Development of a methodology to determine the best grid sampling in precision coffee growing

Abstract

Precision agriculture is based on a set of techniques that explore the spatial variability of properties related to a determined area. The aim of this study was to develop and test a methodology to evaluate the quality of grid sampling. The experiment was performed in three areas of 112, 50 and 26 ha, in coffee plantations (Coffea arabica) with cultivar Catuai 144, in the Três Pontas Farm, located in Presidente Olegário, MG, Brazil, in 2014 and 2015. A total of 224, 100, and 52 georeferenced points (2.0 points/ha) were plotted in the areas regarding the soil chemical properties, respectively: phosphorus, potassium, calcium and magnesium. For the application methodology the standardized accuracy index (SAI), the standardized precision index (SPI) and the standardized optimal grid indicator (SOGI) were developed and tested. From grid 1 (2 points/ha), another three sampling grids (1.0, 0.7 and 0.5 point/ha) were adopted. The indexes were important to analyze the grid quality, whereas the SOGI allowed selecting the grid that best represented the properties.