by Ted Nield
The resulting Apparent Polar Wander (APW) curves trace out distinctive signatures; and if two continental blocks, which may now be widely separated because of subsequent continental drift, can be shown to have shared APW curves at a certain time, it is a fair assumption that they were once joined together and moved as one – possibly within a supercontinent.
The magnetic field of the Earth can be thought of as a big bar magnet more or less aligned with the planet’s axis of rotation. The field’s force-lines (which connect the south magnetic pole to the north, describing a giant virtual apple shape in the space around the planet) intersect the Earth’s surface at different angles depending on latitude: vertical at the poles and near-horizontal towards the Equator. When you measure the fossil magnetism in rocks, the inclination of the field (as it is known) gives you the rock’s original latitude when it formed. The direction of north also enables you to tell the continent’s orientation at the time.
But how can you fix the continent’s position with respect to longitude at a given moment? Correlation with other pieces in the jigsaw may help. If there are fossils available (as there are after the base of the Cambrian), it may be possible to infer one continent’s distance from another at a particular time by the similarity (or lack of it) between assemblages. Obviously, the more similar two assemblages, the more migration was possible between their respective habitats and thus the closer together they were on the Earth’s surface. You can also look at sediments. If distinctive mineral grains occur in them, it may be possible to say exactly from where they were eroded. That source may then tie in older rocks that have since become widely separated by later continental drift from the sediments they gave rise to.
A mineral widely used in this sort of study is the humble zircon, the silicate of the element zirconium, which we have encountered already in the nuclei of John Joly’s pleochroic haloes. The fact that zircons contain radioactive elements also makes them very suitable subjects for radiometric dating. Zircon grains, eroded from an original mass of crystalline rocks, will give the same age as other zircons still in situ in the original source rock. You may find sediments containing zircons with quite different ages, suggesting that they were derived from elsewhere, plucked from rocks that have since been rifted off by subsequent drift after the break-up of the original supercontinent.
The Earth and its magnetic field. The lines of force of the field describe a shape in space rather like an apple of which the planet is the core. Note that the angle at which lines of force intersect the surface of the Earth varies from vertical at the pole to near horizontal around the Equator. By finding the inclination of fossil remanent magnetism frozen into rocks, palaeomagnetists (aka ‘palaeomagicians’) are able to work out continents’ ancient latitudes.
Marriage and divorce
Supercontinents, like people, get married and they get divorced. They may also repeat this process more than once. When continents rift, however, they leave traces in the geological record, even after the oceans that the divorcing continents leave in their wake have long been destroyed and forgotten.
Some pieces of evidence, such as the clustered radiometric dates of mountain-building episodes, show us when continents joined together. Different types of evidence have to be used if we are to date the moment when supercontinents broke up. For example, when rocks undergo tension they crack along lines at right angles to the pulling force. Molten magma may rise up and fill these cracks, creating what geologists colourfully term ‘dyke swarms’ (a dyke being a flat, sheetlike, cross-cutting body of once-molten rock). The dykes’ orientation betrays the direction of the tension.
Another consequence of tension is rifting, just as we saw happening in the North Sea Basin at the end of the Permian, when the Atlantic Ocean was beginning to open. In rifting, rock in the valley floor drops down like a keystone in an opening arch, and sediment rushes in to fill the space. Looking at the distribution of rift valley deposits and dyke swarms can help geologists work out how a supercontinent broke apart.
The leading configuration for the supercontinent Rodinia was published by Paul Hoffman in 1991 in the American journal Science, and it is based on the assumption that the Grenville-age mountain belts of the world were all created by the elimination of oceans. Hoffman’s configuration placed the west coast of North America against East Antarctica, a solution known as SWEAT (South West north America-East AnTarctica), and which was originally proposed by the Scottish-American geologist Ian Dalziel. However, such is the uncertainty in these reconstructions that there is as yet no general agreement among scientists even about this basic configuration. Palaeomagnetic evidence may be geophysics, but it doesn’t tell you everything; the results leave a lot of room for interpretation, at least as regards longitude. Also, very ancient rocks lack the fossil controls that can be used to help determine the relative positions of Cambrian and younger continental fragments at given times; and the geological controls (matching mountain belts, dyke swarms, rift systems and the like) – as Wegener himself found – need not of themselves compel particular solutions to the jigsaw.
For example, the chief rival configuration to SWEAT puts Australia alongside North America (and is known as AUSWUS, for Australia-Western United States). Other possible configurations are also occasionally dropped into this alphabet soup at international meetings, all serving to demonstrate the extreme difficulty of solving the Rodinia jigsaw puzzle from the scanty evidence of the ultimate palimpsest.
For supercontinents even older than Rodinia the situation is predictably even worse, though just to show that controversy does not necessarily increase proportionately with age, many geologists believe (with Ian Dalziel) that in between Rodinia and Pangaea another supercontinent, Pannotia, was created. In this vision of events present-day Australia, East Antarctica and India rifted off en masse from Rodinia about 760 million years ago and became reattached to the eastern side of Africa and Arabia. However, whether Pannotia qualifies as a true supercontinent depends on whether this event did any more than build the megacontinent Gondwanaland. Opinion on this remains resolutely divided. One recent textbook on the subject, for example, makes no mention at all of Pannotia among the panoply of pre-Pangaean supercontinents.
We have seen how supercontinents may form by two processes, for which geologists have borrowed the psychological terms ‘introversion’ and ‘extroversion’. Introversion is another name for the Wilson Cycle, sometimes also called ‘accordion tectonics’, whereby a continent rifts apart, forms an ocean within itself and then closes again along the same line, destroying the interior ocean and forming a new range of mountains more or less where an older range once stood. Extroversion simply envisages this rifting continuing, so that the original supercontinent is turned inside out and all its fragments meet one another along their leading edges somewhere else on the planet.
Tales from topographic oceans
The solution to the question of continental drift did indeed lie at the bottom of the ocean, as many geologists suspected. The problem that geologists have in putting pre-Pangaea supercontinents like Rodinia back together (in other words, in distinguishing between such possible solutions as SWEAT and AUSWUS) is the lack of ocean floor from that time, because it has all long since been destroyed. However, it would be an immense help if, even without that ‘road map’, they could somehow tell whether a given supercontinent (whose existence and approximate date of fusion we should be able to tell from such evidence as clustered radiometric ages) formed by introversion or extroversion. The question is, how?
Consider this. In the case of an interior ocean (like the modern Atlantic, opening between fragments of a disintegrating Pangaea) all the ocean floor that has formed is obviously younger than the break-up of Pangaea. If the two sides of the Atlantic should decide to close again and form Chris Scotese’s Pangea Ultima, the ages of all ocean-floor rock that will have to be consumed will fall (at their oldest) between the dates of Pangaea’s initial break-up and (at their youngest) its eventual
reunification.
But on the other hand if Roy Livermore’s vision comes true and 250 million years from now his Novopangaea forms by the opposite process of extroversion, the ocean floor that is consumed in forming the new supercontinent (mainly the Pacific Ocean) will all have lain outside Pangaea at its state of maximum packing: the floor of an ‘exterior’ ocean called Panthalassa. Nearly all of this ocean floor therefore formed before Pangaea was fully assembled; and therefore nearly every piece of it, especially the very first pieces to be destroyed, would yield ages older than the break-up of that supercontinent.
In other words, in introversion, all ocean floor consumed in making a supercontinent will be younger than the break-up of the previous supercontinent, while in extroversion it will be older.
This presents a potential method of telling which of the two mechanisms broke apart and then formed supercontinents older than Pangaea. But how useful could it be? It relies, after all, on dating ocean floor that is consumed in the creation of a new supercontinent. But surely, you may ask, subduction consumes all ocean floor, so it is no longer available for sampling. If there is indeed none left, how can this idea move us forward?
The answer lies in remembering the difference between perfect models and imperfect reality. It may be true in textbook diagrams that subduction destroys all ocean floor, but it is not so in real life. Real subduction is not the clear-cut business represented in these tectonic cartoons; and sometimes, instead of diving down into the Earth like they should, something goes wrong and pieces of ocean crust become scraped off on to the continents to become parts of mountains: true ‘topographic oceans’.
Geologists have long recognized these distinctive rocks. They consist of three basic elements: basalts, erupted underwater and thus forming characteristic pillow shapes; the vertical dykes that fed these submarine eruptions with lava; and the glassy mineral chert (silica, or silicon dioxide) sitting between the pillows. The pillows typically have chilled margins (small crystals, or even glass) where the hot magma met the ocean and cooled very quickly. Below, the dykes that fed their eruption formed as tension at the mid-ocean spreading ridge opened up long, parallel cracks at right angles to the direction of tension. The cherty sediments between the pillows, rich in silica, partly precipitated from solution and partly derived from the skeletons of such creatures as sponges and the microscopic diatom. (There are few other microfossils because, in the low temperatures of the deep sea, those with skeletons made of calcium carbonate dissolve away.)
Cartoon representing plate tectonics. Ocean floor, produced at the mid-ocean ridge, is pulled back down into the crust at subduction zones. However, the process is not this neat in nature, and bits of the ocean floor get scraped off to survive on top of the continental plate, as ‘ophiolite’.
Pillow lavas, dykes and cherts form a classic threesome first noted in 1905 by Gustav Steinmann (1856–1929) and grandly named the ‘Steinmann Trinity’ in his honour by the eccentric Scots geologist Sir Edward Battersby Bailey.
But the Steinmann Trinity is only part of what geologists now call an ‘ophiolite suite’, an ocean-floor remnant that may run to thicknesses of three to five kilometres. Deeper still within the sequence, below the sheeted dykes, come massive, coarsely crystalline rocks called gabbros. These rocks are the solidified remains of the magma chambers that fed the dykes and that cooled more slowly because of their greater volume and depth, and thus formed rock with the same chemistry but larger crystals. Lying below the gabbros are the deepest rocks of all, including peridotite (which has sometimes been altered by seawater to form a well-known rock called serpentine, often used for ornaments, ashtrays and cheeseboards). These dark-green rocks are slices of the Earth’s mantle.
Only with the coming of plate tectonics were these distinctive rock sequences recognized for what they are: the last remnants of long-vanished ocean crust, scraped off on to the continent but destroyed everywhere else by a subduction process that once drew two continental crust blocks together in a mountain-building episode. If enough ophiolite formed during the accretion of a particular supercontinent could be dated, the spread of results should enable us to tell if that supercontinent formed by introversion or extroversion.
However, there is a big problem with this idea, and it has to do with resetting of radiometric clocks. Ophiolites commonly have three important ‘ages’. There is the date they were created, their ‘magmatic age’. Their second radiometric age is an overprint that dates from the point at which they began to be involved in subduction processes, as the increase in heat and pressure began to alter their constituent minerals. Then there is their third age, which is when they were scraped off on to the oncoming continental crust. For this method to work, geologists need to know the rock’s first, true age: the date when it was first born from the mantle.
Pannotia
In 1991 Paul Hoffman wrote a paper with the title ‘Did the breakup of Rodinia turn Gondwanaland inside out?’ According to this model of how Rodinia fragmented, about 760 million years ago a megacontinental landmass made up of the continents we now know as Australia and Antarctica rifted off from Rodinia along a line that now defines the western edge of North America. (This is not the present-day coast of North America, because since Pangaea split up, the USA and Canada have ridden over much of the ancient Pacific Ocean floor, colliding with many small landmasses on the way. These have accreted as what geologists call terranes to the west coast, and built up the mountainous western seaboard of North America in a process akin to the way that litter collects against the top of an ‘up’ escalator, when the steps are finally subducted into the bowels of the machine.)
In making this move, the Australia-Antarctica continent opened up an interior ocean that became the ancestral Pacific. Ancient ocean floor encircling the fragmenting Rodinia was subducted, and the process continued until Australia-Antarctica had swung round and collided with another continent consisting of South America-Africa, which at that time were still joined along the line that would one day open to form the present South Atlantic. Australia-Antarctica thus became fused with South America-Africa, creating a megacontinent we have seen before: Suess’s Gondwanaland. Many geologists also believe (with Ian Dalziel) that at this time other pieces of continental crust were also very close together, possibly fused, and have given this supercontinent assemblage the name ‘Pannotia’ (‘all southern continents’).
Because the ocean that was consumed in this process was all ‘exterior’ to Rodinia, it provides a known example of extroversion; and a possible test for the dating method, because the maximum ages obtained from any ophiolite remnants of the consumed ocean floor should pre-date the break-up of Rodinia.
When Pannotia subsequently split, about 550 million years ago, the interior oceans created by this event, such as Iapetus, Tuzo Wilson’s so-called ‘proto-Atlantic’ separating present-day North America from Western Europe, were subsequently destroyed as the next supercontinent (Pangaea) was created. They were, we know, destroyed by the accordion tectonics of the Wilson Cycle process – that is, by introversion. Dating fragments of those vanished ocean floors should therefore yield ages younger than Pannotia’s break-up.
If a large enough sample of true ages is gathered from ophiolites preserved during the formation of Pannotia, they should fall in a period entirely before the date of Rodinia’s break-up because the ocean floor they represent was all exterior to that supercontinent. Conversely, if the same is done for oceans that formed between the break-up of Pannotia and the formation of Pangaea, the ages obtained should fall within that interval of time because all the ocean floor they represent formed as an interior ocean. If these predictions hold up, our method should show that Rodinia extroverted to form Pannotia and Pannotia introverted to form Pangaea.
Geologists Professor Brendan Murphy of the Tectonics Research Laboratory at St Francis Xavier University, Nova Scotia, and Professor R. Damian Nance of Ohio University have pursued this technique with great success. Their elegant joint
research has, since 1985, resulted in a much clearer picture of how supercontinents assemble. Murphy and Nance have looked at rocks associated with the assembly of Pannotia (about 600 million years old) from the Borborema Belt of Brazil, and the Trans-Saharan and Mozambique Belts of North and East Africa. Rodinia began to split apart about 760 million years ago. So, if Pannotia formed by the consumption of exterior ocean surrounding Rodinia, the formation dates should come out at between 760 million years and about 1100 million years.
Taking on the mantle
But I still haven’t answered the main question: just how, exactly, do you find out such dates reliably? Simple radiometric dating, as we have already seen, allows you to find out when the atomic clock was last reset. But rocks from the floors of vanished oceans, now anomalously preserved in the mountain belts that replaced those oceans when they closed, have all been involved in mountain building and had their clocks reset. Simple radiometric dating would reveal the date the mountains formed, but that is not what we want. We want to get at the time these ocean floors formed at a mid-ocean ridge. We want the birthday, not the date of the funeral, or the mid-life facelift. We want to know the very first time those ocean-crust rocks were created by volcanic melts, erupting at a mid-ocean ridge; the very moment they were derived from the mantle and became part of the crust.
To find out that crucial birth-time, Murphy and Nance have developed a method that combines radiometric-dating techniques with the tendency of isotopes of elements to separate out: become differentiated during natural processes because of their different atomic weights. The technique is complex and beautiful, and it involves using isotopes of two unusual elements: samarium and its daughter element neodymium.