Objectives and Expected Outputs
develop variable rate technology for N fertilizer application by
trying different methods of obtaining a field map for variable
application of N; to determine the economic benefits of variable rate
technology for N; and to determine the change in potential nitrate
loading to the groundwater from variable field application of N
fertilizer compared to constant-rate application.
field variations of crop response to applied fertilizer N, estimates
of the spatial distribution of the entire N response curve, the
acquisition and testing of a variable rate fertilizer applicator
linked to a Global Positioning System, and mapping of the N soil test,
soil texture, soil organic matter, soil water regime and landform
shape are expected to meet the objectives outlined above.
Contribution Agreement, University
$70 K; 94-95: $70 K; 95-96: $37.4 K; Total: $177 K
The goal of the study is to
determine the feasibility of variable rate technology for N fertilizer
application, to maximize economic crop response while minimizing
environmental impacts on water quality. Specific objectives include:
To assess different
methods of obtaining the field map for variable application N of
Determine the economic
benefits of variable application of N fertilizers, and
Determine the change in
potential nitrate loading to the groundwater from variably applying N
fertilizer. Two sites were established in the spring of 1993 in Huron
Co. near Londesboro, Ontario on the farm of Bruce Shillinglaw.
Each site consisted of 4 adjacent blocks of no-till planted corn. Each block consisted of 2
treatments; (1) Fertilizer added (F) at 160 kg N ha-1, and (2)
No fertilizer N added (NF). Each treatment was 8 rows of corn with 75 cm
row spacing and a length of approximately 325 m. Spatial patterns of yield
with fertilizer added and yield with no fertilizer were obtained from
detailed hand harvesting (approx. 250 hand yield samples per field). Yield
patterns were also obtained using a commercial on-the-go yield sensor
attached to a combine. Soil cores were taken in a dense grid from each
field to obtain the spatial pattern of the soil N test. Extensive soil
sampling to a 90 cm depth was also carried out in the fall period to
obtain the spatial patterns of residual mineral soil N, and the subsequent
loss of N by leaching. All of the instrumentation and sampling was
referenced to a detailed elevation map of the site obtained from a
detailed survey of each site.
The data from the year of
results was used in the second year to construct two variable rate maps
for fertilizer N application for each of the two field sites (S1, S2). The
2 variable rate maps were based on 1) the N soil test and 2) a
differential yield map (fertilizer yield - check yield). Thus, in 1994
each of the two field sites had N fertilizer rate treatments consisting
of: check (0 Kg N ha-1), constant rate at 150 Kg N ha-1,
variable rate from soil test prediction, and variable rate from the
differential yield map. In 1995 the site was planted to soybeans and hand
sampled yields were taken in representative areas to examine any
carry-forward influences from the variable N treatments. In 1996 the site
was seeded to barley and fertilizer applied (67 kg N ha-1) in
the original blocks that had fertilizer in 1993. Hand sampled yields were
taken in the same locations as in 1993. Throughout the study soil samples
were taken to examine the influence of the fertilizer treatments on soil N
storage and losses. Soil solution samplers were installed to measure the
concentration of nitrate in drainage water.
Major findings of the study
A yield map based on one
fertilizer N application rate is not enough information to determine the
spatial pattern of N application for site specific management.
A new yield index called
the Delta yield dYF was developed to estimate the
spatial pattern of N fertilizer response. It is based on the difference
between yield with and without fertilizer N.
Yield measurements with
an on-the-go combine monitor were significantly correlated to hand yield
measurements, but the monitors may not be accurate enough to estimate dYF
. Robust spatial interpolation methods are needed for yield monitor
Site specific application
based on dYF used 40 $ ha-1 less N fertilizer, but
resulted in a decrease in crop yield valued at 36 $ ha-1.
Application based on soil test used 63 $ ha-1 less N
fertilizer, but lost 92 $ ha-1 from crop yield decline.
The soil N test map did
not stay constant with time. The dYF pattern was quite
constant and 1996 YF values for barley were similar to 1994
dYF values for corn.
Variable N application
significantly decreased subsequent soybean yields in areas with low or
no fertilizer N. This cost must be incorporated into economic models of
site specific management.
application resulted in nitrate in drainage water that exceeded the
Ontario drinking water objectives. However, at one site the water
standards were exceeded where no fertilizer N was applied. Drainage
losses increased in sites with high spatial variability.
Fertilizer N applied
using site specific methods was used more efficiently by the crop
compared to constant fertilizer N application. Drainage N losses were
reduced with site specific N application proportional to the decrease in
average fertilizer N applied.
relationship with recommended N and the spatial patterns of YF
measured at the two study sites indicate it may be better to only vary N
application for major changes in N requirements. A simple 3-rate (0 %,
50 %, 100 % of recommended) fertilizer N applicator was built.
May 16, 2011 10:36:04 PM