Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Published May, 2002

Corn Yield Response to Nitrogen Fertilizer Timing and Deficiency Level


Peter C. Scharf,* William J. Wiebold, and John A. Lory

ABSTRACT and positively impact water quality. Understanding the


Nitrogen fertilizer is typically applied to corn (Zea mays L.) shortly effect of delayed N application on yield is critical to the
before planting, but there are several reasons why later N applications use of any of these management tools.
may be of interest: to spread work away from the busy planting season, Fertilizer N recovery by the crop may sometimes be
to avoid the frequent wet field conditions in spring, to reduce or greater when N application is delayed compared with
remedy in-season N loss in wet years, or to allow use of in-season application at planting (Russelle et al., 1983; Jokela and
diagnostic tools. One of the obstacles to the use of later N applications Randall, 1997). This is probably due to greater exposure
is the fear that irreversible yield loss will occur due to N stress. Our
of N applied at planting to a range of possible loss
objective was to evaluate the yield impact of delaying N applications
until the late vegetative growth stages and as far as silking. We con-
processes (immobilization, leaching, denitrification, and
ducted a total of 28 experiments with timing of a single N application clay fixation) at a time when N uptake rates are rela-
as the experimental treatment. We found little or no evidence of tively low. Rate of N uptake as the corn plant develops is
irreversible yield loss when N applications were delayed as late as affected by weather, planting date, and time of fertilizer
stage V11, even when N stress was highly visible. There was weak application but is generally greatest between V8 and
evidence of minor yield loss (about 3%) when N applications were silking (Russelle et al., 1983). When N fertilizer applica-
delayed until stage V12 to V16. Only 3 of the 28 experiments had N tions were delayed until V16, the highest rate of N
applications later than V16—all were at silking and relative yields uptake was generally delayed until after silking (Rus-
were 0.71, 0.89, and 0.95. Though full yield was not achieved when
selle et al., 1983). This would suggest that applying N
N applications were delayed until silking, yield was still highly respon-
sive to N application at this stage—yield response exceeded 2.2 Mg
fertilizer until at least the silking stage may be a reason-
ha⫺1 in all three experiments. able management option.
Most of the research on N application after planting
has compared applications at or before planting to side-
dress applications at growth stage V8 or earlier. These
N itrogen fertilizer is typically applied to corn
shortly before planting, but there are several rea-
sons why later N applications may be of interest: to
applications can typically be accomplished with tractor-
drawn equipment, whereas later applications require
spread work away from the busy planting season, to avoid high-clearance vehicles. In many cases, there are no
the frequent wet field conditions in spring, to reduce or yield differences between preplant and sidedress N ap-
remedy in-season N loss in wet years, or to allow use plications (Jokela and Randall, 1989; Roth et al., 1995).
of in-season diagnostic tools. Sidedress N applications sometimes give small yield in-
Planting time is one of the busiest and most time- creases (Bundy et al., 1992; Reeves and Touchton, 1986;
sensitive periods of the year for grain producers and is Welch et al., 1971) or small yield decreases (Stecker et
an inconvenient time to apply N fertilizer. Moving N al., 1993). However, even when grain yield is not af-
applications away from this busy period is the main fected, total dry matter production may be reduced with
motivation for fall applications of N. Unfortunately, fall late sidedress N applications (Jokela and Randall, 1989).
N application creates a substantial risk of losing N and Delayed N applications may not be appropriate for corn
yield (Blackmer, 1996). Nitrogen applications after plant- silage production.
ing offer an alternative way to move N applications Investigation of N applications later than the V8
away from the busy planting season but are seldom used growth stage has been limited. The studies that have
by producers. There are many reasons for this, one of been conducted have been confined mostly to irrigated
which is the fear that irreversible yield loss will occur systems. Irrigation presents a readily available delivery
due to N stress, especially if weather causes delays in system for N at stages later than V8 and also creates
the planned application timing. risk that N fertilizer will be leached out of the root zone,
Delaying application of some or all N fertilizer until especially on sandy soils.
after planting may allow for precise diagnosis of N For irrigated corn grown on sandy soils, sidedress
needs, by either in-season soil testing (Blackmer et al., applications tend to produce higher yields than preplant
1989), sensing crop color (Varvel et al., 1997; Blackmer applications (Bundy et al., 1983; Rehm and Wiese,
et al., 1996), or estimating weather effects on soil N 1975). Supplemental N applications later than normal
availability (Honeycutt, 1994). Application rates that sidedress time can produce yields higher than those
precisely match crop needs could result in less residual obtained with preplant or sidedress applications alone
soil NO3⫺ available for leaching (Andraski et al., 2000) (Evanylo, 1991; Gascho et al., 1984; Rehm and Wiese,
1975). Jung et al. (1972) observed equivalent yields
Dep. of Agron., 210 Waters Hall, Univ. of Missouri, Columbia, MO when a single N application was made from 5 to 8 wk
65211. Contrib. from the Missouri Agric. Exp. Stn. Journal Ser. No. after planting, but yields began to decline when N appli-
13 150. Received 25 June 2001. *Corresponding author (scharfp@ cation was delayed until the ninth week or later. They
missouri.edu).
did not report growth stages, but 5 to 8 wk after planting
Published in Agron. J. 94:435–441 (2002). is likely to correspond to the V5 to V12 growth stages.
435
436 AGRONOMY JOURNAL, VOL. 94, MAY–JUNE 2002

results are not reported due to drought. Timing of N fertilizer


was the experimental variable. A single application of 180 kg
N ha⫺1 was applied either at planting, growth stage V7 (Ritchie
et al., 1993), V14, or silking. Nitrogen was hand-applied as
ammonium nitrate (NH4NO3 ). Care was taken to apply N
below leaves to avoid burn.
The experimental design was a randomized complete block
with four replications. Four hybrids (DeKalb 668, Pioneer
3163, Pioneer 3394, and Ciba 4575) were arranged in a com-
plete factorial design with the N-timing treatments. Reported
means for each N-timing treatment are averaged over hybrids.
No-tillage production practices were used. Plots were four
rows wide with dimensions of 3 by 7.5 m. Planting population
was 60 000 seeds ha⫺1.
Plots were end-trimmed to 6 m long before harvest. The
center two rows of each plot were harvested with a plot com-
bine. Yields were corrected to a moisture content of 150 g kg⫺1.
Fig. 1. Locations of N-timing experiments.
Yield when N was applied at planting was defined as the
standard yield for these experiments. Relative yield for later
Conclusions are similar for irrigated corn on heavy- applications was calculated by dividing mean yield for N ap-
textured soils—delaying N applications as late as V16 plied at a given time by the yield with N applied at planting.
usually produces yields equivalent to or greater than
those with earlier applications. On irrigated silty clay On-Farm Experiments
loam soils, Olson et al. (1986) measured higher grain Experiments were established in production corn fields in
yields (averaged over 15 yr) when N was applied at the Missouri in 1997, 1998, and 1999 with a total of 25 site-years.
11- to 12-leaf stage than when it was applied at planting, Experiments were located in all four major corn-producing
and Russelle et al. (1983) measured higher yields when regions of Missouri (Fig. 1) and represented a broad cross
N was applied at V8 or V16 than when it was applied section of soils, hybrids, climate, and management practices
at planting or V4. In contrast, Binder et al. (2000) found that are typical for corn production in Missouri (Table 1). Our
that corn yields declined when N applications were de- findings should thus apply to a broad population of corn fields
layed—the earliest stage associated with significant in Missouri and potentially to other states with similarities in
yield reduction was V6 in the first year of the study and climate and soils.
Timing of N fertilizer was the experimental variable. A
VT in the second year. single application of 225 kg N ha⫺1 was applied either at
Late N applications in irrigated systems can be moved planting, growth stage V6 (Ritchie et al., 1993), or two or
into the soil and the root zone by irrigation water. Late three later times. Crop growth stage was determined at the
N applications in dryland production systems might be- time of application. The latest N application timing at any
have differently, for example, being ineffective when experimental location was V15.5. Additional plots received N
rainfall is limited after N applications. Only a few re- rates from 0 to 335 kg N ha⫺1 at planting to characterize the
ports of N applications later than V8 in dryland systems magnitude of the yield response to N and verify the sufficiency
are available. Randall et al. (1997) found equivalent of the 225 kg N ha⫺1 rate. Plots were four rows wide and 12 m
corn yields with all N fertilizer applied at planting or long. At planting and growth stage V6, ammonium nitrate
with 30% applied at planting and 70% delayed until was surface-applied with a custom-built hand-push Gandy
spreader equipped with drop tubes, which was calibrated im-
V16. Miller et al. (1975) obtained a large yield response
mediately before treatment application. Later applications
to N and equivalent yields regardless of whether N was were hand-applied ammonium nitrate. The experimental de-
applied in May, June, July, or a May–June split. sign was completely randomized with three replications. All
In summary, most trials with N applications later than cultural practices other than N fertilizer application and har-
V8 have found little or no evidence of grain yield reduc- vest were performed by cooperating producers.
tion with delayed N applications. Jung et al. (1972) ob- Minolta SPAD 502 chlorophyll meter readings were taken
served yield reductions when N applications were de- from midway along the uppermost collared leaf each time N
layed until 9 wk after planting on an irrigated sandy was applied. Readings were taken on 10 plants per plot in 10
soil, and Binder et al. (2000) observed yield reductions zero-N plots and in 10 plots that had received a high rate of
when N applications were delayed until V6 to VT, de- N (ⱖ225 kg N ha⫺1 ) at planting. Chlorophyll meter readings
pending on the year. were not taken in the two 1999 experiments nor at V14 to
Very few studies have been conducted with N applica- V15 in the four 1997 experiments that were fertilized at those
times. In both 1997 and 1998, the latest chlorophyll meter
tions later than V8 under dryland conditions. Our objec- readings were taken at V13.
tive was to evaluate the yield impact of delaying N The center two rows of each plot were hand-harvested (1.5
applications for dryland corn until the late vegetative by 6 m). Harvest population was also determined in this area.
growth stages and as far as silking. Grain was shelled and weighed, and grain moisture was deter-
mined with a hand-held moisture meter. Yields were corrected
MATERIALS AND METHODS to a moisture content of 150 g kg⫺1.
A quadratic-plateau function was used to describe yield
On-Station Experiments response to N rate at planting and to obtain a plateau yield.
Experiments were established at the Bradford Research Least-squares optimization was performed with PROC NLIN
Farm in Boone County, MO, from 1995 to 1998, but 1997 in SAS to obtain these functions. Relative yield was then
SCHARF ET AL.: CORN YIELD RESPONSE TO N FERTILIZER TIMING AND DEFICIENCY LEVEL 437

Table 1. Background information for N-timing experiments.


Soil
Planting Previous Mineral N to 90 cm
Location Year Series Texture† Great Group Hybrid date Tillage‡ crop§ Manure¶ at planting
kg ha⫺1
1 1995 Putnam sil Albaqualfs Four# 23 May NT SB None NA††
2 1996 Mexico sil Epiaqualfs Four# 20 May NT SB None NA
3 1998 Mexico sil Epiaqualfs Four# 11 May NT SB None NA
4 1997 Mexico sil Epiaqualfs Pioneer 32-31 6 May NT SB Hog lagoon 152
effluent
5 1997 Higginsville sil Argiudolls Cargill 7997 9 May T Corn Hog lagoon 157
effluent
6 1997 Contrary sil Eutrudepts Stine 9706 13 May T SB Hog pit 160
slurry
7 1997 Libourn fsl Fluvaquents Pioneer 3335 9 Apr. T SB None 58
8 1997 Sibley sil Argiudolls Northrup King 7070 9 May NT DC SB None 73
9 1997 Putnam sil Albaqualfs Pioneer 3394 15 May NT SB None 51
10 1997 Luton c Haplaquolls Asgrow 801 12 May NT SB None 92
11 1997 Contrary sil Eutrudepts Pioneer 3335 13 May NT SB None 70
12 1997 Mexico sil Epiaqualfs Pioneer 3260 6 May T SB Hog lagoon 255
effluent
13 1997 Putnam sil Albaqualfs Cargill 7997 17 May T SB None 84
14 1997 Moville sil Fluvaquents Garst 8342 29 Apr. T SB None 42
15 1997 Caruthersville sl Udifluvents Pioneer 3335 8 Apr. T Corn None 46
16 1997 Sibley sil Argiudolls DeKalb 626 29 Apr. T Corn Hog pit 604
slurry
17 1998 Mexico sil Epiaqualfs Pioneer 33Y09 12 May T SB Hog lagoon 246
effluent
18 1998 Mexico sil Epiaqualfs Pioneer 31K62 25 Apr. T SB Hog pit 175
slurry
19 1998 Mexico sil Epiaqualfs Pioneer 33K61 24 Apr. NT SB None 64
20 1998 Dundee sil Endoaqualfs Pioneer 3326 31 Mar. T SB None 4
21 1998 Putnam sil Albaqualfs Patriot 7170 18 May NT SB None 21
22 1998 Nodaway sil Udifluvents DeKalb 668 5 May T Wheat None 10
23 1998 Monona sil Hapludolls Unknown 1 May NT SB None 22
24 1998 Monona sil Hapludolls Pioneer 33Y09 18 Apr. NT SB Hog lagoon 29
sludge 1997
25 1998 Knox sil Hapludalfs Unknown‡‡ 21 Apr. NT SB None 1
26 1998 Caruthersville sl Udifluvents Pioneer 3394 17 Apr. T Corn None 54
27 1999 Farrenburg fsl Hapludalfs Unknown‡‡ 7 Apr. T DC SB None 21
28 1999 Wiota sil Argiudolls Pioneer 33A14 12 Apr. NT SB None 96
† sil, silt loam; sl, sandy loam; fsl, fine sandy loam; c, clay.
‡ T, tilled; NT, not tilled.
§ SB, soybean; DC SB, double-crop soybean.
¶ Manure applied within 6 mo before planting, except as noted.
# Four corn hybrids planted: DeKalb 668, Pioneer 3163, Pioneer 3393, and Ciba 4575.
†† NA, not available.
‡‡ White corn.

calculated by dividing mean yield for N applied at a given 12, 16, and 17 had high levels of soil mineral N at planting
time by the plateau yield for that experiment. Magnitude of time (Table 1), and we would have predicted no need
yield response was calculated by subtracting mean yield of the for additional N at these locations. Average yield of
zero-N plots from the mean yield for N applied at a given time. nonresponsive experiments was 10.4 Mg ha⫺1, which
Linear regression analyses were performed using PROC
was not different than the average plateau yield of re-
REG in SAS. Quadratic-plateau and linear-plateau regres-
sions were performed using PROC NLIN in SAS. When time sponsive experiments. Rainfall distribution and amount
of N application was used as the independent variable in was in general good for these experiments, with some
regression analysis, it was defined as the vegetative stage (from drought stress in July 1997 and July–August 1999.
0–15.5), with silking assigned a value of 20 (because there are When experiments were analyzed individually using
approximately 20 leaves and silking occurs shortly after the linear regression, N application time was a significant
emergence of the last leaf and the tassel). A t-test was used (␣ ⫽ 0.05) predictor of yield for two of the three on-
to estimate the probability that mean relative yield ⫽ 1.00 for station experiments (Locations 1 and 3) and 2 of the 25
several groups of data with later N application times. on-farm experiments (Locations 26 and 27). In all four
experiments with a significant response, yield decreased
RESULTS AND DISCUSSION as N application delay increased. The frequency of yield
loss with delayed N application was greater in the on-
Yield as a Function of Nitrogen
station experiments because longer delays were included
Application Timing
in the experiments. When the latest N application time
Average plateau yield of these 28 experiments was (silking) was not included in the analysis, N application
10.3 Mg ha⫺1, and average yield response to N fertilizer time was not a significant predictor of yield for any of
was 3.1 Mg ha⫺1. This value includes nonresponsive the on-station experiments.
sites. Corn yield responded to added N fertilizer in 22 A similar approach was used to examine the possibil-
of 28 experiments. The nonresponsive locations were 5, ity that delaying N applications would delay develop-
6, 10, 12, 16, and 17, all of which received manure in ment and result in increased grain moisture at harvest.
the year of the study, except for Location 10. Locations Regression of grain moisture against time of N applica-
438 AGRONOMY JOURNAL, VOL. 94, MAY–JUNE 2002

deviations are purely experimental error (most data


points represent mean yield of only three plots), or
whether there is a balance between true yield increases
and yield decreases as N applications are delayed.
Lower variance of relative yield with N applications
at planting (Fig. 2) suggests the latter. However, the
standard deviation of relative yield for nonresponsive
locations is 0.055, and for responsive locations, it is 0.062
(excluding data from N applications at silking) (Fig.
2). Because the standard deviation of relative yield for
nonresponsive locations should represent pure experi-
mental error with no treatment effect, these values sug-
gest that most of the variability in relative yield of re-
sponsive locations is due to experimental error rather
than treatment effects.
Another way to examine the data is to use t-tests. At
growth stages V12 and later, more data points in Fig. 2
are below 1.00 than are above it. A t-test shows that
there is a high probability of a true yield reduction (P ⫽
0.012 for H0: relative yield ⫽ 1.00) for all N applications
delayed until V12 or later. The mean relative yield for
these data is 0.95. If only data for N applications from
V12 to V16 are considered, P ⫽ 0.042 and mean relative
yield is 0.97. Mean relative yield when N was applied
at silking was 0.85, with values of 0.71, 0.89, and 0.95
Fig. 2. Relative yield as a function of timing of a single N application.
for the three experiments receiving N applications at
Conclusions and models are nearly identical for (a) all 28 experiments silking.
or (b) the 22 N-responsive experiments. Simple linear models are Overall, our data suggest that risk of a small but real
not statistically significant. (a) Quadratic-plateau or linear-plateau yield reduction occurs when N applications are delayed
models are highly significant (P ⫽ 0.001) but are not significant until the range V12 to V16 and that the risk and the
when data from N applied at silking are omitted. Pre-emergence
N applications were randomly assigned vegetative stage values yield reduction are larger when N applications are de-
between 0 and 1 to make the data easier to see. layed until silking. We found little or no evidence of
irreversible yield loss when N applications were delayed
tion was significant (␣ ⫽ 0.10) at only one location, as late as stage V11. This agrees with the findings of
suggesting that N timing had a minimal effect on grain Olson et al. (1986), Russelle et al. (1983), and Jung et
moisture. al. (1972) but contrasts with one of the site-years of
When data from all 28 experiments are pooled, there Binder et al. (2000) where yield losses were observed
is little evidence of yield reductions when N applications when N applications were delayed only until V6.
were delayed as late as growth stage V16 (Fig. 2). Linear Though full yield was not achieved in our experiments
regression analysis did not indicate a significant trend when N applications were delayed until silking, yield was
in relative yield when N was applied between planting still highly responsive to N application at this stage—yield
and V16 for all sites (P ⫽ 0.72) or for responsive sites response exceeded 2.2 Mg ha⫺1 in all three experiments.
only (P ⫽ 0.63) (Fig. 2b). Neither was linear regression A N rate of 225 kg N ha⫺1 was sufficient to maximize
significant (P ⫽ 0.23) with data from N applications at yield at all locations except one. Data from Location
silking included. 21 are not included in any of the figures or collective
Either linear-plateau or quadratic-plateau regression statistical analyses. The soil at this location was poorly
analysis of all data in Fig. 2 produces significant models drained, and approximately 30 cm of precipitation fell
(P ⫽ 0.001) indicating a drop in relative yield with delayed during May and June, resulting in apparently severe
N application. The break point in the linear-plateau denitrification loss of N. Yield response to N applied
model is at stage V15, separating only the three N appli- at planting was linear up to 335 kg N ha⫺1, the highest
cations at silking as being different from all others. This N rate applied, so calculating relative yield from the
model essentially becomes a t-test comparing N applica- 225 kg N ha⫺1 applications at different timings was prob-
tions at silking to all others. The quadratic-plateau lematic. Regression of yield vs. N application time (from
model (Fig. 2a) predicts yield loss beginning at V10, planting to V12) for 225 kg N ha⫺1 was not significant
with 3.5% yield loss by V15 and 14% yield loss by at this location. Average yield response to 225 kg N
silking. Most of the data fall in the range from planting ha⫺1 was 4.2 Mg ha⫺1.
to V15, where predicted timing effects are small or none,
limiting the suitability of this analysis for describing Yield and Yield Response as a Function
the data. of Nitrogen Stress
Many observations of relative yield deviated substan- Binder et al. (2000) report that yield losses with de-
tially from 1.00 (Fig. 2). We don’t know whether these layed N applications could be predicted by both applica-
SCHARF ET AL.: CORN YIELD RESPONSE TO N FERTILIZER TIMING AND DEFICIENCY LEVEL 439

Fig. 3. Degree of N stress observed, as indicated by relative SPAD chlorophyll meter readings, was not a significant predictor of relative yield
achieved when N fertilizer was applied at the time of the reading. This was true regardless of the time of the reading and N fertilizer application
within the range from growth stage V6 to V13. For all of the above groups, P ⬎ 0.40 for simple linear regression. Chlorophyll meter readings
were not taken after V13 through a few experiments received later N fertilizer applications.

tion timing and the magnitude of N stress that the crop resulted in a 6.2 Mg ha⫺1 yield response and the attain-
was experiencing at the time of the delayed application. ment of full yield. The equation developed by Binder
They used relative chlorophyll meter reading (the read- et al. (2000) predicts relative yield of 0.90 for corn with
ing of unfertilized plots divided by the reading from a relative chlorophyll meter reading of 0.72 at stage V13.
well-fertilized plots) as their measure of N stress. In 22 While relative chlorophyll meter readings were not
experiments for which we have chlorophyll meter data useful for predicting the magnitude of irreversible yield
(from V5.5 to V13 only), neither N application timing, loss in these experiments, they were useful for predicting
relative chlorophyll meter reading, their squares, or the magnitude of yield gain in response to N application
their cross product is a significant predictor of relative (Fig. 5). This supports earlier research documenting that
yield in a regression model. Sometimes real effects are chlorophyll meters can be useful for predicting response
lost in a coarse model like this, so we examined finer to N (Piekielek and Fox, 1992). Although relative chlo-
subsets of data. When grouped by stage of N application, rophyll meter readings sometimes varied substantially
there is little indication that relative chlorophyll meter over time within an experiment, this variation did not
reading is useful for predicting relative yield (Fig. 3). correspond to differences in relative yield or yield gain
Similarly, when data are grouped by N stress, as indi- to N applied at these times within a location. The signifi-
cated by relative chlorophyll meter reading, there is cant relationship that we observed between relative
little evidence that delaying N application reduces yield chlorophyll meter reading and yield gain was the result
at any stress level (Fig. 4). of differences between locations, rather than between
The most extreme combination of N stress and timing times within a location.
occurred at Location 7 where relative chlorophyll meter A relative chlorophyll meter reading ⱕ0.95 has been
reading decreased from 0.92 at V6 to 0.88 at V8 to suggested as indicating N deficiency and a high likeli-
0.72 at V13. Full yield was achieved with a single N hood of yield response to additional N fertilizer (Varvel
application at V13 in this experiment—relative yield et al., 1997). While relative chlorophyll meter readings
was 1.00, yield was 13.1 Mg ha⫺1, and yield response to were generally high in the six nonresponsive site-years,
N was 6.5 Mg ha⫺1. This is similar to the observation we observed values of ⱕ0.95 in 5 of 17 samplings. The
of Miller et al. (1975) that N application to severely lowest value observed was 0.93. Five of the six nonre-
N-deficient and previously unfertilized corn at tasseling sponsive site-years had received manure applications,
440 AGRONOMY JOURNAL, VOL. 94, MAY–JUNE 2002

Fig. 4. Timing of a single N fertilizer application from V6 to V13 did not significantly influence relative yield, regardless of N stress level observed
at the time of fertilizer application (as indicated by relative SPAD chlorophyll meter reading). For all of the above groups, P ⬎ 0.25 for
simple linear regression.

so manure N mineralization after chlorophyll meter relative chlorophyll meter reading decreased over time
readings may account for the lack of yield response. for 1997 data (P ⫽ 0.15) and increased over time for
Relative chlorophyll meter reading was not necessar- 1998 data (P ⫽ 0.09). This may indicate that weather
ily consistent over time at a given location, but for our plays a role. Relatively wet weather around V6 in 1998
full data set, there was no evidence (P ⫽ 0.91) that led to greater apparent N stress and lower relative chlo-
relative chlorophyll meter reading tended to increase rophyll meter readings than in 1997 at the same stage.
or decrease as the season progressed. Thus, a reading This stress may have lessened as soils dried and mineral-
at any given time may represent an equally valid snap- ization resumed.
shot of crop N status. There was weak evidence that
Extrapolation of Our Results
to Other Environments
Climate may affect the relative risk of yield loss with
delayed N applications. Reduced early season growth,
as might be expected with delayed N applications, might
be expected to have less effect in areas with longer
growing seasons. Yield losses recorded in the literature
in association with delayed N are too few to evaluate
this idea, but the observation of Russelle et al. (1983)
that delayed N gave more favorable results with early
planted than late-planted corn would tend to support
it. It may also help to explain the attainment of full yield
in Kentucky (Miller et al., 1975) with N applications to
severely N-stressed corn at tasseling while we generally
observed yield reductions in central Missouri when ap-
Fig. 5. SPAD chlorophyll meter reading of unfertilized plots, relative
plications were delayed that long.
to the reading of well-fertilized plots, was a highly significant pre- Concern is sometimes expressed that late N applica-
dictor of the magnitude of corn yield response to N fertilizer. tions will not be effective in dry weather due to posi-
SCHARF ET AL.: CORN YIELD RESPONSE TO N FERTILIZER TIMING AND DEFICIENCY LEVEL 441

tional unavailability. Most of our 1997 experimental lo- Proc. Integrated Crop Manage. Conf., Ames, IA. 19–20 Nov. 1996.
Iowa State Univ. Ext., Ames.
cations had ⬍3 cm of rain in July and experienced Blackmer, A.M., D. Pottker, M.E. Cerrato, and J. Webb. 1989. Corre-
considerable water stress. Nine of these experiments lations between soil nitrate concentrations in late spring and corn
received surface N applications in July (as late as 9 July) yields in Iowa. J. Prod. Agric. 2:103–109.
that produced yields as high as those with earlier appli- Blackmer, T.M., J.S. Schepers, G.E. Varvel, and G.E. Meyer. 1996.
cations. Analysis of aerial photography for nitrogen stress within corn fields.
Agron. J. 88:729–733.
Bundy, L.G., T.W. Andraski, and T.C. Daniel. 1992. Placement and
timing of nitrogen fertilizers for conventional and conservation
SUMMARY tillage corn production. J. Prod. Agric. 5:214–221.
Bundy, L.G., K.A. Kelling, D.R. Keeney, and R.P. Wolkowski. 1983.
Averaged over 28 N-timing experiments, we saw no Improving nitrogen efficiency on irrigated sands using a nitrifica-
evidence of yield reduction when N applications were tion inhibitor. p. 165. In 1983 agronomy abstracts. ASA, Madi-
delayed as late as V11, weak evidence of small yield son, WI.
reductions (3%) when N applications were delayed until Evanylo, G.K. 1991. No-till corn response to nitrogen rate and timing
V12 to V16, and moderate yield reductions (15%) when in the Middle Atlantic Coastal Plain. J. Prod. Agric. 4:180–185.
Gascho, G.J., J.E. Hook, and G.A. Mitchell. 1984. Sprinkler-applied
N applications were delayed until silking. The risk of and side-dressed nitrogen for irrigated corn grown on sand. Agron.
yield loss associated with N applications delayed into J. 76:77–81.
the mid- to late-vegetative stages of corn growth appears Honeycutt, C.W. 1994. Linking nitrogen mineralization and plant
to be acceptable and may be less than the risk associated nitrogen demand with thermal units. p. 49–79. In J.L. Havlin and
J.S. Jacobsen (ed.) Soil testing: Prospects for improving nutrient
with fall N applications. Agronomically, delaying N ap- recommendations. SSSA Spec. Publ. 40. SSSA, Madison, WI.
plications to spread work load away from planting sea- Jokela, W.E., and G.W. Randall. 1989. Corn yield and residual soil
son, or allow in-season assessment of N need, appears nitrate as affected by time and rate of nitrogen application. Agron.
to be feasible. Extrapolation of these conclusions into J. 81:720–726.
Jokela, W.E., and G.W. Randall. 1997. Fate of fertilizer nitrogen as
regions with a shorter growing season should be done affected by time and rate of application on corn. Soil Sci. Soc. Am.
with caution. When weather causes unplanned delays J. 61:1695–1703.
in N application, or when severe in-season N loss occurs, Jung, P.E., Jr., L.A. Peterson, and L.E. Schrader. 1972. Response of
our data suggest that economical response to N is likely irrigated corn to time, rate, and source of applied N on sandy soils.
Agron. J. 64:668–670.
to be obtained until silking. Availability of high-clear- Miller, H.F., J. Kavanaugh, and G.W. Thomas. 1975. Time of N appli-
ance equipment remains a factor limiting delayed N cation and yields of corn in wet, alluvial soils. Agron. J. 67:401–404.
applications for many producers, but availability of this Olson, R.A., W.R. Raun, Yang Shou Chun, and J. Skopp. 1986. Nitro-
equipment appears to be increasing. gen management and interseeding effects on irrigated corn and
sorghum and on soil strength. Agron. J. 78:856–862.
Piekielek, W.P., and R.H. Fox. 1992. Use of a chlorophyll meter to
ACKNOWLEDGMENTS predict sidedress nitrogen requirements for maize. Agron. J. 84:
59–65.
We thank Emmett and Jim Burke; John Waggoner; David Randall, G.W., T.K. Iragavarapu, and B.R. Bock. 1997. Nitrogen
Bentley; Russell Flair; Fred and Ryland Utlaut; Bob Zeysing; application methods and timing for corn after soybean in a ridge-
Melvin Keehart; David Copeland; Johnny Haer; Max Kurtz; tillage system. J. Prod. Agric. 10:300–307.
Duane Biermann; Karl Noellsch; Dale Goers; Randall Smoot; Reeves, D.W., and J.T. Touchton. 1986. Subsoiling for nitrogen appli-
cations to corn grown in a conservation tillage system. Agron.
Alan Adam; Larry Abell; Kenny Brinker; and Tom, Bill, and
J. 78:921–926.
John Becker for their interest, cooperation, comments, and Rehm, G.W., and R.A. Wiese. 1975. Effect of method of nitrogen
generosity in allowing us to work on their farms. For help application on corn (Zea mays L.) grown on irrigated sandy soils.
with field work and data organization, we are grateful to Tom Soil Sci. Soc. Am. Proc. 39:1217–1220.
Anderson, Dave Hoehne, Pieter Los, and Andrea Peltzer. Ritchie, S.W., J.J. Hanway, and G.O. Benson. 1993. How a corn plant
Thanks to Alan Olness for his careful reading of and comments develops. Spec. Rep. 48. Iowa State Univ., Ames.
on the manuscript. This work was made possible by financial Roth, G.W., D.D. Calvin, and S.M. Lueloff. 1995. Tillage, nitrogen
support from the Missouri Agricultural Experiment Station timing, and planting date effects on Western Corn Rootworm injury
to corn. Agron. J. 87:189–193.
and the Missouri Commercial Agriculture Program. Russelle, M.P., R.D. Hauck, and R.A. Olson. 1983. Nitrogen accumu-
lation rates of irrigated corn. Agron. J. 75:593–598.
REFERENCES Stecker, J.A., D.D. Buchholz, R.G. Hanson, N.C. Wollenhaupt, and
K.A. McVey. 1993. Application placement and timing of nitrogen
Andraski, T.W., L.G. Bundy, and K.R. Brye. 2000. Crop management solution for no-till corn. Agron. J. 85:645–650.
and corn nitrogen rate effects on nitrate leaching. J. Environ. Varvel, G.E., J.S. Schepers, and D.D. Francis. 1997. Ability for in-
Qual. 29:1095–1103. season correction of nitrogen deficiency in corn using chlorophyll
Binder, D.L., D.H. Sander, and D.T. Walters. 2000. Maize response meters. Soil Sci. Soc. Am. J. 61:1233–1239.
to time of nitrogen application as affected by level of nitrogen Welch, L.F., D.L. Mulvaney, M.G. Oldham, L.V. Boone, and J.W.
deficiency. Agron. J. 92:1228–1236. Pendleton. 1971. Corn yields with fall, spring, and sidedress nitro-
Blackmer, A.M. 1996. Losses of fall-applied nitrogen. p. 55–59. In gen applications. Agron. J. 63:119–123.

You might also like