La Candelaria is the largest of the iron oxide Cu-Au(-Zn-Ag) deposits in the Punta del Cobre belt, which also hosts the Punta del Cobre district, sensu strictu. The Punta del Cobre belt lies within an Early Cretaceous continental volcanic arc/marine back-arc basin terrane. The volcanic arc and marine carbonate back-arc sequences are intruded by Early Cretaceous granitoid plutons that form part of the Chilean Coastal Batholith. The deposits of the Punta del Cobre belt occur along the eastern margin of the batholith within (e.g., La Candelaria) or just outside the contact metamorphic aureole (e.g., the Punta del Cobre district). Andesitic volcanic and volcaniclastic host rocks are intensely altered by biotite-quartz-magnetite. This style of alteration extends much further to the east of the intrusive contact than the metamorphic mineral associations in the overlying rocks that are clearly zoned outboard. Local areas of intense calcic amphibole veining that overprints all rock types occur within the contact metamorphic aureole.

Gold-copper-bismuth deposits of the Tennant Creek district, Northern Territory, Australia, are distinctive as some of the highest grade deposits within the Fe-oxide Cu-Au global family. They are unified by an association with epigenetic magnetite ± hematite - rich 'ironstones' that are hosted by a sequence -I860 Ma, low metamorphic grade, Fe-oxide rich greywacke, siltstone and shale. While many of the high grade gold orebodies are dominated by magnetite - chlorite ± minor hematite, muscovite and pyrite, there are significant variations representing a spectrum of styles from reduced (pyrrhotite-bearing) Cu-Au-Bi deposits to oxidised hematitic Au-Bi(Cu) deposits. Shear-hosted Au-Cu mineralisation outside ironstones further adds to the diversity of styles present in the district. Ironstones predated syn- to late-deformational ~1825-1830 Ma introduction of Au, Cu and Bi in ~-300-350°C, acidic, low-moderate salinity or hypersaline fluids, which were in places carbonic and nitrogenous. The very wide range of oxidation-reduction conditions during ore deposition across the district is interpreted as the product of both reduced (magnetite ± pyrrhotite stable, H2S > S04=) and oxidised (hematite stable, S04= > H2S) fluids reacting with ironstones and/or mixing. Oxygen and hydrogen isotope data point to an hybrid ore fluid source with input of evolved surficial or formation waters, whereas Sm-Nd reconnaissance data and sulfur isotope compositions are consistent with contributions from igneous sources.

Cu-Au(-Mo) mineralisation at the Portia-North Portia prospect is located under cover on the eastern flank of the Benagerie Ridge Magnetic Complex, within the Curnamona Province and approximately 125 km WNW of Broken Hill. The mineralisation is located within rocks which have a SHRIMP U-Pb zircon age of 1703 + 6 Ma, which is similar to Willyama Supergroup ages obtained from the Broken Hill and Olary Domains. The meta-sedimentary unit that hosts the mineralisation is approximately 200m thick and is overlain by carbonaceous phyllite, and underlain by a unit which is dominated by albite-magnetite-hematite. The host units contain numerous carbonate-rich domains intercalated with meta-evaporitic sediments. The sequence had undergone low-grade metamorphism and fabric development and was subsequently intensely albitised. Hydrothermal monazites formed during this albitisation event give SHRIMP II in situ U-Pb ages of ~1630 Ma. The albitised meta-sediments proved to be an excellent host to the later Cu-Au(-Mo) mineralisation. Abundant monazite associated with the mineralisation yield SHRIMP II ages of ~1605 Ma. It is possible that the numerous, highly fractionated and altered diorite bodies known to be present on the Benagerie Ridge, may have produced some of the metals. The "Hiltaba age" (-1585-1590 Ma) granites of the area cannot be considered as a source of the metals.

The >1510-1500 Ma Ernest Henry Fe-oxide-Cu-Au orebody is a hydrothermal deposit hosted in K-feldspar altered ca 1740 Ma plagioclase phyric volcanic rocks in the Cloncurry district, Mount Isa Inlier. Mineralization occurred late in a post-peak metamorphic hydrothermal system, and the ore is mainly hosted in an infill-supported hydrothermal breccia that grades to crackle veining at the margins. The orebody has a > 1km down dip extension, and is structurally-controlled between two shear zones that trend NE-SW and dip -35° to the SE. The ore is mainly composed of subrounded clasts separated by a fine- to medium-grained infill composed of magnetite, calcite, pyrite, biotite, K-feldspar, chalcopyrite, hematite, garnet, barite, fluorite, quartz and molybdenite.

The Cumamona Province (South Australia/New South Wales) and Cloncurry district (NW Queensland) are both extensively metasomatised terrains containing hydrothermal iron oxide copper-gold and related deposits. Structural timing criteria and geochronological data suggest that the deposits formed at 1630-1600 Ma (Cumamona) and 1540-1500 Ma (Cloncurry). The Cloncurry deposits have a close temporal association with I-type granitoids and limited data suggest a similar relationship exists in the Cumamona Province. The majority of deposits are hosted by metamorphosed Palaeoproterozoic supracrustal rocks of varying age, composition and metamorphic grade. Mineralisation was localised by a range of brittle-ductile and brittle structures and produced vein, stockwork, breccia and replacement orebodies. Variations of fluid chemistry, host rocks and physical conditions produced mineralogically-diverse alteration zones, varying Cu:Au ratios, many different minor element associations, and inconsistent spatial relationships between magnetite and ore metals. Regional-scale alteration systems are dominated by Na-(Fe-Ca)-rich assemblages in which the most characteristic mineral is albite. Most of the ore deposits are specifically associated with pre- to synmineralisation alteration assemblages composed of medium to high temperature K-Fe-(Ca-Mg)-rich minerals together with late-stage parageneses containing carbonates. The deposits formed in deep-seated (> 5km) environments by a variety of different geochemical mechanisms from complex H,0-C02±CH4±N2-saIt fluids of magmatic and/or metamorphic derivation .

The -1590 Ma Olympic Dam Cu-U-Au-Ag-REE deposit is located in the Stuart Shelf geological province of South Australia, on the eastern margin of the Gawler Craton. The deposit is hosted by the Olympic Dam Breccia Complex, a large hydrothermal breccia system wholly contained within the Roxby Downs Granite, a Proterozoic age granitoid inteipreted to be part of the Hiltaba Suite. Initial hydrothermal activity within the Olympic Dam Breccia Complex was probably localised by structures in a dextral fault jog environment. Subsequent development of the complex involved repetitive and overprinting physical, chemical and volcanic brecciation mechanisms, resulting in a highly variable array of irregularly shaped and distributed breccia zones with widely differing and gradational lithologies. A complex pattern of hydrothermal alteration dominated by hematite and sericite, with lesser chlorite, siderite and quartz is associated with the breccia zones. Mineralisation within the deposit is intimately associated with iron-oxide alteration of the granitoid, which dominantly occurs as hematite, with lesser magnetite at depth and on the periphery of the breccia complex. The principal copper minerals within the deposit show a broad lateral and vertical, hypogene zonation pattern grading from chalcopyrite on the margins to bornite, then chalcocite adjacent to a central barren core. Gold and silver are mainly associated with the copper sulfides, while uranium dominantly occurs in pitchblende disseminated throughout the hematitic breccia zones. Overall, mineralisation grade generally correlates with the degree of hematite alteration and is largely dependent on copper sulfide tenor. Minor brittle faulting post-dates breccia development and appears to have exploited existing anisotropies within the complex. Late-stage fault movements are associated with barite-fluorite vein arrays which overprint the orebody. The deposit formed in a high level volcanic environment, venting to the surface and possibly forming a composite phreatomagmatic eruption crater, which has subsequently been completely eroded. Mafic and felsic dykes intruded the breccia complex, locally producing diatreme structures. Tectonism, hydrothermal activity, dyke intrusion, brecciation, alteration and mineralisation within the system were broadly concurrent and interdependent. Hydrothermal fluids and metals have a dominantly magmatic source, probably associated with the Middle Proterozoic volcano-plutonic event correlated with the Gawler Range Volcanics and Hiltaba Suite intrusives.

Iron oxide copper (-gold) deposits consist of dominant magnetite or haematite, with one or more copper sulphides and pyrite, with associated K-feldspar or sericite or albite or biotite and chlorite predominant in the ore host rocks. The deposits display a unique association with host successions characterised by an absence of, or by very minor occurrence of, elemental carbon or reduced-carbon compounds and reduced-sulphur minerals. The relatively oxidised nature of the ore host succession is reflected in the "magnetically active" signature that usually defines iron oxide copper (-gold) mineralised districts. This signature shows that magnetite is ubiquitous and variably abundant within ore host successions. Host successions with discrete domains respectively characterised by (a) by rocks with an absence or rarity of carbon or reduced carbon minerals, and (b) by a predominance of rocks containing carbon or reduced carbon minerals, however, contain iron sulphide-copper (-gold) deposits on or near the boundaries of the domains. Examples of the iron sulphide-copper (-gold) deposits are the Mt Isa and Gunpowder deposits, many small occurrences in the Eastern Fold Belt (Mt Isa Inlier), the El Soldado deposit, and others.

Brines may be generated in sedimentary, magmatic or metamorphic settings, and they change chemistry extensively as they move through rocks and interact with them. The primary constraint on their metal carrying capacity is their salinity, but they may carry very variable amounts of S in solution, depending on their source and the rocks that they have encountered. Sulphur availability and oxidation state are also major controls on which metals will be transported and which precipitated. Availability of fluid inclusion brine analyses is making possible the characterisation of a much wider range of brine types than was hitherto possible, and providing information about metal contents in a wide range of settings, as well as tracer analyses. Iron contents of brines are broadly temperature dependent, and are much higher in magmatic brines than in sedimentary ones, but basinal fluids may still carry sufficient Fe in solution to precipitate iron oxides at an oxidation front, and may be much more voluminous. Brines of different origins can often be distinguished on the basis of conservative halogen tracers unaffected by wall rock interactions: Br/CI ratios used in conjunction with I/Cl ratios or d"Cl values separate residual bittern fluids from re-dissolved evaporites, with igneous brines forming an intermediate, but somewhat distinct, grouping.

No single model satisfactorily accounts for the diverse characteristic of Fe-oxide-rich hydrothennal systems. Consideration of a spectrum of geologically reasonable models gives insight into the origins of variability among these deposits. Key features that need to be rationalized by any model are the abundance of hydrothermal magnetite and/or hematite, the chemically distinct suite of elements (REE-Cu-Co-Au-Ag-U), the variability of associated magmas, the distributions and volumes of associated hydrothennal alteration, and the broader geologic setting(s).

Geologic and geochemical evidence show that the ore-fonning fluids are brines, but the source of the brines is controversial. Multiple sources are possible, indeed likely. The identity and consequences of alternative sources - magmatic and non-magmatic are considered: First, by a review of plausible fluids and general consequences, second, by examination of the system characteristics, third, by specific consideration of the consequences of alternative models, and fourth, by consideration of selected systems where non-magmatic brines must play a major role.

We review some of the key characteristics of different types of hydrothermal Fe-oxide-rich(-Cu-Au) systems. Some are economic; many are only geochemically anomalous. Two end-members and several variants on these end members are proposed. One group is typified by relatively high-temperature mineralization, and relatively high K/Na and Si/Fe in the alteration. We suggest that these features (and others) are distinctive of magmatic fluid sources and that this group overlaps with porphyry Cu-Au and related deposit types. A second, broad group is typified by more oxide-rich, sulfide-poor mineralization, low Si/Fe ratios, and voluminous alkali-rich alteration where sodic types commonly exceed K-rich varieties. We suggest that the key features of this group reflect involvement of non-magmatic brines and that ore grades are less common as the metals are less easily trapped. Hybrid examples, where fluids of both types are involved, are expected (and known).

Conceptual and quantitative models of magmatic and non-magmatic fluid sources yield insight into the expected differences, the source and distribution of metals within these systems, and possible controls on ore deposition. These models highlight the difference between magma-sourced fluids and non-magmatic fluids. The former tend to be focused at the tops of magma chambers and have a built-in depositional mechanism -cooling. The latter require a different type of focusing mechanism - structural or stratigraphic, and different traps - mixing, specialized host rocks, and/or boiling. These models predict consistent differences in mineral assemblages, metal contents, alteration volumes, zoning, paragenesis, and geochemistry. For magmatic fluid sources, the models reproduce the key characteristics of that group, notably the porphyry-related systems. For non-magmatic brine sources, predicted characteristics match well with observations of Fe-oxide-rich systems in environments where these fluids are known to dominate, including mafic igneous systems and modem analogs such as the Salton Sea. These cases show that Fe-oxide (-Cu-Au-REE-Co) enrichments can result from non-magmatic sources. In other environments (e.g., with intermediate to felsic igneous rocks; settings deeper than ~5 km) the relative importance of various fluid sources and their consequences for mineralization remain to be fully explored. Young systems in the American Cordillera and elsewhere can help unravel the many threads that relate these enigmatic mineral deposits.

Fe-oxide-Cu-Au deposits typically formed in continental arc and intracratonic tectonic settings, predominantly in the Proterozoic, but also during the Phanerozoic. The lack of reported evaporitic sequences in several major Fe-oxide-Cu-Au districts including the Gawler Craton and Stuart Shelf, Tennant Creek, Great Bear Magmatic Zone, and Carajas districts suggests that the presence of evaporites is not a prerequisite for the formation of these deposits.

Fluid inclusion studies of Fe-oxide-Cu-Au deposits typically indicate the presence of coexisting hypersaline and C02-rich fluid inclusions that may have originated by unmixing of an original H,0-C02-salts fluid. At the Lightning Creek prospect in the Cloncurry district, these fluid types were generated during crystallization of granitic sills that are associated with a major magnetite-rich vein system. Stable isotope data from Lightning Creek and a number of Cu-Au deposits in the Cloncurry district and elsewhere are compatible with formation of these deposits principally from magmatic-hydrothermal fluids. Syn- to post-mineralization Na-Ca-rich fluids are present in many deposits, and may represent meteoric and/or connate fluids that mixed with the hypersaline magmatic fluids during or after mineralization.