Assessment of Multi-Year Variation of Corn Yields and Possible Causes On the Des Moines Lobe of Iowa.

Since the early 1990s, yield monitors have become an essential component of many precision farming operations.  Yield monitors produce yield maps which can be analyzed to depict the magnitude and location of yield variability within a field.  Yield variability maps provide feedback for determining effects of weather, soil properties and management practices.  Through an understanding of within soil map unit variation, agriculturalists will benefit by gaining access to new methods that will potentially increase soil sustainability and yields.

Due to increased adoption of precision farming methods, large volumes of yield data are generated annually.  Unfortunately analyses, interpretation and use of the data to explore underlying factors of within field and within soil map unit yield variation are consistently lagging behind the rate of data collection.

Long term geo-referenced corn yield data were obtained from three fields totaling approximately 500 acres.  Soil Map unit and management information were also assembled.  Using “Anselin Local Moran’s I” (cluster analyses) as a statistical spatial interpolation tool, corn yield variations have been examined.  Preliminary analysis shows a consistency in spatial yields.

As a first step, yield data were filtered to omit extreme values which are defined as negative values, zero values and values greater than a predetermined maximum yield cut off value.  Secondly, the cluster analyses identified statistically significant hot spots, cold spots and spatial outliers within fields and soil map units using Euclidean distance Inverse distance weighting.  Cluster analyses were also performed after combining several consecutive years of yield data in order to identify locations within the fields and soil map units which were consistently variable. Analyses are currently underway to identify weather, soil and management factors responsible for the identified variation.