Journal of Petrology Volume 41 Number 9 Pages 1413-1438 2000
© Oxford University Press 2000
Early Cretaceous Basaltic and Rhyolitic Magmatism in Southern Uruguay Associated with the Opening of the South Atlantic
1DEPARTMENT OF EARTH SCIENCES, THE OPEN UNIVERSITY, WALTON HALL, MILTON KEYNES MK7 6AA, UK
2FACULTEIT DER AARDWETENSCHAPPEN, VRIJE UNIVERSITEIT, 1085 DE BOELELAAN, 1081 HV AMSTERDAM, THE NETHERLANDS
3DANISH LITHOSPHERE CENTRE, ØSTER VOLDGADE 10, L, DK-1350, COPENHAGEN K, DENMARK
4DEPARTMENT OF GEOLOGICAL SCIENCES, UNIVERSITY OF CAPE TOWN, RONDEBOSCH, 7700 SOUTH AFRICA
5DEPARTMENTO DE GEOFÍSICO, INSTITUTO ASTRONÔMICO E GEOFÍSICO, UNIVERSIDADE DE SÃO PAULO, RUA DO MATÃO 1226, CEP 05508-900, SÃO PAULO SP, BRAZIL
Received February 18, 1999; Revised typescript accepted January 25, 2000
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
The Early Cretaceous volcanic rocks of southern Uruguay comprise mafic and felsic volcanics. The position of these outcrops at the southern edge of the Paraná–Etendeka continental flood basalt province provides an opportunity to investigate possible lateral variations in both mafic and more evolved rock types towards the margins of such an area of plume-related magmatism. The mafic lavas are divided into two compositionally distinct magma types. The more voluminous Treinte Y Trés magma type is similar to the low-Ti basalts of the Paraná flood basalt province. The Santa Lucía magma type is a distinct and rare basalt type with ocean-island basalt type asthenospheric affinities (high Nb/La, low 87Sr/86Sri). The felsic volcanics are divided into two series, the Lavalleja Series and the Aigüa Series. The Lavalleja Series are chemically and isotopically similar to the Paraná–Etendeka low-Ti rhyolites, and are considered to be related to the Treinte Y Trés lavas by extensive fractionation and crustal assimilation. The Aigüa Series have low 143Nd/144Ndi and low 87Sr/86Sri and unlike the rhyolites of the Paraná, are interpreted as melts of pre-existing mafic lower crust that subsequently underwent extreme fractionation. The differences observed in the felsic suites may be linked to differences in the volumes of the associated basalts and the amounts of extension.
KEY WORDS: South America; flood basalts; felsic volcanics; crustal melts; plume
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
Mesozoic magmatism associated with the opening of the South Atlantic Ocean is widespread throughout southern Brazil, Uruguay, eastern Paraguay and Argentina, and SW Africa. The extensive Paraná volcanic field in South America and the less voluminous Etendeka volcanic field of Namibia, which formed a single continental flood basalt (CFB) province before South Atlantic rifting (Erlank et al., 1984
; Bellieni et al., 1986), were erupted during the Early Cretaceous. This large igneous province is linked to the Tristan mantle plume via the Rio Grande Rise and the Walvis Ridge (Fig. 1a inset; O’Connor & Duncan, 1990). In addition, small-volume Early and Late Cretaceous alkalic provinces are widely scattered in Brazil, Paraguay, Angola and Namibia (Gibson et al., 1995; Milner et al., 1995).
|
This paper focuses on a volumetrically minor, but compositionally diverse suite of mafic and felsic igneous rocks in southern Uruguay. They overlie a significant positive gravity anomaly that has been modelled as a large mafic body within the crust (S. Hallinan, unpublished data, 1992). Recent 40Ar–39Ar ages (Stewart et al., 1996
) demonstrate that this bimodal volcanism was contemporaneous with the main Paraná–Etendeka magmatism, and does not overlap in age with earlier extensive Jurassic rhyolitic magmatism found throughout Argentina and Chile to the south (Pankhurst & Rapela, 1995; Pankhurst et al., 1998). Furthermore, its position at the edge of the extensive Paraná–Etendeka CFB province is of considerable interest in attempts to establish the lateral variations in source compositions and degrees of partial melting across an area of plume-related magmatism. This magmatism provides direct chemical evidence of mantle plume involvement in South America during the Early Cretaceous, previously undocumented in the Paraná province further north. In addition, the range of rhyolite compositions provides new insights into the continuing controversy of crustal melting vs assimilation and fractional crystallization as an origin for rhyolites associated with CFB. We present new compositional data on the mafic and felsic volcanics of southern Uruguay to assess their origins and their relationship to the Paraná–Etendeka magmatism. |
FIELD RELATIONSHIPS |
|---|
|
TOPABSTRACTINTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
The Mesozoic magmatic rocks of Uruguay are at present preserved in two principal regions. The main Paraná basalt lava field extends into northwestern Uruguay, covering an area of 100 km x 200 km. These basalts are intercalated with aeolian sandstones, and are referred to as the Arapey Formation (Caorsi & Goñi, 1958
; Bossi & Navarro, 1988). Analyses of surface samples and borehole material from western Uruguay indicate that the lavas are chemically similar to those in the Paraná basin, with basalts of the low-Ti/Y Gramado magma type overlying high-Ti/Y Paranápanema magmas (Turner et al., 1999b).
In southern Uruguay, Mesozoic volcanic rocks extend from the Santa Lucía Basin to the Laguna Merín Basin (Fig. 1a). The Santa Lucía Basin (Riccardi, 1988), which is also known as the Canelones Basin (Urien & Zambrano, 1973), lies in a system of faulted basement blocks trending 
N60°E (Fig. 1a). This basin, together with the similarly orientated Laguna Merín Basin and the ESE-trending Salado and Colorado Basins of Argentina (Fig. 1a), were successively infilled with sedimentary and volcanic sequences during Late Jurassic and Early Cretaceous times (Riccardi, 1988). The Santa Lucía and Laguna Merín regions are underlain by a positive gravity anomaly of +90 mGal (Fig. 1b), with an estimated mass excess of 7 (± 2) x 1015 kg (S. Hallinan, unpublished data, 1992), assuming a density contrast of 200 kg/m3. This is interpreted as a large mafic body with a volume of 35 000 km3. The shape is a crude oblong 120 km long and, assuming the density contrast is effective over 15 km of crustal thickness, the width is 20 km. It is orientated in a rough E–W direction similar to many of the other structural features related to the internal deformation of the South American plate during the Jurassic rifting related to the early opening of the South Atlantic Ocean (Nürnberg & Müller, 1991).
The volcanic rocks are bimodal in composition, with the more mafic rocks ranging from basalts to andesites. These mafic lavas occur as numerous isolated outcrops, principally in the north of the area, at Treinte Y Trés and Piraraja, with minor exposures at Aigüa, Lascano and Paso de los Talas (Fig. 1b). It is likely that they originally covered a larger area, comparable in extent with the underlying gravity anomaly. A borehole in eastern Uruguay (DDH 502: 33·2°S, 53·6°W, Fig. 1b) recovered a sequence of 1 km thickness of similar basalts overlain by 267 m of sedimentary rocks (Turner et al., 1999a).
Felsic rocks dominate the landscape, forming topographic highs from Arequita to Lascano. As there is no obvious post-eruption deformation, it is likely that the felsic rocks overlie the mafic lavas, or overstep them to lie directly on basement, although no clear contact between mafic and felsic suites has been observed in the field. This is consistent with new 40Ar–39Ar ages in that mafic lavas yield ages of 134–130 Ma, and felsic volcanics yield ages of 132–124 Ma (Stewart et al., 1996; Kirstein, 1997). The southern Uruguayan basalts, basaltic andesites and andesites are termed the Puerto Gómez Formation, and the trachydacites, trachytes and rhyolites the Arequita Formation, respectively. The Puerto Gómez Formation lavas were previously considered to be Jurassic in age (Walther, 1927; Preciozzi et al., 1980; Sprechmann et al., 1981).
An intrusive syenite plug and ring dyke complex, which forms the Valle Chico suite (Bossi & Navarro, 1988), is exposed in the regions of Cerro Partido, Alferez, Paso de los Talas and Mariscala (Fig. 1b). The syenite outcrops are extensively weathered, with the result that exposures are featureless flat areas with no obvious structures. The syenite has been dated by 40Ar–39Ar at 132 ± 2 Ma (Stewart et al., 1996). Similar alkalic complexes, broadly contemporaneous with the main flood basalt event, occur both in South America and in Namibia (e.g. Toyoda et al., 1994; Milner et al., 1995; Renne et al., 1996). The composition of the Valle Chico syenite is discussed in the context of the petrogenesis of the trachydacites, trachytes and rhyolites. Milner & Ewart (1989) demonstrated that the Awahab Formation rhyolites of the Etendeka were erupted from the Messum complex, which is a large mixed tholeiitic and alkalic centre with intrusive nepheline syenites, quartz syenites, gabbros and granodiorites. In this context, it is notable that there is evidence of fragmented rhyolites and vent breccia in the Alférez region near the syenite complex, indicating that a source of the Uruguay felsic volcanics was nearby and that it might similarly be connected with the alkalic complex.
|
BASALT–ANDESITE PETROGRAPHY AND MINERAL COMPOSITIONS |
|---|
|
TOPABSTRACTINTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
The mafic rocks are well exposed at Treinte Y Trés (Fig. 1b), where the basalts–andesites are dark grey, feldspar-phyric (5–30%, up to 1 cm long), and often vesiculated with zeolites and calcite amygdales. Rare samples are found in which scoria fragments are incorporated in a basaltic matrix. The scoria fragments are angular, highly vesiculated and variable in size (1–8 cm in length), and suggest an explosive origin, but no vents were recognized in the field. The state of preservation of the mafic volcanics is poor. These mafic volcanics, where fresh, have an anhydrous primary mineralogy of clinopyroxene and plagioclase feldspar together with iron oxides ± olivine. Most of the samples are porphyritic (up to 40%), and they contain variably scaled intergrowths of euhedral, lamellar twinned plagioclase (1–3 mm), with subhedral, smaller (0·2–0·8 mm) pyroxene phenocrysts. There are some similarities with the plagioclase-rich Albin basalts and the Tafelberg basalts of the Etendeka (Erlank et al., 1984
), although the Uruguay rocks tend to have higher plagioclase contents than the latter. In contrast, many of the Paraná basalts are virtually aphyric (Comin-Chiaramonti et al., 1988).
Electron microprobe analyses of mineral compositions from selected samples are presented in Table 1. The anorthite contents of most of the plagioclase phenocrysts analysed from the basalts and basaltic andesites vary from An50 to An82, whereas andesine (An30–50) is common in the more evolved andesitic lavas (Fig. 2). Alteration of pyroxenes means that only a few microprobe analyses gave reliable compositions, and these were principally augites, which vary from ferroan to sub-calcic. Fresh olivine, not previously recognized in Paraná basalts (Comin-Chiaramonti et al., 1988), is seen in the more primitive samples. Generally olivines are highly fractured with iron oxides infilling, and the compositions vary from Fo71 to Fo77. These are within the range of olivine compositions of Fo48–84 reported from the Etendeka (Erlank et al., 1984). The Fe/Mg ratios of the olivine and lava compositions (mg-number 69–77 in olivine, and 29–37 in the lavas) indicate disequilibrium between the phenocrysts and host magma, with the liquid from which the olivines crystallized having been more Mg rich (mg-number 56–62).
|
|
, mafic samples, Treinte Y Trés; 
, mafic samples, Santa Lucía; grey triangles, trachydacites and dacites; 
, feldspars from rhyolites.
Iron oxides form between 3 and 8% of the modal composition. The grains are highly variable, with growth textures, exsolution and complex lamellae all identified using back-scattered electron images. Titano-magnetite, magnetite and ilmenite occur as both primary and secondary minerals.
|
FIELD CHARACTERISTICS AND PETROGRAPHY OF THE FELSIC VOLCANICS |
|---|
|
TOPABSTRACTINTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
The outcrops of the felsic volcanic rocks consist either of a number of individual cooling units (each up to 1 m thick) with ignimbritic flow textures, which give a layered appearance in the field, or massive units between 1 and 5 m thick. Weathering is extensive, with alteration of feldspars to clay minerals. Some of the felsic volcanics are highly vesiculated with zeolite and calcite amygdales. Phenocrysts, when present, are primarily of feldspar and quartz (<1 mm to 1 cm), and they tend to be aligned parallel to the flow direction.
The Uruguayan felsic volcanics can be subdivided into two petrographic types defined primarily on the presence or absence of quartz phenocrysts. These two types tend to have different silica contents, with the exception of two of the quartz-phyric samples that have unusually low SiO2 contents of 67 wt % (Fig. 3a). Two chemically defined series are recognized within the felsic rocks, and the presence or absence of quartz is independent of which chemical series the rocks are in. These chemical series are the Lavalleja Series, which has higher Ti/Zr and 
Ndi and lower Ta/Th and Nb abundances, and the Aigüa Series, with lower Ti/Zr and Ndi, and higher Nb and Ta/Th (Fig. 3a; see also Figs 6 and 9a, c, d, below). Felsic volcanics of both petrographic types are encountered in all the main exposures throughout the region, and so they do not appear to represent mappable units confined to particular areas. The regional dip of the rhyolites is 20° to the SW.
|
|
|
Type 1 feldspar phyric
These are predominantly the lower-silica felsic volcanics (62–73 wt % SiO2) (Fig. 3a). Feldspar is the dominant mineral phase, forming
60% of the rock, and felsic volcanics of this petrographic type occur principally in the Minas, Aigüa, Cerro Partido and Paso de los Talas regions (Fig. 1b). The type locality identified here is in the Parque Municipal de Salamanca, in Paso de los Talas. Locally there is evidence of effusive flow in these felsic rocks, with flow folds identified in some outcrops, and there are silicic dyke intrusions.
Type 2 quartz phyric
These are recognized primarily on the presence of variably sized (1–6 mm), euhedral to anhedral quartz phenocrysts. They tend to have higher SiO2 contents (70–78 wt %), with the exception of two samples from near Mariscala, which have 67 wt % SiO2 (Fig. 3a). Ignimbritic textures with flattened pumice fragments and shards are common but not ubiquitous. Subhedral and embayed alkali feldspar phenocrysts are more common than plagioclase phenocrysts. The phenocrysts are set in a devitrified, cryptocrystalline matrix within which a regular alignment of flattened fragments gives rise to a eutaxitic texture. The type locality is at Aigüa (Fig. 1b) in the Maldonado region, where phenocryst-rich felsic flows overlie vent material. Brecciated material containing large angular pumice clasts and sandstone fragments (from subaerial dune deposits) incorporated in a siliceous flow suggests a local source for these rhyolites, and eruption in an aeolian environment as seen elsewhere in the Paraná–Etendeka province (Riccardi, 1988; Milner et al., 1992; Mizusaki et al., 1992).
The two petrographic types share a number of features, particularly at Aigüa where the felsic volcanics tend to be layered and vesiculated. The primary mineralogy of both petrographic types of felsic volcanics is anhydrous, consisting of plagioclase, alkali feldspar, clinopyroxene and iron oxide, with quartz present in the more evolved type 2 high-silica rocks. Accessory minerals include zircon and apatite.
Plagioclase
Plagioclase is common both as larger phenocrysts (megacrysts: 1 cm) and smaller phenocrysts (0·5–2 mm) particularly in the lower-SiO2, type 1 rocks. Compositions vary between An30 and An60, and show a slight provinciality in that more sodic plagioclases tend to be found in the rhyolites of the Paso de los Talas region, with more calcic plagioclases in the Lascano region (Fig. 1b).
Alkali feldspar
In general, the abundance of alkali feldspar increases with increasing silica content and it is therefore one of the principal modal components in the more evolved type 2 felsic volcanics. Phenocrysts of alkali feldspar, up to 5 mm in size, occur both discretely and intergrown with plagioclase and quartz, and range in composition from anorthoclase to sanidine in the type 1 lower-silica felsic rocks. The alkali feldspar compositions in the higher-silica, type 2 rocks tend towards sanidine and with higher potassium abundances (Table 1). There appears to be a regional variation in the alkali feldspar compositions in the felsic volcanics analysed, with alkali feldspars from the Aigüa tending to be more sodic than those from the Lascano region (Fig. 2).
Clinopyroxene
Clinopyroxenes of augite composition are present in the less evolved trachydacite rock types (type 1), and there are also rare pigeonitic pyroxenes (Table 1). In general, the pyroxenes are altered and replaced predominantly by iron oxides. Iron oxides are ubiquitous as a primary phase and also as secondary phases. The compositions range between magnetite and titanomagnetite with rare ilmenite.
|
PETROLOGICAL COMPARISON WITH THE MAIN PARANÁ–ETENDEKA RHYOLITES |
|---|
|
TOPABSTRACTINTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
The Uruguay felsic volcanics can be distinguished from the rhyolites of the main Paraná–Etendeka lava field in both petrography and mineralogy. The Paraná rhyolites have been divided into two compositionally distinct groups that are also petrographically and spatially distinct (Bellieni et al., 1986
; Garland et al., 1995), and their equivalents are found in the Etendeka (Milner et al., 1992, 1995). The high-Ti Chapecó rhyolites in the north are plagioclase rich (up to 25% phenocrysts), whereas the low-Ti Palmas rhyolites in the south are sparsely porphyritic (<10% phenocrysts) with a basic mineralogy of plagioclase, pyroxene and iron oxides, primarily titanomagnetite and more rarely ilmenite (Bellieni et al., 1986; Milner et al., 1992). Both groups lack alkali feldspar and quartz, which are the main components of the more evolved Uruguayan felsic volcanics. This is consistent with their comparatively lower silica contents (Chapecó: 64–68 wt % SiO2; Palmas: 64–72 wt % SiO2).
Interestingly, the quartz-phyric felsic volcanics are similar in appearance to the rheomorphic rhyolitic rocks of the Erongo volcanic complex, west–central Namibia, which have a similar mineralogy (quartz and alkali feldspar) and textures (ignimbritic), although the Erongo rocks are considered to be late Jurassic in age (Pirajno, 1990).
The Paraná–Etendeka rhyolites record relatively high eruption temperatures (950–1100°C) from a number of pyroxene and plagioclase geothermometers (Bellieni et al., 1986; Milner et al., 1992; Garland et al., 1995). These contrast with the relatively low temperatures (800–900°C) calculated from feldspar and apatite geothermometers for the Uruguayan rhyolites (Kirstein et al., 2000). True ignimbritic textures have been recognized only occasionally in the rhyolitic outcrops of the Paraná–Etendeka province (Milner et al., 1992; Garland et al., 1995). This, together with the large lateral extent of individual eruptive units, has led to their interpretation as rheoignimbrites, within which the high eruption temperatures caused intense rewelding of the pyroclastic fragments, largely destroying any ignimbritic textures (Milner et al., 1992, 1995). The Uruguay felsic volcanics are therefore unusual with respect to the Paraná rhyolites in that many of the higher-silica type 2 rocks have quartz phenocrysts, classic ignimbritic textures and lower magmatic temperatures. The thickness of the eruptive units of the Uruguay felsic volcanics (<5 m) is typically much less than that of the Paraná–Etendeka rhyolites (20–480 m thick).
|
GENERAL GEOCHEMISTRY |
|---|
|
TOPABSTRACTINTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
Major and trace element abundances were determined by X-ray fluorescence and the analyses are presented in Table 2 [see Ramsey et al. (1995)
for details of the analytical techniques]. Selected samples were analysed for additional trace elements by instrumental neutron activation analysis (INAA) and for Sr and Nd isotopes by thermal ionization mass spectrometry (TIMS) (Table 3). All the analyses were performed at the Open University, apart from those for O isotopes, which were performed at the University of Cape Town, and the analytical details and error analyses are given in the table footnotes. Loss on ignition (LOI) values were determined on separate powders and were used to screen for more altered mafic samples: in particular, those with an LOI >3% were discarded and are not discussed further. The felsic volcanics have been affected by post-emplacement processes, as is evident from the exceptionally high SiO2 (>78 wt %) contents of some samples. This is considered to be the result of silicification, and movement of the alkali elements (K and Na) is suspected from the devitrification textures noted in thin sections. Devitrification and low-temperature hydration of glassy rhyolites are associated with gains in K and losses in Na (Lipman, 1965). Thus, although the total alkali–silica (TAS) diagram (Fig. 3b) may be used to classify the main sample array, individual analyses of alkali elements should be treated with caution.
|
|
Most of the mafic lavas fall in the basalt, basaltic andesite and andesite fields on the TAS classification diagram, with just a few classifying as basaltic trachy-andesites. The subdivision of volcanic rocks into alkaline and subalkaline series using the boundary lines of either Kuno (1966) or Irvine & Barager (1971) places the majority of the Uruguay basic rocks in the subalkaline–tholeiitic field, although a few plot clearly in the alkaline series (see dashed line in Fig. 3b). The felsic volcanics plot as trachytes, trachydacites and rhyolites in Fig. 3b. It should be noted that although the felsic rocks associated with the main Paraná lava field are loosely termed ‘rhyolites’ in the literature, in fact they are virtually all dacites according to the TAS diagram classification and therefore compositionally distinct from most of the Uruguayan felsic rocks (see Fig. 3b). Further, the bimodal nature of the Uruguay lavas is best seen from their total alkali contents in that most of the basalts and andesites with <60 wt % SiO2 have (K2O + Na2O) <6%, whereas the trachydacites and rhyolites with >60 wt % SiO2 have >7 wt % (K2O + Na2O). This is in marked contrast to the main Paraná–Etendeka province, where the lavas are bimodal in silica abundances with a gap at 60–64 wt % SiO2, but with an overlap in total alkalis (Fig. 3b: Piccirillo et al., 1988; Peate et al., 1992). Two important implications are therefore that the petrogenetic relations between the mafic and felsic rocks, and the origins of the latter, may be different in Uruguay.
|
MAFIC LAVA GEOCHEMISTRY |
|---|
|
TOPABSTRACTINTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
All the basaltic samples analysed are relatively evolved with <7 wt % MgO, and <150 ppm Ni and Cr. On major and trace element variation diagrams (Fig. 4a–e), the mafic lavas define two compositionally distinct groups, termed the Treinte Y Trés and the Santa Lucía magma types. There is no apparent geographic significance in the distribution of the two magma types and they are not associated in the field. Only five samples have been obtained from what is inferred to be the volumetrically minor Santa Lucía magma type, and these were collected from Minas in the SW to Lascano in the north (Fig. 1b). The Treinte Y Trés magma type is widespread throughout southern Uruguay, and it is named after the exclusively basaltic region of Treinte Y Trés where all the lavas are of this magma type.
|
Relative to the Treinte Y Trés magma type, which comprises most of the lavas sampled, the Santa Lucía magma type tends towards lower SiO2 contents, but with higher Fe2O3 and P2O5, and similar ranges in MgO, Al2O3, CaO and K2O. In terms of the more immobile high field strength elements (HFSE; Zr, Y, Ti and Nb), the Santa Lucía basalts have higher Nb (>11 ppm) and TiO2 (>1·5 wt %) at similar MgO values (5–7 wt %), and in particular lower Zr/Nb than the Treinte Y Trés rocks (Fig. 4a, c, d). The differences between the two magma types are most striking on a mantle-normalized incompatible element diagram (Fig. 5). The Santa Lucía basalts have relatively smooth incompatible element signatures, broadly similar to those of asthenosphere-derived ocean-island basalts (OIB), whereas the Treinte Y Trés lavas have distinctive negative anomalies at Nb, P and Ti. These differences are potentially related to differences in their source composition(s), to crustal contamination, or possibly to differences in the residual phases perhaps linked to the degrees of melting.
|
Both Uruguay basalt magma types are enriched in light rare earth elements (LREE) relative to the heavy rare earth elements (HREE), and the Treinte Y Trés rocks have slightly lower LREE abundances than the Santa Lucía basalts. Negative Eu anomalies develop with increasing SiO2 in the Treinte Y Trés lavas and are associated with increasing plagioclase fractionation. The two magma types also have different Sr and Nd isotopic ratios (Fig. 6), with the Treinte Y Trés rocks having a much wider range in Ndi from -7·1 to -10·6. 87Sr/86Sri in the Treinte Y Trés rocks increases from 0·7089 to 0·7200 as SiO2 increases from 52 to 60 wt % (Fig. 7), whereas 87Sr/86Sri in the Santa Lucía basalts is 0·7046–0·7085 at lower SiO2 contents (48–49 wt %). There are a number of clear compositional affinities between the Treinte Y Trés magma type and the low-Ti/Y Gramado magma type of the Paraná (Peate et al., 1992), although the former tends to have higher 87Sr/86Sri and lower Ndi. The rare Santa Lucía basalts, however, have Sr and Nd isotope ratios that overlap those of the high-Ti rocks of the northern Paraná (Fig. 6).
|
Petrogenesis of the Treinte Y Trés magma type
The Treinte Y Trés lavas (TYT) do not represent primary mantle melts, as is shown by their low modal percentages of olivine, their low magnesian numbers (mg-number <54) and low Ni (<139 ppm) contents (Fig. 4a, b, e). Thus these magmas have undergone significant amounts of fractional crystallization, perhaps accompanied by crustal contamination, and the gravity anomaly illustrated in Fig. 1b may reflect a relatively shallow-level magma chamber. Major element trends indicate that fractionation of an essentially basaltic mineral assemblage (olivine, clinopyroxene, plagioclase) had a major control on the evolution of the system (Fig. 4a, b). Least-squares analyses, using one of the more primitive TYT lavas (93L27) as the parental magma and observed mineral compositions from Table 1, demonstrate that the TYT data can be modelled by up to 40% fractionation of olivine–plagioclase–clinopyroxene–magnetite in the proportions of 10:61:21:8 [note the high proportion of magnetite; see Cox (1980)]. However, the large variation in Sr–Nd isotope ratios, and in particular the increase in 87Sr/86Sri with increasing SiO2, Th/Nb and Rb/Ba, and decreasing MgO and Ti/Y (Figs 7 and 8b), is clear evidence for significant open-system differentiation. The range in 87Sr/86Sri from 0·7089 to 0·7200 and the associated increase in SiO2 from 52 to 60 wt % is readily modelled by assimilation and fractional crystallization (AFC; De Paolo, 1981). Simple mixing with a high-silica, high-Th/Nb, -Rb/Ba endmember such as the low-Ti Palmas rhyolites of the Paraná can be ruled out, as TiO2 increases slightly with decreasing MgO in the Treinte Y Trés samples, and yet the Palmas rhyolites have lower TiO2 contents (average TiO2 content 0·71 wt %) (Fig. 4a).
|
Th/Nb ratios of the Treinte Y Trés lavas are far removed from those of mid-ocean ridge basalt (MORB) (0·04–0·07) and OIB (0·08) (Sun & McDonough, 1989
) (Fig. 8a), and thus they might reflect contributions from lithospheric material in the continental crust or the uppermost mantle. On a plot of Th/Nb vs Th (Fig. 8a) fractional crystallization results in straight lines parallel to the x-axis, as illustrated by line (i). Yet, the Treinte Y Trés data define a positive array above such a trend and are best modelled by either bulk mixing or AFC [curves (ii) and (iii), Fig. 8a]. In the modelling of the AFC curves, the rate of assimilation to fractional crystallization (r) was set to 0·25. The data trend towards high Th/Nb (and to high Rb/Ba 
0·32) and they are therefore not simply modelled by bulk addition of upper- or mid-crustal compositions (Th/Nb 0·4 and 0·7; Rb/Ba 0·04 and 0·16, respectively). The most primitive Treinte Y Trés sample (93L27) lies at the low-Th/Nb end of the array, and the preferred contaminant used in the construction of curves (ii) and (iii) (Fig. 8a) is a leucogranite from the Pan-African of southern Brazil (RSM 54, Cascata leucogranite; May, 1990). This highlights how the contaminant needs to be a crustal melt, with fractionated Th/Nb and Rb/Ba ratios, rather than a sample of average bulk continental crust.
Th/Nb increases systematically with increasing 87Sr/86Sri (Fig. 8b), reaffirming that the within-suite variations in Th/Nb reflect open-system differentiation. However, the question of a contribution from distinct sources within the lithospheric mantle remains. The mafic potassic and ultrapotassic rocks of the Paraná province are thought to be small-degree melts of the sub-continental lithospheric mantle (Gibson et al., 1995). Thus, they provide a useful endmember for small-volume lithosphere-derived melts, which might have interacted with asthenosphere-derived magmas. Importantly, Fig. 8b shows that the Treinte Y Trés lavas cannot be generated by simple mixing of MORB-type magmas (Th/Nb 0·04–0·07) with such small-degree melts of the lithospheric mantle. The mafic potassic rocks of the Paraná are an unsuitable lithospheric endmember because their Th/Nb (<0·28) and Ti/Y (>400) ratios are higher than those of MORB (Fig. 8c), whereas the Treinte Y Trés lavas have higher Th/Nb but lower Ti/Y (184–260) than either MORB or OIB (272 and 593, respectively; Sun & McDonough, 1989) (Fig. 8c). The least evolved TYT lava also has lower Th/Nb than the least evolved mafic potassic melt. Peate (1997) pointed out that the low-Ti Gramado basalts of the southern Paraná still had elevated Sr isotope ratios when the data array was extrapolated back to Th/Nb ratios similar to those in oceanic basalts (i.e. 0·1). Similarly, the data array in Fig. 8b indicates that the Treinte Y Trés lavas have 87Sr/86Sri 0·705 at Th/Nb <0·16, and that they differ from the Santa Lucía basalts, which trend towards lower Th/Nb and 87Sr/86Sri.
It is clear from Figs 5–7 that the Treinte Y Trés lavas are very similar, in terms of major and trace elements and Sr–Nd isotopes, to the low-Ti/Y Gramado magma type of the Paraná (Peate & Hawkesworth, 1996). The Gramado basalts dominate the southern half of the Paraná province, including its extension into NW Uruguay (the Arapey Formation), and equivalent basalts form most of the Etendeka lava field in Namibia (e.g. Erlank et al., 1984; Bossi & Navarro, 1988; Peate et al., 1992; Turner et al., 1999b). There is widespread agreement over the importance of open-system behaviour in the petrogenesis of the low-Ti Gramado magma type given the systematic variation in incompatible element and isotopic ratios (Sr, Nd, O) (e.g. Mantovani et al., 1985; Piccirillo et al., 1988; Harris et al., 1989; Mantovani & Hawkesworth, 1990; Peate & Hawkesworth, 1996). However, the source of these low-Ti basalts remains controversial given their distinctive isotope and trace element ratios, and the close association of this magmatic province with the Tristan da Cunha plume.
Arndt et al. (1993) invoked extensive melting in the spinel stability field beneath thinned lithosphere to generate the low Ti/Y and Tb/Yb ratios that are such a feature of the low-Ti/Y Paraná basalts. Melting beneath thick lithosphere is more likely to take place in the presence of residual garnet leading to relatively incompatible-element-enriched basalts with high Ti/Y and Tb/Yb ratios (Arndt et al., 1993). However, the least contaminated Gramado rocks have other isotope and trace element ratios (low Nb/La <0·8, low Ndi <0) that are different from those commonly observed in MORB and OIB. Moreover, aspects of their major element compositions also suggest that their mantle source regions were more depleted in major elements than are the source regions of typical oceanic basalts (Hergt et al., 1991; Turner & Hawkesworth, 1995). Thus, Hergt et al. (1991) and Turner & Hawkesworth (1995) concluded that the parental Gramado magmas were derived from an Fe-depleted source, presumably as a result of previous melt extraction events, argued to be in the continental lithospheric mantle. At similar wt % MgO, the FeO contents of the Treinte Y Trés lavas tend to be even lower than those of the Gramado magma type. For example, at 6 wt % MgO, the Treinte Y Trés lavas have <10·5 wt % FeO and the Gramado lavas have an average FeO = 10·9 wt %, suggesting that both magma types were derived from similarly depleted source regions.
An alternative approach is to evaluate the conditions under which partial melting might have occurred in the lithosphere and/or in the underlying asthenosphere (McKenzie & Bickle, 1988; Gallagher & Hawkesworth, 1994; Hawkesworth et al., 2000). The results indicate that melting beneath thick lithosphere will not occur unless there are exceptional temperatures in the upper mantle, and that the lithosphere can melt by a combination of conductive heating and decompression in response to extension.
In summary, the preferred model for the generation and evolution of the Treinte Y Trés magma type is essentially that proposed for the Gramado magma type by Peate & Hawkesworth (1996). The parental magmas appear to have been very largely derived from within the lithospheric mantle, and the within-suite variations were dominated by extensive open-system fractional crystallization [see also Peate (1997)]. In addition to the petrographic differences, such as the porphyritic nature of the Uruguay rocks, there are some minor compositional differences between the Treinte Y Trés and the Gramado magma types. Thus, the Treinte Y Trés lavas tend to have lower Ndi and Nb/La and higher Nb/Zr than most Gramado lavas, and they also plot at slightly higher 87Sr/86Sri on a SiO2 vs 87Sr/86Sri diagram. Such differences appear to be regionally significant, and may reflect minor differences in the lithospheric mantle source regions or in the crustal contaminants. Peate & Hawkesworth (1996) documented similar small regional compositional differences in southeastern Brazil within the Gramado magma type.
Petrogenesis of the Santa Lucía magma type
The Santa Lucía magma type contains some of the lowest-silica rocks (47–49 wt % SiO2) associated with the Paraná CFB province. Compared with the more voluminous Treinte Y Trés basalts, the Santa Lucía rocks have much higher Nb/La and Nb/Zr ratios of 0·8–1·05 and 0·08–0·18, respectively, and in general lower Th/Nb (0·08–0·15) and 87Sr/86Sri (Figs 4, 6 and 8). Thus, the Santa Lucía basalts plot as distinct fields in Figs 4, 6 and 8, and cannot be simply related to the Treinte Y Trés basalts by fractional crystallization and/or crustal assimilation processes [vector (i) in Fig. 8a, and vector AFC in Fig. 8d]. These vectors were constrained using the fractional crystallization and AFC modelling of the Treinte Y Trés magmas discussed in the previous section, where the contaminant was shown to be a crustal melt with Nb/La > 0·6 and Th/Nb > 0·7, rather than bulk crust.
The relatively smooth primitive mantle-normalized incompatible element signature of the Santa Lucía magma type (Fig. 5) is similar to those of OIB. This suggests that, like many OIB, the Santa Lucía magmas are partial melts of the sub-lithospheric mantle and that they have been relatively little modified by interaction with the continental lithosphere. However, their initial Sr and Nd isotope ratios are higher and lower respectively than most OIB, which is consistent with at least a contribution from old, trace element enriched material as found in the continental lithosphere. Because of the limited number of samples, it is difficult to identify significant trends within the Santa Lucía dataset, but there are apparent correlations between radiogenic isotopes and ratios of highly incompatible trace elements (e.g. Nd isotopes and Nb/La: Fig. 8d). This indicates that mixing processes were important in the evolution of the Santa Lucía magmas: one endmember has high Ndi and Nb/La (inferred to be the Tristan plume), and the other endmember has low Ndi and Nb/La (inferred to be from the continental lithosphere). The latter might be mafic potassic melts from the lithospheric mantle (Gibson et al., 1995), perhaps modified by subsequent crustal contamination. The range in 87Sr/86Sri, for example, of the Santa Lucía magmas is greater than that reported for the Paraná mafic potassic magmas (Gibson et al., 1995) (Fig. 8b) which might reflect contamination at shallow levels involving crustal material.
Figure 8d illustrates two models for mixing between an asthenosphere-derived melt, as from Tristan da Cunha (TDC-91, le Roex et al., 1990), and (i) a mafic potassic lithosphere-derived magma from Brazil (93SOB24; Gibson et al., 1995), and (ii) Brazilian basement, on a plot of Ndi vs Nb/La. The calculated curves demonstrate that simple mixing can produce the variation in Nb/La and Ndi seen in the Santa Lucía magmas, and that those magmas might reflect 25–40% bulk mixing between a Tristan component and either a potassic melt or bulk crust (basement). However, other trace element ratios (La/Yb, La/Sm) cannot be modelled using the potassic melt as an endmember, and the high implied contributions from a crustal endmember are in conflict with the relatively low SiO2 contents (47–49 wt %) of the Santa Lucía magmas. Thus, this magma type is more simply generated by mixing between plume-derived melts and crustally contaminated low-Ti magmas, such as the Treinte Y Trés basalts.
The relative ages of the Treinte Y Trés and Santa Lucía magma types are uncertain in the field and recent 40Ar–39Ar results indicate that their ages are analytically indistinguishable (Stewart et al., 1996; Kirstein, 1997). Peate & Hawkesworth (1996) invoked mixing between MORB and previously contaminated Gramado magmas to explain the compositional variations within the late-stage Esmeralda rocks of the Paraná in southern Brazil, because the asthenospheric component was characterized by a relatively depleted isotope and trace element signature (Fig. 8d). As seen in Fig. 8c, such a MORB-like component would not be suitable for the Santa Lucía magma type, irrespective of the preferred choice of the lithospheric endmember (crust or contaminated basalt).
More recently, Ewart et al. (1998) described some mildly alkaline to tholeiitic basalts dominated by olivine, clinopyroxene and plagioclase phenocryst assemblages, from the southern part of the Etendeka province in the Goboboseb Mountains near Messum Crater. These lavas have some of the compositional features of the Santa Lucía rocks and they were erupted very close to the present Santa Lucía outcrop before break-up. The Etendeka lavas have been divided into two distinct series on the basis of their Ti/Zr ratios into the LTZ.H (higher Ti/Zr, >80) and LTZ.L (lower Ti/Zr, <50) series. The LTZ.H basalts were previously termed the Tafelkop basalts by Milner & le Roex (1996), and this term is used here. Relative to the Tafelkop and LTZ.L basalts, the Santa Lucía basalts have intermediate Ti/Zr ratios of between 60 and 80, and initial Sr and Nd isotopic ratios of 0·7046–0·7085 and 0·51218–0·51234, respectively (Fig. 6). However, both magma series have similar ranges in Nb, Zr and LREE abundances.
The Tafelkop (LTZ.H) melts were interpreted by Ewart et al. (1998) as mixtures of a dominant plume component (Tristan) and material from either mafic lower crust and/or subcontinental lithospheric mantle (SCLM). Isotopic and trace element ratios were further used by Ewart et al. (1998) to suggest that the Tafelkop melts have geochemical characteristics intermediate between E-type MORB and OIB, albeit closer to OIB. Thus, the model for the Tafelkop lavas invoked by Ewart et al. (1998) is similar to that presented here for the Santa Lucía basalts from southern Uruguay. The Santa Lucía rocks show no evidence of a MORB component, but they can be modelled by simple two-component mixing between a lithospheric component, as in the Treinte Y Trés magmas, and OIB similar to that erupted on Tristan da Cunha. The Santa Lucía magmas have lower Tb/Yb (0·27–0·37) than the Tafelkop rocks (Tb/Yb = 0·43–0·55) perhaps reflecting less of a role for residual garnet, and hence more melting in the spinel stability field for the plume-derived endmember.
The striking difference between the Esmeralda and both the Santa Lucía and Tafelkop magma types is that the former appears to have involved a MORB-like component, whereas the Santa Lucía and Tafelkop rocks contain a contribution from the Tristan plume. The Santa Lucía and Tafelkop basalts were almost adjacent before rifting, whereas the Esmeralda lavas were erupted several hundred kilometres north in southern Brazil.
|
GEOCHEMISTRY OF THE FELSIC VOLCANICS |
|---|
|
TOPABSTRACTINTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
Representative major and trace element results for the Uruguayan felsic rocks are presented in Table 2. On a major element variation diagram of TiO2 vs SiO2 (Fig. 9a) it is clear that the felsic volcanic rocks from Uruguay are much more variable than either the high- or low-Ti rhyolites from the Paraná. The Uruguayan rocks have been subdivided into two geochemical series, the Lavalleja Series and the Aigüa Series, on the basis of major and trace element data. As discussed earlier, the higher-silica rocks of both series are distinguished by the presence of modal quartz and increased alkali feldspar (Fig. 3a). The division into two series is based primarily on differences between minor and trace elements such as Nb, Zr, Ti and Y. The rocks of the Lavalleja Series tend to have higher Ti/Zr (3–19) (Fig. 9d) and lower Nb/Y (<1), and in general they form a tight Nb–Zr array compared with the Aigüa Series, which has lower Ti/Zr (<6·65, Fig. 3a) and Nb/Y up to three. Each series can be subdivided into a trachydacite–dacite suite and a rhyolitic suite primarily on the basis of SiO2, but also on isotopes and trace elements. This subdivision is linked to extreme within-suite fractionation of a feldspar-dominated assemblage for the Aigüa Series (Fig. 9c, e). Although the Lavalleja rocks appear to represent a distinct series, their within-series isotope and trace element variations suggest that the trachydacites–dacites and the rhyolites are not simply related but rather reflect contributions from different source rocks (Figs 6, 7 and 9). This is discussed below after a further comparison of the Aigüa and the Lavalleja Series.
Rocks of the Lavalleja Series are also distinguished by having higher CaO, TiO2, Sr, Ba and Eu, and lower Al2O3, K2O, Na2O, Zr, Nb, Rb, Ta and Th abundances than the Aigüa rocks. There is more overlap in other elements such as SiO2 and Y (Table 2). Overall, CaO, Na2O, Al2O3, P2O5, TiO2 and Fe2O3 tend to decrease with increasing SiO2 (Table 2, Fig. 9a), and such trends are broadly consistent with fractionation of an assemblage consisting of plagioclase, alkali feldspar, pyroxene and iron oxide. Some of the Uruguayan rocks have very high silica contents (>78 wt % SiO2), but these may reflect late-stage silicification, not least because the most evolved rhyolites previously described from, for example, Glass Mountain, California, contain 77·6 wt % SiO2 (Metz & Mahood, 1985). No felsic samples were omitted from the dataset because of their LOI values, although certain samples are likely to have been affected by alkali mobility as indicated by their high K2O/Na2O ratios (e.g. 93L28 and 93L80, Table 2).
High-silica rocks are often classified on the basis of their alkali index, and subdivided into peralkaline, peraluminous and metaluminous (Hildreth, 1981; Leat et al., 1986; Davies & MacDonald, 1987). Peraluminous and metaluminous rhyolites are common in continental margin and ancient orogenic belt settings, whereas peralkaline silicic rocks are more common in stable continental settings and they also occur in oceanic regions. Some of the Uruguayan felsic volcanics show evidence of alteration and alkali mobility, which would affect any classification based on alkali indices. In an attempt to overcome such problems in altered rocks, Leat et al. (1986) proposed a revised system based on HFSE abundances, in particular Nb and Zr. Low-Zr (<300 ppm) rocks were classed as subalkaline (peraluminous or metaluminous) and high-Zr (>350 ppm) were termed peralkaline. However, for the Uruguay rocks all but two are peraluminous and metaluminous in that they have [(K + Na)/Al] < 1, and yet the majority would be classified as peralkaline on the criterion of >350 ppm Zr (Leat et al., 1986). This discrepancy may reflect the effects of alteration on the alkali contents of the Uruguay rocks, or the problems of redefining major element terms on trace element criteria.
A log–log plot of Nb vs Zr shows the large variations in the Uruguayan felsic rocks (Fig. 9b). The Lavalleja trachydacites–dacites form a tight array with <350 ppm Zr and <75 ppm Nb, whereas the rhyolites have up to 710 ppm Zr and plot close to the dashed line indicating Zr/Nb = 10. The Aigüa Series rocks show much more variation in both Nb and Zr, and typically have Zr/Nb <10. A Zr/Nb ratio of 10 is considered to divide the samples into particular tectonic settings with Zr/Nb <10 being a feature of within-plate magmatism (Pearce et al., 1984; Leat et al., 1986). High concentrations of Zr (>700 ppm), as seen in the trachydacites–trachytes of the Aigüa Series, are characteristic of ‘peralkaline’ magmas, where the excess of alkalis over alumina increases the solubility of Zr and results in the formation of alkali-zircono-silicate complexes (Collins et al., 1982). This also implies that the relatively high alkali contents of the trachytic samples in Fig. 3b is a primary feature of the magmas and is not due to secondary alteration. Relatively low Zr contents at high SiO2 occur when alkali-zircono-silicate complexes are not formed and zircon crystallization occurs earlier in the sequence (Collins et al., 1982). The low Ti/Zr ratios of the Lavalleja rhyolites appear to reflect such early zircon crystallization (Fig. 9d), and zircons are observed petrographically.
The Nb abundances are very variable and within the Aigüa Series high Nb abundances occur in rocks with very different Zr contents (Fig. 9b). Nb = 32–330 ppm in the Aigüa Series, and 13–53 ppm with a mean of 28 ppm in the Lavalleja Series, consistent with a within-plate continental origin (Pearce et al., 1984). The plot of Ta/Th vs Ti/Zr (Fig. 9d) highlights the high Ta/Th ratios of the Aigüa relative to the Lavalleja Series, consistent with the low Zr/Nb of the former (Fig. 9b) and a within-plate affinity.
Sr and Nd isotope ratios have been measured on selected samples from the Lavalleja and Aigüa Series, and the initial values were calculated for an eruption age of 130 Ma (Figs 6 and 7). The isotope data clearly distinguish between the rhyolites of the two series, with the Lavalleja having 87Sr/86Sri = 0·7158–0·7248 and Ndi = -11·3 to -4·7, and the Aigüa with 87Sr/86Sri = 0·7056–0·7073 and Ndi = -17·5 to -15·8. The trachydacite–dacite samples from the two series have overlapping initial 87Sr/86Sr ratios (Lavalleja 0·7085–0·7089; Aigüa 0·7054–0·7084), but those from the Aigüa Series tend to have lower Ndi (Aigüa -15·8 to -11·2; Lavalleja -12·1 to -10·1) (Fig. 6). Overall, the Aigüa and Lavalleja Series rocks plot in different fields in Fig. 6, with the former being characterized by relatively low Sr and Nd isotope ratios. Within the Lavalleja Series the trachydacites–dacites and the rhyolites also have significantly different isotope ratios (Figs 6 and 7), suggesting that they are not simply interrelated. In general, the shift to lower Sr and Nd isotope ratios from the Lavalleja rhyolites to the Aigüa rocks is accompanied by an increase in within-plate character as evidenced by higher Ta/Th and lower Zr/Nb (Fig. 9). The samples of the Valle Chico syenite overlap those of the Aigüa trachytes–trachydacites in Ndi and 87Sr/86Sri, suggesting a close petrogenetic link between the two rock types (Figs 6 and 7).
Oxygen isotope ratios have been measured on quartz phenocrysts from rhyolites of the Lavalleja and Aigüa Series using conventional fluorination techniques (see Harris & Erlank, 1992; Harris, 1995). The results are presented in Table 4 and Fig. 10 along with representative analyses of both Paraná and Etendeka silicic volcanics from Harris et al. (1990). There is 1
difference in 
18O values between the two Uruguayan series, and the 18O (magma) values of the Lavalleja rhyolites plot within the field for the low-Ti Palmas rhyolites. The Aigüa rhyolites are intermediate between the two Paraná rhyolite fields in Fig. 10 but similar to the data from the silica-oversaturated Damaraland complexes of Namibia.
|
|
The Paraná and Etendeka basaltic rocks have a range in 18O from +6 to +8 (Harris et al., 1989; Iacumin et al., 1991). The latter values overlap with those of the high-Ti Chapecó rhyolites but they are lower than the 18O values for the low-Ti Palmas rhyolites (Fig. 10). The Lavalleja rhyolites plot within the field for the low-Ti Palmas rhyolites in the diagram of 18O vs Zr (Fig. 10). The single basalt sample from Uruguay analysed to assess 18O in potentially underplated basalts has a magma 18O value of +7·2. This is much lower than those measured for the rhyolites, suggesting that underplated basalts are not a suitable source rock for the rhyolites, as also indicated by their Nd and Sr isotopes (Fig. 6). One basement rock from the area (gneiss sample 93L20) has 18O +8, slightly more similar to the rhyolite 18O values (Table 4).
Petrogenesis of Uruguay felsic volcanic rocks
There are two main points that need to be addressed regarding the petrogenesis of the Uruguay high-silica rocks: (1) the nature of the intra-suite variations and the conditions of closed- or open-system differentiation from the least-evolved samples; (2) the generation of the least-evolved magmas with regard to potential links with the basalts of the Puerto Gómez Fm. and partial melting of particular crustal rocks.
Lavalleja Series
The major and trace element variations in the Lavalleja Series are qualitatively consistent with fractional crystallization of the observed phases (Fig. 9, Table 2). In particular, decreasing Sr and Eu abundances and Rb/Sr ratios up to seven indicate extensive feldspar fractionation. In Fig. 11a representative samples from the trachydacites and the rhyolites of the Lavalleja Series are normalized to MORB, as are average Chapecó and Palmas rhyolite samples (Garland et al., 1995) and the most evolved Treinte Y Trés lava (93L125). The trachydacites have higher LILE and REE abundances than the rhyolites even though the trachydacites represent less fractionated compositions. It is therefore difficult to explain the sequence in terms of a single magmatic lineage, and the different Sr and Nd isotope ratios of the Lavalleja trachydacites and rhyolites highlight their different origins. The Lavalleja trachydacites also have higher Ba, Sr, HFSE and REE than the most evolved Treinte Y Trés lava (93L125, 60 wt % SiO2), but simple assimilation and fractional crystallization models are unable to generate the measured concentrations of the Lavalleja trachydacites from the Treinte Y Trés lava. Rather, the trachydacites share certain features with the Aigüa trachydacites, notably their higher Ta/Th ratios together with low 87Sr/86Sri (<0·7089) and negative Ndi (-10) (Fig. 6), perhaps indicative of source regions in the lower crust.
|
The initial 87Sr/86Sr ratios of the Lavalleja rhyolites increase from 0·7158 to 0·7248 with increasing SiO2 (Fig. 7), and their 87Sr/86Sri and
Ndi values overlap with and extend the Nd–Sr isotope array of the Treinte Y Trés lavas (Fig. 6). Such links between the Treinte Y Trés basalts and the Lavalleja rhyolites are similar to those between the low-Ti basalts and rhyolites of the southern Paraná, modelled by open-system fractional crystallization by Garland et al. (1995). Using the same crustal endmember as that invoked for contamination of the Treinte Y Trés lavas, i.e. a Brazilian leucogranite with fractionated Th/Nb and Rb/Ba ratios and elevated 87Sr/86Sr (RSM 54; May, 1990), the within-suite variations for the Lavalleja rhyolites suite can be modelled by 45% fractionation and a rate of assimilation to crystallization (r) of 0·4. This rate of assimilation is higher than that required to explain the variation within the Treinte Y Trés lavas (r = 0·25) reflecting the greater upper-crustal influence in the petrogenesis of the Lavalleja rhyolites. The fractionating assemblage is of clinopyroxene, plagioclase, alkali feldspar, titano-magnetite, apatite and zircon (5:49:30:15:0·5:0·1). In plots of 87Sr/86Sri vs SiO2 (Fig. 7), the Lavalleja rhyolites are displaced to high SiO2 contents, relative to the trend of increasing 87Sr/86Sri and SiO2 in the Treinte Y Trés basalts. A similar displacement is observed between the Gramado basalts and the low-Ti Palmas rhyolites in the Paraná, consistent with the more rapid increase in SiO2 in the residual liquid at the onset of magnetite fractionation (Garland et al., 1995).
The Lavalleja rhyolites have broadly similar major and trace element compositions to the low-Ti rhyolites of the Paraná (Figs 9 and 11). There is also considerable overlap in their isotope ratios, with 87Sr/86Sri = 0·7158–0·7248 and Ndi = -4·7 to -11·3 in the Lavalleja, and 0·7137–0·7274 and -6 to -8·8, respectively, in the Paraná low-Ti rhyolites (Figs 6 and 7). In detail, most of the Lavalleja rhyolites have lower Ndi values than the low-Ti rhyolites of the southern Paraná (Fig. 7), apart from three samples from the same area (33·8°S, 54·3°W), which have higher Ndi. Such differences presumably reflect regional differences in the Nd isotope ratios of the crustal contaminants, consistent with the large variations in Nd130 in the Brazilian mobile belts and cratonic areas (Nd130 = -30 to -10 in the basement rocks, Mantovani et al., 1987). At present, no such data are available from the basement rocks in Uruguay.
Aigüa Series
The Aigüa felsic volcanics have a similar range in 87Sr/86Sri to the Santa Lucía basalts, but much lower Ndi (-17·5 to -11·2 compared with -5·4 to -2·3 in the basalts) (Figs 6 and 7). It is concluded that the Aigüa Series felsic volcanics were not generated by fractional crystallization from the associated basaltic magmas or by remelting of underplated Mesozoic basalts of either the Santa Lucía or the Treinte Y Trés magma types.
As noted previously, the Aigüa Series rocks have very different major and trace element and REE concentrations from either the Lavalleja Series or the Paraná rhyolites (Figs 9 and 11). The Aigüa rocks have more fractionated minor and trace element compositions, with higher incompatible element (Rb, Nb, La, Ce, Nd, Sm, Zr, Y and Yb), and lower compatible element (Ba, Sr, P and Ti) abundances (Fig. 11b). Eu/Eu* is also much lower in the Aigüa Series than in the Lavalleja Series (Fig. 9e), and Rb/Sr and Rb/Ba are much higher (Fig. 9c), consistent with significant fractionation of plagioclase and alkali feldspar, in addition to apatite and magnetite. The Aigüa Series has lower and more restricted 87Sr/86Sri values than the Lavalleja Series rhyolites (Fig. 6). There is some overlap between 87Sr/86Sri in the Aigüa Series trachydacites, the Lavalleja trachydacites and the high-Ti Chapecó rhyolites, but the rocks of the Aigüa Series have much lower Ndi (-11·2 to -15·8) than those of, for example, the high-Ti Paraná rhyolites (-3·6 to -6). Thus, the source regions for the Aigüa suite would appear to have been different from those of the other high-silica rocks (Fig. 6).
Within the Aigüa Series there is no clear change of 87Sr/86Sri with increasing SiO2, as observed in the Lavalleja Series (Fig. 7), despite their often extreme Rb/Sr and Rb/Ba ratios (Fig. 9c). In detail, there is no systematic increase in, for example, Nb with increasing Zr (Fig. 9b), but variations in Eu/Eu*, Sr, Ba, P and Ti with increasing Th (as in Fig. 9e) and Rb are broadly consistent with extreme fractional crystallization (70%) of an assemblage of 45% plagioclase, 40% alkali feldspar, 15% magnetite and 0·5% apatite. This is twice the amount of fractionation estimated for the chemical variations within the Lavalleja Series. However, the latter have been modelled as the products of open-system fractional crystallization of basaltic magma, and so overall they reflect >60% fractional crystallization from the initial basaltic magma, plus crustal assimilation.
The Aigüa trachytes–trachydacites overlap the composition of the Valle Chico syenite in terms of Ta/Th, Th/Nb, 87Sr/86Sri and Ndi values (Figs 6, 9 and 11b, Table 2). However, the syenite is much less fractionated in that it has lower incompatible element and higher compatible element abundances than the rocks of the Aigüa Series (Fig. 11b). Thus, they would appear to have had broadly similar source regions, but the Aigüa Series experienced much greater degrees of fractional crystallization.
The generation of the least evolved felsic magmas
The least evolved felsic samples in the Aigüa and the Lavalleja Series are relatively low-alumina transitional trachytes–trachydacites (Figs 3 and 9). The Lavalleja rhyolites have been modelled as the products of upper-crustal assimilation and fractional crystallization from the associated basaltic Treinte Y Trés magmas, but the less evolved rocks from Aigüa and the Lavalleja Series have distinctive isotope and trace element ratios and are inferred to reflect partial melting at deeper levels in the crust. The Uruguay felsic volcanics also have relatively low Al contents (9–17 wt % Al2O3), consistent with partial melting of an igneous, rather than a metasedimentary protolith (Chappell, 1999). The initial Nd- and Sr-isotope ratios of the trachytes–trachydacites (Fig. 6) indicate that they were derived from source rocks with low time-integrated Sm/Nd and Rb/Sr ratios. Other studies have suggested that broadly similar magmas may be generated by remelting granulite residua (Collins et al., 1982), tonalites (Creaser, 1989) and metaluminous granites (Davies & MacDonald, 1987).
Melting models were used to constrain the generation of the least evolved trachytes–trachydacites of both series. The source was meta-igneous, with low Sm/Nd, Th/Nb, Th/Ta and Rb/Sr ratios, and no Eu anomaly. The time-integrated Rb/Sr ratio of the source, as calculated from the average model Nd age of the Aigüa rocks (1·5 Ga), is 0·06. The lower crust is characterized by such compositions, and so the lower-crustal composition of Rudnick & Fountain (1995) was chosen as the source for the melt calculations (Fig. 12).
|
Partial melting is thought to have occurred in the middle to the lower crust, given the present-day gravity anomaly, under fluid absent conditions (anhydrous mineralogy) and at temperatures between 850 and 950°C (Kirstein et al., 2000). These temperatures were determined by feldspar and apatite thermometry for the southern Uruguay felsic volcanics (Kirstein et al., 2000), and are well within the experimental fields for partial melting of amphibolites and tonalites (Rushmer, 1991; Skjerlie et al., 1993). Using the major and trace element compositions of the lower crust from Rudnick & Fountain (1995), phase proportions were estimated from the normative mineralogies of possible source rocks. Bulk D values were calculated assuming 50% plagioclase + 25% orthopyroxene + 25% clinopyroxene in the source residue with published mineral melt partition coefficients for basaltic andesite magma compositions (Pearce & Norry, 1979; Mahood & Hildreth, 1983) (Table 5). Modelling was undertaken to determine the approximate minor and trace element composition of the source (Co). It was assumed that the least evolved sample in each of the series (CL) was produced by 20% partial melting (Table 5), which is consistent with the Rb and Ba contents of the least evolved rocks, and may be what is required for a granitic melt to separate from its source, given realistic viscosities (McKenzie, 1985). Modal melting was not assumed, and mineral proportions entering the melt were constrained by the petrography, extraction calculations and the phase diagram of En–Di–Qz–Ab (Yoder, 1976).
|
Figure 12 shows MORB-normalized incompatible element patterns for the Aigüa and Lavalleja trachydacites–trachytes, the Valle Chico syenite, mafic lower crust (Rudnick & Fountain, 1995) and 5–20% partial melts of the lower crust (Table 5). Broad similarities are evident between the Valle Chico syenite and both suites of trachydacites and trachytes; however, the Aigüa trachydacites–trachytes have higher incompatible element abundances, and more extreme anomalies in Nb, Sr, P, Zr and Ti. The data also show similar patterns of incompatible element abundances between the calculated lower-crustal melts and the Lavalleja and Valle Chico felsic volcanics. The Aigüa trachydacites have higher incompatible element abundances and more pronounced Sr and Ti anomalies, consistent with significant fractional crystallization. Some elements, such as Zr, Nb and some of the REE, have high and variable abundances (Fig. 9), which may be due to volatile complexing (Hildreth, 1981). If a large volatile (CO2, F) component was dissolved in the magma this would suppress the crystallization of plagioclase, resulting in an alkali-enriched magma (Collins et al., 1982). Overall it appears that the least evolved Aigüa and Lavalleja trachydacites–trachytes, and the Valle Chico syenite, are likely to have been generated by up to 20% melting of Proterozoic mafic lower crust at temperatures higher than 850°C. In detail, there were small differences in, for example, Nd and Sr isotope ratios of the sources of the Aigüa and Lavalleja trachydacites–trachytes. Similar models have previously been proposed for incompatible element enriched magmas from southeastern Australia (Collins et al., 1982) and Kenya (Davies & MacDonald, 1987).
Comparison of the Uruguay rhyolites with those of the Paraná–Etendeka
The felsic volcanics from southern Uruguay show a greater compositional diversity, and over a much smaller area, than those of the Paraná. The Uruguayan felsic rocks are different both petrographically and chemically from those of the Paraná. In particular, the Uruguayan rhyolites have lower calculated eruption temperatures of 850–950°C with well-developed ignimbritic textures (Kirstein et al., 2000), compared with 950–1150°C for the Paraná (Bellieni et al., 1986). Ignimbritic textures are extremely rare in the Paraná, and this has been attributed to rewelding as a result of the high eruption temperatures (Garland et al., 1995). The Uruguayan rhyolites are also considerably more porphyritic (>40%) with phenocrysts that include alkali feldspar and quartz in the higher-silica rocks, and accessory phases that include zircon and apatite. In contrast, the low-Ti Palmas rhyolites of the Paraná contain <5% phenocrysts, and the high-Ti Chapecó rhyolites have <25% phenocrysts of plagioclase, pyroxene and titanomagnetite (Bellieni et al., 1986).
Chemically, the Uruguayan felsic volcanic rocks extend to higher SiO2, K2O and Na2O, and lower Fe2O3 and TiO2. Trace element and REE concentrations vary widely but overall there are trends to higher Rb, Zr and REE concentrations, lower Sr, and higher Rb/Sr and Rb/Ba. Initial 87Sr/86Sr in the Lavalleja Series overlap with those of the low-Ti Palmas rhyolites of the Paraná, and as do those of the Aigüa Series with the high-Ti Chapecó rocks of the Paraná (Fig. 6). However, the initial 143Nd/144Nd ratios of the Aigüa Series are very much lower than those of any of the Paraná rhyolites. The high-Ti Paraná rhyolites have 18O values that are lower (6·3) than those of both the low-Ti Paraná rhyolites (10) and the Uruguayan rhyolites (9·1–10·8). There is no evidence that the Uruguay felsic volcanics were generated by partial melting of underplated Mesozoic basalts (of either the Treinte Y Trés or Santa Lucía magma types) as has been proposed for the genesis of the high-Ti Paraná rhyolites (Piccirillo et al., 1987; Garland et al., 1995).
The Lavalleja Series rhyolites have many similarities with the low-Ti Palmas rhyolites of the Paraná, and both have upper-crustal isotopic signatures (high 87Sr/86Sri and 18O). They have been modelled as the products of open-system fractional crystallization of the associated Treinte Y Trés and Gramado magmas, respectively [this work, and Garland et al. (1995)]. In contrast, the trachydacites and trachytes of both the Aigüa and the Lavalleja Series have a number of similarities with the Valle Chico intrusive syenite. They all have relatively low initial Sr and Nd isotope ratios not observed in the Paraná rocks (Fig. 6), and they have been modelled by partial melting of igneous source rocks in the lower crust. The Aigüa rhyolites then represent up to 70% fractional crystallization of alkali and plagioclase feldspar dominated assemblages.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
The Early Cretaceous volcanic rocks of southern Uruguay are part of the Paraná–Etendeka igneous province in which magmatism was associated with continental break-up and the opening of the South Atlantic Ocean. It has been argued elsewhere that most of the magmas were derived from within the mantle lithosphere, and that there were close links between regional extensional tectonics and melt generation (e.g. Hawkesworth et al., 2000
). The Uruguay rocks are bimodal, particularly in total alkalis, and the more voluminous magma types have close similarities with the Paraná rocks in southern Brazil. In the more mafic rocks SiO2 = 48–60 wt %, and two magma types have been identified primarily on the basis of their different HFSE abundances and Nd and Sr isotopes. The magma types are named Santa Lucía and Treinte Y Trés, and the latter is broadly similar to the low-Ti Gramado lavas of the Paraná. There is much evidence of open-system differentiation and assimilation of fractionated upper-crustal material, with 87Sr/86Sri increasing from 0·7089 to 0·7200 as SiO2 increases from 52 to 60 wt %. The Treinte Y Trés magmas were therefore modelled as up to 40% fractional crystallization, and assimilation with r = 0·25, of parental magmas that may have had 87Sr/86Sr of 0·705–0·706 derived primarily from the mantle lithosphere. Compared with the Gramado rocks of the southern Paraná, the Treinte Y Trés basalts have slightly higher TiO2 contents but 10–20% lower La/Yb and Dy/Yb ratios at 6·5% MgO. Thus, the evidence for possible regional variations in degrees of partial melting is ambiguous, but the REE data would suggest that the Treinte Y Trés basalts may reflect slightly higher degrees of melting than the Gramado rocks.
In contrast, the Santa Lucía basalts bear little resemblance to any previously described magma types in the Paraná province. Their initial Nd and Sr isotopes (0·51218–0·51234 and 0·7046–0·7085) overlap those of the high-Ti Paraná basalts but they tend to have higher Nb/Zr and Nb/La. Their trace element patterns are similar to those in plume-derived oceanic basalts, and consequently the Santa Lucía basalts have been modelled as mixtures between melts from within the Tristan plume and crustally contaminated low-Ti magmas, such as the Treinte Y Trés basalts. This is the first documented occurrence of basalts with a Tristan-like component within the main Cretaceous flood basalt province in South America and it highlights the involvement of the Tristan da Cunha hotspot in the generation of this part of the CFB province. Similar compositions have been recognized in the Etendeka Tafelkop basalts, which would have been erupted in a similar geographic location before rifting (Ewart et al., 1998). Such magma types involve an OIB component, whereas the Esmeralda lavas, which were erupted several hundred kilometres north in southern Brazil, involved a MORB-like component.
The bimodal nature of the Puerto Gómez and Arequita Formations has some similarities with bimodal CFB magmatism elsewhere, but the felsic rocks are unusual in the diversity of rocks preserved in a comparatively small province, and in the range of inferred magma source compositions. Exceptionally in the context of the Paraná–Etendeka province, flow features are common and the higher-silica rocks have variably sized, euhedral to anhedral quartz phenocrysts. The eruption temperatures for the Uruguayan rhyolites (850–950°C, Kirstein et al., 2000) are lower than those for the Paraná (950–1150°C, Bellieni et al., 1986). The felsic rocks are subdivided into two series, the Lavalleja and the Aigüa Series, on the basis of isotope and trace element ratios. Each series consists of both trachydacites–dacites and rhyolites, although the isotope and trace element variations within the Lavalleja Series suggest that these rock types are not simply related but rather reflect contributions from different source rocks. In general, the Lavalleja Series has higher Ti/Zr (3–19) and lower Nb/Y (<1), and higher Sr and Nd isotope ratios than the Aigüa Series. The isotope data most clearly distinguish the rhyolites of the two series, as the Lavalleja have 87Sr/86Sri = 0·7158–0·7248 and Ndi = -11·3 to -4·7, whereas in the Aigüa 87Sr/86Sri = 0·7056–0·7073 and Ndi = -17·5 to -15·8.
The general trend is from broadly upper-crustal isotope and trace element ratios, particularly in the more evolved rocks in the Lavalleja Series, to deeper-crustal signatures in the Aigüa rocks (e.g. Fig. 6). Both series include rocks of significantly higher silica values than those commonly observed in the Paraná, and the Aigüa Series extends to more fractionated compositions than the Lavalleja (e.g. higher Rb/Sr and Rb/Ba ratios and Th contents; Fig. 9). The Lavalleja rhyolites are similar to the low-Ti rhyolites of the Paraná–Etendeka, and they have been modelled by open-system differentiation from low-Ti Treinte Y Trés parental magmas [see also Garland et al. (1995)]. Lower-crustal melts similar to the Aigüa Series are not found in the Paraná, although the high-Ti rhyolites are attributed to partial melting of underplated basalts. The Aigüa Series contains more fractionated compositions than the Lavalleja, but the latter is thought to have evolved from more mafic initial magmas and so each series represents similar overall amounts of fractional crystallization (60–70%).
High- and low-Ti magmatic provinces are recognized in the mafic and rhyolite rocks of the Paraná–Etendeka (Erlank et al., 1984; Bellieni et al., 1986; Peate, 1997), and in ultrapotassic rocks in southern Brazil (Gibson et al., 1995). Within the basalts the rates of melt generation increase from the high- to the low-Ti magmas, with the latter having been erupted through broadly coast-parallel dykes associated with greater amounts of extension (Stewart et al., 1996; Hawkesworth et al., 2000). Within the rhyolites there is a shift from deep-crustal melts (of either underplated basalts or pre-existing crust) nearer the margins, towards more upper-crustal melting and the generation of rhyolites by open-system differentiation from basaltic magma nearer the plume centre. The Cretaceous rocks of Uruguay were generated marginal to the plume, and the felsic rocks are characterized by a diversity of rock types and inferred source regions, and lower magmatic temperatures.
|
ACKNOWLEDGEMENTS |
|---|
This research was undertaken whilst L.K. was in receipt of an Open University studentship. O-isotope analyses were funded by an FRD grant to C.H. We are grateful to the Uruguayan Geological Survey, particularly Fernando Tabó, J. M. Filippini and Hermes Rizeizo, for assistance during fieldwork in 1993. We are also indebted to the Namibian Geological Survey and Shell Namibia during follow-up fieldwork in 1995. M.S.M.M. gratefully acknowledges FAPESP and CNPq for financial support. Simon Milner, Roger Swart and Roy McG. Miller are thanked for useful field discussions. Frances Garland is especially thanked for stimulating discussions, and Tim Elliott for constructive draft reading. Reviews by Anton le Roex, Phil Leat and Sally Gibson substantially improved the manuscript and are very much appreciated. The manuscript was prepared by Janet Dryden.
|
FOOTNOTES |
|---|
*Corresponding author. Present address: Faculteit der Aardwetenschappen, Vrije Universiteit, 1085 De Boelelaan, 1081 HV Amsterdam, The Netherlands. Telephone: +31-20-4447316. Fax: +31-20-6462457. e-mail: kirl{at}geo.vu.nl
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONFIELD RELATIONSHIPSBASALT-ANDESITE PETROGRAPHY AND...FIELD CHARACTERISTICS AND...PETROLOGICAL COMPARISON WITH THE...GENERAL GEOCHEMISTRYMAFIC LAVA GEOCHEMISTRYGEOCHEMISTRY OF THE FELSIC...CONCLUSIONSREFERENCES |
|---|
Arndt, N. T., Czamanske, G. K., Wooden, J. L. & Fedorenko, V. A. (1993). Mantle and crustal contributions to continental flood volcanism. Tectonophysics 223, 39–52.
Bellieni, G., Comin-Chiaramonti, P., Marques, L. S., Melfi, A. J., Nardy, J. R., Papatrechas, C., Piccirillo, E. M., Roisenberg, A. & Stolfa, D. (1986). Petrogenetic aspects of acid and basaltic lavas from the Parana Plateau (Brazil): geological, mineralogical and petrochemical relationships. Journal of Petrology 27, 915–944.
Bossi, J. & Navarro, R. (1988). Geologia del Uruguay. Montevideo: Institute of Geology.
Caorsi, J. H. & Goñi, J. C. (1958). Geología Uruguaya. Instituto Geológico Uruguay Boletín 37, 9–73.
Chappell, B. W. (1999). Aluminium saturation in I- and S-type granites and the characterization of fractionated haplogranites. Lithos 46, 535–551.
Collins, W. J., Beams, S. D., White, A. J. R. & Chappell, B. W. (1982). Nature and origin of A-type granites with particular reference to southeastern Australia. Contributions to Mineralogy and Petrology 80, 189–200.[Web of Science]
Comin-Chiaramonti, P., Bellieni, G., Piccirillo, E. M. & Melfi, A. J. (1988). Classification and petrography of continental stratoid volcanics and related intrusives from the Paraná Basin (Brazil). In: Piccirillo, E. M. & Melfi, A. J. (eds) Mesozoic Flood Volcanism of the Paraná Basin—Petrogenetic and Geophysical Aspects. São Paulo: São Paulo University, pp. 47–72.
Cox, K. G. (1980). A model for flood basalt vulcanism. Journal of Petrology 21, 629–650.
Creaser, R. A. (1989). Depth and mineralogy of the magma source or pause region for the Carboniferous Liberty Hill Pluton, South Carolina; discussion. Geology 17, 482–483.
Davies, G. R. & Macdonald, R. (1987). Crustal influences in the petrogenesis of the Naivasha basalt–comendite complex: combined trace element and Sr–Nd–Pb isotope constraints. Journal of Petrology 28, 1009–1031.
De Paolo, D. J. (1981). Trace element and isotopic effects of combined wallrock assimilation and fractional crystallisation. Earth and Planetary Science Letters 53, 189–202.[Web of Science]
Erlank, A. J., Marsh, J. S., Duncan, A. R., Miller, R. McG., Hawkesworth, C. J., Betton, P. J. & Rex, D. C. (1984). Chemistry and petrogenesis of the Etendeka volcanic rocks from SWA/Namibia. In: Erlank, A. J. (ed.) Petrogenesis of the Volcanic Rocks of the Karoo Province. Geological Society of South Africa, Special Publication 13, 195–245.
Ewart, A., Milner, S. C., Armstrong, R. A. & Duncan, A. R. (1998). Etendeka volcanism of the Goboboseb mountains and Messum igneous complex, Namibia. Part I: Geochemical evidence of Early Cretaceous Tristan plume melts and the role of crustal contamination in the Paraná–Etendeka CFB. Journal of Petrology 39, 191–225.
Gallagher, K. & Hawkesworth, C. J. (1994). Mantle plumes, continental magmatism and asymmetry in the South Atlantic. Earth and Planetary Science Letters 123, 105–117.
Garland, F., Hawkesworth, C. J. & Mantovani, M. S. M. (1995). Description and petrogenesis of the Paraná rhyolites, Southern Brazil. Journal of Petrology 36, 1193–1227.
Gibson, S. A., Thompson, R. N., Dickin, A. P. & Leonardos, O. H. (1995). High-Ti and low-Ti mafic potassic magmas: key to plume–lithosphere interactions and continental flood-basalt genesis. Earth and Planetary Science Letters 136, 149–165.
Harris, C. (1995). The oxygen isotope geochemistry of the Karoo and Etendeka volcanic provinces of Southern Africa. South African Journal of Geology 98, 126–139.[Abstract]
Harris, C. & Erlank, A. J. (1992). The production of large volume low-18O rhyolites during the rifting of Africa and Antarctica: the Lebombo Monocline. Geochimica et Cosmochimica Acta 56, 3561–3570.
Harris, C., Smith, H. S., Milner, S. C., Erlank, A. J., Duncan, A. R., Marsh, J. S. & Ikin, N. P. (1989). Oxygen isotope geochemistry of the Mesozoic volcanics of the Etendeka Formation, Namibia. Contributions to Mineralogy and Petrology 102, 454–461.
Harris, C., Whittingham, A. M., Milner, S. C. & Armstrong, R. A. (1990). Oxygen isotope geochemistry of the silicic volcanic rocks of the Etendeka–Paraná province: source constraints. Geology 18, 1119–1121.
Harris, C., Faure, K., Diamond, R. E. & Scheepers, R. (1997). Oxygen and hydrogen isotope geochemistry of S- and I-type granitoids; the Cape Granite Suite, South Africa. Chemical Geology 143, 95–114.[Web of Science]
Hawkesworth, C. J., Gallagher, K., Kirstein, L., Mantovani, M. S. M., Peate, D. W. & Turner, S. P. (2000). Tectonic controls on magmatism associated with continental break-up: an example from the Paraná–Etendeka Province. Earth and Planetary Science Letters (submitted).
Hergt, J. M., Peate, D. W. & Hawkesworth, C. J. (1991). The petrogenesis of Mesozoic Gondwana low-Ti flood basalts. Earth and Planetary Science Letters 105, 134–148.[Web of Science]
Hildreth, W. (1981). Gradients in silicic magma chambers: implications for lithospheric magmatism. Journal of Geophysical Research 86, 10153–10192.
Humphris, S. E. & Thompson, G. (1982). A geochemical study of rocks from the Walvis Ridge, South Atlantic. Chemical Geology 36, 253–274.[Web of Science]
Iacumin, P., Piccirillo, E. M. & Longinelli, A. (1991). Oxygen isotopic composition of Lower Cretaceous tholeiites and Precambrian basement rocks from the Paraná basin (Brazil): the role of water–rock interaction. Chemical Geology 86, 225–237.[Web of Science]
Irvine, T. N. & Baragar, W. R. A. (1971). A guide to the chemical classification of the common volcanic rocks. Canadian Journal of Earth Science 8, 523–548.
Kirstein, L. A. (1997). Magmatism in southern Uruguay and the early rifting of the South Atlantic. Ph.D. Thesis, Open University, 376 pp.
Kirstein, L. A., Garland, F. G. & Hawkesworth, C. J. (2000). Relationship of temperatures and textures in felsic magmatism from Uruguay and the Paraná flood basalt province. Contributions to Mineralogy and Petrology (submitted).
Kuno, H. (1966). Lateral variation of basalt magma types across continental margins and island arcs. Bulletin of Volcanology 29, 195–222.
Leat, P. T., Jackson, S. E., Thorpe, R. S. & Stillman, C. J. (1986) Geochemistry of bimodal basalt–subalkaline/peralkaline rhyolite provinces within the Southern British Caledonides. Journal of the Geological Society, London 143, 259–273.
Le Maitre, R. W., Bateman, P., Dudek, A., Keller, J., Lameyre, J., Le Bas, M. J., Sabine, M. A., Schmid, R., Sorensen, H., Streckeisen, A., Woolley, A. R. & Zanettin, B. (1989). A Classification of Igneous Rocks and Glossary of Terms. Oxford: Blackwell.
Le Roex, A. P., Cliff, R. A. & Adair, B. J. I. (1990). Tristan da Cunha, South Atlantic: geochemistry and petrogenesis of a basanite–phonolite lava series. Journal of Petrology 31, 779–812.
Lipman, P. W. (1965). Chemical comparison of glassy and crystalline volcanic rocks. US Geological Survey Bulletin 1201, D1–D24.
Mahood, G. & Hildreth, W. (1983). Large partition coefficients for trace elements in high silica rhyolites. Geochimica et Cosmochimica Acta 47, 11–30.[Web of Science]
Mantovani, M. S. M. & Hawkesworth, C. J. (1990). An inversion approach to assimilation and fractional crystallisation processes. Contributions to Mineralogy and Petrology 105, 289–302.
Mantovani, M. S. M., Marques, L. S., DeSousa, M. A., Civetta, L., Atalla, L. & Innocenti, F. (1985). Trace element and strontium isotope constraints on the origin and evolution of the Paraná continental flood basalts of Santa Catarina State (Southern Brazil). Journal of Petrology 26, 187–209.
Mantovani, M. S. M., Hawkesworth, C. J. & Basei, M. A. S. (1987). Nd and Pb isotope studies bearing on the crustal evolution of southeastern Brazil. Revista Brasileira de Geociencias 17, 263–268.
May, S. E. (1990). Pan-African magmatism and regional tectonics of South Brazil. Ph.D. Thesis, Open University.
McKenzie, D. (1985). The extraction of magma from the crust and mantle. Earth and Planetary Science Letters 74, 81–91.
McKenzie, D. & Bickle, M. J. (1988) The volume and composition of melt generated by extension of the lithosphere. Journal of Petrology 29, 625–679.
Metz, J. M. & Mahood, G. A. (1985). Precursors to the Bishop Tuff eruption: Glass Mountain, Long Valley, California. Journal of Geophysical Research 90, 11121–11126.
Milner, S. C. & Ewart, A. J. (1989). The geology of the Goboboseb Mountain volcanics and their relationship to the Messum Complex, Namibia. Communications of the Geological Survey Namibia 5, 31–40.
Milner, S. C. & le Roex, A. P. (1996). Isotope characteristics of the Okenyenya igneous complex, northwestern Namibia: constraints on the composition of the early Tristan plume and the origin of the EM1 mantle component. Earth and Planetary Science Letters 141, 277–291.
Milner, S. C., Duncan, A. R. & Ewart, A. (1992). Quartz latite rheoignimbrite flows of the Etendeka Formation, north western Namibia. Bulletin of Volcanology 54, 200–219.[Web of Science]
Milner, S. C., le Roex, A. P. & O’Connor, J. M. (1995). Age of Mesozoic igneous rocks in northwestern Namibia and their relationship to continental breakup. Journal of the Geological Society, London 152, 94–104.
Mizusaki, A. M., Petrini, R., Bellieni, G., Comin-Chiaramonti, P., Dias, J., DeMin, A. & Piccirillo, E. M. (1992). Basalt magmatism along the passive continental margin of SE Brazil (Campos Basin). Contributions to Mineralogy and Petrology 111, 143–160.
Nürnberg, D. & Müller, R. D. (1991). The tectonic evolution of the South Atlantic from late Jurassic to present. Tectonophysics 191, 27–53.
O’Connor, J. M. & Duncan, R. A. (1990). Evolution of the Walvis Ridge–Rio Grande Rise hot spot system: implications for African and South American plate motions over plumes. Journal of Geophysical Research 95, 17475–17502.
Pankhurst, R. J. & Rapela, C. R. (1995). Production of Jurassic rhyolite by anatexis of the lower crust of Patagonia. Earth and Planetary Science Letters 134, 23–36.
Pankhurst, R. J., Leat, P. T., Srouga, P., Rapela, C. W., Márques, M., Storey, B. C. & Riley, T. R. (1998). The Chon–Aike silicic igneous province of Patagonia and related rocks in West Antarctica: a silicic LIP. Journal of Volcanology and Geothermal Research 81, 113–136.
Pearce, J. A. & Norry, M. J. (1979). Petrogenetic implications of Ti, Zr, Y and Nb variations in volcanic rocks. Contributions to Mineralogy and Petrology 69, 33–47.[Web of Science]
Pearce, J. A., Harris, N. B. W. & Tindle, A. G. (1984). Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology 25, 956–983.
Peate, D. W. (1997). The Paraná–Etendeka Province. In: Large Igneous Provinces: Continental, Oceanic, and Planetary Flood Volcanism. Geophysical Monograph, American Geophysical Union 100, 217–245.
Peate, D. W. & Hawkesworth, C. J. (1996). Lithospheric to asthenospheric transition in Low-Ti flood basalts from southern Paraná, Brazil. Chemical Geology 127, 1–24.
Peate, D. W., Hawkesworth, C. J. & Mantovani, M. S. M. (1992). Chemical stratigraphy of the Parana lavas (South America): classification of magma types and their spatial distribution. Bulletin of Volcanology 55, 119–139.[Web of Science]
Piccirillo, E. M., Raposa, M. I. B., Melfi, A. J., Comin-Chiaramonti, P., Bellieni, G., Cordani, U. G. & Kawashita, K. (1987). Bimodal fissural volcanic suites from the Paraná Basin (Brazil): K–Ar age, Sr isotopes and geochemistry. Geochimica Brasiliensis 1, 53–69.
Piccirillo, E. M., Comin-Chiaramonti, P., Melfi, A. J., Stolfa, D., Bellieni, G., Marques, L. S., Giaretta, A., Nardy, A. J. R., Pinese, J. P. P., Raposa, M. I. B. & Roisenberg, A. (1988). Petrochemistry of continental flood basalt–rhyolite suites and related intrusives from the Paraná basin (Brazil). In: Piccirillo, E. M. & Melfi, A. J. (eds) The Mesozoic Flood Volcanism of the Paraná Basin: Petrogenetic and Geophysical Aspects. São Paulo: University of São Paulo, pp. 94–106.
Pirajno, F. (1990). Geology, geochemistry and mineralisation of the Erongo Volcanic Complex, Namibia. South African Journal of Geology 93, 485–504.[Abstract]
Preciozzi, F. S., Spoturno, J., Heinzen, W. & Rossi, P. (1980). Carta del Uruguay, 2nd edn. Montevideo: Uruguayan Geological Survey.
Ramsey, M. H., Potts, P. J., Webb, P. C., Watkins, P., Watson, J. S. & Coles, B. J. (1995). An objective assessment of analytical method precision: comparison of ICP-AES and XRF for the analysis of silicate rocks. Chemical Geology 124, 1–19.
Renne, P. R., Glen, J. M., Milner, S. C. & Duncan, A. R. (1996). Age of Etendeka flood volcanism and associated intrusions in southwestern Africa. Geology 24, 659–662.
Riccardi, A. C. (1988).The Cretaceous system of southern South America Geological Society of America, Memoir 168, 1–161.
Rudnick, R. L. & Fountain, D. M. (1995). Nature and composition of the continental crust: a lower crustal perspective. Reviews of Geophysics 33, 267–309.[Web of Science]
Rushmer, T. (1991). Partial melting of two amphibolites, contrasting experimental results under fluid-absent conditions. Contributions to Mineralogy and Petrology 107, 41–59.[Web of Science]
Skjerlie, K. P., Patiño-Douce, A. E. & Johnston, A. D. (1993). Fluid absent melting of a layered crustal protolith, implications for the generation of anatectic granites. Contributions to Mineralogy and Petrology 114, 365–378.
Sprechmann, P., Bossi, J. & Da Silva, J. (1981). Cuencas del Jurásico y Cretácico del Uruguay. Cuencas Sedimentarias del Jurásico y Cretácico América del Sur 1, 239–270.
Stewart, K., Turner, S., Kelley, S., Hawkesworth, C., Kirstein, L. & Mantovani, M. (1996). 3D, 40Ar–39Ar geochronology in the Paraná continental flood basalt province. Earth and Planetary Science Letters 143, 95–109.[Web of Science]
Sun, S. M. & McDonough, W. F. (1989). Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: Saunders, A. D & Norry, M. J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publication 42, 313–345.
Taylor, S. R. & McLennan, S. M. (1985). The Continental Crust: its Composition and Evolution. Oxford: Blackwell Scientific.
Toyoda, K., Horiuchi, H. & Tokonami, M. (1994). Dupal anomaly of Brazilian carbonatites: geochemical correlations with hotspots in the South Atlantic and implications for the mantle source. Earth and Planetary Science Letters 126, 315–331.
Turner, S. & Hawkesworth, C. (1995). The nature of the sub-continental mantle: constraints from the major-element composition of continental flood basalts. Chemical Geology 120, 295–314.
Turner, S., Kirstein, L., Hawkesworth, C., Peate, D., Hallinan, S. & Mantovani, M. (1999a). Petrogenesis of an 800 metre lava sequence in Eastern Uruguay: insights into magma chamber processes beneath the Paraná flood basalt province. Journal of Geodynamics 28, 471–487.
Turner, S., Peate, D., Hawkesworth, C. & Mantovani, M. (1999b). Chemical stratigraphy of the Paraná basalt succession in western Uruguay: further evidence for the diachronous nature of the Paraná magma types. Journal of Geodynamics 28, 459–469.
Urien, C. M. & Zambrano, J. J. (1973). The geology of the basins of the Argentine Continental Margin and Malvinas Plateau. In: Nairn, A. E. M. & Stehli, F. G. (eds) The Ocean Basins and Margins. New York: Plenum, pp. 135–169.
Walther, K. (1927). Contribucíon al conocimiento de las rocas ‘basálticas’ de la formación de Gondwana en la América del Sur. Instituto Geología y Perforaciones Uruguay 9, 1–43.
Yoder, H. S. (1976) Generation of Basaltic Magmas. Washington, DC: National Academy of Sciences.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. J. Mahoney, A. D. Saunders, M. Storey, and A. Randriamanantenasoa Geochemistry of the Volcan de l' Androy Basalt-Rhyolite Complex, Madagascar Cretaceous Igneous Province J. Petrology, June 1, 2008; 49(6): 1069 - 1096. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. A. Kirstein, L. A. KIRSTEIN, S. KELLEY, C. HAWKESWORTH, S. TURNER, M. MANTOVANI, and J. WIJBRANS Protracted felsic magmatic activity associated with the opening of the South Atlantic Journal of the Geological Society, July 1, 2001; 158(4): 583 - 592. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||