Assessing phosphorus availability in paddy soils: the importance of integrating soil tests and plant responses

Abstract Phosphorus (P) cycling in paddy soil is closely related to iron (Fe) redox wheel; its availability to rice has thus generally been ascribed to Fe minerals reductive dissolution. However, the literature aimed to identify the best method for predicting rice available P does not uniformly point to Fe reductants. Rice plants can indeed solubilize and absorb P through many strategies as a function of P supply, modifying the chemical environment. Therefore, this study aims to estimate P availability in paddy soils coupling the redox mechanisms driving P cycling with concurrent plant responses. Soil available P was estimated in three groups of paddy soils with low, medium, or high P content assessing easily desorbable pools (0.01 M calcium chloride, Olsen, Mehlich-III, anion exchanging resins) and Fe-bound P pools (EDTA, citrate-ascorbate, and oxalate). Rice P uptake and responses to P availability were assessed by a mesocosm cultivation trial. Although P released in porewater positively correlated with dissolved Fe(II), it did not with plant P uptake, and readily desorbable P pools were better availability predictors than Fe-bound pools, mainly because of the asynchrony observed between Fe reduction and plant P demand. Moreover, in low P soils, plants showed higher Fe(II) oxidation, enhanced root growth, and up-regulation of P root transporter encoding genes, plant responses being related with changes in P pools. These results indicate the generally assumed direct link between Fe reduction and rice P nutrition in paddy soils as an oversimplification, with rice P nutrition appearing as the result of a complex trade-off between soil redox dynamics, P content, and plant responses.