A comparison of soil texture measurements using mid-infrared spectroscopy (MIRS) and laser diffraction analysis (LDA) in diverse soils

Summary

This comparative study evaluated the performance of mid-infrared spectroscopy and laser diffraction analysis against conventional sieve-pipette texture measurements in soils ranging from European to Kenyan origins. MIRS showed superior clay prediction (R² = 0.83) compared to LDA (R² = 0.36), though both techniques performed well for sand content. The analysis revealed that organic carbon content significantly affects clay prediction accuracy, particularly above 5% OC, and that LDA's widely used 8 µm clay threshold was found to be too high; a 4 µm threshold proved more accurate.

UK applicability

These findings are directly applicable to UK agricultural soil testing, as UK soils typically fall within the recommended range of <5% organic carbon and <60% clay content where both techniques provide reliable estimates. UK laboratories and advisory services could adopt either technique for faster, cheaper soil texture determination whilst maintaining adequate accuracy for typical agricultural soils.

Key measures

Soil texture fractions (sand, silt, clay) measured by MIRS, LDA, and conventional sieve-pipette methods; organic carbon content; calibration set R² values for clay and sand predictions

Outcomes reported

The study compared mid-infrared spectroscopy (MIRS) and laser diffraction analysis (LDA) against conventional sieve-pipette methods for measuring soil texture across diverse European and Kenyan soils. Both techniques were evaluated for accuracy in predicting sand and clay content across varying organic carbon levels.

Topic tags