Imagens multiespectrais orbitais: ferramenta para discriminar manejos de fitonematoides do cafeeiro

• Autor(a) principal: Orlando, Vinicius Silva Werneck
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFU
• Texto Completo: https://repositorio.ufu.br/handle/123456789/31129; http://doi.org/10.14393/ufu.di.2020.845
• topic: Sensoriamento Remoto; Remote Sensing; Controle Biológico; Biological Control; Bacillus spp.; Bacillus spp.; Aprendizado de máquina; Machine learning; Árvores aleatórias; Random trees; CNPQ::CIENCIAS EXATAS E DA TERRA::GEOCIENCIAS::GEOFISICA::SENSORIAMENTO REMOTO; CNPQ::CIENCIAS AGRARIAS::AGRONOMIA::FITOSSANIDADE::FITOPATOLOGIA; Cafeeiro; Imagens multiespectrais; Sensoriamento remoto

Remote sensing based on multispectral imaging may be a useful tool to detect vegetation stress responses. In this study, we assessed the potential of orbital multispectral imaging in discriminating the most effective strategies in reducing plant-parasitic nematode populations and in increasing the yield of coffee. The plants were treated with seven Bacillus isolates separately (B. subtilis isolates B18, B202 and B33; B. thuringiensis isolate B22; B. safensis isolate B53; B. amyloliquefaciens isolate B266; B. methylotrophicus isolate B05); commercial biological product based on B. subtilis + B. licheniformis (BC); combined application of abamectin + CB (QBC); commercial chemical nematicide based on fluensulfone (QC); Water was applied as a negative control. The initial and final nematode population in the soil, coffee yield, and other agronomical traits were assessed. Radiometric data were obtained from the Planet orbital multispectral sensor. The data were classified using the random tree algorithm. The population of plant-parasitic nematodes was reduced by 50.45% and 65.12% after the application of the isolates B33 and B266. Plants treated with QBC had the highest yield. Multispectral imaging was more accurate in discriminating the most nematicidal treatments than the agronomical parameters, with global accuracy (GA) of 80% versus 75.75%, respectively. The agronomic parameters were the most reliable estimators for coffee yield (GA = 100%) in comparison to multispectral imaging, which had a maximum GA of 83.63%. Thus, orbital multispectral imaging can discriminate the most effective treatments used for nematode management in coffee.