Megan Williams and Anthony Reid
Geological Survey of South Australia, Department for Energy and Mining
Download this article as a PDF (11 MB); cite as MESA Journal 94, pages 4–18
Published April 2021
Introduction
Archean to earliest Paleoproterozoic rocks of the Gawler Craton are exposed in the central Gawler Craton, where they form a lithostratigraphic package known as the Mulgathing Complex (Fig 1; Daly and Fanning 1993). The Mulgathing Complex contains a diverse set of metamorphosed and deformed lithologies. These include komatiite, gabbro, amphibolite, tonalite and other granitoids, volcanic and volcaniclastic lithologies, and metasedimentary schists to migmatitic gneisses. As a result, the Mulgathing Complex has the potential to host a diverse set of mineral commodities, including gold, nickel–copper – platinum group elements (PGE) and iron (magnetite). The most significant economic commodity of the Mulgathing Complex has been gold; however, there has also been considerable interest in the mafic–ultramafic lithologies for associated nickel mineralisation, and the magnetite gneisses for iron ore (Davies 2000; Iron Road Ltd 2014).
The majority of the rock types within the Mulgathing Complex have undergone amphibolite to granulite facies metamorphism (e.g. Halpin and Reid 2016), with the exception of the eastern region where rocks of the Devils Playground Volcanics have been metamorphosed to greenschist facies (Reid and Daly 2009; Reid et al. 2014). Metamorphism in the complex occurred during the Sleafordian Orogeny (c. 2475–2390 Ma) with peak metamorphism reached at c. 2470–2440 Ma (Jagodzinski et al. 2009; Tomkins and Mavrogenes 2002). Variable reworking during younger Proterozoic events has affected the Mulgathing Complex, largely in the form of shear zones (e.g. Reid et al. 2007; Stewart and Betts 2010).
The Mulgathing Complex comprises a series of elongate, northeast-trending fault-bound blocks, some of which contain quite distinct lithologies (Daly and Fanning 1993; Doublier et al. 2015; Reid and Dutch 2015). These blocks are subparallel to the foliation in the quartzofeldspathic rocks and the trend of marker units (e.g. iron formations and mafic units; Daly and Fanning 1993; Hoatson et al. 2005).
The purpose of this article is to provide a review of the lithostratigraphic units of the Mulgathing Complex and discuss some aspects of the mineral potential of the region. This is motivated by a recent review of the lithostratigraphy as part of the Geological Survey of South Australia’s Digital Geology project, aiming to provide comprehensive and up-to-date stratigraphic notes for South Australia. As part of the stratigraphic review, it has become clear that the recent years have produced a raft of new literature on the Mulgathing Complex and this article seeks to review this and provide a summary of the key units as a resource for mineral exploration in the region. New units identified in the review are included in the discussion and formally defined. While significant challenges in defining the stratigraphy of the Mulgathing Complex remain, continued geological investigations in the region coupled with improved capture of legacy data and application of emerging data science tools will shed new light on this old region of the Gawler Craton.
Stratigraphic framework of the Mulgathing Complex
The Mulgathing Complex was originally defined by Daly (1986) and is named after Mulgathing Homestead northeast of Tarcoola (see historical review by Reid and Daly 2009). The original definition included 10 units, which is expanded here following further investigation, particularly the introduction of widespread geochronology (Table 1). The different lithostratigraphic packages tend to occur in different areas with limited overlap. This means that the Mulgathing Complex can be separated into 3 main areas: undifferentiated Gawler Craton/eastern Mulgathing Complex; Harris Greenstone Domain; and the Christie and Wilgena domains (Fig 1a).
Table 1 Summary of stratigraphic units from the Mulgathing Complex
Stratigraphic unit | Lithology and mineral potential | Age | Domain | SA Geodata map symbol |
Devils Playground Volcanics | Bimodal calc-alkaline volcanics, metamorphosed to greenschist facies. Potential for volcanic-related sulfide and/or epithermal gold mineralisation. | c. 2553 Ma (magmatic)1 | Eastern Mulgathing Complex | ALmd |
Sloan Hill Tonalite | Massive, hematite-stained tonalite with zones of sericite–chlorite alteration. | c. 2529 Ma (magmatic)1 | Eastern Mulgathing Complex | ALm8* |
Lake Harris Komatiite | Bright green metakomatiite with illite, tremolite, chlorite and serpentine. Saprolite float. Potential for nickel, copper and PGE. | Older than c. 2510 Ma (magmatic)2 | Harris Greenstone | ALml |
Hopeful Hill Basalt | Medium- to fine-grained metatholeiitic basalt with plagioclase, hornblende and diopside. | Older than c. 2458 Ma (magmatic)2 | Harris Greenstone | ALmh |
South Lake Gabbro | Plagioclase-rich metagabbro with hornblende, diopside and quartz. | c. 2440–2430 Ma (metamorphic)3 | Harris Greenstone | ALms |
Kenella Gneiss | Orange-pink to reddish-pink volcanic to volcaniclastic gneiss with weak compositional zoning. Minor banded iron formation, calc-silicate rock and metasediment. Hosts gold mineralisation in quartz veins. |
c. 2520–2470 Ma (magmatic)1, 4, 5 c. 2475–2390 Ma (metamorphic)1, 4 | Harris Greenstone and Wilgena | ALmk |
Glenloth Granite | Pink-brown to grey granite to granodiorite and quartz monzonite with quartz, microcline and plagioclase. Magmatic foliation. Hosts gold mineralisation in quartz veins. |
c. 2509–2507 Ma (magmatic)2, 6 c. 2436 Ma (metamorphic)3, 4 | Harris Greenstone | ALmg |
Mobella Tonalite | Grey, coarse-grained, poorly foliated tonalite–quartz monzonite. | c. 2490 Ma (magmatic)6 | Christie and Wilgena | ALmm |
Burden Metagabbro | Gabbroic gneiss with plagioclase, hornblende and possible magnetite. Poikioblastic hornblende with abundant plagioclase inclusions (former pyroxene). Previously included in the Aristarchus Metaperidotite (Peridotite). | c. 2488 Ma (magmatic)6 | Christie and Wilgena | Alma* |
Christie Gneiss | Predominantly metasedimentary gneiss with migmatitic layers. Paragneiss, magnetite gneiss (George Hill Iron Member), marble, calcsilicate and quartzite with minor volcaniclastic and mafic lithologies. Intruded by the Aurora Tank Suite. Hosts Challenger and Golf Bore gold deposits. Hosts iron mineralisation. |
c. 2485–2480 Ma (maximum deposition)4, 6 c. 2470–2415 Ma (metamorphic)6 | Christie and Wilgena | ALmc |
George Hill Iron Member | Coarse-grained magnetite gneiss. Member of the Christie Gneiss. | Unknown | Christie and Wilgena | ALmc2* |
Aurora Tank Suite | Granitic gneiss and tabular feldspar granite with feldspar, quartz, biotite and locally hornblende. |
c. 2475–2470 Ma (magmatic)1, 4, 6 c. 2470–2390 Ma (metamorphic)4, 7 | Christie and Wilgena | Not assigned* |
Aristarchus Metaperidotite | Amphibolite dykes. Intrudes the Aurora Tank Suite. Contains elevated nickel contents. | Younger than c. 2475 Ma (magmatic)4, 6 | Christie and Wilgena | ALma |
Skuse Hill Metapyroxenite | Garnet-bearing mafic gneiss, metagabbro, metapyroxenite and metanorite. Concordant with carbonate and iron formation and paragneiss. Traces of nickel–copper sulfide mineralisation. |
c. 2460 Ma (magmatic)2 c. 2435–2425 Ma (metamorphic)2 | Christie and Wilgena | ALmb* |
Unnamed mafic lithology | Undifferentiated metamafic rocks. | Unknown | Christie and Wilgena | ALm2 |
Age reference: 1 Jagodzinski and Reid (2017); 2 Fanning et al. (2007); 3 Fanning (1997); 4 Jagodzinski et al. (2009); 5 Swain et al. (2005); 6 Reid et al. (2014); 7 McFarlane (2006).
* Current assignment to be revised.
Source: modified from Daly and Fanning (1993) and Reid et al. (2014).
Eastern Mulgathing Complex
In the far east of the Mulgathing Complex (‘undifferentiated Gawler Craton’ in the terminology of Ferris et al. 2002; Fig 1a), 2 formations have been intersected in drillholes: the Devils Playground Volcanics (Fig 2a) and the Sloan Hill Tonalite (new name; Fig 2b; App). The Devils Playground Volcanics comprise c. 2553 Ma bimodal calc-alkaline volcanics that were metamorphosed to greenschist facies (Cowley and Fanning 1991; Jagodzinski and Reid 2017; Reid et al. 2009; Reid et al. 2017). The volcanics have potential for volcanic-related sulfide and/or epithermal gold mineralisation (Eromanga Uranium Ltd 2008). The Sloan Hill Tonalite is a massive granite–tonalite emplaced at c. 2529 Ma. It has extensive hematite staining and zones of sericite–chlorite alteration (Jagodzinski and Reid 2017). These 2 units are the oldest in the Mulgathing Complex.
Harris Greenstone Domain
The Harris Greenstone Domain contains the next oldest lithologies of the Mulgathing Complex (Fig 1a). This region comprises the Kenella Gneiss (c. 2520 Ma; Fanning et al. 2007; Jagodzinski et al. 2009; Swain et al. 2005), which is interlayered with greenstones of the Lake Harris Komatiite (minimum c. 2510 Ma), Hopeful Hill Basalt (minimum c. 2458 Ma) and South Lake Gabbro (minimum c. 2440 Ma), and intruded by the Glenloth Granite (c. 2510 Ma; Fanning et al. 2007) and younger, unnamed granitoids (c. 2460 Ma; Fanning 1997). Bedrock drilling in the area uncovered a range of mostly igneous lithologies which are not captured by the current stratigraphic framework (Davies 2003).
The Kenella Gneiss is generally interpreted as a predominantly volcanic and volcaniclastic unit (characteristically red-pink; Fig 2c), but could alternatively be a massive, source proximal, quartz-rich sediment (Daly and van der Stelt 1992). The lithology is interbedded with minor metasediments, including arenaceous and aluminous gneisses, iron formation and carbonates (Daly and Fanning 1993). These metasediments are sometimes attributed to the Christie Gneiss, but limited description and detailed mapping of the different lithologies makes the distinction between the Christie and Kenella gneisses somewhat unclear (see discussion below). The volcaniclastic rocks have equivalents further north in the Christie Domain, which are generally included in the Kenella Gneiss, but occasionally in the predominantly metasedimentary Christie Gneiss (Daly and Fanning 1993; McFarlane et al. 2007).
In both the Harris Greenstone and Christie domains, bright-green clay and chalcedony are characteristic of shallow weathered mafic–ultramafic rocks, which rarely outcrop (Daly and van der Stelt 1992). Fresh examples of the Lake Harris Komatiite (Fig 2d) are serpentinised and preserve ponded flows and channel facies including spinifex textures (characteristic extrusive textures; Daly and van der Stelt 1992; Davies 2003). The Hopeful Hill Basalt shows pillow structures, characteristic of subaqueous deposition (Daly and Fanning 1993). The Lake Harris Komatiite and Hopeful Hill Basalt are thought to be genetically related, with the Hopeful Hill Basalt forming through fractional crystallisation from the same source as the Lake Harris Komatiite (Coppens 1997). It is unclear how the komatiite and basalt stratigraphically relate to Kenella Gneiss (Daly and van der Stelt 1992; Davies 2003). The other mafic unit in the Harris Greenstone Domain, the South Lake Gabbro, consists of thin (1–2 m) sill-like metagabbro bodies (Daly and Fanning 1993). The intrusive age and emplacement style of the South Lake Gabbro are not well constrained, making it unclear whether the South Lake Gabbro is associated with the Lake Harris Komatiite and Hopeful Hill Basalt. Although the 3 units are often grouped together, the Lake Harris Komatiite is more extensively studied than the Hopeful Hill Basalt and the South Lake Gabbro and therefore dominates discussion. In addition, the ages of these 3 units are not tightly constrained, making it problematic to definitively correlate or group them.
The Glenloth Granite intrudes the Kenella Gneiss and crosscuts the mafic units of the Harris Greenstone Belt in drill core (Fanning et al. 2007). The granite is foliated and contains schlieren, interpreted as primary magmatic flow textures (Fig 2e; Reid et al. 2007).
The Harris Greenstone Domain is prospective for nickel (Ni–Cu and Ni–PGE) within mafic–ultramafic units (Daly and Fanning 1993; Hoatson et al. 2005), although no significant mineralised zones have yet been intersected (Hoatson et al. 2005; Price and Price 2015). The Harris Greenstone Domain is also host to historical gold mines, with gold-bearing quartz veins, thought to be related to the Mesoproterozoic Hiltaba Suite or Gawler Range Volcanics (Daly and Fanning 1993; Ferris and Schwarz 2003).
Christie and Wilgena domains
By far the most extensive formation in the Mulgathing Complex is the Christie Gneiss. This unit occurs over >25,000 km2, forming an arcuate belt that extends from north of Tarcoola to the western edge of the Gawler Craton. The basement rocks are largely covered by a veneer of Cenozoic sediment. The Christie Gneiss is predominately composed of felsic migmatitic garnet–biotite gneisses (Fig 2f), which generally have metasedimentary protoliths, with the exception of the ‘Challenger Gneiss’ which has a volcaniclastic protolith (included in the Christie Gneiss through convention; Fig 3a Daly and Fanning 1993; McFarlane et al. 2007). Other volcanic or volcaniclastic lithologies in the area are attributed to the Kenella Gneiss, although they can be younger than the Kenella Gneiss in the Harris Greenstone Domain (c. 2502 Ma; Daly and Fanning 1993; Jagodzinski et al. 2009). The formation has a maximum depositional age of c. 2485–2480 Ma and was metamorphosed to amphibolite–granulite facies conditions at c. 2470–2440 Ma (Halpin and Reid 2016; Jagodzinski et al. 2009; Reid et al. 2014). Given the lack of exposure of the metasedimentary rocks in this region and the area over which they are distributed, it is likely that there are multiple, as yet undifferentiated packages of sedimentary rock within the Christie Gneiss (Reid et al. 2014). The felsic gneisses host several gold deposits, the most significant of which are Challenger and Golf Bore (Birt and Reid 2007; see discussion below).
Interlayered with the felsic gneisses are coarse-grained magnetite gneiss of the George Hill Iron Member (new name; Fig 3b; App), olivine-bearing carbonate and calc-silicate lithologies, all strongly metamorphosed and deformed multiple times (Daly and Fanning 1993). The George Hill Iron Member is prospective for iron ore and has several named deposits near Mount Christie (Iron Road 2014). The Christie Gneiss also includes minor mafic gneisses (Christie Gneiss units 1 and 3, and some unassigned units; Reid et al. 2007; Tomkins and Mavrogenes 2001, 2002).
The Christie Gneiss is intruded by multiple phases of magmatism which have intrusive ages c. 2490–2470 Ma (Figs 3c, d, e; Jagodzinski and Reid 2017; Jagodzinski et al. 2009). Near Mobella Station (northwestern Mulgathing Complex), the Christie Gneiss is intruded by the c. 2490 Ma Mobella Tonalite (grey, coarse-grained, poorly foliated tonalite–quartz monzonite; Fig 3c; Daly and Fanning 1993; Jagodzinski et al. 2009). Near Mount Christie (southwestern Mulgathing Complex), the Christie Gneiss is intruded by the c. 2490 Ma Burden Metagabbro (new name; plagioclase–hornblende gabbroic gneiss; Fig 3d; App), the younger Aristarchus Metaperidotite (formerly Aristarchus Peridotite; maximum age c. 2475 Ma; Fig 3e; App) and the c. 2460 Ma Skuse Hill Metapyroxenite (redefined name; metamorphosed mafic–ultramafic ?sills; App; Jagodzinski et al. 2009; Reid et al. 2007). The Skuse Hill Metapyroxenite is here renamed because the former name, the Blackfellow Hill Pyroxenite (Daly and Fanning 1993), is considered culturally exclusive and socially inappropriate. The former Aristarchus Peridotite (now Aristarchus Metaperidotite and Burden Metagabbro) was interpreted as a layered mafic intrusion (Daly and Fanning 1993; Daly and van der Stelt 1992). These mafic–ultramafic lithologies are now known to be different ages and so this interpretation is no longer valid (Jagodzinski et al. 2009). The Skuse Hill Metapyroxenite has also been interpreted to be a layered mafic complex (Daly and van der Stelt 1992). Outcrops of Mulgathing Complex rocks are known from the far western Christie Domain and include c. 2486 Ma metanorite (Reid et al. 2014), but these units have received very limited attention and are not assigned to lithostratigraphic units (mapped as undifferentiated Mulgathing Complex). The Aurora Tank Suite (new name; App) consists of granitic lithologies intruding the Christie Gneiss and Burden Metagabbro which have magmatic ages c. 2475–2470 Ma (Fig 3f; Jagodzinski and Reid 2017; Jagodzinski et al. 2009). These include tabular feldspar granite at Aurora Tank (Fig 3f), syenogranite gneiss northwest of Commonwealth Hill Outstation and red to grey potassic-altered granite gneiss with variable assimilation of the Burden Metagabbro near Aristarchus Paddock.
Discussion
Mineral potential – key themes
Gold
Gold mineralisation in the Mulgathing Complex occurs in at least 2 distinct styles (Gum 2019): metamorphosed gold mineralisation (e.g. Challenger); and vein-style gold mineralisation (e.g. Earea Dam and Glenloth goldfield). Post-metamorphic orogenic gold has been suggested as a third distinct style (e.g. Birt and Reid 2007) although the prevalence of this style is still debated, and there is also potential for epithermal gold or volcanic-related sulfide mineralisation in the Devils Playground Volcanics.
Challenger is interpreted as a metamorphosed gold deposit with the distribution and grade of pre-existing gold mineralisation strongly affected by high-temperature deformation, partial melting and melt migration during the Sleafordian Orogeny (Fig 4a; McFarlane 2006). Gold was remobilised and is associated with leucosomes containing abundant coarse-grained retrograde biotite that occur in the short limbs of asymmetric folds (Fig 4b; Tomkins and Mavrogenes 2002). Mineralisation is thought to pre-date the granulite-facies metamorphism due to a lack of hydrothermal alteration minerals associated with gold mineralisation (Tomkins and Mavrogenes 2001, 2002). The protolith for the mineralised host is thought to be a hydrothermally altered volcaniclastic lithology (Christie Gneiss; McFarlane et al. 2007). Challenger is not exposed at the surface and was discovered by calcrete geochemical sampling (Poustie et al. 2002).
Golf Bore is less well understood than Challenger, but is most likely a lower grade version of the Challenger-style metamorphosed gold. The host rock for mineralisation is a mid-amphibolite facies, unmelted, sillimanite–biotite–muscovite (aluminous) metasediment (Fig 4c) interlayered with more massive, arenaceous biotite hornfels with garnet porphyroblasts (Fig 4d; both Christie Gneiss), intruded by garnet- and pyrite-bearing leucosomes (Fig 4e). The garnet in the leucosomes is heavily retrogressed to biotite, possibly the result of fluid in the leucosomes themselves (Fig 4e). Golf Bore has a different protolith to Challenger and is lower metamorphic grade (Greene 2016). The timing of gold mineralisation at Golf Bore is not well constrained, with the initial mineralisation occurring either before metamorphism as at Challenger (e.g. Greene 2016), or during post-metamorphic fluid flow and retrogression in a more typical orogenic gold style (e.g. Birt and Reid 2007). At Golf Bore, a 2–5 m thick, low-grade supergene blanket occurs consistently throughout the deposit. Alteration assemblages consist mainly of chlorite and lesser sericite and pyrite, within the host foliated biotite–sillimanite gneiss (Greene 2016).
Other orogenic gold deposits occur further south in the Christie and Harris Greenstone domains, hosted by iron formations and mafic–ultramafic (greenstone) lithologies (e.g. South Hilga, Black Knight, Double Dutch and Boomerang; Birt and Reid 2007; Gum 2019).
In the Harris Greenstone Domain, quartz-vein gold mineralisation, with minor hematite, is derived from Hiltaba Suite intrusions and hosted in both the Kenella Gneiss and the Glenloth Granite (e.g. Earea Dam and Glenloth goldfield). Ore grade is sporadic and patchy and gold is associated with copper, lead, arsenic, magnesium and iron, and locally associated with tin and silver (Crettenden 1990; Daly and Fanning 1993; Whitten et al. 1978). In the Glenloth goldfield, gold also occurs within altered Paleoproterozoic mafic dykes intruding the Glenloth Granite in the Monarch mine, with associated copper, molybdenum, lead, zinc, arsenic and tin (Blissett 1985). As the gold-bearing quartz veins in the area are related to Hiltaba Suite magmatism, they are therefore much younger than the gold mineralisation in the Christie Domain (Blissett 1985; Gum 2019).
The low metamorphic grade Devils Playground Volcanics are potentially convergent margin volcanics making them a conceptually appealing exploration target, with potential for volcanic-related sulfide and or epithermal gold mineralisation (Reid et al. 2009). Although the 3 available drillholes into these rocks are unmineralised, the volcanics show pervasive chlorite and sericite alteration in the handful of drillhole intersections, suggesting that economic mineralisation may occur elsewhere in areas subjected to more intense hydrothermal activity, or adjacent to major structures (Cowley and Fanning 1991; Eromanga Uranium 2008). In addition, 40Ar/39Ar geochronology of sericite and K-feldspar from the Sloan Hill Tonalite suggest hydrothermal alteration in this rock type occurred at c. 1.6 Ga, consistent with the timing of Hiltaba Suite and Gawler Range Volcanics related mineralisation elsewhere in the eastern Gawler Craton (Reid et al. 2017).
Nickel
There is potential for nickel mineralisation within the mafic–ultramafic lithologies of the Harris Greenstone and Christie domains. In the Harris Greenstone Domain, the Lake Harris Komatiite and Hopeful Hill Basalt have locally anomalous Ni–Cu and Ni–PGE (Ni >1,000 ppm, Cu >200 ppm; Price 2014). Samples analysed thus far are predominantly low in sulfur and have high Pd + Pt (5–30 ppb) contents, indicating sulfur undersaturation which appears to have inhibited development of massive sulfides (Hoatson et al. 2005). However, only a fraction of the inferred >300 km strike length of the mafic–ultramafic sequence in the Harris Greenstone Domain has been tested and the potential for geochemical variation along strike is high (Davies 2003; Hoatson et al. 2005). High-resolution regional magnetics available in the new Gawler Craton Airborne Survey dataset could be useful to identify more deformed zones within the mafic–ultramafic lithologies and develop new targets in the area.
In the Christie Domain, nickel mineralisation in the Aristarchus Metaperidotite is magmatic in origin and occurs as disseminated, interstitial sulfides including pentlandite and violarite (Ni <1,880 ppm, Au <10 ppb, Pt <38 ppb; Daly and van der Stelt 1992; McKinnon-Matthews 2007). Aristarchus Metaperidotite comprises pargasite–tremolite–serpentine–olivine amphibolite interpreted as metaperidotite (lherzolite; Daly and van der Stelt 1992). The metaperidotite also contains chromite, ilmenite and magnetite. The host metaperidotite has no precise age constraints but is younger than c. 2475 Ma and is only weakly foliated (Jagodzinski et al. 2009). The Skuse Hill Metapyroxenite also contains accessory sulfide and elevated nickel and chromium contents (Ni <880 ppm, Cr <2,750 ppm, Au <18 ppb; Daly and van der Stelt 1992).
Iron
The George Hill Iron Member comprises prominent outcrops of iron-rich gneiss, which contain 30–60% magnetite–martite. The gneiss has a characteristic aeromagnetic signature which indicates strike lengths of 40 km, with complex fold structures (Daly and Fanning 1993). The magnetite gneiss is of particular interest due to the typically large grains of magnetite (100–500 µm) which improves ore potential (Iron Road 2014). Significant prospects occur at Mount Christie, Fingerpost Hill, Commonwealth Hill, Ibis and Sequoia (Davies 2000).
Unravelling the mafic intrusions
While the mafic–ultramafic units described above seem similar in their simplified descriptions, there is strong evidence for at least 3 suites/sequences of mafic–ultramafic rocks in the Mulgathing Complex. The oldest sequence includes the c. 2510 Ma Lake Harris Komatiite and probably the Hopeful Hill Basalt (older than c. 2458 Ma). These units are inferred to be genetically related (Coppens 1997; Hoatson et al. 2005) and are grouped on the basis that they occur within parallel bands with similar magnetic characteristics. They are therefore both interpreted to be Neoarchean (c. 2510 Ma; Hoatson et al. 2005) on the basis of the minimum age obtained from the Lake Harris Komatiite. The South Lake Gabbro (older than c. 2440 Ma) may be an intrusive equivalent of the Hopeful Hill Basalt but occurs within the main package of the Kenella Gneiss and has significantly higher calcium than other mafic units in the Mulgathing Complex (SA Geodata sample 9694 compared with geochemistry in Daly and van der Stelt 1992). Further work is needed to clarify the stratigraphic relationships between the mafic–ultramafic lithologies in the Harris Greenstone Domain.
The mafic units in the Christie Domain appear to be distinctly younger. The Burden Metagabbro and Skuse Hill Metapyroxenite have magmatic ages of c. 2490 Ma and c. 2460 Ma, respectively. Mafic rocks at Skuse Hill are weakly magnetic and associated with iron formation ± carbonate rocks (Daly and van der Stelt 1992). In contrast, the mafic–ultramafic rocks at Aristarchus are more magnetic and not associated with the Christie Gneiss. Instead, they are associated with granite gneiss of the Aurora Tank Suite. Nickel mineralisation has not been intersected in the Burden Metagabbro (c. 2490 Ma; Aristarchus prospect), so improving understanding of the mafic stratigraphy in the area may help to target future exploration (McKinnon-Matthews 2007; Reid et al. 2014).
The Aristarchus Metaperidotite is known from lithostratigraphic correlations to be younger than the Burden Metagabbro. The Aristarchus Metaperidotite crosscuts granitic gneiss of the Aurora Tank Suite, dated at c. 2474 Ma (e.g. drillcore DMDD-002; Jagodzinski et al. 2009; McKinnon-Matthews 2007). The metaperidotite dykes have sharp contacts with the granite gneiss which is distinctly different to the diffuse and sometimes gradational contacts between the Aurora Tank Suite and Burden Metagabbro in this location. The granite provides a maximum age for the Aristarchus Metaperidotite, but no minimum age exists and therefore the unit could be correlated with the c. 2460 Ma Skuse Hill Metapyroxenite, or be a younger intrusion. Additional mafic bodies are thought to occur elsewhere in the Christie Domain, but their affinity is unknown at this stage (Daly and van der Stelt 1992).
Challenges of developing a stratigraphy of poorly outcropping and lithologically complex regions
The Mulgathing Complex is dominated by felsic lithologies with low magnetic response and the gravity features in the region are dominantly long wavelength encompassing wide regions of lithostratigraphy (Fig 1). These lithologies often have very similar mineralogy (quartz–feldspar–garnet–biotite), but varied protoliths including felsic igneous, intermediate volcaniclastic and sedimentary rock types (Halpin and Reid 2016; McFarlane et al. 2007; Reid et al. 2007). This makes identifying stratigraphic units within the Mulgathing Complex a challenge, particularly with poor and sporadic outcrop and relatively sparse diamond drilling. However, the example of the Challenger deposit, where mineralisation is hosted in altered volcaniclastic lithologies and not surrounding metasedimentary lithologies, shows how important the distinction between these rock types can be (McFarlane et al. 2007).
Currently, much of the stratigraphic correlation is made through geochronology (Fanning et al. 2007; Jagodzinski and Reid 2017; Jagodzinski et al. 2009). This technique has been used to identify several previously unknown stratigraphic units in the past decade or so (e.g. Aurora Tank Suite, Burden Metagabbro). The felsic lithologies which have been studied in detail show some differences in rock textures and mineralogy which may be related to different protoliths. This interpretation seems to be backed up by small variations in age spectra of different lithologies (Fanning et al. 2007; Jagodzinski et al. 2009; Reid et al. 2007). However, the relatively small number of samples that have been analysed mean wider conclusions about lithostratigraphic variation within units such as the Christie Gneiss are difficult at present.
Iron formations and some mafic units within the complex have a high magnetic response and can be used to trace these lithologies due to the sharp contrast with surrounding rocks. However, here has been a tendency to correlate distinctive lithologies (e.g. iron formations) or packages of lithologies across large areas. This can be problematic when the correlation is carried over faults or large distances as the continuity of the lithostratigraphic package becomes uncertain. The relative timing of sedimentary, igneous and metamorphic events in the Mulgathing Complex have also been identified by careful study of textural and structural relationships in core which extends the value of sparse geochronology analyses.
Future work
Geochronology and lithogeochemistry are currently the most useful tools for differentiating lithologies, and geochronology is the only reliable way to provide correlation across similar looking lithologies. These are key to understanding which horizons have mineral potential and how they can be reliably recognised across the region. Future work should aim to expand the available geochronology and lithogeochemistry for the Mulgathing Complex to aid in clarifying the stratigraphy. Recent advancements in machine learning techniques applied to geoscience show promise as tools to distinguish between similar rock types based on chemistry and could also provide useful insights into the Mulgathing Complex and other poorly characterised terranes. Interpretation of the Gawler Craton Airborne Survey aeromagnetic dataset will bring a clearer picture of the structures in the area which should aid in regional stratigraphic correlation. Future geophysics data acquisition, field mapping and legacy data capture and analysis could assist in understanding the deeper crustal structure of the Mulgathing Complex leading to a better understanding of mineral potential in the region.
Stratigraphic drilling has revealed much of what we know about the Mulgathing Complex. However, of the 22,886 drillholes over the Mulgathing Complex region (many of which do not penetrate the cover sequences), only 853 are diamond drillholes, 638 of which are at Challenger mine and surrounds. There are 59 diamond drill cores in the South Australia Drill Core Reference Library which have significant (>5 m) intersections of Mulgathing Complex stratigraphy, that cover an area of >6,000 km2 (Fig 1c; Daly and Fanning 1993). While this is only a fraction of the total number of cores drilled, only a handful of these cores have been studied in any detail. Therefore, more can be done to reveal the stratigraphic variation within these cores and extract maximum value from legacy drilling samples. With the increasing use of data science and machine learning to predict mineralisation in the Gawler Craton, quality stratigraphic input data is essential to retrieve meaningful outputs from these techniques.
Conclusion
The Gawler Craton is South Australia’s most economically significant geological terrane. Iron oxide – copper–gold deposits in the region account for 3 major mines – i.e. Olympic Dam, Prominent Hill and Carrapateena. However, the Gawler Craton has potential for a range of other commodities including gold, nickel–copper–PGE, and magnetite, all of which are represented to varying degree within the Archean to earliest Paleoproterozoic Mulgathing Complex. Understanding the stratigraphy of the Mulgathing Complex and the distribution of the units at greater than deposit-scale has the potential to open up new areas and new mineral systems to exploration.
Appendix: Stratigraphic definitions
Sloan Hill Tonalite
New name
Derivation of name. Sloan Hill, BILLA KALINA, SH53 07, 1:250,000 map sheet. Located at GDA2020, MGA53, 565570 mE, 6681525 mN.
Synonymy. Informally known as ‘hematite-stained tonalite’ (Jagodzinski and Reid 2017) and ‘Mulgathing Complex unit 8’.
Parent unit. Mulgathing Complex.
Type locality. Drillhole (DDH) SH 7C (SA Geodata drillhole 237530), interval 1,427–1,467.5 m. Drilled by Uranium Exploration Australia Ltd in 2009. Located at GDA2020, MGA53, 569219 mE, 6682981 mN.
Distribution. Intersected in drillholes in the eastern Mulgathing Complex near Sloan Hill. No known outcrop.
Lithology. Massive biotite-bearing tonalite, with abundant plagioclase, minor K-feldspar occurring as rims on plagioclase and minor muscovite. Granite with equigranular quartz, K-feldspar and lesser plagioclase also occurs. Variably hematite altered with zones of sericite–chlorite alteration (Jagodzinski and Reid 2017; Reid et al. 2017).
Thickness. The minimum thickness of the Sloan Hill Tonalite is 40.5 m, as intersected in the type drillhole (base of lithology not intersected).
Relationships and boundary criteria. Unconformably overlain by Pandurra Formation. Relationship with other Mulgathing Complex units unknown.
Age. Neoarchean; magmatic age c. 2530 Ma (SHRIMP U–Pb zircon; Jagodzinski and Reid 2017).
Correlation. The Sloan Hill Tonalite may be a corelative of a c. 2526 Ma granite gneiss intersected in GOMA 4 DDH (Jagodzinski et al. 2013).
Burden Metagabbro
New name
Derivation of name. Burden Tank, TARCOOLA, SH53 10, 1:250,000 map sheet. Located at GDA2020, MGA53, 366804 mE, 6642670 mN.
Synonymy. Previously an unnamed lithology within the Aristarchus Peridotite (now Aristarchus Metaperidotite).
Parent unit. Mulgathing Complex.
Type locality. Drillhole DMDD-002 (234168), interval 96–318 m (total depth). Drilled by Dominion Mining in 2007. Located at GDA2020, MGA53, 367793 mE, 6634851 mN.
Distribution. Occurs in drillholes near Aristarchus Paddock in the southern Christie Domain.
Lithology. Gabbroic gneiss with plagioclase, hornblende and opaques. Hornblende is poikiloblastic and contains abundant plagioclase, likely representing metamorphosed pyroxene (Reid et al. 2014).
Thickness. Occurs over >220 m in type drillhole, with individual layers 0.05 to >2 m thick.
Relationships and boundary criteria. Intruded by the Aurora Tank Suite with partially diffuse contacts.
Age. Paleoproterozoic; magmatic crystallisation at c. 2488 Ma (SHIMP U–Pb zircon; Reid et al. 2014).
Correlation. Temporal correlative of the Mobella Tonalite (Mulgathing Complex).
George Hill Iron Member
New name
Derivation of name. George Hill, TARCOOLA, SH53 10, 1:250,000 map sheet. Located at GDA2020, MGA53, 360010 mE, 6643684 mN.
Synonymy. Informally known as ‘Christie Gneiss unit 2’ (Rankin et al. 1996).
Parent unit. Christie Gneiss.
Type locality. Mount Christie (Rankin et al. 1996). Located at MGA53, 357389 mE, 6646645 mN.
Reference drillhole: GWL023 (266872), interval 24–138 m (total depth). Drilled by Dominion Gold Operations Pty Ltd in 2010 at the Boomer prospect. Located at GDA2020, MGA53, 350902 mE, 6645938 mN.
Distribution. Outcrops extensively as discontinuous bands in the southern Christie Domain, particularly near Mount Christie. Outcrops and drillhole intersections predominantly occur in a northeast–southwest-trending band, closely associated with distinct strong magnetic highs.
Lithology. Iron-rich, quartz magnetite gneiss. The predominant lithology is a medium- to coarse-grained, quartz–magnetite–diopside–hypersthene–amphibole gneiss, with magnetite, partly altered to martite, and trace sulfides. The lithology is strikingly banded green and black on a millimetre scale with local massive sections. The lithology displays complex folding with abundant open to right dextral-vergence folds (Rankin et al. 1996). The magnetite-rich gneiss contains thin interlayers and laminae of carbonate with accessory garnet, clinopyroxene and olivine and quartzofeldspathic gneiss (Daly and Fanning 1993; Rankin et al. 1996). Outcrops are weathered to goethite and have a distinctive shiny black appearance (Rankin et al. 1996).
Thickness. Typically 10–50 m thick, although significantly thicker zones are present in areas of structural thickening by isoclinal folding (Daly and Fanning 1993; Iron Road 2012).
Relationships and boundary criteria. Commonly associated with olivine-bearing metacarbonate and interlayered with quartzofeldspathic gneiss (Christie Gneiss) and metamafic lithologies (Skuse Hill Metapyroxenite; Daly and Fanning 1993).
Age. Paleoproterozoic; maximum depositional age c. 2480 Ma; metamorphosed at c. 2460–2415 Ma with ages derived from adjacent quartz–feldspar–biotite gneiss (Jagodzinski et al. 2009).
Correlation. Iron-rich lithologies also occur within the Harris Greenstone Domain, within the Kenella Gneiss. Sedimentary lithologies in the Harris Greenstone Domain may be slightly older than the George Hill Iron Member, but the ages of both lithologies are not well constrained.
Aurora Tank Suite
New name
Derivation of name. Aurora Tank, COOBER PEDY, SH53 06, 1:250,000 map sheet. Located at GDA2020, MGA53, 404643 mE, 6715420 mN.
Synonymy. Previously informally attributed to the Kenella Gneiss, Glenloth Granite or undifferentiated Mulgathing Complex.
Parent unit. Mulgathing Complex.
Type locality. Four kilometres west of Aurora Tank. Located at GDA2020, MGA53, 400471 mE, 6715483 mN.
Reference drillhole: DMDD-002 (234168), interval 196–307 m. Drilled by Dominion Mining NL in 2007. Located at GDA2020, MGA53, 367793 mE, 6634851 mN.
Distribution. The Aurora Tank Suite outcrops in the northern Gawler Craton near Aurora Tank and Commonwealth Hill Outstation and occurs in drillholes in the Aristarchus prospect.
Lithology. Metamorphosed granitoids including tabular K-feldspar granite, hematite-stained syenogranite gneiss and granoblastic granitic gneiss. At the type locality, the Aurora Tank Suite comprises porphyritic, feldspar-rich granite with elongate K-feldspar phenocrysts (Reid et al. 2014). At the reference locality, the Aurora Tank Suite comprises potassic-altered, granoblastic granitic gneiss. The granitic gneiss has variably assimilated older mafic material, and locally grades to granodiorite. The reference Aurora Tank Suite granitoid is typically orange-red. Locally, unaltered grey gneiss contains red alteration veins.
Thickness. Intersection of 313 m of Aurora Tank Suite interlayered with Burden Metagabbro in the reference drillhole (base not intersected), with individual granitoid layers 0.1 to 15 m thick.
Relationships and boundary criteria. Intrudes Christie Gneiss and Burden Metagabbro. Variable assimilation of the Burden Metagabbro with generally diffuse contacts in reference drillhole. Intruded by and occurs as enclaves within Aristarchus Metaperidotite.
Age. Paleoproterozoic; magmatic ages c. 2475–2470 Ma (SHRIMP U–Pb zircon; Jagodzinski and Reid 2017; Jagodzinski et al. 2009; Reid et al. 2014).
Correlation. Tentatively correlated to other granitoid lithologies of unknown age in the Aristarchus prospect which intrude the Christie Gneiss. Likely temporal correlative of Kiana Granite (Dutton Suite, Sleaford Complex).
Aristarchus Metaperidotite
Redefined name
Derivation of name. Aristarchus Rise, TARCOOLA, SH53 10, 1:250,000 map sheet (Daly and Fanning 1993). Located at GDA2020, MGA53, 367500 mE, 6634473 mN.
Synonymy. Previously known as the Aristarchus Peridotite. Previous definition also included the Burden Metagabbro (new name).
Parent unit. Mulgathing Complex.
Type locality. Drillhole Aristarchus 1 (3239), interval 99.5–106.1 m. Drilled by the South Australia Department of Mines and Energy, SADME, in 1991. Located at GDA2020, MGA53, 367200 mE, 6633919 mN.
Reference drillhole: DMDD-002 (234168), interval 307.44–307.89 m. Drilled by Dominion Mining in 2007. Located at GDA2020, MGA53, 367793 mE, 6634851 mN.
Distribution. Occurs in drillholes near Aristarchus Paddock in the southern Christie Domain.
Lithology. In the type drillhole, the Aristarchus Metaperidotite comprises pargasite–tremolite–serpentine–olivine amphibolite interpreted as metaperidotite (lherzolite; Daly and van der Stelt 1992). Tremolite and pargasite are interpreted to replace pyroxene oikocrysts, with a fine-grained cumulus groundmass of olivine, tremolite, serpentine, magnetite, chromite and sulfide. Hornblende–biotite schist with minor plagioclase also occurs, interpreted as a metapyroxenite (Daly and van der Stelt 1992). Samples in the type drillhole are typically heavily weathered to bright green clay, whereas samples from the reference drillhole are fresher.
In the reference drillhole, the Aristarchus Metaperidotite comprises granular hornblende–biotite metapyroxenite with sparse plagioclase, carbonate, ilmenite and titanite, and up to 8% disseminated sulfides including pyrrhotite, pentlandite, pyrite and lesser chalcopyrite (McKinnon-Matthews 2007).
Thickness. Up to 7 m, but typically 0.2–0.4 m thick in type drillhole.
Relationships and boundary criteria. Intrudes the Aurora Tank Suite and Burden Metagabbro with sharp, irregular contacts.
Age. Undated; maximum magmatic age c. 2474 Ma constrained by granite gneiss of the Aurora Tank Suite (DMDD-002; Jagodzinski et al. 2009).
Correlation. Potential correlative of the c. 2460 Ma Hill Metapyroxenite.
Skuse Hill Metapyroxenite
Redefined name
Derivation of name. Skuse Hill, TARCOOLA, SH53 10, 1:250,000 map sheet. Located at GDA2020, MGA53, 360175 mE, 6645030 mN.
Synonymy. Formerly known as ‘Blackfellow Hill Pyroxenite’ (Daly and Fanning 1993).
Parent unit. Mulgathing Complex.
Type locality. Drillhole Blackfellow Hill 1 (3243), interval 86–170 m (total depth). Drilled by the South Australia Department of Mines and Energy, SADME, in 1991. Located at GDA2020, MGA53, 359420 mE, 6643868 mN.
Reference drillhole: Skuse Hill 2 (3242), interval 18–76 m. Drilled by SADME in 1991. Located at GDA2020, MGA53, 360686 mE, 6644988 mN.
Distribution. Occurs in drillholes near Skuse Hill in the southern Christie Domain.
Lithology. The Skuse Hill Metapyroxenite comprises metapyroxenite and metagabbro–norite. Metapyroxenite occurs as pale green tremolite–phlogopite schist with some actinolite ± hornblende, variable plagioclase, minor chlorite and trace oxides and apatite. Tremolite may preserve textures of primary pyroxene. In some zones, original orthopyroxene (bronzite) is preserved forming a distinctive crescumulate texture and in others, sericitised plagioclase up to 8 mm occurs (Daly and van der Stelt 1992).
The metagabbro–norite comprises amphibole, plagioclase, quartz and biotite with minor oxides and trace zircon and apatite. Recrystallised amphibole aggregates replaced pyroxene. Biotite and zircons in this lithology are interpreted to indicate some crustal contamination. Locally, augen of plagioclase up to 1 mm occur and elsewhere garnet occurs in layers with biotite (Daly and van der Stelt 1992).
Thickness. Intersections of the Skuse Hill Metapyroxenite are up to 58 m thick in the reference drillhole, but are significantly less in other drillholes (Daly and van der Stelt 1992).
Relationships and boundary criteria. The pyroxenite lithology of the Skuse Hill Metapyroxenite occurs structurally below the discontinuous George Hill Iron Member, also associated with olivine-bearing metacarbonate (Christie Gneiss; Daly and van der Stelt 1992).
Age. Early Paleoproterozoic; magmatic crystallisation at c. 2460 Ma (SHRIMP U–Pb zircon; Fanning et al. 2007).
Correlation. Potential correlative of the Aristarchus Metaperidotite (see above).
Acknowledgements
Mark Pawley and Rian Dutch (Geological Survey) are thanked for their reviews of this paper. The lithological descriptions in this paper are based on the reports of numerous workers, including Sue Daly, Marc Davies, Alan Purvis, Liz Jagodzinski and many others.
References
Coppens KT 1997. A geochemical study of the Hopeful Basalt and Lake Harris Komatiite, the relationship between them and their economic potential for nickel and gold. BSc (Hons) thesis, University of Adelaide.
Ferris GM, Schwarz MP and Heithersay P 2002. The geological framework, distribution and controls of Fe-oxide and related alteration, and Cu-Au mineralisation in the Gawler Craton, South Australia. Part I – Geological and tectonic framework. In Porter TM ed, Hydrothermal iron oxide copper-gold and related deposits: a global perspective. PGC Publishing, Adelaide, pp 9–31.
Halpin JA and Reid AJ 2016. Earliest Paleoproterozoic high-grade metamorphism and orogenesis in the Gawler Craton, South Australia: the southern cousin in the Rae family? Precambrian Research 276:123–144. doi:10.1016/j.precamres.2016.02.001.
Hoatson DM, Sun SS, Duggan MB, Davies MB, Daly SJ and Purvis AC 2005. Late Archean Lake Harris komatiite, central Gawler Craton, South Australia: geologic setting and geochemistry. Economic Geology 100(2):349–374. doi:10.2113/gsecongeo.100.2.349
McFarlane CRM 2006. Palaeoproterozoic evolution of the Challenger Au deposit, South Australia, from monazite geochronology. Journal of Metamorphic Geology 24:75–87. doi:10.1111/j.1525-1314.2005.00622.x.
McFarlane CRM, Mavrogenes JA and Tomkins AG 2007. Recognizing hydrothermal alteration through a granulite facies metamorphic overprint at the Challenger Au deposit, South Australia. Chemical Geology 243:64–89. doi:10.1016/j.chemgeo.2007.05.010.
Reid AJ, Jagodzinski EA, Fraser GL and Pawley MJ 2014. SHRIMP U–Pb zircon age constraints on the tectonics of the Neoarchean to early Paleoproterozoic transition within the Mulgathing Complex, Gawler Craton, South Australia. Precambrian Research 250:27–49. doi:10.1016/j.precamres.2014.05.013.
Reid AJ, Jourdan F and Jagodzinski EA 2017. Mesoproterozoic fluid events affecting Archean crust in the northern Olympic Cu–Au Province, Gawler Craton: insights from 40Ar/39Ar thermochronology. Australian Journal of Earth Sciences 64:103–119. doi:10.1080/08120099.2017.1263806.
Stewart JR and Betts PG 2010. Implications for Proterozoic plate margin evolution from geophysical analysis and crustal-scale modeling within the western Gawler Craton, Australia. Tectonophysics 483:151–177. doi:10.1016/j.tecto.2009.11.016.
Swain G, Woodhouse A, Hand M, Barovich K, Schwarz M and Fanning CM 2005. Provenance and tectonic development of the late Archaean Gawler Craton, Australia; U-Pb zircon, geochemical and Sm-Nd isotopic implications. Precambrian Research 141:106–136. doi:10.1016/j.precamres.2005.08.004.
Tomkins AG and Mavrogenes JA 2001. Redistribution of gold within arsenopyrite and löllingite during pro-and retrograde metamorphism: application to timing of mineralization. Economic Geology 96:525–534. doi:10.2113/gsecongeo.96.3.525.
Tomkins AG and Mavrogenes JA 2002. Mobilization of gold as a polymetallic melt during pelite anatexis at the Challenger deposit, South Australia: a metamorphosed Archean gold deposit. Economic Geology 97:1249–1271. doi:10.2113/gsecongeo.97.6.1249.