Spatial Correlation between Eucalyptus Diameter at Breast High and Particle Size Fractions of an Oxisol
Julião Soares de Souza Lima *
Department of Rural Engineering, Federal University of Espírito Santo, Alegre-ES, CEP: 29500-00, Brazil
Diego Antonio Ottonelli de Bona
Department of Rural Engineering, Federal University of Espírito Santo, Alegre-ES, CEP: 29500-00, Brazil
Nilton Cesar Fiedler
Department of Rural Engineering, Federal University of Espírito Santo, Alegre-ES, CEP: 29500-00, Brazil
Vagner Mauri Quinto
Department of Rural Engineering, Federal University of Espírito Santo, Alegre-ES, CEP: 29500-00, Brazil
Samuel de Assis Silva
Department of Rural Engineering, Federal University of Espírito Santo, Alegre-ES, CEP: 29500-00, Brazil
*Author to whom correspondence should be addressed.
Abstract
Field growth of Eucalyptus is influenced by physical attributes of soils, which justifies comprehensive studies on these relationships within a forest system. This work was aimed at studying the spatial variability in the diameter at breast high (DBH) of Eucalyptus, trees cultivated over three years and its relationships with the particle size fractions of a Oxisol. The study was conducted on a 33.0 x 33.0 m sampling grid, totalling 94.0 georeferenced field spots. Soil samples were collected from the 0 – 0.20 m and 0.20 – 0.40 m depth layers to determine the particle size fractions and DBH was measured in 5.0 neighbour Eucalyptus trees, with average DBH centered at each georeferenced spot. Data were submitted to descriptive, multivariate and geostatistical analyses. Maps were built using interpolated ordinary kriging and cokriging, while principal component analysis (PCA) was performed on the clay and total sand fractions from both soil layers. The correlation coefficients between the original variables and the first principal component (PC1) were large and negative and positive in relation to the clay and total sand fractions, respectively. The primary variable DBH was successfully estimated by cokriging using the PC1 score as a secondary variable. The simple and cross semivariograms of the DBH adjusted to the spherical model. The cokriging technique was found to be an efficient tool in estimating variables within forest production systems with reduced sampling cost.
Keywords: Geostatistic, cokriging, multivariate analysis