Performance of commercial corn hybrids in Illinois, 1978 / 1166

Planting methods.All trials were planted by machine.All test fields except those at DeKalb, Macomb, Brownstown, Elwood, and Urbana were part of larger cornfields and thus were bordered by other corn.Each hybrid plot was overplanted 30 percent and later thinned to desired stands.Each plot was four rows wide and 25 feet long.The center two rows of each plot were harvested to determine yields.Fertilization.All test fields were at a high level of fertility.Additional fertilizer was plowed down or sidedressed as needed to assure top yields.Method of harvest.All plots were harvested with a self-propelled combine.Shelled corn from each plot was collected, weighed, and tested for moisture content.No allowance was made for corn that might have been lost in harvest.Measuring Performance Grain moisture.Occasionally, hybrids too late in maturity for a given area are entered in these tests.Such hybrids are often high in yield, but their moisture content may make them poor choices for farm use un- less proper drying or storage facilities are available.Yield of grain.Shelled-corn weight and moisture percentage were measured for each plot of a hybrid and converted to bushels per acre of No. 2 shelled corn (15.5-percent moisture).An electronic moisture tester was used for all moisture readings.Erect plants.The number of erect plants in each plot of a hybrid was counted at harvest time.Any plant leaning at an angle of more than 45 degrees or broken below the ear was considered lodged.Plants broken above the ear were considered erect.Population.In late June, plants in all plots on all fields were counted and population computed.Plots with over 100 percent of the desired population were thinned at that time.Stand differences may be caused by failure to germinate or by damage from diseases, insects, cultivation, or animal pests.Comparing hybrids.It is impossible to measure performance exactly in any test of plant material.Harvesting efficiency may vary, soils may not be uniform, and many other conditions can produce variability.Results of repeated tests, like those reported here, are more reliable than those of a single-year or a single- strip test.In general, a yield difference of a few bushels per acre is not significant in these tests.When one hybrid consistently outyields another at several test locations and over several years of testing, the chances are good that this difference is real and 'should be a consideration in choosing a hybrid.Rut yield alone is not enough.Consider also the grain moisture content, per- centage of erect plants, percentage of stand, or the number of plants per acre in comparing yields.A number of statistical tests are available for com- paring hybrids.One of these tests, the least significant difference (L.S.D.), when used in the manner sug- gested by Carmer and Swanson, 1 is quite simple to 1 Carmer, S. G. and M. R. Swanson."An Evaluation of Ten Pairwise Multiple Comparison Procedures by Monte Carlo Methods."