Does Timing of Nitrogen Application Still Matter After Silking?
Background and motivation
In corn, late-season nitrogen (N) management is a topic of practical relevance for growers and agronomists alike. Modern hybrids show increased N uptake during reproductive stages and a delayed pattern of N remobilisation to the grain — raising the question of whether N applications at or even after silking can still influence final grain yield.
Grain yield is the product of two numerical components: grain number per unit area (GN), set primarily around flowering, and final grain weight (GW), accumulated during the post-silking grain filling period. While GN is generally considered the dominant yield determinant, GW variation under N-limiting conditions can be agronomically significant.
This story focuses on one specific question: does the timing of N application — at flowering versus two weeks after — still matter for the grain filling process?
Does Nitrogen Timing After Silking Still Matter for Grain Filling in Corn?
Javier A. Fernandez 
Ignacio A. Ciampitti
Kansas State University, 2018
The baseline: what N stress does to grain filling
Under zero-N conditions, we can see the complete grain filling trajectory without any late-season N support.
Each point represents a kernel sampled every 3–4 days from R2 to black layer — ten kernels from the central ear portion, tracked for dry weight accumulation across the entire grain filling period.
Without nitrogen, kernels fill at a steady linear rate until they stop — abruptly. The curve flattens into a plateau well before the equivalent fertilised treatments.
This is the signature of N-limited grain filling: not a slower rate, but an earlier termination.
The model: fitting a linear-plateau to grain filling data
To quantify grain filling parameters, we fitted a linear-plateau (bi-linear) model to each treatment × hybrid combination using nls() in R, with the self-starting function SSlinp from the nlraa package.
The linear-plateau model captures a biologically meaningful curve: kernels fill at a constant linear rate until physiological maturity, at which point dry weight plateaus. This breakpoint corresponds to black layer formation.
What the parameters mean
The SSlinp function fits three biologically interpretable parameters:
a— intercept: initial grain dry weight at silking (mg). Small positive value representing the lag phase.b— grain filling rate (GFR): the slope of the linear phase, in mg grain⁻¹ day⁻¹. This is a largely genotype-determined trait.jp— join-point: grain filling duration (GFD), the number of days after silking at which the plateau begins — i.e., when black layer forms and filling stops.
The final grain weight is derived as a + b × jp.
This distinction between rate and duration is central to interpreting the N treatment effects.
Adding nitrogen: the effect on grain filling
Now compare the zero-N trajectory against the N-fertilised treatments.
Both fertilised plots received N at planting and at V6 — the late-season application was the differentiating factor. Here we add the N at flowering treatment.
The slopes — the b parameter, GFR — are nearly identical across treatments (~7.1–7.3 mg grain⁻¹ day⁻¹). Nitrogen did not change the rate of grain filling.
What changed is the join-point (jp): the fertilised treatment extended grain filling duration by approximately four days relative to zero-N — from ~45 to ~49 days after silking.
At a filling rate of ~7 mg grain⁻¹ day⁻¹, four additional days translates to a 21% increase in final grain weight (from ~252 to ~305 mg grain⁻¹).
The key finding: does timing after silking still matter?
The critical agronomic question is whether a late-N application — two weeks after silking, when kernels are already in active linear filling — can still influence the grain filling process.
The answer, for both grain filling rate and duration, is no.
The late-N treatment — Late-N applied 14 days post-silking — produced grain filling curves that were statistically indistinguishable from those receiving N at flowering. The b (GFR) and jp (GFD) parameters did not differ between the two fertilised treatments.
This means that by two weeks after silking, the grain filling trajectory is already determined. Late-N does not extend the duration further, nor does it accelerate the rate. The join-point is set.
The implication for management is direct: the grain filling window is effectively closed by ~R2. N applications after this point should not be expected to improve grain weight. The yield response to N in this experiment — a ~95 bu/a difference between fertilised and zero-N — was driven entirely by GFD differences established before or at silking, not by the precise timing of the late reproductive application.
library(nlraa)
# Fit linear-plateau model per treatment
# SSlinp(x, a, b, jp):
# a = intercept (initial grain weight, mg)
# b = grain filling RATE (mg grain⁻¹ day⁻¹) #<<
# jp = join-point = grain filling DURATION (days after silking) #<<
fit_zero <- nls(
gw ~ SSlinp(days, a, b, jp),
data = zero_n
)
fit_flower <- nls(
gw ~ SSlinp(days, a, b, jp),
data = n_flower
)
fit_late <- nls(
gw ~ SSlinp(days, a, b, jp),
data = late_n
)
# Final grain weight (plateau) = a + b * jp
coef(fit_zero)
Take-home messages
1. N supply affects grain weight through duration, not rate. Grain filling rate (GFR, ~7.1–7.3 mg grain⁻¹ day⁻¹) was consistent across N treatments and is predominantly a genotype-determined trait. Grain filling duration (GFD, the jp parameter) was the N-sensitive component — shorter under zero-N, extended by ~4 days with adequate N supply.
2. The effect of N on grain filling is established before or at silking. Both N-fertilised treatments — whether N was applied at silking or two weeks later — produced identical grain filling trajectories. Late-N applications at R2 + 14 days did not extend GFD or increase GFR. The grain filling window is effectively closed by early R2.
3. Practical implication for N management. The ~95 bu/a yield penalty under zero-N was driven by reduced grain number and reduced grain weight. The grain weight component was entirely attributable to GFD shortening. For growers: if adequate N is not available by silking, a late post-silking application will not recover grain weight losses. The window for N to influence grain filling is at or before R1–R2.