(Un)saturated soil water transport characteristics of the spodic horizon

(1)

(Un)saturated soil water transport characteristics of the spodic horizon

The inspiration of the SoSEAL project

Bachelor project Earth Sciences June 2017

Boonman, J. 10735143

Supervisors: dr. Jansen, B MSc. Brock, O.P. MSc. Zhou, J.

(2)

2

Abstract

The SoSEAL project is based on the permeability reduction in highly permeable layers by aluminium and dissolved organic matter (DOM) precipitation and complexation, which is observed in the upper illuvial spodic Bh horizon of podzols (e.g. De Coninck, 1980; Farmer, 1982; Lundström et al. 2000). The DOM-Al complexation in the highly permeable soil layers is meant to manipulate soil water transport

characteristics in such a way that permeability of the layers is reduced and water leakage is prevented. The SoSEAL method could be the ideal solution for increasing the safety, sustainability and effectiveness of key water management elements with very specific targeting (Heimovaara et al., 2014). Although the project is built on the hydraulic properties of the Bh horizon, data are scarce. Within this research, we aim to assess the change in (un)saturated soil water transport properties (the saturated hydraulic conductivity (Ksat) and water retention curve) after the natural formation of spodic horizons to gain knowledge on the quality of the SoSEAL project, hydrodynamics and soil formation processes in areas with well-developed Bh horizons. Two well developed spodic (Bh) horizons and two corresponding parent material (C, Bs/C) horizons were tested on Ksat and water retention. Results indicated that the development of a Bh horizon is responsible for changing the soil pore structure, due to the formed Al/DOM complexes. The hydraulic conductivity of the layer is reduced up to 94.82% and for-plant-available water holding capacity increased with 28.95%. These results are in line with earlier results found by Mecke & Ilvesniemi (1999) and indicate a potential agricultural appliance of an ‘artificial’ Bh horizon, for which further research is suggested. Finally, the SoSEAL method has proven to be more effective in permeability reduction than is realized within the development of the Bh horizon of podzols. However, little is known about the stability of the formed complexes in the SoSEAL project so no long term conclusion can be drawn. Further research on the stability of the formed complexes in the SoSEAL project is suggested.

Cover page picture: Microscope picture (x60) of a cemented spodic Bh horizon with continuous cracked coatings covering the sand grains (De Coninck, 1980; Edafologia (n.d.).

(3)

3

Acknowledgements

First of all, I especially would like to thank dr. Jan van Mourik, who helped me with the fieldwork preparation, transport, locations and observations. Not to mention all the lovely stories he told; he could start a radio station. Second, I’d like to thank the two PhD’s, Olaf P. Brock and Jiani Zhou for their true involvement. You spend lots of time inspiring and correcting me. Olaf helped me to gain knowledge on the most important topics before starting my research proposal. He also played a very important role discussing his ideas about the research design. Then, thanks to Jiani, the underlying (un)saturated soil water transport physics were clarified. Without him, the days in the lab and the lunches would have been much more boring. Finally, I thank Boris Jansen for his sincere involvement and the sharing of this involvement in the SoSEAL project. Furthermore, his enthusiastic scientific expertise has been a

fundament in my career. Together with Olaf he provided very specific feedback, which made certainly sense in developing my research and presentation skills.

(4)

4

Introduction

This bachelor research project is part of the Soil Sealing by Enhanced Aluminium and DOM Leaching (SoSEAL) project. The aim of the SoSEAL project, is ”to develop a bio-based geo-engineering technology for in-situ permeability reduction that will be applied to reduce the infiltration loss or the seepage burden due to unwanted flow of water through highly permeable layers in the sub-surface’’

(Heimovaara et al., 2014). The SoSEAL project is inspired by the permeability reduction of soil layers due to the precipitation of aluminium and dissolved organic matter (DOM), which is observed in the spodic (upper illuvial, Bh) horizon of podzols (e.g. De Coninck, 1980; Farmer, 1982; Mecke & Ilvesniemi, 1999). This spodic horizon has been formed by natural soil formation processes over hundreds to thousands of years and can be strongly cemented (Lundström et al. 2000). In the SoSEAL project, the process of aluminium and DOM precipitation is artificially imitated by injecting aluminium and organic matter simultaneously in a highly permeable soil layer (for example as in figure 1). The precipitation and complexation process is happening in the soil layer after the groundwater flow mixes the compounds. The DOM-Al complexation in the highly permeable soil layers is meant to change soil water transport characteristics in such a way that permeability of the layers is reduced and water leakage is prevented. The unwanted seepage can be prevented with very precise targeting (Heimovaara et al., 2014). In the SoSEAL pilot project ‘Veersedijk’ (figure 1) the saturated hydraulic conductivity in a sand layer has been reduced with 97.5%. If the formed complexes are stable in the long term, the technique will be a cost effective, sustainable, ecological friendly solution for, for example, seeping dyke bodies, lakes and sewage systems. The SoSEAL method could be the ideal solution for increasing the safety, sustainability and effectiveness of key water management elements (Heimovaara et al., 2014).

For the development of the SoSEAL project, crucial knowledge on three topics is required:

1. To what extend the (im)mobilization and complexation of metals and (D)OM is controlled by pH, grain size, metal-to-carbon ratio, substance concentration and redox potential.

2. To what extend the stability of the formed metal-(D)OM complexes is controlled by these factors.

3. To what extend the (un)saturated soil water transport characteristics (saturated hydraulic conductivity (Ksat) and water retention) of these soil layers are influenced by the factors mentioned above.

This study is related to the third topic. We aim to assess hydraulic properties (Ksat and water retention curve) of two well-developed spodic horizons that have been formed by natural podzolization. The main question of the research is: To which extent are the soil water transport characteristics influenced by the formation of a spodic horizon? The research contributes to knowledge of hydrodynamics and soil formation processes in areas with a well-developed illuvial Bh horizon. Moreover, other applications of OM/Al injection are proposed. Finally, the influence on Ksat of the formation of an upper illuvial horizon is compared with the influence on Ksat of Al/DOM injection in the highly permeable layers of the SoSEAL pilot ‘Veersedijk’, presenting an indication of the quality of the method.

5

(6)

6

Research background

Podsolization process

A podzol is defined by having a spodic horizon starting =< 200 cm from the mineral soil surface (IWG,

2014). This reddish-brown to black illuvial horizon (B) is enriched in organic material, Al and Fe which

originates from the horizon above. This weathered ash-grey albic eluviation horizon (E) contains less Fe and Al than the parent material (C), situated beneath the B horizon, and is depleted from the organic matter that has been leached from biological input (Lundström, 2000; Scheffer & Schachtschabel, 2015). The E horizon is overlain by a darker A horizon and an organic surface horizon (O). In the process of podzolisation, Fe, Al and organic matter have been mobilized in the albic (E) horizon and immobilized in the spodic B horizon. This mobilization is mainly caused by humic and fulvic acids originating from soil organic matter input and a high precipitation level or wet environment. The humic and fulvic acids are able to suspend in the available water, infiltrate into the soil and mobilize other forms of organic carbon and heavy metals such as iron and aluminium. Partly because of the lower solubility (higher pH) in the deeper B (or Bsh) horizon, the cations (metals) are filling up the highly available binding sites of the organic matter (Scheffer & Schachtschabel, 2015). There are several, much more detailed theories about the processes which control the (im)mobilization of Al and Fe or podsolization (Buurman et al., 2005; Lundström, 2000); however, these theories are beyond the scope of this research. A distinction in illuvial horizons of a well-developed podzol can be made: the upper black illuvial horizon (Bh), consisting of a relatively high organic carbon, aluminium and moisture content and a rather low bulk density and the lower red-brown illuvial horizon (Bs), consisting of a relatively high iron content (Hirsch et al., 2017). Depending on the degree cementation by illuvial iron and/or organic matter, unconsolidated to slightly cemented Orterde to massive Ortstein horizons can be distinguished (De Conick, 1980; Hirsch et al, 2017).

(Un)saturated soil water transport characteristics

The (un)saturated soil water characteristics are defined as water retention and saturated hydraulic conductivity (figure 2, 3). Both soil properties are important considering water infiltration rates and hydraulic conductivity. Water retention is directly related to water content, which is related to effective saturation and relative conductivity (Heimovaara, 2017). Moreover, the water retention curve (here defined as the van Genuchten equation, a combination of equations 1 and 3) holds information on capillary rise above the groundwater level, water holding capacity, pore space and minimum (θres)and

maximum (θsat )soil water content. The factors mentioned above are important considering soil water

dynamics and by knowing them it is possible to model hydraulic dynamics for a whole area. Seff (hc) = [1+(a*hc)n_]-m_{for hc >0} ₍₁₎

Seff (hc) = 1 for hc =< 0 (2)

Seff =

(θ

w

– θ

res

)/( θ

sat

-

θ

res

)

(3)

K

rel

= S

eff3

[0,1]

(4)

Equations 1 to 4. With capillary head hc [cm], scale parameter related to the inverse of the air-entry value a [cm-1], shape

parameter related to the rate of water extraction from the soil once the air-entry value has been exceeded n, ], m, 1/1-n

[-],effective saturation Seff [-], volumetric water content θw [-], residual volumetric water content θres [-],saturated volumetric

(7)

7

Causes of change in water transport characteristics

Mecke & Ilvesniemi (1999) suggested that the oxides of iron and aluminium, fine weathering products and non-decomposable organic compounds build coatings on the sand particles and glue the particles together while immobilization takes place. On the cover page (figure description on page 2), the cracked coatings on sand grains in a spodic horizon are captured. De Conick (1980) describes the ‘gluing’, which influences hydraulic soil properties, in a more detailed way: in the very first stage of immobilisation of Al/Fe and organic matter, the complexes are not in a hard, but more flexible ‘gel’ state. Three processes that influence the dehydration of complexes are described: “(1) decreases in charge of the organic matter at the moment of the reaction with cations, (2) increases in concentration inside voids of the ions originally present in free form in solution and (3) (partial) desiccation.’’ The dehydration

(desiccation) takes place because charge decreases. This results in smaller distance between different metal-OM complexes and subsequently leads to the formation of Van der Waals bonds and protonic bridges. As a result, the complexes become larger. According to Mecke & Ilvesniemi (1999), as a consequence, a structure with irregular, small- and medium-sized pores is created. In such a way, tortuosity is increased and porosity is decreased reducing hydraulic conductivity. Moreover, the smaller pores increase suction and water holding capacity in the soil (Bouma, 1983; Fitts, 2002). As hydraulic properties differ for each soil texture, particle size and bulk density, those factors are also considered to be of importance in relation to changes in (un)saturated soil water transport characteristics within the formation of spodic horizons (Fitts, 2002).

Current knowledge

Studies on actual hydrological properties of spodic horizons are scarce. In the research of Bormann (2008), saturated- and unsaturated hydrological conductivity capacity of the top and subsoils of three different podzols and stagnosols are measured. Unfortunately, no distinction in horizons has been made so no conclusions could be drawn about change in hydraulic properties of the spodic horizon. In other research, Ksat values of two Bh horizons have been extrapolated. Results indicate that, compared to the parent material, the spodic horizon can reduce the Ksat up to 95,93% (factor 25) (Mecke & Ilvesniem, 1999). However, these results have to be treated carefully as extrapolation increases uncertainty. Even though most literature describes a reduced hydraulic conductivity of the Bh horizon (compared to the parent material), also an increased conductivity has been found for the horizon (Mehuys, 1976). Considering water retention, Mecke & Ilvesniem (1999) found a higher water holding capacity for the horizon. All results considered, there is still a research gap of soil water transport characteristics of spodic horizons.

Hypothesis

Because hydraulic properties will be influenced by many factors, the research focus will be on the change in (un)saturated soil transport properties due to the formation of Al/Fe OM complexes in the cemented spodic horizon.

The cemented spodic horizon with parent material of medium to fine sand is expected to behave as a sandy loam soil, as was found in the research of Mecke & Ilvesniemi (1999). The water holding capacity of a soil that is available for plants, defined as the difference in volumetric water content at pF 4.2 and pF 2.0, is higher for a loamy soil compared to a sandy soil (figure 2). Moreover, with reference to figure 3, the reduction of Ksat of the upper illuvial Bh horizon could fall in between one or two orders of magnitude.

(8)

8

Figure 2. Soil water retention curves (Lajos,2008) Figure 3. Material Ksat values for medium bulk density

(NRCS (n.d.)

Methodology

In this research, two different podzols in the province Noord-Brabant, the Netherlands were

investigated (maps can be found in appendix A). Both podzols have a clear, cemented, thick (>15 cm), well-developed spodic Bh horizon, which is buried not too deep. The two particular podzols were chosen because of their contrast in activity. Experimental tests were executed to measure (un)saturated soil water transport characteristics of the two Bh and C (Bs/C) horizons. Photos of the horizons of interest can be found in appendix B.

The first soil is a non-active carbic podzol located in the Bedafse Bergen. It had developed in cover sand and is buried under a haplic arenosol with drift sand as parent material. The soil material texture was medium to fine sand for the whole profile. The second soil was a still active, younger podzol, located in the Schaijkse Heide. This podzol has formed in the upraised Peelhorst. On top of the profile (A, E, Bh horizon) the soil texture was medium to fine sand (cover sand). Lower in the profile, fluvial deposited material could be found. The material of the Bs/C horizon consisted of mainly loamy sand (Brock, 2017– in preparation). Pine forestation could have disturbed the upper soil (about 0.5 m).

First, a horizontal plane (appendix B, figure 16) had been created in the field. Then, a field test had been conducted to give an indication of infiltration rate. After this, samples were collected, labeled and taken to the lab to be further analyzed with the KSAT and HYPROP devices. The testing methods had been selected to be time effective and cheap. Moreover, the test had to be suitable for determining low Ksat values ranging from 10-5_{– 10}-8 _{cm/s (layered and unweathered clay, sandstone), which could be} expected for a cemented spodic horizon (Bear, 1972).

(9)

9

Field

On the one side of the horizontal plane, an infiltration test had been conducted. The infiltration was measured with the single ring infiltrometer test (FAO, 1979; Hsin-ya Shan, n.d.). The main testing material was a transparent PVC ring with a diameter of 80 mm and height of approximately 120 mm. Measuring tape had been attached on the tube. The ring was put 1 cm in the horizon with a vibration free hammer to prevent lateral surface flow and soil disturbance (Ankeny et al., 1989). The inside of the ring was covered with plastic foil. As much tap water as possible was poured into the ring. The plastic was removed and after each minute the water level was recorded. If there was enough time available, the test was repeated four times.

On the other side of the plane, samples of the horizon were taken. To take the samples, UMS sampling rings (250 mL, inside ⌀ 80 mm)were driven straight into the soil with a vibration free hammer (UMS, 2015). The samples were taken out of the soil with a knife and top and bottom were flattened. The samples were put in foam UMS transport boxes to prevent vibration and soil disturbance. The major challenge in testing samples on hydraulic

properties is the quality and representativeness of the soil samples. When taking samples, the original soil is always disturbed. However, one moment of inattention within the sampling process could cause considerable changes in pore size and abundance. This leads to

inaccurate hydraulic property estimations in the lab (Dirksen, 1999; Fitts, 2002; UMS, 2012). To anticipate to the (unpreventable normal and possible major) soil disturbance, 4 to 5 samples of each horizon were taken with the UMS rings. Finally, two samples of each horizon were taken in a bag to determine moisture content.

Lab

Hydraulic conductivity- UMS KSAT

To ensure saturation of the samples, the sample rings were placed on a porous plate in a container. The container was filled to 5 mm beneath the sample edge with degassed water in two time steps. As a consequence of capillary suction, the soil filled itself with water without trapping air (UMS, 2012). The saturated hydraulic conductivity is approximated by a falling head test with the UMS KSAT. A burette filled with water was connected with the bottom of the sample. After opening the burette, water could flow through the sample. The flow was established by the pressure imbalance caused by the difference in hydraulic head dh [m] of the water in the burette and the water on top of the sample (figure 4) (Fitts, 2002; UMS, 2012).

The burette was filled with water and when opened, the experiment started automatically. The device recorded the decrease in hydraulic head over time. With Darcy’s law (equation 5) and a fitted

exponential function, the Ksat of the sample was estimated by the device. Each soil sample had been measured two to three times. Outliers had been removed and extra measurements were done if needed.

(10)

10 𝐾𝐾 =𝑄𝑄_𝐴𝐴_𝑑𝑑ℎ𝐿𝐿 and 𝑄𝑄 = −𝑎𝑎𝑑𝑑ℎ_{𝑑𝑑𝑡𝑡}

Equation 5. Darcy’s law, with the hydraulic conductivity K [m/s] discharge Q [m3/s], the sampling area through which the water flows A [m2_{], the distance through the sample L [m], burette cross-sectional area a [m}2_{], change in water level in the burette dh}

[m] and change in time dt [s] (Darcy, 1858; Fitts, 2002).

Water retention- UMS HYPROP

One saturated sample of each horizon of the Bedafse Bergen profile had also been tested with the UMS HYPROP (figure 5). Because the soil surface is open, water was able to evaporate from the soil sample. Two tensiometers were measuring the tension in the pores of the sample at two different depths, while the water was evaporating out of the sample. The mass of the sample was also recorded over time. When air had entered both tensiometers and evaporation stopped, the measurement was finished.

After the measurement period, the samples were put in the oven at 105 °C for 24 hours in order to weight the dry bulk density. Within the HYPROP measurement, the sample mass was weighted at

each step in time. The difference in mass between the dry bulk density and the mass of the sample at time t represented the water mass. With the mass and density of the water, the volume of the water at each step in time was calculated. With the volume of the water and the volume of the sample, the volumetric water content (

θ

i

)

in the soil sample was determined for each time step. The two water tensions had also been measured for each time step. By taking the average of the two tensions, the capillary head (hc) is determined. In this way, the water content and capillary head are calculated for each step in time.

The measured values then were empirically fitted in MatLab (version R2016B; MathWorks) with a combination of the van Genuchten equation (equations 1 and 3), by changing α [cm-1_{] and n [-]. α and n} are the shaping and fitting parameters of the water retention curve. α is primarily related to the inverse of the entry value. n is primarily related to the rate of water extraction from the soil once the air-entry value has been exceeded (Fredlund et al., 2012).

Moisture content

The two samples of each horizon that were taken in a bag were used to determine moisture content.

Approximately ten grams of each sample was put in an oven at 105 °C for 24 hours. The soil mass was

weighted before and after. The difference in mass represented the mass of the water that had evaporated. The water mass was divided by the dry sample weight to obtain the moisture content.

(11)

11

Results

The C horizon of the Schaijkse Heide podzol was more difficult to reach than expected. Therefore, the C horizon had been replaced by the shallower Bs/C horizon. Furthermore, due to a lack of time it was not possible to determine the water retention curves of the Schaijkse Heide horizons. In this section, the results for each location will be presented separately.

Bedafse Bergen

Figure 6.

Above: the moisture content and the mean infiltration rates with standard error of the Bh and C horizon podzol located in Bedafse Bergen. Below: infiltration of all tests on the Bh and C horizon.

The mean infiltration rates of the Bh (0.006 cm/s SE 0.0027) and C horizon (0.058 cm/s SE 0.0041) differed with almost factor ten (figure 6). The mean of the moisture content of the Bh horizon was 41.56% with a standard error of 4.72%. The mean moisture content of the C horizon was 4.81% with a standard error of 0.36%. The soil moisture of the Bh horizon was about 8 times higher than of the C horizon. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Hei gh t [ cm ] Time [s] Infiltration experiment Bedafse Bergen 'B horizon' 'C horizon' 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 Bh C In fil tr at io n r at e [ cm /s ] Mean infiltration 0 10 20 30 40 50 Bh C M oi st ur e c on ten t [ % ] Moisture content

(12)

12 0 100 200 300 400 500 600 700 800 900 1000 1100 B C Av er age K sa t [ cm /d ay]

Saturated hydraulic conductivity

The average Ksat of the Bh horizon (47.2 cm/day SE 7.94) was nearly 20 times lower than of the C horizon (910.5 cm/day SE 134.45), which corresponds to a reduction of Ksat of 94.82% (figure 7).

Table 1 shows the most important values and parameters of the fitted water retention curve. The α and n values are shaping the water retention curve in figure 8. The approximation of the fitting can be found in appendix C. The water available for plant root uptake (moisture content between pF 2- 4.2) in the Bh horizon was 29%, in the C horizon 0.066%. Moreover, a graph representing relative conductivity against saturation is presented in figure 9.

Figure 7. Average saturated hydraulic conductivity (Ksat) values with corresponding standard errors of the Bh and C horizon of the podzol located in Bedafse Bergen.

Table 1 . Results of the HYPROP measurements and manual parameter estimation process.

Horizon

Bh

C

α [cm

-1

_]

_1.3

_1.9

n

1.5

7 Saturated

water content

0.52

0.39 Residual water

content

0.087

0.097

Figure 8. The water retention curves of the Bh and C horizon that have been fitted manually with results of the HYPROP test. On the y-axis the pF value, which is defined as log-(water head [cm]). The prediction plotted actual measured data can be found in appendix A.

(13)

13 0 20 40 60 80 100 120 140 Bh Bs/C Av er age K sa t [ cm /d ay] Horizon

Saturated hydraulic conductivity

Figure 9. Relative K values of the Bh and C horizon. The curves originate from the fitted water retention curves. Schaijkse Heide

At the Bs/C horizon, red spots ( ⌀ >10 cm) were visible on the horizontal plane. This indicated there is an iron concentration heterogeneity in the layer. A more red surface indicates a higher concentration of precipitated iron complexes. Moreover, fine to medium, very few to few roots were found in the Bh layer (size and abundance estimated according to Jahn et al. (2006)).

The mean infiltration rate of the Bs/C horizon (0.011 cm/s SE 0.0027) was significantly higher (α= 0.05) than the rate of the Bh horizon (0.0040 cm/s SE 0.00055) (figure 11). The soil moisture of the Bh horizon (33.26% SE 8.64) was about 5 times higher than the soil moisture of the C horizon (7.32% SE 0.094). The average Ksat of the Bs/C horizon (12.33 cm/day SE 1.45) was roughly 8 times lower than of the Bh horizon (104.4 cm/day SE 16.33), which corresponds to a reduction of 88.18% (figure 10).

Figure 10. Average saturated hydraulic conductivity (Ksat) values with standard errors of the Bh and Bs/C horizons of the podzol located in Schaijkse Heide.

(14)

14 Figure 11. Infiltration rates of the Bh and Bs/C horizons of the podzol located in Schaijkse Heide, the moisture content and the mean infiltration rate with standard error per horizon.

00 02 04 06 08 10 12 14 0 500 1000 1500 2000 2500 3000 Hei gh t [ cm ] Time [s] Infiltration experiment

Schaijkse Heide B horizon

C horizon 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 Bh B/C In fil tr at io n r at e [ cm /s ]

Mean infiltration rate

0 5 10 15 20 25 30 35 40 45 Bh Bs/C M oi st ur e c on ten t [ % ] Moisture content

(15)

15

Discussion

Bedafse Bergen

The results of the Bedafse Bergen profile show large variety in the infiltration rates of the Bh horizon (factor 20). This could either be due to heterogeneity in the soil, to disturbance of the soil or to a difference in soil moisture content. Unfortunately, the testing infiltration tube was not ideally prepared, which could have caused disturbance while it was hammered into the soil. In the results of the Ksat test, Bh samples showed less variance. This indicates that soil disturbance due to the experiment could be responsible for the different infiltration rates of the Bh layer. Moreover, a high spatial soil moisture variance could cause a different infiltration rate as relative conductivity is directly related (figure 9). However, in a very extreme situation of spatial moisture variance (difference of 10%) a maximum change of factor 3 in conductivity is reached, so the infiltration variance could not be completely explained by soil moisture irregularity.

The infiltration rates of the C horizon consist of a much lower variance compared to the mean. The infiltration rate of the Bh horizon is around ten times lower, even though the moisture content is higher. In a normal situation, a higher moisture content would result in a higher infiltration rate (Fredlund et al., 2012). It is expected that due to the specific properties in the Bh horizon –the Al/Fe DOM complexes changing pore size and abundance – the moisture content is high and the Ksat is reduced. Furthermore, the lab measurements on the Bedafse Bergen soil samples confirm the field findings of reduced

permeability of the Bh horizon in saturated conditions. The Bh horizon is responsible for an average Ksat reduction of almost 20 times.

The differences in (un)saturated water transport characteristics of the Bh and C horizon are considerable (table 1, figure 8). Besides the differences in shapes of the water retention curve, high differences in saturated water content (13%) have been found. The parameter values result in an ideal situation for plant water uptake, as 29% of the pore volume can be used for this purpose.

In the upper illuvial horizon of the Bedafse Bergen profile high TOC and aluminium values were found by Brock (2017 – in preparation). Both the results that have been found for the Bedafse Bergen profile in the research of Brock (2017 – in preparation) and in this research are in line with the results of Mecke & Ilvesniemi (1999). Namely, Mecke & Ilvesniemi (1999) also found a lower conductivity of at least one order of magnitude, a relatively high TOC and a relatively high aluminium content and a high water holding capacity for a Bh horizon in medium coarse sand.

Schaijkse Heide

The infiltration rates of the Bs/C horizon showed unexpected high variance. It seems that the difference in the results is caused by the iron heterogeneity in the layer. Unfortunately, only two tests have been done due to an error in the measuring process; when the measurement was prepared, the infiltration tubes were disturbed by accidentally touching them with a foot.

The overall water infiltration was higher for the Bs/C horizon than for the Bh horizon. Nevertheless, the Ksat value is roughly ten times lower for the Bs/C horizon compared to the Bh horizon. On one hand, high iron concentration is expected to play an important role in Ksat reduction of Bs/C horizon; by the

precipitation of iron compounds, the pores are clogged and hydraulic conductivity is decreased. On the other hand, root and forestation disturbance is likely to have increased the Ksat of the Bh horizon. It is

(16)

16 expected that as a consequence of root and forestation disturbance, the average Ksat of this Bh horizon found to be two times higher than the average Ksat of the Bh horizon of the Bedafse Bergen profile.

General

As described in earlier research, the water holding capacity of a spodic horizon is very high (Mecke & Ilvesniemi, 1999). The moisture content of the Bh horizon is up to eight times higher than the moisture content of the C or Bs/C horizon. This is visualized by figure 6 and 11 (moisture content). Moreover, also findings from Brock (2017) indicate a relatively high moisture content for the Bh horizons investigated in the research, including the two of this research. The water retention curve of the Bedafse Bergen upper illuvial horizon gives more insight in the water holding capacity (figure 8). The Bh horizon is found to be able to hold 28.95% more water that is available for plants. The increased water holding capacity of the Bh layer could be of interest for agricultural purposes. In fact, the root abundance in the profile of Schaijkse Heide might suggest that the horizon is preferred by plants.

Additionally, the achieved permeability reduction quality in the SoSEAL pilot ‘Veersedijk’ is high

compared to natural permeability reduction of a spodic horizon in a podzol. The naturally developed Bh horizon of the Bedafse Bergen reduces the saturated hydraulic conductivity of the soil with 94.82% (factor 20). In the research of Mecke & Ilvesniemi (1999), a saturated conductivity reduction of the upper illuvial horizon of 95.93% (factor 25) was found in a podzol located in Hyytiala, southern Finland. Within the first SoSEAL pilot at Veersedijk, the Netherlands, the saturated conductivity has been reduced with 97.5% (factor 40) in the field (Zhou, 2017). Even though the injection and precipitation time is low (only a few hours) compared to the natural podsolization process, it seems that the SoSEAL method has proven to be even more effective in saturated permeability reduction. The high reduction rate of the SoSEAL method is likely to be caused by the high concentrations of organic matter and aluminium. Besides, no (ancient) root disturbance will be of influence within the first SoSEAL pilot layers. However, it has to be kept in mind that the permeability reduction has been measured directly after injection. There are no results of permeability reduction after months or years. This will be influenced by the stability of the complexes. The comparison with a podzol, which has developed over centuries, is therefore only giving an indication. In addition, properties of parent material, metal carbon

concentration and ratio have great influence on the formation of the heavy metal-OM complexes and the effectivity of the complex’ permeability reduction.

Further research

To develop multiple linear regression models based on chemical soil compound (TOC, Al, Fe)

concentration, bulk density and particle size as predictors and the Ksat and van Genuchten parameters (α, n, θres, θsat) as responses, more podzol profiles should be analyzed more extensively. Further research is suggested within this area. Moreover, further research is suggested considering agricultural purposes and developing an ‘artificial’ horizontal Bh horizon to improve water holding capacities of soils. Also, research on long-term permeability reduction of the SoSEAL method is suggested. In fact, studies on complex’ stability are already taking place within the SoSEAL project. Finally, the cementation of the Bh and Bs/C horizons of the podzols had not been classified. In further research, the horizon cementation could be classified according to the method proposed by Hirsch et al. (2017). The degree of cementation could then also be linked with hydraulic properties in a regression model.

(17)

17

Research improvements

Improvements of the research can mainly be found in the fieldwork preparation. First of all, the testing tubes were not ideally prepared on the first field day. The reasons for this were mainly the lack of time and connections. Furthermore, too few notes on the exact field location of soil properties –such as rooting abundance, horizon thickness- were taken in the field, as was assumed that previously found data would exactly describe the profiles. This is considered as a mistake, since some observations, such as rooting abundance and horizon thickness, differed from earlier observations.

Finally, the parameter estimation of the water retention curve was done manually. The reason for this was that a simulated fit was not giving appropriate results. This could be a result of non-homogeneous layers in the Bh horizon. However, the main differences between the Bh and C horizon are evident even though the estimation was done empirically.

Conclusion

The development of a spodic horizon, with Al/Fe-OM complexes, is changing the hydraulic properties of a soil layer. The cause of this change is expected to be a changed pore structure, due to the formed complexes. The pore structure of a spodic horizon, defined as irregular with small- and medium-sized pores, is likely to increase tortuosity and decrease porosity, which reduces hydraulic conductivity up to 94.82% and increases the suction -consequently the water holding capacity- of the soil. The Bh horizon is found to be able to hold 28.95% more for-plant-available-water than the parent material. This indicates that an ‘artificial’ horizontal spodic horizon could increase the water holding capacity of a soil, which is interesting for agricultural purposes. Further research on this topic is suggested. Finally, the SoSEAL method has proven to be more effective in permeability reduction than is realized within the

development of the Bh horizon of podzols, despite differences in parent material, Al/OM concentration and Al/OM ratio. However, the results were maintained just after the injection and still few is known about the stability of the formed complexes in the SoSEAL project so no long term conclusion can be drawn. Further research on the stability of the formed complexes in the SoSEAL project is suggested.

(18)

18

References

Ankeny, M. D., Ahmed, M., Kaspar, T. C., & Horton, R. (1991). Simple field method for determining unsaturated hydraulic conductivity.Soil Science Society of America Journal,55(2), 467-470.

Bear, J. (1972). Dynamics of fluids in porous media. Courier Corporation.

Bormann, H., & Klaassen, K. (2008). Seasonal and land use dependent variability of soil hydraulic and soil hydrological properties of two Northern German soils. Geoderma, 145(3), 295-302.

Bouma, J. (1983). Use of soil survey data to select measurement techniques for hydraulic conductivity. Agricultural Water Management, 6(2-3), 177-190.

Buurman, P., & Jongmans, A. G. (2005). Podzolisation and soil organic matter dynamics. Geoderma, 125(1), 71-83.

Darcy, H. (1856). Les fontaines publiques de la ville de Dijon. Dalmont, Paris.

De Coninck, F. (1980). Major mechanisms in formation of Spodic horizons. Geoderma, 24:101-128. Dirksen C. (1999). Soil Physics Measurements. Catena Verlag, Reiskirchen.

Edafologia (n.d.). Podzolization. Retrieved from http://www.edafologia.net/miclogia/podzolw.htm

FAO, 1979. Soil survey investigations for irrigation. Soils Bulletin 42. Rome. Fitts, C. R. (2002). Groundwater science. Academic press.

Fredlund, D. G., Rahardjo, H., & Fredlund, M. D. (2012). Unsaturated soil mechanics in engineering

practice. John Wiley & Sons.

G. Rout, S. Samantaray, P. Das. Aluminium toxicity in plants: a review. Agronomie, EDP Sciences, 2001, 21 (1), pp.3-21. .

Heimovaara, T. J., Kalbitz, K., Jansen, B. (2014). Soil Sealing by Enhanced Aluminium and DOM Leaching (SoSEAL). WATER2014, STW.

Heimovaara, T. J. (2017). Lecture 03: Unsaturated Flow (CIE4365 Period 4, 2017) [Powerpoint slides]. Section Geo-Engineering, Delft University of Technology.

Hirsch, F., Raab, T., Heller, S., Bauriegel, A. (2017). Classification of pedogenic cementation in Podzols by pocket penetrometry. Journal of Plant Nutritions and Soil Science. 2017, 000, 1–6

Hsin-yu Shan. Hydraulic Conductivity Tests for Soils [Powerpoint slides] (n.d.). Department of Civil IWG, W. (2014). World reference base for soil resources 2014. International soil classification system for

naming soils and creating legends for soil maps. World soil resources reports, (106).

(19)

19 Lajos, B., (2008). Soil science. Retrieved on 13-06-2017 from

http://www.tankonyvtar.hu/en/tartalom/tamop425/0032_talajtan/ch07s05.html

Mecke, M., & Ilvesniemi, H. (1999). Near-saturated hydraulic conductivity and water retention in coarse podzol profiles. Scandinavian journal of forest research, 14(5), 391-401.

Mehuys, G. R., & De Kimpe, C. R. (1976). Saturated hydraulic conductivity in pedogenetic characterization of podzols with fragipans in Quebec. Geoderma, 15(5), 371-380.

NRCS (n.d.). 618.88 Guide for Estimating Ksat from Soil Properties. Retrieved on 13-04-2017 from https://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/ref/?cid=nrcs142p2_054224

phpBB (2009). Soil porosity and water retention. Retrieved on 13-06-2017 from http://cactiguide.com/forum/viewtopic.php?t=13438

Lundström, U. S., van Breemen, N., & Bain, D. (2000). The podzolization process. A review. Geoderma, 94(2), 91-107.

van Genuchten, M. T., A closed-form equation for predicting the hydraulic conductivity of unsaturated soils, Soil Sci. Soc. Am. J., 44, 892-898, 1980

Wösten, J. H. M., Bannink, M. H., De Gruijter, J. J., & Bouma, J. (1986). A procedure to identify different groups of hydraulic-conductivity and moisture-retention curves for soil horizons. Journal of

hydrology, 86(1-2), 133-145.

Blume, H. P., Brümmer, G. W., Fleige, H., Horn, R., Kandeler, E., Kögel-Knabner, I., ... & Wilke, B. M. (2015). Scheffer/Schachtschabel soil science. Springer.

UMS (2012). Operation Manual KSAT. UMS GmbH, Gmunder Str. 37, 81379 Munich, Germany. UMS (2015). Operation Manual HYPROP. UMS GmbH, Gmunder Str. 37, 81379 Munich, Germany. Zhou, J., (2017). Pilot Veersedijk [Powerpoint slides].

(20)

20

Appendix A –Podzol locations

(21)

21

(22)

22

Appendix B – Photos of the fieldwork

Figure 14. The Bh and C horizon of the Bedafse Bergen podzol.

(23)

23

Figure 16. An example of a horizontal plane that has been created on a horizon in the field. While the picture was taken, the infiltration test on the Bh horizon of the Bedafse Bergen podzol took place.

(24)

24

Appendix C

Figu re 1 7. T he fit te d w ate r r et en tio n c ur ve w ith t he m ea sur ed a nd pr edi cted da ta po int s o f s am ple P1 B2 , t ha t is repr es en ting the B h ho riz on o f t he B eda fse B erg en po dz ol pr ofi le.

(25)

25 Fig ure 1 8. T he f itted w ater ret en tio n c ur ve w ith t he m ea sur ed a nd pr edi cted da ta p oin ts o f sa m ple P 1B 2, t ha t is re pre se nti ng th e C h ori zo n o f t he B ed afse B erg en p od zo l p ro file .

(Un)saturated soil water transport characteristics of the spodic horizon