Agriculture Reference
In-Depth Information
grain. In humid regions, site-specifi c
residue maps
can be useful for controlling the
depth of stubble cultivation (see Sect.
7.4
)
.
The relation between the crop residue mass and the recorded grain mass can be
derived from the harvesting inde
x
, which is the fraction of the total aboveground
biomass that is allocated to the grain. For most crops and its varieties, this harvest-
ing index is well known or can easily be determined (Kemanian et al.
2007
). The
residue index - the biomass-fraction of the aboveground residues - is simply 1 (one)
minus the harvesting index. Furthermore it holds:
yield of residue mass
yield of grain mass
residue index
harvesting if
1
−
harvesting index
harvesting index
≈
≈
ndex
So if the mass of grain is mapped and the harvesting index is known, it is rather
easy to get from this to a residue map. And using this for precisely controlling the
site-specifi c depth of stubble cultivation can avoid waste of energy. However, an
accurate map about crop residues requires considering the time that the straw as
well as the grain spend within the combine, hence distance lags must be compen-
sated for in the georeferencing processes.
It has been expected that
yield maps
could be useful for predicting the site-
specifi c performance of crops constantly. This expectation has been based on the
assumption that permanent soil properties such as texture, organic matter and topog-
raphy would always behave in the same way each year. If the site-specifi c yields
were temporally stable, hence the yield maps would be a useful management tool.
The input of agrochemicals could be oriented at such yield maps. But so far, this
expectation of a
constant site-specifi c yield pattern
per fi eld has not been met.
Blackmore et al. (
2003
) tried to predict the actual site-specifi c yield map pattern
from yields in previous harvests in four fi elds and over 6 years for wheat, barley
and rape (colza) in rainfed areas of England. The site-specifi c cell size was
20 m × 20 m. The mean yields of the whole fi elds differed from year to year due to
good or bad weather or because of varying disease or pest attacks. However, the
inter-year offsets that resulted from these different average yields were neutralized
by the data processing.
The coeffi cients of determinations between the site-specifi c yields of the previ-
ous harvests and the respective actual harvest of the fi elds went from a very low
level of r
2
= 0.02 up to a maximum of r
2
= 0.43 with an average for 11 comparisons
of only r
2
= 0.20. It was concluded that yield maps cannot predict the corresponding
site-specifi c yield in the following year. However, spatial and temporal maps could
help to identify larger
homogeneous management zones
. The authors recom-
mend based on these fi ndings that the growing crop should be managed according
to its current needs. And these current needs would have to be derived from the
actual crop itself.
Experiments from Rothamsted Research in England (Milne et al.
2012
) confi rm
these fi ndings. These experiments were limited to winter-wheat that was grown in
3 years - 2002, 2004 and 2006 - in a rotation with rape and peas. Only the results
with wheat are shown (Fig.
13.4
) for site-specifi c cell sizes of 10 m × 10 m. Some
