This is Part 3 of a multi-part series on On-Farm Variety Trials. Part 2: “Planning It Out” can be found here.
Last month, I explored planning your on-farm variety trial. I began with identifying goals. The goals you outlined will guide the next several decisions in executing your variety trial, the design of your experiment.
In order to determine if certain varieties perform better than others with respect to your goals, the trial must account for or “control” for differences in the field, soil, or management conditions. Anyone who has been farming or gardening for at least a season has likely seen how different spots in the field affect plant growth. That wet spot, or shady row will affect performance of your varieties. This is called the “field effect,” or the influence that variable field conditions have on the performance of plants in the field. More examples of field effects include slope, drainage, soil type, pH, wind direction, cold pockets, and irrigation placement. In this blog entry, we will explore how to design the trial so that it emphasizes differences in varieties rather than differences in growing conditions.
There are two main designs of field trials, a “screening trial” and a “replicated trial.” They are characterized by the way they handle field variation through replication of your chosen varieties in multiple plots across the field. Screening trials are unreplicated, which means each test variety is grown in only one plot. Screening trials are easier to design and manage, but are more limited in the questions they can answer because they allow for more influence by field effects. In replicated trials, each variety is planted in multiple places at the same time, or “replicated.” This design requires more attention and effort, but provides more accurate results and greater assurance that the results are truly the result of variation between varieties and not the result of field effects. Choosing the experimental design for your trial is guided by the information you are trying learn, the goals you set in part 2 of this blog series.
Screening trials, because they are not replicated, are more limited in the information they provide about the tested varieties. To conduct a screening trial, select the 2-10 varieties you would like to compare. Then choose a standard or “check” variety to compare the test varieties to. This check variety will be planted intermittently between test varieties and serves as an indicator of the field uniformity of the experiment site. If you find major differences in outcomes between the plantings of the check variety, that is an indication that the differences you see in the other varieties may be the result of field effects and not the characteristics of the varieties. If you stands of the check variety are performing similarly, you can be more confident that the differences between varieties are actually the result of genetic differences, and not field variation. The check variety also gives a point of comparison when analyzing the performance of the test varieties in the characteristics you are trying to measure.
Because a screening trial does not control for field effects, it is especially important to pay attention to field conditions and management consistency. This will help ensure that the results of the trial are reliable. Scout the potential areas where you may hold your trial. Look for a place that is as uniform as possible in water infiltration, irrigation access, slope, and sunlight access. Try to avoid problem places such as areas with compaction, changing soil type, wet spots, etc. Consider locating your trial in the middle of your production area for this crop to ensure that management of the trial is consistent with your typical management practices.
Once your trial location is chosen, create a planting map of your trial. First, determine the gradient of field variation present at your trial site. For example, this may be slope, or it may be soil type. Orient your plots so that they lay with the gradient of variation (see the figure below). Next, distribute your check variety evenly among your test varieties. And finally, plug in your test varieties in a random way. It is essential that the test variety locations are assigned randomly. This can be done using an online random number generator. Assign a number to each variety (e.g. 1-7 if there are seven varieties) and the random number generator will randomly pick the order for you. It is also important to plant a border around your trial area to reduce edge effects.
Once your varieties and location are chose, and your map is created, you are ready to determine your planting density and method, and to plot your trial at your location. We will discuss plot size, plants per plot, and edge effects in a later installment in this blog series.
Though screening trials are more limited in the information they can provide because they do not control for field effects, they are useful for evaluating more limited questions like:
whether a variety merits the greater effort of a replicated trial;
checking for trueness-of-type or seed quality concerns, overall plant composition, and uniformity;
Identifying potential strengths and weaknesses of a variety that can be tested in a replicated trial, and;
Trials with a core question that is less likely to be influenced by field effects, like how well a variety sells at market.
Next month, we will explore designing the more comprehensive “replicated trial”. Until then, happy farming!