David Bullock wants to see on-farm experimentation become a routine part of how strip-tillers farm.

“Ideally, you’re always doing little experiments — or maybe big experiments — in parts of the field,” says the University of Illinois professor. “You’re always gathering information about how your yield is responding to different things. You start farming data intensively.”

Bullock leads the Data Intensive Farm Management (DIFM) research team, a multidisciplinary group founded in 2016 that is working to “revolutionize farm management.”

Funded by USDA grants, DIFM works with researchers and farmers to create large-scale, on-farm checkerboard-style field trials using precision ag technology. These trials generate vast amounts of data that give DIFM insight about how crop yields respond to weather, input application rates and field characteristics. The information helps farmers increase their profits by making data-driven management decisions based on the experiments they’ve conducted on their own fields.

4 Key Findings

At the 2022 National Strip-Tillage Conference, Bullock shared four general findings that DIFM has gathered based on the data generated from participating on-farm trials.

One of those, which Bullock admits growers probably already know, is that in general corn farmers are not using more nitrogen (N) than is economically optimal.

“There’s a lot of talk out there that farmers are using too much N, and they could cut back and not lose any money,” he says. “When we started this, that’s what we hoped was the case because we could tell farmers ‘put on less, and you’ll make more money.’ But that’s not what we’re finding in general, which is not surprising.  

He says some farmers put on too much, others put on too little, and a lot of it depends on the weather. But the data doesn’t support the notion that farmers in general are putting on more than necessary. There’s not a general recognizable pattern showing that farmers would make more money if they lowered their N rates.

However, Bullock says farmers may have an opportunity to reduce their costs by lowering their soybean seeding rates.

“We’re finding fairly often that a lot of farmers can lower their soybean seeding rate densities and make more money,” he says.

Bullock doesn’t have the data analyzed to offer specifics on how much a soybean grower could lower their rates by, adding that it’s not a guarantee every grower would be able to.

“You’ll want to do experiments to see,” he says.

Another general finding is that precision site-specific management often doesn’t pay on uniform fields, like the ones in central Illinois that are “flat and black.” Bullock says there’s not enough variability across the field to justify paying for equipment like variable-rate application. Similarly, commercial prescriptions are often not worth the money either, according to Bullock.

“We're finding a lot of these commercial prescriptions that farmers are buying are really not very good,” Bullock says.

This is because the data used to create those prescriptions is usually not coming from a farmer’s own fields.  

“Some prescription-writing methods have more valid research behind them than others, but even those based on data are not based on data from the field there are providing recommendations for,” he says.

Working with DIFM

The best way strip-tillers can figure out what works best for their fields is by doing their own on-farm experiments. They can conduct these experiments with the help of DIFM.

DIFM will work with farmers to design and execute checkerboard-style trials on their own fields. Farmers will need precision ag equipment — GPS guidance, a yield monitor and variable-rate controller — and ideally a field at least 80 acres in size. DIFM will work with farmers anywhere in the U.S. with any crops, not just corn and soybeans.

Strip-tillers can come to DIFM with an experiment in mind, but Bullock says that even if they don’t have one and they’re just interested in doing some on-farm research, DIFM will work with them to come up with a trial that matches their interests.

It’s a low-risk investment, as DIFM will pay to compensate farmers who lose money by conducting trials. DIFM has received a total of $8 million in grants from the USDA, which funds the payments.

“If your experiment has plots with some low nitrogen rates, you might lose money on that plot. If you have really high nitrogen rates, you might lose money. Either way, we’ll write you a check,” Bullock says.

Once a trial has been completed, the grower will send their data to DIFM researchers, who will process and clean it up. Farmers are also paid $1,000 for sending in their data, and the data still belongs to the farmer. Bullock says DIFM will never sell or give away a farmer’s data, as the group’s interest is to use it to write academic research articles.

After DIFM has processed the data, the researchers will talk with the farmer about what actionable steps they can take from the results.  

“We’re not just looking to just give them a prescription. We want them to talk to us about what they know. We want to hear what they’re thinking, and we’d like to trade ideas with them,” Bullock says.

DIFM may still create a prescription after that conversation and discuss how the prescription would change based on other variables, such as more or less precipitation.

Making On-Farm Research More Accessible

DIFM is working to create a cyber infrastructure that will allow for tens of thousands of acres of on-farm trial research every year — without the assistance of someone with a doctoral degree.

So far, DIFM has created a software for farmers and crop consultants to design trials themselves. The program allows them to create their own checkerboard field trial layout (see sidebar).

The group is also working to create a decision support system, so instead of being given a set prescription of what to do after a trial, farmers and their crop consultants can work with the program to make decisions from it.

“We’re trying to build something that’s interactive, so that the farmer and the consultant could ask the software questions like, ‘What if it rained more in the second week of July, how would that change the results?’” Bullock says. “'What if the price of corn were much higher? How would that change the results?’”