In an effort to deal with these inconsistencies, SDSU’s Carlson is developing variable-rate seeding algorithms based on long-term average yields and the degree of yield variation within the field. He says growers should think of productivity potential as a continuum, rather than as discrete zones.

At Advanced Cropping Systems, Mieth combines yield maps and other layers of information, such as aerial imagery, topography and slope, to estimate yield potential. In eastern Nebraska, for example, “It’s almost inevitable that south- and southwest-facing slopes have a lower yield potential than southeast-, northwest- or north-facing slopes. We process all the data available on the field, and let that determine productivity potential on a 5-meter resolution.”

Despite the high-tech analysis, though, there’s a great deal of subjective knowledge that goes into defining these zones, Mieth says, especially in areas with inconsistent yields. “We sit down and discuss what the grower knows about the field. There’s a lot of agronomic judgment involved.”

Farmer Alverson agrees. Until just a few years ago, the family varied seeding rates by an educated guess. “It was all visual, based on the slope of the ground, the color of the soil, and our knowledge of the field. It was an art.”

They’ve gone digital now, adjusting planting rates from 24,000 to 36,000 seeds/acre on pretty much every planter pass, Alverson says. Still, he calls their GPS-based productivity zones, which are delineated from yield and topography maps, a work in progress. “They’re not as accurate as what we can see out in the field. The challenge is getting what’s in our eyes and heads into our prescription maps.”

Crop consultant Kim Retzlaff, Applied Agronomics, Aberdeen, S.D., sorts through 80-100 satellite images to locate high- and low-potential areas within fields. He says satellite imagery is often a better indicator of productivity differences than yield monitor data, which can be misleading if not collected and filtered correctly.

Fields that show little yield variation will not be good candidates for VRS, adds Joel Wipperfurth, Winfield agronomist in Owatonna, Minn. “In general, you need to see more than 5% yield variability.” Look for areas of the field that yield consistently above or below average, he says.