Soil texture is a key soil physical property for soil quality and used in modeling studies through pedotransfer functions (PTF) for the prediction of physical, e.g. hydraulic, soil properties. Soil texture is quantified by a particle size distribution (PSD) of the fine earth fraction and often translated into a texture class using defined separates of clay (0 - 2 µm), silt (2 µm to 20 µm, 50 µm or 63 µm) and sand (20 µm, 50 µm or 63 µm up to 2 mm) illustrated in a texture triangle.
Until now pretreatment methods (e.g. humus and carbonate removal and dispersion) followed by standardised sedimentation and sieving methods have been well-defined. From literature and a mini-survey, we know already that laser diffraction is a commonly used analytical method for soil PSD determination in scientific environmental studies that involve soils. A body of literature has documented that colloid-sized fraction results obtained by laser diffraction analysis of fine-textured soil samples are not comparable to those obtained with sedimentation and sieving methods, when translating to the traditional particle size limits clay, silt and sand. Also, operating procedures for pretreatment of soil samples are variable, and the analyzed sample volumes are small, adding to uncertainty.
In this study we first compared PSD’s from three different instruments for a set of soil samples to study reproducibility using the analytical operating procedures developed by the owner institutions (Malvern Mastersizer 2000, University of Copenhagen, Coulter LS230, University of Helsinki, and Sympatec Helos, Aarhus University). Secondly, we compared the influence of 1 mm sieving and found decreased fraction standard deviation and improved repeatability of the PSD determination by laser diffraction on the Coulter LS230. 1 mm sieving should be corrected for if the mass is more than a few percent, but depending on study purpose. Thirdly, the laser diffraction PSD’s were compared with PSD’s obtained by sieving and hydrometer analysis showing well-known underestimation of colloids and fine fractions, that increased with colloid content.
We conclude that PSD’s obtained by the laser diffraction method are repeatable and mostly reproducible given standardised pretreatment. Translation to texture class using traditional separates does not work well, and more work and new PTF’s for soils are needed that can translate a laser diffraction PSD into a texture class and its associated physical properties for further use in modeling studies.