Funções de pedotransferência para manejo da irrigação em argissolo amarelo dos tabuleiros costeiros

Abstract

Brazil is the largest producer of sugarcane, Saccharum officinarum L. and the world's largest producer of sugar and ethanol. In recent years, the poor distribution and fluctuation of rainfall volumes have been a constant, bringing losses to agriculture. In this sense, irrigation management should be used to increase the efficiency of water use, and pedo-frequency functions (FPTs) have helped this function. The available water capacity in soils can be predicted, with good precision, from physical-water data of the soil favoring the irrigation management. The objective of this work is to generate pedotransfer functions to estimate water content in soils based on soil fraction and soil density data on soils cultivated with irrigated sugarcane in the northern forest area of Pernambuco. The experiment was carried out on lands cultivated with sugar cane located in the city of Goiana - PE. A total of 168 structured samples and 168 deformed samples were collected at depths of 0 - 0.2; 0.2 - 0.4 and 0.4 - 0.6 m. Through the structured samples, soil moisture was determined on the stress table under the potentials of - 1; - 6 and - 10 kPa, and in the Richards chamber under - 33.3 and - 1500 kPa. The deformed samples were used for particle size and particle size analysis. The data compiled were submitted to the analysis of descriptive statistics, correlations and regressions, using the SAEG statistical package. Pedotransfer functions were generated from the multiple linear regression method by applying the stepwise backward technique. The performance of the pedotransfer functions was analyzed graphically by the 1:1 ratio between the observed and predicted data using the following statistical indices: Mean error (EM), Square root mean square error (RMSE), Residual mass coefficient (CRM), correlation coefficient (r), agreement (d) and performance index (c). The results show that soil structural attributes, besides textural attributes, are fundamental for the constitution of FPTs, reinforcing the importance of the structure in the estimation of available water capacity. The use of a few independent variables facilitates the knowledge of the water storage capacity in the soil, which represents an important physical-hydro property in the study of soil water dynamics and management. The most correlated variables were microporosity, soil density and clay, for the 0 - 0.2 m layer. For the 0 - 0.6 m layer were microporosity, mesoporosity, sand and clay content.