Prediction of soil solution concentrations and leaching losses of cadmium in agricultural soils

Cadmium (Cd) leaching is often estimated in Cd balance models using the product of drainage water (precipitation excess) and soil solution Cd concentration. However, Cd concentrations are seldom available but rather predicted using empirical models. Despite the availability of empirical models that estimate soil solution Cd concentrations, they have limitations for use in New Zealand where total Cd in agricultural soils is low and organic matter (OM) high. This study derived a Freundlich-type model from desorption data and a soil-liquid partitioning coefficient (K_D ) model based on sorption data to predict soil solution Cd concentrations from commonly measured soil parameters that could be used to calculate Cd leaching fluxes. Independent soil solution Cd concentrations and Cd fluxes measured in drainage water from field trials were used to validate the predictive capacity of the models. It was found that soil pH and OM content were the most important factors controlling soil solution Cd, along with total Cd. Both models explained 83% of the variation in measured soil solution Cd concentrations in an independent dataset. Comparisons between Cd fluxes predicted using the Freundlich-type model and measured fluxes were within 25% of each other at 6 of 19 field sites studied. However, physical and chemical nonequilibrium conditions in soils and uncertainty in measured values likely contributed to differences between predicted and measured Cd fluxes at other sites. To unravel the impact of nonequilibrium and soil physical conditions on Cd concentrations in drainage water, more data are required on Cd concentrations collected under field conditions. This will allow better validation of the approach used in Cd balance models to calculate Cd leaching from soils.