The objective of this work was to evaluate the effect of selection criteria for reference populations on establishing DRIS norms for cotton crop. A database was created with data on yield and macro- and micronutrient contents in leaves obtained in average 100 ha plots. The criteria for establishing the high yield subpopulations were plots with: yield above average (4,380 kg ha-1); yield above average + 2/3 standard deviation (above 4,650 kg ha-1); yield above average + 1 standard deviation (above 4,785 kg ha-1); and yield above average + 4/3 standard deviation (above 4,920 kg ha-1). The selection criteria for the reference population provide low agreement frequency between nutritional relations, for each evaluated DRIS norm. Independently of the criterion used, data on average yield and on the nutritional balance index are significantly related; and this relation is intensified with the increase in the rigor of criterion for the inclusion of crops in the reference population.
Gossypium hirsutum; high productivity; nutritional status; nutritional balance index; mineral nutrition; diagnosis integrated system
