Enrichment In Incompatible Elements Research Articles

The Pariquera Açu K-alkaline complex and southern Brazil lithospheric mantle source characteristics

Four of the most important K-alkaline magmatic complexes of southern Brazil are associated with two main NW-trending structures. The Lages occurrence is linked to the uplift of a large crustal block, the Lages Dome, tectonically controlled by old NW-trending fault zones. The Jacupiranga, Juquiá and Pariquera Açu (PGAM) occurrences are related to the Ponta Grossa Arch, a NW-trending uplift structure that produced tensional tectonic lineaments with the same direction and influenced the emplacement of both alkaline rocks and dyke swarms associated with Paraná flood basalt activity. The Pariquera Açu complex, petrographically and geochemically described here for the first time, comprises coarse- to fine-grained rocks representing two main suites with different degrees of alkalinity: theralite to essexite and syenogabbro to syenodiorite. In both series highly evolved products are lacking. The presence of carbonatites is inferred. In the Jacupiranga complex (ca. 130 Ma), clinopyroxenite (jacupirangite) cumulates with minor dunites are associated with ankaratrites, alkali gabbros and leucocratic rocks such as syenites, nepheline syenites and peralkaline phonolites. Many phases of both calcic and magnesian carbonatite intrusions, to which relevant fenitization processes of jacupirangite are related, characterize the district. The Juquiá complex (130–135 Ma) consists of prevalent olivine clinopyroxenite cumulates and nepheline syenites, the latter arranged in subanular concentric patterns together with subordinate ijolites–melteigites–urtites, essexites and syenodiorites. The central part of the intrusion is occupied by a small Mg-carbonatite body. The Lages complex (78–63 Ma) is made up of mafic–ultramafic rock-types (olivine melilitites, olivine nephelinites, basanites) and prevalent leucocratic bodies mainly consisting of peralkaline phonolites. Kimberlitic breccias and carbonatites are also present. Petrography, mineral and bulk-rock chemistry and mass balance calculations show that shallow-level fractional crystallization processes controlled the evolution from mafic to leucocratic rocks in all four complexes. Starting from a mantle source composition as suggested for the eastern Paraguay alkaline (ASU) magmatism, mass balance calculations indicate that primary magmas with high LILE contents and high La/Yb ratios are consistent with a garnet peridotite or phlogopite-bearing garnet peridotite solid residuum and less than 6% melting. The mantle sources are inferred to have been derived by different incompatible element enrichment probably associated with two late Proterozoic metasomatic processes: ASU (1.1 Ga) and Alto Paranaı́ba Igneous Province (APIP; 0.5 Ga). Mainly on the basis of geochemical evidence (e.g. Nb–Ta anomaly), mantle sources for PGAM and Lages Ol-melilitites and Ol-nephelinites seem to have been affected by an APIP-like event. On the other hand, an ASU-like metasomatic process is suggested for the Lages basanite mantle source, which may be indicative of vertical mantle heterogeneity. Geochemical data also point to a large-scale northward increase in concentration of incompatible elements (e.g. Lages Ol-melilitites and Ol-nephelinites vs APIP kamafugites). An attempt to correlate the chemical composition of silicate rocks and carbonate metasomatism shows that the mantle sources for PGAM and Lages occurrences plot on a mixing curve linking depleted peridotites and the average Ca-carbonatite. This fact supports the view that carbonatitic fluids/melts have been involved in the metasomatic processes responsible for variable incompatible element enrichment of the lithospheric mantle of SE Brazil.

Re-Os isotope characteristics of postorogenic lavas: Implications for the nature of young lithospheric mantle and its contribution to basaltic magmas

Re-Os isotopes have been measured on postorogenic potassic lavas from the Tibetan Plateau, the Betic domain of southeastern Spain, and the Colorado Plateau of the southwestern United States. Previous work has established that these lavas were all derived from parts of the subcontinental lithospheric mantle that had undergone metasomatic enrichment in incompatible elements, following various degrees of melt depletion. Cratonic depleted subcontinental lithospheric mantle peridotites typically have subchondritic 187 Os/ 188 Os; however, the postorogenic lavas are characterized by radiogenic 187 Os/ 188 Os ratios (0.139‐0.559). Simple modeling shows that only very large degrees of melt depletion (>25%‐30%) can lower source Re/Os ratios sufficiently to permit timeintegrated development of subchondritic 187 Os/ 188 Os ratios. Such processes may have been largely restricted to the older Precambrian, and the peridotite component of the postorogenic lavas source was probably depleted by <25%. The more radiogenic values may reflect increasing contributions from metasomatic components or possibly crustal contamination. Our findings imply the need for caution in the use of Os isotopes as a diagnostic tracer of subcontinental lithospheric mantle contributions to lavas erupted through younger Proterozoic and Phanerozoic lithosphere.

Review of geochemical variation in Lower Palaeozoic metabasites from the NE Bohemian Massif: intracratonic rifting and plume-ridge interaction

Abstract During early Palaeozoic time the Cadomian basement of the northern margin of Gondwana underwent extensive rifting with the formation of various crustal blocks that eventually became separated by seaways. Attenuation of the continental lithosphere was accompanied by the emplacement of anatectic granites and extensive mafic-dominated bimodal magmatism, often featuring basalts with an ocean crust chemistry. Intrusive metabasites in deep crustal segments (associated with granitic orthogneisses) or extrusive submarine lavas at higher levels (associated with pelagic and carbonate basinal sediments) show a wide range of chemical characteristics dominated by variably enriched tholeiites. Most crustal blocks show the presence of three main chemical groups of metabasites: Low-Titholeiitic metabasalts, Main Series tholeiitic metabasalts and alkalic metabasalt series. They differ in the degree of incompatible element enrichment (depleted to highly enriched normalized patterns), in selected large ion lithophile (LIL) to high field strength element (HFSE) ratios, and abundances of HFSE and their ratios. Both the metatholeiite groups are characterized by a common enrichment of light REE-Th-Nb-Ta. High Th values (or Th/Ta ratios) and associated low ε Nd values (especially in the Low-Ti tholeiitic metabasalts) reflect sediment contamination in the mantle source rather than at crustal levels, although this latter feature cannot be ruled out entirely. The range of chemical variation exhibited is a consequence of the melting of (a) a lithospheric source contaminated by a sediment component (which generated the Low-Ti tholeiites), and (b) a high-level asthenospheric mid-ocean ridge basalt (MORB)-type source that mixed with a plume component (which generated the range of enriched Main Series tholeiites and the alkali basalts). It is considered that a plume played an important role in the generation of both early granites and the enriched MORB-type compositions in the metabasites. Its significance for the initial fragmentation of Gondwana is unknown, but its presence may have facilitated deep continental crust melting and the fracturing into small crustal blocks. The early-mid-Jurassic plume-instigated break-up of the southern Gondwana supercontinent is considered to be a possible tectonic and chemical analogue for Early Palaeozoic Sudetic rifting and its magmatic products.

Petrology and geochemistry of late-stage intrusions of the A-type, mid-Proterozoic Pikes Peak batholith (Central Colorado, USA): implications for petrogenetic models

The ∼1.08 Ga anorogenic, A-type Pikes Peak batholith (Front Range, central Colorado) is dominated by coarse-grained, biotite±amphibole syenogranites and minor monzogranites, collectively referred to as Pikes Peak granite (PPG). The batholith is also host to numerous small, late-stage plutons that have been subdivided into two groups (e.g. Wobus, 1976. Studies in Colorado Field Geology, Colorado School of Mines Professional Contributions, Colorado): (1) a sodic series (SiO 2=∼44–78 wt%; K/Na=0.32–1.36) composed of gabbro, diabase, syenite/quartz syenite and fayalite and sodic amphibole granite; and (2) a potassic series (SiO 2=∼70–77 wt%; K/Na=0.95–2.05), composed of biotite granite and minor quartz monzonite. Differences in major and trace element and Nd isotopic characteristics for the two series indicate different petrogenetic histories. Potassic granites of the late-stage intrusions appear to represent crustal anatectic melts derived from tonalite sources, based on comparison of their major element compositions with experimental melt products. In addition, Nd isotopic characteristics of the potassic granites [ ε Nd(1.08 Ga)=−0.2 to −2.7] overlap with those for tonalites/granodiorites [ca 1.7 Ga Boulder Creek intrusions; ε Nd(1.08 Ga)=−2.4 to −3.6] exposed in the region. Some of the partial melts evolved by fractionation dominated by feldspar. The late-stage potassic granites share geochemical characteristics with most of the PPG, which is also interpreted to have an anatectic origin involving tonalitic crust. The origin of monzogranites associated with the PPG remains unclear, but mixing between granitic and mafic or intermediate magmas is a possibility. Syenites and granites of the sodic series cannot be explained as crustal melts, but are interpreted as fractionation products of mantle-derived mafic magmas with minor crustal input. High temperature and low oxygen fugacity estimates (e.g. Frost et al., 1988. American Mineralogist 73, 727–740) support a basalt fractionation origin, as do ε Nd values for sodic granitoids [ ε Nd(1.08 Ga)=+2.2 to −0.7], which are higher than ε Nd values for Colorado crust at 1.08 Ga (ca −1.0 to −4.0). Enrichments in incompatible elements (e.g. rare earth elements, Rb, Y) and depletions in compatible elements (e.g. Cr, Sr, Ba) in the sodic granitoids compared to coeval mafic rocks are also consistent with fractionation. Accessory mineral fractionation, release of fluorine-rich volatiles and/or removal of pegmatitic fluids could have modified abundances of Ce, Nb, Zr and Y in some sodic granitoid magmas. Gabbros and mafic dikes associated with the sodic granitoids have ε Nd(1.08 Ga) of −3.0 to +3.5, which are lower than depleted mantle at 1.08 Ga, and their trace element characteristics suggest derivation from mantle sources that were previously affected by subduction-related processes. However, it is difficult to characterize the mantle component in these magmas, because assimilation of crust during magma ascent could also result in their observed geochemical features. The Pikes Peak batholith is composed of at least two petrogenetically different granite types, both of which exhibit geochemical characteristics typical of A-type granites. Models proposed for the petrogenesis of the granitoids imply the existence of mafic rocks at depth and addition of juvenile material to the crust in central Colorado at ∼1.1 Ga.

Geochemistry of paleozoic basalts from the Juchatengo complex of southern Mexico: tectonic implications

Analyses of Lower Permian or older basalts and associated dykes of the Juchatengo sequence indicate that they are rift tholeiites that formed in a continental rift or back-arc tectonic setting. Age constraints include a Middle Permian fossil recovered from the tectonically overlying sediments and a cross-cutting, post-tectonic pluton dated by K/Ar on hornblende at 282±6 Ma. A location adjacent to the Oaxacan Complex or other old continental crust is suggested by (1) an εNd i isotopic value of −8.95 and a T DM age of 1487 Ma in the overlying sediments, which are similar to the Oaxacan Complex; (2) enrichment of incompatible elements in the lavas, suggesting old crustal contamination; and (3) the presence of Permian–Triassic calc-alkaline plutons that stitch the Juchatengo–Oaxaca boundary. The possible tectonic models depend on the age of the Juchatengo basalts, which requires future geochronological work. If the Juchatengo basalts are Permo-Carboniferous, they could have formed near the eastern edge of a back-arc basin: the contemporaneous arc would have rifted away to the west. Eastward migration of the arc magmatism indicated by the Permian–Triassic calc-alkaline plutonism may reflect shallowing of the dip of the subduction zone, which probably also produced the deformation of the Juchatengo sequence.

Virginia Dale intrusion, Colorado and Wyoming: Magma-mixing and hybridization in a Proterozoic composite intrusion

The Virginia Dale intrusion is a granitic and monzogranitic composite intrusive body at the southern extent of the Sherman batholith in southeastern Wyoming and adjacent areas of Colorado. Within the study area, at the margin of the intrusion, mafic rock (diorite) is commingled with felsic rock (monzogranite) and intermediate rocks. Resulting features include pillows (some with zones of felsic or mafic enrichment at their margins), enclaves, mantled quartz grains, and alkali feldspar megacrysts. Intermediate rocks interpreted as hybrids include: (1) granodioritic hybrid; (2) main-phase hybrid; (3) granitoid hybrid; and (4) mafic hybrid. Each of the hybrids shows distinct textural and mixing relations. Mixing and hybridization models were evaluated using major-element, trace-element, and isotopic data. Major-element compositions of the hybrid rocks generally plot along simple mixing lines between mafic and felsic rocks. Incompatible element concentrations of the intermediate rocks are similar to, or greater than, concentrations in the felsic end-member. Trace-element data suggest that the variety of intermediate rocks resulted from chemical diffusion in addition to bulk mixing of heterogeneous end-members. Mafic rocks show a possible differentiation trend, in which the most primitive members are cumulates. The more evolved members show progressive LREE and incompatible element enrichment, growth of a negative europium anomaly (to Eu/Eu * = 0.61), and a correlated decrease in Mg-number (to 0.35). Virginia Dale isotopic data (ϵ Nd = 0± 1) are similar to values for the adjacent Sherman batholith, suggesting that both of these bodies formed by partial melting (at 1.43 Ga) of ca. 1.8-Ga crust in the Colorado province. Initial 87 Sr/ 86 Sr values are more heterogeneous, perhaps due to chemical diffusion.

Origin and interaction of mafic and felsic magmas in an evolving late orogenic setting: the Early Paleozoic Terra Nova Intrusive Complex, Antarctica

The Abbott Unit (∼508 Ma) and the Vegetation Unit (∼475 Ma) of the Terra Nova Intrusive Complex (northern Victoria Land, Antarctica) represent the latest magmatic events related to the Early Paleozoic Ross Orogeny. They show different emplacement styles and depths, ranging from forcible at 0.4–0.5 GPa for the Abbott Unit to passive at ∼0.2 GPa for the Vegetation Unit. Both units consist of mafic, felsic and intermediate facies which collectively define continuous chemical trends. The most mafic rocks from both units show different enrichment in trace element and Sr-Nd isotopic signatures. Once the possible effects of upper crustal assimilation-fractional crystallisation (AFC) and lower crustal coupled AFC and magma refilling processes have been taken into account the following features are recognised: (1) the modelled primary Abbott Unit magma shows a slightly enriched incompatible element distribution, similar to common continental arc basalts and (2) the modelled primary Vegetation Unit magma displays highly enriched isotope ratios and incompatible element patterns. We interpreted these major changes in magmatic affinity and emplacement style as linked to a major change in the tectonic setting affecting melt generation, rise and emplacement of the magmas. The Abbott Unit mafic melts were derived from a mantle wedge above a subduction zone, with subcontinental lithospheric mantle marginally involved in the melting column. The Vegetation Unit mafic melts are regarded as products of a different source involving an old layer of subcontinental lithospheric mantle. The crustal evolution of both types of mafic melts is marked by significant compositional contrasts in Sr and Nd isotopes between mafic and associated felsic rocks. The crustal isotope signature showed an increase with felsic character. Geochemical variations for both units can be accounted for by a similar two-stage hybridisation process. In the first stage, the most mafic magma evolved mainly by fractional crystallisation coupled with assimilation of metasedimentary rocks having crustal time-integrated Sr and Nd compositions similar to those of locally exposed metamorphic basement. The second stage involves contaminated products mixing with independently generated crustal melts. Petrographic, geochemical and isotope data also provide evidence of significant compositional differences in the felsic end-members, pointing to the involvement of metaigneous and metasedimentary source rocks for the Abbott granite and Vegetation leucogranite, respectively.

Enrichment of basalt and mixing of dacite in the rootzone of a large rhyolite chamber: inclusions and pumices from the Rattlesnake Tuff, Oregon

A variety of cognate basalt to basaltic andesite inclusions and dacite pumices occur in the 7-Ma Rattlesnake Tuff of eastern Oregon. The tuff represents ∼280 km3 of high-silica rhyolite magma zoned from highly differentiated rhyolite near the roof to less evolved rhyolite at deeper levels. The mafic inclusions provide a window into the processes acting beneath a large silicic chamber. Quenched basaltic andesite inclusions are substantially enriched in incompatible trace elements compared to regional primitive high-alumina olivine tholeiite (HAOT) lavas, but continuous chemical and mineralogical trends indicate a genetic relationship between them. Basaltic andesite evolved from primitive basalt mainly through protracted crystal fractionation and multiple cycles (≥10) of mafic recharge, which enriched incompatible elements while maintaining a mafic bulk composition. The crystal fractionation history is partially preserved in the mineralogy of crystal-rich inclusions (olivine, plagioclase ± clinopyroxene) and the recharge history is supported by the presence of mafic inclusions containing olivines of Fo80. Small amounts of assimilation (∼2%) of high-silica rhyolite magma improves the calculated fit between observed and modeled enrichments in basaltic andesite and reduces the number of fractionation and recharge cycles needed. The composition of dacite pumices is consistent with mixing of equal proportions of basaltic andesite and least-evolved, high-silica rhyolite. In support of the mixing model, most dacite pumices have a bimodal mineral assemblage with crystals of rhyolitic and basaltic parentage. Equilibrium dacite phenocrysts are rare. Dacites are mainly the product of mingling of basaltic andesite and rhyolite before or during eruption and to a lesser extent of equilibration between the two. The Rattlesnake magma column illustrates the feedback between mafic and silicic magmas that drives differentiation in both. Low-density rhyolite traps basalts and induces extensive fractionation and recharge that causes incompatible element enrichment relative to the primitive input. The basaltic root zone, in turn, thermally maintains the rhyolitic magma chamber and promotes compositional zonation.

The Dir-Utror metavolcanic sequence, Kohistan arc terrane, northern Pakistan

The Dir-Utror volcanic series forms a NE–SW trending belt within the northwestern portion of the Kohistan island arc terrane in the western Himalayas of northern Pakistan. The Kohistan arc terrane comprises a diverse suite of volcanic, plutonic, and subordinate sedimentary rocks of late Mesozoic to Tertiary age, developed prior to and after suturing of the Indo-Pakistan and Asiatic continental blocks. The Dir-Utror volcanic series near Dir is dominated by basaltic-andesite and andesite, with subordinate basalt, high-MgO basalt, dacite, and rhyolite. Porphyritic textures are dominant, with less common aphyric and seriate textures. Plagioclase is the dominant phenocryst in mafic to intermediate rocks, K-feldspar and quartz phenocrysts predominate in the dacites and rhyolites. Chlorite, epidote, albite, and actinolite are the most common metamorphic phases; blue-green amphibole, andesine, muscovite, biotite, kaolinite, sericite, carbonate, and opaques are widespread but less abundant. Phase assemblages and chemistry suggest predominant greenschist facies metamorphism with epidote-amphibolite facies conditions attained locally. Whole rock major element compositions define a calc-alkaline trend: CaO, FeO, MgO, TiO 2, Al 2O 3, V, Cr, Ni, and Sc all decrease with increasing silica, whereas alkalis, Rb, Ba, and Y increase. MORB-normalized trace element concentrations show enrichment of the low-field strength incompatible elements (Ce, La, Ba, Rb, K) and deep negative Nb, P, and Ti anomalies—patterns typical of subduction related magmas. Mafic volcanic rocks plot in fields for calc-alkaline volcanics on trace element discrimination diagrams, showing that pre-existing oceanic crust is not preserved here. All rocks are LREE-enriched, with La=16–112×chondrite, La/Lu=2.6–9.8×chondrite, and Eu/Eu*=0.5–0.9. Dacites and rhyolites have the lowest La/Lu and Eu/Eu* ratios, reflecting the dominant role of plagioclase fractionation in their formation. Some andesites have La/Lu ratios which are too high to result from fractionation of the more mafic lavas; chondrite-normalized REE patterns for these andesites cross those of the basaltic andesites, indicating that these lavas cannot be related to a common parent. The high proportion of mafic lavas rules out older continental crust as the main source of the volcanic rocks. The scarcity of more evolved felsic volcanics (dacite, rhyolite) can be explained by the nature of the underlying crust, which consists of accreted intra-oceanic arc volcanic and plutonic rocks, and is mafic relative to normal continental margins. Andesites with high La, La/Lu, K 2O, and Rb may be crustal melts; we suggest that garnet-rich high-pressure granulites similar to those exposed in the Jijal complex may be restites formed during partial melting of the crust.

THE PHLOGOPITE PERIDOTITE OF THE FINERO ULTRAMAFIC COMPLEX (IVREA ZONE, NW ALPS): NEW EVIDENCE OF A METASOMATIZED MANTLE SLAB

THE PHLOGOPITE PERIDOTITE OF THE FINERO ULTRAMAFIC COMPLEX (IVREA ZONE, NW ALPS): NEW EVIDENCE OF A METASOMATIZED MANTLE SLAB

EARLY PARTIAL MELTING AND METASOMATISM N MANTLE XENOLITHS FROM NORTH VICTORIA LAND (ANTARCTICA): EVIDENCE FROM MAJOR AND TRACE ELEMENT OF SILICIC GLASSES AND PYROXENES

Cenozoic alkali basaltic magmas from Greene Point and Baker Rocks (Victoria Land, Antarctica) carry abundant ultramafic xenoliths of spinel harzburgite, spinel lherzolite and some composite xenoliths. In many harzburgitic xenoliths whole rock and trace element patterns reveal the occurrence of cryptic metasomatic enrichments in incompatible elements; pyroxenite veins are also thought to be linked to metasomatic effects (Perinelli et al., 1998). Melt pods are widespread in these material and reveal complex relation with metasomatic events and in situ partial melting. Clinopyroxenes and glasses of these rocks have been analyzed by electron microprobe for major elements and by Secondary Ion Mass Spectrometry for a set of 21 trace elements; 40 trace elements on whole rocks were determined by Inductively Coupled Mass Spectrometry to try to define the nature of metasomatism and unravel the origin of glasses. Glass compositions are silica rich (SiO2 59-70 wt%), exhibit high Mg# (0.53-0.69) and are Q-normative; in Greene point harzburgite glass pods are crystal free, while at Baker Rocks compositions of mineral coexisting with acidic glasses are in strong contrast with the host paragenesis: their olivine is Fo-rich (~0.92) while clinopyroxene has higher Al, Ca, Cr and Ti contents respect to the host rock. REE chondrite-normalized patterns of glasses in Greene Point harzburgite are characterized by LREE enrichment and show unfractionated HREE abundance, by contrast glasses in Baker Rocks harzburgite show weakly REE fractionated pattern. In coexisting clinopyroxene/glass pairs Nerst distribution coefficients were determined: at Greene Point Cpx/liqDHREE result to be ³ 1 suggesting a that unfractionated HREE of glass are linked to very low degree of partial melting of host lherzolite in accordance with experimental observation of Blundy et al. (1998). At Baker Rocks in situ partial melting is to be excluded on account of textural relations and different Cpx/liqDHREE calculated between glass and their clinopyroxene. On basis of glass and associated clinopyroxenes chemistry, the silicate melts of Baker Rocks are interpreted as product of metasomatic event in which carbonate rich melt reacted with xenolith mineralogy.

ARMALCOLITE-BEARING, TI-RICH METASOMATIC ASSEMBLAGES IN HARZBURGITIC XENOLITHS FROM THE KERGUELEN ARCHIPELAGO (INDIAN OCEAN): IMPLICATIONS FOR THE MANTLE BUDGET OF HIGH-FIELD STRENGTH ELEMENTS

Some mantle clinopyroxene- and spinel- harzburgite xenoliths from the Kerguelen Islands contain an unusual metasomatic mineral association consisting of feldspar, olivine II, chromite II and Ti-oxides (rutile, ilmenite and armalcolite). The metasomatic minerals occur in veins along the grain boundaries of the olivine, orthopyroxene and clinopyroxene but may cut across these anhydrous silicates as well as across pre-existing amphibole and the phlogopite. To our knowledge, it is the first time that such a mineral association representing a distinctive type of mantle metasomatism has been reported from peridotite samples in the oceanic mantle. This assemblage is distinct from those commonly attributed to potassium-rich hydrous fluids in cratonic mantle lithosphere or to carbonated Fe-Ti-rich silicate melts in non-cratonic mantle lithosphere. A similar mineral association has been recently described in off-cratonic continental mantle xenolith from southern Siberia (Russia, Ionov et al., 1999). The vein-forming minerals are inferred from experimental and compositional constraints to have crystallized at a minimum pressure of about 1.3 GPa and 1150-1200 °C from a strongly alkaline magma with low water activity buffered by the wall-rock harzburgites. The oxygen fugacity is well-constrained at NNO - 2.5 to - 3.0 log units. The event that formed these Ti-rich oxide veins was very recent as the veins crosscut metasomatic assemblages inferred to have formed in two previous metasomatic episodes that produced firstly anhydrous dunites and secondly Mg-augite ± phlogopite ± amphibole-bearing peridotites. In addition, the persistence of the high-energy skeletal crystal forms indicates that mantle residence time was short. Exchange reactions of the parental melt with the peridotite wall-rock released silica (from opx), chromium (from cpx and original Cr-Al spinel) and magnesium (from opx and cpx); this element exchange was essential to forming the armalcolite paragenesis. Such Ti-rich veins result in strong local enrichment in some incompatible trace elements (figure 1), including LILE that are largely concentrated in feldspar (eg Ba, Sr, LREE) and HFSE (Ti, Nb, Zr) in rutile and armalcolite, if they comprise just 0.05 wt% of the rock (Gregoire et al., 1999). At such low modal abundances, such veins could be easily overlooked or misidentified during conventional microscopic study.

CARBONATITE METASOMATISM IN THE UPPER MANTLE BENEATH THE BAKONY-BALATON HIGHLAND AND LITTLE HUNGARIAN PLAIN VOLCANIC FIELDS, WESTERN HUNGARY: EVIDENCE FROM UPPER MANTLE XENOLITHS

Lithospheric upper mantle xenoliths hosted in Neogene alkaline basalts from the Bakony-Balaton Highland and Little Hungarian Plain Volcanic Fields of the Carpathian-Pannonian region have been studied extensively petrologically, geochemically and isotopically (e.g., Embey-Isztin et al., 1989; Downes et al., 1992; Szabo et al., 1995). Based on these studies, remarkable incompatible trace element enrichment and variable eNd and eSr values of the deformed Type I. xenoliths have been recognized and interpreted by Downes et al. (1992) as being due to the host alkaline magmas and subduction-related calc-alkaline magmas or fluids. However, apparent evidence for mantle metasomatism such as presence of amphiboles and phlogopites, melt pockets, veins and fluid and/or silicate melt inclusions occurring in these xenoliths have been reported very rarely (Embey- Isztin, 1976;. Embey-Isztin et al., 1989; Szabo et al., 1995). Here we present the results of detailed petrographic and major element analyses of melt pockets and veins found in six selected upper mantle samples (four peridotites and two pyroxenites) collected at the best known xenolith locations (Szentbekkalla and Gerce) of the studied volcanic fields. The goal of this work is to characterize the mantle melts and/or fluids formed and migrated in the mantle. According to the petrography, the melt pockets (up to 4.5 mm in diameter) occur as partially crystallized multicomponent aggregates replacing partially or totally melted Cr-diopsidic clinopyroxenes or rarely pargasitic amphiboles. The melt pockets are composed of silicate glass, newly formed clinopyroxene, olivine, spinel, carbonates, ± vesicles ± sulfide blebs. The melt veins (up to 1.5 mm in thickness) occur frequently as parallel cracks crosscutting the whole xenolith. They consist mostly of silicate glass and carbonates. Newly formed clinopyroxenes, olivines and spinels are very rare. Based on electron microprobe analyses, the composition of carbonates is calcite and Mg-calcite with small amounts of FeO (up to 0.54 wt%) and MnO (up to 1.5 wt%). Compositions of silicate glasses show relatively wide silica range (50 to 59 wt%), high alumina (18 to 25 wt%) and high total alkalis (5.5 to 7.8 wt%), relatively low MgO (1.8 to 4.5 wt%) and FeO (2.4 to 6.8 wt%) content. To estimate the bulk compositions of the melt pockets and veins, we have performed mass balance calculation, based on the modes and chemistry of the melt pockets and veins. The bulk compositions of the melt pockets are basaltic and trachybasaltic somehow resembling the composition of the host basalt reported by Embey-Isztin et al. (1993) and Harangi et al. (1995). The bulk compositions of the veins fall in fields of andesite and basaltic trachyandesite close to the large area of “andesitic” silicate melt inclusions occurring in Nograd-Gomor xenoliths studied by Szabo et al. (1996).

A multielement geochronologic study of the Great Dyke, Zimbabwe: significance of the robust and reset ages

New Sm–Nd, U–Pb, and Pb–Pb age determinations indicate that the Great Dyke of Zimbabwe, an elongate intrusion of mafic and ultramafic rocks some 550 km long and between 3 and 10 km wide, is over 100 Ma older than previously believed based on Rb–Sr ages. The intrusion was emplaced as a series of subchambers with similar stratigraphy, comprising a lower ultramafic sequence with cyclic layering of dunite or harzburgite grading upwards into bronzitite, the top sections of which include Pt-enriched sulfide zones, and an upper mafic sequence of pyroxenites capped by olivine gabbro and gabbronorite. The Sm–Nd method has yielded a combined mineral/whole-rock isochron of 2586±16 Ma and ε Nd( t) of +1.1 for samples from the Darwendale, Sebakwe, and Wedza Subchambers as well as the satellite East Dyke. This isochron age is in excellent agreement with the U–Pb age for three concordant rutile fractions extracted from a feldspathic pyroxenite of the Selukwe Subchamber with an error-weighted mean at 2587±8 Ma. Two zircon fractions from the same feldspathic pyroxenite sample as the rutile are discordant, and although not well constrained, suggest Pb loss from the zircons at ca. 830 Ma. This may be related to the onset of the widespread and diachronous Pan-African tectonothermal event in southern Africa. Whole–rock samples and clinopyroxene and plagioclase separates from a Darwendale Subchamber drill core yielded a 207Pb/ 204Pb vs. 206Pb/ 204Pb isochron age of 2596±14 Ma, which is in agreement with the Sm–Nd isochron and the rutile U–Pb crystallization age. This new age information shows that emplacement of the Great Dyke and its satellite dikes closely followed the amalgamation of the Kaapvaal and Zimbabwe Cratons, and was contemporaneous with emplacement of the youngest of the trondhjemite–tonalite–granodiorite granitoid suite in the Zimbabwe Craton. Assuming that amalgamation of the Kaapvaal and Zimbabwe Cratons was largely by NNW-directed convergence, it follows that the source of the Great Dyke was asthenospheric mantle hydrated and enriched in incompatible elements by subduction processes. Isochrons of 206Pb/ 204Pb vs. 238U/ 204Pb and 207Pb/ 204Pb vs. 235U/ 204Pb yield ages with large errors, but well constrained initial Pb ratios ( 206Pb/ 204Pb = 14.15±0.30 and 207Pb/ 204Pb = 15.04±0.06). Assuming a two-stage model for common lead evolution, this result yields a μ value of 9.5. Along with the calculated initial Sr and Nd isotopic compositions, these data are consistent with derivation of the Great Dyke magmas by large volume melting of a mantle that has been hydrated and enriched by subduction. While a small amount of crustal contamination of magma derived from depleted mantle could produce the composition of the Great Dyke, the uniformity of initial ratios between subchambers supports the notion of enrichment in incompatible elements being an intrinsic characteristic of the mantle source.

Post-collisional potassic granitoids from the southern and northwestern parts of the Late Neoproterozoic East African Orogen: a review

Potassic metaluminous granitoids with enrichments of HFS elements constitute part of widespread post-collisional magmatism related to the Late Neoproterozoic Pan-African orogeny in northeastern Africa (Sudan, Ethiopia, Somalia) and Madagascar. The plutons were emplaced between 580 and 470 Ma and comprise both subsolvus and hypersolvus biotite–granite, biotite–hornblende–granite, quartz–monzonite and quartz–syenite. Pyroxene-bearing granitoids are subordinate. Basic dikes and enclaves of monzodioritic composition are locally associated with the granitoid plutons. Granitoids emplaced in pre-Neoproterozoic crust have Sr i -ratios between 0.7060 and 0.7236 and ε Nd( t) values between −15.8 and −5.6 while those emplaced in, or close to the contact with, juvenile Neoproterozoic crust have lower Sr i -ratios (0.7036–0.7075) and positive ε Nd( t) values (4.6). However, it is unlikely that the potassic granitoids represent products of crustal melting alone. The association with basic magmas derived from subduction-modified enriched mantle sources strongly suggests that the granitoids represent hybrid magmas produced by interaction and mixing of mantle and crust derived melts in the lower crust. The most intense period of this potassic granitoid magmatism occurred between 585 and 540 Ma, largely coeval with HT granulite facies metamorphism in Madagascar and with amphibolite facies retrogression in northeastern Africa (Somalia, Sudan). Granitoid magmatism and high-grade metamorphism are probably both related to post-collisional lithospheric thinning, magmatic underplating and crustal relaxation. However, the emplacement of potassic granites continued until about 470 Ma and implies several magmatic pulses associated with different phases of crustal uplift and cooling. The potassic metaluminous granites are temporally and spatially associated with post-collisional high-K calc-alkaline granites with which they share many petrographical, geochemical and isotopical similarities, except the incompatible element enrichments. The resemblance indicates a strongly related petrogenesis of both granite associations.

Isotopic (O, Sr, Nd) and trace element geochemistry of the Laouni layered intrusions (Pan-African belt, Hoggar, Algeria): evidence for post-collisional continental tholeiitic magmas variably contaminated by continental crust

The three layered intrusions studied in the Laouni area have been emplaced within syn-kinematic Pan-African granites and older metamorphic rocks. They have crystallized at the end of the regional high-temperature metamorphism, but are free from metamorphic recrystallization, revealing a post-collisional character. The cumulate piles can be interpreted in terms of two magmatic liquid lines of descent: one is tholeiitic and marked by plagioclase–olivine–clinopyroxene cumulates (troctolites or olivine bearing gabbros), while the other is calc-alkaline and produced orthopyroxene–plagioclase rich cumulates (norites). One intrusion (WL (West Laouni)-troctolitic massif), shows a Lower Banded Zone where olivine-chromite orthocumulates are interlayered with orthopyroxene-rich and olivine–plagioclase–clinopyroxene cumulates, whereas the Upper Massive Zone consists mainly of troctolitic and gabbroic cumulates. The other two massifs are more homogeneous: the WL-noritic massif has a calc-alkaline differentiation trend whereas the EL (East Laouni)–troctolitic massif has a tholeiitic one. Separated pyroxene and plagioclase display similar incompatible trace element patterns, regardless of the cumulate type. Calculated liquids in equilibrium with the two pyroxenes for both noritic and troctolitic cumulates are characterized by negative Nb, Ta, Zr and Hf anomalies and light REE enrichment inherited from the parental magmas. Troctolitic cumulates have mantle-derived δ 18 O (+5 to +6‰), initial 87 Sr/ 86 Sr (Sr i=0.7030 to 0.7054), ε Nd (+5 to −1) values whereas noritic cumulates are variably enriched in δ 18 O (+7 to +9‰), show negative ε Nd (−7 to −12) and slightly higher Sr i (0.7040–0.7065). Based on field, isotopic ratios are interpreted as resulting from a depleted mantle source (Sr i=0.7030; ε Nd=+5.1; δ 18 O =+5.1‰) having experience short term incompatible element enrichment and variable crustal contamination. The mantle magma was slightly contaminated by an Archaean lower crust in troctolitic cumulates, more strongly and with an additional contamination by an Eburnian upper crust in noritic cumulates. Lower crust input is recorded mainly by Sr and Nd isotopes and upper crust input by O isotopes. This is probably due to the different water/rock ratios of these two crust types. Assimilation of low amounts (<10%) of quartz-bearing felsic rocks, coming from both lower and upper crust, can explain the rise of SiO 2 activity, the enrichment in 18 O and 87 Sr and the lowering of ε Nd in the noritic cumulates compared to troctolitic ones. The geodynamic model proposed to account for the Laouni tholeiitic magmatism involves a late Pan-African asthenospheric rise due to a rapid lithospheric thinning associated with functioning of shear zones, which allowed tholeiitic magmas to reach high crustal levels while experiencing decreasing degrees of crustal contamination with time.

Large-Scale Separation of K-frac and REEP-frac in the Source Regions of Apollo Impact-Melt Breccias, and a Revised Estimate of the KREEP Composition

Mafic impact-melt breccias (IMB) from the Apollo landing sites—particularly Apollo 14, Apollo 15, Apollo 16, and Apollo 17—are abundant and form compositionally distinct groups. These groups exhibit a range of major-element compositions and incompatible-element enrichments. Although concentrations of incompatible elements span a significant range, inter-element ratios vary little and have been used in the past to infer a common KREEP component (KREEP = rich in potassium, rare-earth elements, phosphorus, and other alkali and high-field-strength elements). On the basis of an extensive, high-precision data set for melt-breccia groups from different Apollo landing sites, variations in trace-element signatures of the mafic impact-melt breccias reflect significant differences in KREEP components of source regions. These differences are consistent with variable enrichment or depletion of source regions in those trace elements that fractionated during the latest stages of residual-melt evolution and are more or less related to “lunar granite.” Compared to other sites, the source region of Apollo 14 impact melts had an excess of the elements that are concentrated in lunar granite, suggesting either than this source region was enriched in such a component (K-frac) or that it lost a corresponding mafic component (REEP-frac). Because these are impact-melt breccias formed in large (probably basin) impacts, the indicated geochemical separations must have occurred on a broad scale. Variations in the incompatible-element concentrations of the IMB groups reported in this paper are used to calculate a revised KREEP incompatible-element composition. On the basis of several extremely enriched lunar samples that retain the incompatible elements in KREEP-like ratios, the KREEP composition is extended to a level of 300 ppm La, or about three times the concentration of high-potassium KREEP as estimated by Warren (1989).

On the origin of incompatible element enrichment in the Venda Nova pluton, State of Espı́rito Santo, southeast Brazil

The Venda Nova pluton, central portion of the Brasiliano Coastal mobile belt, southern region of the State of Espı́rito Santo, SE Brazil, is part of a late- to post-collisional magmatic arc, formed at the end of the Pan-African–Brasiliano orogeny. These rocks form a high-K calc-alkalic suite, including gabbro-norites, monzodiorites and syeno-monzonites. Minor granitic bodies at the border zone of the pluton are probably related to melting of paragneisses during igneous emplacement, as demonstrated by their geochemistry. Concentrations of Ba, Sr and LREE in the syeno-monzonites, monzodiorites and gabbro-norites are higher than those found in ocean island basalts (OIB) and might be attributed to direct or indirect crustal contamination. Primitive-mantle normalised spidergrams indicate Rb and K depletion in the more basic rocks, and HFSE and HREE depletion for all rocks. Such geochemical signatures are also observed for many other similar igneous rocks and basic occurrences in the Brasiliano Coastal mobile belt. Ba, Sr and LREE contents of granites and regional gneisses are lower than in the Venda Nova complex and together with other geochemical characteristics cannot be regarded as the source for the enrichment in these elements observed in the Venda Nova pluton. It is suggested that the source for the Venda Nova pluton as well as for other similar plutons and basic rocks from this region, was an enriched primitive mantle. Long-lived mantle anomalies in the region have been identified by a number of studies of Phanerozoic volcanic rocks.

Temporal-compositional-isotopic trends in rejuvenated-stage magmas of Kauai, Hawaii, and implications for mantle melting processes

Primitive, low-silica and high-alkali magmas that erupt late in the evolution of most ocean island volcanoes are highly enriched in incompatible trace elements, yet their isotopic compositions require a time-integrated mantle source history of incompatible-element depletion. Reconciling these observations has traditionally required either extremely low degrees of partial melting of depleted mantle (commonly less than 0.2%), invoking an unusual mantle source recently enriched in incompatible elements or extensive melt-mantle interaction. Analyses of stratigraphic sequences of rejuvenated-stage volcanics of Kauai, Hawaii show previously unrecognized isotopic-trace element correlations, as well as temporal variations within monogenetic lava sequences that provide evidence of variable degrees of melting of several distinct mantle sources. Inter-element and isotopic-trace-element correlations indicate little or no chromatographic effects on melt compositions, inconsistent with the expected effects of significant melt-mantle reaction as the source of their incompatible-element enrichment. Trace-element compositions of rejuvenated-stage magmas can be produced by melting of typical depleted mantle sources only if they are mixtures of small- and large-degree (0.1% and 2–15%, depending on source mineralogy) melts of isotopically distinct sources. The simplest model for the Koloa Volcanics, however, consistent with previous interpretations of other Hawaiian lavas, is that they are derived from a range of incompatible-element enriched mantle sources variably metasomatized by small-degree melts of depleted mantle. Isotopic-trace-element trends in the Koloa magmas (of the opposite sense as the overall Hawaiian trend) are best explained by a positive correlation between the extent of source metasomatism and degree of melting to produce the Koloa magmas. Systematic decreases in incompatible element concentrations within individual eruption sequences probably represent sequential eruption of progressively larger-degree melt, possibly caused by vertical zonation in extent of melting in the source regions, or extraction of low-degree melts from surrounding mantle by early-ascending magma batches.

Formation of juvenile continental crust in the Arabian–Nubian shield: evidence from granitic rocks of the Nakasib suture, NE Sudan

Major and trace element data, U–Pb zircon ages, and initial isotopic compositions of Sr, Nd, and Pb are reported for ten granitic and one rhyolitic rock sample from the neo-Proterozoic Nakasib suture in NE Sudan. Chemical data indicate that the samples are medium- to high-K, "I-type" granitic rocks that mostly plot as "volcanic arc granites" on discriminant diagrams. Geochronologic data indicate that rifting occurred 790±2 Ma and constrain the time of deformation associated with suturing of the Gebeit and Haya terranes to have ended by approximately 740 Ma. Isotopic data show a limited range, with initial 87Sr/86Sr=0.7021 to 0.7032 (mean=0.7025), eNd(t) =+5.5 to +7.0 (mean=+6.4), and 206Pb/204Pb = 17.50–17.62. Neodymium model ages (TDM; 0.69–0.85 Ga; mean = 0.76 Ga) are indistinguishable from crystallization ages (0.79–0.71 Ga; mean=0.76 Ga), and the isotopic data considered together indicate derivation from homogeneously depleted mantle. The geochronologic data indicate that the terrane accretion to form the Arabian–Nubian shield began just prior to 750 Ma. The isotopic data reinforces models for the generation of large volumes of juvenile continental crust during neo-Proterozoic time, probably at intra-oceanic convergent margins. The data also indicate that crust formation was associated with two cycles of incompatible element enrichment in granitic rocks, with an earlier cycle beginning approximately 870 Ma and culminating approximately 740 Ma, and the second cycle beginning after pervasive high-degree melts – possibly hot-spot related – were emplaced approximately 690–720 Ma.