Elsevier

Geoderma

Volume 375, 1 October 2020, 114543
Geoderma

Soil particle density as affected by soil texture and soil organic matter: 2. Predicting the effect of the mineral composition of particle-size fractions

https://doi.org/10.1016/j.geoderma.2020.114543Get rights and content

Highlights

  • PTF predicts the particle density of clay, silt and sand fractions.

  • The mineral composition differs in the clay, silt and sand fractions of soil.

  • An appropriate PTF considers the effect of mineral composition on soil particle density.

Abstract

The particle density of soil (ρS) represents one of the basic physical properties of soil. However, measurement of this parameter is not part of common routine soil inventories in most countries. Therefore, pedotransfer functions (PTFs) were developed to calculate ρS. Here, we used a complex, hierarchically structured PTF to calculate ρS, separating the soil mineral substance (SMS) into clay, silt and sand fractions as well as separating the soil organic matter (SOM) into heavy-density and low-density fractions. This PTF was recently published, and here, we introduced an additional hierarchical level to consider the particle size fraction-dependent effect of the mineral composition on ρS. This extended PTF was calibrated and validated using data from soils of 16 German long-term experiments contrasting in soil texture and in soil mineral composition. Χ-ray diffraction analysis was applied to identify the mineralogical composition of the clay, silt and sand fractions. We fitted the particle densities of the identified minerals occurring in the respective particle size fractions by minimising the squared differences between the measured and predicted ρS. The model performed very well (RMSE = 0.011 Mg m−3 soil). According to the mechanistic base of the model and its hierarchical structure, it is very easy to include available information about the composition of any fraction or subfraction of soil mineral substances and to use the model to calculate the ρS corresponding to the specific site conditions.

Introduction

The particle density of soil (ρS) represents one of the basic physical properties of soil and is defined as the mass per unit volume of solid soil components, i.e., excluding voids and water. This parameter depends on the composition of both the mineral and the organic soil components. Therefore, ρS varies for different soils, e.g., within a group of mineral soils, ρS usually ranges from 2.4 to 2.9 Mg m−3 soil; in organic soils, ρS can decrease to approximately 1.5 Mg m−3 soil. Because ρS is not part of common routine soil inventories in most countries, pedotransfer functions (PTFs) were developed to calculate ρS. The basic approach of Adams (1973) comprises a mixture ratio between the soil mineral substance (SMS) and soil organic matter (SOM), both on a mass basis, and the corresponding mean particle densities of SMS (ρSMS) and SOM (ρSOM)). This approach was the starting point of the stepwise development of ρS predicting PTFs that still persists today, as described by Ruehlmann (2020). In that contribution, a complex PTF to calculate ρS was introduced, separating the soil mineral substance (SMS) into clay, silt and sand fractions as well as separating the soil organic matter (SOM) into heavy-density and low-density fractions (Eq. (1)):ρS=SMSclayρclay+siltρsilt+sandρsand-1+SOMSOMhdρSOMhd+SOMldρSOMld-1-1where clay, silt, and sand are mass proportions of SMS given in kg kg−1 minerals and ρclay, ρsilt and ρsand are the corresponding particle densities given in Mg m−3 fraction. High density (SOMhd) and low density (SOMld) are mass proportions of SOM and given in kg kg−1 SOM; ρSOMhd and ρSOMld are the corresponding particle densities given in Mg m−3 SOM. The calculations of SOMhd and SOMld are explained in detail by Ruehlmann (2020).

This ρS predicting model has a mechanistic base and a hierarchical structure to consider the soil’s main and subcomponents. The mass proportions of the main and subcomponents are strictly related to the corresponding particle densities. The model was applied to a dataset from locations worldwide, covering the full range of possible soil organic matter contents, diverse textures, and soil parent materials; thus, the model was able to predict the mean particle densities of the five subcomponents, as mentioned in Eq. (1), with ρclay=2.76, ρsilt=2.69, ρsand=2.66, ρSOMhd=1.43 and ρSOMld=1.27, always given in Mg m−3 fraction (Ruehlmann, 2020).

Different rocks, as basic materials of SMS, differ in particle density according to their mineralogy and according to their chemical composition and crystal structure. The particle density of single minerals can significantly differ from the abovementioned mean particle densities of the clay, silt and sand-size fractions. The ranges of selected clay minerals as reported by Deer et al. (1966) include smectites 2.0–2.6, illites 2.6–2.9, kaolinites 2.61–2.68, chlorites 2.6–3.3 and muscovites 2.77–2.88 Mg m−3 mineral. Totten et al. (2002) summarised that the published mineral densities span a rather wide range — from 2.0 (low end for smectites) to 3.3 (high end for chlorites). In heavy minerals, the particle density varies between 2.9 and 4.0 Mg m−3 mineral, respectively (Schachtschabel et al. 1992). These values are directly affected by the mineral type, e.g., biotite and haematite have values of 2.80–3.20 and 4.80–5.30 Mg m−3 mineral, respectively (Skopp, 2012). If one mineral type is dominant in a single soil particle size class, the particle density of this particle size class might be similar to that of the dominant mineral. McBride et al. (2012) calculated the linear regressions ρS = 2.560 + 0.024 (%clay) and ρS = 2.800–0.024 (%sand + %silt) and discussed the differences in the y-intercepts, 2.560 Mg m−3 for the lighter sand + silt fraction and 2.800 Mg m−3 for the heavier clay fraction, against the background of the soil’s mineral composition. Schjønning et al., 2017a, Schjønning et al., 2017b reported similar findings. Χ-ray diffraction analysis, as an appropriate experimental method to analyse the soil’s mineral composition, was applied to soils of 17 long-term experiments located in Germany (Rühlmann et al., 2006). They used the experimentally estimated proportions of different minerals to predict the particle density of the bulk soil’s mineral substances. However, to our knowledge, no findings are available on the effect of the mineral inventories occurring in different particle size fractions on ρS. Therefore, the aim of this paper is to analyse the mineral inventories of soil particle size fractions and to generate an appropriate structured ρS predicting PTF that allows us to consider the particle size fraction-dependent effect of mineral compositions on ρS.

Section snippets

Generating the PTF

To generate an appropriate structured ρS predicting PTF that considers the specific composition of the particle-size fractions, we wanted strictly to follow the principle of the Adams (1973) approach to relate the mass proportions of soil components to the corresponding particle densities. Therefore, we extended the SMS term in Eq. (1) by an additional hierarchical level – now the fractions of clay, silt, and sand are divided in each case into the proportions of the different mineral types

Results and discussion

The densities of the 8 mineral types (excluding calcite, with a constant density) and of the amorphous phase that were identified by Χ-ray diffraction analysis and calculated by the model calibration are shown in comparison to the minimum and maximum mineral densities given in the literature (Table 3).

Regarding the amorphous material, we set its density range according to Varajao et al. (2002) between the relatively low values of the aluminosilicate allophane (1.9 Mg m−3 mineral) and poor

Conclusions and perspectives

A hierarchically structured model approach that has been previously used to analyse the effects of clay, silt and sand fractions (as well as of SOM) on ρS (Ruehlmann, 2020) was successfully extended by a further hierarchical level – the mineral compositions of the clay, silt and sand fractions. Each parameter fitted as mentioned above represents the particle density of the particular fraction. The model performed very well, although all the fitted mineral densities were restricted by their

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

We thank Joachim Luckert (Landesamt für Geowissenschaften und Rohstoffe Brandenburg) for implementing the Χ-ray diffraction analyses and interpretation of their results. Furthermore, we thank the two anonymous reviewers of a former version of the manuscript for their helpful comments and suggestions.

This research did not receive any specific grants from funding agencies in the public, commercial, or not-for-profit sectors.

References (19)

There are more references available in the full text version of this article.

Cited by (0)

View full text