Agriculture Reference
In-Depth Information
65
EC = 1.13 clay% + 0.24
r
2
= 0.70
45
25
field Dikopshof
field Frankenforst
5
0
10
20
30
40
50
mean clay content in %
65
EC = -1.03 silt% + 92.9
field Dikopshof
field Frankenforst
r
2
= 0.71
45
25
EC = 0.13 silt% + 7.8
r
2
= 0.65
5
0
20
40
60
80
mean silt content in %
Fig. 5.8
Electrical conductivities depending on the clay- and the silt contents of two soils near
Bonn, Germany. For both soils, the effects of the clay on the conductivities were similar. Therefore,
the results are presented in a common regression (
top
). Data points in
brackets
were excluded from
the calculation. The effects of the silt were quite different (
bottom
). A comment to this is in the text
(From Mertens et al.
2008
, altered)
However, the sensing method does not provide for signals about different layers
within the depth response curves, it supplies averaged data for the whole sensed
depth. The topsoils of both fields are mainly loams that originated from loess with
varying thicknesses between some cm and 1.5 m. The subsoil of both fields is quite
different, for the field Dikopshof it is sandy and for the field Frankenforst it is
clayey. These differences in the subsoil mainly cause the shifting of the clay points
for Frankenforst to a much higher range. So the textures of both soils were different
not only horizontally but especially in vertical directions.
Theoretically, the silt content should have only a small- and the sand content
almost no influence, since the conductivity is mainly defined by ions of the soil
constituents. However,
autocorrelation
with the clay content must be considered.
Sand content is defined when clay- and silt content are known since these three
texture classes add up to about 100 %. And apart from sand, even the silt content
alone can depend on the clay content. When the clay content is very high, there is
less space left for silt. So if total autocorrelation for the sand content holds when
