ALFALFA VARIETY TESTING

Many questions arise over time about variety testing programs, such as the following:   

• Do they accurately predict yield?   
• How are the sites chosen for testing?   
• Who decides which varieties are tested and how do they decide?   
• What do the results tell?   
• How can producers use results from the trials?   
• Why don't we have more trials?   
• How long do we keep the trials?   

These are good questions, and I will try to answer these with regards to Alfalfa. Variety testing programs generally have about the same approach but many mechanics and philosophies differ from crop to crop.   

The overall purpose of alfalfa variety tests in Oklahoma is to compare the performance of varieties grown under identical conditions. We can identify those varieties from both public and private breeding programs that have the greatest potential for use in this area. Absolute yields are not nearly as important to us as relative yields, comparing one variety to another. Consequently our yields are not supposed to predict what a farmer in the area would produce. Sometimes commercial production is higher, and sometimes lower.   

We recognize that management is as important as the variety, or more so, in determining yield. Therefore we do everything exactly the same for each variety in a test -- management is exactly the same. All varieties in a test are sown in the same seed bed within an hour. The fertilizer is the same, and the timing of harvests is the same for all varieties in a test. On a field-scale basis, this is impossible to do. There will always be at least a little difference in management of fields, no matter how hard a person tries.   

We choose testing sites on research stations. Normally the stations represent important alfalfa production areas. This is primarily for financial reasons. The costs of conducting tests on farmers' fields is much higher than on research stations.   

Picking varieties to include in a trial is not the decision of a single person. All the major alfalfa seed production companies, especially those with breeding activities, are contacted prior to establishing new trials. The company representatives choose which varieties to enter into which test. They normally pick the tests where they believe they will get the most definitive results. The companies pay a fee (for each variety at each site) to help cover the costs of the tests.   

Varieties developed by public institutions are included in tests without a fee. Once a variety (public or private) has proven to have consistently high yield and persistence, we ask the developer to provide seed of the variety for tests without a fee. These varieties become checks to compare others against.   

The results from variety trials provide information about a varieties' general adaptation to the area and how it compares with competing varieties. No positive conclusions along these lines are made until several years of data are accumulated from several sites. A very poor variety will usually show up to be poor in the first test, but it takes several observations to say a particular variety is good or well adapted to a certain part of the state.   

Breeders make decisions based on these results as to whether a variety should be released. Marketing people use our data, along with data from other states, to determine where a particular variety should be marketed. Reputable companies do not market a variety outside its proven area of adaptation. But, watch out for the guy with a warehouse full of seed! He may not have a reputation to preserve.   

Producers can use the results from our variety tests to decide which varieties they should try in their production. It is normally advisable to sow two or three good varieties. One may be better for a particular farmer than another, even though they perform similarly in on-station tests.   

Some people think we should have alfalfa variety tests in every county that produces alfalfa. This is just too expensive and we don't have the man power to conduct such a program. In reality, varieties that perform consistently well in our tests will probably do well within a 100-mile radius -- assuming management is good.   

We record yield from each harvest for three years. The trials are usually maintained for several additional years and persistence observations are made. Experience indicates that the yields during the first three years show real differences, if they exist. Collecting yield data after that just confirms the earlier observations. Extending the yield collection to 5 or 6 years becomes prohibitive from a cost stand point.      
 
 Results of Recent Alfalfa Variety Tests