ESTIMATING EFFECTIVE LEVELS OF PHOSPHORUS APPLIED AS SUPERPHOSPHATE AND ROCK PHOSPHATE

The effectiveness of different phosphates can be measured in field exp eriments in two ways: (i) different levels of each fertilizer are appl ied once only at the start of the experiment and the effectiveness of the fertilizer residues is measured in subsequent years relative to th e effectiveness of freshly-applied fertilizer; (ii) different levels o f each fertilizer are applied so that the same level is applied cumula tively to the same plot each year. Both methods can be used to compare rock phosphates with superphosphate. The first method can be used to measure the residual value in kg ha-1 of each fertilizer from the rela tionship between yield and the level of P applied. The residual value is calculated by dividing the amount of P as freshly-applied superphos phate needed to produce a given yield by the amount of P from the fert ilizer residues required to produce the same yield. This ratio is then used to estimate the proportion of the P applied, as kg P ha-1, that is effective for plant growth each year. Provided the experiment is bi g enough to have sufficient plots for the freshly-applied superphospha te, the residual value of rock phosphates applied in any number of yea rs can be estimated. While this method allows an estimate of how the a vailabilty of the residues from any one application changes with time, it gives no information on whether there is any additive benefit from residues from cumulative applications. The second method more closely resembles fertilizer practice on farms but it can not be used to meas ure the residual value of the fertilizers because of the problem of kn owing what to plot on the x axis, the amount of P applied. It could be kg P ha-1 applied each year, or kg total P ha-1 applied up to that ye ar. This problem may be resolved by using the first method to calculat e the residual value of each individual application of fertilizer and adding the individual residual values, in kg ha-1, for the appropriate number of years. That is, the sum of the residual values measured usi ng the first method could be used to determine the ''effective'' level of P for the second method. This approach was tested in a field exper iment in Western Australia that measured the residual value of two roc k phosphates and superphosphate. When the residual value measured usin g the first method was used to calculate the 'effective' level of P ap plied as each P fertilizer, the relationship between yield and the 'ef fective' level of P applied could be described by the same equation fo r the three P fertilizers for P applied at the start of the experiment only or cumulatively each year.