The widespread adoption of precision agriculture has allowed growers to generate and collect more data than ever before. The big question now is, what to do with all this data? Why not put those tools to work in your own on-farm research trials to determine if a new product or practice will work in your environment.
The first step in setting up an on-farm research project is to determine the practice or product that will be analyzed. Some of the more common research topics are:
Once you have identified a topic that interests you, confirm that you have the right equipment to set up and measure your trial and that you have the proper field site for it.
Placement and layout of your research trial is vitally important. This is also one of the reasons that doing your own research can be beneficial – because you know exactly how and where the trial was carried out. After the proper site is identified, we will use blocks, replication, and randomization to remove field variability from the equation and increase our confidence in the results.
The block is the treatment strip(s) compared against the “control” or untreated check strip. Blocks should be placed and sized so there is little field variability within the block. If you want to research starter fertilizer, for example, you could set up a block like this:
Next we want to replicate, or repeat, our blocks so that we will have multiple looks at the same comparison. This increases our confidence and lessens the risk of the trial failing due to the loss of a block. Usually, we like to see 4-6 replications (reps) across the field.
To further reduce biases and field variability impacts, we want to randomize the block. Randomization is the process of determining the order of the treatments within the block by chance, such as flipping a coin. Our final, randomized, four-replication starter fertilizer trial will now look something like this:
Example of a randomized, replicated block
The final steps in the research process involve collecting relevant data, analyzing the results using statistical methods and drawing conclusions. The main data we are after are the final yield results, of course, but it is useful to also collect information such as planting and harvest dates, soil types, rainfall and other weather data, and any other applicable information.
Once we have the final data, we can use statistical methods to calculate our Least Significant Differences and confidence intervals, which can determine if the differences in the data are due to the treatments or due to chance. We can then start to draw conclusions to our research and decide whether new products and practices should be adopted more broadly in the operation.
One specific trial we encourage our dealers to utilize is our P3 trials. These are a fixed block of products within a maturity range. These are not randomized within the block, but the same set ends up being replicated across a broad geography many times. Since we will be collecting soil samples at all of our P3 trials, we can start to analyze how our products perform across different soil types and management practices. It is important that all of the products in the P3 set are planted, as this will maintain the integrity of the blocks across our entire brand footprint.
The University of Nebraska has an excellent resource for more information on on-farm research – the Nebraska On-Farm Research Network. There is more in-depth information on setting up research trials, as well as the results of on-farm research trials that have been carried out in Nebraska.
As you can see, on-farm research trials are important for the Producers Hybrids brand and other growers, but more importantly, they are important for YOU. By putting in the time and effort, you can determine what works best on your farm, and that will give you a real return on investment in the future.