Crystal fractionation and partial melting in the petrogenesis of a Proterozoic high-MgO volcanic suite,Ungava, Québec

There is little concensus on the relative importance of crystal fractionation and differential partial melting to the chemical diversity observed within most types of volcanic suites. A resolution to this controversy is best sought in suites containing high MgO lavas such as the Chukotat volcanics of the Proterozoic Cape Smith foldbelt, Ungava, Quebec. The succession of this volcanic suite consists of repetitive sequences, each beginning with olivine-phyric basalt (19-12 wt% MgO), grading upwards to pyroxene-phyric basalt (12-8 wt% MgO) and then, in later sequences, to plagioclase-phyric basalt (7-4 wt% MgO). Only the olivine-phyric basalts have compositions capable of equilibrating with the upper mantle and are believed to represent parental magmas for the suite. The pyroxene-phyric and plagioclase-phyric basalts represent magmas derived from these parents by the crystal fractionation of olivine, with minor chromite, clinopyroxene and plagioclase. The order of extrusion in each volcanic sequence is interpreted to reflect a density effect in which successively lighter, more evolved magmas are erupted as hydrostatic pressure wanes. The pyroxene-phyric basalts appear to have evolved at high levels in the active part of the conduit system as the eruption of their parents was in progress. The plagioclase-phyric basalts may represent residual liquids expelled from isolated reservoirs along the crust-mantle interface during the late stages of volcanic activity.A positive correlation between FeO and MgO in the early, most basic olivine-phyric basalts is interpreted to reflect progressive adiabatic partial melting in the upper mantle. Although this complicates the chemistry, it is not a significant factor in the compositional diversification of the volcanic suite. The preservation of such compositional melting effects, however, suggests that the most basic olivine-phyric basalts represent primitive magmas. The trace element characteristics of these magmas, and their derivatives, indicate that the mantle source for the Chukotat volcanics had experienced a previous melting event.     

Low- and high-alumina komatiites from a Late Archaean sequence,Newton Township,Ontario

Late Archaean komatiitic lavas from Newton Township, Ontario, consist of 6 chemically distinct magma types: 3 komatiites and 3 komatiitic basalts. The succession is unusual in containing both Al- and HREE-depleted komatiites and Al- and HREE-undepleted komatiites. The two types form distinct stratigraphic units separated by komatiitic basalts. Two komatiite types are strongly LREE depleted, whilst the third and the associated komatiitic basalts range from mildly depleted to enriched. Of the six magma types, only the two strongly LREE depleted komatiites represent primary mantle melts. The other komatiite type and the komatiitic basalts were derived from the primary komatiite magmas by combinations of olivine (+chromite) fractionation, assimilation of continental crust, and magma mixing. The two primary magmas may have been derived from similar sources, their contrasting chemistry being due to differing degrees of garnet segregation during melting. A generally applicable conclusion is that a wide range of komatiitic magma types can be generated from a relatively homogeneous depleted mantle, under conditions likely to prevail during the eruption of late Archean greenstone belt sequences.     

Physical volcanology and geochemistry of Palaeoarchaean komatiite lava flows from the western Dharwar craton,southern India: implications for Archaean mantle evolution and crustal growth

ABSTRACT

Palaeoarchaean (3.38–3.35 Ga) komatiites from the Jayachamaraja Pura (J.C. Pura) and Banasandra greenstone belts of the western Dharwar craton, southern India were erupted as submarine lava flows. These high-temperature (1450–1550°C), low-viscosity lavas produced thick, massive, polygonal jointed sheet flows with sporadic flow top breccias. Thick olivine cumulate zones within differentiated komatiites suggest channel/conduit facies. Compound, undifferentiated flow fields developed marginal-lobate thin flows with several spinifex-textured lobes. Individual lobes experienced two distinct vesiculation episodes and grew by inflation. Occasionally komatiite flows form pillows and quench fragmented hyaloclastites. J.C. Pura komatiite lavas represent massive coherent facies with minor channel facies, whilst the Bansandra komatiites correspond to compound flow fields interspersed with pillow facies. The komatiites are metamorphosed to greenschist facies and consist of serpentine-talc ± carbonate, actinolite–tremolite with remnants of primary olivine, chromite, and pyroxene. The majority of the studied samples are komatiites (22.46–42.41 wt.% MgO) whilst a few are komatiitic basalts (12.94–16.18 wt.% MgO) extending into basaltic (7.71 – 10.80 wt.% MgO) composition. The studied komatiites are Al-depleted Barberton type whilst komatiite basalts belong to the Al-undepleted Munro type. Trace element data suggest variable fractionation of garnet, olivine, pyroxene, and chromite. Incompatible element ratios (Nb/Th, Nb/U, Zr/Y Nb/Y) show that the komatiites were derived from heterogeneous sources ranging from depleted to primitive mantle. CaO/Al₂O₃ and (Gd/Yb)_N ratios show that the Al-depleted komatiite magmas were generated at great depth (350–400 km) by 40–50% partial melting of deep mantle with or without garnet (majorite?) in residue whilst komatiite basalts and basalts were generated at shallow depth in an ascending plume. The widespread Palaeoarchaean deep depleted mantle-derived komatiite volcanism and sub-contemporaneous TTG accretion implies a major earlier episode of mantle differentiation and crustal growth during ca. 3.6–3.8 Ga.     

Petrogenesis of Archaean ultrabasic and basic volcanics: Evidence from rare earth elements

Rare earth element (REE) and major element data are presented on 44 Archaean samples which include spinifex textured ultramagnesian lavas (STPK) spinifex textured basalts (STB) and low MgO tholeiites. The samples come from the Yilgarn and Pilbara Blocks (W. Australia), Barberton (South Africa), Belingwe and Que Que (Rhodesia), Abitibi (Canada) and the 3.7 b.y. Isua Belt of Western Greenland. In addition REE data are given on three near primitive mid-ocean ridge basalts (MORB) and a glassy MORB-type basalt from Taiwan. We suggest that REE patterns, particularly the light REE and Eu, can be affected by metamorphism, but argue that the consistency of pattern from samples both within and between areas enables recognition of primary patterns. La/Sm ratios of 2.7 b.y. STPK are characterised by being lower than those of associated basalts. The 3.5 b.y. STPK Barberton material does not show this feature but instead displays significant heavy REE depletion. The separation of garnet from these liquids is suggested as a possible mechanism for the high CaO/Al₂O₃ ratios, (Al loss) and the heavy REE and Sc depletion. The REE data on Barberton material is equivocal on the derivation of the so-called basaltic komatiites from the peridotitic komatiites. However, REE analyses on STPK and high magnesian lavas from elsewhere suggests that crystal fractionation is not a viable mechanism to produce one from the other. We suggest instead, that varying amounts of partial melting of different sources is responsible for the spectrum of compositions. The STB appear to be an easily recognised rock type within the Archaean. They are characterised by quench (clinopyroxene) textures and a light REE enriched pattern. It is suggested that these are near primary melts and that their REE patterns mirror their mantle source. We propose a two stage model for the 2.7 b.y. mafic complexes, in which, prior to the generation of ultrabasic magmas, the source underwent a small amount of partial melting which resulted in the removal of a melt enriched in incompatible elements. The depletion process could be achieved either during mantle diapirism or by upward migration of interstitial melts into an Archaean low velocity zone. The spread of La/Sm ratios in STPK and STB is used as an argument that the Archaean mantle was chemically heterogeneous and that the degree of heterogeneity was similar to that observed in modern ocean volcanics. As a result, partial melting of the mantle under different P-T conditions produced a spectrum of magma types. The information presently available on Archaean mafic and silicic magmatism and the incompleteness of geochemical data on present day tectonic environments are two major obstacles in formulating Archaean tectonic models. In addition a comparison of present day and Archaean ultramafic and silicic rocks suggests that plate tectonic models as presently understood may not be suitable analogues for all Archaean tectonic environments.     

The Mafic and Ultramafic Lavas of the Belingwe Greenstone Belt, Rhodesia

The Belingwe Greenstone Belt (2.8 x 10⁹ yrs old) contains a7 km succession of mafic and ultramafic lavas and high-levelintrusions which overlie a thin sedimentary formation, itselfunconformable on a granitic basement. The lavas range in compositionfrom andesites (4 per cent MgO) to peridotitic komatiites (32per cent MgO). The mineralogy and textures of the most magnesianlavas demonstrate that they were extruded in a completely liquidstate. If the source mantle had an MgO content around 40 percent, then partial melts in the range 35 per cent to 55 percent would be required to produce the most magnesian liquidsobserved. Chemical constraints on the petrogenesis of the ultramafic lavasallow estimates of source mantle composition. In particular,if the source had an MgO content around 40 per cent, then theoverall source composition would be similar to that of garnetIherzolite nodules in kimberlites. The calculated REE contentsof the source are close to chondritic. If all the ultramaficlavas were derived from the same source then the variation inliquid composition may have been controlled by orthopyroxeneas well as olivine during partial melting at depth. The evolutionof the less magnesian komatiites, basalts, and andesites canbe explained by lower degrees of partial melting of a commonsource, and by high-level fractionation of parent liquids similarto those extruded as ultramafic lavas. Physical constraints on the origin of the lavas imply derivationfrom a depth of 150 km or more, at temperatures of 1600–2000°C.     

http://www.sciencedirect.com/science/article/pii/S1674987116000311

Greenstone basalts and komatiites provide a means to track both mantle composition and magma generation temperature with time.Four types of mantle are characterized from incompatible element distributions in basalts and komatiites:depleted,hydrated,enriched and mantle from which komatiites are derived.Our most important observation is the recognition for the first time of what we refer to as a Great Thermal Divergence within the mantle beginning near the end of the Archean,which we ascribe to thermal and convective evolution.Prior to 2.5 Ga,depleted and enriched mantle have indistinguishable thermal histories,whereas at 2.5-2.0 Ga a divergence in mantle magma generation temperature begins between these two types of mantle.Major and incompatible element distributions and calculated magma generation temperatures suggest that Archean enriched mantle did not come from mantle plumes,but was part of an undifferentiated or well-mixed mantle similar in composition to calculated primitive mantle.During this time,however,high-temperature mantle plumes from dominantly depleted sources gave rise to komatiites and associated basalts.Recycling of oceanic crust into the deep mantle after the Archean may have contributed to enrichment of Ti,Al,Ca and Na in basalts derived from enriched mantle sources.After 2.5 Ga,increases in Mg~# in basalts from depleted mantle and decreases in Fe and Mn reflect some combination of growing depletion and cooling of depleted mantle with time.A delay in cooling of depleted mantle until after the Archean probably reflects a combination of greater radiogenic heat sources in the Archean mantle and the propagation of plate tectonics after 3 Ga.     

Late Archaean Ti–rich,Al–depleted komatiites and komatiitic volcaniclastic rocks from the Murchison Terrane in Western Australia

Greenstone belts in the northern Murchison Terrane of the Yilgarn Craton contain an extensive suite of 2.9–3.0 Ga, porphyritic komatiites and komatiitic volcaniclastic rocks. These unusual Ti–rich Al–depleted komatiites have been sampled at Gabanintha and are characterised by higher incompatible‐element abundances than most suites of Barberton‐type Al–depleted komatiites. They form a petrogenetically related group with similar Ti– and incompatible‐element‐rich, Al–depleted porphyritic komatiites and komatiitic volcaniclastic rocks from Karasjok in Norway, Dachine in French Guiana and Steep Rock‐Lumby Lake in Canada (here called Karasjok‐type komatiites). Their Al–depletion results from magma generation at depths of >250 km in the presence of residual majorite‐garnet. The porphyritic textures and abundance of amygdales and volcaniclastic rocks typical of this type of komatiite are features of hydrous ultramafic magmas. The incompatible‐element‐rich ultramafic rocks from Dachine contain diamonds that were most likely picked up as parent magmas interacted with mantle lithosphere that had been hydrated and chemically modified. Consequently the interaction of Karasjok‐type komatiite magmas with thick, island arc or continental mantle lithosphere may have resulted in their elevated water and incompatible‐element contents. The occurrence of Karasjok‐type komatiite lavas and volcaniclastic rocks in the northern Murchison Terrane suggests that during the Late Archaean that terrane had a hydrated, metasomatised or subduction‐modified mantle lithosphere.     

Petrology and geochemistry of komatiites and tholeiites from Gorgona Island,Colombia

Komatiitic rocks from Gorgona Island, Colombia, in contrast to their Archaean counterparts, occur as rather structureless flows. In addition, textural and mineralogical features indicate that the Gorgona komatiites may have crystallized from superheated liquids. Komatiitic rocks have MgO contents which range from 24 to 11 wt.% and plot on well-defined olivine (Fo₉₀) control lines. Calculations show that potential evolved liquids (MgO<11 wt%) will be SiO₂-poor. Komatiites, in this case, cannot be regarded as parental to the associated tholeiitic basalt sequence.On the basis of REE concentrations and Sr, Nd isotopic compositions, the associated basalts are found to be of two types. One type (K-tholeiite) is characterized by noticeably fractionated REE patterns and relatively primitive isotopic compositions similar to those of the komatiites. K-tholeiites, together with komatiites, are regarded as comprising a distinctive komatiitic suite. REE patterns within this suite show progressive depletion in the LREE from K-tholeiites to komatiites, and represent increasingly higher degrees of melting of the same mantle source region. The other type (T-tholeiite), representative of the bulk of the exposed basalt sequence, has flat REE patterns and relatively evolved isotopic compositions. This tholeiitic suite is clearly genetically unrelated to the komatiitic suite.     

Geochemistry of contaminated komatiites from the Umburanas greenstone belt,Bahia State,Brazil

The late Archaean Umburanas greenstone belt (UGB) is located in the São Francisco Craton, southwest of Bahia State, Brazil. The lower unit of UGB comprises basal komatiite lavas and tholeiitic basalts intercalated with felsic volcanic rocks. The regional crystalline basement rocks, the Gavião block, predominantly consist of granitic, granodioritic and migmatitic gneiss along with tonalite-trondhjemite-granodiorite (TTG) associations.Petrographic studies of UGB komatiites reveal characteristic spinifex igneous texture although primary mineralogy is rarely preserved. Based on textural relationships, komatiites are divided into cumulate, spinifex, and massive types. The MgO content varies in the range 31.5–40.4 wt%. The MgO–SiO₂ negative correlation in komatiites suggests olivine fractionation trend. The UGB komatiites are of Al-undepleted type, characterized by Al₂O₃/TiO₂ (21–48) ratio, enriched in highly incompatible LILE relative to moderately incompatible HFSE and distinct negative Nb, Sr and Eu anomalies. Also shows depletion of light rare earths, convex-downward rare earth patterns typically not observed in komatiites world-wide, and primitive mantle normalized Gd/Yb (1.03–1.23) and La/Sm (2.36–4.99) ratios. The negative Eu anomaly is attributed to the circulation of H₂O-rich fluid, whereas the negative Nb and Sr anomalies are attributed to contamination from granitic basement rocks of the Gavião block.The UGB komatiites are most likely derived from adiabatic decompressional melting of a mantle plume. The melting took place at liquidus temperatures in the range 1572–1711 °C, which is consistent with mantle-plume origin invoked for several other komatiites in Archaean greenstone belts elsewhere. The melts were more likely generated at a depth shallower than 100 km (pressure < 2.5 GPa) where garnet was absent in the source mineralogy. Geochemical characteristics suggest contamination of primary melts with granitic basement rocks either during ascent of melt or during emplacement of magma in a continental basin setting. Greenschist to low-T amphibolite facies metamorphism at ∼2Ga may also have played a role in modifying the original komatiite petrography and composition.     

Implications of Nb/U, Th/U and Sm/Nd in plume magmas for the relationship between continental and oceanic crust formation and the development of the depleted mantle

The Nb/U and Th/U of the primitive mantle are 34 and 4.04 respectively, which compare with 9.7 and 3.96 for the continental crust. Extraction of continental crust from the mantle therefore has a profound influence on its Nb/U but little influence on its Th/U. Conversely, extraction of midocean ridge-type basalts lowers the Th/U of the mantle residue but has little influence on its Nb/U. As a consequence, variations in Th/U and Nb/U with Sm/Nd can be used to evaluate the relative importance of continental and basaltic crust extraction in the formation of the depleted (Sm/Nd enriched) mantle reservoir.This study evaluates Nb/U, Th/U, and Sm/Nd variations in suites of komatiites, picrites, and their associated basalts, of various ages, to determine whether basalt and/or continental crust have been extracted from their source region. Emphasis is placed on komatiites and picrites because they formed at high degrees of partial melting and are expected to have Nb/U, Th/U, and Sm/Nd that are essentially the same as the mantle that melted to produce them. The results show that all of the studied suites, with the exception of the Barberton, have had both continental crust and basaltic crust extracted from their mantle source region. The high Sm/Nd of the Gorgona and Munro komatiites require the elevated ratios seen in these suites to be due primarily to extraction of basaltic crust from their source regions, whereas basaltic and continental crust extraction are of subequal importance in the source regions of the Yilgarn and Belingwe komatiites. The Sm/Nd of modern midocean ridge basalts lies above the crustal extraction curve on a plot of Sm/Nd against Nb/U, which requires the upper mantle to have had both basaltic and continental crust extracted from it.It is suggested that the extraction of the basaltic reservoir from the mantle occurs at midocean ridges and that the basaltic crust, together with its complementary depleted mantle residue, is subducted to the core-mantle boundary. When the two components reach thermal equilibrium with their surroundings, the lighter depleted component separates from the denser basaltic component. Both are eventually returned to the upper mantle, but the lighter depleted component has a shorter residence time in the lower mantle than the denser basaltic component. If the difference in the recycling times for the basaltic and depleted components is ∼1.0 to 1.5 Ga, a basaltic reservoir is created in the lower mantle, equivalent to the amount of basalt that is subducted in 1.0 to 1.5 Ga, and that reservoir is isolated from the upper mantle. It is this reservoir that is responsible for the Sm/Nd ratio of the upper mantle lying above the trend predicted by extraction of continental crust on the plot of Sm/Nd against Nb/U.     

Geochemistry of rare high-Nb basalt lavas: Are they derived from a mantle wedge metasomatised by slab melts?

Compositionally, high-Nb basalts are similar to HIMU (high U/Pb) ocean island basalts, continental alkaline basalts and alkaline lavas formed above slab windows. Tertiary alkaline basaltic lavas from eastern Jamaica, West Indies, known as the Halberstadt Volcanic Formation have compositions similar to high-Nb basalts (Nb > 20 ppm). The Halberstadt high-Nb basalts are divided into two compositional sub-groups where Group 1 lavas have more enriched incompatible element concentrations relative to Group 2. Both groups are derived from isotopically different spinel peridotite mantle source regions, which both require garnet and amphibole as metasomatic residual phases. The Halberstadt geochemistry demonstrates that the lavas cannot be derived by partial melting of lower crustal ultramafic complexes, metasomatised mantle lithosphere, subducting slabs, continental crust, mantle plume source regions or an upper mantle source region composed of enriched and depleted components. Instead, their composition, particularly the negative Ce anomalies, the high Th/Nb ratios and the similar isotopic ratios to nearby adakite lavas, suggests that the Halberstadt magmas are derived from a compositionally variable spinel peridotite source region(s) metasomatised by slab melts that precipitated garnet, amphibole, apatite and zircon. It is suggested that high-Nb basalts may be classified as a distinct rock type with Nb > 20 ppm, intraplate alkaline basalt compositions, but that are generated in subduction zones by magmatic processes distinct from those that generate other intraplate lavas.     

The petrology of some volcanic rocks from the British Tertiary province: The islands of Rhum,Eigg, Canna,and Muck

On Rhum, Eigg, Canna and Muck Tertiary volcanics rest upon a Mesozoic or Pre-Mesozoic basement. Aphyric, olivine-phyric, and plagioclase-phyric basalts are recognized. The aphyric basalts are mildly alkaline or transitional types with either a few percent normative nepheline or normative hypersthene. They have anomalously low concentrations of Rb, Sr and K₂O compared to Tertiary tholeiites from the same province.Aphyric hawaiites, and mugearties are found on all the islands, but are particularly abundant on Rhum. The volcanics from Bloodstone Hill, Rhum, originally described as mugearites are anomalous in that they are quartz normative and contain both augite and hypersthene, in contrast to the normal one-pyroxene rocks of the alkali basalt-trachyte association (Muir and Tilley, 1961). These volcanics have closer affinities to the icelandites, the presence of basic plagioclase xenocrysts suggesting an hybrid origin.Olivine and plagioclase are involved in the low-pressure fractionation of the transitional basalts, whereas pyroxene and titanomagnetie play only minor roles. Consequently, the suppression of titanomagnetite crystallization results in an initial trend towards iron enrichment. The presence of both oversaturated and undersaturated derivitives following the hawaiite stage of differentiation, reflects variation in the amount of extracted pyroxene and titanomagnetite.Felsites and pitchstones intrude the volcanic pile on Eigg. The felsites carry corroded quartz crystals and rare alkali feldspar. The more crystal rich pitchstones generally contain augite, hypersthene, zoned plagioclase and titanomagnetite. One from Rudh an Tancaird contains alkali feldspar, titanomagnetite and ferrohedenbergite.Whole rock analyses and microprobe analyses of feldspars and pyroxenes indicate that the acid volcanics are not genetically related to the basalt-hawaiite-mugearite lineage. The felsites appear to have been derived from Torridonian arkose by partial melting, but the pitchstones could only be derived by anatexis of Lewisian gneiss basement (see Dunham, 1968) substantially more basic than that outcropping on Rhum.It is suggested that the low concentrations of Rb, Sr, and K₂O in the alkaline and transitional basalts, mitigates against extensive pre-eruptive differentiation. Possibly the basalts could have been derived by partial melting of a mantle depleted in these elements.     

The primary geochemistry of the monaro alkaline volcanics,southeastern Australia — evidence for upper mantle heterogeneity

The Mg numbers [100 Mg/(Mg+Fe) atomic ratios] of the Tertiary Monaro alkaline volcanics in southeastern Australia indicate that many of these alkali basalts, basanites and nephelinites have undergone only limited crystal fractionation, and that a few may represent unmodified, unfractionated primary magmas. Fractionation involves essentially olivine and clinopyroxene; fractionation trends are identified by plotting trace-element abundances against Mg number, and are then extrapolated linearly back into the primary magma field to yield estimates of the primary geochemistry of the three rock types.The nephelinites, basanites and alkali basalts are interpreted as a partial melting series derived from a peridotitic upper mantle. The estimated primary abundances of Pb,Th, Ga, V, Cr and Ni are essentially the same for the three rock types, but the estimated primary abundances of K₂O, Rb, Sr, Ba, TiO₂, Zr, Nb, P₂O₅, La Ce, Pr Nd, Y, Cu, and Zn-overlap considerably between the three rock types and the coherence of the incompatible elements as a group is not preserved in the overlaps. These patterns are best accounted for by postulating a patchy distribution of accessory phases such as amphibole, mica and apatite, in the source regions for the Monaro volcanics. Heterogeneities of this kind allow local variation in the volume of partial melt generated at the solidus, and offer a possible solution to magma segregation problems in the upper mantle.     

Petrogenesis of Cenozoic Basaltic Rocks from Jiangsu Province,China:Evidence from Geochemical Constraints

Cenozoic（Miocene to Pleistocene） basaltic rocks in Jiangsu province of eastern China include olivine tholeiite and alkali basalt.We present major,trace element and Sr-Nd isotopic data as well as Ar-Ar dating of these basalts to discuss the petrogenesis of the basalts and identify the geological processes beneath the study area.On the basis of chemical compisitions and Ar-Ar dating of Cenonoic basaltic rocks from Jiangsu province,we suggest that these basalts may belong to the same magmatic system.The alkali basalts found in Jiangsu province have higherΣFeO,MgO,CaO,Na₂O, TiO₂ and P₂O₅ and incompatible elements,but lower Al₂O₃ and compatible elements contents than olivine tholeiite which may be caused by fractional crystallization of olivine,pyroxene and minor plagioclase.In Jiangsu basaltic rocks the incompatible elements increase with decreasing MgO/ΣFeO ratios.The primitive mantle-normalized incompatible elements and chondrite-normalized REE patterns of basaltic rocks found in Jiangsu province are similar to those of OIB.Partial loss of the mantle lithosphere accompanied by rising of asthenospheric mantle may accelerate the generation of the basaltic magma.The ¹⁴³Nd/¹⁴⁴Nd vs.⁸⁷Sr/⁸⁶Sr plot indicates a mixing of a depleted asthenospheric mantle source and an EMI component in the study area.According to Shaw’s equation,the basalts from Jiangsu province may be formed by l%-5%partial melting of a depleted asthenospheric mantle source.On the basis of Ar-Ar ages of this study and the fractional crystallization model proposed by Brooks and Nielsen（1982）,we suggest that basalts from Jiangsu province may belong to a magmatic system with JF-2 as the primitive magma which has undergone fractional crystallization and evolved progressively to produce other types of basalts.     

The concentrations of the noble metals in Southern African flood-type basalts and MORB: implications for petrogenesis and magmatic sulphide exploration

Concentrations of the platinum-group elements have been determined in several suites of southern African flood-type basalts and mid-ocean ridge basalt (MORB), covering some 3 Ga of geologic evolution and including the Etendeka, Karoo, Soutpansberg, Machadodorp, Hekpoort, Ventersdorp and Dominion magmas. The magmas cover a compositional range from 3.7 to 18.7% MgO, 26–720 ppm Ni, 16–250 ppm Cu, and <1–255 ppb total platinum-group elements (PGE). The younger basalts (Etendeka, Karoo) tend to be depleted in PGE relative to Cu, while most of the older basalts (Hekpoort, Machadodorp, Ventersdorp, Dominion) show no PGE depletion relative to Cu. Further, the younger basalts tend to have lower average Pt/Pd ratios than the older basalts, and the MORBs have lower average Pt/Pd than the continental basalts within the broad groupings of "old" and "young" basalts. This may reflect (1) a decreasing degree of mantle melting through geologic time, and (2) source heterogeneity, in that the MORBs are derived from predominantly asthenospheric mantle, whereas the continental basalts also contain a lithospheric mantle component enriched in Pt. In addition to these factors, some PGE fractionation also occurred during differentiation of the magmas, with Pd showing incompatible behaviour and the other PGE variably compatible behaviour. The examined southern African flood-type basalts and MORB appear to offer limited prospects for magmatic sulfide ores, largely because they show little evidence for significant chalcophile metal depletion that could be the result of sulphide extraction during ascent and crystallization.Editorial responsibility: I. Parsons     

Hf-Nd isotope evidence for contemporaneous subduction processes in the source of late Archean arc lavas from the Superior Province, Canada

Volcanic suites from Wawa greenstone belts in the southern Superior Province comprise an association of typical late Archean arc volcanic rocks including adakites, magnesian andesites (MA), niobium-enriched basalts (NEB), and ‘normal’ tholeiitic to calc-alkaline basalts to rhyolites. The adakites represent melts from subducted oceanic crust and all other suites were derived from the mantle wedge above the subducting oceanic lithosphere. The magnesian andesites are interpreted to be the product of hybridization of adakite melts with arc mantle wedge peridotite. The initial ?_Hf values of the ∼2.7 Ga Wawa adakites (+3.5 to +5.2), magnesian andesites (+2.6 to +5.1), niobium-enriched basalts (+4.4 to +6.6), and ‘normal’ tholeiitic to calc-alkaline arc basalts (+5.3 to +6.4) are consistent with long-term depleted mantle sources. The niobium-enriched basalts and ‘normal’ arc basalts have more depleted ?_Hf values than the adakites and magnesian andesites. The initial ?_Nd values in the magnesian andesites (+0.4 to +2.0), niobium-enriched basalts (+1.4 to +2.4), and ‘normal’ arc tholeiitic to calc-alkaline basalts (+1.6 to +2.9) overlap with, but extend to lower values than, the slab-derived adakites (+2.3 to +2.8). The lower initial ?_Nd values in the mantle-wedge-derived suites, particularly in the magnesian andesites, are attributed to recycling of an Nd-enriched component with lower ?_Nd to the mantle wedge. As a group, the slab-derived adakites plot closest to the 2.7 Ga depleted mantle value in ?_Nd versus ?_Hf space, additionally suggesting that the Nd-enriched component in the mantle wedge did not originate from the 2.7 Ga slab-derived melts. Accordingly, we suggest that the enriched component had been added to the mantle wedge at variable proportions by recycling of older continental material. This recycling process may have occurred as early as 50-70 Ma before the initiation of the 2.7 Ga subduction zone. The selective enrichment of Nd in the sources of the Superior Province magmas can be explained by experimental studies and geochemical observations in modern subduction systems, indicating that light rare earth elements (e.g., La, Ce, Sm, Nd) are more soluble than high field strength elements (e.g., Zr, Hf, Nb, Ta) in aqueous fluids that are derived from subducted slabs. As a corollary, we suggest that the recycled Nd-enriched component was added to the mantle source of the Wawa arc magmas by dehydration of subducted sediments.     

Magnesian andesites, Nb-enriched basalt-andesites, and adakites from late-Archean 2.7 Ga Wawa greenstone belts, Superior Province, Canada: implications for late Archean subduction zone petrogenetic processes

Magnesian andesites (MA) occur with 'normal' tholeiitic to calc-alkaline basalt-andesite suites in four greenstone belts of the 2.7 Ga Wawa subprovince, Canada. Collectively, the magnesian andesites span ranges of SiO₂=56-64 wt%, Mg-number=0.64-0.50, with Cr and Ni contents of 531-106 and 230-21 ppm, respectively. Relative to 'normal' andesites, the magnesian andesites form distinct trends on variation diagrams, with relatively high Th and LREE contents, uniform Yb over a range of MgO, more fractionated HREE, and lower Nb/Th_pm and Nb/La_pm ratios. Niobium-enriched basalts and andesites (NEBA; Nb=7-16 ppm), and an Al-enriched rhyolite (adakite) suite are associated in space and time with the magnesian andesites. Nb-enriched basalts and andesites are characterized by high TiO₂, P₂O₅, Th, and Zr contents, variably high Zr/Hf (36-44) ratios, and more fractionated HREE (Gd/Yb_cn=1.3-4.1) compared to the 'normal' tholeiitic to calc-alkaline basalt-andesite suites. The adakite suite has the high Al (Al₂O₃=16-18 wt%), high La/Yb_cn (21-43), and low Yb (0.4-1.2 ppm) of Archean tonalite-trondhjemite-granodiorite (TTG) suites and Cenozoic adakites, indicative of liquids derived mainly from slab melting. The basalt-andesite suites are not characterized by normal tholeiitic or calc-alkaline fractionation trends of major or trace elements. Rather, compositional trends can be accounted for by some combination of fractional crystallization and variable degrees of metasomatism of the source of basalt and/or andesites by adakitic liquids. The occurrence of magnesian andesites, Nb-enriched basalts/andesites, and adakites has been described from certain Phanerozoic arcs featuring shallow subduction of young and/or hot oceanic lithosphere. Adakites likely represent slab melts, magnesian andesites the product of hybridization of adakite liquids with mantle peridotite, and Nb-enriched basalts/andesites melts of the residue from hybridization. Geological similarities between the late-Archean Wawa greenstone belts and certain Cenozoic transpressional orogens with the MA-NEBA-adakite association suggest that subduction of young, hot oceanic lithosphere may have played an important role in the production of this arc-related association in the late Archean.     

Tertiary or Mesozoic komatiites from Gorgona Island,Colombia: Field relations and geochemistry

An exceptional occurrence of ultramafic lavas within the volcanic member of the Mesozoic (or younger) Gorgona Igneous Complex represents the first known komatiites of post-Precambrian age. Gorgona komatiites are virtually unaltered and display typical spinifex textures, with 7–10 cm long plates of olivine (Fo 88 to 91) surrounded by acicular aluminous augite, subordinate plagioclase (An 56 to 78), basaltic glass, and two spinel phases. The MgO contents of the komatiites range from 15 to 22 wt.%. Sr and Nd isotopic compositions are indicative of depletion of incompatible elements in the mantle source region, as is the case for normal mid-ocean ridge basalts. The komatiites are low in total REE abundances and extremely depleted in LREE. They represent primary melts generated by high degree of partial melting of the mantle. Eruption temperatures are estimated at 1,450° to 1,500° C.     

Petrology and geochemistry of metamorphosed komatiites and basalts from the Sula Mountains greenstone belt, Sierra Leone

The Sula Mountains greenstone belt is the largest of the late-Archaean greenstone belts in the West African Craton. It comprises a thick (5 km) lower volcanic formation and a thinner (2 km) upper metasedimentary formation. Komatiites and basalts dominate the volcanic formation and komatiites form almost half of the succession. All the volcanic rocks are metamorphosed to amphibolite grade and have been significantly chemically altered. Two stages of alteration are recognised and are tentatively ascribed to hydrothermal alteration and later regional amphibolite facies metamorphism. Ratios of immobile trace elements and REE patterns preserve, for the most part, original igneous signatures and these are used to identify five magma types. These are: low-Ti komatiites – depleted in light REE; low-Ti komatiites – with flat REE patterns; high-Ti komatiitic basalts – with flat REE; low-Ti basalts – depleted in light REE; high-Ti basalts – with flat REE patterns. Much of the variation between the magma types can be explained in terms of different melt fractions of the mantle source, although there were two separate mantle sources one light REE depleted, the other not. The interleaving of the basalts and komatiites produced by this melting indicates that the two mantle sources were melted simultaneously. The simplest model with which to explain these complex melting processes is during melting within a rising mantle plume in which there were two different mantle compositions. The very high proportion of komatiites in the Sula Mountains relative to other greenstone belts suggests either extensive deep melting and/or the absence of a thick pre-existing crust which would have acted as a “filter” to komatiite eruption. Received: 10 February 1998 / Accepted: 28 July 1998     

The petrogenesis of the Lac Guyer komatiites and basalts and the nature of the komatiite-komatiitic basalt compositional gap

A sequence of ultramafic rocks in the Lac Guyer Archean greenstone belt exhibit brecciated flow tops, pillow structures, and spinifex textures testifying to their volcanic origin. Massive, spinifex-textured and differentiated flows in the sequence have the chemical characteristics of peridotitic komatiite, with MgO ranging from 19–25 wt.%. Associated pillowed flows have compositions that straddle the conventional boundary between komatiite and komatiitic basalt with MgO contents ranging from 16 to 19 wt.% MgO and are best termed pyroxenitic komatiites. Unlike other komatiitic occurrences, the peridotitic and pyroxenitic komatiites at Lac Guyer constitute a continuous chemical spectrum with no evidence of population minimum near 18 wt.% MgO. The contrasting behaviour of highly compatible elements, such as Ni and Cr, versus incompatible elements, such as Zr, indicate that this compositional spectrum was produced by a variation in the extent of partial melting (10–40%) of a garnet lherzolite source in the Archean mantle. The pyroxenitic komatiites represent liquids produced during lower (10–20%) degrees of melting during which garnet remained in the mantle residue. However, a change in slope in the distribution of Zr vs. Y between the pyroxenitic and the peridotitic komatiites indicates that garnet was completely consumed at the more extensive degrees of melting which produced the peridotitic komatiites. The Lac Guyer volcanic rocks display a population minimum at 15 wt.% MgO separating komatiitic magmas whose compositions are controlled by partial melting from basalts whose composition is controlled by crystal fractionation. The population minimum near 18 wt.% MgO which is taken as the boundary between komatiite and komatiitic basalt may have a similar origin.