The Data-Intensive Farm Management project was recently awarded a $4 million grant from the U.S. Department of Agriculture’s Natural Resources Conservation Service to work with farmers across the country to improve the efficiency of planting, management and input decisions, integrating technological advancements into agricultural practices.
“We placed a focus on what we thought was most important for Montana, which was to see what information we could get from precision technologies that were becoming commonplace on the farm, including yield monitors and protein sensors,” said Maxwell. “We saw this as a big opportunity to determine: Can Montana farmers increase profits by using this technology?”
Because of diverse geography and climate across Montana, variables such as fertilizer application and irrigation can be fine-tuned using modern technologies to conduct experiments in every field, with minimal extra effort by the producer. Maxwell and Sheppard had been collecting that type of data for several years through MSU’s Montana Research & Economic Development Initiative, which ran from 2016 to 2017. Because of that initiative, many Montana farmers already have the tools to conduct site-specific evaluations of fertilizer, weed control and seeding rate management strategies.
Maxwell has been researching precision agriculture since 1998 and works closely with farmers around Montana to design on-farm experiments and assess crop production. He will lead the on-farm experimentation component of the DIFM project and develop tools and procedures for making field treatment prescriptions.
Paul Broyles, who farms winter wheat, malt barley and legumes in Stillwater County with his father, Gary, has worked with Maxwell for the last six years collecting data on their land.
“We feel like as farmers we’ve learned a lot through the experiment, just from sitting in the seat of the sprayer or the combine,” Broyles said. “It’s been helpful. It’s been interesting to see how the changes we have made in application rates, based on our observations, have increased productivity.”
Since 2015, Sheppard has explored artificially intelligent tools to design on-farm experiments and predict crop yield. He’ll lead the development of an “analytical engine” that will use complex modeling and machine learning to identify trends in years’ worth of data and make recommendations based on variation across the 15 project member states — even at scales down to individual sections of a particular site.
“By introducing artificial intelligence and machine learning technologies focused specifically on farmers’ needs and practices, we hope to enable farmers to make data-informed decisions tailored to their farms,” said Sheppard. “While the main goal is to increase farmer profit, we also hope to help farmers make decisions to reduce negative environmental impact.”
The DIFM project, led by University of Illinois agricultural economist David Bullock, is the largest of 14 grants awarded by the NRCS to help partners implement and evaluate innovative conservation practices on their farmland. The funding is provided through On-Farm Conservation Innovation Trials, a component of the Conservation Innovation Grants program first authorized in the 2018 Farm Bill. Awardees are required to evaluate the conservation and economic outcomes from the practices and systems they study, giving partners, producers and NRCS critical information to inform future conservation work.
For Maxwell and Sheppard, that also includes advising Montana farmers in methods that they can put to use.
“I think most of our collaborating producers are excited about it and feel like they’re getting some really good information,” said Maxwell. “They’re experimenting, and they like this idea of utilizing our approach to learn more in the context of their operation. Now they’re expanding their experiments on their own, which is ultimately what we’d hoped for them to be able to do. It’s our job to synthesize that information, put it together in a way that helps them make better decisions.”
For the Broyles family, the core value of the research lies in the ability to take what Maxwell and Sheppard develop and put it into use on a daily basis.
“We had a better time producing accurate data when we learned how to do more of this in house,” said Gary Broyles. “We have more confidence now. In the development of this project, we feel more confident that what we were learning is reliable.”
The Data-Intensive Farm Management project is recruiting producers in Arkansas, Idaho, Illinois, Louisiana, Texas, Michigan, Minnesota, Montana, Nebraska, North Dakota, Ohio, South Dakota and Washington, but cotton, corn, soybean and wheat farmers from any state can apply to participate. More information can be found at the DIFM project homepage.