Degree Type


Date of Award


Degree Name

Doctor of Philosophy



First Advisor

Alfred M. Blackmer

Second Advisor

Robert Horton


Soil testing for nitrate when corn plants are 15 to 30 cm tall is recognized as a valuable tool for estimating N fertilizer needs in humid portions of the United States. Although there is growing appreciation for the importance of spatial variability in soil nutrient levels, high-density sampling is not practical for the soil nitrate test. In this document we report initial studies to identify optimal sampling densities for non-fertilized corn after soybean in Iowa. Soil nitrate concentrations were measured in 24 cornfields in production agriculture during 1995, 1996, and 1997. The preceding crop on all fields was soybean, which did not receive fertilizer N. The mean spring soil nitrate concentration was 8.2 mg N kg-1. Essentially all samples had concentrations below the critical value 25 mg N kg -1, which is often used as the optimal level for corn production. An analysis of variance showed that a simple model, which included the variables Field, Test area within Field, and Sample, could explain 81% of the variation in soil nitrate concentrations. Linear regression analyses showed that much of the variation (78%) in soil nitrate concentrations within fields was explained by soil organic matter concentrations. Results show that soil nitrate concentrations can be predicted with reasonable accuracy even with extremely low-density sampling if soil organic matter concentrations are used to guide the sampling.;Nitrogen (N) fertilizer needs for corn usually are estimated by assuming that fertilizer needs are proportional to yields, but the validity of this assumption has been difficult to evaluate by using experimental data. We evaluated this assumption for corn grown after soybean in Iowa. Nitrogen fertilizer treatments were applied in replicated strips that crossed several soil types within each of 17 fields. The strips were harvested by using combines equipped with yield monitors and GPS receivers. A geographic information system (GIS) was used to calculate mean yields and yield responses to fertilizer N for fields and individual soil survey map units within fields. Analyses showed that neither observed yields nor published yield potentials provided a reasonable basis for predicting yield responses. However, these yield responses showed significant relationships with soil survey map units and soil organic matter concentrations measured within soil map units (SMU). Soil organic matter concentrations were positively correlated to yields and negatively correlated with yield response to applied N. These observations suggest that increases in soil organic matter had dual effects, increasing yields and supplies of N. These observations also explain why fertilizer needs should not be estimated from expected yield or attained yields under the conditions studied and suggest that soil organic matter deserves more attention when estimating N fertilizer needs.



Digital Repository @ Iowa State University,

Copyright Owner

Susan Eileen White



Proquest ID


File Format


File Size

73 pages