What explains yield variability within a field, and from one field to another? That question is at the heart of all precision data analysis, says Dan Frieberg, president of Premier Crop Systems, West Des Moines, Iowa, a precision agriculture software company. Finding answers involves a process that Frieberg describes as “a collision between technology and agronomy.”

Turn data into knowledge

The first step in turning geo-referenced field data into useful knowledge is comparing productivity across years and crops, says Kevin Kruize, precision ag manager at Central Valley Cooperative in Owatonna, Minn.

In this 62.5-acre field, Kruize measured the productivity of each 60 x 60-foot square using Premier Crop Systems software. The question he and the grower wanted to answer: Would this field benefit from site-specific management?

To compare productivity variations across years and crops, Kruize charted yields in each square as a percentage of the top yield in the field.

Six years of relative yield analysis revealed significant spatial variability. In addition, variability was fairly consistent from year-to-year and from crop-to-crop, Kruize says. “First we look for the areas that are consistently high and low-producing.” In this field, he found that for both corn and soybeans, the lowest yielding 10% of the field (red squares) were only one-third to one-half the yield the highest yielding 10% of the field (dark green squares).

Differences in soil moisture-holding capacity explain much of the yield variability in this field, Kruize says. The sandy-soiled center section of the field usually dries out in July and August, stressing the crop. The higher organic-matter soils in the upper and left portions of the field have better moisture-holding capacity.

Significant and consistent variability from a known cause makes this field an excellent candidate for site-specific management, Kruize says. Using yield and soil data, plus input from the grower, he defined four management zones within this field: A (high productivity), B (above average), C (below average) and D (low productivity). Two to four zones per field is typical, he adds.