The influence of seasonal conditions,plant species and fertilizer type on the prediction of plant yield using the Colwell bicarbonate soil test for phosphate

Testing for soil phosphate (P) using the Colwell procedure is widely used in south-western Australia to estimate fertilizer applications required for crops and pastures. The relationship between plant yield, expressed as a percentage of the maximum yield, and soil test values is assumed to be constant in different years for the same soil type and plant species. Data from 11 long-term field experiments in south-western Australia show that regardless of whether percentage of maximum or absolute yield is used, the relationship between yield and soil test values is different (1)in different years, for the same site and where the same P fertilizer type has been used. This occurred irrespective of whether the same or different plant species were grown in different years; (2)where different types of P fertilizer had been used, for the same site, same year and same plant species; (3)for different plant species, for the same site, same year, and same type of P fertilizer. We conclude that considerable errors in the recommendation of fertilizer rates may result from the assumption that there is a constant relationship between soil test and yield.