Journal of Petrology Advance Access originally published online on December 26, 2006
Journal of Petrology 2007 48(3):537-562; doi:10.1093/petrology/egl071
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Geochemical Evidence for Slab Melting in the Trans-Mexican Volcanic Belt
1Centro de Geociencias, Universidad Nacional Autónoma de México, Querétaro 76230, Mexico
2Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138, USA
3Lamont–Doherty Earth Observatory and Department of Earth and Environmental Sciences, Columbia University, 61 RT. 9W, Palisades, NY 10964, USA
4Instituto de Geología, Universidad Nacional Autónoma de México, Ciudad Universitaria, Mexico City 04510, MEXICO
RECEIVED MARCH 17, 2006; ACCEPTED NOVEMBER 2, 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONTECTONIC AND GEOLOGICAL SETTINGSAMPLES AND ANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONS AND IMPLICATIONSREFERENCES |
|---|
Geochemical studies of Plio-Quaternary volcanic rocks from the Valle de Bravo–Zitácuaro volcanic field (VBZ) in central Mexico indicate that slab melting plays a key role in the petrogenesis of the Trans-Mexican Volcanic Belt. Rocks from the VBZ are typical arc-related high-Mg andesites, but two different rock suites with distinct trace element patterns and isotopic compositions erupted concurrently in the area, with a trace element character that is also distinct from that of other Mexican volcanoes. The geochemical differences between the VBZ suites cannot be explained by simple crystal fractionation and/or crustal assimilation of a common primitive magma, but can be reconciled by the participation of different proportions of melts derived from the subducted basalt and sediments interacting with the mantle wedge. Sr/Y and Sr/Pb ratios of the VBZ rocks correlate inversely with Pb and Sr isotopic compositions, indicating that the Sr and Pb budgets are strongly controlled by melt additions from the subducted slab. In contrast, an inverse correlation between Pb(Th)/Nd and 143Nd/144Nd ratios, which extend to lower isotopic values than those for Pacific mid-ocean ridge basalts, indicates the participation of an enriched mantle wedge that is similar to the source of Mexican intraplate basalts. In addition, a systematic decrease in middle and heavy rare earth concentrations and Nb/Ta ratios with increasing SiO2 contents in the VBZ rocks is best explained if these elements are mobilized to some extent in the subduction flux, and suggests that slab partial fusion occurred under garnet amphibolite-facies conditions.
KEY WORDS: arcs; mantle; Mexico; sediment melting; slab melting
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONTECTONIC AND GEOLOGICAL SETTINGSAMPLES AND ANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONS AND IMPLICATIONSREFERENCES |
|---|
Slab-derived fluxing plays an important role in the global geochemical cycle, and gives rise to the distinctive chemical compositions observed in arc magmas (Gill, 1981
). However, the physical mechanism of element recycling at subduction zones remains controversial: is the subduction component a hydrous fluid, derived from the dehydration of the altered portions of the oceanic crust and its sedimentary cover? Or is it a water-rich silicate melt derived from the partial fusion of the subducted materials? Or, more complicated still, is it a combination of both hydrous fluids and silicate melts? Can we distinguish the chemical fingerprints of these two processes of element recycling in the geochemistry of arc magmas? And if a distinction can be drawn, then which reactions are involved, and how can we relate them to the thermal and tectonic conditions of a subduction zone? Arcs constructed over thick continental crusts provide additional complications because mantle-derived melts could stagnate, differentiate and assimilate pre-existing rocks at various depths along their journey to the surface, in a process that can obscure the chemical imprints of the subduction environment. Is it possible to distinguish the chemical contributions of the continental crust from those derived from the subduction flux? Answering all these questions is essential for understanding and quantifying the solid-earth geochemical cycle on a global scale, and they have implications for more complicated questions such as the origin of the continental crust. In this contribution we provide new geochemical data from the Valle de Bravo–Zitácuaro volcanic field (VBZ), a Pliocene to Quaternary monogenetic field located in the central sector of the Trans-Mexican Volcanic Belt (TMVB). The rocks in this area have major element compositions that are typical of arc volcanic rocks worldwide, and are similar to andesitic rocks erupted from long-lived Mexican stratovolcanoes. They are thus representative of the current tectonic and geological conditions that generate andesitic magmas in the central sector of the TMVB, and are an excellent suite to address magma generation processes in a thick-crusted continental arc. We will show that crystal fractionation and/or crustal contamination did not play a significant role in their petrogenesis, but that the geochemical signatures are better explained by different proportions of slab-derived silicate melts reacting with the sub-arc mantle wedge.
|
TECTONIC AND GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONTECTONIC AND GEOLOGICAL SETTINGSAMPLES AND ANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONS AND IMPLICATIONSREFERENCES |
|---|
The Mexican subduction zone is a natural laboratory to test the efficiency with which chemical elements are subducted and ultimately recycled into arc magmas because parameters that govern its thermal structure gradually change along-strike (Fig. 1). The age and convergence rate of the subducted Rivera and Cocos plates gradually increase eastward along the trench, as the dip angle decreases to a sub-horizontal plane (Pardo & Suárez, 1995
). Although the Wadati–Benioff zone below central Mexico is poorly defined, it has been long considered that the current volcanic front emerges when the subducted slabs reach 
100 km depth, despite significant differences in the subduction angle. Recent thermal models and new focal data compilations have suggested, however, that the slab depth below central Mexico could be as shallow as 70–80 km, and that slab melting could occur at those depths (Manea et al., 2004, 2005). Indeed, slab melting has recently been invoked to explain the origin of some Miocene and Quaternary rock sequences along the arc (Luhr, 2000; Gómez-Tuena et al., 2003; Martínez-Serrano et al., 2004).
|
The compositional variability and the mass of the sediment fill that is subducted are not well known, but evidence from gravimetric profiles indicates that the sediment column could be as thin as
17 m above the young Rivera plate, and that the sediment thickness gradually increases eastwards along the trench (Manea et al., 2003). In addition, Leg 66 of the Deep Sea Drilling Project (DSDP) retrieved six cores off the coast of Acapulco (Moore et al., 1982). DSDP Site 487 (location shown in Fig. 1) reached into the basaltic basement. The sedimentary column of Site 487 has been described and geochemically analyzed (Moore & Shipley, 1988; Plank & Langmuir, 1998; Verma, 2000b; LaGatta, 2003), and consists of two units: 100 m of continentally derived Quaternary hemipelagic sediments that rest above 70 m of late Miocene–Pliocene pelagic sediments. Fragments of altered oceanic crust (AOC) from Site 487 have been described and analyzed, and are mainly composed of an olivine- and plagioclase-bearing basalt with minor amounts of an aphyric basalt (Verma, 2000b; LaGatta, 2003).
The VBZ is currently one of the best studied areas of the TMVB. Recent publications have mapped and addressed the local geology, stratigraphy, petrography, and the major and trace element geochemistry of the volcanic successions (Capra et al., 1997; Blatter & Carmichael, 1998b; Blatter et al., 2001). Furthermore, primitive andesites from the VBZ have also been studied experimentally (Blatter & Carmichael, 2001). In summary, the volcanism in the area is characterized by a myriad of cinder cones, domes and isolated fissural lava flows of andesite–dacite compositions with ages that vary between 5 ka and 6 Ma. The experimentally determined hydrous phase equilibria of the VBZ andesites are suggestive of pre-eruptive H2O contents of 2·5–6 wt %, and the varied phenocryst assemblages in the lavas appear to be the result of water loss during decompression and ascent (Blatter & Carmichael, 2001). The excellent match between the amount of crystals in the natural rocks and the experiments strongly suggests that the lavas are true liquid compositions. Importantly, Blatter & Carmichael (2001) suggested a direct mantle origin for some of the VBZ andesites on the basis of the high Ni and Cr abundances of the volcanic rocks, the presence of nodules of amphibole-bearing peridotite in one of the andesitic lava flows (Blatter & Carmichael, 1998a), and the similarity of compositions between the VBZ lavas and experimental water-saturated lherzolite melts (Hirose, 1997).
|
SAMPLES AND ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONTECTONIC AND GEOLOGICAL SETTINGSAMPLES AND ANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONS AND IMPLICATIONSREFERENCES |
|---|
In this contribution we report comprehensive geochemical data for the youngest volcanic sequences of the Valle de Bravo Zitácuaro volcanic field (< 3·77 Ma), as geologically characterized by Capra et al. (1997
) and Blatter et al. (2001). We deliberately exclude the older and more altered volcanic successions (6–3·77 Ma).
Major element compositions for most samples were obtained by X-ray fluorescence (XRF) using a Siemens SRS-3000 instrument at Instituto de Geología, UNAM, following procedures described elsewhere (Lozano et al., 1995). Trace element abundances were determined by inductively coupled plasma mass spectrometry (ICP-MS) using a VG-PQ2+ at the Lamont–Doherty Earth Observatory (LDEO), following sample preparation and measurement procedures described by Gómez-Tuena et al. (2003). Additional major and trace element data were obtained by ICP-MS at Activation Laboratories of Ancaster, Canada (http://www.actlabs.com/) employing lithium metaborate/tetraborate fusions. Boron concentrations on selected samples were obtained by prompt gamma neutron activation (PGNA) at Activation Laboratories. Reproducibility is better than 10% for Li, Be, Sc, Cr, Ni, Cu, Zn, Mo, Sb, Cs, Tl and Pb, and better than 3% for all other elements. Typical precision for B is 100% at 0·5 ppm, 15% at 5 ppm, and better than 5% at 50 ppm.
Sr, Nd and Pb isotopic compositions were determined by thermal ionization mass spectrometry (TIMS) at LDEO using a VG sector 54-30 system equipped with nine Faraday collectors, following sample preparation and measurement procedures described by Gómez-Tuena et al. (2003). 87Sr/ 86Sr ratios are normalized to 86Sr/ 88Sr = 0·1194 and adjusted to an NBS987 standard ratio of 0·710230. During two separated intervals of analyses, the measured values of the NBS987 standard were 87Sr/ 86Sr = 0·710245 ± 0·000016 (2
, n = 4) and 0·710271 ± 0·000014 (2, n = 6). 143Nd/144Nd ratios were normalized to 146Nd/144Nd = 0·72190 and corrected to a La Jolla Standard value of 143Nd/144Nd = 0·511860. Over the course of this study, the measured 143Nd/144Nd ratio of La Jolla standard was 0·511836 ± 0·000013 (2, n = 15). Pb isotopic ratios were corrected for mass fractionation using the LDEO 207Pb–204Pb double spike. Using this technique, samples are analyzed twice: once mixed with the double spike and once without it. The fractionation corrected values were adjusted to NBS981 standard values of 206Pb/204Pb = 16·9356, 207Pb/204Pb = 15·4891, and 208Pb/204Pb = 36·7006 (Todt et al., 1996). Over the course of this study, the double-spike fractionation corrected Pb isotope ratios of the NBS-981 standard were 206Pb/204Pb = 16·9356 ± 0·0048 (143 ppm), 207Pb/204Pb = 15·4912 ± 0·0047 (152 ppm), and 208Pb/204Pb = 36·7025 ± 0·014 (191 ppm) (2, n = 13). Total procedural blanks between 300 and 500 pg for Pb were negligible compared with the concentrations of this element in the dissolved rock samples.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONTECTONIC AND GEOLOGICAL SETTINGSAMPLES AND ANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONS AND IMPLICATIONSREFERENCES |
|---|
Detailed petrographic descriptions of the VBZ rock suite have been reported in several previous publications (Capra et al., 1997
; Blatter & Carmichael, 1998a, 1998b; Blatter & Carmichael, 2001; Blatter et al., 2001), and the main phenocryst assemblage and modal proportions of the rocks analyzed in this study are given in Table 1. Major and trace element abundances and Sr–Nd–Pb isotopic compositions are reported in Tables 2 and 3. Overall, the VBZ rocks vary from highly porphyritic to nearly aphyric, with the least evolved andesites having a phenocryst mineralogy of olivine (Ol) and orthopyroxene (Opx), or clinopyroxene (Cpx) and Opx. More evolved andesites and dacites usually contain Opx, Cpx, occasional hornblende (Hbl) and rare biotite (Bi). Importantly, most of the lavas do not contain plagioclase as a phenocryst, a remarkable feature that was previously noted by Blatter & Carmichael (1998b). Nevertheless, plagioclase is commonly observed in the groundmass, forming a trachytic texture along with glass and opaque minerals. All kinds of rocks often have small but ubiquitous quartz xenocrysts; some quartzites and schists, which were apparently entrained from the locally exposed basement, are also included as xenoliths. These nodules are often surrounded by a pyroxene shield and/or a glassy aureole, but sharp edges along the contact with the host magma are also commonly observed. These features indicate that the xenoliths were rapidly incorporated into the ascending and cooling magma shortly before eruption.
Like many arcs around the world, the igneous rocks from the VBZ belong to the subalkaline magmatic series (Fig. 2a) and do not display FeO-total enrichment during differentiation, plotting within the calc-alkaline field in an AFM diagram (not shown). The major element compositions of the VBZ rocks are very similar to those observed in the large Mexican stratovolcanoes, classified mostly as andesites and dacites, but some of the least evolved rocks from the VBZ tend to have higher K2O contents at relatively low SiO2 contents (Fig. 2b). These latter lavas were classified as shoshonites by Blatter et al. (2001).
|
|
|
|
VBZ trace elements show relative enrichments of the large ion lithophile elements (LILE) over the high field strength elements (HFSE) that are typical of arc lavas (Fig. 3); nonetheless, a close examination of the trace element systematics allows the recognition of two chemically different rock suites, herein named as VBZ high rare earth element (REE) series (VBZ-HRE) and VBZ low REE series (VBZ-LRE). The VBZ-HRE rocks tend to have higher Sr, Ba and REE contents and show distinctively strong Zr–Hf negative anomalies, an unusual feature that has not been previously observed in other TMVB rock suites (Figs 2f and 3b). VBZ-LRE andesites tend to have high Sr/Y ratios at low Y contents (Fig. 4a), a feature often observed in volcanic arcs where young and hot slabs are being subducted, and where a slab-melt component has been interpreted as the subduction agent (Defant & Drummond, 1990
; Martin, 1999). VBZ-HRE rocks share the high Sr/Y ratios but they also extend to higher Y contents. Although the VBZ rocks display light and middle rare earth element (LREE–MREE) enrichments over the heavy rare earth elements (HREE) (Fig. 4b), this fractionation is not as strong as observed in ‘adakite-like’ magmas of the Miocene TMVB (Gómez-Tuena et al., 2003) or the Andean Austral Volcanic Zone (Stern & Kilian, 1996); but it is comparable with that in high-Mg andesites from Adak island in the Aleutians (Kay, 1978) and from Mount St. Helens in the Cascades (Smith & Leeman, 1996).
|
|
Relationships between Sr, Nd, and Pb isotope ratios and possible end-member components are shown in Fig. 5. Overall, the Sr–Nd isotopic ratios of the VBZ are bracketed between the compositions of East Pacific Rise mid-ocean ridge basalts (EPR-MORB) and an enriched ‘crustal component’, similar in composition to Mexican lower crustal xenoliths and high-grade metamorphic terranes, or to the bulk subducted sediment sampled at DSDP Site 487. Pb isotopes form a tight positive correlation bracketed by EPR-MORB and our estimate of the average value for Site 487 sediments (which could also represent a crustal component) (Fig. 5c and d). Sr and Pb isotopic compositions largely overlap within the suites, but the Nd isotope compositions of the VBZ-HRE rocks appear to be slightly shifted towards higher values at a given 87Sr/86Sr ratio. Interestingly, rocks from the VBZ tend to overlap the compositions observed at Popocatépetl stratovolcano but they also extend to relatively lower Sr and Pb isotope ratios at similar 143Nd/144Nd ratios.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONTECTONIC AND GEOLOGICAL SETTINGSAMPLES AND ANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONS AND IMPLICATIONSREFERENCES |
|---|
The overall trace element patterns of VBZ lavas are typical of arc-related andesites, for example with Nb–Ta depletions, Pb enrichments, and high Sr/Nd and Ba/LREE (Fig. 3). However, different VBZ rock suites with distinct trace element patterns and isotopic compositions erupted concurrently in the area. In addition, VBZ lavas display compositions that are different from other Mexican stratovolcanoes (Fig. 4). For example, VBZ lavas tend to have higher Sr/Y and Sm/Y ratios, but lower Cs/La and Rb/Ba ratios than Popocatépetl lavas at similar fractionation levels. The geochemical differences observed within the VBZ rocks, and at different geographical positions within the arc, might be related to variations in the subduction geochemical fluxes, or they might be governed by different components and processes that operate during magma ascent through the upper plate (i.e. crystal fractionation, crustal melting and/or contamination). In this section we discuss the different components and processes that controlled the compositions of the VBZ rocks and provide a quantitative model for their petrogenesis.
Crystal fractionation and crustal contamination
Several workers have recognized that a significant part of the geochemical variability observed in continental arcs could be attributed to a process of crustal melting and/or assimilation, such that the chemical imprints derived from the subduction environment are largely obscured (Hildreth & Moorbath, 1988). Indeed, based on isotopic and trace element modeling, some workers have suggested that crustal contamination has played an important role in the petrogenesis of differentiated magmas emplaced across the TMVB (Verma, 1999, 2000a; Lassiter & Luhr, 2001; Chesley et al., 2002; Schaaf et al., 2005). Therefore, evaluating the role of magma differentiation and crustal contamination in the genesis of the studied rocks is important if we aim to recognize and quantify the processes that take place at deeper levels in the subduction zone.
The geochemical trends observed in the major and trace element variation diagrams (Figs 2 and 4), together with the variability in the isotopic compositions of the two VBZ rock suites (Fig. 5), rule out simple crystal fractionation from a common primitive magma as the dominant petrogenetic process. For instance, whereas the major element compositions of both rock suites display coherent trends, Sr and REE contents are significantly higher in the VBZ-HRE than in the VBZ-LRE (Fig. 2). Clearly, these compositional differences cannot be simply accounted for by crystal fractionation from a cogenetic primitive magma, but should instead be a reflection of distinct magma source regions for both rock suites.
Atherton & Petford (1993) showed that some rocks that display slab-melt geochemical features, such as high LREE/HREE and Sr/Y ratios, could be the result of partial fusion of newly underplated, garnet-bearing, basaltic crust. Indeed, several workers have recognized that melting of a thickened crust by underplating of mantle-derived magmas can form SiO2-rich liquids with trace element characteristics that are indistinguishable from those related to partial fusion of the subducted oceanic crust (Conrey et al., 2001; Hou et al., 2004; Wang et al., 2005). The VBZ rocks, however, display significantly higher MgO contents (to 9 wt %) and Mg-number (75 to 50) that are unlike those attributed to direct partial melting of a mafic garnet-bearing lower continental crust (Mg-number <50, Fig. 2c). In addition, high Sr/Y ratios are coupled with less radiogenic Pb and Sr isotopic compositions (Fig. 6), and thus the isotope ratios of the higher Sr/Y lavas trend towards the composition of Pacific MORB, and away from the isotopic compositions of lower crustal xenoliths and Mexican basement terranes (Fig. 5). Therefore, if the VBZ rocks with high Sr/Y ratios were derived from melting or assimilation of the pre-existent continental crust, the isotopic composition of the contaminant must be more like Pacific MORB than currently known Mexican basement.
|
Correlations between isotopic compositions and indices of fractionation are often regarded as the best indicators of crustal contamination (assimilation–fractional crystallization; AFC) because this process involves the incorporation of a magma derived from the partial fusion of the crust, and the concurrent fractionation of crystalline phases (DePaolo, 1981
). However, the isotopic compositions of the VBZ are largely uncorrelated with Mg-number (Fig. 6c and d), precluding a simple AFC process. Furthermore, VBZ rocks display a puzzling decrease in MREE, HREE (e.g. Fig. 2e), and Nb/Ta ratios (Fig. 7a) with increasing SiO2 contents. Correlations like these are a common feature of many rock suites emplaced throughout the TMVB (Luhr & Carmichael, 1985b; Verma, 1999, 2000b; Straub & Martin-Del Pozzo, 2001; LaGatta, 2003; Siebe et al., 2004; Schaaf et al., 2005). Fractional crystallization, with or without crustal assimilation, that involves the formation of a typical low-pressure mineral assemblage (Ol, Plag, Cpx and Opx) is unable to explain such trends because the HFSE and the REE are not significantly incorporated in these mineral species. Fractionation of Fe–Ti oxides from a basaltic magma will decrease the Nb and Ta contents of the derivative liquids, but it will also increase their Nb/Ta ratios because DNb/DTa < 1 in these phases (Green & Pearson, 1987).
|
It has also been proposed that amphibole fractionation at shallow levels (Castillo et al., 1999
), or fractionation of amphibole and garnet from hydrous basalts at high pressures (Müntener et al., 2001), can explain the slab-melt geochemical features of some adakites (Castillo et al., 1999; Garrison & Davidson, 2003; MacPherson et al., 2006). Amphibole fractionation at either low or high pressures could in principle be responsible for the low MREE, Nb and Nb/Ta ratios in more evolved lavas, but in this case enormous quantities of fractionating amphibole (> 85%) would be required to explain the observed variations (Fig. 7b and c). High-pressure fractionation experiments involving hydrous basalts essentially form residual websterites, with amphibole and garnet appearing at higher degrees of crystallization (Müntener et al., 2001). High-pressure differentiation will also produce corundum-normative derivative liquids (Müntener et al., 2001), and the strongest garnet signatures will probably be observed in the most evolved rocks (MacPherson et al., 2006). Except for one relatively altered sample (ZIT-99-19), none of the VBZ rocks studied here are corundum normative. Furthermore, the strongest garnet signatures are observed in the relatively silica-poor VBZ-HRE (e.g. higher Sm/Y, Fig. 4b), in contrast to what would be expected if both suites were genetically related to deep crystal fractionation. In addition, even if a common parental basalt for the Mexican andesites exists at depth, the trace element and isotopic variations observed within the VBZ suite (Figs 3 and 5), and at different locations along the arc (Fig. 4), will be hard to explain by differentiation alone, and probably require a compositionally heterogeneous primary source. In conclusion, although some fractionation and interaction with the continental crust is clearly unavoidable, as the magmas must traverse at least 30 km of continental crust before eruption, and there is unequivocal evidence of this interaction in the form of xenocrysts and xenoliths, the overall geochemical variations of the VBZ rock suites cannot easily be accounted for by crystal fractionation from a common primitive magma, nor by crustal assimilation or direct crustal anatexis; therefore, the following discussion assumes that their compositions largely represent mantle-derived and subduction-related phenomena.
Insights into the composition of the Mexican mantle wedge
The concentrations and ratios of the HFSE and the MREE–HREE are often used to infer the characteristics of the mantle source of arc lavas, because these elements are largely insoluble in aqueous fluids and thus are not significantly incorporated into the subduction flux (Class et al., 2000; Hochstaedter et al., 2001; Straub et al., 2004; Pearce et al., 2005). Based on this principle, arc magmas are generally considered to be derived from the partial fusion of a depleted N-MORB-like mantle source. In Mexico the composition of the pre-subduction mantle wedge is elusive because of the compositional heterogeneity of primitive basalts (Luhr & Carmichael, 1985a; Wallace & Carmichael, 1999; Verma, 2000b; Petrone et al., 2003).
Nonetheless, alkaline magmas with negligible subduction signatures and higher Nb contents and Nb/Ta ratios than MORB have been recognized all across the TMVB (Fig. 7). These rocks, often regarded as ‘intraplate’ or ‘ocean island basalt (OIB)-like’ magmas, also have isotopic compositions more ‘enriched’ than MORB (Verma, 2000b; Gómez-Tuena et al., 2003; Petrone et al., 2003). The presence of ‘intraplate-like’ rocks in close space–time association with more typical calc-alkaline arc volcanic rocks has been taken as indicating a highly heterogeneous mantle wedge in the Mexican subduction zone (Luhr & Carmichael, 1985a; Wallace & Carmichael, 1999; Verma, 2000b; Petrone et al., 2003).
Some researchers have suggested that these enriched mantle domains could have been dragged from the back-arc region into the arc source by mantle convection and corner flow (Luhr & Carmichael, 1985a; Wallace & Carmichael, 1999). Others have considered the existence of a mantle plume (Márquez et al., 1999), or association with decompression melting by rifting (Verma, 2002). More recently, Ferrari (2004) suggested that these enriched mantle domains were probably infiltrated into the arc source region following a slab detachment event. However, regardless of the tectonic explanations that have been suggested for the existence of ‘intraplate’ rocks in the Mexican arc, their presence is often considered to be an anomaly that was somehow introduced into the more typical ‘depleted mantle’, because their volume is appreciably small when compared with the more abundant arc-related calc-alkaline sequences.
The decrease of HFSE and MREE–HREE contents, together with Nb/Ta ratios, with increasing SiO2 contents (Fig. 7) is a puzzling characteristic that has been observed in many rock sequences emplaced all across the TMVB (Luhr & Carmichael, 1985b; Verma, 1999, 2000b; Straub & Martin-Del Pozzo, 2001; Siebe et al., 2004; Schaaf et al., 2005). Notably, intraplate volcanic rocks with negligible subduction signals often plot as an enriched end-member in a continuum trend with the rest of the arc-like sequences (Fig. 7). As mentioned above, and as pointed out by previous workers, correlations like these cannot be easily accounted for by a simple liquid line of descent triggered by the crystallization of a typical mineral assemblage, at either low or high pressures, even one extensively modified by crustal assimilation (Fig. 7). In addition, Nb/Ta ratios should not be significantly fractionated by moderate to high extents of mantle melting, because partitioning between these elements and the dominant mantle mineralogy are very similar. Therefore the variations observed in Fig. 7 could, in principle, be attributed to the inherited compositional heterogeneity of the sub-arc mantle wedge beneath Mexico that has been fluxed to various extents by the subduction agents. This interpretation will also mean, however, that rocks with high SiO2 contents may still represent direct mantle melts because Nb/Ta ratios correlate inversely with SiO2, and this variation does not appear to be an artifact of fractionation or crustal contamination.
Water-saturated melting experiments on peridotite have been able to produce high-MgO andesites (Hirose, 1997), and experimental evidence of lavas from the VBZ (Blatter & Carmichael, 2001), together with the presence of amphibole-bearing peridotite nodules in at least one andesitic lava flow (Blatter & Carmichael, 1998a), appears to support the possibility of andesite formation by direct partial melting of the mantle. Nonetheless, to our knowledge, no hydrous melting experiment of peridotite has produced melts with such high silica content as those observed here. The conundrum is thus clear: either there is a mechanism to form very high silica andesites by direct partial melting of the mantle, or the Nb/Ta ratios in this case are simply not reflecting the characteristics of the mantle source.
The usefulness of HFSE ratios as indicators of the mantle source in arc settings relies heavily on the assumption that these elements are not mobilized and/or fractionated by the subduction agents; that is, they behave as ‘conservative’ elements. Support for the insoluble nature of the HFSE in aqueous fluids is clearly established by the experimental evidence (McCulloch & Gamble, 1991; Brenan et al., 1994, 1995c), and the relative low HFSE contents observed in arc rocks with large slab-derived fluid signatures alone (Straub et al., 2004). Nonetheless it has also been increasingly recognized that the HFSE could be mobilized to some extent if the subduction component is in the form of a partial melt coming from the metamorphosed subducted basalt or its sedimentary cover (Stolz et al., 1996; Bourdon et al., 2002). The mobility of these elements, and the possibility to fractionate them, will depend on the residual mineral assembly of the partially melted slab. It has been long considered that rutile should be a residual phase during partial melting of the metamorphosed oceanic crust and sediments, because otherwise the typically high LILE/HFSE ratios observed in modern adakites and Archaean tonalite–trondhjemite–granodiorite (TTG) sequences could not be preserved (Kelemen et al., 2004). Nonetheless, experimental evidence indicates that rutile preferentially incorporates Ta over Nb (DNb/DTa < 1), so that high Nb/Ta ratios might be expected in the partial melts when a rutile-bearing residual assemblage is formed (Green, 1995; Schmidt et al., 2004). In contrast, recently determined partition coefficients between these elements and low-MgO amphiboles indicate that DNb/DTa > 1, and therefore low Nb/Ta ratios might be expected in the slab-derived partial melts when amphibole remains in the residue and the melt fraction is low enough (Tiepolo et al., 2000). Notably, even if some rocks with slab-melt geochemical characteristics have the high Nb/Ta ratios indicative of residual rutile (Stolz et al., 1996; Gómez-Tuena et al., 2003), the vast majority of the studied adakites, as well as the Archean TTG sequences and the bulk continental crust, display subchondritic Nb/Ta ratios (Foley et al., 2002). This characteristic has been taken as evidence for the importance of residual amphibole during partial melting of the subducted slab (Foley et al., 2002), although other investigators have argued that the low Nb/Ta ratios could still be formed with rutile-bearing eclogite residues if the original lithologies involved in the partial melting event had low Nb/Ta ratios to begin with (Rapp et al., 2003).
Although the specific residual mineralogy involved during a slab melting event is hard to define, the experimental results provide insights into the origin and compositional characteristics of the mantle wedge and subduction components involved in the petrogenesis of the Mexican arc. The Nb/Ta variations (Fig. 7) might be explained if the slab-derived chemical flux affecting the arc source region is a water-rich silicate melt derived from the partial fusion of the subducted basalt and sediments, rather than an aqueous fluid in which the HFSE and the HREE are immobile. Low HFSE–HREE contents, together with low Nb/Ta ratios, strong subduction signatures and high SiO2 contents, are all typical characteristics of experimental slab melts and adakites worldwide, and are consistent with melting of garnet amphibolite. These siliceous melts may either react or mix with mantle peridotite and/or their melts during ascent through the mantle and crust (Kelemen, 1998; Prouteau et al., 2001; Tatsumi, 2001). Indeed, Fig. 7c shows that the Nb–Ta concentrations and ratios can be explained by mixing of a slab-derived silicate melt with an enriched ‘intraplate’ magma. The model uses the AOC (altered ocean crust) from DSDP Site 487 as a starting composition (Nb 0·49 ppm, Ta 0·034 ppm, Nb/Ta = 14·5), and assumes garnet amphibolite in the residue (DNb = 0·775 and DTa = 0·555; see Table 4) at 1% batch melting (AOC melt: Nb 0·63 ppm, Ta 0·0607, Nb/Ta = 10·4). Even though very low quantities of residual rutile (< 1%) can significantly affect the Nb/Ta ratios of the resulting slab melt, and sediment contributions are not considered in this case, this exercise provides a feasible explanation for a typical geochemical feature displayed by several volcanic successions of the Mexican arc.
|
The above-mentioned hypothesis thus implies that the low HREE–HFSE contents and Nb/Ta ratios observed in the silica-rich calc-alkaline rocks could represent the inherited compositional characteristics of a slab-derived silicate melt that had a small contribution from the background mantle wedge (or its partial melts). The greater the interaction of the slab melts with the mantle wedge, or the higher the proportion of the ‘intraplate’ magmatic component in a possible magma mixing event, the larger the dilution effect of the slab-melt component in the resulting magmas. In this scenario, the enriched ‘intraplate’ basalts should best represent the composition of the pre-subduction background mantle wedge in Mexico. This will not imply, of course, that the mantle wedge below Mexico is compositionally homogeneous and identical to an OIB-like source. More research will be needed to test this hypothesis. However, the evidence does indicate that the geochemical tools that have proved useful in other arcs for identifying the composition of the mantle wedge are probably not applicable for the Mexican arc because the HFSE and the MREE–HREE have become an intrinsic ingredient of the subduction flux.
Subduction agents: the effect of AOC and sediment melts
As mentioned above, the VBZ rock suites display contrasting trace element patterns, slightly different isotopic compositions, and often follow diverging trends in several incompatible element ratio plots when compared with other Mexican volcanoes (Figs 3–5
). If the geochemical compositions are not significantly controlled by crystal fractionation and/or additions from the continental crust, then one can safely assume that they are controlled by the contributions from the subducted slab and the mantle wedge.
Element ratios between Th, Pb, Nd and Sr are useful indicators of slab-derived chemical fluxes because the partitioning of these elements between slab-derived fluids or melts have been extensively studied (Defant & Drummond, 1990; Drummond & Defant, 1990; Brenan et al., 1995a; Elliott et al., 1997; Johnson & Plank, 1999; Class et al., 2000; Foley et al., 2002; Kessel et al., 2005). Th and Nd are fluid immobile (Brenan et al., 1994, 1995c; Stadler et al., 1998; Kessel et al., 2005), whereas Pb and Sr are soluble in aqueous fluids (Brenan et al., 1994, 1995c; Keppler, 1996; Adam et al., 1997; Ayers et al., 1997; Stadler et al., 1998; Kessel et al., 2005). Therefore, Th and Nd budgets in arc magmas should be mainly controlled by their relative abundances in the mantle wedge plus the slab-melt additions (sediment + AOC). The Pb and Sr contributions, on the other hand, can be controlled by fluids and/or slab-melt additions to the mantle wedge because they are highly soluble, and fairly incompatible during partial melting of eclogite and amphibolite sources (Defant & Drummond, 1990; Tiepolo et al., 2000; van Westrenen et al., 2001; Barth et al., 2002; Foley et al., 2002; Kessel et al., 2005). In addition, because the altered (or fresh) subducted oceanic crust and overlying sediments have different Sr, Nd and Pb isotopic compositions (Verma, 2000b; Gómez-Tuena et al., 2003; LaGatta, 2003), the isotopic variations observed in the magmatic rocks can be used to differentiate between different fluids or melts originating from different portions of the subducted slab.
VBZ samples show a positive correlation between Th/Nd and Pb/Nd, indicating that Pb and Th are incorporated at a constant ratio in the subduction flux (Fig. 8a). The coupling of a fluid-mobile element (Pb) with an insoluble one (Th), and a correlation trending towards higher Th/Nd ratios than those of the subducted sediments, are indications that sediment additions into the mantle source are in the form of a silicate melt rather than an aqueous fluid. This interpretation is further confirmed in the rough but discernible negative correlation of Pb/Nd and Nd isotopes (Fig. 8b), a characteristic that has been often cited as indicating a sediment-melt component that contributes Th, Pb and unradiogenic Nd (Elliott et al., 1997; Class et al., 2000; Hochstaedter et al., 2001; Gómez-Tuena et al., 2003; Kelemen et al., 2004). However, low Pb/Nd ratios extend to lower 143Nd/144Nd values (Fig. 8b) than the EPR-MORB, indicating that the mantle wedge component contributing to the magmas has to be more isotopically enriched (i.e. lower 143Nd/144Nd) than the EPR-MORB source, and more similar to ‘OIB-like’ intraplate basalts.
|
The simple scenario of sediment-melt additions into an enriched mantle wedge is further complicated when the contribution of Sr, and the isotopic compositions of Sr and Pb, are taken into account. Figure 8c shows that the VBZ rocks form an inverse correlation between Sr/Pb and Pb isotope ratios that trend toward a DSDP Site 487 sediment end-member with high Pb isotope ratios on one hand, but whose other end-member is an isotopically depleted (low Pb isotope ratio) MORB-like component with Sr/Pb ratios that are higher than EPR-MORB. This component is different from what would be expected from an aqueous fluid derived from the dehydration of AOC, but is consistent with a silicate melt derived from its partial fusion. Although Pb and Sr are soluble elements, experimental data indicate that Pb is significantly more incompatible than Sr in aqueous fluids extracted from a plagioclase-free source, and that it is only when partial melting is achieved that Sr and Pb are incorporated in the melt at similar rate (Brenan et al., 1995b
, 1995c; Kessel et al., 2005). Supercritical fluids will preferentially incorporate Sr over Pb, but this will only happen at much higher pressures (Kessel et al., 2005). In addition, Fig. 8d shows that MORB or AOC, bulk Site 487 sediment, and the VBZ data form a hyperbolic trend for Sr–Pb isotope ratios that is consistent with binary mixing. The curvature of the trend can only be reproduced if the depleted MORB-like component is a partial melt with a relatively higher Sr/Pb ratio than that of a fluid extracted from the same source.
Modeling subduction and mantle contributions
Figure 8 shows that the overall compositions of the studied rock suites can be explained by mixing of two chemically distinct subduction components derived from the partial fusion of the subducted sediments and underlying oceanic crust. Participation of an enriched mantle wedge, with lower Nd isotope ratios than EPR-MORB, is also apparent from Fig. 8b. Although the precise chemical compositions of these end-members are difficult to constrain, they can be calculated to a reasonable extent with the geological and compositional information available (Table 4).
We have shown above that the entire trace element budget of the VBZ rocks could have been affected by contributions from the subducted slab. This complicates calculation of the composition of the background mantle wedge in the region. Nonetheless, the data indicate that its composition is more similar to the enriched OIB-like mantle that gives rise to the Mexican intraplate basalts than to the source of EPR-MORB (Figs 7 and 8). For this reason, we calculated the trace element composition of the mantle wedge component by inverting the trace element contents of a MgO-rich basalt from the Chichinautzin volcanic field (Wallace & Carmichael, 1999) that shows very little indication of a subduction-derived component (Table 4). The concentrations of Ba, Pb and Sr in the original sample were slightly reduced to have a smooth trace element pattern in a MORB-normalized trace element variation diagram, but no further corrections were made to account for fractionation. We thus emphasize that the modeled slab contributions to the calculated mantle wedge should be considered as maxima in this case. The Nd isotopic composition of the mantle (143Nd/144Nd = 0·51301) was calculated by extrapolation of the trend observed in Fig. 8b, and is similar to that observed in intraplate basalts from other parts of Mexico (Luhr et al., 1989; Pier et al., 1989; Wallace & Carmichael, 1999; Lassiter & Luhr, 2001; Gómez-Tuena et al., 2003; LaGatta, 2003; Petrone et al., 2003). Because the Pb and Sr isotopic compositions of the studied rock suites are strongly controlled by the slab-derived agents (Fig. 8d), the values for the mantle wedge were assumed to be slightly more enriched than the average EPR-MORB (87Sr/86Sr = 0·7029, 206Pb/204Pb = 18·68, 207Pb/204Pb = 15·58, 208Pb/204Pb = 38·40), and are also similar to the values observed in Mexican intraplate basalts (Table 4).
The composition of the sediment melt was calculated using the weighted bulk sediment composition from DSDP Site 487 (Gómez-Tuena et al., 2003; LaGatta, 2003) and assuming 5% batch melting, using the partition coefficients reported by Johnson & Plank (1999) (Table 4). The models consider no major changes in fluid-mobile element contents during fore-arc devolatilization, and assume that melting of the bulk subducted sediment occurs at arc front depths.
A major uncertainty in this modeling is the composition of the altered oceanic crust. The AOC at DSDP Site 487 (Verma, 2000b) is an obvious possibility; however, these particular samples have extremely depleted REE patterns and higher 143Nd/144Nd ratios when compared with most East Pacific Rise samples [PETDB: Petrological Database of the Ocean Floor, 2006, http://www.petdb.org/(Lehnert et al., 2000)]. Therefore, it is unclear to what extent the Site 487 AOC samples represent the bulk oceanic crust that is being subducted along the Middle American Trench (Gómez-Tuena et al., 2003). Nonetheless, if an average composition of a fresh EPR-MORB is chosen as the starting material, then there will be a large uncertainty on the amount of alteration and mobility of key elements that might be transferred to the arc source region during subduction. For these reasons, the models below use both the composition of the pristine AOC at DSDP Site 487 and a hypothetical AOC calculated by mixing 20% of an average EPR-MORB with the DSDP 487 AOC (Table 4). The compositions of the AOC melts were calculated using the average and/or best-fit partition coefficients of andesitic and dacitic liquids, and assuming 1% batch melting with a garnet amphibolite residuum (Table 4).
Natural and experimental slab melts are not in equilibrium with mantle olivine and should react with mantle peridotite during ascent (Rapp et al., 1999). Various experiments and models indicate that slab-melt–peridotite interaction forms Opx-rich lithologies at the expense of olivine, but other phases have also been identified: Al-augite, garnet, plagioclase and hydrous minerals such as amphibole and/or phlogopite (Kelemen, 1998; Prouteau et al., 1999; Rapp et al., 1999; Tatsumi & Hanyu, 2003; Kelemen et al., 2004). Thus modeling such a reaction strongly depends on the partition coefficients between trace elements and the mineral species that are consumed and created. Some workers have attempted to model these reactions using the AFC formulation of DePaolo (1981), and assuming that the slab melts heat as they rise and decompress while assimilating mantle peridotite (Tatsumi & Hanyu, 2003; Kelemen et al., 2004). Others have modeled the process using a simple mixing approach, considering that small amounts of slab-derived additions will not have a significant effect in modifying the mantle mineral paragenesis (Elliott et al., 1997; Class et al., 2000; Gómez-Tuena et al., 2003). Both kinds of models, although conceptually distinct, tend to provide similar results because the incompatible trace element patterns will be essentially governed by the slab-melt additions and the amount of mantle that could be assimilated or mixed.
Given the large uncertainty in the amount and compositions of the minerals that could be consumed and precipitated during melt–peridotite interaction, we have chosen to use the simple mixing approach. Nonetheless, we acknowledge that the ability for a slab melt to remain liquid during ascent will depend on the so-called ‘effective’ melt/peridotite ratio (Rapp et al., 1999). If this ratio remains high, then slab melts could escape significant modification of their incompatible element contents (Rapp et al., 1999). Relatively high bulk Mg-number and Ni for given SiO2 (see Fig. 2c) will probably be the only reliable indicators of slab-derived andesitic–dacitic melts interacting with the mantle wedge (Kelemen et al., 2003; Kessel et al., 2005). In contrast, if the melt/peridotite ratio is relatively low, then melts could solidify completely and form a modally metasomatized, Opx-rich, hybrid mantle that could be later remelted during decompression and ascent. As the partition coefficients for Opx and Ol and most of the elements considered here are extremely low, then moderate to large extents of melting of such a metasomatized mantle will not significantly change the trace element patterns of the resulting melts, and they will be similar to the composition of the bulk slab component itself, unless the proportions of accessory mineral phases such as amphibole, phlogopite or rutile are significantly large. The absolute concentrations, on the other hand, will be controlled by the slab-melt/peridotite ratio, the residual mineralogy, and the extent of melting.
Modeling results
Figure 9 shows modeling results for slab-melt and mantle interactions in Sr, Nd and Pb isotopic variation diagrams. If a melt of the AOC sampled at DSDP Site 487 is used as the depleted MORB-like component, the mixing line in Sr–Nd isotopic space (Fig. 9a) passes over the isotopic compositions of Popocatépetl volcano but is largely unable to reproduce the VBZ data. This probably indicates that the AOC component contributing to VBZ has a slightly different composition from the one contributing to Popocatépetl. For this reason, we slightly adapted the AOC isotopic compositions to fit the observed VBZ geochemical trends, while keeping the bulk-sediment constituent fixed. The isotopic values for this new ‘AOC’ are within the range of those observed in EPR-MORB for Pb and Nd but have a slightly higher 87Sr/86Sr ratio and Sr contents to account for alteration. The rest of the trace elements represent a mixture of 80% AOC from DSDP Site 487 and 20% of an average fresh sample from the EPR (Table 4).
|
).By using this hypothetical AOC the VBZ-LRE rock suite samples plot within a mixture made of 80:20 and 60:40 AOC melt:sediment melt (Fig. 9). An average VBZ-LRE rock could be explained by a mixture of about 70:30 AOC melt:sediment melt (large filled triangle in Fig. 9), with very little contribution from the mantle wedge. The trace element pattern of such a mixture is very similar to an average VBZ-LRE rock, except for the fluid-mobile elements (Rb, Ba, U and Pb) (Fig. 10a). Several possible explanations can account for these differences. They might reflect the above-mentioned uncertainty in the composition of the AOC that is being subducted, or it could be that some of these elements have been retained in hidden accessory mineral phases that were not considered in the models, such as residual mica [e.g. Elburg et al. (2002
)]. Nonetheless, the remarkable match in the rest of the elements, and the fact that the discrepancy affects only the most fluid-mobile elements, makes these interpretations unlikely. We consider it a more plausible explanation that a portion of these highly soluble elements has been simply lost in a fluid phase prior to melting at arc-front depths. This is conceivable because the solidus temperatures of wet basalt or sediments are significantly higher than the stability field of several of the hydrated mineral species that become unstable at relatively low metamorphic grades during subduction (Johnson & Plank, 1999; Poli & Schmidt, 2002). Thus, it seems inevitable that fluid release from the subducted slab should precede anatexis. As the slab component contributing to VBZ-LRE is clearly a composite, it is difficult to quantify precisely the extent of element retention or depletion that affected the sediments and/or the altered oceanic crust. Nevertheless, if the modeled VBZ lavas are true liquid compositions, as previously indicated by the experimental evidence (Blatter & Carmichael, 2001), or have undergone a small amount of fractionation or trace element dilution during mantle interaction, then 66% of the Rb, 65% of the Ba, 50% of the U and 40% of the Pb must have been removed (or retained) from the available fluid-mobile element budget before being incorporated into the calculated bulk-slab component.
|
The VBZ-HRE rocks tend to have higher MgO and Ni contents, more radiogenic Nd isotope compositions, and lower Ba/La, Th/Nd and Pb/Nd ratios than the VBZ-LRE suite, features that are all indicative of a relatively higher contribution from the mantle wedge (Fig. 8). Figures 9c and d shows the effects of mixing a bulk-slab component (made of 60:40 AOC melt:sediment melt) into the calculated mantle wedge. The proportion of slab component needed to reproduce the isotopic compositions of an average VBZ-HRE rock can be constrained to be
10%. Batch melting of such a metasomatized mantle (90% mantle: 10% slab-melt component) can reproduce fairly well the overall trace element patterns of an average VBZ-HRE rock if the fluid-mobile element contents of the bulk slab are adapted for depletion (or retention) in the proportions described above (Fig. 10b). Nevertheless, a model that assumes a peridotitic (or even pyroxenitic) residual mantle mineralogy is unable to reproduce the strong negative Nb–Ta and Zr–Hf anomalies of the VBZ-HRE rock suite because the partitioning of the HFSE and their neighboring elements in the normalized trace element diagrams are considered to be very similar. The HFSE should remain unaffected during the fluid loss that could have affected the LILE because these elements are insoluble in aqueous fluids. In addition, even though negative Zr–Hf anomalies are a common feature of sediments at DSDP Site 487, and are especially strong in the lower pelagic (non-terrigenous) section of the column, these sediments also have MORB-like Pb isotopic compositions (LaGatta, 2003), and their preferential incorporation into the VBZ-HRE magmas should be evident in plots involving Pb isotopes. As the VBZ-LRE and VBZ-HRE rocks are practically indistinguishable in their Pb isotopic compositions, the origin of the Zr–Hf anomalies observed in the VBZ-HRE cannot be related to the preferential incorporation of DSDP Site 487 pelagic sediments.
The presence of residual amphibole and/or rutile is often considered to play a significant role in the formation of Nb–Ta anomalies in slab-derived melts, but very little has been written about the effect on Zr and Hf. As mentioned before, the low Nb/Ta contents and ratios of the VBZ-LRE probably derive from the effect of residual amphibole during slab melting, but this same mineral has also been used to explain the absence of fractionation between Zr–Hf and the REE, because low-Mg amphibole favors the MREE over Zr and Hf (Foley et al., 2002). At first sight it might appear that a different residual mineralogy involving rutile (or zircon) during slab melting explains the negative Zr–Hf anomalies observed in the VBZ-HRE suite. In detail, however, this is not sufficient to elucidate why these rocks appear to have a higher contribution from the mantle wedge, or why those more likely representatives of pristine slab melts (the VBZ-LRE) do not display negative Zr–Hf anomalies.
A possible solution to this dilemma is that the Zr–Hf anomalies have been formed during partial fusion of a parcel of mantle that has been intensively modified by the addition of slab-derived silicate melts. Indeed, if slab melting does take place, then strong reaction of these melts with the surrounding mantle will occur unless the melt/peridotite ratio remains high (Rapp et al., 1999). If the effective melt/peridotite ratio of an average VBZ-HRE magma is assumed to be as low as that constrained by the isotope diagrams (90% mantle: 10% slab melt), then it is very likely that the slab-derived melts were readily consumed and transformed into pyroxene-rich metasomatic lithologies. If these veins were later remelted in the presence of a residual accessory phase (e.g. rutile or zircon), the resulting magmas will display trace element patterns with unusual anomalies in those elements that became more compatible. In fact, Fig. 10b shows that a very small amount of residual rutile (0·3%) during partial melting of such a modally metasomatized mantle can account for the strong Nb–Ta and Zr–Hf negative anomalies that are characteristic of the VBZ-HRE rocks. This could also explain the relatively high Nb/Ta ratios of the VBZ-HRE suite at relatively low Nb contents (Fig. 7).
A question that remains is whether an accessory titaniferous phase such as rutile can precipitate and remain stable in the mantle wedge after a partial melting event. Experimental studies and calculations have shown that rutile is not a likely residual phase in the source of basaltic melts, as the TiO2 contents in basalts are too low (Green & Pearson, 1986; Ryerson & Watson, 1987); however, other studies have indicated that rutile could become stable in peridotite–melt systems under hydrous, silica-rich conditions (Schiano et al., 1994, 1995; Bodinier et al., 1995). In addition, accessory rutiles have also been directly observed and analyzed in modally metasomatized peridotites from Siberia (Ionov et al., 1999; Kalfoun et al., 2002). These studies have shown that rutile could become an important HFSE-bearing phase at least in the conditions of the upper mantle, and thus provide independent support for the presence of residual rutile during partial melting of a modally metasomatized mantle wedge.
|
CONCLUSIONS AND IMPLICATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONTECTONIC AND GEOLOGICAL SETTINGSAMPLES AND ANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONS AND IMPLICATIONSREFERENCES |
|---|
The geochemical data and modeling results provided here indicate that slab melting plays a key role in the petrogenesis of andesitic rocks of Valle de Bravo–Zitácuaro, in the central Trans-Mexican Volcanic Belt. Although these magmas have traversed a 30–40 km thick continental crust, the geochemical results do not support simple crystal fractionation or crustal contamination as the main petrogenetic processes. Instead, the geochemical variations within the studied rock suites can be largely attributed to contribution of partial melts derived from the subducted basalt and its overlying veneer of sediment interacting with the mantle wedge. The new geochemical evidence thus supports previous experimental studies that suggested a mantle origin for andesites in the TMVB (Blatter & Carmichael, 2001
; Carmichael, 2002). Nonetheless, the new data further suggest that a slab-derived siliceous melt interacting with the mantle wedge is probably a more effective mechanism to form high-SiO2 andesites than water-saturated partial melting of the upper mantle. The recognition that slab melting plays an important role in the Mexican arc, together with the conspicuous presence all across Mexico of ‘intraplate’ magmas with enriched (low Nd, high Sr, Pb) isotope ratios, yields a new perspective on the chemical character of the Mexican mantle wedge. Indeed, the results presented here indicate that the HFSE do not remain immobile during slab melting, and thus their concentrations and ratios in subduction-related lavas do not necessarily resemble those of the mantle source. Instead, the depleted mantle signatures (i.e. the low Nb/Ta ratios), together with strong subduction signals and high SiO2 contents, could reflect the inherited character of the slab-melt component itself.
The models provided here also support results from experimental petrology (Johnson & Plank, 1999; Poli & Schmidt, 2002) indicating that the composition of subducted slabs changes during their journey into the mantle, and that they undergo intense modifications of their water content and fluid-mobile element budget as a result of continuous fluid release. In turn, we can distinguish the effects of different mechanisms of element recycling (hydrous fluids vs silicate melts) in the geochemistry of arc magmas because they imprint different geochemical characteristics on the volcanic products. Furthermore, because these differences should be ultimately linked to variations in P–T gradients, they provide an insight into the thermal and tectonic structure of a subduction zone. For instance, the low Nb/Ta ratios of the VBZ-LRE rocks indicate that slab melting probably occurred in the presence of amphibole, independently constraining slab depth to 75 km, equivalent to < 25 kbar, because amphibole is unstable at higher pressures (Schmidt & Poli, 1998). If true, then it is possible that the Cocos slab remains flat for a longer distance, and melts at a significantly shallower depth than the hypothetical 100 km below the central part of the arc (Pardo & Suárez, 1995). A detailed seismic tomography of central Mexico would be capable of testing this hypothesis.
Slab-melt–peridotite reactions also appear to be viable mechanisms for the formation of accessory mineral assemblages in the mantle wedge. As silicate melts are about 10 times more effective as metasomatic agents to form hydrous minerals than aqueous fluids (Schneider & Eggler, 1986), it does not seem surprising that the only reported locality in Mexico with amphibole-bearing lherzolites is the VBZ (Blatter & Carmichael, 1998a), and that their host is a VBZ-HRE rock (our sample ZIT-99-10). The presence of residual accessory phases such as amphibole, phlogopite or even rutile not only can selectively imprint unusual geochemical signals on the arc magmas, but could also represent important carriers of key trace elements in the convecting mantle to the temperatures and pressures at which they remain stable.
|
ACKNOWLEDGEMENTS |
|---|
A.G.-T. thanks Alex LaGatta for fruitful discussions on the petrogenesis of the TMVB. Invaluable help during sample preparation and analysis was provided by Rick Mortlock, Marty Fleisher, Jean Hanley and Rufino Lozano. We thank Marlina Elburg and Colin MacPherson for reviews that led to significant improvements in the manuscript. Editorial handling by Professor John Gamble is also highly appreciated. This work was funded by CONACyT grants 39785 to A.G.-T. and 27642-T to F.O.-G.; and by the NSF grant EAR 96-14782 to C.H.L. and S.L.G.
*Corresponding author. Telephone: (5255) 5623-4104 ext. 120. Fax: (5255) 5623-4101. E-mail: tuena{at}geociencias.unam.mx
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONTECTONIC AND GEOLOGICAL SETTINGSAMPLES AND ANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONS AND IMPLICATIONSREFERENCES |
|---|
Adam J, Green TH, Sie S, Ryan C. (1997) Trace element partitioning between aqueous fluids, silicate melts and minerals. European Journal of Mineralogy 9:569–584.
Atherton M and Petford N. (1993) Generation of sodium-rich magmas from newly underplated basaltic crust. Nature 362:144–146.[CrossRef]
Ayers J, Dittmer S, Layne G. (1997) Partitioning of elements between silicate melt and H2O–NaCl fluids at 1·5 and 2·0 GPa pressure; implications for mantle metasomatism. Geochimica et Cosmochimica Acta 59:4237–4246.
Barth M, Foley S, Horn I. (2002) Partial melting in Archean subduction zones: constraints from experimentally determined trace element partition coefficients between eclogitic minerals and tonalitic melts under upper mantle conditions. Precambrian Research 113:323–340.[CrossRef][Web of Science]
Blatter D and Carmichael I. (1998a) Hornblende peridotite xenoliths from central Mexico reveal the highly oxidized nature of subarc upper mantle. Geology 26:1035–1038.
Blatter D and Carmichael I. (1998b) Plagioclase-free andesites from Zitácuaro (Michoacán), México: petrology and experimental constraints. Contributions to Mineralogy and Petrology 132:121–138.[CrossRef][Web of Science]
Blatter D and Carmichael I. (2001) Hydrous phase equilibria of a Mexican high-silica andesite: a candidate for mantle origin? Geochimica et Cosmochimica Acta 65:4043–4065.[CrossRef][Web of Science]
Blatter D, Carmichael I, Deino A, Renne P. (2001) Neogene volcanism at the front of the central Mexican volcanic belt: basaltic andesites to dacites, with contemporaneous shoshonites and high-TiO2 lava. Geological Society of America Bulletin 113:1324–1342.
Bodinier J, Merlet C, Bedini R, Simien F, Remaidi M, Garrido C. (1995) Distribution of niobium, tantalum, and other highly incompatible trace elements in the lithospheric mantle: The spinel paradox. Geochimica et Cosmochimica Acta 60:545–550.
Bourdon B, Eissen J, Monzier M, Robin C, Martin H, Cotten J, Hall M. (2002) Adakite-like lavas from Antisana Volcano (Ecuador): evidence for slab melt metasomatism beneath the Andean Northern Volcanic Zone. Journal of Petrology 43:199–217.
Brenan J, Shaw H, Phinney D, Ryerson F. (1994) Rutile–aqueous fluid partitioning of Nb, Ta, Hf, Zr, U and Th: implications for high field strength element depletions in island-arc basalts. Earth and Planetary Science Letters 128:327–339.[CrossRef][Web of Science]
Brenan J, Shaw H, Ryerson F. (1995a) Experimental evidence for the origin of lead enrichments in convergent-margin magmas. Nature 378:54–56.[CrossRef]
Brenan J, Shaw H, Ryerson F, Phinney D. (1995b) Experimental determination of trace-element partitioning between pargasite and a synthetic hydrous andesitic melt. Earth and Planetary Science Letters 135:1–11.[CrossRef][Web of Science]
Brenan J, Shaw H, Ryerson F, Phinney D. (1995c) Mineral–aqueous fluid partitioning of trace elements at 900°C and 2· 0 GPa: constraints on the trace element chemistry of mantle and deep crustal fluids. Geochimica et Cosmochimica Acta 59:3331–3350.[CrossRef][Web of Science]
Capra L, Macías J, Garduño V. (1997) The Zitácuaro Volcanic Complex, Michoacán, México: magmatic and eruptive history of a resurgent caldera. Geofísica Internacional 36:161–179.
Carmichael I. (2002) The andesite aqueduct: perspectives on the evolution of intermediate magmatism in west–central (105°–99°W) Mexico. Contributions to Mineralogy and Petrology 143:641–663.[Web of Science]
Castillo P, Janney P, Solidum R. (1999) Petrology and geochemistry of Camiguin Island, southern Philippines: insights to the source of adakites and other lavas in a complex arc setting. Contributions to Mineralogy and Petrology 134:33–51.[CrossRef][Web of Science]
Chesley J, Ruiz J, Righter K, Ferrari L, Gómez-Tuena A. (2002) Source contamination versus assimilation: an example from the Trans-Mexican Volcanic Arc. Earth and Planetary Science Letters 195:211–221.[CrossRef][Web of Science]
Class C, Miller DM, Goldstein SL, Langmuir CH. (2000) Distinguishing melt and fluid subduction components in Umnak Volcanics, Aleutian Arc. Geochemistry, Geophysics, Geosystems 1: doi:10.1029/1999GC000010.
Conrey R, Hooper P, Larson P, Chesley J, Ruiz J. (2001) Trace element and isotopic evidence for two types of crustal melting beneath a High Cascade volcanic center, Mt. Jefferson, Oregon. Contributions to Mineralogy and Petrology 141:710–732.[Web of Science]
Defant M and Drummond M. (1990) Derivation of some modern arc magmas by melting of young subducted lithosphere. Nature 347:662–665.[CrossRef]
DePaolo D. (1981) Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization. Earth and Planetary Science Letters 53:189–202.[CrossRef][Web of Science]
Donnelly K. (2002) The genesis of E-MORB: extensions and limitations of the hot spot model. Ph.D. thesis. (Columbia University, New York).
Drummond M and Defant M. (1990) A model for trondhjemite–tonalite–dacite genesis and crustal growth via slab melting: Archean to modern comparisons. Journal of Geophysical Research 95:21505–21521.
Elburg M, Van Bergen M, Hoogewerff J, Foden J, Vroon P, Zulkarnain I, Nasution A. (2002) Geochemical trends across an arc–continent collision zone: magma sources and slab-wedge transfer processes below the Pantar Strait volcanoes, Indonesia. Geochimica et Cosmochimica Acta 66:2771–2789.[CrossRef][Web of Science]
Elliott T, Plank T, Zindler A, White W, Bourdon B. (1997) Element transport from slab to volcanic front at the Mariana arc. Journal of Geophysical Research 102:14991–15019.[CrossRef]
Ferrari L. (2004) Slab detachment control on mafic volcanic pulse and mantle heterogeneity in central Mexico. Geology 32:77–80.
Foley S, Barth M, Jenner G. (2000) Rutile/melt partition coefficients for trace elements and an assessment of the influence of rutile on the trace element characteristics of subduction zone magmas. Geochimica et Cosmochimica Acta 64:933–938.[CrossRef][Web of Science]
Foley S, Tiepolo M, Vannucci R. (2002) Growth of the early continental crust controlled by melting of amphibolite in subduction zones. Nature 417:837–840.[CrossRef]
Garrison J and Davidson J. (2003) Dubious case for slab melting in the Northern volcanic zone of the Andes. Geology 31:565–568.
Gill J. (1981) Orogenic Andesites and Plate Tectonics(Springer, Berlin).
Gómez-Tuena A, LaGatta A, Langmuir C, Goldstein S, Ortega-Gutiérrez F, Carrasco-Núñez G. (2003) Temporal control of subduction magmatism in the Eastern Trans-Mexican Volcanic Belt: mantle sources, slab contributions and crustal contamination. Geochemistry, Geophysics, Geosystems 4: doi:10.1029/2003GC000524.
Green T and Pearson N. (1987) An experimental study of Nb and Ta partitioning between Ti-rich minerals and silicate liquids at high pressure and temperature. Geochimica et Cosmochimica Acta 51:55–62.[CrossRef][Web of Science]
Green TH. (1995) Significance of Nb/Ta as an indicator of geochemical processes in the crust mantle system. Chemical Geology 120:347–359.[CrossRef][Web of Science]
Green TH and Pearson N. (1986) Ti-rich accessory phase saturation in hydrous mafic–felsic compositions at high P, T. Chemical Geology 54:185–201.[CrossRef][Web of Science]
Hart S and Dunn T. (1993) Experimental cpx/melt partitioning of 24 trace elements. Contributions to Mineralogy and Petrology 113:1–8.[CrossRef][Web of Science]
Hildreth W and Moorbath S. (1988) Crustal contributions to arc magmatism in the Andes of central Chile. Contributions to Mineralogy and Petrology 98:455–489.[CrossRef][Web of Science]
Hirose K. (1997) Melting experiments on lherzolite KLB-1 under hydrous conditions and generation of high-magnesian andesites. Geology 25:42–44.
Hochstaedter A, Gill J, Peters R, Broughton P, Holden P. (2001) Across-arc geochemical trends in the Izu–Bonin arc: contributions from the subducting slab. Geochemistry, Geophysics, Geosystems 2: doi:10.1029/2000GC000105.
Hou Z, Gao Y, Qu X, Rui Z, Mo X. (2004) Origin of adakitic intrusives generated during mid-Miocene east–west extension in southern Tibet. Earth and Planetary Science Letters 220:139–155.[CrossRef][Web of Science]
Ionov D, Grégoire M, Prikhod’ko V. (1999) Feldspar–Ti-oxide metasomatism in off-cratonic continental and oceanic upper mantle. Earth and Planetary Science Letters 165:37–44.[CrossRef][Web of Science]
Irvine T and Baragar W. (1971) A guide to the chemical classification of the common volcanic rocks. Canadian Journal of Earth Sciences 8:523–548.
Johnson K. (1994) Experimental cpx/and garnet/melt partitioning of REE and other trace elements at high pressures; petrogenetic implications. Mineralogical Magazine 58A:454–455.
Johnson M and Plank T. (1999) Dehydration and melting experiments constrain the fate of subducted sediments. Geochemistry, Geophysics, Geosystems 1: doi:10.1029/1999GC000014.
Kalfoun F, Ionov D, Merlet C. (2002) HFSE residence and Nb/Ta ratios in metasomatised, rutile-bearing mantle peridotites. Earth and Planetary Science Letters 199:49–65.[CrossRef][Web of Science]
Kay R. (1978) Aleutian magnesian andesites: melts from subduction Pacific Oceanic crust. Journal of Volcanology and Geothermal Research 4:117–132.[CrossRef][Web of Science]
Kelemen P, Hanghøj K, Greene A. (2003) One view of the geochemistry of subduction-related magmatic arcs, with emphasis on primitive andesite and lower crust. In Rudnick R (Ed.). Treatise on Geochemistry(Elsevier, New York) pp. 593–659.
Kelemen P, Yogodzinski G, Scholl D. Eiler J. (2004) Along-strike variation in lavas of the Aleutian Island Arc: implications for the genesis of high Mg-number andesite and the continental crust. Inside the Subduction Factory. Geophysical Monograph, American Geophysical Union 138:223–276.
Kelemen PB. (1998) Silica enrichment in the continental upper mantle via melt/rock reaction. Earth and Planetary Science Letters 164:387–406.[CrossRef][Web of Science]
Kelemen PB, Shimizu N, Dunn T. (1993) Relative depletion of niobium in some arc magmas and the continental crust: partitioning of K, Nb, La and Ce during melt/rock interaction in the upper mantle. Earth and Planetary Science Letters 120:111–134.[CrossRef][Web of Science]
Keppler H. (1996) Constraints from partitioning experiments on the composition of subduction-zone fluids. Nature 380:237–240.[CrossRef]
Kessel R, Schmidt M, Ulmer P, Pettke T. (2005) Trace element signature of subduction-zone fluids, melts and supercritical liquids at 120–180 km depth. Nature 437:724–727.[CrossRef][Medline]
LaGatta A. (2003) Arc magma genesis in the eastern Mexican volcanic belt. Ph.D. thesis(Columbia University, New York).
Lassiter J and Luhr J. (2001) Osmium abundance and isotope variations in mafic Mexican volcanic rocks: evidence for crustal contamination and constraints on the geochemical behavior of osmium during partial melting and fractional crystallization. Geochemistry, Geophysics, Geosystems 2: doi:10.1029/2000GC000116.
Lehnert K, Su Y, Langmuir C, Sarbas B, Nohl U. (2000) A global geochemical database structure for rocks. Geochemistry, Geophysics, Geosystems 1: doi:10.1029/1999GC000026.
Le Maitre R. (1989) A Classification of Igneous Rocks and Glossary of Terms(Blackwell, Oxford).
Lozano R, Verma S, Giron P, Velasco F, Morán-Zenteno D, Viera F, Chávez G. (1995) Calibración preliminar de fluorescencia de rayos X para análisis cuantitativo de elementos mayores en rocas ígneas. Actas INAGEQ 1:203–208.
Luhr J. (2000) The geology and petrology of Volcan San Juan Nayarit, Mexico and the compositionally zoned Tepic Pumice. Journal of Volcanology and Geothermal Research 95:109–156.[CrossRef][Web of Science]
Luhr J and Carmichael I. (1985a) Contemporaneous eruptions of calc-alkaline and alkaline magmas along the volcanic front of the Mexican Volcanic Belt. Geofísica Internacional 24:203–216.
Luhr J and Carmichael I. (1985b) Jorullo Volcano, Michoacán, Mexico (1759–1774): the earliest stages of fractionation in calc-alkaline magmas. Contributions to Mineralogy and Petrology 90:142–161.[CrossRef][Web of Science]
Luhr J, Allan J, Carmichael I, Nelson S, Hasenaka T. (1989) Primitive calc-alkaline and alkaline rock types from the western Mexican volcanic belt. Journal of Geophysical Research 94:4515–4530.
MacPherson C, Dreher S, Thirlwall MF. (2006) Adakites without slab melting: high pressure differentiation of island arc magma, Mindanao, the Philippines. Earth and Planetary Science Letters 243:581–593.[CrossRef][Web of Science]
Manea M, Manea V, Kostoglodov V. (2003) Sediment fill in the Middle America Trench inferred from gravity anomalies. Geofísica Internacional 42:603–612.
Manea V, Manea M, Kostoglodov V, Currie C, Sewell G. (2004) Thermal structure, coupling and metamorphism in the Mexican subduction zone beneath Guerrero. Geophysical Journal International 158:775–784.[CrossRef][Web of Science]
Manea V, Manea M, Kostoglodov V, Sewell G. (2005) Thermo-mechanical model of the mantle wedge in Central Mexican subduction zone and a blob tracing approach for the magma transport. Physics of the Earth and Planetary Interiors 149:165–186.[CrossRef][Web of Science]
Márquez A, Oyarzún R, Doblas M, Verma S. (1999) Alkalic (oceanic-island basalt type) and calc-alkalic volcanism in the Mexican volcanic belt: a case for plume-related magmatism and propagating rifting at an active margin? Geology 27:51–54.
Martin H. (1999) The adakitic magmas: modern analogues of Archaean granitoids. Lithos 46:411–429.[CrossRef][Web of Science]
Martínez-Serrano R, Schaaf P, Solís-Pichardo G, Hernández-Bernal M, Hernández-Treviño T, Morales-Contreras J, Macías J. (2004) Sr, Nd and Pb isotope and geochemical data from the Quaternary Nevado de Toluca volcano, a source of recent adakitic magmatism, and the Tenango Volcanic Field, Mexico. Journal of Volcanology and Geothermal Research 138:77–110.[CrossRef][Web of Science]
McCulloch M and Gamble J. (1991) Geochemical and geodynamical constraints on subduction zone magmatism. Earth and Planetary Science Letters 102:358–374.[CrossRef][Web of Science]
Moore G and Shipley T. (1988) Mechanisms of sediment accretion in the Middle America Trench. Journal of Geophysical Research 93:8911–8927.
Moore J, Watkins J, Bachman S, Beghtel F, Butt A, Didyk B, Foss G, Leggett J, Lundberg N, McMillan N, Niitsuma N, Shephard L, Stephen J, Shipley T, Strander H. (1982) Facies belts of the Middle America Trench and forearc region, southern Mexico: results from Leg 66 DSDP. In Legget JK (Ed.). Trench–Forearc Geology: Sedimentation and Tectonics on Modern and Ancient Active Plate Margins(Special Publications, Geological Society, London) 10: pp. 77–94.
Müntener O, Kelemen P, Grove T. (2001) The role of H2O during crystallization of primitive arc magmas under uppermost mantle conditions and genesis of igneous pyroxenites: an experimental study. Contributions to Mineralogy and Petrology 141:643–658.[Web of Science]
Pardo M and Suárez G. (1995) Shape of the subducted Rivera and Cocos plate in southern Mexico: seismic and tectonic implications. Journal of Geophysical Research 100:12357–12373.[CrossRef]
Pearce J, Stern C, Bloomer S, Fryer P. (2005) Geochemical mapping of the Mariana arc–basin system: implications for the nature and distribution of subduction components. Geochemistry, Geophysics, Geosystems 6: doi:10.1029/2004GC000895.
Petrone C, Francalanci L, Carlson R, Ferrari L, Conticelli S. (2003) Unusual coexistence of subduction-related and intraplate-type magmatism: Sr, Nd and Pb isotope and trace element data from the magmatism of the San Pedro–Ceboruco graben (Nayarit, Mexico). Chemical Geology 193:1–24.[CrossRef][Web of Science]
Pier J, Podosek F, Luhr J, Brannon J, Aranda-Gómez J. (1989) Spinel-lherzolite-bearing Quaternary volcanic centers in San Luis Potosi, Mexico 2. Sr and Nd isotopic systematics. Journal of Geophysical Research 94:7941–7951.
Plank T and Langmuir C. (1998) The chemical composition of subducting sediment and its consequences for the crust and mantle. Chemical Geology 145:325–394.[CrossRef][Web of Science]
Poli S and Schmidt M. (2002) Petrology of subducted slabs. Annual Review of Earth and Planetary Sciences 30:207–235.[CrossRef][Web of Science]
Prouteau G, Scaillet B, Pichavant M, Maury R. (1999) Fluid-present melting of oceanic crust in subduction zones. Geology 27:1111–1114.
Prouteau G, Scaillet B, Pichavant M, Maury R. (2001) Evidence for mantle metasomatism by hydrous silicic melts derived from subducted oceanic crust. Nature 410:197–200.[CrossRef]
Rapp R. (1995) The amphibole-out phase boundary in partially melted metabasalt, and its control over melt fraction and composition, and source permeability. Journal of Geophysical Research 100:15601–15610.[CrossRef]
Rapp R and Watson E. (1995) Dehydration melting of metabasalt at 8–32 kbar: implications for continental growth and crust–mantle recycling. Journal of Petrology 36:891–931.
Rapp R, Shimizu N, Norman M, Applegate G. (1999) Reaction between slab-derived melts and peridotite in the mantle wedge: experimental constraints at 3·8 GPa. Chemical Geology 160:335–256.[CrossRef][Web of Science]
Rapp R, Shimizu N, Norman M. (2003) Growth of early continental crust by partial melting of eclogite. Nature 425:605–609.[CrossRef][Medline]
Rollinson H. (1993) Using Geochemical Data(Longman, Harlow).
Ruiz J, Patchett P, Arculus R. (1988a) Nd–Sr isotope constraints of lower crustal xenoliths—evidence for the origin of mid-Tertiary felsic volcanics in Mexico. Contributions to Mineralogy and Petrology 99:36–43.[CrossRef][Web of Science]
Ruiz J, Patchett P, Ortega-Gutiérrez F. (1988b) Proterozoic and Phanerozoic basement terranes of Mexico from Nd isotopic studies. Geological Society of America Bulletin 100:274–281.
Ryerson F and Watson E. (1987) Rutile saturation in magmas—implications for Ti–Nb–Ta depletion in island-arc basalts. Earth and Planetary Science Letters 86:225–239.[CrossRef][Web of Science]
Salters V and Longhi J. (1999) Trace element partitioning during the initial stages of melting beneath mid-ocean ridges. Earth and Planetary Science Letters 166:15–30.[CrossRef][Web of Science]
Schaaf P, Heinrich W, Besch T. (1994) Composition and Sm–Nd isotopic data of the lower crust beneath San Luis Potosí, central Mexico: evidence from a granulite-facies xenolith suite. Chemical Geology 118:63–84.[CrossRef][Web of Science]
Schaaf P, Stimac J, Siebe C, Macías J. (2005) Geochemical evidence for mantle origin and crustal processes in volcanic rocks from Popocatépetl and surrounding monogenetic volcanoes, central Mexico. Journal of Petrology 46:1243–1282.
Schiano P, Clocchiatti R, Shimizu N, Weis D, Mattielli N. (1994) Cogenetic silica-rich and carbonate-rich melts trapped in mantle minerals in Kerguelen ultramafic xenoliths: implications for metasomatism in the oceanic upper mantle. Earth and Planetary Science Letters 123:167–178.[CrossRef][Web of Science]
Schiano P, Clocchiatti R, Shimizu N, Maury R, Jochum K, Hoffman A. (1995) Hydrous, silica-rich melts in the sub-arc mantle and their relationship with erupted arc lavas. Nature 377:595–600.[CrossRef]
Schmidt M and Poli S. (1998) Experimentally based water budgets for dehydrating slabs and consequences for arc magma generation. Earth and Planetary Science Letters 163:361–379.[CrossRef][Web of Science]
Schmidt M, Dardon A, Chazot G, Vannucci R. (2004) The dependence of Nb and Ta rutile–melt partitioning on melt composition and Nb/Ta fractionation during subduction processes. Earth and Planetary Science Letters 226:415–432.[CrossRef][Web of Science]
Schneider M and Eggler D. (1986) Fluids in equilibrium with peridotite minerals: implications for mantle metasomatism. Geochimica et Cosmochimica Acta 50:711–724.[CrossRef][Web of Science]
Sen C and Dunn T. (1994) Dehydration melting of a basaltic composition amphibolite at 1·5 and 2·0 GPa: implications for the origin of adakites. Contributions to Mineralogy and Petrology 117:394–409.[CrossRef][Web of Science]
Siebe C, Rodríguez-Lara V, Schaaf P, Abrams M. (2004) Geochemistry, Sr–Nd isotope composition, and tectonic setting of Holocene Pelado, Guespalapa and Chichinautzin scoria cones, south of Mexico City. Journal of Volcanology and Geothermal Research 2712:1–30.
Smith D and Leeman W. (1996) The origin of Mount St. Helens andesites. Journal of Volcanology and Geothermal Research 55:271–303.[CrossRef]
Stadler R, Foley S, Brey G, Horn I. (1998) Mineral–aqueous fluid partitioning of trace elements at 900–1200°C and 3· 0–5·7 GPa: new experimental data for garnet, clinopyroxene, and rutile, and implications for mantle metasomatism. Geochimica et Cosmochimica Acta 62:1781–1801.[CrossRef][Web of Science]
Stern C and Kilian R. (1996) Role of the subducted slab, mantle wedge and continental crust in the generation of adakites from the Andean Austral volcanic vone. Contributions to Mineralogy and Petrology 123:263–281.[CrossRef][Web of Science]
Stolz A, Jochum K, Spettel B, Hofmann AW. (1996) Fluid- and melt-related enrichment in the subarc mantle: evidence from Nb/Ta variation in arc basalts. Geology 24:587–590.
Straub S and Martin-Del Pozzo A. (2001) The significance of phenocryst diversity in tephra from recent eruptions at Popocatépetl Stratovolcano (central Mexico). Contributions to Mineralogy and Petrology 140:487–510.[CrossRef][Web of Science]
Straub S, Layne G, Schmidt A, Langmuir C. (2004) Volcanic glasses at the Izu arc volcanic front: new perspectives on fluid and sediment melt recycling in subduction zones. Geochemistry, Geophysics, Geosystems 5: doi:10.1029/2002GC000408.
Sun S and McDonough W. (1989) Chemical and isotopic systematics of oceanic basalts: implications for mantle compositions and processes. In Saunders AD and Norry MJ (Eds.). Magmatism in the Ocean Basins(Special Publications, Geological Society, London) 42: pp. 313–345.
Tatsumi Y. (2001) Geochemical modeling of partial melting of subducting sediments and subsequent melt–mantle interaction: generation of high-Mg andesites in the Setouchi volcanic belt, southwest Japan. Geology 29:323–326.
Tatsumi Y and Hanyu T. (2003) Geochemical modeling of dehydration and partial melting of subducting lithosphere: toward a comprehensive understanding of high-Mg andesite formation in the Setouchi volcanic belt, SW Japan. Geochemistry, Geophysics, Geosystems 4: doi:10.1029/2003GC000530.
Tiepolo M, Vannucci R, Oberti R, Foley S, Bottazzi P, Zanetti A. (2000) Nb and Ta incorporation and fractionation in titanian pargasite and kaersutite: crystal-chemical constraints and implications for natural systems. Earth and Planetary Science Letters 176:185–201.[CrossRef][Web of Science]
Todt W, Cliff R, Hanser A, Hofmann AW. (1996) Evaluation of a 202Pb–205Pb double spike for high-precision lead isotope analysis. In Basu A and Hart S (Eds.). Earth Processes: Reading the Isotopic Code. Geophysical Monograph, American Geophysical Union 95: pp. 429–437.
van Westrenen W, Blundy J, Wood B. (2001) High field strength element/rare earth fractionation during partial melting in the presence of garnet: implications for identifications of mantle heterogeneities. Geochemistry, Geophysics, Geosystems 2: doi:10.1029/2000GC000133.
Verma S. (1999) Geochemistry of evolved magmas and their relationship to subduction-unrelated mafic volcanism at the volcanic front of the central Mexican Volcanic Belt. Journal of Volcanology and Geothermal Research 93:151–171.[CrossRef][Web of Science]
Verma S. (2000a) Geochemical evidence for a lithospheric source for magmas from Los Humeros caldera, Puebla, Mexico. Chemical Geology 164:35–60.[CrossRef][Web of Science]
Verma S. (2000b) Geochemistry of the subducting Cocos plate and the origin of subduction-unrelated mafic volcanism at the front of the central Mexican Volcanic Belt. In Delgado-Granados H, Aguirre-Díaz G, Stock J (Eds.). Cenozoic Tectonics and Volcanism of Mexico Special Papers Geological Society of America 334: pp. 1–28.
Verma S. (2002) Absence of Cocos plate subduction-related basic volcanism in southern Mexico: a unique case on Earth? Geology 30:1095–1098.
Wallace P and Carmichael I. (1999) Quaternary volcanism near the Valley of Mexico: implications for subduction zone magmatism and the effects of crustal thickness variations on primitive magma compositions. Contributions to Mineralogy and Petrology 135:291–314.[CrossRef][Web of Science]
Wang Q, McDermott F, Xu J, Bellon H, Zhu Y. (2005) Cenozoic K-rich adakitic volcanic rocks in the Hohxil area, northern Tibet: lower-crustal melting in an intracontinental setting. Geology 33:465–468.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E. R. Johnson, P. J. Wallace, H. Delgado Granados, V. C. Manea, A. J. R. Kent, I. N. Bindeman, and C. S. Donegan Subduction-related Volatile Recycling and Magma Generation beneath Central Mexico: Insights from Melt Inclusions, Oxygen Isotopes and Geodynamic Models J. Petrology, September 1, 2009; 50(9): 1729 - 1764. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Vigouroux, P. J. Wallace, and A. J. R. Kent Volatiles in High-K Magmas from the Western Trans-Mexican Volcanic Belt: Evidence for Fluid Fluxing and Extreme Enrichment of the Mantle Wedge by Subduction Processes J. Petrology, September 1, 2008; 49(9): 1589 - 1618. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Gomez-Tuena, L. Mori, N. E. Rincon-Herrera, F. Ortega-Gutierrez, J. Sole, and A. Iriondo The origin of a primitive trondhjemite from the Trans-Mexican Volcanic Belt and its implications for the construction of a modern continental arc Geology, June 1, 2008; 36(6): 471 - 474. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Straub Uniform processes of melt differentiation in the central Izu Bonin volcanic arc (NW Pacific) Geological Society, London, Special Publications, January 1, 2008; 304(1): 261 - 283. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Marchesi, C. J. Garrido, D. Bosch, J. A. Proenza, F. Gervilla, P. Monie, and A. Rodriguez-Vega Geochemistry of Cretaceous Magmatism in Eastern Cuba: Recycling of North American Continental Sediments and Implications for Subduction Polarity in the Greater Antilles Paleo-arc J. Petrology, September 1, 2007; 48(9): 1813 - 1840. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||