Agriculture Reference
In-Depth Information
with the traditional soil analyses in laboratories for nitrogen - the Nmin method - in
humid regions the samples are collected and averaged from a 90 cm deep profile
(Marschner
2008
). Since the relocation of nitrate-N will depend largely on the pre-
cipitation, site-specific spectral sensing should at least be adjustable for the depth
from the soil surface.
For on-the-go sensing in fields, the mid-infrared radiation would also have to
cope with varying soil moisture (Sect.
9.3.2
). A discrete narrow waveband might be
better suited for this than a full spectrum. However, the verification of this assump-
tion has not yet been provided.
9.4
Fertilizing Nitrogen Based on In-Season Crop Properties
Basing the control of fertilizing on
past crop properties
instead of those of the soil
is state of the art when the nutrients that were removed by the yields of previous
crops are relied on (Sect.
9.1
) However, because of the time scale involved, this
method lends itself for detecting rather long-term supplies of those nutrients, whose
availability temporally does not change fast such as
e.g.
phosphorus.
Contrary to this, basing the fertilizing on
current crop properties
is a method
that can provide signals for rather immediate, short-term control and hence with a
high temporal precision. The ideal application for this method is associated with
online and on-the-go proximal sensing and control in real-time. This in turn means
that this method fits well to nutrients, for which the supply of crops changes rather
fast. The classical macronutrient for this method hence is
nitrogen
, and so this sec-
tion is devoted to the application of this nutrient. This limitation, however, does not
imply that sensing on the basis of current crop properties might not be useful for
other applications as well.
It is obvious that any use of
c
rop properties
for the control of farming opera-
tions is also a control that is oriented on soil properties - at least partly. This is
simply because the properties that the plants have are not independent of the soil on
which they grow. Therefore a very important advantage of basing the control on
properties of the crops is that the signals that are thus obtained depend on an
inter-
action
between the plants and its soil. This means that
e.g.
the influence of the
respective root development on the nutrient uptake is automatically taken care of.
All nutrient extraction methods that are used in soil laboratories so far aim at a
simulation of this influence of the soil to crop relation on the availability of nutri-
ents. However, this simulation of the extraction of nutrients by crops in laboratories
has to cope with changing situations, e
.g.
varying root lengths during the develop-
ment of plants. The best solution to this problem of fertilizing control certainly
simply is to use the respective
site-specific soil to crop relation
in the field as it
exists by an intelligent sensing system.
Fertilizing based on current crop properties can provide this. It can be oriented at
the actual nutrient uptake of crops, assuming that suitable surrogate properties for
this are available and selected.
