This work was carried out to investigate the multivariable statistics/cluster analysis as a tool to identify the similarity of irrigated fields. The investigation was conducted in three areas. One was an undisturbed native area (MN) and the others, two irrigated field in the Irrigated District of Jaguaribe-Apodi (DIJA) situated in the municipality of Limoeiro do Norte and Quixeré, Ceará State, Brazil. Soil was sampled monthly, from December/1999 to December/2000, and also from September to December of 2001 in the depths of 0-30, 30-60 cm. The attributes considered in this study were: EC (electric conductivity of soil solution), Ca2+ + Mg2+, Cl-, Na+ and k+ and SAR (sodium adsorption ratio). The Cluster Analysis application identified the similarity of studied attributes and three homogeneous groups were identified. The studied attributes values from MN were statistically different (p < 0.001) from those of irrigated fields. According to the results, Cluster Analysis showed up as a suitable tool to define the similarity independent of sample position in time and space. This fact expresses that Cluster Analysis technique can be applied to identify the similarity in areas with high or little salinization risk.
salinity; semiarid; cluster analysis; irrigation-impact
