Journal of Petrology Volume 41 Number 5 Pages 605-625 2000
© Oxford University Press 2000
Episodic Silicic Volcanism in Patagonia and the Antarctic Peninsula: Chronology of Magmatism Associated with the Break-up of Gondwana
1BRITISH ANTARCTIC SURVEY, NATURAL ENVIRONMENT RESEARCH COUNCIL, HIGH CROSS, MADINGLEY ROAD, CAMBRIDGE CB3 0ET, UK
2NERC ISOTOPE GEOSCIENCES LABORATORY, KEYWORTH, NOTTINGHAM NG12 5GG, UK
3RESEARCH SCHOOL OF EARTH SCIENCES, THE AUSTRALIAN NATIONAL UNIVERSITY, CANBERRA, A.C.T. 0200, AUSTRALIA
4DEPARTMENT OF EARTH SCIENCES, THE OPEN UNIVERSITY, WALTON HALL, MILTON KEYNES MK7 6AA, UK
Received February 22, 1999; Revised typescript accepted October 5, 1999
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSTRATIGRAPHY AND PETROLOGY OF...PREVIOUS GEOCHRONOLOGYANALYTICAL METHODSNEW RESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
New SHRIMP U–Pb zircon, Rb–Sr whole-rock, and 40Ar–39Ar data are presented for the Jurassic silicic volcanic rocks and related granitoids of Patagonia and the Antarctic Peninsula. U–Pb is the only reliable method for dating crystallization in these rocks; Rb–Sr is prone to hydrothermal resetting and Ar–Ar is additionally affected by initial excess 40Ar. Volcanism spanned more than 30 My, but three episodes are defined on the basis of peak activity: V1 (188–178 Ma), V2 (172–162 Ma) and V3 (157–153 Ma). The first essentially coincides with the Karoo–Ferrar mafic magmatism of South Africa, Antarctica and Tasmania. The silicic products of V1 are lower-crustal melts that have incorporated upper-crustal material. The geochemistry of V2 and V3 ignimbrites is more characteristic of destructive plate margins, but the presence of inherited zircon still points to a crustal source. The pattern of volcanism corresponds in space and in time to migration away from the Karoo mantle plume towards the proto-Pacific margin of Gondwana during rifting and break-up. The heat required to initiate bulk crustal fusion may have been supplied by the spreading plume-head, but thinning of the crust during continental dispersion would also have facilitated anatexis.
KEY WORDS: Antarctic Peninsula; ignimbrites; Jurassic; Patagonia; U–Pb; zircon
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSTRATIGRAPHY AND PETROLOGY OF...PREVIOUS GEOCHRONOLOGYANALYTICAL METHODSNEW RESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Jurassic magmatism in western Gondwana constituted the most voluminous episode of continental volcanism in the Phanerozoic era. During Early–Middle Jurassic time, some (2·5–3) x 106 km3 of basalt and, to a lesser extent, rhyolite were erupted onto the supercontinent in its early stages of break-up. Recent high-precision geochronology (U–Pb and Ar–Ar) has shown that much of the basalt volcanism occurred during a very short period around 183–184 My ago (Encarnación et al., 1996
; Duncan et al., 1997; Minor & Mukasa, 1997), but comparable studies of the silicic volcanism have so far been lacking. This paper is concerned with the chronology of the largest of these silicic outbursts, the Jurassic volcanic province of Patagonia and the Antarctic Peninsula. Although the emphasis here is on the geochronology and its significance for the history of volcanism, full geochemical and isotopic data will be presented in a companion paper (Riley et al., 2000), which will also deal more completely with the petrogenesis of the rocks of this province.
General accounts of the Jurassic silicic volcanism of Patagonia have been given by Gust et al. (1985) and Pankhurst et al. (1998), the latter workers introducing the all-encompassing term Chon Aike Province. Much of extra-Andean Patagonia is covered by undeformed rhyolitic ignimbrites, which are exposed in the structural highs of the North Patagonian and Deseado massifs (centred at latitudes of 
43°S and 47°S, respectively; see Fig. 1). Borehole data show that these rocks extend beneath the intervening Cretaceous sedimentary basins. Andesites and basaltic andesites occur within the province, particularly in the west of the North Patagonian Massif, but are volumetrically less important. Similar Jurassic volcanic rocks, also largely rhyolitic in composition, occur in the southern Andes and in the Antarctic Peninsula; in both areas they are variably folded, tectonized and altered. Together, these volcanic rocks constitute one of the largest felsic magmatic fields known (235 000 km3, Pankhurst et al., 1998). Until now, geochronological controls have not been sufficient to test either the assumed contemporaneity of the various volcanic groups within the province, or their precise relationship to Jurassic magmatism in other parts of southwestern Gondwana. Such tests are fundamental to assessing the significance of Jurassic magmatism in the rifting and subsequent break-up of the supercontinent, particularly with respect to any possible relationship to the Karoo mantle plume (see Storey et al., 1992; Storey, 1995).
|
This paper presents and compares new data acquired by Rb–Sr, Ar–Ar and U–Pb zircon methods (including 21 ion microprobe ages), for samples from southern Patagonia and the Antarctic Peninsula; only the U–Pb method is considered fully reliable for these rocks. The results show that volcanism occurred over an extended period of time (c. 185–155 Ma), but was apparently concentrated in three discrete episodes (Early Jurassic, Middle Jurassic and Late Jurassic). The first of these is roughly coincident with the peak of Karoo–Ferrar basaltic volcanism.
|
STRATIGRAPHY AND PETROLOGY OF THE CHON AIKE PROVINCE AND RELATED ROCKS |
|---|
|
TOPABSTRACTINTRODUCTIONSTRATIGRAPHY AND PETROLOGY OF...PREVIOUS GEOCHRONOLOGYANALYTICAL METHODSNEW RESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The silicic volcanic outcrops of Patagonia and the Antarctic Peninsula are shown in Figs 1 and 2, respectively. In both regions the rocks have been subdivided into more localized formations or informal groups with individual characteristics [see Pankhurst et al. (1998)
and Riley & Leat (1999) for reviews]. Brief summaries are given here.
|
The Marifil Formation (after Malvicini & Llambías, 1974) covers a large area in the northeast of Patagonia, mostly to the north of the Gastre fault zone (Rapela & Pankhurst, 1992). It largely consists of thick (25–100 m), flat-lying, strongly welded pink or red ignimbrite units and lesser rhyolite lava flows. Tuffs, lapilli-tuffs and volcanic agglomerates are interbedded with the ignimbrites, and the sequence is cut by sub-volcanic intrusions, including dacitic and andesitic dykes, which are considered to be co-magmatic with the ignimbrites (Rapela & Pankhurst, 1993).
The Chon Aike Formation (see Sruoga & Palma, 1984), up to 300 m thick, covers an area of some 100 000 km2 in the Deseado Massif. Ignimbrites predominate, with subordinate epiclastic deposits, air-fall tuffs and intercalated lavas: the ignimbrites contain a wide variety of lithic clasts. The more highly welded examples are massive, with coarse columnar jointing; flattened fiamme occur but rheomorphic textures are rare. Some features suggest emplacement over wet sediments or subaqueously. The ignimbrites are phenocryst-poor rhyolites or leucocratic dacites. The principal minerals are quartz, K-feldspar, plagioclase and biotite, with accessory magnetite, ilmenite, apatite, zircon and monazite. The vitrophyres are characterized by perlitic texture; a variety of devitrification textures are displayed. Oxidation, silicification and hydrothermal alteration are widespread. The common alteration assemblage is quartz, sericite, calcite, albite and clay minerals.
The El Quemado, Ibañez and Tobífera formations of western Patagonia have been generally considered to be equivalent and consist of mainly silicic volcanic rocks that have been faulted, tilted and thrust as a result of Andean (Cretaceous–Tertiary) deformation. The Tobífera Formation is largely a sub-surface feature, revealed in boreholes in the Magallanes basin. The rocks of the El Quemado and Ibañez formations are similar to those of the Chon Aike Formation: rhyolitic ignimbrites associated with epiclastic sequences, air-fall tuffs, and breccias of various origins. They also contain some intercalated andesitic lava flows. The ignimbrites have a high content of lithic clasts and mineralogically a high content of plagioclase and Fe-rich biotite. In general, the degree of hydrothermal alteration, propylitic and chloritic, is significantly higher than in the Chon Aike Formation.
Thomson & Pankhurst (1983) assigned the majority of volcanic rocks in the Antarctic Peninsula to the Antarctic Peninsula Volcanic Group (APVG; Jurassic–Cenozoic), regarded as the long-term product of magmatic arc volcanism at the palaeo-Pacific margin. On the east coast of the northern Antarctic Peninsula (Graham Land), the Mesozoic sequences rest unconformably upon variably deformed quartzose metasedimentary rocks of the Trinity Peninsula Group (TPG; Permian–Triassic). In the southern Antarctic Peninsula (Palmer Land), they are associated with Middle to Late Jurassic sedimentary rocks.
The Mapple Formation (Riley & Leat, 1999), with a maximum observed thickness of 1 km, represents the most widespread development of silicic volcanic rocks in the northern Antarctic Peninsula. Eruption was almost entirely subaerial, and only locally subaqueous. Ignimbrite units (typically 5–20 m thick) predominate and, as in the Chon Aike Formation, exhibit wide variation in their degree of welding and lithic content. They are intercalated with minor air-fall units, lag breccias and lava flows. The Mapple Formation was metamorphosed up to greenschist facies and deformed, probably during the end-Jurassic Palmer Land compressional event (Kellogg & Rowley, 1989). Steeply dipping cleavage is developed in epiclastic and many mudflow deposits, and increases in intensity westward. The ignimbrite units are poorly phyric, with an assemblage of embayed quartz, K-feldspar, plagioclase, biotite, magnetite, apatite, orthopyroxene, titanite, rutile and zircon. Feldspar is extensively replaced by calcite, sericite and clay minerals, and biotite is typically altered to chlorite. Spherulites in the rhyolitic ignimbrites indicate high-temperature devitrification.
The Brennecke Formation (Wever & Storey, 1992) comprises silicic metavolcanic units, cropping out at several localities in eastern Palmer Land (Fig. 2). At Brennecke Nunataks a sequence of massive dacitic to rhyolitic lava flows are interbedded with more foliated, welded pyroclastic rocks and black shales. A bimodal association has been recognized with an 150 m thick succession of basaltic lavas (Hjort Formation; Wever & Storey, 1992).
The Mount Poster Formation (Rowley et al., 1982) crops out in the southern Antarctic Peninsula and eastern Ellsworth Land (Fig. 2). It comprises dacitic to rhyodacitic pyroclastic rocks and lava flows, which reach a maximum total thickness of 1–2 km. The succession is dominated by crystal-rich, grey or black welded ignimbrite units and minor lava flows. The ignimbrites contain abundant feldspar crystals; locally abundant lithic and pumice fragments confirm a pyroclastic origin. In the Sweeney Mountains (Fig. 2), strongly welded ignimbrite units with oblate flattened pumice occur in association with rheomorphic ignimbrites, which have a well-defined parataxitic texture. The rhyolites contain abundant phenocrysts of plagioclase, sanidine, hornblende, embayed quartz and Fe–Ti oxides, with an alteration assemblage of sericite, clay minerals and calcite. The Mount Poster Formation is metamorphosed to chlorite grade (Rowley et al., 1982).
|
PREVIOUS GEOCHRONOLOGY |
|---|
|
TOPABSTRACTINTRODUCTIONSTRATIGRAPHY AND PETROLOGY OF...PREVIOUS GEOCHRONOLOGYANALYTICAL METHODSNEW RESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Early attempts to date these rocks used the K–Ar whole-rock method, with highly variable results, compounded in most cases by a lack of good stratigraphical control. A comprehensive review by Cortés (1981)
showed an age range of 240–125 Ma in Patagonia, albeit with a peak in the interval 165–155 Ma. From the northern Antarctic Peninsula, Rex (1976) determined three K–Ar ages of 190, 160 and 88 Ma. Subsequently, a single Rb–Sr isochron of 174 ± 2 Ma was published (Pankhurst, 1982), and an Sm–Nd isochron of 156 ± 6 Ma for garnets from a sill and volcanogenic sediments at the base of part of the succession (Millar et al., 1990).
The first systematic dating in the Chon Aike province was by the Rb–Sr whole-rock method for the Marifil Formation (Rapela & Pankhurst, 1993). These data are shown in Fig. 1, together with other published and new high-precision results, all now calculated to one significant decimal place and with 2
errors for comparability. Apart from any other conclusions, this shows that the Rb–Sr whole-rock technique in rhyolitic rocks is capable of precision equivalent to that of the other two methods. Rb–Sr isochrons for four localities yielded a tight range of 182·6 ± 1·5 to 178·4 ± 1·3 Ma, showing essential synchroneity in this area. Extension of the Rb–Sr method to the remainder of the Marifil Formation (Pankhurst & Rapela, 1995) slightly increased this range with further isochrons of 187·7 ± 1·3 Ma and 181·2 ± 2·2 Ma (the latter for two samples from a suite that was mostly reset at 174·1 ± 2·4 Ma). A significantly younger result of 168·5 ± 2·1 Ma was obtained for a suite from the Río Chubut section in the southwest of the outcrop area (Fig. 1); in this case, there was no evidence for resetting. However, the Ar–Ar results of Alric et al. (1996) from Río Chubut (176·9 ± 1·6 Ma and 178·5 ± 1·2 Ma, errors doubled from their originally quoted 1 level) are within the range of the remaining Rb–Sr and Ar–Ar ages for the Marifil Formation (Fig. 1).
Pankhurst et al. (1993) obtained an Rb–Sr whole-rock age of 168·0 ± 1·9 Ma for samples of the Chon Aike Formation to the west of Puerto Deseado in southern Patagonia (Fig. 1), in agreement with a less precise result of 162 ± 11 Ma by the same method [De Barrio (1993), recalculated]. Alric et al. (1996) presented an Ar–Ar age of 177·6 ± 0·7 Ma from Puerto Deseado, significantly older than the Rb–Sr ages for the Chon Aike Formation.
In Tierra del Fuego, Mukasa & Dalziel (1996) determined a U–Pb zircon age (upper intercept) of 164·1 ± 1·7 Ma for a pluton in the peraluminous Darwin granite suite, thought to be correlative with the Tobífera Formation.
Subsequent to the Rb–Sr whole-rock isochron data of Pankhurst (1982), the only application of modern high-precision techniques in the Antarctic Peninsula is in the Mount Poster Formation (Fig. 2), where preliminary microprobe U–Pb zircon ages for rhyolites have been reported by Fanning & Laudon (1997, 1999) as 167 ± 3, 188 ± 3 and 189 ± 3 Ma. There is a clear need for new analytical work to resolve the discrepancies in the data for Patagonia and to amplify the data for the Antarctic Peninsula. The present study is an attempt to do this, and includes data both for the volcanic rocks and for a number of subvolcanic plutons inferred to be related to the volcanism.
|
ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSTRATIGRAPHY AND PETROLOGY OF...PREVIOUS GEOCHRONOLOGYANALYTICAL METHODSNEW RESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
In this study we have relied chiefly on U–Pb zircon and Ar–Ar dating, methods that have the potential to yield meaningful information in rocks that might have suffered deformation, low-grade metamorphism and/or alteration. Samples were selected as far as possible to be representative of the various formations, with more intense sampling in the critical areas.
U–Pb dating was carried out using a sensitive high-resolution ion microprobe (SHRIMP II) at The Australian National University, Canberra, following the procedures of Compston et al. (1992). All zircon concentrates contained recognizable igneous crystals of a form associated with volcanism, e.g. 50–200 µm needles with aspect ratios of 3:1 to 10:1, simple prismatic terminations and often with hollow gas cavities along the centre of the grain. Many of the needles were broken, possibly as a result of explosive eruption of the ignimbrites. Cathodoluminescence examination showed regular zoning with no distinguishable cores or overgrowths. Analysis spots, mostly within the well-zoned ends of grains, were chosen to avoid cracks and inclusions. The final analytical data were treated as indicated below using Isoplot/Ex (Ludwig, 1999) and are presented in Table 1.
|
40Ar–39Ar dating was carried out by laser step-heating at The Open University, using procedures described by Kelley (1995). Most of the analysed material was hand-picked K-feldspar. Clear grains with feldspar cleavage were chosen from most samples. Occasionally, only milky grains could be found, indicating alteration; this was particularly so of the samples from the Andean outcrops of Patagonia. One basaltic andesite from the Bajo Pobre Formation contained primary biotite, which was analysed in preference to altered plagioclase feldspar in the same sample. The data are available from the Journal of Petrology web site, at http://www.petrology.oupjournals.org.
Localities from which the U–Pb and 40Ar–39Ar data were obtained were also sampled for Rb–Sr whole-rock analysis at NIGL, Keyworth, UK, using the combined X-ray fluorescence and mass-spectrometry method employed by Rapela & Pankhurst (1993) and Pankhurst & Rapela (1995). This was less successful than in the previous work in the Marifil Formation, in that statistically valid isochrons were not obtained, probably because of the higher degree of alteration exhibited by the volcanic rocks of southwestern Patagonia and the Antarctic Peninsula. The data are available from the Journal of Petrology web site.
|
NEW RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSTRATIGRAPHY AND PETROLOGY OF...PREVIOUS GEOCHRONOLOGYANALYTICAL METHODSNEW RESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Rb–Sr whole-rock
The new Rb–Sr data for the Chon Aike Formation did not yield any well-constrained age. Possible errorchrons are highly dependent on the combination and exclusion of individual points, and range from 160 to 180 Ma even for the restricted type locality area around Puerto Deseado. In contrast to the previous concordant isochron results for the Marifil Formation, these data clearly indicate inhomogeneous or disturbed isotope systems.
Although somewhat more consistent, the new Rb–Sr analyses of whole rocks from the Mapple Formation of the Antarctic Peninsula fail to define a precise isochron; the full dataset (n = 53) has a mean square weighted deviation (MSWD) of 30. If the scatter is accommodated by expanding errors by the square-root of MSWD, an ‘age’ of 159 ± 5 Ma results. There is clear visual evidence that the scatter within local sub-groups is less marked (Fig. 3a), but these also provide only errorchrons. Thus samples from station R.6908 in the Mapple Glacier give 153 ± 5 Ma (MSWD = 6·9) and 18 samples from the southeastern Stubb Glacier give 163 ± 7 Ma (MSWD = 4·4). Initial 87Sr/86Sr ratios for these errorchrons are in the range 0·7066–0·7076, comparable with previous values from the Marifil and Chon Aike formations.
|
Finally, a granitic pluton on the west coast of the Antarctic Peninsula around Cape Roquemaurel gave an isochron age of 159·6 ± 3·1 Ma (Fig. 3b). A Jurassic age is thus established for this body. This contradicts previous estimates based on four K–Ar biotite ages averaging 141 ± 3 Ma [Rex (1976)
, recalculated to modern decay constants] and a conventional U–Pb zircon discordia, interpreted as indicating Palaeozoic inheritance combined with crystallization at 127+13/-25 Ma (Tangeman et al., 1996).
U–Pb zircon
A total of nine samples from Patagonia and 12 samples from the Antarctic Peninsula were analysed. The U–Pb isotope data obtained, uncorrected for common Pb, are plotted in Tera–Wasserburg concordia diagrams in Figs 4–6. In these plots, concordant data should lie on a chord between the crystallization age and the coeval composition of common Pb, enabling us to distinguish points for zircons that contain older inheritance (displaced to the left of the main group) or that have suffered post-crystallization Pb loss (displaced to the right).
|
|
|
Clear examples of inherited grains are shown by ignimbrites from Canal Morla Vicuña (Tobífera Formation; 1144 and 490 Ma), Río Pinturas (?Chon Aike Formation; 306 Ma), Toth Nunataks (Brennecke Formation; 301, 224 and 215 Ma), Rachel Glacier (Mapple Formation; 473 and 195 Ma), and Camp Hill (Mapple Formation; 280 Ma). Similar inheritance is common in the Antarctic Peninsula granites: 186 and 185 Ma in the Bildad Peak granite, 244 and 183 Ma in the Mapple Glacier granite, and a variety of ages in the Cape Roquemaurel granite. The sample of the Mapple Glacier granite also contains one grain derived from a Neoproterozoic source (with a 207Pb/206Pb age of 2386 Ma). It must be emphasized that, as the focus of this work was on dating crystallization, these instances cannot be taken as representative of the age or degree of inheritance in the volcanic rocks. On the other hand, involvement of continental crust in the petrogenesis of the rhyolites is established by this inheritance. Obvious Pb loss is shown by one grain in the Pequod Glacier ignimbrite, Mapple Formation (156 Ma). Other rather more subjective cases of inheritance and/or Pb loss are shown by a number of samples; all excluded points are indicated in Table 1, which gives a summary of the SHRIMP ages: more complete data are available from the Journal of Petrology web site.
The inferred crystallization ages are calculated from the remaining concordant data in each case, as the weighted mean 238U–206Pb age after correction for common Pb, assessed using the 207Pb/206Pb ratio and model values according to Cumming & Richards (1975).
The U–Pb crystallization ages from Patagonia range from 154 to 168 Ma, but are individually precise. The oldest age, 178·4 ± 1·4 Ma, is for a sample of the Tobífera Formation from a depth of 2518 m in a borehole that hit basement at 4040 m. A second sample from this formation, from the western side of the Andes, gave a somewhat younger age of 171·8 ± 1·2 Ma. The Chon Aike Formation yielded ages of 168·4 ± 1·6 Ma from Cabo Dañoso and 162·7 ± 1·1 Ma from Bajo San Julián. Five samples from west of the Deseado massif (Fig. 1) give consistent ages of 153–156 Ma. One of these is an ignimbrite from the type locality of the El Quemado Formation at Estancia La Unión, and another from a caldera complex at Sierra Colorada. The latter had previously given an Rb–Sr whole-rock errorchron age of 136 ± 6 Ma (Pankhurst et al., 1993); it is now clear that this cannot be the age of crystallization. A third sample in this age group is from the outcrop at Río Pinturas, which is generally attributed to the Chon Aike Formation, as the volcanic rocks here are flat-lying like those near the Atlantic coast. However, continuity of outcrops is not easily established, and it seems that at least some volcanism in the western Deseado Massif may have been synchronous with the El Quemado Formation. The youngest age of 153·0 ± 1·0 Ma is for a sample of the Ibañez Formation, which is generally considered as the Chilean equivalent of the El Quemado. The final sample in this age group is for the hornblende-phyric Sobral granodiorite of the Monte San Lorenzo massif (Ramos & Palma, 1981). This pluton was previously thought to be of Palaeozoic age but is now confirmed as penecontemporaneous (and cogenetic?) with the Jurassic volcanism of the Andes.
In the Antarctic Peninsula, two ignimbrite samples from the Brennecke Formation gave consistent ages of 184·2 ± 2·5 and 183·9 ± 1·7 Ma, comparable with the Rb–Sr ages of the Marifil Formation of Patagonia, with which it shares some geochemical features. As shown in Fig. 2, these are also similar to the older ages from the Mount Poster Formation at the southern end of the Antarctic Peninsula (Fanning & Laudon, 1997, 1999). However, the main group of new ignimbrite ages, from the Mapple Formation in the area around the Mapple Glacier (Fig. 2), fall in the range 168–173 Ma. The oldest of these is given by a sample from near the base of the sequence (R.6619.4, Rachel Glacier) and the youngest by a sample from near the top (R.6632.10, Stubb Glacier). Samples from the classic localities of Mount Flora and Camp Hill, where the volcanic rocks overlie sedimentary rocks of the Botany Bay Group, gave ages of 162·2 ± 1·1 and 166·9 ± 1·6 Ma, respectively. The granitic body intruded into the Mapple Formation ignimbrites, and considered to be magmatically associated with them (R.6906.3, Mapple Glacier), gave an age at the younger limit of this range. A similar granite from Bildad Peak gave an even younger age of 164·2 ± 2·0 Ma, which is concordant with Rb–Sr isochron ages of 166·7 ± 2·1 Ma and 163·0 ± 1·7 Ma previously obtained for this body (Pankhurst, 1982) and noted in Fig. 2. The Cape Roquemaurel granite gave an age of 164·3 ± 1·7 Ma, based on only nine measurements [this sample was primarily analysed for its inherited component, which is clearly more complex than the single upper intercept of the conventional U–Pb data reported by Tangeman et al. (1996)]. This age is concordant with that of the Bildad Peak granite and just within error of the Rb–Sr age reported above for Cape Roquemaurel. Finally, the Jurassic age of the Cape Monaco granite from the western side of the peninsula (156·0 ± 1·1 Ma) was unsuspected, but is consistent with continuation of plutonic activity following the end of the east coast volcanism. It is also indistinguishable from the age of the Sobral granodiorite of the southern Andes.
Ar–Ar
Six samples were analysed, all from Patagonia (feldspar from five ignimbrites and biotite from one basaltic andesite). Ar–Ar temperature release patterns are shown in Fig. 7. Some of these seem relatively straightforward in that they show good medium- to high-temperature plateaux (five or more adjacent steps with the same age within experimental error). The result from Puerto Deseado (169·1 ± 1·6 Ma) appears reliable and is in agreement with the U–Pb result from Cabo Dañoso. The plateau for the latter locality itself (sample PAT.019.5) is more difficult to assess, as the large number of steps, resulting from overestimation of the necessary quantity of material, vary beyond expected error. The apparent age (177·8 ± 1·8 Ma) is significantly greater than the U–Pb age. It is possible that excess 40Ar in fluid and solid inclusions was released during fracturing as a result of heterogeneous heating by the laser. Plateau ages were obtained from two samples from Estancia La Unión (169·5 ± 2·0 and 144·6 ± 1·4 Ma), but the calculated ages do not correspond to the U–Pb age of 154·5 ± 1·4 Ma for this locality, nor are they mutually consistent. In rapidly cooled volcanic rocks, the observed patterns of high initial ages, followed by a low and then climbing to a plateau, would normally be interpreted as gas release from altered material in the early few steps and from unaltered pure feldspar in the later steps. However, in the present case there is no correlation between the ages and 37Ar/39Ar (Ca/K) or 38Ar/39Ar (Cl/K) ratios, as might be expected if chemically altered material provided the gas released in the early steps. On the other hand, if subsolidus exsolution has taken place, these patterns may be more akin to the Ar-release patterns of plutonic K-feldspars or a mixture of K-feldspar and plagioclase, and the results might be interpreted as representing a cooling history, perhaps also affected by hydrothermally produced excess 40Ar. The critical difference between interpreting a release pattern as a complex K-feldspar or a simple volcanic sanidine may be variation in chemistry with the progressive release. This is again true of the sample from Río Pinturas, which seems to show excess 40Ar in the higher-temperature steps; the release shows a noisy increasing spectrum and no obvious variation in chemistry. The lower ages in the first half of the spectrum correspond more closely to the U–Pb age of 156·2 ± 1·8 Ma. Finally, data for the biotite separate from the Bajo Pobre basaltic andesite were treated by forcing an isochron fit though the composition of atmospheric Ar. With two points excluded, this yielded an age of 150·6 ± 2·0 Ma, within error of Alric et al.’s result of 156·7 ± 4·6 Ma for a sample from this formation.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSTRATIGRAPHY AND PETROLOGY OF...PREVIOUS GEOCHRONOLOGYANALYTICAL METHODSNEW RESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Evaluation of the geochronological data
Rb–Sr data on these rocks have yielded statistically acceptable isochrons for only the very fresh ignimbrites of the Marifil Formation of northeast Patagonia and the granite plutons sampled from the Antarctic Peninsula. Concordance with published Ar–Ar ages (in the case of the former) and U–Pb data (for the latter) is taken as indicating that these Rb–Sr isochron ages are both reliable and equally precise. In all other cases, Rb–Sr data for the volcanic rocks scatter about rather crudely defined errorchrons. The scatter could be due either to initial heterogeneity in the volcanic magmas inherited from a heterogeneous source, or to post-crystallization disturbance. As the scatter is almost non-existent for the relatively fresh samples of the Marifil Formation and greatest for the obviously altered rocks of the Andean outcrops, we believe that the latter is the predominant cause. Some of the errorchrons give results that are concordant with the Ar–Ar or U–Pb ages, e.g. the previous Rb–Sr result of 168·0 ± 1·9 Ma for the Chon Aike Formation; although even here the errorchron is not substantiated by the new data obtained from the new sample collections. In general, the Rb–Sr errorchrons give apparent ages that are too young in comparison with the other methods. Their consistency at about 155–165 Ma for the well-sampled Mapple Formation suggests that they do, however, record a significant event, presumably the gradual closure of Rb–Sr systems to post-crystallization hydrothermal equilibration (Riley & Leat, 1999
).
The U–Pb zircon data obtained in this study, which cover all main rock groups except the Marifil Formation of Patagonia, show a very consistent pattern of high-precision ages. In Patagonia, the two ages obtained from the Tobífera Formation are just outside error of each other and suggest an east–west younging from 178 to 172 Ma. Ages from the Chon Aike Formation in the eastern tracts of the Deseado Massif, and from the main volcanic outcrops of the northern Antarctic Peninsula, fall in the range 163–173 Ma, with a clear peak at 168–169 Ma. The 162 Ma age from Mount Flora, at the northern end of the Antarctic Peninsula, is slightly younger, but compares well with the age of the southernmost Chon Aike Formation sample at Bajo San Julián. Equivalence of the Chon Aike and Mapple formations seems very probable. The granites intruding the volcanic rocks of the Mapple Formation give ages of 169 Ma (Mapple Glacier) and 164 Ma (Bildad Peak), suggesting that emplacement was partially synchronous with the volcanism, but consistent with the field evidence that they are somewhat younger than the volcanic rocks. The oldest ages, obtained from the Brennecke Formation of the southern Antarctic Peninsula (184 Ma), are consistent with ages obtained for the Mount Poster Formation, which crops out even farther to the south (Fanning & Laudon, 1997, 1999). Contemporaneity of these southernmost ignimbrite outcrops with the Marifil Formation of Patagonia is established by comparison with the Rb–Sr and Ar–Ar ages of the latter. Outcrops of the ignimbrites in western Patagonia, including both the El Quemado Formation in the Andes and the westernmost outcrops of the Deseado Massif around Río Pinturas, yield significantly younger U–Pb ages of 155–156 Ma. This suggests that the Río Pinturas outcrops, although flat-lying and undeformed, should be considered correlative with the El Quemado Formation of the Andean region, rather than part of the Chon Aike Formation of the Deseado Massif. Small granite plutons, one in the Andes (San Lorenzo) and one on the west coast of the Antarctic Peninsula (Cape Monaco), have also given very similar ages of 155 Ma, confirming their synchroneity, and probable genetic connection, with this later phase of silicic volcanism.
The new Ar–Ar ages obtained are in part concordant with the data of Alric et al. (1996) and in part with U–Pb, notably in the case of the Río Pinturas and Puerto Deseado samples (Fig. 1). There are, however, some notable discrepancies. The most worrying of these is the result of 177·8 ± 0·8 Ma for an ignimbrite from Cabo Dañoso, which is significantly older than the U–Pb age of 168·4 ± 1·6 Ma for the same locality. It is concordant with a similar Ar–Ar age of 177·6 ± 1·4 Ma from Puerto Deseado (Alric et al., 1996), and also with the Rb–Sr isochron age of 178·4 ± 1·3 Ma from the southernmost outcrop of the Marifil Formation, at Península Camarones (Fig. 1). However, other Ar–Ar ages from the Chon Aike Formation are either within range of, or younger than, the U–Pb ages. We conclude that the samples that have yielded Ar–Ar feldspar ages older than U–Pb zircon ages for nearby volcanic rocks must contain excess 40Ar, presumably incorporated during crystallization. The inconsistent Ar–Ar ages obtained from Estancia La Unión, also mentioned above, are taken as evidence that the K–Ar isotopic system has been disturbed, involving redistribution of Ar in a manner not resolved by the Ar–Ar method. The Ar–Ar age of 150·6 ± 2·0 Ma for biotite from the Bajo Pobre Formation could suggest that this is broadly of the same age as the western Patagonian ignimbrites.
In conclusion, we believe that the most reliable geochronological results for the age of crystallization of the silicic rocks of the province are those given by the U–Pb zircon method. In certain cases, these results are supported by data from the Rb–Sr whole-rock and Ar–Ar methods (and extended geographically, as in the case of the Marifil Formation). In general, however, ages determined by the latter methods are subject to uncertainty related to hydrothermal effects or alteration.
The age pattern of volcanic activity
Figure 8 summarizes the data that we consider most reliable, as judged above, for the chronology of the silicic volcanic activity of Patagonia and the Antarctic Peninsula. Figure 9 illustrates the case for considering the activity to be episodic, with three principal phases identified. The Marifil Formation represents the initial burst of volcanism in the interval 178–188 Ma [mostly Toarcian according to the time-scale of Gradstein & Ogg (1996)]. At this stage, V1 could be considered a relatively long-lived episode, although it may be significant that no Marifil Formation rocks have yet been dated by the U–Pb method. The similar ages from the Mount Poster and Brennecke formations imply that a contemporaneous event is recorded in the southern Antarctic Peninsula. Interestingly, some of the subsequent Antarctic Peninsula granites show occasional evidence of zircon inheritance at 185 Ma (Fig. 6), so that V1 volcanic rocks may be more widespread at depth than indicated by the present outcrop areas.
|
|
A second ignimbrite eruption event, V2, occurred throughout the eastern part of the Deseado Massif and along the eastern coast of the northern Antarctic Peninsula, at 162–172 Ma (with a marked peak at 168–170 Ma), including the Chon Aike and Mapple formations. In stratigraphical terms, this age corresponds to Bajocian–Bathonian, reasonably consistent with the best stratigraphical evidence. Stipanicic & Bonetti (1970) described a fossil flora of Callovian age from the volcanogenic sediments of the La Matilde Formation of Patagonia, which overlies and/or interdigitates with the Chon Aike Formation. A broader range of Middle–Upper Jurassic was ascribed to the flora by De Barrio et al. (1982). Eruption is usually considered to have occurred during the Upper Bajocian–Callovian interval (Di Giusto et al., 1980). The ages obtained from Bajo San Julián and Mount Flora (162·7 ± 1·1 and 162·2 ± 1·1 Ma, respectively) would in fact be Callovian, so that a degree of local diachronism within the formation is implied.
Finally, we can identify a third event at 152–157 Ma (Oxfordian–Kimmeridgian), comprising the eastern Andean outcrops of ignimbrite and associated granite intrusions, and referred to here as V3. This includes both the El Quemado Formation and the supposedly correlative Ibañez Formation on the Chilean side of the Andes, for which K–Ar biotite ages in the range 152–135 Ma have been reported (Suárez & De La Cruz, 1997). At this stage we remain sceptical of the significance of previously reported Rb–Sr and K–Ar ages of less than 150 Ma for this group (e.g. Pankhurst et al., 1993), as the clear evidence of this paper is that such ages are prone to resetting, probably by hydrothermal activity. On the basis of the limited Ar–Ar data, it is possible that the Bajo Pobre Formation of basaltic andesites belongs to the V3 episode.
Two results obtained from the Tobífera Formation are slightly discrepant, perhaps indicating diachronism within this unit. The one from the borehole sample, farther east, gave an age at the younger limit of V1, whereas the other gave an age within the main interval of V2.
These data also establish the migration trends within the silicic province as a whole, as illustrated in Fig. 10. The earliest activity occurred in northeast Patagonia and the southern Antarctic Peninsula. During the V2 event, the locus of activity moved westwards in the Deseado Massif and northwards in the Antarctic Peninsula. The final volcanic–plutonic episode marked a clear westward migration of activity toward the proto-Pacific margin, both in Patagonia and in the Antarctic Peninsula. In the reconstruction shown, representing a possible configuration during V1, these trends constitute a continuous migration towards the Pacific margin. The final ignimbrite event appears to merge in time with the earliest plutonism of the Patagonian batholith, perhaps without any significant overlap (U–Pb zircon ages of 150 Ma, Bruce et al., 1991; Parada et al., 1997). The main phase of Early Cretaceous growth in the Antarctic Peninsula batholith began 142 My ago (Leat et al., 1995).
|
Relationship to Jurassic magmatism elsewhere in Gondwana
As shown in Fig. 8, the age of V1 is essentially coincident with the major episode of pre-Gondwana break-up mafic magmatism represented in Antarctica (the Ferrar Supergroup of the Transantarctic Mountains and the Dronning Maud Land mafic igneous province), Australia (Tasmanian dolerites) and South Africa (Karoo province). High-precision geochronological data are available for all three phases of the Ferrar Supergroup: the Ferrar Dolerite (hypabyssal), the Kirkpatrick Basalt (volcanic) and the Dufek gabbro (plutonic). The Ferrar Dolerite has been dated using U–Pb zircon and baddeleyite at 183·6 ± 1·0 Ma (Encarnación et al., 1996), and at 176·7 ± 3·6 Ma (Ar–Ar; Fleming et al., 1997), although the latter age would be 180·2 Ma using the calibration standard of Renne et al. (1998). Ar–Ar ages of the Kirkpatrick Basalt suggest an age of 176·6 ± 3·6 Ma (Heimann et al., 1994), but may be similarly recalculated to 180·5 ± 3·6 Ma. The age of emplacement of the Dufek Intrusion, the third component of the Ferrar Supergroup, has been determined as 182·7 ± 0·4 to 183·9 ± 0·3 Ma (U–Pb zircon; Minor & Mukasa, 1997) and 182·5 ± 4·8 Ma (Ar–Ar; Brewer et al., 1996). [It should be noted that all previously published Ar–Ar ages referred to in this paper specified errors at the 1 level; these have been doubled to compare with the standard 2 errors on the U–Pb and Rb–Sr ages.]
The Jurassic magmatism of Dronning Maud Land, East Antarctica, has yielded two Ar–Ar ages (Brewer et al., 1996) of 182·4 ± 3·8 Ma (dolerite sill) and 172·4 ± 4·2 Ma (basalt lava), whereas dolerite sills from the Theron Mountains farther south gave a range essentially between these two limits. The older age has been confirmed by Duncan et al. (1997), who carried out an Ar–Ar study on Kirwan basalts from Dronning Maud Land, which provided plateau ages of 181–183 Ma. Available age data for the mafic rocks of Tasmania and southeast Australia are restricted to K–Ar whole-rock ages, which give a range of 170–190 Ma (Hergt et al., 1991).
A detailed study of both mafic and silicic volcanic rocks from the Karoo province indicates a total age range of 179–184 Ma (Ar–Ar; Duncan et al., 1997), with a magmatic peak at 183 Ma. A new age for the Karoo dolerites of 183·7 ± 0·6 Ma (U–Pb zircon and baddeleyite; Encarnación et al., 1996) confirms a very close temporal association between Ferrar and Karoo rocks, with a very short-lived episode of basic magmatism and intrusive activity (no more than 3 My). This may be seen as consistent with the Karoo mantle plume hypothesis (Brewer et al., 1992; Cox, 1992).
The close correspondence in eruptive age between the V1 rhyolites of Patagonia and the Antarctica Peninsula on the one hand and the Karoo plume rocks on the other is emphasized by their relative geographical locations. In Fig. 9 it is clear that these Early Jurassic volcanic rocks were all erupted in a position close to the supposed position of the plume-head, whereas the subsequent V2 and V3 volcanic rocks represent migration away from the plume towards the palaeo-Pacific margin of Gondwana. This pattern clearly reflects the kinematics of break-up and dispersal of the smaller plates resulting from it.
The silicic volcanic rocks in the Lebombo–Nuanetsi region in the eastern part of the Karoo province, which are thought to have been generated by crustal melting over a lithospheric thin spot produced by the plume, show many similarities to the silicic rocks of the Marifil and Brennecke formations. These latter formations are older than the active margin rocks of South America and the Antarctic Peninsula, with a peak in eruptive age at c. 180–183 Ma that coincides precisely with the major Karoo mafic volcanic event. The V1 rhyolites exhibit a ‘within-plate’ geochemical signature, e.g. a tendency to high Nb and Zr contents. Their negative 
Nd values of -4 to -9 (Wever & Storey, 1992; Pankhurst & Rapela, 1995; Pankhurst et al., 1998) indicate a high level of continental crust involvement. Wever & Storey (1992) interpreted the Brennecke Formation as partly generated by crustal melting in an ensialic back-arc setting during lithospheric attenuation, possibly over a rising mantle diapir. For the Marifil Formation magmas, Pankhurst & Rapela (1995) used isotopic and geochemical modelling to deduce a lower-crustal granulitic source of Proterozoic age. Further geochemical data for the Mount Poster Formation (Riley et al., 2000) indicate a variable degree of upper-crustal contamination for these rocks.
The silicic volcanic rocks erupted in Patagonia and the Antarctic Peninsula during the V2 and V3 events (Fig. 8) have no obvious counterparts in the plume-related mafic igneous activity of Gondwana, except that part of the Dronning Maud Land basalt sequence may be as young as the V2 event (Brewer et al., 1996). These ignimbritic rocks generally have rather low Nb contents (<20 ppm), and little large ion lithophile element enrichment (Pankhurst et al., 1998). They also have only slightly negative Ndt values in the range c. -2 to -5 (R. J. Pankhurst & T. R. Riley, unpublished data, 1999). Thus they are geochemically more typical of subduction-related rhyolites, although zircon inheritance, as well as the isotopic and geochemical data, still suggest a primary origin by anatexis of the continental crust [see further arguments given by Riley et al. (2000)]. In time and space, their eruption coincides with a significant shift in the locus of volcanism towards the active Andean margin. In particular, the age of V3 (157–153 Ma) merges into the range of the initial plutonic emplacement of the Patagonian batholith, a coast-parallel feature clearly related to subduction of the Pacific ocean-floor beneath the South American continent.
The possible relationship of this volcanic history to earlier events in Patagonia is not clear at this stage. Unpublished data cited by Rapela et al. (1996) record an Early Triassic ignimbrite at Lihue Calel (Fig. 1). The succession here has given a Rb–Sr whole-rock isochron of 240 ± 2 Ma, Zr contents of up to 700 ppm and Ndt values of -5·4 and -8·3. A Late Triassic Rb–Sr age of 222 ± 2 Ma at Los Menucos, within the North Patagonian Massif, was supported by the associated occurrence of Dichroidium-bearing sedimentary rocks. The latter ignimbrites have Ndt values of -8 to -9·5. These Triassic within-plate rhyolites are thus geochemically comparable with the V1 episode, but their eruption pre-dates the Karoo plume by 40–60 My.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONSTRATIGRAPHY AND PETROLOGY OF...PREVIOUS GEOCHRONOLOGYANALYTICAL METHODSNEW RESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The high-precision age determinations presented in this paper have largely resolved the uncertainties in the geochronology of the Jurassic silicic province of Patagonia and the Antarctic Peninsula. The most reliable and consistent dataset is that obtained by U–Pb dating of volcanic zircons, using the SHRIMP. Rb–Sr whole-rock dating has proved very satisfactory for the fresh, highly welded ignimbrites of the Marifil Formation and for syn- to late-volcanic plutons. Less welded and more altered ignimbrites south of the Gastre fault zone and in the Antarctic Peninsula yield Rb–Sr ages that are
10 My too young, probably as a result of hydrothermal resetting. Ar–Ar dating of feldspars has also been suspect in the most altered rocks from the Andean outcrops and, moreover, appears to reflect initial excess 40Ar in some samples of the Chon Aike Formation that have yielded apparent plateau ages older than the U–Pb zircon ages. The new data show unambiguously that silicic volcanic activity in the region extended throughout Jurassic time, but was concentrated in three main episodes. The first of these (V1, Early Jurassic, 188–178 Ma) resulted in the eruption of a vast volume of rhyolitic ignimbrites with within-plate affinities in northeast Patagonia and the southern Antarctic Peninsula. This was coincident with the far more extensive emplacement of mafic igneous rocks in Gondwana generally associated with the Karoo mantle plume. The Middle Jurassic V2 event (172–162 Ma) represented a shift in the focus of eruption to southern Patagonia and the northern Antarctic Peninsula. The chemical characteristics suggest that magma generation occurred in continental crust with less evolved composition. The poorly welded nature of the ignimbrites suggests a lower eruptive temperature than for the highly welded ignimbrites of the Marifil Formation. The final event identified was of Late Jurassic age (V3, 157–153 Ma). It is represented by the Andean outcrops of rhyolitic ignimbrite, following a significant western shift in volcanic activity. These rocks are predominantly active-margin volcanic products and are associated with granitoids. The V3 event almost overlaps with the earliest known intrusions of the Patagonian and Antarctic Peninsula batholiths. Increased activity at this destructive margin could account for the higher degree of deformation and hydrothermal alteration in these volcanic rocks compared with those of the preceding events.
At this stage it is clear that the Jurassic volcanic province of the region is a complex polygenetic one, formed during multiple events over 35 My and probably generated by a variety of mechanisms, among which the melting of pre-existing continental crust was a dominant process. These reflect changes in the tectonic regime of Gondwana break-up, and the pattern of changing volcanism appears to be associated with the spread of heat away from the Karoo mantle plume towards the Pacific margin. However, the extended period of silicic volcanism covered the actual break-up of the region into smaller plates and their active dispersal (Fig. 9). Thus the large amount of crustal extension and thinning that must have been involved could also have been a major cause of the silicic magmatism through decompression melting. All these events preceded Early Cretaceous separation and opening of the Atlantic Ocean.
Note added in proof
The ages reported in this paper, and the associated westward migration of volcanism, are fully concordant with the results and conclusions for eastern Patagonia of a new Ar–Ar study [Feraud et al., Earth and Planetary Science Letters, 172 (1999), 83–96].
|
ACKNOWLEDGEMENTS |
|---|
The main fieldwork on which this study is founded was carried out with invaluable assistance from Carlos Rapela, Marcelo Márquez and Patricia Sruoga (in Argentina); from Simon Abrahams, Mike Austin and BAS support staff (in Antarctica); and from Pancho Hervé (in Chile: Proyecto Fondecyt 1980741). Jorge Skarmeta and Ricardo Fuenzalida were of help in obtaining the ENAP borehole sample. The contributions of Phil Leat and Bryan Storey, both during fieldwork and in the writing of this paper, are gratefully acknowledged. Ian Millar was of great assistance during SHRIMP analyses. Helpful reviews were supplied by C. J. Hawkesworth, S. M. Kay and M. Suárez. This paper is a contribution to IGCP Project 436 (Pacific Gondwana Margin).
|
FOOTNOTES |
|---|
*Corresponding author. Present address: NERC Isotope Geosciences Laboratory, Keyworth, Nottingham NG12 5GG, UK. Telephone: +44-115-936-3263. Fax: +44-0115-936-3302. e-mail: r.pankhurst{at}bas.ac.uk
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSTRATIGRAPHY AND PETROLOGY OF...PREVIOUS GEOCHRONOLOGYANALYTICAL METHODSNEW RESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Alric, V. I., Haller, M. J., Féraud, G., Bertrand, H. & Zubia, M. (1996). Cronología 40Ar/39Ar del volcanismo jurásico de la Patagonia extrandina. XIII Congreso Geológico Argentino y III Congreso de Exploración de Hidrocarburos, Buenos Aires, Actas V. Buenos Aires: Asociación Geológica Argentina, pp. 243–250.
Brewer, T. S., Hergt, J. M., Hawkesworth, C. J., Rex, D. & Storey, B. C. (1992). Coats Land dolerites and the generation of Antarctic continental flood basalts. In: Storey, B. C., Alabaster, T. & Pankhurst, R. J. (eds) Magmatism and the Causes of Continental Break-up. Geological Society, London, Special Publication 68, 185–208.
Brewer, T. S., Rex, D., Guise, P. G. & Hawkesworth, C. J. (1996). Geochronology of Mesozoic tholeiitic magmatism in Antarctica: implications for the development of the failed Weddell Sea rift system. In: Storey, B. C., King, E. C. & Livermore, R. A. (eds) Weddell Sea Tectonics and Gondwana Break-up. Geological Society, London, Special Publication 108, 45–61.
Bruce, R. M., Nelson, E. P., Weaver, S. G. & Lux, D. R. (1991). Temporal and spatial variation in the southern Patagonian batholith; constraints on magmatic arc development. In: Harmon, R. S. & Rapela, C. W. (eds) Andean Magmatism and its Tectonic Setting. Geological Society of America, Special Paper 265, 1–12.
Compston, W., Williams, I. S., Kirschvink, J. L., Zhang, Z. & Ma, G. (1992). Zircon U–Pb ages from the early Cambrian time-scale. Journal of the Geological Society, London 149, 171–184.
Cortés, J. M. (1981). El substrato precretácico del extremo noreste de la Provincia del Chubut. Revista de la Asociación Geológica Argentina 26, 217–235.
Cox, K. G. (1992). Karoo igneous activity, and the early stages of the break-up of Gonwanaland. In: Storey, B. C., Alabaster, T. & Pankhurst, R. J. (eds) Magmatism and the Causes of Continental Break-up. Geological Society, London, Special Publication 68, 137–148.
Cumming, G. L. & Richards, J. R. (1975). Ore lead isotope ratios in a continuously changing earth. Earth and Planetary Science Letters 28, 155–171.[ISI]
De Barrio, R. E. (1993). El volcanismo ácido jurásico en el noroeste de Santa Cruz, Argentina. XII Congreso Geológico Argentino y II Congreso de Exploración de Hidrocarburos, Mendoza, Actas IV. Buenos Aires: Asociación Geológica Argentina, pp. 189–198.
De Barrio, R. E., Arrondo, O., Petriella, B. & Artabe, A. (1982). Estudio geológico y paleontológico de los alrededores de la estancia Baja Pellegrini, Provincia de Santa Cruz. Revista de la Asociación Geológica Argentina 37, 285–299.
De Giusto, J., Di Persia, C. A. & Pezzi, E. (1980). Nesocraton del Deseado. In: Geología Regional Argentina, Vol. II. Córdoba, Argentina: Academia Nacional de Ciencias, pp. 1389–1430.
Duncan, R. A., Hooper, P. R., Rehacek, J., Marsh, J. S. & Duncan, A. R. (1997). The timing and duration of the Karoo igneous event, southern Gondwana. Journal of Geophysical Research 102B, 18127–18138.
Encarnación, J., Fleming, T. H., Elliot, D. H. & Eales, H. V. (1996). Synchronous emplacement of Ferrar and Karoo dolerites and the early breakup of Gondwana. Geology 24, 535–538.
Fanning, C. M. & Laudon, T. S. (1997). Mesozoic volcanism and sedimentation in eastern Ellsworth Land, West Antarctica; conflicting evidence for arc migration? Geological Society of America, Abstracts with Program 29, A51517.
Fanning, C. M. & Laudon, T. S. (1999). Mesozoic volcanism, plutonism and sedimentation in eastern Ellsworth Land, West Antarctica. Eighth International Symposium of Antarctic Earth Sciences, Wellington, New Zealand, 4–8 July 1999 (Abstract). Wellington: The Royal Society of New Zealand.
Fleming, T. H., Heimann, A., Foland, K. A. & Elliot, D. H. (1997). 40Ar/39Ar geochronology of Ferrar dolerite sills from the Transantarctic Mountains, Antarctica: implications for the age and origin of the Ferrar magmatic province. Geological Society of America Bulletin 109, 533–546.
Gradstein, F. M. & Ogg, J. (1996). A Phanerozoic time scale. Episodes 19, 3–5.[ISI]
Gust, D. A., Biddle, K. T., Phelps, D. W. & Uliana, M. A. (1985). Associated middle to late Jurassic volcanism and extension in southern South America. Tectonophysics 116, 223–253.