Prediction of bulk density in Croatian forest Pseudogleys based on contents of soil organic matter and clay

Abstract

Soil bulk density (BD) is often used to assess soil properties related to soil structure and general soil quality, or to convert weight units to volume/area units. Because collecting undisturbed soil samples (cores) and directly measuring BD is laborious and time-consuming, data on BD are often lacking. Pedotransfer functions (PTFs) can be used to predict BD from more readily available (easily measured) properties. However, specific PTFs for specific range of soils should be developed. In this study, soil organic matter (SOM) and clay contents of 90 forest Pseudogleys horizons, distributed across 11 locations in continental Croatia, were used to develop a multiple linear regression equation that predicts BD. The results have shown that the increase in BD due to the unit increase in clay content is lower than the decrease in BD due to the unit decrease in SOM content. PTF performance was relatively high (R2adj=77.5%, RMSE=0.10 g cm-3, MAE=0.08 g cm-3), showing that missing data on soil BD of Croatian Pseudogleys (or other similar soils, formed on similar parent materials) can be estimated reasonably well by using it. However, care should be taken when dealing with arable soils, which have different properties, compared with the natural soils considered herein.

Keywords

• continental Croatia
• forest soils
• multiple linear regression
• pedotransfer function (PTF)
• soil bulk density estimation
• Stagnosols