Soil sampling depth and phosphorus extraction method for phosphorus in leguminous pastures

Abstract

The current guidelines and recommendations for P fertilization in Uruguay can be improved by adjusting the sampling depth and analytical methods of testing soil P in pastures with different fertilization history. A set of field trials was conducted between 2008 and 2012 on 14 sites of Uruguay for the purpose of comparing two sampling depths and three methods for assessing P availability by their correlation with annual dry matter yield response. The trials had a randomized complete block design and were sown with Trifolium repens or Lotus corniculatus. Soil P availability was determined by sampling 0-7.5 and 0-15 cm depth using three analytical methods: Bray I, cationic resins, and citric acid. Rock phosphate and triple superphosphate were applied at five P rates. (0-240 kg kg P2O5 ha-1). Annual forage yield was recorded. Relative yield was calculated as measured yield/maximum yield observed within a block. The correlation between relative yield and soil P availability was studied using the modified arcsine-logarithm calibration curve (ALCC) and analyzing the resulting correlation coefficient, root-mean-square error (RMSE) of the soil P test, and RMSE of the relative yield. The 0-7.5 cm soil sampling depth did not show a better fit than the deeper sampling (0-15 cm), with the latter having less variation in soil P test values. When rock phosphate was used, the correlation coefficients were 0.50 for cationic resins, 0.53 for citric acid, and 0.38 for Bray I. When triple superphosphate was used, the correlation coefficients were 0.37 for cationic resins, 0.44 for citric acid and 0.43 for Bray I. For both P sources, the citric acid method with the sampling of 0-15 cm soil depth is the preferred soil P test method for management of P in leguminous pastures for the soils studied.