Relações lineares multivariadas entre variáveis químicas do solo e componentes de rendimento da soja cultivada sobre área pastejada no inverno

Abstract

Soybean is of global economic importance due to its nutritional and industrial characteristics and for adapting to different latitudes, soils and climatic conditions. Therefore, new soil management practices are needed to reach satisfactory levels of yield and grain quality, which go beyond no-tillage. The integrated system of agricultural production (SIPA) has been studied and is already being adopted by several farmers in order to increase their income through the insertion of cattle in the pasture in winter, which are removed for the sowing of grain crops, in the summer. However, there is great complexity in the relationships between soil chemical variables and grain yield components in agricultural crops. Thus, the simple statistical relationships, which are bivariate relationships, are of great importance, however, they do not serve to show the exact interrelationships between groups of variables. Thus, the canonical correlation analysis makes it possible to verify which variables of one group (soil) determine the best performance of the other (grain yield). Thus, the objective of this work was to verify which are the chemical variables of the soil that provide an increase in the yield components and in the yield of soybean cultivated in an integrated crop-livestock system. The experiment was carried out on a rural property located in the municipality of Vitorino-PR in a Distroferric Red Latosol. The experiment was organized in a two-factor scheme (2 x 4) in split plots with four replications. In the main plots, two ways of sowing the soybean crop were used, being double discs and a rod-type furrower. In the subplots, four forms of soil correction/conditioning were applied: 1) without correction; 2) recommended dose of limestone (2000 kg ha-1); 3) recommended dose of gypsum (1000 kg ha-1); and, 4) limestone mixture (2000 kg ha-1) + gypsum (1000 kg ha-1). Soil chemical variables analyzed were: K, Ca, Mg, Mo, Al, H+Al, V%, S-Al, CTC, CTC EF and the components of soybean grain yield: NVP, NGV, NGP, AIV, AP, PMG, PGP and REND itself. To increase soybean grain yield, taller plants should be recommended, but with a lower insertion height of the first pod and with a greater number of grains per plant, as well as a thousand grain weight, greater amounts of calcium at depths 0- 5 cm and 10-20 cm and Magnesium at 0-5 cm depth are determinant in the increase of grain yield components and in the crop grain yield. Furthermore, soils with higher base saturation at depths of 5-10 cm and 10-20 cm are determinants of plants with lower first pod insertion height, higher plant height, higher thousand-grain weight, and higher yield. And soils with higher CEC 10-20 cm, effective CEC 5-10 cm and higher CEC 0-5 cm, as well as lower CEC 5-10 cm and effective CEC 10-20 cm are determinants of soybean plants with higher yield, lower number of grains per pod and higher grain weight per plant.