Prognóstico da toxidez por ferro em arroz irrigado a partir da análise química do solo

Abstract

The iron toxicity is one of the most important abiotic stress in lowland soils. After the appearance of toxicity symptoms in the field, only palliative measures can be taken to lessen the effect of the problem. The aims of this work were to verify if the interpretation criteria for predicting the risk of iron toxicity occurrence in rice plants by flooding based on the percentage saturation of CEC by Fe2+ (PSFe2+), which is calculated from the estimated exchangeable iron from the determination of iron extracted by ammonium oxalate pH 6.0 in a sample collected prior to flooding, are valid for a group of lowland soils from of Rio Grande do Sul. Furthermore, if the inclusion of other variables improves the estimation and prognosis, and if the relation of the classes of risks is consistent with the occurrence of toxicity in plants. For this, four experiments were conducted with plants, using different soils in a greenhouse and an experiment in nutrient solution in laboratory. The results of the experiment in nutrient solution have shown that the emergence of symptoms of iron toxicity in rice plants occurred only for the 0.45 and 0.60 molar ratio of iron to divalent cations, different from the 0.40 molar ratio of iron in solution defined as critical to the toxicity by the method. However, no change was detected for the iron levels on plant tissue when varying the molar ratio of iron in the nutrient solution. With the results of experiments with plants in soil one could verify that the higher the PSFe2+ of a soil sample the greater the probability of toxicity by iron, and for samples with low PSFe2+ almost no symptoms of toxicity occurred, indicating that the method was efficient in predicting the occurrence of iron toxicity symptoms in rice plants, for the group of soils. It was also found that the use of the variables: organic C, iron and manganese extracted by ammonium oxalate pH 6 improve the accuracy of the exchangeable iron estimation, when compared to estimating only by the iron extraction. However, it was not reflected in better prognostic of the risk of iron toxicity occurrence in rice plants, for the group of soils. It was observed that the correlations with the two forms of estimating exchangeable iron-related attributes of toxic iron coefficients were very close, making impossible the definition of the best method. By the small number of variables to be determined, ie, by the easiness of implementation, the estimation of exchangeable iron obtained only by the iron extracted by ammonium oxalate pH 6.0 is the most efficient method to predict the risk of iron toxicity for rice crop.