Crop performance is measured by yield, a combined result of the interaction of product genetics with the growing conditions under which it was tested. Many different variables contribute to the performance, or yield, of a product. A high-yielding product with the best average across multiple locations will not have the highest yield at every location and even the best product does not ‘win’ all the time.
It is always exciting to watch the first yield data roll in at harvest, but as the season goes on there are sometimes changes to the products we see winning or how they rank in performance. Why is that? Often, the initial data reported does not offer the full picture of actual average performance of a product. As more data is accumulated, a product’s rank will likely become more consistent and provide a better estimate of its true yield potential.
In addition to having a large quantity of data, it is also important to have data that is representative of the entire geography for which the product is marketed. If data presented is from only three-fourths of the geography in which the product is available, the remainder may come from quite a different geography which could result in an inconsistent outcome. Differences in seasonal stress and geography can influence harvest timing and when yields are reported.
Most genetic traits that affect yield are quantitative traits, controlled by multiple genes that each contribute a portion to the overall characteristic. The interaction of environment and genetics can result in variations in actual yield because the environmental conditions can have an effect on each of the genes independently and in different ways. When you observe the actual yields of one product planted across multiple plots, the data will show that most observations are very close to the mean, or average. However, some observations, usually close to 5% or less, are very different from the mean. These values represent the natural variation that can be seen in any population. Plant-to-plant variability can be expected regardless of yield level, so it makes sense that product averages will also be variable.
Figure 1. Yield distribution and individual yields of eleven individual plots among two seed products.
When comparing two products, the yield observations for each will fall into a bell-shaped curve around their means as shown in Figure 1. While the means of Product A and Product B are different, there is some overlap between the two products. You can see that while Product A may have a higher average yield, Product B may have some yield observations that are higher than Product A.
Probability is another factor to consider when a winner is determined in a comparison. If two products have the same yield potential, the odds of one winning over another are similar to the 50:50 odds of heads or tails when you toss a coin. If you toss a coin 10 times, it will not necessarily result in five heads and five tails. The same is true of a yield trial. If two products are equal in yield, the odds of either one winning are 50%. If one product demonstrates superiority, it will win more often, but a win is still not guaranteed in every comparison. Even industry-leading products do not “win” every time.