obtain with Skytrain Ice Rise, we predicted that, if the drop was real in the atmosphere as in Law Dome, we should see the drop in Skytrain, too,” said Thomas Bauska of the British Antarctic Survey, a co-author of the new study.

The ice core was cut into 80-centimeter (31-inch) lengths, put into insulated boxes, and shipped to the UK, all the while held at -20°C (-4°F) to prevent it from melting and releasing its precious cargo of air from millennia ago. “That's one thing that keeps us up at night, especially as gas people,” said Bauska.

In the UK they took a series of samples across 31 depth intervals spanning the period from 1454 to 1688 CE: “We went in and sliced and diced our ice core as much as we could,” said Bauska. They sent the samples, still refrigerated, off to Oregon State University where the CO₂ levels were measured.

The results didn’t show a sharp drop of CO₂—instead, they showed a gentler CO₂ decline of about 8 ppm over 157 years between 1516 and 1670 CE, matching the other West Antarctic ice core.

“We didn't see the drop,” said Bauska, “so we had to say, OK, is our understanding of how smooth the records are accurate?”

To test if the Skytrain ice record is too blurry to show a sharp 1610 drop, they analyzed the levels of methane in the ice. Because methane is much less soluble in water than CO₂, they were able to melt continuously along the ice core to liberate the methane and get a more detailed graph of its concentration than was possible for CO₂. If the atmospheric signal was blurred in Skytrain, it should have smoothed the methane record. But it didn’t.

“We didn't see that really smoothed out methane record,” said Bauska, “which then told us the CO₂ record couldn't have been that smoothed.”

In other words, the gentler Skytrain CO₂ signal is real, not an artifact.

Does this mean the sharp drop at 1610 in the Law Dome data is an artifact? It looks that way, but Bauska was cautious, saying, “the jury will still be out until we actually get either re-measurements of the Law Dome, or another ice core drilled with a similarly high accumulation.”

Disease, population collapse, reforestation

Scientists estimate that about 60 million people inhabited the Americas before European contact. There’s archaeological evidence for numerous cities and settlements, such as miles of now-overgrown urban sprawl that was recently mapped in Amazonian Ecuador, or the city of Cahokia in Illinois, which is estimated to have been larger than London was at that time, or Llanos de Mojos in Bolivia. The Spanish conquistador Francisco de Orellana also described seeing cities in the Amazon in 1542.

Even today in overgrown parts of the Amazon, vegetation carries the imprint of past occupation in an overabundance of cultivated species such as Brazil Nut trees.

A century after the first European contact, some 56 million people had died according to one widely cited estimate. “What we're looking at here is first contact, and [then] 100 years when 90 percent of the population, basically, dies,” said Professor Mark Maslin of University College London, who was not involved in King’s study. They succumbed to wave after wave of pandemics, as smallpox, measles, influenza, bubonic plague, malaria, diphtheria, typhus, and cholera spread through populations with no natural immunity. People who survived one disease outbreak died in the next. With too few people to work them, cities and farms were abandoned and overgrown.

King, Bauska, and colleagues used a carbon cycle model to test estimates of the scale of land abandonment to see if they could reproduce the CO₂ decline observed in the new ice core. The new data does now support large-scale abandonment: “It does seem to agree with those high-end scenarios,” said Bauska.

The smoother decline of CO₂: “actually makes much more sense in the Earth system,” said Maslin.

A lack of ground data?

The study shows reforestation is a feasible explanation for the CO₂ decline. Yet evidence for reforestation at that time is lacking in the central Amazon region, according to Professor Mark Bush of the Florida Institute of Technology, who also was not involved in the King et al study.

Bush has spent years documenting the influence Indigenous people had on Amazon forests over the last 1,000 years or so, as recorded by pollen and charcoal that settled in lake sediments. His team now has 23 “high quality” sediment records from lakes in the Amazon region, nine of which were described in a paper published in 2021.

“So far, those 23 sites, none of them have given us a rising [forest cover] after 1400,” Bush said. Instead, his data points to an earlier period of abandonment between about 1000 and 1200 CE. Its cause remains a mystery, but it was a time of conflict: “There's a lot more evidence of fighting, and there's more cracked skulls and defensive structures showing up at that time,” said Bush.

This matters because it’s hard to explain the CO₂ decline without extensive reforestation in the Amazon, because growth in the Amazon locks away much more carbon than it does in Mexico, the Andes, and other drier environments. With forestation there, “you don't change the carbon budget worth a bean,” Bush remarked.

Maslin agrees that there probably weren't that many people living in the central Amazon basin: “If you look at the empires, they are not sitting in the middle of the Amazon basin. They are the Incas, the Aztecs, those big empires which are [in areas that are] partly forested, partly savanna,” he said.

As for Francisco de Orellana’s reports of cities in the Amazon, Bush thinks he may have lied to save his skin: “they were trying to get in the good graces of the Spanish court, having just deserted,” he said. But that’s not universally accepted; the authors of the paper describing pre-Columbian urban sprawl in Amazonian Ecuador say explicitly: “Orellana did not lie.”

Scaling to net zero

The new study “gives us a real scaling of how much impact [reforestation] would have,” said Maslin.

If reproduced today, the eight parts-per-million CO₂ reduction recorded in the Skytrain ice core would only unwind about four years of CO₂ increase in our atmosphere. “It's quite sobering to think that we had an 8 ppm drop, and we potentially had population decreases in the new world of up to 90 percent,” said Bauska.

Even if the CO₂ decline is puny compared to modern emissions, the Skytrain record’s implications for reforestation are large: “incredibly important when we talk about net zero,” Maslin said. That’s because some sectors of the economy are hard to decarbonize, “and that's where the reforestation, the rewilding, maybe even direct air capture comes in—to actually get the ‘net’ of net zero,” said Maslin.

“I think that reforestation is a very valuable tool in the conservation toolkit so long as it's the right kind of tree,” said Bush, but: “the better solution is to stop producing carbon.”

Bauska agrees: “We need to preserve what we have in terms of forest and land carbon stocks, but the main thing is to curb emissions.”

Nature Communications, 2024.  DOI: 10.1038/s41467-024-45894-9