SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow An erratum has been published
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sommer, M.
Right arrow Articles by Stahr, K.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Sommer, M.
Right arrow Articles by Stahr, K.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Sommer, M.
Right arrow Articles by Stahr, K.
Soil Science Society of America Journal 64:1434-1442 (2000)
© 2000 Soil Science Society of America

DIVISION S-5-PEDOLOGY

Lateral Podzolization in a Granite Landscape

M. Sommera, D. Halmb, U. Wellera, M. Zareib and K. Stahrb

a Inst. für Biomathematik und Biometrie, GSF-Forschungszentrum für Umwelt and Gesundheit, Ingolstädter Landstr. 1, 85764 Neuherberg, Germany
b Inst. für Bodenk. und Standortslehre, Univ. Hohenheim, Emil-Wolfe Str. 27, 70593 Stuttgart, Germany

sommer{at}gsf.de


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and methods
 Relief, Soil Mapping, and...
 Results
 Discussion
 Conclusions
 Appendix
 REFERENCES
 
Analysis of soil pattern, especially the pattern of depletion and accumulation zones, is a powerful tool to decipher pedogenic processes at the landscape scale. To clarify the pedogenesis of a distinct pattern of podzolized soils in the cool, perhumid Black Forest (Germany) we performed a study in the forested upper part of a granitic catchment (Bärhalde). From detailed soil mapping we selected a typical catena of four pedons, which were analyzed for physical (bulk densities and particle-size distribution), chemical (pH, organic C, pyrophosphate, oxalate, dithionite, and total Al, Mn, Fe), and mineralogical (clay minerals) properties. Standard mass balance calculations were modified to include a two-component system with regard to parent material. Results showed a shift from two-mica granite to granite–porphyry downslope. Soil pattern revealed podzolized soils with thick E horizons and thin spodic horizons developed in the upslope areas, whereas in downslope soils the reverse was found (thick spodic B and thin E horizons). Soil chemical and mineralogical properties were linked to soil morphology in that contents of organic C, pedogenic oxides, hydroxy-interlayered vermiculites (HIV), and pH increased from eluvial to illuvial horizons as well as from up- to downslope soils. Mass balances of Fe and Al showed negative fluxes in upslope soils and positive fluxes in downslope soils during pedogenesis. We concluded from these results that a catenary eluvial–illuvial sequence (lateral podzolization) develops as a result of upslope mobilization followed by a (partially) lateral transport and subsequent immobilization in downslope zones, probably because the base-richer parent material built up a geochemical barrier.

Abbreviations: d [subscript], dithionite-extractable • HIV, hydroxy-interlayered vermiculite • o [subscript], oxalate-soluble • p [subscript], pyrophosphate-soluble • XRD, x-ray diffraction • XRF, x-ray fluorescence


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and methods
 Relief, Soil Mapping, and...
 Results
 Discussion
 Conclusions
 Appendix
 REFERENCES
 
PODZOLIZATION

is one of the major soil forming processes on earth. Spodosols, the soil unit showing the most intense podzolization characteristics, cover {approx}5 x 106 km2 of the earth's surface (Eswaran et al., 1993) and are the dominant soils of the boreal zone in the northern hemisphere. Up to now podzolization has been viewed mainly as a vertical translocation of Al, Fe, and Corg at the pedon scale (McKeague et al., 1983; Buol et al., 1989; Ross, 1989), that is, as an intrapedon translocation. This process creates bleached, depleted eluvial and bright-colored, Al-, Fe-, Corg-enriched illuvial horizons. At a landscape scale podzolization was mostly looked at as a juxtaposition of different podzolized pedons. Only a few authors discussed the pattern of podzolization as a result of lateral translocation processes (Schlichting, 1963; Glazovskaya, 1968; Lucas and Chauvel, 1992; Sommer et al., 1997).

Soil pattern comprises much of the information about the pedogenic processes at the landscape scale (Fridland, 1976; Hole and Campbell, 1985). From the existence and spatial arrangement of depletion and accumulation zones, conceptual process models about soil forming processes on a landscape scale can be deduced (Sommer and Schlichting, 1997). Morphological indicators of podzolization are very helpful because it is widely accepted that E horizons have undergone element losses, whereas spodic B horizons have undergone gains of translocated matter. Although two-dimensional, catenas have been widely used and accepted as a powerful tool for soil landscape analysis (reviewed in Sommer and Schlichting, 1997). The first papers on catenas inferred soil forming or destructing processes from pattern (Milne, 1936, 1947; Greene, 1947).

The objective of this study was to decipher podzolization processes in a landscape from pattern, with special emphasis on genetic linkages of soils by lateral translocations. Soils in a granite catena in a cool, perhumid climate were studied with respect to the patterns of soil morphological, physical, chemical, and mineralogical properties. Mass balance calculations of a two-component system also were developed.


   Materials and methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
Relief, Soil Mapping, and...
 Results
 Discussion
 Conclusions
 Appendix
 REFERENCES
 
Environmental Setting of the Area
The study was conducted in the cirque-like catchment Bärhalde, which is located in the southern part of the Black Forest (980–1300 m above sea level). This cirque is composed mainly of two-mica granites with intrusions of granite–porphyry. Mean values of the mineralogical composition of the two-mica granite were 34% quartz, 33% orthoclase, 25% plagioclase (An 2–10), and 4% of biotite as well as muscovite (Keilen, 1978). The granite–porphyry showed a mineralogical composition of quartz, orthoclase, and albite. Only biotite, no muscovite, was detected. Calcium, Mg, K, Al, Ti, and Fe contents were much higher in the granite–porphyry than in the granite (Keilen, 1978; Wimmenauer, 1985). With the exception of a few rock outcrops, soil development occurred in physically preweathered (cryoclastic) and mixed (solifluction) periglacial debris. Nevertheless all soils are of Holocene age. The time of soil development is assumed to be {approx}10000 yr. The mean annual air temperature is 4.5°C, and the mean annual precipitation is 1950 mm. Hydrological investigations showed a mean annual discharge of 1500 mm (Heyn, 1989); that is, only 25% of the precipitation is leaving the system through evapotranspiration. Meesenburg (1997) quantified the amount of soilborne water (lateral subsurface flow) at the weir by hydrochemical discharge separation for 1 yr and found two-thirds of the discharge was soil water and one-third came from the groundwater. Consequently, lateral subsurface flow (interflow) was a major component in the hydrological flow system. The vegetation is dominated by Norway spruce forests [Picea abies (L.) Karsten].


   Relief, Soil Mapping, and Classification
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Relief, Soil Mapping, and...
Results
 Discussion
 Conclusions
 Appendix
 REFERENCES
 
From a reconnaissance soil survey covering the total watershed (Segatz and Thees, 1979), a northeast-facing slope was chosen for detailed soil mapping. This landscape was divided into three segments for data analysis and discussion: an upslope segment extending from the watershed divide to {approx}100 m downslope, a midslope segment extending from {approx}100 to 150 m downslope, and a downslope segment extending from {approx}150 to 250 m downslope. The longitudinal slope profile of the upslope segment was mainly linear, with some curvature, and averaged 45% in gradient. The cross-slope profile of the upslope segment was mainly concave, which resulted in convergent water and element flux directions (Hall and Olson, 1991). The longitudinal slope profiles of the midslope and downslope segments had more concave and convex curvature than the upslope segment and averaged 30% in gradient. The cross-slope profiles of the midslope and downslope segments were mainly linear, although distinct microrelief was apparent.

Soils were augered down to 1-m depth in three parallel transects running from the watershed divide to a spring at the base of the downslope segment. The middle transect runs through the deepest points of the concave cross-slope curvature of the upslope segment. Distances between augering points and transects were {approx}25 m. This sampling plan resulted in a total of 33 observations in the three landscape segments. In addition to these auger borings, several soil pits were excavated down to the parent material (0.8–1.5 m depth) and used as mapping calibration points. For detailed soil analysis three representative pedons were described and sampled in the middle transect of the slope (G-1, G-3, G-4 in Fig. 1a) . The three profiles were arranged in a line with one further pedon (G-2), which was previously analyzed by Keilen (1978) and Heyn (1989).



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 1 (a) Topographic position of the selected pedons and landscape segments and (b) illuvial/eluvial ratio of podzolization (IERpodzol). Boxplots with 5, 25, 50, 75, 95 percentiles and mean values (filled square); numbers in brackets = numbers of augering points in the landscape segments

 
According to U.S. soil taxonomy (Soil Survey Staff, 1998) G-1 and G-2 must be classified as loamy-skeletal, mixed, acid, frigid Typic Udorthents; G-3 and G-4 are classified as loamy-skeletal, mixed, frigid Typic Haplorthods.

Analytical Methods
Rock fragments (>2 mm) were estimated in the pits on a volume percentage basis. Bulk density was either measured by using 100-cm3 cores (oven dry, 105°C) or estimated for horizons high in coarse fragments. The estimation was based on a combination of soil texture, soil structure, and penetration resistance (with a knife) determined in the field. Blocks of Cr horizons were sampled with a pick and crushed first with a hammer and then with a mill. Soil samples were taken from each horizon (excluding fragments >20 mm), air dried, carefully crushed, and passed through a 2-mm mesh. The contents of the 2- to 20-mm fraction were determined by weighing. Standard soil analyses were performed on the <2-mm fraction of the soils and the ground Cr horizons. Equivalents of the fine earth (<2 mm) and of Cr material were oven dried at 105°C to recalculate all results on an oven-dry basis (except pH). Pretreatment for particle-size analysis included destruction of humus (10% H2O2, v/v), dispersion with sodium pyrophosphate, and 15 min of ultrasonic treatment. With a combined wet sieving (>=20 µm) and pipette method (<20 µm) the sand, silt, and clay percentages of the humus-free fine earth were determined (Schlichting et al., 1995: 5.4.1.3). The pH of the soils was determined in 0.01 M CaCl2 (1:2.5 soil/water) and measured electrometrically. Organic C was measured conductometrically after dry combustion (Schlichting et al., 1995: 5.6.1.4). Dithionite-extractable Fe (Fed) and Mn (Mnd) were analyzed after Mehra and Jackson (Schlichting et al., 1995: 5.5.5.3), oxalate-soluble Fe (Feo) and Al (Alo) according to Tamm (Schlichting et al., 1995: 5.5.5.2). Pyrophosphate-soluble Fe (Fep) and Al (Alp) were determined in selected profiles at pH 10 and subsequent high-speed centrifugation (Schlichting et al., 1995: 5.6.6.4). Iron, Mn, and Al concentrations of all extracts were measured by atomic absorption spectroscopy (Model 3100, Perkin Elmer, Norwalk, CT). Total elemental analysis (Al, Fe, Mn) was performed with x-ray fluorescence (XRF) using a Siemens x-ray spectrometer (Model SRS 200, Cr-k{alpha} radiation). The clay fraction (<2 µm) was separated from the silt and sand fraction by wet sieving and sedimentation. The clay preparation included K and Mg saturation, heat treatment of the K-saturated probes to 400°C, and glycerol solvation of the Mg-saturated samples. X-ray diffraction (XRD) analysis was performed on parallel oriented specimens using Cu-k{alpha} radiation (Siemens D 500-Diffractometer, Cherry Hill, NJ). Percentages of the clay minerals (5% intervals) were calculated with the software package Diffrac AT 3.3 from Siemens Co. Because percentages were calculated by integration of peak areas without internal standards, clay mineral assemblages have to be interpreted as semiquantitative data.

Calculations
Morphological Index
A dimensionless illuviation/eluviation ratio of podzolization (IERpodzol) was calculated by dividing the sum of the illuvial horizon thicknesses by the sum of the eluvial horizon thicknesses in podzolized soils:

(1)
where IERpodzol is the illuviation/eluviation ratio of podzolization; Ii is the thickness (cm) of the illuvial horizon i, including transitional horizons such as Bhs, BsC, and CBs; Ej is the thickness (cm) of the eluvial horizon j, including transitional horizons such as AE and EA; m is the number of illuvial horizons; and n is the number of eluvial horizons.

Mass Balances
In order to calculate mass balances of elements one has to know the initial composition of the parent material (Schlichting et al., 1995). Mass balance calculations often assume that one bedrock is the parent material (Spodosols: Brimhall and Dietrich, 1987; Jersak et al., 1995). In our case we faced a mixture of two rock components, namely granite and granite–porphyry. These rocks differ in the content of stable Ti and Zr (Keilen, 1978; Wimmenauer, 1985), which was confirmed by Cr horizons probed in G-1 and G-4 (see Table 1) . We assumed both elements to be inert under the prevailing weathering conditions, which was reasonable for the relatively short duration of soil development (Milnes and Fitzpatrick, 1989). At least on a pedon scale no gains and losses should occur; that is, the actual Ti and Zr masses (g m-2 profile depth-1) equal those of the initial rock mass to be weathered. On the basis of these assumptions, we used Ti and Zr as index elements and calculated the following set of equations (all on pedon scale):

(2)

(3)

(4)
where a is the relative portion of granite–porphyry; b is the relative portion of granite; c is the weathering coefficient of the pedon (c > 0: mass gains; c < 0: mass losses); Zrgp and Zrgr are the Zr contents of the granite–porphyry and granite; Tigp and Tigr are the Ti contents of the granite–porphyry and granite; and Zrfe and Tife are the mean Zr and Ti contents of the pedon's fine earth (Zr, Ti mass divided by fine earth mass).


View this table:
[in this window]
[in a new window]
  		Table 1 Selected morphological, physical, and chemical properties of the pedons

 
Solving the three linear equations with three unknowns resulted in (see Appendix for details):

(5)
and

(6)

Inserting the Ti and Zr contents of the two Cr horizons (Table 1) gave and .

With the actual pedon masses of Ti and Zr we calculated the relative portions of the two components of the parent material, namely the granite–porphyry (a) and the granite (b), the latter of which equals 1 - a in the two-component system. Then initial element contents of the parent material (xp) could be easily computed by

(7)

The masses of parent material (Mt0,) that resulted in the actual fine earth mass through pedogenic processes were calculated by

(8)
where Mt0 is the initial mass of parent material to be weathered to the actual fine earth mass (kg m-2); Zrpedon is the Zr mass of the pedon's fine earth (g m-2 pedon depth-1); Zrp is the Zr content of the parent material, calculated by Eq. [7] (mg kg-1 fine earth); and Mt1 is the actual pedon fine earth mass (kg m-2).

Total fluxes j of element x during soil development (pedon scale) were calculated by the difference between the actual mass of element x, M(x)t1, and the initial mass of element x, M(x)t0, according to

(9)

(10)
where j(x)t1-t0 is the flux of element x between t1 and t0 (kg m-2 pedon depth-1); t1 is the time of probing; t0 is the time of (assumed) beginning of soil development (10000 yr before present); Mi is the fine earth mass of horizon i (kg m-2); xi is the content of element x in horizon i at time t (g kg-1 fine earth); n is the number of horizons to depth of C horizon at time t1; Mt0 is the initial mass of parent material (kg m-2); xp is the initial content of element x in the parent material from Eq. [7] (g kg-1 fine earth).

Pedon fine earth masses were calculated without organic surface layers, because of (i) missing total element contents (XRF not suitable for horizons rich in organic matter), and (ii) neglectible masses (except Corg), because of low bulk densities.


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Relief, Soil Mapping, and...
 Results
Discussion
 Conclusions
 Appendix
 REFERENCES
 
Soil Pattern
The upslope landscape segment was dominated by strongly podzolized soils with thick E horizons (up to 100 cm) and thin (10 cm) or even missing spodic horizons (E Spodosols). Further downslope the E horizons were thinner and the thicknesses of the spodic horizon increased. Soils in the downslope segment had the thinnest E and thickest Bhs horizons (Bs Spodosol). The IERpodzol values ranged from the maximum of 13 (9 in G-4, Table 2) in downslope soils to a minimum of 0 (no Bs, but E) in upslope soils. Box plots of the IERpodzol values for all podzolized soils in each of three slope segments revealed a clear increase in the ratio from up- to downslope areas, despite the spatial variability in each unit (Fig. 1b).


View this table:
[in this window]
[in a new window]
  		Table 2 Chemical properties (<2 mm) and clay minerals (<2 µm) of the pedons

 
Physical, Chemical, and Mineralogical Properties
Pedons were characterized by a significant portion of rock fragments (Table 1), which were dominated by cobbles and gravels. No catenary trend could be detected. Bulk densities were low throughout the catena even in the subsoils. Fine earth textures were uniformly sandy loams, both with depth and throughout the landscape (Table 1). There was no evidence for clay translocation, as depth functions of clay percentages were fairly uniform. Medium/coarse sand ratios between 0.4 and 0.6 indicated a low degree of intrapedon heterogeneity of the parent material. Depth functions of Ti/Zr ratios indicated small intrapedon heterogeneity as well, but in the catena the ratios increased from G-1 to G-4. Thus, we concluded a homogenous mixture of granite and granite–porphyry at the pedon scale, but different relative portions of the two rocks at the landscape scale.

Soils were very acid throughout, with pH values (CaCl2) <= 4.3 (Table 2). The pH increased with depth in most pedons. Both organic surface layers and eluvial mineral horizons had pH values below 3.5, whereas the illuvial horizons and the C horizons were characterized by pH values above 3.5. A downslope increase in pH could be observed as well, because less acidic illuvial horizons were nearer to the soil's surface and made up greater proportions of pedon masses. In general, acidity corresponded with the morphology of the horizons and their catenary gradient.

Vertical translocation of organic C could be inferred from the depth functions of most pedons (Table 2). Although the differences were small compared to the overall Corg contents, there was an increase from any AE or E to the next illuvial Bh with the exception of G-4. Due to the increase in thickness of humus-richer, illuvial horizons downslope, the organic C of the pedons increased in lateral direction; Corg mass densities in the mineral soil material increased along the catena as well (G-1–G-4: 3.3, 12.5, 15.7, 16.1 kg Corg m-2 pedon depth-1).

Soils showed depth functions of pedogenic oxides typical of the podzolization process: eluvial, Fe- and Al-depleted horizons overlying illuvial, and Fe- and Al-enriched horizons (Table 2), although the intensity of element redistribution was different among catena members. Lowest contents of Feo, Fed, and Alo were found in upslope soils. Downslope pedons were higher in pedogenic oxides, and the increase from E to Bs developed more gradually. The vertical and lateral trends in Mnd (54% of the total Mn by mean) were similar to the distributions of pedogenic Fe and Al with depth and slope. All three elements accumulated in the organic surface layers, probably due to the element cycling of the vegetation (bioaccumulation). Pyrophosphate extracted 65% of Feo and 86% of Alo (mean values of all horizons). Therefore we concluded that most of the amorphous Fe and Al was organically bound. Excluding humus-rich topsoils (Corg > 4%) the mean active Fe ratio (Feo/Fed) increased from eluvial (0.17) to spodic horizons (0.53), as has been observed for Spodosol horizons from other studies (Blume and Schwertmann, 1969). On a pedon basis the ratio increased in downslope direction as well (0.05 -> 0.46).

In Fig. 2 typical XRD patterns of the clay fraction (<2 µm) are shown for an eluvial, an illuvial, and a C horizon of the catena. Besides the main components illite (1.00 nm) and kaolinite (0.72 nm), the C horizon was characterized by a peak at 1.43 nm, which remained stable under treatment by glycerol and K. The peak disappeared only through heating (400°C); therefore, we concluded aluminous chlorite (i.e., strongly interlayered HIV) to be the relevant clay mineral (Barnhisel and Bertsch, 1989). The XRD patterns of the eluvial and illuvial horizon resembled each other in that both horizons showed clear 1.00-nm (illite) and 0.7-nm (kaolinite) peaks. In the E horizon (Fig. 2a) no peak was developed at 1.4 nm, instead a bulge occurred between 1.0 and 1.4 nm. The bulge did not expand through glycerol solvation, but disappeared in the K-saturated sample. This pattern was interpreted as illite weathering, namely as an intergrade between illite and vermiculite. In the spodic horizon (Fig. 2b) a 1.4-nm peak can be detected, which did not expand in the gycerol-solvated sample and did not disappear completely in the K-saturated probe, but did disappear at 400°C. This pattern is typical for HIV (Barnhisel and Bertsch, 1989). Hydroxy-interlayered vermiculate is a common clay mineral in Bs horizons, as has been shown elsewhere (Harris et al., 1987; Dahlgren et al., 1989; Bain et al., 1990).



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2 X-ray diffraction patterns of the clay fraction (<2 mm) in three typical horizons: (a) E1 in G=1, (b) Bsh in G=4, and (c) 2 Cr in G=4

 
Illite, illite-vermiculite intergrades, vermiculite, and kaolinite occurred in similar amounts in the clay fractions of all eluvial soil horizons in the landscape (Table 2). Aluminous chlorites were found only in one C horizon. We believe they were not detected in other soil horizons because of the masking effect of organic matter. Hydroxy-interlayered-vermiculite was detected only in spodic horizons. This is thought to be due to a mobilization of Al-hydroxo-complexes in the eluvial horizons (dechloritization) and subsequent translocation into the illuvial horizons, which means a continuous supply of Al-hydroxo-complexes for clay mineral transformations in subsoil horizons. No catenary trend in the clay minerals was evident, except the downslope increase in HIV, which corresponded with the increase in thickness of the illuvial horizons.

Mass Balance Calculations
Different portions of bedrocks were incorporated into the parent material by periglacial mixing: granite–porphyry increased from 16% in the uppermost to 81% in the lowermost pedon (a in Table 3) . The results of mass balance calculations for the relevant elements in podzolization processes matched the morphological and chemical data (weight basis). Fluxes of Al and Fe were negative for the upslope area, which indicates Fe and Al losses during soil development (Table 3). Upslope soils lost {approx}40% of their Fe and 25% of the initial mass of Al (Fig. 3) . In contrast, pedons of the lower part of the catena showed positive fluxes (i.e., gains in Fe and Al). Relative gains of Al were higher than those of Fe. Gains of Al in G-3 resembled those in G-4, whereas Fe accumulation was highest in the lowermost pedon. On a molar basis the Al budget was far more relevant for total soil mass changes than Fe. Aluminum fluxes were one magnitude higher than the Fe fluxes. Beside the pH-dependent precipitation of Fe– and Al–organic complexes, a part of the Al will be retained by interlayering into vermiculite to build HIVs. All soils were depleted in Mn. One-half of the total Mn was already leached from the pedons (Fig. 3). Probably there was an influx of Mn with lateral flowing water in G-3 or G-4 (see Mn losses of upslope soils), but here no precipitation occurred. This was probably due to the high mobility of Mn under the very acid conditions (Gilkes and McKenzie, 1988; Schlichting and Sparrow, 1988).


View this table:
[in this window]
[in a new window]
  		Table 3 Mass balance calculations (fine earth, without organic surface layers).{dagger}

 


View larger version (36K):
[in this window]
[in a new window]
  		Fig. 3 Results of mass balance calculations plotted as relative gains and losses of Fe, Al, and Mn (in percentage of initial mass)

 

   Discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Relief, Soil Mapping, and...
 Results
 Discussion
Conclusions
 Appendix
 REFERENCES
 
In the study area a regular pattern of podzolization was obvious—E horizons dominated pedons in the upslope landscape segment and Bs horizons made up almost the entire pedon of soils in the downslope landscape segment. In general the chemical analysis as well as mass balance calculations matched the morphological findings in the catena. Eluvial zones were more acidic, lower in Corg, and showed a higher degree of leaching of Fe and Al. In contrast, the (morphological) illuviated zones in downslope soils corresponded with higher Corg masses as well as with gains of Fe and Al. Because the thick accumulation zones in downslope pedons cannot be explained by vertical translocation processes (masses of the actual eluvial horizons are too low), two pathways of soil genesis must be taken into account for the patterns observed, namely erosional and solutional translocation processes.

Erosional processes may have led to a removal of (former thicker) eluvial horizons in downslope catenary soils. However, our experience suggests that erosion can be neglected in those catchments for several reasons: Forests prevailed for long times in the study area (no change in land use) and the rates of erosion under forests were less than the rates of soil development. Steeper slopes characterize the zones of thick and strongly depleted E horizons, whereas more gentle slopes dominate the illuviation areas. If erosion was an important process in those landscapes the steeper areas of E Spodosols should be subject to highest rates of soil removal, because of the higher relief energy. This was not the case, as E horizon thickness revealed. In addition, no correlated sediments (very acid, Fe-impoverished) have been found in the potential depositional area. Experience from actual catchment studies did not show any substantial sediment load in the drainage systems, even during storm events (Meesenburg, 1997). Moreover, patterns of erosional processes cannot explain the thin or even missing spodic horizons in the upslope areas.

Two explanations are left, namely vertical upward or lateral downslope translocation processes of soil solutions. Because groundwater occurred fairly deep in the bedrocks, there was no evidence of a vertical transport of solutes via capillary rise. Consequently, lateral translocation processes must be taken into account. Conditions were favorable for lateral solute transport, because of the very dense subsoil/soil–rock interface. Lateral outflow at this boundary could be observed regularly in soil pits under wet conditions. Rainfall was very high, and the pedons of the upslope areas were located in convergent relief positions, which means additional flowthrough water of high reactivity. This is supported by a high portion of interflow water on the catchment scale, as hydrological studies have shown (Heyn, 1989; Meesenburg, 1997). Unfortunately, no direct measurements of interflow water were available on a pedon scale. But Heyn (1989) measured soil water chemistry and element fluxes with suction ceramic plates for 2 yr in G-2 (Zweiseenblick) as well as in a nearby Spodosol (200 m outside the transects) and the catchment outlet (Table 4) . From these results at least the actual mobility of both Al and Fe in the eluvial zone could be concluded, although the directions of the actual fluxes (vertically, laterally) were not determined by Heyn (1989).


View this table:
[in this window]
[in a new window]
  		Table 4 Water chemistry

 
Under the assumption of similar hydrological conditions during the Holocene we calculated the overall mean concentrations of elements in soil water during the time of soil development. Dividing the fluxes achieved through mass balance calculations (Table 3) by the total amount of seepage water throughout the Holocene (1500 L m-2 yr-1 x 10000 yr = 1.5 x 107 L m-2) resulted in very similar element concentrations compared with the actual concentrations (Table 4). The actual process intensities in G-2 resembled the mean intensities during pedogenesis at least in the case of Si, Fe, and Mn. Although soil development was a nonlinear process, at least the eluvial soil seemed to be in equilibrium with actual solutional processes. Aluminum concentrations doubled in recent soil solution compared with the overall concentrations, which might be due to higher pHs at the beginning of soil development (soils far away from the Al buffering range) and consequently lower Al concentrations at that time. Aluminum and Fe concentrations of spring water (Fig. 1a, Table 4) were lower than in the eluvial soil horizons, which supported the findings of Fe and Al immobilization in downslope soils. Although mobile in the upslope landscape segments, only a part of Fe and Al reached the drainage system. This is because of the pedochemical barrier of Bs Spodosols in downslope landscape segments. Higher Fe and Mn values in the catchment outflow ({approx}600 m further downslope from G-4) probably resulted from an element mobilization in hydromorphic soils of the lowermost part of the catchment.

From the patterns observed we developed our concept of lateral podzolization (Fig. 4) . By analogy to podzolization on the pedon scale (Spodosol, left), lateral podzolization (tilted Spodosol, right, idealized) is characterized by (i) an upslope mobilization of Fe and Al, probably by organic acids; (ii) vertical or lateral leaching in the upper parts of the catena; and (iii) a downslope (partly) immobilization of the soil compounds. This hypothesis of a catenary process sequence was supported by its resulting properties, namely (i) morphology (E -> Bs Spodosols), (ii) mass balances (upslope negative fluxes of Fe, Al, downslope positive fluxes), (iii) patterns of Corg accumulation (increase downslope), (iv) clay mineralogy (catenary increase in hydroxy-interlayered vermiculite), and in part by (v) the actual concentrations of the soil solution (recent mobility of Al, Fe in eluvial zones). The controlling factor for the patterns observed was the pattern of the parent material. There was a catenary trend in that admixtures of granite–porphyry become more important in the downslope direction, which means an increasing admixture of a base-richer rock. Consequently, buffering capacities were higher and an immobilization of Fe– and Al–organic complexes will take place, probably by higher pH values (DeConinck, 1980; Buurman and van Reeuwijk, 1984; Farmer, 1984).



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 4 The concepts of vertical (left) and lateral podzolization (right)

 

   Conclusions
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Relief, Soil Mapping, and...
 Results
 Discussion
 Conclusions
Appendix
 REFERENCES
 
We regarded the morphological, mineralogical, and chemical patterns of the catena as an eluviation–illuviation sequence on a landscape scale, analogous to the eluvial–illuvial sequence on the pedon scale. Principally, podzolization processes may also prograde laterally up- (Chauvel et al., 1987; Bravard and Righi, 1988, 1989) or downslope (Lucas and Chauvel, 1992; Boulet et al., 1997) as a result of clay destruction or removal. But in the case of lateral podzolization sensu stricto up- and downslope soils were pedogenetically connected by interpedon translocations (Schlichting and Schweikle, 1980), as eluvial and illuvial horizons were connected by intrapedon-translocations on the pedon scale. According to the systematics of Sommer and Schlichting (1997), interpedon translocations indicate translocation catenas. Further studies on soil processes at the landscape scale should emphasize the aspects of depletion and enrichment zones as well as possible flowpaths of solutes.Schoenberger Wysocki Benham Broderson 1998


   ACKNOWLEDGMENTS
 
The authors would like to thank Herb Huddleston for very helpful comments and discussions. In addition we are grateful to three anonymous reviewers.

Received for publication January 8, 1999.
   Appendix
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Relief, Soil Mapping, and...
 Results
 Discussion
 Conclusions
 Appendix
REFERENCES
 
Derivation of the equations for the determination of the coefficients a, b, c:

(A1)

(A2)

(A3)
where a is the relative portion of granite–porphyry; b is the relative portion of granite; c is the weathering coefficient of the pedon (c = Mt1Mt0-1; c > 0: mass gains, c < 0: mass losses); Zrgp and Zrgr are the Zr contents of the granite–porphyry and granite; Tigp and Tigr are the Ti contents of the granite–porphyry and granite; and Zrfe and Tife are the mean Zr and Ti contents of the pedon's fine earth (Zr, Ti mass divided by fine earth mass).

Resolving Eq. [3] for b and inserting b into Eq. [1] and [2] results in:

(A4)

(A5)
which is equivalent to

(A6)

(A7)

Multiplying Eq. [6] with (Tigp - Tigr) and Eq. [7] with (Zrgp - Zrgr) gives

(A8)


(A9)

Substracting Eq. [9] from Eq. [8] results in

(A10)
which equals

(A11)

Resolving Eq. [4] for a results in:

(A12)


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Relief, Soil Mapping, and...
 Results
 Discussion
 Conclusions
 Appendix
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
A. S. Hartshorn, R. J. Southard, and C. S. Bledsoe
Structure and Function of Peatland-Forest Ecotones in Southeastern Alaska
Soil Sci. Soc. Am. J., September 1, 2003; 67(5): 1572 - 1581.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow An erratum has been published
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sommer, M.
Right arrow Articles by Stahr, K.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Sommer, M.
Right arrow Articles by Stahr, K.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Sommer, M.
Right arrow Articles by Stahr, K.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome