What controls the height of mountains?

‘Everybody knows’ that mountains grow: the question is, ‘How?’ There is a tale that farmers once believed that they grew from pebbles: ‘every year I try to rid my field of stones, but more are back the following year, so they must grow’… Geoscientists know better – or so they think[!] – and for 130 years have referred to ‘orogeny’, a classically-inspired term (from the Ancient Greek óros and geneia – high-ground creation’) adopted by the US geologist Grove Gilbert. It incorporates the concept of crustal thickening that results from lateral forces and horizontal compression. Another term, now rarely used, is ‘epeirogeny’ (coined too by G.K. Gilbert), wherein the continental surface rises or falls in response to underlying gravitational forces. That could include: changing mantle density over a hot, rising plume; detachment or delamination into the mantle of dense lower lithosphere; loading or unloading by ice during glacial cycles. Epeirogeny is bound up with isostasy, the maintenance of gravitational balance of mass in the outermost Earth.

A small part of the High Himalaya (credit: Access-Himalaya)

In 1990, Peter Molnar and Philip England pointed out that the incision of deep valleys into mountain ranges results in stupendous and rapid removal of mass from orogenic belts, which adds a major isostatic force to mountain building (Molnar, P. & England, P. 1990. Late Cenozoic uplift of mountain ranges and global climate change: chicken or egg? Nature, v. 346, p. 29–34; DOI: 10.1038/346029a0). In their model, the remaining peaks are driven higher by isostasy. They, and others, coupled climate change with compressional tectonics in a positive feedback that drives peaks to elevations that they would otherwise never achieve. Molnar and England’s review saw complex interplays contributing to mountain building, accompanying chemical weathering even changing global climate by sequestering atmospheric CO2 into the minerals that it produces. As well as the height of peaks in active zones of crustal shortening and thickening, such as the Himalaya, Molnar and England’s theory explained the aberrant high peaks at the edge of high plateaus that are passively subject to erosion. Examples of the latter are the isolated peaks beyond the eastern edge of the Ethiopian Plateau that locally have the greatest elevation than the flood basalts that form the plateau: unloading around these peaks has caused them to rise isostatically.

Thirty years on, this paradigm is being questioned, at least as regards active orogens (Dielforder, A. et al. 2020. Megathrust shear force controls mountain height at convergent plate margins. Nature, v. 582, p. 225–229; DOI: 10.1038/s41586-020-2340-7). Armin Dielforder and colleagues at the German Research Centre for Geosciences in Potsdam and The University of Münster consider that overall mountain height is sustained by interactions between three forces. 1. They are prevented from falling apart under their own weight or being pushed up further against gravity by lateral tectonic force. 2. Climate controlled erosion limits mountain height by removing material from the highest elevations. 3. Isostasy keeps the mountains ‘afloat’ above the asthenosphere. The authors have attempted to assess and balance all three major forces that determine the overall elevation of mountain belts.

At a convergent plate margin where one plate is shoved beneath another, the megathrust above the subduction zone behaves in a brittle fashion, with associated friction, towards the surface. At depth this transitions to a zone of ductile deformation dominated by viscosity. A major assumption in this work is that stress in the crust below a mountain belt is neutral; i.e. horizontal, tectonic compression is equal to the weight of the mountains themselves and thus to their height. So, the greater the tectonic compressive force the higher the mountain range that it can support. The test is to compare the actual elevation with that predicted from plate-tectonic considerations. For 10 active orogenic belts there is a remarkable correspondence between the model and actuality. the authors conclude that variation over time of mountain height reflects log-term variations in the force balance, in which they find little sign of a climatic/erosional control. But that doesn’t resolve the issue satisfactorily, at least for me.

The study focuses on the mean elevation, and this leaves out the largest mountains; for instance, their maximum mean elevation for the Himalaya is about 5.46 km (in fact for a narrow  NE-SW swath that may not be representative of the whole range). Yet the Himalaya contains 10 of the world’s highest mountains, all over 8 km high and 50 peaks that top 7 km, adjacent to the Tibetan Plateau. The mean elevation of the whole Himalayan range is 6.1 km. Consequently, it seems to me, the range’s maximum mean elevation must be somewhat higher than that reported by Dielforder et al.  The difference suggests that non-tectonic forces do contribute significantly to Himalayan terrain

See also:  Wang, K. 2020. Mountain height may be controlled by tectonic force, rather than erosion. Nature, v. 582, p. 189-190; DOI: 10.1038/d41586-020-01601-4

Geochemistry and the Ediacaran animals

Hopefully, readers will be fairly familiar with the sudden appearance of the Ediacaran fauna – the earliest abundant, large animals – at the start of the eponymous Period of the Neoproterozoic around 635 Ma. If not, use the Search Earth-logs box in the side bar to find extensive coverage since the start of the 21st century. A June 2019 Earth-logs review of the general geochemical background to the Ediacaran Period can be found here. Ten years ago I covered the possible role of the element phosphorus (P) – the main topic here – in the appearance of metazoans (see: Phosphorus, Snowball Earth and origin of metazoans – November 2010).

One of the major changes in marine sedimentation seen during the Ediacaran was a rapid increase in the deposition on the ocean floor of large bodies of P-rich rock (phosphorite), on which a recent paper focuses (Laakso, T.A. et al. 2020. Ediacaran reorganization of the marine phosphorus cycle. Proceedings of the National Academy of Sciences, v. 117, p. 11961-11967; DOI: 10.1073/pnas.1916738117). It has been estimated that on million-year time scales phosphorites remove only a tiny amount of the phosphorus carried into the oceans by rivers. So, conversely, an increase in deposition of marine P-rich sediment would have little effect on the overall availability of this essential nutrient from the oceans. The Ediacaran boost in phosphorites suggests a connection between them and the arrival of totally new ecosystems: the global P-cycle must somehow have changed. This isn’t the only change in Neoproterozoic biogeochemistry. Thomas Laakso and colleagues note signs of slightly increased ocean oxygenation from changes in sediment trace-element concentrations, a major increase in shallow-water evaporites dominated by calcium sulfate (gypsum) and changes in the relative proportions of different isotopes of sulfur.

Because all marine cycles, both geochemical and those involving life, are interwoven, the authors suggest that changes in the fate of dead organic matter may have created the phosphorus paradox. Phosphorus is the fifth most abundant element in all organisms after carbon, hydrogen, nitrogen and oxygen, followed by sulfur (CHNOPS), P being a major nutrient that limits the sheer bulk of marine life. Perhaps changes to dead organic matter beneath the ocean floor released its phosphorus content, roughly in the manner that composting garden waste releases nutrients back to the soil. Two chemical mechanisms can do this in the deep ocean: a greater supply of sinking organic matter – essentially electron donors – and of oxidants that are electron acceptors. In ocean-floor sediments organic matter can be altered to release phosphorus bonded in organic molecules into pore water and then to the body of the oceans to rise in upwellings to the near surface where photosynthesis operates to create the base of the ecological food chain.

Caption The Gondwana supercontinent that accumulated during the Neoproterozoic to dominate the Earth at the time of the Ediacaran (credit: Fama Clamosa, at Wikimedia Commons)

There is little sign of much increase in deep-ocean oxygen until hundreds of million years after the Ediacaran. It is likely, therefore, that increased availability of oxidant sulfate ions (SO42-) in ocean water and their reduction to sulfides in deep sediment chemically reconstituted the accumulating dead organic matter to release P far more rapidly than before. This is supported by the increase in CaSO4 evaporites in the Ediacaran shallows. So, where did the sulfate come from? Compressional tectonics during the Neoproterozoic Era were at a maximum, particularly in Africa, South America, Australia and Antarctica, as drifting continental fragments derived from the break-up of the earlier Rodinia supercontinent began to collide. This culminated during the Ediacaran around 550 Ma ago with assembly of the Gondwana supercontinent. Huge tracts of it were new mountain belts whose rapid erosion and chemical weathering would have released plenty of sulfate from the breakdown of common sulfide minerals.

So the biological revolution and a more productive biosphere that are reflected in the Ediacaran fauna ultimately may have stemmed from inorganic tectonic changes on a global scale

Changing conditions of metamorphism since the Archaean

Metamorphic petrologists have known since their branch of geology emerged that the intensity or ‘grade’ of metamorphism varies with position in an orogenic belt. This is easily visualised by the sequence mudstone-shale-slate-phyllite-schist-gneiss that results from a clay-rich starting material as metamorphic grade increases. Very roughly speaking, the sequence reflects burial, heat and pressure, and must have been controlled by temperature increasing with depth and pressure: the geothermal gradient. In turn, that depends on internal heat production, geothermal heat flow and the way in which heat is transferred through the deep crust: by thermal conduction or mechanical convection. A particular rock composition gives rise to different metamorphic mineral assemblages under different temperature and pressure conditions.

George Barrow was the first to recognise this in the Southern Highlands of Scotland as a series of zones marked by different index minerals. For instance, in once clay-rich sediments he recognised a succession of new minerals in the sequence chlorite; biotite; garnet; staurolite; kyanite; sillimanite in rocks of progressively higher metamorphic grade. Barrow found that once basaltic lavas interleaved with the sediments displayed zones with different characteristic minerals. Other metamorphic terrains, however, revealed different index minerals. Experimental mineralogy eventually showed that Barrow’s zones and others reflected a wide range of chemical reactions between minerals that reach equilibrium over different combinations of pressure and temperature. This enabled geologists to distinguish between metamorphism that had occurred under conditions of high-pressure and low-temperature, low-P and high-T and intermediate conditions (see diagram). This suggested that metamorphic rocks can form in areas with different heat flow and geothermal gradients. Geochemical means of assessing the actual temperatures and pressures at which particular rocks had reached mineralogical equilibrium, known as ‘thermobarometry’, now enable such variations to be assessed quantitatively.

The latest division in pressure-temperature space of different styles of metamorphism (colours) and the main mineral equilibria (dashed lines) that define them

It has long been suspected that the average T/P conditions revealed by metamorphic rocks have varied over geological time, as well as from place to place at any one time. A recent paper has analysed thermobarometric data from the earliest Archaean to recent times (Brown, M. et al. 2020, Evolution of geodynamics since the Archean: Significant change at the dawn of the Phanerozoic: Geology, v. 48, p. 488–492; DOI: 10.1130/G47417.1) They conclude that from the Archaean to the start of the Neoproterozoic the average P/T ratio was more than twice as high as it was in the following billion years. At about 2 Ga they suggests a relatively sudden decrease that correlates with what they regard as the first major assembly of continental crust: the Columbia (Nuna) supercontinent. The Mesoproterozoic Era, occupied by the disassembly of Columbia and the eventual creation of the Rodinia supercontinent, retained a high mean T/P. That began to decline with the break-up of Rodinia and a succession of tectonic cycles of ocean opening and closing during the Neoproterozoic and the Phanerozoic. This phase of truly modern plate textonics saw first the assembly of Gondwana and then the all-encompassing Pangaea, followed by its break up as we witness today. There are other correlations with the T/P variations, but they need not detain us.

The raw metamorphic data (564 points spanning 3.5 Ga) are by no means evenly spaced in time, and four dense clusters of points show a very wide spread of T/P – up to 2 orders of magnitude. Yet the authors have used locally weighted scat­terplot smoothing (LOWESS) to reduce this to a smoothed curve with a zone of uncertainty that is a great deal narrower than the actual spread of data. Frankly, I do not believe the impression of systematic change that this approach has produced, though I am not a statistician. To a lesser extent than me, it seems that neither does Peter Cawood, who comments on the paper in the same issue of Geology: more clearly than do the authors themselves.

Peter Cawood’s ‘take’ on the relationship between tectonic development and other important variables in the Earth-system with the estimate by Brown et al. of the mean metamorphic T/P (‘thermobaric’) variation through Earth history

Cawood’s view is that it was all due to a steady fall in mantle temperature and related broad changes in tectonic processes. But metamorphic rocks form in only the outermost 100 km of the Earth. The post-800 Ma examples include a much greater proportion of those formed under high- and ultrahigh pressures – blueschists and various kinds of eclogite – than do the earlier metamorphic belts. This weights the post-800 Ma record to lower mean T/P. Such rocks form in subduction zones and their high density might seem to doom them to complete resorption into the deep mantle. Yet large chunks now end up embedded in continents, interleaved with less extreme materials. Cawood suggests, as do others, that cooling of the mantle has enabled deeper break-off of subducted slabs to meet their end at the core-mantle boundary. The retained low T/P lithosphere since 800 Ma may have been sliced into the continents by increased underthrusting during continent-continent collisions that dominate the more modern orogenic-metamorphic belts.

See also:  Cawood, P.A. 2020 Earth Matters: A tempo to our planet’s evolution: Geology, v. 48, p. 525–526; DOI: 10.1130/focus052020.1

Earliest direct evidence of plate motions

There are two ways that we recognise the movement of tectonic plates. Since the latter half of the Mesozoic Era, following break up of the Pangaea supercontinent, it bests manifests itself in the magnetic ‘stripes’ on the ocean floor. They result from alternating polarisation of the geomagnetic field as new oceanic lithosphere is generated at constructive plate boundaries to drive sea-floor spreading. The oldest remaining stripes date back to the early Jurassic. For earlier times geologists have to turn to the continental crust.  Lavas and some sedimentary rocks undergo magnetisation at the time of their formation and retained that imprint. Such remanent, palaeomagnetism reveals the original latitude at which it was imprinted, together with the subsequent rotation of a drifting continent relative to an assumed N to S axis joining the opposed magnetic poles. The apparent ‘wandering’ of the pole through time when successive ancient pole positions of different ages are plotted in relation to the present position of a continent is a good guide to its history of drifting as a result of plate tectonics. Comparing the polar-wander paths of two continents allows the time when they were formerly united to be estimated. So palaeomagnetic pole data makes it possible to reconstruct not just Pangaea but a whole series of earlier supercontinents, ancient magnetic data being supplemented by other geological evidence such as reconnecting the trends on different continents of ancient mountain belts.

Apparent polar wander paths for two continents for a period when they were united then split and were separated by sea-floor spreading, eventually to collide and reunite

The further back in time the fewer palaeomagnetic pole positions have been estimated, and the more uncertain are the apparent polar wander paths and the more complex each continent’s accumulated geological history. One of the reasons for such uncertainty is that episodes of metamorphism can reset a rock’s remanent magnetisation, hundreds of million years after it originally formed. Thus, the harder it becomes to be certain about early supercontinents that have been suggested, of which there are quite a few. The earliest that has been proposed is Vaalbara, albeit on grounds of geological similarity, that supposedly united the Kaapvaal and Pilbara Cratons of southern Africa and Western Australia, respectively. Its duration is suggested to have been between 3.6 to 2.8 Ga (billion years ago). The oldest supercontinents with sound palaeomagnetic records date from the end of the Archaean Eon (2.5 Ga). It is the lack or uncertainty of earlier palaeomagnetic evidence that makes the start of plate tectonics the subject of so much debate.

However, geophysicists continually strive to improve the detection of ancient magnetisation, and advances have been made recently to unravel original magnetisation signals from those that have been superimposed later. The fruits of these developments are borne out by a study of a sequence of mafic lavas from the Pilbara Craton that formed about 3.2 Ga ago (Brenner, A.R. et al. 2020. Paleomagnetic evidence for modern-like plate motion velocities at 3.2 Ga. Science Advances, v. 6, article eaaz8670; DOI: 10.1126/sciadv.aaz8670). Alec Brenner and colleagues from several US universities measured palaeomagnetism in more than 200 diamond drill cores from two localities in this sequence and combined their data with others from the Pilbara to cover a roughly 600 Ma period between 3.35 to 2.77 Ga. The palaeopoles form a polar wander path that spans roughly 50 degrees of palaeolatitude. From this they have been able to estimate, in considerable detail, the rate at which the Pilbara Craton had moved in Mesoarchaean. In the first 170 Ma the average horizontal motion was about 2.5 cm per year, falling rapidly to 0.4 cm per year over the following 410 Ma. The earlier speed is comparable with the average of modern plate motions. Data from the later period suggests relative stagnation. Motions over the entire ~600 Ma could be due to episodic operation of plate tectonics on the global scale, or a local slowing in the rate of plate growth.

Earliest plate tectonics tied down?

Papers that ponder the question of when plate tectonics first powered the engine of internal geological processes are sure to get read: tectonics lies at the heart of Earth science. Opinion has swung back and forth from ‘sometime in the Proterozoic’ to ‘since the very birth of the Earth’, which is no surprise. There are simply no rocks that formed during the Hadean Eon of any greater extent than 20 km2. Those occur in the 4.2 billion year (Ga) old Nuvvuagittuq greenstone belt on Hudson Bay, which have been grossly mangled by later events. But there are grains of the sturdy mineral zircon ZrSiO4)  that occur in much younger sedimentary rocks, famously from the Jack Hills of Western Australia, whose ages range back to 4.4 Ga, based on uranium-lead radiometric dating. You can buy zircons from Jack Hills on eBay as a result of a cottage industry that sprang up following news of their great antiquity: that is, if you do a lot of mineral separation from the dust and rock chips that are on offer, and they are very small. Given a laser-fuelled SHRIMP mass spectrometer and a lot of other preparation kit, you could date them. Having gone to that expense, you might as well analyse them chemically using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) to check out their trace-element contents. Geochemist Simon Turner of Macquarie University in Sydney, Australia, and colleagues from Curtin University in Western Australia and Geowissenschaftliches Zentrum Göttingen in Germany, have done all this for 32 newly extracted Jack Hills zircons, whose ages range from 4.3 to 3.3 Ga (Turner, S. et al. 2020. An andesitic source for Jack Hills zircon supports onset of plate tectonics in the HadeanNature Communications, v. 11, article 1241; DOI: 10.1038/s41467-020-14857-1). Then they applied sophisticated geochemical modelling to tease out what kinds of Hadean rock once hosted these grains that were eventually eroded out and transported to come to rest in a much younger sedimentary rock.

Artist’s impression of the old-style hellish Hadean (Credit : Dan Durday, Southwest Research Institute)

Zircons only form duuring the crystallisation of igneous magmas, at around 700°C, the original magma having formed under somewhat hotter conditions – up to 1200°C for mafic compositions. In the course of their crystallising, minerals take in not only the elements of which they are mainly composed, zirconium, silicon and oxygen in the case of zircon , but many other elements that the magma contains in low concentrations. The relative proportions of these trace elements that are partitioned from the magma into the growing mineral grains are more or less constant and unique to that mineral, depending on the particular composition of the magma itself. Using the proportions of these trace elements in the mineral gives a clue to the original bulk composition of the parent magma. The Jack Hills zircons  mainly  reflect an origin in magmas of andesitic composition, intermediate in composition between high-silica granites and basalts that have lower silica contents. Andesitic magmas only form today by partial melting of more mafic rocks under the influence of water-rich fluid driven upwards from subducting oceanic lithosphere. The proportions of trace elements in the zircons could only have formed in this way, according to the authors.

Interestingly, the 4.2 Ga Nuvvuagittuq greenstone belt contains metamorphosed mafic andesites, though any zircons in them have yet to be analysed in the manner used by Turner et al., although they were used to date those late-Hadean rocks. The deep post-Archaean continental crust, broadly speaking, has an andesitic composition, strongly suggesting its generation above subduction zones. Yet that portion of Archaean age is not andesitic on average, but a mixture of three geochemically different rocks. It is referred to as TTG crust from those three rock types (trondhjemite, tonalite and granodiorite). That TTG nature of the most ancient continental crust has encouraged most geochemists to reject the idea of magmatic activity controlled by plate tectonics during the Archaean and, by extension, during the preceding Hadean. What is truly remarkable is that if mafic andesites – such as those implied by the Jack Hills zircons and found in the Nuvvuagittuq greenstone belt – partially melted under high pressures that formed garnet in them, they would have yielded magmas of TTG composition. This, it seems, puts plate tectonics in the frame for the whole of Earth’s evolution since it stabilised several million years after the catastrophic collision that flung off the Moon and completely melted the outer layers of our planet. Up to now, controversy about what kind of planet-wide processes operated then have swung this way and that, often into quite strange scenarios. Turner and colleagues may have opened a new, hopefully more unified, episode of geochemical studies that revisit the early Earth . It could complement the work described in An Early Archaean Waterworld published on Earth-logs earlier in March 2020.

An Early Archaean Waterworld

In Earth-logs you may have come across the uses of oxygen isotopes, mainly in connection with their variations in the fossils of marine organisms and in ice cores. The relative proportion of the ‘heavy’ 18O isotope to the ‘light’ 16O, expressed by δ18O, is a measure of the degree of fractionation between these isotopes under different temperature conditions when water evaporates. What happens is that H216O, in which the lighter isotope is bound up, slightly more easily evaporates thus enriching the remaining liquid water in H218O. As a result the greater the temperature of surface water and the more of evaporates, the higher is its δ18O value. Shells that benthonic (surface-dwelling) organism secrete are made mainly of the mineral calcite (CaCO3). Their formation involves extracting dissolved calcium ions and CO2 plus an extra oxygen from the water itself, as calcite’s formula suggests. So plankton shells fossilised  in ocean-floor sediments carry the δ18O and thus a temperature signal of surface water at the place and time in which they lived. Yet this signal is contaminated with another signal: that of the amount of water evaporated from the ocean surface (with lowered  δ18O) that has ended up falling as snow and then becoming trapped in continental ice sheets. The two can be separated using the δ18O found in shells of bottom-dwelling (benthonic) organisms, because deep ocean water maintains a similar low temperature at all time (about 2°C). Benthonic δ18O is the main guide to the changing volume of continental ice throughout the last 30 million year or so. This ingenious approach, developed about 50 years ago, has become the key to understanding past climate changes as reflected in records of ice volume and ocean surface temperature. Yet these two factors are not the only ones at work on marine oxygen isotopes.

Artistic impression of the Early Archaean Earth dominated by oceans (Credit: Sci-news.com)

When rainwater flows across the land, clays in the soil formed by weathering of crystalline rocks preferentially extract 18O and thus leave their own δ18O mark in ocean water. This has little, if any, effect on the use of δ18O to track past climate change, simply because the extent of the continents hasn’t changed much over the last 2 billion years or so. Likewise, the geological record over that period clearly indicates that rain, wet soil and water flowing across the land have all continued somewhere or other, irrespective of climate. However, one of the thorny issues in Earth science concerns changes of the area of continents in the very long term. They are suspected but difficult to tie down. Benjamin Johnson of the University of Colorado and Boswell Wing of Iowa State University, USA, have closely examined oxygen isotopes in 3.24 billion-year old rocks from a relic of Palaeoarchaean ocean crust from the Pilbara district of Western Australia that shows pervasive evidence of alteration by hot circulating ocean water (Johnson, B.W. & Wing, B.A. 2020. Limited Archaean continental emergence reflected in an early Archaean 18O-enriched ocean. Nature Geoscience, v. 13, p. 243-248; DOI: 10.1038/s41561-020-0538-9). Interestingly, apart from the composition of the lavas, the altered rocks look just the same as much more recent examples of such ophiolites.

The study used many samples taken from the base to the top of the ophiolite along some 20 traverses across its outcrop. Overall the isotopic analyses suggested that the circulating water responsible for the hydrothermal alteration 3.2 Ga ago was much more enriched in 18O than is modern ocean water. The authors’ favoured explanation is that much less continental crust was exposed above sea level during the Palaeoarchaean Era than in later times and so far less clay was around on land. That does not necessarily imply that less continental crust existed at that time compared with the Archaean during the following 700 Ma , merely that the continental ‘freeboard’ was so low that only a few islands emerged above the waves. By the end of the Archaean 2.5 Ga ago the authors estimate that oceanic δ18O had decreased to approximately modern levels. This they attribute to a steady increase in weathering of the emerging continental landmasses and the extraction of 18O into new, clay-rich soils as the continents emerged above sea level. How this scenario of a ‘drowned’ world developed is not discussed. One possibility is that the average depth of the oceans then was considerably less than it was in later times: i.e. sea level stood higher because the volume available to contain ocean water was less. One possible explanation for that and the subsequent change in oxygen isotopes might be a transition during the later Archaean Eon into modern-style plate tectonics. The resulting steep subduction forms deep trench systems able to ‘hold’ more water. Prior to that faster production of oceanic crust resulted in what are now the ocean abyssal plains being buoyed up by warmer young crust that extended beneath them. Today they average around 4000 m deep, thanks to the increased density of cooled crust, and account for a large proportion of the volume of modern ocean basins.

Metamorphic evidence of plate tectonic evolution

The essence of plate tectonics that dominates the Earth system today is the existence of subduction zones that carry old, cold oceanic lithosphere to great depths where they become denser by the conversion of the mineralogy of hydrated basalt to near-anhydrous eclogite. Such gravitational sinking imparts slab-pull force that is the largest contributor to surface plate motions. Unequivocally demonstrating the action of past plate tectonics is achieved from the striped magnetic patterns above yet-to-be-subducted oceanic lithosphere, the oldest being above the Jurassic remnant of the West Pacific. Beyond that geoscientists depend on a wide range of secondary evidence that suggest the drifting and collision of continents and island arcs, backed up by palaeomagnetic pole positions for various terranes that give some idea of the directions and magnitudes of horizontal motions.

Occasionally – the more so further back in time – metamorphic rocks (eclogites and blueschists) are found in linear belts at the surface, which show clear signs of low-temperature, high pressure metamorphism that created the density contrast necessary for subduction. Where such low T/P belts are paired with those in which the effects of high T/P metamorphism occurred they suggest distinctly different geothermal conditions: low T/P associated with the site of subduction of cold rock; high T/P with a zone of magmagenesis – at island- or continental arcs – induced by crustal thickening and flux of volatiles above deeper subduction. Such evidence of geothermal polarity suggests a destructive plate margin and also the direction of relative plate motions. The oldest known eclogites (~2.1 Ga) occur in the Democratic Republic of the Congo, but do they indicate the start of modern-style plate tectonics?

Interestingly, ‘data mining’ and the use of statistic may provide another approach to this question. Determination of the temperatures and pressures at which metamorphic rocks formed using the mineral assemblages in them and the partitioning of elements between various mineral pairs has built up a large database that spans the last 4 billion years of Earth history. Plotting each sample’s recorded pressure against temperature shows the T/P conditions relative to the thermal gradients under which their metamorphism took place. Robert Holder of Johns Hopkins University and colleagues from the USA, Australia and China used 564 such points to investigate the duration of paired metamorphism (Holder, R.M. et al. 2019. Metamorphism and the evolution of plate tectonics. Nature, v. 572, p. 378–381; DOI: 10.1038/s41586-019-1462-2).

The 109 samples from Jurassic and younger metamorphosed terranes that demonstrably formed in arc- and subduction settings form a benchmark against which samples from times devoid of primary evidence for tectonic style can be judged. The post-200 Ma data show a clear bimodal distribution in a histogram plot of frequency against thermal gradient, with peaks either side of a thermal gradient of 500°C GPa-1 (~17°C km-1); what one would expect for paired metamorphic belts. A simple bell-shaped or Gaussian distribution of temperatures would be expected from metamorphism under a similar geothermal gradient irrespective of tectonic setting.

Metc PvT
Pressure-temperature data from Jurassic and younger metamorphic rocks (a) pressure vs temperature plot; (b) Frequency distribution vs log thermal gradient. (Credit: Holder et al. 2019, Fig. 1)

Applying this approach to metamorphic rocks dated between 200 to 850 Ma; 850 to 1400 Ma; 1400 to 2200 Ma, and those older than 2200 Ma, Holder and colleagues found that the degree of bimodality decreased with age. Before 2200 Ma barely any samples fell outside a Gaussian distribution. Also, the average T/P of metamorphism decreased from the Palaeoproterozoic to the present. They interpret the trend towards increased bimodality and decreasing average T/P as an indicator that the Earth’s modern plate-tectonic regime has developed gradually since the end of the Archaean Eon (2500 Ma). Their findings also tally with the 2.1 Ga age of the oldest eclogites in the DRC.

Plate tectonics is primarily defined as the interaction between slabs of lithosphere that are rigid and brittle and move laterally above the ductile asthenosphere. Their motion rests metaphorically on the principle that ‘what comes up’ – mantle-derived magma – ‘must go down’ in the form of displaced older material that the mantle resorbs. That is more likely to be oceanic lithosphere whose bulk density is greater than that supporting the thick, low-density continental crust. Without the steeper subduction and slab pull conferred by the transformation of hydrated basalt to much denser eclogite, subduction would not result in low T/P metamorphism paired with that resulting from high T/P conditions in magmatic arcs. But, while ever lithosphere was rigid and brittle, plate tectonics would operate, albeit in forms different from that which formned terranes younger than the Jurassic

The effect of surface processes on tectonics

Active sedimentation in the Indus and Upper Ganges plains (green vegetated) derived from rapid erosion of the Himalaya (credit: Google Earth)

The Proterozoic Eon of the Precambrian is subdivided into the Palaeo-, Meso- and Neoproterozoic Eras that are, respectively, 900, 600 and 450 Ma long. The degree to which geoscientists are sufficiently interested in rocks within such time spans is roughly proportional to the number of publications whose title includes their name. Searching the ISI Web of Knowledge using this parameter yields 2000, 840 and 2700 hits in the last two complete decades, that is 2.2, 1.4 and 6.0 hits per million years, respectively. Clearly there is less interest in the early part of the Proterozoic. Perhaps that is due to there being smaller areas over which they are exposed, or maybe simply because what those rocks show is inherently less interesting than those of the Neoproterozoic. The Neoproterozoic is stuffed with fascinating topics: the appearance of large-bodied life forms; three Snowball Earth episodes; and a great deal of tectonic activity, including the Pan-African orogeny. The time that precedes it isn’t so gripping: it is widely known as the ‘boring billion’ – coined by the late Martin Brazier – from about 1.75 to 0.75 Ga. The Palaeoproterozoic draws attention by encompassing the ‘Great Oxygenation Event’ around 2.4 Ga, the massive deposition of banded iron formations up to 1.8 Ga, its own Snowball Earth, emergence of the eukaryotes and several orogenies. The Mesoproterozoic witnesses one orogeny, the formation of a supercontinent (Rodinia) and even has its own petroleum potential (93 billion barrels in place in Australia’s Beetaloo Basin. So it does have its high points, but not a lot. Although data are more scanty than for the Phanerozoic Eon, during the Mesoproterozoic the Earth’s magnetic field was much steadier than in later times. That suggests that motions in the core were in a ‘steady state’, and possibly in the mantle as well. The latter is borne out by the lower pace of tectonics in the Mesoproterozoic. Continue reading “The effect of surface processes on tectonics”

Tectonics and glacial epochs

Because the configuration of continents inevitably affects the ocean currents that dominate the distribution of heat across the face of the Earth, tectonics has a major influence over climate. So too does the topography of continents, which deflects global wind patterns, and that is also a reflection of tectonic events. For instance, a gap between North and South America allowed exchange of the waters of the Pacific and Atlantic Oceans throughout the Cenozoic Era until about 3 Ma ago, at the end of the Pliocene Epoch, although the seaway had long been shallowing as a result of tectonics and volcanism at the destructive margin of the eastern Pacific. That seemingly minor closure transformed the system of currents in the Atlantic Ocean, particularly the Gulf Stream, whose waxing and waning were instrumental in the glacial-interglacial cycles that have persisted for the last 2.5 Ma. This was partly through its northward transport of saltier water formed by tropical evaporation that cooling at high northern latitudes encouraged to sink to form a major component of the global oceanic heat conveyor system.   Another example is the rise of the Himalaya following India’s collision with Eurasia that gave rise to the monsoonal system  dominating the climate of southern Asia. The four huge climatic shifts to all-pervasive ice-house conditions during the Phanerozoic Eon are not explained so simply: one during the late-Ordovician; another in the late-Devonian; a 150 Ma-long glacial epoch spanning much of the Carboniferous and Permian Periods, and the current Ice Age that has lasted since around 34 Ma. Despite having been at the South Pole since the Cretaceous Antarctica didn’t develop glaciers until 34 Ma. So what may have triggered these four major shifts in global climate?

Five palaeoclimatologists from the University of California and MIT set out to find links, starting with the most basic parameter, how atmospheric greenhouse gases might have varied. In the long term CO2 builds up through its emission by volcanoes. It is drawn down by several geological processes: burial of carbon and carbonates formed by living processes; chemical weathering of silicate minerals by CO2 dissolved in water, which forms solid calcium carbonate in soil and carbonate ions in seawater that can be taken up and buried by shell-producing organisms. Rather than comparing gross climate change with periods of orogeny and mountain building, mainly due to continent-continent collisions, they focused on zones that preserve signs of subduction of oceanic lithosphere – suture zones (Macdonald,F.A. et al. 2019. Arc-continent collisions in the tropics set Earth’s climate state. Science, v. 363 (in press); DOI: 10.1126/science.aav5300 ). Comparing the length of all sutures active at different times in the Phanerozoic with the extent of continental ice sheets there is some correlation between active subduction and glaciations, but some major misfits. Selecting only sutures that were active in the tropics of the time – the zone of most intense chemical weathering – results in a far better tectonic-climate connection. Their explanation for this is not tropical weathering of all kinds of exposed rock but of calcium- and magnesium-rich igneous rocks; basaltic and ultramafic rocks. These dominate oceanic lithosphere, which is exposed to weathering mainly where slabs of lithosphere are forced, or obducted, onto continental crust at convergent plate margins to form ophiolite complexes. The Ca- and Mg-rich silicates in them weather quickly to take up CO2 and form carbonates, especially in the tropics. Through such weathering reactions across millions of square kilometres the main greenhouse gas is rapidly pulled out of the atmosphere to set off global cooling.

407458aa.2
Top – variation in the total length of active, ophiolite-bearing sutures during the Phanerozoic; middle – length of such sutures in the tropics; bottom – extent of Phanerozoic glaciers. (Credit: Macdonald et al. 2019; Fig.2

Rather than the climatic influence of tectonics through global mountain building, the previous paradigm, Macdonald and colleagues show that the main factor is where subduction and ophiolite obduction were taking place. In turn, this very much depended on the configuration of continents on which ophiolites can be preserved. The most active period of tectonics during the Mesozoic, as recorded by the global length of sutures, was at 250 Ma – the beginning of the Triassic Period – but they were mainly outside the tropics, when there is no sign of contemporary glaciation. During the Ordovician, late-Devonian and Permo-Carboniferous ice-houses active sutures were most concentrated in the tropics. The same goes for the build-up to the current glacial epoch.

Read more on Palaeoclimatology and Tectonics

How does subduction start?

Robert Stern of the University of Texas at Dallas, USA, and Taras Gerya of ETH, Zurich, have produced a masterly review of how subduction gets started from place to place, and from time to time in geological history (Stern, R.J. & Gerya, T. 2018. Subduction initiation in nature and models: A review. Tectonophysics, v. 744 (in press); (PDF). It is the foundering of oceanic lithosphere into the mantle and gravity that give modern plate tectonics the bulk of energy that drives it along by slab pull. Yet the mantle’s consumption of a lithospheric slab somehow has to be set in motion from the symmetrical spreading of ocean floor as occurs either side of a constructive margin. It could not happen were the lithosphere to retain its low bulk density relative to mantle peridotite for all time. Moreover, it wouldn’t last for long were the lithosphere not to retain its strength through hundreds of kilometres depth as it sinks into the mantle. Active subduction zones have consumed vast amounts of oceanic lithosphere, for more than 65 million years, especially in fast-spreading ocean basins such as the western and eastern Pacific. The record is held by the destructive margin on the west flank of South America where more than 150 million years-worth of eastern Pacific lithosphere has been swallowed. Yet in order for oceanic lithosphere, which is stronger than that beneath the continents, somehow to fail and begin to sink a linear weak zone must develop at the interface between two incipient new plates. On top of that, all subduction on Earth is one-sided. A simple mechanism involving just thermal convection predicts that both plates either side of a break would have similar density so both should sink, more or less symmetrically.

subduction types
Various ways in which subduction may start. (Credit: Stern and Gerya 2018 – in press – Figure 4)

Geophysical observations reveal that terrestrial subduction can be divided into that which is induced by plate motions and changes in force balance within spreading plates, or spontaneously due to unique conditions developing along the line of initiation. In the first class are cases where a microcontinent is driven into another continental margin and extinguishes the subduction responsible, while spreading continues behind the accreted microcontinent drive older lithosphere beneath the suture (this may have happened in the past but is not seen today). Another, similar, induced case occurs where an oceanic island arc accretes by subduction beneath it so that subduction flips in polarity to consume the driving sea-floor spreading. The loading of oceanic lithosphere by sediments piled onto it by erosion of a continental margin may spontaneously collapse to result in subduction beneath the sedimentary wedge and the continent (again, not happening today, but inferred from examples inferred by earlier geological history). Spontaneous failure may also occur where old, cold lithosphere is juxtaposed with younger by transform faulting, or where a mantle plume heats up lithosphere to create a thermally weakened zone.

Stern and Gerya do not leave the issue at simple mechanics but discuss how plates may develop weak zones or inherit them from earlier tectonic events. The role of water released by metamorphism of descending materials may encourage the observed one-sidedness of subduction by reducing frictional resistance and plate strength and make the process self-sustaining. The paper also discusses the various permutations and combinations that affect the style of induced destructive margins in compressional and extensional environments and a whole variety of nuanced cases of spontaneous initiation. Numerical modelling of the subduction process plays an important, though somewhat bewildering role in discussion, as do considerations of the forces likely to be at play. Applying theoretical considerations to actual examples from the geological record are sublimely enlivening, as are speculations about the future evolution of the passive margins of the Atlantic. Clearly, there is a healthy future for field and mathematical study on the processes at destructive plate margins, such as building in the aspects of magmagenesis. Since Stern has built his career on study of long dead collusions zones, products of arc accretion etcetera, development of their understanding is undoubtedly the main thrust of his and Gerya’s tour de force. Stern provides a full PDF at his University of Texas website for the benefit of anyone who wants to delve deeper than space at Earth-pages and my limited intellect permit!

A fully revised edition of Steve Drury’s book Stepping Stones: The Making of Our Home World can now be downloaded as a free eBook