Andrew Y. Glikson

Early Precambrian tonalite-trondhjemite sialic nuclei

Abstract Early Precambrian batholiths evolved by diapiric intrusion of near-liquidus to superheated tonalitic and trondhjemitic magmas into an early greenstones crust. Distribution patterns of enclaves and xenolith screens derived from the latter provide reference markers which define the internal geometry and detailed structure of the “gregarious batholiths” (Macgregor) as polydomal multi-lobal bodies. Near-liquidus temperatures are suggested by the digestion of vast volumes of ultramafic—mafic crust by the acid magmas. Tracing of xenolith trains between low and high grade metamorphic terrains provide key evidence for coeval relations between granite—greenstone type terrains and amphibolite to granulite facies infracrustal root zones of the latter. The formation of the plutonic tonalite—trondhjemite suite was accompanied by dacitic to rhyolitic extrusions, the acid volcanic lenses being located above early greenstone units intruded by the batholiths and below upper greenstone sequences which postdate these intrusions. The geochemical characteristics of high-level and deep-level tonalites and trondhjemites are compared. Both suites display very wide compositional spectra, but data from high-grade terrains tend to define a more basic field than data from granite—greenstone terrains. Effects of source compositions on the geochemistry of the acid plutonic rocks are pointed out. Tonalites dominate in South African terrains whereas trondhjemites dominate in Western Australian terrains — a difference conceivably related to the more ultramafic composition of source rocks represented by early greenstone units in southern Africa. Granodiorites and potassic granites form a comparatively minor component of Archaean batholiths, and may occur in the following forms: (1) bands of augen gneiss in high-grade terrains; (2) components of trondhjemitic to granitic gneisses in high-level plutons; and (3) discrete post-tectonic intrusions typically emplaced at high levels of the batholiths and along older tonalite—greenstone contacts. Migmatites characteristically form in close spatial association with xenolith-rich zones, probably due to depression of the solidus consequent on water addition related to dehydration of the xenoliths. A derivation of the acid sodic magmas by anatexis of sialic materials is inconsistent with geochemical evidence and petrological theory. In contrast, the commonly low to very low LIL element levels and REE evidence indicate derivation by about 30–50% melting of basic rocks. Marked trace element anomalies are characteristic of some Archaean plutonic suites, e.g. very high Sr in some Western Australian rocks, low Rb in some Lewisian (Scotland) and South African rocks, U depletion in South African and southwestern Greenland suites, high Li in some Pilbara rocks and high Zr in some southwestern Greenland rocks. However, the only consistent anomaly observed to date is a well-pronounced depletion in Y and heavy REE, suggesting extensive equilibration of the acid melts with eclogite and/or amphibolite. Uniformitarian interpretations of the Archaean are questioned in the light of the evidence for high temperature and pressure, the unique tectonic style of diapirism and the low initial87Sr/86Sr as compared to Proterozoic plutonic suites. The diachronous nucleation of tonalite—trondhjemite plutons during the Archaean is seen as the major process effecting a transformation of an early Archaean sima into sial.

Early Precambrian Evidence of a Primitive Ocean Crust and Island Nuclei of Sodic Granite

The primordial crust consisted of a stratiform oceanic-type ultrabasic-basic assemblage, relicts of which are retained at the base of Archaean sequences of eastern Transvaal, Rhodesia, and Western Australia. Commonly isochemical metamorphism renders the chemistry of well-retained segments of the meta-igneous rocks significant to their original compositions. Archaean tholeiitic metabasalts comply with the principal criteria for oceanic tholeiites, but tend to have higher Fe/Mg and Mn, and lower Al, Ti, K/Rb, and Fe+ 3 /Fe+ 2 than average recent oceanic tholeiites. These geochemical characteristics can be interpreted in terms of either shallow-level differentiation and (or) a mantle relatively depleted in certain siderophile and transition elements, which possibly indicates a lesser degree of core segregation in the Archean. Metabasalts occurring within basic to acid calc-alkaline volcanic cycles at higher stratigraphic levels than the ultrabasic-basic oceanic assemblages are chemically distinct from metabasalts incorporated in the latter; these rocks have higher K, Sr, Zr, and Y values than recent oceanic tholeiites, and may have originated from high degrees of melting of an underlying oceanic substratum. A model is suggested whereby the evolution from an oceanic crust to greenstone belts proceeded through the rippling of the oceanic crust and the subsidence of linear zones; partial melting below the troughs resulted in the emergence of sodic granites and in cyclic calc-alkaline volcanism. High Ni/Mg and Cr/Mg ratios in the earliest sodic granites support their derivation from the oceanic crust. The granites display cyclic decrease in Na/K, K/Rb, and Sr/ Rb with time, trends which reflect crustal thickening. The plutonism resulted in the superposition of sialic nuclei on early linear structural trends, and in a consequent formation of spatially discrete volcanic-sedimentary depositories, or greenstone belts, in which sequences show an Alpine-type trend of evolution from ophiolites into turbidites and late-stage conglomerates. Notwithstanding this similarity, important differences between Archaean greenstone belts and Alpine or island-arc belts are indicated. Significant similarities exist between this model and the evolutionary pattern of the Fijian archipelago.

Vertical zonation and petrogenesis of the early Precambrian crust in Western Australia

Abstract Variations in metamorphic grade, structural style, isotopic ages and granite geochemistry observed within the Yilgarn craton, and between the Yilgarn and Pilbara cratons, Western Australia, are interpreted in terms of vertical zonation of the Archaean crust. We correlate the gneiss-granulite suite of the Wheat Belt (southwestern Yilgarn) with concealed coeval infracrustal roots of the low-grade granite—greenstone Kalgoorlie terrain (eastern Yilgarn). Differences between the Pilbara, Southern Cross and Laverton granite—greenstone blocks and the downfaulted linear greenstone belts of the Kalgoorlie block are interpreted in terms of deeper-level exposure in the first three blocks. Ultramafic—mafic volcanic sequences in the Yilgarn craton can be divided into at least two major groups — the lower greenstones, regarded as relicts of a once extensive simatic crust, and the significantly younger upper greenstones, believed to have formed within linear troughs following the intrusion of Na-rich granites. At least three major Archaean granite phases occur in Western Australia: (1) 3.1-2.9 b.y. old (recognized to date only in the western Yilgarn and in the Pilbara craton); (2) 2.8-2.7 b.y. old, and (3) 2.6 b.y. old (the two latter phases can only be separated from each other in the eastern Yilgarn, and phase (3) is also identified in the Pilbara). In the main, granites of phases (1) and (2) are Na-rich and those of phase (3) are K-rich. There is evidence for a secular increase in Rb levels and initial 87Sr/86Sr ratios. It is suggested that the K-rich granites grade down into Na-rich granites, and the former were generated by ensialic anatexis resulting in upward migration of K, Rb, U, and Th-enriched magmas. A review of data from several Archaean cratons in other continents suggests that evidence from these regions can be interpreted in terms of the general model of crustal evolution proposed for Western Australia. Implications of this model concerning petrogenesis of Archaean plutonic and volcanic suites, geothermal gradients and tectonic evolution of greenstone belts are discussed. Partial melting associated with mantle diapirism is thought to have given rise to the ultramafic—mafic volcanic cycles. Widespread subsidence and partial melting of this crust yielded Na-rich acid magmas. The development of the upper greenstones was confined to linear belts in a partly cratonized crustal environment. About 2.6 b.y. ago a rise in the geothermal gradient resulted in regional metamorphism and crusctal anatexis which gave rise to the K-rich granites.

Trace element geochemistry and origin of early Precambrian acid igneous series, Barberton Mountain Land, Transvaal

Abstract Sixteen new major and trace element analyses of granitic and acid volcanic rocks from the Barberton Mountain Land, eastern Transvaal, are reported. The data indicate very low abundances of LIL elements in albite porphyries incorporated in the lower ultramafic to mafic succession of the volcanic Onverwacht Group, and in the ‘ancient tonalites’ plutonic suite which intruded the volcanics at about 3.2–3.4 aeons. Some of the albite porphyries display highly fractionated REE patterns, which indicate equilibration with garnet and therefore high-pressure origin. The REE patterns of the ‘ancient tonalites’ are characterized by pronounced positive Eu anomalies, and their derivation by partial melting of eclogite within the depth range of 30–50 km is suggested. Some of the albite porphyries may represent shallow-level hypabyssal equivalents of the ‘ancient tonalites’. The Dalmein and Bosmanskop plutons are regarded as late members of the ‘ancient tonalites’ suite, derived by decreasing partial melting of basic source rocks. A contrasting model of origin is suggested for the ca . 3.0 aeons old Nelspruit migmatite and Hood granite, which are thought to have been derived by anatexis of the ‘ancient tonalites’ and incorporated ultramafic to mafic xenoliths of the Onverwacht Group. The chemistry of greywackes of the Fig Tree Group, which overlie the Onverwacht Group, is consistent with the erosion of a terrain consisting of Onverwacht Group and ‘ancient tonalites’—including late K-rich members of the latter suite. The data lend support to models involving a secular transformation from an ensimatic to an ensialic tectonic environment, but important geochemical differences are indicated between the Archaean rocks and acid igneous suites of island arcs and oceanic domains.

The astronomical connection of terrestrial evolution: crustal effects of post-3.8 Ga mega-impact clusters and evidence for major 3.2?0.1 Ga bombardment of the Earth-Moon system

Abstract It is accepted that during the Late Heavy Bombardment (LHB—4.2–3.8 Ga) in the Solar system, as preserved on the Moon, the terrestrial upper mantle-crust system was dominated by interactions between internal mantle processes and extraterrestrial impacts, and that following this period, the impact flux decreased by two orders of magnitude, from 4–9×10 −13 km −2 year −1 to 3.8–6.3×10 −15 km −2 year −1 (for asteroids Dc >=18 km), a rate consistent with a cratering rate of 5.9±3.5×10 −15 km −2 year −1 estimated for near-Earth asteroids (NEA) and comets. Geology being a geocentric science, the assumption is generally made that from about 3.8 Ga the impact factor can be neglected in the context of models of crustal evolution, including the emergence of early continental nuclei and plate tectonics. This paradigm is questioned in this paper. From the observed minimum number of 6 continental impact structures with Dc >=100 km [Vredefort (300 km); Sudbury (250 km); Chicxulub (170 km); Woodleigh (120 km); Manicouagan (100 km); Popigai (100 km)], assuming an Earth surface occupied by time-integrated >=80% ocean crust since 3.8 Ga, the lower limit of post-LHB impacts is deduced at >=30 craters with Dc >=100 km and >=10 craters with Dc >=250 km. From the lunar crater counts and the present-day asteroid flux the impact incidence was likely to have been higher by an order of magnitude, with a possible decline in the impact frequency of the largest bodies Dp >=20 km. Evidence for maria-scale impact basins in the Archaean emerges from 3.24 Ga-old impact vapor condensation-fallout layers in the Barberton greenstone belt, Transvaal, pointing to multiple oceanic impact basin of Dc >=400 km. This impact cluster falls within error from the c. 3.18 Ga impact peak documented by lunar spherules—suggesting a mid-Archaean impact cataclysm in the Earth–Moon system. Models of crustal evolution need to account for the inevitable magmatic and tectonic consequences of these events, particularly on impacted geothermally active oceanic crust and lithosphere. A combination of the internal heat engine and the impact factors is capable of accounting for the early sialic nuclei, for the spatial and temporal localization of major faulting and rifting events, and for ensuing plate tectonic patterns.

Woodleigh, Carnarvon Basin, Western Australia: a new 120 km diameter impact structure

Abstract The Woodleigh multi-ring structure, buried by Cretaceous and, at its centre, Lower Jurassic lacustrine sediments, east of Hamelin Pool, Carnarvon Basin, Western Australia, is identified as an impact structure, the largest discovered to date on the Australian continent. An impact origin is indicated by: a central core of uplifted granitoid basement probably less than 25 km in diameter, which displays shock-induced planar deformation features in quartz, pervasive diaplectic vitrification of feldspar and penetrative pseudotachylite veining; and an inner ring syncline containing a ∼70 m thick thermally modified diamictite overlain by ∼380 m of Lower Jurassic lacustrine deposits. An outermost diameter of 120 km, defined by gravity, magnetic and surface drainage, indicates a ring fault that sharply intersects the NS-striking regional structure. At the centre of the basement uplift shock metamorphosed granitoid was intersected at a depth of 171 m, at least 1800 m higher than the gravity-modelled level of regional basement. Pseudotachylite vein systems within the shocked granitoid are strongly enriched in Al, Ca, Mg, Ni, Co, Cr, V and S, and depleted in K and Si, suggesting chemical fractionation attendant on shock volatilisation, enrichment by an injected and volatilised meteoritic component, and potentially of sulfide mineralisation. The impact age is constrained by overlying Lower Jurassic strata, reworked Early Permian palynomorphs in the Jurassic lacustrine section, and deformed Lower Devonian and older units. A regional thermal event identified by apatite fission track at 280–250 Ma hints at a possible Permian–Triassic boundary age for the impact, although the lack of Triassic fossils in the crater fill favours a late Triassic age.

Oceanic mega-impacts and crustal evolution

Lunar mare crater counts, the terrestrial impact flux, and astronomical observations of asteroids and comets define a consistent impact rate of 4–6 ṁ 10 −15 km −2 ṁ yr −1 within the inner solar system since the end of the late heavy bombardment ∼3.8 Ga. Coupled with the observed crater size vs. cumulative crater size frequency relationship of N ∝ D c −1.8 ( N = cumulative number of craters of diameter > D c ), these rates imply formation on Earth of more than 450 D c ≥ 100-km-diameter craters, more than 50 D c ≥ 300-km-diameter craters, and more than 20 D c ≥ 500-km-diameter craters. Geochemical and isotopic constraints require that more than 80% of the projectiles impacted on time-integrated oceanic crust since the late heavy bombardment. The injection of shock energies calculated at >10 8 Mt TNT equivalent by a D p >10-km-diameter projectile may result in propagating fractures and rift networks, thermal perturbations, and ensuing magmatic activity. Examinations of the geologic record for correlated impact and magmatic fingerprints of such events remain inconclusive in view of isotopic age uncertainties. Potential but unproven connections may be represented by the (1) Cretaceous-Tertiary boundary (ca. 65 Ma) impact(s), onset of the Carlsberg Ridge spreading, Deccan volcanism, and onset of the mantle plume of the Emperor-Hawaii chain; (2) Jurassic-Cretaceous boundary (ca. 145 Ma) impacts, onset of Gondwana breakup, including precursors of the East African rift structures; (3) Permian-Triassic boundary (ca. 251 Ma) impact(s), Siberian Norilsk traps, and Early Triassic rifting; and (4) the 3.26 Ga basal Fig Tree Group (east Transvaal) Ir-rich and Ni-rich quench spinel-bearing impact spherules and contemporaneous igneous-tectonic activity. Tests of the theory require further identification and isotopic dating of distal ejecta, impact spherule condensates, and meteoritic geochemical anomalies.

Use of geochemistry as a guide to platinum group element potential of mafic-ultramafic rocks: examples from the west Pilbara Block and Halls Creek Mobile Zone, Western Australia

Integrated chemical and isotope studies of late Archaean mafic-ultramafic layered complexes and associated extrusive volcanic rocks in the west Pilbara Block, Western Australia, suggest that they were derived from komatiite magmas through processes of crustal assimilation followed by fractional crystallization. On the basis of Nd isotope data the contaminants are believed to be felsic basement rocks of mainly early Archaean age. Parent magmas of these layered complexes are siliceous high magnesian basalts (SHMB) and can be regionally correlated with two types of komatiites. The Cooya Pooya Dolerite and the Negri Volcanics could be related to group 1 komatiite, characterized by near chondritic Al2O3/TiO2 (∼21), Ti/Sc (∼74) and flat heavy rare earth element (REE) patterns, whereas the Munni Munni, Mt. Sholl, Dingo, Maitland and Andover complexes could be related to the group 2 komatiites characterized by depletion in Al (Al2O3/TiO2∼10), heavy REE and Sc. Komatiites are products of very large degrees (e.g. 30–50%) partial melting of the convective mantle. They commonly have Pd∼8−10 ppb, Au ∼ ppb, Pt 10−14 ppb ppb and are strongly undersaturated in sulfur (∼ 500 ppb?). Archaean SHMB commonly have higher Pd(∼ 15 ppb), but are still S-undersaturated (< 1000 ppm) unless contaminated by S-rich crustal rocks. Komatiites and SHMB commonly have Ti/Pd (∼2−3 × 105), and Pd/Cu (∼−3 × −4) similar to estimates of the primitive mantle. The west Pilbara SHMB have PGE contents (Pd ∼10−15 ppb) similar to the estimated parent magma of the Bushveld complex, and other late Archaean SHMB. Some late Archaean major layered complexes and early Proterozoic dykes in the Yilgarn Block, Western Australia arc inferred to have similar values. These magmas are believed to have potential for PGE mineralization through sulfide saturation processes in magma chambers and feeder dykes. Early Proterozoic mafic-ultramafic intrusives in the Halls Creek Mobile Zone, have parent magmas with a chemical affinity to tholeiites generated in continental rift environments. Among them, parent magmas of the Woodward Dolerite, Lamboo Complex, and Panton Sill have high inferred abundances of Pd (up to 15 ppb), Pt and Au. Abundances of some incompatible elements (Ti, Zr, Y and P) in these rocks are lower than in mid-ocean ridge basalts suggesting magma generation by large degrees (>/25%) of mantle melting. Their primitive parent magmas, when first scaparated from the mantle, are inferred to have been S-undersaturated and to have had Pd, Pt and Au contents similar to the Archaean SHMB. With the exception of the Woodward Dolerite, they were generally saturated with sulfur prior to emplacement resulting in occurrences of minor basal chromitite layers with PGE-enriched sulfides generated during injection of new pulses of Saturated magma into the chamber (e.g. Panton Sill). Such rocks are therefore not believed to be favourable hosts for Merensky Reef style PGE mineralization. Feeder dykes of these intrusives and the Woodward Dolerite may have potential for mineralization similar to the NiCuCuPGE sulfide deposits in feeder dykes of Permian-Triassic continental flood basalts in west Siberia. The key factor for PGE potential in mafic-ultramafic magmas is S-undersaturation which can be achieved either by high temperature (> 1400°C) and ⩾10% partial melting under high pressure (⩾ 25kb) and/or large degrees (⩾25%) melting of MORB type mantle source under low pressures. Their mantle sources appear to have fairly constant Pd (∼4 ppd), Pt (∼6 ppb), Au (∼1 ppb) and S (∼250 ppb) from Archaean to the present. S-undersaturation may also be achieved by smaller degrees of meltung of residual, refractory mantle poor in Ca, Al, Ti, Fe and S ⪡ 250 ppm) (with Pd, Pt ⪡ 4 ppb, Au ⩽ 1 ppb) of the continental lithosphere or wedge above a subduction zone. Rocks of the first association include komatiites, continental flood basalts and picrites, which are believed by some to be related to mantle plume activity. These PGE rich magmas could be further enhanced by melting of the refractory lithospheric mantle, whereas the second type of rocks include boninites, low-Ti tholeiites, some lamprophytes and shoshonites.

Geosynclinal evolution and geochemical affinities of early Precambrian systems

Abstract A type-section of the Archaean Kalgoorlie System, Western Australia, displays a trend of evolution from the eugeosynclinal ophiolite stage into the turbidite-deposition stage, which grades into a molasse-like conglomerate stage representing the termination of geosynclinal deposition. The increase in the energy of sedimentation went hand in hand with a weakening of the associated igneous activity. A transition from predominantly ultrabasic and basic rocks into intermediate and acid rocks with stratigraphic level can be demonstrated. Similar trends of evolution have been observed in Archaean sequences in Canada, South Africa, and India. Primary textural, mineralogical, and geochemical features have been extensively retained in rocks of the Kalgoorlie System, as well as in rocks of the Canadian and South African Archaean. The commonly isochemical nature of the low-grade regional metamorphism which affected these systems is indicated by the consistent geochemical characteristics of the metabasalts, as well as by the systematic variations shown by the volcanic sequences. Archaean meta-igneous suites are almost invariably calc-alkaline to calcic. The meta-basalts are very low in K2O, have moderate to low alumina contents and low Fe 2 O 3 FeO ratios, and are comparable with tholeiitic basalts of oceanic domains. High-alumina basalts may be present, whereas spilites and alkaline basalts are very rare. Archaean porphyries in Western Australia are predominantly high in soda and low in potash, although potassic varieties may be present. The geochemical trends shown by Archaean sequences reflect an evolution from a thin primordial oceanic crust to a geosynclinal pile. The crustal subsidence associated with this development resulted in magma generation at progressively greater depth and pressure. This trend is represented by the predominance of oceanic tholeiites at low stratigraphic levels, and the occurrence of high-alumina basalts and alkaline volcanics at intermediate and high stratigraphic levels, respectively, reported from Canada. The absence of spilites could be interpreted in terms of a low soda concentration of the primordial oceans.

K–Ar evidence from illitic clays of a Late Devonian age for the 120 km diameter Woodleigh impact structure, Southern Carnarvon Basin, Western Australia

Abstract Woodleigh is a recently discovered impact structure with a diameter of 120 km, and thereby represents the third largest proven Phanerozoic impact structure known after Morocweng and Chicxulub. K–Ar isotopic studies of fine-grained authigenic illitic clay minerals (