Steinmergelkeuper Forest Soils in Luxembourg: Properties and Pedogenesis of Soils with an Abrupt Textural Contrast

Abstract

The process of clay dispersion and soil formation is investigated for an ‘in situ’ soil under semi-natural deciduous forest in Luxembourg on Steinmergelkeuper marls. We studied the genesis of these soils with a characteristic abrupt textural contrast between the topsoil and the subsurface soil. During and after rainfall, subsurface flow is generated, laterally transporting and exporting dispersed clay downslope at the interface between the silty topsoil and the clayey Bg horizon. This process resulted in a relative coarse silty and shallow topsoil, directly on a clayey Bg horizon. On hillslope positions, the clay content increased from 23% clay in the topsoil to 50% in the subsurface horizons. This abrupt textural contrast was less pronounced in soils on water divides and the colluvial valley bottom. The non-calcareous soils had an AEh–EAh–Bg–Bw–C horizon sequence, a near neutral pH, a very high base saturation (>96%) dominated by Ca and Mg, and a relative organic matter rich topsoil. They are classified as Luvic Planosols. The main mechanism involved in the development of the textural contrast in the soil was the swelling and dispersion of (fine) clay from the top of the Bg horizon, and its lateral transport in macropores downslope over the almost impervious Bg horizon. The high macroporosity and hydraulic conductivity finds its origin in bioturbation, and swelling and shrinking of the topsoil. The lateral transported (fine) clay was dispersed from the top of the Bg horizon, as indicated by the similarity in the clay mineralogy of clay of the Bg horizon and dispersed clay sampled in subsurface flow and stream. The absence of stabilizing agents in the Bg horizon, such as organic matter, carbonates and pedogenic sesquioxides, allowed slaking of macro-aggregates, followed by the swelling of micro-aggregated clay. This swelling was caused due to the very low electrolyte content of the soil solution, which was always below the flocculation value of the clay of the Bg horizon. In addition, the dispersion domain of this soil material was enlarged by soluble humic substances, which were adsorbed at the clay particles. This lateral eluviation process of clays, or subsurface erosion of the Bg horizon, was the main process explaining the sharp abrupt textural contrast in these soils.

Appendices

Appendix 1

Description and particles size distribution, pH and organic carbon content of soil profile I, II and III, as presented in Fig. 9.4. The grain size distribution data are given in Table 9.10.

Table 9.10 Grain size distribution, soil pH and organic carbon content of profiles I, II and III

9.1.1 Description Profile I

Date of description: April 9, 1987

Author: Theo van den Broek

Classification:

Aquic Dystric Eutrochrept, fine silty, mixed, over clayey, mixed, mesic (Soil Survey Staff 1975); Luvic Planosol (IUSS 2015)

Location:

2 km NE of Schrondweiler, in the forest at the crest 49° 49′ 16″N 6° 10′ 50″E

Altitude:

328 m

Relief:

Gently undulating macro-relief

Slope:

0° (watershed)

Drainage:

Poorly to imperfectly drained

Vegetation:

Oak (Quercus robur L), hornbeam (Carpinus betulus L) and beach (Fagus sylvatica L)

Parent rock:

Steinmergelkeuper

All colours are for moist soil, according to the Munsell scale

AEh 0–10 cm:

Dull yellowish brown (10 YR-5/4) wet silt loam with few, faint, fine, diffuse mottles; moderate fine to medium crumb; slightly sticky, slightly plastic; many fine, common medium and a few coarse pores; very frequent fine and common medium roots; gradual, smooth boundary to

EAh 10–18 cm:

Dull yellowish brown (10 YR-5/4) wet silt loam with few, faint, fine, diffuse mottles; moderate fine to medium crumb; slightly sticky, slightly plastic; many fine, common medium and a few coarse pores; very frequent fine and common medium roots; gradual, smooth boundary to

Bg 18–30 cm:

Grayish yellow brown (10 YR-5/2) wet silty clay loam with many, medium, distinct, clear yellowish (10 YR-5/8) mottles; strong medium sub-angular blocky; slightly sticky, very plastic; common fine and few coarse pores; few fine and a very few coarse roots; gradual boundary to

Bw 30–45 cm:

Grey (5 Y-6/1) wet silty clay loam with common, fine, distinct, diffuse brown (10 YR-4/4) mottles; moderate, medium, sub-angular blocky; slightly sticky, very plastic; few fine and very fine pores; few coarse roots; gradual, wavy boundary to

C1 45–60 cm:

Grayish olive (5 Y-5/2) wet silt loam with no mottles; massive structure; slightly sticky, very plastic; few very fine pores; no roots; few large, hard, weathered Steinmergelkeuper fragments; gradual boundary to

C2 60–85 cm:

Grayish olive (5 Y-5/2) wet silt loam with no mottles; massive structure; slightly sticky, very plastic; few very fine pores; no roots; large, hard, weathered Steinmergelkeuper fragments

9.1.2 Description of Profile II. The Reference Profile

Date of description: March 24, 1987

Author: Theo van den Broek

Classification:

Aquic Dystric Eutrochrept, fine silty, mixed, over clayey, mixed, mesic (Soil Survey Staff 1975); Luvic Planosol (IUSS 2015)

Location:

2 km NE of Schrondweiler, in the forest halfway on the slope 49° 49′ 16″N 6° 10′ 50″E

Altitude:

325 m

Relief:

Gently undulating macro-relief

Slope:

4°

Aspect:

North facing

Drainage:

Poorly to imperfectly drained

Vegetation:

Oak (Quercus robur L), hornbeam (Carpinus betulus) and beach (Fagus sylvatica L)

Parent rock:

Steinmergelkeuper

All colours are for moist soil, according to the Munsell scale

AEh 0–10 cm:

Dark brown (10 YR-3/4) wet silt loam with few, fine, faint, diffuse mottles; moderate fine to medium crumb; slightly sticky, non-plastic; many fine, common medium and a few coarse pores; very frequent fine and common medium roots; gradual, wavy boundary to

EAh 10–20 cm:

Dark brown (10 YR-3/4) wet silt loam with few, fine, faint diffuse mottles; moderate medium crumb; slightly sticky, non-plastic; many fine, common medium pores; frequent, fine and common, medium roots; abrupt, smooth boundary to

Bg 20–34 cm:

Brownish grey (10 YR-5/1) wet clay with many, medium, distinct, clear yellowish brown (10 YR-5/8) mottles; strong, medium/coarse angular blocky; slightly sticky, very plastic; few very fine pores; few, very fine and very few medium roots; gradual boundary to

Bw 34–50 cm:

Brownish grey (10 YR-6/1) wet silty clay with common, fine, distinct, diffuse brown (10 YR-4/4) mottles; moderate, medium sub-angular blocky; slightly sticky, plastic; few very fine pores; few, very fine and very few medium roots, clear, irregular boundary to

Cl 50–58 cm:

Dark reddish brown (5 YR-3/4) wet silt loam, no mottles; moderate, medium sub-angular blocky; slightly sticky, plastic; very few, very fine pores; no roots; clear, irregular boundary to

C2 58–75 cm:

Grayish yellow (2.5 Y-6/2) wet silt loam, no mottles; medium sub-angular blocky; slightly sticky, plastic, no pores; no roots; few weathered Steinmergelkeuper marl flakes

C3 75–90 cm:

Grayish yellow (2.5 Y-6/2) wet silt loam, no mottles; medium sub-angular blocky; slightly sticky, plastic, no pores; no roots; weathered Steinmergelkeuper marl flakes.

9.1.3 Description of Profile III

Date of description: November 1, 1988

Author: Theo van den Broek

Classification:

Aquic Dystric Eutrochrept, fine silty, mixed, over clayey, mixed, mesic (Soil Survey Staff 1975); Luvic Planosol (IUSS 2015)

Location:

2 km NE of Schrondweiler, in the forest at footslope, near the stream; 49° 49′ 16″N, 6° 10′ 50″E

Altitude:

323.5 m

Relief:

Gently undulating macro-relief 2

Slope:

2°

Aspect:

North facing

Drainage:

fairly well drained

Vegetation:

Oak (Quercus robur L), hornbeam (Carpinus betulus L) and beach (Fagus sylvatica L)

Parent rock:

Colluvium on Steinmergelkeuper

All colours are for moist soil, according to the Munsell scale

AEh 0–9 cm:

Brown (10 YR-4/4) moist silt loam; weak fine to medium crumb; friable; many fine, medium and coarse pores; many roots, faint boundary to

EAh 9–0 cm:

Dull yellowish brown (10 YR-5/4) moist silty clay loam with few, fine, faint, diffuse mottles; moderate sub-angular blocky; firm; many fine, medium and coarse pores; many roots; faint boundary to

2Bg 20–32 cm:

Dull yellowish brown (10 YR-5/4) moist clay with common, medium, faint yellowish brown (2.5 Y-5/6) mottles; strong angular blocky; very firm; few fine pores; many fine roots; faint boundary to

2C1 32–45 cm:

Grayish yellow brown (10 YR-6/2) moist clay with common, medium, faint yellowish brown (2.5 Y-5/6) mottles; strong angular blocky; very firm; few fine pores; few fine roots.

Appendix 2

Clay mineralogy of the total, coarse and fine clay of the reference profile (Profile II, Appendix 1):

9.2.1 Overall View of the Clay Mineralogy of the Soil Profile

Illite and some kaolinite is present in the whole of the profile in as well the fine, coarse as total clay fraction.

The interstratification chlorite–smectite is found in all horizons in as well the fine as well the total clay fraction. In the coarse clay fraction this swelling interstratification is lacking in the surface horizons, and is only present in the Bw and C1 horizons (the last probably due to a non-ideal separation of the fine and coarse clay during the pretreatment of the samples, caused by lithogenic aggregation).

The interstratification chlorite–vermiculite is present in the fine, coarse and total clay fraction in the whole of the profile. Chlorite is present in particular in the surface horizons of the total and coarse clay fractions.

9.2.2 Total Clay (<2 µm)

Summary

Illite and some kaolinite is present in all the horizons of the profile. The interstratification chlorite–smectite is present in particular in the Bg, Bw and Cl horizons. Chlorite is only shown in the AEh and EAh horizons, whereas the interstratification vermiculite–chlorite is present in the whole of the profile, although the most pronounced in the deeper horizons. Quartz is found in the surface horizons.

Throughout the complete profile, strong and sharp reflections show up around 10, 4.9 and 3.34 Å, being the 001, 002 and 003 reflections respectively of a good crystalline illite.

Saturation with Mg²⁺ yields for whole the profile a peak at 14 Å, which shifts after a glycerol treatment towards 16 Å for particularly the Bg, Bw and C horizons, implying a swelling, interstratified mineral. This, together with the shift of the 29–30 Å peak to a 32–34 Å peak after treatment with glycerol points towards the interstratification chlorite–smectite. The14 Å peak is shown for the Mg as well as the K-saturation. Saturation with K⁺ gives a peak at 14 Å. In the surface horizons it is a good, sharp crystalline peak, whereas deeper in the profile the peak becomes broader.

After heating to 300 °C, the peak collapses towards 10 Å the interstratification of chlorite–vermiculite. At 550 °C the 14 Å peak increases in intensity, mainly in the AEh and EAh horizons: chlorite, whereas the intensity of the 7.07, 4.7 and 3.54 Å peaks decreases, Chlorite peaks show up at 14.1, 7.07, 4.72 and 3.54 Å, being the 001, 002, 003 and 004 reflections respectively. At 550 °C the 14 Å peak of chlorite intensifies, whereas the 7 Å peak disappears: kaolinite. A Quartz peak at 4.26 Å shows up in particular in the surface horizons.

9.2.3 Coarse Clay (2–0.2 µm)

Summary

Illite and some kaolinite is present in all horizons, whereas the interstratification chlorite–smectite shows up mainly in the deeper (Bg), Bw and C1 horizons The chlorite–vermiculite interstratification is shown in the complete profile, whereas chlorite is noticed particular in the upper part of the profile. Quartz shows up in the surface horizons too.

Throughout the complete profile sharp reflections show up around 10, 4.9 and 3.34 Å, being the 001, 002 and 003 reflections respectively of a good crystalline illite.

Saturation with Mg²⁺ results in a 14 Å reflection in all the horizons. After a glycerol treatment, the 14 Å peak shifts towards 16 Å for only the Bw and Cl horizons, and in a lesser degree for the Bg horizon. It implies the presence of a swelling mineral, being the interstratification chlorite–smectite¹. Saturation with K⁺ results in a 14 Å reflection in all the horizons. Compared to the Mg-peaks at 14 Å, the K-peaks are smaller: vermiculite. After heating to 300 °C, the 14 Å peak collapses, particular in the deeper horizons, towards 10 Å: the interstratification chlorite–vermiculite. At 550 °C the 14 Å peak increases in intensity especially in the surface horizons, whereas the intensity of the 7.07, 4.7 and 3.54 Å peaks decreases: chlorite, mainly in the AEh and EAh horizons. Chlorite peaks show up at 14.1, 7.07, 4.72 and 3.54 Å, being the 001, 002, 003 and 004 reflections respectively. A small Quartz peak at 4.26 Å shows up in particular in the surface horizons.

9.2.4 Fine Clay (<0.2 µm)

Summary

Illite and some kaolinite are present in the complete profile, even as the swelling interstratification chlorite–smectite, although the last is more pronounced in the deeper part of the profile. The interstratification chlorite–vermiculite is present in the whole of the profile. In the deeper horizons the chlorite component in this interstratification is larger than in the surface horizons.

Throughout the complete profile, broad reflections show up around 10, 4.9 and 3.34 Å, being the 001, 002 and 003 reflections respectively of illite. Considering the broad peaks, the illite in the fine clay has a lower degree of crystallinity compared to the illite in the total and coarse clay fractions.

Saturation with Mg²⁺ results in 14 Å peaks for whole the profile. After the glycerol treatment, these peaks shift towards 14.7 and 15 Å in the surface horizons, and to 16 Å in the deeper horizons. In the deeper horizons a clear peak shows up around 33 Å. In the Bw and C1-horizons the swelling mineral could well be the interstratification chlorite–smectite. In the surface horizons the contribution of this swelling mineral is less.

Saturation with K⁺ gives a 14 Å peak in all the horizons. These peaks are lower than the Mg-peaks: vermiculite. After heating to 300 °C, the 14 Å peak shifts towards 10 Å: the interstratification chlorite–vermiculite. At 550 °C no intensified peaks show up at exactly 14 Å. In the Bw and C1-horizons the chlorite-vermiculite interstratification contains more chlorite than in the surface horizon. At 550 °C the 14 Å peak of chlorite intensifies, whereas the 7 Å peak disappears: kaolinite.

Appendix 3

Clay mineralogy of the dispersed material in stream and subsurface flow: X-ray diffraction patterns and their interpretation.

The view over the year (Fig. 9.16) of the non-fractionated samples shows illite, with reflections around 10, 4.9 and 3.34 Å. Saturation with Mg²⁺ results in a 14 Å peak. After the glycerol treatment, the 14 Å peak shifts towards 15–16 Å, whereas a peak around 29 Å shifts to 32–33 Å. This implies the presence of the interstratification chlorite–smectite. After the glycerol treatment a reflection at 14 Å stays: a chlorite–vermiculite interstratification.

The fractionated sample ‘Lux 10’ shows in the fine as well as the coarse clay an illite peak. After Mg²⁺ saturation and a glycerol treatment, no shift of the 14 Å peak in the coarse clay fraction is shown, whereas in the fine clay a swelling is noticed to 16 Å: the interstratification chlorite–smectite. K-saturation and heating shows no chlorite, only an interstratification of chlorite–vermiculite. In other stream and subsurface flow samples a small chlorite peak shows up after K-saturation and heating to 550 °C. The peak at 7 Å disappears after heating to 550 °C: kaolinite.

The presence of a 14 Å peak in the dispersed stream and flow material implies that not only fine, but also coarse clay is transported. This conclusion is based on the absence of a 14 Å peak after glycerol treatment in the fine clay fraction of the reference profile.

Appendix 4: Dispersibility of the Soil Material

This section focusses on the dispersibility of soil material in the topsoil and subsurface Bg horizon of the forested Steinmergelkeuper soils.

In dispersion experiments, the potential dispersibility of these horizons has been compared. The effect on dispersion of the natural-, Ca-, and Mg-saturated adsorption complex was tested. In addition, tests have been carried out to determine the minimum concentration of an electrolyte which causes flocculation. This minimum is described as the flocculation value, or as the critical coagulation concentration (van Olphen 1977).

9.4.1 Dispersion Experiments

Soil material of AEh-,EAh- and Bg horizons of soil profile II (Appendix 1) has been used for these experiments. For detailed methods of these experiments see van den Broek (1989).

AEh-, EAh- and Bg material were used with its natural adsorption complex composition. These materials were also transformed into a completely Ca- or Mg-saturated adsorption complex.

In series with increasing soil/water ratios (1–10 g fine earth with 75 ml of demineralized water) and 24 h end-over-end shaking (21 rpm) in tubes, centrifugation was carried out to give suspensions with particles <1 and <0.2 µm respectively. To measure the degree of dispersion, the nephelometric turbidity, in the clay suspensions was estimated. Also, the electrical conductivity, pH and cation- and anion concentrations were determined in the supernatant of the centrifuged clay suspensions. Cation exchange capacity (CEC), and the main exchangeable cations, Ca and Mg, have been determined Table 9.11).

Table 9.11 CEC and saturation of the exchange complex of samples used for the dispersion experiments (van den Broek 1989)

In Fig. 9.17 the relation is shown between the increase in soil/water ratio and turbidity. Presented are the clay fractions <1 and <0.2 µm, originating from the AEh and Bg horizons, with a natural exchange complex composition (Nat), a Ca-saturated (Ca) and a Mg-saturated complex (Mg). For all the 12 subsamples the turbidity rises with increasing soil/water ratio. This implies that in the whole range of increasing soil/water ratio, even to 10 g fine earth in 75 ml water, the clay disperses. Flocculation would cause a decrease in turbidity, as there are less particles to scatter the incident light.

Fig. 9.17

Relationship between turbidity and soil/water ratio for a the clay fractions <1 µm, and b the fine clay fractions (<0.2 µm) with Ca-, Mg- or natural (Nat) exchange complex composition (van den Broek 1989)

The increase in soil/water ratio implies an enhanced release of water-soluble elements, and therefore an increase in electrolyte concentration. Figure 9.17 shows that even 10 g fine earth shaken in 75 ml of demineralized water did not release enough water-soluble elements to cause flocculation, so the flocculation value is not reached.

The clay fractions from the Bg horizon show a higher turbidity than the fractions in the AEh horizon. This absolute difference is ascribed primarily to the higher clay content of the Bg horizon. It does not automatically imply that the clay in the subsurface horizon has a higher potential dispersibility compared to the clay in the surface horizon.

The low turbidity of the samples containing organic matter shows the stabilizing effect of the organic fraction on soil aggregates in the surface horizons. Here organic matter prohibits, to a certain extent, dispersion. The effect of ‘free’ organic matter, not linked to soil particles, is not explicitly shown by this experiment. The contribution of water-soluble organic compounds on clay dispersion is discussed in Appendix 5.

The effect of a Ca or Mg-saturated exchange complex on turbidity shows up, in particular for the fine clay fraction (Fig. 9.17). The fine clay saturated with Mg has a higher turbidity compared to the Ca-saturated material. This is in agreement with (Australian) literature. Soils with a high magnesium content at the exchange complex were noticed for their bad physical structure and properties (Koenigs and Brinkman 1964; van Schuilenborgh and Veenenbos 1951). Often the dispersive effect of magnesium was only noticed with illites, and not with montmorillonites and swelling interstratifications.

Nevertheless, Shainberg and Letey (1984), Alperovich et al. (1981), Dontsova and Norton (2002) and Zhang and Norton (2002) found that at very low electrolyte concentration magnesium has a dispersive effect on clays, especially on montmorillonites. This supports the results of the experiment here, as the fine clay fractions composed of illite as well as of a swelling 16 Å mineral and the electrolyte concentration of the soil solution was always very low, a Ca- and Mg concentration around 0.3 mM (Table 9.8). The dispersive effect of the magnesium compared to the calcium ion can be explained by the larger diameter of the hydrated magnesium ion compared to the calcium ion, being 0.47 and 0.42 nm respectively (Stern and Amis 1959). Tucker (1985) showed that exchangeable Mg is slightly less strongly held at the clay surface than Ca.

A different effect of adsorbed Mg at the exchange complex for clay dispersion, is more indirect in nature. Kreit et al. (1982) studied the rate of hydrolysis of hectorite, a trioctahedral smectite. The rate of hydrolysis of the Mg-saturated hectorite was lower than that of its Ca equivalent. This is ascribed to the rate-reducing effect of exchangeable Mg on the hydrolysis of hectorite, so the presence of Mg, at the exchange complex stabilizes this clay mineral. Consequently, the hydrolysis of the Mg-clay also decreases, which prevents severe dissolution of the clay minerals. Due to the low rate of hydrolysis, the electrolyte concentration of the soil solution in equilibrium with the Mg clay is low, and dispersion will probably take place.

The soils in this study are characterized a.o. by a fine clay fraction which has much Mg adsorbed at the exchange complex (Table 9.2), which contains a swelling 16 Å interstratification (Table 9.5), and which is high in structural Mg. This fine clay fraction in particular is transported laterally in a dispersed state. The low electrolyte concentration of the soil solution and subsurface flow can be explained by the relatively low solubility of the Mg containing trioctahedral clay. Mg at the exchange complex prohibits severe dissolution of the clay minerals, and consequently the electrolyte concentration of the soil solution is kept at a low level, probably governed by the exchange reactions, which is favorable to clay dispersion.

Figures 9.17 and 9.18 show that even in a ratio of 10 g soil in 75 ml of demineralized water, not enough water extractable cations are released to cause flocculation of the dispersed clay. The potential dispersibility of the Bg horizon compared to the overlying topsoil is high due to the absence of aggregating agents, and its high Mg content at the adsorption complex, and the stabilizing effect of organic matter on the clay in the topsoil. The potential dispersibility of the Bg horizon exceeds the potential dispersibility of the topsoil also because of a higher fine clay content.

Fig. 9.18

Relationship of the concentration M²⁺ (Ca and Mg) of the solution before the start of the EOE-shaking with turbidity (a) and relative turbidity (b) of the fractions <1 and <0.2 µm of the AEh, EAh and Bg horizons (van den Broek 1989)

The adsorbed Ca/Mg ratio of the samples with a natural exchange complex (Nat) is in between the adsorbed Ca/Mg ratio of the Ca and Mg-saturated samples (Table 9.11). Figure 9.17 shows clearly that the samples with the natural exchange complex (Nat) never show a turbidity to be expected according to their Ca/Mg ratio, viz. in between the Ca and Mg-saturated samples. This “inconsistent” behaviour of the Nat-samples for turbidity and their Ca/Mg ratio is apparently due to the pretreatment of the samples and the organo-mineral bonds of the soil material. At higher organic matter levels, as in the AEh horizon (Fig. 9.17), the turbidity of the Nat-samples is the lowest. It is suggested that by saturation with Ca or Mg, the bonds between the clay, the divalent cation and the organic matter, are (partly) broken temporarily by the pretreatment. The new bonds between the freshly prepared Ca or Mg clay and the organic matter, are apparently not as strong as the bonds between the clay and the organic matter in the Nat-sample. Consequently, the Ca and Mg-saturated clays show a higher turbidity compared to the Nat-samples.

The samples with very low organic matter levels, being all horizons in Fig. 9.17 except the AEh horizons, show an opposite trend. Here the samples with a natural exchange complex have always a higher turbidity compared to the Ca and Mg-saturated clays. As is described in Appendix 5, the samples with a natural exchange complex (Nat) have been freeze-dried once, whereas the Ca and Mg-saturated clays have been freeze-dried twice. By comparing the turbidity of air dried and freeze-dried material of the Bg horizon in a separate dispersion test, it appeared that the freeze-dried samples showed a slightly lower turbidity.

9.4.2 Flocculation Test

In this test the minimum electrolyte concentration necessary to flocculate the clay fractions <1 and <0.2 µm of the fine earth of the AEh, EAh and Bg horizon has been determined. This minimum concentration is defined as the flocculation value, or as the critical coagulation concentration (van Olphen 1977; Sposito 1984).

The electrolyte used to determine the flocculation value is approximately equal to the soil solution with regard to composition and concentration. Series of samples were prepared of 6.00 g fine earth in 75 ml electrolyte solution with increasing concentration, in 100 ml polyethylene tubes. The electrolyte concentration ranges from 0 to approximately 2 mM, of an equal molar solution of CaCl₂ and MgCl₂. The soil solution contains mainly divalent cations, approximately equal amounts of Ca and Mg (Ca/Mg ratio around 1.4). The range of the electrolyte concentration in the flocculation experiment (0–2 mM) clearly exceeds the concentration of the soil solution, as the average concentration of divalent cations in the soil solution ranges from 0.15 to 0.67 mM (Table 9.8).

After 24 h End-Over-End shaking (EOE), 21 rpm, the tubes were centrifuged for the fraction <1 and <0.2 µm to measure the nephelometric turbidity of these fractions. The separation of the clay fractions and the turbidity measurement followed procedures identical to the procedures of the dispersion experiment (van den Broek 1989). Also, the turbidity and the electrical conductivity have been determined in the centrifuged suspensions too.

In Fig. 9.18a nephelometric turbidity values are plotted against the concentration of the added electrolyte solution used for the flocculation experiment. In Fig. 9.19a the same turbidity values are presented for the electrical conductivity (EC₂₅). As the release of water extractable elements is very low, there are only very small differences between the relationship ‘turbidity-added electrolyte concentration before EOE’ and ‘turbidity-measured EC₂₅ after EOE’.

Fig. 9.19

Relationship of the electrical conductivity (EC₂₅) of the solution after EOE-shaking with turbidity (a) and the relative turbidity (b) of the fractions <1 and <0.2 µm of the AEh, EAh and Bg horizons (van den Broek 1989)

The decrease of the nephelometric turbidity with increasing electrolyte concentration or EC₂₅ shows up in particular in the clay fraction of the Bg horizon. Here a distinct threshold concentration is exceeded. The S-shape decrease of turbidity over a short interval of increasing electrolyte concentration agrees well with the classical DLVO-theory (van Olphen 1977). The flocculation values of the fractions <1 and <0.2 µm are almost identical, and around an EC₂₅ of 200 µS cm⁻¹, or a divalent cation concentration of 0.8 mM. This critical coagulation concentration is in the appropriate range of 0.5–2.0 mM for clays with divalent cations at the adsorption complex (van Olphen 1977).

The nephelometric turbidity of the AEh and EAh horizons decreases too with increasing electrolyte concentration. However, this decrease is not characterized by an S-shape as the fractions of the Bg horizon are. This implies that the clay in the AEh and EAh horizons behaves less according to the concept of concentration induced flocculation.

A comparison between A and B-horizons is hard to make, because of the different clay contents. To overcome this difference in texture, measured nephelometric turbidity values have been expressed relative to the turbidity of the sample containing zero CaCl₂ and MgCl₂ (Figs. 9.18b and 9.19b). This relative turbidity is equal to the measured turbidity divided by the turbidity of the suspension in the presence of only demineralized water. The relative turbidity ranges from 1.00 (dispersion due to the shaking with demineralized water) to almost 0, being the flocculated situation due to the added concentrated electrolyte solution. These relative turbidity curves show again an ‘ideal’ S-shape decrease for the clay fractions in the Bg horizon. This in contrast to the relative turbidity values of the AEh and EAh horizons. Apparently, the clay fractions in the surface horizons are not strongly affected by the increase in electrolyte concentration as the Bg horizon is. As the Bg horizon flocculates completely (relative turbidity of 0.04) at an EC₂₅ of 200 µS cm⁻¹ or at a concentration of 0.8 mM, the AEh and EAh horizons show more dispersion, as is indicated by their higher relative turbidity values, around 0.30, at that concentration level.

This high relative turbidity implies that the flocculating effect of the concentrated electrolyte solution is opposed. This is ascribed to the stabilizing effect of organo-mineral bonds in the surface horizons, and/or the dispersive mechanism of organic anions. The aggregation in the surface horizons due to (solid) organic matter prohibits dispersion, as is shown in Fig. 9.17. The contribution to the negative charge by organic anions in the AEh and EAh horizons is considerable (ranging from 30 to 50%, Table 9.4). The dispersive effect, as stated by Oades (1984), of negatively charged organic compounds in the surface horizons, might play a role in prohibiting a complete flocculation. This will be discussed in Appendix 5.

9.4.3 Summary

The high potential dispersibility of the soil material and the flocculation values of the clay fractions are important. The potential dispersibility of the soil material in the silty surface horizons is far below the potential dispersibility of the material in the heavy clayey Bg horizon. This is ascribed to the higher (fine) clay content in the Bg horizon, and to the stabilizing effect of (solid) organic matter in the surface horizons. The potential dispersibility of the Bg-clay is relatively high, because the low amounts of water extractable elements which are released by shaking the material in water, and its low Ca/Mg ratio at the adsorption complex. The positive effect of exchangeable Mg on clay dispersion of the Steinmergelkeuper soil material compared to exchangeable Ca is also shown. The flocculation value of the clay fraction <1 and <0.2 µm of the Bg horizon is around 0.8 mM CaCl₂ and MgCl₂. These values are above the highest concentration of the subsurface flow and soil solution at the contact between EAh and Bg horizons, as sampled on the forested Steinmergelkeuper slopes (Table 9.8).

Appendix 5: Effects of Water-Soluble Organic Compounds on Clay Dispersion

9.5.1 Dispersion/Flocculation Experiments with Water-Soluble Organic Compounds Additions

Data in Table 9.4 show that the relative contribution of organic anions in various soil/water extracts to the charge balance, is considerable. These relative high concentrations of organic anions, and the DOC values of soil solution and subsurface flow, are all of a concentration level which appears to increase the resistance of clay against flocculation considerably. Based on the results and data of the laboratory experiment, an indirect effect on the dispersion of clay is ascribed to water-soluble humic substances in soil solution and subsurface flow (see also Appendix 5, as well as Bloomfield 1956; Durgin and Chaney 1984; Oades 1984). A mechanism which is likely to contribute to the dispersion domain of clay is the adsorption of soluble humic substances at clay particles. This causes steric stabilization of clay particles, which resists flocculation to a certain extent.

During wetting, forces developed inside and outside the “specific structural units” of the soil material. Soil aggregates which do not fall apart when wetted are regarded as “water stable”. In mineral soils, most of the humified material is associated with inorganic materials, particularly clays. Cementing agents, like free sesquioxides and calcium carbonate, and organic bindings are held responsible for this water stability. However, the first two cementing agents are unimportant in the solum of the Steimergelkeuper soils and only organic matter plays a role in the topsoil.

Unfortunately, up until now, no detailed research was carried out on the stabilizing effect of (solid) organic matter, e.g. on the interactions of organic compounds, with mineral (clay) material, in our Steinmergelkeuper soils.

Moreover, the destabilizing effect of water-soluble organic matter on soil aggregates and especially on dispersion of clays is not completely understood. There are some indications that these humic substances are involved in the process of clay dispersion. If bonds between clay and organic matter are destroyed by mechanical forces, for example by the kinetic impact of a heavy rainstorm, organic matter becomes attached to one simple clay particle only. This creates an extra negative charge and therefore more repulsion, so organic matter will act as a dispersant (Emmerson 1968). The dispersive effect due to adsorption of organic anions on positively charged edges of clay minerals is well known (Bloomfield 1956; Gilman 1974; Durgin and Chaney 1984; Oades 1984). Dispersion is increased due to blocking of the positively charged mineral sites by negative organic anions (Shanmuganathan and Oades 1983). In addition, complexation or chelation of polyvalent cations in the soil solution by dissolved organic matter might occur, reducing the activities of these cations and resulting in a more extended diffuse double layer around clay particles and they will disperse more easily.

In this section, the results of some experiments with natural and synthetic water-soluble organic compounds on the dispersion and flocculation on 2:1 clays with a divalent cation occupation will be discussed. In addition, a link to field conditions was made by using water extracts of forest litter, collected in autumn in the subcatchment in the experiments.

A dispersion/flocculation experiment with increasing salt level was carried out with the addition of various water-soluble organic compounds. The clay used in the experiments originated from the Bg-horizon of the reference profile (profile II, Appendix 1), and was saturated with Ca.

A fulvic acid extracted from a spodosol, and a synthetic citric, salicylic or p-hydroxybenzoic acid were added in various concentrations to the clay suspensions. Flocculation curves of the Ca-clay with these organic compounds in various concentrations, were determined. A range of 50 ml polyethylene tubes with the Ca-clay was prepared, containing an increasing concentration of CaCl₂ (0.1–2 mM). The tubes, with a final volume of 40 ml and a clay concentration of 1 g L⁻¹, were shaken end-over-end. After a certain settling time, the fraction <2 µm has been pipetted from the tubes in a cuvet. The nephelometric turbidity of the suspension in the cuvet was measured.

By plotting the measured turbidity data against the increasing electrolyte concentration in the tubes, flocculation curves of the Ca-clay have been determined, with and without the water-soluble organic compounds mentioned above.

In Fig. 9.20 flocculation curves of the Ca-clay are presented for relevant levels of fulvic acids. The curves show the expected S-shaped relationship between increasing salt concentration and turbidity, corresponding to the flocculation curves of the clay from the Bg horizon (Fig. 9.18). In the absence of fulvic acid the flocculation value of the Ca-clay is 1.0 mM CaCl₂ which is in agreement with earlier results (Appendix 4) and other findings (van Olphen 1977; Sposito 1984). In the presence of fulvic acid, a higher salt concentration is required to flocculate the Ca-clay compared to the clay suspensions without fulvic acid. Addition of fulvic acid to the Ca-clay causes an increase in turbidity and dispersion. From these results, it can be concluded that in this experimental design the natural water-soluble organic compounds have a dispersive effect on the Ca-clay.

Fig. 9.20

Flocculation curves of the Ca-clay in presence of fulvic acid (van den Broek 1989)

When a small amount of organic polyelectrolyte is added to a clay suspension, the salt tolerance of that clay suspension is considerably increased, even to the extent that the clay remains in dispersion in a concentrated salt solution.

The synthetic low molecular water-soluble organic compounds have, in contrast with fulvic acids, not only a dispersive, for the carboxylic(citric) acid, but also a flocculating effect on the clay suspensions for the aromatic acids, e.g. salicyclic and parahydroxybenzoic acids. In Fig. 9.21 the flocculation curves of the Ca-clay in presence of the various synthetic organic acids are shown.

Fig. 9.21

Flocculation curves of the Ca-clay in presence of synthetic acids (phb acid = p-hydroxybenzoic acid) (van den Broek 1989)

In Fig. 9.22 the relation between added acids and the turbidity of the clay suspension is presented. A small amount of fulvic and citric acid added to the clay suspension is responsible for a large increase in turbidity, e.g. the dispersive domain, whereas aromatic acids additions cause flocculation for the complete concentration range. Light absorbance by the brown coloured fulvic acid solution (without clay) is held responsible for the somewhat decreasing turbidity at high fulvic acid concentrations, as resulted from a separate test.

Fig. 9.22

The influence of the water soluble organic acids on the dispersion of the Ca-clay, at a concentration of 0.5 mM CaCl₂ (van den Broek 1989)

The two aromatic acids, salicylic and p-hydroxybenzoic acid, have a distinct flocculating effect. This in contrast to the citric acid, which shows an increase in dispersion of the Ca-clay for low concentrations. The dispersive whether flocculating effect of a water-soluble organic compound on clay depends on the interaction of the humic substance with the solution in which clay dispersion takes place, and on the bond between the humic substance and the clay mineral. The limited set of data, the complexity of humic substances and the many possible interactions between clay and organic compounds, exclude hard conclusions about the dispersive mechanism of humic substances. Therefore, the interpretation of the experimental data concerning the mechanism of humic substances involved in the dispersion of clay, has just a tentative character and will be summarized.

With regard to the increased dispersion of the Ca-clay caused by addition of fulvic and citric acid, cation complexation and ion–dipole interaction cannot be excluded beforehand. However, the similar response of the Na-clay (data not presented) compared to the Ca-clay on citric and fulvic acid rules out complexation and ion–dipole interaction as major mechanisms involved in the dispersion of the (Na as well as the) Ca-clay.

Another mechanism which can be involved in the interaction between water-soluble organic compounds and clay minerals is the adsorption of humic substances at the clay surface. Well known is the charge reversal at the edges of clay minerals by adsorption of small amounts of (organic) anions by ionization of the carboxylic groups (Shanmuganathan and Oades 1983). The results of the experiments here agree well with this model, as small amounts of citric and fulvic acid give a considerable increase in the dispersion domain of the clay (Fig. 9.23). However, the clay mineralogy of the soil material from the Bg horizon (Table 9.1) points towards a relatively small edge charge. Therefore, another adsorption mechanism then charge reversal is worthwhile considering. Results of the experiments show that the dispersive effect of the low molecular citric acid holds only for low concentrations (Figs. 9.21 and 9.22). On the contrary, fulvic acids (Figs. 9.20 and 9.21), which are high in molecular weight, have also a dispersive effect at high concentrations. A bond between the clay mineral and the relatively high molecular fulvic acid is likely.

In addition to this mechanism of surface adsorption, the intercalation of uncharged parts of the fulvic acid in the interlayer of the clay mineral has also to be considered, as the pH of the solution in the experiments was around 4. The optimum pH for interlayering fulvic acids in Na-montmorillonite is around 2.5 (Schnitzer and Kodama 1966), at the prevailing pH conditions in the experiments interlayer adsorption cannot be ruled out on forehand. Notwithstanding that, no differences in X-ray diffraction patterns showed up between clay samples shaken with fulvic acids and those that were not. This implies that intercalation of fulvic acids in the soil material from the Bg horizon, can be excluded.

Summarizing, it is very likely that high molecular fulvic acids are adsorbed at the external surface of the clay. The non-flexible tails of the fulvic acid extend into the solution, and prevent the mutual attraction of particles. In this way, steric stabilization resists flocculation and contributes indirectly to the dispersion of the Ca-clay.

Adsorption of organic acids at the clay is supported by the data presented in Table 9.12. Fulvic and citric acid was added to the Ca-clay suspensions. After shaking, the water-soluble organic compounds were determined, and expressed as dissolved organic carbon (DOC). The clay suspensions with added fulvic acid contain approximately 5 mg C L⁻¹ (DOC), whereas the clay suspensions with added citric acid contain approximately 7 mg C L⁻¹ (DOC). From this data and the blank, containing about 8.5 and 9.0 mg C L⁻¹(DOC), it is concluded that a relative considerable amount of the high molecular fulvic acid is adsorbed at the clay surface. The adsorption of the small citric acid at the clay is less (see also Oades 1984).

Table 9.12 DOC in the Ca-clay suspensions after EOE-shaking (van den Broek 1989)

9.5.2 Dispersion/Flocculation Experiments with Forest Litter Extract Additions

The preceding experiments show a.o. the positive effect of a fulvic acid extracted from a spodosol, on the dispersion of clay from the Bg horizon. In addition, clay dispersion under so-called field conditions was also carried out. Therefore, beech-oak litter lying on the forest floor at the experimental field site was collected and used also in a similar dispersion experiment.

Litter collected in autumn, was extracted with water in a ratio of 1:10 (w/v). This litter extract was diluted and contained 100 mg C L⁻¹ (DOC).

Chromatographic separation by gel filtration of this extract with a Cu(II) solution as eluent, in order to get also information on the complexation capacity, was carried out to separate the water-soluble organic compounds on molecular weight. The fraction eluated first contained the relatively large molecules (M > 5000 D). The second fraction contained the smaller molecules with a non-aromatic character (M < 5000 D). To this last group the category of a.o. fulvic, citric and oxalic acid belongs. Large amounts of Cu(II) were complexated by the small as well as the large water-soluble humic substances.

Gel filtration of an extract of soil material from the AEh horizon in a soil/water ratio of 1:2 (w/v), showed only minor amounts of small and large organic molecules, with a small complexation of Cu(II). Gel filtration of the soil solution, sampled by means of a porous cup at the interface between EAh and Bg horizons, did not show organic molecules at all. Consequently, the soil solution did not contain Cu-complexating water-soluble organic compounds.

Considering the DOC values of the extract of the soil material of the surface horizon and the soil solution, 34 and 20 mg C L⁻¹ respectively, it is not so surprising that gel filtration comes up with small amounts of water-soluble organic compounds and low values of Cu-complexation. In addition, part of the humic substances is probably already complexated or adsorbed at the clay surface.

Anyway, the litter extract was considered as a kind of concentrated soil solution with regard to the composition of water-soluble organic compounds. Chromatographic separation showed that the litter extract contained Cu-complexating water-soluble organic compounds with a low as well as with a high molecular weight. Therefore, it was assumed that the litter extract has a similar effect on the dispersion of clay as the soluble organic compounds used in the preceding experiments. These experiments showed that relatively large natural organic compounds and small, synthetic organic molecules with charged carboxyl groups, contribute considerably at the prevailing pH conditions to the dispersion of the Ca-clay (Figs. 9.20, 9.21 and 9.22). The litter extract contributes to the dispersion of clay, which is confirmed indeed by results of the additional dispersion experiment.

In Fig. 9.23a the dispersive effect of a diluted litter extract is clearly shown.

Fig. 9.23

a Flocculation curve of the Ca-clay in the presence of the diluted litter extract; b the influence of this diluted litter extract on the dispersion of the Ca-clay at 0.5 mM CaCl₂ (van den Broek 1989)

Even at the flocculation value of the Ca-clay, 1 mM CaCl₂, a low concentration of humic substances (3 mg C L⁻¹ DOC) causes a considerable increase in clay dispersion. Figure 9.23b shows even more clearly that very small amounts of litter extract cause a considerable increase in the dispersion of clay. The content of dissolved organic carbon in the soil solution ranges from 11 to 20 mg C L⁻¹ (Table 9.8). In Figs. 9.22 and 9.23 it is shown that humic substances in such a concentration range contribute to the dispersion of clay. Based on these results, a positive (indirect) contribution to the dispersion of clay is ascribed to water-soluble organic compounds with charged carboxylic groups under near neutral pH conditions.