For the last three years, Doug Applegate's corn and soybean fields have been striped like a zebra.

Applegate split-plants about one-third of his acreage. To test what hybrids and varieties work best in his fields, he fills half his planter boxes with one hybrid or variety and chooses a different hybrid or variety for the other half. As he plants, the entire field becomes a test plot to see how the two cultivars perform under the varying soil type and conditions.

At harvest, his yield monitor tracks yields and computer software converts the data to show the competitive advantage of each type of seed.

"We don't strip-plant all our acres," says the Oakland, IA, farmer. "Each farm will have some kind of test. That doesn't mean they'll all get harvested that way. Usually we end up with 500 acres of tests.

"Sometimes it's glaringly obvious that one hybrid is better than another. If they're close, we use our data to supplement with other yield data," Applegate says. "It keeps us ahead of the curve. I figure it's worth 11/2 bu of beans and 5 bu of corn per acre over what we would have if we hadn't done the strips."

Some agronomists argue that it takes at least three years of data to get meaningful results. And by then new hybrids have replaced the ones tested.

But even one year's data can help farmers gain confidence with their seed selection, says Tom Doerge, precision farming specialist at Pioneer Hi-Bred International. "Growers will allow one year's data to influence their decisions. The more comparisons they have the more confidence they have in their ability to understand the data." Strip-planted fields yield a lot more than just yield data, however. "We get data that's specific to our farm and how we do things," says Applegate. "It's a way to test how hybrids and varieties respond to different tillage and management."

If you're looking for irrefutable results from split-planter trials, you're likely to be disappointed. There are just too many variables that are unknowns at this point. "We're still in the research phase with this technology," says Monsanto's Tracy Blackmer.

"One year's data isn't really sufficient for anything more than a guide. It takes quite a bit of data to determine if a relationship exists and whether or not it's predictable over time," he says. "We've tried to work with groups where 10 or more farmers do the same trial. When you can start to look at a number of side-by-side comparisons over a wider area, you can learn more."

Over time, split-planter trials may tell you more about your farm than about individual hybrids or varieties. "In the future, farmers who have used split-planter trials may be able to define the different growing environments on their farms and then do strip tests for those environments," Blackmer says.

Some different "environments" could include areas of the field where high-pH soils lead to chlorosis and yield loss, or where soybean cyst nematode (SCN) infestations have reached economic levels.

Split-planter tests conducted by Monsanto have shown that there's definitely a yield advantage to SCN-resistant varieties where there's significant infestation. That advantage is lost where SCN infestations are light or non-existent. A computerized yield map can help you decide if it's worth planting a field with an SCN-tolerant variety, even though it yields less in areas of minimal infestation.

Ohio State University plant pathologist Mac Riedel saw similar results in trials sponsored by Pioneer. "The extremely good fit of nematode populations to soybean yield, presented as a whole field map, will visualize for growers the damage this nematode causes in farm situations," he says.

The data isn't always as clear-cut as it might appear, warns Blackmer. "Cyst tolerance is only one of the genetic differences between two varieties," he says. "There may be other genetic differences in the same comparison that are also affecting yield that aren't as obvious."

That's one of the dangers that Jim Schepers sees with the yield advantage maps generated from split-planter trials that farmers use to make decisions.

"I'm afraid farmers will end up with pretty yield maps and won't take the initiative to figure out why the differences occurred," says Schepers, a USDA-ARS researcher at the University of Nebraska. "You need to ask why one hybrid is better than the other. You also need to realize that a yield map is a postmortem of what happened. It can't tell you when that difference occurred. You need to incorporate remote sensing to figure that out."

Schepers warns that there is some inherent inaccuracy with yield monitor data.

"Your yield monitor may be accurate, but the flow through isn't, in relation to where that grain came from," he says. "Due to combine flow dynamics, you need to harvest plots in the same direction to more accurately determine where the differences occurred. An aerial photograph is far more accurate."

In spite of some shortcomings, Schepers sees strip-planting as a valuable tool for farmers. "It gives us an opportunity to examine the extremes in the fields more realistically.

"For most people it's a learning tool. After two to three years they start to figure out how to use it," Schepers says. "It will only make you money if you have the ability to integrate the yield data with other information. You have to understand why the variation is there before it can truly become a management tool."

Split trials allow farmers to test other crop inputs. "The split-planter method can be used to compare two seed treatments, two fertilizer treatments, two pesticides, two tillage treatments or any other two agronomic treatments that can be placed in parallel strips," says Doerge. "Split-planting trials help you put a number on the yield differences."

The software to compute those yield advantages is available free from Pioneer and Agris at this Web site: www.agris.com. Farmers who plant at least 20 acres of Asgrow or Dekalb products can have their yield data analyzed at no cost by Monsanto. Both programs provide yield difference maps that show how each hybrid, variety or crop input treatment performed.