Only commercial seed should be used for testing of released varieties. Experimental and commercial varieties may be tested within the same experiment but results should be reported separately.
Site selection: Select sites with uniform soils and topography.
Soil fertility: Soil pH and fertility (P, K, and micronutrients) should meet or exceed recommended levels for the region.
Stands: Uniform stands should be established in rows (12 inches or less apart) or by broadcasting to provide a minimum of 20 plants/ft2 in the seeding year. If data from tests with substandard stands are used, stand density should be noted.
Plot size: Minimum of 3 to 5 feet wide by 12 to 25 feet long.
Borders: Areas of at least 3 feet around tests. Individual plots do not need border areas.
Culture
Irrigation: Not required, but if applied, irrigation should be uniform and irrigation strategy noted.
Pest control: Alfalfa should be scouted and insecticides and herbicides should be applied when needed and noted in reports. In particular, grass and other weeds should be controlled so that only alfalfa yield is reported.
Harvesting
Harvest scheduling: Schedule by stage of maturity, such as first flower, or by date. Harvest
schedules should be similar to those commonly used in the testing region. When reporting data,
provide information on harvest scheduling criteria and seasonal harvest number.
Stubble height: 2 to 3 inch minimum.
Harvest area: Minimum of 2 feet wide, 3 feet wide preferred, by 12 to 25 feet long. The entire plot can be harvested.
Harvest equipment: Sickle bar or flail mower.
Dry matter determination: Subsamples should be clipped from standing forage or hand grabbed from harvested materials. Minimum sample size should be 300 g wet weight. With uniform soil moisture and growing conditions, about 5% of the entries per trial should be sampled with a minimum of 5 samples collected. When wide dormancy and/or maturity stages exists, increase the number of subsamplings. Dry samples at 140o F in a forced air oven for about 48 hours. For drying, cloth, perforated paper bags, or trays can be used.
Weed invasion: Sometimes, persistent weeds may invade alfalfa trials despite best efforts of weed control. Reported yield data under weed invasion should not be adjusted. Weed invasion should be quantified and reported.
Data analysis and statistics:
Experimental designs:
Experimental designs vary with experiences of the investigator. The following designs are
recommended:
For evaluation of entry performance, trials should be conducted at a minimum of two environments.
Rejection tests:
Currently, there are no tests recommended to determine whether data from an individual year or
location are too variable to accept. A subcommittee (Mark Sulc, Dan Undersander, Mike Peterson,
Thad Busbice, and John Reich) is studying this issue. Criteria for rejection include:
Publication of data.
Data from seeding and production years should be reported. Investigators should provide the
following statistics for comparison of data and to measure the variability within a trial: LSD (Least
Significant Difference method at 0.05 level), CV (Coefficient of Variation), MCV (100
LSD/mean), LSR [100(LSD)/Range].
Data collected from experimental alfalfa populations will not be distributed to the public. These data will be distributed through NOT FOR PUBLICATION documents like the annual CAIC Variety Tests.
Data on experimental populations that are named for release to farmers may be published, but only if the newly named variety is entered again in the same University Variety Trials using commercial seed.
Data collected from experimental seed should be deleted from published data after 2 years of data are collected from commercial seed.
Check Cultivars
Currently, there are no universal check cultivars. A committee has initiated a study to develop a
strategy for identification of check cultivars. (Dan Undersander, Neal Martin, Charles Brummer,
and Jon Reich)