Near infrared spectroscopy and principal components analysis for investigation of soils submitted to different land uses in the Brazilian Eastern Amazon

Abstract

The study of soil is important due to its direct relation to the food supply, and the maintenance and improvement of environmental conditions. In this work, an analytical method was developed based on near infrared spectroscopy (NIR) and principal component analysis (PCA) to study an Oxisol submitted to different land uses (primary forest; secondary forest; fruit growing; horticulture; degraded pasture; improved pasture and annual agriculture.) in the Brazilian Eastern Amazon. The results of PCA using diffuse reflectance spectra showed 97 % of the total variance explained in PC1 and PC2, indicating the formation of four sample groups related to different uses and linked to the physico-chemical attributes (contents of carbon, nitrogen, potassium and density) of the soil. The analytical method developed, based on NIR spectra and PCA, is a promising way to identify soils subjected to different uses.