HOW do you swallow an island? It’s a conundrum faced over and again by the
great ocean trenches that skirt many of the world’s continents, where the vast
oceanic plates of the seabed are gulped down into the Earth’s interior. The
plates don’t slip down easily. From the moment they are dished up on the seabed
thousands of kilometres away they are pockmarked and scarred, and by the time
they begin their journey back into the depths of the Earth they are dotted with
islands that stick in the trenches’ throats. The only way to force the islands
down is to slice them into billion-tonne chunks, some of which are spat out to
become flotsam on the planet’s surface.
It’s messy, it’s violent, and we’re just beginning to get a measure of the
tortured complexity of it all. But eventually this geological feasting could
answer a vexed question that has plagued Earth scientists for decades: how
exactly are the world’s continents born?
The rocky skin of our planet comes in two kinds—continental and
oceanic. Continental crust is considerably more complex than oceanic, and much
less dense. Like blocks of cork floating on water, the continents are too
buoyant to be dragged back into the deep interior. This means that some parts of
our continents date back to the planet’s infancy, nearly 4 billion years ago. In
contrast, the oldest oceanic crust is barely 200 million years old.
Advertisement
Over geological aeons, the continents have been repeatedly deformed as the
cork-like blocks crashed into each other. Preserved within these rocks are the
marks of many of the geological processes that have shaped our planet over the
past 4 billion years—so interpreting them is like reading the history of
the globe. There’s a downside to the sheer richness of this archive, though:
many of the events that shaped our continents have obliterated most of the
subtle evidence that would tell us about their origin.
So to understand how the continents were created, we first need to understand
the life cycle of the simpler, younger crust beneath the oceans. Oceanic crust
is formed when the continents drift apart, releasing the pressure on the
underlying mantle and allowing parts of it to melt. The magmas produced spill
out through the cracks of mid-ocean ridges and freeze into solid rock, which
becomes brand new ocean floor.
Where two continents approach each other, though, oceanic crust is forced to
sink back into the mantle, diving beneath the continental crust in a process
called subduction. This sinking is anything but smooth. As an ocean plate makes
its brief earthly journey—a couple of hundred million years or less from
birth to oblivion—it accumulates various kinds of scar tissue. Stressed
beyond endurance by having to accommodate to the Earth’s curved surface, its
ridges splinter into parallel fractures. Hot magma welling up from beneath the
plate can generate vast volcanic edifices like the Hawaiian Islands, which rise
nearly 30,000 metres above the sea floor, and are the world’s tallest mountains.
Larger outpourings can form structures like the Ontong Java Plateau—50
million cubic kilometres of thickened oceanic crust in the Pacific Ocean.
By the time it reaches the ocean trenches or “subduction zones”, the crust is
coated with sediment, sometimes gloopy ooze made from the skeletons of
microscopic sea creatures, sometimes thick layers of detritus eroded from
mountain chains. Stuff all this into a subduction zone and the result is an
almighty mess.
Geologists are beginning to understand why large amounts of this material
never make it back down into the Earth’s interior. As the slab of crust moves
down, sediment and islands are scraped off and smeared, or “accreted”, onto the
edge of the overriding continental plate to form an “accretionary wedge”. Within
the wedge is a jumble of rock types evocatively called melange. Melanges now
cover much of western North America, southern Europe, Japan and South-East Asia
and, at a conservative estimate, make up 20 per cent of the continental crust.
Understand their formation and you understand the origins of much of the
continental mass.
The sedimentary part of the melange is particularly complicated. The mud is
squeezed and smeared out as it is scraped off a subducting plate, and at the
same time it is churned by the large volumes of water escaping to the surface.
In 1990, Daniel Orange from the Monterey Bay Aquarium Research Institute in
California discovered the effects of the subterranean equivalent of a
15-million-year-old high-pressure hosepipe. The grinding plates had squeezed out
water from sediments a few kilometres down. This water then forced itself back
through the sediments that had only just started on the downward path, leaving a
trail of destruction (see Diagram).
Orange uncovered the internal geometry of a stunning melange in Washington
state. He traced the outline of a gigantic inverted cone at least 50 kilometres
across at the top. This, he concluded, had formed when massive volumes of water
passed through the rock, flushing out much of the material and destroying most
of the original sedimentary layering. The inverted cone provides a snapshot of a
hosepipe-like system where a channel of water splays out and surges to the
surface through progressively sloppier layers of sand and mud.
The squeezed-out water can also work in gentler, subtler ways among the
buried layers of sediment. In sediments from the modern accretion complex off
the coast of Peru, networks of mud-lined fractures cut across the sedimentary
layering. The mud in the fractures was flushed in by water passing through the
sediment. Such delicate structures would normally be destroyed over time by the
churning mill of the circulating crust. But last year, working with Ryszard
Kryza from Wroclaw University in Poland, we found wonderfully preserved,
350-million-year-old examples in communist-era boreholes drilled through the
Sudetan Mountains on Poland’s border with the Czech Republic.
We can only guess how these structures escaped destruction. Perhaps
subduction was short-lived and was over before the accretionary complex fully
developed. Whatever the mechanism for their preservation, pressurised water has
squeezed mud into some fractures, while flushing muddy particles out of others.
These structures are all part of the plumbing of the subduction zone.
So the sediment is scraped off the subducting slab, mixed, squeezed and
distorted by water. But that needn’t be the end of the story. In 1993, Roland
von Huene from the Research Centre for Marine Geosciences in Kiel, Germany, and
his colleagues analysed sonar images of the Costa Rican margin. They showed that
huge deposits of partially formed melange periodically collapse and cascade down
the steep submarine slope, depositing a mess of rock and mud back into the
bottom of the trench. There, lying on top of the descending oceanic plate, much
of this jumbled avalanche debris will be scraped off again and reincorporated
into the accretionary wedge.
When Africa met Asia
These tortured sediments are an important part of the rich chowder that, over
time, develops into continental crust. But what about another of the key
ingredients: islands? These lumpy bits of the ocean floor create their own brand
of havoc as they bulldoze into the subduction zone. For the past five years, one
of us (Alan Collins) has been working with Alastair Robertson of Edinburgh
University and colleagues studying the effects of this process in Turkey. Here,
massive piles of sediment and lava form part of an ancient accretionary melange.
Rising as mountains out of their chaotic surroundings, they were created when
the unstoppable force of Africa colliding with Asia overwhelmed the marine
mountains or “seamounts” that lay between.
These island seamounts were shorn off at their base as the ocean floor they
sat upon was subducted. Slices of island, ground progressively smaller as the
crust was pushed ever deeper, were tectonically kneaded and mixed with the sea
floor mud like volcanic raisins in a sedimentary dough. Look at the mineralogy
of these volcanic slices and blocks, and you discover that some of them were
pushed to depths of 30 kilometres or more before being spat out again.
Track forward in time, and move a little sideways, and you can find the same
process taking place today in the Mediterranean. In 1995, a drilling research
cruise south of Cyprus, led by Robertson, discovered thrust faults that were
chopping the Erathosthenes seamount into kilometre-sized blocks as it began to
collide with the Cypriot subduction trench. Erathosthenes is being transferred
piecemeal from the subducting plate to the overriding continental plate, adding
to the land mass of Cyprus.
In 1998, the effect of a seamount colliding with a subduction zone, and the
turmoil created in the accretionary wedge, were ingeniously modelled by Stephane
Dominguez’s group at the Geophysics and Tectonics Laboratory in Montpellier.
According to this model, the seamount first carves a recess in the overriding
accretionary wedge, with the wedge sediment thrust away as a “bow wave” in front
of the seamount. Then the overlying sediment collapses as the seamount dives
down to deeper levels of the accretionary wedge. Whether the seamount is sliced
off and incorporated into the overriding plate (as in the case of
Erathosthenes), or is fully subducted, depends partly on its size and robustness
and partly on the strength of the accretionary wedge material into which it is
ploughing.
Sliced-up ancient seamounts are now being recognised throughout the world.
And it is only a small jump in scale from a seamount to microcontinental
blocks—fragments of continental crust bobbing on a sea of denser oceanic
crust. These large obstructions choke the subduction zone, forcing it to shift
sometimes hundreds of kilometres oceanward.
This appears to have happened in the Aegean when the Menderes-Tauride block,
a large chunk of continent that forms most of south-west Turkey, collided with
the Mediterranean subduction zone about 50 million years ago. Just like the
seamounts, the buoyant block ploughed into the accretionary wedge and got stuck.
However, it was too big to be sliced off by the subduction zone. Instead, a new
subduction zone formed 200 kilometres to the south. The effect of all this has
been to transfer south-west Turkey from the subducting African plate to the
overriding Eurasian plate. As a result of this single step, Eurasia grew by
about 6000 square kilometres.
More and more geologists now believe that large regions of the world,
including the majority of Europe and large swathes of Asia, have been assembled
as a complex collage of such blocks, held in swathes of melange the way bricks
are held in mortar. So what we have among the turmoil of tortured rock seems to
be nothing less than a building site for continents. Catching this enlargement
in the act is difficult, particularly at deeper levels of the subduction system,
where rocks begin to melt and lose the delicate crystalline structures that
reveal their origin.
In 1997, S. Ye and his colleagues from the marine geosciences centre in Kiel
found tantalising hints that subducting material was being plastered onto the
edge of the North American continent at depths greater than 30 kilometres. They
performed a seismic survey off the coast of Alaska and found a deep mass of rock
through which the seismic waves travelled unusually slowly. Seismic waves travel
more slowly through lightweight continental crust than through dense oceanic
crust. So the researchers decided that they were seeing a fragment of subducting
continental crust that hadn’t quite made it into the mantle.
At the same time, Rosalind White, then a PhD student at the University of
Leicester, was finding a new way in which old oceanic plateau might be converted
into continental crust. On the Caribbean island of Aruba, she found evidence
that an ancient oceanic plateau had completely blocked subduction at the Central
American intra-oceanic trench. The oceanic rocks of the plateau then melted and
recrystallised to form a light, continental-style rock called tonalite. In other
words, continental rocks were created exclusively from oceanic rocks.
So did the huge continental masses grow on the sea floor simply as a
by-product of the messy, imperfect mechanics of plate tectonics, much as a
carpenter’s workshop slowly fills with wood shavings? These shavings would have
needed something to anchor themselves to in the first place. White’s work on
Aruba shows that with the help of old oceanic plateaux or seamounts, new
continental material can grow entirely within the oceanic realm, creating this
necessary anchor. Once they have differentiated into this lighter material,
these early continental rafts would be very hard to swallow. Being buoyant, they
would form the overriding plate in later subduction zones and act as nuclei for
additional seamounts to collide with and accrete to, just as happened at
Erathosthenes and in Alaska.
In these early oceans, sediments would also have been accreted to the
continents, just as they do today, and been deformed into melanges. As a number
of continental nuclei grew, they would have collided with each other and formed
larger continental masses, just as Eurasia grew with the addition of what is now
south-west Turkey. Magma that erupted from deep in the subduction zones, where
the subducting slab was heated until hot enough to melt, would also have
contributed to these young continents’ growth.
It seems we have subduction processes to thank for much of the dry land where
we live. At last we have a good idea of where the continents came
from—which is good news for geologists. But it could be bad news for the
oceans. If the theory is right, how long will it be before the shavings fill up
the workshop, and there are no ocean basins left at all?