2010 angerer&hagemann_ structure kooly

7/27/2019 2010 Angerer&Hagemann_ Structure Kooly

http://slidepdf.com/reader/full/2010-angererhagemann-structure-kooly 1/29

0361-0128/10/3905/917-29 917

Introduction

THE STRUCTURAL control of high-grade iron ore (58−68 wt %Fe) hosted in banded iron formation (BIF) is considered to beone of the most important factors that influence the locationand geometry of iron ore deposits (Dalstra and Rosière, 2008,and references therein). However, the details of ore-formationprocesses and the relative timing of deformation and iron oxideenrichment is still contentious for most known iron ore de-posits globally. Syngenetic models propose that synsedimentary or diagenetic structures, such as extensional faults or boudi-nage (Findlay, 1994), led to lithofacies variations or diagenetic

modification of BIF and the production of chert-free iron for-mation (e.g., Lascelles, 2007a). Supergene and supergene-metamorphic models (Morris, 1985) regard existing struc-tures as high-permeability zones for the circulation of ancientor recent meteoric fluids through BIF, causing an upgrade iniron by the leaching of gangue minerals from BIF to form ore(Ramanaidou, 2009) or by the pseudomorphic replacement of gangue minerals with goethite (Morris et al., 1980). In hypo-gene models, hydrothermal fluid flow associated with defor-mation is interpreted to be important for the localization of

iron oxide mineralization in low mean stress zones within struc-tures (e.g., Dalstra and Rosière, 2008). Hypogene-supergenemodels are proposed to explain multistage upgrading processesof iron in BIF (e.g., Barley et al., 1999; Taylor et al., 2001;Thorne et al., 2004). Whereas structural models are well estab-lished for many Proterozoic BIF-hosted iron ore deposits inthe Hamersley basin in the Pilbara craton (e.g., Powell et al.,1999; Taylor et al., 2001; Thorne et al., 2004; Dalstra, 2006),the same level of understanding is lacking for BIF-relatedhigh-grade magnetite and hematite iron ore deposits in BIFin Archean granite-greenstone belts (e.g., Lascelles, 2006b).

The goethite-hematite iron ore deposits located in theKoolyanobbing greenstone belt are among the most signifi-cant accumulations of high-grade iron ore in the Yilgarn cra-ton (Greentree and Lord, 2007). The premining iron ore re-sources of the Koolyanobbing greenstone belt deposits were~150 Mt at 58 percent iron cut-off grade, the K deposit atKoolyanobbing with >100 Mt premining resource is consid-ered to be the largest single deposit in the Yilgarn. Indicatedmineral resources, as in 2008, are 32.4 Mt at 58 percent ironcut-off grade (Portman, 2008).

This paper presents a structural model for the Koolyanob-bing iron ore deposits in the context of the geologic evolution

The BIF-Hosted High-Grade Iron Ore Deposits in the Archean KoolyanobbingGreenstone Belt, Western Australia: Structural Control on Synorogenic- and

Weathering-Related Magnetite-, Hematite-, and Goethite-rich Iron Ore

THOMAS ANGERER† AND STEFFEN G. HAGEMANN

Centre for Exploration Targeting, Department of Earth and Environmental Science,University of Western Australia, M006, 35 Stirling Highway, Crawley, WA 6009, Australia

Abstract

The Koolyanobbing banded iron formation (BIF)-hosted iron ore deposits (total premining resources ~150million metric tons (Mt), indicated reserves ~32 Mt) are located in the Mesoarchean lower succession BIF of the Koolyanobbing greenstone belt, Younami terrane, Yilgarn craton in Western Australia. In the Koolyanob-bing greenstone belt a deformation sequence that broadly correlates with the proposed deformation history of most greenstones belts within the Southern Cross domain includes: D1 structures (mainly small-scale F1a andF1b folds, formed in a north-south to northwest-southeast compressional regime), a ductile to brittle deforma-tion sequence, D2 to D4 (generated during east-west compression) and, a late-stage brittle segmentation of BIFand reactivation of faults, attributed to D5.

The formation of the seven known medium- (45−58 wt % Fe) to high-grade (58−68 wt % Fe) magnetite-,

martite-, specularite-, and goethite-bearing orebodies can be subdivided into four Archean stages and one weathering-related upgrade from the Permian and/or Mesozoic to recent times. The Archean ore-formingstages comprise: (1) early Fe-Mg ± Ca metasomatism causing local ferroan carbonate and ferroan talc alter-ation of the metamorphosed quartz-magnetite BIF protolith; (2) sequential syn-D2a (coaxial) to syn-D4 (trans-pressional) tight folding-driven removal of carbonate, quartz and minor ferroan talc by solution and mechanicaltransfer, producing residual enrichment of medium- to high-grade magnetite ore; (3) magnetite mineralizationin syn-D2b and syn-D4 breccias and fractures, forming medium-grade ore zones, or overprint magnetite in BIFand first-stage magnetite ore; and (4) mineralization of hydrothermal specularite and locally associated ferroandolomite-quartz alteration, and local oxidation of magnetite in and near brittle D 4 faults, fractures, and reacti-

vated F1 and F2a fold cores.Modern weathering-related leaching of carbonate (and minor quartz), pseudomorphic goethite replacement

of existing iron oxides and gangue, and coeval or subsequent to oxidation in the vadose zone formed goethite-martite ore with local relics of specularite or magnetite and/or kenomagnetite. The intensity and localization of this supergene modification is, in most deposits at Koolyanobbing, controlled by existing hypogene magnetite,specularite-rich medium- to high-grade ore zones and/or carbonate-altered BIF at depth. The existence of

high-grade ore below the weathering horizons suggests the possibility of further concealed magnetite- and/orspecularite-rich orebodies within the deposits and region.

† Corresponding author: e-mail, [email protected]

©2010 Society of Economic Geologists, Inc.Economic Geology, v. 105, pp. 917–945

Submitted: December 8, 2009 Accepted: May 1, 2010



of the Koolyanobbing greenstone belt and proposes a multi-stage structurally controlled iron ore genesis, which resulted inmagnetite ore (which is currently uneconomic), coarse- or fine-crystalline hematite ore (specularite and microspecularite, re-spectively), and goethite-martite ore. The basis for this inter-pretation is comprehensive mapping of the A, B, D, and K pits,field mapping in undeveloped deposits (C, E, and F deposits),and petrographical and mineralogical studies of drill cores.

Geologic Setting of the Koolyanobbing Greenstone Belt

The Koolyanobbing greenstone belt is situated 350 km eastof Perth, Western Australia. The northwestern-striking, elon-gated belt is exposed for approximately 35 km between the

Lake Deborah and Lake Seabrook salt lakes (Fig. 1A). Theactual length of the approximately 8-km-wide greenstone beltis unknown due to lacustrine cover. Regional airborne mag-netic images (Geological Survey of Western Australia cover-age: Barlee-Jackson and Southern Cross sheets) indicate con-tinuation of a BIF unit to within a few kilometers south of Lake Seabrook. The surrounding rocks of the Koolyanobbinggreenstone belt in the southwest are gneisses of the Ghooliand Lake Deborah domes (Chin and Smith, 1983) and in thenortheast banded gneisses. Almost the entire southwesternboundary of the belt is defined by the northwest-trendingKoolyanobbing shear zone (KSZ; Fig. 1A), which is markedby a 6- to 14-km-wide mylonite zone (Libby et al., 1991). The

918 ANGERER AND HAGEMANN

0361-0128/98/000/000-00 $6.00 918

50

60

53

63

75

80

71

55-80

6600000 mN

730000mE

740000mE

N750000mE

5 km

6591000 mN

N o r t h R

a n g e

S o u t h

R a n g e

Lake Seabrook granite

Banded gneisses

Koolyanobbing shear zone

Goethite ore, undifferentiated

Amphibolite

Ultramafic rock, undifferentiated

Mafic rock, undifferentiated

Quartzite (after chert)

Goethite-martite ore

Quartz-martite/magnetite BIF

Proterozoic Dike (undiff.)

Fault (approximate)

Anticline / syncline (inferred)

Shear zone boundary (approx.)

BIF bedding

Mafic schist cleavage

Drill hole (azimuth indicated)

Open pit / ore reserve (deposit)

JD-D3

Ironore

citamgaMrocks

serutcurtS

Mylonitic S-C fabric (sinistral)

noisseccusrewoL

Psammite, pelite

Massive pyrite

56

56

56

200 km

Perth

JD-D3

Koolyanobbing

K

ABCDE

F

Lake Seabrook

Lake Deborah

L a k e D e b o r a h

d o m e

G h

o o l i d

o m

e Granitoid and gneissGreenstone belt

Fault, shear zoneProvince boundaryTerrane boundary

K S Z

Murchison

domain EasternGoldfields

Youanmi terrane

Southern

Cross

domainSouth-

western gneissterrane

Koolyanobbing

Narryer gneiss terrane

B

A

lower BIF (0 - 100 m)

*

m0006.aclatot

upper BIF (0- 30 m)

banded gneisses

ca. 800 m *

ca. 6000 m **

SENW

middle BIF unit(max. 200 m) K A D F

FIG. 1. Overview maps. A. Simplified geologic map of the Koolyanobbing greenstone belt (inset: Yilgarn craton with itslocation in Australia). The map is based on existing 1:250.000 scale regional map (GSWA Jackson map sheet: Chin and Smith,1983), maps produced by Cliffs Asia Pacific Iron Ore Ltd., airborne magnetic and gravity imagery (carried out by Cliffs AsiaPacific Iron Ore Ltd.), and local mapping by the first author. B. Lithostratigraphic column of the lower succession in theKoolyanobbing greenstone belt, mainly based on Griffin (1981) and unpublished cores and maps of Cliffs Asia Pacific IronOre Ltd. and Western Areas NL (*Griffin, 1981; **Libby, 1991).



Koolyanobbing shear zone dips moderately toward the north-east (Drummond et al., 1993). Locally, a mostly undeformedmonzogranite (the Lake Seabrook granite) intrudes the Ghoolidome gneisses, Koolyanobbing shear zone mylonites, andKoolyanobbing greenstone belt succession.

Lithostratigraphy

The Koolyanobbing greenstone belt exhibits rocks of thelower greenstone succession (Cassidy et al., 2006), which isassumed to be the oldest (minimum age 3023 ± 10 Ma: Nel-son, 1999) sedimentary succession within the Youanmi ter-rane (Chen et al., 2003). Throughout the Youanmi terrane,the lower greenstone succession consists of discontinuousbasal, clastic quartzites, followed by thick tholeiitic basaltflows, high-magnesium basalts, minor basic tuffites and ko-matiites, several BIFs, and minor clastic sedimentary rocks(e.g., Chen et al., 2003; Wyche et al., 2004; Cassidy et al.,2006).

A lithostratigraphic column of the Koolyanobbing green-stone belt has been established assuming a uniform strati-graphic younging of the inclined lithostratigraphic units toward the northeast (Fig. 1B). The sequences of the lowergreenstone succession in the Koolyanobbing greenstone beltcontain less acid to intermediate lithologic units than mostother greenstone belts in the Yilgarn craton (Chin and Smith,1983). The estimated thickness of the tholeiitic volcanic se-quence in the Koolyanobbing greenstone belt is 6 km (Grif-fin, 1981). The basalts and tuffites were metamorphosed tohornblende- and actinolite-dominated amphibolites and chlo-rite schists (locally talc bearing). Ultramafic layers are abun-dant in the lowermost and in the upper section of the lithos-tratigraphic column and have a total thickness of about 800 m(Griffin, 1981). These originally peridotitic komatiite flows(Nesbitt, 1971) are now carbonate-bearing talc-, chlorite-,

tremolite-, antigorite-rich schists (Griffin, 1981). Exposedclastic sedimentary rocks of the Koolyanobbing greenstonebelt include quartzite and pelites (Griffin, 1981). Three BIFunits within the Koolyanobbing greenstone belt define topo-graphical ridges, which strike broadly parallel to the Kool- yanobbing greenstone belt and have thicknesses between 50and 180 m, locally up to 260 m. The middle BIF unit is themost prominent, with a strike continuation throughout theentire length of the greenstone belt.

The middle BIF unit comprises mainly layered quartz-magnetite rock (BIF s.str.), which is weathered to quartz-martite ± goethite BIF from the surface to depths of about70 m. Laminated or massive metacherts, which contain re-crystallized quartz, are intercalated in the BIF, especially in

the North Range. Within the middle BIF unit, particularly in the K deposit, rocks are locally magnesium-rich (talcschist, layered talc-magnetite BIF, talc-martite BIF,dolomite-magnetite BIF). Amphibole ± carbonate-rich BIFis observed in the lower and middle BIF unit, e.g., at theLake Seabrook and Jock’s Dream prospects (Fig. 1A).Siderite-magnetite BIF occurs in the A and F deposits. Athin (approx 10-m) layer of chlorite schist, probably repre-senting a metamorphosed tuff, is intercalated (boudinaged)in the middle BIF unit. Massive pyrite bodies with thick-nesses between 5 and 70 m are locally present at the strati-graphic footwall of the BIF.

Magmatism and metamorphism: A large proportion of theupper crust in the Southern Cross domain comprises multi-phase granitoid batholiths that were emplaced during 2.80and 2.67 Ga, resulting in separation of the greenstone belts(Mueller and McNaughton, 2000, and references therein).These granitoids were emplaced and deformed duringYouanmi terrane-wide, pre- to synorogenic plutonic phases(Gee, 1979; Chen et al., 2001, 2004). One of these batholiths,the Ghooli dome, is located adjacent to the Koolyanobbinggreenstone belt in the southwest and constitutes a series of sequentially intruded granitoids with intrusion ages (young-ing inward) from 2775 ± 10 to 2691 ± 9 Ma (Dalstra et al.,1998; Qiu et al., 1999; Mueller and McNaughton, 2000). Peakmetamorphism of rocks in most greenstone belts of theSouthern Cross domain was associated with the intrusion of these granitoid batholiths. An increase in metamorphic gradefrom the center (subgreenschist to greenschist) to the bound-ary (up to amphibolite facies) of greenstone belts is thoughtto be the result of regional-scale contact metamorphism(Ahmat, 1986).

A series of postorogenic and postmetamorphic plutons(mostly monzogranites) intruded gneiss domes and green-stone belts throughout the Youanmi terrane between 2.66and 2.60 Ga (Chen et al., 2004, and references within). They most likely formed as a result of anatexis of the second gen-eration plutons (Mueller and McNaughton, 2000). The LakeSeabrook magnetite series monzogranite has been dated at2656 ± 3 Ma (Qiu et al., 1999). A cummingtonite-hornfelscontact aureole in BIF and amphibolites next to the slightly deformed granite is observed in the Koolyanobbing green-stone belt at Lake Seabrook.

Proterozoic mafic and ultramafic dikes in east-trending ten-sion fractures represent the latest stage of crustal-scale intru-sion processes in the central Yilgarn craton (Hallberg, 1987).

Regional structural settingA four-stage deformation sequence is proposed for the

Southern Cross domain of the central Yilgarn craton (Table 1;Libby et al., 1991; Eisenlohr et al., 1993; Dalstra, 1995; Dal-stra et al., 1999; Greenfield and Chen, 1999; Chen et al.,2001, 2004). D1 was associated with a north-south compres-sional paleostress regime and resulted in upright to recum-bent tight to isoclinal folds and localized thrusts. Subsequenteast-west shortening included a deformation sequence char-acterized by coaxial D2 and inhomogeneous transpressionalD3. This east-west shortening is considered to be the mainorogenic deformation period in the central Yilgarn craton re-lated to amalgamation of most of the Yilgarn craton terranes

(Gee, 1979). All deformed granitoids and greenstone belts inthe Youanmi terrane are characterized by D2 deformation,mainly involving the generation of a pervasive foliation ingranitoids and upright folds in greenstone belts. The entireKoolyanobbing greenstone belt succession represents thenortheast-dipping limb of a larger F2 fold (cf. Griffin, 1981;Chin and Smith, 1983). D3 produced megascale moderate- tosteep-dipping shear zones with transcurrent kinematics, ei-ther northwest trending with sinistral movement or northeasttrending with dextral movement (Chen et al., 2001, 2004).The L-S mylonites of the Koolyanobbing shear zone, showingtranscurrent sinistral kinematics, are products of D3 (Libby et

BIF-HOSTED IRON ORE DEPOSITS, ARCHEAN KOOLYANOBBING GREENSTONE BELT, WA 919

0361-0128/98/000/000-00 $6.00 919



al., 1991; Chen et al., 2001, 2004). Because the LakeSeabrook granite intruded the Koolyanobbing shear zone, itsintrusion age (2656 ± Ma: Qiu et al., 1999) may be the mini-mum age of the Koolyanobbing shear zone ductile deforma-tion event. The youngest ductile deformation within theSouthern Cross domain has been reported from the LakeJohnston greenstone belt (2629 ± 1Ma: Joly et al., 2010).Post-D3 brittle faults have varied orientations in the Southern

Cross domain and are interpreted to be results of east-west−to east-northeast-west-southwest−shortening (D4: Dalstra etal., 1999; also Wyche, 1999), and north-south−shorteningcompression with localized brittle reactivation and thrusting(D5: Dalstra et al., 1999).

Petrography of BIF and Iron Ore inthe Koolyanobbing Greenstone Belt

Throughout the Koolyanobbing greenstone belt, the leastaltered, nonweathered, protolith BIF is a quartz-magnetiteBIF. Minor cummingtonite-quartz BIF crops out at LakeSeabrook. The BIF is host to five high-grade iron ore types, which are distinguished based on the dominant iron oxide:

magnetite, martite, specularite, goethite, and goethite-mar-tite ore. The latter is volumetrically the most important type(about 80% of the total iron ore resources). Typically, orelithologic units show mixtures of the main iron oxides (e.g.,magnetite-martite, specularite-martite, specularite-martite-goethite ores). This complexity prohibits an accurate quan-tification of individual ore types. Information about iron orecontaminants, P, Si, Mg, Al, S, are given qualitatively foreach ore type, i.e., low, medium, high, referring to the com-mon terminology used in the iron ore industry. Bulk geo-chemical values of contaminants for the deposits are shownin Table 2.

Quartz-magnetite/martite/goethite BIF

The lower succession of the Koolyanobbing greenstone beltconsists mainly of least altered quartz-magnetite BIF, weath-ered quartz-martite-goethite BIF, and locally hydrothermally oxidized quartz-martite BIF. The texture of the BIF is char-acterized by micro- to mesolayers of iron oxide, intercalated with quartz mesolayers. The average iron oxide mesolayer

thickness increases from 2 to 5 mm, from northwest to south-east within the middle BIF unit (Griffin, 1981). The layer tex-ture is typically anastomosing and centimeter- to microscalestructures are common, including boudinage, shear zones,harmonic isoclinal or tight folds, disharmonic folds, and dis-crete faults and veins that crosscut layering.

Iron oxides: Magnetite commonly occurs as amalgamatedanhedral grains (< 0.05 mm) and is the main constituent of iron oxide layers in least altered BIF, which is characterizedby dark gray magnetite and white or light gray quartz layers.Quartz-martite BIF displays bluish-gray martite layers and lo-cally, typically at microfaults, reddish quartz layers. Weath-ered siliceous BIF is commonly present as quartz-martite-goethite BIF with gray-brown martite-goethite layers and in

many localities ochreous quartz layers. Therefore, the specificcolors of the quartz layers reflect the presence, and the oxi-dation-hydration state, of iron oxide grains. Goethite partially replaces quartz, martite, and amphiboles, and commonly fills voids. Elevated contents of goethite in BIF are observedcharacteristically proximal to goethite and goethite-martiteore.

Quartz: Siliceous BIF layers are characterized by recrystal-lized quartz (~0.025 mm). Medium-grained granoblasticquartz shows grain boundary area reduction in the contact au-reole of the Lake Seabrook granite and in zones of euhedralmagnetite mineralization, close to iron ore.


0361-0128/98/000/000-00 $6.00 920

TABLE 1. Comparison of the Structural Evolutions of the Central Southern Cross Domain1 and the Koolyanobbing Greenstone Belt

Central Southern Cross domain Koolyanobbing greenstone belt

Tectonic regime Tectonic event Structures and P-T conditions Tectonic event Structures and P-T conditions

N-S compression D1 Regional detachments, D1a Regional recumbent isoclinals folds,Moderate- to low-P-T conditions, Low- to moderate-P-T conditions

Localized recumbent folds, moderate- D1b Localized upright to inclined folds,to high-P-T conditions Low- to moderate-P-T condition

E-W compression D2 Regional upright to inclined folds, D2a Regional upright to inclined folds,Low- to high-P-T conditions Low- to moderate-P-T conditions

D2b Localized upright to inclined folds,brittle-ductile reverse shearing

Low-P-T conditions (hydrothermal)

E-W transpression D3 Regional ductile transcurrent shearing, D3 Localized ductile shearing (KSZ),Low- to high-P-T conditions Low- to moderate-P-T conditions

D4 Regional transcurrent faulting, D4 Localized brittle transcurrent faulting,Moderate- to low-P-T conditions Regional folds around vertical axis,

Regional cleavage,Low-P-T conditions (hydrothermal)

N-S compression D5 Localized brittle reactivations and thrusting, D5 Localized brittle reactivations,

low-P-T conditions Low-P-T conditions

1 Summarized in Dalstra et al. (1999)



Grunerite and cummingtonite

The magnesium/iron ratio of amphiboles is high (i.e., cum-mingtonite composition) in BIF that occurs in the SouthRange and low (i.e., grunerite) in the North Range (Griffin,1981). Abundant coarse-grained grunerite is observed in de-formed quartz-magnetite BIF located close to the Koolya-

nobbing shear zone, and a quartz-cummingtonite BIF is ob-served at Lake Seabrook.Minor and accessory minerals: Minor minerals in BIF in-

clude disseminated siderite, ferroan talc, chlorite, iron sul-fides, and garnet (Davis, 1972; Sullivan, 1973; Griffin, 1981).

Magnetite ore

Laminated magnetite ore and ore breccia: Medium-, locally high-grade magnetite ore (45−63 wt % Fe, high Mg, Si, S, Pcontamination) is commonly laminated and characterized by primary magnetite mesolayer and layers of fine-grained an-hedral magnetite, which are microporous (0.02 mm; Fig. 2A,B). Laminated magnetite ore locally displays microfolds, whichcommonly show a “brick-and-mortar” fabric. Magnetite ore

breccias are mineralogically similar to laminated magnetite orebut display a texture of brecciated layers in a fine-grained, cat-aclastic, magnetite matrix. Common gangue minerals in mag-netite ore are quartz, ferroan talc, and Fe carbonate. Pyrite (upto 10 vol %) is present as disseminated, coarse, euhedral crys-tals, clusters, or decimeter-wide bands. Granular magnetite oreis a local variety of medium-grade magnetite ore, which is richin sub- to euhedral, coarse-grained, partially martitized, mag-netite crystals and pressure solution seams (Fig. 2C). Locally,magnetite is replaced by disseminated specularite crystals ortruncated by specularite veins. In specularite-rich areas, mag-netite crystals show evidence of intense martitization.

Martite ore

Martite ore (58−67 wt % Fe, medium to high P, low S) is ei-ther massive or vuggy laminated. In the weathering zone andalong faults, martite ore shows typically goethite replacementto form goethite-martite ore.

Massive laminated and martite ore breccia: Laminatedmartite ore is a common ore type and shows textural charac-teristics that are similar to laminated magnetite ore, such asmartite layers, intercalated with cataclastic martite layers, or“brick-and-mortar” fabrics (Fig. 2E). Martite ore breccia tex-tures show similarities to magnetite ore breccias.

Vuggy laminated martite ore: This ore type is characterizedby martite layers with interstitial layer-parallel voids (Fig. 2F).This indicates minimal lithostatical pressure with limitedcompaction and/or collapse during and after leaching. There-

fore, this ore type is most likely a product of mineral leachingduring recent near-surface weathering.

Specularite ore

Specularite ore: Zones of monomineralic hydrothermalspecularite (65−68% wt Fe, low P, low S) are only observed insome fault zones in the K, C, and D deposits, where it is lo-cally preserved as large, pseudohexagonal crystal blades withbasal planes of up to 10-cm diameter (Fig. 2H) and morecommonly as friable masses of fine-grained, brecciated spec-ularite. In the latter example, specularite probably experi-enced grain-size reduction during the reactivation of faults.

Specularite-martite ore: Specularite with commonly >5-mm grain sizes occur as disseminated crystals, veins, or podsin some martite ore and goethite-martite ore at the K, C, andD deposits (Fig. 2G). Locally, microspecularite with <1-mmgrain sizes is a major constituents of specularite-martite ore(63−68 wt % Fe, medium P, low S), where it partially replacesquartz and martite layers (Fig. 2I).

Goethite-martite ore

High-grade goethite-martite ore (58−63 wt % Fe, mediumP, low S) is locally present in the weathering zone, i.e., withinapproximately 70 m of the present surface in all iron ore de-posits. Goethite typically replaces martite, specularite, organgue minerals.

Massive layered goethite-martite ore: This ore type is com-mon in small- to medium-size orebodies, such as the D and Edeposits, where no deep-seated martite or magnetite orebody is present. The mineralogical difference between this oretype and the laminated ore described above is that the gangueminerals are replaced by microcrystalline goethite to form alayered rock of martite and goethite. Proximal to this layeredore, quartz-martite BIF is characteristically goethite rich.

Massive goethite-martite breccia ore: Goethite-martitebreccia ore is characterized by brecciated, martite mesolayersand locally also microspecularite layers that are enveloped by a groundmass of microcrystalline goethite. Proximal to theBIF host rock, these breccia matrices are siliceous.

Goethite ore

Vitreous goethite: Tabular orebodies of massive and vuggy, vitreous goethite (58−60 wt % Fe, high Al, high Si, medium-high P, low S) are located at the present surface, typically above goethite-martite ore and locally above BIF. This orecommonly contains minor quartz and clay and lacks BIF tex-

ture. Locally, within the vitreous goethite zone, spongy, botry-oidal, stalactite, or tubular goethite is present. These goethitictextural variations also occur as infills in voids within BIF andgoethite-martite ore and are observed in the weathering zoneof the A, C, F, and K deposits but are volumetrically minor.

Gossanous goethite: Deep-seated massive pyrite bodies atbasal parts of the middle BIF unit (Ellis, 1958) are expressedas gossanous goethite (55−59 wt % Fe, high Al, high Si, low P, low S) near the surface. They display a high porosity andcharacteristic boxwork textures after dissolved cubic pyritecrystals.

Ochreous goethite ore (limonite): Unconsolidated ochreousgoethite (50−58 wt % Fe, high Al, high Si, low P, low S) is aproduct of massive, vitreous, or gossanous goethite weather-

ing. Ochreous goethite zones in the Koolyanobbing green-stone belt are commonly controlled by deep weathering alongfault zones within goethite-martite orebodies. Ochreousgoethite is also a minor constituent in some weathered BIFand goethite-martite ore.

Mineralized BIF and BIF breccia

Mineralized BIF and BIF breccia is a medium-grade ore(45−58 wt % Fe) characterized by abundant iron oxide-richpods or veins of decimeter to several meters in size. Below the weathering front, the mineralogy of iron-rich pods andbreccia matrices includes mainly amalgamated, euhedral to


0361-0128/98/000/000-00 $6.00 921




0361-0128/98/000/000-00 $6.00 922

1 cm

A

B DC

0.2 mm

py

talc

cataclastic mag layer

primary mag layer

coarsely brecciatedprimary mag

cataclasticmag zone

primary maglayer

cataclasticmag layer

coarsely brecciated mag

C

mart

py

chl

cataclasticmag

mar

kmag/mag

D4(?)-cleavageplane

goe

qtz

D4(?)-cleavageplane

por

0.2 mm0.2 mm

E G

I

F

H

1 cm

0.2 mm1 cm

primary meso-layer(mag/mar)

shem

mar meso-lay er

mar micro-layers

goe-layer

shem

layerboundary

qtz

mar meso-layer

mar

shem

void+ogoe+hhem

lens(2.5cm)

goe

mar meso-layer

FIG. 2. Ore textures. A. Laminated magnetite ore from the K deposit (core sample KPDDH012-2), characterized by pri-mary layers of coarse magnetite (light gray) intercalated with layers of fine magnetite (dark gray). B. Photomicrograph re-flected + transmitted light, showing details of laminations described in (A), pyrite replaces locally fine magnetite. C. Pho-tomicrograph reflected + transmitted light, showing coarse-grained magnetite ore that is rich in euhedral magnetite grains andpressure solution seams that define a cleavage. D. Photomicrograph reflected + transmitted light of a BIF breccia matrix atthe C deposit, showing a (keno-) magnetite- and/or martite-mineralized zone, which replaced quartz matrix; the zone is de-formed by a pressure-solution cleavage (D4); in the left part quartz is replaced by goethite, seeding on martite grains (sampleC-6). E. Massive laminated martite ± goethite ore from the K pit, showing a “brick-and-mortar” fabric of primary mesolayers,

which are intercalated with fine cataclastically deformed martite (sample KP-50). F. Vuggy martite ± goethite ore at the K pit with secondary ochreous goethite and reddish hydrohematite in voids. G. Martite-specularite ore at the D deposit. H. Handspecimen of specularite ore from a fault zone in the K pit (sample KP-920). I. Photomicrograph reflected + transmitted lightof martite-microspecularite ore from north of the K deposit (sample KN-1). Abbreviations: goe = goethite, hhem = hydrohe-matite, kmag = kenomagnetite, mar = martite, ogoe = ochreous goethite, py = pyrite, qtz = quartz, shem = specularite.



anhedral magnetite and/or martite grains. Locally, medium-grade mineralized BIF is overprinted by hydrothermal spec-ularite. Within the weathering zone, iron oxide is mostly al-tered to kenomagnetite and martite and replaced by massivegoethite (Fig. 2D).

These mineralized BIF and BIF breccias are typically asso-ciated with strongly folded or faulted zones (i.e., satellite fea-tures) proximal to high-grade orebodies. At the A, B, C, andF deposits, meter- to decameter-wide mineralized pods arehosted in siliceous breccias or in BIF. These mineralized podstypically display breccia textures that are characterized by gangue minerals, which are replaced by iron oxides (Fig. 2D).Decimeters to meters away from structures, mineralized BIFretains its layered fabric but shows iron oxide replacement of quartz layers.

Descriptions of the Koolyanobbing Iron Ore Deposit

History of exploration and mining at Koolyanobbing

The Koolyanobbing greenstone belt hosts seven knownhigh-grade iron ore deposits, which are hosted in the south-ern section of the middle BIF unit and named the A, B, C,D, E, F, and K deposits (Fig. 1A). The A, B, D, and K de-posits are exploited by open-pit mines. We use the term “de-posit” to refer to all high-grade orebodies exploited by openpits or existing as undeveloped ore reserves in 2009. The K,A, and C deposits are described in this paper, representing

the economically most important deposits in the Koolyanob-bing greenstone belt.

Iron ore mining at Dowd’s Hill (K deposit) started in 1948 when the government operated the mine to supply charcoaliron to the Wundowie iron smelter. Between the mid 1960sand 1983, BHP operated the K and A mines to supply theKwinana blast furnace. After a ten-year period of care andmaintenance, Portman Iron Ore Ltd., in a joint venture witha Chinese partner, bought the operation and mining recom-menced in 1994. Since 2007, the mines have been operatedby the Koolyanobbing Alliance, a joint venture between Port-man and BGC Contracting. Since 2009, Portman has comeunder 100 percent ownership of Cliffs Natural ResourcesAsia Pacific Iron Ore Ltd.

Mineral exploration during the 1950s led to the discovery of massive sulfide bodies, sulfide-hosted goethite gossan, andmedium- to high-grade magnetite ore below the high-gradehematite-goethite ore at the K and A deposits (Ellis, 1958).Sullivan (1973) described the hematite-goethite ore in theKoolyanobbing greenstone belt “South Range” (south of 6591000 mN) as a product of supergene enrichment of oxideand sulfide facies BIF. The lithostratigraphy, structural, andmetamorphic setting, and the nature of iron ore in theKoolyanobbing greenstone belt have been described by Grif-fin (1981). He concluded that the BIF-hosted iron ore is acomplex product of an initial hydrothermal mineralizationstage related to structurally controlled hydrothermal fluid


0361-0128/98/000/000-00 $6.00 923

TABLE 2. Summary of the Characteristics of the Koolyanobbing Iron Ore Deposits, Showing Their Resources, Average Ore Grades andContaminants, Indicated Ore Protoliths and Ore Types, and the Spatial Controls of the Ore Types

Indicated ore typesTons Fe P SiO2 AI2O3 S LOl

Deposit Resource (Mt) (wt %) (wt %) (wt %) (wt %) (wt %) (wt %) Ore protolith mag mar s-m g-m goss

K Indicated 10.4 62.25 0.103 3.24 0.61 0.063 5.47 Quartz, carbonate, x x x x xTotal 10.4 62.25 0.103 3.24 0.61 0.063 5.47 (talc) magnetite BIF,massive pyrite

A Indicated 4.6 61.71 0.097 3.54 0.62 0.348 6.29 Quartz and carbonate- x x x xTotal 4.6 61.71 0.097 3.54 0.62 0.348 6.29 magnetite BIF,

massive pyrite

B Indicated 3.6 60.96 0.018 4.20 0.59 0.073 6.29 Quartz-magnetite BIF x xInferred 0.2 60.51 0.039 4.97 0.40 0.079 6.88Total 3.8 60.94 0.019 4.25 0.58 0.073 6.33

C Indicated 6.9 61.50 0.033 2.68 0.71 0.066 7.27 Quartz-magnetite BIF, x x x xInferred 0.2 61.35 0.009 3.42 0.36 0.064 7.02 massive pyriteTotal 7.1 61.49 0.032 2.69 0.71 0.067 7.26

D Indicated 0.3 60.67 0.031 4.47 1.46 0.019 6.62 Quartz-magnetite BIF, x x x x

Total 0.3 60.86 0.031 4.46 1.44 0.019 6.63 massive pyrite

E Indicated 0.6 60.54 0.035 4.90 0.39 0.050 7.37 Quartz-magnetite BIF x xTotal 0.6 60.54 0.035 4.90 0.39 0.050 7.37

F Indicated 5.6 61.36 0.042 2.87 0.78 0.084 8.01 Quartz- and carbonate- x xInferred 0.1 60.66 0.022 3.41 0.73 0.012 8.08 magnetite BIF

Total Indicated 32.4 61.66 0.064 3.29 0.66 0.108 6.57Inferred 0.5 60.92 0.024 4.04 0.46 0.060 7.16Total 32.9 61.65 0.064 3.30 0.66 0.108 6.58

Notes: Resource data from Portman (2007); medium- to high-grade “magnetite ore” is not included in tonnage and grades; ore abbreviations: g-m =goethite-martite ore, goss = gossanous goethite ore, mag = magnetite ore, mar = martite ore, s-m = specularite-martite ore



flow, followed by subsequent upgrading by recent weather-ing-related quartz and carbonate leaching. In addition to theupgrading effects of supergene leaching, early formation of syngenetic iron ore has also been suggested (Lascelles,2007b).

K deposit

The K deposit is located in the central part of theKoolyanobbing greenstone belt, about 1.5 km north of thetown of Koolyanobbing (Fig. 1A). This deposit is the most sig-nificant in the Koolyanobbing greenstone belt in terms of itssize and ore grade, however the iron ore is phosphorous rich(Table 2). As of 2008, the K deposit has an indicated mineralresource of 10.4 Mt goethite-martite and specularite-martiteore with an average grade of 62.25 percent iron. A zone of massive magnetite ore, informally named the “K deeps” de-posit (Guarin et al., 2009), is located beneath near-surfacegoethite-martite ore.

Lithostratigraphy and BIF host rocks: The footwall of theK deposit consists of chlorite ± talc schists and decameter- wide lenses of massive metabasalts and metatuffites (Fig. 3).The chlorite ± talc schist is overlain by a 230-m-wide BIF thathosts the iron ore. The northern hanging wall to the BIF com-prises chlorite schists, which are petrographically similar tothe footwall schists. In the western part of the open pit, re-crystallized chert and cherty BIF are the dominant basalrocks of the BIF unit. There is an increase of martite bandsand a corresponding increase in the iron oxide/quartz ratio with distance to the north and upsequence in the BIF unit.The quartz-magnetite BIF displays considerable composi-tional variation in the south and southeastern pit walls fromcarbonate-magnetite to talc-magnetite BIF. Several strike-parallel, 5- to 10-m-thick, en echelon lenses of chlorite schistbands are located in the upper part of the BIF unit. An ap-

proximately 300-m-long and 50- to 70-m-thick lens of massivepyrite is located between the footwall mafic rocks and theBIF (Fig. 4).

Hydrothermal mineral assemblages: Large parts of BIF,magnetite ore, and fault zones at the K deposit are affected by hydrothermal talc and carbonate alteration (Fig. 5). Talc-magnetite BIF is common in the K deposit, and it is charac-terized by ferroan talc mesolayers, magnetite meso- and mi-crolayers, and talc-magnetite mesolayers. Talc-magnetiteschists contain ferroan talc matrices with up to 50 percentmagnetite-rich lenses, which define a weak foliation (D2b,Fig. 6D). This rock is interpreted to be deformed talc BIF,because magnetite lenses are formed by deformation of pri-mary iron oxide layers. In the weathering zone, talc schist has

martite-rich schlieren. Lithologic transitions between talc-rich and siliceous BIF that are oblique to the bedding planessupports the interpreted metasomatic formation of ferroantalc (Fig. 6G). Chlorite ± talc schists are common in the footwall and hanging wall of the K deposit but uncommon as acountry rock within the Koolyanobbing greenstone belt and,therefore, may also be a result of this hydrothermal alterationevent. The origin and timing of the Mg metasomatism is con-tentious. Foliated and crenulated fabrics in talc-magnetiteBIF and schist and syndeformational ferroan talc fibrous inboudin necks of “brick-and-mortar” fabrics (Fig. 6D) suggestthat ferroan talc formed prior to D4.

Magnesium- and Mn-bearing siderite is observed as finegrains interlocked with and overgrowing recrystallized quartzin transitional talc-carbonate-quartz BIF and as inclusions inlargely undeformed hydrothermal sparitic ferroan dolomiteand euhedral magnetite. Magnesium- and Mn-bearingsiderite and quartz in transitional talc-carbonate-quartz BIFare, in many expamles, replaced by ferroan talc.

A late-stage assemblage of sparitic ferroan dolomite (Fig.6B, E), granoblastic and euhedral quartz, specularite (Fig.6F), minor ferroan talc, euhedral magnetite and pyrite, re-placed magnetite BIF, magnetite ore, and talc-schist, orformed breccia pods and veins (Fig. 6E). Minerals are mostly undeformed but locally granoblastic quartz shows unduloseextinction and subgrains, carbonate crystals are locally twinned and undulose, and the hydrothermal pods occuroften in BIF fold cores. These features suggest that there were several stages of metasomatism, or one single stage of metasomatism began during the end of the folding stages but was mostly a postfolding event.

Structures: The deposit strikes west-northwest in its west-ern part and northwest to north-northwest in its eastern andsoutheastern parts, showing a bend in between (Fig. 3A). TheBIF at the K deposit has been structurally thickened to about230 m by the following several deformational processes:

1. Decimeter-scale, tight to isoclinal F1 folding, which shows various orientations, located in limbs of mesoscale F2a folds.

2. East- to southeast-plunging, mesoscale (meter to de-cameter) open and tight F2a folding, which displays S, Z, andM shapes and fold axial planes that dip steeply toward thenorth-northeast to northeast (Fig. 7E1). The mesoscale F2a

fold cores are typically complexly deformed, showing hinge-crosscutting faults and/or fold tightening (Fig. 5B). The foldaxes in BIF and schists are rotated around a subvertical F4

axis, especially in the southeastern pit area, west of the east-ern lithon (Fig. 7E2, E3).3. A large-scale sinistral duplex and imbricate fan system

(cf. Woodcock and Fischer, 1986) also allowed structuralthickening; this is indicated by en echelon, steeply north-northeast to east-northeast dipping D4 faults that are parallelto, and crosscutting, the BIF unit at a low angle (Fig. 7E2,E3). The imbricate fan is indicated by two lithons (a westernand eastern) consisting largely of goethite-martite ore (Fig.3). Both lithons are enclosed by moderately to steeply eastdipping D4 faults. Subhorizontal and dip-slip structural lin-eations on D4 faults and structurally reactivated S2a cleavageplanes in mafic rock and talc-rich rocks are common withinthe duplex system (Fig. 7E5). The chlorite schists in the foot-

wall and hanging wall display two sets of cleavages: one setdips northeast, whereas the other dips east (Fig. 7E4). Thesecond cleavage (S4), characteristically a discrete pressure-so-lution cleavage, overprints the first one (S2a), which is definedby chlorite and quartz elongation (Fig. 8). Strike-slip and dip-slip fault movements, fold rotation, and cleavage orientationssuggest a complex oblique transpressional deformation (cf.Jones et al., 2004). D5 faults crosscut the D4 duplex and D4

imbricate fan structures or reactivate older faults.

Iron ore setting: The K deposit orebody can be subdividedinto five distinct zones according to their iron ore types (Fig.


0361-0128/98/000/000-00 $6.00 924




0361-0128/98/000/000-00 $6.00 925

6 5 9 0 2 0 0 m N

6 5 9 0 3 0 0 m N

6 5 9 0 4 0 0 m N

6 5 9 0 5 0 0 m N 7 4 0 4 0 0 m E

7 4 0 5 0 0 m E

7 4 0 6 0 0 m E

7 4 0 7 0 0 m E

7 4 0 8 0 0 m E

7 4 0 9 0 0 m E

7 4 1 0 0 0 m E

2 2 2 m R

L p i t b o t t o m ( 2

0 0 9 )

c u t b a

c k ( 2 7

0 m R L )

r a m p

A ’

A

60

68

6556

80

867184

46

49

70

3055

13

47 60

47

58

549

37

73

6670

44

40

4853

5

3

40

32

80

51

55

37

66

68

50

494

50

50

44

49

60

50

50

64

48

5055

77

50 52

60

60

5638

83

70

70

56

75

75

6646

6568

60

59

82

60

40

43

50

49

70

58

68

49

50

70

70

63

46

75

70

63

525

49

45

6 0

6 8

6 5

5 6

8 0

8 6

7 1

8 4

4 6

4 9

7 0

3 0

5 5

1 3

4 7

6 0

4 7

5 8

5 4 1

9

3

7

7 3

6 6

7 0

4 4

4 0

4 8

5 3

5

3

4 0

3 2

8 0

5 1

5 5

3 7

6 6

6 8

5 0

4 9 5 4

5 0

5 0

4 4

4 9

6 0

5 0

5 0

6 4

4 8

5 0

5 5

7 7

5 0

5 2

6 0

6 0

5 6

3 8

8 3

7 0

7 0

5 6

7 5

7 5

6 6

4 6

6 5

6 8

6 0

5 9

8 2

6 0

4 0

4 3

5 0

4 9

7 0

5 8

6 8

4 9

5 0

7 0

7 0

6 3

4 6

7 5

7 0

6 3

5 2

5

4 9

4 5

O c h r e o u s g o e t h i t e

m

e d i u m - g r a d e

( l i m o n i t e ) ,

M a r t i t e ± s p e c u l a r i t e o r e ( > 6 3 %

F e )

G o e t h i t e - m a r t i t e o r e ± s p e c u l a r i t e ( > 5 8 % F e )

G o e t h i t e o r e ( > 5 8 % F e )

I r o n o r e

M a g n e t i t e o r e ( > 5 5 % F e )

M i n e r a l i z e d B I F ( 4 5 - 5 5 % F e )

M e d i u m g r a d e ( 4 5 - 5 8 % F e )

S p e c u l a r i t e o r e ( > 6 3 % F e )

S p e c u l a r i t e - q u a r t z b r e c c i a ( l o c a l l y f a u l t - h o s t e d )

H a r d o r e

P h y l l i t i c c h l o r i t e s c h i s t

T a l c - m a g n e t i t e s c h i s t / t a l c - m a

r t i t e s c h i s t

L o w e r s u c c e s s i o n

M e t a b a s a l t

M a s s i v e P y r i t e

Q u a r t z - m a g n e t i t e B I F / q u a r t z - m a r t i t e B I F

M e t a c h e r t

C a r b o n a t e - m a g n e t i t e B I F

T a l c - m a g n e t i t e B I F / t a l c - m a r t i t e B I F

T a l c - q u a r t z B I F

C h l o r i t e ± t a l c s c h i s t

B I F u n i t M a f i c r o c k s

S t r u c t u r e s

B e d d i n g s u r f a c e s i n B I F

C l e a v a g e i n s c h i s t s w i t h

S - s h a p e / Z - s h a p e f o l d l i m b

M i n o r f o l d

F a u l t ( d e f i n e d , i n t e r p r e t e d )

R e v e r s e f a u l t

F o l d h i n g e l i n e ( i n c l i n e d )

S y n c l i n e / a n t i c l i n e

S l i c k e n l i n e a n d s t r e t c h i n g l i n e a t i o

n

S h e a r e d r o c k ( h a t c h u r e o n t o p o f

c o l o r )

D

w e s t e r n l i t h o n

4

D

e a s t e r n l i t h o n

4

D

f a u l t

5

D

f a u l t

5

f o o t w

a l l f a

u l t

t l u a f x e l p u d

D 4

h o r i z o n t a l

F ( d r a g )

4

F 4

S h e a r e d r o c k ( h a t c h u r e o n t o p o f

c o l o r )

I r o n c o n t e n t

> 5 8 - 6 8 % ( o r e )

> 4 5 - 5 8 % ( s u b - o r e )

> 1 0 - 4 5 %

> 1 - 1 0 %

> 0 - 1 %

0 %

5 8

8 0

5 4

4 4

3 2

8 0

6 0

8 6 7 5

7 5 6 5 6 8

4 6

7 5

6 3

5 ( K - d e e p s )

3 4

1 , 2

P i t f l o o r ( 2 2 2 m R

L ) F e a s s a y m a p

Z o n e s o f K d e p o s i t

A

F e a s s a y m a p

C

B

F I G . 3 . G e o l o g i c m

a p o f t h e K

d e p o s i t p i t ( A ) , t h e f i v e o r e z o n e s i n t h e d e p o s i t ( B ) , a s s a y s m a p o f t h e p i t b o t t

o m ( s e e b o x i n A f o r l o c a t i o n o f m a p ) f o r i r o n

- g r a d e

c o n t r o l , s h o w i n g B I F - t o - o r e r e l a t i o n s h i p s a n d f a u l t s g e o m e t r i e s ( C ) .




0361-0128/98/000/000-00 $6.00 926

A ’

?

?

A

400m RL

300m RL

200m RL

100m RL

0m RL

6590100mN

6590300mN

740900mE

6590500mN

741000mE

Martite±specularite ore (> 63 % Fe)

Goethite-martite ore (> 58 % Fe)

Goethite ore (> 58 % Fe)

Ironore

Magnetite ore (>55 % Fe)

ero-buS

)eF%85-54(

Specularite ore (> 63 % Fe)

Specularite-quartz breccia

erodraH

Talc-magnetite schist / talc-martite schist

noisseccusrewoL Massive Pyrite

Quartz-magnetite BIF / quartz-martite BIF

Carbonate-magnetite BIF

Talc-magnetite BIF / talc-martite BIF

Talc-quartz BIF

Mafic volcanic rock (undifferentiated)

tinuFIB

Fault, inferred

Drill hole

cifam rock

?

t l u a f

- D /

D 4

5

? t l u

a f

D 1

present pit floor (2008)

final pit floorF2

t l u

a f - D /

D 4

5

D

f a u l t

2 b

t l uaf D 4

weathering front

FIG. 4. Interpreted cross section A-A' across the K deposit (see Fig. 3 for location of section).

chlorite schistmetabasalt

chlorite schistmetabasaltqtz-mar BIF

mag ore c h

e r t

qtz-marBIF

fault

f a u l t

fault

qtz-marBIF qtz-mag

BIF mag ore

goe ore

massive pyrite

qtz-magBIF

talc schisttalc-mag BIF

phylliiticchlorite schist

shem ore

ogoe

ogoe

qtz-marBIF

tc-mag BIF

carb±tc-mag BIF

mar+shemore

goe-hem oreqtz-mar

BIF

qtz-mar BIF

talc-marBIF

mar+shemore

f a u l t

shemore

b r e

c c i a

talc-marBIF

talc-magBIF

chert

tc-marBIF

goe-spec ore

goe ore

tc-mar BIF

tc-fold core

massive mag ore

brecciatedshearedfold core

talc-schistfold core

foldcore

massive mag ore qtz-mar-shem BIF

tluaftluaf(shem-rich)

shem pod

BIF magore

BIF

5 m

BIF

FIG. 5. Lithologic interpretation of the K pit south wall (as in August 2008). Note in the left part the talc-rich easternlithon in light color and the black magnetite ore to its right showing talc-rich fold cores. Abbreviations: carb = carbonate,goe = goethite, mag = magnetite, mar = martite, ogoe = ochreous goethite, py = pyrite, qtz = quartz, shem = specularite,tc = talc.



3C). Zone 1 is a goethite ± martite ± specularite zone that ex-tends to about 70 m below the current surface. Zone 2 is amartite-microspecularite ± goethite zone situated below zone1 and includes the western lithon. Zone 3 is a 15- to 100-m- wide, quartz-martite BIF and goethite-martite zone, locatedat the southern end of the open pit. Zone 4 is situated in the

eastern lithon and is characterized in the north by two paral-lel goethite-martite ± specularite ore zones with minorquartz-martite BIF, and in the south by talc-rich rocks. Zone5 is a magnetite zone in the south wall, and in the central partof the open pit, and is characterized by medium-gradequartz-, talc-, and carbonate-magnetite BIF enveloping mag-netite orebodies.

The main magnetite ore zone is located either underneaththe martite orebody or is juxtaposed against it by steep dip-ping D4 and/or D5 faults (Fig. 4). The transition from mag-netite to dominantly martite in BIF and ore zones is matched with a decrease in magnetic susceptibility from >1 to 1× 10−3

SI units. The southern wall of the open pit exposes patchy magnetite ore located in fold limbs, boudins and anticlinalhinges of micro- and mesoscale folded talc-, quartz-, and car-bonate-rich BIF. Locally, magnetite ore is granular.

High-grade specularite ore and specularite-quartz brecciasubore are spatially controlled by the steeply north-northeast

to east-northeast dipping D4 faults of the sinistral duplex sys-tem and their adjacent wall rocks. The footwall contact is de-fined by a <10-m-thick zone of brecciated specularite and a10-m-thick zone of brecciated goethite-martite ore and chertbreccia. Mostly undeformed specularite veins are abundantin the eastern lithon.

A deposit

The A deposit, located 2.5 km south of Koolyanobbing, with about 10 Mt premining resources (4.6 Mt remaining asin 2008), is the second largest mine in the Koolyanobbinggreenstone belt (Table 2). This approximately 700-m-long,


0361-0128/98/000/000-00 $6.00 927

1 cm

1 cm

1 cm

A

DC

B

E F G

1 cm

2 cm

weathered talc

qtz-(talc-)martite BIF

qtz shem

py

goe

shem

mag. ore

dol

qtz

fold

dol

mag

mag

talc

qtz

FIG. 6. The K deposit host rocks. A. Fine laminated quartz-magnetite BIF from KKDDH017 diamond core. B. Dolomite-magnetite BIF from KKDDH017, brittle deformed. C. Talc-magnetite BIF from KKDDH017, displaying tight folding withan associated “brick-and-mortar” fabric. D. Talc-magnetite schist from KKDDH017, showing a foliation with schlieren of magnetite in talc matrix. E. Dolomite-specularite breccia pod hosted in magnetite ore in the K pit. F. Granoblastic quartz-specularite fault infill with pyrite flakes and goethite replacement. G. Alteration lenses of talc in quartz-martite, located prox-imal to eastern lithon in the K pit. Abbreviations: dol = ferroan dolomite, goe = goethite, mag = magnetite, mar = martite,py = pyrite, qtz = quartz, shem = specularite.



90-m-wide, and 50-m-deep open pit contains goethite-mar-tite ore and gossanous goethite with an average ore grade of 61.71 percent Fe. Massive magnetite ore is situated beneathgoethite-martite ore (Fig. 9).

Lithostratigraphy and BIF host rocks: Chlorite schist is thestratigraphic footwall rock to BIF in the A deposit. Explo-ration drilling intersected a 6- to 15-m-thick band of massivepyrite between the footwall chlorite schist and BIF. The BIF


0361-0128/98/000/000-00 $6.00 929

6585900 mN

6586000 mN

6586100 mN

6586200 mN

6586300 mN

742400mE

742500mE

742600mE

742700mE

> 58 - 68 % Fe (ore)

> 45 - 58 % Fe (sub-ore)

> 10 - 45 % Fe

> 1 - 10 % Fe

> 0 - 1 % Fe

0 % Fe (or not assayed)

Fe assay map: at 356 m

assay map

F2b

62

42

48

30

45

55

50

55

45

2865

53

17

26

48

63

60

55

66

65

62

42

48

30

45

55

50

55

45

2865

53

17

26

48

63

60

55

66

65

A

A’

?

?

?

300 m RL 300 m RL

350 m RL350 m RL

742200mE

742250mE

742300mE

6586350mN

6586400mN

6586450mN

6586500mN

A A’

?

t l u a f

D5

t l u a f

D

5

Mineralized BIF breccia(goethite matrix)Mineralized mafic breccia(goethite matrix)

Martite ore (> 63 % Fe)


Goethite ore (> 58 % Fe)

Canga (medium-grade ore)

Ironore

Magnetite ore (>55 % Fe)

Mineralized BIF

edarg-muidem

)eF%85-54(

Metatuffite

noisseccusrewoL

Metabasalt

Quartz-magnetite BIF / quartz-martite BIFMetachert

Carbonate-magnetite BIF

Chlorite schist

tinuFIB

skcorcifaM

serutcurtS

Bedding surfaces in BIF

Cleavage in schists

Minor fold

Fault (defined, interpreted)

Reverse fault, thrust

Fold axis trace

Brecciated rock (undifferentiated)

A B

C

55

55

Drill hole

D / D

f a u l t

2 b

4

D fault4

D

f a u l t

2 b

F2b

D f a u l t

2

F2b

F2b

t l u a f D 2

weathering front

FIG. 9. Geologic map of the A deposit pit (A), iron assay map of the pit bottom in the north pit (B), and northeast ori-ented cross section A-A' (C).



consists of quartz-magnetite ± martite and siderite-magnetitedomains (Fig. 9C). A fine-grained crystalline quartzitic rock(recrystallized chert) occurs as a lens near the footwall contactin the north part of the A deposit. A thin (<10-m) band of chlorite schist and layered felsic tuffite (weathered to abanded kaolinite-rich rock in the pit) is intercalated with theBIF. The stratigraphic hanging wall to the BIF unit includesmafic rocks and chlorite schist, which are similar to the footwall rocks.

Hydrothermal siderite-magnetite-pyrite BIF: Siderite is amajor rock constituent in siderite-magnetite BIF below the weathering zone at the A and F deposits. It forms a vein net- work within medium-grade magnetite ore. Close to the weathering zone, goethite needles replace siderite.

Structures: The deposit extends north-northwest in thenorthern half and northwest in the southern half (Fig. 9A).The S1a foliation in the BIF and lamination in the ore dipmoderately toward the east-northeast to northeast, and thedominant cleavage in mafic rocks and chlorite schists (S2a) ismoderately east or northeast to north-northeast dipping. TheBIF and mafic rocks are folded around subhorizontal north-northwest-plunging D2a and/or D2b fold axes. Folds show S-shape symmetries with axial planes dipping moderately tosteeply toward east-northeast (Fig. 7A1, A2). Several steepnortheast- to east-northeast−dipping brittle D2b and/or D4

faults, locally with brecciated zones of several meters thick-ness, truncate D2a and/or D2b fold limbs and reactivate theboundary between BIF and mafic rocks (Fig. 9C). The south-ern end of the A deposit is defined by a north-south−trend-ing, steep dipping D5 fault that crosscuts the BIF unit.

Iron ore setting: Goethite-martite ore and the quartz-mar-tite BIF host rock are common above the present ground- water table at about a 70-m depth (Hoppe, 2005). Below the weathering front quartz-magnetite BIF and siderite-mag-

netite BIF are closely associated with, and envelope at, themeter- to decameter-scale, patchy magnetite ore (Fig. 9C).Larger (decameter-wide) magnetite ore pods have been in-tersected in several drill holes. Similar to the K deposit, mag-netite ore and magnetite-carbonate BIF represent the base tothe goethite-martite ore. A spatial relationship of magnetiteiron ore with decameter-scale folds, as is the case in the K de-posit, is not clearly manifested in the shallow A pit but may beinterpreted from grade-control assay maps (Fig. 9B). Brecciasin D2b fault zones in the northern pit wall contain BIF orchlorite schist clasts and iron oxide (mostly goethite) matri-ces, suggesting at least a local brittle structural control on ironoxide mineralization. Toward the north end of the open pit,goethite-martite ore pinches out and displays a lateral transi-

tion to quartz-martite BIF. At the south end, the ore zone istruncated by the north-south−trending steep D5 fault that isbordered by a thin zone of goethite-martite ore along its east-ern margin. The vuggy gossanous goethite band at the footwall contact is interpreted to be weathered massive pyrite(Hoppe, 2005).

C deposit

The C deposit is located 4.5 km southwest of Koolyanob-bing and has an indicated mineral resource of 6.9 Mt goethiteand goethite-martite ore at an average grade of 61.5 percentFe (Table 2).

Lithostratigraphy and BIF host rocks: The BIF unit at theC deposit has an apparent thickness of approximately 350 m,although the true thickness is unclear due to complex foldingand faulting (Fig. 10). Stratigraphic footwall and hanging-wallrocks include intensely foliated chlorite schists. At least oneapproximately 10-m-wide band of mafic schist is intercalated with the BIF. The host rock of goethite-martite ore is quartz-magnetite and/or martite BIF.

Structures: Cleavages in chlorite schist (S2a, S4) are variably orientated (Fig. 7B2). Most of the C deposit is positioned inthe northeast-trending parasitic short limb of a large-scale(100-) northwest-trending S-shape fold (F4) that has a subvertical axis orientation and a northwest-trending axial plane(Fig. 7B1). In the northwest, the moderately southeast dip-ping boundary between BIF and the footwall chlorite schistsis defined by a reverse fault (D2b?). The BIF in the deposit isstructurally thickened, mainly because of (1) moderately tosteeply plunging, small-scale, M-shape folds (F1b); (2) up todeposit-scale subhorizontal-plunging parasitic F2a folds; and(3) the northwest-trending crosscutting D4 faults, which trun-cate F2a folds. The F1 fold axes scatter around a vertical F4

axis (Fig. 7B1). A <50-m wide, irregularly shaped BIF brec-cia zone is present in an F2b limb and shows siliceous and ironoxide-rich matrices.

Iron ore setting: The main ore zone, located along the west-ern margin of the deposit, comprises goethite-martite ore andis controlled by an approximately east-dipping, reverse fault.The ore is partly concealed underneath BIF owing to theshallow dip of the fault, where BIF and stratigraphic hanging- wall mafic schist are juxtaposed against each other (Fig. 10B).Goethite ore in the southwest of the deposit locally displaysgossanous textures and may indicate a massive pyrite body lo-cated at depth. Matrices of the siliceous breccia in the northwestern part are locally rich in martite and/or goethite and are

overprinted by specularite to the west. Specularite breccias, which contain BIF clasts, represent a medium iron-graderock (Hoppe, 2005). The specularite is considered to be con-trolled by a northwest-trending fault.

Deformation and Metamorphism History of the Koolyanobbing Greenstone Belt

Structures observed in the Koolyanobbing iron ore de-posits and in most parts of the Koolyanobbing greenstonebelt formed during five main deformation events (D1 to D5:Table 1). Some of these main events have been subdividedinto subevents according to different strain characteristics within the same tectonic regime (D1a, D1b and D2a, D2b).Structures are described below in order of their interpreted

formation.D1: Early coaxial folding

Two sets of early folds, F1a and F1b, are developed in all ironore deposits and in parts of the Koolyanobbing greenstonebelt. These are the earliest structures recorded within theKoolyanobbing greenstone belt, however, it is speculated thatearliest extensional faults controlled the localization of thickmassive pyrite bodies, which are abundant throughout theKoolyanobbing greenstone belt along the footwall contact be-tween BIF and mafic rocks. Such a D1 fault, although not ob-served, has been inferred for the K deposit (Fig. 4).


0361-0128/98/000/000-00 $6.00 930



F1a folds and foliation in BIF: Ductile microstructural fea-tures, such as recrystallized, anastomosing, and boudinagedquartz layers, are ubiquitously developed in BIF in theKoolyanobbing greenstone belt (Fig. 11A, B). Recrystallizedquartz typically has a preferred crystallographic orientationand minor elongation aligned parallel to fold limbs. This sug-gests that recrystallization was associated with isoclinal fold-ing. Commonly, there is a lack of recovery textures, such asgrain boundary area reduction and crystal growth, which may indicate that recrystallization ceased before the end of defor-mation. The earliest recognizable folds in the BIF aredecimeter scale, tight to isoclinal with vertical or reclined

attitudes (Fig. 11C). These folds occur locally in moderately to steeply dipping BIF, showing axial planes and fold axes thatare parallel to the BIF foliation (Fig. 7B1, C1, D1). Layeredtalc-magnetite BIF at the K deposit shows localized isoclinalF1a folds with an associated bedding-parallel (i.e., fold limb-parallel) continuous cleavage (i.e., S1b). The S1b crosscuts F1a

fold hinges.F1b open to tight folds in BIF: Throughout the middle BIF

unit, rocks are folded with up to 100-m-long wavelengths.These generally low-amplitude folds typically show internaldecimeter- to meter-scale parasitic folds with Z, S, and Msymmetries (Fig. 11E). Folds are characterized by steep axial


0361-0128/98/000/000-00 $6.00 931

6584800 mN

6584900 mN

6585000 mN

6585100 mN

6585200 mN

743600mE

743700mE

743800mE

744000mE

743900mE

744100mE

8558

318974

39

85

76

56

82

57

76

76

73

70

7080

35

46

54

88

68

73

75

81

74

60

78

56

14

86

80

82

74

7878988

8885

66

67

62

90

85

75

82

90

8864

75

87

65

70

67

63

72

75

60

30

67

63

26

30

53

35

45

70

45

50

80

50

45

40

85

70

80

70

70

45 45

8558

3189

74

39

85

76

56

82

57

76

76

73

70

7080

35

46

54

88

68

73

75

81

74

60

78

56

14

86

80

82

74

78787988

8885

6667

62

90

8575

8290

8864

75

87

65

70

67

63

72

75

60

30

67

63

26

30

53

35

45

70

45

50

80

50

45

40

85

70

80

70

70

45 45

A

A’

300m RL

500m RL6584800mE

6584900mE

6585000mE

6585100mE

6585200mE

300m RL

400m RL

500m RL

400m RL

A A’

▲▲ ▲ ▲ ▲ ▲

▲ ▲ ▲

▲▲

▲▲

▲▲

▲▲ ▲ ▲ ▲ ▲

▲▲

▲ ▲

▲

▲▲

▲ ▲▲

▲▲

▲▲

▲

45

Mineralized BIF breccia(martite matrix)

Ochreous goethitemedium-grade

(limonite)

Martite ore (> 63 % Fe)


Goethite ore (> 58 % Fe)Ironore

Mineralized BIF

edargmuideM

)

eF%85-54(

Specularite rich BIF-breccia

erodraH

Phyllitic chlorite schist

noisseccusrewoL

Quartz-martite BIF

Quartzite (after chert)

Chlorite schist

tinuFIB

cifaM rocks

serutcurtS

Bedding surfaces in BIF

Cleavage in schists

S-shape / Z-shape fold limb

Minor fold

Fault (defined, interpreted)

Thrust

Fold axis trace

brecciated rock (undifferentiated)

sheared rock (undifferentiated)

A

B

Drill hole

45

F2a

D 4 f a

u l t

D 4 f a u l t

D 4

f a u l t

D t h r u s t

2

D thrus t2

D4

f a u l t

weathering front

FIG. 10. Geologic map of the C deposit (A), and north-south−oriented cross section A-A' (B).




0361-0128/98/000/000-00 $6.00 932

0.2 mm

B

mar

mag

goe

mar

qtz

C D

F axialplane

1b

F axial plane1a

boudin lines and “brick-and-mortar” fabric

SE

NW

LstrLbud

F1, Lstr

mineralizedBIF-breccia

E 1 dm

1 cm

qtz

mag

F

A

qtz meso-layer

mar meso-layer

mar micro-layer

mar meso-layer

qtz meso-layer

post-kinematic gru

G

0.2 mm

pre-to syn-kinematic gru

mag

FIG. 11. BIF textures and structures. A. Quartz-martite-goethite BIF from the North Range displays dominantly ironoxide microlayers intercalated with quartz mesolayers (sample KB-1). B. Photomicrograph of quartz-martite BIF from theF deposit, showing iron oxide meso- and microlayers intercalated with quartz mesolayers (sample F-7; left plane-polarized,right reflected light). C. Isoclinally D1a-folded BIF south of the F deposit, characterized by north-south−trending axial plane,

which is parallel with the local BIF unit strike and buckling by D1b open folds with east-west axial planes, which are orthog-onal to the local BIF unit strike. D. Limb of a D1a-folded quartz-martite BIF at the B deposit, showing martite layer boud-inage and a “brick-and-mortar” fabric. E. Open to tight M-shape D1b folds at the C deposit, truncated by a small-scale ironoxide-mineralized breccia, which probably has been generated during D3. F. Sheared BIF close to the Koolyanobbing shearzone, from JD-D3 at the Jock’s Dream prospect. G. Photomicrograph plane-polarized light, showing pre- to synkinematicgranular grunterite and postkinematic grunerite porphyroblasts in domains of sheared BIF from diamond core JD-D3 at theJock’s Dream prospect. Abbreviations: cum = cummingtonite, goe = goethite, gru = grunerite, mag = magnetite, mar = mar-tite, py = pyrite, qtz = quartz, shem = specularite.



planes that are oriented subparallel to orthogonal to the gen-eral northwest trend of the BIF unit. An overprinting rela-tionship between folds in the BIF at Lake Seabrook (Fig.11C) indicates that F1b folds formed after F1a isoclinal folds.The F1b folds are most abundant in areas proximal to all ironore deposits, e.g., in the North Range, but are less commonin BIF distal to ore. Centimeter- to decimeter-scale mi-crothrusts that crosscut BIF layers at a low angle show asso-ciated drag folds with axes that are oriented subparallel to F1b.These thrusts may have formed during D1, however, there areno indications for large-scale thrusts related to D1 in theKoolyanobbing greenstone belt.

Fracture cleavages associated with F1b folds are rare in theBIF and do not show consistent orientations. There is noclear evidence for an S1 in mafic rocks, likely it was obliter-ated by subsequent polyphase cleavage formation.

D 2: Main horizontal coaxial compression

F 2a upright folds: Horizontal coaxial shortening during D2a

resulted in the folding of the entire Koolyanobbing green-stone belt succession around a subhorizontal northwest-trending fold axis. All lower succession units have been tiltedto a moderately to steep northeast dipping orientation (Fig.1A), presumably associated with emplacement of the sur-rounding granitoid domes (Gee, 1979; Ahmat, 1986; Chen etal., 2001, 2004). It is likely that the entire Koolyanobbinggreenstone belt represents an upright F2a mega fold limbbased on (1) observation that the dominant dip orientation isuniformly to the northeast throughout the Koolyanobbinggreenstone belt , (2) subregional-scale fold closures are inter-preted as localized S-shape parasitic F2a folds (Fig. 1A), and(3) there are no unambiguous indications of a regional-scalelithostratigraphic duplication in the belt. At the deposit scale,F2a folds are expressed as meter- to decameter-scale, pre-

dominantly S-shape parasitic folds with shallowly (A deposit,Fig. 9) or moderately (K deposit, Fig. 3, 4, 7E1) east to south-southeast plunging axes and axial planes that are subparallelto the strike of the BIF unit. The regional inclination variationof the F2a axes is either a result of hinge line variation duringfolding (cf. Ramsey, 1962), or it was likely caused by subse-quent deformations. The F2a fold limbs locally preserve boud-inaged F1a or F1b fold cores, which are present in layered talcBIF and siliceous BIF.

S 2a: Cleavage in mafic rocks: During regional horizontalcompression, mafic-ultramafic rocks, and layered talc BIF inthe K deposit developed a steeply northeast dipping S2a fab-ric (Fig. 7E3, E4). Adjacent to the BIF unit, the orientationof the cleavage in rocks is parallel to the axial plane of F2a in

BIF. This indicates that the S2a is genetically associated withthe northwest-trending, regional F2a and generally upright F2a

axial planes. The S2a is characterized by preferred chlorite andquartz orientation (Fig. 8B).

D 2b: Reverse shearing: This deformation, which was typi-cally cataclastic in BIF and ductile in chlorite and talc schists,resulted in reverse shear zones parallel to the footwall con-tacts between BIF and mafic rocks or massive pyrite (Fig. 4),locally within the BIF unit, as observed at the A, C, K, and Fdeposits. Shear zones in talc-rich rocks in the K deposit trun-cate F2a fold hinges, and internal shear facoids show sub-decimeter-scale folds with F2b fold axes that are oriented

subparallel to the F2a (Fig. 7E2). These orientations suggestthat shear zones formed as D2b progressively after F2a. The S2a

cleavage planes in chlorite schists are drag folded proximal tothe D2b fault planes. These F2b drag-fold axes in chloriteschists are oriented parallel to the F2a fold axes in the A andK deposits (Fig. 7A1, E2). Chlorite and talc schists that weredeformed during D2b tend to display phyllitic S2b cleavageplanes. At the C deposit, D2b faults are truncated by specu-larite-bearing D4 faults, which provide a minimum relativetiming (Fig. 10).

D 2b-D4: BIF and ore boudinage: Magnetite and martitemesolayers in tightly folded BIF have been structurally frag-mented by layer-parallel extension. Laminated magnetite andmartite ore display similar boudinage fabrics. This deforma-tion caused either the formation of boundin lines parallel tothe direction of fold axes (i.e., σ 2) and orthogonal to σ 3, or arather irregular “brick-and-mortar” fabric (Fig. 11D), when σ 2 value equals σ 3 value (Ramsey, 1967). Pressure shadows inboudin necks are filled with granoblastic quartz, fibrous fer-roan talc, or sparitic ferroan dolimite, depending on the sur-rounding gangue minerals. These mineral growth zones in thepressure shadows indicate a syndeformational solution trans-fer, either originated in the surrounding, compacted, ganguelayers (Hippertt et al., 2001), or from farther areas away.Quartz is not recrystallized in strain shadows in boudin necks, which suggests that boudinage took place later than the duc-tile stage of D1 and D2a, after dynamic quartz recrystallizationceased. The formation of the layer boudinage may also be re-lated to fold tightening by D2a, D2b, or D4, considering that F1

axial planes are subparallel to the compressional axes of D2a toD4 in many places throughout the BIF unit. Sparitic carbon-ate and talc “mortar” in “brick” strain shadows in BIF of theK deposit indicate at least one stage of tightening associated with D4-related hydrothermal processes.

D3: Ductile transpressionThe transpressional Koolyanobbing shear zone (Libby et

al., 1991) formed during D3. The main shear sense in my-lonites of the Koolyanobbing shear zone is sinistral (Libby etal., 1991), based on horizontal stretching lineations, asym-metric folds around vertical axes, and the orientation of S-Cand C' fabrics (cf. Passchier and Trouw, 1996). Minor, 10-m-long and 0.1-m-wide, north-northeast−striking ductile shearzones display a dextral shear sense (similar shear zone indica-tors as the main sinistral shear zones) and truncate the sinis-tral mylonite zones. Although strain increases toward the cen-ter of the mylonite zone, based on the high density andincreasing width of mylonite zones, there is a localization of

high strain especially along, and up to 100 m away from, the western side of the greenstone belt and/or mylonite zoneboundary. Mylonites exist in the granitoids on the easternside, whereas inside the Koolyanobbing greenstone belt,shear zones are defined by bands of well-foliated amphibo-lites, chlorite-actinolite schists, and talc-chlorite-carbonateschists. Diamond hole JD-D3 (western areas NL, see Fig. 1A)intersects the lowermost sequences of the Koolyanobbinggreenstone belt and the mylonite and shows that the lowerBIF unit displays strongly ductile shear zones, characterizedby disharmonic boudinage and tight folds (Fig. 11F), dynamicquartz recrystallization, and grunerite growth. Grunerite is


0361-0128/98/000/000-00 $6.00 933



present as pre- or synkinematic fine-grained masses and un-deformed postkinematic porphyroblasts (Fig. 11G). The duc-tile deformation is weaker in the middle BIF unit and inmafic rocks located away from the boundary.

D4: Cataclastic strike-slip and reverse structures

Rocks within the Koolyanobbing greenstone belt experi-enced internal deformation, which is described below as D4

strike-slip and reverse faults, F4 folds, and S4 cleavages.Imaging of airborne magnetic data (carried out by Cliffs AsiaPacific Iron Ore Ltd.) defines kilometer-scale, north- andnorthwest-striking faults, which may form conjugate sets(Fig. 1A). Several faults that are oriented at a low angle tothe BIF foliation led to the local duplication of BIF units.The most significant north-striking regional-scale faults arethose located to the northwest and southeast of the K de-posit, which display north-south−trending dextral displace-ments of several hundred meters, thus truncating theKoolyanobbing shear zone.

Deposit-scale, brittle strike-slip faults are observed at theK, C, and D deposits and are locally rich in specularite. Thesespecularite-rich faults and specularite veins truncate the ironoxide-rich D2b BIF breccia in the C deposit.

At the K deposit, horizontal to shallowly plunging slickenfibers and mineral stretching lineations on moderately tosteeply dipping S2a cleavage and fault planes in mafic rocks of the footwall, hanging wall, internal lenses, and in talc BIF andschist indicate strike-slip and reverse faulting, which post-dated D2 (Fig. 7E5). Major steep faults striking at a low angleto the BIF locally duplicated the BIF unit at the K deposit.Two BIF and goethite-martite ± specularite ore lithons areindented in the hanging-wall rocks by displacement alongfaults, which trend north to northwest in a high angle to theBIF unit (Fig. 3). This caused a north-northeast to north

strike of the lithons, deviating from the overall east-northeaststrike of the main BIF unit. Meter-scale drag folding of mul-tiphase-foliated chlorite schist around a horizontal F4 axis par-allel to the eastern lithon boundary fault is displayed in thesouthern pit wall (Fig. 7E5). Its geometry suggests reversemovement, which is probably a late-stage deformation incre-ment following sinistral strike-slip displacement and S4 cleav-age development. Fault planes west of the eastern lithoncrosscut small- to mesoscale F2a fold cores and meter- to de-cameter-scale boudins of magnetite ore are located in foldedtalc-magnetite BIF and talc-magnetite schist.

F4 rotation around vertical axes: All BIF units in theKoolyanobbing greenstone belt show local deviations fromthe uniform northwest trend of the belt (Fig. 12A) and pre-

D4 structures are folded (see small circles in Fig. 7A2, B1,E1). This suggests that F4 folding, or block rotation, around asubvertical axis took place. The rotation of BIF blocksthroughout the Koolyanobbing greenstone belt commonly ex-hibits an anticlockwise direction; an extreme form is the F4 S-shape parasitic fold developed in the C deposit (Fig. 10, andsmall-circle distribution in Fig. 7C1). The BIF in the K de-posit is slightly folded around a vertical axis, indicated by localsmall-circle distribution of F2a axes (Fig. 7E2, E3, E4).

S4 cleavage in mafic rocks: Throughout the Koolyanobbinggreenstone belt, schistose mafic rocks show a steeply east dipping fracture or chlorite cleavage, with minor scattering

between east-southeast to east-northeast (Fig. 7E4). This S4

cleavage overprints the S2a cleavage (Fig. 8B). The relativeorientation of this S4 is characteristically at a low angle(20°−40°) to the general northwest-southeast strike of theBIF units, and subparallel to the S plane orientation in theKoolyanobbing shear zone mylonites located proximal to theboundary of the Koolyanobbing greenstone belt (Fig. 1;Libby et al., 1991). The S4 is locally folded by late incrementsof vertical F4 (Fig. 7E4) and horizontal F4 drag folds at theeastern lithon (Fig. 7E5).


0361-0128/98/000/000-00 $6.00 934

0°0°

0°

K-W

K-SE

K-SE (F2)

K-W (F2)

K-SE (F2)

K-W (F2)

0°

Dol

cw22° F2a/2b

K-SE B-C

A

DN’K E

K-W

F

C

ccw22°

ccw67°

F

E

D BK-W (F2)

Dol

K-SE (F2)

BIF unit (11)

F1 (7)

F2a at K deposit (2)

F4 (2)

alternative:

A

B1

1

1

F4c

0°

F2a/2b

F4c

F2a/2b

F4c

D stress field1

D -Dstress field

2 4

alternative:

F1

F1

F1

F1

1(alter-native)

1(alter-native)

B2

C1 C2

FIG. 12. Back rotation of folded BIF unit orientations. A. Present-day ori-entation of BIF unit at the deposits and some prospects. B. East-west−trend-ing 1 σ orientation during D2 to D4 displayed by back rotation of BIF unitaround F4 to a general north-south trend (cf. Chen et al., 2001, 2004). TheSchmidt net on the right shows an alternative southwest-northeast orienta-tion of 1 σ (alternative) by back rotation of the BIF units to a general north-

west trend (right). C. Second stage of BIF back rotation around horizontal F2

axis to achieve a horizontal orientation of F1b axes displays a north-southtrending 1 σ for D1. The alternative model shows a northwest-trending 1 σ .Abbreviations: B-C = between B and C deposits, Dol = Dolphin prospect, K-SE = southeastern end of K deposit, K-W = western K deposit.



Inclined transpressional system in the K deposit: It is inter-preted that the D4 structures in the K deposit (i.e., D4 strike-slip and reverse slip faults forming the duplex and imbrica-tion, subvertical F4, and S4 cleavage) result from progressivedeformation during inclined transpression (cf. Jones et al.,2004). The initial stage of the system (Fig. 7E2) is the north-east-dipping strike slip duplex, followed by imbrications of the western and then the eastern lithons (Fig. 7E3), with eachdeformation facilitated by strike-slip and locally reverse slipfaults. The slightly different orientations of the lithons arelikely the result of a continuous anticlockwise vertical blockrotation of the K deposit during the transpression phaseunder a stationary east-west stress field.

D5: Brittle faulting and boudinage

The topographic expression of the BIF units and airbornemagnetic images (carried out by Cliffs Asia Pacific Iron OreLtd.) indicate a lateral northwest-southeast extensional boud-inage and pinch-and-swell structures, and displacement of BIF along faults that crosscut the BIF units at low to high angles, which is attributed to D5. It is unlikely that earlier de-formation stages, such as D1 to D2b, formed the boudins, be-cause they are generally not laterally imbricated by transpres-sional D4 strike-slip deformation.

At the K deposit, the footwall and hanging-wall contactshave been reactivated during D5 with the western and easternlithons truncated by dextral D5 faults (Fig. 3). Indications forsuch a reactivation are cataclastic-deformed specularitemasses and local displacement of the eastern and westernlithon faults. Locally, F5 drag folds in the S4-deformed phyl-litic chlorite schist are associated with this faulting with thefold axes displaying the same east plunge as F2a. Generally,the regional and deposit-scale dextral faults may be reactivated sinistral D4 faults. It is likely that many of the D4 faultsthroughout the Koolyanobbing greenstone belt have been re-activated during D5.

Paleostress field

A five-stage deformation (D1 to D5) sequence has been es-tablished (Fig. 13). The paleostress axes determination for D1

is based on a two-stage back rotation of the BIF layers arounda vertical axis to achieve a north-south trend (i.e., a pre-D2b

stage; cf. Chen et al., 2001, 2004); and a horizontal axis toachieve a horizontal orientation (pre-D2a stage). The pale-ostress axes estimation of D2a and D2b is based on a one-stageback rotation of the BIF layers around a vertical axis toachieve a pre-D4 stage north-south trend (Fig. 12). The in-ferred pre-D2b, north-south trend of the Koolyanobbing


0361-0128/98/000/000-00 $6.00 935

N-S compression or E-W extension withassociated dextral faults, locally reactivatingD faults.4

S2 F2 F1a/b

?

massive pyrite

?

D : N-S buckling1b

early crustal shortening

D : intrafoliation folding1a

deposition

Vertical compaction and F 1a

isoclinal folding, followed by F1b

open to tight folding. Low-gradebackground metamorphism

W (SW)E (NE)

W (SW)

E (NE)

BA

Deposition of footwallmafic rocks, massivesulfide(?), chert-magnetiteBIF, and hanging wallmafic rocks.

C

orogenic stage: compression

D

E

orogenic stage: transpression

S at an4

angle with S2 horiz-ontalF4HF2

F4

Koolyanobbingshear zone

F5

post-orogenic stage

D2b faults

E

W E

N S

F

D1a

x

x

x

x

x x

x

x

x

x

x

x

x

x

x

x

NW

F1b

F1a

growth fault?

60°

x

x

x x

x

x

x

x

x

xW E

imbricated fan structures (i.e., K deposit) and S-shapefolding (i.e., at C deposit). Formation of S 4.

Metamorphism in KGB: lower amphibolite-facies close tothe KSZ, greenschist facies in the middle BIF and iron oredeposits.

Formation ofanticlockwise

formation of large

KSZ and associated or subsequent rotationof the KGB to a NW trend. Internally in KGB,rotation of BIF units around vertical F , locally4

accommodated by NNW to NW trending faults.At the deposit scale duplex and

Progressive shearing forming brittle to ductile D2b

shear zones and breccias truncating F . Further2a

tightening of F and F .1 2a

Folding around regional horizontal F fold axis,2

tilting of BIF to a west-dipping steep orientation.Cleavage S formation. Local BIF thickening by2a

parasitic F .2a

W

talc-schist

hanging wall mafic rocks

footwallmafic rocks

chert-magnetite BIF

D : KSZ3

F2

F4

NW

N

N WN

F : vertical rotation4

5 km

K deposit

duplex

K depositimbricate fan

xspecularitemineralization

(D5 )

E1

E2

E3

S

N

S2b

S2a

FIG. 13. Generalized structural model of the Koolyanobbing greenstone belt (KGB) (at both regional and iron ore de-posits scale).



greenstone belt (Chen et al., 2001, 2004) represents an ex-treme anticlockwise rotation of the Koolyanobbing green-stone belt during D4 (Fig. 12B1, C1); an alternative is theminimum rotation around F4 (Fig. 12B2, C2). The true 1 σ trend that occurs during D1 to D4 probably has a position be-tween both end-member possibilities.

D1 paleostress: Compressional structures associated withD1 formed within an overall north to north-northwest com-pressional regime (Fig. 12C). A north-south compressionalregime during D1 has been observed throughout the South-ern Cross domain (Dalstra, 1995; Chen et al., 2001, 2004). Inthe low-grade metamorphosed BIF of the Marda greenstonebelt, which is located 100 km north-northwest of theKoolyanobbing greenstone belt, north-south contraction with associated folds is well observed in fold interferencepatterns in outcrop and at a regional scale (D1a and D1b of Dalstra, 1995; Dalstra et al., 1999; Chen et al., 2003). TheBIFs in the Marda-Diemals greenstone belt are character-ized by thicker mesolayers (>5 mm) and less intense internalD1 deformation than those in the Koolyanobbing greenstonebelt. This suggests that BIF in the Koolyanobbing green-stone belt experienced a more intense coaxial compression.The preservation of early structures is mostly due to less in-tense structural overprint during subsequent D2 and D3.This is not the case in the Southern Cross greenstone belt, where D2 folding obliterated D1 structures older than D2

(Bloem et al., 1994; Dalstra, 1995). Even though isoclinalfolding and stratigraphic thinning is observed in BIF and in-terpreted as a D1a feature, the finite strain during D1b com-pression was rather weak; evidence heretofore includes: (1)there is no evidence of early thrust tectonics in theKoolyanobbing greenstone belt during D1 north-south com-pression; (2) the amplitudes of regional-scale D1b folds arenot high, showing wavelength/amplitudes ratios of <10, oth-

erwise the BIF units would show a strong curvature; instead,the lower, middle, and upper BIF units are characterized by a roughly linear trend; and (3) the BIF displays just localizedinternal D1a folds, and the amplitudes appear to decrease with proximity to footwall and hanging-wall mafic rocks. It islikely that the north-south−directed shortening of theKoolyanobbing greenstone belt succession was accommo-dated by minor shearing, without thrusting, along the footwall and/or hanging-wall contacts, and by folding in the in-ternally layered BIF.

D 2a and D 2b paleostress: After one-stage back rotation, theD2a and progressive D2b event occurred during a compres-sional regime characterized by an east-west− to northeast-southwest−trending 1 σ axis (Fig. 12B). D2a is compatible with

a D2 that has been described for the Southern Cross domain(Bloem et al., 1994; Dalstra, 1995; Chen et al., 2003). D2b isonly expressed in the Koolyanobbing greenstone belt but hasnot been observed regionally. Because D2b involved only minor faulting, or reactivation of lithologic contacts duringprogressive deformation subsequent to D2a, it is still compat-ible with the deformation history for the remainder of theSouthern Cross domain.

D4 paleostress: The development of S4 in mafic rocks asso-ciated with sinistral D4 strike-slip deformation in BIF indi-cates overall lateral east-west− to northeast-southwest−di-rected horizontal shortening (Fig. 7E5), which coincides with

the stress field of the Koolyanobbing shear zone (Libby et al.,1991). The formation of the Koolyanobbing shear zone andsimilar transcurrent shear zones in the Southern Cross do-main has been assigned to a regional D3 in an east-west com-pressive stress field (Libby et al., 1991; Chen et al., 2001,2004). The Koolyanobbing shear zone was likely generatedduring an earlier stage of the D3 transpressional tectonics, whereas the discrete C' fabrics, which indicates deformationin the brittle-ductile transitional stage and the minor north-east-trending dextral shear zones that both overprinted theductile S-C fabrics (Libby et al., 1991) suggest prolonged de-formation during a cooler stage. This stage was probably co-eval with, or in a transitional phase to, the mainly cataclasticdeformation during D4.

The field stress trajectories for D2a, D2b, D3, and D4 aresimilar to each other, therefore, it is possible that D2a to D4

represents a progressive deformation sequence in an east- west−directed paleostress field, commencing with coaxialshortening during D2a and D2b, changing to transpressionduring D4, resulting in strike-slip deformation (Fig. 14). D3, which is characterized by ductile deformation (Koolyanob-bing shear zone), is preceded and followed by two mostly brittle deformation phases, D2b and D4, respectively. This in-dicates that there were separate deformation events. At thescale of the Koolyanobbing greenstone belt this timing prob-lem is not solvable due to the spatial separation of theKoolyanobbing shear zone (D3) and D2 and D4 structures inthe Koolyanobbing greenstone belt , and lack of absolute agedates other than the date for the apparently post-D3 LakeSeabrook granite (2656 Ma: Qiu et al., 1999). However, dat-ing of deformed and undeformed pegmatites in the LakeJohnston greenstone belt (south of the Koolyanobbing green-stone belt ) have been used to constrain the age of ductile D3

deformation associated with the Koolyanobbing shear zone,

and these ages suggest that D3 deformation, at least locally,did not cease before 2629 Ma (Joly et al., 2010). Gold-relatedD4 faults in the Southern Cross greenstone belt (50 km SW of the Koolyanobbing greenstone belt) have similar ages, sup-porting the suggestion of a progressive D3-D4 deformation sequence.

D5 paleostress: The late-stage D5 faulting reactivated earlierfaults and shear zones that display extensional or dextralstrike-slip movement. This suggests a late-stage north-southcompression, or probably east-west extension, and is compat-ible with observations in other parts of the Southern Crossdomain (Dalstra et al., 1999).

Metamorphic events

Peak metamorphism (syn-D 2a to syn-D4 ): Peak metamor-phism in the Koolyanobbing greenstone belt was related toemplacement of regional batholiths, which caused regional-scale contact metamorphism (Ahmat, 1986). Granite em-placement took place during the main orogenic east-westcompression phase (Gee, 1979), starting with the onset of D2. Close to the Koolyanobbing shear zone in the NorthRange, grunerite is an abundant metamorphic mineral inquartz-magnetite BIF, indicating that the upper greenschistfacies was reached. The Koolyanobbing shear zone sepa-rates amphibolite facies metamorphic-grade rocks (to thesouthwest) from greenschist to subamphibolite facies-grade


0361-0128/98/000/000-00 $6.00 936



domains (to the northeast; Libby et al., 1991), hence peakmetamorphism probably ceased during D3 transpressionaltectonics. Within the middle BIF unit (i.e., 1−3 km away from the Koolyanobbing shear zone) mafic footwall andhanging-wall rocks are characterized by a chloritic cleavage.This indicates a decrease in metamorphic grade from amphi-bolite (grunerite) to lower greenschist facies (chlorite zone)

with distance from the surrounding gneisses. Such a meta-morphic zonation has been described from several green-stone belts in the Southern Cross domain and is consideredto be the result of decreasing temperature gradients acrosscontact aureoles (Ahmat, 1986).

Lake Seabrook contact metamorphism (late to post-D4 ):The Lake Seabrook granite experienced syn- to postemplace-ment deformation; the biotite cleavage suggests upper green-schist facies metamorphism (T. Angerer, 2009, unpub. data).Quartz-cummingtonite BIF in the adjacent contact aureolesuggests a cummingtonite-hornfels facies contact metamor-phism (Klein, 1983, 2005).

Ore-Forming Stages and their Structural Control

Ore stage 1: Iron-rich carbonate and talc alteration(pre-D 2a )

Iron ore stage 1 was an early Fe-Mg(±Ca?) metasomatismthat locally altered quartz-magnetite BIF to Fe-rich carbonate-magnetite BIF, replacing silica layers (Fig. 15A). This carbon-

ate alteration is evident in the A, F (siderite), and K (ferroandolomite) deposits. Ferroan talc alteration is closely associated with carbonate alteration in the K deposit and may be the prod-uct of the same metasomatic event. The hydrothermally al-tered BIF is typically enriched in iron (~45−55 wt % Fe) with-out showing significant volume reduction, which implies thatiron was added to the BIF host rock by metasomatism.

In general, dolomite dissolution is facilitated by fluid flow under varied conditions, such as moderately elevated temper-ature of 100° to 250°C (Pokrovsky and Schott, 2001) and neu-tral to low pH (Zhang et al., 2007). Therefore, it is suggestedthat ore stage 1 “prepared” BIF for subsequent dissolution of


0361-0128/98/000/000-00 $6.00 937

deposition oflower succession

1st generationplutons: onlypreserved asrafts in 2ndbatholites

lower succession3010±7 (1)

x xx x

2nd generationbanded gneisses(Ghooli dome,Lake Deborahdome

Ghooli dome:2775±10 (2) to2691±7 (3)

Lake Seabrookgranite,intruded Ghoolidome,KSZ, and KGB

x xx xx xx x

D and D are main deformation2

phases throughout Southern Crossdomain, affecting all greenstone beltsand batholiths(i.e., forming

3

synkinematicbanded gneisses)

r e gi on al c ont a ct M.

N-Scompressiontranstension?D 4

relative ageconstrain

x xx xx xx x

x xx xx xx x

x xx xx xx x

am b i ent b ur i al M et am or ph i s m (

@

)

> 3 0 0 MP a

1 4 k m

reactivation

upright ductilefolding

ductile folding(and thrusting?)

l ow- gr a d ei ni nn er K GB ( < 5 k m )

m e d i um- gr a d ei n o u t er K GB ( 0 . 5 -2 k m )

r e g i o n a l o r

c o n t a c t

M .

strike-slip andreverse dip-slipfaults, vertical-axis rotation

l o c a l c o n

t a c t M .

( 0 - 5

0 m

a u r e o

l e )

d e n u d a t i o n

2.8 Ga

3.0 Ga

2.6 Ga

D - D 2a 2b


D5KGB


p l u t o n s 2 . 6 6 - 2 . 6 0 G a

p l u t o n s

2 . 8 - 2 . 6 7 G a

p l u t o n s 2 . 9 9 - 2 . 9 2 G a

Southern Crossgold skarns(Mueller & McNaughton, 2000)

2620+-6 (5)

r e s i d u a l m a g n e t i t e

s p e c u l a r i t e

m a g n e t i t e m i n e r a l i z a t i o n

age dating:1) Pidgeon, 1990 (U\Pb on Zr in Fsp-porphyry sill)2) Mueller & McNaughton, 2000 (U/Pb on Zr)3) Dalstra et al., 1998 (U/Pb on Zr)4) Qiu et al., 1999 (U/Pb on Zr)

5) Bloehm et al., 1995 (U/Pb on Zr)6) Joly et al., 2010 (U/Pb on Zr)

E-Wcoaxialcompression

N-S or NW-SEcompression

E-Wtranspression

D1a

D4

D2a

D1b

D2b

D KSZ3

D1KGB

relative ageconstrain ductile

recumbent folding(and thrusting?)

britte-ductilereverse faulting

1 4 k m d e p t h

( 3 0 0 M P a )

ages and timingdeposition / magmatism

metamorphism deformationvery low highlow

time-scale

stress fieldiron ore genesis(and gold for comparison)

2 n d m a r t i t e

g o e t h i t euplift

recent

D1regional

1 s t m a r t i t e

~250 Ma

D3


D5

Lake Seabrookgranite: 2656±3 (4)

D :2629±1 (6)3

FIG. 14. Synthesis of the geologic evolution of the Koolyanobbing greenstone belt (KGB), including relative timing of theiron ore stages in relationship to the regional tectonometamorphic and magmatic stages in the central Southern Cross do-main.




0361-0128/98/000/000-00 $6.00 938

boudinage:layer parallelextension

talc-altered folds withoutsignificant gangue removal

mag-layer

qtz/tc/carb-layer

void

process: mass transfer by gangue removal mostly from fold limbs

low strain

D - to D -associated tightening of

folds: partitioning of material2 4

Gangue mineral leaching

within fault:pure specularitemineralization(K deposit)

fold scale: decimeter to decameter

core is talc-carbonate-quartz rich

residualmagnetite

mineralized layer

main fault withsilicious breccia

late stage S4

process: breccia matrix and BIF replacement

controllingstructure

mineralizedbreccia

NW SE

present-daygroundwaterlevels

saprolite

canga

gossan

any steep structurecan enhance depthof weathering

(vertically exaggerated)

BIF detritus

folds controlmagnetite

enrich-ment

main fault controls ironore enrichment

martite BIF

leached martite BIF(medium-grade)

magnetite BIF

pseudomorphic goethite-martitemineralized BIF (medium-grade)

any steep structure cancontrol deep weathering

(vertically exaggerated)

specularite-bearing BIF

medium-gradesatellite features

massivemartite ore

magnetite ore

pseudomorphicgoethite-martite ore

vitreous goethite ore (Al+Si enriched)

vertical zoning in BIF(unaltered and carbonate-altered)

specularite-bearing ore

vitreous goethite ore (Al+Si enriched)

ore stage 2: residual magnteite by gangue leaching

ore stage 3: magnetite mineralization

Ore stage 4: specularite mineralization

(quartz or carbonate)

B

C

D

Ore stage 5: weathering-related modifications

vertical zoning in ore

F2

mineralized breccia

mineralizedminor f ault

mineralizedminor structures

E

g o e ±

m a r

q u a r t z

m a r t i t e

m a r ± g

o e

mar±goe layer

goe

goe±mag±mar goe

mar

mag

qtz

10 m

compare with examplein Fig. 3F

Goethite replacement of gangue and iron oxides Martitization

former qtz-layer

mar±goe layer

carbonatealteration

NW SE carbonatealteration

ore stage 1: carbonate alteration (talc-alteration?) A

F2

mineralized breccia

BIF clasts

altered la yer

process: replacement of quartz-layers

qtz

mag

Fe-carb

siderite-magnetite BIF

process: breccia matrix and BIF replacement

m a g + m

a r + g o e

Processes:

? ?

?

?

t a l c

fluid-feederstructure ?

1 cm

brittle Dstructures

controlspecularite

4



carbonate gangue during ore stage 2. Since dolomite deformseasily under moderately hot (<300°C) and wet conditions(Newman and Mitra, 1994), early stage of carbonate alter-ation was probably also important for strain localization of shear zones and faults in which magnetite and speculariteprecipitated during ore stages 3 and 4, respectively.

The structural control of the Fe-rich carbonate ± talc alter-ation is difficult to reconstruct because subsequent deforma-tion (D2-D4) and alteration were so intense that the originalgeometry of the alteration zone has been obliterated. Thetiming of ore stage 1 may have been as early as diagenetic,and some of the talc-rich rocks are probably related to hy-drothermal activities at or near the seafloor, which may havealso led to the formation of the massive pyrite body (cf. Costaet al., 1980, 1983). At the small scale, a syngenetic depositionof ferroan talc and magnetite, leading to intercalated talcschist, talc-magnetite BIF, and medium- to high-grade mag-netite ore, may have occurred. This syngenetic genesis may explain talc-enriched fold cores and talc-depleted fold limbsat a small scale. However, quartz in transitional talc-carbon-ate-quartz BIF (i.e., in zones of incomplete replacement)shows predominantly carbonate and ferroan talc overgrowingrecrystallized textures and this suggests that replacementpostdated dynamic recrystallization during D1 metamor-phism. The existence of deformed and undeformed carbon-ate and talc veins, syndeformational strain shadows at mag-netite, juxtaposition of different talc and carbonategenerations in talc-schists and late-stage carbonate breccias,respectively, point to a complicated multistage alteration his-tory. A detailed investigation of the genesis of talc and car-bonate alteration is the subject of ongoing research.

Ore stage 2: Residual magnetite enrichment (D 2a-D4 )

Iron ore stage 2 includes formation of laminated magnetite

ore and magnetite ore breccia (Fig. 15B). The known mag-netite orebodies in the A and K deposits are located predom-inantly in tight D1b and/or D2a folded zones. These controllingfolds strike subparallel to the BIF unit, i.e., north-northwestto west-northwest and plunge subhorizontally (in the A de-posit) to moderately steeply (in the K deposit). Talc-schistsand medium-grade talc-carbonate-magnetite BIF occur fromthe meter to decameter scale surrounding medium- to high-grade magnetite ore in the K deposit. The appearance of magnetite ore is patchy in the talc zone but prominent in thecenter of the K deposit, where ore is enveloped mostly by quartz- and carbonate-magnetite BIF. High-grade magnetiteore is also prominent in the A deposit, often enveloped by asiderite-rich BIF protolith. As a consequence, it is proposed

here that formation of the larger bodies of high-grade mag-netite ore was a result of removal of carbonate and, to a minorextent, quartz.

Gangue removal was by selective mobilization of preexist-ing carbonate, quartz, and minor ferroan talc by solution orminor mechanical transfer, observed in high mean stresszones such as tight fold limbs or faults that truncate folds at various scales. The location of fibrous talc, granoblasticquartz, and sparitic ferroan dolomite in strain shadows of boudin necks (talc: Fig. 6C) supports the suggestion that thegangue was mobile during deformation. Deformation-con-trolled solution transfer of quartz during layer-parallel boudi-nage associated with tight folding of quartz-calcite-hematiteBIF has been described by Hippertt et al. (2001). Similarprocesses may have occurred in the Koolyanobbing depositsunder fluid conditions favoring carbonate dissolution, such asmoderately elevated temperatures of 100° to 250°C(Pokrovsky and Schott, 2001) and neutral to low pH (Zhanget al., 2007). Magnetite remained mostly in place during thisprocess, suggesting that fluids were iron saturated, which would inhibit iron oxide dissolution. As a consequence of gangue removal, microporosity increased and caused mag-netite microlayers to disintegrate forming the typical fine-grained collapse zones intercalated with the mechanically more stable magnetite mesolayers in laminated and brecciaore (Fig. 2A, B). Locally, the entire BIF texture disintegratedduring gangue removal and ore brecciated textures formed.

Microscale boudinage fabrics (Fig. 2B) are observed inboth BIF and laminated magnetite ore. These fabrics weregenerated by competency contrasts between mineralogically distinct layers but are rather unlikely in monomineralic rocks where no competency contrast between layers exists, such asa syngenetic, metamorphosed magnetite ore (i.e., chert-freeBIF: Lascelles, 2006a, b). Thus, gangue removal occurred at

least during, probably later than, layer fragmentation in foldlimbs and locally fold hinges. Talc in alteration zones sur-rounding magnetite orebodies may be explained by the reac-tion of Mg in solution from mobilized Fe-Mg carbonate withresidual quartz in less-deformed magnetite BIF (cf. Klein,1974; Costa et al., 1980; Moine et al., 1989). In transitionaltalc-carbonate-quartz BIF this reaction is “halfway frozen” intime. A talc-forming reaction would imply lower to medium(about 300°−400°C) conditions such as greenschist faciesmetamorphism (Klein, 1974). Considering that controllingfolds were formed during synmetamorphic F2a, regional-scalemetamorphism may be a relevant process. Similar conditionsmay also be generated by localized, high-temperature(>300°C) hydrothermal activity unrelated to metamorphism.


0361-0128/98/000/000-00 $6.00 939

FIG. 15. Ore-forming stages and enrichment processes. A. Ore stage 1: carbonate (and talc?) alteration of siliceous BIF.B. Ore stage 2: Residual magnetite enrichment by carbonate removal, controlled by F 2 and tightened F1 fold limbs andhinges and probable associated talc alteration. C. Ore stage 3: Magnetite mineralization of BIF, magnetite ore, and brecciasproximal to, or within, stage 2 magnetite ore. D. Ore stage 4: Specularite in faults, fractures, reactivated fold cores, brecciapods, BIF, and in stages 1 and 2 ore. Synchronous with specularite is a localized first stage of martitization of magnetite oreand quartz-Fe-carbonate alteration. E. Ore stage 5: Three processes occurred: (1) leaching of gangue in martite BIF, form-ing high-grade ore; (2) goethite precipitation in the vadose zone, either by replacing quartz, carbonate, and iron oxides in un-altered and carbonate-altered BIF, and medium-grade stage 2-4 ore, or by precipitation in macroporosity in leached BIF;clay- and silica- enriched vitreous goethite hard cap in the uppermost zone; (3) martitization of magnetite ore to martite oreand magnetite BIF to martite BIF.



In any case, the D2 event most likely represents the maximumrelative age for gangue removal and hence for magnetite oregenesis. A continuation of ore stage 2 during postmetamor-phic D2b and D4 fold reactivation is suggested by the intersti-tial microporosity in and between anhedral magnetite grainsand the nearly complete lack of amalgamation of magnetitegrains. These textural observations are incompatible withmetamorphic overprint and deformation. The minimum rela-tive age for the magnetite ore formation in the K deposit isconstrained by the onset of D4 based on (1) boudinage of magnetite bodies and shearing of adjacent talc-schists as aconsequence of the reverse and strike-slip faulting thatshaped the eastern lithon, (2) development of a discontinuouspressure solution cleavage that overprinted magnetite ore andeuhedral magnetite (Fig. 2C; i.e., formation of granular mag-netite ore), and (3) localized hydrothermal specularite ±quartz ± ferroan dolomite ± pyrite ± talc ± chlorite assem-blage, which replaced the magnetite ore.

Ore stage 3: Magnetite mineralization in brittle (-ductile) structures (D 2a-D4 )

Ore stage 3 includes formation of magnetite in brittle andbrittle-ductile structures (such as D2b faults, fractures, andbreccias) in reactivated D2a fold cores, and locally in wall rockand quartz veins adjacent to mineralized brittle structures(Fig. 15C). Structures that control ore stage 3 mineralizationare (1) in the K deposit, the boundary fault between massivesulfide and quartz-, and carbonate-magnetite BIF (Fig. 4); (2)in the A deposit linear, BIF unit-parallel, fault breccias trun-cating BIF and mafic rocks (Fig. 9); (3) a linear, BIF unit-par-allel, breccia body likely controlling the F deposit and smallerbodies of BIF breccias proximal to this deposit; (4) irregularly shaped breccias proximal to the high-grade ore in the C de-posit (Fig. 10); and (5) meter-scale fractures in BIF proximal

to high-grade ore-forming zones of mineralized BIF (in alldeposits, best displayed in the D pit). Magnetite mineraliza-tion adds only a minor fraction to the overall ore formation, asit is restricted to the brittle structures and bordering wallrocks. The main process of magnetite formation during orestage 3 involved replacement of quartz and localized carbon-ate (Fig. 15B2).

The granular magnetite ore in the K deposit (Fig. 2C) is in-terpreted to have been upgraded during ore stage 2 but mod-ified by minor deformation and mineral growth during orestage 3. In magnetite BIF in the K deposit only minor andcentimeter- to decimeter-scale brittle-ductile shear zones ormineralized fold cores contain stage 2 magnetite mineraliza-tion. The local overprint of magnetite and/or martite crystals

by a weak pressure solution cleavage and the replacement of specularite indicate that D4 represents the minimum age forore stage 3.

Mineralized D2b breccias in proximity to goethite-martitehigh-grade ore are commonly medium grade, due to elevatedquartz content in the matrices and/or clasts of BIF or maficrock. The textures of these mineralized zones are typically produced by subhedral to euhedral magnetite and/or martitecrystals that are partially amalgamated (Fig. 2D) and form ce-ments around BIF clasts (Fig. 15B4). The mineralized matri-ces locally replaced a primarily silica-rich breccia matrix (Fig.15B5).

The densely distributed euhedral martite that commonly occurs in goethite matrices in goethite-martite ore in small- tomedium-scale orebodies (i.e., the C, D, E, and F deposits)implies a two-stage iron oxide formation: sub- to euhedralmagnetite growth replacing siliceous breccia matrices andBIF layers, followed by partial to complete goethite replace-ment of magnetite and/or martite and upgrade by goethite re-placement of remaining quartz.

Ore stage 4: Specularite mineralization and first martitization (synpost-D4 )

Specularite mineralization: Specularite mineralization, con-trolled by D4 faults, is observed in the K, C, and D depositsand defined as ore stage 4 (Fig. 15D). Based on analyses of the fault and/or fracture pattern of the Koolyanobbing green-stone belt, it is likely that the major controlling structures arelarge strike-slip faults with a north-northwest to west-northwest trend. Deposit-scale examples include the strike-slipfaults in the K deposit, such as the footwall and horse faults of the duplex system, and strike-slip and reverse shear zones inthe eastern lithon (Fig. 3). Directly north of the K deposit,outside the strike-slip duplex, specularite is also present, mostlikely controlled by the west-northwest− to north-northwest−striking footwall fault. Specularite in the C deposit is lo-cated within a northwest-striking fault at the limb of a large-scale S-shaped drag fold in that accommodated verticalfolding or in faults that caused imbrications at the D deposit.Specularite formation is locally associated with hydrothermalquartz ± ferroan dolomite ± pyrite ± talc ± chlorite alterationand is most significant in the K deposit within reactivatedbreccia pods, brittle faults, and tension gashes, which allcrosscut D1 and D2a folds. Locally, tightened or faulted D1

and/or D2a fold cores are also loci for a hydrothermal over-print, and replacement of neighboring wall-rock BIF by hy-

drothermal assemblages is common. The precipitation of specularite-bearing hydrothermal assemblages in BIF wasnot an effective ore upgrading process, and in magnetite orecarbonate-specularite veins and pods replacing magnetite orecaused a downgrading. However, carbonate replacement of quartz in siliceous BIF was an important ground preparationfor weathering-related leaching during ore stage 5 (seebelow).

The A, B, E, and F deposits do not show any specularite orassociated hydrothermal overprint, although they are trun-cated by brittle faults. It is, therefore, interpreted that not allfootwall faults are coeval with specularite formation or werenot hydrologically connected with structures that acted asfluid feeder. For example, the mineralized faults in the A de-

posit are interpreted to be reverse slip fault that formed dur-ing D2b and controlling ore stage 3 but were inactive duringore stage 4.

Specularite formation is characterized by the replacementof gangue in laminated microspecularite-martite ore andmedium-grade BIF and by coarse crystalline massive and dis-seminated specularite in brittle faults, breccias, and voids.Disseminated specularite and specularite (±carbonate ±quartz) veins overprint stage 1 and 2 magnetite ore. Most hy-drothermal, ore stage 4-related alteration is late to postdefor-mational, i.e., it took place after D4-related brittle deforma-tion ceased, probably during the relaxation phase of faulting.


0361-0128/98/000/000-00 $6.00 940



This is indicated by the lack of internal deformation in thelate-stage hydrothermal alteration assemblage. Deformedtranstensional gashes in the eastern lithon with crystal-plasticcarbonate and chlorite deformation indicates a localized syn-deformational alteration.

Martitization: Martitization in wall-rock magnetite associ-ated with specularite veins is minor. Existing juxtaposition of martite against magnetite ore along steep D5 faults at the Kdeposit suggests an earlier stage of martitization with regardto faulting. Hydrothermal martitization has been described inseveral hypogene ore deposits (Lobato et al., 2008). It is thusinterpreted that localized, structurally controlled, martitiza-tion related to specularite has been overprinted by ubiquitousmartitization related to weathering.

The spatial relationship of martite with a hydrothermalmineral assemblage suggests that martitization commencedduring the hydrothermal alteration. There is no evidence forgangue leaching taking place during ore stage 4 nor for a preferred replacement of Fe-rich talc or carbonate in the Kdeposit.

Controls by post-D4 deformationSubsequent to the D4 strike-slip fault movement and con-

current hydrothermal ore stage 4, deformation took place without associated ore formation (i.e., upgrading) processes.During D5, preexisting structures such as D4 faults were re-activated during north-south compression or east-west exten-sion and (further) block rotation around a vertical axis may have taken place. It appears that magnetite ore and specular-ite-overprinted magnetite orebodies did not experience D5

block segmentation. However, D5 faulting led to localized re-orientation and displacement. For example, dextral move-ment along the footwall (reactivating sinistral D4 faults) andhanging-wall contact at the K deposit caused decoupling of

the western lithon from its imbricate fan basal fault (Fig.13E3). The southern end of the A deposit is defined by a D5

fault crosscutting the middle BIF unit, which juxtaposedmagnetite ore (now mostly goethite-martite ore) with quartz-martite BIF. This fault likely originated earlier and was an im-portant fluid pathway causing siderite alteration of the BIFprior to residual magnetite ore enrichment.

Ore stage 5: Supergene modifications

Fluids responsible for ore stage 5 are interpreted to be of supergene origin (cf. Clout, 2003), which is supported by atopdown zoning of orebody characteristics. Two types of weathering profiles are developed in the middle BIF: a thickzone of goethite-martite ore in the upper part underlain by a

deep zones of stage 2 to 3 ore, which is characteristic for thelarger iron ore deposits; and a thin zone of goethite ore un-derlain by oxidized, siliceous or slightly hydrothermally al-tered BIF (Fig. 15E). The main zones of the weathering pro-files are, from the base to the top, a magnetite BIF or orezone showing no weathering, followed by a martite zone, which indicates partial to complete gangue leaching (i.e., up-grade in BIF) and minor leaching of massive martite (aftermassive magnetite ore), followed by a goethite-rich martitezone, which shows partial to complete pseudomorphicgangue replacement of carbonate ± siliceous BIF or brecciasand goethite replacement of martite forming goethite-martite

ore, followed by a massive to vuggy, clay- and secondary sil-ica-rich, goethite zone, which represents the hard cap. Thiszoning is similar to vertical depth profiles of supergene (i.e., weathering-related) modifications in iron ore deposits in theHamersley basin (Clout, 2003). The maximum age of this weathering-related overprint in the Koolyanobbing green-stone belt may be as old as Permian, corresponding to re-golith ages throughout the central Yilgarn craton (Pillans,2000, 2004; Anand and Paine, 2002).

Gangue leaching: A second stage of removal of carbonate ±quartz from medium-grade ore of stages 2, 3, or 4 at depthled to an upgrade to residual high-grade ore. These weather-ing-related leached ores are typically martite rich. Locally,gangue leaching is associated with volume loss of the orebody,as the voids that were produced are subject of vertical col-lapse and horizontal contraction. However, leached martiteore is observed that shows no collapse texture. This macrop-orous martite ore is readily distinguishable from martitizedmagnetite ore of ore stage 2, because the latter has a charac-teristically massive laminated texture with no macro- but amicroporosity in the cataclastic layers.

Pseudomorphic goethite mineralization: Goethite ore ischaracterized by textural modification of the BIF layering andshows clearly replacement of magnetite, martite, and specu-larite. Goethite replacement affected rocks to 70 m below thepresent surface, with deeper weathering zones observed insteep fault zones.

Martitization: Textural similarities between massive mag-netite and martite ore and common relics of magnetite inmartite grains suggest that, principally, massive laminatedmartite ore in the Koolyanobbing greenstone belt representspseudomorphically altered massive magnetite ore of ore stage2. Martite is commonly interpreted as a weathering-relatedphase (e.g., Morris, 1980, 1985); however, as mentioned be-

forehand, some martite formed already during earlier D4-re-lated fluid-rock interaction. The localized nature of magnetitecrystals in goethite matrix next to completely martitized mag-netite crystals in BIF suggests that martitization ceased whensurrounded by a massive “mantle” of goethite. Magnetite ingoethite-magnetite-martite ore is locally described from sur-face outcrops at the Mount Gibson deposit (Lascelles, 2006b)and may have been preserved by a similar process describedhere.

Discussion

The relative significance of ore-forming stages andBIF alteration

The intensity of BIF alteration and hypogene ore stages 1to 4 varies throughout the Koolyanobbing greenstone belt,and there is a positive correlation between deposit size andnumber of developed BIF ore stages within the deposit. It isevident that in all BIF-hosted deposits within the Koolyanob-bing greenstone belt, at least one of the four hypogene (i.e.,deep-seated) ore stages occurred. The importance of theresidual magnetite ore formation during ore stage 2 in termsof generating ore volume or tonnage is unclear, because weathering-related gangue leaching may have overprintedareas affected by this ore formation stage. The significance of magnetite mineralization associated with ore stage 3 for the


0361-0128/98/000/000-00 $6.00 941



formation of the high-grade orebodies at Koolyanobbing re-mains contentious. Despite common occurrence of euhedralmagnetite and/or martite in BIF proximal to ore zones, de-posit-scale high-grade magnetite and/or martite ore thatformed during stage 3 mineralization is not identified. How-ever, medium-grade martite ore in the K deposit and high-grade goethite-martite ores in the D, E, and F deposits show euhedral magnetite and/or martite (in case of goethite-mar-tite in a goethite matrix), which originated from magnetitegrowth during ore stage 3. Ore stage 4 developed only locally,but it was a very effective iron upgrading process. At the Kdeposit the high density of duplex and imbrications faultscaused a pervasive specularite alteration in BIF and brecciasin the lithons (wall rock) between faults, resulting in (micro-)specularite-martite ore to be volumetrically very important.The dominant occurrence of goethite-martite ore (approx80%) throughout the Koolyanobbing greenstone belt iron oredeposits is a result of extensive surface-related (supergene)processes that took place during ore stage 5. Supergene mod-ification and upgrade of ore stage 1 carbonate-altered BIF islikely to be an economically important process in the evolu-tion of the iron ore deposits in the Koolyanobbing greenstonebelt. This is in accordance with observations of early parage-netic stages of carbonate in several Archean and ProterozoicBIF-hosted high-grade iron ore deposits (Barley et al., 1999;Taylor et al., 2001; Dalstra and Guedes, 2004; Gutzmer et al.,2005; Lascelles, 2006b; Thorne et al., 2008).

The commonly observed martite and goethite replacementof magnetite and specularite ore and the spatial coincidenceof replacement fronts of ore stage 3 magnetite and/or martiteand ore stage 5 goethite pseudomorphs (Fig. 15B1, B2) sug-gest that residual magnetite, magnetite-mineralized, andprobably also specularite-mineralized zones were importantprecursors for the localization of late-stage martite and

goethite replacement of magnetite and/or martite and quartz.This is petrographically supported by the frequently observedpreferred nucleation of goethite on magnetite and/or martiteand specularite crystals (Fig. 2D).

Fluid flow

Fluids parental to carbonate (and talc?) alteration duringore stage 1 BIF must have been hot, Fe + Mg + CO2 /CO3

rich, and silica undersaturated in order to replace all quartz with BIF-hosted carbonate. Magmatic fluids, sourced frombatholiths intruded during the D2 event (i.e., the Ghoolidome; Ahmat, 1986) and likely enriched in Fe + Mg duringchanneling through the surrounding mafic rocks (i.e., mixing with metamorphic fluids), may have been a suitable fluid

source.It is assumed for ore stage 2 that deformation took placeduring infiltration of carbonate-undersaturated, hydrother-mal fluids through strained zones in BIF units, with prefer-entially upgraded BIF zones that were subject to early car-bonate alteration. The main feeder channels for these fluidshave not yet been detected; they may be D2a thrusts or localD2b reverse faults. The fluids that caused gangue leachingduring ore stage 2 were considerably different from the fluidsthat led to ore stage 1 carbonate (and talc?) alteration in theBIF. However, also the fluid-rock interaction associated withore stage 2 coincided with emplacement of granitoids and

associated metamorphism. Orogenic deformation, graniteemplacement, and contact metamorphism of mafic rocks inthe Koolyanobbing greenstone belt lasted several tens of mil-lions years (Fig. 14; cf. Pidgeon et al., 1990; Dalstra et al.,1998; Qiu et al., 1999; Mueller and McNaughton, 2000). Con-sidering the length of this fluid-forming event, such a changein physicochemical fluid conditions is probable and may beinduced by heat changes, varied fluid mixing, or introductionsof other fluid types such as meteoric water. Typical silica sat-uration in metamorphic water may have been the cause forpreferred carbonate dissolution, whereas strongly foldedquartz-magnetite BIF in the Koolyanobbing greenstone beltand in other greenstone belts, such as the Southern Crossgreenstone belt, did not experience leaching during meta-morphism.

Several contrasting interpretations of magnetite mineral-ization are possible: (1) iron oxide precipitation may havebeen a response of Fe carbonate dissolution nearby duringore stage 2, considering that iron-saturated fluids were re-sponsible for selective gangue leaching; (2) the hydrothermalfluid event that caused magnetite mineralization is indepen-dent of ore stage 2 and occurred syn-D2b /pre-D4; and/or (3)replacement textures of magnetite in D2b breccias do not nec-essarily imply coeval mineralization and deformation, there-fore the mineralizing fluid event may have been genetically related to the subsequent hydrothermal specularite formingore stage 4. An evolution of the fluid characteristics fromlesser oxidizing (magnetite) to more oxidizing (specularite) would have taken place.

Fluids related to the specularite alteration stage 4 may havebeen channeled by major north-trending D4 faults close tothe K deposit (Fig. 1). These brittle structures crosscut theKoolyanobbing shear zone and truncate 2656 Ma old LakeSeabrook granite. It is possible that the brittle deformation

stage channeled magmatic fluids from this granite. The post-ductile deformation and late to postmagmatic timing of thespecularite stage in the Koolyanobbing greenstone belt allow speculation about the synchronicity with hydrothermalquartz-hematite veins (Dalstra et al., 1999) and iron ore-re-lated fluid flow some 50 km to the west, close to Mount Cor-rell in the northern Southern Cross greenstone belt. A minoramount of gold has been exploited at Koolyanobbing (cumu-lative 2.6 Kt until 1951; source: Minedex, DMP), mostly hosted in mafic rocks in the footwall of the middle BIF. Thus,there may be a genetic relationship between specularite-quartz-carbonate alteration and gold-bearing hydrothermalfluids in the central Southern Cross domain (Fig. 14).

Further controls on high-grade iron ore

We have shown that iron ore formation is controlled by de-formation-controlled processes. This structural control is re-quired to explain the absence of exposed iron ore in the lowerand upper BIF and furthermore in many other greenstonebelts in the Southern Cross domain. Generally, the localiza-tion of one or more of the following geologic features andprocesses have been discussed in the literature to control thespatial distribution of processes of BIF-hosted high-gradeiron ore formation: (1) changes in the depositional facies of BIF, particularly carbonate facies BIF (Beukes and Gutzmer,2008) or high iron/silica ratio; (2) hydrothermal alteration of


0361-0128/98/000/000-00 $6.00 942



the gangue minerals, i.e., hydrothermal carbonate replace-ment of quartz, as an important stage forming a proto-ore thatis more amendable to dissolution than quartz during subse-quent hypogene or supergene upgrade processes (Barley etal., 1999; Chown et al., 2000; Taylor et al., 2001; Dalstra andGuedes, 2004; Thorne et al., 2004, 2008; Lobato et al., 2008);(3a) structures 1: the existence of brittle or brittle-ductilestructures as a ground preparation for a channeled infiltrationof certain fluids that lead to iron oxide mineralization and/organgue leaching (Dalstra and Rosière, 2008); (3b) structures2: active deformation-like folding, fold reactivation, or variousstyles of brittle (-ductile) faulting, creating pressure gradients within the structures leading to iron oxide mineralization inlow mean stress zones (Rosière et al., 2008) and/or to selec-tive removal of gangue in high-strain zones (this study); and(4) exposure to the surface leading to weathering of initially nonaltered BIF (Morris, 1985; Ramanaidou and Morris,2009) or hydrothermally altered (e.g., carbonate-replacedBIF) or hypogene iron ore.

The contacts between BIF (gangue carbonate and talc) andlaminated ore (almost free of gangue) are often perpendicu-lar to bedding, which shows that the massive layered mag-netite rock did not form during sedimentation. However, re-moval of gangue minerals may have started at an early paragenetic stage, such as during diagenesis, prior to foldingevents and metamorphism. Although the stratiform distribu-tion of magnetite lenses supports a chert-free iron formation(cf. Lascelles, 2006b, 2007a, b), the fold-controlled nature of early magnetite ore, showing porous, cataclastic, microtex-tures, indicates a synorogenic, late to postmetamorphic, up-grade of the iron ore by gangue removal.

Besides the structural framework that controlled the loca-tion of iron ore at the deposit scale, the conjunction of spe-cific regional-scale geologic elements and processes was a key

combination that facilitated an effective and sequential up-grade of BIF to iron ore at Koolyanobbing. These key ele-ments and processes include: (1) an originally high iron con-tent in BIF, (2) localized preore hydrothermal alteration of BIF, (3) progressive compressive deformation characterizedby fold reactivation and brittle deformation, and (4) long-last-ing syn- to postdeformation hydrothermal fluid flow favoringgangue leaching, channeled by large structures. It is note- worthy that in other greenstone belts throughout the South-ern Cross domain that lack one or more of these features,high-grade iron ore deposits did not develop. For instance,BIF in the Southern Cross belt experienced very intense duc-tile deformation during D2 (Bloem et al., 1994; Dalstra,1995), however, here, the silicate-rich BIF (iron phase is pre-

dominantly amphibole) lacks any apparent early carbonate al-teration and did not develop synorogenic magnetite ore. Syn-deformational metamorphism may also play a significant rolein ore formation, if selective mobilization of (dolomitic)gangue, while iron oxides remain in the solid phase, is facili-tated within a distinct temperature range (low temperatures,?100°−250°C: Pokrovsky and Schott, 2001).

Conclusions

This study establishes a deformation history and a multi-stage structural control for the magnetite, martite, specular-ite, and goethite ore in the Koolyanobbing greenstone belt. A

five-stage deformation sequence has been established, whichis mostly in accordance with structural models proposed forthe Southern Cross domain (Dalstra et al., 1999; Chen et al.,2003). The formation of medium- to high-grade iron ore isclosely associated with Late Archean stages of deformationand associated hydrothermal fluid flow. The stages includecarbonate (and talc?) alteration during Fe-Mg ± Ca metaso-matism (ore stage 1), leaching associated with ductile to brit-tle deformation syn- to postgreenschist facies metamorphismor mesothermal metasomatism (ore stage 2), and magnetiteand specularite mineralization controlled by brittle deforma-tion and associated hydrothermal alteration (ore stages 3 and4, respectively). Hydrothermal ore formation was followedafter a long period of uplift by weathering-related upgradefrom the Permian to recent times (ore stage 5). Hydrothermalore formation during the Late Archean at Koolyanobbing istherefore one of the oldest known, structurally controlled ironore deposits in the world. Fluid sources are still contentiousand the subject of ongoing research, but the importance of magmatic and metamorphic fluids for carbonate alteration,subsequent leaching, and magnetite and specularite mineral-ization is suggested by the proposed timing relationship withmagmatic and tectonometamorphic processes in the country rocks.

Permian to recent weathering-related upgrade by martitiza-tion, gangue leaching, and goethite replacement in ore deposits was in most deposits spatially related to existing Archeanmagnetite and specularite ± martite-rich ore. This implicationhas a significant impact on the understanding of the super-gene-modified hypogene iron ore types, because it suggeststhat the localization of supergene enrichment processes may be preferably controlled by early stages of medium-grade oreformation, rather than by structures in unaltered BIF. Puresupergene ore (i.e., massive, vitreous goethite hosted in BIF)

are minor in the Koolyanobbing greenstone belt.The results of this study may influence exploration for ironore in Archean BIF, because the strong structural control oniron ore distribution that has been described for theKoolyanobbing greenstone belt implies the general necessity of evaluating the structural features and evolution of an ex-ploration district to fully understand its prospectivity. Theclose spatial relationship of the surface-related upgrade by goethite precipitation and gangue leaching to existingmedium- to high-grade magnetite/martite/specularite ore in-dicates the possibility of existing blind magnetite- and/orspecularite-rich orebodies within the BIF units. The martiti-zation during specularite stage may also have produced local-ized high-grade specularite-martite ore. In order to delineate

these blind orebodies, an evaluation of the prospectivity of structures in the BIF unit, combined with application of geo-physical surveys, especially gravity and magnetic inversion,are essential. The occurrence of (ferroan) talc and carbonateproximal to high-grade ore, as observed in the A and K de-posits, may be utilized as footprint indicators for high-grademagnetite-, martite-, and specularite-rich ore.

Acknowledgments

This work has been funded and logistically supported by Cliffs Natural Resources Asia Pacific Iron Ore Ltd. The project was initiated by Nick Payne and Dave Fielding whose


0361-0128/98/000/000-00 $6.00 943



foresight and trust in using applied science as an aid for ex-ploration is highly appreciated. The authors are grateful forthe support from Cliffs Exploration/Resource Developmentteam, especially Nick Payne, Camilo Guarin, Jr., Dave Field-ing, Nigel Maund, and Kaye Hodgkiss. We very much appre-ciate the thorough reviews of the manuscript from SEG re- viewers Hilke Dalstra and Jens Gutzmer.

REFERENCES

Ahmat, A.L., 1986, Metamorphic patterns in the greenstone belts of theSouthern Cross province, Western Australia: Geological Survey of WesternAustralia Professional Paper 19, p. 1−21.

Anand, R., and Paine, M., 2002, Regolith geology of the Yilgarn craton, West-ern Australia: Implications for exploration: Australian Journal of Earth Sci-ences, v. 49, p. 3−162.

Barley, M.E., Pickard, A.L., Hagemann, S.G., and Folkert, S.L., 1999, Hy-drothermal origin for the 2 billion year old Mount Tom Price giant iron oredeposit, Hamersley province, Western Australia: Mineralium Deposita, v.34, p. 784−789.

Beukes, N.J., and Gutzmer, J., 2008, Origin and paleoenvironmental signifi-cance of major iron formations at the Archean-Paleoproterozoic boundary:Reviews in Economic Geology, v. 15, p. 5−47.

Bloem, E.J.M., Dalstra, H.J., Groves, D.I., and Ridley, J.R., 1994, Metamor-

phic and structural setting of Archean amphibolite-hosted gold depositsnear Southern-Cross, Southern-Cross province, Yilgarn block, Western-Australia: Ore Geology Reviews, v. 9, p. 183−208.

Cassidy, K.C., Champion, D.C., Krapez, B., Barley, M.E., Brown, S.J.A.,Blewett, R.S., Groenewald, P.B., and Tyler, I.M., 2006, A revised geologi-cal framework for the Yilgarn craton, Western Australia: Geological Survey of Western Australia Record 2006/8, p. 8.

Chen, S.F., Libby, J.W., Greenfield, J.E., Wyche, S., and Riganti, A., 2001,Geometry and kinematics of large arcuate structures formed by impinge-ment of rigid granitoids into greenstone belts during progressive shorten-ing: Geology, v. 29, p. 283−286.

Chen, S.F., Riganti, A., Wyche, S., Greenfield, J.E., and Nelson, D.R., 2003,Lithostratigraphy and tectonic evolution of contrasting greenstone succes-sions in the central Yilgarn craton, Western Australia: Precambrian Re-search, v. 127, p. 249−266.

Chen, S.F., Libby, J.W., Wyche, S., and Riganti, A., 2004, Kinematic natureand origin of regional-scale ductile shear zones in the central Yilgarn cra-

ton, Western Australia: Tectonophysics, v. 394, p. 139−153.Chin, R.J., and Smith, R.A., 1983, Jackson: Geological Survey of Western

Australia, 1:250,000 series Geological Map, and Explanatory Notes, 30 p.Chown, E.H., N’Dah, E., and Mueller, W.U., 2000, The relation between

iron-formation and low temperature hydrothermal alteration in an Archean volcanic environment: Precambrian Research, v. 101, p. 263−275.

Clout, J.M.F., 2003, Upgrading processes in BIF-derived iron ore deposits:Implications for ore genesis and downstream mineral processing: Transac-tions of the Institution of Mining and Metallurgy, v. 112, sec. B, p.B89−B95.

Costa, U.R., Barnett, R.L., and Kerrich, R., 1983, The Mattagami Lakemine Archean Zn-Cu sulfide deposit, Quebec: Hydrothermal co-precipi-tation of talc and sulfides in a sea-floor brine pool—evidence from geo-chemistry, 18O/ 16O, and mineral chemistry: ECONOMIC GEOLOGY, v. 78, p.1144−1203.

Costa, U.R., Fyfe, W.S., Kerrich, R., and Nesbitt, H.W., 1980, Archaean hy-drothermal talc evidence for high ocean temperatures: Chemical Geology,

v. 30, p. 341−349.Dalstra, H.J., 1995, Metamorphic and structural evolution of greenstone

belts of the Southern Cross-Diemals region of the Yilgarn Block, WesternAustralia, and its relationship to gold mineralisation: Unpublished Ph.D.thesis, University of Western Australia, 219 p.

——2006, Structural controls of bedded iron ore in the Hamersley province, Western Australia—an example from the Paraburdoo Ranges: Transactionsof the Institution of Mining and Metallurgy, v. 115, sec. B, p. B139−B140.

Dalstra, H.J., and Guedes, S.T., 2004, Giant hydrothermal hematite deposits with Mg-Fe metasomatism: A comparison of the Carajas, Hamersley, andother iron ores: ECONOMIC GEOLOGY, v. 99, p. 1793−1800.

Dalstra, H.J., and Rosière, C.A., 2008, Structural controls on high-grade ironores hosted by banded iron formation: A global perspective: Reviews inEconomic Geology, v. 15, p. 73−106.

Dalstra, H.J., Bloem, E.J.M., Ridley, J.R., and Groves, D.I., 1998, Diapirismsynchronous with regional deformation and gold mineralisation, a new con-cept for granitoid emplacement in the Southern Cross province, WesternAustralia: Geologie En Mijnbouw, v. 76, p. 321−338.

Dalstra, H.J., Ridley, J.R., Bloem, E.J.M., and Groves, D.I., 1999, Meta-morphic evolution of the central Southern Cross province, Yilgarn cra-ton, Western Australia: Australian Journal of Earth Sciences, v. 46, p.765−784.

Davis, G.J., 1972, Geology of the North Range, Koolyanobbing, WesternAustralia: Unpublished B.Sc.(Honours) thesis, University of Western Aus-tralia, 121 p.

Drummond, B.J., Goleby, B.R., Swager, C., and Williams, P.R., 1993, Con-straints on Archaean crustal composition and structure provided by deepseismic sounding in the Yilgarn Block: Ore Geology Reviews, v. 8, p.17−124.

Eisenlohr, B.N., Groves, D.I., Libby, J., and Vearncombe, J.R., 1993, The na-ture of large scale shear zones and their relevance to gold mineralisation,Yilgarn Block: Minerals and Energy Research Institute of Western Aus-tralia Report 122, 161 p.

Ellis, H.A., 1958, Report on the exploratory drilling of the Koolyanobbingiron ore deposits for pyrite: Geological Survey of Western Australia Bul-letin, v. 111, p. 1−140.

Findlay, D., 1994, Diagenetic boudinage, an analogue model for the controlon hematite enrichment iron ores of the Hamersley iron province of West-ern Australia, and a comparison with Krivoi Rog of Ukraine, and Nimba

Range, Liberia: Ore Geology Reviews, v. 9, p. 311−324.Gee, R.D., 1979, Structure and tectonic style of the Western Australian

Shield: Tectonophysics, v. 58, p. 327−369.Greenfield, J.E., and Chen, S.F., 1999, Structural evolution of the Marda-

Diemals area, Southern Cross province: Geological Survey of Western Aus-tralia Annual Review 1998-99, p. 68−73.

Greentree, M.R., and Lord, D., 2007, Iron mineralization in the Yilgarn cra-ton and future potential: Geoscience Australia Record 2007/14, p. 70−73.

Griffin, A.C., 1981, Structure and iron-ore deposition in the ArcheanKoolyanobbing greenstone belt, Western Australia: Geological Society of Australia Special Publication 7, p. 429−438.

Guarin, C.P., Jr., Angerer, T., Maund, N.H., Cowan, D.R., and Hagemann,S.G., 2009, The K Deeps magnetite mineralisation at Koolyanobbing,

Western Australia: AusIMM Iron Ore Conference 2009, 27−29 July 2009,Perth, Western Australia, Proceedings, p. 95−104.

Gutzmer, J., Beukes, N.J., de Kock, M.O., and Netshiozwi, S.T., 2005, Originof high-grade iron ores at the Thabazimbi deposit, South Africa: AusIMM

Iron Ore Conference 2005, 19−21 September 2005, Freemantle, WesternAustralia, Proceedings, p. 57−65.

Hallberg, J.A., 1987, Postcratonization mafic and ultramafic dykes of the Yil-garn block: Australian Journal of Earth Sciences, v. 34, p. 135−149.

Hippertt, J., Lana, C., and Takeshita, T., 2001, Deformation partitioning dur-ing folding of banded iron formation: Journal of Structural Geology, v. 23,p. 819−834.

Hoppe, F.E.P., 2005, The Koolyanobbing project: Deposit geology summary:Portman Iron Ore Ltd., Deposit Notes, 40 p.

Joly, A., Miller, J., and McCuaig, T.C., 2010, Archaean polyphase deforma-tion in the Lake Johnston greenstone belt area: Implications for the under-standing of ore systems of the Yilgarn craton: Precambrian Research, v.177, p. 181−198.

Jones, R.R., Holdsworth, R.E., Clegg, P., McCaffrey, K., and Tavarnelli, E.,2004, Inclined transpression: Journal of Structural Geology, v. 26, p.1531−1548.

Klein, C., 1974, Greenalite, stilpnomelane, minnesotaite, crocidolite, and

carbonates in very low-grade metamorphic Precambrian iron-formation:Canadian Mineralogist, v. 12, p. 475−498.

——1983, Diagenesis and metamorphism of Precambrian iron-formations, in Trendall, A.F., and Morris, R.C., eds., Iron-formation: Facts and prob-lems: Amsterdam, Elsevier, p. 417−469.

——2005, Some Precambrian banded iron-formations (BIFs) from aroundthe world: Their age, geologic setting, mineralogy, metamorphism, geo-chemistry, and origin: American Mineralogist, v. 90, p. 1473−1499.

Lascelles, D.F., 2006a, The genesis of the Hope Downs iron ore deposit,Hamersley province, Western Australia: ECONOMIC GEOLOGY, v. 101, p.1359−1376.

——2006b, The Mount Gibson banded iron formation-hosted magnetite de-posit: Two distinct processes for the origin of high-grade iron ore: ECO-NOMIC GEOLOGY, v. 101, p. 651−666.


0361-0128/98/000/000-00 $6.00 944



——2007a, Black smokers and density currents: A uniformitarian model forthe genesis of banded iron-formations: Ore Geology Reviews, v. 32, p.381−411.

——2007b, Genesis of the Koolyanobbing iron ore deposits, Yilgarnprovince, Western Australia: Transactions of the Institution of Mining andMetallurgy, v. 116, p. 86−93.

Libby, J., Groves, D.I., and Vearncombe, J.R., 1991, The nature and tectonicsignificance of the crustal-scale Koolyanobbing shear zone, Yilgarn craton,

Western Australia: Australian Journal of Earth Sciences, v. 38, p. 229−245.Lobato, L.M., Figueiredo e Silva, R.C., Hagemann, S.G., Thorne, W.S., andZuchetti, M., 2008, Hypogene alteration associated with high-gradebanded iron formation-related iron ore: Reviews in Economic Geology, v.15, p. 107−128.

Moine, B., Fortune, J.P., Moreau, P., and Viguier, F., 1989, Comparative min-eralogy, geochemistry, and conditions of formation of two metasomatic talcand chlorite deposits: Trimouns (Pyrenees, France) and Rabenwald (East-ern Alps, Austria): ECONOMIC GEOLOGY, v. 84, p. 1398−1416.

Morris, R.C., 1980, A textural and mineralogical study of the relationship of iron ore to banded iron-formation in the Hamersley iron province of West-ern Australia: ECONOMIC GEOLOGY, v. 75, p. 184−209.

——1985, Genesis of iron ore in banded iron-formation by supergene andsupergene-metamorphic processes—a conceptual model, in Wolf, K.H.,ed., Handbook of strata-bound and stratiform ore deposits: Amsterdam, El-sevier, p. 73−235.

Morris, R.C., Thornber, M.R., and Ewers, W.E., 1980, Deep-seated iron ores

from banded iron-formation: Nature, v. 288, p. 250−252.Mueller, A.G., and McNaughton, N.J., 2000, U-Pb ages constraining

batholith emplacement, contact metamorphism, and the formation of goldand W-Mo skarns in the Southern Cross area, Yilgarn craton, Western Aus-tralia: ECONOMIC GEOLOGY, v. 95, p. 1231−1257.

Nelson, D.R., 1999, Compilation of geochronology data, 1998: GeologicalSurvey of Western Australia Record 1999/2, 222 p.

Nesbitt, R.W., 1971, Skeletal crystal forms in the ultramafic rocks of the Yil-garn block, Western Australia: Evidence for an Archaean ultramafic liquid:Journal Geological Society of Australia Special Publication 3, p. 331−347.

Newman, J., and Mitra, G., 1994, Fluid-influenced deformation and recrys-tallization of dolomite at low temperatures along a natural fault zone,Mountain City window, Tennessee: Geological Society of America Bulletin,

v. 106, p. 1267−1280.Passchier, C.W., and Trouw, R.A.J., 1996, Microtectonics: Berlin, Heidel-

berg, New York, Springer, 289 p.Pidgeon, R.T., Wilde, S.A., Compston, W., and Shield, M.W., 1990, Archean

evolution of the Wongan Hills greenstone-belt, Yilgarn craton, WesternAustralia: Australian Journal of Earth Sciences, v. 37, p. 279−292.

Pillans, B., 2000, Palaeomagnetic dating of Phanerozoic weathering imprints,Mount Percy mine, Kalgoorlie, Western Australia, in Klootwijk, and Chris,eds., Australian palaeomagnetism, rockmagnetism, and environmentalmagnetism 2000: Canberra, A.C.T., Australia, Australian Geological Survey Organisation 390 p.

——2004, Geochronology of the Australian regolith, in Anand, R.R., andBroekert, P., eds., A compilation of regolith-landscape case studies andlandscape evolution models: Perth W.A., CRC Leme, etr p.

Pokrovsky, O.S., and Schott, J., 2001, Kinetics and mechanism of dolomitedissolution in neutral to alkaline solutions revisited: American Journal of Science, v. 301, p. 597−626.

Portman, 2008, Portman Iron Ore Ltd., Annual report 2007, 12 p.Powell, C.M., Oliver, N.H.S., Li, Z.-X., Martin, D.M., and Ronaszeki, J.,

1999, Synorogenic hydrothermal origin for giant Hamersley iron oxide ore-bodies: Geology, v. 27, p. 175−178.

Qiu, Y.M., McNaughton, N.J., Groves, D.I., and Dalstra, H.J., 1999, Ages of

internal granitoids in the Southern Cross region, Yilgarn craton, WesternAustralia, and their crustal evolution and tectonic implications: AustralianJournal of Earth Sciences, v. 46, p. 971−981.

Ramanaidou, E.R., 2009, Genesis of lateritic iron ore from banded iron-for-mation in the Capanema mine (Minas Gerais, Brazil): Australian Journal of Earth Sciences, v. 56, p. 605−620.

Ramanaidou, E.R., and Morris, R.C., 2009, Comparison of supergenemimetic and supergene lateritic iron ore deposits: AusIMM Iron Ore Con-ference 2009, 27−29 July 2009, Perth, Western Australia, Canberra, A.C.T.,Australia, Proceedings, p. 243−246.

Ramsey, J.G., 1962, The geometry of conjugate fold systems: GeologicalMagazine, v. 99, p. 516−526.

——1967, Folding and facturing of rocks: New York, McGraw-Hill, 562 p.Rosière, C.A., Spier, C.A., Rios, F.J., and Suckau, V.E., 2008, The iabirites of

the Quadrilátero Ferrífero and related high-grade iron ore deposits: AnOverview: Reviews in Economic Geology, v. 15, p. 223−254.

Sullivan, C.M., 1973, Geology of the South Range, Koolyanobbing, Western

Australia: Unpublished B.Sc.(Honours) thesis, University of Western Aus-tralia, 84 p.

Taylor, D., Dalstra, H.J., Harding, A.E., Broadbent, G.C., and Barley, M.E.,2001, Genesis of high-grade hematite orebodies of the Hamersley province, Western Australia: ECONOMIC GEOLOGY, v. 96, p. 837−873.

Thorne, W.S., Hagemann, S.G., and Barley, M., 2004, Petrographic and geo-chemical evidence for hydrothermal evolution of the North deposit,Mt.Tom Price, Western Australia: Mineralium Deposita, v. 39, p. 766−783.

Thorne, W.S., Hagemann, S.G., Webb, A., and Clout, J., 2008, Banded ironformation-related iron ore deposits of the Hamersley province, WesternAustralia: Reviews in Economic Geology, v. 15, p. 197−222.

Woodcock, N.H., and Fischer, M., 1986, Strike-slip duplexes: Journal of Structural Geology, v. 8, p. 725−735.

Wyche, S., 1999, Geology of the Mulline and Riverina 1:100 000 sheets: Ge-ological Society of Western Australia, 1:100 000 Geological Series Explana-tory Notes, 28 p.

Wyche, S., Nelson, D.R., and Riganti, A., 2004, 4350−3130 Ma detrital zir-

cons in the Southern Cross granite-greenstone terrane, Western Australia:Implications for the early evolution of the Yilgarn craton: Australian Jour-nal of Earth Sciences, v. 51, p. 31−45.

Zhang, R., Hu, S., Zhang, X., and Yu, W., 2007, Dissolution kinetics of dolomite in water at elevated temperatures: Aquatic Geochemistry, v. 13,309–338.


2010 angerer&hagemann_ structure kooly

Documents