The crop model record: promise or poor show?

At a symposium held on the occasion of the 50th anniversary of agricultural research in Israel, Professor C.T. de Wit gave a survey of achievements in worldwide agricultural research. He maintained that 500 years was a more appropriate period to review, because the last major contribution was Liebig's Chemistry in its Application to Agriculture and Physiology published in 1840. Many would contest this thesis, but whatever other minor achievements there may have been in the interim, can Theoretical Production Ecology be relegated to the same bleak category? Or should we regard de Wit's contribution as the remodelling and development of an age-old discipline traceable to Joseph's long-term yield predictions that were based on esoteric theory and flimsy data but were successfully applied to guide strategic food-security planning? More recently in 1735, Reaumur had the idea of relating day-degrees to phenological development and so conceived one of the more robust 'summary models' (or 'conservative relations' sensu Monteith, Chapter 1) that lives on to this day (Aim et al., 1988; Hesketh et al., 1988). Within this ancient discipline, the year 1958 could mark the beginning of the modern era (if not the revival) of theoretical production ecology when de Wit, in Transpiration and Crop Yields first defined the now well-known 'conservative relationship' underlying the mass of empirical data accumulated by Briggs & Shantz (1913) and others. This was soon followed by Photosynthesis of Leaf Canopies in 1965 which ushered in the computer as the instrument for simulating crop growth. In the following years, crop models proliferated in a worldwide endeavour to describe the growth processes and explain the behaviour and yield potential of crops. At first, the motivation was probably scientific curiosity and a desire to exploit the possibilities offered by the modern computer. This was soon followed by the expectation (or rationalization?) that comprehensive explanatory analysis of growth processes would contribute to better research, plant breeding, crop