Basin provenance and its control on mineralisation within the Early Devonian Cobar Basin, western Lachlan Orogen, eastern Australia

Abstract

The Cobar Basin, in central New South Wales, is an Early Devonian extensional basin that formed in the western Lachlan Orogen. The basin was filled with shallow- to deep-water sequences of the Cobar Supergroup and hosts small to large polymetallic deposits. Detrital zircon geochronology and whole-rock geochemical data collected from the Amphitheatre Group of the Cobar Supergroup provide constraints on the history of basin fill and illustrate the dynamic interplay between basin provenance and mineralisation, corresponding to the evolving tectonic regime. Data reveal provenance dissimilarity between the southern and northern parts of the basin. In the south, units of the Amphitheatre Group received abundant detritus from ca 430–410 Ma magmatic rocks situated to the southwest and southeast of the basin. By contrast, the northern successions were predominately sourced from recycled Ordovician basement found to the northwest, north, northeast of the basin, along with contributions associated with the Macquarie Arc. This spatial provenance variation, however, is less significant in the younger formations: the northern and southern sequences both exhibit an increase in older recycled detritus upwards with time. This reflects a progressive modification of basin paleogeography, during the transition from rift phase to sag phase. The rift-phase basin geography is characterised by fault-restricted deposition with predominant sediments derived from local proximal sources. The subsequent sag-phase subsidence exhibits a uniform depositional system with more homogenised basin input. This provenance variation is coeval with the stratigraphically controlled mineralisation features within the Amphitheatre Group successions, implying a provenance influence on mineralisation. Data suggest the different sediment source regions have produced distinct detrital mineral compositions between the major mineral-hosting and mineral-barren formations. The enrichment of some detrital minerals in the mineral-hosting units, such as feldspar, muscovite, Ti-minerals (and carbonate), is suggested to be an important factor for mineralisation in the basin.

KEY POINTS

1. Detrital zircon geochronology and whole-rock geochemistry illustrate the history of basin fill for the Cobar Basin.

2. Spatial and temporal variation of basin provenance reflects a modification of basin geography, corresponding to the evolving tectonic regimes.

3. The change in basin source regions is one of key controls on the mineralisation within the Cobar Basin.

Keywords:

• Cobar Basin
• detrital zircon geochronology
• whole-rock geochemistry
• provenance analysis
• tectonics
• mineralisation

Introduction

The Cobar Basin, in central New South Wales, eastern Australia, is one of a series of late Silurian to Early Devonian extensional basins that formed in the western Lachlan Orogen (Glen, 1987, 1994; MacRae, 1987). The Cobar Basin was filled with a set of shallow- to deep-water siliciclastic-dominated sequences with lesser volcanic material; together these form the Cobar Supergroup (Glen, 1987, 1994; MacRae, 1987). The deposition of Cobar Superbasin sequences has been divided into two phases: an early rift phase (ca 420 Ma to ca 415 Ma), marked by a period of rapid basin deepening and faulting accompanied by syn-rift and rift–sag transition deposition, and a late sag phase (ca 410 Ma to ca 400 Ma), a period of basin inversion coeval with the deposition of deep to shallow marine successions (Fitzherbert & Downes, 2021; Glen, 1994).

The Cobar Supergroup hosts small to large polymetallic (Cu–Au–Pb–Zn–Ag) deposits and is one of the major historical and active mining provinces in New South Wales, Australia (Downes et al., 2011; Fitzherbert et al., 2021; Fitzherbert & Downes, 2021; Stegman, 2001). Previous metallogenic studies in collaboration with basin architecture, metamorphic mapping and structural analysis illustrate a comprehensive mineralising system that is closely related to the evolution of the Cobar Basin (David, 2008, 2018; Downes et al., 2011, 2013; Fitzherbert & Downes, 2021; Fitzherbert et al., 2017, 2021; Glen et al., 1994; Stegman, 2001). Importantly, all the major deposits (e.g. Peak, Chesney, New Occidental, Great Cobar, New Cobar, CSA, Endeavor, Nymagee-Hera, Mallee Bull) within the Cobar Supergroup are hosted within the pre-sag-phase stratigraphy (rift phase and rift–sag transition phase deposition), whereas little mineralisation is hosted in the sag-phase successions (David, 2018; Fitzherbert et al., 2019, 2020, 2021; Fitzherbert & Downes, 2021 and references therein). Further, mineralised sections commonly parallel lithological boundaries, indicative of stratigraphic influences on mineralisation (David, 2018; Fitzherbert et al., 2019, 2020; Fitzherbert & Downes, 2021). The direct connections between stratigraphy and mineralisation suggest that the history of basin fill played an important role during mineralising processes. Glen et al. (2016) and Parrish et al. (2018) used detrital zircon U–Pb data to illustrate the possible sources for the northern part of the basin. However, to date, limited work has been conducted to constrain the evolution of basin provenance, and further, its influences on the Cobar mineralisation.

This study focuses on the Amphitheatre Group of the Cobar Supergroup and attempts to determine the relationship between mineralisation and provenance evolution. Here we present new detrital zircon U–Pb geochronology and whole-rock geochemical data collected from various stratigraphic units within the Amphitheatre Group. These data, compiled with previously published data collected from the Amphitheatre Group, are used to identify sedimentary sources for each sedimentary unit across various parts of the basin. The subsequent spatial and temporal provenance analysis recognises the dissimilarity in source rocks between the mineral-hosting formations (pre-sag-phase sequences) and the mineral-barren sequences (sag-phase deposition), which illustrates the provenance controls on the mineralising processes.

Geological background

Cobar Supergroup, Amphitheatre Group

The Early Devonian Cobar Basin is an extensional basin located in the central western Lachlan Orogen, New South Wales, Australia (Figure 1). The basin comprises shallow- to deep-water siliciclastic-dominated sequences developed over a basement of Silurian granite and foliated Lower–Middle Ordovician metasedimentary (i.e. the Girilambone and Adaminaby groups) and lesser metamorphosed mafic igneous rocks (Figure 2; Glen, 1987, 1994; MacRae, 1987). The basin has a complex fault-controlled geometry where the deep-water and shallow-water successions are abruptly juxtaposed (Figure 1).

Figure 1. Geological map showing major tectonostratigraphic units and sample locations from this study and previous studies. *Samples from Parrish et al. (2018); **samples from Glen et al. (2016); note that only the samples from the Amphitheatre Group are discussed. Extent of mapped rock units and fault locations are from the NSW Seamless Geology dataset (Colquhoun et al., 2021).

Figure 2. Stratigraphic diagram showing the framework of the Cobar Supergroup after Folkes et al. (2022) and references therein: (a) north Cobar Basin, and (b) south Cobar Basin. Unit abbreviations: Duci, Wilgaroon Granite; Suwe, Erimeran Granite; Suwl, Thule Granite; Duci, Wilgaroon Granite; Durn, Nyora Granite; Duwb, Boolahbone Granite; Skow, Weethalle Granite; Sui, unassigned Silurian intrusions; Suwe, Erimeran Granite; Suwk, Gilgunnia Granite; Suwl, Thule Granite; Suwp, Mount Allen Granite; Suwt, Tinderra Granite.

The deep-water successions of the Cobar Supergroup comprise fine- to medium-grained siliciclastic-dominated turbiditic sequences; these successions have been divided into two groups: the Nurri Group and the Amphitheatre Group (Figure 2; Glen, 1987, 1994; MacRae, 1987). The Nurri Group is a set of upward-fining sequences from basal coarse-grained clastic material to siltstone/mudstone material higher in the sequence. The Nurri Group is mainly deposited in the northern part of the Cobar Basin, whereas the coeval to overlying Amphitheatre Group is more extensively distributed over most of the basin (Figure 1). The Amphitheatre Group, the focus of this study, comprises three main subdivisions, including the (informally named) lower Amphitheatre Group, the Shume Formation and the correlative Biddabirra Formation, and the informally named upper Amphitheatre Group (Figure 2). The lower and upper Amphitheatre groups both contain thin-bedded siltstone–mudstone with minor sandstone; they are commonly separated by a set of basin-wide lithic quartz to quartz-rich sandstone horizons, namely the Shume Formation in the southern Cobar Basin and its equivalent Biddabirra Formation in the north (Figure 2; Glen, 1987, 1994; MacRae, 1987). The CSA Siltstone, mainly deposited in the northern Cobar Basin, is considered a part of the lower Amphitheatre Group (Figure 2; Glen, 1987, 1994). Interestingly, considering the long mining history of the Cobar Basin region and the large body of research focused on describing and characterising aspects of these various deep-water successions, little systematic basin-wide sandstone–siltstone petrography has been conducted on these rock units. Although beyond the scope of this research, petrographic data such as this would provide useful additional information regarding sedimentary provenance studies for rock units in this mineralised region.

The deep-water successions are flanked by shallow-water sequences to the east (the Kopyje and Mouramba groups) and the west (the Winduck Group; Figure 1). The Kopyje and Mouramba groups have been interpreted to represent the shallow-water shelf deposit equivalents of the deep-water Amphitheatre and Nurri groups (Glen, 1987, 1994; MacRae, 1987). Alternatively, these shallow-water successions have also been interpreted to be overlain by the deep-water Amphitheatre and Nurri groups; this model suggests that the shallow- and deep-water sequences were fault-bounded by east-dipping thrusts (i.e. the Great Chesney, Cobar, and Myrt faults) activated during the early Late Devonian (Burton, 2016, 2023). The shallow marine to fluviatile Yarra Yarra Creek Group is classified as part of the Kopyje Group, although a potential 8–10 Ma depositional hiatus has been recognised between the Yarra Yarra Creek Group and the underlying Kopyje Group (Fitzherbert & Downes, 2021). The Winduck Group is thought to overlie or interfinger with the upper Amphitheatre Group, likely representing the final phase of basin fill (Glen, 1987, 1994; MacRae, 1987; Sherwin, 2013). To the south and southwest of the basin, the Cobar Supergroup interfingers with (or is fault-bound against) two volcanogenic troughs—the Mount Hope and the Rast troughs (Figure 1). The Mount Hope Trough comprises the volcanic-dominated Mount Hope Group and the fine- to coarse-grained clastic-dominated Broken Range Group. The Rast Trough was filled with shallow- to deep-marine sediments of the Rast Group and the volcanic-dominated Ural Volcanics. Geochronological dating of the Mount Hope Group and Ural Volcanics yielded similar ages of ca 420–409 Ma, and geochemical analysis reveals bimodal characteristics, suggesting an intracontinental rift tectonic setting (Bull et al., 2008; Bull & Fitzherbert, 2022).

The depositional age of the Cobar Supergroup is constrained by fossil data. Corals and conodonts from the Elura Limestone Member of the Brookong Formation suggest a Lochkovian age for deposition of the Kopyje Group (Zhen & Fitzherbert, 2021). Consistently, conodonts recovered from the Booth Limestone in the Kopyje Group indicate an early Pragian age (MacRae, 1987). The conodont assemblage from the Amphitheatre Group suggests a Lochkovian and Pragian age (Zhen et al., 2023), which is consistent with the allochthonous Lerida Limestone Member of the Biddabirra Formation that hosts conodonts and trilobites of Lochkovian to Pragian age (Figure 2; Ebach, 2002; Mathieson et al., 2016). Moreover, the brachiopod species identified from the Shume Formation range in age from Wenlock to Lochkovian and possibly early Pragian (Zhen et al., 2023). Further, trilobite fragments discovered from the Winduck Group suggest an Early Devonian age (Sherwin & Meakin, 2010). These paleontological age constraints agree with the radiometric dating results (ca 421–409 Ma) of the syn-rift intrusions and volcanics within and adjacent to the Cobar Basin (Bull et al., 2008; Bull & Fitzherbert, 2022; Downes et al. 2016; Jones et al., 2020; Parrish et al., 2018), suggesting that the rifting tectonic regime and sediment deposition likely extended to the Pragian.

Tectonic background and mineralisation

The late Silurian to Devonian history of the eastern Lachlan Orogen is characterised by prolonged extension punctuated by discrete episodes of shortening (Collins, 2002; Collins et al., 2019; Collins & Richards, 2008; Fergusson, 2009, 2010). The Lower–Middle Ordovician metasedimentary basement to the Cobar Supergroup was strongly deformed during the Benambran Orogeny—a regional east–west shortening event in the latest Ordovician or early Silurian (Folkes & Stuart, 2020; Glen et al., 1994, 2007). The Benambran Orogeny was followed by extensive intrusion of granitic magma varying in age from ca 430 Ma to ca 410 Ma, that changed in composition from S-type to I-type, and to A-type with time (Bull et al., 2008; Collins et al., 2019; Downes et al., 2016; Glen et al., 2007; Jones et al., 2020). These protracted magmatic events that accompanied deposition of the Cobar Supergroup and growth faulting along the basin margins have been interpreted to reflect a prolonged period of backarc extension (Tabberabberan extension) associated with a retreating subduction zone, likely in response to ongoing subduction in the eastern Lachlan Orogen (Collins, 2002; Folkes & Stuart, 2020; Parrish et al., 2018; Rosenbaum, 2018).

The inversion of the Cobar Basin is thought to have been initiated at ca 400–395 Ma with regional foliation developing at ca 390 Ma, produced by the mid-Devonian Tabberabberan Orogeny (Downes et al., 2016; Fitzherbert et al., 2021; Glen et al., 1992, 1994). In the Cobar region, the Cobar Supergroup is unconformably (or locally conformably) overlain by a set of non-marine sandstone and conglomerate with minor mudstone and shale—the Mulga Downs Group (Figure 2; Glen et al., 1994; Trigg, 2018). The depositional age for the Mulga Downs Group remains unknown and has been considered variably from early Middle Devonian (Downes et al., 2016; Trigg, 2018) to Late Devonian (Asmussen et al., 2023; Barry, 2016).

The Cobar Basin hosts small to large polymetallic deposits related to different tectonostratigraphic units from the late Silurian to the Early Devonian (David, 2008, 2018; Downes et al., 2011; Fitzherbert et al., 2021; Fitzherbert & Downes, 2021 and references therein; Stegman, 2001). Fitzherbert and Downes (2021) link the Cobar mineral systems with the late Silurian to Devonian tectonic evolution and identify three stages for the mineralisation: (1) ca 420–415 Ma rift phase, including syn-rift magmatic-related epithermal, skarn and potentially volcanic-associated massive sulfide systems; (2) ca 413–400 Ma transitional rift to sag phase, comprising structurally focused distal intrusion-related Cu–Au and skarn systems; and (3) ca 390–380 Ma inversion phase, characterised by the sedimentary rock-hosted low temperature Pb–Zn–Ag mineralisation.

Methodology

Whole-rock geochemistry

A total of 18 samples of slate, meta-siltstone and meta-sandstone were collected for whole-rock geochemistry from the Amphitheatre Group, including three from the lower Amphitheatre Group, five from the CSA Siltstone, five from the Shume Formation and five from the upper Amphitheatre Group. Detailed sample information is provided in Supplemental data (Appendices 1 and 2). The samples were ground in a Tema mill with a chrome steel crusher at the University of Newcastle and sent to Intertek Genalysis Laboratory, Adelaide, for major and trace elements analysis. The geochemical data were obtained by inductively coupled plasma optical emission spectrometry (ICP-OES) and laser ablation-inductively coupled plasma mass spectrometry (LA-ICP-MS). Analytical precision, reproducibility, and detection limits are given in the Supplemental data (Appendix 2). Standards Oreas 100a, DC19005, GSP-2 and SY-4 were used during the analysis of the samples. The La/Th–Hf, Al–Ti–Hf and Th/Sc–Zr/Sc diagrams were plotted by Igpet. Principal-component Analysis (PCA) is used in the study to visualise chemical differences among individual samples. The PCA is statistical procedure that reduces dimensions of large datasets, allowing samples that have multiple variables (e.g. element contents) to be more easily compared and analysed. The PCA plot was conducted by using statistical analysis packages in R programming language.

Detrital zircon geochronology

A total of 17 sandstone/siltstone samples were collected for detrital zircon chronology analysis. Detailed sample information is provided in the Supplemental data (Appendices 1 and 2). Sample preparation was conducted at the University of Newcastle. Rock samples were crushed, milled and sieved to collect material <250 μm in size. Zircon grains were then separated using a Wilfley table, Frantz magnetic separator and heavy liquid separation. Individual zircons were then handpicked and mounted in epoxy resin. Fast laser ablation-inductively coupled plasma mass spectrometry (LA-ICP-MS) was performed at the GeoHistory Facility, John de Laeter Centre, Curtin University, Perth, Australia, following the method proposed by Clark et al. (2023). Ablations utilised an ASI RESOlution-SE 193 nm excimer laser controlled by GeoStar μGIS™ software. Laser fluence was calibrated above the sample cell using a hand-held energy meter, and subsequent analyses were performed in constant energy mode. The Laurin Technic S155 sample cell was flushed by ultrahigh-purity He (320 mL/min) and N₂ (1.2 mL/min), both of which were passed through inline gold sand Hg traps. High purity Ar was used as the ICP-MS carrier gas (flow rate ∼1 L/min). All measurements were performed using an Agilent 8900 QQQ quadrupole ICP-MS operated in single quad mode. Each analytical session consisted of initial gas flow and ICP-MS ion lens tuning for sensitivity with the signal smoothing device (‘squid’) connected. Additional tuning adjusted flow parameters for robust plasma conditions (²³⁸U/²³²Th ∼1; ²⁰⁶Pb/²³⁸U ∼0.2; and ²³⁸UO/²³⁸U <0.004). Finally, pulse-analogue (P/A) conversion factors were determined on the NIST 610 reference glass by varying laser spot sizes and/or laser repetition rate to yield 1–2 Mcps per element. After tuning, the signal smoothing device was removed, and the laser was connected to the mass spectrometer via a shorter (∼60 cm) length of Teflon tubing. Using this setup, signal wash-in and wash-out times were typically around 200 and 600 ms, respectively. Following Chew et al. (2019), laser parameters were chosen as 20 µm spot diameter, 50 Hz repetition rate, and on-sample laser energy of 2 Jcm⁻². The GeoStar software sequence for fast pulse analysis consisted of a 3 s delay after the TTL pulse for baseline acquisition, 4 s ablation and 7 s delay for the 8900 to complete data handling and return to a ‘wait for trigger’ state. ⁹¹Zr, ²⁰⁶Pb, ²⁰⁷Pb, ²⁰⁸Pb, ²³²Th and ²³⁸U were collected with dwell times of 3, 15, 20, 6, 6 and 9 ms, respectively, for a total acquisition time of 10 s, and a duty cycle (data acquisition time/total time) of about 66%. The time-resolved mass spectra were reduced using the ‘U–Pb Geochronology’ data-reduction scheme in Iolite 4.3.5.3 (Paton et al., 2011 and references therein) and presented in Supplemental data (Appendix 2).

Analyses of unknowns were bracketed (every ∼20 unknowns) with analyses of the primary and secondary zircon standards to monitor and correct for mass fractionation and instrumental drift. GEMOC GJ-1 (²⁰⁶Pb/²³⁸U age: 600.7 ± 1.1 Ma; Jackson et al., 2004) was used as the primary standard. The secondary standards are Plešovice (²⁰⁶Pb/²³⁸U 337.13 ± 0.37 Ma; Sláma et al., 2008), 91500 (²⁰⁷Pb/²⁰⁶Pb age: 1065.4 ± 0.3 Ma; ²⁰⁶Pb/²³⁸U age: 1062.4 ± 0.4 Ma; Wiedenbeck et al., 1995), Maniitsoq (²⁰⁷Pb/²⁰⁶Pb age: 3008.70 ± 0.72 Ma; Marsh et al., 2019) and OG1 (²⁰⁷Pb/²⁰⁶Pb age: 3465.4 ± 0.6 Ma; Bodorkos et al., 2009). Analytical results of standards are presented in the Supplemental data (Appendix 2). Individual age concordance is calculated by (²⁰⁶Pb/²³⁸U age/²⁰⁷Pb/²³⁵U age*100) for preferred ages younger than 1000 Ma and by (²⁰⁶Pb/²³⁸U age/²⁰⁷Pb/²⁰⁶Pb age*100) for preferred ages older than 1000 Ma. The IsotopeR program (Vermeesch, 2018) was used for visualisation of the zircon age data as kernel density estimates (KDE) and multi-dimensional scaling (MDS).

Results

Geochemistry

Full chemical analyses of the samples from the Amphitheatre Group are presented in the Supplemental data (Appendix 2). A variety of discrimination diagrams based on relatively immobile elements were plotted to give an indication of sedimentary provenance and tectonic settings. The La/Th–Hf plot (Figure 3a) shows that all lower Amphitheatre Group samples, and most of the Shume Formation and the CSA Siltstone samples plot in the felsic arc field. This contrasts with those of the upper Amphitheatre Group that exhibit a spread towards the passive margin field, indicative of higher contribution of upper continental crust (Floyd & Leveridge, 1987). In the ternary Al–Ti–Hf diagram (Figure 3b), most samples from the lower Amphitheatre Group, the Shume Formation and the CSA Siltstone form distinct clusters, whereas the upper Amphitheatre Group samples contain a spread of Zr values and are plotted along a recycling array (Garcia et al., 1991). The Th/Sc–Zr/Sc plot reveals the compositional variation among the different sedimentary units (McLennan et al., 1993). Note that several samples contain Sc values that are below the ICP-MS detection limit (10 ppm); these samples are plotted using a maximum Sc value of 10 ppm with arrows pointing the directions with lower Sc values (Figure 3c). In Figure 3c, most of the lower Amphitheatre Group, the Shume Formation and the CSA Siltstone samples plot together, clustering around a granodioritic composition. The upper Amphitheatre Group samples plot on the recycling array (Figure 3c), indicative of high compositions of recycled materials within these samples. Note that one CSA Siltstone sample (NP1047-1) shows strong enrichment of Zr and Hf that is not concordant with the other CSA Siltstone samples, probably owing to an elevated number of zircon grains.

Figure 3. Geochemical discrimination diagrams for stratigraphic units of the Amphitheatre Group: (a) La/Th vs Hf plot of Floyd and Leveridge (1987), (b) Zr–Al₂O₃–TiO₂ plot of Garcia et al. (1991), and (c) Th/Sc vs Zr/Sc plot of McLennan et al. (1993). Note that some samples contain Sc values lower than the detection limit (10 ppm); these samples are plotted using a maximum Sc value of 10 ppm with arrows pointing in the directions of lower values.

Geochronology

Detailed sample information and analysis results are given in the Supplemental data (Appendices 1 and 2). Note that several samples have zircon grains yielding ages much younger than the biostratigraphic age constraints (Mathieson et al. 2016; Zhen & Fitzherbert, 2021). Most of these younger ages cluster in age at ca 380 Ma and ca 310 Ma. They are interpreted to be influenced by post-deposition events; the ca 380 Ma ages overlap with the ca 390–380 Ma Tabberabberan Orogeny (Downes et al., 2016), whereas the ca 310 Ma ages might represent a younger event that affected the region. These younger ages (<ca 390 Ma) are not discussed in this study.

Lower Amphitheatre Group

The three samples of the lower Amphitheatre Group were collected from the southern part of the Cobar Basin (Figure 1), including two field samples (NSWSJAF03O and MXSCSTJ0311) and one drill core sample (LN-09). Detrital zircon grains from the lower Amphitheatre Group commonly exhibit sector and oscillatory zonation. Most of these grains are euhedral to subhedral with minor rounded to sub-rounded grains (Supplemental data, Appendix 1, Figure A1). Some grains are broken fragments of large zircon crystals; they commonly exhibit sharp broken edges (Supplemental data, Appendix 1, Figure A1). In sample LN-09, 225 near-concordant (<10% discordance) analyses yielded ages ranging from 3476 Ma to 400 Ma; these analyses form a dominant peak at ca 410 Ma and four subsidiary maxima at ca 3000 Ma, ca 1080 Ma, ca 570 Ma and ca 490 Ma (Figure 4). Seventy-one near-concordant analyses were obtained from sample MXSCSTJ0311. These analyses yielded ages ranging from 1203 Ma to 403 Ma with a major cluster at ca 415 Ma, two minor peaks at ca 560 Ma and ca 490 Ma, and a wide age spread between ca 1200 Ma and ca 830 Ma (Figure 4). Sample NSWSJAF03O yielded ages (from <10% discordance analysis) ranging from 2336 to 404 Ma (n = 110) with one major peak at ca 420 Ma and a secondary peak at ca 490 Ma (Figure 4). Of the three lower Amphitheatre Group samples, the youngest single analysis, obtained from sample LN-09, yielded a ²⁰⁶Pb/²³⁸U date of 400 ± 25 Ma (2σ).

Figure 4. Kernel distribution estimates of detrital zircon age spectra of the lower Amphitheatre Group, the CSA Siltstone and the Biddabirra Formation.

CSA Siltstone

The two samples of the CSA Siltstone (ENRC-1 and NP1235-1) were obtained on drill core from drillholes located in the northern part of the Cobar Basin. The majority of zircon grains from this formation display a stubby, slightly elongate to elongate, rounded to euhedral crystal form, with sector, oscillatory, or cloudy zoning patterns (Supplemental data, Appendix 1, Figure A1). Minor grains are fragments from broken zircon grains (Supplemental data, Appendix 1, Figure A1). The zircons from sample ENRC yielded ages in the range 3330–396 Ma (n = 258), exhibiting a prominent age peak at ca 500 Ma, a minor shoulder at ca 480 Ma, a wide age group ranging from ca 1200 Ma to 800 Ma, and scattered individual analyses between ca 3300 Ma and ca 1200 Ma (Figure 4). Analyses from sample NP1235 yielded ages ranging from 3215 Ma to 412 Ma. The age spectrum of this sample exhibits two major maxima at ca 520 Ma and ca 470 Ma, with a spread of ages ranging from ca 1200 Ma to 600 Ma (Figure 4). The youngest single analysis from the CSA Siltstone, obtained from sample ENRC-1, has an ²⁰⁶Pb/²³⁸U age of 396 ± 29 Ma (2σ).

Biddabirra Formation

The two samples are both field samples collected from similar locations. Sample MXNCLMC0024.01A was from the Biddabirra Formation, and sample MXNCLMC0014.01C was collected from the Alley Sandstone Member located at the top of the Biddabirra Formation (Figure 2). Zircon grains from the Biddabirra Formation/Alley Sandstone Member are commonly rounded or sub-rounded with elongate grain shapes and exhibit oscillatory, broad banding, or sector zoning patterns (Supplemental data, Appendix 1, Figure A1). Zircon age spectra of these two samples are consistent, with dominant age peaks at ca 500 Ma, minor peaks at ca 560 Ma, and wide age groups ranging from ca 1200 Ma to 600 Ma (Figure 4). The youngest analysed single grain is from sample MXNCLMC0024.01A with a ²⁰⁶Pb/²³⁸U age of 406 ± 23 Ma (2σ).

Shume Formation

The seven samples from the Shume Formation are all drillcore samples collected from drillhole LN001 (samples LN-06, LN-02 and LN-01), drillhole 37S-2S (samples 37S-2D-05 and 37S-2D-02) and drillhole MBDD007 (samples MBDD007-9 and MBDD007-8). Zircon grains from the Shume Formation are commonly euhedral to subhedral and range from equant to more elongate crystals, with fewer that are more rounded to sub-rounded (Supplemental data, Appendix 1, Figure A1). Some grains are fragments and have sharp broken edges. Most zircon grains have sector-zoned or broad banding patterns. In sample LN-06, 246 near-concordant analyses yielded ages that range from 3471–398 Ma, with a major peak at ca 490 Ma and minor peaks at ca 1050 Ma, ca 570 Ma and 410 Ma (Figure 5). One-hundred and seventy-eight near-concordant analyses from sample LN-02 yielded ages in the range 3317–400 Ma, with a dominant age peak at ca 490 Ma, two minor peaks at ca 580 Ma and ca 415 Ma, and a spread of ages from ca 1200 Ma to 800 Ma (Figure 5). Zircon grains from sample LN-01 yielded near-concordant ages mainly in the range 2731–399 (n = 184). The age spectrum of this sample is characterised by three major peaks at ca 580 Ma, ca 490 Ma and ca 410 Ma (Figure 5). The zircon grains from sample 37S-2D-05 yielded 257 near-concordant analyses. These analyses range from 3369 to 397 Ma and form two prominent age peaks at ca 490 Ma and ca 415 Ma with a minor peak at ca 560 Ma, and a broad age spread ranging from ca 1200 Ma to 800 Ma (Figure 5). Sample 37S-2D-02 yielded ages ranging from 3607 to 397 Ma (n = 245). They form a major peak at ca 410 Ma, two secondary peaks at ca 570 Ma and ca 490 Ma, and a wide age spread from ca 1200 Ma to 800 Ma (Figure 5). In sample MBDD007-9, 293 near-concordant analyses form a broad range from 2818 Ma to 400 Ma with two major peaks at ca 520 Ma and ca 420 Ma and a wide age group from ca 1200 Ma to 800 Ma (Figure 5). The 254 near-concordant analyses from sample MBDD007-8 yielded ages in range of 3199–402 Ma, with two major peaks at ca 560 Ma and ca 510 Ma and minor peaks at ca 1180 Ma, ca 920 Ma and ca 420 Ma (Figure 5). The youngest single zircon analysis from the Shume Formation yielded a ²⁰⁶Pb/²³⁸U age of 397 ± 19 Ma (2σ) from sample 37S-2D-02.

Figure 5. Kernel distribution estimates of detrital zircon age spectra of the Shume Formation and the upper Amphitheatre Group.

Upper Amphitheatre Group

The three upper Amphitheatre Group samples includes one from the south Cobar Basin (MBDD007-7) and two from the north (DD87GA1-1 and MXNCLMC0012.02A). Zircon grains from this unit are predominantly rounded to sub-rounded with minor euhedral to subhedral grains. Most grains are moderate luminescent and sector/patchy-zoned (Supplemental data, Appendix 1, Figure A1). In sample MBDD007-7, the 294 near-concordant analyses, ranging from 2806 to 403 Ma, form a dominant age peak at ca 510 Ma and a minor peak at ca 430 Ma (Figure 5). The zircon grains from sample DD87GA1-1 yielded ages in the range 2942–401 Ma (n = 91), with a major peak at ca 560 Ma and minor peaks at ca 1050 Ma and ca 430 Ma. In samples MXNCLMC0012.02A, 99 near-concordant analyses yielded a spread of ages ranging from ca 3226 Ma to 407 Ma, with two major maxima at ca 500 Ma and ca 560 Ma and a board age spread from ca 1200 Ma to ca 700 Ma (Figure 5). Of the four upper Amphitheatre Group samples, the youngest single grain age was obtained from sample DD87GA1 with a ²⁰⁶Pb/²³⁸U age of 401 ± 23 Ma (2σ).

Discussion

Basin provenance

Detrital zircon U–Pb age data from this research are compiled with the published Amphitheatre Group data from previous studies conducted by Glen et al. (2016) and Parrish et al. (2018) that include two samples from the Alley Sandstone Member of the Biddabirra Formation (Alley 1 and MP015) and one sample from the lower Amphitheatre Group (MP007). These data, together, are used to identify the source regions for each sedimentary unit examined in this study. The Amphitheatre Group sequences from the south and the north Cobar Basin exhibit different zircon age spectra and are discussed separately below.

South Cobar Basin

A total of 12 samples from the Amphitheatre Group in the south Cobar Basin are discussed, including three from the lower Amphitheatre Group, seven from the Shume Formation and two from the upper Amphitheatre Group. The lower Amphitheatre Group samples have similar zircon age spectra, and their combined age spectrum exhibits a prominent age peak at ca 415 Ma, two minor clusters at ca 560 and ca 490 Ma, and a wide spread of ages ranging from ca 1200 Ma to ca 800 Ma (Figure 6). The dominant ca 415 Ma ages match well with magmatic rocks that are underlying, interfingering with or adjacent to Cobar Supergroup successions, implying a proximal origin. This interpretation is consistent with the commonly euhedral or sub-euhedral morphology of these zircon grains (Supplemental data, Appendix 1, Figure A1), indicative of short-distance transportation. We suggest possible sources for these zircon grains might be the magmatic rocks that are extensively exposed to the southwest and southeast of the basin (Figure 1), including the ca 434–424 Ma Erimeran Granite (Downes et al., 2016), the ca 427–424 Ma Thule Granite (Chisholm et al., 2014; Downes et al., 2016), the ca 428–417 Ma Mineral Hill Volcanics (Downes et al., 2016), the ca 415–409 Ma Mount Hope Volcanics and the ca 412 Ma Ural Volcanics (Bull et al., 2008). This agrees with previous petrographic studies, suggesting that subangular plutonic quartz crystals are abundant in the lower Amphitheatre Group (Glen, 1987; MacRae, 1987). Moreover, volcanic detritus enriched layers have been identified in outcrop (Glen, 1987; MacRae, 1987). The minor age populations, represented by the ca 1200–800 Ma and ca 560 Ma zircon age groups, overlap with the late Mesoproterozoic to Tonian Musgrave Orogeny and the Ediacaran Petermann Orogeny in central Australia (Maboko et al., 1992; Smithies et al., 2011), and the ca 490 Ma detrital zircon ages could be correlated with the ca 520–490 Ma Delamerian-Ross Orogeny of the eastern margin of Gondwana (Foden et al., 2006; Greenfield et al., 2011). Despite the match with these orogenic events, a preferred interpretation in this research is that these ca 1200–800 Ma, ca 560 Ma and ca 490 Ma zircon grains were all derived through erosion and recycling of local Lower–Middle Ordovician sedimentary rocks. Exposed basement sequences of Ordovician age, such as the Girilambone Group and the Adaminaby Group, are commonly located to the north, east and southwest of the basin, and have likely been covered by younger successions to the west (Figure 1). Glen et al. (2011, 2017) presented detrital zircon U–Pb age data from the Girilambone and Adaminaby groups with the main zircon age populations at ca 1200–800 Ma, ca 560 Ma and ca 490 Ma, consistent with those observed in the lower Amphitheatre Group (Figure 6). We suggest the lower Amphitheatre Group in the south Cobar Basin was predominantly sourced from ca 430–410 Ma magmatic rocks to the south with minor contributions that are likely to be derived from the Ordovician basement sequences.

Figure 6. Kernel density estimation plots and pie charts of combined detrital zircon ages from different stratigraphic units of the Amphitheatre Group in different parts of the Cobar Basin (top three panels). Coloured columns are bands of equivalent age ranges with major groups highlighted. Data from the Lower–Middle Ordovician turbiditic successions (lower panel) are from Glen et al. (2011, 2016, 2017). *Compiled detrital zircon data from this research and from previous studies (Glen et al., 2016; Parrish et al., 2018); **published data from Parrish et al. (2018).

Age spectra of the combined samples from the Shume Formation and the upper Amphitheatre Group show similarity to that of the lower Amphitheatre Group, with the main age groups coherently identified at ca 1200–800 Ma, ca 560 Ma, ca 490 Ma and ca 415 Ma (Figure 6). The consistent age spectra indicate similar source regions. However, compared with the lower Amphitheatre Group, the Shume Formation, and the upper Amphitheatre Group both contain significantly higher proportions of ‘older’ grains. Specifically, the percentages of the ca 560 Ma and ca 490 Ma zircon populations are 8.9% and 13.8% in the lower Amphitheatre Group, whereas these numbers rise to 12.5% and 21.9% for the Shume Formation, and 13.5% and 18.1% for the upper Amphitheatre Group, respectively (Figure 6). Similarly, the percentage of the ca 1200–800 Ma age group is 15.3% in the lower Amphitheatre Group, rising to 20.7% and 28.7% in the Shume Formation and the upper Amphitheatre Group, respectively (Figure 6). These older grains are interpreted to be derived from the Lower–Middle Ordovician turbiditic successions, consistent with those seen in the lower Amphitheatre Group. The increased numbers of ‘older’ grains reflect a progressive increase in sediment contributions from the Lower–Middle Ordovician turbiditic basement over time. This interpretation is also supported by petrographic observations, showing that the Shume Formation and the upper Amphitheatre Group samples contain more lithic components derived from (meta)sedimentary rocks, such as quartzite, chert and slate detritus compared with the lower Amphitheatre Group (Supplemental data, Appendix 1, Figure A2a–d).

Although limited paleocurrent data are available from the lower Amphitheatre Group, measurements from the Shume Formation and the upper Amphitheatre Group reveal a northwest to eastern flow direction (Figure 7; Glen, 1987, 1994; MacRae, 1987). This reflects a complex drainage system for the south Cobar Basin, with catchments located to the west and south-southeast (Glen, 1987, 1994; MacRae, 1987). The paleocurrent data are consistent with the detrital zircon interpretation that suggests the Shume Formation and the upper Amphitheatre Group from the south Cobar Basin both received mixed detritus from the ca 430–410 magmatic rocks and the Ordovician basement sequences. The ca 430–410 magmatic rocks are predominately located to the southeast and southwest of the basin. The metasedimentary basement to the west likely contributed material to the basin through east-directed flows; these basement rocks were then covered by younger sequences, for example, the Winduck Group (Figure 1).

Figure 7. Summary of published paleocurrent measurements from the Amphitheatre Group of the Cobar Basin. Data were compiled from MacRae (1987) and Glen (1987, 1994). Note that only measurements that contain more than 10 readings are plotted and discussed.

North Cobar Basin

For the north Cobar Basin successions, detrital zircon age data were collected from the CSA Siltstone (lower part of the Cobar Supergroup), the Biddabirra/Alley formations and the upper Amphitheatre Group.

The combined age spectrum of CSA Siltstone samples exhibits two major peaks at ca 510 Ma and ca 470 Ma, with a broad spread of ages ranging from ca 1200 Ma to ca 800 Ma (Figure 6). The ca 1200 Ma to ca 800 Ma and ca 520 Ma detrital zircon groups are comparable with those seen in the south Cobar Basin successions and are interpreted to be derived from similar source regions—the Lower–Middle Ordovician turbiditic basement sequences. The ca 470 Ma group in the CSA Siltstone is similar in age to zircon data reported in the Kopyje Group in the north Cobar Basin and the lower Amphitheatre Group in the west of the basin; these zircons are interpreted to be derived from the ca 480–460 Ma Macquarie Arc (Phase 2) in the east Lachlan Orogen (Crawford et al., 2007; Parrish et al., 2018). We suggest that the CSA Siltstone received mixed detritus from both the Macquarie Arc and the Lower–Middle Ordovician turbiditic basement. Consistently, detrital zircons from the CSA Siltstone exhibit various grain shapes, from euhedral/sub-euhedral to sub-rounded/rounded (Supplemental data, Appendix 1, Figure A1); the euhedral/sub-euhedral grains with ca 470 Ma ages are likely to be derived from the Macquarie Arc.

The combined age spectrum of the Biddabirra Formation and Alley Sandstone Member comprises four samples, including two samples from this study, one sample from Parrish et al. (2018) and one sample from Glen et al. (2016). The age spectrum is characterised by a dominant age peak at ca 510 Ma with a minor shoulder at ca 560 Ma and a wide age group of ca 1200–800 Ma (Figure 6). These ages are consistent with those seen in the underlying CSA Siltstone as well as the turbiditic basement rocks (e.g. Girilambone and Adaminaby groups; Figure 6). Further, the paleocurrent data demonstrate dominant south to southeast flow directions for the Biddabirra/Alley formations, except one measurement that indicates a west to southwest direction of flow (Figure 7). We suggest that the Biddabirra Formation, including the Alley Sandstone Member, was predominately sourced from the redeposited sediments from the Lower–Middle Ordovician turbiditic basement units that were located to the north, northwest and northeast of the basin. Moreover, in contrast to the CSA Siltstone, the ca 470 Ma ages are less common in the Biddabirra Formation and Alley Sandstone Member (Figure 6), indicative of limited contribution from the Macquarie Arc.

Most of detrital zircons from the upper Amphitheatre Group yielded ages in the ranges of ca 1200–830 Ma and ca 610–400 Ma (Figure 6). The dominant ca 1200–800 Ma, ca 560 Ma and ca 510 age populations match well with those seen in the Ordovician basement sequences, implying that this formation was predominately sourced from the Lower–Middle Ordovician turbiditic successions. The two paleocurrent measurements reveal southeast flow directions, suggesting the sediments of the upper Amphitheatre Group were likely to be derived from sources that were located to the northwest of the basin. Note that the CSA Siltstone, Biddabirra Formation/Alley Sandstone Member, and the upper Amphitheatre Group from the north Cobar Basin all preserve minor ca 430–400 Ma aged zircons (∼2–7%). These zircon grains could have been sourced from magmatic rocks to the southwest, south and southeast, similar to the ca 430–400 Ma zircons seen in the southern Cobar Basin, or alternatively, from more proximal magmatic rocks that are sporadically exposed to the north and northwest of the basin, such as the ca 410 Ma Wilgaroon Granite (Deyssing & Carlton, 2020; Fraser et al., 2014; Pogson & Hilyard, 1981). No matter where these source rocks were located, they provided limited contributions to the north Cobar Basin.

Cobar Basin tectonic geography

The spatial and temporal provenance analysis presented above helps to reconstruct the paleogeography of the Cobar Basin and to understand the tectonic history of the region.

In the south Cobar region, detrital zircon data demonstrate a stratigraphic shift of basin provenance within the Amphitheatre Group. This is characterised by an up-section decrease in the ca 415 Ma zircons: the population of the ca 415 Ma zircons makes up ∼34% of the lower Amphitheatre Group but decreases to ∼15% and ∼12% in the Shume Formation and the upper Amphitheatre Group, respectively (Figure 6). By contrast, the populations of older zircons, such as the ca 1200–800 Ma, ca 610–550 Ma and ca 530–470 Ma ages, grow rapidly with younger stratigraphic units (Figure 6). We suggest that this stratigraphic trend implies a continued increase in detritus from the Lower–Middle Ordovician turbiditic basement units. As a result, influx of older material diluted the proportion of younger (ca 415 Ma) zircon grains that were derived from magmatic rocks located to the southeast–southwest. This stratigraphic provenance shift is well reflected in the MDS plot (Figure 8). In Figure 8, the three lower Amphitheatre Group samples plot close to the ca 415 Ma age peak (a synthetic age peak representing sources of the collective ca 430–400 Ma magmatic rocks), whereas the Shume Formation and the upper Amphitheatre Group samples plot closer to the Ordovician basement units (e.g. the Girilambone and Adaminaby groups). These samples, together, exhibit a stratigraphic curve showing younger units that plot closer to the Lower–Middle Ordovician basement units. Consistently, the geochemical plots reveal that the upper Amphitheatre Group shows a chemical affinity with the upper continental crust and exhibits strong recycling signatures, differing from underlying formations in the Cobar Supergroup (Figure 3).

Figure 8. Non-metric multi-dimensional scaling plot of sedimentary samples and potential source areas (black dots). The synthetic peak (black star) represents sources of the collective ca 430–400 Ma magmatic rocks.

However, this stratigraphic provenance variation is not well recorded by the detrital zircon data from the north Cobar samples. In this region, detrital zircon data suggest that the Amphitheatre Group successions (at all stratigraphic levels) from the north Cobar Basin, including the CSA Siltstone, the Biddabirra Formation/Alley Sandstone Member and the upper Amphitheatre Group, were all dominated by detritus derived from the Lower–Middle Ordovician sedimentary rocks, although the CSA Siltstone may have also received detritus from the Macquarie Arc. Moreover, on the MDS plot, the north Cobar samples all plot close to the Girilambone and Adaminaby groups, and exhibit little stratigraphic differences (Figure 8). Although the zircon data from the north Cobar successions did not record the provenance shift observed in the south, the whole-rock geochemical data show a similar pattern. The geochemical data show that the upper Amphitheatre Group samples from the north Cobar Basin exhibit strong recycling signatures, differing from the underlying CSA Siltstone samples (Figure 3). This stratigraphic trend, with younger units that exhibit stronger recycling signatures, is similar to that seen in the south Cobar Basin and is interpreted to reflect a progressive increase in recycled detritus from the Lower–Middle Ordovician turbidites.

This stratigraphic shift of basin provenance within the Amphitheatre Group coincides with the change in basin phases, from rift phase to sag phase (Fitzherbert & Downes, 2021). It is likely that the changed basin paleogeography during the switch of these basin phases modified basin drainage patterns, affecting sediments built up in the Cobar Basin. The lower Amphitheatre Group samples exhibit significant spatial difference in sediment sources—the samples collected from the southern part of the basin exhibit distinctive detrital zircon age populations compared with those from the northern and the southern parts of the basin (Figure 6). This is consistent with the rift basin geography where deposition would have been restricted by a series of grabens or half-grabens during extension. The syn-rift deposition also received significant sediment contributions from proximal volcanic sources or magmatic basement; this is particularly obvious in the south Cobar Basin that exhibited frequent magmatic activity during this rifting event. The spatial provenance dissimilarity, however, was not observed within the later sag-phase sequences; the upper Amphitheatre Group samples from the south and north Cobar Basin are all dominated by detritus that was derived from the Lower–Middle Ordovician turbiditic basements to the west and northwest of the basin. We suggest that the sag-phase basin geography is likely to be a wider uniform depositional system that has a more consistent sediment dispersal pattern with main catchments located to the northwest and the west.

The transition of basin provenance overlaps with the Bindian Event that is suggested to be ca 415–410 Ma (Fergusson, 2010). The Bindian Event was associated with the south-southeast tectonic transport with accompanying contractional deformation and shortening in the eastern part of the Lachlan Orogen (Fergusson, 2010, 2017; Folkes & Stuart, 2020). In the western Lachlan Orogen, contraction during the Bindian Event produced a series of strike-slip and/or dip-slip reverse faults in the Cobar region, Kopyje Shelf and in the southern part of the Wagga-Omeo Zone (Figure 8; Fitzherbert & Downes, 2021; Folkes & Stuart, 2020). The kinematic analysis of these faulting activities suggests the main direction of contraction was broadly northeast–southwest (Folkes & Stuart, 2020). In the Cobar Basin, Fitzherbert and Downes (2021) suggested that the formation of the Shume–Biddabirra sandstone package, a set of basin-wide, coarse-grained sandstones that separate the fine-grained, lower and upper Amphitheatre groups, might represent a localised uplift, in response to fault activities during the Bindian Event. Although direct correlations between the Bindian Event and the provenance shift within the Cobar Basin remain unclear, it is possible that Bindian-related tectonism modified the basin paleogeography, resulting in Lower–Middle Ordovician basement units becoming new topographic highs. Subsequently, the enhanced denudation of these basement rocks provided more detritus to the basin (Figure 9).

Figure 9. Tectonic, fault activity and deposition history of the Amphitheatre Group during (a) earlier rift-phase and (b) later sag-phase (modified after Fitzherbert & Downes, 2021, and Folkes & Stuart, 2020). Fault locations and their event kinematics from the NSW Seamless Geology dataset (Colquhoun et al., 2021). Filled arrows represent major tectonic stress directions; thin arrows show the directions of sediment fill.

Provenance controls on mineralisation

Previous research noted that major ore deposits in the Cobar Supergroup are preserved within rift- and rift–sag transition phases, whereas fewer ore deposits are hosted in the younger sag-phase successions (David, 2018; Fitzherbert & Downes, 2021; Fitzherbert et al., 2019, 2020, 2021 and references therein). For the Amphitheatre Group, most mineral deposits are hosted at the level of the Shume Formation, the CSA Siltstone and the lower Amphitheatre Group, whereas few are hosted in the later sag-phase deposition—the upper Amphitheatre Group (Fitzherbert & Downes, 2021; Fitzherbert et al., 2021 and references therein). The abrupt change in mineralising style is synchronous with the transition of basin provenance, from proximal magmatic sources to recycled metasedimentary basement (the Lower–Middle Ordovician turbiditic successions). The coevally changing basin fill and mineralising behaviour indicate a possible correlation between basin provenance and mineralisation. Here, we use sedimentary whole-rock geochemical data to provide complementary constraints on basin provenance. The provenance difference between the mineral-hosting units (the rift and rift–sag transition phases successions, including the Shume Formation, the lower Amphitheatre Group and the CSA Siltstone) and the mineral-barren formation (the later sag-phase stratigraphy—the upper Amphitheatre Group) reveals genetic relationships between basin provenance and mineralisation. PCA is used to show the chemical difference within the Cobar sedimentary rocks (Figure 10). Note that only elements that are less easily affected during post-deposition alteration (e.g. metasomatism) are plotted and discussed, as they better reflect the nature of source regions.

Figure 10. Principal-component analysis showing chemical variations for selected elements between the mineral-hosting formations (bold and italic text) and the mineral-barren formations. Ce* and Eu* denote positive cerium and europium anomalies, respectively.

In Figure 10, most samples from the upper Amphitheatre Group contain relatively high SiO₂, Zr and Hf contents, which is distinctive from the underlying Cobar Supergroup successions. The higher silica contents likely reflect a more quartz-dominated lithology, which is consistent with previous microscopic descriptions (Glen, 1987, 1994; MacRae, 1987). The higher Zr and Hf contents are also observed on the geochemical plots, implying recycling processes becoming more dominant (Figure 3). Moreover, the samples from the upper Amphitheatre Group show positive cerium anomalies (Figure 10). This is interpreted to have resulted from Ce⁴⁺–Zr⁴⁺ substitution in the zircon lattice (Thomas et al., 2003); the higher zirconium contents thus led to relatively higher Ce contents, resulting in the positive cerium anomalies.

In contrast, the mineral-hosting units, exhibit different chemical characteristics. Most samples from the Shume Formation, the lower Amphitheatre Group and the CSA Siltstone closely overlap with each other on the PCA plot (Figure 10). However, sample NP1047-1 from the CSA Siltstone is discordant with the other CSA Siltstone samples. This might be because of (a) possible chemical alteration after deposition or (b) the CSA Siltstone being locally chemically heterogeneous. Another sample from the Shume Formation (sample MBDD003-7) exhibits significant chemical inconsistencies with all the other Cobar Basin samples. It plots off the shown extent of Figure 10 and is interpreted to have been affected by post-depositional alteration. The rest of the samples from the mineral-hosting formations are consistent, exhibiting higher Al₂O₃, TiO₂, Ta and Nb contents than the mineral-barren formations, and possess negative Eu anomalies (Figure 10). The higher Al₂O₃ contents seen in the mineral-hosting formations are interpreted to result from increased weathering of aluminium enriched minerals in source rocks, such as feldspar and (or) muscovite. This is supported by hyperspectral data for these rock units suggesting that most mineral-hosting units include a significant component of detrital feldspar and white mica (Fitzherbert et al., 2020). Also, these minerals have been frequently observed microscopically, and detrital feldspar-enriched intervals have been noted in outcrop (Supplemental data, Appendix 1, Figure A2e–h; Glen, 1994). The higher feldspathic components might have led to the negative europium anomaly seen in these units, as Eu²⁺ is preferentially incorporated into feldspar minerals that have higher calcium contents, for example plagioclase (Weill & Drake, 1973). The higher TiO₂ contents are consistent with the abundant Ti-minerals in these formations (e.g. detrital titanite and rutile; Fitzherbert & Downes, 2021; Fitzherbert et al., 2021). Moreover, previous studies have shown Ti enrichment within ore zones in the Cobar region, which has been linked to the breakdown of detrital Ti-phase minerals (Fitzherbert & Downes, 2021; Fitzherbert et al., 2021; McKinnon & Fitzherbert, 2017). These detrital Ti-phase minerals are replaced by hydrothermal titanite near orebodies (Fitzherbert & Downes, 2021; Fitzherbert et al., 2021). LA-ICP-MS dating of the hydrothermal titanites yielded ²⁰⁶Pb/²³⁸U dates of ca 403 Ma and ca 384 Ma, respectively, constraining the ages of mineralisation (Fitzherbert et al., 2021). Further, these higher Ti contents might have resulted in the enrichment of Ta and Nb, owing to the substitution of Ti by Ta and Nb in Ti in the structure of Ti-bearing minerals. Previous metallogenic studies in the Cobar region suggest that metals were transported in reduced (generally pyrrhotite stable), moderately acidic pH (4.6–7.5), and low- to moderate-salinity (<10 wt% NaCl, range 2–15 wt% NaCl) fluids (Fitzherbert & Downes, 2021; Fitzherbert et al. 2021 and references therein). The reaction between the acidic reduced metal-bearing fluids and detrital minerals (e.g. feldspar, mica and/or carbonates) would alter the chemical conditions (e.g. pH and/or sulfidation), which is one of the key triggers for some Au–Cu and Pb–Zn–Ag deposits in the Cobar region (Fitzherbert & Downes, 2021; Fitzherbert et al., 2021). The detrital zircon and whole-rock geochemical data presented in this study demonstrate the different source regions for the mineral-hosting and mineral-barren formations in the Cobar Basin. The mineral-hosting formations are all rich in detrital feldspar, white mica and Ti-minerals, whereas these are minor components of the quartz dominant mineral-barren unit (the upper Amphitheatre Group). We suggest that the lower Amphitheatre Group and the Shume Formation received more detritus from the ca 430–410 Ma magmatic rocks in the Cobar region. These magmatic rocks or magmatic associated successions likely provided the detrital feldspar, muscovite and Ti-phase minerals to the basin. The CSA Siltstone preserves few ca 430–410 Ma aged zircon grains and is also feldspar enriched; these feldspars are likely to be derived from the Macquarie Arc associated sources. In contrast, sediments that formed the upper Amphitheatre Group were predominately sourced from the Ordovician turbiditic basement successions (e.g. the Adaminaby Group) that are quartz enriched (Offler & Fergusson, 2016). In the Cobar Basin, the major mineral deposits are located on the eastern basin margins and spatially overlap with major fault traces or intense shear zones (David, 2008, 2018; Fitzherbert et al., 2021; Fitzherbert & Downes, 2021). It has been suggested that faulting movements, in response to various tectonic events (e.g. Tabberabberan extension, the Bindian Event and the Tabberabberan Orogeny), provided pathways that facilitated the transport of metal-rich fluids (David, 2008, 2018; Fitzherbert & Downes, 2021; Glen et al., 1994). In the transitional rift to sag and the basin inversion phases, syn-rift volcanism and volcanic-related mineralisation waned, and the basin was characterised by distal intrusion-related Cu–Au and skarn systems and low-temperature sedimentary hosted mineralisation. Acidic metal-bearing fluids travelling along these faults would stall and concentrate where abrupt changes of lithology, rheology, resistibility or permeability occurred, for example the boundaries between the Shume Formation and the upper Amphitheatre Group (Fitzherbert & Downes, 2021). The subsequent reactions between the metal-bearing fluids and some detrital minerals, such as feldspar, muscovite, Ti-phase minerals (and carbonate), would result in mineral precipitation at specific stratigraphic intervals. Importantly, a general reduction in feldspar abundance in many mineralised intervals has been observed (Fitzherbert et al., 2020). In contrast, quartz-dominated lithologies would result in limited fluid–mineral reaction and could help explain the paucity of mineralisation within the upper Amphitheatre Group.

The Cobar mineral systems host episodic stages for the mineralisation that are closely related to the late Silurian to Devonian tectonic evolution of the basin (Fitzherbert & Downes, 2021). This changed tectonic setting activated the migration of various metal-bearing fluids that controlled the timing and style of the mineralisation. The reaction between these fluids and the detrital minerals resulted in mineralisation being hosted at specific stratigraphic levels (e.g. pre-sag-phase formations) for some mineral deposits.

Conclusions

Detrital zircon geochronology and whole-rock geochemical data presented in this study provide new constraints on sediment provenance and tectonic geography of the Cobar Basin. We also demonstrate a possible genetic relationship between basin provenance and mineralisation within the extensive Amphitheatre Group of the Cobar Basin.

Detrital zircon data reveal significant provenance dissimilarity between the northern and southern parts of the Cobar Basin. The Amphitheatre Group successions from the south Cobar Basin all received mixed detritus from the ca 430–410 Ma magmatic rocks exposed to the south, southwest and southeast of the basin (e.g. the Erimeran Granite, the Thule Granite, the Mineral Hill Volcanics, the Volcanics and Mount Hope Volcanics) and the Lower–Middle Ordovician turbiditic basement units. However, the north Cobar successions were predominately sourced from the Lower–Middle Ordovician turbiditic sequences with minor contributions from the Macquarie Arc to the east of the basin. Moreover, similar upward shifts of basin provenance have been observed in the Amphitheatre Group successions from both northern and southern parts of the basin. This stratigraphic provenance transition is evidenced by younger formations containing more detritus derived through recycling of the Lower–Middle Ordovician turbiditic rocks. The change in basin sources is interpreted to reflect modified basin paleogeography during the transition from rift phase to sag phase. The earlier rift-phase basin geography exhibits significant spatial variation in basin input, reflecting a fault-controlled deposition with contributions from local sources. During the later sag phase, the basin geography is characterised by a uniform deposition system with more homogenised basin fill derived from the west and northwest of the basin.

The Cobar Basin hosts mineral deposits that commonly occur within the earlier pre-sag-phase stratigraphy (i.e. lower Amphitheatre Group and Shume Formation) but are absent in the younger upper Amphitheatre Group rock units. This stratigraphic control on mineralising behaviour coincides with a shift in basin sources, implying provenance controls on mineralisation. The chemical difference between the mineral hosting and mineral-barren formations is interpreted to have resulted from different detrital mineral compositions derived from varying sources. The mineral-barren units contain higher SiO₂, Zr and Hf contents, reflecting a quartz-rich lithology with strong recycling signatures. In contrast, the mineral hosting units generally contain higher Al₂O₃, TiO₂, Ta and Nb contents with negative Eu anomalies, which we interpret to represent mineral compositions with more detrital feldspar, muscovite and Ti-bearing minerals. The reactions between these detrital minerals and metal-bearing fluids are considered as one of the key influences for some post-rift epigenetic mineralisation in the Cobar Basin.

Acknowledgements

The work has been supported by the Mineral Exploration Cooperative Research Centre whose activities are funded by the Australian Government’s Cooperative Research Centre Program. Steven Trigg, Liann Deyssing and Lorraine Campbell are thanked for their assistance during fieldwork and sample collection. GSNSW staff at the WB Clarke Geoscience Centre—Londonderry Drillcore Library are thanked for their help with drill core sampling. We thank Sun Yanyan for help with thin-section preparation. Chris Fergusson and Bill Collins are thanked for their helpful reviews of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the Supplemental data (Appendices 1 and 2).

Additional information

Funding

This research was funded by MinEx CRC and the Geological Survey of New South Wales (GSNSW). This is MinEx CRC Document 2023/323.