|
Discussion |
1 1Geological Survey, Primary Industries and Resources, South Australia, 101 Grenfell St, Adelaide, South Australia 5000, Australia
2 2Predictive Minerals Discovery CRC, Research School of Earth Sciences, Australian National University, Canberra, ACT 0200, Australia
3 3Geological Survey New South Wales, Department of Primary Industries, 32 Sulphide St, Broken Hill, NSW 2880, Australia
4 4School of Earth Sciences, The University of Melbourne, Victoria 3010, Australia
5 5Earth Sciences, University of New England, Armidale, NSW 2531, Australia
6 6Geoscience Australia, Canberra, ACT 2601, Australia
7 7Research School of Earth Sciences, Australian National University, Canberra, ACT 0200, Australia
 |
Introduction |
|---|
 TOP Introduction AcknowledgementsReferences |
|---|
The essence of the model is that a low-angle, extensional ‘detachment’ formed along the boundary between the Broken Hill and Sundown Groups in the Broken Hill Domain, and between the Bimba and Plumbago Formations in the Olary Domain (Gibson & Nutman, fig.1). The ‘detachment’ was alleged to have occurred as part of a sillimanite-grade D1 deformation between 1690 and 1670 Ma, and to have provided a channel for hydrothermal fluids that could have been responsible for regional scale Na (–Fe) alteration and formation of Pb–Zn mineral deposits, including the Broken Hill orebody. As pointed out below, most of the evidence is misinterpreted, misleading and/or ambiguous, and the geochronological case advanced by Gibson & Nutman for such an early (1690–1670 Ma) high-grade event has no foundation.
Ambiguities in the model.
Gibson & Nutman have difficulty in defining both the timing and position of their ‘detachment’. In Gibson & Nutman (fig. 3) the ‘detachment’ developed anywhere between 1720 and 1670 Ma. In their figure 1, the proposed ‘detachment’ is illustrated at the base of the Broken Hill Group, but in figure 2 the ‘detachment’ postdates the topmost unit of Broken Hill Group, the 1685 ± 3 Ma metavolcanic Hores Gneiss (Page et al. 2000; Stevens & Barron 2002). Similarly, Gibson & Nutman (p. 60) describe the location of their sample 96-BH-13 as ‘within or just below the detachment surface’, but the amphibolite was emplaced immediately below the Broken Hill Group, much lower than the proposed ‘detachment’.
Gibson & Nutman (p. 65) infer that the c.1675 Ma Pb model age (Carr & Sun 1996) for the Broken Hill orebody ‘is more closely aligned with the inferred age of peak M1 metamorphism (1670–1690 Ma) than the age of sedimentary deposition in the lower plate (
1705 Ma by Page et al. 2000).’ But it is the 1685 ± 3 Ma Hores Gneiss (Page et al. 2000), host to the Broken Hill orebody, that constrains the age of the Gibson & Nutman ‘detachment’, not the 1705 Ma age derived from lower in the succession. Additionally the c.1675 Ma Pb model age is now recalculated at c.1690 Ma (Page et al. 2005), and this closely approximates the depositional age of Hores Gneiss.
Importantly, if the ‘lower plate’ has moved upwards as a core complex, the ‘lower and upper plates’ should be unrelated, having originated from different crustal depths. The presence of the thin, precisely dated, 1693 ± 3 Ma Plumbago Formation (Page et al. 2000, 2005; Conor 2004) in both the proposed ‘upper plate’ and ‘lower plate’, indicates that there has been no such displacement, and that both ‘plates’ are part of a single stratigraphic succession. Willis et al. (1983) indicated this to be the case, a conclusion in accord with recent geochronology (Page et al. 2000).
Erroneous map detail.
Gibson & Nutman utilize their only detailed map (fig. 5, c.45 km2 Tommie Wattie Bore–Ameroo Hill area) as evidence for their extensional detachment, which they claim juxtaposes the Wiperaminga Subgroup in their ‘lower plate’ against Saltbush Subgroup in their ‘upper plate’. In disregarding much previous mapping (e.g. Laing 1996), figure 5 illustrates the weakness of the detachment model, because Gibson & Nutman have mistakenly placed their ‘detachment’ between the lower psammitic and upper pelitic parts of a single, c.2 km thick, upwards-fining package, i.e. Tommie Wattie Formation within the Wiperaminga Subgroup, i.e. ‘lower plate’.
Bimodal magmatism.
Gibson & Nutman claim abundant bimodal magmatism in the ‘lower plate’ over the period 1710–1670 Ma, but the only bimodal magmatism was the mafic-dominated 1690–1680 Ma event of the Broken Hill Domain (Gibson & Nutman; Page et al. 2005), and the equivalent c.1685 Ma Lady Louise Suite in the Olary Domain (Conor & Fanning 2001). Neither the A-type Basso Suite of the Olary Domain (Ashley et al. 1996; Conor 2000), comprising c.1715–1710 Ma felsic volcanic and intrusive rocks (Page et al. 2000), nor the 1704 ± 3 Ma granitic Alma Gneiss (Page et al. 2000) include confirmed mafic components.
The most plausible reason for the absence of amphibolites above the Broken Hill Group is that they predate deposition of the post-1685 ± 3 Ma Sundown Group. The conformable amphibolites in the Broken Hill Group are interpreted as sub-seafloor sills, with the dykes and sills in the lower Willyama Supergroup being feeders (Stevens 1998). The observed geometry of the amphibolites is not evidence for extensional tectonism, because, where the Broken Hill Group is at relatively low metamorphic grade, amphibolite sills and adjacent bedding show no evidence of transposition (e.g. Parnell Formation west of Mt Gipps Homestead), and conformable amphibolites are not restricted to the Broken Hill Group but also occur in the Wiperaminga Subgroup.
Metamorphic grade and abundance of migmatites.
Gibson & Nutman propose that granulite-facies metamorphism took place in the range 1691 ± 9 Ma to 1674 ± 8 Ma, but this encompasses the depositional time of the Broken Hill and Sundown Groups. The proposition leads to the untenable conclusion that either granulite-facies metamorphism occurred essentially synchronously with deposition, or that burial to c.18 km depth happened immediately after deposition of the Hores Gneiss. This contradicts the tectonic style required to form metamorphic core complexes: one of exhumation, not burial.
Gibson & Nutman claim there is a contrast in metamorphic grade across their alleged ‘detachment’, but regional mapping does not support such a metamorphic-grade contrast at either the Broken Hill Group–Sundown Group or the Bimba Formation–Plumbago Formation boundaries. Examples of such field evidence are given below.
(1) Around Broken Hill City and to the southeast, the S1 schistosity is defined by sillimanite and biotite in both Broken Hill and Sundown Groups (i.e. both above and below the ‘detachment’).
(2) Similarly in the Mundi Mundi Creek area 30 km northwest of Broken Hill, Broken Hill Group, Sundown Group and part of the overlying Paragon Group are metamorphosed to sillimanite grade.
(3) In the Yanco Glen area 30 km north of Broken Hill, the first appearance of sillimanite coincides with Broken Hill Group–Sundown Group boundary (Corbett 1979), but this coincidence is not the regional norm, and simply represents a gradational metamorphic progression.
(4) The pre-D2 andalusite shown in Gibson & Nutman figure 4a is purported to exemplify the lower-grade assemblage of their upper plate, but the sample is from the Tommie Wattie Formation (Wiperaminga Subgroup), and therefore from their ‘lower plate’. Not only have Gibson & Nutman wrongly identified the stratigraphy, but also the local M1 metamorphic grade of their ‘lower plate’ at this locality is at only mid-amphibolite facies.
(5) In the central and northern part of the Olary Domain, examination of numerous continuous drill hole intersections across the interval from the Curnamona Group to the Strathearn Group has shown no evidence of abrupt metamorphic grade change.
Migmatitic rocks are not restricted to the ‘lower plate’; partial melting is found in all parts of the Willyama Supergroup of appropriate bulk rock composition that experienced sillimanite-grade metamorphic conditions. Locally, at upper sillimanite or granulite grade, not only Thackaringa and Broken Hill Groups (i.e. ‘lower plate’), but also the Sundown and even Paragon Groups (‘upper plate’), exhibit abundant partial melting (Phillips 1980; Stevens et al. 1994). For example, Nutman & Ehlers (1998, fig. 14) illustrate a Sundown Group outcrop with folded D1 partial melt layers and less abundant axial plane D2 melt veins. Significantly, there are large tracts of lower Willyama Supergroup that are unaffected by partial melting, including that shown in the western part of Gibson & Nutman figure 5. This argues against the Gibson & Nutman model which requires the ‘lower plate’ to be of high metamorphic grade.
Oxidation state and Na(–Fe) alteration.
A critical component of the Gibson & Nutman model is the concentration of iron as magnetite and sodium as albite along the upper part of the ‘lower plate’ below their purported detachment structure. At a scale approximating 1:8 million (Gibson & Nutman, fig. 1) it is possible to generalize that the lower part of the Willyama Supergroup is both relatively oxidized and albitized, but the detail does not support a change attributable to any particular surface. The separate changes in both oxidation state and albitization are known to occur well below the stratigraphic position of the Gibson & Nutman (fig. 2) ‘detachment’. For example, in the western part of Gibson & Nutman, fig. 5, points ‘A’ and ‘B’ mark only a magnetite-bearing unit within the Tommie Wattie Formation. The measured base of the Tommie Wattie Formation, below which there is intense migmatization and albitization, is more than 2 km below the position of the proposed ‘detachment’. Increased magnetic susceptibility below the Bimba Formation in the Olary Domain is locally attributed to the presence of particular iron-rich lithostratigraphic units (Lottermoser & Ashley 1996), including the Peryhumuck Formation (Laing 1996). In the Broken Hill Domain, the Broken Hill Group, several hundred metres thick and immediately below the alleged ‘detachment’, is not albitized. Moreover, in the Broken Hill Domain there is no clear-cut redox boundary.
Na(–Fe) alteration affects Willyama Supergroup rocks both below and to a lesser degree above the proposed ‘detachment’. Field evidence suggests that Na(–Fe) alteration has a protracted history, commencing during diagenesis and continuing spasmodically throughout the Olarian Orogeny. Apart from the isotopic evidence constraining the Olarian Orogeny to c.1600 Ma or later (Page et al. 2004), alteration has been directly dated at c.1630 Ma from monazite (Teale & Fanning 2000), and 1588–1583 Ma from titanite associated with albitization (Skirrow & Ashley 2000). Similar ages have been obtained for both intense alteration (Kent et al. 2000) and Cu–Au–Mo mineralization (Teale 2000; Teale & Fanning 2000; Skirrow & Ashley 2000). All the above ages are well after the proposed timing of ‘detachment’ at c. 1690–1670 Ma, refuting the assertion by Gibson & Nutman (p. 64) of a temporal link between ‘detachment’, Na(–Fe) alteration and associated mineralization.
The cause of the regionally distinct redox change in the northern part of the Olary Domain, reflected by a prominent aeromagnetic gradient, is demonstrated by drilling at several mineral prospects (e.g. Polygonum, Hunters Dam, Kalkaroo, Portia). It is due to the change from sulphide (pyrite, pyrrhotite) ± graphite to oxide (magnetite) within a conformable package of greenschist to lower amphibolite facies metasiltstones. U–Pb dating of the succession above the redox boundary at North Portia gave an age of 1702 ± 6 Ma (Teale 2000), significantly older than either the 1693 Ma Plumbago Formation or the 1685 Ma Hores Gneiss. Thus the local redox boundary is well below the expected position of the proposed ‘detachment’.
Mineralization.
The spatial relationship between the alleged ‘detachment’ and mineralization, implied from Gibson & Nutman figure 1, is meaningless, given that they have drawn the wrong boundary on that figure, i.e. the top of the Thackaringa Group rather than the top of the Broken Hill Group. Leyh & Conor (2000) depicted a broad, strata-parallel mineral control (Pb–Zn and Cu–Au) within the Willyama Supergroup that Gibson & Nutman relate to their ‘detachment’ structure, but Pb–Zn mineralization is present throughout the interval from the Thackaringa Group to Paragon Group (Barnes 1988; Stevens et al. 2003), i.e. both ‘upper and lower plates’.
The alleged detachment.
A detachment structure of the dimension proposed by Gibson & Nutman would be expected to be obvious, especially in the northern areas where the Olarian overprint is of lower metamorphic grade. In the Olary Domain we have examined the stratigraphy containing the proposed ‘detachment’ over 300 km strike length. Mapping and logging of numerous drill holes showed no stratigraphic, structural or metamorphic evidence for a detachment surface at this stratigraphic boundary (Ashley et al. 1998). Where observed, the contact of the Bimba and Plumbago Formations (the alleged ‘detachment’) is the knife-sharp contact of a volcaniclastic unit overlying calc-silicate rock. Although Gibson & Nutman assert that the Bimba Formation is a mylonite, no supporting evidence is presented.
One of us (I.R.P.) supervised drilling of the central part of the Broken Hill ore deposit in 2002–2004, wherein the sequence from the Thackaringa Group to the Sundown Group was intersected. Neither open pit exposures nor over 100 diamond core holes provide evidence of a detachment surface.
Any detachment tectonics during deposition of the Willyama Supergroup should have left a sedimentological record. Sedimentation above a rising metamorphic core complex would have been affected by major tilting, listric faulting and erosion, leading to angular unconformities and deposition of coarse clastic rocks. In the Broken Hill Domain, the Sundown Group conformably overlies the Broken Hill Group, and is in turn overlain without angular unconformity by the Paragon Group, None of these units contain coarse clastics (Willis et al. 1983). Similarly in the Olary Domain, neither the Saltbush Subgroup, nor the Mt Howden Subgroup contains significant coarse clastics (Ashley et al. 1998; Conor 2000).
Timing of the proposed metamorphic core complex.
The Gibson & Nutman model calls for a tectonic break within the Willyama Supergroup and imposes a high-grade metamorphic imprint early in the basin development (1690–1670 Ma). However there is no known isotopic record for a metamorphic event at 1690–1670 Ma. Gibson & Nutman’s contention of a metamorphic origin for 1690–1670 Ma zircon is predicated on their belief that metamorphic hornblende and other mineral grains were trapped as inclusions within a few zircons in one of their four dated amphibolites. If 1690–1670 Ma magmatism and high-grade metamorphism were linked, might we not expect zircons in many of the 1690–1670 Ma bimodal magmatic rocks to contain metamorphic inclusions, and should we not also find regional evidence of metamorphic zircon or monazite of that age? There is no such evidence. Gibson & Nutman propose that the evidence was obliterated (by younger events), but this proposition is in conflict with a plethora of Curnamona Province U–Pb zircon data that include consistent populations of pre-metamorphic igneous and detrital zircon defining patterns from late Archaean, c. 1860 Ma, c.1820 Ma, c.1770–1790 Ma, and c.1730 Ma. (Nutman & Gibson 1998; Page et al. 2000).
To sustain their case ‘that these hornblende inclusions formed in textural and metamorphic equilibrium with their zircon hosts’, the authors need textural evidence that the metamorphic grain(s) are wholly and unequivocally sheathed by zircon. They have not provided this. The four zircon images in figure 7 show five inclusion–host relationships. All so-called inclusions are breached to an external surface of the host zircon. Without three-dimensional control on inclusion–host relationships, their so-called inclusion textures are ambiguous at best. In their dismissal of other possibilities, they claim that the zircon host is not fractured. This is not correct. Not only do the photomicrographs supplied by Gibson & Nutman show zircon populations with a normal density of cracks and fractures, but also three of the four zircons shown in figure 7 are strongly embayed and infilled by hornblende, plagioclase, or quartz. Grain 14 is multiply fractured and breached by a plagioclase ‘inclusion’.
To further support their case that hornblende grains are inclusions in textural and metamorphic equilibrium with a 1690–1670 Ma zircon host, Gibson & Nutman state that the Clevedale amphibolite zircons are not recrystallized. However, their figure 7 and other CL images provided to us by Gibson & Nutman demonstrate that a large proportion of their zircon suite is partly recrystallized. Relic prismatic zoning is truncated by younger discontinuous, high-U rims (e.g. grains 12, 14 in figure 7). Such high-U metamorphic selvedges are ubiquitous in zircons affected by the c.1600 Ma Olarian high-grade event (Page et al. 2005). Together with the ambiguity of the ‘inclusion-host’ relationships and widespread zircon fracturing, the existence of this younger zircon seriously erodes any notion that the ‘metamorphic inclusions’ formed in equilibrium with 1690–1670 Ma zircon host. The failure by Gibson & Nutman to recognize or acknowledge the existence of a younger, recrystallized zircon generation in the three Clevedale amphibolites contrasts with the evidence for metamorphic recrystallization at 1593 ± 19 Ma in their fourth site (Readymix) about 8 km west of the Clevedale samples. But at the latter site, we are astonished by their conclusion that ‘dyke emplacement and/or initial metamorphism is 1658 Ma’. The source of this 1658 Ma number (from a set of about 17 apparent 207Pb/206Pb ages between c.1660 and 1600 Ma, and one more discordant analysis at 1678 ± 28 Ma) is inexplicable.
Gibson & Nutman incorrectly maintain that a 1660 Ma 40Ar/39Ar cooling age from metamorphic hornblende provides independent evidence in support of high-grade metamorphism at 1690–1670 Ma: Harrison & McDougall (1981), the primary source of these data, discount any geological significance for this result. The 1660 Ma apparent age is but a single step of a poorly resolved age spectrum; any plateau is totally obscured by excess argon.
The tectonic model proposed by Gibson & Nutman for an early-formed metamorphic complex in the Curnamona Province is underpinned by their interpretation of a 1690–1670 Ma metamorphic age. We believe that the 1690–1670 Ma zircons provide minimum age(s) for mafic magmatism only. The claim of a metamorphic paragenesis remains unsupported, and so the timing evidence, a crucial plank of their overall model, has no foundation.
Allen Nutman kindly provided a set of photomicrographs and CL images of the studied zircons. This discussion is published with permission of the Deputy Director General, New South Wales Primary Industries Mineral Resources, and Executive Director, Minerals and Energy, Primary Industries and Resources, South Australia.
6 September 2004
George Gibson & Allen Nutman reply: Ongoing uncertainties about the origin and timing of economically important Pb–Zn mineralization in the Palaeoproterozoic Willyama Supergroup have ensured that any reinterpretation of the tectonothermal history of the Broken Hill and neighbouring Olary regions will attract close scrutiny and intense debate. It is therefore no surprise that Conor et al. should take issue with the main conclusions of our paper, viz. that: (1) the Willyama Supergroup incorporates one or more detachment surfaces and (2) underwent regional metamorphism and related deformation as early as 1670–1690 Ma.
Their broadest criticism is that we have written a model-driven paper, full of unsubstantiated assertions and scientifically persuasive only because we ignored much published and unpublished literature containing evidence to the contrary. While it is important to acknowledge that our study of early deformation and related low-pressure–high-temperature metamorphism in the Broken Hill area was informed by modern developments in numerical modelling and theoretical petrology (e.g. De Yoreo et al. 1991), we emphasize that this did not prejudice the manner in which we conducted our research or interpreted the results. The supposed lack of corroborating evidence for our interpretation of the regional geology only comes about because, our work aside, Conor et al. are unduly dismissive of several independent structural and geochronological studies documenting (1) pre-1600 Ma thermal events and fabrics (e.g. Teale & Fanning 2000; Skirrow et al. 2000) and (2) disruption of the regional stratigraphy by early, layer-parallel, high-temperature shear zones (White et al. 1995, 1996; Raetz et al. 2002). Conor et al. make little or no mention of these layer-parallel shear zones in their many publications on the Willyama Supergroup (e.g. Stevens et al. 1988; Parr & Plimer 1993; Ashley et al. 1997, 1998) and evidently have yet to be persuaded that these structures exist.
Opinion may differ as to the exact timing and origin of these shear zones (extensional or contractional), but there is no denying their existence or that some of them formed early because they are deformed by what we and others (e.g. White et al. 1995; Gibson et al. 2004) have interpreted to be folds of D2 and D3 age. We attributed a D1 age to the early shear zones and associated layer-parallel fabrics, and argued that they are extensional in origin because they locally excise stratigraphy and emplaced younger on older rocks (cf. White et al. 1995, 1996). In contrast, Conor et al. continue to subscribe to an interpretation where the layer-parallel (S1) fabric is axial planar to as yet unrecognized regional-scale D1 recumbent folds and/or nappes (Conor 2004).
Ambiguities in the model.
As rightly pointed out by Conor et al., our proposed detachment does not always occur at the same stratigraphic position in the Broken Hill Domain, and nor should it. Depending on the degree of crustal excision, the boundary between our two metasedimentary sequences (fig. 1) may lie within the Broken Hill Group or as high as the base of the Sundown Group. We stated as much on p. 59 of our paper and tried to encapsulate these structural relations in figure 2 by representing the Broken Hill Group as a wedge of variable thickness caught up between rocks belonging unequivocally to our upper and lower plates. This is consistent with the observation that the Broken Hill Group (and contained Ettlewood Calc-silicate member) in some areas is completely missing so that Sundown Group rocks rest directly on Thackaringa Group (White et al. 1996). Such excisions of stratigraphy are strongly suggestive of extensional processes and are no less common at Ameroo Hill in the Olary Domain where the Plumbago and Bimba Formations are progressively cut out over a strike length of several kilometres (fig. 5). Just as the Ettlewood Calc-silicate member is not always structurally preserved in the Broken Hill Domain, neither is the Bimba Formation ubiquitously developed across the Olary Domain (e.g. west of Ameroo Hill; fig. 5). There is no inconsistency between our two figures, only recognition that a complete stratigraphy need not be preserved everywhere.
In view of such considerations, we fail to understand why the field relations of either the Plumbago Formation or Hores Gneiss in the upper part of the Broken Hill Group provide grounds for rejecting our central tenet that the upper and lower plates originally developed at different crustal levels. The Plumbago Formation overlies the Bimba and Ettlewood Calc-silicate units (Conor 2004), and along with Hores Gneiss, might be reasonably included in our upper plate. The Hores Gneiss locally preserves textural evidence for a volcanic or hypabyssal origin (Page & Laing 1992), and formed at shallower crustal levels than older parts of the Willyama Supergroup where intrusive rocks predominate. If there is ongoing debate about the exact position of our proposed detachment within the Broken Hill Group, it is only because this part of stratigraphy has been strongly tectonized and, along with the underlying Thackaringa Group, extensively intruded by magmatic rocks so that the boundary between our two metasedimentary sequences is no longer immediately obvious.
Unique interpretation of Pb isotopic data in polymetamorphic terranes with mixed sources is notoriously difficult. We therefore remain unconvinced that the recalculated 1690 Ma model age for the Broken Hill orebody has any real geological significance and note that in a more recent study of Pb isotopes in the Willyama Supergroup, Parr et al. (2004) refrained from assigning any age to mineralization (cf. CSIRO–Geoscience Australia model age of 1670 Ma).
Erroneous map detail.
The stratigraphic affinities of the Tommie Wattie Formation have yet to be resolved, although Grady et al. (1989), like ourselves, correlated chiastolite-bearing pelitic schists making up the upper part of this unit with the Saltbush Group. Their boundary between lower and upper Tommie Wattie Formation is identical to ours and it is debatable whether these schists should be included in the same upward-fining psammopelitic sequence as the lower part of the formation as advocated by Laing (1996) and endorsed by Conor et al.
Bimodal magmatism.
Our interpretation does not require bimodal magmatism before 1700 Ma, only that the bulk of it takes place at 1690–1670 Ma. Moreover, unless there are serious deficiencies in the published geological maps, it is difficult to sustain the argument that all basaltic magmatism in the Willyama Supergroup occurred between 1690 and 1670 Ma as Conor et al. contend. On the Pinnacles 1:25 000 geological map, amphibolite dykes terminate abruptly against granitic orthogneiss mapped as part of the 1705 Ma Alma Gneiss (Brown 1984) whereas elsewhere dykes of similar composition cut through this unit as well as 1680 Ma granitic orthogneiss (Rasp Ridge Gneiss). These field relations indicate to us that basaltic dyke intrusion commenced before intrusion of the 1705 Ma Alma Gneiss and continued episodically until after 1680 Ma. Bimodal magmatism in the Willyama Supergroup has a longer history than Conor et al. envision.
It is debatable whether the absence of amphibolite dykes and sills in the Sundown Group is simply due to a cessation of basaltic magmatism before this unit was deposited. Our alternative interpretation that basaltic intrusion was controlled by crustal thinning and restricted to deeper crustal levels (lower plate) is equally plausible and better explains why magmas of both basaltic and felsic composition reached a comparable position of neutral buoyancy in the crust: both were emplaced into the same crustal-scale shear zone.
Metamorphic grade and abundance of migmatites.
In judging that there is no contrast in metamorphic grade across our proposed detachment, Conor et al. make little or no distinction between the effects of three successive tectonothermal events, any one of which could have produced the conditions necessary for partial melting and sillimanite-grade metamorphism. We have observed migmatitic melts injected along the axial surfaces of both D2 and D3 folds in the Broken Hill Domain but these are largely irrelevant to the issues debated here, indicating only that some parts of the sequence were carried to deeper crustal levels and higher temperatures during post-D1 crustal thickening. The key observation is that sillimanite and migmatites of D1 age are overwhelmingly restricted to our lower plate, along with the bulk of the magmatic rocks, as might be expected of a situation where partial melting, sillimanite-grade metamorphism and magmatic intrusion are all manifestations of the same tectonothermal process. This, coupled with the fact that contacts between migmatized and non-migmatized parts of the Willyama Supergroup in the Olary Domain are typically sharp and coincident with high-strain shear zones, led us to conclude that our two variably migmatized sequences had been tectonically juxtaposed (see also Davies & Anderson 2000).
We accept that the lower part of Tommie Wattie Formation is less intensely migmatized than other parts of our lower plate. Its inclusion in the lower plate is justified on the grounds that, along strike to the south, rocks with the same magnetic character and degree of Na–Fe alteration are overlain by calc-silicate rocks (Ethiudna Subgroup) belonging undisputedly to the lower metasedimentary sequence.
Pre-D2 andalusite shown in figure 4a of our paper is from graphite-bearing schist in the upper part of Tommie Wattie Formation and therefore belongs to the more weakly metamorphosed upper plate (fig. 5). Peak metamorphism in the two plates took place at different times and under different conditions. Swapp & Frost (2003) have shown that early metamorphism in lower plate rocks of the Broken Hill Domain followed a clockwise P–T–t path with pressures initially in excess of 600 MPa before subsequent decompression and lower pressure metamorphism (500–600 MPa). They also calculated that temperatures during peak M1 metamorphism may have been as high as 850 °C, consistent with either extreme crustal thinning and/or voluminous intrusion of basaltic magma. In contrast, prograde metamorphism in the overlying Sundown Group (upper plate) is characterized by isobaric heating and an anticlockwise P–T–t path wherein maximum pressures did not exceed 450 MPa, and M1 mineral assemblages are dominated by andalusite rather than sillimanite (Hobbs et al. 1984; our paper). Such differences in metamorphic conditions and P–T–t paths are difficult to explain in the absence of any structural break or discontinuity between the two sequences.
Contrasting P–T–t paths such as these match the thermal structure induced by extension in a two-layer crustal model (De Yoreo et al. 1991). In this model, crustal thinning and ductile deformation only take place in the lower crust, which initially undergoes isothermal decompression followed by near isobaric cooling. Conversely, crust above the detachment (c. 15 km) undergoes nearly isobaric heating accompanied by brittle faulting with little or no uplift. Moreover, if the conditions for low-pressure–high-temperature metamorphism are to be realised at mid-crustal levels without geologically unreasonable amounts of crustal thinning, it is unavoidable that metamorphism be accompanied by copious amounts of magmatic intrusion and related heating (De Yoreo et al. 1991) as was the case at Broken Hill. Alternative models for low-pressure–high-temperature metamorphism in this region based on crustal thickening are beset by greater problems because there is no obvious heat source at 1600 Ma to generate the anomalously high thermal gradients such metamorphism requires.
Oxidation state and Na(–Fe) alteration.
It is not disputed that there was more than one pulse of Na–Fe alteration or that later pulses affected rocks in both the upper and lower plates. We simply observed that there is a strong structural control on the earliest recognisable phase of Na–Fe alteration and that the transformation of pre-existing psammopelitic rocks into magnetite-bearing albitites is particularly intense in the upper part of our proposed lower plate, at least in the Olary Domain. Conor et al. may debate the exact position of our proposed detachment at some localities, but this does not invalidate our suggestion of a fundamental first-order tectonic control on the distribution of Na–Fe alteration throughout the Willyama Supergroup (see also Davies & Anderson 2000).
Mineralization.
The tectonic boundary between our two metasedimentary sequences is not a single shear surface, but a zone of distributed D1 strain and associated mylonite formation. In the Broken Hill Domain, this zone encompasses much of the lower Broken Hill Group and immediately underlying Thackaringa Group (see also White et al. 1995, 1996). The Broken Hill Line of Lode lies within this zone of intense mylonitization and cannot be interpreted in isolation of either its regional tectonic setting or its position within such an important structure. The issue is not whether there is a spatial relationship between mineralization and deformation as Conor et al. contend, but to what degree the present concentration of ore is the result of pre- or syn-D1 processes. We drew a link between albitization, metal depletion and mylonitization, leading us to speculate that formation of the Broken Hill Line of Lode (as opposed to any mineralization originally present) was the result of dynamically induced fluid flow in an extensional environment.
Alleged detachment.
Layer-parallel, high-temperature shear zones and mylonites have only recently been recognized in the Broken Hill Domain and in some cases have reportedly been misidentified as stratigraphic units (White et al. 1996). Notwithstanding the extent to which mylonitic fabrics are developed throughout the Broken Hill and Thackaringa Groups, Conor et al. insist that there is no evidence for a detachment in drill-core examined by one of them (I.R.P.) from either stratigraphic unit. We believe such conclusions are inconsistent with the available structural evidence and point out that comparable amounts of layer-parallel shearing occur at about the same stratigraphic level at Ameroo Hill in the Olary Domain (fig. 5).
Rocks caught up in the Ameroo Hill shear zone include the Bimba Formation as well as veined migmatitic gneisses immediately below this formation that have been transformed into equigranular biotite–plagioclase mylonitic schist (mapped informally as salt n’ pepper schist by Gibson 2002). Contacts between this schist and the underlying migmatitic rocks are typically sharp, and even though S–C fabrics and other mylonitic structures are only indifferently preserved due to recrystallization accompanying later events, there is no mistaking the tectonic character of this boundary. Based on their own mapping and drill-core logging, Conor et al. insist that no major structural break exists at this stratigraphic level even though there is clear evidence for the progressive excision of lithological units along strike (fig. 5). Near identical structural relations have been observed by the senior author (G.M.G.) elsewhere in the Olary Domain, including several well-known localities described in the excursion guides of Ashley et al. (1997, 1998). We can only surmise that because these mylonitic horizons are layer-parallel, they have gone unrecognized like those at Broken Hill and been previously mapped as part of the regional stratigraphy.
While no photograph of mylonitized Bimba Formation was provided in our paper, three of our critics (C.H.H.C., R.W.P., and W.V.P.) have examined the mylonites at Ameroo Hill on more than one occasion with the senior author.
Timing of proposed metamorphic core complex.
Our own geochronological data notwithstanding, Conor et al. claim that there is no isotopic record for metamorphism at 1670–1690 Ma. They point out that metamorphic overgrowths on older zircons consistently give 1600 Ma ages (e.g. Page et al. 2000) in keeping with their view that there was no major tectonothermal event before the Mesoproterozoic Olarian orogeny. We have always agreed (e.g. Nutman & Ehlers 1998; Gibson et al. 2004) that there was major orogenesis at 1600 Ma but in Gibson & Nutman (2004) we argue that this was not responsible for either initial low-pressure–high-temperature metamorphism or formation of the layer-parallel fabrics (S1). A pre-1600 Ma age for the onset of deformation and metamorphism is supported by (1) a 1630 Ma monazite age for albitization that overprints and thus postdates two deformational fabrics (Teale & Fanning 2000) and (2) our own 1670–1690 Ma ages obtained from zircon in which grains of metamorphic amphibole and epidote are embedded. Conor et al. are not convinced that these embedded crystals represent either true inclusions or that they ever formed in metamorphic equilibrium with their zircon hosts (cf. younger overgrowths). Yet many of the 1670–1690 Ma ages obtained by us derive from zircon immediately adjacent to these inclusions. Moreover, given the number of grains in any one mount, there is no justification for the assumption that all inclusions identified by us originally breached the surface, as implied by Conor et al. This is best illustrated by laser ablation ICP–MS analysis burrowing to a depth of c. 30 µm to establish whether minerals seen associated with zircon in transmitted light are genuine inclusions. The results of one such analysis are presented in Figure 1, demonstrating unequivocally that the mineral (most likely hornblende because it contains c. 10 wt% CaO and 10 wt% Al2O3 as well as FeO and MgO) is fully sheathed by its zircon host. We stand by our conclusion that at least some zircon dated at 1670–1690 Ma in the Broken Hill amphibolites is of metamorphic rather than magmatic origin.
|
It is incorrect to suggest that the younger metamorphic overgrowths on zircon were not recognized by us; they are clearly visible in CL images reproduced in Gibson & Nutman figure 7, but are too narrow for meaningful age analysis. No attempt was made to date these narrow rims and it is for this reason alone that we did not report any younger ages from three of our four amphibolite samples.
The zircon fractures mentioned by Conor et al. (visible in the CL image we provided) cut across both the inclusions and zircon host alike and, consequently, have no bearing on the issues at hand. They are unequivocally late features, quite unrelated to the radial fractures that might have formed if there had been a volumetric change accompanying the replacement of pre-existing pyroxene by amphibole.
Concluding statement.
Opposing views and strongly held opinions are inevitable in an area as complex and economically important as Broken Hill. It is therefore pleasing that some of the longstanding problems concerning the regional structure and tectonothermal evolution of this important mineral province are being debated in a prominent international forum. A voluminous amount of literature has been published on this sequence (over 1000 papers, abstracts and theses since 1994 alone) and it is unavoidable that pressure on journal space will lead to restrictions regarding the number of references cited. As is common practice in the international literature, we cited only readily accessible sources, and make no apology for the fact that we chose to cite a disproportionately large number of key structural and metamorphic papers, some of which Conor et al. have disregarded or failed to take into account (e.g. White et al. 1995).
|
Acknowledgements |
|---|
|
TOPIntroductionAcknowledgementsReferences |
|---|
25 November 2004
|
References |
|---|
|
TOPIntroductionAcknowledgementsReferences |
|---|
Ashley, P.M., Cook, N.D.C. & Fanning, C.M. 1996. Geochemistry and age of metamorphosed felsic igneous rocks with A-type affinities in the Willyama Supergroup, Olary Block, South Australia and implications for minerals exploration. Lithos, 38, 167–184.[CrossRef][ISI][GeoRef]
Ashley, P.M., Lawie, D.C., Conor, C.H.H. & Plimer, I.R. 1997. Geology of the Olary Domain, Curnamona Province, South Australia and field guide to 1997 excursion stops.. Department of Mines and Energy, South Australia, 97/17.
Ashley, P.M., Conor, C.H.H. & Skirrow, R.G. 1998. Geology of the Olary Domain, Curnamona Province, South Australia. In: Field Guidebook to the Broken Hill Exploration Initiative Excursion 13–15 October 1998. Primary Industries and Resources South Australia.
Barnes, R.G., 1988. Metallogenic studies of the Broken Hill and Euriowie Blocks, New South Wales. 1. Styles of Mineralization in the Broken Hill Block. Geological Survey of New South Wales Bulletin, 32(1), 1–115.
Brown, R.E., 1984. Pinnacles 1:25 000 Geological Sheet 7133-1-N. New South Wales Geological Survey, Sydney.
Carr, G.R., Sun, S-S. 1996. Lead isotope models applied to Broken Hill style terrains – syngenetic vs epigenetic metallogenesis. In: Pongratz, J. & Davidson, G.J. (eds) New Developments in Broken Hill-type Deposits. CODES Special Publication, 1, 77–87.
Conor, C.H.H., 2000. Definition of major sedimentary and igneous units of the Olary Domain, Curnamona Province. MESA Journal, 19, 51–56.[GeoRef]
Conor, C.H.H., 2004. Plumbago Formation – a critical chronostratigraphic marker of the Willyama Supergroup, connecting the Olary and Broken Hill Domains in the southern Curnamona Province. MESA Journal, 34, 26–31.[GeoRef]
Conor, C.H.H. & Fanning, C.M. 2001. Geochronology of the Woman-in-White amphibolite, Olary Domain. MESA Journal, 20, 41–43.[GeoRef]
Corbett, G.J., 1979. Structure, metamorphism and stratigraphy of part of the Willyama Complex near Yanco Glen, N.S.W.. PhD Thesis, University of New South Wales.
Davies, B.J. & Anderson, H. 2000. Olarian waves: deformation, fluids and alteration in the western Willyama. Australian Geological Survey Organisation Record, 2000/10, 29–32.
De Yoreo, J.J., Lux, D.R. & Guidotti, C.V. 1991. Thermal modelling in low pressure/high temperature metamorphic belts. Tectonophysics, 188, 209–238.[CrossRef][ISI][GeoRef]
Gibson, G.M. & Nutman, A.P. 2004. Detachment faulting and bimodal magmatism in the Palaeoproterozoic Willyama Supergroup, south-central Australia: keys to recognition of a multiply deformed Precambrian metamorphic core complex. Journal of the Geological Society, London, 161, 55–66.
Gibson, G.M., 2002. Notes for a Broken Hill Exploration Initiative Field Workshop. In: Fabris, A.J. (ed.) Curnamona Province information package for Olary and Mingary Field workshop, October 2002. Mineral Exploration Data Package, South Australia Department of Primary Industries and Resources.
Gibson, G.M., Peljo, M. & Chamberlain, T. 2004 Evidence and timing of crustal extension versus shortening in the early tectonothermal evolution of a Proterozoic continental rift sequence at Broken Hill, Australia. Tectonics, 23, doi: 10.1029/2003TC001552.
Grady, A.E., Flint, D.J. & Wiltshire, R.J. 1989. Excursion guide for Willyama Supergroup and related rocks, Olary district, S.A.. Department of Mines and Energy, South Australia, Report Books, 89/23.
Harrison, T.M. & McDougall, I. 1981. Excess 40Ar in metamorphic rocks from Broken Hill, New South Wales: implications for 40Ar/39Ar age spectra and the thermal history of the region. Earth and Planetary Science Letters, 55, 123–149.[CrossRef][ISI][GeoRef]
Hobbs, B.E., Archibald, N.J., Etheridge, M.A. & Wall, V.J. 1984. Tectonic history of the Broken Hill Block, Australia. In: Kroner, A. & Greiling, R. (eds) Precambrian Tectonics Illustrated. E. Schweizerbart'sche, Stuttgart, 353–368.
Kent, A.J.R., Ashley, P.M. & Fanning, C.M. 2000. Metasomatic alteration associated with regional metamorphism: an example from the Willyama Supergroup, South Australia. Lithos, 54, 33–62.[CrossRef][ISI][GeoRef]
Laing, W.P., 1996. Stratigraphic subdivision of the Willyama Supergroup – Olary Domain, South Australia. MESA Journal, 2, 39–48.[GeoRef]
Leyh, W.R. & Conor, C.H.H. 2000. Stratigraphically controlled metallogenic zonation associated with the regional redox boundary of the Willyama Supergroup — economic implications for the southern Curnamona Province. MESA Journal, 16, 39–47.[GeoRef]
Lottermoser, B.G. & Ashley, P.M. 1996. Geochemistry and exploration significance of ironstones and barite-rich rocks in the Proterozoic Willyama Supergroup, Olary Block, South Australia. Journal of Geochemical Exploration, 57, 57–73.[GeoRef]
Nutman, A.P. & Ehlers, K. 1998. Evidence for multiple Palaeoproterozoic thermal events and magmatism adjacent to the Broken Hill Pb–Zn–Ag orebody, Australia. Precambrian Research, 90, 203–238.[CrossRef][ISI][GeoRef]
Nutman, A.P. & Gibson, G.M. 1998. Zircon ages from metasediments, granites and mafic intrusions: reappraisal of the Willyama Supergroup. Australian Geological Survey Organisation Record, 1998/25, 86–88.
Page, R.W. & Laing, W.P. 1992. Felsic metavolcanic rocks related to the Broken Hill Pb–Zn–Ag orebody, Australia: geology, depositional age, and timing of high-grade metamorphism. Economic Geology, 87, 2138–2168.[Abstract][ISI][GeoRef]
Page, R.W., Stevens, B.P.J., Gibson, G.M. & Conor, C.H.H. 2000. Geochronology of Willyama Supergroup rocks between Olary and Broken Hill, and comparison to northern Australia. Australian Geological Survey Organisation Record, 2000/10, 72–75.
Page, R.W., Stevens, B.P.J. & Gibson, G.M. 2005 Geochronology of the sequence hosting the Broken Hill Pb–Zn–Ag orebody, Australia. Economic Geology, in press.
Parr, J.M. & Plimer, I.R. 1993. Models for Broken Hill-type lead–zinc–silver deposits. In: Kirkham, R.V., Sinclair, W.D., Thorpe, R.I. & Duke, J.M. (eds) Mineral Deposit Modeling. Geological Association of Canada Special Papers, 40, 253–288.
Parr, J.M., Stevens, B.J.P., Carr, G.R. & Page, R.W. 2004. Subseafloor origin for Broken Hill Pb–Zn–Ag mineralisation, New South Wales, Australia. Geology, 32, 589–592.
Phillips, G.N., 1980. Water activity changes across an amphibolite–granulite facies transition, Broken Hill, Australia. Contributions to Mineralogy and Petrology, 75, 377–386.[CrossRef]
Raetz, M., Krabbendam, M. & Donaghy, A.G. 2002. Comparison of U–Pb zircon data from the Willyama Supergroup, Broken Hill region, Australia: evidence for three tectonostratigraphic successions and four magmatic events? Australian Journal of Earth Sciences, 49, 965–984.[CrossRef][ISI][GeoRef]
Skirrow, R.G. & Ashley, P.M. 2000. Proterozoic Cu–Au systems of the Curnamona Province – members of a global family? MESA Journal, 19, 48–50.[GeoRef]
Skirrow, R.G., Ashley, P.M., McNaughton, N.J. & Suzuki, K. 2000. Time–space framework of Cu–Au–(Mo) and regional alteration systems in the Curnamona Province. Australian Geological Survey Organisation Record, 2000/10, 83–86.
Stevens, B.P.J., 1998. The origins of Broken Hill rock types and their relevance to mineralisation. Australian Geological Survey Organisation Record, 1998/25, 109–114.
Stevens, B.P.J. & Barron, L.M. 2002. Volcanic textures in the Palaeoproterozoic Hores Gneiss, Broken Hill, Australia. Quarterly Notes of the Geological Society of New South Wales, 113, 1–22.
Stevens, B.P.J., Barnes, R.G., Brown, R.E., Stroud, W.J. & Willis, I.L. 1988. The Willyama Supergroup in the Broken Hill and Euriowie Blocks, New South Wales. Precambrian Research, 40/41, 297–327.[CrossRef]
Stevens, B.P.J., Willis, I.L., Brown, R.E. & Often, M. 1994. Purnamoota–Mundi Mundi Creek 1:25,000 Geological Sheet 7134-I-S and 7134-IV-S. Geological Survey of New South Wales, 7134.
Stevens, B.P.J., Barratt, R.M. & Conor, C.H.H. 2003. Could there be a Mt Isa type orebody in the Curnamona Province? Australian Geological Survey Organisation Record, 2003/13, 170–173.
Swapp, S.M. & Frost, B.R. 2003. Evidence of high-pressure metamorphism in the granulites of the Broken Hill area. Geoscience Australia Record, 2003/13, 174–175.
Teale, G.S. & Fanning, C.M. 2000. The timing of Cu–Au mineralisation in the Curnamona Province. Australian Geological Survey Organisation Record, 2000/10, 98–100.
Teale, G.S., 2000. Known and potential mineralisation styles, Benagerie Ridge Magnetic Complex. MESA Journal, 19, 43–45.[GeoRef]
White, S.H., Rothery, E., Lips, A.L.W. & Barclay, T.J.R. 1995. Broken Hill area, Australia, as a Proterozoic fold and thrust belt: implications for the Broken Hill base-metal deposit. Transactions of Institute of Mining and Metallurgy, 104, B1–17.
White, S.H., Rothery, E., Lips, A.L.W. & Barclay, T.J.R. 1996. Broken Hill area, Australia, as a Proterozoic fold and thrust belt: implications for the Broken Hill base-metal deposit: comments and reply. Transactions of Institute of Mining and Metallurgy, 105, B90–B98.
Willis, I.L., Brown, R.E., Stroud, W.J. & Stevens, B.P.J. 1983. The Early Proterozoic Willyama Supergroup: stratigraphic subdivisions and interpretations of high- to low-grade metamorphic rocks in the Broken Hill Block, New South Wales. Geological Society of Australia Journal, 30, 195–224.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
