Nutritional status of mango trees and soil relationships in the Sub-Middle Sao Francisco
Valley
Mangifera indica L. DRIS. CND. Multivariate analysis.
The objective of this research was to evaluate the nutritional status of the mango cultivars
Tommy Atkins, Kent and Palmer using the DRIS and CND methods, and to verify the
correlation of the DRIS indices with attributes of different soil types in the Sub-Middle region
of the Sao Francisco Valley. The study was carried out in commercial mango orchards on five
farms of the company Agrobras Agricola Tropical do Brasil Ltda. 85 mango plots were
selected, 33 of which were from Tommy Atkins, 32 from Kent and 20 from Palmer. The
database consisted of productivity and leaf contents of the nutrients N, P, K, Ca, Mg, S, B, Cu,
Fe, Mn, Zn, Mo and Ni. To separate the reference population, three criteria were used as the
database cutoff point, based on the mean productivity + standard deviation: criterion 1
(mean); criterion 2 (mean + 0.5 standard deviation); and criterion 3 (mean + standard
deviation). The high-productivity population was used to establish the DRIS standards. The
means, standard deviations and variances of each nutrient and of all possible bivariate
relationships between nutrients were calculated. All variance ratios between low and high
productivity groups were also calculated. The bivariate relationship with the highest variance
ratio was selected as the DRIS norm. After establishing the DRIS norms, the DRIS functions
and finally the DRIS indices were calculated. To use the CND method, the following were
calculated: the filling value, the geometric mean of the nutrients studied, the multivariate
relationship between nutrients, the CND norms and the CND indices. The average nutritional
balance index (IBNm) was also calculated. The DRIS and CND indices were interpreted
according to the Fertilization Response Potential (FRP). Optimal levels of each nutrient were
estimated by regression analysis between the nutrient concentration in the leaf and its
respective DRIS and CND index. The optimal ranges were established based on the optimal
content ± 2/3 of the standard deviation of the nutrient in the reference population. The chi-
square test was applied to assess whether there was an association between the DRIS and
CND diagnoses. Principal component analysis (PCA) was used to verify the relationship
between DRIS indices and soil attributes of three farms. Criterion 2 proved to be the best
method for dividing the database. Specific DRIS standards for each of the studied cultivars
were established. Optimal nutrient contents and optimal ranges, at the leaf level, were
established for each cultivar. The DRIS and CND methods showed high agreement in
nutritional diagnosis, but the DRIS method proved to be more sensitive in detecting
nutritional imbalances, since it was more correlated with the nutritional indices compared to
CND. The most limiting nutrients by deficiency were Cu, Mn, Zn and Mo, while the most
limiting nutrients by excess were Cu and Mn. The DRIS indices ICu and IK, IS, IMn and INi
showed positive and negative correlations with the P, K, Ca, Mg, Cu, Mn, Zn and T contents,
respectively, for the more clayey (Agrobras 01) and more sandy (Hidrotec) soil conditions.
Future studies should be conducted with the aim of establishing specific DRIS standards
considering not only the cultivar, but also each soil type. Experiments should focus on
validating the established DRIS standards.