When looking at data from independent sources like universities and FIRST, how do you compare hybrids?

One way is to look at the LSD, or least signficant difference.

This is key, especially with the amount of hybrids tested in some of these locations.

Let’s take FIRST for example. In a given location, there are 60-90 hybrids tested in three replications. Six of these locations make up a regional summary. Then there are even zone summaries released in the FIRST testing book, made up of 2-4 regional locations. That’s a lot of data!

Back in the local FIRST plot, say 81 hybrids are tested in three replications (Iowa Falls, IA full-season test). These replications are averaged to get your local result. At the bottom of that report, the LSD (least significant difference – 18.6 bu/A) is listed. For those of you who slept through statistics class (like I did part of the time), this is a measure of the variability of a test.

Say the field was planted all to one hybrid. Due to environmental, pest, soil, nutrient or other factors, the yield of that hybrid changes across the field. The LSD is an indicator of how much variability is due to field variation and how much is due to the hybrid or variety. The lower the LSD, the more precise are the measurements of genetics factors and less on the changes in the environment.

So, the LSD in the plot linked above is 18.6 bushels per acre. The #1 hybrid yielded 232.5 bu/A. That means any hybrid yielding 213.9 bushels or better is statistically the same as the #1 hybrid. In this case, that includes the entire Top-30 hybrids reported. This is a rather high LSD. It also does not only apply to a top hybrid, thus a hybrid that yields 200 bu/A is statistically the same performance as one at 218 or 182 bu/A.

Most independent tests that are replicated in nature show this variability. The replication is the key to determining the LSD for a plot.

One other thing to note, especially in the FIRST trials, is the amount of different genetics getting tested. Many seed companies are pulling hybrid genetics from the same sources. This leads to these trials containing many similar genetics under different names. Look at performance, traits and maturities of these hybrids and note any differences. We’ve seen in some cases a hybrid may finish 20th but there are only two or three genetic families ahead of the 20th place hybrid (and in many cases the LSD was more than the difference). Do your research.

Great Lakes has more than 70% of our production in hybrids that are not available outside of AgReliant Genetics. You can be sure that our hybrids are unique to others in the industry, thus removing a level of risk.
For more on our production, research or plot data, follow the highlighted links.