Andean salars: Reflections of Earth and climate history

A strong sun sends its rays over the white expanse reaching far to the horizon between the Andean summits.

Thick crusts of salt, accumulated over several millennia owing to the combined low rainfall and extremely high evaporation which prevail in these high-altitude tropical environments, completely cover the whole landscape. These desert expanses are called salars. They formed under the arid climate that has predominated in the central Andes for several million years. There are hundreds of them, distributed over Bolivia, Chile and Argentina.

Chemical compositions are complex and varied, characteristics with high interest for industry. They contain elements and salts such as boron, potash, lithium and sodium carbonate. What are the origins of the salars' chemical profiles? To find out, an IRD geologist and his research partners have traced back the history of their formation and brought out evidence of the events leading to their chemical composition.

They studied and analysed 80 salt lakes, scattered over 1000 km of Andean cordillera in Bolivia and northern Chile. Most of them had hitherto never been studied scientifically.

Data from hundreds of samples and analyses were used to reconstruct the processes that formed the present lakes. Two broad categories of elements and salts emerged, depending on their origin: those from volcanic rocks (lithium, boron) and those coming from previous, ancient salars (chlorides and sulphates). During a period of intense volcanic activity, older salars were covered over by material emitted by volcanoes. Subsequently, infiltrating waters became enriched in chemical elements released by rock weathering.

In parallel, the old salars were dissolved, resulting in bodies of groundwater ranging in character from brackish to strongly saline. The waters stemming from the volcanic rocks and from the salars then mixed together, in their turn feeding into new basins. The present salt lakes then formed. Now, the ancient salars, which used to be buried underneath, have been completely dissolved.

The research team also conducted evaporation modelling and changes in salt concentrations of the salar interstitial waters and compared these theoretical results with measured concentrations. The simulations took no account of any geochemical interaction with the sediments or dissolution of older saliferous formations. They were based solely on the chemical composition of inflows to the salars. Significant divergences then appeared between the results of the model and the measured values, particularly in Bolivia.

Input of additional groundwater, interaction with sediments, solution of pre-existing saliferous formations and so on. The divergences reveal phenomena that the researchers had not hitherto identified. For example, in Bolivia, analysis of the results revealed a major anomaly: several salars, which should have been rich in sodium carbonate, and therefore basic, are in fact neutral or even slightly acid, with high sulphate content. This unexpected composition is the result of aeolian erosion of sulphur deposits associated with volcanic activity in the neighbouring areas.

The salars are moreover invaluable water supply resources, for both Andean populations and large-scale mining in the region, especially for copper mines. It was therefore also important to characterise them to assess the potential of these water resources, establish reference data on the latest hydrochemistry and thus prepare the bases for monitoring the impact of industrial activity on these fragile highly particular ecosystems.

Besides the contaminated water it can generate, the mining industry uses great quantities of fresh and brackish water for mineral extraction and processing. These activities pump massive volumes from the groundwater bodies that feed the salars and in doing so they alter their chemical composition. They also compete with local communities who need supplies for drinking water. This problem is crucial in the central Andes and their Pacific -facing slopes.