When crocodiles roamed the Arctic

New Scientist – 18 June 2008

WHEN Ernest Shackleton and his men marched towards the South Pole in December 1908, they came across something entirely unexpected. After scaling the vast Beardmore glacier on the edge of the polar plateau, they found seams of coal amid the snow and ice. They also found impressions of leaves in sandstone boulders nearby and even fossilised wood from a coniferous tree.

The conclusion was extraordinary but inescapable: Antarctica was once warm and forested, conditions that could hardly be more different to the far-below-freezing midsummer weather that forced Shackleton’s team to turn back before reaching the pole. How was this possible?

Four years later, Alfred Wegener put forward his theory of continental drift which, it was later realised, could explain the balmy climate: Antarctica had been warmer because it was once much closer to the equator. Even today, some schoolchildren are taught that continental drift accounts for all the evidence for a warmer Antarctica.

However, the fossil trees Shackleton’s team discovered grew around 250 million years ago, when Antarctica was barely closer to the equator than it is today. What’s more, the continent reached its current position roughly 100 million years ago, and an ever-growing list of fossil finds date from 100 to 40 million years ago. During this time, when dinosaurs roamed the almost subtropical forests of an ice-free Antarctic, conditions on the other side of the planet were even more remarkable: the Arctic Ocean was a gigantic freshwater lake infested with crocodile-like reptiles.

As the modern world warms, there has been a surge of interest in this “hothouse” period. What sustained such high temperatures for tens of millions of years? If the poles were so warm, what were the tropics like? Recent findings provide a fascinating insight into our past – and perhaps also a glimpse of our future.

Icehouse phase

The Earth’s climate is currently in an “icehouse” phase: the polar ice sheets are not as extensive as they were during ice ages, when the sea level fell by as much as 120 metres, but some ice has remained even between ice ages. Before about 34 million years ago, though, the planet went through a prolonged hothouse phase with no ice at all. Sea level was more than 70 metres higher than today, covering vast swathes of what is now dry land. For instance, an inland ocean divided North America in two. This period lasted from the middle of the Cretaceous era until well into the Eocene – about 100 million to 50 million years ago.

One of the earliest signs that the poles were ice-free and warm 100 million years ago was the discovery at the turn of the 20th century of fossil breadfruit trees from the Cretaceous in Greenland; today such trees are at home in places like Hawaii. Since then, even more extraordinary finds have been made.

The most evocative image of a warm Arctic has emerged from the work of John Tarduno of the University of Rochester, New York. For more than a decade, Tarduno has been hunting for fossils on Axel Heiberg Island in the Canadian Arctic, just west of Greenland. The island was already well within the Arctic Circle 90 million years ago.

His team has found bones and even partial skeletons of a crocodile-like creature called a champsosaur from this period. The champsosaur was a fish-eating reptile up to 2.4 metres long that probably looked much like the gharials of India. Because these reptiles would have relied on their environment to stay warm, conditions in the far north must have been far hotter than today. “These fossils speak volumes,” says palaeoclimatologist Paul Wilson of the University of Southampton in the UK.

Warmth-loving reptile

Last year, Tarduno’s team reported that most champsosaur remains are of juveniles, meaning the animals not only lived but bred in the Arctic. As hatchlings and juveniles could not have survived if winter temperatures came anywhere close to freezing, this means it was not only warm, but warm all year round.

Modern crocodiles are found no further north than the lower Yangtze and North Carolina. If the champsosaurs’ temperature requirements were similar, the Axel Heiberg locality must have had mean annual temperatures of at least 14 °C, and the average temperature during the coldest month could not have fallen below 5.5 °C. The region would not even have had ice in winter.

The champsosaur was not the only warmth-loving reptile to live inside the Arctic Circle. Tarduno’s team has found an abundance of fossils of four kinds of turtles at Axel Heiberg Island, again pointing to a mean annual temperature of at least 14 °C.

Most recently, the team has found fossils of a family of turtles called Macrobaenidae on Axel Heiberg Island (the details have yet to be published). These turtles originally lived in Asia, but from the late Cretaceous onwards appeared in North America too. Because turtles are very sensitive to climate, the researchers think they could have survived the migration only if they moved along a route in the far north that was warm all year round. More significantly, these turtles – like the champsosaurs – were freshwater creatures. “They would have required a non-marine connection,” says team member Donald Brinkman of the Royal Tyrrell Museum in Drumheller, Alberta, Canada. “If the Arctic was a big freshwater lake, that would have made it possible.”

Biggest lake

Fresh water in the Arctic Ocean? As far-fetched as it seems, there is now strong evidence that as recently as 50 million years ago, at the start of the middle Eocene, at least the surface of the Arctic Ocean was fresh. This picture has emerged only recently because it is extremely hard to access the records of the ocean’s history, says Kathryn Moran of the University of Rhode Island in Kingston, a member of a 2004 expedition to drill sediment cores from the Arctic seabed.

Drill ships have to stay exactly above their chosen site to prevent the drill from snapping, yet in the Arctic drifting chunks of sea ice up to several kilometres wide make normal drilling operations impossible. “They can easily knock a ship off location,” says Moran. “So what we had to do was break that ice.” The task fell to two icebreakers. “The ships are really big and powerful, and they basically had to learn how to dance together,” says Moran.

Dance they did, and in 2004 the team collected a core of sediment that had been deposited over tens of millions of years on the Lomonosov ridge, just 250 kilometres from the North Pole. One study of the core revealed that a freshwater fern called Azolla grew abundantly in the Arctic Ocean for 800,000 years about 50 million years ago (Nature, vol 441, p 606). At the time the Arctic Ocean was largely isolated from other oceans, and fresh water from rivers would have floated on top of denser salt water. “It might have been, at least in the surface waters, one of the biggest lakes on the planet,” says Moran.

Surprisingly warm

The waters of this mega-lake were a surprisingly warm 10 °C, but that’s nothing to the temperatures reached a few million years earlier during the hottest part of the Eocene, when the ocean was salty. According to another study of the core the surface water 55 million years ago was around 18 °C, peaking at an incredible 23 °C – more than warm enough for a pleasant swim at the North Pole!

What about the Antarctic? Here too gathering evidence is far from easy. Ice cores from Antarctica’s kilometres-thick ice sheets are no help, for even the oldest ice is a mere million years old. It’s the land beneath the ice that holds the secrets. “We don’t want the Antarctic ice sheet to disappear, for there is 67 metres of sea level stored there, but gosh, it would be lovely, from a palaeoclimate perspective, to know what’s under all that ice,” says Wilson. “In particular, because Antarctica has certainly been in a polar position back through the Cretaceous.”

Fossil hunters on the mainland are limited to a few exposed sites. But on the Antarctic Peninsula, a finger of land that juts north towards South America, enough rock is exposed to give explorers a glimpse not just of Antarctica’s ancient flora and fauna, but of the nature of the seas around it.

About 150 to 100 million years ago, the peninsula was a mountain range similar to the Andes, and its rivers drained into a massive basin, now called the James Ross basin. Over millions of years, the basin filled up with sediment and later the rocks it formed were uplifted. Today these rocks lie exposed on islands off the Antarctic Peninsula and contain a treasure trove of fossils from the Cretaceous, including silvery slivers of shells of ocean-dwelling ammonites and gastropods. In the late Antarctic summer, these fragments glint as they catch the sun which barely rises above the horizon. “It looks like the surface is covered in jewels,” says palaeoclimatologist Jane Francis of the University of Leeds, UK, a veteran of 12 expeditions to the poles.

Ferns and cycads

Besides ammonites and gastropods, Francis and her colleagues have found abundant fossils of sea urchins and lobsters that lived on the sea floor, shark teeth, and even massive marine reptiles with rib bones about half a metre long. Oxygen isotopes in the shell fragments show that the waters around Antarctica 100 million years ago were a balmy 15 °C, compared with -2 to 0 °C today.

Dinosaur bones, which must have been washed down off the peninsula into the sea, have also been found in the marine sediments (see “Dinosaurs at the poles”). Plant fossils unearthed by Francis and her students show that 100 million years ago the peninsula was lush with ferns and cycads, along with conifers resembling the monkey puzzle tree. Analysis of the shape and size of fossil leaves has led Francis to conclude that the peninsula was very warm during the mid-Cretaceous, with a mean annual temperature of about 17 to 19 °C, similar to that of South Africa today. “That’s almost sub-tropical,” says Francis.

Growth rings in one fossil tree trunk suggest trees thrived despite complete darkness in mid-winter. “In tree-ring terms, the tree was very happy, it wasn’t growing in any kind of stress, there’s no sign of frost rings and there’s no sign of drought,” Francis says.

Her team has also found fossil flowers dating back to about 85 million years ago. These include flowers resembling those of Siparunaceae, tropical vines found in the Amazon, as well as those of the Australian eucalyptus and Winteraceae trees such as the Tasmanian mountain pepper.

Sweltering greenhouse

It’s abundantly clear that both the Arctic and the Antarctic were ice-free and warm from about 100 million to 40 million years ago. But until a decade ago, climate scientists struggled to explain how the Earth could have become so warm at the poles. Their models suggested it could only have happened if levels of carbon dioxide in the atmosphere were very high – turning the Earth into a sweltering greenhouse – but this would also have made the tropics extremely hot. Isotope ratios in marine shells, however, suggested that tropical waters were not much hotter than they are today.

As it turns out, the models were right and the shell studies were flawed. Recent and more careful studies by Wilson and colleagues (Geology, vol 30, p 299) suggest that tropical seas were indeed hotter during the hothouse phase, with the surface waters being as warm as 34 °C compared with 29 °C today, says Raymond Pierrehumbert of the University of Chicago, a climate researcher and contributor to the RealClimate blog.

Despite this advance, climate modellers face a new problem. While pumping up atmospheric levels of CO₂ in the models creates ice-free poles and warmer tropical waters, the land in the tropics becomes unbearably hot. “The temperatures are so high that unless land plants behave differently from modern types, you would be beyond their temperature tolerance,” says Pierrehumbert. “We are talking of temperatures on land of an average of 40 °C, and with seasonal fluctuations they might even go up to 50 °C. It would kill off just about anything on land.” Today, annual mean temperatures rarely exceed 30°C.

As outlandish as these simulations seem, the models might yet again prove to be right. Researchers such as Matthew Huber of Purdue University in West Lafayette, Indiana, have only recently begun to look for evidence of plant dieback in the tropics at this time. No one had thought to look before.

Too cold

There is yet another serious problem for climate modellers. The one place the models suggest did get cold during the hothouse episode is the interior of continents at high latitudes – regions like Siberia. This doesn’t fit with the evidence.

In rocks from the late Cretaceous in Siberia, Robert Spicer of the Open University in Milton Keynes in the UK and his colleagues have found plenty of evidence for ferns and flowering plants, and even possibly the pollen of palm trees (Earth and Planetary Science Letters, vol 267, p 228). Their analysis suggests that at that time Siberia’s mean annual temperature was about 13 °C, rarely touching freezing even in the winter months. “All the climate models give you very, very cold continental interiors [at high latitudes] in the winter time, so cold that you would certainly freeze palm trees and kill them off,” Pierrehumbert says.

One answer to this puzzle is to keep pumping up the CO₂ levels. Models predict that the interiors of continents at high latitudes would not have frozen during the winter if CO₂ levels were higher – but this means the tropics would have got even hotter.

Huber has suggested a possible answer to this dilemma: what if much more heat from the tropics was somehow carried to the poles, keeping the tropics from boiling over. He and Ryan Sriver, also at Purdue, think they have found one possible mechanism.

Hurricane-ridden

They studied conditions in tropical waters before and after the passage of present-day cyclones. They found that cyclones mix up the upper layers of oceans, moving heat downward. They argue that ocean currents then transport this heat towards the poles, reducing the temperature gradient between the tropics and the polar regions ( Nature, vol 447, p 577). Many researchers think the intensity, frequency and duration of tropical cyclones increase with higher temperatures. If so, the amount of heat transported to the poles by cyclones would increase greatly as temperatures rise. In a hurricane-ridden hothouse Earth, this could have kept the tropics below 35 °C, while the poles simmered in subtropical heat.

However, Pierrehumbert thinks that the cyclonic heat-pump idea needs more work, and that explaining the warm interiors of continents remains a challenge. “This is now the most mysterious and toughest looking part of the problem,” he says.

Others might beg to differ. A few lines of evidence point to something seemingly impossible: ice sheets during the warmest phase of the Cretaceous. “Nobody can imagine that we had these high temperatures and at the same time we had some large glaciers in the Antarctica,” says André Bornemann of the University of Leipzig in Germany. Indeed, models cannot replicate these conditions.

One recent study by Bornemann’s team suggests that for a 200,000-year period around 91 million years ago, there were ice sheets at least half the size of the ones that blanket Antarctica today. The evidence comes from oxygen isotope ratios in shells from the Atlantic seabed ( Science, vol 319, p 189).

However, a similar study by Wilson’s team found no evidence of glaciation ( Geology, vol 35, p 615), so this issue is far from settled. But if ice sheets can grow suddenly even during hothouse periods, Wilson point out, it means the climate can change more suddenly and dramatically than anyone thought. “That really demands being understood.”

High volcanic activity

Despite these vexing issues, there is a growing consensus that the hothouse climates were due to high levels of CO₂ in the atmosphere. But where did it come from?

Among other things, the amount of CO₂ in the atmosphere depends on the balance between volcanic activity and the weathering of rocks. High volcanic activity during the Cretaceous might have kept the level of CO₂ high, says Wilson. Later on, volcanic activity may have fallen and weathering increased as the Himalayas began to form, pushing Earth into an icehouse phase.

However, while CO₂ levels up to a million years ago can be directly measured from bubbles of air trapped in ice sheets, it’s much harder working out what they were 100 million years ago. Researchers have to rely on proxies such as the number of pores in fossil leaves, and there are still big uncertainties. Pinning down these numbers is critical, for this would tell us just how sensitive the climate is to rises in CO₂.

Some models suggest CO₂ levels were 16 times as high as pre-industrial levels during the Cretaceous and Eocene hothouses, while others suggest eight times. Despite the uncertainties, eight times fits in far better with the proxy data, suggesting that the climate is highly sensitive to rises in CO₂.

This does not bode well for us, given the amounts of CO₂ we are dumping into the atmosphere. CO₂ levels look set to double from pre-industrial levels and if we keep failing to curb emissions, they could quadruple within 200 years. “Then we are half way towards the CO₂ levels that turned the world into the Cretaceous hothouse,” says Pierrehumbert.