Journal of Petrology Advance Access originally published online on June 6, 2008
Journal of Petrology 2008 49(7):1319-1342; doi:10.1093/petrology/egn027
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Generation of Rear-arc Magmas Induced by Influx of Slab-derived Supercritical Liquids: Implications from Alkali Basalt Lavas from Rishiri Volcano, Kurile Arc
1The Pheasant Memorial Laboratory for Geochemistry and Cosmochemistry, Institute for study of the earth's interior, Okayama University, Misasa, Tottori 682-0193, Japan
2Department of Earth and Planetary Materials Science, Graduate School of Science, Tohoku University, Sendai, Miyagi 980-8578, Japan
RECEIVED SEPTEMBER 13, 2007; ACCEPTED MAY 1, 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHY AND MINERALOGYWHOLE-ROCK COMPOSITIONDISCUSSIONCONCLUSIONAPPENDIXREFERENCES |
|---|
Magma generation processes were investigated for alkali basalt lavas from Rishiri Volcano, located at the rear of the Kurile arc, using major and trace elements and Sr, Nd, Pb and Th isotopic data. The Numaura and the Araragiyama lava flows, investigated in this study, show a significant variation in TiO2 contents (1· 0–1· 4 wt %) despite a limited variation in SiO2 content (48·5–50·0 wt %); TiO2 contents correlate positively with 143Nd/144Nd and negatively with 87Sr/86Sr, 206Pb/204Pb, 207Pb/204Pb and 208Pb/204Pb. The compositional variations of the lavas cannot be explained by magma chamber processes, such as fractional crystallization, crustal assimilation and magma mixing, and they are suggested to have formed principally during magma generation. The variation of the TiO2 contents essentially reflects a variation of the degree of partial melting (from
2 to 3%) of the source mantle, and it is inferred that the melting degree correlated positively with amounts of slab-derived materials influxed into the melting region. The melting appears to have occurred progressively under isothermal and isobaric conditions, as slab-derived materials were continuously supplied. The geochemical variations in the lavas can be explained by mixing of depleted mid-ocean ridge basalt source mantle with slab-derived materials consisting of an altered oceanic crust component and a sediment component. The slab-derived materials are likely to have contained not only Sr, Ba, Pb and U, but also significant amounts of Nd and Th that are not highly soluble in aqueous fluids. The materials are thus suggested to have been supercritical liquids, and it is suggested that magma generation occurred at depths greater than that at which supercritical liquids were decomposed into aqueous fluid and silicate melt components. The lava samples show 238U–230Th disequilibrium with 10–20% of 230Th excess; this 230Th enrichment resulted primarily from the high-Th nature of the slab-derived materials. KEY WORDS: flux melting; rear-arc magmas; slab-derived materials; supercritical liquids; 230Th excesses
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHY AND MINERALOGYWHOLE-ROCK COMPOSITIONDISCUSSIONCONCLUSIONAPPENDIXREFERENCES |
|---|
Materials and energy have been continuously transported from the Earth's interior by magmatism throughout Earth's evolution, and convergent margins have been one of the most important tectonic settings for such transport. There is now a consensus that water-rich materials released from the subducting slab play a fundamental role in the generation of arc magmas (e.g. Gill, 1981
). However, debate has continued about the details of subduction-zone processes, such as the nature and chemical compositions of slab-derived materials, processes and time-scales of material transport from the slab to the source mantle, the relative importance of decompression melting to fluid-fluxed melting, and so on (e.g. Elliott et al., 1997; Sisson & Bronto, 1998; Turner et al., 2000; Grove et al., 2002; Thomas et al., 2002; Bourdon et al., 2003; Manning, 2004; Portnyagin et al., 2007).
To assess these issues, field-based constraints, supported by high-quality geochemical data, have been accumulated by approaches such as studies on the origin of across-arc and along-arc variations in a number of volcanic arc systems (e.g. Sakuyama & Nesbitt, 1986; Ishikawa & Nakamura, 1994; Ryan et al., 1995; Hoogewerff et al., 1997; Taylor & Nesbitt, 1998; George et al., 2003; Kimura & Yoshida, 2006; Portnyagin et al., 2007) and investigations of magma genesis on lavas from a single volcano (e.g. Clark et al., 1998; Grove et al., 2002; Yokoyama et al., 2003; Handley et al., 2007). However, detailed studies on rear-arc lavas are still scarce, partly because the number of rear-arc volcanoes is limited owing to lower magma production rates in the rear-arc side than along the volcanic front. On the basis of observed correlations between magma chemistries and depths to the slab surface, it has been suggested that contributions of slab-derived materials to the generation of rear-arc magmas are significantly lower than those of the arc front magmas (e.g. Ishikawa & Nakamura, 1994); however, the magma generation processes are still not well understood (Turner et al., 2003). Generation of rear-arc magmas represents one end-member of arc magmatic processes. Therefore, rear-arc lavas have the potential to provide important information for general understanding of subduction-zone processes.
In this study, we investigate magma generation processes for alkali basalt lavas from Rishiri Volcano, located towards the rear of the Kurile arc system (in this paper we use the term ‘rear-arc’ to distinguish the setting of Rishiri Volcano from others in true back-arc basins, distinguished by rifting or spreading, such as the Marianas Trough or Lau Basin). The depth to the subducting slab is 300 km, and it is known that the Rishiri lavas have chemistries that are characteristic of island arc magmas (Nakamura et al., 1985; Shibata & Nakamura, 1997; Kuritani et al., 2005). Therefore, they can provide useful information on the generation processes of rear-arc magmas. Petrological descriptions of the lavas investigated in this study have already been given by Ishizuka & Nakagawa (1999). However, they mainly focused on the overall petrological evolution of the volcano. In this study, we examine magma generation processes in detail, including the relative contribution of fluid-fluxed melting to decompression melting, the coupling of geochemical components (i.e. depleted source mantle, subducted altered oceanic crust, and subducted sediment) in primary magmas, and the nature of the slab-derived materials. The origin of 230Th excesses observed in the lavas is also discussed.
|
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHY AND MINERALOGYWHOLE-ROCK COMPOSITIONDISCUSSIONCONCLUSIONAPPENDIXREFERENCES |
|---|
Rishiri Island belongs to the Kurile arc, and is
300 km above the Wadati–Benioff Zone (Fig. 1). The geology of Rishiri Volcano, mainly consisting of Quaternary alkali basalt and calc-alkaline andesite, has been described in detail by Kobayashi (1987) and Ishizuka (1999). The volcanic activity was divided into Early, Middle, and Late stages by Ishizuka (1999). Volcanic products of the Early stage are mainly andesitic lavas and pyroclastic-flow deposits, and dacitic lava domes. Those of the Middle stage are lava and pyroclastic-flow deposits of calc-alkaline andesite, forming the main stratovolcano. The Late stage has been subdivided into two stages; L-1 and L-2. Volcanic products of the L-1 stage are lava flows of high Na/K alkali basalt and trachytic andesite, whereas those of the L-2 stage are mainly lava flows of low Na/K alkali basalt with minor dacitic and rhyolitic pumice-fall deposits (Ishizuka, 1999).
|
The Numaura and the Araragiyama lava flows, investigated in this study, belong to the L-2 stage and erupted sequentially from the southern flank of the volcano (Fig. 1). The Araragiyama lava flows directly cover the Numaura lava flows at the southern coast of the island, indicating that the latter predates the former. These flows are also covered partly by later L-2 stage lava flows (Ishizuka, 1999
). Estimated volumes of the Numaura and the Araragiyama lavas (including scoria) are 0·03 km3 and 0·07 km3, respectively (Ishizuka, 1999). The eruption ages of the two lava flows have not been determined. However, they are much younger than 20 ka, because the Tanetomi lava flows (L-1 stage), which occurred before the activity of the Numaura lava flows, have been dated at 20·2 ± 3·1 ka (Kuritani et al., 2007) and a soil layer is present between the Tanetomi lava flows and an overlying scoria layer of the L-2 stage (Kuritani, 1995). On the other hand, the Numaura and the Araragiyama lava flows are not younger than 8·2 ka, because the main activity of the L-2 stage finished by 8·2 ka (Miura, 1995). Considering that these flows occurred in the early L-2 stage (Ishizuka, 1999), the eruption ages are much older than 8·2 ka. On the basis of these constraints, in addition to the observation that the Numaura lava flows are covered directly by the Araragiyama lava flows, the time interval between the two eruptions might have been much shorter than 12 ka, although the exact time-scale and the duration of each eruption are not clear. |
ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHY AND MINERALOGYWHOLE-ROCK COMPOSITIONDISCUSSIONCONCLUSIONAPPENDIXREFERENCES |
|---|
Whole-rock major and trace element analyses, and Sr, Nd, Pb, Th and U isotopic analyses were carried out at the Pheasant Memorial Laboratory, Institute for Study of the Earth's Interior, Okayama University at Misasa (Nakamura et al., 2003
). Samples used in this study were collected randomly from the Numaura and the Araragiyama lava flows. Trace element and isotopic analyses were undertaken under class 100 clean conditions. Rock specimens were crushed using a jaw crusher to coarse chips 3–5 mm in diameter. The chips were rinsed with deionized water in an ultrasonic bath at least three times, and then they were dried at 100°C for 12 h. The washed chips were ground using an alumina puck mill. About a half of the samples used in this study were collected from the coastal area. Therefore, to remove the effect of seawater alteration, an acid leaching step was used prior to the Sr, Nd and Pb isotope analyses for all the samples (see the Appendix). In this process, rock powders were leached with 6M HCl at 110°C for 1 h, and then they were rinsed twice with distilled water. On the other hand, so as not to cause artificial elemental fractionation, we cannot perform acid leaching for trace element analyses. Therefore, we have carefully used trace element data of samples from the coastal area (Appendix). We did not perform U and Th isotopic analyses on the coastal samples.
Concentrations of major elements, Ni and Cr were obtained by X-ray fluorescence spectrometry (XRF) on glass beads prepared by fusing with a lithium tetraborate flux (10 to 1 dilution of sample) using a Phillips PW2400 instrument (Takei, 2002). Trace elements were analyzed by inductively coupled plasma mass spectrometry (ICP-MS) using a quadrupole-type Yokogawa Agilent 7500cs system, by the methods of Makishima & Nakamura (2006) and Lu et al. (2007). The analytical procedures for chemical separation and mass spectrometry followed those of Yoshikawa & Nakamura (1993) and Makishima & Nakamura (1991) for Sr and Nd isotopic analyses, respectively. Isotopic ratios of these elements were determined by thermal ionization mass spectrometry (TIMS), using a Finnigan MAT 262. Internal fractionation correction used 86Sr/88Sr = 0·1194 and 146Nd/144Nd = 0·7219 as normalization factors. We corrected instrumental discrimination of the Sr and Nd isotopes by repeated analyses of standard materials during the same analytical campaign, and the corrected data were finally normalized to 87Sr/86Sr = 0·710240 for NIST SRM987 and 143Nd/144Nd = 0·511839 for La Jolla (Makishima & Masuda, 1994). For Pb isotopic analyses, the chemical separation method of Kuritani & Nakamura (2002) was followed, and isotopic compositions were measured by TIMS (Finnigan MAT 261), using the 207Pb–204Pb double spike method described by Kuritani & Nakamura (2003). Long-term averages of the NIST SRM981 standard in our laboratory are 206Pb/204Pb = 16·942, 207Pb/204Pb = 15·500 and 208Pb/204Pb = 36·727. Isotope analyses of U and Th were performed by TIMS (Finnigan MAT262 with RPQplus). The analytical techniques, including the procedures of sample digestion and chemical separation, were described by Yokoyama et al. (1999, 2001, 2003, 2006).
All of the major and trace element analyses were duplicated for each sample; replicate analyses always had less than 0·5% and 3–5% relative per cent difference, for major and trace elements, respectively. Analytical reproducibility (2
) for natural rock samples is typically 0·002% for 87Sr/86Sr, 0·001% for 143Nd/144Nd, 0·008%, 0·006%, and 0·006% for 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb, respectively, 0·33% for (230Th/232Th), 0·54% for Th concentration, 0·13% for (234U/238U), 0·64% for U concentration and 0·45% for (238U/232Th). Isotopic ratios in parentheses represent activity ratios throughout this paper. Decay constants of U and Th nuclides used for calculations in this study were 
238U = 1· 55125 x 10–10, 234U = 2· 8263 x 10–6, 232Th = 4·9475 x 10–11, and 230Th = 9·158 x 10–6 (Le Roux & Glendenin, 1963; Jaffey et al., 1971; Cheng et al., 2000).
Compositions of olivine phenocrysts were determined using a JEOL JXA-8800 electron microprobe, located at the Institute for Study of the Earth's Interior. An accelerating voltage of 25 kV and a beam current of 20 nA were adopted, with a counting time of 100 s. Both oxide and natural mineral standards were used, and data were obtained using a ZAF correction procedure.
|
PETROGRAPHY AND MINERALOGY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHY AND MINERALOGYWHOLE-ROCK COMPOSITIONDISCUSSIONCONCLUSIONAPPENDIXREFERENCES |
|---|
Both the Numaura and the Araragiyama lavas are mostly aphyric with phenocryst contents of about 5 vol. %. The principal phenocryst phase in both lava flows is olivine. Olivine phenocrysts, up to 3 mm in diameter, are commonly euhedral and contain spinel inclusions, but olivines with skeletal morphology are also observed. Glass inclusions are rare and they are commonly small (<10 µm). Small plagioclase crystals (< 1 mm) are sparsely present. In some Araragiyama samples (TiO2 < 1· 1 wt %; shown below), relatively large amounts of plagioclase phenocrysts occur, with modal abundance of up to 2 vol. %. They are <2 mm in length, and commonly have many small glass inclusions, showing typical features of partial dissolution (Tsuchiyama, 1985
; Nakamura & Shimakita, 1998).
Olivine phenocrysts are mostly homogeneous in Mg-number [100 x Mg/(Mg + Fe)] and Ni content except for the marginal part of the crystals. Figure 2 shows NiO contents of olivine phenocrysts at the crystal core, plotted against Mg-number. The Mg-number of olivine phenocrysts is in the range 84–86 and 84–87 in the Numaura and the Araragiyama lavas, respectively. It is notable that the compositions of olivine phenocrysts are close to those of upper mantle olivines (‘mantle olivine array’; Takahashi, 1986).
|
|
WHOLE-ROCK COMPOSITION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHY AND MINERALOGYWHOLE-ROCK COMPOSITIONDISCUSSIONCONCLUSIONAPPENDIXREFERENCES |
|---|
Compositions of major and trace elements and Sr, Nd and Pb isotopic ratios of representative samples of the Numaura and the Araragiyama lava flows, as well as those of some samples of the later L-2 stage lava flows, are reported in Table 1. Figure 3 shows Harker variation diagrams for some major oxides [TiO2, Fe2O3* (total Fe as Fe2O3), MgO, CaO, Na2O, P2O5] plotted against SiO2 content. MgO contents of all lavas are higher than 7· 5 wt %. Samples from the two lava flows can be discriminated clearly in the TiO2–SiO2 diagram, and the Numaura and the Araragiyama lavas are characterized by high and low TiO2 contents, respectively. With increasing SiO2 content, Na2O and P2O5 contents increase abruptly in the Numaura lava flows, whereas they are mostly constant in the Araragiyama lava flows. The Araragiyama samples containing plagioclase phenocrysts have low TiO2 contents (< 1· 1 wt %; crossed open squares in Fig. 3). The later L-2 stage lava flows have major element compositions distinctly different from those of the Numaura and the Araragiyama lava flows.
|
|
Figure 4 shows variation diagrams for Sr, Zr and Ba plotted against SiO2 content. The Numaura lavas are more enriched in Sr than the Araragiyama lavas, and the Sr concentration of the Araragiyama samples is mostly constant irrespective of the SiO2 content. Similar to TiO2, the Zr concentrations of the Numaura samples are significantly higher than those of the Araragiyama samples. With increasing SiO2 content, the Ba content tends to decrease significantly in the Numaura lavas, whereas it increases in the Araragiyama lavas. A normal mid-ocean ridge basalt (N-MORB) normalized multi-element concentration diagram is shown in Fig. 5. The pattern is characterized by negative anomalies of Nb and Ta, and positive spikes in Pb, Sr and Li. These features are characteristic of island arc magmas.
|
|
Variations of Sr, Nd and Pb isotopic compositions with SiO2 content are shown in Fig. 4. The isotopic compositions of the two lava flows are clearly distinctive, and the Numaura lava flows are low in 87Sr/86Sr and 206Pb/204Pb and high in 143Nd/144Nd compared with the Araragiyama lava flows. Interestingly, in the Numaura samples, 87Sr/86Sr and 206Pb/204Pb tend to decrease and 143Nd/144Nd increase with increasing SiO2 content. In contrast, in the Araragiyama samples, 87Sr/86Sr and 206Pb/204Pb correlate positively with SiO2 content.
Isotopic data for U and Th for the studied samples are listed in Table 2, and they are shown in a U–Th equiline diagram (Fig. 6). All the samples plot on the left-hand side of the equiline. The low-TiO2 Araragiyama samples have significantly lower (230Th/232Th) ratios than those of the other lava samples. In Fig. 6, compositions of basaltic lavas of the L-1 stage (Kuritani et al., 2007) are shown for comparison. The samples of the L-1 stage also show 230Th excesses, similar to those of the L-2 stage lavas. Although the eruption ages of the Araragiyama and the Numaura lava flows are not well constrained, the original (230Th/232Th) ratios of the lava samples are higher than the age-corrected (230Th/232Th) ratio of the L-1 stage basalt (Fig. 6).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHY AND MINERALOGYWHOLE-ROCK COMPOSITIONDISCUSSIONCONCLUSIONAPPENDIXREFERENCES |
|---|
The Numaura lava flows and the Araragiyama lava flows erupted sequentially from the southern flank of the volcano during the L-2 stage. However, the geochemical characteristics of these two lava flows, such as TiO2 contents and isotopic compositions, are distinctly different (Figs 3 and 4). In the following sections, the origins of the chemical variations of these lava flows are considered. Then, magma generation processes beneath Rishiri Volcano are discussed on the basis of the geochemical data of the lavas.
Origin of the Numaura magma
Whole-rock compositions of the Numaura lava samples show tight trends, although the range of variation is limited (e.g. 48·6–49·1 wt % in SiO2). The isotopic compositions of the lava samples change systematically with major element composition (Fig. 4). Therefore, one possible process to have produced the compositional trends of the lava flow is assimilation and fractional crystallization (AFC) in a crustal magma chamber. The abrupt increase of TiO2 and Na2O contents and the decrease of MgO content with increasing SiO2 content (Fig. 3) may be consistent with processes involving extensive fractionation of mafic phases such as olivine. However, this possibility is ruled out because Ba concentration tends to decrease with SiO2 content (Fig. 4c). Although plagioclase can contain significant amounts of Ba, plagioclase fractionation cannot explain the variation of Ba, because the distribution coefficient (plagioclase/melt) for Ba does not exceed unity regardless of the anorthite content (Blundy & Wood, 1991). Therefore, it is unlikely that the compositional trends were produced by processes involving extensive fractional crystallization.
The AFC hypothesis is also inconsistent with the Sr isotopic variation. The 87Sr/86Sr ratio of the lava samples tends to decrease with increasing SiO2 content (Fig. 4d). If AFC was responsible for the formation of the compositional trends of the lavas, the assimilant would have had a significantly lower 87Sr/86Sr ratio than the lava samples (< 0·7030). However, 87Sr/86Sr ratios of the basement rocks of the volcano and granodioritic rocks that are considered to be upper crustal materials are 0·7034 and 0·7061, respectively (Kobayashi, 1989; Kuritani et al., 2005). In addition, 87Sr/86Sr ratios of lower crustal xenoliths, including hornblende pyroxenite, hornblendite, hornblende gabbro and gabbro, range from 0·7031 to 0·7034 (Kobayashi, 1989). Therefore, we conclude that AFC was not the principal process to have formed the compositional trends of the lava samples.
The linear compositional variation of the lavas may be consistent with two-component mixing between a mafic magma (i.e. the least differentiated lava samples) and a felsic magma. However, if this was the case, the felsic end-member magma would have had an extremely high Na2O content (> 8 wt % when its SiO2 content was > 53 wt %) and fairly low Ba and 206Pb/204Pb. No volcanic materials with lower Ba concentration and lower 206Pb/204Pb ratio than the Numaura samples have been found in Rishiri Volcano. Therefore, the magma mixing hypothesis may also be ruled out. For these reasons, we conclude that the compositional variations of the Numaura lava flows were not produced primarily by magmatic processes in crustal magma chambers.
Origin of the Araragiyama magma
The range of the whole-rock compositional variation of the Araragiyama lava flows is somewhat larger than that of the Numaura lava flows (e.g. 48·4–50·0 wt % in SiO2), and the Araragiyama samples also form tight compositional trends. Like the Numaura samples, the Araragiyama isotopic compositions change systematically with the major element compositions (Fig. 4), which is suggestive of the formation of the trends by AFC in a crustal magma chamber. Contrasting with the Numaura lava flows, the 87Sr/86Sr ratios of the Araragiyama samples increase with SiO2 content (Fig. 4d), and this observation may be consistent with the formation of the trend by assimilation of crustal materials. The chemical variation of the lava samples is characterized by mostly constant P2O5 contents and a slight decrease of Zr concentration with increasing SiO2 content (Figs 3f and 4b). Because the distribution coefficients for these elements are lower than unity for possible fractionation phases, including Cr-rich spinel, olivine, plagioclase, clinopyroxene and hornblende, the compositional trend cannot have been formed by crystal fractionation-dominated AFC.
Next, we examine the possibility that the variation was produced by crustal assimilation-dominated AFC (the extreme case is a two-component magma mixing). In this hypothesis, the compositional variation of the Araragiyama samples was essentially produced by mixing of a low-SiO2 end-member (i.e. the least differentiated lava samples) with a high-SiO2 end-member. The high-SiO2 end-member would have had a low TiO2 content, because the TiO2 contents of the lava samples decrease abruptly with increasing SiO2 content (Figs 3a and 7a). In Fig. 7b, possible trends of olivine and hornblende fractionation are shown. The hornblende fractionation trend represents the steepest possible case (i.e. fractionation of the highest-TiO2 hornblende crystals found in Rishiri Volcano; Kobayashi, 1989), and therefore, the actual trend may be gentler than that shown in the figure. Although not shown, clinopyroxene and plagioclase fractionation trends are positive, similar to the olivine fractionation trend, because TiO2 is incompatible in these phases (e.g. Hauri et al., 1994; Bédard, 2006). If fractional crystallization occurred simultaneously with crustal assimilation, the assimilant composition must have been to the lower side of the lava trend, because the fractional crystallization trend cannot be steeper than the lava trend (Fig. 7b). If crystal fractionation was not involved, the assimilant composition would have been on the extrapolation of the lava trend (double circle). In any case, the assimilant must have been fairly low in TiO2 (< 1 wt %) and SiO2 (< 54 wt %) (Fig. 7a and b).
|
}, Compositions of experimental partial melt of garnet-free amphibolites (Rapp & Watson, 1995Now we consider the possible composition of the assimilant. Because the SiO2 contents of the basement rocks and granodioritic rocks (possible upper crustal materials) exceed 60 wt % (Kobayashi et al., 1987
; Kuritani et al., 2005), SiO2 contents of their partial melts would be much higher than 60 wt %, and thus, they are not a viable source for the assimilant. Some lower crustal materials beneath Rishiri Volcano have a low SiO2 content of 40 wt % (hornblendite; Kobayashi et al., 1987). However, partial melts of hornblendite are commonly felsic (> 55 wt % in SiO2) even when the degree of melting is 30% (e.g. Sen & Dunn, 1994; Wolf & Wyllie, 1994; Rapp & Watson, 1995). Melt composition tends to decrease in SiO2 as the melting degree increases, but the melt also becomes enriched in TiO2 (Fig. 7a). In fact, TiO2 contents of the hornblendite samples from Rishiri Volcano exceed 3 wt % (Kobayashi et al., 1987). Therefore, it might also be difficult to form the compositional trends by crustal assimilation-dominated AFC.
In Fig. 7a, whole-rock TiO2 and SiO2 contents of other Rishiri lavas (Kobayashi et al., 1987; Ishizuka & Nakagawa, 1999; Kuritani et al., 2005; Kuritani & Nakamura, 2006) are plotted. The figure shows that there are no high-SiO2 rocks on the extrapolation of the Araragiyama trend, which may suggest that such a low-SiO2 and low-TiO2 eruptible magma cannot have been produced beneath Rishiri Volcano. We conclude therefore, that here, and also in the case of the Numaura lava flows, crustal magma chambers were not the principal site for the formation of the compositional variations.
As described above, the low-TiO2 Araragiyama samples (TiO2 < 1· 1 wt %) typically contain plagioclase phenocrysts with disequilibrium textures. This observation suggests that, for these samples (crossed open squares in Figs 3 and 4), magma mixing subsequently took place, in addition to the fundamental magma generation processes that created the primary melts. We will show below that geochemical features of these samples are consistent with the magma mixing hypothesis.
Generation of heterogeneous primary magmas
We have shown that magmatic processes in crustal magma chambers were not the main cause of producing the compositional variations of the Numaura and the Araragiyama lava flows, and it is likely that the variations resulted primarily from mantle processes. One possible process responsible for the formation of the compositional trends is interaction between the magmas and the surrounding mantle peridotite. For example, magma composition can change through reactions involving olivine and orthopyroxene (e.g. Kelemen et al., 1998). However, it is difficult to produce the coupling between major element compositions and isotopic compositions (Fig. 4) solely by melt–mantle interaction, because Pb concentrations, for example, in mantle peridotite are fairly low and it may not be easy to significantly modify the isotopic compositions of the melt. Thus, it is reasonable to consider that the geochemical heterogeneity of the two lava flows was produced primarily during magma generation. This consideration is further supported by geochemical features of the lavas, as shown below.
The trace element concentrations of the lava samples is characterized by relatively flat pattern for middle and heavy rare earth elements (MREE and HREE) (Fig. 5). The (Gd/Yb)sample/(Gd/Yb)MORB ratio of the lava samples is 1· 1–1· 2 [element concentrations of MORB are from Sun & McDonough (1989)]. If the source mantle was garnet peridotite, the (Gd/Yb)sample/(Gd/Yb)MORB ratio would be much higher than unity. For example, the (Gd/Yb)melt/(Gd/Yb)MORB ratios of batch melts of garnet peridotite can be >3 when the melting degree is lower than 20% (Kelemen et al., 2003). Therefore, the source mantle is considered to have been garnet-free spinel peridotite.
Figure 8 shows elemental ratios (CaO/Al2O3 and Ti/Zr) and Na2O and Ba contents of the lava samples as a function of TiO2 content. It is well known that partial melts tend to decrease in TiO2 content with increasing degree of melting of a single source mantle peridotite, and melt TiO2 content has been used as an index of the melting degree (e.g. Stolper & Newman, 1994; Kelley et al., 2006). Except for the low-TiO2 Araragiyama samples (crossed open squares), CaO/Al2O3 of the lava samples increases with decreasing TiO2 content (Fig. 8a). Because the CaO/Al2O3 ratio of a mantle-derived partial melt generally increases with increasing degree of melting of spinel peridotite (Hirose & Kushiro, 1993; Baker & Stolper, 1994; Hirose & Kawamoto, 1995; Wasylenki et al., 2003), the observed trend is consistent with the formation of the geochemical variations by a series of melting events (melting degree: Araragiyama magma > Numaura magma). This hypothesis is also supported by the increase of Ti/Zr ratio with decreasing TiO2 content (Fig. 8b) (e.g. Woodhead et al., 1993) and lower Na2O contents of the Araragiyama samples than those of the Numaura samples (Fig. 8c). The SiO2 contents of the lava samples initially decrease and then increase as the TiO2 content decreases (Fig. 3); such features have also been produced in melting experiments (e.g. Baker & Stolper, 1994). The low-TiO2 Araragiyama samples containing plagioclase phenocrysts deviate from the main trends formed by the other samples in the CaO/Al2O3–TiO2 and Ti/Zr–TiO2 diagrams (Fig. 8a and b), which may reflect magma mixing.
|
On the basis of the ferric–ferrous ratio of the melts calculated using the model of Danyushevsky & Sobolev (1996
) and an equilibrium (Fe2+/Mg)olivine/melt distribution coefficient of 0·3 (Roeder & Emslie, 1970), the Numaura and the Araragiyama magmas (excluding the low-TiO2 samples) could have equilibrated with olivine of Mg-number 87–88, suggesting that the magmas were differentiated from primary magmas. This is also suggested by the observation that compositions of most olivine phenocrysts have less MgO and NiO than those of the mantle olivines (Fig. 2), and by the Ni concentrations measured in the whole-rock samples. Therefore, there must have been another process, in addition to mantle melting, to produce the observed compositional range. Considering that the phenocryst phase of the two lava flows is principally olivine, olivine crystals, formed as a liquidus phase in the ascending magmas, might have been fractionated from the host magmas during the ascent to the surface. Rapid growth of olivine crystals as implied by their skeletal morphology is consistent with their crystallization during magma ascent, because rapid decompression can induce skeletal growth of crystals (Nelson & Montana, 1992). Calculations show that, by addition of 2–4 wt % equilibrium olivine to single lava samples, the magmas can equilibrate with mantle olivine with Mg-number > 89.
Involvement of slab-derived materials
It is widely accepted that primary arc magmas normally consist of three geochemical components; depleted MORB source mantle (DMM), altered oceanic crust (AOC) and the overlying sediment layer (SED) and associated fluids in a subducting slab (White & Dupré, 1986; Ellam & Hawkesworth, 1988; Ishikawa & Nakamura, 1994; Elliott, 2003, and references therein). In Fig. 9, Pb isotopic compositions of the lava samples (excluding the low-TiO2 Araragiyama samples), along with those of the three geochemical components, are shown. Possible compositions of the DMM (light gray field) are from Zindler & Hart (1986), and the AOC and the SED components of Pacific Ocean crust are taken from Hauff et al. (2003). The compositions of the lava samples lie within the triangular field defined by the three components, suggesting that the magmas can indeed be represented as ternary mixtures of the DMM, AOC and SED components.
|
Contributions of AOC and SED components are also suggested from other geochemical characteristics of the lava samples. The Rishiri lavas are characterized by distinctive positive Pb and Sr spikes in the trace element concentration pattern diagram (Fig. 4), which is suggestive of involvement of fluids that could have been derived mainly from AOC (e.g. Brenan et al., 1995
; Elliott, 2003). The lava samples show a significant variation in Nd isotopic ratio (Fig. 4). It is recognized that Nd is not highly mobile in aqueous fluids (e.g. Brenan et al., 1995; Ayers et al., 1997). Therefore, a low 143Nd/144Nd component must have been present in addition to the AOC component. Figure 10 shows the relationship between Th/Nd and 143Nd/144Nd for the lava samples. The 143Nd/144Nd ratios of the samples tend to decrease with increasing Th/Nd ratio. This observation suggests that the low 143Nd/144Nd component is enriched in Th relative to Nd; such Th-enriched, low 143Nd/144Nd materials were, most plausibly, subducted oceanic sediments (e.g. Plank & Langmuir, 1993; Elliott et al., 1997; Hawkesworth et al., 1997). In fact, sediments from the Pacific Ocean have a significantly lower 143Nd/144Nd of 0·5123 (Hauff et al., 2003) than the lava samples (> 0·5129). On the other hand, the 143Nd/144Nd of AOC from the Pacific Ocean is higher ( 0·5131; Hauff et al., 2003) than that of the lava samples (< 0·5130), demonstrating that the lower 143Nd/144Nd ratios of the lava samples compared with DMM (Fig. 10) cannot be explained solely by the contribution of the AOC component.
|
Coupling of AOC and SED components
To a first-order approximation, the lavas data define a linear array in 207Pb/204Pb–206Pb/204Pb compositional space. There are three possibilities to produce the observed trend: (1) the DMM and the SED components were mixed heterogeneously, and then the mixture was mixed with the AOC component (Fig. 11a); (2) the DMM and the AOC components were mixed incompletely, and the mixture was further mixed with the SED component (Fig. 11b); (3) the AOC and the SED component was mixed homogeneously, and the mixture was subsequently mixed with the DMM component (Fig. 11c). Case (1) may be realized when mantle melting is induced by an influx of AOC fluids into source mantle (DMM) that has already been metasomatized by the SED component (e.g. Elliott et al., 1997
). Case (2) may occur when magma generation is caused by addition of the SED component to source mantle that has already been infiltrated by the AOC fluid. Case (3) may occur when melting of the source mantle (DMM) is induced by influx of materials consisting of the AOC and the SED components (e.g. Ishikawa & Nakamura, 1994; Shibata & Nakamura, 1997; Taylor & Nesbitt, 1998; Hochstaedter et al., 2001).
|
For cases (1) and (2), to form the observed linear trend (continuous bold line), the mixing ratio between the homogeneous component [AOC for case (1) or SED for case (2)] and the spatially heterogeneous mantle [the DMM–SED mixture for case (1) or the DMM–AOC mixture for case (2)] needs to vary linearly with the mantle composition. However, this mixing scenario may be less likely to occur. For example, magma generation induced by influx of a SED component into the AOC-metasomatized source mantle has been suggested in Java (Handley et al., 2007
); however, the data for the volcanic products point to a mixture between the DMM component and the AOC component, rather than to the DMM component. On the other hand, for case (3), the observed linear trend can be produced by mixing of the homogeneous DMM component with a homogenized AOC–SED mixture (note that the linear data array of the lava samples extends away from the DMM component; Fig. 9). The homogeneous mixing between the SED and the AOC components could occur naturally (Morris et al., 1990) by interaction with fluids ascending from dehydration of AOC into the overlying sediment layer or by metasomatism of AOC by fluids in equilibrium with SED beneath fore-arc regions (e.g. Usui et al., 2006; Bebout, 2007).
For these reasons, the observed linear trend is considered to have been formed primarily by mixing of the DMM component with a homogenized AOC–SED mixture; a slight heterogeneity of the DMM component is suggested as discussed below. Generation of the magmas was, therefore, induced essentially by a single-stage fluid addition event, similar to that advocated for the Izu–Bonin arc (Hochstaedter et al., 2001) and some lavas from the Aleutian chain (Jicha et al., 2004). In this case, the point at which the extension of the lava trend intersects the AOC–SED mixing line in Fig. 9 indicates the composition of slab-derived materials (SDM) consisting of AOC and SED components (Ishikawa & Nakamura, 1994; Taylor & Nesbitt, 1998).
Figure 9 demonstrates that the Araragiyama magmas contain a greater amount of the slab-derived component than the Numaura magmas, although the Araragiyama magmas are also suggested to have originated by higher degrees of melting. This is clearly illustrated in Fig. 12, which shows the variation of isotopic composition with TiO2 content. Given that the TiO2 content represents an index of the degree of partial melting of the source mantle, the isotopic composition changes systematically with the degree of melting. The 87Sr/86Sr and 206Pb/204Pb ratios tend to increase systematically with decreasing TiO2 content (Fig. 12a and c), suggesting that these ratios increase with increasing degree of melting. Similarly, the primary magmas produced by higher degrees of melting have lower 143Nd/144Nd ratios (Fig. 12b). Therefore, it is inferred that the generation of primary magmas by higher degrees of melting involves a greater amount of the slab-derived component.
|
The Pb isotopic compositions of the relatively SiO2-rich (> 49·5 wt %) Numaura samples (gray squares in Fig. 9) deviate from the linear trend in the 207Pb/204Pb–206Pb/204Pb diagram, and cannot be explained by mixing of the same end-member components (i.e. DMM and SDM) as those on the major trend. This observation suggests that either the SDM component or the DMM component was heterogeneous during the generation of the SiO2-rich Numaura magmas. Given that slab-derived materials are likely to have been chemically homogenized during transport in the wedge mantle, the source mantle, rather than the SDM, might have been heterogeneous. In this case, the curved compositional trend of the Numaura samples can be explained by gradual change of the DMM composition (from the gray circle to the filled circle in Fig. 9) during magma generation.
Nature of slab-derived materials
As discussed above, the SED and the AOC components were coupled in the slab-derived materials that triggered the generation of the Numaura and the Araragiyama magmas. The SDM component appears to have been an aqueous fluid-like material as suggested from the positive spikes of Sr, Pb and Li in the multi-element concentration diagram (Fig. 5). This is also supported by the increase of Ba concentration with increasing degree of melting (i.e. decreasing TiO2 content; Fig. 8d), because Ba is effectively transported by aqueous fluids released from the AOC, which is the major source of Ba in arc lavas (e.g. Elliott, 2003). However, the Nd isotopic variation of the lavas (Fig. 12b) requires that the SDM component contained significant amounts of a silicate melt component. Therefore, the slab-derived materials were either silicate-bearing H2O-rich fluids or supercritical liquids.
Of these possibilities, supercritical liquids may be more likely, considering that the depth to the Wadati–Benioff Zone is 300 km at Rishiri Volcano (e.g. Bureau & Keppler, 1999; Kawamoto, 2006). In fact, it has been shown that such supercritical liquids can contain significant amounts of Nd, as well as Ba, Sr and Pb (Kessel et al., 2005). During the ascent of supercritical liquids to shallow levels in the wedge mantle, they could separate into aqueous fluid and hydrous silicate melt components. However, such unmixing may not occur at deeper levels of the stability field of spinel lherzolite when the temperature is higher than 1000°C (Bureau & Keppler, 1999), and, therefore, magma generation beneath Rishiri Volcano could have been induced by influx of supercritical liquids. In a later section, we show that the involvement of supercritical liquids in the magma generation process is consistent with estimated trace element compositions of the SDM.
Flux melting beneath Rishiri Volcano
To understand the magma generation processes more quantitatively, we evaluate flux melting of the mantle based on incompatible trace element data, using the open-system melting model of Ozawa & Shimizu (1995). Then, importance of decompression-driven melting relative to fluid-fluxed melting and the compositions of slab-derived materials are evaluated.
Critical open-system melting model
As is discussed above, the geochemical heterogeneity of the two lava flows is suggested to have formed essentially by a series of progressive melting events. Therefore, we evaluate magma generation processes using the instantaneous melt model, rather than the accumulated melt model, of Ozawa & Shimizu (1995). The model considers a situation in which melting of the mantle occurs by continuous influx of fluid (or melt), and the melt generated is removed continuously from the partially melting region in a constant melt fraction. According to Ozawa & Shimizu (1995), the composition of the instantaneous melt (element i) being removed from the melting region, 
, is
|
| (1) |
is the initial concentration of the element i in the source, 
is the concentration of the element in influxing fluid, 
is the initial bulk solid partition coefficient, Pi is the bulk partition coefficient weighted for proportion of phases entering melt, 
is the mass ratio of retained melt to solid residue after the critical degree of melting (FC) is reached, and β is mass flux rate, defined as the influxing mass fraction (relative to the initial solid) divided by the degree of melting. This equation is valid when β is constant, irrespective of degree of melting. Fortunately, this condition appears to have been mostly met beneath Rishiri Volcano.
The model is applied to the samples whose Pb isotopic compositions plot on the linear trend in the 207Pb/204Pb–206Pb/204Pb diagram (i.e. black squares and open squares in Fig. 9). The samples collected from the coastal area were not used, because their Pb concentrations may have been modified by seawater alteration (see the Appendix). The source mantle may or may not have been depleted compared with typical DMM by a previous melt extraction event. Figure 13 shows the Nb/Zr ratios of the lavas, as well as the variations of the Nb/Zr ratios of partial melts of the source mantle as a function of the melting degree of the source (F) and the degree of prior melt removal from the DMM source (f), calculated according to Kelley et al. (2006). The comparison of the Nb/Zr ratios of the lavas (0·041–0·044) with the calculated results suggests that the source mantle beneath Rishiri Volcano might not have experienced a significant melt extraction event (f < 0·001; note that F was higher than 0·02 as is shown below). From this result, in the open-system melting model, the trace element compositions of the source mantle are assumed to be the representative DMM compositions of Salters & Stracke (2004) (
= 23·2 ppb). Partition coefficients of Pb between melt and minerals are from Kelemen et al. (2003). Using the pMELTS model with Adiabat_1ph program (Ghiorso et al., 2002; Smith & Asimow, 2005), the initial modal abundance of the source is estimated to be 55· 0 wt %, 27· 5 wt %, 13· 6 wt % and 3· 9 wt % for olivine, orthopyroxene, clinopyroxene and spinel, respectively, at assumed conditions of 1250°C and 2 GPa (within the stability field of spinel lherzolite). The initial bulk solid partition coefficient, 
, was then obtained from the mineral–melt partition coefficients and the initial modal abundance. PPb was calculated using the partition coefficients and the assumed melting mode of 7 wt %, 41 wt %, 30 wt % and 22 wt % for olivine, orthopyroxene, clinopyroxene and spinel, respectively, which is also obtained from the pMELTS model. The fraction of trapped melt in the source mantle peridotite is suggested to be lower than 1 wt % (e.g. Nakano & Fujii, 1989; Johnson & Dick, 1992), and a value of the parameter FC [= /( + β + 1)] of 0·01 is assumed in this study.
|
Model results
Now the remaining unknown variables are
, 
, β and F. We can estimate F and β as a function of 
for each sample, because we have information on the Pb isotopic compositions of SDM, DMM and the samples, and the initial Pb concentration of DMM. The Pb concentration of each primary magma (
) is assumed to be that of the sample, because the effect of olivine fractionation from the primary magma is smaller (2–4%) than the analytical error for Pb ( 5%). Figure 14 shows calculated variations of F and β as a function of 
for the sample Nm-43 (Table 1). The estimated degree of melting, F, does not vary greatly with 
. On the other hand, the β value depends strongly on the Pb concentration of SDM, and tends to decrease significantly with increasing 
.
|
We can estimate uniquely the values of F, β and
, if one more condition is fixed. However, β and 
are strongly coupled (β · 
constant), and β (or 
) cannot be fixed independently by mass balance constraints using Sr and Nd isotopes. Previous studies have utilized the water contents in magmas to estimate parameters regarding flux melting (e.g. Stolper & Newman, 1994; Kelley et al., 2006; Portnyagin et al., 2007). Unfortunately, we could not estimate the H2O contents in the primary magmas of the lava samples, because glass inclusions in olivine phenocrysts are rare and they are too small to permit determination of H2O contents. Kuritani (2000) estimated that the water content in a primary L-1 stage magma was 3 wt % using constraints of crystal–melt thermodynamic equilibrium. Therefore, the melting parameters are constrained assuming that this was also true for the primary Nm-43 magma. Using a water content of 3 wt %, we can constrain the β value using equation (1), given that the H2O content in SDM (possibly, supercritical liquid) was > 60 wt % (Kessel et al., 2005). Then, we obtained the upper limit of the β value to be 0·07, assuming that H2O content in DMM is 116 ppm (Salters & Stracke, 2004) and the behavior of H2O in terms of mineral–melt partitioning is similar to that of Ce (e.g. Michael, 1995; Kelley et al., 2006). The β value cannot be lower than 0·04, because the H2O content in SDM cannot exceed 100 wt %. Thus, 
is constrained to be from 30 ppm to 45 ppm (Fig. 14).
Figure 15a shows the TiO2 contents of the lava samples plotted as a function of the degree of melting calculated at a given 
of 40 ppm (i.e. the SDM compositions are assumed to have been constant throughout the generation of the Numaura and the Araragiyama magmas). Except for one sample of the Numaura lava, the F value increases systematically with decreasing TiO2 content. The result suggests that the primary Numaura and the Araragiyama magmas represent relatively small-degree (2–3%) partial melts of spinel peridotite. Figure 15b shows calculated relationships between the amount of SDM in the source and the degree of partial melting for 
of 30, 40 and 50 ppm. The amounts of SDM involved in the generation of the primary magmas tend to increase with increasing melting degree, irrespective of the compositions of the SDM. These features are similar to those reported by Stolper & Newman (1994) and Kelley et al. (2006) for Mariana Trough magmas and other back-arc basin magmas. The positive correlations between source water contents and the degrees of melting have also been reproduced by experiments and thermodynamic calculations for peridotite melting conducted at isothermal and isobaric conditions (e.g. Gaetani & Grove, 1998; Hirschmann et al., 1999). In the diagram, the calculated relationships using the parameterized hydrous mantle melting model of Katz et al. (2003) at an isothermal and isobaric condition are also shown (SDM is assumed to be H2O). The lava trends can be approximately reproduced by the model calculations (even if we use different pressure conditions from 2 GPa, the lava data can be similarly fitted at different temperature conditions from those shown in Fig. 15b). This observation may suggest that the primary Numaura and the Araragiyama magmas were produced essentially by flux melting, without a significant contribution of decompression melting.
|
We can calculate the SDM compositions from equation (1) using the mineral–melt partition coefficients of Kelemen et al. (2003
) and the compositions of Nm-43 (Table 1), because the relationship between F and β is known as a function of 
. The results of our calculations (
= 40 ppm) are listed in Table 3. It is suggested that the SDM contained appreciable amounts of Sr, Ba, Th, U and light REE (LREE), but high field strength elements (HFSE) and HREE were depleted. These features hold even when we use different 
from 40 ppm. In Table 3, the experimental slab–fluid partition coefficients of Kessel et al. (2005) are also shown. A notable feature of these data is that the partition coefficients of Ce, Nd and Th are higher than unity for supercritical liquids, whereas they are lower than unity for aqueous fluids. It is noteworthy that the SDM is estimated to have contained appreciable amounts of Ce, Nd and Th. Furthermore, for a supercritical liquid, the partition coefficients of the elements that are likely to have been depleted in the SDM (i.e. HFSE and HREE) are also lower than unity. This is consistent with the inference that the SDM was a supercritical liquid. Involvement of supercritical liquids in the genesis of rear-arc magmas has also been suggested in the Kamchatka arc, by Portnyagin et al. (2007); those workers also suggested that the slab-derived materials contained large amounts of large-ion lithophile elements, LREE and Th.
|
On the basis of the results described above, the generation of the Numaura and the Araragiyama magmas is considered to have been induced principally by influx of slab-derived supercritical liquids to the source region under isothermal and isobaric conditions (Fig. 16). Considering that the lavas record heterogeneity of the primary magmas (generated by increasing degrees of melting), the magmas are suggested to have migrated continuously upward from the melting region as soon as they were generated in response to continuing influx of the SDM (Fig. 16). This is consistent with the order of eruption of the lavas: eruption of the Numaura magmas produced by lower degrees of melting was followed by eruption of the Araragiyama magmas formed by higher degrees of melting.
|
230Th-excess signatures in the lavas
The Numaura and the Araragiyama lavas show 230Th enrichment with respect to 238U (Fig. 6). One means of producing the 230Th-excess signatures of the primary melt is magma generation in upwelling mantle (i.e. dynamic melting; McKenzie, 1985
; Williams & Gill, 1989). However, this mechanism may not have operated beneath Rishiri Volcano, because the geochemical variations of the lava samples can be principally explained by fluid-fluxed melting at an invariant point in P–T space, as discussed above.
Another possibility is magma generation by involvement of slab-derived materials having 230Th excesses. On the basis of the estimation of the SDM composition (Table 2), the (238U/232Th) is considered to have been 0·68 (Fig. 17). It is notable that the elemental ratio (i.e. U/Th) of the estimated SDM is independent of 
. Kimura & Yoshida (2006) estimated the (238U/232Th) of the slab source materials to be 0·9 for lavas of the rear NE Japan arc. If the slab source materials had a similar composition beneath the rear Kurile arc (note that Rishiri Volcano is close to the NE Japan arc; Fig. 1) and if they were in U–Th radioactive equilibrium, the initial (230Th/232Th) of the SDM would have been 0·9. Given that the transport rate of slab-derived materials is higher than 102 m/year (Turner et al., 2001; Yokoyama et al., 2003; Hawkesworth et al., 2004), the (230Th/232Th) of the SDM would have been > 0· 89. In this case, the SDM must have been significantly enriched in 230Th with respect to 238U ( 30%) (Fig. 17). Calculations show that 65% of the Th and 57% of the U in the primary magmas were derived from the SDM. Therefore, the 230Th enrichment of the SDM is inferred to have played an important role in the formation of the 230Th-excess signature of the lavas. If the melting rate was low, the same melting effect as that of dynamic melting (i.e. ingrowth of 230Th towards 238U in the residual source mantle) could have additionally contributed to the formation of the Th isotopic variation.
|
) is from Kimura & Yoshida (2006The 230Th-excess feature of the slab-derived materials beneath Rishiri Volcano contrasts with involvement of U-rich fluids in magma generation for frontal-arc lavas (e.g. Elliott et al., 1997
; Turner et al., 2000; Yokoyama et al., 2003). The difference may reflect that, with increasing pressures, the behavior of fluids released from the subducting slab changes from aqueous fluids into which U is preferentially partitioned relative to Th to supercritical liquids that can contain significant amounts of Th (Kessel et al., 2005; Portnyagin et al., 2007). In addition, the significant 230Th excesses of the Rishiri lavas resulted from the condition that the depth of magma generation was greater than that at which supercritical liquids would decompose into aqueous fluid and silicate melt components. |
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHY AND MINERALOGYWHOLE-ROCK COMPOSITIONDISCUSSIONCONCLUSIONAPPENDIXREFERENCES |
|---|
To understand the generation processes of rear-arc magmas, geochemical and petrological studies have been carried out for the Numaura and the Araragiyama lava flows from Rishiri Volcano, located at the rear of the Kurile arc. The primary magmas of these lavas are suggested to have been produced by a series of progressive melting events of mantle peridotite with the degree of melting increasing from
2 to 3%. The primary magmas appear to have been produced at an isobaric and isothermal condition, and the degree of melting increased proportionally with the amounts of slab-derived materials influxed into the melting region. In some arcs, two separate additions of slab components (i.e. an altered oceanic crust component and a sediment component) to the sub-arc mantle have been suggested (e.g. Elliott et al., 1997; Handley et al., 2007). However, beneath Rishiri Volcano, the two components are likely to have been coupled, and only single-stage transport of slab materials to the source mantle is required to explain the geochemical features of the lavas. The inference that the Nd isotopic compositions of the primary magmas change with the degree of melting requires that the slab-derived materials had significant amounts of a silicate melt component, and thus that these materials are likely to have been supercritical liquids. The slab-derived materials are estimated to have contained appreciable amounts of Sr, Ba, LREE, Pb, Th and U, and the Th-rich nature (relative to U) of the supercritical liquids resulted in the 230Th-excess signature of the lavas. |
APPENDIX |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHY AND MINERALOGYWHOLE-ROCK COMPOSITIONDISCUSSIONCONCLUSIONAPPENDIXREFERENCES |
|---|
Evaluation of the effect of seawater alteration on geochemical data
Some samples used in this study were collected from the coastal area and, unfortunately, their original trace element and isotopic compositions could have been modified by seawater alteration (e.g. Jochum & Verma, 1996
; Yokoyama et al., 2003). In this Appendix, the effects of seawater alteration on isotopic and trace element data are evaluated.
In this study, to remove the effect of seawater alteration, an acid leaching step was used prior to the Sr, Nd and Pb isotope analyses. We have also measured the Sr, Nd and Pb isotopic compositions of the unleached Sn-25 sample, and the compositions are as follows: 87Sr/86Sr = 0·703026, 143Nd/144Nd = 0·512994, 206Pb/204Pb = 18·122, 207Pb/204Pb = 15·526 and 208Pb/204Pb = 38·074. In comparison with the data for the leached sample listed in Table 1, the Sr isotopic composition of the unleached sample is slightly higher, but the difference is not significant. Nd isotopic composition does not change by the leaching process within the analytical error. On the other hand, lead isotopic composition changes significantly by the leaching process and, in particular, the 206Pb/204Pb and the 207Pb/204Pb ratios of the leached samples are much higher and lower, respectively, than those of the unleached sample. When volcanic rocks are isotopically heterogeneous on sub-mineral scales (e.g. Davidson et al., 2007), acid leaching could change the original bulk isotopic composition, because the leaching preferentially removes Sr and Pb from glass. However, this is not problematic in the studied lavas, because the lava samples are phenocryst-poor ( 5 vol. % phenocrysts) and the phenocryst phase is solely olivine with little Sr and Pb. Therefore, we believe that the leaching process removes only the effect of seawater alteration and gives the original isotopic compositions of the rocks before seawater alteration.
For trace element analyses, we did not perform acid leaching, because it causes artificial elemental fractionation. Therefore, we need to use the trace element data of the coastal samples cautiously. In Fig. A1, the trace element concentrations of the coastal samples are compared with those of the samples collected far from the coast for two pairs, Sn-14–Nm43 from the Numaura lava flows and Sn-12–Nm-13 from the Araragiyama lava flows (note that the major element compositions of the samples in each pair are mostly similar). The coastal samples are significantly (> 10%) enriched in B and Rb in both pairs. Li, Pb and U are more abundant in the coastal sample in only one of the two pairs. Interestingly, Cs is more enriched in the coastal sample in one pair, but it is more depleted in the coastal sample in the other pair. The concentrations of Sr and Nd in the coastal samples are similar to those of the samples collected far from the coast, which is consistent with the observation that Sr and Nd isotopic compositions of the leached sample are similar to those of the unleached sample. On the basis of these results, Li, B, Rb, Cs, Pb and U concentrations of the samples collected from the coastal area are not used in this paper.
|
– 
)/
, in which 
and 
are the concentration of element i in the coastal sample and the sample collected far from the coast, respectively. If the absolute value of the relative difference of element i is larger than 10% (twice the analytical error), the concentration of element i of the coastal samples is considered to have been modified by seawater alteration. |
ACKNOWLEDGEMENTS |
|---|
We thank Ryoji Tanaka and all other members of the Pheasant Memorial Laboratory (Okayama University), Takeyoshi Yoshida (Tohoku University) and Jun-ich Kimura (Shimane University) for useful discussions. Editorial handling by J. Gamble and critical improvement of the manuscript by J. Davidson, B. Bourdon and an anonymous reviewer are greatly appreciated. We are grateful to Kazuhito Ozawa for fruitful comments on the manuscript. We also acknowledge T. Moriyama, H. Kitagawa and N. Kuritani for helping T.K. collect rock samples at Rishiri Volcano. This work was supported by the Ministry of Education, Culture, Sports, Science, and Technology of the Japanese Government [Grant-in Aid for Young Scientists (B)] to T.K., and also by the program for the ‘Center of Excellence for the 21st Century in Japan’ to ISEI, Okayama University and to Graduate School of Science, Tohoku University.
|
FOOTNOTES |
|---|
Present address: Department of Earth and Planetary Sciences, Graduate School of Science and Engineering, Tokyo Institute of Technology, Tokyo 152-8551, Japan. 
*Corresponding author: Fax: +81-22-795-7760. E-mail: kuritani{at}mail.tains.tohoku.ac.jp
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHY AND MINERALOGYWHOLE-ROCK COMPOSITIONDISCUSSIONCONCLUSIONAPPENDIXREFERENCES |
|---|
Ayers JC, Dittmer SK, Layne GD. Partitioning of elements between peridotite and H2O at 2· 0–3· 0 GPa and 900–1100°C, and application to models of subduction processes. Earth and Planetary Science Letters (1997) 150:381–398.[CrossRef][Web of Science]
Baker MB, Stolper EM. Determining the composition of high-pressure mantle melts using diamond aggregates. Geochimica et Cosmochimica Acta (1994) 58:2811–2827.[CrossRef][Web of Science]
Bebout GE. Metamorphic chemical geodynamics of subduction zones. Earth and Planetary Science Letters (2007) 260:373–393.[CrossRef][Web of Science]
Bédard JH. Trace element partitioning in plagioclase feldspar. Geochimica et Cosmochimica Acta (2006) 70:3717–3742.[CrossRef][Web of Science]
Blundy JD, Wood BJ. Crystal-chemical controls on the partitioning of Sr and Ba between plagioclase feldspar, silicate melts, and hydrothermal solutions. Geochimica et Cosmochimica Acta (1991) 55:193–209.[CrossRef][Web of Science]
Bourdon B, Turner S, Dosseto A. Dehydration and partial melting in subduction zones: constraints from U-series disequilibria. Journal of Geophysical Research (2003) 108. doi:10.1029/2002JB001839.
Brenan JM, Shaw HF, Ryerson FJ, Phinney DL. Mineral–aqueous fluid partitioning of trace elements at 900°C and 2· 0 GPa: constraints on the trace element chemistry of mantle and deep crustal fluids. Geochimica et Cosmochimica Acta (1995) 59:3331–3350.[CrossRef][Web of Science]
Bureau H, Keppler H. Complete miscibility between silicate melts and hydrous fluids in the upper mantle: experimental evidence and geochemical implications. Earth and Planetary Science Letters (1999) 165:187–196.[CrossRef][Web of Science]
Cheng H, Edwards RL, Hoff J, Gallup CD, Richards DA, Asmerom Y. The half-lives of uranium-234 and thorium-230. Chemical Geology (2000) 169:17–33.[CrossRef][Web of Science]
Clark SK, Reagan MK, Plank T. Trace element and U-series systematics for 1963–1965 tephras from Irazú Volcano, Costa Rica: implications for magma generation processes and transit times. Geochimica et Cosmochimica Acta (1998) 62:2689–2699.[CrossRef][Web of Science]
Cousens BL, Allan JF. A Pb, Sr, and Nd isotopic study of basaltic rocks from the Sea of Japan, LEGS 127/128. In: Proceedings of the Ocean Drilling Program, Scientific Results, 127–128.—Tamaki K, Suyehiro K, Allan J, McWilliams M, eds. (1992) College Station, TX: Ocean Drilling Program. 805–817.
Danyushevsky LV, Sobolev AV. Ferric–ferrous ratio and oxygen fugacity calculations for primitive mantle-derived melts: calibration of an empirical technique. Mineralogy and Petrology (1996) 57:229–241.[CrossRef][Web of Science]
Davidson JP, Morgan DJ, Charlier BLA, Harlou R, Hora JM. Microsampling and isotopic analysis of igneous rocks: implications for the study of magmatic systems. Annual Review of Earth and Planetary Sciences (2007) 35:273–311.[CrossRef][Web of Science]
Ellam RM, Hawkesworth CJ. Elemental and isotopic variations in subduction related basalts: evidence for a three component model. Contributions to Mineralogy and Petrology (1988) 98:72–80.[CrossRef][Web of Science]
Elliott T. Tracers of the slab. Inside the Subduction Factory. Geophysical Monograph, American Geophysical Union—Eiler J, ed. (2003) 138:23–45.
Elliott T, Plank T, Zindler A, White W, Bourdon B. Element transport from slab to volcanic front at the Mariana arc. Journal of Geophysical Research (1997) 102:14991–15019.[CrossRef]
Gaetani GA, Grove TL. The influence of water on melting of mantle peridotite. Contributions to Mineralogy and Petrology (1998) 131:323–346.[CrossRef][Web of Science]
George R, Turner S, Hawkesworth C, Morris J, Nye C, Ryan J, Zheng S-H. Melting processes and fluid and sediment transport rates along the Alaska–Aleutian arc from an integrated U–Th–Ra–Be isotope study. Journal of Geophysical Research (2003) 108. doi:10.1029/2002JB001916.
Ghiorso MS, Hirschmann MM, Reiners PW, Kress VC. The pMELTS: a revision of MELTS for improved calculation of phase relations and major element partitioning related to partial melting of the mantle at 3 GPa. Geochemistry, Geophysics, Geosystems (2002) 3. doi:10.1029/2001GC000217.
Gill JB. Orogenic Andesites and Plate Tectonics. (1981) Heidelberg: Springer.
Grove TL, Parman SW, Bowring SA, Price RC, Baker MB. The role of an H2O-rich fluid component in the generation of primitive basaltic andesites and andesites from the Mt. Shasta region, N. California. Contributions to Mineralogy and Petrology (2002) 142:375–396.[Web of Science]
Handley HK, Macpherson CG, Davidson JP, Berlo K, Lowry D. Constraining fluid and sediment contributions to subduction-related magmatism in Indonesia: Ijen Volcanic Complex. Journal of Petrology (2007) 48:1155–1183.
Hauff F, Hoernle K, Schmidt A. Sr–Nd–Pb composition of Mesozoic Pacific oceanic crust (Site 1149 and 801, ODP Leg 185): implications for alteration of oceanic crust and the input into the Izu–Bonin–Mariana subduction system. Geochemisty, Geophysics, Geosystems (2003) 4. doi:10.1029/2002GC000421.
Hauri EH, Wagner TP, Grove TL. Experimental and natural partitioning of Th, U, Pb and other trace elements between garnet, clinopyroxene and basaltic melts. Chemical Geology (1994) 117:149–166.[CrossRef][Web of Science]
Hawkesworth CJ, Turner SP, McDermott F, Peate DW, van Calsteren P. U–Th isotopes in arc magmas: implications for element transfer from the subducted crust. Science (1997) 276:551–555.
Hawkesworth C, George R, Turner S, Zellmer G. Time scales of magmatic processes. Earth and Planetary Science Letters (2004) 218:1–16.[CrossRef][Web of Science]
Hirose K, Kawamoto T. Hydrous partial melting of lherzolite at 1 GPa: the effect of H2O on the genesis of basaltic magmas. Earth and Planetary Science Letters (1995) 133:463–473.[CrossRef][Web of Science]
Hirose K, Kushiro I. Partial melting of dry peridotites at high pressures: determination of compositions of melts segregated from peridotite using aggregates of diamond. Earth and Planetary Science Letters (1993) 114:477–489.[CrossRef][Web of Science]
Hirschmann MM, Asimow PD, Ghiorso MS, Stolper EM. Calculation of peridotite partial melting from thermodynamic models of minerals and melts. III. Controls on isobaric melt production and the effect of water on melt production. Journal of Petrology (1999) 40:831–851.[CrossRef][Web of Science]
Hochstaedter A, Gill J, Peters R, Broughton P, Holden P, Taylor B. Across-arc geochemical trends in the Izu–Bonin arc: contributions from the subducting slab. Geochemistry, Geophysics, Geosystems (2001) 2. doi:10.1029/2000GC000105.
Hoogewerff JA, van Bergen MJ, Vroon PZ, Hertogen J, Wordel R, Sneyers A, Nasution A, Varekamp JC, Moens HLE, Mouchel D. U-series, Sr–Nd–Pb isotope and trace-element systematics across an active island arc–continent collision zone: implications for element transfer at the slab–wedge interface. Geochimica et Cosmochimica Acta (1997) 61:1057–1072.[CrossRef][Web of Science]
Ishikawa T, Nakamura E. Origin of the slab component in arc lavas from across-arc variation of B and Pb isotopes. Nature (1994) 370:205–208.[CrossRef]
Ishikawa T, Tera F. Source, composition and distribution of the fluid in the Kurile mantle wedge: constraints from across-arc variations of B/Nb and B isotopes. Earth and Planetary Science Letters (1997) 152:123–138.[CrossRef][Web of Science]
Ishizuka Y. Eruptive history of Rishiri Volcano, northern Hokkaido, Japan. Bulletin of Volcanological Society of Japan (1999) 44:23–40. (in Japanese with English abstract).
Ishizuka Y, Nakagawa M. Petrological evolution of Rishiri volcano, northern Hokkaido, Japan. Journal of Mineralogy, Petrology and Economic Geology (1999) 94:279–294. (in Japanese with English abstract).[CrossRef]
Jaffey AH, Flynn KF, Glendenin LE, Bentley WC, Essling AM. Precision measurement of half-lives and specific activities of 235U and 238U. Physical Review (1971) C4:1889–1906.[CrossRef][Web of Science]
Jicha BR, Singer BS, Brophy JG, Fournelle JH, Johnson CM, Beard BL, Lapen TJ, Mahlen NJ. Variable impact of the subducted slab on Aleutian island arc magma sources: evidence from Sr, Nd, Pb, and Hf isotopes and trace element abundances. Journal of Petrology (2004) 45:1845–1875.
Jochum KP, Verma SP. Extreme enrichment of Sb, Tl and other trace elements in altered MORB. Chemical Geology (1996) 130:289–299.[CrossRef][Web of Science]
Johnson KTM, Dick HJB. Open system melting and temporal and spatial variation of peridotite and basalt at the Atlantis II fracture zone. Journal of Geophysical Research (1992) 97:9219–9241.
Katz RF, Spiegelman M, Langmuir CH. A new parameterization of hydrous mantle melting. Geochemisty, Geophysics, Geosystems (2003) 4. doi:10.1029/2002GC000433.
Kawamoto T. Hydrous phases and water transport in the subducting slab. Reviews in Mineralogy and Geochemistry (2006) 62:273–289.
Kelemen PB, Hart SR, Bernstein S. Silica enrichment in the continental upper mantle via melt/rock reaction. Earth and Planetary Science Letters (1998) 164:387–406.[CrossRef][Web of Science]
Kelemen PB, Yogodzinski GM, Scholl DW. Along-strike variation in the Aleutian island arc: genesis of high Mg# andesite and implications for continental crust. In: Inside the Subduction Factory.—Eiler J, ed. (2003) 138. Geophysical Monograph, American Geophysical Union. 223–276.
Kelley KA, Plank T, Grove TL, Stolper EM, Newman S, Hauri E. Mantle melting as a function of water content beneath back-arc basins. Journal of Geophysical Research (2006) 111. doi:10.1029/2005JB003732.
Kessel R, Schmidt MW, Ulmer P, Pettke T. Trace element signature of subduction-zone fluids, melts and supercritical liquids at 120–180 km depth. Nature (2005) 437:724–727.[CrossRef][Medline]
Kimura J, Yoshida T. Contributions of slab fluid, mantle wedge and crust to the origin of Quaternary lavas in the NE Japan arc. Journal of Petrology (2006) 47:2185–2232.
Kobayashi T. Geology of Rishiri Volcano. Journal of Geological Society of Japan (1987) 93:749–760. (in Japanese with English abstract).
Kobayashi T. Geology and petrology of Rishiri Volcano. In: Ph.D. thesis (1989) Hokkaido, Japan: Hokkaido University, Sapporo.
Kobayashi T, Yoshida T, Fukuoka T, Aoki K. Geochemical research of Rishiri Volcano, Hokkaido. Research Reports of Laboratory of Nuclear Science, Tohoku University (1987) 20:216–232. (in Japanese).
Kuritani T. Origin of alkali basalt and trachytic andesite from Rishiri Volcano. (1995) Geological Institute of the University of Tokyo. Graduate thesis.
Kuritani T. Thermal and chemical evolution of a crustal magma chamber: constraints from differential information on erupted materials. In: Ph.D. thesis (2000) Graduate School of Science, University of Tokyo.
Kuritani T, Nakamura E. Precise isotope analysis of nanogram-level Pb for natural rock samples without use of double spikes. Chemical Geology (2002) 186:31–43.[CrossRef][Web of Science]
Kuritani T, Nakamura E. Highly precise and accurate isotopic analysis of small amounts of Pb using 205Pb–204Pb and 207Pb–204Pb, two double spikes. Journal of Analytical Atomic Spectrometry (2003) 18:1464–1470.[CrossRef][Web of Science]
Kuritani T, Nakamura E. Elemental fractionation in lavas during post-eruptive degassing: evidence from trachytic lavas, Rishiri Volcano, Japan. Journal of Volcanology and Geothermal Research (2006) 149:124–138.[CrossRef][Web of Science]
Kuritani T, Kitagawa H, Nakamura E. Assimilation and fractional crystallization controlled by transport process of crustal melt: implications from an alkali basalt–dacite suite from Rishiri Volcano, Japan. Journal of Petrology (2005) 46:1421–1442.
Kuritani T, Yokoyama T, Nakamura E. Rates of thermal and chemical evolution of magmas in a cooling magma chamber: a chronological and theoretical study on basaltic and andesitic lavas from Rishiri Volcano, Japan. Journal of Petrology (2007) 48:1295–1319.
Le Roux LJ, Glendenin LE. Half-life of 232Th. In: Proceedings of the National Meeting on Nuclear Energy, Pretoria, South Africa. (1963) 83–94.
Lu Y-H, Makishima A, Nakamura E. Coprecipitation of Ti, Mo, Sn and Sb with fluorides and application to determination of B, Ti, Zr, Nb, Mo, Sn, Sb, Hf and Ta by ICP-MS. Chemical Geology (2007) 236:13–26.[CrossRef][Web of Science]
Makishima A, Masuda A. Ce isotope ratios of N-type MORB. Chemical Geology (1994) 118:1–8.[CrossRef][Web of Science]
Makishima A, Nakamura E. Precise measurement of cerium isotope composition in rock samples. Chemical Geology (1991) 94:1–11.[Web of Science]
Makishima A, Nakamura E. Determination of major, minor and trace elements in silicate samples by ICP-QMS and ICP-SFMS applying isotope dilution-internal standardisation (ID-IS) and multi-stage internal standardisasion. Geostandards and Geoanalytical Research (2006) 30:245–271.[CrossRef][Web of Science]
Manning CE. The chemistry of subduction-zone fluids. Earth and Planetary Science Letters (2004) 223:1–16.[CrossRef][Web of Science]
McKenzie D. 230Th–238U disequilibrium and the melting processes beneath ridge axes. Earth and Planetary Science Letters (1985) 72:149–157.[CrossRef][Web of Science]
Michael P. Regionally distinctive sources of depleted MORB: evidence from trace elements and H2O. Earth and Planetary Science Letters (1995) 131:301–320.[CrossRef][Web of Science]
Miura H. A proposal for the source vent of the Rishiri–Wankonosawa tephra (Rs–Wn), the late Pleistocene marker tephra in northern Hokkaido. Geographical Reports of Tokyo Metropolitan University (1995) 30:133–145.
Morris JD, Leeman WP, Tera F. The subducted component in island arc lavas: constraints from Be isotopes and B–Be systematics. Nature (1990) 344:31–36.[Medline]
Nakamura E, Campbell IH, Sun S-S. The influence of subduction processes on the geochemistry of Japanese alkaline basalts. Nature (1985) 316:55–58.[CrossRef]
Nakamura E, Makishima A, Moriguti T, Kobayashi K, Sakaguchi C, Yokoyama T, Tanaka R, Kuritani T, Takei H. Comprehensive geochemical analyses of small amounts (< 100 mg) of extraterrestrial samples for the analytical competition related to the sample return mission MUSES-C. Institute of Space Astronautical Science Report SP (2003) 16:49–101.
Nakamura M, Shimakita S. Dissolution origin and syn-entrapment compositional change of melt inclusion in plagioclase. Earth and Planetary Science Letters (1998) 161:119–133.[CrossRef][Web of Science]
Nakano T, Fujii N. The multiphase grain control percolation: its implication for a partially molten rock. Journal of Geophysical Research (1989) 94:15653–15661.[CrossRef]
Nelson ST, Montana A. Sieve-textured plagioclase in volcanic rocks produced by rapid decompression. American Mineralogist (1992) 77:1242–1249.[Abstract]
Ozawa K, Shimizu N. Open-system melting in the upper mantle: constraints from the Hayachine–Miyamori ophiolite, northeastern Japan. Journal of Geophysical Research (1995) 100:22315–22335.[CrossRef]
Plank T, Langmuir CH. Tracing trace elements from sediment input to volcanic output at subduction zones. Nature (1993) 362:739–743.[CrossRef]
Portnyagin M, Hoernle K, Plechov P, Mironov N, Khubunaya S. Constraints on mantle melting and composition and nature of slab components in volcanic arcs from volatiles (H2O, S, Cl, F) and trace elements in melt inclusions from Kamchatka Arc. Earth and Planetary Science Letters (2007) 255:53–69.[CrossRef][Web of Science]
Rapp RP, Watson EB. Dehydration melting of metabasalt at 8–32 kbar: implications for continental growth and crust–mantle recycling. Journal of Petrology (1995) 36:891–931.
Roeder PL, Emslie RF. Olivine–liquid equilibrium. Contributions to Mineralogy and Petrology (1970) 29:275–289.[CrossRef][Web of Science]
Ryan JG, Morris J, Tera F, Leeman WP, Tsvetkov A. Cross-arc geochemical variations in the Kurile arc as a function of slab depth. Science (1995) 270:625–627.
Sakuyama M, Nesbitt RW. Geochemistry of the Quaternary volcanic rocks of the Northeast Japan arc. Journal of Volcanology and Geothermal Research (1986) 29:413–450.[CrossRef][Web of Science]
Salters VJM, Stracke A. Composition of the depleted mantle. Geochemisty, Geophysics, Geosystems (2004) 5. doi:10.1029/2003GC000597.
Sen C, Dunn T. Dehydration melting of a basaltic composition amphibolites at 1· 5 and 2· 0 GPa: implications for the origin of adakites. Contributions to Mineralogy and Petrology (1994) 117:394–409.[CrossRef][Web of Science]
Shibata T, Nakamura E. 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 (1997) 102:8051–8064.[CrossRef]
Sisson TW, Bronto S. Evidence for pressure-release melting beneath magmatic arcs from basalt at Galunggung, Indonesia. Nature (1998) 391:883–886.[CrossRef]
Smith PM, Asimow PD. Adiabat_1ph: a new public front-end to the MELTS, pMELTS, and pHMELTS models. Geochemistry, Geophysics, Geosystems (2005) 6. doi:10.1029/2004GC000816.
Stolper E, Newman S. The role of water in the petrogenesis of Mariana trough magmas. Earth and Planetary Science Letters (1994) 121:293–325.[CrossRef][Web of Science]
Sun S-S, McDonough WF. Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: Magmatism in the Ocean Basins.—Saunders AD, Norry MJ, eds. (1989) 42. Geological Society, London: Special Publications. 313–345.
Takahashi E. Origin of basaltic magmas—implications from peridotite melting experiments and an olivine fractionation model. Bulletin of Volcanological Society of Japan (1986) 30:S17–S40. (in Japanese with English abstract).
Takei H. Development of precise analytical techniques for major and trace element concentrations in rock samples and their applications to the Hishikari Gold Mine, southern Kyushu, Japan. In: Ph.D. thesis (2002) Graduate School of Natural Science and Technology, Okayama University.
Taylor RN, Nesbitt RW. Isotopic characteristics of subduction fluids in an intra-oceanic setting, Izu–Bonin Arc, Japan. Earth and Planetary Science Letters (1998) 164:79–98.[CrossRef][Web of Science]
Thomas RB, Hirschmann MM, Cheng H, Reagan MK, Edwards RL. (231Pa/235U)–(230Th/238U) of young mafic volcanic rocks from Nicaragua and Costa Rica and the influence of flux melting on U-series systematics of arc lavas. Geochimica et Cosmochimica Acta (2002) 66:4287–4309.[CrossRef][Web of Science]
Tsuchiyama A. Dissolution kinetics of plagioclase in the melt of the system diopside–albite–anorthite, and origin of dusty plagioclase in andesites. Contributions to Mineralogy and Petrology (1985) 89:1–16.[CrossRef][Web of Science]
Turner S, Bourdon B, Hawkesworth C, Evans P. 226Ra–230Th evidence for multiple dehydration events, rapid melt ascent and the time scales of differentiation beneath the Tonga–Kermadec island arc. Earth and Planetary Science Letters (2000) 179:581–593.[CrossRef][Web of Science]
Turner S, Evans P, Hawkesworth C. Ultrafast source-to-surface movement of melt at island arcs from 226Ra–230Th systematics. Science (2001) 292:1363–1366.
Turner S, Bourdon B, Gill J. Insights into magma genesis at convergent margins from U-series isotopes. Uranium-Series Geochemistry. Geochemical Society, Mineralogical Society of America. Reviews in Mineralogy and Geochemistry—Bourdon B, Henderson GM, Lundstorm CC, Turner SP, eds. (2003) 52:255–315.
Usui T, Nakamura E, Helmstaedt H. Petrology and geochemistry of eclogite xenoliths from the Colorado Plateau: implications for the evolution of subducted oceanic crust. Journal of Petrology (2006) 47:929–964.
Wasylenki LE, Baker MB, Kent AJR, Stolper EM. Near-solidus melting of the shallow upper mantle: partial melting experiments on depleted peridotite. Journal of Petrology (2003) 44:1163–1191.
White WM, Dupré B. Sediment subduction and magma genesis in the Lesser Antilles: isotopic and trace element constraints. Journal of Geophysical Research (1986) 91:5927–5941.
Williams RW, Gill JB. Effects of partial melting on the uranium decay series. Geochimica et Cosmochimica Acta (1989) 53:1607–1619.[CrossRef][Web of Science]
Wolf MB, Wyllie PJ. Dehydration-melting of amphibolites at 10 kbar: the effects of temperature and time. Contributions to Mineralogy and Petrology (1994) 115:369–383.[CrossRef][Web of Science]
Woodhead J, Eggins S, Gamble J. High field strength and transition element systematics in island arc and back-arc basin basalts: evidence for multi-phase melt extraction and a depleted mantle wedge. Earth and Planetary Science Letters (1993) 114:491–504.[CrossRef][Web of Science]
Yokoyama T, Makishima A, Nakamura E. Separation of thorium and uranium from silicate rock samples using two commercial extraction chromatographic resins. Analytical Chemistry (1999) 71:135–141.
Yokoyama T, Makishima A, Nakamura E. Precise analysis of 234U/238U ratio using UO2+ ion with thermal ionization mass spectrometry for natural samples. Chemical Geology (2001) 181:1–12.[CrossRef][Web of Science]
Yokoyama T, Kobayashi K, Kuritani T, Nakamura E. Mantle metasomatism and rapid ascent of slab components beneath island arcs: evidence from 238U–230Th–226Ra disequilibria of Miyakejima volcano, Izu arc, Japan. Journal of Geophysical Research (2003) 108. doi:10.1029/2002JB002103.
Yokoyama T, Kuritani T, Kobayashi K, Nakamura E. Geochemical evolution of a shallow magma plumbing system during the last 500 years, Miyakejima volcano, Japan: constraints from 238U–230Th–226Ra systematics. Geochimica et Cosmochimica Acta (2006) 70:2885–2901.[CrossRef][Web of Science]
Yoshikawa M, Nakamura E. Precise isotope determination of trace amounts of Sr in magnesium-rich samples. Journal of Mineralogy, Petrology, and Economic Geology (1993) 88:548–561.
Zindler A, Hart S. Chemical geodynamics. Annual Review of Earth and Planetary Sciences (1986) 14:493–571.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||