Geração de mapas de produtividade e sua utilização na geração de zonas de manejo

Abstract

The need to increase agricultural production while reducing environmental impact has driven the adoption of technologies such as Precision Agriculture (PA). This approach aims to provide inputs more precisely, improving efficiency and reducing costs. However, its implementation poses a challenge for small farmers due to the high investment required in variable-rate application equipment. A more financially accessible solution is to divide agricultural areas into management zones (MZs), allowing for more specific management than traditional agriculture but without high investments. In this context, monitoring productivity through thematic maps is essential, as it enables the identification of productivity at each location within the field and its correlation with soil and plant attributes. This helps identify the causes of productivity level variations and allows for the implementation of appropriate management practices for each zone. To carry out these procedures, the use of specialized software for efficient farm management is essential. This study aims to filter and generate thematic maps from productivity data collected over 11 harvests in five agricultural areas, define management zones, and evaluate them using specific indices. At the end of the study, it was verified that data filtering was essential to improve the quality of the information and reduce inconsistencies in yield maps. The interpolation methods adequately represented spatial variability, while the Fuzzy C-Means algorithm enabled the delineation of management zones with spatial coherence and statistically significant differences among them. The evaluation indices indicated good performance of the generated zones, demonstrating that the methodology is effective as a decision-support tool, contributing to input optimization, increased productivity, and the sustainability of agricultural systems.