Unambiguously pristine and largely unpolluted sediments from the Late Weichselian and Holocene infillings of the Meuse residual channels in northern Limburg (the Netherlands) have been sampled to determine the natural compositional variation of the river’s suspended load.
Bulk geochemical and granulometric analyses demonstrate that about 70% of the variation can be ascribed to hydrodynamic mineral sorting. Clay- and fine silt-sized phyllosilicates are the most important deterministic features, hosting the bulk of AI2O3, TiO2, K2O, MgO and trace element variability (notably Ba, Cr, Ga, Rb and V). Quartz is abundant in the fine and coarse sand fractions. Na2O and the Zr-Nb-Nd-Y quartet relate to albitic feldspars and heavy minerals, respectively, in the coarse silt fraction. The granulometry should therefore be quantified if geochemical baseline data for a particular geological unit or region are drawn up and for the evaluation of potentially polluted sediments.
Although provenance has not changed, the composition of Meuse sediments cannot be considered constant over a time frame of 1000–10,000 years, due to climatic change. Weathering of phyllosilicates in both interstadial and interglacial soil environments and changing relative source-area contributions alter the detrital clay-mineral supply to raise the AI2O3 and lower the K2O and MgO contents in Holocene Meuse sediments. Early diagenetic siderite and vivianite formation in gyttjas causes relative accumulations of Fe2O3, MnO, P2O5, Co, Ni and notably Zn above the phyllosilicate background values. These accumulations are natural and show that sediments with elevated trace metal contents are not necessarily polluted. Very early atmospheric pollution in relation to ore mining and smelting activities in the Roman era, however, probably caused elevated Pb contents in Subatlantic humic clays and peat samples, long before the historic pollution of the Industrial Revolution started.
The A12O3, Fe2O3 and CaO contents are used to predict the trace-element values as a function of sample granulometry, siderite/vivianite and lime content, respectively. As such, they can provide a sound basis for environmental researchers to determine baseline values of heavy metals in bulk samples of fine-grained fluvial sediments.