Rare earth elements as statistical sentinels of pollution and paleoenvironments?: Application to a highly polluted estuary in southwestern Spain.

Multivariate analyses have been applied to the REE contents of three cores collected in the Tinto estuary, SW Spain, an extremely polluted area. Results indicate an extremely correlation between all REE, which behave as a single variable. A slight natural pollution peak and three anthropogenic pollution peaks are identified, related with the first mining activities, the Roman period and a recent intensive mining accompanied by a heavy industrial pollution. In all these peaks, the increase of Cu is parallel to that of MREE, which are configured as the best indicators of pollution among REE. Statistical analyses clearly differentiate four groups, each consisting of samples from different environments. Although grain size and this strong pollution alter the study of REE as environmental indicators, it is possible to recognise groups of samples with a common origin or to identify the surface extent of a given pollution peak.