Agriculture Reference
In-Depth Information
would need to be pre-processed for real-time site-specific buffer corrections. Such
a combination of real-time sensing on the one hand and simultaneous use of pre-
processed georeferenced maps with long-term soil properties on the other hand
might be useful for an intelligent control of several field operations.
9.3.3
Sensing the Phosphorus Requirement
Phosphorus exists in soils within different compounds, especially in phosphates of
calcium, aluminium and iron. These phosphate compounds are mainly a feature of
natural soil properties, yet to some extent also originate from fertilizers, hence from
human activities. In moderately acid soils as well as in alkaline soils, calcium-
phosphates dominate. When the soil pH goes down, the proportion of aluminium-
and iron-phosphates increases. All these phosphate compounds have very different
optical spectra. They can be sensed spectrally rather easily with a very low classifi-
cation error (Bogrekci and Lee
2005
).
This information, however, only indicates that there are calcium-, aluminium- or
iron phosphates in a soil. It discloses nothing about the availability of phosphorus to
plants. In fact, the important criterion for the supply of crops and hence for site-
specific control of fertilizing is the
plant available phosphorus
.
This exists in soil
solutions either as hydrogen-phosphate anions (HPO
4
2−
) or as dihydrogen-phosphate
anions (H
2
PO
4
−
). The former anion dominates in weakly alkaline soil solutions, and
the latter instead in slightly acid situations. Compared to the amounts of phosphorus
in the soil phosphates, those of plant available phosphorus in the soil solution usu-
ally are very small. This is because phosphorus does not remain in solution for long
in soils. The anions that crops extract from the soil solution are usually replenished
from calcium-phosphates, provided the soil has a reservoir of these.
In soil laboratories, the plant available phosphorus is defined by chemical extrac-
tion. A standard extractant is the Olsen solution of sodium-bicarbonate, though also
other extractants are used. For highly fertilized soils, sometimes simply water is
recommended as an extractant (Finck
1991
).
Sensing the plant available phosphorus by
reflectance
in laboratories tradition-
ally has been done on the basis of
dried soil samples
. However, Maleki et al.
(
2006
) and Mouazen et al. (
2006
) concluded that the spectral prediction of plant
available phosphorus in
fresh, wet soils
is better than in dried samples whilst
portending that the sensing takes place for anions in a soil solution. Their hypoth-
esis is that hence this phosphorus fraction in the water phase of the soil correlates
well with spectral signals and consequently more water implies also more plant
available phosphorus.
The spectra in Fig.
9.12
are based on sensing of fresh, moist soil samples from
several fields in Belgium. But since the water content of the sensed soil samples
varied, the contents of available phosphorus were presented on the basis of dry soil
in order to allow for a precise comparison. Yet this does not alter the original state
of the soil samples at the time of sensing. The higher the soil content of available
