Naturwissenschaften (2002) 89:43–46
DOI 10.1007/s00114-001-0281-z
S H O R T C O M M U N I C AT I O N
W. Schimmack · K. Auerswald · K. Bunzl
Estimation of soil erosion and deposition rates
at an agricultural site in Bavaria, Germany, as derived
from fallout radiocesium and plutonium as tracers
Received: 12 January 2001 / Accepted: 15 October 2001 / Published online: 16 November 2001
© Springer-Verlag 2001
Abstract In the near future, the use of 137Cs from global
fallout (Cs) as a tracer for erosion studies will no longer
be possible in areas with a substantial deposition of
Chernobyl-derived 137Cs. Therefore, we have used
239+240Pu from global fallout (Pu) as a tracer as well as
137Cs in order to determine long-term soil redistribution
rates for an agricultural field (inclination about 20%, area
approx. 3 ha) in Scheyern, Bavaria. The mean erosion
and deposition rates derived from Cs were –37 and
+52 t·ha–1·year–1, respectively; those from Pu were –32
and +39 t·ha–1·year–1. We found no statistical difference
between the means obtained by the two tracers. In contrast to Pu, however, the rates obtained by Cs were not
accurate enough to assure the presence of a net soil redistribution. Modeling of soil translocation in the field
by water and tillage erosion resulted in estimates which
were in reasonable agreement with the rates derived
from Pu.
Introduction
Many erosion studies have been performed worldwide
using 137Cs from global fallout (Cs, main period:
1953–1964) as a tracer of soil movement (see Pimentel
et al. 1995). In Germany, however, no erosion study
using the radiotracer method has been published to date,
although erosion is regarded as a major threat to soils
(Auerswald and Schmidt 1986; Frede and Dabbert
1998). In a few years’ time, it will not be possible to use
137Cs as a retrospective tracer in regions like southern
W. Schimmack (✉) · K. Bunzl
GSF-National Research Center of Environment and Health,
Institute for Radiation Protection, PO Box 1129,
85758 Neuherberg, Germany
e-mail: [email protected]
Tel.: +49-89-31872370, Fax: +49-89-31873323
K. Auerswald
Department of Plant Science,
Technical University München/Weihenstephan,
85350 Freising, Germany
Bavaria with a substantial deposition of Chernobylderived 137Cs, both because the two sources of 137Cs can
no longer be distinguished and because total 137Cs is not
suitable for replacing Cs (for details see Schimmack et
al. 2001). Therefore, we have recently started to investigate the potential of 239+240Pu from global fallout (Pu) as
a tracer of soil movement in a steep field (inclination
20%) near Scheyern in Bavaria. Plutonium was also deposited by the Chernobyl fallout, but in Germany the
amount of this fraction is negligible (<1% of the total
inventory, Hötzl et al. 1987). In an earlier paper
(Schimmack et al. 2001), we compared the activity patterns of Cs and Pu. In the present paper, erosion and deposition rates are calculated from these inventories and
compared with estimates of model calculations including
tillage erosion. There have been only very few attempts
to use Pu as an erosion tracer (for a review see Foster
and Hakonson 1987), and in these studies no soil redistribution rates were reported. Thus, to our knowledge,
the rates given in this study are the first ones derived
from plutonium as a tracer.
Materials and methods
Site and soil
The field investigated (ca. 3 ha) is characterized by steep slopes,
approx. 20%, and is located within the Scheyern farm (11°27′ E,
48°30′ N), about 40 km north of Munich, Germany, in a hilly
landscape. The main soil type is Dystric Eutrochrept (US soil taxonomy). The field was ploughed for the first time in 1970 and for
the last time in 1993. Conventional soil mapping (Fig. 1) was carried out by visual inspection and descriptions of soil cores obtained by hand-augering along several transects from the top to the
valley floor. In 1994, after the transect mapping, 48 points within
the field and 12 points at a reference site were sampled at the
nodes of a rectangular grid, grid edge length 25 m. The lowermost
part of the slope (see Fig. 2) could not be sampled, because the
field was bounded there by a road. At each grid node, an undisturbed soil core, 1.2 m in length and 0.1 m in diameter, was taken
and subdivided into several depth increments. The soil was airdried and fractionated into stones (>2 mm) and fines (<2 mm).
The stones were washed in order to remove their fines (for details
see Schimmack et al. 2001).
44
Determination of radionuclides
The amounts of 134Cs and 137Cs in the soil samples were determined by direct gamma-spectrometry using a high-purity germanium detector and a multichannel analyzer. 137Cs from global fallout
(Cs) was determined as the difference between total 137Cs and
Chernobyl-derived 137Cs obtained by multiplying the activity of
134Cs by the isotopic ratio of 137Cs:134Cs in the Chernobyl fallout.
239+240Pu was determined by α-spectrometry after radiochemical
separation and purification. The method has been described by
Bunzl and Kracke (1994). The experimental error (1 SD) of Cs
was in general <30%, and that of Pu <5%.
Determination of soil redistribution rates by the tracer method
Fig. 1 Generalized result of soil mapping along slope transects.
Arrows indicate positions of radiotracer sampling. Open circles indicate a considerable contribution of rock fragments to the soil
material. Depth of the soil horizons is exaggerated by a factor of 5
as compared to the height axis
Fig. 2 Soil erosion (a, top) and
deposition rates (b, bottom) as
derived from 137Cs (Cs, left)
and 239+240Pu (Pu, right) from
global fallout, shown as bars
on the topography of the field
with the layout of the sampling
points (distances refer to the
Gauss Krueger coordinates
44 57186 E and 53 71964 N).
As an example, the estimated
error (1 SD) of the erosion and
deposition rates is shown at
one point for each radionuclide.
The scale of the erosion and
deposition rates given at the
bottom, right, applies to all
graphs of this figure. A vertical cross-section of the area
along a transect (1,100 m north)
is shown in Fig. 1
The inventories of Cs and Pu in the field and at the reference site
were published in Schimmack et al. (2001). The reference values
were 2,790±730 (1 SD) Bq·m–2 for Cs and 58.8±8.1 Bq·m–2 for
Pu. Field points were considered as ‘erosion’ points if the inventory of Cs or Pu was smaller than the reference value minus one
standard deviation (SD), and as ‘deposition’ points if the inventory of Cs or Pu was greater than the reference value plus 1 SD. Soil
45
(t·ha–1·year–1)
erosion rates
were calculated from radiotracer loss
using the mass balance model 2 of Walling and He (1997). The
particle size correction factors for 137Cs in the eroded material
were about 1.1 (depending on slope position). These were based
on measurements undertaken by Weigand et al. (1998) at Scheyern
during single events, and take account of the relationship of the
enrichment between long-term average annual soil loss and single
event soil loss (Auerswald 1989). The same mass balance model
was used for the Pu data from the field, taking into account the
different half-lives of 137Cs and 239+240Pu. The enrichment of Pu in
the eroded sediment was considered as for 137Cs, because both
radionuclides are concentrated into the finer size soil fractions in
a similar way (Livens and Baxter 1988). Soil deposition rates
(D; t·ha–1·year–1) were estimated from the difference between the
thickness of the total Ap horizon in the sampling year 1994 (X94)
and the thickness of the Ap horizon in 1970 (X70), at the beginning
of the erosion period. X70 (in mm) was calculated from the tracer
concentration (C; Bq·kg–1) in the Ap horizon of the field point and
the tracer inventory AR (Bq·m–2) at the reference site, assuming
(1) that C and the dry bulk density (BD; g·cm–3) at the sampling
point were constant throughout the erosion period (this is a simplification, but the resulting error should be small) and (2) that the
reference inventory measured in 1994 was representative for the
inventories over the entire field in 1970. Then X70=AR/(BD×C)
and D=10×BD(X94–X70)/24. The error of the redistribution rates
(1 SD) was estimated by taking into account the errors of the reference value and of the inventories at the field point. For the rates
derived from Cs the error was on average ~90%, for Pu ~30%.
Determination of soil redistribution rates
by the ‘translocation model’
Soil erosion by water was calculated using the Universal Soil Loss
Equation (USLE, Wischmeier and Smith 1978), which has been
extensively tested and adapted for this landscape (for references
see Schwertmann et al. 1987). It predicts long-term average soil
loss from rain erosivity, soil erodibility and factors accounting for
the influences of slope gradient, slope length, cropping methods
and tillage direction. A modification which differentiates soil loss
along the slope was used (dUSLE, Bork and Hensel 1988; Flacke
et al. 1990), resolving the research site into 1,868 partitions of
14 m2 on average. In this version, the influence of slope length is
replaced with the upslope watershed area per unit flow width to
account for diverging or converging slopes and, instead of computing a field average, the calculations are carried out successively
downslope to take into account local changes of erosion. The rain
erosivity was determined from long-term rainfall data for a location 1 km distant from the experimental site, the topographic factors from a geodetic survey, the soil erodibility from soil properties measured at the sampling locations with standard methods and
the crop and cover factor from cropping records during the
24 years of arable soil use. Soil translocation by tillage was calculated using the equation given by Lindstrom et al. (1992) for tillage up and down slope and a generalized elevation model. This
equation predicts that soil movement during an individual tillage
operation increases linearly with the gradient. The soil loss or gain
at a certain location is then the difference between the soil import
from upslope and the soil export downslope, which are not balanced where slope gradient changes along a slope.
Results and discussion
Soil redistribution rates derived from Cs and Pu
The soil erosion rates derived from Cs and Pu are shown
in Fig. 2a, and the corresponding deposition rates in
Fig. 2b. The mean erosion rate derived from Cs was
–37 t·ha–1·year–1 (coefficient of variation CV=57%),
corresponding to –2.6 mm/year, and that from Pu
–32 t·ha–1·year–1 (CV=44%) or –2.2 mm/year. The mean
deposition rate derived from Cs was 52 t·ha–1·year–1
(CV=28%), corresponding to 3.5 mm/year, that from
Pu 39 t·ha–1·year–1 (CV=51%) or 2.7 mm/year. According
to the results of t- and U-tests, the difference between the
means for Cs and Pu is statistically not significant [erosion rates, t-test: t=0.955, P=0.35; U-test: U=93.0,
P=0.56; n=7 (Cs) and 31 (Pu), respectively; deposition
rates, t-test: t=1.63, P=0.12; U-test: U=23.5, P=0.094;
n=15 (Cs) and 6 (Pu), respectively]. This would indicate
similar quality of both tracers. However, the estimated errors (1 SD) of the rates derived from Cs were on average
about 90%, while those for Pu were only around 30%. As
a consequence, the 95% confidence interval (2 SD) of the
rates derived from Cs always includes zero. This demonstrates that the rates obtained from Cs are not accurate
enough to assure the presence of a net soil redistribution
in this field. In contrast, the rates derived from Pu are
more reliable, because (1) their 95% confidence interval
does not include zero for most points, and (2) their spatial
pattern agrees much better than that of Cs with expectations based on soil mapping (Fig. 1). The field is extremely steep, and erosion and deposition areas are clearly separated. We will thus not consider Cs any further but
discuss only the results obtained with Pu. Since no soil
redistribution rates derived from Pu have been reported in
the literature, we compared our results with those of
many studies performed worldwide using Cs as a tracer.
In terms of both the range of the values as well as their
means, the redistribution rates derived from both tracers
are of the same order of magnitude (see Montgomery et
al. 1997; Walling and He 1999).
Comparison with results of calculations based
on a prediction model (‘translocation model’)
For these calculations, the field points along the horizontal
transects [east coordinates (m) 850, 875, 900, 925, 950,
and 975, see Fig. 2, corresponding to mean heights above
sea level of 464, 466, 471, 477, 480, and 482 m] were
pooled, resulting in a vertical transect of mean Pu inventories. Evaluating them by the methods described above, the
resulting ‘transect’ redistribution rates can be compared
with results of the model calculations, which include the
differentiated Universal Soil Loss Equation (dUSLE) and
a tillage model. The total soil redistribution rates agree
reasonably well with those derived from Pu (Table 1). The
mean erosion rate provided by the translocation model
was –29 t·ha–1·year–1 and, thus, quite close to the mean of
the four erosion rates derived from Pu (weighted by the
number of sampling points), –33 t·ha–1·year–1. While the
tracer method yields ‘overall’ redistribution rates, which
cannot be attributed to a specific process, the translocation
model allows us to differentiate between water and tillage
erosion. Recent studies have shown that tillage erosion is
an important soil redistribution process in agricultural
fields (Quine 1999). Table 1 indicates that, in the field investigated, tillage produced twice as much erosion as rain
and created most of the deposition, because in contrast to
46
Table 1 Comparison of soil redistribution rates derived from
239+240Pu as a tracer (tracer method) with estimates based on a soil
translocation model including rain and tillage erosion. For the
tracer method, the Pu inventories of the field points along six transects perpendicular to the slope at the field site, characterized by
their mean height (m a.s.l.), were pooled. Negative redistribution
rates represent erosion, positive values deposition, of soil
Mean transect
height (m)
Soil redistribution rate (t·ha–1·year–1)
Tracer method
482
480
477
471
466
464
–19
–36
–44
–18
0
+43
Acknowledgements We would like to thank Professor D.E.
Walling for providing the software for the calculation of soil erosion from tracer loss. This work was financed in part by the
Bundesministerium für Bildung, Wissenschaft, Forschung und
Technologie (BMBF No. 0339370) and the Bayerische Staatsministerium für Unterricht und Kultus, Wissenschaft und Kunst.
References
Translocation model
Total
Water
Tillage
–13
–46
–35
–13
+14
+43
–8
–10
–14
–10
–2
+8
–5
–36
–21
–3
+16
+35
water erosion, no soil was transported beyond the field
borders. Similar results have been reported by other authors (Quine 1999). The modeling also explains the result
of soil mapping which found that colluviation had already
started in the middle of the slope, where the curvature
changed from convex to concave and thus induced an accumulation of material moved by tillage (Fig. 1). In contrast, sheet erosion by rain should be large at this steepest
part of the slope (Table 1). The accumulation of rock fragments (up to 50% weight) in the plow horizon at this slope
position possibly indicates this extensive erosion of the
fines by sheet erosion, which will not move the rock fragments. The modeling result for both erosion processes
hence matches the findings of mapping.
Conclusions
The soil redistribution rates derived from 137Cs (Cs) or
239+240Pu (Pu) from global fallout varied considerably, even
on a relatively small scale of 20–50 m. Great caution is
hence required in interpreting the estimates of soil redistribution obtained from tracer values for individual or a small
number of points. However, the rates derived from Pu were
more accurate than those derived from Cs and agreed
much better with expectations from modeling and mapping. Therefore, Pu can serve as an appropriate erosion
tracer instead of Cs in areas where the global fallout fraction of the total 137Cs cannot be distinguished from the
Chernobyl fraction with sufficient precision. The quantification of soil translocation from the Pu movement showed
that erosion is a significant factor in soil degradation in this
landscape. Tillage erosion contributed considerably more
than water erosion to the translocation. The close association of eroded and colluvial soils within the landscape can
only be explained by tillage erosion. This is the first time,
to our knowledge, that tillage erosion has been quantified
under German conditions. The measurement agreed reasonably well with the predictions from modeling water and
tillage erosion, indicating that the modeling tools are suitable for guiding soil conservation in this landscape.
Auerswald K (1989) Predicting nutrient enrichment from longterm average soil loss. Soil Technol 2:271–277
Auerswald K, Schmidt F (1986) Atlas der Erosionsgefährdung in
Bayern – Karten zum flächenhaften Bodenabtrag durch Regen.
GLA-Fachberichte 1, Munich
Bork HR, Hensel H (1988) Computer-aided construction of soil
erosion and deposition maps. Geol Jahrb A 104:357–371
Bunzl K, Kracke W (1994) Efficient radiochemical separation
for the determination of plutonium in environmental samples,
using a supported, highly specific extractant. J Radioanal Nucl
Chem Lett 186:401–413
Flacke W, Auerswald K, Neufang L (1990) Combining a modified
Universal Soil Loss Equation with a digital terrain model for
computing high resolution maps of soil loss resulting from
rain wash. Catena 17:383–397
Foster GR, Hakonson TE (1987) Erosional losses of fallout plutonium. DOE Symp Ser 59:225–53
Frede H-G, Dabbert S (eds) (1998) Handbuch zum Gewässerschutz in der Landwirtschaft. Ecomed, Landsberg, Germany
Hötzl H, Rosner G, Winkler R (1987) Ground depositions and air
concentrations of Chernobyl fallout radionuclides at MunichNeuherberg. Radiochim Acta 41:181–190
Lindstrom MJ, Nelson WW, Schumacher TE (1992) Quantifying
tillage erosion rates due to moldboard plowing. Soil Tillage
Res 24:243–255
Livens FR, Baxter MS (1988) Particle size and radionuclide levels
in some west Cumbrian soils. Sci Total Environ 70:1–17
Montgomery JA, Busacca AJ, Frazier BE, McCool DK (1997)
Evaluating soil movement using caesium-137 and the revised
universal soil loss equation. Soil Sci Soc Am J 61:571–579
Pimentel D, Harvey C, Resosudarmo P, Sinclair K, Kurz D,
McNair M, Crist S, Shpritz L, Fitton L, Saffouri R, Blair R
(1995) Environmental and economic costs of soil erosion and
conservation benefits. Science 267:1117–1123
Quine TA (1999) Use of caesium-137 data for validation of spatially distributed erosion models: the implications of tillage
erosion. Catena 37:415–430
Schimmack W, Auerswald K, Bunzl K (2001) Can 239+240Pu replace 137Cs as an erosion tracer in agricultural landscapes
contaminated with Chernobyl fallout? J Environ Radioactiv
53:41–57
Schwertmann U, Vogl W, Kainz M (1987) Bodenerosion durch
Wasser – Vorhersage des Abtrags und Bewertung von Gegenmaßnahmen. Ulmer Verlag, Ulm, Germany
Walling DE, He Q (1997) Models for converting 137Cs measurements to estimates of soil redistribution rates on cultivated and
uncultivated soils (including software for model implementation). A contribution to the IAEA Coordinated Research Programmes on Soil Erosion (D1.50.05) and Sedimentation
(F3.10.01), Department of Geography, Exeter, UK
Walling DE, He Q (1999) Improved models for estimating soil
erosion rates from cesium-137 measurements. J Environ Qual
28:611–622
Weigand S, Schimmack W, Auerswald K (1998) The enrichment
of 137Cs in the soil loss from small agricultural watersheds.
Z Pflanzenernaehr Bodenkd 161:479–484
Wischmeier WH, Smith DD (1978) Predicting rainfall erosion
losses: a guide to conservation planning. (USDA agriculture
handbook 537) US Department of Agriculture, Washington,
D.C.