Pedotransfer Functions for Estimating Soil Bulk Density Using Image Analysis of Soil Structure.

Area of Science:

• Soil Science
• Agricultural Science
• Environmental Science

Background:

• Soil bulk density is a critical soil property influencing water and air movement.
• Existing pedotransfer functions (PTFs) often lack soil structure information, limiting prediction accuracy.
• Soil structure significantly impacts soil bulk density, yet is rarely integrated into predictive models.

Purpose of the Study:

• To develop novel PTFs for predicting soil bulk density by incorporating soil macrostructure data from image analysis.
• To evaluate the performance of new PTFs against existing models.
• To demonstrate the utility of image analysis for soil bulk density assessment.

Main Methods:

• Laboratory measurements of soil bulk density (BD), texture, and total organic carbon.
• Image analysis to assess soil macroporosity (BDim) and classify macropore presence (MP5).
• Development and comparison of PTFs using BDim and MP5, including models like BD~BDim + MP5 and BD~BDim.

Main Results:

• The proposed PTFs achieved mean prediction errors between 0 and -0.02 Mg m⁻³.
• Modelling efficiency ranged from 0.17-0.39, with prediction R² values from 0.35-0.41.
• PTFs incorporating MP5 showed improved prediction of boundary bulk density values, despite increased scatter for intermediate values.

Conclusions:

• Image analysis of soil macrostructure provides valuable data for developing accurate soil bulk density PTFs.
• The new PTFs enable bulk density estimation using only soil structure images, without requiring other soil data.
• Incorporating soil structure parameters into PTFs enhances their predictive capability for soil bulk density.