Soil texture analysis by laser diffraction and sedimentation and sieving–method and instrument comparison with a focus on Nordic and Baltic forest soils

Abstract

Purpose: Laser diffraction (LD) for determination of particle size distribution (PSD) of the fine earth fraction appeared in the 1990s, partly substituting the Sieving and Sedimentation Method (SSM). Whereas previous comparison between the two methods predominantly encompasses agricultural soils, less attention has been given to forest soils, including pre-treatment requirements related to their highly variable contents of carbon and Alox+ Feox. In this small collaborative learning study we compared (1) national SSM results with one type/protocol of LD analysis (Coulter), (2) LD measurements performed on three different LD instruments / laboratories, and (3) the replication error for LD Coulter analysis of predominantly sandy and loamy forest soils. Methods: We used forest soil samples from Denmark, Norway and Lithuania and their respective national SSM protocols / results. LD analyses were performed on Malvern Mastersizer 2000, Sympatec HELOS version 1999, and Coulter LS230, located at University of Copenhagen, Aarhus University and Helsinki University, respectively. The protocols differed between laboratories, including the use of external ultrasonication prior to LD analysis. Results: The clay and silt fractions content (<20 μm) from the LD analysis were not comparable with SSM results, with differences ranging from −0.5 to 22.3 percentage points (pp) for clay. Preliminary results from loamy samples with spodic material suggested inconsistent effects of external ultrasonication to disperse aggregates. The comparison between the three LD instruments showed a range in the clay and silt fractions content of 1.9–5.3 and 6.2–8.1 pp, respectively. Differences may be related to the instruments, protocols, and content of a given particle size fraction. The replication error of the Coulter LD protocol was found to be <3 pp in sandy soils, but up to 10 pp in loamy soils. Conclusion: Differences in the clay fraction results partly affected the classification of soil types. The fast replication of the LD analysis enables more quality control of results. The pedological evaluation of non-silicate constituents and optional pre-treatment steps (e.g., soil organic matter or sesquioxides) remains the same for LD and SSM. For comparison of results, detailed descriptions of the analytical protocol including pre-treatments are needed irrespective of instrument and theoretical approach.