Document [original]

Entwicklung und Anwendung der Selbst-Integrierenden

Akkumulatoren: Eine Methode zur Erfassung der

Sickerfrachten umweltrelevanter Stoffe

Development and Applications of the Self-Integrating

Accumulators: A Method to quantify the Leaching Losses

of Environmentally Relevant Substances

von Diplom-Geoökologe

Wolf-Anno Bischoff

aus Stuttgart

von der Fakultät VI – Planen Bauen Umwelt

der technischen Universität Berlin zur Erlangung des akademischen Grades

Doktor der Naturwissenschaften

- Dr. rer. nat. -

genehmigte Dissertation

Promotionsausschuss:

Vorsitzender: Prof. Dr. G. Franz

Gutachter: Prof. Dr. M. Kaupenjohann

Gutachter: Prof. Dr. B.-M. Wilke

Tag der wissenschaftlichen Aussprache: 20.11.2007

Berlin 2007

D83

Inhaltsverzeichnis / Content

1 Scope and Summary of Results................................................................1

1.1 Introduction...................................................................................................1

1.1.1 Need for methods to determine leaching losses...........................................1

1.1.2 Problems of common methods for the determination of leaching losses......2

1.2 Objectives.....................................................................................................7

1.3 Method development and validation.............................................................8

1.3.1 Preliminary soil column experiment..............................................................9

1.3.2 Validation with field experiments.................................................................10

1.3.3 Method extension for the use with organic chemicals.................................11

1.4 SIA method applications.............................................................................13

1.4.1 Nitrate leaching and implications for solute transport heterogeneity...........13

1.4.2 Risk assessment: Application of sewage sludge

containing tin organics on agricultural fields ...............................................14

1.5 Leaching risk estimate of pathogens from sewage sludge..........................15

2 Quantification of Nitrate Leaching in Soils

using an Adsorber Method ......................................................................22

2.1 Introduction.................................................................................................23

2.2 Materials and Methods................................................................................24

2.2.1 Self Integrating Accumulators (SIA)............................................................25

2.2.2 Experiment I: Chloride tracer mass balance under field conditions ............27

2.2.2.1 Soils and site description............................................................................28

2.2.2.2 Experimental...............................................................................................29

2.2.2.3 Calculations................................................................................................30

2.2.2.4 Statistics .....................................................................................................30

2.2.3 Experiments II a), II b) and II c): Sampling of preferential flow paths..........32

2.2.3.1 Experiment II a): Brilliant Blue tracer experiment for the

identification of preferential flow paths........................................................32

2.2.3.2 Experiment II b): Mass flux experiment with the sorbing solute tributyltin...32

2.2.3.3 Experiment II c): Mass fluxes of nitrate during a dry vegetation period.......34

2.2.4 Experiment III: Nitrate losses under different land uses..............................35

2.3 Results and Discussion...............................................................................36

2.3.1 Experiment I: Tracer Recovery...................................................................36

2.3.2 Experiments II a, b ,c: Qualitative evidence for sampling

of preferential flow ......................................................................................39

2.3.2.1 Experiment II a): Brilliant Blue tracer experiment for

the identification of preferential flow paths..................................................40

2.3.2.2 Experiment II b): Mass flux experiment with the sorbing solute tributyltin...42

2.3.2.3 Experiment II c): Mass fluxes of nitrate during a dry vegetation period.......43

2.3.3 Experiment III: Plausibility comparison of different land uses.....................45

2.4 Conclusions ...............................................................................................46

3 Stoffeintrag ins Grundwasser – Feldmethodenvergleich

unter Berücksichtigung von preferential flow........................................51

3.1 Einleitung....................................................................................................52

3.2 Stand der Forschung..................................................................................53

3.3 Messprinzip und potenzielle Fehlerquellen

der Selbst-Integrierenden Akkumulatoren (SIA) .........................................55

3.4 Validierung..................................................................................................57

3.4.1 Tracerversuch Beckum...............................................................................57

3.4.2 Tracerversuch auf verschiedenen Waldstandorten.....................................59

3.4.3 Tracerversuch Moor....................................................................................60

3.4.4 Farbtracerversuche.....................................................................................62

3.5 Anwendungsbeispiele.................................................................................63

3.5.1 Nährstoffbilanzierung landwirtschaftlicher Flächen in Nordvietnam............63

3.5.2 Grundwassergefährdung durch Nitrat im Mainbecken................................64

3.5.3 Weitere Anwendungsgebiete der Methode.................................................66

3.6 Zusammenfassende Diskussion.................................................................66

3.7 Schlussfolgerungen....................................................................................69

4 Applications of the SIA-Method: Variability of Mass Fluxes

under Field Conditions and Summary of Nitrate Losses

under different Land Uses .......................................................................73

4.1 Introduction.................................................................................................73

4.2 Materials and Methods................................................................................75

4.2.1 General.......................................................................................................75

4.2.2 Sites............................................................................................................76

4.2.3 Measurements............................................................................................77

4.2.4 Nitrate – general flux transfer function........................................................77

4.2.5 Test for (Log)Normal Distribution................................................................78

4.2.6 Statistical design.........................................................................................78

4.3 Results and Discussion...............................................................................80

4.3.1 Nitrate losses under different land uses......................................................80

4.3.2 Nitrate as a tracer for water fluxes..............................................................81

4.3.2.1 Water flux variability represented by nitrate data:

Mathematical derivation and discussion .....................................................81

4.3.2.2 Variability calculation results and discussion..............................................84

4.3.2.3 Influence of the sample scale .....................................................................88

4.4 Summarizing remarks and conclusions ......................................................88

5 Passivsammler als neue Methode zur in situ Bestimmung

von Pestizidversickerung.........................................................................92

5.1 Einleitung....................................................................................................93

5.2 Material und Methoden...............................................................................95

5.2.1 Messsystem................................................................................................95

5.2.2 Batch-Versuche..........................................................................................97

5.2.3 Abbauversuche...........................................................................................97

5.2.4 Versuche mit dem Passivsammler..............................................................97

5.2.5 Säulenversuche..........................................................................................98

5.2.6 Probenaufbereitung und Messung..............................................................99

5.3 Ergebnisse..................................................................................................99

5.3.1 Batch-Versuche..........................................................................................99

5.3.2 Abbauversuche.........................................................................................101

5.3.3 Versuche mit dem Passivsammler............................................................102

5.3.4 Säulenversuche........................................................................................103

5.4 Diskussion ................................................................................................104

6 Transport stark sorbierender Stoffgruppen

am Beispiel der Zinnorganika................................................................108

6.1 Einführung ................................................................................................108

6.2 Material und Methoden.............................................................................111

6.2.1 Standorte..................................................................................................111

6.2.2 Böden .......................................................................................................112

6.2.3 Stoffe ........................................................................................................113

6.2.4 Pflanzen....................................................................................................114

6.2.5 Anlage Feldversuch..................................................................................114

6.2.6 Probennahme Boden und Pflanze............................................................116

6.2.7 Probennahme Sickerwasser.....................................................................117

6.3 Ergebnisse................................................................................................119

6.3.1 Böden .......................................................................................................119

6.3.2 Pflanzen....................................................................................................120

6.3.3 Sickerwasser ............................................................................................124

6.4 Diskussion und Bewertung .......................................................................125

7 Gefährdungsabschätzung für den Transport von

pathogenen Keimen aus Klärschlamm in das Grundwasser.............132

7.1 Problematik und Ziel.................................................................................132

7.1.1 Mögliche Kontamination des Grundwassers durch Klärschlamm-

ausbrigung in der Landwirtschaft..............................................................132

7.1.2 Ziel............................................................................................................132

7.2 Stand der Kenntnis: Vorkommen von pathogenen Keimen

in Klärschlämmen.....................................................................................132

7.2.1 Einleitung..................................................................................................132

7.2.2 Bakterien...................................................................................................134

7.2.3 Viren .........................................................................................................136

7.2.4 Parasiten...................................................................................................139

7.2.5 Hefen und Pilze ........................................................................................140

7.2.6 Verlagerbarkeit von Keimen aus der Darmflora im Boden........................140

7.3 Material und Methoden.............................................................................143

7.3.1 Säulenversuche........................................................................................143

7.3.2 Feldversuche............................................................................................145

7.4 Ergebnisse................................................................................................146

7.4.1 Säulenversuch..........................................................................................146

7.4.2 Feldversuch..............................................................................................148

7.5 Diskussion ................................................................................................149

7.6 Bewertung.................................................................................................151

7.7 Ausblick ....................................................................................................153

8 Synthesis and Perspectives ..................................................................159

8.1 Method Development of the Self-Integrating accumulators (SIA).............159

8.2 Applications ..............................................................................................160

8.3 Perspectives.............................................................................................163

9 SUMMARY...............................................................................................165

10 ZUSAMMENFASSUNG............................................................................168

Abbildungsverzeichnis

List of Figures

Fig. 2 1: Installation and functional principle of the Self Integrating

Accumulator (SIA): The SIA is located below the undisturbed soil

and surrounded by material with the same hydraulic properties.

Solutes are extracted out of the water by (specific) sorption, while

water passes through. ..............................................................................26

Fig. 2 2: Results of an exact combinatorial analysis. Relation between SIA

method precision and number of replications. At least 10 replicates

are needed to achieve 20 % accuracy with a probability of > 90 %..........38

Fig. 2 3: Soil Profile (Bern) after a Brilliant Blue dye tracer experiment.

Black and dark grey show presence of dye. Squares show the

closed entrances of the side tunnels under the undisturbed soil in

15, 30 and 60 cm depth, where the SIA are installed. ..............................40

Fig. 2 4: SIA in 60 cm depth after a Brilliant Blue dye tracer experiment

(Muenster), partly excavated. Left: Black and white conceptual

picture. Right: Color Picture. Black and dark grey (left) or blue

(right) show presence of dye. Shaded area (up and right) is not yet

excavated. Black quarter circle is the casing of the SIA and

separates outside and inside....................................................................41

Fig. 2 5: Daily precipitation during a field experiment for the leaching

ofnitrate on nine farm fields with two fertilizer treatments. Marked

rainstorm events at the end of July...........................................................43

Fig. 2 6: N-Leaching (60 cm below surface) under dry conditions. Higher N-

losses in a small number of SIA may be explained by the sampling

of preferential flow. (Total N = 180) ..........................................................44

Fig. 2 7: Average N-Losses by leaching under different land uses. Data

from 7 sites (5 crop fields; 1 long term fallow; 1 forest) and 3 years

of continuous measurement (Summer: N=219; Winter: N=191)...............45

Abb. 3 1: Natrium- (links) und Chlorid- (rechts) Tiefenprofile der Vorräte über

den SIA ein Jahr nach Tracerapplikation (Versuchsstandort

Beckum). ..................................................................................................58

Abb. 3 2: Chlorid-Tiefenprofil der Vorräte über den SIA ein Jahr nach

Tracerapplikation (Mittelwerte und Standardabweichungen aus

vier Wiederholungen; Versuchsstandort Niedermoor Gütersloh)..............60

Abb. 3 3: Zusammenfassung der Chlorid-Tracerversuche von Kaupen-

johann in Beckum (unveröffentlicht), (Lang and Kaupenjohann

2004) auf vier Waldstandorten in Baden-Württemberg und

Siemens (unveröffentlicht) in einem Niedermoor bei Gütersloh................62

Abb. 3 4: Flächenausträge von Nitrat in den Versuchsgliedern (Mittelwerte

aus zehn Wiederholungen mit Standardabweichung; Versuchs-

standort Mainterrasse)..............................................................................65

Fig. 4 1: Situation of the regions with at least three independent fields..................76

Fig. 4 2: Nitrate-N losses [kg*ha-1] under different land uses and their

variability. Error bars = Std. Dev.. Total N = 3048.....................................80

Fig. 4 3: Variation of nitrate fluxes within and between the 7 German

regions......................................................................................................85

Fig. 4 4: Summary about the heterogeneity of water fluxes at different

scales. Heterogeneity expressed as coefficient of variation (CV

[%]). Error Bars = Standard Deviation ......................................................86

Abb. 5 1: Passivsammler (Skizze)

Fig. 5 1: Passive Sampler (sketch).........................................................................96

Abb. 5 2: Sorptionskapazität der getesteten Adsorbermaterialien gegenüber

PSM. Konzentrationen 3 und 5 µg l-1, je 4 Wdh.; Mittelwerte

berechnet relativ zur Kontrolle (%)

Fig. 5 2: Sorption results of different sorbers with respect to four pesticides.

Results from two con-centrations (3 und 5 µg l-1) with four

replicates; Means of recovery calculated relative to control (%).............100

Abb. 5 3: Vergleich der Wiederfindung von vier Pestiziden zwischen Florisil

und C18 nach 30 tägiger Inkubationszeit (in % zur Kontrolle);

Mittelwerte und Standardabweichungen aus 4 Wiederholungen

Fig. 5 3: Comparison of recoveries of four pesticides after 30 days of

incubation between C18 and Florisil (in % of control); Means and

standard deviations of 4 replicates .........................................................101

Abb. 5 4: Wiederfindung im Passivsammler nach Perkolation von 33 mm

Lösung mit 25 mg/l je PSM innerhalb einer halben Stunde

(schneller Fluß); Mittelwerte und Standardabweichungen aus 6

Wiederholungen

Fig. 5 4: Recoveries within the passive sampler after percolation of 33 mm

solution containing 25 mg/l per pesticide within 30 min (fast flow);

Means and standard deviations of 6 replicates.......................................102

Abb. 5 5: Wasserbilanz von Passivsammlern und Böden bei 7 Bodensäulen

äulen mit unregelmäßiger Perkolation (mal trockener, mal

feuchter) nach 2,5 Monaten

Fig. 5 5: Water balance of passive sampler outflow and soil outflow from 7

soil columns under a changing percolation regime (alternating

drier and wetter conditions) after 2,5 months..........................................103

Abb. 6 1: Anlage der Versuchsfelder .....................................................................115

Abb. 6 2: Schematische Zeichnung zum Einbau von SIA......................................117

Abb. 6 3: Ansicht des Standorts Forchheim zum Ende des Versuchs...................120

Abb. 6 4: Ansicht des Standorts Oberndorf zum Ende des Versuchs....................121

Abb. 6 5: Hakenwuchs der Karotten bei schwerem Boden (Standort

Oberndorf)..............................................................................................121

Abb. 6 6: Tomatenpflanze, die aus einem nicht abgetöteten Samen nach

Ausbringung des Klärschlamms gekeimt ist (Standort Forchheim).........124

Abb. 6 7: Schematische Vorstellung von bevorzugtem Transport auf

Standorten mit geringer Leitfähigkeit gegenüber solchen mit hoher

Leitfähigkeit (Konzentrationseffekt).........................................................129

Abb. 7 1: Ergebnisse der Sickerwasserproben (Angaben in %).............................146

Abb. 7 2: Vergleich der Fallzahlen positiv getesteter Keime im Sickerwasser

zwischen Oberböden (oben) und Unterböden (unten)............................147

Abb. 7 3: Horizontalschnitte in 15 cm (links) und 70 cm (rechts) nach

einmaliger Beregnung mit 30 mm Brilliant Blue/Kaliumbromid-

Tracer-Lösung ........................................................................................148

Abb. 7 4: Tiefenprofil eines konservativen Tracers im Feldversuch.......................149

Tabellenverzeichnis

List of Tables

Table 2 1: Selected soil properties of the tracer experiment sites.........................28

Table 2 2: Selected soil properties of the tributyltin experimental sites.................33

Table 4 1: Summary table of heterogeneity expressed as the

normalized coefficient of variation (CV [%]) and its

corresponding standard deviations (Std.Dev.) with respect to

scale (rows) and season (columns)......................................................84

Tabelle 6 1: Bodenkenndaten Parabraunerde; Standort Oberndorf.......................112

Tabelle 6 2: Bodenkenndaten Braunerde, Standort Forchheim .............................113

Tabelle 6 3: Anfangsgehalt von OT und MKW im Klärschlamm Nr. 19

(Angaben in mg/kg TS).......................................................................113

Tabelle 6 4: Zeitplan...............................................................................................116

Tabelle 6 5: Gehalte von MKW und OT in Karotten und Karottenkraut

vom Standort Oberndorf.....................................................................122

Tabelle 6 6: Massenspektroskopische Analyse des Karottenkrauts vom

Standort Oberndorf / Identifizierte Substanzen...................................123

Tabelle 6 7: Mittelwerte des Austrags von MKW und OT mit dem

Sickerwasser (80 cm Tiefe)................................................................125

Tabelle 6 8: Prozentualer Austrag von MKW und OT mit dem

Sickerwasser (80 cm Tiefe) bezogen auf die mit

Klärschlamm (KS) applizierte Menge .................................................126

Tabelle 7 1: Übersicht über Pathogene, die aus Fäkalien in den

Klärschlamm gelangen können (Umweltbundesamt 1995) ................133

Tabelle 7 2: Tenazitätsfaktoren im Boden (nach Gerba et al., 1975; Gerba

and Bitton, 1984; Sorber and Moore, 1987; Krannich, 1990) .............142

Tabelle 7 3: Standort und Kennwerte des Bodenprofils.........................................158

1. Scope and Summary of Results 1

1 Scope and Summary of Results

1.1 Introduction

1.1.1 Need for methods to determine leaching losses

The diffuse loss of nutrients and contaminants like nitrate, pesticides or tin organics

from agricultural fields is not only an economic problem for farmers, but is one major

cause for the non point contamination and eutrophication of groundwater (Lake,

Lovett et al. 2003), rivers, lakes, coastal regions and the ocean (Bouwman, Van

Drecht et al. 2005). The need to react to diffuse pollution is also recognized by the

European Commission in a strategy paper on the development of a soil protection

strategy (2002), the development of a soil framework directive proposal (2006) and

the integration of environmental issues in the common agricultural policy (CAP)

(2003).

In the same context, the EU Water Framework Directive (2000) recognises “the need

for action to avoid long-term deterioration of freshwater quality” and calls for the

determination of significant contaminations through anthropogenic diffuse sources.

These problems are currently addressed by economic and ecological projects on a

larger scale (Rinaudo, 2003; Thresholds (Integrated Project), started 2005). In

Germany, for example, close to 50 % of the nitrate input into surface water is by

influx of groundwater previously loaded under agricultural fields (Umweltbundesamt

2002). Nitrate is also the greatest problem for the German drinking water quality

(Umweltbundesamt 2002) and of major concern in the USA (Nolan, Ruddy et al.

1998).

The effectiveness of management practices, which are introduced to reduce the

groundwater pollution by diffuse pollutants, must be controlled. It is therefore

1. Scope and Summary of Results

desirable to monitor diffuse leaching losses on a field scale under practical farming

and natural boundary conditions (Bischoff, Siemens et al. 1999). Unfortunately, the

soil – groundwater path is not easy to monitor. Therefore, quantitative information on

leaching losses of nutrients and diffuse pollutants to groundwater up to now is only

available at great costs and with several limitations to the applied methods (Bischoff,

Siemens et al. 1999; Bischoff and Kaupenjohann 2007).

Efforts are currently made to set standards for the measurement of solutes in the soil

water and to introduce a good practice guide for different methods (Pütz, Deurer et

al. 2007; DIN_V_19715 in prep.).

1.1.2 Problems of common methods for the determination of leaching losses

Field methods to monitor leaching are tile drains (Elliott, Cessna et al. 1998;

Kladivko, Grochulska et al. 1999; Pampolino, Urushiyama et al. 2000; Kladivko,

Frankenberger et al. 2004), lysimeters (Bergström 1990; Jemison and Fox 1992;

Jene 1998; Pampolino, Urushiyama et al. 2000), suction plates (Dressel 2002;

Siemens 2003; Kosugi and Katsuyama 2004), suction cups (Kung 1990; Grossmann

and Udluft 1991; Selker, LeClerq et al. 1992; Elliott, Cessna et al. 1998; Pampolino,

Urushiyama et al. 2000), wick samplers (Brandi-Dohrn, Dick et al. 1996; Brandi-

Dohrn, Dick et al. 1996) and soil coring (Pampolino, Urushiyama et al. 2000; Wang,

Magesan et al. 2004). Many subtypes exist and their limitations are also discussed in

a review (Litaor 1988).

The work and cost problem is mainly related to the repeated sampling and analysis

necessities. For lysimeters high investment and maintenance costs add to the limited

application, which is mainly on experimental stations. The scientific limitations are

mostly due to the heterogeneity of solute transport in the field and the transfer of soil

1. Scope and Summary of Results 3

or water concentrations into mass fluxes. This is also related to the problems of

measuring local water fluxes in an open system.

Tile drain field experiments are a good reference for average leaching losses (Elliott,

Cessna et al. 1998) but are limited to soils with water logging, where tiles often exist,

or experimental fields, where drains can be installed for scientific purposes.

Moreover, drains do not collect the total drainflow of a field. Their collection efficiency

depends on soil properties and drain spacing (Kladivko, Grochulska et al. 1999;

Kladivko, Frankenberger et al. 2004). Therefore, the total solute loss per area is

unknown, unless drainage concentrations are coupled to local water balance models

or groundwater recharge data with the same time resolution.

Monolyth lysimeters are an excellent mean to get full mass balances, but are very

expensive and therefore limited to exemplary studies. Also, undisturbed large soil

monoliths should be preferred over refilled soil columns, but are not easily sampled

(Bergström 1990).Pan lysimeters are more readily installed in the field. Zero tension

pan lysimeters collect only free drainage water, however. They disturb the native flow

regime and are inherently unsuitable for the measurement of mass fluxes under

natural conditions. The collection efficiency has been stated to be in the range of 45

% - 58 % (Jemison and Fox 1992). A large plate with suction cups in a grid pattern,

which sampled defined areas was installed by (Jene 1998) and yielded recoveries of

48 % of a bromide tracer compared to 77 % in a monolith lysimeter.

In two 2-day irrigation experiments (Siemens 2003) found that small tension

controlled suction plates yielded good chloride tracer recoveries (mean: 118 %;

standard deviation: 70 %) in a sandy soil at two different tensions and irrigation

regimes. In contrast, wick samplers gave satisfactory results only at flow conditions

close to saturation, because they apply suction only up to -50 – -60 hPa. Dressel

1. Scope and Summary of Results

(2002) got an overestimation of 30 – 40 % by using tension controlled suction plates

in a bromide mass balance tracer experiment.

The most common field method to date is the use of suction cups in combination with

some kind of water balance, e.g. modelled water fluxes from tensiometer and TDR

measurements. Suction cups sample the soil solution by applying a continuous or

intermittent, fixed or variable suction to the soil. They may not be able to capture

water from preferential flow pathways, which contains relevant concentrations e.g. of

pollutants for groundwater recharge (Kung 1990; Selker, LeClerq et al. 1992).

Problems to sample representatively may arise from non-continuous sampling, a

small cross sectional area of the cup and sampling from an ill defined soil volume and

other factors (Litaor 1988; Li, Skogley et al. 1993).

In a comparison study, the suitability of suction cups, pan lysimeters, tile drains and

resin capsules to describe nitrate movement was tested (Pampolino, Urushiyama et

al. 2000). Suction cups showed higher nitrate concentrations in the top soil and too

low concentrations in the subsoil compared to the pan lysimeters and the tile drains.

This is in agreement with the theoretical consideration that suction cups do not

sample the fast flow, which in the top soil will have lower nitrate concentrations than

the matrix water. In contrast, in the subsoil tile drains sample higher nitrate

concentrations from the preferential pathways. For the subsoil it could be concluded

that on this site preferential flow is also an important transport mechanism for nitrate.

Grossmann and Udluft (1991) criticise that the sampled water inside the suction cups

poorly represents the actual soil solution. Elliott, Cessna et al. (1998) use the

concentration difference of the herbicide clopyralid in tile drains and suction cups as

a measure for preferential flow. They also state that suction cups are not able to

sample fast flows and that fast flow accounted for 40 % of the water transport. Also,

1. Scope and Summary of Results 5

suction cups may suffer from artefacts due to temperature effects and bypass flow

along the tube (Buchter, Hinz et al. 1999).

Changes of the concentrations of solutes in soil core samples over time are often

used to follow the dissipation of chemicals. The main problem is that many processes

may contribute to an observed decrease in concentration. In the case of pesticides,

leaching, microbial degradation, volatilisation and immobilisation may induce an

observed loss of pesticides between two sampling dates. For nitrate, leaching,

chemical changes to several other N-compounds, microbial and plant uptake have to

be considered. Soil cores therefore have to be used in combination with other

methods to yield information on leaching losses. Pampolino, Urushiyama et al. (2000)

advise not to use destructive sampling for the determination of the transport of mobile

solutes such as nitrate, because the amount at only one point of time is known and

repeated sampling in the same spot is impossible. Another problem is due to the

space distribution of preferential flow patterns in unstructured soils (Wang, Wu et al.

2003). The ‘wavelengths’ of fingering due to fluid instability may be large compared to

the sampling area of the soil cores.

Skogley (1992) developed a method, in which ion exchange resin is packed tightly

into plastic nets with small grid spacing. This allows for diffusive movement of

exchangeable ions to the resin, which acts as a total sink. This method also captures

a part of the convection. The two processes cannot be separated, so that the amount

of leached N or plant available diffusive N cannot be calculated.

Yang and Skogley (1992) reported that the movement of nutrients to a resin capsule

was diffusion-controlled, sensitive to solution concentration and soil dependent.

A comparison of resin capsules with suction cups in a soil column Bromide tracer

experiment showed that suction cups could not react to fast spontaneous movement

1. Scope and Summary of Results

of Bromide under intermittent unsaturated flow conditions, whereas resin capsules

showed these mass flow peaks, when exchanged daily (Li, Skogley et al. 1993).

In a comparison study (Pampolino, Urushiyama et al. 2000) resin capsules showed a

fairly good correlation with the tile drains and could be used as a relative, but not

absolute measure for nitrate losses.

To sum up, preferential flow and spatial variability influence solute flux, which causes

further methodic bias for all the above methods (Beven and German 1982; Jury and

Flühler 1992; Williams, Dowd et al. 2003). Many authors show that major quantities

of soil water passes by the soil matrix and drains fast to the subsoil through

macropores (Bouma and Dekker 1978; Brusseau and Rao 1990; Dekker and

Ritsema 1995). Although these ideas are widely accepted they are not well

considered in the setup of field experiments. Many field methods can not account for

either of the phenomena.

Fast water transport through macropores (> 0.1 – 0.2 mm) (Scotter 1978) does not

only occur in structured soils, but also in macroscopically homogeneous soils.

Several authors (Hill and Parlange 1972; Selker, LeClerq et al. 1992; Selker,

Parlange et al. 1992; Hart and Lowery 1996) described typical infiltration patterns in

homogeneous sandy soils called ‚fingering’. Instability of the wetting front was

attributed to differences in soil density, hydrophobicity of surfaces and the water

conductivity of the soil (Stagnitti, Parlange et al. 1995).

In contrast, Wang, Wu et al. (2003) attributed fingering in unstructured soils to fluid

instability at the wetting front by redistribution after the end of rainfall or irrigation.

A review (Knoblauch 1996) concluded that shallow soil depth and high preciptitation

are the main factors increasing the contribution of preferential flow to transport of

chemicals. But also groundwater in great depth can be contaminated by only a few

1. Scope and Summary of Results 7

preferential flow paths (Gießl 1998). Preferential flow paths do not only present a

fundamental problem to transport modelling but also for the monitoring and detection

of chemical transport in the field (Gölz-Huwe, Simon et al. 1989).

Overall, it can be concluded that there is a limited range of methods to quantify mass

transport under field conditions all bearing their limitations, which are available until

now. None of them can measure quantitatively mass fluxes per area directly. Such a

method would be desirable for many applied studies including risk assessment,

success monitoring after introduction of new agricultural management techniques

and others.

Therefore, this work is concerned with developing and using a method to measure

the absolute leaching losses on a [mass area-1 time-1] basis, which closes the mass

balance for the soil - groundwater path and closes a gap of field methods in soil

science.

1.2 Objectives

The first objective of this work was to develop and test an accumulative self-

integrating method to detect and quantify solute transport and leaching in the soil for

a wide range of solutes under field conditions. The possible advantage would be a

low effort method, which does not interfere with the surface activities. It could be

used with high replication numbers to allow for intensive statistical evaluation.

The second objective was to apply the method in field studies with varying target

chemicals and scientific objectives:

1. Quantification of nitrate leaching losses under different land uses

2. Quantification of scale factors for soil water flux heterogeneity under field

conditions from the profile to the region

1. Scope and Summary of Results

3. Importance of preferential flow for sorbing and non-sorbing solutes

4. Risk assessment for the application of sewage sludge containing tin organic

compounds on agricultural fields

The third objective was to extend the assessment of leaching risks of sewage sludge

to pathogens in soils. So far, no satisfactory sorbent for the integrative collection of

pathogens in the soil has been found. Therefore, a worst case combined method of

field dye tracer and soil column experiments was used instead.

1.3 Method development and validation

I wanted to develop and test a method with the following characteristics:

The measurement devices should not need regular sampling during the installation.

Therefore, the target substances have to be collected on appropriate adsorber

materials and the water must not be sampled or stored but rather stripped off the

target substance during the passage through the measurement devices. The

sampled amount of target substance should be representative for its mass flux at the

studied soil depth. This requires a more or less undisturbed water flow at the soil –

device interface. It also requires undisturbed soil conditions above the area of

measurement, because altered soil materials would lead to altered secondary

structures with consequent changes in hydraulic properties.

The validation of the so called Self-Integrating Accumulator (SIA) method had to

concentrate on experiments, which are close to natural conditions.

First, the SIA method has to be representative for the transport through a certain soil

volume or cross sectional area at a certain soil depth. Therefore, the water balance

should be similar within the SIA and in the soil. It is also important to know, whether

preferential flow paths are sampled as well. So the main problem is a practical

1. Scope and Summary of Results 9

solution to the highly various and water content dependent water conductivity

function of different soils.

Second, the target substances should accumulate fully and be protected from

chemical or biological alteration within the accumulators. For cations and anions

appropriate resins already existed. Therefore, only the breakthrough behaviour of

target substances like pesticides and tin organics had to be tested for promising

adsorber materials to be used within the SIA. Also, I tested the degradation of the

sorbed target substances.

The SIA method was then applied to several field studies to assess groundwater

leaching risks of different target substances.

Its wide application for nitrates can be used to gather information on spatial variability

of solute transport depending on soil, management and regional differences.

1.3.1 Preliminary soil column experiment

In a preliminary experiment I tested a first guess for the installation of Self Integrating

Accumulators (SIA) in different soils and its impact on the water flux balance.

Several methods and techniques have been applied so far to quantify diffuse

leaching losses under agricultural fields each of them bearing its limitations. The SIA

method is proposed to overcome some of these limitations.

In a preliminary experiment (Bischoff and Kaupenjohann 1998) I tested the water flux

amounts inside and outside the SIA. For this, I used seven undisturbed soil columns

of 30 cm length and 25 cm diameter with different soil types. The SIA (10 cm

diameter) were installed in the centre. The installation method used in these

experiments was to press the SIA from below into the undisturbed soil. Different

irrigation intensities were applied over a three month period. Soil water drainage and

1. Scope and Summary of Results

SIA drainage were collected separately. A comparison showed that with this type of

installation SIA drainage overestimated soil drainage 2 - 10 times. I concluded that I.

the type of installation had to be changed and II. the boundary effects of the soil

column experiment (Schwartz, McInnes et al. 1999) may influence the scientific value

of the results for a validation experiment and decided to validate under field

conditions.

1.3.2 Validation with field experiments

In a second approach the suitability to measure solute transport correctly and the

ability to collect preferential flow paths had to be shown under field conditions

(Bischoff, Siemens et al. 1999; Bischoff and Kaupenjohann 2007).

In a next step (Bischoff and Kaupenjohann 2007) I changed the installation

procedure. Instead of pressing the SIA into the undisturbed soil, I surrounded them

from all sides with sand and silt quartz of the same hydraulic conductivity during the

installation. Thus, I minimized the pressure on the above undisturbed soil and

drained the excess water along the outsides of the SIA, leading only the inner

cylinder of the flow field through the SIA. To test the suitability of this design I did

three kinds of experiments:

I. With a chloride mass balance experiment on six agricultural fields with five

replicates I determined the correctness of the method and the variation between

replicates.

II. To examine the capability to intercept preferential flow paths I used three

qualitative approaches: IIa. A dye tracer experiment to mark the flow paths in the soil

and around the SIA. IIb. An experiment with a strongly sorbing solute (tributyltin),

which presumably is transported only preferentially. IIc. Nitrate leaching during a very

dry summer, where matric flow is very unlikely.

1. Scope and Summary of Results 11

III. In a plausibility experiment I measured the mean nitrate losses under different

land uses: forest, long term fallow and agriculture. These are compared with

published results.

The results from I. (mass balance) show that with 30 replicates 92 % of the validation

value from the mass balance were recovered. The variation is such that ten

replicates would estimate the true mean within an error of 20 %. From II. it could

qualitatively be shown that SIA intercept and sample from preferential flow paths. In

the dye tracer experiment (IIa.) it could be seen that dye colouring preferential flow

paths continued its trace into the SIA. A strongly sorbing solute (IIb.) was found in 90

cm depth in the SIA, where it could most probably be transported only by preferential

flow. Also nitrate (IIc.) was detected in a small percentage of SIA during a period,

where leaching could only occur due to preferential flow after a few rainstorm events

in a generally very dry period. From III. (plausibility), mean values for N-losses were

7, 5, 21 [kg ha-1] for forest, long term fallow and agriculture respectively.

I concluded that the Self Integrating Accumulator (SIA) method is valid to measure

leaching losses of nitrate and chloride on a mass per area basis in our experiments

and is a promising method also for preferentially transported solutes. These results

are supported by several smaller tracer and field experiments in an earlier stage

(Bischoff, Siemens et al. 1999).

1.3.3 Method extension for the use with organic chemicals

In a third approach the method was extended to organic pollutants. Several

pesticides with a wide range of Octanol-Water distribution coefficients (Kow) were

tested.

Because the hydraulic problem could be solved for ions and the water fluxes are

represented adequately, then the method could be extended to organic chemicals as

1. Scope and Summary of Results

well, if effective adsorbers exist. Our objective was to identify adsorbers for organics

that sorb effectively and protect the target substances from degradation under flux

conditions.

In a screening (Bischoff and Kaupenjohann 1998) seven adsorbers were tested for

the pesticides isoproturon, terbutylazin, metolachlor and pendimethalin, which

represent a large range of hydrophobicity from Kow ~ 150 to > 10 000. The experi-

ment had three steps: I. Recovery from simple batch experiments at two

concentrations. II. With the successful adsorbers: Recovery after 30 days incubation

time in a suspension inocculated with fresh soil material and microorganisms at

294 K. III. With the most promising adsorber: Breakthrough experiment with fast flow

simulation (30 mm in 30 min.) on a filled SIA column with the adsorber – sand / silt

mixture ratio of the nitrate SIA.

The octadecyl phase and the Mg, Al-silicate showed the highest recoveries (50 –

100 %) for the four pesticides. Both performed well in the incubation experiment with

recoveries of 60 – 70 % for the Mg, Al silicate and 70 – 90 % for the octadecyl phase,

but for the octadecyl phase I had a marked loss in adsorber material, which I

attributed to biodegradation of the adsorber. In the breakthrough experiment 70 –

80 % of the pesticides were recovered from the SIA and the breakthrough water did

not contain detectable quantities of any pesticide.

I concluded that the Mg, Al silicate is an appropriate adsorber for a wide range of

organic substances from different chemical groups.

1. Scope and Summary of Results 13

1.4 SIA method applications

1.4.1 Nitrate leaching and implications for solute transport heterogeneity

In a fourth approach the method was applied to several field studies as a monitoring

device for nitrate fluxes from agricultural fields. Conclusions can be drawn from the

results on field heterogeneity and spatial variation, which may be valuable for our

understanding of natural variability (Bischoff and Schwarz 2006).

The spatial and short term variability of Nmin-results is well known (Schweigert and

Zimmermann 2003), but the Nmin method only gives the residual concentration at one

point in time and space. Our objective was to identify land use, scale and seasonal

differences in mass fluxes integrated over time.

The downward nitrate mass flux at a depth of 60 cm (horticulture) or 90 cm

(agriculture) was measured on 50 fields with ten replicates per field for several

seasons (ca.: vegetation period April – September, drainage period October –

March). Overall, more than 1000 replications were used for the data analysis. The

variability was expressed either as mean and standard deviation (SD) or as the

normalized coefficient of variation (CV) to facilitate inter-field and inter-seasonal

comparisons of the variation.

At the field scale and within one season mass fluxes were highly variable from point

to point and varied with two to three orders of magnitude. Losses during the

vegetation period (mean: 13 kg N*ha-1) were about 25 % of the losses during the

main drainage period (mean: 51 kg N*ha-1).

Under forest, approx. 10 kg N*ha-1 were lost, under agriculture 35 kg N*ha-1 and

under horticulture 140 kg N*ha-1.

1. Scope and Summary of Results

The variability increased from a mean CV of 100 % at the 10 cm-scale to > 200 %

between regions with a SD for the CV between 30 – 100 %, also increasing with

scale.

I conclude that losses during the vegetation period could not be neglected in an

overall mass balance. Different land uses were reflected by different means of nitrate

losses. Even time integrated mass fluxes are highly variable at short distance.

Variability increases with distance. This should affect the design of all field

experiments concerned with leaching losses, where many measurement replications

are required.

1.4.2 Risk assessment: Application of sewage sludge containing tin organics

on agricultural fields

In a fifth approach our experience with pesticides was further extended to tin organics

(mono-, di-, tributyltin) to be used in a field study on the environmental fate of these

substances after the application of sewage sludge to agricultural fields (Hartmann,

Bischoff et al. 2004). Little information is available on the environmental behaviour of

this substance class. From the chemical structure it can be deduced that hydrophobic

and ionic bonding are possible and the tin organics are highly sorptive.

I conducted preliminary incubation and breakthrough experiments similar to those for

the pesticides with the SIA and could not detect any tin organics in the leachate.

Therefore, the adsorbers were appropriate. After that SIA were installed from four pits

with three replicates (= twelve per plot) on two sites with two plots (with and without

sewage sludge application). Site 1 is a sandy Eutric Cambisol (FAO) in the Rhine

valley near Karlsruhe. Site 2 is a loamy Luvisol derived from loess near Tübingen.

The fields were ploughed and then the sewage sludge was applied to two plots at

5 t*ha-1 dry mass. All fields were then grubbed and harrowed. Afterwards carrots

1. Scope and Summary of Results 15

were sawn. 3 months later the carrots were harvested, depth profiles of the soil were

sampled and the SIA recovered from the soil. From these samples, plant transfer, soil

residues and fluxes of tin organics could be determined.

Tributyltin was not detectable in soil and plant samples, but about 1 % of the applied

mass was recovered in the SIA as the integrated flux amount at both treated plot

sites. I assume that the tributyltin concentration was diluted below the detection limit

by grubbing and harrowing in the soil samples, whereas the accumulated flux had a

lower detection limit and was in the order of magnitude known for common pesticides

(Flury 1996).

It could be concluded that organic contaminants from sewage sludge could be

measured on their leaching passage and that sewage sludge application on

agricultural fields may pose a risk to groundwater even for strongly sorbing solutes.

1.5 Leaching risk estimate of pathogens from sewage sludge.

Sewage sludge often contains pathogens, which have to be considered for a

comprehensive risk assessment. However, adequate adsorbers to measure the

leaching of microorganisms are currently not available. Therefore, I considered

alternative possibilities to assess the risk of microorganism leaching to groundwater

experimentally.

The discussion about the use of sewage sludge on agricultural fields is controversial

(BMVEL and BMU 2002; VDLUFA and ATV-DVWK 2003). Sewage sludge does not

only effect the soil – groundwater path by chemical contaminants, but also by

pathogens.

1. Scope and Summary of Results

The objective was to assess the risk of pathogen leaching from sewage sludge to

groundwater in a catchment with 0.6 – 1 m loess deposits overlying karstic

calcareous rock (Muschelkalk).

The working assumptions were

1. If the pathogens reach the calcareous rock, there will be no more effective

barrier against them reaching the groundwater.

2. The microorganisms are ‘big’ and tend to sorb strongly to the soil solid phase

compared to the chemical transport considered before.

From 1. I can deduce that I only had to consider the transport in the loess layer. From

2. I can deduce that transport through macropores or preferential flow is the most

probable mechanism.

Therefore, I split the problem in two experiments. In a dye tracer experiment in the

field I tested whether continuous macroporous transport system exists, which

facilitates the bypass flow of the pathogens. In this experiment, 30 mm of a brilliant

blue dye tracer solution was applied to the field once. Then, the horizontal and

vertical cross sections down to 1 m were analysed for the tracer with an optical

software from high resolution photographs.

In an undisturbed soil column leaching test with four indicator groups (Escherichia

coli, Coliformae, Enterococcus faecalis spec., Bacillus subtilis) I tested whether the

indicator organisms survive the passage through the soil. For this, sewage sludge

enriched with these organisms was applied to three top soil (0 - 30 cm) and three

subsoil (30 – 60 cm) columns with a diameter of 20 cm and irrigated for six weeks

with varying intensity. The leachate was collected and given to specific nutrition

media to detect survival of the indicator organisms.

1. Scope and Summary of Results 17

The tracer test showed that even after one application tracer could be found in 1 m

depth. The column experiments showed that in 44 % of the samples at least one

indicator organism survived.

It could be concluded that in this case sewage sludge posed a pathogenic risk to the

groundwater, because the microorganisms survive the transport and an effective

bypass flow structure is existent in the soil.

After this experience, the combination of a soil structural field method and a soil

biological laboratory experiment seems a valuable tool for the risk assessment of

pathogen leaching.

1. Scope and Summary of Results

Literature

(2000). Richtlinie 2000/60/EG zur Schaffung eines Ordnungsrahmens für

Maßnahmen der Gemeinschaft im Bereich der Wasserpolitik. 2000/60/EC: 72.

(2006). Directive establishing a framework for the protection of soil and amending

Directive 2004/35/EC (Proposal). 2006/0086(COD): 30.

Bergström, L. (1990). "Use of lysimeters to estimate leaching of pesticides in

agricultural soils." Environmental Pollution 67: 325-347.

Beven, K. and P. German (1982). "Macropores and water flow in soils." Water

Resources Research 18: 1311-1325.

Bischoff, W.-A. and M. Kaupenjohann (1998). "Passivsammler als neue Methode zur

in situ Bestimmung von Pestizidversickerung." Zeitschrift für Pflanzenkrankheiten und

Pflanzenschutz Sonderheft XVI: 779-786.

Bischoff, W.-A. and M. Kaupenjohann (2007). "Quantification of Diffuse Leaching

Losses of Nitrate under Agricultural Fields using the Self Integrating Accumulator

(SIA) Method." J. Environ. Qual. submitted.

Bischoff, W.-A. and A. Schwarz (2006). "Zeitliche und räumliche Heterogenität von

Messungen des Stoffflusses am Beispiel Nitrat und ihre Bedeutung für den

vorsorgenden Grundwasserschutz." Forum der Geoökologie 3/06: 29-33.

Bischoff, W.-A., J. Siemens, et al. (1999). "Solute leaching into groundwater - A

comparison of field methods considering preferential flow." Wasser & Boden 51(12):

37-42.

BMVEL and BMU (2002). Gute Qualität und sichere Erträge - Wie sichern wir die

langfristige Nutzbarkeit unserer landwirtschaftlichen Böden? - Vorschlag zur

Begrenzung des Eintrags von Schadstoffen bei der Düngung landwirtschaftlicher

Nutzflächen. Berlin, Bundesministerium für Verbraucherschutz, Ernährung und

Landwirtschaft.

Bouma, J. and L. W. Dekker (1978). "A case study on infiltration into dry clay soil. I.

Morphological Observations." Geoderma 46: 3-11.

Bouwman, A. F., G. Van Drecht, et al. (2005). "Exploring changes in river nitrogen

export to the world's oceans." Global Biogeochemical Cycles 19: 1-14.

Brandi-Dohrn, F. M., R. P. Dick, et al. (1996). "Field evaluation of passive capillary

samplers." Soil Sci. Soc. Am. J. 60: 1705-1713.

Brandi-Dohrn, F. M., R. P. Dick, et al. (1996). "Suction cup sampler bias in leaching

characterization of an undisturbed field soil." Water Resources Research 32, No. 5:

1173-1182.

Brusseau, M. L. and P. S. C. Rao (1990). "Modeling solute transport in structured

soils: A review." Geoderma 46: 169-192.

1. Scope and Summary of Results 19

Buchter, B., C. Hinz, et al. (1999). "Evaluation of temperature and bypass flow

sensitivity of tensiometers in a field soil." Geoderma 87: 281-291.

Dekker, L. W. and C. J. Ritsema (1995). "Fingerlike wetting patternsin two water

repellent loam soils." J. Environ. Qual. 24(2): 324-333.

DIN_V_19715 (in prep.). "Probenahme von Bodenwasser zur Bestimmung der

Inhaltsstoffe und Abschätzung von Sickerfrachten." DIN-Norm.

Dressel, J. (2002). Transport von Ethidimuron und Methabenzthiazuron in einer

Parabraunerde - Experimente und Modellierung. Institute for Chemistry and

Dynamics of the Geosphere. Jülich, University of Bonn.

Elliott, J. A., A. J. Cessna, et al. (1998). "Soil processes and chemical transport:

Leaching and preferential flow of clopyralid under irrigation: Field observations and

simulation modeling." J. Environ. Qual. 27: 124-131.

European_Commission (2002). Towards a thematic strategy for soil protection.

Brüssel, European Commission: 35.

European_Commission, D.-G. f. A. (2003). Agriculture and the environment,

European Commission, Directorate-General for Agriculture.

Flury, M. (1996). "Experimental Evidence ofTransport of Pesticides through Field

Soils - A review." J. Environ. Qual. 25: 25-45.

Gießl, H. (1998). Über das Vorkommen ausgewählter Pflanzenschutzmittel in Wasser

unter besonderer Berücksichtigung des Grundwassers. Institut für Phytomedizin.

Stuttgart, Hohenheim: 153.

Gölz-Huwe, H., W. Simon, et al. (1989). "Zum jahreszeitlichen Nitratgehalt und zur

Nitratauswaschung von landwirtschaftlich genutzten Böden in Baden-Württemberg."

Zeitschrift für Pflanzenernährung und Bodenkunde 152: 273-280.

Grossmann, J. and P. Udluft (1991). "The extraction of soil water by the suction cup

method: A review." J. Soil Sci. 42: 83-93.

Hart, G. L. and B. Lowery (1996). "Partitioned flowdomainsof three Wisconsin soils."

Soil Sci. Soc. Am. J. 60: 1686-1696.

Hartmann, E., W.-A. Bischoff, et al. (2004). Untersuchung von Klärschlamm auf

ausgewählte Schadstoffe und ihr Verhalten bei der landwirtschaftlichen

Klärschlammverwertung. Berlin, Umweltbundesamt (Environmental Protection

Agency of Germany): 225.

Hill, D. E. and J.-Y. Parlange (1972). "Wetting front instability in layered soils." Soil

Sci. Soc. Am. J. 36: 697-702.

Jemison, J. M. and R. H. Fox (1992). "Estimation of zero tension pan lysimeter

collection efficiency." Soil Science 154: 85-94.

1. Scope and Summary of Results

Jene, B. (1998). Transport of Bromide and Benazolin in Lysimeters and a field plot

with grid suction bases in a sandy soil. Institute of Soil Science. Stuttgart,

Hohenheim.

Jury, W. A. and H. Flühler (1992). "Trabsport of chemicals through the soil:

Mechanisms, models and field applications." Advances in Agronomy 47: 141-202.

Kladivko, E. J., J. R. Frankenberger, et al. (2004). "Nitrate leaching to subsurface

drains as affected by drain spacing and changes in crop production system." Journal

of Environmental Quality 33(5): 1803-1813.

Kladivko, E. J., J. Grochulska, et al. (1999). "Pesticide and nitrate transport into

subsurface tile drains of different spacings." J. Environ. Qual. 28: 997-1004.

Knoblauch, S. (1996). "Wasser- und Stofftransport über präferentielle Fließbahnen in

Böden - eine Literaturübersicht." Wasserwirtschaft 86: 598-602.

Kosugi, K. and M. Katsuyama (2004). "Controlled-suction period lysimeter for

measuring vertical water flux and convective chemical fluxes." Soil Science Society

of America Journal 68(2): 371-382.

Kung, K.-J. S. (1990). "Preferential Flow in a sandy vadose zone: 1.Field

Observation." Geoderma 46: 51-58.

Lake, I. R., A. A. Lovett, et al. (2003). "Evaluating factors influencing groundwater

vulnerability to nitrate pollution: developing the potential of GIS." Journal of

Environmental Management 68(3): 315-328.

Li, Z. M., E. O. Skogley, et al. (1993). "Resin adsorption for describing bromide

transport in soil under continuous or intermittent unsaturated water flow." J. Environ.

Qual. 22: 715-722.

Litaor, M. I. (1988). "Review of soil solution samplers." Water Resources Research

24, NO. 5: 727-733.

Nolan, B. T., B. C. Ruddy, et al. (1998). "A national look at nitrate contamination of

groundwater." Issue of Water Conditioning and purification 39(12): 76-79.

Pampolino, M. F., T. Urushiyama, et al. (2000). "Detection of nitrate leaching through

bypass flow using pan lysimeter, suction cup, and resin capsule." Soil Science and

Plant Nutrition 46 (3): 703-711.

Pütz, T., M. Deurer, et al. (2007). In situ - Erfassung von Bodenlösung, DWA: 38.

Schwartz, R. C., K. J. McInnes, et al. (1999). "Boundary effects on solute transport in

finite soil columns." Water Resources Research 35(3): 671-681.

Schweigert, P. and P. Zimmermann (2003). "The N-min content of the soil as an

agricultural environmental indicator of water pollution with nitrate." Berichte Uber

Landwirtschaft 81(2): 192-207.

1. Scope and Summary of Results 21

Scotter, D. R. (1978). "Preferential solute movement through larger soil voids I. Some

computations using simple theory." Australian Journal of Soil Research 16: 257-267.

Selker, J., P. LeClerq, et al. (1992). "Fingered Flow in two Dimensions: 1.

Measurement of matric potential." Water Resources Research 28: 2513-2521.

Selker, J., J.-Y. Parlange, et al. (1992). "Fingered flow in two dimensions II. Predicted

finger moisture profile." Water Resources Research 28: 2523-2528.

Siemens, J. (2003). Controls of carbon, nitrogen, and phosphorus fluxes in vadose

zone and groudwater of protected watersheds in Münster (Germany). Fakultaet VII -

Architektur, Umwelt, Gesellschaft-. Berlin, Technische Universität Berlin: 104.

Skogley, E. O. (1992). "The universial bioavailability environment/soiltest: Unibest."

Communications in Soil Science and Plant Analysis 23: 2225-2246.

Stagnitti, F., J.-Y. Parlange, et al. (1995). "Transport of moisture and solutes in the

unsaturated zone by preferential flow." Environmental Hydrology: 193-224.

Umweltbundesamt (2002). "Umweltdaten Deutschland 2002." 43, 48, 52.

VDLUFA and ATV-DVWK (2003). Gütesicherung zur landbaulichen Verwertung von

Klärschlämmen. Hennef, DWA.

Wang, H. L., G. N. Magesan, et al. (2004). "An overview of the environmental effects

of land application of farm effluents." New Zealand Journal of Agricultural Research

47(4): 389-403.

Wang, Z., L. Wu, et al. (2003). "A field study of unstable preferential flow during soil

water redistribution." Water Resources Research 39(4): 1075-1089.

Williams, A. G., J. F. Dowd, et al. (2003). "Preferential flow variability in a well-

structured soil." Soil Science Society of America Journal 67(4): 1272-1281.

Yang, J. E. and E. O. Skogley (1992). "Diffusion kinetics of multinutrient

accumulation by mixed-bed ion-exchange resins." Soil Sci. Soc. Am. J. 56: 408-414.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 22

2 Quantification of Nitrate Leaching in Soils using an

Adsorber Method

Abstract

Several methods and techniques have been applied so far to quantify diffuse

leaching losses under agricultural fields each of them bearing its limitations. The Self

Integrating Accumulator (SIA) method is proposed to overcome some of these

limitations. Basically SIA is an infinite-sink based method which integrates the solute

fluxes over time of installation in the soil. To test the suitability of the SIA I followed

three different approaches: I. With a chloride mass balance experiment on 6

agricultural fields with 5 replicates I determined accuracy and precision of the

method. II. To examine the capability to intercept preferential flow paths I used 3

qualitative approaches: IIa. A dye tracer experiment to mark the flow paths in the soil

and around the SIA. IIb. An experiment with a strongly sorbing solute (tributyltin),

which presumably is transported only preferentially. IIc. Nitrate leaching during a very

dry summer, where matric flow is very unlikely. III. In a plausibility experiment I

measured the mean nitrate losses under different land uses: forest, long term fallow

and agriculture. These were compared with published results. With 30 replicates

92 % of the validation value from the mass balance were recovered (Exp. I). The

variation is such that 10 replicates would estimate the true mean within an error of

less than 20 %. From II. it could be shown that SIA intercept and sample from

preferential flow paths. Mean values for N-losses were 7, 5, 21 kg ha-1 for forest,

fallow and agriculture respectively (Exp. III). I conclude that the Self Integrating

Accumulator (SIA) Method is valid to measure leaching losses of nitrate on a mass

per area basis in our experiments and is a promising method also for preferentially

transported solutes.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 23

2.1 Introduction

The loss of nutrients like nitrate and phosphate is not only an economic problem for

farmers, but it also causes eutrophication of rivers, lakes and coastal regions. In

Germany, close to 50 % of the nitrate input into surface water is due to influx of

groundwater previously loaded under agricultural fields. Nitrate is still the greatest

problem for German drinking water quality (Umweltbundesamt 2002) and of major

concern in many other countries, e.g. the USA (Nolan, Ruddy et al. 1998). Besides,

diffuse contamination by agriculture has been recognised as a major problem in the

European Community and is tackled on a legislatory level (2000; 2006).

It is therefore desirable to monitor leaching losses on a field scale under practical

farming and natural boundary conditions to estimate the effectiveness of

management practices, which are introduced to reduce the groundwater pollution by

diffuse pollutants. However, quantitative information on leaching losses of nitrate and

other diffuse pollutants to groundwater is only available at great costs and with

several limitations to the applied methods so far. This has been shown for tile drains

(Elliott, Cessna et al. 1998; Kladivko, Grochulska et al. 1999; Pampolino, Urushiyama

et al. 2000; Kladivko, Frankenberger et al. 2004), lysimeters (Jemison and Fox 1992;

Jene 1998; Pampolino, Urushiyama et al. 2000), suction plates (Dressel 2002;

Siemens 2003; Kosugi and Katsuyama 2004), suction cups (Kung 1990; Grossmann

and Udluft 1991; Selker, LeClerq et al. 1992; Elliott, Cessna et al. 1998; Netto 1999;

Pampolino, Urushiyama et al. 2000) and soil coring (Netto 1999; Pampolino,

Urushiyama et al. 2000).

The work and cost problem is mainly related to the repeated sampling and analysis

necessities. For lysimeters high investment and maintenance costs add to the limited

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 24

applicability. The technical limitations are mostly due to the heterogeneity of solute

transport in the field and the transfer of soil solution concentrations into mass fluxes.

Infinte-sink methods seem appropriate to avoid frequent sampling. Skogley (1992)

developed a method, in which ion exchange resin is packed tightly into plastic nets

with small grid spacing. This allows for diffusive movement of exchangeable ions to

the resin, which acts as an infinite sink. This method also captures a part of the

solute transported through convection. The two processes can not be separated,

however, so that the amount of leached N or plant available diffusive N can not be

calculated. Still, in comparison to suction cups the ion-exchange method was more

appropriate to detect fast spontaneous movement of bromide under intermittent

unsaturated flow conditions (Yang and Skogley 1992; Li, Skogley et al. 1993).

I took up the idea of ion-exchange resin to capture solutes, but wanted a method to

measure the absolute leaching losses on a [mass area-1 time-1] basis, which closes

the mass balance for the soil - groundwater path. Therefore, it was necessary to

differentiate between convective and diffusive transport to the resins. In addition, the

water and solute flux had to be representative for a defined cross sectional area of

the soil under a range of relevant environmental conditions.

The objectives of this paper are (i) to present the method developed which I call SIA

(Self Integrating Accumulator) and (ii) to demonstarte potentials and limitation of the

method. Fore the latter objective I tested the following hypothesis: (I) The method

collects the flux of conservative ionic solutes quantitatively. (II) The method collects

solutes transported mainly through preferential flow. (III) The method yields

reasonable solute-flux data under field conditions.

2.2 Materials and Methods

In order to test the hypothesis I conducted the following experiments:

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 25

I estimated accuracy and precision of the method with a chloride mass balance

experiment on 6 agricultural fields with 5 replicates.

To examine the capability to intercept preferential flow paths I used 3 qualitative

approaches: IIa. A dye tracer experiment to mark the flow paths in the soil and

around the SIA. IIb. An experiment with a strongly sorbing solute (tributyltin), which

presumably is transported only preferentially. IIc. Nitrate leaching during a very dry

summer, where matric flow is very unlikely.

I measured the mean nitrate losses under different land uses: forest, long term fallow

and agriculture.

2.2.1 Self Integrating Accumulators (SIA)

The Self Integrating Accumulators (SIA) consist of a cylinder (10 cm height; 10 cm

diameter) with a fine net on the bottom side filled with a mixture of quartz sand,

quartz silt and an anion-exchange resin (Figure 2-1; right side) or a hydrophobic

adsorbent for tributyltin. All materials were provided by TerrAquat, Stuttgart. The

mixture develops its own suction under drying soil conditions, so that leaching in

unsaturated soils is also intercepted.

The soil water containing the target solutes enters the open top of the SIA, passes

through and flows out at the lower end. During the passage the solute is sorbed to

the adsorber and immobilised. In this way, all transported target substances are

accumulated inside the SIA. Therefore, the SIA yields an integrated total value of the

leached amount for a defined cross sectional area of the soil during the

measurement period. This would allow for a projection to larger areas, if the

measured values are representative for the same cross sectional area of the soil as

of the SIA. This is to be tested in experiment I.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 26

It should be stressed that I do not measure residual concentrations or masses in the

soil, but only solutes transported by convection through the cross section of the top of

the SIA and that I do not intend to get information on how much water carried the

solutes that are captured.

The SIA are installed such that the soil on top of them is undisturbed (Figure 2-1, left

side). Hence the pore system, which is crucial for solute transport under natural

conditions, is preserved.

Figure 2-1: Installation and functional principle of the Self Integrating Accumulator

(SIA): The SIA is located below the undisturbed soil and surrounded by material with

the same hydraulic properties. Solutes are extracted out of the water by (specific)

sorption, while water passes through.

To that end soil profiles are dug to 20 cm below the envisaged installation depth.

From the wall of the profile side tunnels are dug under the undisturbed soil, in which

the SIA is installed afterwards. The side tunnel and the profile are closed after

installation. Thus, after installation soil management like ploughing, sowing,

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 27

harvesting etc. can be done without limitation, because no continous sampling of

solution is necessary. The installation under an undisturbed soil led to the idea that

the mass flux by preferential flow (like in earthworm holes, fissures) may also be

sampled successfully, because these structures simply end on the SIA-surface. A

qualitative answer shall be given by the experiments II a,b,c.

The installation time of the SIA may typically be 6 months or a vegetation period.

After this, the SIA are recovered and the resin-substrate mixture within the cylinders

is then split into 3 fractions, the upper 5 cm, middle 1 cm and bottom 4 cm. The

upper and middle layer fractions are analysed (see experimental sections)

separately. The bottom fraction normally is discarded. Only the upper 5 cm are used

for the interpretation of mass flow from above. The middle fraction is used as a

control whether the exchange capacity of the top layer was sufficient and no target

substance travelled deeper than 5 cm from the top. The lowest layer is discarded,

because its use is as a barrier for a mixture of possible upward fluxes like diffusion or

capillary rise, which should not be mingled with the downward flow.

2.2.2 Experiment I: Chloride tracer mass balance under field conditions

The tracer application took place in October 1999 on six fields owned by local

farmers and previously used for vegetable cultivation. Some of the fields were

already tilled; on others the plant residues from the previous crop were still visible.

None of the fields contained vegetation during the measuring period. Therefore,

chloride uptake by plants can be excluded.

At these six field sites between 20 and 250 km apart in SW-Germany with varying

soil management and soil properties I installed 5 SIA per site and applied the chloride

tracer. The idea was – given the limited resources – an experiment on several sites

with fewer replicates would be more meaningful than the same experiment on only

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 28

one site with more replications. Statistically, if we include more heterogeneity

(different fields) and still get a sensible tracer recovery, we have more information on

the validity of the method than with limited heterogeneity (one field).

2.2.2.1 Soils and site description

Site Karlsruhe is situated on the lower terrace of the river Rhine and developed on

sandy alluvial material. All other sites are influenced by loess deposits to some

degree, which is most typical for arable land in South West Germany and large parts

of Central Germany.

The average temperature is between 7 - 8.5°C and the mean annual precipitation

varies from 700 to 800 mm. During winter approximately 150 – 250 mm of water are

leached from the unsaturated zone towards groundwater. The sites are between

80 m and 350 m above sea level and the inclination of all experimental fields did not

exceed 2° (Table 2-1).

Table 2-1: Selected soil properties of the tracer experiment sites

Site Soil type

(FAO)

Texture

(USDA)

A-horizon /

B-horizon

pH (0.01 CaCl2)

A-horizon /

B-horizon

CECpot

[mmolc/kg]

A-horizon /

B-horizon

Bad Wimpfen Stagnic

Luvisol

Silty clay loam /

Silty clay

6.4/

6.1

245/

295

Ulm Calcic Luvisol Silt loam /

Silty clay loam

5.5/

5.7

185/

243

Tuebingen Haplic Luvisol Silt loam /

Silt loam

6.6/

7.2

300/

290

Stuttgart Stagnic

Luvisol

Silty clay loam/

Silty clay loam

6.5/

6.0

210/

220

Ludwigsburg Eutric Regosol Silt Loam/

Silt Loam

6.7/

6.8

158/

145

Karlsruhe Eutric

Cambisol

Loamy sand/

Loamy sand

5.3/

5.5

77/

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 29

2.2.2.2 Experimental

At these six field sites I installed 5 SIA per site in 60 cm depth in a transect of 150 cm

length.

On top of these plots I applied 1 kg of NaCl on an area of 2 m * 1.5 m (5.7 Mol*m-2)

evenly through a fine net. The tracer was then dissolved with an evenly distributed

irrigation of 2.5 mm.

Before the application of the tracer I collected soil samples from 0 - 15; 15 - 30; 30 -

50; 50 - 60 cm soil depth at each site, which were analysed for chloride background

concentrations.

The field sites were then left untreated for six months to allow leaching of chloride by

the winter rains.

At the end of March 2000 the soil body on top of the installed SIA was sampled on all

sites at depth of 0 - 15; 15 - 30; 30 - 50; 50 - 60 cm to allow for chloride residue

analysis. This was done by collecting 3 soil core samplers (100 m³) and about 1 kg of

loose soil material at every depth. I sampled vertically at the profile wall and

horizontally on plains at several depths to avoid singularities at the profile wall. In

addition, large undisturbed samples (~ 10 kg / site) were taken from the top 15 cm of

the soils, because the largest residual amounts and largest bulk density

heterogeneities were expected in this soil depth. I also took samples from the

disturbed area in the refilled pit and from undisturbed soil close by to account for

anomalies or later chloride application.

At the same time the SIA were collected for analysis. The resin-substrate mixture

within was sampled in three layers (Top 5 cm; Middle 1 cm; Bottom 4 cm). Top and

middle layer were analysed for chloride after a 0.5 M H2SO4 extraction at a water-

solid mass ratio of 4:1.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 30

Soil cores were analysed for bulk density and water content. All soil samples were

analysed for chloride content of homogenised samples in 2 replicates to control

homogenisation. Laboratory analysis was performed with a chloride selective

electrode against a Pt reference electrode (WTW, Weilheim) in a range of 10-1 to 10-6

mg Cl*L-1.

2.2.2.3 Calculations

Performance of the SIA was calculated as:

Recovery [%] = (Recovered amount SIA [mol * m-²] /(Total amount applied [mol * m-²]

– Residual amount in top 60 cm of soil profile [mol * m-²])) * 100 %

The residual amount is given by:

Residual amount [mol * m-²] = Σi {Concentration [gCl * L-1] * sample weight-1 [L * gsoil -1]

* layeri [m] *bulk densityi [gsoil * m-³]*molar mass-1 [mol * gCl -1]} for each soil layer i

from 0 – 60 cm.

2.2.2.4 Statistics

The data were normally distributed for the whole set (Kolmogorov-Smirnov-Test).

Therefore, an analysis of variance was performed to distinguish between

experimental sites.

I planned the design to cover a wide range of external factors like management,

microclimate, soil etc. and were mainly interested in the general applicability of the

SIA method regardless of sites. Therefore, it was interesting to treat the whole data

set as one experiment. This is also justified statistically by the fact that the analysis of

variance did not yield significant differences between the experimental sites.

I used combinatorial analysis to calculate the error of estimate for all numbers of

replications. Combination mathematics deals with the question: How many

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 31

possibilities are there to draw at random distinguishable pieces (e.g. single

measurements) from a given number of pieces (e.g. a data set) without putting them

back. Probabilities can then be calculated by dividing the chance (number of possible

draws) to draw a certain subset or a number of subsets by the total possible

combinations.

The results can then be used for statistical interpretation. For example, if an

experimentator chooses to measure leaching triplicate with SIA, what would be the

mean error and how would this improve, if he chose to replicate his measurements

15 times with this method?

To calculate e.g. triplicate replication I drew every possible combination of 3

measurements from our data set of 30, which makes 30 * 29 * 28 or 30! / 27! =

24,360 possible combinations. To calculate this for all sensible replications (2 – 29)

about 1 billion calculations were necessary. All results were then grouped in 1 %

steps to get cumulative frequencies of tracer recovery with the mean experimental

recovery set as 100 %. The experimental mean was set as 100 % instead of the real

mean, because we can not know a priori, whether we will approach 100 % of the real

value with more replicates due to possible systematic errors of the method. These

cumulative frequencies were then used to see how many combinations fall into a

certain recovery range, e.g. between 80 % and 120 %, and how many do not.

The advantage of this method is that confidence levels can be obtained directly from

the data of the validation experiment without any statistical assumption on the data

distribution.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 32

2.2.3 Experiments II a), II b) and II c): Sampling of preferential flow paths

2.2.3.1 Experiment II a): Brilliant Blue tracer experiment for the

identification of preferential flow paths

I wanted to test qualitatively, whether SIA are able to intercept preferential flow paths

in the soil. Therefore, I used the dye Brilliant Blue in two experiments at Muenster,

NW-Germany, and Bern, Switzerland. In Muenster I installed SIA at 35 and 60 cm

depth on a Fimic Anthrosol with a loamy sand texture and applied an irrigation of 30

mm on a 3* 2 m² plot on top of the SIA. The irrigation water contained approximately

1 g*L-1 of the dye Brilliant Blue, which colours the actual flow paths, through which it

is moving.

After 2 hours I dug to the SIA and evaluated optically the presence of Brilliant Blue in

the soil profile and on top of the SIA.

A similar experiment was performed on a site near Bern (Inforama Rütti, Zollikofen).

The soil was an Eutric Cambisol derived from a heterogeneous morainic silt loam to

clay loam material with < 5 % of stones. Here, the installation depths of the SIA were

15, 30 and 60 cm.

2.2.3.2 Experiment II b): Mass flux experiment with the sorbing solute

tributyltin

In a field experiment of 4 months (April 2000 – July 2000) I applied sewage sludge

containing a known amount of tributyltin on 2 agricultural fields afterwards sown with

carrots.

Soils and Sites

Both sites had an inclination < 10. Site Forcheim is situated on the lower terrace of

the river Rhine and developed on sandy alluvial material. Its elevation is 116 m

above sea level, the mean annual temperature is 9.50C and the mean annual

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 33

precipitation is 650 mm. Site Rottenburg is situated on an elevation of 440 m above

sea level SW of Stuttgart and developed on loess deposits. Its mean annual

temperature and precipitation are 80C and 750 mm (Table 2-2).

Table 2-2: Selected soil properties of the tributyltin experimental sites

Site Soil type

(FAO)

Texture

(USDA)

A-horizon /

B-horizon

pH (0.01 CaCl2)

A-horizon /

B-horizon

CECpot

[mmolc/kg]

A-horizon /

B-horizon

Forchheim Dystric

Cambisol

Sandy loam /

Loamy sand

5.1/

5.5

48/

Rottenburg Calcic Luvisol Silty clay loam

/Silt loam

7.3/

7.1

243/

196

Experimental

At both sites SIA were installed in two plots (100 m²) at a depth of 80 cm. The SIA

were distributed evenly in 5 pits containing 2 SIA (10 replicates) at each plot. Wet

sewage sludge equivalent to 5 t ha-1 dry substance was applied at one plot at each

site. The other plot received no sewage sludge and served as a control. Carots

(Daucus carota ssp. Sativa ‘Nantaise’) were sown and harvested 3 months later.

After the harvest, SIA were recovered and taken to separate analysis.

Analysis

Tributyltin was analysed by the Wave Ltd. Laboratory (Stuttgart) in a modified

DIN 38 407-13 procedure with a hexane extraction, derivatisation with ethylborate

and measurement with GC-MS using tripropyltin as an internal standard.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 34

2.2.3.3 Experiment II c): Mass fluxes of nitrate during a dry vegetation

period

General

Nitrate leaching loss mass fluxes were measured in a 3-year field experiment at 9

sites and 2 fertilizer treatment plots per site. 10 SIA per plot (180 total) were used to

monitor continuously nitrate leaching losses during vegetation (March to September)

and winter period (October to February). I present the results from the vegetation

period 2003 because of its unusual weather conditions.

Soils and sites

All nine sites are situated in a hilly region of shell-limestone underlying weathered

loess deposits. The loess loam cover averages 60 cm. The site inclination ranges

between < 10 to 40. The mean annual precipitation is 650 mm and mean annual

temperature is 70C.

The soils are mostly Calcic Luvisols (FAO) or similar. The pH is in the range of 6.3 –

7 for the top and subsoils. The CEC is in the range of 200 – 350 mmolc kg-1. The

texture is silty loam to silty clay loam.

Experimental

Each site was treated with two fertilizers with exactly the same content of N.

CULTAN (Controlled Uptake Long Term Ammonia Nutrition) is a concentrated

Ammonium-Nitrate-Urea solution fertilizer, which was placed in thin lines close to the

seeds, whereas KAS (Calcareous NH4NO3) is a granulate, which was spread evenly

over the field.

Each site has the same basic 3-year crop rotation of winter rape (Brassica napus L.

var. napus) - winter wheat (Triticum aestivum L.) – phacelia intercrop (Phacelia

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 35

tanacetifolia Benth.) - summer barley (Hordeum vulgare L.). In each of the 3

experimental years each crop was represented by 3 sites, so that after the

experiment each site had the same rotation, but with a different starting crop.

The nitrate leaching losses were monitored with 10 replicate SIA in 60 cm depth

(average loess cover depth) on each site and treatment (180 total). The SIA on all

fields were exchanged within 3 - 4 consecutive days in March and September.

Basic weather data (Precipitation and temperatures) were collected at two sites

within the experimental area.

Analysis

Each SIA was split into 3 layers (see above) and the top and middle layers were

analysed separately. The samples were homogenized and an aliquot of 15 g was

extracted with 60 mL of a 1 M NaCl solution. The measurement was performed on a

N rapid flow analyser (SANplus, Skalar, Breda, The Netherlands).

2.2.4 Experiment III: Nitrate losses under different land uses

General

A four year monitoring experiment was performed to highlight the possible

contributions of different land uses to the high nitrate loads in an aquifer in central

Germany. Three agricultural fields with regional crops, a long term fallow and a forest

site were monitored for nitrate losses with SIA.

Soil and site description

The sites are situated on an upper terrace of the river Main (Central Germany)

covered with coarse to medium sized sandy eolian deposits (flat dunes) with a depth

of 1 – 2 m. The soils are Dystric Cambisols with a pH of 5 (forest) to 6.5 (agriculture)

in the top soil. All sites have an inclination of < 1 °. The elevation is 130 m above sea

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 36

level, the mean annual temperature is 9°C and the mean annual precipitation is

680 mm.

Experimental

At all 5 sites SIA were installed at a depth of 90 cm. The SIA were distributed

diagonally across the fields (3 crops and fallow) in three pits containing 10 SIA and a

transect of 4 pits within the forest, also containing 10 replicates. SIA were changed at

the end of September and in April from Sept. 1998 to Sept. 2002.

Analysis

Analysis was the same as in 2.2.3.

2.3 Results and Discussion

2.3.1 Experiment I: Tracer Recovery

Precision

The chloride tracer recovery with the SIA method ranged between a mean of 63 %

with a standard deviation (Std. dev.) of 44 % for site Stuttgart and a mean of 128 %

(Std. dev.: 15 %) for Karlsruhe with an overall mean of 92 % (Std. dev.: 39 %) for all

30 replicates. The mean recoveries were 66 % for Tübingen, 89 % for Bad Wimpfen,

104 % for Ulm and 104 % for Ludwigsburg. This is a better recovery for all sites than

in the reported examples for other methods, where chloride or bromide were used as

a tracer. Dressel (2002) overestimated by 30-40 % with suction plates, Jemison and

Fox (1992) recovered 45 -58 %, but Boll, Selker et al. (1991) only 6.5 % in pan

lysimeters. Jene (1998) recovered 48 % with a suction cup grid and 77 % in a

monolith lysimeter. Passive capillary or wick samplers (Boll, Selker et al. 1991;

Brandi-Dohrn, Dick et al. 1996) recovered between 29 – 63 % of a bromide tracer.

Only Siemens (2003) got similarly good results (118 % recovery) with tension

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 37

controlled suction plates in a sandy soil, but his experiment was limited to a 2-day

irrigation experiment.

Therefore, the SIA method with a total mean of 92 % is more efficient in measuring

chloride leaching losses in a long term leaching experiment than any other reported

field method.

Chloride and bromide have been used to evaluate conservative solute transport and

to serve as a good approximation for water movement (e.g. Li, Skogley et al. 1993;

Brandi-Dohrn, Dick et al. 1996; Brown, Hollis et al. 2000; Wang, Wu et al. 2003). It

has been shown that both are equally suitable as tracers (Saffigna 1977), even

though anion exclusion may be responsible for a slightly faster movement compared

to tritium tracer studies (Logsdon, Keller et al. 2002).

Therefore, our understanding is that chloride as a conservative tracer with

insignificant adsorption in the investigated soils is a good indicator for the water flux

in the soil. If the chloride mass balance is correct, the water balance is supposed to

be correct as well. I conclude that I did not only measure solute flux within an error of

10 %, but also the water flux through the SIA was represented within the same range

of error.

This is important to decide whether the method may also be suitable for sorbing

and/or non-conservative tracers. The effectiveness of the method with these

substances depends on the validity of the water balance and suitable adsorber

materials.

Accuracy

From the statistical evaluation of the samples from 6 field sites (Figure 2-2) one can

calculate the amount of replications necessary to obtain a certain accuracy of the

estimated mean.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 38

Combinatorial analysis

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 2 4 6 8 1012141618202224262830

Number of replicate measurements

Probability to be inside

confidence interval

95 to 105 90 to 110 80 to 120

Confidence interval [%]:

Figure 2-2: Results of an exact combinatorial analysis. Relation between SIA method

precision and number of replications. At least 10 replicates are needed to achieve

20 % accuracy with a probability of > 90 %.

According to combinatorial analysis two replicates estimate the overall mean within

an error of 20 % in less than 50 % of the six month measurements, whereas 10

replicates yield a value within the 20 % confidence interval in more than 90 % of the

measurements. A high accuracy of the mean (e.g. +/- 5 % estimates) can only be

obtained with high replication numbers > 27.

It also has to be considered that I compare relative precision to the overall mean of

92 % recovery of the true value. The method may be able to approach 100 %

recovery with higher replications. In this case, the prediction from our analysis may

be wrong for n close to 30 and therefore also for high accuracies.

Brandi-Dohrn, Dick et al. (1996) calculated that 25 passive capillary (wick) samplers

or 37 suction cups were required to estimate the mean within an error of 30 % for a

0.05 confidence interval. This compares with 12 SIA replicates for an error of 20 % at

the 0.05 confidence interval and 10 replicates at the 0.10 confidence interval. This

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 39

difference to our results may be due to the wider suction range of the SIA compared

to wick samplers and the better defined sampling area compared to the suction cups.

All discussed methods need high replication numbers due to the heterogeneity even

within one ‘homogeneous’ soil unit. For example, Netto (1999) showed in a field

study that for none of the tracers bromide, nitrate and bentazon a distance

dependent autocorrelation existed, which is a clear indicator of inherent

heterogeneity.

Thus, the SIA shows a favourable precision and spread to measure the flow of the

non-sorbing tracer chloride and I deduce that the water flux through the SIA is

approximately representative for the flux through the same cross sectional area in the

soil.

2.3.2 Experiments II a, b ,c: Qualitative evidence for sampling of

preferential flow

It is a priori clear that every measurement disturbs the very system it tries to

describe. Thus, it is crucial to know, whether certain main processes are excluded or

included in the measurement. It is quite common to differentiate solute transport into

normal/matric/Richards/capillary and fast/preferential/macropore flow, even though

the concepts to discern them differ.

Therefore, I needed to test, if the flow captured by the SIA method includes part or all

of possible preferential flows. Unfortunately, the reasons and circumstances for

preferential flow are manifold. Therefore, I picked 3 different situations, where

preferential flow supposedly is the main explanation for solute transport.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 40

2.3.2.1 Experiment II a): Brilliant Blue tracer experiment for the

identification of preferential flow paths

Figure 2-3 is taken from the dye tracer experiment in Bern, Switzerland. Black and

dark shades represent regions with visible Brilliant Blue dye. Squares show the

locations of the side tunnels in the profile, which are still closed.

Figure 2-3: Soil Profile (Bern) after a Brilliant Blue dye tracer experiment. Black and

dark grey show presence of dye. Squares show the closed entrances of the side

tunnels under the undisturbed soil in 15, 30 and 60 cm depth, where the SIA are

installed.

It can be seen that in this short term experiment the colour of the dye is not only

distributed over the soil profile, but also flows out of the side tunnels, where the SIA

are installed. Therefore, the SIA and refilling of the tunnel did not pose an obstacle to

the applied water flow. The lowest SIA (60 cm depth) also intercepted a preferential

flow path, which is not visible in the upper part of the profile, but lies deeper in the

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 41

body of the soil. Here, only the right side of the tunnel is coloured indicating a smaller

flow path than for the upper SIA at 15 and 30 cm depth.

Figure 2-4 is taken from the dye tracer experiment in Muenster, Germany. Black and

dark shades represent regions with visible Brilliant Blue dye. The plain on top of a

SIA, which stuck in a side tunnel during the experiment, is prepared, but the SIA only

partly uncovered. About one quarter of the SIA is visible. The casing is marked as

thick black line. The rest is covered under the undisturbed soil marked as the shaded

and striped region on the upper and right hand side. The lower and left hand side

show formerly undisturbed soil and filling material of the tunnel in the same plain

outside the SIA. The dye is found inside and outside the SIA and the pattern is

continuous, e.g. ‘south’ of the border between the shaded area and the prepared

plain Figure 2-4.

Figure 2-4: SIA in 60 cm depth after a Brilliant Blue dye tracer experiment

(Muenster), partly excavated. Left: Black and white conceptual picture. Right: Color

Picture. Black and dark grey (left) or blue (right) show presence of dye. Shaded area

(up and right) is not yet excavated. Black quarter circle is the casing of the SIA and

separates outside and inside.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 42

Dye tracers have been used extensively to mark preferential flow paths (e.g.

(Ghodrati and Jury 1990; Roth, Jury et al. 1991; Wang, Wu et al. 2003). In both of

our experiments dye could be found in the SIA. Thus, the SIA intercept preferential

flow at least qualitatively in short term dye experiments.

2.3.2.2 Experiment II b): Mass flux experiment with the sorbing solute

tributyltin

Tributyltin is a sorbing substance with a reported KOW of 251 (O'Loughlin, Traina et

al. 2000) and a very high affinity to Aldrich Humic Acids with a KOC of 106.1. It is

thought to be little mobile in a KOW - classification for pesticides by Wilson, Duarte-

Daidson et al. (1996) grouping pesticides in 4 classes between very mobile and little

mobile and highly sorptive to humic substance.

On site Forchheim (sandy loam over loamy sand) 0.5 % of the applied tributyltin were

leached out of the rooting zone into the SIA. Only one out of ten SIA contained a

detectable amount of tributyltin. In Rottenburg (silty clay loam over silt loam), 2.2 %

of the applied amount was lost by leaching. Four out of ten SIA contained detectable

amounts of tributyltin. The SIA at both control plots contained no tributyltin.

In the experiment, microbial degradation and other dissipation processes could not

be quantified and residues in the soil, if any, were below the detection limit of

1 µg*kgsoil-1. Therefore, the experiment gives no quantitative information on the

performance of the SIA method. The experiment was performed in late spring and

summer, a period when the soil becomes dry in Germany and leaching of water is

related to heavy rains. The chemical properties of tributyltin can be compared to

those of pesticides, for which preferential flow is thought to be the main transport

mechanism (Flury 1996; Elliott, Cessna et al. 1998).

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 43

Preferential transport is the most likely process for the detection of any tributyltin

mass flux and that the SIA were capable to detect it. Furthermore, the experiment

shows that by accumulation in the SIA tributyltin was detectable, whereas no

residues could be found in the soil probably due to the lower detection limit.

2.3.2.3 Experiment II c): Mass fluxes of nitrate during a dry vegetation

period

The measurement from March to September 2003 took place under extremely dry

conditions in the study area. The precipitation sum during that period was 237 mm

compared to an average of 400 mm in the years 1977 to 1995. It was a period, when

potential evapotranspiration always exceeded the precipitation. Still, some rainstorm

events at the end of July or a wetter period in May (Figure 2-5) may have induced

preferential flow.

01.03.2003

15.03.2003

29.03.2003

12.04.2003

26.04.2003

10.05.2003

24.05.2003

07.06.2003

21.06.2003

05.07.2003

19.07.2003

02.08.2003

16.08.2003

30.08.2003

13.09.2003

27.09.2003

Date

Precipitation [mm]

Figure 2-5: Daily precipitation during a field experiment for the leaching of nitrate on

nine farm fields with two fertilizer treatments. Marked rainstorm events at the end of

July.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 44

The N leaching data (Figure 2-6) from 180 SIA show that in 56 % of the SIA N-

transport was not detectable, but for several measurement points in space, leaching

losses were considerable. Three SIA (1.8 %) received N-losses of more than

10 kg N/ha during the installation period. About 20 % of the SIA received more than

1 kg N/ha. Each of the nine fields had at least two out of 20 SIA with detectable N-

losses. On one field with notably higher clay content due to a mixing of the loess with

shell limestone residual clay every SIA contained detectable amounts of nitrate,

which I attribute to the higher frequency of shrinkage cracks visible at the end of the

measurement period. Field heterogeneity leading to heterogeneous flux has also

been reported by Netto et al. (1999), who showed that no autocorrelation existed

between adjacent nitrate solution measurements which were repeated during a

summer period in SE France.

100

120

01234567111625

N-Loss [kg/ha]

Frequency

Figure 2-6: N-Leaching (60 cm below surface) under dry conditions. Higher N-losses

in a small number of SIA may be explained by the sampling of preferential flow.

(Total N = 180)

Preferential flow in only a few points in space, probably induced by scattered

rainstorm events is the most likely explanation for our data. Again, we see this as

qualitative evidence for the capability of the SIA to sample preferential flow paths. In

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 45

this case, the sampling took place under very dry soil conditions, where e.g. suction

cups would not detect anything.

2.3.3 Experiment III: Plausibility comparison of different land uses

It is well known that different land uses contribute differently to nitrate or other solute

leaching from the root zone. This should be reflected by the SIA measurement.

In the average of a 4-year measurement, forest looses 7 kg N ha-1 (Std. error:

3 kg N ha-1), a long term fallow 5 kg N ha-1 (Std. error: 1 kg N ha-1) and three

agricultural fields 21 kg N ha-1 per year (Std. Error: 3 kg N ha-1, Figure 2-7).

234

Forest Fallow Agriculture

Land use

N [kg/ha]

Winter Summer Error bar = Std. Error of Mean

Figure 2-7: Average N-Losses by leaching under different land uses. Data from 7

sites (5 crop fields; 1 long term fallow; 1 forest) and 3 years of continuous

measurement (Summer: N=219; Winter: N=191).

The area has rather poor soils for the farmers. The limiting factor is the small water

retention capacity of the sandy soils. The farmers therefore apply lower N-amounts

(e.g. 100 kg N*ha-1 for winter wheat) than in other areas (Winter wheat: 120 –

160 kg N*ha-1). The N-losses of ~ 20 kg N*ha-1*year-1 are in good agreement with

long term experiments in Eastern Germany on a Haplic Phaeozem with sandy loam

texture (Graz et al., 1997). In these, leaching losses were calculated as the

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 46

difference between long term inputs and removal by the harvest in a one hundred

year rye (Secale cereale) experiment. For several treatments of N, P, K fertilizers the

surplus is in the range of 18 - 36 kg N ha-1 year-1. Atmospheric N-deposition or

potential denitrification were not considered in this case.

Zhu, Fox et al. (2003) report average N-losses of 17, 39 and 112 kg N ha-1 year-1 with

N-fertilizer treatments of 0, 100 and 200 kg N ha-1 year-1 respectively. These studies

show that with comparable N-inputs comparable losses occur under agricultural

fields in the long run. Cavero, Beltran et al. (2003) state that in two watersheds in the

Ebro valley (Spain) under high quality irrigation management N-losses ranged

between 8 and 22 % of the N-inputs, which is in the percent range of our data.

The groundwater recharge in this region is about 200 mm year-1 and the groundwater

wells produce water with nitrate concentrations of 40 – 60 mg L-1 in areas where the

recharge is predominantly from agricultural land. An extrapolation of our measured

mean to a whole recharge area would yield a concentration of 50 mg L-1 for the

leaching water, which builds up the groundwater. Other measurements (own data,

unpublished) show that denitrification can be neglected at least for some wells, so

that a direct comparison is possible.

Thus the data produced in this 4-year experiment are plausible on the background of

other literature and the hydrological data in the study area.

2.4 Conclusions

A new method (Self-Integrating Accumulators, SIA) has been suggested to measure

directly the total leaching losses of solutes. It combines the adjustment to the soil

matric potential with the accumulation of a range of solutes onto suitable adsorbers.

The SIA method lowers labour and analysis costs compared to soil solution

sampling, but it has no time resolution during the installation period.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 47

The validation of the SIA with a tracer mass balance experiment has shown that the

method is capable of measuring the long term leaching losses of solutes under field

conditions. The error of the measurement is < 10 % with 30 replicates. No less than

10 replications per ‘homogeneous’ field are recommended to hit the mean +/- 20 %

with > 90 % probability.

Evidence from a short term dye tracer experiment at two sites, a 6-month field

experiment with nitrate under dry soil conditions and a 4-month field experiment with

tributyltin leads to the conclusion that the SIA method is capable of measuring solute

transport by preferential flow. However, the evidence is qualitative, because all

experiments lacked a proper mass balance. Therefore, it cannot be decided as yet,

whether and with how many replications the SIA capture preferential flow

quantitatively.

An applied groundwater catchment study with nitrate has shown the ability of the SIA

to get plausible information on the relative contributions of different land uses to the

groundwater pollution.

I conclude that the method may be an alternative to existing field methods, if the

main objective is the quantification of a mass area-1 solute flux.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 48

Literature

(2000). Richtlinie 2000/60/EG zur Schaffung eines Ordnungsrahmens für

Maßnahmen der Gemeinschaft im Bereich der Wasserpolitik. 2000/60/EC: 72.

(2006). Directive establishing a framework for the protection of soil and amending

Directive 2004/35/EC (Proposal). 2006/0086(COD): 30.

Boll, J., J. Selker, et al. (1991). Water quality sampling under preferential flow

conditions. Int. Symposium on lysimetry, American Society of Civil Engineers.

Brandi-Dohrn, F. M., R. P. Dick, et al. (1996). "Field evaluation of passive capillary

samplers." Soil Sci. Soc. Am. J. 60: 1705-1713.

Brandi-Dohrn, F. M., R. P. Dick, et al. (1996). "Suction cup sampler bias in leaching

characterization of an undisturbed field soil." Water Resources Research 32, No. 5:

1173-1182.

Brown, C. D., J. M. Hollis, et al. (2000). "Leaching of pesticides and a bromide tracer

through lysimeters from five contrasting soils." Pest Management Science 56: 83-93.

Cavero, J., A. Beltran, et al. (2003). "Nitrate exported in drainage waters of two

sprinkler-irrigated watersheds." Journal of Environmental Quality 32(3): 916-926.

Dressel, J. (2002). Transport von Ethidimuron und Methabenzthiazuron in einer

Parabraunerde - Experimente und Modellierung. Institute for Chemistry and

Dynamics of the Geosphere. Jülich, University of Bonn.

Elliott, J. A., A. J. Cessna, et al. (1998). "Soil processes and chemical transport:

Leaching and preferential flow of clopyralid under irrigation: Field observations and

simulation modeling." J. Environ. Qual. 27: 124-131.

Flury, M. (1996). "Experimental Evidence ofTransport of Pesticides through Field

Soils - A review." J. Environ. Qual. 25: 25-45.

Ghodrati, M. and W. A. Jury (1990). "A field study using dyes to characterize

Preferential flow of water." Soil Sci. Soc. Am. J. 54: 1558-1563.

Grossmann, J. and P. Udluft (1991). "The extraction of soil water by the suction cup

method: A review." J. Soil Sci. 42: 83-93.

Jemison, J. M. and R. H. Fox (1992). "Estimation of zero tension pan lysimeter

collection efficiency." Soil Science 154: 85-94.

Jene, B. (1998). Transport of Bromide and Benazolin in Lysimeters and a field plot

with grid suction bases in a sandy soil. Institute of Soil Science. Stuttgart,

Hohenheim.

Kladivko, E. J., J. R. Frankenberger, et al. (2004). "Nitrate leaching to subsurface

drains as affected by drain spacing and changes in crop production system." Journal

of Environmental Quality 33(5): 1803-1813.

2. Quantification of Nitrate Leaching in Soils using an Adsorber Method 49

Kladivko, E. J., J. Grochulska, et al. (1999). "Pesticide and nitrate transport into

subsurface tile drains of different spacings." J. Environ. Qual. 28: 997-1004.

Kosugi, K. and M. Katsuyama (2004). "Controlled-suction period lysimeter for

measuring vertical water flux and convective chemical fluxes." Soil Science Society

of America Journal 68(2): 371-382.

Kung, K.-J. S. (1990). "Preferential Flow in a sandy vadose zone: 1.Field

Observation." Geoderma 46: 51-58.

Li, Z. M., E. O. Skogley, et al. (1993). "Resin adsorption for describing bromide

transport in soil under continuous or intermittent unsaturated water flow." J. Environ.

Qual. 22: 715-722.

Logsdon, S. D., K. E. Keller, et al. (2002). "Measured and predicted solute leaching

from multiple undisturbed soil columns." Soil Sci. Soc. Am. J. 66: 686-695.

Netto, A. M. P., R.A.; Gaudet, J.P. (1999). "Field study on the local variability of soil

water content and solute concentration." Journal of Hydrology 215: 23-37.

Nolan, B. T., B. C. Ruddy, et al. (1998). "A national look at nitrate contamination of

groundwater." Issue of Water Conditioning and purification 39(12): 76-79.

O'Loughlin, E. J., S. J. Traina, et al. (2000). "Association of organotin compounds

with aquatic and terrestrial humic substances." Environmental Toxicology and

Chemistry 19(8): 2015-2021.

Pampolino, M. F., T. Urushiyama, et al. (2000). "Detection of nitrate leaching through

bypass flow using pan lysimeter, suction cup, and resin capsule." Soil Science and

Plant Nutrition 46 (3): 703-711.

Roth, K., W. A. Jury, et al. (1991). "Transport of chloride through an unsaturated field

soil." Water Resources Research 27: 2533-2541.

Saffigna, P. G. K., D.R.; Tanner, C.B. (1977). "Lysimeter and Field Measurements of

Chloride and Bromide Leaching in an Uncultivated Loamy Sand." Soil Sci. Soc. Am.

J. 41: 478-483.