Analysis of field-moist Cd contaminated paddy soils during rice grain fill allows reliable prediction of grain Cd levels

Research undertaken over the last 40 years has confirmed that the long-term consumption of cadmium (Cd) contaminated rice contributes to human Cd disease. Rice is the staple of millions throughout South and Southeast Asia. Therefore, the ability to accurately assess the risk of rice grain Cd uptake in areas of elevated soil Cd would be a prerequisite to protecting public health and regional export security. During 2001-2002, 308 concomitant soil and rice grain samples were collected from a Cd/Zn co-contaminated site in Western Thailand and determined for aqua regia digested soil Cd and rice grain Cd. No significant relationship was observed between total soil Cd and rice grain Cd (r(2)=0.117). This intuitively is to be expected since total soil Cd bears no relationship to phyto-available Cd. Similarly no relationship was observed between 0.005 M DTPA extractable soil (air-dry) Cd and rice grain Cd (r(2)=0.165). Again this result could have been predicted as the phyto-availability of Cd in paddy soils is a function of the complex interaction between soil pH, redox conditions and the presence of competing ions. Consequently, in 2003 a further study was undertaken to assess the effectiveness of commonly utilized soil extractants namely, 0.1, 0.05 and 0.01 M CaCl2 solutions at a soil extractant ratio of 1: 5 and 1 M NH4NO3 for 2 h or 4 h extractions times at a soil/extractant ratio of 1:2.5. Soil samples were collected at the critical rice grain fill stage and sub-divided into Portion A which was subjected to conventional air-drying and sample preparation procedures and Portion B which was maintained at Field Condition (FC) and stored at <4 degrees C until extractions were undertaken. Concomitant rice grain samples were collected at maturity. The results indicate that air-dried soil samples subjected to conventional soil preparation procedures were totally ineffective at predicting the uptake of Cd by rice stem, leaf or grain, regardless of extractant. Further, the results indicate that the Stepwise Regression model incorporating 0.1 M CaCl2 extractable Cd and soil pH(w) determined on field moist samples accounts for 63.8% of the variability in rice grain Cd.