CONTROLS ON THE ISOTOPIC COMPOSITION OF SOIL CARBONATE IN A HYPERARID ENVIRONMENT: A CASE STUDY FROM THE ATACAMA DESERT, CHILE

The Atacama Desert of northern Chile spans over 6000m in elevation from the Pacific to the High Andes, yet precipitation is generally < 20 mm/yr and reaches a maximum of 200 mm/yr above 4000m in elevation. Soil carbonate is present along the Coast below 1000m and between 2500-4000m along the Pacific slope of the Andes. In order to understand the controls on the stable isotopic composition of soil carbonate in this extreme hyperarid environment, we studied the relationship between per cent cover of vegetation, the d13C composition of vegetation, and the d13C and d18O composition of soil carbonate along three elevational transects that run from 130m to 4000m in elevation. We then compared these results with reported d18O composition of Andean summer precipitation, Pacific Coastal Fog, and Pacific rainfall. Isotopic values of soil carbonate at depth (> 45 cm) range from –8.2 to +7.9‰ d13C VPDB and from –6.0 to +7.0 ‰ d18O VPDB. Vegetation at nearly all sites is C3-dominated. d13C values from most sites are too high to have formed in equilibrium with overlying vegetation. This suggests significant mixing of atmospheric- with plant-derived-CO2, a result of the very sparse vegetation cover at most sites. d13C values tend to be lowest at the biotically most productive sites (along the coast and at 3200-3500m) and to increase towards the margins of current Absolute Desert, where vegetation cover is thinnest. d13C values >+5 ‰ are too high to be explained by previous simple steady-state diffusion models, and appear to require non-steady state loss of CO2 from the soil. The d18O isotopic values of soil carbonate display almost no trend with elevation (R2 0.08). d 18O values of soil carbonate between 3700-4000 m range from –6.0‰ to +2.0‰, between 3.0 and 11.0 ‰ more enriched than predicted equilibrium values for carbonate precipitated from meteoric water at average annual temperatures. The difference between predicted and measured d18O values decrease with elevation. We suggest that the lack of a trend of decreasing d18O values with lower elevation is mainly the result of 1) evaporative enrichment associated with intense sublimation of high elevation snowfall (>3500m), 2) evaporative enrichment of soil water, and 3) increased winter precipitation from the Pacific that occurs at lower elevations (<3000m).