Application of Near-Infrared Spectroscopy to Detect Modification of the Cation Exchange Properties of Soils from European Beech and Silver Fir Forest Stands in Poland

This study investigated changes in the composition of the cation exchange capacity of soil samples caused by the acid leaching of soil cations under laboratory conditions. Furthermore, near-infrared (NIR) spectroscopy was used to evaluate the properties of forest soils. The potential influence of the species composition of stands (beech and fir) was also investigated. Eighty soil samples from the topsoil of plots located in central Poland were analyzed. Soil samples were leached 0 (non-leached), 5, 10, and 15 times and then analyzed to determine the contents of cations (Al 3+ , Ca 2+ , K + , and Mg 2+ ), the total carbon content, and the pH. From NIR spectra obtained by scanning 54 samples and measurement results for soil sample properties, a calibration model was developed. The model was validated using 26 independent samples. The results showed that acid leaching decreased the pH of soil solutions and the carbon content. The amounts of Al 3+ , Ca 2+ , K + , and Mg 2+ decreased with an increasing number of leaching treatments, but most leaching had occurred after five treatments. Data analysis showed that leaching with hydrochloric acid depleted alkaline cations and Al 3+ in the soil, which reduced the stability of organic matter, causing its release. Modification of ion exchange properties is observable based on the analysis of the NIR spectra. Good calibration results were achieved for all tested parameters (R 2 C ≥ 0.89). The best validation results were obtained for Al 3+ and C contents under fir stands, and for the pH and Al 3+ content of soils under beech stands (R 2 V > 0.8). However, the differences between the measured and estimated mean values of the investigated soil were relatively small (no significant difference, p > 0.05). The species composition of stands (beech and fir) had no impact on the developed mathematical models. Soil assessment using NIR spectroscopy allowed calibration models to be obtained, which were successfully used to calculate soil properties at a much lower cost and in a much shorter time compared with other laboratory methods. The results of the paper affirmed that using a relatively small number of samples (3–4) to calculate an average of soil content properties provided satisfactory results.