Mantle heterogeneities in rifting-related and supra-subduction settings: examples from External Ligurian and New Caledonia ophiolites

1 UNIVERSITA’ DEGLI STUDI DI PARMA DOTTORATO DI RICERCA IN SCIENZE DELLA TERRA CICLO XXXIV MANTLE HETEROGENEITIES IN RIFTING-RELATED AND SUPRA-SUBDUCTION SETTINGS: EXAMPLES FROM EXTERNAL LIGURIAN AND NEW CALEDONIA OPHIOLITES Coordinatore: Chiar.mo Prof. Marco Roveri Tutore: Chiar.ma Prof.ssa Alessandra Montanini Co-Tutore: Chiar.mo Prof. Riccardo Tribuzio Dottorando: Elisa Ferrari Anni Accademici 2018/2019 – 2020/2021 2 3 Index Chapter 1 6 1.1. Introduction: contents and aims of the Thesis 6 1.2. References 7 Chapter 2-Rifting evolution of the lithospheric subcontinental mantle: New insights from the External Ligurian ophiolites (Northern Apennine, Italy) 8 2.1. Introduction 8 2.2. Geological setting 9 2.3. Field relationships 12 2.3.1. Monte Gavi mantle section 12 2.3.2. Monte Sant’Agostino mantle section 13 2.4. Petrographic characteristics and major element mineral chemistry 15 2.4.1. Monte Gavi peridotites 15 2.4.2. Monte Gavi Type-I pyroxenites 15 2.4.3. Monte Gavi Type-II pyroxenites 16 2.4.4. Monte Sant’Agostino deformed peridotites 17 2.4.5. Monte Sant’Agostino spinel-plagioclase websterites 18 2.5. Geothermobarometric estimates 24 2.5.1. Monte Gavi mantle section 24 2.5.2. Monte Sant’Agostino mantle section 25 2.6. Discussion 25 2.6.1. Monte Gavi: a mantle section recording reactive melt infiltration in the plagioclase stability field 25 2.6.2. Monte Sant'Agostino: a deformed mantle section 27 2.6.3. The diverse nature of pyroxenites from the External Ligurian mantle 31 sections 2.6.4. The heterogeneity of the External Ligurian subcontinental mantle 33 2.6.5. Constraints on the mantle evolution in the Western Tethys 35 ocean-continent transition 2.7. Conclusions 39 2.8. References 40 4 2.9. Tables 48 Chapter 3-Insights on the Monte Gavi plagioclase-facies melt-rock reaction event 56 3.1. Introduction 56 3.2. Geological and petrological framework 57 3.3. Methodology 58 3.4. Whole rock compositions 59 3.5. Mineral trace element compositions 63 3.6. Geothermometric estimates 64 3.7. Nd-Hf isotopic compositions 70 3.8. Discussion 73 3.8.1. Geochemical response to plagioclase-facies melt infiltration 73 3.8.2. Thermal evolution of the Monte Gavi sequence 74 3.8.3. Isotopic constraints on the melt infiltration event 75 3.8.3.1. The Upper Triassic melt impregnation event 76 3.9. Conclusions 77 3.10. References 79 3.11. Tables 84 Chapter 4-Ouassé (New Caledonia) pyroxenites witness mantle heterogeneity in young subduction systems 92 4.1. Introduction 92 4.2. Geological and petrological setting 92 4.3. Field relationships and sampling 96 4.4. Methodology 97 4.5. Petrography and mineralogy 100 4.5.1. Harzburgites 100 4.5.2. Orthopyroxenites 101 4.5.3. Amphibole-bearing websterites 101 4.6. Whole rock compositions 108 4.6.1. Major elements 108 4.6.2. Trace elements 108 4.7. Mineral trace elements composition 109 5 4.7.1. Orthopyroxenites 109 4.7.2. Amphibole-bearing websterites 113 4.8. Geothermometry 117 4.9. 40Ar/39Ar amphibole dating 118 4.10. Discussion 118 4.10.1. Nature and evolution of hosting peridotites 118 4.10.2. Origin of the pyroxenite layers 120 4.10.2.1. Orthopyroxenites 120 4.10.2.2. Amphibole-bearing websterites 122 4.10.3. Insights on the New Caledonia Paleocene-Eocene subduction initiation 124 4.10.4. Evolution of the percolating melts/fluids in the forearc mantle 127 4.10.5. Hints for reinterpreting the significance of the Bogota Peninsula Shear Zone 128 4.11. Conclusions 129 4.12. References 129 4.13. Tables 138 Chapter 5 149 5.1. Concluding remarks 149 6 Chapter 1 1.1. Introduction: contents and aims of the Thesis The notion of mantle heterogeneity has become entrenched in the scientific community since the pioneering studies in the 1960s, which evidenced geochemical and isotopic variabilities in the oceanic basalts, one of the main products of mantle melting (Hedge and Walthall, 1963; Gast et al., 1964; Tatsumoto et al., 1965; Schilling and Winchester, 1967, 1969; Gast, 1968; Hart, 1971). Studying Mid-Ocean Ridge Basalts (MORB) and Ocean Island Basalts (OIB) has always been a widely used method to characterize mantle sources, although it is an indirect investigation approach. A direct approach is represented by sampling mantle rocks from the ocean floor, by dredging or drilling, and ophiolitic exposures, besides analysing mantle xenoliths. Mantle heterogeneities are generally explained by the presence of pyroxenite and eclogite layers, which, to a large extent, could derive from processes of crustal recycling into the mantle (i.e., subduction and lithospheric delamination). The guiding theme of this PhD Thesis is the creation of geochemical and isotopic heterogeneities in mantle sequences from distinct geodynamic settings, in particular from rifting-related and young supra-subduction settings. Specific petrogenetic processes act in these different environments: while lithospheric extension and asthenosphere upwelling dominate the rifting-related setting, production of melts from variably depleted mantle sources modified by slab-derived components are believed to characterize the subduction onset. Two distinct ophiolitic exposures are considered in the present study: the External Ligurian ophiolites, chosen as an example of rifting-related mantle section and the New Caledonia ophiolite, chosen as an example of supra-subduction zone mantle section. Chapter 2 and Chapter 3 illustrate the External Ligurian case study. In particular, the main objectives of Chapter 2 are: (i) to describe the petrographic, petrological and geochemical characteristics of two poorly known mantle bodies (Monte Gavi and Monte Sant’Agostino), and (ii) to provide new insights into the behaviour of the External Ligurian subcontinental mantle during the Mesozoic rifting phase. The contents of Chapter 2 represent a paper already published: Ferrari E., Montanini A., Tribuzio R., 2022. Rifting evolution of the lithospheric subcontinental mantle: New insights from the External Ligurian ophiolites (Northern Apennine, Italy). Lithos, 410-411: 106571. The article is reported unchanged in the Thesis, except for the layout. Chapter 3 focuses on: (i) the melt-rock reaction process recorded by Monte Gavi mantle body, and (ii) the origin of the impregnating melt. 7 Chapter 4 illustrates the New Caledonia case study. The main purpose is to define the compositions of the melts percolating the forearc mantle during the subduction initiation. 1.2. References Gast P.W., 1968. Trace element fractionation and the origin of tholeiitic and alkaline magma types. Geochimica et Cosmochimica Acta, 32:1057-1086 Gast P.W., Tilton G.R., Hedge C., 1964. Isotopic composition of lead and strontium from Ascension and Gough Islands. Science, 145:1181-1185 Hart S.R., 1971. K, Rb, Cs, Sr, Ba contents and Sr isotope ratios of ocean floor basalts. Philosophical Transactions of the Royal Society Series A, 268:573-587 Hedge C.E., Walthall F.G., 1963. Radiogenic strontium 87 as an index of geological processes. Science, 140:1214-1217 Schilling J.G., Winchester J.W., 1967. Rare-earth fractionation and magmatic processes. Mantles of Earth and Terrestrial Planets, 267-283, Runcon SK, London: Interscience Publishers Schilling J.G., Winchester J.W., 1969. Rare earth contribution to the origin of Hawaiian lavas. Contributions to Mineralogy and Petrology, 40:231 Tatsumoto M., Hedge C.E., Engel A.E.J., 1965. Potassium, rubidium, strontium, thorium, uranium, and the ratio of strontium-87 to strontium-86 in oceanic tholeiitic basalt. Science, 150:886-888 8 Chapter 2 Rifting evolution of the lithospheric subcontinental mantle: New insights from the External Ligurian ophiolites (Northern Apennine, Italy) 2.1. Introduction Magma-poor ocean-continent transitions (OCT), such as those exposed along the IberiaNewfoundland (e.g., Whitmarsh et al., 2001) and the Australian-Antarctic systems (e.g., McCarthy et al., 2020), are characterized by wide areas of subcontinental mantle uplifted and denuded in conjunction with the late-stage rifting evolution. A key point to elucidate the rifting process leading to basin opening is therefore the tectonic, magmatic and metamorphic evolution of the subcontinental lithospheric mantle. In particular, limited direct information are available on the high temperature shearing and mode of mantle exhumation before the late coupling stage, where brittle faults cut through the extremely thinned continental crust and penetrate into the underlying mantle (Sutra and Manatschal, 2012). Samples drilled or dredged along modern OCTs testify the occurrence of heterogeneous subcontinental lithospheric mantle, which preserves geochemical signatures acquired before onset of rifting, as well as evidence for syn-rift melt-rock interaction and deformation (e.g., Chazot et al., 2005; Müntener and Manatschal, 2006). However, the relatively limited sampling compared with the regional extent of the involved areas, and the diffuse alteration of the recovered samples hampered to envisage a conceptual model for the mantle behaviour from the beginning of lithospheric thinning to crustal breakup. Fragments of magma-poor ocean-continent transitions of Jurassic age are exposed along the Alpine-Apennine belt (e.g., Manatschal and Müntener, 2009). These fossil records provided relevant information about the rifting-related history undergone by the subcontinental mantle during lithosphere thinning and exhumation (e.g., Picazo et al., 2016). For instance, the investigation of the Platta ophiolites from Central Alps documented that the subcontinental mantle was percolated and refertilized by MORB-type melts during rifting (e.g., Müntener et al., 2010). In addition, the central body of the Lanzo mantle massif from Western Alps provides evidence for melt focusing within rifting-related plagioclase-facies shear zones (Kaczmarek and Müntener, 2010; Kaczmarek and 9 Tommasi, 2011). The subcontinental mantle bodies from the External Ligurian units (Northern Apennines, Fig. 1) are also part of a Jurassic magma-poor ocean-continent transition (Marroni et al., 1998). Despite some of these mantle bodies were thoroughly investigated from petrological, geochemical and microstructural viewpoints (e.g., Borghini et al., 2016; Hidas et al., 2020; Montanini et al., 2006; Rampone et al., 1995), a comprehensive rifting-related geological scenario is lacking for the External Ligurian mantle section. The primary aim of this contribution is to provide new constraints on the evolution of the subcontinental mantle involved in a rifting process. We thus present new field, microstructural, mineral chemistry and geothermobarometric data obtained for peridotites and enclosed pyroxenites from kilometre-sized subcontinental mantle bodies from the External Ligurian units, locally recognized as Monte Gavi and Monte Sant'Agostino. To our knowledge, no petrographic and petrological data are available in the literature for these mantle sections. We document that the different mantle bodies from the External Ligurian units display distinct tectonic and thermal evolutions in the plagioclase stability field. Hence, we relate the heterogeneous behaviour of the subcontinental mantle under relatively low pressure conditions to the rifting process that ultimately developed a magma-poor Jurassic OCT. A comparison with ophiolitic mantle sections from the Alpine belt finally allowed us to shed light on the rifting-related evolution of the subcontinental lithospheric mantle at a regional scale. 2.2. Geological setting The ophiolitic bodies exposed along the Alpine-Apennine belt are lithospheric remnants of the Western Tethys (or Ligurian-Piedmontese basin). This basin formed in the Middle to Upper Jurassic in conjunction with the opening of the Central Atlantic Ocean and separated the Europe-Iberia from the Africa-Adria plate (e.g. Schettino and Turco, 2011). The opening of Western Tethys basin was associated with uplift and denudation of subcontinental lithospheric mantle, which is presently exposed along the Alpine-Apennine belt. During rifting and oceanization, the subcontinental lithospheric mantle most likely underwent thermochemical erosion driven by reaction with ascending melts of asthenospheric origin (e.g., Müntener et al., 2010; Piccardo et al., 2004, 2007; Rampone et al., 2020). The Alpine-Apennine ophiolites are locally associated with continental crust material and frequently include mantle sections retaining a subcontinental origin (e.g., Manatschal and Müntener, 2009; Marroni et al., 1998). The best examples of preserved subcontinental mantle were reported for the Malenco-Totalp bodies from Central Alps (Müntener et al., 2004, 2010) and for the External Ligurian units from the Northern Apennine (e.g., Montanini et al., 2006, 2012; Rampone et al., 1995). 16 homogeneous TiO2 (0.7–0.9 wt%). The coarse green spinels (Cr# = 2–4, TiO2 < 0.10 wt%) are rimmed by 1–2 mm thick plagioclase coronas (Fig. 2.5c). The selected sample also includes smaller, submillimetre to millimetre-scale spinel grains with dark brown colour and embayed shape, rimmed by up to 1 cm wide aggregates made up of altered plagioclase and minor olivine (Fig. 2.4a). These spinels have variable and significantly higher Cr# (12–34) and TiO2 (0.15–1.02 wt%). Notably, minor amounts of relatively large chlorite grains displaying prismatic shape and ghost cleavages, presumably representing pseudomorphs after orthopyroxene, are also present. Hence, we envision that the original mineral assemblage of these pyroxenites was mostly composed of clinopyroxene, associated with minor amounts of Al-spinel and orthopyroxene. Type-Ib pyroxenites (samples GA1 and MG6) are characterized by the occurrence of distinct pyroxeneand plagioclase-bearing domains, which are both up to 2 cm long (Fig. 2.3d). The pyroxene-rich domains are composed of clinopyroxene (40–50 vol%), orthopyroxene (35–40 vol%) and minor plagioclase (15–20 vol%), and locally include relics of pre-existing coarse clinopyroxenes (Fig. 2.5d). Thick plagioclase + orthopyroxene lamellae are present along the cleavages of these coarse clinopyroxenes, typically forming symplectitic intergrowths (Fig. 2.5e). The coarse clinopyroxene has relatively low Mg# (79–84) and variable Al2O3 (5.2–9.7 wt%), TiO2 (1.2–2.3 wt%) and Na2O (0.4–1.7 wt%), with no clear systematic zoning (Fig. 2.8). The chemical compositions of the secondary clinopyroxenes associated with plagioclase and orthopyroxene fall within these intervals. Orthopyroxene has low Mg# (81–83), 1.9–3.7 wt% Al2O3, 0.6–1.3 wt% CaO and relatively high TiO2 (0.3–0.5 wt%). Accessory brown amphibole (titanian pargasite according to the nomenclature of Leake et al., 1997) occurs as thin rims mantling resorbed clinopyroxenes and has 76–80 Mg# and 2.7–3.8 wt% TiO2. The plagioclase-rich domains include ~30 vol% olivine and, in their inner portions, accessory amounts of small spinel grains with brownish-black colour (Fig. 2.5f). Plagioclase (71–79 mol% anorthite) and olivine (82–83 mol% forsterite) form nearly polygonal aggregates and are variably altered into sericite and serpentine, respectively. Spinel within the plagioclase-rich domain has variable Cr# (4–30) and TiO2 (0.10–0.56 wt %, Fig. 2.7). Tiny ilmenite and altered Fe-Ni-Cu sulfides with rounded morphology are also in places present within plagioclase. Towards the contact with the pyroxene-rich domains, coarse-grained vermicular olivine + plagioclase intergrowths typically occur. 2.4.3. Monte Gavi Type-II pyroxenites Type-II pyroxenites are mediumto coarse-grained olivine websterites. They display moderate to extensive serpentinization, which makes in places difficult to decipher their primary texture. However, in the most preserved samples, an allotriomorphic granular texture is discernible. Their 17 original mineral assemblage is composed of clinopyroxene (30–40 vol%), orthopyroxene (25–30 vol%), olivine (up to 25 vol%), greenish to brown spinel (5–10 vol%) and plagioclase (10–15 vol %). These pyroxenites are characterized by the presence of exsolved orthopyroxene porphyroclasts, forsterite-rich (89 mol%) olivine (Fig. 2.5g) that is locally kinked, and clinopyroxene-rich domains. These domains consist of irregularly shaped clinopyroxene grains with plagioclase + orthopyroxene lamellae (Fig. 2.5h), associated with interstitial plagioclase and trails of elongated spinels rimmed by plagioclase. Clinopyroxene from Type-II pyroxenites has 88–90 Mg#, relatively high Cr2O3 (0.4–0.9 wt%) and a wide TiO2 range (0.7–1.8 wt%; Figs. 2.8 and 2.9). Al2O3 and Na2O in clinopyroxene vary from 4.5 to 7.4 wt% and from 0.5 to 0.9 wt%, respectively. Orthopyroxene displays high Mg# (87–89), 4.0–5.8 wt% Al2O3 and 0.6–1.0 wt% CaO. Spinel has variable Cr# (6–16) and 0.1–0.6 wt% TiO2 (Fig. 2.7). Plagioclase is in most places replaced by sericite, but small preserved domains with ~65 mol% anorthite were found in sample IC2/3. Accessory brown amphibole (titanian pargasite) occurs as rim around pyroxenes and is characterized by 85–87 Mg# and 3.1–4.0 wt% TiO2. 2.4.4. Monte Sant’Agostino deformed peridotites The Monte Sant'Agostino peridotites mostly consist of plagioclase-facies mylonitic lherzolites including small (cmto dm-sized) domains made up of spinel-facies tectonites. Microstructural evidence for a spinel-facies recrystallization event are preserved in these low-strain domains (Fig. 2.10a), where up to 1 cm in size porphyroclasts of clinopyroxene and orthopyroxene are surrounded by fine-grained granoblastic aggregates made up of pyroxenes + light brown spinel ± orange-brown amphibole. Mutual exsolutions are displayed by the orthopyroxene and the clinopyroxene porphyroclasts (Fig. 2.10a-b), and the overall clinopyroxene mode varies between 10 and 15 vol%. In the less serpentinized samples, mm-sized olivine porphyroclasts surrounded by neoblastic olivine are still recognizable. Holly leaf brown spinels are also locally preserved. The spinel-facies assemblage is widely overprinted by a plagioclase-facies mylonitic fabric. The mylonite microstructure is characterized by porphyroclasts of orthopyroxene and clinopyroxene aligned along the foliation (Fig. 2.10b–c). Cr-spinel and rare brown amphibole may also occur as porphyroclasts. Porphyroclastic orthopyroxene typically displays up to ~10 aspect ratio (Fig. 2.10c). Porphyroclastic clinopyroxene is smaller and more equant than associated porphyroclastic orthopyroxene. Both pyroxene porphyroclasts provide evidence for deformation, like undulose extinction, bending and kinking, and are set into a polyphase olivine-rich matrix with extremely fine grain-size (20–50 μm, Fig. 2.10b-c-d). Besides olivine, this matrix includes neoblastic clinopyroxene and orthopyroxene, plagioclase and, in places, Cr-spinel. Plagioclase also occurs as thin rims mantling 18 the spinel porphyroclasts. The extremely fine-grained matrix is dominant in the ultramylonites (Fig. 2.10e), which contain sporadic pyroxene and brown amphibole porphyroclasts with rounded shape (Table 2.3). Both porphyroclastic and matrix olivine have 89–90 mol% forsterite and 0.4–0.5 wt% NiO. Porphyroclastic clinopyroxene has 90–91 Mg# and high Al2O3 and Na2O contents (7.0–8.4 wt% and 1.6–2.1 wt%, respectively), slightly decreasing from the core to the rim (Fig. 6). Clinopyroxene neoblasts have higher Mg# values (91–92) and lower Al2O3 and Na2O (2.3–3.2 and ~0.6 wt%, respectively) than porphyroclasts (see also Fig. 2.6). Orthopyroxene porphyroclasts have 89–90 Mg#, 4.4–5.6 wt% Al2O3 and 0.7–0.9 wt% CaO. The tiny orthopyroxene from the mylonitic matrix is distinct in the low Al2O3 (0.5–2.4 wt%) and CaO (0.25–0.36 wt%). Spinel has low Cr# (9–16) and TiO2 ≤ 0.20 wt% (Fig. 2.7). Plagioclase in the mylonitic matrix has 38–50 mol% anorthite. Porphyroclastic amphibole is titanian pargasite, with 84–86 Mg# and 3.1–4.6 wt% TiO2. Accessory amphibole with relatively low TiO2 (~1.8 wt%, Table 2.3) is also present in the matrix of the ultramylonites. 2.4.5. Monte Sant’Agostino spinel-plagioclase websterites These pyroxenites typically have protomylonitic texture (Fig. 2.10f-g). Porphyroclasts of exsolved clinopyroxene and orthopyroxene (up to 0.5 cm in size), and of Al-spinel are set into a finegrained neoblastic matrix displaying an average grain size of ~0.1 mm (Fig. 2.10f-g). Comparable amounts of clinopyroxene and orthopyroxene porphyroclasts are generally observed. The clinopyroxene porphyroclasts (Fig. 2.10g) locally display thin exsolution lamellae of orthopyroxene ± plagioclase, and the orthopyroxene porphyroclasts are frequently stretched along the main foliation. The neoblastic mineral assemblage consists of clinopyroxene and orthopyroxene associated with minor to accessory amounts of green spinel, plagioclase, olivine and brown amphibole. The neoblastic minerals commonly display grain boundary alignment parallel to the mylonitic foliation. Websterite MP7 also includes rounded symplectitic domains (Fig. 2.10h) made up of vermicular Crpoor spinel (Cr# ~1), Al-poor orthopyroxene (Al2O3 = 2.8–3.9 wt%) and minor plagioclase (52–54 mol% anorthite). Taken as a whole, clinopyroxenes and orthopyroxenes have 84–88 Mg# and 83–88 Mg#, respectively (Table 2.4). Clinopyroxene porphyroclast cores have high amounts of Al2O3 (9–10 wt%) and, accordingly, high proportions of Ca-Tschermak substitution (up to ~15 mol%), which are coupled with 0.8–1.1 wt% Na2O (Fig. 8). Orthopyroxene is also Al2O3-rich (4.9–6.8 wt%), with rather uniform CaO (0.5–0.7 wt%). For both pyroxenes, a decrease of Al2O3 is observed in porphyroclast rims and in neoblastic grains with respect to porphyroclastic cores. Green spinel has generally low 19 Cr# (< 0.15 wt% TiO2, Fig. 2.7). Olivine and plagioclase in the mylonitic matrix have 80–88 mol% forsterite and 54–55 mol% anorthite. Brown amphibole is titanian pargasite to kaersutite with 79–83 Mg# and 3.2–5.4 wt% TiO2. Fig. 2.4. Thin section (4.6 mm × 2.6 mm) images of: (a) Type-Ia pyroxenite (sample GAV2), (b) protomylonitic lherzolite (sample IC12–3), (c) mylonitic lherzolite (sample MGO1). 20 Fig. 2.5. Thin section photomicrographs under cross-polarized (a-b-d-e-f-g-h) and plane-polarized light (c) of Monte Gavi samples. (a) embayed clinopyroxene partially replaced by orthopyroxene + (altered) plagioclase, peridotite MGA7; (b) orthopyroxene + (sericitized) plagioclase replacing clinopyroxene, Type-Ia pyroxenite GAV2; (c) Al-spinel rimmed by plagioclase, Type-Ia pyroxenite GAV2; (d) pyroxene-rich domain, Type-Ib pyroxenite MG6; (e) symplectitic orthopyroxeneplagioclase intergrowth in clinopyroxene, Type-Ib pyroxenite MG6; (f) plagioclase-rich domain, Type-Ib pyroxenite GA1; (g) Mg-rich olivine, Type-II pyroxenite IC2–3; (h) Orthopyroxene and plagioclase coarse lamellae in clinopyroxene, Type-II pyroxenite IC2–3. 21 Fig. 2.6. Variation of Na2O vs. Al2O3 for clinopyroxenes from Monte Gavi and Monte Sant'Agostino peridotites; pfc = porphyroclast, neo = neoblast. Data sources for other External Ligurian peridotites: Borghini et al., 2011, Borghini et al., 2013; Montanini et al. (2006); Rampone et al. (1995). North Lanzo data from Kaczmarek et al. (2008); Upper Platta, Malenco and Totalp data from Müntener et al. (2010). Fig. 2.7. Variation of Cr # (100 ⋅[Cr/(Cr + Al)]) vs. TiO2 for spinel from Monte Gavi and Monte Sant'Agostino peridotites and pyroxenites. Data sources for other External Ligurian peridotites and pyroxenites: Borghini et al. (2011, 2013, 2016); Montanini et al. (2006); Rampone et al. (1995). Data for North Lanzo peridotites from Kaczmarek and Müntener (2008); Upper Platta, Lower Platta, Malenco and Totalp peridotites after Müntener et al. (2010). Data for Monte Maggiore pyroxenites (Corsica) from Basch et al. (2019). 22 Fig. 2.8. Variation of Na2O and TiO2 vs. Al2O3 for clinopyroxenes from Monte Gavi and Monte Sant'Agostino pyroxenites; pfc = porphyroclast, neo = neoblast, 1 = cpx in primary garnet-bearing assemblage, 2 = cpx in retrograde, garnet-free assemblage. Data sources for other External Ligurian pyroxenites: Borghini et al. (2016), Montanini et al. (2006). Data for Monte Maggiore pyroxenites (Corsica) from Basch et al. (2019). Fig. 2.9. Variation of Cr2O3 vs. Mg# (100 ⋅[Mg/(Mg + Fetot)]) for clinopyroxenes from Monte Gavi and Monte Sant'Agostino peridotites and pyroxenites; pfc = porphyroclast. 23 Fig. 2.10. Thin section photomicrographs under cross-polarized light (a-b-c-e-f-g), plane-polarized light (h) and BSE (back-scattered electron) image (d) of Monte Sant'Agostino samples. (a) Preserved porphyroclastic domain in peridotite with deformed orthopyroxene mantled by fine-grained neoblastic aggregates of orthopyroxene + clinopyroxene + spinel (protomylonite IC12–3); (b-c) Exsolved clinoand orthopyroxene porphyroclasts in polyphase plagioclase-bearing mylonitic matrix (lherzolite IC8–1); (d) BSE image of polyphase plagioclase-bearing mylonitic matrix (lherzolite sample IC 8–3); (e) peridotite ultramylonite with rounded pyroxene porphyroclasts cut by serpentine veins (sample IC11); (f) Orthopyroxene porphyroclasts in websterite with protomylonitic texture (sample AM489); (g) Exsolved clinopyroxene porphyroclast in websterite with protomylonitic texture (sample MP7); (h) Symplectitic orthopyroxene-spinel-plagioclase domain after garnet (sample MP7). 24 2.5. Geothermobarometric estimates Temperature evaluations for the different equilibration stages recorded by the Monte Gavi and Monte Sant'Agostino peridotites and enclosed pyroxenite layers were obtained using conventional pyroxene geothermometry (Table 2.5). In particular, the two-pyroxene geothermometer (Brey and Kohler, 1990 and Taylor, 1998, referred to as TBK90 and TTa98, respectively) and Ca-in-orthopyroxene geothermometer (Brey and Kohler, 1990, referred to as TCa-in-Opx) were applied. For the Monte Gavi Type-I pyroxenites and the Monte Sant'Agostino websterites, we also calculated the equilibration temperatures (THB94) based on the amphibole-plagioclase method of Holland and Blundy (1994). A valuable tool applicable to the plagioclase-bearing peridotites of the present study is the geobarometer of Fumagalli et al. (2017), which relies on the pressure-sensitive equilibrium Forsterite + Anorthite = Ca-Tschermak + Enstatite (FACE) and has a standard error of 0.05 GPa. Notably, the application of the FACE geobarometer to pyroxenites has been recently evaluated by Basch et al. (2020), who obtained slightly higher pressure estimates for websterites than for enclosing peridotites. In particular, a systematic pressure difference of ~0.1 GPa was related to the lower Cr# of the websterites compared with the peridotites, which led to higher Ca-Tschermak activity in the websterite clinopyroxene. Considering the uncertainties reported by Basch et al. (2020), we applied the FACE geobarometer to not only the peridotites but also the enclosed pyroxenites (Table 2.5). 2.5.1. Monte Gavi mantle section Temperature estimates for Monte Gavi samples were calculated for: (i) Large orthopyroxene-clinopyroxene grains from peridotite MGA7 and Type-II pyroxenite IC2/3. The cores of exsolved orthopyroxene porphyroclasts yielded TCa-inOpx of 950–1110 °C, assuming a confining pressure of 1.5 GPa. Temperature estimates obtained from two-pyroxene geothermometry for the peridotite MGA7 are shifted to lower values (897–938 °C). (ii) Non-exsolved orthopyroxenes formed in conjunction with plagioclase in Type-I and Type-II pyroxenites. The obtained temperatures (TCa-in-Opx) fall in a wide range (910– 1065 °C), regardless of different textural occurrences of orthopyroxene. The temperatures intervals obtained from adjacent clinopyroxene-orthopyroxene pairs (TTa98 = 845–983 °C, TBK90 = 876–1015 °C) are slightly lower than TCa-in-Opx. However, the average TCa-in-Opx (974 ± 47 °C) are within error of the average TTa98 (913 ± 52 °C) and TBK90 (937 ± 47 °C). 25 (iii) Coexisting amphibole and plagioclase in Type-I pyroxenite GA1, which gave temperature estimates of 1030 °C using the method of Holland and Blundy (1994). Pressure estimates by FACE geobarometer (Fumagalli et al., 2017) were only computed for the Monte Gavi pyroxenites, as the enclosing peridotites lack preserved plagioclase. We used the same clinopyroxene-orthopyroxene pairs employed for two-pyroxene geothermometric calculations and obtained pressure values ranging from 0.67 to 0.78 GPa. 2.5.2 Monte Sant’Agostino mantle section Exsolved porphyroclast cores from mylonitic lherzolites gave relatively low clinopyroxeneorthopyroxene (TTa98 = 832–886 °C, TBK90 = 857–950 °C) and Ca-in-Opx temperature estimates (930–970 °C). Temperature evaluations falling in the same intervals or slightly shifted towards higher values were obtained for the websterite porphyroclastic assemblage (TTa98 = 914–958 °C, TBK90 = 940–960 °C and TCa-in-Opx = 919–985 °C). The extremely fine-grained pyroxenes from the matrix of mylonitic and ultramylonitic lherzolites yielded TTa98 and TBK90 of 756–873 °C and TCa-in-Opx of 750–780 °C. In addition, an amphiboleplagioclase pair from ultramylonitic lherzolite IC11 gave a value of 778 °C based on the Holland and Blundy (1994) method. In the websterites, we obtained: (i) TTa98 = 856–880 °C, TBK90 = 905–938 °C and TCa-in-Opx = 870–930 °C from the neoblastic pyroxenes associated with plagioclase, and (ii) 880– 907 °C based on the plagioclase-amphibole method by Holland and Blundy (1994). Application of the FACE geobarometer (Fumagalli et al., 2017) indicate shallow conditions for the development of plagioclase-bearing assemblage in the mylonitic lherzolites. Pressure values of 0.54–0.55 GPa and 0.34 GPa were obtained from lherzolites IC8–1 and IC8–3 and IC11, respectively. The websterites gave pressure stimates of 0.77–0.90 GPa. 2.6. Discussion 2.6.1. Monte Gavi: a mantle section recording reactive melt infiltration in the plagioclase stability field The Monte Gavi harzburgite-pyroxenite association is nearly undeformed and characterized by extensive crystallization of plagioclase ± orthopyroxene at the expense of spinel and clinopyroxene. In both the harzburgites and the enclosed pyroxenites, plagioclase not only occurs as coronas around spinel but also within large clinopyroxene grains as patches or thick lamellae associated with orthopyroxene (Fig. 2.5a-b-h), locally forming symplectitic intergrowths (Fig. 2.5e). An exsolution origin of this plagioclase-orthopyroxene association is unlikely, given the high proportions of these 32 bodies (Montanini et al., 2006). The Monte Sant'Agostino websterites also have highly aluminous clinopyroxene porphyroclasts with plagioclase + orthopyroxene exsolutions, similar to those derived from retrogression of primary Na-Al-rich clinopyroxene originally in equilibrium with garnet. In addition, both the Monte Sant'Agostino and the Rio Strega-Monte Prinzera websterites (see also Montanini and Tribuzio, 2015) are characterized by similar two-pyroxene equilibration temperatures obtained for the porphyroclastic spinel-facies assemblage (i.e., TBK90 of 940–960 °C and 900–980 °C, respectively), and protomylonitic textures developed in the plagioclase stability field. The Monte Gavi pyroxenites do not display high temperature deformation microstructures and provide evidence for pervasive melt infiltration under plagioclase-facies conditions, contrary to what was observed for the other External Ligurian mantle bodies. On the other hand, the presence of pyroxenites with distinct thickness and Mg#, as suggested by the clinopyroxene chemistry (79–87 and 88–90 in Type-I and Type-II pyroxenites, respectively; Fig. 2.9), was also recognized for the Rio Strega-Monte Prinzera lherzolite-pyroxenite section (Montanini et al., 2006). The primary modal composition of Type-I (thick, Fe-rich) pyroxenites is difficult to establish, due to the extensive transformations related to the process of melt-rock reaction. We speculate that the Type-I pyroxenite protoliths were coarse-grained spinel-rich orthopyroxene-poor websterites. In addition, the low Mg# of clinopyroxene suggests crystallization in a melt-dominated system, with a negligible mass contribution from enclosing lherzolites. Type-II (thin, Mg-rich) pyroxenites were spinel-bearing olivine websterites. The presence of deformed forsterite-rich olivine and Mg-rich exsolved orthopyroxene in Type-II pyroxenites, similar to those occurring in the host peridotites, documents formation by melt/peridotite hybridization, presumably through a process leading to olivine consumption and concomitant crystallization of Mg-rich clinopyroxene + spinel. Contrary to the pyroxenites from the Rio Strega-Monte Prinzera mantle bodies, no paragenetic or textural evidence for the presence of an earlier garnet-bearing assemblage was observed for the Monte Gavi mantle pyroxenites. Assuming that the Mg# value of the clinopyroxene cores from Type-I pyroxenites was not significantly modified by interaction with the infiltrating melts, and no Fe–Mg subsolidus exchange with associated mafic minerals occurred, we calculated the Mg# of the clinopyroxene equilibrium melts following the equation of Wood and Blundy (1997). Computed melt Mg# (49–64) are substantially lower than the Mg# of melts in equilibrium with a peridotite source. The high Al contents observed in some clinopyroxene cores of Type-I pyroxenites and the high modal proportion of Al-rich spinel (Fig. 2.8) also indicate that the pyroxenite-forming melts were characterized by low Si/Al ratios (Della-Pasqua et al., 1995). We speculate that these melts formed by partial melting of precursor mafic/pyroxenite lithology or of a mixed pyroxenite/peridotite source. 33 In summary, the Monte Sant'Agostino websterites are structurally and compositionally similar to the websterites, in places retaining garnet relics, from the Rio Strega-Monte Prinzera mantle bodies (Montanini et al., 2006). The Monte Gavi pyroxenites have no apparent analogues in the other mantle bodies so far described for the External Ligurian ophiolites, although new geochemical data are needed to establish their origin. 2.6.4. The heterogeneity of the External Ligurian subcontinental mantle The mantle bodies from the External Ligurian ophiolites represent different domains of a largescale subcontinental mantle section that was variably affected by retrogression, deformation and chemical modifications during lithospheric extension and concomitant asthenospheric ascent in response to the rifting process that led to opening of the Jurassic Western Tethys (Hidas et al., 2020; Montanini et al., 2006; Piccardo et al., 2004). Based on the main characteristics of investigated External Ligurian mantle bodies, we subdivide this subcontinental mantle section into three main domain types. 1Plagioclase-bearing spinel tectonite domain (ST), represented by the Suvero mantle body (Fig. 2.1). The Suvero lherzolites and the enclosed pyroxenites record a deformation event developed under spinel-facies and high temperature conditions (1000–1050 °C), which yielded porphyroclastic textures associated with a penetrative tectonic foliation in lherzolites (Borghini et al., 2011; Rampone et al., 1995). The Suvero lherzolites underwent melt infiltration forming pyroxenitic heterogeneities during the early Paleozoic (Borghini et al., 2013, 2016). Melt migration and pyroxenite formation was synto late-kinematic with respect to the spinel-facies deformation and produced a localized metasomatic imprint in the host lherzolites (see also Borghini et al., 2021; Hidas et al., 2020). This event was inferred to have occurred under relatively high pressure conditions (>1.5 GPa) based on the geochemical evidence for a primary garnet-bearing assemblage in the pyroxenites, which was completely replaced during the subsequent decompression evolution. The Suvero mantle body provides evidence for a plagioclase-facies re-equilibration under static conditions, in response to a relatively cold decompression (Hidas et al., 2020). In particular, two recrystallization stages were recognized, from 0.6–0.7 GPa and 890–910 °C to 0.4–0.5 GPa and 800–840 °C (Fumagalli et al., 2017, see Fig. 2.11). Internal Sm-Nd isochrons (clinopyroxene-plagioclase-bulk rock) obtained from four pyroxenites gave an average age of 178 ± 8 Ma for the plagioclase-facies recrystallization (Borghini et al., 2016). 2Plagioclase mylonite (PM) domain, including the Monte Sant'Agostino (this study) and the Rio Strega-Monte Prinzera (Fig. 2.1) mantle sections. In the lherzolites from these mantle bodies, a tectonitic texture developed under spinel-facies conditions was widely overprinted by a dynamic 34 plagioclase-facies recrystallization associated with development of a pervasive mylonitic foliation. Similar to the lherzolites, the enclosed pyroxenites record different extents of spineland plagioclasefacies deformation. Using the mineral chemistry data reported in Montanini et al. (2006), we obtained pressure and temperature estimates of 0.4–0.5 GPa and 850–870 °C (TCa-in-Opx) for the plagioclasefacies mylonitic deformation of the Rio Strega-Monte Prinzera lherzolites (Table 2.5). These pressure-temperature evaluations may be reconciled with the pressure-temperature evolution recognized for the Monte Sant'Agostino mantle section (Fig. 2.11), assuming that the ~0.5 GPa mantle uplift in the plagioclase stability field was associated with a polyphase deformation (Fig. 2.11). Radiometric constraints on the age of the plagioclase-facies shearing affecting the Rio Strega-Monte Prinzera mantle bodies were placed by a Lu–Hf mineral isochron of 220 ± 13 Ma obtained from a garnet pyroxenite (Montanini et al., 2006). This isochron was interpreted as a cooling age, implying that the garnet pyroxenite passed through the Lu–Hf blocking temperature of 800–900 °C (i.e., under plagioclase facies conditions) at ~220 Ma. Accordingly, we relate the polyphase plagioclase-facies deformation recorded by the PM mantle domain to large-scale lithospheric shear zones developed in the late Triassic. An early Jurassic Sm-Nd mineral isochron of 186 ± 2 Ma (defined by plagioclase, clinopyroxene, garnet and whole-rock) was also acquired from the Rio Strega-Monte Prinzera garnet pyroxenite, and interpreted to reflect a process of slow cooling in the plagioclase stability field (Montanini et al., 2006). 3Plagioclase-impregnated domain (PI), represented by the Monte Nero mantle body (Fig. 2.1) and the Monte Gavi mantle section (this study). The former mainly consists of weakly foliated lherzolites characterized by interstitial crystallization of plagioclase and pyroxenes, and extensive replacement of Cr-spinel by plagioclase (Rampone et al., 1995). According to Piccardo et al. (2004), the formation of plagioclase in the lherzolites was related to reaction with a pervasively migrating melt. The application of the FACE geobarometer (Fumagalli et al., 2017) to (i) plagioclase-facies granoblastic aggregates, and (ii) rims of large orthopyroxenes and clinopyroxenes, and associated plagioclase (data after Rampone et al., 1995) yielded pressures of 0.6–0.7 GPa and TCa-in-Opx of 960980 °C (Table 2.5). We also applied the geothermometer of Liang et al. (2013), based on slowly diffusing trace elements like rare earth elements and Y, to clinopyroxene-orthopyroxene pairs from the Monte Nero plagioclase-facies granoblastic assemblage (Piccardo et al., 2004). The obtained temperatures (1120–1140 °C) are significantly higher than those recorded by conventional pyroxene thermometers. Such results are consistent with impregnation by asthenospheric melts, followed by slow cooling (Dygert and Liang, 2015) associated with subsolidus recrystallization. The infiltration of asthenospheric melts during the rifting-related exhumation at Monte Nero mantle body might be related to the local occurrence of cm-scale gabbroic irregular veins (Montanini et al., 2008). Notably, 35 this mantle body also encloses up to meter-scale gabbroic intrusions and is in places crosscut by basalt dykes, with both gabbros and basalts displaying MORB-type chemical signature (Montanini et al., 2008). The Sm–Nd clinopyroxene-plagioclase two-point alignments of 163 ± 20 Ma and 165 ± 20 Ma obtained by Rampone et al. (1995) presumably record the timing of cooling after MORB impregnation associated with plagioclase formation. The Monte Nero and Monte Gavi mantle bodies share similar pressure and TCa-in-Opx estimates for the event of melt-rock reaction under plagioclasefacies conditions (Fig. 2.11), and provide no evidence for shearing in the plagioclase stability field. Contrary to the Monte Nero peridotites, however, the Monte Gavi peridotite-pyroxenite association is not intruded by gabbros nor is crosscut by basalt dikes. New geochronological constraints are therefore required to elucidate if the melt impregnation event recorded by the Monte Gavi mantle body occurred in the Jurassic or during a preceding stage (e.g., late Triassic) of the rifting evolution. Fig. 2.11 compares the plagioclase-facies pressure-temperature evolution of the mantle sections considered in this study and of the Suvero mantle body (Borghini et al., 2011; Fumagalli et al., 2017). It also reports the pressure-temperature estimates obtained in the present study for the Rio StregaMonte Prinzera and the Monte Nero peridotites. The ST and PM mantle domains document a relatively cold plagioclase-facies history characterized by a significant pressure decrease. We thus propose that these domains shared a common evolution in the late Triassic-early Jurassic. In this framework, the Suvero mantle body represents a nearly undeformed domain (ST), whereas plagioclase-facies strain localization produced the PM mantle domain characterized by extensive mylonitic deformation (Fig. 2.12). The PI mantle domain records higher plagioclase-facies thermal conditions than the ST-PM mantle domains (Fig. 2.11). We relate the distinct plagioclase-facies thermal gradients recorded by the External Ligurian mantle bodies to melt-present vs. melt-absent conditions during their decompression history in the extending lithosphere. The three different mantle domains were part of a subcontinental mantle section that was involved into the development of an ocean-continent transition, towards the Adria plate, in the Middle Jurassic (Marroni et al., 1998; Montanini et al., 2006). Tectonic dismembering during convergence-related formation of sedimentary melanges in Upper Cretaceous (Marroni et al., 2017) hampered the preservation of primary relationships among the different mantle domains. We speculate that the PI domain enclosing the MOR-type crustal rocks (i.e., the Monte Nero mantle body) was located oceanward with respect to the ST-PM domains in the Middle Jurassic. 2.6.5. Constraints on the mantle evolution in the Western Tethys ocean-continent transition 36 Here we provide a brief overview of the lithospheric mantle bodies belonging to the oceancontinent transitional domains of the Western Tethys. In particular, we compare the rifting evolution of the External Ligurian mantle section with that inferred from the investigations carried out for the subcontinental mantle bodies from Western (Lanzo massif) and Central Alps (Malenco, Platta and Totalp). Similar to the External Ligurian mantle section, these bodies were exhumed along the OCT of the Adria margin (e.g., Guarnieri et al., 2012; Kaczmarek and Müntener, 2008; Manatschal and Müntener, 2009; Müntener et al., 2004, 2010). Their evolution from the post-Varisic orogenic collapse to the Mesozoic extensional phases and basin opening were reviewed by Picazo et al. (2016), who pointed out how their pristine compositional heterogeneity was amplified during the rifting and oceanization processes. They proposed a conceptual model with two main types of lithospheric mantle sections, namely (1) “inherited” and (2) “refertilized”, exemplified by the Malenco-TotalpUpper Platta and the Lower Platta bodies, respectively. In their view, the refertilized mantle is a subcontinental mantle section located oceanwards from the inherited subcontinental mantle. The inherited mantle is dominated by fertile, amphibole-bearing spinel peridotites, typically interlayered with pyroxenites that are in places garnet-bearing (Müntener and Hermann, 1996). The inherited mantle is interpreted as old lithospheric mantle that was beneath the continental crust before onset of rifting. This interpretation is supported by the preservation of primary contacts between the Malenco mantle body and underplated gabbros of early Permian age (Hermann et al., 1997; Müntener and Hermann, 1996). In the inherited peridotites, no evidence for plagioclase formation, either of subsolidus or melt-rock reaction origin, was reported. Notably, the Malenco, Totalp and Upper Platta mantle bodies are characterized by similar results of REE-in-two-pyroxene and major element pyroxene thermometers (850–950 °C, Dygert and Liang, 2015), as expected for a slowly cooling, subcontinental lithospheric mantle. In the refertilized lithospheric mantle, which also includes the Lanzo South body (Picazo et al., 2016), the peridotites mainly are plagioclase-enriched clinopyroxene-poor lherzolites (Müntener et al., 2004, 2010; Piccardo et al., 2007). These rocks enclose: (i) dunite-harzburgite bodies of replacive origin that are up to tens of meters in scale, and (ii) cmto m-scale gabbroic veins and dykes (see also Sanfilippo et al., 2014, 2019). Pyroxene thermometry applied to spinel-facies peridotite minerals gave equilibration temperatures of 1030–1180 °C (Müntener et al., 2010), higher than those calculated for the inherited mantle. The plagioclase formation was related to infiltration of MORB-type melts associated with crystallization of new orthopyroxene after olivine and/or partial replacement of clinopyroxene by orthopyroxene and plagioclase (see also Kaczmarek and Müntener, 2008). The refertilized plagioclase-peridotites and the inherited fertile spinel-peridotites have distinct mineral compositions (Picazo et al., 2016). For instance, clinopyroxene from the refertilized peridotites has 37 lower Al2O3 and Na2O than clinopyroxene from the inherited fertile peridotites. In addition, spinel has higher Cr# and TiO2 in the refertilized peridotites than in the inherited fertile peridotites. The Lanzo massif shows a peculiar transition from a domain mostly composed of old inherited lithospheric mantle in the northern part (e.g., Guarnieri et al., 2012) to a deeply melt-modified mantle section in the southern part (Piccardo et al., 2007; Sanfilippo et al., 2014, 2019). The two domains are separated by a major shear zone (up to ~1.5 km thick), in which porphyroclastic peridotites provide evidence for plagioclase-facies melt-rock reactions, in turn overprinted by plagioclase mylonites recording solid state deformation (e.g., Kaczmarek and Müntener, 2008; Kaczmarek and Tommasi, 2011). The mylonites were interpreted as the footwall of an extensional detachment fault that accomodated exhumation of the Lanzo mantle and acted as a permeability barrier between the inherited (northern Lanzo) and refertilized (Southern Lanzo) domains. We applied the FACE geobarometer to the neoblastic plagioclase-bearing assemblage of the Lanzo mylonitic peridotites (Kaczmarek and Müntener, 2008), thereby yielding remarkably low pressure values of 0.3 GPa (Table 2.5, Fig. 2.11). These pressure estimates are coupled with TCa-in-Opx (Brey and Kohler, 1990) of 840– 870 °C, indicating a high geothermal gradient, presumably related to extreme lithospheric thinning and asthenosphere ascent. We obtained slightly higher pressure-temperature values from the porphyroclastic peridotites of 0.4 GPa and 890 °C sampled along the margins of the shear zone. Picazo et al. (2016) placed the External Ligurian mantle bodies in the refertilized lithospheric mantle, because of the widespread occurrence of plagioclase-bearing peridotites. However, only some mantle bodies from the External Ligurian units include plagioclase originated by reaction with infiltrating melts (i.e., the PI domain defined in the preceding section). The refertilized mantle of Picazo et al. (2016) and the External Ligurian PI domain actually share several structural and compositional similarities, namely: (i) the presence of microstructures indicating melt-rock reactions in the plagioclase stability field, (ii) the clinopyroxene and spinel composition (Figs. 2.6 and 2.7), and (iii) the relatively high temperatures computed for the plagioclase re-equilibration (Fig. 2.11). Conversely, the External Ligurian plagioclase-bearing spinel tectonites and plagioclase mylonites (ST and PM domains, respectively) record formation of plagioclase in response to subsolidus decompression, with no melt involvement. However, the reconstruction of the Jurassic OCT by Picazo et al. (2016) does not include the occurrence of a plagioclase-bearing mantle section with no melt-rock reaction evidence. The fertile spinel peridotites considered as inherited subcontinental mantle by Picazo et al. (2016) bear several similarities with the spinel tectonite and mylonite protoliths of the External Ligurian mantle bodies. Taken as a whole, these peridotites are characterized by a fertile geochemical signature, presence of Aland Na-rich clinopyroxene (Fig. 2.6) and disseminated Ti-amphibole. 38 Pyroxenite layers, in places garnet-bearing, are also widespread in both mantle section types. Following the subdivision of the Tethyan OCT mantle into two main types (Picazo et al., 2016), we therefore place the ST and PM mantle domains of the External Ligurian units into the inherited subcontinental mantle. The Malenco-Totalp-Upper Platta mantle bodies and the External Ligurian ST and PM mantle domains underwent distinct pressure-temperature evolutions during the post-Variscan lithospheric extension and the subsequent Mesozoic rifting (Fig. 2.11). A cold exhumation path was depicted for the Malenco mantle body, considered as the type locality of inherited mantle (Müntener et al., 2000). The Malenco peridotites underwent a near-isobaric cooling to ~600 °C in the late Paleozoic, followed by near-isothermal decompression from ~0.8 to ~0.4 GPa, associated with formation of chlorite and Mg-hornblende, locally along mylonitic shear zones (see also Müntener and Hermann, 2001). Based on 40Ar/39Ar amphibole investigations (Villa et al., 2000), the decompression was inferred to have started in the late Triassic (~225 Ma) and was related to the onset of the rifting evolution leading to opening of the Western Tethys. The External Ligurian PM domain provides evidence for a two-stage high temperature decompression starting from the base of continental lithosphere (Montanini et al., 2006). The early decompression from ~2.8 GPa and ~1100 °C to the spinel lherzolite stability field was associated with cooling to ~950 °C and was followed by a polyphase shearing under plagioclase-facies conditions ending at 0.3–0.6 GPa and 750–780 °C (Fig. 2.11). Whereas the age of the high temperature decompression to spinel-facies lherzolite conditions is unconstrained, the Lu–Hf cooling age of 220 ± 13 Ma (Montanini et al., 2006) indicate that the spinel to plagioclase-facies transition was related to the Mesozoic rifting event (see also §2.6.3). The similar 40Ar/39Ar and Lu–Hf cooling ages obtained respectively for the Malenco mantle body and the External Ligurian PM domain suggest different thermal conditions in the late Triassic, namely the Malenco was ~250–300 °C colder than the External Ligurian PM domain. The distinct exhumation histories might reflect different mantle depths at the time of the rifting inception. In this view, the External Ligurian PM domain was exhumed from deeper (and hotter) conditions than the spinel peridotites from the Malenco mantle body (Fig. 2.12). The plagioclase-facies shear zones within the External Ligurian PM domain document a major deformation event that accommodated exhumation of the spinel-facies subcontinental mantle in the plagioclase stability field. These shear zones cannot be considered as counterparts of those occurring in the central Lanzo massif. The porphyroclastic spinel protoliths of the External Ligurian mylonites do not provide evidence for melt infiltration and refertilization before or during shearing. The Lanzo plagioclase-facies shear zones are also distinct for being crosscut by MOR-type gabbros and basalts. 39 Notably, available geochronological data for the replacive harzburgites and the gabbros of the Lanzo Massif cluster in the Middle Jurassic (Kaczmarek et al., 2008; Sanfilippo et al., 2019). Hence, the Lanzo shear zones were most likely related to an advanced stage of the rifting evolution, which ultimately formed an oceanward OCT sector. In summary, the onset of the Tethyan rifting involved uplift of subcontinental mantle in the late Triassic, as documented by the Malenco body (e.g., Müntener and Hermann, 2001) and the External Ligurian PM domain (Montanini et al., 2006). This scenario is consistent with the information acquired from the lower continental crust section of the Ivrea-Verbano Zone, which formed a distal sector of the extending Adria continental margin in the Middle Jurassic (e.g., Beltrando et al., 2015). The Ivrea-Verbano Zone actually documents that the rifting evolution started in the late Triassic (see also Ferrari et al., 2021) and was associated with localized infiltration of mantle-derived melts with overall alkaline signature (Bonazzi et al., 2020; Schaltegger et al., 2015; Stahle et al., 1990). Notably, the alkaline magmatism in the Ivrea-Verbano Zone continued until the early Jurassic (see also Galli et al., 2019 and Grieco et al., 2001). From the late Triassic, the rifting evolution probably comprised asthenosphere ascent, melt production and refertilization of the subcontinental mantle, with a Middle Jurassic climax developing MORB-type melts and opening of the Western Tethys. 2.7. Conclusions The subcontinental mantle section from the External Ligurian units (Northern Appennine) consists of three distinct mantle domains, developed in response to the rifting evolution that ultimately formed a Middle Jurassic ocean-continent transition: (1) a spinel tectonite domain that underwent no significant deformation and melt-rock reaction under plagioclase-facies conditions, characterized by static plagioclase development under decreasing pressure conditions, (2) a plagioclase mylonite domain experiencing melt-absent deformation under decreasing pressure conditions, and (3) a nearly undeformed domain that underwent melt infiltration and reaction under plagioclase-facies conditions. We relate mantle domains (1, 2) to a rifting-driven uplift in the late Triassic accommodated by largescale shear zones consisting of plagioclase mylonites. We distinguish different mantle exhumation styles for the onset of the Tethyan rifting. A cold mantle exhumation is exemplified by the Malenco mantle body (Central Alps), in which low temperature hydrous mylonites developed at the expense of porphyroclastic spinel lherzolites. The External Ligurian mantle domains (1, 2) highlight mantle exhumation accommodated by large-scale shear zones consisting of anhydrous, melt-absent plagioclase-facies mylonites. The role of these shear zones in mantle exhumation along modern and fossil OCTs has been overlooked and needs further reconsideration in future studies. From the late Triassic, the Tethyan rifting evolution probably 40 involved asthenosphere ascent, melt production and refertilization of the subcontinental lithospheric mantle. The crustal breakup was associated with a major event of formation of MORB-type melts in the Middle Jurassic, with the mantle exhumation partly driven by a melt-present plagioclase-facies deformation, as documented by the Lanzo mantle body from Western Alps. 2.8. References Basch V., Borghini G., Fumagalli P., Rampone E., Gandolfo A., Ferrando C., 2020. Plagioclasefacies thermobarometric evolution of the External Liguride pyroxenite-bearing mantle (Suvero, Italy). Ofioliti 45, 1-12 Basch V., Rampone E., Borghini G., Ferrando C., Zanetti A., 2019. Origin of pyroxenites in the oceanic mantle and their implications on the reactive percolation of depleted melts. Contributions to Mineralogy and Petrology 174, 97 Beccaluva L., Macciotta G., Piccardo G.B. and Zeda O., 1984. Petrology of lherzolitic rocks from the Northern Apennine ophiolites. Lithos 17, 299-316 Beltrando M., Stockli D.F., Decarlis A. and Manatschal G., 2015. A crustal-scale view at rift localization along the fossil Adriatic margin of the Alpine Tethys preserved in NW Italy. Tectonics 34, 1927-1951 Bodinier J.-L., Garrido C.J., Chanefo I., Bruguier O. and Gervilla F., 2008. Origin of pyroxeniteperidotite veined mantle by refertilization reactions: Evidence from the Ronda peridotite (Southern Spain). Journal of Petrology 49, 999-1025 Bonazzi M., Langone A., Tumiati S., Dellarole E., Mazzucchelli M., Giovanardi T. and Zanetti A., 2020. Mantle-Derived Corundum-Bearing Felsic Dykes May Survive Only within the Lower (Refractory/Inert) Crust: Evidence from Zircon Geochemistry and Geochronology (IvreaVerbano Zone, Southern Alps, Italy). Geosciences 10, 281 Borghini G., Fumagalli P. and Rampone E., 2010. The Stability of Plagioclase in the Upper Mantle: Subsolidus Experiments on Fertile and Depleted Lherzolite. Journal of Petrology 51, 229-254 Borghini G., Fumagalli P. and Rampone E., 2011. The geobarometric significance of plagioclase in mantle peridotites: A link between nature and experiments. Lithos 126, 42-53 Borghini G., Rampone E., Class C., Goldstein S., Cai Y., Cipriani A., Hofmann A.W., Bolge L., 2021. Enriched Hf–Nd isotopic signature of veined pyroxenite-infiltrated peridotite as a possible source for E-MORB. Chemical Geology 586, 120591 41 Borghini G., Rampone E., Zanetti A., Class C., Cipriani A., Hofmann A.W. and Goldstein S.L., 2013. Meter-scale Nd isotopic heterogeneity in pyroxenite-bearing Ligurian peridotites encompasses global-scale upper mantle variability. Geology 41, 1055-1058 Borghini G., Rampone E., Zanetti A., Class C., Cipriani A., Hofmann A.W. and Goldstein S.L., 2016. Pyroxenite Layers in the Northern Apennines’ Upper Mantle (Italy)-Generation by Pyroxenite Melting and Melt Infiltration. Journal of Petrology 57,625-653 Borghini G., Rampone E., Zanetti A., Class C., Fumagalli P., Godard M., 2020. Ligurian pyroxeniteperidotite sequences (Italy) and the role of melt-rock reaction in creating enriched-MORB mantle sources. Chemical Geology 532, 119252 Brey G.P. and Köhler T., 1990. Geothermobarometry in four phase lherzolites II. New thermobarometers, and practical assessment of existing thermobarometers. Journal of Petrology 31, 1353-1378 Chazot G., Charpentier S., Kornprobst J., Vannucci R. and Luais B., 2005. Lithospheric Mantle Evolution During Continental Break-Up: The West Iberia Non-Volcanic Passive Margin. Journal of Petrology 46, 2527-2568 Conti P., Cornamusini G. and Carmignani L., 2020. An outline of the geology of the Northern Apennines (Italy), with Geological Map at 1:250,000 scale. Italian Journal of Geosciences 139, 149-194 Della-Pasqua F.N., Kamenetsky V.S., Gasparon M., Crawford A.J. and Varne R., 1995. Al-spinels in primitive arc volcanics. Mineralogy and Petrology 53, 1-26 Dygert N. and Liang Y., 2015. Temperatures and cooling rates recorded in REE in coexisting pyroxenes in ophiolitic and abyssal peridotites. Earth and Planetary Science Letters 420, 151-161 Elter P., Ghiselli F., Marroni M. and Ottria G., 1997. Note illustrative del Foglio 198 "Bobbio" della Carta Geologica d'Italia in scala 1:50.000 Ferrari E., Tribuzio R., Bosch D. and Bruguier O., 2021. Constraints on the post-Variscan thermal evolution of the Ivrea crustal section (Italian-Swiss Alps) from U-Pb dating of relict rutile in middle crust amphibolites. Lithos 406-407, 106500 Fumagalli P., Borghini G., Rampone E. and Poli S., 2017. Experimental calibration of Forsterite– Anorthite–CaTscherm.ak–Enstatite (FACE) geobarometer for mantle peridotites. Contributions to Mineralogy and Petrology 172, 38 48 2.9. Tables Table 2.1. Summary of microstructural and mineralogical features of Monte Gavi and Monte Sant'Agostino mantle rocks Samples Lithology Thickness Microstructure Main petrographic features Monte Gavi mantle sequence Peridotites MGA7 Cpx-rich Spl-Pl harzburgite aNearly isotropic; coarse-grained, impregnated Evidence of melt-rock reaction: Cr-Ti-rich Spl (Cr#= 33-35, TiO2 =0.3-0.5 wt.%) with Pl coronas; Pl + Opx as replacement of Cpx Type Ia pyroxenites GAV2 Spl websterite* ~10-80 cm Isotropic, coarse-grained, granular allotriomorphic Coarse Al-Spl (Cr# = 2-4, TiO2 < 0.10 wt %) with thin Pl rims and smaller Cr-Spl (Cr# = 12-34, TiO2 = 0.15-1.0 wt%); embayed coarse Cpx with Pl + Opx lamellae; Mg# of coarse Cpx = 84-86 Type Ib pyroxenites GA1, MG6 Spl websterite* ~10-80 cm Isotropic, coarse-grained, granular allotriomorphic Pyroxene-rich (Cpx + Opx + Pl) and Pl-rich (Pl + Ol + Cr-Spl + Ilm + sulfide) domains relict coarse Cpx replaced by Opx + Pl as rims and lamellae; Mg# of Cpx = 79-85 Type II pyroxenites MG5, IC2/3, MG9 Ol-Spl websterite* ~10 cm Isotropic, medium-to coarsegrained, granular allotriomorphic Relict exsolved Opx and Fo-rich Ol (89 mol%) from the host peridotite + Cpx-rich domains with interstitial Pl (embayed Cpx with lamellae of Pl + Opx); spinels with variable Cr# (6-16) and TiO2 (0.1-0.6 wt%) rimmed by Pl Monte Sant'Agostino mantle sequence Peridotites MGO1, IC8-1, IC8-3, IC11 SplPl lherzolite aFoliated, protomylonitic to ultramylonitic with spinel tectonite relics Opx+Cpx (+Spl+Ti-Amp) porphyroclasts in a ultrafine-grained Ol+Opx+Cpx+Pl ± Amp ± Spl polyphase matrix; highly stretched Opx. Al2O3 (7.0-8.4 wt%) and Na2Orich (1.6-2.1 wt%) Cpx porphyroclasts, Cr-and Ti-poor Spl (Cr# = 8-16, TiO2 < 0.2 wt%) Pyroxenites AM489, MP7, IC10-1 Spl websterites ~1-10 cm Protomylonitic Porphyroclasts of Cpx + Opx + Al-Spl set into a matrix of Cpx + Opx + Ol + Pl + TiAmp; Opx + Al-Spl + Pl symplectitic intergrowths after garnet * protolith before melt-rock interaction; sample locations are reported in Fig. 2.2 49 Table 2.2. Representative major element compositions (wt%) of minerals from Monte Gavi selected samples. SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O K2O NiO Total Mg# Cr# An MGA7 Harzburgite Ol 40.61 - - - 9.57 0.08 49.42 0.09 - - 0.50 100.27 90 - - Opx core 54.49 0.09 5.40 0.67 7.20 0.11 32.38 0.79 - - - 101.12 89 - - Opx rim 55.48 0.15 3.59 0.79 7.11 0.17 32.79 0.87 - - - 100.95 89 - - Cpx core 49.90 0.58 7.29 0.92 3.49 0.06 14.77 21.65 0.88 - - 99.53 88 - - Cpx rim 51.66 0.83 4.69 1.13 2.88 - 15.98 22.19 0.75 - - 100.11 91 - - Spl - 0.50 36.92 28.47 19.30 0.22 14.53 - - - - 99.95 57 34 - GAV2 Type-Ia pyroxenite Coarse Cpx core 48.60 0.77 8.66 0.18 4.73 0.01 13.88 21.92 0.70 - - 99.45 84 - - Coarse Cpx rim 51.25 0.67 5.22 0.24 4.50 0.18 15.18 22.74 0.49 - - 100.47 86 - - Small Cpx core 49.77 0.79 7.41 0.24 4.65 0.30 15.43 21.62 0.66 - - 100.86 86 - - Small Cpx rim 51.11 0.80 4.97 0.30 4.54 0.04 15.76 22.74 0.44 - - 100.69 86 - - Al-Spl - - 63.43 2.12 15.73 0.14 18.67 - - - - 100.10 71 2 - Cr-Spl - 1.02 32.98 25.40 27.43 - 12.28 - - - - 99.11 53 34 - GA1 Type-Ib pyroxenite Pyroxene-rich domain Coarse Cpx1 core 46.48 2.33 9.71 0.14 6.85 0.12 14.43 17.64 1.73 - - 99.43 79 - - Coarse Cpx2 core 48.44 2.01 7.88 0.20 5.40 0.20 15.58 17.75 1.40 - - 98.86 84 - - Coarse Cpx2 rim 50.46 1.66 6.21 0.13 5.15 0.08 14.61 22.50 0.46 - - 101.27 83 - - Small Cpx 49.24 1.58 6.65 0.11 5.32 0.24 14.23 21.69 0.66 - - 99.73 83 - - Opx (rim of Cpx) 54.62 0.38 2.52 0.07 12.28 0.40 29.01 0.65 - - - 99.92 81 - - Opx (coarse lamella in Cpx) 54.13 0.50 2.85 0.05 11.65 0.30 29.13 0.95 - - - 99.56 82 - - Amp 42.57 3.67 12.88 0.38 7.74 0.11 15.21 11.65 3.26 - - 97.48 78 - - Plagioclase-rich domain Pl 47.89 - 34.01 - 0.33 - - 15.42 2.30 - - 99.94 - - 79 Ol 39.87 - - - 17.39 0.30 42.52 - - - 0.24 100.32 81 - - Spl - 0.61 35.81 23.16 28.25 0.37 11.00 - - - - 99.20 48 30 - 50 MG6 Type-Ib pyroxenite Pyroxene-rich domain Coarse Cpx core 48.79 1.69 7.49 0.12 5.39 0.18 15.45 20.69 0.77 - - 100.57 84 - - Coarse Cpx rim 50.23 1.51 5.54 0.11 4.70 0.16 15.54 22.11 0.40 - - 100.30 85 - - Small Cpx 49.67 1.58 6.16 0.19 4.81 - 15.49 21.65 0.52 - - 100.07 85 - - Opx (symplectite with Pl) 54.13 0.40 2.91 0.11 10.99 0.31 30.07 0.96 - - - 99.88 83 - - Plagioclase-rich domain Pl 47.42 - 33.70 - - - 14.43 2.95 - - - 98.50 - - 73 Ol 39.47 - - - 17.59 0.10 43.70 - - - 0.19 101.05 82 - - Spl - 0.79 45.02 14.48 26.33 0.15 12.19 - - - - 98.96 51 18 - IC2/3 Type-II pyroxenite Ol 40.57 - - - 10.63 0.10 48.20 - - - 0.40 99.90 89 - - Exsolved Opx 54.71 0.24 5.54 0.55 7.88 0.13 30.46 0.97 0.37 - - 100.85 87 - - Cpx 50.78 1.02 7.41 0.41 3.30 0.18 15.34 22.53 0.56 - - 101.53 89 - - Opx (coarse lamella in Cpx) 55.68 0.26 2.07 0.46 7.44 0.25 33.39 0.72 - - - 100.27 89 - - Spl - 0.55 52.49 15.03 15.06 0.24 16.63 - - - - 100.00 66 16 - Pl 51.82 0.01 31.16 0.05 0.04 - 0.11 13.89 3.70 - - 100.78 - - 67 Major element compositions of minerals were analysed using a JEOL-6400 electron microprobe equipped with a LINK-ISIS energy dispersive microanalytical system at the Department of Chemistry, Life Sciences and Environmental Sustainability of Parma University. The electron beam was produced at an accelerating voltage of 15 kV and probe current of 0.25 nA. Both natural minerals and synthetic compounds were used as standards. pfc=porphyroclast, neo=neoblast, (-) below detection limit (< 0.10 wt%). Mineral abbreviations after Whitney and Evans (2010). 51 Table 2.3. Representative major element compositions (wt%) of minerals from Monte Sant'Agostino peridotites SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O K2O NiO Total Mg# Cr# An MGO1 Mylonitic lherzolite Ol pfc 40.97 - - - 10.70 0.10 47.95 0.14 - - - 99.86 89 - - Cpx2 pfc core 51.24 0.42 8.42 0.99 2.70 0.10 14.21 19.60 2.06 - - 99.74 90 - - Cpx pfc rim 51.63 0.65 5.92 0.85 2.66 - 15.11 22.52 0.93 - - 100.27 91 - - Opx pfc core 54.84 0.33 4.73 0.56 6.65 - 32.83 0.70 - - - 100.64 90 - - Spl - 0.15 59.59 9.90 11.55 0.29 18.70 - - - - 100.18 74 10 - IC8-1 Mylonitic lherzolite Ol pfc 41.27 - - - 10.31 0.07 48.37 - - - 0.49 100.51 89 - - Cpx pfc core 50.45 0.48 7.69 1.45 2.60 0.38 13.87 21.25 1.71 - - 99.88 90 - - Opx pfc core 54.87 0.11 4.50 0.44 7.33 0.12 32.12 0.60 - - - 100.09 89 - - Spl pfc core - - 60.90 8.38 11.95 0.02 19.02 - - - - 100.27 74 8 - Amp 41.54 4.62 14.78 1.20 4.65 0.20 15.14 11.92 4.03 0.10 - 98.18 85 - - Ol neo 41.21 - - - 9.55 0.14 49.91 - - - 0.44 101.25 90 - - Cpx neo 52.80 0.51 2.94 0.27 2.75 0.10 17.00 23.20 0.56 - - 100.13 92 - - Opx neo 55.82 0.10 1.62 0.25 7.23 0.32 35.61 0.34 - - - 101.29 90 - - Pl neo 55.41 - 27.78 - 0.27 - - 8.75 7.04 - - 99.25 - - 41 IC8-3 Mylonitic lherzolite Cpx pfc core 50.65 0.52 8.02 0.55 2.95 0.17 13.83 20.69 1.76 - - 99.14 89 - - Cpx pfc rim 51.56 0.42 7.39 0.76 2.52 - 14.95 20.46 1.44 - - 99.51 91 - - Opx pfc core 54.59 0.25 5.01 0.44 7.00 0.17 32.10 0.70 - - - 100.26 89 - - Spl pfc core - 0.15 57.60 11.43 12.72 0.37 17.91 - - - - 100.18 72 12 - Amp 42.19 3.59 14.54 1.02 6.00 0.17 17.17 10.91 2.81 0.26 - 98.66 84 - - Ol neo 40.84 - - - 9.86 0.18 49.31 0.10 - - 0.30 100.59 90 - - Cpx neo 54.41 0.58 3.06 0.62 2.58 - 17.71 22.54 0.70 - - 102.20 92 - - Opx neo 56.71 0.16 1.06 - 7.02 0.2 35.42 0.32 - - - 100.89 90 - - Pl neo 56.64 - 27.52 - 0.60 - - 7.75 7.14 - - 99.64 - - 38 IC11 ultramylonitic lherzolite Cpx pfc core 54.75 0.56 6.54 0.83 3.93 0.10 15.14 21.11 1.44 - - 104.40 87 - - Opx pfc core 55.35 0.31 3.65 0.14 6.56 0.22 33.42 0.56 - - - 100.21 90 - - Spl pfc core - - 57.74 10.52 11.93 0.25 19.52 - - - 0.36 100.32 76 11 - Amp pfc 43.37 2.42 14.74 0.86 4.24 - 17.17 11.95 3.47 - 98.22 90 - - Ol neo 41.04 - - - 9.79 0.13 49.73 - - - 0.45 101.14 90 - - Cpx neo 54.77 0.23 2.23 0.25 2.39 - 17.92 23.37 0.54 101.70 93 - - Opx neo 55.70 - 0.74 0.10 7.06 0.14 35.65 0.28 - - - 99.67 90 - - Pl neo 54.17 - - - 0.29 - - 10.41 5.82 - - 70.69 - - 50 52 Spl (neo) - - 55.13 10.49 13.47 0.12 20.49 - - - 0.28 99.98 80 11 - pfc=porphyroclast, neo=neoblast, (-) below detection limit (< 0.10 wt%). Mineral abbreviations after Whitney and Evans (2010). 53 Table 2.4. Representative major element compositions (wt%) of minerals from Monte Sant'Agostino pyroxenites SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O K2O NiO Total Mg# Cr# An MP7 Spinel websterite Cpx pfc core 49.64 0.66 9.41 0.23 4.68 0.26 13.40 20.40 1.27 - - 99.92 84 - - Cpx pfc rim 49.81 0.77 8.33 0.51 3.99 0.15 14.17 20.99 0.97 - - 99.68 86 - - Opx pfc core 51.81 0.13 6.86 0.07 10.67 0.43 28.97 0.62 - - - 99.56 83 - - Opx pfc rim 53.52 0.21 4.49 0.19 9.98 0.28 30.49 0.57 - - - 99.74 84 - - Cpx neo 50.24 0.77 7.33 0.18 5.07 0.20 14.25 21.72 0.77 - - 100.52 83 - - Opx neo 54.30 0.20 2.72 - 11.07 0.36 30.43 0.48 - - - 99.57 83 - - Spl neo - 0.09 65.28 1.30 16.58 0.15 17.96 - - - - 101.35 66 1 - Ol neo 38.42 - - - 17.07 0.31 42.79 - - - - 98.59 82 - - Pl neo 52.48 - 29.72 - - - - 10.25 4.92 - - 97.37 - - 54 Amp 41.32 4.23 14.36 0.64 6.71 0.13 15.06 11.91 2.93 0.45 - 97.74 80 - - Opx sym 54.02 0.22 3.57 - 10.91 0.35 30.44 0.52 - - - 100.02 83 - - Spl sym - 0.09 63.55 0.74 16.86 0.27 16.95 - - - - 98.46 64 1 - Pl sym 52.90 - 30.18 - - - - 10.19 5.19 0.18 - 98.64 - - 52 AM489 Spinel websterite Cpx pfc core 49.79 0.63 9.17 0.34 3.35 0.21 13.51 21.82 0.73 - - 99.55 88 - - Cpx pfc rim 52.61 0.58 4.42 0.57 2.64 0.15 15.83 23.02 0.52 - - 100.35 91 - - Opx pfc core 54.07 - 6.52 0.23 7.57 0.19 31.40 0.51 - - - 100.47 88 - - Opx pfc rim 55.86 - 4.10 0.18 7.65 0.28 32.39 0.55 - - - 101.01 88 - - Ol neo 40.20 - - - 11.71 0.33 47.62 - - - 0.21 100.07 88 Cpx neo 50.85 1.26 5.73 3.76 0.21 16.99 21.16 0.67 - - 100.62 89 - - Opx neo 55.40 0.26 2.88 7.41 0.29 33.06 0.67 - - - 99.97 88 - - Spl neo - - 62.50 6.42 13.21 0.45 18.07 - - - - 100.65 71 6 - Pl neo 52.54 - 30.98 - - - - 10.89 4.93 - - 99.34 - - 55 Amp 42.51 5.39 14.99 0.78 5.45 0.28 14.13 11.82 3.29 0.21 - 98.85 83 - - 54 IC10-1 Spl websterite Cpx pfc core 53.94 0.30 4.75 0.15 - 9.41 0.28 30.58 0.74 100.13 85 Opx pfc core 48.88 0.99 8.41 0.37 - 4.37 0.17 14.70 20.76 1.11 99.76 86 Ol neo 39.86 - - - 13.19 0.07 45.16 0.21 - - - 98.48 86 Opx neo 55.35 0.23 3.07 0.30 8.56 0.23 32.45 0.69 - - - 100.88 87 Cpx neo 50.86 1.05 5.65 0.35 4.00 0.26 15.93 21.92 0.85 - - 100.86 88 Pl neo 52.93 - 30.06 - 0.47 - - 10.96 5.06 - - 99.47 - 54 pfc=porphyroclast, neo=neoblast, sym=symplectite, (-) below detection limit (< 0.10 wt%). Mineral abbreviations after Whitney and Evans (2010). 55 Table 2.5. Geotermobarometric estimates for the Monte Gavi and Monte Sant'Agostino mantle sections Monte Gavi TCa-in-Opx TTa98 TBK90 THB94 PFACE (GPa) Peridotites Exsolved Opx(1)- Cpx (core) 950-1110 878-916 897-938 Type Ipyroxenites Non-exsolved Opx associated with Pl (2) 920-1030 Opx - adjacent Cpx in Cpx-rich domain 850-954 876-989 Pl-Amp 1030 0.67-0.78 Type II-pyroxenites Exsolved Opx(1) 1001-1047 Non-exsolved Opx (2) associated with Pl 910-1065 Opx - adjacent Cpx 970-983 996-1015 0.72 Monte S. Agostino Mylonitic peridotites Porphyroclastic Opx(1)- Cpx 930-970 832-886 857-950 Neoblastic Pl-bearing mylonitic matrix (3) 750-780 756-825 788-873 778 0.34-0.55 Spl-Pl websterites Porphyroclastic Opx(1)- Cpx 920-985 914-958 940-960 Neoblastic Pl-bearing matrix (4) 870-930 856-880 905-938 0.76-0.90 Pl-Amp 880-907 Assumed P for TCa-in-Opx : (1) 1.5 GPa, (2) 0.7 GPa, (3) 0.4 GPa, (4) 0.8 GPa 56 Chapter 3 Insights on the Monte Gavi plagioclasefacies melt-rock reaction event 3.1. Introduction The mantle rock bodies from the External Ligurian units (Northern Apennines, Italy) record distinct pressure-temperature evolutions, as well as deformation processes during the Mesozoic rifting stage that preceded the Jurassic Western Tethys opening. A relatively cold decompression evolution linked to the extensional tectonics is widely reported for several of these mantle bodies. Exhumation to plagioclase-facies conditions occurred statically in the Suvero mantle body (e.g., Hidas et al., 2020). Conversely, the Rio Strega-Monte Prinzera (Montanini et al., 2006) and Monte Sant’Agostino (PhD Thesis, Chapter 2) bodies provide evidence for dynamic plagioclase-facies recrystallization along hectometre-scale lithospheric shear zones. Furthermore, plagioclase-facies melt-rock reaction processes are recorded by the Monte Nero mantle body (Piccardo et al., 2004). Plagioclase-facies melt-rock reaction processes were also reported in this PhD Thesis for the Monte Gavi mantle body (see Chapter 2). This mantle section consists of harzburgites interlayered with spinel pyroxenites levels, which were subdivided into Type-I and Type-II. Because of the meltrock interaction event, both peridotites and pyroxenites underwent different replacement extents of: (i) clinopyroxene, by orthopyroxene + plagioclase, and (ii) spinel, by plagioclase. Type-I pyroxenites were subdivided in two sub-types (Ia and Ib) based on the different replacement extents. In particular, Type-Ib pyroxenites are characterized by (i) pyroxene-rich domains in which coarse clinopyroxene relics are associated to newly formed secondary orthopyroxene + plagioclase + clinopyroxene, and (ii) plagioclase-rich domains constituted by plagioclase + olivine + Cr-spinel presumably representing the reaction products between the percolating melt and original Al-spinel-rich domains. Conversely, in Type-Ia pyroxenites coarse clinopyroxene and spinel grains are well preserved, despite the local crystallization of orthopyroxene and plagioclase. Along with the petrographic evidence, mineral major element compositions also registers the effects of melt impregnation. In particular, an increase of Cr# values (up to 34) and TiO2 contents (up to 1 wt%) is recorded by spinels from pyroxenites and enclosing peridotites. Moreover, reacted clinopyroxenes from pyroxenites display an increase in TiO2 contents (up to 2.3 wt%). Overall, the percolating melt should have been 57 silica(orthopyroxene)-saturated and Ti-rich. According to the FACE geobarometer of Fumagalli et al. (2017), the plagioclase-facies melt-rock reaction process occurred at 0.7-0.8 GPa. The aim of this Chapter is to assess the Monte Gavi melt impregnation event. New geochemical data on whole rocks and mineral phases will give additional insights on the nature of the percolating melt. Nd-Hf isotopic data from pyroxenite samples will be used to discuss the possible sources of this melt, besides providing age constraints on the melt-rock reaction event. 3.2. Geological and petrological framework The Alpine-Apennine ophiolites represent lithospheric remnants of the Jurassic Western Tethys basin (also referred to as Ligurian-Piedmontese basin). Some ophiolites from the Central Alps (Malenco and Totalp; Müntener et al., 2004, 2010) and Northern Apennines (External Ligurian units; Rampone et al., 1995; Montanini et al., 2006, 2012) expose mantle bodies showing a subcontinental origin and were interpreted to be exhumed along the Western Tethys magma-poor ocean-continent transition. Other ophiolites from the Alpine-Apennine system (e.g. those from the Internal Ligurian Units) show affinities with oceanic lithosphere from slow and ultra-slow spreading ridges (Sanfilippo and Tribuzio, 2011; Tribuzio et al., 2014). The External Ligurian (EL) ophiolites occur as slide blocks within Late Cretaceous sedimentary melanges formed during the convergence phase that led to the closure of the Western Tethys basin (e.g., Marroni et al., 2017). The cmto km-sized slide blocks comprise subcontinental mantle rocks (e.g., Piccardo et al., 2004), Jurassic MOR-type gabbros and basalts (Tribuzio et al., 2004, 2016; Montanini et al., 2008) and associated pelagic sediments, Late Carboniferous-Early Permian granitoids (Ferrara and Tonarini, 1985) and mafic granulites (Meli et al., 1996). Mantle rocks are mainly constituted by fertile spinel-plagioclase lherzolites (e.g., Rampone et al., 1995) locally bearing accessory amphibole (kaersutite to titanian pargasite). Peridotites are in places interlayered with garnet and spinel pyroxenites (Montanini et al., 2006, 2012; Montanini and Tribuzio, 2015; Borghini et al., 2016). As reported in the previous Chapter 2, the External Ligurian subcontinental mantle sequences could be grouped in three main domains according to their rifting-related evolution preceding the opening of the Jurassic Western Tethys basin. A plagioclase-bearing spinel tectonite (ST) domain, exemplified by the Suvero section (e.g., Rampone et al., 1995), recorded a high-temperature (10001050 °C) spinel-facies deformation. A subsequent plagioclase-facies recrystallization occurred in Jurassic times under melt-absent static conditions (Borghini et al., 2016; Hidas et al., 2020). Two recrystallization stages were recognized, from 0.6-0.7 GPa and 890-910 °C to 0.4-0.5 GPa and 800840 °C (Fumagalli et al., 2017) and were ascribed to the rifting-related decompression evolution. A 64 depletion (LaN/SmN = 0.1-0.3) coupled to variably enriched HREE (GdN/YbN = 0.4-0.9) and no to weak negative Eu anomalies (Eu/Eu* = 0.7-1.1). Strong negative Sr anomalies and small negative Hf anomalies are also visible. Orthopyroxenes display progressively increasing concentrations from LREE to HREE (from ~0.01 to ~10 times chondritic values). No compositional variations were observed between core and rim of pyroxenes from both lherzolites and websterites. 3.6. Geothermometric estimates Trace element analyses of pyroxenes and plagioclase were employed to calculate temperatures based on slowly diffusing elements (REE, Y) according to the calibrations of Liang et al. (2013) and Sun and Liang (2017). The former is based on the REE-Y partitioning between coexisting clinopyroxene and orthopyroxene, whilst the latter is based on the REE-Y partitioning between coexisting clinopyroxene and plagioclase. The obtained estimates are reported in Table 3.4. Porphyroclastic clinopyroxene-orthopyroxene pairs from Monte Gavi harzburgite MGA7 yielded values of 1245 ± 90 °C. Calculated temperatures for coarse clinopyroxene and associated orthopyroxene in Type-Ib MG6 pyroxenite returned 1278 ± 32 °C. By using the calibration of Sun and Liang (2017) for Type-Ib coarse clinopyroxene grains and plagioclase, temperatures of 1257 ± 48 (MG6) and 1170 ± 14 °C (GA1) were obtained. The geothermometer of Liang et al. (2013) was also applied on the porphyroclastic clinopyroxene-orthopyroxene pairs from Monte Sant’Agostino samples for comparative purposes. Three mylonitic lherzolites (MGO1, IC8-1, IC8-3) returned a weighted average temperature of 1044 ± 43 °C (MSWD = 0.39). The spinel websterite AM489 yielded values of 1103 ± 78 °C. 65 a) b) Fig. 3.4. Representative Rare Earth Element compositions normalized to chondrite (normalizing values after Sun and McDonough, 1989) of clinopyroxene cores from Monte Gavi (a) peridotites and (b) pyroxenites. Data from Suvero (Borghini et al., 2016, 2020), Rio Strega-Monte Prinzera (Montanini et al., 2012; Montanini and Tribuzio, 2015; Montanini unpublished data) and Monte Nero (Piccardo et al., 2004) External Ligurian mantle sequences are shown for comparison. Representative clinopyroxene REE compositions from Monte Sant’Agostino section are also reported. Noteworthy, clinopyroxenes from Monte Sant’Agostino lherzolites have REE patterns similar to Suvero and Rio Strega-Monte Prinzera peridotites; clinopyroxenes from Monte Sant’Agostino spinel websterites have REE patterns comparable to the other External Ligurian websterites. 66 0.1 1 10 100 1000 La Ce Sr Pr Nd Zr Hf Sm Eu Gd Tb Dy YHo Er Tm Yb Lu Cpx/chondrite a) 0.1 1 10 100 1000 La Ce Sr Pr Nd Zr Hf Sm Eu Gd Tb Dy YHo Er Tm Yb Lu Cpx/chondrite b) Fig. 3.5. Representative extended incompatible elements compositions normalized to chondrite (normalizing values after Sun and McDonough, 1989) of clinopyroxene cores from Monte Gavi (a) peridotites and (b) pyroxenites. Representative clinopyroxene REE compositions from Monte Sant’Agostino section are also reported for comparative purposes. 67 4 5 6 7 8 9 10 11 0.05 0.15 0.25 0.35 0.45 Zr/Nd Ce/Y a) 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.05 0.15 0.25 0.35 0.45 Hf/Nd Ce/Y b) 0 5 10 15 20 25 0 0.5 1 1.5 Zr/Nd Ce/Y c) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.5 1 1.5 Hf/Nd Ce/Y d) Fig. 3.6. Zr/Nd and Hf/Nd vs. Ce/Y plots of clinopyroxenes from Monte Gavi (a, b) peridotites and (c, d) pyroxenites. Data from Monte Sant’Agostino, Suvero, Rio Strega-Monte Prinzera and Monte Nero mantle sequences are shown for comparison. 68 0.001 0.01 0.1 1 10 100 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Opx/chondrite a) 0.001 0.01 0.1 1 10 100 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Opx/chondrite b) Fig. 3.7. Representative Rare Earth Element compositions normalized to chondrite (normalizing values after Sun and McDonough, 1989) of orthopyroxene cores from Monte Gavi (a) peridotites and (b) pyroxenites. Data from Monte Sant’Agostino are shown for comparison. 69 Fig. 3.8. Representative Rare Earth Element compositions normalized to chondrite (normalizing values after Sun and McDonough, 1989) of Monte Gavi plagioclase from Type-Ib pyroxenites. 0.1 1 10 100 La Ce Pr Nd PmSm Eu Gd Tb Dy Ho Er Tm Yb Lu Pl/chondrite 70 3.7. Nd-Hf isotopic compositions Sm-Nd isotope analyses were performed on 6 whole rock powders (GAV2, GA1, GA1 Px-Rich Fraction, GA1 Pl-Rich Fraction, MG6, IC2/3) and 4 clinopyroxene separates (GAV2, GA1, MG6, MG5) from both Type-I and Type-II pyroxenite samples. Additional Lu-Hf isotopic compositions were obtained for the same clinopyroxene separates (Tab. 3.5). Present-day 143Nd/144Nd ratios of Type-Ia whole rock and clinopyroxene (sample GAV2) describe a narrow range (0.513148 and 0.513133, respectively) coupled to similar 147Sm/144Nd ratios (0.27 and 0.22, respectively). A positive correlation between 143Nd/144Nd and 147Sm/144Nd ratios is depicted by Type-Ib samples (Fig. 3.9) which show a wide range of 143Nd/144Nd (0.512816-0.513252) and 147Sm/144Nd (0.17-0.50) values. The lowest Sm-Nd ratios are displayed by GA1 Plagioclase-Rich Fraction. Type-II pyroxenites have 143Nd/144Nd and 147Sm/144Nd values ranging from 0.512896 to 0.513166 and from 0.22 to 0.43, respectively. Type-II pyroxenite samples plot along the alignment described by Type-Ib samples. A consistent errorchron of 210 ± 20 Ma (n = 8; initial εNd = +4.25; MSWD = 9.8) was obtained by plotting all Type-Ib and Type-II samples together (Fig. 3.10). The four GA1 samples (whole rock, Plagioclase-Rich Fraction, Pyroxene-rich Fraction and clinopyroxene separate) define an errorchron (Fig. 3.11) which yielded a similar age (217 ± 36 Ma, initial εNd = +3.75; MSWD = 9.2). Type-Ib clinopyroxenes show 176Hf/177Hf values ranging from 0.282934 to 0.283008 and 176Lu/177Hf values varying within a small range (0.03-0.04). Type-II clinopyroxene has 176Hf/177Hf ratios (0.282983) at 176Lu/177Hf values (0.03) similar to Type-Ib clinopyroxenes. Type-Ia clinopyroxene shows higher 176Hf/177Hf ratios (0.283339) at nearly similar 176Lu/177Hf ratios (0.04) compared to Type-Ib samples. Initial Nd-Hf compositions of clinopyroxene separates were then computed at 210 Ma, which is the value returned by the Sm-Nd errorchron defined by Type-Ib and Type-II pyroxenites. Type-Ib and Type-II samples show a narrow interval of εNd(210) values (+3.97 to +4.18) at εHf(210) values varying from +5.29 to +8.80. Overall, the Nd-Hf isotopic composition of the Monte Gavi pyroxenites recalculated for the age provided by the Sm-Nd errorchron are poorly radiogenic and plot close or slightly below the mantle Nd-Hf array (Fig. 3.12). In contrast, Type-Ia clinopyroxene GAV2 has distinctly higher εNd(210) (+8.93) and εHf(210) (+19.25) values, falling in the more depleted portion of the MORB field, above the mantle array. 71 Fig. 3.9. Sm-Nd plot for whole rocks and clinopyroxene separates of Monte Gavi pyroxenite samples (clinopyroxene separates are distinguished by a cross). Fig. 3.10. 143Nd/144Nd vs. 147Sm/144Nd correlation diagram for all Type-Ib and Type-II pyroxenites. Data-point error ellipses are 2σ. 72 Fig. 3.11. 143Nd/144Nd vs. 147Sm/144Nd correlation diagram for GA1 samples. Datapoint error ellipses are 2σ. Fig. 3.12. εHf-εNd diagram for clinopyroxene separates of Monte Gavi pyroxenite samples. εHf and εNd values are calculated at 210 Ma. MORB (dark grey dots), OIB (light grey dots) and HIMU fields and the mantle Nd-Hf line are from Stracke (2012). The Depleted Mantle (DM) values are after Faure (1986) and Griffin et al. (2000). 73 3.8. Discussion 3.8.1. Geochemical response to plagioclase-facies melt infiltration Melt-rock reaction in the Monte Gavi mantle section triggered some major elements modifications in spinels (Cr# values and TiO2 contents) and clinopyroxenes (TiO2 contents) from both peridotites and pyroxenites (see §2.6.1. Monte Gavi: a mantle section recording reactive melt infiltration in the plagioclase stability field, Chapter 2). Trace element compositions of clinopyroxene bear further evidence of the melt infiltration process. Clinopyroxenes from Monte Gavi harzburgites display distinct REE patterns compared to the clinopyroxenes of the other External Ligurian peridotites not affected by plagioclase-facies melt impregnation (e.g., Suvero, Rio Strega-Monte Prinzera and Monte Sant’Agostino mantle sequences; Fig. 3.4). In particular, Monte Gavi clinopyroxenes are more LREE-depleted and show higher HREE concentrations. Their REE patterns also differ from those of Monte Nero peridotites, the other External Ligurian plagioclase-impregnated mantle domain (see §2.6.4. The heterogeneity of the External Ligurian subcontinental mantle, Chapter 2). This suggests that the Monte Gavi and Monte Nero mantle sections interacted with geochemically distinct melts. Clinopyroxenes from Monte Gavi peridotites testify the melt-rock reaction event also in having relatively high Zr contents coupled with relatively low Sr contents (Fig. 3.13). Following Müntener et al. (2010), who studied plagioclase peridotites from the Lower Platta unit, the high Zr/Sr ratios of Monte Gavi clinopyroxenes cannot be explained by a peridotite melting model. Instead, they are consistent with refertilization models, i.e. interaction with an infiltrating melt. The plagioclaseimpregnated Monte Nero peridotites follow this trend as well, whilst the other External Ligurian peridotites generally exhibit higher Sr contents coupled with lower Zr contents. Monte Gavi Type-I and Type-II pyroxenites recorded different extents of melt-rock interaction. Specifically, textural and mineral major elements evidence indicate that Type-Ia pyroxenites were not significantly affected by reaction with a percolating melt (see §2.6.1. Monte Gavi: a mantle section recording reactive melt infiltration in the plagioclase stability field, Chapter 2). Therefore, the REE patterns of Type-Ia clinopyroxenes presumably represent the composition prior to melt-rock reaction. Melt impregnation extensively modified the REE compositions of Type-Ib and Type-II clinopyroxenes, producing strong depletions in LREE, marked Eu anomalies and HREE enrichments (Fig. 3.4). Type-Ib and Type-II clinopyroxenes share analogous patterns, implying that the two distinct types of Monte Gavi pyroxenites reacted with the same percolating melt. Such clinopyroxene REE patterns are not displayed by the other External Ligurian pyroxenites (Fig. 3.4). The geochemical 80 Ferrara G. and Tonarini S., 1985. Radiometric geochronology in Tuscany: Results and problems. Rendiconti della Società Italiana di Mineralogia e Petrologia, 40: 111-124 Ferrari E., Montanini A., Tribuzio R., 2022. Rifting evolution of the lithospheric subcontinental mantle: New insights from the External Ligurian ophiolites (Northern Apennine, Italy). Lithos, 410-411:106571 Fumagalli P., Borghini G., Rampone E. and Poli S., 2017. Experimental calibration of Forsterite Anorthite–CaTscherm.ak–Enstatite (FACE) geobarometer for mantle peridotites. Contributions to Mineralogy and Petrology, 172:38 Griffin W.L., Pearson N.J., Belousova E., Jackson S.E., van Achterbergh E., O’Reilly S.Y., Shee S.R., 2000. The Hf isotope composition of cratonic mantle: LAM-MC-ICPMS analysis of zircon megacrysts in kimberlites. Geochimica et Cosmochimica Acta, 64:133-147 Guarnieri L., Nakamura E., Piccardo G.B., Sagaguchi C., Shimizu N., Vannucci R., Zanetti A., 2012. Petrology, trace element and SR, Nd, Hf isotope geochemistry of the North Lanzo peridotite massif (Western Alps, Italy). Journal of Petrology, 53(11):2259-2306 Hidas K., Borghini G., Tommasi A., Zanetti A. and Rampone E., 2020. Interplay between melt infiltration and deformation in the deep lithospheric mantle (External Liguride ophiolite, North Italy). Lithos, 380-381:105855 Jeffries T.E., Jackson S.E., Longerich H.P., 1998. Application of a frequency quintupled Nd:YAG source (λ=213 nm) for laser ablation inductively coupled plasma mass spectrometric analysis of minerals. Journal of Analytical Atomic Spectrometry, 13:935-940 Le Roux V., Bodinier J.-L., Alard O., O’Reilly S.Y., Griffin W.L., 2009. Isotopic decoupling during porous melt flow: A case-study in the Lherz peridotite. Earth and Planetary Science Letters, 279:76-85 Liang Y., Sun C., Yao L., 2013. A REE-in-two-pyroxene thermometer for mafic and ultramafic rocks. Geochimica et Cosmochimica Acta, 102:246-260 Marroni M., Meneghini F. and Pandolfi L., 2017. A revised subduction inception model to explain the Late Cretaceous, double-vergent orogen in the precollisional western Tethys: Evidence from the Northern Apennines. Tectonics, 36:2227-2249 Meli S., Montanini A., Thoni M. and Frank W., 1996. Age of mafic granulite blocks from the external Liguride units (Northern Apennines, Italy). Memorie di Scienze Geologica, 48: 65-72 81 Montanini A. and Tribuzio R., 2015. Evolution of recycled crust within the mantle: Constraints from the garnet pyroxenites of the External Ligurian ophiolites (northern Apennines, Italy). Geology, 43:911-914 Montanini A., Tribuzio R. and Anczkiewicz R., 2006. Exhumation History of a Garnet Pyroxenitebearing Mantle Section from a Continent-Ocean Transition (Northern Apennine Ophiolites, Italy). Journal of Petrology, 47:1943-1971 Montanini A., Tribuzio R. and Thirlwall M., 2012. Garnet clinopyroxenite layers from the mantle sequences of the Northern Apennine ophiolites (Italy): Evidence for recycling of crustal material. Earth and Planetary Science Letters, 351-352:171-181 Montanini A., Tribuzio R. and Vernia L., 2008. Petrogenesis of basalts and gabbros transition (external Liguride from an ancient continent-ocean ophiolites, northern Italy). Lithos, 101:453479 Müntener O., Manatschal G., Desmurs L., Pettke T., 2010. Plagioclase Peridotites in Ocean-Continent Transitions: Refertilized Mantle Domains Generated by Melt Stagnation in the Shallow Mantle Lithosphere. Journal of Petrology, 51:255-294 Müntener O., Pettke T., Desmurs L., Meier M. and Schaltegger U., 2004. Refertilization of mantle peridotite in embryonic ocean basins: trace element and Nd isotopic evidence and implications for crust-mantle relationships. Earth and Planetary Science Letters, 221:293-308. Piccardo G.B., Müntener O., Zanetti A. and Pettke T., 2004. Ophiolitic Peridotites of the AlpineApennine System: Mantle Processes and Geodynamic Relevance. International Geology Reviews, 46(12):1119-1159 Piccardo G.B., Padovano M., Guarnieri L., 2014. The Ligurian Tethys: Mantle processes and geodynamics. Earth-Science Reviews, 138:409-434 Rampone E., Borghini G., Basch V., 2020. Melt migration and melt-rock reaction in the AlpineApennine peridotites: Insights on mantle dynamics in extending lithosphere. Geoscience Frontiers, 11:151-166 Rampone E., Hofmann A.W., 2012. A global overview of isotopic heterogeneities in the oceanic mantle. Lithos, 148:247-261 82 Rampone E., Hofmann A.W., Piccardo G.B., Vannucci R., Bottazzi P. and Ottolini L., 1995. Petrology, Mineral and Isotope Geochemistry of the External Liguride Peridotites (Northern Apennines, Italy). Journal of Petrology, 36:81-105 Sanfilippo A. and Tribuzio R., 2011. Melt transport and deformation history in a nonvolcanic ophiolitic section, northern Apennines, Italy: Implications for crustal accretion at slow spreading settings. Geochemistry Geophysics Geosytems, 12(7) Snow J.E., Schmidt G. and Rampone E., 2000. Os isotopes and highly siderophile elements (HSE) in the Ligurian ophiolites, Italy. Earth and Planetary Science Letters, 175:119-132 Stracke A., 2012. Earth's heterogeneous mantle: A product of convection-driven interaction between crust and mantle. Chemical Geology, 330-331:274-299 Stracke A., Genske F., Berndt J., Koornneef J.M., 2019. Ubiquitous ultra-depleted domains in Earth’s mantle. Nature Geoscience, 12(10):851-855 Stracke A., Snow J.E., Hellebrand E., von der Handt A., Bourdon B., Birbaum K., Günther D., 2011. Abyssal peridotite Hf isotopes identify extreme mantle depletion. Earth and Planetary Science Letters, 308:359-368 Sun C., Liang Y., 2017. A REE-in-plagioclase-clinopyroxene thermometer for crustal rocks. Contributions to Mineralogy and Petrology, 172:24 Sun S.-s, McDonough W.F., 1989. Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. Geological Society Special Publications, 42:313-345 Tribuzio R., Garzetti F., Corfu F., Tiepolo M. and Renna M.R., 2016. U-Pb zircon geochronology of the Ligurian ophiolites (Northern Apennine, Italy): Implications for continental breakup to slow seafloor spreading. Tectonophysics, 666:220-243 Tribuzio R., Renna M.R., Dallai L. and Zanetti A., 2014. The magmatic-hydrothermal transition in the lower oceanic crust: Clues from the Ligurian ophiolites, Italy. Geochimica Cosmochimica Acta, 130:188-211 Tribuzio R., Thirlwall M. and Vannucci R., 2004. Origin of the Gabbro---Peridotite Association from the Northern Apennine Ophiolites (Italy). Journal of Petrology, 45:1109-1124 83 Villa I.M., Hermann J., Müntener O. and Trommsdorff V., 2000. 39Ar−40Ar dating of multiply zoned amphibole generations (Malenco, Italian Alps). Contributions to Mineralogy and Petrology, 140: 363-381 84 3.11. Tables Table 3.1. Bulk rock major and trace element compositions of Monte Gavi pyroxenites. Major element compositions are calculated on anhydrous basis and normalized to 100. Major elements are expressed in wt%, trace elements in ppm. Sample GAV2 GA1 GA1 Pl rich fraction GA1 Cpx rich fraction MG6 MG5 IC2-3 Description Type Ia pyroxenite Type Ib pyroxenite Type Ib pyroxenite Type Ib pyroxenite Type Ib pyroxenite Type II pyroxenite Type II pyroxenite SiO2 41.77 46.95 45.29 50.23 47.30 40.84 46.11 TiO2 0.25 0.42 0.14 0.82 0.48 0.38 0.37 Al2O3 14.02 14.44 20.66 11.04 14.06 12.27 7.09 FeO 6.92 8.58 7.03 7.09 8.84 7.49 7.77 MnO 0.14 0.16 0.12 0.20 0.17 0.16 0.13 MgO 24.17 20.04 15.74 18.73 19.02 32.82 30.10 CaO 12.50 6.99 7.82 10.06 7.47 5.92 8.21 Na2O 0.15 0.54 0.82 0.63 0.49 0.10 0.19 K2O 0.09 1.88 2.37 1.20 2.18 0.01 0.02 P2O5 bdl bdl bdl bdl bdl 0.01 bdl LOI 8.46 4.12 4.2 2.18 4.04 11.11 7.03 Mg# 86 81 80 82 79 89 87 SiO2/MgO 1.73 2.34 2.88 2.68 2.49 1.24 1.53 V 115 155 76 273 198 155 129 Cr 2570 790 890 650 620 3300 2400 Ni 900 560 620 260 530 990 1420 Co 53 54 56 35 52 55 69 Sc 17 36 6.00 71 39 35 22 Sr 15 84 118 61 92 9.00 16 Y 4.90 14 1.20 35 16 9.80 8.40 Nb bdl bdl bdl bdl bdl bdl bdl Zr 10 20.00 3.00 59.00 21.00 15.00 17.00 La 0.37 0.23 bdl 0.19 0.47 0.22 0.33 85 Ce 1.40 0.93 0.66 1.37 1.31 0.99 1.46 Pr 0.22 0.20 0.12 0.37 0.26 0.18 0.29 Nd 1.26 1.54 0.76 3.51 1.88 1.12 1.81 Sm 0.48 1.03 0.22 2.45 1.22 0.51 0.66 Eu 0.23 0.46 0.39 0.74 0.49 0.26 0.29 Gd 0.74 1.79 0.27 4.56 2.14 1.00 1.11 Tb 0.13 0.39 0.04 0.93 0.42 0.22 0.22 Dy 0.92 2.50 0.22 6.28 2.90 1.57 1.45 Ho 0.19 0.53 0.04 1.29 0.60 0.37 0.31 Er 0.52 1.48 0.11 3.56 1.76 1.10 0.91 Tm 0.08 0.22 0.02 0.51 0.24 0.18 0.14 Yb 0.56 1.39 0.11 3.08 1.69 1.17 1.02 Lu 0.09 0.21 0.02 0.48 0.26 0.18 0.16 Hf 0.30 0.70 bdl 1.50 0.70 0.40 0.50 Ta bdl bdl 0.13 0.01 bdl bdl bdl Th bdl bdl bdl bdl bdl bdl bdl U bdl bdl bdl 0.01 0.09 bdl bdl LOI = Loss On Ignition Mg# = [100 × Mg/(Mg+Fetot)] bdl = below detection limits 86 Table 3.2. Representative trace element compositions (ppm) of minerals from Monte Gavi selected samples. Sample MGA7 GAV2 GA1 MG6 MG5 Rock type Harzburgite Type-Ia pyroxenite Type-Ib pyroxenite Type-Ib pyroxenite Type-II pyroxenite Mineral Opx Cpx Cpx (core) Cpx (rim) Cpx (core) Cpx (int) Cpx (rim) Pl Cpx (core) Cpx (rim) Opx Pl Cpx (core) Sc 28 64 40 45 148 138 220 1.58 154 199 69 3.34 199 Ti 1100 4400 3871 3499 11197 11343 10945 345 10499 9850 3556 553 10233 V 137 312 259 247 543 493 687 2.21 750 867 402 23 491 Cr 3570 6064 2279 2586 757 726 952 bdl 673 1048 628 24 3605 Co 51 22 26 24 24 24 22 0.08 24 23 63 0.80 17 Ni bdl bdl 275 274.57 190 198.61 169 0.67 189 166 348 3.84 265 Rb bdl bdl 0.08 0.08 0.03 bdl bdl 0.06 0.05 0.05 0.06 0.09 0.09 Sr 0.06 7.61 8.20 7.34 1.53 1.20 1.06 25 1.88 1.89 0.38 40 2.72 Y 2.77 31 16 15 123 128 126 1.10 84 90 7.91 1.75 104 Zr 4.36 49 29 27 189 191 221 bdl 126 142 10 1.33 178 Nb bdl bdl 0.25 0.16 0.04 0.02 bdl bdl 0.03 0.02 bdl bdl 0.02 Ba bdl bdl 0.51 1.00 0.11 0.07 0.04 0.16 0.10 0.05 0.18 0.27 0.22 La bdl 0.47 0.69 0.50 0.26 0.25 0.18 0.24 0.24 0.20 0.01 0.15 0.48 Ce 0.03 3.34 3.72 3.15 2.69 2.65 2.39 0.96 2.57 2.23 0.02 0.90 3.91 Pr 0.01 0.87 0.76 0.65 1.00 1.17 0.97 0.18 0.93 0.89 0.01 0.19 1.23 Nd 0.10 6.36 4.77 3.95 11 11 11 0.96 9.19 9.32 0.12 1.26 11 Sm 0.07 2.90 1.75 1.53 9.23 9.04 8.76 0.23 6.13 6.31 0.14 0.38 7.42 Eu 0.04 1.11 0.79 0.67 2.29 2.23 2.02 0.62 1.73 1.80 0.03 0.60 1.72 Gd 0.20 4.19 2.49 2.42 16 17 17 0.29 11 12 0.42 0.33 13 Tb 0.05 0.77 0.47 0.42 3.47 3.40 3.54 0.05 2.33 2.46 0.11 0.06 2.79 Dy 0.39 5.42 3.00 2.87 23 24 26 0.22 15 17 1.03 0.26 20 Ho 0.09 1.12 0.61 0.57 4.95 5.10 5.00 0.04 3.30 3.46 0.27 0.08 4.54 Er 0.34 3.17 1.66 1.59 13 13 13 0.09 8.56 9.23 1.14 0.16 12 Tm 0.06 0.45 0.24 0.23 1.74 1.78 1.72 bdl 1.21 1.26 0.18 0.04 1.62 Yb 0.55 2.75 1.67 1.47 11 11 11 bdl 7.99 8.18 1.82 bdl 10 87 Lu 0.09 0.38 0.25 0.22 1.39 1.50 1.43 bdl 1.08 1.06 0.34 0.04 1.37 Hf 0.13 1.38 0.91 0.79 5.26 3.80 7.91 bdl 4.51 6.54 0.469 bdl 6.72 Ta bdl bdl 0.05 0.11 0.04 0.01 bdl 0.06 0.03 0.05 0.11 0.21 0.02 Pb bdl bdl 0.04 bdl 0.03 bdl 0.03 bdl 0.04 0.03 0.03 bdl 0.07 Th bdl bdl 0.03 bdl 0.01 bdl bdl bdl 0.01 0.02 bdl bdl 0.02 U bdl bdl 0.01 0.02 0.02 bdl bdl 0.02 0.01 bdl bdl bdl 0.01 LaN/SmN - 0.10 0.25 0.21 0.02 0.02 0.01 0.69 0.03 0.02 0.04 0.26 0.04 GdN/YbN 0.31 1.26 1.23 1.36 1.20 1.24 1.27 - 1.11 1.18 0.19 - 1.02 Eu/Eu* 1.00 0.97 1.15 1.06 0.57 0.55 0.51 7.29 0.65 0.64 0.43 5.12 0.54 bdl = below detection limits 88 Table 3.3. Representative trace element compositions (ppm) of minerals from Monte Sant'Agostino selected samples. Sample MGO1 IC8-1 IC8-3 AM489 Rock type Mylonitic lherzolite Mylonitic lherzolite Mylonitic lherzolite Spl websterite Mineral Opx (core) Opx (rim) Cpx (core) Cpx (rim) Opx (core) Cpx (core) Cpx (rim) Opx (core) Cpx (core) Opx (core) Cpx1 (core) Cpx1 (rim) Cpx2 (core) Cpx2 (rim) Sc 19 20 65 65 21 91 71 15 55 35 81 89 99 123 Ti 952 836 3689 4021 815 5872 4057 835 4187 728 3181 3425 3252 3468 V 102 99 254 254 101 366 283 93 293 161 341 348 273 280 Cr 2468 2138 5246 5040 1780 5599 5308 2261 4253 1136 1869 1492 1356 1188 Co bdl bdl bdl bdl 56 106 24 55 22.99 49 18 15 15 16 Ni bdl bdl bdl bdl 670 1978 371 677 356 396 169 136 179 180 Rb bdl bdl bdl bdl 0.18 0.22 0.06 0.07 0.05 0.23 0.18 0.08 bdl 0.08 Sr 0.22 0.10 60 66 0.67 67 77 0.47 69.57 0.50 5.25 3.95 6.42 6.00 Y 1.24 1.19 18 19 1.29 28 22 1.05 18.61 2.23 29 32 42 43 Zr 1.78 1.57 30 32 2.21 54 48 1.83 41.65 2.49 30 35 37 40 Nb 0.01 0.01 0.15 0.10 0.02 0.38 0.05 0.01 0.12 0.01 0.40 0.35 0.32 0.34 Ba 0.004 0.01 0.01 0.005 0.14 0.56 3.92 0.02 0.04 1.01 0.40 0.54 0.28 0.58 La 0.002 0.001 0.87 0.90 0.03 1.68 1.74 0.01 1.50 0.03 0.58 0.42 0.63 0.70 Ce 0.01 0.01 3.15 3.20 0.04 5.71 5.20 0.03 5.34 0.03 2.96 2.85 3.64 3.44 Pr 0.003 0.002 0.62 0.62 0.02 1.11 0.90 0.01 0.94 0.01 0.67 0.68 0.72 0.72 Nd 0.02 0.02 3.87 3.98 0.10 6.17 5.58 0.05 5.73 0.09 4.65 4.82 4.93 4.82 Sm 0.02 0.02 1.60 1.59 bdl 2.56 2.05 0.03 2.09 bdl 2.20 2.10 2.19 2.21 Eu 0.01 0.01 0.65 0.66 bdl 1.06 0.84 0.01 0.93 0.03 0.88 0.90 0.63 0.74 Gd 0.05 0.04 2.27 2.43 0.08 3.43 3.00 0.07 2.74 0.11 3.28 3.29 3.49 3.24 Tb 0.01 0.01 0.44 0.46 0.02 0.76 0.55 0.01 0.52 0.03 0.64 0.67 0.78 0.85 Dy 0.15 0.14 3.19 3.44 0.14 4.70 3.80 0.09 3.41 0.23 4.60 4.72 7.25 7.17 Ho 0.04 0.04 0.68 0.72 0.04 1.12 0.79 0.04 0.72 0.07 1.07 1.19 1.72 1.69 Er 0.18 0.18 2.04 2.08 0.19 3.36 2.41 0.15 2.00 0.36 3.19 3.68 4.97 5.02 Tm 0.04 0.04 0.28 0.32 0.04 0.43 0.34 0.03 0.27 0.07 0.50 0.56 0.76 0.79 Yb 0.29 0.33 1.91 2.14 0.39 3.08 2.25 0.24 2.08 0.58 3.51 3.86 5.60 5.47 89 Lu 0.05 0.06 0.27 0.29 0.09 0.45 0.39 0.04 0.29 0.12 0.52 0.55 0.82 0.79 Hf 0.06 0.05 0.89 1.00 0.10 1.61 1.24 0.08 1.21 0.10 0.69 0.92 0.91 1.27 Ta 0.02 0.01 0.02 0.02 0.02 0.09 0.04 0.01 0.03 0.02 0.04 0.02 0.05 0.03 Pb 0.09 0.08 0.05 0.05 bdl 0.08 bdl 0.01 0.06 0.55 0.23 0.22 0.15 0.26 Th 0.0002 0.0004 0.03 0.02 0.02 0.06 bdl 0.01 0.03 0.01 0.03 0.06 0.05 0.11 U 0.0003 0.0003 0.01 0.005 0.02 0.03 bdl 0.01 0.03 0.02 0.04 0.03 0.03 0.04 LaN/SmN 0.06 0.05 0.35 0.36 - 0.42 0.55 0.21 0.46 - 0.17 0.13 0.19 0.21 GdN/YbN 0.14 0.11 0.98 0.94 0.17 0.92 1.10 0.25 1.09 0.15 0.77 0.70 0.52 0.49 Eu/Eu* 0.83 0.92 1.04 1.03 - 1.09 1.03 0.94 1.19 - 1.00 1.05 0.69 0.85 bdl = below detection limits 96 4.3. Field relationships and sampling Samples were collected in the Ouassé Bay, in the Bogota Peninsula shear zone (Fig. 4.1). The Bogota Peninsula is located in the central-eastern part of the New Caledonia island and represents the north-western extension of the Massif du Sud. The Bogota Peninsula Shear Zone (BPSZ) is a 25-km-wide region of mantle harzburgites characterized by a strain gradient increasing towards two distinct 1 to 3 km-wide mylonitic zones (located in Ouassé and east of Kouaoua, respectively). Localization of deformation is marked by coarse granular textures ranging from protomylonite to mylonite, olivine grain size reduction, increasingly stretched orthopyroxene grains (aspect ratio up to 25:1) and presence of olivine-rich bands concordant to the foliation (Titus et al., 2011; Chatzaras et al., 2020). Strain increase also influenced fabric orientation, rotating strike foliation from NW-SE to NNE-SSW and steepening it from ~ 60° to ~ 90°. For comparison, the average foliation of the undeformed portion of the Massif du Sud body is ~ 20° E-W to NW-SE (Titus et al., 2011). Dextral sense of motion was attributed to the BPSZ and mylonite zones were regarded as the mantle areas corresponding to single faults in a transform system (Prinzhofer and Nicolas, 1980; Titus et al., 2011). Chatzaras et al. (2020) identified microdeformation zones (MDZs) in all deformed harzburgites. MDZs are µmto mm-wide and are made by recrystallized olivine or orthopyroxene locally dislocating grains of the primary mineral assemblage. MDZs were interpreted to record stress change through time, from seismic-related brittle deformation to ductile deformation and dynamic recrystallization. Interstitial pargasite modally increasing towards the mylonite zones suggests high-temperature fluids percolating the shear zone, likely favouring strain localization and mylonitization. Serpentinite shear zones concordant with the high-temperature foliation would indicate ocean water infiltration during cooling of the shear zone (Teyssier et al., 2016). Prinzhofer and Nicolas (1980) first reported the existence of several dikes in the BPSZ. Based on crosscutting relationships, the sequence of intrusion is (i) pyroxenite, (ii) feldspathic pyroxenite, (iii) hornblende gabbro, (iv) plagiogranite and (v) diabase, from the oldest to the youngest. A careful structural study of pyroxenite and diabase dikes was made by Titus et al. (2011). Some pyroxenitic bodies show constant thicknesses while others are necked or boudinaged. Locally, there are also folded layers. Pyroxenite foliation rotated and steepened with increasing strain, i.e. going towards the mylonite zone, and the dikes underwent greater stretching. Ultra-elongated boudins were locally dismembered leaving relict orthopyroxene blebs. Diabase dikes are mainly found in the mylonite zone. Due to their undeformed nature, they were considered as the last magmatic event. 97 In the Ouassé Bay, porphyroclastic to mylonitic harzburgites cut by pyroxenitic dikes crop out. Up to 1 cm-sized aligned and/or stretched orthopyroxene grains define the peridotite foliation. Tiny black spinel is also visible (Fig. 4.2a-b). Pyroxenites occur as 5-15 cm thick layers, mostly intersecting at high angle the harzburgite foliation and showing sharp contacts (Fig. 4.2c-d). Discordant low angle (Fig. 4.2e) and rare concordant (Fig. 4.2f), locally boudinaged, layers occur too. Our sampling focused on pyroxenites and the host harzburgites. Here, we report data for three websterites (two discordant high angle: OU1 and OU4; one discordant low angle: OU7B), two orthopyroxenites (one concordant: OU5A; one discordant high angle: OU8) and three harzburgites. 4.4. Methodology Major element compositions of minerals were analysed using a JEOL-6400 electron microprobe equipped with a LINK-ISIS energy dispersive microanalytical system at the Department of Chemistry, Life Sciences and Environmental Sustainability of Parma University. The electron beam was produced at an accelerating voltage of 15 kV and probe current of 0.25 nA. Both natural minerals and synthetic compounds were used as standards. Whole rock powders were prepared to analyse major and trace element compositions. Before making the powders, alteration was carefully removed from each rock sample using the diamond saw. Then, each rock sample was ground with a mortar until reaching a size less than 2 mm. The obtained granulate was reduced to powder (grain size < 2 µm) using an agate mill. The mortar, the sieves and the agate mill were cleaned with water and alcohol between one sample and another to avoid any contamination. Whole rock major and trace elements were analysed at Activation Laboratories (Ancaster, Ontario, Canada) by inductively coupled plasma (ICP) optical emission spectroscopy and ICP mass spectrometry. Precision and accuracy are estimated to be better than 2% for SiO2, Al2O3, Fe2O3 and MgO and better than 5% for the other major elements. Precision and accuracy of trace element analyses are assessed to be within 10%. In situ trace element analyses of mineral phases were carried out using laser ablation ICP mass spectrometry at Istituto di Geoscienze e Georisorse - C.N.R., Unità di Pavia. The laser probe consisted of a Q-switched Nd:YAG laser, model Quantel (Brilliant), whose fundamental emission in the nearIR region (1064 nm) was converted into 266 nm wavelength using three harmonic generators (Jeffries et al., 1998). Spot diameter was typically ~50 µm. The ablated material was analysed by using an Elan DRC-e quadrupole mass spectrometer. Helium was utilized as carrier gas and mixed with Ar downstream of the ablation cell. NIST 612 was used as external standard. The CaO content determined by electron microprobe was utilized as internal standard, scaled on the 44Ca+ signal. Precision and accuracy were assessed from repeated analyses of the BCR2 standard and resulted 98 better than 10% at ppm concentration level. Detection limits were typically in the range of 1.0-0.5 ppm for Cr and Ti, 0.5-0.1 ppm for Sc, 100-10 ppb for Sr, Zr, Ba, Rb, V and Gd, 10-1 ppb for Y, Nb, REE, Hf and Ta. Amphibole separates from two distinct amphibole-bearing websterites (OU1 and OU7B) were obtained in order to get 40Ar/39Ar data. After grinding and sieving of rock samples, amphiboles were separated from the 250/180 µm granulate fractions by handpicking. The 40Ar/39Ar analytical data are organized to comply with FAIR data reporting norms (see for instance Schaen et al., 2020). Each amphibole separate was irradiated at the Oregon State University for 14 hours in the CLICT position. Fish Canyon sanidine was used as the neutron flux monitor and has an assigned an age of 28.201 Ma (Kuiper et al., 2008) using a 40K decay constant of 5.463e-10/a (Min et al., 2000). After irradiation, the sample was step-heated with a diode laser with a heating time of 45 seconds per step followed by 45 seconds of gas clean up with a SAES GP-50 getter operated at 2 A. Gas was also exposed to a cold finger operated at ~-140°C. Argon isotopes were measured using a Thermo-Fisher Scientific Helix MC-plus multi-collector mass spectrometer. Isotopes 40Ar, 39Ar, 38Ar and 37Ar were measured on Faraday collectors with 40Ar and 37Ar using a 1012 Ohm resistor, 39Ar using a 1013 Ohm resistor, and 38Ar using a 1014 Ohm resistor. 36Ar was measured on a compact discrete dynode (CDD) ion counter that has a dead time of 20 ns. Due to the low K/Ca of the hornblende, CaF2 was analysed at the same time as the sample to help ensure accurate correction for production of 36Ar and 39Ar from irradiation of Ca. Calibration gases of air and a gas mixture enriched in radiogenic 40Ar along with 39Ar were analysed interspersed with the sample to monitor instrument drift and determine detector intercalibration factors. Extraction line blanks were analysed before and after data collection and are reported in the intensity data worksheet. All data collection was conducted with the in-house Pychron software and data reduction utilized MassSpec version 7.875. The reported plateau age is derived from the weighted mean of the chosen steps with the weighting factor being the inverse variance (e.g., Taylor, 1982) and the error is the square root of the sum of 1/σ2 values. The plateau age error is reported at 1σ and includes the J-factor error and irradiation correction factor uncertainties. 99 Fig. 4.2. Ouassé outcrop. (a,b) Harzburgites with rounded or stretched orthopyroxene porphyroclasts (red circles) and tiny black spinel grains (blue circle). (c,d) Discordant high angle pyroxenites The thin whitish veins are made of serpentine. (e) Discordant low angle pyroxenite. (f) Concordant pyroxenite. 100 4.5. Petrography and mineralogy Representative mineral major element compositions are listed in Table 4.1. 4.5.1. Harzburgites The studied harzburgites consists of 70-80 vol% olivine, 20-30 vol% orthopyroxene (enstatite) and accessory Cr-spinel. Textures range from porphyroclastic to mylonitic (Fig. 4.3a-b). The foliation is defined by aligned and/or stretched orthopyroxene, with aspect ratio up to 12:1 (Fig. 4.3c) and olivine porphyroclasts. Olivine and orthopyroxene porphyroclasts have refractory compositions (Fo = 91-92 mol% and Mg# = 91-92, respectively). Up to 1 cm-sized orthopyroxene porphyroclasts locally show undulous extinction and bending. Thin exsolution lamellae of clinopyroxene are also present. Orthopyroxene is characterised by relatively low Al2O3 contents (1.3-1.9 wt%; Fig. 4.6), slightly diminishing towards the rim (1.1-1.8 wt%). Cr2O3 and CaO contents show a broad decrease from core to rim (0.6-0.8 to 0.3-0.6 wt% and 0.6-1.1 to 0.4-0.7 wt%, respectively). Up to 1.5 mm dark brown to black spinel grains have anhedral to subhedral shapes and are locally organised in trails (Fig. 4.3d) concordant to the peridotite foliation. They have Cr# and Mg# ranging between 55-65 and 38-54 (Fig. 4.8), respectively, and low TiO2 contents (≤0.13 wt%). Porphyroclasts are set into a neoblastic matrix made of fine-grained (mostly 100-400 µm) olivine + orthopyroxene ± spinel displaying a polygonal texture. Neoblastic olivine has 91-93 mol% forsterite. Orthopyroxene neoblasts are mainly rounded crystals lacking exsolution lamellae. Overall, they display slightly lower Al2O3 (0.9-1.4 wt%) and Cr2O3 (0.3-0.7 wt%) contents compared to porphyroclasts (Fig. 4.6); CaO contents range between 0.5 and 1.1 wt%. Spinel shows Cr# values of 52-56 and Mg# values of 44-47 with TiO2 contents up to 0.1 wt%. Fine-grained (up to 200 µm) clinopyroxene (diopside) was rarely found, locally associated to neoblastic olivine in small embayments on porphyroclastic orthopyroxene rim. It displays (Fig. 4.7) very high Mg# values (9495) coupled to Cr2O3 contents ranging from 0.4 to 0.8 wt% and low Al2O3 contents (1.0-1.5 wt%). Tiny amphibole with pseudo-prismatic to rounded shape may be associated with the neoblastic assemblage. It has an edenitic composition (according to Leake et al., 1997) and Mg# values of 9192, high Cr2O3 (2.0-2.6 wt%) and low TiO2 contents (≤0.2 wt%). Up to 0.5 cm-thick veins made of talc ± calcite ± magnetite crosscut the harzburgite foliation. Close to the vein, harzburgitic orthopyroxenes may be replaced by the secondary mineral vein assemblage. Further away, talc + calcite occasionally occur as small veins cutting the orthopyroxene crystals and substitute the neoblastic aggregates. 101 4.5.2. Orthopyroxenites The orthopyroxenite samples are composed of ~95 vol% enstatitic orthopyroxene and accessory Cr-spinel, but differ for the minor components, as well as for the texture. Orthopyroxenite OU5A shows a disequigranular texture made of mediumto coarse-grained (up to 0.5 cm) orthopyroxenes set into a fine-grained granoblastic polygonal matrix (Fig. 4.4a). Orthopyroxenes have a stubby prismatic shape and bear very thin Ca-rich exsolution lamellae. They have irregular rims and embayments formed by the granoblastic polygonal assemblage (Fig. 4.4b). Locally, some crystals are truncated by the fine-grained assemblage. The latter is mainly composed of subhedral orthopyroxene grains, ~5 vol% clinopyroxene, frequently showing triple junction points (Fig. 4.4c) and accessory amphibole. Granoblastic pyroxenes do not exhibit exsolution lamellae. The orthopyroxene grains have high Mg# values (90-91; Fig. 4.6). Al2O3, Cr2O3 and CaO contents slightly diminish from core (1.8-2.1, 0.7-1.0 and 0.5-1.0 wt%, respectively) to rim (1.6-1.9, 0.5-1.0 and 0.40.6 wt%). Clinopyroxene is a diopside up to 100 µm in size. It is characterised by high Mg# values (93), low Al2O3 contents (2.2-2.5 wt%), Na2O and Cr2O3 contents of 0.4-0.5 and 1.0-1.5 wt%, respectively (Fig. 4.7). Accessory amphibole occurs as up to 100 µm in size crystals in the granoblastic assemblage, substituting clinopyroxene grains. It is a Cr-rich (Cr2O3 = 1.9-2.3 wt%) edenite showing nearly constant TiO2 and Na2O contents (~0.4 and ~1.9 wt%, respectively) and remarkably high Mg# values (92-93). Fine-grained (up to ~250 µm) dark brown to black chromian (Cr# = 57-61; Fig. 4.8) spinel occurs throughout the sample. It has cubic to anhedral habitus and relatively low TiO2 contents (~0.1 wt%) coupled to Mg# values of 45-48. The sample is crosscut by up to ~1.5 mm-sized veins of secondary talc + calcite ± magnetite. Sample OU8 is an olivine-bearing orthopyroxenite (Ol ~5 vol%), showing a mediumto coarsegrained cumulitic texture. Orthopyroxene mainly has a stubby shape (Fig. 4.4d) and is up to 1.5 cm in size. It is frequently highly fractured, but cleavages are still well identifiable. Orthopyroxene (Fig. 4.6) has high Mg# values (92), Al2O3, Cr2O3 and CaO contents comparable to those of the sample OU5a (1.5-1.9, 0.5-0.8 and 0.4-0.9 wt%, respectively). Olivine (Fig. 4.9; Fo = 92 mol%; NiO = 0.50.6 wt%) shows rounded to anhedral habitus and is up to 1.5 mm. Locally, it is included at the orthopyroxene rim (Fig. 4.4e). Spinel is a dark brown to black crystal up to 500 µm in size, displaying cubic, prismatic or anhedral shapes. Its TiO2 contents reach up to 0.1 wt%. Mg# values range from 56 to 64 and Cr# values cluster at ~52 Veins consisting of talc + calcite cut the orthopyroxenite. Locally, anhedral, elongated spinel is associated to the secondary assemblage. 4.5.3. Amphibole-bearing websterites 102 Enstatitic orthopyroxene (~ 25-70 vol%) and augitic to diopsidic clinopyroxene (~ 25 vol%) are the main constituents of the sampled websterites. Edenitic amphibole occurs in variable amounts (~ 5-40 vol%) and plagioclase is locally present (up to ~ 3 vol% in sample OU1). Spinel occurs as an accessory phase. In samples OU1 and OU4, the texture is disequigranular and is made of mediumto coarse-grained (up to ~ 2.5 mm) pyroxenes set into a fineto medium-grained granoblastic polygonal assemblage (Fig. 4.5a). Mediumto coarse-grained pyroxenes have prismatic to subhedral habitus and octagonal basal sections are locally observed. They locally show mutual thin exsolution lamellae and are characterised by numerous inclusions. The latter are mostly concentrated at the core region and mainly consist of rounded, tabular, or anhedral amphibole. Orthopyroxene and spinel (TiO2 = 0.6 wt%, Cr# = 0.7 and Mg# = 22) are minor inclusions. Clinopyroxene twins can also be easily found (Fig. 4.5b). The granoblastic polygonal matrix (down to ~ 100 µm) is constituted by orthopyroxene ± clinopyroxene ± amphibole ± plagioclase (Fig. 4.5c) and frequently shows triple junction points. Interlobate, irregular boundaries among crystals are locally observed. Recrystallized pyroxenes lack exsolution lamellae and are sometimes substituted by amphibole at their rims. Amphibole is green to light brown crystal showing a pseudo-prismatic to anhedral habitus. It mainly occurs as an interstitial phase and locally displays a poikilitic texture surrounding pyroxenes (Fig. 4.5d). Cubic spinel grains are rarely included (TiO2 = 2.3 wt%, Cr# = 66-73 and Mg# ~ 8). A websterite sample (OU7B) consists of coarse-grained (up to ~ 2 cm) subhedral clinopyroxene, orthopyroxene and amphibole displaying mutually interfingered crystals. They mostly share interlobate boundaries and mutual inclusions (Fig. 4.5e). Up to 1 mm amphibole and apatite grains are found associated to medium-grained pyroxenes collectively intersecting a cm-sized amphibole crystal. Apatite has a prismatic or rounded shape and is associated to amphibole (Fig. 4.5f). Rare FeNi sulphides occur as cubic crystals in amphibole. Edenitic amphibole and pyroxenes are commonly substituted by secondary tremolite (Al2O3 = 1.3-3.6 wt%, Mg# = 92-95), notably towards the contact with the host harzburgite. Locally, mafic minerals may be partially altered into a mixture of chlorite + serpentine + talc ± smectite. Websterite OU1 shows interstitial patches made of sericite. Mineral major element compositions are overall homogeneous among the samples and texturerelated chemical variations were not observed. Orthopyroxene has relatively low Al2O3 contents (0.61.2 wt%; Fig. 4.6). CaO contents range from 0.6 to 0.9 wt%. Cr2O3 contents (0.1-0.4 wt%) increase at decreasing Mg# values (78-89). Clinopyroxene shows relatively low Al2O3, Cr2O3 (Fig. 4.7) and Na2O contents (0.6-2.1, 0.1-0.3 and 0.1-0.4 wt%, respectively). Higher TiO2 contents (0.1-0.2 wt%) 103 are coupled to progressively lower Mg# values (81-90). Edenite largely displays increasing TiO2 contents (0.3-1.2 wt%) with increasing Al2O3 contents (7.5-11 wt%). Mg# values range from 78 to 88, whilst Na2O contents mainly cluster at ~1.5-2.0 wt%. Plagioclase has 82-86 mol% anorthite. Fig. 4.3. Thin section photomicrographs under cross-polarized light of Ouassé harzburgites. (a) Porphyroclastic texture. (b) Mylonitic texture with coarse-grained orthopyroxene porphyroclast associated with the neoblastic assemblage. (c) Highly stretched orthopyroxene porphyroclast, showing aspect ratio of 12:1. (d) Trails of spinel grains. Mineral abbreviations are after Whitney and Evans (2010). Ol Opx Opx Ol Spl 104 Fig. 4.4. Thin section photomicrographs under cross-polarized (a,b,d,e) and plane-polarized (c) light of Ouassé orthopyroxenites. (a,b) Mediumand coarse-grained orthopyroxenes set into a fine-grained granoblastic polygonal matrix in orthopyroxenite OU5A. (c) Granoblastic polygonal crystals showing triple junction points in orthopyroxenite OU5A. (d) Coarse-grained orthopyroxene in orthopyroxenite OU8. (e) Olivine crystal included at the orthopyroxene rim in orthopyroxenite OU8. Mineral abbreviations are after Whitney and Evans (2010). Opx Cpx Opx Cpx Opx Opx Cpx Opx Spl Opx Ol 105 Fig. 4.5. Thin section photomicrographs under cross-polarized (a,b,e,f) and plane-polarized (c,d) light of Ouassé amphibole-bearing websterites. (a) Mediumto coarse-grained pyroxenes set into a granoblastic polygonal assemblage in websterite OU4. (b) Clinopyroxene twinned crystal in websterite OU1. (c) Amphibole crystal associated with the granoblastic polygonal assemblage in websterite OU1. (d) Poikilitic amphibole surrounding pyroxenes in websterite OU1. (e) Texture of websterite OU7B. (f) Apatite crystal associated with amphibole in websterite OU7B. Mineral abbreviations are after Whitney and Evans (2010). Opx Cpx Cpx Amp Pl Cpx Opx Amp Amp Apt Amp Cpx Opx Apt 112 Fig. 4.12. Rare Earth Elements compositions of Ouassé pyroxenites normalized to chondrite (normalizing values after Sun and McDonough, 1989). Field for New Caledonia harzburgites includes data from Marchesi et al. (2009), Ulrich et al. (2010), Secchiari et al. (2020). Fig. 4.13. Extended incompatible trace element diagram for Ouassé pyroxenites normalized to Primitive Mantle (normalizing values after Sun and McDonough, 1989). Field for New Caledonia harzburgites includes data from Marchesi et al. (2009), Ulrich et al. (2010), Secchiari et al. (2020). 113 4.7.2. Amphibole-bearing websterites Clinopyroxene has REE contents ranging from ~2 to 10 times chondritic values. It shows variably depleted LREE (LaN/SmN = 0.1-0.6), weak to moderate negative Eu anomalies (Eu/Eu* = 0.7-0.9) and nearly flat HREE. In addition, negative anomalies are observed for Pb, Sr, Zr and Ti elements (Fig. 4.14). Orthopyroxene has very low REE contents in OU1 and OU4 websterites (~ 0.1-1 times chondritic values), whilst higher concentrations are displayed by sample OU7B (~1-4 times chondritic values). The orthopyroxene is characterized by increasing concentrations from LREE to HREE. Orthopyroxene from sample OU7B has a positive Eu anomaly (Eu/Eu* = 1.8). Amphibole displays high REE contents, up to 30 times chondritic values. It has a convex REE pattern characterized by a strong depletion in LREE (LaN/SmN = 0.2-0.7) and HREE (GdN/YbN = 1.92.0) and a very weak negative Eu anomaly (Eu/Eu* = 0.9). Amphibole from sample OU7B displays a positive LREE fractionation over HREE, Fig. 4.15). In Primitive Mantle normalized trace element diagram all amphiboles show negative peaks in Pb, Sr and Zr. The edenitic amphibole from the recrystallized matrix of harzburgite sample OU7A shows REE contents up to ~10 times chondritic values, characterized by a strong LREE enrichment (LaN/SmN = 1.6) and MREE positively fractionated over HREE (GdN/YbN = 1.3). It broadly mirrors the amphibole REE pattern of the host websterite OU7B, but at lower absolute concentrations. Plagioclase from pyroxenite OU1 has high LREE concentrations (mostly ~1-5 times chondritic values; Fig. 4.16), whilst Mand HREE contents are <1 times chondritic values. It is characterized by LREE enrichment (LaN/SmN = 1.2-4.4) and prominent positive Eu anomalies (Eu/Eu* = 3.3-11). Apatite from websterite OU7B has extremely high REE contents (~20 to ~1100 times chondritic values; Fig. 4.16), progressively decreasing from LREE to HREE. It shows a steep Lto MREE fractionation and nearly flat HREE. 114 Fig. 4.14. (a) Rare Earth Elements and (b) extended incompatible trace elements diagrams for clinopyroxenes from Ouassé pyroxenites normalized to chondrite and Primitive Mantle, respectively (normalizing values after Sun and McDonough, 1989). Data from other Ouassé websterites studied by Xu et al. (2021) are shown for comparison. b) a) 115 Fig. 4.15. (a) Rare Earth Elements and (b) extended incompatible trace elements diagrams for amphiboles from Ouassé pyroxenites normalized to chondrite and Primitive Mantle, respectively (normalizing values after Sun and McDonough, 1989). b) a) 116 Fig. 4.16. Rare Earth Elements compositions for plagioclase and apatite from Ouassé amphibolebearing websterites OU1 and OU7B, respectively. Chondritic normalizing values are after Sun and McDonough (1989). Pl Apt 117 4.8. Geothermometry Equilibrium temperatures calculated for host harzburgites and enclosed pyroxenite layers are listed in Table 4.4. Pressure conditions were assumed equal to 1 GPa, considering the spinel-facies conditions of the host harzburgites. The applied methods are the following: • two-pyroxene thermometers of Taylor (1998; referred to as TTa98) and Brey and Köhler (1990; referred to as TBK90), based on the Fe2+-Mg exchange between coexisting clinopyroxene and orthopyroxene • Ca-in-orthopyroxene thermometer of Brey and Köhler (1990; referred to as TCa-in-Opx), based on the Ca content of orthopyroxene • olivine-spinel thermometer of Jianping et al. (1995; referred to as TOl-Spl), based on the Fe2+-Mg exchange between coexisting olivine and spinel • Ca-in-olivine thermometer of De Hoog et al. (2010; referred to as TCa-in-Ol), based on the Ca content of olivine • amphibole-plagioclase thermometer of Holland and Blundy (1994; referred to as THB94), based on the compositions of coexisting amphibole and plagioclase • thermometers based on the partitioning of slowly diffusing elements (REE-Y), between (i) clinopyroxene and orthopyroxene (Liang et al., 2013; referred to as TREE-Px), and (ii) clinopyroxene and plagioclase (Sun and Liang, 2017; referred to as TREE-Cpx-Pl). The porphyroclastic assemblage of Ouassé harzburgites yielded Ca-in-Opx temperatures ranging from 920 to 970 °C at the orthopyroxene core. Remarkably lower values (730-750 °C) were obtained for olivine-spinel cores. Slightly lower temperatures were recorded by the neoblastic assemblage (TCain-Opx = 920-945 °C, TTa98 = 880-920 °C, TBK90 = 890-930 °C, TOl-Spl = 720 °C). Orthopyroxenite OU5A recorded Ca-in-Opx temperatures of 920-930 °C, TTa98 = 894 °C and TBK90 = 910 °C. Orthopyroxene cores from sample OU8 recorded the lowest Ca-in-Opx temperatures (870 °C). Olivine-spinel and Ca-in-olivine thermometers gave values of 865 and 861 °C, respectively, consistent with the Ca-in-Opx estimate. Thermometers based on slowly diffusing elements (REE, Y) were applied on the websteritic assemblages. Sample OU1 yielded values of 1251 ± 37 °C for the clinopyroxene-orthopyroxene pair and values of 1310 ± 14 °C were provided by clinopyroxene and plagioclase. Comparable temperatures were obtained from the core of the coarse clinopyroxene-orthopyroxene (1264 ± 50 °C) and the core of the granoblastic pyroxenes (1249 ± 52 °C) from sample OU4. Conventional geothermometry applied to pyroxene cores provided TCa-in-Opx = 965-995 °C, TTa98 = 930-970 °C and 118 TBK90 = 915-965 °C. These values are in the range of those from the host harzburgites. The amphiboleplagioclase geothermometer applied on sample OU1 gave consistent temperatures of 970 °C. 4.9. 40Ar/39Ar amphibole dating Two magmatic amphibole (edenite) separates from websterites OU1 and OU7B were analysed for 40Ar/39Ar dating and the results are reported in Table 4.5 and Figure 4.17. The two samples show low amounts of excess argon (40Ar/36Ar = ~295-297) and display good concordances between the plateau ages and the isochron ages (Tab. 4.5). Well-defined plateau ages can be seen in Figure 4.17. The plateau ages of the two samples (OU1 = 56.01 ± 0.19 Ma; OU7B = 56.05 ± 0.17 Ma) are nearly undistinguishable and cluster at 56 Ma. 4.10. Discussion 4.10.1. Nature and evolution of hosting peridotites Overall, major element mineral and whole rock compositions of the Ouassé harzburgites are in the range of the other New Caledonia mantle harzburgites which were extensively studied in previous works (Marchesi et al., 2009; Ulrich et al., 2010; Secchiari et al., 2020). In particular, according to Secchiari et al. (2020), the harzburgites are ultra-refractory peridotites recording a two-stage melting process including an anhydrous melting event which possibly occurred in a marginal basin and a subsequent fluid-assisted melting in a forearc environment. Post-melting reactive percolation of depleted melts led to crystallization of secondary pyroxenes (e.g., Secchiari et al., 2020) and produced variable enrichments in several incompatible trace elements (Land MREE, LILE and some HFSE, see also Xu et al., 2021a). The highly refractory nature of Ouassé harzburgites is supported by (i) the absence of primary clinopyroxene, (ii) the very low bulk rock Al2O3 (0.4-0.5 wt%) and CaO (0.4-0.5 wt%) contents coupled to high MgO contents (46-47 wt%), (iii) the high Cr# values (55-65) coupled to low TiO2 contents (≤0.1 wt%) in spinels, (iv) the high forsterite content in olivine (Fo = 91-92 mol%) and (v) the high Mg# values (91-92) of orthopyroxene porphyroclasts associated with low Al2O3 contents (1.1-1.9 wt%). 119 OU1 Hornblende Fig. 4.17. 40Ar/39Ar age spectra for two amphibole separates from Ouassé websterites OU1 and OU7B. Fig. 4.18. Zr/Nb vs. Ti/Nb of amphiboles from Ouassé samples. Fields of suprasubduction and intraplate amphibole are after Coltorti et al. (2007). 120 Remarkably, the Ouassé harzburgites were affected by intense deformation, that produced mylonitic textures and highly stretched orthopyroxene porphyroclasts (Fig. 4.3c). Deformation in the study area is linked to the Bogota Peninsula Shear Zone, one of the three regional-scale high temperature shear zones of the Peridotite Nappe (see §Geological and petrological setting). Deformation was accompanied by recrystallization and grain size reduction, as testified by the finegrained (~ 100-400 µm) neoblastic assemblage (ol + opx ± spl). Secondary clinopyroxene crystallized among neoblasts, suggesting deformation synchronous to percolation of melt or fluids capable to precipitate clinopyroxene. Amphibole of edenitic composition was also locally found within the neoblastic assemblage. The presence of amphibole relatively enriched in incompatible elements (Table 4.3, Fig. 4.15) in a highly depleted harzburgite points to a metasomatic origin. Its Ti/Nb and Zr/Nb ratios (Fig. 4.18) are coherent with an origin in a supra-subduction zone (Coltorti et al., 2007). Teyssier et al. (2016) also reported the presence of fine-grained interstitial pargasite in Bogota mylonites. They proposed that the high temperature amphibole originated from percolating fluids initially derived from the subduction context and subsequently related to ocean water infiltration. Ca-in-orthopyroxene equilibration temperatures of neoblastic pyroxenes range from 920 to 945 °C. These estimates suggest fairly high temperature conditions for the deformation. They are comparable to those calculated for the porphyroclastic orthopyroxene (TCa-in-Opx = 890-970 °C), in agreement with the absence of significant differences in major element compositions between porphyroclasts and neoblasts. These temperatures fall in the range of those reported by other authors (Teyssier et al., 2016; Chatzaras et al., 2020) for the Bogota Peninsula Shear Zone. Significantly lower temperatures were obtained for neoblastic (720 °C) and porphyroclastic (730-750 °C) olivinespinel pairs (Jianping et al., 1995). Considering that the olivine-spinel thermometer is based on faster element diffusion compared to the Ca-in-orthopyroxene, the remarkable temperature difference detected by the two geothermometers suggest a relatively slow cooling and slow thermal equilibration of the studied mantle sector. 4.10.2. Origin of the pyroxenite layers Whole rocks and minerals chemistry of Ouassé orthopyroxenites and amphibole-bearing websterites are consistent with the compositions of supra-subduction zone pyroxenites reported in the literature (see Figs. 4.6,4.7,4.8,4.9,4.10,4.11). 4.10.2.1. Orthopyroxenites The presence of orthopyroxenite layers in mantle suites is generally ascribed to percolating silicarich melts. Pyroxenes from Ouassé orthopyroxenites are characterized by very low Al2O3 contents 121 (1.5-2.5 wt%) coupled to very high Mg# values (90-93), suggesting an origin from a supra-subduction agent, as reported, for instance, by Müntener et al. (2001), Morishita et al. (2003) and Arai et al. (2006) for rocks crystallized from hydrous silica-rich melts showing high Mg# andesite to boniniteaffinity. The presence of clinopyroxene in orthopyroxenite OU5A allowed the calculation of parental liquids in equilibrium with this mineral. Trace element analysis performed on clinopyroxene and cpx/liquid partition coefficients from the compilation of Hart and Dunn (1993) returned the composition of the equilibrium melt (Tab. 4.6). Its REE pattern (Fig. 4.19a) shows contents up to ~ 20 times chondrite values, with LREE positively fractionated over MREE (LaN/SmN = 3.2) and nearly flat HREE (GdN/YbN = 0.9). Positive Ba, Pb and Sr anomalies and negative Ti anomaly are displayed in a N-MORB normalized extended trace element diagram (Fig. 4.19b). Enrichments in LREE, LILE (Ba, Sr) and Pb are typical of slab-derived hydrous silicate melts, rather than aqueous fluids (Zheng, 2019). Mg# of the equilibrium melt (computed according to the equation of Wood and Blundy, 1997) is equal to 80. The high Mg# suggest an origin from a refractory source (e.g., Pearce, 1982). According to this, both pyroxenes have refractory compositions with high Mg# values (90-93). Boninite-like liquids could be a good candidate for a silica-rich melt in equilibrium with such a depleted source (Pearce and Reagan, 2019), as the New Caledonia Peridotite Nappe is crosscut by dikes with boninite affinity (Cluzel et al., 2006; Xu et al., 2021a) and boninite rocks were emplaced at its base (Cluzel et al., 2016). Fig. 4.19 shows a good match between the calculated liquid representing the parental melt of orthopyroxenite OU5A and the field of New Caledonia boninites, as well as the pattern of pre-obduction boninite-like dikes. Boninitic rocks from New Caledonia were interpreted as the product of low degree partial melting of a depleted peridotite source previously enriched by slab-derived felsic melts (Cluzel et al., 2016). The small percentage of clinopyroxene (~ 5 vol%) in sample OU5A likely reflects the Ca-poor composition of the whole rock which could be explained by the low-Ca nature of New Caledonia boninites (Cluzel et al., 2016). Textural evidence show that the amphibole grew at the expense of clinopyroxene. Its geochemical features (e.g. Ti/Nb and Zr/Nb ratios Fig. 4.18) are consistent with crystallization from a subduction-related agent (Coltorti et al., 2007). The boninite-like melt was injected in the deforming harzburgite, as evidenced by the orientation of the orthopyroxenite dike concordant to the peridotite foliation. In addition, equilibration temperatures gave values consistent with those of the host harzburgite (TCa-in-Opx = 920930 °C). The orthopyroxenite OU8 has peculiar features, as (i) the high Ni contents (2890 ppm), which exceeds the values displayed by all the New Caledonia harzburgites and (ii) the extremely low REE contents (< 0.1 times chondritic values). Most of the elements appearing in the bulk rock extended 128 these rocks and further field surveys could enhance the studies in remote areas. Upper arc crust is missing (e.g., Cluzel et al., 2012b), thus rendering difficult to determine the existence of a FAB crust. 4.10.5. Hints for reinterpreting the significance of the Bogota Peninsula Shear Zone The Bogota Peninsula Shear Zone (BPSZ) was generally considered as an oceanic paleotransform fault characterized by a dextral sense of motion (Prinzhofer and Nicolas, 1980; Titus et al., 2011; Chatzaras et al., 2020). The new data on Ouassé mantle rocks could provide new constraints on the BPSZ origin. Ouassé mylonitic harzburgites developed secondary clinopyroxene and amphibole in the neoblastic assemblage, suggesting mylonitization concurrent to infiltration of slab-derived fluids/melts. In this view, the BPSZ was established in a supra-subduction environment. Accordingly, the subduction-related melt forming orthopyroxenite OU5A was injected in the harzburgites during deformation, as the pyroxenite layer is concordant with the host peridotites foliation. As discussed in the previous section, the subsequent intrusion of amphibole-bearing websterites is connected to a brittle behaviour of the mantle. Different subduction-related scenarios could be invoked to explain the origin of the Bogota Peninsula Shear Zone: • It could represent a transform fault related to a short-lived spreading center developed in the forearc region shortly after subduction initiation. A similar mechanism of seafloor spreading was hypothesised by Reagan et al. (2019) for the Izu-Bonin-Mariana system • New Caledonia Paleocene-Eocene subduction inception is thought to occur near a spreading ridge and oblique subduction has been envisaged (Cluzel et al., 2021). The resulting transpressive forces could have deformed the hot and buoyant lithospheric mantle, thus creating the BPSZ • The BPSZ could represent an old (Late Cretaceous-Paleocene) transform fault or at least a weak lithospheric mantle region of the South Loyalty Basin reactivated during the beginning of the Paleocene-Eocene subduction Some considerations can be made in order to give at least some preferences over the above options. First, mylonitization occurred after partial melting of peridotites, not prior as partial melting could influence or hide deformation textures. Since Ouassé peridotites, like the other New Caledonia harzburgites, are interpreted to record subduction-related fluid-assisted partial melting, deformation is presumed to have occurred after this event and thus the third option can be discarded, as it would include a pre-existing shear zone preceding the subduction-related partial melting event. The first two 129 options seem to be more plausible and additional data are needed to better interpret the origin of the BPSZ. 4.11. Conclusions The Ouassé mantle section shows a rare variety of subduction-related pyroxenites crosscutting the host harzburgites. In particular, orthopyroxenites and amphibole-bearing websterites were recognized. Each studied sample displays peculiar mineralogical characteristics, whereby, for instance, Ol-bearing and Cpx-bearing orthopyroxenites were distinguished. In addition, the websteritic rocks have variable amounts of primary amphibole (~5-40 vol%). From a chemical point of view, the orthopyroxenites show more refractory compositions compared to the amphibole-bearing websterites. Beyond the characterization of the pyroxenitic samples, the main achievements of this PhD Thesis were (i) the identification of their parental liquids and (ii) their significance in the overview of the Eocene subduction. The main conclusions are: • Equilibrium melts of the orthopyroxenites have a boninite-affinity. • Equilibrium melts of the amphibole-bearing websterites are more enriched liquids compared to those of the orthopyroxenites. A derivation from enriched boninitic liquids and/or from adakite-like liquids is plausible. • The 40Ar/39Ar cooling age of ~56 Ma referred to the websterites suggest that Ouassé pyroxenites represent one of the earliest magmatic products of melt percolation in the New Caledonia forearc mantle reported in the literature. The results also allowed to give new insights on the significance of the Bogota Peninsula Shear Zone (BPSZ). In particular, the concordant orthopyroxenite layer found in the mylonitic harzburgites indicates the injection of boninitic liquids concurrent to the activity of the BPSZ. As the boninite-like liquids are typically subduction-related products, the shear zone should have been developed in a supra-subduction setting. Thus, it cannot be considered as a paleotransform fault merely related to a normal oceanic spreading ridge, as proposed by Prinzhofer and Nicolas (1980), Titus et al. (2011) and Chatzaras et al. (2020). 4.12. References Arai S., 1994. Characterization of spinel peridotites by olivine-spinel compositional relationships: review and interpretation. Chemical Geology, 113:191-204 130 Arai S., Shimizu Y., Morishita T., Ishida Y., 2006. A new type of orthopyroxenite xenolith from Takashima, Southwest Japan: silica enrichment of the mantle by evolved alkali basalt. Contributions to Mineralogy and Petrology, 152:387-398 Bénard A., Nebel O., Ionov D.A., Arculus R.J., Shimizu N., Métrich N., 2016. Primary Silica-rich Picrite and High-Ca Boninite Melt Inclusions in Pyroxenite Veins from the Kamchatka Sub-arc Mantle. Journal of Petrology, 57:1955-1982 Berly T.J., Hermann J., Arculus R.J., Lapierre H., 2006. Supra-subduction Zone Pyroxenites from San Jorge and Santa Isabel (Solomon Islands). Journal of Petrology, 47:1531-1555 Brey G.P., Köhler T., 1990. Geothermobarometry in four-phase lherzolites II. New thermobarometers, and practical assessment of existing thermobarometers. Journal of Petrology, 31:1353-1378 Carroll M.R., Wyllie P.J., 1989. Experimental phase relations in the system tonalite-peridotite-H2O at 15 kb; implications for assimilation and differentiation processes near the crust-mantle boundary. Journal of Petrology, 30:1351-1382 Chatzaras V., Tikoff B., Kruckenderg S.C., Titus S.J., Teyssier C., Drury M.R., 2020. Stress variations in space and time within the mantle section of an oceanic transform zone: Evidence for the seismic cycle. Geology, 48:569-573 Choi S.H., Shervais J.W., Mukasa S.B., 2008. Supra-subduction and abyssal mantle peridotites of the Coast Range ophiolite, California. Contributions to Mineralogy and Petrology, 156:551-576 Cluzel D., 2020. Subduction erosion: contributions of footwall and hanging wall to serpentinite mélange; field, geochemical and radiochronological evidence from the Eocene HP-LT belt of New Caledonia. Australian Journal of Earth Sciences, 68:99-119 Cluzel D., Aitchison J.C., Picard C., 2001. Tectonic accretion and underplating of mafic terranes in the Late Eocene intraoceanic fore-arc of New Caledonia (Southwest Pacific): geodynamic implications. Tectonophysics, 340:23-59 Cluzel D., Boulvais P., Iseppi M., Lahondère D., Lesimple S., Maurizot P., Paquette J.-L., Tarantola A., Ulrich M., 2020. Slab‑derived origin of tremolite–antigorite veins in a supra‑subduction ophiolite: the Peridotite Nappe (New Caledonia) as a case study. International Journal of Earth Sciences, 109:171-196 131 Cluzel D., Iseppi M., Chen Y., 2021. Eocene preand syn-obduction tectonics in New Caledonia (Southwest Pacific), a case for oblique subduction, transcurrent tectonics and oroclinal bending; structural and paleomagnetic evidence. Tectonophysics, 811 Cluzel D., Jourdan F., Meffre S., Maurizot P., Lesimple S., 2012a. The metamorphic sole of New Caledonia ophiolite: 40Ar/39Ar, U-Pb, and geochemical evidence for subduction inception at a spreading ridge. Tectonics, 31 Cluzel D., Maurizot P., Collot J., Sevin B., 2012b. An outline of the geology of New Caledonia; from Permian-Mesozoic southeast Gondwanaland active margin to Cenozoic obduction and supergene evolution. Episodes, 35:72-86 Cluzel D., Meffre S., Maurizot P., Crawford A.J., 2006. Earliest Eocene (53 Ma) convergence in the Southwest Pacific; evidence from pre-obduction dikes in the ophiolite of New Caledonia. Terra Nova, 18:395-402 Cluzel D., Ulrich M., Jourdan F., Meffre S., Paquette J.-L., Audet M.-A., Secchiari A., Maurizot P., 2016. Early Eocene clinoenstatite boninite and boninite-series dikes of the ophiolite of New Caledonia; a witness of slab-derived enrichment of the mantle wedge in a nascent volcanic arc. Lithos, 260:429-442 Coltorti M., Bonadiman C., Faccini B., Grégoire M., O’Reilly S.Y. and Powell W., 2007. Amphiboles from suprasubduction and intraplate lithospheric mantle. Lithos, 99:68-84 De Hoog J.C.M., Gall L., Cornell D.H., 2010. Trace-element geochemistry of mantle olivine and application to mantle petrogenesis and geothermobarometry. Chemical Geology, 270:196-215 Defant M.J., Drummond M.S., 1990. Derivation of some modern arc magmas by melting of young subducted lithosphere. Nature, 347:662-665 Dick H.J.B., Bullen T., 1984. Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86:54-76 Eissen J.-P., Crawford A.J., Cotten J., Meffre S., Bellon H., Delaune M., 1998. Geochemistry and tectonic significance of basalts in the Poya Terrane, New Caledonia. Tectonophysics, 284:203219 Eslami A., Borghini G., Montanini A., Grieco G., Marchesi C., 2021. Petrological constraints on the origin of pyroxenite dykes in the lithospheric mantle of the Cheshmeh-Bid ophiolitic massif, Southern Iran. Ofioliti, 46(1):63-81 132 Falloon T.J., Danyushevsky L. V., Crawford T. J., Maas R., Woodhead J. D., Eggins S. M., Bloomer S. H., Wright D. J., Zlobin S. K., Stacey A. R., 2007. Multiple mantle plume components involved in the petrogenesis of subduction-related lavas from the northern termination of the Tonga Arc and northern Lau Basin: Evidence from the geochemistry of arc and backarc submarine volcanics. Geochemistry, Geophysics and Geosystems, 8:1-45 Ferré E.C., Belley F., Tikoff B., Martín-Hernández F., Nzokwe G., Ward C., 2004. Anatomy of an oceanic mantle shear zone deduced from high-field magnetic anisotropy: the Humboldt corridor, New Caledonia. Eos Trans. AGU 85 (47). Fall Meeting Supplement, Abstract GP23B-04 Foden J., Elburg M., Turner S., Clark C., Blades M.L., Cox G., Collins A.S., Wolff K., George C., 2020. Cambro-Ordovician magmatism in the Delamerian orogeny: Implications for tectonic development of the southern Gondwanan margin. Gondwana Research, 81:490-521 Hart S.R., Dunn T., 1993. Experimental cpx/melt partitioning of 24 trace elements. Contributions to Mineralogy and Petrology, 113:1-8 Hermann J., Spandler C., Hack A., Korsakov A.V., 2006. Aqueous fluids and hydrous melts in highpressure rocks: Implications for element transfer in subduction zone. Lithos, 92:399-417 Holland T., Blundy J., 1994. Non-ideal interactions in calcic amphiboles and their bearing on amphibole-plagioclase thermometry. Contributions to Mineralogy and Petrology, 116:433-447 Jeffries T.E., Jackson S.E., Longerich H.P., 1998. Application of a frequency quintupled Nd:YAG source (λ=213 nm) for laser ablation inductively coupled plasma mass spectrometric analysis of minerals. Journal of Analytical Atomic Spectrometry, 13:935-940 Jianping L., Kornprobst J., Vielzeuf D., 1995. An improved experimental calibration of the olivinespinel geothermometer. Chinese Journal of Geochemistry, 14:68-77 Kelemen P.B., 1995. Genesis of high Mg# andesites and the continental crust. Contributions to Mineralogy and Petrology, 120:1-19 Kuiper K.F., Deino A., Hilgen F.J., Krijgsman W., Renne P.R., Wijbrans J.R., 2008. Synchronizing the rock clocks of Earth history. Science, 320:500-504 Leake B.E., Woolley A.R., Arps C.E., Birch W.D., Gilbert M.C., Grice J.D., Hawthorne F.C., Kato A., Kisch H.J., Krivovichev V.G., Linthout K., Laird J., Mandarino J., Maresch W.V., Nickel E.H., Tock N.M.S., Schumacher J.C., Smith D.C., Stephenson N.C.N., Ungaretti L., Whittaker E.J.W., Youzhi G., 1997. Nomenclature of amphiboles; report of the Subcommittee on 133 Amphiboles of the International Mineralogical Association Commission on new minerals and mineral names. Mineralogical Magazine, 61 (405):295-310 Li H., Arculus R.J., Ishizuka O., Hicjey-Vargas R., Yogodzinski G.M., McCarthy A., Kusano Y., Brandl P.A., Savov I.P., Tepley F.J., Sun W., 2021. Basalt derived from highly refractory mantle sources during early Izu-Bonin-Mariana arc development. Nature Communications, 12:1723 Liang Y., Sun C., Yao L., 2013. A REE-in-two-pyroxene thermometer for mafic and ultramafic rocks. Geochimica et Cosmochimica Acta, 102:246-260 Lin K.-Y., Wang K.-L., Chung S.-L., Bingöl A.F., Iizuka Y., Lee H.Y., 2020. Tracking the magmatic response to subduction initiation in the forearc mantle wedge: Insights from peridotite geochemistry of the Guleman and Kızıldağ ophiolites, Southeastern Turkey. Lithos, 376-377 Manning C.E., 2004. The chemistry of subduction zone fluids. Earth and Planetary Science Letters, 223:1-16 Marchesi C., Garrido C.J., Godard M., Belley F., Ferré E., 2009. Migration and accumulation of ultradepleted subduction-related melts in the Massif du Sud ophiolite (New Caledonia). Chemical Geology, 266:171-186 Min K., Mundil R., Renne P.R., Ludwig K.R., 2000. A test for systematic errors in 40Ar/39Ar geochronology through comparison with U–Pb analysis of a 1.1 Ga rhyolite. Geochimica et Cosmochimica Acta, 64:73-98 Morishita T., Arai S., Green D.H., 2003. Evolution of low-Al orthopyroxene in the Horoman peridotite, Japan: an unusual indicator of metasomatzing fluids. Journal of Petrology, 44:12371246 Müntener O., Kelemen P.B., Grove T.L., 2001. The role of H2O during crystallization of primitive arc magmas under uppermost mantle conditions and genesis of igneous pyroxenites: an experimental study. Contributions to Mineralogy and Petrology, 141:643-658 Nekvasil H., Dondolini A., Horn J., Filiberto J., Long H., Lindsley D.H., 2004. The Origin and Evolution of Silica-saturated Alkalic Suites: an Experimental Study. Journal of Petrology, 45:693-721 Nicolas A., 1989. Structure of Ophiolites and Dynamics of Oceanic Lithosphere. Kluwer, Dordrecht, p. 367 134 Pearce J.A., 1982. Trace element characteristics of lavas from destructive plate boundaries. In: Thorpe, R.S. ed., Orogenic andesites and related rocks, Chichester, England: John Wiley and Sons, 528-548 Pearce J.A., Lippard S.J., Roberts S., 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites. Geological Society, London, Special Publications, 16:77-94 Pearce J.A., Reagan M.K., 2019. Identification, classification, and interpretation of boninites from Anthropocene to Eoarchean using Si-Mg-Ti systematics. Geosphere, 15:1008-1037 Pirard C., Hermann J. and O’Neill H.ST.C., 2013. Petrology and Geochemistry of the Crust-Mantle Boundary in a Nascent Arc, Massif du Sud Ophiolite, New Caledonia, SW Pacific. Journal of Petrology, 54:1759-1792 Price A.A., Jackson M.G., Blichert-Toft J., Blusztajn J., Conatser C.S., Konter J.G., Koppers A.A.P., Kurz M.D., 2016. Geochemical evidence in the northeast Lau Basin for subduction of the CookAustral volcanic chain in the Tonga Trench. Geochemistry, Geophysics, Geosystems, 17:16941724 Prinzhofer A., Nicolas A., 1980. The Bogota peninsula, New Caledonia: a possible oceanic transform fault. The Journal of Geology, 88:387-398 Prouteau G., Scaillet B., Pichavant M., Maury R., 2001. Evidence for mantle metasomatism by hydrous silicic melts derived from subducted oceanic crust. Nature, 410:197-200 Reagan M.K., Heaton D.E., Schmitz M.D., Pearce J.A., Shervais J.W., Koppers A.A.P., 2019. Forearc ages reveal extensive short-lived and rapid seafloor spreading following subduction initiation. Earth and Planetary Science Letters, 506:520-529 Rogkala A., Petrounias P., Tsikouras B., Hatzipanagiotou K., 2017. New Occurrence of Pyroxenites in the Veria-Naousa Ophiolite (North Greece): Implications on Their Origin and Petrogenetic Evolution. Geosciences, 7,92 Schaen A.J., Jicha B.R., Hodges K.V., Vermeesch P., Stelten M.E., Mercer C.M., Phillips D., Rivera T.A., Jourdan F., Matchan E.L., Hemming S.R., Morgan L.E., Kelley S.P., Cassata W.S., Heizler M.T., Vasconcelos P.M., Koppers A.A.P., Mark D.F., Niespolo E.M., Sprain,C.J., Benowitz J.A., Hames W.E., Kuiper K.F., Turrin B.D., Renne P.R., Ross J., Nomade S., Guillou H., Laura E., Webb L.E., Cohen B.A., Calvert A.T., Joyce N., Morgan Ganderød M., Wijbrans J., Ishizuka, O., He H., Ramirez A., Pfänder J.A., Lopez-Martínez M., Huaning Qiu H., Brad S. Singer B.S., 2020. 135 On the reporting and interpretation of 40Ar/39Ar geochronologic data, Geological Society America Bullettin, 133(3-4):461-487 Schiano P., Clocchiatti R., Shimizu N., Maury R.C., Jochum K.P., Hofmann A.W., 1995. Hydrous silica-rich melts in the sub-arc mantle and their relationships with erupted arc lavas. Nature, 377:595-600 Secchiari A., Montanini A., Bosch D., Macera P., Cluzel D., 2016. Melt extraction and enrichment processes in the New Caledonia lherzolites: Evidence from geochemical and Sr–Nd isotope data. Lithos, 260:28-43 Secchiari A., Montanini A., Bosch D., Macera P., Cluzel D., 2018. The contrasting geochemical message from the New Caledonia gabbronorites: insights on depletion and contamination processes of the sub-arc mantle in a nascent arc setting. Contributions to Mineralogy and Petrology, 173:66 Secchiari A., Montanini A., Bosch D., Macera P., Cluzel D., 2020. Sr, Nd, Pb and trace element systematics of the New Caledonia harzburgites: Tracking source depletion and contamination processes in a SSZ setting. Geoscience Frontiers, 11:37-55 Soret M., Agard P., Dubacq B., Vitale-Brovarone A., Monié P., Chauvet A., Whitechurch H., Villemant B., 2016. Strain localization and fluid infiltration in the mantle wedge during subduction initiation: Evidence from the base of the New Caledonia ophiolite. Lithos, 244:1-19 Steiger R.H., Jäger E., 1977. Subcommission on geochronology: Convention on the use of decay constants in geoand cosmochronology. Earth and Planetetary Science Letters, 36:359-362 Stern R.J., 2002. Subduction zones. Reviews of Geophysics, 40:1-42 Stern C.R., Kilian R., 1996. Role of the subducted slab, mantle wedge and continental crust in the generation of adakites from the Andean Austral Volcanic Zone. Contributions to Mineralogy and Petrology, 123:263-281 Sun C., Liang Y., 2017. A REE-in-plagioclase-clinopyroxene thermometer for crustal rocks. Contributions to Mineralogy and Petrology, 172:24 Tamura A., Arai S., 2006. Harzburgite–dunite–orthopyroxenite suite as a record of supra-subduction zone setting for the Oman ophiolite mantle. Lithos, 90:43-56 136 Taylor J.R., 1982. An Introduction to Error Analysis: The Study of Uncertainties in Physical Measurements. University Science Books, Mill Valley, California, 270 p Taylor W.R., 1998. An experimental test of some geothermometer and geobarometer formulations for upper mantle peridotites with application to the thermobarometry of fertile lherzolite and garnet websterite. Neues Jb Miner Abh, 172: 381-408 Teyssier C., Chatzaras V., Von Der Handt A., 2016. Microfabrics in depleted mantle plaeotransform (New Caledonia). Geophysical Research Abstracts, EGU2016-11489, 18 Tilhac R., Ceuleneer G., Griffin W.L., O’Reilly S.Y., Pearson N.J., Benoit M., Henry H., Girardeau J., Grégoire M., 2016. Primitive Arc Magmatism and Delamination: Petrology and Geochemistry of Pyroxenites from the Cabo Ortegal Complex, Spain. Journal of Petrology, 57:1921-1954 Titus S.J., Maes S.M., Benford B., Ferré E.C., Tikoff B., 2011. Fabric development in the mantle section of a paleotransform fault and its effect on ophiolite obduction, New Caledonia. Lithosphere, 3:221-244 Ulrich M., Picard C., Guillot S., Chauvel C., Cluzel D., Meffre S., 2010. Multiple melting stages and refertilization as indicators for ridge to subduction formation: The New Caledonia ophiolite. Lithos, 115:223-236 van de Lagemaat S.H.A., van Hinsbergen D.J.J., Boschman L.M., Kamp P.J.J., Spakman W., 2018. Southwest Pacific Absolute Plate Kinematic Reconstruction Reveals Major Cenozoic TongaKermadec Slab Dragging. Tectonics, 37:2647-2674 Varfalvy V., Hébert R., Bédard J.H., Laflèche M.R., 1997. Petrology and geochemistry of pyroxenite dykes in upper mantle peridotites of the North Arm mountain massif, Bay of Islands ophiolite, Newfoundland: implications for the genesis of boninitic and related magmas. The Canadian Mineralogist, 35:543-570 Villa I.M., Grobéty B., Kelley S.P., Trigila R., Wieler R., 1996. Assessing Ar transport paths and mechanisms for McClure Mountains Hornblende. Contributions to Mineralogy and Petrology, 126:67-80 Vogt J., Podvin P., 1983. Carte Géologique à l’échelle du 1 / 50 000 et notice explicative: feuille Humboldt-Port-Bouquet. Territoire de Nouvelle-Calédonie-Bureau de Recherches Géologiques et Minières, 1-68 137 Whattam S.A., Stern R.J., 2011. The ‘subduction initiation rule’: a key for linking ophiolites, intraoceanic forearcs, and subduction initiation. Contributions to Mineralogy and Petrology, 162:1031-1045 Whitney D.L., Evans B.W., 2010. Abbreviations for names of rock-forming minerals. American Mineralogist, 95:185-187 Wood B.J., Blundy J.D., 1997. A predictive model for rare earth element partitioning between clinopyroxene and anhydrous silicate melt. Contributions to Mineralogy and Petrology, 129:166181 Xu Y., Liu C.-Z., Lin W. 2021a. Melt extraction and reaction in the forearc mantle: Constraints from trace elements and isotope geochemistry of ultra-refractory peridotites of the New Caledonia Peridotite Nappe. Lithos, 380-381 Xu Y., Liu C.-Z., Shi X.-.F., Lin W., 2021b. Petrogenesis of Eocene mafic and felsic magmas in the New Caledonia ophiolite: geochemistry and geochronology constraints. International Geology Review Zhao J.-H., Asimow P.D., 2014. Neoproterozoic boninite-series rocks in South China: A depleted mantle source modified by sediment-derived melt. Chemical Geology, 388:98-111 Zheng Y.-F., 2019. Subduction zone geochemistry. Geoscience Frontiers, 10:1223-1254 144 Table 4.3. Representative trace element compositions (ppm) of minerals from Ouassé pyroxenites. Sample OU1 OU4 OU5A OU7B OU8 Rock type Amp-websterite Amp-websterite Orthopyroxenite Amp-websterite Orthopyroxenite Mineral Opx Cpx Amp Pl Opx Cpx Amp Opx Cpx Amp Opx Cpx Amp Apt Opx Ol Sc 27 90 115 3.58 23 89 125 20 62 71 6.33 68 73 16 14 1.88 Ti 649 1168 8711 41 442 904 7592 203 465 1781 144 638 6323 872 17 3.80 V 72 357 588 0.88 64 198 587 94 216 398 19 257 461 33 87 1.03 Cr 1007 706 1846 6.78 1329 1271 4216 5792 7729 11484 41 539 1167 148 4681 38 Co 101 39 65 0.45 92 37 69 68 36 49 70 32 48 6.56 61 164 Ni 433 312 557 1.45 526 290 753 934 646 1197 661 408 779 21 944 4179 Zn 128 30 45 2.73 89 21 40 38 15 18 39 19 33 165 36 42 Rb 0.14 0.08 0.31 bdl bdl 0.10 0.36 0.12 0.09 0.60 0.20 0.05 0.24 31 0.04 0.04 Sr 0.18 20 67 690 0.16 25 63 0.46 17 21 0.34 28 56 669 0.11 0.14 Y 0.83 10 20 0.11 0.58 5.58 15 0.36 2.78 5.23 0.77 5.03 15 23 0.05 bdl Zr 0.58 8.64 26 bdl 0.58 7.50 26 0.40 3.21 5.98 0.69 10 63 100 bdl 0.04 Nb 0.06 0.03 1.06 bdl 0.06 0.05 0.50 0.07 0.04 0.21 0.12 0.04 3.31 3.02 0.03 0.03 Cs 0.04 0.01 0.02 0.11 0.01 0.03 0.02 0.02 0.01 bdl 0.06 bdl 0.02 1.10 0.01 0.01 Ba 0.04 0.14 3.55 4.52 0.03 0.27 5.12 0.25 0.22 3.09 0.10 0.11 6.14 220 0.08 0.06 La 0.02 0.39 1.17 0.70 bdl 0.45 1.43 bdl 0.22 0.45 bdl 1.10 4.06 161 0.02 0.01 Ce bdl 2.47 6.25 1.43 0.07 2.39 6.89 0.06 0.71 1.64 0.08 4.78 18 713 0.01 bdl Pr 0.01 0.59 1.39 0.15 bdl 0.49 1.34 bdl 0.11 0.23 0.08 0.82 3.12 47 0.01 bdl Nd 0.10 4.17 10 0.42 bdl 2.76 8.01 0.07 0.75 1.32 0.39 3.92 15 113 0.10 0.06 Sm 0.05 1.78 3.78 0.20 bdl 1.03 2.51 0.08 0.24 0.34 0.31 1.12 3.59 18 0.02 0.07 Eu 0.04 0.41 1.20 0.29 0.02 0.34 0.95 0.07 0.12 0.17 0.20 0.27 1.07 4.97 0.04 0.02 Gd 0.14 2.01 4.57 0.17 bdl 1.36 3.47 0.17 0.35 0.75 0.37 1.22 3.38 4.45 0.07 0.06 Tb bdl 0.32 0.66 0.05 0.02 bdl 0.53 0.03 bdl 0.13 0.10 0.17 0.46 1.08 bdl bdl Dy bdl 2.12 4.67 0.54 0.15 1.23 3.06 bdl 0.49 0.91 bdl 1.10 3.01 6.30 0.01 bdl Ho 0.04 0.43 0.83 bdl 0.03 bdl 0.61 0.04 0.12 0.20 bdl 0.23 0.57 1.11 bdl 0.01 Er 0.15 1.08 2.19 0.31 0.12 0.67 1.45 bdl 0.34 0.55 0.34 bdl 1.65 5.36 0.07 0.05 Tm bdl 0.16 0.31 0.12 bdl 0.10 0.23 0.02 bdl 0.10 0.07 0.09 0.22 bdl 0.01 bdl 145 Yb 0.22 0.94 2.04 bdl 0.12 0.61 1.54 0.16 bdl 0.54 bdl 0.56 1.39 bdl 0.12 0.08 Lu 0.05 0.13 0.24 0.02 0.02 0.09 0.20 0.02 0.04 0.08 0.09 0.08 0.20 1.10 0.01 bdl Hf 0.15 0.49 1.32 1.08 0.04 0.42 1.01 0.07 0.18 0.15 0.52 0.32 2.38 3.54 0.29 bdl Ta 0.01 0.02 0.07 0.02 bdl 0.04 0.07 0.02 0.03 0.01 0.07 0.01 0.19 1.75 bdl 0.02 Pb 0.10 0.05 0.10 0.40 0.02 0.09 0.07 0.04 0.06 bdl bdl 0.06 0.11 25 0.03 0.03 Th 0.02 0.02 0.03 0.02 bdl 0.04 0.12 0.03 0.01 0.01 0.14 0.04 0.06 6.25 0.02 0.01 U 0.03 0.01 0.04 0.05 bdl 0.03 0.02 0.01 0.02 0.01 bdl 0.03 0.05 6.43 0.01 0.01 bdl = below detection limits 146 Table 4.4. Temperature estimates (°C) for Ouassé peridotites and pyroxenites. Calculations were made considering the core analysis of each grain and pressure conditions of 1 GPa. TREE-Pxa TREE-Cpx-Plb TCa-in-Opxc TTa98d TBK90c TAmp-Ple TOl-Splf TCa-in-Olg Harzburgite OU2 porphyroclast 956 747 neoblast 923 883 892 721 Harzburgite OU3 porphyroclast 973 neoblast 945 923 934 721 Harzburgite OU7A porphyroclast 923 728 Amp-websterite OU1 mediumto coarse-grained pyroxenes 1251 ± 37 990 958 923 granoblastic assemblage 1310 ± 14 997 992 946 970 Amp-websterite OU4 mediumto coarse-grained pyroxenes 1264 ± 50 965 941 927 granoblastic assemblage 1249 ± 52 971 927 914 Orthopyroxenite OU5A mediumto coarse-grained pyroxenes 924 granoblastic assemblage 932 894 911 Amp-websterite OU7B 993 970 965 Orthopyroxenite OU8 869 865 861 a= Liang et al. (2013) b= Sun and Liang (2017) c= Brey and Köhler (1990) d= Taylor 1998 e= Holland and Blundy (1994) f= Jianping et al. (1995) g= De Hoog et al. (2010) 147 Table 4.5. Summary of 40Ar/39Ar results for hornblende analysis Plateau Age Isochron Age Integrated age Sample Irrad min Calculation n %39Ar MSWD Age(Ma) ± 1s 40Ar/36Ar ± 1s Age(Ma) ± 1s n Age(Ma) ± 1s OU7B NM-312A Hornblende WMA 10 97.0 0.44 56.05 ± 0.17 297.3 ± 4.4 55.87 ± 0.47 12 56.57 ± 0.24 OU1 NM-312A Hornblende WMA 12 100.0 0.50 56.01 ± 0.19 295.4 ± 0.8 56.03 ± 0.22 12 55.98 ± 0.30 Irradiation = NM-312A WMA = Weighted mean age 148 Table 4.6. Trace elements compositions and Mg# values of calculated liquids in equilibrium with clinopyroxenes from Ouassé pyroxenites. Sample OU1 OU4 OU5A OU7B Rock type Amp-websterite Amp-websterite Orthopyroxenite Amp-websterite Ti 3040 2355 1211 1662 V 115 64 70 87 Rb 448 566 491 253 Ba 212 352 565 162 Th 122 176 28 182 U 23 53 33 53 Nb 3.99 6.19 5.45 4.91 La 7.34 8.35 4.11 20 Ce 29 28 8.28 56 Pb 0.80 1.12 0.81 0.88 Pr 4.22 3.54 0.80 5.88 Sr 157 194 95 222 Nd 22 15 3.99 21 Sm 6.12 3.55 0.82 3.84 Zr 70 61 26 78 Hf 2.07 1.65 0.71 1.25 Eu 1.17 0.97 0.34 0.76 Gd 5.02 3.39 0.88 3.04 Dy 4.79 2.78 1.11 2.49 Y 21 12 5.96 11 Er 2.79 1.74 0.88 1.22 Yb 2.18 1.42 0.74 1.30 Lu 0.29 0.21 0.08 0.20 Sr/Y 7.38 16 16 21 La/Yb 3.36 5.88 5.53 16 Mg# 58 67 80 69 149 Chapter 5 5.1. Concluding remarks The present PhD Thesis focused on mantle sequences from the External Ligurian and New Caledonia ophiolites. The External Ligurian ophiolites expose a lithospheric mantle of subcontinental origin that underwent a decompression evolution during the Mesozoic rifting stage that preceded the opening of the Jurassic Western Tethys basin. The New Caledonia ophiolites expose a lithospheric mantle of oceanic nature that experienced melt extraction in a forearc environment during the Late Eocene subduction phase. Despite these differences, both studied mantle sections are characterized by a variety of pyroxenite layers. In addition, large-scale lithospheric shear zones were detected both in the External Ligurian (Monte Sant’Agostino) and New Caledonia (Bogota Peninsula Shear Zone) ophiolites. The Monte Sant’Agostino shear zone accommodated the deformation of the subcontinental mantle during a rifting-related exhumation event. The Bogota Peninsula Shear Zone has been reinterpreted as a paleotransform fault developed during the subduction phase. Besides the specific objectives of each case study, the results of the Thesis could be read in terms of the heterogeneities exhibited by the mantle sequences. Listed below some observations (abbreviations: EL = External Ligurian case study; NC = New Caledonia case study): • EL: Considering the studied mantle bodies (Monte Gavi and Monte Sant’Agostino) and the others case studies reported in the literature, the Thesis has shown a heterogeneous rifting-related plagioclase-facies evolution, characterized by presence or absence of mylonitic to ultramylonitic deformation, and presence or absence of melt-rock reaction events. • EL: The Monte Gavi mantle body documents the occurrence of a plagioclase-facies melt impregnation event in the late Triassic. The melt impregnation event recorded by the Monte Nero mantle body conversely most likely developed in the middle-late Jurassic. These age differences highlight the polyphase nature of the rifting-related mantle evolution, which presumably occurred in a time span of ~60 Ma. • NC: The studied mantle sequence (near Ouassé, Bogota Peninsula) offers an extreme pyroxenite heterogeneity within few hundred meters. These pyroxenites provide evidence for: (i) the presence of compositionally diverse subduction-related magmas, and (ii) a variable geochemical contribution of the sinking slab in the forearc mantle.