Geoscience Reference
In-Depth Information
13.2.2 eM-ec
a
d
e v i c e
The device used to measure EM-EC
a
was the Geonics soil conductivity meter EM38 (Mississauga,
ON, Canada; www.geonics.com), which is the sensor most often used for such measurements in
agriculture (Sudduth et al., 2001). The device may be used in one of two measuring modes, in verti-
cal mode (coil axes perpendicular to soil surface, denoted EM
V
) or in horizontal mode (coil axes
parallel to soil surface, denoted EM
H
). The respective effective measuring depths are approximately
1.5 m and 0.75 m. When measuring, the device was placed in a plastic cylinder, which was mounted
on a plastic sledge and towed behind an off-road vehicle for continuous measurements (ca. 200
measurements ha
−1
). A differentical Global Positioning System (DGPS) receiver (Ceres, CSI Wire-
less Inc., Canada) was attached on the sledge to deliver the geographical coordinates of the EC
a
measurements.
13.2.3 M
e a s u R e M e n t s
On 30 April 2003, 310 soil samples were taken from the topsoil (0 to 20 cm) in a 20 × 20 m grid,
covering about 38 percent of the field (12 ha) and analyzed for ignition loss, pH, P-AL (AL denotes
the ammonium lactate/acetic acid mixture used for extraction; Égner et al., 1960), K-AL, Mg-AL,
Ca-AL, Na-AL, and K-HNO
3
. A selection of 154 soil samples were additionally analyzed gravi-
metrically for water content, and mechanical analysis of texture was performed on forty of these
samples. From the latter group, twenty-one samples were extracted with 1 M ammonium acetate,
and exchangeable Ca
2+
, K
+
, Mg
2+
, and Na
+
were analyzed spectrometrically (ICAP 1100, Thermo
Jarrell Ash Corp, US). Exchangeable H
+
was determined by titration (NaOH) and cation exchange
capacity (CEC) calculated as the sum of the five cations. Soil organic matter (SOM) was determined
from ignition loss, by correcting for clay content, where the clay content was estimated by the finger
method in those samples that lacked ordinary texture analysis. The same day as the soil sampling,
EM-EC
a
was measured in the field with the EM38 in horizontal mode only. The rationale for using
this mode is that the relative contribution to the signal from the topsoil is larger for EM
H
than for
EM
V
(McNeill, 1980), and that EM
H
has been found to be superior to EM
V
in practice, when using the
EM-EC
a
technique to map topsoil properties on fairly comparable soils (Korsaeth, 2005a, 2005b).
13.2.4 s
t a t i s t i c s
a n d
d
a t a
a
n a l y s e s
The EC
a
data were not transformed, as this is seldom done in commercial practice. The statistical
software package MINITAB (Release 14.13, www.minitab.com) was used for basic statistics (cal-
culating means, standard deviations, and correlations), one-way analysis of variance (ANOVA), and
Fisher's least significant difference (LSD) test for comparisons. ArcView GIS (version 3.2, www.
esri.com) was used to produce the maps.
Soil properties that showed a reasonably strong correlation with EM
H
(|r| > 0.5), and from which
at least forty observations were available (thus excluding the exchangeable cations), were selected
to test the standard procedure used for commercial soil survey. On the basis of measured EC
a
, the
selected data were grouped separately into classes with intervals of 2 mS m
−1
(groups with less than
three observations were excluded). Thereafter, one-way ANOVA and Fisher's LSD test were used to
test for differences between groups with respect to each of the selected variables. The rationale for
using an interval of 2 mS m
−1
was that experience from commercial mapping of a range of different
soil types in Norway has shown that an interval of this size is normally most suitable.
13.3 ReSUltS
Measured EM
H
correlated significantly with sixteen of the twenty-two measured soil properties
(Table 13.1). Properties with a correlation coefficient (r) larger than 0.8 were ranked in descend-
ing order: CEC > soil water content > ignition loss > exchangeable Ca > SOM. All significant
