Soybean Phosphorus and Potassium Requirement Evaluation by Three M-DRIS Data Bases
- W. B. Hallmark ,
- R. B. Beverly,
- M. E. Sumner,
- C. J. deMooy,
- H. F. Morris,
- John Pesek and
- J. D. Fontenot
- I beria Res. Stn., Louisiana State Univ. Agric. Ctr., P.O. Box 466, Jeanerette, LA 70544
H orticulture Dep., Univ. Georgia, Griffin, GA 30212
D ep. Agronomy, Univ. Georgia, Athens, GA 30602
D ep. Agronomy, Colorado State Univ., Fort Collins, CO 80532
F eed Pert. Lab., Louisiana State Univ. Agric. Ctr., Baton Rouge, LA 70803
D ep. Agronomy, Iowa State Univ., Ames, IA 50011
Modified-diagnosis and recommendation integrated system (MDRIS) soybean (Glycine max L.) norms are now available for select, large, and regional data bases. It is unclear which of these data bases is more accurate in evaluating nutritional needs of soybean. A study was conducted to compare the ability of select, narrow-based (S); large, broad-based (L); Midwest (MW); and Southeast (SE) MDRIS data bases to diagnose P and K nutrition in four soybean cultivars grown in a limestone by a P by K soil fertility study in Iowa. Greater accuracy by L, MW, and SE vs. S in evaluating plant P and K status was attributed to spuriously high P and K concentrations in S, which resulted in excessive P and K deficiency diagnoses. Diagnostic accuracy for L and MW did not differ. Greater accuracy of MW vs. SE in detecting P deficiencies was attributed to the larger norm mean for P. The SE performed better than L or MW in determining K nondeficiencies, resulting in superior overall K diagnoses and overall P plus K diagnoses. This was attributed to relatively larger CVs for norms involving K (vs. the average CV for nutrient expressions involving DM) in the SE data base compared to those in the L and MW data bases, which resulted in less incorrect diagnoses by SE where K was nondeficient. Diagnoses by MW were not superior to those for L and SE; therefore, data do not support the suggestion that regional data bases will improve soybean diagnoses for P iihd K. Data do support the assumption that MDRIS norms from large, broad-based (L) data sets are more accurate than norms from select, narrow-based (S) data sets in evaluating plant nutrient requirements.
Copyright © . .