Analysis of the sedimentary regime of the Pleistocene paleo lake where dental remains of the equidae Hippidion devillei were found in Mendoza highlands, Argentina

Abstract

The hydrochemical and sedimentary features of water bodies are directly linked to climate conditions during genesis and evolution (Williamson et al., 2008). Lacustrine deposits record climatic signals spatially and temporally, so their study allows the reconstruction of these environmental conditions and their evolution (Lapridaet al., 2014). The high resolution of the lake sedimentary record has a key role in paleoenvironmental and paleoclimate understanding (Gierlowski-Kordesch and Kelts, 1994; Bohacs et al., 2000). Hence, paleolimnology is an interdisciplinary science developed from chemical, physical and biological information of sedimentary records of aquatic bodies that allows paleo-environmental reconstruction (García-Rodríguez, 2002). In particular, these studies are performed in current lakes by drilling and core extraction, which limits the interpretation of macro structures given that many cores are required to perform a comprehensive analysis of past environmental and climatic conditions. In this work we analyze the sedimentary sequence of the exposed profile of a paleolake where remains of an equine belonging to the Hippidion devillei species were found (Cerdeño et at., 2008). These remains correspond to dental material consisting in premolars and molars that constitute the first fossil record of a Pleistocene horse found in the province of Mendoza (Argentina) (32º S).

The objective of our work is to understand the sedimentary regime and evolution of this high altitude-paleolake (2680 m) associated with the megafauna of South America. This paleolake, named as DL3, is preserved along the Arroyo Minero located in the northwest of Mendoza province, specifically in the north extreme of the El Plata range (Central Andes) (Fig. 1). This paleolake was naturally dammed by the PB3 rock avalanche (Fig. 2). The chronology of this collapse was established by terrestrial cosmogenic nuclides (TCN) dating, performed on surficial big boulders on the top of rock avalanche deposits. 36Cl isotope was used because of the volcanic composition of these blocks. An~ 115±23 Ka age (Moreiras et al., 2015) was determined after discarding a younger result from 4 dated samples (PB4.01 in Table 1).

The total lake succession reaches 29 m in thickness, for this reason several profiles were analyzed considering the abrupt topography (Fig. 3). Each level was described considering grain size, sedimentary structures, and main facies according to the lithofacies code of Miall, (1978) (Table 3). Color of layers was determined with field humidity and was described according to the Munsell chart. Descriptions were documented by pictures taken with a digital camera (Sony DSC, zoom 15X). Three sections were a priori established in the whole lake sequence profile. A Lower section corresponding to the lowest 9 m.; the Medium section comprehends the following 10 m up, and the Upper section involves the last 10 m of the sequence. Data of the revealed profiles were used to draw the lacustrine stratigraphic sequence (Fig. 4).

Sedimentary analysis shows a thick succession of alternating levels of sands and silts with thicker subordinate boulders and gravels in the profile base. The facies association a is composed by matrix supported gravels, stratified sands and laminated silt-clays (Gm-Sh-Fl) (Table 2) representing the lower 10 meters of the profile. It consists of the interbedding of clays and silts layers of 50-70 mm (Fig. 5a) with 60 mm – thick sand layers. The clays and silts are laminated (facies Fl), with reddish, light brown and green sheets of 0.5 to 1 mm in thickness (Fig. 5b and c); they can be massive (Fm) and present organic matter (OM) content. Sand levels are medium to fine in grain with yellow color. They can be massive (Sm facies) or show planar lamination (Sh facies) within 1-2 cm thick layers. The Sm facies can present CaCO3 nodules (Fig. 5e), associated with oxidized levels showing rain drop marks. The Sh facies presents lags of rhyolitic angular clasts of 5-30 cm in diameter (Fig. 5d), from local outcrops. Towards the top of the lower section we can observe matrix supported gravels (Gm facies) with angular to subangular blocks of 0.07 m in diameter of rhyolitic porphyries and vulcanites within a sabulitic matrix of coarse sand of equal composition, its base is erosive. The facies b association is composed by stratified sands and laminated silt-clays (facies Sh-Sr-Sp-Fl-Fr) that extend over 9 m of the exposed profile (Fig. 6a). The Sh facies is characterized in this case by lentiform levels (Fig. 6b) that may have superimposed load structures and silt cracks (Fig. 6e and f). The Sr facies have ripples of 2 cm in width and 1 cm in height (Fig. 6c). The Sp facies are characterized by 7 cm planar cross-bedding structures forming 0.5-1 cm sets and 2 cm cosets containing the found remains of the Hippidion devillei (Fig. 6d). Silts and clays (facies Fl) contain sand dikes (Fig. 6g) and mud drapes with cracks drying out at the tops. They can form ripples of 0.5 mm in height on inclined surfaces between 10-15°. We identified rework, cut and fill structures supported by silts in diffuse contact. In certain strata, the presence of a reverse fault is observed towards the top of this section.

Concerning interpretation of the facies associations, the alternation of facies Sh-Sm with facies Fl suggests a zone of interaction of tractive flows probably sheet flood with processes of deposition by suspension (Young et al., 2003, Fisher et al., 2008). The presence of massive sandstones arranged in coarsening-thickening upward pattern and covered by levels rich in carbonate nodules indicates alteration of the fabric by pedogenetic processes (Retallack, 1994). This indirectly raises the possibility of a system with low sediment input. The deposit of clasts, possibly dealing with a hyper concentrated flow without lamination, came from the outside or edge of this level. It is ruled out that it is a local fall since no deformation is observed at the base of said level. The angular blocks originating from the Choiyoi Group reinforce the hypothesis of sediment supply from the basin margin. Taking into account the described facies, and the scarce presence of organic matter, one can deduce a sub-environment of ephemeral sandy plain (Smoot and Lowenstein, 1991). While, interpretation of the facies b association with an alternation of facies Sh and Fl with greater contribution of fine sediments allows inferring a more central position in the system with greater dominion of decantation processes and lower contribution of coarse sediments from the margin of the basin through tractive processes (Gierlowski-Kordesch and Rust, 1994). The presence of Sr suggests tractive processes associated with sheet flood flows. The Fr facies suggest ripple formation by wave action in possible bodies of water seasonally generated (Gierlowski-Kordesch, 1998). Loading structures are interpreted as a result of alternating sediments of different densities and with different degrees of water saturation (Van Loon, 2009). The presence of wafers suggests the development of a thin sediment cover during the cessation of water and sediment supply (mud drapes) and subsequent erosion (wafers) during the subsequent flood event of the system (Fisher et al., 2008). The thin covering added to the presence of desiccation cracks is a robust evidence of the system’s ephemeral regime (Smoot and Lowenstein, 1991). Sand dams are associated with possible seismites (Obermeier, 1996, Rodríguez-Pascua et al., 2000, Alfaro et al., 2010). The reverse type fault corresponds to the most central area of the playa (Fig. 4i) to which sliding structures (slumps) could have been extended. The dominion of the facies Fl towards the top of the succession allows inferring an ephemeral muddy plain (Smoot and Lowenstein, 1991).

Facies associations correspond to two sub-environments: sandy and muddy plains that conform an ephemeral playa lake. The normal gradation trend of layers with presence of thick sediments (Gm) at the base and finer layers (Fl) towards the top of the succession suggest a decrease in the sedimentary contribution over time and a more central position in the basin. These findings are significant for Quaternary paleo environment record of the Central Andes. The predominance of arid climate conditions restricts development and preservation of organic matter. Thus, the use of commonly used techniques such as pollen determination or radiocarbon dating in Quaternary records is practically unfeasible in this region. For this reason, the study of paleolakes if they show well preserved successions, may be a response to this limitation in paleoclimate reconstruction in the Arid Andes.