Relationships between phosphate retention,olsen and truog soil tests and the estimated short term phosphate fertiliser requirements of well established pasture

Abstract

Abstract Available phosphorus (QT) determined by the Truog and Olsen methods and P retention (R) were calibrated with field trials against the mean quantity of P required annually in the three years following sampling to maintain 90% of the yield obtainable with very high rates of P. The field sites fell mainly into two groups; 15 were on yellow–brown loams of the North Island with high R levels and 34 were on yellow–grey and yellow–brown earths and gley podzols in both the South and North Islands with low R levels. Most experiments with medium R values, i.e., on yellow–brown pumice soils and on mature soils derived from basalt, were excluded because of internally inconsistent results. A model in which R and QT were treated as a composite soil test accounted for 69% and 56% of the variation in P requirements for the Truog and Olsen tests respectively. Treating soils with high R as a separate group, 73% and 48% of the variation in P requirements were accounted for by the Truog and Olsen tests respectively but on soils with low R the corresponding values were only 40% and 32%. In the low R group the negative correlation between QT and P requirements was weak, partly because on many sites low QT values were associated with low P requirements. For this reason a few ‘elite’ trials rather than QT will continue to exert the major influence on P fertiliser recommendations in this group.