Journal of Petrology

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (7) Request Permissions Google Scholar Articles by FORNELLI, A. Articles by ACQUAFREDDA, P. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

Journal of Petrology | Volume 43 | Number 12 | Pages 2191-2217 | 2002
© Oxford University Press 2002

Multi-stage Melting in the Lower Crust of the Serre (Southern Italy)

A. FORNELLI^1,*, G. PICCARRETA¹, A. DEL MORO² and P. ACQUAFREDDA¹

¹DIPARTIMENTO GEOMINERALOGICO, BARI UNIVERSITY, VIA E. ORABONA, 4, 70124 BARI, ITALY
²ISTITUTO DI GEOCRONOLOGIA E GEOCHIMICA ISOTOPICA (CNR), VIA ALFIERI, 1, 36 56100 PISA, ITALY

Received January 7, 2001; Revised typescript accepted May 6, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION GEOLOGICAL BACKGROUND ANALYTICAL METHODS PETROGRAPHY WHOLE-ROCK CHEMISTRY DISCUSSION CONCLUSION AND GEOLOGICAL... REFERENCES

 
The lower-crustal section exposed in the Serre, southern Italy, consists mainly of Al-rich metasediments, which underwent granulite-facies metamorphism, partial melting and melt extraction. The paper considers the formation of melts in metapelites and metagreywackes. Leucosomes and host rocks have been studied to investigate the melting process. Biotite-rich and biotite-free melanosomes with scarce felsic components are present; the biotite-rich types are widespread in the upper part of the section and the two types may occur side by side in the lower part. Na-rich and K-rich leucosomes including residual phases are interspersed within the metasediments; on the whole they do not show geochemical signatures suggestive of magmatic fractionation. Leucotonalitic types prevail among the sampled leucosomes, which generally are rare earth element (REE) depleted with positive Eu anomalies whereas the host rocks are REE enriched with overall negative Eu anomalies. Melanosomes and migmatites show restitic chemistries. The precursor metagreywackes underwent depletion in Na₂O and enrichment in K₂O. The precursor metapelites document general depletion in Na₂O and they may be enriched or depleted in K₂O. All the characteristics of the migmatites and of their components reflect a two-stage melting: (1) H₂O-present melting, involving mainly plagioclase, and (2) dehydration melting of micas. All the metasediments underwent H₂O-present melting, forming mostly sodic melts which, owing to their removal from the source as fast as they formed, did not accumulate in such proportions as to allow migration and mostly remained within the lower-crustal metasediments; metapelites also underwent variable dehydration melting, depending on chemical features and physical conditions, forming larger volumes of mobile granitic melts, most of which migrated far from the source. Extractions of 57–66 vol. % of total melts (sodic + potassic) from the most residual metapelitic melanosomes and of about 27–44 vol. % of potassic melts from metapelitic migmatites have been calculated. Higher volumes of the extracted melts have been calculated for the metapelites of the lower part of the section; the most depleted metagreywackes underwent melt extraction of about 9–13 vol. %. The two-stage melting occurred during the prograde metamorphism and continued during the isothermal decompression.

KEY WORDS: Calabria; lower crust; multi-stage melting

	INTRODUCTION

TOP ABSTRACT INTRODUCTION GEOLOGICAL BACKGROUND ANALYTICAL METHODS PETROGRAPHY WHOLE-ROCK CHEMISTRY DISCUSSION CONCLUSION AND GEOLOGICAL... REFERENCES

 
Studies on anatectic terranes often focus on obtaining information about the intensive parameters controlling high-grade metamorphism and can provide a deep understanding of processes responsible for leucosome formation. Two fundamental processes have been proposed for the origin of migmatites: partial melting and subsolidus migmatization. The knowledge of the geological context, such as defining the role played by P–T and fluid composition as well as by the composition of the protoliths and their mineralogical components, are all needed to decipher the leucosome-forming processes. The lower continental crust is generally considered as a fertile reservoir for the production of granitic melts (e.g. Clemens & Vielzeuf, 1987; Le Breton & Thompson, 1988; Vielzeuf & Holloway, 1988; Vielzeuf et al., 1990; Patiño Douce & Johnston, 1991; Skjerlie & Johnston, 1996). Over the temperature range between 700 and 900°C, metapelites, metagreywackes and amphibolites can all melt. The proportions of melts range from 20 to 60% depending on the compositions of protoliths and P–T–X_H2O conditions (e.g. Clemens & Vielzeuf, 1987; Patiño Douce & Johnston, 1991; Patiño Douce & Beard, 1995).

Under fluid-absent conditions, granitic melts form through incongruent dehydration-melting reactions, from metapelites and metagreywackes (e.g. Le Breton & Thompson, 1988; Vielzeuf & Holloway, 1988; Vielzeuf et al., 1990; Patiño Douce & Johnston, 1991; Thompson & Connolly, 1995). Under fluid-present conditions and in the presence of muscovite, leucotonalitic–trondhjemitic melts may form from Al-rich metasediments at P >6 kbar (e.g. Conrad et al., 1988; Patiño Douce & Harris, 1998).

The partial melting process affecting the lower continental crust is constrained by: (1) the fact that partial melting occurs under conditions characteristic of upper amphibolite–granulite facies; (2) the widespread occurrence of migmatites; (3) occurrence of rocks more or less depleted in incompatible elements interpreted as residues of partial melting complementary to granitic magmas (e.g. Clemens, 1990; Vielzeuf et al., 1990). Once produced, melts can either crystallize in situ, or migrate over variable distances from their source. The extent to which melt migration from the source occurs plays a key role in the differentiation of the continental crust. In addition, there is a growing body of evidence pointing to the occurrence of leucosomes having moderate rare earth element (REE) contents and positive Eu anomaly in high-grade terranes (e.g. Barbey et al., 1990; Watt & Harley, 1993; Carrington & Watt, 1995; Johannes et al., 1995). Among these leucosomes, leucotonalitic and trondhjemitic compositions are frequent in granulitic terranes (e.g. Conrad et al., 1988; Sawyer, 1991; Mazzucchelli et al., 1992; Whitney & Irving, 1994).

The purpose of this paper is: (1) to investigate the nature and origin of the leucosomes and of melanosomes occurring in a portion of the lower-crustal section exposed in the Serre, southern Italy (Fig. 1); (2) to derive geological implications combining the results of this study with P–T–t lower-crustal path. Petrographic and geochemical data as well as mass balance computations are considered. This work builds on previous geological and petrological studies that address the composition and the metamorphic evolution of these lower-crustal rocks as well as the occurrence of partial melting and extraction of melts (e.g. Paglionico & Piccarreta, 1978; Maccarrone et al., 1983; Schenk, 1984, 1989, 1990; Caggianelli et al., 1991; Del Moro et al., 2000). Melanosomes, mesosomes and leucosomes, including several leucosome–host-rock pairs, have been studied with attention placed on metasediments having pelites, semipelites and greywackes as protoliths, which are interlayered with felsic granulites and form most of the metapelite unit that overlies such felsic granulites (Fig. 1). The samples have been taken along north–south profiles (Fig. 1): (1) north of the univariant zone Grt–Crd–Bt–Sil–Kfs–Qtz mapped by Schenk (1990); (2) straddling and south of the univariant zone. These are distinguished as lower and upper samples, respectively. South of the univariant zone only one profile has been sampled (Fig. 1).

View larger version (36K): [in this window] [in a new window]   Fig. 1. Geological sketch map of the Serre. (1) Lower-crustal rocks; (2) calc-alkaline and peraluminous granitoids. Rectangles indicate the sampled zones.

 

	GEOLOGICAL BACKGROUND

TOP ABSTRACT INTRODUCTION GEOLOGICAL BACKGROUND ANALYTICAL METHODS PETROGRAPHY WHOLE-ROCK CHEMISTRY DISCUSSION CONCLUSION AND GEOLOGICAL... REFERENCES

 
In the Serre, southern Italy, a section of the Hercynian continental crust is exposed (e.g. Schenk, 1980; Rottura et al., 1990). This section, consisting of intermediate–lower-crustal and shallow crustal units sutured by huge masses of calc-alkaline and peraluminous granitoids (Fig. 1), overthrusts phyllonites of former amphibolite-facies rocks, which, in turn, lie upon a phyllitic unit (Paglionico & Piccarreta, 1976). The calc-alkaline granitoids form deep to shallow intrusions from north to south (Fig.1). Caggianelli et al. (1997), using the ‘Al-in hornblende’ geobarometer, showed that the foliated tonalites (Fig. 1) were emplaced at a depth of 20–25 km whereas the southernmost plutons were emplaced at a depth of a few kilometres. Also, the peraluminous granites intruded at different depths with different characteristics: (1) the Cittanova peraluminous granite (south of the mapped area in Fig. 1) intrudes, with sharp contact, the shallowest calc-alkaline plutonics (Crisci et al., 1979); (2) the Petrizzi peraluminous granite (Fig.1) intrudes the partially molten tonalites (Caggianelli, 1988); (3) small and well-defined stocks of cordierite microgranodiorite are present within the granodiorites of the central Serre (Fornelli, 1991; Fornelli et al., 1994); (4) crustal melts contaminated the granodiorites, which include fragments of Crd–Sil–Spl–Crn residues of melting (Fornelli, 1991). Within the border zone of the tonalites, peraluminous granites, gabbros and gabbrodiorites form small masses (e.g. Caggianelli, 1988).

From the lowest section, namely, from the highest metamorphic grade, the exposed lower crust consists of felsic and mafic granulites (with minor intercalations of metapelites, metagreywackes and large bodies of layered metagabbros as well as slices of meta-ultramafites), which underlie migmatitic metapelites and metagreywackes (with minor intercalations of mafic granulites and marbles). Concordant and discordant leucosomes (Fig. 2) are interspersed along the section.

View larger version (144K): [in this window] [in a new window]   Fig. 2. Concordant and discordant leucosomes in migmatic metapelites.

 

Metagabbros, granulites and overlying migmatites are structurally concordant and show the same high-grade foliation (see also Schenk, 1984; Kruhl & Huntemann, 1991). The rocks have experienced a complex pre-Alpine history (e.g. Dubois, 1976; Paglionico & Piccarreta, 1978) interpreted in terms of a simple Hercynian clockwise P–T–t path by Schenk (1984, 1990). According to Schenk (1980, 1990) the metamorphic peak was reached under static conditions at 300 ± 10 Ma (U–Pb monazite and zircon ages) and is roughly contemporaneous with the intrusion of calc-alkaline granitoids [293–295 Ma according to Schenk (1990)]; it was followed by isothermal decompression and finally by isobaric cooling. Peak metamorphic conditions of 7–8 kbar and 800°C are estimated for the base of the section and of 5–6 kbar and 700°C for the top; the isothermal decompression was characterized by a pressure decrease of 2 kbar (e.g. Schenk, 1984, 1990). The present volume of leucosomes in migmatitic paragneisses increases from 10 to 40 vol. % towards the upper part of the section (Caggianelli et al., 1991) where, in areas close to tonalites, they form larger bodies including randomly distributed centimetre- to decimetre-sized blocks of biotite–sillimanite restite (solid residue of partial melting). Quantitative modelling has been performed utilizing major elements (Schenk, 1990) and also trace elements (Caggianelli et al., 1991), and assuming extraction of melts having the composition of Calabrian peraluminous granites (Schenk, 1990; Caggianelli et al., 1991). The computation indicates 40 vol. % of peraluminous granitic melts extracted from migmatites (Schenk, 1990) and 60 vol. % of total melts (in situ leucosomes plus missing granite) extracted from the most residual melanosomes (Caggianelli et al., 1991). The leucosomes occurring in the migmatites of this study are one- or two-feldspar bearing (e.g. Maccarrone et al., 1983; Caggianelli et al., 1991).

	ANALYTICAL METHODS

TOP ABSTRACT INTRODUCTION GEOLOGICAL BACKGROUND ANALYTICAL METHODS PETROGRAPHY WHOLE-ROCK CHEMISTRY DISCUSSION CONCLUSION AND GEOLOGICAL... REFERENCES

 
Representative samples of all the studied rocks were selected for chemical analyses. The >4 cm thick leucosomes were carefully separated from melanosomes, distinguishing discordant and concordant leucosomes and considering their geometric position along the sampled segments of the lower-crustal section (Fig. 1). The mineral analyses were performed using a Cambridge S360 electron microscope equipped with a LINK AN 10.000 Si (Li) energy dispersive detector. Operating conditions were 15 kV accelerating potential and a probe current of about 1 nA. EDS intensities were converted into wt % oxides by ZAF 4/FLS quantitative analysis software support.

Major and trace element analyses were performed on all samples by means of X-ray fluorescence (XRF) at Bari University. REE concentrations were determined by inductively coupled plasma mass spectrometry (ICP-MS) at the CRPG laboratory in Nancy, France. They show uncertainties of ±5% except for Lu (16%). Analytical methods for the determination of ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd ratios have been given by Del Moro et al. (2000).

	PETROGRAPHY

TOP ABSTRACT INTRODUCTION GEOLOGICAL BACKGROUND ANALYTICAL METHODS PETROGRAPHY WHOLE-ROCK CHEMISTRY DISCUSSION CONCLUSION AND GEOLOGICAL... REFERENCES

 
Ashworth’s (1985) migmatite nomenclature is used in this paper. Mesosome, leucosome and melanosome are used as descriptive terms; mesosomes and melanosomes were distinguished on the basis of field appearance and petrography. Abbreviations after Kretz (1983) have been used for minerals. On the whole, the exposed migmatites along the portion of the lower-crustal section of the Serre studied here (Fig. 1) are stromatic with a dominant planar structure surviving partial melting and with mineralogically distinct layers. The majority of leucosomes are parallel to the pre-migmatization structure (e.g. layering) in their host rocks, although some are cross-cutting. The migmatites are composed of millimetre- to decimetre-spaced discontinuous bands formed by alternating melanocratic, mesocratic and leucocratic bodies (Fig. 2). The leucocratic bodies, on the whole, are more abundant in the upper part of the investigated crustal section (Fig. 1). The presence or absence of sillimanite in the mesocratic and melanocratic bodies suggests mostly derivation from pelite to greywacke protoliths or from mixtures of these. All the studied metasediments are primary muscovite free. Garnet (up to 40–50 vol. %)–sillimanite (up to 40 vol. %)-rich and biotite-poor or -free mesosomic–melanosomic metapelites are more common towards the base of the metapelite unit; biotite-rich types (up to >30–40 vol. %) are more common in the upper part; rock-types derived from the metagreywackes do not show preferential spatial distribution along the studied profiles (Fig. 1).

Leucosomes
The leucosomes are fine to coarse grained. When undeformed, they show magmatic textures (Figs. 3a and b); when deformed, they show the same foliation as the host rock. They are leucotonalitic to granitic in composition; intermediate granodioritic compositions that occur in some localities are due to interaction of granitic and leucotonalitic end-members (LFC12), as some strings and layers consist of distinct portions that are either rich in Kfs or in Pl or are made up of Qtz + two feldspars. A given leucosome composition may form concordant and discordant or deformed and undeformed bodies. Leucosomes are composed of quartz, plagioclase and K-feldspar in different proportions: Pl may be the only feldspar in some leucotonalitic types; Kfs exceeds 50 vol. % within some granitic bodies; Qtz, independently of the composition of the rocks, may exceed 60 vol. %. Biotite, garnet, sillimanite and, more rarely, cordierite may occur in leucosomes associated with rock-types derived from metapelites; biotite, garnet and, rarely, orthopyroxene are present in leucosomes hosted by melanosomic and mesosomic metagreywackes. Zircon, monazite, apatite, ilmenite and rutile are enclosed within the large crystals of biotite and garnet. Biotite and garnet are also present as small crystals. Most of these minerals, in single crystals or in small clusters having the same assemblage as the host rock, represent phases entrained in the melts on the basis of the textural (size and kind of inclusions) and compositional (see below) similarities in the leucosome–host-rock pairs. This seems to indicate that melt segregation was not completely efficient and that melt did not migrate over long distances; some migration did occur, however, as indicated by the discordant leucosomes and by the variable concentration of leucosomes, at outcrop scale, owing to melt loss or gain.

View larger version (102K): [in this window] [in a new window]   Fig. 3. Granitic (a) and leucotonalitic (b) leucosomes showing magmatic textures.

 

The granitic bodies associated with metapelites are composed of large subhedral and embayed crystals of Kfs and more rarely Pl with interstitial fine- to medium-grained Qtz, Pl and Kfs. The large crystals of Kfs include small euhedral crystals of Grt (as single grains or as clusters), Bt, Sil and Pl. Within the granitic leucosome LFC6, associated with a mesosomic metagreywacke, there are no entrained mafic phases and the large crystals of Kfs include myrmekitic plagioclase.

The leucotonalitic leucosomes are medium grained or heterogranular; the latter are characterized by large subhedral crystals of plagioclase within a medium-grained ‘matrix’ composed of Qtz + Pl ± Kfs. They contain minor Sil, Bt and Grt, except for sample KIS16.

Melanosomes and mesosomes
The melanosomes comprise (1) massive or foliated rock-types having metapelitic precursors and (2) foliated rocks derived from metagreywackes.

The sampled melanosomic metapelites consist of garnet, biotite, sillimanite and sometimes cordierite and minor feldspars (mostly plagioclase) and quartz. They include rock-types containing abundant biotite and scarce feldspars and rocks in which biotite and/or feldspars are no longer present. Garnet forms large crystals including accessories together with remains of biotite, sillimanite, plagioclase and quartz and small crystals or intergranular extensions.

Bt + Pl + Qtz + Grt ± Opx occur in the melanosomic metagreywackes. Biotite is ubiquitous, quartz and plagioclase occur in variable quantities; garnet, biotite and sometimes orthopyroxene may form thin layers.

Zircon, monazite, apatite, rutile, ilmenite, pyrrhotite and graphite are the common accessory phases; hercynite may be present in melanosomic metapelites. Zircon, monazite, apatite, ilmenite and rutile form inclusions in large crystals of biotite and garnet; hercynite may form inclusions in the large crystals of sillimanite.

The mesosomic metapelites and metagreywackes have the same structural and compositional characteristics as the melanosomic types, except for the interspersed leucosomes. The sampled mesosomes and melanosomes bear evidence of later retrograde rehydration as documented, for instance, by the formation of biotite from garnet. The retrograde rehydration appears more intense within the migmatites of the upper part of the section, where also the leucosomes are more abundant. Hence, the melanosomic types may be cumulations of refractory material, peritectic phases left behind by fluxing melts and retrograde phases owing to release of fluids during their solidification.

Mineral chemistry
Biotite, garnet and plagioclase have been analysed both in leucosomes and in host rocks to compare their compositions and to define their significance.

Biotite forms large and poikilitic crystals within both host rocks and leucosomes; in the latter few small crystals intergrown with quartz and feldspars may also occur. Biotite is generally Ti rich (0·38–0·79 a.p.f.u.). The ternary plot of Fig. 4 shows that the biotites within host rocks derived from metapelites and within the associated leucosomes are richer in Al₂O₃ than those occurring within mesocratic and melanocratic metagreywackes and the associated leucosomes. The X_Mg [Mg/(Mg + Fe)] in the large crystals present both in the host rocks and in the leucosomes is variable within the same range (0·67–0·78). The X_Mg of the small crystals intergrown with quartz and feldspars within the leucosomes is lower (0·49–0·67). Therefore the former can be interpreted as a relict phase, and the latter as formed during migmatization or as a magmatic phase.

View larger version (12K): [in this window] [in a new window]   Fig. 4. Ternary plot showing the chemical composition of biotite in host rocks and leucosomes: •, large crystals within the metagreywacke–leucosome pair; , large crystals within the metapelite–leucosome pair; , , composition of the small crystals within the leucosomes.

 

Garnet forms large crystals (1–6 mm) including Bt, Pl, Qtz, Rt–Ilm and accessories in all the studied rocks; in the melanosomic and mesosomic metapelites and within the associated leucosomes it forms also euhedral small crystals (<1 mm). Within the granitic leucosomes from the lower metapelitic migmatites, the small euhedral garnet may be included, as single crystals or aggregates of crystals, within the large feldspars. Within the host-rock–leucosome pairs garnet shows variable X_Mg (0·24–0·40) and Ca (0·08–0·25 a.p.f.u.) between samples but similar compositions in individual pairs (Fig. 5). The large crystals show: (1) a moderate zoning within the metapelite–granitic leucosome pair sampled at the bottom of the section, with rim–core X_Mg ranging from 0·27 to 0·40 and Ca from 0·08 to 0·16 (a.p.f.u.), and fairly homogeneous compositions within the pairs sampled at the top of the section; (2) X_Mg ranging from 0·34 to 0·40 and Ca from 0·14 to 0·25 (a.p.f.u.) within the restitic metagreywacke–leucosome pairs. The small crystals analysed in the metapelite–granitic leucosome pair from the lower part of the section, including also those occurring within the feldspars, have X_Mg 0·24–0·34. The small crystals analysed within the leucotonalitic leucosomes associated with metapelites have roughly the same composition as the core of the large crystals (Fig. 5). The core–rim compositional variation in the larger crystals and the similarity of compositions of the rims and of the small crystals (Fig. 5), within metapelite–granitic leucosome pair from the lower part of the section, are interpreted as reflecting formation of garnet as subsolidus to peritectic phase; the presence of peritectic garnet is compatible with the participation of biotite in the melting process (see below). Garnets occurring within the leucotonalitic leucosomes associated with the metagreywackes and metapelites are interpreted as relict phases. In fact they, on the whole, have the same textural and compositional features as in the host rocks (Fig. 5) and, as will be shown below, the leucotonalitic melts of the studied section formed within the stability field of biotite without production of peritectic garnet.

View larger version (22K): [in this window] [in a new window]   Fig. 5. XMg vs Ca (a.p.f.u.) plot showing the composition of garnet within the samples from the lower portion of the studied section: 1, core (c) and rim (r) of large crystals within the metapelite–granitic leucosome pair; 2, large crystals within the metapelite–leucotonalitic leucosome pair; 3, small crystals within the metapelite–granitic leucosome pair; 4, small crystals within the metapelite–leucotonalitic leucosome pair; 5, garnet of metagreywacke–leucotonalitic leucosome pair.

 

Plagioclase composition clusters around An₄₀ in the melanosomic and mesosomic metapelites, whereas in the associated leucosomes, compositional clusters around An_30–40 and around An_15–25 are present. In the restitic metagreywackes plagioclase An_45–50 is present whereas within the associated leucosomes a more sodic plagioclase (An_12-30) also occurs. The more sodic plagioclase is interpreted as a magmatic phase whereas the others are considered to be relict.

	WHOLE-ROCK CHEMISTRY

TOP ABSTRACT INTRODUCTION GEOLOGICAL BACKGROUND ANALYTICAL METHODS PETROGRAPHY WHOLE-ROCK CHEMISTRY DISCUSSION CONCLUSION AND GEOLOGICAL... REFERENCES

 
Leucosomes, mesosomes and melanosomes have been analysed for major and trace elements and REE. Analyses are given in Tables 1–4, where the rock-types sampled within the lower and upper portions of the investigated section have been distinguished.

View this table: [in this window] [in a new window]   Table 1: Selected chemical analyses of leucosomes

 

View this table: [in this window] [in a new window]   Table 2: Chemical analyses of mesosomic and melanosomic metapelites of the lower part of the studied section

 

View this table: [in this window] [in a new window]   Table 3: Chemical analyses of mesosomic and melanosomic metapelites of the upper part of the studied section

 

View this table: [in this window] [in a new window]   Table 4: Chemical analyses of mesosomic and melanosomic metagreywackes

 

Leucosomes
On the basis of their chemical compositions, the concordant and discordant leucosomes studied here (Table 1) are leucotonalitic, granodioritic and granitic (Fig. 6). Among the sampled rocks (24) the leucotonalitic types prevail. Most of the leucosomes have SiO₂ contents around 75 wt %; some quartz-rich samples (>60 vol. %) have very high SiO₂ contents (>80%); some leucosomes have high CaO, FeO and Al₂O₃ reflected in the presence of plagioclase, sillimanite and garnet. Attention has been focused on leucosomes potentially representative of melt compositions that have been affected as little as possible by entrainment of restitic material (Watt & Harley, 1993), and as similar as possible to experimental melts; so the samples having extreme compositions will not be considered from now on. The selected samples are leucosomes having SiO₂ ranging from 69 to 78 wt % and CaO contents <4 wt % ; the chemical analyses of the selected samples are given in Table 1.

View larger version (15K): [in this window] [in a new window]   Fig. 6. Normative composition of the selected leucosomes associated with metagreywackes (), upper metapelites () and lower metapelites (•).

 

In the leucotonalitic leucosomes CaO content varies from 2·35 to 3·77 wt % and it is higher in those associated with mesosomic and melanosomic metagreywackes (Table 1: samples KIS11A and MFS11); Na₂O and K₂O contents average 3·20 wt % and 0·77 wt %, respectively (Fig. 7), the A/CNK [Al₂O₃/(CaO + Na₂O + K₂O) molar ratio] ranges from 1·1 to 1·74 and the Mg-number [100MgO/(MgO + FeO^*) molar proportions] from 17·9 to 44·8 (Fig. 8; Table 1); the Mg-number of sample KIS16, containing negligible quantities of MgO and FeO as a result of the absence of entrained mafic phases, is an exception. Mg-number is higher for the leucotonalitic leucosomes associated with mesocratic and melanocratic metagreywackes and upper metapelites (Table 1). The wide variation of A/CNK and CaO (Table 1) reflects variable proportions of entrained sillimanite and plagioclase. From Table 1 it appears that the leucosomes are generally poor in Cr, Ni, V, Zr and P₂O₅ and have very low Rb/Sr ratio (0·016–0·185).

View larger version (26K): [in this window] [in a new window]   Fig. 7. Na2O–MgO, K2O–Na2O and MgO–SiO2 plots for the host rocks and leucosomes: lower (1) and upper (2) metapelites; metagreywackes (3); the samples labelled M represent the respective melanosomes. The numbers 4, 5 and 6 indicate leucotonalitic, granodioritic and granitic leucosomes, respectively, and the tie-lines connect leucosome–host rocks. Also, the compositions of Qtz-poor greywackes (7) and PAAS (8, Post Archean Australian shales) taken from Taylor & McLennan (1985) have been plotted. The compositions of the Bt, Grt and Sil are indicated by bars or arrows. Bt and Grt compositions are from the present study; Sil composition is from Deer et al. (1967).

 

View larger version (21K): [in this window] [in a new window]   Fig. 8. A/CNK vs Mg-number for the studied host rocks; symbols as in Fig. 7.

 

In Fig. 9 the REE chondrite-normalized patterns of leucotonalitic leucosomes are given; they show different concentrations but similarly shaped patterns for light REE (LREE), which are moderately fractionated (La_N/Sm_N = 3·94–5·67), and differently shaped heavy REE (HREE) patterns. The specimens LCF8 and S87 from the lower and upper mesosomic metapelites show the highest LREE concentrations and flat HREE from Dy to Lu (Fig. 9); these REE patterns differ in the kind of Eu anomaly, which is negative in sample S87. Sample KIS18, associated with an upper metapelite, shows marked fractionating HREE (Gd_N/Yb_N = 17·06), potentially reflecting melting in the presence of garnet and the lack of garnet entrained into the leucosome. Samples KIS11A and MFS11, associated with mesosomic metagreywackes, have lower concentrations of LREE.

View larger version (50K): [in this window] [in a new window]   Fig. 9. Chondrite-normalized REE patterns (McDonough & Sun, 1995) of leucosomes and host rocks. Each metasedimentary group includes one melanosome. The numbers 1, 2 and 3 indicate leucosomes associated with lower metapelites, upper metapelites and metagreywackes, respectively.

 

The granodioritic leucosomes are associated with mesosomic metagreywackes (LFC12 and MFS12) and metapelites (LFC7). CaO abundances range from 2·30 to 3·37% (Table 1; Fig. 10); K₂O and Na₂O average 2·31% and 2·84%, respectively (Fig. 7); A/CNK ratio varies from 1·17 to 1·37 (Table 1); their Mg-numbers range from 27 to 50 and are higher for leucosomes associated with metagreywackes (Table 1). They have low abundances of Cr, Ni, V, Zr and P₂O₅ (Table 1) and very low Rb/Sr ratio (<0·2). Granodioritic leucosomes, like the leucotonalitic ones, suggest the presence of abundant relict plagioclase in the light of their high CaO contents (up to 3·37%). The chondrite-normalized patterns show different REE concentrations and different Eu/Eu^* ratios, otherwise they are similarly shaped (Fig. 9). The specimens MFS12 and LFC12, associated with mesosomic metagreywackes, have the lowest and the highest REE concentrations as well as the highest and the lowest Eu/Eu^* ratios, respectively. The slightly negative Eu anomaly (Eu/Eu^* = 0·87) of sample LFC12 may be due to enrichment of both LREE and HREE caused by entrained phases hosting accessories (e.g. Bea & Montero, 1999). Sample LFC7, associated with a lower mesosomic metapelite, has the highest La_N/Sm_N ratio (8·58).

View larger version (24K): [in this window] [in a new window]   Fig. 10. Al2O3, K2O and CaO vs SiO2, EuN vs K2O and Rb/Sr vs Y for the selected leucosomes. Leucotonalitic leucosomes associated with lower mesosomic metapelites (1), upper mesosomic metapelites (2) and with mesosomic metagreywackes (3); granodioritic leucosomes associated with lower mesosomic metapelites (4) and with mesosomic metagreywackes (5); granitic leucosomes associated with lower mesosomic metapelites (6), upper mesosomic metapelites (7) and with mesosomic metagreywackes (8).

 

The granitic leucosomes MFS13, LFC5 and MFS10 are associated with rock-types derived from metapelites and sample LFC6 is associated with a mesosomic metagreywacke (LFC6A). They have high contents of K₂O (>4·8 wt %; Figs 7 and 10) and low contents of CaO (<1·31 wt %; Fig. 10), which is highest in sample LFC6 (1·31 wt %; Fig. 10, Table 1); A/CNK ratio ranges from 1·11 to 1·61 and the Mg-number from 31 to 39, except for sample LFC6, which is Mg free (Table 1). Ba contents are high whereas P₂O₅ and transition element abundances are low (Table 1). In the granitic leucosomes Rb/Sr ratios vary from 0·23 to 0·60. Samples MFS13 and MFS10, as well as sample LFC5, associated with mesosomic upper and lower migmatic metapelites, respectively, have higher K₂O, FeO^* and MgO contents, and low SiO₂ contents that may reflect the presence of entrained garnet and biotite and higher K-feldspar contents (Table 1, Fig. 10).

The chondrite-normalized REE patterns of granitic leucosomes (Fig. 9) show different concentrations but similarly shaped LREE (La_N/Sm_N = 4·03–5·18) as well as variable abundances and different degrees of differentiation of HREE (Gd_N/Yb_N = 0·71–8·26). Samples MFS10 and MFS13 have higher HREE when compared with sample LFC5, which is associated with a lower migmatic metapelite, and with sample LFC6, which is associated with a mesosomic metagreywacke. Sample LFC5 has the highest Gd_N/Yb_N ratio (8·26) and the highest LREE contents, features that are considered to reflect mica breakdown allowing the hosted accessory phases to be available to the melt. As a result of lower garnet contents, sample LFC6 has the lowest REE contents and the highest Eu/Eu^* ratio (Table 1; Fig. 9). All the granitic leucosomes have positive Eu anomaly (Eu/Eu^* = 1·38–3·41) reflecting LREE and HREE depletion or concentration of feldspars.

Mesosomes and melanosomes
The mesosomes and melanosomes have variable compositions owing to the variability of the proportions of their components, the composition of the protoliths and, potentially, the extents of the melt loss or gain. The chemical compositions are given in Table 2.

Most of the mesosomes derived from metapelites are richer in MgO, FeO^*, TiO₂, Al₂O₃, Cr, Ni, V, Zr and Nb, and poorer in SiO₂ (Table 2; Fig. 7) than common shales (Taylor & McLennam, 1985). The contents of K₂O, Na₂O, CaO, Rb, Sr and Ba are variable (Table 2; Fig. 7) owing to the variable biotite and plagioclase contents; K₂O/Na₂O may be as high as nine and CaO/MgO as low as 0·1. Within this variability, on average, the mesosomic metapelites from the upper part of the studied section are poorer in SiO₂ and richer in K₂O (2·96 wt % vs 2·06 wt %), Na₂O (1·46 wt % vs 0·41 wt %), CaO (1·61 wt % vs 0·40 wt %), Ba (954 ppm vs 560 ppm), Rb (91 ppm vs 67 ppm) and Sr (212 ppm vs 88 ppm) than the mesosomic metapelites from the lower part; Fig. 8 shows that A/CNK (2·73 vs 7·65) and the Mg-number (44 vs 32) are different, too. These differences reflect different concentration and nature of the residual and peritectic phases as well as different concentration of leucosomes and of retrograde phases; they may also imply differences among the protoliths (see also Caggianelli et al., 1991). This is supported by the different Mg-numbers and Al₂O₃/TiO₂ ratios involving strongly compatible elements during partial melting (Table 5). On the whole it seems that the lower metapelitic mesosomes derived from pelites similar to PAAS (Post-Archean Australian Shale, Taylor & McLennan, 1985) whereas the upper mesosomes include more mafic types (Tables 2 and 3, and Fig. 7). The mesosomic metapelites are richer in both LREE and HREE than common shales (Taylor & McLennan, 1985). Chondrite-normalized REE patterns (Fig. 9) show differentiated LREE and flat HREE with general negative Eu anomalies; the overall shape and steepness (La_N/Sm_N = 2·99–4·61; Gd_N/Yb_N = 0·87–2·15) of these patterns are similar to those of the PAAS (Taylor & McLennan, 1985).

View this table: [in this window] [in a new window]   Table 5: Some chemical features of migmatites and their components of the lower-crustal section of the Serre

 

Most of the migmatitic metapelites have depleted chemistries (Fig. 7; see also Schenk, 1990; Caggianelli et al., 1991) relative to the common shales (Taylor & McLennan, 1985), a feature that is suggestive of melt loss.

The melanosomic metapelites are strongly enriched in Al₂O₃, FeO^*, MgO and TiO₂ and generally depleted in SiO₂ when compared with mesosomes (Tables 2 and 3; Fig. 7); K₂O may be either very low (1 wt % within the lower portion) or abundant (>5 wt % within the upper portion), depending on biotite consumption or concentration; Na₂O is low (0·14–0·81 wt %), as is CaO (0·27 wt % in sample KIS21), owing to plagioclase consumption. Generally, these melanosomes have low P₂O₅ contents (<0·09 wt %). The garnet–sillimanite-rich melanosomic metapelite LFC1, derived from the lower stratigraphic part of the section, is enriched in Zr, Nb, V and Cr and depleted in Ba, Rb, Sr and LREE (Tables 2–5) when compared with the mesosomes sampled in the same part of the section; the very low abundances of Rb (25 ppm) and Ba (241 ppm) in sample LFC1 reflect the very low abundances of host phases for these elements, such as biotite and alkali feldspar. The very low abundance of plagioclase explains the depletion of Eu (0·48 ppm) and Sr (3 ppm). Its chondrite-normalized pattern (Fig. 9) shows weakly fractionated LREE (La_N/Sm_N = 1·70), flat HREE (Gd_N/Yb_N = 1·17) and a very strong negative Eu anomaly (Eu/Eu^* = 0·15). The biotite-rich melanosomic metapelite KIS15A, sampled in the upper part of the section, is enriched in Zr, Nb, V, Cr, Ni, Rb and Ba relative to the mesosomes sampled within the same area (Table 3). It shows more fractionated LREE (Fig. 9) when compared with the melanosome LFC1 (La_N/Sm_N = 4·14; Gd_N/Yb_N = 0·91) and a pattern similar to those of the mesosomes sampled in the upper section, with a strong Eu anomaly (Eu/Eu^* = 0·53).

The mesosomic and melanosomic metagreywackes on the whole reflect derivation from Qtz-poor greywackes (Taylor & McLennan, 1985) owing to their low SiO₂ and very high TiO₂, MgO and CaO contents (Table 4 and Fig. 7), except for samples MFS11A, MFS12A and LFC6A, which document leucosome gain.

The mesosomic metagreywackes have higher TiO₂, Na₂O, CaO and Sr contents and lower Al₂O₃ and FeO^* contents than those mesosomes derived from pelitic precursors (Table 2 and Fig. 7). The A/CNK and Mg-number average 1·19 and 49, respectively (Table 5). From Fig. 9 it appears that a subset of samples is richer in REE and shows chondrite-normalized patterns similar to those of the migmatitic metapelites and PAAS.

The melanosomic metagreywacke (LFC10A) is enriched in MgO, FeO, K₂O, V, Zr, Nb and P and depleted in Na₂O relative to mesosomic metagreywackes (Table 4; Fig. 7); K₂O/Na₂O is >2 and CaO/MgO 1 (Table 5). Its chondrite-normalized pattern (Fig. 9) is similar to those of mesosomic metagreywackes poor in REE; its Eu/Eu^* ratio is 0·85 (Table 2).

Variability in A/CNK and Mg-number is evident in Fig. 8 for the different rock-types but it appears that the rock-types derived from metagreywackes and those derived from metapelites sampled in the upper portions of the studied section have, on the whole, lower A/CNK values and higher Mg-numbers.

With respect to the Sr and Nd isotope systematics of the Serre migmatites (Del Moro et al., 2000), it appears that within both leucosomes and their host rocks there is a wide variability of (⁸⁷Sr/⁸⁶Sr) and of Nd values calculated at 290 Ma (Tables 1–4); the values of (⁸⁷Sr/⁸⁶Sr)_{290 Ma} and of Nd_{290 Ma} in leucosomes are generally distinct from those of their host rocks (Tables 1–4); this is consistent with the inference that the leucosomes are in general not sourced from their immediate host material. As a consequence, in Sr and Nd evolution diagrams isochrons cannot be defined either for leucosomes or for melanosomes.

	DISCUSSION

TOP ABSTRACT INTRODUCTION GEOLOGICAL BACKGROUND ANALYTICAL METHODS PETROGRAPHY WHOLE-ROCK CHEMISTRY DISCUSSION CONCLUSION AND GEOLOGICAL... REFERENCES

 
The lower-crustal portion of the Serre here investigated (Fig. 1) experienced at the metamorphic peak and during decompression T between about 700 and 800°C and P between 4 and 8 kbar (Schenk, 1984, 1990). The metasediments underwent partial melting, which resulted in their chemical variability and the production of sodic and potassic melts. The composition of mesosomes and melanosomes reflects their refractory and residual character, indicating that a large fraction of potassic melts is ‘missing’ (e.g. Maccarrone et al., 1983; Schenk, 1990; Caggianelli et al., 1991). The behaviour of K₂O, which may increase or decrease dramatically, and the tendency to consume or concentrate biotite within the melanosomes indicate that varied melting processes did operate. The variability of compositions of the leucosomes in migmatitic terranes may be caused by fractional crystallization and/or melt–residuum unmixing (e.g. Watt & Harley, 1993; Brown, 1994; Sawyer, 1994). In Table 5 some parameters that are useful to decipher the melting process experienced by the precursor metasediments of the Serre are summarized. To provide general information, the mean compositions of the migmatites and of their components are given. We should expect: (1) residues of H₂O-fluxed melting, with large involvement of plagioclase (Patiño Douce & Harris, 1998), characterized by decrease of Na₂O, CaO, CaO/MgO, Al₂O₃/TiO₂ and Eu/Eu^* and increase of K₂O, K₂O/Na₂O and Rb/Sr; (2) residues of the mica-dehydration melting characterized by decrease of K₂O, K₂O/Na₂O, Rb/Sr and Al₂O₃/TiO₂ and increase of CaO (Table 5; Patiño Douce & Harris, 1998; Otamendi & Patiño Douce, 2001). REE have no definite behaviour as their carriers (accessories) may form either intergranular crystals or inclusions within the porphyroblasts involved (or not) in melting reactions and may be utilized in the reactions forming major minerals (Bea & Montero, 1999). The melanosomes LFC10A, in which Bt concentrates, and LFC1, in which it is nearly absent, have very different P₂O₅ contents and both show distinctly lower LREE than the leucosomes (Table 5). When the two processes take place in a given rock, the elemental patterns vary depending on the degree of melting and the efficiency of melt–residuum separation in each of them, as well as on the relative proportion of the phases involved in total melting. In the following these problems will be elucidated.

Fractional crystallization vs residuum–melt unmixing
Leucotonalitic to granitic leucosomes are at present interspersed along the studied portion of the lower-crustal section of the Serre. Among the sampled rocks the leucotonalitic types prevail. Petrographic observations and mineral chemistry demonstrate that they are mixtures of melt and entrained phases, as generally is considered to occur in anatectic systems (e.g. Solar & Brown, 2001). Figure 10 shows the relationships between CaO, K₂O, Al₂O₃ and SiO₂, between Eu_N and K₂O, and between Y and Rb/Sr in the selected samples. There do not appear to be any distinct correlations between CaO, K₂O, Al₂O₃ and SiO₂ that would be indicative of fractional crystallization of the melts to produce residual melts richer in SiO₂ and K₂O. An exception to this, possibly, is the granitic leucosome LFC6 associated with mesosomic metagreywacke LFC6A which, in addition, is very poor in FeO and totally devoid of MgO (Table 1), owing to the absence of mafic phases. The Eu and Y contents of the leucosomes point to the same conclusion; fractionated melts should be depleted in Eu and Y in comparison with less fractionated ones, owing to the separation of plagioclase, whereas, in the studied case, the granitic leucosomes have Eu and Y contents similar to or higher than those of the leucotonalitic leucosomes, again with the exception of sample LFC6 (Table 1). It is evident, however, that the variability both among the leucosomes and within each of them is influenced by the kind and proportion of entrained phases as discussed by Watt & Harley (1993): (1) the selected granitic leucosomes associated with mesosomic metapelites have low SiO₂ contents (Fig. 10), high FeO, TiO₂ and MgO values (Table 1) as well as high Ba, Rb and Y contents reflecting the influence of entrained garnet and biotite on their compositions; (2) the leucotonalitic leucosomes show high and variable CaO (2·35–3·7 wt %), FeO (0·51–2·07 wt %) and MgO (0·01–0·79 wt %) contents and variable K₂O (0·13–1·57 wt %), reflecting various proportions of entrained plagioclase, garnet and biotite. Thus the potassic and sodic leucosomes do not show geochemical signatures indicative of a link owing to crystal fractionation, so they were probably produced by independent melts formed through distinct melting reactions leading to different melt and residue compositions. An exception may be the granitic leucosome LFC6 associated with the metagreywackes, which may have evolved through magmatic fractionation.

Origin of the leucosomes interspersed within the migmatites
The leucotonalitic to granitic leucosomes show some textures suggestive of melt interaction producing intermediate granodiorite compositions. Thus they may be coeval or formed during steps of a continuous process. The origin of the leucotonalitic leucosomes will be discussed first.

In Fig. 8 it appears that: (1) the rock-types derived from the melting of metapelites sampled along the lower part of the investigated section and those derived from metagreywackes define distinct fields; (2) the mesosomic and melanosomic metapelites from the upper part of the section are scattered in composition. Melanosomic and mesosomic metagreywackes as well as some melanosomic and mesosomic metapelites sampled in the upper part lie between the field of the leucotonalitic leucosomes and the field defined by the projected compositions of Grt and Bt, consistent with residuum–melt unmixing, the melanosome being characterized both by the biotite that survived the melting process and by the biotite that formed during the retrograde rehydration. Mesosomic and melanosomic metapelites from the lower part of the section, together with some samples of the upper part, mostly lie between the potassic leucosomes and the projected compositions of Grt and Sil, reflecting loss of biotite (Fig. 7). Hence, a complex melting history must be envisaged for the lower-crustal metasediments under study. In detail, the melanosomic metagreywacke LFC10A shows strong depletion in Na₂O and increase in K₂O, K₂O/Na₂O and Rb/Sr (Table 5) consistent with what emerges from Fig. 7; that is, with the consumption of abundant plagioclase during melting. Such a process, however, should determine decrease of CaO, which, on the contrary, increases in sample LFC10A; this is consistent with the more calcic plagioclase (An_45–50) and Ca presence in the garnet (Fig. 5). The most depleted metapelites (LFC1 and KIS15A) show CaO low contents (Table 5) that are not explicable in terms of dehydration melting alone, which should produce concentration of CaO within the residues (e.g. Patiño Douce & Harris, 1998; Otamendi & Patiño Douce, 2001). Many depleted rocks have very high K₂O/Na₂O and very low CaO/MgO ratios (Tables 2 and 3) reaching values three times higher and a fifth lower, respectively, than those of typical fine-grained sediments (Taylor & McLennan, 1985). Overall, these geochemical signatures indicate that both the metagreywackes and the metapelites of the Serre underwent a melting process that consumed abundant plagioclase.

Experimental results (e.g. Conrad et al., 1988; Patiño Douce & Harris, 1998) have shown that hydrous melting of metapelites and metagreywackes, in a T range of 700–800°C, produces sodic melts. The temperature of the beginning of melting in the presence of free water is independent of the percentage of water, but the amount of melt produced depends on the amount of added H₂O (Castro et al., 2000). Experiments on metapelites by Patiño Douce & Harris (1998) were performed at 6 and 10 kbar, giving liquids with Na₂O/K₂O ratios of 1·7 and 3–4, respectively. During prograde metamorphism and subsequent isothermal decompression the investigated portion of the lower-crustal section of the Serre experienced P = 8–4 kbar and T = 700–800°C (e.g. Schenk, 1990), conditions that partly overlap the physical conditions explored by Patiño Douce & Harris (1998). The sodic leucosomes have variable Na₂O/K₂O ratios, as a result of the variable proportions of entrained plagioclase. However, in samples KIS16, LFC8 and S87 Na₂O/K₂O is in the range of 2–4, which is to be considered somewhat higher than in pure melts, as the relict or entrained plagioclase is still present. Thus in the case under study, the presence of free H₂O [H₂O-fluxed melting of Patiño Douce & Harris (1998)] may have promoted the production of leucotonalitic melts and caused stronger involvement of plagioclase with respect to micas in the melting process. In the light of the geochemical features noted above, a reaction such as

may have been responsible for the formation of the sodic leucosomes. The very low Rb/Sr ratio (Table 5) in the leucotonalites, which has been used as a discriminator of water-fluxed melting vs dehydration melting (e.g. Harris & Inger, 1992, 1993), supports the participation of H₂O in their production, although care must be exercised as some degree of disequilibrium may be involved in the melting process, which may reduce the Rb/Sr ratio of the melt regardless of the amount of H₂O involved during melting (Harris et al., 1993).

The generally low abundances of REE in leucosomes are usually attributed to the failure of accessory phases to equilibrate with melts, because (1) melt removal is rapid or (2) accessories are shielded during melting and segregation or (3) accessories have limited dissolution, as occurs during the production of H₂O-undersaturated melts (e.g. Barbey et al., 1990; Watt & Harley, 1993; Bea et al., 1994; Whitney & Irving, 1994; Carrington & Watt, 1995). In the case under consideration, the relatively low REE contents (Table 5) of the leucotonalitic leucosomes must be attributed to rapid melt removal in response to fluid ingress combined with shielding of accessories included within garnet and biotite, which remain stable during H₂O-present melting; the HREE could also have been trapped in the residual garnet. The tendency of LREE and HREE to decrease in the leucosomes (Figs 9–11, Table 5) accounts for their general positive Eu anomalies. In detail, however, the proportion of plagioclase in the melts and the entrainment of the relict plagioclase will also have control over the kind of Eu anomaly, because either some leucosomes (Fig. 11) or their average (Table 5) show Eu contents higher than in host rocks. Thus the leucotonalitic leucosomes at present distributed in the lower-crustal segment represent disequilibrium melts especially in terms of their REE contents (residue did not equilibrate with liquid before melt removal). This is also supported by other chemical characteristics: low P₂O₅ and low and variable Zr, Nb and Y contents in leucosomes and different (⁸⁷Sr/⁸⁶Sr)_{290 Ma} and Nd values in leucosome–host-rock pairs (Tables 1–4; Del Moro et al., 2000).

View larger version (26K): [in this window] [in a new window]   Fig. 11. Chondrite-normalized REE patterns (McDonough & Sun, 1995) of leucosome–host-rock pairs.

 

The presence within the migmatites also of minor potassic leucosomes, however, implies that potassic phases were involved in the melting process. As discussed above, potassic and sodic melts mixed, producing locally granodioritic leucosomes; hence the formation of these distinct melts may be framed within a continuous melting process. In fact, H₂O-present melting [reaction (1)] at rising temperature and decreasing H₂O activity leads to more muscovite reacting out to draw the melts toward granitic composition (Patiño Douce & Harris, 1998).

Origin of melanosomes—inferences
Let us consider now the most depleted melanosomic metapelites and metagreywackes. These include: (1) Grt–Sil-rich melanosome LFC1 with little biotite (partially retrograde) and plagioclase, which is strongly depleted in SiO₂, K₂O and Na₂O (Fig. 7, Table 5) and has very high A/CNK and low Mg-number (Fig. 8, Table 5); (2) the melanosome KIS15A with abundant biotite but scarce felsic minerals, which is depleted in SiO₂, Na₂O and CaO, enriched in K₂O (Fig. 7, Table 5), and characterized by lower A/CNK and higher Mg-number than LFC1 (Table 5); (3) melanosomic metagreywacke LFC10A, which contains biotite, garnet and scarce plagioclase and quartz, is depleted in SiO₂ and Na₂O and enriched in K₂O (Fig. 7, Table 5), and is characterized by the lowest A/CNK (Table 5). Samples LFC1 and LFC10A come from the base and sample KIS15A from the top of the studied lower-crustal portion (Fig. 1). Within both the lower and the upper sections there are residual metapelites that exhibit (Tables 2 and 3) depletions in CaO and Na₂O that are even stronger than the type examples LFC1 (e.g. sample LFC7B) and KIS15A (e.g. samples KIS20 and KIS21).

We propose that: (1) the metagreywackes experienced H₂O-present melting producing sodic melts and, as a result of their removal, Bt–Grt-rich and Pl-poor melanosomes; (2) the metapelites underwent H₂O-present melting (as discussed above) and a variable degree of mica-dehydration melting depending on their Mg-number and on physical conditions, producing sodic and potassic melts and, after their removal, Bt–Grt–Sil-rich melanosomes poor in plagioclase as well as Bt-free types rich in Grt and Sil and poor in plagioclase.

To demonstrate the above points we consider first the Grt–Sil-rich metapelitic melanosome LFC1 and the melanosomic metagreywacke LFC10A, occurring within the lowermost metapelites and within the felsic granulites, respectively (Fig. 1). They experienced the highest and broadly similar metamorphic conditions as well as extraction of sodic melts but probably were derived from precursors having different Mg-numbers. Stevens et al. (1997) have demonstrated, starting from metapelites and metagreywackes having Mg-number 49, that: (1) at P <8 kbar, biotite dehydration melting in metapelites occurs at T (780–740°C) lower than in metagreywackes (>800°C) under fluid-absent conditions, and (2) the Mg-number exerts a moderate control on melting and a strong control on the maximum thermal stability of biotite, i.e. the lower the ratio the lower its thermal stability. This implies that the metapelites having lower Mg-number and producing such Grt–Sil-rich rocks as LFC1 underwent biotite dehydration melting, whereas the metagreywackes did not experience biotite dehydration melting. This is also seen from inspection of Fig. 7, where the migmatitic and melanosomic metapelites of the lower part of the section show a tendency towards the loss of biotite, whereas the mesosomic and melanosomic metagreywackes show maintenance of biotite. Furthermore, the dramatic depletion in both K₂O and Na₂O within sample LFC1 (Table 5) and the enrichment in K₂O and depletion in Na₂O within sample LFC10A (Table 5) are consistent with the removal of both sodic and potassic melts from metapelites, whereas only sodic melts are removed from the melanosomic metagreywacke.

We now consider the melanosomic metapelite KIS15A sampled south of the Grt–Crd univariant zone (Fig. 1). According to Schenk’s estimates (1984, 1990), within this area metapelites experienced P 5·5–6 kbar and T 700°C during the prograde metamorphism. Sample KIS15A is rich in Bt and Sil, as confirmed by the high K₂O and Al₂O₃ contents (Table 3 and Fig. 7). To a first approximation this sample, with relatively low Na₂O and CaO, very high K₂O and K₂O/Na₂O, and very low CaO/MgO ratios (Table 5), seems to monitor only H₂O-present melting. However, mass balance computations (see below) indicate extraction of abundant granitic melts from both mesosomes and melanosomes and imply that the precursor metapelites underwent also dehydration melting. The potassic melts must in this case, and given the P–T conditions, have been mostly formed as a result of Ms-dehydration melting, probably following the reaction (Patiño Douce & Harris, 1998)

Textural evidence such as myrmekitic intergrowths of biotite and quartz within the melanosomes suggests also contribution by Bt-dehydration melting. This is also seen from Fig. 12, where the P–T path at the metamorphic peak and during decompression enters muscovite and biotite dehydration melting conditions. The contribution of biotite in the dehydration melting, however, should not be abundant owing to the generally higher Mg-number of the metapelites (Fig. 8 and Table 5) and the high concentration of biotite (>40 vol. %) within the melanosomes even if in part Bt has been produced by melt–wall-rock back-reaction. Hence, even the upper metapelites are deduced to have undergone hydrous melting and mica-dehydration melting.

View larger version (36K): [in this window] [in a new window]   Fig. 12. Partial P–T grid in the NaKFMASH system for metapelites. Melting and dehydration reactions are from Spear et al. (1989). Al2SiO5 triple point after Pattison (1989), in Spear et al. (1999). Arrows show inferred P–T paths of the upper (a) and lower (b) part of the section of the Serre after Schenk (1984). The retrograde path has not been reported completely.

 

The above scenario accounts for the coexistence, side by side, of rock-types either rich in or lacking in biotite, and of rock-types more or less depleted in felsic components depending on the nature and chemistry of their precursors and on fluid availability.

Interestingly, two melting stages or events have been proposed for the metasediments of the Ivrea–Verbano zone in the Western Alps (Vavra et al., 1996), with which the Calabrian lower-crustal rocks have been compared (Moresi et al., 1979; Schenk, 1981).

Estimation of melt fraction
Estimations of melting degree and of the extents of melt extraction from the most depleted rocks have been performed following different strategies. Melting processes have been modelled in two stages. In the first stage we have modelled the partial melting of metagreywackes undergoing H₂O-present melting; in the second stage we have modelled the partial melting of the metapelites recording both H₂O-present and dehydration melting.

The melanosomic metagreywacke LFC10A and the melanosomic metapelite LFC1 have been taken into account to model the melting at the base of the lower-crustal section; the melanosomic metapelite KIS15A has been considered to estimate the melting of metapelites at the top of the investigated portion (Fig. 1). Owing to (1) disequilibrium melting in the production of the leucosomes, (2) the multi-stage melting affecting metapelites, (3) the presence of entrained phases in the leucosomes as well as the presence of peritectic phases and of biotite formed by melt–wall-rock back-reaction in the host rock, and (4) the problematic identification of the protoliths and of the extracted melts, a crude approach to estimate the melt fraction has been adopted by modelling distribution of major elements and some trace elements as an equilibrium process and making some assumptions for the starting materials and for the melt compositions. Obviously the results are to be considered as indicative rather than fully quantitative. The maximum permissible error Thj at a 95% confidence level (Bea, 1989) has been calculated for the various elements to test the statistical significance of the resulting estimates (Tables 6–8).

View this table: [in this window] [in a new window]   Table 6: Results of mass balance computations (major elements) to estimate the degree of melting followed by melt extraction for the melanosomic metagreywackes

 

View this table: [in this window] [in a new window]   Table 7a: Results of mass balance computations (major elements) to estimate the degree of melting followed by melt extraction (melt composition calculated combining 87% Petrizzi granite and 13% leucotonalitic leucosome KIS16; see Table 6) for the lower metapelites

 

View this table: [in this window] [in a new window]   Table 7b: Results of mass balance computations (major elements) to obtain the volume of melts extracted from migmatitic metapelites of the lower portion of the studied section

 

View this table: [in this window] [in a new window]   Table 8a: Calculated source for the upper metapelites considering the melanosome KIS15A as the depleted rock after the extraction of total melt calculated as combination of 70% Petrizzi granite and 30% leucotonalitic leucosome KIS16

 

View this table: [in this window] [in a new window]   Table 8b: Results of mass balance computations (major elements) to estimate the degree of melting followed by melt extraction producing the Petrizzi granite

 

To model the hydrous melting: (1) mesosomic metagreywacke KIS11 has been considered as the source of the melanosomic metagreywacke LFC10A, as it matches the composition of Qtz-poor greywackes (Taylor & McLennan, 1985); (2) the leucotonalitic leucosome KIS16 has been chosen as representative of the composition of the extracted melt because of the Na₂O/K₂O ratio (2·35), which is similar to those of the melts obtained by Patiño Douce & Harris (1998) at P <10 kbar and because of the absence of entrained mafic phases as indicated by low total FeO + MgO.

The results of mass balance computations based on these assumptions are given in Table 6a. To control these results, the composition of a plausible restite (solid residue after melt removal) has been modelled using the composition of minerals present in a melanosomic metagreywacke (Table 6b). It is apparent from Table 6 that 9–13 vol. % of leucotonalitic melt would have to be removed from a composition like that of KIS11 to produce a residual metagreywacke like LFC10A or somewhat similar to that. On the whole, the results may be considered with some confidence especially when they are compared with the field estimate of 10 vol. % of leucosomes in the lowermost portion of the migmatitic metapelites (Caggianelli et al., 1991).

The melt fraction of 13 vol. % is likely to approximate the maximum percentage of the leucotonalitic melt that was also removed from the melanosomic metapelite LFC1, as plagioclase is generally more abundant in metagreywackes. To model the total melting (H₂O-present plus dehydration melting) of this sample, the compositions of PAAS as starting material and of the total removed melt [combining the maximum proportion 13 vol. % of leucotonalitic melt KIS16 with a complementary fraction 87 vol. % of Petrizzi peraluminous granite, of Caggianelli (1988)] were assumed (Table 7a). A strategy similar to that adopted for the metagreywackes, considering the chemical compositions of the restitic phases to model a possible restite, has been followed as alternative approach (Table 7b). These calculations lead to an estimate of 57–66 vol. % for the total melt extracted, leaving behind residual rocks similar to those occurring in the lower part of the section. Subtraction of the highest quantity of melts extracted from the metagreywacke (13 vol. %) from the total melt extracted from the melanosomic metapelite LFC1 suggests that 44–53 vol. % of melts formed through mica-dehydration melting was removed from the biotite-free or -poor pelitic melanosomes at the base of the lower-crustal section of the Serre.

With respect to the upper migmatites, Caggianelli et al. (1991) estimated up to 40 vol. % of leucosomes in the field. The biotite-rich melanosomic metapelite KIS15A has been considered to represent the most depleted rock after the extraction of total melt (the present leucosomes plus ‘missing’ granites). A melt composition obtained by combining 70% of peraluminous granite occurring within the foliated tonalites (Caggianelli, 1988) and 30% of leucotonalitic melt (KIS 16) was added in stepwise fashion to KIS15A until the composition of a plausible metapelitic source was obtained (Table 8). This procedure indicated that probably 58 vol. % of total melt has been removed from KIS15A; that is, 28 vol. % peraluminous granitic melt, most of which migrated far from the source, and 30 vol. % of leucotonalitic melt, which remained within the metasediments.

The following strategy has been adopted to estimate the fraction of melt extracted from the migmatitic metapelites: the mean values of upper (Table 8b) and lower migmatites (Table 7b) have been used as residues; the composition of PAAS and the composition obtained in Table 8 have been used as starting materials of the lower and upper metapelites, respectively. As an alternative approach, the various migmatites have been considered as potential residues and those giving the minimum values of the sum of squares of residuals have been considered as possible residues; they are samples LFC5B for the lower migmatites (Table 7b) and samples KIS18A and MFS10A for the upper ones (Table 8b). It is apparent that 36–44 vol. % and 27–40 vol. % of peraluminous granitic melt have been extracted from lower and upper migmatites, respectively.

Computations using Rb, Ba, Sr and Y hosted in the phases involved in melting have also been made assuming equilibrium melting (Allègre & Minster 1978) and C_restite/C_melt as distribution coefficients. The results are unrealistic probably as a result of the retrograde rehydration, which could have modified the budget of these elements. The volumes of the missing peraluminous granitic melt here approximated are broadly in agreement with Schenk’s results (1990), which indicate extraction of 40 vol. % of peraluminous granitic melt from the metapelitic unit as a whole. The large volume fraction of the melts formed by dehydration melting allowed large-scale mobility. The occurrences of dikes and masses of S-type peraluminous granites in the upper portion of the lower-crustal section as well as in the foliated tonalites have a bearing in this interpretation. The rapid melt removal in response to fluid ingress, instead, did not allow the collection of the sodic melts in such proportions as to permit large-scale mobility (e.g. Sawyer, 1994; Otamendi & Patiño Douce, 2001). The granitic leucosomes interspersed within the metapelitic migmatites could represent (1) a fraction of peraluminous granitic melts, which formed by dehydration melting and were left behind within the migmatites, or (2) melts formed in presence of free H₂O at a temperature higher than when the sodic melts formed, as discussed above. The first possibility does not appear viable, as the potassic leucosomes have positive Eu anomalies whereas the peraluminous granite associated with the tonalites has a negative Eu anomaly (Eu/Eu^* = 0·66, Caggianelli, 1988). The granitic leucosomes associated with mesosomic metapelites have very high K contents (up to 8 wt %) and the entrained biotite is insufficient to account for this feature owing to its low modal content (<5 vol. %). An explanation for K enrichment and, accordingly, for the positive Eu anomaly (Figs 9–11) is the production of peritectic Kfs formed by melting in systems with high H₂O/K₂O ratio (e.g. Carrington & Watt, 1995) and the subsequent incorporation of this phase into the removed melts. This possibility is supported by the absence of Kfs within melanosomes such as LFC1 and by its scarceness within those such as KIS15A, and is consistent with hydrous melting at rising temperature leading to more Ms reacting out to draw the melts toward granitic composition (Patiño Douce & Harris 1998).

Even if the data presented here are to be considered as indicative, owing to the assumption we have made, the missing information about the quantities of Ms and Bt within the precursor metapelites and the retrograde formation of Bt, it should be emphasized that comparable volumes of leucosomes, which formed in the presence of free H₂O, and of missing granites, which formed by dehydration melting, have been estimated elsewhere (e.g. Otamendi & Patiño Douce 2001) within anatectic crustal sectors, at somewhat higher temperatures.

	CONCLUSION AND GEOLOGICAL IMPLICATION

TOP ABSTRACT INTRODUCTION GEOLOGICAL BACKGROUND ANALYTICAL METHODS PETROGRAPHY WHOLE-ROCK CHEMISTRY DISCUSSION CONCLUSION AND GEOLOGICAL... REFERENCES

 
Examination of the leucosomes occurring in the anatectic migmatites and of the melanosomes of the Serre shows that: (1) the precursor metagreywackes underwent hydrous melting producing leucotonalitic melts; (2) the precursor metapelites underwent H₂O-present melting and both Ms- and Bt-dehydration melting producing leucotonalitic and peraluminous granitic melts (Fig. 12). The migmatitic metapelites, the most widespread lithologies in the lower-crustal section of the Serre, have restitic chemistries (see also Schenk, 1990; Caggianelli et al., 1991). They appear as the sources of anatectic peraluminous granites after the extraction of 32 vol. % and of 40 vol. % in the upper and lower part of the section, respectively. The most depleted metapelites (Grt + Sil + Rt ± Qtz ± Bt ± Pl), at the base of the section, are the products of extraction of 60 vol. % of total melts (10 vol. % formed by hydrous melting plus 50 vol. % of missing granitic melt formed by mica-dehydration melting). Hence, there is a link between migmatites and granites (see also Maccarrone et al., 1983; Schenk, 1990; Caggianelli et al., 1991), which migrated far from their sources. In contrast, the melts formed through H₂O-present melting reactions were removed rapidly and in small batches from their sources and, consequently, did not coalesce into large melt pockets permitting mobility over large distances (e.g. Sawyer, 1994; Otamendi & Patiño Douce, 2001).

It is possible to infer estimates of the relative timing of dehydration melting producing the ‘missing’ peraluminous granites within the context of the P–T–t path of the lower-crustal rocks of the Serre. From Fig. 12 it appears that the lowermost metapelites during the prograde metamorphism entered the Ms- and Bt-dehydration melting regime and the consumption of biotite continued during the decompression, being completed within the Mg-poor types (Stevens et al. 1997) through the reactions

At the metamorphic peak the upper metapelites (Fig. 1) were mostly in the field of Ms-dehydration melting (Fig. 12). Schenk (1990) showed that Ms stability within the upper part of the lower-crustal section was overstepped at the metamorphic peak. During the decompression the upper metapelites underwent Bt-dehydration melting limited by their common higher Mg-number (Table 5) and the P–T conditions through reaction (3) and, going downward, at rising temperature, through the univariant reaction (4). Dehydration melting during decompression is suggested also by: (1) the intrusion of the anatectic peraluminous granites into the foliated tonalites, probably in an extensional tectonic regime (Caggianelli et al., 1997; Di Battista, 1999; Del Moro et al., 2000), while the latter were still partially molten (Caggianelli, 1988); (2) the occurrence of Crd + Sil + Spl + Crn residual rocks within the contaminated calc-alkaline granitoids of the central Serre; (3) the intrusion of cordierite microgranodiorite (Fornelli, 1991). Hence, it appears that peraluminous granitic melts formed both earlier than and during the decompression.

The H₂O-present melting stage cannot be unambiguously placed within the evolutionary path of the lower-crustal rocks as no field evidence of the chronological relations with the dehydration melting has been found so far. In the simplest model, however, the fluids generated during prograde metamorphism, i.e. during the amphibolite–granulite transition, might have promoted H₂O-fluxed melting. This is in agreement with the conclusions of Patiño Douce & Harris (1998) that H₂O-fluxed melting producing sodic melts occurs at lower temperatures before dehydration melting and is favoured by higher pressures. Hence, the hydrous melting in the Serre probably occurred during the prograde metamorphism, earlier than dehydration melting, and the lower-crustal metasediments were subjected to two sequential episodes (Fig. 12) of melting under progressively decreasing water activity.

	ACKNOWLEDGEMENTS

 
We are indebted to J. R. Ashworth, L. Barbero, V. Schenk, K. P. Skjerlie and M. Thöni for their constructive criticism on the previous version of the paper. Comments and suggestions from I. Braun, V. Schenk and an anonymous reviewer, together with editorial suggestions from S. Harley, helped us to greatly improve the present version of the paper, and are most appreciated. Financial support was provided by MURST cofin (1997–1998) and Bari University.

	FOOTNOTES

 
^*Corresponding author. Telephone: +39-80-5442661. Fax: +39-80-5442591. E-mail: a.fornelli{at}geomin.uniba.it

	REFERENCES

TOP ABSTRACT INTRODUCTION GEOLOGICAL BACKGROUND ANALYTICAL METHODS PETROGRAPHY WHOLE-ROCK CHEMISTRY DISCUSSION CONCLUSION AND GEOLOGICAL... REFERENCES

 
Allègre, C. J. & Minster, J. F. (1978). Quantitative models of trace element behavior in magmatic process. Earth and Planetary Science Letters 38, 1–25.[Web of Science]

Ashworth, J. R. (1985). Introduction. In: Ashworth, J. R. (ed.) Migmatites. Glasgow: Blackie, pp. 1–35.

Barbey, P., Macaudiere, J. & Nzenti, J. P. (1990). High-pressure dehydration melting of metapelites: evidence from the migmatites of Yaoundé (Cameroon). Journal of Petrology 31, 401–427.[Abstract/Free Full Text]

Bea, F. (1989). A method for modelling mass balance in partial melting and anatectic leucosome segregation. Journal of Metamorphic Geology 7, 619–628.[Web of Science]

Bea, F. & Montero, P. (1999). Behavior of accessory phases and redistribution of Zr, REE, Y Th, and U during metamorphism and partial melting of metapelites in the lower crust: an example from the kinzigite formation of Ivrea–Verbano, NW Italy. Geochimica et Cosmochimica Acta 63(7–8), 1133–1153.[Web of Science]

Bea, F., Pereira, M. D. & Stroh, A. (1994). Mineral/leucosome trace-element partitioning in a peraluminous migmatite (a laser ablation-ICP-MS study). Chemical Geology 117, 291–312.[Web of Science]

Brown, M. (1994). The generation, segregation ascent and emplacement of granite magma: the migmatite-to-crustally derived granite connection in thickened orogens. Earth-Science Reviews 36, 83–130.

Caggianelli, A. (1988). Le tonaliti di Squillace–Petrizzi (Serre Calabre): aspetti petrogenetici. Ph.D. thesis, University of Bari, 91 pp.

Caggianelli, A., Del Moro, A., Paglionico, A., Piccarreta, G., Pinarelli, L. & Rottura, A. (1991). Lower crustal granite genesis connected with chemical fractionation in the continental crust of Calabria (Southern Italy). European Journal of Mineralogy 3, 159–180.[Abstract/Free Full Text]

Caggianelli, A., Prosser, G. & Di Battista, P. (1997). Textural features and fabric analysis of granitoids emplaced at different depths: the example of the Hercynian tonalites and granodiorites from Calabria. Mineralogica et Petrografica Acta XL, 11–26.

Carrington, D. P. & Watt, G. R. (1995). A geochemical and experimental study of the role of K-feldspar during water-undersaturated melting of metapelites. Chemical Geology 122, 59–76.[Web of Science]

Castro, A., Corretgé, L. G., El-Biad, M., El-Hmidi, H., Fernàndez, C. & Patiño Douce, A. E. (2000). Experimental constraints on Hercynian anatexis in the Iberian massif, Spain. Journal of Petrology 41(10), 1471–1488.[Abstract/Free Full Text]

Clemens, A. (1990). The granulite–granite connection. In: Vielzeuf, D. & Vidal, Ph. (eds) Granulites and Crustal Evolution. Dordrecht: Kluwer, pp. 25–36.

Clemens, J. D. & Vielzeuf, D. (1987). Constraints on melting and magma production in the crust. Earth and Planetary Science Letters 86, 287–306.[Web of Science]

Conrad, W. K., Nicholls, I. A. & Wall, V. J. (1988). Water saturated and undersaturated melting of metaluminous and peraluminous crustal compositions at 10 kb: evidence for the origin of the silicic magmas in the Taupo volcanic zone, New Zealand, and other occurrences. Journal of Petrology 29, 765–803.[Abstract/Free Full Text]

Crisci, G. M., Maccarrone, E. & Rottura, A. (1979). Cittanova peraluminous granites (Calabria, Southern Italy). Mineralogica et Petrographica Acta 23, 279–302.

Deer, W. A., Howie, R. A. & Zussman, J. (1967). An Introduction to the Rock Forming Minerals. Harlow: Longman.

Del Moro, A., Fornelli, A. & Piccarreta, G. (2000). Disequilibrium melting in granulite facies metasedimentary source of the northern Serre (Calabria–Southern Italy). Mineralogy and Petrology 70, 89–104.[Web of Science]

Di Battista, P. (1999). Studio petrologico e strutturale delle tonaliti erciniche di Squillace–Petrizzi (Serre–Calabria). Ph.D. thesis, University of Bari, 69 pp.

Dubois, R. (1976). La suture calabro-apénninique Crétace–Eocéne et l’ouverture thyrrénienne Neogene: étude pétrographique et structurale de la Calabre centrale. Ph.D. thesis, Université Pierre et Marie Curie, Paris, 557 pp.

Fornelli, A. (1991). Il plutonismo tardo ercinico nelle Serre orientali (Calabria Meridionale): aspetti petrologici. Ph.D. thesis, University of Bari, 133 pp.

Fornelli, A., Caggianelli, A., Del Moro, A., Bargossi, G. M., Paglionico, A., Piccarreta, G. & Rottura, A. (1994). Petrology and evolution of the central Serre granitoids (Southern Calabria–Italy). Periodico Mineralogia 63, 53–70.

Harris, N. & Inger, S. (1992). Trace element modelling of pelite-derived granites. Contributions to Mineralogy and Petrology 110, 46–56.

Harris, N., Inger, S. & Massey, J. (1993). The role of fluids in the formation of High Himalayan leucogranites. In: Treloar, P. J. & Searle, M. P. (eds) Himalayan Tectonics. Geological Society, London, Special Publications 74, 391–400.

Inger, S. & Harris, N. B. W. (1993). Geochemical constraints on leucogranite magmatism in the Langtang Valley, Nepal Himalaya. Journal of Petrology 34, 345–368.[Abstract/Free Full Text]

Johannes, W., Holtz, F. & Moller, P. (1995). REE distribution in same layered migmatites: constraints on their petrogenesis. Lithos 35, 139–152.[Web of Science]

Kretz, R. (1983) Symbols for rock-forming minerals. American Mineralogist 34, 345–368.

Kruhl, J. H. & Huntemann, T. (1991). The structural state of the former lower continental crust in Calabria (S. Italy). Geologische Rundschau 80(2), 289–302.

Le Breton, N. & Thompson, A. B. (1988). Fluid absent (dehydration) melting of biotite in metapelites in the early stages of crustal anatexis. Contributions to Mineralogy and Petrology 99, 226–237.[Web of Science]

Le Maitre, R. W. (1982). Numerical Petrology, Developments in Petrology 8. Amsterdam: Elsevier, 281 pp.

Maccarrone, E., Paglionico, A., Piccarreta, G. & Rottura, A. (1983). Granulite–amphibolite facies metasediments from the Serre (Calabria, Southern Italy): their protoliths and the processes controlling their chemistry. Lithos 16, 95–111.[Web of Science]

Mazzucchelli, M., Rivalenti, G., Vannucci, R., Bottazzi, P., Ottolini, L., Hofmann, A. W. & Parenti, M. (1992). Primary positive Eu anomaly in clinopyroxenes of low-crust gabbroic rocks. Geochimica et Cosmochimica Acta 59, 2363–2370.

McDonough, W. F. & Sun, S. s. (1995). The composition of Earth. Chemical Geology 120, 223–253.[Web of Science]

Moresi, M., Paglionico, A., Piccarreta, G. & Rottura, A. (1979). The deep crust in Calabria (Polia Copanello Unit): a comparison with Ivrea–Verbano zone. Memorie della Società Geologica, Padova 33, 233–242.

Otamendi, J. E. & Patiño Douce, A. E. (2001). Partial melting of aluminous metagreywackes in the Northern Sierra de Comechingones, Central Argentina. Journal of Petrology 42, 1751–1772.[Abstract/Free Full Text]

Paglionico, A. & Piccarreta, G. (1976). Le Unità del Fiume Pomo e di Castagna nelle Serre settentrionali (Calabria). Bollettino della Società Geologica Italiana 95, 1–11.

Paglionico, A. & Piccarreta, G. (1978). History and petrology of a fragment of the deep crust in the Serre (Calabria, Italy). Neues Jahrbuch für Mineralogie und Petrographie, Mitteilungen Monatshefte 9, 385–396.

Patiño Douce, A. E. & Beard, J. S. (1995). Dehydration-melting of biotite gneiss and quartz amphibolite from 3 to 15 kbar. Journal of Petrology 36, 707–738.[Abstract/Free Full Text]

Patiño Douce, A. E. & Harris, N. (1998). Experimental constraints on Himalayan anatexis. Journal of Petrology 39(4), 689–710.[Web of Science]

Patiño Douce, A. E. & Johnston, A. D. (1991). Phase equilibria and melt productivity in the pelite system: implications for the origin of peraluminous granitoids and aluminous granulites. Contributions to Mineralogy and Petrology 107, 202–218.[Web of Science]

Rottura, A., Bargossi, M. G., Caironi, V., Del Moro, A., Maccarrone, E., Macera, P., Paglionico, A., Petrini, R., Piccarreta, G. & Poli, G. (1990). Petrogenesis of contrasting Hercynian granitoids from the Calabrian Arc, Southern Italy. Lithos 24, 97–119.[Web of Science]

Sawyer, E. W. (1991). Disequilibrium melting and the rate of melt residuum separation during migmatization of mafic rocks from the Grenville Front, Quebec. Journal of Petrology 32, 701–738.[Abstract/Free Full Text]

Sawyer, E. W. (1994). Melt segregation in the continental crust. Geology 22, 1019–1022.[Abstract/Free Full Text]

Schenk, V. (1980). U–Pb and Rb–Sr radiometric dates and their correlation with metamorphic events in the granulite-facies basement of the Serre, Southern Calabria (Italy). Contributions to Mineralogy and Petrology 73, 23–38.[Web of Science]

Schenk, V. (1981). Synchronous uplift of the lower crust of the Ivrea Zone and of Southern Calabria and its possible consequences for the Hercynian orogeny in Southern Europe. Earth and Planetary Science Letters 56, 305–320.[Web of Science]

Schenk, V. (1984). Petrology of felsic granulites, metapelites, metabasics, ultramafics, and metacarbonates from Southern Calabria (Italy): prograde metamorphism, uplift and cooling of a former lower crust. Journal of Petrology 25, 255–298.[Abstract/Free Full Text]

Schenk, V. W. (1989). P–T–t path of the lower crust in the Hercynian fold belt of southern Calabria. In: Daly, J. S., Cliff, R. A. & Yardley, B. W. D. (eds) Evolution of Metamorphic Belts. Geological Society, London, Special Publications 43, 337–342.

Schenk, V. (1990). The exposed crustal cross section of Southern Calabria, Italy: structure and evolution of a segment of Hercynian Crust. In: Salisbury, M. H. & Fountain, D. M. (eds) Exposed Cross-Sections of the Continental Crust. Dordrecht: Kluwer Academic, pp. 21–42.

Skjerlie, K. P. & Johnston, A. D. (1996). Vapor absent melting from 10 to 20 kbar of crustal rocks that contain multiple hydrous phases: implications for anatexis in the deep to very deep continental crust and active continental margins. Journal of Petrology 37(3), 661–691.[Abstract/Free Full Text]

Solar, G. S. & Brown, M. (2001). Petrogenesis of migmatites in Maine, USA: possible source of peraluminous leucogranite in plutons? Journal of Petrology 42(4), 789–823.[Abstract/Free Full Text]

Spear, F. S., Kohn, M. J. & Cheney, J. T. (1999). P–T paths from anatectic pelites. Contributions to Mineralogy and Petrology 117, 375–393.

Stevens, G., Clemens, J. D. & Droop, G. T. R. (1997). Melt production during granulite-facies anatexis: experimental data from ‘primitive’ metasedimentary protoliths. Contributions to Mineralogy and Petrology 128, 352–370.[Web of Science]

Taylor, S. R. & McLennan, S. M. (1985). The Continental Crust: its Composition and Evolution. Oxford: Blackwell Scientific, pp. 1–312.

Thompson, A. B. & Connolly, J. A. D. (1995). Melting of the continental crust: some thermal and petrological constraints on anatexis in continental collision zones and other tectonic settings. Journal of Geophysical Research 100, 15565–15579.

Vavra, G., Gebauer, D., Schimd, R. & Compston, W. (1996). Multiple zircon growth and recrystallization during polyphase late Carboniferous to Triassic metamorphism in granulites of the Ivrea Zone (Southern Alps): an ion microprobe (SHRIMP) study. Contributions to Mineralogy and Petrology 122, 337–358.[Web of Science]

Vielzeuf, D. & Holloway, J. R. (1988). Experimental determination of the fluid absent melting relations in the pelitic system. Consequences for crustal differentation. Contributions to Mineralogy and Petrology 98, 257–276.[Web of Science]

Vielzeuf, D., Clemens, J. D., Pin, C. & Moinet, E. (1990). Granites, granulites, and crustal differentiation. In: Vielzeuf, D. & Vidal, Ph. (eds) Granulites and Crustal Evolution. Nato ASI Series. Dordrecht: Kluwer Academic, pp. 59–85.

Watt, G. R. & Harley, S. L. (1993). Accessory phase controls on the geochemistry of crustal melts and restite produced during water-undersaturated partial melting. Contributions to Mineralogy and Petrology 114, 550–556.[Web of Science]

Whitney, D. L. & Irving, A. J. (1994). Origin of K-poor leucosomes in a metasedimentary migmatite complex by ultrametamorphism, syn-metamorphic magmatism and subsolidus processes. Lithos 32, 173–192.[Web of Science]

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				Y. Tatsumi and T. Suzuki Tholeiitic vs Calc-alkalic Differentiation and Evolution of Arc Crust: Constraints from Melting Experiments on a Basalt from the Izu-Bonin-Mariana Arc J. Petrology, August 1, 2009; 50(8): 1575 - 1603. [Abstract] [Full Text] [PDF]

				P. Fiannacca, I. S. Williams, R. Cirrincione, and A. Pezzino Crustal Contributions to Late Hercynian Peraluminous Magmatism in the Southern Calabria-Peloritani Orogen, Southern Italy: Petrogenetic Inferences and the Gondwana Connection J. Petrology, August 1, 2008; 49(8): 1497 - 1514. [Abstract] [Full Text] [PDF]

				S. Tommasini, A. Heumann, R. Avanzinelli, and L. Francalanci The Fate of High-Angle Dipping Slabs in the Subduction Factory: an Integrated Trace Element and Radiogenic Isotope (U, Th, Sr, Nd, Pb) Study of Stromboli Volcano, Aeolian Arc, Italy J. Petrology, December 1, 2007; 48(12): 2407 - 2430. [Abstract] [Full Text] [PDF]

				P. ACQUAFREDDA, A. FORNELLI, A. PAGLIONICO, and G. PICCARRETA Petrological evidence for crustal thickening and extension in the Serre granulite terrane (Calabria, southern Italy) Geological Magazine, March 1, 2006; 143(2): 145 - 163. [Abstract] [Full Text] [PDF]

				J. DUFEK and G. W. BERGANTZ Lower Crustal Magma Genesis and Preservation: a Stochastic Framework for the Evaluation of Basalt-Crust Interaction J. Petrology, November 1, 2005; 46(11): 2167 - 2195. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (7) Request Permissions Google Scholar Articles by FORNELLI, A. Articles by ACQUAFREDDA, P. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

Online ISSN 1460-2415 - Print ISSN 0022-3530

Copyright © 2009 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics