Variabilidade espaço-temporal da pluviometria no semiárido brasileiro e sua relação com a produtividade do milho.

Abstract

The productivity of various agricultural crops is strongly dependent on the rainfall supply, as well as on its frequency and intensity. The objective of this work was to evaluate the spatio-temporal behavior of corn rain and yield in the Brazilian semi-arid region (BSA). The Principal Component Analysis (PCA), Grouping Analysis (AA) and Mann-Kendall tests were used to identify trends of the variables for the period from 1990 to 2014. The results showed that there is no uniformity in the distribution of precipitation, given that several factors, whether the performance of atmospheric systems or geographic position influence the region's total rainfall. In most of the BSA territory maize productivity is less than 1000 kg / ha. AA allowed the BSA region to be divided into 4 groups. Groups 1 and 4 presented a small negative trend in precipitation, but with positive values in maize productivity trend. Groups 2 and 3 presented positive trends in both rain and corn yield. The average maize yield in the last decades has been varied and is closely related to climatic variability. The effect of rainfall on maize productivity was statistically positive, that is, in most of the BSA region, grain yield efficiency is dependent on local rain.