| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology | Volume 43 | Number 5 | Pages 907-942 | 2002
© Oxford University Press 2002
Petrogenesis of Early Neogene Magmatism in the Northern Puna; Implications for Magma Genesis and Crustal Processes in the Central Andean Plateau
1INSTITUTO DE GEOLOGÍA Y MINERÍA, UNIVERSIDAD NACIONAL DE JUJUY-CONICET, AVDA. BOLIVIA 1661, SAN SALVADOR DE JUJUY, ARGENTINA
2GEOFORSCHUNGSZENTRUM-POTSDAM, TELEGRAFENBERG 14473, POTSDAM, GERMANY
Received January 22, 2001; Revised typescript accepted November 30, 2001
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTIONNew compositional data and petrogenetic models are presented for pre-Upper Miocene volcanism in the northern Puna of Argentina (22°S–24°S). Two phases of volcanism produced small dome complexes of mainly silicic andesite to low-SiO2 rhyolite. The Upper Oligocene–Lower Miocene phase (UOLM, 20–17 Ma), produced two distinct groups of rocks. The UOLM-1 group is metaluminous and mainly andesitic, with isotopic compositions like those of the recent arc (87Sr/86SrT
0·706; 
NdT -3). The UOLM-2 group is more silicic and peraluminous, and has isotopic compositions indicating a substantial crustal contribution (87Sr/86SrT 0·713; NdT -8). The Mid-Miocene phase (MM: 15–12 Ma) produced rocks similar in composition to those of the UOLM-2 group (87Sr/86SrT 0·710; NdT -7) but with higher incompatible element contents. Ratios of Ba/Nb and Zr/Nb in the UOLM group rocks are uniform and similar to those of the current arc, whereas the ratios in MM centres show a mixed arc and back-arc affinity. This suggests that the westward shift in the arc began in the northern Puna in the mid-Miocene. Neither the exposed Palaeozoic felsic basement nor the lower-crustal granulites known from xenolith suites are compositionally suitable as protoliths for the UOLM and MM magmas. The preferred petrogenetic model for the magmas involves hybridization of a depleted arc basalt with partial melts of the felsic basement. Geochemical modelling and thermal arguments rule out magma mixing as the process of hybridization. Successful assimilation–fractional crystallization (AFC) solutions indicate an increase in crustal assimilation from 15–25% in UOLM-1, to 40–60% in the case of UOLM-2 and MM group rocks. Assuming the same end-member compositions, the modelling suggests genesis of the MM magmas at higher pressure than the UOLM-2 centres (
10 kbar vs 7 kbar), which may reflect the influence of crustal thickening in the plateau region by the mid-Miocene. The felsic dome complexes of this study are compositionally similar to the large-volume, caldera-sourced felsic ignimbrites that dominated volcanism in the region from 10 to 2 Ma and our results suggest that there is no fundamental difference in magma genesis between them. The differences in the volumes and the mode of eruption reflect changes in the stress and thermal regime with time. KEY WORDS: Cenozoic volcanism; Central Andes–Puna Plateau; crustal assimilation; dacitic magmas; geochemical modelling
|
INTRODUCTION |
|---|
Cenozoic magmas erupted in the high plateau region (Puna–Altiplano) of the central Andes (Fig. 1) have been the subject of intense investigation since the late 1980s. The focus has primarily been on the Upper Miocene to Recent magmatism, which includes large-volume caldera-sourced ignimbrites (Schneider, 1985
; de Silva, 1989a; Francis et al., 1989; Ort et al., 1996; Lindsay et al., 2001), frontal-arc composite volcanoes (Davidson et al., 1991; Feeley & Davidson, 1994; Matthews et al., 1999), and small-volume back-arc or transverse mafic centres (Kay et al., 1994a; Davidson & de Silva, 1995; Redwood & Rice, 1997). In comparison, the initial phases of Andean volcanism in the Altiplano–Puna region (Late Oligocene to Middle Miocene) have not been investigated in detail, and have been considered mainly in papers of a regional or reconnaissance nature (Halls & Schneider, 1988; Coira et al., 1993; Kay et al., 1994b; Trumbull et al., 1999).
|
). The study area is outlined by the box. Grey circles are 20–17 Ma centres with arc-like Ba/Nb and Zr/Nb ratios. Open stars are 14–12 Ma magmas with arc-like signatures; filled stars indicate centres with back-arc affinity. The El Toro lineament divides the northern and the southern Puna, and is marked with a bold continuous line. Other NW–SE lineaments in the southern Puna are represented as dashed lines.
This study focuses on the Late Oligocene–Middle Miocene magmatism in the northern Puna region of Argentina (22–24·5°S). This period coincides with important episodes of compressive deformation in the Central Andes, which are thought to be the main reason for crustal thickening (up to 60 km; Yuan et al., 2000) beneath the Puna region (Isacks, 1988; Sheffels, 1990; Allmendinger et al., 1997; Kley et al., 1999). It has also been suggested that the subduction angle steepened during this time, leading to a narrowing and westward migration of the volcanic front (Coira et al., 1993; Allmendinger et al., 1997, and references therein). The northern Puna Middle Miocene volcanic centres form the southern extension of the Bolivian Tin and Polymetallic Belt and some of them have a close relationship with Pb–Zn–Ag ± Sn ore deposits (Coira, 1994; Caffe & Coira, 1999). Therefore, a better understanding of the genesis and development of these centres is important for many broader issues relating to geodynamic controls on Andean magma generation, and may have implications for the relationship between magmatism and ore associations.
We present new petrographic, mineralogical, geochemical, isotopic and geochronological data for Upper Oligocene to Middle Miocene volcanic dome complexes and related igneous rocks from the northern Puna. The data are used to identify potential magma source components within the mantle and crust. We suggest that even in the most crustal-like magmas the involvement of mantle sources cannot be ruled out. Geochemical and isotopic modelling favor hybridization of arc magmas in the lower crust, and are consistent with a broadly felsic composition of the lower crust suggested by Lucassen et al. (1999b).
|
TECTONIC SETTING AND REGIONAL GEOLOGY |
|---|
The northern Puna region of Argentina (22–24·5°S) is structurally separated from the southern Puna (25–28°S) by the NW–SE-trending El Toro lineament, whereas its border with the Bolivian Altiplano to the north (
21·5°S) lacks any major structural expression. Volcanism in the period from 26 to 18 Ma shows a characteristic distribution. North of 22°S in the Bolivian Altiplano volcanic centres are spread out across the arc and back-arc regions, whereas south of 25°S (southernmost Puna) they are confined to the arc front. This distribution and the overall scarcity of volcanic rocks of this age in the 22–25° segment led to the suggestion of a spatial gap in magmatism in the northern Puna between the Upper Oligocene and Lower Miocene (Coira et al., 1993; Allmendinger et al., 1997). In contrast, Middle Miocene to Pliocene volcanic centres are spread out over the whole Argentine Puna (Fig. 1). On the basis of age determinations and trace element geochemistry, Coira et al. (1993) and Kay et al. (1999) suggested progressive shallowing of the angle of subduction from Oligocene to Recent times south of 28°S, and steepening north of 25°S. This interpretation was questioned by Kley et al. (1999), who concluded that no major changes took place in the subduction regime before 10–7 Ma.
Cenozoic volcanism
Cenozoic volcanic rocks in the Puna province (22–28°S) and southern Bolivia occur across a broad area underlain by Ordovician sedimentary and volcanic rocks (Acoite Formation; Turner, 1960; and Faja Eruptiva de la Puna; Coira et al., 1999, respectively), scarce Cretaceous to Paleocene rocks (Salta Group; Salfity, 1982), and Eocene to Lower Miocene redbed deposits associated with the sedimentary fill of the first Andean foreland basins (Jordan & Alonso, 1987). Cenozoic volcanic phases identified in the Puna (Coira et al., 1993) are roughly coincident with those proposed for the Bolivian Altiplano and Eastern Cordillera by Soler & Jiménez (1993), and comprise four main periods: Upper Oligocene–Lower Miocene, Mid-Miocene, Upper Miocene–Pliocene, and Pleistocene (Table 1). Each of these volcanic phases coincided with the end of a major tectonic phase. Thus, the oldest Cenozoic magmatic rocks (Upper Oligocene–late Lower Miocene) are interbedded with thick conglomeratic sequences, indicating that they were erupted before or during the Pehuenche tectonic phase. This phase involved thrusting of Ordovician metasediments onto Paleogene redbeds. Thrusting ceased at 18 Ma in southern Bolivia (Hèrail et al., 1994; Kley et al., 1996), whereas activity continued to 16 Ma some 80 km to the south (Cladouhos et al., 1994; Coira et al., 2002).
|
Like the deformation, the first phase of volcanism also shows a diachronous distribution from north to south. Magmatism in the northern Puna may have started in the Upper Oligocene (28 Ma) as in the Altiplano, but it was mainly active during the late Lower Miocene (19–17 Ma). The age of these later eruptions overlaps somewhat with the second volcanic phase (Coira et al., 1993; Allmendinger et al., 1997); however, because the eruptions occurred during the Pehuenche tectonic event, we ascribe the rocks to the first volcanic phase. The eruptive volumes of this early magmatic phase were similar to, or even higher than, those of the second volcanic phase but volcanism was more restricted geographically, being confined to the areas around the Laguna de Pozuelos (Fig. 2). The first volcanic phase was not accompanied by ore formation.
|
The second volcanic phase (Fig. 2, Table 1) developed completely within the Middle Miocene (16–12 Ma). Main centres erupted within this phase have been classified as volcanic dome complexes or dacitic stocks (Caffe & Coira, 1999), and form a group of regionally extensive, conspicuous magmatic features in the northern Puna, Eastern Cordillera and the Bolivian Altiplano (Coira et al., 1993). Most of the centres are closely associated with mineralization, being part of the Bolivian Tin or Polymetallic Belts (see Cunningham et al., 1991; Coira, 1994). These complexes and some volcaniclastic units interbedded in the Middle Miocene Tiomayo Formation (Coira et al., 2002) display only slight deformation. However, seismic reflection data from the Laguna de Pozuelos basin (Fig. 2) show syntectonic sedimentation structures and thrusting and folding of mid-Miocene strata (Gangui, 1998), which suggests that there was intense compressive strain in the northern Puna at this time.
At 11–8 Ma, the Quechua tectonic phase (Table 1) resulted in marked elevation of the northern Puna region and development of the broad San Juan del Oro erosional surface (Gubbels et al., 1993). The third volcanic phase (10–4 Ma) overlaps with the Quechua phase, and is dominated by undeformed, large-volume ignimbrites (Fig. 2, Table 1). As in the Altiplano, the end of the Quechua phase (at 8 Ma) in the Puna is marked by changes in strain directions and deformational behaviour of the upper–middle crust, passing from a compressional regime into one that was dominated by strike-slip or extensional faulting (Cladouhos et al., 1994).
|
DESCRIPTION OF THE UNITS SAMPLED |
|---|
First volcanic phase: Upper Oligocene to Lower Miocene (henceforth UOLM centres)
Casa Colorada dacite dome complex (22°19’S–66°20’W)
The Casa Colorada dacite dome complex is a small extrusive magmatic centre (
1·5 km3) located west of Sierra de Rinconada at the intersection of north–south- and east–west-trending faults (Fig. 2). Three successive volcanic events have been recognized (Caffe, 1996): deposition of a tuff ring; collapse of growing lava domes; and effusion of flow-banded, crystal-rich, dacitic lavas. The 17·3 ± 0·7 Ma K–Ar age of the centre (Coira et al., 2002) indicates that the main dacite lava dome was extruded contemporaneously with the nearby Cabreria Formation ignimbrites.
Minuyoc dacite dome complex (22°32'S–66°14'W)
The Minuyoc dacite dome complex is the smallest early Miocene volcanic feature in the northern Puna (0·15 km3). It was emplaced at the intersection of NW–SE and NNE–SSW fracture zones at the eastern border of Sierra de Rinconada (Fig. 2). The eruptive sequence at Minuyoc includes a basal explosion breccia followed by massive hydromagmatic breccias and tuffs, crystal-rich dacitic lavas and finally, phreatic tuffs. This complex has not been dated, but on the basis of its trace element and isotopic composition, it is correlated with other dated first cycle volcanic rocks.
Pirurayo volcanic complex (22°21'S–65°52'W)
The Pirurayo volcanic complex consists of a succession of block and ash flow deposits, lavas and lahar bodies within the middle levels of the conglomeratic Moreta Formation (Soler, 1996) (Fig. 2). Ages obtained for the unit range between 28 ± 3 Ma and 20 ± 2 Ma (Linares & González, 1990). Soler (1996) estimated the volume of the Pirurayo deposits at 13 km3. The composition is mainly andesitic to dacitic, although scarce low-silica rhyolites are also present. The complex is strongly deformed, with open folds prevailing in the northern exposures, tight folds and thrust faults at central locations, and a monoclinal form in the south.
Cabreria Formation ignimbrites and volcaniclastic deposits (probable location of vents 22°14'S–66°20'W)
The western margin of Sierra de Rinconada (Fig. 2) is flanked by a thick conglomeratic sequence (400–1200 m thick) known as the Cabreria Formation (Coira et al., 2002). The upper member (Quebrada Grande) is composed of 20 volcaniclastic units, of which nine are crystal-rich, poorly welded, dacitic ignimbrites (each 2–15 m thick). According to thickness variations of the pyroclastic sequence, the source of the primary deposits may be located to the north of Casa Colorada, near 22°15'S, from which they thin out both to the north and south. Homogeneous major and trace element compositions suggest that the entire volcanic volume represented in these units (15 km3, primary plus reworked) erupted from the same source. Coira et al. (2002) determined a 17·4 ± 0·8 Ma K–Ar age for the middle part of the sequence.
Laguna de Pozuelos volcaniclastic sequence (22°30'S–22°40'S; 66°08'W–66°15'W)
This sequence of mixed pyroclastic and reworked volcanic rocks is located SW of the Laguna de Pozuelos basin (Fig. 2). Primary volcanic units include pumiceous, crystal-rich dacite to rhyodacite ignimbrites, massive air-fall tuffs, pyroclastic surge deposits, and lahar bodies up to 30 m thick. Chernicoff et al. (1996) suggested that a ring fracture (ancient caldera structure) at the southern border of the Laguna de Pozuelos basin could have been the source of eruptions. A new K–Ar biotite age for pumice from an ignimbrite in the middle section is 18·6 ± 1 Ma (P. J. Caffe, unpublished data, 2000), which is consistent with its stratigraphic position below mid-Miocene tuffs of the Cara Cara strata (Tiomayo Formation), dated by Cladouhos et al. (1994). The greatest abundance of pyroclastic material is restricted to exposures to the west of the Pan de Azucar mine (Fig. 2), and the total preserved eruptive volume is 6 km3. This value is probably much too low, as the base of the sequence is not exposed in most outcrops, and because post-10 Ma cover in the Laguna de Pozuelos basin has been extensive. Thus, the Laguna de Pozuelos volcaniclastic sequence studied here may correspond to the basal part of an 1200 m sequence interpreted from seismic reflection data by Gangui (1998).
El Morro (23°11'S–66°54'W)
This centre is located slightly to the east of the Argentina–Chile border near Jama (Fig. 2). It is a small quartz andesite stock, which intrudes the Ordovician basement close to the mid-Miocene Aguiliri volcanic complex (see below). This centre has not been dated, but a number of factors, including greater degree of erosion than the Aguiliri dacitic domes, and its overall geochemical characteristics, suggest the El Morro centre has closer affinities with magmatic rocks of the first phase rather than the second.
Second volcanic phase: Middle Miocene (henceforth MM centres)
Pan de Azucar (22°32'–22°38'S; 66°01'–66°07'W)
The well-known Pan de Azucar Ag–Pb–Zn ore deposit, which has been mined from colonial times until the early 1990s, is situated in a volcanic dome complex of the same name at the southwestern margin of the Laguna de Pozuelos basin (Fig. 2). The complex comprises a variety of lava flows, autobreccias, subvolcanic dacites, lava domes and diverse pyroclastic rocks. Available K–Ar ages (12 ± 2 Ma, 13 ± 1 Ma; Coira, 1979) constrain the eruptive events to the Middle Miocene. Caffe (1999) identified a minimum outcropping volume of 1·5 km3, which was erupted in three events. The first event began with plinian eruptions followed by alternating growth–collapse episodes of lava domes. The second event involved phreatomagmatic–vulcanian eruptions, late dacite-dome lavas and emplacement of a late-stage shallow subvolcanic body. The third event involved collapse of a flow-banded lava dome to the south and eruption of grey dacitic dome lavas. Eruptions were localized by NW–SE extensional fractures crossing the Laguna de Pozuelos basin, most of which are related to strike-slip components of the thrust faults bounding the high ranges (Fig. 2) to the east and west of the complex (Chernicoff et al., 1996).
Chinchillas (22°30'S–66°15'W)
Like Minuyoc, the Chinchillas volcanic dome complex is a small volcanic centre (0·26 km3) located in the Sierra de Rinconada. The complex hosts a mineralized hydrothermal breccia, with ores enriched in Zn and minor Sn, which were mined intermittently (Caffe & Coira, 1999). The centre has been dated by only a single K–Ar age determination of 13 ± 1 Ma (Linares & González, 1990). Volcanic units comprise co-ignimbrite breccias and massive low-volume pyroclastic flow deposits, followed by block and ash flow deposits and a lava dome confined to the south. Volcanism was related to the same NW–SE fault along which the Minuyoc complex was emplaced.
Aguiliri (23°11'–23°17'S; 66°51'–66°57'W)
The Aguiliri lava dome complex is the southernmost eruptive centre studied (Fig. 2). The centre forms a cluster of one dacite stock and three dacitic lava domes, which intruded basement and Tertiary sediments near the Chile–Argentina border at Jama. A new K–Ar biotite age indicates a Middle Miocene age of 12·7 ± 1·3 Ma (P. J. Caffe, unpublished data, 2000). The Aguiliri intrusive stock has anomalously high concentrations of U, and weak anomalies of Pb and Ag. Aniel (1987) concluded that the U enrichment was hydrothermal, not magmatic.
|
PETROGRAPHY AND MINERAL COMPOSITIONS |
|---|
The sampled dome complexes and related volcanic units in the northern Puna have dominantly dacitic to rhyodacitic compositions except for the El Morro intrusive rocks and some flows from the Pirurayo complex, which are andesitic. The dacitic rocks generally share similar petrographic characteristics. Lavas and pumice fragments are porphyritic, with a variably devitrified glassy groundmass. Crystal contents range from 5 to 47 vol. % in pumice and from 19 to 60 vol. % in lavas (Table 2). Shallow intrusive facies have the same crystallinity as the lavas.
|
The dominant mineral (Table 2) in the dacitic and rhyodacitic rocks is plagioclase, followed in abundance by biotite and quartz. Amphibole, clinopyroxene, orthopyroxene, (Ti-) magnetite and ilmenite occur in minor and variable proportions. Sanidine is extremely rare, and allanite, apatite, zircon and monazite occur as accessories. Small (millimetre-sized) clots of the two or three dominant mineral phases in lavas and pumices are interpreted as crystal accumulation textures (Table 2). The main petrographic difference shown by andesitic lavas (in Pirurayo and El Morro) is the relative abundance of mineral mafic phases, amphiboles and pyroxenes being more abundant than biotite.
Some samples from the mid-Miocene centres contain sillimanite, ilmenite, spinel and biotite as inclusions in plagioclase phenocrysts and in the cores of plagioclase–biotite clots. These are interpreted as restitic phases derived from partial melting of assimilated crustal rocks. Inclusions of oval or blocky holocrystalline quartz-dioritic to tonalitic enclaves (diameter 15–2 cm) are also common. These inclusions have almost the identical mineral assemblage to the host lavas (Table 2) but are finer grained. Like the mineral clots mentioned above, we interpret them as products of crystal accumulation, probably at the magma chamber–wallrock contacts (de Silva, 1989b).
Mineral compositions were obtained by electron microprobe analysis from the UOLM centres Casa Colorada and Minuyoc, and from the MM Pan de Azucar and Aguiliri complexes [the latter from Aniel (1987)]. Representative microprobe analyses of plagioclase, biotite, amphibole, pyroxenes and Fe–Ti oxides are reported in Tables 3
–7.
|
|
|
|
|
Plagioclase
Plagioclase compositions vary from sodic andesine to sodic labradorite in the various units. Plagioclase in cumulate clots and tonalitic inclusions is significantly richer in the anorthite component (Fig. 3). Phenocrysts in UOLM complexes are normally zoned from An76–55 cores to An55–35 rims. Plagioclase phenocrysts from the MM centres show either moderate normal zonation (in Aguiliri, An50–37), or oscillatory to slightly inverse zonation (in Pan de Azucar; Fig. 3 and Table 3). The highest anorthite contents (An62–48) in the Pan de Azucar samples occur in intermediate zones of grains showing sieve texture. Plagioclase from mineral clots (Pan de Azucar), which include relict aluminous minerals, is more calcic than lava phenocrysts (Fig. 3).
|
The Sr contents in plagioclase mirror the prominent whole-rock Sr variations observed in the UOLM to MM volcanic rocks (see below). Thus, plagioclase from MM centres generally has higher Sr contents than the UOLM equivalents (Table 3).
Biotite
The mg-number [100 x Mg/(Mg + Fe)] of biotite phenocrysts from UOLM centres ranges between 45 and 49 (Table 4), except for biotite rims around amphibole (in Minuyoc), which are more magnesian (mg-number 56–62). In MM centres, biotite mg-numbers are in the range of 50–60 for Aguiliri (Aniel, 1987) and 45–62 for Pan de Azucar. Biotite from tonalitic to dioritic inclusions has mg-numbers between 57 and 61. The TiO2 content of biotite is higher in MM centres [Ti 0·20–0·27 cations per formula unit (p.f.u.)] than in UOLM lavas (Ti 0·18–0·21 cations p.f.u.; Table 4), suggesting lower crystallization temperatures (Patiño Douce, 1993) for the UOLM group.
Amphibole
Amphibole in the UOLM centres classifies as tschermakite to Mg-hornblende and commonly displays disequilibrium textures. For example, hornblende from Casa Colorada lavas is corroded, and in Minuyoc samples, hornblende is replaced by a reaction rim composed of biotite–orthopyroxene–ilmenite–plagioclase. Amphibole phenocrysts from the MM samples classify mainly as Mg-hastingsite or Mg-hornblende (Table 5) and, in contrast to the UOLM rocks, they show no signs of instability in the melt before eruption. Amphibole mg-numbers are generally higher in UOLM than in MM lavas (72–84 vs 58–75, respectively; Fig. 4). Hornblende–plagioclase geothermometry (Holland & Blundy, 1994) performed on rim compositions from MM samples lacking reaction texture suggested pre-eruption temperatures of 830–900°C (Caffe, 1999).
|
Pyroxenes
Pyroxene is a rare phase in the dome complexes (Table 2). Orthopyroxene (En68–47 Fs51–29 Wo1·4–2·5) was noted in Minuyoc lavas, and it shows decreasing mg-number from reaction rims on amphibole to phenocryst cores to rims (Table 6). Augitic clinopyroxene (En41Fs16Wo43; mg-number 73) is restricted to rare occurrences in lavas from Pan de Azucar and Aguiliri dome complexes.
Fe–Ti oxides
Magnetite is the predominant oxide phase in UOLM lavas and holocrystalline cumulates (Table 7); ilmenite is subordinate. Among the MM domes, Pan de Azucar contains abundant magnetite (TEL in Table 7), either in the groundmass or included in plagioclase and biotite, whereas ilmenite is restricted to inclusions in feldspar cores (SEL and TEL in Table 7). The Aguiliri intrusive stock lacks magnetite, but lavas from the dacitic domes in Aguiliri (Cerro Chingolo–Norte) contain both oxides.
Almost all magnetite crystals have developed oxidation–exsolution lamellae. Among the UOLM volcanic domes, the best preserved magnetite compositions (Casa Colorada) have ulvöspinel contents up to 18–21 mol %. Among the MM centres, non-exsolved, primary grains were observed in a few lavas and dioritic cumulates from Pan de Azucar (TEL in Table 7), ranging from 23 to 29 mol % ulvöspinel. Reintegrated primary Ti-magnetite compositions [method of Mathison (1975)] suggest original ulvöspinel contents of up to 36–44 mol %.
UOLM rocks show variable composition of primary ilmenite. In Minuyoc, ilmenite is poor in the hematite component (3 mol %; Table 7), but in Casa Colorada hematite ranges from 7 to 19 mol %. Lavas from the MM Pan de Azucar complex contain nearly pure ilmenite (XHem 2–6 mol %), whereas in the holocrystalline mineral cumulates (AH-51 in Table 7) ilmenite crystals are replaced by a hemo-ilmenite–rutile intergrowth. Aniel (1987) reported ilmenite from the MM Aguiliri intrusive rocks, which exhibits variable oxidation textures and hematite contents ranging from 7 to 50 mol %.
|
GEOCHEMISTRY |
|---|
Major elements
In general, the UOLM to MM volcanic rocks in the northern Puna classify (Fig. 5) as high-K dacites to low-SiO2 rhyolites in the K2O–SiO2 diagram of Peccerillo & Taylor (1976)
. A few matrix glass analyses from Pan de Azucar lavas have high-silica rhyolite compositions. Dioritic enclaves plot in the medium- to high-K andesite fields, as do the El Morro intrusive rocks and the least evolved rocks from the Pirurayo volcanic complex (high-SiO2 andesites).
|
In terms of the alumina saturation index [A/CNK = molar Al2O3/(CaO + Na2O + K2O)], both the UOLM (A/CNK = 0·83–1·26) and MM groups (A/CNK = 0·91–1·39) are highly variable (Fig. 6). Among the UOLM centres, two sub-groups can be differentiated using A/CNK ratios. The UOLM-1 sub-group, formed by El Morro and Pirurayo (Fig. 6), is characterized by A/CNK ratios that increase with increasing SiO2 (58–71%) from metaluminous quartz-andesites and dacites to peraluminous dacites and low-SiO2 rhyolites (A/CNK up to 1·14). On the other hand, UOLM-2 centres (Pozuelos, Casa Colorada, Cabreria Formation and Minuyoc), have A/CNK ratios that show no correlation with SiO2 (Fig. 6), and many more samples are peraluminous. The MM group of rocks from the northern Puna have the same characteristics as UOLM-2 rocks with respect to A/CNK and SiO2.
|
The overall major element characteristics of the studied rocks are typical for central Andes back-arc silicic volcanic rocks, with Al2O3 ranging from 14 to 17%, K2O >2·5%, K2O/Na2O ratios generally >1 and the sum of the alkalis >4–9%. TiO2 concentrations are low (<1·6%), and MgO contents are <3% (Table 8). Concentrations of Al2O3, TiO2, FeOt, MgO and CaO decrease with increasing silica (Fig. 5), K2O increases with increasing SiO2, and Na2O is variable, except for the Pan de Azucar matrix glasses, where it correlates negatively with SiO2. The Harker diagrams in Fig. 5 show that, for the same SiO2 content, MM rocks are slightly richer in Na2O (especially Aguiliri, whose K2O/Na2O ratios are 1). Chinchillas is an exception, and its low Na2O (and CaO) contents probably reflect hydrothermal remobilization. The P2O5–SiO2 variation trends (not shown) are specific to each centre, with kinked trends in Aguiliri, strong negative correlation in Pozuelos, and constant P2O5 level in the remaining centres, which suggest different apatite fractionation histories. The dioritic and tonalitic enclaves show similar variation trends to their volcanic hosts, the differences reflecting the particular mineralogy accumulated (e.g. high P2O5 indicates apatite accumulation).
|
Trace elements
As for the major elements, the trace element characteristics of the early Neogene volcanic rocks in the northern Puna are typical of Andean back-arc magmatism (see Kay et al., 1994b), in particular with respect to Ba (370–1224 ppm), Sr (153–849 ppm), Rb (65–328 ppm), Th (6–22 ppm), U (1–7 ppm), the rare earth elements (REE)* (La 21–56 ppm, Ce 33–107 ppm, Sm 4·3–9·4 ppm, Eu 1·1–2·7 ppm, Yb 1·4–3·3 ppm) and high field strength elements (HFSE) Hf (3·4–9·8 ppm), Zr (110–240 ppm) and Nb (7–21 ppm).
Selected trace element variation diagrams are plotted in Fig. 7. Rb concentrations exhibit an overall weak positive correlation with SiO2. Sr concentrations are more variable and specific to each centre, but for similar SiO2, MM rocks (14–12 Ma) are richer in Sr (430–849 ppm) than those from the older centres (153–480 ppm). Ba and Zr concentrations are also rather scattered and, like Sr, concentrations are higher in MM rocks than in UOLM centres. Within each centre, Nb shows a flat variation trend with SiO2 and tends toward higher values in the MM group than in UOLM volcanic rocks. In contrast, for the same SiO2 level, Y contents decrease from the older, UOLM volcanic rocks to the MM rocks.
|
The Ba/Nb and Zr/Nb ratios (or equivalent Ba/Ta and Zr/Ta) have been used to distinguish between supra-subduction zone (frontal arc) and back-arc magmas (Hildreth & Moorbath, 1988; Kay et al., 1994b; Davidson & de Silva, 1995). In terms of the Ba/Nb ratio and Nb concentration (Fig. 8 and Table 8), the northern Puna centres show both a temporal difference (i.e. UOLM vs MM) and a difference according to geographic location. The UOLM centres have Ba/Nb ratios and Nb values in the field of CVZ arc magmas (Fig. 8) and the eastern UOLM centres have slightly higher Nb contents than the western UOLM, at similar Ba/Nb ratios. The MM samples show generally higher Nb contents than both UOLM groups, and Ba/Nb ratios that vary with geographic position. Those erupted eastward (i.e. Pan de Azucar, and the 12 Ma Huayra Huasi porphyry, located east of Aguiliri) have a more back-arc affinity, with Ba/Nb generally <50, and those erupted to the west (Chinchillas and Aguiliri) have ratios >50, which are typical for the frontal arc.
|
Figure 9 shows a comparison of the REE contents of the UOLM and MM rocks. Generally, the chondrite-normalized patterns are light REE (LREE) enriched, and there is no significant correlation between LREE contents or LREE/heavy REE (HREE) ratios and SiO2 values. The dioritic and tonalitic enclaves have higher HREE concentrations and flatter chondrite-normalized REE patterns than the lavas (Table 8, Fig. 9). Negative Eu anomalies are weak, which at a first glance seems incompatible with major-element fractionation models (see below), which suggest that plagioclase was a major fractionating phase. As suggested by Davidson & de Silva (1995), the weak Eu anomalies could be due to high fO2 conditions in Andean magmas and subsequently low Eu2+/Eu3+ ratios. An outstanding feature of the northern Puna volcanism is a steepening in slope of the REE patterns from the UOLM to the MM centres (Fig. 10a). This steepening involves both a slight increase in LREE contents, raising the La/Sm ratio from 4–7 for UOLM samples to 5–7·5 for MM rocks, and a decrease in total HREE, producing a change in the Sm/Yb ratio (from 2–3·3 for UOLM to 3·5–5 for MM). The overall effect is expressed in the La/Yb ratios, which are 11–23 for the UOLM rocks and 23–34 for the MM centres (Fig. 10b).
|
|
Radiogenic isotopes
Radiogenic Nd, Sr and Pb isotopic data from northern Puna UOLM and MM volcanic rocks are listed in Table 9. Selected samples represent the full range of compositions present at each centre, except in the case of the Pozuelos volcaniclastic sequence and the Cabreria Formation ignimbrites, where only one analysis for each centre was obtained (from the same pumice samples as used for dating). Analytical details are given in Appendix A.
|
UOLM volcanic rocks can be divided into two groups according to their Sr–Nd isotopic compositions (Fig. 11; Table 9), and these correlate with the UOLM-1 and UOLM-2 groups as distinguished by A/CNK ratios (compare Fig. 6). The Pirurayo volcanic complex (UOLM-1) has the lowest 87Sr/86SrT values and the highest NdT (0·706 and -3, respectively), very close to values reported for Neogene andesites from the Central Volcanic Zone (CVZ base-line compositions; Davidson et al., 1991). The UOLM-1 rocks represent the isotopically most depleted signatures (in terms of same compositions) reported from the northern Puna except for the Plio-Pleistocene monogenetic mafic centres (Kay et al., 1994a). The two Pirurayo samples represent the full range of chemical variation at the complex (see Table 9) and their nearly identical 87Sr/86SrT and 143Nd/144NdT ratios support the concept of closed-system fractional crystallization during the magmatic evolution of this complex. The UOLM-2 group (Fig. 6) has much more radiogenic initial Sr isotopic ratios, ranging between 0·7128 and 0·7140, and lower NdT values (–7·6 to -9), similar to those of the Panizos ignimbrites (Ort et al., 1996). Samples from the mid-Miocene dome complexes have ratios intermediate between the two UOLM groups (87Sr/86SrT 0·709–0·710, NdT -6·9 to -8·4) and they more closely resemble the Chilean APVC ignimbrites (Fig. 11).
|
Figure 12 shows the variation of Sr content vs 87Sr/86Sr ratio. Negative correlations in this diagram are typical for CVZ volcanic rocks and are generally taken as evidence for upper-crustal contamination of arc magmas (Harmon et al., 1984; Hildreth & Moorbath, 1988; Davidson et al., 1991; Feeley & Davidson, 1994). Following this interpretation, if we take the 87Sr/86Sr ratios of the Pirurayo complex as the most primitive in the northern Puna, the isotopically more enriched UOLM-2 magmas show a crustal assimilation trend typical of CVZ centres, with lower Sr and higher 87Sr/86SrT ratios. On the other hand, the northern Puna MM rocks show an increase in 87Sr/86SrT relative to Pirurayo, but they also have higher Sr concentrations (90–150 ppm more Sr than Pirurayo). This relationship between high Sr contents and high 87Sr/86SrT ratios is also seen in Aguiliri, the Huayra Huasi porphyry (12 Ma) and the 14–15 Ma Tiomayo Formation ignimbrites (B. L. Coira & Ch.-H. Chen, unpublished data, 2000). Rogers & Hawkesworth (1989) also observed increasing 87Sr/86SrT with rising Sr concentration in CVZ magmas sampled on a west-to-east traverse and attributed it to increased influence of radiogenic, late-Proterozoic subcontinental mantle lithosphere as volcanism migrated progressively eastward. In the case of the UOLM and MM volcanic rocks, the locus of magmatism was roughly the same, so this explanation is unlikely. Instead, we consider that this trend suggests that Sr was not removed by fractional crystallization during crustal assimilation and remained incompatible in the melt. As discussed in the modelling section below, the incompatible behaviour of Sr in this case can be explained by suppressed fractionation of plagioclase as a result of a higher pressure of crystallization. It should be noted that the large-volume Upper Miocene dacitic ignimbrites (dashed oval in Fig. 12), share the more usual trend of more radiogenic Sr isotope ratios accompanied by decreasing Sr concentrations.
|
The Nd isotope composition of the UOLM-1 Pirurayo centre is considerably more radiogenic, with an NdT value of -3, than both the UOLM-2 and MM centres, which overlap considerably in the range of NdT -6·7 to -9). The latter values are similar to reported values for dacitic to rhyolitic Upper Miocene–Pliocene ignimbrites (Ort et al., 1996; Schmitt et al., 2001). Calculated Nd model ages (tDM) for the studied rocks are Proterozoic (1–1·56 Ga; Table 9) and indistinguishable from those of Late Palaeozoic granites from northern Chile (Lucassen et al., 1999a) but somewhat lower than the tDM values derived for Puna lower-crustal felsic granulites (Lucassen et al., 1999b), Chilean basement gneisses (Lucassen et al., 1999a), and southern Puna Lower Palaeozoic metamorphic basement (Becchio et al., 1999).
The Pb-isotope ratios from the studied centres are rather radiogenic in contrast to typical Nazca Plate basalts (Fig. 13). Like other northern Puna igneous rocks (e.g. 6–7 Ma Panizos ignimbrites and shoshonite minor centres), their compositions correspond to the Eastern Cordillera–Southern Altiplano Pb domain identified by Aitcheson et al. (1995). Metamorphic rocks of the southern Puna basement (Becchio et al., 1999) have similar 207Pb/204Pb ratios (18·1–18·7), but lower 208Pb/204Pb (<38·6), suggesting a crustal component with higher time-integrated Th/Pb ratios in the northern Puna magmas. Felsic granulite xenoliths (Lucassen et al., 1999b) plot in the lower range of our 206Pb/204Pb data (Fig. 13). The Pb isotopic provinciality in Andean magmas is generally interpreted to reflect a dominance of crustal lead over mantle-derived lead (Wörner et al., 1992; Aitcheson et al., 1995; Kay et al., 1999) and this is also the case in our study, thus the Pb isotopic variations in the 20–12 Ma northern Puna magmas probably reflect mixing of isotopically variable crustal source rocks.
|
|
DISCUSSION: MAGMA SOURCES AND EVOLUTION |
|---|
Petrographic and geochemical results obtained may be briefly summarized as follows. The first-phase UOLM volcanic rocks (20–17 Ma) form two distinct groupings. The metaluminous to weakly peraluminous UOLM-1 group is mainly andesitic but includes minor dacitic compositions and shows the isotopic signatures with the least crustal affinity, similar to the ‘CVZ base-line’ composition of Davidson et al. (1991)
. The UOLM-2 group is composed exclusively of dacitic to rhyolitic rocks that have variable A/CNK ratios and crustal-like isotopic ratios. MM volcanic centres from the second volcanic phase (14–12 Ma) make up a third compositional group. These are dacitic to rhyolitic in composition and are similar in many respects to the UOLM-2 group. Their radiogenic isotopic ratios are intermediate between those of UOLM-1 and -2. The A/CNK ratios are variable, as in the UOLM-2 group. Concentrations of Y, Sc and the HREE are lower, and the slopes of chondrite-normalized REE patterns are steeper than in UOLM magmas, and the incompatible trace element contents in MM rocks are the highest of all the rocks examined in this study.
The petrogenetic modelling discussed in this section aims to explore two different aspects of the pre-Upper Miocene magmatism in the northern Puna: (1) the role of upper-crustal fractional crystallization or assimilation in causing the observed compositional variability; (2) identification of possible source constraints for the three magma groups of northern Puna.
Fractionation in upper-crustal magma chambers
The petrographic characteristics and major element compositional variations in pre-Upper Miocene volcanic centres from this study, along with the limited isotopic variation within individual centres, are compatible with fractional crystallization as the dominant process in magma evolution.
Table 10 shows the results of least-squares modelling designed to test the fractional crystallization hypothesis for selected centres of the UOLM (Pirurayo, Cabreria) and MM (Pan de Azucar) groups. The results suggest that 20–50% crystallization can account for the major element variations observed in these centres. The model mineral proportions are similar to those observed in diorite inclusions, supporting the inference that the inclusions represent crystal accumulations. The fractionation probably took place in upper-crustal magma chambers, judging from the dominance of plagioclase in the fractionating assemblages (Table 10) and its abundance as a phenocryst phase in the rocks.
|
Ideal fractional crystallization is an oversimplification in some cases. The Pan de Azucar centre, in particular, shows petrographic and compositional evidence for local disequilibrium (zones with sieve texture in plagioclase containing high-Mg biotite), which we attribute to influx of a more mafic magma before eruption. The minor decrease in initial Sr isotope ratios and increase in Nd isotope ratio between the second-event and third-event lavas at Pan de Azucar is consistent with this hypothesis (PA-6 and Iso-2 in Table 9, respectively). Caffe (1999) showed that the isotopic variations are consistent with 5–7 wt % admixing of a mafic magma similar to the Maquinas basalt (see below). Therefore even in this case, fractional crystallization is the controlling process for within-centre chemical variability.
Source constraints: arc magma contamination vs crustal melting
Fractional crystallization in pre-eruptive magma chambers can explain most of the within-centre chemical variations, but differences in trace element contents and isotope ratios between centres, and in particular the differences between the UOLM and MM magmatic groups as a whole, must represent specific characteristics of magma source and/or magma evolution in the northern Puna back-arc. A meaningful appraisal of potential sources and/or contaminants for these magmas depends on information about the compositional and isotopic characteristics of basement rocks and mantle in the Central Volcanic Zone. This information is at present limited by the lack of detailed basement studies in the northern Puna and, more fundamentally, by a lack of deep exposures and paucity of mantle-derived rocks. Nevertheless, existing information allows some useful constraints to be made, as follows:
- outcrops of pre-Cenozoic basement in the southern Puna and in northern Chile show dominantly felsic compositions, with an estimated abundance of <5 vol. % metabasites (Becchio et al., 1999; Lucassen et al., 1999a, 2001). This is also expected to be the case for the lower crust, on the basis of the dominance of felsic compositions among lower-crustal xenoliths recovered from Salta Rift volcanic rocks (Lucassen et al., 1999b) and from gravity and seismic velocity data (Wigger et al., 1994; Zandt et al., 1994; Götze & Kirchner, 1997; Graeber & Asch, 1999; Swenson et al., 2000). Compilations of chemical and isotopic data from well over 200 analyses of basement granitoids and felsic gneisses (Lucassen et al., 2001) give a good estimate of mid- to upper-crust composition in this region.
- The basement in the northern Puna is also mostly intermediate to felsic in composition and of Lower Palaeozoic age but it has been comparatively poorly studied in terms of its geochemical and isotopic composition. The few data available (Coira & Barbieri, 1989; Becchio et al., 1999; Coira et al., 1999), suggest low Sr contents (25–200 ppm) and radiogenic 87Sr/86Sr ratios (0·720–0·760) that are similar to the average values of Lucassen et al. (2001). The only reported Nd isotopic data for northern Puna basement rocks are from crustal xenoliths included in ignimbrites from Coranzuli (Becchio et al., 1999) and Panizos (Ort et al., 1996), which range between Nd -8 and -15. The few Sr and Nd values available are similar to those of the better-studied basement in the southern Puna (Becchio et al., 1999) and contiguous region of northern Chile (Fig. 11). Further evidence for a broadly homogeneous felsic upper-crustal composition in this region is given by a study of siliciclastic sedimentary rocks from Ordovician basins (Bock et al., 2000) in the northern Puna and Altiplano.
- The felsic lower-crustal granulites from the Salta Rift reported by Lucassen et al. (1999b) have moderate 87Sr/86Sr ratios (0·7131–0·7143) and very low Rb/Sr (0·04) compared with the outcropping basement, but similar Nd and Pb isotope ratios (144Nd/143Nd = 0·5121; 206Pb/204Pb = 18·4–18·5; 208Pb/204Pb = 39·5).
- Neogene basalts are exceedingly rare in the Central Volcanic Zone and there is little direct evidence available to assess the composition of the mantle component involved in magma genesis. Mafic magmas erupted after 20 Ma typically have basaltic andesite compositions, and their trace element and isotopic compositions show clear evidence for crustal contamination (Kay et al., 1994a; Davidson & de Silva, 1995). The least-contaminated mafic magmas known in the greater region pre-date the Andean orogenic cycle (Upper Oligocene Chiar Khollu basalt, Bolivia: Davidson & de Silva, 1995; Cretaceous basanites from the Salta Rift, Argentina, and the Oligocene Máquinas alkaline basalt from 31°S in Chile: Kay et al., 1999). The compositions of these rocks suggest that the mantle source was isotopically depleted at the beginning of the Andean orogenic cycle (see Fig. 11).
On the basis of the evidence cited above, neither the exposed basement rocks nor the felsic granulite xenoliths, which may reflect lower-crustal compositions, could represent the sole source for the northern Puna volcanic rocks from this study. The main problem with the mid- to upper-crustal basement rocks is their highly radiogenic Sr isotopic compositions. The felsic granulite xenoliths from the Salta Rift have appropriate Sr isotope ratios for potential source material for the UOLM-2 volcanic rocks, but other compositional features of the granulites are inconsistent with this hypothesis. For example, the Nd and Pb isotopic ratios of the volcanic rocks (144Nd/143Nd 0·5122; 206Pb/204Pb 18·7; 208Pb/204Pb 38·9) differ from those of the granulites (144Nd/143Nd = 0·5121; 206Pb/204Pb = 18·4–18·5; 208Pb/204Pb = 39·5). Furthermore, the UOLM-2 rocks have only slightly higher incompatible element contents than the felsic basement and xenoliths, and a very high degree of partial melting (>60–70%) would be needed if these lithologies were the sole magma source. It seems unlikely that this extent of melting could be achieved thermally, apart from the likelihood that melt would separate from its source well before 50% melting is achieved (Vigneresse et al., 1996).
The 87Sr/86Sr ratios of the MM centres Pan de Azucar and Chinchillas (0·709–0·710) are considerably lower than those of the UOLM-2 rocks (Figs 11 and 12), making a pure crustal source even less likely. The rare metabasic rocks from the basement overlap with the MM rocks in terms of Sr isotope ratios but their initial 144Nd/143Nd ratios are much higher (>0·5124, Becchio et al., 1999; Lucassen et al., 1999b) and rule them out as a potential source.
Therefore we conclude that none of the magmas involved in the northern Puna volcanic centres could be derived from the basement by crustal melting alone. They must be hybrid magmas and the next section presents bulk mixing and assimilation–fractional crystallization (AFC) models to explore probable components involved and their relative proportions.
Bulk mixing and AFC models
In the context of this study, the bulk mixing process is envisioned to involve mantle-derived arc basaltic magmas interacting with partial melts of the lower- or mid–upper-crustal lithologies. Pure mixing is a limiting case unlikely to occur in practice, as the crustal melts will be cooler than basaltic magma and some crystallization of the latter is expected to accompany the mixing. The AFC models take this crystallization into account.
The mafic end-member composition used in the modelling is based on the Maquinas basalts (Kay et al., 1999) except for the Pb isotope ratios, which are considered to have been affected by crustal contamination. For Pb, we use isotope ratios and concentrations from the Cretaceous basanites from the Salta Rift (Kay et al., 1999). Two crustal components are considered. One represents partial melts of the mid- to upper crust and for this we use the mean composition of Palaeozoic granites from northern Chile from Lucassen et al. (1999a), which are considered to represent crustal melts of the local basement. The second component is based on an average composition of felsic granulites from the Salta Rift, which may be more appropriate for assimilation of lower crust. The trace element concentrations and isotope ratios chosen for the crustal components are shown in Appendix B and Table 11, respectively.
|
Mixing models using these end-member compositions can match the isotopic ratios of the different groups of northern Puna volcanic rocks (e.g. 30 wt % crustal component in UOLM-1, 50 wt % in MM and 70 wt % crust in UOLM-2 samples; see Table 11, models D, G and J, and Table 12) but they generally fail to account for the observed trace element concentrations and we consider mixing to be an unsatisfactory explanation for the hybridization process. A possible exception to this is the UOLM-2 rocks, where the mixing and AFC results (below) both yield reasonable solutions.
|
Conventional AFC modelling followed the equations of De Paolo (1981) and the refinement by Aitcheson & Forrest (1994). The bulk distribution coefficients (D values) used in the calculations were based on the stable mineral assemblages predicted for crystallization of the Maquinas basalt composition using the MELTS program of Ghiorso & Sack (1995). Specifically, D values were derived for the solid assemblage after 40–60 wt % crystallization at various pressure conditions and oxygen fugacity fixed at the Ni–NiO buffer (Appendix B). The bulk D values of Sr, in particular, are sensitive to the pressure of crystallization, as this affects the proportion of plagioclase in the assemblage. The graphical method of Aitcheson & Forrest (1994) was used to find reasonable estimates of the variables r (ratio of assimilation to crystallization rates) and 
(proportion of assimilated material) by simultaneous solution of the AFC equations for several trace elements and isotope ratios. Figure 14 shows results for the UOLM-1 samples, where mutual intersections of the AFC curves indicate values of both r and between 0·2 and 0·3. The compositions of UOLM-2 and MM group samples are such that the AFC curves intersect at a lower angle than for UOLM-1 and the estimates for r and are allowed to vary from 0·3 to 0·7. To avoid confusion, we point out that the values for crustal assimilation given below express the mass of assimilant relative to the mass of original magma [Ma/Mm° in the nomenclature of DePaolo (1981)]. The amount of crustal material in the final hybrid magma is greater than this by a factor depending on the degree of fractional crystallization.
|
The AFC model for Pirurayo rocks, as representative for the UOLM-1 group, is consistent with derivation from a Maquinas-type basalt with 15–25% crustal assimilation. Both the felsic granulite (lower-crust) and granite (mid- to upper-crust) compositions yield similar solutions (Table 11, models A–C). The bulk D values for Sr, Ba, Rb and Nd that fit the Pirurayo data are those for a crystallization assemblage at a pressure of 7 kbar, where the ratio of clinopyroxene to plagioclase is about 4:1. At lower pressure (<5 kbar), the proportion of plagioclase increases at the expense of clinopyroxene and this results in low model Sr concentrations and high Ba, Rb, Nd concentrations (Table 11, model C, and Appendix B). The data for Sm, Zr and Nb are not satisfied by any of the AFC models. Modelled Sm is too low, perhaps because bulk D has been overestimated, and the modelled Zr and Nb contents are too high, perhaps because the crystallization of accessory minerals is not considered.
The UOLM-2 compositions can be reasonably explained by AFC models using the same end-members as for UOLM-1, but with a higher ratio of assimilation to fractionation (r = 0·6–0·8). The Aitcheson & Forrest (1994) solution suggests values of between 1·2 and 1·6 (Table 11, model E) for the lower-crustal assimilant, corresponding to 55–60 wt % crustal component in the hybrid magma, or 40 wt % (0·75) if the more radiogenic, Sr-poor granite composition is used for the assimilant. Like the UOLM-1 solutions, bulk D values calculated for fractionation at 7 kbar give the best fit to data overall. The AFC solutions for La, Nd, Yb, Ba and Rb match or slightly exceed observed values regardless of whether granulitic or granitic assimilants are assumed (see Table 12) but the model Sr concentrations are too low if granite is used as the assimilant (except for Casa Colorada, which has very low Sr). As in the UOLM-1 case, the model Sm contents are too low, and Zr and Nb are too high.
The AFC solutions for the MM rocks (Table 11, models H and I) indicate proportions of crustal material in the hybrid magma of 50% using a granulite composition for the assimilant and 40% for the more radiogenic granitic composition. One important difference between the MM and UOLM-2 models concerns the bulk D values for Sr. The Sr concentrations in MM rocks are higher than UOLM-2, and if we assume the same end-member compositions for both, the values for DSr must be lower for the MM case than for UOLM models (appropriate ranges of DSr are 0·54–0·73 for MM, 0·90–1·13 for UOLM, Appendix B). The MELTS calculations show that the fractionating assemblage of Maquinas basalt at 10 kbar (F = 0·5) is 0·94 clinopyroxene, 0·03 plagioclase and 0·03 garnet (Appendix B). The corresponding values for DSr, DNd, DPb and DRb produce a good fit to the observed MM data (Table 12). The model results are reasonable for both granulitic or granitic assimilants but the granulite composition yields a better fit for Yb and Sr.
In all the AFC models calculated for UOLM and MM groups, the solutions failed to match Zr, Nb and Sm concentrations although the fit to other trace elements is acceptable (Table 12). We suggest that this is due to an oversimplification in formulating the bulk distribution coefficients. Only clinopyroxene, plagioclase and garnet were considered (Appendix B), and neglecting accessory phases or oxides that host these elements would lead to bulk D values that are too low, and thus melt concentrations that are too high.
Preferred models
The model results show that the composition of Pirurayo andesites (UOLM-1) can be explained by fractional crystallization of a mafic arc magma like the Maquinas basalt with a moderate degree assimilation (15–25%) of crustal material similar compositionally to the Palaeozoic granites or lower-crustal granulites of the known local basement. Simple mixing of basaltic and crustal components does not fit the observed data. The El Morro quartz-andesitic intrusive rocks have the same trace element characteristics as the Pirurayo andesites and may have formed in a similar way. The observed UOLM-1 compositions are best fitted by an AFC model with low r values (0·3) and pressures (7 kbar) representing mid- to lower-crustal depths.
The model results for the UOLM-2 group do not allow a clear distinction between AFC and bulk mixing. Both solutions give a reasonable fit to the data and the proportion of assimilated crust is roughly similar for both (40–60% and 70%, respectively). However, AFC is considered the more realistic process because the temperature difference between pristine basaltic magma and partial melts of the crust should cause the former to cool and crystallize during the mixing process. The AFC modelling indicates a high assimilation to fractionation rate (r = 0·6, Table 11), which implies a small thermal contrast between mafic magma and assimilated crust in the source region. The D values for models that best fit the observed data correspond to a fractionation assemblage stable at 7 kbar according to the MELTS solution for Maquinas basalt, which implies roughly the same mid- to lower-crustal depths as for the UOLM-1 group.
The MM dacite compositions cannot be explained by bulk mixing and the acceptable AFC solutions (Tables 11 and 12) are similar to those for the UOLM-2 rocks (40–50% assimilation). The modelling is consistent with the concept that both magma groups could have been generated from the same basaltic end-member and similar crustal components. If this is the case, the observed differences in trace element concentrations can be interpreted in terms of different physical conditions of the AFC process. According to the MELTS crystallization modelling, the higher degree of incompatibility for Sr, Ba and LREE in the MM group magmas compared with the UOLM-2 group, and the greater compatibility of HREE and Y, can be caused by fractionation of parental basalts at higher pressures (10 kbar vs 7 kbar). The 15 Ma South Lípez ignimbrites [Morokho and Bonete centres at 21°30'S, Fornari et al. (1993) and Fig. 10] and the northern Tiomayo Formation tuffs at 22°S (Coira et al., 2002), have similar values of Sr, Ba, Nb, Zr, La/Sm and Sm/Yb to those of the contemporary MM rocks studied here, and we suggest that their origin may be similar.
It should be emphasized that the AFC models used to generate Tables 11 and 12 are simple in the sense that they consider only mass-balance and assume that the rate of assimilation and fractionation (the r parameter) is constant during the AFC process. Alternative and potentially more realistic AFC models by Spera & Bohrson (2001) consider energy constraints as well as mass balance. In these energy-constrained AFC models (EC-AFC), assimilation is assumed to occur only via partial melts of the wallrocks, and therefore the r parameter depends on the solidus temperature and melt productivity in the wallrocks. A full comparison of the two methods is beyond the scope of this paper and may not be meaningful for the northern Puna example because the EC-AFC models require constraints on melt productivity and extraction for the crustal component, and these are highly uncertain. We performed a limited comparison of the two methods based on Sr concentration and isotope ratios for the cases presented in models B (UOLM-1) and F (UOLM-2) of Table 11. For the EC-AFC calculation heat capacity and enthalpy values for basaltic magma and crustal assimilant were taken from Bohrson & Spera (2001, table 4), the basaltic magma temperature was 1280°C and the initial crustal temperature was 600°C. The crustal assimilant was assumed to have solidus and liquidus temperatures of 700°C and 900°C, respectively, and a linear increase in melt production between them. Bulk DSr for crustal melting was set at unity, as the assimilant composition used (Palaeozoic granite average) already represents a crustal partial melt. The EC-AFC solution indicates 20% crystallization before assimilation begins. This causes slight changes in element concentrations in the magma, and once assimilation begins, the compositional trends of the hybrid magma are similar to those for conventional AFC models. A useful parameter for comparing the two model results is the degree of assimilation needed to achieve a given isotopic ratio in the hybrid magma. For the AFC model B (Table 11), a ratio of 87Sr/86Sr = 0·7059 is reached after 15 wt % assimilation (proportion of crust added to the original magma) and the EC-AFC solution yields 20 wt % assimilation. For model F in Table 11, the AFC calculation gives 87Sr/86Sr = 0·7134 after 40 wt % crustal contamination and the EC-AFC solution requires 55 wt % assimilation to reach the same ratio. Considering the uncertainty of assumptions inherent in both models, we conclude that the agreement between them is good and that the conventional AFC model appears adequate for this application. More insights on the hybridization process might be gained from a full EC-AFC treatment, but this would require a better understanding of the composition and thermal state of the crust than is now available.
Regional implications
The use of incompatible trace element ratios (Ba/Nb, Ba/Ta, La/Ta, Zr/Nb) to discriminate frontal-arc from back-arc magmas is well established (e.g. Kay et al., 1994a; Davidson & de Silva, 1995) but its application to the UOLM and MM rocks is problematic because their trace element ratios may reflect variations in degree or type of crustal assimilation as well as source magma variations. Nevertheless, the UOLM and MM samples do show systematic differences in Ba/Nb (Fig. 8) and Zr/Nb (Table 8) ratios as described above. In particular, the UOLM magmas have Ba/Nb ratios in the range of the recent CVZ arc and this is also true for those MM centres located in the western part of the study area. The eastern MM centres (Pan de Azucar and Huayra Huasi) have lower Ba/Nb ratios in the range of recent back-arc mafic centres. If the effects of crustal assimilation and fractionation on the Ba/Nb ratios are minor or if they affected the UOLM and MM magmas equally, then these difference in Ba/Nb are evidence that the frontal arc migrated westward with time and that the back-arc setting which the study area now occupies (Fig. 1) became established in the mid-Miocene. This conclusion is somewhat speculative but we note that Ba/Nb ratios of andesitic UOLM-1 rocks are similar to those of dacitic UOLM-2 centres with at least 50% crustal material. This suggests that the effect of crustal assimilation and fractionation on the Ba/Nb ratio may in fact be minor.
The AFC models discussed above imply that the MM group magmas may have undergone hybridization in the crust at a higher pressure than the UOLM-1 and UOLM-2 magmas (>10 kbar vs 7 kbar). This inference depends on several assumptions that cannot be directly tested but it is consistent with tectonic interpretations, which consider the Middle Miocene as the main time for intracrustal shortening and thickening of the Andean orogen (Gubbels et al., 1993; Cladouhos et al., 1994; Hèrail et al., 1994; Kley et al., 1996; Allmendinger et al., 1997). On the other hand, MM rocks lack the high LREE/HREE ratios found in several post-Miocene centres from the Central Andes, which are suggested to indicate extreme crustal thickness (e.g. the Vilama Caldera in northern Puna with La/Yb >26–45: Coira et al., 1996; or the Cordillera Blanca batholith in southern Peru with La/Yb >27–116: Petford & Atherton, 1996). The moderate ratios of MM dacites (La/Yb = 23–33) may indicate that current crustal thicknesses were reached after the Late Middle Miocene (Isacks, 1988; Cladouhos et al., 1994; Allmendinger et al., 1997; Okaya et al., 1997) or, as argued by McMillan et al. (1993), that the effect of crustal thickening is recorded in hybrid magma compositions only after a significant delay.
The large-volume Upper Miocene to Pliocene dacitic ignimbrites erupted in the Puna plateau and southern Altiplano are compositionally similar to the UOLM and MM centres studied here and they have also been suggested to form through AFC processes of crustal assimilation by mafic arc magma (e.g. de Silva et al., 1993; Ort et al., 1996; Lindsay et al., 2001; Schmitt et al., 2001). The degree of assimilation concluded by different workers varies, but there is consensus that the hybrid magmas contain of the order of 50% crust or more. The Upper Miocene Panizos ignimbrite (Ort et al., 1996), for example, has geochemical and isotopic compositions almost identical to those of the UOLM-2 magmas (Fig. 11). Other caldera-sourced dacitic ignimbrites in the area have somewhat different geochemical and isotopic features [i.e. Vilama Caldera in Fig. 10 and in the study by Coira et al. (1996)] but the first-order difference between the northern Puna centres from this study and the Upper Miocene–Pliocene ignimbrites from the same region is the enormous contrast in magma volumes. The increase in volume of hybrid magmas produced since the late Miocene could be a consequence of several processes that are interrelated: (1) a progressive increase in the volume of asthenospheric mantle and extension in the back-arc, both caused by arc migration to the west and steepening of the subduction dip and/or by lithospheric delamination following crustal thickening (Coira et al., 1993; Allmendinger et al., 1997); (2) changes in regional stress conditions (from compressive to tensional or strike-slip) in the northern Puna crust since 10–8 Ma (Allmendinger et al., 1997; Riller et al., 2001), which enhanced and focused emplacement of mafic magmas into the crust; (3) bulk intracrustal convection causing efficient heat transport into compositionally fertile regions of the mid-crust (Babeyko et al., 2002).
|
CONCLUSIONS |
|---|
The pre-Upper Miocene volcanism in the northern Puna of Argentina produced small dome complexes, which are grouped by age into the UOLM (20–17 Ma) and MM (15–12 Ma) volcanic phases. The main results of geochemical and isotopic analyses of representative centres from these two phases can be summarized as follows:
- the UOLM phase produced two distinct groups of rocks, one (UOLM-1) mostly andesitic and metaluminous and the other (UOLM-2) more silicic and with higher and more variable A/CNK ratios. The MM phase (14–12 Ma) produced dacitic to low-Si rhyolitic rocks similar to the UOLM-2 group but with higher concentrations of incompatible trace elements. Some of these centres are associated with Ag–Pb–Zn mineralization.
- Petrography, mineral chemistry and whole-rock compositions indicate that the chemical variations within individual dome complexes reflect low-pressure fractional crystallization before eruption. This is consistent with minor isotopic variability within single centres.
- Sr- and Nd-isotopic compositions of UOLM-1 rocks are distinctive and like those of present CVZ arc andesites (87Sr/86SrT 0·706; NdT -3), whereas the ratios of UOLM-2 and MM rocks overlap and have considerable crustal affinity (87Sr/86SrT 0·710–0·713; NdT -7 to -8). Pb isotope ratios are similar in the three groups and are dominated by crustal Pb. Values overlap with those of other CVZ volcanoes and basement in the Eastern Cordillera–Southern Altiplano domain of Aitcheson et al. (1995).
- Partial melting of known basement lithologies cannot be the sole process for generating the UOLM or MM magmas. Most isotopic and trace element characteristics of the magmas can be explained by an AFC process involving different degrees of fractional crystallization of an arc basaltic magma similar in composition to the Oligocene Maquinas basalt (Kay et al., 1999) and assimilation of felsic crustal melts. Successful AFC solutions suggest that the proportion of crustal material increased from 15–25% in UOLM-1 magmas to 40–60% for UOLM-2 and MM magmas, but cannot, in general, distinguish between a felsic granulite (lower crust) or granitic assimilant (mid- to upper crust).
- The dependence of bulk D values on pressure, especially for Sr, suggests that the depth of magma genesis by AFC was greater for the MM group magmas than for the UOLM group (10 vs 7 kbar). This increase in pressure may reflect the influence of tectonic shortening, which was active since at least the mid-Miocene (Allmendinger et al., 1997).
- Ratios of Ba/Nb and Zr/Nb in the UOLM group are fairly homogeneous and resemble values from the present arc, whereas the MM centres show arc-like ratios in a western group and intraplate affinity in an eastern group. This suggests that the westward shift of the arc towards its present position in the western Cordillera began in this region from 20–17 Ma and 14–12 Ma, and that the back-arc position of the northern Puna was probably established since the mid-Miocene.
We suggest that there is no fundamental difference, compositionally, between magma genesis of the dacitic centres reported here and the large ignimbrite sheets that erupted from caldera complexes in the northern Puna since the Upper Miocene. The difference in volume and mode of eruption may instead reflect differences in crustal stress and thermal state. In the pre-Upper Miocene, the stress pattern was uniformly compressive and this may have restricted intrusion of mafic magmas to the lower crust. In contrast, the extension and transpressive stresses that prevailed since 8 Ma allowed mafic magmas to intrude at more variable depths within a thickened crust. Recent thermal–mechanical modelling by Babeyko et al. (2000, 2002) showed that mafic intrusions alone are not enough to explain the large-scale crustal melting represented by the ignimbrites and present-day geophysical anomalies in the Altiplano–Puna crust. Instead, a higher mantle heat flow is needed, and possible reasons for this include slab retreat and detachment of overthickened mantle lithosphere (Kay et al., 1994).
|
APPENDIX A: ANALYTICAL TECHNIQUES |
|---|
Major and trace elements
Electron microprobe analyses of minerals (Tables 3–7) were performed in a JEOL-JXA 8900-M Superprobe at the Universidad Complutense de Madrid (Spain). Operating conditions were 15 kV and 20 nA, with 10 s counting times and a beam diameter of
5 µm. Standards were from the Smithsonian Institution. The ZAF correction was applied.
Most of the whole-rock major and trace element data (Table 8) were obtained by X-ray fluorescence (XRF) at the laboratory of the Instituto de Geologia y Mineria, Universidad Nacional de Jujuy, Argentina, on a Rigaku FX2000 spectrometer with a Rh tube. Ground and homogenized samples were fused with lithium tetraborate as a flux for major element analyses. Ba, Sr, Rb, Zr, Nb, Hf, Y, Th, and U determinations were performed on rock powder pellets mixed with methyl methacrylate, and pressed at 20 t. Operating conditions were 50 kV and 45 mA. Major and trace elements were analysed by standard methods, using standards from the US Geological Survey and the Japan Geological Survey. Those samples whose LOI calculations are not shown have been analysed for major elements following the fundamental parameters method (Rousseau, 1984).
REE and Sc concentrations were determined by inductively coupled plasma atomic emission spectrometry (ICP-AES) in GeoForschungsZentrum-Potsdam, Germany, following the separation method of Zuleger & Erzinger (1990), and using a Varian Liberty 200 spectrometer. Twenty runs of standard JA-2 showed an analytical precision of 5% for all REE, except Tb (11%) and Tm (8%). Trace elements in ignimbrite samples from the Cabreria Formation (shown in Table 8) were analysed by Activation Laboratories, Canada, using a combination of instrumental neutron activation analysis (INAA) and ICP. One sample shown in Table 8 (CCl-1) was analysed by INAA at the Ward Laboratory in Cornell University following the methods described by Kay et al. (1987).
Radiogenic isotopes
Whole-rock Sr, Nd and Pb isotopic compositions (Table 9) were determined at the GeoForschungsZentrum laboratories using procedures described by Romer et al. (2001). Samples (0·116–0·123 g) were dissolved with 52% HF for 4 days at 160°C on a hot plate. Digested samples were dried and taken up in 6N HCl. Sr and Nd were separated and purified using cation-exchange chromatography. Pb was separated using the HBr–HCl anion-exchange procedure of Tilton (1973). 87Sr/86Sr and 143Nd/144Nd were obtained on a Finnigan MAT262 multi-collector mass spectrometer operated in static mode. Ratios were normalized to 86Sr/88Sr = 0·1194 and 146Nd/144Nd = 0·7219, respectively. Multiple measurement of NBS 987 Sr reference material and La Jolla Nd reference material gave 0·710249 ± 0·000004 (n = 12) and 0·511892 ± 0·000007 (n = 13), respectively. Static 143Nd/144Nd values were adjusted to the value obtained for dynamic measurements (0·511850 ± 0·000004, n = 14). Analytical uncertainties are reported as 2
of the mean. 87Sr/86Sr(T) and Nd(T) were calculated for known K–Ar ages (except sample Min-9, for which an age of 17 Ma was assumed), using 
87Rb = 1·42E -11 y-1 and 147Sm = 6·54E -12 y-1, (147Sm/144Nd)0CHUR = 0·1966, and (143Nd/144Nd)0CHUR = 0·512638, respectively, and the concentration data given in Table 8. Rb and Sr data were determined with XRF with an estimated precision in Rb/Sr better than 1%. Sm and Nd were determined with ICP-AES and the estimated precision is better than 0·5% for Sm/Nd. Lead isotope analyses were performed on a Finnigan MAT262 mass spectometer using static multicollection. Data were corrected for mass discrimination with 0·1%/a.m.u. Reproducibility at the 2 level is better than 0·1%.
|
APPENDIX B |
|---|
|
|
|
ACKNOWLEDGEMENTS |
|---|
We thank C. Casquet and J. Erzinger, who permitted the access to some of the analytical facilities used for this work. P. Zambrana, O. Orosco, H. Justi and A. J. Pérez are gratefully acknowledged for their help in field trips, as are E. G. Baldo, P. Flores, P. Cachizumba, R. Liquín, E. Kramer, C. Schulz and K. Hahne for their assistance in various parts of the analytical work. We are also grateful to J. Lindsay, A. Schmitt, W. Siebel and W. Schnurr for many helpful discussions about magmatic processes and evolution of the Andean CVZ. The German exchange program DAAD supported the stay of P.J.C. in Potsdam during 1998. The fieldwork and part of the geochemical analyses were financed by the Argentine CONICET (National Council of Scientific and Technological Research), both with Ph.D. fellowships to P.J.C. and a grant to B.L.C. (PIP-5017). Additional funding for geochemical studies came from the Agencia de Promoción Científica y Tecnológica de Argentina (PICT-00511, to B.L.C.), and the Special Research Program SFB-267 ‘Deformation processes in the Andes’ supported by the Deutsche Forschungsgemeinschaft (DFG). Reviews by Shan de Silva, Todd Feeley and Chris Nye, as well as comments and editorial suggestions by Marjorie Wilson, highly improved the quality of the original manuscript. Finally, we wish to thank the people from the Puna of Jujuy for their hospitality, especially the personnel and teachers of many rural schools.
|
FOOTNOTES |
|---|
*Corresponding author. Telephone: 54 388 4221593. Fax: 54 388 4232957. E-mail: pabcaf{at}idgym.unju.edu.ar
|
REFERENCES |
|---|
Aitcheson, S. J. & Forrest, A. H. (1994). Quantification of crustal contamination in open magmatic systems. Journal of Petrology 35, 461–488.
Aitcheson, S. J., Harmon, R. S., Moorbath, S., Schneider, A., Soler, P., Soria Escalante, E., Steele, G., Swainbank, I. & Wörner, G. (1995). Pb isotopes define basement domains of the Altiplano, central Andes. Geology 23, 555–558.
Allmendinger, R. W., Jordan, T. E., Kay, S. M. & Isacks, B. L. (1997). The evolution of the Altiplano–Puna of the Central Andes. Annual Review of Earth and Planetary Sciences 25, 139–174.[Web of Science]
Aniel, B. G. (1987). Comportement naturel de l’uranium en milieu volcanique et approche expérimentale. Ph.D. thesis, Université de Nancy.
Babeyko, A. Yu., Sobolev, S. V., Trumbull, R. B., Lavier, L. & Oncken, O. (2000). Thermomechanical modelling of large-scale crustal melting in the middle crust of the Altiplano–Puna plateau. AGU Fall Meeting, San Francisco, 15–19 December. EOS Transactions, American Geophysical Union 81, F1136–F1137.
Babeyko, A. Y., Sobolev, S. V., Trumbull, R. B., Oncken, O. & Lavier, L. (2002). Numerical models of crustal scale convection and partial melting beneath the Altiplano–Puna plateau. Earth and Planetary Sciences Letters (submitted).
Bacon, C. R. & Druitt, T. H. (1988). Compositional evolution of the zoned calcalkaline magma chamber of Mount Mazama Crater Lake, Oregon. Contributions to Mineralogy and Petrology 98, 224–256.
Becchio, R., Lucassen, F., Kasemann, S., Franz, G. & Viramonte, J. (1999). Geoquímica y sistemática isotópica de rocas metamórficas del Paleozoico inferior: Noroeste de Argentina y Norte de Chile (21°–27°S). Acta Geológica Hispánica 34, 273–299.
Bock, B., Bahlburg, H., Wörner, G. & Zimmermann, U. (2000). Tracing crustal evolution in the southern Central Andes from Late Precambrian to Permian using Nd and Pb isotopes. Journal of Geology 108, 515–535.
Bohrson, W. A. & Spera, F. J. (2001). Energy-constrained open-system magmatic processes II: application of energy-constrained assimilation–fractional crystallization (EC–AFC) model to magmatic systems. Journal of Petrology 42, 1019–1041.
Caffe, P. J. (1996). La dacita de Casa Colorada. Complejo volcánico dómico del Terciario superior en Puna Norte, Argentina. Memorias del XII Congreso Geológico de Bolivia (Tarija, Bolivia) 3, 1019–1030.
Caffe, P. J. (1999). Complejos volcánicos dómicos del Terciario superior de Puna Norte: sus implicancias magmatotectónicas y metalogenéticas. Ph.D. thesis, Universidad Nacional de Córdoba, Argentina, 421 pp.
Caffe, P. J. & Coira, B. (1999). Complejos de domos volcánicos del Mioceno medio de Puna Norte. Un modelo geológico y metalogenético para yacimientos epitermales de metales de base ricos en Ag (Sn). In: Zappettini, E. (ed.) Recursos Minerales de la República Argentina. Instituto de Recursos Minerales, SEGEMAR. Anales 35, 1569–1578.
Chernicoff, C. J., Garea, G., Hongn, F., Seggiaro, R., Zappettini, E., Coira, B., Caffe, P. J., Chayle, W., Rodríguez, G. A., Pérez, A., Soler, M. M. & Rankin, L. (1996). Interpretación geológica del relevamiento aeromagnético de la Puna Septentrional, Jujuy y Salta. Dirección Nacional del Servicio Geológico, Serie Contribuciones Técnicas 1, 46 pp.
Cladouhos, T. T., Allmendinger, R. W., Coira, B. & Farrar, E. (1994). Late Cenozoic deformation in the Central Andes: fault kinematics from the Northern Puna, Northwestern Argentina and Southwestern Bolivia. Journal of South American Earth Science 7, 209–228.
Coira, B. (1979). Descripción Geológica de la Hoja 3c, Abra Pampa, Pcia. de Jujuy. Boletín del Servicio Geológico Argentino 170, 1–90.
Coira, B. (1994). Metallogenic events in the framework of magmatic–tectonic evolution of the Northern Puna of Argentina during the late Cenozoic. Comunicaciones 45, 67–76.
Coira, B. & Barbieri, M. (1989). Relaciones isotópicas 87Sr/86Sr de vulcanitas cenozoicas (22°S–24°30’S) Puna, Argentina. Revista de la Asociación Geológica Argentina 64, 78–88.
Coira, B., Kay, S. M. & Viramonte, J. G. (1993). Upper Cenozoic magmatic evolution of the Argentine Puna—a model for changing subduction geometry. International Geology Review 35, 677–720.
Coira, B., Caffe, P. J., Kay, S. M., Díaz, A. & Ramirez, A. (1996). Complejo volcánico Vilama—Sistema caldérico del Cenozoico superior en Puna, Jujuy. XIII Congreso Argentino de Geología, Buenos Aires, Actas 3, 603–620.
Coira, B. L., Kay, S. M., Pérez, B., Woll, B., Hanning, M. & Flores, P. (1999). Magmatic sources and tectonic setting of Gondwana margin Ordovician magmas, northern Puna of Argentina and Chile. In: Ramos, V. & Keppie, D. (eds) Laurentian–Gondwanan Connections before Pangea. Geological Society of America, Special Paper 336, 1–26.
Coira, B., Caffe, P. J., Ramírez, A., Chayle, W., Díaz, A., Rosas, S. A., Pérez, A., Pérez, E. M. B., Orosco, O. & Martínez, M. (2002). Hoja Geológica 2366-I Mina Pirquitas (1:250 000). Buenos Aires: SEGEMAR, Secretaría de Minería de la Nación (in press).
Cunningham, C., McNamee, J., Vázquez Pinto, J. & Ericksen, G. (1991). A model of volcanic dome-hosted precious metal deposits in Bolivia. Economic Geology 86, 415–421.
Davidson, J. P. & de Silva, S. L. (1995). Late Cenozoic magmatism of the Bolivian Altiplano. Contributions to Mineralogy and Petrology 119, 387–408.[Web of Science]
Davidson, J. P., Harmon, R. S. & Wörner, G. (1991). The source of central Andean magmas: some considerations. In: Harmon, R. S. & Rapela, C. W. (eds) Andean Magmatism and its Tectonic Setting. Geological Society of America, Special Publication 265, 233–243.
DePaolo, D. J. (1981). Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization. Earth and Planetary Science Letters 53, 189–202.[Web of Science]
de Silva, S. L. (1987). Large volume explosive silicic volcanism in the Central Andes of North Chile. Ph.D. thesis, Open University, Milton Keynes, 409 pp.
de Silva, S. L. (1989a). The Altiplano–Puna volcanic complex of the central Andes. Geology 17, 1102–1106.
de Silva, S. L. (1989b). The origin and significance of crystal rich inclusions in pumices from two Chilean ignimbrites. Geological Magazine 12, 159–175.
de Silva, S. L., Davidson, J. P., Croudace, I. W. & Escobar, A. (1993) Volcanological and petrological evolution of volcan Tata Sabaya, S.W., Bolivia. Journal of Volcanology and Geothermal Research 55, 305–335.[Web of Science]
Droop, G. (1987). A general equation for estimating Fe3+ concentrations in ferromagnesian silicates and oxides from microprobe analyses, using stoichiometric criteria. Mineralogical Magazine 51, 431–435.
Ewart, A. & Griffin, W. L. (1994). Application of proton-microprobe data to trace-element partitioning in volcanic rocks. Chemical Geology 117, 251–284.
Feeley, T. C. & Davidson, J. P. (1994). Petrology of calc-alkaline lavas at Volcán Ollagüe and the origin of compositional diversity at Central Andean stratovolcanoes. Journal of Petrology 35, 1295–1340.
Fornari, M., Pozzo, L., Soler, P., Bailly, L., Leroy, J. & Bonhomme, M. G. (1993). Miocene volcanic centers in the southern Altiplano of Bolivia, the Cerro Morokho and Cerro Bonete area (Sur lípez). Second International Symposium on Andean Geogynamics, Oxford (UK), Abstract Volume. Paris: ORSTOM, pp. 363–366.
Francis, P. W., Sparks, R. S. J., Hawkesworth, C. J., Thorpe, R. S., Pyle, D. M., Tait, S. R., Mantovani, M. S. & McDermott, F. (1989). Petrology and geochemistry of volcanic rocks of the Cerro Galan caldera, northwest Argentina. Geological Magazine 126, 515–547.[Abstract]
Gangui, A. (1998). Seismic stratigraphy of the Pozuelos basin: Northern Puna, Argentina. X Congreso Latinoamericano de Geología, Buenos Aires, Actas 1, 12–17.
Ghiorso, M. S. & Sack, R. O. (1995). Chemical mass transfer in magmatic processes. IV. A revised and internally consistent thermodynamic model for the interpolation and extrapolation of liquid–solid equilibria in magmatic systems at elevated temperatures and pressures. Contributions to Mineralogy and Petrology 119, 197–212.[Web of Science]
Götze, H. J. & Kirchner, A. (1997). Interpretation of gravity and geoid in the Central Andes between 20° and 29°S. Journal of South American Earth Sciences 10, 179–188.
Graeber, F. M. & Asch, G. (1999). Three-dimensional models of P-wave velocity and P- to S- velocity ratio in the central Andes by simultaneous inversion of local earthquake data. Journal of Geophysical Research 104, 20237–20256.
Gubbels, T. L., Isacks, B. L. & Farrar, E. (1993). High level surfaces, plateau uplift, and foreland development, Central Bolivian Andes. Geology 21, 695–698.
Halls, C. & Schneider, A. (1988). Comentarios sobre la génesis de los yacimientos del cinturón estannífero boliviano. Revista Geológica de Chile 15, 41–56.
Harmon, R. S., Barreiro, B. A., Moorbath, S., Hoefs, J., Francis, P. W., Thorpe, R. S., Derruelle, B., McHugh, J. & Viglino, J. A. (1984). Regional O-, Sr-, and Pb- isotope relationships in late Cenozoic calc alkaline lavas of the Andean Cordillera. Journal of the Geological Society, London 141, 803–822.
Hauri, E. H., Wagner, T. P. & Grove, T. L. (1994). Experimental and natural partitioning of Th, U, Pb and other trace elements between garnet, clinopyroxene and basaltic melts. Chemical Geology 117, 149–166.[Web of Science]
Hèrail, G., Baby, P. & Soler, P. (1994). El contacto Cordillera Oriental–Altiplano en Bolivia: evolución tectónica, sedimentaria y geomorfológica durante el Mioceno. VII Congreso Geológico Chileno, Concepción, Actas 1, 62–66.
Hickey, R., Frey, F. & Gerlach, D. (1986). Multiple sources for basaltic arc rocks from the Southern Volcanic Zone of the Andes (34°–41°S): trace element and isotopic evidence for contributions from subducted oceanic crust, mantle, and continental crust. Journal of Geophysical Research 91, 5963–5983.
Hildreth, W. & Moorbath, S. (1988). Crustal contributions to arc magmatism in Andes of Central Chile. Contributions to Mineralogy and Petrology 98, 455–499.[Web of Science]
Holland, T. & Blundy, J. (1994). Non-ideal interactions in calcic amphiboles and their bearing on amphibole–plagioclase geothermometry. Contributions to Mineralogy and Petrology 116, 433–447.
Irving, A. J. & Frey, F. A. (1978). Distribution of trace elements between garnet megacrysts and host volcanic liquids of kimberlitic to rhyolitic composition. Geochimica et Cosmochimica Acta 42, 743–770.
Isacks, B. L. (1988). Uplift of the central Andean plateau and bending of the Bolivian orocline. Journal of Geophysical Research 93, 3211–3231.
Jordan, T. E. & Alonso, R. (1987). Cenozoic stratigraphy and basin tectonics of the Andes Mountains. 20°–28°, South latitude. AAPG Bulletin 71, 49–56.[Abstract]
Kay, S. M., Maksaev, V., Moscoso, R., Mpodozis, C. & Nasi, C. (1987). Probing the evolving Andean lithosphere: mid–late Tertiary magmatism in Chile (29°–30°30’S) over the modern zone of subhorizontal subduction. Journal of Geophysical Research 92, 6173–6189.
Kay, S. M., Coira, B. & Viramonte, J. G. (1994a). Young mafic backarc volcanic rocks as indicators of continental lithospheric delamination beneath the Argentine Puna Plateau, Central Andes. Journal of Geophysical Research 99, 24323–24339.
Kay, S. M., Mpodozis, C., Tittler, A. & Cornejo, P. (1994b). Tertiary magmatic evolution of the Maricunga Belt in Chile. International Geology Review 36, 1079–1112.
Kay, S. M., Mpodozis, C. & Coira, B. (1999). Neogene magmatism, tectonism, and mineral deposits of the Central Andes (22°S to 33°S). In: Skinner, B. (ed.) Geology and Ore Deposits of the Central Andes. Society of Economic Geologists Special Publication 7, 27–59.
Kley, J., Muller, J., Tawackoli, S., Jacobshagen, V. & Manutsoglu, E. (1996). Pre-Andean and Andean-age deformation in the Eastern Cordillera of Southern Bolivia. Journal of South American Earth Sciences 10, 1–19.
Kley, J., Monaldi, C. R. & Salfity, J.A. (1999). Along-strike segmentation of the Andean foreland: causes and consequences. Tectonophysics 301, 75–94.
Kretz, R. (1983). Symbols for rock-forming minerals. American Mineralogist 68, 277–279.[Abstract]
Linares, E. & González, R. R. (1990). Catálogo de edades radimétricas de la República Argentina 1957–1987. Asociación Geológica Argentina, Publicaciones Especiales 19, 628 pp.
Lindsay, J., Schmitt, A., Trumbull, R., de Silva, S.L., Siebel, W. & Emmermann, R. (2001). Magmatic evolution of the La Pacana Caldera system, Central Andes, Chile: compositional variation and contrasting eruption mechanisms of two cogenetic, large volume felsic ignimbrites. Journal of Petrology 42, 459–486.
Lucassen, F., Franz, G., Thirlwall, M. F. & Mezger, K. (1999a). Crustal recycling of metamorphic basement: Late Paleozoic granitoids of northern Chile (22°S). Implications for the composition of the Andean crust. Journal of Petrology 40, 1527–1551.
Lucassen, F., Lewerenz, S., Franz, G., Viramonte, J. & Mezger, K. (1999b). Metamorphism, isotopic ages and composition of lower crustal granulite xenoliths from the Cretaceous Salta rift, Argentina. Contributions to Mineralogy and Petrology 134, 325–341.
Lucassen, F., Becchio, R., Harmon, R., Kasemann, S., Franz, G., Trumbull, R., Wilke, H-G., Romer, R. L. & Dulski, P. (2001). Composition and density model of the continental crust at an active continental margin—the Central Andes between 21° and 27°S. Tectonophysics 341, 195–223.
Marinovic, N. (1979). Geología de los cuadrángulos Zapaleri y Nevados de Poquis, II Región, Antofagasta. B.Sc. thesis, Universidad de Chile, Santiago, 75 pp.
Mathison, C. I. (1975). Magnetites and ilmenites in the Somerset Dam layered basic intrusion, southeastern Queensland. Lithos 8, 93–111.
Matthews, S. J., Sparks, R. S. J. & Gardeweg, M. C. (1999). The Piedras Grandes–Soncor eruptions, Lascar volcano, Chile; evolution of a zoned magma chamber in the Central Andean upper crust. Journal of Petrology 40, 1891–1919.
McMillan, N. J., Davidson, J. P., Wörner, G., Harmon, R. S., Moorbath, S. & López-Escobar, L. (1993). Influence of crustal thickening on arc magmatism: Nevados de Payachata volcanic region, northern Chile. Geology 21, 467–470.
Okaya, N., Tawackoli, S. & Giese, P. (1997). Area-balanced model of the Late Cenozoic tectonic evolution of the Central Andean arc and back arc (lat 20°–22°S). Geology 25, 367–370.
Ort, M., Coira, B. & Mazzoni, M. (1996). Geochemical and isotopic evidence for magma sources and contamination in the crystal rich dacitic volcanic rocks of Cerro Panizos, central Andes. Contributions to Mineralogy and Petrology 123, 308–322.
Patiño Douce, A. E. (1993). Ti substitution in biotite: an empirical model with applications to thermometry, O2 and H2O barometries, and consequences for biotite stability. Chemical Geology 108, 133–162.
Peccerillo, R. & Taylor, S. R. (1976). Geochemistry of Eocene calc-alkaline volcanic rocks from the Kastamonu area, northern Turkey. Contributions to Mineralogy and Petrology 58, 63–81.[Web of Science]
Petford, N. & Atherton, M. (1996). Na-rich partial melts from newly underplated basaltic crust: the Cordillera Blanca Batholith, Peru. Journal of Petrology 37, 1491–1521.
Pichavant, M., Kontak, D. J., Brique, L., Valencia Herrera, J. & Clark, A. H. (1988). The Miocene Macusani volcanics, SE Perú II. Geochemistry and origin of a felsic peraluminous magma. Contributions to Mineralogy and Petrology 100, 325–338.
Redwood, S. D. & Rice, C. M. (1997). Petrogenesis of Miocene basic shoshonitic lavas in the Bolivian Andes and implications for hydrothermal gold, silver and tin deposits. Journal of South American Earth Sciences 10, 203–221.[Web of Science]
Riller, U., Petrinovic, I., Ramelow, J., Strecker, M. & Oncken, O. (2001). Late Cenozoic tectonism, collapse caldera and plateau formation in the central Andes. Earth and Planetary Science Letters 188, 299–311.
Rogers, G. & Hawkesworth, C. J. (1989). A geochemical transverse across the North Chilean Andes: evidence of crust generation from the mantle wedge. Earth and Planetary Science Letters 91, 271–285.
Romer, R. L., Foerster, H.-J. & Breitkreuz, Ch. (2001). Intracontinental extensional magmatism with a subduction fingerprint: the late Carboniferous Halle Volcanic Complex (Germany). Contributions to Mineralogy and Petrology 141, 201–221.
Rousseau, N. (1984). Fundamental algorithm between concentration and intensity in XRF analysis. X Ray Spectrometry 13, 115–120.
Salfity, J. A. (1982). Evolución paleogeográfica del Grupo Salta (Cretácico–Eógeno). V Congreso Latinoamericano de Geología, Carlos Paz, Argentina, Actas 2, 11–26.
Schmitt, A. K., de Silva, S. L., Trumbull, R. B. & Emmermann, R. (2001). Magma evolution in the Purico ignimbrite complex, northern Chile: evidence for zoning of a dacitic magma by injection of rhyolitic melts following mafic recharge. Contributions to Mineralogy and Petrology 140, 680–700.
Schneider, A. (1985). Eruptive processes, mineralization and isotopic evolution of the Los Frailes–Kari Kari region, Bolivia. Ph.D. thesis, Imperial College of Science and Technology, London, 281 pp.
Schwab, K. & Lippolt, H. (1974). K–Ar mineral ages and late Cenozoic history of the Salar Cauchari area (Argentine Puna). Proceedings of Symposium on Andean and Antarctic Volcanology, Santiago, 693–714.
Sheffels, B. M. (1990). Lower bound on the amount of crustal shortening in the Central Bolivian Andes. Geology 18, 812–815.
Soler, M. M. (1996). Geología del Complejo Volcánico Pirurayo, Puna de Jujuy. B.Sc. thesis, Universidad Nacional de Córdoba, 119 pp.
Soler, P. & Jiménez, C. (1993). Magmatic contraints upon the evolution of the Bolivian Andes since late Oligocene times. Second International Symposium on Andean Geodynamics, Oxford (UK), Abstract Volume. Paris: ORSTOM, pp. 447–451.
Spera, F. J. & Bohrson, W. A. (2001). Energy-constrained open-system magmatic processes I: general model and energy-constrained assimilation and fractional crystallization (EC-AFC) formulation. Journal of Petrology 42, 999–1018.
Swenson, J. L., Beck, S. L. & Zandt, G. (2000). Crustal structure of the Altiplano from broadband regional waveform modeling: implications for the composition of thick continental crust. Journal of Geophysical Research 105, 607–621.
Tilton, G. R. (1973). Isotopic lead ages of chondritic meteorites. Earth and Planetary Science Letters 19, 321–329.
Tilton, G. R. & Barreiro, B. (1980). Origin of lead in Andean calc-alkaline lavas, southern Peru. Science 210, 1245–1247.
Trumbull, R. B., Wittenbrink, R., Hahne, K., Emmermann, R., Büsch, W. & Gerstenberger, H. (1999). Evidence of Late Miocene to Recent contamination of arc andesites by crustal melts in the Chilean Andes (25°–26°S) and its geodynamic implications. Journal of South American Earth Sciences 12, 135–155.
Turner, J. C. M. (1960). Estratigrafía de la Sierra de Santa Victoria y adyacencias. Boletín de la Academia Nacional de Ciencias de Córdoba 41, 163–169.
Unruh, D. M. & Tatsumoto, M. (1976). Lead isotopic composition and uranium, thorium and lead concentrations in sediments and basalts from the Nazca plate. In: Initial Reports of the Deep Sea Drilling Project, 34. Washington, DC: US Government Printing Office, pp. 341–347.
Vigneresse, J.-L., Barbey, P. & Cuney, M. (1996). Rheological transitions during partial melting and crystallisation with application to felsic magma segregation and transfer. Journal of Petrology 37, 1579–1600.
Wigger, P. J., Schmitz, M., Araneda, M., Asch, G., Baldzuhn, S., Giese, P., Heisohn, W. D., Martinez, E., Ricaldi, E., Rüwer, P. & Viramonte, J. (1994). Variation in the crustal structure of the Southern Central Andes, deduced from seismic refraction investigations. In: Reutter, K. J., Scheuber, E. & Wigger, P. J. (eds) Tectonics of the Southern Central Andes. Heidelberg: Springer, pp. 23–48.
Wörner, G., Harmon, R. S., Davidson, J. P., Moorbath, S., Turner, D. L., McMillan, N., Nye, C., López Escobar, L. & Moreno, H. (1988). The Nevados de Payachata volcanic region (18°S/69°W, N. Chile). 1—Geological, geochemical, and isotopic observations. Bulletin of Volcanology 50, 287–303.
Wörner, G., Moorbath, S. & Harmon, R. S. (1992). Andean Cenozoic volcanic centers reflect basement isotopic domains. Geology 20, 1103–1106.
Yuan, X., Sobolev, S. V., Kind, R., Oncken, O., Bock, G., Asch, G., Schurr, B., Graeber, F., Rudloff, A., Hanka, W., Wylegalla, K., Tibi, R., Haberland, Ch., Rietbrock, A., Giese, P., Wigger, P., Rower, P., Zandt, G., Beck, S., Wallace, T., Pardo, M. & Comte, D. (2000). Subduction and collision processes in the Central Andes constrained by converted seismic phases. Nature 408, 958–961.[Medline]
Zandt, G., Velasco, A. A. & Beck, S. L. (1994). Composition and thickness of the southern Altiplano crust, Bolivia. Geology 22, 1003–1006.
Zentilli, M., Doe, B., Hedge, C., Alvarez, O., Tidy, E. & Daroca, J. (1988). Isótopos de Pb en yacimientos de tipo pórfiro cuprífero comparados con otros depósitos metalíferos en los Andes del N de Chile y Argentina. V Congreso Geológico Chileno, Santiago, pp. 331–369.
Zuleger, E. & Erzinger, J. (1990). Determination of REE and Y in silicate materials with ICP-AES. Fresenius Zeitschrift für Analytische Chemie 332, 140–143.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. T. Drew, M. N. Ducea, and L. M. Schoenbohm Mafic volcanism on the Puna Plateau, NW Argentina: Implications for lithospheric composition and evolution with an emphasis on lithospheric foundering Lithosphere, October 1, 2009; 1(5): 305 - 318. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. De Silva, G. Zandt, R. Trumbull, J. G. Viramonte, G. Salas, and N. Jimenez Large ignimbrite eruptions and volcano-tectonic depressions in the Central Andes: a thermomechanical perspective Geological Society, London, Special Publications, January 1, 2006; 269(1): 47 - 63. [Abstract] [PDF]
|
|
|
|
 |
|
 T. KURITANI, H. KITAGAWA, and E. NAKAMURA Assimilation and Fractional Crystallization Controlled by Transport Process of Crustal Melt: Implications from an Alkali Basalt-Dacite Suite from Rishiri Volcano, Japan J. Petrology, July 1, 2005; 46(7): 1421 - 1442. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||