Assessment on the source of geochemical anomalies in the sediments of the Changjiang river (China), using a modified enrichment factor based on multivariate statistical analyses

Rivers can be sinks for potential toxic elements (PTEs) inputted in their systems by both natural and anthropic processes. Many indices have been proposed to assess the contamination degree of sediments and the environmental conditions of surficial water bodies. Above all, enrichment factor (EF) is the most used tool, but also it is the most debated for its limitations. The need for a reference element and for a background/baseline composition makes the EF method dependent on the researcher's expertise, implying that its repeatability may not be granted. Starting from the awareness that geochemical processes, bringing to compositional changes in the environmental matrices, involve multiple elements rather than individual variables, we developed a modified EF (mEF) based on the use of elemental associations. Different multivariate statistical methods (i.e. Robust Principal Component Analysis and Fuzzy Clustering), in a compositional data analysis (CoDA) perspective, were used to set all the terms of the mEF. The mEF was applied to 101 sediment samples collected from a 2 m-long core, covering a sedimentation period of about 150 years (1850–2007), located in the lower Changjiang River (China). The method resulted effective in recognizing most of the signals proceeding from the main natural and anthropogenic events which affected the lower river basin in the considered timespan. The largest geochemical variations recorded fit well the flooding events occurred; besides, the effects produced on the system by the recent socio-economic development (following the end of the civil war in 1949 and the beginning of economic reforms in 1978) and the start-up of the Three Gorges Dam (the world's largest power station since 2012) were also intercepted. The proposed method represents a step forward to enhance the effectiveness of the EF in discriminating geochemical anomalies that may be significant to assess the human historical impact on the environment.