Origin of two types of olivine from the Ogi Picritic Dolerite Sill, northeast Japan

Abstract

A Middle Miocene picritic dolerite sill is exposed to the southwest of the Sado Island, the largest island in the Sea of Japan. This picrite has two types of phenocrystic olivine: high-Fo and -Ca normally zoned olivine with abundant Cr-spinel inclusions (type-1) and low-Fo and -Ca reverse-zoned olivine without Cr-spinel inclusions (type-2). Type-1 olivine and type-2 olivine mantle are in equilibrium with the chilled margin of the sill, but the type-2 olivine core is clearly in disequilibrium with that. This suggests that the types-1 and -2 olivine crystals may be autocrysts and antecrysts, respectively. These two types of olivines contain multiphase solid inclusions in their cores. These inclusions are also classified into the following two types based on the mineral assemblage and average bulk composition: plagioclase + clinopyroxene + hornblende assemblage with basaltic composition (O1-type) and orthopyroxene + clinopyroxene + hornblende (± plagioclase) assemblage with high-magnesian andesitic composition (O2c-type). The composition of glasses coexisting with plagioclase in multiphase solid inclusions indicates that plagioclase crystallized at the early stage of crystallization in the O1-type inclusions but did not crystallize until the last stage in the O2c-type. The mineral assemblages suggest that the O1- and O2c-types multiphase solid inclusions represent the trapped basaltic and high-Mg andesitic melts, respectively. Our results suggest that the type-1 olivine is autocrysts crystallized from a primitive basaltic magma and the type-2 olivine core is antecrysts crystallized from an H₂O-rich high-Mg andesitic magma. After the type-2 olivine core crystallized at the deeper parts (>130-190 MPa), it was then incorporated into the primitive basaltic magma. Consequently, the high-Fo and -Ca olivine mantles grew around the type-2 olivine cores under the condition of moderate undercooling at a shallower depth. The high-Mg andesitic magmatism inferred from the O2c-type multiphase solid inclusions in the picrite suggest that the highly depleted oceanic mantle infiltrated by slab-derived melts/fluids may have existed ubiquitously under the back-arc basin during the Miocene back-arc opening of the Japan Sea.

INTRODUCTION

Picrites are common volcanic rocks in the areas of upwelling mantle plumes such as Hawaii (e.g., Wilkinson and Hensel, 1988) and Réunion (e.g., Albaréde and Tamagnan, 1988), and large igneous provinces such as Paraná-Etendeka (e.g., Thompson et al., 2001) and Emeishan (e.g., Kamenetsky et al., 2012). However, they also rarely occur in arc settings, such as in Aleutians (e.g., Nye and Reid, 1986), Solomon (e.g., Ramsay et al., 1984; Rohrbach et al., 2005), Vanuatu (e.g., Eggins, 1993), and Japan (e.g., Yamamoto, 1988; Yokoyama et al., 1992; Ninomiya and Arai, 1993; Ishiwatari and Imasaka, 2002). The petrogenesis of magnesian mafic igneous rocks is important for constraining the conditions and processes of mantle melting (e.g., Révillon et al., 1999; Herzberg et al., 2007; Clarke and Beutel, 2020). Some researchers have considered that picrites are primary magmas generated by high-degree of melting of mantle peridotite (e.g., Jaques and Green, 1980; Takahashi and Kushiro, 1983). Melt inclusions in early crystallization phases, such as olivine and Cr-spinel, have often been used to understand the petrogenesis of mafic igneous rocks because they sometimes preserve more primitive compositions than groundmass glasses (e.g., Shimizu, 1998; Schiano, 2003; Laubier et al., 2007; Sorbadere et al., 2013; Barker et al., 2020). In hypabyssal and plutonic rocks, daughter minerals crystallize from trapped melt inclusions due to slow cooling enough to form multiphase solid inclusions (MSIs). The investigation of petrological characteristics of MSIs can provide important insights into the origin of magmas and magmatic processes, such as crystal differentiation and magma mixing.

The Ogi Picritic Dolerite (OPD) belongs to the Ogi Basalt Member (Yanagisawa, 2012) which was active immediately after the opening of the Sea of Japan in the Miocene and is a picritic dolerite sill which formed at the western end of the Ogi Peninsula, southwestern Sado Island, in the Sea of Japan (Fig. 1a). There are two types of olivine crystals in the OPD: normally zoned olivine with a high-Fo [Fo = 100*Mg/(Mg + Fe²⁺) on molar basis] core and reversely zoned olivine with a low-Fo core, high-Fo mantle, and low-Fo rim (Yokoyama et al., 1992). Yokoyama et al. (1992) suggested that the low-Fo core and high-Fo mantle crystallized at low- and high-temperature conditions, respectively, of a single basaltic magma chamber where the magma was stored before the emplacement. On the other hand, Fujibayashi et al. (2016) proposed that low-Fo core and high-Fo mantle crystallized from high-Ca boninite and olivine basalt magmas, respectively, as a result of the mixing between these two magmas. These two hypotheses differ from each other in magma type involved in the OPD formation.

Figure 1. (a) Location maps of the OPD. Modified from the map of the Geospatial Information Authority of Japan. (b) A lower contact between the tuffaceous mudstone (country rock) and the OPD at the northern coast. The glassy zone, which might have been formed by the partial melting of the country rock, is present between them. The dotted green line indicates the boundary between the country rock and the OPD. (c) Close-up of the 10-50 cm thick glassy zone.

We investigated the origin of the two olivine types, focusing on the zoning profiles of olivine and MSIs in olivine in order to clarify the formation process of the OPD, and the characteristics of magmatism in the back-arc basin related to the opening of the Sea of Japan during the Miocene.

OGI PICRITIC DOLERITE SILL

Yamakawa and Chihara (1968) described the petrological characteristics of the Ogi Basalt and illustrated a geological map of the Ogi Peninsula. The Ogi Basalt Member is widely distributed on the central part of the Ogi Peninsula, and the overlying mudstone of the Tsurushi Formation is distributed on the northeastern side (Yamakawa and Chihara, 1968; Ogi Collaborative Research Group, 1986; Fujibayashi and Sakai, 2003). The Ogi Basalt Member mainly consists of Middle Miocene basaltic-andesitic volcanic rocks and volcaniclastic rocks (Fujibayashi and Sakai, 2003). The OPD intrudes into tuffaceous mudstone intercalated in the lowermost part of the Ogi Basalt Member.

Shinmura et al. (1995) provided whole-rock K-Ar ages of 14.0 ± 1.5-10.70 ± 0.43 Ma for the lavas and associated dykes of the Ogi Basalt Member and 15.5 ± 2.2 Ma for the OPD. However, they pointed out that the older age for the OPD might be due to contamination of olivine phenocrysts with excess Ar. Yanagisawa (2012) reported the diatom biochronological age of mudstones intercalated in the Ogi Basalt Member, concluding that the age of volcanic activity of the Ogi Basalt Member is 14.1-12.7 Ma.

There are two picritic dolerite sills, which are 100 m apart from each other to the east and west at Sawasaki, on the western tip of the Ogi Peninsula, Sado Island. The western region (Miko-iwa; Fig. 1a) was investigated in the present study. The OPD is exposed on the coast with a thickness of at least 20 m. The OPD is emplaced into tuffaceous mudstone, and the lower contacts are exposed at the northern and southeastern parts of the sill (Fig. 1b). The contacts with tuffaceous mudstone are gently tilted southward at the northern part but are vertical at the southern part. The directions of columnar joints also change from vertical to horizontal toward the south (Supplementary Fig. S1; Fig. S1 is available online from https://doi.org/10.2465/jmps.231002), indicating that the intrusion thrusts upward at the south. A 10-50 cm thick glassy zone (71-74 wt% whole-rock SiO₂), which was formed by the melting of tuffaceous mudstone (Fujibayashi et al., 2014), occurs between the tuffaceous mudstone and the lower chilled margin of the OPD (Fig. 1c). The glassy zone shows perlitic textures with glossy surfaces near the contact with the chilled margin, but it becomes more greenish near the tuffaceous mudstone. The upper contact of the OPD is not exposed because it is covered by vegetation. The columnar joints and ‘vesicle layering’ (Toramaru et al., 1996) perpendicular to the columnar joints can be clearly observed on the eroded surface.

PETROGRAPHY

Chilled marginal basalt and picritic basalt

The chilled marginal basalt within 30 cm of the contact shows a porphyritic texture with olivine and plagioclase phenocrysts with Cr-spinel inclusions (Figs. 2a and 2b). Euhedral olivine phenocrysts (0.7-2.2 mm in size) are almost altered. The modal abundance of phenocrystic olivine is about 10 vol%. Plagioclase phenocrysts (0.5-1.0 mm in size) are subhedral to euhedral and show tabular shapes. Cr-spinels [Cr# = Cr/(Cr + Al) on molar basis; 0.43-0.63] occur not only as small inclusions in olivine but also as phenocrysts. The Cr-spinel phenocrysts (∼ 0.3 mm in size) commonly contain crystallized melt inclusions, which are composed mainly of augite and glass with minor amounts of plagioclase, hornblende, pigeonite, and Fe-Ni sulfide. The groundmass is composed of augite (∼ 0.6 mm in size), plagioclase (∼ 0.2 mm in size), ilmenite, titaniferous magnetite, and amygdules (former vesicles). The contact between the chilled marginal basalts and the glassy zone of country rocks is sharp but commonly shows curved boundaries (Figs. 1b and 2a). The glassy zone formed by the melting of country rock is composed of resorbed phenocrysts of clinopyroxene, orthopyroxene, and plagioclase and a matrix of altered glass with a perlitic texture (Figs. 1c and 2a).

Figure 2. Photomicrographs of the OPD. (a) Boundary between the chilled marginal basalt (0806-3c) and the black glassy zone shown by green broken line taken under cross-polarized light and (b) taken under plane-polarized light. (c) Picritic basalt (0806-3d) taken under cross-polarized light and (d) taken under plane-polarized light. (e) Picritic dolerite (OGP2m). Groundmass phases (clinopyroxene, plagioclase, and amygdule) fills the interstitial spaces of phenocrystic olivine (surrounded by green lines). Taken under cross-polarized light. (f) Olivine with kink bands in picritic dolerite (XPL). Taken under cross-polarized light. Amyg, amygdule; Cpx, clinopyroxene; Cr-Spl, Cr-spinel; Ol, olivine; Pl, plagioclase.

The picritic basalts are porphyritic rocks within 4 m from the chilled margins (Figs. 2c and 2d). The mineral assemblage is the same as that of the chilled marginal basalts, but the modal abundance of the olivine phenocrysts is higher (approximately 15 vol%) than that of the chilled marginal basalts. Euhedral to subhedral olivine phenocrysts (1.3-3.8 mm in size) commonly contain Cr-spinel grains. The groundmass is composed of augite, plagioclase, ilmenite, titaniferous magnetite, and amygdules.

Picritic dolerite

Picritic dolerite is the main lithology of the OPD (Figs. 2e and 2f). The picritic dolerite contains coarse-grained phenocrystic olivines in a groundmass consisting of fine-grained olivine, clinopyroxene, plagioclase, titaniferous magnetite, and amygdules (Fig. 2e). The morphology of phenocrystic olivine (Figs. 3a and 3b) is related to the grain size; smaller (1-2 mm) grains tend to be subhedral to euhedral, but larger grains (>2 mm) vary from euhedral to anhedral. Anhedral olivine often exhibits a branched shape (Fig. 3b). The modal abundance of phenocrystic olivine shows the variation of approximately 60-70 vol%. Cr-spinel is often included in olivine but is also found as microphenocrysts (<0.5 mm) in the groundmass. Also, some phenocrystic olivine grains show kink bands irrespective of their types (Fig. 2f). Moreover, olivine and spinel sometimes contain MSIs. Clinopyroxene (0.4-2.0 mm) and plagioclase (0.1-1.0 mm) in groundmass show ophitic texture.

Figure 3. Back-scattered electron images and zoning profiles of two types of olivine in the picritic dolerite. Note the abundance of Cr-spinel inclusions surrounded by yellow circles. (a) Type-1 olivine with Cr-spinel inclusions distributed throughout the crystal. (b) Type-2 olivine, which contains less Cr-spinel inclusions in core and more Cr-spinel inclusions in mantle to rim. The branched shape of olivine is traced by green line. MSI, multiphase solid inclusion. Abbreviations for minerals are the same as in Figure 2.

ANALYTICAL METHODS

Whole-rock composition

The OPD samples for the whole-rock geochemical analyses were cleaned by removing surface contaminants, using a vibrating cup mill made of tungsten carbide, and were powdered using an agate mill. Loss on ignition (LOI) was calculated for sample powders heated in a muffle furnace at 900 °C for 2 h, after cooling in a desiccator. The powdered samples after ignition were mixed with lithium borate flux (sample powders = 0.8000 g, Li₂B₄O₇ = 3.0000 g, and LiBO₂ = 1.0000 g) in a platinum crucible, and 10 µL each of oxidizing (LiNO₃) and release agents (LiI) was added. The powdered samples were fused at 1100 °C to make glass beads using a bead sampler (Tokyo Kagaku Co., Ltd.). The fused glass beads were analyzed using a wavelength dispersive X-ray fluorescence spectrometer ZSX Primus II (Rigaku Co., Ltd.) installed at the Graduate School of International Resource Sciences, Akita University.

Mineral composition

The chemical compositions of the olivine and daughter minerals in the olivine-hosted MSIs were analyzed using an electron probe microanalyzer (EPMA; JXA-8230, JEOL) installed at the Graduate School of International Resource Science, Akita University. The analyses were conducted at 15 kV accelerating voltage, 20 nA probe current, and 2-5 µm probe diameters with counting times of 20 s on peak and 10 s on background.

To obtain the bulk composition of the MSIs, we used the raster-scan mode of a scanning electron microprobe (JSM-IT300, JEOL) equipped with an energy dispersive X-ray spectrometer (X-Max^N, Oxford Instruments) at the Graduate School of International Resource Sciences, Akita University. The analyses were conducted at 15 kV accelerating voltage and 2.2 nA probe current. The outlines of MSI were traced using the freehand tool of the AZtec software, and then the X-ray counts of the traced area were obtained by the raster-scan mode (e.g., Hoshide and Obata, 2012; Rollinson et al., 2018). The mean value and the standard deviation of the bulk MSI composition were calculated from the 10 analysis values obtained by repeating the above procedure 10 times. Olivine may have grown from the wall of melt inclusions after melt entrapment, but it is difficult to determine the width of olivine growth. Therefore, the composition of each inclusion was determined by tracing the rim of the MSIs excluding the host olivine.

RESULTS

Whole-rock compositions

The representative whole-rock compositions of the OPD are presented in Table 1. The whole-rock MgO content monotonically increases from the chilled marginal basalt to the picritic dolerite along the olivine control line (Fig. 4). The chilled marginal basalts, which show the most differentiated compositions in the sill, contain approximately 10 wt% MgO and approximately 52 wt% SiO₂. The Mg number [Mg# = MgO/(MgO + FeOt) on a molar basis] is 0.72-0.74. The MgO content gradually increases toward the inner part of the sill from the chilled marginal basalt through the picritic basalt (∼ 15 wt%) to the picritic dolerite (24.5-28.5 wt%).

Table 1. Representative whole-rock major element compositions from the Ogi Picritic Dolerite Sill

Sample:	0806-3c	0807-3a	0807-3d	OGP1m	OGP2m	OGP4m	OGP6m	OGP8m	OGP10m	OGP11m	OGP14m	OGP16m	OGP17m
Sampling site:	2	1	1	3	3	3	3	3	3	3	3	3	3
Stratigraphic level:	<3 cm	<3 cm	70 cm	1 m	2 m	4 m	6 m	8 m	10 m	11 m	14 m	16 m	17 m
Lithology:	Chilled marginal basalt	Chilled marginal basalt	Picritic basalt	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite
SiO2	51.44	52.07	49.04	42.48	44.77	45.60	45.92	45.92	44.55	45.65	45.14	45.13	46.64
TiO2	0.84	0.87	0.78	0.38	0.48	0.45	0.47	0.40	0.44	0.44	0.49	0.46	0.58
Al2O3	16.10	16.26	14.63	8.14	8.44	9.09	8.15	8.72	8.52	8.92	8.30	9.19	9.47
FeOt	7.02	6.42	8.37	10.51	10.98	10.99	11.21	10.91	10.97	9.92	10.00	9.77	9.79
MnO	0.11	0.10	0.15	0.25	0.14	0.13	0.13	0.12	0.14	0.16	0.15	0.13	0.13
MgO	10.01	10.40	14.98	27.70	28.52	26.38	27.08	25.27	28.00	27.95	28.10	26.60	24.52
CaO	11.56	10.76	9.78	9.43	5.12	5.69	5.73	5.76	5.43	5.28	5.17	5.64	6.30
Na2O	1.71	2.04	1.37	0.53	0.69	0.81	0.78	0.75	0.69	0.61	0.62	0.61	0.81
K2O	0.53	0.72	0.49	0.13	0.19	0.34	0.32	0.28	0.24	0.25	0.32	0.18	0.36
P2O5	0.16	0.18	0.15	0.09	0.11	0.11	0.10	0.09	0.10	0.09	0.11	0.10	0.13
Total	99.48	99.82	99.74	99.64	99.44	99.59	99.89	98.22	99.08	99.27	98.40	97.81	98.73
H2O(-) (wt%)	1.10	1.29	0.80	n.d.	n.d.	n.d.	n.d.	n.d.	n.d.	0.96	n.d.	n.d.	n.d.
LOI (%)	2.90	2.42	4.82	n.d.	n.d.	n.d.	n.d.	n.d.	n.d.	4.70	n.d.	n.d.	n.d.
Mg#	0.72	0.74	0.76	0.82	0.82	0.81	0.81	0.81	0.82	0.83	0.83	0.83	0.82

Total iron is calculated as FeOt (FeOt = Fe₂O₃t*0.9). LOI, Loss on ignition. Mg# = MgO/(MgO + FeOt) on molar basis. Sampling sites correspond to the numbers in Figure 1a. Stratigraphic levels indicate the detailed locations of samples at each sampling site: For the chilled marginal basalts and the picritic basalts, they indicate the distances from the contact of the country rock. For the picritic dolerites, they indicate the distances from bottom of exposed part of the OPD because the lower contact is submerged. n.d., not determined.

Figure 4. Whole-rock MgO versus SiO₂ wt% diagram of the OPD. All plotted data were recalculated to 100 wt% total. The olivine control line (Fo89) is shown by the gray dashed line.

Two types of olivine and associated spinel

Representative olivine compositions are listed in Table 2. Phenocrystic olivines can be classified into two types based on their grain sizes, zoning profiles, and inclusion abundances. Type-1 olivine is medium-grained (1-2 mm on the major axis), subhedral to euhedral crystals (Fig. 3a) with normal zoning from homogeneous core (Fo88.0-89.8, NiO = 0.25-0.30 wt%) to rim (Fo65.0-84.6, NiO = 0.01-0.23 wt%) and high-Ca content (0.21-0.26 wt%). They are also characterized by containing abundant Cr-spinel inclusions and MSIs. Type-2 olivine is coarse-grained, anhedral to euhedral crystals (>2 mm along the major axis) with reverse zoning from low-Fo, -Ni, and -Ca core (Fo83.5-85.0, NiO = 0.13-0.24 wt%, CaO = 0.13-0.20 wt%) toward high-Fo, -Ni, and -Ca mantle (Fo88.1-89.3, NiO = 0.25-0.30 wt%, CaO = 0.22-0.25 wt%). The width of the reverse-zoned part (where the Fo content increases outward) in the type-2 olivine is positively correlated with the grain size (Fig. 5). The Fo, NiO, and CaO contents decrease from the mantle to the rim (Fo65.0-72.1, NiO = 0.10-0.17 wt%, CaO = 0.16-0.21 wt%). Some type-2 olivine crystals exhibit a branched shape (Fig. 3b). Cr-spinel inclusions and MSIs are present in the type-2 olivine cores but are more abundant in the mantles and rims. The ratio of modal abundances of the type-1 and -2 olivines is about 1:3 in the picritic dolerite, but all olivine phenocrysts are classified as the type-1 in the chilled marginal basalt.

Table 2. Representative microprobe analyses of olivine

Sample:	0807-3d	0807-3d	OGP2m	OGP2m	OGP2m	OGP2m	OGP2m	OGP2m	OGP2m	OGP2m
Lithology:	Picritic basalt	Picritic basalt	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite	Picritic dolerite
Olivine type:	1	1	1	1	2	2	2	2	2	2
	Core	Rim	Core	Rim	Core	Core	Mantle	Mantle	Rim	Rim
SiO2	41.32	38.54	40.79	37.94	39.77	39.91	40.82	40.67	36.99	38.24
TiO2	0.00	0.00	0.10	0.00	0.00	0.06	0.00	0.00	0.00	0.00
Al2O3	0.01	0.01	0.02	0.00	0.02	0.00	0.04	0.02	0.03	0.00
Cr2O3	0.06	0.02	0.08	0.00	0.00	0.04	0.07	0.02	0.00	0.01
FeO	10.25	23.66	10.29	24.91	15.56	14.35	10.33	11.15	30.54	24.92
MnO	0.18	0.44	0.16	0.43	0.26	0.21	0.20	0.21	0.58	0.54
MgO	48.53	37.15	48.57	36.19	44.24	45.12	48.58	47.71	31.79	36.20
CaO	0.24	0.22	0.24	0.21	0.20	0.20	0.24	0.24	0.16	0.21
NiO	0.30	0.16	0.31	0.10	0.14	0.13	0.30	0.27	0.13	0.10
Total	100.89	100.19	100.55	99.77	100.17	100.02	100.58	100.30	100.22	100.22
Cation (O = 4)
Si	1.007	1.008	0.998	1.003	1.000	1.000	0.999	1.002	1.002	1.006
Ti	0.000	0.000	0.002	0.000	0.000	0.001	0.000	0.000	0.000	0.000
Al	0.000	0.000	0.001	0.000	0.000	0.000	0.001	0.001	0.001	0.000
Cr	0.001	0.000	0.002	0.000	0.000	0.001	0.001	0.000	0.000	0.000
Fe	0.209	0.517	0.211	0.551	0.327	0.301	0.211	0.230	0.692	0.549
Mn	0.004	0.010	0.003	0.010	0.005	0.004	0.004	0.004	0.013	0.012
Mg	1.762	1.448	1.772	1.427	1.659	1.685	1.773	1.752	1.283	1.420
Ca	0.005	0.005	0.005	0.004	0.004	0.004	0.005	0.005	0.004	0.004
Ni	0.005	0.003	0.006	0.002	0.003	0.002	0.005	0.005	0.003	0.002
Total	2.993	2.992	2.999	2.997	2.999	2.999	3.000	2.998	2.998	2.994
Fo	89.4	73.7	89.4	72.1	83.5	84.9	89.3	88.4	65.0	72.1

Fo = 100*Mg/(Mg + Fe) on molar basis.

Figure 5. Relationships between the width of the Fe-Mg reverse-zoned part and the grain size of the type-2 olivine in picritic dolerite (OGP2m). The width of the Fe-Mg reverse-zoned part represents the length of one side of the symmetric reverse-zoned part in a line analysis of a single olivine crystal (e.g., B-B′ section in Fig. 3b). Grain size is the total length of line analysis of a single olivine crystal.

Cr-spinel in the type-1 olivine has higher Al₂O₃ contents (25-29 wt%) and Mg# (0.66-0.67) and lower TiO₂ contents (0.3-0.7 wt%). On the other hand, Cr-spinel in the type-2 olivine core has lower Al₂O₃ contents (23-27 wt%) and Mg# (0.58-0.59) and higher TiO₂ contents (0.5-0.9 wt%). Cr-spinel in the type-2 olivine mantles has broad compositional ranges that partly overlap with those of Cr-spinel inclusions in the above olivines (23-26 wt% Al₂O₃, Mg# = 0.59-0.65 and 0.4-0.9 wt% TiO₂). However, Cr# (0.46-0.54) is not much different among inclusions in any types of olivine. In the Al₂O₃ versus TiO₂ diagram, all Cr-spinel compositions plot on the back-arc basin basalt (BABB) and mid-ocean ridge basalt (MORB) fields (Fig. 6a; Lenaz et al., 2000; Kamenetsky et al., 2001). In the Mg# versus Cr# diagram, they are plotted on the abyssal and arc-related peridotite fields (Fig. 6b). Additionally, the olivine Fo and Cr-spinel Cr# are plotted in the vicinity of the olivine-spinel mantle array (OSMA) field (Arai, 1994), but those of the type-2 olivine core are away from the field (Fig. 6c).

Figure 6. (a) Spinel Al₂O₃ (wt%) versus TiO₂ (wt%) diagram. Compositional fields of OIB (ocean island basalts), MORB, and IAB (island arc basalts) are from Kamenetsky et al. (2001). The field of BABB is from Lenaz et al. (2000). (b) Spinel Mg# versus Cr# diagram. Compositional fields of arc-related peridotites (Arai and Ishimaru, 2008; Arai et al., 2011), abyssal peridotites (Dick and Bullen, 1984; Arai, 1994; Arai et al., 2011), boninite and back-arc peridotite (Stern et al., 2004) are shown. (c) Olivine Fo content versus spinel Cr# diagram. OSMA is after Arai (1994).

MSIs in olivine

The backscattered electron images of the two types of MSI in olivine are shown in Figure 7. The bulk and glass compositions of the MSIs and their representative major element variations in terms of SiO₂ contents are presented in Table 3 and Figure 8, respectively. Some MSIs have MgO and SiO₂ contents as high as those of boninites, but they contain more than 0.5 wt% TiO₂ and are not classified as boninites (Le Bas, 2000). In this study, we refer to as ‘high-Mg andesite’ with >52 wt% SiO₂, >8 wt% MgO, and >0.5 wt% TiO₂ (e.g., Ohba et al., 2007). Most MSIs in the type-1 olivine (O1-MSI) are composed of clinopyroxene (Mg# = 0.65-0.81) with a minor amount of hornblende and plagioclase [An# = 100*Ca/(Ca + Na + K) on molar basis; 74-95]. Glasses and vesicles are present in various proportions in interstitial spaces among crystals (Figs. 7a-7c). In the crystalline O1-MSIs, plagioclase and silicic glasses (67-72 wt% SiO₂) are contained (Figs. 7b and 8). The glass-rich O1-MSIs (Fig. 7a) commonly consist of clinopyroxene and glass (62-63 wt% SiO₂) with minor amount of hornblende and dendritic plagioclase. Except for one MSI with high-Mg andesite composition, the bulk MgO and SiO₂ contents of most O1-MSI samples are less than 8 and 60 wt%, respectively, which are equivalent to those of ordinary basalt to andesite.

Figure 7. Back-scattered electron images of the two types of MSI in olivine. (a) Low-crystallinity O1-MSI (OGP5m Ol002 incl1). (b) Medium-crystallinity O1-MSI (OGP2m Ol005 incl1). (c) High-crystallinity O1-MSI (OGP10m Ol003 incl1). (d) Low-crystallinity O2c-MSI (OGP10m Ol004 incl1). (e) Medium-crystallinity O2c-MSI (OGP2m Ol008 incl1). (f) Low-crystallinity O2c-MSI (OGP6m Ol005 incl1). Hbl, hornblende; Opx, orthopyroxene; Ilm, ilmenite; V, void. Other abbreviations for minerals are the same as in Figure 2. Each scale bar in the lower right represents 10 µm. The L, M, and H in the upper left indicate the crystallinities of the MSI (low, medium, and high, respectively).

Table 3. Major element compositions of bulk MSI and associated glass from the OPD

Sample:	OGP2m	OGP5m	OGP10m	OGP10m	OGP10m	OGP10m	OGP10m	OGP10m	OGP10m	OGP2m	OGP2m	OGP2m	OGP2m	OGP4m	OGP6m	OGP10m
Daughter mineral assemblage:	Cpx	Hbl, Cpx, Opx	Hbl, Cpx	Hbl, Cpx, Opx	Pl, Hbl, Cpx	Hbl, Cpx, Opx	Hbl, Cpx	Pl, Hbl, Cpx	Pl, Cpx	Pl, Hbl, Cpx	Hbl, Cpx, Opx	Cpx	Pl, Hbl, Cpx	Hbl, Cpx, Opx	Pl, Hbl, Cpx, Opx	Hbl, Cpx, Opx
MSI type:	O1	O1	O1	O1	O1	O1	O1	O1	O1	O2m	O2c	O2c	O2c	O2c	O2c	O2c
Bulk MSI
SiO2	56.07	56.74	56.93	53.31	57.95	54.76	56.29	56.22	60.95	57.33	55.54	59.55	60.47	50.93	51.87	53.29
TiO2	1.06	1.49	0.81	0.66	0.66	1.24	1.29	1.25	0.99	1.03	0.41	1.22	1.15	1.06	1.01	0.79
Al2O3	18.30	17.98	19.09	14.75	20.41	16.53	18.21	19.50	17.45	21.21	14.71	17.82	15.88	16.23	14.14	16.94
FeO	3.02	2.97	2.88	5.59	2.57	4.92	4.07	3.10	2.71	1.99	6.06	2.92	3.44	7.80	6.25	6.16
MnO	0.07	0.06	0.06	0.10	0.06	0.10	0.06	0.06	0.07	0.14	0.14	0.07	0.07	0.20	0.15	0.13
MgO	3.74	4.61	5.15	12.95	3.44	7.98	5.04	4.34	3.15	2.31	12.42	2.27	5.91	10.22	12.81	9.08
CaO	12.80	8.90	10.75	9.06	9.84	10.11	9.73	10.67	8.31	11.28	5.74	10.82	6.50	9.69	9.73	9.12
Na2O	2.99	3.68	2.93	2.43	2.70	3.14	3.23	3.06	3.98	3.27	2.51	3.38	3.92	2.40	2.38	2.36
K2O	1.26	1.58	1.15	0.90	2.01	0.89	1.67	1.46	2.07	1.05	2.18	1.19	1.99	1.07	1.03	0.94
P2O5	0.69	2.01	0.25	0.26	0.35	0.32	0.42	0.33	0.32	0.39	0.28	0.78	0.67	0.39	0.62	1.19
Bulk Mg#	0.69	0.73	0.76	0.81	0.70	0.74	0.69	0.71	0.67	0.60	0.76	0.58	0.75	0.70	0.78	0.72
Bulk MSI 1σ
SiO2	0.39	0.11	0.30	0.41	0.11	0.13	0.29	0.08	0.25	0.15	0.36	0.23	2.59	0.24	0.18	0.19
TiO2	0.03	0.04	0.03	0.03	0.03	0.05	0.03	0.04	0.06	0.08	0.07	0.04	0.09	0.09	0.04	0.04
Al2O3	0.14	0.12	0.11	0.11	0.07	0.10	0.18	0.08	0.09	0.10	0.36	0.14	0.07	0.32	0.10	0.16
FeO	0.12	0.10	0.03	0.10	0.09	0.08	0.12	0.05	0.06	0.66	0.23	0.04	0.05	0.37	0.12	0.15
MnO	0.04	0.03	0.05	0.04	0.04	0.03	0.04	0.04	0.03	0.31	0.06	0.04	0.03	0.06	0.04	0.03
MgO	0.33	0.36	0.13	0.17	0.19	0.27	0.46	0.13	0.19	0.21	0.90	0.15	0.16	0.85	0.07	0.41
CaO	0.24	0.06	0.09	0.05	0.09	0.13	0.13	0.12	0.15	0.13	0.30	0.10	0.24	0.25	0.11	0.09
Na2O	0.07	0.07	0.05	0.03	0.03	0.03	0.06	0.03	0.04	0.05	0.09	0.05	0.33	0.04	0.03	0.05
K2O	0.06	0.03	0.03	0.04	0.04	0.03	0.04	0.04	0.03	0.37	0.07	0.04	0.04	0.04	0.04	0.03
P2O5	0.04	0.06	0.03	0.04	0.05	0.03	0.06	0.04	0.04	0.05	0.06	0.05	0.19	0.06	0.05	0.03
Matrix glass
SiO2	63.10	71.90	63.25	67.03	69.62	63.33	62.87	70.52	68.38	72.92	69.40	68.44	73.01	64.08	71.27	61.25
TiO2	0.36	0.18	0.01	0.12	0.17	0.17	0.19	0.23	0.30	0.47	0.07	0.60	0.17	0.06	0.07	0.07
Al2O3	22.02	15.52	18.91	17.67	14.32	19.64	19.91	14.57	15.78	14.78	18.85	16.83	15.34	18.25	16.63	20.11
FeO	0.90	0.43	0.80	0.66	0.55	1.07	1.34	0.72	0.75	0.82	1.06	1.28	0.55	0.86	0.79	1.61
MnO	0.08	0.07	0.06	0.07	0.06	0.00	0.13	0.00	0.00	0.11	0.01	0.00	0.00	0.07	0.00	0.05
MgO	0.20	0.19	0.33	0.19	0.10	0.35	0.29	0.00	0.05	0.22	0.37	0.10	0.14	0.17	0.29	0.60
CaO	7.63	0.86	5.02	2.55	1.39	3.84	4.42	1.00	1.44	1.48	2.93	3.27	1.34	3.39	2.15	5.58
Na2O	3.22	3.95	3.00	3.74	2.85	4.01	3.37	3.46	3.63	3.83	3.21	4.31	3.68	2.76	4.08	2.85
K2O	2.04	2.63	1.60	2.02	3.48	1.47	2.39	2.56	2.66	2.48	3.98	2.35	3.14	2.23	3.05	1.67
P2O5	0.20	0.28	0.25	0.37	0.42	0.48	0.66	0.24	0.34	0.02	0.00	0.33	0.08	0.75	0.69	1.59
Total	99.75	96.01	93.23	94.42	92.96	94.36	95.57	93.30	93.33	97.13	99.88	97.51	97.45	92.62	99.02	95.38

Bulk MSI data are recalculated to 100 wt% total.

Mg# = MgO/(MgO + FeO) on a molar basis. The bulk MSI composition is shown as the mean value and standard deviation (1σ) of 10 analyses. See text for details.

Figure 8. Representative oxides versus SiO₂ variation diagrams for the bulk and glass compositions of the MSI. All plotted data were recalculated to 100 wt% total. The compositional field for picrite is after Le Bas (2000). See text for the compositional definition of high-Mg andesite.

Most MSIs in the type-2 olivine cores (O2c-MSI) contain clinopyroxene (Mg# = 0.71-0.81), hornblende, and orthopyroxene (Mg# = 0.77-0.84) as daughter minerals, with interstitial glasses and a substantial amount of bubble (Fig. 7e). Both O1- and O2c-MSIs contain Al-spinel, Fe-Ti oxide, apatite, and Fe-Ni sulfide as accessory minerals. Al-spinel (>60 wt% Al₂O₃) commonly occurs between glass and Cr-spinel grain or as isolated grains (Fig. 7d). Plagioclase occurs rarely in the O2c-MSIs (Figs. 7d-7f) and is found as a daughter mineral only in some highly crystalline MSIs. The bulk compositions of the O2c-MSIs are, on average, lower in CaO content and higher in MgO and FeOt contents than those of the O1-MSIs (Fig. 8), some of which are equivalent to the compositions of picrite to high-Mg andesite. However, the ranges of the bulk compositions of the O1- and O2c-MSIs entirely overlap each other, and there is no significant difference between them.

DISCUSSION

Compositional difference between the chilled margins and interior of the sill

The whole-rock MgO content of OPD varies proportionally with the modal abundance of olivine. If the whole-rock MgO content and olivine mode of the chilled margin represent those of the magma at the time of intrusion and the variation of whole-rick MgO content results only from simple crystal settling of olivine, the average MgO content of the entire sill should be consistent with that of the chilled margin (Gibb and Henderson, 1992). However, since the interior of the OPD (picritic basalt and picritic dolerite) at any stratigraphic level is richer in whole-rock MgO content than the chilled marginal basalt (Fig. 4), it is clear that the average MgO content of the entire sill is higher than that of the chilled margin. This characteristic cannot be explained by gravitational settling of olivine in a magma with the same composition as the chilled margin.

There are two possible mechanisms for the increase in modal olivine and whole-rock MgO content in the interior of the sill compared to that in the chilled margin. One is the injection of another olivine-richer magma after the intrusion of magma with the composition of the chilled marginal basalt. As no internal chilled margin (e.g., Henderson and Gibb, 1987) is observed inside the OPD, the following magma pulse may have intruded soon after the intrusion of the preceding magma. For example, one of possible scenarios may be sequential injections of magmas ejected from a magma chamber with a gradient in olivine modal composition.

The other possible mechanism is flow differentiation due to the Bagnold effect (Bagnold, 1954) during the intrusion of sills (Bhattacharji and Smith, 1964; Simkin, 1967; Gibb, 1968). Flow differentiation can occur even in low-viscosity fluid, such as basaltic melt (Bhattacharji and Smith, 1964; Bhattacharji, 1967). Barriére (1976) argued that the Bagnold effect, which is a major effect in flow differentiation, is operative in the case of large dykes and pipes with a width or a diameter less than 100 m. Therefore, flow differentiation could have occurred during the intrusion of the 20 m-thick OPD sill. The flow differentiation may have sorted the coarse-grained type-2 olivine from the margin toward the interior of the sill, leaving only the fine-grained type-1 olivine at the chilled margin.

Origin of two types of olivine

The type-1 olivine and the type-2 olivine mantle in the OPD have high-Fo (88.0-89.8 and 88.1-89.3, respectively) and -Ca (CaO = 0.21-0.26 and 0.22-0.25 wt%, respectively) composition (Table 2 and Fig. 3). As mantle olivine is generally characterized by low-Ca contents (CaO < 0.10 wt%) (e.g., Thompson and Gibson, 2000; Lustrino et al., 2019), both types of olivine in the OPD are not derived from mantle peridotites. The Fo content of the type-2 olivine cores are much lower (Fo83.5-85.0) than typical mantle olivine (e.g., Arai, 1994). The Ca content of the type-2 olivine core (CaO = 0.13-0.20 wt%) is also higher compared to mantle-derived olivine but is slightly lower than other magmatic olivines. The value is comparable to those of olivines crystallized from boninite (∼ 0.2 wt% CaO; Kamenetsky et al., 2006) or H₂O-rich magma (∼ 0.07 wt% CaO; Altunkaynak et al., 2019).

Chilled margins may preserve the magmatic composition at the time of emplacement (e.g., Hoshide et al., 2006). Therefore, we estimated liquidus olivine composition in chemical equilibrium with chilled marginal basalt using Fe-Mg partition coefficient (assuming K_D = 0.30 ± 0.02; Roeder and Emslie, 1970) and Ca partition coefficient ($\text{LnD}{*}_{\text{CaO}}^{\text{Olivine-Melt}} = -1.24X_{\text{Fo}}^{3} + 3.33X_{\text{Fo}}^{2} - 6.55X_{\text{Fo}} + 2.05$; Libourel, 1999) between olivine and melt. We assumed that Fe³⁺/ΣFe ratio is 0.15 for parent melt (Cottrell and Kelley, 2011). Estimated Fo and minimum CaO contents of olivine were Fo90.3-91.4 and 0.19-0.23 wt% CaO, respectively. These values are only slightly different from the compositional ranges of the type-1 olivine and the type-2 olivine mantle. In addition, the chilled marginal basalt contains only the Cr-spinel inclusion-rich euhedral olivine pseudomorphs. Although the original chemical composition is unknown because it is completely altered to chlorite, the other characters are similar to those of the type-1 olivine from the picritic dolerite. Thus, type-1 olivine seems to be ‘autocryst’ that crystallized from a basaltic magma comparable to the chilled marginal basalt in terms of chemical composition. The chemical and petrographic characteristics of the type-2 olivine mantle are similar to those of type-1 olivine. Therefore, the type-2 olivine mantle must have been cogenetic with the type-1 olivine crystals.

On the other hand, the type-2 olivine core (Fo83.5-85.0, CaO = 0.13-0.20 wt%) is clearly in disequilibrium with chilled marginal basalt. Yokoyama et al. (1992) advocated that reversely zoned olivine phenocrysts in the OPD crystallized at the low-temperature condition in a magma reservoir and were subsequently brought to the high-temperature condition, resulting in the reverse zoning of Fe and Mg. In the type-2 olivine, low-Fo parts crystallized at low-temperatures are found in both the core and rim (Fig. 3b). However, there are differences in the amounts of Cr-spinel inclusion and CaO content between the type-2 olivine core (poor in Cr-spinel inclusion and CaO content) and rim (rich in them). These observations cannot be explained by crystallization at different temperatures within a single magma reservoir proposed by Yokoyama et al. (1992).

Gavrilenko et al. (2016) proposed an olivine geohygrometer based on the water content dependence of Ca partitioning between olivine and melt. By applying the whole-rock composition of the chilled marginal basalt (samples 0806-3c and 0807-3a; Table 1) and the average CaO content of type-1 olivine (= 0.24 wt%) to the hygrometer, the water content of the magma was estimated to be 0.9-1.5 wt% (Fig. 9). Calculations using the most primitive bulk composition of the O1-MSIs (SiO₂ = 53.3 wt%; Table 3) and the average CaO content of the type-1 olivine, the water content was estimated to be 0.5 wt%. Though the CaO content of the type-2 olivine core varies between 0.13-0.20 wt% among grains, by applying the more primitive bulk compositions of the O2c-MSI (SiO₂ = 50.9-53.3 wt%) and the minimum CaO content of the type-2 olivine core (CaO = 0.13 wt%), the estimated water contents reached 3.6-4.5 wt%. In addition, comparing MSIs with similar crystallinities, the O2c-MSIs have higher modal hornblendes than the O1-MSIs (Fig. 7). These results suggest that the magma that crystallized the type-2 olivine core was more hydrous than the magma that crystallized the type-1 olivine.

Figure 9. Bulk MgO contents versus calculated $\text{D}_{\text{CaO}}^{\text{Olivine-Liquid}}$ diagram for the MSIs and chilled marginal basalts. The prediction lines of H₂O content are after Gavrilenko et al. (2016).

The O2c-MSIs differ from the O1-MSIs in the assemblage of daughter minerals and the bulk composition. The O2c-MSIs commonly contain orthopyroxene and hornblende, whereas the O1-MSIs lack orthopyroxene and contain plagioclase formed at the early stage of crystallization (coexisting with glass with 63 wt% SiO₂; Fig. 8). The decreasing trend of the Al₂O₃ content with increasing SiO₂ in the glass of the O1-MSIs could be due to the crystallization of plagioclase and minor hornblende. In contrast, in the O2c-MSIs, hornblende crystallized at the early stage and a small amount of plagioclase crystallized only at the late stage of crystallization, when the SiO₂ content of the glass reached 75 wt% (Fig. 8).

The mineral assemblage and bulk composition of the O2c-MSIs and the estimated high-water content of magma are similar to the characters of boninitic magma (Figs. 8 and 9) (>4.7 wt% H₂O: Umino, 1986; Umino and Kushiro, 1989; Dobson et al., 1995). Therefore, we conclude that the type-2 olivine core is ‘antecryst’, which crystallized from another magma close to high-Mg andesite. From the H₂O-saturation pressure of basaltic melts (Moore, 2008) and the magmatic H₂O contents obtained from the CaO contents of the type-2 olivine core, we can estimate that the formation pressure of the type-2 olivine core was higher than 130-190 MPa (Fig. 10). After the formation of the type-2 olivine core, it was entrapped in the basaltic magma in equilibrium with high-Fo olivine (i.e., type-1 olivine and type-2 olivine mantle) and was overgrown with high-Fo olivine (Fig. 3b). The two types of olivine phenocrysts in the OPD are probably explained by the assimilation of an olivine basaltic magma with unconsolidated cumulates that formed by a high-Mg andesitic magma, as proposed by Fujibayashi et al. (2016).

Figure 10. Relationships between dissolved H₂O concentrations and pressure in pure H₂O-saturated basaltic melt at 1200 °C (modified after Moore, 2008). The basalt H₂O data at 1200 °C is after Hamilton et al. (1964) and Dixon et al. (1995).

Interpretation of reverse zoning in the type-2 olivine

We verified the effects of diffusion and crystal growth for the Fe-Mg reverse zoning profile of the type-2 olivine to consider the formation process and environment of the type-2 olivine. If the Fe-Mg reverse zoning was caused by diffusion, the timescale of Fe-Mg diffusion between the type-2 olivine core and mantle can be estimated using the following method. The crystallization temperatures of the type-2 olivine mantle are estimated at 1250-1210 °C from the bulk compositions of the O1-MSI using crystallization simulation in MELTS (Gualda and Ghiorso, 2015). We obtained the Fe-Mg diffusion coefficients D = 8.36 × 10⁻¹⁷ (m²/s) (along [001]) and 1.39 × 10⁻¹⁷ (m²/s) (along [100] and [010]) at 1250 °C and D = 5.16 × 10⁻¹⁷ (m²/s) (along [001]) and 8.60 × 10⁻¹⁸ (m²/s) (along [100] and [010]) at 1210 °C using a global equation by Dohmen and Chakraborty (2007).

Generally, the solution to the diffusion equation for elemental diffusion between two substances is expressed as follows:

  

\begin{equation*} C\ (x, t) = \frac{C_{0}}{2}\left[1 + \text{erf}\left(\frac{x}{2\sqrt{Dt}}\right) \right] \end{equation*}

where C (x, t) is the concentration at distance x (meter) from the surface at time t (second), and C₀ is the initial concentration.

Assuming the initial compositional profile prior to the diffusion, the above equation provides the Fe-Mg diffusion profile in olivine at a given time. The reversed zoning profile of the longest width (= 1.32 mm; Fig. 11) best fits with the diffusion profiles of ∼ 45-270 years at 1250 °C or ∼ 75-440 years at 1210 °C. On the other hand, the reversed zoning profile of the shortest width (= 0.3 mm; Fig. 5) best fits with the diffusion profiles of ∼ 1.6-10 years at 1250 °C or ∼ 3-16 years at 1210 °C. As the diffusion time obtained from the longer reverse-zoned part is two orders of magnitude greater than the time for the cooling and solidification of the 20 m-thick basaltic sill (about 2-3 years; Holness, 2014), the diffusion might have occurred before intruding as the sill into the shallow part. However, the obtained diffusion time is two to three orders of magnitude greater than those obtained from the compositional zoning of olivine in most previous studies (e.g., Costa et al., 2020). It is highly doubtful that the magma can hold temperatures high at a shallow part for such prolonged periods. It is also difficult to explain the large variation in diffusion time for each type-2 olivine grain.

Figure 11. Fe-Mg diffusion modeling in the type-2 olivine (OGP2m Ol008) of Figure 3b. The yellow line represents the best fit for diffusion curves, whereas the black dashed line shows the assumed initial compositional profile prior to the diffusion. See text for details.

Rather, the correlation between the grain size and the width of the reverse-zoned part suggests that the reverse-zoned part (outward increase of Fo content) has formed during crystal growth. When the low-Fo olivine antecryst (the type-2 olivine core) is incorporated into a Mg-rich melt that crystallizes high-Fo olivine, the Mg-rich melt should mix with the melt surrounding the low-Fo olivine. The mixing may have made the melt around the low-Fo olivine progressively Mg-richer in composition, resulting in the growth of high-Fo mantle.

The low-Fo olivine that crystallized from a high-Mg andesitic magma was incorporated by the basaltic magma at the deeper part (>130-190 MPa; Fig. 12a), and then, the type-1 olivine might have been formed at the shallow depth around 20 MPa (Fig. 12b). The overgrowth of the type-1 olivine around the type-2 olivine may have been succeeded by the exsolution of H₂O during magma ascent toward the shallower part.

Figure 12. A schematic model for the formation processes of the OPD. (a) Basaltic (forming the type-1 olivine and type-2 olivine mantle) and high-Mg andesitic (forming the type-2 olivine at pressures >130-190 MPa) magmas were generated by the partial melting of the back-arc mantle beneath the Sea of Japan. (b) Ascending basaltic magma captured the type-2 olivine core. The type-1 olivine and the type-2 olivine mantle crystallized from the basaltic magma at a shallower depth, and finally emplaced as the OPD sill.

Geotectonic implications of the OPD

Most high-Mg andesites in southwest Japan occur in the fore-arc region in the Miocene (Tatsumi and Ishizaka, 1982; Shimoda et al., 1998). However, in southwest Japan and Hokkaido, some high-Mg andesites occur in the back-arc region (e.g., Lopéz and Ishiwatari, 2002; Sato et al., 2014; Yamada et al., 2023). The Middle Miocene basalts in the back-arc region of northeastern Japan including the OPD and Ogi basalts show depleted SrI and NdI isotopic compositions (Shuto et al., 2006; Okamura et al., 2016). This suggests that the Middle Miocene back-arc basalts were generated by the melting of upwelling-depleted asthenospheric mantle during the opening of the Japan Sea. The chilled marginal basalts of the OPD were generated from such depleted basaltic magmas (Fig. 12a).

Ninomiya et al. (2007) found the following two types of highly depleted mantle peridotite xenoliths from the Takeshima Basin in the Japan Sea, which are similar to abyssal peridotite (type-1) and continental lithospheric peridotite (type-2), respectively. Ichiyama et al. (2016) also reported that peridotite xenoliths derived from the mantle beneath the Japan Basin have depleted geochemical signatures. Ninomiya et al. (2007) and Morishita et al. (2020) also suggested that the infiltration of LREE-rich melts or fluids affect the melting of the source mantle beneath the Japan Sea back-arc basin based on the geochemical characteristics of the type-1 peridotite and back-arc basin basalts. Furthermore, the peridotite xenoliths from the Japan Basin show the track of the infiltration of silica-rich melts (Ichiyama et al., 2017). The signatures of high-Mg andesitic magmatism in the type-2 olivine and O2c-MSI of the OPD suggest that the highly depleted oceanic mantle infiltrated by slab-derived melts/fluids such as the type-1 peridotite may have existed ubiquitously beneath the back-arc basin and was partially melted by heat from the upwelling asthenosphere during the Miocene back-arc opening of the Japan Sea.

CONCLUSIONS

The Ogi Picritic Dolerite Sill (OPD) contains two types of olivine. Type-1 olivine has high Fo and CaO contents and exhibits normal zoning. Type-2 olivine has a low-Fo and low-Ca core and a high-Fo and high-Ca mantle and exhibits reverse zoning.

Both types of olivine have multiphase solid inclusions (MSIs), which can be also classified into two types. O1-MSIs included in the type-1 olivine mainly consists of clinopyroxene, minor hornblende, and plagioclase; the bulk compositions are basaltic to andesitic. O2c-MSIs included in the type-2 olivine core mainly consists of clinopyroxene, orthopyroxene, and hornblende; the bulk compositions are picritic to high-Mg andesitic. The water content estimated from olivine and bulk-MSI compositions suggests that the type-2 olivine core may have crystallized from a more hydrous magma than the type-1 olivine, indicating that the type-2 olivine core crystallized at higher pressures than the type-1 olivine.

Therefore, the type-1 olivine is an autocryst crystallized from primitive basaltic magma such as chilled marginal basalt and the type-2 olivine core is an antecryst crystallized from high-Mg andesitic magma. After the type-2 olivine core crystallized at the deeper part (>130-190 MPa), it was then incorporated into primitive basaltic magma that crystallized the type-1 olivine. Subsequently, the high-Fo and high-Ca olivine grew around the type-2 olivine core during the crystallization from the primitive magma at a shallower depth.

Miocene high-Mg andesitic magmatism inferred from the type-2 olivine and O2c-MSIs in the OPD might have been derived from highly depleted oceanic mantle infiltrated by slab-derived melts/fluids which was probably distributed beneath the Japan Sea back-arc basin.

ACKNOWLEDGMENTS

This work is part of the bachelor’s theses of the first (AC) and third (ST) authors, and the master’s thesis of the first author undertaken at Akita University. This work was partly supported by JSPS KAKENHI Grant Number JP22K03750 (TH) and Akita University SPRING Program (AC). We would like to thank Prof. Tsukasa Ohba and Dr. Osamu Nishikawa for the fruitful discussions. We are grateful to thank Mr. Mitsuhisa Aida (Sado Island Geopark Promotion Office) for his kind assistance with fieldwork and rock sample collection procedures. We would like to thank Editage (www.editage.jp) for English language editing. We also thank two anonymous reviewers for their very thorough and thoughtful reviews and Dr. Yuji Ichiyama for his constructive suggestions and editorial handling.

SUPPLEMENTARY MATERIAL

Supplementary Figure S1 is available online from https://doi.org/10.2465/jmps.231002.

REFERENCES

•  Albaréde,  F. and  Tamagnan,  V. (1988) Modelling the recent geochemical evolution of the Piton de la Fournaise Volcano, Réunion Island. Journal of Petrology, 29, 997-1030.
•  Altunkaynak,  Ş.,  Ünal,  A.,  Howarth,  G.H.,  Aldanmaz,  E. and  Nývlt,  D. (2019) The origin of low-Ca olivine from ultramafic xenoliths and host basaltic lavas in a back-arc setting, James Ross Island, Antarctic Peninsula. Lithos, 342, 276-287.
•  Arai,  S. (1994) Characterization of spinel peridotites by olivine-spinel compositional relationships: Review and interpretation. Chemical Geology, 113, 191-204.
•  Arai,  S. and  Ishimaru,  S. (2008) Insights into petrological characteristics of the lithosphere of mantle wedge beneath arcs through peridotite xenoliths: a review. Journal of Petrology, 49, 665-695.
•  Arai,  S.,  Okamura,  H.,  Kadoshima,  K.,  Tanaka,  C., et al (2011) Chemical characteristics of chromian spinel in plutonic rocks: Implications for deep magma processes and discrimination of tectonic setting. Island Arc, 20, 125-137.
•  Bagnold,  R.A. (1954) Experiments on a gravity-free dispersion of large solid spheres in a Newtonian fluid under shear. Proceedings of the Royal Society of London, Series A, Mathematical and Physical Sciences, 225, 49-63.
•  Barriére (1976) Flowage differentiation: limitation of the “Bagnold effect” to the narrow intrusions. Contributions to Mineralogy and Petrology, 55, 139-145.
•  Barker,  S.J.,  Rowe,  M.C.,  Wilson,  C.J.N.,  Gamble,  J.A., et al (2020) What lies beneath? Reconstructing the primitive magmas fueling voluminous silicic volcanism using olivine-hosted melt inclusions. Geology, 48, 504-508.
•  Bhattacharji,  S. (1967) Mechanics of flow differentiation in ultramafic and mafic sills. The Journal of Geology, 75, 101-112.
•  Bhattacharji,  S. and  Smith,  C.H. (1964) Flowage differentiation. Science, 145, 150-153.
•  Clarke,  D.B. and  Beutel,  E.K. (2020) Davis Strait Paleocene picrites: Products of a plume or plates? Earth-Science Reviews, 206, 102770.
•  Costa,  F.,  Shea,  T. and  Ubide,  T. (2020) Diffusion chronometry and the timescales of magmatic processes. Nature Reviews Earth & Environment, 1, 201-214.
•  Cottrell,  E. and  Kelley,  K.A. (2011) The oxidation state of Fe in MORB glasses and the oxygen fugacity of the upper mantle. Earth and Planetary Science Letters, 305, 270-282.
•  Dick,  H.J.B. and  Bullen,  T. (1984) Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 54-76.
•  Dixon,  J.E.,  Stolper,  E.M. and  Holloway,  J.R. (1995) An experimental study of water and carbon dioxide solubilities in mid-ocean ridge basaltic liquids. Part I: Calibration and solubility models. Journal of Petrology, 36, 1607-1631.
•  Dobson,  P.F.,  Skogby,  H. and  Rossman,  G.R. (1995) Water in boninite glass and coexisting orthopyroxene: concentration and partitioning. Contributions to Mineralogy and Petrology, 118, 414-419.
•  Dohmen,  R. and  Chakraborty,  S. (2007) Fe-Mg diffusion in olivine II: point defect chemistry, change of diffusion mechanisms and a model for calculation of diffusion coefficients in natural olivine. Physics and Chemistry of Minerals, 34, 409-430.
•  Eggins,  S.M. (1993) Origin and differentiation of picritic arc magmas, Ambae (Aoba), Vanuatu. Contributions to Mineralogy and Petrology, 114, 79-100.
•  Fujibayashi,  N. and  Sakai,  U. (2003) Vesiculation and eruption processes of submarine effusive and explosive rocks from the Middle Miocene Ogi Basalt, Sado Island, Japan. Washington DC American Geophysical Union Geophysical Monograph Series, 140, 259-272.
•  Fujibayashi,  N.,  Asakura,  K.,  Hattori,  T. and  Allen,  S. (2014) Pillow lava and spasmodic submarine fire fountaining in the middle Miocene marginal basin, Sado Island, Japan. Island Arc, 23, 344-364.
• Fujibayashi, N., Takaoka, M. and Sasaki, Y. (2016) Compositional reverse zoning displayed by olivine phenocrysts in picritic dolerite from the Ogi Basalt, and its possible interpretation. 123rd Annual Meeting of Geological Society of Japan, Abstract, 44 (in Japanese).
•  Gavrilenko,  M.,  Herzberg,  C.,  Vidito,  C.,  Carr,  M.J., et al (2016) A calcium-in-olivine geohygrometer and its application to subduction zone magmatism. Journal of Petrology, 57, 1811-1832.
•  Gibb,  F.G. (1968) Flow differentiation in the xenolithic ultrabasic dykes of the Cuillins and the Strathaird Peninsula, Isle of Skye, Scotland. Journal of Petrology, 9, 411-443.
•  Gibb,  F.G. and  Henderson,  C.M.B. (1992) Convection and crystal settling in sills. Contributions to Mineralogy and Petrology, 109, 538-545.
•  Gualda,  G.A. and  Ghiorso,  M.S. (2015) MELTS_Excel: A Microsoft Excel-based MELTS interface for research and teaching of magma properties and evolution. Geochemistry, Geophysics, Geosystems, 16, 315-324.
•  Hamilton,  D.L.,  Burnham,  C.W. and  Osborn,  E.F. (1964) Solubility of water and effects of oxygen fugacity and water content on crystallization in mafic magmas. Journal of Petrology, 5, 21-39.
•  Henderson,  C.M.B. and  Gibb,  F.G.F. (1987) The petrology of the Lugar sill, SW Scotland. Earth and Environmental Science Transactions of The Royal Society of Edinburgh, 77, 325-347.
•  Herzberg,  C.,  Asimow,  P.D.,  Arndt,  N.,  Niu,  Y., et al (2007) Temperatures in ambient mantle and plumes: Constraints from basalts, picrites, and komatiites. Geochemistry, Geophysics, Geosystems, 8, Q02006.
•  Holness,  M.B. (2014) The effect of crystallization time on plagioclase grain shape in dolerites. Contributions to Mineralogy and Petrology, 168, 1076.
•  Hoshide,  T.,  Obata,  M. and  Akatsuka,  T. (2006) Crystal settling and crystal growth of olivine in magmatic differentiation - the Murotomisaki Gabbroic Complex, Shikoku, Japan. Journal of Mineralogical and Petrological Sciences, 101, 223-239.
•  Hoshide,  T. and  Obata,  M. (2012) Amphibole-bearing multiphase solid inclusions in olivine and plagioclase from a layered gabbro: Origin of the trapped melts. Journal of Petrology, 53, 419-449.
•  Ichiyama,  Y.,  Morishita,  T.,  Tamura,  A. and  Arai,  S. (2016) Peridotite xenoliths from the Shiribeshi Seamount, Japan Sea: insights into mantle processes in a back-arc basin. Contributions to Mineralogy and Petrology, 171, 86.
•  Ichiyama,  Y.,  Morishita,  T.,  Tamura,  A. and  Arai,  S. (2017) Adakite metasomatism in a back-arc mantle peridotite xenolith from the Japan Sea. American Mineralogist, 102, 341-346.
•  Ishiwatari,  A. and  Imasaka,  M. (2002) Picritic basalt from the Miocene Yoka Formation in the Tango Peninsula, Kyoto Prefecture, southwestern Japan. The Journal of the Geological Society of Japan, 108, 671-684 (in Japanese with English abstract).
•  Jaques,  A.L. and  Green,  D.H. (1980) Anhydrous melting of peridotite at 0-15 kb pressure and the genesis of tholeiitic basalts. Contributions to Mineralogy and Petrology, 73, 287-310.
•  Kamenetsky,  V.S.,  Crawford,  A.J. and  Meffre,  S. (2001) Factors controlling chemistry of magmatic spinel: an empirical study of associated olivine, Cr-spinel and melt inclusions from primitive rocks. Journal of Petrology, 42, 655-671.
•  Kamenetsky,  V.S.,  Elburg,  M.,  Arculus,  R. and  Thomas,  R. (2006) Magmatic origin of low-Ca olivine in subduction-related magmas: Co-existence of contrasting magmas. Chemical Geology, 233, 346-357.
•  Kamenetsky,  V.S.,  Chung,  S.L.,  Kamenetsky,  M.B. and  Kuzmin,  D.V. (2012) Picrites from the Emeishan Large Igneous Province, SW China: a compositional continuum in primitive magmas and their respective mantle sources. Journal of Petrology, 53, 2095-2113.
•  Laubier,  M.,  Schiano,  P.,  Doucelance,  R.,  Ottolini,  L. and  Laporte,  D. (2007) Olivine-hosted melt inclusions and melting processes beneath the FAMOUS zone (Mid-Atlantic Ridge). Chemical Geology, 240, 129-150.
•  Le Bas,  M.J. (2000) IUGS reclassification of the high-Mg and picritic volcanic rocks. Journal of Petrology, 41, 1467-1470.
•  Lenaz,  D.,  Kamenetsky,  V.S.,  Crawford,  A.J. and  Princivalle,  F. (2000) Melt inclusion in detrital spinel from the SE Alps (Italy-Slovenia): a new approach to provenance studies of sedimentary basins. Contributions to Mineralogy and Petrology, 139, 748-758.
•  Libourel,  G. (1999) Systematics of calcium partitioning between olivine and silicate melt: implications for melt structure and calcium content of magmatic olivines. Contributions to Mineralogy and Petrology, 136, 63-80.
•  Lopéz,  J.C. and  Ishiwatari,  A. (2002) Petrogenesis of the tholeiitic basalt, calc-alkaline basaltic andesite and high magnesian andesite lava succession of the Oligo-Miocene Anamizu Formation in northeastern Noto Peninsula, central Japan. Journal of Mineralogical and Petrological Sciences, 97, 85-113.
•  Lustrino,  M.,  Luciani,  N. and  Stagno,  V. (2019) Fuzzy petrology in the origin of carbonatitic/pseudocarbonatitic Ca-rich ultrabasic magma at Polino (central Italy). Scientific Reports, 9, 9212.
•  Moore,  G. (2008) Interpreting H₂O and CO₂ contents in melt inclusions: constraints from solubility experiments and modeling. Reviews in Mineralogy and Geochemistry, 69, 333-362.
•  Morishita,  T.,  Hirano,  N.,  Sumino,  H.,  Sato,  H., et al (2020) Alkali basalt from the Seifu Seamount in the Sea of Japan: post-spreading magmatism in a back-arc setting. Solid Earth, 11, 23-36.
•  Ninomiya,  A. and  Arai,  S. (1993) Textural characteristics of olivine crystals in the Oshima-Ōshima picrite basalt and their genetical implications. The Science Reports of the Kanazawa University, 38, 25-49.
•  Ninomiya,  C.,  Arai,  S. and  Ishii,  T. (2007) Peridotite xenoliths from the Takeshima seamount, Japan: an insight into the upper mantle beneath the Sea of Japan. Japanese Magazine of Mineralogical and Petrological Sciences, 36, 1-14 (in Japanese).
•  Nye,  C.J. and  Reid,  M.R. (1986) Geochemistry of primary and least fractionated lavas from Okmok volcano, central Aleutians: implications for arc magmagenesis. Journal of Geophysical Research: Solid Earth, 91, 10271-10287.
• Ogi Collaborative Research Group (1986) Late Cenozoic Group in the southern part of Kosado Mountains, Niigata Prefecture. Earth Science (Chikyu Kagaku), 40, 417-436.
•  Ohba,  T.,  Kimura,  Y. and  Fujimaki,  H. (2007) High-magnesian andesite produced by two-stage magma mixing: A case study from Hachimantai, northern Honshu, Japan. Journal of Petrology, 48, 627-645.
•  Okamura,  S.,  Inaba,  M.,  Adachi,  Y. and  Shinjo,  R. (2016) Miocene-Pliocene mantle depletion event in the northern Fossa Magna, western NE Japan. Journal of Geodynamics, 97, 42-61.
•  Ramsay,  W.R.H.,  Crawford,  A.J. and  Foden,  J.D. (1984) Field setting, mineralogy, chemistry, and genesis of arc picrites, New Georgia, Solomon Islands. Contributions to Mineralogy and Petrology, 88, 386-402.
•  Révillon,  S.,  Arndt,  N.T.,  Hallot,  E.,  Kerr,  A.C. and  Tarney,  J. (1999) Petrogenesis of picrites from the Caribbean Plateau and the North Atlantic magmatic province. Lithos, 49, 1-21.
•  Roeder,  J.P. and  Emslie,  R.F. (1970) Olivine-liquid equilibrium. Contributions to Mineralogy and Petrology, 29, 275-289.
•  Rohrbach,  A.,  Schuth,  S.,  Ballhaus,  C.,  Münker,  C., et al (2005) Petrological constraints on the origin of arc picrites, New Georgia Group, Solomon Islands. Contributions to Mineralogy and Petrology, 149, 685-698.
•  Rollinson,  H.,  Mameri,  L. and  Barry,  T. (2018) Polymineralic inclusions in mantle chromitites from the Oman ophiolite indicate a highly magnesian parental melt. Lithos, 310, 381-391.
•  Sato,  M.,  Shuto,  K.,  Nohara-Imanaka,  R.,  Takazawa,  E., et al (2014) Repeated magmatism at 34 Ma and 23-20 Ma producing high magnesian adakitic andesites and transitional basalts on southern Okushiri Island, NE Japan arc. Lithos, 205, 60-83.
•  Schiano,  P. (2003) Primitive mantle magmas recorded as silicate melt inclusions in igneous minerals. Earth-Science Reviews, 63, 121-144.
•  Shimizu,  N. (1998) The geochemistry of olivine-hosted melt inclusions in a FAMOUS basalt ALV519-4-1. Physics of the Earth and Planetary Interiors, 107, 183-201.
•  Shimoda,  G.,  Tatsumi,  Y.,  Nohda,  S.,  Ishizaka,  K. and  Jahn,  B.M. (1998) Setouchi high-Mg andesites revisited: geochemical evidence for melting of subducting sediments. Earth and Planetary Science Letters, 160, 479-492.
•  Shinmura,  T.,  Kobayashi,  Y.,  Arakawa,  Y. and  Itaya,  T. (1995) K-Ar ages of Neogene basaltic rocks in the Ogi Peninsula, Sado Island. Journal of Mineralogy, Petrology, and Economic Geology, 90, 403-409 (in Japanese with English abstract).
•  Shuto,  K.,  Ishimoto,  H.,  Hirahara,  Y.,  Sato,  M., et al (2006) Geochemical secular variation of magma source during Early to Middle Miocene time in the Niigata area, NE Japan: Asthenospheric mantle upwelling during back-arc basin opening. Lithos, 86, 1-33.
• Simkin, T. (1967) Flow differentiation in the picritic sills of north Skye. In Ultramafic and Related Rocks (Wyllie, P.J. Ed.). John Wiley and Sons, Inc., New York, 64-69.
•  Sorbadere,  F.,  Schiano,  P. and  Métrich,  N. (2013) Constraints on the origin of nepheline-normative primitive magmas in island arcs inferred from olivine-hosted melt inclusion compositions. Journal of Petrology, 54, 215-233.
• Stern, R.J., Johnson, P.R., Kröner, A. and Yibas, B. (2004) Neoproterozoic ophiolites of the Arabian-Nubian Shield. In Precambrian Ophiolites and Related Rocks Developments in Precambrian Geology vol. 13 (Kusky, T.M. Ed.). Elsevier, Amsterdam, 95-128.
•  Takahashi,  E. and  Kushiro,  I. (1983) Melting of a dry peridotite at high pressures and basalt magma genesis. American Mineralogist, 68, 859-879.
•  Tatsumi,  Y. and  Ishizaka,  K. (1982) Origin of high-magnesian andesites in the Setouchi volcanic belt, southwest Japan, I. Petrographical and chemical characteristics. Earth and Planetary Science Letters, 60, 293-304.
•  Thompson,  R.N. and  Gibson,  S.A. (2000) Transient high temperatures in mantle plume heads inferred from magnesian olivines in Phanerozoic picrites. Nature, 407, 502-506.
•  Thompson,  R.N.,  Gibson,  S.A.,  Dickin,  A.P. and  Smith,  P.M. (2001) Early Cretaceous basalt and picrite dykes of the southern Etendeka region, NW Namibia: windows into the role of the Tristan mantle plume in Paraná-Etendeka magmatism. Journal of Petrology, 42, 2049-2081.
•  Toramaru,  A.,  Ishiwatari,  A.,  Matsuzawa,  M.,  Nakamura,  M. and  Arai,  S. (1996) Vesicle layering in solidified intrusive magma bodies: a newly recognized type of igneous structure. Bulletin of Volcanology, 58, 393-400.
•  Umino,  S. (1986) Magma mixing in boninite sequence of Chichijima, Bonin Islands. Journal of Volcanology and Geothermal Research, 29, 125-157.
• Umino, S. and Kushiro, I. (1989) Experimental studies on boninite petrogenesis. In Boninites and Related Rocks (Crawford, A.J. Ed.). Unwin Hyman, London, 89-111.
•  Wilkinson,  J.F.G. and  Hensel,  H.D. (1988) The petrology of some picrites from Mauna Loa and Kilauea volcanoes, Hawaii. Contributions to Mineralogy and Petrology, 98, 326-345.
•  Yamada,  R.,  Takahashi,  T. and  Ogita,  Y. (2023) Petrogenesis of Oligocene to Miocene volcanic rocks from the Toyama basin of the SW Japan arc: Temporal change of arc volcanism during the back-arc spreading in the Japan Sea. Journal of Mineralogical and Petrological Sciences, 118, 221219a.
•  Yamakawa,  M. and  Chihara,  K. (1968) Petrological study of the Ogi Basalt, Sado Island. Contributions from Department of Geology and Mineralogy, Niigata University, 2, 41-80 (in Japanese with English abstract).
•  Yamamoto,  M. (1988) Picritic primary magma and its source mantle for Oshima-Ōshima and back-arc side volcanoes, Northeast Japan Arc. Contributions to Mineralogy and Petrology, 99, 352-359.
•  Yanagisawa,  Y. (2012) Age of the Miocene basalt distributed in Ogi Peninsula, Sado Island, Niigata Prefecture, Japan: reexamination by diatom biochronology. Open-File Report of the Geological Survey of Japan, AIST, 569, 1-19 (in Japanese with English abstract).
•  Yokoyama,  K.,  Tiba,  K. and  Chihara,  K. (1992) Picritic sill in the Sado island, Japan. Bulletin of the National Science Museum. Series C, 18, 1-11.