Norm Range Size Effects in Calculating Diagnosis and Recommendation Integrated System Indices
- H. J. Savory and
- D. L. Robinson
Calculation of Diagnosis and Recommendation Integrated System (DRIS) indices using a norm range may minimize incorrectly diagnosed deficiencies and maximize correct diagnoses of nutrient balance. This study evaluated the effects of size of the norm range on DRIS P and K overall-diagnostic and response-prediction accuracies in order to identify an optimum norm range for practical application. Field data were obtained from P and K rate experiments with Louisiana S-l white clover (Trifolium repens) on Dexter loam (fine-silty, mixed, thermic Ultic Hapludalf) and Providence silt loam (fine-silty, mixed, thermic Typic Fragiudalf) soils of adequate and relatively low, respectively, P and K levels. The DRIS indices were calculated at norm ranges of 0/3 (the norm value ±0), 4/3, 8/3,12/3, and 16/3 a and verified for correctness. The use of a norm range in the calculation of DRIS indices increased P and K overall-diagnostic accuracies as much as 23 and 27%, respectively, on the Dexter soil. for the Providence soil these values were 5 and 21%, respectively. Widest norm ranges resulted in a decrease in P and K responseprediction accuracies of 10 and 23%, respectively, on the responsive Providence soil. The importance of using a norm range for the calculation of DRIS indices was demonstrated by these results. It appears that the greatest benefit may be realized in nonresponsive to moderately responsive systems. The norm range of 8/3 a appears to be an acceptable compromise for calculation of DRIS indices at sites where the nutrients of interest are in relatively low supply. A wider norm range of 12/3 or 16/3 a was indicated at sites with adequate nutrient supplies.
Copyright © . .