Potassium variation across the New Britain volcanic arc

Late Cainozoic volcanoes of the New Britain island arc overlie an inclined Benioff zone that extends to a depth of at least 580 km. The rocks are tholeiitic basalt, andesite, dacite, and rhyolite. Unlike many other examples of island arcs described in the literature, K₂O contents in rocks with the same SiO₂ content do not increase progressively as depth,h, to the New Britain Benioff zone increases. The most complex relationships between K₂O, SiO₂, andh are shown by volcanoes overlying the deeper part of the Benioff zone. In these, the K₂O contents of rocks containing more than about 60% SiO₂,decrease as depth to the Benioff zone increases. The New Britain volcanic arc provides a striking exception to the generalisation thatK-h relationships are essentially similar in all island arcs.     

Alkali mapping of the japanese quaternary volcanic rocks

Of ca. 200 Quaternary volcanic centers in the Japanese islands, ca. 120 are characterized by K₂O and Na₂O concentrations of their ejecta normalized to SiO₂ = 60% K₂O values give a geographically well-defined trend of regularly increasing away from the volcanic front. Na₂O values show no geographical trend. Relationships between the SiO₂-normalized K₂O and the depth of the deep seismic zone (K-h relation) are different for different arcs. The Southwest Japan arc shows distinctly higher K₂O and less regular geographic trend than the others probably due to crustal contamination. In the normal segments of the Northeastern Japan belt, K₂O increases ca. 2.5 to 3.0% per 100 km depth of the seismic zone. Groups of volcanic centers near the junction between the Northeast Honshu and Izu-Mariana arcs give characteristically lower K₂O while Rishiri volcano far behind the arc junction between the Kurile and Honshu arcs also gives a low K₂O value.     

The magmatic evolution of the Seiland province,and Caledonian plate boundaries in northern Norway

The synorogenic intrusive activity of the eastern part of the Seiland province evolved from tholeiitic basalt low in K and Ti, through high-K calc-alkaline magmas and possible transitional basalt, to alkaline olivine basalt and picrite; finally, highly differentiated alkaline magmas and carbonatites were emplaced. For intrusive rocks with equivalent SiO₂ contents, K₂O, K₂O + Na₂O and K₂O/Na₂O increased with time, and the degree of iron enrichment in basaltic suites dimished. The western part of the province shows no equivalent evolution, tholeiitic magmas being emplaced at the same time as calc-alkaline magmas to the east.The magmatism is believed to have stemmed from a diapiric complex established in the mantle above a Benioff zone dipping to the east beneath the deforming Andean-type margin of the Baltic plate. Tectonic shortening of the continental edge and rearward movement of the underthrust plate relative to the asthenosphere resulted in migration of the plate junction, steepening of the seismic zone, and increasing depth to the magmagenetic region.     

The Clarno Formation of central Oregon,U.S.A. — Volcanism on a thin continental margin

The Clarno Formation (mostly Eocene) of central Oregon, U.S.A., was formed as North America moved westward over subducting Pacific Ocean crust. The Clarno is a volcanic and volcanogenic assemblage whose flow rocks show: a calc-alkaline pattern on a Harker diagram, K₂O-SiO₂ diagram, alkali-SiO₂ diagram, and AFM diagram; and a pattern transitional between calc-alkaline and tholeiitic on a SiO₂-FeO^*/MgO diagram. Its basalts are chemically similar to those of intra-oceanic island arcs (e.g., K₂O of 0.30%), but subaerial deposition of the entire formation plus differentiation to rocks of high SiO₂ and alkali contents indicate that the Clarno was formed on a continental margin. Comparison of the Clarno with other Pacific-margin volcanic suites indicates that the Clarno was formed on thin (20–30 km) continental crust overlying a subduction zone of about 120 km depth.     

Temporal variation in the lavas of Mayon volcano, Philippines

Chemical and petrographic analyses of 51 sequential lava flows from the central vent of Mayon volcano show cyclical variation. In the two most recent cycles, from 1800 to 1876 and from 1881 to the present, one to three basaltic flows are followed by six to ten andesitic flows. Modal and whole-rock chemical parameters show the most regular cyclical variation; calculated groundmass chemical parameters vary less regularly. There is also a long-term trend, over approximately 1700 years of exposed section, toward more basic compositions.The cyclical variation in modes and the chemical composition of the lavas apparently results from periodic influxes of basaltic magma from depth into a shallow magma system. Fractional crystallization of olivine, augite, hypersthene, calcic plagioclase, magnetite and pargasitic hornblende produces successively more andesitic lavas until the next influx of basaltic magma. Differentiation in a deep zone of magma generation is not excluded by the data, but is more likely responsible for the overall change toward more basic compositions than for the cyclical variation.Three points in a cycle — the beginning of basaltic lavas, the beginning of andesitic lavas and a leveling-off of SiO₂, K₂ O and K₂O/Na₂O values — correspond roughly to the beginning of frequent effusive eruptions (with or without an early Plinian eruption), frequent weak to moderately explosive (Strombolian) eruptions, and less frequent explosive (Vulcanian) eruptions, respectively. Recognition of the current stage in a cycle can give a qualitative indication of the nature of forthcoming eruptions. Changes in several specific parameters may precede basaltic lavas and allow early detection of basaltic influxes. These include minima in the glass inclusion/plagioclase phenocryst and phenocryst/groundmass ratios, vesicularity and groundmass TiO₂, a decrease in hypersthene phenocrysts, and constant values for the whole-rock K₂O/Na₂O ratio. The Mayon area is densely populated, making prediction of eruption type important for safety and land-use planning.     

Calc-alkaline and shoshonitic lavas from five Andean volcanoes (between latitudes 21° 45′ and 24°30′S) and the distribution of the Plio-Quaternary volcanism of the south-central and southern Andes

The Plio-Quaternary volcanic rocks of the south-central Andes (southward from latitude 18°S) contain two associations: calc-alkaline and shoshonitic which coincide with seismic belts as geographically distinct zones aligned parallel to the oceanic trench. There is a continuous gradation from calc-alkaline to shoshonitic associations. The shoshonitic association appears to the north of latitude 26°S; southwards, the calc-alkaline association directly abuts against the continental (Argentinian) alkaline association.Thirty-one lavas from the Plio-Quaternary calc-alkaline Socompa, Lascar, Sairecabur and Tocorpuri and shoshonitic Sierra de Lipez volcanoes were studied. The lavas are porphyric with abundant glass. The distribution and the nature of the phenocrysts vary according to the chemistry of the calc-alkaline lavas. Petrographic evidence for crystal fractionation has been observed. Occasional phenocrysts of alkali feldspars occur in the shoshonitic lavas. The K₂O and SiO₂ contents increase from calc-alkaline to shoshonitic lavas with distance away from the oceanic trench. In lavas from Socompa, Lascar, Sairecabur and Tocorpuri calc-alkaline volcanoes, K₂O, Li and Rb increase and K/Rb and Sr decrease with increasing SiO₂; Ba increases with decreasing Sr, probably as a result of plagioclase fractionation. In lavas from Sierra de Lípez shoshonitic volcano, SiO₂ is high, K₂O is high and rather constant and Li, Rb, Ba and Sr increase with increasing SiO₂. Bolivian shoshonitic lavas appear to be genetically related to the calc-alkaline suite.The calc-alkaline lavas may be derived by crystal fractionation from a parental magma of andesitic nature that originated in or above the subjacent Benioff zone.     

K‐ and Si‐rich glasses in harzburgite from Damaping,north China

A study of potassium‐ and silica‐rich glass (SiO₂ = 65.3–67.4%, K₂O = 7.1–9.8%, Na₂O = 4.4–6.5%) in spongy clinopyroxene rims from anhydrous spinel harzburgite, collected from Damaping (Hannuoba, north China), is reported here. The corroded surface of clinopyroxene along with clear chemical zonation (homogeneous core and partially melted rim) suggests that incongruent melting of primary clinopyroxene is responsible for generating the Si‐rich glass in clinopyroxene rims. The degree of clinopyroxene melting is estimated to be higher than 15%. In order to generate glass with K₂O contents of 7.0 to 9.8% by clinopyroxene melting at a degree of 15% in a closed system, K₂O contents in the primary clinopyroxene should be greater than 1.0 wt%, suggesting a very deep origin for the Damaping harzburgite.     

Si- and alkali-rich melt inclusions in minerals of mantle peridotites from eastern China: Implication for lithospheric evolution

CO2 fluid inclusions in mantle minerals are an im-portant source for us to get the information of mantle fluids. Fluid inclusions are mainly composed of CO2, with minor CO, H2O, CH4, N2, H2S, SO2, F, etc., which were demonstrated by lots of Raman spec-trometer analyses in recent years. In contrast, there are very few researches on CO2-bearing melt inclusions since it is more difficult to do so. The available studies have found that the primary CO2-bearing melt inclu-sions are basaltic …     

Oxygen isotope fractionation in mantle minerals

The increment method is adopted to calculate oxygen isotope fractionation factors for mantle minerals, particularly for the polymorphic phases of MgSiO₃ and Mg₂SiO₄. The results predict the following sequence of¹⁸O-enrichment:pyroxene (Mg, Fe, Ca)₂Si₂O₆>olivine (Mg, Fe)₂SiO₄ > spinel (Mg, Fe)₂SiO₄> ilmenite (Mg, Fe, Ca) SiO₃>perovskite (Mg, Fe, Ca) SiO₃. The calculated fractionations for the calcite-perovskite (CaTiO₃) System are in excellent agreement with the experimental calibrations. If there would be complete isotopic equilibration in the mantle, the spinel-structured silicates in the transition zone are predicted to be enriched in¹⁸O relative to the perovskite-structured silicates in the lower mantle but depleted in¹⁸O relative to olivines and pyroxenes in the upper mantle. The oxygen isotope layering of the mantle might result from differences in the chemical composition and crystal structure of mineral phases at different mantle depths. Assuming isotopic equilibrium on a whole earth scale, the chemical structure of the Earth’s interior can be described by the following sequence of¹⁸O-enrichment:upper crust>lower crust>upper mantle>transition zone>lower mantle>core. Project supported by the National Natural Science Foundation of China and the Chinese Academy of Sciences.     

Petrography and geochemistry of upper Triassic sandstones from the Tumengela Formation in the Woruo Mountain area,North Qiangtang Basin,Tibet: Implications for provenance,source area weathering,and tectonic setting

The petrography and major and trace element concentrations of the sandstones from the Tumengela Formation in the Woruo Mountain area, North Qiangtang Basin, are studied to determine their provenance, intensity of weathering and tectonic setting. The detrital compositions of the Tumengela sandstone samples are dominated by quartz (58.0–70.1 %, average 64.7 %) and lithic fragments (21.8–35.9 %, average 27.3 %), but low in feldspar content (4.9–12.9 %, average 8.0 %). The sandstones can be classified as litharenite and feldspathic litharenite according to their detrital compositions, which is consistent with the geochemical data. The detrital modal compositions reflect that these sandstones are probably derived from a recycled orogenic source. The index of chemical variability (ICV) and SiO₂/Al₂O₃ ratio values suggest that the compositional maturity and recycling were moderate. The weathering indices such as the chemical index of alteration (CIA), plagioclase index of alteration (PIA), chemical index of weathering (CIW), and Al₂O₃–(CaO* + Na₂O)–K₂O (A–CN–K) diagram indicate that the intensities of weathering in the source area were moderate. The Al₂O₃/TiO₂, Th/Co, La/Sc, La/Co, Th/Sc, Cr/Th ratio values and the discriminant function of the Tumengela sandstones indicate that the sediments were mainly derived from felsic source rocks, while also mixed with intermediate source rocks. The comparison of rare earth element patterns and its Eu anomalies to the probable source rocks infer that the sandstones were derived from the combination of granite, rhyolite, dacite, and gneisses. The proximal central uplift belt was probably the primary provenance area as evidenced by the petrographical and geochemical features of the Tumengela sandstones. The multidimensional tectonic discrimination diagram based on major elements show a collision setting (80 %) combined with a rift setting (20 %) for the Tumengela sandstones, which is consistent with the general geology of the study areas.     

A Statistical Framework for Calculating and Assessing Compositional Linear Trends Within Fault Zones: A Case Study of the NE Block of the Clark Segment,San Jacinto Fault,California, USA

Utilizing chemical data derived from the various fault zone architectural components of the Clark strand of the San Jacinto fault, southern California, USA, we apply for the first time non-central principal component analysis to calculate a compositional linear trend within molar A–CN–K space. In this procedure A–CN–K are calculated as the molar proportions of Al₂O₃ (A), CaO* + Na₂O (CN), and K₂O (K) in the sum of molar Al₂O₃, Na₂O, CaO*, and K₂O. CaO* is the molar CaO after correction for apatite. We then derive translational invariant chemical alteration intensity factors, t, for each architectural component through orthogonal projection of analyzed samples onto the compositional linear trend. The chemical alteration intensity factor t determines the relative change in composition compared to the original state (i.e., the composition of the altered wall rocks). It is dependent on the degree of intensity to which the process or processes responsible for the change in composition of each architectural component has been active. These processes include shearing, fragmentation, fluid flow, and generation of frictional heat. Non-central principal component analysis indicates that principal component 1 explains 99.7 % of the spread of A–CN–K data about the calculated compositional linear trend (i.e., the variance). The significance level for the overall one-way analysis of variance (ANOVA) is 0.0001. Such a result indicates that at least one significant difference across the group of means of t values is different at the 95 % confidence level. Following completion of the overall one-way ANOVA, the difference in means t test indicated that the mean of the t values for the fault core are different than the means obtained from the transition and damage zones. In contrast, at the 95 % confidence level, the means of the t values for the transition and damage zones are not statistically distinguishable. The results of XRD work completed during this study revealed that the <2 µm fraction is composed primarily of illite/smectite with ~15 % illite in the damage zone, of illite/smectite with ~30 % illite in the transition zone, and of discreet illite with very minor smectite in the fault core. These changes parallel the increasing values of the chemical alteration intensity factors (i.e., t). Based on the above results, it is speculated that when fault zones are derived from tonalitic wall rocks at depths of ~400 ± 100 m, the onset of the illite/smectite to illite conversion will occur when t values exceed 0.20 ± 0.12, the average chemical alteration intensity factor calculated for the transition zone. Under such conditions during repeated rupturing events, frictional heat is produced and acidic fluids with elevated temperatures (≥ ~125 °C) are flushed through the fault core. Over time, the combination of shearing, fragmentation, and frictionally elevated temperatures eventually overcomes the kinetic barrier for the illite/smectite to illite transition. Such settings and processes are unique to fault zones, and as a result, they represent an underappreciated setting for the development of illite from illite/smectite. The success of non-central principal component analysis in this environment offers the first statistically rigorous methodology for establishing the existence of compositional linear trends in fault zones. This method also derives quantifiable alteration intensity factors that could potentially be used to compare the intensity of alteration at different segments of a fault, as well as offer a foundation to interpret the potential driving forces for said alteration and differences therein.     

Geochemistry of eclogite‐ and blueschist‐facies rocks from the Bantimala Complex,South Sulawesi,Indonesia: Protolith origin and tectonic setting

We present the first data on bulk‐rock major and trace element compositions for a suite of eclogite‐ and blueschist‐facies rocks from the Bantimala Complex, Indonesia, with the aim of better constraining the protolith origins and nature of the subducted crust. The eclogites can be classified into two groups: glaucophane‐rich eclogite and glaucophane‐free eclogite, whereas the blueschists are divided into albite–epidote glaucophanite and quartz–glaucophane schists. SiO₂ contents of the eclogites are 43.3–49.6 wt%, with Na₂O + K₂O contents 3.7–4.7 wt%. The blueschists show a wider range of compositions, with SiO₂ = 40.7–63.8 wt% and Na₂O + K₂O = 2.7–4.5 wt%. Trace element data suggest that the eclogite protoliths include both enriched and normal mid‐oceanic ridge basalt (E‐MORB and N‐MORB) and also gabbroic cumulates. The blueschists show more variation in protoliths, which include N‐MORB, Oceanic Island Basalt (OIB) and Island Arc Basalt (IAB). Plots of element concentrations against the immobile Zr show considerable mobility of large ion lithophiles but not of high field‐strength elements during high‐pressure metamorphism, and indicate that the high SiO₂ content of some blueschists is probably due to metasomatism by a LILE‐rich siliceous aqueous fluid. Strong correlations between K, Rb, Ba and Cs suggests that enrichment of these elements occurred by a single process. All the protoliths were subducted, metamorphosed to blueschist/eclogite‐facies and subsequently exhumed. It is noteworthy that the samples deduced to have come from thicker‐crust environments (OIB, IAB) were subducted to shallower depths (blueschist‐facies) than MORB‐derived samples, all except one of which reached eclogite‐facies conditions. The geochemical data of this study demonstrate the variety of ocean floor types that were subducted under the southeast margin of Sundaland in the late Jurassic period.     

On the lateral variations in chemical composition and volume of quaternary volcanic rocks across Japanese arcs

The Fe/Mg+Fe) ratios (X_Fe) of the Quaternary basalts (SiO₂ < 53 wt.%) in the Japanese arcs were examined. The X_XFe of relatively magnesian basalts decreases from the volcanic front toward the Japan Sea across the arcs. Based on the partition coefficient of Mg-Fe²⁺ between olivine and liquid, it is suggested that all the basalts near the volcanic front, which are mostly tholeiitic basalts, are significantly fractionated, whereas many basalts near the Japan Sea, which are mostly alkali basalts, are little fractionated. The K₂ O content in the primary basalt magmas increases toward the Japan Sea. Combining the X_Fe and K₂ O data, it is suggested that relatively large amounts of tholeiitic magmas are produced near the volcanic front, but they fractionate during their ascent, whereas smaller amounts of alkali basalt magmas are formed near the Japan Sea, but they can ascend with less fractionation. The density of primary tholeiite magma is significantly larger than that of primary alkali basalt magmas. It is most likely that primary tholeiite magmas cannot ascend beyond the upper crust and would fractionate to produce less dense tholeiitic magmas near the volcanic front, whereas primary alkali basalt magmas can ascend through the upper crust without fractionation, as far as buoyancy is the principal ascending force. In the Japanese arcs, the stress field may be less compressional near the Japan Sea than near the volcanic front, so that magmas can ascend more rapidly in the latter region than in the former. These two factors may be responsible for the above mentioned chemical variations of basalt magmas across the arcs. The variation in volume of the Quaternary volcanic rocks across the arcs can be explained by the presence of a melt-rich zone above but nearly parallel to the subducted slab.     

Discovery of double-peaking potassic volcanic rocks in Langshan Group of the Tanyaokou hydrothermal-sedimentary deposit,Inner Mongolia,and its indicating significance

Since the mid-1980s,Tanyaokou large Zn-Cu-Fe sulfides deposit,located at the southwest end of Langshan-Zhaertaishan-Bayan Obo Mesoproterozoic metallogenic belt in the west section of the northern margin of the North China Platform[1?9](Fig.1),has been confirmed to be submarine volcanic exhalative-sedimentary metamorphosed deposit hosted in the miogeosynclinal mud-carbonaceous formation of the Langshan Group(LG)[1],or submarine volcanic exha-lative-deposition-altered deposit[2]or stratabo…     

Magma types of volcanic rocks and continental margin of the Taiwan upper Paleozoic Geosyncline

Greenrocks are very common in the Tananao Schist of eastern Taiwan where the known fossils are of Permian in age. Fourty-four greenrock samples were chemically analysed and their magma types studied. The chemical composition of the greenrocks have marked variation common in volcanic rock series. The greater parts of the greenrocks belong to basalt and a smaller portion to basaltic andesite (SiO₂ 53 %–58 %). They are probably isochemical with their original igneous rocks except for volatile components. No striking Fe-enrichment exists in a MgO-ΣFeO-(Na₂O=K₂O) diagram. Based on (Na₂O=K₂O)-Al₂O₃-SiO₂ diagrams afterKuno (1960), the parent magma of the rocks mostly belong to the high-alumina basalt series and only a few to alkali olivine basalt series. The high-alumina basalt can be looked upon as having an incipient trend for the calc-alkaline or the hypersthene series ofKuno (1959). The average K/Rb ratio of 460, the average TiO₂ percentage of 1.5 %, and low K₂O of around 0.5 % seem to warrant a conclusion that the basaltic rocks were poured out in the upper Paleozoic eugeosyncline on an embryonic continental crust. Considering the rock association of amphibolite plus serpentine (dismembered ophiolite), meta-graywacke, metachert, crystalline limestone, metaarkose, and metabasites in the Tananao Schist, the most probable site for the eugeosyncline may been an extensional trough near the fragmented paleo-Asiatic margin.     

A worldwide comparison of alkali olivine basalts and their differentiation trends

Worldwide alkali olivine basalts (AOB) and their differentiation series have been subdivided into continental, oceanic, or island-arc assemblages according to the inferred crustal environment at their time and place of eruption. No systematic differences have been found in major element composition of the AOB's from these three different environments. As plotted on (Na₂O + K₂O) vs. SiO₂ and AMF diagrams, AOB differentiation trends also show no differences between environments. Thus, AOB appears to be a primary magma generated at sufficient depth in the mantle that its major element content is unaffected by chemical or thermal differences between mantle regions underlying continents, ocean basins, or island arcs. The major element chemistry of AOB is also apparently unaffected by passage through different types of crust.     

Petrology and geochemistry of clastic sedimentary rocks as evidences for provenance of the Late Palaeozoic Madzaringwe Formation,Tshipise-Pafuri Basin,South Africa

Petrography and geochemistry(major, trace and rare earth elements) of clastic rocks from the Late Palaeozoic Madzaringwe Formation, in the Tshipise-Pafuri Basin, Northern South Africa, have been investigated to understand their provenance. Sandstone petrography and detrital modes indicates that the Late Palaeozoic succession was derived from craton interior and recycled orogen provenance. Sandstones in the Madzaringwe Formation are sub-arkosic to sub-litharenite. The sediments may represent a recycled to craton interior provenance. The geochemical data of major elements show that sandstone and shales have the same source. The study of paleoweathering conditions based on modal composition, chemical index of alteration(CIA) and A-CN-K(Al2O3-Ca O+Na2O-K2O) relationships indicate that probably chemical weathering in the source area and recycling processes have been more important in shale and sandstone rocks. The relatively high CIA values(70–90%) indicates moderate to high weathering conditions of the samples and the paleoclimate of the source area was warm. K2O/Na2 O versus Si O2 and Na2O-Ca O-K2 O tectonic setting discrimination plots, suggest a passive continental margin. In the study of trace elements, triangular Th-Sc-Zr/10 and La-Th-Sc plots both suggest a passive margin setting of the basin. Petrographic and geochemical results of the samples suggest uplifted basement source areas dominated by sedimentary rocks and/or granite-gneiss rocks. The source rocks might have been the recycled pre-Soutpansberg Karoo Supergroup rocks and the metasedimentary rocks of the Soutpansberg Group. Other source rocks may have been the pre-Beit-Bridge basement rocks(granites and gneisses).     

Major-element petrochemistry of some extrusive rocks from the volcanically active Mariana Islands

Volcanic rocks from six of the currently or recently active volcances of the Mariana Island are show little variation in major element abundances. SiO₂ content averages 51.5 wt.%. The flows are high in Al₂O (mean 17.7 wt.%) and Fe oxides (mean 10.1 wt.% calculated as FeO only), and moderate in MgO content (mean 4.7 wt.%), Na₂O (mean 2.7 wt.%), and K₂O (mean 0.7 wt.%). Only the rocks from Farallon de Pajaros, the northernmost of the Mariana Islands, deviate slightly from the average of the analyses. Three analyses from this island are slightly higher in SiO₂ (about 54 wt.%) and Al₂O₃, and are lower in total Fe oxides and MgO. According to preferred classification, the lavas of the Mariana Islands can be termed mela-andesites, high-alumina basalts, or calc-alkaline (orogenic) basalts. The K₂O values (mean 0.7 wt.%) obtained from lavas of the Mariana Islands are significantly higher than the K₂O values (about 0.33 wt.%) from volcanics of the Izu chain to the north. Inasmuch as the substantial scatter in location of earthquake foci beneath both arcs prevents accurate delineation of the upper boundary of the Benioff zone, it presently cannot be determined whether this discrepancy in K₂O values reflects a difference in depth from the volcanic are to the dipping seismic zone or relates to other phenomena. The older volcanic islands within the Mariana-Bonin island chain apparently defined an island arc system during Eocene to Miocene time. This indicates that the present plane of convergence between the Pacific plate and the Philippine Sea plate has defined the convergence between these plates since Eocene time.     

Petrology and geochemistry of basaltic rocks of the Lau Basin

The Lau Basin is a marginal sea, located between the Tonga and Lau Ridges, in the southwestern Pacific. The basin is on the “inner” or concave side of the Tonga Trench-Arc system and is situated above the deep seismic zone dipping westward from the Tonga Trench. The Tonga Trench-Arc system is undoubtedly located above a zone of crustal shortening as evidenced by the deep seismicity and vulcanism. However, the geological and geophysical data give strong support to the contention that the Lau Basin has formed by crustal dilation.Rocks dredged from ridges and seamounts in the basin are sub-alkaline basalt. The average major element composition of least altered samples is: SiO₂ 48.8%, TiO₂ 1.2%, K₂O 0.18%, P₂O₅ 0.08%, H₂O+ 0.30%, Fe^III/Fe^II = 0.26,CaO/Al₂O₃ = 0.77. The data for Lau Basin basalt (LBB) show close similarity to data of typical oceanic ridge basalt (ORB). Trace element abundances (ppm): Ni 160, Cr 390, Sr 100, Ba < 31, Rb < 1 also resemble ORB values. K/Rb in a least altered and unfractionated sample is 860, Ba/Sr is 0.1, Ba/Rb is 8. Strontium isotope data show the only marked variance from ORB chemistry with LBB values ranging from ⁸⁷Sr/⁸⁶Sr=0.7020 to 0.7051. The low Sr abundances in the samples suggest the possibility of crustal Sr contamination to explain the radiogenic Sr enrichment. An alternate possibility is that the mantle source rocks were enriched in ⁸⁷Sr. Variation within dredge hauls and between dredge sites may be explained by low-pressure fractional crystallization of magmas separated from the mantle at about 50 km depth.The basin probably began to open in middle to late Miocene time either by the disruption of a single andesitic island arc by splitting along its axis or by dilation of the area between two closely spaced concentric arcs. Mantle counterflow in the asthenosphere above the downgoing oceanic lithosphere slab is the probable driving force for dilation and has provided a continuous supply of parent material for the basalt of the basin floor.     

Mineralogy and chemistry of modern orogenic lavas — some statistics and implications

Island arc and continental margin (i.e. western Americas) lavas are divided (based on raw data from literature) into basalts (defined by absence of Ca-poor pyroxene, dominated by quartz-normative tholeiites); basaltic andesites and andesites (subdivided on basis of breaks in SiO₂ histogram and taken as <56% and 56–63% SiO₂; Ca-poor pyroxene present; amphibole and biotite absent); and hornblende (±biotite) lavas, which prove to be mainly relatively silicic andesites. Relative proportions of these types are (576 samples): 23% basalts, 29% basaltic andesites; 30% andesites; 18% hornblende andesites. The compilation emphasizes the dominance of calcic plagioclase (labradorite-anorthite) amongst the phenocryst phases. Pyroxenes are largely augite and hypersthene (En_60–75); olivine (Fo_65–85) is common through all compositions. There is an overall close similarity in chemistry and mineralogy between continental margin and island arc lavas, although small consistent differences are apparent (e.g. K₂O, TiO₂, P₂O₅).Modal data indicate that 70% of lavas are phenocryst-rich (20–60 vol.%), and that phenocryst contents show a bimodal distribution. Statistically and petrologically significant correlations are found between mineralogy and rock chemistry, most notably between total rock Al₂O₃ and modal phenocrystic plagioclase (found in all data groups, except hornblende andesites). This, and related data and correlations, indicate that the majority of orogenic magmas are modified by crystal fractionation (including crystal accumulation) processes dominated by plagioclase, and interpreted to occur under relatively low pressures. Dominance of plagioclase suggests phenocryst precipitation occurs typically in water-undersaturated magmas.