Plate tectonics – the large-scale movement of Earth’s lithosphere or outer layers – started around three billion years ago, but how those movements started was a bit of a mystery – until today.
With colleagues from the University of Sydney and Claude Bernard University Lyon 1, I show how the spreading primitive continents could have kick-started plate tectonics on Earth in a paper published today in Nature.
Before I go into the modelling behind our results, let’s have a look at plate tectonics today.
At the Earth’s surface, tectonic plates are pushed away at mid-oceanic ridges to make space for the new basaltic crust that comes from the melting of the shallow mantle at a depth no greater than 40km.
In the diagram below, you can see a continent extending, forming a rift valley.
The continent breaks and a mid-oceanic ridge forms above a zone of partially molten mantle (in pink) which produces the basaltic oceanic crust (in purple).
The continental blocks move away from each other. The newly formed oceanic plate cools as it moves away from the mid-oceanic ridge.
As the oceanic plate grows older and colder it also becomes heavier until …
… it sinks into the hot mantle. This is called a subduction zone.
Due to the inverse relationship between temperature and density, tectonic plates become progressively denser as they move away from the mid-oceanic ridge and, no more than 200 million years after their formation, they sink at subduction zones into the hotter convective mantle.
The thickness of the crust is critical to the subduction of plates, and therefore to the operation of plate tectonics. Basalts are less dense than mantle rocks, so a thicker oceanic crust makes plates more buoyant, and prevents their subduction into the mantle.
In the Archaean (2.5-4 billion years ago), the Earth’s interior was much hotter than today and larger volumes of basalts were produced through partial melting of the mantle. As a result, the Earth was covered by a layer of basalt 10-20km thick, making subduction and therefore plate tectonics difficult – if not impossible.
Without plate tectonics, the rigid 100-150km outer shell of the Earth was immobile, and the surface of the Earth was flat, with no mountains or rift valleys.
This explains the lack of volcanic rocks produced by partial melting of subducted basaltic crusts, as well as the absence of sediments coming from the erosion of granites that typically form above subduction zones. Continental sediments and volcanic rocks produced by subducted basaltic crusts appear only about three billion years ago.
Getting the rock rolling
So what initiated plate tectonics? This central question has kept the geoscience community busy over the past decade.
One line of investigation focused on the hypothesis that the increasing torque on the immobile plate by a slowly cooling convective mantle would, at some stage, become strong enough to break the stagnant plate and entrain pieces of it into the convective mantle.
But computer modelling suggests that this idea is only viable when the immobile plate has some pre-existing zones of weakness.
In our Nature paper today, we called upon another mechanism to force the immobile plate into the convective mantle and to kick-start episodes of subduction. On the present-day Earth, we know that the basaltic crust forming at mid-oceanic ridges is in average 5-7km thick.
However, exceptional episodes of partial melting in the deep mantle can led to the accumulation of up to 20km of basalt forming thick oceanic plateaux.
The Ontong Java plateau in the South West Pacific ocean is one example. In the much hotter Archaean Earth, these exceptional episodes of partial melting – typically occurring in raising gigantic bubbles of deep hotter mantle, or alternatively following major meteoritic impacts – may have led to the formation of 50-60km-thick basaltic plateaux.
The residual mantle bubbles, depleted of its basaltic fraction, would have formed the deep roots of these thick oceanic plateaux. The lower section (deeper than 40km) of the thick basaltic crusts may have re-melted to produce the first granitic continental crusts, leading to the first continents.
These first continents are buoyant with respect to both the underlying hotter and more fertile mantle and the surrounding plate. This means that early continents had a natural tendency to spread horizontally under their own weight.
Our computer models suggest that these spreading early continents would have imposed an horizontal stress strong enough to force at their edges the subduction of the stagnant plate.
As the continent slowly spreads …
… its triggers subduction …
… but then the slab detaches, and a stagnant lid is re-established.
These episodes would have been short-lived until subduction became self-sustaining, once the Earth was cold enough to produce basaltic crusts less than 10km thick.
Because plate tectonics is largely responsible for the Earth’s topography and bathymetry (the underwater equivalent of topography), the rise of plate tectonics around three billion years ago would have drastically changed the Earth’s landscapes.
The erosion of high mountain chains, continuously produced by steady-state plate tectonics, provided a constant flow of key nutrients such as phosphorus allowing life to migrate from deep, geographically restricted, ecosystems powered by hydrothermal vents, to more widespread shallow marine environments along continental margins.