Pot and field experiments were conducted to investigate the effects of water regimes on the speciation and accumulation of arsenic（As） and cadmium（Cd） in Brazilian upland rice growing in soils polluted with both As and Cd. In the pot experiment constant and intermittent flooding treatments gave 3–16 times higher As concentrations in soil solution than did aerobic conditions but Cd showed the opposite trend. Compared to arsenate, there were more marked changes in the arsenite concentrations in the soil solution as water management shifted, and therefore arsenite concentrations dominated the As speciation and bioavailability in the soil. In the field experiment As concentrations in the rice grains increased from 0.14 to 0.21 mg/kg while Cd concentrations decreased from 0.21 to 0.02 mg/kg with increasing irrigation ranging from aerobic to constantly flooding conditions. Among the various water regimes the conventional irrigation treatment produced the highest rice grain yield of 6.29 tons/ha. The As speciation analysis reveals that the accumulation of dimethylarsinic acid（from 11.3% to 61.7%） made a greater contribution to the increase in total As in brown rice in the intermittent and constant flooding treatments compared to the intermittent-aerobic treatment. Thus, water management exerted opposite effects on Cd and As speciation and bioavailability in the soil and consequently on their accumulation in the upland rice. Special care is required when irrigation regime methods are employed to mitigate the accumulation of metal（loid）s in the grain of rice grown in soils polluted with both As and Cd.
Highly precise and reliable determination of heavy metals in soil micro-arthropod tissues remains a challenge because of the small size of the animals and their typical low abundance in metal-contaminated agricultural soils. The present study sought to develop a method for cadmium （Cd） determination in soil micro-arthropods by optimizing the sample digestion procedure, reducing sample weight, modifying sample pre-treatment and validating the methodology with field samples. The optimized digestion conditions comprised a sample mass of 50-150 μg, digestion reagent of nitric acid：hydrogen peroxide （3：1）, digestion temperature of 105 ℃, digestion period of 3 h and digestion volume of 30 μL. Defecation of the standard Collembola Folsomia candida （92 h） and the indigenous Collembola Onychiurus yodai （42 h） and ultrasonic cleaning of F. candida increased the accuracy of Cd determination. The recovery of Cd using the refined procedure was 98.9% and the limits of detection and quantification were 0.002 and 0.008 μg L-1, respectively. The within-batch precision values were 〈 3%. The Cd concentrations in the tissues of the Collembola Isotorna sp. collected from a range of metal-contaminated fields determined by the improved method were consistent with the Cd concentrations in the field soils. The results indicate that the optimized method can be used for more accurate or reliable determination of Cd concentrations in soil micro-arthropod tissues.
