©2014 Society of Economic Geologists, Inc. Special Publication 18, pp. 153–175 Chapter 8 The Mineral System Concept: The Key to Exploration Targeting T. Campbell McCuaig1,† and Jon M. A. Hronsky1,2 1 Centre for Exploration Targeting and Australian Research Council Centre of Excellence for Core to Crust Fluid Systems, School of Earth and Environment, University of Western Australia 6009, Australia 2 Western Mining Services, Suite 26, 17 Prowse Street, West Perth, Western Australia 6005, Australia Abstract To aid conceptual targeting, the past two decades have seen the emergence of the mineral systems concept, whereby ore deposits are viewed as small-scale expressions of a range of earth processes that take place at different temporal and spatial scales. The mineral systems approach has been spurred by three main drivers: the recognition of patterns of mineralization in increasingly available large geoscience datasets; advances in geographic information system (GIS) technologies to spatially query these datasets; and marked advances in understanding the evolution of earth systems and geodynamics that provide context for mineralization patterns. An understanding of mineral systems and the scale-dependent processes that form them is important for guiding exploration strategies and further research efforts. Giant ore deposits are zones of focused mass and energy flux. Advances in understanding of the physics of complex systems—self organized critical systems—leads to a new understanding of how fluid flow is organized in the crust and how high-quality orebodies are formed. Key elements for exploration targeting include understanding and mapping threshold barriers to fluid flow that form extreme pressure gradients, and mapping the transient exit pathways in which orebodies form. It is proposed that all mineral systems comprise four critical elements that must combine in nested scales in space and time. These include whole lithosphere architecture, transient favorable geodynamics, fertility, and preservation of the primary depositional zone. Giant mineral deposits have an association with large, longlived deeply penetrating and steeply dipping structures that commonly juxtapose distinctly different basement domains. These structures are vertically accretive in nature, often having limited or subtle expressions at or above the level of ore deposition. Three transient geodynamic scenarios are recognized that are common to many mineral systems: anomalous compression, initial stages of extension, and switches in the prevailing far-field stress. In each of these scenarios, ”threshold barriers” are established which produce extreme energy and fluid/magma pressure gradients that trigger self-organized critical behavior and ore formation. Fertility is defined as the tendency for a particular geologic region or time period to be better endowed than otherwise equivalent geologic regions. Fertility comprises four major components: secular Earth evolution (variations in the Earth’s atmosphere-hydrosphere-biosphere-lithosphere through geologic history that result in formation of deposits), lithospheric enrichment, geodynamic context, and paleolatitude (in specific mineral systems). The primary depositional zone is usually within the upper 10 km of the Earth’s surface, where large P-T-X gradients can be established over short distances and time scales. The variable preservation of this zone through subsequent orogeny explains the secular distribution of many ore deposit types. The mineral system approach has advantages in exploration targeting compared to approaches that use deposit models. Emphasizing common ore-forming processes, it links many large ore systems (e.g., VMS-epithermal, porphyry-orogenic gold) that are currently considered disparate deposit models and relates these ore systems in a predictable way to their large-scale geodynamic context. Moreover, it focuses mineral exploration strategies on incorporating primary datasets that can map the critical elements of mineral systems at a variety of scales, and particularly the regional to camp scales needed to make exploration decisions. Introduction For centuries the characteristics of mineral deposits have been studied with two main objectives: to better understand their geometry and nature to enable their efficient exploitation, and to better understand their genesis to aid prediction of the location of additional mineral resources (Agricola, 1556; Loughlin and Behre, 1933). The outcomes of these research efforts are currently embedded in the literature as deposit models, and current studies tend to assign deposits to major groups or subgroups (Hedenquist et al., 2005, and references † Corresponding author: e-mail, [email protected] therein). These models serve as useful frameworks in which we can compare and contrast deposits and deposit styles, better understand common links in ore genesis, particularly at the site of deposition, and design detection techniques to find further analog mineral concentrations (e.g., Hedenquist et al., 2005). One important measure of the utility of a model is its ability to predict the location and quality of undiscovered resources. Most economic geologists would agree that our understanding of deposit-scale controls on mineral deposition and ore genesis, albeit incomplete, has improved substantially concomitant with the large amount of research effort directed 153 154 MCCUAIG AND HRONSKY toward understanding them. A corollary should be that we are now better at finding new high-quality mineral deposits. Yet recent reviews of exploration success indicate that this is not the case, with exploration effectiveness and the quality of the resource project pipeline declining (e.g., McKeith et al., 2010). Furthermore, credit for the role of predictive geologic targeting concepts in mineral discoveries has been quite limited, with most discoveries described as the result of surface prospecting (Sillitoe, 2004). It appears that there is a gap between our current understanding of ore deposits and our ability to translate this knowledge into a predictive framework to find further high-quality resources (McCuaig et al., 2010). A theme that has emerged over the past two decades to aid conceptual targeting is the mineral system concept, whereby mineral systems are viewed as small expressions of a much larger set of geologic processes that align to concentrate minerals (Wyborn et al., 1994; Knox-Robinson and Wyborn, 1997). The approach places the mineral deposit within the rapidly increasing understanding of evolving earth systems that has advanced substantially over the past two decades (Groves et al., 2005; Kerrich et al., 2005; Reddy and Evans, 2009; Goldfarb et al., 2010; Cawood and Hawkesworth, 2013). This multiscale approach to mineral systems has substantial predictive power compared to the standard deposit model concept (McCuaig et al., 2010). The purpose of this paper is twofold: to outline the current understanding of the mineral system as an organizational framework to both understand ore genesis and predict the location of high-quality deposits, and to establish a coherent language of mineral systems for economic geologists. First, the evolution of the understanding of mineral systems is reviewed. Secondly, the principle of ore deposits as the loci of large fluxes of mass and energy is discussed in terms of the physics of fluid flow and self-organizing critical systems. Thirdly, the principal elements of a mineral system, here defined as whole lithosphere architecture, transient geodynamic triggers, fertility and preservation of the primary depositional zone, are reviewed with examples taken from many mineral districts globally across a range of commodities. Finally, links between currently disparate ore deposit models are illustrated to highlight the potential power of the mineral system framework over the traditional application of the deposit model framework. Evolution of Understanding of Mineral Systems Deposit models A traditional tool for understanding ore deposit genesis is the deposit model. No matter what mineral concentration is studied, inevitably it is compared to one or more “models” of deposit styles. Deposit models commonly refer to ”type” deposits, e.g., Carlin-type Au (Cline et al., 2005), Kambaldatype NiS (Gresham and Loftus-Hills, 1981), Bushveld-type PGE (Barnes and Lightfoot, 2005), and Witswatersrand-type Au (Frimmel et al., 2005; Law and Phillips, 2005). These models are built dominantly from deposit-scale observations. In the past two decades, more sophisticated deposit models have included links between different deposit types to make more comprehensive ”unified” deposit models or model spectrums, e.g., the epithermal deposit spectrum (Simmons et al., 2005), the broader porphyry-related ore environment (Fig. 1; Sillitoe, 2010), the crustal continuum model for Archean gold deposits (Groves, 1993), orogenic gold (Groves et al., 1998; Goldfarb et al., 2005), and the clastic-dominated (CD), Mississippi Valley-type (MVT) spectrum of lead-zinc deposits (Leach et al., 2005, 2010). The above mentioned models are excellent summaries of variations between deposit types and can be excellent syntheses of deposit-scale processes. From these models has come an understanding of analog structural, chemical, and mineralogical footprints of mineralization. Structural analogs have been useful in mine-scale targeting of similar ore shoots (e.g., Archean orogenic gold at Norseman, Western Australia; Campbell, 1990), whereas the chemical and mineralogical understanding of alteration halos has been very important in helping discover further resources in known mineralized provinces (e.g., porphyry deposits in the American Cordillera, Fig. 1; Lowell and Guilbert, 1970; Sillitoe, 2000; Sillitoe and Perelló, 2005; Sillitoe and Thompson, 2006). The way deposit models are traditionally applied to regional exploration is to search for geologic situations analogous to those defined by the deposit model. However, there are several challenges associated with this. 1. Too strict a focus on the analog can result in deposits being missed because their geologic setting does not have all the features of the deposit model. However, those missing features may not be fundamental to the process of ore formation. For example, in the Eastern Goldfields of the Archaean Yilgarn craton of Western Australia, the earliest gold discoveries in the 1890s were dominantly in mafic rock types. As a result, exploration targeting up until the 1980s in this region was heavily biased toward this rock type. This mafic host rockcentric targeting model blinded many explorers to significant gold concentrations in other rock types. However, in the Yilgarn, the major discoveries in the past three decades have been in other settings such as late conglomerate sequences (Wallaby, 3 Moz; Kanowna Belle, 8.5 Moz), sedimentary/volcaniclastic basins (Sunrise, 7 Moz), and intrusive stocks (Boddington, 35 Moz; data from Guj et al., 2011)—all in previously low-ranked regions. Clearly, the presence of a mafic host rock is not a fundamental requirement for a major gold deposit. 2. Conversely, targeting focused on the analog characteristics of the model may generate many “false positives” (McCuaig et al., 2009). For example, in the case of targets in mafic rocks for gold discussed in the example above, there are many more targets without gold than with gold. 3. For some commodities, there is an overabundance of defined deposit models. Commonly this develops over time, as every major new discovery that differs significantly from previously established analogs results in the definition of a new variation of the deposit model. Uranium models are an excellent example of this. Until recently, uranium deposit classification had 14 models and 22 submodels (International Atomic Energy Agency, 2000), almost a different model for each major deposit. In such a classification, there are often too many variations on a theme for practical application to exploration (Kreuzer et al., 2010). 4. Most importantly, deposit models struggle to differentiate between large or high-quality mineral concentrations and 155 THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING High sulfidation epithermal disseminated Au+ Ag + Cu Phreatic breccia V V V V V Rocks V V V V KEY Intermediate sulfidation Au-Ag V Volcanic rocks Carbonate rocks V V V V V V High sulfidation lode Cu-Au+ Ag V V + + + + + 1 Km + Porphyry intrusion Alteration Steam heated Intermediate argillic Quartz-kaolinite Quartz pyrophyllite Sericitic Chloritic + Skarn Potassic + + + Vuggy quartz + silicification + Porphyry Cu+ Au+ Mo + Quartz-alunite + 1 Km V V Marble front + Marble front V V V V V V V V V V V V V V V Propylitic + Fig. 1. Example of a deposit model showing linkages between porphyry Cu ± Au ± Mo deposits, epithermal deposit types, and peripheral vein or skarn systems depicting the spatial variation in deposit style with depth, host-rock composition, and distance from the intrusion. Also depicted is the traditional view of deposit-scale footprints delineated by chemical and mineralogical changes caused by fluid-rock interaction around the deposit. Note that although porphyry Cu deposits are known to have some of the largest footprints, their halos of alteration are only detectable on the deposit to camp scale. Modified from Sillitoe (2012). small or low-quality mineral concentrations. The past half century of intensive ore deposit research has demonstrated that big deposits look very similar to small deposits at the deposit scale—they are just bigger. Small deposits have the same deposit-scale structural and lithologic settings, fluid or sequences of fluids, alteration, and metal anomalism as large metal concentrations. 5. Deposit models also focus on describing how and why mineralization occurs, but rarely on the spatial prediction of large, high-quality deposits. Therefore they are commonly of little use in early-stage exploration environments with little geologic data (the norm for most undercover exploration). Therefore, there is a disconnection between deposit models and their successful practical application to mineral exploration. It is proposed that a large portion of this disconnection relates to the issue of scale and particularly the identification of elements of the deposit models that are relevant at different scales of exploration targeting. The deposit model framework often focuses on features at the deposit scale, whereas many of the critical decisions in mineral exploration occur at larger scales, prior to the employment of detection technology at the deposit scale (McCuaig and Hronsky, 2000; Hronsky and Groves, 2008; McCuaig et al., 2010). At these larger scales, the deposit models have had limited predictive power (Sillitoe, 2004; Simmons et al., 2005; Sillitoe and Thompson, 2006). This scale consideration is best illustrated from the perspective of footprints of mineralization. For the purposes of this paper, footprints are defined as the recognizable and mappable expressions of mineral deposits, and it is from the understanding of deposit footprints that the prediction and detection methods employed by the exploration industry are derived. Traditional deposit footprints are formed by the circulation of fluids in surrounding wall rocks that leave a chemical and mineralogical expression and, in certain cases, a geophysical expression that can be used to vector toward mineralization. These traditionally defined footprints are commonly small, on the order of tens of square kilometers, and often much smaller (Fig. 1; Large et al., 2001; Seedorf et al., 2005; Sillitoe, 2010), and even when a footprint is recognized, one cannot determine subeconomic or barren footprints from well-mineralized and economic counterparts. The search for new high-quality mineral districts is increasingly focused in technically challenging areas below cover where traditional detection technology is both expensive and 156 MCCUAIG AND HRONSKY less effective. In these areas, the traditional footprints discussed above will be of limited use in exploration targeting. It is therefore imperative to understand and recognize the larger scale footprints of the entire mineral system in order to effectively target mineral exploration. The largest scale footprint of a mineral system is commonly at the continental (Fig. 2) or transcrustal scale (e.g., Drummond et al., 2006; Snyder, 2013). Such large-scale footprints are never defined by traditional deposit-scale studies. Therefore, there is a need for an organizing framework to aid exploration targeting for new high-quality mineral deposits. The framework must encompass the expression of mineral systems at a range of scales appropriate to the range of scales of critical exploration targeting decisions. industry-driven approach, as academics did not initially have easy access to these datasets that were proprietary and closely guarded by industry when first available. Second, advances in GISs and the ability to systematically interrogate large datasets spurred the need to find common characteristics of mineral deposits that could be queried to produce conceptual targets alongside empirically driven approaches (Wyborn et al., 1994; Knox-Robinson and Wyborn, 1997). Finally, advances in understanding the evolution of earth systems and geodynamics provided a multiscale context for understanding various expressions of mineralization (Groves et al., 2005; Kerrich et al., 2005; Goldfarb et al., 2010; Leach et al., 2010). A mineral system was defined by Wyborn et al. (1994, p. 109) as “all the geologic factors that control the generation and preservation of mineral deposits [stressing] the processes that are involved in mobilizing ore components from a source, transporting and accumulating them in more concentrated form, and then preserving them throughout the subsequent geologic history.” This seminal paper recognized that ore deposits were the expression of a large number of geologic processes aligning to trigger ore accumulation. Furthermore, they emphasized A systems approach to understanding mineralization The systems approach to understanding mineralization emerged as a result of three concurrent factors. First, the advent in the 1980s of large-scale image-processed geophysical datasets allowed recognition of regional patterns correlating with mineralization (Woodall, 1984, 1994). This was an Canada 0 Galena-Decorah-Platteville formations 500 1000 km Outline of Upper Miss District Zn in deep oil wells Viburnum Trend Old Lead Belt Zn in deep oil wells Bonneterre Fm Knox Group NE Arkansas Davis K x no Gr o East Middle Tennessee Tennessee District District up Sinking Valley Timberville Freidensville Atlantic Ocean Dolomite Fronts and associated Zn/Pb mineralization mid-continent USA Gulf of Mexico Dolomite Front (hachures on dolomitized side of front) Dolomitized rock behind Dolomite Front Ore occurrences/significant mineralization Fig. 2. Example of a continental-scale footprint of a mineral system. All MVT Pb-Zn deposits in the Phanerozoic basins of North America occur at the dolomite front (Harper and Borrok, 2007). Therefore, the dolomite front is an example of a highorder mappable expression of the mineral system footprint that can be used to narrow the search space for Pb-Zn deposits at an early stage of exploration strategy. THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING that by understanding the entire system, one had the potential to expand the mappable footprint of a deposit. The mineral systems approach was originally driven by industry. The earliest documented successful application of a mineral systems approach (although not called that at the time) was the discovery by Western Mining Corporation (WMC) of the Yeelirrie uranium deposit in Western Australia in 1972 (Woodall, 1994), followed by the discovery of Olympic Dam, South Australia in 1975 (Woodall, 1994; Haynes, 2006). These conceptually driven exploration programs were actually aimed at finding existing orebody types—sandstone-hosted uranium in the case of Yeelirrie, and sedimentary-hosted copper in the case of Olympic Dam. In the Yeelirrie case, the concept of having a source of U-rich granites and a sedimentary system with redox gradients to precipitate U was transferred to the Archean Yilgarn craton of Western Australia. This program resulted in the discovery of paleochannel uranium, the first deposit of this kind discovered globally. In the case of Olympic Dam, WMC geologists had broken down the critical elements of models for sedimentary rock-hosted copper deposits to a required source of Cu (in this case weathered basalts), large structural pathways to transport the fluids through the crust (lineaments recognized in large geophysical datasets), and an overlying sedimentary basin that could host sequences of reduced rocks to cause Cu deposition. What the exploration program discovered was an entirely different orebody type—and a style never before discovered—an iron-oxide copper gold uranium orebody in the basement beneath the sedimentary rocks. These examples highlight a key strength of the mineral systems concept. Because the mineral systems concept focuses on the critical processes required to generate a large concentration of metal, it sees through deposit-style variations to common links between ore systems, in this case lithosphere-scale architecture, metal source regions, and depositional zones. No evidence of significant metal concentration was required, nor entered into, in the targeting process. Therefore, this approach to targeting mineralization has the potential to lead exploration geoscientists into regions with no empirical evidence of mineralization and to find the yet undiscovered styles of mineral systems (e.g., Woodall, 1994). The mineral system approach gained favor in parallel with advances in GISs, which allowed systematic querying of datasets for conceptual and empirical expressions of mineral systems in GIS platforms (Wyborn et al., 1994; Knox-Robinson and Wyborn, 1997; Kreuzer et al., 2010). The original approach mimicked that used by the petroleum industry since the 1970s (Magoon and Beaumont, 1991). Three main aspects of ore formation were considered: sources (of ligands, metals, fluid, heat, magma), pathways (permeable strata or actively deforming structures that create permeability and transport ore-forming fluids/magmas through the crust), and traps (areas of metal deposition). The key aspects of this approach were that it was process based (rather than descriptive based), the processes considered were critical for ore formation, and the processes were independent of each other. Recognizing the difference between petroleum systems (a mass trapping system) and mineral systems (a mass scrubbing process), McCuaig et al. (2010) modified the mineral system framework to consider the critical elements of source, pathway, physical throttle (putting a large amount of fluid/magma through 157 a small volume of rock), and mass scrubbing (triggering metal deposition from the transporting medium, but allowing flowthrough of the fluid medium). Contrast this mineral system approach to that of the mafic-hosted model for Au in the Yilgarn, as discussed previously. The latter focused on geologic characteristics of known deposits, but failed to articulate that the differentiated dolerites were just one (albeit very good) example of a physical throttle (brittle dolerite attracts fracturing and permeability) and a chemical scrubber (high Fe/Fe + Mg ratio triggers sulfidation and Au deposition; McCuaig et al., 2010). Yet there are many other favorable depositional sites in the Yilgarn, as shown by discoveries of the past three decades noted previously. There have been many versions of mineral systems proposed (cf. Wyborn et al., 1994; Knox-Robinson and Wyborn, 1997; Lord et al., 2001; Price and Stoker, 2002; Kreuzer et al., 2008, 2010; McCuaig et al., 2010; Murphy et al., 2011), all aimed at compositing observations from mineral deposit models to develop a set of broad encompassing principles that could be used for conceptual targeting. The challenge in application of the mineral system approach is typically scale. McCuaig et al. (2010) outlined the scale dependency of targeting parameters, highlighting that processes critical to mineralization on the regional scale have little relevance on the deposit scale and vice versa. Yet this understanding is rarely carried through in targeting exercises. In practice, targeting exercises on the regional scale almost always become a structural-kinematic analysis, intersecting with the spatial distribution of favorable host rocks, as these are usually the only elements we can map with any confidence in our regional datasets. On the belt to prospect scale, targeting exercises also become a chemical (host rock, alteration, metal anomalism) and structural targeting exercise. Human nature draws attention to areas with abundant data and where evidence for mineralization exists (anomalies), irrespective of whether they are likely to host large concentrations of minerals, and biases against data-poor areas with no historic evidence of mineralization (McCuaig et al., 2009). Yet at the scale of a craton, the processes at the site of deposition are not relevant for exploration targeting (McCuaig et al., 2010). Conceptual targeting is difficult at the larger scale because the processes of interest may occur within the deep crust or lithosphere, cannot be easily observed, and are much more uncertain. Therefore, a review of the critical elements of a mineral system is warranted. Giant Ore Systems as Zones of Focused Mass and Energy Transfer What are the critical elements that must come together to form a mineral system? The original definition by Wyborn et al. (1994) emphasized that a mineral system requires sources of the mineralizing fluids and transporting ligands, sources of the metals and other ore components, a migration pathway for the fluids, a thermal gradient to drive fluid flow, an energy source, a mechanical and structural focusing mechanism at the trap site, and chemical or physical traps for mineralization. However, although this is a good description of the various components of a mineral system, this description does not explain why these components (which collectively are common in orogens) have sometimes interacted to form a giant 158 MCCUAIG AND HRONSKY mineral deposit. Because high-quality orebodies are very rare, the coincident conditions that create them also must be rare. with the hottest magma and largest magma channel recorded on Earth (Barnes et al., 1988; Barnes, 2006). The giant Olympic Dam deposit is hosted by a brecciated rock volume of >20 km3 (Ehrig et al., 2012), which must rank it as one of the largest breccia complexes globally. There are some basic physical constraints on what can constitute an ore-forming fluid. First, the fluid needs to have low viscosity to enable significant mass and energy transfer over ore-forming time scales (104−105 yrs, Arribas et al., 1995; Repetski and Narkiewicz, 1996; McInnes et al., 2005; Hickey et al., 2014). There are three important low-viscosity fluids in the crust: water, mafic-ultramafic magmas, and hydrocarbons. Currently economic metal deposits are only known to form from the first two of these, even though the capacity of hydrocarbons to transport metals has been demonstrated (Emsbo et al., 2009). Second, the fluid needs to be available in large quantities over these geologically short time frames. This places significant constraints on what constitutes a viable process to drive ore-forming fluids for many systems. For example, it is now widely accepted that compaction-driven basin dewatering cannot move fluid at high enough rates to form sedimentary basin-hosted metal deposits and that instead topographically driven flow of fluid, sourced from the hydrosphere, is Physical constraints on ore fluids The most essential constraint on the ore formation process is that it must take elements at low concentration in large volumes of source rock and deposit them at high concentration in small volumes of rock (Kerrich, 1983). The only plausible mechanism to do this is through large-scale advective mass flux (Fig. 3). This in turn requires the presence of a transporting medium (hydrothermal fluid or magma, both generically referred to here as “fluid”). This implies that giant ore deposits must be the foci of large-scale systems of energy and mass flux. In many cases of magmatic and hydrothermal deposits, an association between giant ore deposits and large-scale energy flux can be very clearly demonstrated. For example, Norilsk is not only the world’s largest nickel deposit, it occurs within the Permian-Triassic Siberian Traps which is the world’s largest recognized continental flood basalt province and is associated with the largest mass extinction event in the geologic record (Rampino and Strothers, 1988). The Perseverance nickel deposit in the Yilgarn, Western Australia, is not only the largest komatiite-hosted nickel-sulfide deposit, it is associated A B FLUID SINK Energy Sink Deposit Scale Energy Flux released in transient “Avalanches” Threshold Barrier Potential Energy Gradient Self-Organised System Camp Scale Focused fluid exit conduit FLUID RESERVOIR Fluid flow barrier Entropy (exported to environment as diffuse heat) Fluid delivery pathway Energy Flux fed into system at a slow rate Regional Scale Energy Source 1. Map architecture PRIMARY FLUID SOURCE REGION Fig. 3. A. Simple model for a self-organized system. Keys to the generation of self-organized critical behavior are that the energy is continually added to the system and that a threshold barrier is present that prevents the energy from dissipating to the sink. The energy is added continually and slowly over long time frames, whereas the energy release is via transient dynamic “avalanches” of much shorter duration. See text for discussion. B. Translation of the self-organized critical system concept to ore genesis as the focus of a scale-hierarchical mass concentrative fluid (including magma) advection system. Different sets of processes are important at different scales, an aspect that must be taken into account in research and exploration strategy. See text for discussion. 2. Then map chemistry onto architecture THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING required (Ingebritsen and Appold, 2012, and references therein). Metamorphism is another process that generates large amounts of fluid, but does so at a significantly lower rate (ca. 10−10 ms−1; Skelton, 2011) than required for mineralization (Cox, 1999), and contrasts with the faster processes of magmatic devolatilization (10−8 ms−1; Cathles and Shannon, 2007; Simmons and Brown, 2008) or topographically driven basinal flow (10−6 −10−5 ms−1; Cathles, 1990). Third, fluid-flow systems must be highly organized to produce the required extreme concentrations of metal. Organized fluids are those that are delivered in a highly focused manner in both space and time. These conditions cannot be common because most crustal fluid flow and magma systems do not produce ore. Although ore fluids may contain metal concentrations that are one to three orders of magnitude greater than background fluids (Yardley, 2005; Kouzmanov and Pokrovski, 2012), they show a continuum in chemical composition with background crustal fluids that do not form ore (Yardley, 2005; Simmons and Brown, 2008). Thus orebodies must represent moments in time when fluid flow was highly organized (and potentially transiently supersaturated?) and released as transient pulses to dramatically focus the mass and energy through the rock mass and trigger ore deposition. Such extreme fluid flow is characteristic of earthquakes, where transient extreme permeability increases of five orders of magnitude above background have been recorded after seismic events (Miller et al., 2004). The topic of the dynamics of how ore-fluid flow systems organize is addressed below. Mineralization as a product of self-organized critical systems A seminal paper by Bak et al. (1987) proposed a model to explain the physics of complex natural systems that show order and organization, despite a wide range of initial starting conditions. These systems were termed self-organizing critical systems. One of the key examples used by Bak (1996) and Bak et al. (1987) was a geologic one: the patterns of earthquake occurrence. Hronsky (2011) proposed that ore-forming fluid-flow systems can also be considered as self-organized critical systems. Key factors that control self-organized critical behavior are that energy is added slowly and continually to the system and that this energy is prevented from dispersing into a sink by a threshold barrier, thereby creating an enhanced energy gradient (Fig. 3A). This energy gradient builds until it reaches a critical point and overwhelms the threshold barrier. Then energy is released in rapid transient pulses termed avalanches, with the potential size of the avalanches proportional to the energy gradient it seeks to disperse (Bak et al., 1987; Schneider and Dorion, 2005). In Bak’s earthquake example, the energy input is the slow but continual buildup of stress due to the motion of the earth’s tectonic plates. The threshold barrier is the brittle upper crust, which prevents stress release through ductile deformation. The avalanches are earthquakes that release energy through brittle failure of rock, with the size distribution of earthquakes proportional to the stress gradient the system seeks to disperse. As long as the energy gradient and threshold barrier remain intact, the system will remain organized around the critical point (in the earthquake example, the brittle-ductile transition). The avalanches in which energy is released show a distinct set of characteristics: they 159 are multiple events with heavy-tailed size-frequency distributions following scale-invariant power law behavior, are fractal in geometry (in that they show similar patterns over a range of scales), and occur on a vastly shorter time scale than the energy input to the system (Bak et al., 1987; Jensen, 1998). Ore deposits also demonstrate the characteristics of selforganized critical systems. They show power-law size-frequency relationships (Folinsbee, 1977; Schodde and Hronsky, 2006; Guj et al., 2011) and their spatial distributions exhibit fractal geometries (Carlson, 1991). Many mineral deposits present evidence that their formation involved multiple transient pulses of intense fluid flow. This is commonly recorded as multiple-overprinting generations of veins and brecciation, or layers within a single vein (e.g., Sibson et al., 1988; Cathles and Adams, 2005; Cox, 2005). Cathles and Smith (1983) calculated that volumetric flux of brine pulses that formed Mississippi Valley-type deposits were more than three orders of magnitude greater than those which could be produced by steady-state basin dewatering, with the most likely mechanism for these pulses being the rupture of an overpressured fluid reservoir. Consistent with increased fluid flow, some fault-related damage zones show direct evidence that transient, extreme fluid permeability, and fluid flow followed seismic events. For example, in the 1997 Umbria Marche earthquake sequence in northern Italy, transient, postseismic permeability for a localized zone in the hanging wall of a major rupture has been estimated as 4 × 10−11m2; this is 105 to 106 times greater than background crustal permeability at that depth (Miller et al., 2004). At a more regional scale, mineral deposits form in narrow time frames within the longer lasting magmatic-deformation-hydrothermal history (energy input) of the region (Goldfarb et al., 2005, 2014; Sillitoe and Perello, 2005; Sillitoe and Mortensen 2010; Maydagán et al., 2014). If we consider ore-forming systems to be examples of selforganized critical systems, they represent the particular case where energy flux occurs primarily as advective fluid (including magma) flux. Self-organized critical behavior is likely to be essential for ore formation in most magmatic-hydrothermal systems because it is the only viable physical mechanism in the crust to produce the concentrated fluid fluxes generally considered to be required for major metal accumulations (e.g., Cathles and Adams, 2005; Cox, 2005). Background permeability is far too low for convective fluid flow throughout most of the crust (Manning and Ingebritsen, 1999). Implications of self-organized critical systems for mineral exploration Fundamentally, the above discussion implies that generation of extreme and anomalous fluid-pressure gradients is important for producing more focused and larger fluxes of mass and energy. Such regions will have larger size-frequency distributions of energy release events and the greatest opportunity for the formation of giant ore deposits. The most important implication of these ideas for economic geology is that a localized threshold barrier to fluid flow is an essential element of ore-forming systems (Fig. 3B). Although intuition suggests that ore deposits will form in the most dilational parts of the crust, this is not supported by observational data. Instead, evidence from a range of deposit types including orogenic gold, porphyry copper, and magmatic Ni-Cu-PGE 160 MCCUAIG AND HRONSKY sulfide deposits commonly indicates exactly the opposite, with anomalous localized compressional geodynamics and/or barriers to fluid flow playing an important role (e.g., Sibson et al., 1988; Czamanske et al., 1995; Cathles and Adams, 2005; Rohrlach and Loucks, 2005; Begg et al., 2010; Sillitoe, 2010). The threshold barriers in mineral systems can be physical features distal to fluid source regions, such as the crystallizing carapaces of intrusions, antiformal culminations, impermeable stratigraphic layers in basins, and the steep-dipping structural margins of sedimentary basins. The threshold barriers can also be geodynamic, such as an arc undergoing transient pulses of increased compression, causing magmas to pond at depth (Fig. 4; e.g., Rohrlach and Loucks, 2005). In many cases, this threshold barrier will have a clear spatial identity and enclose an overpressured fluid or magma reservoir that is episodically ruptured. An important component of any targeting strategy then becomes the identification of potential paleo-overpressured fluid or magma reservoir sites. These are likely to be much larger targets than the ore environment itself. It is proposed here that the scale of these overpressured reservoirs defines the scale of associated mineral deposit camps (i.e., a cluster of closely related deposits, see below). For example, the typical scale of clusters of porphyry copper deposits is 5 to 30 km, similar to the scale of underlying magmatic reservoirs inferred to drive these systems (Tosdal et al., 2009; Sillitoe, 2010; Fig. 4). The energy-release events in ore-forming self-organized critical systems comprise pulses of overpressured fluid. The self-organized critical system concept predicts that these focused pulses of ore fluid will nucleate at sites where the structural architecture is most conducive to failure driven by fluid pressure build-up (i.e., “the weakest link” in the barrier). These pulses will create their own conduits (although usually, but not always, utilizing existing structural weaknesses) and form pipe-like networks between their source and sink. Thus, the permeability creation and fluid flow is not driven by deformation on active structures (Cox, 2005) but by extreme fluid pressures opening up weaknesses in preexisting architecture beyond the threshold barrier. Depending on the details of the chemistry of ore deposition in a particular system, ore deposits will form either within the conduit or where the fluid discharges into its fluid sink (Figs. 3, 4). There are consistent spatial patterns for large mineral deposits that are commonly observed across a range of mineral deposit types. These include a regional camp-scale spatial association with zones of localized complexity along long-lived, large-scale structures and a direct association between ore-fluid conduits and low bulk-strain fracture networks (Ridley, 1993). It is also quite common for ore deposits widely separated in time to form in the same volume of rock (e.g., Kambalda NiS and St. Ives Au, Miller et al., 2010; Mt. Isa Pb-Zn and Cu, Swager, 1985; Carlin district Au, Bettles, 2002). These coincidences must have an underlying explanation and it is proposed that only certain structural architectures may be favorable for the generation of an ore-forming self-organized critical system. Determining the nature of such architectures should be a target for further research. In summary, some key implications for exploration targeting emerge from this self-organized critical concept, with different relevance at different scales. At the broadest scale, Limit of lithocap v v v v v v v v v v v Base of degraded volcanic edifice Paleosurface v v KEY v Late-mineral Intermineral Early Internal Threshold Barrier (carapace) Fluid Exit Conduit Porphyry Stock Parental pluton Composite precursor pluton v v Comagmatic volcanic rocks Subvolcanic basement 5 Km 5 Km Fluid Reservoir External Threshold Barrier (transient anomalous compression) Fig. 4. Diagram illustrating how porphyry deposits may form in exit conduits above a threshold barrier. In this example the threshold barrier is a combination of a local physical barrier (the carapace of previous crystallized phases of intrusion) and geodynamic (the transient anomalous compression due to increased far-field stress that clamps the vertical permeability of the system, shutting off volcanism and causing magmas to pond at depth). Modified from Sillitoe (2010); see also Figure 9. 161 THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING recognition of transient geodynamic episodes that create threshold barriers is critical. At a camp scale, the ability to define these threshold barriers in space and time becomes an important element in mineral exploration. Finally, at the deposit scale, understanding the 3-D architecture is critical for predicting possible exit conduits. Although ore is formed within exit conduits, it is not formed everywhere within them. Therefore we need to recognize pathways when they are not ore. An important exploration strategy becomes first locating the fluid exit pathways, then finding the ore within them. Critical Elements of a Mineral System McCuaig et al. (2010) proposed a pragmatic methodology for applying the mineral system concept to exploration targeting. The purpose of this method is to relate the highest order, generic elements of a mineral system (termed “critical elements”) to practical observations that can be made in available exploration data sets. The choice of critical elements is important because it organizes the entire process of applying mineral system concepts to exploration targeting. As discussed above, different combinations of critical elements have been applied by various authors and mineral explorationists, including the classic trinity of “source-transport-trap” borrowed from the petroleum industry (Lord et al., 2001; Kreuzer et al., 2008, 2010; McCuaig et al., 2010). Although these are valid, we advocate that the most practical and useful set of critical elements are lithosphere architecture, transient favorable geodynamics, fertility, and preservation of primary depositional zone (Table 1; Fig. 5). In this framework, ore formation results from the conjunction of these critical independent elements. Fertility Favorable Whole-Lithosphere Architecture + Preservation (of Primary Depositional Zone) Favorable (Transient) Geodynamics Ore Genesis Fig. 5. Critical elements of a mineral system. Ore deposition occurs as a conjunction of whole lithosphere architecture, favorable transient geodynamics, and fertility. Postmineralization preservation of the primary depositional zone is a critical element for ore deposit discovery. Table 1. Critical and Constituent Elements in the Formation of Gold Deposits1 Critical elements —————————————————— Scale ————————————————— Fertility Favorable architecture Primary depositional zone Ore-shoot N/A at this scale N/A at this scale Localized dilatant zone in conduit-hosting structure Deposit N/A at this scale Pipelike rock volume more favorable for fracturing by fluid-exit pulse (either local structural complexity or pipe of more competent rock) 2nd order—pressure drops; 3rd order—favorable substrate (chemical reaction) Camp N/A at this scale Period of low active tectonic strain, e.g., stress switch causing transient neutral stress state causing fluid system to selforganize; areas of greatest uplift favored (provides stress switch and high rates of energy and mass transfer) Major heterogeneity (e.g., cross-structure intersection) along trend of inverted rift-axial (or rift-marginal) fault with associated physical seal (e.g., antiformal culmination or unconformity) Province Discrete Au-enriched upper lithospheric domain, particularly near its margins; potentially mantle lithosphere enriched by small volume partial melts prior to termination of orogeny Terminal phase of synore orogenic event (e.g., the transition to incipient extension associated with the termination of collision and locus of subduction retreating oceanward) Inverted retroarc rift; preferably developed at a continental margin, or margin of deep mantle lithosphere root; long-lived “vertically accretive” structure Continental A major collisional orogenic event within the history of an evolving accretional orogen; the major collision that actually terminates a long-lived (>200 Ma) accretionary orogen is most prospective and usually associated with a peak of supercontinent formation Major subcontinental scale lineament (representing longlived zone of transverse dislocation within accretionary orogen); long-lived “vertically accretive” structure Currently unclear but the occurrence of the western US gold superprovince suggests that some control at this scale exists Favorable geodynamics 1st order—upper 10 km of crust at the time of mineralizing event where fluid pressure (+T, X) gradients are greatest; preserved through multiple orogenic cycles Notes: Note the scale-dependent nature of the constituent elements; these elements require translation into features that can be mapped directly in existing or obtainable geoscience datasets at the appropriate scale to generate targets (e.g., McCuaig et al, 2010) 162 MCCUAIG AND HRONSKY Favorable whole lithosphere architecture The control of structure on hydrothermal mineral deposits has long been recognized, from the scale of ore shoots (Conolly, 1936) to the broad regional scale (Billingsley and Locke, 1935, 1941; O’Driscoll, 1986). Traditionally, in the academic economic geology community, the focus on structure has been at the ore shoot-orebody-camp scales, and regional studies have focused on upper-crustal aspects of major fault systems and their role as permeability pathways or structural traps (e.g., anticlines). The most important and consistent structural pattern in giant mineral deposit targeting, however, is a spatial relationship between large deposits and inferred fundamental basement structures. This pattern has been recognized since at least the 1930s and widely applied in the industry (O’Driscoll, 1986; Woodall, 1994). These structures tend to be difficult to recognize in surface geologic mapping, and historically were often only recognized as linear arrays of mineral deposits associated with subtle linear patterns of structural discontinuities (i.e., “lineaments”). Therefore, this association tended to be treated with skepticism by the academic community. However, this changed in the last decade with the availability to the academic community of large-scale geophysical datasets that could image these features to great depth (e.g., Chernicoff et al., 2002; Crafford and Grauch, 2002; Murphy et al., 2008). The giant polymetallic skarn deposit at Antamina, Peru, is an excellent example of deep architecture with only cryptic surface expression (Fig. 6). Antamina formed at the intersection of an apparently minor thrust and a cryptic transfer fault in the western Andes at ca. 9 Ma as a result of multiple stress switches during the Quechuan orogeny (Love et al., 2004; Fig. 6A, B). However, as shown by Love et al. (2004), the faults reactivated a thrust architecture that was established at ca. 42 Ma during the Incaian orogeny. Furthermore, the Incaian orogeny involved reactivation of structures that had been established in the Jurassic, at the time when the host sedimentary sequences were deposited in a back-arc basin off the western coast of South America. Recent work in the Pataz region of the eastern Andes by Witt et al. (2013) indicates that these E-NE-trending orogen-transverse structures were present on the western margin of South America at least as early as 300 m.y. ago. Therefore, the transfer fault is a fundamental lithospheric structure that was active for a long period of time in geologic history, yet is marked only as a series of brittle joints at the scale of the deposit, and by subtle changes in structural trends and thickness of stratigraphic units on a regional scale (Figs. 6C, D, 7). The Motherlode orogenic gold province of California is another excellent example. At surface, the host orogen-parallel fault network is an anastomosing series of moderate displacement high-angle reverse faults. However, this network overlies a fundamental feature in the deep lithosphere as imaged by Bouguer gravity and seismic tomography data, corresponding to the margin between older Precambrian mantle lithosphere to the east and Phanerozoic mantle lithosphere to the west (Bierlein et al., 2008; Griffin et al., 2013). In the northern Chilean porphyry Cu belt, the important ore-controlling Domeyko (or West) fault can be imaged by magnetotellurics as penetrating through the mantle lithosphere (Lezaeta, 2001). It is a small displacement structure at surface, yet marks a fundamental translithospheric structure that has been active periodically since at least the Mesozoic (Padilla-Garza et al., 2001). Whole-lithosphere architecture also exerts control on the location of many deposits traditionally viewed as upper-crustal fluid systems. For example, unconformity uranium deposits in the Athabasca basin align along a reactivated basement structure corresponding to the margin of the Paleoproterozoic Trans-Hudson orogeny (Mercadier et al., 2013). The Century Zn deposit and Kupfershiefer Cu-Pb-Zn deposits have also been demonstrated to be associated with deep-seated structures (Murphy et al., 2008; Borg et al., 2012). Even Precambrian banded iron-hosted Fe oxide deposits, which show evidence for hypothermal upgrades before supergene modification (Angerer et al., 2014), correlate with early lithospherescale architecture (Fig. 8; Mole et al., 2013). More recently there has been the recognition in the academic community (although long known in industry) that the spatial and genetic relationship of deposits to lithospheric structures does not just apply to hydrothermal deposits but also to orthomagmatic deposits at a range of scales (Fig. 8; Begg et al., 2009, 2010; McCuaig et al., 2010; Griffin et al., 2013; Mole et al., 2013, 2014). The clearest example of the control of mantle lithosphere architecture on ore systems is kimberlites. Kimberlites are notorious for having little correlation with upper-crustal structure, but a strong correlation with deep structures, as suggested by tomographic and isotopic images of the mantle lithosphere (Begg et al., 2009; Jelsma et al., 2009). There are a number of key features that these fundamental ore-controlling structures share that are important for targeting: (1) they are both strike and depth extensive, usually penetrating into or through the lithospheric mantle as imaged by geophysical data or interpreted from changes in isotopic composition of magmatic rocks (e.g., Carlin Trend, Crafford and Grauch, 2002; Domeyko fault zone in northern Chile, Lezaeta, 2001; Yilgarn gold and nickel, Mole et al., 2013, 2014; Fig. 8); (2) they are relatively difficult to trace in map patterns (at least at the structural level of ore formation) and are not the obvious structures at or above the level of mineralization (e.g., Antamina, Love et al., 2004; Fig. 6; St. Ives, Western Australia, Miller et al., 2010; Fig. 7); (3) they are never classic flat-dipping thrust zones even though such structures are some of the biggest crustal faults on Earth, and usually not the continuous major shear zones that are obvious in regional maps; (4) they have an anomalously low ratio of displacement to strike length, in comparison to the typical scaling relationships in neoformed faults (cf. Peacock, 2003), implicating formation by reactivation of underlying ancestral structures; (5) they can be demonstrated to be multiply-reactivated (commonly with variable senses of movement) faults with a very long (commonly hundreds of millions of years or more) history (Love et al., 2004; Mole et al., 2014; Miller et al., 2010); and (6) they commonly juxtapose distinctly different basement domains as imaged by magma chemistry (Loucks, 2014) or isotope chemistry (Mole et al., 2013; Fig. 8). These important translithospheric ore-controlling structures have what is termed here as ”vertically accretive growth” histories (Fig. 7). Major structures at depth can be overlain by younger volumes of rock that are deposited or obducted into place; 163 THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING Deposit scale N A B Transfer Zone 1 km 1 km Regional scale D Camp scale 10 km C 81°W Peru 8°S Pasto Bueno Magistral Trujillo LIMA PACIFIC OCEAN 9° 30'S 18°S RD Basin Margin BL A CA ER AN ILL Chimbote Mineral Deposits RD 9°S A CO ER ILL Fig. 6A,B 69°W CO ANTAMINA 0°S GR Casma NE 9°40'S Pierina Fig. 6C Antamina Huaraz A 10°S 77°20'W 77°10'W 77°00'W Pliocene - Quaternary clastic sediments Middle Eocene - Upper Miocene granitic rocks Middle Eocene - Middle Miocene volcanic rocks Albian - Upper Cretaceous carbonate rocks Upper Jurassic - Lower Cretaceous siliciclastic rocks pre-Ordovician Marañón Metamorphic Complex anticline syncline plunge direction thrust fault normal fault strike-slip fault approx. locus of fold plunge changes, and strike of folds and faults 0 100 Huarmey km 79 °W 78° W Cenozoic clastic sediments Lower - Upper Miocene granitic rocks Eocene - Pliocene volcanic rocks Cretaceous - Paleogene intrusive rocks 77°W Albian - Cretaceous carbonate rocks Cretaceous volcanic rocks Mississippian - Cretaceous siliciclastic rocks pre-Ordovician rocks Fig. 6. Multiscale architectural controls on the Antamina polymetallic skarn deposit, Peru. Formed at ca. 9 Ma during a period of multiple stress switches, the deposit lies at the intersection of an inverted margin of a basin-controlling orogenyparallel fault with a transfer zone that has only a cryptic expression at surface but has been active since at least the Mesozoic. A. Premineralization deposit-scale architecture established by 10 Ma (McCuaig et al., 2003). B. Synmineralization intrusion follows preexisting architecture of thrusts and fold axes at mine scale (McCuaig et al., 2003). C. Camp-scale architecture showing cross orogen, vertically accretive structure that has only cryptic expression at surface expressed as discontinuities in structural patterns across it (number of thrusts and folds, bends in fold axes); after Love et al. (2004). D. Regional-scale expression of cross orogen, vertically accretive structures showing control on the distribution of stratigraphic units through time, indicating that the structures have been active at least since deposition of host rocks in the Mesozoic-Cenozoic; after Love et al. (2004). reactivation of the underlying shear zone initially produces complex anastomosing fractures in the overlying rock volume as the structure propagates through the overlying strata. Thus they may have only a subtle en echelon or apparently unconnected (soft-linked) brittle fracture pattern at the current surface of the Earth. Transient favorable geodynamics The increasing availability of high-resolution geochronological data over the last decade or so has led to what we consider one of the most significant scientific discoveries in the field of economic geology: the realization that ore deposit formation occurred during very narrow time intervals within 164 MCCUAIG AND HRONSKY + + + + + + + + + + + + + + + + + V V + + + V + + + + + + + + Re-activation and Upward Propagation of Basement Structure Newly Deposited Volcanic/ Sedimentary Strata Fundamental Basement Structure V V V V + + + V V + + V + + V + + V V + + + + + V V + + + + + + V V + + + + V + + + + GEOLOGICAL TIME Fig. 7. Schematic diagram illustrating the concept of vertically accretive structures. Deeply rooted structures in the lithosphere may have only cryptic expressions at or above the level of ore formation. See text for discussion. Sm-Nd (εNd): -9.3 - -6.2 -6.2 - -3.9 -3.9 - -2.2 -2.2 - -1.0 -1.0 - -0.2 -0.2 - 0.5 0.5 - 0.9 0.9 - 1.6 1.6 - 2.4 2.4 - 3.6 Gold deposit Iron deposit Nickel deposit Fig. 8. Epsilon neodymium isotopic map of the Yilgarn craton of Western Australia at 2.7 to 2.6 Ga, showing distribution of ca. 2.7 Ga nickel deposits, ca. 2.65 Ga gold deposits, and BIF-hosted iron oxide deposits. A strong N-NW-trending gradient in the center of the image is interpreted as representing the margin of the paleocraton at ca. 2.7 Ga, with evolved crust to the west and juvenile crust to the east. Other areas of more juvenile crust in the western portion of the craton are indicated. Gold deposits are largely restricted to the margins of more juvenile domains. Komatiite-hosted nickel-sulfide deposits are also controlled by this whole lithosphere architecture, where high melt volumes were channeled around the SCLM lithospheric root to the palecraton margin. BIF-hosted iron-oxide deposits are restricted to the margins of stable cratonic roots (domains of evolved crust), where thicker BIF sequences were present and fluids were channeled to produce hypogene upgrades of the BIF. Such isotopic maps enable imaging of whole lithosphere architecture through time. After Mole et al. (2013). THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING the much broader evolution of their host terranes. Even more remarkably, in many cases it can be demonstrated that disparate deposit types, commonly separated by large distances (100s to, in some cases, 1,000s of kilometers) form in the same narrow time windows. For example, dating of alteration minerals has narrowed the age of formation of all deposits in the Juneau orogenic gold system in Alaska to a ~1-m.y. interval at about 55 Ma (Goldfarb et al., 1991; Fig. 9E, F). Within the Carlin Trend, Hickey et al. (2014) have shown through apatite fission track dating that the Betze-Post gold deposits (ca. 40 Moz of Au) formed in less than 15 to 45 k.y. Yamada and Yoshida (2011) showed that the Kuroko VHMS deposits formed within a ~1-m.y. interval at about 14 Ma. There are now also several good examples of major oreforming events occurring simultaneously over broad areas. In southeast Australia, orogenic Au mineralization in the western Lachlan orogen of Victoria took place at the same time (440 Ma) as the formation of the porphyry Au-Cu deposits of the eastern Lachlan orogen in New South Wales (Squire and Miller, 2003), which also correlates with the timing of major gold mineralization in northern Kazakhstan (Goldfarb et al., 2014). In the southwestern U.S., all of the Au deposits in the Carlin trend, the Cripple Creek alkalic intrusion-associated Au deposit and the Bingham Canyon porphyry Cu deposits, generally considered to represent very different deposit types, formed in a narrow window of ca. 40 to 30 Ma (Presnell, 1992; Kelley and Ludington, 2002; Muntean et al., 2011). Furthermore, major porphyry deposits in northern Chile (e.g., Chuquicamata, Escondida; Sillitoe and Perelló, 2005) and in southwest China (Lu et al., 2013) also formed in this narrow time period, suggesting that this important event was global in nature. There is only one plausible explanation for these observations: these critical time horizons must reflect unusual regional-scale geodynamic settings that are favorable for mineralization and that these favorable geodynamic settings must be transient, lasting for only short periods of geologic time. Furthermore, the apparently generic link (i.e., independent of deposit type) between periods of favorable geodynamics and ore formation implies that fundamental physical processes must be involved. As predicted, there is an association between these favorable metallogenic time periods and major periods of geodynamic reorganization. In some cases, this can be shown to be at a global scale. For example, in the case of the late Eocene to early Oligocene (40−30 Ma) epoch in the southwestern U.S., the period of ore formation corresponds closely with a major geodynamic transition in this region from a long period of compressional tectonics (the Laramide orogeny) to a period of slab roll-back and extensional tectonics that ultimately produces the Basin and Range province (e.g., Muntean et al., 2011). However, the simultaneous timing of mineralization in northern Chile and southwest China and the fact that this period coincides in time with formation of the bend in the Emperor Seamount-Hawaian Island chain (O’Connor et al., 2013) suggest a more global-scale control. Another example is provided by the Jiaodong gold province of the North China craton and Motherlode orogenic gold province of California, both of which coincide with the emergence of the Ontong-Java Plateau and plate reorganization 165 in the Pacific at ca. 110 Ma (Goldfarb et al., 2014). Rohrlach (2002) elegantly demonstrated that the formation of the giant Neogene Tampakan porphyry and high-sulfidation Cu-Au deposit on Mindanao formed within a few 100-Kyr period during an anomalous peak of compression that was the result of a subduction zone flip. Saunders et al. (2008) showed that the formation of the Miocene northern Nevada Bonanza gold province (including important deposits such as Sleeper) was closely associated with the initial impact of the Yellowstone hot spot; no mineralization is associated with the subsequent hot-spot track. Goldfarb et al. (2005) demonstrated that the formation of the Paleocene Juneau orogenic gold deposits was associated with a large-scale tectonic switch from orthogonally convergent to transpressional tectonics (Fig. 9E, F). The geodynamic context of the period of formation of the Miocene Kuroko VHMS deposit is a tectonic reorganization that terminates back-arc rifting in the Kuroko rift (Yamada and Yoshida, 2011). Jelsma et al. (2009) showed that kimberlite swarms correlate with periods of major tectonic plate reorganization. Rosenbaum et al. (2005) correlated the temporal and spatial distribution of ore deposit formation in the central Andes mountains to the oblique subduction of the Nazca ridge (Fig. 9C, D). It is easier to recognize the precise geodynamic context of these metallogenically favorable events in young rocks. However, Idnurm (2000) showed that the periods of formation of the giant Mount Isa and Olympic Dam Proterozoic metal deposits in Australia corresponded with periods of major “bends” in the apparent polar wander paths for Australia that might plausibly be interpreted to represent major periods of tectonic reorganization. Robert et al. (2005) demonstrated that Archean orogenic gold deposits tend to occur close in time to the termination of their host orogenic belts. What are the processes that link periods of tectonic reorganization with ore formation? This is a critical question and we propose a hypothesis that relates these observations to the self-organized critical system concept discussed above. We propose that these favorable periods are times when the prevailing geodynamic conditions impose strong threshold barriers to fluid flow, causing fluid-flux systems to become highly organized (e.g., Fig. 3; Hronsky, 2011). These are cases where there is both a strong fluid supply and a lack of active, pervasive vertical permeability, the latter of which is due either to tectonic strain that clamps vertical permeability, or simply to the absence of active pervasive fracturing of the rock mass to generate many possible fluid escape paths (Fig. 9B, D, F). In these scenarios, the only way fluids can escape is by the buildup of extreme fluid pressure and subsequent organized and focused fluid expulsion through transient exit pathways, using preexisting weaknesses in the rock mass. This situation leads to the extreme fluid focusing over short timeframes (104−105 yrs) required to form ore deposits and is likely quite rare in the history of orogens. Central Italy provides a good example for the regional-scale relationship between fluid-flux patterns and tectonic regime. This region is characterized by a significant background flux of mantle-derived CO2. Chiodini et al. (2004) presented the first map of regional-scale variation in this CO2 flux and showed that a broad, diffuse flux characterized the extensional setting of the Tyrrhenian back-arc basin. The anomalous flux of CO2 166 MCCUAIG AND HRONSKY A B Bismarck Sea Spreading Ridge Inactive Outer Melanesian Arc Transient zone of incipient rifting (determines extent of “camp”) Established rifting limited high-quality ore formation Epithermal (e.g. Lihir) Sea level Lihir Grasberg Ok Tedi Porgera VMS (e.g. Solwara) Limited active permeability creation Extreme energy and fluid pressure gradients produced System selforganises and high-quality ore bodies produced Solwara-1 Crust SCLM Melt zone Small volume alkalic melts Asthenosphere C 80 W 300 Inverted orogen-parallel vertical Orogen-parallel limit of accretive structure controls ‘camp’ anomalous compression Fig 1 limits extent of ‘camp’ Anomalous transient compression Crust Moho Magma ponds at depth: volatiles + Oc ea metal content nic increases pl Large Hg deposit tion 10 S D Large Cu deposit duc Km 70 W KEY Large Zn deposit sub 0 8 - 6 Ma e zon Arc volcano ate / ri dg e au Ri OLM e LAB Melt zone zc Asthenosphere Moho LAB Na E Vertical permeability ‘clamped’ Extreme vertical energy gradient produced SCLM a 15 S dg Fig 4 F Fig. 9. Examples of common transient dynamic settings that trigger ore formation. A. Incipient rifting in the Bismark Sea triggering both Au-rich VMS deposits (Solwara) and epithermal Au (Lihir). After Hronsky et al. (2012). B. A schematic section showing how incipient rifting causes self organization of the fluid-flow system. C. Reconstruction of the subduction history of the Nazca Ridge (thick black outline) at 8 to 6 Ma imposing transient anomalous compression in the arc, cessation of volcanism, and the formation of ore deposits. Symbols represent volcanoes and ore deposits formed at this time. Note that this wave of anomalous compression and accompanying mineralization has migrated along the arc due to the oblique subduction of the ridge. From Rosenbaum et al. (2005). D. Schematic cross section illustrating how transient anomalous compression causes self organization of the fluid flow system. Relative position of Figures 1 and 4 are shown. LAB = lithosphere-aesthenosphere boundary, OLM = oceanic lithospheric mantle, SCLM = subcontinental lithospheric mantle. E and F. Example of a stress switch controlling Au mineralization in the Juneau belt in Alaska, after Goldfarb et al. (2005). During the stress switch, active permeability ceases, high fluid pressure gradients form, and the system self organizes to form ore. After the stress switch, active permeability creation resumes in the new stress regime, and ore formation ceases. Common to all of these situations is a transient period where vertical permeability is clamped or active permeability creation ceases, yet energy and fluid is still supplied to the system. The result is the build-up of extreme fluid pressure and energy gradients, triggering extreme self-organized critical behavior and the formation of ore during energy avalanches. The areal extent of the threshold barrier intersecting vertically accretive structures determines the scale of camps. 167 THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING Diffuse Low-Grade Metal Anomalism Ore Formation Diffuse Low-Grade Metal Anomalism SINK SINK SINK FLUID FLUX Transient focused ore-forming flux FLUID FLUX is absent in the adjacent Apennine orogenic belt even though it is the cause of most of the seismicity which characterizes this part of Italy. This is because at depth, the CO2 accumulates in crustal traps, generating overpressurized reservoirs which induce seismicity. This is analogous to the self-organized critical process discussed above. At least three physical scenarios are relevant to this discussion (Fig 9): (1) the initial stages of extensional events and/ or plume/hot-spot impacts (e.g., Fig. 9A, B; Kelley and Ludington, 2002; Saunders et al., 2008) are settings where pervasive crustal-scale permeability has not yet been established; (2) transient anomalous compression, which effectively seals an active system (e.g., Fig. 9C, D; Rohrlach, 2002; Rohrlach and Loucks, 2005), is a setting that is important in forming porphyry deposits; and (3) switches in the prevailing far-field stress that transiently result in a neutral stress field and result in the absence of pervasive deformation-induced permeability (Richard Tosdal; pers. commun., 2009). All of these scenarios are likely to occur in the context of much larger scale patterns of geodynamic reorganization. The significance of threshold barriers may also explain the general absence of significant mineralization from strongly extensional environments, which lack the large threshold barriers compared to those established in compressional settings. A well-documented example is the Andean copper belt, where the three most productive belts with the largest and highest grade deposits correlate with brief periods of anomalously strong compression. The deposits associated with periods of more extensional and less compressional tectonics are smaller and of lower grade (Sillitoe and Perelló, 2005). In this context, it is important to distinguish between actively extending and previously extended settings. Many deposits form in rifts but this does not necessarily indicate they form during periods of significant active rifting. For example, although VHMS deposits are considered to be rift associated, deposits within the classic Neogene Kuroko VHMS province (arguably among the best constrained because of their young age) formed during a period of geodynamic reorganization that actually marks the termination of the extensional history of its host rift (Yamada and Yoshida, 2011). This hypothesis of transient geodynamic triggers for ore formation predicts that for most of the history of a potential oreforming system, ore deposits do not form and that the only manifestation of the system might be broad regional alteration and metal anomalism (Fig. 10). When the geodynamic threshold barrier is established, extreme fluid pressure gradients are produced, and the system self organizes to form ore (Figs. 3, 9B, D). When the barrier is removed, ore formation ceases and the system reverts to broad regional fluid flow, alteration, and metal anomalism (Fig. 10). This punctuated history of ore formation within a broader background context is best documented for the Central Andean porphyry province. Although this region has continuously been an active site of subduction and arc magmatism since at least the Mesozoic, porphyry Cu mineralization in Chile is restricted to several brief windows on the order of millions of years, preceded and succeeded by longer periods of barren magmatism, deformation, fluid flow, and alteration (Sillitoe and Perelló, 2005). Importantly, this model for fluid flow implies that broad regional alteration and metal anomalism, currently considered a vector to ore, may THRESHOLD BARRIER SOURCE SOURCE SOURCE Temporal Evolution of System Fig. 10. Schematic diagram showing that ore deposition occurs during brief intervals within much longer episodes of deformation, magmatism, fluid flow, and alteration that is largely barren. Mineral systems only transiently self organize to form ore. have no direct relationship to the formation of high-quality ore deposits. The system needs to have undergone self organization to form high-quality ore. Therefore, exploration geologists must focus on identifying at various scales the factors that represent geodynamic threshold barriers and the resulting highly organized fluid flow that forms high-quality ore. The key to applying the geodynamic concept in regional targeting is to identify geologic proxies for metallogenically favorable geodynamic epochs. The most practical way to do this is to determine the spatial distribution of rocks that most closely correlate in time with the mineralizing event of interest. For example, Robert et al. (2005) discussed the spatial and temporal relationship between formation of late tectonic sedimentary basins and Archean orogenic gold deposits. Although these sedimentary rocks are not necessarily host rocks to the gold, they represent the unit that formed closest in time to the gold deposits. In this example, these deformed late tectonic sedimentary rocks are the best available proxy for synmineralization tectonic activity. This concept implies that improved precision in geochronology will increase our predictive targeting capability. Within compressional tectonic scenarios, there is commonly an empirical correlation between uplift and the location of large mineral systems. For example, within the young porphyry belt of West Papua New Guinea, the Grasberg deposit (largest Cu deposit in the belt) is located near the highest elevation. Similarly, the Miocene porphyry systems of Central Chile are located near the highest elevations within the Andes Mountains. These areas of highest elevation may indicate areas of stronger anomalous compression. Ancient deposits may show similar relationships; for example, the Kalgoorlie and St. Ives goldfields in Western Australia are located along the axis of maximum uplift, as measured by the exposure of basal stratigraphy in the middle of the Kalgoorlie terrane (Robert et al., 2005). Similarly, the Timmins gold camp in the Superior province of Canada is located where the lowermost stratigraphy is uplifted and structurally repeated in the center of the Abitibi southern volcanic zone, indicating an area of major uplift along the translithospheric PorcupineDestor fault (Robert et al., 2005). In these ancient cases, areas of relatively higher uplift within the primary depositional zone 168 MCCUAIG AND HRONSKY may have corresponded with areas of higher elevation and stronger anomalous compression. Fertility It has become increasingly obvious that there are specific tectonic scenarios that are triggers for major ore-forming events. Yet during these events, not all regions contain worldclass mineralization. Therefore, another critical element must be present, and that is—fertility. Fertility is defined here as the tendency for a particular geologic region or geologic time period to be systematically better endowed than otherwise equivalent geologic environments. Fertility is usually the highest order (largest scale) control on endowment potential. Four factors may contribute to fertility: secular Earth evolution, lithosphere enrichment, large-scale geodynamic processes, and paleolatitude (in specific cases). Secular earth evolution: Secular patterns in ore deposit distribution have long been recognized (Turneaure, 1955; Meyer, 1981, 1985, 1988; Barley and Groves, 1992; Goldfarb et al., 2010; Cawood and Hawkesworth, 2013; O’Neill et al., 2013b) and are summarized in a special issue of Economic Geology (Goldfarb et al., 2010). Key drivers of secular variation in mineral systems include cooling of the earth, evolution of the atmosphere-hydrosphere, evolution of the biosphere, and evolution of global geodynamics and supercontinent cycles. The evolution of the Earth’s lithosphere-hydrospherebiosphere-atmosphere controls the availability or mobility of metals (Hazen et al., 2008; Hazen, 2010), and the geodynamic triggers for fluid/magma flow and mineralization. Although there are conflicting arguments, planetary physical models and the rock record suggest that the Earth has been cooling over time; for example, since the Archean, there has been a decrease in the abundance and calculated melting temperature of ultramafic magmas reaching the earth’s surface (Davies, 1995). The mineral systems most directly affected by this cooling are orthomagmatic deposits such as NiS and PGEs. For example, komatiite-hosted NiS deposits first enter the rock record at ca. 3.0 Ga, peak at 2.7 Ga, and are not found in rocks younger than 1.9 Ga (Naldrett, 2010). The atmosphere and hydrosphere have undergone dramatic and rapid changes through Earth’s history. A rise in oxygen occurred at the end of the Archean (the Great Oxidation Event) and again at the end of the Neoproterozoic (Kump, 2008; Farquhar et al., 2010). This oxygenation of the atmosphere had three major effects on mineral systems, including influencing (1) the available valence states for metals of interest, (2) the stability of host minerals, and (3) the availability of ligands (Hazen et al., 2008; Hazen, 2010). For example, uranium is generally mobile in +6 valence state, and immobile in +4 valence state. Prior to the Great Oxidation Event, primary uranium minerals (e.g., pitchblende) were stable in the relatively reducing hydrosphere and atmosphere, and as a consequence, survived in placer deposits (e.g., Witswatersrand, Elliot Lake). After the Great Oxidation Event, primary uranium minerals were subject to weathering and oxidation, and uranium with +6 valence became available for transport in an oxidizing hydrosphere (Cuney, 2010). It is therefore no coincidence that the highest grades of uranium globally (unconformity uranium deposits) occur in Paleoproterozoic basins. This was the first time that major large sedimentary basins were stable (at the terminal stage of a supercontinental cycle) with U-rich detritus and basement rocks that had never before been leached of their uranium. The oxidation of the hydrosphere and particularly its effect on the redox state of sulfur had a major impact on the availability of metal transporting ligands (Farquhar et al., 2010). Before the Great Oxidation Event, the atmosphere and hydrosphere were relatively reduced and the hydrosphere was Fe rich, with sulfur largely sequestered as sulfide in shales. After the Great Oxidation Event, the oceans were scrubbed of their iron and there was enhanced oxidation of sulfides in the crust that provided sulfate to the hydrosphere. This chemical transition had a major effect on some mineral systems. For example, an important factor in the development of large clastic-dominated or SEDEX Pb-Zn deposits after the Great Oxidation Event was the compositional and redox gradients in the ocean and the availability of metals and sulfate provided by the oxidative weathering of the crust (Leach et al., 2010). Evolution of the biosphere was coupled to the evolution of the atmosphere and was characterized by periodic “spurts” of life. Examples include the organic bloom at 2.2 to 2.0 Ga, marked by widespread generation of organic-rich shales (Papineau, 2010), and the emergence of reef-building carbonate-secreting organisms in the Phanerozoic (Leach et al., 2010). The former had an effect on the expression of sediment-hosted deposits, forming thick sequences of reduced shale. Moreover, these organic-rich rocks may have preferentially sequestered metals from the seawater column (Tomkins, 2013), subsequently forming preferential source rocks for metals when buried and metamorphosed (Tomkins, 2013). The latter formed host carbonate sequences for MVT-style deposits (Leach et al., 2010). Lithosphere enrichment: Some areas of the lithosphere appear to be intrinsically enriched in a target metal. This concept has long been applied empirically in the mineral exploration industry. It is clearest in systems where metal is sequestered from nearby protoliths, such as a uranium-rich hinterland for sediment-hosted uranium mineralization, or substantial sequences of banded iron formation for sedimenthosted Fe oxide deposits. Some researchers have also speculated on lithosphere enrichment for more complex mineral systems. For example, Hodgson (1993) suggested that gold endowment in Archean greenstone belts was related to the abundance of komatiites that preenriched the succession in gold. The first rigorous analysis of the lithosphere enrichment concept was provided by Titley (2001). Further strong support for this idea was provided by Sillitoe (2008), who showed that the bulk of the gold endowment of the American (north and south) Cordillera, regardless of age, is restricted to a few segments of this long-lived orogenic system of approximately 20,000-km strike length. Hronsky et al. (2012) developed these ideas further, proposing an integrated process model that related regions of long-term metasomatic enrichment of gold in the lithospheric mantle to regions of persistent upper crustal Au endowment. Pettke et al. (2010) have also convincingly demonstrated that lithosphere enrichment may be relevant to the genesis of the giant Bingham Canyon porphyry Cu deposit. Zhang et al. (2008) and Griffin et al. (2013) have suggested that lithospheric mantle enrichment may also play a key role in the formation of orthomagmatic Ni-Cu-PGE 169 THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING deposits, although this currently remains controversial (cf. Arndt, 2013). Whatever the precise mechanisms, it appears that large-scale lithosphere control on fertility is real. This enrichment is independent of deposit type, in that different styles of mineralization are generated from the same seemingly enriched volume of lithosphere in many cases, e.g., southwest U.S., eastern Papua New Guinea (Fig. 9A; Sillitoe, 2008; Hronsky et al., 2012). The enrichment process also transcends association with any one ore-forming event, as some terranes show multiple periods of mineralization in the same location (Robert et al., 2005; Hronsky et al., 2012). The implications of preferential lithosphere fertility are significant for our understanding and targeting of ore deposits. For regions where lithosphere fertility may be a factor, metallogenic potential is not solely a property of a particular magmatic or tectonostratigraphic event. Figure 11 shows this concept schematically. For these regions, we can no longer rank an entire belt or province as prospective or nonprospective. Instead, areas along strike of a poorly endowed segment of a belt might host large ore systems, and vice versa, if there is a significant change in the fertility of the underlying lithosphere. These zones of differing fertility will be bounded and defined by the whole lithosphere architecture, hence the importance of constructing maps of the deep lithosphere to aid mineral potential assessment. Fertility and large-scale geodynamic context: It is instructive to view the critical element of fertility in a geodynamic context. Concomitant with the cooling of the Earth has been a change in the mechanism by which the Earth transfers its heat to space. It has been increasingly accepted that the early Earth consisted of a relatively stagnant lithosphere periodically breached by plume tectonics and affected by meteorite bombardment, which eventually gave way to more steady-state Infertile Lithospheric Domain subduction and plate tectonics (Debaille et al., 2013; O’Neill et al., 2013a, b). Within this evolution the earth has undergone major geodynamic cycles, often referred to as supercontinent cycles, although their exact nature prior to 2.7 Ga is strongly debated (O’Neill et al., 2013a). The link between mineral deposits and such cycles has been recognized and gave rise to the first empirical observations of the association between different mineral deposit types and specific tectonic regimes (Meyer, 1981, 1985, 1988; Barley and Groves, 1992; Groves et al., 2005; Kerrich et al., 2005; O’Neill et al., 2013b). Examples include the association of global ”orogenic” Au formation with the terminal stages of accretionary orogenic events and supercontinent formation (Goldfarb et al., 2005; Groves et al., 2005; Kerrich et al., 2005; Robert et al., 2005) and the association of mafic-hosted NiS deposits with peaks in supercontinent formation (Begg et al., 2010). The most important period of orthomagmatic sulfide mineralization and associated plume activity (e.g., Norilsk, Russia, and the Emeishan flood-basalt associated deposits of southwest China) and the most important period of orogenic gold deposit formation (e.g., the Altaids province of Central Asia) in the Phanerozoic occurred broadly coeval with the final assembly of Pangea (ca. 300−250 Ma). In the case of orthomagmatic deposits, this association was most likely related to the accumulation of subducted lithosphere on the core-mantle boundary during supercontinent assembly, which was subsequently reactivated as a series of major plumes (Begg et al., 2009, 2010). In the case of orogenic gold provinces, the precise reason for this particular geodynamic association remains unclear. An example of a long-lived fertile geodynamic setting is the western margin of South America, beginning in the mid-Cretaceous. Rapid opening of the Atlantic ocean in the mid Cretaceous (~100 Ma) resulted in the Andean margin switching Strongly Fertile Lithospheric Domain Weakly Fertile Lithospheric Domain Orogen-parallel vertically accretive structure BARRE N Outline of Metallogenic Province (eg, Magmatic Arc) STRONGLY ENDOWED WED Y ENDO WEAKL BARRE Orogen-transverse vertically accretive structures N STRONGLY ENDOWED WED Y ENDO WEAKL Cover Fig. 11. Schematic diagram illustrating how fertility affects targeting strategy. 170 MCCUAIG AND HRONSKY from an extensional (i.e., formation of the Tarapaca back-arc basin during the Jurassic) to an increasingly compressional mode (e.g., Chen et al., 2013). The western margin of South America is currently characterized by the largest extent of flat subduction on Earth (Lallemand et al., 2005; Syracuse and Abers, 2006), which exemplifies the extent of anomalous compression. Following this tectonic switch, a series of arc-related Cu-Au metallogenic events occurred along this margin (e.g., Sillitoe and Perelló, 2005). Each of the metallogenic events correlates with a brief local period of even stronger compression due to factors such as subduction of anomalously thick or buoyant lithosphere (e.g., Fig. 9C; Rosenbaum et al., 2005). Therefore, the metallogenic events may be controlled by nested scales of geodynamic control: (1) a broadly anomalous compressional plate margin, and (2) local, transient increased compression resulting in clamping of vertical permeability, extreme self-organized critical behavior, and high-quality ore formation (Fig. 9C, D). Paleolatitude: Paleolatitude may have an effect on fertility for some mineral systems, particularly those that require highsalinity fluids. For example, Leach et al. (2010) showed that Phanerozoic clastic-dominated Pb-Zn deposits, when spatially reconstructed to time of formation, dominantly formed between 5° to 30° of the equator—the ”arid zone” with high evaporation rates conducive to the formation of evaporites and high-salinity brines. These dense residual brines penetrate deep into the crust and leach metals along their flow paths until expelled by tectonic triggers where they deposit metals upon encountering reducing horizons (e.g., organic matterrich shales). High-salinity ore fluids responsible for deposition of other deposit types may have formed by leaching of evaporites by meteoric fluids after evaporite formation. Regardless of how such brines formed, there is increasing evidence of a role for evaporate-related brines in other near-surface metal accumulations such as BIF-hosted Fe deposits (Evans et al., 2013), or sediment-hosted uranium systems (Kish and Cuney, 1981; Richard et al., 2011). Preservation of primary depositional zone There are factors about ore deposition that are common for a range of deposit styles, within and among commodity types. In both hydrothermal and magmatic systems, metal deposition involves destabilization of the metal-ligand complexes in the transporting fluid. This destabilization is generally triggered by a physical or chemical change, for example, pressure drop, cooling, or rapid interaction with other rocks or fluids out of equilibrium with the transporting fluid (e.g., M ­ cCuaig and Kerrich, 1998). The focus of metal deposition rapidly into a small volume of rock involves large P-T-X gradients over relatively short distances and times (e.g., Knox-Robinson and Wyborn, 1997). Therefore, the primary depositional zone for most metal systems resides in the upper portions of the Earth’s crust, broadly defined as the upper 10 km, as this is the zone with both the highest geothermal and pressure gradients (and in particular the zone where phase separation becomes more common in magmas and fluids to drive dramatic chemical changes in the transporting media). Within this primary depositional zone, high-quality ore formation may be a natural consequence of the self-organized critical nature of the system. In other words, if enough mass and energy is focused in avalanches above a threshold barrier, metal will probably deposit somewhere along the exit pathway. Thus, determining sites of self-organized critical behavior is the critical goal in regional targeting. Preferred depositional sites are only pertinent at the camp to prospect scales (Table 1; McCuaig et al., 2010). In addition to predicting potential sites of mass and energy transfer, self-organized critical behavior, and potential formation of high-quality ore deposits, it is necessary to know if the potential deposits are exhumed close to the current surface of the Earth but not exhumed to the extent that they are eroded and not preserved. Currently available exploration technology dictates that, outside the immediate vicinity of major mines, deposits can only be located if they are present within a few hundred meters of the land surface. Given that most metal deposits form at significantly greater depths below their paleosurface, this implies that mineral discovery also requires a favorable postdepositional geologic history. A general empirical observation is that different deposit styles have different preservation potentials. For example, Kessler and Wilkinson (2006) demonstrated that an increasing statistical probability of erosional removal with age is consistent with the temporal distribution of porphyry and epithermal deposits (which are largely restricted to the Cenozoic). However, this does not explain the temporal distribution of orogenic gold and VHMS deposits, which instead may be related to the different tectonic settings of formation of these different deposit types (see summary in Hronsky et al., 2012). Those deposits that form in actively uplifting regions (e.g., porphyries) have low-preservation potential, whereas those deposits that form in rifts (e.g., VHMS deposits) or at significant depths late in the history of the orogen (e.g., orogenic gold deposits) have relatively higher preservation potential (Groves et al., 2005; Kerrich et al., 2005). Lithospheric architecture is also an important factor in deposit preservation. Deposits that either form within thick stable cratonic lithosphere or are incorporated into such lithosphere at the end of their host orogenic cycle, without exhumation and erosion, are likely to have a high probability of long-term preservation (e.g., Groves et al., 2005; Kerrich et al., 2005). Implications for Grouping of Ore Deposits to Aid Targeting One important implication of the mineral system concept is how it highlights connections between different deposit types. Of critical importance is to recognize that apparently disparate deposit types can exist as a continuum, related in a predictable way based on the processes controlled by the host geodynamic environment. For example, it has become increasingly clear that a continuum exists between rift-related epithermal deposits and VHMS deposits, relating to degree of extension and water depth in the host rift (Fig. 9A, B; see summary in Hronsky et al., 2012). The progression from epithermal to VMS-epithermal hybrid to classic VMS mineralization with increasing crustal thinning and water depth is exemplified by the Alexander Triassic metallogenic belt in Alaska and British Columbia (Taylor et al., 2008). An important pragmatic implication of understanding this particular continuum is that Aurich hybrid VMS-epithermal systems, such as Henty (Halley and Roberts, 1997) or Eskay Creek (Sherlock et al., 1999), are THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING likely to be associated with the transition from the epithermal to VMS environment. Viewing ore deposits within a broader geodynamic context allows large-scale metallogenic patterns to be discerned. For example, MVT Pb-Zn deposits and orogenic gold deposits would normally be considered very different ore types and almost certainly have very different fluid sources. However, because MVT deposits form in orogenic forelands, which represent the more distal parts of orogens (Bradley and Leach, 2003), and orogenic gold deposits form in the more proximal parts of orogens (e.g., Goldfarb et al., 2005), it is plausible that both deposit types could have formed broadly synchronously during the same metallogenic event. Possible examples of such are MVT deposits of the Earaheedy basin of Western Australia that have been dated at about 1.79 Ga, which is also the main period of Paleoproterozoic orogenic gold mineralization in Western Australia (Pirajno, 2004). Similarly, the Solwara Au-rich sea-floor VMS deposit and the giant epithermal Au Lihir deposit, two very different deposit types, are potentially related to the same volume of enriched lithosphere and same transient geodynamic trigger (Hronsky et al., 2012). Conclusions The understanding of mineral deposits has evolved from describing ore specimens at hand-sample scale, to identifying specific favorable host rocks, to determining structural control on fluid flow (Conolly, 1936; McKinstry, 1941, 1955), and to understanding broad metasomatic processes (Korzhinskii, 1968; Bohlke, 1989). This evolution led to the development of analog deposit models that incorporate the processes of ore formation. More recently, deposit models have evolved into a systems-based understanding of mineralization processes within the larger context of planetary geodynamics and secular Earth evolution (Groves et al., 2005; Kerrich et al., 2005; Goldfarb et al., 2010; Leach et al., 2010; Cawood and Hawkesworth, 2013; O’Neill et al., 2013b). Thus, the understanding of the mineral system has evolved from observation to taxonomical classification to a predictive understanding of processes, much like that for many other natural systems (e.g., tectonics, biology). The mineral systems framework presented here builds on decades of mineral deposit research. This framework views ore-forming processes as a conjunction of the critical elements of metal fertility, whole lithosphere architecture, and transient geodynamic events to trigger organization of extreme fluid-flow and metal deposition, followed by preservation of the primary depositional zone to allow discovery (Table 1; Fig. 5). These critical elements operate in nested scales of both time and space. This mineral systems approach has significant predictive power compared to the traditional deposit model paradigm, particularly as mineral exploration moves into regions under cover and to greater depths. It is hoped that the mineral system concept will inform both future exploration strategies and research efforts. From a research perspective, the mineral systems approach allows identification of the highest value fields on which to focus limited intellectual and financial resources. Key areas that emerge are (1) better integration of geology and geophysics to image multiscale whole lithosphere architecture; (2) better methodologies for determining the time scales of mineral 171 systems (geochronological methods)—and the geodynamic triggers to ore-forming episodes that transect commodities and deposit types; (3) a better understanding of terrane fertility, that is, the regions and time periods of Earth’s geologic record that are more conducive for mobilizing large volumes of metals in fluids; and (4) distinct methods for identifying self-organizing critical systems, and for understanding how fluid flow is organized in the crust to generate ore deposits. A particularly fruitful avenue of research is on the use of detrital minerals at an early stage of mineral exploration to identify transient geodynamic events, lithosphere character, and terrane fertility. From an exploration perspective, emphasis should be placed on identification of multiscale footprints of mineral systems that are nested in space and time, which can be mapped in geoscience datasets at early stages of exploration and in covered terranes where traditional deposit-scale footprints are not obvious with current detection technologies. These mineral system footprints will have different expressions at differing scales, and therefore require the collation of primary datasets that can map the critical elements of the mineral system at the scales appropriate to the exploration decision being considered. Acknowledgments The authors wish to acknowledge input and discussions with colleagues over the years, including Graham Begg, Nick Hayward, Steve Beresford, John Miller, Marco Fiorentini, the past and current staff and students of the Centre for Exploration Targeting, the ARC Centre of Excellence for Core to Crust Fluid Systems, CSIRO Minerals Down Under Flagship, and the predictive mineral discovery Cooperative Research Centre. In particular, the authors acknowledge the influence of Roy Woodall, David Groves, and Rob Kerrich in the application of multiscale mineral-systems thinking to ore deposit geology. David Leach and Neil Williams are thanked for helpful reviews of earlier versions of the manuscript, and formal reviews from Karen Kelley, Dick Tosdal, and Noel White greatly improved the manuscript. McCuaig acknowledges receipt of ARC Linkage grant LP110100667. This is contribution 467 from the ARC Centre of Excellence for Core to Crust Fluid Systems (http://www.ccfs.mq.edu.au). REFERENCES Agricola, G., 1556, De Re Metallica: Translation from the original in Latin into English by Hoover, H.C., and Hoover, L.H., 1950: New York, Dover Publications, 638 p. Angerer, T., Duuring, P., Hagemann, S.G., Thorne, W., and McCuaig, T.C., 2014, A mineral system approach to iron ore in Archaean and Palaeoproterozoic BIF of Western Australia: Geological Society, London, Special Publication 393, DOI 10.1144/SP393.X. Arribas, A., Jr., Hedenquist, J.W., Itaya, T., Okada, T., Concepcion, R.A., and Garcia, J.S., Jr., 1995, Contemporaneous formation of adjacent porphyry and epithermal Cu-Au deposits over 300 ka in northern Luzon, Philippines: Geology, v. 23, p. 337–340. Arndt, N., 2013, The lithospheric mantle plays no active role in the formation of orthomagmatic ore deposits: Economic Geology, v. 108, p. 1953–1970. Bak, P., 1996, How nature works: the science of self-organized criticality: Copernicus, New York, Springer-Verlag, 212 p. Bak, P., Tang, C., and Weisenfeld, K., 1987, Self-organized criticality: An explanation of 1/f noise: Physical Review Letters, A38, p. 364–374. Barley, M.E., and Groves, D.I., 1992, Supercontinent cycles and the distribution of metal deposits through time: Geology, v. 20, p. 291–294. 172 MCCUAIG AND HRONSKY Barnes, S.J., 2006, Komatiite-hosted nickel sulfide deposits: Geology, geochemistry, and genesis: Society of Economic Geologists Special Publication 13, p. 51–97. Barnes, S.J., and Lightfoot, P.C., 2005, Fformation of magmatic nickel sulfide ore deposits and processes affecting their copper and platinum group element contents: Economic Geology 100th Anniversary Volume, p. 179–214. Barnes, S.J., Hill, R.E.T., and Gole, M.J., 1988, The Perseverance ultramafic complex, Western Australia: Product of a komatiite lava river: Journal of Petrology, v. 29, p. 305–331. Begg, G.C., Griffin, W.L., Natapov, L.M., O’Reilly, S.Y., Grand, S., O’Neill, C.J., Hronsky, J.M.A., Poudjom Djomani, Y., Deen, T., and Bowden, P., 2009, The lithospheric architecture of Africa: Seismic tomography, mantle petrology and tectonic evolution: Geosphere, v. 5, p. 23–50. Begg, G.C., Hronsky, J.M.A., Arndt, N.T., Griffin, W.L., OReilly, S.Y., and Hayward, N., 2010, Lithospheric, cratonic and geodynamic setting of NiCu-PGE sulfide deposits: Economic Geology, v. 105, p. 1057–1070. Bettles, K., 2002, Exploration and geology, 1962 to 2002, at the Goldstrike Property, Carlin Trend, Nevada: Society of Economic Geologists Special Publication 9, p. 275–298. Bierlein, F.P., Northover, H.J., Groves, D.I., Goldfarb, R.J., and Marsh, E.E., 2008, Controls on mineralization in the Sierra Foothills gold province, central California, USA: a GIS-based reconnaissance prospectivity analysis: Australian Journal of Earth Sciences, v. 55, p. 61–78. Billingsley, P., and Locke, A., 1935, Tectonic position of ore districts in the Rocky Mountain region: American Institute of Mining and Metallurgical Engineers Transactions, v. 115, p. 59–68. ——1941, Structure of ore districts in the continental framework: American Institute of Mining and Metallurgical Engineers Transactions, v. 144, p. 9–64. Bohlke, J.K., 1989, Comparison of metasomatic reactions between a common CO2-rich vein fluid and diverse wall rocks: Intensive variables, mass transfers, and Au mineralization at Alleghany, California: Economic Geology, v. 84, p. 291–327. Borg, G., Piestrzyński, A., Bachmann, G.H., Püttmann, W., Walther, S., and Fiedler, M., 2012, An overview of the Kupferschiefer deposits: Society of Economic Geologists Special Publication 16, p 455–486. Bradley, D.C., and Leach, D.L., 2003, Tectonic controls of Mississippi Valleytype lead-zinc mineralization in orogenic forelands: Mineralium Deposita, v. 38, p. 652–667. Campbell, J.D., 1990, Hidden gold: the Central Norseman story: Australian Institute of Mining and Metallurgy Monograph 16, 68 p. Carlson, C.A., 1991, Spatial distribution of ore deposits: Geology, v. 19, p. 111–114. Cathles, L.M., 1990, Scale and effects of fluid flow in the upper crust: Science, v. 248, p. 322–329. Cathles, L.M., and Adams, J.J., 2005, Fluid flow and petroleum and mineral resources in the upper (<20-km) continental crust: Economic Geology 100th Anniversary Volume, p. 77–110. Cathles, L.M., and Shannon, R., 2007, How potassium silicate alteration suggests the formation of porphyry ore deposits begins with the nearly explosive but barren expulsion of large volumes of magmatic water: Earth and Planetary Science Letters, v. 262, p. 92–108. Cathles, L.M., and Smith, A.T., 1983, Thermal constraints on the formation of Mississippi Valley-type lead-zinc deposits and their implications for episodic basin dewatering and deposit genesis: Economic Geology, v. 78, p. 983–1002. Cawood, P.A., and Hawkesworth, C.J., 2013, Temporal relations between mineral deposits and global tectonic cycles: Geological Society, London, Special Publications, first published October 24, 2013; DOI 10.1144/SP393.1 Chen, H., Cooke, D.R., and Baker M., 2013, Mesozoic iron oxide coppergold mineralization in the Central Andes and the Gondwana Supercontinent breakup: Economic Geology, v. 108, p. 37–44. Chernicoff, C.J., Richards, J.P., and Zappettini, E.O., 2002, Crustal lineament control on magmatism and mineralization in northwestern Argentina: Geological, geophysical, and remote sensing evidence: Ore Geology Reviews, v. 21, p. 127–155. Chiodini, G., Cardellini, C., Amato, A., Boschi, E., Caliro, S., Frondini, F., and Ventura, G., 2004, Carbon dioxide Earth degassing and seismogenesis in central and Southern Italy: Geophysical Research Letters, v. 31, L07615, DOI 10.1029/2004GL019480. Cline, J.S., Hofstra, A.H., Muntean, J.L., Tosdal, R.M., and Hickey, K.A., 2005, Carlin-type gold deposits in Nevada: Critical geologic characteristics and viable models: Economic Geology 100th Anniversary Volume, p. 451–484. Conolly, H.J.C., 1936, A contour method of revealing some ore structures: Economic Geology, v. 31, p. 259–271. Cox, S.F., 1999, Deformational controls on the dynamics of fluid flow in mesothermal gold systems: Geological Society, London, Special Publication 155, p. 123–140. ——2005, Coupling between deformation, fluid pressures, and fluid flow in ore-producing hydrothermal systems at depth in the crust: Economic Geology 100th Anniversary Volume, p. 39–75. Crafford, A.E.J., and Grauch, V.J.S., 2002, Geologic and geophysical evidence for the influence of deep crustal structures on Paleozoic tectonics and the alignment of world-class gold deposits, north-central Nevada, USA: Ore Geology Reviews, v. 21, p. 157–184. Cuney, M., 2010, Evolution of uranium fractionation processes through time: Driving the secular variation of uranium deposit types: Economic Geology, v. 105, p. 553–569. Czamanske, G.K., Zen’ko, T.E., Fedorenko, V.A., Calk, L.C., Budahn, J.R., Bullock, J.H., Fries, T.L., King, B.S., and Siems, D.F., 1995, Petrography and geochemical characterization of ore-bearing intrusions of the Noril’sk type, Siberia, with discussion of their origin: Resource Geology Special Issue 18, p. 1–48. Davies, G.F., 1995, Punctuated tectonic evolution of the Earth: Earth and Planetary Science Letters, v. 136, p. 363–379. Debaille, V., O’Neill, C., Brandon, A. D., Haenecour, P., Yin, Q. Z., Mattielli, N., and Treiman, A.H., 2013, Stagnant-lid tectonics in early Earth revealed by 142Nd variations in late Archean rocks: Earth and Planetary Science Letters, v. 373, p. 83–92. Drummond, B., Lyons, P., Goleby, B., and Jones, L., 2006, Constraining models of the tectonic setting of the giant Olympic Dam iron oxide-copper-gold deposit, South Australia, using deep seismic reflection data: Tectonophysics, v. 420, p. 91–103. Ehrig, K., McPhie, J., and Kamenetsky, V., 2012, Geology and mineralogical zoning of the Olympic Dam iron oxide Cu-U-Ag deposit, South Australia: Society of Economic Geologists Special Publication 16, p. 237–267. Emsbo, P., William Jones, A.E., Koenig, A.E., and Wilson, S.A., 2009, Petroleum as an agent of metal transport: Metallogenic and exploration implications: Smart Science for Exploration and Mining, Biennial SGA Meeting, 10th, Townsville, 2009, Proceedings, v. A1, p. 99–101. Evans, K.A., McCuaig, T.C., Leach, D., Angerer, T., and Hagemann, S.G., 2013, Banded iron formation to iron ore: A record of the evolution of Earth environments: Geology, v. 41, p. 99–102. Farquhar, J., Wu, N., Canfield, D.E., and Oduro, H., 2010, Connections between sulfur cycle evolution, sulfur isotopes, sediments, and base metal sulfide deposits: Economic Geology, v. 105, p. 509–533. Folinsbee, R.E., 1977, World’s view—from Alph to Zipf: Bulletin of the Geological Society of America, v. 88, p. 897–907. Frimmel, H.E., Groves, D.I., Kirk, J., Ruiz, J., Chesley, J., and Minter, W.E.L., 2005, The formation and preservation of the Witwatersrand goldfields, the world’s largest gold province: Economic Geology 100th Anniversary Volume, p. 769–788. Goldfarb, R.J., Snee, L.W., Miller, L.D., and Newberry, R.J., 1991, Rapid dewatering of the crust deduced from ages of mesothermal gold deposits: Nature, v. 354, p. 296–298. Goldfarb, R.J., Baker, T., Dubé, D., Groves, D.I., Hart, C.J.R., and Gosselin, P., 2005, distribution, character, and genesis of gold deposits in metamorphic terranes: Economic Geology 100th Anniversary Volume, p. 407–450. Goldfarb, R.J., Bradley, D., and Leach, D.L., 2010, Secular variation in economic geology: Economic Geology, v. 105, p. 459–465. Goldfarb, R.J., Taylor, R.D., Collins, G.S., Goryachev, N.A., and Orlandini, O.F., 2014, Phanerozoic continental growth and gold metallogeny of Asia: Gondwana Research, v. 25, p. 48–102. Gresham, J.J., and Loftus-Hills, G.D., 1981, The geology of the Kambalda nickel field, Western Australia: Economic Geology, v. 76, p. 1373–1416. Griffin, W.L., Begg, G.C., and O’Reilly, S.Y., 2013, Continental-root control on the genesis of magmatic ore deposits: Nature Geoscience, DOI 10.1038/ NGEO1954. Groves, D.I., 1993, The crustal continuum model for late-Archaean lode-gold deposits of the Yilgarn Block, Western Australia: Mineralium Deposita, v. 28, p 366–374. Groves, D.I., Goldfarb, R.J., Gebre-Mariam, M., Hagemann, S.G., and Robert, F., 1998, Orogenic gold deposits: A proposed classification in the context of their crustal distribution and relationship to other gold deposit types: Ore Geology Reviews, v. 13, p. 7–27. THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING Groves, D.I., Condie, K.C., Goldfard, R.J., Hronsky, J.M.A., and Vielreicher, R.M., 2005, Secular changes in global tectonic processes and their influence on the temporal distribution of gold-bearing mineral deposits: Economic Geology, v. 100, p. 203–224. Guj, P., Fallon, M., McCuaig, T.C., and Fagan, R., 2011, A time-series audit of Zipf’s law as a measure of terrane endowment and maturity in mineral exploration: Economic Geology, v. 106, p. 241–259. Halley, S.W., and Roberts, R.H., 1997, Henty: A shallow-water gold-rich volcanogeneic massive sulfide deposit in western Tasmania: Economic Geology, v. 100, p. 203–224. Harper, D.D., and Borrok, D.M., 2007, Dolomite fronts and associated zinclead mineralization, USA: Economic Geology, v. 102, p. 1345–1352. Haynes, D., 2006, The Olympic Dam ore deposit discovery—a personal view: Society of Economic Geologists Newsletter 66, p. 1, 8–15. Hazen, R.M., 2010, Evolution of minerals: Scientific American, v. 302, p. 58–65. Hazen, R.M., Papineau, D., Bleeker, W., Downs, R.T., Ferry, J.M., McCoy, T.J., and Yang, H., 2008, Mineral evolution: American Mineralogist, v. 93, p. 1693–1720. Hedenquist, J.W., John, F.H., Thompson, J.F.H., Goldfarb, R.J., and Richards, J.P., 2005, eds., Economic Geology 100th Anniversary Volume, 1,136 p. Hickey, K.A., Barker, S.L., Dipple, G.M., Arehart, G.B., and Donelick, R.A., 2014, The brevity of hydrothermal fluid flow revealed by thermal halos around giant gold deposits: Implications for Carlin-type gold systems: Economic Geology, v. 109, p. 1461–1487. Hodgson, C.J., 1993, Mesothermal lode-gold deposits: Geological Association of Canadian Special Paper 40, p. 635–678. Hronsky, J.M.A., 2011, Self-organized critical systems and ore formation: The key to spatial targeting?: Society of Economic Geology Newsletter 84, p. 14–16. Hronsky, J.M.A., and Groves, D.I., 2008, Science of targeting: Definition, strategies, targeting and performance measurement: Australian Journal of Earth Science, v. 55, p. 3–12. Hronsky, J.M.A., Groves, D.I., Loucks, R.R., and Begg, G.C., 2012, A unified model for gold mineralization in accretionary orogens and implications for regional scale targeting methods: Mineralium Deposita, v. 47, p. 339–358. Idnurm, M., 2000, Towards a high resolution Late Palaeoproterozoic earliest Mesoproterozoic apparent polar wander path for northern Australia: Australian Journal of Earth Science, v. 47, p. 405–429. Ingebritsen, S.E, and Appold, M.S., 2012, The physical hydrogeology of ore deposits: Economic Geology, v. 107, p. 559–584. International Atomic Energy Agency, 2000, Methods of exploitation of different types of uranium deposits: Vienna, International Atomic Energy Agency IAEA-TECDOC-1174, 75 p. Jelsma, H., Barnett, W., Richards, S., and Lister, G., 2009, Tectonic setting of kimberlites: Lithos, v. 112, p. 155–165. Jensen, H.J., 1998, Self-organized criticality: Emergent complex behavior in physical and biological systems: Cambridge Lecture Notes in Physics, Cambridge University Press, v. 10, 152 p. Kelley, K.D., and Ludington, S., 2002, Cripple Creek and other alkalinerelated gold deposits in the southern Rocky Mountains, USA: Influence of regional tectonics: Mineralium Deposita, v. 37, p. 38–60. Kerrich, R., 1983, Geochemistry of gold deposits in the Abitibi greenstone belt: Canadian Institute of Mining and Metallurgy Special Paper 27, 75 p. Kerrich, R., Goldfarb, R.J., and Richards, J.P., 2005, Metallogenic provinces in an evolving geodynamic framework: Economic Geology 100th Anniversary Volume, p. 1097–1136. Kesler, S.E., and Wilkinson, B.H., 2006, The role of exhumation in the temporal distribution of ore deposits: Economic Geology, v. 101, p. 919–922. Kish, L., and Cuney, M., 1981, Uraninite-albite veins from the Mistamisk Valley of the Labrador trough, Québec: Mineralogical Magazine, v. 44, p. 471–483. Knox-Robinson, C.M., and Wyborn, L.A.I., 1997, Towards a holistic exploration strategy: Using geographic information systems as a tool to enhance exploration: Australian Journal of Earth Sciences, v. 44, p. 453–463. Korzhinskii, D.S., 1968, The theory of metasomatic zoning: Mineralium deposita, v. 3, p. 222–231. Kouzmanov, K., and Pokrovski, G.S., 2012, Hydrothermal controls on metal distribution in porphyry Cu (-Mo-Au) systems: Society of Economic Geologists Special Publication 16, p. 573–618. Kreuzer, O.P., Etheridge, M.A., Guj, P., McMahon, M.E., and Holden, D., 2008, Linking mineral deposit models to quantitative risk analysis and decision-making in exploration: Economic Geology, v. 103, p. 829–850. 173 Kreuzer, O.P., Markwitz, V., Porwal, A.K., and McCuaig, T.C., 2010, A continent-wide study of Australia’s uranium potential: Part I: GIS-assisted manual prospectivity analysis: Ore Geology Reviews, v. 38, p. 334–366. Kump, L.R., 2008, The rise of atmospheric oxygen: Nature, v. 451, p. 277–278. Lallemand, S., Heuret, A., and Boutelier, D., 2005, On the relationships between slab dip, back-arc stress, upper plate absolute motion, and crustal nature in subduction zones: Geochemistry Geophysics Geosystems, v. 6, Q09006, DOI 10.1029/2005GC000917. Large, R.R., McPhie, J., Gemmell, J.B., Herrmann, W., and Davidson, G.J., 2001, The spectrum of ore deposit types, volcanic environments, alteration halos, and related exploration vectors in submarine volcanic successions: Some examples from Australia: Economic Geology, v. 96, p. 913–938. Law, D.M., and Phillips, G.N., 2005, Hydrothermal replacement model for Witwatersrand gold: Economic Geology 100th Anniversary Volume, p. 799–812. Leach, D.L., Sangster, D.F., Kelley, K.D., Large, R.L., Garven, G.G., Allen, C.,R., Gutzmer, J., and Walters, S., 2005, Sediment-hosted lead-zinc deposits: a global perspective: Economic Geology 100th Anniversary Volume, p. 561–608. Leach, D.L., Bradley, D.C., Huston, D., Pisarevsky, S.A., Taylor, R.D., and Gardoll, S. J., 2010, Sediment-hosted lead-zinc deposits in Earth history: Economic Geology, v. 105, p. 593–625. Lezaeta, P., 2001, Distortion analysis and 3-D modeling of magnetotelluric data in the southern Central Andes: Ph.D. dissertation, Fachbereich Geowissenschaften, Freie Universität Berlin, 196 p. Lord, D., Etheridge, M.A., Willson, M., Hall, G., and Uttley, P.J., 2001, Measuring exploration success: An alternative to the discovery-cost-per-ounce method of quantifying exploration success: Society of Economic Geologists Newsletter 45, p. 10–16. Loucks, R.R., 2014, Distinctive composition of copper-ore-forming arcmagmas: Australian Journal of Earth Sciences, p. 1–13, DOI 10.1080/08120099. 2013.865676. Loughlin, G.F., and Behre, C.H., Jr., 1933, Classification of ore deposits, in Finch, J.W., Bateman, A.M., Butler, B.S., Loughlin, G.F., McLaughlin, D.H., and Singewald, J.T., Jr., eds., Ore deposits of the western states (Lindgren volume): New York, American Institute Mining and Metallurgical Engineers, p. 17–55. Love, D.A., Clark, A.H., and Glover, J.K., 2004, The lithologic, stratigraphic, and structural setting of the giant Antamina copper-zinc skarn deposit, Ancash, Peru: Economic Geology, v. 99, p. 887–916. Lowell, J., and Guilbert. J.M., 1970, Lateral and vertical alteration mineralization zoning in porphyry ore deposits: Economic Geology, v. 65, p. 373–406. Lu, Y.J., Kerrich, R., Kemp, A.I., McCuaig, T.C., Hou, Z.Q., Hart, C.J., and Evans, N.J., 2013, Intracontinental Eocene-Oligocene porphyry Cu mineral systems of Yunnan, Western Yangtze craton, China: Compositional characteristics, sources, and implications for continental collision metallogeny: Economic Geology, v. 108, p. 1541–1576. Magoon, L.B., and Beaumont, E.A., 1991, Petroleum systems, in Beaumont, E.A., and Foster, N.H., eds., Exploring for oil and gas traps: Treatise of Petroleum Geology, Handbook of Petroleum Geology, p. 3-1–3-34. Manning, C.E., and Ingebritsen, S.E., 1999, Permeability of the continental crust: The implications of geothermal data and metamorphic systems: Reviews of Geophysics, v. 37, p. 127–150 Maydagán, L., Franchini, M., Chiaradia, M., Dilles, J., and Rey, R., 2014, The Altar porphyry Cu-(Au-Mo) deposit (Argentina): A complex magmatichydrothermal system with evidence of recharge processes: Economic Geology, v. 109, p. 621–641. McCuaig, T.C., and Hronsky, J.M.A., 2000, The current status and future of the interface between the exploration industry and economic geology research: Reviews in Economic Geology, v. 13, p. 553–559. McCuaig, T.C., and Kerrich, R., 1998, P-T-t-deformation-fluid characteristics of lode gold deposits: Evidence from alteration systematic: Ore Geology Reviews, v. 12, p. 381–453. McCuaig, T.C., Guibal, D., Roberts, M., and Cowan, J., 2003, Structural review of the Antamina mine, Peru: SRK report (unpub.) for Compania Minera Antamina, April, 2003, 37 p. McCuaig, T.C., Porwal, A., and Gessner, K., 2009, Fooling ourselves: Recognizing uncertainty and bias in exploration targeting: Centre for Exploration Targeting Newsletter, March 2009, p. 1, 5–8. McCuaig, T.C., Beresford, S., and Hronsky, J., 2010, Translating the mineral systems approach into an effective exploration targeting system: Ore Geology Reviews, v. 38, p. 128–138. 174 MCCUAIG AND HRONSKY McInnes, B.I., Evans, N.J., Fu, F.Q., and Garwin, S., 2005, Application of thermochronology to hydrothermal ore deposits: Reviews in Mineralogy and geochemistry, v. 58(1), p. 467–498. McKeith, T.D., Schodde, R.C., and Baltis, E.J., 2010, Gold discovery trends: Society of Economic Geologists Newsletter 81, p. 1, 20–26. McKinstry, H.E., 1941, Structural control of ore deposition in fissure veins: American Institute of Mining Engineers Transactions, v. 144, p. 65–95. ——1955, Structure of hydrothermal ore deposits: Economic Geology, v. 50, p. 170–225. Mercadier, J., Annesley, I.R.J., McKechnie, C.L., Bogdan T.S., and Creighton, S., 2013, Magmatic and metamorphic uraninite mineralization in the western margin of the Trans-Hudson orogen (Saskatchewan, Canada): A uranium source for unconformity-related uranium deposits?: Economic Geology, v. 108, p. 1037–1065. Miller, J., Blewett, R., Tunjic, J., and Connors, K., 2010, The role of early formed structures on the development of the world class St Ives goldfield, Yilgarn, WA: Precambrian Research, v. 183, p. 292–315. Miller, S.A., Colletini, C., Chiaraluce, L., Cocco, M., Barchi, M., and Klaus, B.J.P., 2004, Aftershocks driven by a high-pressure CO2 source at depth: Nature, v. 247, p. 724–727. Meyer, C., 1981, Ore-forming processes in geologic history: Economic Geology, v. 75, p. 6–41. ——1985, Ore metals through geologic history: Science, v. 227, p. 1421–1428. ——1988, Ore deposits as guides to geologic history of the Earth: Annual Reviews of Earth and Planetary Science, v. 16, p. 147–171. Mole, D.R., Fiorentini, M.L., Cassidy, K.F., Kirkland, C.L., Thebaud, N., McCuaig, T.C., and Miller, J., 2013, Crustal evolution, intra-cratonic architecture and the metallogeny of an Archaean craton: Geological Society, London, Special Publication 393, SP393-8; DOI 10.1144/SP393.8. Mole, D.R., Fiorentini, M.L., Thebaud, N., Cassidy, K.F., McCuaig, T.C., Kirkland, C.L., Romano, S.S., Doublier, M.P., Belousova, E.A., Barnes, S.J., and Miller, J., 2014, Archean komatiite volcanism controlled by the evolution of early continents: Proceedings of the National Academy of Sciences (in press). Muntean, J.L., Cline, J.S., Simon, A.C., and Longo, A.A., 2011, Magmatic hydrothermal origin of Nevada’s Carlin-type gold deposits: Nature Geoscience, v. 4, p. 122–127. Murphy, F.C., Cull, J., Lee, T.J., Lee, S.K., and Song, Y., 2008, Magnetotelluric soundings and crustal architecture at Century mine, northern Australia: Australian Journal of Earth Sciences, v. 55, p. 5, 655–667. Murphy, F.C., Hutton, L.J., Walshe, J.L., Cleverley, J.S., Kendrick, M.A., McLellan, J., Rubenach, J., Oliver, N.H.S., Gessner, K., Bierlein, F.P., Jupp, B., Ailleres, L., Laukamp, C., Roy, I.G., Miller, J.McL., Keys, D., and Nortje, G.S., 2011, Mineral system analysis of the Mt Isa-McArthur River region, northern Australia: Australian Journal of Earth Sciences, v. 58, p. 849–873. Naldrett, A.J., 2010, Secular variation of magmatic sulfide deposits and their source magmas: Economic Geology, v. 105, p. 669–688. O’Connor, J.M., Steinberger, B., Regelous, M., Koppers, A.A., Wijbrans, J.R., Haase, K.M., Stoffers, P., Jokat, W., and Garbe-Schönberg, D., 2013, Constraints on past plate and mantle motion from new ages for the HawaiianEmperor Seamount Chain: Geochemistry, Geophysics, Geosystems, v. 14, p. 4564–4584. O’Driscoll, E.S.T., 1986, Observations of the lineament-ore relation: London, Philisophical Transactions of the Royal Society, A317, p. 195–218. O’Neill, C., Debaille, V., and Griffin, W., 2013a, Deep earth recycling in the Hadean and constraints on surface tectonics: American Journal of Science, v. 313, p. 912–932. O’Neill, C., Lenardic, A., and Condie, K.C., 2013b, Earth’s punctuated tectonic evolution: Cause and effect: Geological Society, London, Special Publication, DOI 10.1144/SP389.4. Padilla-Garza, R.A., Titley, S.R., and Pimental, F.B., 2001, Geology of the Escondida porphyry copper deposit, Antofagasta region, Chile: Economic Geology, v. 96, p. 307–324. Papineau, D., 2010, Global biogeochemical changes at both ends of the Proterozoic: Insights from phosphorates: Astrobiology, v. 10, p. 165–181. Peacock, D.C.P., 2003, Scaling of transfer zones in the British Isles: Journal of Structural Geology, v. 25, p. 1561–1567. Pettke, T., Oberli, F., and Heinrich, C.A., 2010, The magma and metal source of giant porphyry-type ore deposits, based on lead isotope microanalysis of individual fluid inclusions: Earth and Planetary Science Letters, v. 296, p.267–277. Pirajno, F., 2004, Metallogeny in the Capricorn orogen, Western Australia, the result of multiple ore-forming processes: Precambrian Research, v. 128, p. 411–439. Presnell, R.D., 1992, Local and regional geology of the Oquirrh Mountains: Utah Geological Survey Miscellaneous Publications, v. 3, p. 293–306. Price, G.P., and Stoker, P., 2002, Australian Geodynamics Cooperative Research Centre’s integrated research program delivers a new minerals exploration strategy for industry: Australian Journal of Earth Sciences, v. 49, p. 595–600. Rampino, M.R., and Stothers, R.B., 1988, Flood basalt volcanism during the past 250 million years: Science, v. 241, p. 663–668. Reddy, S.M., and Evans, D.A.D., 2009, Palaeoproterozoic supercontinents and global evolution: Correlations from core to atmosphere: Geological Society, London, Special Publication 323, p. 1–26. Repetski, J.E., and Narkiewicz, M., 1996, Conodont color and surface textural alteration in the Muschelkalk (Triassic) of the Silesian-Cracow Zn-Pb district, Poland: Prace Państwowego Instytutu Geologicznego, v. 154, p. 113–121. Richard, A., Banks, D.A., Mercadier, J., Boiron, M.C., Cuney, M., and Cathelineau, M., 2011, An evaporated seawater origin for the ore-forming brines in unconformity-related uranium deposits (Athabasca Basin, Canada): Cl/ Br and δ37Cl analysis of fluid inclusions: Geochimica et Cosmochimica Acta, v. 75, p. 2792–2810. Ridley, J., 1993, The relations between mean rock stress and fluid flow in the crust: With reference to vein-and lode-style gold deposits: Ore Geology Reviews, v. 8, p. 23–37. Robert, F., Poulsen, K.H., Cassidy, K.F., and Hodgson, C.J., 2005, Gold metallogeny of the Superior and Yilgarn craton: Economic Geology 100th Anniversary Volume, p. 1001–1033. Rohrlach, B.D., 2002, Tectonic evolution, petrochemistry, geochronology and palaeohydrology of the Tampakan porphyry and high sulphidation epithermal Cu-Au deposit, Mindanao, Philippines: Unpublished Ph.D. thesis, Canberra, Australia, Australian National University, 499 p. Rohrlach, B.D., and Loucks, R.R., 2005, Multi-million year cyclic ramp-up of volatiles in a lower crustal magma chamber trapped below the Tampakan copper-gold deposit by Mio-Pliocene crustal compression in the southern Philippines, in Porter T.M., ed., Super porphyry copper and gold deposits: A global perspective: Adelaide, PGC Publishing, v. 2, p. 270–313. Rosenbaum, G., Giles, D., Saxon, M., Betts, P.G., Weinberg, R.F., and Duboz, C., 2005, Subduction of the Nazca Ridge and the Inca Plateau: Insights into the formation of ore deposits in Peru: Earth and Planetary Science Letters, v. 239, p. 18–32. Saunders, J.A., Unger, D.L., Kamenov, G.D., Fayek, M., Hames, W.E., and Utterback, W.C., 2008, Genesis of middle Miocene Yellowstone hotspotrelated bonanza epithermal Au-Ag deposits, northern Great Basin, USA: Mineralium Deposita, v. 43, p.715–734. Schneider, E.D., and Dorion, S., 2005, Into the cool: Energy flow, thermodynamics and life: Chicago, IL, University of Chicago Press, 362 p. Schodde, R.C., and Hronsky, J.M.A., 2006, The role of world-class mines in wealth creation: Society of Economic Geologists Special Publication 12, p. 71–90. Seedorff, E., Dilles, J.H., Proffett, J.M., Jr., Einauidi, M.T., Zurcher, L., Stavast, W.J.A., Johnson, D.A., and Barton, M.D., 2005, Porphyry deposits: Characteristics and origin of hypogene features: Economic Geology 100th Anniversary Volume, p. 251–298. Sherlock, R.L., Roth, T., Spooner, E.T.C., and Bray, C.J., 1999, Origin of the Eskay Creek precious metal-rich volcanogenic massive sulfide deposit: Fluid inclusion and stable isotope evidence: Economic Geology, v. 94, p. 803–824. Sibson, R.H., Robert, F., and Poulsen, K.H., 1988, High-angle reverse faults, fluid-pressure cycling and mesothermal gold deposits: Geology v. 14, p. 551–555. Sillitoe, R.H., 2000, Gold-rich porphyry deposits: Descriptive and genetic models and their role in exploration and discovery: Society of Economic Geologists Reviews in Economic Geology, v. 13, p. 315–345. ——2004, Musings on future exploration targets and strategies in the Andes: Society of Economic Geologists Special Publication 11, p. 1–14. ——2008, Major gold deposits and belts of the North and South American Cordillera: Distribution, tectonomagmatic settings, and metallogenic considerations: Economic Geology, v. 103, p. 663–688. ——2010, Porphyry copper systems: Economic Geology, v. 105, p. 3–41. ——2012, Copper provinces: Society of Economic Geologists Special Publication 16, p. 1–18. THE MINERAL SYSTEM CONCEPT: THE KEY TO EXPLORATION TARGETING Sillitoe, R.H., and Mortensen, J.K., 2010, Longevity of porphyry copper formation at Quellaveco, Peru: Economic Geology, v. 105, p. 1157–1162. Sillitoe, R.H., and Perelló, J., 2005, Andean copper province: Tectonomagmatic settings, deposits types, metallogeny, exploration and discovery: Economic Geology 100th Anniversary Volume, p. 845–890. Sillitoe, R.H., and Thompson, J.F.H., 2006, Changes in mineral exploration practice: Consequences for discovery: Society of Economic Geologists Special Publication 12, p. 193–219. Simmons, S.F., and Brown, K.L., 2008, Precious metals in modern hydrothermal solutions and implications for the formation of epithermal ore deposits: Society of Economic Geologists Newsletter 72, p. 1, 9–12. Simmons, S.F., White, N.C., and John, D.A., 2005, Geological characteristics of epithermal precious and base metal deposits: Economic Geology 100th Anniversary Volume, p. 485–522. Skelton, A., 2011, Flux rates for water and carbon during greenschist facies metamorphism: Geology, v. 39, p. 43–46. Snyder, D.B., 2013, Imaging Archaean-age whole mineral systems: Precambrian Research, v. 229, p. 125–132. Squire, R.J., and Miller, J.M.M., 2003, Synchronous compression and extension in East Gondwana: Tectonic controls on world-class gold deposits at 440 Ma: Geology, v. 31, p. 1073–1076. Swager, C.P., 1985, Syndeformational carbonate-replacement model for the copper mineralization at Mount Isa, Northwest Queensland: A microstructural study: Economic Geology, v. 80, p. 107–125. Syracuse, E.M., and Abers, G.A., 2006, Global compilation of variations in slab depth beneath arc volcanoes and implications: Geochemistry, Geophysics, Geosystems, v. 7, DOI 10.1029/2005GC001045. Taylor, C.D., Premo, W.R., Meier, A.L., and Taggart, J.E., 2008, The metallogeny of Late Triassic rifting of the Alexander terrane in southeastern Alaska and northwestern British Columbia: Economic Geology, v. 103, p. 89–115. 175 Titley, S.R., 2001, Crustal affinities of metallogenesis in the American Southwest: Economic Geology, v. 96, p. 1323–1342. Tomkins, A.G., 2013, A biogeochemical influence on the secular distribution of orogenic gold: Economic Geology, v. 108, p. 193–197. Tosdal, R.M., Dilles, J.H., and Cooke, D.R., 2009, From source to sinks in auriferous magmatic-hydrothermal porphyry and epithermal deposits: Elements, v. 5, p. 289–295. Turneaure, F.S., 1955, Metallogenetic provinces and epochs: Economic Geology, v. 50, p. 38. Witt, W.K., Hagemann, S.G., Villanes, C., and Zeng, Q., 2013, New geochronological results and structural evolution of the Pataz gold mining district: Implications for the timing and origin of the batholith-hosted veins: Ore Geology Reviews, v. 50, p. 143–170. Woodall, R., 1984, Success in mineral exploration: Confidence in science and ore deposit models: Geoscience Canada, v. 11, p. 127–133. ——1994, Empiricism and concept in successful mineral exploration: Australian Journal of Earth Sciences, v. 41, p. 1–10. Wyborn, L.A.I., Heinrich C.A., and Jaques A.L., 1994, Australian Proterozoic mineral systems: Essential ingredients and mappable criteria: Australian Institute of Mining and Metallurgy Annual Conference, Melbourne, Proceedings, p. 109–115. Yamada, R., and Yoshida, T., 2011, Relationships between Kuroko volcanogenic massive sulfide (VMS) deposits, felsic volcanism, and island arc development in the northeast Honshu arc, Japan: Mineralium Deposita, v. 46, p. 431–448. Yardley, B.W.D., 2005, Metal concentrations in crustal fluids and their relationship to ore formation: Economic Geology, v. 100, p. 613–632. Zhang, M., O’Reilly, S.Y., Wang, K-L., Hronsky, J., and Griffin, W.L., 2008, Flood basalts and metallogeny: The lithospheric mantle connection: Earth Science Reviews, v. 86, p. 145–174.
