Agriculture Reference
In-Depth Information
thrown back, whereas the fluorescence comes from the photosynthetic apparatus of
plants as surplus light energy or as a by-product. Artificial induction of fluorescence
often is done by pulsed red laser beams.
In the red wavelength range, fluorescent light is emitted from the leaves with two
peaks at 680 and 735 nm. It has been shown (Lichtenthaler
1996
) that the relation-
ship of the fluorescent intensities at these two peaks is an indicator of the
chloro-
phyll concentration
in leaves. This indication of chlorophyll concentration is
explained by differences in the re-absorption of fluorescent light at these two emis-
sion peaks in the red and near-infrared range. The red fluorescent light at 680 nm
wavelength is partly re-absorbed by chlorophyll for photosynthesis. This re-absorption
depends on the chlorophyll concentration within the leaves. On the other hand, the
fluorescent light at 735 nm wavelength is above the range of absorbed light. It there-
fore is barely reduced by absorption. For that reason, the
ratio
of the
two red fluo-
rescence peaks
can be used to sense the chlorophyll concentration and thus also the
nitrogen concentration in the leaves. So the background is chlorophyll estimation
from
fluorescence absorption
. But what is the correlation between this fluores-
cence ratio and the nitrogen supply?
The fluorescence ratio was recorded by means of a handheld instrument (Thiessen
2002
). Nitrogen top dressings for winter-cereals and winter-rape were applied three
times in the growing season, beginning in early March. Readings were taken at the
time of the third dressing and 3-4 weeks after this date. The actual dates for the
dressings depended on the crop species, but typical dates for Northern Germany
were used (Fig.
9.30
).
The basic assumption is that with increasing nitrogen supply by the preceding
dressing, the chlorophyll concentration in the leaves should go up and thus the
fluorescence ratio should drop. At the time of the third dressing - which is in mid-
May for winter-barley or in early June for winter-wheat - this assumption is sup-
ported up to a preceding nitrogen supply of 120 kg/ha. Beyond this level of the
nitrogen supply, the fluorescence ratio does not drop any more. On the contrary, it
rises (Fig.
9.30
, top).
A similar trend showed up 3-4 weeks after the third dressing with winter-barley,
with winter-wheat and with winter-rape. However, at this time, the fluorescence
ratio decreased up to a preceding nitrogen supply of 160 kg/ha (Fig.
9.30
, bottom),
but above this rate, it also rose. In conclusion, at both dates and with all crops, there
was no unidirectional relationship, which is a prerequisite for a simple control sys-
tem. One might argue, that a unidirectional relationship up to a range of 120-160 kg/ha
suffices, since higher pre-applications at earlier dressings seldom occur. So sensing
of nitrogen by fluorescence absorption might be a feasible option except perhaps for
very high dressing rates. For further results about nitrogen sensing by fluorescence
see Thiessen (
2002
), Schächtl et al. (
2005
), Thoren (
2007
), Thoren and Schmidhalter
(
2009
), Thoren et al. (
2010
).
Contrary to signals from reflectance, fluorescence sensing is never associated
with erroneous information from
non-vegetated soil
. Because bare soil does not
emit fluorescence. Neither any soil nor its plant residues or its stones influence the
signals. The fluorescence ratio is solely based on the chlorophyll concentration in
the leaves. This can be a distinct advantage when the crop canopy in not closed.
