Lithos 154 (2012) 130–146

Contents lists available at SciVerse ScienceDirect

Lithos journal homepage: www.elsevier.com/locate/lithos

Co-eruption of carbonate and silicate magmas during volcanism in the Limagne graben (French Massif Central) Gilles Chazot a,⁎, Juliette Mergoil-Daniel b a Université Européenne de Bretagne, France et Université de Brest, CNRS, UMR 6538 Domaines Océaniques, Institut Universitaire Européen de la Mer, Place Copernic, 29280 Plouzané, France b Laboratoire Magmas et Volcans, OPGC, Université Blaise Pascal, CNRS, IRD, 5 rue Kessler, 63038 Clermont-Ferrand cedex, France

a r t i c l e

i n f o

Article history: Received 13 March 2012 Accepted 30 June 2012 Available online 10 July 2012 Keywords: Peperite Carbonatite French Massif Central Geochemistry Basalt Volcanism

a b s t r a c t Peperites in the Limagne graben (French Massif Central) have been interpreted up to now as phreatomagmatic eruptions involving fragmentation of sedimentary rocks during magma–water interaction. We report about the possible magmatic origin of the carbonates involved in the peperites from Gergovie and Pileyre localities. In these two volcanic deposits, carbonates are manly dolomitic in composition, while the Limagne sediments are almost exclusively detrital formations or limestones. In the studied rocks, the carbonates can be found in different habits, but are sometimes closely associated with Cr-diopsides and Cr-spinels, minerals never found in the Limagne Miocene volcanic rocks and related to deep mantle processes. Dykes crosscutting the layered deposits are mainly composed of euhedral dolomite grains. Trace elements and Sr isotopes confirm that the carbonates are chemically different compared to the sedimentary rocks from the Limagne and therefore that the studied volcanic rocks are not a mixing between silicate magma and the local sediments. Based on our new petrological and geochemical data, we interpret these carbonates as evidence for the eruption of carbonatitic magmas associated with the silicate magmas found in the deposits. The high Sr isotopic ratios of these rocks imply the melting of an enriched metasomatised mantle for the source of the carbonates, as already advocated for many ultrapotassic or carbonatitic volcanic rocks in Europe. These results call for a reappraisal of the origin of the volcanism in the Limagne volcanic province and more generally of the carbonatitic volcanic provinces in Western Europe. © 2012 Elsevier B.V. All rights reserved.

1. Introduction The term peperite was used for the first time by Lecoq and Bouillet (1830) to describe and illustrate the volcaniclastic rocks from the Limagne area in the French Massif Central, when for the same rocks Scrope (1827, 1858) used the term of “peperino”. These rocks were generally interpreted as a mixture between sedimentary rocks (mainly limestones) and lava fragments. For a long time, the main debate was on the place of the volcanic event, during or after the sedimentation. A comprehensive view on the Limagne peperites petrography was elaborated by Michel (1953). We do not report here on the numerous and various historical hypotheses which have been expressed on these formations. The term peperite is now commonly used to describe rocks formed by intrusion of magma into wet sediments (Skilling et al., 2002; White et al., 2000), while the peperites from Limagne are now described as resulting from subaerial phreatomagmatic eruptions during which sedimentary rocks were pulverised after magma–water interaction (Goër de Herve, 2000).

⁎ Corresponding author. Tel.: +33 685356221; fax: +33 298498760. E-mail address: [email protected] (G. Chazot). 0024-4937/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.lithos.2012.06.032

A new look at similar deposits in the Calatrava province in Spain have recently shown that the carbonates associated with silicate magma in this volcanic province are of igneous origin, calling for a new interpretation of this kind of eruptions (Bailey et al., 2005; Stoppa et al., 2012). Following these observations, Bailey et al. (2006) assigned an igneous origin to the carbonates in the Puy de Crouel volcano, a typical diatreme from the Limagne graben, close to Clermont-Ferrand. Here we report on two other peperite volcanoes from the Limagne area, from which we present evidences of co-eruption of carbonate and silicate magmas from field, petrographic and geochemical observations. 2. Regional setting The Limagne graben is a component of the Cenozoic West European Rift System where the main distensive phases occurred during Oligocene and upper Miocene (Merle and Michon, 2001). The Oligocene sedimentation is variable in thickness and can reach up to 2500 m at the Riom trench in the north western part of the graben but only 750 m at the south western part near Gergovie. The clastic deposits are thicker along the western flank. Sedimentary deposits are made by several sequences, each one beginning by detrital formations, then by shales, clay limestones and ending by limestones (Briot and Poidevin, 1998; Wattinne, 2004). On the basis of associated faunae

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

and florae, the most recent deposits are interpreted as the products of a upper Miocene and Aquitanien lacustrine sedimentation (Wattinne, 2004). A widespread volcanism is present in the graben and along the two borders (Fig. 1). K–Ar and Ar–Ar dating (Bellon, 1971; Chantepie, 1990) show that these eruptions occurred between 25 and 10 Ma, even if some events are still not well dated and are possibly younger. According to these data, volcanism is for its main part posterior to sedimentation. Sr isotopic data on limestone rocks and bioclasts yield to the conclusion that there is no isotopic trace of synsedimentary volcanism in the Limagne area (Briot et al., 2001).

131

The volcanic activity consists mostly in small monogenic edifices. More than 230 (Nehlig et al., 2001) eruptions have been identified so far. Among them, more than 120 contain peperites which occurred as diatremes or pyroclastic deposits. Volcanic remnants are present as lava flows, dykes, necks, sills or lava lakes. From the total alkali silica (TAS) diagram (Le Bas et al., 1986) and normative compositions, most of the lavas (flows, dykes or sills) from the Limagne graben plot in the basanite and tephrite fields (Cheguer, 1996). Some rare samples are phono-tephrites and tephri-phonolites. Among the samples with the lowest silica content plotting in the U1 or F fields of the TAS diagram, some are nephelinites (normative

Fig. 1. Simplified geological map showing the main tectonic features of the Limagne graben, as well as the location of the volcanic rocks belonging to the Limagne volcanic province and to the Chaine des Puys. Gergovie and Pileyre volcanic systems are highlighted in an open square. Modified from Cheguer (1996).

132

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

Fig. 2. Photomicrographs of peperites from Gergovie. (a) Dolomite clast containing euhedral Cr-spinels in a fine-grained dolomitic matrix in sample G1. (b) Platy dolomite crystals with inclusions of Cr-diopside and Cr-spinel in a fine-grained dolomitic matrix (Sample G1). (c) Association of Cr-spinels and Cr-diopsides embayed in large dolomitic crystals showing a diffuse contact with the matrix (Sample G7). (d) Enlarged area from (c) seen in backscattered electron mode. Dol = dolomite, Di = diopside, Sp = spinel.

nepheline> 20%), but most of them have normative nepheline lower than 20% (between 14 and 20%) and very low normative albite (b 5%). They correspond to the melanephelinite rocks chemically defined in the IUGS nomenclature (Le Maitre et al., 2002). They also contain normative olivine and may be compared to the olivine-poor nephelinites described by Le Bas (1987) in relation with carbonatitic complexes. For the purpose of this study, we choose to work on two different peperite localities: 1) The Gergovie Plateau, on the West side of the

Limagne graben, because it represents the type-locality for the peperite definition in the Massif Central and shows nice outcrop conditions; 2) Pileyre peperites, in the centre of the graben, which show very fresh rocks with silicate–carbonate associations. The detailed structure of the Gergovie volcanic complex is still matter of debate. Several eruptive centres are superimposed, separated by sedimentary layers and affected by several erosion episodes. A recent survey (Degeai and Pastre, 2008) identified three different volcanic events with

Fig. 3. Photomicrographs of peperites from Gergovie and Pileyre. (a) Anhedral to euhedral dolomite crystals in a fine-grained isotropic matrix in sample G11. (b) Enlargement of a dolomitic crystal showing large chemical zoning in sample G11 (backscattered electron image). (c) Euhedral dolomitic crystal containing Cr-spinel grains, surrounded by melanephelinitic glass in sample PIL8. (d) Same area in backscattered electron mode. Dol = dolomite, Sp = spinel.

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

the formation of maars that they ascribed to phreatomagmatic eruptions. The possibility of CO2 dry-magmatic origin of the maars, as proposed by Stoppa et al. (2003, 2012) was not considered at that time. Paleontological evidences indicate that the first two maars were emplaced during lower Aquitanian, around 23 Ma ago. The last eruptive event is the formation of the Puy Mardoux, on the Eastern side of the Gergovie plateau and is contemporaneous with the lava flow covering the plateau. The lava flow and some intrusive basaltic lavas from Puy Mardoux have been dated and gave ages between 15 and 19 Ma for the flow and between 15 and 18.2 Ma for the Puy Mardoux (Bellon, 1971; Bout et al., 1966; Camus et al., 1969; Cantagrel and Boivin, 1978). The peperites appear to be very complex and show different facies types. The main part is well stratified with variable grain size. The amount of lava fragments is highly variable and in some layers they are totally absent. Small basaltic sills are present in the stratified peperites. In different places, dykes and sills mainly made of carbonate are interstratified or crosscut the peperites. They have been interpreted as lithified sediments injected into the peperites during the eruption (Bessonneau, 1997; Peterlongo, 1978). The composition of the silicate magmas at Gergovie volcano ranges from basanite to tephrite and phonotephrite in Puy Mardoux. The Pileyre area is also a complex volcanic edifice. Massive peperites from the Puy de Challas are crosscut by numerous small dykes. They are interpreted as an intrusive facies while coarsely stratified peperites form the large plateau between Vertaizon and Chauriat localities (Gouhier, 1973). The top of the Puy Pileyre (530 m) is overlaid by siliceous travertine and calcareous deposits with algae bioconstructions and Trichoptera larvae often described in the late Oligocene Limagne sedimentation. The studied samples come from the intrusive and stratified parts of the Pileyre volcano. No age information is available for the emplacement of the Pileyre peperite which is most probably contemporaneous to the main peperite volcanism in the Limagne area.

3. Analytical techniques Whole-rock major elements have been analysed by ICP-AES in the Laboratoire Magmas et Volcans (LMV), Clermont-Ferrand, France. X-ray diffraction analyses (XRD) were performed on a CGR S2080 diffractometer (LMV) using λKα1 Cu (focusing quartz monochromator on the incident beam) and numerical acquisition (100 steps/° θ with 5 s counting time by step). Major element composition of the minerals have been analysed using a CAMECA SX100 electron microprobe (LMV, Clermont-Ferrand, France), turned at 15 kV voltage,10 nA current and 20 s peak counting time with Na measured first. For the carbonates, the beam size was enlarged to 10 μm. Trace elements in glasses and carbonates were determined by laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) at Institut Universitaire Européen de la Mer, Plouzané, France. The analyses were performed under a He atmosphere using an Excimer (193 nm wavelength) laser ablation system (Geolas Pro102), connected to a Thermo Element 2 spectrometer operated in low resolution mode. Analytical and data reduction procedures generally followed those described by Agranier and Lee (2007). The laser beam diameter was fixed at 60 μm. We used a laser repetition rate of 6 Hz and the power output of the laser was approximately 15 J/cm2. Results were normalised to CaO abundance measured by electron microprobe as an internal standard to account for variable ablation yield. For all data, the NIST 612, BIR-1G and BCR-2G glass standards were used for external calibration of relative element sensitivities, using values given by Gao et al. (2002). External reproducibility is generally better than 5% for most of the elements. Strontium isotopic ratios were measured with a VG ISOMASS 54E mass spectrometer (LMV, Clermont-Ferrand, France). The average value of the NBS987 standard Sr isotope composition measured during this study is 0.710238±12 (n=2).

133

4. Petrology and mineralogy of the peperites We selected 5 facies types from Gergovie and 2 from Pileyre as representative of the most important features of the silicate–carbonate association, and for which we acquired mineralogical and chemical data. The Gergovie samples G1, G4, G7, G11 and G14 were collected from the eastern side of the Gergovie Plateau, along the section between Gergovie village and Aubière via the Puy Mardoux. According to Degeai and Pastre (2008) they belong to the first volcanic episode; G7 was sampled at the Puy Mardoux and belongs to the last eruptive phase. G1 is a layered peperite mainly constituted by a fine grained carbonate matrix (grain size average b 1 μm) which contains various inclusions. XRD analyses on the whole rock detected dolomite as the main and exclusive carbonate component, associated with minor amounts of smectite and analcime. The most important feature of this sample is the presence of large clasts (up to 1 mm) of polycrystalline or monocrystalline dolomite with angular shape (Fig. 2a). The contact with the matrix is often enhanced by an analcime-rich zone. These clasts frequently contain Cr-spinel grains (which are never found in the matrix of the rock). Another characteristic feature is the occurrence of spots (up to 500 μm) with very irregular boundaries and made of dolomite crystals, looking like quench-textured dolomite carbonates (Andersen, 2008). Every spot is surrounded by an isotropic analcime mantle, as the dolomite clasts described above. Pale green clinopyroxene crystals also occur as clasts and are in some cases partially replaced by dolomite. Another interesting feature of this sample is the occurrence of platy dolomite crystals up to 300 μm long (Fig. 2b). Locally some oval or round carbonate lapilli with narrow cryptocrystalline brown margins are observed and may be up to 450 μm long. They are mainly constituted by the same material as the matrix but some are composite, with a kernel of a dolomite clast or a gathering of carbonate crystals larger than the average size of the matrix crystals (about 4 μm). This sample is a carbonate ash tuff. G7 from Puy Mardoux is mainly composed of a matrix made of small (b 20 μm long) dolomite grains and numerous spheres of dark and isotropic material (b 200 μm) identified as analcime. This mineral may be of primary origin but is more probably a replacement of a former silicate glass. In some places, the association of diopside, dolomite and spinel crystals forms millimetre-sized aggregates where spinel crystals can be up to 100 μm long (Fig. 2c, d). Large fragments (up to 1 cm) of lava and carbonate aggregates are scattered in the matrix, as well as patches or streaks of larger carbonate grains (up to 150 μm) without any associated silicates and showing diffuse contact with the matrix. The lava fragments have a sharp contact with the matrix. Mainly constituted by a red to yellow altered glass, they are stretched and display scoriaceous textures with elongated voids filled with analcime and carbonates. G14 is a composite sample. One part is a 3 cm large lava fragment made of glass, clinopyroxene and plagioclase with large vesicles filled with secondary products with fan texture. The other part is composed of pyroclasts up to 5 mm large containing fresh silicate glass and displaying smooth sinuous shape. The silicate glass does not contain any carbonate globules but many and large fresh clinopyroxene crystals (zoned phenocrysts and microphenocrysts) and some vesicles filled with secondary products. These pyroclasts are embayed in a matrix of large dolomite crystals (up to 300 μm) containing polycrystalline dolomite clasts (up to 1 mm) with sharp contact and including Cr-spinel. Some of these clasts are also found in the pyroclasts. G4 comes from a dyke crosscutting the layered Gergovie peperites, while G11 is a small sill interbedded in the same peperites. XRD analyses on the whole rocks show that dolomite is the main mineral phase with only traces of analcime. After HCl attack, XRD analyses detected potassic feldspar as the main mineral phase and a well crystallised

134

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

Table 1 Representative analyses of carbonates from Gergovie and Pileyre peperites. Sample

G1

Analysis no.

rd3-26J

rd7-31J

rd4-35J

Rd1-11J

Rd1-14J

Rd1-6J

z13-23A

z3-12A

z3-18A

z1-6A

z1-7A

z1-8A

rd5-45J

rd5-46J

1

1

1

2

2

2

3

4

4

4

4

4

4

4

ppm

31.60 16.73 3.6 0.11 45.35 97.39 1377

31.85 18.15 1.54 0.10 45.83 97.48 1692

32.49 18.00 2.22 0.15 46.61 99.47 1309

30.45 18.60 3.09 0.11 46.17 98.42 2950

28.23 18.97 4.36 0.04 45.58 97.19 3460

53.94 1.24 0.11 0.00 43.76 99.05 425

31.95 19.69 1.25 0.06 47.39 100.34 3621

31.24 18.66 2.91 0.10 46.74 99.65 5466

31.82 19.51 2.19 0.11 47.70 101.34 4233

31.26 20.08 1.72 0.71 47.96 101.73 2516

32.17 17.93 3.02 0.08 46.73 99.92 1921

32.93 20.14 1.07 0.01 48.51 102.67 1233

32.48 16.96 2.82 0.13 45.83 98.22 1700

33.11 18.76 0.34 0.07 46.73 99.00 1853

Dol

1.09 0.81 0.10 0.00 Dol

1.09 0.86 0.04 0.00 Dol

1.09 0.84 0.06 0.00 Dol

1.04 0.88 0.08 0.00 Dol

0.97 0.91 0.12 0.00 Fe-Dol

1.94 0.06 0.00 0.00 Cc

1.06 0.91 0.03 0.00 Dol

1.05 0.87 0.08 0.00 Dol

1.05 0.89 0.06 0.00 Dol

1.02 0.91 0.04 0.02 Dol

1.08 0.84 0.08 0.00 Dol

1.07 0.91 0.03 0.00 Dol

1.11 0.81 0.08 0.00 Dol

1.11 0.88 0.01 0.00 Dol

Location CaO MgO FeO MnO CO2 calc. Total Sr

wt.% wt.% wt.% wt.% wt.%

Cations per 6 O Ca Mg Fe Mn Carbonate type

G7

G4

G11

Dol = dolomite, Fe-Dol = ferroan dolomite, Mg-Sdte = Magnesiosiderite, Cc = calcite.Location: 1: Clast, 2: Matrix, 3: Globule, 4: Euhedral crystals, 5: Clusters in the matrix. PIL5

PIL8

62/1

74/1

75/1

44/1

11/1

90/1

91/1

78/1

z1-154

z1-153

z6-162

z6-163

81/1

101/1

3

4

4

5

5

3

3

4

1

1

4

4

2

2

mica. The two samples have similar texture: they are mainly composed of small (up to 400 μm) anhedral to euhedral dolomite rhombs embayed in a fine-grained isotropic matrix containing very small dolomite grains, K-feldspar and small grains of mica (Fig. 3a, b). In some places, the dolomite grains are in contact, without any matrix. G4 sample also contains the same spinel-bearing dolomite clasts as already described in sample G1. These samples do not contain any recognisable fossils or fragments of sediment and are difficult to interpret as lithified sediments mixed with a basaltic magma. Chemical and isotopic arguments developed below will confirm that the dolomite crystals in these samples are probably of magmatic origin. Two samples were selected in the Pileyre volcanic complex. PIL5 was sampled in an ancient quarry located at the upper part of the Puy de Challas vent. PIL8 is representative of the stratified peperites. PIL5 is mainly constituted by adjoined heterometric pyroclasts (from 1 μm to several millimetres in size) in a cryptocrystalline matrix. The pyroclasts, even the smaller ones, have sinuous boundaries displaying, in some places, open rounded shapes (25 to 60 μm in diameter) showing that these pyroclasts broke into pieces. They are made of fresh silicate glass and include numerous monocrystalline or polycrystalline dolomite globules. Most of them have a round shape, but some are coalescent and have an amoeboid shape as observed before in the Chabrières tuff (Chazot et al., 2003) and also described in the Laguna Bianca tuff from Calatrava volcanic field in Spain (Stoppa et al., 2012). They also contain numerous clinopyroxene crystallites (up to 2 μm) and some euhedral mm-sized clinopyroxene crystals. The pyroclasts also contain Cr-spinel-bearing dolomite clasts, very similar to those found in Gergovie samples, as well as orthopyroxene crystals with reaction rims. Magnesite is the carbonate phase of the cryptocrystalline matrix and appears, in some places, concentrated under mm-sized “bothryoid spots”. Whole rock XRD analyses confirm that magnesite and dolomite are the main carbonate phases, with only minor amounts of calcite. PIL8 contains large glassy pyroclasts (millimetre sized) embayed in a carbonate matrix of large calcite crystals (up to 1 mm). Some large fragments of crustal rocks (polycrystalline quartz and granitic feldspars) are also present. The pyroclasts are similar to those of PIL5 but have a smoother shape and the micropyroclasts (μm sized) are missing. As in PIL5, the silicate glass is generally fresh, contains clinopyroxene and the pyroclasts include numerous monocrystalline or polycrystalline carbonate globules (made of dolomite or dolomite/calcite). More spectacular is the presence of euhedral carbonate

crystals in the silicate glass (Fig. 3c, d). They contain Cr-spinel and show no sign of reaction with the silicate melt. 5. Chemical data 5.1. Major and trace elements of minerals Carbonates, Cr-spinels and pyroxenes are the main minerals of the studied samples. There chemical compositions are presented in Tables 1, 2 and 3. Other minerals (Table 4) are present in small amounts or in specific places and they will be discussed in their specific context. They are: small smectite area in silicate glass and around some dolomite clasts in G1; sporadic analcime in Gergovie peperites except in the dolomitic matrix of the Mardoux peperite (G7) rich in analcime microspherules; K-rich minerals in dolomitic dykes G4 and G11 in Gergovie (potassic feldspars, Or95–97), micas and zeolite (offretite) with K and Mg as dominant extra framework cations. 5.1.1. Carbonates In Gergovie and Pileyre peperites, dolomite is the main carbonate phase except in some specific locations. The Sr content of dolomite from Gergovie is high and varies from 1200 to 5400 ppm (Tables 1 and 6). In the dykes G4 and G11, euhedral dolomite crystals are frequently zoned. While Mn, Sr and Ba systematically decrease from core to rim, Ca, Fe and Mg variations are more complex. In most cases, Mg is high in the core and in the rim and lower in the intermediate part of the crystals while Fe follows a reverse pattern. Ca content is even more variable, with an increase from core to rim in some crystals or with higher content in the intermediate part in other crystals. In Gergovie peperites, calcite is present in some secondary infillings with a low Sr content (~ 400 ppm). In Pileyre peperites, the matrix enclosing the dolomite-bearing silicate pyroclasts is different according to the geological setting; calcite in the layered peperites (PIL8); magnesite in the intrusive peperites (PIL5). This magnesite (according to Buckley and Woolley (1990) with Mg/Mg + Fe > 0.75) contains some amount of Ca. The large-size Ca ion has a significant effect on the cell parameters in comparison with Fe and Mg ions. The cell parameters calculated from the XRD data (a= 4.66 Å; c = 15.15 Å) are in agreement with the theoretical cell including the Ca, Fe and Mg carbonate components calculated from the chemical analyses. Ca element belongs to the structure and

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

135

Table 1 (continued) PIL5

PIL8

62/1

74/1

75/1

44/1

11/1

90/1

91/1

78/1

z1-154

z1-153

z6-162

z6-163

81/1

101/1

3

4

4

5

5

3

3

4

1

1

4

4

2

2

26.68 23.11 1.62 0.04 47.20 98.66 –

30.00 16.90 4.75 0.09 44.98 96.72 –

30.53 17.94 4.42 0.13 46.35 99.37 –

2.55 40.89 5.78 0.00 50.19 99.41 –

2.567 36.30 11.24 0.29 48.73 99.13 –

33.66 12.68 7.85 0.11 45.15 99.45 –

31.27 19.65 1.30 0.19 46.92 99.33 –

34.48 11.17 7.57 0.00 43.91 97.13 –

33.63 12.62 8.19 0.13 45.28 99.84 –

57.12 0.43 0.10 0.26 46.08 104.89 –

35.49 12.42 6.79 0.13 45.66 100.49 –

57.45 0.10 0.40 0.11 45.52 103.59 –

53.64 0.28 0.88 0.00 42.95 97.74 –

52.54 0.60 0.59 0.02 42.28 96.04 –

1.03 0.85 0.12 0.00 Fe-Dol

0.08 1.78 0.14 0.00 Mg-sdte

0.08 1.63 0.28 0.01 Mg-sdte

1.17 0.61 0.21 0.00 Fe-Dol

1.05 0.91 0.03 0.00 Dol

1.23 0.56 0.21 0.00 Fe-Dol

1.17 0.61 0.22 0.00 Fe-Dol

1.95 0.02 0.03 0.01 Cc

1.22 0.59 0.18 0.00 Fe-Dol

1.98 0.00 0.01 0.00 Cc

1.96 0.01 0.03 0.00 Cc

1.95 0.03 0.02 0.00 Cc

Cations per 6 O 0.89 1.05 1.07 0.82 0.04 0.13 0.00 0.00 Dol Fe-Dol

not to calcitic or dolomitic micro inclusions, so the representative formula for these carbonates is Mg1,7 Fe0,2 Ca0,1 (CO3)2. As demonstrated recently by Rosatelli et al. (2010), the geochemical discrimination between igneous and sedimentary carbonates is sometimes difficult. In order to characterise the trace element composition of the different types of carbonates present in the peperite samples as well as in the underlying sediments, the trace element composition of the different carbonates has been obtained by laser ablation ICP-MS with a direct control of the location and nature of the analysed carbonates. For the Gergovie underlying sediment G0, a large laser beam diameter has been used to get an average composition of the sediment

(Table 5). G0 carbonates are enriched in light rare earths (LREE) compared to heavy rare earths (HREE, Fig. 4). They have a well pronounced negative Eu anomaly. The four analyses are very homogeneous, with high Li, Rb, Sr, Cs and Ba content. They also contain up to 3 ppm Th. REE patterns of the Gergovie dolomites (Table 6) are not very different compared to the sediments (Fig. 4). They have the same shape, and the Gergovie dolomites can contain either higher or lower amounts of rare earths. The negative Eu anomaly is only present in the dolomite from the G4 and G11 dykes even if a small anomaly can be observed in some crystals from G7, mostly in the matrix carbonates. As a whole, Gergovie dolomites have heterogeneous trace element compositions,

Table 2 Representative analyses of spinels from Gergovie and Pileyre peperites. Sample

G1

G1

G1

G7

G7

G14

G14

PIL5

PIL8

PIL8

Analysis no.

122

125

126

131

136

167

174

28/1

145

146

148

Location

1

1

1

2

2

1

2

1

1

1

1

SiO2 Al2O3 TiO2 Cr2O3 Fe2O3calc FeOcalc MnO MgO CaO Total

0.08 28.70 2.32 25.06 11.65 18.73 0.34 12.56 0.16 99.59

0.09 22.51 2.49 24.71 16.88 21.31 0.29 9.67 0.77 98.74

0.05 27.57 3.26 22.35 13.15 17.47 0.25 12.77 1.09 97.96

0.11 27.79 1.89 28.71 10.24 16.65 0.18 13.77 0.04 99.39

0.11 28.00 1.92 27.57 10.50 16.79 0.21 13.49 0.12 98.70

0.08 30.41 2.16 24.76 10.75 16.10 0.22 14.25 0.26 98.98

0.05 29.03 2.27 24.65 11.65 17.78 0.19 13.02 0.30 98.94

0.11 19.20 1.55 38.20 9.04 13.41 0.33 13.91 0.13 95.89

0.10 22.85 1.46 36.47 8.89 13.67 0.25 14.00 1.03 98.72

0.05 27.91 1.67 30.10 8.76 14.61 0.28 13.98 0.88 98.23

0.09 26.43 1.70 30.92 9.59 14.18 0.25 14.07 1.00 98.23

Cations on the basis of 4 Si Al Cr Fe3+ Ti Sum Mg Fe2+ Mn Ca Sum Total Fe2+/Fe2+ + Mg Cr/Cr + Al

O and 3 cations 0.002 0.003 1.024 0.846 0.600 0.623 0.265 0.405 0.053 0.060 1.944 1.937 0.567 0.460 0.474 0.568 0.009 0.008 0.005 0.026 1.055 1.062 3.000 2.999 0.46 0.55 0.37 0.42

0.001 0.999 0.543 0.304 0.075 1.922 0.585 0.449 0.007 0.036 1.077 2.999 0.43 0.35

0.003 0.989 0.685 0.233 0.043 1.953 0.620 0.420 0.005 0.001 1.046 2.999 0.40 0.41

Location: 1: In carbonate clasts, 2: In diopside–dolomite–spinel associations.

0.003 1.003 0.662 0.240 0.044 1.952 0.611 0.427 0.005 0.004 1.047 2.999 0.41 0.40

0.002 1.070 0.584 0.242 0.048 1.946 0.634 0.402 0.006 0.008 1.050 2.996 0.39 0.35

0.002 1.037 0.591 0.266 0.052 1.948 0.588 0.451 0.005 0.010 1.054 3.001 0.43 0.36

0.004 0.728 0.971 0.219 0.037 1.959 0.667 0.361 0.009 0.005 1.042 3.000 0.35 0.57

0.003 0.831 0.889 0.206 0.034 1.963 0.644 0.353 0.007 0.034 1.038 3.001 0.35 0.52

0.001 0.999 0.722 0.200 0.038 1.960 0.632 0.371 0.007 0.029 1.039 2.999 0.37 0.42

PIL8

0.003 0.950 0.746 0.220 0.039 1.958 0.640 0.362 0.007 0.033 1.042 3.000 0.36 0.44

136

Table 3 Representative analyses of pyroxenes from Gergovie and Pileyre peperites. Sample

G1

G1

G4

G14

G7

G7

G7

G7

G7

PIL5

PIL5

PIL5

PIL5

PIL5

PIL5

PIL5

PIL8

PIL8

104

123

56

177

19

20

141

142

137

8/1

6/1

36/1

41/1

60/1

61/1

65/1

86/1

89/1

Location

1

1

1

2

2

2

2

2

3

4

4

2

2

2

2

2

2

2

SiO2 TiO2 Al2O3 Fe2O3calc Cr2O3 FeOcalc MnO MgO CaO Na2O K2O Total

43.90 3.71 9.70 4.72 0.00 4.88 0.20 10.49 21.71 0.99 0.00 100.29

43.48 4.35 9.12 4.24 0.03 3.84 0.09 11.29 23.20 0.47 0.00 100.11

49.29 1.74 5.29 2.47 nd 2.82 0.09 14.37 22.76 0.61 0.00 99.43

49.93 1.63 4.39 2.27 0.24 4.20 0.13 14.60 22.30 0.50 0.00 100.17

47.75 2.01 6.88 3.18 nd 2.62 0.12 13.55 22.72 0.62 0.02 99.47

46.25 3.13 7.44 3.69 nd 3.82 0.02 12.69 22.52 0.61 0.01 100.18

48.68 1.02 4.67 5.09 0.01 8.79 0.61 8.62 21.24 1.55 0.00 100.28

48.77 2.14 7.33 2.12 0.36 4.63 0.17 13.17 21.86 0.86 0.00 101.41

54.00 0.18 1.52 0.59 0.34 2.01 0.01 16.35 25.81 0.11 0.00 100.92

52.20 0.27 5.02 0.72 1.02 2.65 0.10 16.36 21.37 0.72 0.00 100.44

55.17 0.01 4.13 0.53 0.39 5.78 0.14 33.04 0.70 0.05 0.01 99.95

55.99 0.04 4.04 0.00 0.51 6.16 0.12 32.68 0.80 0.12 0.02 100.48

53.65 0.41 2.10 0.00 2.08 3.52 0.09 18.12 18.55 0.92 0.03 99.46

52.90 0.24 4.82 1.31 0.65 1.75 0.10 16.17 22.32 0.91 0.00 101.18

40.32 5.75 11.90 4.70 0.06 2.56 0.06 10.66 23.42 0.39 0.00 99.83

49.53 1.76 5.52 1.59 0.43 3.31 0.09 15.13 21.98 0.43 0.09 99.86

47.51 2.31 5.67 3.10 0.05 3.38 0.11 13.42 23.46 0.31 0.00 99.31

41.63 5.03 10.20 4.44 0.23 2.58 0.09 11.33 23.14 0.36 0.05 99.08

1.824 0.176 2 0.054 0.048 0.069 0.000 0.792 0.087 0.003 1.053 0.902 0.044 0.000 0.946 3.999 Di.

1.841 0.159 2 0.032 0.045 0.063 0.007 0.802 0.129 0.004 1.082 0.881 0.036 0.000 0.917 3.999 Di.

1.772 0.228 2 0.073 0.056 0.089 0.000 0.749 0.081 0.004 1.052 0.903 0.045 0.001 0.949 4.001 Di.

1.72 0.28 2 0.046 0.088 0.103 0.000 0.703 0.119 0.001 1.060 0.897 0.044 0.001 0.942 4.002 Di.

1.85 0.15 2 0.059 0.029 0.146 0.000 0.488 0.279 0.020 1.021 0.865 0.114 0.000 0.979 4.000 Aeg.aug.

1.78 0.22 2 0.095 0.059 0.058 0.010 0.716 0.141 0.005 1.084 0.855 0.061 0.000 0.916 4.000 Di.

1.954 0.046 2 0.018 0.005 0.016 0.010 0.881 0.061 0.000 0.991 1.001 0.008 0.000 1.009 4.000 Di.

1.887 0.113 2 0.101 0.007 0.02 0.029 0.881 0.080 0.003 1.121 0.828 0.051 0.000 0.879 4.000 Di.

1.906 0.094 2 0.074 0.000 0.014 0.011 1.701 0.167 0.004 1.971 0.026 0.003 0.001 0.03 4.001 En.

1.923 0.077 2 0.086 0.001 0.000 0.014 1.672 0.177 0.003 1.953 0.029 0.008 0.001 0.038 3.991 En.

1.952 0.048 2 0.042 0.011 0.000 0.060 0.983 0.107 0.003 1.206 0.723 0.065 0.001 0.789 3.995 Endi.

1.896 0.104 2 0.100 0.007 0.035 0.019 0.864 0.052 0.003 1.080 0.857 0.063 0.000 0.920 4.000 Endi.

1.519 0.481 2 0.048 0.163 0.133 0.002 0.598 0.081 0.002 1.027 0.945 0.029 0.000 0.974 4.001 Fass.Aug.

1.821 0.179 2 0.060 0.049 0.044 0.013 0.829 0.102 0.030 1.127 0.866 0.030 0.040 0.936 4.063 Di.

1.777 0.223 2 0.027 0.065 0.087 0.001 0.748 0.106 0.003 1.037 0.940 0.023 0.000 0.963 4.000 Di.

1.576 0.424 2 0.032 0.143 0.127 0.007 0.639 0.082 0.003 1.033 0.939 0.027 0.002 0.968 4.001 Fass.Aug.

Cations on the basis of 6 oxygens Si 1.644 1.632 AlIV 0.356 0.368 Sum 2 2 AlVI 0.072 0.035 Ti 0.104 0.123 3+ Fe 0.133 0.120 Cr 0.000 0.001 Mg 0.585 0.631 2+ 0.153 0.120 Mn 0.006 0.003 Sum 1.053 1.030 Ca 0.871 0.933 Na 0.072 0.034 K 0.000 0.000 Sum 0.943 0.970 Total 3.996 4.000 Pyroxene Di. Di.

Di. = diopside, Aeg.aug. = aegyrine augite, En. = enstatite, Endi. = endiopside, Fass.Aug. = fassaite-augite. Location: 1: In dolomite clasts, 2: In glass, 3: In diopside–dolomite–spinel associations, 4: In the carbonate matrix.

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

Analysis no.

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

137

Table 4 Representative analyses of accessory mineral (wt.%) phases in Gergovie peperites. Sample

G7

G11

G11

G1

G1

G7

Analysis no.

An1

An.44

z4

An.32

#4v‐9

#1v‐4

SiO2 Al2O3 TiO2 Cr2O3 FeOtot MnO MgO CaO Na2O K2O BaO P2O5 Total

64.93 18.80 – – 0.02 0.02 0.01 0.16 1.81 14.49 0.41 – 100.64 K-Feld.

64.88 18.31 – – 0.04 0.05 0.00 0.05 0.35 16.67 0.09 – 100.44 K-Feld.

49.78 29.42 – – 1.46 0.02 2.06 0.17 0.04 9.34 – – 92.29 Dioct. Mica

45.02 9.19 0.71 – 12.18 0.01 9.42 1.87 0.04 1.56 – – 79.99 Smectite

55.04 23.10 0.01 – 0.02 0.04 0.04 0.16 13.71 0.35 – – 92.48 Analcime

55.55 25.13 0.02 – 0.06 0.00 0.01 0.15 10.80 0.15 – – 91.88 Analcime

Cations per 8 oxygens Si Al Mg Fe Mn Ca Ba Na K Total

2.977 1.016 0.001 0.001 0.001 0.008 0.007 0.160 0.848 5.019

2.996 0.996 – 0.001 0.002 0.003 0.002 0.032 0.982 5.013

An Ab Or

1 15 84

0 3 97

G1

G4

G11

48.93 16.61 0.56 – 4.94 0.03 3.90 0.77 0.12 5.32 – – 81.16 Zeolite

48.08 13.56 – – 4.82 0.00 5.03 0.94 0.42 4.94 – – 77.79 Zeolite

47.48 15.42 – – 6.15 0.00 4.27 0.99 0.19 5.06 – – 79.55 Zeolite

Cations per 22 Oxygens

Cations per 6 oxygens

Cations per 72 oxygens

Si AlIV Sum AlVI Fe Mn Mg Ti Sum K Na Ca Sum

Si Al Fe2+ Mn

2.005 0.992 0.001 0.001

2.004 1.068 0.002 0

Si Al Ti Fe3+

24.395 9.761 0.209 2.060

25.043 8.32 0 2.098

24.285 9.295 0 2.630

Ti Mg Ca Na K Total

0 0.002 0.006 0.968 0.016 3.991

0 0 0.006 0.755 0.007 3.842

Mn Mg Ca Na K

0.011 2.895 0.410 0.113 3.381

0 3.909 0.526 0.423 3.282

0.001 3.258 0.542 0.183 3.299

Si/Al Si/Si + Al Si/Al + Si + Fe3+ DEC SEC m H2O

2.50 0.71 0.67 K, Mg Ca, Na 31

3.01 0.75 0.71 Mg, K Ca, Na 39

2.61 0.72 0.67 K, Mg Ca, Na 35

6.758 1.242 8 3.465 0.166 0.002 0.417 4.05 1.618 0.01 0.025 1.653

7.09 0.91 8 0.79 1.60 0.00 2.21 0.09 4.69 0.32 0.01 0.31 0.64

G1 An.108 44.43 10.39 1.29 0.04 16.30 0.34 10.99 11.74 1.24 0.86 0.12 0.12 97.73 Amphibole Cations per 23 oxygens Si AlIV Sum AlVI Ti Cr Fe3+ Mg Fe2+ Mn Sum Ca Na Sum Na K Sum

6.555 1.445 8.000 0.362 0.143 0.005 0.532 2.417 1.478 0.043 4.980 1.856 0.144 2 0.191 0.161 0.352

K-Feld. = potassic feldspar, Dioct. Mica = dioctahedral mica. DEC = dominant extra framework cations, SEC = subsidiary extra framework cations.

with Lan ranging from 6 to 100. Sr content is similar to the sedimentary carbonates, but the dolomites are clearly distinguished from the sediments by their lower Rb, Ba, Nb, Li and Cs content. In most cases, Hf, Ta and Pb are also less abundant in the dolomite than in the G0 sediments. REE patterns from the Pileyre carbonates (Table 7) are also roughly parallel to those observed in the sediments and in Gergovie samples (Fig. 4), but the concentrations are always lower, and no Eu negative anomaly is observed. Carbonates from the matrix (calcite), probably of secondary origin, have high REE content. As for Gergovie, the Pileyre carbonates have lower Rb, Nb, Ba, Li and Cs content than the sediments. They also have very low Y and Zr content.

5.1.2. Cr-Spinels In Gergovie and Pileyre peperites, Cr-spinel crystals occur within carbonate dolomite clasts which are embayed in the silicate glass or in the microcrystalline carbonate matrix. They all belong to the Mg–Al chromite group (Table 2). In the most usual representations, Cr/Fe3+/ Al or Fe2+/Fe2+ + Mg vs. Cr/Cr + Al or Fe3+/Fe3+ + Al+ Cr (Barnes and Roeder, 2001), they are different from the spinels analysed in the lavas from Limagne volcanism (Cheguer, 1996) and from the spinels included in pyroxenes from the alkali basalts from the French Massif Central (Boumedhi, 1988) which all plot in the titanomagnetite range compositions. In the Cr/Cr + Al vs. Fe 2+/Fe 2+ + Mg diagram (Fig. 5), they plot in the large composition field of the spinels from mantle peridotite xenoliths (Barnes and Roeder, 2001; Matsukage and Kubo, 2003), but they plot outside the field of the spinels from Massif Central mantle peridotites (Chazot, unpublished data) but in the field of the more Cr/

Cr+ Al-rich Cr-spinels from harzburgite peridotites (Matsukage and Kubo, 2003). The Cr-content is slightly higher in spinels from the Pileyre samples than from Gergovie. There is no significant chemical zoning in spinel crystals but in some grains, Cr2O3 can vary from 36 wt.% in the core to 31 wt.% in the rim.

5.1.3. Pyroxenes Peperites display several pyroxene types in specific locations and associations (Table 3). – Enstatite (0.4–0.5% Cr2O3) is only present in the Pileyre intrusive peperite (PIL5) as xenocrysts in the matrix and also within the silicate glass where it appears corroded and rimmed by a symplectic association made of endiopside, magnesite and silica. Compared to enstatite, endiopside contains Na and more Cr (0.8–2% Cr2O3). – In the dolomite/Cr-sp/pyroxene agglomerates from G7 peperite, the pyroxene is pure diopside with very low Al and Ti contents. – Pyroxenes in the peperites are diopsides (Morimoto, 1988) or diopsidic extending to ferroan diopsides and magnesian augites (Rock, 1990), with an Al/Ti ratio about 3 to 4. The Al versus Ti domain ranges from Al 0.2–Ti 0.05 to Al 0.5–Ti 0.16 (on the basis of 6 O). They contain significant amounts of Na (up to 1.5% Na2O) and the Cr2O3 content in the Pileyre pyroxenes is higher than in Gergovie. Diopside with higher Al and Ti contents is found at the rim of Cr-diopsides and corresponds to a late crystallisation stage. This is well illustrated in G7 sample where a Na-rich pyroxene crystal (11.4% aegyrine component) is also rimmed by a light colour diopside.

138

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

100

G0-1

G0-2

G0-3

G0-4

Li Be B Ti Cr Mn Co Ni Rb Sr Y Zr Nb Cs Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Hf Ta Pb Th U

473.7 1.48 133.9 618.9 10.37 371.9 0.64 6.84 92.55 1313.4 4.19 13.06 3.05 15.04 206.6 7.01 12.68 1.41 5.15 0.96 0.16 0.95 0.12 1.06 0.14 0.40 0.33 0.05 0.32 0.26 9.35 3.03 0.84

420.8 1.32 63.1 474.1 15.20 455.7 2.14 7.41 78.19 1329.3 4.21 10.99 2.41 12.05 202.0 6.85 12.39 1.38 5.16 0.99 0.16 0.94 0.13 0.80 0.14 0.39 0.35 0.05 0.28 0.21 6.64 2.93 0.85

436.0 1.28 200.3 537.5 12.56 437.9 2.74 7.12 78.88 1292.8 4.20 17.84 2.58 12.53 180.6 6.47 12.16 1.33 4.82 0.94 0.15 0.88 0.12 0.79 0.14 0.39 0.37 0.05 0.40 0.22 5.78 2.83 0.94

480.9 1.45 229.0 483.9 13.58 504.5 2.52 7.87 83.33 1297.5 4.06 11.11 2.37 13.80 190.4 6.51 12.28 1.39 4.99 0.94 0.16 0.87 0.13 0.78 0.14 0.38 0.36 0.05 0.28 0.19 5.90 2.82 0.87

5.2. Whole rocks and silicate glass geochemistry 5.2.1. Whole-rocks Whole rock chemical analyses were performed on some Gergovie peperites and the Oligocene limestone (G0) sampled just below the Gergovie volcanic deposits (Table 8). As expected from the whole rock XRD analyses, the sedimentary sample G0 is mainly calcitic and contains low amounts of MgO (4.2 wt.%) and more than 16 wt.% SiO2, in agreement with the presence of some quartz and phyllites identified from XRD analyses. This limestone contains around 1400 ppm Sr. The peperites from Gergovie have very different chemical composition. Their SiO2 content ranges from 10 to 23 wt.% while MgO ranges from 12 to 16 wt.%. Their Sr content is similar to the G0 limestone, ranging from 757 ppm in G1 up to 1658 ppm in G11. The weight % dolomite contents, calculated from CaO and MgO content and taking into account the matrix composition is in agreement with the loss of ignition (LOI) measured in these samples. It is about 50–55% in G1, 60–70% in G7 and about 70–80% in the dykes G4 and G11. The residual compositions after removal of the dolomitic component from Mardoux peperites are characterised by a very high Na/K ratio (about 32), that corresponds to the abundance of analcime. The peperites from dykes G4 and G11 contain 6 to 10 times more K and 4 times less Na, so the Na/K ratio is low (0.5 to 1.5); this composition corresponds to a potassic matrix (about 20% of the rocks) containing K-feldspar, micas and a K–Mg-rich zeolite (offretite type, Table 4), probably corresponding to the K–Al–Si interstitial residua reported in dolomitic carbonatites from Zambia, France and Spain (Bailey and Kearns, 2011). 5.2.2. Silicate glass Fresh silicate glass from Pileyre pyroclasts is melanephelinite (IUGS classification in Le Bas et al., 1986; Le Maitre et al., 2002)

G0 Sediments G7 crystals G7 matrix crystals

10

1

0.1

La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu

100

Sample/Chondrite

G0

Analysis no.

10

1 G0 Sediments G1 G11

0.1

La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu

1000 Pileyre glasses

Sample/Chondrite

Sample

Sample/Chondrite

Table 5 Trace element composition (ppm) of Gergovie sediments.

100

G0 Sediments PIL5 PIL8 globules PIL8 matrix

10 1 0.1 0.01

La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu

Fig. 4. Rare earth element compositions of the carbonates, glasses and sediments obtained by LA-ICP-MS from the Gergovie and Pileyre peperites (chondrite from Anders and Grevesse, 1989).

(Table 9). In Gergovie peperites, the glass is more altered and good chemical analyses were only obtained in pyroclasts from fine grained layers in sample G14. In this sample the glass has a phonotephritic composition, similar to the Puy Mardoux lava (Cheguer, 1996). In G7 sample from Puy Mardoux, the glass has a high LOI and is highly enriched in Na2O, approaching an analcime composition. Microspheres of analcime observed in the dolomitic matrix probably correspond to the alteration of silicate glass. In G1 peperite, the glass is also highly altered and shows high LOI values, but has very low Na2O content. Its composition corresponds, in some places, to smectite (saponite). In G1 as in the dykes G4 and G11, the glass is mainly altered to a zeolite, offretite type, displaying K and Mg as dominant extra framework cations (Table 4). This type of zeolite is often described as an alteration product of alkaline silicate glass (Weisenberger and Spurgin, 2009). Glasses have been analysed for trace elements in PIL5 and PIL8 (Table 10). They are similar in the two samples, with REE patterns parallel to the sediments, but with higher REE content and no negative Eu anomaly (Fig. 4). They have high Nb, Ba, Ta, Th and U content but low Cs content compared to the sediments.

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

5.3. Isotopic data Sr isotopic data have been obtained on whole rocks for several samples as well as on separated minerals from G11 peperite (Table 11). The G0 limestone has an 87Sr/86Sr ratio of 0.71307. This value is in the low range of isotopic values published for sediments from the Limagne area with values ranging from 0.711 up to 0.720 (Briot and Poidevin, 1998; Briot et al., 2001). Whole rock peperites from Gergovie (4 samples) as well as from Pileyre (2 samples) have Sr isotopic values ranging from 0.70717 up to 0.70966. Separated dolomite grains from sample G11 have Sr isotopic ratio of 0.70960, very close to the value measured for the whole rock (0.70966), thus confirming that the Sr from the rock is dominated by the dolomitic content. C and O isotopic compositions have also been obtained on separated dolomite grains from G11. δ13C value is −2‰. This value plots in the primary carbonatite box and is very similar to the values that have been obtained in dolomite carbonatite from the Velay volcanic province in the French Massif Central (Chazot et al., 2003). δ18O value is +27‰ and plots to the right of the carbonatite box. Such high oxygen isotopic compositions are often ascribed to low temperature alteration and have also been observed in the Velay carbonatites (Chazot et al., 2003) as well as in other extrusive carbonatites in Europe (Hubberten et al., 1988; Riley et al., 1999; Rosatelli et al., 2000). 6. Discussion Carbonates mixed with lava fragments in the peperites from the Limagne graben have been interpreted as pieces of local sediments fragmented and incorporated into the silicate magma during phreatomagmatic explosions close to the surface. Here we present evidences for the coexistence of two liquids, a silicate one and a carbonate one emplaced all together during explosive eruptions, as well as evidences for a magmatic origin of the carbonates observed in the peperites. 6.1. Incorporation of local sediments In the Limagne graben, the sedimentary limestones are mainly calcitic, and only minor amounts of dolomitic sediments have been reported by Briot et al. (2001) and Wattinne (2004). In agreement with these observations, the G0 sample just below Gergovie peperites is calcitic as well as other sediments outcropping nearby Gergovie. On the opposite, the most representative and major carbonate phase in peperites is dolomite. In Gergovie, except some secondary calcitic infilling of vesicles in lava fragments from Mardoux volcano (G7), dolomite is the only carbonate phase detected. In the Pileyre volcano, dolomite is also the main carbonate phase as globules or crystals inside the silicate glass pyroclasts or as subangular clasts. The dominant dolomitic composition of the carbonates in the peperites has already been observed in the Puy de Crouel peperites (Bailey et al., 2006) and this observation is also valid for many other peperites in the area. Major element compositions confirm these observations in showing that the Gergovie peperites clearly represent a mixing between silicate magma and a dolomitic component. In the diagrams involving CaO, MgO and SiO2 whole rock compositions (Fig. 6), the Gergovie peperites form a linear array along a line joining the Limagne basalts and the dolomitic carbonates analysed in the peperites. No calcitic component is involved in these rocks. The chemical composition of the peperites only represent a mixing between a basaltic magma and carbonates of dolomitic composition, ruling out a mixing with the local sediments. Even if some blocks or small fragments of sedimentary (calcitic) rocks are sometimes observed and easily identified in the peperites, these peperites are clearly not a mechanical mixture between silicate magma and the most abundant sediments in the Limagne graben. Secondary dolomitisation process could have been at the origin of the dolomite in these samples, but in the Pileyre samples dolomite

139

clasts and globules within silicate glass are often sealed by a calcite or magnesite matrix, indicating that the dolomite was the first carbonate to crystallise. Sr isotopic compositions also argue against a mixing between sediments and silicate magma. Whole rock peperites have high 87Sr/86Sr isotopic ratios. A simple mixing model between the local sediment G0 and high Sr basalt from Limagne (Cheguer, 1996) fails to reproduce the Gergovie peperite compositions (Fig. 7). For this calculation, we took a basalt composition with a high Sr content (1200 ppm), and the G0 sediment which has a Sr isotopic composition on the low side of the Limagne sediments. To correctly fit the peperite data, the sediments must have a very high Sr content, around 2300 ppm, which is higher than any measured composition in the Limagne basin. Even with this extreme Sr composition, the mixing model only fits the data by involving 50 to 75% of basaltic component. This is very different to what is observed in the peperites in which the basaltic part is always very low, especially in the dykes G4 and G11 where the amount of dolomite has been calculated to be around 70% from the whole rock major element compositions. These calculations rule out the possibility for the peperites to result from a simple mixing between basaltic magma and the sediments from the Limagne graben. It is to be noted that in sample G11, separated dolomite grains have the same Sr isotopic composition as the whole rock and the other peperites, confirming that the measured values reflect the isotopic composition of the carbonates, and that the whole rock Sr isotopic composition is dominated by the composition of these carbonates. This whole rock isotopic composition is clearly different from the composition of the carbonate sediments analysed in the Limagne. It is to be noted that when it was observed (Wenzell et al., 2001, 2002), assimilation of dolomitic sediments by mafic magma produces a calcitic liquid while we observe dolomitic carbonates in the peperites. 6.2. Dolomite associated with Cr-spinels and diopside Cr-spinels are relatively rare in intrusives carbonatites, but they have been frequently cited as xenocrysts in extrusive carbonatites of small volcanic edifices from various areas in Germany, Italy, Morocco and Zambia (review in Woolley and Church, 2005). More specifically, the association of Cr-spinels with dolomitic extrusive carbonatite has been described in Zambia (Bailey and Kearns, 2011). Mantle xenoliths from the melilitic tuffs from Mt. Vulture contain Cr-spinels with Cr2O3 ranging from 18 to 40% (Jones et al., 2000; Stoppa et al., 2009). Cr-spinels are also present in the Crouel peperites (Bailey et al., 2006). In Pileyre and Gergovie samples, Cr-spinels occur in dolomite clasts or in close association with dolomite and Cr-diopside. In the dolomite clasts, they are frequently euhedral. They have intermediate Al2O3 content between mantle spinels and oxides from the Limagne basalts and very high Cr2O3 content. Interestingly, they have higher Ca, Ti and Mn content than Cr-spinel from mantle peridotites and their chemical composition is very similar to spinels analysed in carbonatites from Tamazert (Mourtada et al., 1997), Rufunsa (Bailey, 1989) and also from the Crouel peperites (Bailey et al., 2006). These Cr-rich spinel compositions are also found in euhedral spinels formed in melt pockets from mantle lherzolites. These melt pockets are frequent in lherzolites metasomatised by carbonatitic melts (Chazot et al., 1996) and have been described in mantle xenoliths from Massif Central (Jakni et al., 1996; Xu et al., 1998). So the chemical composition of the spinels in the peperites and their close association with Cr-diopside and dolomite are strong arguments for a magmatic origin of these minerals. The clasts or the associations with Cr-diopside and dolomite we observed in the peperite samples are always surrounded by a dolomitic carbonate matrix, implying that these Cr-spinels associations crystallised early, maybe at mantle depth, and have been carried by the carbonatite magma up to the surface. It is to be noted that in the hypothesis of assimilation of limestones by basaltic magmas, the spinels are specifically Cr-poor spinels (Wenzell et al., 2001, 2002).

140

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

Table 6 Trace element composition (ppm) of Gergovie carbonates. Sample

G1

Analysis no.

G1-1

G1-2

G1-3

G1-4

G1-5

G1-4.2

G1-6.2

G1-7.2

G1-8.2

G7 G7-1

G7-2

G7-3

Li Be B Ti Cr Mn Co Ni Rb Sr Y Zr Nb Cs Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Hf Ta Pb Th U

37.54 2.95 14.56 10.21 12.11 926.0 1.90 1.22 – 1474.0 7.56 27.04 0.04 – 42.84 13.03 17.31 2.05 8.06 1.68 0.47 1.69 0.26 1.62 0.31 0.87 0.93 0.14 0.048 – 0.05 0.55 0.39

38.90 3.13 14.74 18.24 11.80 911.9 1.83 0.86 0.00 1424.0 7.39 27.57 0.09 0.013 39.54 12.59 16.51 1.98 7.88 1.58 0.44 1.59 0.24 1.52 0.29 0.83 0.91 0.13 0.047 – 0.05 0.45 0.35

34.08 3.58 16.53 – 8.96 824.4 2.27 0.40 – 1473.9 7.66 25.82 – – 37.68 14.16 19.65 2.33 8.98 1.81 0.51 1.72 0.26 1.57 0.32 0.79 0.89 0.13 0.062 – – 0.32 0.38

26.76 2.93 14.67 – 5.00 1341.9 1.89

28.60 2.90 13.35 – 19.93 1171.6 2.31 0.29 – 1702.4 8.30 12.07 – – 54.75 16.13 23.36 2.77 10.47 2.14 0.63 1.86 0.28 1.81 0.35 1.03 1.07 0.16 – – – 0.93 0.62

35.50 1.47 3.89 7.56 9.60 1060.9 2.82 25.40 0.01 1054.4 6.15 26.07 0.05 0.000 27.20 9.10 11.08 1.19 4.53 0.79 0.21 0.83 0.12 0.79 0.16 0.48 0.52 0.07 0.030 0.001 0.87 0.05 0.15

28.99 1.53 3.11 7.15 9.18 961.8 0.15 0.66 0.02 980.7 6.50 25.15 0.04 0.002 26.87 10.72 14.29 1.55 5.90 1.10 0.27 1.06 0.14 0.92 0.17 0.52 0.50 0.07 0.028 0.001 0.15 0.13 0.17

33.28 3.23 2.86 10.86 23.10 943.2 0.11 0.54 0.09 1171.9 5.05 29.76 0.05 0.005 36.33 8.09 10.54 1.21 4.51 0.78 0.27 0.91 0.11 0.76 0.15 0.40 0.42 0.07 0.037 0.002 0.10 0.10 0.14

37.35 0.97 3.92 8.40 9.21 1155.5 0.02 0.13 0.00 1130.1 3.90 23.79 0.02 0.001 26.03 4.77 6.26 0.66 2.45 0.43 0.12 0.48 0.06 0.44 0.09 0.29 0.31 0.05 0.021 0.000 0.18 0.03 0.09

54.30 0.66 15.65 575.4 26.20 369.1 3.37 4.97 0.06 1658.2 4.39 11.93 1.16 0.001 20.23 5.28 10.11 1.27 5.25 1.16 0.26 1.07 0.15 0.88 0.15 0.38 0.34 0.05 0.188 0.165 0.16 3.13 0.19

50.52 0.54 12.88 309.4 19.54 365.4 4.80 8.32 0.02 1872.6 5.32 6.75 0.80 – 28.08 9.09 16.53 1.93 7.57 1.52 0.34 1.32 0.18 1.03 0.18 0.44 0.33 0.05 0.078 0.068 0.17 5.20 0.21

44.84 0.60 11.52 164.2 7.35 423.8 1.52 1.42 0.07 1442.1 5.14 5.42 0.38 0.004 16.77 11.57 18.51 2.03 7.77 1.45 0.35 1.28 0.16 0.89 0.15 0.37 0.25 0.03 0.067 0.039 0.13 3.82 0.37

– 1900.3 8.68 8.29 – – 64.22 16.04 24.19 2.75 10.24 2.03 0.67 1.97 0.30 1.99 0.39 1.05 1.13 0.17 – – – 1.25 0.59

All the analysed carbonates are dolomites. G7 G7-4

G11 G7-11

G7-13

G7-15

G7-17

G7-18

G7-19

6.3. Euhedral dolomite crystals in dykes The fine-grained dyke G4 crosscutting the layered peperites and the similar sample G11 are mainly constituted by anhedral to euhedral dolomite crystals with some zoning which is different from the zoning appearing in dolomitisation process when foreign matter is incorporated as microinclusions. Dolomite crystals are embayed in a K-rich amorphous silicate matrix including dolomite, K-feldspar and phyllite microcrystals. This textural relation is very similar to those displayed by the fine-grained intrusive subvolcanic alvikites (calcite) or beforsites (dolomite) carbonatites described in the Wasaki–Ruri complex from western Kenya (Le Bas, 1977) and in the Sinaï Peninsula (Shimron, 1975), respectively. From their geological setting, their chemical and mineral compositions and textures, these dykes may be interpreted as magmatic and emplaced simultaneously or soon after the layered peperites. 6.4. Matrix — cement carbonates In Gergovie area, the carbonate matrix in the peperites is always a microcrystalline dolomite. In Pileyre peperites, while the carbonate mineral within the silicate glass is dolomite, other carbonate phases occur in the matrix and may be related to a rapid and specific crystallisation stage. 1) In the peperite PIL5, microcystallised magnesite carbonate Mg1,7 Fe0,2 Ca0,1 (CO3) 2 is the main constituent between the dolomite bearing pyroclasts, and appears as bothryoid clusters or as a matrix matter where it is mixed with amorphous silica. Magnesite carbonate crystallisation does not result from any usual

G11-4

G11-10

G11-11

G11-12

G11-13

G11-15

sedimentary process, so it cannot be interpreted as a sedimentary remnant. On the contrary, magnesite with Cpx is a well known stable product of the decarbonation of dolomitic liquids reacting with Opx in mantle conditions (>21 kbars) or, as a classical product of the carbonation of basic material in relation with mantle fluids (Hansen et al., 2005; Lutkov et al., 2007). In Pileyre, the occurrence of dolomite clasts, the reaction associations Opx/Cpx/ carbonate in glassy pyroclasts and the presence of magnesite can be accounted for by the occurrence of a dolomitic primary liquid and its interaction through decarbonation reaction with the basic material. 2) In the layered peperite PIL8, the same dolomite bearing glass pyroclasts but with smooth shapes are enclosed in a white matrix made of large calcite crystals without any inclusions. This matrix displays no cementation nor recrystallisation evidences, as banded calcite crystals around the pyroclasts (Barker, 2007) nor dedolomitisation marks as iron oxides microinclusions issued from some Fe element of a supposed primitive dolomite. The major part of the globules and carbonate crystals within the pyroclasts are dolomite but in some places they are constituted by a calcite–dolomite association without any distinctive limit between the two species, as if they were synchronous or pseudomorphosed. Two hypotheses may explain this association: a) co-crystallisation of the two carbonates from a carbonatitic dolomitic liquid in response to changes in temperature and pressure as the magma moves through the crust and solidifies (Gittins et al., 2005); b) hydrothermal calcitic cementation invading some dolomite crystals in pyroclasts. In Gergovie as in Pileyre area, the effects of hydrothermal fluids or melt–fuel–coolant interactions (Hooten and Ort, 2002) may be

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

141

Table 6 (continued) G7

G11

G7-4

G7-11

G7-13

G7-15

G7-17

G7-18

G7-19

G11-4

G11-10

G11-11

G11-12

G11-13

G11-15

55.36 0.94 12.15 356.2 24.04 573.2 4.72 7.04 0.13 2317.5 9.78 5.20 1.22 – 99.44 16.06 30.06 3.33 12.57 2.40 0.53 2.08 0.29 1.83 0.34 0.91 0.82 0.12 0.063 0.060 0.15 4.96 0.25

33.04 0.61 13.25 256.3 539.87 442.2 6.97 47.42 1.80 1137.0 1.54 8.31 0.59 0.110 57.32 3.14 4.66 0.58 2.16 0.42 0.11 0.38 0.05 0.29 0.05 0.13 0.10 0.02 0.144 0.062 1.06 0.18 0.09

58.40 0.69 11.19 134.0 15.31 341.7 0.76 1.24 0.03 1703.4 1.98 3.65 0.32 – 21.23 3.38 7.29 0.81 3.38 0.65 0.18 0.62 0.08 0.47 0.07 0.16 0.12 0.01 0.021 0.023 0.37 1.59 0.14

23.26 −0.05 11.87 42.2 41.56 329.3 0.37 1.17 0.09 1948.7 0.58 1.69 0.08 – 17.21 1.45 2.93 0.43 1.79 0.44 0.09 0.32 0.04 0.15 0.02 0.04 0.02 0.00 0.017 0.005 0.17 0.93 0.09

42.63 −0.37 11.61 101.2 9.71 384.7 34.03 59.92 0.60 1776.0 1.46 3.05 0.24 0.072 29.42 2.55 5.36 0.66 2.56 0.58 0.15 0.49 0.06 0.36 0.06 0.12 0.13 0.01 0.028 0.008 1.22 0.81 0.11

54.46 0.54 10.60 471.8 13.50 355.3 2.09 4.69 0.28 1723.9 5.79 7.84 0.78 0.010 22.03 11.36 19.08 2.16 8.26 1.70 0.38 1.48 0.20 1.10 0.19 0.46 0.33 0.04 0.119 0.099 0.49 4.75 0.33

37.42 0.56 11.76 828.6 256.54 559.3 3.53 15.53 1.28 1214.3 2.95 7.60 4.67 0.124 34.90 3.29 7.63 1.08 4.34 0.91 0.22 0.69 0.10 0.56 0.09 0.24 0.20 0.02 0.110 0.260 0.47 0.66 0.13

34.41 0.47 20.90 207.6 0.19 341.2 2.31 3.34 14.34 1291.1 3.85 14.58 0.80 1.311 76.58 4.95 8.88 1.07 4.19 0.92 0.17 0.80 0.12 0.71 0.13 0.34 0.26 0.04 0.178 0.063 20.16 1.32 0.29

42.53 1.13 15.28 141.7 2.71 930.7 1.05 0.41 9.29 1580.8 12.61 10.29 0.63 1.150 246.5 21.91 36.17 4.52 16.34 3.39 0.67 2.84 0.42 2.48 0.46 1.18 1.02 0.14 0.129 0.058 1.57 7.95 0.87

37.95 0.97 14.46 157.4 1.52 901.6 2.60 1.21 6.57 1605.1 11.85 8.71 0.56 0.764 224.0 19.95 33.23 4.18 15.51 3.19 0.64 2.61 0.39 2.37 0.44 1.15 0.96 0.14 0.098 0.054 1.81 7.23 0.84

78.65 1.55 26.79 670.3 19.75 344.9 5.18 11.62 64.63 1236.3 3.45 59.79 3.03 7.715 89.08 5.85 10.19 1.32 5.00 1.02 0.22 0.90 0.13 0.75 0.14 0.36 0.31 0.04 0.947 0.291 135.36 2.30 0.55

97.92 1.86 28.55 1096.2 25.99 326.4 4.12 8.48 87.50 1108.1 2.51 71.15 5.52 11.117 84.20 3.99 7.33 0.94 3.61 0.75 0.13 0.58 0.08 0.53 0.10 0.29 0.29 0.04 1.290 0.532 12.93 1.85 0.61

43.11 0.75 16.84 234.2 5.12 522.2 2.21 1.60 19.26 1319.0 6.97 19.32 1.06 2.385 74.23 9.81 16.69 2.19 8.41 1.79 0.37 1.56 0.23 1.46 0.27 0.72 0.59 0.09 0.305 0.105 3.50 3.22 0.64

superimposed. Such effects have been inferred for the carbonatitic formations of Veseloe (Doroshkevitch et al., 2007) in Brazil. In Gergovie, these effects may explain the massive alteration of the silicate glass and the abundance of analcime and, in Pileyre (PIL8), the large-grain calcitic matrix, embaying the dolomite bearing pyroclasts. 6.5. Trace elements in carbonates, silicate glass and sediments Trace element compositions are not very discriminating between carbonates, silicate glass and local sediments. REE patterns are all parallel, confirming La/Yb ratios very similar in these different phases. However, some elements show large differences between the carbonates and both the glass and the sediments. In the Nb vs. Rb diagram (Fig. 8), the carbonates from Gergovie and Pileyre plot in the lower left of the diagram with very low Nb and Rb concentrations. The G0 sediment has very high Rb concentration, and also high Nb content. Silicate glasses from PIL5 and PIL8 also have high Rb content (more than 55 ppm) and very high Nb content compared to the sediment and the carbonates. Only a few analyses from G11 are displaced towards the sediment composition, maybe reflecting some interaction with the local sediments. The same remarks is also valid with Li, with very high concentrations in the local sediment, while the peperite carbonates have very low Li content and only a few analyses from G11 plot towards the silicate glass composition. In any case, the carbonates in the peperites plot on a mixing trend between the glasses and the sediments. In detail, it is sometimes possible to distinguish between different types of carbonates in a single sample on the basis of their trace element compositions. In G7, large dolomite crystals associated with spinels have low La, Y and Th content. Carbonates from the matrix have

higher concentrations of these elements, similar to the composition of the G0 sediment or even higher (Fig. 4). Similarly in PIL8, carbonates in globules have higher Co and Ni content, in agreement with a magmatic origin, compared to the carbonates analysed from the calcitic matrix which have compositions very close to G0 sediment. In their study of the Oricola carbonatites, Stoppa et al. (2005) have already emphasised the similarities between the composition of carbonates and the associated silicate glass, showing parallel REE patterns but with higher REE concentrations in the carbonates. In Pileyre, the REE patterns of the different carbonates are also parallel to those of the silicate glasses, but in these samples, glasses have higher REE content than the carbonates. More surprising is the similarity of REE patterns between the silicate glasses and the G0 sediment, with higher REE content in the glasses. They have also similar Th and U concentrations, but very different Zr, Nb and Ta contents. These observations confirm that it is difficult to use trace element compositions to decipher between magmatic or sedimentary origin of carbonates, except for some elements such as Rb, Nb or Li which confirm that in our case the analysed carbonates in the peperites are not of sedimentary origin. 6.6. Origin of the carbonates All the observations presented above argue against a sedimentary origin for the carbonates found in the Gergovie and Pileyre peperites. They are dolomitic, they have different major and partly trace element compositions, they contain minerals (diopside and spinel) not found in sedimentary carbonates, they have low δ13C value and different Sr isotopic composition. We interpret these carbonates as evidence for the eruption of carbonatitic magmas directly associated with the basaltic magmas found in Gergovie and Pileyre. In both localities, the two types of magmas

142

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

Table 7 Trace element composition of Pileyre carbonates. Analysis no.

PIL5-1

PIL8-10

PIL8-12

PIL8-13

PIL8-15

PIL8-16

PIL8-17

PIL8-18

PIL8-20

Location

E.D.

Globule

Matrix

Matrix

Globule

Globule

Globule

Globule

Matrix

Li Be B Ti Cr Mn Co Ni Rb Sr Y Zr Nb Cs Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Hf Ta Pb Th U

47.00 0.00 20.05 86.92 214.9 138.3 63.11 732.12 1.06 2571.0 0.34 0.69 0.16 0.28 20.74 0.38 0.67 0.08 0.33 0.07 0.01 0.07 0.01 0.08 0.01 0.04 0.04 0.01 0.02 0.02 0.06 0.02 0.01

3.25 84.67 5.88 0.21 0.003 0.37 0.01 0.00 – 43.45 0.38 3.55 0.07 – 1.95 3.45 2.56 2.23 1.90 1.22 0.99 0.74 0.59 0.51 0.41 0.34 0.32 0.34 1.26 0.01 0.03 1.45 1.91

– 14.94 7.11 0.07 0.002 0.39 0.00 0.00 0.19 10.46 0.95 5.39 0.39 0.03 1.57 5.55 3.69 3.31 2.86 1.70 1.45 1.36 1.07 0.99 0.90 0.85 0.80 0.71 2.07 0.64 0.11 3.48 18.13

– 50.43 5.23 0.00 0.000 0.37 0.00 0.00 – 22.84 1.16 5.70 0.00 – 0.05 7.06 5.02 4.58 3.95 2.44 2.23 1.96 1.42 1.20 0.92 0.75 0.51 0.49 1.14 0.08 0.03 0.08 2.18

4.37 50.81 6.95 0.16 0.003 0.36 0.01 0.00 0.00 47.92 0.38 2.85 0.14 – 2.29 2.87 2.28 1.99 1.62 0.88 0.85 0.62 0.48 0.42 0.34 0.32 0.25 0.29 0.57 0.01 0.01 1.64 3.31

4.69 21.40 7.64 0.38 0.015 0.37 0.00 0.00 0.07 47.24 0.24 3.05 1.59 0.07 2.51 2.54 2.08 1.66 1.44 0.95 0.87 0.55 0.33 0.30 0.25 0.28 0.20 0.20 0.67 1.43 0.01 1.92 2.36

5.02 27.50 9.86 0.23 0.001 0.34 0.00 0.00 0.01 46.06 0.23 3.01 0.53 – 2.46 2.18 1.70 1.42 1.20 0.75 0.58 0.44 0.33 0.36 0.22 0.20 0.14 0.27 0.63 0.46 0.01 1.65 1.16

3.19 220.8 6.94 0.13 0.001 0.32 0.01 0.00 0.00 27.56 0.26 5.38 0.25 – 1.40 2.15 1.48 1.24 1.03 0.66 0.56 0.40 0.31 0.31 0.26 0.22 0.28 0.22 1.40 0.19 0.02 1.74 3.73

– 17.76 6.54 0.03 0.003 0.34 0.00 0.00 – 8.09 1.06 13.54 0.01 – 0.11 6.05 4.21 3.62 3.26 2.06 1.70 1.48 1.11 1.11 0.95 0.99 0.83 0.84 7.06 0.00 0.01 2.29 18.07

E.D. = euhedral dolomite, globule = dolomite, matrix = calcite.

are contemporaneous and were emplaced during the same eruption. Association between carbonate and silicate magmas is often interpreted as immiscibility from a single silicate mantle-derived magma. However, the high Sr isotopic composition of the carbonates indicates that at least for Gergovie, this magma is not in equilibrium with the silicate one. The implication is that the carbonate melts cannot originate by an immiscibility process from the silicate magmas because this process is unable to

1.0

Rufunsa Puy de Crouel Pileyre spinels Gergovie spinels

Cr/(Cr+Al)

0.8 Mantle spinels

0.6

Table 8 Whole rock major element composition of sediments and peperites from Gergovie.

0.4 0.2

Limagne basalts Massif central xenoliths

0.0 0.0

modify the Sr isotopic composition. Carbonate globules are also frequently presented as a strong argument for liquid immiscibility, but Valentini et al. (2010) have shown that carbonate globules can form by instantaneous fragmentation of carbonatite viscous fingers subsequent to the injection in a magma chamber containing silicate magma. Furthermore, recent experimental work aimed at reassessing the silicate–carbonate two-liquid solvus has proposed that carbonate globules in silicate glass may represent a solid phase precipitated in the magma instead of drops of immiscible carbonate magma (Brooker and Kjarsgaard, 2011). For Gergovie, no carbonate globules are observed, and Sr isotopic disequilibrium between silicate magma and carbonates precludes an immiscible phenomenon to explain the contemporaneous eruption of the two magmas. In Pileyre, no isotopic data are available and the carbonate globules observed in some samples may well illustrate the incapacity of

0.2

0.4

0.6

0.8

1.0

1.2

Fe2+/(Fe2++Mg) Fig. 5. Plot of Cr/(Cr + Al) vs. Fe2+/(Fe2+ + Mg) for spinels analysed in Gergovie and Pileyre peperites. They are compared to mantle spinels (Barnes and Roeder, 2001 and Matsukage and Kubo, 2003) and spinels from Massif Central mantle xenoliths (Chazot, unpublished data) and also to spinels from the Limagne basalts (Cheguer, 1996). Grey and black squares show the composition of spinels from Rufunsa carbonatites (Bailey, 1989) and Puy de Crouel peperites (Bailey et al., 2006).

Sample

G0

G1

G4

G7

G7b

G11

SiO2 Al2O3 Fe2O3 MgO CaO Na2O K2O TiO2 MnO P2O5 Ba % Sr % LOI

16.04 3.38 1.18 4.24 39.00 0.37 1.02 0.12 0.06 0.06 0.02 0.15 34.36

23.49 7.11 4.39 12.66 18.80 2.01 1.01 0.91 0.10 0.20 0.01 0.08 29.23

10.82 3.42 2.21 15.75 26.58 0.95 0.92 0.29 0.07 0.06 0.02 0.14 38.77

19.91 7.32 3.87 12.28 21.55 3.26 0.15 0.74 0.07 0.20 0.04 0.15 30.46

16.52 6.19 3.94 13.52 22.39 2.72 0.13 0.69 0.08 0.14 0.02 0.17 33.49

13.41 4.59 2.23 15.10 25.39 0.58 1.75 0.21 0.06 0.03 0.01 0.17 36.48

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146 Table 9 Representative analyses of glass from Gergovie and Pileyre peperites. Sample

PIL5

PIL5

PIL5

PIL8

PIL8

Analysis 175-Z4 no.

G14

26-rd1

31-rd1

77-rd5

151-z1

155-z1

SiO2 Al2O3 TiO2 FeOtot MnO MgO CaO Na2O K2O P2O5 Total IUGS name

43.28 15.32 3.51 8.83 0.12 5.38 13.36 3.93 3.77 nd 97.50 Melaneph.

43.44 15.75 3.31 9.31 0.17 5.27 12.92 4.03 3.01 nd 97.21 Melaneph.

43.46 16.06 3.35 8.92 0.19 5.24 12.69 4.34 3.08 nd 97.32 Nephel.

43.80 16.30 3.36 9.36 0.19 5.09 12.35 4.24 3.30 1.03 99.01 Melaneph.

43.46 16.26 3.37 9.62 0.21 5.03 12.21 4.29 3.43 1.05 98.92 Melaneph.

48.92 18.64 2.78 8.38 0.16 3.11 7.21 5.01 3.91 0.87 98.99 Phonotephr.

Phonotephr. = phonotephrite, Melaneph. = melanephelinite, Neph. = nephelinite. Nd = not determined.

the two magmas to mix instead of an origin by immiscibility. As in many other examples, more observation and data are needed to fully understand the origin of this carbonate–silicate association, but in the present case we prefer to rely on isotopic data to discard immiscible processes. Many experimental works have shown that dolomitic melts can be produced by partial melting of a carbonated lherzolite at great depths which may correspond to the upper asthenosphere or the deeper part of the lithospheric mantle (e.g. Dalton and Wood, 1993; Gudfinnsson and Presnall, 2005; Wallace and Green, 1988). Such melts will be eliminated during ascent by reaction with the surrounding mantle as they

Table 10 Trace element composition (ppm) of Pileyre glasses. Sample

PIL5

Analysis no.

PIL5-8

PIL5-9

PIL5-10

PIL8 PIL8-5

Li Be B Ti Cr Mn Co Ni Rb Sr Y Zr Nb Cs Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Hf Ta Pb Th U

56.63 1.39 19.38 10,732 1060.3 1247.2 54.83 282.7 59.09 889.1 13.77 155.1 53.71 3.57 421.8 33.87 61.74 6.71 25.28 4.61 1.36 3.73 0.49 2.82 0.49 1.25 1.05 0.15 3.32 3.01 2.28 3.72 1.17

8.93 1.01 12.19 13,028 2647.9 1123.4 27.88 51.1 55.73 679.6 18.41 183.3 79.63 0.78 608.4 49.82 90.39 9.54 35.87 6.26 1.89 5.01 0.66 3.87 0.65 1.64 1.38 0.19 3.18 4.01 2.86 5.48 1.55

8.11 1.49 12.22 15,805 101.4 1201.0 30.11 38.1 68.65 836.3 23.86 226.4 98.19 0.99 771.6 63.58 118.6 12.57 46.96 8.35 2.50 6.64 0.88 5.13 0.88 2.25 1.90 0.26 4.05 5.01 3.66 7.20 1.97

60.65 1.58 8.08 11,068 329.6 479.7 40.58 157.5 57.19 282.5 16.28 141.6 53.63 8.21 39.37 41.22 94.51 10.77 41.18 7.95 2.31 5.67 0.77 4.28 0.70 1.69 1.34 0.17 2.81 2.63 1.63 4.81 1.23

143

cross the decarbonation reaction but several authors have argued that a rapid ascent should prevent equilibrium reaction to occur and allow these magmas to reach crustal levels and ultimately the Earth surface (Brooker and Kjarsgaard, 2011; Harmer and Gittins, 1997, 1998). Combining these observations with our data, and in particular the Sr isotopic results, we can argue that at least for Gergovie, the carbonate magma formed from a mantle source different from the source of the associated silicate one, and that the two magmas ascended and erupted simultaneously. This was probably also the case in Pileyre but isotopic arguments are not available at the moment. Sr isotopic values found in the whole rocks or in the separated carbonates are high, but still lower than any sedimentary carbonate analysed in the Limagne area. These values are in fact similar to many exotic rocks emplaced in different parts of Europe since the Oligocene (Fig. 9). Many lamproites from Spain, from the Alps, from Italy or from Central Europe and Turkey have high Sr isotopic compositions, with 87 Sr/86Sr ratios ranging from 0.706 to more than 0.720 (Prelevic et al., 2008; Tommasini et al., 2011). Many ultrapotassic rocks from Italy also have high Sr isotopic compositions, usually ranging from 0.704 to 0.711 (Pecerillo, 2005 and references therein). Carbonatites from the Mt. Vulture Pleistocene volcano have 87Sr/86Sr values ranging from 0.7058 to 0.7062 (D'Orazio et al., 2007; Rosatelli et al., 2007) close to the lower range of values displayed by the Limagne peperites. C and O isotopic compositions measured on dolomite grains from G11 argue also for a magmatic origin for these carbonates. C isotopic value is slightly higher than the values measured in Vulture carbonatites (−2‰ in G11 instead of −4 to −10‰ in Vulture samples, D'Orazio et al., 2007) but are similar to the dolomitic carbonates from the nearby Velay volcanic province (−2.6 to −2.9‰, Chazot et al., 2003) and is at the upper limit of the “magmatic carbonate” box (Sheppard and Dawson, 1973). This observation reinforces the link between the carbonates analysed in the peperites with those emplaced in the French Massif Central, in Italy as well as in the Calatrava province in Spain. From the data presented above, one can propose that the carbonates found in the Gergovie and Pileyre peperites crystallised from a carbonatitic melt formed at mantle depth, along with silicate melanephelinitic magma. Their different Sr isotopic composition implies that these magmas are not cogenetic but formed from different mantle sources. The source of the carbonate magmas has obviously and “exotic” composition and probably a long history of enrichment through metasomatic events. Such events can be related to ancient subductions during Hercynian orogeny or to Tertiary and Quaternary magmatic activity in the Massif Central. Lamprophyric rocks emplaced from 325 to 310 Ma ago in the Massif Central have initial Sr isotopic values ranging from 0.706 to 0.712 (Fig. 9, Chazot et al., unpublished data). Even if crustal contamination probably plays a role in the genesis of these rocks, their high Sr isotopic compositions as well as their peculiar major and trace element composition imply the formation, during Palaeozoic times, of an enriched lithospheric mantle related to fluid percolation associated with subductions. Parts of this mantle may have survived since the end of the orogenic events. Silicate melt circulation in the lithospheric mantle can then trigger melting of these enriched lithologies to form carbonatitic melts, thus explaining the systematic association of silicate and carbonate magmas in the peperites localities. An alternative hypothesis has been recently put forward to explain the genesis of undersaturated rocks and associated carbonatites in Italy and more generally in Western Europe (Bell et al., 2005; Lavecchia et al., 2006). Mantle metasomatism is still the central mechanism to form an enriched mantle from which carbonate and silicate magmas can be produced. In this new model, a large plume head is trapped in the mantle transition zone and releases H2O–CO2-rich fluids which can percolate the above asthenosphere and modify its chemical and isotopic composition. This model is an alternative explanation for the fragmentation of the European and African continents and the genesis of carbonated and potassic silicate magmas in Europe, and is also consistent with carbonatitic

144

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

Table 11 Strontium isotopique composition of Gergovie sediments and peperites. Sample

87

Sr/86Sr

G0

G4

G7a

G7b

G11

G11

G11

PIL4

PIL8

WR

WR

WR

WR

Dol.

Dol.

WR

WR

WR

0.713065 ± 19

0.707317 ± 13

0.707979 ± 63

0.708495 ± 16

0.709600 ± 13

0.709587 ± 44

0.709663 ± 12

0.707173 ± 276

0.708379 ± 15

WR = whole rock, Dol. = dolomite crystals.

metasomatism observed in mantle xenoliths from Northern Africa (Raffone et al., 2009 and references therein). Obviously our data do not provide much information about these deep mantle processes but confirm the presence of a CO2-rich mantle with enriched isotopic composition beneath a large part of the Western Europe and thus represent a step towards understanding the origin of the magmatism in this part of the continent. 7. Conclusions

by a simple mixing between basaltic magma and the Limagne sediments, as it appears also from the major elements. – Dolomite appears under different textural occurrences but is sometimes closely associated with Cr-spinel and diopside, minerals which are absent from the Limagne sedimentary limestones. – Petrography and geochemistry provide evidences for a magmatic origin of the carbonates, and for the eruption of carbonate magma contemporaneously with silicate magma in both localities. The Sr isotopic composition of the carbonates is similar to many volcanic

Petrographic and geochemical data on the volcanic rocks from Gergovie and Pileyre provide key information about the origin of the carbonates in these rocks and allow us to put forward the following observations:

Gergovie sediment

40

Gergovie sediment

0.712 87Sr/86Sr

– Volcanic rocks from the studied areas contain mostly dolomitic carbonates rarely described in the sedimentary record of Limagne basin. Sr isotopic composition of these rocks is different from the composition of the local sediments and cannot be accounted for

0.716

0.708

0.704 Pileyre and Gergovie carbonates

Limagne basalts

CaO

30

0.700 500

1000

1500

2000

2500

Srppm 20 Pileyre glasses

Limagne volcanic rocks

10 Gergovie glasses

0

0

5

10

15

20

25

Fig. 7. 87Sr/86Sr vs Sr diagram for the whole rock peperites (Open circles), the Gergovie sediments (G0, black square) and the Limagne basalts (Grey field, Cheguer, 1996). Two mixing calculations have been performed, one between the most Sr-rich basalt and the G0 sediment, and the other one between a Limagne basalt and an hypothetic sediment with the same Sr isotopic composition than G0, but with 2000 ppm Sr. The dashes on the mixing lines represent 10% increments.

30

MgO + FeO 60

7

Gergovie glasses

Gergovie sediments

Limagne volcanic rocks

50 Pileyre glasses

30

Nb

SiO2

40

20 Gergovie sediment

10

Range of Pileyre and Gergovie carbonate composition

0

Gergovie carbonates

6

0

5

10

15

20

25

Pileyre carbonates

5

1.5

4

1.0

3

0.5

2

0

Pileyre glasses (Nb from 53 to 92 ppm)

0

2

4

1 30

MgO + FeO

0

0

20

40

60

80

100

Rb Fig. 6. CaO vs. MgO and SiO2 vs. MgO plots for whole rock peperites (Open circles). The black square represents the composition of Oligocene limestone (G0) sampled just below the Gergovie peperites. The grey fields represent the composition of the Limagne basalts (Cheguer, 1996) and the carbonates and glasses analysed by microprobe in the different Gergovie and Pileyre peperites.

Fig. 8. Nb vs. Rb plot comparing the composition of the Gergovie sediments (G0, Black squares) with the peperite carbonates from Gergovie and Pileyre. Glasses from the Pileyre peperites contain around 60 ppm Rb, but very high Nb. The inset shows an enlargement of the lower left part of the diagram.

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

G0 sediment

Limagne sediments Peperites Limagne basalts Massif Central lamprophyres Mediterranean lamproites Campania lavas Roman lavas Tuscany lavas Roman carbonatites Vulture carbonatites

0.700

0.705

0.710

0.715

0.720

0.725

87Sr/86Sr Fig. 9. Sr isotopic composition of the Gergovie and Pileyre peperites (Open circles) compared to the composition of the Limagne sediments (Briot and Poidevin, 1998; Briot et al., 2001), and to the composition of volcanic rocks. Limagne basalts are from Cheguer (1996); Massif Central lamprophyres are from Chazot (unpublished data); Mediterranean lamproites are from Prelevic et al. (2008) an Tommasini et al. (2011); Italian lavas are from Peccerillo (2005 and references therein) and Boari et al. (2009); Roman carbonatites are from Castorina et al. (2000); and Vulture carbonatites are from D'Orazio et al. (2007) and Rosatelli et al. (2007).

rocks erupted in Europe since Hercynian times and we propose that carbonate magmas are formed in the mantle by low degree partial melting of an enriched lithology. This partial melting can be triggered by silicate magma percolation in the mantle, thus explaining the association of silicate and carbonate magmas in the volcanic rocks while not being in isotopic equilibrium. Many silicate magma–carbonate associations in the Limagne graben have been described as magma–sediment mixing during eruption. Our conclusions about the magmatic origin of the carbonates in Gergovie and Pileyre volcanic complexes impose to reevaluate the origin of the Limagne volcanic province and to assess with more details the importance of carbonatitic magmas in the French Massif Central volcanism. Acknowledgements We thank Ken Bailey who initiated the work on Limagne peperites and who, along with Frances Wall, stimulated many fruitful discussions about this volcanism. Dmitri Ionov is thanked for providing stable isotope data. We thank Ken Bailey and Francesco Stoppa for their contructive and stimulating reviews. References Agranier, A., Lee, C.T.A., 2007. Quantifying trace element disequilibria in mantle xenoliths and abyssal peridotites. Earth and Planetary Science Letters 257, 290–298. Anders, E., Grevesse, N., 1989. Abundances of the elements: meteoritic and solar. Geochimica et Cosmochimica Acta 53, 197–214. Andersen, T., 2008. Coexisting silicate and carbonatitic magmas in the Qassiarsuk Complex, Gardar rift, Southwest Greenland. The Canadian Mineralogist 46 (4), 933–950. Bailey, D.K., 1989. Carbonate melt from the mantle in the volcanoes of south-east Zambia. Nature 388, 415–418.

145

Bailey, K., Kearns, S., 2011. Dolomitic volcanism in Zambia: Cr and K signatures and comparisons with other deep mantle melts. In: Beccaluva, L., Bianchini, G., Wilson, M. (Eds.), Volcanism and evolution of the African lithosphere: Geological Society of America Special Paper, 478, pp. 211–222. Bailey, D.K., Garson, M., Kearns, S., Velasco, A.P., 2005. Carbonate volcanism in Calatrava, central Spain: a report on the initial findings. Mineralogical Magazine 69, 907–915. Bailey, K., Kearns, S., Mergoil, J., Mergoil-Daniel, J., Paterson, B., 2006. Extensive dolomitic volcanism through the Limagne Basin, central France: a new form of a carbonatite activity. Mineralogical Magazine 70, 1–6. Barker, D.S., 2007. Origin of cementing calcite in ‘carbonatite’ tuffs. Geological Survey of America Bulletin 35, 371–374. Barnes, S.J., Roeder, P.L., 2001. The range of spinel compositions in terrestrial mafic and ultramafic rocks. Journal of Petrology 42, 2279–2302. Bell, K., Lavecchia, G., Stoppa, F., 2005. Reasoning and beliefs about Italian geodynamics. Bollettino della Societa Geologica Italiana 5, 119–127. Bellon, H., 1971. Datation absolue de laves d'Auvergne par la méthode potassium– argon. PhD. University Orsay Paris‐Sud (France). 68 pp. Bessonneau, D., 1997. Les buttes pépéritiques dans la Limagne de Clermont-Ferrand (France). Memoire Géologue IGAL, No.76. 214 pp. Boari, E., Tommasini, S., Laurenzi, M.A., Conticelli, S., 2009. Transition from ultrapotassic kamafugitic to sub-alkaline magmas: Sr, Nd, and Pb isotope, trace element and 40 Ar–39Ar age data from the Middle Latin Valley volcanic field, Roman magmatic province, Central Italy. Journal of Petrology 50, 1327–1357. Boumedhi, A., 1988. Les clinopyroxènes dans les basaltes alcalins continentaux (Massif Central français). PhD University Clermont-Ferrand (France). 174 pp. Bout, P., Frechen, J., Lippolt, H.-.J., 1966. Datations stratigraphiques et radiochronologiques de quelques coulées basaltiques de Limagne. Revue d'Auvergne 80, 207–231. Briot, D., Poidevin, J.L., 1998. Stratigraphie 87Sr/86Sr de quelques laminites carbonatées du Rupélien supérieur du fossé de Limagne: incursions marines dans le rift du Massif central français? Comptes rendus de l'Académie des Sciences de Paris 326, 479–483. Briot, D., Poidevin, J.L., Hugueney, M., 2001. Apports de l'étude isotopique Sr et Nd des sédiments cénozoïques de Limagne à la compréhension du fonctionnement du rift du Massif central français. Bulletin de la Societe Geologique de France 172, 17–24. Brooker, R.A., Kjarsgaard, B.A., 2011. Silicate–carbonate liquid immiscibility and phase relations in the system SiO2–Na2O–Al2O3–CaO–CO2 at 0.1–2.5 GPa with applications to carbonatite genesis. Journal of Petrology 52, 1281–1305. Buckley, H.A., Woolley, A.R., 1990. Carbonates of the magnesite–sidérite series from four carbonatites complexes. Mineralogical Magazine 54, 413–418. Camus, G., Cantagrel, J.-M., Taupinard, J.M., Vialette, Y., 1969. Mesure par la méthode à l'argon, de l'âge de quelques rochesvolcaniques de Limagne (Massif central français). Comptes rendus de l'Académie des Sciences de Paris 269, 2513–2516. Cantagrel, J.-M., Boivin, P., 1978. Datation K–Ar de quelques basaltes du socle du Massif Central au Nord-Est de Clermont-Ferrand. Réunion annuelle des Sciences de la Terre, Orsay, France, 25–27 avril 1978. Société Géologique de France, Paris, p. 89. Castorina, F., Stoppa, F., Cundari, A., Barbieri, M., 2000. An enriched mantle source for Italy's melilitite–carbonatite association as inferred by its Nd–Sr isotope signature. Mineralogical Magazine 64, 625–639. Chantepie, M., 1990. Le volcanisme basaltique miocène et pliocène dispersé dans la région de Clermont et sur le plateau des Dômes : étude pétrologique et géochronologique. Mémoire DEA, University Clermont-Ferrand (France). 73 pp. Chazot, G., Menzies, M., Harte, B., 1996. Silicate glasses in spinel lherzolites from Yemen: origin and chemical composition. Chemical Geology 134, 159–179. Chazot, G., Bertrand, H., Mergoil, J., Sheppard, S.M.F., 2003. Mingling of immiscible dolomite carbonatite and trachyte in tuffs from Massif Central (France). Journal of Petrology 44, 1917–1936. Cheguer, L., 1996. Les laves miocènes de la Limagne d'Allier et des plateaux bordiers (Massif Central français). PhD University Clermont-Ferrand (France). 201 pp. Dalton, J.A., Wood, B.J., 1993. The composition of primary carbonate melts and their evolution through wallrock reaction in the mantle. Earth and Planetary Letters 119, 511–525. Degeai, J.P., Pastre, J.F., 2008. Evolution morphostructurale du plateau volcano-sédimentaire de Gergovie au Miocène inférieur: implications géodynamiques sur la phase tarditectonique du rift de Limagne (Massif central, France). Revue canadienne des Sciences de la Terre 45, 641–650. D'Orazio, M., Innocenti, F., Tonarini, S., Doglioni, C., 2007. Carbonatites in a subduction system: the Pleistocene alvikites from Mt. Vulture (southern Italy). Lithos 98, 313–334. Doroshkevitch, A.G., Wall, F., Ripp, G.S., 2007. Calcite-bearing dolomite carbonatite dykes from Veseloe, North Transbaikalia, Russia and possible Cr-rich mantle xenoliths. Mineralogy and Petrology 90, 19–49. Gao, S., Liu, X.M., Yuan, H.L., Hattendorf, B., Gunther, D., Chen, L., Hu, S.H., 2002. Determination of forty two major and trace elements in USGS and NIST SRM glasses by laser ablation-inductively coupled plasma-mass spectrometry. Geostandards Newsletter 26, 181–196. Gittins, J., Harmer, R.E., Barker, D.S., 2005. The bimodal composition of carbonatites: reality or misconception? Lithos 85, 129–139. Goër de Herve, A., 2000. Peperites from the Limagne Trench (Auvergne, French Massif Central): a distinctive facies of phreatomagmatic pyroclastic. History of a semantic drift. In: Leyrit, H., Montenant, C. (Eds.), Volcaniclastic Rocks from Magmas to Sediments. Gordon and Breach Science Publishers, Amsterdam, pp. 91–110. Gouhier, J., 1973. Contribution à l'étude des pépérites de la région du Puy de Mur (Limagne) PhD, Université Orsay (France). Gudfinnsson, G.H., Presnall, D.C., 2005. Continuous gradations among primary carbonatitic, kimberlitic, melilititic, basaltic, picritic, and komatiitic melts in equilibrium with garnet lherzolite at 3–8 GPa. Journal of Petrology 46, 1645–1659. Hansen, L.D., Dipple, G.M., Gordon, T.M., Kellett, D.A., 2005. Carbonated serpentinite (listwanite) at Atlin, British Columbia: a geological analogue to carbon dioxide sequestration. The Canadian Mineralogist 43, 225–239.

146

G. Chazot, J. Mergoil-Daniel / Lithos 154 (2012) 130–146

Harmer, R.E., Gittins, J., 1997. The origin of dolomitic carbonatites: field and experimental constraints. Journal of African Earth Sciences 25, 5–28. Harmer, R.E., Gittins, J., 1998. The case for primary, mantle-derived carbonatite magma. Journal of Petrology 39, 1895–1903. Hooten, J.A., Ort, M.H., 2002. Peperite as a record of early-stage phreatomagmatic fragmentation processes: an example from the Hopi Buttes volcanic field, Navajo Nation, Arizona, USA. Journal of Volcanology and Geothermal Research 114, 95–106. Hubberten, H.-W., Katz-Lehnert, K., Keller, J., 1988. Carbon and oxygen isotope investigations in carbonatites and related rocks from the Kaiserstuhl, Germany. Chemical Geology 70, 257–274. Jakni, B., Dautria, J.M., Liotard, J.M., Briqueu, L., 1996. Mise en évidence d'un manteau carbonaté à l'aplomb du Bas-Languedoc: les xenolites péridotitiques du complexe volcanique de Grand-Magnon (Lodévois). Comptes Rendus de l'Académie des Sciences de Paris 323, 33–40. Jones, A.P., Kostoula, T., Stoppa, F., Woolley, A.R., 2000. Cr spinels in mantle xenoliths in a carbonate rich melilitic tuff from Vulture. Mineralogical Magazine 64, 593–613. Lavecchia, G., Stoppa, F., Creati, N., 2006. Carbonatites and kamafugites in Italy: mantlederived rocks that challenge subduction. Annals of Geophysics 49, 389–402. Le Bas, M.J., 1977. Carbonate–Nephelinite Volcanism; An African Case History. In: John Wiley and Sons (Ed.), . 348 pp. Le Bas, M.J., 1987. Nephelinites and carbonatites. In: Fitton, J.G., Upton, B.G.J. (Eds.), Alkaline Igneous Rocks: Geological Society Special Publication, 30, pp. 53–83. Le Bas, M.J., Le Maitre, R.W., et al., 1986. A chemical classification of volcanic rocks based on the total alkali–silica diagram. Journal of Petrology 27, 745–750. Le Maitre, R.W., Streckeisen, A., Zanettin, B., Le Bas, J., Bonin, B., Bateman, G., Bellieni, G., Dudek, A., Efremova, S., Keller, J., Lameyre, J., Sabine, P.A., Schmid, R., Sorensen, H., Woolley, A.R., 2002. Igneous Rocks: A Classification and Glossary of Terms. In: Le Maitre, R.W. (Ed.), Cambridge University Press. 236 pp. Lecoq, H., Bouillet, J.B., 1830. Vues et coupes des principales formations géologiques du Département du Puy-de-Dome. Imp.Thibaut-Lambriot, Clermont-Ferrand. 266 pp. Lutkov, V.S., Mogarovskii, V.V., Lutkova, V.Y., 2007. Petrogeochemistry and genesis of listwaenite xenoliths in the alkali basalts of the southern Tien. Geochemistry International 45, 733–743. Matsukage, K.N., Kubo, K., 2003. Chromian spinel during melting of experiments of dry peridotite KLB-1 at 1.0–2.5 GPa. American Mineralogist 88, 1271–1278. Merle, O., Michon, L., 2001. The formation of the West European rift: a new model as exemplified by the Massif Central area. Bulletin de la Societe Geologique de France 172, 213–221. Michel, R., 1953. Contribution à l'étude pétrographique des pépérites et du volcanisme tertiaire de la grande Limagne. Mémoire de la Société d'Histoire Naturelle d'Auvergne, 5. 140 pp. Morimoto, N., 1988. Nomenclature of pyroxenes. Mineralogical Magazine 52, 535–550. Mourtada, S., Le Bas, M.J., Pin, C., 1997. Pétrogénèse des magnésio-carbonatites du complexe de Tamazert (Haut Atlas marocain). Comptes Rendus de l'Académie des Sciences de Paris 325, 559–564. Nehlig, P., Boivin, P., Goer de Hervé, A., Mergoil, J., Prouteau, G., Thieblemont, D., 2001. Les volcans du Massif Central. Géologues 130–131, 66–91. Pecerillo, A., 2005. Plio-Quaternary Volcanism in Italy. Springer, Berlin-Heiderlberg. Peterlongo, J.M., 1978. La Limagne. Guides géologiques régionaux. Masson, Massif Central (France), pp. 93–107. Prelevic, D., Stracke, A., Foley, S., Romer, R.L., Conticelli, S., 2008. Mediterranean tertiary lamproites derived from multiple source components in postcollisional geodynamics. Geochimica et Cosmochimica Acta 72, 2125–2156. Raffone, N., Chazot, G., Pin, C., Vannucci, R., Zanetti, A., 2009. Metasomatism in the lithospheric mantle beneath Middle Atlas (Morocco) and the origin of Fe- and Mg-rich wehrlites. Journal of Petrology 50, 197–249. Riley, T.R., Bailey, D.K., Harmer, R.E., Liebsch, H., Lloyd, F.E., Palmer, M.R., 1999. Isotopic and geochemical investigation of a carbonatite–syenite–phonolite diatreme, West Eifel (Germany). Mineralogical Magazine 63, 615–631.

Rock, N.M.S., 1990. Pyroxene nomenclature scheme: computerization and its consequences. Mineralogy and Petrology 43, 99–119. Rosatelli, G., Stoppa, F., Jones, A., 2000. Intrusive calcite–carbonatite occurrence from Mt. Vulture volcano, southern Italy. Mineralogical Magazine 64, 615–624. Rosatelli, G., Wall, F., Stoppa, F., 2007. Calcio-carbonatite melts and metasomatism in the mantle beneath Mt. Vulture (Southern Italy). Lithos 99, 229–248. Rosatelli, G., Wall, F., Stoppa, F., Brilli, M., 2010. Geochemical distinctions between igneous carbonate, calcite cements, and limestone xenoliths (Polino carbonatite, Italy): spatially resolved LAICPMS analyses. Contributions to Mineralogy and Petrology 160, 645–661. Scrope, G.P., 1827. Memoir on the Geology of Central France, including the volcanic formation of Auvergne. In: Longman (Ed.), The Velay and the Vivarais. London, 182 pp. Scrope, G.P., 1858. The Geology and Extinct Volcanos of Central France. In: Murray (Ed.), London, 258 pp. Sheppard, S.M.F., Dawson, J.B., 1973. 13C/12C and D/H isotope variations in “primary igneous carbonatites”. Fortschritte des Mineralogie 50, 128–129. Shimron, A.E., 1975. Petrogenesis of the Tarr albitite–carbonatite complex Sinai Peninsula. Mineralogical Magazine 40, 13–24. Skilling, I.P., White, J.D.L., McPhie, J., 2002. Peperite: a review of magma–sediment mingling. Journal of Volcanology and Geothermal Research 114, 1–17. Stoppa, F., Lloyd, F.E., Rosatelli, G., 2003. CO2 as the propellant of carbonatite–kamafugite cognate pairs and the eruption of diatremic tuffisite. Per.Mineral. Special Issue, Eurocarb. 72, 205–222. Stoppa, F., Rosatelli, G., Wall, F., Jeffries, T.E., 2005. Geochemistry of carbonatite–silicate pairs in nature: a case history from Central Italy. Lithos 85, 26–47. Stoppa, F., Jones, A.P., Sharygin, V.V., 2009. Nyerereite from carbonatite rocks at Vulture Volcano: implications for mantle metasomatism and petrogenesis of alkali carbonate melts. Central European Journal of Geosciences 1, 131–151. Stoppa, F., Rosatelli, G., Schiarzza, M., Tranquilli, A., 2012. Hydrovolcanic vs. magmatic processes in forming maars and associated pyroclasts: the Calatrava-Spain-case history. In: Stoppa, F. (Ed.), Updates in Volcanology — A Comprehensive Approach to Volcanological Problems. InTech — Open Access Publisher. ISBN: 978-953-307-434-4, pp. 3–26. Tommasini, S., Avanzinelli, R., Conticelli, S., 2011. The Th/La and Sm/La conundrum of the Tethyan realm lamproites. Earth and Planetary Science Letters 301–305, 469–478. Valentini, L., Moore, K.R., Chazot, G., 2010. Unravelling carbonatite–silicate magma interaction dynamics: a case study from the Velay province (Massif Central, France). Lithos 116, 53–64. Wallace, M.E., Green, D.H., 1988. An experimental determination of primary carbonatite magma composition. Nature 335, 343–346. Wattinne, A., 2004. Evolution d'un environnement carbonaté lacustre à bioconstructions en Limagne bourbonnaise (Oligo-Miocène, Massif Central, France). PhD Museum National d'Histoire Naturelle de Paris, 189 pp. Weisenberger, T., Spurgin, S., 2009. Zeolites in alkaline rocks of the Kaiserstuhl Volcanic Complex, SW Germany — new microprobe investigation and the relationship of zeolite mineralogy to the host rock. Geologica Belgica 12, 75–91. Wenzell, T., Baumgartner, L.P., Brügmann, G.E., Konnikov, E.G., Kislov, E.V., Orsoev, D.A., 2001. Contamination of mafic magma by partial melting of dolomitic xenoliths. Terra Nova 13, 197–202. Wenzell, T., Baumgartner, L.P., Brügmann, G.E., Konnikov, E.G., Kislov, E.V., 2002. Partial melting and assimilation of dolomitic xenoliths by mafic magma: the Ioko–Dovyren intrusion (North Baikal Region, Russia). Journal of Petrology 43, 2049–2074. White, J.D.J., McPhie, J., Skilling, I.P., 2000. Peperite: a useful genetic term. Bulletin of Volcanology 62, 65–66. Woolley, A.R., Church, A.A., 2005. Extrusive carbonatites: a brief review. Lithos 85, 1–14. Xu, Y.G., Menzies, M., Bodinier, J.L., Bedini, R.M., Vroon, P., Mercier, J.-C., 1998. Melt percolation and reaction atop a plume: evidence from the poïkiloblastic peridotite xenoliths from Borée (Massif Central, France). Contribution to Mineralogy and Petrology 132, 65–84.