Funções de pedotransferência para estimar a umidade em solos cultivados com cana-de-açúcar na zona da mata de Pernambuco

Abstract

This work is related to knowledge of Pedotransfer functions (PTF), enabling an understanding about water soil content influenced by soil properties and analyzes to generate PTFs. This study aimed was proposed pedotransfer functions (PTF) based on hydro-physical soil properties and organic carbon to estimate soil moisture. The study was conducted in the area owned by Santa Teresa Plant in Goiana City, PE (7º36'56 "S 34º59'19" W) with sugarcane crop (Saccharum spp) irrigated by Lateral Move System. Field capacity (θfc) and wilting point (θwp) were assumed by dependent variables. Sand, silt, clay, soil bulk density and organic carbon were adopted as independent variables. Pedotransfer functions generated in this study estimate soil moisture to matric potential at -10 kPa, equivalent to θfc, in sandy soils and at -33 kPa to clay soils and matric potential at -1.500 kPa to θwp in 0-20; 20-40 e 40-60cm soil depth. These functions were classified as “Point PTF”, since estimated specific points to matric potential to soil water retention. Statistical indices and graphic analyses (1:1 ratio) between model-predicted and observed data were used to evaluate PTFs performance. The following statistical indices were used: Correlation coefficient (r), Concordance index (d), Maximum error (ME), Square root mean squared error normalized (RMSE), Coefficient of residual mass (CRM), Mean absolute error (MAE), Performance Index (c) and Determination coefficient (R²). Statistical indices that evaluate generated models to estimate soil water content indicated high performance for most of PTFs and good correlation between observed and estimated values at matric potential studied in all soil layers analyzed. Thus, retention water estimated by PTFs can be an interesting alternative to support irrigation management at sugarcane.