Geological relationships, relative garnet Lu-Hf chronology, absolute PT calculations and spatial database of metamorphic parameters across all domains, constrain the metamorphic evolution of the Eastern Goldfields Superterrane (EGST). The post-2720 Ma thermo-barometric evolution of the EGST is divided into 5 broad periods of contrasting metamorphic style, conditions and possible thermo-mechanical/tectonic setting. Unlike stress fields, thermal event switching involves slow rates, conductive delays and a history less punctuated than the structural evolution; necessitating broad thermal events and continuums between them, particularly from M2 to M3a and M3b (see below).

The period of late crust formation and volcanic stratigraphy involved both multiple events and a spatially complex mosaic of subduction, magmatic arcs, back-arc and plume related environments and processes. Widespread near surface metamorphic parageneses that would have formed at this time are sea floor alteration (e.g. Phillips. 1986). These early alteration assemblages would have been largely obliterated during subsequent events, except possibly in the lowest grade parts of the EGST. Nevertheless, the resulting altered bulk compositions are preserved and these have strongly influenced the subsequent developed metamorphic parageneses in the EYC (e.g. Purvis, 1984).

Granulite metamorphism is recognised in only two localities: a cluster in the northern part of the Southern Duketon Domain and one location in the northern part of Gindalbie Domain. These granulites are statically metamorphosed gabbros, preserving ghosted ophitic textures, recrystallised to granoblastic two pyroxene assemblages. Peak metamorphic conditions of ~730 ºC at low pressures between 2.5-5.0 kbar, indicate very high average thermal gradients (G – simply the ratio T/depth assuming a density of 2.8 gm/cm3) between 45-80 ºC/km. These high-T, high-G conditions in the upper-crust are most possibly formed in a high heat flow magmatic arc environment (Ma). Such a scenario would result in small high-grade domains, widely scattered and of different ages, reflecting multiple magmatic arcs. In contrast, formation in either an extensional or plume setting would result in widespread distribution.

[TOP]

Rare, but widely distributed medium-P (>6 kbar), high-T (>600 ºC) and uniquely low average thermal gradient (G ≤20 ºC/km) metamorphic parageneses are found almost exclusively associated with high-strain shear zones (e.g. Ida, Ockerburry, Celia and Hootanui Fault systems). A single Lu-Hf garnet age from one of these high-grade parageneses is 15 Ma older than Lu-Hf ages for garnets from regional metamorphic parageneses and 30-35 Ma older than Lu-Hf ages for metamorphic garnet in a post-volcanic late basin (see below). Consequently, these medium-P high-T parageneses are early formed, prior to pervasive regional metamorphic events. This is substantiated by overprinting of medium-P rocks within shear zones by lower grade shear assemblages and the formation of these rocks in thermal regimes entirely distinct from all younger events (i.e. within late basins) and over-printing (i.e. regional metamorphism and low grade shear events), all of which are characterized by high average thermal gradients (G ≥40 ºC/km). Peak-T was attained at peak-P and was followed by near isothermal decompression, either in the same metamorphic cycle or during a later exhumation event.

The tectonic setting for early, medium-P parageneses is unknown, but can be explained by two competing end-member processes. (1) Either these medium-P, low-G conditions were widespread regional metamorphism and the current restricted distributions are due to late-stage exhumation bringing only a portion into the shallow crustal (i.e. by M3a extensional telescoping). This scenario is supported by exposures within shear zones but nevertheless would predict more widespread relict medium-P parageneses, such as in garnet cores, boudins and xenoliths. (2) Or alternatively, medium-P, low-G conditions were spatially restricted at the time of formation. In which case the current distribution is significant and the tectonic setting would be responsible for discrete, thin, arcuate zones of deep burial followed by rapid exhumation. Possible tectonic settings that match the metamorphic conditions and the apparent restricted spatial distribution (i.e. alternative 2) are shallow and/or slow subduction (typical subduction has G <15ºC/km) or downward advection of greenstone margins by diapiric overturn (e.g. Collins & Van Kranendonk, 1999).

Currently, both the tectonic setting and relative timing of M1 and Ma metamorphic parageneses are unknown. Though both metamorphic events have entirely distinct thermal regimes, it is possible that both formed at similar times in different parts of a tectonic system. For example, both thermal regimes are developed in subduction / magmatic arc pairs; Ma metamorphic conditions at shallow-crustal levels in magmatic arcs and M1 metamorphic conditions at deeper-crustal levels in an associated subduction zone setting.

[TOP]

Lu-Hf garnet ages are consistent with stratigraphy; and suggest that regional metamorphic, main phase matrix parageneses in greenstones grew 17-23 Ma before assemblages within turbiditic post-volcanic late basins (PVLB). The regional M2 thermal anomaly possibly continued over a protracted period and into continuum with the thermal event associated with M3a metamorphism of the PVLB’s. Outside of the PVLB’s, it is locally difficult to distinguish the M3a thermal over-print from regional M2 metamorphic parageneses. M2 parageneses range in temperatures from sub-greenschist to mid-amphibolite facies, 300-550 ºC. The pressure range is limited from 3.5 kbar at low grades to 4.0 kbar at highest grades, with elevated average thermal gradients of 30-40 ºC/km. Metamorphic temperatures increase toward granite/gneiss domes (Mikucki & Roberts, 2003) and the moderately elevated thermal gradients are consistent with a “regional-contact” type of metamorphism dominated by conduction and heat supplied by large volumes of (65%) High-Ca granitoid. M2 metamorphic mineral growth accompanied deformation conditions ranging widely from essentially no strain to syn-kinematic well-developed fabrics. M2 metamorphism broadly overlaps in age with the heterogeneously distributed D2 compressional deformation. Ridley (1993) recognised that the universally low pressure and elevated average thermal gradient, mitigate against widespread crustal over-thickening, such as in a collisional orogen as has been previously suggested for the Yilgarn Craton (Groves & Phillips, 1987). The spatial association of M2 field gradients and distribution of High-Ca granites of broadly similar age, infer that metamorphism was in a thermally inversely stratified crust with elevated thermal gradients in the upper crust due to neutral buoyancy pooling of large volumes of granite.

Metamorphism of turbiditic post-volcanic late basin (PVLB), such as Mt Belches, is distinctive, well constrained and characteristic of extensional settings. Peak metamorphism was at 500-580 ºC and 4±1 kbar, indicating high average thermal gradients (G) between 40 and 50 ºC/km. Peak metamorphism accompanied a static stress regime, subsequent to formation of the early burial, bedding-parallel mica foliation. P-T paths are tight, heating followed by cooling paths. Most have up-P prograde-burial trajectories, from early cordierite-andalusite, to peak garnet and post-peak staurolite growth, indicating anticlockwise P-T loops. Recognition of a post-2665 Ma high-heat flow regime accompanying turbiditic basinal infill, has implications for driving basinal brine circulation, and release of large volumes of hydrous fluid by dewatering and dehydration reactions.

The PVLB’s are a crucial time marker that “samples” the PT conditions of the thermal regime in operation between 2665-2650 Ma. Metamorphic parageneses in all older stratigraphy (volcanics) could potentially be a combination of both M2 and M3a metamorphic mineral growth – depending on the extent of M3a thermal effects. There are two possibilities for the extent of the M3a thermal overprint. (1) Either a widespread regional thermal overprint, in which case the EGST metamorphic pattern is a combination of both M2-M3a effects. (2) Or alternatively, M3a thermal effects are restricted to spaced elongate domains of high heat flow associated with maximum rifting, coinciding with exposed and eroded PVLB’s (Fig. 1). Outlining the extent of M3a rifting and thermal overprint is hampered by there being only remnants of PVLB’s and no other widespread stratigraphic marker between M2 and M3a parageneses. If exhumation of medium-P M1 parageneses and old (>2735 Ma) stratigraphic units occurred as a result of M3a extensional telescoping, these rock units will indicate the lower-plate margins to zones of extreme M3a extension. M3a lithospheric extension would be locally controlled by dip-slip reactivation of growth faults that control development of PVLB depositional troughs, some of which may have had earlier histories controlling syn-volcanic late basins (SVLB). Metamorphic patterns are consistent with extensional telescoping of the crust during M3a, with transport away from lower-plate domal cores, stepping outwards to successively younger extensional structures. These geometries result in exhumation of panels of deeper crustal levels and juxtaposition against shallow crustal level rocks, as movements on different extensional structures migrate across this margin. The presence of old (>2735 Ma) stratigraphic units and medium-P M1 parageneses indicate the extended margins of lower-plate domain and PVLB’s occupy upper-plate settings.

Crucially, metamorphism of the PVLB’s by high average thermal gradients (40-50 ºC/km) at 10-12 km depth is diagnostic of lithospheric extensional settings. Other extensional basins such as the Kanmantoo trough and Irindina Province and thermal modelling show that basin fill, burial and metamorphism can all occur in short, 2-10 Ma, time frames. Consequently, M3a metamorphism is interpreted to have essentially accompanied ongoing PVLB fill, with metamorphism peaking somewhere between 2665 and 2650 Ma. The thermal peak recorded in the PVLB may have been more widespread but obscured by the pre-existing M2 parageneses that formed at very similar conditions. If PVLB’s occupied elongate, growth-fault controlled settings coincident with maximum extension, the zones of highest M3a high heat will also be spaced elongate domains mirroring the distribution of exposed and eroded PVLB’s. The ultimate driver for increased M3a lithospheric extension is necessarily speculative, but does post-date volcanism and may be related to cessation of down-going subduction and consequent gravity driven slab roll back.

[TOP]

The M3b period, subsequent to metamorphism of the PVLB’s, is characterised by a switch from dip-slip extensional kinematics to multiple strike-slip kinematic episodes (Blewett & Czarnota, 2005) through a protracted period of high but waning thermal gradients and multiple Au-mineralization events. Metamorphic mineral growth was within hydrothermal alteration assemblages that over-print regional metamorphic parageneses (M2-M3a) and was strongly partitioned into restricted domains and is associated with veining. Hydrothermal alteration associated with main phase Au-mineralization (~2650-2640 Ma) occurred across a wide range of temperatures 250-500 ºC but similar crustal depths of 3.0-3.5 kbar, across most parts of the EGST, indicating moderately elevated thermal gradients of 30-50 ºC/km. This represents crustal levels only 3 km shallower than peak metamorphism during M2 and M3a. Multiple fluid generations and fluid types, ranging from H2O-rich to CO2±CH4-rich are evident in different alteration systems. Alteration assemblages vary on scales too small to have experienced significant differences in T and P and are primarily controlled by pH and redox (e.g. Mikucki, 1997; John Walsh pers. comm.). Later episodes of alteration and Au-mineralization (~2630-2620 Ma) occurred at pressures as low as 1 kbar, while still at relatively elevated alteration temperatures (250-350 ºC) indicating very high average thermal gradients (G >70ºC/km). This event coincides with a secondary peak of shallow crustal, Low-Ca granites, which possibly supplied the heat necessary to drive fluid circulation and gold remobilisation.

Multiple mineral growth events throughout M3b, define a broad continuum of elevated average thermal gradient that persisted from M3a. This long-lived high thermal budget is due to a combination of factors, such as conductive thermal lag from M3a, continuing lithospheric extension or transtension and emplacement of significant (35%) volumes of Low-Ca granites. Low-Ca crustal melts were initiated during M3a lithospheric extension, followed by pooling, emplacement and crystallization in the upper crust ~10 Ma later, representing a causal link between M3a and M3b thermal periods. Similarly, M3a fluid production by dewatering of PVLB brines and dehydration of hydrous minerals in both PVLB sediments and low-grade regional (M2) metamorphic domains sourced the large volumes of hydrous fluid responsible for much of M3b alteration. This causal link is supported by stable isotope and fluid inclusion salinity data indicating meteroric fluid that has undergone exchange in post-volcanic sedimentary basins. (e.g. Mikucki, 1997). The large ambient fluid volumes produced were possibly important in setting up permeability pathways and alteration systems where mixing with secondary, metal-rich, fluids could occur. A third causal link between M3a and M3b is the formation of (1) mantle-derived magmas (Mafic granites and Syenites) and (2) pathways for dry mantle fluids in response to M3a lithospheric extension; both being possible sources for bringing Au into the upper crust.

The strong causal relationships between M3a and M3b intimately links Au-mineralization with the thermal evolution of the crust. This period follows on from a significant switch in lithospheric response to far field tectonic processes: from crustal growth (subduction-arc accretion?) to lithospheric extension. Lithospheric extension was possibly first initiated at ~2685 Ma with the formation of growth fault controlled syn-volcanic late basins (SVLB), steep increase in the volume of High-Ca granitoids and regionally extensive elevated thermal gradient (M2). Coincident with the termination of volcanism, accelerated lithospheric extension from ~2665 Ma generated the M3a thermal perturbation and PVLB rifts. Lithospheric extension may have generated mantle magmas that both transported Au and possibly set up conduits for dry metal-rich fluids to bring Au into the upper crust. M3a rifting generated both the turbiditic PVLB’s that brought large hydrous fluid reservoirs down into the upper crust and resulted in the high thermal gradients that permitted both dehydration and dewatering of juvenile and pre-existing low-grade rocks, and the thermal energy to drive fluid circulation. Modelling indicates that lithospheric extension is conducive to downward fluid flow trajectories, further facilitating movement of basinal brine fluids down into the upper-crust (Sheldon et al., This Volume). The thermally weakened and thinned crust, with a structural architecture established by M2 and M3a growth faults, was conducive to strike-slip and thrust reactivation in appropriate stress fields (D4 to D6), generating dynamic fluid pumping, permeability pathways and deposition sites for Au-mineralization. By comparison to Phanerozoic earth history, a plausible first-order cause for the switch from continental growth to lithospheric thinning is sudden plate reconfiguration caused by the final assembly of a super-continent (e.g. Squire & Miller, 2003). Termination of super-continent assembly may have resulted in slab roll back and lithospheric thinning at ~2665 Ma with similar lagging causal effects resulting in Au-mineralization events from ~2650-2620 Ma in other Neoarchaean terranes worldwide.

Thanks to the sponsors for their ongoing commitment to the pmd*CRC Y4 project. Special thanks to GSWA for access to legacy thin sections, databases, vehicle for the fieldwork and ongoing support for this project. John Walsh, Martin Van Kranendonk, John Miller, Heather Sheldon, Ned Stolz, Kevin Cassidy and Joe Ogierman are thanked for the very helpful discussions. Angus Netting, Ian Pontifex, Liz Webber, Greg Hitchen and Mike Verrall are thanked for help with lapidary, XRD and mineral analyses. Published with permission of the CEO’s of the pmd*CRC, GSWA and GA.