Iron Compounds and the Color of Soils in the Sakhalin Island

Abstract

Numerical parameters of soil color were studied according to the CIE-L*a*b color system before and after the Tamm’s and Mehra-Jackson’s treatments; we also determined the total Fe content in the samples from the main genetic horizons of the alluvial gray-humus soil, two profiles of burozems, and two profiles of podzols in the Sakhalin Island. In the analyzed samples, the numerical color parameters L* (lightness), a* (redness) and b* (yellowness) are found to vary within 46–73, 3–11, and 8–28, respectively. A linear relationship is revealed between the numerical values of a* parameters and Fe content in the Mehra-Jackson extracts; the regression equations are derived with the determination coefficients (R²): 0.49 (typical burozem), 0.79 (podzolized burozem), 0.96 (shallow-podzolic mucky podzol), 0.98 (gray-humus gley alluvial soil). For the surface-podzolic mucky podzol contaminated with petroleum hydrocarbons, R² was equal to only 0.03. In the gray humus (AY) and structural-metamorphic (BM) horizons of the studied soils, a* and b* parameters decrease after their treatment with the Tamm’s reagent by 2 points on average. After the Mehra-Jackson treatment, the a* parameter decreased by 6 (AY) and 8 (BM) points; whereas b* parameter, by 10 and 15 points, respectively. In the E horizons of podzols, the Tamm’s treatment increased a* and b* parameters by 1 point; whereas the Mehra-Jackson’s treatment decreased these parameters by only 1 and 3 points, respectively. The redness (a*) decreased maximally in the lower gley horizon of the alluvial gray humus soil, i.e., by 6 (in the Tamm’s extract) and 10 points (in the Mehra-Jackson’s) extract. Yellowness (b*) decreased by 12 and 17 points, respectively. The revealed color specifics in the untreated samples and the color transformation under the impact of reagents in the studied soils and horizons may serve as an additional parameter that characterizes quantitatively the object of investigation in the reference databases.