Madsen, H. B. and Jensen, N. H. 1985: The establishment of
pedological soil data bases i Denmark. Geografisk Tidsskrift 85:
1-8, Copenhagen, Oct. 1985.

The construction of a main gas pipeline from the North Sea across Denmark offered a unique possibility of taking a subsequent number of soil profiles. Roughly 835 profiles have been described in detail according to a system very similar to FAO's Guidelines for Soil Profile Description, and samples from each horizon have been analysed in the laboratory. Based on easily detectable stable characteristics of the soil, about 8000 soil classifications were made down in the trench. The applied classification system, the analyses, the data storing system, and the principal soil conditions found along the pipelines are described in the paper.

H. B. Madsen, Institute of Geography, Østervoldgade 10, DK-1350
K. N. H. Jensen, Ministry of Agriculture, Bureau
of Land Data, Enghavevej 2, DK-7100 Vejle.

Keywords: Soil Data Base, Denmark

In 1975-80 a nationwide Danish soil classification of the farmland was carried out comprising soil sampling at about 35,000 sites. At every site a sample was taken from 0-20 cm depth, and at selected sites also from 35-55 cm depth. Texture, organic matter, and content of calcium carbonate were determined for all samples. On the basis of these analyses, soil maps in scale 1:50,000 were constructed, showing principally the texture at 0-20 cm depth (Mathiesen 1980). All soil maps and sampling sites have been digitized and hereafter stored in a computer system at the Ministry of Agriculture, Bureau of Land Data (ADK).

The computerized data have been widely used in agricultural planning. For this purpose transfer models have been elaborated to calculate e.g. the actual and potential drainage requirements within larger regions and the amount of nitrate leached from the farmland to the watercourses. courses.To run these transfer models a need has arisen for more detailed information about the different soil types, especially with regard to drainage classes, soil water retention, root development, variation in texture and the pedological development within the uppermost 2 m of the profile.

The establishment of the main gas pipeline system from the North Sea gasfields across Denmark in 1981-83, fig. 1, represented a unique possibility of studying soil profiles and to build up computerized pedological soil data bases comprising systematical soil classifications, profile descriptions and analytical data.

This paper describes the pedological investigations carried out along the main gas pipeline system in 1981-83 and examples of combining the results with existing computerized soil- and landscape maps.

FIELD INVESTIGATIONS

The field investigations were carried out during the constructionperiod
of the main gas pipeline system, that

means May to November in 1981-83. Through the farmlanda 20-m wide working belt was established. First the topsoil was romoved and placed along the one side of the working belt. The pipe sections were connected and an approximately 2-m deep narrow trench was dug, asymmetricallyplaced in the working belt, fig. 2. The subsoil was placed opposite to the topsoil. In forest areas the working belt was 12 meter wide and there was no separationbetween top- and subsoil.

The field investigations comprised pedological soil classifications for every 25 meter along the trench and detailed profile descriptions and soil sampling at the borderline between working belt and undisturbed agricultural land. Furthermore, some detailed profile investigations and soil samplings were carried out in nearby forests for comparing the pedological development in uncultivated soils with cultivated ones which are ploughed, limed, and manured.

In total, 835 detailed profile descriptions and 8440 pedological
soil classifications were carried out.

Soil classifications in the trench

After the pipe had been installed in the trench bottom and before replacement of the subsoil to close the trench, soil classifications were carried out for every 25 meter over long distances in the roughly 2-m deep trench. Only where this followed highways, the soil profiles were frequently too much desturbed to be classified, and in a few places the trench was closed immediately after the pipe had been laid down so no time was left for soil classifications.

Because the topsoil was removed before the trench was dug, the soil classification was frequently based on only a thin A1 layer and on B and C horizons, and in some cases only on B and C horizons. A description of the Al horizon can therefore not be complete, and problems of soil depth or depth to horizons with different soil characteristics arise. By the assessment of soil depth it was assumed that 30 cm of topsoil had been removed. If a soil layer e.g. began 40 cm below the surface of the trench, it was assu-

med to begin 70 cm below the surface of the undisturbed
profile, and it was classified as beginning at 70 cm depth.

The field classifications must be based on parameters which are easily detectable on location, because no analytical data are or will be available for these classifications. Thus it has been necessary to build up a Danish pedological soil classification system based on soil characteristics which can be easily recognized and quantified in the field. These are

The organic matter in the Al-horizon

The genesis, thickness, and colour of the B and C horizons

Soil depth to horizon with gley features

Soil depth to horizon with different amounts of calcium
carbonate

Soil depth to bedrock

The presence of fragipans or cemented soil layers
within the uppermost 120 cm of the profile
The presence of sodium-rich soil layer within the uppermost
120 cm of the profile

The presence of degradated B horizons
Texture

The system developed is an uncontrolled parametric system where no parameter excludes other parameters when present. In that way statistical treatment of the material will be favoured. The system is hierarchical with four levels: order, group, series, and phase. The soils are defined through the presence or absence of diagnostic horizons or characteristics. In table 1 the three highest levels are shown with a brief, but not complete definition. The system is discussed in detail in Madsen (1983).

The soil names are constructed according to the following

- Only one name is given at order level, e.g. lessive.
- At group level up to three designations may be added to

the order-name. The first group-name, e.g. typi, is always
added, but the second and third group-name are
only used if the defined paramenters are present.

- Order- and group-names are written as one word begin

ning with capital letter, e.g. Pseudogleytypilessive.
- The designation at series level is only used if the de

fined parameters are present. The series-names are written in front of the group-order-name as individuals beginning with non-capital letter e.g. gleyey entic Pseudogleytypilessive.

In the trench the soil classification was carried out to maximum series level but at least to group-level and the data were stored in the ADK computer system in numerical form. The figures are given in table 1, and in table 2 different soil names and their translation into numerical form are given.

The location of the field classifications was determined
from distance-marks on the pipe and then plotted on aerialphotographs
in scale 1:4000 or on working plans in

scale 1:1000. These plots were later readjused to »as built plans« showing the precise location of the pipeline, and they were digitized with reference to the UTM-coordinate system, so the soil classifications can be combined with already existing soil maps and other information available in the ADK-computer system (Mathiesen 1984, Madsen 1984).

Soil profile investigations

Based on the soil classifications in the trench, sites were selected for detailed profile investigations and sampling for soil chemical and physical analyses. The selection of the profiles were based on differences in parent material, drainage class or on the presence of distinct pedological features such as placic horizon, fragipan or bog iron. The profiles were frequently situated as toposequences.

The pits were dug on the borderline between the working belt and the agricultural land, so the profile description was carried out on an undisturbed profile wall outside the working belt. The pits were 1.5 to 2.0 meter deep even in areas with permanent ground water near the surface. This was possible because the ground water was pumped down during the construction period.

The profiles were described according to a system similar to FAO »Guideline for Soil Profile Description«, but at some points readjusted to fit Danish conditions. The profile description comprises the following characteristics:

- Type and depth of horizons
- Soil colours (matrix and mottles)

- Borderlines between horizons (form and sharpness)
- The geology of the parent material

- Gley features

- Texture and organic matter

- Stoniness (shape, size, amount, sort, and state of weathering)

- Calcium carbonate (amount and sort e.g. nodules, powder
or shells)

- Structure and consistency

- Soil pores (diameter above 1 mm)
- Roots (amount and thickness)
- Nodules and coatings

- Cementations and pans (type and continuity) Furthermore general notations about date, name of surveyor, UTM-coordinates, slope, elevation, vegetation, and drainage class are made.

The profile descriptions were entered in standard schemes, and the quantification of the different parameters was given in numerical form to be stored in the ADKcomputer system. It is thus easy to search on single features and combine different parameters in the profile description, e.g. finding soils having horizons with platy structure and gley features.

In order to make the profile descriptions easily available for different users, programs have been elaborated transforming the numerical soil profile descriptions into words (Munk 1984).

Soil samples have been taken from all major horizons. In average six samples have been taken from each profile. For chemical and textural analyses the samples were first air-dried and then sieved. Only the fraction finer than 2 mm was used for the chemical analyses. For determination of soil water retention and root densities undisturbed samples were used. These were collected in tubes of different size. For determination of soil water retention 3 samples were collected from each horizon in tubes of volume 100 cm³. For determination of root densities one sample per 10-cm section was taken down through the profile using tubes 7cm high and diameter 11 cm.

ANALYSES

Table 3 shows the different analyses and the number carriedout during the investigation period. Texture, organic matter, calcium carbonate, and pH(CaC!₂) have been determinedon nearly all samples, while the other analyses have only been caried out on a minor part of the samples. The CEC was mainly determined on samples from profilessituated in forest because of heavy liming of the farmland.Soil water retention and root determinations have been carried out solely on samples from profiles situated on farmland, because these parameters will be of great value especially in irrigation planning. Pyrophosphatesolubleiron and aluminium were mainly determined on sandy samples, while KCI-soluble aluminium only were determined on acid subsoils. KCI-soluble aluminium was determined for the evaluation of Al-toxidity in the root zone. The electric conductivity was only determined on

samples from marine deposits, because all other deposits are leached for easily soluble salt due to heavy surplus of rain during the wintertime. The total nitrogen and phosphorouscontents were mainly determined on samples from or just below the topsoil, while citric-acid soluble phosphorous were determined exclusively on Al-material which fullfilled all other specifications to a mollic epipedon.

The analytical methods are described in detail in Madsen et al. (1985), but the following short comments should be made. Texture analyses were carried out by sieving of the sand fractions and the hydrometer method for determination of fine silt and clay. The following grain sizes were determined 2 n, 20 /x, 63 n, 125 n, 200 /*, 500 n, and 2000 ju. The content of organic matter was determined by an IR-Leco apparatus and the pH was determined potentiometrically in a suspension of soil and 0.01 M CaCl₂. Calcium carbonate content was determined by treating the sample with 4N HCI, catching the developed CO₂ in 0. IN Ba(OH)₂ followed by a titration with 0. IN HCI. The exchangeable bases were determined by AAS after treating the samples with IN NH4Ac, while the exchangeable acidity was determined by Pipers method. The CEC-value was determined as the sum of exchangeable bases and exchangeable acidity, except in lime-containing samples where the CEC-value was determined by treating the samples with IN NaAc, thereafter washing the samples with alcohol and exchange the sodium with IM NH4Ac. Dithionite-citrate and pyrophosphate soluble iron and aluminium were determined by AAS after treating the samples over night with a dithionite- citrate solution or with a pyrophosphate solution (Soil Conservation Service 1972). The KCI-soluble aluminium was determined by mixing the soil samples with IM KCI for one hour, filtration and titration with 0.02N NaOH. The total content of nitrogene was determined by the Kjeldahl method and the citric-acid soluble phosphorous content was determined spectrophotometrically after treating the samples for 18-20 hours with 1% citric-acid. The total phosphorous content was determined spetrophotometrically by ashing the samples and thereafter treat them with 12N H₂SO₄.

The soil conductivity was determined in the liquid
phase after treating 25 g sample with 50 ml pure water for
16 hours.

The soil water characteristic curve was determined by the pressure plate apparatus method, and the water content was determined at: pF 1.0, pF 2.0, pF 3.0 and pF 4.2. The root densities were determined as cm root/ccm soil. The roots were separated from the soil by washing and their lengths measured on a grid with mesh-wide 1.27 cm.

After finishing the analyses the remaining part of the
samples were stored in a soil sample library at ADK. This
library contains today more than 40,000 soil samples.

THE PRESENTATION AND USE OF DATA FROM THE PEDOLOGICAL INVESTIGATIONS

The database for soil profile descriptions is connected with the analytical databases, and programs have been developed to produce outputs giving the user a complete profile description in Danish and all corresponding analytical data. Fig. 4 shows an output for a single profile, a Pseudogleytypilessive developed in till.

In the profile description the exterior characteristics are mentioned first followed by a description of the interior characteristics for each horizon. In the first scheme a list of samples is given with sampling depth, sampling number, and ring numbers, the latter if samples for soil water retention have been collected. The second scheme shows the texture and the content of organic matter and calcium carbonate. The third scheme shows ppm dithionite-citrate and pyrophosphate soluble iron and aluminium. The fourth scheme shows the pH-values, exchangeable bases and acidity, CEC-values, base saturation, and electric conductivity, while the fifth scheme shows the nitrogene, phosphorous and carbon content. Root densities and soil water retention have not been determined.

The data from the profile investigations are today frequentlyused
in combination with existing computerized

soil- and landscape maps. Through transfer models the profile data are used for irrigation planning (Madsen et al. 1983), drainage-class assessments (Holst 1985), and calculationof soil erodibility (Madsen et al 1985).

Soil profile data combined with computerized landscape maps

The soil profile descriptions combined with the international laboratory methods make it possible to classify the profiles according to international systems like the FAO- Unesco system (FAO-Unesco 1974) and the American system »Soil Taxonomy« (Soil Survey Staff 1975). The classifications according to the FAO-Unesco system combined with existing soil maps give opportunities to readjust and progress existing soil maps covering Denmark, e.g. FAO soil map in scale 1:5000000 or EC-soil map in scale 1:1000000.

For classification according to FAO-Unesco and Soil
Taxonomy the following assumptions were made
- Danish soils have a mesic soil temperature regime and

a udic or aquic soil moisture regime

- The citric-acid soluble phosphorous content is less than
250 ppm. This excludes antropic epipedons

- If base saturation has not been determined, pH is used.
If pH (H2O) = 4,0 the base saturation is 0%, when pH
(H2O) = 8,3 the base saturation is 100%.

Fig. 3 shows the main gas pipeline system in relation to landforms in southern Jutland and Funen. The computerized landscape map (Hoist 1985) divides Denmark into 9 different landforms of which six are present in fig. 3. In table 4 and 5 the number of investigated profiles within the different landforms are given and their classification according to FAO-Unesco and Soil Taxonomy.

On outwash plains and in dune sand areas, the parent material is exclusively sandy. The well drained soils are mainly podzolized, nearly 90% of the investigated profiles in the dune sand area and 80% of the well drained profiles on the outwash plains were podzols or spodosols.

In the Saale glaciation landscape approximately % of the well drained profiles have developed in sandy parent material. Podzols dominate just as found in the outwash plains. Among the relatively well drained soils developed in clayey parent material soils with clay illuviation dominate. These old soils are strongly leached and exclusively classified as acrisols, frequently gleyey because of pseudogley in the subsoil. According to Soil Taxonomy the major part are ultisols, but few alfisols are present.

In the Weichsel glaciation landscape soils developed in clayey parent material dominate. These soils are more or less leached and most of them have clay illuviation. In Jutland acrisols dominate, while on Funen luvisols are most pronounced. On Zealand acrisols are nearly not present on farmland. According to Soil Taxonomy alfisols dominate, but ultisols frequently occur in Jutland. In sandy parent material within the Weichsel glaciation landscape podzols dominate in Jutland and arenosols on Funen. This shows that the leaching of the profiles decreases towards the east and thereby the frequency of acrisols and podzols, while the frequency of luvisols and arenosols

Resumé

I perioden 1981-84 er der udført pedologiske undersøgelser langs den udgravning, der er foretaget af Dansk Olie og Naturgas A/S (DONG) til etablering af hovedtransmissionsledninger for naturgas.I det omhandlede tidsrum er gasledninger etableret og de pedologiske undersøgelser foretaget på strækningen Frøslev- Egtved, Egtved-Storebælt, på tværs af Sjælland og Egtved-Vesterhavet.Arbejdet har omfattet detaljerede profilbeskrivelser med ledsagende prøveudtagning til analyse fra 835 profiler gravet i kanten af arbejdsbæltet eller i nærliggende skove, samt registreringaf profiludviklingen for hver 25 m i rørgraven over store strækninger. På næsten alle de udtagne prøver er der bestemt

textur, humus, pH og calciumcarbonat. På en del af prøverne er der tillige blevet bestemt total nitrogen, kationadsorptionskapacitet,indhold af ombyttelige kationer, pyrophosphat- og dithionitcitratopløseligtjern og aluminium, kaliumchloridekstraherbartaluminium, uorganisk og organisk bundet phosphor, citronsyreekstraherbartphosphor, vandretention samt rodintensiteter. Alle resultater findes på databaser ved Landbrugsministeriets Arealdatakontor, hvor de indgår i udbygningen af den danske jordklassificering.

Kortfattet kan man på baggrund af undersøgelsens resultater give følgende generelle beskrivelse af jordbundsforholdene langs traceen. De sandede jorde i Jylland er i overvejende grad podzolerede på nær tæt ved vestkysten, hvor helt unge jorde uden nævneværdig profiludvikling forekommer. En stor del af sandjordene på hedesletterne har gleypræg inden for de øverste 120 cm af profilen. På Fyn og Sjælland er sandjorde ikke så udbredte som i Jylland, og kun et fåtal af de fynske og sjællandske sandjorde er podzolerede. De lerede jorde er i overvejende grad præget af lernedslemning og pseudogleydannelse. De vestjyske lerjorde er foruden lernedslemning og pseudogleydannelse præget af periglaciale processer som jordflydning. Disse jorde vil ofte være stærkt udvaskede, og selvom de er blevet kalket, vil underjorden dog stadig ofte være stærkt sur. I Østjylland i det unge moræneområde er jordene almindeligvis stærkt udvaskede, og kalk findes kun yderst sjældent i de øverste 120 cm af profilen. Hyppigheden af stærkt ud vaskede morænelerjorde falder kraftigt over Fyn, og frit kalk begynder at optræde i profilen. På Sjælland optræder frit kalk hyppigt i profilen.

I relation til internationale klassifikationssystemer vil dette sige, at podzols (spodosols) dominerer i de sandede vestjyske jorde. Acrisols (ultisols) dominerer på lerede aflejringer i Syd- og Vestjylland, mens luvisols (alfisols) dominerer på Fyn og Sjælland.

ACKNOWLEDGEMENTS

Financial support was granted by the Danish Natural Science Research Council, the Danish Agricultural and Veterinary Research Council, and the Ministry of Agriculture. The authors are indebted to Mrs. Kirsten Winter, Geographical Institute, Copenhagen, for improving the English text.

References

FAO-Unesco (1974): Soil map of the world, vol. 1, legend.
Unesco Paris.

Hoist K. Å. (1985): The elaboration of drainage class maps for
agricultural planning in Denmark. Ministry of Agriculture, Bureau
of Land Data, Vejle.

Madsen, H. B. (1983): A pedological soil classification system for
Danish soils. Pedologie XXXIII, 2: 171-197.

Madsen, H. B. (1984): Soil mapping in Denmark. Soil Survey
and Land Evaluation 4,3: 57-62.

Madsen, H. 8., Hasholt, B. & Platou S. W. (1985): The development
of a computerized erodibility map covering Denmark.
EEC-Meeting, Cesena Italy, Oct. 1985.

Madsen, H. B. & Jensen, N. H. (1985): Jordprofilundersøgelsen
(In Danish with English summary). Ministry of Agriculture,
Bureau of Land Data, Vejle.

Madsen, H. B. & Platou S. W. (1983): Land use planning in
Denmark. Nordic Hydrology 14,5: 267-276.

Mathiesen, F. D. (1980): Soil classification in Denmark. Its results
and applicability. EEC report on land resource evaluation.
EUR 6875.

Mathiesen, F. D. (1984): Soil classification in Denmark and its adjustment in relation to land use planning. In J. C. F. M. Haans, G. G. L. Steur & G. Heide (ed.): Progress in land evaluation. A. A. Balkema, Rotterdam.

Munk, I (1985): Program systems for handling pedological soil
data. Ministry of Agriculture, Bureau of Land Data, Vejle.

Soil Conservation Service (1972): Soil survey laboratory methods
and procedures for collecting soil samples. U. S. Dep. Agric.
SSIR 1. U.S. Govt. Print. Office, Washington, D. C.

Soil Survey Stajf(\97s): Soil Taxonomy. U. S. Dep. Agric., U. S.
Govt. Print. Office, Washington, D. C.