Correlations and Principal Components Analysis for Defining Management Zones in Cotton / Correlações e análise de componentes principais para definir zonas de gerenciamento em algodão

• Autor(a) principal: Baio, Fabio Henrique Rojo
• Data de Publicação: 2020
• Outros Autores: Faraun, Rafael da Silva, Teodoro, Paulo Eduardo, Silva, Alexandra Fagioli da, Neves, Danilo Carvalho, Azevedo, Gileno Brito de
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Revista Veras
• Texto Completo: https://ojs.brazilianjournals.com.br/ojs/index.php/BRJD/article/view/6899
• topic: Gossypium hirsutum; precision agriculture; multivariate analysis.

One approach for using variable rate fertilizer applications in precision agriculture is to divide an area into management zones. The objectives were: (i) to identify the chemical, physical and phenological properties that have the highest correlation with the yield; (ii) to use principal component analysis (PCA) to identify what physical, chemical, and phenological properties contribute to greater spatial variability; (iii) and to use these variables in the establishing management zones (MZ) for cotton through fuzzy k-means clustering analysis, associated with the geostatistics technique by the ordinary kriging method. The experiment was carried out in a cotton field in the Chapadões region in 2015. Phenological variables of cotton (plant height, number of bolls, number of capsules, opening percentage and Red Edge vegetation index) and chemical (pH, Ca, Mg, H+Al, V%, Ca/Mg, CEC, K, Al3+ and P) and physical (total soil porosity, soil density, soil moisture, soil mechanical resistance to penetration, clay content, and macro and micro-porosity) attributes of the soil were evaluated to define management zones. The variables that showed the highest correlation with cotton yield were pH, phosphorus, soil moisture measured at 39 and 70 days after cotton emergence (DAE), number of bolls at 107 DAE and red edge vegetation index at 53 DAE. The map with four MZ has a better representation, being the most indicated in the management of agricultural inputs applications at variable rates aiming to increase the cotton yield in the Brazilian Cerrado.