Comparison of MIR Spectral Signatures of Soils from Acacia, Eucalyptus, and Cassava Plantations in the Tropical Region of the Batéké Plateau (Republic of Congo)

This study examines the effects of Acacia, Eucalyptus, and Cassava plantations on soil chemical composition in the Batéké Plateau, a tropical savanna region in Congo-Brazzaville. Using mid-infrared (MIR) spectroscopy, the research reveals distinct spectral signatures for each type of plantation, highlighting notable variations in soil chemical properties. Soils under Acacia exhibit increased organic nitrogen content, enhancing fertility through atmospheric nitrogen fixation. In contrast, soils under eucalyptus contain lignified organic matter, which is less biodegradable and limits immediate nutrient access but may contribute to long-term carbon storage. Cassava soils show high levels of silicates and labile components, indicating a risk of rapid nutrient depletion without proper management. The results underscore the importance of carefully selecting plant species to achieve sustainable soil management objectives and suggest that a combined MIR and NIR spectroscopy approach could enrich analyses by providing a more comprehensive view of soil variations based on vegetation cover.

Rapid recognition of Chinese herbal pieces of Areca catechu by different concocted processes using Fourier transform mid-infrared and near-infrared spectroscopy combined with partial least-squares discriminant analysis

Rapid and sensitive recognition of herbal pieces according to different concocted processing is crucial to quality control and pharmaceutical effect. Near-infrared （NIR） and mid-infrared （MIR） technology combined with supervised pattern recognition based on partial least-squares discriminant analysis （PLSDA） was attempted to classify and recognize six different concocted processing pieces of 600 Areca catechu L. samples and the influence of fingerprint information preprocessing methods on recognition performance was also investigated in this work. Recognition rates of 99.24%, 100% and 99.49% for original fingerprint, multiple scatter correct （MSC） fingerprint and second derivative （2nd derivative） fingerprint of NIR spectra were achieved by PLSDA models, respectively. Meanwhile, a perfect recognition rate of 100% was obtained for the above three fingerprint models of MIR spectra. In conclusion, PLSDA can rapidly and effectively extract otherness of fingerprint information from NIR and MIR spectra to identify different concocted herbal pieces ofA. catechu.