Site-Specific Farming — Number 3
Yield Mapping
SF-1176 (3), June 1999
Dr. Dave Franzen, NDSU Extension Soil Specialist
Yield monitors are available for use in most grain crops and for some bulk crops. A
yield monitor enables a producer to measure yield and grain moisture at many locations in
a field. These data can be processed with computer mapping programs into a yield map. The
yield map is not only a record of yield variability within the field, but also allows a
producer to research the question of why certain areas produced the way they did. Years of
yield map data can be compared for similarities, resulting in a change of management or
inputs to those areas that may increase yields or decrease input costs. Thousands of
points in a yield map can be used to evaluate treatments in strip trials instead of just
one or two as in old standard weigh wagon trials.
Yield monitors do not actually weigh grain. They operate on a principle of force
against a pressure sensitive plate as the grain moves up the clean grain auger into the
grain hopper (Figure 1A).
Moisture is measured continuously in the clean grain elevator (Figure 1B). The force
measurements and moisture are compared against the calibration performed early in the
season by the on-board computer (Figure 1C). The GPS signal is combined with the resulting
yield information to provide an address for each measurement.
Figure 1. Principle components of a yield monitor. (8KB
color illustration)
What does each measurement mean?
Each yield measurement is pounds or bushels of grain for an area equal to the combine
width times the length the combine travels in the time the operator sets the measurement
for (one, two, three or more seconds). The computer tallies the force/yield measurement
for the designated time periods and presents a mean (average) value for that length of
time. For a combine traveling at 4 mph, the distance traveled per second is 5.87 feet. For
a 20-foot head, the area estimate is for about 117 square feet, or 0.003 acres. There may
be considerable error in each yield measurement. Grain does not flow evenly through a
combine. It travels through in spurts. Grain yields are variable on a small scale, and the
flow tends to even out some of the small-scale yield variability as it passes through the
combine cleaning system. It also takes time for grain yield to be measured from the time
the grain enters the combine head to the time it is measured. If it takes four seconds to
reach the force plate, the unit has traveled almost 24 feet before yield is recorded.
The combination of lag time, unevenness of flow through the combine and evening out
some small scale variability makes each individual measurement of the yield map full of
error and inaccuracy. It is therefore desirable when producing a yield map that several
adjoining measurements be averaged together to produce a larger zone of yield. This larger
zone is more representative of the area yield than some of the measurements it contains.
The problems associated with error as the grain moves through the combine are also
important when comparing treatments in field-size studies. A recent study at NDSU suggests
that the width of strip trials is not particularly important, but that the combine must
travel about 400 feet before error is minimized in the yield estimate. Therefore, if
yields are to be determined by yield monitor, the strips should be at least 400 feet long
in each treatment. Strips shorter than this tend to average some of the preceding or
following treatments within the interior of the other treatments.
What can be seen from a yield map?
In a trial at Oakes, yield was measured with a yield monitor in successive years. In
both years there were similar patterns of yield. The south half of the field tended to
yield better than the north half (Figure 2A). There was a strip of particularly low yields
along the south of the northernmost east-west lane through the field (Figure 2B). However,
in 1994 there was a pie-shaped area in the southwest that does not appear in the 1995 map
(Figure 2C). Without ground-truthing, it is not possible to determine what caused these
differences. A yield map is an indicator of what has happened and can be a predictor of
upcoming events if causes of high and low yield are identified through careful scouting,
sampling and record-keeping.
Figure 2. A quarter-section irrigation pivot in corn,
1994-1995, Oakes. (Knighton) (25KB color image)
Through soil sampling and analysis, the reason for the consistently higher yields in
the south half of the field was determined to be due to better drainage and lower salt
levels. The water table is higher in the north part of the field and rises closer to the
surface as the groundwater flow moves toward the west, resulting in a semi-permanent
wetland in the northwest corner. Perhaps alfalfa barriers in the north part of the field
could be seeded to lower the water table and decrease salt over time in the field. The
persistently low yields on the south side of the north lane are caused by intensive weed
pressure. The corners of the field are lower in yield because the irrigation water in the
center-pivot does not reach these areas. The pie-shaped area in the southwest was caused
by running out of a pivot applied insecticide for European corn-borer that summer. The map
indicates that insect control resulted in a 30-40 bu/acre benefit, and also indicates that
it would have paid to finish the application over the entire field.
Figure 3. Identification of problem areas at Oakes, 1994.
(Knighton) (21KB color image)
At Valley City, there are several areas with yields of about 10 bu/acre. With
supporting soil test data and good record keeping by the growers, it is clear that two of
the areas are saline seeps. The remainder of the low yielding areas were avoided by the
combine in the initial mapping because they were seriously lodged, and when the areas were
finally combined, the yield monitor was left off. These areas contained significant yields
but were not measured in the mapping process. Other low testing areas may have been
affected by low micronutrient levels and are being investigated in another study.
Figure 4. Valley City yields and lodging. (Franzen)
(8KB color image)
Benefits of yield mapping
There are several benefits of yield mapping. One is simply creating a yield record by
location for each field. These records can be shown to landowners to support a reasonable
change in rental agreement, show the effects of management change, or to support a change
in management such as additional drainage. The yield map also is a chance to pass down
management and yield records to heirs, land buyers or renters. The information is useful
and valuable. In the future, farms with yield map records should be more valuable than
those without, similar to the tendency that machinery with good maintenance records is
worth more than machinery without.
The yield map increases curiosity about the field and serves to point out areas with
exceptionally high or low yields. Once areas are identified, ground-truthing will help to
point out the reasons behind the differences. It is also possible to create a
profitability map. By importing the yields into a spreadsheet program and adding input
costs at each location, a map of profitability can show where profit is made or lost
within the field. It may become obvious after a few years that certain parts of the field
are money pits that do not support their operation. Putting these areas into grass and
farming the rest may increase profitability, all because of a yield map. In other areas,
some easily changed input or practice can be imposed that turns the picture around and
produces profit in areas that did not pull their weight previously.
Yield mapping also allows better on-farm testing than simply using a limited number of
measurements with a weigh-wagon. Replicated test strips with different treatments can
supply thousands of yield measurements. If the test is designed to cross several different
soil areas, the results from each area can be compared. With a little effort, the yields
from the treatment can be imported into a spreadsheet program and a simple statistical
test can compare the means of each treatment. The sheer number of yield measurements from
each treatment make this statistical comparison possible and valid. Although not as solid
as controlled University testing, it nevertheless provides the grower much more confidence
on the effect of the treatment than was possible using other testing methods. Inputs that
can be tested in this manner include varieties, fertilizers and fertilizer rates,
non-conventional products and amendments, fungicides, herbicides, insecticides-
practically anything that might have an influence on yield.
Can a yield map be used to direct a variable-rate fertilizer application?
Yield maps alone have not been found to consistently identify nutrient zones. This is
because yields are an integration of over 40 different environmental and cultural factors.
It is not reasonable to assume that yields are solely dependent on nutrients. In some
cases where nutrients and water availability are both strongly related to landscape, there
may be some correlation in some years. However, because of the effects of insects,
disease, weed pressure and cultural practices, it should not be assumed that yield and
nutrient levels go together. If yield maps, remote imagery, electrical conductivity
detector mapping, detailed soil maps and topography show similar patterns, then yield
mapping can help confirm that management zone boundaries are appropriate and in years
without pests may be used to support these zones. However, yield maps should not be used
alone to identify nutrient management zones.
SF-1176 (3), June 1999
|
