Journal of Petrology Advance Access published online on February 10, 2008
Journal of Petrology, doi:10.1093/petrology/egn004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Lithospheric Origin of Oligocene–Miocene Magmatism in Central Chile: U–Pb Ages and Sr–Pb–Hf Isotope Composition of Minerals
1Universite De Nice–Sophia Antipolis, Geosciences Azur (UMR 6526), PARC Valrose, F-06108 Nice, France
2Universidad De Chile, Departamento De Geologia, Plaza Ercilla 803, Casilla 13518, Correo 21, Santiago, Chile
Received May 16, 2007; Revised typescript accepted January 9, 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL BACKGROUNDANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Establishing the petrogenesis of volcanic and plutonic rocks is a key issue in unraveling the evolution of distinct subduction-related tectonic phases occurring along the South American margin. This is particularly true for Cenozoic times when large volumes of magma were produced in the Andean belt. In this study we have focused on Oligo-Miocene magmatism in central Chile at 33°S. Our data include field and petrographic observations, whole-rock major and trace element analyses, U–Pb zircon dating, and Pb, Sr, and Hf isotope analyses of plagioclase, clinopyroxene, and zircon mineral separates. Combined with earlier dating results the new zircon ages define a 28·8–5·2 Ma period of plutonic and volcanic activity that ceased as a consequence of flattening subduction of the Nazca–Farallon plate. Rare earth elements patterns are variable, with up to 92 times chondrite concentrations for light rare earth elements yielding (La/Yb)N between 3·6 and 7·0, and an absence of Eu anomalies. Initial Pb isotope signatures are in the range of 18·358–19·023 for 206Pb/ 204Pb, 15·567–15·700 for 207Pb/ 204Pb and 38·249–39·084 for 208Pb/ 204Pb. Initial 87Sr/ 86Sr are mostly in the range of 0·70369–0·70505, with two more radiogenic values at 0·7066. Initial Hf isotopic compositions of zircons yield exclusively positive
Hfi ranging between + 6·9 and + 9·6. The newly determined initial isotope characteristics of the Oligo-Miocene magmas suggest that the mantle source lithologies are different from both those of Pacific mid-ocean ridge basalt and ocean island basalt, plotting in the field of reference values for subcontinental lithospheric mantle, characterized by moderate large ion lithophile element–high field strengh element depletion and high 238U/ 204Pb. A Hf model age of 2 Ga is estimated for the formation of the subcontinental mantle–continental crust assemblage in the region, suggesting that the initial Sr and Pb isotope ratios inferred for the source of the Oligo-Miocene parental magmas are the result of later Rb and U enrichment caused by mantle metasomatism. A time-integrated model Rb/Sr of 
0·039 and µ 16 are estimated for the source of the parental magmas, consistent with ratios measured in peridotite xenoliths from continental areas. Evolution from predominant (>90%) basaltic–gabbroic to andesitic–dioritic magmas seems to involve a combination of (1) original trace element differences in the metasomatized subcontinental mantle, (2) different degrees of partial melting and (3) fractional crystallization in the garnet- and spinel-peridotite stability fields. The genesis of more differentiated magmas reaching rhyolitic–granitic compositions most probably also includes additional crystal fractionation at both shallow mantle depths and within the crust, possibly leading to some very minor assimilation of crustal material. KEY WORDS: calc-alkaline magmatism; Oligo-Miocene; U–Pb dating; Sr–Pb–Hf isotopes; central Chile
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Mesozoic and Cenozoic construction of the Andean belt is characterized by eastwards migration of magmatic arcs and crustal thickening, leading to the formation of high mountain chains generated during different tectonic phases and settings, related to varying plate configurations (e.g. Ramos, 2000
). During Late Eocene and Early Oligocene times, subduction occurred at a rate of 6 cm/year, increasing to 15 cm/year during Late Oligocene and Early Miocene times. At 
20 Ma the roughly orthogonal subduction direction rotated by about 10° in a northwards direction, associated with a decrease in convergence rate. This study is focused on the determination of the precise age and origin of Cenozoic volcanism and plutonism in the central part of the Andean belt at 33°S, where magmatism ceased about 5 Myr ago, associated with progressive flattening of the subducted slab (e.g. Ramos, 2000; Deckart et al., 2005). U–Pb zircon dating was undertaken to clarify ambiguities in the interpretation of earlier K–Ar and 40Ar/39Ar dates from the region, which are often affected by low-grade metamorphism and hydrothermal activity, both post-dating volcanic and plutonic rock emplacement. It is important to note that the rocks analyzed here were previously investigated by 40Ar/39Ar mineral dating (P. Montecinos, unpublished data), which revealed complex patterns from which no reliable ages could be derived.
Both zircon dating and isotope measurements were undertaken on fresh hand-picked minerals extracted from the least metamorphosed rocks, out of a series of 162 samples. U–Pb and Hf isotope analyses of zircon, and Pb and Sr isotopic analyses of feldspar and clinopyroxene were made on the same rocks. Some whole-rocks were also analyzed for major and trace elements. Because (1) these rocks are young, (2) Lu concentrations in zircon are extremely low, and (3) U and Rb concentrations are very low in plagioclase and clinopyroxene (e.g. Schärer, 1991; Schärer et al., 1997), the measured isotopic ratios directly define the time-integrated initial ratios of the source lithologies at the time of melting.
|
GEOLOGICAL BACKGROUND |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Figure 1a shows a schematic geological map of central Chile (33–36°S), two cross-sections, and a more detailed map of the study area (Fig. 1c) with sample locations. The Cenozoic volcanic–plutonic rocks form part of a north–south-oriented belt, trending parallel to the continental margin and extending over about 1300 km from 23 to 35°S (Nyström et al., 2003
). The thickness of the volcanic and plutonic units varies between about 2000 and 6000 m, constituting both the Principal and Frontal Cordilleras. Based on radiometric ages, these are subdivided into the Early Miocene Abanico Formation and the Middle to Late Miocene Farellones Formation (e.g. Nyström et al., 2003; Kay et al., 2005); however, this distinction is hampered by the fact that both formations contain similar lithologies. All the rock types contain secondary minerals such as prehnite, pumpellyite, actinolite, chlorite, sericite, calcite, zeolite, and epidote, produced during non-deformational low-grade metamorphism and fluid circulation (Levi et al., 1989; Aguirre et al., 2000). Tectonic activity associated with the emplacement of earlier Oligocene magmatic rocks was extensional, whereas subsequent activity during Miocene times changed to a compressional regime, reflecting convergence rates increasing from 6 to 15 cm/year (Charrier et al., 2002; Kay et al., 2005).
|
The Abanico and Farellones Formations comprise basalts, basaltic andesites, and andesites, and some more differentiated dacites and rhyolites (Nyström et al., 2003
; Kay et al., 2005). Corresponding plutonic bodies range from gabbro to diorites, granodiorites and minor granites. Associated minor lithologies are pyroclastic flows and breccias, tuffs, and rhyolitic ashes. Basalts–gabbros and andesites–diorites are by far the most abundant lithologies, making up at least 90% of the magmatic rocks along the belt. There seems to be an evolutionary trend from the lower to the upper part of the series, with 80% basalts and 20% andesites in the basal part and 75% andesites, 20% basalts, and 5% dacites in the upper part. Intrusion of the largest (15 km wide) granodioritic stocks into the andesites has produced up to 3 km wide epidote–actinolite–hornblende contact metamorphic aureoles.
Major and trace element geochemistry
The calc-alkaline character of the rocks composing the Oligo-Miocene magmatic belt has been documented in a series of studies (Fig. 2; Vergara et al., 1988; Nyström et al., 2003; Kay et al., 2005; Muñoz et al., 2006). They contain between 46·2 and 74·5 wt% SiO2, correlating with the observed lithological variations of the volcanic and plutonic members, and have (La/Yb)N ratios between three and 16. Plagioclase fractionation in a few lithologies is suggested by Eu anomalies reaching Eu/Eu* = 0·71. Heavy rare earth element (HREE) concentrations are between eight and 20 times chondrite. In total alkali–silica (TAS) and AFM diagrams (Fig. 2) the rocks follow the typical differentiation trend of medium-K calc-alkaline magmas. Nyström et al. (2003) concluded that magmas of the apparently older Abanico Fm. were formed at shallower mantle depths than those of the Farellones Fm., with the Abanico parental magmas having segregated in the spinel peridotite stability field, whereas the younger Farellones mafic magmas show evidence of residual garnet in their source. Based on trace element ratios (e.g. Th/Yb vs Ta/Yb, Ta/Yb vs Ba/Th, Ba/La vs La/Yb) a minor contribution of minor crustal material was suggested, possibly including sediments (Nyström et al., 2003; Deckart & Godoy, 2006).
|
Geochronology
K–Ar whole-rock dating of the Abanico and Farellones volcanic rocks (Fig. 1) has yielded ages between 20 and 4·1 Ma (Drake et al., 1976
; Vergara & Drake, 1979; Munizaga & Vicente, 1982; Beccar et al., 1986; Vergara et al., 1988). Roughly 100 km to the south, whole-rock K–Ar ages lie between 27·7 and 6·0 Ma (Kay et al., 2005). Granitoid intrusions into these volcanic rocks yield K–Ar whole-rock ages ranging from 19·5 to 18·5 Ma (Munizaga & Vicente, 1982). Moreover, amphibole and biotite 40Ar/39Ar dates from the volcanic rocks give ages ranging from 28·8 to 18·6 Ma (Aguirre et al., 2000; Fuentes et al., 2002) and 50 km to the south, plagioclases from porphyritic basalts were dated at 34–18 Ma (Muñoz et al., 2006). Amphibole and biotite 40Ar/39Ar ages of a granitoid intrusion a further 100 km to the south are 21·6 and 5·5 Ma (Kurtz et al., 1997). Close to our study area, U–Pb dating of zircon has yielded ages of 17·2 and 16·8 Ma for two andesitic flows, 12·0 and 8·4 Ma for two intruding granodiorites, and 6·3–5·2 Ma for four porphyric dacitic rocks (Deckart et al., 2005). Further U–Pb ages on zircons from two granodiorites about 50 km to the south yield ages of 11·3 and 10·3 Ma (Deckart & Godoy, 2006).
Sr–Nd–Pb isotope geochemistry
The petrogenesis of the Oligo-Miocene volcanic–plutonic series was previously addressed through major and trace element and Pb–Sr–Nd isotope measurements on whole-rock samples (Vergara et al., 1999; Nyström et al., 2003; Kay et al., 2005; Muñoz et al., 2006). The resulting models for the magmatism implicate varying mantle source regions and depths of magma generation (Nyström et al., 2003; Kay et al., 2005). Moreover, some basalts and andesites with ages of 20–18 Ma were interpreted to be generated by the interaction of mantle-derived magmas with the lower continental crust (e.g. Kay et al., 2005).
Data fields for these earlier data are shown for reference in the Figs 7–11 (summarized in Table 7). Initial 87Sr/86Sr ratios for both extrusive and intrusive lithologies lie in a very narrow range between 0·7033 and 0·7044 (Vergara et al., 1999; Nyström et al., 2003; Kay et al., 2005; Deckart & Godoy, 2006; Muñoz et al., 2006). For the same series of whole-rock samples from central Chile (33–36°S) initial epsilon Nd values (Ndi) define a relatively narrow range between +3·0 and +6·2. Similarly homogeneous isotope characteristics are also observed for initial Pb ratios, yielding 18·453–18·588 for 206Pb/204Pb, 15·548–15·610 for 207Pb/204Pb, and 38·210–38·478 for 208Pb/204Pb (Vergara et al., 1999; Nyström et al., 2003; Kay et al., 2005). For zircon of two granodioritic intrusions initial epsilon Hf values (Hfi) lie at +4 and +8 (Deckart & Godoy, 2006).
Based on Ndi vs initial 87Sr/86Sr isotopic ratios (Sri) Nyström et al. (2003) suggested a mantle source enriched in large ion lithophile elements (LILE) and high field strength elements (HFSE) compared with asthenospheric mantle. Moreover, Deckart & Godoy (2006) used Ndi vs Hf to propose mixing between such enriched mantle and pelagic sediments; however, Kay et al. (2005) showed that Ndi is independent of SiO2, ruling out significant crustal contamination. Apart from petrological arguments suggesting increasingly deeper mantle melting with time (Nyström et al., 2003) a correlated change from more to less LILE/HFSE depleted mantle sources was suggested (Kay et al., 2005).
|
ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Major and trace element analyses were performed by inductively coupled plasma atomicc emission spectrometry (ICP-AES) using a Perkin Elmer P400 instrument at the Geology Department of the University of Chile. Mineral compositions were determined on carbon-coated polished thin-sections using the wavelength-dispersive spectrometry (WDS) system of a CAMECA SX100 electron microprobe at the University of Montpellier II, France, calibrated with natural and synthetic standards. Results are considered to be accurate to within 1–3% for major elements. Mineral separation was carried out using a Frantz isodynamic magnetic separator, heavy liquids (CHBr3 and CH2I2), and hand-picking under a binocular microscope. All U–Pb, Pb and Sr isotope analyses were performed at the University of Nice–Sophia Antipolis.
Initial Pb isotopic compositions were measured for primary magmatic feldspar. These data were also used to correct for common Pb in zircon as well as in the determination of 238U/204Pb–206Pb/204Pb isochron ages. Overall analytical uncertainties (2
-STERR) for the U–Pb dates are 2–4% for 206Pb/238U, and 3–10% for 207Pb/235U, including in-run precision and correction for blanks, mass-fractionation and common Pb. Lead blanks are 13–17 pg per analysis, with 5–7 pg coming from the PTFE® capsules and 8–10 pg from the chemical reagents. Typical ratios of sample to blank Pb lie around 10. Error ellipses in Figs 3–5 correspond to the above uncertainties with correlation coefficients between about 0·3 and 0·5. Analytical uncertainties are much smaller for the isochron data (no common Pb correction) at 0·10–2·0% for 206Pb/204Pb and 0·30–3·0% for 238U/204Pb (Table 3).
|
|
|
Prior to dissolution of zircon in >50% HF at 215°C for 3 days in PTFE® Teflon bombs, the solutions were spiked with a mixed 205Pb/235U/233U solution, followed by the separation of U and Pb from Zr and Hf (Krogh, 1973
). To eliminate 176Yb interferences on 176Hf, a separation step on AGW-50X cation resin® was used (Patchett & Tatsumoto, 1980). Hafnium isotope measurements were performed on a Nu plasma 1700 instrument at the Ecole Normale Supérieure at Lyon using the JMC-475 Hf standard for calibration, run every fourth sample (Blichert-Toft et al., 1997). Hafnium isotope measurements were normalized to 179Hf/177Hf = 0·7325 (Patchett & Tatsumoto, 1980). The chemical procedures for the combined U–Pb–Hf analyses of zircon have been described elsewhere (Bodet & Schärer, 2000).
Hand-picked feldspars and clinopyroxenes were washed in 6N HCl, ground in an agate mortar and leached with 1% HF/HBr 1N for a few minutes in an ultrasonic bath (see Schärer, 1991). Dissolution was performed in >50% HF overnight at 120°C in 2 ml Savilex beakers®. To confirm the low U abundance in these minerals, some analyses were spiked with the same 233U–235U–205Pb solution as the zircons. For feldspar and clinopyroxene, a modified HBr procedure (Manhès et al., 1978) was used to separate and purify Pb and U from major elements and Sr, followed by Sr separation from major elements and Rb using Eichrom Sr-Spec. resin®.
All U–Pb and Pb isotope analyses were carried out on single Re filaments (H3PO4/Si-gel load) using a single secondary electron multiplier on a Thomson 206 mass spectrometer. Mass-fractionation of 0·10 ± 0·05%/a.m.u. was controlled by repeated analyses of the NBS-981 standard, which yielded average ratios of 16·941 ± 0·004 (2-STERR) for 206Pb/204Pb, 15·501 ± 0·004 for 207Pb/204Pb, and 36·728 ± 0·009 for 208Pb/204Pb. For concordia and isochron plots, and linear regression calculations we used the program ISOPLOT-3 (Ludwig, 2003).
Strontium isotopic compositions were measured on a VG-Sector instrument using single Re filaments with a H3PO4/TaF5 load. The NBS-987 standard was regularly run to control the accuracy of Sr measurement yielding an average (87Sr/86Sr)norm. of 0·702912 ± 0·000015 (2-STERR). All Sr isotope measurements were normalized to 86Sr/88Sr = 0·1194.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
The samples
A series of thin-sections from 162 rocks were studied to select samples with minimum metamorphic overprint, with well-preserved primary magmatic mineral assemblages. Sixteen samples of fresh plagioclase and clinopyroxene were chosen for Pb–Sr analysis. Eight rocks were chosen for U–Pb zircon dating, from which 13 zircon populations were selected for Hf isotope measurements. Eight whole-rocks were characterized for major and trace element and mineral compositions.
The samples analyzed are as follows.
- An olivine-phyric basalt (Ab-99) with a glomero-porphyritic texture containing between 5 and 20% euhedral to subhedral 1–2 mm olivine crystals. Augite crystals are 1–3 mm in size (2–5 vol.%,) and plagioclase (An88–92) 1–5 mm (10–40 vol.%). The groundmass is composed of very small grains (< 0·2 mm) of plagioclase, clinopyroxene and minor olivine. A chlorite–epidote–calcite assemblage represents altered augite.
- A gabbro sill (Ab-142) with a few per cent of 1–2 mm fresh augite and rare pigeonite, less than 2% olivine, and 30–50% fresh 1–4 mm plagioclase (An72–86). The groundmass is composed of altered glass with clinopyroxene and plagioclase crystals 0·1 mm in size. Apatite is an accessory mineral. Secondary minerals are calcite and chlorite in some augites, and phengite–sericite in the most altered plagioclase crystals.
- Three samples (Ab-154, -156, -159) were collected from different clinopyroxene-carrying andesitic flows, also characterized by glomero-porphyric textures. They are composed of fresh to intensely altered andesine (An45–49) and glomerocrysts of augite (Wo39En39Fs18 to Wo40En42Fs21) in a groundmass of altered glass, locally showing pilotaxitic plagioclase. Micro-phenocrysts (<0·2 mm) of clinopyroxene and plagioclase also occur in the groundmass. Calcite, smectite and albite are secondary phases occurring in the most altered domains of the andesine phenocrysts. Epidote, calcite, and chlorite are secondary minerals formed after clinopyroxene.
- Two dacitic sills (Ab-133, -143) and an andesitic sill (Ab-152) were sampled from within the volcanic sequence. They exhibit porphyritic textures, composed of plagioclase phenocrysts (1·5–3 mm) in a groundmass of fine-grained feldspar and minor quartz. The dacitic sills contain primary phenocrystic muscovite (0·5–2 mm) and altered amphibole (1–2 mm). Secondary calcite and sericite are observed in strongly altered plagioclase. Accessory minerals are apatite, zircon and titanite.
- Two samples, a micro-gabbro (Ab-153) and a gabbro (Ab-157), were collected from two fine- to medium-grained intrusions, respectively. Plagioclase is fresh in both rocks and shows a poikilitic texture with interstitial augite, whereas clinopyroxene has some minor secondary epidote.
- A tremolite-andesite hornfels sample (Ab-139) was collected from the metamorphic contact aureole around a granodioritic body intruding intermediate volcanic rocks; this has a fine-grained granoblastic texture. The most common contact metamorphic minerals are tremolite and biotite occurring together with primary andesine. The contact metamorphic rocks are fresh, showing some minor epidote, sericite, chlorite and pumpellyite in rare altered domains.
- Three samples (Ab-138, -132, -135) were taken from different homogeneous medium-grained granodioritic stocks that are much less altered than the volcanic members. Their constituent biotite and amphibole is replaced to varying degrees by chlorite, epidote and calcite. Plagioclase shows minor secondary smectite and calcite. Accessory phases are apatite, titanite, and zircon.
- Two fine- to medium-grained diorites (Ab-134, -136) were collected from the uppermost part of the series; these contain fresh plagioclase, clinopyroxene (augite) and orthopyroxene with epidote–chlorite alteration. Accessory phases are apatite and zircon.
Major and trace elements
Table 1 lists the major and trace element compositions of eight rocks taken from different levels of the volcanic sequence; Fig. 2 shows these data in a TAS diagram (a), an AFM diagram (b), and chondrite-normalized REE patterns (c). Basaltic to trachyandesitic rocks contain between 46 and 62 wt% SiO2, and plot within the field defined by all previously analyzed rocks from the Abanico and Farellones Formations at 33–36°S. Their REE patterns are enriched in light REE (LREE) relative to HREE, and are characterized by (La/Yb)N ratios ranging between 3·6 and 7·0 (Fig. 2c). LREE enrichment abundance reaches about 90 times chondrite; the HREE in the basaltic rocks lie at about six times chondrite abundance, whereas the more differentiated lithologies, such as trachyandesites, reach an enrichment factor of 24 for HREE (Fig. 2c) with a somewhat flatter pattern than the less differentiated rocks. A notable observation is the total absence of significant Eu anomalies in all samples.
|
U–Pb dating
Table 2 summarizes the zircon U–Pb analytical results used for the concordia plots, and Table 3 shows those samples for which 206Pb/204Pb (
)–238U/204Pb (µ) isochrons were calculated, including initial Pb ratios measured in cogenetic feldspars (Table 4). These results were obtained on 50 size fractions of zircon. Figures 3–5
display the concordia and isochron diagrams. Size-fractions of the 0·05 and 0·20 mm long zircon crystals vary in weight between 0·042 and 0·949 mg, representing between 20 and 100 grains. Cracks were absent and all grains analyzed were euhedral prisms ranging in width to length ratio between 1:1 and 4:1. Most zircons were pinkish translucent grains, and only the dacitic sill (Ab-143) had transparent colorless grains. Opaque and translucent inclusions, excluded from analysis as far as possible, were present in all populations. Independent of grain size and lithology, both types of inclusion reach about 5 vol. %, with most opaque inclusions being of clinopyroxene, and translucent grains of apatite.
|
|
|
Nine zircon fractions from a dacitic sill (Ab-143) lying conformably within olivine-basalt lavas yield both concordant and differently discordant data. Scatter in 206Pb/238U prevents the use of the
–µ isochron plot. On the other hand, seven of the fractions define a regression line that intercepts the concordia curve at 28·1 ± 1·5 (2) Ma (MSWD = 2·6). Uranium concentrations lie between 320 and 704 ppm. Total common Pb is 28–370 pg and measured 206Pb/204Pb ratios range from 48·9 to 457.
A gabbroic intrusive body (Ab-157) occurring within basaltic flows of the lowermost Abanico Formation yields three identically concordant zircon fractions, and a slightly discordant age. All 206Pb/238U values are identical, whereas 207Pb/235U shows up to 5% scatter. The mean value of the five zircon 206Pb/238U ratios defines an age of 22·20 ± 0·15 Ma, and the corresponding –µ isochron age, including feldspar, is 22·25 ± 0·1 (2) Ma (MSWD = 1·0), identical to the concordia age. Uranium concentration is between 307 and 683 ppm, total common Pb is 159–495 pg, and measured 206Pb/204Pb lies between 54·9 and 307.
Five zircon fractions from a micro-gabbro (Ab-153) emplaced into andesitic lava flows of the uppermost Abanico section produced two concordant and one slightly discordant dates, with the two remaining fractions showing scatter in 207Pb/235U. One of these fractions also has a slightly younger 206Pb/238U age. In using the four zircons with identical 206Pb/238U, an average age of 22·13 ± 0·23 Ma is obtained and an age of 21·97 ± 0·50 Ma (MSWD = 2·0) is given by the –µ isochron plot. Uranium concentrations are between 335 and 361 ppm, total common Pb ranges from 166 to 251 pg, and measured 206Pb/204Pb lies between 42 and 92.
A granodiorite (Ab-138) cutting the uppermost andesitic lavas of the Abanico Formation yields two identically concordant zircon analyses, a slightly discordant date, and two analyses that show significant scatter in 207Pb/235U. The five identical 206Pb/238U ratios yield an average age of 17·72 ± 0·47 Ma, and together with its cogenetic plagioclase they give an –µ isochron age of 17·2 ± 1·0 Ma (MSWD = 2·2). Uranium ranges between 213–521 ppm, total common Pb 188–724 pg, and measured 206Pb/204Pb from 33·4 to 80·3.
From a diorite (Ab-136), emplaced into the uppermost layers of the Abanico Formation, four of seven zircon fractions plot concordantly, whereas the remaining three fractions exhibit up to 12% scatter in 207Pb/235U. The average 206Pb/238U age of the five most concordant fractions is 12·82 ± 0·40 Ma, and together with plagioclase they give a –µ isochron age of 13·45 ± 0·42 Ma (MSWD = 10·2). Uranium concentrations are 233–401 ppm, total common Pb is 66·0–404 pg, and measured 206Pb/204Pb lies between 31·7 and 171.
Five zircon fractions of a granodiorite (Ab-135) cutting andesitic flows in the upper Abanico Fm. yield a tight cluster of 206Pb/238U ratios but 207Pb/235U again shows scatter, reaching 7%. The average 206Pb/238U age of the five zircon fractions is 13·02 ± 0·25 Ma, identical to their zircon–plagioclase –µ isochron age of 13·41 ± 0·34 Ma (MSWD = 2·4). Uranium concentrations are homogeneous at 223–282 ppm, total common Pb is 75·4–292 pg and measured 206Pb/204Pb lies between 40·2 and 106.
Sample Ab-134 is a diorite emplaced into basaltic flows of the lower Abanico series. All 206Pb/238U ratios are identical but 207Pb/235U shows large scatter of up to 20%. An age of 12·28 ± 0·15 Ma is defined by the mean value of the six 206Pb/238U ratios, and the –µ isochron gives an age of 12·13 ± 0·05 (MSWD = 1·3). Uranium concentrations range from 259–318 ppm, total common Pb is 94·1–458 pg and measured 206Pb/204Pb ranges between 36·6 and 97·1.
Sample Ab-132 is a granodiorite intruding an olivine-basalt lava flow in the lower part of the Abanico Fm. Eight zircon fractions yield identical 206Pb/238U ages, but again 207Pb/235U shows scatter. The mean value of samples with identical 206Pb/238U ratios define an age of 11·53 ± 0·19 Ma, and the zircon–plagioclase –µ isochron age is 11·40 ± 0·17 Ma (MSWD = 2·4). Uranium concentrations are 178–278 ppm, total common Pb is 25·2–302 pg, and measured 206Pb/204Pb lies between 37·0 and 233.
Pb, Sr, and Hf isotope data
Tables 4–6 list Pb, Sr, and Hf isotope data for 26 plagioclase separates and two clinopyroxene fractions for Pb, 18 plagioclase and two clinopyroxene fractions for Sr, and 20 zircon fractions for Hf. Duplicate analyses were performed for all isotope systems, using different size fractions of the same mineral population, to trace any potential heterogeneity in the initial isotope signatures on the rock scale; that is, minerals extracted from a few kilograms of rock. Figure 6 shows the Hfi values plotted relative to the evolution of a chondritic uniform reservoir (CHUR), and the model for a continuously LILE-depleted mid-ocean ridge basalt (MORB) source mantle (DePaolo & Wasserburg, 1976; Patchett et al., 1981). All Hfi values are positive, ranging between +6·9 and +9·6. Figures 7–11 display correlation diagrams for Pb, Sr, and Hf isotope data, compared with previous data from the Oligo-Miocene magmatic belt (O-M belt), Pacific MORB source mantle, ocean island basalt (OIB) mantle sources, Pacific sediments, and reference values for subcontinental lithospheric mantle. This last field includes the initial isotope signatures of group I kimberlitic magmas (Smith, 1983; Weis & Demaiffe, 1985; Davies et al., 2001; Schmidberger et al., 2001), megacrystic clinopyroxene extracted from kimberlites (Kramers et al., 1983; Davies et al., 2001), magmatic perovskite (Heaman, 1989) and clinopyroxene from peridotite xenoliths in the volcanic rocks of northern Chile (Lucassen et al., 2005).
|
|
|
|
|
|
For samples lacking Hf isotope analyses, Nd isotope signatures were translated to
Hfi using the relationship Hfi 2Ndi for MORB sources, and Hfi = 1·33Ndi + 3·19 for OIB mantle sources (Patchett, 1983; Vervoort et al., 1996, 1999). Differences in Hfi between the two conversion models are relatively small (typically of the order of 1 Hfi unit) but they exceed our analytical errors, which are of the order of 0·1–0·3 units (Table 6). Figure 7b also shows the initial isotopic compositions of lower crustal mafic granulite xenoliths brought to the surface by Miocene volcanic rocks of the pre-Cordillera in Argentina (32°S; Kay et al., 1996). The field of initial isotopic compositions of MORB reflects present-day ratios of their mantle sources. Shifts in isotope signatures, extrapolated back to Oligocene times, are negligible, given the low U/Pb, Rb/Sr, and Lu/Hf of MORB source mantle. The same would apply if melting of early Cenozoic oceanic lithosphere was considered, for which residence times are too short to produce significantly different isotope signatures compared with the MORB sources from which they were extracted. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
U–Pb dating
The following issues have to be considered concerning the U–Pb dating of the Oligo-Miocene zircons: (1) they contain very low amounts of radiogenic Pb produced since crystallization of the mineral; (2) relative to such low abundances of radiogenic Pb, initial common Pb is high; (3) ion beams must be sufficiently strong to be measured with sufficient accuracy, in particular for 204Pb; (4) long dissolution times at high temperatures induce high Pb blanks. As a consequence of the low abundance of radiogenic Pb, the minimum weights of zircon fractions used were between 0·04 and 0·95 mg, requiring the use of relatively large amounts of ion-exchange resin adding a second Pb blank component. Although the Pb isotopic compositions of cogenetic feldspars were measured to allow common Pb corrections (Table 4), additional uncertainties are twofold: the Pb isotopic compositions of feldspar fractions of the same population can be different (Table 4), and inherited old Pb may be associated with different common initial Pb. Uncertainty in the common Pb correction is therefore the dominant analytical uncertainty for U–Pb dating using the concordia diagram. On the other hand, the determination of alternative isochron ages does not require this correction.
As revealed by the concordia plots, the scatter of 207Pb/235U in zircons from a given sample exceeds the expected uncertainties, such as those integrated into the error propagation (size of ellipses). This means that the 207Pb/204Pb ratios of initial common Pb cannot be determined with sufficient precision. From a purely analytical point of view, scatter of 207Pb/235Pb could also be due to interferences on the 207Pb peak or bad peak shapes; however, both these parameters have been systematically controlled, as well as the quality of the mass base line. Because the 238U–206Pb chronometer is much less sensitive to common Pb corrections, no corresponding scatter is observed, and as a consequence all U–Pb ages are derived from the 206Pb/238U ratios. Alternative use of different common Pb isotopic compositions measured for feldspar fractions of the same rock does not change the concordia or isochron ages beyond the given analytical precisions. The new U–Pb ages range from 28·1 to 11·5 Ma, covering an Oligo-Miocene period of 16·6 Myr.
The 20 Hf isotope analyses (Hfi) performed on the eight dated zircon populations, as well as on a further five zircons, yield Hfi ranging between +6·9 and +9·6 (Fig. 6). A slight difference in Hfi is observed even among zircon fractions separated from the same population (Ab-138; Ab-153, Table 6). Two of the previously published Hfi values for 11–10 Ma granodiorite zircons from the same area are identical to those of this study, whereas one analysis has a less radiogenic value at +4·0 (Fig. 6; Deckart & Godoy, 2006).
Initial Pb isotope ratios (Pbi) of plagioclase and clinopyroxene show a large range with all values being significantly more radiogenic than Pacific MORB. They are also distinct from Pacific OIB sources (Fig. 7). In the 206Pb/204Pb () vs 207Pb/204Pb (β) diagram (Fig. 7a), most Pbi values lie in the range of model values for average and upper continental crust, with some Pbi plotting below average crust. In the vs 208Pb/204Pb (
) diagram (Fig. 7b) they lie at the upper end of the average crust model but they are distinct from the initial Pb isotopic compositions of Andean lower crustal xenoliths (Kay et al., 1996). Previous Pb isotope data obtained for whole-rocks of the 28·8–5·2 Ma Oligo-Miocene magmatic sequence are identical to the least radiogenic values measured here, and again are distinct from both Pacific MORB and OIB sources. One exception is sample Ab-154, which has high 207Pb (Fig. 7a). In the Pbi vs Sri correlation diagram of Fig. 8 all initial isotope signatures are again distinct from any Pacific MORB source, and is also different from Pacific OIB sources, whereas β lies very close to OIB. Large differences also exist relative to Pacific sediments. On the other hand, as already noted for Fig. 7, the data plot in the field of reference values for subcontinental lithospheric mantle.
Figure 9 displays Pbi vs Hfi data substantiating the same isotope differences already observed for , β and as well as Pbi–Sri correlations (Figs 7 and 8). For subsequent discussion it is important to note that for the initial Hf isotope ratios of subcontinental mantle only measured Hf isotope data were used. Lu/Sm fractionation in the subcontinental mantle is not sufficiently constrained to translate Nd isotope ratios into Hfi (Nowell et al., 2004). A particular observation is that in the –Hfi diagram, our Oligo-Miocene samples lie in, or very close to, the field of initial isotope signatures of zircon and baddeleyite megacrysts from the Central African Mbuji Mayi kimberlite (Weis & Demaiffe, 1985; Schärer et al., 1997). In the β–Hfi diagram the O-M belt zircons overlap with data from SE Pacific OIB sources, whereas they are distinct from any Pacific MORB source. In the –Hfi and Sri–Hf correlation diagrams significant differences exist relative to both MORB or OIB mantle sources (Figs 10 and 11).
U–Pb ages
As already emphasized, the new U–Pb zircon dating was undertaken to determine unambiguous emplacement ages for the Oligo-Miocene magmatic rocks. Thus far, only U–Pb zircon ages could be considered to date the age of primary crystallization (Deckart et al., 2005; Deckart & Godoy, 2006); whereas the interpretation of most previous K–Ar and 40Ar/39Ar whole-rock and mineral dates from central Chile remains uncertain because of the omnipresence of later low-grade metamorphism and hydrothermal activity (Drake et al., 1976; Vergara & Drake, 1979; Munizaga & Vicente, 1982; Beccar et al., 1986; Vergara et al., 1988; Kurtz et al., 1997; Fuentes et al., 2002; Deckart et al., 2005; Kay et al., 2005; Muñoz et al., 2006). For some older ages around 34 Ma, argon degassing patterns reveal the presence of significant excess 40Ar, which rules out the determination of any geologically meaningful age (‘maximum ages’; Gana & Wall, 1997; Muñoz et al., 2006). In some cases, no age interpretation could be proposed; for example, for 40Ar/39Ar dating results of secondary minerals (Deckart et al., 2005).
The new and all previous radiometric ages of the Oligo-Miocene magmatic rocks are summarized in Table 7 together with initial Sr, Pb, Nd, and Hf isotopic ratios, and model Th/U of the magma sources. Table 7 also distinguishes plutonic and volcanic lithologies. Previous zircon U–Pb ages (17–5 Ma; Deckart et al., 2005) were also exclusively based on the 238U–206Pb chronometer, with the difference that no feldspars could be measured for Pbi because of the strong metamorphic overprint of the rocks. The full set of radiometric ages substantiates basaltic–gabbroic to rhyolitic–granitic magmatic activity between 28·8 and 5·2 Ma, covering a period of 24 Myr during Oligo-Miocene times.
Abanico vs Farellones Formation
Based on field geology, radiometric ages, and isotope ratios it appears that both the volcanic and plutonic lithologies represent a continuous period of calc-alkaline magmatism, emplaced during Nazca–Farallon plate subduction beneath the South American continental margin. The distinction between the Abanico and Farellones Formations was originally based on differences in lithologies, observed between the lower and upper part of the volcanic–plutonic series (Hoffstetter et al., 1957). The Farellones Formation was redefined by Rivano et al. (1990) with a distinction into a lower rhyolitic–dacitic to ignimbritic part, and an upper section composed of basaltic andesites, intruded by rhyodacitic domes. It was concluded that Abanico magmatism lasted from Oligocene times until 20 Ma, followed by emplacement of the Farellones Formation (Munizaga & Vicente, 1982; Vergara et al., 1988; Fuentes et al., 2002; Deckart et al., 2005; Kay et al., 2005). On the other hand, an overlap in emplacement ages was proposed for the same series about 100 km to the south of our study area, indicating that the lower Farellones and upper Abanico Formations have a similar age (Charrier et al., 2002). A local angular unconformity at the base of the Farellones Formation was suggested (e.g. Aguirre, 1960), but other workers have questioned the existence of such an unconformity and proposed either a thrust at the base of the Farellones Formation or a stratigraphic contact (e.g. Godoy et al., 1999). Our own detailed observations along the transition from the Abanico to the Farellones Formation do not reveal any unconformity or tectonic contact, and the zone seems to be a simple transition from older to younger magmatic series showing the same lithological spectrum. Given the uncertainties concerning both the distinction and age of the Abanico vs Farellones Formations, we consider that the dated rocks represent a continuum of magmatic activity.
Magma sources
The question arises whether lithological differences or isotope signatures are correlated in space and time. From our age data we cannot identify any clear compositional progression with time. The oldest (28·1 Ma) and youngest rocks (11·5 Ma) are both differentiated (i.e. a dacite and a granodiorite, respectively) and gabbros and diorites cover almost the full period of magmatism (22·2–12·3 Ma). An absence of correlation is also observed for isotope signatures and ages; however, previous geochemical data from the same magmatic belt suggested a continuous evolution from primitive rocks with (La/Yb)N = 3, (87Sr/86Sr)i = 0·7033, and Ndi = +6·2 to slightly more evolved members having (La/Yb)N = 16, (87Sr/86Sr)i = 0·7044, and Ndi = +3·0 (Nyström et al., 2003; Kay et al., 2005). Limited differences are also suggested by the Pb isotopes signatures, with the older members being more homogeneous than the younger volcanic and plutonic lithologies (Nyström et al., 2003). This is consistent with the differences in lithology between the lower and upper parts of the succession (i.e. 80% basalt and 20% andesites in the lower part, and 20% basalt, 75% andesite, and 5% dacite in the upper part). All the isotopic data lie within the reference field for subcontinental lithospheric mantle.
Heterogeneities on the data from this study are also observed, with two dacites having more radiogenic Sri than the main cluster of data, and a pyroxene-andesite being more radiogenic in 207Pb (Figs 7–11). For Sr in the dacites this deviation could be due either to a lithospheric mantle source that has evolved to a higher Rb/Sr but low U/Pb, or to the incorporation of a highly radiogenic crustal component. This latter hypothesis is in contradiction with the low radiogenic Pbi measured in the same feldspar. Concerning the high 207Pb, we necessarily need an old Precambrian source component, either in the metasomatized subcontinental mantle or as strongly Rb-depleted crustal material (lower crust?), to explain the absence of more radiogenic Sri.
Previous and new (this study) U–Pb studies of zircon from the magmatic belt reveal the presence of a small amount of older radiogenic Pb, essentially seen through the 235U–207Pb chronometer (Figs 3–5). Such inherited components reflect relict zircon extracted by the magmas from the melted source lithologies, present either in the subcontinental lithospheric mantle or in the overlying continental crust. In addition to zircon in kimberlites, the presence of old zircon in the lithospheric mantle has been demonstrated for plagiogranite dikes intruding the lithospheric peridotites of the European plate margin (Borsi et al., 1996). In either case, the ultimate source of inheritance is continental crust material, either integrated into the subcontinental mantle during ancient subduction or directly extracted from the overlying crust by the ascending mantle-derived magmas. Such inheritance is consistent with internal differences in Hfi, such as observed in two zircon populations extracted from a granodiorite and a micro-gabbro (Ab-138, -153, Fig. 6).
Further corroboration for the incorporation of different components is revealed by differences in Pbi and Sri, in both plagioclase and clinopyroxene fractions from the same rock (samples Ab-133; -134; -135; -153; -157; Tables 4 and 5). This could reflect mixing of crystals from partly crystallized magmas or extraction of xenocrystic grains from the wall-rocks. Whatever explanation is preferred for these intra-rock isotope differences, it should be recognized that they are very small and therefore these contributions have not significantly changed the isotopic composition of the Oligo-Miocene magmas. This is well illustrated by the preservation of primary basaltic–gabbroic compositions even for those that show such isotopic heterogeneities.
Concerning potential crustal contamination of the parental mantle-derived magmas, none of our Pbi values exceeds the reference field for subcontinental lithospheric mantle (Figs 7, 8, and 10a). The most sensitive tracer for continental material is Pb, for which 1–2% of crustal material would significantly shift the Pbi signatures of the mantle-derived magmas (e.g. Schärer, 1991). Other isotope tracers such as Sr, Nd and Hf are more robust to crustal contamination. To further quantify potential crustal contamination, we have used Sr, which is the most sensitive crustal isotopic tracer after Pb. Crustal end-members can be estimated from the isotopic composition of Paleozoic granites and basement rocks of the region (33°S, Fig. 1; Parada et al., 1999), yielding a 87Sr/86Sr of 0·7090 and a Sr concentration of 219 ppm. To estimate the mantle component, we used data for spinel-peridotite xenoliths found in Miocene volcanic rocks in Argentina that gave 87Sr/86Sr of 0·7038 and a Sr concentration of 22·3 ppm (Conceição et al., 2005). In this case, 2% of a crustal melt would shift the initial 87Sr/86Sr of the mantle-derived magmas to 0·704, whereas 5% contamination would produce an initial Sr isotope ratio of 0·7066, reflecting the most radiogenic composition measured in our Oligo-Miocene samples. In most of our samples initial 87Sr/86Sr lies around 0·704–0·705 (Table 5), significantly less radiogenic than the value for 5% crustal components. If we used a more radiogenic crustal component in the model (e.g. 87Sr/86Sr = 0·714), the potential crustal melt contributions would be much lower.
|
|
|
A further argument for a very small crustal component, or even its absence, is given by comparison of our Pbi values with those of lower crustal rocks such as pyroxene–garnet granulites underlying the area (Kay et al., 1996
). In the –β diagram, these granulite compositions plot outside the frame of Fig. 7a (206Pb/204Pb: 17·06–17·80), whereas they are included in the – plot of Fig. 7b. In both cases these lower crust reference values are distinct from the initial isotopic signatures of the O-M belt, and such contamination cannot be seen. As mentioned above, the absence of any correlation between initial Nd isotope ratios and SiO2 in these rocks also constrains crustal contamination to be insignificant (Kay et al., 2005). Initial hafnium isotope ratios indicate magma sources that are about 50% less depleted in LILE/HFSE than MORB sources (Fig. 6) and initial 87Sr/86Sr values are significantly more radiogenic (Table 5). A major difference is also observed for Th/U model ratios of the mantle source, which range from 3·9 to 4·1 (Table 4); these are higher than any value for Pacific MORB source mantle (Th/U = 3·7–3·8).
From a tectonic point of view it has been proposed that melts from asthenospheric mantle sources (i.e. melts from MORB-source mantle) provide the major component of the Oligo-Miocene magmas; however, neither our new or previous isotope data (Figs 7–11) are consistent with this hypothesis. As a consequence, the model of Kay et al. (2005) for the petrogenesis of the Oligo-Miocene magmas by decompression melting of the asthenoshere underneath extending continental lithosphere has to be questioned (e.g. Kay et al., 2005). All the Pbi, Sri, Hfi, and Th/U signatures of the Oligo-Miocene magmatic rocks in central Chile are consistent with their derivation from subcontinental lithospheric mantle. It is therefore plausible that the parental magmas to rock types ranging from basalt–gabbro to rhyolite–granite are derived from slightly LILE/HFSE-depleted lithospheric mantle reservoirs with high U/Pb.
Using an average value of +8·0 for Hfi (full range = +6·9 to +9·6; Table 6), a model age of 2·0 Ga is obtained for formation of the subcontinental mantle and corresponding continental crust. At 2 Ga, we assume that the difference between the depleted and primitive mantle was still small. Initial 87Sr/86Sr isotope ratios of the primitive mantle would have been 0·70123 (Rb/Sr = 0·027), and initial 206Pb/204Pb 12·685 (for µ = 7·13) using the Rb/Sr and U/Pb ratios of the primitive mantle from Hofmann (1988) and Stacey & Kramers (1975). Because crust extraction around 2·0 Ga would necessarily cause Rb and U depletion in the residual mantle, the Rb/Sr at that time would have been around 0·017, and µ at about 5·3, yielding a present-day 87Sr/86Sr of 0·702 and a 206Pb/204Pb of 14·613. Both these model ratios are significantly lower than the initial isotope ratios determined for the Oligo-Miocene mantle sources. In consequence, the 2·0 Ga depleted mantle underneath the South American continent must have been re-enriched in Rb and U, most probably through metasomatism occurring in relation to ancient subduction events, previously suggested to explain the inherited zircon components. This metasomatized mantle would have evolved with an average model Rb/Sr of about 0·039, and a µ of about 16·3. It should be noted that both these ratios are similar to those measured in some garnet-peridotite xenoliths from South Africa (Kimberley) and the Canadian Arctic (Hawkesworth et al., 1990; Schmidberger et al., 2001). The 2·0 Ga Hf model age is consistent with a 2·1–1·8 Ga upper zircon intercept and Nd model ages obtained for Paleozoic and Cenozoic rocks in the central Andes (Franz et al., 2006).
To induce Oligo-Miocene melting within the Andean metasomatized lithospheric mantle, the most plausible model is fluid release from the subducting Nazca–Farallon plate. Petrological data show that such melting occurred successively in the garnet and spinel stability fields (Nyström et al., 2003), reflecting an increasing depth of melting from Oligocene to Miocene times. A crucial question is how the large variation of lithologies was produced. A first-order observation is that basalts (gabbros) and andesites (diorites) make up more than 90% of the magmatic rocks. The evolution from olivine-basalts to andesites may be explained by a combination of (1) original chemical differences in the metasomatized subcontinental mantle source, (2) differences in the degree of mantle melting, and (3) crystal fractionation in the garnet and spinel stability fields. To explain further chemical differentiation towards rhyolitic–granitic compositions, further fractional crystallization must have played a role, occurring during magma ascent beneath and within the plagioclase stability field. For the latter, this is indicated by the presence of Eu anomalies previously reported for some samples (Nyström et al., 2003; Kay et al., 2005).
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
- The new U–Pb zircon ages confirm a period of about 16 Myr of calc-alkaline magmatism in central Chile, from 28·1 to 11·5 Ma. Together with previous dating results from the same region, the full period of continuous magmatism seems to have occurred between 28·8 and 5·2 Ma.
- Fluid-induced magma generation from old metasomatized subcontinental lithospheric mantle satisfies all the isotope signatures of the magmatic products. These lithospheric mantle sources are characterized by weak to moderate LILE/HFSE depletion and high 238U/204Pb. The Hf model age for major lithospheric mantle formation at about 2 Ga is in agreement with earlier model ages from the region.
- Genesis of basalts–gabbros and andesites–diorites constituting 90% of the magmatic belt seems to reflect a combination of (1) different degrees of partial melting of the mantle source, (2) original chemical (metasomatic) differences within the subcontinental mantle, and (3) crystal fractionation in both the spinel- and garnet-peridotite stability fields. The petrogenesis of the more evolved magmas (rhyolites–granites) most probably involves differentiation through fractional crystallization at mantle depths and within the crust, possibly inducing some very minor crustal melting.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
For technical assistance we thank J.-P. Goudour and M. Manetti, and we appreciated helpful advice during field work by A. Demant, E. Godoy, J. J. Verdugo, S. Calderon, F. C. Fuentes, J. Vargas and F. Rodriguez. For critical reading and helpful suggestions on earlier versions of the manuscript we are indebted to K. Deckart. Detailed reviews by S. Noble, T. Waight, G. Wörner and J. Gamble have helped to considerably improve the manuscript. For funding the project we thank FONDECYT (numbers 1020809 and 1061266) and ECOS-CONICYT (C03U01). P. Montecinos thanks CONICYT for providing a 3 years grant for his Ph.D. thesis.
*Corresponding author. Telephone: +33-04-92-07-68-11. Fax: +33-04-92-07-68-16. E-mail: scharer{at}unice.fr
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL BACKGROUNDANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Aguirre L. Geología de los Andes de Chile central (provincia de Aconcagua). Instituto de Investigaciones Geológicas, Santiago, Boletín (1960) 9:70.
Aguirre L, Robinson D, Bevins RE, Morata D, Vergara M, Fonseca E, Carrasco J. A low-grade metamorphic model for the Miocene volcanic sequences in the Andes of central Chile. New Zealand Journal of Geology and Geophysics (2000) 43:83–93.
Bach W, Erzinger J, Dosso L, Bollinger C, Bougault H, Etoubleau J, Sauerwein J. Unusually large Nb–Ta depletions in North Chile ridge basalts at 36°50' to 38°56'S: major element, trace element, and isotopic data. Earth and Planetary Science Letters (1996) 142:223–240.[CrossRef][Web of Science]
Beccar I, Vergara M, Munizaga F. Edades K–Ar de la Formación Farellones, en el cordón del cerro La Parva, Cordillera de los Andes de Santiago, Chile. Revista Geológica de Chile (1986) 28–29:109–113.
Ben Othman D, White WM, Patchett J. The geochemistry of marine sediments, island arc magma genesis, and crust–mantle recycling. Earth and Planetary Science Letters (1989) 94:1–21.[CrossRef][Web of Science]
Blichert-Toft J, Albarède F. Lu–Hf isotope geochemistry of chrondrites and the evolution of mantle–crust system. Earth and Planetary Science Letters (1997) 148:243–258.[CrossRef][Web of Science]
Blichert-Toft J, Chauvel C, Albarède F. Separation of Hf and Lu for high-precision isotope analysis of rock samples by magnetic sector–multiple collector ICP-MS. Contributions to Mineralogy and Petrology (1997) 127:248–260.[CrossRef][Web of Science]
Bodet F, Schärer U. Evolution of the SE-Asian continent from U–Pb and Hf isotopes in single grains of zircon and baddeleyite from large rivers. Geochimica et Cosmochimica Acta (2000) 64:2067–2091.[CrossRef][Web of Science]
Borsi L, Schärer U, Gaggero L, Crispini L. Age, origin and geodynamic significance of plagiogranites in lherzolites and gabbros of the Piedmont–Ligurian basin. Earth and Planetary Science Letters (1996) 140:227–241.[CrossRef][Web of Science]
Charrier R, Baeza O, Elgueta S, Flynn JJ, Gans P, Kay SM, Muñoz N, Wyss AR, Zurita E. Evidence for Cenozoic extensional basin development and tectonic inversion south of the flat-slab segment, southern Central Andes, Chile (33°–36°S.L.). Journal of South American Earth Sciences (2002) 15:117–139.[CrossRef][Web of Science]
Chauvel C, Blichert-Toft J. A hafnium isotope and trace element perspective on melting of the depleted mantle. Earth and Planetary Science Letters (2001) 190:137–151.[CrossRef][Web of Science]
Conceição R, Mallmann G, Koester E, Schilling M, Bertotto G, Rodriguez-Vargas A. Andean subduction related mantle xenoliths: isotopic evidence of Sr–Nd decoupling during metasomatism. Lithos (2005) 82:273–287.[CrossRef][Web of Science]
Davies GR, Spriggs AJ, Nixon PH. A non-cognate origin for the Gibeon Kimberlite megacryst suite, Namibia: Implications for the origin of Namibian kimberlites. Journal of Petrology (2001) 42:159–172.
Deckart K, Godoy E. Barren igneous intrusives in the Central Andes: U/Pb geochronology and Nd–Hf isotope geochemistry. In:. In: Extended Abstracts Volume, V, South American Symposium on Isotope Geology (2006) Montevideo: SSAGI (CD).
Deckart K, Clark A, Aguilar C, Vargas R, Bertens A, Mortensen J, Fanning M. Magmatic and hydrothermal chronology of the giant Rio Blanco Porphyry Copper Deposit, central Chile: Implications of an integrated U–Pb and 40Ar/39Ar. Economic Geology (2005) 100:905–934.
DePaolo DJ, Wasserburg GJ. Nd isotopic variations and petrogenetic models. Geophysical Research Letters (1976) 3:249–252.[Web of Science]
Devey CW, Hémond C, Stoffers P. Metasomatic reactions between carbonated plume melts and mantle harzburgite: the evidence from Friday and Domingo seamounts (Juan Fernandez chain, SE Pacific). Contributions to Mineralogy and Petrology (2000) 139:68–84.[CrossRef][Web of Science]
Drake RE, Curtis G, Vergara M. Potassium–argon dating of igneous activity in the Central Chilean Andes—latitude 33°S. Journal of Volcanology and Geothermal Research (1976) 1:285–295.[CrossRef][Web of Science]
Franz G, Lucassen W, Trumbull R, Romer RL, Wilke H, Viramonte J, Becchio R, Siebel W. Crustal evolution at the Central Andean continental margin: a geochemical record of crustal growth, recycling and destruction. In: The Andes—Active Subduction Orogeny—Oncken O, Chong G, Franz G, Götze H, Ramos V, Strecker M, Wigger P, eds. (2006) Berlin: Springer. 45–64.
Fuentes F, Vergara M, Aguirre L, Féraud G. Relaciones de contacto de unidades volcanicas terciarias de los Andes de Chile central (33°S): una reinterpretación sobre la base de dataciones 40Ar/39Ar. Revista Geológica de Chile (2002) 29(2):207–225.[Web of Science]
Gana P, Wall R. Evidencias geocronológicas 40Ar/39Ar y K–Ar de un hiatus Cretácico superior–Eoceno en Chile central (33–33°30'S). Revista Geológica de Chile (1997) 24:145–163.[Web of Science]
Gerlach DC, Hart SR, Morales VW, Palacios C. Mantle heterogeneity beneath the Nazca Plate: San Felix and Juan Fernandez islands. Nature (1986) 322:165–169.[CrossRef]
Godoy E, Yáñez G, Vera E. Inversion of an Oligocene volcano-tectonic basin and uplifting of its superimposed Miocene magmatic arc in the Central Chilean Andes: First seismic and gravity evidence. Tectonophysics (1999) 306:217–236.[CrossRef][Web of Science]
Hawkesworth CJ, Erlank AJ, Kempton PD, Waters FG. Mantle metasomatism: isotope and trace element trends in xenoliths from Kimberley, South Africa. Chemical Geology (1990) 85:19–34.[CrossRef][Web of Science]
Heaman LH. The nature of the subcontinental mantle from Sr–Nd–Pb isotopic studies on kimberlite perovskite. Earth and Planetary Science Letters (1989) 92:323–334.[CrossRef][Web of Science]
Hofmann AW. Chemical differentiation of the Earth: the relationship between mantle, continental crust and oceanic crust. Earth and Planetary Science Letters (1988) 90:297–314.[CrossRef][Web of Science]
Hoffstetter R, Fuenzalida H, Cecioni G. Léxique stratigraphique international, 7, Amérique Latine, Chili–Chile (1957) Paris: Centre National de la Recherche Scientifique. 444.
Holmes A. An estimate of the age of the Earth. Nature (1946) 157:680–684.[CrossRef]
Jaffey H, Flynn KF, Glendenin LE, Bentley WC, Essling AM. Precision measurements of half-lives and specific activities of 235U and 238U. Physical Reviews (1971) C4:1889–1906.[CrossRef]
Karsten JL, Klein EM, Sherman SB. Subduction zone geochemical characteristics in ocean ridge basalts from the southern Chile Ridge: Implications of modern ridge subduction systems for the Archean. Lithos (1996) 37:143–161.[CrossRef][Web of Science]
Kay SM, Orrell S, Abbruzzi JM. Zircon and whole-rock Nd–Pb isotopic evidence for a Grenville age and a Laurentian origin for the basement of the Precordillera in Argentina. Journal of Geology (1996) 104:637–648.[Web of Science]
Kay SM, Godoy E, Kurtz A. Episodic arc migration, crustal thickening, subduction erosion, and magmatism in the south–central Andes. Geological Society of America Bulletin (2005) 117:67–88.
Kramers JD, Roddick JCM, and Dawson JB. Trace element and isotope studies on veined, metasomatic and ‘MARID’ xenoliths from Bultfontein, South Africa. Earth and Planetary Science Letters (1983) 65:90–106.[CrossRef][Web of Science]
Krogh TE. A low contamination method for hydrothermal decomposition of zircon and extraction of U and Pb for isotopic ages determination. Geochimica et Cosmochimica Acta (1973) 37:485–494.[CrossRef][Web of Science]
Kurtz A, Kay SM, Charrier R, Farrar E. Geochronology of Miocene plutons and exhumation history of the El Teniente region, Central Chile (34°–35°S). Revista Geológica de Chile (1997) 24:75–90.[Web of Science]
Le Bas MJ, Le Maitre RW, Streckeisen A, Zanettin B. A chemical classification of volcanic rocks based on the total alkali–silica diagram. Journal of Petrology (1986) 27:745–750.
Levi B, Aguirre L, Nyström JO, Padilla H, Vergara M. Low-grade regional metamorphism in the Mesozoic–Cenozoic volcanic sequences of the Central Andes. Journal of Metamorphic Geology (1989) 7:487–495.[Web of Science]
Lucassen F, Franz TG, Viramonte J, Romer RL, Dulski P, Lang A. The late Cretaceous lithospheric mantle beneath the Central Andes: Evidence from phase equilibria and composition of mantle xenoliths. Lithos (2005) 82:379–406.[CrossRef][Web of Science]
Ludwig KR. Isoplot 3.0, a Geochronological Toolkit for Microsoft Excel. Berkeley Geochronology Center Special Publications (2003) 4:74.
Manhès G, Minster JF, Allègre CJ. Comparative U–Th–Pb and Rb–Sr study of the Saint Séverin amphoterite: consequence for early Solar System chronology. Earth and Planetary Science Letters (1978) 39:14–24.[CrossRef][Web of Science]
Munizaga F, Vicente JC. Acerca de la zonación plutónica y del volcanismo miocénico en los Andes de Aconcagua (Lat. 32–33°S): datos radiométricos K–Ar. Revista Geológica de Chile (1982) 16:3–21.
Muñoz M, Fuentes F, Vergara M, Aguirre L, Nyström J, Féraud G, Demant A. Abanico East Formation: petrology and geochemistry of volcanic rocks behind the Cenozoic arc front in the Andean Cordillera, central Chile (33°50'S). Revista Geológica de Chile (2006) 33(1):109–140.[Web of Science]
Nowell GM, Pearson DG, Bell DR, Carlson RW, Smith CB, Kempton PD, Noble SR. Hf isotope systematics of kimberlites and their megacrysts: new constraints on their source regions. Journal of Petrology (2004) 45:1583–1612.
Nyström JO, Vergara M, Morata D, Levi B. Tertiary volcanism during extension in the Andean foothills of central Chile (33°15'–33°45'S). Geological Society of America Bulletin (2003) 115:1523–1537.
Parada M, Nyström JO, Levi B. Multiple sources for the Coastal Batholith of central Chile (31–34°S): geochemical and Sr–Nd isotopic evidence and tectonic implications. Lithos (1999) 46:505–521.[CrossRef][Web of Science]
Patchett PJ. Importance of the Lu–Hf isotopic studies of planetary chronology and chemical evolution. Geochimica et Cosmochimica Acta (1983) 47:81–91.[CrossRef][Web of Science]
Patchett PJ, Tatsumoto M. A routine high-precision method for Lu–Hf isotope geochemistry and chronology. Contributions to Mineralogy and Petrology (1980) 75:263–267.[CrossRef][Web of Science]
Patchett PJ, Kouovo O, Edge CE, Tatsumoto M. Evolution of continental crust and mantle heterogeneity: Evidence from Hf isotopes. Contributions to Mineralogy and Petrology (1981) 78:279–297.[Web of Science]
Cordani UG, Milani EJ, Thomaz Filho A, Campos DA. The Southern Central Andes. In: Tectonic Evolution of South America. Abstracts, 31st International Geologic Congress (2000) Rio de Janeiro: International Union of Geological Sciences (IUGS). 561–604.
Rivano S, Godoy E, Vergara M, Villaroel R. Redefinición de la Formación Farellones en la Cordillera de los Andes de Chile Central (32°–34°S). Revista Geológica de Chile (1990) 17:205–214.
Schärer U. Rapid continental crust formation at 1·7 Ga from a reservoir with chondritic isotope signatures, eastern Labrador. Earth and Planetary Science Letters (1991) 102:110–133.[CrossRef][Web of Science]
Schärer U, Corfu F, Demaiffe D. U–Pb and Lu–Hf isotopes in baddeleyite and zircon megacrysts from the Mbuji-Mayi kimberlite: Constraints on the subcontinental mantle. Chemical Geology (1997) 143:1–16.[CrossRef][Web of Science]
Schmidberger SS, Simonetti A, Francis D. Sr–Nd–Pb isotope systematics of mantle xenoliths from Somerset Island kimberlites: Evidence for lithosphere stratification beneath Arctic Canada. Geochimica et Cosmochimica Acta (2001) 65:4243–4255.[CrossRef][Web of Science]
Sguigna AP, Larabee AJ, Waddington JC. The half-life of 176Hf by – coincidence measurements. Canadian Journal of Physics (1982) 60:361–364.
Smith CB. Pb, Sr and Nd isotopic evidence for sources of southern African Cretaceous kimberlites. Nature (1983) 304:51–54.[CrossRef]
Stacey JS, Kramers JD. Approximation of terrestrial lead isotope evolution by a two-stage model. Earth and Planetary Science Letters (1975) 26:207–221.[CrossRef][Web of Science]
Steiger RH, Jäger E. Subcommission on geochronology: convention on the use of decay constants in geo- and cosmochronology. Earth and Planetary Science Letters (1977) 36:359–362.[CrossRef][Web of Science]
Tatsumoto M, Knight RJ, Allègre CJ. Time differences in the formation of meteorites as determined from the ratio of 207-lead to 206-lead. Science (1973) 180:1279–1283.
Vergara M, Drake R. Edades K/Ar en secuencias volcánicas continentales postneocomianas de Chile Central; su depositación en cuencas intermontanas restringidas. Revista de la Asociación Geológica Argentina (1979) 34:42–52.
Vergara M, Charrier R, Munizaga F, Rivano S, Sepúlveda P, Thiele R, Drake R. Miocene volcanism in the central Chilean Andes. Journal of South American Earth Sciences (1988) 1:199–209.[CrossRef]
Vergara M, Morata D, Hickey-Vargas R, López-Escobar L, Beccar I. Cenozoic tholeiitic volcanism in the Colbun area, Linares Precordillera, Central Chile. Revista Geológica de Chile (1999) 26:23–41.[Web of Science]
Vervoort JD, Patchett PJ, Gehrels GE, Nutman AP. Constraints on early Earth differentiation from hafnium and neodymium isotopes. Nature (1996) 379:624–627.[CrossRef]
Vervoort JD, Patchett PJ, Blichet-Toft J, Albarède F. Relationships between Lu–Hf and Sm–Nd isotopic systems in the global sedimentary system. Earth and Planetary Science Letters (1999) 168:79–99.[CrossRef][Web of Science]
Weis D, Demaiffe D. A depleted mantle source for kimberlites from Zaire: Nd, Sr and Pb isotopic evidence. Earth and Planetary Science Letters (1985) 73:269–277.[CrossRef][Web of Science]
White WM, Patchett J, Ben Othman D. Hf isotope ratios of marine sediments and Mn nodules: evidence for a mantle source of Hf in seawater. Earth and Planetary Science Letters (1986) 79:46–54.[CrossRef][Web of Science]
Zartman RE, Doe BR. Plumbotectonics—the model. Tectonophysics (1981) 75:135–162.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||