Precision Agriculture, 1, 167᎐183 1999
ᮊ 1999 Kluwer Academic Publishers. Manufactured in The Netherlands.
Soil Nutrient Relationships with Topography as
Influenced by Crop
D. W. FRANZEN, V. L. HOFMAN, L. J. CIHACEK email@example.com
AND L. J. SWENSON
Abstract. Variable-rate fertilizer application is often based on grid soil sample data from a single year
of data in an annual crop rotation. The objectives of this study were to determine if crop history
influences spatial dependence the degree of spatial variability of nutrients in a rotation including both
annual crops and alfalfa, and to compare grid-based and topography-based sampling strategies for
representing within-field nutrient levels. A site in the Red River Valley of North Dakota was observed
over three years from 1994᎐1996. The site was divided into one field of continuous alfalfa Medica
a L. and an adjacent field seeded to spring wheat Triticum aesti
um L. in 1994, barley Hordeum
ulgare L. in 1995 and alfalfa in 1996. Samples were taken from a 16.2-ha site each fall in a 33-m grid
and analyzed for NO -N, P, SO -S and Cl. Topography was determined by measuring elevation in a
33-m grid with a laser-surveying device. Spatial dependence was determined by calculating the
semivariogram and using regression analysis to assess the relationship between the semivariogram and
the semivariogram model. Spatial dependence of NO -N and P was strongest following spring wheat and
barley, while spatial dependence for SO -S and C1 was strongest for vigorous stands of alfalfa. When
the continuous alfalfa stand declined following winter kill, NO -N and P spatial dependence intensified.
Topography based sampling was correlated with the 33-m grid by giving each 33-m sampling location its
value as directed by a topography sampling, then correlating that topography based value with the
original 33-m sampling value. Topography-based sampling was correlated with the 33-m sampling grid
for all nutrients following spring wheat and barley, but not in continuous alfalfa until the stand began to
decline in vigor. Following alfalfa seeding in the annual crop field, topography relationships with NO -N
and P decreased, while topography relationships with SO -S and Cl increased. Topography samplings of
sulfate-S and chloride were most highly correlated to 33-m grid values in vigorous alfalfa. Lack of
NO -N spatial dependence in the vigorous alfalfa stands suggests that a composite or field average soil
test might be sufficient to provide soil NO -N information under similar conditions.
Keywords: topography, crop rotation, soil sampling
The density and pattern of soil sampling influences the quality of soil nutrient
information gathered for a site-specific fertilizer application Franzen and Peck,
1995; Gotway et al. 1996 . Wollenhaupt et al. 1994 suggested a 60-m sampling
density for P and K levels in Wisconsin. Franzen and Peck 1995 suggested a 66-m
grid sampling density for initial investigation of P, K, and soil pH levels in Illinois.
Sampling recommendations for variable-rate N application on potato in Washing-
ton state is also approximately 66-m Hammond, 1993 . Methods of grid sampling
include square aligned grid sampling Franzen and Peck, 1995 and systematic
unaligned sampling Wollenhaupt, 1996 .