Greg Bartley wants more producers to take research into their own hands.
An on-farm specialist with the Manitoba Pulse and Soybean Growers, Bartley has been working with farmers to develop research trials in the commercial field, rather than the research plot.
That research may become a producer’s greatest tool in measuring the actual value of a product or management practice for their specific farm, he argued Jan. 22. Bartley helped open this year’s Ag Days speaking schedule in Brandon with his pitch for on-farm research.
Why it matters: Farmers rely on research to maximize field performance, but testing products and practices on farm may also give some farm-specific insights and fill in the gaps left by the usual research streams.
Bartley’s arguments centre around variable product results from farm to farm. Data derived on farm will be better suited to make management decisions for that farm, and take in differences on how a given product behaved in that environment, he told the room. At the same time, he argued, there are a wealth of products on the market that have little to no third-party research backing up their efficacy.
Taking in the basics
Planning is crucial to successful research, the Ag Days speaker said.
Farmers should plan for test strips at least 1,000 feet long and at least one harvest width wide, he said, although a test strip the full length of the field would be even more effective. A uniform field will give the best results, he added, although increasing replications can offset a field’s natural gradients and soil variability.
There is no such thing as a perfectly uniform field, he admitted, something that he says drives home the need for replication and randomization, since farmers otherwise risk skewing studies. The farmer risks measuring the field’s variability rather than any change due to treatments if that scientific method isn’t followed, Bartley argued.
“I think the first thing to consider is, are you randomizing and replicating your treatments?” he said. “The replication is really important for on-farm trials just to manage the variability in the field and allow you to do statistical analysis.”
Two replications will not be enough to make management decisions, he warned. His own studies through the Manitoba Pulse and Soybean Growers use six replications, and farmers looking to do their own research should aim for at least four.
Adam Gurr, founder of field-scale research firm Agritruth Research, also cited proper scientific method, although he puts the minimum number of replications at three.
“I guess you need to take into account field space and everything like that as well,” he said, adding the number may change when producers become more accustomed to the research and the variability across a specific field.
“You get comfortable with the amount of variability that’s there and (those are) maybe (fields) where I would compromise the amount of replicates for the opportunity to do more treatments,” he said.
The Brandon-area producer has been integrating field-scale research into his normal commercial operation for years, drawn from Gurr’s desire for more relatable data. He wanted concrete data on farm management on a field scale, as well as data that was tailored for his farm’s circumstances. His multi-year farm-level trials soon became a hit with his fellow Manitoba farmers, leading Gurr to turn his trials into an independent research company.
The company has since been tapped for third-party product testing by companies like Agrotain or Proline, providing some of the independent efficacy data that both Gurr and Bartley say is sorely needed.
Bartley also warned producers to vary only one or two treatments and keeping all else — other products, seeding, etc. — constant.
“We still need small-plot research,” he said. “It’s really important and it allows us to really understand the fundamentals of why or what is going on in the trial. In many cases, you can test multiple different things and you can figure out maybe a combination of treatments… when you’re doing on-farm testing, you need to keep it simple.”
Producers looking to stress test a specific treatment or product are prime candidates for field-scale trials, Bartley said, “but if you really don’t know what’s going on, or if there’s multiple different treatments that are going on, that’s more suitable for a small-plot setting.”
Research and farm flow
Bryce MacMillan, of Marquette, says his two years of on-farm research have been largely complication free.
“When we first got into it, I was a little bit hesitant,” he said. “I was sort of concerned by how much time was going to be taken putting in the crop and taking the crop off, but after our second experience doing it, (it was) virtually flawless — no problems as far as harvesting the crop in the fall.”
In retrospect, MacMillan says he might have been more concerned about cleaning plots up in fall his first year, something that might have sped up harvest.
Harvest delay might stem some enthusiasm, Bartley admitted to the Ag Days audience. At the same time, he argued, the few minutes’ delay to properly measure and record data may pay big dividends in informed management decision.
Gurr says his own process is relatively streamlined by now, despite the scale of on-farm research that his operation presents on a yearly basis.
“Really, for us, the amount of time that we’re sacrificing is really just where the grain cart (is) standing still rather than unloading on the go,” he said.
For a basic trial, he estimates that lag at about 15 minutes a load.
“You’ve got to focus on the economics of these trials you’re doing,” he said. “I know harvest can be a hectic time and if you’re not able to do it in a timely fashion, I know that can be a barrier, for sure.”
Farmers may also want to consider how harvest can skew data, according to Bartley. Harvesting in a single day with the same combine sidesteps any variability between equipment or weather conditions, he said, while lodged crops should be harvested in the same direction.
Both Bartley and MacMillan advised farmers to start small, but not to back away from the scientific process if they want to start their own trials.