Volume 17, Issue 6 p. 2298-2322
Research Article
Free Access

Melt-rock interactions and fabric development of peridotites from North Pond in the Kane area, Mid-Atlantic Ridge: Implications of microstructural and petrological analyses of peridotite samples from IODP Hole U1382A

Yumiko Harigane

Corresponding Author

Yumiko Harigane

Research Institute of Geology and Geoinformation, Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology, Tsukuba, Japan

Correspondence to: Y. Harigane, [email protected]Search for more papers by this author
Natsue Abe

Natsue Abe

Research and Development Center for Ocean Drilling Science/Mantle and Continental Crust Drilling Research Group, Japan Agency for Marine-Earth Science and Technology, Yokosuka, Japan

Search for more papers by this author
Katsuyoshi Michibayashi

Katsuyoshi Michibayashi

Institute of Geosciences, Shizuoka University, Shizuoka, Japan

Search for more papers by this author
Jun-Ichi Kimura

Jun-Ichi Kimura

Department of Solid Earth Geochemistry, Japan Agency for Marine-Earth Science and Technology, Yokosuka, Japan

Search for more papers by this author
Qing Chang

Qing Chang

Department of Solid Earth Geochemistry, Japan Agency for Marine-Earth Science and Technology, Yokosuka, Japan

Search for more papers by this author
First published: 21 May 2016
Citations: 8

Abstract

North Pond is an isolated sedimentary pond on the western flank of the Kane area along the Mid-Atlantic Ridge. Drill-hole U1382A of IODP Expedition 336 recovered peridotite and gabbro samples from a sedimentary breccia layer in the pond, from which we collected six fresh peridotite samples. The peridotite samples came from the southern slope of the North Pond where an oceanic core complex is currently exposed. The samples were classified as spinel harzburgite, plagioclase-bearing harzburgite, and a vein-bearing peridotite that contains tiny gabbroic veins. No obvious macroscopic shear deformation related to the formation of a detachment fault was observed. The spinel harzburgite with a protogranular texture was classified as refractory peridotite. The degree of partial melting of the spinel harzburgite is estimated to be ∼17%, and melt depletion would have occurred at high temperatures in the uppermost mantle beneath the spreading axis. The progressive melt–rock interactions between the depleted spinel harzburgite and the percolating melts of Normal-Mid Ocean Ridge Basalt (N-MORB) produced the plagioclase-bearing harzburgite and the vein-bearing peridotite at relatively low temperatures. This implies that the subsequent refertilization occurred in an extinct spreading segment of the North Pond after spreading at the axis. Olivine fabrics in the spinel and plagioclase-bearing harzburgites are of types AG, A, and D, suggesting the remnants of a mantle flow regime beneath the spreading axis. The initial olivine fabrics appear to have been preserved despite the later melt–rock interactions. The peridotite samples noted above preserve evidence of mantle flow and melt–rock interactions beneath a spreading ridge that formed at ∼8 Ma.

Key Points

  • Melt-rock interactions formed plagioclase-bearing harzburgite and vein-bearing peridotite
  • Melt-impregnated harzburgite preserved olivine fabric of the host spinel harzburgite
  • Peridotite preserved records of mantle flow and melt–rock interactions beneath spreading ridge at ∼8 Ma

1 Introduction

North Pond is an isolated sedimentary pond about ∼8 × 15 km2 in size that is located on the western flank of the Mid-Atlantic Ridge Kane (MARK) area at 22°45'N, 46°05'W (Figure 1). This area exhibits a normal magnetic polarity that has been interpreted as magnetic anomaly 4 [Melson et al., 1978], corresponding to a basement age of between 7.43 and 8.07 Ma using the timescale of Cande and Kent [1995]. Sediment thicknesses range up to a maximum of 300 m in the southernmost part of the pond. Holes 395 and 395A at Site 395 of DSDP Leg 45 penetrated the southeastern part of the pond (Figure 1; 22°45.35'N, 46°04.90'W) at water depths of 4414–4483 m [Expedition 336 Scientists, 2012; Edwards et al., 2014]. The basement lithologies are dominated by several massive pillow lavas that are separated from each other by sedimentary breccias [Melson et al., 1978; Bartetzko et al., 2001]. The sedimentary breccias are debris flow deposits that contain cobbles of gabbro and serpentinized peridotite. A peridotite–gabbro complex several meters thick with brecciated contacts was cored in Hole 395 [Melson et al., 1978; Arai and Fujii, 1979; Sinton, 1979].

Details are in the caption following the image

(a) Regional map of the study area on the Mid-Atlantic Ridge (data from ETOPO1 and Amante and Eakins [2009]). The black square indicates our study area. (b) Location map of the North Pond in the Mid-Atlantic Ridge Kane area. Data are from the Global Multiresolution Topography Data Portal (http://www.marine-geo.org/portals/gmrt/; Ryan et al. [2009]). The red square indicates the North Pond area. Black circles indicate previous ocean drilling sites along the spreading axis. The Kane fracture zone is shown as a broken black line. The spreading axis is marked roughly by a double black line. (c) Bathymetric map of North Pond showing the locations of DSDP Hole 395A, ODP Hole 1074A, and Sites U1382, U1383, and U1384 [Expedition 336 Scientists, 2012]. The bathymetry map data for North Pond were obtained from Cruise Conrad 30-01 in 1989 (Co-Chief Scientists R. Detrick and J. Mutter) (data available from www.marine-geo.org/tools/search/entry.php?id=RC3001) [see Schmidt-Schierhorn et al., 2012]. The red star indicates the present study site.

In 2011, microbiological investigations as part of IODP Expedition 336 were conducted at the North Pond. This expedition recovered cores from three sites (U1382, U1383, and U1384) (Figure 1) [Expedition 336 Scientists, 2012; Edwards et al., 2014]. Fresh peridotite, partially serpentinized peridotite, serpentinite, and gabbro samples were recovered from the sedimentary breccia unit in Hole U1382A, which was drilled close to Hole 395A [Expedition 336 Scientists, 2012; Edwards et al., 2014]. Many geological and geophysical surveys in the MARK area have been conducted by a number of cruises, including DSDP, ODP, and IODP (Figure 1) [see Cannat et al., 1995a, 1995b; Ghose et al., 1996; Dick et al., 2008, 2010; Cann et al., 2015]. The crustal and mantle structures in the MARK area are therefore well understood, but the North Pond has not yet been studied in detail from a petrological point of view, especially with regard to the crustal and mantle structures in the gabbros and peridotites. Accordingly, the petrological relationships between the peridotites from North Pond and the other abyssal peridotites in the MARK area are unknown.

North Pond is isolated and located on an extinct spreading segment, far from the current spreading axis and transform zone (Figure 1) [Cann et al., 2015]. Thus, the peridotite samples collected from North Pond during IODP Expedition 336 are suitable for investigating the mantle structure beneath an extinct spreading segment, and for comparing this mantle structure with that beneath the current spreading axis of the MARK area. In this paper we present rock descriptions, analyses of mineral fabrics, and detailed geochemical analyses of the minerals in fresh peridotite samples recovered from Hole U1382A. Our new data allow us to discuss melt–rock interactions and fabric development beneath the North Pond extinct spreading segment.

2 Geological Setting of Hole U1382A

The basement at Hole U1382A was cored from 110 to 210 mbsf (meters beneath sea floor) with an overall recovery of 31% [Expedition 336 Scientists, 2012; Edwards et al., 2014]. The recovered core material was classified into eight main lithologic units (Figure 2a), with boundaries defined by contacts between pillow lava, massive lava, and interlayered sedimentary units [Expedition 336 Scientists, 2012; Edwards et al., 2014]. Although the basalts include seven distinct chemical types, the main lithologies are highly phyric plagioclase–olivine basalts and aphyric basalts. Each basalt unit consists of basaltic lavas, pillow sequences, and intrusive and massive units of diabase (Figure 2a). There are three massive aphyric units (Units 1, 3, and 8) and one highly phyric unit (Unit 6) that lacks obvious contact and glassy zones (Figure 2a). The phyric basalts below the brecciated unit correspond stratigraphically to basalts encountered in Hole 395A [Expedition 336 Scientists, 2012; Edwards et al., 2014]. The aphyric pillow basalts in Units 2, 4, and 7 are petrographically similar to each other, but the aphyric pillow basalts of Unit 7 may be derived from a different parental magma, judging from their Zr/Y and Ti/Zr ratios [Expedition 336 Scientists, 2012].

Details are in the caption following the image

(a) Core recovery and lithology of the basement section in Hole U1382A. Separate core sections are labeled 2R to 12R. Black intervals in the ‘Core Recovery’ column indicate sections of recovered core. (b) Close-up view (half core scan) of the sedimentary breccia (interval Exp.336-U1382A-8R-4-31-43) in Unit 5. (c) Close-up view (half core scan) of peridotite cut by carbonate veins (interval Exp.336-U1382A-9R-1-90-106) in the sedimentary breccia zone at Unit 5. The centers of the carbonate veins contain strings of round and/or elongate manganese oxide grains.

The sedimentary breccia unit (Unit 5) in Cores 8R and 9R was encountered between 161.3 and 173.2 mbsf in Hole U1382A (Figure 2a). This unit shows a variety of clasts (Figures 2b and 2c) that include sediments, lithic fragments, basalts (phyric and aphyric), gabbroic rock, and peridotite (Figure 2c) [Expedition 336 Scientists, 2012; Edwards et al., 2014]. The clasts vary in size from very fine to pebble or boulder (Figures 2b and 2c) [Expedition 336 Scientists, 2012], and they are poorly sorted. The matrix is mainly pelagic nannofossil ooze with tiny fragments of basaltic and gabbroic rock. The subunits in Unit 5 were assigned to the different lithologies identified in the sedimentary breccia, but the lack of contacts and the likely disturbance during drilling precluded the determination of depth-related relationships between the numerous basaltic and gabbroic cobbles associated with the sedimentary layers [Expedition 336 Scientists, 2012].

We collected six fresh peridotites from the sedimentary breccia unit (Unit 5). These samples lack any Mn coating, and neither foliation nor lineation is visible to the naked eye. The samples are weakly serpentinized and exhibit minor cataclastic deformation that led to the development of a network of carbonate-filled veins, along which the rocks have been subjected to oxidative alteration, with olivine breaking down to clays, oxides, and carbonates [Expedition 336 Scientists, 2012]. The carbonate veins are variable in thickness (<0.5–4.0 mm), and they cut across the microstructures of the peridotites (Figure 2c). The centers of the carbonate veins contain strings of round and/or elongate opaque grains (Figure 2c).

3 Microstructures

The microstructures of six peridotites were studied under an optical microscope at the Geological Survey of Japan (GSJ)/AIST and under a scanning electron microscope (SEM) HITACHI S-3400N (Hitachi High-Technologies Co., Tokyo, Japan) at Shizuoka University, Japan. We also used a scanning X-ray analytical microscope M4 Tornado (Bruker AXS, Berlin, Germany) at the GSJ-Lab of the Geological Survey of Japan/AIST in order to calculate the modal composition of the vein-bearing peridotite. The peridotites consist of spinel harzburgite, plagioclase-bearing harzburgite, and vein-bearing peridotite (Figure 3 and Table 1). Most of the peridotites lack any obvious macroscopic evidence of shear deformation.

Details are in the caption following the image

Representative photomicrographs of the analyzed peridotites. (a) Photomicrograph of spinel harzburgite (sample U1382A-8R-4-60-63). Cross-polarized light. (b) Photomicrograph of spinel harzburgite (sample U1382A-9R-1-117-122). This sample is cut by carbonate veinlets (white arrow). Cross-polarized light. (c) Protogranular texture in spinel harzburgite (sample U1382A-9R-1-79-83). Orthopyroxene and clinopyroxene grains contain exsolution lamellae. Cross-polarized light. (d) Olivine subgrain boundaries in spinel harzburgite (sample U1382A-8R-4-60-63). Cross-polarized light. (e) Vermicular shapes of spinel grains in spinel harzburgite (sample U1382A-8R-4-60-63). Orthopyroxene and olivine grains surround the spinel grains. Cross-polarized light. (f) Representative photomicrograph of plagioclase-bearing harzburgite (sample U1382A-8R-4-72-76). Cross-polarized light. (g) Plagioclase surrounding spinel grains in the plagioclase-bearing harzburgite. Cross-polarized light. (h) Photomicrograph of vein-bearing peridotite (sample U1382A-9R-1-71-74). The gabbroic vein in the center of the image intruded the peridotite. Cross-polarized light. (i) Poikilitic texture in vein-bearing peridotite comprises oikocrysts of clinopyroxene and small chadacrysts of olivine and orthopyroxene. Cross-polarized light. (j) Spinel grains in vein-bearing peridotite are black in color, anhedral to vermicular in shape (∼1 mm in size), and surrounded by dusty plagioclase (red arrow). Plane-polarized light. Ol, olivine; Opx, orthopyroxene; Cpx, clinopyroxene; Sp, spinel; Pl, plagioclase.

Table 1. Modal Compositions and Seismic Properties of the Analyzed Peridotites in IODP Exp.336
Core Section Interval (cm) Lithology Ol (%) Opx (%) Cpx (%) Sp (%) Pl (%) Othersa (%) VPx VPy VPz Vp anisotropy VPx/VPy VPy/VPz K r Fabric-Index Angle (FIA)
U1382A-
8R-4 60-63 Spinel harzburgite 73.0 23.7 0.9 2.3 0.3 8.84 8.20 8.06 9.23 1.08 1.02 4.49 0.08 77
9R-1 79-83 Spinel harzburgite 82.0 16.4 0.9 0.9 0.0 8.56 8.37 8.15 4.91 1.02 1.03 0.84 0.04 40
9R-1 97-102 Spinel harzburgite 50.9 37.9 8.6 1.1 1.6 8.61 8.29 8.20 4.88 1.04 1.01 3.52 0.04 74
9R-1 117-122 Spinel harzburgite 66.7 27.4 5.7 0.3 0.0 8.48 8.40 8.20 3.36 1.01 1.02 0.39 0.03 21
8R-4 72-76 Plagioclase-bearing harzburgite 87.9 7.5 1.7 1.4 1.6 0.1 8.57 8.34 8.12 5.39 1.03 1.03 1.02 0.04 46
9R-1 71-74 Vein-bearing peridotite 50.3 12.6 22.8 0.8 11.5b 2.0
  • a Others (%) includes serpentine and void.
  • b Pl includes Pl of gabbroic vein.

Four specimens of spinel harzburgite from DSDP 395 (U1382A-8R-4-60-63, U1382A-9R-1-79-83, U1382A-9R-1-97-102, and U1382A-9R-1-117-122) consist mainly of olivine (50.9–82.0 vol%), orthopyroxene (16.4–37.9 vol%), clinopyroxene (0.9–8.6 vol%), and spinel (0.3–2.3 vol%), together with accessory serpentine and magnetite (Figures 3a and 3b; Table 1). The mineral assemblage and modal compositions are consistent with those of harzburgites reported by Michael and Bonatti [1985]. The spinel harzburgites display a protogranular texture [Mercier and Nicolas, 1975] that consists of coarse-grained olivine (ca. 1–20 mm) and medium-grained orthopyroxene and clinopyroxene (2–15 mm) (Figure 3c). Similar microstructures were also observed in serpentinized peridotite from Hole 395, as reported by Boudier [1979]. The olivine grains show evidence of weak intracrystalline deformation (e.g., wavy extinction and subgrain boundaries) (Figure 3d), and the orthopyroxene and clinopyroxene grains have kinked or distorted lamellae (Figure 3c). The spinels are pale brown, dark brown, or black in color, and are anhedral to vermicular in shape (∼1 mm in size) (Figure 3e). Serpentine has replaced olivine and is developed in networks. Carbonate veinlets in several samples (e.g., U1382A-9R-1-97-102 and U1382A-9R-1-117-122) cut across the main microstructures (Figure 3b).

The plagioclase-bearing harzburgite (U1382A-8R-4-72-76) consists mainly of olivine (87.9 vol%), orthopyroxene (7.5 vol%), clinopyroxene (1.7 vol%), spinel (1.4 vol%), and plagioclase (1.6 vol%), along with accessory serpentine and magnetite (0.1 vol%) (Figure 3f and Table 1). The plagioclase-bearing harzburgite also shows a protogranular texture [Mercier and Nicolas, 1975]. Coarse-grained olivine grains (ca. 1–20 mm in size) show signs of weak intracrystalline deformation, and the medium-grained orthopyroxene and clinopyroxene grains (2–15 mm) show kinked or distorted lamellae. The spinels are anhedral to vermicular in shape and black in color (Figure 3g). Plagioclase grains usually surround the spinels (Figure 3f) and also occur between olivine and pyroxene. The plagioclases are 0.1 to ∼1.5 mm in grain size and are partially altered. Serpentine has replaced olivine and is developed in networks.

The vein-bearing peridotite (U1382A-9R-1-71-74) was intruded by a vein of gabbroic rock (Figure 3h), and the boundary between vein and peridotite host is indistinct (Figures 3h and 4). The vein-bearing peridotite contains olivine (50.3 vol%), orthopyroxene (12.6 vol%), clinopyroxene (22.8 vol%, including the clinopyroxene in the gabbroic vein), spinel (0.8 vol%), and plagioclase (11.5 vol%, including the plagioclase in the gabbroic vein), along with secondary serpentine and magnetite (2.0 vol%) (Table 1). This sample was termed the vein-bearing peridotite because we cannot distinguish gabbroic vein portions from the host peridotite in terms of modal composition. The microstructure is poikilitic with clinopyroxene oikocrysts enclosing small chadacrysts of olivine and orthopyroxene (Figures 3i and 4). Olivine grains show weak intracrystalline deformation in the form of wavy extinction or subgrain boundaries. Orthopyroxene and clinopyroxene grains have kinked or distorted lamellae. Most of the clinopyroxene grains are found in the vicinity of the gabbroic vein; they show no compositional zoning (Figure 4), and it is often unclear whether they belong to the gabbroic vein or the host peridotite. Black-colored spinel in the peridotite is anhedral to vermicular (∼1 mm in size), and surrounded by dusty plagioclase and aggregates of Si–Al–Mg or Si–Al–Ca rich minerals (Figure 3j). The spinel grains also contain inclusions of the plagioclase and Si–Al–Mg or Si–Al–Ca rich minerals. The plagioclases are <1 to ∼4 mm in diameter and are locally altered. Fine-grained plagioclases occur between olivines and pyroxenes and around spinel grains (Figure 4). Coarse-grained plagioclases are found in the gabbroic vein (Figure 4).

Details are in the caption following the image

Ca, Fe, Al, and Mg element distribution maps for the entire thin section of vein-bearing peridotite (sample U1382A-9R-1-71-74) shown in Figure 3h. We measured the sample with a scanning X-ray analytical microscope using the following analytical conditions: 500 Pa in vacuum, 30 kV in tube voltage, 340 µA in tube current, 25 µm beam spot size, and 50 mm/s scan speed per pixel. The imaging area was 27 mm × 31 mm with 10 µm per pixel. Ol, olivine; Opx, orthopyroxene; Cpx, clinopyroxene; Sp, spinel; Pl, plagioclase.

4 Crystallographic Preferred Orientations

The crystallographic preferred orientations (CPOs) of olivine, orthopyroxene, clinopyroxene, and plagioclase were analyzed in highly polished thin sections of the peridotite samples using an SEM (HITACHI S-3400N) equipped for electron back-scattered diffraction (EBSD, with HKL Channel5 software), at Shizuoka University, Japan. The orientations of the crystallographic axes of olivine, orthopyroxene, clinopyroxene, and plagioclase are presented in equal-area, lower-hemisphere projections (Figure 5).

Details are in the caption following the image

Crystallographic preferred orientation (CPO) data for olivine, orthopyroxene, clinopyroxene, and plagioclase in the spinel harzburgites, the plagioclase-bearing harzburgite, and the vein-bearing peridotite. The data are plotted on equal-area, lower-hemisphere projections. Contours for the olivine CPOs are multiples of uniform density. N is the number of analyzed grains of olivine, orthopyroxene, clinopyroxene, and plagioclase.

We manually measured the crystallographic orientations of 40–142 olivine grains, 22–58 orthopyroxene grains, 9–28 clinopyroxene grains, and 18 plagioclase grains, visually checking the computerized indexation of each diffraction pattern and merging the data from two or three thin sections of the analyzed samples [cf. Harigane et al., 2011]. Analyses of these peridotite samples, except for the vein-bearing peridotite, were performed in two or three vertical sections to exclude the effects of large mineral grains on the CPO results (Figure 5).

The orientations of foliation and lineation were detected by calculating P-wave velocity properties in the analyzed peridotites (Table 1). This P-wave velocity was calculated from a single crystal elastic constant, density, and CPO of olivines. We inputted the parameters of a 100% olivine modal composition to define the characteristics of the olivine fabric, the EBSD-measured CPO of olivine, and the elastic constants of olivine, and we used a Voigt-Reuss-Hill average scheme [Mainprice et al., 2000]. The olivine elastic constant used in our calculations was that of Abramson et al. [1997]. The results of calculating the P-wave velocity properties are shown in Table 1. For this study, we positioned the lineation in an E–W direction on the pole figure diagrams, based on the results from the calculated P-wave velocity properties.

Although the olivines were generally observed to have weak CPO concentrations, olivine (010)[100] patterns are present in the spinel harzburgites, with an alignment of [100]-axes and with [010]-axes perpendicular to the (100) plane (Figure 5). Boudier [1979] also reported similar olivine fabrics in serpentinized peridotite. On the other hand, the vein-bearing peridotite is characterized by a poorly defined olivine CPO. We have not been able to determine clearly whether the orthopyroxenes, clinopyroxenes, and plagioclases in the peridotites have CPOs (Figure 5) because of the small number of grains and because their relationships to the macroscopic mineral shape fabrics are unknown.

5 Mineral Chemistry

5.1 Major Elements

Major element analysis of the constituent minerals in the peridotites was carried out with a JXA-8800 (JEOL, Tokyo, Japan) electron probe micro-analyzer (EPMA) at the GSJ-Lab at the Geological Survey of Japan/AIST. The analyses were undertaken with an acceleration voltage of 15 kV, a specimen current of 12 nA, and a beam diameter of 2 µm. Natural and synthetic mineral standards provided by JEOL were employed for data calibrations. The X-ray peak of Ni was counted for 30 s, whereas other elements were counted for 20 s. Data were corrected using the methods of Bence and Albee [1968]. The average major element compositions of each mineral are listed in Tables 2-6.

Table 2. Average Major Element Compositions (in wt %) of Olivine in the Analyzed Peridotites
Sample No. Lithology N SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O Cr2O3 NiO Total Fo#a
U1382A-8R-4-60-63 Spinel harzburgite 19 40.39 0.01 0.00 9.37 0.14 50.20 0.07 0.00 0.00 0.01 0.38 100.58 90.52
stdb 0.35 0.01 0.01 0.29 0.02 0.31 0.03 0.01 0.01 0.02 0.03 0.42 0.27
U1382A-9R-1-79-83 10 40.56 0.00 0.00 9.19 0.15 50.12 0.07 0.00 0.01 0.01 0.40 100.51 90.67
std 0.39 0.01 0.01 0.35 0.03 0.37 0.02 0.00 0.01 0.01 0.05 0.38 0.36
U1382A-9R-1-97-102 17 40.40 0.01 0.01 9.01 0.13 50.02 0.07 0.01 0.00 0.01 0.41 100.06 90.82
std 0.19 0.01 0.01 0.31 0.02 0.22 0.02 0.01 0.01 0.01 0.03 0.43 0.28
U1382A-9R-1-117-122 9 40.50 0.00 0.00 9.20 0.14 50.05 0.06 0.00 0.00 0.01 0.39 100.35 90.66
std 0.29 0.01 0.00 0.17 0.02 0.20 0.02 0.00 0.00 0.02 0.04 0.40 0.16
U1382A-8R-4-72-76 Plagioclase-bearing harzburgite 8 40.28 0.01 0.00 11.36 0.18 48.45 0.05 0.00 0.00 0.01 0.34 100.68 88.38
std 0.19 0.01 0.00 0.44 0.04 0.39 0.02 0.00 0.00 0.01 0.06 0.27 0.47
U1382A-9R-1-71-74 Vein-bearing peridotite 13 40.14 0.01 0.00 11.25 0.18 48.49 0.05 0.00 0.01 0.01 0.36 100.50 88.48
std 0.26 0.01 0.01 0.70 0.02 0.70 0.03 0.00 0.01 0.02 0.03 0.34 0.77
  • a Fo#=100×Mg/(Mg+Fe) ratio.
  • b std Standard deviation (1 sigma).
Table 3. Average Major Element Compositions (in wt %) of Orthopyroxene in the Analyzed Peridotites
Sample No. Lithology n SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O Cr2O3 NiO Total Mg#a
U1382A-8R-4-60-63 Spinel harzburgite 12 55.38 0.05 3.18 6.08 0.15 33.32 1.57 0.00 0.00 0.78 0.09 100.60 90.72
stdb 0.31 0.02 0.22 0.16 0.04 0.27 0.35 0.01 0.01 0.10 0.02 0.33 0.23
U1382A-9R-1-79-83 16 55.45 0.04 3.22 5.95 0.15 33.07 1.86 0.01 0.00 0.79 0.10 100.64 90.83
std 0.50 0.02 0.35 0.31 0.02 0.61 0.63 0.01 0.00 0.13 0.03 0.47 0.39
U1382A-9R-1-97-102 13 54.94 0.03 3.40 5.97 0.16 33.07 1.62 0.00 0.00 0.85 0.10 100.14 90.80
std 0.41 0.02 0.29 0.17 0.03 0.80 0.89 0.01 0.00 0.13 0.03 0.44 0.14
U1382A-9R-1-117-122 12 55.01 0.04 3.54 5.64 0.14 32.36 2.62 0.01 0.01 0.87 0.10 100.34 91.09
std 0.52 0.02 0.33 0.34 0.03 1.47 1.82 0.01 0.01 0.14 0.03 0.39 0.33
U1382A-8R-4-72-76 Plagioclase-bearing harzburgite 4 54.96 0.09 3.07 7.14 0.17 31.63 2.44 0.04 0.01 1.03 0.09 100.66 88.76
std 0.21 0.07 0.63 0.47 0.05 0.56 1.12 0.03 0.00 0.08 0.02 0.13 0.64
U1382A-9R-1-71-74 Vein-bearing peridotite 14 55.64 0.25 2.15 7.21 0.20 32.60 1.50 0.02 0.00 0.71 0.09 100.38 88.96
std 0.30 0.07 0.31 0.59 0.03 0.33 0.45 0.02 0.01 0.10 0.03 0.52 0.85
  • a Mg#=100×Mg/(Mg+Fe) ratio.
  • b std Standard deviation (1 sigma).
Table 4. Average Major Element Compositions (in wt %) of Clinopyroxene in the Analyzed Peridotites
Sample No. Lithology n SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O Cr2O3 NiO Total Mg#a
U1382A-8R-4-60-63 Spinel harzburgite 6 51.84 0.08 4.24 2.86 0.10 17.33 22.57 0.06 0.00 1.29 0.05 100.43 91.52
stdb 0.45 0.02 0.50 0.25 0.03 0.40 0.75 0.01 0.01 0.21 0.04 0.51 0.55
U1382A-9R-1-79-83 5 51.96 0.10 4.13 2.89 0.08 17.20 22.88 0.07 0.00 1.23 0.05 100.59 91.40
std 0.59 0.02 0.45 0.18 0.01 0.61 0.88 0.01 0.00 0.20 0.02 0.69 0.22
U1382A-9R-1-97-102 8.00 51.76 0.10 4.25 2.87 0.11 17.38 22.28 0.07 0.00 1.25 0.06 100.12 91.54
std 0.38 0.02 0.55 0.21 0.03 0.44 1.00 0.02 0.00 0.22 0.03 0.56 0.51
U1382A-9R-1-117-122 8.00 51.47 0.10 4.57 2.70 0.09 17.00 22.95 0.07 0.01 1.37 0.05 100.37 91.82
std 0.25 0.03 0.36 0.13 0.02 0.39 0.43 0.02 0.01 0.13 0.04 0.39 0.32
U1382A-8R-4-72-76 Plagioclase-bearing harzburgite 6.00 51.92 0.13 4.39 3.75 0.14 17.36 21.14 0.29 0.00 1.49 0.05 100.65 89.22
std 0.26 0.10 0.35 0.45 0.03 0.75 1.19 0.06 0.00 0.08 0.02 0.20 0.89
U1382A-9R-1-71-74 Vein-bearing peridotite 19 52.21 0.57 3.37 3.66 0.13 17.90 20.55 0.40 0.00 1.31 0.06 100.17 89.74
std 0.45 0.16 0.43 0.60 0.02 1.76 2.30 0.07 0.01 0.19 0.03 0.43 0.75
  • a Mg#=100xMg/(Mg+Fe) ratio.
  • b std Standard deviation (1 sigma).
Table 5. Average Major Element Compositions (in wt %) of Spinel in the Analyzed Peridotites
Sample No. Lithology n SiO2 TiO2 Al2O3 FeOa MnO MgO CaO Na2O K2O Cr2O3 NiO Total Mg#b Cr#c
U1382A-8R-4-60-63 Spinel harzburgite 9 0.05 0.08 36.14 15.49 0.25 16.13 0.01 0.00 0.00 32.15 0.22 100.52 64.99 37.38
stda 0.13 0.02 0.75 0.48 0.03 0.23 0.01 0.01 0.01 0.81 0.04 0.42 0.88 1.05
U1382A-9R-1-79-83 10 0.00 0.07 37.26 15.89 0.23 15.85 0.02 0.01 0.00 31.01 0.21 100.56 64.02 35.84
std 0.00 0.03 1.54 0.78 0.03 0.24 0.02 0.01 0.01 1.78 0.04 0.26 1.41 2.23
U1382A-9R-1-97-102 14 0.01 0.06 36.71 15.74 0.23 15.85 0.01 0.01 0.00 31.37 0.21 100.20 64.23 36.44
std 0.01 0.02 0.71 0.30 0.02 0.23 0.01 0.01 0.01 0.72 0.03 0.55 0.70 0.93
U1382A-9R-1-117-122 6 0.00 0.07 39.01 15.13 0.23 16.43 0.01 0.01 0.01 29.44 0.19 100.54 65.94 33.61
std 0.01 0.02 1.38 0.50 0.04 0.15 0.01 0.01 0.01 1.80 0.04 0.34 0.82 2.14
U1382A-8R-4-72-76 Plagioclase-bearing harzburgite 4 0.02 0.35 23.54 23.34 0.35 11.42 0.00 0.01 0.01 41.37 0.12 100.53 46.57 54.11
std 0.01 0.06 0.53 0.62 0.03 0.67 0.00 0.01 0.01 0.79 0.02 0.46 1.97 0.89
U1382A-9R-1-71-74 Vein-bearing peridotite 3 0.01 1.34 21.60 23.26 0.34 11.47 0.00 0.01 0.01 41.73 0.13 99.90 46.80 56.48
std 0.02 0.66 1.54 2.12 0.02 1.01 0.00 0.02 0.01 0.94 0.06 0.60 4.46 1.28
  • a std Standard deviation (1 sigma).
  • b Mg#=100xMg/(Mg+Fe) ratio.
  • c Cr#=100xCr/(Cr+Al).
Table 6. Average Major Element Compositions (in wt %) of Plagioclase in the Analyzed Peridotites
Sample No. Lithology n SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O Cr2O3 NiO Total An%a Remarks
U1382A-8R-4-72-76 Plagioclase-bearing harzburgite 9 46.48 0.00 34.08 0.22 0.01 0.03 17.90 1.68 0.04 0.05 0.01 100.50 85.29 Plagioclase around spinel
stdb 1.04 0.00 0.50 0.04 0.01 0.01 0.59 0.38 0.01 0.02 0.01 0.56 3.18
U1382A-9R-1-71-74 Vein-bearing peridotite 4 51.44 0.05 30.99 0.23 0.01 0.20 13.83 3.70 0.05 0.04 0.02 100.56 67.16 Plagioclase around spinel
std 0.90 0.02 0.23 0.04 0.01 0.29 0.65 0.19 0.01 0.05 0.02 0.36 2.17
U1382A-9R-1-71-74 Vein-bearing peridotite 8 52.07 0.04 30.20 0.14 0.00 0.04 13.38 4.12 0.04 0.01 0.01 100.07 64.06 Plagioclase of gabbroic vein
std 0.52 0.02 0.32 0.03 0.01 0.02 0.38 0.22 0.01 0.02 0.01 0.49 1.78
  • a An%=100xCa/(Ca+Na+K) ratio.
  • b std Standard deviation (1 sigma).

5.1.1 Spinel Harzburgite

The forsterite contents (Fo# = 100 × Mg/(Mg + Fe) molar ratio) of the olivines in the spinel harzburgites range from 90.1 to 92.5 (Figure 6a and Table 2). Sinton [1979] and Niida [1997] reported similar ranges of forsterite (Fo# = 90.0–91.0 at Hole 395, and Fo# = 90.3–91.5 at Hole 920D, respectively). The NiO contents of the olivines in the spinel harzburgites vary from 0.30 to 0.48 wt%, and this fits the compositional range of mantle olivines [Takahashi et al., ] (Figure 6a and Table 2). The values of Mg number (Mg# = 100 × Mg/(Mg + Fe)) for the orthopyroxenes in the spinel harzburgites range from 90.3 to 91.7 (Table 3).

Details are in the caption following the image

Compositional variations of olivine, clinopyroxene, and spinel in the spinel harzburgites (Sp. Hz.), the plagioclase-bearing harzburgite (Pl-bg. Hz.), and the vein-bearing peridotite (Vein-bg. Peridotite) from Hole U1382A in the North Pond. (a) Nickel versus forsterite (Fo) contents in the olivines. (b) Na2O contents versus TiO2 contents in the clinopyroxenes. (c) Cr# versus Mg# in the spinels. The field of abyssal peridotites (gray shading) is from Dick et al. [2010]. (d) Cr# versus TiO2 in the spinels.

The orthopyroxenes have wide ranges of Al2O3 (2.3–4.0 wt%) and Cr2O3 (0.55–1.1 wt%) contents and low TiO2 contents (∼0.1 wt%) (Table 3). The values of Mg# in the clinopyroxenes of the spinel harzburgites range from 90.8 to 92.4 (Table 4). The compositional ranges of clinopyroxenes in the spinel harzburgites are comparable to those reported by Sinton [1979] and Ghose et al. [1996] for Hole 395 and Niida [1997] and Ross and Elthon [1997] for Site 920, the locations of which are situated on the western wall of the median valley of the Mid-Atlantic Ridge (MAR). The Al2O3 and Cr2O3 contents of the clinopyroxenes vary within slightly larger ranges (3.3–4.9 wt% Al2O3 and 0.83–1.5 wt% Cr2O3) than those of the orthopyroxenes (Tables 3 and 4). In contrast, the Na2O (0.02–0.11 wt%) and TiO2 (0.06–0.14 wt%) contents are typically low, and these contents are among the lowest of any abyssal peridotite (Figure 6b and Table 4). Similar low Na and Ti contents were reported by Ghose et al. [1996] for Hole 395, Niida [1997] and Ross and Elthon [1997] for Site 920, Tamura et al. [2008] for Site U1309 of the Atlantis Massif, Constantin et al. [1995] for the Terevaka transform fault, and for Moll et al. [] Hole 1274A.

The values of Mg# in the spinels of the spinel harzburgites vary from 61.9 to 66.6. The values of the Cr number (Cr# = 100 × Cr/(Cr + Al)) vary from 30.1 to 38.6 (Figure 6c and Table 5), and the TiO2 contents are very low (0.03–0.08 wt%) (Figure 6d and Table 5). Although the compositions of the spinels in the spinel harzburgites are slightly offset toward iron-rich compositions, these compositions lie within the range for abyssal peridotites from mid-ocean ridges (Figure 6c and Table 5) [e.g., Dick and Bullen, 1984; Dick et al., 2010].

5.1.2 Plagioclase-Bearing Harzburgite and Vein-Bearing Peridotite

Olivines in the plagioclase-bearing harzburgite and the vein-bearing peridotite have lower contents of Fo# (Fo# = 87.6–90.0) and NiO (0.29–0.42 wt%) than the olivines in the spinel harzburgites (Figure 6a and Table 2), and the orthopyroxenes and clinopyroxenes of the plagioclase-bearing harzburgite and the vein-bearing peridotite also have lower Mg# values than in the spinel harzburgites (Table 3). The Cr2O3 (0.96–1.1 wt%), Al2O3 (2.3–3.9 wt%), and TiO2 (0.03–0.18 wt%) contents of the orthopyroxenes in the plagioclase-bearing harzburgite are similar to those in the spinel harzburgites (Table 3) whereas the orthopyroxenes of the vein-bearing peridotite show lower Cr2O3 (0.56–0.96 wt%) and Al2O3 (1.6–2.7 wt%) contents than those in the spinel harzburgites and plagioclase-bearing harzburgite (Table 3). The Cr2O3 (1.4–1.6 wt%), Al2O3 (3.8–4.7 wt%), and TiO2 (0.08–0.34 wt%) contents of clinopyroxenes in the plagioclase-bearing harzburgite resemble those in the spinel harzburgites. Although the TiO2 contents of these clinopyroxenes are almost always uniformly low (<0.15 wt%), some parts of the crystals show higher contents up to 0.34 wt%. In contrast, the Na2O (0.21–0.39 wt%) contents of the clinopyroxenes in the plagioclase-bearing harzburgite are higher than those in the spinel harzburgites (Figure 6b and Table 4). The TiO2 contents of clinopyroxenes in the vein-bearing peridotite also show a wide range from 0.15 to 0.72 wt% (Figure 6b and Table 4), whereas the Cr2O3 (1.1–1.8 wt%) and Al2O3 (2.6–4.5 wt%) contents are relatively constant (Table 4). Of all the analyzed clinopyroxenes, those in the vein-bearing peridotite have the highest contents of TiO2 and Na2O (Figure 6b).

Spinels in the plagioclase-bearing harzburgite and the vein-bearing peridotite have high values of Cr# (53.2–57.5) and high contents of TiO2 (>0.28 wt%) when compared with those in the spinel harzburgites (Figures 6c and 6d). The Ti contents of the spinels in the vein-bearing peridotite are typically much higher than those of the spinels in the plagioclase-bearing harzburgite (which rarely exceed 0.5 wt% TiO2) (Figure 6d).

Plagioclases in the plagioclase-bearing harzburgite have high anorthite contents (An80.8–89.3; Table 6). In contrast, the plagioclase grains that occur around spinel grains in the vein-bearing peridotite have moderate values of An (An64.6–70) (Table 6), and plagioclases in the gabbroic veins of the vein-bearing peridotite have even lower An values (An61.1–66.5) (Table 6).

5.2 Trace Elements

The trace element (Ti, Sr, Y, Zr, Nb, Ba, Hf, Pb, Th, and U) and rare earth element (REE) compositions of the clinopyroxenes in the six peridotite samples were determined using a 260 nm ultraviolet femtosecond laser OK-Fs2000K (OK Laboratory Ltd., Tokyo, Japan) coupled with a sector field inductively coupled plasma–mass spectrometer (ICP–MS) Element XR (ThermoFisher Scientific, Bremen, Germany) at the Department of Solid Earth Geochemistry in the Japan Agency for Marine-Earth Science and Technology, Japan. All analyses were performed by ablating 50 µm diameter spots for the clinopyroxene grains. The BHVO-2G glass issued by the United States Geological Survey (USGS) was used as a standard, and BCR-2G was analyzed to provide quality control of the analyses (Table 7). The analytical techniques used followed those described by Kimura and Chang [2012]. The laboratory biases measured for BCR-2G were better than a 10% relative deviation compared with the reference values for all elements analyzed (Table 7). The typical lower limits of detection (LLD) are shown in Table 7, and were calculated in terms of 3-sigma deviations of the elemental concentrations measured from the near LLD gas blank intensities. The average trace element compositions of the clinopyroxenes in the analyzed peridotites, with one-standard deviation (std) errors, are listed in Tables 7-9. The chondrite-normalized trace-element patterns of the clinopyroxenes in the spinel harzburgites, the plagioclase-bearing harzburgite, and the vein-bearing peridotite are shown in Figure 7.

Details are in the caption following the image

Chondrite-normalized trace-element compositions of clinopyroxenes in each of the peridotite samples. Chondrite values are from Anders and Grevesse [1989]. The black and gray solid lines indicate the cores and rims of clinopyroxenes, respectively. The four analyzed spinel harzburgite (Sp. Hz.) specimens are shown in the upper and central rows. The analyzed plagioclase-bearing harzburgite (Pl-bg. Hz.) and vein-bearing peridotite (Vein-bg. Peridotite) are shown in the bottom row.

Table 7. Average Trace Element Compositions of Clinopyroxene in the Spinel Harzburgites
Sample No. Clinopyroxene n Tia Sr Y Zr Nb Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Pb Th U
U1382A-8R-4-60-63 core 8 455 0.091 3.563 0.014 5.593b bdlc 0.703 1.162 bdl 11.73 41.30 21.82 175 53.40 488 133 450 72.68 478 74.83 11.55 15.25 bdl 1.000
stdd 29.95 0.042 0.242 0.005 3.861 0.852 0.658 2.371 13.02 2.416 15.30 3.857 35.07 9.134 33.51 6.209 38.99 5.915 2.534 4.114 2.288
rim 8 510 0.181 3.665 0.014 5.517 bdl bdl 1.316 bdl 11.97 38.26 24.85 210 56.71 531 140 460 78.80 512 76.54 13.02 19.47 0.561 4.443
std 84.93 0.116 0.177 0.010 4.155 0.845 5.109 9.367 4.789 28.03 6.386 58.63 9.954 29.76 6.128 27.62 6.555 3.554 6.832 0.736 8.782
U1382A-9R-1-79-83 core 7 463 0.099 4.232 0.023 9.038 bdl 1.323 0.991 bdl 21.48 59.57 36.10 259 72.03 634 161 514 80.13 525 81.81 15.65 14.34 bdl bdl
std 63.28 0.111 0.385 0.004 1.548 2.035 0.600 4.914 13.54 4.987 38.26 9.224 66.82 16.85 51.53 6.515 35.93 6.587 4.690 2.280
rim 6 463 0.310 4.175 0.023 8.351 bdl 4.260 4.387 bdl 19.28 63.57 35.24 258 74.16 627 159 517 81.88 527 80.20 15.55 24.92 bdl 0.564
std 136.70 0.354 0.292 0.003 1.856 6.315 5.969 7.125 22.86 6.397 54.14 14.81 88.32 21.21 44.47 11.11 46.20 4.689 4.165 22.97 0.745
U1382A-9R-1-97-102 core 6 487 0.155 3.999 0.018 7.030 bdl 0.608 bdl bdl 25.51 69.37 37.71 294 75.05 650 157 490 79.30 516 78.85 18.10 19.45 bdl bdl
std 85.50 0.193 0.158 0.007 1.599 1.519 8.239 12.30 4.150 47.87 10.86 65.79 13.84 21.06 6.659 37.29 8.555 4.706 3.214
rim 3 470 0.153 4.206 0.019 6.772 bdl bdl bdl bdl 25.06 70.79 38.73 301 77.84 673 160 509.18 80.76 534 82.63 18.02 21.03 bdl bdl
std 81.48 0.144 0.044 0.001 0.227 3.728 13.15 2.296 42.03 6.284 39.89 8.530 22.35 7.72 26.09 0.990 1.482 3.838
U1382A-9R-1-117-122 core 6 409 0.071 4.031 0.012 12.25 bdl bdl bdl bdl bdl 15.98 57.28 32.26 237 67.99 603 156 504 81.03 525 81.15 14.29 7.842 bdl
std 31.70 0.018 0.144 0.006 6.262 3.781 5.850 1.705 14.72 2.948 17.13 5.967 11.74 1.890 18.45 3.049 2.729 0.765
rim 6 469 0.223 4.146 0.013 12.85 bdl bdl bdl bdl bdl 20.75 64.11 33.91 269 74.23 639 161 525 86.57 546 83.37 13.77 8.600 bdl
std 68.17 0.078 0.170 0.004 4.032 2.536 7.302 4.771 29.56 6.754 45.22 7.807 18.71 5.527 22.37 3.778 1.554 1.952
BCR-2G 10 14276 347 39 196 13028 687048 25978 53280 7024 29599 6770 2036 7044 1080 6727 1384 3911 572 3633 543 5108 10618 6359 1621
std 513 14 3 8 1034 19006 1247 1201 211 1085 232 85 250 48 238 73 176 29 198 22 168 451 237 41
BHVO-2G 16557 396 26 170 18300 131000 15200 37600 5350 24500 6100 2070 6160 920 5280 980 2560 340 2010 279 4320 1700 1220 403
Lower limit of detection (LLD) 0.874 0.018 0.001 0.004 1.212 58.434 0.548 0.907 1.228 3.107 4.870 2.121 7.662 0.965 3.451 0.322 2.435 0.578 2.685 0.771 2.440 3.076 0.463 0.253
Chondrite value by Anders and Grevesse [1989] 436 7.8 1.56 3.94 234.7 603.2 89.1 452.4 147.1 56 196.6 36.3 242.7 55.6 158.9 24.2 162.5 24.3
  • a From Ti to Zr element concentrations are given in ppm an from Nb to U are also given in ppb.
  • b Suspicious data due to seawater contamination are shown in the italics.
  • c bdl is data below lower limit of detection.
  • d std Standard deviation (1 sigma).
Table 8. Trace Element Compositions of Clinopyroxene in the Plagioclase-Bearing Harzburgite
Analysis No. Clinopyroxene Tia Sr Y Zr Nb Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Pb Th U
8R4-72-76-02 core 611 0.50 3.64 0.057 36.84b bdlc 17.23 26.53 3.25 18.02 51.67 26.01 242 75.34 581 153 480 77.52 546 77.03 13.34 10.52 2.06 1.43
8R4-72-76-03 451 1.17 5.65 1.49 75.31 bdl 118 408 60.68 266 138 61.44 325 89.10 760 209 663 107 780 122 18.08 11.29 8.39 1.99
8R4-72-76-05 522 0.57 5.49 1.64 39.46 bdl 58.49 207 36.87 196 138 59.38 340 96.04 796 206 648 104 699 104 22.77 8.53 2.32 1.27
8R4-72-76-10 535 0.44 4.72 0.90 65.40 bdl 71.48 207 28.96 139 97.24 49.65 319 88.91 724 175 550 102 634 91.26 16.74 12.28 5.87 1.74
8R4-72-76-11 432 0.19 4.17 0.17 60.97 bdl 32.28 79.82 8.90 41.76 59.01 28.85 237 67.22 580 145 493 75.54 550 89.47 10.78 9.61 5.18 2.43
8R4-72-76-14 578 0.24 3.91 0.102 32.58 bdl 14.48 25.58 2.31 26.16 58.44 28.33 260 69.89 593 149 466 81.41 547 82.19 16.26 13.58 1.48 1.07
8R4-72-76-21 660 0.23 3.79 0.028 33.78 bdl 3.14 3.61 1.23 32.39 90.35 40.14 293 75.35 581 138 460 75.23 488 68.85 17.67 8.42 1.23 1.59
8R4-72-76-24 521 0.73 4.12 0.71 106 bdl 154 399 54.79 217 107 48.88 300 74.74 652 148 495 76.83 510 79.78 15.39 8.06 10.45 3.87
8R4-72-76-01 rim 637 1.89 7.97 5.42 60.19 bdl 178 805 151 910 406 149 719 154 1229 292 919 133 925 126 135 12.85 6.38 1.88
8R4-72-76-04 775 1.55 7.90 5.36 53.51 bdl 185 648 135 766 433 138 721 170 1233 290 893 145 875 122 142 10.43 5.48 1.57
8R4-72-76-06 471 0.86 5.63 1.88 55.67 96.57 73.06 296 51.12 253 136 63.19 357 95.59 780 203 673 98.97 734 105 22.87 26.07 4.42 1.98
8R4-72-76-09 670 1.03 6.74 4.35 65.01 bdl 210 777 151 862 388 128 686 150 1125 262 796 121 771 112 146.42 13.25 5.15 2.37
8R4-72-76-12 585 0.97 6.91 4.68 76.30 bdl 200 865 163 870 343 121 627 131 1028 235 786 126 770 119 106.05 11.11 7.34 2.18
8R4-72-76-13 499 0.79 5.22 2.04 85.52 bdl 139 534 83.06 407 161 74.47 344 89.64 782 186 609 99.74 683 95.96 36.25 90.47 9.18 3.15
8R4-72-76-15 792 2.15 7.19 5.53 75.24 203.11 354 1090 221 1180 411 149 773 165 1183 263 766 137 797 112 206 29.48 6.04 2.74
8R4-72-76-16 432 0.080 4.79 0.70 34.26 bdl 24.40 72.54 10.08 41.83 77.18 35.21 246 67.16 640 174 570 86.80 624 97.50 14.23 7.23 3.96 3.13
8R4-72-76-17 483 0.16 5.17 1.61 32.31 bdl 26.22 93.74 17.88 111 101 48.31 300 86.42 738 188 598 100 626 97.54 17.37 5.31 2.88 2.23
8R4-72-76-22 520 0.73 4.42 0.49 104 bdl 107 291 33.67 140 91.29 45.08 307 67.57 613 160 506 84.43 539 88.03 19.33 12.81 6.65 2.90
8R4-72-76-23 532 1.06 5.08 1.14 128 bdl 238 701 95.39 364 142 63.00 341 83.06 728 187 590 99.16 641 92.99 17.27 8.99 12.18 4.16
  • a From Ti to Zr element concentrations are given in ppm an from Nb to U are also given in ppb.
  • b Suspicious data due to seawater contamination are shown in the italics.
  • c bdl is data below lower limit of detection.
Table 9. Trace Element Compositions of Clinopyroxene in the Vein-Bearing Peridotite
Analysis No. Clinopyroxene Tia Sr Y Zr Nb Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Pb Th U
9R1-71-74-02 core 2652 9.00 15.93 9.16 57.40b 168 380 1197 325 2071 1237 455 2100 423 2851 606 1701 265 1515 218 375 12.71 4.68 1.86
9R1-71-74-04 2752 9.13 16.57 10.36 58.09 bdlc 256 1228 314 2187 1194 440 2051 410 2915 618 1767 260 1513 218 406 12.87 4.60 1.28
9R1-71-74-05 2931 10.45 20.18 12.08 96.96 194 448 1459 374 2508 1430 540 2363 537 3496 805 2220 345 1997 292 445 8.69 7.18 2.41
9R1-71-74-08 2642 10.24 14.98 9.30 47.94 98 260 950 251 1766 1056 427 1972 399 2582 577 1645 250 1485 219 392 7.75 3.95 1.61
9R1-71-74-09 2052 9.57 13.19 6.96 39.07 128 248 825 220 1464 912 378 1721 350 2357 508 1426 230 1336 199 262 18.50 2.11 1.90
9R1-71-74-10 2554 9.35 15.95 10.37 35.84 163 362 1173 316 2140 1244 464 2115 424 2901 612 1681 269 1567 231 392 7.11 1.37 1.50
9R1-71-74-11 4197 11.65 25.72 19.81 119 418 726 2053 554 3509 2102 672 3368 706 4408 980 2619 406 2186 324 728 693 11.34 3.50
9R1-71-74-14 2873 10.09 19.09 10.70 96.17 203 437 1423 366 2358 1441 529 2389 507 3266 716 2034 309 1834 268 401 11.57 7.67 3.16
9R1-71-74-17 536 0.62 4.19 0.19 15.96 bdl 23.92 79.65 14.15 72.74 79.49 45.96 275 75.60 597 149 470 88.06 542 81.60 23.58 10.73 1.05 0.53
9R1-71-74-20 557 2.34 7.84 2.20 61.91 bdl 182 703 129 656 286 141 550 135 1079 277 922 160 1058 157 23.57 10.71 5.05 2.20
9R1-71-74-21 3366 10.10 18.79 14.78 77.93 127 444 1701 426 2868 1579 518 2592 481 3228 682 1912 283 1608 232 625 18.10 4.70 2.61
9R1-71-74-22 4161 11.30 23.04 16.79 94.40 190 529 1873 492 3452 1921 618 3177 634 4116 873 2362 346 2008 281 719 15.98 7.11 2.94
9R1-71-74-24 3834 10.36 23.46 12.47 54.40 205 447 1597 440 3052 1818 629 3084 632 4136 884 2503 387 2183 320 542 12.55 3.34 2.07
9R1-71-74-26 3524 10.02 21.05 11.46 50.63 245 494 1436 402 2710 1729 598 2805 603 3773 801 2148 349 1925 273 483 5.44 3.86 1.85
9R1-71-74-28 2684 9.74 18.71 10.66 69.58 262 405 1474 355 2354 1255 513 2178 470 3148 695 1981 300 1893 265 393 177 4.82 2.32
9R1-71-74-30 4318 9.40 24.79 20.00 76.92 bdl 317 1887 518 3746 1978 664 3360 647 4250 939 2688 379 2224 309 946 8.94 4.33 1.80
9R1-71-74-31 3346 9.33 21.24 14.27 62.17 bdl 312 1527 399 2930 1618 553 2781 546 3760 797 2301 335 1984 284 591 10.69 4.93 2.41
9R1-71-74-34 3082 10.11 19.03 12.13 76.33 60.23 314 1531 383 2748 1526 528 2446 485 3313 713 2000 286 1706 248 547 16.19 8.41 3.20
9R1-71-74-35 4135 11.43 23.36 20.96 131 139 512 2018 517 3382 1866 612 3183 599 4115 856 2337 355 1926 298 941 21.49 9.25 3.08
9R1-71-74-01 rim 3175 10.42 20.03 12.20 82.25 437 646 1471 435 2653 1598 550 2754 574 3710 815 2099 359 1830 261 454 11.51 7.15 2.94
9R1-71-74-03 3043 12.26 18.59 12.78 92.98 145 421 1597 381 2552 1417 522 2353 486 3339 731 1954 294 1673 243 454 14.43 9.00 3.86
9R1-71-74-06 3444 9.44 21.72 14.13 106 217 479 1657 425 2868 1691 606 2907 599 3800 817 2265 350 1923 280 600 18.60 9.06 2.93
9R1-71-74-07 2553 8.32 16.42 10.03 65.92 bdl 222 1007 250 1840 1072 437 1953 402 2818 621 1775 255 1579 227 419 17.79 4.86 1.56
9R1-71-74-12 3840 9.64 24.84 20.05 121 181 539 1970 520 3368 1888 610 3216 624 4241 905 2558 385 2231 321 686 13.06 7.32 2.59
9R1-71-74-13 3018 9.49 21.31 15.37 105 96.54 456 1853 471 3088 1588 546 2677 515 3561 780 2156 339 2024 292 472 15.95 10.40 2.77
9R1-71-74-15 3408 10.37 20.63 10.60 57.93 215 404 1308 375 2561 1588 542 2751 570 3589 795 2085 328 1902 285 503 12.54 7.19 2.70
9R1-71-74-16 937 3.54 6.76 2.08 49.99 166 255 571 125 584 370 149 629 163 1096 265 774 153 871 115 24.34 12.36 3.85 1.68
9R1-71-74-19 3307 8.20 20.23 12.75 51.33 197 389 1344 361 2551 1547 546 2647 560 3448 751 2117 350 1926 264 568 10.07 3.51 1.46
9R1-71-74-23 2766 10.49 16.82 10.06 58.43 bdl 240 1147 292 2137 1221 452 2016 414 2799 601 1766 269 1612 226 437 239 2.86 1.09
9R1-71-74-25 3241 8.28 21.29 12.05 45.99 bdl 236 1512 383 2832 1599 559 2664 536 3601 760 2243 304 1906 259 521 7.68 3.85 1.05
9R1-71-74-27 3525 8.89 21.84 14.27 57.09 60.86 319 1652 444 3030 1722 618 2816 552 3746 816 2331 329 1947 279 616 5.93 3.94 1.86
9R1-71-74-29 3709 8.43 21.91 14.46 50.22 90.58 341 1553 424 2973 1692 588 2824 557 3819 816 2262 334 1995 284 607 10.40 3.79 1.57
9R1-71-74-32 3608 10.02 22.96 15.01 65.73 115 403 1704 454 3138 1778 599 3103 609 4119 841 2376 350 1993 292 609 11.99 5.63 2.45
9R1-71-74-33 3671 9.11 21.81 16.42 66.83 bdl 362 1799 464 3328 1759 583 2903 570 3918 817 2296 324 1937 275 696 12.39 4.73 2.62
9R1-71-74-39 2931 7.22 19.04 15.86 72.98 78.00 316 1580 395 2774 1455 490 2480 486 3344 715 2063 297 1749 254 617 10.38 4.35 2.34
9R1-71-74-40 3410 8.52 19.70 16.57 77.01 bdl 307 1779 427 3063 1635 550 2777 513 3518 751 2086 283 1765 252 707 15.86 6.70 2.38
  • a From Ti to Zr element concentrations are given in ppm an from Nb to U are also given in ppb.
  • b Suspicious data due to seawater contamination are shown in the Italics.
  • c bdl is data below lower limit of detection.

Nb, Ba, Pb, Th, and U contents were also determined quantitatively for all the peridotite samples, and the results are presented in Tables 7-9. Among these, the high field strength elements (HFSE), especially Nb, are all above the LLD. The Ba, Pb, Th, and U contents are usually above the LLD, but the contents of these elements may include the effects of seawater contamination along the cleavage planes of the clinopyroxenes [e.g., Kimura and Sano, 2012]. The sampled peridotites all come from the sedimentary breccia zone (Figure 2) in the North Pond, and this zone is interpreted to be a site of recharging hydrothermal circulation [Langseth et al., 1992; Edwards et al., 2014], which means these samples could easily have been contaminated. The elements (Nb, Ba, Pb, Th, and U) have therefore been discarded from the following discussions.

5.2.1 Spinel Harzburgite

Clinopyroxenes in the four specimens of spinel harzburgite show simple patterns with a steep slope from the heavy REEs (HREEs) to the middle REEs (MREEs) (Figure 7). The core and rim of any one grain commonly show similar trace-element contents (Figure 7). All the clinopyroxenes in the four spinel harzburgites exhibit negative anomalies of Ti and Zr relative to REEs, and have lower Ti and Zr contents than clinopyroxenes in the plagioclase-bearing harzburgite and vein-bearing peridotite (Figures 7 and 8a). The spinel harzburgites have a concentrated area of extremely low Zr contents when plotted on Figure 8a.

Details are in the caption following the image

Compositional variations of the clinopyroxenes in the spinel harzburgites (Sp. Hz.), the plagioclase-bearing harzburgite (Pl-bg. Hz.), and the vein-bearing peridotite (Vein-bg. Peridotite) from Hole U1382A in North Pond. (a) Ti concentrations (in ppm) versus Zr concentrations (in ppm). (b) (Sm/Yb)N ratios versus YbN contents. (c) Average Cr# values of the spinels versus the YbN contents of the clinopyroxenes. The gray area is the field of residual abyssal peridotites from Hellebrand and Snow [2003].

5.2.2 Plagioclase-Bearing Harzburgite and Vein-Bearing Peridotite

The clinopyroxenes in the plagioclase-bearing harzburgite and the vein-bearing peridotite have a wide range of trace element compositions (Figure 7). The light-REEs (LREEs) in the clinopyroxenes of the plagioclase-bearing harzburgite shift from low abundances ((Sm/Yb)N = 0.10–0.23) in the cores to high abundances in the rims ((Sm/Yb)N = 0.14–0.57) (Figures 7 and 8b). These clinopyroxenes have negative Ti, Zr, and Sr anomalies (Figure 7). In addition, the trace element compositions of the clinopyroxenes show a slight Eu negative anomaly in the vein-bearing peridotite, and the rims of the clinopyroxenes in the plagioclase-bearing harzburgite show a similar anomaly (Figure 7). In the plagioclase-bearing harzburgite the clinopyroxenes display almost constant Ti abundances whereas the Zr contents increase from cores to rims (Figures 7 and 8a).

The clinopyroxenes in the vein-bearing peridotite show a gentle slope from HREEs to LREEs, and have relatively constant abundances of HREEs ((Sm/Yb)N = 0.16–1.08) (Figure 7). The core and rim in any one grain of clinopyroxene show similar trace-element compositions. These clinopyroxenes also have negative Ti, Zr, and Sr anomalies (Figure 7). Nevertheless, it is generally the clinopyroxenes in the vein-bearing peridotite that have the highest Ti and Zr abundances (Figure 8a).

6 Discussion

Here, we discuss the origin of the spinel harzburgite, plagioclase-bearing harzburgite, and vein-bearing peridotite from the North Pond of the MARK area, and we base our discussion on the fabric analyses and the major and trace element analyses of the minerals. First we will discuss the geological origin of the peridotite samples, and then we will examine the petrogenesis of the depleted harzburgites and the process of refertilization by melt–rock interactions during which process the other two peridotite lithologies were formed. Our penultimate discussion will concern the significance of the olivine fabrics in these peridotites. Finally, by bringing all the petrochemical and textural data together, the origin of the North Pond peridotites will be discussed.

6.1 Origin of the Peridotites at the North Pond

IODP Expedition 336 sank drillholes at three new sites (U1382, U1383, and U1384) in the northern and southern parts of the North Pond [Expedition 336 Scientists, 2012; Edwards et al., 2014] (Figure 1b). Although several kinds of basalt were encountered at all three sites [Expedition 336 Scientists, 2012; Edwards et al., 2014], only two sites (U1382 and 395) in the southern area of the North Pond provided samples of gabbros and peridotites, which were present as clasts in sedimentary breccias [Melson et al., 1978; Expedition 336 Scientists, 2012; Edwards et al., 2014].

Previous expeditions of ODP Leg 174B drilled Hole 1074A, located 4.5 km to the northwest of Hole 395A (Figure 1b), and they used Hole 395A at the North Pond for logging operations, packer testing, and borehole fluid sampling, where they identified sedimentary units containing nannofossil ooze with varying amounts of foraminifera, clay, radiolaria, and sand, as well as clays and zones of aphyric basalt [Becker et al., 1998a, 1998b]. Sites U1383 and U1384 in the northern area of the North Pond (Figure 1b) also revealed a basaltic basement, but no gabbros or peridotites [Expedition 336 Scientists, 2012].

Given that gabbros and peridotites have only been found at Hole 395 in the southern area of the North Pond as clasts in a debris flow deposit [Melson et al., 1978; Expedition 336 Scientists, 2012; Edwards et al., 2014], our peridotite specimens from Hole U1382A were almost certainly derived by mass wasting from the southern area of the North Pond. Furthermore, pebble-sized peridotites and gabbroic clasts were obtained from Hole U1382B in sediments that lie between the surface and the basement [Expedition 336 Scientists, 2012; Edwards et al., 2014], which suggests that future debris flows might continue to supply clasts from the outcrops around the North Pond.

With regard to the southern area of the North Pond, Cann et al. [2015] gathered together the existing geophysical (bathymetric, gravity, and magnetic) data for the MARK area in order to reconstruct the history of accretion. As a result, they identified the massif in the southern area of North Pond as an oceanic core complex at the end of an extinct spreading segment (corresponding to segment 7W––Cann et al., [2015]), indicating that this is likely related to a lower magma supply at this end of the segment [cf. Cann et al., 2015, Figures 1, 7, and 8]. Therefore, the clasts of peridotite in the North Pond area were most probably derived from the oceanic core complex that is exposed in the southern part of the Pond, which suggests that the collapse of the oceanic core complex caused an inflow of many samples of gabbro and peridotite.

6.2 Depleted Harzburgites and Melt-Rock Interactions

6.2.1 Spinel Harzburgite as Residual Ridge Mantle in Origin

The four samples of spinel harzburgite from the North Pond have features that are comparable to those of refractory abyssal peridotites, including the mineral assemblages, the intermediate Cr# values and low TiO2 contents of the spinels (Figure 6d), the extremely low Na and Ti abundances of the clinopyroxenes (Figure 6b), and the negative Ti and Zr anomalies and steep slopes from the HREEs to MREEs in the clinopyroxene trace element patterns (Figure 7). Constantin et al. [1995] reported low Na contents (average Na2O = 0.06%) in the clinopyroxenes of harzburgite, and they suggested this indicates a pure mineral residue of melting without any refertilization. Similarly depleted clinopyroxene compositions have also been reported form ultra-depleted harzburgites in both abyssal and ophiolite peridotites (e.g., sample BH5 from Site U1309 of the Atlantis Massif [Tamura et al., 2008] and the MOR ophiolite of the Red Hills in New Zealand [Sano and Kimura, 2007]). The geochemical features of our North Pond samples of spinel harzburgite correspond to those of residual peridotites formed by simple partial melting of the upper mantle.

Hellebrand et al. [2001] proposed that the nature of the correlation between HREE abundances in clinopyroxenes and Cr# values in spinels can indicate the degree of partial melting related to residual abyssal peridotites. Abyssal peridotites show negative correlations between these parameters (Figure 8c), and the analyzed North Pond spinel harzburgites plot in the same field (Figure 8c). The lower Zr and Ti contents in spinel harzburgites also indicate a melting trend [Johnson et al., 1990] and the North Pond spinel harzburgites have similarly low contents (Figure 8a). These trace element features are also consistent with the proposal that the spinel harzburgites are residual peridotites.

The degree of melting (F) is estimated to be 13.1–14.2% assuming a depleted Mid-Ocean Ridge Basalt (MORB) mantle (DMM) as a source, and using the Cr# of the spinels according to the formula F = 10ln(Cr#) + 24, as proposed by Hellebrand et al. [2001, Figure 3]. Hellebrand et al. [2001] also showed that the most depleted peridotite in the MARK area corresponds to 15% melting. Since the F calculation accounts only for melting in the spinel stability field, additional studies on melting in the garnet field are required [e.g., Hellebrand et al., 2002; Barth et al., 2003; Tamura et al., 2008].

The extremely LREE-depleted clinopyroxenes in the spinel harzburgites (Figure 7) are traditionally interpreted as originating from the fractional melting of a peridotite that was present in both the garnet and spinel peridotite fields [Johnson et al., 1990; Hellebrand et al., 2002]. Model calculations using REEs are useful for examining the melting processes through a range of depths. We have calculated the behavior of REEs during melting of a DMM [Workman and Hart, 2005] using the adiabatic melting model Ocean Basalt Simulator ver. 1 (OBS1) [Kimura and Kawabata, 2015]. The OBS1 model combines the thermodynamics of melting of a dry to water-bearing peridotite under adiabatic conditions, the petrological modal changes during adiabatic melting, and the elemental partitioning between the melts and the residual mantle at different depths [Kimura and Kawabata, 2015].

In the calculations, we assumed water in the DMM (XH2O) to be 300 ppm and the mantle potential temperature (Tp) to be 1350°C (Figure 9a). With these conditions, melting first occurs at P = 3.7 GPa in the garnet field. The garnet–spinel transition occurs at 2.5 GPa, and the lherzolite–harzburgite transition (clinopyroxene-out) occurs at 1.2 GPa (Figure 9b). The degree of melting is 4% at the garnet–spinel transition, but this increases to 15% at the lherzolite–harzburgite transition (clinopyroxene-out), and reaches 17% at ∼1 GPa (Figure 9b). The composition of the melt generated at 1 GPa (the accumulated melt) corresponds to that of D-MORB [Jenner and O'Neill, 2012] (Figure 9c). The REE compositions of the residual clinopyroxenes are calculated using instantaneous melt compositions for each incremental melting step. This assumption is reasonable because the residual clinopyroxenes finally equilibrated with the instantaneous melts which were eventually extracted from the melting portion, whereas the accumulated melt compositions correspond to the bulk melts that were extracted during the fractional melting [Kimura and Kawabata, 2015].

Details are in the caption following the image

(a–d) Origin of clinopyroxenes in spinel harzburgite and vein-bearing peridotite. Adiabatic fractional melting of a DMM under a mantle potential temperature of Tp = 1350°C leaves various clinopyroxenes at different depth levels. Color-coded thin lines show clinopyroxene compositions with different depth ranges and phase stabilities in the host peridotite. The observed North Pond clinopyroxenes from the spinel harzburgite can form at P = ∼1.0 GPa in the spinel to plagioclase stability field at T = 1340–1320°C and with a degree of partial melting F = 14–17% (consistent with F = 13–14% using the Cr# values of spinel). The melt generated under this condition is a D-MORB. (e) Clinopyroxenes in the vein-bearing peridotite can form under equilibrium crystallization from N-MORB at a temperature of 1180°C. The calculations were performed using Ocean Basalt Simulator ver.1.2 [Kimura and Kawabata, 2015]. D- and N-MORB compositions are from Jenner and O'Neill [2012]. REE concentrations were normalized by primitive mantle (PM) [Sun and McDonough, 1988]. Clinopyroxene data were from the averaged core and rim data in Tables 7–9. Gar, garnet; Cpx, clinopyroxene; Opx, orthopyroxene; Pl, plagioclase.

The results of the calculations show that the North Pond clinopyroxenes in the spinel peridotites could have formed in residual harzburgites at depths equivalent to around 1.0 GPa (Figure 9d). The results also indicate the strong effects of residual garnet from 3.7 to 2.5 GPa, residual clinopyroxene from 2.5 to 1.3 GPa, and residual orthopyroxene at less than 1.3 GPa, during the formation of the spinel peridotites (Figure 9d). The estimated degrees of melting (using spinel Cr# values) are F = 13.1–14.2% (see above). This is consistent with the total amount of incremental melt in the spinel field of ∼13% using the OBS1 model with the preceding ∼4% melting in the garnet field. In petrological mode, our estimates for the spinel harzburgites (Table 1) are compatible with the calculated mineral modes using the OBS1 model, which resulted in a large orthopyroxene mode and a lesser clinopyroxene mode (Figure 9b). Note that the incremental depletion of the source during fractional melting of the DMM was considered, and that the thermodynamic heat balance was internally consistent with the petrological modes, and that the partition coefficients of the REEs are T-dependent based on the lattice strain model and experimental results [Kimura and Kawabata, 2014]. Therefore, the calculated clinopyroxene compositions are internally and quantitatively consistent.

6.2.2 Plagioclase-Bearing Harzburgite and Vein-Bearing Peridotite Originating by Off-Ridge Refertilization

Our sample of plagioclase-bearing peridotite contains plagioclase grains that surround spinel grains and occur between olivine and pyroxene. Plagioclases are also present in the cross-cutting vein of gabbro in the vein-bearing peridotite (Figures 3e–3j). In terms of their major element compositions, the olivines and pyroxenes in the plagioclase-bearing peridotite tend to be iron-rich compared with those in the spinel harzburgites (Figure 6a), the clinopyroxenes have higher contents of Na and Ti than those in the spinel harzburgites (Figure 6b), and the spinels have higher Cr# and TiO2 contents than those in the spinel harzburgites (Figures 6c and 6d).

The high Na2O contents of clinopyroxenes in abyssal peridotites are a result of the refertilization of peridotite by entrapped melts [Elton, 1992; Hellebrand et al., 2002; Tamura et al., 2008]. Hellebrand et al. [2002] also concluded that melt–rock reactions increase LREE concentrations much more effectively than MREE and HREE concentrations, because LREE concentrations in the residual peridotite are extremely low. The plots of YbN in the clinopyroxenes of our two samples of plagioclase-bearing peridotite versus the Cr# values in spinels of the spinel harzburgites are clearly off the trend of partial melting for residual abyssal peridotites (Figure 8c), indicating that these samples were possibly produced by melt–rock interactions that overprinted the depleted residual spinel harzburgite. These petrological features are also common in plagioclase-bearing peridotites at mid-ocean ridges, where they have been interpreted to be the result of impregnation by trapped or transient MORB-like melts [Dick and Bullen, 1984; Dick, 1989; Seyler and Bonatti, 1997; Taratrotti et al., 2002; Dick et al., 2010].

The REE patterns of the clinopyroxenes in the vein-bearing peridotite show a gentle slope from HREEs to LREEs, and relatively constant values of HREEs from cores to rims. These are similar to the REE patterns of the clinopyroxene rims in the plagioclase-bearing harzburgites (Figure 7). The patterns of these clinopyroxenes that are enriched in REEs are similar to those of the gabbros in the MARK area [Coogan et al., 2000a, 2000b]. The microstructures of the vein-bearing peridotite include poikilitic textures in the gabbroic vein (Figure 3i). The vein-bearing peridotite contains more plagioclase and clinopyroxene than the plagioclase-bearing harzburgite (Table 1). The Zr and Ti contents of the vein-bearing peridotite are higher than the spinel harzburgites and the plagioclase-bearing harzburgite (Figure 8a). Furthermore, the values of (Sm/Yb)N and YbN in the vein-bearing peridotite are higher than in the plagioclase-bearing harzburgite (Figure 8b). Hellebrand et al. [2002] also noted that a large amount of melt (>1%) was needed to significantly increase YbN in both depleted and fertile peridotites. These observations suggest that the formation of the vein-bearing peridotite took place after the formation of the plagioclase-bearing harzburgite. A progressive melt–rock interaction between the plagioclase-bearing harzburgite and melt related to the intruded vein of gabbroic material could have resulted in the formation of the vein-bearing peridotite. The differences between the plagioclase-bearing harzburgite and the vein-bearing peridotite can be accounted for by differences in the amount of melt that interacted with the spinel harzburgite.

The clinopyroxenes in the vein-bearing peridotite were precipitated simply from N-MORB melts. The equilibrated clinopyroxene compositions were calculated from N-MORB basalt compositions [Jenner and O'Neill, 2012] using the T-dependent partition coefficients of clinopyroxene and melt [Kimura and Kawabata, 2015]. To reproduce the clinopyroxene compositions found in the North Pond vein-bearing peridotite, the results show that the best fit occurs at T = 1180°C (Figure 9e). Slight negative anomalies of Eu in the North Pond samples are the effect of co-existing plagioclase that is not assumed in the calculations.

6.3 Fabrics and Tectonic Implications

6.3.1 Vp-Flinn Diagram

We found it difficult to identify the type of olivine fabric in the spinel harzburgites and the plagioclase-bearing harzburgite due to the weakness of the CPOs and the lack of a kinematic framework such as foliation and lineation. Thus, we applied the fabric-index angle (FIA) method with a Vp-Flinn diagram to our fabric analyses, and used the P-wave properties of olivines following the method described by Michibayashi [2015] and Michibayashi et al. [2016]. We calculated three P-wave velocities (V1, V2, and V3) and analyzed the anisotropy of the olivine in the peridotite samples by plotting V2/V3 on the horizontal axis and V1/V2 on the vertical axis of a Flinn diagram, with the origin set at (1, 1) [Michibayashi, 2015; Michibayashi et al., 2016]. The angle of inclination between the point of origin (1, 1) and a point on the Flinn diagram is defined by K, as follows:
urn:x-wiley:15252027:media:ggge21036:ggge21036-math-0001
Furthermore, because it is inconvenient to use K (which commonly has a value of 150), as used in structural geology, Michibayashi [2015] and Michibayashi et al. [2016] defined a new measure, the fabric-index angle (FIA: θ):
urn:x-wiley:15252027:media:ggge21036:ggge21036-math-0002

A set of CPOs can be expressed as a single FIA angle in a range between 0° and 90° on the Flinn diagram, where 0° is the AG-type, 90° is the D-type, and 63° is a single-maximum CPO pattern such as an A-type. Our calculated results for the spinel harzburgites and the plagioclase-bearing harzburgite gave a range of FIA between 21° and 77° (Figure 10 and Table 1), indicating that there are two fabric types in the analyzed samples. Two spinel harzburgites are situated at a midway point between A- and D-types. The olivine fabrics of the three samples (two spinel harzburgites and the plagioclase-bearing harzburgite) are distributed between AG- and A-types (Figure 10). One spinel harzburgite sample in particular (U1382A-9R-1-117-122) corresponds to the AG-type (Figure 10). Thus, the analyzed spinel harzburgites and the plagioclase-bearing harzburgite from the North Pond are characterized by [010] fiber patterns (i.e., AG-type), (010)[100] patterns (i.e., A-type), and {0kl}[100] patterns (i.e., D-type) (Figure 10 and Table 1).

Details are in the caption following the image

Vp-Flinn plot for the olivine fabrics. The broken lines are Vp anisotropy values defined by V1 and V2. FIA = Fabric-Index Angle.

6.3.2 Fabric Development in the Uppermost Mantle

The Hilti mantle section of the Oman ophiolite provides an exposure of typical lithospheric mantle beneath the crust–mantle boundary area, and it shows that horizontal flow took place in the uppermost mantle away from a diapir below the oceanic spreading center [e.g., Ildefonse et al., 1995; Michibayashi et al., 2000]. The deformed high-temperature peridotites in the Hilti mantle section are dominated by AG-type olivine fabrics, but with a few A-types [Michibayashi and Mainprice, 2004; Michibayashi et al., 2016].

Achenbach et al. [2011] reported A-type olivine fabrics in the peridotites from ODP Leg 209 Hole 1274A in the Fifteen–Twenty Fracture Zone area of the MAR, and discussed them in terms of a mantle flow model beneath a spreading axis. They concluded that the olivine (010)[100] and orthopyroxene (100)[001] fabrics are consistent with crystal-plastic deformation related to mantle flow at high temperatures (∼1250°C) beneath a mid-ocean ridge.

Our data (Figures 5 and 10) are similar to those reported for the Hilti mantle section of the Oman ophiolite, and also similar to some abyssal peridotites at the MAR, as noted above. Our specimens of spinel harzburgite and plagioclase-bearing harzburgite exhibit a protogranular texture and show evidence of weak intracrystalline deformation; there is no obvious foliation and lineation (Figure 3). The grain sizes of each constituent mineral and the microstructure in our peridotites appear to be similar to those described by Achenbach et al. [2011] and Michibayashi and Mainprice [2004]. It is reasonable, therefore, to assume that the olivine fabrics observed in our specimens of spinel harzburgite and plagioclase-bearing harzburgite might represent the nature of the mantle flow below a spreading axis. Although the plagioclase-bearing harzburgite was formed by the impregnation of melts into the spinel harzburgite, as noted above, the microstructures of the plagioclase-bearing harzburgite are similar to those of the spinel harzburgite (Figure 3).

The olivine fabrics of the plagioclase-bearing harzburgite and spinel harzburgites are similar, and are distributed between AG- and A-type fabrics. The olivine fabric of the plagioclase-bearing harzburgite (Table 1 and Figure 5) may therefore simply represent the fabric of a spinel harzburgite that was preserved despite static interactions with minor amounts of infiltrating melt. In contrast, the vein-bearing peridotite has an indistinct CPO compared with the other peridotites (Table 1 and Figure 5), and since this peridotite would have been subjected to progressive melt–rock interactions, as discussed above, the indistinct fabric may in this case be the result of the reconstruction of the mineral fabric during those interactions.

6.4 Tectonomagmatic History of the North Pond Extinct Segment of the MARK Area

In the MARK area, Cann et al. [2015] showed that the 1W segment [see Cann et al., 2015, Figures 1, 7, and 8] near the Kane Fracture Zone shows evidence of detachment faulting along the entire western side of the median valley from north to south, revealing the existence of oceanic core complexes that include the Kane oceanic core complex [Dick et al., 2008, 2010] and Sites 920–924 of ODP drilling Leg 153 [Cannat et al., 1995a]. Stephens [1997] also documented at Site 920 the presence of both residual and refertilized peridotites. In the Kane oceanic core complex, Dick et al. [2010] provided evidence for local melt accumulation in the shallow mantle near the base of the crust, and for plural refertilization processes such as melt–rock reactions, dunite formation, and melt impregnation.

Our study has shown that the spinel harzburgites preserve mantle flow structures generated beneath the MAR. These could have formed in the residual spinel harzburgite at high temperatures along the ridge axis (T = 1320–1340°C, P = 1.0–0.8 GPa) after an intensive D-MORB-like melt extraction with mantle depletion by removing a ∼17% melt component from the source DMM peridotite. Later, the residual mantle left the ridge axis, but the temperature was still high enough (T = ∼1180°C) to allow the infiltration of channel melt flows of N-MORB melt to form plagioclase-bearing harzburgite and vein-bearing peridotite with increasing amounts of melt (i.e., refertilization). This refertilization could have occurred under extinct (off-) ridge conditions. In this case, the refertilization took place with N-MORB-like melts so that the location was not that far from the ridge axis. Furthermore, the initial olivine fabrics that were formed at the ridge axis were preserved under this temperature condition, and the indistinct fabric in the vein-bearing peridotite may be due to the deconstruction of these mineral fabrics during the process of refertilization.

The preliminary results of Expedition 336 Scientists [2012] indicated the presence of mylonites in the gabbroic rocks from Hole U1382A, and these mylonites could be related to the development of the oceanic core complex. Since our specimens of peridotite do not display mylonitic or cataclastic deformation, we consider that they were unaffected by the detachment faulting related to the core complex. Dick et al. [2008] mentioned that the crustal and mantle rocks in the oceanic core complex directly constrained the crustal architecture and the pattern of melt-flow from the mantle to the lower crust and within the lower crust. We suggest, therefore, that our peridotite samples preserve structures that could represent mantle flow and melt–rock interactions that occurred below the spreading axis when it was active ∼8 Ma ago. Previous studies of ophiolites of mid-ocean ridge origin have included an examination of the refertilization processes that involved residual harzburgites beneath an ancient ridge [e.g., Kelemen, 1990; Kelemen et al., 1990, 1995; Bedini et al., 1997; Sano and Kimura, 2007]. The lithologies in North Pond are similar to the Red Hills ophiolite with its matrix of harzburgite (with ultra-depleted clinopyroxenes) and its channels of dunite or dikes of gabbro (with enriched clinopyroxenes from MORB melts) [Sano and Kimura, 2007; Kimura and Sano, 2012]. The peridotites in the North Pond sedimentary breccia were derived from similar sources.

Refertilization processes such as melt penetration into residual peridotites have also been documented at Site U1309 in the Atlantis Massif, which is one of the typical oceanic core complexes located at MAR 30°N [Tamura et al., 2008]. The analyzed peridotites from the two holes at Site U1309 in the Atlantis Massif were respectively in contact with the gabbro sampled at Hole 1309B, and surrounded by the gabbro sampled at Hole 1309D [Tamura et al., 2008]. The formation of the Atlantis Massif is interpreted to have involved the emplacement of a large intrusive gabbro body into a host made predominantly of peridotite, and that the intrusive gabbro resulted from an enhanced supply of melt from the underlying mantle, and/or enhanced melting at the spreading center and transport of the melt along the ridge axis [Ildefonse et al., 2007; Blackman et al., 2011]. Tamura et al. [2008] concluded that the initial peridotites in the Atlantis Massif were possibly derived from the upper mantle, and that they were subsequently refertilized during the emplacement of the large intrusive gabbroic body into the underlying mantle. Moreover, it was suggested that the injected melt might have been generated at the ridge-segment center, and that it subsequently moved and evolved toward the segment end beneath the oceanic core complex.

The mutual relationships of these lithologies are unclear with respect to the North Pond reworked breccia samples, and we cannot discuss any direct relationship to the emplacement of a large intrusive gabbroic body into the underlying mantle. Cann et al. [2015] implied that the magmatic activity of the ridge axis during the period 7–10 Ma in the southern parts of the MARK area took place close to that of the current ridge axis. In addition, for the Kane oceanic core complex, Dick et al. [2008] revealed that melt flow from the mantle was focused in local magmatic centers, creating plutonic complexes within the ridge segment in positions that varied in space and time, rather than being fixed at a single central point. Therefore, the extinct ridge segment around the North Pond might possibly have developed not only in relation to the along-axis melt transport from the segment center to the segment end, but also in relation to the creation of plutonic complexes at localized centers.

7 Conclusions

We analyzed six peridotites collected from the sedimentary breccia zone in the North Pond at Site U1382A of the IODP Expedition 336, in order to investigate the histories of deformation and magmatism at an extinct spreading ridge. The six peridotite samples were classified as spinel harzburgite, plagioclase-bearing harzburgite, and vein-bearing peridotite.

Major and trace element compositions of the constituent minerals indicate that the spinel harzburgites are refractory peridotites, similar to those observed in the MARK area. The estimated depletion in the source DMM peridotite caused by D-MORB-like melting, and using the adiabatic melting model OBS1, was ∼17%. The residual spinel harzburgites indicate that melt depletion occurred in the uppermost mantle during active spreading at ∼8 Ma. The plagioclase-bearing harzburgite and the vein-bearing peridotite formed by melt–rock interactions between the depleted spinel harzburgites and the percolating N-MORB-like melts, and they show evidence of refertilization. The differences between the plagioclase-bearing harzburgite and the vein-bearing peridotite can be explained by differences in the amount of melt that interacted with the spinel harzburgite. Therefore, the plagioclase-bearing harzburgite and vein-bearing peridotite provide evidence for progressive melt–rock interactions after the formation of residual spinel harzburgites in the upper mantle.

We evaluated the character of the olivine fabrics in the analyzed peridotites, which are characterized generally by weak CPOs and the lack of a kinematic framework such as foliation and lineation. We were able to recognize several olivine fabric types (AG-, A-, and D-types) in the spinel harzburgites and the plagioclase-bearing harzburgite with a range of fabric-index angle values between 21° and 77°, whereas the vein-bearing peridotite lacks any obvious olivine CPO. The olivine fabrics observed are common in abyssal peridotites from Mid-Ocean Ridges, and they may record flow deformation in the mantle beneath the spreading axis. The fabrics are preserved in the plagioclase-bearing harzburgite, even though it was subjected to impregnation by melts. On the other hand, in the vein-bearing peridotite, the more extensive impregnation by melts left only indistinct CPOs for all the constituent minerals. These observations indicate that there was a progressive reconstruction of the mineral fabrics during the progressive impregnation by melts.

The southern area of the North Pond is adjacent to the oceanic core complex that lies at the extinct end of the segment. Therefore, the peridotite fragments in the North Pond sedimentary breccia of Hole U1382A were probably derived from the oceanic core complex at the end of the extinct ridge segment by a process of mass wasting (i.e., debris flow deposits). None of the peridotite samples exhibited obvious macroscopic evidence of shear deformation, indicating that these peridotites were not affected by the detachment faulting in the oceanic core complex. Although the mutual relationships between lower crust and mantle are unclear for the North Pond reworked breccia samples, our peridotite samples preserve structures in the MARK area that could represent mantle flow and melt–rock interactions beneath the spreading axis at ∼8 Ma. Moreover, such an extinct ridge segment around the North Pond might possibly have developed as a result of the along-axis transport of melts from the segment center to the segment end, and also in relation to the creation of plutonic complexes at localized centers along the axis, similar to the situation along the present slow-spreading axis.

Acknowledgments

For this research we used samples and data provided by the Integrated Ocean Drilling Program (IODP) and the shipboard parties of Expedition 336. We thank K. J. Edwards, W. Bach, A. Klaus, the shipboard scientists, the captain and crew of R/V Joides Resolution, and the technical staff of the Integrated Ocean Drilling Program for their support and cooperation during Expedition 336. Figures illustrating the CPOs were made using the interactive programs of D. Mainprice, Université Montpellier II, France. We were able to make use of analytical instruments housed at the GSJ-Lab (JEOL JXA-8800 electron probe micro-analyzer) of the Geological Survey of Japan, AIST. We also thank the sample preparation team (A. Owada, T. Sato, K. Fukuda, and E. Hirabayashi) of the Geological Survey of Japan, AIST, for technical assistance in preparing thin sections. We are indebted to T. Sato for drawing the bathymetry map of the North Pond in Figure 1c. E. Hellebrand, U. Faul, and an anonymous reviewer provided thoughtful comments. This study was supported by research grants from the Japan Drilling Earth Science Consortium (J-DESC) and JSPS grants 15H02148 and 16H001123 for J.-I.K. The maps used in Figure 1 were generated by Generic Mapping Tool (GMT version 4.5.9) [Wessel and Smith, 1998] with digital elevation data of ETOPO1 from data in Amante and Eakins [2009] , the Global Multiresolution Topography Data Portal (http://www.marine-geo.org/portals/gmrt/) [Ryan et al., 2009], and Cruise Conrad 30-01 in 1989 (Co-Chief Scientists R. Detrick and J. Mutter) (data available from www.marine-geo.org/tools/search/entry.php?id=RC3001). The core recovery and lithological data for Hole U1382A used in Figure 2a are from the literature [Expedition 336 Scientists, 2012; Edwards et al., 2014]. Closed core photo data from IODP Expedition 336, used in Figures 2b and 2c, can be obtained through the expedition's implementing organization USIO (http://iodp.tamu.edu/database). Photographs and mapping data used in Figures 3 and 4 are from our new data. Mineral fabric data used in Figure 5 are also from our new data. The data used in Figure 6 are from our new data given in Tables 2-6. Data used in Figures 7 and 8 are from our own data given in Tables 7-9. Data supporting Figure 10 are available in Table 1.