1 Department of Geological Sciences, California State University, Long Beach, CA 90840 USA. E–mail: scfinney@csulb.edu

2 Department of Geological Sciences, University of Michigan, Ann Arbor, MI 48109 USA.

3 Department of Geosciences, University of Arizona, Tucson, AZ 85721 USA.

4 CONICET, Universidad Nacional de San Juan, 5400 Rivadavia, San Juan, Argentina.

5 Department of Geology, Washington State University, Pullman WA 99164 USA.

Key words: Geochronology. Paleogeography. Upper Ordovician. Precordillera. Argentina.

Considerable geological evidence indicates that the Precordillera terrane of northwest Argentina is exotic to Gondwana and that it rifted from the Ouachita embayment of Laurentia in early Cambrian time, drifted as a microcontinent across the Iapetus Ocean, and then collided with the proto–Andean margin of Gondwana in middle to late Ordovician time (Astini et al., 1995; Dalziel et al., 1996; Thomas and Astini, 1996). When the Precordillera terrane docked with Gondwana, according to this model (Astini et al., 1995; Thomas and Astini, in press), it collided with the Famatina system, which had formed as a westward–facing subduction complex and volcanic arc along the proto–Andean margin between 490 and 470 Ma, as a result of the convergence of the Precordillera terrane. In contrast, however, are the interpretations of Rapela et al. (1998) and Keller (1999) that the Precordillera terrane did not arrive at its position adjacent to the Famatina system until after late Silurian time, which raise questions about the history of the Precordillera during Cambrian and Ordovician time. In fact, Finney et al. (2003) concluded, on the basis of U/Pb age populations of detrital zircons in the upper Lower Cambrian Cerro Totora Formation and the lower Upper Ordovician Las Vacas Formation, that the Precordillera terrane is of Gondwanan affinity, not Laurentian, and that it was not adjacent to the Famatina system until after the Late Ordovician. In this paper, we present detrital zircon analyses from additional samples in the Precordillera, from the Upper Ordovician La Cantera and Empozada formations, that further support these conclusions.

The Upper Ordovician La Cantera and Las Vacas formations stratigraphically overlie the black shale of the Middle Ordovician Gualcamayo Formation that, in turn, overlies the Cambrian to Middle Ordovician platform carbonate succession that distinguishes the eastern tectofacies of the Precordillera. These formations are composed of conglomerate, sandstone and fine–grained siliciclastic sediment that filled fault–bounded basins (Astini, 1998a, 1998b). In contrast, the Upper Ordovician Empozada Formation is composed of shale, sandstone, and debris flow beds and unconformably overlies the Los Sombreros Formation, which is composed of shale with some carbonate and a variety of mass flow deposits including sandstone, conglomerate, breccia, diamictite, and olistoliths. These lithologies accumulated primarily through hemipelagic settling, turbidity flows, and debris flows in a continental slope to rise setting on the western (present coordinates) margin of the carbonate platform of the Precordillera terrane (Heredia et al., 1990; Gallardo and Heredia, 1995; Keller, 1999).

Figure 1. Index map of northwest Argentina showing location of samples from (1) La Cébila Formation in the Sierra de Ambato, (2) Las Vacas Formation at Quebrada de Las Plantas in the northern Precordillera near Guandacol, (3) La Cantera Formation at the Don Braulio Creek section in the Villicum Range, and (4) Empozada Formation at Quebrada de San Isidro in the southern Precordillera.

The new results from the La Cantera and Empozada formations are compared to those of the Las Vacas Formation, and the results from these three Upper Ordovician formations from the Precordillera, in turn, are compared to those of the Upper Cambrian? to Lower Ordovician La Cébila Formation of the Sierras Pampeanas.

Locations of samples from the La Cébila and Las Vacas formations (Figure 1) are described in Finney et al., (2003). Two samples were taken from the La Cantera Formation in the well–known section at Don Braulio Creek in the Villicum Range (Figure 1): a lower sample from bed N/1 and an upper sample from bed N/6 in Peralta’s stratigraphic column (Peralta, 1990, Figure 1). Both samples are within the Nemagraptus gracilis graptolite Zone. The sample from the Empozada Formation was collected from the section on the south side of Quebrada San Isidro (Figure 1) from the lower part of the predominately shale interval that contains graptolites of the Climacograptus bicornis Zone (Cuerda and Alfaro, 1992; Toro and Brussa, 2001).

As with the Las Vacas sample, samples from the La Cantera and Empozada formations are dominated by detrital zircon populations largely of Mesoproterozoic age (Figure 2), and zircon populations of the Las Vacas and La Cantera formations are very similar to the Mesoproterozoic age population of the La Cébila sample, which defines a broad group from 0.9 to 1.5 Ga with a large peak at 1.1–1.3 Ga and progressively fewer grains from 1.3 to 1.5 Ga (Figure 2). The curve for the Empozada sample differs significantly in that the great majority of zircon grains are 1.3–1.5 Ga. As in the Las Vacas sample, zircons of Neoproterozoic and early Paleozoic age are rare in the La Cantera and Empozada samples. Two grains in the upper La Cantera sample have ages of 645.2±5.9 Ma and 613.8±4.6 Ma. The Empozada sample includes one grain with an age of 528.8±7 Ma and another with an age of 470.9±22.7 Ma.

Figure 2. Relative age–probability curves showing U–Pb individual detrital zircon age spectra for La Cébila Formation (Upper Cambrian? to Lower Ordovician), lower La Cantera Formation and upper La Cantera Formation (Nemagraptus gracilis graptolite Zone; lower Upper Ordovician), Las Vacas Formation (Climacograptus bicornis graptolite Zone; lower Upper Ordovician), and Empozada Formation (Climacograptus bicornis graptolite Zone; lower Upper Ordovician), all discussed in this paper. Number of grains analyzed in each sample is shown in parentheses. Zircon analyses and method for plotting age populations are those described in Finney et al. (2003).

As earlier reported for the Las Vacas sample (Finney et al., 2003), zircons with ages of magmatic units in the Famatina orogeny (ca. 490–470 Ma, Rapela et al., 1998; Pankhurst et al., 2000) are virtually absent from the La Cantera and Empozada samples. In fact, only one zircon grain out of 328 dated from the four Upper Ordovician samples is time correlative with Famatinian magmatism. From this, we conclude that the basins in which the Las Vacas, La Cantera, and Empozada formations accumulated during the interval of 460–455 Ma (i.e., the age of the N. gracilis and C. bicornis graptolite zones as calibrated by Cooper, 1999) received negligible sediment from the Famatina arc. This is surprising given 1) the present–day proximity of the Las Vacas and La Cantera depositional basins to the Famatina system; 2) the abundance of possible Famatina–age zircons in the La Cébila sample (11 grains with ages of 480–515 Ma), which is evidence of surface exposure and erosion of Famatina granitoids or volcanics in the Early Ordovician, 3) the interpretations of Astini et al., (1995) and Astini (1998a) that the depositional basins of the Las Vacas and La Cantera formations formed by extension after collision of the Precordillera terrane against the Famatina magmatic arc, which stood as a topographic high immediately to the east, and 4) the interpreted eastern source area, including the Famatina arc, for granitoid clasts in the Las Vacas Formation and equivalent formations (Astini, 1998b; Thomas et al., 2002; Thomas and Astini, in press).

In contrast to the wide range of detrital zircon age populations in the La Cébila sample, the narrow range in the La Cantera, Las Vacas, and Empozada sandstones indicates that their immediate provenances were of limited variability. The provenance of the Mesoproterozoic grains in the La Cantera and Las Vacas sandstones was likely the same as that of the Mesoproterozoic population in the La Cébila sample, given their similar plots (Figure 2). This provenance would have been the Gondwanan continent, given the depositional setting of the La Cébila sample, and possibly the Sunsas (1.0–1.3 Ga) and Rondonian–San Ignácio (1.3–1.55 Ga) geochronological provinces of the Amazonian craton (Tassinari and Macambira, 1999). The Rondonian–San Ignácio province, in particular the Santa Helena batholith with U/Pb ages of 1.42–1.45 (Geraldes et al., 2001), is a possible source area for the sandstone of the Empozada Formation. The system delivering sediment for the Empozada Formation must have been distinct from that of the La Cantera and Las Vacas formations. It did pick up zircon grains possibly from the Famatina system, but, in comparison to the abundance of possible Famatinian grains in the La Cébila sample, the Famatina system was a negligible source area for the Empozada Formation and probably very far removed. It is of interest that of the three Upper Ordovician formations sampled, a Famatina–age grain occurs only in the formation that would have been farthest and most geographically isolated from the Famatina system in the paleogeographic reconstruction of Astini et al. (1995) and Thomas and Astini (in press).

Our results indicate that the Precordillera terrane was far from the Famatina system in the Late Ordovician, yet received extrabasinal sediment largely from Mesoproterozoic orogenic belts of the Gondwana continent. Such a scenario is difficult to reconcile with a Laurentian origin of the Precordillera terrane because it would require a collision of the Precordillera terrane with Gondwana in the Mid to Late Ordovician followed much later by rifting and migration to its present position. Instead, we propose that the Late Ordovician extensional basins were pull–apart basins that developed along a transcurrent fault along which the Precordillera migrated before later arriving at its position adjacent to the Sierra de Famatina.

Acknowledgement is made to the Donors of the Petroleum Research Fund, administered by the American Chemical Society, for support of this research. The Radiogenic Isotope Lab at the University of Arizona funded sample preparation and analysis. Universidad Nacional de San Juan provided field vehicle and equipment support.

Accepted: June 15, 2003