Diagnóstico da composição nutricional como ferramenta para avaliação do estado nutricional da soja em solos arenosos

Abstract

The objective of this study was to apply the Compositional Nutrient Diagnosis (CND), using the centered log-ratio (CLR) transformation of nutrient concentrations, to establish a reference population. Based on this, the study aimed to integrate foliar nutrition data and soil chemical attributes to identify factors associated with high soybean productivity cultivated in Quartzipsamments. The research was conducted in commercial soybean fields during the 2021/22 growing season, from October to March, on three farms located in the state of Mato Grosso do Sul, Brazil. The soils in the study areas were classified as Quartzipsamments, with clay content ranging from 8% to 14%. Sampling grids were implemented with spacing of 100 x 100 m, and at each sampling point, foliar and soil data were collected. The samples were classified into subpopulations based on Mahalanobis distance and the Cate-Nelson method, resulting in the categories: True Negative (TN), True Positive (TP), False Negative (FN), and False Positive (FP). This classification was validated through the evaluation of Positive Predictive Value (PPV), Negative Predictive Value (NPV), Accuracy (Acc), Sensitivity, and Specificity. CND norms were established, consisting of the centered means and coefficients of variation of nutrients, along with the calculation of the Nutrient Balance Index (NBI) based on foliar concentrations, allowing for the identification of nutritional patterns associated with different yield levels. Multivariate statistical analyses were also applied, including Multiple Discriminant Analysis (MDA), Receiver Operating Characteristic (ROC) curves, and Principal Component Analysis (PCA), in order to identify the most discriminant variables among the predicted groups, assess the nutritional status of the plants, and understand the interaction between nutrients. PCA was also applied to the soil data to investigate the factors influencing productivity. The results indicated that soybean yield in Quartzipsamment soils is strongly related to the nutritional balance of the plants, supported by soils with a pH close to 6,25 and adequate macronutrient availability. On the other hand, lower yields were associated with deficiencies of nitrogen (N), phosphorus (P), magnesium (Mg), and sulfur (S), as well as relative accumulation of micronutrients such as copper (Cu), iron (Fe), and zinc (Zn), indicating imbalance and possible stress. These findings highlight the importance of integrated soil and nutrient management to maximize soybean performance in environments with high physical and chemical vulnerability.