Limitations to the diagnostic information obtained from soil analyses

The limitations of various methods of soil analysis are discussed. The total analysis of an element gives a simple figure which is, however, useless for diagnositic purposes since nutrient availability differs in various soil fractions. Since the transfer coefficient applied to soil pollutants is based on total soil analysis, it also has no diagnostic value. The use of soil extractants is empirical and is only as good as the calibration curve. It should, therefore, only be applied to those soils for which a good calibration curve has been shown to exist. The scientific approach of Schofield in proposing nutrient potentials is basically correct, but it fails to fully characterize the nutrient availability in soils. Difficulties are of a fundamental nature or are soil or plant related.A possible means of improvement may be a soil classification which is geared towards soil fertility. Soils which react similarly to nutrients should be combined in classification units. For these units, detailed information on the reaction behavior must be obtained. A few simple analyses might suffice to show how a specific field fits into this reaction pattern. Based on the thorough knowledge of the soil (especially its reaction pattern) it is then possible to quantitatively predict for a specific crop plant the changes that occur when nutrients are removed by uptake or added by fertilization.