The Russell Glacier in Greenland. After ancient glaciers scoured away soil and reached bedrock, they may have grown thicker—triggering a global cooling that shifted the ice age cycles.

Jason Edwards/National Geographic/Getty Images

Seafloor cores suggest sticky, thick glaciers caused mysterious shift in ice age rhythms

BOSTON—About 1 million years ago, one of Earth's most important metronomes mysteriously shifted: Ice ages went from occurring every 40,000 years to every 100,000 years. At the same time, the "conveyor belt" of warming currents in the North Atlantic Ocean slowed sharply. Last week, scientists here at the Goldschmidt Conference presented a clue to these twin mysteries: evidence that glaciers in the Northern Hemisphere suddenly began to stick to their beds. Growing thicker, they might have triggered a cooling that disrupted the conveyor belt and allowed the 100,000-year cycle that we see today to take root.

"The system basically crashed," says Steve Goldstein, an ocean geochemist at Columbia University who led the study. Other scientists welcome the new clues to the transition. "This is really exciting new evidence," says Henrieka Detlef, a paleoclimatologist at Cardiff University in the United Kingdom. But she and others aren't sold yet on the long causal chain that Goldstein's team posits.

Scientists have long known that tiny changes in Earth's orbit around the sun, called Milankovitch cycles, drive the planet in and out of ice ages. But nothing changed in those orbital patterns 1 million years ago. Recently, Goldstein and his colleagues found signs of a possible contributor to the ice age transition: a near-collapse of the Atlantic meridional overturning circulation (AMOC). The AMOC shepherds shallow warm water to the North Atlantic, where it cools and sinks before returning south along the sea floor to the Southern Ocean to meet Pacific Ocean waters.

Goldstein's group deduces the overall strength of the AMOC from geochemical markers in ocean sediment cores. The researchers take advantage of a ratio between two isotopes of neodymium that varies with the age of their source rocks: ancient crust runs negative, whereas younger rocks are more positive. As it happens, the North Atlantic is surrounded by ancient crust, whereas the Pacific, thanks to its volcanic Ring of Fire, tilts younger. The neodymium-carrying grit ends up incorporated into the shells of single-celled foraminifera or fish teeth, both of which accumulate over time on the sea floor. Changes in the isotope ratio record the wax and wane of intruding North Atlantic or Pacific waters.

Earlier this decade, the Columbia group tested its approach on two archived sediment cores from the South Atlantic. About 950,000 years ago, they saw the isotopic signals shoot up, reflecting an incursion of Pacific waters, with little evidence of returning North Atlantic waters—suggesting a stark "AMOC crisis." The slowdown could have sharply cooled the North Atlantic region—and might have lengthened the ice age rhythm.

Now, the team has analyzed five other ocean cores that also show signs of a weak AMOC. Two of the cores, from the North Atlantic, suggest a possible trigger for the AMOC crisis. In the millennia leading up to it, the neodymium signal sharply trended negative before abating—a sign that an influx of older and older grit from the North Atlantic region had suddenly stopped.

The only plausible explanation, they say, is a long-standing hypothesis advanced by Peter Clark, a glaciologist at Oregon State University in Corvallis, and several others: that the northern ice sheets had finally ground their way to bedrock. Before Earth's current ice age cycles began 3 million years ago, a long warm period had allowed a thick soil layer to build up on northern landmasses. At first, the soil acted as a grease that caused early ice sheets to collapse before they could thicken much. But repeated glaciations gradually scoured this grit away, and meltwater swept it into the ocean. As the glaciers dug deeper into older rock, the neodymium signal in ocean sediment became more negative. Eventually, the glaciers reached bedrock and began to stick to their base, allowing them to grow thicker—leading to a more profound and persistent cooling that somehow caused the AMOC to crash and the glacial cycle to lengthen. "We think we're seeing the trigger," says Maayan Yehudai, the Columbia graduate student who presented the work. (Scientists believe pronounced global warming—like the warming underway now—could also disrupt the AMOC.)

A slowing pulse

Earth's orbital patterns cannot explain a shift in the ice ages' frequency that began about 1 million years ago. Thickening glaciers and crippled ocean currents could be to blame.

2.5 3 3.5 4 4.5 5 5.5 Age (million years ago) 1.6 1.4 0.2 present 0.4 0.6 0.8 1 1.2 40,000-yearcycles 100,000-yearcycles Mid-Pleistocenetransition Temperature proxy

The neodymium evidence supports this geological story, Clark says. "It's a pretty clear signal that you should see." Detlef notes, however, that there is no conclusive evidence that northern ice sheets were increasing in thickness prior to the AMOC slowdown. But she accepts that something important happened in the North Atlantic leading up to the AMOC crisis.

One hypothesis that does seem ruled out, however, is the notion that the growth of Antarctic ice sheets 900,000 years ago played a pivotal role in the tempo change. "Everything that's happening in the North Atlantic is happening before [that]," Yehudai says.

The AMOC and the glaciers may not have been the only factors in the transition, however. Some scientists have suggested that a small drawdown of carbon dioxide (CO2), perhaps driven by a dust-fertilized plankton bloom in the Southern Ocean, would have been enough to shift the ice age rhythm. Yair Rosenthal, a paleo-oceanographer at Rutgers University in New Brunswick, New Jersey, thinks a CO2 drop, thickening ice sheets, and a weak AMOC could have all played a role. "I'm not a fan of single triggers of anything."