Petrology of the 120 ka Caldera-Forming Eruption of Kutcharo Volcano, Eastern Hokkaido, Japan: Coexistence of Multiple Silicic Magmas and their Relationship with Mafic Magmas

Petrology of the 120 ka Caldera-Forming Eruption of Kutcharo Volcano, Eastern Hokkaido, Japan:... Abstract Petrological and geochemical examination of the largest caldera-forming eruption (Kp IV) of Kutcharo volcano, eastern Hokkaido, Japan, were undertaken in order to understand the magma genesis and eruptive processes of a large silicic magma system. The eruption started with an ash and pumice fall, followed by voluminous pyroclastic flows. Juvenile materials are mainly porphyritic pumice and a small amount of heterogeneous, nearly aphyric, scoria is contained in the pyroclastic flow deposits. On the basis of mineral, whole-rock and matrix glass chemistry, two silicic magmas (rhyolitic and dacitic) and three intermediate (andesitic) magmas were identified. Temporal variations in the whole-rock and matrix glass chemistry of the juvenile materials indicate that the activity originated from a large zoned silicic magma chamber in which dacitic magma had stagnated beneath more voluminous rhyolitic magma. During the generation of pyroclastic flows, andesitic magmas were sequentially injected into the zoned chamber, resulting in the eruption of small amounts of heterogeneous products along with voluminous silicic magmas. The rhyolite-MELTS program, mass-balance calculations, and Rayleigh fractionation models cannot explain the generation of two silicic magmas by fractional crystallization of coexisting andesitic magmas. In addition, the higher 87Sr/86Sr ratio of dacitic pumice suggests that the dacitic magma was not a parent of the rhyolitic magma. Therefore, we infer that both the rhyolitic and dacitic magmas were produced by the accumulation of interstitial melts generated from crustal materials with heterogeneous Sr isotopic compositions. It may be common that large silicic magma systems that produce caldera-forming eruptions are composed of multiple silicic magmas that are produced from an extensive area of heterogeneous crust. Mg and Ti diffusion profiles in Fe–Ti oxide phenocrysts indicate that three different andesitic magmas were successively injected into the zoned chamber hours to weeks before the eruptions. Their injection may have triggered the Kp IV eruptive activities. INTRODUCTION Understanding the magma system and eruptive processes of large, caldera-forming eruptions is important, not only to clarify the formation and modification processes of the continental crust, but also to prepare against these catastrophic events. Many studies of caldera volcanism have revealed the presence of a zoned magma chamber in which small amounts of intermediate or mafic magma are episodically injected into a predominantly silicic resident magma (e.g. Bacon & Druitt, 1988; Streck & Grunder, 1999; Milner et al., 2003; Wilson et al., 2006; Deering et al., 2011b; Wilcock et al., 2013). Studies have also investigated the petrogenetic relationships between silicic and mafic magmas (e.g. Streck & Grunder, 2008; Deering et al., 2011a) and the eruptive activities caused by mafic injections during caldera formation (e.g. Leonard et al., 2002; Milner et al., 2003). Previous studies have typically assumed that the silicic magma involved in caldera-forming eruptions is either homogeneous or has compositional and thermal gradients imposed by differentiation (e.g. Streck & Grunder, 1997; Brown et al., 1998; Wilson et al., 2006; Hildreth & Wilson, 2007; Chamberlain et al., 2015). Other studies have reported evidence of silicic magma mixing (e.g. Folkes et al., 2011; Wotzlaw et al., 2014) and the timescale of mixing (e.g. Druitt et al., 2012; Allan et al., 2013), and have proposed schematic models of magma reservoirs beneath caldera volcanoes (e.g. Hildreth, 2004; Cashman & Giordano, 2014). Several studies have documented the existence in caldera systems involving multiple, separately evolved silicic magmas (e.g. Brown et al., 1998; Reubi & Nicholls, 2005; Cooper et al., 2012; Bégué et al., 2014; Ellis et al., 2014; Wotzlaw et al., 2014). Moreover, caldera volcanoes and their associated silicic magma systems may vary in different tectonic settings (e.g. Bachmann & Bergantz, 2008; Chamberlain et al., 2015). A fuller understanding of the common processes and characteristic features of large silicic magma systems thus requires further comprehensive studies of the chemical and structural variations within the silicic magma chambers of caldera volcanoes from different tectonic settings. Kutcharo volcano, in eastern Hokkaido, Japan, is one of the Quaternary caldera volcanoes in the southern part of the Kurile arc (Fig. 1; Hasegawa & Nakagawa, 2007; Hasegawa et al., 2009, 2012). The volcano started its activity at c. 400 ka and then was quiescent until c. 200 ka (Hasegawa et al., 2011). Since 200 ka, at least eight major episodes of explosive activity have occurred (named Kp VIII to Kp I in chronological order), of which the Kp IV eruption at c. 120 ka, associated with the Kutcharo Pumice Flow Deposit IV (Machida & Arai, 2003), was the largest and most explosive. The Kp IV eruption, which produced c. 175 km3 (bulk unit volume) of eruptive products, can be divided into four phases after minor preceding activity. The Kp IV juvenile materials consist of rhyolitic pumice and a small amount of scoria. Although rhyolitic magma was predominant throughout the Kp IV eruption, small amounts of intermediate magma erupted just after they were injected into the rhyolitic magma (Hasegawa et al., 2016). Fig. 1. View largeDownload slide Index map of eastern Hokkaido, Japan (after Hasegawa et al., 2016). The Kutcharo Pumice Flow Deposit IV (Kp IV) is distributed throughout eastern Hokkaido (gray shaded areas). The broken line shows the border between the NW and SE areas detailed in Fig. 2. Fig. 1. View largeDownload slide Index map of eastern Hokkaido, Japan (after Hasegawa et al., 2016). The Kutcharo Pumice Flow Deposit IV (Kp IV) is distributed throughout eastern Hokkaido (gray shaded areas). The broken line shows the border between the NW and SE areas detailed in Fig. 2. In this study, we report the petrological and geochemical features of juvenile materials from the Kp IV eruption and document the presence of two distinct silicic magmas in the main silicic magma body. We also present evidence bearing on the petrogenetic relationships between these magmas to clarify the formation of the magma plumbing system for the caldera-forming eruption. Finally, we estimate the timescales between mixing of mafic magmas and eruption using geospeedometry for Fe–Ti oxides, to discuss the trigger of caldera-forming eruption. KP IV ERUPTION AND SAMPLES In a recent publication, we described the Kp IV eruption sequence in detail and documented a previously unrecognized small eruptive deposit (Pre-Kp IV) beneath the Kp IV eruptive deposits, that is topped by a thin paleosol (Hasegawa et al., 2016). The Kp IV eruptive products are divided into the following four units in ascending order: Unit 1, a widespread ash fall deposit (phreatoplinian eruption); Unit 2, a pumice fall deposit (subplinian eruption); Unit 3, a voluminous pyroclastic flow deposit (caldera-forming eruption); Unit 4, a small-scale scoria-rich pyroclastic flow deposit (Fig. 2). A small amount of scoria is included in Unit 3 (<0·85 km3 within 170 km3 bulk unit volume), and a greater proportion is present in Unit 4 (c. 0·5 km3 within 1 km3 bulk unit volume) (Hasegawa et al., 2016). Units 1, 2, and 3 are sequential deposits, suggesting that the eruptive activity reached a climactic phase (Unit 3) almost immediately. There is, however, a short, but well-defined, break between Unit 3 and Unit 4 (Hasegawa et al., 2016). Fig. 2. View largeDownload slide Generalized columnar section of Kp IV deposits (Hasegawa et al., 2016). Units 1 to 3 were deposited successively, whereas Unit 4 cuts into Unit 3 as shown by the truncation of gas segregation pipes. The matrix color of Unit 3 changes upsection from white to brown or black. Northwest of the caldera (NW area), Unit 3 includes a few scoria clasts, whereas Unit 4 is highly scoriaceous. In contrast, southeast of the caldera (SE area), no scoria clasts are found in Unit 3. Abbreviations: pfl, pyroclastic flow; pfa, pumice fall; wp, white pumice; bp, brown pumice; sc, scoria. Fig. 2. View largeDownload slide Generalized columnar section of Kp IV deposits (Hasegawa et al., 2016). Units 1 to 3 were deposited successively, whereas Unit 4 cuts into Unit 3 as shown by the truncation of gas segregation pipes. The matrix color of Unit 3 changes upsection from white to brown or black. Northwest of the caldera (NW area), Unit 3 includes a few scoria clasts, whereas Unit 4 is highly scoriaceous. In contrast, southeast of the caldera (SE area), no scoria clasts are found in Unit 3. Abbreviations: pfl, pyroclastic flow; pfa, pumice fall; wp, white pumice; bp, brown pumice; sc, scoria. Juvenile materials of the Kp IV eruption can be divided into white pumice, brown pumice and scoria. The white and brown pumices have indistinguishable petrographic characteristics and matrix glass chemistry (Hasegawa et al., 2016), and are simply referred to as ‘pumice’ hereafter. The scoria, our designation for heterogeneous samples including mafic portions, can be divided according to the P2O5 content of the matrix glass into low-P2O5 (LP: SiO2 = 58–74 wt % and P2O5 = 0·1–0·3 wt %), medium-P2O5 (MP: SiO2 = 52–59 wt % and P2O5 = 0·.4–0·8 wt %), and high-P2O5 (HP: SiO2 = 59–71 wt % and P2O5 = 0·3–0·9 wt %) scoria (Fig. 3). In Unit 3, HP scoria is the main component, but small amounts of LP and MP scorias are also found. In contrast, only MP scoria occurs in Unit 4. HP scoria in Unit 3 is finely inter-banded with pumice, and MP scoria in Unit 4 commonly contains silicic inclusions resembling pumice in hand specimen and thin section. Both LP and MP scorias in Unit 3 are relatively homogeneous (Hasegawa et al., 2016; see Supplementary Data; supplementary data are available for downloading at http://www.petrology.oxfordjournals.org). Fig. 3. View largeDownload slide P2O5 vs SiO2 plot for matrix glass chemistry of Kp IV juvenile materials (Hasegawa et al., 2016). Pumice has relatively homogeneous compositions, regardless of eruptive unit. Scoria exhibits wide variations, diverging toward low SiO2 and converging toward high SiO2. Scoria can be divided into three types on the basis of its P2O5 content: lower P content (LP), medium P content (MP), and higher P content (HP). The compositional gaps between them suggest the existence of three distinct magmas. Abbreviations: pm, pumice; sc, scoria. Fig. 3. View largeDownload slide P2O5 vs SiO2 plot for matrix glass chemistry of Kp IV juvenile materials (Hasegawa et al., 2016). Pumice has relatively homogeneous compositions, regardless of eruptive unit. Scoria exhibits wide variations, diverging toward low SiO2 and converging toward high SiO2. Scoria can be divided into three types on the basis of its P2O5 content: lower P content (LP), medium P content (MP), and higher P content (HP). The compositional gaps between them suggest the existence of three distinct magmas. Abbreviations: pm, pumice; sc, scoria. ANALYTICAL METHODS We collected samples from all units at representative outcrops on all sides of the Kutcharo caldera (Supplementary Data). All analyses were conducted at Hokkaido University and all the analytical results are listed in the Supplementary Data. Modal compositions were determined on thin sections using an automatic point-counter on the basis of 3000 counts per thin section. Mineral compositions were determined using a JEOL-8800R electron probe microanalyser under the following operating conditions: 15 kV accelerating voltage, 10 nA beam current for plagioclase, and 20 nA beam current for pyroxene and Fe–Ti oxides. The glass data are the same as those of Hasegawa et al. (2016), using 10 nA beam current and a 2 µm beam scanning an area 10 µm square. To avoid Na migration, X-ray data for Na were counted only for the first 30 s. All analyses were corrected using the oxide ZAF method. Whole-rock compositions were determined by X-ray fluorescence using a Spectris MagiX PRO system with a Rh tube. Major and trace elements were measured using glass beads prepared by fusing the sample with an alkali flux (a 4:1 mixture of lithium tetraborate and lithium metaborate); major and trace elements were measured in 181 samples diluted to 1:2 and only major elements were measured in 32 samples diluted to 1:10. Trace and rare earth elements (REE) were determined in 19 samples by inductively coupled plasma mass spectrometry (ICP-MS) using a Thermo Electron X series instrument, following the method of Eggins et al. (1997). Of these 19 samples, eight less silicic samples were prepared using the acid digestion method. About 50 mg of powdered sample was digested in a 2:1 mixture of HF and HClO4. After drying, the residual samples were dissolved with HNO3 and small amounts of HCl and HF. The remaining 11 silicic samples were prepared by the alkali fusion method (Roser et al., 2000). About 100 mg of powder was digested in a 1:1 mixture of HCl and HF. After drying, the residual samples were fused with 500 mg of Na2CO3 at 1100°C, then dissolved with HNO3, HCl, and a small amount of HF. Solutions for ICP-MS were prepared in 5% HNO3 and small amounts of HCl and HF with a dilution factor of 10 000 (acid digestion) and 20 000 (alkali fusion). The measurement precision and accuracy were monitored by repeated analyses of the JB-1a standard. Reproducibility of measurements was generally within less than 10% (see the Supplementary Data). Isotopic analyses of 15 samples were carried out using a Finnigan MAT262 mass spectrometer according to the methods reported by Orihashi et al. (1998). The 87Sr/86Sr and 143Nd/144Nd ratios were corrected to 86Sr/88Sr = 0·1194 and 146Nd/144Nd = 0·7219, respectively. The analytical accuracy and reproducibility in this study were monitored using the NIST-SRM987 standard for Sr and the GSJ-JNdi-1 standard for Nd. The results were 87Sr/86Sr = 0·710220 (2σ = 0·000042; n = 3) and 143Nd/144Nd = 0·512129 (2σ = 0·000019; n = 3). The house standard samples show good standard deviation values (87Sr/86Sr = 0·704239, 2σ = 0·000022, n = 3; 143Nd/144Nd = 0·512823, 2σ = 0·000021, n = 3). Pumice clasts in Unit 1 and MP and LP scoria clasts in Unit 3 were too small to determine their mineral and whole-rock chemistry (<1 cm diameter). In addition, the whole-rock chemistry results for pumice from Pre-Kp IV and Unit 2 were disregarded because the material was slightly altered (>3 wt % loss-on-ignition). PETROGRAPHY AND MINERAL CHEMISTRY Crystal contents of the pumice range from 2 to 16 vol %. The scorias contain less than 6 vol % crystals and crystallinity decreases with decreasing whole-rock SiO2 content (Fig. 4, Table 1). The crystal assemblage in all juvenile materials consists of plagioclase, orthopyroxene, Fe–Ti oxides, and minor clinopyroxene. The matrix of the pumice is glassy, varies from colorless to brownish and has a relatively homogeneous chemical composition (Hasegawa et al., 2016; Fig. 3). The groundmass of the HP scoria has a hyalopilitic texture and consists of dark brown glass and plagioclase. In both the MP and LP scoria the groundmass displays an intersertal texture and is composed of plagioclase, orthopyroxene and dark brown glass. Table 1: Representative modal compositions of Kp IV and Pre-Kp IV juvenile materials Sample No. 271 29 229 254 90 355 99 118 114 Occurrence wp wp wp bp HP sc HP sc wp MP sc MP sc Eruptive unit Pre-KpIV 2 3 3 3 3 4 4 4 Phenocryst mode (vol· %) Pl 12·0 4·4 8·2 3·1 4·0 1·5 7·4 3·2 0·3 Opx 0·5 0·1 0·2 0·3 0·4 0·9 0·5 tr 0·2 Cpx tr tr tr 0·9 tr tr 0·2 tr tr Opq 0·2 0·2 0·3 0·5 0·4 0·2 0·8 0·1 0·3 Total 12·7 4·7 8·7 4·8 4·9 2·6 8·9 3·3 0·7 Sample No. 271 29 229 254 90 355 99 118 114 Occurrence wp wp wp bp HP sc HP sc wp MP sc MP sc Eruptive unit Pre-KpIV 2 3 3 3 3 4 4 4 Phenocryst mode (vol· %) Pl 12·0 4·4 8·2 3·1 4·0 1·5 7·4 3·2 0·3 Opx 0·5 0·1 0·2 0·3 0·4 0·9 0·5 tr 0·2 Cpx tr tr tr 0·9 tr tr 0·2 tr tr Opq 0·2 0·2 0·3 0·5 0·4 0·2 0·8 0·1 0·3 Total 12·7 4·7 8·7 4·8 4·9 2·6 8·9 3·3 0·7 wp, white pumice; bp, brown pumice; sc: scoria. tr : <0·1. Phenocryst abbreviations: Pl, plagioclase; Opx, orthopyroxene; Cpx, clinopyroxene; Opq, opaques. Table 1: Representative modal compositions of Kp IV and Pre-Kp IV juvenile materials Sample No. 271 29 229 254 90 355 99 118 114 Occurrence wp wp wp bp HP sc HP sc wp MP sc MP sc Eruptive unit Pre-KpIV 2 3 3 3 3 4 4 4 Phenocryst mode (vol· %) Pl 12·0 4·4 8·2 3·1 4·0 1·5 7·4 3·2 0·3 Opx 0·5 0·1 0·2 0·3 0·4 0·9 0·5 tr 0·2 Cpx tr tr tr 0·9 tr tr 0·2 tr tr Opq 0·2 0·2 0·3 0·5 0·4 0·2 0·8 0·1 0·3 Total 12·7 4·7 8·7 4·8 4·9 2·6 8·9 3·3 0·7 Sample No. 271 29 229 254 90 355 99 118 114 Occurrence wp wp wp bp HP sc HP sc wp MP sc MP sc Eruptive unit Pre-KpIV 2 3 3 3 3 4 4 4 Phenocryst mode (vol· %) Pl 12·0 4·4 8·2 3·1 4·0 1·5 7·4 3·2 0·3 Opx 0·5 0·1 0·2 0·3 0·4 0·9 0·5 tr 0·2 Cpx tr tr tr 0·9 tr tr 0·2 tr tr Opq 0·2 0·2 0·3 0·5 0·4 0·2 0·8 0·1 0·3 Total 12·7 4·7 8·7 4·8 4·9 2·6 8·9 3·3 0·7 wp, white pumice; bp, brown pumice; sc: scoria. tr : <0·1. Phenocryst abbreviations: Pl, plagioclase; Opx, orthopyroxene; Cpx, clinopyroxene; Opq, opaques. Fig. 4. View largeDownload slide Plot of phenocryst content vs whole-rock SiO2 content. Scoria is nearly aphyric, whereas pumice is porphyritic with phenocryst content decreasing with decreasing SiO2. Abbreviations are the same as those in Fig. 2. Solid and broken lines are regression lines of juvenile materials from Unit 3 and Unit 4, respectively. The whole-rock data for pumice from Unit 2 and Pre-Kp IV are shown for reference, but were not used in our analysis because of their high loss on ignition (>3 wt %). Fig. 4. View largeDownload slide Plot of phenocryst content vs whole-rock SiO2 content. Scoria is nearly aphyric, whereas pumice is porphyritic with phenocryst content decreasing with decreasing SiO2. Abbreviations are the same as those in Fig. 2. Solid and broken lines are regression lines of juvenile materials from Unit 3 and Unit 4, respectively. The whole-rock data for pumice from Unit 2 and Pre-Kp IV are shown for reference, but were not used in our analysis because of their high loss on ignition (>3 wt %). Plagioclase Plagioclase is a major phenocryst in all samples, often forming crystal clots with Fe–Ti oxides. Phenocrysts in the Kp IV pumice are usually clear and euhedral to subhedral (<1 mm long). The anorthite contents (An = Ca × 100/(Ca + Na + K)) in the cores of the phenocrysts in pumice and scoria exhibit a clear peak at 40–44, although a few phenocrysts have higher An contents (An = 48–80; Fig. 5). Pumice samples with lower SiO2 contents have more anorthitic phenocrysts. The scoria samples contain a few grains of calcic plagioclase (An = 80–95; Fig. 5b) that are usually <0·5 mm long and surrounded by brown glass with an intersertal and hyalopilitic texture. Fig. 5. View largeDownload slide Histograms of core compositions of phenocryst minerals in (a) pumice and (b) scoria samples. Italic numbers are whole-rock SiO2 contents (wt%). The arrows indicate outlying values for Plagioclase An content and Mg/Mn in magnetite and ilmenite, as discussed in the text. Abbreviations: An = Ca × 100/(Ca + Na + K); Mg# = Mg × 100/(Fe + Mg). Fig. 5. View largeDownload slide Histograms of core compositions of phenocryst minerals in (a) pumice and (b) scoria samples. Italic numbers are whole-rock SiO2 contents (wt%). The arrows indicate outlying values for Plagioclase An content and Mg/Mn in magnetite and ilmenite, as discussed in the text. Abbreviations: An = Ca × 100/(Ca + Na + K); Mg# = Mg × 100/(Fe + Mg). The high-An plagioclase (An = 48–80) in pumice and scoria often shows normal zoning and the low-An plagioclase exhibits slight reverse zoning (Fig. 6). Most of the calcic plagioclase shows weak zoning. Fig. 6. View largeDownload slide Core and rim compositions of phenocrystic minerals in Kp IV pumice and scoria samples. In all samples, An-rich plagioclase and magnesian pyroxene show normal zoning, and An-poor plagioclase and less magnesian pyroxene show reverse zoning. High-An plagioclase phenocrysts in scoria appear to be weakly zoned. For Fe–Ti oxides, low-Mg phenocrysts in scoria exhibit slight reverse zoning and those in pumice show weak zoning. Abbreviations: N, normal zoning; R, reverse zoning. Fig. 6. View largeDownload slide Core and rim compositions of phenocrystic minerals in Kp IV pumice and scoria samples. In all samples, An-rich plagioclase and magnesian pyroxene show normal zoning, and An-poor plagioclase and less magnesian pyroxene show reverse zoning. High-An plagioclase phenocrysts in scoria appear to be weakly zoned. For Fe–Ti oxides, low-Mg phenocrysts in scoria exhibit slight reverse zoning and those in pumice show weak zoning. Abbreviations: N, normal zoning; R, reverse zoning. The core compositions of phenocrysts in the Pre-Kp IV pumice show a broad peak, distinctly different from those in Kp IV pumice, with An values of 35–58, plus a few outliers at An = 67 and 77 (Fig. 5a). Pyroxene Phenocrysts of clear, euhedral to subhedral orthopyroxene and clinopyroxene (<1 mm long) are present in all samples and often form crystal clots with Fe–Ti oxides. Orthopyroxene phenocrysts often include apatite, whereas clinopyroxene phenocrysts often contain melt inclusions. The cores of orthopyroxene phenocrysts in Kp IV pumice and scoria range in magnesium number (Mg# = Mg × 100/(Fe + Mg)) from 67 to 76, with a clear peak at Mg# = 67–69 (Fig. 5a). The more magnesian orthopyroxene phenocrysts (Mg# > 69) are often found in less silicic pumice and some scoria samples (Fig 5), and sometimes form crystal clots with magnesian clinopyroxene (Mg# = 73–78). Those more magnesian orthopyroxene phenocrysts often show normal zoning and a few show weak zoning (Fig. 6), whereas the less magnesian ones (Mg# < 69) exhibit weak to slightly reverse zoning. Clinopyroxene phenocrysts in Kp IV samples have core compositions of Mg# = 68–78, and most are around Mg# = 76 (Fig. 5). The phenocrysts with Mg# < 72 exhibit reverse zoning, whereas those with Mg# > 73 have normal to weak zoning (Fig. 6). In Pre-Kp IV pumice, orthopyroxene phenocrysts have slightly more iron-rich compositions than in Kp IV pumice (Mg# = 64–71; Fig. 5a), and clinopyroxene phenocrysts have homogeneous core compositions of Mg# = 72–74. Fe–Ti oxides Phenocrysts of Fe–Ti oxides (<0·6 mm long) are composed of Ti-magnetite and ilmenite (Fig. 5). Pumice and scoria from Kp IV include both oxide minerals as major phenocryst phases. Ti-magnetite phenocrysts in all Kp IV rocks have homogeneous core compositions in the range Mg/Mn = 2·4–2·9, although some phenocrysts in the less silicic Unit 3 pumice are more magnesian (Mg/Mn = 2·9–3·4 and 5·0–7·5 in sample No. 168; Fig. 5a). Samples of HP scoria from Unit 3 also have magnesian magnetite in the range Mg/Mn = 3–17 (Fig. 5b). MP scoria samples from Unit 4 also include a few highly magnesian phenocrysts (Mg/Mn = 11–25). In pumice samples, low-Mg magnetite (Mg/Mn = 2·4–3·4) shows weak zoning and higher-Mg magnetite (Mg/Mn > 3·4) has normal zoning (Fig. 6). In scoria, low-Mg magnetite shows weak to slightly reverse zoning. In Pre-Kp IV pumice samples, the cores of magnetite phenocrysts are slightly less magnesian than in Kp IV pumice samples (Mg/Mn = 2·3–2·7; Fig. 5a). Ilmenite phenocrysts have similar compositional variations. In Kp IV rocks their compositional range is Mg/Mn = 3·5–4·2. The less silicic pumice in Unit 3 (sample No. 168), pumice sample No. 99 in Unit 4, and some HP scoria samples from Unit 3 contain ilmenite phenocrysts with Mg/Mn = 4·3–13 (Fig. 5). In pumice samples, low-Mg phenocrysts (Mg/Mn < 5) show weak zoning and higher-Mg ones (Mg/Mn > 6) show normal zoning (Fig. 6). In scoria, the low-Mg phenocrysts (Mg/Mn < 5) exhibit weak to slightly reverse zoning and higher-Mg ones (Mg/Mn > 5) show normal zoning. In Pre-Kp IV pumice, ilmenite cores have less magnesian compositions than in Kp IV pumice samples (Mg/Mn = 3·0–3·5 and 4·2; Fig. 5a). GEOCHEMISTRY Major and trace elements The SiO2 contents of matrix glass range from 74 to 80 wt % in Kp IV pumice and from 52·5 to 74 wt % in Kp IV scoria (Fig. 3). These data do not form a single linear trend in a SiO2 vs P2O5 plot. Whole-rock SiO2 contents of Kp IV pumice samples are in the range 72·9–74·8 wt % (Fig. 7a, Table 2). Although pumice samples from Unit 3 vary more widely in composition than those from Unit 4, the data from both units form a single linear trend in SiO2 variation diagrams (Fig. 7b). Whole-rock SiO2 contents are 63·4–72 ·5 wt % in HP scoria and 58·5–70·7 wt % in MP scoria (Fig. 7a). Although both types of scoria lie on the same linear trend in SiO2 variation diagrams with FeO*, Al2O3 and Y, they form two distinct linear trends in diagrams with P2O5, V and Sr. HP scoria have higher P2O5 contents than MP scoria at all SiO2 contents, consistent with their matrix glass chemistry (Fig. 3). Although the whole-rock compositions of these scorias extend to the SiO2 range of the pumice, their linear trends converge at the dacitic rather than rhyolitic end of the pumice compositions (Fig. 7a). Fig. 7. View largeDownload slide Variation diagrams for the whole-rock compositions of Kp IV juvenile materials. Each diagram includes a 2σ error symbol. (a) Harker diagrams of representative major oxides and trace elements for all samples. Data for Kp IV juvenile materials form two linear trends on several diagrams, converging toward high SiO2 values. These trends correspond to the matrix glass compositions of the HP and MP scoria. (b) Details of selected diagrams emphasizing pumice samples. Pumice samples form single linear trends that differ from those of scoria. The scoria trends converge toward the dacitic rather than the rhyolitic end of the pumice data. (c) Ratio diagrams of incompatible elements plotted against Rb content. Ratios are approximately constant with increasing incompatible element content, except for Y/Rb. Fig. 7. View largeDownload slide Variation diagrams for the whole-rock compositions of Kp IV juvenile materials. Each diagram includes a 2σ error symbol. (a) Harker diagrams of representative major oxides and trace elements for all samples. Data for Kp IV juvenile materials form two linear trends on several diagrams, converging toward high SiO2 values. These trends correspond to the matrix glass compositions of the HP and MP scoria. (b) Details of selected diagrams emphasizing pumice samples. Pumice samples form single linear trends that differ from those of scoria. The scoria trends converge toward the dacitic rather than the rhyolitic end of the pumice data. (c) Ratio diagrams of incompatible elements plotted against Rb content. Ratios are approximately constant with increasing incompatible element content, except for Y/Rb. Incompatible element ratios differ between pumice and scoria (Fig. 7c). The Y/Rb ratios of all samples decrease with increasing Rb content in a continuous trend, whereas the Zr/Rb and Ba/Rb ratios are approximately constant with increasing Rb content. Rare earth element patterns and isotopic compositions Chondrite-normalized REE patterns of all samples display weak light-REE enrichment and a negative Eu anomaly, more so in pumice than in scoria (Fig. 8, Table 2). The REE concentrations in pumice increase with increasing whole-rock SiO2 content. The REE patterns of MP and HP scorias are sub-parallel, but the HP scoria have higher average REE concentrations. Table 2: Representative geochemical compositions of Kp IV and Pre-Kp IV juvenile materials Sample No. 271* 17* 168 150 99 229 355 85 140 297 90 114 112 118 121 110 HB-2 Occurrence pm pm pm pm pm pm sc HP sc HP sc HP sc HP sc HP sc MP sc MP sc MP sc MP sc MP Eruptive unit Pre-Kp IV 2 3 3 4 3 3 3 3 3 3 4 4 4 4 4 House standard 2σ Major elements (wt %, XRF) N=14 SiO2 73·4 73·8 73·5 74·7 75·3 75·4 63·2 64·0 65·8 69·2 71·8 58·2 60·1 64·3 67·4 69·2 53·6 0·32 TiO2 0·49 0·52 0·53 0·49 0·46 0·47 0·93 0·87 0·81 0·68 0·61 1·02 0·95 0·83 0·70 0·67 0·76 0·01 Al2O3 14·9 14·1 14·2 13·9 13·5 13·5 15·3 14·8 15·0 14·3 14·2 15·6 15·2 14·9 15·3 14·2 17·4 0·12 FeO* 2·45 2·56 2·60 2·30 2·30 2·14 6·23 5·98 5·17 4·19 3·30 9·21 8·17 6·47 4·83 4·53 9·29 0·19 MnO 0·12 0·14 0·15 0·13 0·14 0·13 0·20 0·20 0·20 0·17 0·15 0·21 0·20 0·18 0·16 0·16 0·18 0·003 MgO 0·62 0·74 0·80 0·59 0·65 0·58 2·47 2·17 1·75 1·30 1·02 3·61 3·20 2·45 1·73 1·62 6·44 0·06 CaO 2·64 2·58 2·80 2·42 2·30 2·17 5·89 5·46 4·96 3·66 3·20 7·39 6·88 5·64 4·83 4·10 9·56 0·17 Na2O 4·07 4·30 4·51 4·51 4·51 4·51 3·80 3·75 4·04 4·12 4·33 3·05 3·30 3·60 3·87 4·13 2·40 0·03 K2O 1·67 1·72 1·64 1·71 1·83 1·76 1·03 1·06 1·16 1·42 1·59 0·77 0·93 1·15 1·28 1·42 0·24 0·005 P2O5 0·10 0·12 0·13 0·10 0·11 0·10 0·62 0·54 0·46 0·31 0·21 0·42 0·42 0·34 0·25 0·25 0·11 0·002 Total 100·37 100·59 100·76 100·87 101·07 100·72 99·63 98·78 99·37 99·32 100·41 99·44 99·36 99·87 100·40 100·23 Trace elements (ppm, XRF) N=14 Sc 17·9 14·6 17·4 16·6 30·0 20·5 21·2 38·8 33·7 31·7 25·0 25·9 33·3 2·3 V 24·0 23·8 21·4 19·7 114 71·5 42·3 253 214 158 95·8 94·9 217 4·6 Cr 3·17 2·66 3·40 2·35 5·12 2·65 3·57 19·1 8·6 17·6 10·7 4·8 170·0 3·8 Ni 0·95 1·06 0·81 1·29 2·43 1·75 1·19 3·99 3·93 2·44 4·77 1·74 54·0 0·9 Rb 38·3 39·9 43·8 43·9 25·5 37·0 37·2 18·9 22·5 27·6 29·0 34·2 7·6 0·7 Zn 80·3 76·1 83·9 74·9 113 107 81·2 102 98·5 93·2 84·8 87·5 83·0 1·4 Sr 238 215 210 201 308 246 240 287 283 271 258 242 256 3·1 Y 45·5 47·2 50·1 52·6 36·0 45·3 45·7 31·8 30·9 36·5 39·8 42·3 18·4 1·3 Zr 134 143 146 151 92 127 130 71 81 98 111 119 45 1·3 Nb 3·5 4·0 3·4 4·2 2·8 3·5 3·9 2·8 2·5 3·0 3·1 3·3 3·6 0·4 Ba 512 546 548 567 347 481 497 266 306 360 410 447 100 6·9 Rare earth elements (ppm, ICP-MS) N=4 La 12·52 13·66 13·09 14·00 9·62 10·41 10·47 12·55 9·83 7·52 8·06 11·25 11·80 3·84 0·12 Ce 29·53 32·84 31·32 32·70 22·86 25·21 24·39 29·12 24·28 18·00 20·00 26·13 28·06 9·58 0·30 Pr 4·14 4·51 4·33 4·60 3·37 3·61 3·57 4·21 3·32 2·61 2·86 3·76 3·97 1·45 0·07 Nd 18·57 20·63 19·87 20·75 15·90 17·08 16·66 19·53 14·94 12·62 13·32 17·56 18·02 7·08 0·30 Sm 5·16 5·60 5·27 5·70 4·54 4·81 4·642 5·44 4·11 3·65 3·84 4·90 4·96 2·13 0·14 Eu 1·42 1·57 1·47 1·47 1·45 1·48 1·47 1·56 1·22 1·19 1·19 1·42 1·46 0·81 0·05 Gd 5·78 6·38 6·22 6·50 5·31 5·63 5·28 6·26 4·46 4·36 4·36 5·56 5·80 2·66 0·13 Tb 1·02 1·08 1·08 1·15 0·92 0·98 0·93 1·10 0·79 0·76 0·78 0·96 1·02 0·47 0·02 Dy 6·47 7·15 7·04 7·36 5·80 6·15 5·84 7·00 4·98 4·79 4·99 6·10 6·47 2·96 0·13 Ho 1·47 1·58 1·58 1·65 1·28 1·37 1·30 1·59 1·11 1·08 1·10 1·35 1·45 0·64 0·02 Er 4·35 4·82 4·74 4·99 3·80 4·07 3·82 4·73 3·37 3·16 3·32 4·00 4·37 1·86 0·07 Tm 0·67 0·73 0·75 0·80 0·57 0·61 0·59 0·73 0·52 0·48 0·51 0·61 0·66 0·28 0·01 Yb 4·59 4·86 4·87 5·30 3·60 3·97 3·79 4·67 3·47 3·08 3·31 3·92 4·42 1·75 0·06 Lu 0·72 0·77 0·78 0·81 0·57 0·62 0·59 0·73 0·54 0·47 0·51 0·62 0·69 0·28 0·01 Isotope ratios (TIMS) N=3 87Sr/86Sr 0·70332 0·70328 0·70327 0·70327 0·70332 0·70332 0·70335 0·70332 0·70336 0·70331 0·70338 0·70330 0·70331 0·70424 2·15E-05 2SE 1·13E-05 7·29E-06 8·34E-06 8·02E-06 1·25E-05 6·34E-06 1·58E-05 8·39E-06 1·53E-05 7·78E-06 1·09E-05 7·91E-06 7·54E-06 143Nd/144Nd 0·51301 0·51301 0·51302 0·51301 0·51301 0·51298 0·51300 0·51302 0·51302 0·51301 0·51301 0·51301 0·51302 0·51282 2·08E-05 2SE 7·14E-06 9·39E-06 7·83E-06 7·18E-06 6·00E-06 9·05E-06 1·08E-05 9·07E-06 9·29E-06 8·14E-06 1·09E-05 1·02E-05 7·72E-06 Sample No. 271* 17* 168 150 99 229 355 85 140 297 90 114 112 118 121 110 HB-2 Occurrence pm pm pm pm pm pm sc HP sc HP sc HP sc HP sc HP sc MP sc MP sc MP sc MP sc MP Eruptive unit Pre-Kp IV 2 3 3 4 3 3 3 3 3 3 4 4 4 4 4 House standard 2σ Major elements (wt %, XRF) N=14 SiO2 73·4 73·8 73·5 74·7 75·3 75·4 63·2 64·0 65·8 69·2 71·8 58·2 60·1 64·3 67·4 69·2 53·6 0·32 TiO2 0·49 0·52 0·53 0·49 0·46 0·47 0·93 0·87 0·81 0·68 0·61 1·02 0·95 0·83 0·70 0·67 0·76 0·01 Al2O3 14·9 14·1 14·2 13·9 13·5 13·5 15·3 14·8 15·0 14·3 14·2 15·6 15·2 14·9 15·3 14·2 17·4 0·12 FeO* 2·45 2·56 2·60 2·30 2·30 2·14 6·23 5·98 5·17 4·19 3·30 9·21 8·17 6·47 4·83 4·53 9·29 0·19 MnO 0·12 0·14 0·15 0·13 0·14 0·13 0·20 0·20 0·20 0·17 0·15 0·21 0·20 0·18 0·16 0·16 0·18 0·003 MgO 0·62 0·74 0·80 0·59 0·65 0·58 2·47 2·17 1·75 1·30 1·02 3·61 3·20 2·45 1·73 1·62 6·44 0·06 CaO 2·64 2·58 2·80 2·42 2·30 2·17 5·89 5·46 4·96 3·66 3·20 7·39 6·88 5·64 4·83 4·10 9·56 0·17 Na2O 4·07 4·30 4·51 4·51 4·51 4·51 3·80 3·75 4·04 4·12 4·33 3·05 3·30 3·60 3·87 4·13 2·40 0·03 K2O 1·67 1·72 1·64 1·71 1·83 1·76 1·03 1·06 1·16 1·42 1·59 0·77 0·93 1·15 1·28 1·42 0·24 0·005 P2O5 0·10 0·12 0·13 0·10 0·11 0·10 0·62 0·54 0·46 0·31 0·21 0·42 0·42 0·34 0·25 0·25 0·11 0·002 Total 100·37 100·59 100·76 100·87 101·07 100·72 99·63 98·78 99·37 99·32 100·41 99·44 99·36 99·87 100·40 100·23 Trace elements (ppm, XRF) N=14 Sc 17·9 14·6 17·4 16·6 30·0 20·5 21·2 38·8 33·7 31·7 25·0 25·9 33·3 2·3 V 24·0 23·8 21·4 19·7 114 71·5 42·3 253 214 158 95·8 94·9 217 4·6 Cr 3·17 2·66 3·40 2·35 5·12 2·65 3·57 19·1 8·6 17·6 10·7 4·8 170·0 3·8 Ni 0·95 1·06 0·81 1·29 2·43 1·75 1·19 3·99 3·93 2·44 4·77 1·74 54·0 0·9 Rb 38·3 39·9 43·8 43·9 25·5 37·0 37·2 18·9 22·5 27·6 29·0 34·2 7·6 0·7 Zn 80·3 76·1 83·9 74·9 113 107 81·2 102 98·5 93·2 84·8 87·5 83·0 1·4 Sr 238 215 210 201 308 246 240 287 283 271 258 242 256 3·1 Y 45·5 47·2 50·1 52·6 36·0 45·3 45·7 31·8 30·9 36·5 39·8 42·3 18·4 1·3 Zr 134 143 146 151 92 127 130 71 81 98 111 119 45 1·3 Nb 3·5 4·0 3·4 4·2 2·8 3·5 3·9 2·8 2·5 3·0 3·1 3·3 3·6 0·4 Ba 512 546 548 567 347 481 497 266 306 360 410 447 100 6·9 Rare earth elements (ppm, ICP-MS) N=4 La 12·52 13·66 13·09 14·00 9·62 10·41 10·47 12·55 9·83 7·52 8·06 11·25 11·80 3·84 0·12 Ce 29·53 32·84 31·32 32·70 22·86 25·21 24·39 29·12 24·28 18·00 20·00 26·13 28·06 9·58 0·30 Pr 4·14 4·51 4·33 4·60 3·37 3·61 3·57 4·21 3·32 2·61 2·86 3·76 3·97 1·45 0·07 Nd 18·57 20·63 19·87 20·75 15·90 17·08 16·66 19·53 14·94 12·62 13·32 17·56 18·02 7·08 0·30 Sm 5·16 5·60 5·27 5·70 4·54 4·81 4·642 5·44 4·11 3·65 3·84 4·90 4·96 2·13 0·14 Eu 1·42 1·57 1·47 1·47 1·45 1·48 1·47 1·56 1·22 1·19 1·19 1·42 1·46 0·81 0·05 Gd 5·78 6·38 6·22 6·50 5·31 5·63 5·28 6·26 4·46 4·36 4·36 5·56 5·80 2·66 0·13 Tb 1·02 1·08 1·08 1·15 0·92 0·98 0·93 1·10 0·79 0·76 0·78 0·96 1·02 0·47 0·02 Dy 6·47 7·15 7·04 7·36 5·80 6·15 5·84 7·00 4·98 4·79 4·99 6·10 6·47 2·96 0·13 Ho 1·47 1·58 1·58 1·65 1·28 1·37 1·30 1·59 1·11 1·08 1·10 1·35 1·45 0·64 0·02 Er 4·35 4·82 4·74 4·99 3·80 4·07 3·82 4·73 3·37 3·16 3·32 4·00 4·37 1·86 0·07 Tm 0·67 0·73 0·75 0·80 0·57 0·61 0·59 0·73 0·52 0·48 0·51 0·61 0·66 0·28 0·01 Yb 4·59 4·86 4·87 5·30 3·60 3·97 3·79 4·67 3·47 3·08 3·31 3·92 4·42 1·75 0·06 Lu 0·72 0·77 0·78 0·81 0·57 0·62 0·59 0·73 0·54 0·47 0·51 0·62 0·69 0·28 0·01 Isotope ratios (TIMS) N=3 87Sr/86Sr 0·70332 0·70328 0·70327 0·70327 0·70332 0·70332 0·70335 0·70332 0·70336 0·70331 0·70338 0·70330 0·70331 0·70424 2·15E-05 2SE 1·13E-05 7·29E-06 8·34E-06 8·02E-06 1·25E-05 6·34E-06 1·58E-05 8·39E-06 1·53E-05 7·78E-06 1·09E-05 7·91E-06 7·54E-06 143Nd/144Nd 0·51301 0·51301 0·51302 0·51301 0·51301 0·51298 0·51300 0·51302 0·51302 0·51301 0·51301 0·51301 0·51302 0·51282 2·08E-05 2SE 7·14E-06 9·39E-06 7·83E-06 7·18E-06 6·00E-06 9·05E-06 1·08E-05 9·07E-06 9·29E-06 8·14E-06 1·09E-05 1·02E-05 7·72E-06 * pm, pumice; sc: scoria. These samples are only for reference because of higher LOI (>3 wt %). Table 2: Representative geochemical compositions of Kp IV and Pre-Kp IV juvenile materials Sample No. 271* 17* 168 150 99 229 355 85 140 297 90 114 112 118 121 110 HB-2 Occurrence pm pm pm pm pm pm sc HP sc HP sc HP sc HP sc HP sc MP sc MP sc MP sc MP sc MP Eruptive unit Pre-Kp IV 2 3 3 4 3 3 3 3 3 3 4 4 4 4 4 House standard 2σ Major elements (wt %, XRF) N=14 SiO2 73·4 73·8 73·5 74·7 75·3 75·4 63·2 64·0 65·8 69·2 71·8 58·2 60·1 64·3 67·4 69·2 53·6 0·32 TiO2 0·49 0·52 0·53 0·49 0·46 0·47 0·93 0·87 0·81 0·68 0·61 1·02 0·95 0·83 0·70 0·67 0·76 0·01 Al2O3 14·9 14·1 14·2 13·9 13·5 13·5 15·3 14·8 15·0 14·3 14·2 15·6 15·2 14·9 15·3 14·2 17·4 0·12 FeO* 2·45 2·56 2·60 2·30 2·30 2·14 6·23 5·98 5·17 4·19 3·30 9·21 8·17 6·47 4·83 4·53 9·29 0·19 MnO 0·12 0·14 0·15 0·13 0·14 0·13 0·20 0·20 0·20 0·17 0·15 0·21 0·20 0·18 0·16 0·16 0·18 0·003 MgO 0·62 0·74 0·80 0·59 0·65 0·58 2·47 2·17 1·75 1·30 1·02 3·61 3·20 2·45 1·73 1·62 6·44 0·06 CaO 2·64 2·58 2·80 2·42 2·30 2·17 5·89 5·46 4·96 3·66 3·20 7·39 6·88 5·64 4·83 4·10 9·56 0·17 Na2O 4·07 4·30 4·51 4·51 4·51 4·51 3·80 3·75 4·04 4·12 4·33 3·05 3·30 3·60 3·87 4·13 2·40 0·03 K2O 1·67 1·72 1·64 1·71 1·83 1·76 1·03 1·06 1·16 1·42 1·59 0·77 0·93 1·15 1·28 1·42 0·24 0·005 P2O5 0·10 0·12 0·13 0·10 0·11 0·10 0·62 0·54 0·46 0·31 0·21 0·42 0·42 0·34 0·25 0·25 0·11 0·002 Total 100·37 100·59 100·76 100·87 101·07 100·72 99·63 98·78 99·37 99·32 100·41 99·44 99·36 99·87 100·40 100·23 Trace elements (ppm, XRF) N=14 Sc 17·9 14·6 17·4 16·6 30·0 20·5 21·2 38·8 33·7 31·7 25·0 25·9 33·3 2·3 V 24·0 23·8 21·4 19·7 114 71·5 42·3 253 214 158 95·8 94·9 217 4·6 Cr 3·17 2·66 3·40 2·35 5·12 2·65 3·57 19·1 8·6 17·6 10·7 4·8 170·0 3·8 Ni 0·95 1·06 0·81 1·29 2·43 1·75 1·19 3·99 3·93 2·44 4·77 1·74 54·0 0·9 Rb 38·3 39·9 43·8 43·9 25·5 37·0 37·2 18·9 22·5 27·6 29·0 34·2 7·6 0·7 Zn 80·3 76·1 83·9 74·9 113 107 81·2 102 98·5 93·2 84·8 87·5 83·0 1·4 Sr 238 215 210 201 308 246 240 287 283 271 258 242 256 3·1 Y 45·5 47·2 50·1 52·6 36·0 45·3 45·7 31·8 30·9 36·5 39·8 42·3 18·4 1·3 Zr 134 143 146 151 92 127 130 71 81 98 111 119 45 1·3 Nb 3·5 4·0 3·4 4·2 2·8 3·5 3·9 2·8 2·5 3·0 3·1 3·3 3·6 0·4 Ba 512 546 548 567 347 481 497 266 306 360 410 447 100 6·9 Rare earth elements (ppm, ICP-MS) N=4 La 12·52 13·66 13·09 14·00 9·62 10·41 10·47 12·55 9·83 7·52 8·06 11·25 11·80 3·84 0·12 Ce 29·53 32·84 31·32 32·70 22·86 25·21 24·39 29·12 24·28 18·00 20·00 26·13 28·06 9·58 0·30 Pr 4·14 4·51 4·33 4·60 3·37 3·61 3·57 4·21 3·32 2·61 2·86 3·76 3·97 1·45 0·07 Nd 18·57 20·63 19·87 20·75 15·90 17·08 16·66 19·53 14·94 12·62 13·32 17·56 18·02 7·08 0·30 Sm 5·16 5·60 5·27 5·70 4·54 4·81 4·642 5·44 4·11 3·65 3·84 4·90 4·96 2·13 0·14 Eu 1·42 1·57 1·47 1·47 1·45 1·48 1·47 1·56 1·22 1·19 1·19 1·42 1·46 0·81 0·05 Gd 5·78 6·38 6·22 6·50 5·31 5·63 5·28 6·26 4·46 4·36 4·36 5·56 5·80 2·66 0·13 Tb 1·02 1·08 1·08 1·15 0·92 0·98 0·93 1·10 0·79 0·76 0·78 0·96 1·02 0·47 0·02 Dy 6·47 7·15 7·04 7·36 5·80 6·15 5·84 7·00 4·98 4·79 4·99 6·10 6·47 2·96 0·13 Ho 1·47 1·58 1·58 1·65 1·28 1·37 1·30 1·59 1·11 1·08 1·10 1·35 1·45 0·64 0·02 Er 4·35 4·82 4·74 4·99 3·80 4·07 3·82 4·73 3·37 3·16 3·32 4·00 4·37 1·86 0·07 Tm 0·67 0·73 0·75 0·80 0·57 0·61 0·59 0·73 0·52 0·48 0·51 0·61 0·66 0·28 0·01 Yb 4·59 4·86 4·87 5·30 3·60 3·97 3·79 4·67 3·47 3·08 3·31 3·92 4·42 1·75 0·06 Lu 0·72 0·77 0·78 0·81 0·57 0·62 0·59 0·73 0·54 0·47 0·51 0·62 0·69 0·28 0·01 Isotope ratios (TIMS) N=3 87Sr/86Sr 0·70332 0·70328 0·70327 0·70327 0·70332 0·70332 0·70335 0·70332 0·70336 0·70331 0·70338 0·70330 0·70331 0·70424 2·15E-05 2SE 1·13E-05 7·29E-06 8·34E-06 8·02E-06 1·25E-05 6·34E-06 1·58E-05 8·39E-06 1·53E-05 7·78E-06 1·09E-05 7·91E-06 7·54E-06 143Nd/144Nd 0·51301 0·51301 0·51302 0·51301 0·51301 0·51298 0·51300 0·51302 0·51302 0·51301 0·51301 0·51301 0·51302 0·51282 2·08E-05 2SE 7·14E-06 9·39E-06 7·83E-06 7·18E-06 6·00E-06 9·05E-06 1·08E-05 9·07E-06 9·29E-06 8·14E-06 1·09E-05 1·02E-05 7·72E-06 Sample No. 271* 17* 168 150 99 229 355 85 140 297 90 114 112 118 121 110 HB-2 Occurrence pm pm pm pm pm pm sc HP sc HP sc HP sc HP sc HP sc MP sc MP sc MP sc MP sc MP Eruptive unit Pre-Kp IV 2 3 3 4 3 3 3 3 3 3 4 4 4 4 4 House standard 2σ Major elements (wt %, XRF) N=14 SiO2 73·4 73·8 73·5 74·7 75·3 75·4 63·2 64·0 65·8 69·2 71·8 58·2 60·1 64·3 67·4 69·2 53·6 0·32 TiO2 0·49 0·52 0·53 0·49 0·46 0·47 0·93 0·87 0·81 0·68 0·61 1·02 0·95 0·83 0·70 0·67 0·76 0·01 Al2O3 14·9 14·1 14·2 13·9 13·5 13·5 15·3 14·8 15·0 14·3 14·2 15·6 15·2 14·9 15·3 14·2 17·4 0·12 FeO* 2·45 2·56 2·60 2·30 2·30 2·14 6·23 5·98 5·17 4·19 3·30 9·21 8·17 6·47 4·83 4·53 9·29 0·19 MnO 0·12 0·14 0·15 0·13 0·14 0·13 0·20 0·20 0·20 0·17 0·15 0·21 0·20 0·18 0·16 0·16 0·18 0·003 MgO 0·62 0·74 0·80 0·59 0·65 0·58 2·47 2·17 1·75 1·30 1·02 3·61 3·20 2·45 1·73 1·62 6·44 0·06 CaO 2·64 2·58 2·80 2·42 2·30 2·17 5·89 5·46 4·96 3·66 3·20 7·39 6·88 5·64 4·83 4·10 9·56 0·17 Na2O 4·07 4·30 4·51 4·51 4·51 4·51 3·80 3·75 4·04 4·12 4·33 3·05 3·30 3·60 3·87 4·13 2·40 0·03 K2O 1·67 1·72 1·64 1·71 1·83 1·76 1·03 1·06 1·16 1·42 1·59 0·77 0·93 1·15 1·28 1·42 0·24 0·005 P2O5 0·10 0·12 0·13 0·10 0·11 0·10 0·62 0·54 0·46 0·31 0·21 0·42 0·42 0·34 0·25 0·25 0·11 0·002 Total 100·37 100·59 100·76 100·87 101·07 100·72 99·63 98·78 99·37 99·32 100·41 99·44 99·36 99·87 100·40 100·23 Trace elements (ppm, XRF) N=14 Sc 17·9 14·6 17·4 16·6 30·0 20·5 21·2 38·8 33·7 31·7 25·0 25·9 33·3 2·3 V 24·0 23·8 21·4 19·7 114 71·5 42·3 253 214 158 95·8 94·9 217 4·6 Cr 3·17 2·66 3·40 2·35 5·12 2·65 3·57 19·1 8·6 17·6 10·7 4·8 170·0 3·8 Ni 0·95 1·06 0·81 1·29 2·43 1·75 1·19 3·99 3·93 2·44 4·77 1·74 54·0 0·9 Rb 38·3 39·9 43·8 43·9 25·5 37·0 37·2 18·9 22·5 27·6 29·0 34·2 7·6 0·7 Zn 80·3 76·1 83·9 74·9 113 107 81·2 102 98·5 93·2 84·8 87·5 83·0 1·4 Sr 238 215 210 201 308 246 240 287 283 271 258 242 256 3·1 Y 45·5 47·2 50·1 52·6 36·0 45·3 45·7 31·8 30·9 36·5 39·8 42·3 18·4 1·3 Zr 134 143 146 151 92 127 130 71 81 98 111 119 45 1·3 Nb 3·5 4·0 3·4 4·2 2·8 3·5 3·9 2·8 2·5 3·0 3·1 3·3 3·6 0·4 Ba 512 546 548 567 347 481 497 266 306 360 410 447 100 6·9 Rare earth elements (ppm, ICP-MS) N=4 La 12·52 13·66 13·09 14·00 9·62 10·41 10·47 12·55 9·83 7·52 8·06 11·25 11·80 3·84 0·12 Ce 29·53 32·84 31·32 32·70 22·86 25·21 24·39 29·12 24·28 18·00 20·00 26·13 28·06 9·58 0·30 Pr 4·14 4·51 4·33 4·60 3·37 3·61 3·57 4·21 3·32 2·61 2·86 3·76 3·97 1·45 0·07 Nd 18·57 20·63 19·87 20·75 15·90 17·08 16·66 19·53 14·94 12·62 13·32 17·56 18·02 7·08 0·30 Sm 5·16 5·60 5·27 5·70 4·54 4·81 4·642 5·44 4·11 3·65 3·84 4·90 4·96 2·13 0·14 Eu 1·42 1·57 1·47 1·47 1·45 1·48 1·47 1·56 1·22 1·19 1·19 1·42 1·46 0·81 0·05 Gd 5·78 6·38 6·22 6·50 5·31 5·63 5·28 6·26 4·46 4·36 4·36 5·56 5·80 2·66 0·13 Tb 1·02 1·08 1·08 1·15 0·92 0·98 0·93 1·10 0·79 0·76 0·78 0·96 1·02 0·47 0·02 Dy 6·47 7·15 7·04 7·36 5·80 6·15 5·84 7·00 4·98 4·79 4·99 6·10 6·47 2·96 0·13 Ho 1·47 1·58 1·58 1·65 1·28 1·37 1·30 1·59 1·11 1·08 1·10 1·35 1·45 0·64 0·02 Er 4·35 4·82 4·74 4·99 3·80 4·07 3·82 4·73 3·37 3·16 3·32 4·00 4·37 1·86 0·07 Tm 0·67 0·73 0·75 0·80 0·57 0·61 0·59 0·73 0·52 0·48 0·51 0·61 0·66 0·28 0·01 Yb 4·59 4·86 4·87 5·30 3·60 3·97 3·79 4·67 3·47 3·08 3·31 3·92 4·42 1·75 0·06 Lu 0·72 0·77 0·78 0·81 0·57 0·62 0·59 0·73 0·54 0·47 0·51 0·62 0·69 0·28 0·01 Isotope ratios (TIMS) N=3 87Sr/86Sr 0·70332 0·70328 0·70327 0·70327 0·70332 0·70332 0·70335 0·70332 0·70336 0·70331 0·70338 0·70330 0·70331 0·70424 2·15E-05 2SE 1·13E-05 7·29E-06 8·34E-06 8·02E-06 1·25E-05 6·34E-06 1·58E-05 8·39E-06 1·53E-05 7·78E-06 1·09E-05 7·91E-06 7·54E-06 143Nd/144Nd 0·51301 0·51301 0·51302 0·51301 0·51301 0·51298 0·51300 0·51302 0·51302 0·51301 0·51301 0·51301 0·51302 0·51282 2·08E-05 2SE 7·14E-06 9·39E-06 7·83E-06 7·18E-06 6·00E-06 9·05E-06 1·08E-05 9·07E-06 9·29E-06 8·14E-06 1·09E-05 1·02E-05 7·72E-06 * pm, pumice; sc: scoria. These samples are only for reference because of higher LOI (>3 wt %). Fig. 8. View largeDownload slide Chondrite-normalized REE concentrations in Kp IV juvenile materials. Chondrite data are from Sun & McDonough (1989). Pumice samples display greater light-REE enrichment and stronger negative Eu anomalies than scoria samples. REE contents are higher in HP scoria than in MP scoria. Fig. 8. View largeDownload slide Chondrite-normalized REE concentrations in Kp IV juvenile materials. Chondrite data are from Sun & McDonough (1989). Pumice samples display greater light-REE enrichment and stronger negative Eu anomalies than scoria samples. REE contents are higher in HP scoria than in MP scoria. Juvenile materials from the Kp IV eruption show relatively wide 87Sr/86Sr variations, whereas their 143Nd/144Nd values are nearly the same (Fig. 9, Table 2). The average 87Sr/86Sr value of scoria is slightly higher than that of pumice. The 87Sr/86Sr values of Kp IV pumice range from 0·70326 to 0·70332, and decrease systematically with increasing SiO2. Among the scorias, MP scoria exhibits a wider variation in 87Sr/86Sr (0·70330–0·70338) than HP scoria (0·70332–0·70335) and can be divided into two types with higher (>0·70336) and lower Sr isotope ratios (<0·70331). Fig. 9. View largeDownload slide Isotopic compositions of Kp IV juvenile materials. (a) 87Sr/86Sr vs SiO2. (b) 143Nd/144Nd vs 87Sr/86Sr. Each diagram includes a 2σ error symbol. The Sr isotopic ratios of pumices are lower than those of scorias, and less silicic pumices have higher 87Sr/86Sr than more silicic pumices. Abbreviations: wp, white pumice; sc, scoria. The data from Unit 2 and Pre-Kp IV pumice are shown for reference but were not used in our analysis because of their high loss on ignition (>3 wt %). Fig. 9. View largeDownload slide Isotopic compositions of Kp IV juvenile materials. (a) 87Sr/86Sr vs SiO2. (b) 143Nd/144Nd vs 87Sr/86Sr. Each diagram includes a 2σ error symbol. The Sr isotopic ratios of pumices are lower than those of scorias, and less silicic pumices have higher 87Sr/86Sr than more silicic pumices. Abbreviations: wp, white pumice; sc, scoria. The data from Unit 2 and Pre-Kp IV pumice are shown for reference but were not used in our analysis because of their high loss on ignition (>3 wt %). RELATIONSHIP BETWEEN PRE-KP IV AND KP IV MATERIALS Pumices from the Pre-Kp IV and Kp IV eruptions are similar in their matrix glass chemistry (Fig. 3), as well as their mineral chemistry and modal phenocryst contents (Fig. 5a). However, the cores of orthopyroxene and ilmenite phenocrysts in Pre-Kp IV pumice can be distinguished from those in Kp IV pumice, which indicates that the rhyolitic magma of Pre-Kp IV is distinct from that of Kp IV. Considering the short time gap between the Pre-Kp IV and Kp IV eruptions (Hasegawa et al., 2016), it appears that the silicic magma system was modified immediately before each eruption. In the following discussion, we focus on the petrological and geochemical features of the Kp IV eruption. DISCUSSION Magma mixing and end-member magmas of the Kp IV eruption Juvenile materials of the Kp IV eruption show heterogeneous textures that indicate mingling of silicic and basic magmas, such as banded pumice and scoria, and silicic inclusions in scoria. Evidence of more intimate magma mixing is present in the form of disequilibrium phenocryst compositions, even in samples of homogeneous pumice (Figs 5 and 6). In this section, we estimate the end-member magmas of the Kp IV eruption on the basis of the juvenile products and their mineral, whole-rock, and matrix glass chemistry. Mafic end-member magmas Hasegawa et al. (2016) suggested that the juvenile materials of the Kp IV eruption consist of pumice and three types of scoria, and that the scorias mingled with pumice to form heterogeneous clasts. Although the whole-rock chemistry of LP scoria could not be determined, the HP and MP scorias form two distinct linear trends in SiO2 variation diagrams (Fig. 7). This suggests that these two types of scoria represent the products of mixing or mingling of relatively mafic and silicic magmas. On the other hand, in a P2O5 vs SiO2 diagram (Fig. 3) the matrix glasses of the three scorias are clearly compositionally distinct, and a >10% SiO2 compositional gap separates the MP scoria and the pumice. In addition, although the glass of the HP scoria forms simple linear trends in all SiO2 variation diagrams, the other two types of scoria do not, again most clearly in the P2O5 vs SiO2 diagram. The evidence suggests that the three types of scoria are distinct products of mingling with the pumice, accounting for the differences in their matrix glasses. The scorias are less porphyritic than the pumice, the LP scoria having no phenocrysts and the HP and MP scorias being sparsely phyric. Although phenocryst minerals with compositional variations similar to those in the pumice are dominant in the scoria, minor amounts of calcic plagioclase (An = 80–95) and magnesian magnetite (Mg/Mn = 7·5–25) are also present in the scoria (Fig. 5, summarized in Fig. 10). Because the latter two types of phenocryst must have crystallized from a magma less silicic than that yielding the dominant phenocrysts, we describe them here as Type-M (mafic) phenocrysts. We further divide the Type-M phenocrysts into Type-M1, with lower Mg/Mn (7·5–17) magnetite, found in the HP scoria, and Type-M2, with higher Mg/Mn (11·5–25) magnetite, found in the MP scoria (Fig. 10). Phenocryst contents in the HP and MP scorias decrease with decreasing whole-rock SiO2 content (Fig. 4), suggesting that the end-member magmas were nearly aphyric andesites, containing only a few modal percent of Type-M phenocrysts. In the following discussion, we refer to these end-member magmas as LP-andesitic, MP-andesitic (containing Type-M2 phenocrysts), and HP-andesitic (containing Type-M1 phenocrysts). Fig. 10. View largeDownload slide Summary of types of phenocryst in Kp IV juvenile materials. Type-R (rhyolitic) and Type-D (dacitic) phenocrysts are dominant in all samples. Type-M1 and Type-M2 (mafic) phenocrysts are found only in HP and MP scoria, respectively. Phenocrysts shown in light gray with question marks (high-An plagioclase in pumice and high-Mg/Mn ilmenite in HP scoria) could not be clearly characterized and were not considered further. Fig. 10. View largeDownload slide Summary of types of phenocryst in Kp IV juvenile materials. Type-R (rhyolitic) and Type-D (dacitic) phenocrysts are dominant in all samples. Type-M1 and Type-M2 (mafic) phenocrysts are found only in HP and MP scoria, respectively. Phenocrysts shown in light gray with question marks (high-An plagioclase in pumice and high-Mg/Mn ilmenite in HP scoria) could not be clearly characterized and were not considered further. Silicic end-member magmas Pumice whole-rock compositions form well-defined linear trends in SiO2 variation diagrams (Fig. 7b). The REE patterns of pumice samples are also nearly parallel to one another (Fig. 8). These features imply that the pumice compositions were not significantly affected by fractionation of minor phenocryst phases (e.g. apatite and zircon) and that pumice is a product of mixing between two end-member magmas. Because the pumice trend is so clearly different from those of the scorias (Fig. 7b), the andesitic end-member magmas that formed the scorias must not have been involved in the formation of the pumice. Instead, two silicic end-member magmas are required. This interpretation is consistent with the compositional distribution of phenocryst minerals. In each pumice sample, the cores of plagioclase, orthopyroxene, and clinopyroxene phenocrysts have a wide compositional range, though not an obviously bimodal distribution, except in the case of Fe–Ti oxide phenocrysts in sample No. 168 (Fig. 5a). This evidence suggests that the pumice is not a product of a simple crystallization process but, rather, can be explained by mixing of end-member magmas containing phenocrysts with similar compositions. Compositional variations of phenocrysts that show reverse zoning also support this hypothesis (Fig. 6). Mafic magma-derived phenocrysts, such as olivine and pyroxenes with Mg# > 80 and calcic plagioclase (An > 80), are absent in the pumice. Therefore, we modeled the two silicic end-member magmas as a rhyolite and a dacite, each with its respective set of phenocrysts (Type R or Type D) as shown in Fig. 10. Individual pumice samples contain both Type-R and Type-D phenocrysts (Fig. 10, Table 3). Type-R phenocrysts are the predominant component and consist of sodic plagioclase (An = 35 to c. 48), ferrous orthopyroxene (Mg# = 67 to c. 70), and Mg-poor magnetite (Mg/Mn = 2·4–2·9) and ilmenite (Mg/Mn = 3·5–4·2), which are in equilibrium with each other (Bacon & Hirschmann, 1988). A small amount of ferrous clinopyroxene (Mg# = 69 to c. 73), which could be in equilibrium with the ferrous orthopyroxene phenocrysts (Baker & Eggler, 1987; Brophy & Dreher, 2000), might be also included. Type-D phenocrysts consist of slightly calcic plagioclase (An = c. 48–70), magnesian orthopyroxene (Mg# = c. 70–76), and magnesian clinopyroxene (Mg# = c. 73–78). The pyroxene phenocrysts are in equilibrium (Baker & Eggler, 1987; Brophy & Dreher, 2000), and some of them form crystal clots. The slightly Mg-rich magnetite (Mg/Mn = 2·8–7·5) and ilmenite (Mg/Mn = 4·2–13) in sample No. 168 might be considered Type-D phenocrysts, but other pumice samples do not contain magnetite and ilmenite of that composition. Table 3: Estimated whole-rock compositions, magmatic temperatures, and phenocryst types of end-member magmas in Kp IV eruptive activity Magma type R1 D1 HP-A MP-A LP-A2 wt % SiO2 75·5 72·0 61·4 57·7 59·4 TiO2 0·45 0·54 1·00 1·07 0·924 Al2O3 12·9 14·7 15·6 15·8 16·54 FeO* 2·06 2·62 7·25 9·65 8·43 MnO 0·12 0·14 0·22 0·21 0·18 MgO 0·54 0·77 2·82 3·71 2·41 CaO 2·04 2·93 6·56 7·73 7·57 Na2O 4·49 4·56 3·56 2·97 3·63 K2O 1·84 1·56 0·90 0·73 0·80 P2O5 0·09 0·12 0·71 0·47 0·19 ppm Sc 16·2 17·8 31·7 38·0 V 17·7 25·0 142 257 Cr 3·2 2·5 5·0 9·5 Ni 0·8 1·3 2·3 3·9 Rb 43·3 37·7 23·5 19·2 Zn 73·5 76·9 112 101 Sr 193 255 311 289 Y 52·3 43·3 34·9 29·6 Zr 148 135 84·7 68·0 Ba 561 518 327 256 Pb 12·0 11·3 13·2 7·5 La 14·2 12·4 9·0 7·2 Ce 33·7 29·7 21·3 17·3 Temperature(oC) 849–8593 856–8723 1044–10624 1063–10814 1033–10484 Phenocryst type R D M1 M2 None Plagioclase (An) 35–48 48–70 80–95 80–95 – Orthopyroxene (Mg#) 67–70 70–76 – – – Clinopyroxene (Mg#) 70–73 73–78 – – – Magnetite (Mg/Mn) 2·4–2·8 2·8–7·5 7·5–17 11·5–25 – Ilmenite (Mg/Mn) 3·5–4·2 4·2–13 – – – Magma type R1 D1 HP-A MP-A LP-A2 wt % SiO2 75·5 72·0 61·4 57·7 59·4 TiO2 0·45 0·54 1·00 1·07 0·924 Al2O3 12·9 14·7 15·6 15·8 16·54 FeO* 2·06 2·62 7·25 9·65 8·43 MnO 0·12 0·14 0·22 0·21 0·18 MgO 0·54 0·77 2·82 3·71 2·41 CaO 2·04 2·93 6·56 7·73 7·57 Na2O 4·49 4·56 3·56 2·97 3·63 K2O 1·84 1·56 0·90 0·73 0·80 P2O5 0·09 0·12 0·71 0·47 0·19 ppm Sc 16·2 17·8 31·7 38·0 V 17·7 25·0 142 257 Cr 3·2 2·5 5·0 9·5 Ni 0·8 1·3 2·3 3·9 Rb 43·3 37·7 23·5 19·2 Zn 73·5 76·9 112 101 Sr 193 255 311 289 Y 52·3 43·3 34·9 29·6 Zr 148 135 84·7 68·0 Ba 561 518 327 256 Pb 12·0 11·3 13·2 7·5 La 14·2 12·4 9·0 7·2 Ce 33·7 29·7 21·3 17·3 Temperature(oC) 849–8593 856–8723 1044–10624 1063–10814 1033–10484 Phenocryst type R D M1 M2 None Plagioclase (An) 35–48 48–70 80–95 80–95 – Orthopyroxene (Mg#) 67–70 70–76 – – – Clinopyroxene (Mg#) 70–73 73–78 – – – Magnetite (Mg/Mn) 2·4–2·8 2·8–7·5 7·5–17 11·5–25 – Ilmenite (Mg/Mn) 3·5–4·2 4·2–13 – – – *Total Fe given as FeO. 1 These compositions are the values at SiO2 = 75·5 and 72·0 wt % on the mixing trend of pumices samples, respectively. The ratios of FeO/total-FeO of melts are calculated on the basis of Sack et al. (1981) and Kilinc et al. (1983). These estimated compositions exhibit compositionally equilibrium with the Type-R and Type-D orthopyroxene phenocrysts approximately. 2 The composition is substituted by the most mafic one of the matrix glass in LP scoria. 3 These values are estimated by Fe–Ti oxides geothermometer in QUILF program (Andersen et al., 1993). 4 These values are the liquidus temperatures of each magma estimated using a rhyolite-MELTS program (Gualda et al., 2012) because they are nearly aphyric. The calculated condition is NNO buffer, 2–3 kbr, and H2O = 4 wt %. This water content is assumed on the basis of the model of Moore et al. (1998). Table 3: Estimated whole-rock compositions, magmatic temperatures, and phenocryst types of end-member magmas in Kp IV eruptive activity Magma type R1 D1 HP-A MP-A LP-A2 wt % SiO2 75·5 72·0 61·4 57·7 59·4 TiO2 0·45 0·54 1·00 1·07 0·924 Al2O3 12·9 14·7 15·6 15·8 16·54 FeO* 2·06 2·62 7·25 9·65 8·43 MnO 0·12 0·14 0·22 0·21 0·18 MgO 0·54 0·77 2·82 3·71 2·41 CaO 2·04 2·93 6·56 7·73 7·57 Na2O 4·49 4·56 3·56 2·97 3·63 K2O 1·84 1·56 0·90 0·73 0·80 P2O5 0·09 0·12 0·71 0·47 0·19 ppm Sc 16·2 17·8 31·7 38·0 V 17·7 25·0 142 257 Cr 3·2 2·5 5·0 9·5 Ni 0·8 1·3 2·3 3·9 Rb 43·3 37·7 23·5 19·2 Zn 73·5 76·9 112 101 Sr 193 255 311 289 Y 52·3 43·3 34·9 29·6 Zr 148 135 84·7 68·0 Ba 561 518 327 256 Pb 12·0 11·3 13·2 7·5 La 14·2 12·4 9·0 7·2 Ce 33·7 29·7 21·3 17·3 Temperature(oC) 849–8593 856–8723 1044–10624 1063–10814 1033–10484 Phenocryst type R D M1 M2 None Plagioclase (An) 35–48 48–70 80–95 80–95 – Orthopyroxene (Mg#) 67–70 70–76 – – – Clinopyroxene (Mg#) 70–73 73–78 – – – Magnetite (Mg/Mn) 2·4–2·8 2·8–7·5 7·5–17 11·5–25 – Ilmenite (Mg/Mn) 3·5–4·2 4·2–13 – – – Magma type R1 D1 HP-A MP-A LP-A2 wt % SiO2 75·5 72·0 61·4 57·7 59·4 TiO2 0·45 0·54 1·00 1·07 0·924 Al2O3 12·9 14·7 15·6 15·8 16·54 FeO* 2·06 2·62 7·25 9·65 8·43 MnO 0·12 0·14 0·22 0·21 0·18 MgO 0·54 0·77 2·82 3·71 2·41 CaO 2·04 2·93 6·56 7·73 7·57 Na2O 4·49 4·56 3·56 2·97 3·63 K2O 1·84 1·56 0·90 0·73 0·80 P2O5 0·09 0·12 0·71 0·47 0·19 ppm Sc 16·2 17·8 31·7 38·0 V 17·7 25·0 142 257 Cr 3·2 2·5 5·0 9·5 Ni 0·8 1·3 2·3 3·9 Rb 43·3 37·7 23·5 19·2 Zn 73·5 76·9 112 101 Sr 193 255 311 289 Y 52·3 43·3 34·9 29·6 Zr 148 135 84·7 68·0 Ba 561 518 327 256 Pb 12·0 11·3 13·2 7·5 La 14·2 12·4 9·0 7·2 Ce 33·7 29·7 21·3 17·3 Temperature(oC) 849–8593 856–8723 1044–10624 1063–10814 1033–10484 Phenocryst type R D M1 M2 None Plagioclase (An) 35–48 48–70 80–95 80–95 – Orthopyroxene (Mg#) 67–70 70–76 – – – Clinopyroxene (Mg#) 70–73 73–78 – – – Magnetite (Mg/Mn) 2·4–2·8 2·8–7·5 7·5–17 11·5–25 – Ilmenite (Mg/Mn) 3·5–4·2 4·2–13 – – – *Total Fe given as FeO. 1 These compositions are the values at SiO2 = 75·5 and 72·0 wt % on the mixing trend of pumices samples, respectively. The ratios of FeO/total-FeO of melts are calculated on the basis of Sack et al. (1981) and Kilinc et al. (1983). These estimated compositions exhibit compositionally equilibrium with the Type-R and Type-D orthopyroxene phenocrysts approximately. 2 The composition is substituted by the most mafic one of the matrix glass in LP scoria. 3 These values are estimated by Fe–Ti oxides geothermometer in QUILF program (Andersen et al., 1993). 4 These values are the liquidus temperatures of each magma estimated using a rhyolite-MELTS program (Gualda et al., 2012) because they are nearly aphyric. The calculated condition is NNO buffer, 2–3 kbr, and H2O = 4 wt %. This water content is assumed on the basis of the model of Moore et al. (1998). Whole-rock composition and temperature of end-member magmas According to the compositional equilibria between melt and pyroxenes, in which the Fe–Mg partition coefficient between mineral and melt is 0·247 for orthopyroxene and 0. ·233 for clinopyroxene (No. SS-40 in Brophy & Dreher, 2000), rhyolitic and dacitic end-member magmas can be assumed with about 75·5 and 72·0 wt % SiO2, respectively, from the linear pumice trends in SiO2 variation diagrams (Table 3, Fig. 11a). Although andesitic magmas should be only slightly porphyritic or aphyric, we can estimate the whole-rock compositions of the andesitic end-member magmas of the HP and MP scorias (HP-andesitic and MP-andesitic magmas, respectively) using the linear relationship between phenocryst contents and whole-rock SiO2 contents in Fig. 4 (Table 3, Fig. 11b). The estimated compositions of these two magmas, SiO2 = 61·4 wt % for HP-andesitic and 57·7 wt % for MP-andesitic, are similar to those of the respective matrix glasses. These SiO2 contents are considered minimum values because these magmas must originally have contained small percentages of phenocrysts, as discussed previously. If we assume the LP-andesitic magma to be aphyric as in the case of the HP and MP scorias, the composition of the matrix glass with the lowest SiO2 content should represent the whole-rock composition of its source andesitic magma (SiO2 = 58·7 wt %; Table 3, Fig. 11b). Fig. 11. View largeDownload slide Estimated whole-rock compositions of Kp IV end-member magmas and their mixing relationships. (a) Al2O3 vs SiO2 emphasizing pumice samples. The single linear trend in pumice, which differs from the trends in scoria samples, suggests mixing of rhyolitic (R) and dacitic (D) end-member magmas. (b) P2O5 vs SiO2. Assuming that the andesitic end-member magmas were aphyric, we can estimate compositions of HP-andesitic (HP-A) and MP-andesitic (MP-A) magmas as similar to the matrix glass compositions of HP and MP scoria, respectively, with the least silica. By extension, LP-andesitic magma is assigned the composition of matrix glass in LP scoria with the least silica. Fig. 11. View largeDownload slide Estimated whole-rock compositions of Kp IV end-member magmas and their mixing relationships. (a) Al2O3 vs SiO2 emphasizing pumice samples. The single linear trend in pumice, which differs from the trends in scoria samples, suggests mixing of rhyolitic (R) and dacitic (D) end-member magmas. (b) P2O5 vs SiO2. Assuming that the andesitic end-member magmas were aphyric, we can estimate compositions of HP-andesitic (HP-A) and MP-andesitic (MP-A) magmas as similar to the matrix glass compositions of HP and MP scoria, respectively, with the least silica. By extension, LP-andesitic magma is assigned the composition of matrix glass in LP scoria with the least silica. We estimated the magmatic temperatures of the silicic end-member magmas from Fe–Ti oxide thermometry using the QUILF program (Andersen et al., 1993; Table 3 and Supplementary Data). Results from average core compositions of Type-R phenocrysts show that the temperature of the rhyolitic end-member magma was 849–859°C. If the magnetite and ilmenite phenocrysts with higher Mg/Mn ratios in sample No. 168 are considered Type-D phenocrysts, the dacitic end-member magma was only slightly hotter than the rhyolitic end-member magma (c. 856–872°C). According to the two-pyroxene geothermometer (Putirka, 2008), magma temperatures indicated by Type-R and Type-D pyroxene cores are 790–823°C and 835–951°C, respectively, making the dacitic end-member magma notably hotter than the rhyolitic magma. By comparison, we can substitute the liquidus temperatures of the HP-, MP-, and LP-andesitic magmas for their magmatic temperatures, because they are nearly aphyric. Using the rhyolite-MELTS program (Gualda et al., 2012) assuming a NNO buffer, 200–300 MPa pressure, and 4 wt % water content, we estimate the liquidus temperatures of the HP-, MP-, and LP-andesitic magmas to be 1044–1062°C, 1063–1081°C, and 1033–1048°C, respectively (Table 3). Petrogenetic relationships among end-member magmas Silicic and andesitic magmas Silicic magma can be produced by partial melting of mafic to intermediate crustal materials (e.g. Smith & Leeman, 1987; Streck & Grunder, 2008), by melt extraction from incompletely crystallized plutons, i.e. crystal mush (e.g. Bachmann & Bergantz, 2004; Hildreth, 2004; Deering et al., 2011b), and by fractional crystallization of basaltic magma (e.g. Miyagi et al., 2012; Parker et al., 2012). We examined the possibility that fractional crystallization of coexisting mafic magmas, as proposed for Kutcharo volcano by Miyagi et al. (2012), can produce rhyolitic as well as dacitic magmas. According to the rhyolite-MELTS model (in which the fractionated assemblage is composed of plagioclase, orthopyroxene, clinopyroxene, Fe–Ti oxides, quartz, hornblende, and apatite), neither the HP- nor the MP-andesitic magma can produce rhyolitic and dacitic magmas, especially their Na2O and P2O5 contents (Fig. 12). We also modeled fractional crystallization based on mass-balance calculations for major elements, assuming that the fractionated phases are the same as the phenocryst minerals in the MP scoria (Table 4). This calculation does not rule out the possibility that fractional crystallization of MP-andesitic magma produced a rhyolitic magma: the sum of squared residuals (R2) is 0·09 and the relative residuals in the rhyolitic end-member magma for each element except for MnO and P2O5 are less than 15% (Table 4). When applying these results to trace elements, however, a Rayleigh fractionation model (CL = Ci * F(D–1), where CL is the composition of the daughter melt, Ci the composition of the parental melt, F the melt fraction, and D the bulk partition coefficient of the fractionating phase, as listed in Table 4) cannot match the trace element contents, especially Y, of the end-member silicic magmas (Fig. 13). In addition, we employed an assimilation and fractional crystallization model (DePaolo, 1981), using the same minerals for the fractionate as those in the simple fractional crystallization model. For the assimilant, we tested two compositions: granite exposed in the Hidaka belt in central Hokkaido (Kamiyama et al., 2007) and a granodiorite xenolith in the Eocene Beppo Formation in eastern Hokkaido (Ogasawara et al., 1998). In both cases, the models give good results for major elements (R2 is 0·03 and the relative residuals in the rhyolitic end-member magma values for each element are less than 26%; Table 4). For trace elements, however, these models cannot duplicate the end-member rhyolitic magma (Fig. 13). Therefore, combined assimilation and fractional crystallization of coexisting andesitic magmas does not account for the Kp IV silicic magma system. This result is also consistent with the higher Sr isotopic ratios in scoria than in pumice (Fig. 9). Table 4: Results of the assimilation and fractional crystallization models for MP-andesitic magma wt % SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O P2O5 ppm Rb Sr Y Zr Ba Parent MP-A 57·70 1·07 15·76 9·65 0·21 3·71 7·73 2·97 0·73 0·47 19·2 289 29·6 68 256 Daughter R 75·50 0·45 12·88 2·06 0·12 0·54 2·04 4·49 1·84 0·09 43·3 193 52·3 148 561 Assimilant Hidaka granite 71·86 0·40 14·61 2·26 0·04 0·59 1·73 4·42 4·04 0·06 139·0 78 43·0 308 488 Beppo granodiorite 62·29 0·57 16·76 5·47 0·10 2·78 5·60 3·48 2·75 0·19 68·1 382 22·5 146 620 Rayleigh fractionation model Fractionated phase: 54·5 wt % 31·5 high-An plg, 22·9 low-An plg, 13·1 opx, 15·6 cpx, 9·6 high-Mg mgt, 5·0 low-Mg mgt, 0·5 ilm, 1·7 apt Result ΣR2 = 0·09 75·56 0·50 12·92 2·11 0·06 0·57 2·08 4·49 1·57 0·15 39·2 202 25·2 140 478 R −0·06 −0·05 −0·04 −0·05 0·06 −0·03 −0·04 0·00 0·27 −0·06 4·09 −8·42 27·16 7·66 83·04 R/Daughter 0·00 −0·11 0·00 −0·02 0·53 −0·05 −0·02 0·00 0·15 −0·68 0·09 −0·04 0·52 0·05 0·15 Assimilation and fractional crystallization model (Hidaka granite) Fractionated phase: 54·2 wt % 31·6 high-An plg, 22·6 low-An plg, 12·5 opx, 15·6 cpx, 13·8 high-Mg mgt, 1·0 low-Mg mgt, 1·0 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·10 75·43 0·40 12·98 2·03 0·09 0·62 2·02 4·49 1·84 0·09 57·2 183 25·0 180 528 R 0·07 0·04 −0·10 0·02 0·03 −0·08 0·02 0·00 0·00 0·00 −13·9 10·0 27·4 −32·2 32·6 R/Daughter 0·00 0·10 −0·01 0·01 0·24 −0·15 0·01 0·00 0·00 0·01 −0·3 0·1 0·5 −0·2 0·1 Assimilation and fractional crystallization model (Beppo granodiorite) Fractionated phase: 57·1 wt % 31·8 high-An plg, 22·7 low-An plg, 12·6 opx, 15·5 cpx, 13·5 high-Mg mgt, 0·9 low-Mg mgt, 0·9 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·14 75·42 0·40 13·00 2·03 0·09 0·63 2·02 4·49 1·84 0·09 55·0 200 23·0 177 607 R 0·08 0·05 −0·12 0·02 0·03 −0·09 0·02 0·00 0·00 0·00 −11·7 −6·9 29·3 −29·6 −46·1 R/Daughter 0·00 0·11 −0·01 0·01 0·26 −0·17 0·01 0·00 0·00 0·02 −0·3 0·0 0·6 −0·2 −0·1 Mineral compositions* Partition coefficient** high-An plg 44·81 0·00 35·06 0·39 0·00 0·09 19·08 0·57 0·01 0·00 0·10 1·50 0·06 0·00 0·20 low-An plg 57·90 0·00 26·28 0·40 0·00 0·03 8·84 6·41 0·14 0·00 0·17 4·04 0·02 0·01 0·47 opx 53·79 0·27 1·74 15·92 0·67 26·27 1·30 0·03 0·00 0·00 0·05 0·01 0·40 0·11 0·02 cpx 53·27 0·30 1·11 8·10 1·02 14·88 20·98 0·34 0·00 0·00 0·07 0·17 2·70 0·17 0·10 high-Mg mgt 0·15 7·70 5·87 81·41 0·33 4·54 0·01 0·00 0·00 0·00 0·01 0·01 0·20 0·10 0·12 low-Mg mgt 0·13 9·35 1·99 85·47 1·22 1·84 0·02 0·00 0·00 0·00 0·01 0·01 0·12 0·24 0·10 ilm 0·03 40·17 0·28 56·03 1·10 2·39 0·01 0·00 0·00 0·00 0·01 0·01 0·27 0·60 0·01 apt 0·00 0·00 0·00 0·00 0·00 0·00 56·80 0·00 0·00 43·20 0·00 2·00 40·0 0·64 0·00 wt % SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O P2O5 ppm Rb Sr Y Zr Ba Parent MP-A 57·70 1·07 15·76 9·65 0·21 3·71 7·73 2·97 0·73 0·47 19·2 289 29·6 68 256 Daughter R 75·50 0·45 12·88 2·06 0·12 0·54 2·04 4·49 1·84 0·09 43·3 193 52·3 148 561 Assimilant Hidaka granite 71·86 0·40 14·61 2·26 0·04 0·59 1·73 4·42 4·04 0·06 139·0 78 43·0 308 488 Beppo granodiorite 62·29 0·57 16·76 5·47 0·10 2·78 5·60 3·48 2·75 0·19 68·1 382 22·5 146 620 Rayleigh fractionation model Fractionated phase: 54·5 wt % 31·5 high-An plg, 22·9 low-An plg, 13·1 opx, 15·6 cpx, 9·6 high-Mg mgt, 5·0 low-Mg mgt, 0·5 ilm, 1·7 apt Result ΣR2 = 0·09 75·56 0·50 12·92 2·11 0·06 0·57 2·08 4·49 1·57 0·15 39·2 202 25·2 140 478 R −0·06 −0·05 −0·04 −0·05 0·06 −0·03 −0·04 0·00 0·27 −0·06 4·09 −8·42 27·16 7·66 83·04 R/Daughter 0·00 −0·11 0·00 −0·02 0·53 −0·05 −0·02 0·00 0·15 −0·68 0·09 −0·04 0·52 0·05 0·15 Assimilation and fractional crystallization model (Hidaka granite) Fractionated phase: 54·2 wt % 31·6 high-An plg, 22·6 low-An plg, 12·5 opx, 15·6 cpx, 13·8 high-Mg mgt, 1·0 low-Mg mgt, 1·0 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·10 75·43 0·40 12·98 2·03 0·09 0·62 2·02 4·49 1·84 0·09 57·2 183 25·0 180 528 R 0·07 0·04 −0·10 0·02 0·03 −0·08 0·02 0·00 0·00 0·00 −13·9 10·0 27·4 −32·2 32·6 R/Daughter 0·00 0·10 −0·01 0·01 0·24 −0·15 0·01 0·00 0·00 0·01 −0·3 0·1 0·5 −0·2 0·1 Assimilation and fractional crystallization model (Beppo granodiorite) Fractionated phase: 57·1 wt % 31·8 high-An plg, 22·7 low-An plg, 12·6 opx, 15·5 cpx, 13·5 high-Mg mgt, 0·9 low-Mg mgt, 0·9 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·14 75·42 0·40 13·00 2·03 0·09 0·63 2·02 4·49 1·84 0·09 55·0 200 23·0 177 607 R 0·08 0·05 −0·12 0·02 0·03 −0·09 0·02 0·00 0·00 0·00 −11·7 −6·9 29·3 −29·6 −46·1 R/Daughter 0·00 0·11 −0·01 0·01 0·26 −0·17 0·01 0·00 0·00 0·02 −0·3 0·0 0·6 −0·2 −0·1 Mineral compositions* Partition coefficient** high-An plg 44·81 0·00 35·06 0·39 0·00 0·09 19·08 0·57 0·01 0·00 0·10 1·50 0·06 0·00 0·20 low-An plg 57·90 0·00 26·28 0·40 0·00 0·03 8·84 6·41 0·14 0·00 0·17 4·04 0·02 0·01 0·47 opx 53·79 0·27 1·74 15·92 0·67 26·27 1·30 0·03 0·00 0·00 0·05 0·01 0·40 0·11 0·02 cpx 53·27 0·30 1·11 8·10 1·02 14·88 20·98 0·34 0·00 0·00 0·07 0·17 2·70 0·17 0·10 high-Mg mgt 0·15 7·70 5·87 81·41 0·33 4·54 0·01 0·00 0·00 0·00 0·01 0·01 0·20 0·10 0·12 low-Mg mgt 0·13 9·35 1·99 85·47 1·22 1·84 0·02 0·00 0·00 0·00 0·01 0·01 0·12 0·24 0·10 ilm 0·03 40·17 0·28 56·03 1·10 2·39 0·01 0·00 0·00 0·00 0·01 0·01 0·27 0·60 0·01 apt 0·00 0·00 0·00 0·00 0·00 0·00 56·80 0·00 0·00 43·20 0·00 2·00 40·0 0·64 0·00 The results of only MP-andesite magma are shown here, because we obtained the similar result using HP-andesite magma. Assimilants are taken from Kamiyama et al. (2007) and Ogasawara et al. (1998). * Mineral compositions are taken from analysed data of the phenocrysts in the most mafic MP-type scoria (No. 114), except for apatite (theoretical value). ** These values are taken from Philpotts & Schnetzler (1970), Arth (1976), Okamoto (1979), Pearce & Norry (1979), Luhr & Carmichael (1980), Gill (1981), Fujimaki (1986), Bacon & Druitt (1988), McKenzie & O’Nions (1991), Sisson (1991), Ewart & Griffin (1994), and Bindeman & Davis (2000). Table 4: Results of the assimilation and fractional crystallization models for MP-andesitic magma wt % SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O P2O5 ppm Rb Sr Y Zr Ba Parent MP-A 57·70 1·07 15·76 9·65 0·21 3·71 7·73 2·97 0·73 0·47 19·2 289 29·6 68 256 Daughter R 75·50 0·45 12·88 2·06 0·12 0·54 2·04 4·49 1·84 0·09 43·3 193 52·3 148 561 Assimilant Hidaka granite 71·86 0·40 14·61 2·26 0·04 0·59 1·73 4·42 4·04 0·06 139·0 78 43·0 308 488 Beppo granodiorite 62·29 0·57 16·76 5·47 0·10 2·78 5·60 3·48 2·75 0·19 68·1 382 22·5 146 620 Rayleigh fractionation model Fractionated phase: 54·5 wt % 31·5 high-An plg, 22·9 low-An plg, 13·1 opx, 15·6 cpx, 9·6 high-Mg mgt, 5·0 low-Mg mgt, 0·5 ilm, 1·7 apt Result ΣR2 = 0·09 75·56 0·50 12·92 2·11 0·06 0·57 2·08 4·49 1·57 0·15 39·2 202 25·2 140 478 R −0·06 −0·05 −0·04 −0·05 0·06 −0·03 −0·04 0·00 0·27 −0·06 4·09 −8·42 27·16 7·66 83·04 R/Daughter 0·00 −0·11 0·00 −0·02 0·53 −0·05 −0·02 0·00 0·15 −0·68 0·09 −0·04 0·52 0·05 0·15 Assimilation and fractional crystallization model (Hidaka granite) Fractionated phase: 54·2 wt % 31·6 high-An plg, 22·6 low-An plg, 12·5 opx, 15·6 cpx, 13·8 high-Mg mgt, 1·0 low-Mg mgt, 1·0 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·10 75·43 0·40 12·98 2·03 0·09 0·62 2·02 4·49 1·84 0·09 57·2 183 25·0 180 528 R 0·07 0·04 −0·10 0·02 0·03 −0·08 0·02 0·00 0·00 0·00 −13·9 10·0 27·4 −32·2 32·6 R/Daughter 0·00 0·10 −0·01 0·01 0·24 −0·15 0·01 0·00 0·00 0·01 −0·3 0·1 0·5 −0·2 0·1 Assimilation and fractional crystallization model (Beppo granodiorite) Fractionated phase: 57·1 wt % 31·8 high-An plg, 22·7 low-An plg, 12·6 opx, 15·5 cpx, 13·5 high-Mg mgt, 0·9 low-Mg mgt, 0·9 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·14 75·42 0·40 13·00 2·03 0·09 0·63 2·02 4·49 1·84 0·09 55·0 200 23·0 177 607 R 0·08 0·05 −0·12 0·02 0·03 −0·09 0·02 0·00 0·00 0·00 −11·7 −6·9 29·3 −29·6 −46·1 R/Daughter 0·00 0·11 −0·01 0·01 0·26 −0·17 0·01 0·00 0·00 0·02 −0·3 0·0 0·6 −0·2 −0·1 Mineral compositions* Partition coefficient** high-An plg 44·81 0·00 35·06 0·39 0·00 0·09 19·08 0·57 0·01 0·00 0·10 1·50 0·06 0·00 0·20 low-An plg 57·90 0·00 26·28 0·40 0·00 0·03 8·84 6·41 0·14 0·00 0·17 4·04 0·02 0·01 0·47 opx 53·79 0·27 1·74 15·92 0·67 26·27 1·30 0·03 0·00 0·00 0·05 0·01 0·40 0·11 0·02 cpx 53·27 0·30 1·11 8·10 1·02 14·88 20·98 0·34 0·00 0·00 0·07 0·17 2·70 0·17 0·10 high-Mg mgt 0·15 7·70 5·87 81·41 0·33 4·54 0·01 0·00 0·00 0·00 0·01 0·01 0·20 0·10 0·12 low-Mg mgt 0·13 9·35 1·99 85·47 1·22 1·84 0·02 0·00 0·00 0·00 0·01 0·01 0·12 0·24 0·10 ilm 0·03 40·17 0·28 56·03 1·10 2·39 0·01 0·00 0·00 0·00 0·01 0·01 0·27 0·60 0·01 apt 0·00 0·00 0·00 0·00 0·00 0·00 56·80 0·00 0·00 43·20 0·00 2·00 40·0 0·64 0·00 wt % SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O P2O5 ppm Rb Sr Y Zr Ba Parent MP-A 57·70 1·07 15·76 9·65 0·21 3·71 7·73 2·97 0·73 0·47 19·2 289 29·6 68 256 Daughter R 75·50 0·45 12·88 2·06 0·12 0·54 2·04 4·49 1·84 0·09 43·3 193 52·3 148 561 Assimilant Hidaka granite 71·86 0·40 14·61 2·26 0·04 0·59 1·73 4·42 4·04 0·06 139·0 78 43·0 308 488 Beppo granodiorite 62·29 0·57 16·76 5·47 0·10 2·78 5·60 3·48 2·75 0·19 68·1 382 22·5 146 620 Rayleigh fractionation model Fractionated phase: 54·5 wt % 31·5 high-An plg, 22·9 low-An plg, 13·1 opx, 15·6 cpx, 9·6 high-Mg mgt, 5·0 low-Mg mgt, 0·5 ilm, 1·7 apt Result ΣR2 = 0·09 75·56 0·50 12·92 2·11 0·06 0·57 2·08 4·49 1·57 0·15 39·2 202 25·2 140 478 R −0·06 −0·05 −0·04 −0·05 0·06 −0·03 −0·04 0·00 0·27 −0·06 4·09 −8·42 27·16 7·66 83·04 R/Daughter 0·00 −0·11 0·00 −0·02 0·53 −0·05 −0·02 0·00 0·15 −0·68 0·09 −0·04 0·52 0·05 0·15 Assimilation and fractional crystallization model (Hidaka granite) Fractionated phase: 54·2 wt % 31·6 high-An plg, 22·6 low-An plg, 12·5 opx, 15·6 cpx, 13·8 high-Mg mgt, 1·0 low-Mg mgt, 1·0 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·10 75·43 0·40 12·98 2·03 0·09 0·62 2·02 4·49 1·84 0·09 57·2 183 25·0 180 528 R 0·07 0·04 −0·10 0·02 0·03 −0·08 0·02 0·00 0·00 0·00 −13·9 10·0 27·4 −32·2 32·6 R/Daughter 0·00 0·10 −0·01 0·01 0·24 −0·15 0·01 0·00 0·00 0·01 −0·3 0·1 0·5 −0·2 0·1 Assimilation and fractional crystallization model (Beppo granodiorite) Fractionated phase: 57·1 wt % 31·8 high-An plg, 22·7 low-An plg, 12·6 opx, 15·5 cpx, 13·5 high-Mg mgt, 0·9 low-Mg mgt, 0·9 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·14 75·42 0·40 13·00 2·03 0·09 0·63 2·02 4·49 1·84 0·09 55·0 200 23·0 177 607 R 0·08 0·05 −0·12 0·02 0·03 −0·09 0·02 0·00 0·00 0·00 −11·7 −6·9 29·3 −29·6 −46·1 R/Daughter 0·00 0·11 −0·01 0·01 0·26 −0·17 0·01 0·00 0·00 0·02 −0·3 0·0 0·6 −0·2 −0·1 Mineral compositions* Partition coefficient** high-An plg 44·81 0·00 35·06 0·39 0·00 0·09 19·08 0·57 0·01 0·00 0·10 1·50 0·06 0·00 0·20 low-An plg 57·90 0·00 26·28 0·40 0·00 0·03 8·84 6·41 0·14 0·00 0·17 4·04 0·02 0·01 0·47 opx 53·79 0·27 1·74 15·92 0·67 26·27 1·30 0·03 0·00 0·00 0·05 0·01 0·40 0·11 0·02 cpx 53·27 0·30 1·11 8·10 1·02 14·88 20·98 0·34 0·00 0·00 0·07 0·17 2·70 0·17 0·10 high-Mg mgt 0·15 7·70 5·87 81·41 0·33 4·54 0·01 0·00 0·00 0·00 0·01 0·01 0·20 0·10 0·12 low-Mg mgt 0·13 9·35 1·99 85·47 1·22 1·84 0·02 0·00 0·00 0·00 0·01 0·01 0·12 0·24 0·10 ilm 0·03 40·17 0·28 56·03 1·10 2·39 0·01 0·00 0·00 0·00 0·01 0·01 0·27 0·60 0·01 apt 0·00 0·00 0·00 0·00 0·00 0·00 56·80 0·00 0·00 43·20 0·00 2·00 40·0 0·64 0·00 The results of only MP-andesite magma are shown here, because we obtained the similar result using HP-andesite magma. Assimilants are taken from Kamiyama et al. (2007) and Ogasawara et al. (1998). * Mineral compositions are taken from analysed data of the phenocrysts in the most mafic MP-type scoria (No. 114), except for apatite (theoretical value). ** These values are taken from Philpotts & Schnetzler (1970), Arth (1976), Okamoto (1979), Pearce & Norry (1979), Luhr & Carmichael (1980), Gill (1981), Fujimaki (1986), Bacon & Druitt (1988), McKenzie & O’Nions (1991), Sisson (1991), Ewart & Griffin (1994), and Bindeman & Davis (2000). Fig. 12. View largeDownload slide Results of fractional and equilibrium crystallization models using rhyolite-MELTS (Gualda et al., 2012). Fractionated phases are plagioclase, orthopyroxene, clinopyroxene, Fe–Ti oxides, quartz, hornblende, and apatite. Although hornblende was included in the model, hornblende crystals were not observed. Only the results for the MP-andesitic (MP-A) magma are shown because results for the HP-andesitic (HP-A) magma are similar. The results for several elements closely match the compositions of rhyolitic (R) and dacitic (D) end-member magmas, but not for Na2O and P2O5. Fig. 12. View largeDownload slide Results of fractional and equilibrium crystallization models using rhyolite-MELTS (Gualda et al., 2012). Fractionated phases are plagioclase, orthopyroxene, clinopyroxene, Fe–Ti oxides, quartz, hornblende, and apatite. Although hornblende was included in the model, hornblende crystals were not observed. Only the results for the MP-andesitic (MP-A) magma are shown because results for the HP-andesitic (HP-A) magma are similar. The results for several elements closely match the compositions of rhyolitic (R) and dacitic (D) end-member magmas, but not for Na2O and P2O5. Fig. 13. View largeDownload slide Rb variation diagrams showing results of assimilation and fractional crystallization models based on mass-balance calculations (Table 4): (a) Y/Rb for the Rayleigh fractional crystallization (FC) model; (b) Y/Rb for the assimilation and fractional crystallization (AFC) model; (c) Zr/Rb for both models; (d) Ba/Rb for both models. Fractionated phases are listed in Table 4. Results are shown only for the MP-andesitic (MP-A) magma using Hidaka granite (No. 9081106D in Kamiyama et al., 2007) as the assimilant; the results in the case of the HP-andesitic (HP-A) magma, as well as eastern Hokkaido granodiorite (No. Y4 in Ogasawara et al., 1998), are similar. Results for Zr/Rb and Ba/Rb values are consistent with the composition of the rhyolitic end-member magma; however, Y/Rb values become strongly depleted with increasing Rb content, and, therefore, the assimilation and fractional crystallization model fails to replicate the rhyolitic magma. Fig. 13. View largeDownload slide Rb variation diagrams showing results of assimilation and fractional crystallization models based on mass-balance calculations (Table 4): (a) Y/Rb for the Rayleigh fractional crystallization (FC) model; (b) Y/Rb for the assimilation and fractional crystallization (AFC) model; (c) Zr/Rb for both models; (d) Ba/Rb for both models. Fractionated phases are listed in Table 4. Results are shown only for the MP-andesitic (MP-A) magma using Hidaka granite (No. 9081106D in Kamiyama et al., 2007) as the assimilant; the results in the case of the HP-andesitic (HP-A) magma, as well as eastern Hokkaido granodiorite (No. Y4 in Ogasawara et al., 1998), are similar. Results for Zr/Rb and Ba/Rb values are consistent with the composition of the rhyolitic end-member magma; however, Y/Rb values become strongly depleted with increasing Rb content, and, therefore, the assimilation and fractional crystallization model fails to replicate the rhyolitic magma. Given these results, we conclude that the silicic magmas were produced from the accumulation of melts generated from a partial melting zone or incompletely crystallized plutons, in which case a heat source (such as basaltic magma) must have underplated the melting zone. The andesitic magmas we have inferred could have acted as such a heat source (e.g. Streck & Grunder, 1999; Hildreth, 2004). Rhyolitic and dacitic magmas In the Kp IV eruption, preexisting rhyolitic and dacitic magmas mixed. In a SiO2 vs 87Sr/86Sr plot (Fig. 9a), the pumice data form a single linear trend in which higher 87Sr/86Sr ratios are strongly correlated with lower SiO2 contents. This result suggests that the Sr isotopic ratios of the silicic end-member magmas that formed the pumice are distinctly different. We infer that the end-member magmas formed simultaneously, either from partial melts of distinct crustal materials or by assimilation of a parental silicic magma. Primary partial melts from crustal melting range in whole-rock chemistry from rhyolitic to dacitic, depending on the degree of melting of mafic to intermediate crustal materials (e.g. Rapp & Watson, 1995; Springer & Seck, 1997). It is also postulated that large silicic magma systems can be generated by accumulation of melts within the crystal mush zone (e.g. Bachmann & Bergantz, 2004; Hildreth, 2004). In either case, crustal materials are heterogeneous enough to generate isotopically distinct melts forming silicic magmas like those of the Kp IV eruption (Matsumoto & Nakagawa, 2010; Cooper et al., 2012; Ellis et al., 2014; Fig. 14a). Accumulation of interstitial melts from large, long-lived crystal mush zones has been recognized in volcanoes with repeated caldera-forming eruptions, such as the Taupo volcanic zone (e.g. Wilson et al., 2006; Bégué et al., 2014) and the Snake River Plain–Yellowstone Plateau province (e.g. Ellis et al., 2014; Wotzlaw et al., 2014, 2015). Kutcharo volcano, where caldera-forming eruptions have occurred since 400 ka (Hasegawa et al., 2011, 2016), may be a similar case. The Sr and Nd isotopic compositions of Cenozoic granodioritic rocks in the area (87Sr/86Sr = 0·7052–0·7071 and 143Nd/144Nd = 0·5126–0·5129 in central Hokkaido; Jahn et al., 2014) are more enriched than those of Kp IV pumices (87Sr/86Sr < 0·7034 and 143Nd/144Nd > 0·5130; Fig. 9), which appears inconsistent with generation of silicic magmas from long-lived crystal mushes that evolve from preexisting crustal materials (e.g. Hildreth et al., 1991). However, in the Kurile arc extending into eastern Hokkaido, basaltic magmas from Quaternary volcanoes show isotopically depleted compositions similar to Kp IV pumices, with low 87Sr/86Sr and high 143Nd/144Nd (Nakamura & Iwamori, 2009). Also, rhyolitic as well as basaltic magmas of Pliocene and Quaternary age share similar unradiogenic compositions (Takanashi et al., 2012). These features imply that crustal materials in eastern Hokkaido have isotopically depleted compositions. Although the exact formation mechanism of the Kp IV magmas is still unclear, we emphasize that the generation and coexistence of multiple silicic magmas are common features in large silicic magma systems. Fig. 14. View largeDownload slide Schematic diagram of the formation and evolution of the large silicic magma system beneath Kutcharo volcano. The left column shows the crustal setting of the silicic magma system, which is detailed in the right column. (a) Formation of the silicic magma chamber. Partial melts of isotopically heterogeneous crustal materials accumulated to form multiple silicic magmas. The dacitic magma (D) was injected into the rhyolitic magma (R), producing a zoned magma chamber in which the dacitic magma stagnated beneath the rhyolitic magma. (b) Status of the magma chamber just before the Kp IV eruption. Multiple andesitic magmas (MP-A, HP-A, and LP-A) ascended from deeper levels and were injected into the zoned magma chamber, triggering the Kp IV eruption. The upper part of the zoned magma was erupted in the early phase, producing Units 1 and 2, and the lower dacitic zone participated in the climactic phase, producing Unit 3. Just before the eruption of Unit 4, a fresh injection of MP-andesitic magma occurred. Fig. 14. View largeDownload slide Schematic diagram of the formation and evolution of the large silicic magma system beneath Kutcharo volcano. The left column shows the crustal setting of the silicic magma system, which is detailed in the right column. (a) Formation of the silicic magma chamber. Partial melts of isotopically heterogeneous crustal materials accumulated to form multiple silicic magmas. The dacitic magma (D) was injected into the rhyolitic magma (R), producing a zoned magma chamber in which the dacitic magma stagnated beneath the rhyolitic magma. (b) Status of the magma chamber just before the Kp IV eruption. Multiple andesitic magmas (MP-A, HP-A, and LP-A) ascended from deeper levels and were injected into the zoned magma chamber, triggering the Kp IV eruption. The upper part of the zoned magma was erupted in the early phase, producing Units 1 and 2, and the lower dacitic zone participated in the climactic phase, producing Unit 3. Just before the eruption of Unit 4, a fresh injection of MP-andesitic magma occurred. Structure and processes of the magma plumbing system The mixing trends between the silicic end-member magmas do not consistently indicate the involvement of andesitic end-member magmas (Figs 3, 7). This lack of consistency strongly suggests that magma mixing between the silicic end-member magmas occurred before mingling with andesitic magmas. Therefore, we infer that the two silicic magmas were part of a large, zoned silicic magma chamber (Fig. 14a). In contrast, the three types of Kp IV scoria yield data plotting along three clearly separate linear trends that converge toward higher SiO2 contents (Figs 7, 11). This feature strongly suggests that the three andesitic, end-member magmas mixed with the silicic magma independently, forming heterogeneous scorias. The HP and MP scorias plot on distinct linear trends that converge toward the composition of the dacitic rather than the rhyolitic pumice (Fig. 7b). This convergence suggests that their corresponding andesitic magmas mixed with less silicic parts of the zoned silicic magma chamber. Therefore, we conclude that a zoned silicic magma chamber, in which dacitic magma lay beneath rhyolitic magma, existed before the Kp IV eruption (Fig. 14a), and that several andesitic magmas were successively injected into the bottom of the chamber (Fig. 14b). Although we cannot determine the sequence of the ascent and injection of multiple andesitic magmas, the following interpretation is consistent with the petrological features of the Kp IV andesitic magmas. The andesitic magmas have similar incompatible element ratios (Fig. 13), and the MP-andesitic magma appears to be capable of producing the HP-andesitic magma by fractional crystallization (Fig. 12). Thus, it is possible that these andesitic magmas were produced separately by similar processes in a common environment, for example, by fractional crystallization of underplating basaltic magma (e.g. Streck & Grunder, 1999). In addition, the Kp IV andesitic magmas are nearly aphyric and their densities (c. 2400–2500 kg/m3, estimated using the model of Bottinga & Weill, 1970) are clearly less than those of crustal materials. Therefore, the Kp IV andesitic magmas may have buoyantly ascended and may have been injected at nearly the same time. Timing of andesitic magma injection and triggering of the caldera-forming eruption Heterogeneous textures in the HP and MP scorias, such as banding and silicic inclusions, suggest that multiple andesitic magmas were injected just before the Kp IV eruption. Because Fe–Ti oxides have much higher elemental diffusion rates than other minerals, the zoning profiles in Fe–Ti oxides can provide information on magmatic processes shortly before the eruption (e.g. Nakamura, 1995; Coombs et al., 2000). We focus on zoning profiles of Fe–Ti oxides in this section. The cores and rims of Fe–Ti oxides, some Type-R and Type-D magnetites in HP scorias from Unit 3, representing the climactic stage of the eruption, exhibit reverse zoning, whereas those of pumices show weak and normal zoning, especially Type-D ones (Fig. 6, Supplementary DataFig. S4). Focusing on their line profiles, several phenocrysts of Type-R and Type-D magnetites in the HP scorias of Unit 3 show clear reverse zoning, with rims enriched in Mg (Fig. 15a), although they have weak zoning in X’usp (the atomic ratio of ulvöspinel in titanomagnetite) (Supplementary DataTable S4). Type-D magnetite phenocrysts in pumices have clear normal zoning in MgO as well as X’usp (Fig. 16). In contrast, there are few grains showing reverse or normal zoning in MP scorias and pumices from Unit 4. Only one ilmenite exhibits reverse zoning, with a Mg-enriched rim (Figs 6, 15b, and Supplementary DataFig. S4). Fig. 15. View largeDownload slide Compositional profiles of Fe–Ti oxide phenocrysts that show reverse zoning; data acquired at intervals of 1–3 µm Diffusion profiles were calculated using the compositionally independent diffusion model of Crank (1975) for a sphere. The diffusion coefficients of Mg are 0·09–1·56 × 10–15 m2/s, estimated using Type-D magnetite (see Fig. 16). Abbreviations: D, partition coefficient of Mg; t, time of diffusion; a, radius of grain. Black solid lines are the best fitting curves for each phenocryst. See the text for discussion. (a) Type-R magnetite phenocrysts from Unit 3. Reverse zoning in these phenocrysts fits curves ranging from Dt/a2 = 0·00125 to Dt/a2 = 0·08, representing very short diffusion times (hours to weeks). (b) Type-R ilmenite phenocryst from Unit 4 with reverse zoning. Although the diffusion rate of Mg in ilmenite is unknown, with the same diffusion coefficient used for magnetite this reverse zoning fits the curves from Dt/a2 = 0·02 to Dt/a2 = 0·04, consistent with magma injection just before the eruption of Unit 4. Fig. 15. View largeDownload slide Compositional profiles of Fe–Ti oxide phenocrysts that show reverse zoning; data acquired at intervals of 1–3 µm Diffusion profiles were calculated using the compositionally independent diffusion model of Crank (1975) for a sphere. The diffusion coefficients of Mg are 0·09–1·56 × 10–15 m2/s, estimated using Type-D magnetite (see Fig. 16). Abbreviations: D, partition coefficient of Mg; t, time of diffusion; a, radius of grain. Black solid lines are the best fitting curves for each phenocryst. See the text for discussion. (a) Type-R magnetite phenocrysts from Unit 3. Reverse zoning in these phenocrysts fits curves ranging from Dt/a2 = 0·00125 to Dt/a2 = 0·08, representing very short diffusion times (hours to weeks). (b) Type-R ilmenite phenocryst from Unit 4 with reverse zoning. Although the diffusion rate of Mg in ilmenite is unknown, with the same diffusion coefficient used for magnetite this reverse zoning fits the curves from Dt/a2 = 0·02 to Dt/a2 = 0·04, consistent with magma injection just before the eruption of Unit 4. Fig. 16. View largeDownload slide Line profiles of X’usp (top) and MgO (bottom) in three Type-D magnetite phenocrysts with normal zoning; electron back-scatter images of these phenocrysts are shown in Supplementary DataFig. S3(f–h). Black solid and broken lines are the best fitting lines for each phenocryst. We estimated the Mg diffusion coefficient for magnetite as follows. First, the diffusion times of X’usp for Type-D magnetite in pumice were estimated using the sphere model of Crank (1975) and the Ti partition coefficient of Freer & Hauptman (1978). The assumed temperature is that of the dacitic end-member magma, 870°C. Second, we fitted the line profiles to Dt/a2 curves. From that result, we estimated the diffusion coefficients using the diffusion times obtained from X’usp. Abbreviations are the same as those in Fig. 15. Fig. 16. View largeDownload slide Line profiles of X’usp (top) and MgO (bottom) in three Type-D magnetite phenocrysts with normal zoning; electron back-scatter images of these phenocrysts are shown in Supplementary DataFig. S3(f–h). Black solid and broken lines are the best fitting lines for each phenocryst. We estimated the Mg diffusion coefficient for magnetite as follows. First, the diffusion times of X’usp for Type-D magnetite in pumice were estimated using the sphere model of Crank (1975) and the Ti partition coefficient of Freer & Hauptman (1978). The assumed temperature is that of the dacitic end-member magma, 870°C. Second, we fitted the line profiles to Dt/a2 curves. From that result, we estimated the diffusion coefficients using the diffusion times obtained from X’usp. Abbreviations are the same as those in Fig. 15. We consider that these zoning profiles were produced mainly by elemental diffusion, due to the narrower width of zonation (10–30 μm) as well as the faster elemental diffusion in magnetite. Therefore, we can estimate the diffusion times of these profiles to evaluate the timescales from magma mixing to eruption. For the estimation, we used a compositionally independent diffusion model assuming a sphere (equation 6.1 in Crank, 1975) and assumed a homogeneous magnetite core mantled by a 1-µm-thick reversely zoned rim, whose compositions are the same as in our data for the initial condition. Lacking published data for the diffusion rate of Mg in magnetite, we estimated the diffusion rate on the basis of the line profiles of MgO and Ti (based on X’usp) in three Type-D magnetite phenocrysts with normal zoning (Fig. 16) as follows. First, we estimated the diffusion times using the line profiles in Xʹusp of these phenocrysts on the basis of the diffusion coefficient of Ti (9·25–9·50 × 10–17 m2/s at 870°C, the approximate temperature of Type-D Fe-Ti oxides; Freer & Hauptman, 1978): c. 20–120 h. Second, we fitted the line profiles in MgO to Dt/a2 curves (D, partition coefficient of Mg; t, time of diffusion; a, radius of grain). From these fitting results and diffusion times of Ti, we estimated the diffusion coefficients of Mg. The corresponding diffusion rate of Mg is 0·09–1·56 × 10–15 m2/s. These values are similar to the diffusion rate of divalent elements (Fe, Co and Mn: 0·44–3·30 × 10–15 m2/s at log fO2 = -11; Aggarwal & Dieckmann, 2002; Van Orman & Crispin, 2010). Using these estimated diffusion coefficients, the diffusion times of Mg in four phenocrysts of Type-R and Type-D magnetite with reverse zonation in Unit 3 range from c. 4 h to 26 days (Fig. 15a, Table 5). Unit 4 yielded no magnetite grains showing compositional zoning (Fig. 6, Supplementary DataFig. S4), but we did characterize a Type-R ilmenite phenocryst (No. 114–56a) that had a thin diffusion mantle (10–20 µm) reversely zoned with respect to Mg (Fig. 15b). The existence of this ilmenite suggests that magma mixing and the eruption of Unit 4 occurred on the same time scale as for Unit 3 (Table 5). Table 5: Estimated diffusion times of Mg in Fe–Ti oxides in scorias of the Kp IV eruption Grain No. Type Unit No. MgO wt % core MgO wt % rim Diffusion time* 355-ox31 Type-D mgt Unit 3 2·08 2·39 4·4 hours – 3·1 days 355-oxad33 Type-R mgt Unit 3 1·70 1·88 6·4 hours – 4·5 days 140-ox46 Type-R mgt Unit 3 1·84 2·19 26·3 hours – 18·3 days 140-ox49 Type-R mgt Unit 3 1·65 1·84 37·0 hours – 25·7 days 114-ox56 Type-R ilm Unit 4 2·32 2·54 17·4 hours – 24·2 days Grain No. Type Unit No. MgO wt % core MgO wt % rim Diffusion time* 355-ox31 Type-D mgt Unit 3 2·08 2·39 4·4 hours – 3·1 days 355-oxad33 Type-R mgt Unit 3 1·70 1·88 6·4 hours – 4·5 days 140-ox46 Type-R mgt Unit 3 1·84 2·19 26·3 hours – 18·3 days 140-ox49 Type-R mgt Unit 3 1·65 1·84 37·0 hours – 25·7 days 114-ox56 Type-R ilm Unit 4 2·32 2·54 17·4 hours – 24·2 days * Diffusion times are estimated using diffusion coefficients of Mg (Kd) = 0·94 x 10–16 and 1·56 x 10–15 m2/s (calculated by Type-D magnetite showing normal zoning in MgO as well as X’usp, see text and Supplementary Data). Table 5: Estimated diffusion times of Mg in Fe–Ti oxides in scorias of the Kp IV eruption Grain No. Type Unit No. MgO wt % core MgO wt % rim Diffusion time* 355-ox31 Type-D mgt Unit 3 2·08 2·39 4·4 hours – 3·1 days 355-oxad33 Type-R mgt Unit 3 1·70 1·88 6·4 hours – 4·5 days 140-ox46 Type-R mgt Unit 3 1·84 2·19 26·3 hours – 18·3 days 140-ox49 Type-R mgt Unit 3 1·65 1·84 37·0 hours – 25·7 days 114-ox56 Type-R ilm Unit 4 2·32 2·54 17·4 hours – 24·2 days Grain No. Type Unit No. MgO wt % core MgO wt % rim Diffusion time* 355-ox31 Type-D mgt Unit 3 2·08 2·39 4·4 hours – 3·1 days 355-oxad33 Type-R mgt Unit 3 1·70 1·88 6·4 hours – 4·5 days 140-ox46 Type-R mgt Unit 3 1·84 2·19 26·3 hours – 18·3 days 140-ox49 Type-R mgt Unit 3 1·65 1·84 37·0 hours – 25·7 days 114-ox56 Type-R ilm Unit 4 2·32 2·54 17·4 hours – 24·2 days * Diffusion times are estimated using diffusion coefficients of Mg (Kd) = 0·94 x 10–16 and 1·56 x 10–15 m2/s (calculated by Type-D magnetite showing normal zoning in MgO as well as X’usp, see text and Supplementary Data). These time scales (hours to weeks) are shorter than those of magma accumulation and silicic mixing (e.g. Druitt et al., 2012; Allan et al., 2013) and are comparable to the timing of mafic injections in previously observed silicic explosive eruptions (e.g. Unzen 1991 eruption, Nakamura, 1995; Soufrière Hills 1995–2002 eruptions, Devine et al., 2003; Shinmoe-dake 2011 eruption, Tomiya et al., 2013). This agreement suggests that injection of mafic magmas triggered the Kp IV caldera-forming eruption. It also indicates that the injection of hot mafic magma could supply volatiles to the silicic magma, creating overpressures that could lead to eruption (e.g. Folch & Martí, 1998). Therefore, our interpretation is that the Kp IV caldera-forming eruption was triggered by andesitic injections. Likewise, the Unit 4 eruption was probably triggered by an injection of andesitic magma after the Unit 3 eruption because the estimated time scale of this process is similar to, or smaller than, the time between eruptions indicated by geologic evidence (several days to two months, as estimated using the degassing model of Riehle et al. (1995) for the Unit 3 pyroclastic flow deposit (Hasegawa et al., 2016)). The change of mafic magma from mainly HP-andesite in Unit 3 to MP-andesite in Unit 4 might also suggest the new injection. Given the number of other reports that mafic magmas mingle with silicic magmas during caldera-forming eruptions (Streck & Grunder, 1999; Reubi & Nicholls, 2005; Wilson et al., 2006; xWright et al., 2011), mafic magma injection may be a common factor in triggering caldera-forming eruptions. CONCLUSIONS We investigated the structure and eruptive processes of the large silicic magma system responsible for the Kp IV caldera-forming activity of Kutcharo volcano. Our major conclusions are as follows: Before the caldera-forming eruption, the magma plumbing system consisted of a large, shallow silicic chamber and a set of deeper, smaller pockets of andesitic magma. The silicic magma chamber was zoned, with rhyolitic magma overlying dacitic magma. The rhyolitic and dacitic magmas could not have been produced by assimilation or fractional crystallization of the andesitic magmas, but formed instead by accumulation of melts generated from crustal materials, for which the andesitic magmas served as a heat source. Likewise, the dacitic magma could not have produced the rhyolitic magma by fractional crystallization, suggesting that distinct silicic magmas were produced contemporaneously from heterogeneous crustal materials. Such generation of multiple silicic magmas may be common in large silicic magma systems. Hours to weeks before the caldera-forming eruption, multiple andesitic magmas were injected into the zoned silicic magma chamber. These injections may have been the final trigger of this catastrophic eruption. ACKNOWLEDGEMENTS We are grateful to Masataka Ikeda for supporting our XRF and EPMA analyses, and to Hidehiko Nomura and Kousuke Nakamura for preparing the thin sections. This manuscript was greatly improved as a result of reviews by Ben Ellis and two anonymous reviewers. We also thank Georg Zellmer for his editorial work. FUNDING This work was supported by JSPS KAKENHI (Grant No. JP15H03745) for M.N. and by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, under its ‘Earthquake and Volcano Hazards Observation and Research Program’ and ‘Integrated Program for Next Generation Volcano Research and Human Resource Development’. SUPPLEMENTARY DATA Supplementary data are available at Journal of Petrology online. REFERENCES Aggarwal S. , Dieckmann R. ( 2002 ). Point defects and cation tracer diffusion in (TixFe1-x)3-δO4. II. Cation tracer diffusion . Physics and Chemistry of Minerals 29 , 707 – 718 . Google Scholar CrossRef Search ADS Allan A. S. R. , Morgan D. J. , Wilson C. J. N. , Millet M.-A. ( 2013 ). From mush to eruption in centuries: assembly of the super-sized Oruanui magma body . Contributions to Mineralogy and Petrology 166 , 143 – 164 . Google Scholar CrossRef Search ADS Andersen D. J. , Lindsley D. H. , Davidson P. M. ( 1993 ). QUILF: a Pascal program to assess equilibria among Fe-Mg-Mn-Ti oxides, pyroxenes, olivine, and quartz. Computers and Geosciences 19 , 1333 – 1350 . Arth J. G. ( 1976 ). Behavior of trace elements during magmatic processes- a summary of theoretical models and their applications . Journal of Research of the U.S. Geological Survey 4 , 41 – 47 . Bachmann O. , Bergantz G. W. ( 2004 ). On the origin of crystal-poor rhyolites: extracted from batholithic crystal mushes . Journal of Petrology 45 , 1565 – 1582 . Google Scholar CrossRef Search ADS Bachmann O. , Bergantz G. W. ( 2008 ). Rhyolites and their source mushes across tectonic settings . Journal of Petrology 49 , 2277 – 2285 . Google Scholar CrossRef Search ADS Bacon C. R. , Druitt T. H. ( 1988 ). Compositional evolution of the zoned calcalkaline magma chamber of Mount Mazama, Crater Lake, Oregon . Contributions to Mineralogy and Petrology 98 , 224 – 256 . Google Scholar CrossRef Search ADS Bacon C. R. , Hirschmann M. M. ( 1988 ). Mg/Mn partitioning as a test for equilibrium between coexisting Fe-Ti oxides . American Mineralogist 73 , 57 – 61 . Baker D. R. , Eggler D. H. ( 1987 ). Compositions of anhydrous and hydrous melts coexisting with plagioclase, augite, and olivine or low-Ca pyroxene from 1atm to 8 kbar: application to the Aleutian volcanic center of Atka . American Mineralogist 72 , 12 – 28 . Bégué F. , Deering C. D. , Gravley D. M. , Kennedy B. M. , Chambefort I. , Gualda G. A. R. , Bachmann O. ( 2014 ). Extraction, storage and eruption of multiple isolated magma batches in the paired Mamaku and Ohakuri eruption, Taupo Volcanic Zone, New Zealand . Journal of Petrology 55 , 1653 – 1684 . Google Scholar CrossRef Search ADS Bindeman I. N. , Davis A. M. ( 2000 ). Trace element partitioning between plagioclase and melt: investigation of dopant influence on partition behavior . Geochimica et Cosmochimica Acta 64 , 2863 – 2878 . Google Scholar CrossRef Search ADS Bottinga Y. , Weill D. F. ( 1970 ). Densities of liquid silicate systems calculatedfrom partial molar volumes of oxide components . American Journal of Science 269 , 169 – 182 . Google Scholar CrossRef Search ADS Brophy J. G. , Dreher S. T. ( 2000 ). The origin of composition gaps at South Sister volcano, central Oregon: implications for fractional crystallization processes beneath active calc-alkaline volcanoes . Journal of Volcanology and Geothermal Research 102 , 287 – 307 . Google Scholar CrossRef Search ADS Brown S. J. A. , Wilson C. J. N. , Cole J. W. , Wooden J. ( 1998 ). The Whakamaru group ignimbrites, Taupo Volcanic Zone, New Zealand: evidence for reverse tapping of a zoned silicic magmatic system . Journal of Volcanology and Geothermal Research 84 , 1 – 37 . Google Scholar CrossRef Search ADS Cashman K. V. , Giordano G. ( 2014 ). Calderas and magma reservoirs . Journal of Volcanology and Geothermal Research 288 , 28 – 45 . Google Scholar CrossRef Search ADS Chamberlain K. J. , Wilson C. J. N. , Wallace P. J. , Millet M.-A. ( 2015 ). Micro-analytical perspectives on the Bishop Tuff and its magma chamber . Journal of Petrology 56 , 605 – 640 . Google Scholar CrossRef Search ADS Coombs M. L. , Eichelberger J. C. , Rutherford M. J. ( 2000 ). Magma storage and magma mixing conditions for the 1953-1974 eruptions of Southwest Trident volcano, Katmai National Park, Alaska . Contributions to Mineralogy and Petrology 140 , 99 – 118 . Google Scholar CrossRef Search ADS Cooper G. F. , Wilson C. J. N. , Millet M.-A. , Baker J. A. , Smith E. G. C. ( 2012 ). Systematic tapping of independent magma chambers during the 1 Ma Kidnappers supereruption . Earth and Planetary Science Letters 313–314 , 23 – 33 . Google Scholar CrossRef Search ADS Crank J. ( 1975 ). The Mathematics of Diffusion . London : Oxford University Press . Deering C. D. , Cole J. W. , Vogel T. A. ( 2011b ). Extraction of crystal-poor rhyolite from a hornblende-bearing intermediate mush: a case study of the caldera-forming Matahina eruption, Okataina volcanic complex . Contributions to Mineralogy and Petrology 161 , 129 – 151 . Google Scholar CrossRef Search ADS Deering C. D. , Bachmann O. , Dufek J. , Gravley D. M. ( 2011a ). Rift-related transition from andesite to rhyolite volcanism in the Taupo Volcanic Zone (New Zealand) controlled by crystal-melt dynamics in mush zones with variable mineral assemblages . Journal of Petrology 52 , 2243 – 2263 . Google Scholar CrossRef Search ADS DePaolo D. J. ( 1981 ). Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization . Earth and Planetary Science Letters 53 , 189 – 202 . Google Scholar CrossRef Search ADS Devine J. D. , Rutherford M. J. , Norton G. E. , Young S. R. ( 2003 ). Magma storage region processes inferred from geochemistry of Fi-Ti oxides in andesitic magma, Soufrière Hills volcano, Montserrat, W.I . Journal of Petrology 44 , 1375 – 1400 . Google Scholar CrossRef Search ADS Druitt T. H. , Costa F. , Deloule E. , Dungan M. , Scaillet B. ( 2012 ). Decadal to monthly timescales of magma transfer and reservoir growth at a caldera volcano . Nature 482 , 77 – 82 . Google Scholar CrossRef Search ADS PubMed Eggins S. M. , Woodhead J. D. , Kinsley L. P. J. , Mortimer G. E. , Sylvester P. , McCulloch M. T. , Hergt J. M. , Handler M. R. ( 1997 ). A simple method for the precise determination of ≥40 trace elements in geological samples by ICPMS using enriched isotope internal standardization . Chemical Geology 134 , 311 – 326 . Google Scholar CrossRef Search ADS Ellis B. S. , Bachmann O. , Wolff J. A. ( 2014 ). Cumulate fragments in silicic ignimbrites: the case of the Snake River Plain . Geology 42 , 431 – 434 . Google Scholar CrossRef Search ADS Ewart A. , Griffin W. L. ( 1994 ). Application of proton-microprobe data to trace-element partitioning in volcanic rocks . Chemical Geology 117 , 251 – 284 . Google Scholar CrossRef Search ADS Folch A. , Martí J. ( 1998 ). The generation of overpressure in felsic magma chambers by replenishment . Earth and Planetary Science Letters 163 , 301 – 314 . Google Scholar CrossRef Search ADS Folkes C. B. , de Silva S. L. , Wright H. M. , Cas R. A. F. ( 2011 ). Geochemical homogeneity of a long-lived, large silicic system; evidence from the Cerro Galán caldera, NW Argentina . Bulletin of Volcanology 73 , 1455 – 1486 . Google Scholar CrossRef Search ADS Freer R. , Hauptman Z. ( 1978 ). An experimental study of magnetite-titanomagnetite interdiffusion . Physics of the Earth and Planetary Interiors 16 , 223 – 231 . Google Scholar CrossRef Search ADS Fujimaki F. ( 1986 ). Fractional crystallization of the basaltic suite of Usu volcano, southwest Hokkaido, Japan, and its relationships with the associated felsic suite . Lithos 19 , 129 – 140 . Google Scholar CrossRef Search ADS Gill J. B. ( 1981 ). Orgenic Andesites and Plate Tectonics . Berlin : Springer . Google Scholar CrossRef Search ADS Gualda G. A. R. , Ghiorso M. S. , Lemons R. V. , Carley T. L. ( 2012 ). Rhyolite-MELTS: a modified calibration of MELTS optimized for silica-rich, fluid-bearing magmatic systems . Journal of Petrology 53 , 875 – 890 . Google Scholar CrossRef Search ADS Hasegawa T. , Nakagawa M. ( 2007 ). Stratigraphy of Early to Middle Pleistocene pyroclastic deposits around Akan caldera, eastern Hokkaido, Japan . The Journal of the Geological Society of Japan 113 , 53 – 72 . (in Japanese with English abstract). Google Scholar CrossRef Search ADS Hasegawa T. , Kishimoto H. , Nakagawa M. ( 2012 ). The eruption history and silicic magma systems of caldera-forming eruptions in eastern Hokkaido, Japan . Journal of Mineralogical and Petrological Sciences 107 , 39 – 43 . Google Scholar CrossRef Search ADS Hasegawa T. , Matsumoto A. , Nakagawa M. ( 2016 ). Evolution of the 120 ka caldera-forming eruption from Kutcharo volcano, east Hokkaido, Japan: Geologic and petrologic evidence for multiple vent systems and rapid generation of pyroclastic flow . Journal of Volcanology and Geothermal Research 321 , 58 – 72 . Google Scholar CrossRef Search ADS Hasegawa T. , Kishimoto H. , Nakagawa M. , Ito J. , Yamamoto T. ( 2009 ). Eruptive history of post-caldera volcanoes of Kutcharo caldera, eastern Hokkaido, Japan, as inferred from tephrostratigraphy in the Konsen and Shari areas for the period 35-12 ka . The Journal of the Geological Society of Japan 115 , 369 – 390 . (in Japanese with English abstract). Google Scholar CrossRef Search ADS Hasegawa T. , Nakagawa M. , Itoh J. , Yamamoto T. ( 2011 ). Depositional ages of Quaternary sequences in Kushiro region, eastern Hokkaido, Japan: Correlations and chronology on the basis of high-resolution tephro-stratigraphy . The Journal of the Geological Society of Japan 117 , 686 – 699 . (in Japanese with English abstract). Google Scholar CrossRef Search ADS Hildreth W. ( 2004 ). Volcanological perspectives on Long Valley, Mammoth Mountain, and Mono Craters: several contiguous but discrete systems . Journal of Volcanology and Geothermal Research 136 , 169 – 198 . Google Scholar CrossRef Search ADS Hildreth W. , Wilson C. J. N. ( 2007 ). Compositional zoning of the Bishop Tuff . Journal of Petrology 48 , 951 – 999 . Google Scholar CrossRef Search ADS Hildreth W. , Halliday A. N. , Christiansen R. L. ( 1991 ). Isotopic and chemical evidence concerning the genesis and contamination of basaltic and rhyolitic magma beneath the Yellowstone Plateau volcanic field . Journal of Petrology 32 , 63 – 138 . Google Scholar CrossRef Search ADS Jahn B.-M. , Usuki M. , Usuki T. , Chung S.-L. ( 2014 ). Generation of Cenozoic granitoids in Hokkaido (Japan): constraints from zircon geochronology, Sr-Nd-Hf isotopic geochemical analyses, and implications for crustal growth . American Journal of Science 314 , 704 – 750 . Google Scholar CrossRef Search ADS Kamiyama H. , Nakajima T. , Kamioka H. ( 2007 ). Magmatic stratigraphy of the tilted Tottabetsu plutonic complex, Hokkaido, North Japan: magma chamber dynamics and pluton construction . Journal of Geology 115 , 295 – 314 . Google Scholar CrossRef Search ADS Kilinc A. , Carmichael I. S. E. , Rivers M. L. , Sack R. O. ( 1983 ). The ferric-ferrous ration of natural silicate liquids equilibrated in air . Contributions to Mineralogy and Petrology 83 , 136 – 140 . Google Scholar CrossRef Search ADS Leonard G. S. , Cole J. W. , Nairn I. A. , Self S. ( 2002 ). Basalt triggering of the c. AD 1305 Kaharoa rhyolite eruption, Tarawera Volcanic Complex, New Zealand . Journal of Volcanology and Geothermal Research 115 , 461 – 486 . Google Scholar CrossRef Search ADS Luhr J. F. , Carmichael I. S. E. ( 1980 ). The Colima volcanic complex, Mexico. I. Post-caldera andesite from Volcán Colima . Contributions to Mineralogy and Petrology 71 , 343 – 372 . Google Scholar CrossRef Search ADS Machida H. , Arai F. ( 2003 ). Atlas of Tephra in and around Japan . Tokyo: University of Tokyo Press . (in Japanese). Matsumoto A. , Nakagawa M. ( 2010 ). Formation and evolution of silicic magma plumbing system: petrology of the volcanic rocks of Usu volcano, Hokkaido, Japan . Journal of Volcanology and Geothermal Research 196 , 185 – 207 . Google Scholar CrossRef Search ADS McKenzie D. , O’Nions R. K. ( 1991 ). Partial melt distributions from inversion of rare earth element concentrations . Journal of Petrology 32 , 1021 – 1091 . Google Scholar CrossRef Search ADS Milner D. M. , Cole J. W. , Wood C. P. ( 2003 ). Mamaku ignimbrite: a caldera-forming ignimbrite erupted from a compositionally zoned magma chamber in Taupo Volcanic Zone, New Zealand . Journal of Volcanology and Geothermal Research 122 , 243 – 264 . Google Scholar CrossRef Search ADS Miyagi I. , Itoh J. , Hoang N. , Morishita Y. ( 2012 ). Magma systems of the Kutcharo and Mashu volcanoes (NE Hokkaido, Japan): petrogenesis of the medium-K trend and the excess volatile problem . Journal of Volcanology and Geothermal Research 231–232 , 50 – 60 . Google Scholar CrossRef Search ADS Moore G. , Vennemann T. , Carmichael I. S. E. ( 1998 ). An empirical model for the solubility of H2O in magmas to 3 kilobars . American Mineralogist 83 , 36 – 42 . Google Scholar CrossRef Search ADS Nakamura H. , Iwamori H. ( 2009 ). Contribution of slab-fluid in arc magmas beneath Japan arcs . Gondwana Research 16 , 431 – 445 . Google Scholar CrossRef Search ADS Nakamura M. ( 1995 ). Continuous mixing of crystal mush and replenished magma in the ongoing Unzen eruption . Geology 23 , 807 – 810 . Google Scholar CrossRef Search ADS Ogasawara M. , Shibata K. , Uchiumi S. ( 1998 ). K-Ar ages and petrological characteristics of granotoid pebbles from the Eocene Beppo Formation in the Kushiro region, Hokkaido, Japan . The Journal of the Geological Society of Japan 104 , 516 – 524 . (in Japanese with English abstract). Google Scholar CrossRef Search ADS Okamoto K. ( 1979 ). Geochemical study on magmatic differentiation of Asama volcano, central Japan . The Journal of the Geological Society of Japan 85 , 525 – 535 . Google Scholar CrossRef Search ADS Orihashi Y. , Maeda J. , Tanaka R. , Zeniya R. , Niida K. ( 1998 ). Sr and Nd isotopic data for the seven GSJ rock reference samples; JA-1, JB-1a, JB-2, JB-3. JG-1a, JGb-1 and JR-1 . Geochemical Journal 32 , 205 – 211 . Google Scholar CrossRef Search ADS Parker D. F. , Ren M. , Adams D. T. , Tsai H. , Long L. E. ( 2012 ). Mid-Tertiary magmatism in western Big Bend National Park, Texas, U.S.A.: evolution of basaltic source regions and generation of peralkaline rhyolite . Lithos 144–145 , 161 – 176 . Google Scholar CrossRef Search ADS Pearce J. A. , Norry M. J. ( 1979 ). Petrogenetic implications of Ti, Zr, Y, and Nb variations in volcanic rocks . Contributions to Mineralogy and Petrology 69 , 33 – 47 . Google Scholar CrossRef Search ADS Philpotts J. A. , Schnetzler C. C. ( 1970 ). Phenocryst-matrix partition coefficients for K, Rb, Sr and Ba, with applications to anorthosite and basalt genesis . Geochimica et Cosmochimica Acta 34 , 307 – 322 . Google Scholar CrossRef Search ADS Putirka K. D. ( 2008 ). Thermometers and barometers for volcanic systems. In: Putirka, K. D. & Tepley F. J., III (eds) Minerals, Inclusions and Volcanic Processes. Mineralogical Society of America and Geochemical Society, Reviews in Mineralogy and Petrology 69 , 61 – 111 . Rapp R. P. , Watson E. B. ( 1995 ). Dehydration melting of metabasalt at 8-32 kbar: implications for continental growth and crust-mantle recycling . Journal of Petrology 36 , 891 – 931 . Google Scholar CrossRef Search ADS Reubi O. , Nicholls I. A. ( 2005 ). Structure and dynamics of a silicic magmatic system associated with caldera-forming eruptions at Batur volcanic field, Bali, Indonesia . Journal of Petrology 46 , 1367 – 1391 . Google Scholar CrossRef Search ADS Riehle J. R. , Miller T. F. , Bailey R. A. ( 1995 ). Cooling, degassing, and compaction of rhyolitic ash flow tuffs: a computational model . Bulletin of Volcanology 57 , 319 – 336 . Google Scholar CrossRef Search ADS Roser B. , Kimura J. , Hisatomi K. ( 2000 ). Whole-rock elemental abundances in sandstones and mudrocks from the Takabe Group, Kii Peninsula, Japan . Geoscience Reports of Shimane University 19 , 101 – 112 . Sack R. O. , Carmichael I. S. E. , Rivers M. , Ghiorso M. S. ( 1981 ). Ferric-ferrous equilibria in natural silicate liquids at 1 bar . Contributions to Mineralogy and Petrology 75 , 369 – 376 . Google Scholar CrossRef Search ADS Sisson T. W. ( 1991 ). Pyroxene-high silica rhyolite trace element partition coefficients measured by ion microprobe . Geochimica et Cosmochimica Acta 55 , 1575 – 1585 . Google Scholar CrossRef Search ADS Smith D. R. , Leeman W. P. ( 1987 ). Petrogenesis of Mount St. Helens dacitic magmas . Journal of Geophysical Research: Solid Earth 92 , 10313 – 10334 . Google Scholar CrossRef Search ADS Springer W. , Seck A. ( 1997 ). Partial fusion of basic granulites at 5 and 15 kbar: implications for the origin of TTG magmas . Contributions to Mineralogy and Petrology 127 , 30 – 45 . Google Scholar CrossRef Search ADS Streck M. J. , Grunder A. L. ( 1997 ). Compositional gradients and gaps in high-silica rhyolites of the Rattlesnake Tuff, Oregon . Journal of Petrology 38 , 133 – 163 . Google Scholar CrossRef Search ADS Streck M. J. , Grunder A. L. ( 1999 ). Enrichment of basalt and mixing of dacite in the rootzone of a large rhyolite chamber: inclusions and pumices from the Rattlesnake Tuff, Oregon . Contributions to Mineralogy and Petrology 136 , 193 – 212 . Google Scholar CrossRef Search ADS Streck M. J. , Grunder A. L. ( 2008 ). Phenocryst-poor rhyolites of bimodal, tholeiitic provinces: The Rattlesnake Tuff and implications for mush extraction models . Bulletin of Volcanology 70 , 385 – 401 . Google Scholar CrossRef Search ADS Sun S.–S. , McDonough W. F. ( 1989 ). Chemical and isotopic systematics of ocean 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 . Google Scholar CrossRef Search ADS Takanashi K. , Kakihara Y. , Ishimoto H. , Shuto K. ( 2012 ). Melting of crustal rocks as a possible origin for Middle Miocene to Quaternary rhyolites of northeast Hokkaido, Japan: constraints from Sr and Nd isotopes and major- and trace-element chemistry . Journal of Volcanology and Geothermal Research 221–222 , 52 – 70 . Google Scholar CrossRef Search ADS Tomiya A. , Miyagi I. , Saito G. , Geshi N. ( 2013 ). Short time scales of magma-mixing processes prior to the 2011 eruption of Shinmoedake volcano, Kirishima volcanic group, Japan . Bulletin of Volcanology 75 , 1 – 19 . Google Scholar CrossRef Search ADS Van Orman J. A. , Crispin K. L. ( 2010 ). Diffusion in oxides . In: Zhang, Y. & Cherniak, D. J. (eds) Diffusion in Minerals and Melts. Mineralogical Society of America and Geochemical Society, Reviews in Mineralogy and Petrology 72 , 757 – 825 . Wilcock J. , Goff F. , Minarik W. G. , Stix J. ( 2013 ). Magmatic recharge during the formation and resurgence of the Valles caldera, New Mexico, USA: evidence from quartz compositional zoning and geothermometry . Journal of Petrology 54 , 635 – 664 . Google Scholar CrossRef Search ADS Wilson C. J. N. , Blake S. , Charlier B. L. A. , Sutton A. N. ( 2006 ). The 26.5 ka Oruanui eruption, Taupo volcano, New Zealand: development, characteristics and evacuation of a large rhyolitic magma body . Journal of Petrology 47 , 35 – 69 . Google Scholar CrossRef Search ADS Wotzlaw J.-F. , Bindeman I. N. , Stern R. A. , D’Abzac F.-X. , Schaltegger U. ( 2015 ). Rapid heterogeneous assembly of multiple magma reservoirs prior to Yellowstone supereruptions . Scientific Reports 5 , doi:10.1038/srep14026. Wotzlaw J.-F. , Bindeman I. N. , Watts K. E. , Schmitt A. K. , Caricchi L. , Schaltegger U. ( 2014 ). Linking rapid magma reservoir assembly and eruption trigger mechanisms at evolved Yellowstone-type supervolcanoes . Geology 42 , 807 – 810 . Google Scholar CrossRef Search ADS Wright H. M. N. , Folkes C. B. , Cas R. A. F. , Cashman K. V. ( 2011 ). Heterogeneous pumice populations in the 2.08-Ma Cerro Galán ignimbrite: imlications for magma recharge and ascent preceding a large-volume silicic eruption . Bulletin of Volcanology 73 , 1513 – 1533 . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Petrology Oxford University Press

Petrology of the 120 ka Caldera-Forming Eruption of Kutcharo Volcano, Eastern Hokkaido, Japan: Coexistence of Multiple Silicic Magmas and their Relationship with Mafic Magmas

Loading next page...
 
/lp/ou_press/petrology-of-the-120-ka-caldera-forming-eruption-of-kutcharo-volcano-zNSH7b0Ft3
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0022-3530
eISSN
1460-2415
D.O.I.
10.1093/petrology/egy043
Publisher site
See Article on Publisher Site

Abstract

Abstract Petrological and geochemical examination of the largest caldera-forming eruption (Kp IV) of Kutcharo volcano, eastern Hokkaido, Japan, were undertaken in order to understand the magma genesis and eruptive processes of a large silicic magma system. The eruption started with an ash and pumice fall, followed by voluminous pyroclastic flows. Juvenile materials are mainly porphyritic pumice and a small amount of heterogeneous, nearly aphyric, scoria is contained in the pyroclastic flow deposits. On the basis of mineral, whole-rock and matrix glass chemistry, two silicic magmas (rhyolitic and dacitic) and three intermediate (andesitic) magmas were identified. Temporal variations in the whole-rock and matrix glass chemistry of the juvenile materials indicate that the activity originated from a large zoned silicic magma chamber in which dacitic magma had stagnated beneath more voluminous rhyolitic magma. During the generation of pyroclastic flows, andesitic magmas were sequentially injected into the zoned chamber, resulting in the eruption of small amounts of heterogeneous products along with voluminous silicic magmas. The rhyolite-MELTS program, mass-balance calculations, and Rayleigh fractionation models cannot explain the generation of two silicic magmas by fractional crystallization of coexisting andesitic magmas. In addition, the higher 87Sr/86Sr ratio of dacitic pumice suggests that the dacitic magma was not a parent of the rhyolitic magma. Therefore, we infer that both the rhyolitic and dacitic magmas were produced by the accumulation of interstitial melts generated from crustal materials with heterogeneous Sr isotopic compositions. It may be common that large silicic magma systems that produce caldera-forming eruptions are composed of multiple silicic magmas that are produced from an extensive area of heterogeneous crust. Mg and Ti diffusion profiles in Fe–Ti oxide phenocrysts indicate that three different andesitic magmas were successively injected into the zoned chamber hours to weeks before the eruptions. Their injection may have triggered the Kp IV eruptive activities. INTRODUCTION Understanding the magma system and eruptive processes of large, caldera-forming eruptions is important, not only to clarify the formation and modification processes of the continental crust, but also to prepare against these catastrophic events. Many studies of caldera volcanism have revealed the presence of a zoned magma chamber in which small amounts of intermediate or mafic magma are episodically injected into a predominantly silicic resident magma (e.g. Bacon & Druitt, 1988; Streck & Grunder, 1999; Milner et al., 2003; Wilson et al., 2006; Deering et al., 2011b; Wilcock et al., 2013). Studies have also investigated the petrogenetic relationships between silicic and mafic magmas (e.g. Streck & Grunder, 2008; Deering et al., 2011a) and the eruptive activities caused by mafic injections during caldera formation (e.g. Leonard et al., 2002; Milner et al., 2003). Previous studies have typically assumed that the silicic magma involved in caldera-forming eruptions is either homogeneous or has compositional and thermal gradients imposed by differentiation (e.g. Streck & Grunder, 1997; Brown et al., 1998; Wilson et al., 2006; Hildreth & Wilson, 2007; Chamberlain et al., 2015). Other studies have reported evidence of silicic magma mixing (e.g. Folkes et al., 2011; Wotzlaw et al., 2014) and the timescale of mixing (e.g. Druitt et al., 2012; Allan et al., 2013), and have proposed schematic models of magma reservoirs beneath caldera volcanoes (e.g. Hildreth, 2004; Cashman & Giordano, 2014). Several studies have documented the existence in caldera systems involving multiple, separately evolved silicic magmas (e.g. Brown et al., 1998; Reubi & Nicholls, 2005; Cooper et al., 2012; Bégué et al., 2014; Ellis et al., 2014; Wotzlaw et al., 2014). Moreover, caldera volcanoes and their associated silicic magma systems may vary in different tectonic settings (e.g. Bachmann & Bergantz, 2008; Chamberlain et al., 2015). A fuller understanding of the common processes and characteristic features of large silicic magma systems thus requires further comprehensive studies of the chemical and structural variations within the silicic magma chambers of caldera volcanoes from different tectonic settings. Kutcharo volcano, in eastern Hokkaido, Japan, is one of the Quaternary caldera volcanoes in the southern part of the Kurile arc (Fig. 1; Hasegawa & Nakagawa, 2007; Hasegawa et al., 2009, 2012). The volcano started its activity at c. 400 ka and then was quiescent until c. 200 ka (Hasegawa et al., 2011). Since 200 ka, at least eight major episodes of explosive activity have occurred (named Kp VIII to Kp I in chronological order), of which the Kp IV eruption at c. 120 ka, associated with the Kutcharo Pumice Flow Deposit IV (Machida & Arai, 2003), was the largest and most explosive. The Kp IV eruption, which produced c. 175 km3 (bulk unit volume) of eruptive products, can be divided into four phases after minor preceding activity. The Kp IV juvenile materials consist of rhyolitic pumice and a small amount of scoria. Although rhyolitic magma was predominant throughout the Kp IV eruption, small amounts of intermediate magma erupted just after they were injected into the rhyolitic magma (Hasegawa et al., 2016). Fig. 1. View largeDownload slide Index map of eastern Hokkaido, Japan (after Hasegawa et al., 2016). The Kutcharo Pumice Flow Deposit IV (Kp IV) is distributed throughout eastern Hokkaido (gray shaded areas). The broken line shows the border between the NW and SE areas detailed in Fig. 2. Fig. 1. View largeDownload slide Index map of eastern Hokkaido, Japan (after Hasegawa et al., 2016). The Kutcharo Pumice Flow Deposit IV (Kp IV) is distributed throughout eastern Hokkaido (gray shaded areas). The broken line shows the border between the NW and SE areas detailed in Fig. 2. In this study, we report the petrological and geochemical features of juvenile materials from the Kp IV eruption and document the presence of two distinct silicic magmas in the main silicic magma body. We also present evidence bearing on the petrogenetic relationships between these magmas to clarify the formation of the magma plumbing system for the caldera-forming eruption. Finally, we estimate the timescales between mixing of mafic magmas and eruption using geospeedometry for Fe–Ti oxides, to discuss the trigger of caldera-forming eruption. KP IV ERUPTION AND SAMPLES In a recent publication, we described the Kp IV eruption sequence in detail and documented a previously unrecognized small eruptive deposit (Pre-Kp IV) beneath the Kp IV eruptive deposits, that is topped by a thin paleosol (Hasegawa et al., 2016). The Kp IV eruptive products are divided into the following four units in ascending order: Unit 1, a widespread ash fall deposit (phreatoplinian eruption); Unit 2, a pumice fall deposit (subplinian eruption); Unit 3, a voluminous pyroclastic flow deposit (caldera-forming eruption); Unit 4, a small-scale scoria-rich pyroclastic flow deposit (Fig. 2). A small amount of scoria is included in Unit 3 (<0·85 km3 within 170 km3 bulk unit volume), and a greater proportion is present in Unit 4 (c. 0·5 km3 within 1 km3 bulk unit volume) (Hasegawa et al., 2016). Units 1, 2, and 3 are sequential deposits, suggesting that the eruptive activity reached a climactic phase (Unit 3) almost immediately. There is, however, a short, but well-defined, break between Unit 3 and Unit 4 (Hasegawa et al., 2016). Fig. 2. View largeDownload slide Generalized columnar section of Kp IV deposits (Hasegawa et al., 2016). Units 1 to 3 were deposited successively, whereas Unit 4 cuts into Unit 3 as shown by the truncation of gas segregation pipes. The matrix color of Unit 3 changes upsection from white to brown or black. Northwest of the caldera (NW area), Unit 3 includes a few scoria clasts, whereas Unit 4 is highly scoriaceous. In contrast, southeast of the caldera (SE area), no scoria clasts are found in Unit 3. Abbreviations: pfl, pyroclastic flow; pfa, pumice fall; wp, white pumice; bp, brown pumice; sc, scoria. Fig. 2. View largeDownload slide Generalized columnar section of Kp IV deposits (Hasegawa et al., 2016). Units 1 to 3 were deposited successively, whereas Unit 4 cuts into Unit 3 as shown by the truncation of gas segregation pipes. The matrix color of Unit 3 changes upsection from white to brown or black. Northwest of the caldera (NW area), Unit 3 includes a few scoria clasts, whereas Unit 4 is highly scoriaceous. In contrast, southeast of the caldera (SE area), no scoria clasts are found in Unit 3. Abbreviations: pfl, pyroclastic flow; pfa, pumice fall; wp, white pumice; bp, brown pumice; sc, scoria. Juvenile materials of the Kp IV eruption can be divided into white pumice, brown pumice and scoria. The white and brown pumices have indistinguishable petrographic characteristics and matrix glass chemistry (Hasegawa et al., 2016), and are simply referred to as ‘pumice’ hereafter. The scoria, our designation for heterogeneous samples including mafic portions, can be divided according to the P2O5 content of the matrix glass into low-P2O5 (LP: SiO2 = 58–74 wt % and P2O5 = 0·1–0·3 wt %), medium-P2O5 (MP: SiO2 = 52–59 wt % and P2O5 = 0·.4–0·8 wt %), and high-P2O5 (HP: SiO2 = 59–71 wt % and P2O5 = 0·3–0·9 wt %) scoria (Fig. 3). In Unit 3, HP scoria is the main component, but small amounts of LP and MP scorias are also found. In contrast, only MP scoria occurs in Unit 4. HP scoria in Unit 3 is finely inter-banded with pumice, and MP scoria in Unit 4 commonly contains silicic inclusions resembling pumice in hand specimen and thin section. Both LP and MP scorias in Unit 3 are relatively homogeneous (Hasegawa et al., 2016; see Supplementary Data; supplementary data are available for downloading at http://www.petrology.oxfordjournals.org). Fig. 3. View largeDownload slide P2O5 vs SiO2 plot for matrix glass chemistry of Kp IV juvenile materials (Hasegawa et al., 2016). Pumice has relatively homogeneous compositions, regardless of eruptive unit. Scoria exhibits wide variations, diverging toward low SiO2 and converging toward high SiO2. Scoria can be divided into three types on the basis of its P2O5 content: lower P content (LP), medium P content (MP), and higher P content (HP). The compositional gaps between them suggest the existence of three distinct magmas. Abbreviations: pm, pumice; sc, scoria. Fig. 3. View largeDownload slide P2O5 vs SiO2 plot for matrix glass chemistry of Kp IV juvenile materials (Hasegawa et al., 2016). Pumice has relatively homogeneous compositions, regardless of eruptive unit. Scoria exhibits wide variations, diverging toward low SiO2 and converging toward high SiO2. Scoria can be divided into three types on the basis of its P2O5 content: lower P content (LP), medium P content (MP), and higher P content (HP). The compositional gaps between them suggest the existence of three distinct magmas. Abbreviations: pm, pumice; sc, scoria. ANALYTICAL METHODS We collected samples from all units at representative outcrops on all sides of the Kutcharo caldera (Supplementary Data). All analyses were conducted at Hokkaido University and all the analytical results are listed in the Supplementary Data. Modal compositions were determined on thin sections using an automatic point-counter on the basis of 3000 counts per thin section. Mineral compositions were determined using a JEOL-8800R electron probe microanalyser under the following operating conditions: 15 kV accelerating voltage, 10 nA beam current for plagioclase, and 20 nA beam current for pyroxene and Fe–Ti oxides. The glass data are the same as those of Hasegawa et al. (2016), using 10 nA beam current and a 2 µm beam scanning an area 10 µm square. To avoid Na migration, X-ray data for Na were counted only for the first 30 s. All analyses were corrected using the oxide ZAF method. Whole-rock compositions were determined by X-ray fluorescence using a Spectris MagiX PRO system with a Rh tube. Major and trace elements were measured using glass beads prepared by fusing the sample with an alkali flux (a 4:1 mixture of lithium tetraborate and lithium metaborate); major and trace elements were measured in 181 samples diluted to 1:2 and only major elements were measured in 32 samples diluted to 1:10. Trace and rare earth elements (REE) were determined in 19 samples by inductively coupled plasma mass spectrometry (ICP-MS) using a Thermo Electron X series instrument, following the method of Eggins et al. (1997). Of these 19 samples, eight less silicic samples were prepared using the acid digestion method. About 50 mg of powdered sample was digested in a 2:1 mixture of HF and HClO4. After drying, the residual samples were dissolved with HNO3 and small amounts of HCl and HF. The remaining 11 silicic samples were prepared by the alkali fusion method (Roser et al., 2000). About 100 mg of powder was digested in a 1:1 mixture of HCl and HF. After drying, the residual samples were fused with 500 mg of Na2CO3 at 1100°C, then dissolved with HNO3, HCl, and a small amount of HF. Solutions for ICP-MS were prepared in 5% HNO3 and small amounts of HCl and HF with a dilution factor of 10 000 (acid digestion) and 20 000 (alkali fusion). The measurement precision and accuracy were monitored by repeated analyses of the JB-1a standard. Reproducibility of measurements was generally within less than 10% (see the Supplementary Data). Isotopic analyses of 15 samples were carried out using a Finnigan MAT262 mass spectrometer according to the methods reported by Orihashi et al. (1998). The 87Sr/86Sr and 143Nd/144Nd ratios were corrected to 86Sr/88Sr = 0·1194 and 146Nd/144Nd = 0·7219, respectively. The analytical accuracy and reproducibility in this study were monitored using the NIST-SRM987 standard for Sr and the GSJ-JNdi-1 standard for Nd. The results were 87Sr/86Sr = 0·710220 (2σ = 0·000042; n = 3) and 143Nd/144Nd = 0·512129 (2σ = 0·000019; n = 3). The house standard samples show good standard deviation values (87Sr/86Sr = 0·704239, 2σ = 0·000022, n = 3; 143Nd/144Nd = 0·512823, 2σ = 0·000021, n = 3). Pumice clasts in Unit 1 and MP and LP scoria clasts in Unit 3 were too small to determine their mineral and whole-rock chemistry (<1 cm diameter). In addition, the whole-rock chemistry results for pumice from Pre-Kp IV and Unit 2 were disregarded because the material was slightly altered (>3 wt % loss-on-ignition). PETROGRAPHY AND MINERAL CHEMISTRY Crystal contents of the pumice range from 2 to 16 vol %. The scorias contain less than 6 vol % crystals and crystallinity decreases with decreasing whole-rock SiO2 content (Fig. 4, Table 1). The crystal assemblage in all juvenile materials consists of plagioclase, orthopyroxene, Fe–Ti oxides, and minor clinopyroxene. The matrix of the pumice is glassy, varies from colorless to brownish and has a relatively homogeneous chemical composition (Hasegawa et al., 2016; Fig. 3). The groundmass of the HP scoria has a hyalopilitic texture and consists of dark brown glass and plagioclase. In both the MP and LP scoria the groundmass displays an intersertal texture and is composed of plagioclase, orthopyroxene and dark brown glass. Table 1: Representative modal compositions of Kp IV and Pre-Kp IV juvenile materials Sample No. 271 29 229 254 90 355 99 118 114 Occurrence wp wp wp bp HP sc HP sc wp MP sc MP sc Eruptive unit Pre-KpIV 2 3 3 3 3 4 4 4 Phenocryst mode (vol· %) Pl 12·0 4·4 8·2 3·1 4·0 1·5 7·4 3·2 0·3 Opx 0·5 0·1 0·2 0·3 0·4 0·9 0·5 tr 0·2 Cpx tr tr tr 0·9 tr tr 0·2 tr tr Opq 0·2 0·2 0·3 0·5 0·4 0·2 0·8 0·1 0·3 Total 12·7 4·7 8·7 4·8 4·9 2·6 8·9 3·3 0·7 Sample No. 271 29 229 254 90 355 99 118 114 Occurrence wp wp wp bp HP sc HP sc wp MP sc MP sc Eruptive unit Pre-KpIV 2 3 3 3 3 4 4 4 Phenocryst mode (vol· %) Pl 12·0 4·4 8·2 3·1 4·0 1·5 7·4 3·2 0·3 Opx 0·5 0·1 0·2 0·3 0·4 0·9 0·5 tr 0·2 Cpx tr tr tr 0·9 tr tr 0·2 tr tr Opq 0·2 0·2 0·3 0·5 0·4 0·2 0·8 0·1 0·3 Total 12·7 4·7 8·7 4·8 4·9 2·6 8·9 3·3 0·7 wp, white pumice; bp, brown pumice; sc: scoria. tr : <0·1. Phenocryst abbreviations: Pl, plagioclase; Opx, orthopyroxene; Cpx, clinopyroxene; Opq, opaques. Table 1: Representative modal compositions of Kp IV and Pre-Kp IV juvenile materials Sample No. 271 29 229 254 90 355 99 118 114 Occurrence wp wp wp bp HP sc HP sc wp MP sc MP sc Eruptive unit Pre-KpIV 2 3 3 3 3 4 4 4 Phenocryst mode (vol· %) Pl 12·0 4·4 8·2 3·1 4·0 1·5 7·4 3·2 0·3 Opx 0·5 0·1 0·2 0·3 0·4 0·9 0·5 tr 0·2 Cpx tr tr tr 0·9 tr tr 0·2 tr tr Opq 0·2 0·2 0·3 0·5 0·4 0·2 0·8 0·1 0·3 Total 12·7 4·7 8·7 4·8 4·9 2·6 8·9 3·3 0·7 Sample No. 271 29 229 254 90 355 99 118 114 Occurrence wp wp wp bp HP sc HP sc wp MP sc MP sc Eruptive unit Pre-KpIV 2 3 3 3 3 4 4 4 Phenocryst mode (vol· %) Pl 12·0 4·4 8·2 3·1 4·0 1·5 7·4 3·2 0·3 Opx 0·5 0·1 0·2 0·3 0·4 0·9 0·5 tr 0·2 Cpx tr tr tr 0·9 tr tr 0·2 tr tr Opq 0·2 0·2 0·3 0·5 0·4 0·2 0·8 0·1 0·3 Total 12·7 4·7 8·7 4·8 4·9 2·6 8·9 3·3 0·7 wp, white pumice; bp, brown pumice; sc: scoria. tr : <0·1. Phenocryst abbreviations: Pl, plagioclase; Opx, orthopyroxene; Cpx, clinopyroxene; Opq, opaques. Fig. 4. View largeDownload slide Plot of phenocryst content vs whole-rock SiO2 content. Scoria is nearly aphyric, whereas pumice is porphyritic with phenocryst content decreasing with decreasing SiO2. Abbreviations are the same as those in Fig. 2. Solid and broken lines are regression lines of juvenile materials from Unit 3 and Unit 4, respectively. The whole-rock data for pumice from Unit 2 and Pre-Kp IV are shown for reference, but were not used in our analysis because of their high loss on ignition (>3 wt %). Fig. 4. View largeDownload slide Plot of phenocryst content vs whole-rock SiO2 content. Scoria is nearly aphyric, whereas pumice is porphyritic with phenocryst content decreasing with decreasing SiO2. Abbreviations are the same as those in Fig. 2. Solid and broken lines are regression lines of juvenile materials from Unit 3 and Unit 4, respectively. The whole-rock data for pumice from Unit 2 and Pre-Kp IV are shown for reference, but were not used in our analysis because of their high loss on ignition (>3 wt %). Plagioclase Plagioclase is a major phenocryst in all samples, often forming crystal clots with Fe–Ti oxides. Phenocrysts in the Kp IV pumice are usually clear and euhedral to subhedral (<1 mm long). The anorthite contents (An = Ca × 100/(Ca + Na + K)) in the cores of the phenocrysts in pumice and scoria exhibit a clear peak at 40–44, although a few phenocrysts have higher An contents (An = 48–80; Fig. 5). Pumice samples with lower SiO2 contents have more anorthitic phenocrysts. The scoria samples contain a few grains of calcic plagioclase (An = 80–95; Fig. 5b) that are usually <0·5 mm long and surrounded by brown glass with an intersertal and hyalopilitic texture. Fig. 5. View largeDownload slide Histograms of core compositions of phenocryst minerals in (a) pumice and (b) scoria samples. Italic numbers are whole-rock SiO2 contents (wt%). The arrows indicate outlying values for Plagioclase An content and Mg/Mn in magnetite and ilmenite, as discussed in the text. Abbreviations: An = Ca × 100/(Ca + Na + K); Mg# = Mg × 100/(Fe + Mg). Fig. 5. View largeDownload slide Histograms of core compositions of phenocryst minerals in (a) pumice and (b) scoria samples. Italic numbers are whole-rock SiO2 contents (wt%). The arrows indicate outlying values for Plagioclase An content and Mg/Mn in magnetite and ilmenite, as discussed in the text. Abbreviations: An = Ca × 100/(Ca + Na + K); Mg# = Mg × 100/(Fe + Mg). The high-An plagioclase (An = 48–80) in pumice and scoria often shows normal zoning and the low-An plagioclase exhibits slight reverse zoning (Fig. 6). Most of the calcic plagioclase shows weak zoning. Fig. 6. View largeDownload slide Core and rim compositions of phenocrystic minerals in Kp IV pumice and scoria samples. In all samples, An-rich plagioclase and magnesian pyroxene show normal zoning, and An-poor plagioclase and less magnesian pyroxene show reverse zoning. High-An plagioclase phenocrysts in scoria appear to be weakly zoned. For Fe–Ti oxides, low-Mg phenocrysts in scoria exhibit slight reverse zoning and those in pumice show weak zoning. Abbreviations: N, normal zoning; R, reverse zoning. Fig. 6. View largeDownload slide Core and rim compositions of phenocrystic minerals in Kp IV pumice and scoria samples. In all samples, An-rich plagioclase and magnesian pyroxene show normal zoning, and An-poor plagioclase and less magnesian pyroxene show reverse zoning. High-An plagioclase phenocrysts in scoria appear to be weakly zoned. For Fe–Ti oxides, low-Mg phenocrysts in scoria exhibit slight reverse zoning and those in pumice show weak zoning. Abbreviations: N, normal zoning; R, reverse zoning. The core compositions of phenocrysts in the Pre-Kp IV pumice show a broad peak, distinctly different from those in Kp IV pumice, with An values of 35–58, plus a few outliers at An = 67 and 77 (Fig. 5a). Pyroxene Phenocrysts of clear, euhedral to subhedral orthopyroxene and clinopyroxene (<1 mm long) are present in all samples and often form crystal clots with Fe–Ti oxides. Orthopyroxene phenocrysts often include apatite, whereas clinopyroxene phenocrysts often contain melt inclusions. The cores of orthopyroxene phenocrysts in Kp IV pumice and scoria range in magnesium number (Mg# = Mg × 100/(Fe + Mg)) from 67 to 76, with a clear peak at Mg# = 67–69 (Fig. 5a). The more magnesian orthopyroxene phenocrysts (Mg# > 69) are often found in less silicic pumice and some scoria samples (Fig 5), and sometimes form crystal clots with magnesian clinopyroxene (Mg# = 73–78). Those more magnesian orthopyroxene phenocrysts often show normal zoning and a few show weak zoning (Fig. 6), whereas the less magnesian ones (Mg# < 69) exhibit weak to slightly reverse zoning. Clinopyroxene phenocrysts in Kp IV samples have core compositions of Mg# = 68–78, and most are around Mg# = 76 (Fig. 5). The phenocrysts with Mg# < 72 exhibit reverse zoning, whereas those with Mg# > 73 have normal to weak zoning (Fig. 6). In Pre-Kp IV pumice, orthopyroxene phenocrysts have slightly more iron-rich compositions than in Kp IV pumice (Mg# = 64–71; Fig. 5a), and clinopyroxene phenocrysts have homogeneous core compositions of Mg# = 72–74. Fe–Ti oxides Phenocrysts of Fe–Ti oxides (<0·6 mm long) are composed of Ti-magnetite and ilmenite (Fig. 5). Pumice and scoria from Kp IV include both oxide minerals as major phenocryst phases. Ti-magnetite phenocrysts in all Kp IV rocks have homogeneous core compositions in the range Mg/Mn = 2·4–2·9, although some phenocrysts in the less silicic Unit 3 pumice are more magnesian (Mg/Mn = 2·9–3·4 and 5·0–7·5 in sample No. 168; Fig. 5a). Samples of HP scoria from Unit 3 also have magnesian magnetite in the range Mg/Mn = 3–17 (Fig. 5b). MP scoria samples from Unit 4 also include a few highly magnesian phenocrysts (Mg/Mn = 11–25). In pumice samples, low-Mg magnetite (Mg/Mn = 2·4–3·4) shows weak zoning and higher-Mg magnetite (Mg/Mn > 3·4) has normal zoning (Fig. 6). In scoria, low-Mg magnetite shows weak to slightly reverse zoning. In Pre-Kp IV pumice samples, the cores of magnetite phenocrysts are slightly less magnesian than in Kp IV pumice samples (Mg/Mn = 2·3–2·7; Fig. 5a). Ilmenite phenocrysts have similar compositional variations. In Kp IV rocks their compositional range is Mg/Mn = 3·5–4·2. The less silicic pumice in Unit 3 (sample No. 168), pumice sample No. 99 in Unit 4, and some HP scoria samples from Unit 3 contain ilmenite phenocrysts with Mg/Mn = 4·3–13 (Fig. 5). In pumice samples, low-Mg phenocrysts (Mg/Mn < 5) show weak zoning and higher-Mg ones (Mg/Mn > 6) show normal zoning (Fig. 6). In scoria, the low-Mg phenocrysts (Mg/Mn < 5) exhibit weak to slightly reverse zoning and higher-Mg ones (Mg/Mn > 5) show normal zoning. In Pre-Kp IV pumice, ilmenite cores have less magnesian compositions than in Kp IV pumice samples (Mg/Mn = 3·0–3·5 and 4·2; Fig. 5a). GEOCHEMISTRY Major and trace elements The SiO2 contents of matrix glass range from 74 to 80 wt % in Kp IV pumice and from 52·5 to 74 wt % in Kp IV scoria (Fig. 3). These data do not form a single linear trend in a SiO2 vs P2O5 plot. Whole-rock SiO2 contents of Kp IV pumice samples are in the range 72·9–74·8 wt % (Fig. 7a, Table 2). Although pumice samples from Unit 3 vary more widely in composition than those from Unit 4, the data from both units form a single linear trend in SiO2 variation diagrams (Fig. 7b). Whole-rock SiO2 contents are 63·4–72 ·5 wt % in HP scoria and 58·5–70·7 wt % in MP scoria (Fig. 7a). Although both types of scoria lie on the same linear trend in SiO2 variation diagrams with FeO*, Al2O3 and Y, they form two distinct linear trends in diagrams with P2O5, V and Sr. HP scoria have higher P2O5 contents than MP scoria at all SiO2 contents, consistent with their matrix glass chemistry (Fig. 3). Although the whole-rock compositions of these scorias extend to the SiO2 range of the pumice, their linear trends converge at the dacitic rather than rhyolitic end of the pumice compositions (Fig. 7a). Fig. 7. View largeDownload slide Variation diagrams for the whole-rock compositions of Kp IV juvenile materials. Each diagram includes a 2σ error symbol. (a) Harker diagrams of representative major oxides and trace elements for all samples. Data for Kp IV juvenile materials form two linear trends on several diagrams, converging toward high SiO2 values. These trends correspond to the matrix glass compositions of the HP and MP scoria. (b) Details of selected diagrams emphasizing pumice samples. Pumice samples form single linear trends that differ from those of scoria. The scoria trends converge toward the dacitic rather than the rhyolitic end of the pumice data. (c) Ratio diagrams of incompatible elements plotted against Rb content. Ratios are approximately constant with increasing incompatible element content, except for Y/Rb. Fig. 7. View largeDownload slide Variation diagrams for the whole-rock compositions of Kp IV juvenile materials. Each diagram includes a 2σ error symbol. (a) Harker diagrams of representative major oxides and trace elements for all samples. Data for Kp IV juvenile materials form two linear trends on several diagrams, converging toward high SiO2 values. These trends correspond to the matrix glass compositions of the HP and MP scoria. (b) Details of selected diagrams emphasizing pumice samples. Pumice samples form single linear trends that differ from those of scoria. The scoria trends converge toward the dacitic rather than the rhyolitic end of the pumice data. (c) Ratio diagrams of incompatible elements plotted against Rb content. Ratios are approximately constant with increasing incompatible element content, except for Y/Rb. Incompatible element ratios differ between pumice and scoria (Fig. 7c). The Y/Rb ratios of all samples decrease with increasing Rb content in a continuous trend, whereas the Zr/Rb and Ba/Rb ratios are approximately constant with increasing Rb content. Rare earth element patterns and isotopic compositions Chondrite-normalized REE patterns of all samples display weak light-REE enrichment and a negative Eu anomaly, more so in pumice than in scoria (Fig. 8, Table 2). The REE concentrations in pumice increase with increasing whole-rock SiO2 content. The REE patterns of MP and HP scorias are sub-parallel, but the HP scoria have higher average REE concentrations. Table 2: Representative geochemical compositions of Kp IV and Pre-Kp IV juvenile materials Sample No. 271* 17* 168 150 99 229 355 85 140 297 90 114 112 118 121 110 HB-2 Occurrence pm pm pm pm pm pm sc HP sc HP sc HP sc HP sc HP sc MP sc MP sc MP sc MP sc MP Eruptive unit Pre-Kp IV 2 3 3 4 3 3 3 3 3 3 4 4 4 4 4 House standard 2σ Major elements (wt %, XRF) N=14 SiO2 73·4 73·8 73·5 74·7 75·3 75·4 63·2 64·0 65·8 69·2 71·8 58·2 60·1 64·3 67·4 69·2 53·6 0·32 TiO2 0·49 0·52 0·53 0·49 0·46 0·47 0·93 0·87 0·81 0·68 0·61 1·02 0·95 0·83 0·70 0·67 0·76 0·01 Al2O3 14·9 14·1 14·2 13·9 13·5 13·5 15·3 14·8 15·0 14·3 14·2 15·6 15·2 14·9 15·3 14·2 17·4 0·12 FeO* 2·45 2·56 2·60 2·30 2·30 2·14 6·23 5·98 5·17 4·19 3·30 9·21 8·17 6·47 4·83 4·53 9·29 0·19 MnO 0·12 0·14 0·15 0·13 0·14 0·13 0·20 0·20 0·20 0·17 0·15 0·21 0·20 0·18 0·16 0·16 0·18 0·003 MgO 0·62 0·74 0·80 0·59 0·65 0·58 2·47 2·17 1·75 1·30 1·02 3·61 3·20 2·45 1·73 1·62 6·44 0·06 CaO 2·64 2·58 2·80 2·42 2·30 2·17 5·89 5·46 4·96 3·66 3·20 7·39 6·88 5·64 4·83 4·10 9·56 0·17 Na2O 4·07 4·30 4·51 4·51 4·51 4·51 3·80 3·75 4·04 4·12 4·33 3·05 3·30 3·60 3·87 4·13 2·40 0·03 K2O 1·67 1·72 1·64 1·71 1·83 1·76 1·03 1·06 1·16 1·42 1·59 0·77 0·93 1·15 1·28 1·42 0·24 0·005 P2O5 0·10 0·12 0·13 0·10 0·11 0·10 0·62 0·54 0·46 0·31 0·21 0·42 0·42 0·34 0·25 0·25 0·11 0·002 Total 100·37 100·59 100·76 100·87 101·07 100·72 99·63 98·78 99·37 99·32 100·41 99·44 99·36 99·87 100·40 100·23 Trace elements (ppm, XRF) N=14 Sc 17·9 14·6 17·4 16·6 30·0 20·5 21·2 38·8 33·7 31·7 25·0 25·9 33·3 2·3 V 24·0 23·8 21·4 19·7 114 71·5 42·3 253 214 158 95·8 94·9 217 4·6 Cr 3·17 2·66 3·40 2·35 5·12 2·65 3·57 19·1 8·6 17·6 10·7 4·8 170·0 3·8 Ni 0·95 1·06 0·81 1·29 2·43 1·75 1·19 3·99 3·93 2·44 4·77 1·74 54·0 0·9 Rb 38·3 39·9 43·8 43·9 25·5 37·0 37·2 18·9 22·5 27·6 29·0 34·2 7·6 0·7 Zn 80·3 76·1 83·9 74·9 113 107 81·2 102 98·5 93·2 84·8 87·5 83·0 1·4 Sr 238 215 210 201 308 246 240 287 283 271 258 242 256 3·1 Y 45·5 47·2 50·1 52·6 36·0 45·3 45·7 31·8 30·9 36·5 39·8 42·3 18·4 1·3 Zr 134 143 146 151 92 127 130 71 81 98 111 119 45 1·3 Nb 3·5 4·0 3·4 4·2 2·8 3·5 3·9 2·8 2·5 3·0 3·1 3·3 3·6 0·4 Ba 512 546 548 567 347 481 497 266 306 360 410 447 100 6·9 Rare earth elements (ppm, ICP-MS) N=4 La 12·52 13·66 13·09 14·00 9·62 10·41 10·47 12·55 9·83 7·52 8·06 11·25 11·80 3·84 0·12 Ce 29·53 32·84 31·32 32·70 22·86 25·21 24·39 29·12 24·28 18·00 20·00 26·13 28·06 9·58 0·30 Pr 4·14 4·51 4·33 4·60 3·37 3·61 3·57 4·21 3·32 2·61 2·86 3·76 3·97 1·45 0·07 Nd 18·57 20·63 19·87 20·75 15·90 17·08 16·66 19·53 14·94 12·62 13·32 17·56 18·02 7·08 0·30 Sm 5·16 5·60 5·27 5·70 4·54 4·81 4·642 5·44 4·11 3·65 3·84 4·90 4·96 2·13 0·14 Eu 1·42 1·57 1·47 1·47 1·45 1·48 1·47 1·56 1·22 1·19 1·19 1·42 1·46 0·81 0·05 Gd 5·78 6·38 6·22 6·50 5·31 5·63 5·28 6·26 4·46 4·36 4·36 5·56 5·80 2·66 0·13 Tb 1·02 1·08 1·08 1·15 0·92 0·98 0·93 1·10 0·79 0·76 0·78 0·96 1·02 0·47 0·02 Dy 6·47 7·15 7·04 7·36 5·80 6·15 5·84 7·00 4·98 4·79 4·99 6·10 6·47 2·96 0·13 Ho 1·47 1·58 1·58 1·65 1·28 1·37 1·30 1·59 1·11 1·08 1·10 1·35 1·45 0·64 0·02 Er 4·35 4·82 4·74 4·99 3·80 4·07 3·82 4·73 3·37 3·16 3·32 4·00 4·37 1·86 0·07 Tm 0·67 0·73 0·75 0·80 0·57 0·61 0·59 0·73 0·52 0·48 0·51 0·61 0·66 0·28 0·01 Yb 4·59 4·86 4·87 5·30 3·60 3·97 3·79 4·67 3·47 3·08 3·31 3·92 4·42 1·75 0·06 Lu 0·72 0·77 0·78 0·81 0·57 0·62 0·59 0·73 0·54 0·47 0·51 0·62 0·69 0·28 0·01 Isotope ratios (TIMS) N=3 87Sr/86Sr 0·70332 0·70328 0·70327 0·70327 0·70332 0·70332 0·70335 0·70332 0·70336 0·70331 0·70338 0·70330 0·70331 0·70424 2·15E-05 2SE 1·13E-05 7·29E-06 8·34E-06 8·02E-06 1·25E-05 6·34E-06 1·58E-05 8·39E-06 1·53E-05 7·78E-06 1·09E-05 7·91E-06 7·54E-06 143Nd/144Nd 0·51301 0·51301 0·51302 0·51301 0·51301 0·51298 0·51300 0·51302 0·51302 0·51301 0·51301 0·51301 0·51302 0·51282 2·08E-05 2SE 7·14E-06 9·39E-06 7·83E-06 7·18E-06 6·00E-06 9·05E-06 1·08E-05 9·07E-06 9·29E-06 8·14E-06 1·09E-05 1·02E-05 7·72E-06 Sample No. 271* 17* 168 150 99 229 355 85 140 297 90 114 112 118 121 110 HB-2 Occurrence pm pm pm pm pm pm sc HP sc HP sc HP sc HP sc HP sc MP sc MP sc MP sc MP sc MP Eruptive unit Pre-Kp IV 2 3 3 4 3 3 3 3 3 3 4 4 4 4 4 House standard 2σ Major elements (wt %, XRF) N=14 SiO2 73·4 73·8 73·5 74·7 75·3 75·4 63·2 64·0 65·8 69·2 71·8 58·2 60·1 64·3 67·4 69·2 53·6 0·32 TiO2 0·49 0·52 0·53 0·49 0·46 0·47 0·93 0·87 0·81 0·68 0·61 1·02 0·95 0·83 0·70 0·67 0·76 0·01 Al2O3 14·9 14·1 14·2 13·9 13·5 13·5 15·3 14·8 15·0 14·3 14·2 15·6 15·2 14·9 15·3 14·2 17·4 0·12 FeO* 2·45 2·56 2·60 2·30 2·30 2·14 6·23 5·98 5·17 4·19 3·30 9·21 8·17 6·47 4·83 4·53 9·29 0·19 MnO 0·12 0·14 0·15 0·13 0·14 0·13 0·20 0·20 0·20 0·17 0·15 0·21 0·20 0·18 0·16 0·16 0·18 0·003 MgO 0·62 0·74 0·80 0·59 0·65 0·58 2·47 2·17 1·75 1·30 1·02 3·61 3·20 2·45 1·73 1·62 6·44 0·06 CaO 2·64 2·58 2·80 2·42 2·30 2·17 5·89 5·46 4·96 3·66 3·20 7·39 6·88 5·64 4·83 4·10 9·56 0·17 Na2O 4·07 4·30 4·51 4·51 4·51 4·51 3·80 3·75 4·04 4·12 4·33 3·05 3·30 3·60 3·87 4·13 2·40 0·03 K2O 1·67 1·72 1·64 1·71 1·83 1·76 1·03 1·06 1·16 1·42 1·59 0·77 0·93 1·15 1·28 1·42 0·24 0·005 P2O5 0·10 0·12 0·13 0·10 0·11 0·10 0·62 0·54 0·46 0·31 0·21 0·42 0·42 0·34 0·25 0·25 0·11 0·002 Total 100·37 100·59 100·76 100·87 101·07 100·72 99·63 98·78 99·37 99·32 100·41 99·44 99·36 99·87 100·40 100·23 Trace elements (ppm, XRF) N=14 Sc 17·9 14·6 17·4 16·6 30·0 20·5 21·2 38·8 33·7 31·7 25·0 25·9 33·3 2·3 V 24·0 23·8 21·4 19·7 114 71·5 42·3 253 214 158 95·8 94·9 217 4·6 Cr 3·17 2·66 3·40 2·35 5·12 2·65 3·57 19·1 8·6 17·6 10·7 4·8 170·0 3·8 Ni 0·95 1·06 0·81 1·29 2·43 1·75 1·19 3·99 3·93 2·44 4·77 1·74 54·0 0·9 Rb 38·3 39·9 43·8 43·9 25·5 37·0 37·2 18·9 22·5 27·6 29·0 34·2 7·6 0·7 Zn 80·3 76·1 83·9 74·9 113 107 81·2 102 98·5 93·2 84·8 87·5 83·0 1·4 Sr 238 215 210 201 308 246 240 287 283 271 258 242 256 3·1 Y 45·5 47·2 50·1 52·6 36·0 45·3 45·7 31·8 30·9 36·5 39·8 42·3 18·4 1·3 Zr 134 143 146 151 92 127 130 71 81 98 111 119 45 1·3 Nb 3·5 4·0 3·4 4·2 2·8 3·5 3·9 2·8 2·5 3·0 3·1 3·3 3·6 0·4 Ba 512 546 548 567 347 481 497 266 306 360 410 447 100 6·9 Rare earth elements (ppm, ICP-MS) N=4 La 12·52 13·66 13·09 14·00 9·62 10·41 10·47 12·55 9·83 7·52 8·06 11·25 11·80 3·84 0·12 Ce 29·53 32·84 31·32 32·70 22·86 25·21 24·39 29·12 24·28 18·00 20·00 26·13 28·06 9·58 0·30 Pr 4·14 4·51 4·33 4·60 3·37 3·61 3·57 4·21 3·32 2·61 2·86 3·76 3·97 1·45 0·07 Nd 18·57 20·63 19·87 20·75 15·90 17·08 16·66 19·53 14·94 12·62 13·32 17·56 18·02 7·08 0·30 Sm 5·16 5·60 5·27 5·70 4·54 4·81 4·642 5·44 4·11 3·65 3·84 4·90 4·96 2·13 0·14 Eu 1·42 1·57 1·47 1·47 1·45 1·48 1·47 1·56 1·22 1·19 1·19 1·42 1·46 0·81 0·05 Gd 5·78 6·38 6·22 6·50 5·31 5·63 5·28 6·26 4·46 4·36 4·36 5·56 5·80 2·66 0·13 Tb 1·02 1·08 1·08 1·15 0·92 0·98 0·93 1·10 0·79 0·76 0·78 0·96 1·02 0·47 0·02 Dy 6·47 7·15 7·04 7·36 5·80 6·15 5·84 7·00 4·98 4·79 4·99 6·10 6·47 2·96 0·13 Ho 1·47 1·58 1·58 1·65 1·28 1·37 1·30 1·59 1·11 1·08 1·10 1·35 1·45 0·64 0·02 Er 4·35 4·82 4·74 4·99 3·80 4·07 3·82 4·73 3·37 3·16 3·32 4·00 4·37 1·86 0·07 Tm 0·67 0·73 0·75 0·80 0·57 0·61 0·59 0·73 0·52 0·48 0·51 0·61 0·66 0·28 0·01 Yb 4·59 4·86 4·87 5·30 3·60 3·97 3·79 4·67 3·47 3·08 3·31 3·92 4·42 1·75 0·06 Lu 0·72 0·77 0·78 0·81 0·57 0·62 0·59 0·73 0·54 0·47 0·51 0·62 0·69 0·28 0·01 Isotope ratios (TIMS) N=3 87Sr/86Sr 0·70332 0·70328 0·70327 0·70327 0·70332 0·70332 0·70335 0·70332 0·70336 0·70331 0·70338 0·70330 0·70331 0·70424 2·15E-05 2SE 1·13E-05 7·29E-06 8·34E-06 8·02E-06 1·25E-05 6·34E-06 1·58E-05 8·39E-06 1·53E-05 7·78E-06 1·09E-05 7·91E-06 7·54E-06 143Nd/144Nd 0·51301 0·51301 0·51302 0·51301 0·51301 0·51298 0·51300 0·51302 0·51302 0·51301 0·51301 0·51301 0·51302 0·51282 2·08E-05 2SE 7·14E-06 9·39E-06 7·83E-06 7·18E-06 6·00E-06 9·05E-06 1·08E-05 9·07E-06 9·29E-06 8·14E-06 1·09E-05 1·02E-05 7·72E-06 * pm, pumice; sc: scoria. These samples are only for reference because of higher LOI (>3 wt %). Table 2: Representative geochemical compositions of Kp IV and Pre-Kp IV juvenile materials Sample No. 271* 17* 168 150 99 229 355 85 140 297 90 114 112 118 121 110 HB-2 Occurrence pm pm pm pm pm pm sc HP sc HP sc HP sc HP sc HP sc MP sc MP sc MP sc MP sc MP Eruptive unit Pre-Kp IV 2 3 3 4 3 3 3 3 3 3 4 4 4 4 4 House standard 2σ Major elements (wt %, XRF) N=14 SiO2 73·4 73·8 73·5 74·7 75·3 75·4 63·2 64·0 65·8 69·2 71·8 58·2 60·1 64·3 67·4 69·2 53·6 0·32 TiO2 0·49 0·52 0·53 0·49 0·46 0·47 0·93 0·87 0·81 0·68 0·61 1·02 0·95 0·83 0·70 0·67 0·76 0·01 Al2O3 14·9 14·1 14·2 13·9 13·5 13·5 15·3 14·8 15·0 14·3 14·2 15·6 15·2 14·9 15·3 14·2 17·4 0·12 FeO* 2·45 2·56 2·60 2·30 2·30 2·14 6·23 5·98 5·17 4·19 3·30 9·21 8·17 6·47 4·83 4·53 9·29 0·19 MnO 0·12 0·14 0·15 0·13 0·14 0·13 0·20 0·20 0·20 0·17 0·15 0·21 0·20 0·18 0·16 0·16 0·18 0·003 MgO 0·62 0·74 0·80 0·59 0·65 0·58 2·47 2·17 1·75 1·30 1·02 3·61 3·20 2·45 1·73 1·62 6·44 0·06 CaO 2·64 2·58 2·80 2·42 2·30 2·17 5·89 5·46 4·96 3·66 3·20 7·39 6·88 5·64 4·83 4·10 9·56 0·17 Na2O 4·07 4·30 4·51 4·51 4·51 4·51 3·80 3·75 4·04 4·12 4·33 3·05 3·30 3·60 3·87 4·13 2·40 0·03 K2O 1·67 1·72 1·64 1·71 1·83 1·76 1·03 1·06 1·16 1·42 1·59 0·77 0·93 1·15 1·28 1·42 0·24 0·005 P2O5 0·10 0·12 0·13 0·10 0·11 0·10 0·62 0·54 0·46 0·31 0·21 0·42 0·42 0·34 0·25 0·25 0·11 0·002 Total 100·37 100·59 100·76 100·87 101·07 100·72 99·63 98·78 99·37 99·32 100·41 99·44 99·36 99·87 100·40 100·23 Trace elements (ppm, XRF) N=14 Sc 17·9 14·6 17·4 16·6 30·0 20·5 21·2 38·8 33·7 31·7 25·0 25·9 33·3 2·3 V 24·0 23·8 21·4 19·7 114 71·5 42·3 253 214 158 95·8 94·9 217 4·6 Cr 3·17 2·66 3·40 2·35 5·12 2·65 3·57 19·1 8·6 17·6 10·7 4·8 170·0 3·8 Ni 0·95 1·06 0·81 1·29 2·43 1·75 1·19 3·99 3·93 2·44 4·77 1·74 54·0 0·9 Rb 38·3 39·9 43·8 43·9 25·5 37·0 37·2 18·9 22·5 27·6 29·0 34·2 7·6 0·7 Zn 80·3 76·1 83·9 74·9 113 107 81·2 102 98·5 93·2 84·8 87·5 83·0 1·4 Sr 238 215 210 201 308 246 240 287 283 271 258 242 256 3·1 Y 45·5 47·2 50·1 52·6 36·0 45·3 45·7 31·8 30·9 36·5 39·8 42·3 18·4 1·3 Zr 134 143 146 151 92 127 130 71 81 98 111 119 45 1·3 Nb 3·5 4·0 3·4 4·2 2·8 3·5 3·9 2·8 2·5 3·0 3·1 3·3 3·6 0·4 Ba 512 546 548 567 347 481 497 266 306 360 410 447 100 6·9 Rare earth elements (ppm, ICP-MS) N=4 La 12·52 13·66 13·09 14·00 9·62 10·41 10·47 12·55 9·83 7·52 8·06 11·25 11·80 3·84 0·12 Ce 29·53 32·84 31·32 32·70 22·86 25·21 24·39 29·12 24·28 18·00 20·00 26·13 28·06 9·58 0·30 Pr 4·14 4·51 4·33 4·60 3·37 3·61 3·57 4·21 3·32 2·61 2·86 3·76 3·97 1·45 0·07 Nd 18·57 20·63 19·87 20·75 15·90 17·08 16·66 19·53 14·94 12·62 13·32 17·56 18·02 7·08 0·30 Sm 5·16 5·60 5·27 5·70 4·54 4·81 4·642 5·44 4·11 3·65 3·84 4·90 4·96 2·13 0·14 Eu 1·42 1·57 1·47 1·47 1·45 1·48 1·47 1·56 1·22 1·19 1·19 1·42 1·46 0·81 0·05 Gd 5·78 6·38 6·22 6·50 5·31 5·63 5·28 6·26 4·46 4·36 4·36 5·56 5·80 2·66 0·13 Tb 1·02 1·08 1·08 1·15 0·92 0·98 0·93 1·10 0·79 0·76 0·78 0·96 1·02 0·47 0·02 Dy 6·47 7·15 7·04 7·36 5·80 6·15 5·84 7·00 4·98 4·79 4·99 6·10 6·47 2·96 0·13 Ho 1·47 1·58 1·58 1·65 1·28 1·37 1·30 1·59 1·11 1·08 1·10 1·35 1·45 0·64 0·02 Er 4·35 4·82 4·74 4·99 3·80 4·07 3·82 4·73 3·37 3·16 3·32 4·00 4·37 1·86 0·07 Tm 0·67 0·73 0·75 0·80 0·57 0·61 0·59 0·73 0·52 0·48 0·51 0·61 0·66 0·28 0·01 Yb 4·59 4·86 4·87 5·30 3·60 3·97 3·79 4·67 3·47 3·08 3·31 3·92 4·42 1·75 0·06 Lu 0·72 0·77 0·78 0·81 0·57 0·62 0·59 0·73 0·54 0·47 0·51 0·62 0·69 0·28 0·01 Isotope ratios (TIMS) N=3 87Sr/86Sr 0·70332 0·70328 0·70327 0·70327 0·70332 0·70332 0·70335 0·70332 0·70336 0·70331 0·70338 0·70330 0·70331 0·70424 2·15E-05 2SE 1·13E-05 7·29E-06 8·34E-06 8·02E-06 1·25E-05 6·34E-06 1·58E-05 8·39E-06 1·53E-05 7·78E-06 1·09E-05 7·91E-06 7·54E-06 143Nd/144Nd 0·51301 0·51301 0·51302 0·51301 0·51301 0·51298 0·51300 0·51302 0·51302 0·51301 0·51301 0·51301 0·51302 0·51282 2·08E-05 2SE 7·14E-06 9·39E-06 7·83E-06 7·18E-06 6·00E-06 9·05E-06 1·08E-05 9·07E-06 9·29E-06 8·14E-06 1·09E-05 1·02E-05 7·72E-06 Sample No. 271* 17* 168 150 99 229 355 85 140 297 90 114 112 118 121 110 HB-2 Occurrence pm pm pm pm pm pm sc HP sc HP sc HP sc HP sc HP sc MP sc MP sc MP sc MP sc MP Eruptive unit Pre-Kp IV 2 3 3 4 3 3 3 3 3 3 4 4 4 4 4 House standard 2σ Major elements (wt %, XRF) N=14 SiO2 73·4 73·8 73·5 74·7 75·3 75·4 63·2 64·0 65·8 69·2 71·8 58·2 60·1 64·3 67·4 69·2 53·6 0·32 TiO2 0·49 0·52 0·53 0·49 0·46 0·47 0·93 0·87 0·81 0·68 0·61 1·02 0·95 0·83 0·70 0·67 0·76 0·01 Al2O3 14·9 14·1 14·2 13·9 13·5 13·5 15·3 14·8 15·0 14·3 14·2 15·6 15·2 14·9 15·3 14·2 17·4 0·12 FeO* 2·45 2·56 2·60 2·30 2·30 2·14 6·23 5·98 5·17 4·19 3·30 9·21 8·17 6·47 4·83 4·53 9·29 0·19 MnO 0·12 0·14 0·15 0·13 0·14 0·13 0·20 0·20 0·20 0·17 0·15 0·21 0·20 0·18 0·16 0·16 0·18 0·003 MgO 0·62 0·74 0·80 0·59 0·65 0·58 2·47 2·17 1·75 1·30 1·02 3·61 3·20 2·45 1·73 1·62 6·44 0·06 CaO 2·64 2·58 2·80 2·42 2·30 2·17 5·89 5·46 4·96 3·66 3·20 7·39 6·88 5·64 4·83 4·10 9·56 0·17 Na2O 4·07 4·30 4·51 4·51 4·51 4·51 3·80 3·75 4·04 4·12 4·33 3·05 3·30 3·60 3·87 4·13 2·40 0·03 K2O 1·67 1·72 1·64 1·71 1·83 1·76 1·03 1·06 1·16 1·42 1·59 0·77 0·93 1·15 1·28 1·42 0·24 0·005 P2O5 0·10 0·12 0·13 0·10 0·11 0·10 0·62 0·54 0·46 0·31 0·21 0·42 0·42 0·34 0·25 0·25 0·11 0·002 Total 100·37 100·59 100·76 100·87 101·07 100·72 99·63 98·78 99·37 99·32 100·41 99·44 99·36 99·87 100·40 100·23 Trace elements (ppm, XRF) N=14 Sc 17·9 14·6 17·4 16·6 30·0 20·5 21·2 38·8 33·7 31·7 25·0 25·9 33·3 2·3 V 24·0 23·8 21·4 19·7 114 71·5 42·3 253 214 158 95·8 94·9 217 4·6 Cr 3·17 2·66 3·40 2·35 5·12 2·65 3·57 19·1 8·6 17·6 10·7 4·8 170·0 3·8 Ni 0·95 1·06 0·81 1·29 2·43 1·75 1·19 3·99 3·93 2·44 4·77 1·74 54·0 0·9 Rb 38·3 39·9 43·8 43·9 25·5 37·0 37·2 18·9 22·5 27·6 29·0 34·2 7·6 0·7 Zn 80·3 76·1 83·9 74·9 113 107 81·2 102 98·5 93·2 84·8 87·5 83·0 1·4 Sr 238 215 210 201 308 246 240 287 283 271 258 242 256 3·1 Y 45·5 47·2 50·1 52·6 36·0 45·3 45·7 31·8 30·9 36·5 39·8 42·3 18·4 1·3 Zr 134 143 146 151 92 127 130 71 81 98 111 119 45 1·3 Nb 3·5 4·0 3·4 4·2 2·8 3·5 3·9 2·8 2·5 3·0 3·1 3·3 3·6 0·4 Ba 512 546 548 567 347 481 497 266 306 360 410 447 100 6·9 Rare earth elements (ppm, ICP-MS) N=4 La 12·52 13·66 13·09 14·00 9·62 10·41 10·47 12·55 9·83 7·52 8·06 11·25 11·80 3·84 0·12 Ce 29·53 32·84 31·32 32·70 22·86 25·21 24·39 29·12 24·28 18·00 20·00 26·13 28·06 9·58 0·30 Pr 4·14 4·51 4·33 4·60 3·37 3·61 3·57 4·21 3·32 2·61 2·86 3·76 3·97 1·45 0·07 Nd 18·57 20·63 19·87 20·75 15·90 17·08 16·66 19·53 14·94 12·62 13·32 17·56 18·02 7·08 0·30 Sm 5·16 5·60 5·27 5·70 4·54 4·81 4·642 5·44 4·11 3·65 3·84 4·90 4·96 2·13 0·14 Eu 1·42 1·57 1·47 1·47 1·45 1·48 1·47 1·56 1·22 1·19 1·19 1·42 1·46 0·81 0·05 Gd 5·78 6·38 6·22 6·50 5·31 5·63 5·28 6·26 4·46 4·36 4·36 5·56 5·80 2·66 0·13 Tb 1·02 1·08 1·08 1·15 0·92 0·98 0·93 1·10 0·79 0·76 0·78 0·96 1·02 0·47 0·02 Dy 6·47 7·15 7·04 7·36 5·80 6·15 5·84 7·00 4·98 4·79 4·99 6·10 6·47 2·96 0·13 Ho 1·47 1·58 1·58 1·65 1·28 1·37 1·30 1·59 1·11 1·08 1·10 1·35 1·45 0·64 0·02 Er 4·35 4·82 4·74 4·99 3·80 4·07 3·82 4·73 3·37 3·16 3·32 4·00 4·37 1·86 0·07 Tm 0·67 0·73 0·75 0·80 0·57 0·61 0·59 0·73 0·52 0·48 0·51 0·61 0·66 0·28 0·01 Yb 4·59 4·86 4·87 5·30 3·60 3·97 3·79 4·67 3·47 3·08 3·31 3·92 4·42 1·75 0·06 Lu 0·72 0·77 0·78 0·81 0·57 0·62 0·59 0·73 0·54 0·47 0·51 0·62 0·69 0·28 0·01 Isotope ratios (TIMS) N=3 87Sr/86Sr 0·70332 0·70328 0·70327 0·70327 0·70332 0·70332 0·70335 0·70332 0·70336 0·70331 0·70338 0·70330 0·70331 0·70424 2·15E-05 2SE 1·13E-05 7·29E-06 8·34E-06 8·02E-06 1·25E-05 6·34E-06 1·58E-05 8·39E-06 1·53E-05 7·78E-06 1·09E-05 7·91E-06 7·54E-06 143Nd/144Nd 0·51301 0·51301 0·51302 0·51301 0·51301 0·51298 0·51300 0·51302 0·51302 0·51301 0·51301 0·51301 0·51302 0·51282 2·08E-05 2SE 7·14E-06 9·39E-06 7·83E-06 7·18E-06 6·00E-06 9·05E-06 1·08E-05 9·07E-06 9·29E-06 8·14E-06 1·09E-05 1·02E-05 7·72E-06 * pm, pumice; sc: scoria. These samples are only for reference because of higher LOI (>3 wt %). Fig. 8. View largeDownload slide Chondrite-normalized REE concentrations in Kp IV juvenile materials. Chondrite data are from Sun & McDonough (1989). Pumice samples display greater light-REE enrichment and stronger negative Eu anomalies than scoria samples. REE contents are higher in HP scoria than in MP scoria. Fig. 8. View largeDownload slide Chondrite-normalized REE concentrations in Kp IV juvenile materials. Chondrite data are from Sun & McDonough (1989). Pumice samples display greater light-REE enrichment and stronger negative Eu anomalies than scoria samples. REE contents are higher in HP scoria than in MP scoria. Juvenile materials from the Kp IV eruption show relatively wide 87Sr/86Sr variations, whereas their 143Nd/144Nd values are nearly the same (Fig. 9, Table 2). The average 87Sr/86Sr value of scoria is slightly higher than that of pumice. The 87Sr/86Sr values of Kp IV pumice range from 0·70326 to 0·70332, and decrease systematically with increasing SiO2. Among the scorias, MP scoria exhibits a wider variation in 87Sr/86Sr (0·70330–0·70338) than HP scoria (0·70332–0·70335) and can be divided into two types with higher (>0·70336) and lower Sr isotope ratios (<0·70331). Fig. 9. View largeDownload slide Isotopic compositions of Kp IV juvenile materials. (a) 87Sr/86Sr vs SiO2. (b) 143Nd/144Nd vs 87Sr/86Sr. Each diagram includes a 2σ error symbol. The Sr isotopic ratios of pumices are lower than those of scorias, and less silicic pumices have higher 87Sr/86Sr than more silicic pumices. Abbreviations: wp, white pumice; sc, scoria. The data from Unit 2 and Pre-Kp IV pumice are shown for reference but were not used in our analysis because of their high loss on ignition (>3 wt %). Fig. 9. View largeDownload slide Isotopic compositions of Kp IV juvenile materials. (a) 87Sr/86Sr vs SiO2. (b) 143Nd/144Nd vs 87Sr/86Sr. Each diagram includes a 2σ error symbol. The Sr isotopic ratios of pumices are lower than those of scorias, and less silicic pumices have higher 87Sr/86Sr than more silicic pumices. Abbreviations: wp, white pumice; sc, scoria. The data from Unit 2 and Pre-Kp IV pumice are shown for reference but were not used in our analysis because of their high loss on ignition (>3 wt %). RELATIONSHIP BETWEEN PRE-KP IV AND KP IV MATERIALS Pumices from the Pre-Kp IV and Kp IV eruptions are similar in their matrix glass chemistry (Fig. 3), as well as their mineral chemistry and modal phenocryst contents (Fig. 5a). However, the cores of orthopyroxene and ilmenite phenocrysts in Pre-Kp IV pumice can be distinguished from those in Kp IV pumice, which indicates that the rhyolitic magma of Pre-Kp IV is distinct from that of Kp IV. Considering the short time gap between the Pre-Kp IV and Kp IV eruptions (Hasegawa et al., 2016), it appears that the silicic magma system was modified immediately before each eruption. In the following discussion, we focus on the petrological and geochemical features of the Kp IV eruption. DISCUSSION Magma mixing and end-member magmas of the Kp IV eruption Juvenile materials of the Kp IV eruption show heterogeneous textures that indicate mingling of silicic and basic magmas, such as banded pumice and scoria, and silicic inclusions in scoria. Evidence of more intimate magma mixing is present in the form of disequilibrium phenocryst compositions, even in samples of homogeneous pumice (Figs 5 and 6). In this section, we estimate the end-member magmas of the Kp IV eruption on the basis of the juvenile products and their mineral, whole-rock, and matrix glass chemistry. Mafic end-member magmas Hasegawa et al. (2016) suggested that the juvenile materials of the Kp IV eruption consist of pumice and three types of scoria, and that the scorias mingled with pumice to form heterogeneous clasts. Although the whole-rock chemistry of LP scoria could not be determined, the HP and MP scorias form two distinct linear trends in SiO2 variation diagrams (Fig. 7). This suggests that these two types of scoria represent the products of mixing or mingling of relatively mafic and silicic magmas. On the other hand, in a P2O5 vs SiO2 diagram (Fig. 3) the matrix glasses of the three scorias are clearly compositionally distinct, and a >10% SiO2 compositional gap separates the MP scoria and the pumice. In addition, although the glass of the HP scoria forms simple linear trends in all SiO2 variation diagrams, the other two types of scoria do not, again most clearly in the P2O5 vs SiO2 diagram. The evidence suggests that the three types of scoria are distinct products of mingling with the pumice, accounting for the differences in their matrix glasses. The scorias are less porphyritic than the pumice, the LP scoria having no phenocrysts and the HP and MP scorias being sparsely phyric. Although phenocryst minerals with compositional variations similar to those in the pumice are dominant in the scoria, minor amounts of calcic plagioclase (An = 80–95) and magnesian magnetite (Mg/Mn = 7·5–25) are also present in the scoria (Fig. 5, summarized in Fig. 10). Because the latter two types of phenocryst must have crystallized from a magma less silicic than that yielding the dominant phenocrysts, we describe them here as Type-M (mafic) phenocrysts. We further divide the Type-M phenocrysts into Type-M1, with lower Mg/Mn (7·5–17) magnetite, found in the HP scoria, and Type-M2, with higher Mg/Mn (11·5–25) magnetite, found in the MP scoria (Fig. 10). Phenocryst contents in the HP and MP scorias decrease with decreasing whole-rock SiO2 content (Fig. 4), suggesting that the end-member magmas were nearly aphyric andesites, containing only a few modal percent of Type-M phenocrysts. In the following discussion, we refer to these end-member magmas as LP-andesitic, MP-andesitic (containing Type-M2 phenocrysts), and HP-andesitic (containing Type-M1 phenocrysts). Fig. 10. View largeDownload slide Summary of types of phenocryst in Kp IV juvenile materials. Type-R (rhyolitic) and Type-D (dacitic) phenocrysts are dominant in all samples. Type-M1 and Type-M2 (mafic) phenocrysts are found only in HP and MP scoria, respectively. Phenocrysts shown in light gray with question marks (high-An plagioclase in pumice and high-Mg/Mn ilmenite in HP scoria) could not be clearly characterized and were not considered further. Fig. 10. View largeDownload slide Summary of types of phenocryst in Kp IV juvenile materials. Type-R (rhyolitic) and Type-D (dacitic) phenocrysts are dominant in all samples. Type-M1 and Type-M2 (mafic) phenocrysts are found only in HP and MP scoria, respectively. Phenocrysts shown in light gray with question marks (high-An plagioclase in pumice and high-Mg/Mn ilmenite in HP scoria) could not be clearly characterized and were not considered further. Silicic end-member magmas Pumice whole-rock compositions form well-defined linear trends in SiO2 variation diagrams (Fig. 7b). The REE patterns of pumice samples are also nearly parallel to one another (Fig. 8). These features imply that the pumice compositions were not significantly affected by fractionation of minor phenocryst phases (e.g. apatite and zircon) and that pumice is a product of mixing between two end-member magmas. Because the pumice trend is so clearly different from those of the scorias (Fig. 7b), the andesitic end-member magmas that formed the scorias must not have been involved in the formation of the pumice. Instead, two silicic end-member magmas are required. This interpretation is consistent with the compositional distribution of phenocryst minerals. In each pumice sample, the cores of plagioclase, orthopyroxene, and clinopyroxene phenocrysts have a wide compositional range, though not an obviously bimodal distribution, except in the case of Fe–Ti oxide phenocrysts in sample No. 168 (Fig. 5a). This evidence suggests that the pumice is not a product of a simple crystallization process but, rather, can be explained by mixing of end-member magmas containing phenocrysts with similar compositions. Compositional variations of phenocrysts that show reverse zoning also support this hypothesis (Fig. 6). Mafic magma-derived phenocrysts, such as olivine and pyroxenes with Mg# > 80 and calcic plagioclase (An > 80), are absent in the pumice. Therefore, we modeled the two silicic end-member magmas as a rhyolite and a dacite, each with its respective set of phenocrysts (Type R or Type D) as shown in Fig. 10. Individual pumice samples contain both Type-R and Type-D phenocrysts (Fig. 10, Table 3). Type-R phenocrysts are the predominant component and consist of sodic plagioclase (An = 35 to c. 48), ferrous orthopyroxene (Mg# = 67 to c. 70), and Mg-poor magnetite (Mg/Mn = 2·4–2·9) and ilmenite (Mg/Mn = 3·5–4·2), which are in equilibrium with each other (Bacon & Hirschmann, 1988). A small amount of ferrous clinopyroxene (Mg# = 69 to c. 73), which could be in equilibrium with the ferrous orthopyroxene phenocrysts (Baker & Eggler, 1987; Brophy & Dreher, 2000), might be also included. Type-D phenocrysts consist of slightly calcic plagioclase (An = c. 48–70), magnesian orthopyroxene (Mg# = c. 70–76), and magnesian clinopyroxene (Mg# = c. 73–78). The pyroxene phenocrysts are in equilibrium (Baker & Eggler, 1987; Brophy & Dreher, 2000), and some of them form crystal clots. The slightly Mg-rich magnetite (Mg/Mn = 2·8–7·5) and ilmenite (Mg/Mn = 4·2–13) in sample No. 168 might be considered Type-D phenocrysts, but other pumice samples do not contain magnetite and ilmenite of that composition. Table 3: Estimated whole-rock compositions, magmatic temperatures, and phenocryst types of end-member magmas in Kp IV eruptive activity Magma type R1 D1 HP-A MP-A LP-A2 wt % SiO2 75·5 72·0 61·4 57·7 59·4 TiO2 0·45 0·54 1·00 1·07 0·924 Al2O3 12·9 14·7 15·6 15·8 16·54 FeO* 2·06 2·62 7·25 9·65 8·43 MnO 0·12 0·14 0·22 0·21 0·18 MgO 0·54 0·77 2·82 3·71 2·41 CaO 2·04 2·93 6·56 7·73 7·57 Na2O 4·49 4·56 3·56 2·97 3·63 K2O 1·84 1·56 0·90 0·73 0·80 P2O5 0·09 0·12 0·71 0·47 0·19 ppm Sc 16·2 17·8 31·7 38·0 V 17·7 25·0 142 257 Cr 3·2 2·5 5·0 9·5 Ni 0·8 1·3 2·3 3·9 Rb 43·3 37·7 23·5 19·2 Zn 73·5 76·9 112 101 Sr 193 255 311 289 Y 52·3 43·3 34·9 29·6 Zr 148 135 84·7 68·0 Ba 561 518 327 256 Pb 12·0 11·3 13·2 7·5 La 14·2 12·4 9·0 7·2 Ce 33·7 29·7 21·3 17·3 Temperature(oC) 849–8593 856–8723 1044–10624 1063–10814 1033–10484 Phenocryst type R D M1 M2 None Plagioclase (An) 35–48 48–70 80–95 80–95 – Orthopyroxene (Mg#) 67–70 70–76 – – – Clinopyroxene (Mg#) 70–73 73–78 – – – Magnetite (Mg/Mn) 2·4–2·8 2·8–7·5 7·5–17 11·5–25 – Ilmenite (Mg/Mn) 3·5–4·2 4·2–13 – – – Magma type R1 D1 HP-A MP-A LP-A2 wt % SiO2 75·5 72·0 61·4 57·7 59·4 TiO2 0·45 0·54 1·00 1·07 0·924 Al2O3 12·9 14·7 15·6 15·8 16·54 FeO* 2·06 2·62 7·25 9·65 8·43 MnO 0·12 0·14 0·22 0·21 0·18 MgO 0·54 0·77 2·82 3·71 2·41 CaO 2·04 2·93 6·56 7·73 7·57 Na2O 4·49 4·56 3·56 2·97 3·63 K2O 1·84 1·56 0·90 0·73 0·80 P2O5 0·09 0·12 0·71 0·47 0·19 ppm Sc 16·2 17·8 31·7 38·0 V 17·7 25·0 142 257 Cr 3·2 2·5 5·0 9·5 Ni 0·8 1·3 2·3 3·9 Rb 43·3 37·7 23·5 19·2 Zn 73·5 76·9 112 101 Sr 193 255 311 289 Y 52·3 43·3 34·9 29·6 Zr 148 135 84·7 68·0 Ba 561 518 327 256 Pb 12·0 11·3 13·2 7·5 La 14·2 12·4 9·0 7·2 Ce 33·7 29·7 21·3 17·3 Temperature(oC) 849–8593 856–8723 1044–10624 1063–10814 1033–10484 Phenocryst type R D M1 M2 None Plagioclase (An) 35–48 48–70 80–95 80–95 – Orthopyroxene (Mg#) 67–70 70–76 – – – Clinopyroxene (Mg#) 70–73 73–78 – – – Magnetite (Mg/Mn) 2·4–2·8 2·8–7·5 7·5–17 11·5–25 – Ilmenite (Mg/Mn) 3·5–4·2 4·2–13 – – – *Total Fe given as FeO. 1 These compositions are the values at SiO2 = 75·5 and 72·0 wt % on the mixing trend of pumices samples, respectively. The ratios of FeO/total-FeO of melts are calculated on the basis of Sack et al. (1981) and Kilinc et al. (1983). These estimated compositions exhibit compositionally equilibrium with the Type-R and Type-D orthopyroxene phenocrysts approximately. 2 The composition is substituted by the most mafic one of the matrix glass in LP scoria. 3 These values are estimated by Fe–Ti oxides geothermometer in QUILF program (Andersen et al., 1993). 4 These values are the liquidus temperatures of each magma estimated using a rhyolite-MELTS program (Gualda et al., 2012) because they are nearly aphyric. The calculated condition is NNO buffer, 2–3 kbr, and H2O = 4 wt %. This water content is assumed on the basis of the model of Moore et al. (1998). Table 3: Estimated whole-rock compositions, magmatic temperatures, and phenocryst types of end-member magmas in Kp IV eruptive activity Magma type R1 D1 HP-A MP-A LP-A2 wt % SiO2 75·5 72·0 61·4 57·7 59·4 TiO2 0·45 0·54 1·00 1·07 0·924 Al2O3 12·9 14·7 15·6 15·8 16·54 FeO* 2·06 2·62 7·25 9·65 8·43 MnO 0·12 0·14 0·22 0·21 0·18 MgO 0·54 0·77 2·82 3·71 2·41 CaO 2·04 2·93 6·56 7·73 7·57 Na2O 4·49 4·56 3·56 2·97 3·63 K2O 1·84 1·56 0·90 0·73 0·80 P2O5 0·09 0·12 0·71 0·47 0·19 ppm Sc 16·2 17·8 31·7 38·0 V 17·7 25·0 142 257 Cr 3·2 2·5 5·0 9·5 Ni 0·8 1·3 2·3 3·9 Rb 43·3 37·7 23·5 19·2 Zn 73·5 76·9 112 101 Sr 193 255 311 289 Y 52·3 43·3 34·9 29·6 Zr 148 135 84·7 68·0 Ba 561 518 327 256 Pb 12·0 11·3 13·2 7·5 La 14·2 12·4 9·0 7·2 Ce 33·7 29·7 21·3 17·3 Temperature(oC) 849–8593 856–8723 1044–10624 1063–10814 1033–10484 Phenocryst type R D M1 M2 None Plagioclase (An) 35–48 48–70 80–95 80–95 – Orthopyroxene (Mg#) 67–70 70–76 – – – Clinopyroxene (Mg#) 70–73 73–78 – – – Magnetite (Mg/Mn) 2·4–2·8 2·8–7·5 7·5–17 11·5–25 – Ilmenite (Mg/Mn) 3·5–4·2 4·2–13 – – – Magma type R1 D1 HP-A MP-A LP-A2 wt % SiO2 75·5 72·0 61·4 57·7 59·4 TiO2 0·45 0·54 1·00 1·07 0·924 Al2O3 12·9 14·7 15·6 15·8 16·54 FeO* 2·06 2·62 7·25 9·65 8·43 MnO 0·12 0·14 0·22 0·21 0·18 MgO 0·54 0·77 2·82 3·71 2·41 CaO 2·04 2·93 6·56 7·73 7·57 Na2O 4·49 4·56 3·56 2·97 3·63 K2O 1·84 1·56 0·90 0·73 0·80 P2O5 0·09 0·12 0·71 0·47 0·19 ppm Sc 16·2 17·8 31·7 38·0 V 17·7 25·0 142 257 Cr 3·2 2·5 5·0 9·5 Ni 0·8 1·3 2·3 3·9 Rb 43·3 37·7 23·5 19·2 Zn 73·5 76·9 112 101 Sr 193 255 311 289 Y 52·3 43·3 34·9 29·6 Zr 148 135 84·7 68·0 Ba 561 518 327 256 Pb 12·0 11·3 13·2 7·5 La 14·2 12·4 9·0 7·2 Ce 33·7 29·7 21·3 17·3 Temperature(oC) 849–8593 856–8723 1044–10624 1063–10814 1033–10484 Phenocryst type R D M1 M2 None Plagioclase (An) 35–48 48–70 80–95 80–95 – Orthopyroxene (Mg#) 67–70 70–76 – – – Clinopyroxene (Mg#) 70–73 73–78 – – – Magnetite (Mg/Mn) 2·4–2·8 2·8–7·5 7·5–17 11·5–25 – Ilmenite (Mg/Mn) 3·5–4·2 4·2–13 – – – *Total Fe given as FeO. 1 These compositions are the values at SiO2 = 75·5 and 72·0 wt % on the mixing trend of pumices samples, respectively. The ratios of FeO/total-FeO of melts are calculated on the basis of Sack et al. (1981) and Kilinc et al. (1983). These estimated compositions exhibit compositionally equilibrium with the Type-R and Type-D orthopyroxene phenocrysts approximately. 2 The composition is substituted by the most mafic one of the matrix glass in LP scoria. 3 These values are estimated by Fe–Ti oxides geothermometer in QUILF program (Andersen et al., 1993). 4 These values are the liquidus temperatures of each magma estimated using a rhyolite-MELTS program (Gualda et al., 2012) because they are nearly aphyric. The calculated condition is NNO buffer, 2–3 kbr, and H2O = 4 wt %. This water content is assumed on the basis of the model of Moore et al. (1998). Whole-rock composition and temperature of end-member magmas According to the compositional equilibria between melt and pyroxenes, in which the Fe–Mg partition coefficient between mineral and melt is 0·247 for orthopyroxene and 0. ·233 for clinopyroxene (No. SS-40 in Brophy & Dreher, 2000), rhyolitic and dacitic end-member magmas can be assumed with about 75·5 and 72·0 wt % SiO2, respectively, from the linear pumice trends in SiO2 variation diagrams (Table 3, Fig. 11a). Although andesitic magmas should be only slightly porphyritic or aphyric, we can estimate the whole-rock compositions of the andesitic end-member magmas of the HP and MP scorias (HP-andesitic and MP-andesitic magmas, respectively) using the linear relationship between phenocryst contents and whole-rock SiO2 contents in Fig. 4 (Table 3, Fig. 11b). The estimated compositions of these two magmas, SiO2 = 61·4 wt % for HP-andesitic and 57·7 wt % for MP-andesitic, are similar to those of the respective matrix glasses. These SiO2 contents are considered minimum values because these magmas must originally have contained small percentages of phenocrysts, as discussed previously. If we assume the LP-andesitic magma to be aphyric as in the case of the HP and MP scorias, the composition of the matrix glass with the lowest SiO2 content should represent the whole-rock composition of its source andesitic magma (SiO2 = 58·7 wt %; Table 3, Fig. 11b). Fig. 11. View largeDownload slide Estimated whole-rock compositions of Kp IV end-member magmas and their mixing relationships. (a) Al2O3 vs SiO2 emphasizing pumice samples. The single linear trend in pumice, which differs from the trends in scoria samples, suggests mixing of rhyolitic (R) and dacitic (D) end-member magmas. (b) P2O5 vs SiO2. Assuming that the andesitic end-member magmas were aphyric, we can estimate compositions of HP-andesitic (HP-A) and MP-andesitic (MP-A) magmas as similar to the matrix glass compositions of HP and MP scoria, respectively, with the least silica. By extension, LP-andesitic magma is assigned the composition of matrix glass in LP scoria with the least silica. Fig. 11. View largeDownload slide Estimated whole-rock compositions of Kp IV end-member magmas and their mixing relationships. (a) Al2O3 vs SiO2 emphasizing pumice samples. The single linear trend in pumice, which differs from the trends in scoria samples, suggests mixing of rhyolitic (R) and dacitic (D) end-member magmas. (b) P2O5 vs SiO2. Assuming that the andesitic end-member magmas were aphyric, we can estimate compositions of HP-andesitic (HP-A) and MP-andesitic (MP-A) magmas as similar to the matrix glass compositions of HP and MP scoria, respectively, with the least silica. By extension, LP-andesitic magma is assigned the composition of matrix glass in LP scoria with the least silica. We estimated the magmatic temperatures of the silicic end-member magmas from Fe–Ti oxide thermometry using the QUILF program (Andersen et al., 1993; Table 3 and Supplementary Data). Results from average core compositions of Type-R phenocrysts show that the temperature of the rhyolitic end-member magma was 849–859°C. If the magnetite and ilmenite phenocrysts with higher Mg/Mn ratios in sample No. 168 are considered Type-D phenocrysts, the dacitic end-member magma was only slightly hotter than the rhyolitic end-member magma (c. 856–872°C). According to the two-pyroxene geothermometer (Putirka, 2008), magma temperatures indicated by Type-R and Type-D pyroxene cores are 790–823°C and 835–951°C, respectively, making the dacitic end-member magma notably hotter than the rhyolitic magma. By comparison, we can substitute the liquidus temperatures of the HP-, MP-, and LP-andesitic magmas for their magmatic temperatures, because they are nearly aphyric. Using the rhyolite-MELTS program (Gualda et al., 2012) assuming a NNO buffer, 200–300 MPa pressure, and 4 wt % water content, we estimate the liquidus temperatures of the HP-, MP-, and LP-andesitic magmas to be 1044–1062°C, 1063–1081°C, and 1033–1048°C, respectively (Table 3). Petrogenetic relationships among end-member magmas Silicic and andesitic magmas Silicic magma can be produced by partial melting of mafic to intermediate crustal materials (e.g. Smith & Leeman, 1987; Streck & Grunder, 2008), by melt extraction from incompletely crystallized plutons, i.e. crystal mush (e.g. Bachmann & Bergantz, 2004; Hildreth, 2004; Deering et al., 2011b), and by fractional crystallization of basaltic magma (e.g. Miyagi et al., 2012; Parker et al., 2012). We examined the possibility that fractional crystallization of coexisting mafic magmas, as proposed for Kutcharo volcano by Miyagi et al. (2012), can produce rhyolitic as well as dacitic magmas. According to the rhyolite-MELTS model (in which the fractionated assemblage is composed of plagioclase, orthopyroxene, clinopyroxene, Fe–Ti oxides, quartz, hornblende, and apatite), neither the HP- nor the MP-andesitic magma can produce rhyolitic and dacitic magmas, especially their Na2O and P2O5 contents (Fig. 12). We also modeled fractional crystallization based on mass-balance calculations for major elements, assuming that the fractionated phases are the same as the phenocryst minerals in the MP scoria (Table 4). This calculation does not rule out the possibility that fractional crystallization of MP-andesitic magma produced a rhyolitic magma: the sum of squared residuals (R2) is 0·09 and the relative residuals in the rhyolitic end-member magma for each element except for MnO and P2O5 are less than 15% (Table 4). When applying these results to trace elements, however, a Rayleigh fractionation model (CL = Ci * F(D–1), where CL is the composition of the daughter melt, Ci the composition of the parental melt, F the melt fraction, and D the bulk partition coefficient of the fractionating phase, as listed in Table 4) cannot match the trace element contents, especially Y, of the end-member silicic magmas (Fig. 13). In addition, we employed an assimilation and fractional crystallization model (DePaolo, 1981), using the same minerals for the fractionate as those in the simple fractional crystallization model. For the assimilant, we tested two compositions: granite exposed in the Hidaka belt in central Hokkaido (Kamiyama et al., 2007) and a granodiorite xenolith in the Eocene Beppo Formation in eastern Hokkaido (Ogasawara et al., 1998). In both cases, the models give good results for major elements (R2 is 0·03 and the relative residuals in the rhyolitic end-member magma values for each element are less than 26%; Table 4). For trace elements, however, these models cannot duplicate the end-member rhyolitic magma (Fig. 13). Therefore, combined assimilation and fractional crystallization of coexisting andesitic magmas does not account for the Kp IV silicic magma system. This result is also consistent with the higher Sr isotopic ratios in scoria than in pumice (Fig. 9). Table 4: Results of the assimilation and fractional crystallization models for MP-andesitic magma wt % SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O P2O5 ppm Rb Sr Y Zr Ba Parent MP-A 57·70 1·07 15·76 9·65 0·21 3·71 7·73 2·97 0·73 0·47 19·2 289 29·6 68 256 Daughter R 75·50 0·45 12·88 2·06 0·12 0·54 2·04 4·49 1·84 0·09 43·3 193 52·3 148 561 Assimilant Hidaka granite 71·86 0·40 14·61 2·26 0·04 0·59 1·73 4·42 4·04 0·06 139·0 78 43·0 308 488 Beppo granodiorite 62·29 0·57 16·76 5·47 0·10 2·78 5·60 3·48 2·75 0·19 68·1 382 22·5 146 620 Rayleigh fractionation model Fractionated phase: 54·5 wt % 31·5 high-An plg, 22·9 low-An plg, 13·1 opx, 15·6 cpx, 9·6 high-Mg mgt, 5·0 low-Mg mgt, 0·5 ilm, 1·7 apt Result ΣR2 = 0·09 75·56 0·50 12·92 2·11 0·06 0·57 2·08 4·49 1·57 0·15 39·2 202 25·2 140 478 R −0·06 −0·05 −0·04 −0·05 0·06 −0·03 −0·04 0·00 0·27 −0·06 4·09 −8·42 27·16 7·66 83·04 R/Daughter 0·00 −0·11 0·00 −0·02 0·53 −0·05 −0·02 0·00 0·15 −0·68 0·09 −0·04 0·52 0·05 0·15 Assimilation and fractional crystallization model (Hidaka granite) Fractionated phase: 54·2 wt % 31·6 high-An plg, 22·6 low-An plg, 12·5 opx, 15·6 cpx, 13·8 high-Mg mgt, 1·0 low-Mg mgt, 1·0 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·10 75·43 0·40 12·98 2·03 0·09 0·62 2·02 4·49 1·84 0·09 57·2 183 25·0 180 528 R 0·07 0·04 −0·10 0·02 0·03 −0·08 0·02 0·00 0·00 0·00 −13·9 10·0 27·4 −32·2 32·6 R/Daughter 0·00 0·10 −0·01 0·01 0·24 −0·15 0·01 0·00 0·00 0·01 −0·3 0·1 0·5 −0·2 0·1 Assimilation and fractional crystallization model (Beppo granodiorite) Fractionated phase: 57·1 wt % 31·8 high-An plg, 22·7 low-An plg, 12·6 opx, 15·5 cpx, 13·5 high-Mg mgt, 0·9 low-Mg mgt, 0·9 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·14 75·42 0·40 13·00 2·03 0·09 0·63 2·02 4·49 1·84 0·09 55·0 200 23·0 177 607 R 0·08 0·05 −0·12 0·02 0·03 −0·09 0·02 0·00 0·00 0·00 −11·7 −6·9 29·3 −29·6 −46·1 R/Daughter 0·00 0·11 −0·01 0·01 0·26 −0·17 0·01 0·00 0·00 0·02 −0·3 0·0 0·6 −0·2 −0·1 Mineral compositions* Partition coefficient** high-An plg 44·81 0·00 35·06 0·39 0·00 0·09 19·08 0·57 0·01 0·00 0·10 1·50 0·06 0·00 0·20 low-An plg 57·90 0·00 26·28 0·40 0·00 0·03 8·84 6·41 0·14 0·00 0·17 4·04 0·02 0·01 0·47 opx 53·79 0·27 1·74 15·92 0·67 26·27 1·30 0·03 0·00 0·00 0·05 0·01 0·40 0·11 0·02 cpx 53·27 0·30 1·11 8·10 1·02 14·88 20·98 0·34 0·00 0·00 0·07 0·17 2·70 0·17 0·10 high-Mg mgt 0·15 7·70 5·87 81·41 0·33 4·54 0·01 0·00 0·00 0·00 0·01 0·01 0·20 0·10 0·12 low-Mg mgt 0·13 9·35 1·99 85·47 1·22 1·84 0·02 0·00 0·00 0·00 0·01 0·01 0·12 0·24 0·10 ilm 0·03 40·17 0·28 56·03 1·10 2·39 0·01 0·00 0·00 0·00 0·01 0·01 0·27 0·60 0·01 apt 0·00 0·00 0·00 0·00 0·00 0·00 56·80 0·00 0·00 43·20 0·00 2·00 40·0 0·64 0·00 wt % SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O P2O5 ppm Rb Sr Y Zr Ba Parent MP-A 57·70 1·07 15·76 9·65 0·21 3·71 7·73 2·97 0·73 0·47 19·2 289 29·6 68 256 Daughter R 75·50 0·45 12·88 2·06 0·12 0·54 2·04 4·49 1·84 0·09 43·3 193 52·3 148 561 Assimilant Hidaka granite 71·86 0·40 14·61 2·26 0·04 0·59 1·73 4·42 4·04 0·06 139·0 78 43·0 308 488 Beppo granodiorite 62·29 0·57 16·76 5·47 0·10 2·78 5·60 3·48 2·75 0·19 68·1 382 22·5 146 620 Rayleigh fractionation model Fractionated phase: 54·5 wt % 31·5 high-An plg, 22·9 low-An plg, 13·1 opx, 15·6 cpx, 9·6 high-Mg mgt, 5·0 low-Mg mgt, 0·5 ilm, 1·7 apt Result ΣR2 = 0·09 75·56 0·50 12·92 2·11 0·06 0·57 2·08 4·49 1·57 0·15 39·2 202 25·2 140 478 R −0·06 −0·05 −0·04 −0·05 0·06 −0·03 −0·04 0·00 0·27 −0·06 4·09 −8·42 27·16 7·66 83·04 R/Daughter 0·00 −0·11 0·00 −0·02 0·53 −0·05 −0·02 0·00 0·15 −0·68 0·09 −0·04 0·52 0·05 0·15 Assimilation and fractional crystallization model (Hidaka granite) Fractionated phase: 54·2 wt % 31·6 high-An plg, 22·6 low-An plg, 12·5 opx, 15·6 cpx, 13·8 high-Mg mgt, 1·0 low-Mg mgt, 1·0 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·10 75·43 0·40 12·98 2·03 0·09 0·62 2·02 4·49 1·84 0·09 57·2 183 25·0 180 528 R 0·07 0·04 −0·10 0·02 0·03 −0·08 0·02 0·00 0·00 0·00 −13·9 10·0 27·4 −32·2 32·6 R/Daughter 0·00 0·10 −0·01 0·01 0·24 −0·15 0·01 0·00 0·00 0·01 −0·3 0·1 0·5 −0·2 0·1 Assimilation and fractional crystallization model (Beppo granodiorite) Fractionated phase: 57·1 wt % 31·8 high-An plg, 22·7 low-An plg, 12·6 opx, 15·5 cpx, 13·5 high-Mg mgt, 0·9 low-Mg mgt, 0·9 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·14 75·42 0·40 13·00 2·03 0·09 0·63 2·02 4·49 1·84 0·09 55·0 200 23·0 177 607 R 0·08 0·05 −0·12 0·02 0·03 −0·09 0·02 0·00 0·00 0·00 −11·7 −6·9 29·3 −29·6 −46·1 R/Daughter 0·00 0·11 −0·01 0·01 0·26 −0·17 0·01 0·00 0·00 0·02 −0·3 0·0 0·6 −0·2 −0·1 Mineral compositions* Partition coefficient** high-An plg 44·81 0·00 35·06 0·39 0·00 0·09 19·08 0·57 0·01 0·00 0·10 1·50 0·06 0·00 0·20 low-An plg 57·90 0·00 26·28 0·40 0·00 0·03 8·84 6·41 0·14 0·00 0·17 4·04 0·02 0·01 0·47 opx 53·79 0·27 1·74 15·92 0·67 26·27 1·30 0·03 0·00 0·00 0·05 0·01 0·40 0·11 0·02 cpx 53·27 0·30 1·11 8·10 1·02 14·88 20·98 0·34 0·00 0·00 0·07 0·17 2·70 0·17 0·10 high-Mg mgt 0·15 7·70 5·87 81·41 0·33 4·54 0·01 0·00 0·00 0·00 0·01 0·01 0·20 0·10 0·12 low-Mg mgt 0·13 9·35 1·99 85·47 1·22 1·84 0·02 0·00 0·00 0·00 0·01 0·01 0·12 0·24 0·10 ilm 0·03 40·17 0·28 56·03 1·10 2·39 0·01 0·00 0·00 0·00 0·01 0·01 0·27 0·60 0·01 apt 0·00 0·00 0·00 0·00 0·00 0·00 56·80 0·00 0·00 43·20 0·00 2·00 40·0 0·64 0·00 The results of only MP-andesite magma are shown here, because we obtained the similar result using HP-andesite magma. Assimilants are taken from Kamiyama et al. (2007) and Ogasawara et al. (1998). * Mineral compositions are taken from analysed data of the phenocrysts in the most mafic MP-type scoria (No. 114), except for apatite (theoretical value). ** These values are taken from Philpotts & Schnetzler (1970), Arth (1976), Okamoto (1979), Pearce & Norry (1979), Luhr & Carmichael (1980), Gill (1981), Fujimaki (1986), Bacon & Druitt (1988), McKenzie & O’Nions (1991), Sisson (1991), Ewart & Griffin (1994), and Bindeman & Davis (2000). Table 4: Results of the assimilation and fractional crystallization models for MP-andesitic magma wt % SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O P2O5 ppm Rb Sr Y Zr Ba Parent MP-A 57·70 1·07 15·76 9·65 0·21 3·71 7·73 2·97 0·73 0·47 19·2 289 29·6 68 256 Daughter R 75·50 0·45 12·88 2·06 0·12 0·54 2·04 4·49 1·84 0·09 43·3 193 52·3 148 561 Assimilant Hidaka granite 71·86 0·40 14·61 2·26 0·04 0·59 1·73 4·42 4·04 0·06 139·0 78 43·0 308 488 Beppo granodiorite 62·29 0·57 16·76 5·47 0·10 2·78 5·60 3·48 2·75 0·19 68·1 382 22·5 146 620 Rayleigh fractionation model Fractionated phase: 54·5 wt % 31·5 high-An plg, 22·9 low-An plg, 13·1 opx, 15·6 cpx, 9·6 high-Mg mgt, 5·0 low-Mg mgt, 0·5 ilm, 1·7 apt Result ΣR2 = 0·09 75·56 0·50 12·92 2·11 0·06 0·57 2·08 4·49 1·57 0·15 39·2 202 25·2 140 478 R −0·06 −0·05 −0·04 −0·05 0·06 −0·03 −0·04 0·00 0·27 −0·06 4·09 −8·42 27·16 7·66 83·04 R/Daughter 0·00 −0·11 0·00 −0·02 0·53 −0·05 −0·02 0·00 0·15 −0·68 0·09 −0·04 0·52 0·05 0·15 Assimilation and fractional crystallization model (Hidaka granite) Fractionated phase: 54·2 wt % 31·6 high-An plg, 22·6 low-An plg, 12·5 opx, 15·6 cpx, 13·8 high-Mg mgt, 1·0 low-Mg mgt, 1·0 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·10 75·43 0·40 12·98 2·03 0·09 0·62 2·02 4·49 1·84 0·09 57·2 183 25·0 180 528 R 0·07 0·04 −0·10 0·02 0·03 −0·08 0·02 0·00 0·00 0·00 −13·9 10·0 27·4 −32·2 32·6 R/Daughter 0·00 0·10 −0·01 0·01 0·24 −0·15 0·01 0·00 0·00 0·01 −0·3 0·1 0·5 −0·2 0·1 Assimilation and fractional crystallization model (Beppo granodiorite) Fractionated phase: 57·1 wt % 31·8 high-An plg, 22·7 low-An plg, 12·6 opx, 15·5 cpx, 13·5 high-Mg mgt, 0·9 low-Mg mgt, 0·9 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·14 75·42 0·40 13·00 2·03 0·09 0·63 2·02 4·49 1·84 0·09 55·0 200 23·0 177 607 R 0·08 0·05 −0·12 0·02 0·03 −0·09 0·02 0·00 0·00 0·00 −11·7 −6·9 29·3 −29·6 −46·1 R/Daughter 0·00 0·11 −0·01 0·01 0·26 −0·17 0·01 0·00 0·00 0·02 −0·3 0·0 0·6 −0·2 −0·1 Mineral compositions* Partition coefficient** high-An plg 44·81 0·00 35·06 0·39 0·00 0·09 19·08 0·57 0·01 0·00 0·10 1·50 0·06 0·00 0·20 low-An plg 57·90 0·00 26·28 0·40 0·00 0·03 8·84 6·41 0·14 0·00 0·17 4·04 0·02 0·01 0·47 opx 53·79 0·27 1·74 15·92 0·67 26·27 1·30 0·03 0·00 0·00 0·05 0·01 0·40 0·11 0·02 cpx 53·27 0·30 1·11 8·10 1·02 14·88 20·98 0·34 0·00 0·00 0·07 0·17 2·70 0·17 0·10 high-Mg mgt 0·15 7·70 5·87 81·41 0·33 4·54 0·01 0·00 0·00 0·00 0·01 0·01 0·20 0·10 0·12 low-Mg mgt 0·13 9·35 1·99 85·47 1·22 1·84 0·02 0·00 0·00 0·00 0·01 0·01 0·12 0·24 0·10 ilm 0·03 40·17 0·28 56·03 1·10 2·39 0·01 0·00 0·00 0·00 0·01 0·01 0·27 0·60 0·01 apt 0·00 0·00 0·00 0·00 0·00 0·00 56·80 0·00 0·00 43·20 0·00 2·00 40·0 0·64 0·00 wt % SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O P2O5 ppm Rb Sr Y Zr Ba Parent MP-A 57·70 1·07 15·76 9·65 0·21 3·71 7·73 2·97 0·73 0·47 19·2 289 29·6 68 256 Daughter R 75·50 0·45 12·88 2·06 0·12 0·54 2·04 4·49 1·84 0·09 43·3 193 52·3 148 561 Assimilant Hidaka granite 71·86 0·40 14·61 2·26 0·04 0·59 1·73 4·42 4·04 0·06 139·0 78 43·0 308 488 Beppo granodiorite 62·29 0·57 16·76 5·47 0·10 2·78 5·60 3·48 2·75 0·19 68·1 382 22·5 146 620 Rayleigh fractionation model Fractionated phase: 54·5 wt % 31·5 high-An plg, 22·9 low-An plg, 13·1 opx, 15·6 cpx, 9·6 high-Mg mgt, 5·0 low-Mg mgt, 0·5 ilm, 1·7 apt Result ΣR2 = 0·09 75·56 0·50 12·92 2·11 0·06 0·57 2·08 4·49 1·57 0·15 39·2 202 25·2 140 478 R −0·06 −0·05 −0·04 −0·05 0·06 −0·03 −0·04 0·00 0·27 −0·06 4·09 −8·42 27·16 7·66 83·04 R/Daughter 0·00 −0·11 0·00 −0·02 0·53 −0·05 −0·02 0·00 0·15 −0·68 0·09 −0·04 0·52 0·05 0·15 Assimilation and fractional crystallization model (Hidaka granite) Fractionated phase: 54·2 wt % 31·6 high-An plg, 22·6 low-An plg, 12·5 opx, 15·6 cpx, 13·8 high-Mg mgt, 1·0 low-Mg mgt, 1·0 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·10 75·43 0·40 12·98 2·03 0·09 0·62 2·02 4·49 1·84 0·09 57·2 183 25·0 180 528 R 0·07 0·04 −0·10 0·02 0·03 −0·08 0·02 0·00 0·00 0·00 −13·9 10·0 27·4 −32·2 32·6 R/Daughter 0·00 0·10 −0·01 0·01 0·24 −0·15 0·01 0·00 0·00 0·01 −0·3 0·1 0·5 −0·2 0·1 Assimilation and fractional crystallization model (Beppo granodiorite) Fractionated phase: 57·1 wt % 31·8 high-An plg, 22·7 low-An plg, 12·6 opx, 15·5 cpx, 13·5 high-Mg mgt, 0·9 low-Mg mgt, 0·9 ilm, 1·8 apt Result ΣR2 = 0·03, r = 0·14 75·42 0·40 13·00 2·03 0·09 0·63 2·02 4·49 1·84 0·09 55·0 200 23·0 177 607 R 0·08 0·05 −0·12 0·02 0·03 −0·09 0·02 0·00 0·00 0·00 −11·7 −6·9 29·3 −29·6 −46·1 R/Daughter 0·00 0·11 −0·01 0·01 0·26 −0·17 0·01 0·00 0·00 0·02 −0·3 0·0 0·6 −0·2 −0·1 Mineral compositions* Partition coefficient** high-An plg 44·81 0·00 35·06 0·39 0·00 0·09 19·08 0·57 0·01 0·00 0·10 1·50 0·06 0·00 0·20 low-An plg 57·90 0·00 26·28 0·40 0·00 0·03 8·84 6·41 0·14 0·00 0·17 4·04 0·02 0·01 0·47 opx 53·79 0·27 1·74 15·92 0·67 26·27 1·30 0·03 0·00 0·00 0·05 0·01 0·40 0·11 0·02 cpx 53·27 0·30 1·11 8·10 1·02 14·88 20·98 0·34 0·00 0·00 0·07 0·17 2·70 0·17 0·10 high-Mg mgt 0·15 7·70 5·87 81·41 0·33 4·54 0·01 0·00 0·00 0·00 0·01 0·01 0·20 0·10 0·12 low-Mg mgt 0·13 9·35 1·99 85·47 1·22 1·84 0·02 0·00 0·00 0·00 0·01 0·01 0·12 0·24 0·10 ilm 0·03 40·17 0·28 56·03 1·10 2·39 0·01 0·00 0·00 0·00 0·01 0·01 0·27 0·60 0·01 apt 0·00 0·00 0·00 0·00 0·00 0·00 56·80 0·00 0·00 43·20 0·00 2·00 40·0 0·64 0·00 The results of only MP-andesite magma are shown here, because we obtained the similar result using HP-andesite magma. Assimilants are taken from Kamiyama et al. (2007) and Ogasawara et al. (1998). * Mineral compositions are taken from analysed data of the phenocrysts in the most mafic MP-type scoria (No. 114), except for apatite (theoretical value). ** These values are taken from Philpotts & Schnetzler (1970), Arth (1976), Okamoto (1979), Pearce & Norry (1979), Luhr & Carmichael (1980), Gill (1981), Fujimaki (1986), Bacon & Druitt (1988), McKenzie & O’Nions (1991), Sisson (1991), Ewart & Griffin (1994), and Bindeman & Davis (2000). Fig. 12. View largeDownload slide Results of fractional and equilibrium crystallization models using rhyolite-MELTS (Gualda et al., 2012). Fractionated phases are plagioclase, orthopyroxene, clinopyroxene, Fe–Ti oxides, quartz, hornblende, and apatite. Although hornblende was included in the model, hornblende crystals were not observed. Only the results for the MP-andesitic (MP-A) magma are shown because results for the HP-andesitic (HP-A) magma are similar. The results for several elements closely match the compositions of rhyolitic (R) and dacitic (D) end-member magmas, but not for Na2O and P2O5. Fig. 12. View largeDownload slide Results of fractional and equilibrium crystallization models using rhyolite-MELTS (Gualda et al., 2012). Fractionated phases are plagioclase, orthopyroxene, clinopyroxene, Fe–Ti oxides, quartz, hornblende, and apatite. Although hornblende was included in the model, hornblende crystals were not observed. Only the results for the MP-andesitic (MP-A) magma are shown because results for the HP-andesitic (HP-A) magma are similar. The results for several elements closely match the compositions of rhyolitic (R) and dacitic (D) end-member magmas, but not for Na2O and P2O5. Fig. 13. View largeDownload slide Rb variation diagrams showing results of assimilation and fractional crystallization models based on mass-balance calculations (Table 4): (a) Y/Rb for the Rayleigh fractional crystallization (FC) model; (b) Y/Rb for the assimilation and fractional crystallization (AFC) model; (c) Zr/Rb for both models; (d) Ba/Rb for both models. Fractionated phases are listed in Table 4. Results are shown only for the MP-andesitic (MP-A) magma using Hidaka granite (No. 9081106D in Kamiyama et al., 2007) as the assimilant; the results in the case of the HP-andesitic (HP-A) magma, as well as eastern Hokkaido granodiorite (No. Y4 in Ogasawara et al., 1998), are similar. Results for Zr/Rb and Ba/Rb values are consistent with the composition of the rhyolitic end-member magma; however, Y/Rb values become strongly depleted with increasing Rb content, and, therefore, the assimilation and fractional crystallization model fails to replicate the rhyolitic magma. Fig. 13. View largeDownload slide Rb variation diagrams showing results of assimilation and fractional crystallization models based on mass-balance calculations (Table 4): (a) Y/Rb for the Rayleigh fractional crystallization (FC) model; (b) Y/Rb for the assimilation and fractional crystallization (AFC) model; (c) Zr/Rb for both models; (d) Ba/Rb for both models. Fractionated phases are listed in Table 4. Results are shown only for the MP-andesitic (MP-A) magma using Hidaka granite (No. 9081106D in Kamiyama et al., 2007) as the assimilant; the results in the case of the HP-andesitic (HP-A) magma, as well as eastern Hokkaido granodiorite (No. Y4 in Ogasawara et al., 1998), are similar. Results for Zr/Rb and Ba/Rb values are consistent with the composition of the rhyolitic end-member magma; however, Y/Rb values become strongly depleted with increasing Rb content, and, therefore, the assimilation and fractional crystallization model fails to replicate the rhyolitic magma. Given these results, we conclude that the silicic magmas were produced from the accumulation of melts generated from a partial melting zone or incompletely crystallized plutons, in which case a heat source (such as basaltic magma) must have underplated the melting zone. The andesitic magmas we have inferred could have acted as such a heat source (e.g. Streck & Grunder, 1999; Hildreth, 2004). Rhyolitic and dacitic magmas In the Kp IV eruption, preexisting rhyolitic and dacitic magmas mixed. In a SiO2 vs 87Sr/86Sr plot (Fig. 9a), the pumice data form a single linear trend in which higher 87Sr/86Sr ratios are strongly correlated with lower SiO2 contents. This result suggests that the Sr isotopic ratios of the silicic end-member magmas that formed the pumice are distinctly different. We infer that the end-member magmas formed simultaneously, either from partial melts of distinct crustal materials or by assimilation of a parental silicic magma. Primary partial melts from crustal melting range in whole-rock chemistry from rhyolitic to dacitic, depending on the degree of melting of mafic to intermediate crustal materials (e.g. Rapp & Watson, 1995; Springer & Seck, 1997). It is also postulated that large silicic magma systems can be generated by accumulation of melts within the crystal mush zone (e.g. Bachmann & Bergantz, 2004; Hildreth, 2004). In either case, crustal materials are heterogeneous enough to generate isotopically distinct melts forming silicic magmas like those of the Kp IV eruption (Matsumoto & Nakagawa, 2010; Cooper et al., 2012; Ellis et al., 2014; Fig. 14a). Accumulation of interstitial melts from large, long-lived crystal mush zones has been recognized in volcanoes with repeated caldera-forming eruptions, such as the Taupo volcanic zone (e.g. Wilson et al., 2006; Bégué et al., 2014) and the Snake River Plain–Yellowstone Plateau province (e.g. Ellis et al., 2014; Wotzlaw et al., 2014, 2015). Kutcharo volcano, where caldera-forming eruptions have occurred since 400 ka (Hasegawa et al., 2011, 2016), may be a similar case. The Sr and Nd isotopic compositions of Cenozoic granodioritic rocks in the area (87Sr/86Sr = 0·7052–0·7071 and 143Nd/144Nd = 0·5126–0·5129 in central Hokkaido; Jahn et al., 2014) are more enriched than those of Kp IV pumices (87Sr/86Sr < 0·7034 and 143Nd/144Nd > 0·5130; Fig. 9), which appears inconsistent with generation of silicic magmas from long-lived crystal mushes that evolve from preexisting crustal materials (e.g. Hildreth et al., 1991). However, in the Kurile arc extending into eastern Hokkaido, basaltic magmas from Quaternary volcanoes show isotopically depleted compositions similar to Kp IV pumices, with low 87Sr/86Sr and high 143Nd/144Nd (Nakamura & Iwamori, 2009). Also, rhyolitic as well as basaltic magmas of Pliocene and Quaternary age share similar unradiogenic compositions (Takanashi et al., 2012). These features imply that crustal materials in eastern Hokkaido have isotopically depleted compositions. Although the exact formation mechanism of the Kp IV magmas is still unclear, we emphasize that the generation and coexistence of multiple silicic magmas are common features in large silicic magma systems. Fig. 14. View largeDownload slide Schematic diagram of the formation and evolution of the large silicic magma system beneath Kutcharo volcano. The left column shows the crustal setting of the silicic magma system, which is detailed in the right column. (a) Formation of the silicic magma chamber. Partial melts of isotopically heterogeneous crustal materials accumulated to form multiple silicic magmas. The dacitic magma (D) was injected into the rhyolitic magma (R), producing a zoned magma chamber in which the dacitic magma stagnated beneath the rhyolitic magma. (b) Status of the magma chamber just before the Kp IV eruption. Multiple andesitic magmas (MP-A, HP-A, and LP-A) ascended from deeper levels and were injected into the zoned magma chamber, triggering the Kp IV eruption. The upper part of the zoned magma was erupted in the early phase, producing Units 1 and 2, and the lower dacitic zone participated in the climactic phase, producing Unit 3. Just before the eruption of Unit 4, a fresh injection of MP-andesitic magma occurred. Fig. 14. View largeDownload slide Schematic diagram of the formation and evolution of the large silicic magma system beneath Kutcharo volcano. The left column shows the crustal setting of the silicic magma system, which is detailed in the right column. (a) Formation of the silicic magma chamber. Partial melts of isotopically heterogeneous crustal materials accumulated to form multiple silicic magmas. The dacitic magma (D) was injected into the rhyolitic magma (R), producing a zoned magma chamber in which the dacitic magma stagnated beneath the rhyolitic magma. (b) Status of the magma chamber just before the Kp IV eruption. Multiple andesitic magmas (MP-A, HP-A, and LP-A) ascended from deeper levels and were injected into the zoned magma chamber, triggering the Kp IV eruption. The upper part of the zoned magma was erupted in the early phase, producing Units 1 and 2, and the lower dacitic zone participated in the climactic phase, producing Unit 3. Just before the eruption of Unit 4, a fresh injection of MP-andesitic magma occurred. Structure and processes of the magma plumbing system The mixing trends between the silicic end-member magmas do not consistently indicate the involvement of andesitic end-member magmas (Figs 3, 7). This lack of consistency strongly suggests that magma mixing between the silicic end-member magmas occurred before mingling with andesitic magmas. Therefore, we infer that the two silicic magmas were part of a large, zoned silicic magma chamber (Fig. 14a). In contrast, the three types of Kp IV scoria yield data plotting along three clearly separate linear trends that converge toward higher SiO2 contents (Figs 7, 11). This feature strongly suggests that the three andesitic, end-member magmas mixed with the silicic magma independently, forming heterogeneous scorias. The HP and MP scorias plot on distinct linear trends that converge toward the composition of the dacitic rather than the rhyolitic pumice (Fig. 7b). This convergence suggests that their corresponding andesitic magmas mixed with less silicic parts of the zoned silicic magma chamber. Therefore, we conclude that a zoned silicic magma chamber, in which dacitic magma lay beneath rhyolitic magma, existed before the Kp IV eruption (Fig. 14a), and that several andesitic magmas were successively injected into the bottom of the chamber (Fig. 14b). Although we cannot determine the sequence of the ascent and injection of multiple andesitic magmas, the following interpretation is consistent with the petrological features of the Kp IV andesitic magmas. The andesitic magmas have similar incompatible element ratios (Fig. 13), and the MP-andesitic magma appears to be capable of producing the HP-andesitic magma by fractional crystallization (Fig. 12). Thus, it is possible that these andesitic magmas were produced separately by similar processes in a common environment, for example, by fractional crystallization of underplating basaltic magma (e.g. Streck & Grunder, 1999). In addition, the Kp IV andesitic magmas are nearly aphyric and their densities (c. 2400–2500 kg/m3, estimated using the model of Bottinga & Weill, 1970) are clearly less than those of crustal materials. Therefore, the Kp IV andesitic magmas may have buoyantly ascended and may have been injected at nearly the same time. Timing of andesitic magma injection and triggering of the caldera-forming eruption Heterogeneous textures in the HP and MP scorias, such as banding and silicic inclusions, suggest that multiple andesitic magmas were injected just before the Kp IV eruption. Because Fe–Ti oxides have much higher elemental diffusion rates than other minerals, the zoning profiles in Fe–Ti oxides can provide information on magmatic processes shortly before the eruption (e.g. Nakamura, 1995; Coombs et al., 2000). We focus on zoning profiles of Fe–Ti oxides in this section. The cores and rims of Fe–Ti oxides, some Type-R and Type-D magnetites in HP scorias from Unit 3, representing the climactic stage of the eruption, exhibit reverse zoning, whereas those of pumices show weak and normal zoning, especially Type-D ones (Fig. 6, Supplementary DataFig. S4). Focusing on their line profiles, several phenocrysts of Type-R and Type-D magnetites in the HP scorias of Unit 3 show clear reverse zoning, with rims enriched in Mg (Fig. 15a), although they have weak zoning in X’usp (the atomic ratio of ulvöspinel in titanomagnetite) (Supplementary DataTable S4). Type-D magnetite phenocrysts in pumices have clear normal zoning in MgO as well as X’usp (Fig. 16). In contrast, there are few grains showing reverse or normal zoning in MP scorias and pumices from Unit 4. Only one ilmenite exhibits reverse zoning, with a Mg-enriched rim (Figs 6, 15b, and Supplementary DataFig. S4). Fig. 15. View largeDownload slide Compositional profiles of Fe–Ti oxide phenocrysts that show reverse zoning; data acquired at intervals of 1–3 µm Diffusion profiles were calculated using the compositionally independent diffusion model of Crank (1975) for a sphere. The diffusion coefficients of Mg are 0·09–1·56 × 10–15 m2/s, estimated using Type-D magnetite (see Fig. 16). Abbreviations: D, partition coefficient of Mg; t, time of diffusion; a, radius of grain. Black solid lines are the best fitting curves for each phenocryst. See the text for discussion. (a) Type-R magnetite phenocrysts from Unit 3. Reverse zoning in these phenocrysts fits curves ranging from Dt/a2 = 0·00125 to Dt/a2 = 0·08, representing very short diffusion times (hours to weeks). (b) Type-R ilmenite phenocryst from Unit 4 with reverse zoning. Although the diffusion rate of Mg in ilmenite is unknown, with the same diffusion coefficient used for magnetite this reverse zoning fits the curves from Dt/a2 = 0·02 to Dt/a2 = 0·04, consistent with magma injection just before the eruption of Unit 4. Fig. 15. View largeDownload slide Compositional profiles of Fe–Ti oxide phenocrysts that show reverse zoning; data acquired at intervals of 1–3 µm Diffusion profiles were calculated using the compositionally independent diffusion model of Crank (1975) for a sphere. The diffusion coefficients of Mg are 0·09–1·56 × 10–15 m2/s, estimated using Type-D magnetite (see Fig. 16). Abbreviations: D, partition coefficient of Mg; t, time of diffusion; a, radius of grain. Black solid lines are the best fitting curves for each phenocryst. See the text for discussion. (a) Type-R magnetite phenocrysts from Unit 3. Reverse zoning in these phenocrysts fits curves ranging from Dt/a2 = 0·00125 to Dt/a2 = 0·08, representing very short diffusion times (hours to weeks). (b) Type-R ilmenite phenocryst from Unit 4 with reverse zoning. Although the diffusion rate of Mg in ilmenite is unknown, with the same diffusion coefficient used for magnetite this reverse zoning fits the curves from Dt/a2 = 0·02 to Dt/a2 = 0·04, consistent with magma injection just before the eruption of Unit 4. Fig. 16. View largeDownload slide Line profiles of X’usp (top) and MgO (bottom) in three Type-D magnetite phenocrysts with normal zoning; electron back-scatter images of these phenocrysts are shown in Supplementary DataFig. S3(f–h). Black solid and broken lines are the best fitting lines for each phenocryst. We estimated the Mg diffusion coefficient for magnetite as follows. First, the diffusion times of X’usp for Type-D magnetite in pumice were estimated using the sphere model of Crank (1975) and the Ti partition coefficient of Freer & Hauptman (1978). The assumed temperature is that of the dacitic end-member magma, 870°C. Second, we fitted the line profiles to Dt/a2 curves. From that result, we estimated the diffusion coefficients using the diffusion times obtained from X’usp. Abbreviations are the same as those in Fig. 15. Fig. 16. View largeDownload slide Line profiles of X’usp (top) and MgO (bottom) in three Type-D magnetite phenocrysts with normal zoning; electron back-scatter images of these phenocrysts are shown in Supplementary DataFig. S3(f–h). Black solid and broken lines are the best fitting lines for each phenocryst. We estimated the Mg diffusion coefficient for magnetite as follows. First, the diffusion times of X’usp for Type-D magnetite in pumice were estimated using the sphere model of Crank (1975) and the Ti partition coefficient of Freer & Hauptman (1978). The assumed temperature is that of the dacitic end-member magma, 870°C. Second, we fitted the line profiles to Dt/a2 curves. From that result, we estimated the diffusion coefficients using the diffusion times obtained from X’usp. Abbreviations are the same as those in Fig. 15. We consider that these zoning profiles were produced mainly by elemental diffusion, due to the narrower width of zonation (10–30 μm) as well as the faster elemental diffusion in magnetite. Therefore, we can estimate the diffusion times of these profiles to evaluate the timescales from magma mixing to eruption. For the estimation, we used a compositionally independent diffusion model assuming a sphere (equation 6.1 in Crank, 1975) and assumed a homogeneous magnetite core mantled by a 1-µm-thick reversely zoned rim, whose compositions are the same as in our data for the initial condition. Lacking published data for the diffusion rate of Mg in magnetite, we estimated the diffusion rate on the basis of the line profiles of MgO and Ti (based on X’usp) in three Type-D magnetite phenocrysts with normal zoning (Fig. 16) as follows. First, we estimated the diffusion times using the line profiles in Xʹusp of these phenocrysts on the basis of the diffusion coefficient of Ti (9·25–9·50 × 10–17 m2/s at 870°C, the approximate temperature of Type-D Fe-Ti oxides; Freer & Hauptman, 1978): c. 20–120 h. Second, we fitted the line profiles in MgO to Dt/a2 curves (D, partition coefficient of Mg; t, time of diffusion; a, radius of grain). From these fitting results and diffusion times of Ti, we estimated the diffusion coefficients of Mg. The corresponding diffusion rate of Mg is 0·09–1·56 × 10–15 m2/s. These values are similar to the diffusion rate of divalent elements (Fe, Co and Mn: 0·44–3·30 × 10–15 m2/s at log fO2 = -11; Aggarwal & Dieckmann, 2002; Van Orman & Crispin, 2010). Using these estimated diffusion coefficients, the diffusion times of Mg in four phenocrysts of Type-R and Type-D magnetite with reverse zonation in Unit 3 range from c. 4 h to 26 days (Fig. 15a, Table 5). Unit 4 yielded no magnetite grains showing compositional zoning (Fig. 6, Supplementary DataFig. S4), but we did characterize a Type-R ilmenite phenocryst (No. 114–56a) that had a thin diffusion mantle (10–20 µm) reversely zoned with respect to Mg (Fig. 15b). The existence of this ilmenite suggests that magma mixing and the eruption of Unit 4 occurred on the same time scale as for Unit 3 (Table 5). Table 5: Estimated diffusion times of Mg in Fe–Ti oxides in scorias of the Kp IV eruption Grain No. Type Unit No. MgO wt % core MgO wt % rim Diffusion time* 355-ox31 Type-D mgt Unit 3 2·08 2·39 4·4 hours – 3·1 days 355-oxad33 Type-R mgt Unit 3 1·70 1·88 6·4 hours – 4·5 days 140-ox46 Type-R mgt Unit 3 1·84 2·19 26·3 hours – 18·3 days 140-ox49 Type-R mgt Unit 3 1·65 1·84 37·0 hours – 25·7 days 114-ox56 Type-R ilm Unit 4 2·32 2·54 17·4 hours – 24·2 days Grain No. Type Unit No. MgO wt % core MgO wt % rim Diffusion time* 355-ox31 Type-D mgt Unit 3 2·08 2·39 4·4 hours – 3·1 days 355-oxad33 Type-R mgt Unit 3 1·70 1·88 6·4 hours – 4·5 days 140-ox46 Type-R mgt Unit 3 1·84 2·19 26·3 hours – 18·3 days 140-ox49 Type-R mgt Unit 3 1·65 1·84 37·0 hours – 25·7 days 114-ox56 Type-R ilm Unit 4 2·32 2·54 17·4 hours – 24·2 days * Diffusion times are estimated using diffusion coefficients of Mg (Kd) = 0·94 x 10–16 and 1·56 x 10–15 m2/s (calculated by Type-D magnetite showing normal zoning in MgO as well as X’usp, see text and Supplementary Data). Table 5: Estimated diffusion times of Mg in Fe–Ti oxides in scorias of the Kp IV eruption Grain No. Type Unit No. MgO wt % core MgO wt % rim Diffusion time* 355-ox31 Type-D mgt Unit 3 2·08 2·39 4·4 hours – 3·1 days 355-oxad33 Type-R mgt Unit 3 1·70 1·88 6·4 hours – 4·5 days 140-ox46 Type-R mgt Unit 3 1·84 2·19 26·3 hours – 18·3 days 140-ox49 Type-R mgt Unit 3 1·65 1·84 37·0 hours – 25·7 days 114-ox56 Type-R ilm Unit 4 2·32 2·54 17·4 hours – 24·2 days Grain No. Type Unit No. MgO wt % core MgO wt % rim Diffusion time* 355-ox31 Type-D mgt Unit 3 2·08 2·39 4·4 hours – 3·1 days 355-oxad33 Type-R mgt Unit 3 1·70 1·88 6·4 hours – 4·5 days 140-ox46 Type-R mgt Unit 3 1·84 2·19 26·3 hours – 18·3 days 140-ox49 Type-R mgt Unit 3 1·65 1·84 37·0 hours – 25·7 days 114-ox56 Type-R ilm Unit 4 2·32 2·54 17·4 hours – 24·2 days * Diffusion times are estimated using diffusion coefficients of Mg (Kd) = 0·94 x 10–16 and 1·56 x 10–15 m2/s (calculated by Type-D magnetite showing normal zoning in MgO as well as X’usp, see text and Supplementary Data). These time scales (hours to weeks) are shorter than those of magma accumulation and silicic mixing (e.g. Druitt et al., 2012; Allan et al., 2013) and are comparable to the timing of mafic injections in previously observed silicic explosive eruptions (e.g. Unzen 1991 eruption, Nakamura, 1995; Soufrière Hills 1995–2002 eruptions, Devine et al., 2003; Shinmoe-dake 2011 eruption, Tomiya et al., 2013). This agreement suggests that injection of mafic magmas triggered the Kp IV caldera-forming eruption. It also indicates that the injection of hot mafic magma could supply volatiles to the silicic magma, creating overpressures that could lead to eruption (e.g. Folch & Martí, 1998). Therefore, our interpretation is that the Kp IV caldera-forming eruption was triggered by andesitic injections. Likewise, the Unit 4 eruption was probably triggered by an injection of andesitic magma after the Unit 3 eruption because the estimated time scale of this process is similar to, or smaller than, the time between eruptions indicated by geologic evidence (several days to two months, as estimated using the degassing model of Riehle et al. (1995) for the Unit 3 pyroclastic flow deposit (Hasegawa et al., 2016)). The change of mafic magma from mainly HP-andesite in Unit 3 to MP-andesite in Unit 4 might also suggest the new injection. Given the number of other reports that mafic magmas mingle with silicic magmas during caldera-forming eruptions (Streck & Grunder, 1999; Reubi & Nicholls, 2005; Wilson et al., 2006; xWright et al., 2011), mafic magma injection may be a common factor in triggering caldera-forming eruptions. CONCLUSIONS We investigated the structure and eruptive processes of the large silicic magma system responsible for the Kp IV caldera-forming activity of Kutcharo volcano. Our major conclusions are as follows: Before the caldera-forming eruption, the magma plumbing system consisted of a large, shallow silicic chamber and a set of deeper, smaller pockets of andesitic magma. The silicic magma chamber was zoned, with rhyolitic magma overlying dacitic magma. The rhyolitic and dacitic magmas could not have been produced by assimilation or fractional crystallization of the andesitic magmas, but formed instead by accumulation of melts generated from crustal materials, for which the andesitic magmas served as a heat source. Likewise, the dacitic magma could not have produced the rhyolitic magma by fractional crystallization, suggesting that distinct silicic magmas were produced contemporaneously from heterogeneous crustal materials. Such generation of multiple silicic magmas may be common in large silicic magma systems. Hours to weeks before the caldera-forming eruption, multiple andesitic magmas were injected into the zoned silicic magma chamber. These injections may have been the final trigger of this catastrophic eruption. ACKNOWLEDGEMENTS We are grateful to Masataka Ikeda for supporting our XRF and EPMA analyses, and to Hidehiko Nomura and Kousuke Nakamura for preparing the thin sections. This manuscript was greatly improved as a result of reviews by Ben Ellis and two anonymous reviewers. We also thank Georg Zellmer for his editorial work. FUNDING This work was supported by JSPS KAKENHI (Grant No. JP15H03745) for M.N. and by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, under its ‘Earthquake and Volcano Hazards Observation and Research Program’ and ‘Integrated Program for Next Generation Volcano Research and Human Resource Development’. SUPPLEMENTARY DATA Supplementary data are available at Journal of Petrology online. REFERENCES Aggarwal S. , Dieckmann R. ( 2002 ). Point defects and cation tracer diffusion in (TixFe1-x)3-δO4. II. Cation tracer diffusion . Physics and Chemistry of Minerals 29 , 707 – 718 . Google Scholar CrossRef Search ADS Allan A. S. R. , Morgan D. J. , Wilson C. J. N. , Millet M.-A. ( 2013 ). From mush to eruption in centuries: assembly of the super-sized Oruanui magma body . Contributions to Mineralogy and Petrology 166 , 143 – 164 . Google Scholar CrossRef Search ADS Andersen D. J. , Lindsley D. H. , Davidson P. M. ( 1993 ). QUILF: a Pascal program to assess equilibria among Fe-Mg-Mn-Ti oxides, pyroxenes, olivine, and quartz. Computers and Geosciences 19 , 1333 – 1350 . Arth J. G. ( 1976 ). Behavior of trace elements during magmatic processes- a summary of theoretical models and their applications . Journal of Research of the U.S. Geological Survey 4 , 41 – 47 . Bachmann O. , Bergantz G. W. ( 2004 ). On the origin of crystal-poor rhyolites: extracted from batholithic crystal mushes . Journal of Petrology 45 , 1565 – 1582 . Google Scholar CrossRef Search ADS Bachmann O. , Bergantz G. W. ( 2008 ). Rhyolites and their source mushes across tectonic settings . Journal of Petrology 49 , 2277 – 2285 . Google Scholar CrossRef Search ADS Bacon C. R. , Druitt T. H. ( 1988 ). Compositional evolution of the zoned calcalkaline magma chamber of Mount Mazama, Crater Lake, Oregon . Contributions to Mineralogy and Petrology 98 , 224 – 256 . Google Scholar CrossRef Search ADS Bacon C. R. , Hirschmann M. M. ( 1988 ). Mg/Mn partitioning as a test for equilibrium between coexisting Fe-Ti oxides . American Mineralogist 73 , 57 – 61 . Baker D. R. , Eggler D. H. ( 1987 ). Compositions of anhydrous and hydrous melts coexisting with plagioclase, augite, and olivine or low-Ca pyroxene from 1atm to 8 kbar: application to the Aleutian volcanic center of Atka . American Mineralogist 72 , 12 – 28 . Bégué F. , Deering C. D. , Gravley D. M. , Kennedy B. M. , Chambefort I. , Gualda G. A. R. , Bachmann O. ( 2014 ). Extraction, storage and eruption of multiple isolated magma batches in the paired Mamaku and Ohakuri eruption, Taupo Volcanic Zone, New Zealand . Journal of Petrology 55 , 1653 – 1684 . Google Scholar CrossRef Search ADS Bindeman I. N. , Davis A. M. ( 2000 ). Trace element partitioning between plagioclase and melt: investigation of dopant influence on partition behavior . Geochimica et Cosmochimica Acta 64 , 2863 – 2878 . Google Scholar CrossRef Search ADS Bottinga Y. , Weill D. F. ( 1970 ). Densities of liquid silicate systems calculatedfrom partial molar volumes of oxide components . American Journal of Science 269 , 169 – 182 . Google Scholar CrossRef Search ADS Brophy J. G. , Dreher S. T. ( 2000 ). The origin of composition gaps at South Sister volcano, central Oregon: implications for fractional crystallization processes beneath active calc-alkaline volcanoes . Journal of Volcanology and Geothermal Research 102 , 287 – 307 . Google Scholar CrossRef Search ADS Brown S. J. A. , Wilson C. J. N. , Cole J. W. , Wooden J. ( 1998 ). The Whakamaru group ignimbrites, Taupo Volcanic Zone, New Zealand: evidence for reverse tapping of a zoned silicic magmatic system . Journal of Volcanology and Geothermal Research 84 , 1 – 37 . Google Scholar CrossRef Search ADS Cashman K. V. , Giordano G. ( 2014 ). Calderas and magma reservoirs . Journal of Volcanology and Geothermal Research 288 , 28 – 45 . Google Scholar CrossRef Search ADS Chamberlain K. J. , Wilson C. J. N. , Wallace P. J. , Millet M.-A. ( 2015 ). Micro-analytical perspectives on the Bishop Tuff and its magma chamber . Journal of Petrology 56 , 605 – 640 . Google Scholar CrossRef Search ADS Coombs M. L. , Eichelberger J. C. , Rutherford M. J. ( 2000 ). Magma storage and magma mixing conditions for the 1953-1974 eruptions of Southwest Trident volcano, Katmai National Park, Alaska . Contributions to Mineralogy and Petrology 140 , 99 – 118 . Google Scholar CrossRef Search ADS Cooper G. F. , Wilson C. J. N. , Millet M.-A. , Baker J. A. , Smith E. G. C. ( 2012 ). Systematic tapping of independent magma chambers during the 1 Ma Kidnappers supereruption . Earth and Planetary Science Letters 313–314 , 23 – 33 . Google Scholar CrossRef Search ADS Crank J. ( 1975 ). The Mathematics of Diffusion . London : Oxford University Press . Deering C. D. , Cole J. W. , Vogel T. A. ( 2011b ). Extraction of crystal-poor rhyolite from a hornblende-bearing intermediate mush: a case study of the caldera-forming Matahina eruption, Okataina volcanic complex . Contributions to Mineralogy and Petrology 161 , 129 – 151 . Google Scholar CrossRef Search ADS Deering C. D. , Bachmann O. , Dufek J. , Gravley D. M. ( 2011a ). Rift-related transition from andesite to rhyolite volcanism in the Taupo Volcanic Zone (New Zealand) controlled by crystal-melt dynamics in mush zones with variable mineral assemblages . Journal of Petrology 52 , 2243 – 2263 . Google Scholar CrossRef Search ADS DePaolo D. J. ( 1981 ). Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization . Earth and Planetary Science Letters 53 , 189 – 202 . Google Scholar CrossRef Search ADS Devine J. D. , Rutherford M. J. , Norton G. E. , Young S. R. ( 2003 ). Magma storage region processes inferred from geochemistry of Fi-Ti oxides in andesitic magma, Soufrière Hills volcano, Montserrat, W.I . Journal of Petrology 44 , 1375 – 1400 . Google Scholar CrossRef Search ADS Druitt T. H. , Costa F. , Deloule E. , Dungan M. , Scaillet B. ( 2012 ). Decadal to monthly timescales of magma transfer and reservoir growth at a caldera volcano . Nature 482 , 77 – 82 . Google Scholar CrossRef Search ADS PubMed Eggins S. M. , Woodhead J. D. , Kinsley L. P. J. , Mortimer G. E. , Sylvester P. , McCulloch M. T. , Hergt J. M. , Handler M. R. ( 1997 ). A simple method for the precise determination of ≥40 trace elements in geological samples by ICPMS using enriched isotope internal standardization . Chemical Geology 134 , 311 – 326 . Google Scholar CrossRef Search ADS Ellis B. S. , Bachmann O. , Wolff J. A. ( 2014 ). Cumulate fragments in silicic ignimbrites: the case of the Snake River Plain . Geology 42 , 431 – 434 . Google Scholar CrossRef Search ADS Ewart A. , Griffin W. L. ( 1994 ). Application of proton-microprobe data to trace-element partitioning in volcanic rocks . Chemical Geology 117 , 251 – 284 . Google Scholar CrossRef Search ADS Folch A. , Martí J. ( 1998 ). The generation of overpressure in felsic magma chambers by replenishment . Earth and Planetary Science Letters 163 , 301 – 314 . Google Scholar CrossRef Search ADS Folkes C. B. , de Silva S. L. , Wright H. M. , Cas R. A. F. ( 2011 ). Geochemical homogeneity of a long-lived, large silicic system; evidence from the Cerro Galán caldera, NW Argentina . Bulletin of Volcanology 73 , 1455 – 1486 . Google Scholar CrossRef Search ADS Freer R. , Hauptman Z. ( 1978 ). An experimental study of magnetite-titanomagnetite interdiffusion . Physics of the Earth and Planetary Interiors 16 , 223 – 231 . Google Scholar CrossRef Search ADS Fujimaki F. ( 1986 ). Fractional crystallization of the basaltic suite of Usu volcano, southwest Hokkaido, Japan, and its relationships with the associated felsic suite . Lithos 19 , 129 – 140 . Google Scholar CrossRef Search ADS Gill J. B. ( 1981 ). Orgenic Andesites and Plate Tectonics . Berlin : Springer . Google Scholar CrossRef Search ADS Gualda G. A. R. , Ghiorso M. S. , Lemons R. V. , Carley T. L. ( 2012 ). Rhyolite-MELTS: a modified calibration of MELTS optimized for silica-rich, fluid-bearing magmatic systems . Journal of Petrology 53 , 875 – 890 . Google Scholar CrossRef Search ADS Hasegawa T. , Nakagawa M. ( 2007 ). Stratigraphy of Early to Middle Pleistocene pyroclastic deposits around Akan caldera, eastern Hokkaido, Japan . The Journal of the Geological Society of Japan 113 , 53 – 72 . (in Japanese with English abstract). Google Scholar CrossRef Search ADS Hasegawa T. , Kishimoto H. , Nakagawa M. ( 2012 ). The eruption history and silicic magma systems of caldera-forming eruptions in eastern Hokkaido, Japan . Journal of Mineralogical and Petrological Sciences 107 , 39 – 43 . Google Scholar CrossRef Search ADS Hasegawa T. , Matsumoto A. , Nakagawa M. ( 2016 ). Evolution of the 120 ka caldera-forming eruption from Kutcharo volcano, east Hokkaido, Japan: Geologic and petrologic evidence for multiple vent systems and rapid generation of pyroclastic flow . Journal of Volcanology and Geothermal Research 321 , 58 – 72 . Google Scholar CrossRef Search ADS Hasegawa T. , Kishimoto H. , Nakagawa M. , Ito J. , Yamamoto T. ( 2009 ). Eruptive history of post-caldera volcanoes of Kutcharo caldera, eastern Hokkaido, Japan, as inferred from tephrostratigraphy in the Konsen and Shari areas for the period 35-12 ka . The Journal of the Geological Society of Japan 115 , 369 – 390 . (in Japanese with English abstract). Google Scholar CrossRef Search ADS Hasegawa T. , Nakagawa M. , Itoh J. , Yamamoto T. ( 2011 ). Depositional ages of Quaternary sequences in Kushiro region, eastern Hokkaido, Japan: Correlations and chronology on the basis of high-resolution tephro-stratigraphy . The Journal of the Geological Society of Japan 117 , 686 – 699 . (in Japanese with English abstract). Google Scholar CrossRef Search ADS Hildreth W. ( 2004 ). Volcanological perspectives on Long Valley, Mammoth Mountain, and Mono Craters: several contiguous but discrete systems . Journal of Volcanology and Geothermal Research 136 , 169 – 198 . Google Scholar CrossRef Search ADS Hildreth W. , Wilson C. J. N. ( 2007 ). Compositional zoning of the Bishop Tuff . Journal of Petrology 48 , 951 – 999 . Google Scholar CrossRef Search ADS Hildreth W. , Halliday A. N. , Christiansen R. L. ( 1991 ). Isotopic and chemical evidence concerning the genesis and contamination of basaltic and rhyolitic magma beneath the Yellowstone Plateau volcanic field . Journal of Petrology 32 , 63 – 138 . Google Scholar CrossRef Search ADS Jahn B.-M. , Usuki M. , Usuki T. , Chung S.-L. ( 2014 ). Generation of Cenozoic granitoids in Hokkaido (Japan): constraints from zircon geochronology, Sr-Nd-Hf isotopic geochemical analyses, and implications for crustal growth . American Journal of Science 314 , 704 – 750 . Google Scholar CrossRef Search ADS Kamiyama H. , Nakajima T. , Kamioka H. ( 2007 ). Magmatic stratigraphy of the tilted Tottabetsu plutonic complex, Hokkaido, North Japan: magma chamber dynamics and pluton construction . Journal of Geology 115 , 295 – 314 . Google Scholar CrossRef Search ADS Kilinc A. , Carmichael I. S. E. , Rivers M. L. , Sack R. O. ( 1983 ). The ferric-ferrous ration of natural silicate liquids equilibrated in air . Contributions to Mineralogy and Petrology 83 , 136 – 140 . Google Scholar CrossRef Search ADS Leonard G. S. , Cole J. W. , Nairn I. A. , Self S. ( 2002 ). Basalt triggering of the c. AD 1305 Kaharoa rhyolite eruption, Tarawera Volcanic Complex, New Zealand . Journal of Volcanology and Geothermal Research 115 , 461 – 486 . Google Scholar CrossRef Search ADS Luhr J. F. , Carmichael I. S. E. ( 1980 ). The Colima volcanic complex, Mexico. I. Post-caldera andesite from Volcán Colima . Contributions to Mineralogy and Petrology 71 , 343 – 372 . Google Scholar CrossRef Search ADS Machida H. , Arai F. ( 2003 ). Atlas of Tephra in and around Japan . Tokyo: University of Tokyo Press . (in Japanese). Matsumoto A. , Nakagawa M. ( 2010 ). Formation and evolution of silicic magma plumbing system: petrology of the volcanic rocks of Usu volcano, Hokkaido, Japan . Journal of Volcanology and Geothermal Research 196 , 185 – 207 . Google Scholar CrossRef Search ADS McKenzie D. , O’Nions R. K. ( 1991 ). Partial melt distributions from inversion of rare earth element concentrations . Journal of Petrology 32 , 1021 – 1091 . Google Scholar CrossRef Search ADS Milner D. M. , Cole J. W. , Wood C. P. ( 2003 ). Mamaku ignimbrite: a caldera-forming ignimbrite erupted from a compositionally zoned magma chamber in Taupo Volcanic Zone, New Zealand . Journal of Volcanology and Geothermal Research 122 , 243 – 264 . Google Scholar CrossRef Search ADS Miyagi I. , Itoh J. , Hoang N. , Morishita Y. ( 2012 ). Magma systems of the Kutcharo and Mashu volcanoes (NE Hokkaido, Japan): petrogenesis of the medium-K trend and the excess volatile problem . Journal of Volcanology and Geothermal Research 231–232 , 50 – 60 . Google Scholar CrossRef Search ADS Moore G. , Vennemann T. , Carmichael I. S. E. ( 1998 ). An empirical model for the solubility of H2O in magmas to 3 kilobars . American Mineralogist 83 , 36 – 42 . Google Scholar CrossRef Search ADS Nakamura H. , Iwamori H. ( 2009 ). Contribution of slab-fluid in arc magmas beneath Japan arcs . Gondwana Research 16 , 431 – 445 . Google Scholar CrossRef Search ADS Nakamura M. ( 1995 ). Continuous mixing of crystal mush and replenished magma in the ongoing Unzen eruption . Geology 23 , 807 – 810 . Google Scholar CrossRef Search ADS Ogasawara M. , Shibata K. , Uchiumi S. ( 1998 ). K-Ar ages and petrological characteristics of granotoid pebbles from the Eocene Beppo Formation in the Kushiro region, Hokkaido, Japan . The Journal of the Geological Society of Japan 104 , 516 – 524 . (in Japanese with English abstract). Google Scholar CrossRef Search ADS Okamoto K. ( 1979 ). Geochemical study on magmatic differentiation of Asama volcano, central Japan . The Journal of the Geological Society of Japan 85 , 525 – 535 . Google Scholar CrossRef Search ADS Orihashi Y. , Maeda J. , Tanaka R. , Zeniya R. , Niida K. ( 1998 ). Sr and Nd isotopic data for the seven GSJ rock reference samples; JA-1, JB-1a, JB-2, JB-3. JG-1a, JGb-1 and JR-1 . Geochemical Journal 32 , 205 – 211 . Google Scholar CrossRef Search ADS Parker D. F. , Ren M. , Adams D. T. , Tsai H. , Long L. E. ( 2012 ). Mid-Tertiary magmatism in western Big Bend National Park, Texas, U.S.A.: evolution of basaltic source regions and generation of peralkaline rhyolite . Lithos 144–145 , 161 – 176 . Google Scholar CrossRef Search ADS Pearce J. A. , Norry M. J. ( 1979 ). Petrogenetic implications of Ti, Zr, Y, and Nb variations in volcanic rocks . Contributions to Mineralogy and Petrology 69 , 33 – 47 . Google Scholar CrossRef Search ADS Philpotts J. A. , Schnetzler C. C. ( 1970 ). Phenocryst-matrix partition coefficients for K, Rb, Sr and Ba, with applications to anorthosite and basalt genesis . Geochimica et Cosmochimica Acta 34 , 307 – 322 . Google Scholar CrossRef Search ADS Putirka K. D. ( 2008 ). Thermometers and barometers for volcanic systems. In: Putirka, K. D. & Tepley F. J., III (eds) Minerals, Inclusions and Volcanic Processes. Mineralogical Society of America and Geochemical Society, Reviews in Mineralogy and Petrology 69 , 61 – 111 . Rapp R. P. , Watson E. B. ( 1995 ). Dehydration melting of metabasalt at 8-32 kbar: implications for continental growth and crust-mantle recycling . Journal of Petrology 36 , 891 – 931 . Google Scholar CrossRef Search ADS Reubi O. , Nicholls I. A. ( 2005 ). Structure and dynamics of a silicic magmatic system associated with caldera-forming eruptions at Batur volcanic field, Bali, Indonesia . Journal of Petrology 46 , 1367 – 1391 . Google Scholar CrossRef Search ADS Riehle J. R. , Miller T. F. , Bailey R. A. ( 1995 ). Cooling, degassing, and compaction of rhyolitic ash flow tuffs: a computational model . Bulletin of Volcanology 57 , 319 – 336 . Google Scholar CrossRef Search ADS Roser B. , Kimura J. , Hisatomi K. ( 2000 ). Whole-rock elemental abundances in sandstones and mudrocks from the Takabe Group, Kii Peninsula, Japan . Geoscience Reports of Shimane University 19 , 101 – 112 . Sack R. O. , Carmichael I. S. E. , Rivers M. , Ghiorso M. S. ( 1981 ). Ferric-ferrous equilibria in natural silicate liquids at 1 bar . Contributions to Mineralogy and Petrology 75 , 369 – 376 . Google Scholar CrossRef Search ADS Sisson T. W. ( 1991 ). Pyroxene-high silica rhyolite trace element partition coefficients measured by ion microprobe . Geochimica et Cosmochimica Acta 55 , 1575 – 1585 . Google Scholar CrossRef Search ADS Smith D. R. , Leeman W. P. ( 1987 ). Petrogenesis of Mount St. Helens dacitic magmas . Journal of Geophysical Research: Solid Earth 92 , 10313 – 10334 . Google Scholar CrossRef Search ADS Springer W. , Seck A. ( 1997 ). Partial fusion of basic granulites at 5 and 15 kbar: implications for the origin of TTG magmas . Contributions to Mineralogy and Petrology 127 , 30 – 45 . Google Scholar CrossRef Search ADS Streck M. J. , Grunder A. L. ( 1997 ). Compositional gradients and gaps in high-silica rhyolites of the Rattlesnake Tuff, Oregon . Journal of Petrology 38 , 133 – 163 . Google Scholar CrossRef Search ADS Streck M. J. , Grunder A. L. ( 1999 ). Enrichment of basalt and mixing of dacite in the rootzone of a large rhyolite chamber: inclusions and pumices from the Rattlesnake Tuff, Oregon . Contributions to Mineralogy and Petrology 136 , 193 – 212 . Google Scholar CrossRef Search ADS Streck M. J. , Grunder A. L. ( 2008 ). Phenocryst-poor rhyolites of bimodal, tholeiitic provinces: The Rattlesnake Tuff and implications for mush extraction models . Bulletin of Volcanology 70 , 385 – 401 . Google Scholar CrossRef Search ADS Sun S.–S. , McDonough W. F. ( 1989 ). Chemical and isotopic systematics of ocean 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 . Google Scholar CrossRef Search ADS Takanashi K. , Kakihara Y. , Ishimoto H. , Shuto K. ( 2012 ). Melting of crustal rocks as a possible origin for Middle Miocene to Quaternary rhyolites of northeast Hokkaido, Japan: constraints from Sr and Nd isotopes and major- and trace-element chemistry . Journal of Volcanology and Geothermal Research 221–222 , 52 – 70 . Google Scholar CrossRef Search ADS Tomiya A. , Miyagi I. , Saito G. , Geshi N. ( 2013 ). Short time scales of magma-mixing processes prior to the 2011 eruption of Shinmoedake volcano, Kirishima volcanic group, Japan . Bulletin of Volcanology 75 , 1 – 19 . Google Scholar CrossRef Search ADS Van Orman J. A. , Crispin K. L. ( 2010 ). Diffusion in oxides . In: Zhang, Y. & Cherniak, D. J. (eds) Diffusion in Minerals and Melts. Mineralogical Society of America and Geochemical Society, Reviews in Mineralogy and Petrology 72 , 757 – 825 . Wilcock J. , Goff F. , Minarik W. G. , Stix J. ( 2013 ). Magmatic recharge during the formation and resurgence of the Valles caldera, New Mexico, USA: evidence from quartz compositional zoning and geothermometry . Journal of Petrology 54 , 635 – 664 . Google Scholar CrossRef Search ADS Wilson C. J. N. , Blake S. , Charlier B. L. A. , Sutton A. N. ( 2006 ). The 26.5 ka Oruanui eruption, Taupo volcano, New Zealand: development, characteristics and evacuation of a large rhyolitic magma body . Journal of Petrology 47 , 35 – 69 . Google Scholar CrossRef Search ADS Wotzlaw J.-F. , Bindeman I. N. , Stern R. A. , D’Abzac F.-X. , Schaltegger U. ( 2015 ). Rapid heterogeneous assembly of multiple magma reservoirs prior to Yellowstone supereruptions . Scientific Reports 5 , doi:10.1038/srep14026. Wotzlaw J.-F. , Bindeman I. N. , Watts K. E. , Schmitt A. K. , Caricchi L. , Schaltegger U. ( 2014 ). Linking rapid magma reservoir assembly and eruption trigger mechanisms at evolved Yellowstone-type supervolcanoes . Geology 42 , 807 – 810 . Google Scholar CrossRef Search ADS Wright H. M. N. , Folkes C. B. , Cas R. A. F. , Cashman K. V. ( 2011 ). Heterogeneous pumice populations in the 2.08-Ma Cerro Galán ignimbrite: imlications for magma recharge and ascent preceding a large-volume silicic eruption . Bulletin of Volcanology 73 , 1513 – 1533 . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of PetrologyOxford University Press

Published: May 3, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off