Relaciones entre componentes morfológicos de palmito y atributos físicos de un inceptisol: Una aproximación multivariada

Abstract

The study of the soiĺs agricultural potential from a multivariate approach can help to interpreter the interactions between its attributes and crop yield components. The aim of this study was to evaluate the relationships between the soiĺs physical attributes and morphological components of the heart palm in an Inceptisol Eutric in the city of Registro (São Paulo, Brazil) using multivariate analyses. The experiment was carried in the crop year 2013/2014 for a crop that was implanted in the year 2012, where yield components of heart palm and the soiĺs physical attributes were analyzed at a depth of -0.10 m, using a grid with 54 georeferenced points. Three multivariate methods were used: hierarchical clustering, non-hierarchical clustering (k-means) and principal components analysis (PCA). The k-means analysis showed larger heart palm stem diameter values in the group of points with higher mechanical resistance to penetration values, and lower gravimetric soil moisture values. The PCA showed that the first two principal components (CP1 and CP2) explained 68.29 % of the variations of the original data set, where soil density and total porosity showed the largest discriminatory power in CP1, and only stem diameter showed significant discriminatory power in PC2. Multivariate statistics was efficient in the discrimination of the physical variables of the soil studied that have a greater relationship with the morphological components of the heart palm plant.