Conducting On-farm Research
Average reading time: 4 minutes
Small-plot research has informed farmers worldwide about various applications, and approaches, from increasing yield to decreasing the soil seed bank. Whether it be the Morrow Plots at the University of Illinois, or Becks Hybrid’s Practical Farm Research, they all have the same end goal of providing information. This type of information can then be processed by the recipient. Often, it also raises many questions. How will this hybrid, technique, or application perform on my farm? My soils? With my hybrids? The answer is simple but must be executed well.
Conducting your own on-farm research first starts with the questions above and can be done, most likely, with your current equipment. A research hypothesis is a statement that the farmer wants to test. For example: “Using the Horsch Maestro; I will compare the seed lock wheel, seed firmer and the control of no firming device”. Notice how this will be utilizing one piece of equipment limiting the variables to only the seed furrow devices.
Variability is the easiest way to contaminate data. To understand it there are two different types. The first being spatial variables. This is the variation in environmental characteristics over distance and depth. The second is temporal variables. These include any variability over time. So, how can any worthwhile data come from a farm field, full of variables?
Replication, randomization and keeping the collection on site are the three main components of quality data. Many farmers choose to split a field 50/50 for a data set. While this is the simplest way of conducting a study, it is not the best data. Imagine an 80-acre field. The left side is our control, and the right 40-acres are the applied product. There can be many differences from one side of the field to the other. To eliminate this the passes must be replicated and at random. With only two data sets replication is priority but if there were a few hundred data sets, like in hybrid categorization, randomization becomes very important.
Sometimes how data is analyzed can be just as important as the way it was arranged and collected. Once again, replication will show the true outliers. If there is one replication, or rep, that is drastically different than the others it demands further investigation. For plot research, I keep detailed notes of every pass. This allows me to figure out why there is an outlier in the data set. For example, rep 3 has a low area that is prone to staying wet throughout the growing season. This would be a simple explanation for why this rep yielded much lower than the rest and should not be used in the overall data. Or “Rep 2 had rows 2 and 5 ran over by post applicator for 300 feet”. You get the idea, but without detailed notes for the entire growing season a greater chance of tainted data can occur.
In conclusion, on-farm research is the best practice to obtain reliable data that fits your operation in your conditions. It is very necessary to still rely on small plot data from others. These off-site sources will give the farmer a starting point. They figure out what rates, and timing of applications, to give an idea of what will yield best. Once the established application, and rate, are established repetitions are placed and finally the most important component to research, uniformity. The same rate, speed, depth etc must be the same. Every pass must be treated identical.