Predicting the bioavailability of sediment-bound uranium to the freshwater midge (Chironomus dilutus) using physicochemical properties

Assessment of uranium (U)-contaminated sediment is often hindered by the inability to accurately account for the physicochemical properties of sediment that modify U bioavailability. The goal of this research was to determine whether sediment-associated U bioavailability could be predicted over a wide range of conditions and sediment properties using simple regressions and a geochemical speciation model, the Windermere Humic Aqueous Model (WHAM7). Data from U-contaminated field sediment bioaccumulation tests, along with previously published bioaccumulation studies with U-spiked field and formulated sediments were used to examine the models. Observed U concentrations in Chironomus dilutus larvae exposed to U-spiked and U-contaminated sediments correlated well (r2 > 0.74, p < 0.001) with the WHAM-calculated concentration of U bound to humic acid (HA), indicating that HA may be a suitable surrogate for U binding sites (biotic ligands) in C. dilutus larvae. Subsequently, the concentration of U in C. dilutus was predicted with WHAM7 by numerically optimizing the equivalent mass of HA per gram of organism. The predicted concentrations of U in C. dilutus larvae exposed to U-spiked and U-contaminated field sediment compared well with the observed values, where one of the regression models provided a slightly better fit (mean absolute error [MAE; mg U/kg d.w.] = 18.1) than WHAM7 (MAE = 34.2). The regression model provides a predictive capacity with a minimal number of variables, while WHAM7 provides additional complementary insight into the chemical variables influencing the speciation, sorption and bioavailability of U in sediment. Our results indicate that physicochemical properties of sediment can be used to account for variability in U bioavailability as measured through bioaccumulation in chironomids exposed to U-contaminated sediments.