Origin of differentiated lavas at Kilauea Volcano,Hawaii: Implications from the 1955 eruption

Kilauea's 1955 eruption was the first major eruption (longer than 2 days) on its east rift zone in 115 years. It lasted 88 days during which 108 × 10⁶ m³ of lava was erupted along a discontinuous, 15-km-long system of fissures. A wide compositional range of lavas was erupted including the most differentiated lavas (5.0 wt% MgO) from a historic Kilauea eruption. Lavas from the first half of the eruption are strongly differentiated (5.0–5.7 wt% MgO); later lavas are weakly to moderately differentiated (6.2–6.7 wt% MgO). Previous studies using only major-element compositions invoked either crystal fractionation (Macdonald and Eaton 1964) or magma mixing (Wright and Fiske 1971) as models to explain the wide compositional variation in the lavas. To further evaluate these models detailed petrographic, mineralogical, and whole-rock, major, and trace element XRF analyses were made of the 1955 lavas. Plagioclase and clinopyroxene in the early and late lavas show no petrographic evidence for magma mixing. Olivines from both the early and late lavas show minor resorption, which is typical of tholeiitic lavas with low MgO contents. Core-to-rim microprobe analyses across olivine, augite, and plagioclase mineral grains give no evidence of disequilibrium features related to mixing. Instead, plots of An/Ab vs distance from the core (D) and %Fo vs (D)^4.5 generated essentially linear trends indicative of simple crystal fractionation. Least-squares, mass-balance calculations for major- and trace-element data using observed mineral compositions yield excellent results for crystal fractionation (sum of residuals squared <0.01 for major elements, and <5% for trace elements); magma mixing produced less satisfactory results especially for Cr. Furthermore, trace-element plots of Zr vs Sr, Cr, and A1₂O₃ generate curved trends indicative of crystal fractionation processes. There is no evidence that mixing occurred in the 1955 lavas. Instead, the data are best explained by crystal fractionation involving a reservoir that extends at least 15 km along Kilauea's east rift zone. A dike was intruded into the rift zone from the summit reservoir eight days after the eruption started. Instead of causing magma mixing, the dike probably acted as a hydraulic plunger forcing more of the stored magma to be erupted.     

Geochemical trends of sequential lava flows from Meseta volcano, Guatemala

The Meseta and Fuego volcanoes closely overlap and collectively are known as the Fuego Volcanic Complex. Historic activity occurs exclusively at Fuego, the southern center, and consists of high-Al basalts. Meseta, the inactive northern center, is predominantly composed of basaltic andesites with minor basalt and andesite. A thick sequence of lava flows and dikes is exposed by a steep collapse escarpment on the east flank of Meseta. The upper 75% of the sequence was sampled from three interfingering stratigraphic sections consisting of 27, 10 and 4 lavas, respectively. Temporal geochemical trends of each section indicates a complex evolutionary history. A major trend toward more evolved compositions upward in the section is consistent with crystal fractionation. This trend is sharply interrupted by the youngest lavas which become distinctly more mafic in composition. Magma mixing is apparently the dominant magmatic evolution process that generated these lavas. The two trends have distinct Sr signatures that suggest a change in parental magma compositions. This abrupt change in composition is interpreted to signal high input rates of mafic magma into the subvolcanic magma chamber. These changes eventually led to sector collapse of Meseta volcano and deposition of the Escuintla debris avalanche. Eruptive activity then migrated to the Fuego volcano where historic activity is similar to that of Meseta immediately prior to its collapse.     

Isotopic geochemistry of Hualalai shield-stage tholeiitic basalts from submarine North Kona region,Hawaii

The geochemical data of Hualalai tholeiitic basalts allow extension of the temporal variations established at Mauna Loa back in time, and provide important information for the long-term temporal variation of the Hawaiian lavas. We report new Hf, Pb, Nd, and Sr isotope compositions for 32 Hualalai tholeiitic basalts collected from deep submarine portions of the North Kona region. The samples were collected from the lower section of the North Kona bench (dives K218 and K219), a submarine stratigraphic section at Hualalai volcano's northwest rift zone (dive S690), and an elongate ridge outboard of the central section of the bench (dive S692), during two JAMSTEC Hawaii cruises in 2001 and 2002. The Hualalai shield-stage tholeiitic basalts have magma source isotopic signatures similar to Mauna Loa. The new data shows temporal Pb and Sr isotope trends that correspond to the long-term temporal variations in Loa-trend lavas, and the Hualalai–Mauna Loa lavas seem to show inter-shield geochemical excursions. Variation in Pb and Sr isotopes at Hualalai appears to take place over a longer time scale than at Mauna Loa. The merged Hualalai–Mauna Loa isotopic trends support models where heterogeneous material in the plume conduit is distributed chaotically, with variable cross-sectional density and length scale.     

Petrology of the hilina formation,Kilauea volcano,Hawaii

The Hilina Formation comprises the oldest sequence of lava flows and tuffs exposed on Kilauea Volcano. These rocks are only exposed in kipukas in younger Puna Formation lavas along cliffs on the south flank of Kilauea Volcano. Locally, tuffs and flows of the Pahala Formation separate the underlying Hilina Formation rocks rom the overlying Puna Formation rocks. Charcoal collected from the base of the Pahala Formation yielded a C¹⁴ age of 22.800±340 years B.P. which defines a minimum age for the Hilina Formation. Hilina Formation lavas crop out over a wide region and probably originated from the summit area and from both rift zones. The Hilina Formation contains both olivine-controlled and differentiated lavas (using the terminology ofWright, 1971). The olivine-controlled lavas of the Hilina Formation are distinguishable mineralogically and geochemically from younger olivine-controlled Kilauea lavas. The younger lavas generally contain discrete low-calcium pyroxene grains. greater glass contents, higher K₂O/P₂O₅ ratios and lower total iron contents. Similar geochemical trends prevail for Manuna Loa lavas, and may typify the early lavas of Hawaiian shield volcanoes. Despite these similarities, the Hilina Formation (and all Kilauea) lavas have higher TiO₂ and CaO, and lower SiO₂ and Al₂O₃ contents than Mauna Loa Lavas. These differences have existed for over 30,000 years. Therefore, it is unlikely that the older lavas of Kilauea are compositionally similar to recent Mauna Loa lavas as was previously suggested. K₂O, TiO₂, Na₂ and Zr contents of lavas from a stratigraphic sequence of Hilina Formation lavas are variable. These variations may be utilized to subdivide the sequence into geochemical groups. These groups are not magma batches. Rather, they represent lavas from batches whose compositions may have been modified by crystal fractionation and magma mixing.     

Subdivision of the Archean lavas of the Abitibi area,Canada, from Fe-Mg-Ni-Cr relations

The lavas of a part of the Archean Abitibi region may be divided into three stratigraphic levels in each of which FeO-MgO-Ni-Cr contents conform to certain broad differentiation trends. Within each stratigraphic level, there is a tendency for rocks to become more felsic upwards. The earliest and stratigraphically lowest subdivision is composed largely of magnesium-rich basaltic lavas called the magnesian suite. In the central part of the pile, where basalts predominate, the lavas contain intermediate MgO concentrations, and display pronounced Fe enrichment in intermediate members similar to conventional tholeiites. In the latest and stratigraphically highest lavas, where andesites predominate, Fe depletion is characteristic; these lavas are grouped into a primitive calcalkaline suite. All of the Abitibi lavas contain unusually high Ni and Cr. Other Archean lava piles appear to be similarly divisible, although all three suites are not always present.Mafic end-members of the three complete differentiation suites are viewed as possible source magmas derived by partial melting in a primitive, olivine-rich parent, probably the Archean mantle. The earliest, and highest temperature magmas precipitated olivine, Al-clinopyroxene, and minor Al-orthopyroxene, and display moderate FeO, TiO₂, MnO, Al₂O₃, and CaO enrichment in more felsic members. The intermediate age lavas, derived originally by less complete melting in the parent, precipitated plagioclase, olivine, and lesser clinopyroxene, and display, as a result, strong Fe enrichment until, in intermediate members, magma volumes became small enough to yield P_f of levels sufficient to form clinopyroxene plus magnetite. The uppermost lavas, derived by relatively small volumetric melting in the parent, contain abundant Fe-Ti oxides in even the most mafic members, along with augite and plagioclase.     

Petrology of volcanic rocks from anjouan,comores archipelago

Anjouan is an extinct and substantially denuded volcanic island, belonging to the Comores Archipelago, a chain of volcanic islands in the northern Mozambique Channel. Volcanism was probably controlled by a complex regional stress pattern, closely implicated with the recent geotectonic evolution of East Africa and the western Indian Ocean. On Anjouan a lengthy period of shield construction was succeeded by fissure-controlled eruptions forming three extended peninsulas to the north, south and west. A ‘rejuvenescent’ phase of activity erupted lavas on to an erosional surface formed during a period of quiescence following build-up of the main volcanic edifice. Lavas of the shield-building stage comprise ankaramites, oceanites and olivine basalts, with minor development of hawaiites and trachytes. The fissure controlled eruptions are also mainly basaltic although generally more alkaline, while rejuvenescent lavas comprise basanites and a significant proportion of derivatives following a trend towards phonolite. Coarse-grained xenoliths are found in lavas of the latter two phases, while a gabbroic intrusion (the ‘Tatinga Intrusion’) is exposed in the centrally-situated Cirque de Bambao in the vicinity of N’Tingui (1595 m). The main chemical trends and petrographic characteristics of the Anjouan lavas are related to the eruptive sequence.     

Flux versus decompression melting at stratovolcanoes in southeastern Guatemala

The major and trace element geochemistry of lavas erupted from four volcanic front (VF) stratovolcanoes in southeastern Guatemala show differences in the relative importance of flux and decompression melting in a continental arc setting. The VF stratovolcanoes exhibit a wide compositional range from basalt to dacite, although modern Pacaya erupts basaltic lavas. The VF basalts have relatively low MgO contents and plot outside the field of primary arc magmas defined by melting experiments on hydrous peridotite. After subtracting the effects of the fractionation, assimilation, and alteration of some VF lavas, separate partial melting and mixing trends were identified for Agua–Pacaya and Tecuamburro–Moyuta.The distinct chemical signatures of the hemipelagic and carbonate sediments subducted off Guatemala provide constraints on material transfer processes that occurred between the slab and mantle wedge. Model fluids and melts from the subducted slab were calculated using recently published mineral–aqueous fluid partition coefficients. Wide separation of the model fluid and melt compositions on a U/La versus Ba/Th diagram creates diagnostic mixing curves with an enriched mid-ocean ridge basalt source. Fluid from mature ocean crust has high U/La, fluid from carbonate sediment has high Ba/Th, and fluid and melt from hemipelagic sediments have both high U/La and Ba/Th. In a simple single-stage model, a mantle metasomatized by fluid originating largely from the oceanic crust with only minor sediment fluid contributions best explains the overall large ion lithophile element composition of the VF lavas. (Th/Rb)_N ratios of ∼1 in the VF lavas from southeastern Guatemala require a component of sediment melting. Therefore, a more realistic two-stage model to describe the Guatemalan arc data involves an initial hemipelagic sediment melt input to the wedge followed by minor fluid additions from the oceanic crust or sediments. Correlation between measures of slab input and extent of melting in the older VF lavas from Tecuamburro and Moyuta favors flux-dominated melting near the base of the mantle wedge. In sharp contrast, the lack of a relationship between slab additions and melting in younger lavas from Agua and Pacaya volcanoes implies a significant role for decompression melting closer to the top of the wedge. In this melting scenario, the rate of crustal extension determines the extent of melting.     

Relation of lava compositions to volcano size and structure in El Salvador

Most of the lavas at the nine volcanic centers along the volcanic front of El Salvador are basalts, basaltic andesites and andesites. The compositional variation within and among these centers can be explained by fractionation processes within the crust. Cognate gabbroic inclusions found in the lavas have appropriate mineralogy (plagioclase, olivine, magnetite and augite) to be cumulates formed by fractional crystallization. Two main variation trends occur, depending on the proportion of plagioclase removal. The more common, or normal, trend has a high (> 55%) proportion of plagioclase being removed. A less common, Al-rich, trend has a low (40%) proportion of plagioclase being removed. The Al-rich trend is found only at volcanoes that lack large negative Bouguer gravity anomalies. These volcanoes are unlikely to have large shallow magma chambers and fractionation probably occurs deeper in the crust where plagioclase removal is inhibited.The incompatible element (Na₂O, K₂O, Rb, Ba) contents of lavas vary systematically with the volume of the volcanic centers. At the same level of SiO₂, large volcanic centers have higher incompatible element contents than small volcanic centers. This suggests that open system fractionation in a periodically refilled chamber is the controlling factor. The large difference in Ba contents of lavas between eastern (low) and western (high) El Salvador suggests a difference in the mantle source region.     

Differentiation and magma mixing on Kilauea's east rift zone

The lavas of the 1955 east rift eruption of Kilauea Volcano have been the object of considerable petrologic interest for two reasons. First, the early 1955 lavas are among the most differentiated ever erupted at Kilauea, and second, the petrographic character and chemical composition of the lava being erupted changed significantly during the eruption. This shift, from more differentiated (MgO=5.0–5.7%) to more magnesian (MgO=6.2–6.8%) lava, has been variously interpreted, as either due to systematic excavation of a zoned, differentiated magma body, or to invasion of the differentiated magma by more primitive magma, followed by rapid mixing and eruption of the resulting hybrid magmas. Petrologic examination of several nearvent spatter samples of the late 1955 lavas shows abundant evidence for magma mixing, including resorbed and/or reversely zoned crystals of olivine, augite and plagioclase. In addition, the compositional ranges of olivine, plagioclase and groundmass sulfide are very large, implying that the assemblages are hybrid. Core compositions of olivine phenocrysts range from Fo85 to Fo77. The most magnesian olivines in these samples must have originally crystallized from a melt containing 8.0–8.5% MgO, which is distinctly more magnesian than the bulk composition of the late 1955 lavas. The majorelement and trace-element data are either permissive or supportive of a hybrid origin for the late 1955 lavas. In particular, the compositional trends of the 1955 lavas on plots of CaO vs MgO, and the virtual invariance of Al₂O₃ and Sr in these plagioclase-phyric lavas are more easily explained by magma mixing than by fractionation. The pattern of internal disequilibrium/re-equilibration in the late 1955 spatter samples is consistent with reintrusion and mixing having occurred at least twice, during the latter part of the 1955 eruption. Plagioclase zonation preserves possible evidence for additional, earlier reintrusion events. Least-squares modelling the mixing of early 1955 bulk compositions with various summit lavas±olivine pick the 1952 summit lava as most like the primitive component. The results also indicate the primitive component had MgO=7.5–8.0%, corresponding to liquidus temperatures of 1165–1175°C. The absence of Fe-Ti oxide phenocrysts in the late 1955 lavas implies that the cooler component of the hybrid had T>1110°C. Thus the thermal contrast between the two components may have been as much as 55–65°C, sufficient to produce the conspicuous disequilibrium effects visible in the spatter samples.     

Remagnetization in tertiary rocks from the Faeroe Islands

Palaeomagnetic investigations of highly oxidized interbasaltic horizons and associated Lower Tertiary lavas from the Faeroe Islads suggest a long history of magnetization. The remanence directions are composed mostly of two deviating directions of opposite polarities aligned along an axis which palaeomagnetically appears to be younger than that of the original field. This, in addition to the observation that both the baking and baked rocks in most cases show almost identical directional trends upon demagnetization, questions the general validity of the baked contact test. It is suggested that to a major extent post-baking processes, such as low-temperature oxidation and precipitation of ferric oxides, have remagnetized the lavas as well as the tuff layers. Only two of the seven interbasaltic horizons investigated appear to have retained the ‘primary’ magnetization. It is concluded that besides a high bulk oxidation state (induced prior to, or as a result of baking) a “hornfelsic” nature of the baked sediments may be a necessary requirement for establishing primary palaeomagnetic directions in many volcanic regions.     

Petrogenesis of boninites in the Ordovician Ballantrae Complex ophiolite, southwestern Scotland

Primitive lava and hyaloclastite with unusual, highly refractory compositions, form part of the Early Ordovician Balcreuchan Group within the ophiolitic Ballantrae Complex, southwestern Scotland. They are identified as likely high-Ca boninites on the basis of new XRF and INAA results and are the first unambiguous boninites to be discovered in the British Isles. The boninites are interbedded with low-Ti tholeiitic lavas with which they share some distinctive geochemical characteristics suggestive of a close petrogenetic relationship. The low-Ti tholeiite lavas have been interpreted as island-arc tholeiites but they also resemble back-arc basin basalts. The newly discovered boninites confirm an intra-oceanic environment of eruption; their distinctive features include relatively high SiO₂, MgO, Cr and Ni but low Al₂O₃ and HFSE abundances, U-shaped REE patterns, low Ti/Zr and high Zr/Hf ratios. Bulk geochemical trends are indicative of low-temperature, seawater-dominated alteration of the lavas but these alteration conditions apparently had little effect on the distribution of critical diagnostic elements such as Zr, Ti, Sc, Ta and the mid-heavy rare earths. We suggest that the Ballantrae boninites and low-Ti tholeiites represent different batch melts derived from a common, depleted mantle source region variably modified compositionally (i.e., made “streaky”) by fluids and/or melts during slab interaction (subduction metasomatism). A contribution from slab-derived pelagic sediments and/or a carbonatite melt is necessary to account for the fractionated, non-chondritic Zr/Hf ratios in the boninites. In view of the close compositional similarity of the Ballantrae lavas to Cenozoic boninite suites, we believe that these interpretations may have wider application to the petrogenesis of boninites in general.     

Isotope geochemistry and origin of calc-alkaline lavas from a caledonian continental margin volcanic arc

Analyses for major and trace elements, including REE, and Sr, Nd and Pb isotopes are reported from a suite of Siluro-Devonian lavas from Fife, Scotland. The rocks form part of a major calc-alkaline igneous province developed on the Scottish continental margin above a WNW-dipping subduction zone. Within the small area (ca. 15 km²) considered, rock types range from primitive basalts and andesites (high Mg, Ni and Cr) to lavas more typical of modern calc-alkaline suites with less than 30 ppm Ni and Cr. There is a marked silica gap between these rocks (< 62%) and the rare rhyolites (> 74%), yet the latter can be generated by fractional crystallization from the more mafic lavas. In contrast, variation in incompatible element concentrations and ratios in the mafic lavas can not be generated by fractional crystallization processes. Increasing SiO₂ is accompanied by increasing Rb, K, Pb, U and Ba relative to Sr and high field strength elements, increasing LREE enrichment and increasing _Sr calculated at 410 Ma, and by decreasing HREE, Eu/Eu*, Sm/Nd and _Nd (410). _Nd and _Sr are roughly anticorrelated and have more radiogenic compositions than the mantle array, in common with data reported elsewhere from this part of the arc. The correlation extrapolates up to cross the mantle array within the composition field of the contemporary MORB source, and extrapolates down towards the probable compositional range of Lower Palaeozoic greywackes, which may form the uppermost 8 km of the crust, or may be supplied to the source by subduction. One sample, however, lies within the mantle array, and closely resembles lavas from northwestern parts of the arc, where a mantle source with mild time-integrated Rb/Sr and LREE enrichment has been inferred. The lavas have relatively high initial ²⁰⁷Pb/²⁰⁴Pb for their ²⁰⁶Pb/²⁰⁴Pb, a feature which has been interpreted elsewhere as the result of incorporation of a sediment component into arc magmas. The systematic changes with increasing SiO₂ in isotopic and chemical parameters can be explained by mixing of a greywacke-derived component with depleted mantle. The various possible mixing mechanisms are discussed, and it is considered most likely that mixing occurred in the mantle source through greywacke subduction. The bulk of the Rb, K, Ba and Pb in the lavas is probably recycled from the crust, whereas less than some 40% of the Sr and Nd is recycled. The calc-alkaline chemical trends are solely a function of mixing with the sediment component.     

Petrochemistry of Kamen volcano: A comparison with neighboring volcanoes of the Klyuchevskoy group

The data on geology, petrography, mineralogy and petrochemistry for Kamen volcano in the Central Kamchatka Depression are presented. A study of the volcano??s rocks and comparison with rocks of neighboring active volcanoes of the Klyuchevskoy group allow the establishment of some relationships. The rocks and minerals of Kamen and Ploskie Sopky volcanoes show systematic differences in the chemistry of rocks and minerals such that they obviously could not be formed from the same primary melts. The rocks of dykes and Kamen stratovolcano on one hand and the rocks of the Klyuchevskoy Volcano on the other hand form differently directed trends on petrochemical diagrams and differ in their compositions of rock-forming minerals, such they also could not originate from the same primary melts. The lavas of the monogenetic cones of Kamen volcano and moderately magnesian basalts of Klyuchevskoy volcano are derivates of the same melts, i.e., the cones situated on the slopes of Kamen are cones of Klyuchevskoy. The rocks of Kamen and Bezymianny stratovolcanoes form a single narrow trend in all petrochemical diagrams in which the lavas of Bezymianny volcano show a silica-rich part, thus indicating a genetic relationship between these two volcanoes.     

Petrology and geochemistry of volcanic rocks of Aegina,Greece

The Plio-Pleistocene cale-alkalic lavas of Aegina are comparable geochemically with similar lavas from elsewhere in the South Aegean arc. At least four differentiation series, largely of basaltic andesite to rhyodacite type, have been recognized in Aegina. Petrographical and geochemical evidence shows that lavas and xenoliths have a common origin. The mineralogy of the rocks suggests an alternation of hydrous and dry conditions during their formation. The chemical composition of the basic rocks, and of basic cognate xenoliths in the lavas, suggest that the parent material of all the lavas was of basic composition. Fractionation of the early formed phases is believed to have been the main process for the formation of the more acidic differentiates.     

Ce anomalies in Gough Island lavas — Trace element characteristics of a recycled sediment component

High precision trace element data obtained by inductively coupled plasma mass spectrometry and Sr–Nd isotope analyses are presented for mafic volcanic rocks from Gough Island, South Atlantic. The new data reveal negative Ce anomalies, with Ce/Ce^? values in Gough lavas extending down to ～ 0.92. Ce is only fractionated from other rare earth elements (REE) due to formation of Ce⁴⁺ under oxidizing conditions of near-surface environments while other REE remain trivalent. Ce anomalies in convergent margin magmas have been shown to indicate a contribution of a subducted sediment component. In contrast, Ce anomalies in intra-plate basalts have been attributed to weathering processes, but can be excluded here based on element–element systematics indicating magmatic trends rather than weathering-induced element mobility. Shallow-level contamination by local marine sediments with negative Ce anomaly inherited from seawater can be excluded because Gough lavas with increasingly negative Ce anomalies do not trend towards low Ce/Pb ratios characterizing such sediments. Rather, it is argued that the negative Ce anomalies in Gough Island lavas are consistent with variable amounts of a sediment component in the mantle plume source. Mixtures between estimates of subducting sediment columns with negative Ce anomaly and mantle capable of giving rise to Gough Island magmas without Ce anomalies reproduce the Gough compositional array with the exception of highly fluid-mobile elements. The calculated trace element composition of the deeply recycled sediment in the Gough plume source is depleted in fluid-mobile elements such as Ba and Pb relative to the composition of some present-day subducting sediments. This loss is attributed to the dehydration or flushing of sediment in the subduction factory, consistent with constraints from arc magmas.     

Flow and emplacement direction determined for selected basaltic bodies using magnetic susceptibility anisotropy measurements

Low-field anisotropy of magnetic susceptibility (AMS) has been determined for a total of 248 basaltic specimens taken from cross sections between the cooling interfaces of 6 subaerial lavas, 6 deep-sea lavas, and 6 intrusives (5 dikes and 1 sill). Statistically significant AMS clusters are exhibited by all the dikes examined and, based upon these clusters, derivation of emplacement direction becomes possible. Two lavas are observed to have statistically significant AMS clusters which can be used for flow direction determinations. The methods of emplacement and flow direction analysis are discussed as well as the statistics used. It is concluded that most of the dikes examined have low angle emplacement directions. A classification scheme for AMS data distributions is presented.The AMS analysis shows that intrusives and deep-sea lavas can be distinguished from subaerial lavas approximately 80% of the time by the random AMS ellipsoid orientations exhibited in subaerial lavas. Contrasts in the fluid properties, degassing, wall effects with subsequent distortion of the fluid, and grain interaction during the extrusion of subaerial lavas can be expected to distort magnetic grain alignment. Further effects such as convection and secondary processes contribute to yield the random distributions observed for most of these bodies.     

Be isotope and B-Be investigations of the historic eruptions of Mt. Vesuvius

Cosmogenic and radioactive ¹⁰Be, stable ⁹Be and B concentrations have been determined for four alkaline lavas from the recent cycle of historical activity at Mt. Vesuvius, in the Central Campania Province (Italy). The goal of this study was to use the Be isotope and Be-B systematics of these lavas in a manner analogous to that used in regions unequivocally related to active subduction, in hopes of being able to document a subduction origin for the Vesuvian lavas. Four lavas measured to date have low ¹⁰Be concentrations as well as low B/Be ratios. While the low ¹⁰Be concentrations could reflect subduction and incorporation of old sediments and/or only contributions from the basaltic part of the subducting plate, the combination of low ¹⁰Be and low B/Be ratios must be interpreted as the absence of a subduction signature in these lavas. Unfortunately, the absence of a subduction signature in the lavas cannot be used unequivocally to argue against recent subduction in the Central Campania region. Subduction of an unusually young, hot slab which has lost its B through prograde metamorphic reactions at shallow levels could explain the absence of a subduction signature, as could extensive crustal contamination. In addition, recent studies in Java show that alkaline lavas in this region of active subduction never show a subduction signature, even when erupting in close proximity to calc-alkaline and tholeiitic lavas which do; by analogy, the Vesuvian lavas studied may have been generated in a part of the sub-arc mantle which either does not experience or does not preserve the chemical signature of subduction modification. The present data set does not allow us to make definitive statements about the role of subduction in the origin of the Central Campania lavas.     

Trace element transport rates in subduction zones: evidence from Th, Sr and Pb isotope data for Tonga-Kermadec arc lavas

Trace element and Th, Sr and Pb isotope data for young lavas from the Tonga-Kermadec arc in the southwest Pacific suggest that geochemical variations in the lavas along the arc are linked to differences in the material being subducted beneath the arc. Lavas from the southern (Kermadec) segment of the arc have relatively radiogenic Pb isotope compositions, which reflects a contribution from subducted sediment. In contrast, much of the Pb in Tonga lavas is derived from the altered oceanic crust in the subducting Pacific Plate, and lavas from the northernmost Tonga islands of Tafahi and Niuatoputapu contain Pb and Sr derived from the subducted part of the Louisville Seamount Chain. The origin of the Pb in the lavas from these two islands can thus be traced to a point on the subducting slab, and this observation is used to estimate the rate at which trace elements are transported beneath the arc. Our calculations suggest that fluid-soluble elements such as U, Sr and Pb are transported from the subducted slab, across the mantle wedge and back to the surface in lavas over a period of approximately 2–3 Ma, and that magmas are erupted at the surface less than 350 ka after the melts are generated in the mantle wedge.     

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.     

Distinguishing strongly rheomorphic tuffs from extensive silicic lavas

High-temperature silicic volcanic rocks, including strongly rheomorphic tuffs and extensive silicic lavas, have recently been recognized to be abundant in the geologic record. However, their mechanisms of eruption and emplacement are still controversial, and traditional criteria used to distinguish conventional ash-flow tuffs from silicic lavas largely fail to distinguish the high-temperature versions. We suggest the following criteria, ordered in decreasing ease of identification, to distinguish strongly rheomorphic tuffs from extensive silicic lavas: (1) the character of basal deposits; (2) the nature of distal parts of flows; (3) the relationship of units to pre-existing topography; and (4) the type of source. As a result of quenching against the ground, basal deposits best preserve primary features, can be observed in single outcrops, and do not require knowing the full extent of a unit. Lavas commonly develop basal breccias composed of a variety of textural types of the flow in a finer clastic matrix; such deposits are unique to lavas. Because the chilled base of an ashflow tuff generally does not participate in secondary flow, primary pyroclastic features are best preserved there. Massive, flow-banded bases are more consistent with a lava than a pyroclastic origin. Lavas are thick to their margins and have steep, abrupt flow fronts. Ashflow tuffs thin to no more than a few meters at their distal ends, where they generally do not show any secondary flow features. Lavas are stopped by topographic barriers unless the flow is much thicker than the barrier. Ash-flow tuffs moving at even relatively slow velocities can climb over barriers much higher than the resulting deposit. Lavas dominantly erupt from fissures and maintain fairly uniform thicknesses throughout their extents. Tuffs commonly erupt from calderas where they can pond to thicknesses many times those of their outflow deposits. These criteria may also prove effective in distinguishing extensive silicic lavas from a postulated rock type termed lava-like ignimbrite. The latter have characteristics of lavas except for great areal extents, up to many tens of kilometers. These rocks have been interpreted as ash-flow tuffs that formed from low, boiling-over eruption columns, based almost entirely on their great extents and the belief that silicic lavas could not flow such distances. However, we interpret the best known examples of lava-like ignimbrites to be lavas. This interpretation should be tested through additional documentation of their characteristics and research on the boiling-over eruption mechanism and the kinds of deposits it can produce. Flow bands, flow folds, ramps, elongate vesicles, and probably upper breccias occur in both lavas and strongly rheomorphic tuffs and are therefore not diagnostic. Pumice and shards also occur in both tuffs and lavas, although they occur throughout ash-flow tuffs and generally only in marginal breccias of lavas. Dense welding, secondary flow, and intense alteration accompanying crystallization at high temperature commonly obliterate primary textures in both thick, rheomorphic tuffs and thick lavas. High-temperature silicic volcanic rocks are dominantly associated with tholeiitic flood basalts. Extensive silicic lavas could be appropriately termed flood rhyolites.