Several changes in the methodology of DRIS indexes calculation were proposed in order to increase the accuracy in the nutritional diagnosis for several crops. The calculation of the functions or standard deviation units can be defined by the methodology originally developed by Beaufils (1973), Jones (1981) or Elwali & Gascho (1984), there are some conflicting results in the literature regarding the effectiveness of each method of calculation. According to Mourao Filho (2004), there is still no clear definition of what would be the best recommendation to calculate the functions or standard deviation units for the DRIS.

According to Serra (2011), the use of the methodology proposed by Jones (1981) when compared with Beaufils (1973) and Elwali & Gascho (1987) showed better efficacy on DRIS index for cotton crop (Gossypium hirsutum r latifolium). The measure of the efficacy used by Serra et al. (2011) was the relation between yield and nutritional balance index (NBI).

Beaufils (1973):

For A/B < a/b;

a/b] 100- A/b ‘ CV%

f(A/B)=0, for A/B = a/b For A/B > a/b;

Jones (1981):

Elwali & Gascho (1984): For A/B < a/b-1s

a/b~ 100 •K

ш’ cv%

f(A/B)=0, to the range between a/b-1s to a/b+1s For A/B > a/b+1s

After defining the functions DRIS, the DRIS index is calculated and for each nutrient a DRIS index is determined, which may have positive or negative values, that represent the arithmetic average of functions in which the nutrient is involved, when the result is negative
(below zero), this means deficiency and when the positive value indicates excess, as proposed by Beaufils (1973):

е/Й)-е/(5

DRIS Index A =—— ———— —

n

n=number of DRIS functions of each dual ratio defined by criteria of chosen of the norms, in that the A nutrient is involved.

The sum of DRIS index in module of the nutrients in a sample diagnosed, generates the nutritional balance index (NBI), in an increasing scale, the higher NBI the greater nutritional imbalance in the plant and consequently low productivity, and the correlation between NBI and yield is considered one measure of the effectiveness of the system DRIS (Beaufils, 1973; Nachtigall & Dechen, 2007b; Guindani et al., 2009).

Mourao Filho (2004) concludes that researches on DRIS are still incipient, therefore, many accurately factors must still be better studied, factors such as the criteria for choosing the reference populations, the combination of methods to be used, so there is a need to more refined studies on these aspects.