“Right now we have a theory, a guess, a gut feeling that variable-rate planting should work, but we don’t have the research on how to write a planting prescription,” says Tracy Blackmer, research director for the Iowa Soybean Association. “It is so much easier to develop the equipment than the recommendation to drive it. Right now, variable-rate seeding is guesswork.”

Blackmer notes that most variable-rate seeding recommendations are based on soil map units and aerial imagery. “Soil map units are a very intuitive way to look at how to manage soils differently, but there is not a lot of data on how to draw boundaries. I think there is great potential, but I think it is important to evaluate where it is doing good or hurting you,” he adds.

To that end, the ISA’s On-Farm Network has conducted two years of research on variable seeding rates. Results from the first year’s tests in 2009 raised questions about whether full-field corn seeding rates used by cooperating farmers were too high in the first place. This may have masked potential benefits from variable-rate seeding. Results from 2010 trials were not available at press time.

Nafziger says it makes intuitive sense to vary seeding rates based on yield potential of differing management zones. He has found that corn yields are positively correlated with the plant population needed to produce those yields. So the higher-potential parts of fields should get more seed.

For fields with average variability, corn seeding rates might be set to vary up to 5,000 or 6,000 seeds/acre across the field. Fields with larger ranges of yield potential should see a wider range of rates, but Nafziger says to be careful not to set rates outside the normal range for the growing environment. Except for very light soils, dropping populations to less than about 25,000 seeds/acre may lower yields, especially if weather turns out to be good, he adds.

Illinois research comparing various seeding rates hints at the potential payoff of variable-rate seeding of corn in relatively uniform soils, and in soils with more variability. To simulate variable-rate seeding, researchers compared optimum seeding rates with the most profitable fixed seeding rate.

In northern Illinois, where soils were relatively uniform, the exercise showed an $8.81/acre higher return for the variable rate compared to a 33,500-seed/acre flat rate. In southern Illinois, where soils were more variable, the theoretical return to variable-rate seeding was $34.52 compared to a 27,300 seed/acre flat rate. Higher returns came from avoiding high populations in areas where high populations decreased yields, as well as higher yields from more productive areas with higher-than-normal seeding rates.

The Illinois research mirrors Millar’s experience in central South Dakota, where it is not uncommon to vary seeding rates from 18,000 to 36,000 seeds/acre in the same dryland field. “We absolutely harvest more grain on the low-end areas because we have fewer barren stalks,” he says. “If your yield varies from 75 bu./acre to 220 bu./acre in the same field, it doesn’t take a rocket scientist to see that variable-rate seeding works. But if all you have is a 15- to 20-bu./acre difference, I don’t think you want variable seeding rates.”