Near-infrared spectroscopic assessment of soil texture in an oxisol of the eastern plains of Colombia

Techniques such as near infrared spectroscopy (NIR) can be used to identify classes and properties of soils with good precision. The aim of this study was to calibrate models to predict the content of clay, silt and sand of a Typic Hapludox by NIR spectroscopy. The study was carried out in the Carimagua Experimental Station located in the municipality of Puerto Gaitán, Meta, Colombia. A grid design was used, 1200 samples were collected in an area of 5100 ha. The development of the models have made by partial least squares regression. A low representatively models were obtained for sand and silt contents, with values of R2 (0.41 and 0.34, respectively).The clay model content showed a high R2 (0.76). For the clay was possible the development of similar digital maps and digital spectrum maps . The results for the clay content indicate that laboratory analysis can be substituted, in large part by spectral models. In the case of sand and silt, in the future, it would be convenient to improve the model for substituting the laboratory analysis for this class of soil.