'09 Public Hybrid Trial Data Sites
Friday, January 1, 2010
filed under: Hybrid Selection/Planting
Hybrid evaluations from several 2009 public trials in the Dakotas and High Plains are now available on the National Sunflower Association’s website and, as well, on related university websites. These trials were conducted through cooperation among university and USDA sunflower specialists and the private seed companies whose hybrids were being tested. (Some 2009 trial data may not yet be posted as of this writing, but will be shortly.)
To find trial data on the NSA website, go to www.sunflowernsa.com. Click on the “Growers” link, then “Yield Trials and Hybrid Disease Ratings.”
The following identify university sources of sunflower hybrid performance information available online for several key production states:
North Dakota State University www.ag.ndsu.edu/varietytrials/sunflower
South Dakota State University plantsci.sdstate.edu/varietytrials/
Colorado State University www.extsoilcrop.colostate.edu/CropVar
Kansas State University www.agronomy.ksu.edu/extension/
University of Nebraska www.varietytest.unl.edu/
For a list of seed suppliers and seed company websites, go to "Buyers" then “Seed Suppliers/Buyers,” and then “Hybrid Seed Suppliers.”
Seed companies previewed new 2010 hybrids in The Sunflower’s December 2009 issue. That article can be seen online at www.sunflowernsa.com. Go to “Sunflower Magazine,” then “View Archives” and finally to “Hybrid Selection/Planting.”
Making Sense of Hybrid Trial Statistics
• Expected mean in plot trial information refers to the average performance number for a particular trait of all hybrids evaluated in the trial.
• Coefficient of variability (CV %), often listed at the bottom of a hybrid data table, is a relative measure of the amount of variation or consistency recorded for a research trial (or a particular trait within a trial), expressed as a percentage of the mean. Generally, trials with low CV rates are more reliable for making hybrid choices than trials with higher CV rates. Trials with CV below 15-20% are generally considered to be reliable for comparing yield.
• To accurately determine whether one hybrid is better than another for a given trait, use the least significant difference value (LSD 5%) provided at the bottom of the table. This is a statistical way to indicate whether a certain trait (e.g., yield) differs when comparing two hybrids. If two hybrids differ by more than the indicated LSD 5% value, they would most likely differ again when grown under similar conditions. If two hybrids differ by less than the LSD, there’s no statistical difference.
For example, if a performance trial table indicates one hybrid yielded 2,600 lbs/acre, compared to another hybrid in the same plot that yielded 2,310 lbs/acre — and the LSD for this plot trial data is 407 lbs/acre — there is no statistical difference in yield between the two varieties.
In another example, if the oil content percentage for one hybrid is 44 compared to 41 for another, and the LSD is 2.3, the first hybrid can be expected to have higher oil content than the second hybrid, under similar growing conditions.
Give more weight to information from trials or fields close to your own growing area. It’s best to compare relative performance over multiple years and locations. Consult an agronomist or your local seed dealer for more-specific information.