The Pleistocene-Recent Tonga-Kermadec Arc Lavas: Interpretation of New Isotopic and Rare Earth Data in Terms of a Depleted Mantle Source Model

New REE data, and new Nd, O, Sr, and Pb isotopic data are presentedand integrated with previous data for this low-K intra-oceanicarc suite. Geochemically, the arc tholeiites and basaltic andesitesrange from extremely HFS element depleted (northern Tonga) tonear N-MORB-like HFS element abundances in L'Esperance (southernKermadecs). LIL elements (Sr, Rb, K, Rb, Ba, Th) show the characteristicselective enrichment generally recognized in arc magmas, andthus indicate decoupling of the HFS and LIL elements. Modellingsuggests a compositionally variable source (mantle wedge) alongthe arc, ranging from restite after remelting an N-type MORBsource (northern end), to progressively less depleted, MORB-likesources southwards. Thus, the low HFS/LIL element ratios areinterpreted in terms of HFS depletion followed by LIL elementenrichment associated with subduction; broad correlations occurbetween Zr/Ba and Sr/Nd ratios (fractionation corrected) and87Sr/86Sr and ¹⁴³Nd/¹⁴⁴Nd ratios. Derivation of the arc magmas from depleted peridotote requiressuperimposed fractional crystallization, which has been modelledthermodynamically using SILMIN (Ghiorso, 1985), utilizing experimentallyproduced partial melts from depleted lherzolite (Jaques &Green, 1980). It is shown that the arc tholeiites and basalticandesites (and also high Mg-andesites) are potentially developedat low pressures 5 kb), from parental magmas also generatedat relatively low pressure ( 10 kb). These data further suggestthat a southward increasing depth of magma segregation (correlatingwith Benioff Zone geometry) could account for differences inchemistry between the Tonga and Kermadec arc segments. The mechanism of LILE enrichment is still highly problematic,but it is suggested that the model of Tatsumi et al. (1986)may account for much of the geochemical data; this involvesrelatively shallow release, via fluids, of LIL elements intooverlying peridotite beneath the fore-arc region. Induced convectionin the mantle wedge moves the metasomatized mantle into thezones of magma generation. The development of the inferred, variably depleted mantle wedgesource is here related to active back-arc spreading which isslightly older and more rapid behind the northern region ofthe arc. The lherzolite restite from this spreading processis interpreted to undergo further partial melting in the metasomatizedmantle wedge overlying the subduction zone, involving inducedconvection. The back-arc island of Niua fo'ou is geochemically quite distinctfrom the arc magmas, being similar to N-MORB in its trace elements,but to OIB in its isotope ratios.     

Origin of some magmas in oceanic and circum-oceanic regions

Partial melting experiments on spinel-lherzolite, a rock which probably occurs in relatively shallow parts of the oceanic upper mantle, demonstrate that alkali basaltic melt is formed at depths of at least 20 kbar whereas tholeiitic melt is formed at lower pressures (< 15 kbar) under anhydrous conditions. The specimen studied was a relatively iron-rich natural spinel-lherzolite (Fe/Mg+Fe=0.15) and the melts produced have ratios comparable to those obtained in basalts. Slight increase of degree of partial melting produces picritic melt over a wide pressure range. Under hydrous (water-excess) conditions, andesitic melt is produced by partial melting of the same natural spinel-lherzolite and a synthetic lherzolite. The melting experiments on two different abyssal tholeiites from the Mid-Atlantic Ridge suggest that the derivation of olivine tholeiite from a more mafic magma or a mantle peridotite (lherzolite) is possible, but is limited to depths shallower than 25 km under essentially anhydrous conditions, whereas plagioclase tholeiite may have been formed by fractional crystallization at depths of about 20 km in the presence of a small amount (～ 2 wt.%) of water.It is suggested that under mid-ocean ridges, partial melting of spinel-lherzolite at depths shallower than 60 km would produce olivine-tholeiitic magma, which differentiates at shallower levels (20–25 km) under either essentially anhydrous or hydrous (but vapor-absent) conditions to produce abyssal tholeiites of olivine-tholeiite type or plagioclase-tholeiite type. It may be also possible that the former olivine-tholeiite is generated by direct partial melting of plagioclase-lherzolite. Alkali basalts in the oceanic region may be generated at depths greater than 50 km by relatively small degree of partial melting. Along island arcs and continental margins, where the subduction zones probably exist, partial melting of lherzolite would take place in the presence of water that may be supplied by breakdown of hydrous minerals in the subducted oceanic crust, thereby producing andesitic magmas. High-alumina basalt magma could be produced by partial melting of the dehydrated oceanic crust in the subduction zone at depths between 40 and 60 km, where garnet is unstable above the solidus.     

A mantle or mafic crustal source for Proterozoic anorthosites?

Calculations of fractional crystallization (FC) and assimilation fractional crystallization (AFC) at 11 kb for a variety of primitive magmatic compositions and a mafic assimilant demonstrate that none of them has a bulk composition suitable to be parental to massif anorthosites. Mafic compositions thought to be parental to massif anorthosites have Mg′ values of 0.6 to 0.4 and form coherent arrays with moderately steep slopes on plots of TiO₂, K₂O, and P₂O₅ versus Mg′. The calculated liquid lines of descent (LLD) of basaltic magmas undergoing FC or AFC processes pass through the arrays of anorthosite parent magma compositions with much shallower slopes than the natural arrays, which indicates that the arrays of natural parental magmas were produced by a process other than FC/AFC. Also, by the time most crystallizing basaltic magmas with or without assimilation reach plagioclase saturation, their residual liquids have Mg′ values that are too low to be parental to anorthosites. MORB-like olivine tholeiites and high-aluminum olivine tholeiites (HAOT) from convergent plate margins do reach plagioclase saturation while sufficiently magnesian, but their Wo (Wollastonite) contents are too high such that they reach plagioclase saturation coexisting only with augite and do not reach orthopyroxene saturation (if at all) until Mg′ is too low. Calculations show it is not possible to produce a high-Al melt from typical mantle peridotites that has sufficient TiO₂ to make andesine-type anorthosite.

Calculation of partial melting for an average mafic crustal composition at 11 kbar provides a much closer match to the array of natural parental compositions in terms of minor element concentrations and proportions of mineral components. However, accounting for the entire array requires a more magnesian source composition. Such compositions exist in several crustal xenolith localities. Similar results were obtained using the bulk composition of the Stillwater Complex, which is used as a model mafic source (here the premise is that overdense crustal intrusions might sink back into the mantle). As with the terrain composition, this particular layered intrusion composition is not sufficiently magnesian, however, the fit improves when mixtures of early and late stage portions of the complex (i.e., the denser portions) were run as potential source regions.     

Contemporaneous Convergent Margin and Intraplate Magmatism, North Island, New Zealand

A convergent margin magma series with characteristic low Nband Ta abundances and enrichments in alkalis and alkaline earthsis intercalated with typical intraplate alkalic basalts in aback-arc setting, 200–250 km above the Wadati-Benioffzone on the North Island, New Zealand. These two contrastingmagma types, together with late-stage K-rich maflc lavas, wereerupted over a short time period (1{dot}60–2{dot}74 Ma)and constitute the Alexandra Volcanics. Field relationshipsindicate that these diverse magma types were contemporaneous,and thus their mantle source regions coexisted, in a singletectonic environment. The convergent margin magma series forms a linear chain of stratovolcanoesaligned at right angles to the present subduction zone. Closed-systempolybaric fractional crystallization models can explain theevolution from ankaramites to transitional olivine basalts toolivine tholeiites to high-Al basalts to medium- and high-Kandesites. The most primitive lavas have geochemical (high LIL/LREEand LIL/HFS element ratios) and Sr, Nd, and Pb isotopic compositionstypical of convergent margin magmas. Calculated source compositionssuggest that three components are involved: a MORB component,a component derived from subducted oceanic crust, and a contributionfrom subducted sediments. The alkalic basalts occur as dispersed monogenetic volcanoesand are intercalated with the larger convergent margin stratovolcanocs.These basalts are enriched in LILE, LREE, Nb, and Ta, and havelow Ba/Nb and Ba/La ratios, all of which are characteristicof ocean island (intraplate) basalts (OIBs). Their relativelyhigh _Nd (+5{dot}5 and low ⁸⁷Sr/⁸⁶Sr(0{dot}703l–0{dot}7036)are also typical of OIBs. These alkalic magmas were derivedfrom the underlying continental lithospheric mantle that hasbeen enriched by upward-migrating silica-undersaturated melts,probably including volatiles, from the low- velocity zone. Asubducted slab component is not required to account for theirincompatible element enriched character. The K-rich mafic lavas, basanites, and absarokites are volumetricallyminor and cap the largest of the stratovolcanoes, Pirongia.The basanites have geochemical and isotopic compositions whichsuggest they are mixtures of multiple source components, includingthe alkalic and convergent margin region.     

Significance of orthopyroxene and major element constraints on the petrogenesis of Ferrar tholeiites from southern Prince Albert Mountains, Victoria Land, Antarctica

The least evolved Jurassic Ferrar tholeiites from southern Prince Albert Mountains (Antarctica) are characterized by the occurrence of orthopyroxene (opx), a mineralogical feature virtually absent in the tholeiites from the large igneous provinces of Karoo (South Africa) and Paraná (Brazil). Petrography suggests that opx is the early phase in the sequence of crystallization and mineral chemistry indicates that it is in equilibrium with the host rock. In general, MELTS modeling predicts that opx is the liquidus phase in the Ferrar tholeiites with MgO higher than ~7 wt% at P=1.5-5 kbar, H₂O=0-1 wt% and % MathType!MTEF!2!1!+- % feaaeaart1ev0aaatCvAUfKttLearuavTnhis1MBaeXatLxBI9gBae % bbnrfifHhDYfgasaacH8YjY-vipgYlH8Gipec8Eeeu0xXdbba9frFj % 0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs0dXdbPYxe9 % vr0-vr0-vqpWqaaeaabiGaaiaacaqabeaabaqaamaaaOqaaiabdAga % MTWaaSbaaeaacqqGpbWtaeqaaWWaaSbaaeaadaWgaaqaaiabikdaYa % qabaaabeaakiabg2da9iabbgfarjabbAeagjabb2eanjabgkHiTiab % igdaXiaaysW7cyGGSbaBcqGGVbWBcqGGNbWzaaa!381A! = QFM - 1 f_{\rm O} _{_2 } = {\rm QFM} - 1\;\log unit conditions. MELTS results also show that the early crystallization of opx is primarily controlled by high SiO₂ and high SiO₂/CaO, chemical characteristics typical of the Ferrar tholeiites, but not shown by the Karoo and Paraná analogs with similar MgO content. Major element geochemistry of the least evolved Ferrar tholeiite has been modeled through fractional crystallization and fractional crystallization coupled with crustal assimilation processes, starting from natural peridotite-derived experimental melts. Mass balance and MELTS modeling support the argument that theoretical magma compositions suitable to be primary to the least evolved Ferrar tholeiites are compatible with hydrous (H₂O=0.3-0.5 wt%) and anhydrous melts obtained at 10-15 kbar by high melting degrees (>25%) of fertile and depleted spinel lherzolites, respectively, and later contaminated by the high-grade metamorphic rocks from the Victoria Land crystalline basement. The genesis of primary Ferrar tholeiites does not necessarily reflect the generally assumed depleted source mantle being also compatible with a fertile one.     

The Volume and Composition of Melt Generated by Extension of the Lithosphere

Calculation of the volume and composition of magma generatedby lithospheric extension requires an accurate mitial geotherm,and knowledge of the variation and composition of the melt fractionas a function of pressure and temperature. The relevant geophysicalobservations are outlined, and geotherms then obtained fromparameterized convective models. Experimental observations whichconstrain the solidus and liquidus at various pressures aredescribed by simple empirical functions. The variation in meltfraction is then parameterized by requiring a variation from0 on the solidus to 1 on the liquidus. The composition of the melts is principally controlled by themelt fraction, though those of FeO, MgO, and SiO² in additionvary with pressure. Another simple parameterization allows theobserved compositions of major elements in 91 experiments tobe calculated with a mean error of 1.1%, and those of TiO₂ andNa₂O to 0.3%. These expressions are then used to calculate theexpected compositions of magma produced by adiabatic upwelling.The mean major element composition of the most magnesium-richMORB glasses resemble the mean composition calculated for amantle potential temperature T_p of 1280?C. Adiabatic meltingduring upwelling of mantle of this temperature generates a meltthickness of 7 km. The observed variations of the MgO and TiO₂concentrations in a large collection of MORB glass compositionssuggest that extensive low pressure fractional crystallizationoccurs, but that its effect on the concentrations of SiO₂, Al₂O₃,and CaO is small. There is no evidence that normal oceanic crustis produced from magmas containing more than 11% MgO. The mantlepotential temperature within hot rising jets is about 1480?Cand can generate 27 km of magma containing 17% MgO. Extension of the continental lithosphere generates little meltunless ß> 2 and T_p> 1380?C. The melts generatedby larger values of ß or of T_p are alkali basalts,and change to tholeiites as the amount of melting increases.Large quantities of melt can be generated, especially at continentalmargins, where estimates of ß obtained from changesin crustal thickness will in general be too small.     

西昆仑普鲁新生代火山岩的矿物化学特征及其对岩浆演化过程的约束

对西昆仑普鲁新生代火山岩的矿物学进行了系统的研究。结果表明：该地区火山岩主要由橄榄石、单斜辉石和斜长石组成,并有少量的斜方辉石、黑云母、角闪石、碱性长石和铁钛氧化物。其矿物学特征指示了岩浆的性质有点类似于碱性岩浆,但与典型的碱性玄武岩又有明显的区别,属于橄榄安粗岩系列。利用橄榄石-熔体平衡原理估算了进入高位岩浆房中的熔体的MgO含量约为6．2％,Mg^#为0．57,说明其不是地幔熔融形成的原始岩浆,而是经历了深部岩浆房的分离结晶过程。由单斜辉石估算的高位岩浆房的深度约7～9km。岩浆在高位岩浆房中发生了较长时间的强烈分离结晶作用,分离结晶相主要为橄榄石、单斜辉石和斜长石以及少量的斜方辉石、黑云母、角闪石、碱性长石和铁钛氧化物。不同时期形成的铁钛氧化物指示了分离结晶过程由相对高温高氧逸度向相对低温低氧逸度演化。与此相对照的是岩浆在深部岩浆房中可能只发生了橄榄石和辉石等铁镁矿物的分离结晶作用,且分异作用时间较短。深部岩浆房可能存在于岩石圈地幔或壳幔过渡带中,岩浆由深部岩浆房上升到高位岩浆房中的过程是近绝热的,从浅部岩浆房到地表是快速上升的过程。     

Olivine Tholeiites from Krafla, Iceland: Evidence for Variations in Melt Fraction within a Plume

Olivine tholeiites (8–10 wt. % MgO) from Krafla show significantcorrelations between major elements (notably Fe) and incompatibletrace elements. In particular, the samples with the highestFe contents are the most enriched in elements such as K, Ti,and light rare earth elements (LREE_s). The observed trends cannotbe explained by fractional crystallization of olivine, plagioclase,or clinopyrox-ene from a single primary magma, nor are theylikely to result from crustal contamination. The simplest explanationfor the compositional variations is that they result from imperfectmixing of primary melts, produced at different levels in theupwelling asthenosphere, which later underwent olivine fractionation.Nd and Sr isotopic data hint at the possibility that some mixingbetween two (plume and non-plume) mantle sources may also berequired. The average olivine tholeiite composition is comparedwith the average compositions of melts, predicted from parameterizationsof melting experiments, produced from mantle with differentpotential temperatures. The predicted compositions were correctedfor fractional crystallization before the comparison was made.The data compare well with the predicted average compositionof melt from mantle with a potential temperature of {small tilde}1580C. Differences between the observed and predicted compositions(notably higher Fe and lower Na in the Krafla basalts) are ascribedeither to errors related to the modelling or to the effect oftemperature- and velocity-structure of the mantle plume beneathIceland. The average REE composition of the olivine tholeiiteswas then inverted to obtain the variation of melt fraction withdepth. The predicted melt fraction rises from 00 at a depthof {small tilde} 140 km (consistent with a potential temperatureclose to 1580 C) to a maximum value of {small tilde} 03 atthe surface. The predicted melt thickness ({small tilde}22 kmwhen corrected for fractional crystallization) is consistentwith geophysical estimates of crustal thickness.     

Petrology of Anorthite-bearing Gabbroic Anorthosite Dykes in northwest Skye

A group of gabbroic anorthosite dykes in northwest Skye containsmegacrysts of calcic plagioclase (An_93–87) and olivine(FO_87–78.5and plagioclase-rich xenoliths set in a matrixof low-alkali, high Ca olivine tholeiite. Field, mineralogicand petrochemical evidence indicates (1) that the dykes arefilled with crystals concentrated from magmas feeding lava flowsand (2) that, although the megacrysts crystallized from low-alkali,high Ca olivine tholeiite magmas at depths <4 km, the megacrystsand the matrix in each dyke are not cognate. The emplacementtemperature was less than 1230 °C. Low-alkali, high Ca olivinetholeiite is an abundant Hebridean magma type, resembling unfractionatedoceanic tholeiites from the Atlantic. The temporal change fromalkali olivine basalt to low-alkali, high Ca olivine tholeiitein the Hebridean proto-spreading centre may mimic compositionalchanges during initial development of the mid-Atlantic ridge.     

Mineralogy and proposed P-T paths of basaltic lavas from Rabaul caldera,Papua New Guinea

Rabaul caldera is a large volcanic depression at the north-east tip of New Britain, Papua New Guinea. The lavas range in composition from basalt to rhyolite and have a calc-alkalic affinity but also display features typical of tholeiites, including moderate absolute iron enrichment in flows cropping out around the caldera. The basalts contain phenocrysts of plagioclase and clinopyroxene with less abundant olivine and titanomagnetite. In the basaltic andesites olivine is rare, while orthopyroxene and titanomagnetite are common along with plagioclase and clinopyroxene. Orthopyroxene is also found mantling olivine in some of the basalts while in both rock types pigeonitic augite is a fairly common constituent of the groundmass. Plagioclase in both basalt and basaltic andesite often exhibits sieve texture and analysis of the glass blebs show them to be of similar composition to the bulk rock. Phenocrystic clinopyroxene is a diopsidic augite in both basalt and basaltic andesite. Al₂O₃ content of the clinopyroxene is moderately high (4%) and often shows considerable variation in any one grain. Calculations show that the microphenocrysts probably crystallised near the surface, while phenocrysts crystallised at around 7 kb (21 km). Neither the basalts nor the basaltic andesites would have been in equilibrium at any geologically reasonable P and T with quartz eclogite. Equilibration between mantle peridotite and a. typical Rabaul basaltic liquid could have occurred around 35 kb and 1270 °C. A basaltic andesite liquid yields a temperature of 1263 °C and a pressure of 28 kb for equilibration with mantle peridotite.Partial melting of sufficient volumes of mantle peridotite at these P's and T's requires about 15% H₂O, but there is no evidence that these magmas ever contained large amounts of water. It is proposed that the Rabaul magmas were initially generated by partial melting of subducted lithosphere and subsequently modified by minor partial melting as they passed through the overlying mantle peridotite.     

Trace-Element Evidence of Multistage Mantle Fusion and Polybaric Fractional Crystallization in the Palaeocene Lavas of Skye, NW Scotland

The Palaeocene magnesian transitional basalts of the Main LavaSeries (SMLS) of Skye, NW Scotland, have concentration rangesof K, Ti, P, Rb, Sr, Ba, Nb, Ta, Zr, Hf, Th and light REE varyingby a factor of up to two at a given value of (FeO + Fe₂O₃)/(FeO+ Fe₂O₃ + MgO). Their chondrite-normalized REE patterns varywidely in slope and cuvature, with (Ce/Yb)_N=2.2–4.7. Theabundances of Ti, P, Zr, Hf, Eu, Gd and Tb correlate negativelywith Si-saturation and are thought to be primary, reflectingvariable localized partial melting (5 per cent) for each magmabatch at about 60 km depth of a spinel-lherzolite upper mantle,leaving a lherzolitic residuum. Y and the heavy REE vary littlewith Si-saturation, due to their partial retention in residualmantle diopside. The large abundance ranges of Rb, Sr, Ba, Nb,Th, La, Ce and Nd in the SMLS basalts, uncorrelated with Si-saturation,may reflect local upper-mantle variability in the concentrationsof the ultra-incompatible elements beneath Skye, caused by thepre-Palaeocene extraction of small quantities of alkalic, incompatible-element-richmagma, such as formed the Permian lamprophyre dykes of westernScotland. The trace element data confirm major-element, least-squaresmodels, which show that fractional crystallization of SMLS magnesianbasalt to less-magnesian basalt residua involved the separationof 10 per cent olivine and 4 per cent plagioclase, whilst thefractionation of SMLS less-magnesian basalt to hawaiite occurredat about 35 km depth by precipitation of 8 per cent olivine,15 per cent plagioclase and 21 per cent aluminous sub-calcicaugite. The variation of Nb and Ta abundances in hawaiites,mugearites and low-Fe intermediate lavas suggests that theseelements partitioned strongly into liquidus titanomagnetitemicrophenocrysts. Zircon fractionation occurred during the finalstages of evolution of benmoreites and trachytes, the latterrepresenting the residuum of at least 90 per cent fractionalcrystallization of SMLS basalt magma. High-Ca, low-alkali olivine tholeiites of the Preshal Mhor magmatype occur near the top of the present lava field erosionalremnant and predominate in the dyke swarm transecting it. Theyhave low incompatible trace-element abundances and REE patternswith (Ce/Yb)_N 0.6, similar to those of many mid-ocean ridgebasalts. Models attempting to explain the genesis and relationsof the contrasting SMLS and Preshal Mhor basalts by postulatingseparate mantle sources, arranged in plumes, blobs or layers,fail to account satisfactorily for: (1) the constant incompatible-elementratio ranges of all Skye basalt lavas, (2) the partial interstratificationof SMLS and Preshal Mhor basalts and (3) the appearance of PreshalMhor magmas at the climax of crustal extension in the dyke swarm.The contrasting REE patterns of SMLS transitional basalts andPreshal Mhor tholeiites, and the high Ca/Al, Ca/Na and Ca/Tiof the latter, can be explained if they were produced by furtherlimited partial fusion of the lherzolitic residuum (with a trappedmelt fraction) from SMLS genesis, leaving a final harzburgiticresiduum. The petrogenesis of the Skye lavas provides a localshort-timespan analogue of worldwide processes involved in thegeneration of mid-ocean ridge basalts.     

Effect of Water on the Composition of Magmas Formed at High Pressures

Portions of the system MgO-CaO-Na₂O-Al₂O₃-SiO₂-H₂O have beenstudied in the pressure range 13–35 kb at near-liquidustemperatures. The liquidus field of forsterite relative to thatof orthopyroxene is considerably wider under anhydrous thanunder anhydrous conditions and it covers part of the plane ofsilica-saturation in a wide pressure range. Partial meltingof simple garnet lherzolite (= forsterite+orthopyroxene+clinopyroxene+garnet)with water produces quartz-normative liquids at pressures upto at least 25 kb regardless of water content. Hydrous mineralsare not encountered at or near the solidus temperatures exceptin a Na-rich part of the system. Microprobe analysis of therun products in this synthetic system shows that the liquid(glass) in equilibrium with the lherzolite mineral assemblageis silica and alumina-rich at 20 kb under vapor-present conditions.With increasing degree of partial melting, the liquid changesits composition, passing into a ‘vapour-absent region’and becoming less silicic. Fractional crystallization of olivinetholeiitic magma under hydrous conditions also produces silica-richmagmas at high pressures. If the system is open to water, andwater pressure is less than total pressure, the compositionof the liquid varies from quartz-normative to olivine (±nepheline)-normativedepending on water pressure. It is suggested that in the presenceof water, silica-rich magmas such as those of calc-alkalic andesiteor dacite may be formed by direct partial melting of the peridotiticupper mantle at depths down to about 80 km. A large degree ofpartial melting of lherzolite under hydrous conditions wouldproduce SiO2 and MgO-rich magmas. The clinoenstatite rock fromCape Vogel, Papua, may have been formed by such a process. Peridotiteswith low CaAl2SiO5/jadeite ratios in the clinopyroxene couldproduce nepheline-normative magma by small degree of partialmelting and tholeiitic magma by large degree of partial meltingunder hydrous conditions.     

Melting and phase relations in the plane tholeiite-lherzolite-nepheline basanite to 40 kilobars with geological implications

Six crystalline mixtures, picrite, olivine-rich tholeiite, nepheline basanite, alkali picrite, olivine-rich basanite, and olivine-rich alkali basalt were recrystallized at pressures to 40 kb, and the phase equilibria and sequences of phases in natural basaltic and peridotitic rocks were investigated.The picrite was recrystallized along the solidus to the assemblages (1) olivine+orthopyroxene+ clinopyroxene +plagioclase+spinel below 13 kb, (2) olivine+orthopyroxene+clinopyroxene+spinel between 13 kb and 18 kb, (3) olivine+orthopyroxene+clinopyroxene+ garnet+spinel between 18 kb and 26 kb, and (4) olivine+clinopyroxene+garnet above 26 kb. The solidus temperature at 1 atm is slightly below 1,100° and rises to 1,320° at 20 kb and 1,570° at 40 kb. Olivine is the primary phase crystallizing from the melt at all pressures to 40 kb.The olivine-rich tholeiite was recrystallized along the solidus into the assemblages (1) olivine+ clinopyroxene+plagioclase+spinel below 13 kb, (2) clinopyroxene+orthopyroxene+ spinel between 13 kb and 18 kb, (3) clinopyroxene+garnet+spinel above 18 kb. The solidus temperature is slightly below 1,100° at 1 atm, 1,370° at 20 kb, and 1,590° at 40 kb. The primary phase is olivine below 20 kb but is orthopyroxene at 40 kb.In the nepheline basanite, olivine is the primary phase below 14 kb, but clinopyroxene is the first phase to appear above 14 kb. In the alkali-picrite the primary phase is olivine to 40 kb. In the olivine-rich basanite, olivine is the primary phase below 35 kb and garnet is the primary phase above 35 kb. In the olivine-rich alkali basalt the primary phase is olivine below 20 kb and is garnet at 40 kb.Mineral assemblages in a granite-basalt-peridotite join are summarized according to reported experimental data on natural rocks. The solidus of mafic rock is approximately given by T=12.5 P _Kb+1,050°. With increasing pressure along the solidus, olivine disappears by reaction with plagioclase at 9 kb in mafic rocks and plagioclase disappears by reaction with olivine at 13 kb in ultramafic rocks. Plagioclase disappears at around 22 kb in mafic rocks, but it persists to higher pressure in acidic rocks. Garnet appears at somewhat above 18 kb in acidic rocks, at 17 kb in mafic rocks, and at 22 kb in ultramafic rocks.The subsolidus equilibrium curves of the reactions are extrapolated according to equilibrium curves of related reactions in simple systems. The pyroxene-hornfels and sanidinite facies is the lowest pressure mineral facies. The pyroxene-granulite facies is an intermediate low pressure mineral facies in which olivine and plagioclase are incompatible and garnet is absent in mafic rocks. The low pressure boundary is at 7.5 kb at 750° C and at 9.5 kb at 1,150° C. The high pressure boundary is 8.0 kb at 750° C and 15.0 kb at 1,150° C. The garnet-granulite facies is an intermediate high pressure facies and is characterized by coexisting garnet and plagioclase in mafic rocks. The upper boundary is at 10.3 kb at 750° C and 18.0 kb at 1,150° C. The eclogite facies is the highest pressure mineral facies, in which jadeite-rich clinopyroxene is stable.Compositions of minerals in natural rocks of the granulite facies and the eclogite facies are considered. Clinopyroxenes in the granulite-facies rocks have smaller jadeite-Tschermak's molecule ratios and higher amounts of Tschermak's molecule than clinopyroxenes in the eclogite-facies rocks. The distribution coefficients of Mg between orthopyroxene and clinopyroxene are normally in the range of 0.5–0.6 in metamorphic rocks in the granulite facies. The distribution coefficients of Mg between garnet and clinopyroxene suggest increasing crystallization temperature of the rocks in the following order: eclogite in glaucophane schist, eclogite and granulite in gneissic terrain, garnet peridotite, and peridotite nodules in kimberlite.Temperatures near the bottom of the crust in orogenic zones characterized by kyanitesillimanite metamorpbism are estimated from the mineral assemblages of metamorphic rocks in Precambrian shields to be about 700° C at 7 kb and 800° C at 9 kb, although heat-flow data suggest that the bottom of Precambrian shield areas is about 400° C and the eclogite facies is stable.The composition of liquid which is in equilibrium with peridotite is estimated to be close to tholeiite basalt at the surface pressure and to be picrite at around 30 kb. The liquid composition becomes poorer in normative olivine with decreasing pressure and temperature.During crystallization at high pressure, olivine and orthopyroxene react with liquid to form clinopyroxene, and a discontinuous reaction series, olivine orthopyroxene clinopyroxene is suggested. By fractional crystallization of pyroxenes the liquid will become poorer in SiO₂. Therefore, if liquid formed by partial melting of peridotite in the mantle slowly rises maintaining equilibrium with the surrounding peridotite, the liquid will become poorer in MgO by crystallization of olivine, and tholeiite basalt magma will arrive at the surface. On the other hand, if the liquid undergoes fractional crystallization in the mantle, the liquid may change in composition to alkali-basalt magma and alkali-basalt volcanism may be seen at a late stage of volcanic activity.Publication No. 681, Institute of Geophysics and Planetary Physics, University of California, Los Angeles.     

Volcanic Rocks from the Nejapa and Granada Cinder Cone Alignments, Nicaragua, Central America

Mafic tholeiitic basalts from the Nejapa and Granada (NG) cindercone alignments provide new insights into the origin and evolutionof magmas at convergent plate margins. In comparison to otherbasalts from the Central American volcanic front, these marietholeiitic basalts are high in MgO and CaO and low in Al₂O_p,K₂O₁, Ba and Sr. They also differ from other Central Americanbasalts, in having clinopyroxene phenocrysts with higher MgO,CaO and Cr₂O₃ concentrations and olivine phenocrysts with higherMgO contents. Except for significantly higher concentrationsof Ba, Sr and ⁸⁷Sr/₈₆Sr, most of the tholeiites are indistinguishable in compositionfrom mid-ocean ridge basalts. In general, phenocryst mineralcompositions are also very similar between NG tholeiites andmid-ocean ridge basalts. The basalts as a whole can be dividedinto two groups based on relative TiO₂-K₂O concentrations. Thehigh-Ti basalts always have the lowest K₂O and Ba and usuallyhave the highest Ni and Cr. All of the basalts have experienced some fractional crystallizationof olivine, plagioclase and clinopyroxene. Relative to otherCentral American basalts, the Nejapa-Granada basalts appearto have fractionated at low P_T and P_H2O. The source of primarymagmas for these basalts is the mantle wedge. Fluids and/ormelts may have been added to the mantle wedge from hydrothermally-altered,subducting oceanic crust in order to enrich the mantle in Sr,Ba and ⁸⁷Sr/⁸⁶Sr, but not in K and Rb. The role of lower crustaicontamination in causing the observed enrichments in Sr, Baand ⁸⁷Sr/⁸⁶Sr of NG basalts in comparison to mid-ocean ridgebasalts, however, is unclear. Rutile or a similar high-Ti accessoryphase may have been stable in the mantle source of the low-TiNG basalts, but not in that of the high-Ti basalts. Mafic tholeiiticbasalts, similar to those from Nejapa and Granada, may representmagmatic compositions parental to high-Al basalts, the mostmafic basalts at most Central American volcanoes. The characterof the residual high-Al basalts after this fractionation stepdepends critically on P_H2O Both high and low-Ti andesites are also present at Nejapa. Likethe high-Ti basalts, the high-Ti andesites have lower K₂O andBa and higher Ni and Cr in comparison to the low-Ti group. Thehigh-Ti andesites appear to be unrelated to any of the otherrocks and their exact origin is unknown. The low-Ti andesitesare the products of fractional crystallization of plagioclase,clinopyroxene, olivine (or orthopyroxene) and magnetite fromthe low-Ti basalts. The eruption that deposited a lapilli sectionat Cuesta del Plomo involved the explosive mixing of 3 components:high-Ti basaltic magma, low-Ti andesitic magma and high-Ti andesiticlava.     

Evolution and Genesis of Magmas from Vico Volcano, Central Italy: Multiple Differentiation Pathways and Variable Parental Magmas

Vico volcano has erupted potassic and ultrapotassic magmas,ranging from silica-saturated to silica-undersaturated types,in three distinct volcanic periods over the past 0·5Myr. During Period I magma compositions changed from latiteto trachyte and rhyolite, with minor phono-tephrite; duringPeriods II and III the erupted magmas were primarly phono-tephriteto tephri-phonolite and phonolite; however, magmatic episodesinvolving leucite-free eruptives with latitic, trachytic andolivine latitic compositions also occurred. In Period II, leucite-bearingmagmas (⁸⁷Sr/⁸⁶Sr_initial = 0·71037–0·71115)were derived from a primitive tephrite parental magma. Modellingof phonolites with different modal plagioclase and Sr contentsindicates that low-Sr phonolitic lavas differentiated from tephri-phonoliteby fractional crystallization of 7% olivine + 27% clinopyroxene+ 54% plagioclase + 10% Fe–Ti oxides + 4% apatite at lowpressure, whereas high-Sr phonolitic lavas were generated byfractional crystallization at higher pressure. More differentiatedphonolites were generated from the parental magma of the high-Srphonolitic tephra by fractional crystallization of 10–29%clinopyroxene + 12–15% plagioclase + 44–67% sanidine+ 2–4% phlogopite + 1–3% apatite + 7–10% Fe–Tioxides. In contrast, leucite-bearing rocks of Period III (⁸⁷Sr/⁸⁶Sr_initial= 0·70812–0·70948) were derived from a potassictrachybasalt by assimilation–fractional crystallizationwith 20–40% of solid removed and r = 0·4–0·5(where r is assimilation rate/crystallization rate) at differentpressures. Silica-saturated magmas of Period II (⁸⁷Sr/⁸⁶Sr_initial= 0·71044–0·71052) appear to have been generatedfrom an olivine latite similar to some of the youngest eruptedproducts. A primitive tephrite, a potassic trachybasalt andan olivine latite are inferred to be the parental magmas atVico. These magmas were generated by partial melting of a veinedlithospheric mantle sources with different vein–peridotite/wall-rockproportions, amount of residual apatite and distinct isolationtimes for the veins. KEY WORDS: isotope and trace element geochemistry; polybaric differentiation; veined mantle; potassic and ultrapotassic rocks; Vico volcano; central Italy     

The petrology and geochemistry of the Azores Islands

Forty lavas from the Azores Islands have been analyzed for ⁸⁷Sr/⁸⁶Sr ratios, major elements, first transition series metals, and LIL elements. The samples belong to the alkali basalt magma series but range from transitional hy-normative basalts from Terceira to basanitoids from Santa Maria. Differentiated lavas include both typical trachytes and comenditic trachytes and comendites. Major and trace element concentrations define smooth trends on variation diagrams, and these trends can be related to phases crystallizing in the rocks. Systematic interisland differences are also apparent in these variation diagrams. LIL element concentrations in island basalts are roughly twice as high as those in tholeiites from the adjacent Mid-Atlantic Ridge which transects the Azores Plateau. ⁸⁷Sr/⁸⁶Sr ratios in lavas from 6 of the 9 islands range from 0.70332 to 0.70354, a range similar to that found in tholeiites from the Mid-Atlantic Ridge transect of the Azores Plateau. This suggests that lavas from these islands and this portion of the Mid-Atlantic Ridge may be derived from a similar source. However, lavas from the islands of Faial and Pico have ⁸⁷Sr/⁸⁶Sr ratios up to 0.70394 and ratios in Sao Miguel lavas range up to 0.70525, suggesting basalts from these islands are derived from a chemically distinct source. Differences in the average LIL element concentrations of the least fractionated ridge tholeiites from the Azores Plateau and alkali basalts from the islands result from differences in extent of partial melting and residual mineralogy. The alkali basalts are derived by roughly half as much melting as are the tholeiites. Trace element concentrations in Azores peralkaline lavas preclude their derivation by partial melting of peridotitic mantle or basaltic crust; rather the data suggest they are produced by fractional crystallization of a basaltic parent.     

Integrated Models of Basalt Petrogenesis: A Study of Quartz Tholeiites to Olivine Melilitites from South Eastern Australia Utilizing Geochemical and Experimental Petrological Data

The Tertiary to Recent basalts of Victoria and Tasmania havemineralogical and major element characteristics of magmas encompassingthe range from quartz tholeiites to olivine melilitites. Abundancesof trace elements such as incompatible elements, including therare earth elements (REE), and the compatible elements Ni, Coand Sc, vary systematically through this compositional spectrum.On the basis of included mantle xenoliths, appropriate 100 Mg/Mg+ Fe₊₂ (68–72) and high Ni contents many of these basaltsrepresent primary magmas (i.e., unmodified partial melts ofmantle peridotite). For fractionated basalts we have derivedmodel primary magma compositions by estimating the compositionalchanges caused by fractional crystallization of olivine andpyroxene at low or moderate pressure. A pyrolite model mantlecomposition has been used to establish and evaluate partialmelting models for these primary magmas. By definition and experimentaltesting the specific pyrolite composition yields parental olivinetholeiite magma similar to that of KilaeauIki, Hawaii (1959–60)and residual harzburgite by 33 per cent melting. It is shownthat a source pyrolite composition differing only in having0.3–0.4 per cent TiO₂ rather than 0.7 per cent TiO₂, isable to yield the spectrum of primary basalts for the Victorian-Tasmanianprovince by 4 per cent to 25 per cent partial melting. The mineralogiesof residual peridotites are consistent with known liquidus phaserelationships of the primary magmas at high pressures and thechemical compositions of residual peridotite are similar tonatural depleted or refractory lherzolites and harzburgites.For low degrees of melting the nature of the liquid and of theresidual peridotite are sensitively dependent on the contentof H₂O, CO₂ and the CO₂/H₂O in the source pyrolite. The melting models have been tested for their ability to accountfor the minor and trace element, particularly the distinctivelyfractionated REE, contents of the primary magmas. A single sourcepyrolite composition can yield the observed minor and traceelement abundances (within at most a factor of 2 and commonlymuch closer) for olivine melilitite (4–6 per cent melt),olivine nephelinite, basanite (5–7 per cent melt), alkaliolivine basalt (11–15 per cent melt), olivine basalt andolivine tholeiite (20–25 per cent melt) provided thatthe source pyrolite was already enriched in strongly incompatibleelements (Ba, Sr, Th, U, LREE) at 6–9 x chondritic abundancesand less enriched (2.5–3 x chondrites) in moderately incompatible(Ti, Zr, Hf, Y, HREE) prior to the partial melting event. Thesources regions for S.E. Australian basalts are similar to thosefor oceanic island basalts (Hawaii, Comores, Iceland, Azores)or for continental and rift-valley basaltic provinces and verydifferent in trace element abundances from the model sourceregions for most mid-ocean ridge basalts. We infer that thismantle heterogeneity has resulted from migration within theupper mantle (LVZ or below the LVZ) of a melt or fluid (H₂O,CO₂-enriched) with incompatible element concentrations similarto those of olivine melilitite, kimberlite or carbonatite. Asa result of this migration, some mantle regions are enrichedin incompatible elements and other areas are depleted. Although it is possible, within the general framework of a lherzolitesource composition, to derive the basanites, olivine nephelinitesand olivine melilitites from a source rock with chondritic relativeREE abundances at 2–5 x chondritic levels, these modelsrequire extremely small degrees of melting (0.4 per cent forolivine melilitite to 1 per cent for basanite). Furthermore,it is not possible to derive the olivine tholeiite magmas fromsource regions with chondritic relative REE abundances withoutconflicting with major element and experimental petrology argumentsrequiring high degrees (15 per cent) of melting and the absenceof residual garnet. If these arguments are disregarded, andpartial melting models are constrained to source regions withchondritic relative REE abundances, then magmas from olivinemelilitites to olivine tholeiites can be modelled if degreesof melting are sufficiently small, e.g., 7 per cent meltingfor olivine tholeiite. However, the source regions must be heterogenousfrom 1 to 5 x chondritic in absolute REE abundances and heterogerieousin other trace elements as well. This model is rejected in favorof the model requiring variation in degree of melting from 4per cent to 25 per cent and mantle source regions ranging fromLREE-enriched to LREE-depleted relative to chondritic REE abundances.     

Geochemistry and petrogenesis of the Proterozoic dykes in Tamil nadu, southern India: another example of the Archaean lithospheric mantle source

Approximately 1650-Ma-old NW/SE and NE/SW-trending dolerite dykes in the Tiruvannamalai (TNM) area and approximately 1800-Ma-old NW/SE-trending dolerite dykes in the Dharmapuri (DP) area constitute major Proterozoic dyke swarms in the high-grade granulite region of Tamil nadu, southern India. The NW- and NE-trending TNM dykes are compositionally very similar and can be regarded as having been formed during a single magmatic episode. The DP dykes may relate to an earlier similar magmatic episode. The dolerites are Fe-rich tholeiites and most of the elemental variations can be explained in terms of fractional crystallisation. Clinopyroxene and olivine are the inferred ferromagnesian fractionation phases followed by plagioclase during the late fractionation stages. All the studied dykes have, similar to many continental flood basalts (CFB), large-ion lithophile element (LILE) and light rare-earth element (LREE) enrichment and Nb and Ta depletion. The incompatible element abundance patterns are comparable to the patterns of many other Proterozoic dykes in India and Antarctica, to the late Archaean (～2.72 Ga) Dominion volcanics in South Africa and to the early Proterozoic (～2.0 Ga) Scourie dykes of Scotland. The geochemical characteristics of the TNM and DP dykes cannot be explained by crustal contamination alone. Instead, they are consistent with derivation from an enriched lithospheric mantle source which appears to have been developed much earlier than the dyke intrusions during a major crustal building event in the Archaean. The dyke magmas may have been formed by dehydration melting induced by decompression and lithospheric attenuation or plume impingement at the base of the lithosphere. These magmas, compared with CFB, appear to be the minor partial melts from plume heads of smaller diameter and of shallow origin (650 km). Therefore, the Proterozoic thermal events could induce crustal attenuation and dyke intrusions in contrast to the extensive CFB volcanism and continental rifting generally associated with the Phanerozoic plumes of larger head diameter (>1000 km) and of deeper origin (at crust mantle boundary).     

Polybaric Evolution of Calc-alkaline Magmas from Nisyros, Southeastern Hellenic Arc, Greece

The lavas of Nisyros were erupted between about 0?2 m.y B.P.and 1422 A.D., and range in composition from basaltic andesiteto rhyodacite. Most were erupted prior to caldera collapse (exactdate unknown), and the post-caldera lavas are petrographically(presence of strongly resorbed phenocrysts) and chemically (lowerTiO₂ K₂O, P₂O₅, and LIL elements) distinct from the pre-calderalavas. The pre-caldera lavas do not form a continuous seriessince lavas with SiO₂ contents between 60 and 66 wt.% are absent.Nevertheless, major element variations demonstrate that fractionalcrystalliz ation (involving removal of olivine, dinopyroxene,plagioclase, and Fe-Ti oxide from the basaltic andesites andandesites and plagioclase, clinopyroxene, hypersthene, Ti-magnetite,ilmenite, apatite, and zircon from the dacites and rhyodacites)played a major role in the evolution of the pre-caldera lavas.Several lines of evidence indicate that other processes werealso important in magma evolution: (1) Quantitative modelingof major element data shows that phenocryst phases of unlikelycomposi tion or unrealistic assemblages of phenocryst phasesare required to relate the dacites and rhyodacites to the basalticandesites and andesites; (2) The proportions of olivine andclinopyroxene required in quantitative models for the initialstages of evolution differ from those observed petrographicallyand this is not likely to reflect either differential ratesof crystal settling or the curvature of cotectics along whichliquids of basaltic andesite to andesite composition lie; (3)The concentrations of Rb, Cs, Ba, La, Sm, Eu, and Th in therhyod.acites are too high for these lavas to be related to thedacites by fractional crystallization alone; and (4) ⁸⁷Sr/⁸⁶Srratios for the andesites and rhyodacites are higher than thosefor the basaltic andesites and dacites, respectively. It isshown that fractional crystallization was accompanied by assimilation,and that magma mixing played a minor role (if any) in the evolutionof the pre-caldera lavas. Trace element and isotopic data indicatethat the andesites evolved from the basaltic andesites by AFCinvolving average crust or upper crust, whereas the rhyodacitesevolved from the dacites by AFC involving lower crust. Additionalevidence for polybaric evolution is provided by the occurrenceof distinct Ab-rich cores of plagioclase phenocrysts in thedacites and rhyodacites, which record a period of high pressurecrystallization, and by the occurrence of both normal and reverse-zonedphenocrysts in the basaltic andesites and andesites. Furthermore,calculated pressures of crystallization are {small tilde}8 kbfor the dacites and rhyodacites and 3?5–4 kb for the basalticandesites and andesites. It is concluded that the dacites andrhyodacites evolved via AFC from basaltic andesites and andesiteslargely in chambers sited near the base of the crust whereasthe basaltic andesites and andesites mostly evolved in chamberssited at mid-crustal levels. Eruption from different chambersexplains the compositional gap in the chemistry of the pre-calderalavas since eruptive products represent a more or less randomsampling of residual liquids which separate (via filter pressing)from bodies of crystallizing magma at various depths. Magmamixing was important in the evolution of the post-caldera lavas,but geochemical data require that these magmas evolved fromparental magmas which were derived from a more refractory sourcethan the parental magmas to the pre-caldera lavas. ^*Present address: Netherlands Energy Research Foundation (ECN), P.O. Box 1, 1755 ZG Petten, The Netherlands     

Magma fractionation systems beneath the mid-Atlantic ridge at 36–37 ° N

Variation of major and trace elements in drilled basalts from the Mid-Atlantic Ridge (DSDP Leg 37) reflects distinct cycles of low pressure fractionation operating independently within a complex network of magma storage reservoirs beneath the crustal spreading axis. Low pressure phase relations are determined by parental magma composition, which varies from An-rich (An/Di > ca. 1.4) to Di-rich (An/Di < ca. 1.4). High An/Di magmas probably formed under slightly hydrous conditions in the mantle. They have low LIL element contents, low P/Y and high Mg/(Mg + Fe) ratios. Zr, P and Y abundance and inter-element ratios are highly diagnostic of primary magma type, and are used to quantify fractional crystallization models.Low pressure fractionation hypotheses were tested by least-squares modelling of whole-rock and phenocryst chemistry, which indicated removal or addition of phenocryst assemblages: ol; pl; ol + pl; ol + pl + cpx; pl + cpx, (± sp). Accumulation of plagioclase or olivine is an important mechanism for generating highly porphyritic rocks. A rare 3-phase (ol + pl + cpx) cumulate resulted from cotectic fractionation of a low An/Di magma type. Olivine and plagioclase cumulates appear to be related to high An/Di magmas. Olivine accumulation has been monitored by comparison of olivine/bulk rock partitioning of Fe and Mg to experimental measurements of the equilibrium K_D value. A single extensive sub-axial magma chamber could not account for the observed chemical variation and would probably be dynamically unstable.