Orthopyroxene–omphacite- and garnet–omphacite-bearing magmatic assemblages, Breaksea Orthogneiss, New Zealand: Oxidation state controlled by high-P oxide fractionation ☆ Timothy Chapman a, ⁎, Geoffrey L. Clarke a , Nathan R. Daczko b,c , Sandra Piazolo b,c , Adrianna Rajkumar a a School of Geosciences, F09, University of Sydney, Sydney, NSW 2006, Australia b ARC Centre of Excellence for Core to Crust Fluid Systems, Department of Earth and Planetary Sciences, Macquarie University, NSW 2109, Australia c GEMOC, Department of Earth and Planetary Sciences, Macquarie University, NSW 2109, Australia abstract article info Article history: Received 19 December 2013 Accepted 21 November 2014 Available online 3 December 2014 Keywords: Omphacite–garnet granulite Orthopyroxene eclogite Omphacite–orthopyroxene granulite REE Igneous omphacite EBSD microstructure The Breaksea Orthogneiss comprises a monzodioritic host partially recrystallised to omphacite–garnet–plagioclase– rutile granulite at 850 °C and 1.8 GPa, with metre to decametre-scale, cognate inclusions ranging from ultramafic through gabbroic to monzodioritic composition. Coarsely layered garnetite and diopsidic clinopyroxenite cumulate preserves igneous textures, whereas garnet–omphacite cumulate shows a partial metamorphic overprint to eclogite. Garnet and omphacite in undeformed to weakly deformed rocks have similar major and rare earth element characteristics reflecting their common igneous origin, pointing to a lack of metamorphic recrystallisation. Inclusions of omphacite–orthopyroxene–plagioclase–ulvöspinel orthogneiss have whole-rock compositions almost identical to the host monzodiorite. Reaction zones developed along contacts between the orthopyroxene-bearing inclusions and host contain metamorphic garnet that is microstructurally and chemically distinct from igneous garnet. The presence of orthopyroxene is interpreted to reflect redox distinctions: early, oxidised magma crystallised orthopyroxene and ulvöspinel at high-P (~1.8 GPa), garnet crystallisation having been suppressed. Progressive frac- tionation of oxygen into early formed phases (ulvöspinel, magnetite, orthopyroxene, ferric iron-rich omphacite and rare garnet) drove the magma to less oxidising conditions, resulting in the more common igneous assemblage of garnet, omphacite and rutile in the main host. © 2014 Elsevier B.V. All rights reserved. 1. Introduction The lower crustal roots of magmatic arcs are exposed in only a few locations, such as the Ladakh–Kohistan arc in Northern Pakistan (Jagoutz and Schmidt, 2012), the Talkeetna arc in Alaska (Behn and Keleman, 2006; DeBari and Coleman, 1989) and the Fiordland arc, New Zealand (Clarke et al., 2000; Daczko et al., 2009, 2012). Crustal thicknesses inferred commonly for magmatic arcs lie in the range of 30 to 40 km (1.0–1.4 GPa at their base), with some overthickened Cor- dilleran arcs reaching thicknesses in excess of 50 km (1.5–2.0 GPa). At such deep crustal levels, intermediate to mafic high-P granulite and eclogite are predicted to be the dominant rock types (O'Brien and Rötzler, 2003; Rudnick and Fountain, 1995). These rare exposures pro- vide unique insights into the dynamics of lower crustal processes, in particular the interplay between magmatic and metamorphic processes, that otherwise can only be directly observed via xenoliths (e.g. Griffin et al., 1979). There are large overlaps in the stability of mineral assemblages com- monly occurring in high-grade orthogneisses and their plutonic protoliths. This commonality can lead to ambiguity in distinguishing metamorphic grains from igneous relicts, and debate concerning petro- logic and tectonic interpretations of lower continental crustal rocks (e.g. Kotková and Harley, 2010; Štípská and Powell, 2005; Williams et al., 2000). The Cretaceous Breaksea Orthogneiss in Fiordland (De Paoli et al., 2012) presents such an example, incorporating rare exposures of intermediate to mafic lower crustal rocks formed in a thickened arc. Igne- ous crystallisation and subsequent metamorphic recrystallisation at high-P and high-T produced omphacite–garnet–plagioclase assemblages in monzodioritic compositions, and eclogite facies assemblages in gab- broic compositions (De Paoli et al., 2012). However, Clarke et al. (2013) interpreted that the majority of observed mineral assemblages in the Breaksea Orthogneiss should be attributed to igneous crystallisation, based on commonality in the rare earth element (REE) content of garnet and clinopyroxene across diverse cogenetic rock types. Mineral assemblages involving orthopyroxene and omphacite occur in both granulite and eclogite components of the Breaksea Orthogneiss, presenting an unusual mix of minerals commonly considered character- istic of low- and high-P granulite facies conditions (e.g. De Paoli et al., 2012; Green, 1970; Green and Ringwood, 1967). Orthopyroxene– Lithos 216–217 (2015) 1–16 ☆ Mineral abbreviations given follow Kretz (1983). ⁎ Corresponding author. Tel.: +61 2 93518199. E-mail address: t.chapman@sydney.edu.au (T. Chapman). http://dx.doi.org/10.1016/j.lithos.2014.11.019 0024-4937/© 2014 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Lithos journal homepage: www.elsevier.com/locate/lithos