ONCE it was a lush green forest, but now it stretches out, yellow and bare,
an endless savannah. Only a few clumps of trees remain as a sad reminder of what
the Amazon rainforest used to be. No, this isn’t a pessimist’s view of the
Advertisement
future. It’s how a long-held theory describes what happened to the region during
the ice ages. And it’s a theory that’s been widely accepted for decades, simply
because it explains one of the greatest mysteries of ecology: why the tropical
forests are so amazingly diverse.
According to this theory, every time the world is plunged into an ice age,
the tropics take a beating. Cold air from the poles alters the patterns of the
prevailing winds and sucks all the moisture out of the equatorial region. The
tropical forests die back, and the parrots and the monkeys run screeching for
the few trees that remain. That creates isolated refuges where plants and
animals evolve away from their fellows. When the forest eventually grows back,
it is full of new species – creating the diverse Amazon rainforest that we
know today. It’s a simple, compelling explanation. No wonder ecologists have put
all their faith in it.
The problem is, it’s probably wrong. For decades, one man has been digging
holes in the jungle to find ancient pollen and test the theory. He concludes
that the Amazon region has been trees, trees and more trees as far back as he
can see. No grasslands, no savannah. Just forest. Now at last the world is
starting to listen to his message. It leaves researchers hunting for a different
explanation for tropical diversity. It also implies that the rainforest is a lot
more resistant to climate change than anyone ever imagined.
The Amazon rainforest is the most biologically diverse place on Earth. Around
80,000 different kinds of plants and 30 million species of animal live there.
And back in 1969, German geologist Jürgen Haffer suggested his idea of ice
age dieback and islands of forest to explain this extraordinary abundance. At
the time, the theory fitted nicely with studies showing that Africa appeared to
have dried out during the last ice age. But the evidence of drought in the
Amazon wasn’t nearly so clear-cut.
Paul Colinvaux, a palaeoecologist based at the Marine Biological Laboratory
in Woods Hole, Massachusetts, decided to test the theory, so he picked up his
drilling equipment and headed for the rainforest. Colinvaux’s idea was simple.
If he could find a spot in the Amazon where soil had been laid down over the
millennia, he could drill a core and see whether it contained pollen from grass
or trees. If there was a drop in tree pollen and a large rise in pollen from
grass that coincided with an ice age Haffer’s theory would be proved true.
The hard part was finding the soil. Most of the forest floor is a deep mass
of overlapping tree roots, while threading through the Amazon basin is a vast
network of rivers which have washed away most of the fresh soil. Much of the
rest is weathered into hard red clay that is devoid of pollen.
Colinvaux needed a lake that had been around for hundreds of thousands of
years, where sediments and pollen grains could have settled undisturbed onto its
floor. Lake Pata, on a hill lying 300 metres above the jungles of north-western
Brazil looked promising. Rain washes debris into Lake Pata almost constantly and
water then seeps out through pores in the rock. There is no outgoing river to
flush away the soil, and the hill is made of hard rock, suggesting the lake has
been there for some time. The only trouble was getting there.
“You take a river boat as far as the river boats will go, which is about
three days,” says Mark Bush, an ecologist from the Florida Institute of
Technology and one of Colinvaux’s companions on the 1990 trek to Pata. “Then you
go into a bus and you drive as far as the bus can drive, which is about another
half a day. And then you get into a small canoe, and you paddle up as far as you
can take a canoe. And then you get dumped out and you’re left standing at the
bottom of this huge hill with no trails up it.” It took Colinvaux and his team
three days just to get their equipment to the top of the hill. The forest floor
was a slippery path of tree roots with potholes a metre deep. Not to mention the
mosquitoes. “We were up there for about two weeks. It was miserable,” says
Bush.
But it was worth it. The researchers bagged a core 6.5 metres long, and
carbon dating showed that the first metre alone held 30,000 years of history,
going right into the heart of the last ice age. Colinvaux and his colleagues
spent painstaking months picking through the mud for pollen grains, sorting them
into some 450 different families. What they found made the refuge hypothesis
look very shaky.
Throughout the core, there was plenty of evidence of trees and no increase in
the amount of grass pollen. And, says Colinvaux, since wind-blown pollen can
travel a long way, his sediment didn’t just represent the vegetation of the
hill. “There are places on that climb where you can see the forest stretching
out before you until it disappears into the haze of the horizon,” says
Colinvaux. “Now if any of that within view had been savannah, I’ll bet anything
we would have seen it in our record.”
The results, published in Science in 1996, put a dent in the refuge
hypothesis, but they didn’t disprove it. It was just possible that by some
extraordinary coincidence Lake Pata might have been smack in the middle of a
refuge. Trees could have remained there while the rest of the forest shrank. So
Colinvaux set out for another lake, 1500 kilometres from the first. The
logistics for this trip were even worse, with a helicopter needed to get in, no
landing site closer than a three-day machete-hack through the jungle, and
malaria running rampant in the nearest village. But Colinvaux eventually bagged
another core. Again, he found nothing but trees going back more than 30,000
years.
The results have yet to be published, but from his conference talks Colinvaux
already knows what the critics will say: “Wow, how odd. You’ve found another
ڳܲ.”
“It’s ridiculous,” he says. “They move the refuge boundaries after our work.”
Areas that were once charted as savannah under the refuge theory are simply
re-inked green to represent forest, he says.
But he’s got yet another, more convincing piece of evidence from the ultimate
drainage site of the forest: the sea floor at the mouth of the Amazon. In 1994,
the Ocean Drilling Program sent a ship to sink a drill into sediment some 400
kilometres from the river mouth. One core they pulled up was more than 35 metres
long, and held sediment more than 50,000 years old. But again it revealed no
increase in grass pollen – just more trees. Says Colinvaux, “even if this is
only draining the lowland basin, that’s exactly where the refugists said there
was most likely to be savannah.”
So just how far back did the forest exist? At first, Colinvaux couldn’t tell.
He had five more metres of mud from the bottom of Lake Pata, but he couldn’t
date it with radiocarbon since that’s only effective to about 50,000 years ago.
Then Bush noticed something strange about the core. Chemical analysis of the
deep mud had revealed that its concentration of potassium varied in a complex
and peculiar zigzag pattern. After years of puzzling over it, Bush finally
realised that this zigzag was marking out a well-known pattern of climate
change.
The more recent ice ages and many other major changes in the Earth’s climate
are generally attributed to wobbles in the Earth’s orbit around the Sun. This
complicated celestial ballet adds up to a wiggly line known as the Milankovitch
cycle, which describes the amount of sunlight reaching the Earth at various
latitudes over time. Bush spotted that the Amazonian potassium data followed the
Milankovitch cycle for the equator.
Why should potassium levels act as a climate marker? The answer, says Bush,
appears to lie in algae. Lake Pata is generally black with dissolved organics,
and light can only penetrate a few centimetres into the water. When the water is
deep that stops any life from growing, since it needs both nutrients from the
bottom of the lake and light from the top. But when the lake levels are low
enough, both light and food come together and the water turns pea green with
algae. Bush believes that the algae take up potassium from the water, and that
this is preserved in the sediment when the algae die. “It’s a peculiarity of
this particular system, it isn’t universal,” says Colinvaux. “We were
ܳ.”
Since the Milankovitch cycle is so well documented, the correlation means the
researchers can date the sediments from their entire core. It turns out that it
goes back 170,000 years – long enough to see into a previous ice age. “We’ve
spotted samples from all the way down the core, and it’s all forested,” says
Bush. “You will not find savannah interludes in it. I’ll go to the bank on that
one.” For Colinvaux, that clinches it. The refuge theory is dead.
Others have yet to be persuaded. Colin Pendry, a botanist from the Royal
Botanical Garden in Edinburgh, is sceptical about the pollen data. “Palynology
is a bit of a blunt tool,” he says, pointing out that trees can’t be identified
at the species level. And even two good sites hardly constitute an extensive
look at an area the size of the continental US. “Paul would make out that
everything’s sewn up and he’s got it right,” says Pendry. “But everything is
still very much up in the air.”
Even so, Pendry, along with some of Colinvaux’s other critics, is beginning
to doubt that the Amazon basin dried up into grassland. “The pendulum has really
swung against the refuge theory,” he says. This group is taking a middle road,
believing that the rainforest became a seasonally dry forest full of legumes
like the Brazil tree.
The most serious objection to Colinvaux’s conclusions comes from an earlier
study of an area called Carajás in the north Brazilian plateau. In 1991,
Maria Lucia Absy and her colleagues found a strong grass pollen signal there
from 24,000 to 11,000 years ago, a period that covers the height of the last ice
age. But Colinvaux dismisses the idea that this was an area of savannah between
refuges. The grass pollen, he claims, is consistent with the marshy plant life
that lives on that high, relatively dry plateau even today.
Generally, it seems, the tide is starting to turn against the refuge theory.
“I really don’t think that is too much in vogue now,” says Joel Cracraft, an
ornithologist at the American Museum of Natural History in New York, who has
written several books about biodiversity in the Amazon. As far as he’s
concerned, shifts in river location could very well explain the clumping of
animal and plant species in the jungle. Though it isn’t obvious, he says, a
river can be a pretty big barrier, even to a bird.
In the end, Colinvaux finds himself coming back to the theory that was pretty
much accepted before Haffer came along – that the Amazon is a generally
pleasant place to live, so species just don’t die out. Given time, he thinks,
the jungle simply accumulates new species.
Does this mean that any future climate change will leave the rainforest
intact? Though researchers are still figuring out the details of tropical
climate during the ice ages, it seems that the forest survived temperature drops
of around 5 °C, and a serious reduction in rainfall (see “Cold and dry”).
Surely, then, with this degree of adaptability, it should be nice and cosy if
things get just a little warmer. “It can take a hell of a lot,” agrees
Colinvaux. “I think the Amazon will almost certainly survive any warming we
throw at it – as long as we don’t chop it down.”
But, he warns, that might be the wrong way to think about things. The ice
ages aren’t really difficult or unusual times. The planet has spent about
four-fifths of the past million years in a deep freeze, making that the normal
state of affairs. “The awful time is now,” he says. “Nature took a normal
glacial period and heated it up and made it horrible. And global warming will
just make it worse.” Just because the Amazon can handle the cold doesn’t mean it
will be able to take the heat. “It may be close to its limit, we just don’t