Multi-proxy reconstruction of the paleo-hydrological cycle during Andean Plateau uplift, NW-Argentina

Tuesday, 16 December 2014
Alexander Rohrmann1, Dirk Sachse1, Manfred R Strecker1, Andreas Mulch2, Heiko Pingel1 and Ricardo N Alonso3, (1)University of Potsdam, Potsdam, Germany, (2)Biodiversity and Climate Research Centre, Frankfurt, Germany, (3)Universidad Nacional de Salta, Departmento de Geología, Salta, Argentina
Controversies around the uplift history of the Central Andes may partly be due to the common use of single stable-isotope elevation proxies that are limited in recording the paleo-hydrological cycle during uplift. Our new, multi-proxy stable isotope data from 68 lipid biomarker n-alkanes, 62 soil carbonates (SC), and 15 volcanic glass shards (VG) from a well-dated section in the intermontane Angastaco Basin in the E Cordillera of NW Argentina (25°45 S, 66 W) provide insight into the paleo-hydrological cycle during uplift and outward growth of the adjacent Puna plateau’s eastern margin. The samples cover a time interval from ~10 to 2 Ma with a resolution of ~ 0.2 Ma. C29 and C31 alkanes yield δD and δ13C values ranging from -95 to -160 ‰ (VSMOW) for δD and -23 to -36 ‰ (PDB) for δ13C. Measured SC range from 18 to 31 ‰ (VSMOW) for δ18O and -4 to -17 ‰ (PDB) for δ13C, whereas VG δD values range from -71 to -95 ‰ (VSMOW). In concert with published clumped-isotope temperature and VG data (Carrapa et al., 2014) our data indicate humid foreland conditions after 10 Ma that became drier at 6.5 Ma as a result of basin uplift and orographic barrier formation farther east. We constructed equidistant time series to calibrate and evaluate each proxy signal to determine which part of the hydrologic cycle is recorded and how each is influenced by regional surface uplift. The results show that water hydrating the VG reflects precipitation, whereas SC and n-alkanes record soil-water compositions. All three proxies show similar short-term trends in δD, δ18O and δ13C on timescales of < ~1 Myr, but deviate significantly over timescales of > ~1 Myr. For example, VG δD values show a gradual depletion during uplift from 6.5 to 4 Ma, whereas C29 and C31 δD values are D-enriched in response to basin aridification, which is in line with a change from C3 to C4 plants as recorded by δ13C C29 and C31. SC δ18O values show a buffered signal with a similar trend as δD of n-alkanes. As no modern C4 plants exist >2000 m (Cotton et al., 2014), an observed shift at 3 Ma in δ13C C29 and C31 alkanes back to C3 vegetation cover suggests that most of the basin was uplifted above 2000 m (to modern elevations) during that time. The reconstruction, calibration, and signal analysis of these proxies on a single record is crucial to understand how each proxy responds to surface uplift.