EN
The missing data on bulk density make impossible the calculations of soil water retention, organic carbon pools and evaluation of the numerical indices of soil and forest habitat fertility and potential productivity. A common reason of skipping of bulk density measurement is the high content of rock fragments in soil, in particular in the subsoil layers, that may hamper the collection of samples with undisturbed soil structure. The aim of present study was to analyse the relationships between bulk density and basic properties of mountain soils, where the problem of high skeleton content is relatively common, selection of the properties of the largest impact on the soil bulk density, and the fitting of the mathematical model (equation) that allows the prediction of soil bulk density at the highest available statistical significance. Forest soils in the Sudeten Mountains and the Sudeten Foreland, SW Poland (total number of samples N=580), situated on various bedrocks (granite, basalts, gneiss, mica schists, sandstones, mudstones, etc.) were selected to analysis. It was found in those soils featured by large variability of soil texture (the range of clay content 0−40%), skeleton content (0−74%), and organic carbon content (0.09−50%) in the soil profile, that the bulk density (having the values 0.10−1.87 g/cm³ in mineral layers and 0.04−0.58 g/cm³ in organic/litter layers) was individually correlated with the organic carbon content, skeleton content, clay content, and the depth of sampling (depth of particular soil layer), but the multifactor analysis has shown, that the organic carbon content has the highest impact on the bulk density and the organic carbon alone is sufficient for reliable prediction of soil bulk density. The best fitting of the measured and predicted bulk densities (R²=0.83) was provided by ‘dose−response’ Hill model: dv=1.72–[(2.36·Corg0.6)/(13.80.6+Corg0.6)], that returned dv values between 0.10 and 1.72 g/cm³ in a range of Corg content between 50 and 0%.