Abstract
Trials of simple design, dispersed through the area in which fertiliser recommendations are to be given, and conducted for several consecutive years, are needed to define fertiliser requirements of up-graded permanent pasture. These requirements should be studied through trials with physiologically balanced fertiliser mixtures based on superphosphate. Given monetary values for fertiliser and for pasture, optimum rates at which such mixtures should be applied in practice may be calculated through production function analyses of trial data. The results of such calculation show the effects of climate and of variation in pasture value on the amount of superphosphate constituting an optimum rate. It is also shown that until seasonal growing conditions can be predicted, an optimum rate calculated from pasture response averaged over several years is the soundest basis for fertiliser recommendations.