The mechanism and timescales of soil formation in the hyper-arid atacama desert, Chile

Fan Wang, Purdue University


The planetary surfaces that evolve in the near absence of water are strikingly different from surfaces where water is abundant, but their formation is poorly documented. This research is an in-depth exemplary work in the Atacama Desert in northern Chile to understand the soil formation in hyper-arid environments. The geochemical, isotopic and mineralogical composition of atmospheric dust deposited along a West-East transect in the Atacama Desert, stretching from the Pacific Ocean to the Andean Altiplano, was evaluated. Results suggested that the deposited salts are mainly sourced from marine aerosol, as well as fog, local entrainment of surface material and secondary reactions. This has led to a proposed soil formation mechanism in hyper-arid environments consisting of four stages of development: 1) initiation of soil development on regional bedrock material induced by salt fracturing, 2) maturation of soil sequence featured by a continuous trapping of atmospheric dust and salts and the salt segregation into discrete ionic zones, 3) termination of soil accretion due to the formation of surface blocky layer, and 4) unexpected alteration of soil accumulation induced by anthropogenic disturbance. A combination of two long-lived radioactive nuclides (10Be and 36Cl) was used to constrain the timescale of a 225-cm-deep soil profile from the hyper-arid core of the Atacama. Considering the differences of 10Be and 36Cl nuclides in the half-life and mobility in the soil, 10Be could indicate the fate of insoluble silicate dust (dominant in the soil matrix) because it readily adsorbs to soil particles, while 36Cl could be used to trace the transport and fate of salts considering the chlorine's solubility. Soil 10Be concentrations showed a systematic decline from surface to 225 cm deep, which has been reproduced using a simple model that assumes the soil matrix, including 10Be, builds up as layers over time while 10Be decays in situ. This concurs with the essence of the soil formation mechanism proposed above, i.e. the net soil accumulation via atmospheric deposition. The model estimates an age of ~6.6±0.4 Ma for the total soil profile. Likewise, the 36Cl/Cl ratios showed a systematic decline with depth and a simple accumulation model that chloride builds up over time via atmospheric inputs and 36Cl radioactively decays in situ also reproduces the data remarkably well. This model suggests the atmospheric origin of soil chloride and the chloride age at 225 cm of 860 (±90) ky. Paleo-precipitation variations, which can potentially impact soil formation, were first detected from the deviations of the observed 10Be concentrations from the above model estimates that are likely mainly due to changes in 10Be delivery rates impacted by invariant precipitation rates. The 10Be data suggested a drying after ~4.7 Ma, likely due to Andean uplift, and the returning to an insignificant wet period at ~1 Ma, possibly connected to global climate change. Similarly, 36Cl discontinuities with depth also suggested the interruption of long-term hyper-aridity by brief wet periods that induced chloride migration. To investigate the timescale of salt accumulation (36Cl age: ~860 ky), a new precipitation proxy (soil NO3- cap-delta 17O) was calibrated using the stable oxygen isotope measurements in nitrates from four deserts with different mean annual precipitation rates. The paleo-precipitation history for the formation of the 225-cm deep trench soil profile was then reconstructed based on the soil NO3- cap-delta 17O proxy, indicating seven wet-dry cycles, likely corresponding to possible glacial-interglacial cycles operating on the 100 ky scale. Crypto-biotic crusts (CBC), a consortium of pioneer species in hostile environments, may play an important role in soil evolution. Preliminary data from a CBC distribution transect indicated that CBCs can enhance dust trapping and physical protection of the dust from wind erosion, leading to the thickened loose soil profile beneath CBCs compared to at nearby sites without CBCs. Fine particle fractions in the soil under CBCs were higher beneath CBC than those in non-CBC sites, and this was probably related to the enhanced small dust particle trapping or in situ weathering. That the profile beneath CBCs had fewer soluble ions compared to sites without CBCs suggested an enhanced leaching at sites with CBCs, though it is still unclear whether the enhanced leaching was the cause or the reason of the presence of CBCs. The bomb spike with the 36Cl/Cl ratio of 398 × 10-15 was preserved on the surface at a CBC site over the past ~60 years, while the subsurface 36Cl/Cl ratios were likely homogenized due to the past more significant wet events.




Michalski, Purdue University.

Subject Area


Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server