|
Background and
Objectives:
To spread their workload, reduce
spring tillage operations, and capitalize on lower fertilizer prices,
many producers in Manitoba prefer to apply nitrogen (N) fertilizer in
the fall rather than in spring. Southern Manitoba historically receives
fall rains that make fieldwork difficult and producers are interested in
applying N fertilizer as soon as possible after harvest, while soil
conditions are still favourable. Unfortunately, early fall applications
of ammonia fertilizers such as urea and anhydrous ammonia are expected
to form more nitrate prior to the soil freezing than fertilizer applied
later in the season (Nyborg et al. 1990), increasing the potential for
over-winter and early spring losses of NO3- via
leaching and denitrification (Yadvinder-Singh et al. 1994).
Therefore, Manitoba Agriculture, Food and Rural Initiatives currently
recommends that fall-applied N fertilizers be banded, as opposed to
broadcast, and that application be delayed until soil temperatures are
below 5°C (Soil Fertility Guide 2001).
Most of the historical research
with fall applied N fertilizer has been with broadcast and incorporated
methods of application. By comparison, banding or nesting nitrogen
fertilizers slows microbial activity within the soil (due to the high
pH, high concentrations of NH3+, and increased
osmotic pressure within the fertilizer band), lowering the risk of N
immobilization, slowing nitrification and reducing N losses by leaching
and denitrification(Harapiak et al.
1993; Yadvinder-Singh et al. 1994).
In western Canada, applying nitrogen in bands or nests has consistently
improved the efficiency of fall-applied fertilizers, with average yield
increases from fall-banded urea double that of fall broadcast and
incorporated urea (Ridley 1976; Ridley 1977; Racz 1979; Malhi and Nyborg
1984; Malhi et al. 1984; Malhi and Nyborg 1985; Malhi and Nyborg 1988;
Malhi and Nyborg 1990; Malhi et al. 1992).
However, in these studies grain yields and N uptake from fall-banded N
were still, on average, lower than spring-applied N. Recent work in
south-western Manitoba reported no differences in grain yield and total
crop N uptake between fall and spring-banded N in 2 of 3 years on a clay
loam soil, and in all 3 years on a drier fine sandy loam
(Bole et al. 1984; Kucey 1986; Malhi et al. 1992),
and when soil moisture contents in the fall and spring are low (Harapiak
1979; Ukrainetz 1984).
Landscape position is another
factor that will influence the efficiency of fall-applied N, through the
accumulation of water in lower lying areas of the field (Hanna et al.
1982). The
effects of landscape position are most significant during the early
spring period, when considerable ponding of snowmelt often occurs.
These flooded soil conditions greatly increase the potential of
denitrification losses. Numerous studies from Saskatchewan have
reported that denitrification rates were higher in the wetter footslope
and low level complexes than in the well-drained upper slope positions(Elliott and de Jong 1992; Pennock et al.
1992; van Kessel et al. 1993; Corre et al. 1995; Corre et al. 1996;
Farrell et al. 1996).
However, no experiments have focused on the impact of landscape
position on the loss of fall-banded N under Western Canadian
conditions.
Fertilizer additives such as
urease inhibitors and nitrification inhibitors have been used in
research trials to improve the efficiency of fall-applied N (Malhi et
al. 1992; Yadvinder-Singh et al. 1994).
Very limited work at three sites in Manitoba indicated that the addition
of N-Serve (nitrapyrin), a nitrification inhibitor, increased the
percent uptake of fall-banded N by 30%, when compared to fall-banded
urea without nitrapyrin (Ridley 1976).
However, the effectiveness of a double inhibitor containing N-(n-butyl)
thiophosphoric triamide (NBPT) and dicyandiamide (DCD) has not been
investigated in fall banding trials in Western Canada.
The objective of this project
was to evaluate the interactive effects of date of application,
landscape position, fertilizer additives, and weather and climate on the
efficiency of fall-banded N fertilizer in Manitoba. This project also
generated fundamental information on the effect of soil moisture and
temperature on the rate of ammoniacal N transformation into nitrate via
the nitrification process, and the amount of fall-applied fertilizer N
lost by leaching and denitrification after the ammoniacal N has
nitrified.
Procedure and Project
Activities:
Site Selection and Description
Field experiments were conducted
over two fertilization/growing seasons; fall 2000 to harvest 2001 (year
1), and fall 2001 to harvest 2002 (year 2). In total, seven small plot
sites were established throughout southern Manitoba (four intensive
sites and three satellite sites). In year 1, one intensive experiment
was established near the town of Kane on Red River-Osborne heavy clay
soil. In the second year of the project, two intensive sites were
situated on Red River-Osborne heavy clay soil near the towns of Kane and
Rosser, while a third intensive site was located on Newdale clay loam
soil at the Agriculture and Agri-Food Canada Brandon Research Centre’s
Phillips Research Farm. The Red River/Osborne and Newdale soil series
represent common soils in eastern and western Manitoba respectively and
provide two distinctly different potentials for N fertilizer loss due to
significant differences in soil texture, topography and climate. To
complement the intensively monitored experimental sites, three
additional satellite sites were established; one site near Oak Bluff in
year 1 and two sites near Oak Bluff and Sperling in year 2. The
satellite trials were all located on Red River-Osborne heavy clay soil,
and employed similar treatments to those of the intensive experiments.
However, only yield and N uptake of the crop was measured.
Experimental Design and Treatments
At the intensive sites, a
split-plot design was utilized with landscape position mainplots and
fertilization treatment subplots. Three of the four intensively
monitored sites were located in the relatively level lacustrine
landscape of the Red River Valley, with elevation differences of less
than 1 m per km within each site. Eight mainplots, consisting of four
plots in high areas and four plots in low areas, were selected
throughout the field using a Total Station and the Surfer grid and
contour software (Surfer 1997).
Each mainplot contained six, 2 x 10 m fertilization treatment subplots,
with all six treatments assigned at random to the subplots within each
mainplot. A more simplistic split-plot design was employed at the
satellite sites. At each satellite site, four complete replicated
blocks of fertilization treatments were placed into one high and one low
landscape position, based on their relative positions in the field to
one another.
The six fertilization treatments
were based on time of fertilizer application and included: early fall
application, early fall application with a double urease and
nitrification inhibitor (NBPT and DCD respectively), mid fall
application, late fall application, and a spring application (mid-row
banded at time of seeding). Nitrogen was applied as urea fertilizer
(46-0-0) banded at a rate of 80 kg N ha-1, with 40 cm
spacing, at a depth of 7.5 cm. Application of the urea was targeted for
September 15, September 30 and October 15 of each year. However, in
year 1, excess moisture caused a delay in application dates, at both
Kane and Oak Bluff, to September 29, October 12, and October 26. In
year two, treatments were applied at Brandon on September 15, October 1,
and October 15; at Rosser, Sperling and Oak Bluff on September 19,
October 1, and October 19; and at Kane on September 26, October 9, and
October 19. During fertilization, band rows were clearly marked with
small wooden stakes and pin flags to ensure precise sampling of the
banded areas, especially in the spring.
Crop Measurements
AC Barrie wheat (Triticum
aestivum) at a rate of 1.5 to 2 bu/acre was grown as the test crop
at all sites. MAP (11-52-0) was applied in the seedrow at a rate of 40
kg MAP (P2O5) ha-1. All pesticides
were applied at recommended rates based on the Manitoba Crop Protection
Guide using a 4 m bicycle sprayer, including a pre-seeding burn off with
Glyphosate. At midseason (50% heading), a 1 m x 2 row sample of above
ground plant tissue was hand harvested from each subplot and dry matter
yield (kg ha-1) was measured. At physiological maturity, a 3
m x 2 row sample of above ground plant tissue was harvested from each
subplot, dried, threshed and weighed for grain and straw yields. Tissue
samples collected at midseason and harvest were analyzed for total N
using a Leco CNS Analyzer.
Soil Sampling and Analyses
To characterize the overall N behaviour in each
subplot, the soil was sampled to 120 cm in mid-September, at seeding and
harvest. The background levels of soil NO3--N in
mid-September, prior to fertilization, are reported in Table 1. In
addition to sampling to 120 cm, separate soil samples of 0-15 and 15-30
cm were gathered three times in the fall (@
2 week interval) from the band zone and between the band zones, to
monitor the transformation of banded fertilizer. Weather and soil
conditions again dictated when soil samples were collected at the
individual sites. In year 1 at Kane, the third fall sampling period was
missed because of snow and frozen soil conditions. In year 2,
fertilized subplots were sampled three times at Kane and Brandon, but
excessive rainfall cancelled the second fall sampling period at Rosser.
Ground soil samples were extracted for water soluble nitrate and
nitrite, exchangeable ammonium, and urea nitrogen. Electrical
conductivity (EC) and pH of all 0-15 and 15-30 cm soil samples were
measured using a 2:1 water to soil extract, an Orion conductivity meter
and a Fisher Accumet pH meter.
Gravimetric soil moisture
contents of 0-7.5, 7.5-15 and 15-30 cm were measured weekly at all
intensive sites from early fall to freeze-up, and from early spring to
planting. Over the same period, soil temperatures were monitored
electronically every 15 minutes using a StowAway® Tidbit®
temperature probe placed directly into one of the fertilizer bands (7.5
cm depth) within each early fall application subplot. Rainfall data was
collected at all intensive sites using a tipping bucket rain gauge and a
HOBO® event driven data logger. Weather conditions,
including precipitation and aerial temperature were obtained from
Agrometeorological Centre of Excellence (ACE) weather collection devices
located near the individual intensive sites.
Results and Discussion
(intensive field sites only):
Crop
Data for Midseason
At midseason, total dry matter biomass was
significantly greater for high landscape positions than for low
landscape positions (Table 2). The high landscape positions also had
greater crop N uptake at midseason than the low landscape positions, but
due to the site year by landscape position interaction the LSD is not
reported. Midseason N uptake was significantly greater in the high
landscape positions than in the low landscape positions at two of the
four sites; Rosser (2001/02) and Brandon (2001/02) (data not
presented). Spring-banded N significantly increased both midseason dry
matter biomass and midseason N uptake, when compared to the fall-applied
fertilization treatments. Comparing the two early fall applications
with and without inhibitors, there were no substantial differences in
midseason biomass and/or midseason crop N uptake.
Crop Data for Harvest
At physiological maturity, mean
grain yield and total N uptake were 20 and 25% greater in the high
landscape positions than in the low landscape positions (Table 3).
However, due to the site year by landscape position interaction,
statistical analyses of the landscape position effects on grain yield
and total crop N uptake are reported at the individual intensive sites
only (Table 4). Grain yields were 265, 996, and 1283 kg ha –1
greater in the high landscape positions than in the low landscape
positions at Kane (2000/01), Rosser (2001/02) and Brandon (2001/02)
respectively. Crop N uptake was an average of 42 kg ha-1
greater in the high landscape positions than in the low landscape
positions over the same three intensive sites. At Kane (2001/02), grain
yield and total crop N uptake appeared to be greater in the low
positions than in the high positions, but the differences were not
significant. The higher grain yield and N uptake in the imperfectly
drained lower positions at Kane (2001/02) was likely due to a prolonged
dry period at this site during July and August, when the high landscape
positions became more drought stressed than the lower landscape
positions.
Spring and late fall-banded N
applications generally increased the mean grain yields and total N
uptake of the crop, compared to the other fall applications, with or
without inhibitors (Table 3). The LSD analysis for grain yield is not
reported because there was a landscape position by fertilization
treatment interaction. Further statistical analyses of the landscape
position by fertilization treatment interaction for grain yield
determined that in the high landscape position there were no significant
differences in crop response among fertilization treatments. However,
in the low positions, spring-banded N significantly increased grain
yields compared to early fall, mid fall and early fall with inhibitors.
Grain yields appeared to be slightly higher for spring-banded than for
late fall-banded N, but statistically they were not different. Grain
yield and total N uptake of individual fertilization treatments
consistently ranked higher in the high landscape positions than in the
low landscape positions (i.e. early fall in high vs. early fall in
low).
Similar trends are seen for both
grain yield increases and fertilizer N use efficiency (NUE), within the
respective landscape positions, as was the case for grain yield (Table
5). In the high landscape positions, there were no real differences in
increased grain yield and fertilizer NUE among the fertilization
treatments. In the low landscape positions, increases in grain yield
from late fall and spring-banded fertilization treatments were
significantly higher than those from early fall, mid fall and early fall
with inhibitors. The fertilizer NUE of late fall and spring-banded N in
the low landscape positions was 12 to 16% higher then that of the early
and mid fall-banded treatments, with and without inhibitors. Higher
soil moisture contents in the low areas during the fall and early
spring, combined with early fall applications of ammonia fertilizers,
increased the potential for over-winter and early spring losses of NO3-
via denitrification. Urea applied later in the season, when soil
temperatures were cool did not convert to nitrate as quickly and was
less subject to over-winter losses. In the high positions, prolonged
water saturation of the soil was not common, even in the spring, and
therefore the potential for N losses were much lower.
Correlation analysis showed that
the effect of date of application on relative grain yields was
significantly different for the high landscape positions compared to the
low landscape positions (Figure 1). Overall, the results suggest that
selection of suitable timing for application of fertilizer N to optimize
crop yields is much more critical for poorly drained fields, and for
poorly drained areas within a field, than for better drained land.
Soil Data
Landscape position did not have
a significant effect on the conversion of banded-urea to nitrate under
the moisture conditions present at the sites. Delaying the date of
application of fall-banded urea fertilizer into the late fall and the
presence of NBPT and DCD slowed nitrification and increased the percent
recovery of fertilizer N as NH4+-N in the soil
prior to freeze-up. Date of application, soil temperature on the date
of application, the accumulation of soil heat units (SHU) and
nitrification heat units (NHU) were all linearly related to the percent
of recovered fertilizer N as NH4+-N (Figures
A1,
A2,
A3,
A4).
Accumulated SHU and NHU best described the relationship with percent of
recovered fertilizer N as NH4+-N at the end of the
fall, with and without inhibitors. The percent recovery of fertilizer N
as NH4+-N prior to the winter was greater for the
early fall-banded urea with NBPT and DCD than for the early fall-banded
urea without inhibitors (Figure
A5).
In the high landscape positions,
the performance of fall-banded urea, relative to spring-banded urea, was
not affected by application date, soil temperature on date of
application, cumulative soil heat units or cumulative nitrification heat
units. This suggests that application date for fall-banded N is not a
factor in better-drained landscape positions and in well-drained
fields. However, in the low landscape positions, delaying application
until late in the fall, when soil temperatures had cooled to 5 or 6ºC,
greatly increased relative grain yields and total N uptake by the crop.
Soil temperature at time of fertilizer application gave the highest
correlation with relative grain yields in the low landscape positions (r
= -0.79**); date of application gave a slightly lower correlation (r =
0.66*). Soil heat units (SHU) and nitrification heat units (NHU)
accumulated from date of application until freeze-up gave inferior
correlations (r = -0.56ns and -0.49ns,
respectively).
Conclusions:
The knowledge acquired in this
project will enable farmers in Manitoba to extract more value from their
investment in N fertilizer while reducing the risk of environmental
problems associated with N loss (e.g., greenhouse gas emissions and
nitrate contamination of groundwater). Several of the key findings
include:
-
delaying the date of fall
banding and adding a nitrification and urease inhibitor (DCD and NBPT,
respectively) to urea slowed nitrification and increased the
proportion of fertilizer N remaining in the ammonium form at freeze-up
-
the proportion of fertilizer N
remaining in the ammonium form at freeze-up was successfully predicted
or monitored by a variety of tools, including soil temperature on date
of application, date of application, soil heat units and nitrification
heat units, in increasing order of accuracy
-
timing of banding (at planting
or during the fall) was not critical for well-drained areas in the
fields, but early fall banding was detrimental to fertilizer
efficiency in low areas, compared to banding in late fall or at
planting. Overall, the average wheat yield increase from early fall
banded N in low areas was 25% less than for spring applied N; in
well-drained areas of the fields the yield increase from early fall
banded N was at least as large as for spring banded N.
-
the benefits of adding urease
and nitrification inhibitors to early fall-banded urea was
inconsistent, showing a significant agronomic benefit in only 1 of 4
site years
Acknowledgements:
We would like to thank Western
Cooperative Fertilizers Limited, the Agri-Food Research and Development
Initiative (ARDI), Agriculture and Agri-Food Canada, and the Natural
Sciences and Engineering Research Council of Canada (NSERC) for the
financial and technical support to make this project possible. In
particular, we appreciate the efforts of Dr. Cindy Grant and Brian
Hadley at the Agriculture and Agri-Food Canada Brandon Research Centre
for establishing and maintaining the intensive field site near Brandon.
Special thanks to Bill Toews, Scott Corbett, Brad Erb and Bill Rempel
for the use of their land, and to the technical support staff and fellow
graduate students at the University of Manitoba for their contributions
to the project.
References:
Bole, J. B., Harapiak, J. T., Malhi, S. S. and Penny,
D. C. 1984. Regional and environmental influences of nitrogen use
efficiency. Proc. Alberta Soils Science Workshop. Edmonton, AB.
Corre, M. D., van Kessel, C. and Pennock, D. J. 1996.
Landscape and seasonal patterns of nitrous oxide emissions in a semiarid
region. Soil Sci. Soc. Am. J. 60: 1806-1815.
Corre, M. D., van Kessel, C., Pennock, D. J. and
Solohub, M. P. 1995. Ambient nitrous oxide emissions from different
landform complexes as affected by simulated rainfall. Commun. Soil Sci.
Plant Anal. 26: 2279-2293.
Elliott, J. A. and de Jong, E. 1992. Quantifying
denitrification on a field scale in hummocky terrain. Can. J. Soil Sci.
72: 21-29.
Farrell, R. E., Sandercock, P. J., Pennock, D. J. and
van Kessel, C. 1996. Landscape-scale variations in leached nitrate:
relationship to denitrification and natural N-15 abundance. Soil Sci.
Soc. Am. J. 60: 1410-11415.
Grant, C. A., Brown, K. R., Racz, G. J. and Bailey,
L. D. 2001. Influence of source, timing and placement of nitrogen on
grain yield and nitrogen removal of durum wheat under reduced- and
conventional-tillage management. Can. J. Soil Sci. 81: 17-27.
Hanna, A. V., Harlan, P. W. and Lewis, D. T. 1982.
Soil available water as influenced by landscape position and aspect.
Agron. J. 74: 999-1004.
Harapiak, J. T. 1979. Comparison of fall and spring
applied nitrogen. Proc. Alberta Soil Science Workshop. Lethbridge, AB.
Harapiak, J. T., Malhi, S. S., Campbell, C. A. and
Nyborg, M. 1993. Fertilizer N application practices. pp 251-313 in D. A.
Rennie et al., eds. Impact of macronutrients on crop responses and
environmental sustainability on the Canadian prairies. Canadian Society
of Soil Science, Ottawa, ON.
Kucey, R. M. N. 1986. Effect of fertilizer form,
method and timing of application on barley yield and N uptake under
dryland conditions in southern Alberta. Can. J. Soil Sci. 66: 615-621.
Malhi, S. S. and Nyborg, M. 1984. Inhibiting
nitrification and increasing yield of barley by band placement of
thiourea with fall-applied urea. Plant and Soil 77: 193-206.
Malhi, S. S. and Nyborg, M. 1985. Methods of
placement for increasing the efficiency of N fertilizers applied in the
fall. Agron. J. 77: 27-32.
Malhi, S. S. and Nyborg, M. 1988. Control of
nitrification of fertilizer nitrogen: effect of inhibitors, banding and
nesting. Plant and Soil 107: 245-250.
Malhi, S. S. and Nyborg, M. 1990. Evaluation of
methods of placement for fall-applied urea under zero tillage. Soil
Tillage Res. 15: 383-389.
Malhi, S. S., Nyborg, M., Harapiak, J. T. and Penny,
D. C. 1984. Efficiency of fall-applied N as influenced by methods of
placement and date of fall application. Proc. Alberta Soil Science
Workshop, Agriculture Secretary, University of Alberta. Edmonton, AB.
Malhi, S. S., Nyborg, M., Solberg, E. D. and Heaney,
D. J. 1992. Fall compared to spring application of nitrogen fertilizers
in Alberta. Technical Bulletin 8E. Research Branch Agriculture Canada,
Lacombe, AB.
Nyborg, M., Malhi, S. S. and Solberg, E. D.
1990. Effect of date of application on the fate of 15N-labelled urea and
potassium nitrate. Can. J. Soil Sci. 70: 21-31.
Pennock, D. J., van Kessel, C., Farrell, R. E. and
Sutherland, R. A. 1992. Landscape-scale variations in denitrification.
Soil Sci. Soc. Am. J. 56: 770-776.
Racz, G. J. 1979. Losses of fertilizer nitrogen as
affected by time and method of application - Manitoba. Proc. Alberta
Soil Science Workshop. Lethbridge, AB.
Ridley, A. O. 1976. Efficiency of nitrogen
fertilizers, time and method of placement. Proc. Annual Conference of MB
Agronomists. Winnipeg, MB.
Ridley, A. O. 1977. Nitrogen fertilizers, time and
method of placement. Proc. Manitoba Soil Science Society Annual General
Meeting. Winnipeg, MB.
Soil Fertility Guide. 2001. Manitoba Agriculture and
Food, Winnipeg, MB.
Surfer. 1997. Surfer TM gridding and contouring
software. Version 6.0.4. Golden Software, Inc., Boulder, Colorado
U.S.A.
Ukrainetz, H. 1984. Fertilizer placement for maximum
effectiveness. Proc. Alberta Soil Science Workshop. Edmonton, AB.
van Kessel, C., Pennock, D. J. and Farrell, R. E.
1993. Seasonal variations in denitrification and nitrous oxide evolution
at the landscape scale. Soil Sci. Soc. Am. J. 57: 988-995.
Yadvinder-Singh, Malhi, S. S., Nyborg, M. and
Beauchamp, E. G. 1994. Large granules, nests or bands: Methods of
increasing efficiency of fall-applied urea for small cereal grains in
North America. Fert. Res. 38: 61-87.
|
HomeContact
UsAbout
ManitobaDepartmentsLinksPrivacy 
