Resumo

ESQUIVEL-VALENZUELA, Berenice et al. Organic carbon and total nitrogen in Mexican forest soils through VIS-NIR spectroscopy. Rev. mex. de cienc. forestales [online]. 2018, vol.9, n.47, pp.295-313. ISSN 2007-1132.  https://doi.org/10.29298/rmcf.v9i47.158.

The sustainable use of the soil resource, as well as the correct decision making regarding its management depends on the knowledge of its properties. Conventional soil analysis methods are laborious, costly and generate large amounts of chemical waste. Due to the need for methods that allow a rapid, reliable analysis of soil properties, the objective of this work was to develop a prediction model for the content of organic carbon (OC) and total nitrogen (TN) in the soil through Visible and Near Infrared regions (VIS-NIR) spectroscopy. The conventional analysis of TC and TN of 599 forest soils from different regions of Mexico was carried out by dry combustion, and the content of inorganic carbon (IC) was determined by Bernard's calcimeter method. The prediction models were developed in a FOSS NIR System 6500. The models generated in the calibration process presented R2 values of 0.93 and 0.88 for OC and TN, respectively. The values of the relationship between the standard error of prediction and the standard deviation of the samples (RPD, for its acronym in English) for both properties were higher than 2. In the case of the validation process, the values of R2 were higher than 0.9, and the RPD was also above 2 in both properties. The results of this study show that VIS-NIR spectroscopy is an alternative technique to the conventional analysis methods of organic carbon and total nitrogen of the soil.

Palavras-chave : Absorbance; spectral analysis; prediction models; chemometrics; reflectance; forest soils.