high-temperature metamorphism and crustal melting: working...

An International Journal ofMINERALOGY, CRYSTALLOGRAPHY, GEOCHEMISTRY,ORE DEPOSITS, PETROLOGY, VOLCANOLOGYand applied topics on Environment, Archaeometry and Cultural Heritage

PERIODICO di MINERALOGIAestablished in 1930

Periodico di Mineralogia (2015), 84, 3B (Special Issue), 591-614 DOI: 10.2451/2015PM0434

High-temperature metamorphism and crustal melting: working with melt inclusions

Omar Bartoli1,*, Antonio Acosta-Vigil1,2 and Bernardo Cesare1

1 Dipartimento di Geoscienze, Università di Padova, Via Gradenigo 6, 35131 Padova, Italy2 Instituto Andaluz de Ciencias de la Tierra, Consejo Superior de Investigaciones Científicas-

Universidad de Granada, Avda. de Las Palmeras nº 4, Armilla 18100, Granada, Spain*Corresponding author: [email protected]

Abstract

The application of melt inclusion (MI) studies to migmatitic and granulitic terranes is a recent, small-scale approach for a better understanding of melting in the continental crust. In order to show the role of anatectic MI in providing a wealth of microstructural and compositional information on high-temperature metamorphism and crustal anatexis, we review a series of studies on the crustal footwall of the Ronda peridotites (Betic Cordillera, S Spain), which consists of an inverted metamorphic sequence with granulite-facies rocks showing extensive melting on top and amphibolites-facies rocks at the bottom. We studied the microstructures and geochemistry of small (2-10 µm) primary MI hosted in peritectic garnet of metatexites at the bottom of the migmatitic sequence and of mylonitic diatexites close to the contact with the mantle rocks. The occurrence of MI is a proof that the investigated rocks were partially melted at some time in their history, despite other microstructures indicating the former presence of melt in diatexites were erased by deformation. MI show a variable degree of crystallization ranging from totally glassy to fully crystallized (nanogranites), consisting of Qtz+Pl+Kfs+Bt+Ms aggregates (often modal Kfs > Pl in diatexites). Piston cylinder remelting experiments led to the complete rehomogenization of nanogranites in metatexites at the conditions inferred for anatexis. Compositions of investigated MI are all leucogranitic and peraluminous and differ from those of coexisting leucosomes and from melts calculated by phase equilibria modeling. Systematic compositional variations have been observed between MI in metatexites and diatexites: the former commonly show higher H2O, CaO, Na2O/K2O and lower FeO. The compositions of MI in metatexites and diatexites are interpreted to record the composition of the anatectic melts produced from a peraluminous greywacke i) on, and immediately after crossing, the fluid-saturated solidus of this metasedimentary rock, and ii) during anatexis via biotite dehydration melting at increasing temperature, respectively.

Periodico di Mineralogia (2015), 84, 3B (Special Issue), 591-614 O. Bartoli et al.592

Introduction

High-temperature (HT) metamorphism and partial melting (anatexis) of the mid- to lower continental crust, together with extraction and ascent of the magma to upper crustal levels, represent the main agents of differentiation of the continental crust (Brown, 2010; Sawyer et al., 2011), and have profound effects on the rheology of the lithosphere and, as a consequence, on its geodynamics (e.g., Vanderhaeghe, 2001; Brown et al., 2011). Anatexis of the metasedimentary crust produces granitic magmas (Brown, 2013) that can segregate from the source regions leaving high grade metamorphic rocks like granulites as residuum (Vielzeuf et al., 1990), and either form S-type granites in the upper crust (e.g. Petford et al., 2000; Brown, 2013) or migrate to the surface originating explosive volcanism (e.g., Pallister et al., 1992). Migmatites in the middle crust represent both zones of melt generation and zones of melt transfer (Sawyer, 2008; Brown et al., 2011).

The wide spectrum of issues related to high-temperature metamorphism and crustal anatexis justifies the increasing attention that these topics have received in the last years (Brown and Rushmer, 2006; Sawyer et al., 2011; Brown, 2013; Brown and Korhonen, 2009; and references therein). From a petrological and geochemical point of view, there is the strong urge for a better chemical characterization of natural crustal melts (Sawyer et al., 2011). In fact, despite the wealth of experimental studies on crustal melting (Clemens, 2006; and references therein), and the important information that

they provide, their direct application to natural contexts is often not straightforward, and sometimes problematic (see White et al., 2011).

Traditionally, representative examples of natural anatectic melts produced by metasedimentary protoliths were identified in S-type granites and rhyolites, and in leucosomes from migmatites (Brown and Rushmer, 2006; Clemens and Stevens, 2012). However, the reliability of the former has been challenged in recent works that document an important contamination of primary melt compositions by entrainment of residual or peritectic material (Stevens et al., 2007; Clemens et al., 2011; Clemens and Stevens, 2012 and references therein). In turn, leucosome chemistry is generally affected by cumulus phenomena and fractional crystallization, or presence of pre-anatectic phases (Sawyer, 2008, 2014; Marchildon and Brown, 2001). It is therefore apparent that the composition of natural anatectic melts remains one of the least constrained parameters in the petrological modelling of high-temperature metamorphism and crustal melting.

However “This situation is changing...” as stated by Sawyer et al. (2011). By studying the glassy melt inclusions (MI) hosted in peritectic minerals of metapelitic enclaves from the Neogene Volcanic Province (SE Spain), Cesare (2008) and Acosta-Vigil et al. (2007, 2010) demonstrated that during their growth, peritectic minerals can entrap small droplets of the coexisting anatectic melt produced during incongruent melting reactions. The re-examination of more conventional, regionally metamorphosed migmatite and granulite

While partial melting at the bottom of the migmatitic sequence likely started in the presence of an aqueous fluid phase, MI data support the fluid-absent character of the melting event in diatexites. Anatectic MI should therefore be considered as a new and important opportunity to understand the partial melting processes.

Key words: crustal anatexis; migmatite and granulite; peritectic phase; melt inclusions; nanogranite.

Periodico di Mineralogia (2015), 84, 3B (Special Issue), 591-614 593High-temperature metamorphism and crustal…

terranes has recently proved the common occurrence of MI also in peritectic phases of these rocks (e.g. Cesare et al., 2009; Ferrero et al., 2012; Bartoli et al., 2013b). In most cases these inclusions totally crystallized upon slow cooling and now contain cryptocrystalline aggregates that have been named “nanogranite” owing to their grain size, texture and chemical/mineralogical composition (Cesare et al., 2009). The findings of nanogranite in many anatectic terranes worldwide (Cesare et al., 2009, 2011; Ferrero et al., 2012; Barich et al., 2014; Bartoli et al., 2013b; Darling 2013), along with the novel experimental approach recently proposed to successfully re-melt these inclusions in order to obtain the original composition of the trapped melt (Bartoli et al. 2013a), open the possibility for the detailed geochemical characterization of natural anatectic melts from different geodynamic settings.

In this paper we review a series of MI studies on the migmatitic terrane underlying the Ronda peridotites (Betic Cordillera, S Spain), in order to show how MI hosted in peritectic minerals of migmatites and granulites can provide a wealth of microstructural and compositional information on HT metamorphism and crustal anatexis.

Geological setting and sampling

The Betic Cordillera (S Spain) represents the westernmost part of the peri-Mediterranean Alpine orogen, formed during the N-S to NW-SE convergence of the African and Iberian plates from Late Cretaceous to Tertiary times (Andrieux et al., 1971; Dewey et al., 1989). This study focuses on former anatectic rocks located structurally below and at the contact with the Ronda peridotites (Figure 1), which represent the largest known exposure of subcontinental lithospheric mantle on the surface of the Earth (≈ 300 km2; Obata, 1980). The peridotites occur as km-thick slabs sandwiched in-between

mostly metasedimentary crustal rocks (Lundeen, 1978; Platt et al., 2013). These crustal units are characterized by their increasing metamorphic grade, extent of melting and intensity of deformation towards the contact with the mantle rocks (Loomis, 1972; Westerhof, 1977; Torres-Roldán, 1981, 1983; Tubía et al., 1997, 2013; Platt et al., 2003; Acosta-Vigil et al., 2001, 2014; Esteban et al., 2008). Mantle and crustal rocks are separated by ductile shear zones (e.g., Tubía et al., 1997, 2013). Different authors (see Sánchez-Rodríguez and Gebauer, 2000; Esteban et al., 2011a; Acosta-Vigil et al., 2014; Massonne, 2014 and references therein) ascribed the timing of HT metamorphism and anatexis in the crustal rocks around the Ronda peridotites to either the Alpine or Variscan orogenies.

In the study area, the mantle rocks are emplaced over the Ojen nappe (Figure 1c) which, in general, is formed by ≈ 30 m of strongly mylonitic rutile-bearing pelitic gneisses at the very contact with the peridotites, and ≈ 200 m of rutile-free mostly quartzo-feldspathic gneisses, that grade downwards into undeformed pelitic and quartzo-feldspathic diatexites and metatexites (Figure 2) (Tubía, 1988; Acosta, 1998). The bottom of the sequence is constituted by amphibolite-facies Sil-bearing schists and marbles (Westerhof, 1977; Tubía, 1988). A low-temperature shear zone divides retrograde migmatites and chlorite-rich schists (Tubía et al., 1997). Decimetric to decametric amphibolite lenses that preserve eclogitic relicts (Figure 2) have been described within the migmatites (Tubía and Gil-Ibarguchi, 1991; Tubía et al., 1997), mostly included within the quartzo-feldspathic metatexites (Acosta, 1998).

This study is primarily focused on MI in the quartzo-feldspathic metatexite (sample ALP1) and mylonitic diatexite (sample ALP13). These samples were collected in the metamorphic footwall of the Sierra Alpujata peridotite massif (Figures 1, 2), roughly in correspondence of the Los Villares transect of Tubía et al. (1997). These


rocks have a bulk rock composition corresponding to that of Ca-poor, Si-rich peraluminous greywacke (Si2O ≈ 72-74 wt%, FeO ≈ 2-3 wt%, Na2O ≈ 2 wt%, K2O ≈ 4.5-5 wt%), with lower alumina content (Al2O3 ≈ 14-14.5 wt%) and aluminium saturation index (ASI ≈ 1.2-1.3) than typical pelitic rocks (Bartoli, 2012). In the field, the metatexite ALP1 shows a stromatic structure (Sawyer, 2008) with thin (≤ 1 cm) and discontinuous leucocratic layers surrounded by a mesocratic matrix (Figure 2). Conversely, the diatexite ALP13 appears as a deformed, gneissic

rock composed of alternating leucocratic bands and mesocratic bands (Figure 2).

Petrography

The quartzo-feldspathic migmatites are composed of varying modal amounts of Qtz, Pl, Kfs, Bt, Sil and Grt, with minor amounts of Gr, Ilm, Ap, Zrn and Mnz (mineral abbreviations after Kretz, 1983). Graphite is randomly distributed in the matrix of rocks.

Metatexites are fine- to medium-grained (≈ 0.2-

Figure 1. (a) Location map of the study area in the S Spain. (b) Simplified geological map of the western sector of the Betic Cordillera (modified after Esteban et al., 2011b). (c) Geological map of the Sierra Alpujata massif. Blue and yellow stars show the location of the studied diatexite ALP13 and metatexite ALP1.


3.0 mm) rocks made of i) a mesocratic matrix (containing Qtz+Pl+Kfs+Bt+Sil+Ilm+Ap) that encloses Grt and Kfs porphyroblasts, and ii) discontinuous, medium-grained leucocratic bands (Figure 2). The main foliation (Sp) is defined by abundant oriented Bt flakes (≈ 8-12 vol.%) generally clustered with fibrolitic Sil (Figure 3a). Apatite is mostly included within biotite and sillimanite aggregates. Along with these minerals, rare (≈ 1vol.%) Ms crystals are present in metatexites, and appear with resorbed shapes, included in K-feldspar porphyroblasts or associated with Bt and Sil (Figure 3b, c).

Notably, fibrolite apparently grew on primary muscovite (Figure 3c). All these textures suggest a prograde exhaustion of muscovite. Alkali feldspar is often poikiloblastic, containing inclusions of Qtz, Pl, Bt and Sil (Figure 3d). Garnet (2-5 vol.%) occurs as small (≈ 50-200 μm in diameter) subhedral to euhedral crystals (Figure 3a). The leucocratic layers (≈ 5 vol.%) contain Qtz, Pl, Kfs, Bt and rare Grt. Here feldspars may show euhedral shapes (Figure 3e).

Compared with metatexite, the diatexite ALP13 is richer in Grt (≈ 5-10 vol.%) and poorer in Bt (≈ 2-5 vol.%). The fine-grained

Figure 2. Schematic section of the crustal footwall of the Ronda peridotites at Sierra Alpujata showing the location of the studied samples (blue and yellow stars as in Figure 1). The field aspect images and photomicrographs show the macro and microstructural evolution of migmatites as a function of distance to the bottom of the Ronda peridotite slab. Red arrows show the location of peritectic garnets. Yellow lines show the traces of the main foliation defined by biotite and/or sillimanite folia.


Figure 3. Microstructures in the stromatic metatexite ALP1 (a-e) and diatexite ALP13 (f-i). (a) Photomicrograph of the mesocratic matrix in which biotite and sillimanite grains define the main foliation. Red arrows: graphite lamellae (b) Resorbed muscovite armoured in K-feldspar porphyroblast. (c) Primary muscovite partially replaced by fibrolite. (d) K-feldspar poikiloblast with inclusions of quartz, plagioclase and biotite. (e) Euhedral faces of feldspar, suggesting crystal growth from melt (Vernon, 2011) in a leucocratic band. (f) Mesocratic matrix showing fabric-forming sillimanite. Red arrows: graphite lamellae. (g) Garnet crystal partially replaced by Bt formed as product of retrograde reactions during cooling. (h, i) Photomicrographs of two leucocratic bands. In (h) feldspars and quartz are deformed and elongated, whereas they display euhedral shape with planar faces in (i).


(grain size of ≈ 20-200 μm), quartzo-feldspathic matrix includes porphyroclasts of Grt (0.5-3 mm in diameter) and Kfs (up to 2 cm in size) (Figure 2). A mylonitic foliation is defined by alignment of Sil folia and minor elongate crystals of Bt and Ilm, oriented ribbons of Qtz and quartzo-feldspathic layers (Figures 2, 3f). Ms is absent, whereas Sil is both fibrolitic and prismatic (Figure 3f, g). Bt and minor Bt+Qtz and Bt+Pl+Qtz intergrowths often grew in the strain shadows partially to totally replacing Grt crystals (Figure 3g). Quartz microstructures, such as chessboard texture, subgrains, irregular grain boundaries and undulose extinction, indicate the occurrence of high strain and dynamic recrystallization at temperatures ≥ 650 °C (Stipp et al., 2002). These mylonitic diatexites are also characterized by, compared to the metatexites, a greater abundance and greater thickness (up to 20 cm) of leucocratic bands (Figure 2), which are mainly composed of Qtz+Kfs+Pl often showing rounded or elongate shapes mantled by finer-grained trails (Figure 3h). Locally, leucocratic bands may contain feldspars displaying euhedral shapes with straight boundaries (Figure 3i).

Microstructures indicating the presence of melt such as mineral pseudomorphs after melt films and pools (Sawyer, 2001; Vernon, 2011) are abundant in the matrix of metatexite, but are very rare in the diatexite (Figure 4). The leucocratic bands of both metatexites and diatexites, mainly composed of Qtz, Pl and Kfs, showing a granitic composition (see below) and containing some igneous microstructures such as euhedral minerals (Figure 3e, i), are interpreted as anatectic leucosomes. The crystallization of these leucocratic portions from an anatectic melt is also supported by the presence of a more albitic, euhedral plagioclase (Sawyer, 2001). Biotite and Bt+Pl+Qtz intergrowths replacing garnet (Figure 3g) are likely to have formed as a result of melt-consuming retrograde reactions (Kriegsman and Hensen, 1998).

Microstructural characterization of MI

MI have been recognized within garnet in both types of quartzo-feldspathic migmatites (Figure 5). In the diatexite, MI-bearing garnets are less abundant (~ 20% of the garnet population) than in the metatexite (~ 90%). In general, MI are clustered, forming groups of tens of inclusions which are often characterized by a similar size. Clusters, generally displaying a subspherical geometry, are preferentially located at the core of small garnets in the metatexite (Figure 5a), whereas they do not have a preferential

Figure 4. Photomicrographs of pseudomorphs after melt films in metatexite. (a) Plagioclase that crystallized as a melt pseudomorph around rounded quartz. (b) K-feldspar with cuspate outlines that has probably crystallized from a pool of melt. The reactant minerals, quartz and plagioclase, are rounded and resorbed. Crossed polars with l plate.


arrangement within garnets of the diatexite (Figure 5b, c). In the latter they may occur both at the core and close to rim, sometimes showing

a sigmoidal to spiral-like geometry (Figure 5c). No compositional discontinuities have been observed between MI-rich and MI-free portions of the garnets, and MI do not form trends along linear discontinuities of the host crystal.

In transmitted light under the optical microscope, most MI appear dark-brownish (Figure 6a) and contain a polycrystalline aggregate of birefringent crystals under cross polarized light (Figure 6b, c). Other MI are transparent in transmitted light and contain a homogeneous isotropic phase, often along with a bubble (Figure 6d). Raman spectroscopy indicated that these bubbles are empty and therefore represent shrinkage bubbles. In the diatexite, some dark-brownish MI mantle fibrolite needles (Figure 6e). The shape of both types of inclusion is isometric and their size does not exceed 15 μm (average size ~ 5 μm).

Owing to the small size of MI, their microstructures can be successfully characterized with back-scattered electron (BSE) imaging, using the new generation of Field Emission Gun (FEG)-based scanning electron microscopes (SEM) and a working distance in the range 7-15 mm. MI-forming crystals, have been identified by acquiring EDS and Raman spectra, and X-ray maps of the major elements (see also Ferrero et al., 2012; Bartoli et al., 2013b). Under SEM investigation, MI appear typically facetted, and often with a well-developed negative crystal shape (Figure 7). They show a variable degree of crystallization, even in the same cluster, ranging from totally (i.e. nanogranites), to partially crystallized, and down to crystals-free (glassy) MI (Figure 7). Glassy MI, common in the diatexite, are very rare in the metatexite. No systematic difference in size between the different types of inclusions is observed. Indeed, the size of glassy MI is often equal to (Figure 7a), and sometimes even larger than (Figure 7b), that of the partially crystallized or nanogranite inclusions.

Crystallized inclusions contain aggregates

Figure 5. Photomicrographs of the MI-bearing garnets in metatexite (a) and diatexite (b, c). MI clusters may show a subspherical geometry (red arrows) or a spiral-like arrangement (red dotted lines).


of Qtz, Bt, Ms, Pl and Kfs with equigranular, hypidiomorphic to allotriomorphic texture (Figure 8). Crystal size ranges from hundreds of nm to a few μm. The SEM investigation of tens of MI highlights some differences in the mineral mode of nanogranites from the two samples. In the metatexite, MI generally contain Qtz+Bt+Ms+Pl and rare Kfs (Figure 8a, b), whereas the assemblage Qtz+Bt+Ms+Kfs and minor Pl is common in garnets from the diatexite (Figure 8c, d). The largest grains within crystallized MI generally consist of subhedral to euhedral micas, ≤ 2 μm in size, which often grew starting from the inclusion walls (Figure 8) and are likely to be the first phases to have crystallized (see also Ferrero et al., 2012). Feldspars form subhedral to anhedral crystals, whereas Qtz occurs as an interstitial phase. Sometimes granophyric to microgranophyric intergrowths of Qtz and feldspars are present (Figure 8b), mostly in MI from the diatexite. Some nanogranites display a variable micro- to nano-porosity that is more developed in the samples from metatexite (Figure 8b). Here, micro-Raman mapping of some crystallized MI located below the Grt surface documented the presence of micro- and nano-pores filled with liquid H2O (Figure 9), suggesting H2O exsolution during crystallization of hydrous

Figure 6. Photomicrographs of melt inclusions. (a) Plane-polarized image of a MI cluster in metatexite. (b, c) Plane-polarized and crossed polars images respectively of a crystallized inclusion in metatexite. (d) Glassy MI containing a shrinkage bubble (red arrow) in diatexite, plane-polarized light. (e) Crystallized MI with a Sil needle (white arrow) that is likely to have favored the entrapment of melt, plane-polarized light.

Figure 7. SEM-BSE images of coexisting crystallized and preserved glassy MI in metatexite (a) and diatexite (b).


melts to nanogranites, and preservation of the H2O within the inclusion.

Partially crystallized inclusions are indistinguishable from nanogranites under the optical microscope because of their small size. The presence of glass together with crystals is only revealed by SEM investigation. Glass

occupies different area percentages of the MI, and commonly coexists with Ms, Bt and Qtz in the partially crystallized MI from the metatexite (Figure 8e). In the diatexite, rare partially crystallized MI generally contain only Ms and/or Bt together with the glass (Figure 8f).

Figure 8. SEM-BSE images of nanogranites and partially crystallized inclusions in metatexite (a, b, e) and diatexite (c, d, f). Red arrows: primary nanoporosity.


Chemical characterization of MI

The large variability in microstructures and grain size within nanogranites, coupled with their small diameters, does not allow a reliable estimate of their modal composition based on image analysis. Only glassy inclusions may be analyzed directly by electron microprobe, but they are very rare in the metatexite. To recover complete and meaningful compositional data, including the volatile contents of melt, nanogranites and partially crystallized MI in metatexite must be remelted to a homogeneous liquid, reversing the phase changes that occurred during natural cooling.

Experimental re-homohenization of nanogranites The first attempts to remelt nanogranites

using the routine technique in igneous petrology (i.e. the one atmosphere heating stage, see Esposito et al., 2012) resulted in extensive inclusion decrepitation and interaction with the host mineral (Cesare et al., 2009). Remelting experiments try to reverse the phase changes (crystallization and exsolution of fluids) that occurred in MI along the cooling path after entrapment (see Figure 9 in Bartoli et al., 2013a), and by using a ‘conventional’ heating stage to conduct these experiments, the internal pressure of inclusions largely exceeds the external pressure (i.e., ambient pressure),

Figure 9. Raman mapping of liquid H2O distribution within a crystallized melt inclusion in metatexite. (a) Representative Raman spectrum obtained from mapping: the peaks at 3620 and 3691 cm-1 correspond to main OH stretching vibrations in muscovite and biotite, respectively. (b) Investigated inclusion below garnet surface. (c) Raman map in the 3200-3400 cm-1 stretching region (bounded by red dotted lines in a) of liquid H2O. The inclusion contains both hydroxylated minerals and free H2O in the pores.


producing decrepitation of MI volatile loss. As a consequence of H2O loss, the complete remelting of nanogranite inclusion occurs at higher temperature than that of entrapment, favoring dissolution of host crystal and melt contamination (Bartoli et al., 2013a). To prevent these drawbacks, we remelted MI in metatexite at high-pressure conditions using a single-stage piston cylinder apparatus. Experiments were run at 700, 750 and 800 °C and 5 kbar, both under dry conditions and with excess H2O (see Bartoli et al., 2013a for details on the experimental procedure). After the experiments at 700 °C, MI reached a complete melt + vapour homogenization (Figure 10a). MI still preserve the original negative crystal shape, suggesting that the host garnet did not dissolve into the melt during heating, and therefore that the trapping temperature was not significantly exceeded. Higher experimental temperatures (750 and 800 °C) resulted in dissolution of the host into the melt, as suggested by irregular walls, and by formation of (i) ≈ 5 μm long decrepitation cracks extending into the host garnet and (ii) one or more bubbles (Figure 10b).

Chemical composition of MIThe micrometre scale of the MI required the

use of a focused EMP beam with size of ≈ 1 μm and analytical conditions chosen to minimize volatile and alkali loss (see details in Bartoli et al., 2013c). All the analyzed glassy MI contain a SiO2-rich, leucocratic and peraluminous melt (SiO2 ≈ 70-78 wt%, FeOTot+MgO+MnO+TiO2 < 3 wt%, ASI ≈ 1.10-1.20) (Table 1). Glassy MI in the metatexite have rather constant composition, with Na2O/K2O from 0.7 to 0.8 and K# ≈ 0.47 [K# = mol. K2O/(Na2O+K2O)]. Conversely, the composition of glassy MI in the diatexite is much more variable, richer in FeO, MgO and P2O5 and lower in CaO. In particular, these inclusions are highly variable in Na and K contents, such that we have classified them into two groups: type I, with K# ≥ 0.6 and

Na2O/K2O < 0.5, and type II, with K# ≤ 0.5 and Na2O/K2O > 0.6 (Table 1). Type II MI have been found only in 2 of the 20 investigated garnets. The analyzed MI correspond to granites based on their normative compositions (Figure 11a,b). When plotted in the Qtz-Ab-Or normative ternary diagram, MI data define two different clusters according to their K# and the composition of type II MI in diatexite overlaps that of glassy MI from metatexite (Figure 11b). All MI plot in the Qtz field, close to the 5 kbar cotectic curve and at some distance from the eutectic melt compositions of the haplogranite system. It should be noted that the involvement of Fe, Ti and Ca moves eutectic points and

Figure 10. SEM-BSE images of MI in metatexite after remelting experiments at 700 °C (a) and 800 °C (b).


cotectic curves toward quartz-richer, albite-poorer compositions, as recently demonstrated by Wilke et al. (2015) at a pressure of 2 kbar.

Normative compositions of MI re-homogenized at 700 °C overlap those of coexisting glassy MI in metatexite (Figure 11c, Table 1). The

Table 1. Major element composition (wt%) of melt inclusions in metatexite ALP1 (data from Bartoli et al., 2013b) and diatexite ALP13 (data from Bartoli et al., submitted), and of coexisting leucosomes (data from Bartoli, 2012). Analyses are shown on an anhydrous basis. Numbers in parentheses refer to 1σ standard deviation.

No. Analyses

Melt inclusions Leucosomes

Metatexite Diatexite Metatexite Diatexite

remelted* Type I Type II

3 13 37 31 8

SiO2 76.79 (0.68) 76.45 (1.08) 78.34 (0.78) 78.88 (1.70) 76.35 (4.48) 75.48TiO2 0.09 (0.15) 0.05 (0.08) 0.05 (0.08) 0.07 (0.07) 0.13 (0.04) 0.07Al2O3 12.98 (0.15) 13.00 (0.84) 11.66 (0.42) 11.74 (0.83) 13.28 (2.60) 14.32FeOT 1.32 (0.12) 1.89 (0.25) 1.56 (0.42) 1.40 (0.36) 0.82 (0.24) 0.64MnO 0.10 (0.10) 0.18 (0.11) 0.06 (0.06) 0.08 (0.08) 0.02 (0.00) 0.02MgO 0.08 (0.03) 0.13 (0.08) 0.15 (0.10) 0.15 (0.16) 0.24 (0.05) 0.21CaO 0.43 (0.21) 0.49 (0.13) 0.07 (0.05) 0.15 (0.18) 0.99 (0.04) 0.82Na2O 3.40 (0.32) 3.09 (0.45) 2.02 (0.36) 3.17 (0.52) 2.21 (0.20) 2.35K2O 4.62 (0.27) 4.45 (0.35) 5.92 (0.39) 4.13 (0.34) 5.74 (1.39) 5.90P2O5 0.20 (0.30) 0.26 (0.22) 0.18 (0.22) 0.23 (0.29) 0.13 (0.02) 0.13

ASIa 1.15 (0.09) 1.20 (0.09) 1.19 (0.10) 1.19 (0.12) 1.13 (0.05) 1.22Na2O/ K2O 0.74 (0.09) 0.70 (0.13) 0.34 (0.06) 0.77 (0.11) 0.40 (0.06) 0.40K#b 0.47 (0.03) 0.49 (0.05) 0.66 (0.04) 0.46 (0.03) 0.63 (0.03) 0.62H2O by diff 9.24 (1.61) 8.67 (2.16) 2.42 (1.60) 3.63 (1.08)

Calculated normative mineralogy

Qtz 34.23 (2.86) 35.70 (3.05) 41.26 (2.68) 41.56 (5.06) 39.07 (8.86) 37.09Crn 1.87 (0.63) 2.50 (0.90) 1.86 (0.89) 1.91 (1.02) 1.92 (0.72) 2.87Or 24.77 (1.34) 24.05 (2.3) 34.02 (2.15) 23.50 (1.95) 33.58 (6.57) 34.43Ab 26.11 (2.07) 23.85 (3.7) 16.56 (2.98) 25.84 (4.29) 18.53 (1.37) 19.60

An 0.83 (0.72) 0.91 (0.80) 0.13 (0.22) 0.18 (0.23) 4.04 (0.23) 3.15

aAlumina Saturation Index [= mol. Al2O3/(CaO+Na2O+K2O)]bK# = mol. K2O/(Na2O+K2O)* experimental run at 700 °C, 5 kbar, 24 h (for details see Bartoli et al., 2013a)


composition of glass from remelted MI at 750 and 800 ºC shows much more scatter than glassy MI and glass from the 700 ºC experiments (Figure 11c). In particular, glasses showing the greatest scatter are those showing clear microstructural and chemical evidence of interaction with the host garnet (e.g. irregular inclusion walls, FeOT > 2.5 wt% and ASI > 1.5; see Bartoli et al., 2013a). The average H2O content estimated by difference (i.e. 100-EMP totals) is much lower in MI from diatexite (≈ 2.4-3.6 wt%) than in MI from metatexite (≈ 9.2 wt%) (Table 1). NanoSIMS and Raman analyses on MI re-homogenized at 700 °C confirmed the high H2O content of these low-temperature melts (up to 9.8 and 7.6 wt% H2O, respectively; Bartoli et al., 2013b, 2014).

Melt inclusions in other migmatites

Although this paper focuses on MI in quartzo-feldspathic metatexites and diatexites, MI are virtually present in all migmatites outcropping below the Ronda peridotite at Sierra Alpujata. Here, we briefly report some interesting microstructural features of MI in other samples from the same anatectic terrane.

In the pelitic metatexite ALPA35.2 - composed of Qtz, Pl, Kfs, Bt, Sil, Crd, Ilm, Gr and rare Grt - MI may occur in ilmenite (100-200 μm across) scattered in the melanosome. MI (1.5-20 μm in diameter) are grouped with no preferred microstructural location and appear typically facetted, sometimes with a well-developed negative crystal shape (Figure 12a). MI are totally crystallized and contain quartz, plagioclase, biotite and rare apatite.In the pelitic gneiss ALP14 - composed of Qtz, Kfs, Bt, Sil, Grt, Crd, Rt, Gr and rare Pl - outcropping at the very contact with the peridotite (Figure 2), rutile needles are present within many of the partially crystallized and glassy MI (2-6 μm in diameter) hosted in garnet (Figure 12b). The EMP analysis of the phase

Figure 11. Pseudoternary diagrams showing the normative An, Qtz, Or and Ab compositions of all analyzed melt inclusions. Compositions of leucosomes in metatexite (white squares) and diatexite (black square) are plotted for comparison. (a) An-Or-Ab diagram (after O’Connor, 1965). (b) Qtz-Ab-Or diagram. Black triangle and lines show eutectic point and cotectic lines for the subaluminous haplogranite system at 0.5 GPa and aH2O = 1; black stars are eutectic points at aH2O = 0.6 and 0.4 (Becker et al., 1998). (c) Qtz-Ab-Or diagram for re-melted MI in metatexite. Data from dry and wet experiments overlap at 700, 750 and 800 °C (not shown; for details see Bartoli et al., 2013a).


enclosed between the rutile needle and the host garnet in Figure 12b provided a granitoid composition with SiO2 = 66.2 wt%, Al2O3 = 12.3 wt%, TiO2 = 2.2 wt%, FeO = 4.1 wt%, Na2O = 0.6 wt% and K2O = 6.7 wt%. Such a composition strongly suggests that this phase is a glass (i.e., the former melt), even though the anomalously high Ti and Fe contents indicate that the EMP measurement has been contaminated by the neighboring rutile and garnet.

Discussion

Nature of the studied melt inclusionsIn the studied quartzo-feldspathic rocks, MI

have been found in garnet, that is considered to be a typical peritectic mineral in anatectic Al-rich metasedimentary rocks (Thompson, 1982). The observed zonal arrangement (Figure 5) is a strong indicator of the primary nature of the MI (i.e., they were trapped when garnet was growing, Roedder, 1984). Hence, these MI represents small samples of the melt coexisting with the garnet during the prograde anatexis of the studied rocks. This mode of occurrence (MI hosted in peritectic phases) is different from that of MI hosted by phenocrysts in lavas, where the host mineral crystallized “from” the melt during cooling. In migmatites, instead, peritectic minerals and melt form at the same time during the prograde history, i.e. the peritectic garnet grows “with” the melt (see Figure 1 in Bartoli et al., 2014 for details).

As in other migmatitic terranes (Ferrero et al., 2012; Cesare et al., 2009; Barich et al., 2014), three type of inclusions were identified in the investigated quartzo-feldspathic migmatites: nanogranites, partially crystallized inclusions and preserved glassy inclusions. The phase assemblage in crystal-bearing MI (quartz, plagioclase, K-feldspar, muscovite and biotite) indicates that the trapped melt likely had a granitic composition, as confirmed by the analyses of remelted MI. In the partially crystallized MI, the glass represents the residual melt after the partial crystallization of the former trapped melt. During the crystallization of melts to nanogranites in the garnets of metatexite, H2O exsolved in micro- and nano-bubbles (Figure 9), but it was also consumed by the crystallizing biotite and muscovite.

Since the composition of nanogranites and glassy MI in the metatexite is comparable, their contrasting behaviour upon cooling is unexpected. Glassy inclusions are common

Figure 12. (a) SEM-BSE image of an ilmenite crystal in the pelitic metatexite ALPA35-2, containing primary MI. (b) SEM-BSE image of rutile-bearing MI hosted in garnet of the pelitic granulite ALP14.


in phenocrysts from volcanic rocks which undergo sudden (minutes to days) cooling from suprasolidus temperatures. Based on the statistical study of the size of MI found in garnets from granulites of the Kerala Khondalite Belt, Cesare et al. (2009) showed a difference between the mean diameter of the preserved glassy inclusions (smaller) and the crystallized nanogranites (larger), and proposed that crystal nucleation was inhibited in the smaller inclusions. The pore size effect is related to the higher interfacial energy (and consequent lower stability) of the crystals in smaller pores with respect to those in larger ones (Holness and Sawyer, 2008). However, in the present study the apparent size of glassy MI is often equal to, and sometimes even larger than, that of the nanogranites (Figure 7). This suggests that this range of dimensions (5-10 μm) may represent a threshold at which additional factors such as the heterogeneous distribution of nucleation sites among inclusions (e.g. presence or absence of irregularities on inclusion walls or trapped minerals) may significantly affect nucleation and melt crystallization (Cesare et al., 2011; Ferrero et al., 2012). In addition, one of the main compositional differences between MI in metatexite and diatexite is their H2O content. High H2O concentrations increase the diffusivities in MI (Lowenstern, 1995), favoring the partial or total crystallization of the majority of MI in metatexite. Conversely, the higher viscosity of the melt in MI from diatexites (due to low H2O content) likely inhibited the crystallization, with the formation of many glassy MI.

Melting conditions and reactions along the migmatitic terrane

The occurrence of MI represents a proof that the investigated rocks were partially melted at some time in their history (Cesare et al., 2011). The abundance of MI in the studied rocks clearly indicates the former occurrence of melt in them,

even though deformation and high temperature annealing in the diatexite have erased most of the classic microstructures indicating the former presence of melt, such as subhedral microstructure and mineral pseudomorphs after melt films and pools (see Holness and Sawyer, 2008; Holness et al., 2011).

The P-T conditions at which MI were trapped, and in turn at which rocks melted, have been investigated by means of thermodynamic modelling of phase equilibria. The reader may refer to Bartoli et al. (2013c) for details regarding the model chemical systems, thermodynamic databases and solution models used in the calculation of the phase diagrams. The relevant compositional isopleths for MI-bearing garnets cross consistently at ≈ 660-700 °C, ≈ 4.5-5 kbar for the metatexite ALP1, and at ≈ 820-830 °C and ≈ 5.5-6.5 kbar for the diatexite ALP13 (Figure 13), indicating that the temperature and, to a lesser extent, pressure of melting increase across the migmatitic terrane towards upper structural levels.

The entrapment of primary MI may be promoted by the presence of fine-grained minerals at the rims of the growing host, which may act as surfaces to which the melt droplet could cling (Roedder, 1984), and such “trapped minerals” may help to constrain the P-T conditions of anatexis (see Barich et al, 2014). In the investigated quartzo-feldspathic rocks, the presence of trapped sillimanite in MI (Figure 6e) supports the inferences obtained from phase equilibria modeling (i.e. garnet and melt were produced in the stability field of sillimanite). Conversely, pre-existing rutile crystals within MI from the pelitic granulites located structurally above the quartzo-feldspathic diatexites (Figure 12b) strongly supports the presence of anatectic melt at higher pressure, in agreement with independent geothermobarometric estimates from previous authors (Figure 13).

In the calculated pseudosections for the quartzo-feldspathic migmatites, after crossing


the fluid-saturated solidus the garnet and melt modes increase towards higher temperatures, supporting the peritectic nature of MI-bearing garnets (Figure 13). Conversely, the amount of biotite decreases (not shown) indicating that anatexis in the migmatitic terrane underlying the Ronda peridotites at Sierra Alpujata largely occurred by a continuous melting reaction consuming biotite, up to the Bt-out curve (Figure 13). Metatexites at the bottom of the anatectic sequence experienced low-temperature melting, and peritectic garnet in these rocks trapped the anatectic melt generated immediately after entering supersolidus conditions (Figure 13). Some microstructures reflect a prograde exhaustion of muscovite (Figure 3b, c) and the melt-in line coincides with the Ms-out curve at the pressure of interest (Figure 13). It

follows that muscovite could have participated as reactant to the melting reaction. However, the trace element contents of MI are needed to understand in detail the role of muscovite during anatexis, and in general the nature of melt-producing reaction in the metatexites (see Acosta-Vigil et al., 2010).

The high H2O contents (up to 9.8 wt%) observed in MI from the metatexite ALP1 and the constraints from phase equilibria modeling (i.e. that melt was produced immediately after crossing the fluid saturated solidus; Figure 13), indicate that partial melting at the bottom of the migmatitic sequence started in the presence of an H2O-rich fluid, likely produced by the subsolidus devolatilization of hydroxylated phases. The H2O contents (≈ 2-4 wt%) of MI from diatexite ALP13, however, are close to the values predicted for H2O-undersaturated granitic melts at conditions of interest (Holtz et al., 2001) and support the fluid-absent character of the melting event at higher structural levels.

An intriguing aspect of our study is that peritectic garnet in metatexite formed at ≈ 660-700 °C, in contrast with results from melting experiments that predict peritectic garnet generally developing above 800 °C by the fluid-absent melting of biotite (see Clemens, 2006 and references therein). Notably, these low temperatures are consistent with the complete experimental re-homogenization of nanogranites at 700 °C (Figure 10a). The results of our research indicate, therefore, that small amounts of peritectic garnet may be produced in natural metasediments starting from as low as 660-700 °C, well below Bt-out conditions. All the above observations renew the importance of studying MI to better constrain melting processes in crystalline basements.

Evolution of melt composition during prograde melting

A point to be addressed is whether MI are representative of the bulk anatectic melt in

Figure 13: Inferred P-T conditions of equilibration for the studied metatexite ALP1 and diatexite ALP13 (with data from Bartoli et al., 2013c). P-T estimates for pelitic granulites (i.e.rutile-bearing diatexites) are reported for comparison (data from Tubía et al., 1997). (b) Simplified section reported in Figure 2, showing the relative stratigraphic position of the rocks reported in (a). See text for explanation.


the system at the time of entrapment. Detailed studies on the geochemistry of MI in migmatites are just in their infancy (Cesare et al., 2011; Ferrero et al., 2012; Bartoli et al., 2013b, 2014), and there is little information available on the significance of MI compositions yet. On the other hand, MI hosted in minerals of anatectic metapelitic enclaves hosted in the felsic peraluminous lavas of the Neogene Volcanic Province of SE Spain, have been widely characterized from the geochemical point of view during the last 15 years (e.g. Cesare et al., 2003, 2007; Cesare, 2008; Acosta-Vigil et al., 2007, 2010, 2012; Ferrero et al., 2011). On the basis of a huge dataset of major and trace elements Acosta-Vigil et al. (2010, 2012) demonstrated that MI compositions do not represent exotic boundary layers, except for trace elements compatible with respect to the host phase, but rather reflect a melt reservoir with a well-defined geochemical signatures. Likewise, the analyses of MI presented here correspond to regular leucogranitic peraluminous compositions, similar to those of the MI in the El Hoyazo enclaves or to those of experimental glasses reported in the literature (see Clemens 2006 and references therein). Si-rich (SiO2 > 70%), peraluminous compositions have been also reported for MI in xenoliths from Vulcano Island (S Italy), and were interpreted as primary anatectic melts formed by melting of basement metamorphic rocks (Frezzotti et al., 2004). In addition the compositions of MI from metatexites and diatexites have a limited variability in the Qtz-Ab-Or normative diagram (Figure 11b) and vary as expected during increasing temperature along the prograde supersolidus path of these rocks (see below). Therefore, it is concluded that the major element compositions of MI presented here represent the bulk composition of the primary anatectic melt at the time of MI entrapment.

In order to compare the different types of anatectic melts observed in the studied

rocks (melt inclusions and leucosomes) and to highlight compositional similarities and differences, a series of bivariant diagrams have been created (Figure 14). Compositions of type II MI in diatexite often overlap those of glassy MI in metatexite (Figure 11, 14), suggesting that these rocks have garnet crystals/domains initially formed at lower T that trapped the earliest low-temperature melts produced in the mylonitic diatexites during their prograde melting. Higher Na2O/K2O values in type II MI of the diatexite and in those of the metatexite may also reflect higher aH2O values at the onset of melting, because the increasing a H2O depresses the plagioclase + quartz solidus more strongly than the stability of micas (Conrad et al., 1988; Patiño-Douce and Harris, 1998), consuming plagioclase in greater proportion than biotite during melting. The higher K contents of diatexite type I MI (Figures 11b, 14a) can be explained by the progressive consumption of biotite with rising temperature that produces an increase in the K2O content of the melt (Patiño-Douce and Johnston, 1991; Gardien et al., 2000). Although the FeO content of glassy MI increases from the metatexite to the diatexite (Figure 14b) in agreement with the increasing melting temperature (e.g. Patiño-Douce and Johnston, 1991), type I MI in the diatexite show a large spread in Fe concentrations (0.8-2.4 wt%) and seem to define two different clusters according to their Fe content (Figure 14b). The lack of a clear positive correlation between ASI, Mg# and FeO content argues against any significant melt-host garnet interaction (Figure 14c, d). Despite the major elements of MI provide important information on melting reactions (see above), a detailed characterization of the trace elements contents is needed to shed light on the controls on MI compositions and the mechanisms of crustal anatexis (e.g. melting reactions, role of accessory phases, equilibrium versus disequilibrium melting; see Acosta-Vigil et al., 2010, 2012).


The crustal sequence of Sierra Alpujata represents the first migmatitic terrane in which both MI and coexisting leucosomes have been analyzed and can be compared: although the compositions of leucosomes are leucogranitic and approach those of MI, there are differences. In the Qtz-Ab-Or normative diagram (Figure 11b), three leucosomes have a peraluminous leucogranitic compositions and plot close to the Qtz-Or cotectic, suggesting that they may approach unmodified (i.e. primary) anatectic melts. A fourth leucosome, far from the cotectic line, likely represents a composition modified by accumulation or fractionation processes. The

observation that the compositions of leucosomes are located away from the eutectic and closer to the Qtz-Or side when compared to the coexisting MI may indicate that MI record the evolution of melt composition during the early stages of anatexis, whereas leucosomes mostly reflect the composition of melt at, or closer to, the peak metamorphic conditions. However, all the investigated leucosomes show remarkably similar compositions in the variation diagrams of Figure 14, often not consistent with primary melts produced at higher temperatures than MI. Leucosome chemistry may be commonly affected by i) cumulus phenomena (Marchildon

Figure 14: Bivariant diagrams showing the compositions of MI and coexisting leucosomes. Symbols as in Figure 11. The compositions calculated by thermodynamic modeling (red and green asterisks) are plotted for comparison. (a) CaO vs. K# [K# =mol. K2O/(Na2O+K2O)]. (b) FeOT vs. K# (with all iron treated as FeO and reported as FeOT). (c) FeOT vs. ASI [ASI = mol. Al2O3/(CaO+Na2O+K2O)]. (d) FeOT vs. Mg# [Mg# =mol. MgO/(FeOt+MgO)].


and Brown, 2001), ii) fractional crystallization (Sawyer, 2008), iii) entrainment of peritectic phases (Stevens et al., 2007), iv) entrainment of crystals from the pre-anatectic framework (Sawyer, 2014), and v) diffusion of components towards the residue (White and Powell, 2010). The high CaO concentrations (Figure 14a, b) suggest that the investigated leucosomes likely contain xenocrysts of An-rich plagioclase from the rock matrix. Indeed, Sawyer (2014) has recently demonstrated that leucosomes may enclose entrained minerals as a consequence of their growth mechanism (i.e. rupture of the bridges of matrix). In addition, one leucosome could have been affected by the presence of residual quartz, as testified by the high content of normative Qtz (Figure 11a).

In the framework of thermodynamic modeling of phase equilibria, the calculation of melt compositions and their reintegration into the analyzed bulk composition has become a routine approach to study the prograde P-T evolution of anatectic rocks that have undergone melt loss (see White et al., 2004). We notice, however, that the melt model compositions calculated at the P-T conditions of interest differ in terms of K2O, Na2O, FeO, MgO and CaO when compared to MI, and that although the compositional departure decreases towards higher temperatures, model compositions never match the analyzed MI compositions (Figure 14). As recently suggested by White et al. (2011), the current melt model needs to be improved, and we suggest that such refinement may take advantage of the analytical database that is being obtained from anatectic MI (e.g. Cesare et al., 2011). In any case, and for those cases where MI are present in the residual anatectic rock, a more precise protolith composition (and therefore more constrained P-T estimations) may be obtained by reintegrating the composition of the MI hosted in peritectic phases.

Concluding remarks

Peritectic phases of migmatites and granulites may contain droplets of the coexisting melt that was being produced during incongruent crustal melting. The microstructural and compositional characterization of glassy and nanogranite inclusions hosted in peritectic garnet of stromatic metatexites and mylonitic diatexites from Sierra Alpujata (Betic Cordillera, S Spain) indicate that partial melting at the bottom of the migmatitic sequence started in the presence of an aqueous fluid phase, immediately after entering supersolidus conditions, and continued, particularly towards higher structural levels, largely under H2O-undersaturated conditions by progressive consumption of biotite, at increasing temperatures and, to a lesser extent, pressure. Notably, compositions of MI differ from those of leucosomes in the host rocks and of melts calculated from phase equilibra modeling.

Based on the study of MI in natural samples, our work confirms the conclusions of previous experimental and theoretical studies that partial melting of the metasedimentary crust produces peraluminous leucogranitic melts. However, our results also document that MI in peritectic minerals represent a unique tool to obtain in situ quantitative information on crustal anatexis, making accessible the precise melt composition for any anatectic terrane. It is increasingly important that petrologic studies of partially-melted crystalline basements integrate classic petrologic tools with results from MI investigation to better constrain melting processes. We believe that many occurrences of MI have been overlooked because they simply were not searched for, and that they will be uncovered by careful re-investigation of migmatites and granulites worldwide.


Acknowledgements

The authors thank Eugenio Fazio, Patrizia Fiannacca, Gaetano Ortolano, Rosalda Punturo (University of Catania) Davide Zanoni and Michele Zucali (University of Milano) for the invitation to participate in the workshop entitled “The art of deciphering structures and compositions: research advancements and investigation strategies in the study of crystalline basements” at the Congresso congiunto SGI-SIMP (2014), and to contribute to this special issue. The authors are grateful to Robert S. Darling, Stefan Jung and to the guest editor Patrizia Fiannacca for their comments, which improved this contribution. This research benefitted from funding from the Italian Ministry of Education, University, Research (grant PRIN 2010TT22SC to BC), from Padova University (Progetti per Giovani Studiosi 2013 to OB and Progetto di Ateneo CPDA107188/10 to BC) and from the Ministerio de Ciencia e Innovación of Spain (grant CGL2007-62992 to AAV). The research leading to these results has received funding from the European Commission, Seventh Framework Programme, under Grant Agreement nº 600376.

References

Acosta A., (1998) - Estudio de los fenómenos de fusión cortical y generación de granitoides asociados a las peridotitas de Ronda. Unpublished PhD Thesis, Universidad de Granada, 305 pp.

Acosta-Vigil A., Buick I., Cesare B., London D. and Morgan G.B. V.I. (2012) - The extent of equilibration between melt and residuum during regional anatexis and its implications for differentiation of the continental crust: a study of partially melted metapelitic enclaves. Journal of Petrology, 53, 1319-1356.

Acosta-Vigil A., Buick I., Hermann J., Cesare B. Rubatto D. London D. and Morgan VI G.B. (2010) - Mechanisms of crustal anatexis: a geochemical study of partially melted metapelitic enclaves and host

dacite, SE Spain. Journal of Petrology, 51, 785-821.Acosta-Vigil, A. Cesare, B. London D. and Morgan

G.B. VI (2007) - Microstructures and composition of melt inclusions in a crustal anatectic environment, represented by metapelitic enclaves within El Hoyazo dacites, SE Spain. Chemical Geology, 235, 450-465.

Acosta-Vigil A., Pereira M.D., Shaw D.M. and London D. (2001) - Contrasting behaviour of boron during crustal anatexis. Lithos, 56, 15-31.

Acosta-Vigil A., Rubatto D., Bartoli O., Cesare B., Meli S., Pedrera A., Azor A. and Tajcˇmanová L. (2014) - Age of anatexis in the crustal footwall of the Ronda peridotites, S Spain. Lithos, 210-211, 147-167.

Andrieux J., Fontbote J.M. and Mattauer M. (1971) - Sur un modèle explicatif de l’arc de Gibraltar. Earth and Planetary Science Letters, 12, 191-198.

Barich A., Acosta-Vigil A., Garrido C.J., Cesare B., Tajcˇmanová L. and Bartoli O. (2014) - Microstructures and petrology of melt inclusions in the anatectic sequence of Jubrique (Betic Cordillera, S Spain): Implications for crustal anatexis. Lithos, 206-207, 303-320.

Bartoli O. (2012) - When the continental crust melts: a combined study of melt inclusions and classical petrology on the Ronda migmatites. Unpublished PhD Thesis, Università degli Studi di Parma, 128 pp.

Bartoli O., Cesare B., Poli S., Acosta-Vigil A., Esposito R., Turina A., Bodnar R.J., Angel R.J. and Hunter J. (2013a) - Nanogranite inclusions in migmatitic garnet: behavior during piston cylinder re-melting experiments. Geofluids, 13, 405-420.

Bartoli O., Cesare B., Poli S., Bodnar R.J., Acosta-Vigil A., Frezzotti M.L. and Meli S. (2013b) - Recovering the composition of melt and the fluid regime at the onset of crustal anatexis and S-type granite formation. Geology, 41, 115-118.

Bartoli O., Cesare B., Remusat L., Acosta-Vigil A. and Poli S. (2014) - The H2O content of granite embryos. Earth and Planetary Science Letters, 395, 281-290.

Bartoli O., Tajcˇmanová L., Cesare B. and Acosta-Vigil A. (2013c) - Phase equilibria constraints on melting of stromatic migmatites from Ronda (S. Spain): insights on the formation of peritectic garnet. Journal of Metamorphic Geology, 31, 775-789.

Becker A., Holtz F., and Johannes W. (1998) - Liquidus temperatures and phase compositions in the system Qz-Ab-Or at 5 kbar and very low water activities. Contributions to Mineralogy and Petrology, 130, 213-224.


Brown M. (2010) - Melting of the continental crust during orogenesis: the thermal, rheological, and compositional consequences of melt transport from lower to upper continental crust. Canadian Journal of Earth Sciences, 47, 655-694.

Brown M. (2013) - Granite: from genesis to emplacement. Geological Society of America Bulletin, 125, 1079-1113.

Brown M. and Korhonen F. J. (2009) - Some remarks on melting and extreme metamorphism of crustal rocks. Physics and Chemistry of the Earth’s Interior, 67-87.

Brown M., Korhonen F.J. and Siddoway C.S. (2011) - Organizing melt flow through the crust. Elements, 7, 261-266.

Brown M. and Rushmer T. (2006) - Evolution and Differentiation of the Continental Crust. Cambridge University Press, Cambridge. 553 pp.

Cesare B. (2008) - Crustal melting: working with enclaves. In: Working with Migmatites. (eds): E.W. Sawyer and M. Brown. Mineralogical Association of Canada, Quebec City, Short Course 38, pp. 37-55.

Cesare B., Acosta-Vigil A., Ferrero S. and Bartoli O. (2011) - Melt inclusions in migmatites and granulites. In: The Science of Microstructure - Part II. (eds): M.A. Forster and J.D. Fitz Gerald, Journal of Virtual Explorer, 38, paper 2, Electronic edition.

Cesare B., Ferrero S., Salvioli-Mariani E., Pedron D. and Cavallo A. (2009) - Nanogranite and glassy inclusions: the anatectic melt in migmatites and granulites. Geology, 37, 627-630.

Cesare B., Maineri C., Baron Toaldo A., Pedron D. and Acosta-Vigil A. (2007) - Immiscibility between carbonic fluids and granitic melts during crustal anatexis: a fluid and melt inclusion study in the enclaves of the Neogene Volcanic Province of SE Spain. Chemical Geology, 237, 433-449.

Cesare B., Marchesi C., Hermann J. and Gomez-Pugnaire M.T. (2003) - Primary melt inclusions in andalusite from anatectic graphitic metapelites: implications for the position of the Al2SiO5 triple point. Geology, 31, 573-576.

Clemens J.D. (2006) - Melting of the continental crust: fluid regimes, melting reactions and source-rock fertility. In: Evolution and Differentiation of the Continental Crust. (eds): M. Brown and T. Rushmer. Cambridge University Press, Cambridge, pp. 297-331.

Clemens J.D., Birch W.D. and Dudley R.A. (2011) - S-type ignimbrites with polybaric crystallisation

histories: the Tolmie Igneous Complex, Central Victoria, Australia. Contributions to Mineralogy and Petrology, 162-6, 1315-1337.

Clemens J.D. and Stevens G. (2012) - What controls chemical variation in granitic magmas? Lithos, 134-135, 317-329.

Connolly J.A.D. (1995) - Phase diagrams for graphitic rocks. Contributions to Mineralogy and Petrology, 119, 94-116.

Conrad W.K., Nicholls I.A. and Wall V.J. (1988) - Water-saturated and -undersaturated melting of metaluminous and peraluminous crustal compositions at 10 kb: evidence for the origin of silicic magmas in the Taupo Volcanic Zone, New Zealand, and other occurrences. Journal of Petrology, 29, 765-803.

Darling R.S. (2013) - Zircon-bearing, crystallized melt inclusions in peritectic garnet from the Western Adirondack Mountains, New York State, USA. Geofluids, 13, 453-459.

Dewey J.F., Helman M.L., Turco E., Hutton D.H.W. and Knott S.D. (1989) - Kinematics of the Western Mediterranean. In: Alpine Tectonics. (eds): M.P. Coward, D. Dietrich, and R.G. Park, Special Publication, Geological Society of London, pp. 265-283.

Esposito E., Klebesz R., Bartoli O., Klyukin Y.I., Moncada D., Doherty A. and Bodnar R.J. (2012) -Application of the Linkam TS1400XY heating stage to melt inclusion studies. Central European Journal of Geosciences, 4/2, 208-18.

Esteban J.J., Cuevas J., Tubía J.M., Sergeev S. and Larionov A. (2011a) - A revised Aquitanian age for the emplacement of the Ronda peridotites (Betic Cordilleras, southern Spain). Geological Magazine, 148, 183-187.

Esteban J.J., Cuevas J., Vegas N. and Tubía J.M. (2008) - Deformation and kinematics in a melt-bearing shear zone from the Western Betic Cordilleras (Southern Spain). Journal of Structural Geology, 30, 380-393.

Esteban J.J., Tubía J.M., Cuevas J., Vegas N., Sergeev S. and Larionov A. (2011b) - Peri-Gondwanan provenance of pre- Triassic metamorphic sequences in the western Alpujarride nappes (Betic Cordillera, southern Spain). Gondwana Research, 20, 443-449.

Ferrero S., Bartoli O., Cesare B., Salvioli-Mariani E., Acosta-Vigil A., Cavallo A., Groppo C. and Battiston S. (2012) - Microstructures of melt inclusions in anatectic metasedimentary rocks. Journal of Metamorphic Geology, 30, 303-322.


Ferrero S., Bodnar R.J., Cesare B. and Viti, C. (2011) - Reequilibration of primary fluid inclusions in peritectic garnet from metapelitic enclaves, El Hoyazo, Spain. Lithos, 124, 117-131.

Frezzotti M.L., Peccerillo A., Zanon V. and Nikogosian I. (2004) - Silica-rich melts in quartz xenoliths from Vulcano Island and their bearing on processes of crustal anatexis and crust-magma interaction beneath the Aeolian Arc, Southern Italy. Journal of Petrology, 45, 3-26.

Gardien V., Thompson A. B. and Ulmer P. (2000) - Melting of biotite + plagioclase + quartz gneisses: the role of H2O in the stability of amphibole. Journal of Petrology, 41/5, 651-666.

Holness M.B., Cesare B. and Sawyer E.W. (2011) - Melted rocks under the microscope: microstructures and their interpretation. Elements, 7, 245-250.

Holness M.B. and Sawyer E.W. (2008) - On the pseudomorphing of melt-filled pores during the crystallization of migmatites. Journal of Petrology, 49, 1343-1363.

Holtz F., Johannes W., Tamic N. and Behrens H. (2001) - Maximum and minimum water contents of granitic melts generated in the crust: a re-evaluation and implications. Lithos, 56, 1-14.

Kretz R. (1983) - Symbols for rock-forming minerals. American Mineralogist, 68, 277-279.

Kriegsman L.M. and Hensen B.J. (1998) - Back reaction between restite and melt: Implications for geothermobarometry and pressure-temperature paths. Geology, 26, 1111-1114.

Loomis T.P. (1972) - Diapiric emplacement of the Ronda high-temperature ultramafic intrusion, southern Spain. Geological Society of America Bulletin, 83, 2475-2496.

Lowenstern J. (1995) - Applications of silicate-melt inclusions to the study of magmatic volatiles. In: Magmas, Fluids and Ore Deposits. (eds): J.F.H. Thompson. Mineralogical Association of Canada Short Course, Victoria. 23, pp. 71-99.

Lundeen, M.T. (1978) - Emplacement of the Ronda peridotite, Sierra Bermeja, Spain. Geological Society of America Bulletin, 89, 172-180.

Massonne H.-J. (2014) - Wealth of P-T-t information in medium-high grade metapelites: example from the Jubrique Unit of the Betic Cordillera, S Spain. Lithos, 208-209, 137-157.

Marchildon N. and Brown M. (2001) - Melt segregation in late-tectonic anatectic migmatites: an example

from the Onawa contact aureole, Maine, U.S.A. Physics and Chemistry of the Earth, A, 26, 225-229.

Obata M. (1980) - The Ronda peridotite: garnet-, spinel-, and plagioclase-lherzolite facies and the P-T trajectories of a high temperature mantle intrusion. Journal of Petrology, 21, 533- 572.

O’ Connor J.T. (1965) - A classification for quartz-rich igneous rocks based on feldspar ratios. U.S. Geological Survey Professionnal Paper 525, 79-84 (B).

Patiño-Douce A.E. and Harris N. (1998) - Experimental constraints on Himalayan anatexis. Journal of Petrology, 39, 689-710.

Patiño-Douce A.E., and Johnston A.D. (1991) - Phase equilibria and melt productivity in the pelitic system: implications for the origin of peraluminous granitoids and aluminous granulites. Contributions to Mineralogy and Petrology, 107, 202-218.

Pallister J.S., Hoblitt R.P., and Reyes A.G. (1992) - A basalt trigger for the 1991 eruptions of Pinatubo volcano. Nature, 356, 426-428.

Petford, A.R. Cruden, K.J.W. McCaffrey and J.-L. Vigneresse (2000) - Granite magma formation, transport and emplacement in the Earth’s crust. Nature, 408, 669-673.

Platt J.P., Argles T.W., Carter A., Kelley S.P., Whitehouse M.J. and Lonergan L. (2003) - Exhumation of the Ronda peridotite and its crustal envelope: Constraints from thermal modelling of a P-T-time array. Journal of the Geological Society of London, 160, 655-676.

Platt J.P., Behr W.M., Johanesen K. and Williams J.R. (2013) - The Betic-Rif arc and its orogenic hinterland: A review. Annual Review of Earth and Planetary Sciences, 41, 14.1-14.45.

Roedder E. (1984) - Fluid inclusions. Mineralogical Society of America, Reviews in Mineralogy, 12, 644 pp.

Sánchez-Rodríguez L. and Gebauer D. (2000) - Mesozoic formation of pyroxenites and gabbros in the Ronda area (southern Spain), followed by Early Miocene subduction metamorphism and emplacement into the middle crust: U-Pb sensitive high-resolution ion microprobe dating of zircon. Tectonophysics, 316, 19-44.

Sawyer E.W. (2001) - Melt segregation in the continental crust: distribution and movement of melt in anatectic rocks. Journal of Metamorphic Geology, 19, 291-309.

Sawyer E.W. (2008) - Atlas of Migmatites. The


Canadian Mineralogist Special Publication 9, NRC Research Press, Ottawa, Ontario, Canada.

Sawyer E.W. (2014) - The inception and growth of leucosomes: microstructure at the start of melt segregation in migmatites. Journal of Metamorphic Geology, 32, 695-712.

Sawyer E.W., Cesare B. and Brown M. (2011) - When the continental crust melts. Elements, 7, 229-234.

Stevens G., Villaros A. and Moyen J.F. (2007) - Selective peritectic garnet entrainment as the origin of geochemical diversity in S-type granites. Geology, 35, 9-12.

Stipp M., Stünitz H., Heilbronner R. and Schmid S.M. (2002) - The eastern Tonale fault zone: a ‘natural laboratory’ for crystal plastic deformation of quartz over a temperature range from 250 to 700 °C. Journal of Structural Geology, 24, 1861-1884.

Thompson A.B. (1982) - Dehydration melting of pelitic rocks and the generation of H2O-undersaturated granitic liquids. American Journal of Science, 282, 1567-1595.

Torres-Roldán R.L. (1981) Plurifacial metamorphic evolution of the Sierra Bermeja peridotite aureole (southern Spain). Estudios Geológicos, 37, 115-133.

Torres-Roldán R.L. (1983) - Fractionated melting of metapelite and further crystal-melt equillibria. The example of the Blanca Unit migmatite complex, north of Estepona (southern Spain). Tectonophysics, 96, 95-123.

Tubía J.M. (1988) - Estructura de los Alpuja_rrides occidentales: Cinematica y condiciones de emplazamiento de las peridotitas de Ronda. Publicaciones Especiales del Boletın Geologico y Minero de Espana, Madrid, 124 pp. (in Spanish).

Tubía J.M., Cuevas J. and Gil-Ibarguchi J.I. (1997) - Sequential development of the metamorphic aureole beneath the Ronda peridotites and its bearing on the tectonic evolution of the Betic Cordillera. Tectonophysics, 279, 227-252.

Tubía J.M., Cuevas J. and Esteban J.J. (2013) - Localization of deformation and kinematics shift during the hot emplacement of the Ronda peridotites (Betic Cordilleras, southern Spain). Journal of Structural Geology, 50, 148-160.

Tubía J.M. and Gil-Ibarguchi J.I. (1991) - Eclogites of the Ojen nappe: a record of subduction in the Alpujarride complex (Betic Cordilleras, southern Spain). Journal of the Geological Society, 148, 801-804.

Vanderhaeghe O. (2001) - Melt segregation, pervasive melt migration and magma mobility in the continental crust: The structural record from pores to orogens. Physics and Chemistry of the Earth, Part A: Solid Earth and Geodesy, 26, 213-223.

Vernon R.H. (2011) - Microstructures of melt-bearing regional metamorphic rocks. In: Origin and Evolution of Precambrian High-Grade Gneiss Terranes, with Special Emphasis on the Limpopo Complex of Southern Africa. (eds): D.D. Van Reenen, J.D. Kramers, S. McCourt, and L.L. Perchuk. Geological Society of America Memoir, 207, 1-11.

Vielzeuf D., Clemens J.D., Pin C. and Moinet E. (1990) - Granites, granulites and crustal differentiation. In: Granulites and crustal differentiation. (eds): D. Vielzeuf and P. Vidal. Dordrecht, Kluwer Academic Publishers, p. 59-85.

Westerhof A.B. (1977) - On the contact relations of high-temperature peridotites in the Serrania de Ronda, southern Spain. Tectonophysics, 39, 579-591.

White R.W. and Powell R. (2010) - Retrograde melt-residue interaction and the formation of near-anhydrous leucosomes in migmatites. Journal of Metamorphic Geology, 28, 579-597.

White R.W., Powell R. and Halpin J.A. (2004) - Spatially-focussed melt formation in aluminous metapelites from Broken Hill, Australia. Journal of Metamorphic Geology, 22 825-845.

White R.W., Stevens G. and Johnson T.E. (2011) - Is the crucible reproducible? Reconciling melting experiments with thermodynamic calculations. Elements, 7, 241-246.

Wilke S., Klahn C., Bolte T., Almeev R., Holtz F. (2015) - Experimental investigation of the effect of Ca, Fe and Ti on cotectic compositions of the rhyolitic system. European Journal of Mineralogy, DOI: 10.1127/ejm/2015/0027-2423.

Submitted, March 2015 - Accepted, September 2015

high-temperature metamorphism and crustal melting: working...

Documents