5Alluvial soils can store a wide range of metal contaminants originating from point and diffuse sources. The biological health of these soils is important as they act as an interface between terrestrial and aquatic environments, therefore playing an important role in maintaining the quality of surface waters. The aim of this work was to examine the lability, solubility and bioavailability of Pb, Zn and Cd in the top (0-15 cm) and sub soil (35-50 cm) of metal contaminated alluvial soils from the Trent catchment, U.K. Samples 10 (n=46) were collected from within 10m of the river bank. Sources of contamination include historical mining, industry, sewage treatment works and energy production. Enrichment factors based on total metal concentrations showed that contamination in soils declined with distance from the mining areas before rising again as a result of general urbanisation and identified point sources (e.g. river dredging activities). Pore waters were extracted and isotopic dilution and single extraction assays were undertaken on the soils 15 to assess the lability and solubility of the metals. Multi-element isotopic dilution assays were used to determine the labile pool or E-value of these metals in the soil. E-value concentrations were found to range between 0.5-14 mg/kg, 11-350 mg/kg and 25-594 mg/kg for Cd, Pb and Zn, respectively. Comparison of the E-value assay with the EU standard extraction assay for trace element availability (0.05M EDTA) showed that EDTA extractions generally over-estimated the E-value for Zn and Pb,with 20 the difference being greater as contamination levels increased. Bioavailability of the metals was assessed by speciating the pore waters [M Sol ] using WHAM 7 to obtain estimates of free ion activities (M 2+ ). Values of (M 2+ ) were compared to published 'median critical limits' for soils that estimate levels of protection for 95% of biological species. For each of the three metals, (M 2+ ) was found to exceed these critical limits at some sites. Solubility of the metals are reported using K d values expressed using both the 25 total and E-value as the solid phase. Finally we examine the use of different metal pools (total, E-value, EDTA-extractable) and different measures of Fe oxide pools (total, free total, free amorphous), in predicting [M Sol ] concentrations and (M 2+ ) using WHAM 7 in assemblage modelling mode. Overall best simultaneous model predictions for the three metals were obtained using the E-values. Larger overestimates of [M Sol ] and (M 2+ ) were produced using the EDTA and total metal pools whereas a better 30 fitting in the prediction was obtained when models used either the total or the free total FeOx pools.
IntroductionAlluvial soils act as archives for fluvially dispersed metal contaminants from numerous sources [1][2][3][4][5] . These soils are often used in intensive agricultural systems. Several studies in the UK 35 have previously noted the potential for metal transfer into cattle, when grazing grass grown on contaminated alluvial soils [6][7] . In a...