| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology | Volume 44 | Number 12 | Pages 2243-2260 | 2003
© Oxford University Press 2003; all rights reserved
Andesites and Dacites from Daisen Volcano, Japan: Partial-to-Total Remelting of an Andesite Magma Body
1 IFREE, JAMSTEC, YOKOSUKA 237-0061, JAPAN
2 FACULTY OF SCIENCES, FUKUOKA UNIVERSITY, FUKUOKA 814-0180, JAPAN
3 OCEAN RESEARCH INSTITUTE, UNIVERSITY OF TOKYO, TOKYO 164-8639, JAPAN
* Corresponding author. Telephone: +81-468-67-9761. Fax: +81-468-67-9625. E-mail: tamuray{at}jamstec.go.jp
RECEIVED AUGUST 20, 2002; ACCEPTED JUNE 16, 2003
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTIONVoluminous andesite and dacite lavas of Daisen volcano, SW Japan, contain features suggesting the reverse of normal fractionation (anti-fractionation), in the sense that magma genesis progressed from dacite to andesite, accompanied by rises in temperature. A positive correlation exists between phenocryst content (0–40 vol. %) and wt % SiO2 (61–67%). Phenocryst-rich dacites contain hornblende and plagioclase that are generally unaltered, clear, and euhedral. However, phenocryst-poor rocks contain sieve-textured plagioclase, resorbed plagioclase, and opacite in which hornblendes are pseudomorphed. Some Daisen rocks contain two coexisting pyroxenes. Many orthopyroxene phenocrysts from two-pyroxene lavas have high-Ca overgrowth rims (up to 50 µm), a feature consistent with crystallization from a higher-temperature magma than the core. Rim compositions are similar from phenocryst to phenocryst in individual samples. Temperatures of 800–900°C are obtained from the cores, whereas temperatures of 1000–1100°C are indicated for the rims. Lavas ranging from aphyric andesite (
61 wt % SiO2) to phenocryst-rich dacite (67 wt % SiO2) have similar 87Sr/86Sr (0·7045–0·7052) and 143Nd/144Nd (0·5127– 0·5128). Isotopic variability within Daisen volcano is likely to be mantle-derived, reflecting isotopic variability within the magma source region associated with a single mantle diapir. The Daisen andesites and dacites have the same trace element signatures as the associated basalts and were probably derived from primary magmas at the same general depth (60 km). Our interpretation is that mantle-derived hydrous magnesian andesite, generated in the same mantle diapir as coexisting basalt magma, may be parental to the hydrous calc-alkaline magmas in Daisen volcano. We suggest a two-stage process, involving mid-crustal solidification of bodies of this calc-alkaline magma followed by varying degrees of partial melting from body to body, to produce the magmatic trends and phenocryst zoning patterns observed. The heat required for this melting, according to our model, was supplied by the intermittent rise of subjacent basaltic magma. KEY WORDS: arc volcanism; Daisen volcano; remelting; andesite genesis; superheating
|
INTRODUCTION |
|---|
Daisen volcano, SW Japan, is made up mostly of andesite and dacite. Basaltic rocks also occur; however, the volcanic association is clearly bimodal with a SiO2 gap of
8 wt % between the basalts and andesites (Tamura et al., 2000
). Many Daisen andesites are aphyric, contrasting sharply with associated dacites that contain 30–40% phenocrysts. Ewart (1976) showed that arc-volcano andesites in general show a bimodal distribution related to phenocryst content; one group consists of aphyric rocks and the other contains 30–40% phenocrysts. The coexistence of aphyric and phenocryst-rich rocks may yield important information relating to the genesis of calc-alkaline andesites and dacites at Daisen volcano and perhaps at arc volcanoes in general. We present evidence here that the absence of phenocrysts in the aphyric magmas suggests that they were superheated in the near-surface environment. We further suggest that basaltic magma input aided the generation of andesites and dacites at Daisen volcano, and that the additional thermal energy of the basalt was essential for this process.
Huppert & Sparks (1988) showed that the melting of the crust or, as envisaged here, solidified silicic magma body, has a strong cooling effect on basaltic magma. Evidence for supercooling of basaltic magmas is well developed at Daisen volcano (Tamura et al., 2000). Arc basalts from Daisen volcano ubiquitously bear olivine phenocrysts (5–11 vol. %) with rare Cr-spinel inclusions, but some lava flows are characterized by the great abundance of skeletal iron-rich olivines having highly irregular shapes. The effects of fractional crystallization were distinguished from those produced by supercooling on the basis of olivine morphology and chemical relationships between olivines and host basalts (Tamura et al., 2000).
We propose that the petrological features of Daisen lavas during the last 1 Myr can be attributed to ‘anti-fractionation’, in which episodes of heating (and remelting) of solidified andesite protolith produced the compositional variations of the volcanic rocks observed at the surface. Our model proposes that andesite and dacite eruptions were triggered by the influx of hot basalt magmas from the mantle, which reheated and softened crustal andesite magma bodies at depth, permitting them to erupt. Basalt magma erupted to the surface only during the initial stage of Daisen volcano (1 Ma), but the repeated eruption of calc-alkaline magmas at later stages suggests that basalt magmas continued to intrude beneath the volcano, providing heat to partially (or completely) melt solidified calc-alkaline magma bodies throughout the life of the volcano.
|
ANALYTICAL METHODS |
|---|
Major and trace elements
After initial splitting and jaw crushing, all samples were pulverized in an agate ball mill. H2O– and loss on ignition were determined at 110°C and 950°C (5 h), respectively. Major and trace elements were determined by X-ray fluorescence (XRF) at the Ocean Research Institute, University of Tokyo. Trace elements were analysed on pressed powder discs, and major elements were determined on fused glass discs. A mixture of
0·5 g of each powdered sample and 5 g of anhydrous lithium tetraborate (Li2B4O7) was used; no matrix correction was applied because of the high dilution involved. Instrumental neutron activation analysis (INAA) was carried out with the Kyoto University reactor (KUR) and a gamma-ray spectrometer with a Ge (Li) detector at the RadioIsotope Center, Kanazawa University, using the procedures of Ishiwatari & Ohama (1997). Microprobe analyses were carried out on the JAMSTEC JEOL JXA-8900 Superprobe equipped with five wavelength-dispersive spectrometers (WDS).
Isotopic compositions
87Sr/86Sr and 143Nd/144Nd ratios were determined with thermal ionization mass spectrometers (Finnigan MAT 262 and MAT 261) at Niigata University. Sr isotopes were normalized to 86Sr/88Sr = 0·1194 and adjusted to 0·710241 for the 87Sr/86Sr value of the NBS-987 standard. Nd isotopic ratios were corrected by normalization to 146Nd/144Nd = 0·7219, and values of 143Nd/144Nd are given relative to a value of 0·512784 for the JB-1a standard. This Nd isotopic ratio of JB-1a corresponds to 143Nd/144Nd = 0·512638 of BCR-1 (Kagami et al., 1989). The blanks for the whole procedure were Rb <0·25 ng, Sr <0·52 ng, Sm <0·025 ng, and Nd <0·22 ng. Detailed isotopic analytical procedures have been reported by Miyazaki & Shuto (1998), Hamamoto et al. (2000) and Yuhara et al. (2000).
|
PREVIOUS STUDIES OF DAISEN VOLCANO |
|---|
Daisen volcano, SW Japan, is a volcanic complex ranging in age from 1·3 to 0·02 Ma, and consists of clustered and overlapping lava domes and associated lava flows and pyroclastic flows (Tsukui, 1984
, 1985; Tamura et al., 2000; Fig. 1). The volcanic association of Daisen volcano is bimodal, consisting of primitive basalts (50 wt % SiO2, 7–10 wt % MgO) and andesites and dacites (60–68 wt % SiO2, 3–1 wt % MgO) (Tamura et al., 2000). Morris (1995) noted Y and heavy rare earth element (HREE) depletions in dacites from Daisen volcano and suggested that the dacites originated by slab melting, leaving an eclogite residue. However, the slab melting scenario is not consistent with the genesis of the Daisen basalts, which have similar 87Sr/86Sr and 143Nd/144Nd ratios and the same general age as the older dacites and andesites of Daisen volcano (Tamura et al., 2000). Daisen basalts also record the signature of residual garnet in their trace element characteristics. Tamura et al. (2000) estimated primary arc basalt magma compositions for Daisen volcano (50 wt % SiO2, 11 wt % MgO and 1 wt % K2O) and suggested a segregation pressure of 18 kbar (60 km). Concurrent generation of basalt and magnesian andesite and the likelihood of garnet in the residual mantle would explain the signature of garnet, or its transitional signature, in the trace element compositions of the Daisen basalts and dacites (Tamura et al., 2000). Tsukui (1985) studied the temporal relations of dacite phenocryst compositions in the Upper Tephra Group (erupted during the last 150 kyr) and showed that magmatic temperatures estimated by the Fe–Ti oxide geothermometer fluctuated between 850°C and 950°C with three temperature cycles during the past 150 kyr. Tsukui (1985) considered that a magma reservoir existed at a shallow depth below Daisen volcano, and received intermittent inputs of hotter magma from depth.
|
|
GEOLOGICAL AND ANALYTICAL FRAMEWORK |
|---|
A geological map of Daisen volcano (Fig. 1) shows the eruption ages of the main lithological units that make up the volcano, as well as the location of some of the samples referred to in this paper. The Daisen volcanic centre began with the eruption of basalt at
1·3 Ma and ended with the extrusion of the Misen dacite lava dome at 17 000 BP (Tsukui et al., 1985). K–Ar ages of porphyritic dacites range continuously from 1·0 to 0·02 Ma (Tsukui et al., 1985; Uto, 1989; Kimura et al., 2003). Eruption ages of the aphyric andesites concentrate at 1 Ma and 0·5 Ma (Tsukui et al., 1985; Kimura et al., 2003). Thus, some lava flows and lava domes of aphyric andesite and porphyritic dacite have overlapping ages. Major element, trace element and isotopic data are presented in Table 1. Representative modal analyses are shown in Table 2. All discussions refer to analyses normalized to 100% volatile free, with total iron calculated as Fe2O3.
|
|
|
PETROGRAPHY OF ANDESITES AND DACITES |
|---|
We define crystals >200 µm in size as phenocrysts, and smaller crystals as groundmass microlites. The 200 µm criterion is somewhat arbitrary for a few samples such as d-28, in which most crystals are relatively small. Most other samples, however, display clear porphyritic textures, and the 200 µm phenocryst–groundmass distinction is readily apparent.
Phenocryst contents
Plots of groundmass content (vol. %) against wt % SiO2 are shown in Fig. 2. Ewart (1976) found a bimodal distribution of phenocryst contents in low- and medium-K arc-related andesites (56–63% SiO2), with the main bulk of eruptives having phenocryst contents in the 30–40% range, but with aphyric and phenocryst-poor rocks (<10% phenocrysts) forming another peak. Thus, the phenocryst content of the Daisen rocks, ranging from 0 to 40 vol. %, is common in arc andesite volcanoes. A negative correlation exists between these petrographic values and wt % SiO2 in Daisen volcano (Fig. 2). Aphyric andesites (0–3 vol. % phenocrysts) have SiO2 contents ranging from 61 to 63·5 wt %. Phenocryst-rich rocks (20–40 vol. % phenocrysts) are biotite-bearing orthopyroxene hornblende dacite or clinopyroxene orthopyroxene andesite and dacite containing 62·5–68 wt % SiO2. Daisen andesites and dacites can thus be divided into three groups by the dashed lines in Fig. 2: (1) aphyric andesites (<3% phenocrysts); (2) phenocryst-poor andesite (7–20% phenocryst); (3) phenocryst-rich andesites and dacites (20–40% phenocrysts) (Fig. 2).
|
|
|
Although there is considerable scatter in the data in Fig. 2, it is noteworthy that aphyric rocks are confined to the mafic part of the silicic rocks (<63·5% SiO2) and no aphyric dacites (>64% SiO2) exist in Daisen volcano. Photomicrographs of representative aphyric andesite (d-22), phenocryst-rich dacites (d-52, d-12 and d-83) and transitional phenocryst-poor andesites (d-30 and d-90) are shown in Fig. 3. The groundmass of the aphyric andesites (Fig. 3a) is usually more coarsely crystalline (microcrystalline) than the cryptocrystalline texture in the phenocryst-rich andesite and dacites (Fig. 3b). It is possible that the phenocryst-rich dacites quenched more readily than the aphyric andesites because of their higher content of groundmass H2O.
Phenocryst textures, assemblages and volume percentages
Phenocryst textures, phase assemblages, and phenocryst volume percentages show important interrelationships (Fig. 3). Phenocrysts of hornblende and plagioclase in phenocryst-rich dacites are generally unaltered, clear, and euhedral (Fig. 3b, g and h). In contrast, aphyric and phenocryst-poor rocks contain sieve-textured plagioclase (Fig. 3c), resorbed plagioclase (Fig. 3d), and opacite in which hornblendes are pseudomorphed (Fig. 3e). Clinopyroxene phenocrysts, relatively minor in volume, seem to be incompatible with fresh hornblendes, but they appear where hornblendes are completely pseudomorphed by opacite (Table 2). Sample d-97 is exceptional, but clinopyroxenes in this sample show marginal resorption. Quartz phenocrysts are almost invariably resorbed. Moreover, quartz in d-22, d-30, d-61, d-72, d-90 and d-97 is jacketed by overgrowths of clinopyroxene. Significantly, quartz is commonly associated with opacitized hornblende, but quartz and fresh hornblende are never observed to coexist (Table 2).
|
PYROXENE THERMOMETRY |
|---|
Some Daisen rocks contain two coexisting pyroxenes (the open squares in Fig. 2), permitting magmatic temperatures to be estimated using the two-pyroxene thermometer of Lindsley & Andersen (1983)
(Fig. 4). Both core and rim compositions of individual phenocrysts are plotted in each sample. Some samples contain pyroxenes having relatively uniform temperatures; pyroxenes in d-50 indicate magmatic temperatures of 1000–1100°C, whereas those in d-45 clearly indicate a lower temperature (Fig. 4a). Many rocks, however, show a remarkable scatter of coexisting pyroxene compositions, which do not indicate unique magmatic temperatures; the orthopyroxene temperatures range from <800°C to >1100°C even within a single thin section (Fig. 4). What causes the wide range of pyroxene temperatures and how are they observed in individual samples?
Figure 5 shows back-scattered electron images (BEI) and distribution maps of Ca and Mg contents of orthopyroxene phenocrysts in samples d-28 and d-104. Chemical profiles of these orthopyroxenes are presented in Fig. 6. Many orthopyroxene phenocrysts from most two-pyroxene-bearing lava flows have overgrowth rims (up to 50 µm) with a high Ca content, a feature consistent with crystallization from a higher-temperature magma than their cores (Fig. 5). Generally, temperatures of 800–900°C are obtained from the cores, whereas 1000–1100°C is indicated for the rims (Figs 4, 5 and 6). Moreover, many orthopyroxene phenocrysts show reversibly zoned patterns in terms of Mg number (Figs 5a–c and 6a, b). The zoning patterns of Wo and Mg number, however, do not necessarily correlate with each other. Magnesian orthopyroxene compositions (Fs20) in d-46 and d-104 (Fig. 4) reflect magnesian core compositions of some orthopyroxene phenocrysts shown in Figs 5d and 6c. Orthopyroxene 5 and orthopyroxene 3 in d-104 have similar rim compositions (Mg number 70, Wo 3), but they have different iron-rich (Mg number 65) and magnesian (Mg number 80) cores, respectively (Figs 5c, d and 6b, c). Interestingly, however, these magnesian cores also have lower Ca contents (Fig. 5d), suggesting crystallization at lower temperatures (900–1000°C) than their rims (1000–1100°C) (Fig. 4).
>
|
|
Rim compositions are similar from phenocryst to phenocryst in individual samples (Fig. 6b and c), thus reverse-zoned crystals having significant differences between core and rim compositions are mainly responsible for the observed scatter of orthopyroxene temperatures.
|
ISOTOPES |
|---|
Lavas ranging from aphyric andesite (
61 wt % SiO2) to phenocryst-rich dacite (67 wt % SiO2) from Daisen volcano have similar 87Sr/86Sr (0·7041– 0·7052) and 143Nd/144Nd (0·5127–0·5128) (Fig. 7). The Sr isotopic variability, however, is fairly large (>0·001), perhaps suggesting some amount of crustal contamination (Tamura & Nakamura, 1996). Figure 8a shows a 143Nd/144Nd vs 87Sr/86Sr variation diagram for Daisen volcano. Andesites and dacites have lower 87Sr/86Sr and higher 143Nd/144Nd than basalts, but their values slightly overlap and the total variation is continuous (Fig. 8). A model where contaminated basalt is derived from andesite is unlikely on both thermal and chemical grounds. Possible theoretical assimilation of crust by Daisen magmas is shown in Fig. 8b. The curved lines show trends of bulk assimilation of representative basement granite (d-54, Table 2). d-54 has isotopic values that are representative of granitoid rocks of the San'in belt, SW Japan (Kagami et al., 1992). Generally, 87Sr/86Sr values of granites in SW Japan are relatively low (Kagami et al., 1992), thus mixing between Daisen magmas and the granites would not cause drastic changes in 87Sr/86Sr and 143Nd/144Nd. Most Daisen basalts are primitive; they contain only olivine phenocrysts and show no evidence of crustal assimilation (Tamura et al., 2000). Thus, the percentages of assimilated granite d-54 of 10–30% are unrealistic. In addition to the unrealistic mixing ratios, the Sr and Nd isotopic variation of andesites and dacites does not follow the mixing line. Similar results are obtained by considering other granitoid rocks in SW Japan (Kagami et al., 1992). In addition, contamination of mantle-derived magma with granulites at the base of the crust is also unlikely. Lower crust-derived granulite xenoliths from SW Japan have 87Sr/86Sr values ranging from 0·7058 to 0·7102 (Kagami et al., 1993), and are not suitable as assimilants for the same reason as the granites. Thus, isotopic variability within Daisen volcano is probably mantle-derived, reflecting either isotopic variability within the magma source region associated with a single mantle diapir, or a series of diapirs with some interaction between them.
|
|
) and andesites and dacites (
) of Daisen volcano. (a) The MORB and OIB fields are from Zindler & Hart (1986) |
MAJOR AND TRACE ELEMENTS |
|---|
Plots of selected major and trace elements vs SiO2 in lavas from Daisen volcano were shown in figs 3 and 4 of Tamura et al. (2000)
. Figure 9 reproduces some of these plots, illustrating that K2O, Rb, Ba and Zr contents are similar in both basalts and andesites, which are separated by a SiO2 gap of 8 wt %.
|
Mid-ocean ridge basalt (MORB)-normalized plots of trace element data for Daisen andesite and dacite lavas are shown in Fig. 10, and the ranges of all analysed lavas are compared in Fig. 11. All lavas show subduction-zone affinities, but the andesites and dacites show a stronger depletion in Nb and a larger enrichment in Pb than the basalts (Fig. 11). La contents are similar among the basalts, andesites and dacites, but the andesites and dacites have systematically lower values of more compatible elements (Sm, Eu, Ti, Y, Yb and Lu) than the basalts (Fig. 11).
|
|
Figure 12 shows a La/Sm–Sm/Yb plot for all Daisen lavas studied. Sm/Yb values are similar amongst the basalts and andesites, but interestingly, there is a clear difference in La/Sm, suggesting hornblende fractionation. Residual garnet would be expected to cause a change in Sm/Yb. Daisen basalts have been interpreted to contain transitional signatures, suggesting melting from the garnet stability field [ocean-island basalt (OIB)-like] to the spinel peridotite field (MORB-like) at a pressure of 18 kbar (60 km) (Tamura et al., 2000
). The Daisen andesites and dacites also have the same Sm/Yb signature of residual garnet. Thus, their primary magmas were probably produced at the same general depth. Moreover, the trace element features of the andesites and dacites are consistent with hornblende fractionation at a shallower level.
|
|
DISCUSSION |
|---|
Reheating and remobilization of calc-alkaline magmas has been envisaged to have occurred in the Adamello massif, Italy (Blundy & Sparks, 1992
), the Lascar volcano, Chile (Matthews et al., 1999) and the Soufrière Hills volcano, Montserrat (Murphy et al., 2000; Couch et al., 2001; Harford & Sparks, 2001). Segregation of partial melt from restite crystals would produce a magma of rhyolitic composition. Hannah et al. (2002) showed that partial melting of previously emplaced, intermediate calc-alkaline rocks can produce the chemical compositions of the silicic group of the voluminous Tiribí Tuff (25 km3) in Central Costa Rica. Tamura & Tatsumi (2002) showed that the Izu–Bonin arc is characterized by bimodal, basalt– rhyolite, magmatism and they concluded that this rhyolite could be produced by dehydration melting of solidified hydrous calc-alkaline andesite. The aphyric andesites and phenocryst-rich dacites at Daisen volcano may represent another case relating to the genesis of calc-alkaline andesites and dacites at arc volcanoes in general.
Anti-fractionation from dacite to andesite
Phenocryst contents of Daisen lavas, ranging from zero to 40%, increase from andesite to dacite (Fig. 2). It should be emphasized again that aphyric rocks are confined to <63·5% SiO2. Ewart (1982) noted a bimodal phenocryst distribution in low- and medium-K andesites (56–63% SiO2). Most of the eruptives he studied contain 30–40% phenocrysts, but aphyric rocks (<4% phenocrysts) also form a major mode (Ewart, 1982). Ewart (1976) interpreted this bimodal distribution to be the result of fractional crystallization and/or flow differentiation processes acting in feeder conduits or in shallow magma chambers. These mechanisms, however, do not provide an adequate explanation for the positive correlation between SiO2 and modal phenocryst content in the Daisen lavas (Fig. 2).
Eruption ages of the aphyric andesites concentrate at 1 Ma and 0·5 Ma (Fig. 1; Tsukui et al., 1985; Kimura et al., 2003) and the aphyric andesites have a more mafic character than most of porphyritic rocks (Fig. 2). Given that aphyric andesite magmas are the first to be generated, crystallization of these magmas and the concurrent removal of phenocryst phases may result in phenocryst-rich dacites. This scenario is consistent neither with the petrographic observations nor mineral chemistry at Daisen volcano (Figs 3–6). Partly melted phenocrysts of plagioclase, resorbed crystals of quartz, opacitized hornblende, and high contents of Ca in orthopyroxene rims (Figs 4–6) suggest heating of the magmas, but are not consistent with fractional crystallization accompanying temperature drop. Breakdown of hornblende can also be triggered by decompression, but at Daisen volcano the differences of phenocryst assemblages between fresh hornblende-bearing rocks and opacite-bearing rocks (Table 2) indicate that heating, accompanied by dehydration, caused the breakdown of hornblende to opacite.
The phenocryst-rich dacite contains the same mineral assemblage as the coeval tephra studied by Tsukui (1985) and had similar magmatic temperatures (850–950°C). Orthopyroxene zoning patterns in two-pyroxene andesites and dacites indicate temperature rises from 800°C to 1100°C (Figs 4–6). Phenocryst-poor andesite and aphyric andesite magmas are thought to have had eruption temperatures of >1100°C, because of the continuous petrographic change from phenocryst-rich dacites to aphyric andesites (Fig. 2) as well as the melting textures of the phenocrysts they contain (Fig. 3). Tsukui (1985) showed that magmatic temperatures of dacite tephras changed cyclically from 850°C to 950°C. We suggest here that, in a broader time scale, the magmatic systems cyclically changed from 800°C to 1100°C, reflecting varying degrees of melting of the andesite protolith and resulting in corresponding changes in the composition of the erupted lavas.
The volcanic products of Daisen volcano are clearly bimodal, and the production of andesites and dacites through crystal fractionation from basalts cannot be supported by major element chemical variation trends and trace element data (Figs 9 and 11). Given the interpretation that the primary magmas of the andesites and dacites are produced at the same pressure as the basalts (60 km) in a different part of the same diapir (Tamura, 1994), the likelihood of garnet in the residual mantle would explain the garnet signature (or the transitional garnet signature) in the trace element compositions of the Daisen basalts, andesites and dacites (Fig. 12). Our contention is that mantle-derived hydrous magnesian andesite, not basalt magmas, may be parental to the calc-alkaline series rocks (Tamura, 1994; Tamura & Tatsumi, 2002).
Both andesites and dacites are fairly strongly differentiated (<3 wt % MgO), and abundant crystals of hornblende or opacites suggest a major role of hornblende in their differentiation. This is consistent with the fractionation of La from Sm and Yb and the constant Sm/Yb between basalts and andesites-plus-dacites (Fig. 12).
We suggest that a two-stage process, involving mid-crustal solidification of calc-alkaline magmas followed by their recurring partial melting, generated the magmatic trends and phenocryst zoning patterns observed at Daisen volcano. The heat required for this melting was, according to our model, supplied by the intermittent rise of subjacent basaltic magma (Tamura & Tatsumi, 2002). Figure 13 shows our model for the evolution of mantle-derived basalt and magnesian andesite in higher-level magma chambers beneath Daisen volcano. In Fig. 13b, a body of hydrous calc-alkaline magma solidifies within the crust (60% SiO2, 3% MgO). These bodies would have extensively evolved from primary magnesian andesite magmas through fractionation involving hornblende. In Fig. 13c, basalt magma is emplaced beneath a solidified andesite magma body, reheating and partially melting the andesite, and producing a phenocryst-rich dacite lava. Finally, in Fig. 13d, basalt magmas completely melt an andesite body, triggering the eruption of aphyric andesite. The sequence of lavas produced shows the reverse of normal fractionation (anti-fractionation), in the sense that the process progresses from dacitic partial melts to andesitic complete melts, and is accompanied by a temperature rise. Dacitic protoliths may not have existed below Daisen volcano, because of the more mafic character of the aphyric andesites and/or the non-existence of aphyric dacites at Daisen volcano.
|
Heat problem
Is it, however, reasonable to argue that a basalt (
1200°C) can cause melting in the crust to produce an andesite at 1100°C?
It is possible that the protolith was already hot and partially molten. Tamura & Tatsumi (2002) suggested that water-saturated andesite magma would solidify in the crust as a result of decompression, rather than as a result of the sudden drop in temperature. In this case, solidified andesite magma bodies could have retained a high temperature before subjacent basalt magmas were emplaced.
Huppert & Sparks (1988) showed that when basalt sills are emplaced into continental crust, or as envisaged here, a highly crystalline magma body, and the rocks have been preheated, a voluminous overlying layer of convecting silicic magma can be expected. For example, for a 500 m basalt sill and a crustal melting temperature of 850°C, the thickness of the silicic magma layer (950°C) generated will be 1000 m, for country rock at a temperature of 850°C prior to sill intrusion (Huppert & Sparks, 1988). The voluminous dacitic partial melts of Daisen volcano may be predicted from their model. Temperature-wise, however, an andesite complete melt (1100°C) would be more difficult to produce. Moreover, simultaneous melting and crystallization are a consequence of the fluid dynamical principles in their model, and the magmas formed may be highly porphyritic (Huppert & Sparks, 1988).
However, there are two factors in Daisen volcano that may have promoted the production of complete andesite melts. First, the country rock of Daisen volcano may act as a refractory insulating container. Isotopic evidence from Daisen volcano (Fig. 8) shows that crustal melts from the country rock did not contribute to the genesis of Daisen andesites and dacites. Therefore, heating would have been localized, so that a large amount of heat could have been transferred into a relatively small andesite body. Second, because a region of partially molten magma provides an effective density barrier, basalt sills may be repeatedly intruded into the same region during an episode of andesite remelting (Fig. 13d). As the eruption ages of the aphyric andesites are concentrated at 1 Ma and 0·5 Ma, in contrast to those of the porphyritic dacites, which range continuously from 1 to 0·02 Ma, multiple intrusions of basalt may have occurred twice in 1·3 Myr. Both refractory insulating wall rocks and repeated injections of basalt sills over a short time scale could be necessary conditions to produce sufficient, localized heat for the generation of andesitic complete melt.
|
CONCLUSIONS |
|---|
(1) The volcanic products of Daisen volcano are clearly bimodal; however, the production of andesites and dacites through crystal fractionation of basalt is not indicated by major element chemical variation trends and trace element data. Daisen volcano is made up mostly of andesite and dacite, with a SiO2 gap of
8 wt % between primitive Daisen basalts and andesites. All andesites and dacites have similar 87Sr/86Sr and 143Nd/144Nd, and isotopic variability within Daisen volcano is likely to be mantle-derived. Moreover, the Daisen andesites and dacites have the same trace element signatures as the associated basalts (Tamura et al., 2000) and were probably derived from primary magmas that were produced at the same general depth (60 km). (2) Voluminous andesite and dacite lavas contain petrographic features suggesting the reverse of normal fractionation (anti-fractionation), in the sense that magma genesis progressed from dacite to andesite, accompanied by rises in temperature. The absence of phenocrysts in the aphyric magmas suggests they were ultimately superheated in the near-surface environment.
(3) Mantle-derived hydrous magnesian andesite, generated in the same mantle diapir as coexisting basalt magma (Tamura, 1994), may be parental to the evolved hydrous calc-alkaline magmas of Daisen volcano. We suggest a two-stage process, involving mid-crustal solidification of bodies of this calc-alkaline magma, followed by varying degrees of partial melting from body to body, to produce the magmatic trends and phenocryst zoning patterns observed in the andesites and dacites. The heat required for this melting was probably supplied by the intermittent rise of subjacent basaltic magma.
|
ACKNOWLEDGEMENTS |
|---|
All field studies and some of the analytical work by Y. Tamura were carried out under the guidance of I. Kushiro, Okayama University (now at IFREE), whose help is much appreciated. H. Kagami and K. Shuto, Niigata University, are thanked for their help with isotope analyses. We particularly thank R. S. Fiske, Smithsonian Institution, and M. Handler, IFREE, for help and comments. A. Ishiwatari, H. Ishida and M. Kondo assisted with INAA analysis at Kyoto and Kanazawa Universities. H. Asada of Okayama University made the many thin sections used in this study. We thank T. H. Green and R. S. J. Sparks for their careful and insightful reviews. Part of this work was supported by a grant from the Ministry of Education, Science, Sports and Culture (Y.T.).
|
REFERENCES |
|---|
Blundy, J. D. & Sparks, R. S. J. (1992). Petrogenesis of mafic inclusions in granitoids of the Adamello massif, Italy. Journal of Petrology 33, 1039–1104.
Couch, S., Sparks, R. S. J. & Carroll, M. R. (2001). Mineral disequilibrium in lavas explained by convective self-mixing in open magma chambers. Nature 411, 1037–1039.[CrossRef][Medline]
Ewart, A. (1976). Mineralogy and chemistry of modern orogenic lavas—some statistics and implications. Earth and Planetary Science Letters 31, 417–432.[CrossRef][Web of Science]
Ewart, A. (1982). The mineralogy and petrology of Tertiary– Recent orogenic volcanic rocks: with special reference to the andesitic–basaltic compositional range. In: Thorpe, R. S. (ed.) Andesites. New York: John Wiley, pp. 25–95.
Hamamoto, T., Yuhara, M., Miyazaki, T., Fukase, M., Kondo, H., Ikawa, T., Ishioka, J., Kagami, H. & Shuto, K. (2000). Rb, Sr, Sm and Nd separation from rocks, minerals and natural water using ion-exchange resin. Science Reports of Niigata University, Series E (Geology) 15, 49–58.
Hannah, R. S., Vogel, T. A., Patino, L. C., Alvarado, G. E., Pérez, W. & Smith, D. R. (2002). Origin of silicic volcanic rocks in Central Costa Rica: a study of a chemically variable ash-flow sheet in the Tiribí Tuff. Bulletin of Volcanology 64, 117–133.[CrossRef][Web of Science]
Harford, C. L. & Sparks, R. S. J. (2001). Recent remobilisation of shallow-level intrusions on Montserrat revealed by hydrogen isotope composition of amphiboles. Earth and Planetary Science Letters 185, 285–297.[CrossRef][Web of Science]
Huppert, H. E. & Sparks, R. S. J. (1988). The generation of granitic magmas by intrusion of basalt into continental crust. Journal of Petrology 29, 599–624.
Ishiwatari, A. & Ohama, H. (1997). Clinopyroxene basalt dikes in the Miocene Iwaine Formation, Hokuriku Province, Japan: various continental arc magmas including shoshonite series and origin of the clinopyroxene phenocrysts. Journal of the Geological Society of Japan 103, 565–578.
Kagami, H., Yokose, H. & Honma, H. (1989). 87Sr/86Sr and 143Nd/144Nd ratios of GSJ rock reference samples, JB-1a, JA-1 and JG-1a. Geochemical Journal 23, 209–214.[Web of Science]
Kagami, H., Iizumi, S., Tainosho, Y. & Owada, M. (1992). Spatial variations of Sr and Nd isotope ratios of Cretaceous–Paleogene granitoid rocks, Southwest Japan Arc. Contributions to Mineralogy and Petrology 112, 165–177.[CrossRef]
Kagami, H., Iizumi, S., Iwata, M. & Nureki, T. (1993). 1. Sr–Nd isotope systematics of xenoliths in Cenozoic volcanic rocks from SW Japan. Proceedings of Japan Academy 69(B), 1–6.
Kimura, J.-I., Kunikiyo, T., Osaka, I., Nagao, T., Yamauchi, S., Kakubuchi, S., Okada, S., Fujibayashi, N., Okada, R., Murakami, H., Kusano, T., Umeda, K., Hayashi, S., Ishimaru, T., Ninomiya, A. & Tanase, A. (2003). Late Cenozoic volcanic activity in the Chugoku area, southwest Japan arc during back-arc basin opening and reinitiation of subduction. Island Arc 12, 22–45.
Lindsley, D. H. & Andersen, D. J. (1983). A two-pyroxene thermometer. Proceedings of 13th Lunar and Planetary Science Conference. Journal of Geophysical Research 88, Supplement, 887–906.[CrossRef]
Matthews, S. J., Sparks, R. S. J. & Gardeweg, M. C. (1999). The Piedras Grandes–Soncor eruptions, Lascar volcano, Chile; evolution of a zoned magma chamber in the central Andean upper crust. Journal of Petrology 40, 1891–1919.[CrossRef][Web of Science]
Miyazaki, T. & Shuto, K. (1998). Sr and Nd isotope ratios of twelve GSJ rock reference samples. Geochemical Journal 32, 345–350.[Web of Science]
Morris, P. A. (1995). Slab melting as an explanation of Quaternary volcanism and aseismicity in southwest Japan. Geology 23, 395–398.
Murphy, M. D., Sparks, R. S. J., Barclay, J., Carroll, M. R. & Brewer, T. S. (2000) Remobilization of andesite magma by intrusion of mafic magma at the Soufrière Hills volcano, Montserrat, West Indies. Journal of Petrology 41, 21–42.
Shibata, T. & Nakamura, E. (1997). Across-arc variations of isotope and trace element compositions from Quaternary basaltic volcanic rocks in northeastern Japan: implications for interaction between subducted oceanic slab and mantle wedge. Journal of Geophysical Research 102, 8051–8064.[CrossRef]
Sun, S.-s. & McDonough, W. F. (1989). Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: Saunders, A. D. & Norry, M. J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications 42, 313–345.
Tamura, Y. (1994). Genesis of island arc magmas by mantle-derived bimodal magmatism: evidence from the Shirahama Group, Japan. Journal of Petrology 35, 619–645.
Tamura, Y. & Nakamura, E. (1996). The arc lavas of the Shirahama Group, Japan: Sr and Nd isotopic data indicate mantle-derived bimodal magmatism. Journal of Petrology 37, 1307–1319.
Tamura, Y. & Tatsumi, Y. (2002). Remelting of an andesitic crust as a possible origin for rhyolitic magma in oceanic arcs: an example from the Izu–Bonin arc. Journal of Petrology 43, 1029–1047.
Tamura, Y., Yuhara, M. & Ishii, T. (2000). Primary arc basalts from Daisen volcano, Japan: equilibrium crystal fractionation versus disequilibrium fractionation during supercooling. Journal of Petrology 41, 431–448.
Tsukui, M. (1984). Geology of Daisen volcano. Journal of Geological Society of Japan 90, 643–658.
Tsukui, M. (1985). Temporal variation in chemical composition of phenocrysts and magmatic temperature at Daisen volcano, southwest Japan. Journal of Volcanology and Geothermal Research 26, 317–336.[CrossRef][Web of Science]
Tsukui, M., Nishido, H. & Nagao, K. (1985). K–Ar ages of the Hiruzen volcano group and the Daisen volcano. Journal of Geological Society of Japan 91, 279–288.
Uto, K. (1989). Neogene volcanism of southwest Japan: its time and space based on K–Ar dating. Ph.D. thesis, University of Tokyo.
Yuhara, M., Hamamoto, T., Kondo, H., Ikawa, T., Kagami, H. & Shuto, K. (2000). Rb, Sr, Sm and Nd concentrations of GSJ, KIGAM and BCR-1 rock reference samples analysed by isotope dilution method. Science Reports of Niigata University, Series E (Geology) 15, 23–34.
Zindler, A. & Hart, S. (1986). Chemical geodynamics. Annual Review of Earth and Planetary Sciences 14, 493–571.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. Tamura, J. B. Gill, D. Tollstrup, H. Kawabata, H. Shukuno, Q. Chang, T. Miyazaki, T. Takahashi, Y. Hirahara, S. Kodaira, et al. Silicic Magmas in the Izu-Bonin Oceanic Arc and Implications for Crustal Evolution J. Petrology, April 15, 2009; (2009) egp017v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Macdonald, H. E. Belkin, J. G. Fitton, N. W. Rogers, K. Nejbert, A. G. Tindle, and A. S. Marshall The Roles of Fractional Crystallization, Magma Mixing, Crystal Mush Remobilization and Volatile-Melt Interactions in the Genesis of a Young Basalt-Peralkaline Rhyolite Suite, the Greater Olkaria Volcanic Complex, Kenya Rift Valley J. Petrology, August 1, 2008; 49(8): 1515 - 1547. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. TATSUMI, T. SUZUKI, H. KAWABATA, K. SATO, T. MIYAZAKI, Q. CHANG, T. TAKAHASHI, K. TANI, T. SHIBATA, and M. YOSHIKAWA The Petrology and Geochemistry of Oto-Zan Composite Lava Flow on Shodo-Shima Island, SW Japan: Remelting of a Solidified High-Mg Andesite Magma J. Petrology, March 1, 2006; 47(3): 595 - 629. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||