Funções de pedotransferência para estimativa de retenção de água em solos da mesorregião do agreste paraibano

• Autor(a) principal: Pequeno, Petrus Luiz de Luna
• Data de Publicação: 2016
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFPB
• Texto Completo: https://repositorio.ufpb.br/jspui/handle/123456789/25544
• topic: Solos representativos; Modelagem; Atributos físicos; CNPQ::CIENCIAS AGRARIAS::AGRONOMIA::CIENCIA DO SOLO

Knowledge of the hydraulic properties of the soil is important for the establishment of good agricultural practices. The soil water retention estimated from mathematical models generated for different edafoclimatics conditions has contributed to high errors in estimates. The research aimed to characterize, in six soils of Mesoregion of the Paraíba state, physical attributes, and, additionally, attributes chemical with influence on the soil water retention, as also to create single and unified database and through them establish for to test and validate pedotransfer functions (FPT) able to estimate water retention in these soils on the 0-20 cm depth. Were studied classes Ferrasol, Lixisol, Acrisol, Arenosol, Leptosol and Planosol. In the first chapter in each soil class were undisturbed soil samples collected to study the physical attributes total porosity, macro and microporosity, aggregate stability, bulk density, water retention curve tensions -0.006, -0.01, -0.033; -0.1; -0.3; -0.5; -1.0; -1.5 MPa and hydraulic conductivity saturated, and deformed samples for determination of particle size attributes, water dispersible clay, flocculation, particle density, addition of chemical soil parameters K+ , Na+ , Ca+2, Mg+2, Al+3 and total organic carbon. On the second chapter, with the data obtained from the physical and chemical attributes formed a individual database for each soil class and a unified database (with all data of the six soil classes). The variables components of each FPT were selected by stepwise method and had their respective predictive efficiencies analyzed by the statistical indicators: coefficient of determination (R²), average error of estimate (ME) and mean square error (RMSE). In those databases were tested other functions/models existing in literature and certain under different conditions. In the first research, the sand fraction was predominant in all soils, where the aggregate stability, represented by the IEA, was shown to be dependent on the contents clay and organic matter, besides suffering influence of Ca and Mg. The structure and texture of soil influenciated the retention of water in the different tensions. Was verified that the -0,033 MPa tension isn't representative of the to field capacity point. The potentials -0.006 and -0.01 MPa represented better the soil field capacity and contributed to a larger volume of water available in all soils. In the second research, there was a positive influence of the structural parameters in the accuracy of the FTP for all studied soil classes and especially in potential higher matrix; the heterogeneity resulting from the union of the individual databases generated in the composition of FPT gave positive impact in improving the accuracy and predictive ability of all FPT in all matric potentials; the pedotransfer functions obtained from the literature showed poor predictive ability for the soil and climatic conditions of Mesoregion of the Agreste Paraibano.