1 CONICET, Insugeo, Facultad de Ciencias Naturales e IML, Miguel Lillo 205, 4000 Tucumán, Argentina. E–mail: acecha@unt.edu.ar

2 CONICET, Igeo, Facultad de Ciencias Exactas y Naturales, Universidad Nacional de San Juan. Av. Ignacio de La Roza y Meglioli, 5400, San Juan. E–mail: jpmilana@hotmail.com

3 CONICET, Museo de Geología y Paleontología, Universidad Nacional de Comahue, Neuquen, Argentina. E–mail: sheredia@uncoma.edu.ar

4 Instituto de Biociências, UNESP, campus de Botucatu, Botucatu, SP. CP 510. E–mail: btsimones@ibb.unesp.br

Key words: Stratigraphy. Biostratigraphy. Tremadocian. Incised valley. Argentina.

The Angosto de la Quesera is a classical locality displaying Cambro–Ordovician sequences in the Cordillera Oriental of Salta province, northwestern Argentina (Figure1).

Keidel (1943) was the first author to focus on this locality, describing the lithological units, the stratigraphy and genesis of the conglomerates, called by him "glacialmarine deposits". This first detailed paper represented the starting point for a long lasting debate on the geological interpretation of this locality.

During the last decade, there was a renowned interest on the Angosto de la Quesera sequences, and the many paleontological data, arguments and models that came out from these debates are helping to clarify this interesting locality (e.g., Aceñolaza, 1997; Moya, 1999, 2002; Hongn et al., 2001a; 2001b; 2003).

In this opportunity, an integrated sedimentological and paleontological analysis has been carried out, presenting an alternative interpretation for the outcrops, sharing some elements with the early referred papers.

The conglomeratic unit of the Angosto de la Quesera is bounded by a major basal erosive surface. The lateral tracing of this surface indicates that conglomerates are found over the second unit of Meson Group at the western side of La Quesera creek (marginal, distinct facies of Campanario Formation), while to the east, the basal erosion removed Meson Group entirely and it is overlying directly the Tastil Granite (517 and 526 Ma: Hongn, 2003). To the north–west, the apparition of Lampazar Formation and a sandy bioturbated unit on top of it, ascribed to Cardonal Formation, suggests that the erosive relief associated to the conglomerate is no less than 300 m. We have not detected a tectonic unconformity, but certain low angularity observed sometimes between the units, probably as a consequence of a slightly different slope of deposition for the fine sandstones of Meson Group and the boulder conglomerates, and not the effect of tectonical rotation (Figures 1, 2).

Conglomerates can easily be divided in two members, separated by a sort of unconformity that indicates a hiatus and a long interval of reworking of the conglomerate member upper surface (Figure 2 A).

Figure 1. Location map of the Angosto de la Quesera and geological units cropping out in the area (Cordillera Oriental of Salta province, NW Argentina). Below: Schematic section of A–B–C showing sequence architecture of the locality. Different contact relations can be seen in the same. Location map partially taken from Hongn et al. (2001b).

Lower member: This member groups all conglomerate and boulder conglomerate facies. Within this lithology we can recognize two main stages of deposition. The lower stage is characterized by the presence of boulder conglomerates, interbedded with coarse badly sorted and well–sorted conglomerates, and minor sand lenses showing high stage plane bed and current–ripple laminations. In addition, wave action is possible to recognize on top of the conglomerate beds. Some intervals show well sorting and a completely open framework, with infilling sandy matrix only near the top of the bed. Boulders in this conglomerate are dominantly quartz sandstone and granite–granodiorites fragments, but other types of sedimentary rocks are found as well as quartz vein fragments. The two dominant types might be larger than 5 m in length, a fact that perhaps conducted to other geologist to think in small bodies of the Tastil Granite intruding these conglomerates. This unit is usually not thicker than 5–6 m but locally can be reduced to 2–3 meters or be missing. Keidel (1943) reported the presence of striated clasts in this conglomerate, and an exhaustive search only provided few clasts with doubtful striations. However, the clast pictured by Keidel shows typical glacial striations, which along together with common faceting of the clasts suggest some glacial inheritance. Primary glacial deposition is ruled out due to the frequent presence of primary tractive structures.

The second part of this conglomerate member is 6 m thick in the type section and it is characterized by pebble and cobble conglomerates with occasional small to medium size boulders, alternating with thick packages of sandstones. Most conglomerates show better internal stratification and irregular tops with evidence of reworking by waves. The sandy intervals are fine grained with some medium–sand beds showing basal current structures and normal gradation. The top of this unit is characterized by a conglomerate 1.7 m thick, containing boulders up to 1.8 m large, protruding from the top. The sandy matrix is scarce, and in the upper 60 cm, the matrix turns into a bioclastic deposit of fragmented lingulid shells. The top is irregular but sharp. A search for proglacial evidences indicated the existence of very rare dropstones in this unit, and only one of them having undoubtly proglacial origin, due to the associated structures.

The upper member lacks completely of conglomerates and starts directly with fine grained sands alternating with turbidite–like medium to coarse sands. The sequence of this upper member is coarsening and thickening up, showing a clear increase on the energy of the system. Many sandstone show unidirectional structures in the lower and middle part, but towards the top, wave structures are clear. Additionally, the hemipelagic–like deposits separating sandy beds show an increasing effect of wave action, turning from well laminated fine silty sand to heterolithic deposits as wavy bedding and lenticular bedding. The top of this member is characterized by a 60 cm unit that shows the gradation from the high– energy sandstones of the top of this complex, to the fine grained greenish–yellowish shales that locally characterizes the Saladillo Formation.

Conodonts, graptolites, trilobites, echinoderms, trace fossils and conulariids were recovered from the strata of the Angosto de la Quesera. Sandstones below the conglomerates of the eastern flank of the Angosto de la Quesera were sterile by means of shelly fauna (Mesón Group). Only a particular association of relatively shallow water trace fossils has been localized (Cruziana isp. Planolites isp., Palaeophycus tubularis. and Spirodesmos isp. A a.o.). Clasts of the conglomeratic unit proved to be highly fossiliferous, with conodonts, trilobites, brachiopods, echinoderms and conulariids mostly belonging to the reworked Lampazar and Cardonal formations. Boulders of the lower sector of the conglomerates yielded Problematiconus perforatus, Oneotodus sp. cf. O. simplex, Teridontus nakamurai and Scolopodus sp., while boulders of the upper sector of the conglomerates yieled very small Nogamiconus sp., Teridontus nakamurai, Scolopodus sp., Scolopodus filosus, Variabiloconus sp. and Drepanoistodus sp.. Conodont fauna suggest a Lower Ordovician, pre–Paltodus deltifer biozone (probably Cordylodus angulatus–base of Rossodus manitouensis). Trilobites were also recovered from conglomerate boulders, and they belong to the Parabolina (N.) frequens argentina biozone. In addition, the presence of Rhabdinopora sp. in the upper conglomeratic unit supports the early referred Lower Tremadocian above the Cambro–Ordovician boundary but below Bryograptus kjerulfi biozone (Aceñolaza, 1997).

Finally, stem fragments of pelmatozoans and the rare presence of conulariids within the conglomeratic unit introduces new elements in the paleontology of NW Argentina. Conularids are common fossils in marine Devonian strata of the Malvinokaffric Realm, particularly from the central andean area (e.g., Bolivia, Babcock et al., 1987) and from the cratonic basins of Brazil (e.g., Paraná Basin, Simões et al., 2000, with references) and Uruguay (Méndez–Alzola and Sprechmann, 1973). Until this new mention of the Angosto de la Quesera, the oldest conulariids recorded in South America were found in the Ordovician/Silurian strata of the Manacapuru Formation (Ramos et al., in press), the Silurian Pitinga Formation (Siviero, 2002), the Amazon Basin and the Silurian occurrences of the Vargas Peña Shale from Paraguay (Babcock et al., 1990). Hence, this Tremadocian conulariid remain of the Cordillera Oriental of Salta is worthwhile to mention.

Figure 2. Lithostratigraphic scheme and facies arrangement of the different units in the area. A. Depositional hiatus separating both members of the conglomerate unit: a lower Conglomerate Member and an upper Coarse Sandstone Member. B. Conglomerates erosive surface contacting all lithological units in the region, from the Cambrian red granites to the Cambro/Ordovician Cardonal Formation. C. Pre–Mesón erosive surface evidenced by stratified lingulid shell beds in the sandstones above the granite.

Above the conglomeratic unit follows the lower sector of the Saladillo Formation, whose paleontological data is quite known (Moya et al., 1994; Moya and Monteros, 2000; Albanesi et al., 2001, a.o.) with Bryograptus kjerulfi as the most remarkable fossil. Skolithos, Palaeophycus and small number of Cruziana are frequent trace fossils in these strata.

It is quite clear that when Keidel (1943) saw the large boulders of the basal conglomerates he could not think in other explanation than glacial. However we propose a different origin, probably by cohesive and uncohesive flows powered by the high slopes associated to the sides of this paleovalley which, as we can see from present–day reconstruction, had at least 20° of slope. Uncohesive, high energy flows, powered by high slopes were probably the responsible for the open–framed conglomerates. This was facilitated by a wave–action environment that was the responsible to keep the deposit clean and of reworking of the top of the conglomerates. Many of the boulders suggest some glacial heritage, which is compatible with the presence of although rare dropstones that unequivocally point to a pro–glacial environment The conglomerates accumulates in an overall transgressive array, and perhaps the reworked top of the last conglomerates with the bioclastic (fragmented shells) matrix, it is pointing to a ravinement surface.

Above this transgressive interval the upper member is deposited as a single prograding unit associate to a relatively stable and high eustatic level. As indicated above, the level of energy of this member, and wave influence increases to the top although still the bases of sandstones show clear directional structures ascribed to the generation of rip–currents, very common in a wave–dominated coastal environments. As a result, the basal part of the sand bed shares many characteristics with turbidites due to the deposition by a decelerating flow but to the top it is clear the association of the beds with wave events (small storms). We did not find clear hummocky bedding which is probably due to the existence of a limiting paleogeography (the incised valley). As sedimentation proceeded, the depth was reducing, and the sequence turned into sandy–rich and showing more energy. This evolution is truncated at its top by the fast inundation that marks the apparition of Saladillo Formation graptolithic shales.

The conglomerates of the Angosto de la Quesera represent a Lower Tremadocian incised valley system with an erosive contact respect to several Cambro–Ordovician units: the Tastil Granite, the entire Mesón Group and the Lampazar and Cardonal Formations of the Santa Victoria Group. This unit is placed below the Saladillo Formation (Figures 1A–C, 2)

The conglomeradic unit displays a basal conglomerate member, deposited by an interaction of high energy unidirectional cohesive and uncohesive flows powered by high slope and wave action. Few clasts in the conglomerate shows inherited glacial features, and some proglacial structures are present (e.g., dropstones).

Conglomerates display an upper ravinement surface, separating the basal conglomerates from a coarse sandstone upper member, here interpreted as a progradational unit, in a siliciclastic wave–dominated shore associated to a Highstand Interval.

Faunal elements and biostratigraphic data support the interpretation here proposed related to the incised valley system eroding several Cambro–Ordovician units, from the Tastil Granite up to the Cardonal Formation.

Authors thank F. Aceñolaza, J. Rossi and A. Toselli for critically reading this contribution. J. De Moraes Leme and S. Coelho Rodrigues (Instituto de Biociências, Botucatu, SP–Brasil) provided valuable help on the taxonomic analysis of the conularid. Line drawings were kindly done by E. Gómez and D. Ruiz Holgado. This contribution was finished by a financial support of the Fundación Antorchas (Argentina).