Suscetibilidade e predição da compactação de solos cultivados com cana-de-açúcar em Pernambuco

Abstract

The intensification of agricultural traffic from planting to several harvests in the cultivation of sugarcane promotes different changes in the physical quality of the soils due to the excessive stresses applied to the soil surface, compromising the agricultural yield and the longevity of the crops. Confined endometrial uniaxial tests make it possible to evaluate and monitor the mechanical-mechanical degradation of soils promoted by intense agricultural traffic. The objective of this study was to evaluate the susceptibility to soil compaction with different textural and management classes cultivated with sugarcane in Pernambuco and to propose mathematical models for the prediction of compaction. Deformed and undisturbed soil samples were collected in nine locations at the depths of 0.00-0.20 m and 0.20-0.40 m under two managements (cane plant and cane soca). The soils were classified as: Yellow Utisol, Spodosol, Entisol and Yellow Latosol. Organic carbon (OC), total porosity (TP), moisture in field capacity (W), silt + clay content (S+A), degree of saturation (DS), degree of compaction (DC), recompression index (RI), compression index (CI) and pre-consolidation stress (σp) were submitted to principal component analysis, in a multivariate approach. Additionally, they were submitted to multiple linear regression analysis considering σp and CI as dependent variables. The significance of the explanatory variables on the dependent variables was considered using the F-test (p-value <0.05). The successive traffic did not interfere in the behavior of the variables, but the S+A levels influenced it. Soils with high levels of fines showed greater load capacity, however, they were more susceptible to the compaction process, especially with high DS. The prediction models of the load capacity and susceptibility of the soil to general compaction and grouped according to the levels of S + A showed acceptable coefficients of determination (≥ 60%). In general, p and CI can be estimated by linear regression models using easily measured soil attributes for a wide range of texture, CO and management.