Data from 135 field experiments on four major New Zealand soil groups have been used to calibrate a slightly modified Olsen bicarbonate phosphorus (P) soil test against the response of clover #x2014; grass pasture to phosphate fertiliser. The calibration has used the approach described by Saunders et al. (1987) whereby probable minimum yields represented by the 10th, 20th, and 50th percentiles of relative yield on each soil test value are related to the Olsen P values. These percentile curves are used to indicate the probability of a required yield being obtained if phosphate fertiliser was not used. The normal regression approach left too much error for it to be a practical approach in calibrating the Olsen P soil test for fertiliser advice. The interpretation of the Olsen P values differed markedly between soil groups and between the slow and fast growth periods of the year. Pastures were less responsive to phosphate fertiliser in the fast growth period of late spring and early summer, than over the slow growth period #x2014; the remainder of the year. Differentiating between the two periods of the year enables phosphate fertilisers to be used more effectively. This method of calibration of the Olsen P soil test and the calibrations for soil groups are incorporated in the New Zealand Ministry of Agriculture and Fisheries Computerised Fertiliser Recommendation Scheme.
展开▼
