The diagnosis and recommendation integrated system (DRIS) for diagnosing the nutrient status of grassland swards: II. Model calibration and validation

Abstract

A DRIS model for perennial ryegrass, based on data collected from a glasshouse experiment, was used to diagnose the nutrient sufficiency status of perennial ryegrass swards growing in field situations. Initially, DRIS overestimated the N and P status of these swards and underestimated their S status. However, by calibrating the model outputs on the basis of actual sward dry matter yield responses to specific fertiliser inputs, correction factors were evaluated, and the nutrient indices modified to reflect the nutritional status of swards in field situations. Modified DRIS diagnoses of the N, P, K and S status of swards were compared with those made using the critical value approach. DRIS proved to be as reliable as the critical value approach at diagnosing N deficiency (both approaches having reliability scorings of 90%), but was superior to the latter at diagnosing P, K and S deficiencies, having reliability scorings of 100% (P), 90% (K) and 70% (S), compared with scorings of only 0% (P), 80% (K) and 30% (S), for the critical value approach.