Multivariate predictive model of minerals content in the basal portion of peach palm heart (Bactris gasipaes Kunth) using agrometeorological data

• Autor(a) principal: Bellettini, Marcelo Barba
• Data de Publicação: 2019
• Outros Autores: Bach, Fabiane, Morón, Miriam Fabiola Fabela, Bespalhok Filho, João Carlos
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Semina. Ciências Agrárias (Online)
• Texto Completo: https://ojs.uel.br/revistas/uel/index.php/semagrarias/article/view/34790
• topic: Peach palm; Mineral content; Agro-meteorological factors; Multivariate statistical analysis; Climatic influence.; Pupunha; Minerais; Fatores agro-meteorológicos; Análise estatística multivariada; Influência climática.

The climatic influence in minerals content of peach palm heart (Bactris gasipaes Kunth) was studied and a quick method was assessed to determine Mg, Cl, K and S in the basal portion of peach palm heart based on multivariate predictive model using agro-meteorological data. A total of 24 samples of B. gasipaes Kunth were collected along 14 to 18 months of cultivation, growing in two types of terrain: hillside and lowland. Principal component analysis (PCA) was used to select principal components. The data were modeled using partial least squares regression (PLS). Low average relative prediction errors (4.60%) confirm the good predictability of the models. The factors that most influence the minerals content prediction model were the rain precipitation and solar radiation. The results show that predictive model can be used as rapid method to determine the mineral content in the basal portion of peach palm heart factories and may help to choose geographical regions suitable for the establishment of new peach palm plantations. The models can provide reductions of cost and time analysis to palm heart without generating laboratory effluents. This is the first time in which multivariate analysis is used to generate models to predict minerals concentration in the basal portion of peach palm hearts, quantifying numerically the intensity of climatic factors in the minerals content.