Over hundreds of millions – if not billions – of years, plate tectonics has sculpted the Earth’s crust and modulated its geodynamics, from the mantle to the atmosphere. This process recycles crust and as such allows for the long-term biogeochemical cycles that support the global biosphere. Nevertheless, the onset of plate tectonics on Earth remains open to debate, with proposed inception ranging from Neoproterozoic times (1,000-542 million years ago) to as far back as the Hadean eon (before about 4,000 million years ago). The oldest rocks, from before 3,600 million years ago, are strongly deformed and recrystallised; direct observations of early Earth’s magmatism and, thus, of the geodynamic regime operating at that time, remain scarce.
Two such Eoarchean (4,000-3,600 million years ago) terranes include the Nuvvuagittuq and Ukaliq supracrustal belts (Figure 1), both located in northern Québec (Canada). These ancient lithologies, locked in Eoarchean to Paleoarchean (3,600-3,200 million years ago) granitoid gneisses of the Northeastern Superior Province, mostly comprise metamorphosed mafic, ultramafic, and sedimentary rocks. Previous studies measuring neodymium isotopes showed that some of these Eoarchean mafic rocks preserve a signature that could be interpreted as being derived from a 4,400-million-year-old crustal source.
As our understanding of the Archean geology has slowly improved in the few last decades, little emphasis has hitherto been placed on direct analysis of rocks that could tell us about the nature of melt differentiation which led to the formation of early continents. It therefore makes sense to interrogate such ancient records of the early Archean world from studies of ancient rocks in places such as the Québec outcrops.
A few years ago, Pierre Bouilhol and Guillaume Caro (University of Lorraine, France) followed by Stephen J. Mojzsis (now at Research Centre for Astronomy and Earth Sciences, Budapest, Hungary) introduced me to such an exciting topic for my Master’s thesis project. Since that time, my goal has been to establish the petrogenesis of the mafic and ultramafic rocks of the Ukaliq locality using careful thin section observations coupled with bulk rock and mineral major and trace element chemistry. From a textural and chemical perspective, the main findings include the surprising result that melt-cumulate relationships are preserved between the mafic and ultramafic rocks forming a damp and a wet series, and the remarkable preservation of original Eoarchean minerals within these cumulates.
Petrology in its broadest sense, that is, phase relationships and bulk rock and in situ chemistry, is widely used to infer magma crystallisation paths and pression-temperature conditions recorded by metamorphic rocks. Following this simple concept, we identified several mineralogical assemblages within the Ukaliq locality leading us to recognise two main series among the mafic and ultramafic rocks and to identify the metamorphic conditions experienced by these rocks. The latter is well-described by phase equilibria of a metasedimentary rock which returned conditions of 600-625°C and 4-5 kbar. This result explains the typical metamorphic assemblages of the mafic and ultramafic rocks. However, on the one hand, some ultramafic samples contain tiny clinopyroxenes included in amphibole porphyroblasts whereas, on the other, large plates of orthopyroxene phenocrysts occur and are reminiscent of a crystal accumulation texture. Such a dichotomy also appears when looking at the bulk rock chemistry for both mafic and ultramafic rocks. The first series, which includes the clinopyroxene-bearing ultramafic samples, comprises moderate to high Al2O3/TiO2 ratios and chondrite normalised flat rare earth element spectra which are usually interpreted as a tholeiitic signature (Figure 2). Conversely, the second series, encompassing the orthopyroxene-bearing ultramafic rocks, has low Al2O3/TiO2 ratios, and U-shaped, depleted rare earth element spectra compared to the tholeiitic series. These chemical signatures are commonly observed in forearc lavas and cumulates and are inferred to have been boninites. We therefore recognise a tholeiitic and a boninitic series of mafic-ultramafic rocks which correspond, based on compatible (e.g., nickel and chromium) and rare earth element abundances, to a melt-cumulate relationship. Additionally, the texture and the chemistry of the orthopyroxenes, comparable to what is observed in modern arc lower crust, and probably the clinopyroxenes, suggest that they directly crystallised from an Eoarchean melt.
Our work also modelled the melt evolution of these two suites considering the two most primitive melts and the crystallising assemblages inferred from observations and bulk rock compositions (Figure 2). This led to the elucidation of a typical damp tholeiitic sequence dominated by clinopyroxene and plagioclase, and a boninitic suite crystallising orthopyroxene, plagioclase, and amphibole. We then coupled these findings with a previous study showing that some mafic rocks, and especially the boninitic melts, have preserved a 142/144-neodymium isotope anomaly which coincides with the damp and wet character of the tholeiitic and boninitic suites, respectively. At this point, we just need the mantle source of the boninitic melts to be wet, while the primary tholeiitic melts remain damp, but how? Regardless of the geodynamic regime, water must be recycled into the mantle. Based on our observations, it appeared to us that the easiest way to form damp tholeiitic and wet boninitic melts requires a horizontal component into the downgoing lithospheric portion, similar to subduction. This would allow the formation of a mantle corner flow leading to decompression melting of tholeiitic liquids, followed by the fluid intake of the descending lithosphere enabling fluid-assisted melting of a probably depleted mantle and, therefore, the genesis of boninitic melts. Alternatively, the vertical delamination of an overthickened crust could also produce such signatures, but it does not provide a direct pathway to producing usually-low-pressure boninitic melts. The main conclusion of our work is that Ukaliq rocks preserve geochemical signatures and 3,800-million-year-old crystals indicative of the type of magmatism operating then. Now that we know such rocks are preserved from these early times, it makes sense to expand our studies to reveal even more about how and when plate tectonics came to be.
Please sign in or register for FREE
If you are a registered user on Nature Portfolio Earth and Environment Community, please sign in