Tracking Deep Lithospheric Events with Garnet-Websterite Xenoliths from Southeastern Australia

Tracking Deep Lithospheric Events with Garnet-Websterite Xenoliths from Southeastern Australia ABSTRACT Pyroxenites provide important information on mantle heterogeneity and can be used to trace mantle evolution. New major and trace element and Sr-, Nd-, and Hf-isotope analyses of minerals and whole-rock samples of garnet websterites entrained in basanite tuffs in Bullenmerri and Gnotuk maars, southeastern Australia, are here combined with detailed petrographic observations to constrain the sources and genesis of the pyroxenites, and to trace the dynamic evolution of the lithospheric mantle. Most garnet websterites have high MgO and Cr2O3 contents, relatively flat light rare earth element (LREE) patterns ([La/Nd]CN = 0·77–2·22) and ocean island basalt-like Sr-, Nd-, and Hf-isotope compositions [87Sr/86Sr = 0·70412–0·70657; εNd(t) = –0·32 to +4·46; εHf(t)=+1·69 to +18·6] in clinopyroxenes. Some samples show subduction-related signatures with strong enrichments in large ion lithophile elements and LREE, and negative anomalies in high field strength elements, as well as high 87Sr/86Sr (up to 0·709), and decoupled Hf- and Nd-isotope compositions [εNd(t) = –3·28; εHf(t) = +11·6). These data suggest that the garnet pyroxenites represent early crystallization products of mafic melts derived from a convective mantle wedge. Hf model ages and Sm–Nd mineral isochrons suggest that these pyroxenites record at least two stages of evolution. The initial formation stage corresponds to the Paleozoic subduction of the proto-Pacific plate beneath southeastern Australia, which generated hydrous tholeiitic melts that crystallized clinopyroxene-dominated pyroxenites at ∼1420–1450°C and ∼75 km depth in the mantle wedge. The second stage corresponds to Eocene (c. 40 Ma) back-arc lithospheric extension, which led to uplift of the former mantle-wedge domain to 40–60 km depths, and subsequent cooling to the ambient geotherm (∼950–1100°C). Extensive exsolution and recrystallization of garnet and orthopyroxene (± ilmenite) from clinopyroxene megacrysts accompanied this stage. The timing of these mantle events coincides with vertical tectonism in the overlying crust. INTRODUCTION Studies of ultramafic xenoliths and massifs have provided much information on the Earth’s upper mantle. In a broad sense, these mantle-derived rocks provide direct evidence that the subcontinental upper mantle consists dominantly of peridotite with minor pyroxenite (<10%; e.g. Downes, 2007; Bodinier & Godard, 2014). Consequently, nearly all petrological studies of mantle melting traditionally start with the assumption that primary mantle melts equilibrate with olivine. More recently, the role of pyroxenites in the generation of mantle-derived magmas has been increasingly considered (Hirschmann & Stolper, 1996; Sobolev et al., 2005; Tuff et al., 2005). Preferential melting of pyroxenite (or eclogite) layers or veins accounts satisfactorily for isotopic variations in oceanic basalts, and potentially for isotopic discrepancies between mid-ocean ridge basalts (MORB) and residual abyssal peridotites (Salters & Dick, 2002). Experimental petrology and isotopic considerations suggest that even less than 2% of pyroxenite in basalt source regions may result in significant enhancements of magma production and may be responsible for a sizeable fraction of MORB melts (Pertermann & Hirschmann, 2003a). Moreover, pyroxenites may represent the missing link between the average andesitic composition of the continental crust and the basic primary arc magmas (e.g. Kelemen, 1995; Rudnick, 1995; Hawkesworth & Kemp, 2006; Lee et al., 2006, 2007; Xiong et al., 2014; Lee & Anderson, 2015; Tilhac et al., 2016). Mantle pyroxenites may thus play a key role in global geodynamic processes, although they represent only a small volume of the upper mantle, and a better understanding of their distribution and origin is critical to constraining the evolution of geochemical heterogeneities of mantle through time. Pyroxenites occur as minor layers and lenses in peridotite massifs (Kornprobst et al., 1990; Pearson et al., 1993; Garrido & Bodinier, 1999; Takazawa et al., 1999; Gysi et al., 2011) and as xenoliths in kimberlites (e.g. Roden et al., 2006; Aulbach et al., 2007; Gonzaga et al., 2010; Smit et al., 2014; Aulbach & Jacob, 2016; Newton et al., 2016) and alkali basalts (e.g. Griffin et al., 1984, 1988; Xu et al., 1996, 1998; Xu, 2002; Ishikawa et al., 2004; Bizimis et al., 2005; Yu et al., 2010; Ackerman et al., 2012). The petrogenesis of these pyroxenites is controversial (Downes, 2007; Bodinier & Godard, 2014). Most are interpreted as high-pressure cumulates that crystallized from a variety of melts migrating through the mantle (Frey, 1980; Irving, 1980; Medaris & Syada, 1999; Fabries et al., 2001; Xu, 2002; Upton et al., 2003; Ishikawa et al., 2004; Bizimis et al., 2005; Berger et al., 2007; Dantas et al., 2007; Medaris et al., 2013; Xiong et al., 2014; Martin et al., 2015), or formed by metamorphism of subducted oceanic crust tectonically incorporated into the mantle (Allègre & Turcotte, 1986; Kornprobst et al., 1990; Morishita et al., 2003; Obata et al., 2006; Yu et al., 2010; Svojtka et al., 2016). Melt–rock and related peridotite-replacement reactions also have been recognized as a viable model for the origin of some mantle pyroxenites (Garrido & Bodinier, 1999; Liu et al., 2005; Bodinier et al., 2008; Tilhac et al., 2016, 2017). Abundant pyroxenite xenoliths are found in the Mesozoic–Cenozoic basalts of southeastern Australia and they have been interpreted as clinopyroxene-dominant cumulates, modified by exsolution and recrystallization during cooling to the ambient geotherm (Irving, 1974; Wilkinson & Kalocsai, 1974; Griffin et al., 1984; Stolz, 1984; Lu et al., 2017). Most previous studies mainly focused on the reconstruction of a paleo-geotherm for southeastern Australia (SEA geotherm; i.e. Griffin et al., 1984; O’Reilly & Griffin, 1985, 1996; O’Reilly, 1989; O’Reilly et al., 1997), and there are few constraints on the source of the pyroxenites, or the conditions and timing of their formation. Therefore, this study has re-examined garnet pyroxenites from the Quaternary maars at Lakes Bullenmerri and Gnotuk (Victoria, southeastern Australia), using new geochemical analytical and imaging technologies, and integrates petrological and major and trace element compositions of minerals and whole-rock samples with Sr-, Nd-, and Hf-isotope compositions. The aims are to better constrain the origin of the parental magmas to the garnet pyroxenites and to determine the dynamic processes that trigger the melt generation and pyroxenite formation. GEOLOGICAL SETTING AND TECTONIC CONTEXT The Australian continent consists of two broad tectonic regimes: the western two-thirds, which comprises dominantly Archean–Proterozoic cratonic terranes overlain in places by Paleozoic cover, and the eastern third consisting of the complex Paleozoic–Mesozoic Tasmanides, a collage of several orogenic belts and an internal Permian–Triassic foreland fold–thrust belt (Fig. 1a). The Tasman Line is the boundary of these two domains and it traces the locus of the ∼600 Ma rifting margin (the younger of two rifting events) during the break-up of Rodinia (Veever, 1984). A series of convergent-margin orogenic belts subsequently docked with cratonic Australia from ∼540 to 200 Ma. Successive SW to NE accretion events formed the southern Tasmanides, now represented by the Delamerian, Lachlan, Thomson and New England fold belts (Fig. 1a;Glen, 2005). The Lachlan and Delamerian fold belts (Fig. 1b) crop out in western Victoria, separated by a major crustal feature, the Moyston Fault (VandenBerg et al., 2000; Cayley et al., 2011). Fig. 1. View largeDownload slide (a) Major tectonic divisions of Australia. Area west of the Tasman Line consists of Archean–Proterozoic cratonic terranes; the eastern part consists of the Paleozoic–Mesozoic Tasmanides (Delamerian, Lachlan, New England, and Thomson orogenic belts. (b) Simplified geological map of western Victoria showing the Lachlan orogenic belt, Delamerian orogenic belt and Otway Basin (compiled from Miller et al., 2002, 2005). Fig. 1. View largeDownload slide (a) Major tectonic divisions of Australia. Area west of the Tasman Line consists of Archean–Proterozoic cratonic terranes; the eastern part consists of the Paleozoic–Mesozoic Tasmanides (Delamerian, Lachlan, New England, and Thomson orogenic belts. (b) Simplified geological map of western Victoria showing the Lachlan orogenic belt, Delamerian orogenic belt and Otway Basin (compiled from Miller et al., 2002, 2005). The Delamerian Orogen formed a west-vergent foreland-type fold-and-thrust belt that overprinted the Neoproterozoic Adelaidean successions from c. 514 to 490 Ma (Betts et al., 2002; Foden et al., 2006). The Lachlan fold belt is considered to be a wide (∼750 km), early to mid-Paleozoic continental margin with repetitive rifting, collision, and subduction–accretion episodes associated with granitic magmatism and basaltic–andesitic volcanism from c. 440 to 340 Ma (Coney, 1992; Foster & Gray, 2000). During the Mesozoic and the Cenozoic, Australia moved northward from Antarctica, forming a series of rift basins (i.e. Otway Basin in western Victoria) along the southern seaboard of the Australian continent (Fig. 1b). The Otway Basin is an east–west trough initiated early in the Mesozoic period and contains a thick sequence of Mesozoic and Cenozoic sedimentary and volcanic rocks (Edwards et al., 1996; Holdgate & Gallagher, 2003; Briguglio et al., 2015). Basaltic lavas, tuffs and other pyroclastic flows are widespread in the Otway Basin. Many are Pleistocene to Recent in age, and are referred to as the ‘Newer Volcanic Province’ (NVP). The NVP can be subdivided into two series, the early Plains (dominant) and the slightly later Cones, on the basis of geomorphology and composition (Wellman & McDougall, 1974). The Plains basalts are mainly tholeiitic to transitional in composition, whereas the Cones comprise silica-undersaturated alkali basalts, commonly erupted as scoria or cinder cones, maars and tuff rings; they are locally very rich in mantle-derived xenoliths (Edwards et al., 1996). The mantle xenoliths used in this study are from Lakes Bullenmerri and Gnotuk (Bullenmerri–Gnotuk), adjacent vents, now maars, <1 km apart (Fig. 1b). The locality is close to the Mortlake discontinuity, defined in the lithospheric mantle by a contrast in the Sr- and Pb-isotopic compositions of NVP Plains basalts; 87Sr/86Sr is low to the west and high to the east (Price et al., 1997). The Re–Os age distribution of the mantle xenoliths also suggests a significant age change from Proterozoic in the west to dominantly Phanerozoic to the east of the Mortlake discontinuity (Handler et al., 1997; Handler & Bennett, 2001). The discontinuity appears to be coincident with the surface expression of the east-dipping Moyston Fault, which separates the Paleozoic Delamerian–Lachlan fold belts (Glen, 1992). The basanite tuffs have compositions within the range of the NVP Cones basalts, with low MgO (∼6·71 wt %) and CaO (∼6·89 wt %) and high alkali contents; they also have limited ranges in Sr and Nd isotope compositions (87Sr/86Sr = 0·7039–0·7041, 143Nd/144Nd= ∼0·51283; Griffin et al., 1988; Stolz & Davies, 1988). PETROGRAPHY The pyroxenite suite, including both spinel- and garnet-bearing pyroxenites, is clearly distinguished from the peridotite suite based on modal abundances, pyroxene compositions and microstructures. Pyroxenites occur separately, or as layers and pods in spinel lherzolite xenoliths, suggesting that they were intrusive into the lherzolites (Griffin et al., 1984). Although Griffin et al. (1984) reported the presence of garnet-free pyroxenites in the Bullenmerri–Gnotuk suite, all the samples studied here contain garnet. Most xenoliths are small (∼5 cm in the longest dimension), but some samples reach 20 cm. They may be sub-spherical or ovoid but more commonly are angular in shape with rectangular to triangular cross-sections. Among the nine garnet websterites and one garnet clinopyroxenite chosen for this study, a wide range of microstructural types has been observed, from coarse to porphyroblastic (Fig. 2; Table 1). Table 1: Summary of petrological description and calculated modal compositions (wt %) for garnet pyroxenites in Lakes Bullenmerri and Gnotuk Sample no. Rock type Texture Mg# T–P estimates Estimated modal abundances (wt %)‡ Primary* Re-equilibration† Cpx Opx Grt Amp Spl Ilm Phl BM96-2 Grt websterite Coarse 78·2 1094°C, 1·8 GPa 52 13 30 4 1 BM99-6 Grt clinopyroxenite Coarse 84·4 1060°C, 1·6 GPa 61 38 <0·5 1 BM99-4 Grt websterite Porphyroblastic 85·1 1450°C, 2·3 GPa 982°C, 1·5 GPa 57 20 18 5 BME-1 Grt websterite Porphyroblastic 82·3 1420°C, 2·4 GPa 1011°C, 1·4 GPa 55 11 33 <0·5 BM99-2 Grt websterite Porphyroblastic 80·7 1450°C, 2·4 GPa 1064°C, 1·7 GPa 55 13 31 BM96-3 Grt websterite Coarse 73·6 1010°C, 1·4 GPa 23 15 60 <0·5 2 BM99-5 Grt websterite Porphyroblastic 76·1 1460°C, 3·0 GPa 1047°C, 1·7 GPa 64 28 5 3 <0·5 DR9748§ Grt websterite Granoblastic 85·2 1045°C, 1·6 GPa 40(41) 44(45) 11(13) 5(0) DR10165§ Grt websterite Porphyroblastic 83·2 970°C, 1·2 GPa 49(46) 25(28) 17(17) 3(0) 6(8) GN35§ Grt websterite Porphyroblastic 78·4 1063°C, 1·6 GPa 54(50) 27(37) 8(0·28) 4(0) 7(12) Sample no. Rock type Texture Mg# T–P estimates Estimated modal abundances (wt %)‡ Primary* Re-equilibration† Cpx Opx Grt Amp Spl Ilm Phl BM96-2 Grt websterite Coarse 78·2 1094°C, 1·8 GPa 52 13 30 4 1 BM99-6 Grt clinopyroxenite Coarse 84·4 1060°C, 1·6 GPa 61 38 <0·5 1 BM99-4 Grt websterite Porphyroblastic 85·1 1450°C, 2·3 GPa 982°C, 1·5 GPa 57 20 18 5 BME-1 Grt websterite Porphyroblastic 82·3 1420°C, 2·4 GPa 1011°C, 1·4 GPa 55 11 33 <0·5 BM99-2 Grt websterite Porphyroblastic 80·7 1450°C, 2·4 GPa 1064°C, 1·7 GPa 55 13 31 BM96-3 Grt websterite Coarse 73·6 1010°C, 1·4 GPa 23 15 60 <0·5 2 BM99-5 Grt websterite Porphyroblastic 76·1 1460°C, 3·0 GPa 1047°C, 1·7 GPa 64 28 5 3 <0·5 DR9748§ Grt websterite Granoblastic 85·2 1045°C, 1·6 GPa 40(41) 44(45) 11(13) 5(0) DR10165§ Grt websterite Porphyroblastic 83·2 970°C, 1·2 GPa 49(46) 25(28) 17(17) 3(0) 6(8) GN35§ Grt websterite Porphyroblastic 78·4 1063°C, 1·6 GPa 54(50) 27(37) 8(0·28) 4(0) 7(12) * CaO–MgO–Al2O3–SiO2 (CMAS) system from Gasparik (2014). † Ellis & Green (1979) Cpx–Grt thermometer and Wood (1974) Al-in-Opx barometer. ‡ Mineral modes were estimated from point-counted modes and converted to wt % from mineral densities using ρCpx = 0·32, ρOpx = 0·32, ρGrt = 3·71, ρAmp = 3·12, ρSpl = 3·6, ρIlm = 4·72, ρPhl = 2·8. Least-squares fitted modes (results in parentheses) were calculated from whole-rock and mineral major element compositions. § Samples are from Griffin et al. (1984, 1988) and O’Reilly & Griffin (1985). Table 1: Summary of petrological description and calculated modal compositions (wt %) for garnet pyroxenites in Lakes Bullenmerri and Gnotuk Sample no. Rock type Texture Mg# T–P estimates Estimated modal abundances (wt %)‡ Primary* Re-equilibration† Cpx Opx Grt Amp Spl Ilm Phl BM96-2 Grt websterite Coarse 78·2 1094°C, 1·8 GPa 52 13 30 4 1 BM99-6 Grt clinopyroxenite Coarse 84·4 1060°C, 1·6 GPa 61 38 <0·5 1 BM99-4 Grt websterite Porphyroblastic 85·1 1450°C, 2·3 GPa 982°C, 1·5 GPa 57 20 18 5 BME-1 Grt websterite Porphyroblastic 82·3 1420°C, 2·4 GPa 1011°C, 1·4 GPa 55 11 33 <0·5 BM99-2 Grt websterite Porphyroblastic 80·7 1450°C, 2·4 GPa 1064°C, 1·7 GPa 55 13 31 BM96-3 Grt websterite Coarse 73·6 1010°C, 1·4 GPa 23 15 60 <0·5 2 BM99-5 Grt websterite Porphyroblastic 76·1 1460°C, 3·0 GPa 1047°C, 1·7 GPa 64 28 5 3 <0·5 DR9748§ Grt websterite Granoblastic 85·2 1045°C, 1·6 GPa 40(41) 44(45) 11(13) 5(0) DR10165§ Grt websterite Porphyroblastic 83·2 970°C, 1·2 GPa 49(46) 25(28) 17(17) 3(0) 6(8) GN35§ Grt websterite Porphyroblastic 78·4 1063°C, 1·6 GPa 54(50) 27(37) 8(0·28) 4(0) 7(12) Sample no. Rock type Texture Mg# T–P estimates Estimated modal abundances (wt %)‡ Primary* Re-equilibration† Cpx Opx Grt Amp Spl Ilm Phl BM96-2 Grt websterite Coarse 78·2 1094°C, 1·8 GPa 52 13 30 4 1 BM99-6 Grt clinopyroxenite Coarse 84·4 1060°C, 1·6 GPa 61 38 <0·5 1 BM99-4 Grt websterite Porphyroblastic 85·1 1450°C, 2·3 GPa 982°C, 1·5 GPa 57 20 18 5 BME-1 Grt websterite Porphyroblastic 82·3 1420°C, 2·4 GPa 1011°C, 1·4 GPa 55 11 33 <0·5 BM99-2 Grt websterite Porphyroblastic 80·7 1450°C, 2·4 GPa 1064°C, 1·7 GPa 55 13 31 BM96-3 Grt websterite Coarse 73·6 1010°C, 1·4 GPa 23 15 60 <0·5 2 BM99-5 Grt websterite Porphyroblastic 76·1 1460°C, 3·0 GPa 1047°C, 1·7 GPa 64 28 5 3 <0·5 DR9748§ Grt websterite Granoblastic 85·2 1045°C, 1·6 GPa 40(41) 44(45) 11(13) 5(0) DR10165§ Grt websterite Porphyroblastic 83·2 970°C, 1·2 GPa 49(46) 25(28) 17(17) 3(0) 6(8) GN35§ Grt websterite Porphyroblastic 78·4 1063°C, 1·6 GPa 54(50) 27(37) 8(0·28) 4(0) 7(12) * CaO–MgO–Al2O3–SiO2 (CMAS) system from Gasparik (2014). † Ellis & Green (1979) Cpx–Grt thermometer and Wood (1974) Al-in-Opx barometer. ‡ Mineral modes were estimated from point-counted modes and converted to wt % from mineral densities using ρCpx = 0·32, ρOpx = 0·32, ρGrt = 3·71, ρAmp = 3·12, ρSpl = 3·6, ρIlm = 4·72, ρPhl = 2·8. Least-squares fitted modes (results in parentheses) were calculated from whole-rock and mineral major element compositions. § Samples are from Griffin et al. (1984, 1988) and O’Reilly & Griffin (1985). Fig. 2. View largeDownload slide Scanned images of thin sections of garnet websterites. (a) BM99-4 shows Cpx and Opx megacrysts and subsolidus recrystallization textures. (b) GN35 shows intergrowths of spinel and orthopyroxene parallel to exsolution lamellae of fine-grained garnet and orthopyroxene in the Cpx megacryst. (c) BM96-3 shows coarse polygonal Grt set in a matrix of recrystallized Cpx and Opx; Phl-bearing pockets surrounding coarse Grt should be noted. Cpx, clinopyroxene; Opx, orthopyroxene; Grt, garnet; Phl, phlogopite; Amp, amphibole; Spl, spinel. Fig. 2. View largeDownload slide Scanned images of thin sections of garnet websterites. (a) BM99-4 shows Cpx and Opx megacrysts and subsolidus recrystallization textures. (b) GN35 shows intergrowths of spinel and orthopyroxene parallel to exsolution lamellae of fine-grained garnet and orthopyroxene in the Cpx megacryst. (c) BM96-3 shows coarse polygonal Grt set in a matrix of recrystallized Cpx and Opx; Phl-bearing pockets surrounding coarse Grt should be noted. Cpx, clinopyroxene; Opx, orthopyroxene; Grt, garnet; Phl, phlogopite; Amp, amphibole; Spl, spinel. Porphyroblastic textures Most samples show porphyroblastic textures with large pyroxene megacrysts. Megacrysts (Fig. 2a and b) dominantly are clinopyroxene (Cpx) with exsolved lamellae of garnet (Grt) + orthopyroxene (Opx) ± ilmenite (Ilm) (Fig. 3a and b); minor Opx megacrysts have exsolution lamellae of Cpx and Grt (Fig. 3c). Amphibole (Amp) also occurs as lamellae and pods in clinopyroxene megacrysts and replaces garnet and orthopyroxene lamellae (Fig. 3d). The matrix shows subsolidus recrystallization textures with an equigranular mosaic of fine-grained (0·5–2 mm) Cpx + Grt + Opx + Amp (Fig. 3e). Cpx and Opx neoblasts are generally free of exsolution and well equilibrated in the texture (Fig. 3e). Discrete garnet grains are polygonal and are locally aligned among the randomly distributed Cpx subgrains and neoblasts, indicating that they probably represent former lamellae exsolved from original coarse Cpx, the grain size of which was reduced to fine neocrysts with increasing degree of recrystallization (Lu et al., 2017). Amphibole is pervasive in the recrystallized assemblages and occurs as subhedral grains in apparent textural equilibrium with the anhydrous phases. Some samples (e.g. DR9748) also contain amphibole veins that crosscut the sample along a set of cracks (Fig. 3f) and locally enclose spinel grains with strongly embayed margins (Fig. 3g). Textural observations indicate at least two episodes of amphibole formation, as the veined amphibole postdates the discrete (pervasive) amphibole in the recrystallized areas. The transition from lamellar to mosaic microstructure is not sharp (Fig. 3h–i), reflecting varying degrees of recrystallization within individual thin sections. Detailed descriptions of the microstructures have been given by Lu (2018). Fig. 3. View largeDownload slide Photomicrographs [a–d and g, back-scattered electron (BSE) images; e, f, plane-polarized light photographs; h, i, cross-polarized light photographs] showing porphyroblastic textures of garnet pyroxenite xenoliths (a, c and i, BM99-4; b, BM99-5; d, BME-1; e, f, DR9748; g, BME-1; h, BM99-2) from Lakes Bullenmerri and Gnotuk, southeastern Australia. (a) Cpx megacryst with Opx lamellae and Grt blebs; (b) Cpx megacryst with Opx, Grt and ilmenite (Ilm) lamellae; (c) Opx megacryst with Cpx and Grt lamellae surrounded by coarser garnet crystals around the grain margin; (d) ‘exsolution-like’ Amp in Cpx megacryst; (e) mosaic textures with discrete Cpx, Grt, Opx and Amp; (f) Amp vein (dashed line) across the sample; (g) veined Amp reaction rim on Spl; (h, i) transition from lamellar to mosaic microstructure. Fig. 3. View largeDownload slide Photomicrographs [a–d and g, back-scattered electron (BSE) images; e, f, plane-polarized light photographs; h, i, cross-polarized light photographs] showing porphyroblastic textures of garnet pyroxenite xenoliths (a, c and i, BM99-4; b, BM99-5; d, BME-1; e, f, DR9748; g, BME-1; h, BM99-2) from Lakes Bullenmerri and Gnotuk, southeastern Australia. (a) Cpx megacryst with Opx lamellae and Grt blebs; (b) Cpx megacryst with Opx, Grt and ilmenite (Ilm) lamellae; (c) Opx megacryst with Cpx and Grt lamellae surrounded by coarser garnet crystals around the grain margin; (d) ‘exsolution-like’ Amp in Cpx megacryst; (e) mosaic textures with discrete Cpx, Grt, Opx and Amp; (f) Amp vein (dashed line) across the sample; (g) veined Amp reaction rim on Spl; (h, i) transition from lamellar to mosaic microstructure. Coarse textures In coarse microstructures, garnets are up to 5 mm across and are set in slightly finer granoblastic Cpx, Opx and Amp (Figs 2c and 4a, b). Coarse-grained garnets are polygonal or rounded, and partially or completely replaced by an extremely fine-grained, symplectitic aggregate (kelyphite) consisting of Opx, spinel (Spl) and Cpx. Rare Grt aggregates also contain small round (0·1–0·5 mm) Spl grains (Fig. 4c) and accompany the intergrowths of vermicular Spl and Opx (Fig. 4d), probably representing the breakdown of Grt lamellae exsolved from Cpx megacrysts under different P–T conditions. In sample GN35, for instance, the intergrowth of Opx and Spl is parallel to the Opx and Grt exsolution lamellae in the Cpx megacryst (Fig. 2b), suggesting that the sample shows two apparently sequential reactions: (1) Cpx I → Cpx II + Grt I ± Opx I; (2) Grt I → Grt II + Spl + Opx II. One sample (BM96-3) contains phlogopite (Phl)-bearing pockets in which subhedral Phl grains are surrounded by open cavities associated with intergrowths of fine-grained Cpx, Spl and Opx (Fig. 4e and f). Phl is in direct contact with the neighboring Cpx and Opx but is separated from coarse Grt by fine-grained Cpx + Opx + Spl, indicating that garnet breakdown occurred following the reaction Grt + fluids → Cpx + Opx + Spl + Phl. Rare apatite (Ap) is also found in some samples. Fig. 4. View largeDownload slide Photomicrographs showing coarse textures and mineral assemblages. (a, b) Coarse Grt and relatively finer Cpx + Opx (plane-polarized light). Amp vein (dashed line) along the boundary of coarse Grt in (b) should be noted. (c, d) Spl–Grt relationship (BSE image); (c) spinel-core in direct contact with coarse garnet; (d) intergrowth of Opx and Spl. (e, f) Phlogopite in direct contact with discrete Cpx and Opx but not Grt in BM96-3 (false-colour SEM phase map). Fig. 4. View largeDownload slide Photomicrographs showing coarse textures and mineral assemblages. (a, b) Coarse Grt and relatively finer Cpx + Opx (plane-polarized light). Amp vein (dashed line) along the boundary of coarse Grt in (b) should be noted. (c, d) Spl–Grt relationship (BSE image); (c) spinel-core in direct contact with coarse garnet; (d) intergrowth of Opx and Spl. (e, f) Phlogopite in direct contact with discrete Cpx and Opx but not Grt in BM96-3 (false-colour SEM phase map). The petrographic features described above suggest that the xenoliths from the Bullenmerri–Gnotuk maars preserve three distinct mineral assemblages. The first generation is composed of clinopyroxene megacrysts and small amounts of spinel and orthopyroxene. The second generation formed by the subsolidus re-equilibration of Cpx megacrysts through exsolution and recrystallization into dominantly Cpx, Grt, Opx and Amp. They were then intruded by veins of Amp and Phl, potentially related to an episode of volatile-rich metasomatism. ANALYTICAL METHODS All analyses were carried out at the Geochemical Analytical Unit (GAU) of the ARC Centre of Excellence for Core to Crust Fluid Systems (CCFS) and GEMOC, Macquarie University, Sydney, Australia. Detailed analytical procedures for elemental X-ray mapping, and the analysis of major element compositions by electron probe microanalysis (EPMA) and of trace element compositions by laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) have already been presented for the same set of xenoliths by Lu et al. (2017). For Sr-, Nd-, and Hf-isotope analysis, about 200 mg of clean clinopyroxene and amphibole, and ∼400 mg of clean garnet were separated from garnet websterites, acid-leached and digested, and processed using ion-exchange chromatographic column techniques described by Pin & Zalduegui (1997) and Blichert-Toft (1999). Sr- and Nd-isotopic ratios were analyzed by thermal ionization mass spectrometry (TIMS) using a Thermo Finnigan Triton system. Detailed analytical procedures have been described by Tilhac et al. (2017). Sr and Nd were loaded on single and double Re filaments, respectively. SRM 987 and JMC321 were measured to check instrument status and sensitivity, and yielded 87Sr/86Sr = 0·710244 ± 7 (2σ) and 143Nd/144Nd = 0·511129 ± 2 (2σ), respectively, which are within error of the GeoReM recommended values (http://georem.mpch-mainz.gwdg.de/). The USGS reference standard BHVO-2 yielded 0·703468 ± 8 (2σ) for 87Sr/86Sr and 0·512985 ± 3 (2σ) for 143Nd/144Nd, also comparable with GeoReM recommended values of 0·703469 ± 34 and 0·512980 ± 12, respectively. Measured 87Sr/86Sr and 143Nd/144Nd ratios were normalized to 86Sr/88Sr = 0·1194 and 146Nd/144Nd = 0·7219 to correct for mass fractionation, respectively. Hf-isotope ratios were analysed using a Nu Plasma multi-collector (MC)-ICP-MS system. Standards JMC475 and BHVO-2 gave average 176Hf/177Hf ratios of 0·282156 ± 8 (2σ) and 0·283108 ± 24 (2σ), within error of the GeoReM recommended values of 0·282159 ± 26 and 0·283109 ± 12, respectively. Ratios were normalized to 179Hf/177Hf = 0·7325. Sr isotopes in clinopyroxene and amphibole were also analyzed in situ using a 193 nm ArF excimer laser system attached to a Nu Plasma MC-ICP-MS system. Typical operating conditions for the laser were a 5 Hz repetition rate and an energy density of 7·5–9·3 J cm–2, which provided the best signal stability and analytical precision. The spot size was 155 μm. Each analysis included 60 s of background followed by 140 s of data collection. The instrument was calibrated to the optimal conditions before analysis by measuring an SRM987 Sr solution. All analyses were done in time-resolved analysis mode, monitoring signals on each mass with the Nu Plasma analysis software to select the most stable interval of each run and avoid mixed ablation. Interference correlations were performed following Donnelly et al. (2012). The Batbjerg Cr-diopside (Batbjerg-1) reference material was run for monitoring instrument drift. In our sessions, analyses of Batbjerg-1 gave an average 87Sr/86Sr of 0·704454 ± 46 (2SD, n = 59), within the uncertainty of the published TIMS results (87Sr/86Sr = 0·704474 ± 17; Neumann et al., 2004). RESULTS Mineral compositions Major element compositions The major element compositions of minerals are listed in Table 2. Compositional profiles across grains of clinopyroxene and garnet indicate that the garnet pyroxenite xenoliths reached chemical equilibrium in terms of major elements (Lu et al., 2017) and they show consistent compositions among different textures within individual samples (Table 2). Table 2: Representative major element compositions of minerals in the garnet pyroxenites from Lakes Bullenmerri and Gnotuk Clinopyroxene Sample: BM 96-2 BM 99-6 BM 99-4 BME-1 BM 99-2 BM 96-3 BM 99-5 DR 9748* GN35* DR10165* Type: M M H M H M H M M H M H M H M SiO2 51·2 51·3 51·0 51·4 51·5 51·8 50·9 51·2 51·4 51·3 50·9 51·1 51·4 51·1 51·4 TiO2 0·37 0·36 0·73 0·79 0·43 0·42 0·40 0·40 1·02 0·80 0·88 0·68 0·62 0·53 0·51 Al2O3 6·62 6·75 7·14 7·51 6·33 6·42 6·43 6·34 6·20 4·87 5·98 7·29 7·11 6·5 6·02 Cr2O3 0·28 0·17 0·30 0·30 0·24 0·26 0·19 0·22 0·11 0·32 0·44 0·37 0·2 0·2 0·17 FeO 5·12 3·25 3·32 3·29 3·58 3·60 4·56 4·43 5·00 6·67 5·86 3·47 3·77 3·61 3·77 MnO 0·09 0·07 0·07 0·09 0·06 0·06 0·08 0·08 0·09 0·13 0·13 0·13 0·12 0·10 MgO 14·5 15·4 14·7 14·3 14·9 14·6 15·0 14·9 14·8 14·2 14·5 14·3 14·5 14·8 15·2 CaO 20·3 21·6 20·8 20·9 21·7 22·0 21·0 21·0 21·1 20·3 21·1 20·9 20·5 22·2 21·5 Na2O 1·55 1·28 1·51 1·58 1·17 1·19 1·41 1·43 1·16 1·26 1·03 1·49 1·5 1·03 1·00 NiO 0·08 0·08 b.d. b.d. 0·07 0·07 b.d. b.d. 0·09 0·08 — — — — — Total 100·1 100·2 99·6 100·2 99·9 100·5 99·9 100·1 100·8 99·9 100·8 99·7 99·8 99·9 99·7 Mg# 83·6 89·5 88·8 88·7 88·2 88·0 85·5 85·8 84·1 79·3 81·5 88·1 87·4 88·1 87·9 Orthopyroxene Sample: BM96-2 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748* GN35* DR10165* Type: M E1 M H E1 M E1 M M E1 M E1 M E1 M SiO2 53·3 54·5 54·8 55·0 55·2 54·4 53·5 53·4 54·9 53·3 53·6 54·2 54·3 54·2 54·4 TiO2 0·08 0·11 0·10 0·12 0·07 0·08 0·06 0·05 0·22 0·25 0·15 0·12 0·10 — 0·10 Al2O3 4·63 4·61 4·72 4·79 4·61 4·60 4·70 4·60 4·12 3·69 4·07 4·99 4·92 4·84 4·77 Cr2O3 0·14 0·14 0·14 0·15 0·12 0·12 0·10 0·11 0·06 0·16 0·21 0·19 0·09 0·1 0·08 FeO 11·6 8·44 9·18 8·57 9·41 9·43 10·5 10·6 11·7 15·4 13·1 8·47 8·71 9·30 9·45 MnO 0·16 0·17 0·19 0·19 0·12 0·09 0·15 0·16 0·15 0·25 0·22 0·21 0·22 0·12 0·16 MgO 29·7 31·2 31·4 31·7 31·1 30·5 30·6 30·9 29·2 26·6 28·6 31·1 31·0 30·7 30·9 CaO 0·75 0·61 0·52 0·51 0·55 0·60 0·67 0·68 0·77 0·86 0·71 0·69 0·63 0·48 0·56 Na2O 0·09 0·08 0·10 0·07 0·05 b.d. 0·09 0·08 0·08 0·11 0·08 — 0·08 — 0·06 NiO 0·12 b.d. b.d. b.d. 0·12 0·13 0·11 0·11 0·14 0·10 — — — — — Total 100·6 100·0 101·1 101·1 101·4 100·0 100·5 100·5 101·4 100·7 100·8 100·0 100·0 99·7 100·5 Mg# 82·0 86·8 85·9 86·8 85·5 85·2 83·8 83·9 81·7 75·5 79·5 86·7 86·4 85·5 85·3 Garnet Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748* GN35* DR10165* Type: C C E1 E2 M E1 C E1 M C C E1 M E1 C M SiO2 41·2 41·8 40·9 40·9 41·0 41·9 41·6 41·4 41·5 41·3 41·7 40·6 41·1 41·9 41·8 41·8 TiO2 0·08 0·04 0·07 0·08 0·07 0·04 0·04 0·04 0·07 0·04 0·14 0·17 0·10 — — — Al2O3 23·2 24·2 23·5 23·5 23·5 23·3 23·6 23·8 23·5 23·9 23·0 22·6 22·9 23·7 23·4 24·0 Cr2O3 0·45 0·18 0·24 0·25 0·25 0·27 0·24 0·21 0·23 0·26 0·12 0·44 0·64 0·32 0·52 0·17 FeO 13·3 9·3 10·7 10·5 10·8 10·9 11·7 12·3 12·6 11·9 13·7 16·6 15·0 10·1 10·6 11·6 MnO 0·41 0·31 0·50 0·51 0·52 0·25 0·30 0·42 0·42 0·42 0·42 0·57 0·56 0·54 0·56 0·44 MgO 17·4 19·6 19·1 19·3 19·0 17·8 18·6 17·8 18·3 17·7 15·9 14·6 15·5 18·6 18·4 17·7 CaO 5·30 5·47 4·28 4·20 4·24 5·80 5·42 5·46 5·39 5·46 5·61 5·79 5·81 5·23 5·16 5·53 Na2O b.d. b.d. b.d. b.d. b.d. 0·24 b.d. b.d. b.d. b.d. b.d. 0·02 — — — — Total 101·3 101·0 99·3 99·2 99·5 100·4 101·5 101·3 102·0 101·0 100·5 101·4 101·6 100·4 100·4 101·3 Mg# 70·2 79·1 76·2 76·7 76·0 74·7 74·1 72·3 72·4 72·9 67·6 61·4 64·7 76·6 75·6 73·1 Clinopyroxene Sample: BM 96-2 BM 99-6 BM 99-4 BME-1 BM 99-2 BM 96-3 BM 99-5 DR 9748* GN35* DR10165* Type: M M H M H M H M M H M H M H M SiO2 51·2 51·3 51·0 51·4 51·5 51·8 50·9 51·2 51·4 51·3 50·9 51·1 51·4 51·1 51·4 TiO2 0·37 0·36 0·73 0·79 0·43 0·42 0·40 0·40 1·02 0·80 0·88 0·68 0·62 0·53 0·51 Al2O3 6·62 6·75 7·14 7·51 6·33 6·42 6·43 6·34 6·20 4·87 5·98 7·29 7·11 6·5 6·02 Cr2O3 0·28 0·17 0·30 0·30 0·24 0·26 0·19 0·22 0·11 0·32 0·44 0·37 0·2 0·2 0·17 FeO 5·12 3·25 3·32 3·29 3·58 3·60 4·56 4·43 5·00 6·67 5·86 3·47 3·77 3·61 3·77 MnO 0·09 0·07 0·07 0·09 0·06 0·06 0·08 0·08 0·09 0·13 0·13 0·13 0·12 0·10 MgO 14·5 15·4 14·7 14·3 14·9 14·6 15·0 14·9 14·8 14·2 14·5 14·3 14·5 14·8 15·2 CaO 20·3 21·6 20·8 20·9 21·7 22·0 21·0 21·0 21·1 20·3 21·1 20·9 20·5 22·2 21·5 Na2O 1·55 1·28 1·51 1·58 1·17 1·19 1·41 1·43 1·16 1·26 1·03 1·49 1·5 1·03 1·00 NiO 0·08 0·08 b.d. b.d. 0·07 0·07 b.d. b.d. 0·09 0·08 — — — — — Total 100·1 100·2 99·6 100·2 99·9 100·5 99·9 100·1 100·8 99·9 100·8 99·7 99·8 99·9 99·7 Mg# 83·6 89·5 88·8 88·7 88·2 88·0 85·5 85·8 84·1 79·3 81·5 88·1 87·4 88·1 87·9 Orthopyroxene Sample: BM96-2 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748* GN35* DR10165* Type: M E1 M H E1 M E1 M M E1 M E1 M E1 M SiO2 53·3 54·5 54·8 55·0 55·2 54·4 53·5 53·4 54·9 53·3 53·6 54·2 54·3 54·2 54·4 TiO2 0·08 0·11 0·10 0·12 0·07 0·08 0·06 0·05 0·22 0·25 0·15 0·12 0·10 — 0·10 Al2O3 4·63 4·61 4·72 4·79 4·61 4·60 4·70 4·60 4·12 3·69 4·07 4·99 4·92 4·84 4·77 Cr2O3 0·14 0·14 0·14 0·15 0·12 0·12 0·10 0·11 0·06 0·16 0·21 0·19 0·09 0·1 0·08 FeO 11·6 8·44 9·18 8·57 9·41 9·43 10·5 10·6 11·7 15·4 13·1 8·47 8·71 9·30 9·45 MnO 0·16 0·17 0·19 0·19 0·12 0·09 0·15 0·16 0·15 0·25 0·22 0·21 0·22 0·12 0·16 MgO 29·7 31·2 31·4 31·7 31·1 30·5 30·6 30·9 29·2 26·6 28·6 31·1 31·0 30·7 30·9 CaO 0·75 0·61 0·52 0·51 0·55 0·60 0·67 0·68 0·77 0·86 0·71 0·69 0·63 0·48 0·56 Na2O 0·09 0·08 0·10 0·07 0·05 b.d. 0·09 0·08 0·08 0·11 0·08 — 0·08 — 0·06 NiO 0·12 b.d. b.d. b.d. 0·12 0·13 0·11 0·11 0·14 0·10 — — — — — Total 100·6 100·0 101·1 101·1 101·4 100·0 100·5 100·5 101·4 100·7 100·8 100·0 100·0 99·7 100·5 Mg# 82·0 86·8 85·9 86·8 85·5 85·2 83·8 83·9 81·7 75·5 79·5 86·7 86·4 85·5 85·3 Garnet Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748* GN35* DR10165* Type: C C E1 E2 M E1 C E1 M C C E1 M E1 C M SiO2 41·2 41·8 40·9 40·9 41·0 41·9 41·6 41·4 41·5 41·3 41·7 40·6 41·1 41·9 41·8 41·8 TiO2 0·08 0·04 0·07 0·08 0·07 0·04 0·04 0·04 0·07 0·04 0·14 0·17 0·10 — — — Al2O3 23·2 24·2 23·5 23·5 23·5 23·3 23·6 23·8 23·5 23·9 23·0 22·6 22·9 23·7 23·4 24·0 Cr2O3 0·45 0·18 0·24 0·25 0·25 0·27 0·24 0·21 0·23 0·26 0·12 0·44 0·64 0·32 0·52 0·17 FeO 13·3 9·3 10·7 10·5 10·8 10·9 11·7 12·3 12·6 11·9 13·7 16·6 15·0 10·1 10·6 11·6 MnO 0·41 0·31 0·50 0·51 0·52 0·25 0·30 0·42 0·42 0·42 0·42 0·57 0·56 0·54 0·56 0·44 MgO 17·4 19·6 19·1 19·3 19·0 17·8 18·6 17·8 18·3 17·7 15·9 14·6 15·5 18·6 18·4 17·7 CaO 5·30 5·47 4·28 4·20 4·24 5·80 5·42 5·46 5·39 5·46 5·61 5·79 5·81 5·23 5·16 5·53 Na2O b.d. b.d. b.d. b.d. b.d. 0·24 b.d. b.d. b.d. b.d. b.d. 0·02 — — — — Total 101·3 101·0 99·3 99·2 99·5 100·4 101·5 101·3 102·0 101·0 100·5 101·4 101·6 100·4 100·4 101·3 Mg# 70·2 79·1 76·2 76·7 76·0 74·7 74·1 72·3 72·4 72·9 67·6 61·4 64·7 76·6 75·6 73·1 Amphibole Phlogopite Ap Sample: BM96-2 BM99-6 BM99-4 BME-1 BM96-3 BM99-5 DR9748* GN35 DR10165* BM96-3 BME-1 Type: M vein M M vein E1 M E1 M vein M M Core Rim1 Rim2 M SiO2 42·2 41·8 42·1 41·9 42·7 42·0 41·9 40·9 42·0 42·1 42·9 42·5 36·3 35·3 33·5 TiO2 1·57 0·92 1·47 2·56 0·84 1·63 3·97 2·24 3·47 1·62 2·60 2·09 6·50 10·2 11·5 Al2O3 15·5 15·9 16·2 15·6 15·8 15·7 13·4 14·2 14·8 15·6 16·0 15·9 15·5 15·6 15·7 Cr2O3 0·29 0·26 0·24 0·43 0·35 0·35 0·13 0·42 0·54 0·40 0·34 0·29 0·14 0·20 — FeO 6·86 7·14 4·65 5·03 5·34 5·33 6·97 9·72 8·54 8·46 5·92 5·38 7·90 9·01 10·8 MnO 0·06 0·06 0·04 0·07 0·10 0·05 0·05 0·10 — 0·09 0·10 — 0·04 0·06 0·08 MgO 16·0 15·9 17·1 16·7 16·9 16·7 14·5 14·0 14·9 14·7 16·8 16·8 16·6 14·0 11·5 CaO 10·6 10·5 11·0 11·0 11·0 11·0 11·4 10·4 10·7 10·6 10·5 11·1 0·05 0·05 0·08 52·5 Na2O 3·28 3·01 3·36 3·93 3·44 3·58 2·50 3·21 3·35 2·80 3·69 3·60 0·51 0·42 0·30 K2O 0·97 1·56 0·57 0·07 0·90 0·10 1·81 1·03 0·30 1·72 0·36 0·14 8·45 8·20 7·64 P2O5 40·8 NiO 0·12 0·14 0·14 b.d. b.d. 0·14 0·16 b.d. 0·14 0·11 0·11 — 0·30 0·22 — Cl 0·07 0·09 0·07 0·04 0·04 b.d. 0·18 0·04 — — — — 0·27 0·10 0·08 0·29 F b.d. b.d. b.d. b.d. b.d. 0·13 b.d. 0·07 — — — — 0·07 0·12 0·12 0·98 Total 97·4 97·3 96·9 97·3 97·3 96·6 97·0 96·4 98·8 98·2 99·3 97·8 92·5 93·5 91·3 94·6 Mg# 80·6 79·9 86·7 85·6 84·9 84·8 78·7 72·0 75·7 75·6 83·5 84·8 78·9 73·5 65·6 Spinel Olivine Ilmenite Sample: BM96-2 BM99-6 BME-1 DR10165* GN35* BM96-2 BM99-5 Type: M M† M‡ M M† M† M M M E1 SiO2 0·06 0·07 0·07 0·09 0·04 — — — 39·5 0·0 TiO2 0·10 0·34 0·08 0·05 0·05 — — 0·12 b.d. 47·1 Al2O3 59·7 47·2 57·7 64·6 61·3 64·2 64·4 64·6 0·03 0·35 Cr2O3 3·74 14·24 5·55 1·94 4·23 2·84 2·31 3·49 b.d. 0·51 FeO 17·8 22·0 18·3 11·3 13·8 13·7 13·7 13·0 18·0 39·4 MnO 0·08 0·11 0·10 0·07 0·06 b.d. 0·05 0·14 0·15 0·35 MgO 17·3 14·7 16·7 20·5 19·1 19·5 19·6 19·6 41·6 7·0 CaO b.d. 0·02 b.d. 0·05 b.d. — — — 0·07 0·08 ZnO 0·23 0·18 0·24 0·13 0·36 — — — — — NiO 0·59 0·42 0·51 0·65 0·65 0·27 0·24 0·33 0·48 0·15 Total 99·6 99·5 99·3 99·3 99·6 100·5 100·3 101·2 99·9 95·3 Mg# 63·3 54·4 62·0 76·3 71·2 71·7 71·9 72·9 80·4 24·0 Cr# 4·04 16·83 6·06 1·97 4·42 2·88 2·35 3·50 Amphibole Phlogopite Ap Sample: BM96-2 BM99-6 BM99-4 BME-1 BM96-3 BM99-5 DR9748* GN35 DR10165* BM96-3 BME-1 Type: M vein M M vein E1 M E1 M vein M M Core Rim1 Rim2 M SiO2 42·2 41·8 42·1 41·9 42·7 42·0 41·9 40·9 42·0 42·1 42·9 42·5 36·3 35·3 33·5 TiO2 1·57 0·92 1·47 2·56 0·84 1·63 3·97 2·24 3·47 1·62 2·60 2·09 6·50 10·2 11·5 Al2O3 15·5 15·9 16·2 15·6 15·8 15·7 13·4 14·2 14·8 15·6 16·0 15·9 15·5 15·6 15·7 Cr2O3 0·29 0·26 0·24 0·43 0·35 0·35 0·13 0·42 0·54 0·40 0·34 0·29 0·14 0·20 — FeO 6·86 7·14 4·65 5·03 5·34 5·33 6·97 9·72 8·54 8·46 5·92 5·38 7·90 9·01 10·8 MnO 0·06 0·06 0·04 0·07 0·10 0·05 0·05 0·10 — 0·09 0·10 — 0·04 0·06 0·08 MgO 16·0 15·9 17·1 16·7 16·9 16·7 14·5 14·0 14·9 14·7 16·8 16·8 16·6 14·0 11·5 CaO 10·6 10·5 11·0 11·0 11·0 11·0 11·4 10·4 10·7 10·6 10·5 11·1 0·05 0·05 0·08 52·5 Na2O 3·28 3·01 3·36 3·93 3·44 3·58 2·50 3·21 3·35 2·80 3·69 3·60 0·51 0·42 0·30 K2O 0·97 1·56 0·57 0·07 0·90 0·10 1·81 1·03 0·30 1·72 0·36 0·14 8·45 8·20 7·64 P2O5 40·8 NiO 0·12 0·14 0·14 b.d. b.d. 0·14 0·16 b.d. 0·14 0·11 0·11 — 0·30 0·22 — Cl 0·07 0·09 0·07 0·04 0·04 b.d. 0·18 0·04 — — — — 0·27 0·10 0·08 0·29 F b.d. b.d. b.d. b.d. b.d. 0·13 b.d. 0·07 — — — — 0·07 0·12 0·12 0·98 Total 97·4 97·3 96·9 97·3 97·3 96·6 97·0 96·4 98·8 98·2 99·3 97·8 92·5 93·5 91·3 94·6 Mg# 80·6 79·9 86·7 85·6 84·9 84·8 78·7 72·0 75·7 75·6 83·5 84·8 78·9 73·5 65·6 Spinel Olivine Ilmenite Sample: BM96-2 BM99-6 BME-1 DR10165* GN35* BM96-2 BM99-5 Type: M M† M‡ M M† M† M M M E1 SiO2 0·06 0·07 0·07 0·09 0·04 — — — 39·5 0·0 TiO2 0·10 0·34 0·08 0·05 0·05 — — 0·12 b.d. 47·1 Al2O3 59·7 47·2 57·7 64·6 61·3 64·2 64·4 64·6 0·03 0·35 Cr2O3 3·74 14·24 5·55 1·94 4·23 2·84 2·31 3·49 b.d. 0·51 FeO 17·8 22·0 18·3 11·3 13·8 13·7 13·7 13·0 18·0 39·4 MnO 0·08 0·11 0·10 0·07 0·06 b.d. 0·05 0·14 0·15 0·35 MgO 17·3 14·7 16·7 20·5 19·1 19·5 19·6 19·6 41·6 7·0 CaO b.d. 0·02 b.d. 0·05 b.d. — — — 0·07 0·08 ZnO 0·23 0·18 0·24 0·13 0·36 — — — — — NiO 0·59 0·42 0·51 0·65 0·65 0·27 0·24 0·33 0·48 0·15 Total 99·6 99·5 99·3 99·3 99·6 100·5 100·3 101·2 99·9 95·3 Mg# 63·3 54·4 62·0 76·3 71·2 71·7 71·9 72·9 80·4 24·0 Cr# 4·04 16·83 6·06 1·97 4·42 2·88 2·35 3·50 M, minerals in the mosaic area; H, host minerals with exsolution textures; E1, minerals as exsolved lamellae in Cpx megacryst; E2, mineral as exsolved lamellae in Opx megacryst; C, Grt in the coarse area; b.d., below detection limit; —, no value. * Data from Griffin et al. (1984) and O’Reilly & Griffin (1985). †Grt -rimmed spinel. ‡ Amp-rimmed spinel. Table 2: Representative major element compositions of minerals in the garnet pyroxenites from Lakes Bullenmerri and Gnotuk Clinopyroxene Sample: BM 96-2 BM 99-6 BM 99-4 BME-1 BM 99-2 BM 96-3 BM 99-5 DR 9748* GN35* DR10165* Type: M M H M H M H M M H M H M H M SiO2 51·2 51·3 51·0 51·4 51·5 51·8 50·9 51·2 51·4 51·3 50·9 51·1 51·4 51·1 51·4 TiO2 0·37 0·36 0·73 0·79 0·43 0·42 0·40 0·40 1·02 0·80 0·88 0·68 0·62 0·53 0·51 Al2O3 6·62 6·75 7·14 7·51 6·33 6·42 6·43 6·34 6·20 4·87 5·98 7·29 7·11 6·5 6·02 Cr2O3 0·28 0·17 0·30 0·30 0·24 0·26 0·19 0·22 0·11 0·32 0·44 0·37 0·2 0·2 0·17 FeO 5·12 3·25 3·32 3·29 3·58 3·60 4·56 4·43 5·00 6·67 5·86 3·47 3·77 3·61 3·77 MnO 0·09 0·07 0·07 0·09 0·06 0·06 0·08 0·08 0·09 0·13 0·13 0·13 0·12 0·10 MgO 14·5 15·4 14·7 14·3 14·9 14·6 15·0 14·9 14·8 14·2 14·5 14·3 14·5 14·8 15·2 CaO 20·3 21·6 20·8 20·9 21·7 22·0 21·0 21·0 21·1 20·3 21·1 20·9 20·5 22·2 21·5 Na2O 1·55 1·28 1·51 1·58 1·17 1·19 1·41 1·43 1·16 1·26 1·03 1·49 1·5 1·03 1·00 NiO 0·08 0·08 b.d. b.d. 0·07 0·07 b.d. b.d. 0·09 0·08 — — — — — Total 100·1 100·2 99·6 100·2 99·9 100·5 99·9 100·1 100·8 99·9 100·8 99·7 99·8 99·9 99·7 Mg# 83·6 89·5 88·8 88·7 88·2 88·0 85·5 85·8 84·1 79·3 81·5 88·1 87·4 88·1 87·9 Orthopyroxene Sample: BM96-2 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748* GN35* DR10165* Type: M E1 M H E1 M E1 M M E1 M E1 M E1 M SiO2 53·3 54·5 54·8 55·0 55·2 54·4 53·5 53·4 54·9 53·3 53·6 54·2 54·3 54·2 54·4 TiO2 0·08 0·11 0·10 0·12 0·07 0·08 0·06 0·05 0·22 0·25 0·15 0·12 0·10 — 0·10 Al2O3 4·63 4·61 4·72 4·79 4·61 4·60 4·70 4·60 4·12 3·69 4·07 4·99 4·92 4·84 4·77 Cr2O3 0·14 0·14 0·14 0·15 0·12 0·12 0·10 0·11 0·06 0·16 0·21 0·19 0·09 0·1 0·08 FeO 11·6 8·44 9·18 8·57 9·41 9·43 10·5 10·6 11·7 15·4 13·1 8·47 8·71 9·30 9·45 MnO 0·16 0·17 0·19 0·19 0·12 0·09 0·15 0·16 0·15 0·25 0·22 0·21 0·22 0·12 0·16 MgO 29·7 31·2 31·4 31·7 31·1 30·5 30·6 30·9 29·2 26·6 28·6 31·1 31·0 30·7 30·9 CaO 0·75 0·61 0·52 0·51 0·55 0·60 0·67 0·68 0·77 0·86 0·71 0·69 0·63 0·48 0·56 Na2O 0·09 0·08 0·10 0·07 0·05 b.d. 0·09 0·08 0·08 0·11 0·08 — 0·08 — 0·06 NiO 0·12 b.d. b.d. b.d. 0·12 0·13 0·11 0·11 0·14 0·10 — — — — — Total 100·6 100·0 101·1 101·1 101·4 100·0 100·5 100·5 101·4 100·7 100·8 100·0 100·0 99·7 100·5 Mg# 82·0 86·8 85·9 86·8 85·5 85·2 83·8 83·9 81·7 75·5 79·5 86·7 86·4 85·5 85·3 Garnet Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748* GN35* DR10165* Type: C C E1 E2 M E1 C E1 M C C E1 M E1 C M SiO2 41·2 41·8 40·9 40·9 41·0 41·9 41·6 41·4 41·5 41·3 41·7 40·6 41·1 41·9 41·8 41·8 TiO2 0·08 0·04 0·07 0·08 0·07 0·04 0·04 0·04 0·07 0·04 0·14 0·17 0·10 — — — Al2O3 23·2 24·2 23·5 23·5 23·5 23·3 23·6 23·8 23·5 23·9 23·0 22·6 22·9 23·7 23·4 24·0 Cr2O3 0·45 0·18 0·24 0·25 0·25 0·27 0·24 0·21 0·23 0·26 0·12 0·44 0·64 0·32 0·52 0·17 FeO 13·3 9·3 10·7 10·5 10·8 10·9 11·7 12·3 12·6 11·9 13·7 16·6 15·0 10·1 10·6 11·6 MnO 0·41 0·31 0·50 0·51 0·52 0·25 0·30 0·42 0·42 0·42 0·42 0·57 0·56 0·54 0·56 0·44 MgO 17·4 19·6 19·1 19·3 19·0 17·8 18·6 17·8 18·3 17·7 15·9 14·6 15·5 18·6 18·4 17·7 CaO 5·30 5·47 4·28 4·20 4·24 5·80 5·42 5·46 5·39 5·46 5·61 5·79 5·81 5·23 5·16 5·53 Na2O b.d. b.d. b.d. b.d. b.d. 0·24 b.d. b.d. b.d. b.d. b.d. 0·02 — — — — Total 101·3 101·0 99·3 99·2 99·5 100·4 101·5 101·3 102·0 101·0 100·5 101·4 101·6 100·4 100·4 101·3 Mg# 70·2 79·1 76·2 76·7 76·0 74·7 74·1 72·3 72·4 72·9 67·6 61·4 64·7 76·6 75·6 73·1 Clinopyroxene Sample: BM 96-2 BM 99-6 BM 99-4 BME-1 BM 99-2 BM 96-3 BM 99-5 DR 9748* GN35* DR10165* Type: M M H M H M H M M H M H M H M SiO2 51·2 51·3 51·0 51·4 51·5 51·8 50·9 51·2 51·4 51·3 50·9 51·1 51·4 51·1 51·4 TiO2 0·37 0·36 0·73 0·79 0·43 0·42 0·40 0·40 1·02 0·80 0·88 0·68 0·62 0·53 0·51 Al2O3 6·62 6·75 7·14 7·51 6·33 6·42 6·43 6·34 6·20 4·87 5·98 7·29 7·11 6·5 6·02 Cr2O3 0·28 0·17 0·30 0·30 0·24 0·26 0·19 0·22 0·11 0·32 0·44 0·37 0·2 0·2 0·17 FeO 5·12 3·25 3·32 3·29 3·58 3·60 4·56 4·43 5·00 6·67 5·86 3·47 3·77 3·61 3·77 MnO 0·09 0·07 0·07 0·09 0·06 0·06 0·08 0·08 0·09 0·13 0·13 0·13 0·12 0·10 MgO 14·5 15·4 14·7 14·3 14·9 14·6 15·0 14·9 14·8 14·2 14·5 14·3 14·5 14·8 15·2 CaO 20·3 21·6 20·8 20·9 21·7 22·0 21·0 21·0 21·1 20·3 21·1 20·9 20·5 22·2 21·5 Na2O 1·55 1·28 1·51 1·58 1·17 1·19 1·41 1·43 1·16 1·26 1·03 1·49 1·5 1·03 1·00 NiO 0·08 0·08 b.d. b.d. 0·07 0·07 b.d. b.d. 0·09 0·08 — — — — — Total 100·1 100·2 99·6 100·2 99·9 100·5 99·9 100·1 100·8 99·9 100·8 99·7 99·8 99·9 99·7 Mg# 83·6 89·5 88·8 88·7 88·2 88·0 85·5 85·8 84·1 79·3 81·5 88·1 87·4 88·1 87·9 Orthopyroxene Sample: BM96-2 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748* GN35* DR10165* Type: M E1 M H E1 M E1 M M E1 M E1 M E1 M SiO2 53·3 54·5 54·8 55·0 55·2 54·4 53·5 53·4 54·9 53·3 53·6 54·2 54·3 54·2 54·4 TiO2 0·08 0·11 0·10 0·12 0·07 0·08 0·06 0·05 0·22 0·25 0·15 0·12 0·10 — 0·10 Al2O3 4·63 4·61 4·72 4·79 4·61 4·60 4·70 4·60 4·12 3·69 4·07 4·99 4·92 4·84 4·77 Cr2O3 0·14 0·14 0·14 0·15 0·12 0·12 0·10 0·11 0·06 0·16 0·21 0·19 0·09 0·1 0·08 FeO 11·6 8·44 9·18 8·57 9·41 9·43 10·5 10·6 11·7 15·4 13·1 8·47 8·71 9·30 9·45 MnO 0·16 0·17 0·19 0·19 0·12 0·09 0·15 0·16 0·15 0·25 0·22 0·21 0·22 0·12 0·16 MgO 29·7 31·2 31·4 31·7 31·1 30·5 30·6 30·9 29·2 26·6 28·6 31·1 31·0 30·7 30·9 CaO 0·75 0·61 0·52 0·51 0·55 0·60 0·67 0·68 0·77 0·86 0·71 0·69 0·63 0·48 0·56 Na2O 0·09 0·08 0·10 0·07 0·05 b.d. 0·09 0·08 0·08 0·11 0·08 — 0·08 — 0·06 NiO 0·12 b.d. b.d. b.d. 0·12 0·13 0·11 0·11 0·14 0·10 — — — — — Total 100·6 100·0 101·1 101·1 101·4 100·0 100·5 100·5 101·4 100·7 100·8 100·0 100·0 99·7 100·5 Mg# 82·0 86·8 85·9 86·8 85·5 85·2 83·8 83·9 81·7 75·5 79·5 86·7 86·4 85·5 85·3 Garnet Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748* GN35* DR10165* Type: C C E1 E2 M E1 C E1 M C C E1 M E1 C M SiO2 41·2 41·8 40·9 40·9 41·0 41·9 41·6 41·4 41·5 41·3 41·7 40·6 41·1 41·9 41·8 41·8 TiO2 0·08 0·04 0·07 0·08 0·07 0·04 0·04 0·04 0·07 0·04 0·14 0·17 0·10 — — — Al2O3 23·2 24·2 23·5 23·5 23·5 23·3 23·6 23·8 23·5 23·9 23·0 22·6 22·9 23·7 23·4 24·0 Cr2O3 0·45 0·18 0·24 0·25 0·25 0·27 0·24 0·21 0·23 0·26 0·12 0·44 0·64 0·32 0·52 0·17 FeO 13·3 9·3 10·7 10·5 10·8 10·9 11·7 12·3 12·6 11·9 13·7 16·6 15·0 10·1 10·6 11·6 MnO 0·41 0·31 0·50 0·51 0·52 0·25 0·30 0·42 0·42 0·42 0·42 0·57 0·56 0·54 0·56 0·44 MgO 17·4 19·6 19·1 19·3 19·0 17·8 18·6 17·8 18·3 17·7 15·9 14·6 15·5 18·6 18·4 17·7 CaO 5·30 5·47 4·28 4·20 4·24 5·80 5·42 5·46 5·39 5·46 5·61 5·79 5·81 5·23 5·16 5·53 Na2O b.d. b.d. b.d. b.d. b.d. 0·24 b.d. b.d. b.d. b.d. b.d. 0·02 — — — — Total 101·3 101·0 99·3 99·2 99·5 100·4 101·5 101·3 102·0 101·0 100·5 101·4 101·6 100·4 100·4 101·3 Mg# 70·2 79·1 76·2 76·7 76·0 74·7 74·1 72·3 72·4 72·9 67·6 61·4 64·7 76·6 75·6 73·1 Amphibole Phlogopite Ap Sample: BM96-2 BM99-6 BM99-4 BME-1 BM96-3 BM99-5 DR9748* GN35 DR10165* BM96-3 BME-1 Type: M vein M M vein E1 M E1 M vein M M Core Rim1 Rim2 M SiO2 42·2 41·8 42·1 41·9 42·7 42·0 41·9 40·9 42·0 42·1 42·9 42·5 36·3 35·3 33·5 TiO2 1·57 0·92 1·47 2·56 0·84 1·63 3·97 2·24 3·47 1·62 2·60 2·09 6·50 10·2 11·5 Al2O3 15·5 15·9 16·2 15·6 15·8 15·7 13·4 14·2 14·8 15·6 16·0 15·9 15·5 15·6 15·7 Cr2O3 0·29 0·26 0·24 0·43 0·35 0·35 0·13 0·42 0·54 0·40 0·34 0·29 0·14 0·20 — FeO 6·86 7·14 4·65 5·03 5·34 5·33 6·97 9·72 8·54 8·46 5·92 5·38 7·90 9·01 10·8 MnO 0·06 0·06 0·04 0·07 0·10 0·05 0·05 0·10 — 0·09 0·10 — 0·04 0·06 0·08 MgO 16·0 15·9 17·1 16·7 16·9 16·7 14·5 14·0 14·9 14·7 16·8 16·8 16·6 14·0 11·5 CaO 10·6 10·5 11·0 11·0 11·0 11·0 11·4 10·4 10·7 10·6 10·5 11·1 0·05 0·05 0·08 52·5 Na2O 3·28 3·01 3·36 3·93 3·44 3·58 2·50 3·21 3·35 2·80 3·69 3·60 0·51 0·42 0·30 K2O 0·97 1·56 0·57 0·07 0·90 0·10 1·81 1·03 0·30 1·72 0·36 0·14 8·45 8·20 7·64 P2O5 40·8 NiO 0·12 0·14 0·14 b.d. b.d. 0·14 0·16 b.d. 0·14 0·11 0·11 — 0·30 0·22 — Cl 0·07 0·09 0·07 0·04 0·04 b.d. 0·18 0·04 — — — — 0·27 0·10 0·08 0·29 F b.d. b.d. b.d. b.d. b.d. 0·13 b.d. 0·07 — — — — 0·07 0·12 0·12 0·98 Total 97·4 97·3 96·9 97·3 97·3 96·6 97·0 96·4 98·8 98·2 99·3 97·8 92·5 93·5 91·3 94·6 Mg# 80·6 79·9 86·7 85·6 84·9 84·8 78·7 72·0 75·7 75·6 83·5 84·8 78·9 73·5 65·6 Spinel Olivine Ilmenite Sample: BM96-2 BM99-6 BME-1 DR10165* GN35* BM96-2 BM99-5 Type: M M† M‡ M M† M† M M M E1 SiO2 0·06 0·07 0·07 0·09 0·04 — — — 39·5 0·0 TiO2 0·10 0·34 0·08 0·05 0·05 — — 0·12 b.d. 47·1 Al2O3 59·7 47·2 57·7 64·6 61·3 64·2 64·4 64·6 0·03 0·35 Cr2O3 3·74 14·24 5·55 1·94 4·23 2·84 2·31 3·49 b.d. 0·51 FeO 17·8 22·0 18·3 11·3 13·8 13·7 13·7 13·0 18·0 39·4 MnO 0·08 0·11 0·10 0·07 0·06 b.d. 0·05 0·14 0·15 0·35 MgO 17·3 14·7 16·7 20·5 19·1 19·5 19·6 19·6 41·6 7·0 CaO b.d. 0·02 b.d. 0·05 b.d. — — — 0·07 0·08 ZnO 0·23 0·18 0·24 0·13 0·36 — — — — — NiO 0·59 0·42 0·51 0·65 0·65 0·27 0·24 0·33 0·48 0·15 Total 99·6 99·5 99·3 99·3 99·6 100·5 100·3 101·2 99·9 95·3 Mg# 63·3 54·4 62·0 76·3 71·2 71·7 71·9 72·9 80·4 24·0 Cr# 4·04 16·83 6·06 1·97 4·42 2·88 2·35 3·50 Amphibole Phlogopite Ap Sample: BM96-2 BM99-6 BM99-4 BME-1 BM96-3 BM99-5 DR9748* GN35 DR10165* BM96-3 BME-1 Type: M vein M M vein E1 M E1 M vein M M Core Rim1 Rim2 M SiO2 42·2 41·8 42·1 41·9 42·7 42·0 41·9 40·9 42·0 42·1 42·9 42·5 36·3 35·3 33·5 TiO2 1·57 0·92 1·47 2·56 0·84 1·63 3·97 2·24 3·47 1·62 2·60 2·09 6·50 10·2 11·5 Al2O3 15·5 15·9 16·2 15·6 15·8 15·7 13·4 14·2 14·8 15·6 16·0 15·9 15·5 15·6 15·7 Cr2O3 0·29 0·26 0·24 0·43 0·35 0·35 0·13 0·42 0·54 0·40 0·34 0·29 0·14 0·20 — FeO 6·86 7·14 4·65 5·03 5·34 5·33 6·97 9·72 8·54 8·46 5·92 5·38 7·90 9·01 10·8 MnO 0·06 0·06 0·04 0·07 0·10 0·05 0·05 0·10 — 0·09 0·10 — 0·04 0·06 0·08 MgO 16·0 15·9 17·1 16·7 16·9 16·7 14·5 14·0 14·9 14·7 16·8 16·8 16·6 14·0 11·5 CaO 10·6 10·5 11·0 11·0 11·0 11·0 11·4 10·4 10·7 10·6 10·5 11·1 0·05 0·05 0·08 52·5 Na2O 3·28 3·01 3·36 3·93 3·44 3·58 2·50 3·21 3·35 2·80 3·69 3·60 0·51 0·42 0·30 K2O 0·97 1·56 0·57 0·07 0·90 0·10 1·81 1·03 0·30 1·72 0·36 0·14 8·45 8·20 7·64 P2O5 40·8 NiO 0·12 0·14 0·14 b.d. b.d. 0·14 0·16 b.d. 0·14 0·11 0·11 — 0·30 0·22 — Cl 0·07 0·09 0·07 0·04 0·04 b.d. 0·18 0·04 — — — — 0·27 0·10 0·08 0·29 F b.d. b.d. b.d. b.d. b.d. 0·13 b.d. 0·07 — — — — 0·07 0·12 0·12 0·98 Total 97·4 97·3 96·9 97·3 97·3 96·6 97·0 96·4 98·8 98·2 99·3 97·8 92·5 93·5 91·3 94·6 Mg# 80·6 79·9 86·7 85·6 84·9 84·8 78·7 72·0 75·7 75·6 83·5 84·8 78·9 73·5 65·6 Spinel Olivine Ilmenite Sample: BM96-2 BM99-6 BME-1 DR10165* GN35* BM96-2 BM99-5 Type: M M† M‡ M M† M† M M M E1 SiO2 0·06 0·07 0·07 0·09 0·04 — — — 39·5 0·0 TiO2 0·10 0·34 0·08 0·05 0·05 — — 0·12 b.d. 47·1 Al2O3 59·7 47·2 57·7 64·6 61·3 64·2 64·4 64·6 0·03 0·35 Cr2O3 3·74 14·24 5·55 1·94 4·23 2·84 2·31 3·49 b.d. 0·51 FeO 17·8 22·0 18·3 11·3 13·8 13·7 13·7 13·0 18·0 39·4 MnO 0·08 0·11 0·10 0·07 0·06 b.d. 0·05 0·14 0·15 0·35 MgO 17·3 14·7 16·7 20·5 19·1 19·5 19·6 19·6 41·6 7·0 CaO b.d. 0·02 b.d. 0·05 b.d. — — — 0·07 0·08 ZnO 0·23 0·18 0·24 0·13 0·36 — — — — — NiO 0·59 0·42 0·51 0·65 0·65 0·27 0·24 0·33 0·48 0·15 Total 99·6 99·5 99·3 99·3 99·6 100·5 100·3 101·2 99·9 95·3 Mg# 63·3 54·4 62·0 76·3 71·2 71·7 71·9 72·9 80·4 24·0 Cr# 4·04 16·83 6·06 1·97 4·42 2·88 2·35 3·50 M, minerals in the mosaic area; H, host minerals with exsolution textures; E1, minerals as exsolved lamellae in Cpx megacryst; E2, mineral as exsolved lamellae in Opx megacryst; C, Grt in the coarse area; b.d., below detection limit; —, no value. * Data from Griffin et al. (1984) and O’Reilly & Griffin (1985). †Grt -rimmed spinel. ‡ Amp-rimmed spinel. In the garnet pyroxenites studied here, Cpx is diopside (En40–49Fs1–8Wo46–52) and has lower Mg# (79·3–89·5) and Cr2O3 (0·11–0·44 wt %) than the clinopyroxene of peridotite xenoliths from the same maars (Fig. 5a), but slightly higher Al2O3 (4·87–7·51 wt %) and similar CaO contents (Fig. 5b). Opx is enstatite (En75–87Fs12–23) and shows a positive correlation between MgO and Al2O3, but a negative correlation between MgO and CaO contents (Fig. 5c and d). Opx compositions are clearly correlated with those of coexisting Cpx. Grt is composed mainly of pyrope (55–71%), almandine (19–30%) and grossular (11–14%), with small amounts of spessartine (∼1%). It shows nearly constant CaO (5·26 wt % on average) contents with variable Cr2O3 contents (0·12–0·45 wt %) (Fig. 5e). Their Mg# and MgO contents show obvious positive correlations with those of coexisting Cpx. Spl is dominantly Mg–Al pleonaste and has lower Mg# at given Al2O3 content (or Cr#) than those from spinel pyroxenite xenoliths, which plot along the melt–peridotite reaction trend defined by composite xenoliths of spinel pyroxenite and spinel lherzolite from this area (Fig. 5f;Griffin et al., 1984). Spinel also has slightly lower Cr2O3 but higher Ni contents at given Al2O3 than those from the spinel pyroxenites and peridotites. In addition, in sample BM96-2, spinel compositions show heterogeneity related to microstructure (Table 2): the Grt-rimming spinel has the highest Cr2O3 (∼14·2 wt %), whereas Amp-rimming spinel has intermediate Cr2O3 (∼5·55 wt %) and discrete spinel grains have the lowest Cr2O3 (∼3·74 wt %). These heterogeneous compositions are different from those produced by subsolidus reaction between spinel pyroxenite and peridotite (Fig. 5f). Sample BM96-2 also contains minor olivine (Ol), which has a much lower Mg# (∼80·4), but higher NiO (∼0·46 wt %), compared with olivine in peridotites (87·1–91·3 and 0·34–0·42 wt %, respectively). Ilmenite was found as exsolution lamellae in one Cpx megacryst; it has an MgO content of ∼7·0 wt %. Fig. 5. View largeDownload slide Compositional variations of Cpx (a, b), Opx (c, d), Grt (e), and Spl (f) in garnet pyroxenites. Spinel-pyroxenitic and peridotitic minerals from this locality are shown for comparison (Griffin et al., 1984; Powell, 2005). Dashed line in (f) represents melt–peridotite reaction observed in composite xenoliths of spinel pyroxenite and lherzolite (GN23; Griffin et al., 1984); dotted line links Spl of sample BM96-2. Fig. 5. View largeDownload slide Compositional variations of Cpx (a, b), Opx (c, d), Grt (e), and Spl (f) in garnet pyroxenites. Spinel-pyroxenitic and peridotitic minerals from this locality are shown for comparison (Griffin et al., 1984; Powell, 2005). Dashed line in (f) represents melt–peridotite reaction observed in composite xenoliths of spinel pyroxenite and lherzolite (GN23; Griffin et al., 1984); dotted line links Spl of sample BM96-2. Amphiboles are all pargasites (Fig. 6a), but differ from the pargasitic amphiboles of peridotite xenoliths, which have higher Mg# and Cr2O3 contents (Fig. 6b). No correlation has been observed between their SiO2 content and that of other oxides (e.g. Al2O3; Fig. 6c). The discrete Amp tends to have higher TiO2 but lower K2O than vein-forming Amp in samples BM96-2 and DR9748 (Fig. 6d). In sample BM96-3, Amp associated with Phl has the highest TiO2 (∼3·97 wt %) and K2O (∼1·81 wt %), and is essentially identical to the amphiboles in the late-stage wehrlite suite of xenoliths from the same locality (Griffin et al., 1984). Phlogopite is found in one sample (BM96-3) and is inferred to be secondary, considering its microstructural position (Fig. 4e and f). It has homogeneous cores with moderate TiO2 (∼6·50 wt %) and Mg# (∼79·1), whereas rims have higher TiO2 (10·2–11·5 wt %) and lower Mg# (65·6–73·5), which represents a transformation to biotite. Rare apatite (Ap) in BME-1 has a F (0·98 wt %) content higher than Cl (0·29%) content. Fig. 6. View largeDownload slide Compositional variations of Amp in the garnet pyroxenite xenoliths from Lakes Bullenmerri and Gnotuk. Amp from spinel pyroxenite and peridotite in this locality are shown for comparison (Griffin et al., 1984; Powell, 2005). Grey shaded areas represent Amp from wehrlite xenoliths in this area (Griffin et al., 1984). Fig. 6. View largeDownload slide Compositional variations of Amp in the garnet pyroxenite xenoliths from Lakes Bullenmerri and Gnotuk. Amp from spinel pyroxenite and peridotite in this locality are shown for comparison (Griffin et al., 1984; Powell, 2005). Grey shaded areas represent Amp from wehrlite xenoliths in this area (Griffin et al., 1984). Trace element compositions The trace element compositions of minerals are listed in Tables 3–5. In most samples, no compositional gradients are observed in Grt and Cpx, except for sample DR10165, which shows obvious differences in heavy rare earth element (HREE) contents and normalized REE patterns related to their microstructural positions (Fig. 7). Table 3: Trace element compositions (ppm) of clinopyroxene in the garnet pyroxenites from Lakes Bullenmerri and Gnotuk Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: M M H M H M H M M H M H M H M1 M2 Li 3·24 1·57 4·62 4·04 2·15 2·58 2·13 2·12 3·09 4·07 3·84 2·21 2·04 — 2·76 2·81 Be 0·17 0·13 0·25 0·28 0·29 0·29 0·11 0·09 0·23 0·48 0·32 b.d. b.d. — b.d. b.d. B 1·87 2·10 10·06 8·31 2·46 1·53 1·73 2·42 1·74 1·74 0·76 9·80 8·60 — 10·09 10·51 Sc 28·4 17·2 37·0 39·6 22·0 30·7 26·9 34·0 34·3 54·6 49·1 43·0 50·5 29·8 31·6 39·2 Ti 2285 2450 4066 4259 2478 2632 2302 2337 6059 4539 5280 4212 3597 3127 3077 3240 V 177 187 351 346 230 241 244 254 487 392 358 277 263 198 200 191 Cr 1816 1173 1794 1924 1407 1601 1245 1455 597 2047 2450 2397 1479 1334 1226 1233 Co 41·0 27·7 25·6 23·9 34·6 36·1 37·6 37·0 43·8 45·8 46·6 23·9 25·6 24·6 23·5 23·2 Ni 472 439 227 227 428 444 410 387 507 385 511 304 314 294 287 259 Cu 1·00 1·82 1·44 1·26 0·75 0·75 1·49 1·41 1·82 1·49 3·47 1·98 1·98 1·05 1·03 1·01 Zn 21·1 6·4 14·5 11·8 18·7 18·6 24·3 23·8 33·7 52·3 47·2 6·6 6·4 6·1 6·3 6·4 Ga 8·74 9·78 11·08 10·59 9·73 10·44 12·45 12·43 16·80 11·47 12·11 5·26 4·09 4·33 3·76 2·94 Rb b.d. b.d. 0·04 b.d. b.d. b.d. 0·03 b.d. b.d. b.d. b.d. b.d. b.d. b.d. b.d. 0·04 Sr 146 137 149 147 126 125 114 115 104 109 135 95 97 128 138 138 Y 1·56 0·72 5·07 5·62 1·68 1·99 1·21 1·36 3·06 6·54 6·23 6·28 7·68 2·82 3·79 6·70 Zr 6·64 14·27 25·35 26·01 16·44 19·90 5·49 6·23 47·95 30·48 32·16 18·51 18·07 21·85 22·78 27·70 Nb 0·14 0·55 0·20 0·17 0·08 0·06 0·18 0·22 0·86 0·07 0·45 0·20 0·22 0·08 0·12 0·15 Ba 0·05 0·12 0·09 0·05 0·09 0·07 0·04 0·05 0·12 0·06 0·07 0·09 0·11 0·11 0·12 0·21 La 2·64 2·24 4·98 4·66 4·03 3·72 1·71 1·74 4·45 2·61 3·77 4·80 4·80 8·07 8·47 8·76 Ce 3·64 4·43 11·73 10·50 16·17 15·82 3·53 3·51 10·54 6·38 10·95 9·49 9·66 24·72 27·28 26·93 Pr 0·46 0·60 1·72 1·50 3·09 3·03 0·48 0·48 1·47 1·13 1·87 1·17 1·16 3·62 4·00 4·06 Nd 2·34 3·03 8·29 7·33 14·89 15·23 2·42 2·41 6·82 6·69 9·59 5·46 5·32 15·28 16·80 17·29 Sm 0·71 0·81 2·32 2·12 2·68 2·97 0·72 0·72 2·15 2·23 2·60 1·69 1·66 2·49 2·78 2·99 Eu 0·31 0·33 0·91 0·89 0·71 0·69 0·26 0·30 0·73 0·79 0·89 0·70 0·71 0·74 0·83 0·90 Gd 0·74 0·64 2·31 2·20 1·49 1·35 0·64 0·68 2·17 2·43 2·44 1·91 1·99 1·66 1·90 2·36 Tb 0·09 0·07 0·30 0·31 0·13 0·15 0·08 0·08 0·25 0·33 0·31 0·23 0·40 b.d. 0·25 0·32 Dy 0·47 0·30 1·78 1·67 0·57 0·61 0·40 0·43 1·06 1·73 1·72 1·58 1·94 0·88 1·10 1·71 Ho 0·07 0·03 0·25 0·26 0·06 0·09 0·05 0·07 0·14 0·28 0·26 0·33 0·32 b.d. 0·16 0·27 Er 0·13 0·05 0·50 0·52 0·12 0·15 0·10 0·11 0·22 0·56 0·52 0·54 0·69 0·22 0·30 0·61 Tm 0·01 b.d. 0·05 0·05 0·01 0·01 0·01 0·01 0·02 0·07 0·06 0·09 0·07 b.d. 0·03 0·07 Yb 0·07 b.d. 0·24 0·27 0·05 0·07 0·04 0·07 0·09 0·35 0·34 0·32 0·38 0·10 0·12 0·35 Lu 0·01 0·01 0·03 0·03 b.d. 0·01 0·01 0·01 0·01 0·04 0·04 0·04 0·04 0·01 0·01 0·04 Hf 0·35 0·58 1·18 1·08 0·49 0·59 0·29 0·36 2·50 1·38 1·41 0·67 0·68 0·61 0·63 0·67 Ta 0·02 0·14 0·02 0·02 0·03 0·02 0·03 0·03 0·05 0·02 0·04 0·01 0·01 0·02 0·02 0·02 Pb 0·42 0·20 0·75 0·69 0·50 0·52 0·22 0·20 0·34 0·29 0·47 0·43 0·42 0·54 0·55 0·54 Th 0·55 0·29 0·71 0·72 0·22 0·22 0·31 0·28 0·56 0·63 0·76 0·61 0·60 0·46 0·43 0·46 U 0·13 0·06 0·17 0·19 0·06 0·04 0·07 0·07 0·13 0·15 0·15 0·14 0·14 0·10 0·09 0·09 Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: M M H M H M H M M H M H M H M1 M2 Li 3·24 1·57 4·62 4·04 2·15 2·58 2·13 2·12 3·09 4·07 3·84 2·21 2·04 — 2·76 2·81 Be 0·17 0·13 0·25 0·28 0·29 0·29 0·11 0·09 0·23 0·48 0·32 b.d. b.d. — b.d. b.d. B 1·87 2·10 10·06 8·31 2·46 1·53 1·73 2·42 1·74 1·74 0·76 9·80 8·60 — 10·09 10·51 Sc 28·4 17·2 37·0 39·6 22·0 30·7 26·9 34·0 34·3 54·6 49·1 43·0 50·5 29·8 31·6 39·2 Ti 2285 2450 4066 4259 2478 2632 2302 2337 6059 4539 5280 4212 3597 3127 3077 3240 V 177 187 351 346 230 241 244 254 487 392 358 277 263 198 200 191 Cr 1816 1173 1794 1924 1407 1601 1245 1455 597 2047 2450 2397 1479 1334 1226 1233 Co 41·0 27·7 25·6 23·9 34·6 36·1 37·6 37·0 43·8 45·8 46·6 23·9 25·6 24·6 23·5 23·2 Ni 472 439 227 227 428 444 410 387 507 385 511 304 314 294 287 259 Cu 1·00 1·82 1·44 1·26 0·75 0·75 1·49 1·41 1·82 1·49 3·47 1·98 1·98 1·05 1·03 1·01 Zn 21·1 6·4 14·5 11·8 18·7 18·6 24·3 23·8 33·7 52·3 47·2 6·6 6·4 6·1 6·3 6·4 Ga 8·74 9·78 11·08 10·59 9·73 10·44 12·45 12·43 16·80 11·47 12·11 5·26 4·09 4·33 3·76 2·94 Rb b.d. b.d. 0·04 b.d. b.d. b.d. 0·03 b.d. b.d. b.d. b.d. b.d. b.d. b.d. b.d. 0·04 Sr 146 137 149 147 126 125 114 115 104 109 135 95 97 128 138 138 Y 1·56 0·72 5·07 5·62 1·68 1·99 1·21 1·36 3·06 6·54 6·23 6·28 7·68 2·82 3·79 6·70 Zr 6·64 14·27 25·35 26·01 16·44 19·90 5·49 6·23 47·95 30·48 32·16 18·51 18·07 21·85 22·78 27·70 Nb 0·14 0·55 0·20 0·17 0·08 0·06 0·18 0·22 0·86 0·07 0·45 0·20 0·22 0·08 0·12 0·15 Ba 0·05 0·12 0·09 0·05 0·09 0·07 0·04 0·05 0·12 0·06 0·07 0·09 0·11 0·11 0·12 0·21 La 2·64 2·24 4·98 4·66 4·03 3·72 1·71 1·74 4·45 2·61 3·77 4·80 4·80 8·07 8·47 8·76 Ce 3·64 4·43 11·73 10·50 16·17 15·82 3·53 3·51 10·54 6·38 10·95 9·49 9·66 24·72 27·28 26·93 Pr 0·46 0·60 1·72 1·50 3·09 3·03 0·48 0·48 1·47 1·13 1·87 1·17 1·16 3·62 4·00 4·06 Nd 2·34 3·03 8·29 7·33 14·89 15·23 2·42 2·41 6·82 6·69 9·59 5·46 5·32 15·28 16·80 17·29 Sm 0·71 0·81 2·32 2·12 2·68 2·97 0·72 0·72 2·15 2·23 2·60 1·69 1·66 2·49 2·78 2·99 Eu 0·31 0·33 0·91 0·89 0·71 0·69 0·26 0·30 0·73 0·79 0·89 0·70 0·71 0·74 0·83 0·90 Gd 0·74 0·64 2·31 2·20 1·49 1·35 0·64 0·68 2·17 2·43 2·44 1·91 1·99 1·66 1·90 2·36 Tb 0·09 0·07 0·30 0·31 0·13 0·15 0·08 0·08 0·25 0·33 0·31 0·23 0·40 b.d. 0·25 0·32 Dy 0·47 0·30 1·78 1·67 0·57 0·61 0·40 0·43 1·06 1·73 1·72 1·58 1·94 0·88 1·10 1·71 Ho 0·07 0·03 0·25 0·26 0·06 0·09 0·05 0·07 0·14 0·28 0·26 0·33 0·32 b.d. 0·16 0·27 Er 0·13 0·05 0·50 0·52 0·12 0·15 0·10 0·11 0·22 0·56 0·52 0·54 0·69 0·22 0·30 0·61 Tm 0·01 b.d. 0·05 0·05 0·01 0·01 0·01 0·01 0·02 0·07 0·06 0·09 0·07 b.d. 0·03 0·07 Yb 0·07 b.d. 0·24 0·27 0·05 0·07 0·04 0·07 0·09 0·35 0·34 0·32 0·38 0·10 0·12 0·35 Lu 0·01 0·01 0·03 0·03 b.d. 0·01 0·01 0·01 0·01 0·04 0·04 0·04 0·04 0·01 0·01 0·04 Hf 0·35 0·58 1·18 1·08 0·49 0·59 0·29 0·36 2·50 1·38 1·41 0·67 0·68 0·61 0·63 0·67 Ta 0·02 0·14 0·02 0·02 0·03 0·02 0·03 0·03 0·05 0·02 0·04 0·01 0·01 0·02 0·02 0·02 Pb 0·42 0·20 0·75 0·69 0·50 0·52 0·22 0·20 0·34 0·29 0·47 0·43 0·42 0·54 0·55 0·54 Th 0·55 0·29 0·71 0·72 0·22 0·22 0·31 0·28 0·56 0·63 0·76 0·61 0·60 0·46 0·43 0·46 U 0·13 0·06 0·17 0·19 0·06 0·04 0·07 0·07 0·13 0·15 0·15 0·14 0·14 0·10 0·09 0·09 M, Cpx in the mosaic area; H, host Cpx with exsolution textures; b.d., below detection limit; —, no value. Table 3: Trace element compositions (ppm) of clinopyroxene in the garnet pyroxenites from Lakes Bullenmerri and Gnotuk Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: M M H M H M H M M H M H M H M1 M2 Li 3·24 1·57 4·62 4·04 2·15 2·58 2·13 2·12 3·09 4·07 3·84 2·21 2·04 — 2·76 2·81 Be 0·17 0·13 0·25 0·28 0·29 0·29 0·11 0·09 0·23 0·48 0·32 b.d. b.d. — b.d. b.d. B 1·87 2·10 10·06 8·31 2·46 1·53 1·73 2·42 1·74 1·74 0·76 9·80 8·60 — 10·09 10·51 Sc 28·4 17·2 37·0 39·6 22·0 30·7 26·9 34·0 34·3 54·6 49·1 43·0 50·5 29·8 31·6 39·2 Ti 2285 2450 4066 4259 2478 2632 2302 2337 6059 4539 5280 4212 3597 3127 3077 3240 V 177 187 351 346 230 241 244 254 487 392 358 277 263 198 200 191 Cr 1816 1173 1794 1924 1407 1601 1245 1455 597 2047 2450 2397 1479 1334 1226 1233 Co 41·0 27·7 25·6 23·9 34·6 36·1 37·6 37·0 43·8 45·8 46·6 23·9 25·6 24·6 23·5 23·2 Ni 472 439 227 227 428 444 410 387 507 385 511 304 314 294 287 259 Cu 1·00 1·82 1·44 1·26 0·75 0·75 1·49 1·41 1·82 1·49 3·47 1·98 1·98 1·05 1·03 1·01 Zn 21·1 6·4 14·5 11·8 18·7 18·6 24·3 23·8 33·7 52·3 47·2 6·6 6·4 6·1 6·3 6·4 Ga 8·74 9·78 11·08 10·59 9·73 10·44 12·45 12·43 16·80 11·47 12·11 5·26 4·09 4·33 3·76 2·94 Rb b.d. b.d. 0·04 b.d. b.d. b.d. 0·03 b.d. b.d. b.d. b.d. b.d. b.d. b.d. b.d. 0·04 Sr 146 137 149 147 126 125 114 115 104 109 135 95 97 128 138 138 Y 1·56 0·72 5·07 5·62 1·68 1·99 1·21 1·36 3·06 6·54 6·23 6·28 7·68 2·82 3·79 6·70 Zr 6·64 14·27 25·35 26·01 16·44 19·90 5·49 6·23 47·95 30·48 32·16 18·51 18·07 21·85 22·78 27·70 Nb 0·14 0·55 0·20 0·17 0·08 0·06 0·18 0·22 0·86 0·07 0·45 0·20 0·22 0·08 0·12 0·15 Ba 0·05 0·12 0·09 0·05 0·09 0·07 0·04 0·05 0·12 0·06 0·07 0·09 0·11 0·11 0·12 0·21 La 2·64 2·24 4·98 4·66 4·03 3·72 1·71 1·74 4·45 2·61 3·77 4·80 4·80 8·07 8·47 8·76 Ce 3·64 4·43 11·73 10·50 16·17 15·82 3·53 3·51 10·54 6·38 10·95 9·49 9·66 24·72 27·28 26·93 Pr 0·46 0·60 1·72 1·50 3·09 3·03 0·48 0·48 1·47 1·13 1·87 1·17 1·16 3·62 4·00 4·06 Nd 2·34 3·03 8·29 7·33 14·89 15·23 2·42 2·41 6·82 6·69 9·59 5·46 5·32 15·28 16·80 17·29 Sm 0·71 0·81 2·32 2·12 2·68 2·97 0·72 0·72 2·15 2·23 2·60 1·69 1·66 2·49 2·78 2·99 Eu 0·31 0·33 0·91 0·89 0·71 0·69 0·26 0·30 0·73 0·79 0·89 0·70 0·71 0·74 0·83 0·90 Gd 0·74 0·64 2·31 2·20 1·49 1·35 0·64 0·68 2·17 2·43 2·44 1·91 1·99 1·66 1·90 2·36 Tb 0·09 0·07 0·30 0·31 0·13 0·15 0·08 0·08 0·25 0·33 0·31 0·23 0·40 b.d. 0·25 0·32 Dy 0·47 0·30 1·78 1·67 0·57 0·61 0·40 0·43 1·06 1·73 1·72 1·58 1·94 0·88 1·10 1·71 Ho 0·07 0·03 0·25 0·26 0·06 0·09 0·05 0·07 0·14 0·28 0·26 0·33 0·32 b.d. 0·16 0·27 Er 0·13 0·05 0·50 0·52 0·12 0·15 0·10 0·11 0·22 0·56 0·52 0·54 0·69 0·22 0·30 0·61 Tm 0·01 b.d. 0·05 0·05 0·01 0·01 0·01 0·01 0·02 0·07 0·06 0·09 0·07 b.d. 0·03 0·07 Yb 0·07 b.d. 0·24 0·27 0·05 0·07 0·04 0·07 0·09 0·35 0·34 0·32 0·38 0·10 0·12 0·35 Lu 0·01 0·01 0·03 0·03 b.d. 0·01 0·01 0·01 0·01 0·04 0·04 0·04 0·04 0·01 0·01 0·04 Hf 0·35 0·58 1·18 1·08 0·49 0·59 0·29 0·36 2·50 1·38 1·41 0·67 0·68 0·61 0·63 0·67 Ta 0·02 0·14 0·02 0·02 0·03 0·02 0·03 0·03 0·05 0·02 0·04 0·01 0·01 0·02 0·02 0·02 Pb 0·42 0·20 0·75 0·69 0·50 0·52 0·22 0·20 0·34 0·29 0·47 0·43 0·42 0·54 0·55 0·54 Th 0·55 0·29 0·71 0·72 0·22 0·22 0·31 0·28 0·56 0·63 0·76 0·61 0·60 0·46 0·43 0·46 U 0·13 0·06 0·17 0·19 0·06 0·04 0·07 0·07 0·13 0·15 0·15 0·14 0·14 0·10 0·09 0·09 Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: M M H M H M H M M H M H M H M1 M2 Li 3·24 1·57 4·62 4·04 2·15 2·58 2·13 2·12 3·09 4·07 3·84 2·21 2·04 — 2·76 2·81 Be 0·17 0·13 0·25 0·28 0·29 0·29 0·11 0·09 0·23 0·48 0·32 b.d. b.d. — b.d. b.d. B 1·87 2·10 10·06 8·31 2·46 1·53 1·73 2·42 1·74 1·74 0·76 9·80 8·60 — 10·09 10·51 Sc 28·4 17·2 37·0 39·6 22·0 30·7 26·9 34·0 34·3 54·6 49·1 43·0 50·5 29·8 31·6 39·2 Ti 2285 2450 4066 4259 2478 2632 2302 2337 6059 4539 5280 4212 3597 3127 3077 3240 V 177 187 351 346 230 241 244 254 487 392 358 277 263 198 200 191 Cr 1816 1173 1794 1924 1407 1601 1245 1455 597 2047 2450 2397 1479 1334 1226 1233 Co 41·0 27·7 25·6 23·9 34·6 36·1 37·6 37·0 43·8 45·8 46·6 23·9 25·6 24·6 23·5 23·2 Ni 472 439 227 227 428 444 410 387 507 385 511 304 314 294 287 259 Cu 1·00 1·82 1·44 1·26 0·75 0·75 1·49 1·41 1·82 1·49 3·47 1·98 1·98 1·05 1·03 1·01 Zn 21·1 6·4 14·5 11·8 18·7 18·6 24·3 23·8 33·7 52·3 47·2 6·6 6·4 6·1 6·3 6·4 Ga 8·74 9·78 11·08 10·59 9·73 10·44 12·45 12·43 16·80 11·47 12·11 5·26 4·09 4·33 3·76 2·94 Rb b.d. b.d. 0·04 b.d. b.d. b.d. 0·03 b.d. b.d. b.d. b.d. b.d. b.d. b.d. b.d. 0·04 Sr 146 137 149 147 126 125 114 115 104 109 135 95 97 128 138 138 Y 1·56 0·72 5·07 5·62 1·68 1·99 1·21 1·36 3·06 6·54 6·23 6·28 7·68 2·82 3·79 6·70 Zr 6·64 14·27 25·35 26·01 16·44 19·90 5·49 6·23 47·95 30·48 32·16 18·51 18·07 21·85 22·78 27·70 Nb 0·14 0·55 0·20 0·17 0·08 0·06 0·18 0·22 0·86 0·07 0·45 0·20 0·22 0·08 0·12 0·15 Ba 0·05 0·12 0·09 0·05 0·09 0·07 0·04 0·05 0·12 0·06 0·07 0·09 0·11 0·11 0·12 0·21 La 2·64 2·24 4·98 4·66 4·03 3·72 1·71 1·74 4·45 2·61 3·77 4·80 4·80 8·07 8·47 8·76 Ce 3·64 4·43 11·73 10·50 16·17 15·82 3·53 3·51 10·54 6·38 10·95 9·49 9·66 24·72 27·28 26·93 Pr 0·46 0·60 1·72 1·50 3·09 3·03 0·48 0·48 1·47 1·13 1·87 1·17 1·16 3·62 4·00 4·06 Nd 2·34 3·03 8·29 7·33 14·89 15·23 2·42 2·41 6·82 6·69 9·59 5·46 5·32 15·28 16·80 17·29 Sm 0·71 0·81 2·32 2·12 2·68 2·97 0·72 0·72 2·15 2·23 2·60 1·69 1·66 2·49 2·78 2·99 Eu 0·31 0·33 0·91 0·89 0·71 0·69 0·26 0·30 0·73 0·79 0·89 0·70 0·71 0·74 0·83 0·90 Gd 0·74 0·64 2·31 2·20 1·49 1·35 0·64 0·68 2·17 2·43 2·44 1·91 1·99 1·66 1·90 2·36 Tb 0·09 0·07 0·30 0·31 0·13 0·15 0·08 0·08 0·25 0·33 0·31 0·23 0·40 b.d. 0·25 0·32 Dy 0·47 0·30 1·78 1·67 0·57 0·61 0·40 0·43 1·06 1·73 1·72 1·58 1·94 0·88 1·10 1·71 Ho 0·07 0·03 0·25 0·26 0·06 0·09 0·05 0·07 0·14 0·28 0·26 0·33 0·32 b.d. 0·16 0·27 Er 0·13 0·05 0·50 0·52 0·12 0·15 0·10 0·11 0·22 0·56 0·52 0·54 0·69 0·22 0·30 0·61 Tm 0·01 b.d. 0·05 0·05 0·01 0·01 0·01 0·01 0·02 0·07 0·06 0·09 0·07 b.d. 0·03 0·07 Yb 0·07 b.d. 0·24 0·27 0·05 0·07 0·04 0·07 0·09 0·35 0·34 0·32 0·38 0·10 0·12 0·35 Lu 0·01 0·01 0·03 0·03 b.d. 0·01 0·01 0·01 0·01 0·04 0·04 0·04 0·04 0·01 0·01 0·04 Hf 0·35 0·58 1·18 1·08 0·49 0·59 0·29 0·36 2·50 1·38 1·41 0·67 0·68 0·61 0·63 0·67 Ta 0·02 0·14 0·02 0·02 0·03 0·02 0·03 0·03 0·05 0·02 0·04 0·01 0·01 0·02 0·02 0·02 Pb 0·42 0·20 0·75 0·69 0·50 0·52 0·22 0·20 0·34 0·29 0·47 0·43 0·42 0·54 0·55 0·54 Th 0·55 0·29 0·71 0·72 0·22 0·22 0·31 0·28 0·56 0·63 0·76 0·61 0·60 0·46 0·43 0·46 U 0·13 0·06 0·17 0·19 0·06 0·04 0·07 0·07 0·13 0·15 0·15 0·14 0·14 0·10 0·09 0·09 M, Cpx in the mosaic area; H, host Cpx with exsolution textures; b.d., below detection limit; —, no value. Table 4: Trace element compositions (ppm) of amphibole in the garnet pyroxenites from Lakes Bullenmerri and Gnotuk Sample: BM96-2 BM99-6 BM99-4 BME-1 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: M vein M M E1 M E1 M vein M M1 M2 Li 1·97 1·97 1·09 2·91 2·06 0·95 3·14 2·29 3·26 0·97 1·59 1·63 Be 0·42 0·53 0·24 0·58 0·58 0·33 0·85 0·49 0·70 0·30 0·30 0·27 B 6·52 6·94 6·4 6·00 7·26 5·41 7·85 0·69 1·34 2·76 2·97 2·85 Sc 19·0 20·3 13·0 28·5 16·6 26 45·8 38·2 37 41·5 28·3 38·4 Ti 8856 5149 8224 14344 9551 21921 17059 20413 16530 16103 13087 12879 V 265 180 248 476 336 649 468 494 386 322 294 269 Cr 1829 1665 1547 2644 2237 850 2866 4553 3045 2302 1944 1510 Co 79·7 77·6 53·6 52·0 67·4 79·4 75·5 79·1 81·0 47·5 46·6 47·2 Ni 1212 1053 1040 613 1072 1274 782 1171 1032 773 723 704 Cu 1·71 1·81 3·31 2·00 1·55 2·04 2·08 3·91 4·56 0·88 0·81 0·75 Zn 45·0 42·9 12·8 22·8 35·1 54·7 91·3 70·9 81·7 9·70 9·10 9·30 Ga 36·1 27·1 36·4 27·4 18·5 73 23·6 27·9 30·8 49·5 19·1 20·9 Rb 9·35 24·4 6·25 1·57 1·01 3·17 8·88 3·88 13·1 8·77 2·14 3·02 Sr 409 327 386 493 371 359 384 525 577 311 530 556 Y 1·96 2·56 1·18 7·75 2·72 4·76 12·36 8·77 11·07 10·71 5·42 11·1 Zr 6·64 71·3 19·6 20·7 14·1 43·3 33·9 25·7 119 16·1 21·6 25·3 Nb 13·0 30·2 41·9 16·3 7·99 33·5 7·27 28·4 62·5 17·6 8·48 11·9 Ba 365 213 371 211 52·0 1109 109 198 193 623 130 168 La 8·4 8·68 2·88 6·17 4·91 6·68 5·13 4·30 9·38 6·00 11·8 12·4 Ce 15·7 13·5 5·7 14·2 20·7 16·2 10·8 12·5 25·1 11·8 35·7 37·7 Pr 1·73 1·34 0·76 2·03 3·73 2·29 1·45 2·2 3·66 1·44 5·43 5·76 Nd 6·68 4·99 3·78 9·97 19·1 11·5 8·00 11·7 17·1 6·60 23·6 25·4 Sm 1·19 1·02 1·06 2·78 3·48 3·42 2·73 3·2 4·12 2·07 4·11 4·46 Eu 0·47 0·4 0·44 1·12 0·93 1·22 1·00 1·16 1·49 0·95 1·23 1·41 Gd 0·93 0·97 0·88 2·94 1·97 3·38 2·97 3·34 3·76 2·79 2·97 3·72 Tb 0·11 0·12 0·1 0·39 0·19 0·37 0·44 0·43 0·51 0·43 0·34 0·48 Dy 0·57 0·65 0·42 2·09 0·84 1·58 2·81 2·29 2·72 2·63 1·58 2·70 Ho 0·09 0·11 0·05 0·32 0·10 0·21 0·51 0·37 0·44 0·44 0·22 0·44 Er 0·16 0·22 0·08 0·64 0·19 0·36 1·18 0·72 1·00 0·96 0·40 0·90 Tm 0·01 0·02 0·01 0·07 0·02 0·03 0·14 0·08 0·11 0·09 0·04 0·09 Yb 0·07 0·12 0·04 0·37 0·09 0·17 0·82 0·4 0·61 0·52 0·18 0·47 Lu 0·01 0·01 0·01 0·04 0·01 0·01 0·11 0·05 0·07 0·06 0·02 0·05 Hf 0·31 1·74 0·76 0·72 0·37 1·87 1·07 1·06 2·85 0·59 0·58 0·57 Ta 0·16 1·59 2·47 0·29 0·25 0·34 0·18 0·69 2·34 0·22 0·34 0·40 Pb 1·64 1·49 1·09 2·84 2·09 1·7 0·93 2·4 2·29 1·97 2·93 2·86 Th 0·64 0·82 0·32 0·57 0·22 0·61 0·78 0·68 0·88 0·59 0·55 0·59 U 0·15 0·17 0·07 0·12 0·05 0·13 0·17 0·14 0·20 0·13 0·10 0·10 Sample: BM96-2 BM99-6 BM99-4 BME-1 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: M vein M M E1 M E1 M vein M M1 M2 Li 1·97 1·97 1·09 2·91 2·06 0·95 3·14 2·29 3·26 0·97 1·59 1·63 Be 0·42 0·53 0·24 0·58 0·58 0·33 0·85 0·49 0·70 0·30 0·30 0·27 B 6·52 6·94 6·4 6·00 7·26 5·41 7·85 0·69 1·34 2·76 2·97 2·85 Sc 19·0 20·3 13·0 28·5 16·6 26 45·8 38·2 37 41·5 28·3 38·4 Ti 8856 5149 8224 14344 9551 21921 17059 20413 16530 16103 13087 12879 V 265 180 248 476 336 649 468 494 386 322 294 269 Cr 1829 1665 1547 2644 2237 850 2866 4553 3045 2302 1944 1510 Co 79·7 77·6 53·6 52·0 67·4 79·4 75·5 79·1 81·0 47·5 46·6 47·2 Ni 1212 1053 1040 613 1072 1274 782 1171 1032 773 723 704 Cu 1·71 1·81 3·31 2·00 1·55 2·04 2·08 3·91 4·56 0·88 0·81 0·75 Zn 45·0 42·9 12·8 22·8 35·1 54·7 91·3 70·9 81·7 9·70 9·10 9·30 Ga 36·1 27·1 36·4 27·4 18·5 73 23·6 27·9 30·8 49·5 19·1 20·9 Rb 9·35 24·4 6·25 1·57 1·01 3·17 8·88 3·88 13·1 8·77 2·14 3·02 Sr 409 327 386 493 371 359 384 525 577 311 530 556 Y 1·96 2·56 1·18 7·75 2·72 4·76 12·36 8·77 11·07 10·71 5·42 11·1 Zr 6·64 71·3 19·6 20·7 14·1 43·3 33·9 25·7 119 16·1 21·6 25·3 Nb 13·0 30·2 41·9 16·3 7·99 33·5 7·27 28·4 62·5 17·6 8·48 11·9 Ba 365 213 371 211 52·0 1109 109 198 193 623 130 168 La 8·4 8·68 2·88 6·17 4·91 6·68 5·13 4·30 9·38 6·00 11·8 12·4 Ce 15·7 13·5 5·7 14·2 20·7 16·2 10·8 12·5 25·1 11·8 35·7 37·7 Pr 1·73 1·34 0·76 2·03 3·73 2·29 1·45 2·2 3·66 1·44 5·43 5·76 Nd 6·68 4·99 3·78 9·97 19·1 11·5 8·00 11·7 17·1 6·60 23·6 25·4 Sm 1·19 1·02 1·06 2·78 3·48 3·42 2·73 3·2 4·12 2·07 4·11 4·46 Eu 0·47 0·4 0·44 1·12 0·93 1·22 1·00 1·16 1·49 0·95 1·23 1·41 Gd 0·93 0·97 0·88 2·94 1·97 3·38 2·97 3·34 3·76 2·79 2·97 3·72 Tb 0·11 0·12 0·1 0·39 0·19 0·37 0·44 0·43 0·51 0·43 0·34 0·48 Dy 0·57 0·65 0·42 2·09 0·84 1·58 2·81 2·29 2·72 2·63 1·58 2·70 Ho 0·09 0·11 0·05 0·32 0·10 0·21 0·51 0·37 0·44 0·44 0·22 0·44 Er 0·16 0·22 0·08 0·64 0·19 0·36 1·18 0·72 1·00 0·96 0·40 0·90 Tm 0·01 0·02 0·01 0·07 0·02 0·03 0·14 0·08 0·11 0·09 0·04 0·09 Yb 0·07 0·12 0·04 0·37 0·09 0·17 0·82 0·4 0·61 0·52 0·18 0·47 Lu 0·01 0·01 0·01 0·04 0·01 0·01 0·11 0·05 0·07 0·06 0·02 0·05 Hf 0·31 1·74 0·76 0·72 0·37 1·87 1·07 1·06 2·85 0·59 0·58 0·57 Ta 0·16 1·59 2·47 0·29 0·25 0·34 0·18 0·69 2·34 0·22 0·34 0·40 Pb 1·64 1·49 1·09 2·84 2·09 1·7 0·93 2·4 2·29 1·97 2·93 2·86 Th 0·64 0·82 0·32 0·57 0·22 0·61 0·78 0·68 0·88 0·59 0·55 0·59 U 0·15 0·17 0·07 0·12 0·05 0·13 0·17 0·14 0·20 0·13 0·10 0·10 M, Amp in the mosaic area; E1, Amp as exsolution-like lamellae in Cpx megacryst. Table 4: Trace element compositions (ppm) of amphibole in the garnet pyroxenites from Lakes Bullenmerri and Gnotuk Sample: BM96-2 BM99-6 BM99-4 BME-1 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: M vein M M E1 M E1 M vein M M1 M2 Li 1·97 1·97 1·09 2·91 2·06 0·95 3·14 2·29 3·26 0·97 1·59 1·63 Be 0·42 0·53 0·24 0·58 0·58 0·33 0·85 0·49 0·70 0·30 0·30 0·27 B 6·52 6·94 6·4 6·00 7·26 5·41 7·85 0·69 1·34 2·76 2·97 2·85 Sc 19·0 20·3 13·0 28·5 16·6 26 45·8 38·2 37 41·5 28·3 38·4 Ti 8856 5149 8224 14344 9551 21921 17059 20413 16530 16103 13087 12879 V 265 180 248 476 336 649 468 494 386 322 294 269 Cr 1829 1665 1547 2644 2237 850 2866 4553 3045 2302 1944 1510 Co 79·7 77·6 53·6 52·0 67·4 79·4 75·5 79·1 81·0 47·5 46·6 47·2 Ni 1212 1053 1040 613 1072 1274 782 1171 1032 773 723 704 Cu 1·71 1·81 3·31 2·00 1·55 2·04 2·08 3·91 4·56 0·88 0·81 0·75 Zn 45·0 42·9 12·8 22·8 35·1 54·7 91·3 70·9 81·7 9·70 9·10 9·30 Ga 36·1 27·1 36·4 27·4 18·5 73 23·6 27·9 30·8 49·5 19·1 20·9 Rb 9·35 24·4 6·25 1·57 1·01 3·17 8·88 3·88 13·1 8·77 2·14 3·02 Sr 409 327 386 493 371 359 384 525 577 311 530 556 Y 1·96 2·56 1·18 7·75 2·72 4·76 12·36 8·77 11·07 10·71 5·42 11·1 Zr 6·64 71·3 19·6 20·7 14·1 43·3 33·9 25·7 119 16·1 21·6 25·3 Nb 13·0 30·2 41·9 16·3 7·99 33·5 7·27 28·4 62·5 17·6 8·48 11·9 Ba 365 213 371 211 52·0 1109 109 198 193 623 130 168 La 8·4 8·68 2·88 6·17 4·91 6·68 5·13 4·30 9·38 6·00 11·8 12·4 Ce 15·7 13·5 5·7 14·2 20·7 16·2 10·8 12·5 25·1 11·8 35·7 37·7 Pr 1·73 1·34 0·76 2·03 3·73 2·29 1·45 2·2 3·66 1·44 5·43 5·76 Nd 6·68 4·99 3·78 9·97 19·1 11·5 8·00 11·7 17·1 6·60 23·6 25·4 Sm 1·19 1·02 1·06 2·78 3·48 3·42 2·73 3·2 4·12 2·07 4·11 4·46 Eu 0·47 0·4 0·44 1·12 0·93 1·22 1·00 1·16 1·49 0·95 1·23 1·41 Gd 0·93 0·97 0·88 2·94 1·97 3·38 2·97 3·34 3·76 2·79 2·97 3·72 Tb 0·11 0·12 0·1 0·39 0·19 0·37 0·44 0·43 0·51 0·43 0·34 0·48 Dy 0·57 0·65 0·42 2·09 0·84 1·58 2·81 2·29 2·72 2·63 1·58 2·70 Ho 0·09 0·11 0·05 0·32 0·10 0·21 0·51 0·37 0·44 0·44 0·22 0·44 Er 0·16 0·22 0·08 0·64 0·19 0·36 1·18 0·72 1·00 0·96 0·40 0·90 Tm 0·01 0·02 0·01 0·07 0·02 0·03 0·14 0·08 0·11 0·09 0·04 0·09 Yb 0·07 0·12 0·04 0·37 0·09 0·17 0·82 0·4 0·61 0·52 0·18 0·47 Lu 0·01 0·01 0·01 0·04 0·01 0·01 0·11 0·05 0·07 0·06 0·02 0·05 Hf 0·31 1·74 0·76 0·72 0·37 1·87 1·07 1·06 2·85 0·59 0·58 0·57 Ta 0·16 1·59 2·47 0·29 0·25 0·34 0·18 0·69 2·34 0·22 0·34 0·40 Pb 1·64 1·49 1·09 2·84 2·09 1·7 0·93 2·4 2·29 1·97 2·93 2·86 Th 0·64 0·82 0·32 0·57 0·22 0·61 0·78 0·68 0·88 0·59 0·55 0·59 U 0·15 0·17 0·07 0·12 0·05 0·13 0·17 0·14 0·20 0·13 0·10 0·10 Sample: BM96-2 BM99-6 BM99-4 BME-1 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: M vein M M E1 M E1 M vein M M1 M2 Li 1·97 1·97 1·09 2·91 2·06 0·95 3·14 2·29 3·26 0·97 1·59 1·63 Be 0·42 0·53 0·24 0·58 0·58 0·33 0·85 0·49 0·70 0·30 0·30 0·27 B 6·52 6·94 6·4 6·00 7·26 5·41 7·85 0·69 1·34 2·76 2·97 2·85 Sc 19·0 20·3 13·0 28·5 16·6 26 45·8 38·2 37 41·5 28·3 38·4 Ti 8856 5149 8224 14344 9551 21921 17059 20413 16530 16103 13087 12879 V 265 180 248 476 336 649 468 494 386 322 294 269 Cr 1829 1665 1547 2644 2237 850 2866 4553 3045 2302 1944 1510 Co 79·7 77·6 53·6 52·0 67·4 79·4 75·5 79·1 81·0 47·5 46·6 47·2 Ni 1212 1053 1040 613 1072 1274 782 1171 1032 773 723 704 Cu 1·71 1·81 3·31 2·00 1·55 2·04 2·08 3·91 4·56 0·88 0·81 0·75 Zn 45·0 42·9 12·8 22·8 35·1 54·7 91·3 70·9 81·7 9·70 9·10 9·30 Ga 36·1 27·1 36·4 27·4 18·5 73 23·6 27·9 30·8 49·5 19·1 20·9 Rb 9·35 24·4 6·25 1·57 1·01 3·17 8·88 3·88 13·1 8·77 2·14 3·02 Sr 409 327 386 493 371 359 384 525 577 311 530 556 Y 1·96 2·56 1·18 7·75 2·72 4·76 12·36 8·77 11·07 10·71 5·42 11·1 Zr 6·64 71·3 19·6 20·7 14·1 43·3 33·9 25·7 119 16·1 21·6 25·3 Nb 13·0 30·2 41·9 16·3 7·99 33·5 7·27 28·4 62·5 17·6 8·48 11·9 Ba 365 213 371 211 52·0 1109 109 198 193 623 130 168 La 8·4 8·68 2·88 6·17 4·91 6·68 5·13 4·30 9·38 6·00 11·8 12·4 Ce 15·7 13·5 5·7 14·2 20·7 16·2 10·8 12·5 25·1 11·8 35·7 37·7 Pr 1·73 1·34 0·76 2·03 3·73 2·29 1·45 2·2 3·66 1·44 5·43 5·76 Nd 6·68 4·99 3·78 9·97 19·1 11·5 8·00 11·7 17·1 6·60 23·6 25·4 Sm 1·19 1·02 1·06 2·78 3·48 3·42 2·73 3·2 4·12 2·07 4·11 4·46 Eu 0·47 0·4 0·44 1·12 0·93 1·22 1·00 1·16 1·49 0·95 1·23 1·41 Gd 0·93 0·97 0·88 2·94 1·97 3·38 2·97 3·34 3·76 2·79 2·97 3·72 Tb 0·11 0·12 0·1 0·39 0·19 0·37 0·44 0·43 0·51 0·43 0·34 0·48 Dy 0·57 0·65 0·42 2·09 0·84 1·58 2·81 2·29 2·72 2·63 1·58 2·70 Ho 0·09 0·11 0·05 0·32 0·10 0·21 0·51 0·37 0·44 0·44 0·22 0·44 Er 0·16 0·22 0·08 0·64 0·19 0·36 1·18 0·72 1·00 0·96 0·40 0·90 Tm 0·01 0·02 0·01 0·07 0·02 0·03 0·14 0·08 0·11 0·09 0·04 0·09 Yb 0·07 0·12 0·04 0·37 0·09 0·17 0·82 0·4 0·61 0·52 0·18 0·47 Lu 0·01 0·01 0·01 0·04 0·01 0·01 0·11 0·05 0·07 0·06 0·02 0·05 Hf 0·31 1·74 0·76 0·72 0·37 1·87 1·07 1·06 2·85 0·59 0·58 0·57 Ta 0·16 1·59 2·47 0·29 0·25 0·34 0·18 0·69 2·34 0·22 0·34 0·40 Pb 1·64 1·49 1·09 2·84 2·09 1·7 0·93 2·4 2·29 1·97 2·93 2·86 Th 0·64 0·82 0·32 0·57 0·22 0·61 0·78 0·68 0·88 0·59 0·55 0·59 U 0·15 0·17 0·07 0·12 0·05 0·13 0·17 0·14 0·20 0·13 0·10 0·10 M, Amp in the mosaic area; E1, Amp as exsolution-like lamellae in Cpx megacryst. Table 5: Trace element compositions (ppm) of garnet in the garnet pyroxenites from Lakes Bullenmerri and Gnotuk Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: C C E1 E2 M E1 C E1 M C C E1 M E1 C E1 M M Li 0·31 b.d. 0·65 0·45 0·28 b.d. 0·33 b.d. b.d. 0·16 0·38 0·52 0·48 b.d. b.d. b.d. b.d. b.d. B 1·49 2·21 12·2 16·9 12·8 1·79 1·75 1·95 2·25 2·93 1·18 1·68 0·95 b.d. 10·8 b.d. 9·78 10·9 Sc 88·0 63·7 89·3 82·8 84·4 58·3 53·3 90·2 83·9 113 87·9 153 110 136 131 78·1 97·1 100 Ti 557 285 368 354 321 222 271 381 423 359 889 1043 978 569 606 356 381 333 V 51·6 38·8 59·8 58·6 55·6 47·7 57·2 62·5 66·0 62·5 133 116 118 67·3 61·4 58·5 44·6 54·1 Cr 2943 1187 1240 1283 1302 1622 1650 1415 1431 1683 747 2908 3191 2094 2091 1538 1312 1519 Co 75·4 58·4 43·3 46·4 42·9 75·1 74·3 69·8 71·2 72·4 75·4 69·3 73·3 47·9 49·6 51·0 51·2 46·7 Ni 48·4 46·7 21·8 22·7 20·6 44·0 40·5 45·9 45·1 38·7 56·6 37·2 54·9 37·2 37·8 30·4 29·4 25·7 Cu 0·17 1·27 0·69 b.d. 0·03 0·27 0·14 b.d. 0·17 0·12 0·15 0·30 0·17 0·13 0·09 0·41 0·12 b.d. Zn 17·77 6·56 14·24 18·05 14·90 18·56 19·12 23·71 24·11 22·19 29·96 41·46 38·60 7·46 6·94 6·14 6·94 6·99 Ga 6·45 4·20 3·78 3·93 3·95 4·63 4·98 6·48 6·90 6·16 8·58 7·10 7·10 2·61 3·46 2·85 1·58 2·36 Rb 0·07 b.d. 0·10 0·01 b.d. b.d. b.d. b.d. b.d. 0·08 b.d. 0·06 b.d. b.d. b.d. b.d. b.d. b.d. Sr 0·13 0·13 0·20 0·07 0·08 0·07 0·12 0·10 0·09 0·09 0·14 0·17 0·14 0·07 0·07 0·37 0·10 0·15 Y 12·88 7·66 35·62 33·27 35·37 14·11 13·38 13·83 12·42 14·97 26·33 57·97 46·71 49·56 43·30 24·08 34·86 59·57 Zr 4·82 13·10 9·46 9·01 10·04 7·30 7·65 3·78 3·80 4·16 25·82 18·04 16·93 9·21 9·85 11·84 12·38 13·77 Nb 0·01 0·08 0·01 0·01 0·01 b.d. b.d. 0·02 0·01 0·02 0·04 0·01 0·03 0·02 0·01 b.d. 0·01 0·01 Ba 0·01 0·22 0·08 b.d. 0·00 0·02 b.d. 0·00 0·01 0·01 0·03 0·08 b.d. b.d. b.d. b.d. 0·09 0·07 La 0·02 0·02 0·03 0·01 0·01 0·02 0·01 b.d. 0·01 0·01 0·02 0·03 0·01 0·02 0·01 b.d. 0·02 0·07 Ce 0·04 0·11 0·06 0·07 0·07 0·12 0·11 0·03 0·03 0·03 0·09 0·10 0·08 0·08 0·08 0·19 0·20 0·22 Pr 0·02 0·02 0·03 0·03 0·03 0·06 0·05 0·02 0·01 0·02 0·04 0·05 0·04 0·02 0·02 0·06 0·09 0·07 Nd 0·18 0·18 0·34 0·40 0·36 0·78 0·58 0·18 0·19 0·14 0·38 0·65 0·52 0·30 0·31 0·70 0·94 0·69 Sm 0·29 0·28 0·52 0·58 0·54 0·67 0·59 0·24 0·22 0·28 0·70 1·08 0·79 0·43 0·52 0·61 0·79 0·69 Eu 0·22 0·23 0·46 0·48 0·47 0·34 0·33 0·20 0·20 0·21 0·42 0·69 0·52 0·36 0·38 0·38 0·47 0·45 Gd 0·92 0·85 2·21 2·27 2·28 1·37 1·37 0·87 0·89 0·91 2·60 3·82 2·80 1·89 1·90 1·51 2·12 2·21 Tb 0·24 0·19 0·63 0·66 0·65 0·29 0·27 0·25 0·22 0·25 0·56 0·96 0·73 b.d. 0·60 0·46 0·55 0·65 Dy 1·98 1·33 6·39 6·74 6·74 2·37 2·10 2·11 1·96 2·17 4·65 8·94 6·94 6·17 5·60 3·48 4·92 6·88 Ho 0·49 0·30 1·68 1·95 1·84 0·51 0·50 0·55 0·48 0·58 1·03 2·27 1·83 b.d. 1·40 0·92 1·10 1·97 Er 1·45 0·84 5·78 6·69 6·34 1·48 1·49 1·63 1·39 2·00 2·79 6·82 5·84 6·73 4·92 2·91 4·10 7·63 Tm 0·21 0·11 0·88 1·04 0·97 0·20 0·22 0·24 0·20 0·34 0·38 1·04 0·90 b.d. 0·71 0·39 0·59 1·21 Yb 1·47 0·77 6·20 7·70 7·29 1·29 1·59 1·82 1·31 2·54 2·40 7·09 6·60 8·47 5·02 3·02 4·50 10·86 Lu 0·22 0·11 0·92 1·14 1·11 0·16 0·23 0·25 0·19 0·40 0·35 1·01 0·98 1·35 0·89 0·48 0·66 1·76 Hf 0·12 0·29 0·19 0·19 0·17 0·08 0·18 0·10 0·09 0·09 0·45 0·27 0·29 0·14 0·20 0·10 0·10 0·15 Ta 0·01 0·01 b.d. b.d. b.d. 0·01 0·01 b.d. b.d. 0·00 b.d. 0·00 b.d. b.d. b.d. b.d. b.d. b.d. Pb b.d. 0·16 b.d. b.d. 0·02 b.d. 0·06 b.d. b.d. b.d. b.d. 0·11 b.d. b.d. b.d. b.d. b.d. b.d. Th 0·02 b.d. 0·02 b.d. 0·01 b.d. 0·01 b.d. b.d. b.d. 0·02 0·02 b.d. b.d. b.d. b.d. b.d. b.d. U 0·02 0·02 0·01 0·04 0·02 b.d. 0·13 0·01 b.d. b.d. 0·04 0·04 0·02 b.d. 0·03 b.d. b.d. b.d. Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: C C E1 E2 M E1 C E1 M C C E1 M E1 C E1 M M Li 0·31 b.d. 0·65 0·45 0·28 b.d. 0·33 b.d. b.d. 0·16 0·38 0·52 0·48 b.d. b.d. b.d. b.d. b.d. B 1·49 2·21 12·2 16·9 12·8 1·79 1·75 1·95 2·25 2·93 1·18 1·68 0·95 b.d. 10·8 b.d. 9·78 10·9 Sc 88·0 63·7 89·3 82·8 84·4 58·3 53·3 90·2 83·9 113 87·9 153 110 136 131 78·1 97·1 100 Ti 557 285 368 354 321 222 271 381 423 359 889 1043 978 569 606 356 381 333 V 51·6 38·8 59·8 58·6 55·6 47·7 57·2 62·5 66·0 62·5 133 116 118 67·3 61·4 58·5 44·6 54·1 Cr 2943 1187 1240 1283 1302 1622 1650 1415 1431 1683 747 2908 3191 2094 2091 1538 1312 1519 Co 75·4 58·4 43·3 46·4 42·9 75·1 74·3 69·8 71·2 72·4 75·4 69·3 73·3 47·9 49·6 51·0 51·2 46·7 Ni 48·4 46·7 21·8 22·7 20·6 44·0 40·5 45·9 45·1 38·7 56·6 37·2 54·9 37·2 37·8 30·4 29·4 25·7 Cu 0·17 1·27 0·69 b.d. 0·03 0·27 0·14 b.d. 0·17 0·12 0·15 0·30 0·17 0·13 0·09 0·41 0·12 b.d. Zn 17·77 6·56 14·24 18·05 14·90 18·56 19·12 23·71 24·11 22·19 29·96 41·46 38·60 7·46 6·94 6·14 6·94 6·99 Ga 6·45 4·20 3·78 3·93 3·95 4·63 4·98 6·48 6·90 6·16 8·58 7·10 7·10 2·61 3·46 2·85 1·58 2·36 Rb 0·07 b.d. 0·10 0·01 b.d. b.d. b.d. b.d. b.d. 0·08 b.d. 0·06 b.d. b.d. b.d. b.d. b.d. b.d. Sr 0·13 0·13 0·20 0·07 0·08 0·07 0·12 0·10 0·09 0·09 0·14 0·17 0·14 0·07 0·07 0·37 0·10 0·15 Y 12·88 7·66 35·62 33·27 35·37 14·11 13·38 13·83 12·42 14·97 26·33 57·97 46·71 49·56 43·30 24·08 34·86 59·57 Zr 4·82 13·10 9·46 9·01 10·04 7·30 7·65 3·78 3·80 4·16 25·82 18·04 16·93 9·21 9·85 11·84 12·38 13·77 Nb 0·01 0·08 0·01 0·01 0·01 b.d. b.d. 0·02 0·01 0·02 0·04 0·01 0·03 0·02 0·01 b.d. 0·01 0·01 Ba 0·01 0·22 0·08 b.d. 0·00 0·02 b.d. 0·00 0·01 0·01 0·03 0·08 b.d. b.d. b.d. b.d. 0·09 0·07 La 0·02 0·02 0·03 0·01 0·01 0·02 0·01 b.d. 0·01 0·01 0·02 0·03 0·01 0·02 0·01 b.d. 0·02 0·07 Ce 0·04 0·11 0·06 0·07 0·07 0·12 0·11 0·03 0·03 0·03 0·09 0·10 0·08 0·08 0·08 0·19 0·20 0·22 Pr 0·02 0·02 0·03 0·03 0·03 0·06 0·05 0·02 0·01 0·02 0·04 0·05 0·04 0·02 0·02 0·06 0·09 0·07 Nd 0·18 0·18 0·34 0·40 0·36 0·78 0·58 0·18 0·19 0·14 0·38 0·65 0·52 0·30 0·31 0·70 0·94 0·69 Sm 0·29 0·28 0·52 0·58 0·54 0·67 0·59 0·24 0·22 0·28 0·70 1·08 0·79 0·43 0·52 0·61 0·79 0·69 Eu 0·22 0·23 0·46 0·48 0·47 0·34 0·33 0·20 0·20 0·21 0·42 0·69 0·52 0·36 0·38 0·38 0·47 0·45 Gd 0·92 0·85 2·21 2·27 2·28 1·37 1·37 0·87 0·89 0·91 2·60 3·82 2·80 1·89 1·90 1·51 2·12 2·21 Tb 0·24 0·19 0·63 0·66 0·65 0·29 0·27 0·25 0·22 0·25 0·56 0·96 0·73 b.d. 0·60 0·46 0·55 0·65 Dy 1·98 1·33 6·39 6·74 6·74 2·37 2·10 2·11 1·96 2·17 4·65 8·94 6·94 6·17 5·60 3·48 4·92 6·88 Ho 0·49 0·30 1·68 1·95 1·84 0·51 0·50 0·55 0·48 0·58 1·03 2·27 1·83 b.d. 1·40 0·92 1·10 1·97 Er 1·45 0·84 5·78 6·69 6·34 1·48 1·49 1·63 1·39 2·00 2·79 6·82 5·84 6·73 4·92 2·91 4·10 7·63 Tm 0·21 0·11 0·88 1·04 0·97 0·20 0·22 0·24 0·20 0·34 0·38 1·04 0·90 b.d. 0·71 0·39 0·59 1·21 Yb 1·47 0·77 6·20 7·70 7·29 1·29 1·59 1·82 1·31 2·54 2·40 7·09 6·60 8·47 5·02 3·02 4·50 10·86 Lu 0·22 0·11 0·92 1·14 1·11 0·16 0·23 0·25 0·19 0·40 0·35 1·01 0·98 1·35 0·89 0·48 0·66 1·76 Hf 0·12 0·29 0·19 0·19 0·17 0·08 0·18 0·10 0·09 0·09 0·45 0·27 0·29 0·14 0·20 0·10 0·10 0·15 Ta 0·01 0·01 b.d. b.d. b.d. 0·01 0·01 b.d. b.d. 0·00 b.d. 0·00 b.d. b.d. b.d. b.d. b.d. b.d. Pb b.d. 0·16 b.d. b.d. 0·02 b.d. 0·06 b.d. b.d. b.d. b.d. 0·11 b.d. b.d. b.d. b.d. b.d. b.d. Th 0·02 b.d. 0·02 b.d. 0·01 b.d. 0·01 b.d. b.d. b.d. 0·02 0·02 b.d. b.d. b.d. b.d. b.d. b.d. U 0·02 0·02 0·01 0·04 0·02 b.d. 0·13 0·01 b.d. b.d. 0·04 0·04 0·02 b.d. 0·03 b.d. b.d. b.d. C, Grt in the coarse area; M, Grt in the mosaic area; E1, Grt as exsolution lamellae in Cpx megacryst; E2, Grt as exsolution lamellae in Opx megacryst; b.d., below detection limit. Table 5: Trace element compositions (ppm) of garnet in the garnet pyroxenites from Lakes Bullenmerri and Gnotuk Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: C C E1 E2 M E1 C E1 M C C E1 M E1 C E1 M M Li 0·31 b.d. 0·65 0·45 0·28 b.d. 0·33 b.d. b.d. 0·16 0·38 0·52 0·48 b.d. b.d. b.d. b.d. b.d. B 1·49 2·21 12·2 16·9 12·8 1·79 1·75 1·95 2·25 2·93 1·18 1·68 0·95 b.d. 10·8 b.d. 9·78 10·9 Sc 88·0 63·7 89·3 82·8 84·4 58·3 53·3 90·2 83·9 113 87·9 153 110 136 131 78·1 97·1 100 Ti 557 285 368 354 321 222 271 381 423 359 889 1043 978 569 606 356 381 333 V 51·6 38·8 59·8 58·6 55·6 47·7 57·2 62·5 66·0 62·5 133 116 118 67·3 61·4 58·5 44·6 54·1 Cr 2943 1187 1240 1283 1302 1622 1650 1415 1431 1683 747 2908 3191 2094 2091 1538 1312 1519 Co 75·4 58·4 43·3 46·4 42·9 75·1 74·3 69·8 71·2 72·4 75·4 69·3 73·3 47·9 49·6 51·0 51·2 46·7 Ni 48·4 46·7 21·8 22·7 20·6 44·0 40·5 45·9 45·1 38·7 56·6 37·2 54·9 37·2 37·8 30·4 29·4 25·7 Cu 0·17 1·27 0·69 b.d. 0·03 0·27 0·14 b.d. 0·17 0·12 0·15 0·30 0·17 0·13 0·09 0·41 0·12 b.d. Zn 17·77 6·56 14·24 18·05 14·90 18·56 19·12 23·71 24·11 22·19 29·96 41·46 38·60 7·46 6·94 6·14 6·94 6·99 Ga 6·45 4·20 3·78 3·93 3·95 4·63 4·98 6·48 6·90 6·16 8·58 7·10 7·10 2·61 3·46 2·85 1·58 2·36 Rb 0·07 b.d. 0·10 0·01 b.d. b.d. b.d. b.d. b.d. 0·08 b.d. 0·06 b.d. b.d. b.d. b.d. b.d. b.d. Sr 0·13 0·13 0·20 0·07 0·08 0·07 0·12 0·10 0·09 0·09 0·14 0·17 0·14 0·07 0·07 0·37 0·10 0·15 Y 12·88 7·66 35·62 33·27 35·37 14·11 13·38 13·83 12·42 14·97 26·33 57·97 46·71 49·56 43·30 24·08 34·86 59·57 Zr 4·82 13·10 9·46 9·01 10·04 7·30 7·65 3·78 3·80 4·16 25·82 18·04 16·93 9·21 9·85 11·84 12·38 13·77 Nb 0·01 0·08 0·01 0·01 0·01 b.d. b.d. 0·02 0·01 0·02 0·04 0·01 0·03 0·02 0·01 b.d. 0·01 0·01 Ba 0·01 0·22 0·08 b.d. 0·00 0·02 b.d. 0·00 0·01 0·01 0·03 0·08 b.d. b.d. b.d. b.d. 0·09 0·07 La 0·02 0·02 0·03 0·01 0·01 0·02 0·01 b.d. 0·01 0·01 0·02 0·03 0·01 0·02 0·01 b.d. 0·02 0·07 Ce 0·04 0·11 0·06 0·07 0·07 0·12 0·11 0·03 0·03 0·03 0·09 0·10 0·08 0·08 0·08 0·19 0·20 0·22 Pr 0·02 0·02 0·03 0·03 0·03 0·06 0·05 0·02 0·01 0·02 0·04 0·05 0·04 0·02 0·02 0·06 0·09 0·07 Nd 0·18 0·18 0·34 0·40 0·36 0·78 0·58 0·18 0·19 0·14 0·38 0·65 0·52 0·30 0·31 0·70 0·94 0·69 Sm 0·29 0·28 0·52 0·58 0·54 0·67 0·59 0·24 0·22 0·28 0·70 1·08 0·79 0·43 0·52 0·61 0·79 0·69 Eu 0·22 0·23 0·46 0·48 0·47 0·34 0·33 0·20 0·20 0·21 0·42 0·69 0·52 0·36 0·38 0·38 0·47 0·45 Gd 0·92 0·85 2·21 2·27 2·28 1·37 1·37 0·87 0·89 0·91 2·60 3·82 2·80 1·89 1·90 1·51 2·12 2·21 Tb 0·24 0·19 0·63 0·66 0·65 0·29 0·27 0·25 0·22 0·25 0·56 0·96 0·73 b.d. 0·60 0·46 0·55 0·65 Dy 1·98 1·33 6·39 6·74 6·74 2·37 2·10 2·11 1·96 2·17 4·65 8·94 6·94 6·17 5·60 3·48 4·92 6·88 Ho 0·49 0·30 1·68 1·95 1·84 0·51 0·50 0·55 0·48 0·58 1·03 2·27 1·83 b.d. 1·40 0·92 1·10 1·97 Er 1·45 0·84 5·78 6·69 6·34 1·48 1·49 1·63 1·39 2·00 2·79 6·82 5·84 6·73 4·92 2·91 4·10 7·63 Tm 0·21 0·11 0·88 1·04 0·97 0·20 0·22 0·24 0·20 0·34 0·38 1·04 0·90 b.d. 0·71 0·39 0·59 1·21 Yb 1·47 0·77 6·20 7·70 7·29 1·29 1·59 1·82 1·31 2·54 2·40 7·09 6·60 8·47 5·02 3·02 4·50 10·86 Lu 0·22 0·11 0·92 1·14 1·11 0·16 0·23 0·25 0·19 0·40 0·35 1·01 0·98 1·35 0·89 0·48 0·66 1·76 Hf 0·12 0·29 0·19 0·19 0·17 0·08 0·18 0·10 0·09 0·09 0·45 0·27 0·29 0·14 0·20 0·10 0·10 0·15 Ta 0·01 0·01 b.d. b.d. b.d. 0·01 0·01 b.d. b.d. 0·00 b.d. 0·00 b.d. b.d. b.d. b.d. b.d. b.d. Pb b.d. 0·16 b.d. b.d. 0·02 b.d. 0·06 b.d. b.d. b.d. b.d. 0·11 b.d. b.d. b.d. b.d. b.d. b.d. Th 0·02 b.d. 0·02 b.d. 0·01 b.d. 0·01 b.d. b.d. b.d. 0·02 0·02 b.d. b.d. b.d. b.d. b.d. b.d. U 0·02 0·02 0·01 0·04 0·02 b.d. 0·13 0·01 b.d. b.d. 0·04 0·04 0·02 b.d. 0·03 b.d. b.d. b.d. Sample: BM96-2 BM99-6 BM99-4 BME-1 BM99-2 BM96-3 BM99-5 DR9748 GN35 DR10165 Type: C C E1 E2 M E1 C E1 M C C E1 M E1 C E1 M M Li 0·31 b.d. 0·65 0·45 0·28 b.d. 0·33 b.d. b.d. 0·16 0·38 0·52 0·48 b.d. b.d. b.d. b.d. b.d. B 1·49 2·21 12·2 16·9 12·8 1·79 1·75 1·95 2·25 2·93 1·18 1·68 0·95 b.d. 10·8 b.d. 9·78 10·9 Sc 88·0 63·7 89·3 82·8 84·4 58·3 53·3 90·2 83·9 113 87·9 153 110 136 131 78·1 97·1 100 Ti 557 285 368 354 321 222 271 381 423 359 889 1043 978 569 606 356 381 333 V 51·6 38·8 59·8 58·6 55·6 47·7 57·2 62·5 66·0 62·5 133 116 118 67·3 61·4 58·5 44·6 54·1 Cr 2943 1187 1240 1283 1302 1622 1650 1415 1431 1683 747 2908 3191 2094 2091 1538 1312 1519 Co 75·4 58·4 43·3 46·4 42·9 75·1 74·3 69·8 71·2 72·4 75·4 69·3 73·3 47·9 49·6 51·0 51·2 46·7 Ni 48·4 46·7 21·8 22·7 20·6 44·0 40·5 45·9 45·1 38·7 56·6 37·2 54·9 37·2 37·8 30·4 29·4 25·7 Cu 0·17 1·27 0·69 b.d. 0·03 0·27 0·14 b.d. 0·17 0·12 0·15 0·30 0·17 0·13 0·09 0·41 0·12 b.d. Zn 17·77 6·56 14·24 18·05 14·90 18·56 19·12 23·71 24·11 22·19 29·96 41·46 38·60 7·46 6·94 6·14 6·94 6·99 Ga 6·45 4·20 3·78 3·93 3·95 4·63 4·98 6·48 6·90 6·16 8·58 7·10 7·10 2·61 3·46 2·85 1·58 2·36 Rb 0·07 b.d. 0·10 0·01 b.d. b.d. b.d. b.d. b.d. 0·08 b.d. 0·06 b.d. b.d. b.d. b.d. b.d. b.d. Sr 0·13 0·13 0·20 0·07 0·08 0·07 0·12 0·10 0·09 0·09 0·14 0·17 0·14 0·07 0·07 0·37 0·10 0·15 Y 12·88 7·66 35·62 33·27 35·37 14·11 13·38 13·83 12·42 14·97 26·33 57·97 46·71 49·56 43·30 24·08 34·86 59·57 Zr 4·82 13·10 9·46 9·01 10·04 7·30 7·65 3·78 3·80 4·16 25·82 18·04 16·93 9·21 9·85 11·84 12·38 13·77 Nb 0·01 0·08 0·01 0·01 0·01 b.d. b.d. 0·02 0·01 0·02 0·04 0·01 0·03 0·02 0·01 b.d. 0·01 0·01 Ba 0·01 0·22 0·08 b.d. 0·00 0·02 b.d. 0·00 0·01 0·01 0·03 0·08 b.d. b.d. b.d. b.d. 0·09 0·07 La 0·02 0·02 0·03 0·01 0·01 0·02 0·01 b.d. 0·01 0·01 0·02 0·03 0·01 0·02 0·01 b.d. 0·02 0·07 Ce 0·04 0·11 0·06 0·07 0·07 0·12 0·11 0·03 0·03 0·03 0·09 0·10 0·08 0·08 0·08 0·19 0·20 0·22 Pr 0·02 0·02 0·03 0·03 0·03 0·06 0·05 0·02 0·01 0·02 0·04 0·05 0·04 0·02 0·02 0·06 0·09 0·07 Nd 0·18 0·18 0·34 0·40 0·36 0·78 0·58 0·18 0·19 0·14 0·38 0·65 0·52 0·30 0·31 0·70 0·94 0·69 Sm 0·29 0·28 0·52 0·58 0·54 0·67 0·59 0·24 0·22 0·28 0·70 1·08 0·79 0·43 0·52 0·61 0·79 0·69 Eu 0·22 0·23 0·46 0·48 0·47 0·34 0·33 0·20 0·20 0·21 0·42 0·69 0·52 0·36 0·38 0·38 0·47 0·45 Gd 0·92 0·85 2·21 2·27 2·28 1·37 1·37 0·87 0·89 0·91 2·60 3·82 2·80 1·89 1·90 1·51 2·12 2·21 Tb 0·24 0·19 0·63 0·66 0·65 0·29 0·27 0·25 0·22 0·25 0·56 0·96 0·73 b.d. 0·60 0·46 0·55 0·65 Dy 1·98 1·33 6·39 6·74 6·74 2·37 2·10 2·11 1·96 2·17 4·65 8·94 6·94 6·17 5·60 3·48 4·92 6·88 Ho 0·49 0·30 1·68 1·95 1·84 0·51 0·50 0·55 0·48 0·58 1·03 2·27 1·83 b.d. 1·40 0·92 1·10 1·97 Er 1·45 0·84 5·78 6·69 6·34 1·48 1·49 1·63 1·39 2·00 2·79 6·82 5·84 6·73 4·92 2·91 4·10 7·63 Tm 0·21 0·11 0·88 1·04 0·97 0·20 0·22 0·24 0·20 0·34 0·38 1·04 0·90 b.d. 0·71 0·39 0·59 1·21 Yb 1·47 0·77 6·20 7·70 7·29 1·29 1·59 1·82 1·31 2·54 2·40 7·09 6·60 8·47 5·02 3·02 4·50 10·86 Lu 0·22 0·11 0·92 1·14 1·11 0·16 0·23 0·25 0·19 0·40 0·35 1·01 0·98 1·35 0·89 0·48 0·66 1·76 Hf 0·12 0·29 0·19 0·19 0·17 0·08 0·18 0·10 0·09 0·09 0·45 0·27 0·29 0·14 0·20 0·10 0·10 0·15 Ta 0·01 0·01 b.d. b.d. b.d. 0·01 0·01 b.d. b.d. 0·00 b.d. 0·00 b.d. b.d. b.d. b.d. b.d. b.d. Pb b.d. 0·16 b.d. b.d. 0·02 b.d. 0·06 b.d. b.d. b.d. b.d. 0·11 b.d. b.d. b.d. b.d. b.d. b.d. Th 0·02 b.d. 0·02 b.d. 0·01 b.d. 0·01 b.d. b.d. b.d. 0·02 0·02 b.d. b.d. b.d. b.d. b.d. b.d. U 0·02 0·02 0·01 0·04 0·02 b.d. 0·13 0·01 b.d. b.d. 0·04 0·04 0·02 b.d. 0·03 b.d. b.d. b.d. C, Grt in the coarse area; M, Grt in the mosaic area; E1, Grt as exsolution lamellae in Cpx megacryst; E2, Grt as exsolution lamellae in Opx megacryst; b.d., below detection limit. Fig. 7. View largeDownload slide Chondrite-normalized REE (a, c and e) and primitive-mantle normalized trace element (b, d and f) patterns of Cpx (a, b), Amp (c, d) and Grt (e, f) in garnet pyroxenite xenoliths from Lakes Bullenmerri and Gnotuk. Chondrite and primitive-mantle compositions from McDonough & Sun (1995). The green lines are minerals from samples BME-1 and DR10165. Fig. 7. View largeDownload slide Chondrite-normalized REE (a, c and e) and primitive-mantle normalized trace element (b, d and f) patterns of Cpx (a, b), Amp (c, d) and Grt (e, f) in garnet pyroxenite xenoliths from Lakes Bullenmerri and Gnotuk. Chondrite and primitive-mantle compositions from McDonough & Sun (1995). The green lines are minerals from samples BME-1 and DR10165. Cpx exhibits flat to variably light REE (LREE)-enriched patterns ([La/Nd]CN = 0·77–2·22; Fig. 7a;Table 3) and fractionated middle REE (MREE)–HREE patterns ([Sm/Yb]CN = 4·82–60·0, but mostly less than 19·2) owing to the exsolution of Grt from Cpx megacrysts. Most clinopyroxenes have slightly positive Eu anomalies (Eu/Eu* = 1·04–1·38), positive Sr anomalies and negative Nb–Ta anomalies, but no obvious Ti anomaly (Fig. 7b). Zr is fractionated from Hf ([Zr/Hf]PM = 0·46–0·74). In contrast, Cpx in samples DR10165 and BME-1 shows higher LREE contents, steeper negative MREE–HREE slopes ([Sm/Yb]CN = 29·6–60·0) and marked negative Sr and Pb anomalies along with strongly negative high field strength element (HFSE) anomalies (Fig. 7a and b), but unfractionated Zr/Hf ([Zr/Hf]PM = 0·90–1·11). Amp displays LREE-enriched and HREE-depleted REE patterns ([La/Yb]CN = 4·50–90·6), similar to the pattern of coexisting Cpx, but at slightly higher contents (Fig. 7c;Table 4). Discrete Amp has highly fractionated Nb/Ta ([Nb/Ta]PM = 1·39–5·63) and displays pronounced positive anomalies in Pb, Sr and Ti (Fig. 7d) and variable negative anomalies in Zr and Hf. In contrast, vein-related amphiboles in samples DR9748 and BM96-2 have slightly higher incompatible-element concentrations than the discrete ones, and they show pronounced positive HFSE anomalies (Fig. 7d) with little Nb–Ta fractionation ([Nb/Ta]PM = 1·07–1·51). Garnet displays LREE-depleted REE patterns ([Sm/La]CN = 15·34–131) with near-flat ([Gd/Yb]CN = 0·92; e.g. BM99-6) to elevated ([Gd/Yb]CN = 0·24; e.g. BM99-4) MREE–HREE slopes (Fig. 7e;Table 5). All garnets also show negative Ti anomalies but variable Ti contents (Fig. 7f). Bulk-rock reconstructions Reconstruction of the bulk compositions of the garnet websterites used the analysed compositions of Cpx, Grt, Opx and Amp and their modal proportions as determined by point counting (Supplementary Data Electronic Appendix 1; supplementary data are available for downloading at http://www.petrology.oxfordjournals.org). Because compositional heterogeneities have been observed in only a few grains, we assumed that most minerals have reached equilibrium. Point-counted modes slightly differ from mineral modes calculated using the least-squares methods modified from MINSQ (Table 1; Herrmann & Berry, 2002), but both methods yielded results that are within 5% relative for most major elements and 20% for compatible and moderately incompatible trace elements (i.e. DR9748; Supplementary Data Electronic Appendix 1) when compared with measured whole-rock compositions (Griffin et al., 1988). Calculated bulk-rock compositions are also comparable with those of primary clinopyroxene before subsolidus re-equilibration (Lu et al., 2017), indicating that Cpx was the main liquidus phase (Griffin et al., 1984; Lu et al., 2017). The Grt websterites have MgO of 17·1–20·8 wt %, CaO of 8·40–15·2 wt %, SiO2 of 45·3–50·7 wt %, Al2O3 of 5·73–13·9 wt %, FeO of 5·63–11·1 wt %, Na2O of 0·28–1·08 wt % and TiO2 of 0·23–0·88 wt % (Fig. 8). These major element compositions are similar to those of high-MgO Grt-pyroxenite cumulates from the Sierra Nevada (Lee et al., 2006). Both suites plot within the field of experimentally produced pyroxenite cumulates from high-P hydrous basaltic andesites (Müntener et al., 2001; Fig. 8a and c). In the pseudoternary forsterite–Ca-Tschermak pyroxene–quartz (Fo–CaTs–Qz) system projected from diopside (Di), the Grt websterites plot along the En–CaTs join, as do pyroxenite cumulates from other localities worldwide (Fig. 9). In contrast, the spinel pyroxenites from the same locality (Griffin et al., 1988), characterized by high MgO and Cr2O3 and low CaO (Fig. 8), plot between the Fo–An join and the En–CaTs join (Fig. 9), suggesting they are the products of melt–peridotite reaction (Garrido & Bodinier, 1999; Rapp et al., 1999; Liu et al., 2005; Bodinier et al., 2008; Lambart et al., 2012; Tilhac et al., 2016). Fig. 8. View largeDownload slide Whole-rock major oxides vs MgO for garnet pyroxenites (red circle; this study), spinel pyroxenites (Griffin et al., 1988) and peridotite xenoliths (O’Reilly & Griffin, 1988; Stolz & Davies, 1988; Powell, 2005) from Lakes Bullenmerri and Gnotuk. High-MgO Grt-pyroxenite cumulates (pink diamond) are from the Sierra Nevada continental arc (Lee et al., 2006); the worldwide (‘Global’) database of pyroxenite cumulates, pyroxenites from metamorphosed oceanic crust (pink shaded area) and from melt–rock reaction (yellow shaded area) are from Xiong et al. (2014). Dashed line (blue) represents fractional crystallization trend defined by experimentally derived pyroxenite cumulates from hydrous basaltic andesite at 1·2 GPa (Müntener et al., 2001). Fig. 8. View largeDownload slide Whole-rock major oxides vs MgO for garnet pyroxenites (red circle; this study), spinel pyroxenites (Griffin et al., 1988) and peridotite xenoliths (O’Reilly & Griffin, 1988; Stolz & Davies, 1988; Powell, 2005) from Lakes Bullenmerri and Gnotuk. High-MgO Grt-pyroxenite cumulates (pink diamond) are from the Sierra Nevada continental arc (Lee et al., 2006); the worldwide (‘Global’) database of pyroxenite cumulates, pyroxenites from metamorphosed oceanic crust (pink shaded area) and from melt–rock reaction (yellow shaded area) are from Xiong et al. (2014). Dashed line (blue) represents fractional crystallization trend defined by experimentally derived pyroxenite cumulates from hydrous basaltic andesite at 1·2 GPa (Müntener et al., 2001). Fig. 9. View largeDownload slide Normative compositions of pyroxenite xenoliths from Lakes Bullenmerri and Gnotuk in the pseudoternary system Fo–CaTs–Qz projected from Di using the method of O’Hara (1968). Nepheline- and hypersthene-normative compositions are separated by the Di–Fo–An plane (blue line). Silica-deficient and silica-excess compositions are divided by the Di–CaTs–En plane (red line). Fo, forsterite; CaTs, Ca-Tschermak pyroxene; Qz, quartz; An, anorthite; En, enstatite; Di, diopside. The symbols and shaded fields are as in Fig. 8. Fig. 9. View largeDownload slide Normative compositions of pyroxenite xenoliths from Lakes Bullenmerri and Gnotuk in the pseudoternary system Fo–CaTs–Qz projected from Di using the method of O’Hara (1968). Nepheline- and hypersthene-normative compositions are separated by the Di–Fo–An plane (blue line). Silica-deficient and silica-excess compositions are divided by the Di–CaTs–En plane (red line). Fo, forsterite; CaTs, Ca-Tschermak pyroxene; Qz, quartz; An, anorthite; En, enstatite; Di, diopside. The symbols and shaded fields are as in Fig. 8. The Grt websterites show a limited range of low Ni contents (mostly <500 ppm) compared with the associated peridotite xenoliths (1279–2111 ppm), suggesting that the abundance of olivine in the original cumulates was low. They also have slightly lower Cr (614–2273 ppm) contents than spinel websterite and peridotite xenoliths (Griffin et al., 1988; Powell, 2005; 2810–5436 ppm and 1680–3644 ppm, respectively). The garnet websterites have relatively flat REE patterns ([La/Yb]CN = 1·15–3·41) with small positive Eu anomalies (Eu/Eu* = 0·90–1·37), although they display variable ΣREE (8·62–37·4) (Fig. 10). The flat REE patterns are similar to those of peridotite xenoliths metasomatized by silicate melts from the same locality (Powell et al., 2004). Sample BM96-3 shows HREE enrichment relative to the flat LREE ([Sm/Yb]CN = 0·71; [La/Nd]CN = 1·14), potentially indicating a slight overestimation of the modal proportion of garnet. Samples BME-1 and DR10165 have convex-upward REE patterns with elevated downward-convex LREE and fractionated MREE–HREE (Fig. 10a). In a multi-element diagram normalized to primitive mantle values (McDonough & Sun, 1995), most Grt websterites exhibit depletion in the highly incompatible elements (Rb, Ba) and negative Nb, Ta and Zr anomalies, but no significant anomalies in Hf and Ti. They also display negative Pb anomalies and positive Sr anomalies, except for samples BME-1 and DR10165, which have pronounced negative Sr anomalies (Fig. 10b). Samples BME-1 and DR10165 also show more obvious negative anomalies in the HFSE and Pb (Fig. 10b). Sample BM96-3 has lower contents of highly incompatible elements than other garnet pyroxenites and also shows positive anomalies in Zr and Hf but a negative Ti anomaly. Fig. 10. View largeDownload slide Chondrite-normalized REE (a) and primitive mantle normalized trace element (b) patterns of garnet websterite xenoliths from Lakes Bullenmerri and Gnotuk. Chondrite and primitive mantle values from McDonough & Sun (1995). Fig. 10. View largeDownload slide Chondrite-normalized REE (a) and primitive mantle normalized trace element (b) patterns of garnet websterite xenoliths from Lakes Bullenmerri and Gnotuk. Chondrite and primitive mantle values from McDonough & Sun (1995). Pressure (P) and temperature (T) estimates Thermodynamic modeling of Cpx–Opx–Grt equilibration in the system CaO–MgO–Al2O3–SiO2 (CMAS; Gasparik, 2014) indicates that the primary clinopyroxene crystallized at 1420–1450°C and 2·3–2·4 GPa (Table 1), with the exception of sample BM99-5 (1460°C and 3·0 GPa; Lu et al., 2017). These values are slightly higher than those estimated in volcanic systems using clinopyroxene-based thermobarometers (1324–1377°C and 2·0–2·3 GPa, respectively; Putirka, 2008). These conditions lie between the average mantle adiabats of Tp = 1350–1430°C (Wang et al., 2002; Putirka et al., 2007; Green, 2015), but lower than that expected from mantle plumes (Tp >1550°C; Herzberg et al., 2007; Putirka et al., 2007). P–T values derived using the Fe–Mg exchange thermometer of Ellis & Green (1979) and Al-in-Opx barometer of Wood (1974) with the recrystallized mineral assemblages suggest that these xenoliths all finally equilibrated at ∼970–1100°C and 1·2–1·8 GPa, thus falling along the xenolith-defined SEA geotherm (e.g. O’Reilly & Griffin, 1985). Thermobarometric estimates for the equilibration of REE between Grt and Cpx (Sun & Liang, 2015) appear to preserve evidence of an intermediate stage (1030°C and ∼2·1 GPa). These P–T results indicate a decompressional cooling path (Lu et al., 2017) before entrainment in the host basalt. Sr-, Nd-, and Hf-isotope geochemistry The Rb–Sr, Sm–Nd and Lu–Hf isotopic compositions of pyroxenitic garnet, clinopyroxene and amphibole from five samples are given in Table 6. Bulk-rock Sr-, Nd-, and Hf- isotope compositions were also reconstructed using counted mineral modes, high-precision in situ elemental analyses (LA-ICP-MS) and solution isotopic data on separated minerals (Table 6). Cpx has 87Sr/86Sr ranging between 0·70412 and 0·70657 (Table 6); one amphibole gives 87Sr/86Sr = 0·70550, similar to that of the coexisting Cpx (∼0·70566). The reconstructed bulk Sr-isotope compositions are controlled by Cpx, because of the negligible Sr contents in Grt. Table 6: Sr–Nd–Hf isotopes of clinopyroxene, amphibole, garnet and whole-rock of garnet pyroxenites from Lakes Bullenmerri and Gnotuk Sample Rb (ppm) Sr (ppm) 87Rb/ 86Sr 87Sr/86Sr (±2SE) Sm (ppm) Nd (ppm) 147Sm/ 144Nd 143Nd/144Nd (±2SE) εNd (t = 40 Ma) TDM (Ma) TDM2 (Ma) Lu (ppm) Hf (ppm) 176Lu/ 177Hf 176Hf/177Hf (±2SE) εHf (t = 40 Ma) TDM (Ma) BHVO-2 0·703468±8 0·512985±3 0·283107±10 BM99-4 Cpx 0·04 149 0·0008 0·705659±3 2·25 7·93 0·179 0·512607±3 –0·32 2361 420 0·03 1·14 0·004 0·282901±3 4·87 538 Amp 1·57 493 0·0090 0·705500±5 2·78 9·97 0·176 0·512699±3 1·48 1816 354 0·04 0·72 0·008 0·282959±7 6·85 504 Grt 0·56 0·38 0·928 0·512846±5 0·47 391 1·13 0·18 0·857 0·283947±6 19·3 46 WR* 1·48 4·88 0·191 0·512611±3 –0·34 3528 421 0·20 0·71 0·040 0·282945±3 5·49 — BME-1 Cpx 127 0·705815±14 2·82 15·22 0·117 0·512436±4 –3·38 1123 532 0·51 0·002 0·283051±5 10·2 Cpx-R 127 0·705848±3 0·512442±3 0·283072±26 11·0 Grt 0·61 0·66 0·586 0·512545±12 –3·64 542 0·35 0·09 0·580 0·284281±8 38·5 102 WR* 1·77 8·71 0·128 0·512444±2 –3·28 1252 528 0·07 0·37 0·026 0·283107±5 11·6 629 BM99-2 Cpx 114 0·704123±4 0·72 2·42 0·188 0·512833±8 4·00 1894 262 0·01 0·30 0·006 0·283290±11 18·6 — Grt 0·27 0·09 0·404 0·284950±21 66·8 248 WR* 0·11 0·23 0·067 0·283507±11 26·0 555 BM99-5 Cpx 109 0·705659±3 2·23 6·69 0·210 0·512862±6 4·46 11683 245 0·04 1·38 0·004 0·282811±8 1·69 682 DR9748† Cpx 0·013 116 0·0003 0·70657±4 2·50 9·00 0·169 0·512738±5 2·25 1391 326 Opx 0·29 0·94 0·185 0·512748±6 2·36 2159 322 Grt 0·79 0·57 0·833 0·512876±50 1·55 352 WR 1·38 4·89 0·171 0·512745±16 2·37 1451 322 GN35† WR 0·29 57·6 0·0146 0·704935±16 0·92 2·76 0·202 0·512838±6 3·87 4069 261 DR10165† WR 0·36 85 0·0123 0·708579±10 1·62 8·91 0·110 0·512413±20 –3·95 1088 547 Sample Rb (ppm) Sr (ppm) 87Rb/ 86Sr 87Sr/86Sr (±2SE) Sm (ppm) Nd (ppm) 147Sm/ 144Nd 143Nd/144Nd (±2SE) εNd (t = 40 Ma) TDM (Ma) TDM2 (Ma) Lu (ppm) Hf (ppm) 176Lu/ 177Hf 176Hf/177Hf (±2SE) εHf (t = 40 Ma) TDM (Ma) BHVO-2 0·703468±8 0·512985±3 0·283107±10 BM99-4 Cpx 0·04 149 0·0008 0·705659±3 2·25 7·93 0·179 0·512607±3 –0·32 2361 420 0·03 1·14 0·004 0·282901±3 4·87 538 Amp 1·57 493 0·0090 0·705500±5 2·78 9·97 0·176 0·512699±3 1·48 1816 354 0·04 0·72 0·008 0·282959±7 6·85 504 Grt 0·56 0·38 0·928 0·512846±5 0·47 391 1·13 0·18 0·857 0·283947±6 19·3 46 WR* 1·48 4·88 0·191 0·512611±3 –0·34 3528 421 0·20 0·71 0·040 0·282945±3 5·49 — BME-1 Cpx 127 0·705815±14 2·82 15·22 0·117 0·512436±4 –3·38 1123 532 0·51 0·002 0·283051±5 10·2 Cpx-R 127 0·705848±3 0·512442±3 0·283072±26 11·0 Grt 0·61 0·66 0·586 0·512545±12 –3·64 542 0·35 0·09 0·580 0·284281±8 38·5 102 WR* 1·77 8·71 0·128 0·512444±2 –3·28 1252 528 0·07 0·37 0·026 0·283107±5 11·6 629 BM99-2 Cpx 114 0·704123±4 0·72 2·42 0·188 0·512833±8 4·00 1894 262 0·01 0·30 0·006 0·283290±11 18·6 — Grt 0·27 0·09 0·404 0·284950±21 66·8 248 WR* 0·11 0·23 0·067 0·283507±11 26·0 555 BM99-5 Cpx 109 0·705659±3 2·23 6·69 0·210 0·512862±6 4·46 11683 245 0·04 1·38 0·004 0·282811±8 1·69 682 DR9748† Cpx 0·013 116 0·0003 0·70657±4 2·50 9·00 0·169 0·512738±5 2·25 1391 326 Opx 0·29 0·94 0·185 0·512748±6 2·36 2159 322 Grt 0·79 0·57 0·833 0·512876±50 1·55 352 WR 1·38 4·89 0·171 0·512745±16 2·37 1451 322 GN35† WR 0·29 57·6 0·0146 0·704935±16 0·92 2·76 0·202 0·512838±6 3·87 4069 261 DR10165† WR 0·36 85 0·0123 0·708579±10 1·62 8·91 0·110 0·512413±20 –3·95 1088 547 WR, whole-rock; Cpx-R, replication clinopyroxene; TDM and TDM2, single- and second-stage depleted mantle model ages, respectively. Accepted decay constants (λSm = 6·54 × 10–12 a–1 and λLu = 1·865 × 10–11 a–1) were used. The present-day chondritic uniform reservoir (CHUR) values of 147Sm/144Nd = 0·1960, 143Nd/144Nd = 0·512630 and176Lu/177Hf = 0·0336, 176Hf/177Hf = 0·282785 are taken from Bouvier et al. (2008); the present-day depleted mantle values of 147Sm/144Nd = 0·21357, 143Nd/144Nd = 0·51315 and 176Lu/177Hf = 0·0384, 176Hf/177Hf = 0·283251 are from DePaolo (1981) and Griffin et al. (2000), respectively. * Whole-rock Sm–Nd and Lu–Hf isotopic compositions were reconstructed from Cpx and Grt, which dominate the Sm, Nd, Lu and Hf budget. † Data from Griffin et al. (1988). Table 6: Sr–Nd–Hf isotopes of clinopyroxene, amphibole, garnet and whole-rock of garnet pyroxenites from Lakes Bullenmerri and Gnotuk Sample Rb (ppm) Sr (ppm) 87Rb/ 86Sr 87Sr/86Sr (±2SE) Sm (ppm) Nd (ppm) 147Sm/ 144Nd 143Nd/144Nd (±2SE) εNd (t = 40 Ma) TDM (Ma) TDM2 (Ma) Lu (ppm) Hf (ppm) 176Lu/ 177Hf 176Hf/177Hf (±2SE) εHf (t = 40 Ma) TDM (Ma) BHVO-2 0·703468±8 0·512985±3 0·283107±10 BM99-4 Cpx 0·04 149 0·0008 0·705659±3 2·25 7·93 0·179 0·512607±3 –0·32 2361 420 0·03 1·14 0·004 0·282901±3 4·87 538 Amp 1·57 493 0·0090 0·705500±5 2·78 9·97 0·176 0·512699±3 1·48 1816 354 0·04 0·72 0·008 0·282959±7 6·85 504 Grt 0·56 0·38 0·928 0·512846±5 0·47 391 1·13 0·18 0·857 0·283947±6 19·3 46 WR* 1·48 4·88 0·191 0·512611±3 –0·34 3528 421 0·20 0·71 0·040 0·282945±3 5·49 — BME-1 Cpx 127 0·705815±14 2·82 15·22 0·117 0·512436±4 –3·38 1123 532 0·51 0·002 0·283051±5 10·2 Cpx-R 127 0·705848±3 0·512442±3 0·283072±26 11·0 Grt 0·61 0·66 0·586 0·512545±12 –3·64 542 0·35 0·09 0·580 0·284281±8 38·5 102 WR* 1·77 8·71 0·128 0·512444±2 –3·28 1252 528 0·07 0·37 0·026 0·283107±5 11·6 629 BM99-2 Cpx 114 0·704123±4 0·72 2·42 0·188 0·512833±8 4·00 1894 262 0·01 0·30 0·006 0·283290±11 18·6 — Grt 0·27 0·09 0·404 0·284950±21 66·8 248 WR* 0·11 0·23 0·067 0·283507±11 26·0 555 BM99-5 Cpx 109 0·705659±3 2·23 6·69 0·210 0·512862±6 4·46 11683 245 0·04 1·38 0·004 0·282811±8 1·69 682 DR9748† Cpx 0·013 116 0·0003 0·70657±4 2·50 9·00 0·169 0·512738±5 2·25 1391 326 Opx 0·29 0·94 0·185 0·512748±6 2·36 2159 322 Grt 0·79 0·57 0·833 0·512876±50 1·55 352 WR 1·38 4·89 0·171 0·512745±16 2·37 1451 322 GN35† WR 0·29 57·6 0·0146 0·704935±16 0·92 2·76 0·202 0·512838±6 3·87 4069 261 DR10165† WR 0·36 85 0·0123 0·708579±10 1·62 8·91 0·110 0·512413±20 –3·95 1088 547 Sample Rb (ppm) Sr (ppm) 87Rb/ 86Sr 87Sr/86Sr (±2SE) Sm (ppm) Nd (ppm) 147Sm/ 144Nd 143Nd/144Nd (±2SE) εNd (t = 40 Ma) TDM (Ma) TDM2 (Ma) Lu (ppm) Hf (ppm) 176Lu/ 177Hf 176Hf/177Hf (±2SE) εHf (t = 40 Ma) TDM (Ma) BHVO-2 0·703468±8 0·512985±3 0·283107±10 BM99-4 Cpx 0·04 149 0·0008 0·705659±3 2·25 7·93 0·179 0·512607±3 –0·32 2361 420 0·03 1·14 0·004 0·282901±3 4·87 538 Amp 1·57 493 0·0090 0·705500±5 2·78 9·97 0·176 0·512699±3 1·48 1816 354 0·04 0·72 0·008 0·282959±7 6·85 504 Grt 0·56 0·38 0·928 0·512846±5 0·47 391 1·13 0·18 0·857 0·283947±6 19·3 46 WR* 1·48 4·88 0·191 0·512611±3 –0·34 3528 421 0·20 0·71 0·040 0·282945±3 5·49 — BME-1 Cpx 127 0·705815±14 2·82 15·22 0·117 0·512436±4 –3·38 1123 532 0·51 0·002 0·283051±5 10·2 Cpx-R 127 0·705848±3 0·512442±3 0·283072±26 11·0 Grt 0·61 0·66 0·586 0·512545±12 –3·64 542 0·35 0·09 0·580 0·284281±8 38·5 102 WR* 1·77 8·71 0·128 0·512444±2 –3·28 1252 528 0·07 0·37 0·026 0·283107±5 11·6 629 BM99-2 Cpx 114 0·704123±4 0·72 2·42 0·188 0·512833±8 4·00 1894 262 0·01 0·30 0·006 0·283290±11 18·6 — Grt 0·27 0·09 0·404 0·284950±21 66·8 248 WR* 0·11 0·23 0·067 0·283507±11 26·0 555 BM99-5 Cpx 109 0·705659±3 2·23 6·69 0·210 0·512862±6 4·46 11683 245 0·04 1·38 0·004 0·282811±8 1·69 682 DR9748† Cpx 0·013 116 0·0003 0·70657±4 2·50 9·00 0·169 0·512738±5 2·25 1391 326 Opx 0·29 0·94 0·185 0·512748±6 2·36 2159 322 Grt 0·79 0·57 0·833 0·512876±50 1·55 352 WR 1·38 4·89 0·171 0·512745±16 2·37 1451 322 GN35† WR 0·29 57·6 0·0146 0·704935±16 0·92 2·76 0·202 0·512838±6 3·87 4069 261 DR10165† WR 0·36 85 0·0123 0·708579±10 1·62 8·91 0·110 0·512413±20 –3·95 1088 547 WR, whole-rock; Cpx-R, replication clinopyroxene; TDM and TDM2, single- and second-stage depleted mantle model ages, respectively. Accepted decay constants (λSm = 6·54 × 10–12 a–1 and λLu = 1·865 × 10–11 a–1) were used. The present-day chondritic uniform reservoir (CHUR) values of 147Sm/144Nd = 0·1960, 143Nd/144Nd = 0·512630 and176Lu/177Hf = 0·0336, 176Hf/177Hf = 0·282785 are taken from Bouvier et al. (2008); the present-day depleted mantle values of 147Sm/144Nd = 0·21357, 143Nd/144Nd = 0·51315 and 176Lu/177Hf = 0·0384, 176Hf/177Hf = 0·283251 are from DePaolo (1981) and Griffin et al. (2000), respectively. * Whole-rock Sm–Nd and Lu–Hf isotopic compositions were reconstructed from Cpx and Grt, which dominate the Sm, Nd, Lu and Hf budget. † Data from Griffin et al. (1988). The high Sr and low Rb contents of the Cpx and Amp made them suitable for in situ analysis. Within each sample, in situ Sr-isotope compositions of the megacrystalline Cpx and the discrete Cpx are indistinguishable within analytical uncertainty (Supplementary Data Electronic Appendix 2); the average data are listed in Table 7. They give 87Sr/86Sr of 0·7044–0·7084, mainly concentrated at <0·706. Compared with Cpx, amphiboles show slightly higher 87Rb/86Sr (0·008–0·083) but similar 87Sr/86Sr (0·7044–0·7087); 87Sr/86Sr values are well correlated with those of Cpx, suggesting that the Cpx and Amp have achieved near-isotopic equilibrium in terms of Sr. A marked Sr-isotope discrepancy is found in sample DR9748, where the discrete grains of Cpx have higher 87Sr/86Sr than the Cpx close to the Amp vein (0·7074 vs 0·7059). This discrepancy is also observed between the discrete Amp (∼0·7069) and the vein Amp (∼0·7056), which is slightly lower but less variable than those of Cpx. Table 7: Mean Sr isotopic compositions of clinopyroxene and amphibole in the garnet pyroxenite from Lakes Bullenmerri and Gnotuk by LA-MC-ICP-MS Sample Mineral Rb Sr 87Rb/86Sr 87Sr/86Sr 1σ BME-1 Cpx (5) 127 0·70610 0·00014 Amp (4) 1·01 371 0·008 0·70605 0·00007 BM99-4 Cpx (7) 0·04 149 0·001 0·70627 0·00008 Amp (4) 1·57 493 0·009 0·70604 0·00004 BM99-2 Cpx (4) 114 0·70447 0·00008 BM99-5 Cpx (4) 109 0·70685 0·00016 Amp (4) 8·88 384 0·068 0·70612 0·00007 BM96-3 Cpx (11) 104 0·70621 0·00009 Amp (3) 3·17 359 0·026 0·70599 0·00003 BM96-2 Cpx (5) 146 0·70444 0·00007 Amp (5) 3·17 359 0·067 0·70443 0·00004 BM99-6 Cpx (11) 137 0·70449 0·00006 Amp (4) 6·25 386 0·048 0·70461 0·00006 DR10165 Cpx (6) 136 0·70844 0·00081 Amp (6) 2·31 540 0·012 0·70872 0·00001 GN35 Cpx (5) 96 0·70563 0·00045 Amp (4) 8·77 311 0·083 0·70549 0·00009 DR9748 Cpx (4) 135 0·70749 0·00022 Cpx-V (4) 0·70598 0·00026 Amp (4) 3·88 525 0·022 0·70698 0·00004 Amp-V (4) 13·06 577 0·066 0·70569 0·00001 Sample Mineral Rb Sr 87Rb/86Sr 87Sr/86Sr 1σ BME-1 Cpx (5) 127 0·70610 0·00014 Amp (4) 1·01 371 0·008 0·70605 0·00007 BM99-4 Cpx (7) 0·04 149 0·001 0·70627 0·00008 Amp (4) 1·57 493 0·009 0·70604 0·00004 BM99-2 Cpx (4) 114 0·70447 0·00008 BM99-5 Cpx (4) 109 0·70685 0·00016 Amp (4) 8·88 384 0·068 0·70612 0·00007 BM96-3 Cpx (11) 104 0·70621 0·00009 Amp (3) 3·17 359 0·026 0·70599 0·00003 BM96-2 Cpx (5) 146 0·70444 0·00007 Amp (5) 3·17 359 0·067 0·70443 0·00004 BM99-6 Cpx (11) 137 0·70449 0·00006 Amp (4) 6·25 386 0·048 0·70461 0·00006 DR10165 Cpx (6) 136 0·70844 0·00081 Amp (6) 2·31 540 0·012 0·70872 0·00001 GN35 Cpx (5) 96 0·70563 0·00045 Amp (4) 8·77 311 0·083 0·70549 0·00009 DR9748 Cpx (4) 135 0·70749 0·00022 Cpx-V (4) 0·70598 0·00026 Amp (4) 3·88 525 0·022 0·70698 0·00004 Amp-V (4) 13·06 577 0·066 0·70569 0·00001 Cpx-V, Amp-V, clinopyroxene and amphibole in the vein, respectively. Numbers in parentheses indicate the number of analyzed grains. Table 7: Mean Sr isotopic compositions of clinopyroxene and amphibole in the garnet pyroxenite from Lakes Bullenmerri and Gnotuk by LA-MC-ICP-MS Sample Mineral Rb Sr 87Rb/86Sr 87Sr/86Sr 1σ BME-1 Cpx (5) 127 0·70610 0·00014 Amp (4) 1·01 371 0·008 0·70605 0·00007 BM99-4 Cpx (7) 0·04 149 0·001 0·70627 0·00008 Amp (4) 1·57 493 0·009 0·70604 0·00004 BM99-2 Cpx (4) 114 0·70447 0·00008 BM99-5 Cpx (4) 109 0·70685 0·00016 Amp (4) 8·88 384 0·068 0·70612 0·00007 BM96-3 Cpx (11) 104 0·70621 0·00009 Amp (3) 3·17 359 0·026 0·70599 0·00003 BM96-2 Cpx (5) 146 0·70444 0·00007 Amp (5) 3·17 359 0·067 0·70443 0·00004 BM99-6 Cpx (11) 137 0·70449 0·00006 Amp (4) 6·25 386 0·048 0·70461 0·00006 DR10165 Cpx (6) 136 0·70844 0·00081 Amp (6) 2·31 540 0·012 0·70872 0·00001 GN35 Cpx (5) 96 0·70563 0·00045 Amp (4) 8·77 311 0·083 0·70549 0·00009 DR9748 Cpx (4) 135 0·70749 0·00022 Cpx-V (4) 0·70598 0·00026 Amp (4) 3·88 525 0·022 0·70698 0·00004 Amp-V (4) 13·06 577 0·066 0·70569 0·00001 Sample Mineral Rb Sr 87Rb/86Sr 87Sr/86Sr 1σ BME-1 Cpx (5) 127 0·70610 0·00014 Amp (4) 1·01 371 0·008 0·70605 0·00007 BM99-4 Cpx (7) 0·04 149 0·001 0·70627 0·00008 Amp (4) 1·57 493 0·009 0·70604 0·00004 BM99-2 Cpx (4) 114 0·70447 0·00008 BM99-5 Cpx (4) 109 0·70685 0·00016 Amp (4) 8·88 384 0·068 0·70612 0·00007 BM96-3 Cpx (11) 104 0·70621 0·00009 Amp (3) 3·17 359 0·026 0·70599 0·00003 BM96-2 Cpx (5) 146 0·70444 0·00007 Amp (5) 3·17 359 0·067 0·70443 0·00004 BM99-6 Cpx (11) 137 0·70449 0·00006 Amp (4) 6·25 386 0·048 0·70461 0·00006 DR10165 Cpx (6) 136 0·70844 0·00081 Amp (6) 2·31 540 0·012 0·70872 0·00001 GN35 Cpx (5) 96 0·70563 0·00045 Amp (4) 8·77 311 0·083 0·70549 0·00009 DR9748 Cpx (4) 135 0·70749 0·00022 Cpx-V (4) 0·70598 0·00026 Amp (4) 3·88 525 0·022 0·70698 0·00004 Amp-V (4) 13·06 577 0·066 0·70569 0·00001 Cpx-V, Amp-V, clinopyroxene and amphibole in the vein, respectively. Numbers in parentheses indicate the number of analyzed grains. Overall, Sr-isotope compositions obtained by solution are less variable than those from in situ analysis, but they are still positively correlated with each other, probably owing to the averaging effect of digesting whole grains. Grt has higher 147Sm/144Nd (0·586–0·928) and more radiogenic 143Nd/144Nd (0·51255–0·51288) than the corresponding Cpx (0·117–0·210 and 0·51244–0·51274, respectively). The reconstructed bulk Nd-isotope compositions range from 0·51244 to 0·51275, essentially identical to that of Cpx; this reflects the much higher Nd and Sm contents in Cpx, and its greater modal abundance compared with Grt. Grt has high 176Lu/177Hf (0·580–0·857) and radiogenic 176Hf/177Hf (0·28395–0·28495), whereas Cpx has a restricted 176Hf/177Hf range of 0·28281–0·28329. The reconstructed bulk Hf-isotope compositions are intermediate between those of Grt (essentially controlling Lu) and Cpx (controlling Hf), but they still show an obvious positive correlation with Cpx compositions. Internal Sm–Nd isochrons of three Grt–Cpx pairs give similar ages with a mean of ∼40 Ma (Fig. 11a); Lu–Hf isochrons yield older ages ranging from 70 to 220 Ma (Fig. 11b). Fig. 11. View largeDownload slide Sm–Nd (a) and Lu–Hf (b) internal isochrons defined by Grt and Cpx pairs in the garnet pyroxenites from Lakes Bullenmerri and Gnotuk. Ages are reported at 2σ level. Fig. 11. View largeDownload slide Sm–Nd (a) and Lu–Hf (b) internal isochrons defined by Grt and Cpx pairs in the garnet pyroxenites from Lakes Bullenmerri and Gnotuk. Ages are reported at 2σ level. Hf depleted-mantle model ages (TDM) of the reconstructed bulk-rocks and clinopyroxenes are 538–682 Ma, with the exception of the reconstructed bulk-rock of BM99-4 and the clinopyroxene in BM99-2, which give negative (future) values (Table 6). Garnets have a wide range of TDM from 46 to 248 Ma. In contrast, most bulk-rocks and clinopyroxenes (this study; Griffin et al., 1988) give old depleted-mantle Nd model ages (mostly >1000 Ma; Table 6). However, second-stage depleted-mantle model ages (TDM2), calculated assuming the absence of radiogenic ingrowth before 40 Ma (i.e. 147Sm/144Nd = 0), yield an average of 388 ± 107 Ma (Table 6), comparable with the TDM2 ages calculated from the compositions of metasomatic peridotite xenoliths in this area (Griffin et al., 1988; Stolz & Davies, 1988). On a plot of age-corrected 87Sr/86Sr vs εNd at t = 40 Ma, most samples are restricted to the field of ocean island basalt (OIB) (Fig. 12a), contrasting with the limited range of compositions reported for the host basalts (87Sr/86Sr = 0·70395–0·70409; εNd = +3·7; Griffin et al., 1988; Stolz & Davies, 1988). However, they overlap with the host peridotites and Cenozoic basalts, which mostly cluster within the MORB–OIB mantle array (87Sr/86Sr < 0·7060; εNd(t) > –4·6), although a few peridotites metasomatized by pyroxenitic melts have 87Sr/86Sr up to 0·7105 and low εNd(t) (–7·7; Stolz & Davies, 1988). On a plot of age-corrected εNd vs εHf at t = 40 Ma, pyroxenitic clinopyroxenes are mostly scattered around the OIB field along the mantle array (Fig. 12b). Sample BM99-4 is within the OIB field, whereas sample BME-1 tends to have more radiogenic Hf at a given εNd(t). In addition, clinopyroxenes in samples BM99-2 and BM99-5 have similar Nd-isotope compositions but very different εHf(t) (+18·6 and +1·69, respectively). Fig. 12. View largeDownload slide Variation of 87Sr/86Sr vs εNd (a) and εNd vs εHf (b) at t = 40 Ma for Cpx, Amp and whole-rocks (WR) of garnet pyroxenites from Lakes Bullenmerri and Gnotuk. In (a), fields for MORB and OIB are from Zindler & Hart (1986); data for peridotites from this area are from Griffin et al. (1988) and Stolz & Davies (1988); data for Cenozoic basalts in easternAustralia are from O’Reilly & Zhang (1995), Vogel & Keays (1997), Zhang & O’Reilly (1997), McBride et al. (2001), Demidjuk et al. (2007), Price et al. (2014) and Oostingh et al. (2016). In (b), fields for MORB, OIB and the mantle array are from Vervoort et al. (1999); field for Fe–Mn crust and seawater array are from Albarède et al. (1998). Fig. 12. View largeDownload slide Variation of 87Sr/86Sr vs εNd (a) and εNd vs εHf (b) at t = 40 Ma for Cpx, Amp and whole-rocks (WR) of garnet pyroxenites from Lakes Bullenmerri and Gnotuk. In (a), fields for MORB and OIB are from Zindler & Hart (1986); data for peridotites from this area are from Griffin et al. (1988) and Stolz & Davies (1988); data for Cenozoic basalts in easternAustralia are from O’Reilly & Zhang (1995), Vogel & Keays (1997), Zhang & O’Reilly (1997), McBride et al. (2001), Demidjuk et al. (2007), Price et al. (2014) and Oostingh et al. (2016). In (b), fields for MORB, OIB and the mantle array are from Vervoort et al. (1999); field for Fe–Mn crust and seawater array are from Albarède et al. (1998). DISCUSSION Origin of the garnet pyroxenites Pyroxenites are an important component in the Earth’s upper mantle, yet their genesis remains controversial. Proposed pyroxenite-forming processes are variable, as pyroxenites are petrologically, chemically and isotopically diverse (e.g. Downes, 2007). We here discuss the origin of the garnet websterites as potentially representing (1) solidified melts, (2) melt–rock reaction products, (3) metamorphosed oceanic crust, or (4) cumulates. If the pyroxenites represent solidified melts, their bulk-rock compositions would be similar to those of primary melts generated by partial melting of peridotite, or more evolved melts. However, their MgO contents are too high to represent typical basaltic melts at mantle depths. In particular, melts with such high MgO contents (i.e. 17·0–20·8 wt %) have been produced experimentally by melting dry fertile mantle compositions at ∼1500–1600°C (e.g. Falloon et al., 1988; Hirose & Kushiro, 1993; Walter, 1998); such melts are associated with Phanerozoic mantle plumes (i.e. Hawaii picrites) or the Archean subcontinental mantle (e.g. komatiites). Picrites or komatiites have either accumulated olivine or are olivine-saturated, even at high pressure, so that they can reach such high MgO and Mg# (Lee et al., 2006). In the present case, the low Ni contents (<500 ppm) and high CaO (Supplementary Data Electronic Appendix 1) contents of the garnet pyroxenites preclude significant olivine accumulation. Experimental studies suggest that the addition of H2O reduces the solidus temperature by 100–150°C compared with anhydrous melting of a lherzolite source, which means that the pyroxenites could alternatively come from solidified magmas derived from the flux melting of a fertile mantle source at ∼1380–1480°C beneath volcanic arcs (Ulmer, 2001). However, the pyroxenites have Mg# higher than any type of primitive arc magmas (Kelemen et al., 2003a). The garnet pyroxenites are therefore unlikely to represent solidified basaltic melts. The second hypothesis is that the garnet pyroxenites are the products of melt–rock reaction. It has been shown on theoretical, observational and experimental grounds that pyroxene-rich mantle assemblages can be generated by reaction of a silicic melt with peridotite (Garrido & Bodinier, 1999; Rapp et al., 1999; Liu et al., 2005; Bodinier et al., 2008; Lambart et al., 2012). Such ‘secondary’ pyroxenites have compositions transitional between peridotites and silicic mantle melts, and thus plot between the En–CaTs and Fo–An joins in the CaTs–Fo–Qz diagram projected from Di (Lambart et al., 2012; Mallik & Dasgupta, 2012; Tilhac et al., 2016). They also commonly have high Mg#, and high concentrations of Ni and other compatible elements released by olivine when reaction with a silicic melt converts olivine to pyroxene to generate pyroxenites or pyroxene-rich peridotites (Garrido & Bodinier, 1999; Liu et al., 2005; Tilhac et al., 2016). These features are shown by the spinel pyroxenites from Lakes Bullenmerri and Gnotuk, which have high Mg# (up to 90; Fig. 8) and Ni contents (760–924 ppm; Griffin et al., 1984). In addition, composite xenoliths of spinel pyroxenites and their host lherzolites accordingly show variable Cr2O3 in spinel across the contact zone (Fig. 5f;Griffin et al., 1984). However, the garnet pyroxenites studied here lie mostly along the En–CaTs join (Fig. 9) and show a limited range of low Ni contents. As noted above, the garnet pyroxenites provide little or no compositional evidence for the involvement of olivine in their formation, indicating limited, if any, contribution of melt–peridotite reactions. Recycling of crustal material as a mechanism for the formation of garnet pyroxenites has attracted much attention [see reviews by Spray (1989) and Downes (2007)]. This type of pyroxenite is expected to inherit its chemical and isotopic compositions from low-P crustal cumulates (e.g. gabbro and MORB). These pyroxenites would have lower Mg# and MgO but higher Al2O3, CaO and Na2O contents than most mantle pyroxenites (Fig. 8), and would lie along the Fo–An join in the Fo–CaTs–Qz (+Di) system (Fig. 9). Their whole-rock trace element patterns also may show positive Eu and Sr anomalies, relatively low HREE and MREE, and high LREE/HREE owing to the accumulation of primary plagioclase. In addition, garnet and clinopyroxene (the two dominant phases) show ‘ghost’ signatures of plagioclase (positive Eu and Sr anomalies), and garnet has depleted HREE patterns, suggesting a metamorphic origin (Morishita et al., 2009; Yu et al., 2010; Svojtka et al., 2016). Despite their positive Eu anomalies, the garnet pyroxenites studied here are clearly distinct from recycled oceanic gabbros and primary MORB (Figs 8 and 9), precluding their origin as recycled mafic crust in the convective mantle. Their reconstructed bulk-rock compositions are indeed similar to those of other pyroxenites recognized as cumulates worldwide (Figs 8 and 9), and they plot among the pyroxenitic cumulates produced experimentally from high-P hydrous basalts (Müntener et al., 2001). Therefore, the most probable hypothesis for the origin of the garnet pyroxenites is that they are Cpx-dominated cumulates (sensu lato) from relatively primitive mantle melts (Bultitude & Green, 1971). Nature of the parental melt(s) and source region To clarify the geochemical affinity of the parental melt(s) of the garnet pyroxenites, we have calculated the melt compositions in equilibrium with the reconstructed primary clinopyroxenes (Supplementary Data Electronic Appendix 3). We used partition coefficients reported by Bodinier et al. (1987) for Cpx at 1400°C and 1·5–2·5 GPa, comparable with our thermobarometric estimates for the initial crystallization conditions, and the procedure of Hanson & Langmuir (1978). Calculated melts have compositions typical of tholeiitic basalts (Supplementary Data Electronic Appendix 3), as indicated by their SiO2 contents (∼50 wt %) and the variations in Mg# (58–64). The melt in equilibrium with the reconstructed Cpx for sample BM99-5 has high MgO (∼10·9%) but lower Mg# (∼48·1), probably indicating Fe enrichment in the source, which is reflected in the exsolution of ilmenite from clinopyroxene megacrysts during cooling. Sample BM99-4 contains primary orthopyroxene, and thus probably represents the most primitive pyroxenite in this study, as Opx is expected to appear early on the liquid line of descent of tholeiites (Demarchi et al., 2001). We also have estimated the trace element compositions of melts in equilibrium with the reconstructed primary Cpx (Fig. 13; Supplementary Data Electronic Appendix 3). The partition coefficients used in the calculations are taken from Adam & Green (2006) except for Ba, Sr, Eu, Dy and Er, which are from Hart & Dunn (1993). Calculated melts for samples BM99-4 and BM99-5 show tholeiite-like or OIB-like trace element characteristics, with moderate enrichment of LREE and the absence of HFSE anomalies (Fig. 13). These garnet websterites also show OIB-like Sr-, Nd-, and Hf-isotope compositions (Fig. 12). Their melts may represent the primary compositions of magmas parental to garnet pyroxenites. For sample BM99-2, calculated melts have REE patterns similar to those of sample BM99-4 but lower REE contents, probably reflecting a differentiation process. Therefore, it seems that the garnet pyroxenites studied here crystallized from primitive to moderately differentiated tholeiitic basalts. Fig. 13. View largeDownload slide REE patterns normalized to chondrites (a) and trace element diagrams normalized to primitive mantle (b) for melts in equilibrium with garnet pyroxenites from Lakes Bullenmerri and Gnotuk. Patterns are calculated using the reconstructed Cpx and mineral–melt partition coefficients listed by Hart & Dunn (1993) and Adam et al. (2006). The data for Newer Volcanic Province (NVP) tholeiitic basalts are from Price et al. (1997), Vogel & Keays (1997) and McBride et al. (2001). The data for OIB are from McDonough & Sun (1995). Fig. 13. View largeDownload slide REE patterns normalized to chondrites (a) and trace element diagrams normalized to primitive mantle (b) for melts in equilibrium with garnet pyroxenites from Lakes Bullenmerri and Gnotuk. Patterns are calculated using the reconstructed Cpx and mineral–melt partition coefficients listed by Hart & Dunn (1993) and Adam et al. (2006). The data for Newer Volcanic Province (NVP) tholeiitic basalts are from Price et al. (1997), Vogel & Keays (1997) and McBride et al. (2001). The data for OIB are from McDonough & Sun (1995). However, experimental studies performed at 2 and 3 GPa on a typical anhydrous tholeiitic basalt (MORB-like eclogite G2; Pertermann & Hirschmann, 2003a, 2003b), suggest that only extremely high degrees of melting (>90%, or, equivalently, extremely low degrees of crystallization from a magma of this composition) could produce the high MgO values characteristic of the garnet pyroxenites studied here (Fig. 14a). Moreover, the garnet pyroxenites tend to have lower CaO and Al2O3 than the melting residues (Fig. 14b–d). It thus seems unlikely that the garnet pyroxenites were derived from anhydrous tholeiitic basalts. Fig. 14. View largeDownload slide Covariation diagrams modelling petrogenetic relationships between garnet pyroxenites (this study), experimental cumulates from hydrous basaltic andesite at 1·2 Ga (Müntener et al., 2001), melt compositions and solid residues from partial melting of anhydrous MORB-like eclogite (similar to a typical anhydrous tholeiitic basalt; Pertermann & Hirschmann, 2003b). Fig. 14. View largeDownload slide Covariation diagrams modelling petrogenetic relationships between garnet pyroxenites (this study), experimental cumulates from hydrous basaltic andesite at 1·2 Ga (Müntener et al., 2001), melt compositions and solid residues from partial melting of anhydrous MORB-like eclogite (similar to a typical anhydrous tholeiitic basalt; Pertermann & Hirschmann, 2003b). The garnet pyroxenites have compositions similar to those of the high-MgO Sierra Nevada arc pyroxenites, which are interpreted as cumulates from mantle-wedge-derived hydrous basalt or basaltic andesite (Müntener et al., 2001; Lee et al., 2006). Müntener et al. (2001) showed that a hydrous high-MgO basaltic andesite (51·7 wt % SiO2, 16·4 wt % Al2O3, 10·8 wt % MgO, 9·67 wt % CaO and Mg# = 71; Baker et al., 1994) could crystallize clinopyroxene–orthopyroxene (± garnet) assemblages at 1·2 GPa. In terms of major elements (Fig. 14), the garnet pyroxenites studied here are closely similar to experimental cumulates from a hydrous basaltic andesite (Fig. 14). Furthermore, discrete amphibole grains with minor Cl contents (Table 6) are abundant in the garnet pyroxenites, and they are in equilibrium with the recrystallized Cpx in the garnet pyroxenites (Fig. 3e), consistent with crystallization of amphibole from the migrating or trapped residual melts (Harte et al., 1993). Minor ‘exsolution-like’ amphibole lamellae and pods (Fig. 3d) and F–Cl-bearing apatite inclusions were also found in the clinopyroxene megacrysts. These results indicate a high volatile budget in the bulk system. In addition, these xenoliths also contain abundant fluid inclusions and cavities (>3 vol. %) trapped at high pressure (Andersen et al., 1984; Griffin et al., 1984; O’Reilly et al., 1990). CO2 is the dominant phase now present, but it is interpreted as the residual fluid left after a post-entrapment reaction from an initial CO2–H2O mixture, which contained small amounts of halogens (F, Cl) and sulfur species (Andersen et al., 1984). The occurrence of halogens may suggest an oxidizing environment. We thus suggest that upon percolation, these volatiles were released into the surrounding wall-rock and most of the H2O reacted to produce amphibole, leaving a CO2-rich fluid. Such a compositional shift from H2O-rich fluids at depth to CO2-rich residual liquids at shallower levels, and associated amphibolization, have also been described in experimental results (Schneider & Eggler, 1986) and in other arc-related pyroxenites (Burg et al., 1998; Tilhac et al., 2016). A mantle-wedge environment is specifically suggested for the genesis of the parental melts. Samples BME-1 and DR10165 show strong enrichment of large ion lithophile elements (LILE) and LREE and marked negative anomalies of HFSE, a typical arc signature (Fig. 10). Furthermore, clinopyroxene and the bulk-rocks in BME-1 and DR10165 have radiogenic Sr-isotope compositions (87Sr/86Sr up to 0·708) and low εNd(t), but relatively high εHf(t), which plot between a depleted-mantle source and a subducted oceanic sediment (Fig. 12). Trace element compositions and decoupled Nd–Hf isotopes probably suggest that these garnet pyroxenites were derived from a metasomatic mantle source contaminated by subduction-related sediments (e.g. Griffin et al., 1988). The melt in equilibrium with the reconstructed Cpx of BME-1 also has highly fractionated LREE/HREE and depleted HFSE (Fig. 13), and a high Sr/Y ratio, similar to the melt produced by metasomatized peridotite (e.g. Rapp et al., 1999). In addition, the estimated crystallization temperature of the primary Cpx (1420–1450°C and 2·3–2·3 GPa) is close to the mantle potential temperature (Table 1), indicating that the melts parental to the garnet pyroxenites studied here did not form in the lithospheric mantle but in the hottest part of the mantle wedge (1350–1450°C, 2·0–2·5 GPa; Ulmer, 2001; Kelemen et al., 2003b), where tholeiitic or high-Al basaltic melts were produced during the initial stage by high degrees of partial melting (10–15%) of a mantle source with a low water content (<0·0 wt %; Baker et al., 1994; Ulmer, 2001; Grove et al., 2006; Kushiro, 2007). Such low H2O contents in the initial lavas are found in the Cascades, western USA and Galunggung, Indonesia (Sisson & Layne, 1993; Sisson & Bronto, 1998). We thus propose that the garnet pyroxenites may best be interpreted as cumulates from tholeiitic magmas produced by hydrous melting of metasomatized fertile peridotite in a convective mantle wedge. Constraints on the age of the garnet pyroxenites No zircons or other phases directly datable by the U–Pb system were found in these pyroxenite xenoliths and thus ages could only be estimated from isochron and model ages calculated from our radiogenic isotopic data. Generally, the Lu–Hf system has a higher closure temperature for Cpx–Grt pairs (Blichert-Toft, 1999; Cheng et al., 2008; Gonzaga et al., 2010; Shu et al., 2014); it thus resists disturbance during thermo-tectonic processes, and may preserve primary age information (Jacob et al., 2004; Gonzaga, 2007; Xiong et al., 2015). The Rb–Sr and Sm–Nd systems both have lower closure temperatures and are more easily disturbed by later events, such as the interaction with fluids (e.g. Gysi et al., 2011; Xiong et al., 2014; Tilhac et al., 2017). Model ages theoretically correspond to the timing of separation (e.g. change of parent–daughter ratios following melt extraction) of a given sample (or domain) from a modelled reservoir, for instance, a convecting upper mantle (depleted mantle source). Depleted-mantle Nd model ages (TDM) of the garnet pyroxenites are over 1000 Ma (Table 6; Griffin et al., 1988), and older than those of peridotite xenoliths from these localities that were metasomatized by pyroxenitic melts (concentrated between 600 and 1000 Ma; Griffin et al., 1988; Stolz & Davies, 1988). This may suggest that the Sm–Nd system in the garnet pyroxenites has been disturbed below its blocking temperature by later events, possibly the infiltration of volatile-rich fluids (leading to the formation of vein amphibole and phlogopite; Figs 3f and 4e, f) and thus Nd model ages could be meaningless. However, second-stage Nd model ages (TDM2) calculated from these garnet websterites are comparable with the estimates (300–500 Ma; Griffin et al., 1988) derived from a mixing model based on the Sr and Nd isotopic compositions of metasomatic peridotites and melt-related garnet pyroxenites, and may represent the intrusive ages of the garnet pyroxenites (Griffin et al., 1988). One garnet pyroxenite from the studied area has given a whole-rock Re–Os model age (TMA) of 560 Ma (McBride et al., 1996) and clinopyroxenes and reconstructed bulk-rocks in this study give Hf model ages of 538–682 Ma (Table 6). Considering the geological history of the area, it is reasonable to conclude that the cumulate precursors of the garnet pyroxenites could have crystallized at ∼600–300 Ma during the extended Paleozoic orogenic history of the area, although we emphasize that these model ages have large uncertainties. Garnet–clinopyroxene pairs give similar Sm–Nd isochron ages with a mean of 40 Ma (Fig. 11a). If this age is meaningful, it would represent a ‘cooling age’ below the blocking temperature of the system rather than the actual timing of pyroxenite formation (Blichert-Toft, 1999; Montanini et al., 2006, 2012; Ishikawa et al., 2007; Gysi et al., 2011; Shu et al., 2014; Xiong et al., 2014; Borghini et al., 2016). Petrographic observations suggest that most garnets have formed by subsolidus exsolution from clinopyroxene megacrysts upon cooling and reconstructed bulk-rock trace element compositions typically show flattened patterns in the MREE–HREE range, indicating little or no contribution of primary garnet to their whole-rock compositions. However, internal Lu–Hf mineral isochrons yield a range of ages older than the Sm–Nd isochron ages (Fig. 11b), suggesting the incomplete resetting of Lu–Hf mineral isochrons during cooling. Le Roux et al. (2016) suggested that Lu/Hf ratios are more uncertain as they were probably affected by subsolidus re-equilibration even in clinopyroxene, as opposed to Nd and Sm, for which subsolidus re-equilibration is barely detectable in clinopyroxene. This may reflect d