Figure 11. Evolutionary diagrams for Isua and Pilbara ultramafic rocks. Ultramafic rocks from both terranes can be interpreted via similar hot stagnant-lid tectonic models. Ultramafic rocks are initially cumulates formed during cooling of magmas in hot stagnant-lid settings that feature voluminous volcanism. These cumulates were then variably deformed and/or metamorphosed during tectonic events that either represent (1) shortening, corresponding to volcanic burial, plate-breaking or plate tectonic subduction (panel a); or (2) intra-crustal diapirism corresponding to gravitational instability (panel b). Later, mostly static (talc/carbonate/serpentine) alterations further modified the petrology and geochemistry.
ConclusionsSome ultramafic rocks preserved in or near the Isua supracrustal belt have been interpreted as tectonically emplaced mantle peridotites that require >3.7 Ga onset of plate tectonics (e.g., Nutman et al., 2020; Van de Löcht et al., 2018). In contrast, this study shows that: (1) the polygonal rock textures of Isua ultramafic samples can also be observed in Pilbara ultramafic rocks which show rock textures of crustal cumulates; (2) the whole-rock major element, trace element and HSE patterns of Isua ultramafic rocks are similar to those of Pilbara ultramafic rocks and/or crustal cumulates; (3) the co-existence of Ti-humite, magnesite, serpentine, olivine, clinopyroxene and perhaps talc may be compatible with crustal conditions; (4) the olivine oxygen isotopic signatures of Isua ultramafic rocks can be explained by mantle-derived or metamorphic fluid fluxing in a hot stagnant-lid setting or a plate tectonic subduction setting; (5) the CPO inferred B-type olivine fabrics are consistent with crustal cumulates; and (6) the spinel geochemistry of Isua ultramafic rocks is only compatible with crustal cumulates. In summary, many petrological and geochemical aspects (e.g., rock and mineral textures, Ti-humite phases, and normalized HSE patterns) of phaneritic ultramafic rocks in early Earth terranes on Earth could be explained in the contexts of tectonically-emplaced mantle slices atop of crustal rocks, but are also consistent with crustal cumulates (cf. Nutman et al., 2021). In contrast, other characteristics of these rocks, such as certain types of spinel geochemistry (e.g., Fe-Ti trends in Cr#-Mg# space, Barnes and Roeder, 2001) as well as cumulate textures, appear to be unique to cumulates. Thus, we conclude that no features preserved in ultramafic rocks of the Isua supracrustal belt and East Pilbara Terrane are diagnostic of plate tectonic-related mantle slices, but instead are compatible with crustal cumulates. We argue that differences between ultramafic rocks within two terranes only reflect contrasting metamorphism, deformation, and/or alteration conditions experienced by these rocks, not necessarily different protoliths (cf. Friend and Nutman, 2011; Nutman et al., 2020). Again, it is important to note that these interpretations do not exclude plate tectonic origins for the formation of the Isua supracrustal belt (e.g., Van Kranendonk, 2010; Nutman et al., 2020), but they permit a hot stagnant-lid tectonic origin for this terrane, consistent with previous studies for the belt (Ramírez-Salazar et al., 2021; Webb et al., 2020; Zuo et al., 2021). Therefore, because the East Pilbara Terrane (e.g., Collins et al., 1998; Van Kranendonk et al., 2007) can also be explained in terms of a hot stagnant-lid setting, no tectonic shift between the Eoarchean and Paleoarchean is required. Short episodes of local plate tectonic processes during the Eo- and Paleoarchean might be possible, as regional stagnant-lid processes may have coexisted with local plate tectonic processes in early terrestrial planets (e.g., Van Kranendonk, 2010; Yin, 2012a; Yin, 2012b). Nonetheless, our findings show that a ≤3.2 Ga initiation of plate tectonics is viable.