Ultramafic Units Research Articles

Mineralogical, petrological, and geochemical studies of the Limahe mafic–ultramatic intrusion and associated Ni–Cu sulfide ores, SW China

The Limahe Ni–Cu sulfide deposit is hosted by a small mafic–ultramafic intrusion (800 × 200 × 300 m) that is temporally associated with the voluminous Permian flood basalts in SW China. The objective of this study is to better understand the origin of the deposit in the context of regional magmatism which is important for the ongoing mineral exploration in the region. The Limahe intrusion is a multiphase intrusion with an ultramafic unit at the base and a mafic unit at the top. The two rock units have intrusive contacts and exhibit similar mantle-normalized trace element patterns and Sr–Nd isotopic compositions but significantly different cumulus mineralogy and major element compositions. The similarities suggest that they are related to a common parental liquid, whereas the differences point to magma differentiation by olivine crystallization at depth. Sulfide mineralization is restricted to the ultramafic unit. The abundances of sulfides in the ultramafic unit generally increase towards the basal contacts with sedimentary footwall. The δ34S values of sulfide minerals from the Limahe deposit are elevated, ranging from +2.4 to +5.4‰. These values suggest the involvement of external S with elevated δ34S values. The mantle-normalized platinum-group element (PGE) patterns of bulk sulfide ores are similar to those of picrites associated with flood basalts in the region. The abundances of PGE in the sulfide ores, however, are significantly lower than that of sulfide liquid expected to segregate from undepleted picrite magma. Cr-spinel and olivine are present in the Limahe ultramafic rocks as well as in the picrites. Mantle-normalized trace element patterns of the Limahe intrusion generally resemble those of the picrites. However, negative Nb–Ta anomalies, common features of contamination with the lower or middle crust, are present in the intrusion but absent in the picrites. Sr–Nd isotopes suggest that the Limahe intrusion experienced higher degrees of contamination with the upper crust than did the picrites. The results of this study permit us to suggest that the parental magma of the Limahe intrusion was derived from picritic magma by olivine fractionation and contamination in a staging chamber at mid-crustal levels. Depletion of PGE in the sulfide ores in the Limahe intrusion is likely due to previous sulfide segregation of the parental magmas in the staging chamber. Sulfide mineralization in the Limahe intrusion is related to second-stage sulfide segregation after the fractionated magmas acquired external S from pyrite-bearing country rocks during magma ascent to the Limahe chamber. The abrupt change in mineralogical and chemical compositions between the ultramafic unit and the overlying unit suggests that at least two separate pulses of magma were involved in the development of the Limahe intrusion. We propose that the Limahe intrusion was once a wider part of a dynamic conduit that fed magma to the overlying subvolcanic dykes/sills or lavas. The ultramafic unit formed by the first, relatively more primitive magma, and the mafic unit formed by the second, relatively more fractionated magma. Immiscible sulfide droplets that segregated from the first magma settled down with olivine crystals to form the sulfide-bearing, olivine-rich rocks in the base of the intrusion. The overlying residual liquids were then pushed out of the chamber by the second magma. Critical factors for the formation of an economic Ni–Cu sulfide deposit in such a small intrusion include the dynamic petrologic processes involved and the availability of external sulfur. The Limahe deposit reminds us that small, multiphase, mafic–ultramafic intrusions in the region should not be overlooked for the potential of economic Ni–Cu sulfide deposits.

Occurrence, composition and growth of polyhedral serpentine

Occurrences, compositions and crystallization of polyhedral serpentine were investigated by SEM, AEM/TEM and μ-XANES analysis of samples from different ultramafic units. Polyhedral serpentines are identified in all of these contexts and form as an alteration product of orthopyroxene (enstatite) and as late veining events. They are always the last serpentine type to crystallize. Their formation requires a combination of three factors: 1) open space, 2) a relatively low temperature ( T < 200–300 °C), and 3) the presence of trivalent cations (Al3+ in this study, ⩾ 0.1 atoms per serpentine formula unit). μ-XANES data at the iron K -edge indicate that Fe is predominantly ferrous and octahedrally coordinated in our Al-rich samples. This microstructure therefore cannot be systematically used as a marker of oxidizing conditions. Textural and microstructural criteria suggest that polyhedral serpentine crystallizes via a “gel” precursor first reorganized into a poorly-crystallized proto-serpentine, in which onion nucleation takes place as nested, discontinuous sheets. Grains expand radially, inwards and outwards, by a layer-by-layer mechanism. Thick layers, made of tens of serpentine sheets, propagate laterally in the (001) plane and result in a pseudo-spherical “onion-like” morphology. By analogy with available clay synthesis experiments, the relatively low temperature conditions under which polyhedral serpentine form may favor a segregation of trivalent cations in the structure. This could create locally dioctahedral components in the structure that may explain the peculiar bending along 〈010〉 responsible for the faceted morphology of polyhedral serpentine.

UPPER TRIASSIC TO CRETACEOUS RADIOLARIA FROM NICARAGUA AND NORTHERN COSTA RICA - THE MESQUITO COMPOSITE OCEANIC TERRANE

We propose a new terrane subdivision of Nicaragua and Northern Costa Rica, based on Upper Triassic to Upper Cretaceous radiolarian biochronology of ribbon radiolarites, the newly studied Siuna Serpentinite Melange, and published 40Ar/39Ar dating and geochemistry of mafic and ultramafic igneous rock units of the area. The new Mesquito Composite Oceanic Terrane (MCOT) comprises the southern half of the Chortis Block, that was assumed to be a continental fragment of N-America. The MCOT is defined by 4 corner localities characterized by ultramafic and mafic oceanic rocks and radiolarites of Late Triassic, Jurassic and Early Cretaceous age: 1. The Siuna Serpentinite Melange (NE-Nicaragua), 2. The El Castillo Melange (Nicaragua/Costa Rica border), 3.The Santa Elena Ultramafics (N-Costa Rica) and, 4. DSDP Legs 67/84. 1. The Siuna Serpentinite Melange contains, high pressure metamorphic mafics and Middle Jurassic (Bajocian-Bathonian) radiolarites in original, sedimentary contact with arc-metandesites. The Siuna Melange also contains Upper Jurassic black detrital chert formed in a marginal (fore-arc?) basin shortly before subduction. A phengite 40Ar/39Ar -cooling age dates the exhumation of the high pressure rocks as 139 Ma (earliest Cretaceous). 2. The El Castillo Melange comprises a radiolarite block tectonically embedded in serpentinite that yielded a diverse Rhaetian (latest Triassic) radiolarian assemblage, the oldest fossils recovered so far from S-Central America. 3. The Santa Elena Ultramafics of N-Costa Rica together with the serpentinite outcrops near El Castillo (2) in Southern Nicaragua, are the southernmost outcrops of the MCOT. The Santa Elena Unit (3) itself is still undated, but it is thrust onto the middle Cretaceous Santa Rosa Accretionary Complex (SRAC), that contains Lower to Upper Jurassic, highly deformed radiolarite blocks, probably reworked from the MCOT, which was the upper plate with respect to the SRAC. 4. Serpentinites, metagabbros and basalts have long been known from DSDP Leg 67/84 (3), drilled off Guatemala in the Nicaragua-Guatemala forearc basement. They have been restudied and reveal 40Ar/39Ar dated Upper Triassic to middle Cretaceous enriched Ocean Island Basalts and Jurassic to Lower Cretaceous depleted Island arc rocks of probable Pacific origin. The area between localities 1-4 is largely covered by Tertiary to Recent arcs, but we suspect that its basement is made of oceanic/accreted terranes. Earthquake seismic studies indicate an ill-defined, shallow Moho in this area. The MCOT covers most of Nicaragua and could extend to Guatemala to the W and form the Lower (southern) Nicaragua Rise to the NE. Some basement complexes of Jamaica, Hispaniola and Puerto Rico may also belong to the MCOT. The Nicoya Complex s. str. has been regarded as an example of Caribbean crust and the Caribbean Large Igneous Province (CLIP). However, 40Ar/39Ar - dates on basalts and intrusives indicate ages as old as Early Cretaceous. Highly deformed Jurassic and Lower Cretaceous radiolarites occur as blocks within younger intrusives and basalts. Our interpretation is that radiolarites became first accreted to the MCOT, then became reworked into the Nicoya Plateau in Late Cretaceous times. This implies that the Nicoya Plateau formed along the Pacific edge of the MCOT, independent form the CLIP and most probably unrelated with he Galapagos hotspot. No Jurassic radiolarite, no older sediment age than Coniacian-Santonian, and no older 40Ar/39Ar age than 95 Ma is known from S-Central America between SE of Nicoya and Colombia. For us this area represents the trailing edge of the CLIP s. str.

Petrogenesis and tectonic significance of Jurassic IAT magma types in the Hellenide ophiolites as deduced from the Rhodiani ophiolites (Pelagonian zone, Greece)

The Rhodiani ophiolites are represented by two tectonically superimposed ophiolitic units: the “lower” Ultramafic unit and the “upper” Volcanic unit, both bearing calcareous sedimentary covers. The Ultramafic unit consists of mantle harzburgites with dunite pods and chromitite ores, and represents the typical mantle section of supra-subduction zone (SSZ) settings. The Volcanic unit is represented by a sheeted dyke complex overlain by a pillow and massive lava sequence, both including basalts, basaltic andesites, andesites, and dacites. Chemically, the Volcanic unit displays low-Ti affinity typical of island arc tholeiite (IAT) ophiolitic series from SSZ settings, having, as most distinctive chemical features, low Ti/V ratios (< 20) and depletion in high field strength elements and light rare earth elements. The rare earth element and incompatible element composition of the more primitive basaltic andesites from the Rhodiani ophiolites can be successfully reproduced with about 15% non-modal fractional melting of depleted lherzolites, which are very common in the Hellenide ophiolites. The calculated residua correspond to the depleted harzburgites found in the Rhodiani and Othrys ophiolites. Both field and chemical evidence suggest that the whole sequence of the Rhodiani Volcanic unit (from basalt to dacite) originated by low-pressure fractional crystallization under partially open-system conditions. The modelling of mantle source, melt generation, and mantle residua carried out in this paper provides new constraints for the tectono-magmatic evolution of the Mirdita–Pindos oceanic basin.

The association of mafic–ultramafic intrusions and A-type magmatism in the Tian Shan and Altay orogens, NW China: Implications for geodynamic evolution and potential for the discovery of new ore deposits

The NW China region is characterised by tectonic and lithostratigraphic domains, such as the Tian Shan and Altay orogens, the Tarim, Junggar and Turpan Basins. The Tian Shan and Altay orogens are part of the Central Asian Orogenic Belt. The NW China region was affected by a series of thermal events that occurred between the Silurian and the Triassic, which resulted in the emplacement of numerous granitic plutons and mafic–ultramafic intrusions. A number of these granitic plutons are of A-type affiliation, which on the basis of the positive εNd values are likely to have been derived from mantle sources. In addition, at least two large igneous provinces (LIPs) can be recognised in NW China, namely the 345–325 Ma Tian Shan LIP and the ca. 270–280 Ma Tarim LIP. Age and field data suggest a spatial and temporal relationship between the mafic–ultramafic intrusions and A-type granites within the LIPs. In this paper we discuss mafic–ultramafic intrusions that host magmatic Ni–Cu sulphide deposits (Kalatongke in the Altay, Huangshan and Poyi–Poshi) in the eastern Tian Shan. These intrusions are typically zoned, characterised by an envelope of early gabbroic rocks that enclose later ultramafic units. These zoned mafic–ultramafic intrusions have some features that are comparable with Alaskan-type complexes. Taking into consideration the spatial–temporal relationship of the mafic, mafic–ultramafic rocks and A-type granites, we suggest that these magmatic events occurred during an extensional regime, possibly related to a mantle superplume event that affected much of central Asia during the Permian, of which the Siberian Traps and the Emeishan continental flood basalts of SW China are part. If the A-type felsic magmatism took place during a superplume event, we also suggest that these rocks may be conducive to host iron–oxide–copper–gold (IOCG) style mineralisation. We conclude with a model that attempts to explain the relationship between the zoned mafic–ultramafic intrusions and mantle plume activity in NW China during the Permian.

Controls on the emplacement and genesis of the MKD5 and Sarah’s Find Ni–Cu–PGE deposits, Mount Keith, Agnew–Wiluna Greenstone Belt, Western Australia

The Mount Keith (MKD5) nickel sulfide deposit is one of the largest komatiite-hosted nickel sulfide deposits in the world; it is hosted by a distinctive spinifex-free, cumulate-rich, ultramafic horizon/unit termed the Mount Keith Ultramafic (MKU). The Mount Keith Ultramafic shows significant variation along its lateral extent. The internal architecture is made up of adcumulate-textured pods and lenses, which are flanked by thinner meso- and orthocumulate-textured units, overlain by pyroxenitic and gabbroic horizons. The lateral and vertical changes in the geometry and internal architecture reflect variations in the lithological association and emplacement conditions along the strike extent of the belt. The chilled margins of the Mount Keith Ultramafic unit contain ∼1,200 ppm Ni. Olivine cumulates average ∼2,500–3,500 ppm Ni, with few exceptions (Ni > 4,500 ppm) reflecting occurrence of minor nickel sulfides, whereas pyroxenites and gabbros generally contain, respectively, ∼1,500–2,000 and ∼100–1,000 ppm Ni. Olivine cumulates generally contain low Cr concentrations ( 5,000 ppm Cr). The internal stratigraphy of the Mount Keith Ultramafic unit may be subdivided into two groups based on rare earth element distribution. The chilled margins and the internal units of the Main Adcumulate domain display LREE-enriched patterns [(La/Sm)n > 1–3] and negative Eu, Hf, Zr, Nb, and Ti anomalies. The internal units in the Western Mineralized Zone generally display flat chondrite-normalized REE patterns and only minor negative Nb anomalies. The pattern of platinum-group element (PGE) distribution varies greatly along the strike extent of the Mount Keith Ultramafic unit. The chilled margins display relatively low absolute concentrations [PGE (excl. Os) ∼16 ppb] and relatively fractionated patterns, with subchondritic Pt/Pd ratios (∼1.5), and superchondritic Pd/Ir ratios (∼3). The PGE trends in the thick adcumulate-textured pods containing widespread nickel sulfide mineralization display positive correlation with sulfide abundance, whereas fractionated pyroxenites and gabbros in the thinner domains display highly depleted PGE concentrations and generally show compatible PGE trends. The nickel sulfide ore typology and style vary greatly along the strike extension of the Mount Keith Ultramafic unit. Basal massive nickel sulfide mineralization (e.g., Sarah’s Find) occurs in the thinner meso- and orthocumulate-textured units, whereas stratabound disseminated nickel sulfide mineralization (e.g., MKD5 Ni Deposit) is hosted in the adcumulate-textured pods. We hypothesize that the very low PGE content of the initial liquid of the Mount Keith Ultramafic unit indicates that the initial magma pulse that penetrated through the dacite host-rock had already equilibrated with sulfides at depth and/or carried entrained immiscible sulfide blebs. We argue that upon emplacement, the intruding magma experienced a significant thermal shock at the contact with water-saturated volcaniclastic breccias. The sudden chilling would have increased the viscosity of the magma, possibly to the point where it was no longer able to sustain the suspension of the immiscible sulfide liquid. As a result, the sulfide blebs coalesced and formed the basal massive sulfide nickel sulfide mineralization at the base of the sill (i.e., Sarah’s Find). Prolonged focused high volume magma flow within the sill resulted in the emplacement of a thick, lens-shaped accumulation of olivine adcumulate. Local variations in intensive parameters other than crustal assimilation (e.g., T, fO2, fS2) may be principally responsible for sulfide supersaturation and controlled the local distribution of stratabound disseminated nickel sulfide mineralization (e.g., MKD5 Ni Deposit), generally localized within the core of the thicker dunite lenses.

Quantitative bedrock geology of Brazil

We quantitatively analyze the area‐age distribution of sedimentary, igneous, metamorphic, and ultramafic bedrock on the basis of data from the digital geologic map of Brazil, published as a GIS map by the Brazilian Geological Survey. Bedrock units exclusively encompassing sedimentary rocks, igneous rocks, or metamorphic rocks cover 40.4%, 31.5%, and 17.7%, respectively, of the total bedrock area. These numbers have to be considered minimum estimates of the areal abundance of sedimentary, igneous, and metamorphic bedrock because polygons defined by mixed lithologies cover ∼8.5–9.5% of the total bedrock area. These mixed units are sedimentary rocks with igneous and/or metamorphic contributions (1.4%), metamorphic rocks with sedimentary contributions (1.2%), metamorphic rocks with igneous contributions (1.5%), igneous rocks with sedimentary and/or metamorphic contributions (4.4%), and ultramafic units with sedimentary, igneous, and/or metamorphic contributions (∼1–2%). The average ages of major lithologic units, weighted according to bedrock area, are as follows: sedimentary rocks (average stratigraphic age of 248 ± 5 [1σ] Myr; median stratigraphic age of 87.5 Myr), igneous rocks (1153 ± 13 [1σ] Myr), metamorphic rocks (1678 ± 30 [1σ] Myr), and ultramafic rocks (∼1227 ± 25 [1σ] Myr). The average bedrock age of Brazil is 946 ± 7 [1σ] Myr. The range in lithologic composition and age structure of the various bedrock units reflects the complex tectonic makeup of Brazil that ranges from Neogene sedimentary cover in the Amazon Basin to Precambrian cratons (Guyana and Brazilian shields) and Transamazonian greenstone belts. The average spatial resolution of the data is 232 km2 polygon−1 and is sufficient to perform area‐age analyses of individual river drainage basins larger than ∼5,000 km2.

Anomalous Reef-Type Platinum-Group Element Mineralisation in the Wiluna Domain, Agnew-Wiluna Greenstone Belt, Western Australia

The ultramafic horizons of the Wiluna domain (Agnew-Wiluna Greenstone Belt, Western Australia) form two main tabular and sheet-like fractionated units, which generally comprise a 5-10m-thick basal pyroxenite layer, a 50-100m-thick meso-to-orthocumulate-textured peridotite horizon, which is overlain by a core of wehrlite, ranging 10-50m in thickness, grading upwards into oikocrystic pyroxenites (30-50m-thick) and upper gabbroic margins (30-50m-thick). Primary textures are exceptionally well preserved: pegmatoidal and dendritic (harrisitic) olivine domains are common in all lithologies, whereas spinifex-textured horizons and flowtop breccias are absent. The fractionation sequence and the geochemical profile of the two ultramafic units are almost indistinguishable, suggesting that the belts may have been structurally duplicated. The ultramafic units display a steady trend towards increasing Fe content up stratigraphy, above a narrow basal zone showing a reversed trend. Calculated Ni contents of cumulus olivines indicate that the magma remained sulfur-undersaturated up to about the middle of the unit, where fractionation of a small proportion of sulfide (R>1000) occurred and depleted the upper fractionated levels in Ni. The presence of various anomalous PGE-enriched horizons (SPt-Pd= 200-300 ppb) in the fractionated pyroxenite layer directly overlying the basal ortho-mesocumulate-textured unit indicates the occurrence of localised weak reef-style PGE mineralisation. Reef-style PGE mineralisation at Wiluna is more similar in stratigraphic setting, style and composition to PGE-rich disseminated Fe-Cu sulfide mineralisation zones within thick differentiated intrusions (e.g. Platinova reef in the Skaergaard intrusion) rather than to other zones in komatiite-hosted systems.

Mapping non-magnetic, talc-bearing ultramafic rocks in the Kalgoorlie region using airborne hyperspectral data

Previously unrecognised ultramafic units in the Kalgoorlie-Kanowna area have recently been discovered using airborne hyperspectral sensing. Some of these units are up to 1 km in length and 200 m in width and were identified through airborne mineral mapping of talc. These areas lack significant magnetic expression, have little or no outcrop or subcrop and comprise a surface dominated by talc-bearing soil pediment. Exploration drill spoil in these areas confirms the presence of weathered talc schists underlying the soils. These results suggest that even in well mapped areas of deeply weathered greenstone belts, airborne hyperspectral sensing can complement traditional geophysical data by mapping non-magnetic, talc-bearing ultramafic rocks, which may be potential hosts to massive nickel and base metal mineralisation.

Geology, Age and Origin of Supracrustal Rocks at Akilia, West Greenland

New mapping, geochronology, and geochemistry of the supracrustal enclave and associated orthogneisses at Akilia, West Greenland, support a ≥3825 ± 6 Ma age for the sequence and a chemical sedimentary origin for the controversial Fe-rich quartz-pyroxene gneisses. Lithologies of the enclave comprise laterally continuous, mappable units of mafic amphibolite, ultramafic rocks, and two Fe-rich quartz-pyroxene units. A minor anthophyllite-garnet rock with chemical characteristics suggesting a sedimentary protolith was identified. The earliest foliation parallels lithologic contacts. This foliation was isoclinally folded about a steep hinge surface, and then tightly refolded about a steep NS hinge plane. The latest folding produced limbs exhibiting contrasting strains. The full deformation history is shared by all lithologies of the enclave. Two previously unidentified orthogneiss sheets preserve deformed magmatic crosscutting relations with amphibolite and ultramafic units in the low-strain limb. The supracrustal anthophyllite-garnet rock experienced zircon growth at ∼3600 and ∼2700 Ma, in common with amphibolites. These ages are interpreted to reflect metamorphism at granulite and amphibolite facies, respectively, consistent with independent regional evidence. The crosscutting orthogneiss sheets preserve a complex history of zircon growth, but only earliest cores indicate Th/U exchange equilibrium with the bulk rock at magmatic conditions. Weighted-mean ²⁰⁷Pb/²⁰⁶Pb ages of cores of 3730 ± 7 Ma and 3825 ± 6 Ma date the crystallization ages of the sheets, indicating that the Akilia supracrustal rocks were deposited ≥3819 Ma. Quartz-pyroxene rocks on Akilia have been interpreted as metamorphosed chemical sediments that may preserve carbon-isotope evidence for life at the time of their formation. An alternative proposal is an origin by metasomatic alteration. Our results show that these rocks experienced a history of deformation and zircon growth identical to other lithologies of the supracrustal body. Metamorphic zircons give a minimum ²⁰⁷Pb/²⁰⁶Pb age of 3589 ± 13 Ma regardless of origin. The two models of formation of quartz-pyroxene rocks were evaluated using field, petrologic and geochemical tests. The distribution of the units, the absence of field or petrologic indicators of metasomatism, and geochemistry (oxygen and nontraditional stable isotopes, immobile elements and REE) together support a sedimentary, but not a metasomatic, origin. Because their protolith was sedimentary, the quartz-pyroxene rocks are a primary part of the supracrustal sequence and have the same minimum age. Despite strong deformation and polymetamorphism, the Akilia supracrustal enclave contains information about terrestrial surficial processes at ≥3819 Ma.

A Review of Scientific Literature Examining the Mining History, Geology, Mineralogy, and Amphibole Asbestos Health Effects of the Rainy Creek Igneous Complex, Libby, Montana, USA

This article reviews the past 90 yr of scientific research directed on multiple aspects of the unique geology and environmental health issues surrounding the vermiculite deposit found at Libby, MT. Hydrothermal alteration and extensive weathering of the ultramafic units resulted in the formation of a rich deposit of vermiculite that was mined for 67 yr and used in numerous consumer products in its expanded form. Later intrusions of alkaline units caused hydrothermal alteration of the pyroxenes, resulting in formation of amphiboles. Some of these amphiboles occur in the asbestiform habit and have been associated with pulmonary disease in former miners and mill workers. Identification of these amphibole asbestos minerals has received little attention in the past, but recent work shows that the majority of the amphibole mineral species present may not be any of the amphibole species currently regulated by government agencies. Epidemiological studies on former miners have, nevertheless, shown that the amphibole asbestos from the Rainy Creek igneous complex is harmful; also, a recent study by the Agency for Toxic Substances and Disease Registry shows that residents of Libby who had not been employed in the vermiculite mining or milling operations also appear to have developed asbestos-related pulmonary diseases at a higher rate than the general public elsewhere. Since November 1999, the U.S. Environmental Protection Agency has been involved in the cleanup of asbestos-contaminated sites in and around Libby associated with the mining and processing of vermiculite.

Primary Platinum Mineralization in the Nizhny Tagil and Kachkanar UltramaficComplexes, Urals, Russia: A Genetic Model for PGE Concentration in Chromite-RichZones

The Nizhny Tagil Uralian-(Alaskan)-type Ultramafic Complex is composed of a dunite core with a rim of wehrlite and pyroxenite. The dunite unit, which has a uniform mineral composition (olivine Fo 89.8–92.1 , avg 0.25 wt % CaO), represents an accumulation of olivine and minor chromite from a mafic magma in an open-system magma chamber. After solidification of the dunite, differentiation began with the formation of the wehrlite (olivine Fo 77.9–88.8 ) and pyroxenite, corresponding to the closure of the magmatic system. Two main types of chromitite occurrence are observed in the dunite body: small (100 ∞ 5 cm) scattered schlieren of chromite crystals within the ultramafic unit and much larger (up to 100 ∞ 5 m) concentrations. The chromite composition in both types is similar and uniform [Cr/(Cr + Al + Fe 3+ ) = 66.02, Mg/(Mg + Fe 2+ ) = 49.05 and Cr/(Cr + Al + Fe 3+ ) = 69.14, Mg/(Mg + Fe 2+ ) = 49.81, respectively]. The larger concentrations of chromite include brecciated and net-textured ore or massive veins that reflect the dynamic accumulation of chromite crystals in cavities along magma conduits inside the dunite body. These cavities are analogous to small magma chambers within the consolidated part of the dunite body and, similar to podiform chromitites in ophiolites, were created by continuous feeding of magma to the system. The platinum-group element (PGE) content of the chromite-free facies of the dunite is relatively low, Pt being the only significant PGE (typically between 0.3–35 ppb but as much as 98 ppb in one sample). In contrast, both types of chromite occurrence are enriched in PGE. Two PGE mineralizing episodes are distinguished: an Ir-Ru-(Rh ± Os)-rich, Pt-Pd-poor episode in which the platinum-group minerals (PGM) are mainly Ir alloys, and a Pt-(Pd)-rich episode in which the PGM are mainly Pt alloys (i.e., Pt 2 (Fe,Ni,Cu), PtFeNi 2 , and tetraferroplatinum- and tulameenite-like compositions, along with rare hollingworthite and an unidentified RhNiAs mineral). The primary PGM have been affected locally by a serpentinization-related low-temperature alteration that is characterized by addition of copper, with alteration to complex Pt-(Fe,Ni,Cu) alloys of varied stoichiometry, including tulameenite and platinian copper. Placer concentrations derived from the Nizhny Tagil Complex have a slightly different assemblage of PGM, with abundant isoferroplatinum, the presence of gold commonly aggregating the PGM, and unusual Au-Sn and Au-Pb alloys. The Kachkanar Complex, situated 120 km north of the Nizhny Tagil Complex, has essentially the same characteristics, apart from a slightly different composition of the ultramafic cumulates suggesting that they were derived from a magma of different composition. The PGM distribution also differs in that it contains more abundant isoferroplatinum than other PGM. The high Pt concentrations are the result of the affinity of platinum minerals for chromite and the mode of chromite deposit formation. In the cavities in which chromite crystallized, PGM tended to be incorporated in chromite or remained attached to chromite crystals. To explain the very high concentrations of chromite-associated PGM we define a CR factor, which is the ratio of the mass of silicate magma to the mass of chromite in contact with the ascending magma. Where the CR is high (i.e., a relatively low proportion of chromite in contact with a large amount of magma crystallizing the Pt minerals), the efficiency of the mechanical collection of Pt minerals is at a maximum, and the Pt concentration (i.e., amount of Pt minerals) will be high. Possible variations due to the collection of PGE by a sulfide liquid have not been established. A model is proposed whereby the observed chromite and associated Pt mineralization can result from the normal magmatic evolution of the complex, provided that a structural (hydrodynamic) trap is available to concentrate the chromite.

Interpretation of a Sub-Audio Magnetic Survey over the lena Shear near Cue, Western Australia

The Lena Shear is a gold-anomalous structure that sits in an Archaean greenstone belt and extends for more than 7 km. It is partially obscured by regolith and an extensive salt lake system over its length. Identification of the shear under cover is hampered by the poor magnetic susceptibility contrasts of the host lithologies. A Sub-Audio Magnetic (SAM) survey over an exposed section of the Lena Shear recorded anomalous equivalent magnetometric resistivity (EQMMR) values proximal to the shear, but not directly associated with the geologically mapped location of the shear as expected. An interpretation of the Sub-Audio Magnetic survey highlights current channelling in structural and geological elements that were previously unknown. The relationship between the shear zone and the anomalous EQMMR response is unresolved because of the stronger response of an adjacent ultramafic unit. It is likely that deep weathering of rock units, contacts and shears has combined to produce the observed anomalies.

Mineralogy and composition of the chromitites and their platinum-group minerals from Ortaca (Mu?la-SW Turkey): evidence for ophiolitic chromitite genesis

A large number of podiform chromitite bodies of massive, disseminated and nodular type have been located in ultramafic units, composed of depleted mantle harzburgite and dunite of the Marmaris Peridotite from Ortaca (Mugla, SW Turkey). The chromite ore bodies are surrounded by dunite envelopes of variable thickness, exhibiting transitional boundaries to harzburgite host rocks. Chromitites, containing a large number of inclusions, i.e. silicates, base metal sulphides and alloys, and platinum-group minerals (PGM) have a wide range of chemical composition. The Cr# [Cr/(Cr + Al)] values of most chromitites are high (0.61–0.81) and Mg# [Mg/(Mg + Fe2 + )] values range between 0.65 and 0.71 with TiO2 content lower than 0.24 wt.%, which may reflect the crystallization of chromites from boninitic magmas in supra-subduction setting environment. Platinum-group minerals (PGM) such as laurite, erlichmanite and Os–Ir alloys, silicates such as olivine, clinopyroxene and amphibole, and base metal sulphides (BM-S), alloys (BM-A) and arsenides (BM-As) are found as inclusions in chromite or in the serpentine matrix. Platinum-group element (PGE) concentrations of the Ortaca chromitites (OC) are low in all samples. Total PGE (Ir + Ru + Rh + Pt + Pd) ranges from 63 ng/g to 266 ng/g and Pd/Ir ratios range between 0.23 and 4.75. PGE content is higher and the Pd/Ir ratio lower in Cr-rich chromitites compared to Al-rich ones. There is a strong negative correlation between the Cr# and Pd/Ir ratios (r=−0.930). The PGE patterns show a negative slope from Ru to Pt and a positive slope from Pt to Pd. The low PGE content in the majority of the OC may reflect a lack of sulphur saturation during an early stage of their crystallization. The laurite compositions show a wide range of Ru–Os substitution caused by relatively low temperature and increasing f(S2) during the chromite crystallization. The high Cr# of and hydrous silicate mineral inclusions in chromite imply that chromite crystallized in a supra-subduction setting.

Forearc structure beneath southwestern British Columbia: A three‐dimensional tomographic velocity model

This paper presents a three‐dimensional compressional wave velocity model of the forearc crust and upper mantle and the subducting Juan de Fuca plate beneath southwestern British Columbia and the adjoining straits of Georgia and Juan de Fuca. The velocity model was constructed through joint tomographic inversion of 50,000 first‐arrival times from earthquakes and active seismic sources. Wrangellia rocks of the accreted Paleozoic and Mesozoic island arc assemblage underlying southern Vancouver Island in the Cascadia forearc are imaged at some locations with higher than average lower crustal velocities of 6.5–7.2 km/s, similar to observations at other island arc terranes. The mafic Eocene Crescent terrane, thrust landward beneath southern Vancouver Island, exhibits crustal velocities in the range of 6.0–6.7 km/s and is inferred to extend to a depth of more than 20 km. The Cenozoic Olympic Subduction Complex, an accretionary prism thrust beneath the Crescent terrane in the Olympic Peninsula, is imaged as a low‐velocity wedge to depths of at least 20 km. Three zones with velocities of 7.0–7.5 km/s, inferred to be mafic and/or ultramafic units, lie above the subducting Juan de Fuca plate at depths of 25–35 km. The forearc upper mantle wedge beneath southeastern Vancouver Island and the Strait of Georgia exhibits low velocities of 7.2–7.5 km/s, inferred to correspond to ∼20% serpentinization of mantle peridotites, and consistent with similar observations in other warm subduction zones. Estimated dip of the Juan de Fuca plate beneath southern Vancouver Island is ∼11°, 16°, and 27° at depths of 30, 40, and 50 km, respectively.

SLAB-RELA TED MANTLE-METASOMATISM: A SIMS INVESTIGATION OF LLE COMPOSITION OF FINERO PHLOGOPITE-PERIDOTITE (W ALPS)

The Finero mafic-ultramafic complex (Ivrea-Verbano Zone, Western Italian Alps) offers a unique opportunity to study km-scale hydrous melt migration on a outcropping mantle body. The ultramafic mantle section is mainly formed by spinel harzburgites, which contain hydrous mineral assemblages (amphibole + phlogopite ± apatite) and show LILE (Cs, Rb, Ba, K, Sr, LREE) enrichments. Moreover, these peridotites are characterised by large radiogenic Sr and Pb and low radiogenic Nd (Voshage et al., 1987; Cummings et al., 1987; Hartmann and Wedepohl, 1993). Such petrochemical features have been supposed to be related to pervasive migration of slab-derived melts through a supra-subduction mantle wedge (e.g. Zanetti et al. 1999; Prouteau et al., 2001; Morishita et al., 2003). Alternative hypotheses involve the upward migration of different kinds of alkaline melts (from K-rich differentiates to carbonatites), in the frame of variable geodynamic environments (Grieco et al., 2001, 2004; Garuti et al., 2001; Zaccarini et al, 2004). A new SIMS investigation has been accomplished on clinopyroxenes (Cpx) from the mantle sequence of the Finero peridotite, in order to place further constraints on the nature of the ascending melt(s) and on the genetic and temporal relationships occurring among the different ultramafic units so far recognised. Micro-analytical characterisation of selected LILE, HFSE and REE was combined with the insitu determination of Li and B concentration (Light Lithophile Elements: LLE), because the fractionation of these elements represents a valuable geochemical marker about the presence of slab-derived components in uprising magmas (Ottolini et al., 2004). Five spinel-facies samples, namely one phlogopite pyroxenite, two apatite-free phlogopite harzburgites, one apatite- bearing phlogopite wehrlite and one chromitite, representative of the main Finero ultramafics were selected for this study. Pyroxenite is a pegmatitic websterite, in which clinopyroxene (the dominant phase), amphibole and orthopyroxene are up to 16, 13 and 6 mm long, respectively. Phlogopite is also present with crystals up to 3 cm long. Minerals do not show deformation. Amphibole and phlogopite normally occur as interstitial phases. Unlike pyroxenite, the peridotites show amphibole modal concentration exceeding that of clinopyroxene. The apatitefree phlogopite harzburgites (the dominant mantle lithology) have porphyroclastic textures. Olivines, and sometimes orthopyroxenes, show evidence of ductile deformation, which is absent in most of clinopyroxenes and in the hydrous minerals. In the apatite-bearing wehrlites, orthopyroxene is lacking or occurs only in trace amounts. They have a mosaic equigranular texture, consisting of crystals of equant dimension often with 120° triple junctions. Modal abundance of amphibole and phlogopite increases up to 25 and 4% by volume, respectively. Apatite occurs as interstitial grains, sometimes associated with dolomite, and within fractures cross-cutting large olivine. Dolomite is also present as inclusion in spinel. Dunite-chromitite bands have secondary coarse granular texture, modally dominated by very large olivine and spinel grains. Clinopyroxene and amphibole are randomly present as strongly anhedral interstitial phases. Locally, phlogopiterich layers occur. The B content is 3.9-9.7 ppm and 1.3-1.6 ppm in Cpx from phlogopite harzburgites and pegmatitic websterites, respectively. Such concentrations are rather large for mantle minerals and lie in the range of the lower continental crust (Leeman and Sisson, 1996). Conversely, Cpx from apatite-bearing wehrlite and dunite-chromitite bands have low B concentrations (~0.8 and ~0.6 ppm, respectively), which are consistent with those of the upper mantle (<1 ppm) (Chaussidon and Jambon, 1994). Li contents show a narrow range. However, Cpx from phlogopite harzburgites have lower Li concentrations (0.5- 1.1 ppm) than pegmatitic websterite (1.5-1.9 ppm) and apatite- bearing peridotite (1.6-1.7 ppm). The highest Li contents occur in Cpx from dunite-chromitite bands (2.0-2.4 ppm). Clinopyroxenes from websterite and phlogopite harzburgites are similar, being enriched in LILE and depleted in moderately incompatible trace elements (e.g. HREE; LaN/YbN = 15-25, YbN = 0.97-1.82 xC1). Differently, Cpx from apatite-bearing wehrlite are characterised by: i) relatively low LREE/HREE fractionation (LaN/YbN = 2, with YbN = 25 xC1); ii) marked negative HFSE anomalies. The combination of the LLE data on Cpx with the other trace elements relevant for geochemical modelling indicates the presence in the Finero mantle sequence of contrasting metasomatic overprints. In particular, the large B/Li ratios and B concentration shown by Cpx from phlogopite harzburgites and websterite confirm that they formed as the result of the migration of a melt having a crust-derived component. By contrast, the occurrence of crustal contributions is not apparent on the basis of the LLE compositions of the Cpx from apatite-bearing wehrlite and chromitite, which are more consistent with a derivation from basaltic melts of upper mantle origin. Further geochemical investigations are in progress to assure whether such different trace element compositions are actually due to the migration of several melts or can be the product of complex differentiation processes affecting a unique melt with crustal component.

Komatiites and nickel sulphide orebodies of the Black Swan area, Yilgarn Craton, Western Australia. 1. Petrology and volcanology of host rocks

The Black Swan Succession is a bimodal association of dacitic and komatiitic volcanic rocks located about 50 km NNE of Kalgoorlie, within the 2.7-Ga Eastern Goldfields greenstone province of the Yilgarn Craton. The komatiite stratigraphy comprises a steep dipping, east facing package about 700 m in maximum thickness and about 2.5 km in strike length (Fig. 1), which hosts a number of economically exploitable Ni sulphide orebodies including the Silver Swan massive ore shoot (approximately half a million tonnes at about 10.5% Ni). The sequence can be subdivided into a Lower Felsic Unit, comprising coherent and autobrecciated facies of multiple dacite lava flows; an upper Eastern and lower Western Ultramafic Unit, each showing marked lateral facies variation, and an Upper Felsic Unit coeval with the Eastern Ultramafic Unit. The komatiite sequence has been metamorphosed at sub-greenschist facies in the presence of high proportions of CO2-rich fluid, giving rise to pervasive talc–carbonate and talc–carbonate–quartz assemblages, with extensive preservation of pseudomorphed igneous textures. Cores of lizardite serpentinite are present in the thickest parts of the ultramafic succession. The degree of penetrative deformation is generally very low, and original stratigraphic relationships are largely intact in much of the sequence. The Eastern Ultramafic Unit and Western Ultramafic Unit are interpreted as components of a single large komatiite flow field, representing overlapping stages in the emplacement of a series of distributory lava pathways and flanking sheet flows. The Western Ultramafic Unit which hosts the bulk of the high-grade massive and disseminated ores is a sequence dominated by coarse-grained olivine cumulates, 2 km wide and up to 500 m thick, with major magma pathways represented by thick, homogenous olivine mesocumulate piles at its northern and southern ends: respectively 400 and 200 m thick. The sequence between the two major pathways consists of olivine orthocumulates (oOC) with minor spinifex-textured intervals. The Unit is capped by a persistent spinifex-textured crust less than 1 m thick, and is locally vesicular. The Eastern Ultramafic Unit contains the Black Swan Cumulate Zone, a 500-m thick sequence of very coarse-grained hopper-textured, locally vesicular oOC containing disseminated sulphides in its lower 200 m. The zone is flanked to the north and south by complexly interdigitated sequence of highly irregular, spinifex-capped, olivine cumulate-rich flow lobes between 1 and 100 m thick, and dacitic lavas and tuffs. The complexity of the 3-D spatial relationship of these units suggests a combination of simultaneous eruption of dacite and komatiite, combined with thermal or thermomechanical erosion. The Eastern and Western Units are interpreted as the result of more or less continuous prolonged eruption of olivine charged komatiite lava, which developed localised thermo-mechanical erosion channels in the dacitic substrate. Komatiite and dacite eruption was synchronous, giving rise to complex interdigitation and extensive contamination and hybridisation.

Geophysical characterisation of the Trident gold deposit, Western Australia

The Trident gold deposit is contained in the Plutonic Well Greenstone Belt, about 900 km northeast of Perth, Western Australia. The deposit is hosted in a highly sheared ultramfic unit overthrusted by granite. This project examined the physical properties of the important rock units to determine whether they contrast sufficiently to allow mapping of mineralisation, geology and structure. It investigated the possible occurrence at Trident of the most important mineralised lithological unit in the district, the `mine mafic'. The project also investigated the nature of the thrusted contact between granites and greenstones, one of the most important structures in the district. Drill core samples representing the major lithologies at Trident underwent a range of petrophysical testing. Gravity profiles were collected over the deposit and these were analysed along with high resolution airborne magnetic and radiometric data. The magnetic and radiometric data were assessed qualitatively with comparisons to maps of geology, regolith and known mineralisation. Gravity and magnetic data were computer modelled using drilling information and petrophysical data. The host ultramafic unit was readily mapped and modelled to depth using magnetic data. Mineralisation itself was expressed as a relative low within the magnetic ultramafic unit. The down dip extension of the mineralised ultramafic presents a good drill target. The gravity profiles are not long enough to allow accurate interpretation, but modeling shows that it is unlikely that the mine mafic is present at Trident. Radiometric data mapped broad geology and regolith trends but did not show an anomaly related to mineralisation. The results from IP and resistivity tests on core samples indicated that these techniques were not well suited for targeting this style of mineralisation.

Alteration and geochemical patterns in the 3.7–3.8 Ga Isua greenstone belt, West Greenland

Volcanic rocks of the 3.7–3.8 Ga Isua greenstone belt have undergone heterogeneous, poly-phase post-magmatic alteration, resulting in mobilisation of many elements including LREE and Th. In contrast, Nb, Ta, Ti, Zr, Y and HREE were relatively immobile during alteration processes. The degree of element mobility appears to have been controlled by chemical composition, primary volcanic structures, carbonate and silica enrichment, and proximity to ultramafic units and shear zones. On the basis of the geochemical characteristics of the least altered samples, two distinct associations have been recognised in structurally separated lithotectonic sequences. These magmatic associations are: a low-HFSE association (TiO 2 =0.20–0.40 wt.%; Zr=12–30 ppm; Nb=0.13–0.80 ppm; Y=6–14 ppm) in the central arc sequence, resembling Tertiary boninites; and a high-HFSE association (TiO 2 =0.50–1.14 wt.%; Zr=34–77 ppm; Nb=1.2–2.7 ppm; Y=11–28 ppm) in the outer and inner arc sequences, comparable to Phanerozoic island arc basaltic to picritic volcanic rocks. These two geochemically distinct sequences were likely to have been juxtaposed as a consequence of Phanerozoic-style plate tectonic processes operating in the early Earth. The Isua volcanic rocks are characterised by sub-chondritic Nb/Ta (10–16) ratios. Post-magmatic alteration can be ruled out as the cause of these low ratios. Rather, these ratios reflect a compositional characteristic of the Paleoarchean mantle source(s). The sub-chondritic Nb/Ta ratios may have resulted from either preferential sink of Nb into the core as a weakly siderophile element , or preferential retention of Nb in the subducted oceanic crust during subduction zone petrogenetic processes. Alternatively, the building blocks of the Earth were not composed of CI-like chondritic material.

GEOCHEMISTRY, PETROGENESIS AND TECTONO-MAGMATIC SIGNIFICANCE OF VOLCANIC AND SUBVOLCANIC ROCKS FROM THE KOZIAKAS MÉLANGE (WESTERN THESSALY, GREECE)

The Koziakas ophiolites have previously been interpreted as a Jurassic complex generated at a mid-ocean spreading ridge, and have been subdivided into a lower volcanic unit and an upper ultramafic unit. The data presented in this paper demonstrate that the lower volcanic unit is, in fact, a tectonic melange consisting of a polygenetic, multiple packet of stacked thrust-bound slices, representing distinct volcanic sequences and radiolarian chert successions. Volcanic sequences include pillowed and massive lava varieties, which are frequently crosscut by dikes of various nature, including boninitic dikes. Both lava varieties and dikes are predominantly aphyric, though a few samples display slightly porphyritic textures, where phenocryst assemblages include olivine and/or plagioclase in pillow and massive lavas, clinopyroxene, plagioclase and sanidine in trachytic dikes, clinopyroxene in boninitic dikes. Geochemical data point out the presence of three compositionally distinct groups of lavas: (1) transitional to alkaline basalts, trachyandesites, trachytes; (2) tholeiitic basalts; (3) very low-Ti (boninitic) basaltic andesites and andesites. The transitional to alkaline rocks displays high Nb/Y ratios and enriched incompatible element characteristics, similar to those of many intraplate oceanic island basalts (OIB). The tholeiitic group display lower abundances of incompatible elements and includes rocks resembling normal mid-ocean ridge basalts (N-MORB) with light REE (LREE) depletion (LaN/SmN = 0.29) and very low Ce/Y, Ta/Hf, Th/Yb ratios, as well as rocks resembling enriched mid-ocean ridge basalts (EMORB) showing moderate LREE enrichment (LaN/SmN = 1.26-1.52) and Ce/Y, Ta/Hf, Th/Yb ratios higher than those of N-MORBs. The very low-Ti group displays greatly depleted incompatible element abundances and the U-shaped REE patterns typical of boninites generated in suprasubduction settings. The Koziakas Melange appears to have formed due to tectonic dismemberment and accretion of material generated in an oceanic environment (MORBs and OIB-type rocks), possibly in an intra-oceanic forearc setting. After their accretion these rocks were affected by widespread boninitic dikes generated by partial melting of depleted peridotites in the fore-arc setting. The record of different lava types in the Koziakas Melange is in accordance with the general geological evolution of the Neo-Tethyan Pindos oceanic basin, from the Permo-Triassic rifting to the Middle-Late Jurassic intra-oceanic convergence phase.