Journal of Petrology | Volume 40 | Number 10 | Pages 1527-1551 | 1999
© Oxford University Press 1999
Crustal Recycling of Metamorphic Basement: Late Palaeozoic Granitoids of Northern Chile (
22°S). Implications for the Composition of the Andean Crust
1 Fachgebiet Petrologie, Technische Universität Berlin Ernst Reuter Platz 1, 10623 Berlin, Germany
2 Department of Geology, Royal Holloway University of London Egham TW20 0EX, UK
3 Mineralogisches Institut, Universität Münster Corrensstrasse 24, 48149 Münster, Germany
Received August 20, 1998; Revised typescript accepted April 30, 1999
 |
ABSTRACT |
|---|
 TOP ABSTRACT IntroductionGeological SettingResultsDiscussionAppendixReferences |
|---|
Upper Palaeozoic silicic magmatism is widespread in the Central Andes, but its origin is poorly constrained. We investigated whole-rock chemical and isotopic composition of Upper Palaeozoic granitoids and their Early Palaeozoic high-grade country rocks in the Chilean Coastal Cordillera and Precordillera at
22°S, in comparison with an Upper Cretaceous granitoid. The age of the Late Palaeozoic granitoids from a Rb–Sr isochron of 300 Ma is consistent with K–Ar cooling ages of hornblende and biotite. Similar major and trace element patterns as well as Nd and Pb isotopic composition of Upper Palaeozoic granitoids and gneisses point to a source of the granitoids that is similar to the gneisses at outcrop. Sr isotope ratios of the Upper Palaeozoic granitoids are less radiogenic than those in many of the gneisses. We propose a stratification of the Early Palaeozoic crust with a Rb-deficient granulitic mid–lower crust, resulting in less radiogenic Sr compared with the upper crust, based on the interpretation of the P–T–t history and isotopic composition of the Lower Palaeozoic metamorphic basement and of the isotopic composition of the Late Palaeozoic granitoids and younger magmatic rocks. Nd isotopic composition is identical in lower and upper crust and in crustal melts from the Late Palaeozoic to Recent. The Cretaceous granitoid evolved from partial melts of a mantle-derived source with considerable contamination by the old crustal component. The crust that formed in the Early Palaeozoic is the major source of material for the Cenozoic tectonic thickening of the Andean crust. KEY WORDS: Central Andes; crustal composition; crustal recycling; granitoid magmatism; isotopic composition
|
Introduction |
|---|
|
TOPABSTRACTIntroductionGeological SettingResultsDiscussionAppendixReferences |
|---|
The magmatic arc in the Puna–Altiplano plateau of the Central Andes rests on a continental crust of a maximum thickness of
70 km. Crustal thickening is mainly attributed to shortening of the crust and not to juvenile magmatic additions from the subduction zone [see review by Allmendinger et al., (1997)
]. This implies a major contribution of pre-Mesozoic material to the present thick crust, because Mesozoic juvenile additions to the crust are focused in the Mesozoic magmatic arcs mainly in the Chilean Coastal Cordillera (Fig. 1, e.g. Scheuber et al., 1994; Lucassen et al., 1996a), which is not involved in the crustal thickening. The last major reorganization of the crust before the Cenozoic was in the Early Palaeozoic orogeny between 500 Ma, the age of peak metamorphism, and 400 Ma, the age of the final exhumation (Damm et al., 1990, 1994; Lucassen et al., 1996b, unpublished data, 1998; Becchio et al., 1997). The widespread Upper Palaeozoic granitoid magmatism starting at 300 Ma was assigned to arc magmatism (e.g. Brown, 1991; Bahlburg & Hervé, 1997) or to melting of the crust at a passive margin [Breitkreuz & Zeil, (1984) and references therein]. The metamorphic history of the pre-Mesozoic crust was studied in detail before (Lucassen et al., 1999a, unpublished data, 1998; summary of previous work given by Damm et al., 1990, 1994; Miller et al., 1994). This study focuses on the composition of both the high-grade metamorphic basement and granitoid intrusions along a traverse from the Chilean Coastal Cordillera to the Precordillera at 22°S. Granitoids are used widely to monitor crustal composition and geodynamic settings. Surprisingly, the granitoids previously assigned to an arc setting are crustal melts from a source very similar to the high-grade basement without traceable influence of mantle-derived material. A comparison of the isotopic data from northern Chile with those of the Lower Palaeozoic high-grade basement of NW Argentina (R. Becchio, unpublished data, 1998) shows the regional relevance of our findings. We propose a stratification of the Early Palaeozoic crust with a Rb-deficient granulitic mid–lower crust resulting in less radiogenic Sr isotope ratios, but with identical Sm–Nd ratios and Nd isotope ratios based on the P–T–t history and the isotopic composition of the basement, the isotopic composition of the granitoids and the geological history of the area. This hypothesis is supported by the isotopic composition of various crustal or contaminated melts from Mesozoic to Recent.
|
|
Geological Setting |
|---|
|
TOPABSTRACTIntroductionGeological SettingResultsDiscussionAppendixReferences |
|---|
Extended periods of magmatism in the Central Andes occurred from Late Precambrian to Early Palaeozoic (
570–400 Ma; e.g. Damm et al., 1990, 1994; Rapela et al., 1990; Coira et al., 1999) and from Late Palaeozoic (300 Ma) to Recent (Rogers & Hawkesworth, 1989; Brown, 1991; Scheuber et al., 1994). The early period was contemporaneous with widespread low-pressure (4–7 kbar) and high-temperature (600–750°C) metamorphism in northern Chile and NW Argentina (Lucassen et al., 1996b, unpublished data, 1998; Becchio et al., 1997). These magmatic–metamorphic events formed the basement for the next period of magmatism. There is no record of magmatic activity in northern Chile from Silurian time (Damm et al., 1990, 1994) until the renewed onset of magmatism in Late Carboniferous time. The post-Silurian history starts with the deposition of clastic Devonian–Lower Carboniferous sediments in a passive margin setting, and a hypothetical landmass in the west has been proposed to explain depositional features (Bahlburg & Breitkreuz, 1991; Bahlburg, 1993; Bahlburg & Hervé, 1997). Late Palaeozoic–Triassic magmatic rocks are mainly felsic and form a broad belt from the Coastal Cordillera to the Altiplano (Fig. 1; Berg et al., 1983; Brown, 1991; Breitkreuz & Zeil, 1994). Their chemistry has been variously interpreted as a result of melting of the crust at a passive margin [Breitkreuz & Zeil, (1984) and references therein] or as a result of interaction of mantle-derived melts from a subduction zone with the crust (Brown, 1991). The latter interpretation implies the transition from a passive to an active margin in Late Carboniferous time (Bahlburg, 1993; Bahlburg & Hervé, 1997).
During the Mesozoic a dramatic change in magmatism occurred, from mainly felsic crustally derived to gabbroic–dioritic compositions with mantle signatures and varying amounts of crustal contamination (Rogers & Hawkesworth, 1989; Pichowiak, 1994; Lucassen & Franz, 1994, and references therein). The centre of Jurassic–Cretaceous magmatism, the Coastal Cordillera, comprises 70% intrusive or volcanic rocks at the surface (Scheuber et al., 1994), and gravity and seismic velocity data indicate the continuation of the prevailing mafic magmatic crust to depths of at least 20 km (Götze et al., 1994; Wigger et al., 1994). All magmatic activity has been attributed to an active continental margin setting with a tectonic regime of prevailing normal extension to transpression during the Mesozoic (e.g. Scheuber et al., 1994; Dallmeyer et al., 1996). Mantle-derived magmatism with considerable contributions from crustal melts and generation of crustal melts (ignimbrites) continued throughout the Cenozoic (e.g. Francis et al., 1989; Ort et al., 1996; Wittenbrink, 1997). The Cenozoic Andean subduction zone is tectonically dominated by crustal shortening and thickening, and has been the subject of numerous studies.
Little is known about the Late Palaeozoic magmatism (21–26°S), and previous studies did not consider the possible source compositions or the effects of contamination by the pre-Silurian basement. The study presented here focuses on the intrusions in the southern Sierra de Moreno (Fig. 1). All but one of the intrusions are of Late Carboniferous to Early Permian age. Included in this study is an Upper Cretaceous pluton in the southernmost Sierra de Moreno, to compare differences in the possible state of the crust and the sources of the different magmatic pulses. Additional Palaeozoic granitoids in the Caleta Loa area of the Coastal Cordillera and at the western slope of Sierra de Limón Verde were sampled to compare the intrusions on a regional scale (Fig. 1). Previous investigations of these rocks are limited to isotopic dating (Damm et al., 1990; Maksaev, 1990). We also investigated isotope chemistry of representative samples from the high-grade basement (Fig. 1). The area of investigation is covered by the geological maps No. 3 Tocopilla (preliminary), No. 58 Calama, No. 51 Quillagua, No. 40 Ollagüe (all published by Servicio Nacional de Geología y Mineria, Santiago, Chile). Granitoid intrusions of similar age occur between Taltal and Chañaral (Fig. 1) in the Coastal Cordillera (Berg et al., 1983; Brown, 1991), the Precordillera (Cordillera Domeyko; Smoje & Marinovic, 1994) and in the Altiplano (Brown, 1991).
|
Results |
|---|
|
TOPABSTRACTIntroductionGeological SettingResultsDiscussionAppendixReferences |
|---|
Petrography and field relations
In Sierra de Moreno we distinguish a northern Palaeozoic intrusive complex at Cerro Negro separated by metamorphic rocks from a southern complex referred to here as SM (north) and SM (south) and a Cretaceous intrusion in the southernmost part at Qda. de Barreras (Fig. 1). When the Upper Palaeozoic intrusions are referred to as a group, the term ‘Palaeozoic granitoids’ is used. Representative analyses are given in Table 1. The Palaeozoic granitoids intruded the high-grade LowerPalaeozoic metamorphic basement (Lucassen et al., 1996b
, unpublished data, 1998) and in the north also intruded presumed Devonian–Early Carboniferous sediments (Fig. 1; Bahlburg & Breitkreuz, 1991).
|
Granitic and granodioritic–tonalitic compositions can be easily distinguished in the field by their colour index. The northern plutons comprise granite–granodiorite,whereas the rocks from the southern block are granite–tonalite. Granodiorites–tonalites are less voluminous than the more evolved granite at SM (north) and are subordinate at SM (south). At SM (north) the granite intruded into the granodiorite, but at SM (south) the different rock types may have been coeval, and pods and dykes ofgranodiorite–tonalite interacted with the unsolidified granite. Also, a few hornblende-rich cumulates were found.
Mineral parageneses and fabrics of the granitoids are very similar at SM (north) and SM (south). Major phases are plagioclase, hornblende, biotite, quartz and potassic feldspar, with generally increasing proportions of the last three minerals with increasing SiO2 content. In the most evolved compositions (SiO2 > 70 wt %) hornblende disappears and biotite is rare. Common minor phases are opaque minerals, titanite, apatite, zircon and monazite, and in the more evolved granitoids magmatic Ce-epidote. Garnet is present as a minor phase in the evolved compositions at SM (north) and at Caleta Loa (west). Chlorite formed in many samples after biotite and/or hornblende. Sericitization and/or saussuritization reach variable extent in the different samples and are commonly restricted to the core region or parts of the zoned plagioclase. Grain sizes generally increase from the mafic to the evolved compositions. The minerals are never strongly deformed and all rocks have a magmatic fabric. In some coarse-grained samples small rims of recrystallized or primary smaller grains occur around quartz and feldspar. Mineral alignment is not found at the microscopic scale. All macroscopic alignment, for example, of feldspar laths, is in magmatic flow textures.
The Upper Cretaceous granitoid in the southermost Sierra de Moreno (Fig. 1) is intruded into Cretaceous and Jurassic sediments in the west (Bogdanic & Espinoza, 1994), Palaeozoic granitoids in the east and the Lower Palaeozoic metamorphic rocks in the southeast. The pluton is mainly granodiorite to monzonite with few fairly mafic compositions. The latter have mainly plagioclase, biotite and clinopyroxene that can be replaced by hornblende. Primary hornblende is found in rocks with >59 wt % SiO2. In the most evolved rocks biotite is the only Fe–Mg mineral and the proportions of potassic feldspar and quartz increase. Ductile deformation is not found in the rocks.
At Sierra de Limón Verde (Fig. 1) a granodiorite pluton intrudes migmatites for which no isotopic dating is available, but which are very similar to those from Sierra de Moreno [Baeza, (1984), including a geological map). At Caleta Loa (Fig. 1), the western pluton intrudes into migmatites of Lower Palaeozoic metamorphic rocks. Apart from local shear zones of unknown age, the relatively homogeneous pluton is undeformed. The eastern pluton intruded into Devonian–Lower Carboniferous sediments (Bahlburg & Breitkreuz, 1991). All but one sample plot into the granite field. The appearance of the rocks is different from the western pluton, with the development of large (centimetre size) K-feldspar megacrysts. Their size and frequency varies, but they are present in most samples.
In general, the intrusions are undeformed and no foliation planes or preferred orientations developed. Contacts between the intrusions and the country rocks are primary or overprinted by post-Jurassic–Recent brittle faults. The contact relations of the Upper Palaeozoic intrusions with Palaeozoic sediments (maximum thickness 3000 m, Bahlburg & Breitkreuz, 1991) and of the Cretaceous intrusion with Jurassic–Cretaceous sediments point to high levels of intrusion (<5 km) at all locations. This is consistent with the absence of primary muscovite, because muscovite is not stable at pressures below 2 kbar. Contact metamorphism with the formation of new minerals was observed in the sediments with contact aureoles of some hundred metres to 1 km.
The high-grade basement comprises abundant gneisses and migmatites (quartz–plagioclase–biotite ± garnet ± K-feldspar). Common minor phases are zircon, magnetite, ilmenite and apatite. Both rock types have the same compositional range and are referred to in the text as gneisses. Rare minor intercalations (decimetre to metre size) of quartzite and calcsilicate point to a sedimentary protolith, but granitoid orthogneisses were also found. Aluminous compositions with quartz–cordierite–plagioclase–micas–garnet–aluminosilicate are rare in the basement of northern Chile. Amphibolites are rare, forming <5% of the metamorphic rocks, and are former dykes or minor volcanic intercalations. Peak metamorphic conditions are of high-T and low-P type at 600–750°C and 4–7 kbar, and migmatization is widespread in these rocks. Metamorphism occurred at 500 Ma, and subsequent uplift and erosion was finished at 400 Ma with the rocks close to the present erosion surface. More details of the geology and metamorphic history of the high-grade basement in northern Chile and NW Argentina have been given by Damm et al., (1990, 1994), Lucassen et al., (1994, 1996b, 1999a, unpublished data, 1998), Miller et al., (1994) and Becchio et al., (1997).
Geochronology
Seven samples of the Palaeozoic granite from SM (north and south; sample 4/5 excluded) define an isochron of 300.4 ± 2.5 Ma, mean squared weighted deviation (MSWD) = 4.3 (Fig. 2, Table 2) with Sri = 0.708589 ± 53, calculated according to the method of York, (1969) using a 0.5% error (1
) on the 87Rb/86Sr ratios and the internal precision on the 87Sr/86Sr ratios in the calculation. The MSWD could be further improved to 2.9, if the external reproduceability of 0.003% from the Sr standard is used as error on the 87Sr/86Sr ratios in the calculation. The 300 Ma age is interpreted as age of intrusion and coincides with a concordant U–Pb zircon age of 298 ± 1.5 Ma from a granodioritic sample from Sierra de Limón Verde (Fig. 1; Damm et al., 1990).
|
for data see
|
The K–Ar ages (Table 3) are considered as cooling ages. Ages of 283 ± 6 Ma for biotite (3/299) from a granite sample and 281 ± 9 Ma for hornblende (3/300) from a diorite sample of SM (south) coincide. The K–Ar ages from SM (north) are 301 ± 8 Ma for biotite (4/417) from a granite and 332 ± 7 Ma for biotite (4/36) from a granodiorite. The difference in age between granite and granodiorite might be significant, though sample 4/36 plots on the Rb–Sr isochron (Fig. 2), because the intrusions are in contact and the granite contains dioritic xenoliths. In the Rb–Sr isochron it is not possible to distinguish between slightly older and younger samples if the Rb/Sr ratio is low.
|
At Sierra de Limón Verde K–Ar ages (5/19) are 270 ± 15 Ma for hornblende and 273 ± 8 Ma for biotite from granite. The K–Ar age at Caleta Loa (east, 5/49) on biotite is 310 ± 7 Ma. A similar age of 318 ± 6 Ma on biotite of the same intrusion further to the south was reported by Maksaev & Marinovic, (1980)
. A granite sample from Caleta Loa (west, 5/40) has a K–Ar age on biotite of 236 ± 5 Ma. Skarmeta & Marinovic, (1981) reported K–Ar ages on biotite from the same intrusion of 322 ± 5 and 320 ± 5 Ma, however. We speculate that our sample was partially reset by the widespread Jurassic–Cretaceous igneous activity in the relevant magmatic arcs. All samples used for K–Ar dating plot on or close to the Rb–Sr isochron (Fig. 2). We conclude—in combination with the contact relations of the granitoids with deformed Devonian–Upper Carboniferous sediments—that all samples belong to the same period of Late Palaeozoic igneous activity between 330 and 270 Ma also known from other occurrences of granitoids in the Precordillera (Brown, 1991; Smoje & Marinovic, 1994) and in the Coastal Cordillera (Berg et al., 1983; Berg & Baumann, 1985).
Hornblende from the Cretaceous pluton in SM (south) yielded an Upper Cretaceous K–Ar age of 82 ± 6 Ma (4/10; Table 3). This cooling age, together with the observation of intrusive contacts into the Cretaceous (Bogdanic & Espinoza, 1994) country rocks, is clear evidence for a Mesozoic intrusion age.
Whole-rock chemical composition
Major elements
The granitoids and the gneisses from the basement have a wide range of chemical composition (Figs 3 and 4 Table 1). The Palaeozoic granitoids of Sierra de Moreno and Sierra de Limón Verde range from 50 to 78 wt % SiO2, whereas at Caleta Loa all but one sample have SiO2 > 70 wt %. The Cretaceous granitoid has a restricted range of 55–65 wt % SiO2. All Palaeozoic granitoid samples are subalkaline, whereas those from the Cretaceous pluton are alkaline (Fig. 3b, d, g). For these rocks K2O is slightly higher than Na2O in many samples and they can be classified as shoshonitic; alkalis and SiO2 show no systematic variation. Most of the Palaeozoic rocks are mildly peraluminous (Fig. 3h) with A/CNK = 1–1.1, the Caleta Loa plutons are more aluminous (A/CNK = 1.2), and many of the Cretaceous samples are metaluminous (A/CNK = 0.9–1). The differences in major element composition between the Palaeozoic granitoids from different localities are minor and roughly follow the common trends of magmatic differentiation within the scatter of the values that are typical for rocks with varying proportions of cumulates (Fig. 3). The basement gneisses are similar in composition and compositional trend to the Palaeozoic granitoids, and most samples plot in the range of the less evolved granitoids (Fig. 3). The average composition of 35 gneiss samples from northern Chile (Tables 1 and 4) matches the composition of upper-crustal rocks as greywackes (Taylor & McLennan, 1985).
|
|
|
Trace elements
The similarity between the basement gneisses and the Palaeozoic granitoids is also evident from the trace element contents (Table 1; selected element variation diagrams in Fig. 4), for example, Ba, V, Sr and Y. The Cretaceous granitoids are mostly similar to the Palaeozoic granitoids, but Sr is on average significantly higher. V and TiO2 are positively correlated (Fig. 4b), as are other transition metals. Ba behaves rather irregularly, though it is in many localities positively correlated with K2O. The samples from the Cretaceous pluton and those from Caleta Loa, however, show no correlation. Rb (not shown) is positively correlated with K2O in all rock types, but not with other incompatible elements such as P. Y is not correlated with SiO2 (Fig. 4d) in general and within the distinct groups. Low Y at Caleta Loa (west) could be related to the presence of garnet in the source seen in the high LaN/YbN (>40); however, other samples from other plutons low in Y have low LaN/YbN ratios (see below).
Rare earth elements
The rare earth elements (REE) of the granitoids and the basement gneisses are generally similar in their range of element contents and their patterns apart from effects of magmatic differentiation in the granitoids (Fig. 5, Table 1). Within the Palaeozoic granitoids, there is a systematic variation of contents and patterns with SiO2 content at different locations. At SM (north) the less evolved rocks with SiO2 < 70 wt % have a slightly negative Eu anomaly, whereas the evolved rocks > 70 wt % SiO2 have a pronounced negative Eu anomaly, lower L(light)REE contents and rather variable H(heavy)REE contents (Fig. 5a). The latter rocks show a Nd–Sm ‘plateau’ as a result of a high Sm/Nd ratio compared with the other samples (sample 4/417 contains garnet). In contrast, at SM (south) (Fig. 5b) the evolved rocks show higher REE contents, and the less evolved rocks lower REE contents with a positive Eu anomaly. These observations, in line with the CaO and Sr variation of these rocks (Figs 3c and 4c) point to fractionation and accumulation, respectively, of plagioclase in the respective magmas. The granitoids from Sierra de Limón Verde and Caleta Loa (east) show the same features as those from Sierra de Moreno; the samples from Caleta Loa (west) show a very steep pattern with LaN/YbN of 43–74. The LaN/YbN ratio in most of the other granitoids (12 samples) varies between two and nine, and three samples have a ratio of 11, 13 and 15. The high LaN/YbN ratio for Caleta Loa west points to garnet that is present in the samples or zircon involvement in the source or during fractionation not evident at the other locations. The basement gneisses have a LaN/YbN ratio between five and nine. REE patterns of the Cretaceous granitoids (Fig. 5d) show no systematic variation with SiO2 content. Their patterns are slightly steeper than those of the Palaeozoic granitoids, with LaN/YbN of 9–16.
|
The Argentine Lower Palaeozoic metamorphic basement is similar in composition, but comprises more evolved pelitic metasedimentary rocks, as seen by the comparison of the average composition of gneisses from northern Chile and NW Argentina (Table 1, Fig. 6). Lower Ca and Sr compared with the Chilean gneisses point to less plagioclase, and the higher K and Rb to more clay minerals and micas in the protoliths, in accordance with the interpretation of more abundant pelitic protoliths in NW Argentina. Apart from these differences the element patterns of Chilean and Argentine gneisses are similar.
|
Sr, Nd and Pb isotopic composition of granitoids and gneisses
We analysed 11 samples from the Upper Palaeozoic granitoids, four from the Cretaceous granitoid and 12 samples from the gneisses (Table 2). Additionally, we included samples from the basement in NW Argentina (our data; R. Becchio, unpublished data, 1998) for regional comparison. We also included Lower Palaeozoic mafic, mantle-derived rocks (Table 2 and Damm et al., 1990
), Mesozoic intrusive and volcanic rocks and lower-crustal xenoliths (Rogers & Hawkesworth, 1989; Lucassen & Franz, 1994; Lucassen & Thirlwall, 1998; Lucassen et al., 1999b; our unpublished data, 1998) and Cenozoic volcanic rocks (Francis et al., 1989; Ort et al., 1996; Wittenbrink, 1997) in our study to constrain the isotopic composition of mantle sources and the possible contamination of the Cenozoic volcanic rocks by the postulated crustal source. For Sr and Nd, whole-rock samples were selected (Table 2), whereas for Pb determination feldspar separates were prepared (Table 4). (For analytical technique, see the Appendix.)
The 87Sr/86Sr ratios and 
Nd of Palaeozoic granitoids and those basement gneisses with low Rb/Sr ratios cover a similar range at 300 Ma (Fig. 7a). The differences in the 143Nd/144Nd300Ma ratio are small (range 0.511733–0.512295 for the gneisses including those with high 87Sr/86Sr ratio; 0.511941–0.512166 for the granitoids; Fig. 7a). The basement in NW Argentina has a similar Nd isotopic composition at 300 Ma; however, the 87Sr/86Sr ratios of many samples are more radiogenic compared with those of granitoids and basement of northern Chile, as a result of the generally higher amounts of metapelitic material with higher Rb contents (Fig. 7a). Recalculated to 500 Ma, the mean age of high-grade metamorphism and isotope homogenization, both areas show the same range of isotopic composition (Fig. 7b). The Early Palaeozoic amphibolites cluster at Nd +5 (Fig. 7a), consistent with the interpretation that they are not metasediments, but metamorphosed igneous dykes or volcanic intercalations.
|
A group of high-SiO2 granites (samples 4/23, 4/25, 4/417) from SM (north) has unusually high 147Sm/144Nd ratios (
0.15, Table 2) compared with other granitoids and the gneisses (0.11–0.13), and higher Nd (Fig. 7a). However, their 143Nd/144Nd300Ma ratios are similar to those of the other samples (Fig. 7a). High Sm/Nd ratios from granites compared with their protoliths are known from other areas (e.g. Harris, 1996; Ayres & Harris, 1997) and ascribed to the preferred dissolution of high-REE phases, preferably monazite and apatite, in the anatectic melt. The 143Nd/144Nd might be also displaced compared with the bulk source if the isotopic equilibration of the source occurred much earlier than the melting (e.g. Ayres & Harris, 1997) and the radiogenic growth of 143Nd continued in the relevant mineral. The concept could be valuable for our samples; however, a quantitative evaluation is not possible without knowledge of the mineral compositions in the source. The three samples are highly fractionated, with relatively low REE abundances (Table 1), and crystallization of respective phases could also fractionate the REE (very minor garnet and monazite are present; see Nd–Sm ‘plateau’ in the REE pattern of the three samples in Fig. 5a). The 143Nd/144Nd300Ma ratio of 0.512003 for sample 5/40 indicates the same isotopic composition of the source for the granitoid at Caleta Loa (west) as for the other Palaeozoic granitoids. However, the 147Sm/144Nd ratio of 0.0898 reflects garnet in the source (Table 2).
The Cretaceous granitoid at 80 Ma has lower 87Sr/86Sr ratios (0.704550–0.704669) and higher 143Nd/144Nd ratios (0.512471–0.512505) than the Palaeozoic granitoids, as a result of the influence of a mantle or mantle-derived source (Fig. 7b; Table 2).
The Nd of the granitoids at time of intrusion is not correlated with SiO2 of the rocks (Fig. 8). This indicates that in both the Upper Palaeozoic and the Cretaceous granitoid fractionation of a homogeneous parental magma or different amounts of partial melting of the same source can be the reason for the variation of SiO2 content. Magma mixing of two sources, such as a mantle and a crustal component, is unlikely in the Palaeozoic granitoids. The magma of the Cretaceous intrusion possibly is a mixture of a depleted mantle source magma, as defined by the mafic rocks of the area (Fig. 7b), and a source similar to the Early–Late Palaeozoic crust, as presented by Lower Palaeozoic gneisses and Upper Palaeozoic granitoids (see discussion below). The very small spread in isotopic composition and the absence of correlation between Nd and SiO2 indicate a well-homogenized magma.
|
Pb isotopic ratios from feldspar separates of the gneisses and of the granitoids are also similar (Fig. 9a, b). 206Pb/204Pb, 207Pb/204Pb and 208Pb/204Pb ratios plot above or near the Pb ore development line (Stacey & Kramers, 1975
). This is in agreement with other Precambrian to Palaeozoic basement rocks from the Central Andes. Pb isotope ratios are not correlated with Nd or Sr isotope ratios (not shown). The similarity of the values for all rocks is consistent with our hypothesis of a homogeneous crustal source. This source is different from the Proterozoic Arequipa Massif and transitional between the fields for young volcanic rocks of the northern and southern Central Volcanic Zone (Wörner et al., 1994), in general agreement with the assumption of a more radiogenic Pb source of contamination for the Cenozoic volcanic rocks south of 21°S.
|
|
Discussion |
|---|
|
TOPABSTRACTIntroductionGeological SettingResultsDiscussionAppendixReferences |
|---|
Source of the Palaeozoic granitoids
Isotope composition
Ratios of radiogenic isotopes are widely used to hypothesize about possible source rocks of igneous rocks. Nd isotope ratios, recalculated to 300 Ma (Fig. 7a), of all Palaeozoic granitoids are similar and in the same range as those of the Lower Palaeozoic gneisses from northern Chile and NW Argentina. Sm–Nd-depleted mantle model ages of the gneisses (Table 2) have an average of
1.65 ± 0.2 Ga with a range of 1.36–1.92 Ga. The Palaeozoic granitoids’ average is 1.58 ± 0.2 Ga with a range of 1.28–1.87 Ga, exclusive of samples 4/23, 4/25, 4/417 and 5/40 (all Palaeozoic granitoids yield 1.60 ± 0.2 Ga).
The Sr isotope ratios (at 300 Ma, Fig. 7a) are low fora crustal source, though similar to the samples of the highgrade metamorphic basement with low Rb/Sr ratios. We propose as a hypothesis a source for the granitoids similar to the currently exposed country rocks at the intrusion level and refer to this as the ‘gneiss-type source’. The low Sr ratios of the granitoids are interpreted considering the metamorphic history of the basement. The P–T conditions of the Early Palaeozoic metamorphism with widespread migmatization at 500 Ma indicate a mid-crustal level of 15–20 km at temperatures of upper amphibolite–lower granulite facies. Therefore it can safely be assumed that the Early Palaeozoic lower crust (the possible source for the Upper Palaeozoic granitoids) had experienced a granulite facies metamorphism. Rb depletion in rocks of granitoid composition is common during granulite facies metamorphism (e.g. Rudnick & Presper, 1990) and this depletion explains the lack of radiogenic Sr. Direct evidence of possible Rb depletion and conservation of low Sr isotope ratios in the lower crust is given by lower-crustal felsic xenoliths from the Cretaceous Salta Rift of NW Argentina, which have the same Nd isotopic composition as the Early Palaeozoic metamorphic basement (Fig. 7b; Lucassen et al., 1999b). The more heterogeneous Sr300Ma isotopic composition of the basement gneisses, compared with the granitoids, is due to the fact that the Rb/Sr of the gneisses varies at a scale of 10–100 cm because of the layered structure of the rocks, with biotite-rich and biotite-poor layers. This could be already the case in the sedimentary protolith of the gneisses, with enrichment of clay minerals in layers.
A prominent contribution of a mantle source to the granites could not be excluded from the isotope composition, if an enriched mantle with Nd of -5 and a 87Sr/86Sr ratio of 0.708 is assumed, similar to, for example, the enriched mantle under South Africa [e.g. Nixon, (1987) and references therein]. This is unlikely for the Andes, however, as all known mafic rocks are strongly depleted (Fig. 7b). Furthermore, the granitic magmatism is voluminous (including the Permian SiO2-rich volcanic rocks; see Fig. 1) and fractionation of a major volume of granitic magma from a basaltic parental magma would leave a considerable gabbroic residue of the order of 6–10 times the granites’ volume (Cox, 1993). There is no indication for such rocks from the geophysical data on the crustal structure (Götze et al., 1994; Wigger et al., 1994) and such rocks are also absent in an exposed lower-crustal section of Permian–Triassic age at Sierra de Limón Verde (Lucassen et al., 1999a).
The Pb isotope ratios confirm the hypothesis of the gneiss-type source, similar to the Early Palaeozoic basement (Fig. 9). Compared with the data given by Tosdal, (1996) for a variety of basement rocks from the area between 17 and 21°S, they are rich in 206Pb and 207Pb, and intermediately radiogenic in 208Pb. This type of basement can be clearly distinguished from the less radiogenic 206Pb and 207Pb, and high radiogenic 208Pb in the Proterozoic Arequipa Massif, which extends from southern Peru to Bolivia. Such a Proterozoic type of basement is therefore unlikely to be the source of the granitoids. The basement rocks from the area between 21 and 26°S from our study have Pb isotope ratios that are intermediate between those from Cenozoic volcanic rocks north of 21°S and south of 21°S (Wörner et al., 1994).
Major and trace element composition
The hypothesis of a gneiss-type source for the Upper Palaeozoic granitoids can be tested by comparison of the chemical composition of the rocks. Major and trace element contents are similar and in the range of typical upper-crustal rocks, compared with, for example, the average upper crust (Fig. 6). The average of the granitoids is slightly higher than the average of the gneisses in SiO2, Na2O, K2O, Rb and Ba, lower in CaO, Fe2O3, MgO, TiO2, V, Cr and Ni, and has an mg-number of 33 compared with a value of 44 in the gneisses, i.e. the granitoids are slightly more felsic and silicic. Also, the average of the Argentine gneisses is similar, except that the lower Ca and Sr compared with the Chilean gneisses points to less plagioclase in the protoliths, in accordance with the interpretation of more abundant pelitic protoliths in NW Argentina. The mg-number of 42 is similar to that for the Chilean gneisses.
The REE patterns of all but one granitoid are similar to those of the gneisses and point to no or minor garnet in the source or as a fractionated mineral. The LaN/YbN is between 2 and 15. This is consistent with our observation that the basement is poor in garnet, and also consistent with the hypothesis that it possibly extends to lower-crustal depth. At an average Mg/(Mg + Fetotal) ratio of the biotite–plagioclase–quartz gneisses of 44, garnet will form in the stability field of plagioclase only in small amounts at T above the solidus and pressures between 5 and 10 kbar (mid–lower crust), as known from experimental work on greywacke composition (e.g. Vielzeuf & Montel, 1994; Patiño Douce & Beard, 1995, 1996; Montel & Vielzeuf, 1997; Stevens et al., 1997). Metabasites, which could have garnet as a major constituent at these pressures, are rare at the surface and in the geophysical images of the crust, and therefore are probably also absent in the source. Only at Caleta Loa (west) are the steep REE patterns and low Y contents typical for a garnet residue or fractionation of garnet. The increasing La/Yb and Nd/Sm with simultaneously decreasing HREE and Y contents (see Fig. 5c, samples 5/36, 5/39, 5/40) mirror the distribution coefficients of garnet for the REE.
Other granitoids from the Central Andes at 26°S in the Taltal–Chañaral area (Fig. 1) from the Permian–Triassic show major and trace element patterns, including REE and Sri isotope ratios between 0.7065 and 0.7115 [Brown, (1991) and references therein], that are very similar to our results. These granitoids could have formed under similar conditions. There are, however, other granitoid intrusions from the area (25°S, Smoje & Marinovic, 1994) of the same age, for which no geochemical data are available.
The mineralogical composition of the gneiss-type source of the granitoids, and the composition and the state of equilibrium of the restite are not directly accessible, and the initial melt composition could be blurred by the possible assimilation or fractionation of material during the ascent of the melt. However, the similarity in chemical and isotopic composition of the granitoids and the gneisses is obvious. Results of melting experiments on rock types similar to the average gneiss from the literature are used to compare compositional trends between starting materials and melts from experiments with differences between the possible source rock and melt from northern Chile (Table 5). The experimentally observed trend that Si, Na and K are higher in the melt than in the source, and Ti, Ca, Fe, Mg and mg-number are lower than in the source is found in the average composition of gneiss and granitoid. Alumina and A/CNK index in the melt is similar to or slightly higher than in the source. A discrimination between S- and I-type granitoids by the A/CNK index alone is impossible, if the sedimentary source already has a composition that produces an I-type signature (Fig. 3h) during partial melting. Most experiments selected in Table 5 investigated fluid-absent melting by breakdown of biotite at 5 and 10 kbar at 850–920°C. Although the melt proportions in the experiments vary considerably (<10–50 wt %) over small temperature intervals depending on the availability of water from the biotite breakdown, the resulting melt compositions are not dramatically different (Patiño Douce & Beard, 1995, 1996; Montel & Vielzeuf, 1997). New minerals formed in the presence of melt, in order of their modal importance, are plagioclase, quartz, biotite, orthopyroxene, garnet and Fe–Ti oxides (Patiño Douce & Beard, 1995, 1996; Montel & Vielzeuf, 1997; Stevens et al., 1997). These mineral assemblages are common in the granulite facies and could form the restite. The restite would still have abundant quartz and plagioclase, and might be indistinguishable from other felsic compositions in geophysical images of the crust. Garnet appears in modally important proportions (10–20% depending on the composition) at pressures >10 kbar (e.g. Patiño Douce & Beard, 1995, 1996). The steep REE pattern of the granitoid at Caleta Loa (west) indicates such a gneiss-type source (isotopic composition of sample 4/40) with abundant garnet at depth of 30 km. The experimental results are in line with the observed compositional relations between the average composition of Lower Palaeozoic gneiss as an approximation to the source and the average composition of the Palaeozoic granitoid as an approximation to melt.
|
Source of the Cretaceous granitoid
In contrast to the Upper Palaeozoic granitoids, the Sr and Nd isotopic ratios of the Cretaceous granitoid clearly show a prominent mantle contribution (Fig. 7b). All samples plot in a position intermediate between the mantle-derived mafic rocks and suggest a mixture of crustal and mantle sources. If an enriched mantle source is ruled out, two possible end member processes remain: the magma could be derived from a depleted mantle source, which was contaminated in the source by mixing with subducted sediments from the Cretaceous continental margin, or the contamination could have occurred during ascent of the melt. The two processes are indistinguishable because clastic sediments in the trench are derived from the continent and therefore mainly from the gneisses. Pb isotope ratios are above the typical values of rocks from a depleted mantle source (Fig. 9), but they are dominated by the high Pb contents in feldspar of old crustal rocks. Fractionation trends (Fig. 3) and homogeneous isotopic composition at variable SiO2 contents (Fig. 8) point to an origin from one shoshonitic parental magma, without significant contamination at the shallow level of intrusion.
Possible geodynamic settings
A possible geodynamic scenario for the Late Palaeozoic magmatism has to explain the prominent thermal anomaly in the crust that caused the large-scale melting in a broad belt >300 km wide and the lack of a mantle component in the granitoids. The Late Palaeozoic magmatism began after a lull in magmatic and tectonic activity that proceeded from the cooling of the Early Palaeozoic orogen during Silurian time (Lucassen et al., 1996b, unpublished data, 1998; Bahlburg & Hervé, 1997). During Devonian–Early Carboniferous time a passive margin basin developed at the continental margin with no or very minor magmatism, and a (hypothetical) land mass is assumed west of the present coast (Bahlburg, 1993; Bahlburg & Hervé, 1997). Compressional tectonics in the Late Carboniferous was not penetrative in many areas and is restricted to the Coastal Cordillera (Bahlburg & Breitkreuz, 1991; Bahlburg & Hervé, 1997). It took place before the onset of intrusive activity, because the Palaeozoic granitoids intruded into deformed Devonian–Late Carboniferous sediments in the Taltal–Chañaral area (292 Ma; Berg et al., 1983) and at Caleta Loa and Sierra de Moreno at 300 Ma. The intrusions were not deformed in the ductile regime and no syn-intrusional or pre-cooling deformation was observed. Neither prominent thinning nor thickening of the crust on a large scale is seen in the sedimentary record: the passive margin sediments of <3 km thickness (Bahlburg & Breitkreuz, 1991) are followed by Lower Permian platform sediments, partly carbonates, which are widespread on the South American continent (Bahlburg & Hervé, 1997), followed by more local, partly lacustrine basins with volcaniclastic sedimentation and volcanic rocks (Breitkreuz & Zeil, 1994). The topography was smooth at elevations close to the sea level during that timespan. For the Late Palaeozoic it could be speculated that a strike-slip system existed with transpression and transtension, the latter giving way for the granitoid intrusions, without a major effect on the thickness of the mainly felsic crust.
A rather similar topography for the active continental margin during Jurassic–Cretaceous time has been described in the Chilean Coastal Cordillera and Precordillera (Prinz et al., 1994). However, this setting was dominated by strong tectonic extension (Scheuber et al., 1994), growth of the crust by considerable amounts of juvenile magmas that compensated for the crustal thinning (Lucassen et al., 1996a), and strong ductile deformation (Scheuber et al., 1994). Extension probably favoured the emplacement of the mantle-derived magmas in the crust, leading to a range of compositions from pure mantle derivatives to contaminated mantle magmas as the Cretaceous intrusion of SM (south).
Temperatures above the solidus at 5–10 kbar in a granitoid crust of normal thickness (30–35 km) with a topography close to sea level are commonly interpreted as caused by magmatic underplating and/or intrusions of mantle magmas. If this process was active in the formation of the granitoids, the heat source (mantle magma) and the crustal melt (granitoid magma) did not mix, a situation that is different for the Cretaceous intrusions or the Recent andesite formation. The geodynamic regimes generating large amounts of mantle magma in the continental realm are arc–back-arc or plume–rift systems, but diagnostic igneous rocks are not known from the Late Palaeozoic and are not found before the onset of the extensional Mesozoic arcs. The Palaeozoic granites in the Taltal–Chañaral area were formerly classified as volcanic arc granites (Brown, 1991) using the discrimination diagrams of Pearce et al., (1984) and the distinction between S- and I-type granites, for example, by the A/CNK index, with many samples having a transitional composition. However, if the proposed ‘primitive’ old gneiss-type source is accepted for the granitoids, the classification is meaningless because the trace element composition of the granitoids mirrors the composition of their crustal source. In summary, a continental margin setting with subduction might be possible and does not contradict the development of the crustal thickness and topography in the Late Palaeozoic which is similar to that in the Mesozoic magmatic arcs, but it cannot be proven by the composition of the magmatic rocks as in the Mesozoic sample because Late Palaeozoic magmatism is dominated by recycling of the pre-existing crust.
The widespread slightly younger Permian volcanic rocks might be better candidates to constrain the geodynamic setting by magma compositions, because the volcanism is bimodal with abundant silicic and (very) minor mafic rocks. Unfortunately, these rocks have not yet been investigated with comparable methods. Zircons from the Permian ignimbrites indicate an inherited Precambrian lead component (Breitkreuz & Van Schmus, 1996) and recycling of Lower Palaeozoic gneisses that contain inherited zircons (Damm et al., 1990; Lezaun et al., 1997). The first appearance of clearly mantle-derived magmas is distinctly later, in Early Jurassic (Figs 1 and 7b). The time-related change from crustal-derived (older) to mantle-derived (younger) melts seems to be common in Andean plutonic rocks (Pankhurst et al., 1988).
Evidence for the possible geotectonic setting from other sections of the Pacific margin is rare and ambiguous. Sediments and igneous rock fragments of the Chañaral Melange were interpreted as a Late Palaeozoic subduction-related accretionary wedge (Bell, 1987), but no unambiguous subduction-related metamorphism or rock compositions are known from the Melange and age constraints are poor. Mpodozis & Kay, (1992) speculated on a subduction–terrane collision scenario for the Late Palaeozoic using the composition of granitoids in Chile between 28 and 31°S. The compositions of these Late Palaeozoic granitoids appear to be rather similar compared with those of the Late Palaeozoic granitoids north of 27°S. A significant change in the granitoid's composition towards a mantle-dominated Sr isotope signature occurs first in the Late Triassic (210 Ma), similar in age to the first occurrences of mantle derivatives north of 27°S (Berg & Baumann, 1985; Pichowiak, 1994). Llambias & Sato, (1995) investigated a suite of granites (ages 329–247 Ma) in the Argentine part of the Cordillera Frontal (29–31°S) and proposed a cessation of subduction in the Permian and a possible transition from a magmatic arc to a collisional regime with crustal thickening and a post-orogenic stage, but they emphasized that the question of the tectonic regime and evolution of the crust during Late Palaeozoic time is not yet resolved in the area. South of 34°S an accretionary complex with high-P–low-T metamorphism is consistent with a subduction regime (Hervé, 1988; Massonne et al., 1996); however, the age of the metamorphism is poorly known in detail and could be Late Palaeozoic and/or Mesozoic (Hervé, 1988; Hervé et al., 1990).
Implications for the structure of the Andean crust
Seismic velocities for the area show the existence of rocks with densities between 2.75 and 2.90 g/cm3 from the Precordillera to the east, and no indications for thick mafic layers or a mafic lower crust with densities >2.9 g/cm3 were detected down to the Moho (Wigger et al., 1994). A large-scale prominent positive gravity anomaly and high seismic velocities are restricted to the Chilean Coastal Cordillera and attributed to the voluminous prevailing mafic magmatism during Jurassic and Cretaceous time (Götze et al., 1994; Wigger et al., 1994). We propose a crustal composition at the beginning of the Mesozoic with rocks similar to the high-grade gneisses as a result of the Early Palaeozoic metamorphism and orogeny. The upper crust comprises gneisses found at the outcrop level, and mid- and lower crust comprise the granulite facies equivalent of these gneisses. The Upper Palaeozoic granitoids are products of recycling of this type of crust without change of the bulk composition. Later changes of this composition are largely restricted to the Jurassic magmatic belt with large amounts of prevailing mantle derived mafic rocks in the Coastal Cordillera (Fig. 1).
The thickening of the currently 70 km thick crust in the Central Andes started in Tertiary time, mainly caused by tectonic processes, and with only a minor amount by magmatic addition (e.g. Allmendinger et al., 1997). The geophysical observations indicate that mafic material from the Coastal Cordillera was not involved in this process and therefore the whole crust must comprise mainly felsic rocks that formed during the Early Palaeozoic from a Proterozoic protolith (tDM 1.65 Ga) or rocks such as the Palaeozoic granitoids (tDM 1.60 Ga) that formed during subsequent periods of crustal recycling. Isotopic compositions of Late Cenozoic magmatic rocks require a crustal source with low Sri ratio, which we interpret as the Early Palaeozoic mid–lower crust, which was depleted in Rb. Our Fig. 7b shows that all magmatic rocks of various ages from the segment of the Andes between 21 and 26°S plot on a well-defined array with an astonishingly small spread also for the SiO2-rich compositions, from the depleted mantle array to a low-radiogenic (Sr) crust. The late Cenozoic ignimbrites are probably melts with a signature dominated by the crust, and, considering the large amount of ignimbrite melts, recycling of the Palaeozoic metamorphic–magmatic basement is an important process still continuing in the Andes.
|
Appendix |
|---|
|
TOPABSTRACTIntroductionGeological SettingResultsDiscussionAppendixReferences |
|---|
Analytical methods
XRF and ICP-AES
Major and trace element contents of 89 magmatic rocks and 35 gneisses or migmatites were determined by XRF techniques on fused glass discs at TU-Berlin using Oxiquant and X-40 Philips software. Representative analyses are shown in Table 1, and the complete set of analyses is available from the authors. REE contents were determined by ICP-AES at the GFZ-Potsdam (Zuleger & Erzinger, 1988
) on a subset of magmatic rocks and gneisses (Table 1).
Sm–Nd, Rb–Sr and Pb isotopes
Mass-spectrometric analytical techniques at University of London Radiogenic Isotope Laboratory, Royal Holloway College have been described by Thirlwall (1982, 1991). The Nd laboratory standard gave 0.511423 ± 7 (2 SD, n = 69) and 0.511419 ± 9 (2 SD, n = 12) during the two periods of measurement, equivalent to a value for the La Jolla standard of 0.511857. 143Nd/144Nd ratios were normalized to 142Nd/144Nd = 1.14187, equivalent to 146Nd/144Nd = 0.72190. Sm and Nd contents were determined by isotope dilution, on an aliquot of the same dissolution as used for 143Nd/144Nd. The assumed error on the 147Sm/144Nd ratio is 0.1%. The Sr standard NBS SRM 987 gave 0.710244 ± 22 (2 SD, n = 14) during the period of study. Rb and Sr contents were determined by XRF at the TU-Berlin, and for the granitoid samples by XRF at Royal Holloway College on powder pellets.
Sr, Sm and Nd of some of the gneisses were measured at Zentrallaboratorium für Geochronologie, WWU-Münster. After dissolution, bulk REE were separated by conventional cation exchange technique. The Sr fraction was eluted from the same aliquot and separated from Rb by standard cation exchange technique. Nd–Sm were separated using quartz glass chromatographic columns with Teflon powder coated with HDEHP (Richard et al., 1976). Nd was eluated using 0.17N HCl, followed by Sm in 0.4N HCl. Sr and Nd isotope ratios and Sm and Nd concentrations were measured on a VG Sector 54 mass spectrometer. During the time of measurements the La Jolla Nd standard yielded 143Nd/144Nd = 0.511835 ± 4 (2 SD; n = 10). An external precision (reproducibility) of 0.1% is assumed for the 147Sm/144Nd ratio from the Sm–Nd concentrations by the isotope dilution technique. Nd isotope ratios were normalized to 146Nd/144Nd = 0.72190. The Sr standard NBS 987 yielded 87Sr/86Sr = 0.710261 ± 11 (2 SD; n = 4) during the period of the measurements.
Pb isotope ratios were analysed on handpicked feldspar separates from a subset of samples at Zentrallaboratorium für Geochronologie, Universität Münster. Feldspars were leached in HF and HNO3 before final dissolution, to remove possible radiogenic Pb (DeWolf & Mezger, 1994). Lead was separated by standard ion exchange techniques using HBr and HCl chemistry. The Pb standard NBS981 yielded 204Pb/206Pb = 0.05923 ± 5, 207Pb/206Pb = 0.91345 ± 18, 208Pb/206Pb = 2.1614 ± 4 (2 SD; n = 5) during the course of the measurements. Lead isotope ratios from all samples were corrected for fractionation based on the values of the NBS standard SRM 981 standard. The calculated average fractionation factor is 1.0015 per mass unit.
K–Ar dating
Minerals were separated by magnetic separator and by handpicking under the binocular microscope. Purified biotite was ground in pure alcohol to remove altered rims that might have suffered a loss of Ar or K. Details of sample preparation, and argon and potassium analyses for the laboratory in Göttingen have been given by Wemmer (1991). Potassium was determined in duplicate by flame photometry using Eppendorf Elex 63/61apparatus. The samples were dissolved in a mixture of HF and HNO3 according to the technique of Heinrichs & Hermann (1990). CsCl and LiCl were added as an ionization buffer and internal standard, respectively. The argon isotopic composition was measured in a Pyrex glass extraction and purification line coupled to a VG 1200 C noble gas mass spectrometer operating in static mode. The amount of radiogenic 40Ar was determined by the isotope dilution method using a highly enriched 38Ar spike from Schumacher, Bern (Schumacher, 1975). The spike was calibrated against the biotite standard HD-B1 (Fuhrmann et al., 1987). The age calculations were based on the constants recommended by the IUGS quoted by Steiger & Jäger (1977). The analytical error for the K/Ar age calculations is given on the 95% confidence level (2) in Table 3.
|
Acknowledgements |
|---|
The research for this paper is a result of Sfb 267 ‘Deformation processes in the Andes’. We thank Gerry Ingram (London), Heidi Baier and Sigmund Rochnowski (both Münster) for help with the isotope analysis, and Klaus Wemmer, Universität Göttingen, who carried out most of the K–Ar age determinations. We thank Maren Krause for sample preparation and Erika Kramer for measuring the REE at the GeoForschungsZentrum Potsdam; and Jose Infanta, Andreas Laber, Sven Lewerenz, Rike Wegener and Hans Wilke, who helped with the field work and sample preparation. Raul Becchio, UNSA Salta, Argentina, is thanked for supplying his unpublished data on the Argentine basement, which allowed us to compare our data on a broader regional scale. Reviews by N. Petford, C. Rapela and an anonymous reviewer are gratefully acknowledged. We thank the Deutsche Forschungsgemeinschaft for financial support and Grant L-501/2 to F.L. Isotopes were analysed in the Radiogenic Isotope Laboratory at Royal Holloway, University of London, and Zentrallaboratorium für Geochronologie, Universität Münster. The laboratory at Royal Holloway is a ULIRS facility.
* Corresponding author. e-mail: luca0938{at}mailszrz.zrz.tu-berlin.de
|
References |
|---|
|
TOPABSTRACTIntroductionGeological SettingResultsDiscussionAppendixReferences |
|---|
Aitcheson S. J., Harmon R. S., Moorbath S., Schneider A., Soira-Escalante E., Steele G., Swainbank I., Wörner G. Pb isotopes define basement domains of the Altiplano, central Andes. Geology (1995) 23:555–558.
Allmendinger W., Jordan T. E., Kay S. M., Isacks B. L. The evolution of the Altiplano-Puna plateau of the Central Andes. Annual Review of Earth and Planetary Sciences (1997) 25:139–174.[Web of Science]
Ayres M., Harris N. REE fractionation and Nd-isotope disequilibrium during crustal anatexis: constraints from Himalayan leucogranites. Chemical Geology (1997) 139:249–269.[Web of Science]
Baeza L. Petrography and tectonics of the plutonic and metamorphic complexes of Limón Verde and Mejillones Peninsula, northern Chile. (1984) Tübingen: Eberhard-Karl Universität. 205. Ph.D. Thesis.
Bahlburg H. Hypothetical southeast Pacific continent revisited: new evidence from the middle Palaeozoic basins of northern Chile. Geology (1993) 21:909–912.
Bahlburg H., Breitkreuz C. Palaeozoic evolution of active margin basins in the southern Central Andes (northwestern Argentina and northern Chile). Journal of South American Earth Sciences (1991) 4:171–188.[Web of Science]
Bahlburg H., Hervé F. Geodynamic evolution and tectonostratigraphic terranes of northwestern Argentina and northern Chile. Geological Society of America Bulletin (1997) 109:869–884.
Becchio R., Lucassen F., Franz G., Viramonte J. Condiciones de P–T del basamento metamorfico de alto grado, borde oriental de la Puna Austral, Argentina. In: VIII Congreso Geologica Chileno, Antofagasta, Chile, Actas Vol. II (1997) Antofagasta, Chile: Universidad Católica de Norte. 1220–1224.
Bell C. M. The origin of the Upper Palaeozoic Chañaral melange of N Chile. Journal of the Geological Society, London (1987) 144:599–610.
Berg K., Baumann A. Plutonic and metasedimentary rocks from the Coastal Range of northern Chile: Rb–Sr and U–Pb isotopic systematics. Earth and Planetary Science Letters (1985) 75:101–115.[Web of Science]
Berg K., Breitkreuz C., Damm K. W., Pichowiak S., Zeil W. The North-Chilean Coast Range—an example for the development of an active continental margin. Geologische Rundschau (1983) 72:715–731.
Bogdanic T., Espinoza S. Tectono-sedimentary evolution of the Cretaceous–Early Tertiary and metallogenic scheme of Northern Chile. In: Cretaceous Tectonics of the Andes—Salfity J., ed. (1994) Braunschweig: Vieweg. 213–265.
Breitkreuz C., Van Schmus W. R. U/Pb geochronology and significance of Late Permian ignimbrites in Northern Chile. Journal of South American Earth Sciences (1996) 9:281–293.[Web of Science]
Breitkreuz C., Zeil W. Geodynamic and magmatic stages on a traverse through the Andes between 20° and 24° (N Chile, S Bolivia, NW Argentina). Journal of the Geological Society, London (1984) 141:861–868.
Breitkreuz C., Zeil W. The Late Carboniferous to Triassic volcanic belt in Northern Chile. In: Tectonics of the Southern Central Andes—Reutter K. J., Scheuber E., Wigger P. J., eds. (1994) Berlin: Springer. 277–292.
Brown M. Comparative geochemical interpretation of Permian–Triassic plutonic complexes of the Coastal Range and Altiplano (23°30' to 26°30'S), northern Chile. Geological Society of America, Special Paper (1991) 265:157–171.
Coira B., Kay S. M., Peréz B., Woll B., Hanning M., Flores P. Magmatic source and tectonic setting of Gondwana margin Ordovician magmas. In: Laurentian-Gondwana Terranes before Pangea—Ramos V. A., Keppie D., eds. (1999) Geological Society of America, Special Paper, in press.
Cox K. G. Continental magmatic underplating. Philosophical Transactions of the Royal Society of London, Series A (1993) 342:155–166.
Dallmeyer R. D., Brown M., Grocott J., Taylor G. K., Treloar P. J. Mesozoic magmatic and tectonic events within the Andean plate boundary zone, 26°-27°30'S, North Chile: constraints from 40Ar/39Ar mineral ages. Journal of Geology (1996) 104:19–40.[Web of Science]
Damm K. W., Pichowiak S., Harmon R. S., Todt W., Kelley S., Omarini R., Niemeyer H. Pre-Mesozoic evolution of the central Andes; the basement revisited. Geological Society of America, Special Paper (1990) 241:101–126.
Damm K. W., Harmon R. S., Kelley S. Some isotope and geochemical constraints on the origin and evolution of the Central Andean basement (19°-24°S). In: Tectonics of the Southern Central Andes—Reutter K. J., Scheuber E., Wigger P. J., eds. (1994) Heidelberg: Springer. 263–275.
DeWolf C. P., Mezger K. Lead isotope analyses of leached feldspars: constraints on the early crustal history of the Grenville Orogen. Geochimica et Cosmochimica Acta (1994) 58:5537–5550.[Web of Science]
Evensen N. M., Hamilton P. J., O'Nions R. K. Rare-earth abundances in chondritic meteorites. Geochimica et Cosmochimica Acta (1978) 42:1199–1212.[Web of Science]
Finger F., Clemens J. D. Migmatization and 'secondary’ granitic magmas: effects of emplacement and crystallization of ‘primary’ granitoids in Southern Bohemia, Austria. Contributions to Mineralogy and Petrology (1995) 120:311–326.[Web of Science]
Francis P. W., Sparks R. S. J., Hawkesworth C. J., Thorpe R. S., Pyle D. M., Tait S. R., Montovani M. S., McDermott F. Petrology and geochemistry of volcanic rocks of the Cerro Galan caldera, northwest Argentina. Geological Magazine (1989) 126:515–547.[Abstract]
Fuhrmann U., Lippolt H. J., Hess J. C. Examination of some proposed K–Ar standards: 40Ar/39Ar analyses and conventional K–Ar data. Chemical Geology (Isotope Geosciences Section) (1987) 66:41–51.
Goldstein S. L., O'Nions R. K., Hamilton P. J. A Sm–Nd study of atmospheric dust and particulates from major river systems. Earth and Planetary Science Letters (1984) 70:221–236.[Web of Science]
Götze H. J., Lahmeyer B., Schmidt S., Strunk S. The lithospheric structure of the Central Andes (20°-26°S) as inferred from interpretation of regional gravity. In: Tectonics of the Southern Central Andes—Reutter K. J., Scheuber E., Wigger P. J., eds. (1994) Heidelberg: Springer. 7–22.
Harris N. Radiogenic isotopes and the interpretation of granitic rocks. Episodes (1996) 19:107–113.[Web of Science]
Heinrichs H., Herrmann A. G. Praktikum der analytischen Geochemie (1990) Berlin: Springer. 669.
Hervé F. Late Paleozoic subduction and accretion in southern Chile. Episodes (1988) 11:183–188.[Web of Science]
Hervé F., Pankhurst R. J., Brook M., Alfaro G., Frutos J., Miller H., Schira W., Amstutz G. C. Rb–Sr and Sm–Nd data from some massive sulfide occurrences in the metamorphic basement of South–Central Chile. In: Stratabound Ore Deposits in the Andes—Fonboté I., Amstutz G. C., Cardozo E., Frutos J., eds. (1990) Heidelberg: Springer. 221–228.
Lezaun J., Beck A., Heber V., Henjes-Kunst F., Zinngrebe E., Lucassen F., Rößling R., Wörner G. Pre-Mesozoic evolution of the Andean metamorphic basement at 18°S. In: VIII Congreso Geologico Chileno, Antofagasta, Chile, Actas Vol. II (1997) Antofagasta, Chile: Universidad Católica de Norte. 1339–1343.
Llambias E. J., Sato A. M. El batolito de Colangüel: transición entre orogénesis y anorogénesis. Revista de la Asociación Geológica Argentina (1995) 50:111–131.
Lucassen F., Franz G. Arc related Jurassic igneous and metaigneous rocks in the Coastal Cordillera of northern Chile/Region Antofagasta. Lithos (1994) 32:273–298.[Web of Science]
Lucassen F., Thirlwall M. F. Sm–Nd ages of mafic rocks from the Coastal Cordillera at 24°S, Northern Chile. Geologische Rundschau (1998) 86:767–774.
Lucassen F., Laber A., Franz G. Geology and petrology of metamorphic rocks from the Preandean basement of northern Chile, 18°-24°. In: VII Congreso Geologico Chileno, Concepción 1994, Actas Vol. I (1994) Concepción, Chile: Universidad de Concepción. 96–100.
Lucassen F., Fowler C. M. R., Franz G. Formation of magmatic crust at the Andean continental margin during early Mesozoic: a geological and thermal model of the North Chilean Coast Range. Tectonophysics (1996a) 262:263–279.[Web of Science]
Lucassen F., Wilke H. G., Viramonte J., Becchio R., Franz G., Laber A., Wemmer K., Vroon P. The Palaeozoic of the Central Andes (18°-26°S)—a metamorphic view. In: Third International Symposium on Andean Geodynamics, St Malo, France, Abstract Volume (1996b) Paris: ORSTOM. 779–782.
Lucassen F., Franz G., Laber A. Permian high pressure rocks—the basement of Sierra de Limón Verde in N-Chile. Journal of South American Earth Sciences (1999a) (in press).
Lucassen F., Lewerenz S., Franz G., Viramonte J., Mezger K. Metamorphism, isotopic ages and composition of lower crustal granulite xenoliths from the Cretaceous Salta Rift, Argentina. Contributions to Mineralogy and Petrology (1999b) 134:325–341.[Web of Science]
Maksaev V. Metallogeny, geological evolution, and thermochronology of the Chilean Andes between latitudes 21° and 26° south, and the origin of the major porphyry copper deposits. (1990) Halifax, N.S: Dalhouse University. 555. Ph.D. Thesis.
Maksaev V., Marinovic N. Cuadrángulos Cerro de la Mica, Quillagua, Cerro Pasada y Oficina Prosperidad, Región Antofagasta. Carta Geológica de Chile, N°45–48 (1980) Santiago: Instituto de Investigaciones Geológicas. 63.
Massonne H.-J., Hervé F., Muñoz V., Willner A. P. New petrological results on high-pressure, low-temperature metamorphism of the Upper Palaeozoic basement of Central Chile. In: Third International Symposium on Andean Geodynamics, St Malo, France, Abstract Volume (1996) Paris: ORSTOM. 783–785.
Miller H., Toselli A. J., Rossi de Toselli J., Aceñolaza F. G. Regional and geochronological development of the metamorphic basement in Northwest Argentina. Zentralblatt für Geologie und Paläontologie, Teil I, Heft 1 und 2 (1994) 263–273.
Montel J. M., Vielzeuf D. Partial melting of metagreywackes. Part II. Compositions of minerals and melts. Contributions to Mineralogy and Petrology (1997) 128:176–196.[Web of Science]
Mpodozis C., Kay S. M. Late Palaeozoic to Triassic evolution of the Gondwana margevidence from the Chilean Frontal Cordilleran batholiths (28°S to 31°S). Geological Society of America Bulletin (1992) 104:999–1014.
Nixon P. H., ed. Mantle Xenoliths (1987) Chichester: John Wiley. 844.
Ort M. H., Coira B. L., Mazoni M. M. Generation of a crust–mantle magma mixture: magma sources and contamination at Cerro Panizoz, central Andes. Contributions to Mineralogy and Petrology (1996) 123:308–322.[Web of Science]
Pankhurst R. J., Hole M. J., Brook M. Isotope evidence for the origin of Andean granites. Transactions of the Royal Society of Edinburgh: Earth Sciences (1988) 79:123–133.
Patiño Douce A. E., Beard J. S. Dehydration-melting of biotite gneiss and quartz amphibolite from 3 to 15 kbar. Journal of Petrology (1995) 36:707–738.
Patiño Douce A. E., Beard J. S. Effects of P, f(O2) and Mg/Fe ratio on dehydration melting of model metagreywackes. Journal of Petrology (1996) 37:999–1024.
Pearce J. A., Harris N. B. W., Tindle A. G. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology (1984) 25:956–983.
Pichowiak S. Early Jurassic to Early Cretaceous magmatism in the Coastal Cordillera and the Central Depression of north Chile. In: Tectonics of the Southern Central Andes—Reutter K. J., Scheuber E., Wigger P. J., eds. (1994) Heidelberg: Springer. 203–218.
Prinz P., Wilke H. G., von Hillebrandt A. Sediment accumulation and subsidence history in the Mesozoic marginal basin of northern Chile. In: Tectonics of the Southern Central Andes—Reutter K. J., Scheuber E., Wigger P. J., eds. (1994) Heidelberg: Springer. 219–232.
Rapela C. W., Toselli A., Heaman L., Saavedra J. Granite plutonism of the Sierras Pampeanas; an inner cordilleran Palaeozoic arc in the southern Andes. Geological Society of America, Special Paper (1990) 241:77–90.
Richard P., Shimizu H., Allègre C. J. 143Nd/146Nd, a natural tracer: an application to oceanic basalts. Earth and Planetary Science Letters (1976) 31:269–278.[Web of Science]
Rogers G., Hawkesworth C. J. A geochemical traverse across the North Chilean Andes: evidence for crust generation from the mantle wedge. Earth and Planetary Science Letters (1989) 91:271–285.[Web of Science]
Rudnick R. L., Presper T. Geochemistry of intermediate- to high pressure granulites. In: Granulites and Crustal Evolution—Vielzeuf D., Vidal Ph., eds. (1990) Dordrecht: Kluwer Academic. 523–550.
Scheuber E., Bogdanic T., Jensen A., Reutter K. J. Tectonic development of the North Chilean Andes in relation to plate convergence and magmatism since the Jurassic. In: Tectonics of the Southern Central Andes—Reutter K. J., Scheuber E., Wigger P. J., eds. (1994) Heidelberg: Springer. 7–22.
Schumacher E. Herstellung von 99.9997% 38Ar für die 40K/40Ar Geochronologie. Geochron. Chimia (1975) 24:441–442.
Steiger R. H., 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.[Web of Science]
Stevens G., Clemens J. D., Droop G. T. R. Melt production during granulite-facies anatexis: experimental data from ‘primitive’ metasedimentary protoliths. Contributions to Mineralogy and Petrology (1997) 128:352–370.[Web of Science]
Skarmeta J., Marinovic N. Hoja Quillagua, Region de Antofagasta. Carta Geologica de Chile N°51 (1981) Santiago: SERNAGEOMIN. 63.
Smoje I., Marinovic N. Intrusivos de Carbonifero Permico en la Cordillera de Domeyko: nuevos antecedentes radiometricos. In: VII Congreso Geologico Chileno 1994, Concepción, Actas Vol. II (1994) Concepción, Chile: Universidad de Concepción. 1213–1216.
Stacey J. S., Kramers J. D. Approximation of terrestrial lead isotope evolution by a two-stage model. Earth and Planetary Science Letters (1975) 26:207–221.[Web of Science]
Taylor S. R., McLennan S. M. The composition and evolution of the continental crust: rare earth element evidence from sedimentary rocks. Philosophical Transactions of the Royal Society of London, Series A (1981) 301:381–399.
Taylor S. R., McLennan S. M. The Continental Crust: its Composition and Evolution (1985) Oxford: Blackwell Scientific.
Thirlwall M. F. A triple-filament method for rapid and precise analysis of the rare-earth elements by isotope dilution. Chemical Geology (1982) 35:155–166.[Web of Science]
Thirlwall M. F. High precision multicollector analysis of low levels of Nd as oxide. Chemical Geology (Isotope Geosciences Section) (1991) 94:13–22.
Tilton G. R., Barreiro B. A. Origin of lead in Andean calc-alkaline lavas, southern Peru. Science (1980) 210:1245–1247.
Tosdal R. M. The Amazon–Laurentian connection as viewed from the Middle Proterozoic rocks in the central Andes, western Peru and northern Chile. Tectonics (1996) 15:827–842.[Web of Science]
Vielzeuf D., Montel J. M. Partial melting of metagreywackes. Part I. Fluid-absent experiments and phase relationships. Contributions to Mineralogy and Petrology (1994) 117:375–393.[Web of Science]
Wemmer K. K/Ar-Altersdatierungsmöglichkeiten für retrograde Deformationsprozesse im spröden und duktilen Bereich—Beispiele aus der KTB-Vorbohrung (Oberpfalz) und dem Bereich der Insubrischen Linie (N-Italien). Göttinger Arbeiten Geologie und Paläontologie (1991) 51:61.
Wilson M. Igneous Petrogenesis (1989) London: Chapman & Hall. 466.
Wigger P. J., Schmitz M., Araneda M., Asch G., Baldzuhn S., Giese P., Heinsohn W. D., Martinez E., Ricaldi E., Rüwer P., Viramonte J. Variation in the crustal structure of the Southern Central Andes deduced from seismic refraction investigations. In: Tectonics of the Southern Central Andes—Reutter K. J., Scheuber E., Wigger P. J., eds. (1994) Heidelberg: Springer. 23–48.
Wittenbrink R. Zeitliche Variationen der Magmengenese miozäner bis quartärer Vulkanite in südlichen Bereich der Zentralen Vulkanischen Zone der Anden (CVZ, 25°-26°S, 67°-69°W). Berliner Geowissenschaftliche Abhandlungen, Reihe A (1997) 193:135.
Wörner G., Moorbath S., Horn S., Entenmann J., Harmon R., Davidson J., Lopez-Escobar L. Large- and fine-scale geochemical variations along the Andean Arc of Northern Chile (17.5°-22°S). In: Tectonics of the Southern Central Andes—Reutter K. J., Scheuber E., Wigger P. J., eds. (1994) Heidelberg: Springer. 77–92.
York D. Least squares fitting of a straight line with correlated errors. Earth and Planetary Science Letters (1969) 5:320–324.[Web of Science]
Zuleger E., Erzinger J. Determination of REE and Y in silicate materials with ICP-AES. Fresenius’ Zeitschrift für Analytische Chemie (1988) 332:140–143.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||