March 2014 LIP of the Month

Mafic-ultramafic intrusions of the Giles Event, Western Australia: Prospectivity for magmatic ore deposits

by

WD Maier1, HM Howard2, RH Smithies2

1School of Earth and Ocean Sciences, Cardiff University, UK
2Geological Survey of Western Australia, Perth

Introduction

The Musgrave Province of central Australia contains one of the largest concentrations of mafic-ultramafic layered intrusions on Earth (Fig. 1). These are referred to as the Giles Complex (Daniels, 1974) or the Giles intrusions (Smithies et al., 2009). The intrusions were emplaced at variable crustal depths (Goode and Moore, 1975; Glickson et al., 1995; Ballhaus and Glickson, 1995), during a c.3 Ma period from c.1078 to 1075 Ma (Howard et al., 2011b). These intrusions, together with earlier and later bimodal magmatism of the Bentley Supergroup and the basic magmas of the Warakurna Large Igneous Province (including the Alcura Dolerite Suite) form the components of the Warakurna Supersuite, formed during the c. 1090 to 1040 Ma Giles Event.


Figure 1: Simplified geological map of the Musgrave Province, highlighting mafic-ultramafic intrusions.

Many aspects of the Giles intrusions remain poorly understood, notably their potential to host magmatic ore deposits. This is partly due to poor exposure, but is also because much of the region forms part of the Ngaanyatjarra-Anangu Pitjantjatjara–Yankunytjatjara Central Reserve, established in the late 1970’s, into which access is strictly regulated. Much of the work documented here arose through a regional-scale geological mapping project jointly co-ordinated through the Ngaanyatjarra Council and the Geological Survey of Western Australia (GSWA), and guided by the local indigenous people in the region between the communities of Wingellina and Warburton.

During the course of the regional mapping program, the Giles intrusions were remapped on a scale of 1:100 000, and are portrayed on the Bell Rock, Holt, Blackstone, Finlayson, and Cooper 1:100 000 scale map sheets, covering a total area of 120 000 km2 (Howard et al. 2006a, 2007, 2011b; Evins et al. 2009, 2010a; Smithies et al. 2009). More than 450 geochemical samples of the Giles intrusions were analysed and most of the data are available from the WACHEM database at http://geochem.dmp.wa.gov.au./geochem/. A detailed discussion of the results is available in Maier et al. (2014).

The large volume of mafic igneous rocks and the enormous size of some of the intrusions (up to several 1000 km2) reflect a significant flux of mantle magma and heat into the crust. Such events tend to be favourable for the formation of a range of ore deposits, including magmatic Cu-Ni-PGE, Cr, Fe, V, Ti, apatite, lateritic Ni, as well as hydrothermal deposits. As yet, two world-class deposits have been discovered, namely the Nebo Babel magmatic Ni-Cu deposit (Seat et al., 2007, 2009) and the Wingellina Ni laterite deposit (Metals X Ltd; http://metalsx.com.au/operations/wingellina_nickel/>), in addition to several other, smaller deposits.

Regional geology

The Musgrave Province is a Mesoproterozoic belt tectonically bound by the Neoproterozoic to Palaeozoic sedimentary rocks of the Amadeus Basin in the north and the Officer Basin in the south (Edgoose et al., 2004). The Province is expressed on geophysical images as a series of east-trending anomalies covering an area up to 800 km long and 350 km wide that straddles the borders between the Northern Territory, Western Australia and South Australia (Fig. 2). The term Musgrave Province is used here to refer to all the high-grade metamorphic rocks affected by the Mesoproterozoic Musgrave Orogeny.


Figure 2: Combined gravity and TMI (total magnetic intensity – grey scale) image of the west Musgrave Province. Grey scales range from high (red) to low (blue).

Main tectonomagmatic events

The basement to the west Musgrave Province remains poorly known. It can only be studied through isotopic data on the detrital components in paragneisses, and on zircon xenocrysts. A recent study of Hf isotopes in zircons from magmatic and sedimentary rocks throughout the Musgrave Province (Kirkland et al., 2013) indicates that the unexposed basement is dominated by two major juvenile crust formation events, one at 1600–1550 Ma and a possibly more significant event at 1950–1900 Ma.

Outcrop in the west Musgrave Province consists largely of granites formed during several Mesoproterozoic events. The oldest of these is a recently identified, but un-named, event that involved intrusion and possible extrusion of felsic calc-alkaline magmas of the Papulankutja Supersuite and contemporaneous redistribution of this material into local sedimentary basins at c. 1400 Ma (Howard et al., 2011b; Kirkland, et al., 2013). The oldest clearly recognizable event in the west Musgrave Province is the Mount West Orogeny. During this period, calc-alkaline granites of the Wankanki Supersuite were emplaced mainly within the central and southeastern part of the Province (Smithies et al., 2009; Evins et al., 2009). Crystallization ages range from c. 1345 to c. 1293 Ma (Gray, 1971; Sun et al., 1996; White et al., 1999; Bodorkos and Wingate, 2008; Smithies et al., 2009), but cluster between c. 1326 and 1312 Ma. Rocks of the Wankanki Supersuite are typically metaluminous, calcic to calc-alkaline granodiorites and monzogranites with strong compositional similarities to the Phanerozoic granites of the Andean continental-arc (Smithies et al., 2010). The Mount West Orogeny may reflect final subduction and accretion during the amalgamation of the North, West and South Australian Cratons (Giles et al., 2004; Betts and Giles, 2006; Smithies et al., 2009; 2010; Kirkland et al., 2013).

The 1220-1150 Ma Musgrave Orogeny involved mylonitic deformation and widespread granulite-facies crustal reworking and is generally regarded as essentially intracratonic (Wade et al., 2008; Smithies et al., 2009; 2010), although Smithies et al. (2013) suggest that it may alternatively reflect an ultra-hot orogen initiated upon the back arc region of the Mount West Orogeny. Orthopyroxene-bearing (charnockitic) and locally rapakivi-textured granites of the Pitjantjatjara Supersuite intruded the mid-crust more or less continuously throughout the Musgrave Orogeny. They are ferroan and typically alkali-calcic granites with significant REE- and HFSE- enrichments. In all areas, the earliest Pitjantjatjara intrusions are strongly Yb-depleted granites formed through deep-crustal melting under garnet-present conditions. A transition from these to Yb-undepleted granites formed at lower melting pressures is diachronous, migrating from the northeast to the southwest of the province from c. 1220 to c. 1200 Ma. This transition has been attributed to progressive removal of the lower crust and lithosphere, previously thickened during the Mount West Orogeny (Smithies et al., 2010, 2011).

The Pitjantjatjara granites were emplaced at temperatures up to 1000 ºC (Smithies et al., 2010, 2011) and intrusion coincided with a 70 - 100 m.y. period of regional ultrahigh-temperature (UHT) metamorphism (Kelsey et al., 2009, 2010; Smithies et al., 2010, 2011), characterised by temperatures in the lower to mid crust of > 1000°C, along a geothermal gradient of ≥35–40°C km-1 (Kelsey et al., 2009, 2010). These thermal conditions are consistent with complete removal of the lithospheric mantle during the early Musgrave Orogeny.

Voluminous mafic to felsic magmas were intruded into and extruded onto the west Musgrave Province during the c. 1090 – 1040 Ma Giles Event (Fig. 3). All igneous rocks that formed during the Giles Event are grouped into the Warakurna Supersuite. These include the giant layered mafic-ultramafic ‘Giles intrusions’(G1), massive gabbro (G2) locally mixed and mingled with granite, various dyke suites including the Alcurra Dolerite suite, granite plutons, as well as mafic and felsic lavas. The latter are grouped into the Bentley Supergroup, which also includes volcaniclastic and sedimentary rocks.


Figure 3: Time–space plot showing GSWA SHRIMP (U–Pb) zircon ages from the west Musgrave Province (after Smithies et al., 2010).

The outcrop extent of the Bentley Supergroup defines the preserved extent of the Bentley Basin (Fig. 4). This basin can be subdivided into several sub-basins in Western Australia, including the Talbot Sub-Basin, in the area west of Jameson Community, and the smaller Blackstone Sub-Basin (Tollu Group), in the area south of Blackstone Community. The lithological range, lithological associations and the distribution of the Warakurna Supersuite and the Bentley Supergroup, and their geological history, is consistent with a long-lived intracontinental rift setting, referred to as the Ngaayatjarra Rift (Evins et al., 2010b; Aitken et al., 2013). Intrusive rocks of the Warakurna Supersuite are particularly abundant within the Mamutjarra and Tjuni Purlka Zone, where they typically occur as tectonically dismembered bodies of layered mafic-ultramafic ‘Giles intrusions (G1)’, ‘massive gabbro (G2)’ and granite.


Figure 4: Interpreted bedrock geology map of the west Musgrave Province.

The Giles Event has been interpreted as the result of a mantle plume, based on the observation that mafic rocks of the Warakurna Supersuite crop out across approximately 1.5 million km2 of central and western Australia, forming the Warakurna Large Igneous Province (Wingate et al., 2004; Morris and Pirajno, 2005). However, mantle-derived magmatism lasted more or less continuously for at least 50 m.y. suggesting that the Giles Event reflects a more protracted and complex geodynamic setting inconsistent with a simple plume model (Smithies et al., 2009, 2013; Evins et al., 2010), as will be discussed in more detail in a later section.

Younger events include the 580-530 Ma intracratonic Petermann Orogeny, which coincides with the global Pan-African plate reorganization that marks the assembly of Gondwana. Granulites and high-grade gneisses of the Musgrave Province were thrust northwards, over or into (interleaved with) rocks of the Neoproterozoic basins (Camacho, 1997; Flöttmann and Hand, 1999; Edgoose et al., 2004), and many of the Giles intrusions were fragmented. The Alice Springs Orogeny was a further major intraplate event, or series of events, that affected much of central Australia from c. 450 to 300 Ma (Collins and Teyssier, 1989; Haines et al., 2001), but there is no firm evidence that the orogeny had any significant effect on the Musgrave region. Additional younger events are indicated through the presence of regional dolerite dyke suites (at c. 1000 Ma, c. 825 Ma and c. 750 Ma) and low-volume felsic magmatism (at c. 995 Ma and c. 625 Ma).

Tectonic subdivisions of the west Musgrave Province

The Musgrave Province has been sub-divided into several distinct zones with different structural and metamorphic characteristics separated by major west- and west-northwesterly trending faults that were last active during the Petermann Orogeny (Camacho, 1989). These faults include the south-dipping Woodroffe Thrust (Fig. 1) that separates the amphibolite facies Mulga Park Domain in the north from the granulite facies Fregon Domain in the south, the latter comprising most of the west Musgrave Province. In the eastern part of the west Musgrave Province, the Fregon Domain shows a marked north to south change in the pressure of granulite-facies metamorphism. To the north, high-pressure (10 to 14 kbar) metamorphism during the Petermann Orogeny has masked the effects of Mesoproterozoic metamorphism (Scrimgeour and Close, 1999). To the south, where metamorphic overprints of Petermann Orogeny age are not as marked, evidence for Mesoproterozoic high-temperature metamorphism, at much lower pressures, is preserved (Clarke et al., 1995). In the west Musgrave Province, the boundary separating these two metamorphic styles lies close to the west-trending and near vertical Mann Fault (Fig. 1). The western part of the Fregon Domain is subdivided into the Walpa Pulka Zone, the Tjuni Purlka Zone, and the Mamutjarra Zone (from northeast to southwest; Smithies, et al., 2009; 2010)(Fig. 4). The Tjuni Purlka Zone is a broad northwest-trending zone of multi-generational (c. 1220, 1075, and 550 Ma) shearing. The extent and intensity of northwest-trending shearing in this zone exceeds that of neighbouring zones. The margins of the Tjuni Purlka Zone, in particular, have formed the focus for felsic and mafic magmas of the Warakurna Supersuite (Fig. 4). These tectonic boundaries were syn-magmatic shear zones throughout much of the Giles Event. The effects of the Petermann Orogeny in this zone are most intense to the northeast, near the contact with the Walpa Pulka Zone, and decrease to the southwest. Further to the north, the Walpa Pulka Zone (Fig. 4) is a deep-crustal domain dominated by c. 1220-1150 Ma high-K granite plutons of the Pitjantjatjara Supersuite emplaced during the Musgrave Orogeny. The zone contains high-pressure metamorphic assemblages preserved by rapid exhumation along east- and northwest-trending mylonitic and migmatitic shear zones related to the Petermann Orogeny (Scrimgeour and Close, 1999; Camacho et al., 1997; Raimondo et al., 2009, 2010). The Mamutjarra Zone, south of the Tjuni Purlka Zone (Fig. 4), is dominated by Giles intrusions and the c. 1345 to 1293 Ma calc-alkaline granites of the Wankanki Supersuite, formed during the Mount West Orogeny. The effects of the Petermann Orogeny in this zone are minimal.

Geology and petrology of mafic and mafic-ultramafic intrusions of the Giles Event

Introduction

The Giles intrusions comprise a range of lithologies, including (i) predominantly ultramafic, (ii) interlayered mafic-ultramafic, and (iii) predominantly mafic bodies. Intrusions with important ultramafic segments include Wingellina Hills, Pirntiri Mulari, The Wart (Fig. 1) and, in South Australia, Ewarara, Claude Hills, Gosse Pile, and Kalka. The ultramafic rocks comprise mainly wehrlite, harzburgite, websterite, and (olivine)orthopyroxenite, with less abundant dunite. Predominantly mafic intrusions in the west Musgrave Province include Hinckley Range, Michael Hills, Latitude Hill, Murray Range, Morgan Range, Cavenagh, Saturn, Blackstone, Jameson, Finlayson, Bell Rock, and several smaller intrusions or fragments of intrusions to the north of the Tjuni Purlka Zone, including the Mount Muir and Lehmann Hills intrusions (Note that the Jameson, Finlayson, Blackstone and Bell Rock group of intrusions are now believed to be tectonically segmented portions of an originally single body, hereafter referred to as the Mantamaru intrusion). These mafic intrusions are predominantly of leucogabbronoritic composition, with > 50 vol. % plagioclase and variable proportions (often clustering around unity) of orthopyroxene to clinopyroxene. Several intrusions contain thick troctolitic sequences where olivine and plagioclase are the main cumulus phases and pyroxenes are either interstitial or form minor cumulus minerals. These intrusions include Cavenagh, Morgan Range, and Mantamaru. Anorthosites occur as relatively thin layers (centimeters – several decimeters) in the “troctolitic” and many gabbronoritic intrusions, but may attain several 10s of meters at Kalka in South Australia. Massif-type anorthosite intrusions do not occur in the west Musgrave Province, but one such body has been delineated in South Australia (Teizi; Gray, 1967). In the following section, we will provide an overview of the geology and petrography of some of the studied intrusions. For a more comprehensive account, the reader is referred to Maier et al. (2014).

Mafic-ultramafic layered intrusions

Pirntiri Mulari intrusion

The width of this wedge shaped intrusion is ~ 5km, and its stratigraphic thickness is ~ 3km. The rocks consist mostly of medium-grained (olivine)websterite, peridotite, (olivine)orthopyroxenite, and (olivine)gabbronorite. Dunite is rare, but this may partly reflect poor preservation and exposure of these often highly altered rocks. The websterites are concentrated at the base and the top of the intrusion where they are interlayered with gabbronorites. Peridotites and orthopyroxenites are concentrated in the centre. The intrusion is slightly overturned. The basal contact of the intrusion is not exposed, but likely located not far beyond the exposed southwestern edge. The top contact of the intrusion is likely of a tectonic nature, as suggested by the mylonitization of the uppermost rocks and by their relatively unevolved chemical composition (Mg# of 0.7, Cr/V >1). In the lower portion of the intrusion, textural evidence suggests a complex intrusive history and substantial syn-magmatic cumulate mobility. Gabbroic layers contain abundant ultramafic and anorthositic schlieren (Fig. 6a+b). The schlieren horizons are concentrated both below and above pyroxenite layers, with the latter also containing gabbroic schlieren. PGE concentrations are mostly < 10 ppb, and the Cu/Pd ratio is mostly around primitive mantle levels (4000-7000), suggesting that the magma was S undersaturated. At a level of ~ 2600m above the base of the intrusion Cu/Pd ratios increase significantly suggesting that the magma reached sulfide saturation at this stage, consistent with elevated Cu contents of almost 500 ppm Cu, but low PGE. The stratigraphic position of this sample in the mafic-ultramafic transition interval is analogous to that of the PGE reefs in the Bushveld Complex and in many other PGE mineralized intrusions.


Figure 5: The lower-central ultramafic portions of the Pirntiri Mulari intrusion


Figure 6: Textures in Pirntiri Mulari (a+b) and Cavenagh intrusions (c+d)

Wingellina Hills intrusion

This intrusion extends for ~12 km along a strike of 110-120° and is up to 3km wide. The exposed stratigraphic thickness amounts to 2.5 km, barring tectonic duplications of stratigraphy. The base of weathering varies considerably depending on rock type and degree of shearing. The outcropping gabbros and pyroxenites tend to be unaltered at surface, but in the peridotites the depth of weathering varies from about 60 m to more than 200 m in shearzones. The basal contact of the intrusion is likely intrusive, as suggested by the fact that the basal pyroxenite shows a well defined basal compositional reversal that is typical of the basal contacts of layered intrusions. Through the next 300m, the basal olivine gabbronorite and pyroxenite are overlain by progressively more magnesian harzburgite and then peridotite and wehrlite. The wehrlite is overlain by ~ 20m of pyroxenite and > 40m of olivine gabbronorite forming the upper portion of the first cyclic unit of the intrusion. Within the pyroxenite occurs a PGE reef that has been traced along a strike of 2-3 km. The remainder of the intrusion consists of interlayered peridotite, wehrlite, pyroxenite and olivine gabbronorite. Megarhythmic layering occurs on a scale of 10s of meters, but centimeter-scale layering may also occur (Ballhaus and Glickson, 1989).

The PGE reef is hosted in websterite which overlies wehrlite with a sharp contact, reflected by a decrease in MgO and an increase in Cr concentrations (Fig. 7). PGE concentrations show an initial peak of almost 100 ppb at, or just below the wehrlite-websterite contact, and then increase through the websterite layer to a point about 5-7 meters beneath the top of the websterite, where concentrations reach up to 2 ppm Pt+Pd+Au over a 1-meter interval, and >1 ppm PGE over a 3-5 meter interval. Above the reef, PGE concentrations fall rapidly across a few meters of stratigraphy. Compared to the Bushveld reefs (~1-2 % S), the Wingellina Hills reefs are relatively S poor, having mostly < 500 ppm S. The reefs are thus invisible in hand specimen, posing an added challenge to exploration. Notably, the total PGE concentrations of the Wingellina Hills reef interval (at least in holes WPRCD0-064 and WPRCD0-083) are similar to those in the combined Merensky Reef and UG2 chromitite of the Bushveld Complex, but the Bushveld reefs are much narrower, i.e., the PGE are more concentrated. Further details on the PGE mineralisation are given in Maier et al. (2014).


Figure 7: Log of diamond drill core WPRCD0-083, Wingellina Hills intrusion (web = websterite, wehr = wehrlite).

The Wingellina Hills laterite comprises a yellow-brown to dark brown ochre material which is composed of goethite, manganese oxides, gibbsite and kaolinite derived from the weathering of dunite and peridotite. This forms the world-class Wingellina laterite Ni deposit, discovered by INCO in 1956. The laterites formed by selective leaching of SiO2 and MgO and the resulting residual concentration of alumina, Fe-oxides and Ni, which is especially pronounced along shear zones. The lateritic material is locally cut by semi-precious, pale green chrysoprase. The lateritic ore is exposed at surface, with an average thickness of 80 meters and a maximum of up to 200 meters. The deposit also has a high aspect ratio and therefore a very low strip ratio.

The Wart intrusion

This is a relatively small mafic-ultramafic body. Its main preserved portion measures ~ 5x2 km (Fig 4). It has been suggested that it may form the lower portion of the Bell Rock intrusion (Ballhaus and Glickson, 1995). The stratigraphic thickness of the intrusion is on the order of 1-2 km. It shares certain lithological and compositional features with the Pirntiri Mullari intrusion. For example, both contain layers of medium-grained mesocumulate wehrlite-peridotite within clinopyroxenite and melagabbronorite adcumulates and mesocumulates. Many ultramafic units have sharp contacts and may represent sills (Ballhaus and Glickson, 1995). Both intrusions contain microgabbro layers or sills. The mineral compositions mostly overlap with those of the Pirntiri Mulari intrusion, as previously noted by Ballhaus and Glickson (1995). However, plagioclase is significantly more calcic than in the Pirntiri Mulari intrusion.

Predominantly mafic intrusions

Latitude Hill-Michael Hills intrusions

The Latitude Hill intrusion is located 5-10 km to the east of The Wart and Bell Rock intrusions (Fig. 4). The body has been interpreted as a folded segment of the Michael Hills gabbro (Ballhaus and Glickson, 1995) forming one of thickest intrusions related to the Giles Event, at ~8000m. The present work indicates that deformation is broadly syn-magmatic, based on mingling textures between gabbro and coeval deformed and undeformed granite (Howard et al., 2011). The intrusion contains numerous layers and lenses of olivine gabbronorite, olivine pyroxenite and rare peridotite, but it is unclear if these all form one body or fragments of several distinct intrusions (Glickson et al., 1995). It also remains uncertain in which direction the layers dip. Abundant fine-grained gabbros were interpreted to represent intraplutonic quench zones (Ballhaus and Glickson, 1995).

Morgan Range intrusion

The Morgan Range intrusion is located within the Tjuni Purlka Zone, some 10 km to the north of Blackstone community (Fig. 4). It is a ~8 km x 3 km (ca 15 km2) body with a stratigraphic thickness in excess of 1 km. Most of the rocks are relatively unaltered olivine gabbronorites and troctolites that show modal layering on a scale of centimeters to meters. The intrusion forms a boat-shaped synclinal structure whose margins dip steeply (up to 80°) to the interior of the massif, whereas in the centre the dip is sub-horizontal. The intrusion is not overturned, and plunges at a relatively shallow angle to the southeast. At the northeastern tip of the body occurs a small (~ 300 x 300 m) interlayered mafic-ultramafic lens, containing dunite, troctolite, and melagabbronorite. The nature of the contact with the main Morgan Range intrusion is either concealed by regolith or highly altered, but is likely a fault. The rocks of the lens strike ~ 100° and dip steeply (80°) to the north. It is presently unclear whether this segment forms part of the Morgan Range intrusion, or whether it is a fragment of a different intrusion that was tectonically attached to the Morgan Range intrusion. The mafic-ultramafic lens shows some compositional similarities to the Wingellina Hills and Pirntiri Mullari intrusions.

Hinckley Range intrusion

This is a large (~30x10 km), strongly deformed, mostly poorly layered body. It has a stratigraphic thickness of  ~5800 meters. The rocks are mostly (olivine)gabbros, troctolites and microgabbros, with some layers or lenses of anorthosite, and minor pyroxenite. The overprint of the Petermann Orogeny is evident in abundant pseudotachylite veins. Many of our samples have relatively high concentrations of K2O and incompatible trace elements, indicative of assimilation of, or mixing with, a granitic component. To the west, the Hinckley Range is in direct contact with the younger West Hinckley Range intrusion. Here, well-developed mingling textures are observed between gabbro and granite related to the Giles Event. The rocks at this locality can locally be described as agmatite or injection migmatite (Fig. 8). In some cases, gabbro forms angular blocks engulfed by granite, or brittle fractures in gabbro are infilled by granite veins. Such relationships indicate that the granite has intruded a largely solidified gabbro. In other cases, contacts between the felsic and mafic components are cuspate indicating that the mafic component was ductile in the presence of the felsic component and that the granitic magmas intruded gabbro that was only partially solidified. Some 20 km to the northwest, at Amy Giles Hill, contacts between the same two mafic and felsic components are typically cuspate, with tear-shaped protrusions of gabbro in granite and of granite in gabbro – reflecting two co-existing liquids.


Figure 8: Mingling textures in the Hinckley intrusion 

In the West Hinckley Range intrusion, locally mingled gabbro forms a km-scale fold with a steep northwest-trending axial plane that has been intruded by syn-deformational leucogranite. A strong ‘gneissic’ fabric has locally formed in mixed or agmatitic rocks as the axial planar fabric continued to develop and this has been again engulfed within subsequent injections of leucogranite. A sample of undeformed leucogranite from within one of these axial planar zones, approximately 2 km to the south of the mingled gabbro-granite described above, yielded a crystallization age of 1075 ± 7 Ma. Syn-mylonitic leucogranite has also pooled into boudin necks in a northwest-trending mylonite immediately south of Charnockite Flats (approximately 2.5 km to the northwest of the West Hinckley Range intrusion), and has been dated at 1075 ± 2 Ma. These data effectively define a very narrow period of intrusion of massive (G2) gabbro and multi-phase intrusion of leucogranites (1078-1074 Ma), northwest-directed folding, and northwest-trending shearing. The relationships between gabbro and granite in the Hinckley Range intrusion confirm earlier suggestions by Clarke et al. (1995b) that substantial deformation occurred during the Giles Event.

Murray Range intrusion

This is a large gabbroic body comprising a > 25 km2 layered portion and extensive segments of massive gabbro, to the east and northeast of the Pirntiri Mulari intrusion (Fig. 4). The body underwent significant tectonic dismemberment due to its location at the tectonic contact between the Tjuni Purlka and Walpa Pulka Zones. As a result, the true stratigraphic thickness and structure of the intrusion remain uncertain. The Murray Range intrusion is also one of the G1 bodies that (like the Hinckley Range intrusion) was pervasively intruded by G2 gabbro at its margins, consistent with the idea that at the stage of G2 emplacement, the contact between the Tjuni Purlka and Walpa Pulka Zones was a syn-magmatic shear zone (Evins et al., 2010). The intrusion also contains abundant stratiform microgabbro layers and lenses (Fig. 9), and medium grained gabbronorites that are commonly cross bedded (Fig.9). Deformation and alteration are usually slightly more pronounced than in the other intrusions.


Figure 9: Layering in the Lehman Hills (b), Murray (a+c) and Jameson intrusions (d).

Cavenagh intrusion

Because access to much of the Cavenagh intrusion is restricted on cultural grounds, this body has remained one of the least studied amongst the Giles intrusions. The Cavenagh intrusion is situated to the south of the Blackstone intrusion (Fig. 4, 10). It can be sub-divided into two main portions. The southern portion, around Mount Morphett, forms a syncline measuring ~ 15 km along a strike of ~100°, with a width of ~7 km, and a stratigraphic thickness of ~ 1 km. The rocks here are predominantly olivine gabbronorites, olivine gabbros, troctolites, and norites. Websterites, anorthosites and microgabbros form bands, discontinuous pods, schlieren, and autoliths (Fig. 6). A major east-trending fault separates the Mount Morphett segment from the northern portion of the Cavenagh intrusion. The latter measures ~ 10x20 km on the surface, strikes at 100° and dips at ~15-30° to the northeast. The stratigraphic thickness of this segment is ~2-4 km (assuming no structural duplication). It consists mostly of relatively homogenous olivine gabbronorite, troctolite and magnetite-bearing olivine gabbronorite. In terms of geochemistry, the southern to central portion of the Cavenagh intrusion is the most primitive. In the northern portion of the intrusion there is a subtle trend towards more differentiated compositions with height, expressed, for example, by decreasing Cr/V ratio and olivine-Fo content to the north-east.


Figure 10:

Fig. 10: The Cavenagh intrusion, view towards south

The microgabbros are conspicuously interlayered with medium-grained gabbronorite, with sharp or gradational contacts. In places, microgabbro has been injected by medium-grained gabbro, but elsewhere the microgabbros may contain autoliths of anorthosite and thin bands, irregular clasts and circular concretions of granular websterite and clinopyroxenite adcumulate. These field relationships indicate that the microgabbros and associated medium-grained mafic-ultramafic rocks intruded broadly contemporaneously. Olivine and pyroxenite grains may form strings, sometimes oriented in a radial configuration, that are interpreted to have resulted from growth of the crystals in a flowing, supercooled magma (Fig. 6). Microgabbros in the Wingellina Hills intrusion and other Giles intrusions were previously explained by intraplutonic quenching of new magma influxes (Ballhaus and Glickson, 1989). The association of microgabbro with fragments and schlieren of pyroxenite suggests that the emplacement of the microgabbro was associated with, and possibly led to, slumping of semi-consolidated cumulates in the chamber, resulting in sorting of crystals to form pyroxenite and anorthosite (cf Maier et al., 2013b). The combined observations from the Cavenagh intrusion thus indicate a dynamic, semi-consolidated magma chamber that was frequently replenished by unevolved magma.

Lehmann Hills, Mount Muir, and other small intrusions to the north of the Blackstone and Wingellina communities

The Lehman Hills intrusion forms several low hills scattered over ~ 2 km2, some 40 km to the north of Blackstone community (Fig. 4). The rocks consist mostly of medium-grained olivine gabbronorites and, in places, interlayered pyroxenite and anorthosite (Fig. 9). These rocks may show a distinct flow structure, containing elongated and aligned lenses and schlieren of anorthosite in pyroxenite and fragments and clots of pyroxenite within anorthosite. The Mount Muir intrusion is a small (3x1 km) mafic intrusion, or an isolated tectonic fragment of a larger body, 25 km to the northwest of the Lehman Hills (Fig. 4). The rocks are well layered olivine gabbronorites and troctolites. Layering occurs at the 1-3 centimeter scale, strikes at ~100°, and dips steeply (80-90°) to the north. Several additional small mafic bodies have been mapped and sampled up to 10 km to the north of Mount Muir. The rocks are mostly gabbronorites and less commonly olivine gabbronorites. We also sampled several small mafic bodies up to 20 km to the north of the Hinckley Range intrusion, in the Mount Gosse - Mount Daisy Bates area. These mafic bodies are metagabbros and metagabbronorites. They commonly have a partial granoblastic texture and garnet forms fine-grained rims on pyroxene and magnetite grains or more rarely forms garnet porphyroblasts. Pyroxene is commonly replaced by hornblende and biotite is also common.

Jameson-Finlayson intrusion

The combined Jameson-Finlayson intrusion extends for 66 km along a strike of ~ 120° and is ~ 30 km wide (Fig. 4). Layering is not overturned and dips at ~20° to 30° to the southwest, indicating a stratigraphic thickness of up to ~10 km. Daniels (1974) recognised four zones within the intrusion. From the base to the top, these are glomeroporphyritic gabbro, banded lherzolite and magnetite-ilmenite-bearing lherzolite, rhythmically layered troctolite and olivine gabbronorite, and a layered sequence of troctolite, olivine gabbro and olivine gabbronorite, with at least 11 major titaniferous magnetite bands. The olivine gabbro and olivine gabbronorite in the upper part of the intrusion contain relatively abundant sulfides, in places > 1 vol.%. These form the only commonly sulfide-bearing units within the layered mafic-ultramafic G1 Giles intrusions, and some have Cu concentrations up to 860 ppm (at 0.12 wt % SO3). This is likely related to the magma reaching S saturation in response to protracted fractionation.

All magnetite layers sampled are of massive Fe-oxide, containing <5% silicates, and have relatively coarse grain sizes (up to 3 mm). These features possibly reflect sintering. The mineralogy of the seams consists essentially of magnetite, granular ilmenite, fine ilmenite lamellae, abundant hematite replacement patches and lamellae (up to 20 vol.%), as well as goethite. No significantly apatite-enriched layers were encountered, but this may reflect poor outcrop. The most reliable observations and interpretations can be made on the basal magnetite seam. It dips ~ 20-30° to the southwest and strikes 100°. It has been traced along strike for about 19 km as an aeromagnetic anomaly with sporadic broken outcrop. It may reach a thickness of 50 meter and up to three sub-seams are developed locally. It is unclear whether this is due to primary magmatic processes or structural duplication. The immediate footwall and hanging wall consist of magnetite-bearing leucotroctolite and anorthosite, showing evidence for deformation. The broader footwall sequence is characterized by pronounced interlayering between magnetite gabbronorite and anorthosite.

Peak PGE concentrations in the basal seam are 2 ppm Pt+Pd+Au, significantly higher than in the Bushveld Main Magnetite Layer, but lower than in magnetite layers from, for example, the Stella intrusion in South Africa (Maier et al., 2003b). The layer has relatively constant V concentrations (up to 7400 ppm V, 1.35wt % V2O5), but PGE concentrations tend to be markedly enriched at the base. Sulfur concentrations are mostly 100-150 ppm, locally reaching 700 ppm. The average Pd/Ir ratio is 34, consistent with a magmatic origin of the PGE mineralization. The Pt/Pd ratio is >1, analogous to other examples of PGE mineralization in magnetitites or magnetite gabbros (see Maier, 2005, and references therein). Phosphorous concentrations are mostly up to ~100 ppm, and Fe concentrations reach 55 wt. %.

The compositional variation through the remainder of the magnetite seams remains somewhat poorly defined, but that the concentrations of vanadium and chalcophile elements (PGE, Cu, Au) decrease sharply in the upper seams, whereas Fe, Cr, and P concentrations increase. A more detailed description of the seams can be found in Maier et al., (2014).

Bell Rock intrusion

This intrusion extends for ~50 km along a strike of ~120°. The exposed width is ~5-6 km, and the rocks dip at 70° to the southwest. Field exposures of graded layers and cross bedding, as well as compositional data indicate younging to the southwest. The exposed stratigraphic thickness is ~ 3800 m, but since most contacts are unexposed this is a minimum estimate. The top of the intrusion is interpreted to be in contact with volcanic rocks of the Bentley Supergroup, thus either the intrusion has been deeply eroded after emplacement and/or the top contact is a fault. At the base of the intrusion are medium- to coarse-grained troctolites and gabbos. These are overlain by magnetite-bearing troctolite in the centre and at the top of the intrusion, containing some centimeter- to decimeter-thick magnetite seams, dunitic layers, numerous microgabbro sills and a few anorthosite layers. Recent drilling suggests that magnetite seams might be present below cover.

Blackstone intrusion

This intrusion forms an elongate body, ~50 km long, with a strike of ~90° and a width of up to 5 km. Layering dips mostly at between 70 and 80° to the south, and is not overturned. The exposed true stratigraphic thickness of the body is ~4 km. Layering is locally pronounced and may be defined by thin (centimeter-scale) magnetite layers, or minor changes in modal proportions of pyroxene, olivine and magnetite visible on weathered surfaces, and variations in grain size. The rocks are mostly (olivine)gabbronorites and troctolites, each constituting approximately 50% of the total mass of the intrusion. The body is interpreted to represent the exposed northern limb of an upright west-trending structural syncline (the Blackstone syncline) with relics of the southern, northward dipping limb sporadically exposed 20 km to the south, immediately north of the Cavenagh intrusion. This would give it a size of ~1400 km2. The intrusion is conformably overlain by felsic volcanic rocks of the Tollu Group (Bentley Supergroup) and several basal contact exposures indicate that intrusion was into the lower basaltic portions of the Kunmarnara Group (Bentley Supergroup). In terms of lithostratigraphy, the body bears strong similarities to the Bell Rock intrusion, but with an additional, more fractionated, portion at its top.

In the upper part of the Blackstone intrusion occur discontinuous magnetite seams several millimeters in thickness within a coarse-grained troctolite. At the southern margin of the intrusion is a ~ 1 meter magnetite layer (Fig. 11) that is relatively V rich (1.5% V2O5), but PGE poor (5 ppb Pt+Pd+Au). Copper concentrations are 250 ppm, suggesting the presence of minor sulfide, common to all magnetite rich rocks in the upper portions of the Blackstone intrusion. It is uncertain whether this layer can be correlated to magnetite layer 1 in the Jameson-Finlayson intrusion, as the latter is PGE rich.


Figure 11: Magnetite layer in the Blacktone intrusion

Alcurra Dolerite suite

The Alcurra Dolerite suite includes the dolerite dykes and sills that form the majority of the regional Warakurna Large Igneous Province, formed between c. 1078 and 1073 Ma (Wingate et al., 2004). Within the west Musgrave Province, rocks compositionally similar to the dolerite dykes additionally form small basic and intermediate bodies as well as dykes typically emplaced near the margins of, or peripheral to, older G1 layered mafic intrusions, G2 massive gabbro and co-mingled gabbro–granite. These have also been included within the Alcurra Dolerite suite (Howard et al., 2009). Contact relationships from the west Musgrave Province broadly constrain the emplacement age of these rocks to < 1078 Ma, and direct dating of some of the more evolved intrusions indicates magmatism continued to at least c. 1067 Ma (Howard et al., 2009). Geochemical data from the west Musgrave Province now indicates that mafic compositions typical of the Alcurra Dolerite suite were likely formed over a much longer period, forming lavas throughout the Bentley Supergroup until at least 1047 Ma (Howard et al., 2009, 2011a; Smithies et al., 2013). Thus, the Alcurra Dolerite suite reflects reasonably long-lived melting of mantle rather than a specific melting event (Smithies et al., 2013).

The c. 1076 Ma mafic to intermediate bodies included here within the Alcurra Dolerite suite typically include fine- to medium-grained olivine gabbro, olivine norite, ferronorite, and ferrodiorite. The rocks are typified by evolved and Fe-rich tholeiitic compositions, resulting in a strong aeromagnetic signature and high specific gravity.

The Alcurra Dolerite suite shows considerable compositional variation. Most rocks are relatively evolved, with low Mg# and Cr/V ratios and elevated incompatible trace element concentrations, but there are also more primitive samples that have MgO concentrations up to 9 wt.%. Other data are discussed in Maier et al. (2014).

Saturn intrusion

The Saturn intrusion is delineated by an elliptical aeromagnetic anomaly ~ 10 km in diameter, between the Cavenagh and Blackstone intrusions (Figs. 1 and 4). A direct date of 1072 ± 8 Ma (U/Pb on baddelyite in olivine gabbro) obtained for the intrusion (Redstone Resources Ltd., written comm., 2007) is within error of the c. 1078 to 1075 Ma age range for both the G1 and G2 phases of mafic magmatism related to the Giles Event. However, a younger age of the Saturn intrusion relative to the Blackstone intrusion is indicated by the fact that the intrusion cuts the layering of the Blackstone intrusion.

The concentric magnetic pattern indicates zones of magnetite enrichment, possibly including massive magnetite layers. None have been found on the surface, but outcrop is very poor. The only exposed rocks consist of scattered massive, medium-grained, leucocratic olivine gabbros, typically containing biotite and magnetite oikocrysts up to 1 centimeter wide that form up to 5% of the rock. The samples have up to 6.7% TiO2 and 800 ppm V, comparable to magnetite gabbros from the Jameson and Blackstone intrusions. The rocks in the centre of the intrusion are somewhat more primitive than those at the margin, having higher Mg# and lower Ti concentrations, consistent with a dome-like structure. In contrast to most other mafic intrusions related to the Giles Event, biotite constitutes up to 5% of the rock, and there are abundant sulfides (pyrrhotite and chalcopyrite), reaching > 1 vol. %. The geochemistry of the rocks indicates relatively evolved compositions with Mg# up to 0.65 and Cr/V ratios <1. PGE and Cu concentrations are generally low. Based on its age, the cross-cutting relationships with rocks in the Blackstone syncline, and the enrichment in mica and sulfide, the intrusion may be transitional between the G1/G2 intrusive phase and the Alcurra Dolerite suite.

Intrusions in the Halleys-Helena-DB Hill area

Mafic rocks outcropping to the northeast and southeast of the Saturn intrusion have been explored at the Halleys, Helena, and DB Hill prospects (Redstone Resources Ltd.). The intrusion(s) can be distinguished from the Cavenagh intrusion – the latter having a strong remanant magnetic signature not seen in the former intrusions. Although contacts are not exposed, the cross-cutting magnetic patterns suggest that these bodies have intruded into the G1 Giles intrusions, and also the volcanic, volcaniclastic, and clastic rocks of the Kunmarnara and lower Tollu Group.

The rocks are mostly medium-grained, leucocratic, pyroxene-rich magnetite ferrogabbros or ferronorites. They have up to 60% clino- and orthopyroxene,  1-2% plagioclase, up to 20% intercumulus or oikocrystic magnetite, up to 5% biotite, and several per cent of sulfide minerals. The whole rock compositions are relatively differentiated, showing lower Mg# (mostly <0.6), and higher V (up to 3000 ppm) and Cu (up to 4000 ppm) concentrations than in the upper portions of the Blackstone or Cavenagh intrusions. The rocks have locally high Cu-PGE-Au concentrations. A pipe-like body has been delineated by drilling, with the best intersections containing 0.33 wt.% Cu and 0.24 ppm PGE over 74 m, and 0.5 wt.% Cu and 0.53 ppm PGE over 16 meters (Redstone Resources, 2008 Report). We considered whether the rocks could comprise the uppermost portions of the Cavenagh or Blackstone intrusions, but the latter have markedly lower mica and sulfide concentrations, lower incompatible trace element concentrations, lower Cr/Vratios, and much lower Au/PGE ratios. Instead, the Halleys rocks have more chemical and petrographic affinities to the Alcurra Dolerite suite.

Nebo Babel intrusion

The Nebo-Babel intrusion is located approximately 25 km south of Jameson community (Fig. 1 and 4). A detailed study of the Nebo-Babel Ni-Cu-PGE deposit was carried out by Seat et al. (2007, 2009) and Seat (2008), and the following section has been compiled mostly from this work. The intrusion has been dated at 1068 ± 4 Ma [Seat, 2008]. It has a tubular shape and can be traced for about 5 km. The cross-section measures 1 km in width x 0.5 km in height. The chonolith is offset by the Jameson Fault which effectively divides it into the Nebo section in the east and the Babel section in the west. Based on geochemistry, the body is interpreted to be overturned. Where contacts are observed, it appears that it was emplaced in felsic orthogneiss of the Pitjantjatjara Supersuite, although our mapping of the area shows that paragneisses belonging to the Wirku Metamorphics are a more common older basement component in the region. The stratigraphy of the intrusion is characterised by a basal breccia zone (MBZ), overlain by a chilled margin (7-9% MgO), variably-textured leucogabbronorite (VLGN), melagabbronorite (mela-GN) and mineralised gabbronorite (MGN, present only in the Babel sector) and barren gabbronorite (BGN), which in the Nebo sector is associated with oxide-apatite gabbronorite (OAGN). The OAGN constitutes about 20 to 30% of the intrusion and is characterised by oxide-rich layers that are from 5 to 30 centimeters thick, with gradational bases and sharp upper contacts. A massive and coarse-grained troctolite unit, about 15 meters thick, occurs only at Babel where it is located between VLGN and BGN in the upper part of the intrusion.

In April 2002, Western Mining Corporation announced a drill intersection of 26 meters containing 2.45% Ni, 1.78% Cu and 0.09% Co at the Nebo-Babel prospect. The deposit was discovered by conventional deflation lag geochemical sampling (Baker and Waugh, 2004). Resource estimates, obtained from 90 drillholes, are 392 Mt at 0.30% Ni and 0.33% Cu (Seat et al., 2007). The mineralised gabbronorite has a uniform grain size (5 to 20 mm) and consists of 55 to 65 vol% plagioclase, 15 to 25 vol% orthopyroxene and 5 to 10 vol% clinopyroxene. Other minerals include ilmenite, magnetite, biotite and apatite. Sulfides are monoclinic pyrrhotite, pentlandite, chalcopyrite and pyrite. The sulfide mineralisation exhibits two styles: massive ores with associated sulfide breccias and stringers, and disseminated ores, generally forming interstitial blebs in the gabbronorite unit (MGN). Sulfur isotopic data show a remarkably narrow range of δ34S values from 0 to +0.8‰. The metal tenors of the sulphides are mostly 5-6% Ni and 2-8% Cu (Cu/Ni ratio around unity), and up to several ppm Pt and Pd each (Seat et al., 2009). The massive sulfides underwent fractional crystallisation of a sulfide liquid, producing a cumulate of monosulfide solid solution relatively enriched in Os, Ir, Ru, and Rh and depleted in Pt, Pd and Au.

Seat et al. (2007) suggested that the intrusion represents a magma conduit. The initial magma was proposed to have intruded along a shear zone or a fault, more or less parallel to the regional foliation. During the first stage, chilled margins and then the VLGN units were emplaced. This order of emplacement is consistent with the progressive coarsening of the chilled margins toward the VLGN and the presence of chilled margin xenoliths within the VLGN rocks. Both chilled margin and VLGN units carry sulfides, interpreted by Seat et al. (2007) to indicate that some of the magmas entrained sulphides. However, other samples of chilled margin (Godel et al., 2011) have ~1000 ppm S, 10-20 ppb Pt and Pd each (i.e., levels similar to fertile basaltic magmas), suggesting that some of the initial magmas intruded in a S undersaturated state. The emplacement of the VLGN was followed by the MGN unit (Stage 2), which was intruded into the hot VLGN core zone, resulting in the splitting of the VLGN unit and in further inflation of the conduit. The MGN magma was also proposed to be sulfide-oversaturated, based on the occurrence of disseminated sulfide blebs. Stage 3 comprises a new pulse of more fractionated magma that was intruded between VLGN and MGN, forming the BGN unit. The flow of MGN magma is assumed to have been from the southwest, because the unit thins out towards the northeast and at the same time becomes progressively more fractionated.

Dykes related to the Giles Event

Studies on the dyke suites in the west Musgrave Province commenced with the work of Nesbitt et al. (1970), Clarke et al. (1995), Zhao and McCulloch (1994), Glikson et al. (1996) and Scrimgeour et al. (1999). Howard et al. (2006b) identified seven suites of dykes. The oldest dykes (c. 1170 Ma, ~8%MgO) belong to the Pitjantjatjarra Supersuite and are the only pre-Giles Event mafic dykes recognized. During the Giles Event, magmatism associated with the emplacement of the Warakurna large igneous province (Wingate et al., 2004) formed the Alcurra Dolerite suite (6-9% MgO, Zhao and McCulloch, 1994; Edgoose et al., 2004). Unnamed plagioclase-rich dolerite dykes (~8% MgO) clearly post-date the G1 Giles intrusions but their timing with respect to other mafic intrusions of the Giles Event is unclear, although some outcrops suggest synchronaity with G2 intrusions.

 Godel et al. (2011) identified 5 distinct dyke suites (NB1-5) in the Nebo-Babel area and investigated these as possible parental magmas to intrusions related to the Giles Event. Types NB1-3 are low-Ti basalts with 5-20% MgO, proposed to be derived from the sub-continental lithospheric mantle (SCLM), whereas types NB4 and NB5 are high-Ti basalts with 5-14% MgO interpreted to be derived from a plume source. The NB1 type is broadly equivalent to the plagioclase-phyric dykes of Howard et al. (2006b) and is a potential canadate for a parental liquid to the G1 intrusions, having c. 10-13% MgO. NB4 was proposed to be equivalent to the Alcurra Dolerite suite, consistent with low PGE concentrations in both suites. The other dyke suites studied by Godel et al. (NB2-3 and 5) are not good candidates for parental magmas to the Giles intrusions. These dykes are either too coarse-grained and MgO-rich (NB2), or too young (NB5). NB3 dykes have intermediate compositions between NB1 and NB4 and thus may represent hybrids of NB1 and the Alcurra Dolerite suite (Godel et al., 2011).

Comparative geochemistry of the intrusions

The state of differentiation of the various intrusions can be compared in a plot of Cr/V ratio vs. Mg# (Fig. 12). The least evolved intrusions are Wingellina Hills, Pirntiri Mulari, The Wart and Morgan Range. These show some overlap with the Ewarara intrusion in South Australia and the Lower Zone of the Bushveld Complex, South Africa, except that the latter has higher Cr/V ratios due to a higher chromite abundance. Intrusions showing intermediate composition are Cavenagh, Murray Range, and Hinckley Range, and the intrusive fragments to the north of Mount Muir and Hinckley Range (“North” and “Northeast”). Relatively evolved intrusions include Mantamaru, Saturn and Halleys, and dykes belonging to the Alcurra Dolerite suite.


Figure 12: Binary variation diagram of Cr/V vs Mg# for the Giles intrusions. Bushveld data are from Maier et al. (2013). “Northern” intrusions include intrusive fragments to the North of Mt Muir and Hinckley Range. Cavenagh data include samples from Straubmann (2010).

The Nd isotopic composition of the intrusions is plotted vs. Ce/Nb ratio in Figure 13. The Mantamaru intrusion has systematically more radiogenic Nd isotopic compositions (εNd = 0 to +2) and lower Ce/Nb ratios (mostly 2-7) than the other intrusions. The least radiogenic Nd isotope compositions are found in the G2 gabbros and the G1 Cavanagh and Morgan Range intrusions (εNd of -1 to -4, Ce/Nb ratios of 3-13). Similarly low εNd values are found at Kalka in South Australia (Wade, 2012). Rocks of the Alcurra Dolerite suite plot at an intermediate position (εNd = -1 to +2, Ce/Nb ratio ~ 3-5). Nebo Babel overlaps with the Alcurra Dolerite suite (εNd = -1.7 to 0.3, Ce/Nb 5-7, except for the contaminated Nebo Babel marginal rocks which have εNd as low as -3). Basement rocks in the Nebo Babel area have εNd of -4.5 to -5 and Ce/Nb ratios of 9 (Seat et al., 2011), whereas the regional Pitjantjatjara granite suite has εNd -2 to -4 and highly variable Ce/Nb (5 to >20).


Figure 13: Plot of εNd vs Ce/Nb. Note that troctolitic G1 intrusions and Alcurra Dolerite suite plot near mantle range, whereas the other intrusions contain an enriched component. Compositional field of Pitjantjatjara granite comprises 10-90 percentile of Ce/Nb data. Data for Kalka intrusion are from Wade (2006).

Most of the intrusions contain <30 ppb Pt+Pd (Fig. 14). Notable exceptions are the pyroxenite-hosted PGE reefs of the Wingellina Hills intrusion, containing up to several ppm PGE, a sulfide-bearing pyroxenitic sample from the Pirntiri Mullari intrusion (200 ppb PGE, not shown in Fig. 14 due to lack of major element data), and the samples from the pipe-like body in the Halleys intrusion (up to 200 ppb PGE). Other PGE rich sulfides not plotted include those from Nebo Babel and Manchego. The upper magnetite-enriched portions of the Jameson intrusion have up to ~ 2 ppm PGE. Scattered PGE enrichment, not necessarily accompanied by sulfide enrichment, has also been found in a few samples from the Morgan Range (up to 80 ppb), The Wart (1 sample with 120 ppb), and Cavenagh (3 samples with between 75 and 100 ppb) intrusions.


Figure 14: Binary variation diagrams vs Mg# of (d) Pt+Pd, (e) Cu/Pd. Primitive mantle value in (e) is from Maier and Barnes (1999). Symbols as in Fig. 34, with the addition of solid green squars for Alcurra Dolerite samples.

Most samples from the Giles intrusions have Cu/Pd ratios above primitive mantle values (~7000), i.e., they are PGE depleted relative to the primitive mantle (Fig. 14). Cu/Pd ratios progressively increase with decreasing Mg#, consistent with sulfide saturation having been reached at a relatively early stage during magmatic fractionation. PGE rich samples (Cu/Pd < primitive mantle) are mostly confined to the Wingellina Hills intrusion, as well as the Pirntiri Mullari, Morgan Range and Cavenagh intrusions, i.e., rocks that have Mg# mostly > 60. The greatest spread of Cu/Pd ratios, with both fertile and depleted samples, is found in the Wingellina Hills intrusion. The Halleys and Saturn intrusions have Cu/Pd ratios mostly above primitive mantle values, despite the observed enrichment in PGE. This could either suggest that the magma was already PGE depleted when sulfide saturation occurred, or that it assimilated Cu-rich crust, or that the mantle source was relatively enriched in Cu. Magnetitite layer 1 from the Jameson intrusion also has low Cu/Pd ratios (< primitive mantle), suggesting that this intrusion did not reach S saturation previously or that its upper portion crystallized from a new fertile magma batch replenishing a PGE-depleted resident magma.

Discussion

Parental magma composition

For the ultramafic portions of the G1 intrusions the fine-grained low-Ti tholeiitic “plagioclase-rich dykes” (Howard et al., 2007) and the NB1 dyke type of Godel et al. (2011), provide a suitable parent liquid. A further potential parental magma type could be the primitive members of the fine grained G2 massive gabbros. These have ~ 10-13 wt% MgO, 12 wt% FeOT, 350 ppm Ni, up to 700 ppm Cr, and 10-15 ppb Pt and Pd each. The parental magma must have been variably contaminated prior to or during emplacement, consistent with variable Nd isotopic signatures of all ultramafic intrusions. Our modeling using the PELE software (Boudreau, 1999) indicates that NB1 has a crystallization order of chromite > olivine+chromite > olivine+chromite +clinopyroxene > chromite +clinopyroxene +plagioclase+orthopyroxene, broadly consistent with petrographic observations on the Pirntiri Mulari and Wingellina Hills intrusions. The basal sequence at the Wingellina Hills and Kalka (South Australia) intrusions, and most of the central portions of Pirntiri Mulari have a crystallization sequence of olivine> opx+olivine+chromite> opx+cpx, indicating that there exist at least two different liquid lines of descent in the intrusions related to the Giles Event, the latter possibly formed from a liquid that was more strongly contaminated with partial melts of the country rocks (Irvine, 1970). The Nebo Babel chilled margins (7-9 wt% MgO and a Mg# of 51-61) are a suitable parent magma to the gabbroic intrusions. Based on similarities in incompatible trace element ratios and common enrichment in mica and sulfide, the Alcurra Dolerite suite provides a good candidate for the parental magma to the Halleys and Saturn intrusions (Howard et al., 2009).

Source of magmas

Godel et al. (2011) suggested that the NB1 magma type parental to the ultramafic intrusions of the Giles Event was derived from the sub-continental lithospheric mantle (SCLM), but Smithies et al. (2010) suggested that the regional SCLM must have been removed at the beginning of the Musgrave Orogeny. If there was SCLM at the beginning of the Giles Event, this would have to have formed after the Musgrave Orogeny. Because UHT metamorphism continued for almost 100 Ma, any SCLM present at the beginning of the Giles Event must have been still young, hot and weak, and any metasomatism of the SCLM could not have resulted in the observed non-radiogenic Nd isotopic compositions of the NB 1 dykes (εNd of -2). Thus, we do not subscribe to the model of SCLM derivation of the NB1 magma. Instead, there is evidence that NB1 is derived from a depleted asthenospheric source, based on the MORB-like Yb-Ti-Zr-Nb concentrations.

We considered whether the magmas represented by the Alcurra Dolerite suite (and NB4 dykes) could have formed through meelting of mantle that contained old, subducted, mafic oceanic crust. This could explain the high Cu/Pd ratios of the Alcurra Dolerite suite relative to other fertile basalts with similar Mg#, and the relatively high Au/PGE ratios, but not the high Pt/Pd ratios. Possibly, the mantle source for the Alcurra Dolerite suite contained both delaminated crust and mantle lithosphere. More work is clearly required to constrain the source of the Alcurra Dolerite suite.

Contamination

The intrusions related to the Giles Event have a range of εNd values from +2 to -5 (Fig. 13). This could be due to either variable contamination in the crust, or melting of several compositionally diverse mantle sources, or both. There is abundant field evidence for in situ contamination in the case of the G2 gabbros and some of the layered G1 intrusions such as Hinckley Range (Fig. 13). Amongst the intrusions in the west Musgrave Province, the lowest εNd values occur in the Cavenagh intrusion (εNd as low as -5), overlapping with the basalts of the Mummawarrawarra and Glyde Formations of the Bentley Supergroup (and Warakurna Supersuite) (Smithies et al. in preparation). Although the Musgrave crust at that stage only showed a range in εNd values from -3 to -6, which would require very substantial contamination of the Cavenagh magma, available contaminants such as granites of the Pitjantjatjara Supersuite are extremely rich in HFSE, potentially significantly reducing the required amounts of contamination (Kirkland et al., 2013). Somewhat more radiogenic values occur in rocks of the Morgan Range (εNd -3), Hinckley Range (εNd -2) and Pirntiri Mullari (εNd -1) intrusions. We have no Nd isotope data for the Wingellina Hills intrusion, but whole rock and mineral compositions are similar to those of the Pirntiri Mullari intrusion, as is the stratigraphic position of the PGE reefs, suggesting a broadly similar contamination history for both intrusions. Neodynium and Sr isotopic data for the troctolitic intrusions (Mantamaru) approximate CHUR (εNd mostly 0 to +2, 87/86ISr ~0.704) and are significantly more radiogenic (for Nd) or less radiogenic (for Sr) than any Musgrave crust present at the time. Together with trace element ratios (for example, Ce/Nb) these data indicate only little (<5%) crustal contamination in the troctolitic intrusions. Rocks of the Alcurra Dolerite suite were most likely derived from relatively shallow melting (<80 km) of a slightly depleted mantle source and have undergone early and very minor (<4%) contamination with highly enriched crustal material, followed by closure of the continuously fractionating system to further contamination. This general model of early contamination followed by fractionation can be applied to all magmas of the Giles Event, consistent with the broad similarity in Nd-isotopic data within individual intrusions, but significant variation in Ce/Nb and La/Sm ratios in the intrusions.

Emplacement and crystallization

It has been suggested that the Giles intrusions represent a rare example of an intrusive suite emplaced at highly variable crustal depths, from the lower crust to the near-surface (Glickson, 1995). Our work confirms that some of the troctolitic intrusions in the southwest were emplaced at relatively shallow levels (sub-volcanic to <4kb) whereas the exposure level deepens to the north and east. Preserved county-rock inclusions and contacts indicate that the Mantamaru intrusion was emplaced at the stratigraphic level of the Mummawarrawarra Basalt (Kunmarnara Group). The low metamorphic grade (greenschist facies) of the basalts indicates an upper crustal and extensional environment for intrusion. Constraints on the crystallisation age of these G1 intrusions are the minimum depositional age of the Kunmarnara Group (defined by granite intrusion at c. 1078 Ma, Sun et al., 1996, Howard et al., 2011), and a direct U–Pb zircon age of 1076 ± 4 Ma (Kirkland et al., in prep) on a layered Giles intrusion gabbro (GSWA 194762).

Primarily in the eastern part of the west Musgrave region (Hinckley Range), massive gabbro (G2) cuts the layered G1 intrusions and typically shows abundant and widespread evidence of co-mingling with leucogranitic magmas. This bimodal magmatism was also accompanied by deformation (shearing and west-northwest folding adjacent to major shear zones) and age constraints on magmatism and deformation lie between 1078 ± 3 and 1074 ± 3 Ma (Howard et al., 2011b). These age constraints are virtually identical to those for the layered Giles (G1) intrusions but where temporal field relationships can be established, G2 intrusions always post-date G1.

South of Blackstone community, in the Blackstone Sub-basin, rhyolites of the Smoke Hill Volcanics directly overly the layered G1 Blackstone intrusion (a component of the Mantamarru intrusion) without an obvious intervening fault. Crystallisation, or depositional, ages for the rhyolites are within analytical error of the emplacement age range of the G1 and G2 intrusions, and rhyolite compositions strongly resemble those of leucogranites associated with G2 intrusions. This requires extensive and rapid crustal uplift, erosion, and exhumation of the layered Giles G1 intrusions, immediately followed by felsic volcanism.

Some of the ultramafic intrusions (eg Ewarara) were proposed to have been emplaced at a higher pressure, up to 10-12 kbar (Goode and Moore, 1975; Ballhaus and Glickson, 1989). If the crust was only ~35 km thick at the end of the Musgrave Orogeny (e.g. Smithies et al., 2011), only 40 m.y. before the Giles Event, emplacement of the Ewarara intrusion was then likely near the base of the crust. Ballhaus and Glickson (1989) proposed a somewhat shallower depth of emplacement for the Wingellina Hills and Pirntiri Mullari intrusions, at 6.5 kbar (~20km). However, since the coeval gabbroic intrusions are up to 10km thick, the original emplacement depth of the ultramafic bodies may have been as low as 10 km. A relatively deep emplacement level of the ultramafic intrusions could explain why they are less abundant than the gabbroic and troctolitic bodies. If the ultramafic intrusions were generally emplaced at a deeper level than the mafic intrusions, there should be proportionally more ultramafic intrusions in South Australia since the crust there is exposed at a deeper level (Goode, 2002). This appears to be valid. Also consistent with this model is the fact that the ultramafic bodies tend to be exposed in the cores of regional folds (i.e. the anticline north of Blackstone community hosting the Pirntiri Mullari and Morgan Range intrusions) or along faults. However, based on PELE modeling, the An contents of plagioclase in the Wingellina Hills and The Wart intrusions are too high for high-pressure crystallization and more consistent with a pressure of 1 kbar. In contrast, the low An contents of plagioclase in the Pirntiri Mulari intrusion would be consistent with a deeper emeplacement relative to Wingellina Hills and The Wart. Viewed solely from a stratigraphic perspective, this idea is inviting, as the former is located closer to the core of the anticline situated to the north of Blackstone community.

The origin of the layering in the studied intrusions remains uncertain. The G1 intrusions contain layers and schlieren of magnetite, anorthosite, ultramafic rocks, and microgabbros, as well as abundant xenoliths and autoliths. Inch-scale layering has been observed in pyroxenite from the Ewarara and Wingellina Hills intrusions (Goode, 1970; Ballhaus and Glickson, 1989). Other features described include convoluted layering (Wingellina Hills, Ballhaus and Glickson, 1989), scour channels (Kalka, Goode, 1970), graded layers (sometimes overturned) and load casts (e.g., in the Olivine Gabbro and Anorthosite zones of the Kalka intrusion, Goode, 1970) and strong alignment of plagioclase and pyroxene. These features are analogous to those described in the Bushveld Complex where they were explained by magma replenishment and cumulate slumping and sorting (Maier et al., 2013b).

Fragmentation of intrusions

Past authors speculated that some or all G1 intrusions represent tectonised remnants of a larger intrusion (Sprigg and Wilson, 1959; Nesbitt and Talbot, 1966, Glickson, 1995, Smithies et al., 2009, Howard et al., 2011; Aitken et al., 2013). In the west Musgrave Province, the Cavenagh and Blackstone intrusions are least deformed (Glickson et al., 1995), whereas the Murray Range intrusion is amongst the most deformed. Significant deformation has also affected the Wingellina Hills, Pirntiri Mullari, Lehman Hills, and Jameson intrusions. In some cases, deformation is focused along the contacts of the intrusions, expressed by mylonite zones, for example, at the top of the Pirntiri Mullari intrusion. Enhanced deformation also occurs along layer contacts such as magnetite layer 1 in the Jameson intrusion. Some intrusions have undergone large-scale folding, for example, the Latitude Hill and Blackstone intrusions. A potential model of the original configuration of many of the Giles intrusion(s) could be the Kalka intrusion (Goode, 1970), a 6 km thick body with a 450 m thick ultramafic basal portion (basal orthopyroxenite, progressively changing to websterite with height) and a gabbroic-troctolitic-anorthositic upper portion. Peridotites are exposed at the nearby Gosse Pile intrusion that has been interpreted to represent the tectonically dismembered lower portion of the Kalka intrusion. In contrast to the above model, Ballhaus and Glickson (1995) suggested that, except for the combined Blackstone-Bell Rock-(The Wart) intrusions, most G1 intrusions represent distinct intrusions.

The present work provides support for connecting some, but not all of the bodies. Specifically, our geophysical, lithological and compositional data strongly indicate that the Jameson-Finlayson, Blackstone and Bell Rock intrusions are fragments of an originally contiguous body (i.e., the Mantamaru intrusion). It has a minimum preserved size of 3400 km2, making it one of the largest mafic-ultramafic intrusions on Earth, after the Bushveld Complex (60000 km2), and being in the same size range as the Great Dyke, Stillwater, Sept Iles and Dufek intrusions (3000-5000 km2). It has been suggested that the Cavenagh intrusion forms the southern limb of the synclinal Blackstone intrusion (Nesbitt and Talbot, 1966; Aitken et al., 2013). This would add at least another 540 km2 to the size of the Mantamarru intrusion. However, there are some significant compositional differences between the Blackstone and Cavenagh intrusions. The Blackstone intrusion is much more differentiated than the Cavenagh intrusion, and it has a massive magnetite layer which seems to be absent in the Cavenagh intrusion. The Cavenagh intrusion has elevated PGE concentrations in its upper portion whereas the Blackstone intrusion is PGE depleted. Finally, the Blackstone intrusion has a significantly more radiogenic Nd isotopic signature than the Cavenagh intrusion. However, it needs to be borne in mind that individual intrusions can have a wide range of trace element and isotiopic compositions, as seen, e.g., in the Bushveld and Kalka intrusions.

In summary, at present the only strong case for dismemberment of an originally contiguous body is the Mantamaru intrusion, consisting of the Jameson, Blackstone and Bell Rock ranges. However, future work may well show that other Giles intrusions also represent fragments of larger precursor intrusions.

Comparison to other large layered intrusions    

In order to evaluate the petrogenesis and mineralization potential of the intrusions emplaced during the Giles Event, it is useful to draw comparisons with the well characterized and highly mineralized Bushveld Complex (Fig. 15). The Pirntiri Mulari, Wingellina Hills and The Wart intrusions represent approximate stratigraphic equivalents of the Lower and Critical zones of the Bushveld Complex. Both the Giles ultramafic intrusions and the Bushveld Lower Zone show basal reversals, thick ultramafic portions, and numerous ultramafic-mafic cyclic units. Bushveld-style PGE reefs have been identified in the Wingellina Hills intrusion, and there are indications a reef may exist in the Pirntiri Mulari intrusion. The prospective horizon of The Wart (and other Giles ultramafic intrusions) intrusion remains poorly studied, partly due to limited access. A major difference between the Giles ultramafic intrusions and the Bushveld Complex is that the former lack chromitite seams. The Wingellina Hills intrusion contains thin chromitite schlieren and lenses, but The Wart and Pirntiri Mularu intrusions are essentially barren of chromitite. This is possibly due to the different crystallization orders of these Giles intrusions. In the Pirntiri Mulari and The Wart intrusions Cr-rich clinopyroxene was an early crystallizing phase which could have led to depletion of the residual magma in Cr. In contrast, the Wingellina Hills intrusion has a thick harzburgitic basal portion, possibly leading to a lesser degree of Cr depletion of the fractionating magma. In addition, late magmatic concentration and sorting processes, as advocated by Maier et al. (2013b) for the Bushveld Complex, may have been relatively less efficient in the Pirntiri Mulari and The Wart intrusions.


Figure 15: Stratigraphic comparison of Giles intrusions with Bushveld Complex (Bushveld log and data from Maier et al., 2013).

The Morgan Range intrusion could be a stratigraphic equivalent to the Upper Critical Zone-Main Zone transition interval of the Bushveld Complex, as it contains some ultramafic rocks at its northern edge, but is otherwise dominated by gabbronoritic rocks of moderately evolved composition. Whether it contains a PGE reef analogous to that at Wingellina Hills remains unknown. The Cavenagh, Michael Hills, Latitude Hill, Hinckley Range, Murray Range, Lehman Hills, and Mount Muir intrusions could be stratigraphic equivalents of the Main Zone, as they have intermediate compositions and, in some cases, are relatively poorly layered.

The Mantamaru intrusion appears to be the approximate stratigraphic equivalent of the Bushveld Upper Zone. Specifically, the upper portion of the Jameson intrusion shares several similarities with the Bushveld Upper Zone. Firstly, both contain several magnetite layers (~25 in the Bushveld, at least 11 at Jameson) that are between a few centimetres to more than 10 meters thick (i.e., layer 1 at Jameson, magnetite layer 21 in the Bushveld) and may contain anorthosite xenoliths. Secondly, most layers have sharp lower contacts, whereas the upper contacts may be gradational. Thirdly, there is a trend of decreasing V concentration with height from the basal to the upper layers, and within individual layers. In the Bushveld Complex, the fourth layer from the base (in the eastern limb) constitutes the Main Magnetite Layer from which >50% of the world’s V production is derived (Crowson, 2001). It combines considerable thickness (up to ~3m) with some of the highest V concentrations amongst Bushveld magnetitites (1.5-2% V2O3, Klemm et al. 1985). However, there are also significant differences between the Bushveld Complex and the upper portion of the Jameson intrusion. For example, the basal magnetite seam in the Jameson intrusion has higher PGE concentrations than the equivalent Bushveld magnetite layer(s). The low PGE concentrations of the Bushveld basal magnetite seams could be due to the formation of the world-class Merensky and UG2 PGE reefs at stratigraphically lower levels, effectively extracting the PGE from the magma. If this model is correct, the implication could be that the Mantamaru intrusion has a low potential to host PGE reefs at depth. Alternatively, the PGE depletion of the Bushveld basal magnetites could reflect their exposure at the margin of the Bushveld lopolith, with relatively PGE rich seams possibly present in the centre of the Bushveld Complex. Another difference between the Bushveld Complex and the upper portion of the Jameson intrusion is that the Bushveld Complex has an apatite-rich layer near the top, whereas no apatite-rich layers have yet been identified in the Jameson intrusion, although there is a trend of P2O5 enrichment in some of the upper magnetite layers (Traka Resources), and up to 0.8wt% P has been intersected by boreholes in magnetite to the south of the Bell Rock Range (P. Polito, written communication). A further difference between the two intrusions is that the V concentrations in the basal magnetite layer of the Jameson intrusion are somewhat lower than in the Bushveld Complex, possibly due to lower R factors (mass ratio of silicate melt to sulfide melt, Campbell and Naldrett, 1979) in the former. This interpretation is possibly consistent with a weak positive correlation between the V concentration of magnetite and the size of the host layered intrusion (Maier et al., 2014).

Tectonic setting

The Musgrave Province has been located between the thick lithospheric keels of the West Australian, South Australian and North Australian cratons since the Mesoproterozoic. Begg et al. (2009) proposed that mantle plumes may be channeled along cratonic keels resulting in enhanced adiabatic mantle partial melting and mafic magmatism at the margins of cratons or within cratonic suture zones. However, the long time span of continuing mantle magmatism and UHT metamorphism in the Musgrave Province is inconsistent with a mantle plume model (Evins et al., 2010; Smithies et al., 2013, 2014). Instead, the latter authors suggested that the Musgrave Province acted as a stationary zone of mantle upwelling for > 200 Ma, resulting in a persisting hot zone characterized by mafic and felsic magmatism and UHT metamorphism. These features appear to point to a plate-driven trigger of magmatism. Mantle melting could have been caused by processes such as lithospheric delamination, volatile transfer from the SCLM or the crust to the asthenosphere, or mantle flow along an irregular base of the lithosphere (Silver et al., 2006). The resulting mantle magmas may have ponded at the base of the thinned lithosphere to be periodically drained during collisional events resulting in transpressional rupturing (Silver et al., 2006). A similar scenario was envisaged for the Ventersdorp, Great Dyke, Bushveld and Soutpansberg continental magmatic events in southern Africa (Foulger, 2010).

In the case of the Musgrave region, events during and after the 1345-1293 Ma Mount West Orogeny likely resulted in crustal thickening, partial melting and densification of lower crust. Evidence from the REE geochemistry of granites related to the subsequent, 1220-1150 Ma, Musgrave Orogeny is that this event began with a change from deep to shallow crustal melting that can be related to delamination of residual lower crust and the underlying lithospheric mantle. The ensuing c. 100 Ma period (1220-1120 Ma) of UHT metamorphism supports a sustained regime of highly thinned crust, with little or no remaining lithosphere. Throughout this period, magmatism was predominantly felsic because the lower crust was a MASH zone preventing ascent of dense mafic magmas. The crustal thermal structure established during the Musgrave Orogeny strongly influenced conditions at the beginning of the Gilse Event (Smithies et al., 2014). The Giles Event was triggered by far-field forces acting on the margins of the West Australian craton (Evins et al., 2011; Smithies et al., 2014). This led to initial subsidence and deposition of the Kunmarnara Group, followed by draining of ponded sub-crustal melts (G1, G2, Alcurra Dolerite suite). The relatively early G1 and G2 magmas underwent variable contamination during ascent into the crust (Fig. 16). Subsequent magmas (Alcurra Dolerite suite) were generally less contaminated because the crust had become more refractory. These magmas are more differentiated because the crust had become thicker, leading to enhanced intra-crustal ponding and fractional crystallization. The composition of Alcurra magmas (including their low PGE concentrations, high Cu/Pd, Pt/Pd, and Au/PGE ratios) could be explained by foundering of crust and new SCLM (Fig. 16), lowering of the mantle solidus, and melting of a hybrid crust-rich mantle.


Figure 16: Schematic model of emplacement of Giles intrusions. See text for discussion.

The formation of the large layered intrusions of the Giles Event can be explained within the context of a prevailing compressive regime, allowing the magma to form inflating sills rather than dykes. Importantly, the Musgrave Province demonstrates that large layered intrusions are not confined to cratons. What is required is a stable, broadly (at least locally and temporarily) compressive, tectonic environment where magmas can ascend in locally extensional, possibly transpressional zones.

Origin of mineralization

PGE reefs within the Wingellina Hills layered intrusion

Layered intrusions host the bulk of the world’s PGE resources, in the form of stratiform layers or so-called reefs. PGE reefs occur in many layered intrusions (see summary in Maier, 2005), but economic examples are presently confined to the Bushveld Complex, the Stillwater Complex and the Great Dyke. In all cases, the reefs consist of <1-3% disseminated sulfides within laterally extensive layers of ultramafic or mafic rocks. The host intrusions are exceedingly sulfur poor suggesting that sulfide saturation of the magma was evantually reached due to fractionation. Many authors consider addition of external S, as normally believed to be responsible for the formation of most Ni-Cu sulfide ores, of little importance, consistent with mantle-like S isotopic signatures in most reefs (Li et al., 2008; Liebenberg, 1970). At least in the case of the Bushveld Complex, mixing between compositionally different magmas was probably also not important in reef formation, because Bushveld magmas were found to be highly S undersaturated (Barnes et al., 2010). Concentration of sulphides likely occurred during cumulate sorting in response to syn-magmatic subsidence of the intrusions (Maier et al., 2013b).

The Wingellina Hills main PGE reef shows many similarities to the Great Dyke and Munni Munni PGE reefs, including the broadly stratiform nature, their occurrence towards the top of the ultramafic portion of the intrusion, and the offset patterns of the chalcophile elements. The total amount of PGE is also in the same range in both the Wingellina Hills reef and the Main Sulfide Zone of the Great Dyke. In both cases, there is no significant change in trace element ratios across the reef. The origin of the Wingellina Hills main reef can be modelled by a process of sulphide saturation in response to fractionation of a NB1-type magma. The latter has been shown to reach S saturation after ~30% fractionation, at about the time plagioclase appears on the liquidus (Godel et al., 2011), consistent with the position of the reef at the top of the ultramafic portion.

Important differences between the PGE reefs at Wingellina Hills and those in the Great Dyke and Bushveld Complex include the fact that the Wingellina Hills reefs have lower PGE grades and contain less sulphide. The low grade could reflect less efficient metal concentration due to faster cooling rates in a relatively small intrusion, whereas the paucity in sulphides could be due to metamorphic S loss, consistent with sub-cotectic sulphide proportions in most Wingelina Hills rocks.

Cu-Ni-PGE-Au mineralization at Halleys

PGE-Cu-Au reef-style mineralization is common to the upper portions of many layered intrusions (see Maier, 2005 for a summary). The enrichment of magnetite together with sulfide at the Halleys prospect could suggest that the Halleys body represents the evolved top portion of a layered intrusion, perhaps being a correlative of the magnetite-rich upper part of the Blackstone intrusion. However, Halleys has much higher Au and S concentartions, and mica contents than the Blackstone intrusion, and field relationships indicate that it is cross cutting the southern segment of the Blackstone intrusion. Thus, if Halleys represents the upper portion of a large intrusion, this would likely be a different body from the Blackstone or Cavenagh intrusions, and instead be related to the Alcurra Dolerite suite to which also the Saturn intrusion belongs. The relatively high Au/PGE ratios are a hallmark of Alcurrra-type magma. However, a direct correlation between the Halleys and Saturn intrusions is inconsistent with the distinct magnetic signature of the intrusions. An alternative model could be that the mineralization at Halleys is more akin to contact-style mineralization in many mafic ultramafic intrusions, including the Platreef of the Bushveld Complex. Such deposits are considered to have formed through a combination of processes, including sulfide saturation in response to fractional crystallization, often accompanied by floor contamination. The proximilty of the floor would have caused a high cooling rate of the magma, producing wide disseminated mineralization rather than narrow sulfide reefs. Sulfides at Halleys have δ34S -0.9, providing little added constraint on the nature of the S source. More work is clearly required to further constrain the petrogenesis of the intrusion and its mineralization.

V and PGE mineralization in magnetite seams of the Jameson intrusion

The petrogenesis of oxide seams in layered intrusions has been reviewed by Maier et al. (2013b). To form massive oxide layers, magnetite must effectively fractionate from the magma as the cotectic porportions of magnetite and silicates are in the 5-30% range (Toplis and Carroll, 1996). One of the main criticisms of crystal fractionation models has been that the high yield strength of basaltic magmas prevents effective segregation of small magnetite crystals to form laterally extensive, massive oxide layers with knife-sharp bottom and top contacts (McBirney and Noyes, 1976). Irvine et al. (1998) proposed that cumulate layers in the Skaergaard intrusion precipitated from density currents of crystal slurries that swept down the chamber walls, but Maier et al. (2013b) rejected this model for oxide layers in the Bushveld Complex because density currents would not preserve the abundant, highly elongated to wispy, sub-horizontally orientated, anorthosite autoliths within chromitite and magnetitite layers. An alternative model invokes a shift in phase stability fields of oxides due to changes in pressure (Cameron, 1980; Lipin, 1993) affecting the entire magma chamber simultaneously. Temporary super-saturation in magnetite could also be achieved by an increase in the oxygen fugacity of the magma, for example, via contamination (Ulmer, 1969), or a combination of magma mixing, pressure-change, and oxidation in response to magma replenishment. One of the most popular models for the formation of massive magnetite layers is saturation of the magma in Fe-oxide liquid (Philpotts, 1967; Naslund, 1983; Zhou et al., 2005). Maier et al. (2013b) proposed that magnetite seams form through crystal sorting of oxide-silicate slurries that slump along the chamber floor during subsidence of large slowly cooling magma chambers. The slurries could inject into the semi-consolidated crystal pile, and locally form transgressive pipes. Evidence for cumulate deformation and slumping has been found in magnetite seams of other layered intrusions.

Magnetite layers in layered intrusions commonly contain PGE-rich sulfides (see Maier, 2005, for a summary of occurrences). The sulfides may have precipitated in response to a rapid depletion in iron of the magma, or magma reduction following crystallization of copious amounts of magnetite (Jugo et al., 2010). As long as the magma still contained PGE at this relatively advanced stage of differentiation, the sulfides would be PGE rich. In the Bushveld Complex, this was evidently not the case, presumably because of the formation of the PGE reefs in the Critical Zone. In the Jameson intrusion, magnetites are relatively rich in PGE, possibly suggesting there is no PGE reef at depth.

Nebo-Babel Ni-Cu deposit

The Nebo-Bable Ni-Cu deposit bears a number of similarities to other magmatic Ni-Cu sulfide deposits. Of particular note is the tubular (chnolithic) shape and the high abundance of sulfides relative to the size of the body. Seat et al. (2007, 2009), argued that the distribution of sulfides would have been controlled by changes in magma velocity, in turn related to changes in the orientation of the conduit. However, in contrast to Noril’sk and Voisey’s Bay, sulphide saturation at Nebo Babel was interpreted to have been achieved largely by mixing of Alcurra-type magma with NB1-type magma, as well as orthogneiss contamination, without addition of external S (Seat et al., 2007; Godel et al., 2011). The magma mixing model is consistent with the fact that there is considerable compositional overlap between NB3 dykes (interpreted to represent hybrid Alcurra-NB1 magma) and the Nebo Babel chilled margins. Furthermore, Nebo Babel appears to be located at the boundary between two compositional crustal domains; To the east of Nebo Babel the mafic rocks are of NB1 and Alcurra Dolerite lineage, whereas to the west, only the Alcurra Dolerite lineage appears to be present.

In terms of mass balance, the orthomagmatic model of sulfur derivation is certainly feasible: extracting 100 ppm S from 1km3 of magma can produce a massive sulfide lens 1 km long, 10 meters high, and 20 meters wide. However, in the recent literature on magmatic Ni-Cu deposits, the prevalent trend has been to explain massive sulfide deposits in general by addition of externally derived S. One of the main reasons for this is that the alternative (i.e. orthomagmatic derivation of the sulfur) would require an extremely effective concentration mechanism (Keays and Lightfoot, 2010), being that the cotectic proportion of sulfide precipitating from S-saturated troctolitic-gabbronoritic magma is very small (perhaps as little as 0.1 wt.%) due to the increase in Fe concentration of fractionating plagioclase-saturated magma (Scoates et al., 2001). Quantitative modeling to resolve this question remains to be done, but the field evidence provides some insight. Massive sulfides are extremely rare in layered intrusions, illustrating that concentration of cotectically precipitating sulfides from convecting magma is difficult. It seems doubtful that it would be easier in fast flowing magmas within magma feeder systems.

The basis for the model of Seat et al. (2007) was that the sulfide ore and its gabbroic host rocks both have a mantle-like sulphur isotopic composition, and that the rocks of the Pitjantjatjara Supersuite, which forms the immediate country-rock to the mineralised gabbro, are sulphur-poor. However, the mantle-like isotopic composition of the Nebo-Babel mineralization, and of the host gabbro, do not exclude a country-rock source of sulphur, they only require that the contaminant was isotopically juvenile and/or that R factors were large (Lesher and Burnham, 2001).

Granites of the Pitjantjatjara Supersuite represent only a very minor lithological component within this part of the Mamutjarra Zone. Main rock-types within this area include rocks of the Wirku Metamorphics, the Winburn Granite and rocks of the Bentley Supergroup. The Wiku Metamorphics comprise a range of pelitic and psammitic metasedimentary rocks and are highly unlikely to have mantle-like S-isotopic compositions. However, the Winburn Granite and volcanic and volcaniclastic rocks of the Bentley Supergroup are co-genetic (Smithies et al., 2013), and in the area around and to the west of Nebo-Babel itself (the Palgrave area), typically contain visible sulfide (mostly pyrite) and are locally sulfide-rich. In addition, our bedrock geological interpretation of the area immediately to the south of Nebo-Babel infers the presence of a deep graben-structure filled by rocks of the Bentley Supergroup. Thus, there exist clearly several locally available potentially sulphur-rich country rocks. Whereas much of the Bentley Supersuite was deposited after c.1065 Ma and could not have interacted with the Nebo-Babel gabbro, deposition of the lower portion (the Mount Palgrave Group and much of the Kaarnka Group), pre-dates intrusion of the Nebo-Babel gabbro (Smithies et al., 2013) and thus provides a potential contaminant. In addition, magmatism associated with the Bentley Supergroup is overwhelmingly dominated by mantle-derived tholeiitic mafic magmas and by their felsic derivatives; The felsic volcanic rocks and the Winburn Granite are dominantly strongly ferroan and alkaline-calcic (A-type) derived through fractionation of mantle-derived mafic magmas, rather than through partial melting of crustal sources (Smithies et al., 2013).The presently available S isotopic data for the units of the Bentley Supergroup indicate a range of compositions, with those from the lower part having δ34S between +1.8 and +7, potentially representing a suitable S-source for the Nebo-Babel deposit.

In summary, considering that there is compositional and lithological evidence for country rock contamination at Nebo Babel, and the fact that the basement contains lithologies that may have provided juvenile S to the magma, we suggest that addition of external S to the Nebo Babel magma remains a possibility. External addition of juvenile volcanic rocks would be consistent with the relatively high Cu/Pd and Cu/Ni ratios (due to addition of crustal Cu) and high Au/Pd ratios (due to addition of crustal Au) at Nebo Babel (see data of Seat et al., 2007).

Thoughts on prospectivity

Large magmatic events dominated by mafic-ultramafic magmas cause enhanced heat flux into the crust, triggering crustal melting, devolatisation and large scale fluid flow. Deposit types favoured by such regimes include magmatic PGE-Cr-Fe-P deposits in large layered intrusions, Ni-Cu sulfide ores formed through assimilation of S-rich strata in magma feeder conduits or at the base of layered intrusions, and hydrothermal deposits of variable style, notably in the roof and side wall of the largest intrusions. Pirajno et al. (2000) drew analogies between the Giles Event and the Bushveld Complex of South Africa. Both are associated in space and time with bimodal volcanic rocks which, in the case of the Bushveld Complex, are well endowed with a wide range of hydrothermal ore deposits, ranging from greisen-style deposits, breccia pipes with Sn-W to epithermal and mesothermal lode-Au and iron oxide-copper-gold (IOCG) deposits. The Musgrave region too has a high potential for such deposits, as highlighted by the discovery of Cu-Au vein style mineralisation in the felsic volcanic rocks of the Tollu Group to the north of the Cavenagh Range (Abeysinghe, 2002). The prospectivity for certain commodities and deposit styles is discussed in more detail in Maier et al., (2014).

Conclusions

The west Musgraves Province is located between the West Australian, South Australian and North Australian cratons. The area was the focus of long-lived mantle upwelling producing large volumes of magnesian basaltic to tholeiitic magma and their felsic derivatives. Magmatism led to crustal melting, lithospheric delamination, and a high crustal heatflux over > 200 Ma. The broadly compressive regime with localized extension favoured the formation of several large mafic-ultramafic layered intrusions. Due to the large size of the bodies, cooling rates were relatively slow. Crustal loading may have led to subsidence and sagging prior to complete solidification. The cumulates unmixed and formed lenses and layers of peridotite and magnetitite that are locally PGE enriched. Syn- to post magmatic tectonism led to fragmentation of many of the intrusions. The degree of crustal contamination was mostly relatively minor (<5%), but locally, basaltic magmas mingled with coeval granitic magmas. The mineralization potential of the Musgrave Province is considerable. The mafic-ultramafic intrusions host significant V, PGE and Ni-Cu deposits, with high potential for Fe, Ti and apatite. The heat flux generated by the large mantle melting event likely led to widespread crustal fluid flux that may have formed hydrothermal mineral deposits of variable style. Paucity of exposure and access creates an exploration challenge, but also an opportunity.

References

Aitken, AR, Dentith, MC, Evans, SF, Gallardo, LA, Joly, A, Thiel, S, Smithies, RH and Tyler, IM 2013, Imaging crustal structure in the west Musgrave Province from magnetotelluric and potential field data: Geological Survey of Western Australia, Report 114, 81p.

Ballhaus, C and Glickon, AY 1989, Magma mixing and intraplutonic quenching in the Wingellina Hills intrusion, Giles Complex, central Australia: Journal of Petrology, v.30, p.1443-1469.

Ballhaus, C and Glickson, AY 1995, The petrology of layered mafic-ultramafic intrusions in the Giles Complex, western Musgrave block, central Australia: AGSO Journal, v. 16, p.69-89.

Barnes, S-J and Maier, WD 1999, The fractionation of Ni, Cu and the noble metals in silicate and sulphide liquids, in Keays, RR, Lesher, CM, Lightfoot, PC, Farrow, CEG, eds., Dynamic processes in magmatic ore deposits and their application to mineral exploration: Geological Association of Canada, Short Course Notes 13, p. 69-106.

Betts, PG and Giles, D 2006, The 1800–1100 Ma tectonic evolution of Australia: Precambrian Research, v. 144, p. 92–125.

Bodorkos, S, and Wingate, MTD 2008, 174594: metamorphosed leucogabbro, Mirturtu Camp; Geochronology Record 716: Geological Survey of Western Australia, 4p.

Boudreau AE (1999) PELE-a version of the MELTS software programme for the PC platform. Computers and Geosciences, 25, 201-203.

Camacho, A 1997, An isotopic study of deep-crustal orogenic processes: Musgrave Block, Central Australia: PhD Thesis (unpublished), The Australian National University.

Camacho, A, Compston, W, McCulloch, M and McDougall, I 1997, Timing and exhumation of eclogite facies shear zones, Musgrave Block, central Australia: Journal of Metamorphic Geology, v. 15, p. 735–751.

Cameron, EN 1980, Evolution of the Lower Critical Zone, central sector, eastern Bushveld Complex: Economic Geology, v.75, p. 845-871.

Campbell, IH and Naldrett, A.J 1979, The influence of silicate:sulfide ratios on the geochemistry of magmatic sulfides: Economic Geology, v. 74, p. 1503-1505.

Clarke, GL, Sun, S-S, and White, RW 1995a, Grenville age belts and associated older terranes in Australia and Antarctica: AGSO Journal of Australian Geology and Geophysics, v. 16, p. 25-39.

Clarke, GL, Buick, IS, Glikson, AY, and Stewart, AJ 1995b, Structural and pressure-temperature evolution of host rocks of the Giles Complex, western Musgrave Block, central Australia: evidence for multiple high-pressure events: AGSO Journal of Australian Geology and Geophysics, v. 16, p 127-146.

Collins, WJ and Teyssier, c, 1989, Crustal scale ductile fault systems in the Arunta Inlier, central Australia: Tectonophysics, 158, 49-66

Crowson, P 2001, Mineral handbook 2000-2001 Mining Journal Books Kent UK Edenbridge 486p.

De Paolo, DJ and Wasserburg, GJ 1976, Inferences about magma sources and mantle structure from variations of 143Nd/144Nd: Geophysical Research Letters, v.3, p. 743–746.

Daniels, JL 1974, The Geology of the Blackstone region, Western Australia: Geological Survey of Western Australia, Bulletin 123, 257pp.

Edgoose, CJ, Scrimgeour, IR and Close, DF 2004, Geology of the Musgrave Block, Northern Territory: Northern Territory Geological Survey Report 15, 48pp

Evins, PM, Smithies, RH, Howard, HM and Maier, WD 2009, Holt, W.A. Sheet 4546: Western Australia Geological Survey, 1:100 000 Geological Series.

Evins, PM, Smithies, RH, Howard, HM, Kirkland, CL, Wingate, MTD and Bodorkos, S 2010a, Redefining the Giles Event within the setting of the 1120–1020 Ma Ngaanyatjarra Rift, west Musgrave Province, Central Australia: Geological Survey of Western Australia, Record 2010/6, 36p.

Evins, PM, Smithies, RH, Howard, HM, Kirkland, CL, Wingate, MTD and Bodorkos, S 2010b, Devil in the detail: The 1150–1000 Ma magmatic and structural evolution of the Ngaanyatjarra Rift, west Musgrave Province, Central Australia: Precambrian Research, v. 183, p. 572–588.

Flöttmann, T and Hand, M 1999, Folded basement-cored tectonic wedges along the northern edge of the Amadeus Basin, central Australia; evaluation of orogenic shortening: Journal of Structural Geology, v. 21, p. 399–412.

Foulger, GR 2010, Plates vs Plumes. Wiley-Blackwell, 328pp.

Giles, D, Betts, PG, and Lister, GS 2004, 1.8-1.5-Ga links between the North and South Australian Cratons and the Early-Middle Proterozoic configuration of Australia: Tectonophysics, v. 380, p. 27-41.

Glikson, AY, (ed) 1995, The Giles mafic-ultramafic complex and environs, western Musgrave Block, central Australia, Thematic issue: AGSO Journal of Geology and Geophysics, v. 16, No. 1-2.

Glikson, AY, Stewart, AT, Ballhaus, GL, Clarke, GL, Feeken, EHT, Level, JH, Sheraton, JW and Sun, S-S 1996, Geology of the western Musgrave Block, central Australia, with reference to the mafic-ultramafic Giles Complex: Australian Geological Survey Organisation, Bulletin 239. Australian Government Publishing Service, Canberra.

Godel, B, Seat, Z, Maier, WD and Barnes, S-J 2011, The Nebo-Babel Ni-Cu-PGE Sulfide Deposit (West Musgrave Block, Australia): Part 2- Constraints on Parental Magma and Processes, with Implications for Mineral Exploration: Economic Geology, v. 106, p. 557-584.

Goode, ADT 1970, The petrology and structure of the Kalka and Ewarara layered basic intrusions, Giles Complex, central Australia. Unpubl PhD thesis, University of Adelaide.

Goode, ADT 2002, The Western Musgrave Block – Australia. In Musgrave Block field guide, 16th Australian Geol Convention

Goode, ADT and Moore, AC 1975, High pressure crystallisation of the Ewarara, Kalka and Gosse Pile intrusions, Giles Complex, central Australia: Contributions to Mineralogy and Petrology, v. 51, p. 77-97.

Gray, C.M., 1971. Strontium isotope studies on granulites: Australian National University, Canberra, Australian Capital Territory, PhD thesis (unpublished), 242p.

Gray CM (1967) The geology, petrology and geochemistry of the Teizi metanorthosite. Unpubl Hns thesis, Univ Adelaide.

Haines, PW, Hand, M and Sandiford, M 2001, Palaeozoic syn-orogenic sedimentation in central and northern Australia: a review of distribution and timing of intracontinental orogens: Australian Journal of Earth Science, v. 48, no. 6, p. 911–928.

 Howard, HM, Smithies, RH, Pirajno, F and Skwarnecki, MS 2006a, Bates, W.A. Sheet 4646: Western Australia Geological Survey, 1:100 000 Geological Series.

Howard, HM, Smithies RH, Pirajno F  2006b, Geochemical and Nd isotopic signatures of mafic dykes in the western Musgrave Complex: Geological Survey of Western Australia Annual Review 2005-06, p. 64-71.

Howard, HM, Smithies, RH, Pirajno, F and Skwarnecki, MS 2007, Bell Rock, W.A. Sheet 4645: Western Australia Geological Survey, 1:100 000 Geological Series.

Howard, HM, Smithies, RH, Kirkland, CL, Evins, PM and Wingate, MTD 2009, Age and geochemistry of the Alcurra Suite in the west Musgrave Province and implications for orthomagmatic Ni-Cu-PGE mineralization during the Giles Event; Geological Survey of Western Australia, Record 2009/16, 16pp.

Howard, HM, Smithies, RH, Werner, M, Kirkland, CL, and Wingate, MTD 2011a, Geochemical characteristics of the Alcurra Dolerite (Giles Event) and its extrusive equivalents in the Bentley Supergroup; Geological Survey of Western Australia, Record 2011/2, p. 27-30.

Howard, H.M., Werner, M., Smithies, R.H., Kirkland, C.L., Kelsey, D.L., Hand, M., Collins, A., Pirajno, F., Wingate, M.T.D., Maier, W.D., Raimondo, T., 2011b. The geology of the west Musgrave Province and the Bentley Supergroup - a field guide. Geological Survey of Western Australia Record 2011/4, pp. 119.

Irvine, TN, Andersen, JCO and Brooks, CK 1998, Included blocks (and blocks within blocks) in the Skaergaard Intrusion: Geological relations and the origins of rhythmic modally graded layers: Geological Society of America, Bulliten 110, p. 1398-1447.

Jugo, PJ, Luth, RW and Richards, JP 2005, An experimental study of the sulfur content in basaltic melts satu-rated with immiscible sulfide or sulfate liquids at 1300 C and 1.0 GPa: Journal of Petrology, v. 46, p. 783-798.

Kelsey, DE, Hand, M, Evins, P, Clark, C and Smithies, H 2009, High temperature, high geothermal gradient metamorphism in the Musgrave Province, central Australia; potential constraints on tectonic setting. Kangaroo Island 2009 conference. Specalist Group in Geochemistry, Mineralogy and Petrology, Geological Society of Australia. Abstracts.

Kelsey, DE, Hand, M, Smithies, H, Evins, P, Clark, C and Kirkland, CL 2010, What is the tectonic setting of long-lived Grenvillian-aged ultrahigh temperature, high geothermal gradient metamorphism in the Musgrave Province, central Australia?: Geological Society of America Abstracts with Programs, v. 42(5), Denver, Colorado.

Kirkland, CL, Smithies, RH, Woodhouse, AJ, Howard, HM, Wingate, MTD, Belousova, EA, Cliff, JB, Murphy, RC and Spaggiari, CV 2013, Constraints and deception in the isotopic record; the crustal evolution of the west Musgrave Province, central Australia: Gondwana Research, v. 23, p.759-781.

Lesher, CM and Burnham, OM 2001, Multicomponent elemental and isotopic mixing in Ni-Cu-(PGE) ores at Kambalda, Western Australia: Canadian Mineralogist, v. 39, p. 421-446.

Li, C, Ripley, EM, Oberthür, T, Miller, JD Jr and Joslin, GD 2008, Textural, mineralogical and stable isotope studies of hydrothermal alteration in the main sulfide zone of the Great Dyke, Zimbabwe and the precious metals zone of the Sonju Lake Intrusion, Minnesota, USA: Mineralium Deposita, v. 43, p. 97-110.

Liebenberg, L 1970, The sulfides in the layered sequence of the Bushveld igneous Complex: Geological Society of South Africa, Special Publication, v. 1, p. 108-208.

Lipin, BR 1993, Pressure increases, the formation of chromite seams, and the development of the Ultramafic series in the Stillwater Complex, Montana: Journal of Petrology, v. 34, p. 955-976.

Maier, WD 2005, Platinum-group element (PGE) deposits and occurrences: mineralization styles, genetic concepts and exploration criteria: Journal of African Earth Sciences, v. 41, p. 165-191.

Maier, WD, Barnes, SJ, Gartz, V and Andrews, G  2003b, Pt-Pd reefs in magnetitites of the Stella layered intrusion, South Africa: a world of new exploration opportunities for platinum group elements: Geology, v. 31, p. 885-888.

Maier, WD, Rasmussen, B, Li, C, Barnes, S-J and Huhma, H 2013a, The Kunene anorthosite complex, Namibia, and its satellite bodies: geochemistry, geochronology and economic potential: Economic Geology (in press).

Maier, WD, Barnes, S-J and Groves, DI 2013b, The Bushveld Complex, South Africa: Formation of platinum-palladium, chrome and vanadium- rich layers via hydrodynamic sorting of a mobilized cumulate slurry in a large, relatively slowly cooling, subsiding magma chamber: Mineralium Deposita, v. 48, p. 1-56.

Maier, W.D., Howard, H.M., Smithies, R.H., Yang, S., Barnes, S.-J., 2014. Mafic–ultramafic intrusions of the Giles Event, Western Australia: petrogenesis and prospectively formagmatic ore deposits. Geological Survey of Western Australia (GSWA Record, in preparation).

McBirney, AR and Noyes, RM 1979, Crystallization and layering of the Skaergaard intrusion: Journal of Petrology, v. 20, p. 487-554.

McDonough, WF and Sun, S.-s 1995, The composition of the Earth: Chemical Geology 120, 223-253.

Morris, PA and Pirajno, F 2005, Geology, geochemistry, and mineralization potential of Mesoproterozoic sill complexes of the Bangemall Supergroup, Western Australia: Western Australia Geological Survey of Western Australia. Report 99.

Nesbitt, RW and Talbot, JL 1966, The layered basic and ultrabasic intrusives of the Giles Complex, central Australia: Contributions to Mineralogy and Petrology, v.13, p.1-11.

Nesbitt, RW, Goode, ADT, Moore, AC and Hopwood, TP 1970, The Giles Complex, central Australia:  a stratified sequence of mafic and ultramafic intrusions: Special Publications of the Geological Society of South Africa, v.1, p. 547-564.

Philpotts, AR 1967, Origin of certain iron–titanium oxide and apatite rocks: Economic Geology, v 62, p. 303–315.

Pirajno F, Smithies, HR and Howard, HM, 2006, Mineralisation associated with the 1076 Ma Giles mafic-ultramafic intrusions, Musgrave Complex, central Australia: a review, SGA News, v. 20: 1-20.

Raimondo, T, Collins, AS, Hand, M, Walker-Hallam, A, Smithies, RH, Evins, PM and Howard, HM 2009, Ediacaran intracontinental channel flow: Geology, v. 37, p. 291-294.

Raimondo, T, Collins, AS, Hand, M, Walker-Hallam, A, Smithies, RH, Evins, PM and Howard, HM 2010, The anatomy of a deep intracontinental orogen: Tectonics, v. 29, TC4024, doi:10.1029/2009TC002504, 2010.

Scoates, J.S., and Mitchell, J.N., 2000, The evolution of troctolitic and high-Al basaltic magmas in Proterozoic anorthosite plutonic suites and implications for the Voisey's Bay massive Ni-Cu sulfide deposit: Economic Geology, v. 95, p. 677-701.

Scrimgeour, IR and Close, DF 1999, Regional high pressure metamorphism during intracratonic deformation: the Petermann orogeny, central Australia: Journal of Metamorphic Geology, v. 17, p. 557-572.

Seat, Z 2008, Geology, Petrology, Mineral and Whole-rock Chemistry, Stable and Radiogenic Isotope Systematics and Ni-Cu-PGE Mineralisation of the Nebo-Babel Intrusion, west Musgrave, Western Australia: University of Western Australia, PhD thesis (unpublished).

Seat, Z, Beresford, SW, Grguric, BA, Waugh, RS, Hronsky, JMA, Gee, MMA, Groves, DI and Mathison, CI   2007, Architecture and emplacement of th Nebo-Babel gabbro-norite-hsoted magmatic Ni-Cu-PGE sulfide deposit, West Musgrave, Western Australia: Mineralium Deposita, v. 42, p. 551-582.

Seat, Z, Beresford, SW, Grguric, BA, Gee, MA and Grassineau, NV  2009, Reevaluation of the role of external sulfur addition in the genesis of Ni-Cu-PGE deposits: evidence from the Nebo-Babel Ni-Cu-PGE deposit: West Musgrave, Western Australia: Economic Geology, v. 104, p. 521-538.

Silver, PG, Behn, MD, Kelley, K, Schmitz, M and Savage, B 2006, Understanding cratonic flood basalts: Earth and Planetary Science Letters, v. 245, p.190-201.

Smithies, RH, Howard, HM, Evins, PM, Kirkland, CL, Bodorkos, S and Wingate, MTD 2009, The west Musgrave Complex – some new geological insights from recent mapping, geochronology, and geochemical studies: Geological Survey of Western Australia, Record 2008/19.

Smithies, RH, Howard, HM, Evins, PM, Kirkland, CL, Kelsey, DE, Hand, M, Wingate, MTD, Collins, AS, Belousova, E and Allchurch, S 2010, Geochemistry, geochronology and petrogenesis of Mesoproterozoic felsic rocks in the western Musgrave Province of central Australia, and implication for the Mesoproterozoic tectonic evolution of the region: Geological Survey of Western Australia, Report 106, 73p.

Smithies, RH, Howard, HM, Evins, PM, Kirkland, CL, Kelsey, DE, Hand, M, Wingate, MTD, Collins, AS and Belousova, E 2011, Mesoproterozoic High Temperature Granite Magmatism, Crust-Mantle Interaction and the Intracontinental Evolution of the Musgrave Province: Journal of Petrology; doi:10.1093/petrology/egr010.

Smithies, RH, Howard, HM, Kirkland, CL, Werner, M, Medlin, CC, Wingate, MTD and Cliff, JB 2013, Geochemical evolution of rhyolites of the Talbot Sub-basin and associated felsic units of the Warakurna Supersuite: Geological Survey of Western Australia, Report 118, 74p.

Smithies, RH, Kirkland, CL, Korhonen, FJ, Aitken, ARA, Howard, HM, Maier, WD, Wingate, MTD, Quentin de Gromard, R and Gessner, K 2014, The Mesoproterozoic thermal evolution of the Musgrave Province in central Australia – Plume vs. the geological record: Gondwana Research, http://dx.doi.org?10.1016/j.gr.2013.12.014

Sprigg, RC and Wilson, RB 1959, The musgrave mountain belt in South Australia: Geologische Rundschau, v.47, p. 531-542.

Staubmann, M 2010, The Petrogenesis and Economic Potential of the Southern Cavenagh Range Intrusion (West Musgraves), Western Australia: BSc Hns thesis (unpubl.), Univ Tasmania.

Sun S.-s., & McDonough WF (1989) Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes in: Saunders, A. D. & Norry, M. J. (eds), 1989, Magmatism in the Ocean Basins, Geological Society Special Publication No. 42, pp. 313-345.

Sun, S-S, Sheraton, JW, Glikson, AY and Stewart, AJ 1996, A major magmatic event during 1050-1080 Ma in central Australia, and an emplacement age for the Giles Complex:  AGSO Journal of Australian Geology and Geophysics, 24, p. 13-15.

Toplis, MJ and Carroll, MR 1995, An experimental-study of the influence of oxygen fugacity on Fe–Ti oxide stability, phase-relations, and mineral-melt equilibria in ferro-basaltic systems: Journal of Petrology, v. 36, p. 1137–1170.

Toplis, MJ and Carroll, MR 1996, Differentiation of ferro-basaltic magmas under conditions open and closed to oxygen: Implications for the Skaergaard intrusion and other natural systems: Journal of Petrology, v. 37, p. 837-858.

Ulmer, GC 1969, Experimental investigation of chromite spinels: Economic Geology, Monograph 4, p 114-131.

Wade, BP 2006, Unravelling the tectonic framework of the Musgrave Province, central Australia. PhD thesis, Univ Adelaide, 219pp.

Wade, BP, Kelsey, DE, Hand, M and Barovich, KM 2008, The Musgrave Province: Stitching north, west and south Australia: Precambrian Research, v. 166, p. 370-386.

White, RW, Clarke, GL and Nelson, DR 1999, SHRIMP U-Pb zircon dating of Grenville-age events in the western part of the Musgrave Block, central Australia: Journal of Metamorphic Geology, v. 17, p. 465-481.

Wingate, MTD, Pirajno, F and Morris, PA 2004, Warakurna large igneous province: a new Mesoproterozoic large igneous province in west-central Australia: Geology, v. 32, p. 105–108.

Zhao, J-X, McCulloch, MT and Korsch, RJ 1994, Characterisation of a plume-related ~ 800 Ma magmatic event and its implications for basin formation in central-southern Australia: Earth and Planetary Science Letters, v. 121, p. 349-367.

Zhou, M-F, Robinson, PT, Lesher, CM, Keays, RR, Zhang, C-J and Malpas, J 2005, Geochemistry, Petrogenesis and Metallogenesis of the Panzhihua Gabbroic Layered Intrusion and Associated Fe–Ti–V Oxide Deposits, Sichuan Province, SW China: Journal of Petrology, v. 46, p. 2253-2280.