Papers

Summary

Predicting and mapping productivity areas allows crop producers to improve their planning of agricultural activities. The primary aims of this work were the identification and mapping of specific management areas allowing coffee bean quality to be predicted from soil attributes and their relationships to relief. The study area was located in the Southeast of the Minas Gerais state, Brazil. A grid containing a total of 145 uniformly spaced nodes 50 m apart was established over an area of 31.7 ha from which samples were collected at depths of 0.00–0.20 m in order to determine physical and chemical attributes of the soil. These data were analysed in conjunction with plant attributes including production, proportion of beans retained by different sieves and drink quality. The results of principal component analysis (PCA) in combination with geostatistical data showed the attributes clay content and available iron to be the best choices for identifying four crop production environments. Environment A, which exhibited high clay and available iron contents, and low pH and base saturation, was that providing the highest yield (30.4l ha−1) and best coffee beverage quality (61 sacks ha−1). Based on the results, we believe that multivariate analysis, geostatistics and the soil–relief relationships contained in the digital elevation model (DEM) can be effectively used in combination for the hybrid mapping of areas of varying suitability for coffee production.