DIAGNÓSTICO E ESTIMATIVA DE PRODUÇÃO DE FORRAGEM EM ZONAS DE MANEJO ESPECÍFICO: ESTUDO DE CASO NA ESCOLA-FAZENDA DO IF GOIANO – CAMPUS CAMPOS BELOS

Abstract

Brazil stands out in cattle production and a large part of this production occurs on pasture. However, we face numerous challenges due to pasture degradation, resulting from common mistakes that make exploited areas unproductive. To reverse this situation, it is crucial to adjust pasture management, finding a balance between soil fertility, plant growth, animal consumption and animal performance. In this way, area diagnosis plays an important role in this process, providing producers and technicians with reliable information to support their decisions and implement more effective management practices. This work aimed to implement a diagnostic methodology using Machine Learning, with the purpose of generating soil fertility maps, forage production estimates and maps of land use and occupation within the scope of agricultural exploration, providing subsidies to support decision-making. The research conducted at the Escola-Fazenda of the Federal Institute Goiano – Campos Belos campus, is a case study that uses programs such as RStudio and Quantum GIS to perform geostatistical and classical statistics analyses, in addition to generating prediction maps and mapping of use and occupation from soil. The study highlights the relevance of an accurate area diagnosis for decision-making. Given the results, the organization starts to consider management techniques that aim to generate savings for producers and technicians interested in sustainable production alternatives and increased productivity. Through the use of information represented on maps with specific management zones and image fusion, it is possible to identify critical zones and concentrate efforts effectively and locally to solve problems, thus strengthening cattle farming in a sustainable way and increasing the productivity of areas of management pastures.