Degree Type


Date of Award


Degree Name

Master of Science




Soil Science (Soil Morphology and Genesis)

First Advisor

Andrew K. Manu


In Iowa, precision agriculture technology continues to be embraced by the agricultural community and large volumes of data are continually being collected. One type of data collected is spatially dense georeferenced yield points that are converted into maps to show the magnitude and variability of corn yield. The first part of this thesis reports on the use of “Anselin Local Moran's I Cluster tool” to quantify spatial corn yield variability in the Loess Hills of Northwest Iowa and then validate the results using a residual method. A 32 hectare site with geo-referenced corn yield data collected from 2006-2013 was identified. Cluster analysis was performed to identify the following three yield cluster types: high, low, and average yielding. Results showed that 2.3% of the site had consistently high and low yielding clusters. To validate the efficiency of the “Anselin Local Moran's I cluster tool” in identifying high and low yielding clusters, a residual method was used on the same dataset. This method identified areas of persistent yield variability previously identified by the cluster tool. Identifying these consistent clusters or areas of persistent yield variability may provide a tool to help create zones in a field for variable rate management.

The second part of this study sought to determine any relationships between soil chemical and/ or physical characteristics associated with persistent long-term high and low yielding clusters previously identified using cluster analysis. It was observed that high yielding clusters were deeper to a maximum depth of mollic colors and contained higher levels of soil test P and K to the depth of 0-25 cm when compared to the low yielding clusters. TC, TN, pH, and particle size distribution were not significantly different across the study site irrespective of cluster type.

These results suggest these fields show little consistent yield variability in the long term. This may be due to the type of parent material as well as consistent management by the farm operator. Based on these findings variable rate management would not be economically justifiable in a loess parent material but results may change with different soil parent materials and management strategies.


Copyright Owner

Thomas J. Lawler



File Format


File Size

113 pages