¿ìè¶ÌÊÓƵ

Suppose God had blown his whistle about 10 million years ago, and said, ‘Let
there be no more change in climate: let there be no more volcanoes; let
tectonic movements cease, and the continents remain exactly where they are.’
Would animals have ceased to evolve? Specifically, would the chimp-like apes
who began to evolve into humans between 7 and 5 million years ago still be
swinging in their trees?

Charles Darwin’s answer would have been no. While he acknowledged that
evolution speeded up when conditions changed, he also supposed that animals
and plants must continue to evolve whatever the conditions. They would
always be competing with each other: leopards becoming ever more cunning,
the better to capture antelopes; antelopes ever swifter, to avoid the
leopards. But some modern biologists feel that evolution would indeed grind
to a halt in the absence of environmental change – especially climatic
change. This controversial idea is rooted in the notion of ‘punctuated
equilibrium’, championed by Niles Eldredge and Stephen Gould in the 1970s.
Its strongest proponent today is Elisabeth Vrba of Yale University.

Vrba postulates that big evolutionary changes – migrations, extinctions and
the creation of new species – are triggered by climate changes. To the
extent that climate changes tend to happen in bursts and at regular
intervals, major evolutionary changes also occur in bursts, or ‘pulses’.
Vrba calls this idea the ‘turnover pulse hypothesis’. She has spent much of
her professional life in Africa studying the fossil record. One aspect that
has intrigued her is the apparently dramatic changes in evolution and
climate which occurred some 2.5 million years ago. At that time, so
increasing evidence shows, there were huge turnovers in faunas worldwide,
and, most important for us, the earliest ancestors of true humans appeared.
What’s more, says Vrba, this burst of evolutionary activity is only one of
several that occurred in Africa and elsewhere.

If Vrba is right, humanity may owe its existence to a relatively short
period of change in the world’s climate. But can her hypothesis be proved?
While some researchers believe that the turnover pulse hypothesis should be
written, at least provisionally, into the textbooks, others profoundly
disagree. Nobody disputes the notion that sudden environmental changes could
trigger bursts of evolutionary upheaval. The contentious question is whether
such bursts could at times be global. Vrba says yes; her critics, no.

‘The turnover pulse hypothesis is effectively dead,’ says Andrew Hill, a
biologist at Yale, who believes that, in arguing for global bursts of
evolutionary change, Vrba is going out on a limb. ‘Yes, environmental shift
leads to evolutionary change. But a series of ‘pulses’ in faunas caused by
episodic shifts of climate? – No.’

Such differences of opinion are perhaps only to be expected, given the
vagaries of the fossil record and the difficulty of reconstructing trends
in the world’s climate over the past few million years. In the last decade
or so, researchers have developed some brilliant techniques for detecting
past climate changes, most based on studies of fossilised pollen and
protozoa. But the endeavour as a whole has suffered some irritating
drawbacks.

Nick Shackleton of the University of Cambridge, for example, studies
sediment cores from the Pacific that contain fossils of the benthic
foramini-fera, or ‘forams’ – single-celled protozoa which live on the
seabed. These fossils contain the two common isotopes of oxygen, ¹⁶0 and
¹⁸0, in different ratios, depending (so Shackleton and others infer) on how
much ice there is in the world at the time. An increase in ¹⁸0 indicates
more ice and therefore a cooler Earth. The deeper the core the older the
fossils, so by measuring the ratio of oxygen isotopes through a foram core,
researchers can trace the Earth’s iciness back through time – and hence
piece together a picture of past global temperature. Shackleton says that
his findings support the idea that the Earth suddenly cooled about 2.5
million years ago, consistent with Vrba’s hypothesis.

Other researchers have reached a similar conclusion from pollen studies.
Cores dug from the bottoms of lakes (or places that used to be the bottoms
of lakes) still contain the pollen of local plants, in various degrees of
fossilisation. Most of the sampled species still exist, or have close modern
relatives, enabling researchers to reconstruct their ecology. This
knowledge, combined with other information about dates (in general, the
pollen layers must correlate with datable strata), reveals what the climate
did in the past, and when.

One proponent of this approach is Henry Hooghiemstra of the University of
Amsterdam, who has studied the sediments of the high plain of Bogot in the
Eastern Cordillera of Colombia. Among other things, concludes Hooghiemstra,
there was ‘a long period of strong climatic fluctuations that started about
2.7 million years ago’. In East Africa, too, there is evidence of cooling in
the highlands 2.5 million years ago, and possibly 3.4 million years ago,
according to pollen studies by Raymonde Bonnefille of the CNRS (France’s
national agency for scientific research at Marseilles. Cooling 2.7 to 2.4
million years ago is also suggested by the appearance of Saharan dust at
around that time, and by the build-up of loess – sediment formed from
wind-blown dust in China. Dust indicates an increase in aridity, which in
turn correlates with cooling. And Michael Archer of the University of New
South Wales suggests that extensive grassland, which is usually linked to a
cooling and drying climate, first appeared in Australia some time after 3
million years ago.

THE PANAMA CONNECTION

‘The evidence for a global climate shift between 2.7 and 2.4 million years
ago is quite strong, and is becoming stronger,’ says Vrba, who links this
shift with major changes in animal life in the fossil record. But there are
complications. Some researchers, including Bill Ruddiman of the University
of Virginia, do not find a sharp fall in temperature 2.5 million years ago
in their foram data. Instead, they perceive a steady decline in temperature
lasting 1 million years and starting 3 million years ago. Furthermore, the
foram research tells you only what happened at the bottom of the sea, not on
land.

Nor is there any firm evidence, says Hill, to suggest that the climate
changes detected on individual continents are part of a global change in the
environment. The case of South America illustrates just how hard it is to
disentangle local and global influences on past climatic upheavals. South
America did not merely grow cooler between 3 and 2 million years ago. Until
3 million years ago it was an island, a piece of Gondwanaland, drifting
through the Southern Ocean, like Australia. But around 3 million years ago
it joined up with North America, via what is now the Isthmus of Panama. This
joining resulted in part from climatic change, since the isthmus appeared
as ice accumulated at the poles, and the sea level fell. But it must itself
have had serious climatic consequences, since it cut off any ocean currents
that might previously have flowed between what are now the Atlantic and the
Pacific. And in East Africa, where our ancestors seem first to have arisen,
the cooling was accompanied by a huge tectonic shift that lifted the
eastern part of the continent to a height of hundreds of metres – itself
enough to produce significant cooling.

‘I don’t doubt that physical changes lead to major evolutionary events,’
says Hill. ‘I am even happy with the idea that evolution might grind to a
stop in the absence of external change. But I do not think that the physical
changes are necessarily climatic; when they are climatic I do not believe
that they are necessarily global.’ However, climatologists like Shackleton
and Ruddiman insist that there was a major shift in the pattern of the
global climate system near 2.5 million years ago.

So everyone agrees that there was widespread cooling between 3 and 2 million
years ago. The uncertainty is over whether the climate changes detected on
individual continents are part of a global change, how sudden that change
might have been, and whether it really precipitated evolutionary events on
different continents. And in theory even this dispute can be resolved as
more data accumulate. What about the crucial part of the turnover pulse
hypothesis? Does the fossil record show that animals changed dramatically,
en masse, 2.5 million years ago?

Hill is again sceptical: ‘Some groups change, like shrews or rodents, and
some, like African pigs, do not.’ With other groups such as hominids, he
says, the data are simply too sparse to make any judgment. ‘In short, you
don’t see a ‘pulse’ of evolutionary change, you see a spread-out response.
It’s difficult to see how this differs from the kinds of stepping up of pace
that Darwin said would happen.’

In Vrba’s view, neither pigs nor hominids can make the case one way or the
other. Her hypothesis does not require that all animal groups would
respond to climatic change at exactly the same moment. It is not surprising,
she says, that animals like shrews and rodents might respond more rapidly
than pigs to climate change. Being small, they would be especially
sensitive to temperature; and having specialised feeding needs, they might
react particularly strongly to a sudden change in vegetation. But the same
may not be true of pigs. Most living pigs are omnivores. Only a few – like
the wart hog of Africa – have become specialist grazers. And even if the
pigs had responded rapidly to climate change, argues Vrba, detecting the
resulting turnover in species would be difficult, because there are so few
species in the fossil record – a mere 13. ‘In such small samples it is very
difficult to see any bursts of change that would stand up to statistical
analysis,’ says Vrba.

But if this should seem like a cop-out, Vrba has looked at the fossil
records of 20 different groups of closely related large animals, from
giraffes to spring hares, and from elephants to African monkeys. The
clearest data come from the fossil remains of the Bovidae, a group of hoofed
animals with an even number of toes, which includes cattle, sheep, goats and
antelopes.

There are hundreds of known species of bovids going back more than 20
million years, and their fossil records are extremely rich. One
particularly dramatic period of innovation occurred between 7 and 5 million
years ago. Cattle migrated into Africa from Eurasia, leading to the first
appearance in Africa of the ancestors of modern buffalo, the Bovini ‘tribe’
of cattle. Grazing antelopes such as reed-buck, waterbuck and rhebok (the
Reduncini) also appeared for the first time. Other grazers – hartebeest,
blesbok and gnu (the Alcelaphini) – evolved for the first time; and so too
did the sable, roan and the oryxes (the Hippotragini), the kudus and nyalas
(the Tragelaphini), and the impala (the single species of the Aepycerotini
tribe). But this period of diversification seems to have given way to a
wave of extinctions and replacements between 3 and 2.5 million years ago. A
huge loss of genera occurred within the various bovid tribes, and today’s
modern genera first evolved.

An earlier turnover of species, between 7 and 5 million years ago, is
intriguing. For one thing, comparable turn-overs seem to have taken place at
the same time in North America and the Siwalik mountains in Pakistan.
Moreover, both these regions, according to fossil evidence, seem to have
experienced previous upheavals in fauna around 11 million years ago and
again 14 million years ago. On the basis of this pattern, Vrba speculates
that there may be a three million year cycle in global climate.

Nobody knows what might cause such long climatic cycles – or indeed,
whether they truly occur. But one possible origin is in the periodic
movements of the Earth’s tectonic plates. Certainly, the cooling has
nothing to do with the kind of shifts in the Earth’s orbit, or in its angle
of tilt – the so-called Milankovitch shifts – that cause ice ages. Ice ages
come and go in a relatively rapid cycle (around 100 000 years), superimposed
on an underlying temperature decline. It is only in the Earth’s recent,
and relatively cool, past that Milankovitch shifts have triggered ice ages.

A DANCE THROUGH TIME

Vrba’s vision of evolutionary pulses leads to various specific predictions.
Vrba predicted that specialist creatures, like shrews, which feed on
insects, slugs and worms which are very sensitive to climate, should respond
to climatic change more quickly than generalists, such as pigs. ‘And that,’
says Vrba, ‘is precisely what you see within the African antelopes.’
Contrast, she says, the explosion of new species that formed between 2.7 and
2.4 million years ago within the Alcelaphini tribe – specialist grazers –
with the impala, which feeds on anything and has remained virtually
unchanged for 3 million years. It matters to the Alcelaphini when the
vegetation alters. They have to change (speciate) or become extinct.
‘Specialist species replace each other as the climate changes, like partners
in a gavotte – a dance through time. The impala simply alters its diet.’

Traditional Darwinism would predict a steady modification of the impala over
3 million years, even without climatic change, because it still needs to
outrun leopards. But, in fact, neither impalas nor leopards have changed
very much. They are both too versatile to be worried by climatic change, and
competition between them and with their own kind does not – as Darwin
supposed – provide sufficient selective pressure to cause them to alter. ‘In
the absence of sufficient extraneous stimulus,’ says Vrba, ‘they have
reached equilibrium.’

It also follows from the turnover pulse hypothesis that animals should
change in ways which correspond to a particular climatic change. In cold
conditions, for instance, natural selection favours large body size in
warm-blooded animals, so that heat can be conserved more easily. Many of
the animals that first appeared 2.5 million years ago were the biggest of
their kind, including giant cattle and antelopes in Africa, and giant deer
(like the Irish elk) in Europe. Even Homo habilis, traditionally regarded as
the first bona fide member of the genus Homo was considerably bigger than
Homo’s likely ancestor, Australopithecus afarensis.

Insights into the environment of our earliest probable ancestor come from
Hank Wesselman of the American River College in Sacramento, California, who
has studied fossil rodents from the same East African deposits, near the
River Omo, that have yielded remains of A. afarensis. He found that about
half of the rodents dating from around 2.9 million years ago – the time of
A. afarensis – are of species associated with moist forest, suggesting, as
does pollen evidence, that A. afarensis arose first in forest. Later
fossils, from around 3 million years ago, suggest drier, more open country.

Yet, if the response to climatic change takes years to unfold, as seems
probable, can we really call it a ‘pulse’? The point, says Vrba, is not how
long the response takes but whether, while it lasts, the rate of change
through extinction, migration and speciation, is faster than at other times,
and whether such discrete periods of rapid change follow shifts in physical
environment. ‘A million years of nothing followed by 100 000 years of
change, and then another million years of stability is a pulse.’ Vrba
maintains that the fossil record might reveal such pulses once it has been
more thoroughly studied, and that modern statistical methods are sensitive
enough to distinguish them from random clustering.

In a nutshell, then, Vrba’s theory is this. Evolutionary pulses should not
be seen as sudden, across-the-board changes. Neither should they be seen as
general increases in evolutionary pace, as Darwin predicted. They should be
seen as discrete periods of evolutionary change that accompany periodic
changes in the physical environment, usually in the climate. Within each
pulse, as the climatic change intensifies, animals are likely to change in a
definite order: specialists first, generalists less readily, perhaps not at
all. The hypothesis further predicts that new species should only form when
initiated by physical changes. So far, Vrba maintains, the evidence is
reinforcing that hypothesis.

If she is right, we have indeed to concede that if God had decided to keep
the world stable five million years ago, then our arboreal ancestors would
not have been forced to change their ways. I would not be writing this and
you would not be reading it. We would still be involved in border disputes
with chimpanzees.

Colin Tudge is a freelance writer. This article was researched at a
‘Conference on Palaeoclimate and Evolution, with Emphasis on Human Origins’,
held in May 1993 at Airlie Conference Centre near Washington DC. Colin
Tudge’s latest book, Last Animals at the Zoo, is now available as an Oxford
paperback.