Spatial variability of soil fertility attributes and productivity in a coffee crop farm

Coffee cultivation is of great importance to Brazilian agribusiness, as coffee occupies extensive production areas and is one of the most exported Brazilian products. To maintain coffee production numbers, productive techniques must be adopted that optimize productive system use. The objective of this work was to apply geostatistical techniques in the evaluation of soil fertility attributes to construct maps of variability in soil fertility parameters and the productivity of a coffee crop in the municipality of Monte Carmelo, Minas Gerais (MG), Brazil. The work was developed with coffee of the cultivar Mundo Novo 379/19, and 19 sample points were georeferenced in Universal Transverse Mercator coordinates. Spatial dependence of the fertility and productivity parameters was analysed via classic semivariogram fitting and interpolation by ordinary kriging using the statistical computer system, R. All parameters evaluated showed high degrees of spatial dependence. The attribute values varied along the sampling points, except for the sodium (Na) contents, which had similar values in all samplings. The studied parameters ranged from 80 to 200 meters. As conclusion, the use of productivity maps linked to soil chemical attributes can be useful for determining the occurrence of variable productivity rates throughout the area, allowing the adoption of corrective practices for subsequent crops and thus making the maps very useful tools for producers.