Remote sensing of canopy reflectance on a field scale has been proposed as a useful tool for diagnosing nitrogen (N) deficiency of corn plants. Differences in leaf color among plants can be quantified by analyzing the canopy reflectance. The approach to

1
Quantifying Temporal Patterns in Symptoms of
Nitrogen Deficiencies in Corn by Remote
Sensing Jun Zhang a, Alfred M. Blackmer a, Peter
M. Kyveryga a, Mark J. Glady b, and Tracy M.
Blackmer b a Agronomy Department, Iowa State
University, Ames, IA 50011 b Iowa Soybean
Association, 4554 114th Street, Urbandale, IA
50322
Introduction
Results and Discussion
Remote sensing of canopy reflectance on a
field scale has been proposed as a useful tool
for diagnosing nitrogen (N) deficiency of corn
plants. Differences in leaf color among plants
can be quantified by analyzing the canopy
reflectance. The approach to making fertilizer
recommendation through remote sensing deserves
attention because factors other than N
deficiencies can also influence canopy
reflectance. When extra N is applied to
reference strips of cornfields and differences in
canopy reflectance between non-fertilized strips
and the reference strips are compared to estimate
N rates for in-season correction, the temporal
patterns in canopy reflectance could have shown
confounding effects of N fertilization and other
factors. We report field-scale studies in which
remote sensing of corn canopy reflectance
produced unexpected evidence for
fertilizer-induced advancement of growth stage
early in the season.
Canopy reflectance values measured early in
the season were mainly affected by the
application time and that measured late in the
season were mainly affected by the application
rate (Fig. 1). The canopy reflectance values
measured late in the season were better
correlated to yield responses to N (Fig. 2).
The difference in yields between the early and
late times of fertilizer applications was not
statistically different (data not shown).
Therefore, temporary shortages of N may produce
symptoms of N deficiency in situations where
subsequent additions of N should not be expected
to increase yields.
Materials and Methods
Three no-till fields under the corn-soybean
rotation were chosen in 2002 within the
Clarion-Nicollet-Webster soil association in
Green County, Iowa. Treatments were applied in
strips going the length of the fields and
consisted of various application dates and rates
of N fertilizer in a split-plot design with 4
replications. The early (June 7, 6, and 5) and
late (June 17, 14, 13) applications of N
fertilizer were as main plots at Sites 1, 2 and
3, respectively. Three N rates (56, 112, and 168
kg N ha-1) applied as UAN were as sub-plots.
Aerial images taken for each site on three
dates are shown in Fig. 1. Reflectance values
were derived from six test areas within each
replication in each site. Relative reflectance
was expressed as a percentage of the mean
reflectance for lower N rates over that for the
highest N rate (168 kg N ha-1). Corn yields in
the 6-row strips were measured by using combine
equipped with a yield monitor at one-second
intervals.
Figure 2. Relationships between canopy
reflectance and yield responses to N.
Conclusions
Remote sensing late in the season has
greater ability to detect yield-limiting
deficiencies of N because supplies of N become
exhausted by growth under such a condition.
The appropriate use of remote sensing requires
distinction between the short term effects of a
deficiency (often only temporarily) and the net
effects as expressed in grain yields at the end
of the season. The key to recognizing the power
of remote sensing is to make these distinctions
and recognize that the growth of corn is a
dynamic process that occurs over several months
and it is divided into clearly different growth
phases.
Figure 1. Aerial images taken on three dates at
the three sites. Strips that received early N
applications are marked by solid lines and strips
that received late N applications are marked by
dotted lines.