Data on a wide range of trace elements, determined by multi-element procedures based on inductively coupled plasma emission and mass spectrometry, are presented for the major rivers entering the Humber estuary. The average trace element chemistry varies significantly across the region with a clear divide between the southern and northern rivers. This pattern is related to lead-zinc-barium mineralisation and flood plain sediment contamination from historic mining activity, and to historic and current industry and urbanisation to the south. For the industrial/urban rivers, most dissolved components dilute with flow; this pattern is most clearly seen for those components with little acid available particulate (AAP) fractions. In the rural/mineralised areas, dissolved components show more variable flow patterns. AAP fractions and dissolved components with significant associated AAP fractions usually increase with flow in all rivers. The different trace elements show multi-linear relationships with one another. These are much tighter than the links between dissolved and AAP components of the same element. Two main groups of closely associated elements emerge, but these groups are different on northern and southern rivers. The first group corresponds to elements which dilute with flow and this group includes significantly more trace elements to the south where industrial and urban inputs dominate. For this group, within-river chemical processes do not seem to be operative as linear relationships with each of the trace elements of the group and chloride are observed: chloride is chemically conserved within river systems and is predominantly found in the point source effluent discharges. The second group corresponds to those determinands which increase with increasing flow for both dissolved and AAP fractions and they have a high AAP fraction: relationships show much more scatter for this group. Links between dissolved and particulate fractions for this second group are weak and are not well described by empirical partition coefficient relationships which are commonly used in environmental modelling studies. Rather, AAP fractions are much more closely linked with suspended solid concentrations than theThe results point to the importance of contributing sources and hydrological controls in determining dissolved and AAP concentrations in the Humber rivers. The role of within-river chemical controls is much less clear-cut and may well be of second order importance. (C) 1997 Elsevier Science B.V. dissolved component. The reasons for the contrasting behaviour between the two groups probably reflects the inability of 0.45 mu m filtration (47 mm diameter cellulose nitrate Whatman sterile filters in this case) to remove all colloidal sized materials. Thus, for this second group, at high flows, when suspended sediments are at their highest, there is the greatest potential for acid available enriched micro-particulates to pass through the filters. This feature provides a fundamental schism for environmental research for this second and wide ranging group: process based water suspended-sediment interaction modelling requires a clear separation between truly dissolved and truly particulate fractions; water industry based environmental sampling and management strategies, as well as legislative water consent controls, are based on a 0.45 mu M separation.
