Multispectral remote sensing in the identification and mapping of biotic and abiotic coffee tree variables

ABSTRACT Multispectral remote sensing is a reliable and feasible methodology to assist farmers in decision making for best management practices, ensuring a more efficient and sustainable agricultural production. The objective of this study was to identify and map stress on coffee caused by biotic and abiotic variables through vegetation indices derived from Landsat-5 Thematic Mapper (TM) multispectral images. The sampling grid was composed of 67 points, with each sampling point consisting of five plants. The analyzes of the incidence of brown eye spot and infestation of the leaf miner in the leaves, pH, organic matter, soil texture and nutrients leaf contents were performed at each of the sampling points and correlated with 16 vegetation indices obtained from images at the time of analysis. The vegetation indices presented a spatial distribution similar to the agronomic variables in the crop. There was a positive correlation of the indices with infestation of the leaf miner, silt and clay content in the soil and concentration of Mg, Cu, B and Mn in the leaves, and negative with the incidence of brown eye spot, pH and soil sand content. Based on these results, it was possible to map and identify the changes in the spectral reflectance of the coffee trees, caused by these agronomic variables.