Geoscience Reference
In-Depth Information
12.3 ReSUltS
At Apelsvoll, measurements of EC
a
were correlated with most of the measured soil properties
(Table 12.2). The contents of gravel and fine sand were the only properties that did not affect EC
a
.
Most correlations were found between EC
a
and soil properties in the depth layer 30 to 50 cm. The
correlations were generally strongest when EC
a
was measured in horizontal mode (EM
H
). This was
especially so in the topsoil layer, where there was no correlation between EM
V
and the measured
soil properties.
At Møystad, most correlations were found between EC
a
and the topsoil properties (Table 12.3).
EC
a
was unaffected by the contents of fine sand, fine silt, and coarse silt, regardless of depth. As
found at Apelsvoll, the topsoil properties were more strongly correlated with EM
H
than with EM
V
.
In contrast to the results at Apelsvoll, however, this trend appeared to change with depth at Møystad,
where the soil properties in the deepest layer (40 to 60 cm) were most strongly correlated with EM
V
.
At both locations, the strongest correlations were found between EM
H
and topsoil clay content.
The total content of topsoil sand was the second most important soil property influencing EM
H
.
The content of clay and total sand were strongly negatively intercorrelated (|r| > 0.580) at both
locations and at all depths. In contrast to the clay content, the sand content was, as expected, nega-
tively related to EM
H
. At Apelsvoll, these two properties were the only ones that had a significant
influence on EC
a
at all three depths. At Møystad, this applied for ignition loss as well, and here the
relative influence of ignition loss appeared to increase with increasing depth.
Even though many of the soil properties were rather strongly correlated with EC
a
, only a few
were included in the regression models using the stepwise procedure (Table 12.4). Topsoil clay
content alone accounted for 57 percent of the variation in measured EM
H
at Apelsvoll and 67 per-
cent of that at Møystad. When expanding the models to include all potential predictors, the cor-
responding degrees of explanations increased up to 72 and 89 percent, respectively. Topsoil clay
tAble 12.2
Correlation Coefficients between Selected Soil properties at three
depths and Apparent Soil electrical Conductivity (log-transformed),
Measured in Autumn 2002 at Apelsvoll (
n
= 18) Using eM38
in either horizontal (eM
h
) or vertical (eM
v
) Modes
depth (cm)
0-20
30-50
60-80
Soil
property
log eM
h
log eM
v
log eM
h
log eM
v
log eM
h
log eM
v
Gravel
a
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
Coarse sand
n.s.
n.s.
−0.51*
−0.55*
n.s.
n.s.
Medium sand
n.s.
n.s.
−0.47*
n.s.
−0.60**
0.58*
Fine sand
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
Total sand
a
−0.63**
n.s.
−0.75***
−0.59**
−0.54*
n.s.
Coarse silt
n.s.
n.s.
0.47*
0.49*
n.s.
n.s.
Medium silt
n.s.
n.s.
0.63**
0.53*
0.49*
n.s.
Fine silt
0.51*
n.s.
0.62**
0.55*
n.s.
n.s.
Total silt
a
n.s.
n.s.
0.60**
0.56*
n.s.
n.s.
Clay
a
0.77***
n.s.
0.77***
0.50*
0.50*
n.s.
Ignition loss
0.62**
n.s.
n.s.
n.s.
n.s.
n.s.
Note:
*
p
< 0.05; **
p
< 0.01; ***
p
< 0.001, n.s. not significant.
a
Particle size of gravel: >2000 μm, total sand: 60 to 2000 μm, total silt: 2 to 60 μm and clay:
<2 μm.
