Mineralogical and geochemical evidence of magma mingling/mixing in the Sierra de las Cruces volcanic range, Mexican Volcanic Belt

Pliocene – Pleistocene lava flows, mainly of dacitic composition, are exposed in the Sierra de las Cruces (SC) volcanic range within the Mexican Volcanic Belt (MVB). SC volcanic rocks are porphyritic, generally containing an assemblage of plagioclase + amphibole + orthopyroxene ± clinopyroxene ± quartz ± Fe-Ti oxides. Most of them exhibit diverse mineralogical and geochemical features that attest a magma mixing and mingling processes with concomitant fractional crystallization in which a small volume of hot andesite magma injects into dacitic magma. Both rock types are probably derived from partial melting of continental crust at different levels. The evidences of magma mixing and mingling include: (a) normal and sieved plagioclases in the same sample, rounded and embayed crystals, and armoured rims over the dissolved crystal surfaces; (b) subrounded, vesicular magmatic enclaves, ranging from a few millimeters to ~20 centimeters in size, with plagioclase + orthopyroxene + amphibole + quartz ± olivine ± Fe-Ti-oxides assemblage; (c) mineral chemistry evidence such as crystals with reaction rims or heterogeneous plagioclase compositions (inverse and oscillatory zoning or normally and inversely zoned crystals) in the same sample; and (d) elemental geochemical variations and trace-element ratio more akin to magma mixing and to some extent diffusion process. These andesitic enclaves could be considered as portions of the intermediate magma that did not mix completely (mingling) with the felsic host lavas, confirming the major role of magma mixing and mingling processes in the overall evolution of the MVB 148 Velasco-Tapia et al. / Journal of Iberian Geology 39 (1) 2013: 147-166


Introduction
The Mexican Volcanic Belt (MVB; Fig. 1) is a geologic province of ~8000 volcanic centers (stratovolcanoes, cinder cones, maars, etc.) of Miocene to present-day, predominantly having an andesitic to dacitic composition.The MVB is ~1000 km long and 50-300 km wide, extending approximately east-west from Veracruz to Puerto Vallarta (Gómez-Tuena et al., 2007).The MVB origin has generally been viewed as related to the subduction of the Cocos and Rivera plates subduction beneath the North American plate (Pardo and Suárez, 1995;Ferrari et al., 1999).However, some geological, geochemical and geophysical observations do not completely agree with a unique subduction-related origin for the MVB, as summarized by Márquez et al. (1999a,b) and Sheth et al. (2000).Alternative models have been proposed to explain the MVB origin and these include: (a) hybrid models involving both mantle and slab sources (Luhr, 1997;Wallace and Carmichael, 1999;Ferrari, 2004); (b) the activity of a west-east propagating mantle plume coexisting with the subduction of the Cocos and Rivera plates (Márquez et al., 1999a); and (c) the ongoing extensional processes along the MVB (Sheth et al., 2000;Verma, 2002Verma, , 2009)).
On the other hand, several geochemical and isotopic studies reveal the crust participation in the genesis of the MVB intermediate and silica-rich magmas.Besch et al. (1995) determined that an assimilation of lower crust can explain the chemical and isotopic differences between primitive mafic and evolved rocks in the eastern MVB.Verma (1999) reported that andesites and dacites from the Sierra de Chichinautzin volcanic field (central MVB) were derived from partial melting of a heterogeneous mafic granulite from the lower crust.Later, Márquez and De Ignacio (2002) suggested that the magmatic diversity of Sierra de Chichinautzin can be explained by magma mixing between mafic magmas derived from a heterogeneous upper mantle and two different felsic crustal magmas.Chelsey et al. (2002), based on geochemical and isotopic data (particularly Os isotopic ratios), reported that the Michoacan-Guanajuato volcanism (central MVB) shows assimilation of the lower crust.Magma mixing processes, between magmas of different composition, have also been inferred from the phenocryst assemblages and geochemistry of the Iztaccíhuatl (Nixon, 1988a,b),Tequila (Wallace and Carmichael, 1994), Tancítaro (Ownby et al., 2011), and Popocatépetl volcanoes (Straub and Martin-Del Pozzo, 2001;Schaaf et al., 2005;Witter et al., 2005;Sosa-Ceballos et al., 2012), as well as in Amealco caldera (Aguirre-Díaz, 2001).
Therefore, it is clear that detailed studies involving mineralogical, geochemical, and isotopic data are required to explain the magmatic diversity observed in the MVB and to try to solve the controversies related to this significant geologic province.This paper describes an initial petrogenetic evaluation of magmatism in the Sierra de las Cruces (SC) volcanic range, based on mineralogical, geochemical, and Sr-Nd isotopic ratios of intermediate and felsic volcanic rocks.The lithological features of SC volcanic range represent an opportunity to test the
Los enclaves andesíticos podrían considerarse como porciones de magma intermedio que no se mezcló con los líquidos dacíticos receptores, lo que confirma la importancia de los procesos de mezcla incompleta de magmas en la evolución magmática del CVM.role of the continental crust in the origin and evolution of magmas in the central MVB.

Regional stratigraphy
Vázquez-Sánchez and Jaimes-Palomera (1989) and García-Palomo et al. (2002) have reported detailed stratigraphic information for the SC surrounding regions (Mexico and Toluca basins), covering the Cretaceous to present-day period.Stratigraphic record begins with a calcareous sedimentary sequence (marine to shelf facies) that was deposited in central Mexico during the Cretaceous (Fries, 1960).Rocks from this sequence include massive limestone with black chert lenses, beds of gypsum, massive to thickly bedded limestones, greywacke interbedded with limolite and shale beds.This Cretaceous sedimentary sequence (thickness ~ 3000 m) was folded and uplifted during the Laramide orogenic event (Fries, 1960) and was subsequently intruded by granitic or granodioritic dykes dated at 50 ± 10 Ma (De Cserna et al. 1974).The Eocene-Oligocene stratigraphy that overlies the Cretaceous sequence consists of calcareous conglomerates, lava flows, sandstones, volcanic siltstones, and lacustrine deposits, with a maximum thickness of 500 m.This sedimentary sequence is unconformably overlain by about 38 to 7.5 Ma rhyolite, rhyodacite, dacitic lava flows and pyroclastic flow deposits (Morán-Zenteno et al. 1998;García-Palomo et al. 2002), and by a Pliocene to Holocene volcanism of the MVB, that include the Las Cruces eruptive period (Delgado-Granados and Martin del Pozzo 1993).

Central MVB Pliocene to Holocene volcanism
The Pliocene to Holocene volcanism in the central MVB was generated during three eruptive periods (Delgado-Granados and Martin del Pozzo, 1993): Las Cruces, Ajusco, and Chichinautzin. in fault and fracture density in this direction (García-Palomo et al., 2008), a favorable condition for magma mingling/mixing processes.The most common enclaves are small, ranging from a few millimeters to 4 centimeters in diameter, although in the northern sector they can reach a maximum dimension of ~20 centimeters.Enclave contacts with their host dacite are generally sharp and crenulated, showing chilled margins.Their morphology and vesicularity of the ME indicate that they were partially molten before they were entrained into and quenched against relatively cooler host dacitic melt.On the other hand, lava flows are sometimes intercalated by pyroclastic deposits (Brecha Piroclástica Cantimplora after Delgado-Granados and Martin del Pozzo, 1983; total thickness = 1-4 m) conformed by gray dacitic blocks (20-30 cm), pumice clasts (<15 cm), and ashes.

Ajusco eruptive period
During this event, in the Middle Pleistocene, the Ajusco volcano was formed by the extrusion of several andesitic domes.Mora Alvarez et al. (1991) reported a K-Ar age of 0.39 ± 0.15 Ma for a sample from the Ajusco southern flank.

Sampling and analytical techniques
On the basis of the morphostructural classification by García-Palomo et al. (2008) and geochronological K-Ar data reported by Mora Alvarez et al.(1991) and Osete et al.(2000), we have divided the SC volcanic range in four sectors (Fig. 2): (a) Northern sector (SCN; 2.9 -3.7 Ma); (b) Central sector (SCC; 1.9 -2.9 Ma); (c) Southern sector (SCS; 0.7 -1.9 Ma), and (d) Las Cruces -Chichinautzin transition sector, which includes the Ajusco volcano (SCT; < 0.6 Ma).Fifty-two samples were collected across the range, being complemented by eight previously collected ones by Osete et al. (2000;Fig. 2 and Table 1 in Appendix A Supplementary Material).Most of the samples collected correspond to massive lava flows several meters thick, although some dykes and dome structures were also considered.Faulting in the area (e.g., García-Palomo et al., 2008) together with the dense vegetal cover, make difficult to clearly identify which of the proposed stratovolcanoes is the source of the sampled

Las Cruces eruptive period
It is considered as the oldest one, from Late Pliocene to Early Pleistocene (Fries, 1960;Schaelpfer, 1968;Sánchez-Rubio, 1984;Delgado-Granados and Martin del Pozzo, 1993).Mora-Alvarez et al. (1991) reported K-Ar dates from 2.87 to 1.92 Ma for this volcanism, whereas Osete et al. (2000), combining magnetostratigraphic and K-Ar radiometric data, established that this eruptive period can be best constrained between 3.6 and 1.8 Ma.The main mass of SC range (Fig. 2) was formed by consecutive episodes of faulting accompanied by the development of volcanic edifices, constructed by lavas and associated pyroclastic products (Mooser et al., 1974;Fries, 1960;Schlaepfer, 1968;Delgado-Granados and Martin del Pozzo, 1993;García-Palomo et al., 2002, 2008).The volcanic range is elongated, extending in a NNW-SSE direction for ~65 km, with a width varying from 47 km to the north and 27 km to the south.The SC constitutes the western margin of the Mexico basin (2220 m.a.s.l.) and the eastern border of the Toluca Valley (2400 m.a.s.l.), both being graben/horst depressions.According to García Palomo et al. (2008), the SC volcanic field has been divided in three sectors: northern, central, and southern, all of them bounded by E-W faults.Each sector is characterized by specific altitudes, slopes, trends of morpholineaments, and drainage patterns.Morphostructure of the northern and central sectors is controlled by N-S and NE-SW faults, whereas E-W faults have ruled the southern sector.An N-S spatial-temporal evolution of SC magmatic activity was confirmed by paleomagnetic and K-Ar data (Mora Alvarez et al., 1991;Osete et al., 2000;Fig. 2).The SC consists of eight overlapped stratovolcanoes, which are named, from south to north: Zempoala, La Corona, San Miguel, Salazar, Chimalpa, Iturbide, La Bufa, and La Catedral (García-Palomo et al., 2008).These volcanic edifices underwent alternated episodes of effusive and explosive activity, during which they were affected by faulting.Effusive products comprise gray andesitic to dacitic porphyritic lava flows (< 4 m thick) with planar fracturing sub parallel to the surface, labeled as Lava Dacítica Apilulco (total thickness ~400 m) by Delgado-Granados and Martin del Pozzo (1993).Occasionally, these lava flows contain magmatic enclaves (see Table 1 in Appendix A), which are dark gray on fresh surfaces and may appear lighter than the host on weathered surfaces.The enclaves comprise <2 to 10 vol% of the rock at a given outcrop.Their shape is commonly spherical to ellipsoidal and only rarely angular on 2D outcrop, with a finely crystalline and vesicular texture in hand specimen.They are randomly distributed over the volcanic range, although their number and size apparently increase towards the north.This fact could be related to an increase  Alvarez et al., 1991;Osete et al., 2000): (a) SCN-northern sector (2.9-3.7 Ma); (b) SCC -central sector (1.9-2.9Ma); (c) SCS -southern sector (0.7-1.9 Ma), and (d) SCT-transition sector that include the Ajusco volcano (< 0.7 Ma).Fig. 2.-Esquema geológico de la Sierra de las Cruces, mostrando litología, principales centros eruptivos y la distribución de muestras colectadas a lo largo de la sierra volcánica (modificado de Instituto Nacional de Estadística, Geografía e Informática (INEGI), 1980, Carta Geológica E14-2 1:250,000).La división del área de estudio en cuatro sectores de N a S se basa en datos radiométricos de K-Ar (estrellas) reportados en la literatura (Mora Alvarez et al., 1991;Osete et al., 2000): (a) SCN -sector norte (2.9-3.7 Ma); (b) SCC -sector central (1.9-2.9Ma); (c) SCS -sector sur (0.7-1.9 Ma) y (d) SCT -sector de transición que incluye el volcán Ajusco (< 0.7 Ma).
structures.In seven of the sampled outcrops ellipsoidal magmatic enclaves, up to several centimetres diameter, have been collected in the field.Modal compositions were determined by point counting on thin sections using a Prior Scientific petrographic microscope.Approximately 500 points per sample were counted in order to obtain a representative mode (Table 1 in Appendix A Supplementary Material).Mineral chemistry was analysed on thin sections of selected sam-ples, using the WDS JXA-8900 JEOL microprobe system of Centro de Microscopía Electrónica, Universidad Complutense (Madrid, Spain).Measurement conditions were 15 kV and 20 nA, with a beam diameter of ~1mm.The apparent concentrations were corrected for atomic number (Z), absorption (A), and fluorescence (F) effects, automatically applying the internal ZAFJEOL software.Microprobe system calibration was carried out using reference minerals from the Smithsonian Institution (Jarose-Trace element concentrations were determined by inductively coupled plasma-mass spectrometry (ICP-MS) with an analytical precision <10% and accuracy typically better than 7% for most elements at the 95% confidence level, based on analysis of diverse GRM.CIPW norms were calculated on a100% anhydrous adjusted basis of major elements, with Fe 2 O 3 /FeO ratios adjusted depending on the rock type (Middlemost, 1989).Rock classification was based on the total alkali-silica (TAS) scheme (LeBas et al., 1986;LeBas, 1989).All computations (anhydrous and iron-oxidation ratio adjustments, norm compositions, and rock classifications) were automatically done using the SINCLAS software (Verma et al., 2002(Verma et al., , 2003)).
Sr and Nd isotope analysis (Table 10 in Appendix A Supplementary Material) for eleven SC rocks were performed at Pacific Centre for Isotopic and Geochemical Research, University of British Columbia (Vancouver, Canada).Analyses were carried out on a Thermo Finnigan Triton thermal ionization mass spectrometer.De- wich et al., 1980).Analytical precision and detection limit data have been reported by Verma et al. (2009).Zoning in different mineral phases was studied by means of profile analysis and scanning probe microanalysis (SEM) images.Amphibole and pyroxene compositions were recalculated for Fe +3 using the charge balance method described by the equation of Droop (1987) equation.Representative analyses of the different mineral types are shown in Tables 2-7 (Appendix A Supplementary Material).
Additionally, FDE rocks exhibit the following distinctive characteristics: (a) diverse disequilibrium textures including sieved plagioclase, normal and sieved plagioclases in the same sample, rounded and embayed crystals, and reaction rims (Fig. 6A-D); (b) occurrence of vesicular, rounded or ellipsoidal andesitic magmatic enclaves (Fig. 6E-F); and (c) unusual mineral compositions, such as heterogeneous plagioclase core and rim compositions, and complex mineral zoning (reverse and oscillatory zoning in crystals or normally and reversely zoned crystals in the same sample).tailed analytical procedures were reported by Weiss et al. (2006).Repeated analysis of SRM987 Sr standard (n = 5) and La Jolla standard (n = 7) gave average values of 0.710248 ± 15 and 0.511854 ± 17.The analytical errors for 87 Sr/ 86 Sr and 143 Nd/ 144 Nd measured ratios are directly quoted for each sample.

Petrography and mineral chemistry
Petrographic and electron microprobe analysis reveal the existence of three mineralogical groups: (a) felsic rocks of dacitic composition (F); (b) intermediate and felsic rocks (andesite to dacite) with disequilibrium evidence (FDE); and (c) chilled magmatic enclaves with an andesitic composition (ME) hosted in some lavas with disequilibrium evidence, whose distinctive features are reported in the next paragraphs.

Felsic rocks with disequilibrium evidence (FDE)
Most of the collected SC rocks (~70% distributed along all sectors;  Plagioclase occurs as euhedral or subhedral phenocrysts (0.5-8 mm), showing Carlsbad or polysynthetic twinning.Crystals with clear and sieved texture can be observed in the same lava (Fig 6C ,D).Core compositions in clear texture plagioclase vary from An 30 to An 51 (Table 2 in Appendix A, and Fig. 3B).They show normal zoning with rims of An 32-43 .Sieved plagioclase shows disequilibrium or resorption rims, although occasionally sieved-cored crystals are also observed.In comparison with clear texture crystals, sieved plagioclase exhibits a restricted range composition of labradorite (An 51-61 ; Table 2 in Appendix A).Additionally, an oscillatory zoning has been detected in several sieved plagioclase crystals (Table 3 in Appendix A), where the more calcic composition is observed in the dusty zones (Fig. 7).Some plagioclase phenocrysts show acicular apatite inclusions.On the other hand, FDE lavas include acicular Na-rich plagioclase microphenocrysts, one of them displaying an Or-rich  2, Fig. 3B).Amphibole phenocrysts (0.8-4.0 mm), with intense green to reddish brown pleochroism, cover a variety of compositions, probably as a product of a polybaric crystallization (Hammarstrom and Zen, 1986).In accordance with Leake et al. (1997) classification scheme, they are calcic, showing Fe-tschermakite, Fe-Mg-hastingsite or K-edenite compositions (Table 4 in Appendix A Supplementary Material, and Fig. 4A,B).Thin opaque rims, presumably formed by oxidation, are found on many crystals and locally Fe-Ti oxides appear to have replaced entire phenocrysts.
Mineralogical assemblage in this group is complemented with anhedral Fe-Ti oxides microphenocrysts (< 250 μm).According to the recalculation method suggested by Stormer (1983), titanomagnetite and ilmenite compositions have the range %mol Usp = 13-19 and %mol Ilm = 67-89, respectively (Table 7 in Appendix A Supplementary Material).They occur separately as well as in intimate contact.However, thermodynamic data have not been calculated as the Mg/Mn partition test of Bacon and Hirschmann (1988) has not been validated by the oxide pairs.SC FDE lavas collected in Cantimplora (SC4 and SC5) and Rancho Agustín localities (SC14 and SC15) from SCT sector have a high content (70-80% in phenocryst mode) of euhedral to subhedral (skeletal) olivine phenocrysts (0.7-2.0 mm).Additionally, olivine (<1.5 mm) has also been observed in Peña de Lobos (SC49 from SCN sector) and Ajusco34 (SC1 from SC-SCh transition sector) outcrops, although its modal abundance is ~5%.Some SC olivine phenocrysts exhibit partial iddingsitization indicated by the development of red-clay and Feoxyhydroxide.Olivine compositions extend from Fo 87-90 (core) to Fo 75-88 (rim), generally showing normal zoning (Table 7 in Appendix A).

Chilled magmatic enclaves (ME)
In the present study, seven ME (SCN sector: SC43a, SC49a, and SC49b; SCC sector: SC35a and SC37a; SCS sector: SC29a and SC57a; SC-SCh transition sector: SC24a; Table 1) were separated from their host dacites.They are darker than their host rock and display a porphyritic texture, with plagioclase, orthopyroxene, amphibole, and quartz phenocrysts in a microlitic groundmass mainly composed of acicular plagioclase and accessory opaque minerals.Some ME (SC49a,b) additionally contain scarce olivine phenocrysts (Fo 88-90 ; Table 7 in Appendix A).
On Harker diagrams, all major elements contents decrease with increasing SiO 2 , except alkalis and Al 2 O 3 (Fig. 9).Na 2 O and Al 2 O 3 do not change whereas K 2 O increases with SiO 2 .It is important to note that the most silicic FDE lavas overlap the F dacites without disequilibrium texture.

Trace elements
F dacites exhibit higher abundances of trace elements compared to most ME, with the exception of compatible elements, Sr, and Y.However, SC37a enclave (SCC sector) represents a special case, due to a marked enrichment in REE, Ba, Sr, Zr and Y compared to the other enclaves.Compatible and LILE variation diagrams versus SiO 2 show that the SC rocks plot along a mixing line, although a considerable scatter exists (Fig. 10).REE (e.g., La, Yb) and HFSE (e.g., Nb, Zr) variation diagrams versus SiO 2 display relatively flat trends showing a significant data dispersion.
Chondrite-normalized REE diagrams of all SC rocks (Fig. 11) are characterized by enrichment in light REE and a flat pattern for heavy REE.F dacites display significantly higher (La/Yb) N ratios (= 5.7 -10.3) than the ME (= 4.7 -7.8), whereas FD-E lavas have ratios (= 4.7 -12.0) that overlap or exceed those of the other groups.All rocks lack a significant negative Eu anomaly, as would be expected by a prolonged fractionation of plagioclase, the most abundant phenocryst of the SC rocks.However, a slightly negative Ce anomaly is observed in some cases, which can be evaluated by the Ce N /Ce* N parameter (= Ce N /[La N *Pr N ] 0.5 ; Seto and Akagi, 2008).Dacites lacking disequilibrium texture display a Ce N /Ce* N = 0.74-0.89,a range comparable to the observed in FDE (= 0.60-0.93)and ME (= 0.81-0.93)rock types.
Primitive mantle-normalized multi-element diagrams for SC groups (Fig. 12) are characterized by a zig-zag pattern, with negative anomalies in HFSE (Th, Nb, Ta, and Ti) and P, and a general trend where the abundance diminishes as compatibility increases.

Nd and Sr isotopic ratios
Available Sr and Nd isotopic information for SC rock types are summarized in Table 10 (in Appendix A Supplementary Material). 87Sr/ 86 Sr values of these rocks range from 0.703917 to 0.704268, whereas 143 Nd/ 144 Nd ratios cover the narrow interval from 0.512812 to 0.512903.All samples fall within the "mantle array", shown schematically by small dashed lines in the 87 Sr/ 86 Sr -143 Nd/ 144 Nd diagram (Fig. 13), overlapping with the Mexican lower crust (Patchett and Ruiz, 1987;Ruiz et al., 1988a,b;Roberts and Ruiz, 1989;Schaaf et al. 1994) and crustal xenoliths of Popocatépetl stratovolcano (Schaaf et al., 2005).

Discussion
Several mineralogical and geochemical features observed in SC rocks strongly indicate the involvement of magma mingling processes during the 3.6 to 0.4 Ma period of development of the SC stratovolcanoes.

Magma mingling mineralogical evidence
Sieve texture, as observed in FDE (for example, Fig. 6D) and ME (for example, Fig. 6F), has been described (Tsuchiyama, 1985;Shelley, 1993) as the small, interconnected inclusions of glass or other matrix material giving the crystal a porous appearance.Heating above the plagioclase liquidus temperature causes fusion of the phenocryst and a rounding off (resorption) of the crystal shape.
Acicular plagioclase microphenocrysts displaying a K-rich composition (Fig. 3C) could be interpreted as micro-crystals produced by quenching, where Na atoms have been partially substituted by ion exchange K atoms at high temperature (e.g., Viswanathan, 1971;Kroll and Bambauer, 1981).K migration (diffusion coefficient ~ 2.10 -7 cm 2 /sec) from felsic to mafic magmas during mingling, before the ME undercooling, has been reported by Kumar and Pieru (2010).
An effective blending is inhibited when temperature and viscosity contrasts between two magmas are large and when the proportion of the mafic/intermediate endmember magma is small (typically < 50%), because the mafic/intermediate magma is undercooled to form isolated magmatic inclusions or enclaves in the felsic magma (Eichelberger, 1980;Bacon, 1986;Sparks and Marshall, 1986;Vernon et al., 1988;Stimac et al., 1990;Clyne, 1999;Gençalioğlu Kuşcu and Floyd, 2001;Coombs et al., 2002;Alpaslan et al., 2005).Some of the FDE rocks (see Table 1 and Fig. 6E,F) include rounded to ellipsoidal magmatic enclaves, which generally maintain coherent and sharp contacts with their host lavas.ME can be interpreted as quenched blobs of an andesitic magma that was injected into relatively cooler dacitic magma, displaying a non-Newtonian behavior within their host magma.A disparity in physical properties between andesitic enclaves and their host dacitic magmas probably leads to fast crys-  tallization of the former magmatic enclaves, as indicated by the acicular character of some crystals (Sparks and Marshall, 1986;Blake and Fink, 2000).

Magma mingling geochemical evidence
Eichelberger et al. (2006) suggested that, rather than a liquid line of descent, the linear trends as those observed in the Harker major-element diagrams for SC lavas (Fig. 9) reflect a spectrum of discrete magma batches product of complex mingling processes between intermediate and silicic end-members.The lack of coherent trends on Harker trace-element diagrams (Fig. 10) also supports the argument that analyzed compositions did not follow simple liquid lines of descent, but rather reflect complex opensystem processes including partial mixing or mingling.
Trace element (including rare earths) Harker patterns, however, show partial to complete equilibration, most likely governed by different degrees of elemental dif-fusion.Additionally, at least one end-member must be compositionally heterogeneous, as reflected by the differences in trace element geochemistry of SC37a and the rest of andesitic enclaves.
It is important to note that, the Harker diagram linear arrays are product of several magma mingling events that occurred during the SC geological history, (e.g., SiO 2 vs. MgO Harker plots, (Fig. 14).Taking into account the morphostructural SC sectors, the best defined linear arrangement is observed in the SCN sector, where: (a) the compositional contrast between the magmatic end-members is more pronounced, and (b) the magmatic enclave density per area is maximum, and its size reaching 20 cm.Harker diagrams of SCC and SCS sectors are characterized by a greater dispersion, as: (a) the difference in composition of ME in relation to F rocks is lower, and (b) the number and size of enclaves decreases southward.
Low HFSE concentrations showed by SC rocks could be related to amphibole and ilmenite fractional crystallization, whereas P negative anomaly could be related to apatite fractionation (Rollinson, 1993).On the other hand, Th/Cs (= 2.5-3.5) and Rb/Cs (= 23-45) ratios in ME are similar to those of the continental crust (Th/Cs = 2.1-4.0,Rb/Cs = 17.1-36.7;Rudnick and Gao, 2003).Crustal ratios are also observed in F lavas (Th/Cs = 1.5-2.8;Rb/Cs = 15-40).This similarity in incompatible trace element ratios strongly suggests the crust as the potential source region for both rock types, although magma generation probably occurred at different depths.Nevertheless, crystal fractionation processes are not sufficient in order to explain the correlation between Th/Cs and Rb/ Cs ratios.Therefore, it is suggested that a partial mixing relationship superimposed on fractional crystallization is required to explain the SC volcanic rocks geochemistry.
Negative Ce anomalies, observed in chondrite-normalized REE diagrams, could result from incipient weathering (e.g., Patino et al., 2003), although no petrographic evidence has been detected in the SC volcanic rocks.Borg and Clyne (1998) suggested that upper-crustal assimilation might explain the negative Ce anomalies observed in felsic calc-alkaline magmas from the southernmost Cascades.However, REE patterns of crustal xenoliths from central MVB (Urrutia-Fucugauchi and Uribe- Cifuentes, 1999;Aguirre-Díaz et al., 2002) do not display negative Ce anomalies.Gómez-Tuena et al. (2006) have reported intermediate to felsic lavas from the Nevado de Toluca volcano characterized by negative Ce anomalies, being interpreted as evidence for slab-derived sedimentary contributions in the petrogenesis of the central MVB.However, REE normalized patterns for pelagic sediments from Site 487 located in the Cocos plate also show a significant negative Eu anomaly (Verma, 2000), which is not observed in the SC rock types.In such a way, negative Ce anomaly remains unexplained, as additional Pb isotopic information for SC lavas is necessary to understand this phenomenon.
Unfortunately, approximations of physical properties of magmas (e.g., pressure and temperature) cannot at present be evaluated because thermodynamic data are not still available for SC host dacitic and intruded andesitic magmas.
The formation of enclaves probably took place later in the mingling process, because the co-mingled lavas exhibit disequilibrium mineral textures and a geochemistry between the andesitic enclaves and the dacites.The partial mixing origin of FDE rocks can also be modeled based on the geochemical compositions of two end-members, particularly using trace elements (Rollinson, 1993).For example, in the Rb/Sr and K/Ti ratios plotted versus Rb diagrams (Fig. 15A,B), lavas with disequilibrium textures (FDE) fall on the representative mixing curves drawn by using average F dacites as a silicic end-member and considering two basic ME end-members (E1: average of SC27a, SC35a, and SC49a,b; E2: SC37a andesitic enclave).Shifting of some points from representative mixing curves can be interpreted as previous or coeval plagioclase, amphibole, and pyroxene fractional crystallization, although may also been affected by mineral sorting and/or diffusion processes.

Proposed magma mingling/mixing scenario
Magmatic processes such as differentiation, eruption style, vesiculation, and fluid flow are mainly ruled by chemical composition and inter-related melt physical properties including temperature, density, viscosity, and crystallinity.When two contrasting magmas interact several scenarios may unfold (Donoghue et al., 1995): (a) the melts may mix physically, thereby forming a hybrid magma; (b) one magma may freeze against the host magma, giving a net-vein complex, and abruptly arresting the mixing process; (c) the melts may mix partially but incompletely, particularly when the two end-member components are porphyritic, leaving phenocrysts from both magmas with strong disequilibrium textures; and (d) the melts may intermingle but not mix, forming a banded rock.Where mixing has been efficient, the only evidence for this process may be disequilibrium textures and/or linear arrays in Harker diagrams.However, where mixing is limited or incomplete, the development of magmatic enclaves or banded rocks may occur, which could retain the identity of the end-members.These phenomena have operated commonly in volcanic and plutonic environments (e.g., Nakamura, 1995;Eichelberger et al., 2000;Coombs et al., 2002;Brown et al., 2006;Kumar and Rino, 2006;Dokukina et al., 2010).andesitic magmas, probably originated in the interface between upper and lower crust, and felsic magmas possibly derived from upper crust partial melting.These lines of evidence include: (a) chilled magmatic enclaves which probably represent the material that mingled with felsic magma, causing thermal and compositional disequilibrium; (b) presence of more and less silicic amphiboles in the same sample, some of them showing reaction textures and inverse zoning; (c) bronzite and hypersthene orthopyroxenes displaying augitic reaction rims, and (d) sieved plagioclase crystals and their occurrence together with clear normal plagioclase, as well as oscillatory zoning of sieved plagioclase.Magma mixing processes in Sierra de las Cruces are also supported by linear arrays in bivariate plots of major elements.Diagrams based on trace element ratios also point out to an origin by mixing between dacitic and andesitic melts for the Sierra de las Cruces volcanic rocks.Results from this study confirm the significance of magma mixing processes in the intermediate-felsic volcanic rocks petrogenesis in the central MVB.
Therefore, the following model is proposed for magma mingling/mixing in Sierra de las Cruces volcanic range, that include the recurrent operation of the (b) and (c) scenarios by Donoghue et al. (1995) from 3.5 to 0.4 Ma.Fix (1975) reported a decrease in the seismic velocities at the base of the crust in the central MVB.Campos-Enríquez and Sánchez-Zamora (2000) have indeed interpreted gravity data to infer the presence of partial melts in the lower crust beneath this region.These magmas might be stored at the base of the crust, transferring heat to shallower crustal levels and provoking their partial melting.SC dacitic rocks were probably generated by partial melting of the upper continental crust (depth at the base ~10 km; Ortega- Gutiérrez et al., 2008), and were subsequently stored in a shallow level magma chamber.The combined Sr-Nd isotope data from the Mexican lower crust and crustal xenoliths from Popocatépetl stratovolcano (Fig. 13) are in general consistent with a significant crustal involvement in the genesis of SC rocks.Sodic plagioclase, Mg-hastingsite amphibole, hypherstene and minor quartz were crystallized prior to the injection of the andesitic magma.Subsequently, a small volume of andesitic magma, probably generated at a deeper crustal level (lower crust (?): 25-45 km; Ortega-Gutiérrez et al., 2008), intruded in the shallow magma chamber, losing heat to the surroundings and starting to vesiculate.Eichelberger et al. (2000) suggested that, if ponding mafic magma operates like a rising piston, co-mingled lavas are slowly expulsed reaching the surface by a volatile-poor and non-explosive effusive eruption.Felsic minerals in the host dacite became round and embayed, plagioclase crystals were sieved, and mafic minerals developed reaction rims as a response to an increase in temperature.On the other hand, quenching of the andesitic magma resulted in discrete enclave development, reflecting an incomplete physical mixing process.Mingling process occurred due to differences in physical properties between the two magmas and the relatively small volume proportion of the intermediate magma.Additionally, limited exchange of xenocrysts (labradorite, edenitic amphibole, and bronzite in dacites, oligoclase and andesine in andesitic enclaves) might have occurred during this entrainment stage.

Conclusions
This study presents detailed petrography, mineral chemistry, and whole-rock geochemical data for Sierra de las Cruces.Magma mingling/mixing processes have played a significant role in the evolution of this volcanic range.Most dacites display several disequilibrium features that imply the involvement of open system magmatic evolution processes over the 3.6 to 0.4 Ma period.Several mineralogical lines of evidence confirm the interaction between
Table 1 in Appendix A) have been integrated in this group.FDE lavas are generally porphyritic, showing flow texture and sometimes vesicles.Phenocrysts are