Features Agronomy Seeding/Planting
How to do on-farm trials

On-farm trialing is a way for growers to test new varieties, products or equipment and to confirm if it works in their operation. “We started doing on-farm trials to help our customers and ourselves assess
different inputs and varieties and make better management decisions,” says David Kaun, who runs a large pedigreed seed operation near Penhold, Alberta with his brother Mark. Their parents started Kauns Seed Farm in 1950, and today the farm grows canola, wheat, barley and peas. “Our first trials were testing Tilt when it was first released, which convinced us that on-farm trials can provide us with useful information for making decisions. We saw some phenomenal yield increases when comparing treated and untreated barley strips.”

Kauns Seed Farm uses a weigh wagon to accurately measure trial yields. Photos Courtesy Of Kauns Seed Farm.

Although most of their land is under pedigreed seed production, Kaun has some fields for commercial production and on-farm trials. “The key is selecting a really uniform portion of the field for conducting the trials,” explains Kaun. It is also important to consider the history of the land and ensure prior trials, such as fertility or seed treatment, will not impact the new trials. “We want the field to be as uniform as possible and to be able to treat every strip exactly the same. This way we eliminate every variable possible, except for the treatment or variety we want to compare. Our philosophy is, if we aren’t going to do these trials the best way we know how, then we shouldn’t be doing them.”

Kaun believes the advantage of on-farm trials, compared to small plot trials, is that they are done on a field scale on land with a known history, using farm scale equipment and common practices farmers use in their operations and quantified. He notes that to ensure meaningful results, using replicated trials is key. Every strip trial within a variety trial is seeded one drill width the same way and on the same day, then they are all treated exactly the same in terms of fertilizer application, herbicide and fungicide applications or any other treatments. The headlands are seeded to one variety.

A replicated trial of each strip is included to provide valid data and information for comparison. “Once we select the land for the trials, we then develop the plan for where each strip will go,” he says. A replicate of the check variety is placed on each side of the plot to make sure there are not any other variables making an impact across the width of the trial. Every strip is harvested and weighed separately for accurate comparison. “When we look at the data, we want to make sure it means something.”

New technology has made the process of setting up and managing on-farm trials easier. “We use GPS and tramlines, which give us an extra marking in the fields. And because of the width of our drills and swather, we don’t have to take the tramlines into the trial area, we harvest without tramlines. They are really useful for trials with fungicides for example, where we can leave one or more tramlines unsprayed as the check. Then at harvest we can go in and harvest the treated areas separately from the untreated strips.”

During the crop year, Kaun regularly monitors and records observations on emergence, maturity, standability and other details. Each strip is swathed when the crop is mature, which can mean going back three or four different times to swath all the strips. “The strips are all swathed in one direction and all combined in the same direction with the same combine. Our drill is 51 feet wide, more than twice as wide as our swaths, so we cut two full 21 foot swaths, leaving the edges and tramlines to be harvested with the headlands.”

Kaun uses a yield monitor as a rough check but does not think it is accurate enough, so he uses a weigh wagon to weigh every strip separately. The weigh wagon is put in one spot in the field and never moved until the trial harvest is complete. It is always emptied and set back to zero after each strip is harvested, and the combine is always checked to ensure it is empty. At the end, Kaun goes in with a straight cut header and finishes harvesting the headlands, strip edges and tramlines. All of these steps do take extra time, it is a big commitment but Kaun believes unless a grower is prepared to do it right they might as well forget about it. Kaun has weighed off as many as 85 different strips in one year, sometimes stretching the patience of his harvest crew.

After switching to a no-till system, Kaun began working with a crop consultant, Errol Schimke of Green Key Solutions on their fertility program. “With the rising prices of fertilizer and the change from conventional to zero-till, we wanted to really sharpen up our fertility program and move away from a nutrient replacement strategy,” explains Kaun. He uses a one-pass seeding system with a Flexi-Coil 5000 air-drill set up with a double shoot nine inch spacing, four inch paired rows and one tank for seed and two for fertilizer. This system makes it fairly easy to apply variable rates of fertilizer on-the-go during the seeding operation. Kaun took advantage of the full field strips and added a fertility trial with the variety trial strips.

“Along with comparing fertilizer rates, we’ve also been trying to quantify fertility requirements on barley in the first and second year after a pea crop. After five years of trials, we’re seeing some very interesting results.” Kaun conducted a yield analysis and an economic analysis to determine the rate of return on investment. “I can look back and say the fertility trialing we’ve done has probably saved our farm between $30,000 and $40,000 per year on fertilizers.”

The fertility trials confirmed that without replicated trials growers cannot accurately estimate the yield difference. “Visually we saw big differences in trials and expected that to translate into big yields. However, we’ve been really surprised on what we think is going to happen compared to what actually happens when we take the yield. In some cases, we were actually just growing straw, it didn’t translate into a lot more bushels.”

Kaun is pleased with the results and returns of his on-farm trial efforts. Anytime a company approaches him to try a new product, variety or equipment, he will do an on-farm trial first to see if it works on his farm under his system and if it pays to use it. Kaun shares the on-farm trial results with his customers so everyone benefits. The crop consultant is also using the information to help make recommendations for other growers in the local area.

Replicated on-farm trials provide powerful information
On-farm trials can range from single demonstration strips, side-by-side comparisons to fully replicated trials. However, to realize truly meaningful results and make decisions, demonstration or single strip trials will not be enough. The most powerful information comes from replicated on-farm trials.

“Replication gives you more power to separate treatment effects and does give a measure of variability in the field to help determine if the difference that occurs is statistically significant or valid,” explains Roger Andreiuk, agronomist with Alberta Reduced Tillage Linkages (RTL) in Leduc. “Once you can confirm the statistical significance, then you can apply economics to see if the variety or product you are testing really does pay.”

Andreiuk explains that replication is the repeated random application of a treatment in a field in order to get a good idea if that new product or technique will work. The reason for the replication is that the treatment (e.g. variety A) can be checked out across the natural variability that exists in all fields. It is important to repeat the treatments randomly to ensure that any one treatment is not biased or favoured in any way. The more times a treatment is replicated, the more confident a grower can be of the results.

Fertility trialing has saved Kauns Seed Farm between $30,000 and $40,000 per year on fertilizers.

Although people are pretty good at estimating yields over a whole field, they usually are not very accurate on individual strips. “It’s difficult to see a five or 10 bushel yield difference in a strip, but if you add up the difference in terms of economics, it can be substantial, says Andreiuk. “There are many effects that you can’t see visually, but they are still financially important.” That is why replicated field scale on-farm trialing can be so useful.

For those who have not done this before, a good place to start is by working with partners such as RTL or other organizations, crop consultants or agri-businesses, to help with the process of planning and implementing the project. “Working with partners can help distill ideas, provide opinions and advice and bring other networks and support services to the project,” says Andreiuk. Consider other management factors that may impact the results, such as access to labour, available equipment, time or other infrastructure factors that may determine the success of on-farm experimenting.

“Some of the more sophisticated equipment makes replicated trials much simpler,” says Andreiuk. “Along with GPS, yield monitors and yield maps, technology such as parallel tracking really helps with field operations. For example, using your randomization plot map, you can parallel track over to the next replicated plot without changing settings. Then you can go back and do the same for each replicated treatment. “Another option is to separate treatment strips by a couple of feet to clearly separate different replications.

“Keep the trials straight forward and simple, and down to one or two variables if you can, even though there is the temptation to add several things,” says Andreiuk. If the trial gets too big and across too many different landscapes, then growers will not likely end up with the results they expected. It is better to divide it and start a new trial. “During installation and through the life of the on-farm trial, make sure every party who is doing anything to that field knows what is happening and which way they should be doing the next operation.”

Andreiuk emphasizes the importance of planning up front and getting assistance if needed for randomization, treatments and other aspects of the trial such as statistics and reporting. During the trial, make observations and measurements and keep notes on all aspects. “Make sure to keep all information on a permanent record, don’t rely on a computer card that could get erased or lost before you have time to analyze the information.”

True on-farm trialing is field scale, scientifically valid research carried out and managed by the grower. It is replicated research that allows growers to explore and evaluate production practices. “Partner with others, carefully define your question, ensure the only variable is the treatment, everything else must be consistent across all replications throughout the season,” says Andreiuk. “There is a way of testing just about anything you want to, but you have to make sure it fits into your operation and infrastructure, and will answer the question you are asking.” -end-

The Bottom Line
It can get frustrating to look at data that companies provide on their products. Each claims that their product is the best and will provide decent returns.

The true test is how each product/treatment/variety will perform under your specific conditions. What might equate to a 10 percent increase in yield for one producer, may not generate anything for the producer in the next field – simply due to differences in production methods. Kenton Possberg, Humboldt, Saskatchewan.

So much more can be learned from field scale trials than from small plots. The crop has the opportunity to react as a field would not as an isolated plot. One great addition to field scale trials is getting aerial or satellite images. Because they’re big enough, the treatment effects really show up and add tons of credibility to your yield map. There are a lot of local aerial photo companies that will shoot a field for a very reasonable price, and it’s money well spent. Like David Kaun, we have been looking at fungicide and fertility interactions in an attempt to optimize oats production and wouldn’t have seen the incredible reduction in lodging with fungicide applications if we hadn’t been field scale. It’s real economics under real field conditions. John Waterer, Winnipeg, Manitoba.

Every year we do yield trials mainly of canola varieties. We co-operate with both Dekalb and Canterra on these trials. They consist of strip trials that are harvested and weighed, and the quality determined. The results from across the prairies are compiled and put together in a summary book that is distributed to producers. All the farming operations are using our equipment and we determine when the different operations are done. We have done comparisons using the yield monitor in the combine, and actual truck scale weights, to determine yield comparisons. We have found that the truck scale weights or weigh wagon weights are more accurate. We find these trials interesting to do as well as a good source of information for our seed customers. James and Peter Galloway, Fort Saskatchewan, Alberta. -end-