Particle Size Distribution of Various Soil Materials Measured by Laser Diffraction—The Problem of Reproducibility

Abstract

Particle size distribution is an important soil parameter—therefore precise measurement of this characteristic is essential. The application of the widely used laser diffraction method for soil analysis continues to be a subject of debate. The precision of this method, proven on homogeneous samples, has been implicitly extended to soil analyses, but this has not been sufficiently well confirmed in the literature thus far. The aim of this study is to supplement the information available on the precision of the method in terms of reproducibility of soil measurement and whether the reproducibility of soil measurement is characterized by a normal distribution. To estimate the reproducibility of the laser diffraction method, thirteen various soil samples were characterized, and results were analysed statistically. The coefficient of variation acquired was lowest (3.44%) for silt and highest for sand (23.28%). Five of the thirteen tested samples were characterized by a normal distribution. The fraction content of eight samples was not characterized by normal distribution, but the extent of this phenomenon varied between soils. Although the laser diffraction method is repeatable, the measurement of soil particle size distribution can have limited reproducibility. The main cause seems to be small amounts of sand particles. The error can be amplified by the construction of the dispersion unit. Non-parametric statistical tests should be used by default for soil laser diffraction method analysis.

Spatial particle size distribution at intact sample surfaces of a Dystric Cambisol under forest use

“…The highest average CV and therefore the lowest reproducibility for both classification systems were found for the sand, silt, and clay fraction of the Bs horizon and for the sand and clay fraction of pretreated samples of the Bsw and BwC horizons (Table 3). Thus, the lower the sand content (Table 2), the lower the reproducibility as also found by Polakowski et al (2021). In our study, the texture classes of the four horizons are very limited (silty, sandy loam) and the previous statements is consistent for the present study.…”

“…The reproducibility of LDM measurements was determined by the coefficient of variation (CV). As the ratio of the standard deviation to the average (%), the lower the CV value, the higher is the reproducibility (e.g., Bieganowski et al, 2018;Polakowski et al, 2021).…”

“…It should be considered that the LDM itself is a repeatable method in contrast to SPM (e.g., Makó et al, 2019) and the reproducibility of the measurements depends on the texture (Goosens, 2008;Polakowski et al, 2021). Nevertheless, reproducible measurements require a clean, stable background, because particulate contaminations cause fluctuations on the background over time (Mastersizer 3000 User Guide, 2020).…”

“…The position in the stirrer as part of the Hydro EV dispersion unit can also limit the reproducibility. A stirrer high above the bottom of the suspension-containing vessel is not efficient enough to move heavy particles (Polakowski et al, 2021). This circumstance would lead to an underestimation of the sand fraction.…”

“…Intact soil aggregates from the same diagnostic horizon can be used to account for a similarity in the pore structure but can differ in OC and texture due to small-scale heterogeneity (Leue et al, 2020). For better description of PSD along heterogeneous structured soil surfaces, methods with lower material usage such as the laser diffraction method (Polakowski et al, 2021;Yang et al, 2019) are required. The reason is the limited amount of soil material that can be separated from thin macropore surfaces (Leue et al, 2017).…”