Comparison of sequential extraction and principal component analysis for determination of heavy metal partitioning in sediments: the case of protected Lagoon El Kelbia (Tunisia)

A sequential extraction procedure applied to surface sediments from El Kelbia Lagoon determined the partitioning of Ba, Co, Cr, Cu, Ni, Sr, V and Zn among evaporites, carbonates, Fe–Mn oxides, organic matter and silicates. To validate the procedure, the sequential extraction results (SER) were compared to principal components analysis (PCA) using major and trace element concentrations and mineralogical quantitative data of surface sediments. SER showed that a part of Sr was highly mobile; Cu, Sr, and Zn and a part of Co and Ni were mobile depending on pH conditions; Cr and V were strongly bound to silicate phases; Co and Ni were partitioned between carbonates, oxides and silicates, and a great part of Ba and Sr were bound to organic matter or sulfides. An agreement was found with PCA in terms of partitioning among minerals for most trace elements. Moreover, the absence of correlation between Ca concentrations and the abundances of calcite, gypsum or dolomite could be explained by an important fraction of Ca bound to organic matter. Also, unexpected negative correlations between abundances of smectite and illite could be explained by a transformation of illite into smectite. Thus, SER and PCA were mostly convergent, which enabled: (1) validation of the extraction procedure used, and (2) refinement of interpretations of the origin and relations between minerals.