Rethinking the principles and practices of nutrient monitoring and field testing in horticultural crops.
In many horticultural cropping systems, leaf sampling with comparison to established critical values (CV’s) represents the primary tool for fertilizer decision-making. In the majority of crop production systems, however, the use of CV’s is inadequate as a result of flaws in both practice and interpretation. Ideally, CV’s will have been established by carefully controlled experiments in which the relationship between yield and nutrient concentration is closely followed. In the majority of crops, with the exception of the major staples and broad acre field crop, trials of this kind are very limited and hence the integrity of the CV's is questionable. The utility of the CV as a management tools is further compromised by the very significant difficulty in obtaining representative tissue samples as a result of within plant, spatial and temporal variability. Further, CV’s established on the basis of the nutrient concentration at which a species’ attains near full yield (physiological CV), provide only limited information on the response of a population of plants under field conditions. The important difference between a ‘physiological’ CV and a ‘population’ CV, is rarely considered in field practice. Managing a field so that the mean plant tissue concentration is at or just above the CV will result in up to 50% of individuals falling below critical value, while managing to ensure that 95% of all individuals exceed the critical value thus ensuring maximal field productivity, will result in over-fertilization of the majority of the field. While attempts to reduce the cost of sampling, to refine the speed or accuracy of monitoring techniques or to create fertilization practices based upon models of crop demand or nutrient budget, can all help improve efficiency incrementally, these approaches alone cannot substantially contribute to enhanced fertilizer use efficiency unless they are combined with attempts to quantify and manage field variability.