ABSTRACT

For sustainable agriculture, efficient input use is crucial. Therefore insight into the relation between inputs and outputs is needed, but the functional form and empirical estimates of production functions are still a matter of debate. Following von Liebig, it is assumed in production ecology that the production function for the single nutrient input, that is, nitrogen (N), phosphate (P) or potassium (K), can be characterized by a linear relation with a plateau (LRP) (de Wit, 1992; Paris, 1994). For any given nutrient, the threshold yield is the yield in the absence of the nutrient. As the level of nutrient is increased, yield is assumed to increase until a plateau is reached, when no further increases in yield can be obtained by additional doses of the nutrient. Of course, the level of the threshold yield, the slope of the response curve and the height of the plateau depend on the applications of other nutrients. Agronomists argue that the LRP for any one nutrient should be estimated under conditions where none of the other major nutrients is limiting (Perrin, 1976). Economists generally assume functional forms with

168 Economics of Agro-Chemicals

decreasing returns for nutrients, but recent economic research points to the superiority of LRP models (Kuhlmann, 1992; Ackello-OgutU et al., 1985; Berck and Helfand, 1990; Frank et al., 1990; Paris, 1992, 1994).