Near infrared spectroscopy-A tool for chemical properties and organic matter assessment of afforested mine soils

The objectives of this study were to assess chemical and microbial properties of different post-mining barrens afforested with Scots pine (Pinus sylvestris) and to evaluate the potential of near infrared spectroscopy (NIRS) to predict these properties in the reclaimed soils. The samples were collected from four sites at different substrates and measured for the contents of organic C (C-org), total N (N-t), C-org-to-N-t ratio, dehydrogenase activity, contents of humified C, humic acid C and fulvic acid C, cation exchange capacity, exchangeable acidity, total exchangeable bases (TEB) and base saturation. The spectra of samples were recorded between 400 and 2500nm. A modified partial least squares method with cross-validation was used to develop calibration models for prediction of chemical and microbial properties of the samples from their spectra. There were significant differences in the contents of C-org, N-t, dehydrogenase activity, pH and sorption properties between the mine soils. The samples from the studied sites had different spectral characteristics suggesting differences in chemical composition of their organic matter. "Excellent" predictive models (coefficient of correlation between the predicted and reference values >= 0.90) based on NIR spectra were developed for N-t, C-org-to-N-t ratio, humified C, humic acid C and exchangeable acidity. For other soil properties (except the TEB), "satisfactory" (r = 0.80-0.90) models useful for fast screening were developed. The results indicate the potential of NIRS to predict several properties of the reclaimed mine soils. (C) 2013 Elsevier B.V. All rights reserved.