Predicting Phosphorus Buffer Coefficients of Thai Flooded Rice Soils Using Soil Properties

Abstract

To quantify a more precise phosphorus (P) requirement prediction by the phosphorus decision support system (PDSS), the most important factor in PDSS equation, the phosphorus buffer coefficient (PBC) was studied. The use of clay content as a predictor of PBC prediction might not be suitable when applied with Thai flooded rice soils. Ten flooded rice soils were collected to study the PBC estimated by Bray-2 (PBC_Bray-2), Mehlich-1 (PBC_Mehlich-1) and iron oxide-impregnated filter paper method or Pitest (PBC_Pi-test). Multiple regression was used to study the relationship between PBC and soil properties  which involved P sorption process. The results indicated that PBC_Bray-2 correlated with organic matter and iron determined by citrate-bicarbonate-dithionite method (Fed) (adjR²=0.87) whereas PBCMehlich-1 correlated with Fe_d (adjR2=0.73). Unlike those two chemical extractions, PBC_Pi-test correlated with iron  determined by ammonium oxalate pH 3 in darkness (Fe_o), aluminum determined by citrate-bicarbonatedithionite method (Al_d) and soil pH (adjR²=0.93). However, the study suggested that the PBC should be estimated by Pi-test so that the model would allow more accurate PBC prediction.