Hydrothermal alteration styles in ancient and ...

New Zealand Journal of Geology & Geophysics, Vol. 52: 11–26 Craw et al.—Hydrothermal alteration in orogenic2009, gold deposits, NZ 0028–8306/09/5201–0011 © The Royal Society of New Zealand 2009

11

Hydrothermal alteration styles in ancient and modern orogenic gold deposits, New Zealand D. Craw P. Upton* D. J. MacKenzie Geology Department University of Otago PO Box 56 Dunedin 9054, New Zealand [email protected] *Present address: GNS Science, Private Bag 1930, Dunedin 9054, New Zealand. Abstract Orogenic hydrothermal systems in the South Island of New Zealand were active during Mesozoic and late Cenozoic collisional deformation and metamorphism of greywacke/schist terranes. Observations on the currently active mountain-building environment yield insights on processes occurring in the upper 5–15 km of the crust, and observations on an adjacent lithologically identical exhumed ancient mountain belt provide information on processes at 10–20 km in the crust. Hydrothermal fluids were mainly derived from metamorphic dehydration reactions and/or circulating topographically driven meteoric water in these mountain belts. Three geochemically and mineralogically different types of hydrothermal alteration and vein mineralisation occurred in these orogenic belts, and gold enrichment (locally economic) occurred in some examples of each of these three types. The first type of alteration involved fluids that were in or near chemical equilibrium with their greenschist facies host rocks. Fluid flow was controlled by discontinuous fractures, and by microshears and grain boundaries in host rocks, in zones from metres to hundreds of metres thick. Vein and alteration mineralogy was similar to that of the host rocks, and included calcite and chlorite. The second type of alteration occurred where the fluids were in distinct disequilibrium with the host rocks. Fracture permeability was important for fluid flow, but abundant host rock alteration occurred as well. The alteration zones were characterised by decomposition of chlorite and replacement by ankeritic carbonate in zones up to tens of metres thick. The mineralising fluid was deep-sourced and initially rock-equilibrated, with some meteoric input. The third type of mineralisation was controlled almost exclusively by fracture permeability, and host rock alteration was minor (centimetre scale). This mineralisation type commonly involved calcite and chlorite as vein and alteration minerals, and mineralisation fluids had a major meteoric water component. The three mineralisation types can be traced spatially and/or temporally from one to another with some overlap. The first type is characteristic of the deeper parts of an orogenic hydrothermal system, and this G08024: Online publication date 27 February 2009 Received 19 September 2008; accepted 16 January 2009

type gave way to the second type formed at shallower crustal levels, locally near to the surface. The third type of alteration is typically a late-stage, shallow-level phenomenon. Keywords gold; orogenic; metamorphism; greenschist; ankerite; alteration; vein; arsenic INTRODUCTION Alteration zones in host rocks around hydrothermal gold deposits are important for economic and scientific reasons. Altered rocks expand the width of the target zone at the exploration stage, and well-defined alteration zones can provide vectors for ore (Bohlke 1988; Eilu & Mikucki 1998; Bierlein & Crowe 2000). Commonly, alteration zones, especially the innermost zones, contain disseminated gold that can substantially increase the mining width and overall tonnage of ore (Bohlke 1988; Eilu & Mikucki 1998; Bierlein & Maher 2001). Mineral assemblages in alteration zones can provide information on temperatures, pressures, and fluid composition during mineralisation, thereby facilitating construction of models of formation of ore deposits (McCuaig & Kerrich 1998; Bierlein & Crowe 2000; Goldfarb et al. 2005). Orogenic (mesothermal) gold deposits are known to have varying degrees of alteration associated with their emplacement, but that alteration can be subtle (Clark et al. 1989; Gao & Kwak 1997; Bierlein & Crowe 2000; Bierlein & Maher 2001). Alteration has been detected around major orogenic gold systems on a scale of hundreds of metres where appropriate rock types are present (Bohlke 1988; Clark et al. 1989; Bierlein & Maher 2001), but other orogenic deposits appear to have no significant alteration at all (Bierlein & Crowe 2000; Goldfarb et al. 2005). Some orogenic deposits have fluid flow controlled by microshears and alteration similar to metamorphic processes (Cox et al. 1991; Witt & Vanderhor 1998), whereas others are strongly controlled by open fracture permeability (Bierlein & Crowe 2000; Goldfarb et al. 2005). Alteration zonation at the deposit scale has been well described (Kishida & Kerrich 1987; Bohlke 1988; Clark et al. 1989; McCuaig & Kerrich 1998), but more general orogenscale alteration zonation is less well defined. The controls on formation, or lack of formation, of hydrothermal alteration zones around orogenic gold deposits in general are poorly understood, and while there are some unifying models for formation of orogenic gold deposits (e.g., Hagemann et al. 1994; Groves et al. 1998; Bierlein & Crowe 2000; Goldfarb et al. 2005), the associated alteration characteristics are less well understood. This paper attempts to fill this gap in knowledge of orogenic gold deposit alteration by examining two different mineralisation systems in southern New Zealand. Both these orogenic gold mineralisation systems affected the same suite of host rocks which are broadly uniform over large areas, so the vagaries of different rock types can be removed from

12

New Zealand Journal of Geology and Geophysics, 2009, Vol. 52 Fig. 1 Digital terrane model of the South Island of New Zealand (from Geographx.co.nz), showing the Southern Alps active mountain belt and the adjacent Mesozoic Otago Schist Belt, made up of Caples, Torlesse, and Aspiring Terrane rocks. Both these orogens have gold deposits with differing alteration styles as indicated. Zones in which ankeritic alteration is common are overprinted with dark grey ornament.

the mix of variables. In addition, there is no evidence for syn-formational igneous effects on the gold mineralisation processes or alteration in either of the orogenic systems on this study. Hence, these orogenic gold systems are ideal for taking the first steps at constructing a large-scale hydrothermal alteration model. Further, one of these orogenic gold systems is currently active, so the tectonic setting for the hydrothermal alteration processes is well-defined. Consequently, this paper describes a combination of tectonic, structural, and geochemical processes that we build into a general model for orogenic gold alteration through space and time. Some of the data presented herein have been published previously, but this paper brings together a wide range of observations over both orogenic systems for the first time. GENERAL GEOLOGY AND TECTONIC SETTINGS The host rocks for both orogenic systems described in this study are variably deformed and recrystallised metasedimentary rocks in a large but monotonous Mesozoic accretionary complex. The metasedimentary rocks fall broadly into three terranes (Fig. 1) that have been juxtaposed during regional metamorphism (Mortimer 1993). The Torlesse Terrane is volumetrically dominant, and this includes a range of variably quartzofeldspathic metagreywackes to the northeast of

the Otago Schist Belt (Mortimer 1993) (Fig. 1). The Caples Terrane also consists of metagreywackes, but with a slightly higher volcanogenic component (Mortimer 1993), to the south of the Otago Schist (Fig. 1). These two terranes increase in strain and metamorphic grade towards the Otago Schist, which has upper greenschist facies rocks exposed in the core. This schist belt consists of highly deformed and possibly interdigitated Caples and Torlesse Terrane rocks, as well as an additional component, the Aspiring Terrane, which contains abundant (locally >5%) metabasite horizons. Collisional deformation and metamorphism that yielded the Otago Schist metamorphic pile occurred in the Jurassic to Early Cretaceous (Mortimer 1993; Gray & Foster 2004). Erosional and extensional tectonic unroofing of the Otago Schist began in the middle Cretaceous as the regional tectonics evolved from collisional to extensional (Gray & Foster 2004). Low-angle normal faults with 1–10 km scale offsets juxtaposed blocks of Otago Schist of differing metamorphic grade (e.g., upper and lower greenschist facies) during uplift and extension (Deckert et al. 2002; Gray & Foster 2004). This regional extension continued until the Oligocene associated with breakup of Gondwana in the Southwest Pacific. Subsidence of thinned crust permitted marine transgression and deposition of a thin veneer of Cretaceous–Oligocene marine sediments unconformably on top of the Otago Schist (Landis et al. 2008).

Craw et al.—Hydrothermal alteration in orogenic gold deposits, NZ Realignment of the Pacific-Australia plate boundary in the Oligocene and Miocene resulted in development of the Alpine Fault (Fig. 1) as a new plate boundary through New Zealand (Cooper et al. 1987). The Southern Alps mountain range developed immediately southeast of this plate boundary (Fig. 1) in response to renewed compressional deformation, with a strong component of dextral motion as well (Cooper et al. 1987). The Southern Alps were initiated in thick Otago Schist rocks in the Miocene and evolved northeastwards in the Pliocene (Craw et al. 2006). The Southern Alps are still undergoing active transpressional deformation (plate vector indicated in fig. 1 from De Mets et al. 1994), with uplift rates varying from as high as 8 mm/ yr near the Alpine Fault to rock T water, minor CO2, salts rock-exchanged, meteoric syn-orogenic, near-surface

Craw et al.—Hydrothermal alteration in orogenic gold deposits, NZ

15

Fig. 2 Summary of mineral assemblages in host greenschist facies rocks and the three alteration types described in this study. Minerals outlined with grey are primarily hydrothermal minerals relevant to gold mineralisation.

of alteration zones are diffuse and irregular in the vicinity of controlling structures. Metamorphic hydrothermal alteration principally involves recrystallisation of most of the greenschist facies assemblage in and near ductile or brittle-ductile deformation zones. Rutile, calcite, and quartz replace titanite, and epidote is replaced by calcite and/or siderite (Table 2). Titanite and epidote are accessory minerals only, so these mineral transformations are volumetrically insignificant. Recrystallisation of muscovite and chlorite results in a new fabric in the rock, and this new fabric can be the most prominent feature of the alteration. Early stages of alteration are most apparent in rocks with ductile folds (Fig. 3A–C). Deformation is focused into narrower zones in the rock, particularly in fold axial surfaces (Fig. 3B). The pre-existing folds become tighter and may be disrupted by offset along the new fold axial surface fabric (Fig. 3B). Fine-grained micrometre-scale microshears with recrystallised muscovite and chlorite, plus scattered rutile, form dark seams through the rock (Fig. 3C). More intensely deformed zones have many closely spaced (mm–cm scale) microshears, which can dominate the rock fabric. Micaceous host rocks are dominated by these microshears, which commonly follow the pre-existing foliation. The Macraes mine (Fig. 1, 4A), is being developed in a major foliation-parallel shear zone (Hyde-Macraes Shear Zone; Craw et al. 1999a), and the hydrothermally altered and variably mineralised zone is up to 250 m thick (Fig. 4A). Shearing and gold mineralisation were focused in micaceous schist, with shears anastomosing around pods of little-mineralised or unmineralised massive feldspathic schist (Fig. 4A,B). The unmineralised pods are hydrothermally altered, however, with the same

alteration mineral transformations as in the sheared micaceous schist zones. The Glenorchy mineralised zone (Fig. 1) is dominated by normal faults at a high angle to foliation (Begbie & Sibson 2006), but alteration extends up to 2 m from the veins (Paterson 1986). The Callery occurrence (Southern Alps; Fig. 1) has irregular quartz veinlets (cm scale) with host rock alteration zones wider than the veinlets (Fig. 5A). Quartz veins, with calcite and associated silicification of host rocks, occur sporadically through the sheared and altered host rocks. Veins are typically thin (cm–m scale) and discontinuous, and many veins have been deformed within associated shear zones. Scheelite occurs in many veins, some of which have been folded during syn-mineralisation ductile deformation (Craw & Norris 1991; Petrie & Craw 2005). Silicification involves replacement of the greenschist facies assemblage with microcrystalline quartz, generally with residual albite, chlorite, and muscovite (Fig. 5B). Albite is locally replaced by hydrothermal muscovite in the Macraes shear-hosted deposit (Craw 2002), and also in the Glenorchy vein-dominated deposit (Paterson 1986), but hydrothermal albite has formed on quartz vein margins in the Invincible deposit (Hay & Craw 1993). Hence, albite may be either consumed or produced as part of this alteration (Table 2). Gold in hydrothermally altered rocks occurs primarily within or adjacent to sulfide minerals arsenopyrite and pyrite. The Callery deposit (Fig. 1, 5A) contains pyrrhotite rather than pyrite. The sulfide minerals have replaced silicates in and adjacent to microshears (Table 2; Fig 4C–E, 5B) and adjacent to quartz veins (Fig. 5A). Sulfide minerals were locally sheared as part of ongoing syn-mineralisation deformation, and sheared sulfide grains contribute to the black

16

New Zealand Journal of Geology and Geophysics, 2009, Vol. 52

Table 2 Comparison of principal mineral alteration reactions in metamor- by recrystallisation of metamorphic muscovite to more phic alteration zones and ankeritic alteration zones (modified after Craw aluminous muscovite (locally sericitic) and/or kaolinite 2002; MacKenzie & Craw 2007). (Table 2; Fig. 2). Albite is locally sericitised, but albite Metamorphic alteration Metamorphic muscovite + albite → hydrothermal muscovite (→ illite) KFe0.2Mg0.2Al2.0Si3.3O10(OH)2 + 0.2 Al3+(albite) = KFe0.05Mg0.05Al2.2Si3.3O10(OH)2) + 0.15Fe2+(pyrite, chlorite) + 0.15Mg2+(chlorite) Metamorphic chlorite → hydrothermal chlorite (negligible compositional change) Albite ↔ muscovite 3NaAlSi3O8 + K+ + 2H+ = KAl3Si3O10(OH)2 + 3Na+ + 6SiO2 (ideal formulae; alteration reaction in either direction) Epidote→ kaolinite + calcite (or siderite) 2Ca2Al3Si3O12(OH) + 4CO2 +5H2O→ 3Al2Si2O5(OH)4 + 4CaCO3 Titanite → rutile + calcite + quartz CaTiSiO5 + CO2 = TiO2 + CaCO3 + SiO2 Muscovite + chlorite + arsenate → pyrite + arsenopyrite 2FeO(silicates) +H3AsO3(aq) + 0.5H2 + 3H2S = FeS2 + FeAsS + 5H2O Ankeritic alteration Metamorphic muscovite + chlorite or albite → hydrothermal muscovite KFe0.2Mg0.2Al2.0Si3.3O10(OH)2 + 0.2 Al3+(chlorite, albite) + 0.3HCO3– = KFe0.05Mg0.05Al2.2Si3.3O10(OH)2) + 0.15FeMg(CO3)2 + 0.3H+ Metamorphic chlorite + CO2 → sideritic ankerite + kaolinite + quartz Fe2.5Mg2.5Al2Si3O10(OH)8 + 5HCO3–- +11H+ = 2.5FeMg(CO3)2 + 2Al3+(kaolinite) + 3SiO2 + 12H2O Sideritic ankerite + calcite → dolomitic ankerite FeMg(CO3)2 + 2CaCO3 = Ca2FeMg(CO3)4 Epidote→ kaolinite + calcite (or ankerite) 2Ca2Al3Si3O12(OH) + 4CO2 + 5H2O → 3Al2Si2O5(OH)4 + 4CaCO3 Titanite → rutile + calcite + quartz CaTiSiO5 + CO2 = TiO2 + CaCO3 + SiO2 chlorite + arsenate → pyrite + arsenopyrite 2FeO(chlorite) + H3AsO3(aq) + 0.5H2 + 3H2S = FeS2 + FeAsS + 5H2O

microshears that pervade the rocks (Fig. 4D,E). The Macraes deposit also has graphite in these microshears (Fig. 4C–E). Gold is typically micrometre-scale grains encapsulated within sulfide grains, but coarser (100 µm scale) gold occurs in the Callery deposit in the Southern Alps (Fig. 1), in close association with coarse (cm scale) arsenopyrite (Fig. 5A). Scattered chalcopyrite, sphalerite, and galena occur within the main sulfide minerals, although no significant enrichment of Cu, Zn, or Pb has occurred (Craw 2002). Ankeritic alteration This type of alteration is widespread and readily recognisable in the field as the secondary ankerite weathers to orangebrown iron oxyhydroxide. Ankerite forms as a vein and replacement mineral, following decomposition of chlorite in the host rock (Fig. 2). This carbonation reaction is accompanied

also occurs locally as a vein mineral with ankerite, so albite consumption or production as part of this alteration is similar to the metamorphic alteration (Table 2). The silicate alteration reactions are commonly accompanied by sulfidation reactions that precipitate pyrite and arsenopyrite in the alteration zone. Alteration is commonly accompanied by silification and quartz veins (Fig. 5C,D). These ankeritic alteration reactions in Otago Schist are described in more detail by MacKenzie et al. (2007). Ankeritic alteration is strongly controlled by structures that enhance permeability, such as foliation, microshears, faults, and fractures. Hydrothermal fluids have commonly penetrated several metres along schist foliation adjacent to faults, especially where the foliation surfaces have been enhanced by microshears. Networks of hydrothermal quartz veins, commonly containing ankerite, pyrite, and arsenopyrite, pervade brittlely deformed rocks, and ankeritic alteration affects less-deformed rocks between fractures (Fig. 5D). Larger veins contain abundant hydrothermal breccia zones, and breccia clasts are strongly altered and mineralised with sulfides (Table 1; Fig. 5C). All these textures are intermixed in the most deformed zones near to key faults that have acted as fluid conduits, and there the rock is pervaded with quartz veins, silicified host, and altered breccia clasts (Fig. 5D). The latter silicified zones are up to 5 m wide, and the whole alteration zone can be up to 100 m wide. Gold occurs encapsulated in sulfides, and as free grains in quartz veins closely associated with sulfides. Stibnite locally accompanies auriferous pyrite and arsenopyrite in veins and breccias. Ankeritic alteration is most prominent in the Southern Alps, where it occurs in broad zones associated with fold zones, faults, and fracture sets (Fig. 1). The ankeritic zones in the Otago Schist portion of the Southern Alps are focused around the Miocene Shotover gold deposits (Fig. 1), and similar alteration accompanies Miocene lamprophyre dikes nearby (Cooper et al. 1987; Craw et al. 2007). Ankeritic alteration zones then occur sporadically near to the crest of the central Southern Alps (Fig. 1) although not all occurrences have sulfides and gold. Zones of ankeritic alteration in the northern Southern Alps occur near to the Alpine Fault, rather than the mountain crest (Fig. 1), and this alteration type is particularly well developed near the junction of the Alpine and Hope Faults (Fig. 1) (Campbell et al. 2004). Scattered ankerite alteration zones occur along an active fault strand near the Hope Fault, at the southern edge of the Marlborough Fault system (Cox River; Fig. 1). The Rise & Shine Shear Zone (Fig. 1) is the earliest formed ankeritic alteration and gold mineralisation zone that is hosted in the Otago Schist. This 50 m thick mineralised shear zone formed in the latter stages of compressional uplift, and is truncated by a low-angle normal fault (Deckert et al. 2002; Cox et al. 2006; MacKenzie & Craw 2007). Other ankeritic alteration zones in the Otago Schist are smaller and more focused than the Rise & Shine zone, and are centred on discrete quartz vein swarms that resemble the Shotover swarm described above (Fig. 1, 5C,D).


17

Shallow veins This mineralisation style involves little or no alteration of host rocks adjacent to auriferous quartz veins. Vein walls are locally silicified on the centimetre scale, and minor calcite veinlets occur. Otherwise, mineralisation is confined to swarms of veins that fill extensional sites in faults, and fractures. Prismatic quartz crystals commonly protrude into open cavities in the veins (Fig. 6). Gold is generally closely associated with pyrite and arsenopyrite, which occur as coarse (mm scale) euhedral crystals and as fine-grained (micrometre scale) aggregates impregnating microcrystalline quartz. Massive and euhedral stibnite occurs in some veins, in close association with prismatic quartz crystals. Chlorite occurs as a vein mineral in some auriferous veins, and is a common accessory mineral coating quartz and calcite crystals in non-auriferous shallow veins in the Southern Alps. Fluid inclusions in many shallow veins show evidence for fluid immiscibility or boiling (Craw et al. 1987; McKeag & Craw 1989). GEOCHEMISTRY All three alteration types have distinct enrichment of arsenic and sulfur associated with gold mineralisation, with both arsenic and sulfur reaching percent levels in many altered rocks. Antimony is enriched in these rocks as well, as Sb occurs in solid solution in arsenopyrite in all alteration types, and commonly occurs as the mineral stibnite in ankeritic alteration zones and shallow veins. Metamorphic alteration zones are commonly enriched in tungsten, with or without gold enrichment. Mercury is enriched to a minor extent in the Macraes deposit, and is locally enriched in some shallow veins in the Otago Schist (MacKenzie & Craw 2005; Pitcairn et al. 2006). Mineralogical changes were minor during metamorphic alteration (Fig. 2), and the geochemical signature of the alteration, apart from the above metals, is correspondingly subtle. Whole rock geochemistry suggests that the metamorphic alteration at Macraes mine is almost isochemical (Craw 2002; Craw et al. 2007). The Macraes deposit has been locally enriched in graphite (Fig. 4C–E), so minor carbon enrichment (to 3 wt%) has occurred. Localised replacement of albite by muscovite has resulted in minor enrichment in K2O and depletion in Na2O, but these effects are smaller than typical host rock variations in these elements at the hand specimen scale. Likewise, silicification in metamorphic alteration zones has increased silica concentrations in some rocks and diluted other less mobile elements such as Al and Ti (Fig. 7A), but this is commonly at similar or lower levels than the inherent variations in host rock silica contents. Mobility, and possible enrichment, of Cr has apparently occurred in the Macraes and Invincible deposits (Fig. 7B) (Hay & Craw 1993; Craw 2002). This Cr mobility has resulted in deviation of altered rocks to higher Cr contents compared to typical background schist Cr-Ni relationships in which these elements are strongly correlated (Fig. 7B). Paterson (1986) noted a distinct decrease in Sr in the Glenorchy alteration zone compared to the host rocks. Ankeritic alteration was also largely isochemical with respect to the rock-forming elements. The main chemical alteration feature is the addition of CO2 to form ankerite (Table 2), as indicated by higher loss-on-ignition (LOI) (Fig. 7C). Unaltered rocks have MgO-LOI relationships reflecting their

Fig. 3 Late metamorphic fold zones in upper greenschist facies Otago Schist, showing progressive development of lower greenschist facies shear zones and associated metamorphic alteration. A, Ductile folds with shallow-dipping fold axial surfaces (FAS). B, Sheared and recrystallised FAS zones in ductile folds. C, Microscopic view (plane polarised light) of FAS shears, with dark shears containing alteration rutile, chlorite, and muscovite anastomosing around metamorphic quartz (Q, white).

muscovite and chlorite contents, as these two minerals are the dominant Mg and H2O bearing phases (Fig. 7C). Many of the rocks from ankeritic alteration zones spread towards higher LOI (Fig. 7B). Some samples with even higher LOI than predicted by the presence of abundant ankerite replacing chlorite (Fig. 7C) contain syn-mineralisation and late stage calcite as well. Silicification of many of these altered rocks has caused dilution of other elements, including both MgO and LOI (Fig. 7C). Ankeritic alteration zones in Shotover deposits have elevated Sr and depleted Ba compared to immediate host

18

New Zealand Journal of Geology and Geophysics, 2009, Vol. 52 Fig. 4 Metamorphic alteration at Macraes mine, Otago Schist (Fig. 1). A, Photograph of the Frasers open pit, showing the scale of metamorphic alteration around the shear zone host. All visible rocks have been altered to some extent. Bench height is 10 m. White lines show foliation; dotted white lines enclose a prominent mineralised shear zone; dashed white lines outline pods of altered but largely unmineralised feldspathic schist pods. B, Outcrop view of the mineralised shear zone, with variably mineralised micaceous schist and well-mineralised black shears (graphitic) anastomosing around pods of massive feldspathic schist (white). C, Microphotograph (plane polarised light) of incipient metamorphic alteration of micaceous schist. Original greenschist facies foliation and segregation (horizontal fabric) has minor parallel development of graphitic microshears. D, Microphotograph (plane polarised light) of moderately mineralised altered rocks, with euhedral sulfides (black) replacing micas, and black graphitic microshears. E, Microphotograph (plane polarised light) of strongly mineralised altered rocks, with microshears disrupting the metamorphic foliation.

rocks (MacKenzie et al. 2007), but these differences were not observed in ankeritic alteration zones at the Rise & Shine deposit (MacKenzie & Craw 2007). Shallow veins have negligible alteration zones associated with them, and the little that is present is dominated by silicification. Silicification is focused on breccia fragments, with little of this alteration extending into intact host rock. Hence, there is no distinctive geochemical signature associated with this mineralisation style. CARBON AND OXYGEN STABLE ISOTOPES Greenschist facies metamorphic rocks that form the host rocks for most of the vein systems in this study have calcite as a metamorphic mineral (Fig. 2). This calcite, which can

form up to 5% of the host rocks, occurs in syn-metamorphic veins, and as scattered grains through the matrix of the rocks. Greenschist facies calcite has a well-defined and restricted range of carbon and oxygen isotopic ratios (Fig. 8A) (Blattner & Cooper 1974). Relatively low δ13C values in calcite in some greenschist facies host rocks (Fig. 8A) reflects coprecipitation of graphite (Fig. 2) from metamorphism of organic matter during prograde metamorphism (Pitcairn et al. 2005). Mineralised shallow veins in the Otago Schist also have a relatively restricted range of oxygen and carbon isotopic ratios (Fig. 8A). The isotopic signature of these mineralised veins arises from remobilisation and recrystallisation of host schist calcite into the hydrothermal fluid, with deposition occurring at a lower temperature than the original metamorphic calcite (Craw 1992). Similar late-stage shallow veins through the Southern Alps have calcite isotopic ratios


19

Fig. 5 Outcrop photographs of altered rocks. A, Metamorphic alteration at the Callery gold occurrence, Southern Alps. Dark micaceous schist (M) and grey quartzofeldspathic schist are truncated by thin quartz-calcite veinlets (QC). Coarse-grained euhedral arsenopyrite (A) replaces silicates in the host schist. B, Schist at the Macraes mine has been altered and silicified, but original ductile folds are still preserved. Auriferous pyrite (P) replaces silicates, especially in micaceous segregations (M). C, Auriferous ankeritic rock in the Shotover area, showing quartz-ankerite vein (QA) with pyrite (P), late stage quartz veins (Q), silicified schist (QS), and ankeritic alteration in host schist (AS). D, Ankeritic host rock to the vein in C, with brecciated and altered schist cemented with quartz veins. Fig. 6 Outcrop photograph of a shallow vein deposit, Nenthorn, Otago Schist. Host schist is hydrothermally unaltered, but has been brecciated and cemented with quartz veins containing prismatic quartz extending into open cavities.

that reflect the same remobilisation of host calcite, with a spread of data to even lower temperatures of recrystallisation (Jenkin et al. 1994; Koons et al. 1998). In contrast, calcite in the metamorphic alteration zone at Macraes mine shows a distinct trend in isotopic ratios towards low δ13C values (Fig. 8A) (Craw 1992; deRonde et al. 2000). These low δ13C values reflect the influence of co-precipitating graphite in the Macraes deposit (Craw 2002). Shotover ankeritic alteration zones contain the best exposures of fresh rock in economically significant gold deposits, so these were examined in detail for variations in carbonate isotopic ratios (Fig. 8B; Table 3). Greenschist facies prograde metamorphic calcites from unaltered Shotover host rocks (Fig. 8B) have similar range of carbon and oxygen isotopic ratios to the regional compilation (Fig. 8A), including the trend towards low δ13C values that accompanies prograde graphite


20 1.0

4

A

0.6

δ13C(PDB), per mil

TiO2, wt%

Shallow Au veins in faults, Otago

0

0.8

silicification

0.4

Shallow hydrothermal calcite, S. Alps

-4

-8

-12

5

Shotover Macraes

4

10 Al2O3, wt%

15

0

B

40 35

Macraes

regional

15

20

25

30

25

30

SHOTOVER AU DEPOSIT Late hydrothermal calcite

B

20

δ13C(PDB), per mil

45

5

10

δ18O(SMOW), per mil

0.0 0

Macraes late metamorphic Au shear

Metamorphic calcite

-16

0.2

30

Ni, ppm

CALCITE

A

Host schist

calcite

-4

Mineralised ankerite

-8 Metamorphic calcite

-12

25 -16

20

5

alteration

15

5

4

100

C

150 Cr, ppm

200

250

50

100

Cox R 100

-8

200

from rockexchanged water

300

Central S. Alps

chlorite -16

carbonation

5

10

15

20

25

30

δ18O(SMOW), per mil

silicification

2

ankerite

1 Rise & Shine Shotover

0 0

from meteoric water

-4

-12

muscovite

3

300

δ13C(PDB), per mil

0

MgO, wt%

ANKERITE Alpine/Hope intersection

0

50

20

C

Invincible

0

15

δ18O(SMOW), per mil 4

10

10

2

4 LOI, wt%

6

8

Fig. 8 Oxygen and carbon stable isotopic data for host schist, and for veins and altered rocks. Data from Table 3 and sources referenced in text. A, Calcite in Otago Schist (black ellipse), shallow auriferous veins in Otago (dotted ellipse), Macraes metamorphic alteration (black squares), and shallow veins in the Southern Alps (open circles). B, Carbonates associated with ankeritc alteration, Shotover, Otago Schist. Calcite in unaltered host rocks (black diamonds), calcite accompanying ankerite from altered rocks (open squares), alteration ankerite (black squares), post-alteration calcite (open circles). C, Ankerite from the Southern Alps. Central Southern Alps (black circles), Cox River (black triangles), intersection of Alpine and Hope Faults (open circles). Model curves show ankerite isotopic compositions deposited from meteoric and rock-exchanged fluids at a range of temperatures (see text).

Fig. 7 Geochemical plots showing some of the most distinctive chemical features of alteration (see text for discussion). A, Silicification causes dilution of relatively immobile Al2O3 and TiO2 at Macraes (metamorphic alteration) and Shotover (ankeritic alteration), from the host schist composition (ellipse, top right). B, Metamorphic alteration effects on Cr and Ni at Macraes (black squares) and Invincible (open diamonds) deposits, Otago Schist. Background schist (small black dots) has a distinctive relationship between Cr and Ni. Alteration causes Cr mobilisation and localised enrichment. C, Ankeritic alteration effects on loss-on-ignition (LOI) and MgO. Addition of ankerite during the alteration causes increase in LOI at constant MgO, but silicification can dilute these effects as indicated.

21

Craw et al.—Hydrothermal alteration in orogenic gold deposits, NZ Table 3 Stable isotopic data for carbonates from Shotover and Cox River alteration zones. Location

Mineral

Generation

δ18O (‰) VSMOW

Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover

calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite

late late late late late late late mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation mineralisation

13.12 14.06 11.23 11.98 12.38 12.45 9.86 15.55 15.86 15.54 13.93 15.33 15.38 16.29 14.46 14.86 16.14 19.46 21.70 19.03 20.09 21.54 18.09 21.81 21.12 22.65 22.53 21.62 19.04 20.81 20.89 20.98 19.81 21.42 22.63 22.35 22.06 22.72 21.44 22.96 21.72 21.11 21.79 21.04 21.38 20.72

δ13C (‰) VPDB

Location

Mineral

Generation

δ18O (‰) VSMOW

–5.76 –8.01 –2.62 –2.76 –3.36 –5.65 1.95 –4.17 –3.89 –4.32 –4.66 –3.78 –3.24 –3.86 –3.98 –3.63 –3.78 –9.98 –4.33 –8.60 –2.02 1.05 –9.96 –5.91 –9.78 –4.35 –7.03 –1.98 –1.44 –2.34 –3.01 2.92 –1.65 –3.99 –2.77 –2.60 –1.60 –2.97 –3.54 –3.03 –3.33 –3.70 –3.68 –7.80 –9.96 –9.86

Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Shotover Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R Cox R

ankerite ankerite calcite calcite calcite calcite calcite calcite calcite calcite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite ankerite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite calcite

mineralisation mineralisation metamorphic metamorphic metamorphic metamorphic metamorphic metamorphic metamorphic metamorphic alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration alteration

21.71 22.30 12.03 12.02 10.49 11.52 12.86 11.98 11.88 12.05 18.69 21.68 22.97 19.91 17.67 18.88 17.32 17.12 17.51 19.54 17.56 19.31 18.91 18.72 17.19 17.28 23.45 20.46 17.97 17.35 15.34 17.54 16.84 16.94 19.36 17.27 18.89 20.17 17.51 19.16 12.32 16.10 14.74 18.64 17.71 18.06

formation. Likewise, the relatively rare syn-mineralisation calcites that accompany ankerite have similar isotopic values (Fig. 8B) to the shallow mineralised veins elsewhere in Otago Schist (Fig. 8A). Late-stage hydrothermal calcites in the Shotover alteration zones have isotopic signatures broadly similar to the host rocks (Fig. 8B). Hydrothermal ankerites have a broad range of δ13C values and relatively narrow range of δ18O values, with δ18O values distinctly higher than co-existing calcite (Fig. 8B). Ankerites from veins in mineralised zones along the active Southern Alps hydrothermal system have a more restricted range of δ13C values and wider range of δ18O values than ankerites in the Shotover deposits (Fig. 8C). Ankerites from the central Southern Alps (Fig. 8C) have isotopic signatures similar to the host schist calcites (Fig. 8A,B). Ankerites in the Marlborough Fault System at Cox River (Fig. 1) have

δ13C (‰) VPDB –6.29 –6.53 –10.13 –5.16 –5.37 –2.50 –14.20 –10.55 –11.98 –12.36 –2.28 –3.59 –2.51 –2.44 –2.25 –2.71 –2.25 –2.17 –3.01 –3.18 –1.86 –3.17 –2.57 –3.29 –2.03 –1.47 –2.64 –1.74 –1.43 –2.95 –3.58 –2.35 –2.71 –3.11 –2.90 –2.52 –2.76 –2.77 –3.31 –1.78 –5.34 –2.87 –1.62 –3.29 –3.36 –3.39

distinctly higher δ18O values than those of the central Southern Alps (Fig. 8C; Table 3), with a similar range to those of the Shotover deposits (Fig. 8B). Hydrothermal ankerites near the intersection of Hope and Alpine Faults, spatially between the central Southern Alps and Cox River (Fig. 1), have δ18O values between –10 and –25‰, spanning the range from central Southern Alps to Cox River data (Fig. 8C; Table 3). The observed ranges of ankerite isotopic data have been interpreted to represent formation from different proportions of meteoric and crustally exchanged fluids, at a range of temperatures (Campbell et al. 2004). Model ankerite compositions resulting from deposition from observed and inferred fluid compositions at a range of plausible temperatures are shown with lines in Fig. 8C (after Upton et al. 2003; Campbell et al. 2004). The observed data fall between these model curves (Fig. 8C).

22

New Zealand Journal of Geology and Geophysics, 2009, Vol. 52 Fig. 9 Schematic cross-section (A, not strictly to scale) of a collisional orogen, based on the active Southern Alps, showing the relative distribution of the three principal alteration zones described in this paper. The characteristic features of goldbearing structures in the three zones, as numbered in the cross-section, are indicated in sketches B, C, and D.

DISCUSSION Orogenic hydrothermal systems The active Southern Alps hydrothermal system is driven by collisional tectonic activity at a major plate boundary (Koons et al. 1998; Craw et al. 2002). We suggest that this is an analogue for processes that have occurred in ancient collisional mountain belts, and that similar hydrothermal alteration zones resulted from this type of hydrothermal activity in ancient belts. The general shapes and spatial relationships of these alteration zones are summarised in Fig. 9A. This model system is based on the Southern Alps and the known fluid saturation zones identified in this orogen by Wannamaker et al. (2002) with magnetotelluric methods, combined with observations of exhumed hydrothermal systems (Craw et al. 1987; Jenkin et al. 1994; Cox et al. 1997). The volumetrically largest portion of this hydrothermal system occurs in the middle crust, where ductile and brittle/

ductile compressional deformation is occurring, and this has resulted in a zone of connected fluid in these low porosity, low permeability rocks (Wannamaker et al. 2002). These fluids have had extensive chemical interaction with the host rocks and so their isotopic signatures are now indicative of at least partial equilibration with their greenschist facies metamorphic hosts (Jenkin et al. 1994; Koons et al. 1998; Campbell et al. 2004). The source of the fluids is unknown, but it is likely to be deep-circulating meteoric water and/or dehydration fluids from the underlying hot metamorphic pile (Jenkin et al. 1994; Upton et al. 1995; Koons et al. 1998; Campbell et al. 2004; Pitcairn et al. 2006). A component of magmatic water has also been suggested for the Macraes deposit in the Otago Schist (de Ronde et al. 2000). Irrespective of the ultimate source of the fluid, gold mobility is an inevitable consequence of this middle crustal hydrothermal activity. Gold was derived, with other metals and metalloids, from the host rock with at least some dehydration fluids (Pitcairn et al. 2006).

Craw et al.—Hydrothermal alteration in orogenic gold deposits, NZ The alteration zones described above develop within the collisional orogen as the middle crustal fluids migrate through the orogen (Fig. 9A). The deepest portion of the hydrothermal system accompanies greenschist facies metamorphism (Fig. 9A), and rising fluids are only marginally out of chemical equilibrium with the enclosing rocks. Fluid flow at these depths is focused by compressional structures such as microshears in folds and shear zones (Fig. 9B), and pervasive fluid flow occurs along grain boundaries adjacent to these structures (cf. Fig. 3A–C, 4A–E). The alteration reactions that result from this process are essentially the same as retrogressive metamorphism in the lower greenschist facies (Craw 2002). Fluid flow under these near-metamorphic conditions is slow, probably millimetres per year (Upton 1998; Craw 2002). The available source of gold, via leaching from the host rocks (Pitcairn et al. 2006), is extremely large, and formation of gold deposits depends primarily on physical and/ or chemical factors that can cause focused gold deposition from the fluid. The exact nature of such depositional factors is not known, but hydrothermal graphite deposition may be one such factor (Craw 2002; Upton & Craw 2008). The middle crustal fluid rises through the orogen and can penetrate to shallow levels, as detected by Wannamaker et al. (2002) and Upton et al. (2003). In addition, topographically driven meteoric water can penetrate to at least 6 km depth within the collisional orogen (Jenkin et al. 1994; Koons et al. 1998) and possibly considerably deeper (Upton et al. 1995). Consequently, the upper crustal zone beneath the mountain belt is a mixing zone for the middle crustal fluid and downward migrating meteoric fluid, and the hydrothermal system involves both fluids in varying proportions (Fig. 8C) (Campbell et al. 2004). This upper crustal portion of the hydrothermal system has fluids that are distinctly out of chemical equilibrium with the host rocks, and ankeritic hydrothermal alteration results from water-rock interaction (Fig. 9A). This alteration is dominated by decomposition of host rock chlorite and replacement by ankerite, using CO2 from the fluid for this carbonation process. Fluid flow is controlled primarily by faults and fractures (Campbell et al. 2004), with some penetration of fluid beyond these structures into the host rock (Fig. 9C). This alteration style occurs at a wide range of depths, as variably exposed in the Southern Alps (Fig. 1, 9A). Ankeritic alteration penetrates almost to the surface along the main mountain crest of the central Southern Alps (Cox et al. 1997) and along active faults of the Marlborough Fault System, especially at the intersection of Hope and Alpine Faults (Campbell et al. 2004), and Cox River area (Fig. 1). Shallow veins form in extensional sites in the near-surface region (Fig. 9A,D). Many of these veins form from mainly meteoric water, although a subordinate component of rockexchanged water may also be involved (Jenkin et al. 1994; Cox et al. 1997; Koons et al. 1998). Fluid flow has occurred in open fractures, and at least some of this fluid flow was rapid and probably episodic, as indicated by fluid inclusion evidence that fluid temperature was greater than rock temperature (Craw 1997). This rapid fluid flow is one reason why host rock alteration is negligible in this portion of the hydrothermal system. Gold mineralisation has occurred in few of the observed shallow veins in the Southern Alps. Likewise, shallow-formed quartz veins without gold are common throughout the Otago Schist, although relative timing of these Otago Schist veins is more poorly constrained than in the active Southern Alps.

23

Space-time relationships of mineralised structures The most notable feature of the hydrothermal alteration system depicted in Fig. 9A is the large scale of the hydrothermal zonation. The whole alteration system is up to 15 km high and >20 km wide in places. This very large scale, combined with the mineralogical subtlety of the metamorphic alteration, means that the metamorphic alteration is extremely difficult to detect in a greenschist facies metamorphic belt. Without good fresh outcrop and obvious mineral indicators such as arsenopyrite and graphite (Fig. 4A–E), this style of alteration is largely invisible. Structural indicators, such as those depicted in Fig. 3B,C, provide other clues, but these are subtle as well. In contrast, ankeritic alteration is more readily recognised because of the distinctive mineralogy, so that regional-scale mapping of zones containing this alteration type is possible (Fig. 1). Shallow veins have strong structural control, but prediction of their occurrence is difficult, and they occur irregularly throughout both the ancient (Otago Schist) and modern (Southern Alps) orogens in this study (Fig. 1). The spatial relationships between the alteration zones depicted in Fig. 9A are broadly predictable because of the structural and chemical nature of fluid flow and fluid-rock interaction that occurs at different crustal levels (Fig. 9B–D). Penetration of ankeritic alteration process to near-surface regions (above) is the principal complicating factor in defining spatial zonation of alteration. The alteration zonation outlined in Fig. 9A presents a snapshot in time of these alteration zones. While the relative positions of the alteration zones in the orogen remain approximately constant, the host rocks are constantly moving, mainly laterally, through the orogen in the manner of a conveyor belt at plate tectonic rates (mm–cm/ yr). New un-dehydrated rocks are constantly added to the metamorphic zone in the middle crust (Fig. 9A, from right), dehydrated rocks are extracted by uplift and erosion (Fig. 9A, left), and topographically driven meteoric water continuously migrates downwards. All alteration styles are formed at the same time in this large hydrothermal system. Preservation, or lack of preservation, of altered and mineralised rocks within the orogen is a result of subsequent uplift and deformation activity. Hence, all alteration and mineralisation zones are syn-orogenic. The model of mineralisation processes and alteration zonation described above is applicable to other mountain belts, ancient and modern. The styles of mineralisation and alteration listed in Table 1 occur in many well-established goldfields worldwide, and broadly similar depth sequences have been proposed by earlier workers without the dynamic component displayed by the Southern Alps (Hagemann et al. 1994; Groves et al. 1998; Goldfarb et al. 2005). Taiwan is a similar active collisional belt to the Southern Alps, and similar gold mineralisation is occurring there (Tan et al. 1991). The Himalaya mountains have many similarities with the Southern Alps, but large amounts of dry crystalline rocks in the most active hydrothermal zones apparently preclude active gold mineralisation of the style described herein (Craw et al. 2002). CONCLUSIONS Collisional orogenesis caused large scale (>10 km) hydrothermal systems to form in ancient and modern mountain belts of southern New Zealand. This hydrothermal activity

24


resulted in hydrothermal alteration zones within the orogen, locally accompanied by orogenic gold mineralisation. There are three distinct alteration zones formed within these orogens (Fig. 9A), and these alteration zones all form at the same time but under different physicochemical conditions (Table 1). The volumetrically largest alteration zone occurs above a large (10 km scale) middle crustal zone of deformation. This metamorphic alteration occurs under lower greenschist facies conditions as greenschist facies rocks are being uplifted through the brittle-ductile transition, and the fluid is only marginally out of equilbrium with the host rocks. Alteration and mineralisation is focused in shear zones where slow-moving fluid (mm/yr) is controlled by microfractures and grain boundary permeability. This style of alteration accompanies disseminated gold mineralisation. Rising fluids become distinctly out of equilbrium with the host rocks, and ankeritic alteration results. This alteration style involves carbonation of the rock with CO2 from the fluid, via replacement of chlorite with ankerite. Ankeritic alteration is controlled by faults and fractures, but abundant water-rock interaction also occurs in adjacent host rock and wide alteration zones (10–100 m scale) can develop. This alteration style is the most readily observed in the field, and areas with abundant structures hosting this alteration are mappable over tens of square kilometres in the active Southern Alps orogen. Ankeritic alteration is best developed in zones where extensive faulting and fracturing of rocks has occurred, and the alteration extends almost to the surface near major active fault zones. Ankeritic alteration is dominated by rock-exchanged fluids from deeper in the orogen, with variable proportions of meteoric fluid also making a contribution. Gold mineralisation occurs in many ankeritic alteration zones. Shallow veins form in swarms of extensional fractures at shallow levels in an orogen. The veins are dominated by quartz and calcite with minor chlorite. Some of these veins contain sulfides and gold. Host rock alteration is minimal, with minor silicification on the centimetre scale. Meteoric and rock-exchanged fluids contribute to formation of this vein type. Fluid flow in these fracture systems is rapid and episodic compared to the other two alteration styles, and fluid temperature is commonly greater than rock temperature. The three alteration styles form over a vertical distance of up to 15 km, and some overlap occurs in alteration styles. The alteration zones form above the main dehydration zone in the collisional orogen, where new rocks are constantly being added at plate tectonic rates while dehydrated rocks are exhumed and removed from the orogen by erosion. Hence, while the orogenic position of alteration zones remains approximately fixed, the host rocks are moving at several centimetres a year through the orogen. All alteration styles and their associated gold-bearing veins are syn-orogenic.

REFERENCES Al-Aasm IS, Taylor BE, South B 1990. Stable isotope analysis of multiple carbonate samples using selective acid digestion. Chemical Geology 80: 119–125. Begbie MJ, Sibson RH 2006. Structural controls on the development of fault-hosted scheelite-gold mineralization in Glenorchy, northwest Otago. In: Christie AB, Brathwaite RL ed. Geology and exploration of New Zealand mineral deposits. Australasian Institute of Mining and Metallurgy Monograph 25: 291–298. Bierlein FP, Crowe D 2000. Phanerozoic orogenic lode gold deposits. In: Hagemann SG, Brown PE ed. Gold in 2000. Reviews in Economic Geology 13: 103–140. Bierlein FP, Maher S 2001. Orogenic disseminated gold in Phanerozoic fold belts: examples from Victoria, Australia, and elsewhere. Ore Geology Reviews 18: 113–148. Blattner P, Cooper AF 1974. Carbon and oxygen isotopic composition of carbonatite dikes and metamorphic country rock of the Haast Schist terrain, New Zealand. Contributions to Mineralogy and Petrology 104: 332–347. Bohlke JK 1988. Carbonate-sulfide equilibria and “stratabound” disseminated epigenetic gold mineralization: a proposal based on examples from Alleghany, California, USA. Applied Geochemistry 3: 499–516. Campbell JR, Craw D, Frew R, Horton T, Chamberlain CP 2004. Geochemical signature of orogenic hydrothermal activity in an active tectonic intersection zone, Alpine Fault, New Zealand. Mineralium Deposita 39: 437–451. Clark ME, Carmichael DM, Hodgson CJ, Fu M 1989. Wall-rock alteration, Victory gold mine, Kambalda, Western Australia: processes and P-T-CCO2 conditions of metasomatism. Economic Geology Monograph 6: 445–459. Cooper AF, Barreiro BA, Kimbrough DL, Mattinson JM 1987. Lamprophyre dike intrusion and the age of the Alpine fault, New Zealand. Geology 15: 941–944. Cox SF, Wall VJ, Etheridge MA, Potter TF 1991. Deformation and metamorphic processes in the formation of mesothermal vein-hosted gold deposits: examples from the Lachlan Fold Belt in central Victoria, Australia. Ore Geology Reviews 6: 391–423. Cox SC, Craw D, Chamberlain CP 1997. Structure and fluid migration in a late Cenozoic duplex system forming the Main Divide in the central Southern Alps, New Zealand. New Zealand Journal of Geology and Geophysics 40: 359–373. Cox L, MacKenzie DJ, Craw D, Norris RJ, Frew R 2006. Structure and geochemistry of the Rise & Shine Shear Zone mesothermal gold system, Otago Schist, New Zealand. New Zealand Journal of Geology and Geophysics 49: 429–442. Craw D 1992. Fluid evolution, fluid immiscibility and gold deposition during Cretaceous–Recent tectonics and uplift of the Otago and Alpine Schist, New Zealand. Chemical Geology 98: 221–236. Craw D 1997. Fluid inclusion evidence for geothermal structure beneath the Southern Alps, New Zealand. New Zealand Journal of Geology and Geophysics 40: 43–52.

ACKNOWLEDGMENTS This research was funded by the New Zealand Foundation for Research, Science and Technology, and the University of Otago. Discussions with Richard Norris and Peter Koons were valuable in developing ideas expressed herein. John Williams assisted with fieldwork, and Damian Walls assisted with laboratory work.

Craw D 2002. Geochemistry of late metamorphic hydrothermal alteration and graphitisation of host rock, Macraes gold mine, Otago Schist, New Zealand. Chemical Geology 191: 257–275. Craw D, Norris RJ 1991. Metamorphogenic Au-W veins and regional tectonics: mineralisation throughout the uplift history of the Haast Schist, New Zealand. New Zealand Journal of Geology and Geophysics 34: 373–383.

Craw et al.—Hydrothermal alteration in orogenic gold deposits, NZ Craw D, Rattenbury MS, Johnstone RD 1987. Structural geology and vein mineralisation in the Callery River headwaters, Southern Alps, New Zealand. New Zealand Journal of Geology and Geophysics 30: 273–286. Craw D, Windle SJ, Angus PV 1999a. Gold mineralization without quartz veins in a ductile-brittle shear zone, Macraes Mine, Otago Schist, New Zealand. Mineralium Deposita 34: 382–394. Craw D, Youngson JH, Koons PO 1999b. Gold dispersal and placer formation in an active oblique collisional mountain belt, Southern Alps, New Zealand. In: Minter WEL, Craw D ed. Special Issue on placer deposits. Economic Geology 94: 605–614. Craw D, Koons PO, Horton T, Chamberlain CP 2002. Tectonically driven fluid flow and gold mineralisation in active collisional orogenic belts: comparison between New Zealand and western Himalaya. In: Labaume, P, Craw D, Lespinasse M, Muchez P ed. Tectonic processes and flow of mineralizing fluids. Tectonophysics 348: 135–153. Craw D, Begbie M, MacKenzie D 2006. Structural controls on Tertiary orogenic gold mineralization during initiation of a mountain belt, New Zealand. Mineralium Deposita 41: 645–659. Craw D, MacKenzie DJ, Pitcairn IK, Teagle DAH, Norris RJ 2007. Geochemical signatures of mesothermal Au-mineralised late-metamorphic deformation zones, Otago Schist, New Zealand. Geochemistry: Exploration, Environment, Analysis 7: 225–232. Craw D, Burridge CP, Upton P, Rowe DL, Waters JM 2008. Evolution of biological dispersal corridors through a tectonically active mountain range in New Zealand. Journal of Biogeography 35: 1790–1802. Deckert H, Ring U, Mortimer N 2002. Tectonic significance of Cretaceous bivergent extensional shear zones in the Torlesse accretionary wedge, central Otago Schist, New Zealand. New Zealand Journal of Geology and Geophysics 45: 537–547. DeMets C, Gordon RG, Argus DF, Stein S 1994. Effect of recent revisions to the geomagnetic reversal time scale on estimates of current plate motions. Geophysical Research Letters 21: 2191–2194. de Ronde CEJ, Faure K, Bray CJ, Whitford DJ 2000. Round Hill shear zone-hosted gold deposit, Macraes Flat, Otago, New Zealand: evidence of a magmatic ore fluid. Economic Geology 95: 1025–1048. Eilu P, Mikucki EJ 1998. Alteration and primary geochemical dispersion associated with the Bulletin lode-gold deposit, Wiluna, Western Australia. Journal of Geochemical Exploration 63: 73–103. Gao ZL, Kwak TAP 1997. The geochemistry of wall rock alteration in turbidite-hosted gold vein deposits, central Victoria, Australia. Journal of Geochemical Exploration 59: 259–274. Goldfarb RJ, Baker T, Dube B, Groves DI, Hart CJ, Gosselin P 2005. Distribution, character and genesis of gold deposits in metamorphic terranes. In: Hedenquist JW, Thompson JFH, Goldfarb RJ, Richards JP ed. Economic Geology 100th Anniversary Volume. Pp. 407–450. Gray DR, Foster DA 2004. 40Ar/39Ar thermochronologic constraints on deformation, metamorphism and cooling/exhumation of a Mesozoic accretionary wedge, Otago Schist, New Zealand. Tectonophysics 385: 181–210. Groves DI, Goldfarb RJ, Gebre-Mariam M, Hagemann SG, Robert F 1998. Orogenic gold deposits: a proposed classification in the context of their crustal distribution and relationship to other gold deposit types. Ore Geology Reviews 13: 7–27.

25

Hagemann SG, Gebre-Mariam G, Groves DI 1994. Surface-water influx in shallow-level Archean lode-gold deposits in Western Australia. Geology 22: 1067–1070. Hay R, Craw D 1993. Syn-metamorphic gold mineralisation, Invincible Vein, NW Otago Schist, New Zealand. Mineralium Deposita 28: 90–98. Jenkin GRT, Craw D, Fallick AE 1994. Stable isotopic and fluid inclusion evidence for meteoric fluid penetration into an active mountain belt; Alpine Schist, New Zealand. Journal of Metamorphic Geology 12: 429–444. Kishida A, Kerrich R 1987. Hydrothermal alteration zoning and gold concentration at the Kerr-Addison Archean lode gold deposit, Kirkland Lake, Ontario. Economic Geology 82: 649–690. Koons PO, Craw D, Cox SC, Upton P, Templeton AS, Chamberlain CP 1998. Fluid flow during active oblique convergence: a Southern Alps model from mechanical and geochemical observations. Geology 26: 159–162. Landis CA, Campbell HJ, Begg JG, Mildenhall DC, Paterson AM, Trewick SA 2008. The Waipounamu Erosion Surface: questioning the antiquity of New Zealand land surface and terrestrial fauna and flora. Geological Magazine 145: 173–197. McCuaig TC, Kerrich R 1998. P-T-t-deformation-fluid characteristics of lode gold deposits: evidence from alteration systematics. Ore Geology Reviews 12: 381–453. McKeag SA, Craw D 1989. Contrasting fluids in gold-quartz vein systems formed progressively in a rising metamorphic belt: Otago Schist, New Zealand. Economic Geology 84: 22–33. MacKenzie DJ, Craw D 2005. The mercury and silver contents of gold in quartz vein deposits, Otago Schist, New Zealand. New Zealand Journal of Geology and Geophysics 48: 265–278. MacKenzie DJ, Craw D 2007. Contrasting hydrothermal alteration mineralogy and geochemistry in the auriferous Rise & Shine Shear Zone, Otago, New Zealand. New Zealand Journal of Geology and Geophysics 50: 67–79. MacKenzie DJ, Craw D, Begbie M 2007. Mineralogy, geochemistry, and structural controls of a disseminated gold-bearing alteration halo around the schist-hosted Bullendale orogenic gold deposit, New Zealand. Journal of Geochemical Exploration 93: 160–176. Mitchell M, Maw L, Angus PV, Craw D 2006. The Macraes gold deposit in east Otago. In: Christie AB, Brathwaite RL ed. Geology and exploration of New Zealand mineral deposits. Australasian Institute of Mining and Metallurgy Monograph 25: 313–318. Mortimer N 1993. Geological map of the Otago Schist and adjacent rocks, scale 1:500,000. Institute of Geological & Nuclear Sciences Geological Map 7. Lower Hutt, New Zealand, Institute of Geological & Nuclear Sciences Ltd. Paterson CJ 1986. Controls on gold and tungsten mineralization in metamorphic-hydrothermal systems, Otago, New Zealand. Geological Association of Canada Special Paper 32: 25–39. Petrie BS, Craw D 2005. Lithological controls on structural evolution of mineralised schist, Macraes gold mine, Otago, New Zealand. New Zealand Journal of Geology and Geophysics 48: 435–446. Pitcairn IK, Roberts S, Teagle DAH, Craw D 2005. Detecting hydrothermal graphite deposition during metamorphism and gold mineralisation. Journal of the Geological Society London 162: 429–432.

26


Pitcairn IK, Teagle DAH, Craw D, Olivo GR, Kerrich R, Brewer TS 2006. Sources of metals and fluids in orogenic gold deposits: insights from the Otago and Alpine Schists, New Zealand. Economic Geology 101: 1525–1546. Rosenbaum J, Sheppard SMF 1986. An isotope study of siderites, dolomites and ankerites at high temperature. Geochimica et Cosmochimica Acta 50: 1147–1150. Simpson GD, Cooper AF, Norris RJ 1994. Late Quaternary evolution of the Alpine Fault Zone at Paringa, South Westland, New Zealand. New Zealand Journal of Geology and Geophysics 37: 49–58. Tan LP, Chen CH, Yeh TK, Takeuchi A 1991. Tectonic and geochemical characteristics of Pleistocene gold deposits in Taiwan. In: Jones M, Cosgrove J ed. Neotectonics and resources. London, Bellhaven Press. Pp. 290–305. Upton P 1998. Modelling localisation of deformation and fluid flow in a compressional orogen: implications for the Southern Alps of New Zealand. American Journal of Science 298: 296–323. Upton P, Craw D 2008. Modelling the role of graphite in development of a mineralised mid-crustal shear zone, Macraes mine, New Zealand. Earth and Planetary Science Letters 266: 245–255.

Upton P, Koons PO, Chamberlain CP 1995. Penetration of deformation-driven meteoric water into ductile rocks: isotopic and model observations from the Southern Alps, New Zealand. New Zealand Journal of Geology and Geophysics 38: 535–543. Upton P, Craw D, Caldwell TG, Koons PO, James Z, Wannamaker PE, Jiracek GJ, Chamberlain CP 2003. Upper crustal fluid flow in the outboard region of the Southern Alps, New Zealand. Geofluids 3: 1–12. Wannamaker PE, Jiracek GR, Stodt JA, Caldwell TG, Gonzalez VM, McKnight JD, Porter AD 2002. Fluid generation and pathways beneath an active compressional orogen, the New Zealand Southern Alps, inferred from magnetotelluric data. Journal of Geophysical Research 107(B6). Wellman H 1979. An uplift map for the South Island of New Zealand, and a model for uplift of the Souhern Alps. In: Walcott RI, Cresswell MM ed. The origin of the Southern Alps. Royal Society of New Zealand Bulletin 18: 13–20. Witt WK, Vanderhor F 1998. Diversity within a unified model for Archean gold mineralization in the Yilgarn Craton of Western Australia: an overview of the late-orogenic structurally controlled gold deposits. Ore Geology Reviews 13: 29–64.